Nanofibers: Fabrication, Performance, and Applications

NANOTECHNOLOGY SCIENCE AND TECHNOLOGY SERIES Safe Nanotechnology Arthur J. Cornwelle 2009. ISBN: 978-1-60692-662-8 National Nanotechnology Initiative: Assessment and Recommendations Jerrod W. Kleike (Editor) 2009. ISBN 978-1-60692-727-4 Nanotechnology Research Collection 2009/2010. DVD edition James N. Ling (Editor) 2009. ISBN 978-1-60741-293-9 Nanotechnology Research Collection 2009/2010. PDF edition James N. Ling (Editor) 2009. ISBN 978-1-60741-292-2 Safe Nanotechnology in the Workplace Nathan I. Bialor (Editor) 2009. ISBN 978-1-60692-679-6 Strategic Plan for NIOSH Nanotechnology Research and Guidance Martin W. Lang (Author) 2009. ISBN: 978-1-60692-678-9 Nanotechnology in the USA: Developments, Policies and Issues Carl H. Jennings (Editor) 2009. ISBN: 978-1-60692-800-4 New Nanotechnology Developments Armando Barrañón (Editor) 2009. ISBN: 978-1-60741-028-7

Electrospun Nanofibers and Nanotubes Research Advances A. K. Haghi (Editor) 2009. ISBN: 978-1-60741-220-5

Nanostructured Materials for Electrochemical Biosensors Umasankar Yogeswaran, S. Ashok Kuma and Shen-Ming Chen 2009. ISBN: 978-1-60741-706-4 Magnetic Properties and Applications of Ferromagnetic Microwires with Amorpheous and Nanocrystalline Structure Arcady Zhukov and Valentina Zhukova 2009. ISBN 978-1-60741-770-5 Electrospun Nanofibers Research: Recent Developments A.K. Haghi (Editor) 2009. ISBN 978-1-60741-834-4 Nanotechnology: Environmental Health and Safety Aspects Phillip S. Terrazas (Editor) 2009. ISBN: 978-1-60692-808-0 Nanofibers: Fabrication, Performance, and Applications W. N. Chang (Editor) 2009. ISBN: 978-1-60741-947-1

Nanotechnology Science and Technology Series

NANOFIBERS: FABRICATION, PERFORMANCE, AND APPLICATIONS

No part of this digital document may be reproduced, stored in a retrieval system or transmitted in any form or by any means. The publisher has taken reasonable care in the preparation of this digital document, but makes no expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No liability is assumed for incidental or consequential damages in connection with or arising out of information contained herein. This digital document is sold with the clear understanding that the publisher is not engaged in rendering legal, medical or any other professional services.

Nanotechnology Science and Technology Series

NANOFIBERS: FABRICATION, PERFORMANCE, AND APPLICATIONS

W. N. CHANG EDITOR

Nova Science Publishers, Inc. New York

Copyright © 2009 by Nova Science Publishers, Inc.

All rights reserved. No part of this book may be reproduced, stored in a retrieval system or transmitted in any form or by any means: electronic, electrostatic, magnetic, tape, mechanical photocopying, recording or otherwise without the written permission of the Publisher. For permission to use material from this book please contact us: Telephone 631-231-7269; Fax 631-231-8175 Web Site: http://www.novapublishers.com NOTICE TO THE READER The Publisher has taken reasonable care in the preparation of this book, but makes no expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No liability is assumed for incidental or consequential damages in connection with or arising out of information contained in this book. The Publisher shall not be liable for any special, consequential, or exemplary damages resulting, in whole or in part, from the readers’ use of, or reliance upon, this material. Any parts of this book based on government reports are so indicated and copyright is claimed for those parts to the extent applicable to compilations of such works. Independent verification should be sought for any data, advice or recommendations contained in this book. In addition, no responsibility is assumed by the publisher for any injury and/or damage to persons or property arising from any methods, products, instructions, ideas or otherwise contained in this publication. This publication is designed to provide accurate and authoritative information with regard to the subject matter covered herein. It is sold with the clear understanding that the Publisher is not engaged in rendering legal or any other professional services. If legal or any other expert assistance is required, the services of a competent person should be sought. FROM A DECLARATION OF PARTICIPANTS JOINTLY ADOPTED BY A COMMITTEE OF THE AMERICAN BAR ASSOCIATION AND A COMMITTEE OF PUBLISHERS. LIBRARY OF CONGRESS CATALOGING-IN-PUBLICATION DATA Nanofibers : fabrication, performance, and applications / W.N. Chang, editor. p. cm. Includes index. ISBN 978-1-61668-288-0 (E-Book) 1. Nanofibers. I. Chang, W. N. TA418.9.F5N36 2009 620'.5--dc22 2009021232

Published by Nova Science Publishers, Inc. Ô New York

CONTENTS Preface Chapter 1

Chapter 2

Chapter 3

Chapter 4

Chapter 5

vii Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers Sk. F. Ahmed and K. K. Chattopadhyay

1

Permeability Studies of Electrospun Chitin and Chitosan Nanofibrous Membranes Jessica D. Schiffman and Caroline L. Schauer

61

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning: Preparation and Biomedical Application D. Paneva, М. Ignatova, N. Manolova and I. Rashkov

73

A Novel Approach for Analysis of Processing Parameters in Electrospinning of Nanofibers M. Ziabari, V. Mottaghitalab and A. K. Haghi

153

Carbon Nano-Fibers and their Applications: Derived from Electrospinning and Vapor Grown Processes S. K. Nataraj, B. H. Kim and K. S. Yang

183

Chapter 6

Carbon Nanofibers as Sensors Sharlene A. Lewis and Charles M. Lukehart

225

Chapter 7

Processing-Structure Relationships of Electrospun Nanofibers Xiangwu Zhang

239

Chapter 8

Glycosylated Nanofibers for Protein Adsorption and Recognition Ai-Fu Che, Ling-Shu Wan and Zhi-Kang Xu

271

Chapter 9

Porphyrinated Polymer Nanofibers by Electrospinning Yuan-Yuan Lv, Jian Wu, Zhen-Mei Liu and Zhi-Kang Xu

303

Chapter 10

A Nanofibrillar Prosthetic Modified with Fibroblast Growth Factor2 for Spinal Cord Repair Sally Meiners, Suzan L. Harris, Roberto Delgado-Rivera, Ijaz Ahmed, Ashwin N. Babu, Ripal P. Patel and David P. Crockett

327

vi Chapter 11

Chapter 12

Chapter 13

Chapter 14

Index

Contents Fabrication, Performance, and Biomedical Application of Collagen-, Gelatin- or Keratin-Containing PHBV Nanofibers Inn-Kyu Kang, Zhi-Cai Xing, Jiang Yuan, Oh Hyeong Kwon, Jung Chul Kim and Yoshihiro Ito

345

Fabrication and Characterization of Polypropylene Fiber Reinforced by Carbon Nanofiber Yuanxin Zhou and Shaik Jeelani

375

Thermal Analysis of Carbon Nanotubes Incorporated Polyurethanes Nanocomposites Shahrul Azam Abdullah and Lars Frormann

395

Carbon Nanotubes Incorporated Polyurethanes Nanocomposites for Thermal and Electrical Conductive Applications Shahrul Azam Abdullah and Lars Frormann

411 425

PREFACE Nanofibers are defined as fibers with diameters on the order of 100 nanometers. They can be produced by interfacial polymerization and electrospinning. Contrastly, carbon nanofibers are graphitized fibers produced by catalystic synthesis. Nanofibers are included in garments, insulation and in energy storage. They are also used in medical applications, which include drug and gene delivery, artificial blood vessels, artificial organs and medical facemasks. This book presents new research in this dynamic field. Chapter 1 - From the authors experimental observation, it is proposed that the nanothermometer can be constructed using MWCNT more easily. As the emission current vary linearly with temperature for a particular applied electric field, so temperature can be directly measured. The sensitivity of the nanothermometer can be adjusted by choosing the area of the MWCNT film or appropriate applied electric field. The study cited shows that the temperature dependent field emission property of CNFs and MWCNTs has potential for development of direct thermal-to-electrical power conversion applications. Continued improvements in the PECVD of CNFs/CNTs and related nanostructures are indeed required to explore the potential utility of these structures in advanced applications and future largescale integration. Chapter 2 - Electrospinning has been utilized to fabricate fibrous membranes composed of polymer nanofibers, which have large surface area-to-volume ratios and small pores. Electrospun nanofibrous membranes have potential uses in a variety of industries such as energy, environment, medicine, packaging, and automotive, with specific applications including air filtration, protective clothing, fuel cells, and nanocomposites. Nanofibrous membranes composed of biopolymers have potential uses that harness their inherent biocompatibility. Chitin, the second most abundant, naturally occurring polysaccharide after cellulose, is found in shells of crabs and shrimp. Chitosan, the acid soluble form of chitin, is a non-toxic, biodegradable, biopolymer consisting primarily of β(1→4) linked 2-amino-2deoxy-β-D-glucopyranose units, and is currently used in tissue engineering, antifouling coatings, separation membranes, stent coatings, enzyme immobilization matrices, and the removal of heavy metals from ground and wastewater. Chitosan is a commercially interesting compound because of its high nitrogen content (6.89%), making it a useful chelating agent for metal ions. Before these chitin or chitosan nanofibrous membranes can be used in the myriad of industries their physical properties, such as permeability, must be known. This chapter focuses on the fabrication and flow cell testing of chitin and chitosan nanofibrous membranes.

viii

W. N. Chang

Additionally, it explores the potential applications of biopolymer and synthetic polymer electrospun membranes. Chapter 3 - At present increased attention is paid on fibrous materials from the natural polymer chitosan because of its numerous beneficial properties (biocompatibility, biodegradability, inherent antibacterial and haemostatic activity). The presence of both hydroxyl and amino groups enables the tailored modification of chitosan into derivatives having targeted properties. The materials containing chitosan or its derivatives are considered as very promising candidates for versatile applications in medicine, pharmacy, food industry, and agriculture. Nowadays the preparation of nanosized fibrous materials is of special interest because of their unique properties, in particular their high surface area-to-volume and aspect ratios. Electrospinning is a cutting edge technique for fabrication of continuous polymer micro- and nanofibers. The basic principles and the effect of the process parameters on the morphology of the electrospun fibers and fibrous materials are briefly discussed in the present Chapter. The first successful attempt to prepare chitosan-containing electrospun materials dates from 2004. This has been achieved by the addition of a non-ionogenic, water-soluble polymer into the spinning solution. The application of this approach for preparation of chitosan-containing fibers is thoroughly discussed in the Chapter. The preparation of neat chitosan nanofibers by electrospinning is outlined as well. The application of suitable chitosan derivatives soluble in water or low toxic organic solvents enables the design of novel non-woven textiles in absence/presence of a non-ionogenic polymer. The preparation of such non-toxic, environmentally friendly materials is detailed. The applied two-step procedures (heat or UV treatment, use of appropriate crosslinking agents) for imparting water-insolubility to the obtained micro- and nanofibrous materials are described. The main approaches that have been used for preparation of electrospun materials combining the beneficial properties of chitosan and aliphatic polyesters based on poly(L-lactide): simultaneous electrospinning or electrospinning of the polyester, followed by coating of the non-woven textile with a thin chitosan layer, are summarized. Moreover, the recently developed routes for preparation of chitosan-containing micro- and nanofibers, such as reactive electrospinning, combination of electrospinning and polyelectrolyte complex formation as well as yarns formation, are discussed. The advantages of the one-step imparting of water-insolubility of chitosan fibers by reactive electrospinning and polyelectrolyte complex formation as compared to the twostep procedures are emphasized. Last but not least the potential biomedical application of the obtained micro- and nanofibers are outlined. Chapter 4 - The precise control of fiber diameter during electrospinning is very crucial for many applications. A systematic and quantitative study on the effects of processing variables enables us to control the properties of electrospun nanofibers. In this contribution, response surface methodology (RSM) was employed to quantitatively investigate the simultaneous effects of four of the most important parameters, namely solution concentration (C), spinning distance (d), applied voltage (V) and volume flow rate (Q) on mean fiber diameter (MFD) as well as standard deviation of fiber diameter (StdFD) in electrospinning of polyvinyl alcohol (PVA) nanofibers. Chapter 5 – Electrospun (ES) and vapor grown carbon nanofibers (VGCFs) are attractive building blocks for functional nanoscale devices. They are promising candidates for various applications, including filtration, protective clothing, polymer batteries and sensors. The continuous progress of nanotechnology in material science has led to the development of nanostructure materials with unique chemical, physical, and thermal properties. Nanofibers

Preface

ix

possesses significant characteristic that exhibit enormous availability of surface area per unit mass. Furthermore, their high surface-to-volume ratio renders them attractive as catalyst supports, energy storage devices, as well as in drug delivery and tissue engineering. Since the discovery of carbon nanotubes in 1991 and based on the results obtained from the characterization of these nanostructures, many other carbon-based nanomaterials have been developed. Of these, carbon nanofibers (CNFs), fullerenes, carbon nanohorns, and nanoporous structures are the subject of extensive experimental and theoretical studies for specific applications. Carbon is a truly remarkable element existing as four allotropes, viz. diamond, graphite, carbynes and fullerenes, each having significant scientific and technological importance. Its most abundant allotrope, graphite, can take many forms with respect to microstructure, amorphous to highly crystalline structure, highly dense with density 2.2 g/cm3 to highly porous with density 0.5 g/cm3 and different shapes. These types of graphites are called synthetic carbons and in technical terms, engineered carbons. Carbon nanofibers are unique in the fact that their whole surface area can be activated. Since carbon nanofibers have a much larger functionalized surface area compared to that of nanotubes, the surface-active groups-to-volume ratio of these materials is much larger than that of the glassy-like surface of the carbon nanotubes. This characteristic, combined with the fact that the number and type of functional groups on the outer surface of the carbon fibers can be well controlled, is expected to allow for the selective immobilization and stabilization of functional biomolecules such as proteins, enzymes, and DNA. Also, the high conductivity of carbon nanofibers seems to be ideal for the electrochemical signal transduction. The oxygen-containing activated sites are ideal for the immobilization and stabilization of biomaterials is an important feature. Industrial applications of these new materials include: polymer and elastomer fillers, commercial hydrogen storage systems, radiowave-absorbing composites, lithium battery electrodes, construction composites, oil additives, gas-distribution layers for fuel cells, absorbents and filters as well as in capacitive deionization (CDI) processes for water treatment. Many applications are being developed for field emission display, electrodes of secondary battery and reinforcement of materials. Among the many future possibilities includes; soft protective vests stronger than Kevlar, bandages that can contract to put pressure on, artificial muscles powered by electricity those expected much lighter than current hydraulics, would make it easier to incorporate electronic sensors and actuators into clothing. All of these possible applications derive from the remarkable properties of carbon nanofiber, the ability to conduct both heat and electricity along with the extreme toughness of the fiber. This chapter presents the various features of carbon nanofibers with elaborated properties description in connection with their different application, produced from electrospinning and vapor grown techniques. Carbon fibers are fibrous carbon materials with carbon content more than 90%. They are transformed from organic matter by 1000-1500oC heat treatment, which are substance with imperfect graphite crystalline structure arranged along the fiber axis. Chapter 6 - Carbon nanofibers (CNFs) are platelet or conical (herringbone) carbon nanostructures consisting of nested cup-shaped or platelet graphene sheets stacked along the long fiber axis. CNFs typically have diameters on the nanometer scale and lengths on the micrometer scale and possess attractive properties, such as large surface area, high electrical conductivity, and good mechanical strength and thermal stability. As-prepared CNFs have surfaces along the long axis that terminate in C(sp2)-H edge sites that are suitable for

x

W. N. Chang

chemical functionalization. The surface charge, wettability, dispersibility, and chemical reactivity of CNFs can be altered through chemical and physical modifications of these CNF surface sites. CNFs or surface-functionalized CNFs have been used as sensor media either as pristine nanofibers or as CNF-based composites. Large changes in electrical properties, such as electrical resistance, are observed depending on the presence or absence of gaseous analytes that adsorb, bind, or electrochemically react with the CNF component. CNFs functionalized with biomolecules, such as enzymes or DNA oligomers, have been used as biosensors to detect complexation of specific proteins or complementary DNA oligomers. CNFs also act as mechanical sensors. Deflection along the vertical axis of CNFs by acoustic fields results in the generation of electrical fields that can be detected. This chapter summarizes diverse applications in which CNFs are used as sensor media. Chapter 7 - Electrospinning is a simple and versatile method for producing nanofibers from various materials including polymers, composites, carbons, ceramics, and metals. One unique and important aspect of electrospinning is its ability to manipulate the structures of nanofibers through careful control of processing parameters, including: i) intrinsic properties of the spinning solution such as rheological behavior, conductivity, surface tension, polymer molecular weight, and solution concentration; and ii) operational conditions such as voltage, solution flow rate, nozzle diameter, spinneret-collector distance, spinneret configuration, and motion of the collector. This chapter addresses the fundamental relationships between processing and structures of electrospun nanofibers and the utilization of such first-principle knowledge to achieve nanofibers with desirable structures. Nanofiber structures that are covered include fiber diameter, primary pore structure, secondary pore structure, and other secondary pore structures. The focus of this chapter is on polymer nanofibers, but electrospun fibers of other materials, such as composites, carbons, ceramics, and metals, are also discussed. Chapter 8 - As a kind of biomacromolecules, carbohydrates are found on the external surface of cell membranes in the forms of glycoproteins, glycolipids and polysaccharides. They play essential roles in biological processes such as cell adhesion, blood coagulation, viral infection, immune response and apoptosis. Numerous biological phenomena are based on the carbohydrate-protein interaction. In nature, carbohydrates always interact with specific proteins through multivalent interaction, namely “cluster glycoside effect”. A great number of glycopolymers have been designed and synthesized to mimic the multivalent functions of natural glycoconjugates. It is expected that nanostructured materials with morphology similar to the native extracellular matrix will be more interesting for the mimicking of “cluster glycoside effect”. Therefore, a series of polyacrylonitrile-based nanofibers with glycosylated surfaces were studied in our laboratory. Two protocols were used to fabricate these glycosylated nanofibers. One is the synthesis of glycopolymers followed by electrospinning and the other is the surface modification of polyacrylonitrile nanofibers having reactive groups. The authors found that the morphology of the glycosylated nanofibers could be modulated by the characteristics of glycopolymers and the parameters of electrospinning and surface modification. These glycosylated nanofibers were studied for protein adsorption and recognition. Concanavalin A (Con A), peanut agglutinin (PNA) and bovine serum albumin (BSA) were used for comparison. Water contact angle measurement confirms that the glycosylated nanofiber surface is hydrophilic which facilitates the resistance to the nonspecific adsorption of proteins. Because of the specific interaction of Con A and glucose

Preface

xi

residues, nanofibers with glucose groups have strong affinity with Con A, but present no binding with PNA or BSA. By contrast, those with galactose groups can selectively recognize PNA instead of Con A or BSA. The results suggest that the glycosylated nanofibers possess the capability to recognize the corresponding protein, which is strongly dependent on the specific carbohydrate-protein interaction. Furthermore, the adsorbed surface can be regenerated by incubation with high concentration of sugar solutions and then be reused. As a consequence, it is believed that the glycosylated nanofibers have potential applications in separation and purification of proteins. Chapter 9 - Electrospinning has been suggested as a useful method to prepare non-woven fabrics of sub-micron or nano-scale fibers, which have high porosity and large surface areato-volume ratio, small pore size between the depositing fibers of the electrospun mats, and vast possibility for surface functionalization. These characteristics make the non-woven fabrics attractive for many applications, such as functional membranes, photocatalysts, biosensors, and nanoelectronics. On the other hand, porphyrins play important roles in biological processes and much attention has been paid to design and synthesize porphyrinfunctionalized polymers for potential applications including molecular recognition or molecular imprinting, sensors, interactions with biological systems, and enzyme mimics for catalysis. Combining the merits of electrospinning with the bioinspired applications of porphyrinated polymers may generate functionalized nanofibers for more multiple purposes. Following this idea, various porphyrinated polymers, which include polyacrylonitrile, polyimide and polypeptide, were synthesized by either physical blending or chemical copolymerization. They were fabricated into nanofibrous membranes by electrospinning process. The authors found these porphyrinated polymer nanofibers not only preserved their nature characteristics but also endowed new spectroscopy properties of porphyrins. On the other hand, it is well known that porphyrins have been used as red emitting materials that have reasonable fluorescence efficiency and good thermal stability. Based on this, fluorescent microspheres or nanofibers with different diameters were prepared from the porphyrinated polymers by changing the parameters for electrospinning process, such as solution concentration and molecular weight of the polymers. Confocal laser scanning microscopy (CLSM) showed that red light emitted uniformly through out the nanofibers in spite of fiber morphologies. Besides being used as emitting materials, the authors expect that these luminescent nanofibers may be latent materials applied in many areas such as catalysis, molecular imprinting, biosensors, and light/energy conversion. Chapter 10 - Thousands of new cases of spinal cord injury occur each year in the USA alone. However, despite recent advances, there is at present no cure for the resulting paraplegia or quadriplegia. This chapter evaluates a spinal cord prosthetic (SCP) developed in our laboratoy that is comprised of longitudinally bundled strips of nanofibers whose surfaces have been modifed with fibroblast growth factor-2 (FGF-2). The SCP is designed to be a prefabricated implant that can be grafted into the lesion site not only to provide structural but also to provide chemical cues that permit regenerating axons to cross the lesion site. For a comparative study, two separate SCPs were produced with one containing unmodified nanofibers and the other containing FGF-2-modified nanofibers. Both SCPs correctly guided regenerating axons across the injury gap created by an over-hemisection to the adult rat thoracic spinal cord and encouraged revascularization of the injury site. Neither SCP initiated glial scarring when implanted into the injured rat spinal cord. However, devices that incorporated nanofibers modified with FGF-2 encouraged more axonal regrowth and

xii

W. N. Chang

significantly better functional recovery than did devices that incorporated unmodified nanofibers as assessed using the Basso, Beattie, Bresnahan (BBB) locomotor rating scale. As such, the FGF-2-modified SCP provides a multi-faceted approach to spinal cord repair. Chapter 11 - Electrospinning has recently emerged as a leading technique for the formation of nanofibrous structures made of synthetic and natural extracellular matrix components. In this chapter, nanofibrous scaffolds were obtained by electrospinning a combination of poly(3-hydroxybutyrate-co-3-hydroxyvalerate)(PHBV) and cell attachment factor such as type-I collagen, gelatin and keratin in 1,1,1,3,3,3-hexafluoro-2-isopropanol (HIFP). The resulting fibers ranged from 300 to 800 nm in diameter. Their surfaces were characterized by attenuated total reflection Fourier transform infrared spectroscopy (ATRFTIR), electron spectroscopy for chemical analysis (ESCA) and atomic force microscopy (AFM). The PHBV and protein components such as collagen and keratin were biodegraded by PHB depolymerase, type-I collagenase and trypsin solution, respectively. The results of cell adhesion experiment showed that NIH 3T3 cells more adhered to the PHBV/protein nanofibrous mats than to the PHBV nanofibrous one. It was also found, from a BrdU assay, that the PHBV/protein nanofibrous mats could accelerate the proliferation of fibroblast cells more effectively than the PHBV nanofibrous mats, suggesting the good scaffolds for tissue engineering. Chapter 12 - In this study, vapor grown carbon nanofiber (CNF) has been used to improve thermal and mechanical properties of polypropylene. The CNFs were first dispersed over the polypropylene particles using sonication coating method, and then extruded into filaments with a single screw extruder. The thermal properties of neat and nanophased polypropylene were characterized by TGA and DSC. TGA thermograms showed that the nanoparticle-infused systems are more thermally stable, and DSC results indicated that CNFs have no effect on melting temperature. Tensile tests were performed on the single filament at a strain rate range from 0.02/min to 2/min. Results indicate that both neat and nanophased polypropylene were strain rate-strengthening material. The tensile modulus and yield strength both increased with increasing strain rate. Experiment results also show that infusing polypropylene with nanofibers increases tensile modulus and yield strength, but decreases ductility. Finally, based on the tensile test results, a nonlinear constitutive equation was developed to describe strain rate-sensitive behavior of neat and nanophased polypropylene. Chapter 13 - Carbon nanotubes (CNTs) have a number of outstanding mechanical and physical properties which make them attractive as reinforcement in polymer matrix. CNTs reinforced polyurethane nanocomposites provide the possibility to tailor the material strength, stiffness and thermal behavior of polyurethane. Multi-walled carbon nanotubes (MWNTs) filled polyurethane composites were prepared by mixing and injection molding and its thermal characteristics were investigated. The analysis of the influences of MWNTs particles and composites mixing methods were done using dynamic mechanical analysis (DMA), differential scanning calorimetry (DSC) and thermogravimetry analysis (TGA). The storage moduli of the composites increased with increasing MWNTs loading which indicate a good matrix/filler adhesion and increased the stiffness of the composites. However, the increase in processing speed has decrease the storage modulus. Addition of MWNTs filler also broadened and lowered the peak of tan δ denotes that the polyurethane composite became more elastic. Thermal stability of the polyurethane was improved with MWNTs loading which is associated to high thermal stability of CNTs. A very high processing speed reduced the composites thermal stability making it easier to degrade. DSC analysis indicated that the

Preface

xiii

inclusion of MWNTs increased the melting temperature and act as restriction sites for the polyurethanes soft segments. The present chapter revealed the potential of MWNTs as agent for better thermal properties of polyurethane nanocomposites and their properties depend strongly on the dispersion and distribution of nanotubes in polyurethane matrix. Chapter 14 - Polyurethane composites filled with multi-walled carbon nanotubes (MWNTs) were prepared by mixing and injection molding and its thermal as well as electrical conductivity characteristics were investigated. The influences of MWNTs addition and mixing methods on thermal and electrical conductivity of MWNT/polyurethane nanocomposites were investigated using a high resistance meter and thermal conductivity analyzer. The electrical resistivity of MWNTs incorporated polyurethanes were decreased in relation to filler concentration which is attributed by the formation of a conductive path made up from MWNTs particle. Increasing the processing speed will further decrease the resistivity because the dispersion of CNTs in polymer is improved. Higher processing speed samples shows resistivity values closer to the theoretical value because of better dispersion of CNTs in polyurethane and more conductive pathway were formed. The result shows that the addition of MWNTs fillers improved the thermal conductivity of the polyurethane composites. Higher filler concentration and higher shear rate results in better thermal conductivity because better formation of thermally conductive networks along polymer matrix to ensure the thermal was conducted through the matrix and the network along the polymer composites. The theoretical thermal conductivity comparisons approximately agree with the experimental measurements for the composites studied. The present study revealed the potential of MWNTs as agent for better thermal and electrical conductivities of polyurethane nanocomposites and their properties depend strongly on the dispersion and distribution of nanotubes in polyurethane matrix.

In: Nanofibers: Fabrication, Performance, and Applications ISBN 978-1-60741-947-1 Editors: W. N. Chang © 2009 Nova Science Publishers, Inc.

Chapter 1

SYNTHESIS AND ELECTRON FIELD EMISSION FROM DIFFERENT MORPHOLOGY CARBON NANOFIBERS Sk. F. Ahmed1,a and K. K. Chattopadhyay2,b 1

Future Fusion Technology Laboratory, Korea Institute of Science and Technology, P.O. Box 131, Cheongryang, Seoul 130-650, Republic of Korea 2 Thin Film and Nanoscience Laboratory, Department of Physics, Jadavpur University, Kolkata, India

1. INTRODUCTION Carbon Nanofiber and Carbon Nanotube Carbon, which belongs to group IV of the periodic table, is the lightest element in this group, and it possesses countless interesting physical and chemical properties. Among the different types of supports used in heterogeneous catalysis carbon materials attract a growing interest due to their specific characteristics which are mainly: (i) resistance to acid / basic media, (ii) possibility to control, up to certain limits, the porosity and surface chemistry and (iii) easy recovery of precious metals by support burning resulting in a low environmental impact. In contrast to Si, Ge and Sn, which have the same number of electrons in the outermost shell as carbon and can only exist in cubic sp3 hybridization, carbon not only exhibits sp3 hybridization (diamond), but also planar sp2 hybridization as in the graphite structure and sp1 hybridization as in carbynes. Each carbon atom has six electrons which occupy 1s2, 2s2 and 2p2 atomic orbitals. The 1s2 orbital contains two strongly bound core electrons. Four more weakly bound electrons occupy the 2s22p2 valence orbitals. In the crystalline phase, the valence electrons give rise to 2s, 2px, 2py and 2pz orbitals which are important in forming covalent bonds in carbon materials. Since the energy difference between the upper 2p energy levels and the lower 2s level in carbon is small compared with the binding energy of the chemical bonds, the electronic wave functions for these four electrons aE-mail: [email protected] (Sk. F. Ahmed). bE-mail: [email protected] (K. K. Chattopadhyay, Corresponding author).

2

Sk. F. Ahmed and K. K. Chattopadhyay

can readily mix with each other, thereby changing the occupation of the 2s and three 2p atomic orbitals so as to enhance the binding energy of the carbon atom with its neighboring atoms. The general mixing of 2s and 2p atomic orbitals is called hybridization, whereas the mixing of a single 2s electron with one, two, or three 2p electrons is called spn hybridization with n = 1,2,3 [1,2]. The bonding structures of diamond, graphite and nanotubes, or fullerenes are shown in Figure 1.1. When a graphite sheet is rolled over to form a nanotube, the sp2 hybrid orbital is deformed for rehybridization of sp2 toward sp3 orbital or σ−π bond mixing. This rehybridization structural feature, together with electron confinement, gives nanofibers/nanotubes unique, extraordinary electronic, mechanical, chemical, thermal, magnetic, and optical properties [3-5]. The physical reason why these nanostructures form is that a graphene layer (defined as a single 2D layer of 3D graphite) of finite size has many edge atoms with dangling bonds, index dangling bonds and these dangling bonds correspond to higher energy states. Therefore the total energy of a small number of carbon atoms (30 -100) is reduced by eliminating dangling bonds, even at the expense of increasing the strain energy, thereby promoting the formation of closed cage clusters such as fullerenes and carbon nanotubes. For example, diamond and layered graphite forms of carbon are well known, but the same carbon also exists also in planar sheet, rolled up tubular, helical spring, rectangular hollow box, and nanoconical forms. Elemental carbon in the sp2 hybridization can form a variety of amazing structures. Apart from the well-known graphite, carbon can build closed and open cages with honeycomb atomic arrangement. First such structure to be discovered was the C60 molecule by Kroto et al. in 1985 [1]. Although various carbon cages were studied, it was only in 1991, when Iijima [6] observed for the first time tubular carbon structures. Two years later, Iijima and Ichihashi [7] and Bethune et al. [8] synthesized single-walled carbon nanotubes (SWNTs). Actually carbon nanotubes (CNTs) are allotropes of carbon. Nanotubes are members of the fullerene structural family, which also includes buckyballs. Whereas buckyballs are spherical in shape, a nanotube is cylindrical, with at least one end typically capped with a hemisphere of the buckyball structure. The nature of the bonding of a nanotube is described by applied quantum chemistry, specifically, orbital hybridization. Nanotubes are composed entirely of sp2 bonds, similar to those of graphite. This bonding structure, which is stronger than the sp3 bonds found in diamond, provides the molecules with their unique strength. Nanotubes naturally align themselves into "ropes" held together by Van der Waals forces. Under high pressure, nanotubes can merge together, trading some sp2 bonds for sp3 bonds, giving great possibility for producing strong, unlimited-length wires through highpressure nanotube linking [9]. A single-wall carbon nanotube (SWCNT) is best described as a rolled-up tubular shell of graphene sheet (Figure 1.2(a)) which is made of benzene-type hexagonal rings of carbon atoms [10-12]. There are many possible orientations of the hexagons on the nanotubes, even though the basic shape of the carbon nanotube wall is a cylinder. A single walled carbon nanotube is a graphene sheet appropriately rolled into a cylinder of nanometer size diameter [13,14]. The planar sp2 bonding, which is characteristic of graphite, plays a significant role in carbon nanotubes. The body of the tubular shell is thus mainly made of hexagonal rings (in a sheet) of carbon atoms, whereas the ends are capped by half-dome shaped half-fullerene molecules. The internal diameter of these structures can vary between 0.4 to 2.5 nm and the length ranges from few microns to several millimeters.

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers

3

Figure 1.1. Bonding structures of diamond, graphite, nanotubes and fullerenes[From ref. 3].

Since the single wall carbon nanotube is only one atom thick and has a small number of atoms around its circumference, only a few wave vectors are needed to describe the periodicity of the nanotubes. These constraints lead to quantum confinement of the wave functions in the radial and circumferential directions, with plane wave motion occurring only along the nanotube axis corresponding to a large number or closely spaced allowed wave

4

Sk. F. Ahmed and K. K. Chattopadhyay

vectors. Thus, although carbon nanotubes are closely related to a 2D graphene sheet, the tube curvature and the quantum confinement in the circumferential direction lead to a host of properties that are different from those of a graphene sheet. A multi-wall carbon nanotube (MWCNT) is a rolled-up stack of graphene sheets of coaxial SWCNTs, with the ends again either capped by half-fullerenes or kept open. Both SWCNTs and MWCNTs have physical characteristics of solids and are nanocrystals with high aspect ratios of 1000 or more, although their diameter is close to molecular dimensions. The number of walls present can vary from two (double wall nanotubes) to several tens, so that the external diameter can reach upto 100 nm. The concentric walls are regularly spaced by 0.34 nm similar to the inter graphene distance evidenced in turbostatic graphite materials. The main difference between nanotubes and nanofibers consists in the lack of a hollow cavity for the latter. The diameters of carbon nanofiber (CNF) are generally higher than the ones presented by nanotubes and can easily reach 500 nm. A nomenclature (n,m), used to identify each single-wall nanotube, refers to integer indices of two graphene unit lattice vectors corresponding to the chiral vector of a nanotube. Chiral vectors determine the directions along which the graphene sheets are rolled to form tubular shell structures and perpendicular to the tube axis vectors. Figure 1.2 shows the schematic representation of the construction of a nanotube by rolling-up an infinite strip of graphite sheet (so called graphene). In Figure 1.2(a) the chiral vector Ch = na1 + ma2 connects two lattice points O and A on the graphene sheet, where n and m are integers, a1 and a2 the unit cell vectors of the two-dimensional lattice formed by the graphene sheets. The direction of the nanotube axis is perpendicular to this chiral vector. An infinite strip is cut from the sheet through these two points, perpendicular to the chiral vector. The strip is then rolled-up into a seamless cylinder. T = t1a1 + t2a2 is the primitive translation vector of the tube [15]. The nanotube is uniquely specified by the pair of integer numbers n, m or by its radius R = Ch / 2π and chiral angle θ which is the angle between Ch and the nearest zigzag of C–C bonds. All different tubes have angles θ between zero and 30o. Special tube types are the achiral tubes (tubes with mirror symmetry): when n = m, the nanotube is called “armchair” type (θ = 0o) [Figure 1.2(b)]; when m = 0, then it is of the “zigzag” type (θ = 30o) [Figure 1.2(c)]. Otherwise, when n = m, it is a “chiral” tube and θ takes a value between 0o and 30o [Figure 1.2(d)]. The value of (n,m) determines the chirality of the nanotube and affects the optical, mechanical and electronic properties. Nanotubes with |n - m| = 3q are metallic and those with |n - m| = 3q ± 1 are semiconducting (q is an integer). Exhaustive studies concerning electronic properties of both SWCNT [16] and MWCNT [17] are available in the literature, whereas carbon nanofiber (CNF) are often considered as conductive substrates that can exert electronic perturbations similar to those of graphite [18]. In the case of SWCNT, studies have demonstrated that they behave like pure quantum wires (1D-system) where the electrons are confined along the tube axis. Electronic properties are mainly governed by two factors: the tube diameter and the helicity, which is defined by the way in which the graphene layer is rolled up (armchair, zigzag or chiral) [8]. In particular, armchair SWCNTs are metallic and zigzag ones display a semi-conductor behavior. Studies on MWCNTs electronic properties have revealed that they behave like an ultimate carbon fiber at high temperature their electrical conductivity may be described by semi-classical models already used for graphite, whereas at low temperature they reveal 2D-quantum transport features [17]. Nanotubes can be as small as 1 nm in diameter and as long as 100,000 nm.

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers

5

Figure 1.2. Schematic representation of the construction of a nanotube by rolling-up an infinite strip of graphite sheet.

These tubes are extremely strong, approaching the strength of diamond, and also dissipate heat better than any other known material. Carbon nanofibers/nanotubes are one of the strongest and stiffest materials known, in terms of tensile strength and elastic modulus respectively.

6

Sk. F. Ahmed and K. K. Chattopadhyay

This strength results from the covalent sp2 bonds formed between the individual carbon atoms. Depending on how they are configured, CNFs/CNTs are good conductors of electricity and can also act as semi-conductors for molecular electronics. CNFs/CNTs are three dimensional as opposed to the current silicon based electronics that are two-dimensional. They appear to be able to extend the miniaturization process by several additional orders of magnitude over current methods. As they conduct electricity better than copper and can also act as semi-conductors. CNFs/CNTs have very good elasto-mechanical properties because the two dimensional arrangement of carbon atoms in a graphene sheet allows large out-of-plane distortions, while the strength of carbon-carbon in-plane bonds keeps the graphene sheet exceptionally strong against any in-plane distortion or fracture. Some of the properties of carbon nanotube, which are forming a driving force for their wide range of applications, are shown in Table -1. Table 1. Some fundamental properties of carbon nanofiber/nanotube Properties Average Diameter: SWNT's MWCNTs CNFs Young's Modulus Maximum Tensile Strength Band Gap: For Metallic For Semi-Conducting Thermal Conductivity (Room Temp.) Carrier mobility Semi-Conducting NT Maximum Current Density Turn-on field Threshold field for CNTs/CNFs

Value 1.2 – 1.4 nm 2 – 100 nm 10 nm – 100 μm ~ 1 TPa ~ 63 Gpa 0 eV 0.18 – 1.8 eV 3000 W/mK 105 cm2/Vsec 109 A/m2 1.5 – 7.5 V/μm 1.5 – 9.5 V/μm

References [19] [20] [21] [22,23] [24] [25] [25] [26] [27] [28] [29,30] [31,32]

History of Carbon Nanofiber and Nanotube We provide here a brief review of the history of carbon fibers, the macroscopic analog of carbon nanotubes, as carbon nanotubes have become the focus of recent developments in carbon fibers. Since last decade, new carbon forms like carbon nanofibers (CNF) or nano filaments and carbon nanotubes (CNT) have generated an interest in the scientific community. However, it has got to be remembered that carbon nano filaments have been synthesized for very long as products from the action of a catalyst over the gaseous species originating from the thermal decomposition of hydrocarbons. One of the first evidence that the nano filaments thus produced could have been nanotubes, exhibiting an inner cavity, can be found in the transmission electron microscope micrographs published by Hillert et al. in the year of 1958 [33]. Radushkevich et el. published clear images of 50 nanometer diameter tubes made of carbon in the year 1952 [34]. This discovery was largely unnoticed, the article was published in the Russian language. The production of graphite nanofibers is even older and the first reports date of more than a century [35,36]. Their efforts were mostly directed toward the study of vapor grown carbon filaments, showing filament growth from the thermal

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers

7

decomposition of hydrocarbons. The second applications-driven stimulus to carbon fiber research came in the 1950’s from the needs of the space and aircraft industry for strong, stiff light-weight fibers that could be used for building lightweight composite materials with superior mechanical properties. This stimulation led to great advances in the preparation of continuous carbon fibers based on polymer precursors, including rayon, polyacrylonitrile (PAN) and later mesophase pitch. The late 1950’s and 1960’s was a period of intense activity at the Union Carbide Corporation, the Aerospace Corporation and many other laboratories worldwide. This stimulation also led to the growth of a carbon whisker [37], which has become a benchmark for the discussion of the mechanical and elastic properties of carbon fibers. The growth of carbon whiskers was also inspired by the successful growth of single crystal whisker filaments at that time for many metals such as iron, non-metals such as Si, and oxides such as Al2O3, and by theoretical studies [38], showing superior mechanical properties for whisker structures [39]. Parallel efforts to develop new bulk synthetic carbon materials with properties approaching single crystal graphite led to the development of highly oriented pyrolytic graphite (HOPG) in 1962 by Ubbelohde and co-workers [40,41], and HOPG has since been used as one of the benchmarks for the characterization of carbon fibers. While intense effort continued toward perfecting synthetic filamentary carbon materials, and great progress was indeed made in the early 1960’s, it was soon realized that long term effort would be needed to reduce fiber defects and to enhance structures resistive to crack propagation. New research directions were introduced because of the difficulty in improving the structure and microstructure of polymer-based carbon fibers for high strength and high modulus applications, and in developing graphitizable carbons for ultra-high modulus fibers. Because of the desire to synthesize more crystalline filamentous carbons under controlled conditions, synthesis of carbon fibers by a catalytic Chemical Vapor Deposition (CVD) process was developed, laying the scientific basis for the mechanism and thermodynamics for the vapor phase growth of carbon fibers in the 1960’s and early 1970’s. In parallel to these scientific studies, other research studies focused on control of the process for the synthesis of vapor grown carbon fiber [42-45], leading to the more recent commercialization of vapor grown carbon fibers in the 1990’s for various applications. Concurrently, polymer-based carbon fiber research has continued worldwide, mostly in industry, with emphasis on greater control of processing steps to achieve carbon fibers with ever-increasing modulus and strength, and on fibers with special characteristics, such as very high thermal conductivity, while decreasing costs of the commercial products. As research on vapor grown carbon fibers on the micrometer scale proceeded, the growth of very small diameter filaments less than 10 nm, was occasionally observed and reported [46-47], but no detailed systematic studies of such thin filaments were carried out. Oberlin et al. clearly showed hollow carbon fibres with nanometer-scale diameters using a vapour-growth technique [48]. The interest in fibrous carbon has since then been recurrent and a significant boost in the research in carbon nanostructure field coincides with the discovery of multiwall carbon nanotubes (MWNT) by Iijima in 1991 [6]. It is likely that carbon nanotubes were produced before this date, but the invention of the transmission electron microscope allowed the direct visualization of these structures. Carbon nanotubes have been produced and observed under a variety of conditions prior to 1991. The arc discharge technique was well known to produce the famed Buckminster fullerene on a preparative scale [49] and these results appeared to extend the run of accidental discoveries relating to fullerenes. The original observation of fullerenes in mass

8

Sk. F. Ahmed and K. K. Chattopadhyay

spectrometry was not anticipated [5], and the first mass-production technique by Kratschmer et al. was used for several years before realising that it produced fullerenes [49].

2. SYNTHESIS AND GROWTH MECHANISM OF CARBON NANOTUBE AND CARBON NANOFIBER BY DIFFERENT PROCESS There have been major developments in the synthesis processes and characterization techniques of CNT and CNF. It is possible to produce CNTs by a wide range of deposition methods. The most popular deposition techniques are arc discharge [6,50], laser ablation [51], chemical vapour deposition (CVD) [52,58-62] plasma enhanced chemical vapour deposition [53] and solvothermal process [54-55]. The earliest approach to produce nanotubes was an arc process [56] as pioneered by Iijima in 1991. This was shortly followed by a laser ablation technique developed at Rice University [57]. Chemical vapor deposition (CVD) has become a common technique to grow nanotubes in the last ten years [58-62]. The figure-of-merit for an ideal growth process depends on the application. For development of composites and other structural applications, the expected metric is the ability to achieve controlled growth of specified thickness on patterns is important for applications in nano electronics, field emission, displays, and sensors. The arc process involves striking a dc arc discharge in an inert gas (such as argon or helium) between a set of graphite electrodes [6,56]. The electric arc vaporizes a hollow graphite anode packed with a mixture of a transition metal (such as Fe, Co or Ni) and graphite powder. The inert gas flow is maintained at 50-600 Torr. Nominal conditions involve 2000 - 3000° C, 100 amps and 20 volts. This produces SWCNTs in mixture of MWCNTs and soot. In the arc-discharge synthesis of nanotubes, Bethune et al. in 1993 used as anodes thin electrodes with bored holes which were filled with a mixture of pure powdered metals (Fe, Ni or Co) and graphite [8]. The electrodes were vaporized with a current of 95-105 A in 100-500 Torr of He. The gas pressure, flow rate, and metal concentration can be varied to change the yield of nanotubes, but these parameters do not seem to change the diameter distribution. Typical diameter distribution of SWCNTs by this process appears to be 0.7 - 2 nm. On the other hand in laser ablation method, a target consisting of graphite mixed with a small amount of transition metal particles such as such as Co, Ni, Fe etc., catalyst is placed at the end of a quartz tube enclosed in a furnace [57]. A neodymium-yttrium-aluminum-garnet laser was employed to vaporize the target and helium or argon carrier gas was flowed through the tube. Argon gas flowing through the reactor, heated about 1200 °C by a furnace, carries the vapor and nucleates the nanotubes which continue to grow. The nanotubes are deposited on the cooler walls of the quartz tube downstream from the furnace. Both the above-mentioned methods were used to synthesize SWNTs in relatively large quantities. They are based on the condensation of hot carbon gases through vaporizing solid carbon, where temperatures of > 3000 K are initialized by either arc or laser. Due to such a high temperature, carbon nanotubes obtained exhibit high straightness and high crystallinity. However, depositions by both approaches are not directly on the substrates and are in a form of either powder or mat. Thus applications require additional manipulation or processes to deposit the CNTs on substrates. Lastly, the products obtained normally contain a large quantity of catalyst and carbon particles so that a purification step becomes necessary.

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers

9

Chemical vapor deposition (CVD) methods have been successful in making carbon fiber and filament since more than 10-30 years ago [46,63,64], but in recent years, chemical vapor deposition (CVD) has progressed rapidly towards the growth of carbon nanotubes and has been adopted worldwide [65]. In this method, transition catalysts such as Fe, Co, Ni are deposited on the substrates (silicon, quartz, molybdenum etc.). A mixture of precursors such as hydrogen, methane, acetylene, and ammonia are flowed into the chamber. Assisted by either direct heating or by an external source such as plasma and hot filament, carbon gases are decomposed at the catalyst surface and carbon nanotubes are grown directly on the substrates at temperatures between ~ 700-1400 K [52]. The advantage of this method is that the nanotubes can be deposited directly on the substrate, which facilitates nanotube applications and mass production. Nevertheless, MWNTs are often obtained by CVD and only until recently SWNT growth by CVD was possible [66]. The drawback of the catalytic CVD-based nanotube production is the inferior quality of the structures that contain gross defects (twists, tilt boundaries etc.), particularly because the structures are created at much lower temperatures (600-1000 oC) compared to the arc or laser processes (~2000 oC). Several plasma based growth techniques have been reported [67-69] and in general, the plasma-grown nanotubes appear to be more vertically oriented than that is possible by thermal CVD. Since the plasma is very efficient in tearing apart the precursors and creating radicals, it is also hard to control and keep the supply of carbon low to the catalyst particles and hence, plasma based growth always results in MWCNTs and filaments. Different CVD systems which employ different approaches to dissociate the precursor gases have been used including thermal CVD [70], hot filament chemical vapour deposition (HFCVD) [53], and plasma enhanced chemical vapor deposition (PECVD) [71]. Several types of plasma systems which have been used including dc-plasma [72], radio frequency (RF) plasma [73], microwave plasma [74], and electron cyclotron resonance (ECR). In CVD technique, CNTs and CNFs are grown using the catalytic decomposition of hydrocarbons over transition metal catalysts [52]. The function of these metals is to facilitate the decomposition of the hydrocarbon gases and the formation of the tubular graphene structures. Chemical composition and particle size of the catalyst is expected to crucially affect the diameter and the number of walls of the carbon nanotubes [75]. The metal catalysts have been prepared by several methods including wet chemical solution [76], thin metal films [71], thick metal films/substrates [77], colloids [75], and solgel techniques [61]. In some cases surface treatments such as wet chemical etching [67], plasma etching [77], ion beam sputtering [62] and annealing [71] have been used to enable the formation of nano-particles before the growth. Effect of catalyst on growth of carbon nanofibers have been studied by Kamada et al. [78]. They have been successfully grown carbon nanofiber (CNF) films on Pd-Se, Fe-Ni, and Ni-Cu alloy catalysts at low temperatures by a thermal chemical vapor deposition method. Among these alloy catalysts, Ni-Cu alloy catalyst was found to be most suitable for low temperature growth of CNF. The CNFs grown using Pd-Se catalyst were found to have more defective structure than that obtained with the other catalysts, and exhibited excellent field emission property with threshold field ~ 1.1 V/μm. Merkulov et al. [79] evaporated Ni on n-type Si by E-gun. Shyu et al. deposited Fe-Ni with various components by e-beam evaporation [80]. Kin et al. [81] prepared copper-nickel powder by coprecipitation of the metal carbonates from mixed nitrate solutions using ammonium bicarbonate and a sequence complex treatment process including drying, calcining and reducing, etc. Carbon nanofibers (CNFs) were grown on a Ni–P alloy catalyst deposited on a silicon substrate in a microwave heating chemical vapor deposition system

10

Sk. F. Ahmed and K. K. Chattopadhyay

with methane gas at 650 oC [82]. The nanosized clusters on the clustered surface of the Ni–P alloy catalyst film directly provided the nucleation sites for CNFs without any pretreatment before the growth of the CNFs. The CNFs grown on the Ni–P alloy catalyst showed random orientation and it composed of parallel graphite planes. Figure 2.1 shows scanning electron microscopy images of CNFs grown at substrate temperature approximately 650 oC for 7 min. It reveals that the growth rate of CNFs is related to the thickness of Ni–P alloy catalyst film. The results indicate that the growth rate of the CNFs decreases as the thickness of the catalyst film increases. The above phenomenon explained by the diffusion of carbon atom into the catalyst particle. The growth of carbon nanostructures, including CNTs and CNFs, occurs by diffusion driven precipitation of carbon atoms from the supersaturated catalyst particles [83]. The size of catalyst particle increases and that causes the diffusion length to increase and the gradient of supersaturation to decrease. These factors will decrease the growth rate of CNFs. So, the thin catalyst film has a larger growth rate than the thick catalyst film. The result also proves that diffusion of carbon through the catalyst particle is the rate-determining step in the growth of carbon nanostructures using a Ni-P alloy catalyst.

Figure 2.1. SEM images of CNFs grown at 650 oC substrate temperature for 7 min. The corresponding thickness of Ni–P alloy catalyst film is 20 and 40 nm in images (a) and (b) respectively [From ref. 82].

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers 11 Figure 2.2 is the SEM micrograph showing the surface morphology of CNFs grown at approximately 650 oC for 10 min. All of these SEM images show that these CNFs grown on the catalyst film with various thicknesses have similar morphology and are not vertically aligned but randomly tangled. The diameter of the CNFs in Figure 2.2(a)-(c) is approximately 30-70, 50-120 and 70-150 nm, respectively. The diameter of the CNFs increases as the size of the catalyst clusters increases with catalyst film thickness. The results show that the diameter of the CNFs is dependent on the initial thickness of the pre-deposited catalyst film. Wei et al. [69], Yudasaka et al. [84] and Bower et al. [85] also reported same type results using Fe, Co and Ni catalyst.

Figure 2.2. SEM images of CNFs grown at 650 oC substrate temperature for 10 min. The corresponding thickness of Ni–P alloy catalyst film is 20, 30 and 40 nm in images (a), (b) and (c), respectively [from ref. 82].

Vertically aligned carbon nanofiber and naotubes were synthesized by different technique and there are different growth mechanism proposed by different groups [86-92]. The physical and chemical characteristics of vertically aligned carbon nanofiber (VACNF) structures, in comparison to ideal multiwalled carbon nanotubes, offer inherent processing advantages imparted by their vertical architecture. The ability to control the VACNF growth rate is an important practical aspect of the synthesis process because some applications require high growth rates, whereas others would benefit from a lower growth rate but a high degree of uniformity and control over the final VACNF length. Thus understanding the factors that

12

Sk. F. Ahmed and K. K. Chattopadhyay

determine the growth rate is essential not only from the fundamental science point of view, but also from the point of view of practical applications. In conventional thermal CVD the growth of carbon nanofibers/nanotubes occurs in three main steps [63]; (i) decomposition of the carbonaceous gas molecules at the surface of the catalyst nanoparticle, (ii) diffusion of the resultant carbon atoms through the catalyst nanoparticle from the nanoparticle/gas interface towards the nanoparticle/nanofiber interface due to the concentration gradient, and (iii) precipitation of carbon atoms at the nanoparticle/nanofiber interface. Each of these steps can be a complex process by itself [93,94] and the whole picture is not completely understood, mainly due to the difficulty of conducting imaging and surface analysis in situ during the growth. In thermal CVD, the carbon feedstock is the molecules of the carbonaceous gas used in the growth process. The growth rate is determined by all of the three steps, but under typical growth conditions appears to be diffusion limited as suggested by the equality of the activation energies for the nanofiber growth rate and for the diffusion of carbon atoms through the catalyst [95]. Chuang et al. [86] prepared carbon nanofibers and carbon nanotubes using CH4 as a precursor material of carbon on Ni/Si and Ni/Ti/Si substrates at 640 oC and 700 oC by thermal chemical vapor deposition method. They have explained the growth of carbon nanofibers (CNF) and carbon nanotubes (CNT) on Ni/Si substrate through tip growth mechanism, and the growth mechanism of carbon nanotubes on Ni/Ti/Si substrate through root growth mechanism [87]. Figure 2.3(a) schematically shows the solid amorphous carbon nanofiber grown on Si substrate at 640 oC under the catalysis of Ni particle. During the heating of Ni/Si substrate from room temperature to 640 oC, the Ni film on Si surface would agglomerate into Ni nanoparticles, which still are in solid state. In the growth chamber the CH4 molecules are adsorbed on the upper surface of Ni particles and decomposed into C and H atoms. These C atoms are absorbed into Ni particles and diffuse to lower part through inside of Ni particles. Due to the difference of both temperature and concentration of carbons between upper and lower parts of Ni particles and the weak adhesion force between Ni particle and Si surface, once the concentration of C atoms exceeds the saturation solubility of Ni, the C atoms would precipitate on the lower surface of Ni particles and form carbon nanofiber. Because of the roughness of Ni particle surface, the precipitated carbons could not form graphene planes on the lower surface and hence only amorphous solid carbon nanofibers are obtained. Figure 2.3(b) schematically displays the hollow carbon nanofiber, i.e. carbon nanotube, is grown on Si substrate at 700 oC under the catalysis of Ni particle. When the growth temperature is raised from 640 - 700 oC, the Ni nanoparticles become fluid-like due to the lower melting temperature of Ni–C alloy and the effects of size and interfacial stress between Ni particle and carbon tube [96]. The surface of fluid-like Ni nanoparticle is much smoother than solid Ni nanoparticle, and the diffusion of C atoms on fluid-like Ni nanoparticle surface is much higher than on the surface of solid Ni nanoparticle. The C atoms are much easy to precipitate on the surface of middle part, not the lower part, of Ni nanoparticle, and form the parallel grapheme planes. Therefore, the carbon nanotubes are grown through the tip growth mechanism. Figure 2.3(c) schematically shows the hollow carbon nanofiber, i.e. carbon nanotube, is grown on Ti/Si substrate at 700 oC under the catalysis of Ni particle. Due to the strong adhesion force between Ni nanoparticle and Ti film surface, the growth force cannot push Ni nanoparticle up. Therefore, the carbon nanotubes are grown through the root growth mechanism. When the temperature of Ni/Ti/Si substrate increased, a part of Ni diffuse through Ti layer and into Si substrate to form NiSi and the rest of Ni would agglomerate to

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers 13 form Ni particles [97]. The C atoms from the decomposition of CH4 under the catalysis of Ni particles diffuse into Ni particles and Ti layer. When ternary alloy of Ni-Ti-C was formed at the lower part of Ni particle due to the interdiffusion of Ni, Ti and C, the ternary alloy layer at the bottom of Ni particle would melt due to the lower melting point of the Ni-Ti-C ternary alloy [98]. Although the eutectic temperature of Ni-C is 1327 °C, there are many reports [99,100] indicating that the Ni catalyst particle had melted or was behaved fluid like during CNTs growth in the temperature of 600-900 °C due to the size effect of catalyst at nanometer level and the interfacial effect between catalyst and carbon. The eutectic temperature of NiTi-C is 1265-1295 °C, lower than 1327 °C, hence the Ni-Ti-C alloy layer at the bottom of Ni particle can melt due to the same reasons for Ni-C. Therefore Ni particles could sink into and firmly adhere to the Ti layer. When C atoms in Ni particles exceed the saturation solubility, they would form CNTs and grow up under the catalysis of Ni particles by root growth mechanism. Once Ni particles were surrounded by CNTs, the possibility for CH4 to touch Ni particles and decompose into C and H would decrease largely. For making CNT growth able to continue, the C atoms supplied from Ti layer. Figure 2.4 schematically shows the layer structure of substrates before and after CNTs growth. The three-layered Ni/Ti/Si substrate becomes four layers structure after CNTs growth. From the AES depth profile measurement of each element and the layer structure of sample they confirmed that the Ti interlayer couldn't prevent the outward diffusion of Si and the inward diffusion of Ni, C and Ti; however it can supply C atoms to continue the CNTs growth at later stage in CNT growth process.

Figure 2.3. Schematic diagrams of one dimensional carbon growth mechanism for solid amorphous carbon nanofiber grown through tip mechanism (a), carbon nanotube grown through tip mechanism (b), and carbon nanotube grown through root mechanism (c) [ From Ref. 86].

14

Sk. F. Ahmed and K. K. Chattopadhyay

Figure 2.4. Schematic diagrams for substrate (a) before CNTs growth and (b) mechanism for CNT growth [From Ref. 87].

Plasma enhanced chemical vapor deposition technique has achieved lower growth temperatures compared to other methods and vertical alignment of the nanotubes, which facilitates the CNT based device fabrication with current silicon technologies. It has been suggested [88] that in PECVD only VACNFs grown from the tip are aligned specifically due to the presence of the plasma electric field in the growth process, whereas VACNFs grown from the base are aligned mainly due to the crowding effect. Consequently, in the case of the base-type growth, deterministic synthesis of isolated VACNFs is expected to be rather difficult. Vertically aligned carbon nanofibers (VACNFs) were synthesized by direct-current plasma enhanced chemical vapor deposition using acetylene and ammonia as the gas source as shown in Figure 2.5 [101]. Generally in the case of PECVD growth the activation energies were found to be different [71] and the growth rate was suggested to be limited by the supply of carbon from the gas phase. Huang et al. showed that the VACNF growth rate depends quite strongly upon the gas mixture and plasma power used in the PECVD process [102]. This was attributed to changes in the chemical composition of the excited gas species. These species include (i) simple radicals created in the plasma as a result of direct dissociation of C2H2 and NH3 molecules and (ii) larger radicals that form due to collision and consequent attachment of radicals to each other as they move towards the substrate. It was suggested that changing the gas mixture or plasma power changes the chemical distribution of the excited species. Since different species are expected to have different decomposition rates at the catalyst surface, the nanofiber growth rate also changes. The fact that reduction of the C2H2 content resulted in a several-fold increase of the growth rate as well as a dramatic increase of the nitrogen content within the nanofibers strongly suggests that the growth occurs mainly due to the species created in the plasma, not due to unexcited C2H2 molecules.

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers 15

Figure 2.5. SEM image of an array of VACNFs grown by PECVD for 5 min and of randomly oriented CNFs/CNTs produced after the plasma was turned off and the growth run continued via purely thermal CVD [From Ref.101].

Figure 2.5 shows the SEM image of a sample for which VACNFs were first grown using a plasma, then the plasma was turned off, the NH3 flow was stopped, and C2H2 was flowed over the sample that was kept at the same high temperature as during the PECVD part of the growth (~700 oC). During the purely thermal phase, the VACNFs did not seem to increase in length, but very long thin non-aligned CNFs/CNTs that were not present during the PECVD part of the growth were produced. The non-aligned CNFs seemed to originate from the bases of VACNF, where perhaps some of the catalyst was still left. This indicates that the catalytic activities are quite different for the base- and tip-type growth modes. While the base-type catalytic growth can produce CNFs in the purely thermal process, the tip-type growth, requires the presence of radicals. The inability to re-grow VACNFs using only thermal CVD confirms the idea of the feedstock for the VACNF growth consisting mainly of radicals and not C2H2 molecules. There is another reason that a carbon shell may form around the catalyst nanoparticles sitting at the tips of VACNFs after the plasma is turned off, which prevents decomposition of carbonaceous species at the catalyst surface and consequently the VACNF growth. Since some applications may require very long VACNFs, it is highly desirable to develop controllable ways to further increase the growth rate. One way to relax the limit for the VACNF growth rate, imposed by the process of radical diffusion towards the substrate, is to change the radical transport mechanism from diffusive to forced flow by applying a pressure gradient perpendicular to the substrate surface to force more radicals to impinge on the surface [103]. The radical flux in this case will be given by F = Cv, where C is the radical concentration and v is an average velocity towards the substrate. Thus, by increasing the gas flow and therefore the radical velocity one can expect to achieve substantial increase in the VACNF growth rate. The high gas flow during the growth can be produced highly conical structures even for dense forests of VACNFs. Isolated VACNFs tend to assume conical shape during the growth due to reactive species emerging for the discharge and attaching to the

16

Sk. F. Ahmed and K. K. Chattopadhyay

sidewalls of the nanofibers [104]. In contrast, dense VACNF forests nominally consist of essentially cylindrical nanofibers due to the shielding of the sidewalls by neighboring VACNFs [79]. However, if the velocity of the incoming radicals is high enough, they will be able to penetrate the inter-fiber space for even densely spaced VACNFs and precipitate at the sidewalls thus forming conical structures [101]. Also the degree of conicity can be modified by changing the gas mixture. Merkulov et al. [88] proposed the following model to explain the vertical alignment of PECVD grown CNFs. The growth of CNFs occurs via decomposition of the carbonaceous gas molecules at the catalyst particle surface or in the glow discharge, diffusion of the carbon atoms through the particle, and subsequent precipitation at the particle/fiber interface [63]. The axis of a CNF growing perpendicular to the substrate coincides with the direction of the applied electrostatic force, resulting in a uniform tensile stress across the entire nanofiber/catalyst particle interface, as shown in Figure 2.6(a) and (d) Consequently, carbon uniformly precipitates across the interface and the fiber continues to grow vertically (perpendicular to the substrate). However, if there were a spatial fluctuation in the C precipitation at the interface, CNF growth would deviate from vertical alignment, as shown in Figure 2.6(c) and (b). In the case of nanofibers growing from the tip (catalyst particle at the tip), the electrostatic force produces a compressive stress at the particle/nanofiber interface where the greater rate of growth is seen [Figure 2.6(c)]. Likewise, a tensile stress is applied to the particle/nanofiber interface where the lesser rate of growth is seen. These opposing stresses favor subsequent C precipitation at the interface experiencing tensile stress and the lesser rate of growth. The net result is stable, negative feedback that acts to equalize the growth rate around the entire periphery of the particle/nanofiber interface, and vertically aligned CNFs are grown. The presence of the preferred direction of C precipitation can be caused by stress-induced diffusion [105] due to the stress gradient in the catalyst particle and possibly by the variation in the stress-dependent sticking of diffusing C atoms to the C side of the Ni-C interface. Since the nanofiber base is attached to the substrate, the stress created at the particle/ nanofiber interface with the greater growth rate is tensile [Figure 2.6(d)] and acts to continue the increased growth rate, thus causing the CNF to bend even further. The inherent instability of positive feedback control systems leads to the wildly varying CNF orientation. Ngo et al. synthesized vertically aligned carbon nanofibers (VACNFs) using palladium as a catalyst by plasma enhanced chemical vapor deposition (PECVD) [106]. Figure 2.7 shows the TEM micrographs of vertically aligned carbon nanofiber grown on thick Pd catalyst. They observed that the thick Pd films lead to a variety of growth morphologies including a hybrid tip growth phenomenon, as well as small cone angles that are imparted by the elongation/wetting of the inner cavity of the CNFs by the Pd catalyst. Huang et al. reported growth of core/shell carbon nanofibers and nanotubes using metal sulfide (FeS, CoS and NiS) as a catalyst by arc discharge technique [107]. Figure 2.8 shows the schematic growth model of the core–shell carbon nanofibers and large-cavity carbon nanotubes. The catalyst seed of metal sulfide come from melted metal sulfides or from the combination reaction of evaporated metal and sulfur ions dissociated from metal sulfides. The linkage of sulfur with metal and carbon, as well as the low-temperature environment prevents the core materials from extruding out of the carbon nanofibers during growth. If the core/shell carbon nanofibers are not exposed to high temperature during and after their growth, the stuffed sulfides will remain in the carbon shell and contract and separate upon cooling.

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers 17

Fig. 2.6 Alignment mechanism of carbon nanofibers. If a CNF grows vertically (along the electric-field lines), electrostatic force F creates a uniform tensile stress across the entire catalyst particle/nanofiber interface, regardless of whether the particle is located at the tip (a) or at the base (b). If during the growth the CNF starts to bend due to spatial fluctuations in carbon precipitation at the particle/nanofiber interface, nonuniform stresses are created at the particle/nanofiber interface. For the nanoparticles at the tip (c) and at the base (d) the stresses are distributed in the opposite way, which leads to the nanofiber alignment in the first (c) but not in the second (d) case. White ellipses indicate the interface regions where the stresses occur [From Ref. 88].

When the temperature is high (> 1600 oC), such as in areas where the arc plasma reaches, the metal sulfide core materials released from the carbon shells and subsequently the carbon nanofibers will be annealed to become carbon nanotubes with large cavities, the main component of the deposit on the bottom of the bowl-like cathode. The intensive gas flow inside the cathode bowl and the gravity of the filled carbon nanofibers could be the force to bring the nanofibers into the plasma environment for the annealing process. On the other hand, in the plasma region, the catalyst seeds of metal sulfides could also lead directly to the growth of large cavity carbon nanotubes. The growth model is similar to that of metal catalyzed growth of carbon nanotubes and core/shell carbon nanofibers are no longer the intermediates of the carbon nanotubes.

18

Sk. F. Ahmed and K. K. Chattopadhyay

Figure 2.7. TEM micrographs of vertically aligned carbon nanofiber [From ref. 106].

Figure 2.8. Schematic growth mechanism proposed for the formation of the core/shell carbon nanofibers and carbon nanotubes [From ref 107].

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers 19 Carbon nanofibers have been synthesized by the thermal decomposition of acetylene with a copper nanocatalyst derived from cupric nitrate trihydrate at a low temperature of 260 oC [108]. Figure 2.9 shows the typical TEM image of as-prepared regularly helical nanofibers with a symmetric growth mode. The copper nanoparticles changed from initial irregular shapes to regular shapes during the growth of nanofibers.

Figure 2.9. TEM image of typical helical carbon nanofibers [From ref. 108].

Figure 2.10 shows the schematic diagram of growth mechanism of helical nanofibers. The mechanism of morphological changes of copper catalysts and the growth of carbon nanofibers are proposed in five steps: (1) the dehydration of cupric nitrate trihydrate, (2) the decomposition of cupric nitrate into Cu oxide, (3) the reduction of Cu oxide, (4) the formation of Cu nanoparticle before the growth of carbon nanofibers, and (5) the reaction with acetylene for the growth of carbon nanofibers.

Figure 2.10. The schematic diagram of growth mechanism of helical carbon nanofibers [From ref. 108].

20

Sk. F. Ahmed and K. K. Chattopadhyay

Hansen et al. reported that copper nanocrystals would undergo dynamic reversible shape changes in response to changes in the gaseous environment, and the shape changes were caused both by adsorbate-induced changes in surface energies and by changes in the interfacial energy [109]. The gas adsorption on the surface of Cu nanoparticles and the surface energy of different crystallographic planes of a single crystal were the main driving force for the gas-induced surface reconstruction and reshaping of the Cu nanoparticles. In a small metal particle, surface energies associated with different crystallographic planes are usually different. The catalyst particles undergo surface reconstruction to form geometrical shapes, which were able to promote the formation of carbon nanofibers with certain growth conditions of catalysts, gas, and temperature [110,111]. The shape changes of copper nanoparticles were induced by the adsorption of gases on the surfaces of particles. During the reaction, the active sites of the copper nanoparticles were changed from one place to another by following the surface reconstruction. The copper nanoparticle size has a considerable effect on the morphology of carbon nanofibers. The helical carbon nanofibers with a symmetric growth mode were grown on copper catalyst nanoparticles with a grain size less than 50 nm. When the catalyst particle size was around 50-200 nm, straight carbon nanofibers were obtained dominantly. It is reasonable to assume that it is possible to control the diameter of carbon nanofibers by controlling the size of the catalyst particle. A novel method for the direct growth of a single carbon nanofiber (CNF) onto the tip of a commercially available scanning probe microscope (SPM) using Ar+ ion irradiation was reported by Tanemura et al. [112]. This method was proposed on the basis of the experimental fact that the Ar+ ion bombardment of carbon coated substrates induced the formation of conical protrusions that possessed a single CNF at their tip. Commercially available Si SPM tips were coated with carbon and then were Ar+ ion bombarded at room temperature and at 200 oC. Figure 2.11(a) and (b) shows SEM image of a commercially available Si SPM chip of the tip region before and after sputtering respectively. The CNF thus grown was ~30 nm in diameter and 1.5 μm in length. The length was controlled between 0.5 and 1.5 μm by varying the sputtering duration.

Figure 2.11. SEM image of Si SPM chip at tip region (a) before and (b) after sputtering [From ref. 112].

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers 21 The formation of ion-induced CNFs explained in terms of erosion and/or growth processes during sputtering. In a case where CNFs are formed by the erosion process alone, so-called ‘‘seed’’ materials, which differ from the surface-constituent materials, are necessary for the CNF formation. These are known as “left-standing” protrusions [113]. The seed materials act as a protection against sputtering during the continuous erosion of the surrounding surface, thus yielding the protrusions tipped with the seed materials [Figure 2.12(b)]. The protrusions thus formed must be linear in shape and possess no conical base. TEM observation confirmed that there was no seed material on the CNF top and possessing CNF-tipped cone structure, which disagree with those of projections formed by the erosion process alone. This clearly suggests that the diffusion process plays a dominant role in the formation of CNFs. Based on the TEM observations, they proposed the following growth mechanism of ion-induced CNFs: (i) Formation of conical protrusions triggered by surface defects such as grain boundaries and small amounts of impurities, (ii) deposition of carbon atoms sputter-ejected from the surface onto the sidewall of the conical protrusions, and (iii) surface diffusion of the deposited carbon atoms toward the tips during sputtering. Since the diffusion is the thermal process, sputtering at elevated temperatures must enhance the diffusion of deposited carbon atoms, thus yielding the longer CNF on the tip of the conical protrusions. Van Vechten et al. demonstrated that carbon atoms readily migrated as far as ~20 mm on the surface during sputtering [114]. In addition to the thermal diffusion, the ionbombardment-enhanced diffusion, which is widely known to occur during sputtering, [115] is likely responsible for the CNF growth. One of the most successful approaches to obtain oriented arrays of nanotubes uses a nano channel alumina template for catalyst patterning [116]. First, aluminum is anodized on a substrate such as Si or quartz, which provides ordered vertical pores. Anodizing conditions are varied to tailor the pore diameter, height and spacing between pores. This is followed by electrochemical deposition of a cobalt catalyst at the bottom of the pores. The use of a template not only provides uniformity but also vertically oriented nanotubes.

Figure 2.12. Schematic representation of the possible formation mechanism of surface projections. (a) left-standing model based on ion etching process alone and (b) a growth model based on the diffusion processes. [From ref. 112].

22

Sk. F. Ahmed and K. K. Chattopadhyay

3. BASIC THEORY OF ELECTRON FIELD EMISSION Field emission (FE) is based on the physical phenomenon of quantum tunneling, in which electrons are injected from the surface of materials into vacuum under the influence of a strong external electric field [117]. The potential barrier is rectangular when no electric field is present, and becomes triangular when a negative potential is applied to the solid. The slope of the latter depends on the amplitude of the local electric field E just above the surface. This local electric field is drastically enhanced if the structure of the emitter is very sharp and protruding (high aspect ratio) as in the case of a CNT. Compared to thermionic electron emission and photo electron emission where electrons have sufficient energy to overcome the potential barrier (work function ϕ ), field emitted electrons tunnel into the vacuum because of a strongly deformed barrier under an electrical field (shown in Figure 3.1).

Figure 3.1. Potential-energy diagram for electrons at a metal surface under an applied electric field, where a strongly deformed potential results from the combination of the applied field and the image charges induced by the emitted electrons.

In the presence of high electric field, flat thin films of some materials emit electrons at macroscopic fields of about 1 to 10 V/μm, although cold field emission of electrons normally occurs only at fields of about 2 V/μm or above. I

This occurs because the thin film is an electrically nanostructured heterogeneous (ENH) material. II Internal nanostructure creates geometrical field enhancement at or near the film or vacuum interface, so local fields are much greater than macroscopic fields. III Electron emission at low macroscopic field strengths is a property of all ENH materials under appropriate conditions. IV Field enhancement is the primary effect, but there also be the secondary effects that contribute to facilitating emission at low fields.

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers 23 A simple mathematical expression for field emission can be obtained from the Heisenberg uncertainty principle: ∆p. ∆x ≅ ħ/2

(3.1)

where ∆p is the uncertainty of the electron momentum and ∆x is the corresponding uncertainty in position. Considering electrons near the Fermi level, the barrier height is the work function φ, so the uncertainty of the electron momentum ∆p = (2m φ) 1/2 and ∆x can be obtained from equation. (3.1) ∆x = h / 2 (2m φ) 1/2

(3.2)

Thus when ∆x is of the order of the barrier width, the probability that electrons would penetrate the barrier is high. The barrier width is x=φ/E

(3.3)

where E is the electrical field. Combining equations 3.2 and 3.3, we can obtain an estimate of the electrical field required for electron field emission: 3

E = 2( 2 m / h 2 )

1 2

ϕ2

(3.4)

e

A more detailed mathematical description has been done with the Wentzel- KramersBrillouin (WKB) method. The emission current density (J) can be obtained as a function of the electrical field, E, and work function, φ, which is described by Fowler-Nordheim (F-N) equation [118]: 3 ⎡ 2 ϕ AE 2 ⎢− Bα ( y ) exp J = ⎢ E ϕ .t 2 ( y ) ⎢⎣

⎤ ⎥ A / cm 2 ⎥ ⎥⎦

(3.5)

where A and B are constants with values of 1.54 x 10-6 and 6.87 x 107, respectively, and 1/ 2 α ( y ) = 0.95 − y 2 , where y = 3 .8 x10 − 4 E , Both t 2 ( y ) and α ( y ) are contributions

φ arising from the image potential, which is due to positive image charges induced by the emitted electrons at the surface. Image charges can cause a further lowering of the barrier 2 2 height by a factor of − e (Figure 3.1). Under typical conditions, t ( y ) and

α ( y ) are close

4x

to unity and normally omitted in practice. The standard physical assumptions of F-N theory are that the metal: (i) has a free-electron band structure; (ii) has electrons that are in thermodynamic equilibrium and obey Fermi-Dirac statistics; (iii) is at zero temperature; (iv) has a smooth flat surface; and (v) has a local work function that is uniform across the emitting

24

Sk. F. Ahmed and K. K. Chattopadhyay

surface and is independent of external field. It is also assumed that: (vi) there is a uniform electric field above the emitting surface; (vii) the exchange-and-correlation interaction between the emitted electron and the surface can be represented by a classical image potential; and (viii) barrier penetration coefficients may be evaluated using the JWKB approximation. In addition, the F-N equation is based on the assumption of a flat surface and only valid at 0 K, therefore, a modified F-N equation has to be used for an irregular surface and at different temperatures. In order to initiate field emission, an extremely large electrical field has to be employed, which is difficult to achieve from flat surface. However, with a tip structure, a high local electrical field can be obtained around high curvature regions. For instance, an electrical field E at a surface of a sphere, with a radius r and a potential V, is E = V/r. When r becomes smaller, E will become larger. Therefore, to account for the geometrical effects on the local electric field, a field enhancement factor β is introduced in the F-N equation as follows:

J = 1.54 x10 −6

3 ⎤ ⎡ 2 (βE ) 7 φ ⎥ ⎢ exp − 6.87 x10 A / cm 2 ⎥ ⎢ ϕ βE ⎥⎦ ⎢⎣ 2

(3.6)

An experimental F-N plot is modeled by the tangent, which can be written in the form [119-121]

ln(

I S )=a− 2 V V

(3.7)

where a is a constant and 3

S =−

6 . 83 x10 7 ϕ 2 d

β

(3.8)

A linear relationship can be obtained when plotting ln(I/V2) vs I / V, and if we know the work function (φ) of the material, specially for carbon nanotubes, we will assume a work function of 5 eV, [122,123] the inter-electrode distance (d), and the slope (S) of the F-N plot, the field enhancement factor β can be obtained. In geometrical configurations resembling a parallel plate capacitor, the macroscopic field EM is defined by: EM = V/d,

(3.9)

where V is the voltage applied across a gap of thickness d. The local field E is the field, close to the emitting surface (within 1-2 nm of the surface atoms), that determines the barrier through which field emitted electrons tunnel. The field E is some times called the barrier field. E is typically a few V/nm, and is often significantly higher than EM. Their ratio defines a field enhancement factor β

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers 25

β = E / EM

(3.10)

Considering the simple physical models of a ‘floating sphere at the emitter plane potential’ and a ‘hemisphere on a post’ the corresponding mathematical expressions have the form:

β=m+h/r

(3.11)

where r is the radius of the sphere or the hemisphere, h is its height above the emitter plane, and m is a constant generally taken as 0, 2 or 3 [124].

4. FIELD EMISSION FROM CARBON BASED MATERIALS Physically the length of CNT is several mm and diameters down to 10 Å, so the nanotubes exhibit a very high geometric aspect ratio (h/r), and has been shown that the applied electric field is concentrated precisely at the nanotube tips [125,126], which results in a large field enhancement factor β, typically ~10,000. In addition to the geometric field enhancement factor shown in equation (3.6), the effect of adsorbates i.e., forming a tunneling state at the surface [127] the tube cap and tip structures such as open or close [128,129] and the surface morphologies and contaminations i.e. catalyst particles, amorphous carbon [130] have been proposed as affecting their extraordinary field emission behavior. Carbon nanotubes have the right combination of properties such as nanometer size diameter, structural integrity, high electrical conductivity, and chemical stability that make good electron emitters. Electron field emission from carbon nanotubes was first demonstrated by Rinzler et al. in 1995 [131], and has since been studied intensively on various carbon nanotube materials. Field emission properties of different types of carbon nanotubes have been reported, including individual nanotubes [131-133], MWCNTs embedded in epoxy matrices [134], MWCNT films [135], SWCNTs [136], aligned MWCNT films [65,137] and hollow carbon nanotubes [132]. Random and aligned MWCNTs were found to have threshold fields slightly larger than that of the SWCNT films and are typically in the range of 3-5 V/µm for a 10 mA/cm2 current density [65]. These values for the threshold field are all significantly better than those from conventional field emitters such as the Mo and Si tips which have a threshold electric field of 50-100 V/µm. It is interesting to note that the aligned MWCNT films do not perform better than the random films. This is due to the electrical screening effect arising from closely packed nanotubes [138]. The case of emission from the large variety of carbon-based materials suggests that the NEA is not a prerequisite, and more general emission models are desired for these systems. Many models have been proposed to discuss the origin of electron emission from ta-C films and other amorphous carbon (a-C) films. Robertson suggested that field emission at a low electric field is due to the low electron affinity of ta-C films [139]. Other models such as space-charge-induced band bending [140], surface dipole-controlled emission [139], field enhancement due to the film microstructure [141] and conductive paths caused by localized sp2 sites were also proposed [142]. Thermal annealing is an effective method to alter structure of DLC films. Ilie et al. [143] and Carey et al. [144] proposed that the presence of sp2 clusters

26

Sk. F. Ahmed and K. K. Chattopadhyay

within the insulating sp3 matrix could give rise to field enhancement in amorphous carbon (aC) films containing large defect densities (>1019 cm−3). It was proposed that the presence of such dielectric inhomogeneity is responsible for field enhancement in these films. Since sp2 clusters will have different dielectric constants, the application of the external field will result in local field enhancements around the clusters and will aid in the emission of electrons. Groning et al. [142] explained the emission mechanism from DLC films in a way that, like a freestanding conductive tip in the vacuum, sp2 bonded carbon clusters are assumed to form a conductive channel in an insulating matrix, which leads to local field enhancement and hence to an enhanced electron emission. Since the sp2 clusters are located at or near the Fermi level and high concentration sp2 carbon clusters in the films play a more important role in determining the electron emission property of the films. The emission can depend on various parameters such as negative electron affinity [145], band gap [146], surface termination [145,147], depletion layers [148] and film thickness [149]. In these cases, the emission can be interpreted in terms of homogeneous films and a well-defined band structure [150]. In amorphous carbon, the sp3 content controls the band gap and electron affinity. Nanostructured carbon, nanocrystalline diamond, and carbon nanotubes are the types of carbon that emit at lowest applied field. In microcrystalline diamond, emission is found to occur from grain boundaries [151,152], that is, nm-scale sp2-bonded regions of positive electron affinity. Similarly, emission from carbon nanotubes [153] occurs from 1 nm curved regions. CNFs films were synthesized by plasma enhanced chemical vapor deposition about 100 nm in diameter and about 10 μm in length using P doped n-type Si (100) wafers and indium tin oxide (ITO) coated glasses as substrates [154]. Figure 4.1(a) shows the SEM image of the CNF film on Si substrate and (b) shows the TEM image of a tip of a CNF where as (c) shows a schematic illustration of the CNF structure. The CNFs ranged 50-100 nm in diameter and over 10 μm in length, which were randomly oriented to the substrate showed electron field emission characteristic (as shown in Figure 4.2(d)). Since the nanofibers were grown to random orientation, electrons can be emitted with any directions from the protrusions on the fiber. The threshold electric field (Eth) was estimated 2.4 V/μm, which is comparable with the threshold field of several carbon nanostructures such as, Eth of single wall carbon nanotube (SWCNT) by arc discharge about 2.2 V/μm, multi wall carbon nanotube (MWCNT) by microwave plasma CVD about 1.8 V/μm and nanostructured carbon film about 3.0 V/μm respectively [155-157]. The reason of the excellent field emission characteristics is due to the CNF film has many protrusions which are 10 nm in width and 30 nm in length, as shown in Figure 4.1(b). Since the nanofibers were grown to random orientation, electrons can be emitted with any directions from the protrusions on the fiber. Aligned carbon nanofibers and hollow carbon nanofibers were grown by micro wave ECR-CVD method using methane and argon mixture gas at a temperature of 550 oC showed good electron field emission (Figure 4.2) [158]. The aligned carbon nanofibers give a high current density 7.25 mA/cm2 at 12.5 V/μm in comparison with the value 0.69 mA/cm2 at 12.5 V/μm of the hollow carbon nanofibers.

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers 27

Figure 4.1.(a) SEM image of the CNF film on Si substrate, (b) TEM image of a tip of a CNF, (c) schematic illustration of the CNF structure and (d) J-E curve and in inset the corresponding F-N plot [From ref. 154].

28

Sk. F. Ahmed and K. K. Chattopadhyay

Figure 4.2. (a) I-V graph of the carbon nanofiber (b) and the hollow carbon nanofiber [From ref. 158].

Effect of catalysts on the field emission property was studied by Kamada et al. [78]. Carbon nanofiber films on Pd-Se, Fe-Ni, and Ni-Cu alloy catalysts have been synthesized at low temperatures by a thermal CVD technique. Figure 4.3(a) shows the field electron emission characteristics of the CNF films grown at 600 oC. The threshold electric fields of the CNFPd-Se, CNFNi-Cu, and CNFFe-Ni films are estimated to be 1.1, 2.8, and 3.8 V/μm, respectively. The CNFs grown using Pd-Se catalyst were found to have more defective structure than that obtained with the other catalysts, and exhibited best field emission property. It is likely that defects play a role as electron emission sites. Figure 4.3(b) shows the threshold electric field obtained for the CNFPd-Se, CNFFe-Ni, and CNFNi-Cu films as a function of growth temperature. The threshold electric field was not strongly dependent on the catalyst type and decreased with growth temperature. Ilie et al. [159] reported that the surface electronic properties introduced by defects could provide a local field enhancement to facilitate the field emission. From this, it is suggested that the excellent field emission property obtained for CNFPd-Se originates from numerous defects in the body of the CNFs. In this sense, good crystallinity of the CNFs is not required to obtain good field emission characteristics. Carbon nanofibers (CNFs) were grown on a Ni-P alloy catalyst deposited on a silicon substrate via MWCVD technique with methane gas [82]. The CNFs grown on the Ni-P alloy catalyst showed random orientation and it composed of parallel graphite planes with defects tilted from their axis. Figure 4.4 shows the electron emission current density versus electric field (J-E) curves of CNFs with different thickness of the catalyst.

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers 29

Figure 4.3. (a) Emission current density as a function of applied electric field and (b) threshold electric field for the CNFPd-Se, CNFFe-Ni, and CNFNi-Cu films as a function of growth temperature [From ref. 78].

For CNF grown on catalyst with thickness 20 nm, the turn-on field was approximately 0.11 V/μm with an emission current density of 10 mA/cm2 and the threshold field was 3.1 V/μm with an emission current density of 10 mA/cm2. CNF grown on catalyst thickness 30 and 40 nm, have almost the same turn-on field approximately 0.22 V/μm, but the threshold field is 3.4 and 4.1 V/μm, respectively.

30

Sk. F. Ahmed and K. K. Chattopadhyay

Figure 4.4. The current density vs. electric field curves for CNFs deposited with different thickness of catalyst and in Inset corresponding F-N plot [From ref 82].

The excellent field emission properties of Ni-P alloy catalyzed-CNFs may be attributed to the random and defects of CNFs. Davydov et al. [160] have also pointed out that perfectly aligned CNTs were less efficient field emitters and had lower field enhancement than chaotic CNTs. Some reports relating defect densities to field emission properties have also been proposed [161,162]. The enhanced emission may originate from the defect-induced energy bands that are formed within the band gap of graphite. The energy barrier that the electrons must tunnel through to be emitted is reduced, so the electrons residing at these defect levels can be emitted directly into vacuum from these bands or be transported to the surface states for emission [163]. Obraztsov et al. [164] have also found that the field emission properties were improved by increasing the density of structural defects. Figure 4.4 also indicates that the field emission properties of CNFs with small diameter are better than those of the CNFs with large diameter. A clear fluctuation of the I-V curve at higher voltages for CNFs is seen in Figure 4.4, indicating some emission sites are damaged or destroyed. It is reasonable to suggest that the CNFs with small diameter are more easily damaged by ion bombardment than the CNFs with large diameter [165]. The CNF is synthesized on the iron-evaporated Si substrate by microwave plasma chemical vapor deposition and nitrogen (N2) plasma treatment is carried out to modify the CNF surface [166]. A reduction in the turn-on electric field is achieved by N2 plasma treatment for the CNF and the stability of the electron emission current is also improved by nitridation of the CNF surface. In Figure 4.5 the field emission characteristics are compared between as grown and N2 plasma-treated samples. Potential barrier height is reduced by nitridation of the CNF surface. The excellent field emission of carbon nanofibers and nanotubes have stimulated their applications as electron sources in x-ray applications [137], and field emitters in electron microscopes [132], with their environmental sensitivities of electrical conduction.

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers 31

Figure 4.5. Field emission characteristics (a) untreated and (b) N2 plasma treated CNFs [From ref. 166].

5. SYNTHESIS AND FIELD EMISSION PROPERTY OF DIFFERENT CARBON NANOSTRUCTURE 5.1. Synthesis and Field Emission Property of Carbon Fibrous Films Among the different techniques for the production of carbon nanotubes, plasma enhanced chemical vapor deposition is a high yield and controllable method for the production carbon nanotubes/nanofibers [167] with mass production. Thin film catalyst layers have been successfully employed in the carbon nanofiber growth. The use of metal catalysts such as Ni, Fe, Co, Pt and some of their alloys has been explored in an effort to control the size and morphology of carbon nanostructures formed through the decomposition of hydrocarbons. We have used Ni as a catalyst for the formation of carbon fibers, because carbon nanostructure formed on Ni are more crystalline than those formed with other catalysts [168,169]. The target used for sputtering was a Ni plate of thickness ~1 mm with a diameter 2.5 cm (with purity 99.99 %, Aldrich). The Ni target has been sputtered on Si substrates via dc sputtering technique to produce thin film of Ni catalyst. The substrates were 10×10 mm2 cleaned Si (400) wafer. The Si substrates were etched in HF (~20%) for 5 minutes to remove the surface oxide layer and finally cleaned in an ultrasonic cleaner. The sputtering was done at a pressure 0.2 mbar sending argon as a sputtering gas with an inter-electrode distance 1.6 cm at room temperature for deposition time 5 minutes, which yielded a Ni film with a thickness ~10 nm, as measured by quartz crystal thickness monitor. For sputtering, we

32

Sk. F. Ahmed and K. K. Chattopadhyay

maintained high voltage 2.5 KV and corresponding current density was 19.5 mA cm-2. Table 5.1 shows the deposition parameters for Ni catalyst thin films. After deposition of the catalyzed film, the sample was immediately transferred in to the CVD chamber where nanofibers growth has been performed. Deposition chamber used for the synthesis of carbon fibers was made of stainless steel (SS). The plasma was produced between two parallel plates SS electrodes as usual. The lower plate was grounded and the upper plate was used as the cathode. The schematic diagram of the DC-PECVD unit is shown in Figure 5.1. The deposition chamber was initially evacuated by a standard rotary and a diffusion pump arrangement up to a base pressure of 10−6 mbar. The substrate (Si) was placed on a molybdenum substrate holder, which could be directly heated. When the chamber pressure attained 10−6 mbar, the Mo substrate holder was started to heat by sending current through it. The substrate temperature could be varied by varying the current through the Mo substrate holder, which was connected to the secondary of a step down transformer. The temperature of the substrate was measured by a disappearing filament type pyrometer (PYROPTO, IT65). Acetylene (C2H2) gas was used in PECVD process as a precursor of carbon. Acetylene (C2H2) gas was allowed to flow maintaining the CVD chamber pressure 50 mbar. Deposition was performed at 2.0 kV DC supply with corresponding current density 25 mA cm-2 for 30 min. Substrate temperature was varied from 700 to 850 oC for different set of experiment. The deposition parameters for the synthesis of carbon fibrous thin films via PECVD have been shown in Table - 5.2.

Figure 5.1. Schematic diagram of the DC -PECVD unit.

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers 33 Table 5.1. Deposition parameters used for dc-sputtered deposited Ni catalyst Deposition parameters Deposition time dc voltage Electrode distance Sputtering gas Gas pressure Substrates used Substrate temperature

Corresponding values 5 min. 2.5 kV 1.6 cm Argon (Ar) 0.2 mbar Si (400) 300 oK

Table - 5.2. Deposition parameters used for the synthesis of carbon fibrous thin films Deposition parameters 1. Deposition time 2. dc voltage 3. Electrode distance 4. Precursor material Gas pressure Substrates used Substrate temperature

Corresponding values 30 min. 2.0 kV 1.4 cm Acetylene (C2H2) 50 mbar Ni catalyzed Si (400) 700 - 850 oC

Figure 5.2 showed the SEM micrographs of the deposited films, which showed the existence of carbon fibers in the films. The morphologies of the films have been changed with the change of substrate temperature. At 700 oC substrate temperature, only particles have been found but at 750 oC, some carbon nanofibers have also been grown. At 800 oC substrate temperature, carbon fibers have been grown with length ~ 1000 nm and the corresponding diameter ~ 400 nm. Finally at 850 oC substrate temperature, the best quality carbon fibers have been grown with length ~ 2000 nm with corresponding diameter ~ 400 nm. It is clear from these studies of substrate temperature variations that at lower substrate temperature only particles are grown and at higher substrate temperature the morphology changes from particles to nanotubes or fibers like structure i.e. quasi one dimensional growth takes place at higher substrate temperature. The electron field emission properties of the CNFs deposited on Si substrates have been studied by our high vacuum (~10-7 mbar) field emission setup as shown in Figure 5.3. Field emission measurements were carried out by using a diode configuration consisting of a cathode (the film under test) and a stainless steel tip anode mounted in a liquid nitrogen trapped rotary-diffusion vacuum chamber with appropriate chamber baking arrangement. The measurements were performed at a base pressure of ~5 x 10-7 mbar and at different temperature, which was controlled with a controller and measured with a thermocouple. The tip-sample distance was continuously adjustable to a few hundred μm by spherometric arrangement with screw-pitch of 10 μm. The anode-sample spacing was set at a particular value by rotating the micrometer screw which served as an anode electrode. Field emission current-voltage measurements were done with the help of an Agilent multimeter (model 3440-1A). Emission characteristics were registered and analyzed with the help of a personal computer.

34

Sk. F. Ahmed and K. K. Chattopadhyay

Figure 5.2. SEM image for different substrate temperature (a) 750 oC, oval shaped particles, (b) 800 oC, fiber like and (c) 850 oC; fibers [From ref. 168].

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers 35

Figure 5.3. Schematic diagram of the field-emission apparatus [From ref. 170].

We have used the simplified F-N equation for field emission analysis. Figure 5.4(a) shows the emission current (I) vs. macroscopic field (E) curves of carbon nanofibers thin film deposited on Si substrate for anode-sample separation (d) of 60 μm. The macroscopic field is calculated from the external voltage applied (V), divided by the anode-sample spacing (d). Theoretically, the emission current I is related to the macroscopic electric field E by

− b vF φ I = A a t φ ( β E ) exp{ βE −2 F

−1

2

3

2

}

(5.1)

where, φ is the local work-function, β is the field enhancement factor, A is the effective emission area, a is the first Fowler-Nordheim Constant (1.541434 x 10-6 A eV V-2), b is the second F-N Constant (6.830890 x 109 eV-3/2 V m-1), and vF and tF are the values of the special field emission elliptic functions [119] v and t, evaluated for a barrier height φ. In so-called Fowler-Nordheim coordinates, this equation takes the form: 3

(v bφ 2 β −1 ) I ln{ 2 } = ln{t F− 2 A a φ −1 β 2 } − F E E

(5.2)

36

Sk. F. Ahmed and K. K. Chattopadhyay

An experimental F-N plot is modeled by the tangent, which can be written in the form [119-121]: 3

I ( s bφ 2 β −1 ) ln{ 2 } = ln{rA a φ −1 β 2 } − E E

(5.3)

where r and s are appropriate values of the intercept and slope correction factors, respectively. Typically, s is of the order of unity, but r may be of order 100 or greater. Both r and s are relatively slowly varying functions of 1/E, so a F-N plot (plotted as a function of 1/E) is expected to be a good straight line. The F-N plot of our sample is shown in Figure 5.4(b). It has been observed that the I-E curve in the present work is closely fitted with straight line. This suggests that the electrons are emitted by cold field emission process. The turn-on field, which we define as the macroscopic field needed to get an emission current I = 0.09 μA, (which corresponds to an estimated macroscopic current density, Jest = 14.5 μA/cm2, where Jest = I/A, A = anode-tip area) were lying in the range 2.57 to 9.71 V/μm for variation substrate temperature films. This value is quite lower than that of nanocrystalline carbon 6.4 V/μm [171] and carbon nanofiber arrays (~3 V/μm) reported by Cao et al. [172]. According to the F-N plot (Figure 5.4(b)), the slope m (given by equation (5.4)) would represent the combined effect of work function and enhancement of local electric field and is given by,

m= −

bφ

3

2

(5.4)

β

The effective work function relation [173]

φE =φ / β

2

3

φ E is related with the true work function φ through the

. Using φ = 5 eV is the work function of CNF [174]; the field

enhancement factor was calculated from the slope of the F-N plot, lies in the range 8090 to 1945 and the corresponding effective work function φC lies in the range 0.0124 to 0.0321 eV for films deposited with different substrate temperature, which is comparable with CNT films [173]. The plots of I-E graph for different electrode distance (d) are shown in Figure 5.5(a) and (b) the corresponding F-N plot. The turn-on field was found to vary in the range 6.87 to 2.87 V/μm for a variation of anode sample spacing 80 - 120 μm for the carbon fibrous film deposited at 850 oC. In the I-E graph (Figure 5.5(a)), we observed a parallel shift of curves with respect to anode-sample separation (d) i.e., for a particular electric field the current density increases with increasing the anode-sample separation. Zhou et al. [175] reported similar type of observation for their β-SiC nanorods. We suppose that this type of shift observed in our sample may be due to the change in the effective emission area of the sample for different anode-sample separation. The change of effective emission area with respect to d may be related to the geometry of the anode.

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers 37

Figure 5.4. Emission current vs. macroscopic field curves of carbon fibrous film deposited for different substrate temperature (a) and corresponding. F-N plot (b) [From ref. 168].

In our experiment we have used a conical shape anode with tip diameter 1 mm, therefore the lines of force immerging from the edge of the anode tip and terminating to the sample surface are diverging in nature, whereas the lines of force immerging from the flat surface of the tip are parallel in nature. Hence, the effective emission area of the sample increases with increasing d as shown in Figure 5.6. Au et al. [176] performed field emission of silicon nanowires using a spherical-shaped stainless steel probe with a tip diameter of 1 mm as an anode. They also found a parallel shift in their I-V curve.

38

Sk. F. Ahmed and K. K. Chattopadhyay

Figure 5.5. I-E curves carbon fibrous film for different anode-sample separation (a) and (b) corresponding F-N plot (b) [From ref. 168].

Okano et al. [177] reported that their macroscopic current density for diamond films was independent of the anode-sample separation. Their field emission apparatus consisted of a parallel plate arrangement of the anode and sample, separated by spacers. So the electric lines of force between the anode and the sample were parallel in nature, hence effective emission area remained independent of the anode-sample spacing.

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers 39

Figure 5.6. Schematic diagram for the dependence of effective emission area as an increasing function of anode-sample separation.

5.2. Synthesis and Field Emission Property of Vertically Aligned CNFS Vertically aligned carbon nanofibers (VACNFs) have been synthesized by direct current PECVD technique and for the synthesis of VACNFs, Ni catalyst was deposited in thin film form on Si substrates via RF magnetron sputtering technique. For sputtering, we maintained RF power 200 watt and corresponding chamber pressure 0.1 mbar. We have synthesized Ni thin film having different thickness varying from 10-20 nm, as measured by quartz crystal thickness monitor (HindHivac, Digital thickness monitor, Model: DTM-101). The deposition parameters for Ni catalyst have been shown in Table- 5.3. After deposition of the Ni film, the substrates were immediately transferred into the CVD chamber where nanotube growth has been performed. Acetylene (C2H2) gas was used in PECVD process as a precursor of carbon and during deposition the chamber pressure was maintained at 30 mbar. Deposition was performed at 2.2 kV DC with corresponding current density 21.5 mA cm-2 for 25 min. Table – 5.3. Deposition parameters used for rf-magnetron sputtered deposited Ni catalyst Deposition parameters RF-power Electrode distance Sputtering gas Gas pressure Substrates used Substrate temperature Deposition time

Corresponding values 200 Watt 3 cm Argon (Ar) 0.2 mbar Si (400) 300 oK 4 -10 min.

40

Sk. F. Ahmed and K. K. Chattopadhyay

An atomic force microscope (AFM-NT-MDT, Solver Pro) was used to analyze the surface topography of the grown CNFs films. Figure 5.7 shows the AFM images of the carbon nanofibrous films deposited on Ni catalyst having different thicknesses. From the figure it is clear that the diameter of the CNFs increase and length decreases with the increase of Ni catalyst film thickness. The diameter and lengths of the CNFs deposited on thinner catalyst (Ni film thickness 10 nm) are ~ 150 nm and 2.5 μm whereas the diameter and length of the CNFs deposited on thicker catalyst (Ni film thickness 20 nm) are ~ 250 nm and 1.0 μm. The morphology of the catalyst film is known to play a critical role in CNF growth. So, the thickness of the Ni catalyst film will affect the growth and the properties of CNFs. The diameter of the CNFs decreased as the thickness of the catalyst film decreased.

Figure 5.7. AFM 3D pictures of VACNFs deposited on different thickness of Ni catalyst (a) for 10 nm, (b) for 15 nm and (c) for 20 nm [From ref. 178].

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers 41 Figure 5.8(a) shows the emission current density (J) vs. macroscopic field (E) curves for the VACNF films deposited at a fixed anode-sample separation (d) of 120 μm. Field emission characteristics of the films were analyzed using the help of simplified Fowler-Nordheim (FN) theory [119-121]. The F-N plots of our sample are shown in Figure 5.8(b).

Figure 5.8. J-E graph of VACNFs for different aspect ratios (a) and corresponding F-N plot (b) [From ref. 178].

42

Sk. F. Ahmed and K. K. Chattopadhyay

It has been observed that all the J-E curves in the present work are satisfactorily fitted with a straight line, which suggests that the electrons are emitted by field emission process. The threshold field, which we define as the macroscopic field needed to get an emission current density J = 10 μA/cm2, were lying in the range 4.3 to 5.4 V/μm for VACNFs deposited on Ni catalyst having different thicknesses. As our deposited CNFs have different aspect ratios so the local field enhancement occurs and changes in the threshold field was observed for different type of VACNFs. The field enhancement factor (β) as well as emission current density is strongly dependent on the aspect ratio of canon nanofibers (shown in Figure 5.9). Tsai et al. [82] showed that the CNFs with small diameter and many defects exhibited excellent field emission properties than the CNFs with larger diameter.

Figure 5.9. The variation of threshold field and emission current density with aspect ratio carbon nanofiber [From ref. 178].

5.3. Synthesis and Field Emission Property of Multiwalled Carbon Nanotubes For the deposition of Ni catalyst in thin film form by dc sputtering process we have used a Ni plate of thickness ~1 mm with a diameter 2.5 cm (with purity 99.99 %, Aldrich). The substrates were 10×10 mm2 cleaned Si (400) wafer. The Si substrates were etched in HF (~20%) for 5 minutes to remove the surface oxide layer and finally cleaned in an ultrasonic cleaner. The sputtering was done at a pressure 0.1 mbar sending argon as a sputtering gas with an inter-electrode distance 1.6 cm at room temperature for deposition time 5 minutes, which yielded a Ni film with a thickness ~ 12 nm, as measured by quartz crystal thickness monitor. For sputtering, we maintained high voltage 3.0 KV and corresponding current

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers 43 density was 27.5 mA cm-2. After deposition of the catalyzed film, the sample was immediately transferred to the CVD chamber where the synthesis of nanotubes was performed. The deposition procedure was described in section 5.1. Table 5.4 shows the deposition parameters for the synthesis of MWCNTs thin films via DC-PECVD technique. Table 5.4. Deposition parameters used for the synthesis of MWCNTs thin films Deposition parameters 1. Deposition time 2. dc voltage 3. Electrode distance 4. Precursor material 5. Gas pressure 6. Substrates used 7. Substrate temperature

Corresponding values 30 min. 2.0 kV 1.4 cm Acetylene (C2H2) 30 mbar Ni catalyzed Si (400) 900 0K

Figure 5.10(a) shows the FESEM micrographs of the deposited films, which showed the existence of carbon nanotubes in the films. The diameter of the carbon nanotubes are ~ 12 nm and few micrometer in length. The transmission electron micrographs of multiwalled carbon nanotubes have been shown in Figure 5.10(b). It could be observed that the carbon nanotubes are multiwalled with diameter ~12 nm.

Figure 5.10. FESEM micrograph (a) and HRTEM lattice image of the MWCNT (b).

44

Sk. F. Ahmed and K. K. Chattopadhyay

Figure 5.11(a) shows the emission current density (J) vs. macroscopic field (E) curves for MWCNT films for different anode-sample separation (d). The F-N plot of our sample is shown in Figure 5.11(b). The threshold fields were found to vary in the range 4.7 - 3.6 V/μm for a variation of anode-sample separation in the range 80 - 150 μm.

Figure 5.11. (a) J-E curves for the MWCNTs for different anode-sample separation (d) and (b) corresponding F-N plots.

In the J-E graph (Figure 5.11(a)), we observed a parallel shift of curves with respect to anode-sample separation (d) i.e., for a particular electric field the current density increases with increasing the anode-sample separation. For example, at a field of 5 V/μm, the J values were found to be 0.2 mA/cm2 (for d = 80 μm), 1.2 mA/cm2 (for d = 110 μm) and 3.7 mA/cm2 (for d = 150 μm). This type of shift observed in our sample is due to the change in the effective emission area of the sample for different anode-sample separation, which is described in section 5.1.

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers 45

6. EFFECT OF TEMPERATURE ON THE ELECTRON FIELD EMISSION FROM VERTICALLY ALIGNED CARBON NANOFIBERS AND MWCNTS The synthesis of vertically aligned carbon nanofiber (VACNF) thin films has been described in previous section 5.2 and for synthesis of MWCNTs, Ni catalyst deposited in thin film form on Si substrates via RF sputtering technique. The target used for sputtering was a Ni plate of thickness ~1 mm with a diameter 2.5 cm (with purity 99.99 %, Aldrich). For sputtering, we maintained RF power 180 watt and corresponding chamber pressure 0.2 mbar. After deposition of the Ni film, the substrates were immediately transferred into the CVD chamber where nanotube growth has been performed. Acetylene (C2H2) gas was used in PECVD process as a precursor of carbon. C2H2 gas was allowed to flow maintaining the CVD chamber pressure 40 mbar. Deposition was performed at 2.5 kV DC supply with corresponding current density 27.5 mA cm-2 for 30 min at 700 oC substrate temperature. AFM image (Figure 6.1(a)) of the carbon nanofibrous films shows that the vertically aligned CNFs are having an average diameter ~ 250 nm and length 2.0 μm. Figure 6.1(b) shows the FESEM micrographs of the carbon nanotubes, which showed that the diameter of the carbon nanotubes are ~ 20 nm and few micrometer in length. The transmission electron micrographs of multiwalled carbon nanotubes have been shown in inset of Figure 6.1(b). It could be observed that the carbon nanotubes are multiwalled with diameter ~ 20 nm.

Figure 6.1.(a) AFM 3D pictures of vertically aligned carbon nanofiber thin films, (b) FESEM micrograph and in inset HRTEM micrograph of the MWCNT [From ref. 179].

46

Sk. F. Ahmed and K. K. Chattopadhyay

For temperature dependent field emission total current density (J = JE + JT, where JE and JT are the field current and thermionic current density respectively) given by simplified F–N equation and Richardson equation as [180,181]:

J = J E+ J T J =a φ

−1

(6.1)

− sb φ ( β E ) exp( β E 2

3

2

φ ⎡ θ ⎤ 2 − KT + ADT e )⎢ ⎥ ⎣ sin( θ ) ⎦

(6.2)

where A is a constant about 120 A/(cmK)2, D is the average transmission coefficient of emitter surface, T is the temperature in Kelvin, φ is the work function of CNF/CNT, k is the Boltzmann constant and θ is the temperature correction factor, and is given by

2.2π (kT / q)φ θ≈ 1.959 E

1

2

(6.3)

For CNF/CNT with a work function 5 eV [174,182] and temperature below 1000 K, the value of [θ/(sin(θ))] in eqn. (6.2) is always 1.0 and in our studied temperature range, (< 400 o C), the highest contribution of thermionic emission is much smaller than the field emission current density i.e., the measured emission property is predominated by field emission current because below 1000 K, the thermionic emission effect is less significant than the field emission effect [183]. Hence eqn. (6.2) reduced as 3

( s bφ 2 β −1 ) J ln{ 2 } = ln{r a φ −1 β 2 } − E E

(6.4)

Figure 6.2(a) and 6.3(a) shows the emission current density (J) vs. macroscopic field (E) curves of carbon nanofibers and MWCNTs thin films respectively, for different temperature and corresponding F-N plot is shown in Figure 6.2(b) and 6.3(b). It has been observed that the J-E curve in the present work is closely fitted with straight line. This suggests that the electrons are emitted by cold field emission process. The threshold field, which we define as the macroscopic field needed to get an emission current density J = 10 μA/cm2, were lying in the range 5.1 to 2.6 V/μm for CNFs and lying in the range 4.0 to 1.4 V/μm for MWCNTs for the variation of temperature from 300 K to 650 K.

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers 47

Figure 6.2. J-E graph of CNFs for different temperature (a) and corresponding F-N plot of CNFs (b) [From ref. 179].

48

Sk. F. Ahmed and K. K. Chattopadhyay

Figure 6.3. J-E graph of MWCNTs for different temperature (a) and corresponding F-N plot of MWCNTs (b) [From ref. 179].

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers 49 It is clear from the Figure 6.4(a) and (b) that with the increase of temperature, threshold field decreases and current density increases for CNFs as well as MWCNTs. The emission current density is strongly dependent on the work function and as well as on the aspect ratio. For an example to get emission current density J = 5 mA/cm2, 9.5 V/μm field needed for CNFs but for MWCNTs 4.5 V/μm field needed for the same current density at 650 K ambient temperature. The work function of materials is temperature dependent. Therefore, the decrease of threshold field with the increase of temperature may be due to the decrease of work function of CNF and MWCNT films.

Figure 6.4. The variation of threshold field and emission current density with temperature (a) for CNFs (b) for MWCNTs [From ref. 179].

50

Sk. F. Ahmed and K. K. Chattopadhyay

Figure 6.5 shows the variation of the effective work function with temperature for CNFs and MWCNTs. It shows that the effective work function decreases with increase of temperature. The variation of the effective work function with temperature is consistent with the variation of the emission current density as observed. From the quantum mechanical tunneling phenomena, we know that the Fermi energy determines the field emission current. The work function (φ) is given by φ = EV - EF, where EV is the fixed vacuum level and EF is the Fermi level [184]. For low temperature emission, Fermi level is lower and the electrons have to transmit through a much boarder barrier as shown in Figure 6.6(a) and for high temperature field emission, Fermi level increase so barrier width decrease as shown in Figure 6.6(b). Hence the emission current increases under same field for high temperature field emission (as shown in Figure 6.4).

Figure 6.5. Variation of the effective work function with temperature for CNFs and MWCNTs [From ref. 179].

Figure 6.6. Schematic diagram of the transmission of electrons from the MWCNTs at low (a) and high (b) temperatures under applied field [From ref. 170].

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers 51 As the field emission characteristic has a complicated dependence on electric field and temperature when these two factors coexist, and cannot be explained simply by F–N or thermionic emission model, the information of β can still provide fruitful information. According to the F-N plot (Figure 6.2(b) and 6.3(b)), the slope m (given by equation (6.5)) would represent the combined effect of work function and enhancement of local electric field and is given by,

m= −

bφ

3

2

(6.5)

β

The effective work function relation [183]

φC = φ / β

2

3

φC is related with the true work function φ through the

. Using φ = 5 eV is the work function of CNF/CNT, the field

enhancement factor was calculated from the slope of the F-N plot, lies in the range 4589 to 9917 and the corresponding effective work function φC lies in the range 0.018 to 0.011 eV for the CNFs with different ambient temperature. For MWCNTs field enhancement factor lies in the range 5838 to 11448 and the corresponding effective work function φC lies in the range 0.015 to 0.010 eV. The field enhancement factor β increases monotonously with the temperature, which explains very well the increase in emission current density with measuring temperature (shown in Figure 6.4). But physically the field enhancement factor should depend on the geometric shape of the emitter rather than temperature. Hence there may be other factors responsible for such temperature dependence of emission current. Although the exact explanation of the observed temperature dependence of the emission current needs more research, the following effects may have strong influence. The presence of defects or surface states is predominant in nanomaterials like CNT or CNF. Wang et al. and Xu et al. also proposed that the field emission property is related with the defect densities [161,162]. These states might have small activation energies and when the temperature in increased, the carriers trapped in these states are activated into the conduction band and more emission currents results. Chen et al. [185] observed the field emission of different oriented CNTs and they discovered that the CNTs oriented parallel to the substrate have a lower onset applied field than those oriented perpendicular to the substrate. They also suggested that the defect emission mechanism is a reason for the low onset electrical field. Obraztsov et al. [164] have also found that the field emission properties were improved by increasing the density of structural defects. The field enhancement factor also depends on aspect ratio of carbon nanostructure. The aspect ratio (h/r, where r is the average radius and h is the length of the tubes, respectively) of our MWCNTs is greater than that of CNFs. From our experimental result we also see that the field emission properties of CNTs are better than those of the CNFs with large diameter. Another possible reason is that the screening effect, which diminishes the electric field near the CNFs. Nilson et al. proposed from their experimental observation that to overcome screening effect the distance between nanotubes will be 1-2 times the tube height [186]. As our deposited vertically aligned CNFs are compact so screening effects is much remarkable. Since it is well known that MWCNTs are mostly

52

Sk. F. Ahmed and K. K. Chattopadhyay

conductive, their conductivity should decrease with higher temperature and thus causes the screen effect less remarkable.

7. APPLICATION OF CARBON NANOFIBER AND CARBON NANOTUBE Along with the improvement of the production and characterization techniques for nanotubes, progress is being made in their applications. MWCNTs have also exhibited ballistic transport [187]. The reasons for high electron conductance in carbon nanotubes are as follows; i) physical perfection: a smooth surface with no chemical dangling bonds and no edges reducing surface states, which affects backscattering; ii) strong covalent bonds in CNT and iii) no low-energy dislocations or defects, which also reduces backscattering and provide stability for high current transport. Carbon nanofibers/nanotubes have many properties from their unique dimensions to an unusual current conduction mechanism that make them ideal components of electrical circuits. Due to their semi conducting properties, nanofibers/nanotubes may be the building blocks for smaller, faster computers. Other potential applications in electronics and computers include, storage devices, It was proposed to use nanotubes as central elements of electronic devices including field-effect transistors, single-electron transistors [188] and rectifying diodes [189] and for logic circuits [134]. The geometric properties of nanotubes such as the high aspect ratio and small tip radius of curvature, coupled with the extraordinary mechanical strength and chemical stability, make them an ideal candidate for electron field emitters [129,186]. CNT field emitters have several industrial and research applications; flat panel displays [131], outdoor displays, traffic signals and electron microscopy. De Heer et al. [135] demonstrated the earliest high intensity electron gun based on field emission from a film of nanotubes. The properties of carbon nanotubes (CNTs) and the less crystalline carbon nanofibers (CNFs) have attracted considerable interest for both scientific and technological issues [17,190]. Their impressive mechanical properties [191], high current carrying ability [192], and field emission performance [135] have opened the way to a number of applications such as field emission devices [193], interconnects [194], sensors [195], super-capacitors [196], fuel cells [197] and battery electrodes [198]. The vertical geometry of carbon nanofibers (CNFs) is particularly useful in technologies such as nanoelectronics [89], electrodes for biosensing/stimulation [90], nanomechanical [91], and thermal interface materials [92]. Achieving near-ohmic contact at the nanotube-metal interface as well as investigating the affect of nanotube crystallinity is critical for evaluating and modeling the electrical performance of on-chip interconnects. Sim et al. reported that the carbon-nanofiber-based (CNF) ionization gas sensing devices on plastic substrates [199]. The device is configured as diode structure with a Cu plate and a CNF film as anode and cathode respectively. For a fixed applied voltage of 600 V, the ionization current of that device exhibits two regions of linearity with respect to gas pressure below and above 5 Pa, suggesting that the device can be employed as vacuum ion gauge. Ngo et al. reported that due to thermal conductance properties CNF-Cu composite material could be use as a thermal interface material in both IC packaging and equipment cooling applications [200]. The concept of nanothermometer using CNT was first proposed by Gao et al. [201,202]. They used gallium filled CNT as nanothermometer and transmission electron microscope is necessary for observation of CNT during temperature measurement.

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers 53 From our experimental observation, it is proposed that the nanothermometer can be constructed using MWCNT more easily. As the emission current vary linearly with temperature for a particular applied electric field, so temperature can be directly measured. The sensitivity of the nanothermometer can be adjusted by choosing the area of the MWCNT film or appropriate applied electric field. The above study shows that the temperature dependent field emission property of CNFs and MWCNTs have potential for development of direct thermal-to-electrical power conversion applications. Continued improvements in the PECVD of CNFs/CNTs and related nanostructures are indeed required to explore the potential utility of these structures in advanced applications and future large-scale integration.

REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13]

[14] [15] [16] [17] [18] [19] [20]

H.W. Kroto, J.R. Heath, S.C. Obrien, S.C. Curl and R.E. Smalley, Nature, 318, (1985) 162. B.T. Kelly, Physics of Graphite, Applied Science, London, (1981). M. Meyyappan, (Eds.) Carbon nanotubes Science and Applications, CRC Press, (2005). M.L. Weeks, T. Rahmanb, P.D. Frymier, S.K. Islam and T.E. McKnight, Sensors and Actuators B, 133, (2008) 53. M. Endo, C. Kim, T. Karaki, Y. Nishimura, M.J. Matthews, S.D.M. Brown and M.S. Dresselhaus, Carbon 37, (1999) 561. S. Iijima, Nature, 354, (1991) 56. S. Iijima and T. Ichihashi, Nature, 363, (1993) 603. D.S. Bethune, C.H. Kiang, M.S.D. Vries, G. Gorman, R. Savoy, J. Vazquez and R. Beyers, Nature, 363, (1993) 605. T. Yildirim, O. Gulseren, C. Kılıc and S. Ciraci, Phys. Rev. B, 62, (2000) 12648. M.S. Dresselhaus, G. Dresselhaus and P.C. Eklund, Science of Fullerenes and Carbon Nanotubes, Academic Press, London, (1996). R. Saito, G. Dresselhaus and M.S. Dresselhaus, Physical Properties of Carbon Nanotubes, Imperial College Press, London (1998). S.C. Tsang, Y.K. Chen, P.J.F. Harris and M.L.H. Green, Nature, 372, (1994) 159. J. Liu, A. Rinzler, H. Dai, J. Hafner, R. Bradley, P. Boul, A. Lu, T. Iverson, K. Shelimov, C. Huffman, F.R. Macias, Y. Shon, R. Lee, D. Colbert and R.E. Smalley, Science, 280, (1998) 1253. P.M. Ajayan and S. Iijima, Nature, 361, (1993) 333. M.S. Dresselhaus, G. Dresselhaus and R. Saito, Phys. Rev. B, 45, (1992) 6234. J.E. Fischer and A.T. Johnson, Solid State Mater. Sci., 4, (1999) 28. J.P. Issi, J.C. Charlier, K. Tanaka, and T. Yamabe and K. Fukui, (Eds.) The Science and Technology of Carbon Nanotubes, Elsevier, Amsterdam, (1999). C.A. Bessel, K. Laubernds, N.M Rodriguez and R.T.K. Baker, J. Phys. Chem., 105, (2001) 1115. J.A. Elliott, J.K.W. Sandler, A.H. Windle, R.J. Young and M.S.P. Shaffer, Phys. Rev. Lett., 92, (2004) 95501. C.S. Kuo, A. Bai, C.M. Huang, Y.Y. Li, C.C. Hu and C.C. Chen, Carbon, 43, (2005) 2760.

54

Sk. F. Ahmed and K. K. Chattopadhyay

[21] M.S. Dresselhaus, G. Dresselhaus and P. Avouris, (Eds.) Carbon nanotubes: Synthesis, properties and applications, Spinger-verlog, (2000). [22] E.W. Wong and P.E. Sheehan, C.M. Lieber, Science, 277, (1997) 1971. [23] P. Poncharal, Z.L. Wang, D. Ugarte and W.A.D. Heer, Science, 283, (1999) 1513. [24] W.D. Callister, Materials science and engineering, an introduction, New York, Wiley, (2003). [25] J.N. Coleman, U. Khan, W.J. Blau and Y.K. Gunko, Carbon, 44, (2006) 1624. [26] M.A. Osman and D. Srivastava, Nanotechnology, 12, (2001) 21. [27] B.M. Kim and A.M.S. Fuhrer, J. Phys. Cond. Matter., 16, (2004) 553. [28] P.L. Mceuen and J.Y. Park, M.R.S. Bulletin, 29, (2004) 272. [29] P.G. Collins and A. Zettl, Appl. Phys. Lett., 69, (1996)1969. [30] L. Song, S. Liu, G.M. Zhang, L.F. Liu, W.J. Ma, D.F. Liu, X.W. Zhao, S.D. Luo, Z.X. Zhang, Y.J. Xiang, J. Shen, J.J. Zhou, G. Wang and W.Y. Zhou, Chinese Phys. 15, (2006) 422. [31] C. Bower, O. Zhou, W. Zhu, A.G. Ramirez, G.P. Kochanski and S. Jin, Mater. Res. Soc. Symp., 593, (1999). [32] Y. Wang, H. Chen, R.W. Hoffman and J.C. Angus, J. Mater. Res. 5 (1990) 2378. [33] ] M. Hillert and N.Z. Lange, Kristallogr, 111, (1958) 24. [34] L.V. Radushkevich and V.M. Lukyanovich, Russian J. of Phys. Chem., 26, (1952) 88. [35] P. Schutzenberger and L.C.R. Schutzenberger, Acad. Sci. Paris, 111, (1890) 774. [36] C. Pelabon and H.C.R. Pelabon, Acad. Sci. Paris, 137, (1903) 706. [37] R. Bacon, J. Appl. Phys., 31, (1960) 283. [38] C. Herring and J.K. Galt, Phys. Rev., 85, (1952) 1060. [39] A.P. Levitt, Whisker Technology, Wiley-Interscience, New York, 1970. [40] A.W. Moore, A.R. Ubbelohde and D.A. Young, J. Appl. Phys., 13, (1962) 393. [41] L.C.F. Blackman and A.R. Ubbelohde, Proc. Roy. Soc.A, London 266, (1962) 20. [42] T. Koyama, Carbon 10, (1972) 757. [43] M. Endo, T. Koyama and Y. Hishiyama, Jap. J. Appl. Phys., 15, (1976) 2073. [44] G.G. Tibbetts, Appl. Phys. Lett., 42, (1983) 666. [45] G.G. Tibbetts, J. Cryst. Growth, 66, (1984) 632. [46] M. Endo, The growth mechanism of vapor-grown carbon fibers, Orleans, France, (1975). [47] M. Endo, PhD thesis, Nagoya University, Japan, (1978). [48] A. Oberlin, M. Endo and T. Koyama, J. Cryst. Growth, 32, (1976) 335. [49] W. Kratschmer, L.D. Lamb, K. Fostiropoulos and D.R. Huffman, Nature, 347, (1990) 354. [50] T.W. Ebbesen, J. Mater. Res., 13, (1998) 2432. [51] A. Thess, R. Lee, P. Nikolaev, H. Dai and J. Robert, Science, 273, (1996) 483. [52] H. Dai, M.S. Dresselhouse, G. Dresselhouse and P. Avouris, (Eds.) Carbon nanotubes: synthesis, structure, properties and application. New York, Springer, (2001). [53] Z.F. Ren, Z.P. Huang, J.W. Xu, J.H. Wang, P. Bush and M.P. Siegal, Science, 282, (1998) 1105. [54] Y. Gogotsi, J.A. Libra and M. Yoshimura, J. Mater. Res., 15, (2000) 2591. [55] J.M.C. Moreno and M. Yoshimura, J. Am. Chem. Soc., 123, (2001) 741. [56] C.H. Kiang, W.A. Goddard, R. Beyers and D.S. Bethune, Carbon, 33, (1995) 903.

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers 55 [57] T. Guo, P. Nikolaev, A. Thess, D.T. Colbert and R.E. Smalley, Chem. Phys. Lett., 243, (1995) 49. [58] H. Dai, A.G. Rinzler, P. Nikolaev, A. Thess, D.T Colbert. and R.E. Smalley, Chem. Phys. Lett., 260, (1996) 471. [59] J. Kong, H.T. Soh, A.M. Cassel, C.F. Quate and H. Dai, Nature, 395, (1998) 878. [60] H. Kind, J.M. Bonard, L. Forro, K. Kern, K. Hernadi, L. Nilsson and L. Schlapbach, Langmuir, 16, (2000) 6877. [61] M. Su, B. Zheng and J. Liu, Chem. Phys. Lett., 322, (2000) 321. [62] L. Delzeit, B. Chen, A. Cassell, R. Stevens, C. Nguyen and M. Meyyappan, Chem. Phys. Lett., 348, (2001) 368. [63] R.T.K. Baker, Carbon, 27, (1989) 315. [64] R.T.K. Baker, L.L. Murrell, (Eds.), Novel Materials in Heterogeneous Catalysis,Washington, DC (1990) 32. [65] C. Bower, W. Zhu, S. Jin and O. Zhou, Appl. Phys. Lett., 77, (2000) 830. [66] W. Kim, A. Javey, O. Vermesh, O. Wang, Y.M. Li and H.J. Dai, Nano Lett., 3, (2003) 193. [67] Y.C. Choi, D.W. Kim, T.J. Lee and Y.H. Lee, Synthetic metals, 117, (2001) 81. [68] Q. Wei, J. Sankos and J. Narayan, Surf. Coat. Technol., 146-147, (2001) 250. [69] Y.Y. Wei, G. Eres, V.I. Merkulov and D.H. Lowndes, Appl. Phys. Lett., 78, (2001) 1394. [70] M. Joseyacaman, M. Mikiyoshida, L. Rendon and J.G. Santiesteba, Appl. Phys. Lett., 62, (1993) 657. [71] M. Chhowalla, K.B.K. Teo, C. Ducati, N.L. Rupesinghe, G.A.J. Amaratunga, A.C. Ferrari, D. Roy, J. Robertson and W.I. Milne, J. Appl. Phys., 90, (2001) 5308. [72] V.I. Merkulov, D.H. Lowndes, Y.Y. Wei, G. Eres and E. Voelkl, Appl. Phys. Lett., 76, (2000) 3555. [73] G.W. Ho, A.T.S. Wee, J. Lin and W.C. Tjiu, Thin Solid Films, 388, (2001) 73. [74] L.C. Qin, D. Zhou, A.R. Krauss and D.M. Gruen, Appl. Phys. Lett., 72, (1998) 3437. [75] C.L. Cheung, H. Park and C.M. Lieber, J. Phys. Chem. B, 106, (2002) 2429. [76] H. Kind, J.M. Bonard, C. Emmenegger, L.O. Nilsson, K. Hernadi, E.M. Schaaler, L. Schlapbach, L. Forro and K. Kern, Adv. Mater., 11, (1999), 1285. [77] Y. Chen, Z.L. Wang, J.S. Yin, D.J. Johnson and R.H. Prince, Chem. Phys. Lett,. 272, (1997) 178. [78] K. Kamada, T. Ikuno, S. Takahashi, T. Oyama, T. Yamamoto, M. Kamizono, S. Ohkura, S. Honda, M. Katayama, T. Hirao and K. Oura, Appl. Surf. Sci., 212, (2003) 383. [79] V.I. Merkulov, A.V. Melechko, M.A. Guillorn, D.H. Lowndes and M.L. Simpson, Appl. Phys. Lett., 80, (2002) 476. [80] Y.M. Shyu and F.C.N. Hong, Diam. Relat. Mater., 10, (2001) 1241. [81] M.S. Kin, N.M. Rodriguez and R.T.K. Baker, J. Catal. Growth., 131, (1991) 60. [82] T.K. Tsai, C.C. Chuang, C.G. Chao and W.L. Liu, Diam. Relat. Mater., 12, (2003) 1453. [83] N.M. Rodriguez, J. Mater. Res., 8, (1993) 3233. [84] M. Yudasaka, R. Kikuchi, T. Matsui, Y. Ohki, S. Yoshimura and E. Ota, Appl. Phys. Lett., 67, (1995) 2477. [85] C. Bower, O. Zhou, W. Zhu, D.J. Werder and S. Jin, Appl. Phys.Lett., 77, (2000) 2767.

56

Sk. F. Ahmed and K. K. Chattopadhyay

[86] C.C. Chuang, W.L. Liu, W.J. Chen and J.H. Huang, Appl. Surf. Sci., 254, (2008) 4681. [87] C.C. Chuang, W. L. Liu, W.J. Chen and J.H. Huang, Surf. Coat. Technol., 202, (2008) 2121. [88] V.I. Merkulov, A.V. Melechko, M.A. Guillorn, D.H. Lowndes and M.L. Simpson, Appl. Phys. Lett., 79, (2001) 2970. [89] X. Yang, M.A.Guillorn, D. Austin, A.V. Melechko, H. Cui, H.M. Meyer III, V.I. Merkulov, J.B.O. Caughman, D.H. Lowndes and M.L. Simpson, Nano Lett, 3, (2003) 1751. [90] J. Li, H.T. Ng, A. M. Cassell, W. Fan, H. Chen, Q. Ye, J. Koehne, J. Han and M. Meyyappan, Nano Lett, 3, (2003) 597. [91] H.J. Qi, K.B.T. Teo, K.K.S. Lau, M.C. Boyce, W.I. Milne, J. Robertson, K.K. Gleason, J. Mech. Phys. Sol., 51, (2003) 2213. [92] H.F. Chuang, S.M. Cooper, M. Meyyappan and B.A. Cruden, J. Nanosci. Nanotechnol., 4, (2004) 964. [93] I. Alstrup, J. Catalysis, 109, (1988) 241 [94] D. Nolan, D.C. Lynch and A.H. Cutler, J. Phys. Chem. B, 102, (1998) 4165. [95] R.T.K Baker and J.J. Chludzinski Jr., J. Catalysis, 64, (1980) 464. [96] C.M. Hsu, C.H. Lin, H.J. Lai and C.T. Kuo, Thin Solid Films, 471, (2005) 140. [97] Q. Jiang, Y.W. Wang and J.C. Li, Appl. Surf. Sci., 152, (1999) 156. [98] C.C. Lin, C.C.I.C. Leu, J.H. Ten and M.H. Hon, Diam. Relat. Mater., 14, (2005)1901. [99] X. Chen, R. Wang, J. Xu and D. Yu, Micro, 35, (2004) 455. [100] K. Bartsch and A. Leonhard, Carbon, 42, (2004) 1731. [101] V.I. Merkulov, A.V. Melechko, M.A. Guillorn, D.H. Lowndes and M.L. Simpson, Chem. Phys. Lett., 361, (2002) 492. [102] Z.P. Huang, J.W. Xu, Z.F. Ren, J.H. Wang, M.P. Siegal and P.N. Provencio, Appl. Phys. Lett., 73, (1998) 3845. [103] F. Jensen, Plasma-enhanced Chemical Vapor Deposition, American Vacuum Society, New York, (1997). [104] V.I. Merkulov, M.A. Guillorn, D.H. Lowndes, M.L. Simpson and E. Voelkl, Appl. Phys. Lett., 79, (2001) 1178. [105] M. Laudon, N.N. Carlson, M.P. Masquelier, M.S. Daw and W. Windl, Appl. Phys. Lett., 78, (2001) 201. [106] Q. Ngo, A. M. Cassell, V. Radmilovic, J. Li, S. Krishnan, M. Meyyappan and C. Y. Yang, Carbon, 45, (2007) 424. [107] H. Huang, H. Kajiura, Y. Murakami and M. Ata, Carbon, 41, (2003) 615. [108] L. Yu, L. Sui, Y. Qin and Z. Cui, Chem. Eng. Journal, 144, (2008) 514. [109] P.L. Hansen, J.B.Wagner, S. Helveg, J.R.R. Nielsen, B.S. Clausen and H. Topsoe, Science, 295, (2002) 2053. [110] N.M. Rodriguez, A. Chmbers and R.T.K. Baker, Langmuir, 11, (1995) 3862. [111] A. Chambers and R.T.K. Baker, J. Phys. Chem. B, 101, (1997) 1621. [112] M. Tanemura, M. Kitazawa, J. Tanaka, T. Okita, R. Ohta1, L. Miao and S. Tanemura, Jpn. J. Appl. Phys., 45, (2006) 2004. [113] O. Auciello and R. Kelly, (Eds.), Ion Bombardment Modification of Surfaces, Elsevier, Amsterdam, (1984). [114] J.A.V. Vechten, W. Solberg, P.E. Batson, J.J. Cuomo and S.M. Rossnagel, J. Cryst. Growth, 82, (1987) 289.

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers 57 [115] P. Sigmund, (Eds.) Fundamental Processes in Sputtering of Atoms and Molecules, Royal Danish Academy of Sciences and Letters, Copenhagen, (1993). [116] J. Li, C. Papadopoulos, J.M. Xu and M. Moskovits, Appl. Phys. Lett., 75, (1999) 367. [117] R. Gomer, Field Emission and Field Ionization, American Institute of Phys., New York, (1993). [118] R.H. Fowler and L.W. Nordheim, Proc. R. Soc. London, series A, 119, (1928) 173. [119] E.L. Murphy and Jr. R.H. Good, Phys. Rev., 102, (1956) 1464. [120] R.G. Forbes, Ultramicroscopy, 79, (1999)11. [121] R.G. Forbes, Proceedings of 4th Moscow International ITEP School of Physics (Eds: A.L. Suvorov, Yu.G. Abov, V.G. Firsov) Moscow, (2001) p.62. [122] L. Nilsson, O. Groening, P. Groening, O. Kuettel and L. Schlapbach, J. Appl. Phys., 90, (2001) 768. [123] R. Gao, Z. Pan and Z.L. Wang, Appl Phys. Lett., 78, (2001) 1757. [124] R.G. Forbes, C.J. Edgcombe and U. Valdre, Ultramicroscopy, 95, (2003) 57. [125] J. Cumings, A. Zettl, M.R. McCartney and J.C.H. Spence, Phys. Rev. Lett., 88, (2002) 56804. [126] X. Zheng, G.H. Chen, Z. Li, S. Deng and N. Xu, Phys. Rev. Lett., 92, (2004) 106803. [127] R. Collazo, R. Schlesser and Z. Sitar, Appl. Phys. Lett., 78, (2001) 2058. [128] S.M. Yoon, J. Chae and J.S. Suh, Appl. Phys. Lett., 84, (2004) 825. [129] S.C. Kung, K.C. Hwang and I.N. Lin, Appl. Phys. Lett., 80, (2002) 4819. [130] S.H. Jo, Y. Tu, Z.P. Huang, D.L. Carnahan, J.Y. Huang, D.Z. Wang and Z.F. Ren, Appl. Phys. Lett., 84, (2004) 413. [131] A.G. Rinzler, J.H. Hafner, P. Lou Nikolaev, L. Kim, S.G. Tomanek, D. Colbert and R.E. Smalley, Science, 269, (1995) 1550. [132] Y. Saito, K. Hamaguchi, K Hata., K. Uchida, Y. Tasaka, F. Ikazaki, M. Yumura, A. Kasuya and Y. Nishina, Nature, 389, (1997) 554. [133] Y. Saito, K. Hamaguchi, T. Nishino, K. Hata, K. Tohji, A. Kasuya and Y. Nishina, Jpn. J. Appl. Phys., 36, (1997) 1340. [134] P.C. Collins, M.S. Arnold and P. Avouris, Science, 292, (2001) 706. [135] W.A. De Heer, A. Chatelain and D. Ugerate, Science, 270, (1995) 1179. [136] W. Zhu, C. Bower, O. Zhou, G. Kochanski and S. Jin, Appl. Phys. Lett., 75, (1999) 873. [137] H. Sugie, M. Tanemura, V. Filip, K. Iwata, K. Takahashi and F. Okuyama, Appl. Phys. Lett., 78, (2001) 2578. [138] J. Robertson, Mater. Res. Soc. Symp. Proc., 498, (1998) 197. [139] J. Robertson and M.J. Rutter, Diam. Relat. Mater., 7, (1998) 620. [140] G.A.J. Amaratunga and S.R.P. Silva, App Phys Lett., 68, (1996) 2529. [141] F.Y. Chuang, C.Y. Sun, H.F. Cheng, W.C. Wang and I.N. Lin, Appl. Surf. Sci., 113, (1997) 259. [142] O. Groning, O.M. Kuttel, P. Groning and L. Schlapbach, Appl. Surf. Sci., 111, (1997) 135. [143] A. Ilie, A.C. Ferrari, T. Yagi and J. Robertson, Appl. Phys. Lett., 7, (2000) 2627. [144] J.D. Carey, R. D. Forrest, R.U. A. Khan and S.R.P. Silva, Appl. Phys. Lett., 77, (2000) 2006. [145] M.W. Geis, J.C. Twichell, J. Macaulay and K. Okano, Appl. Phys. Lett., 67, (1995) 1328.

58

Sk. F. Ahmed and K. K. Chattopadhyay

[146] B.S. Satyanarayana, A. Hart, W.I. Milne and J. Robertson, Appl. Phys. Lett., 71, (1997) 1430. [147] A. Hart, B.S. Satyanarayana, J. Robertson and W.I. Milne, Appl. Phys. Lett., 74, (1999) 1594. [148] M.W. Geis, J.C. Twichell and T.M. Lyszczarz, J. Vac. Sci. Technol. B, 14, (1996) 2060. [149] R.D. Forrest, A.P. Burden, S.R.P. Silva, L.K. Cheah and X. Shi, Appl. Phys. Lett., 73, (1998) 3784. [150] J. Robertson, Mater. Res. Soc. Symp. Proc. 417, (1997) 217. [151] N.S. Xu, Y. Tzeng and R.V. Latham, J. Phys. D, 26, (1993) 1776. [152] A.V. Karabutov, V.D. Frolov, S.M. Pimenov and V.I. Konov, Diam. Relat. Mater., 8, (1999) 763. [153] O.M. Kuttel, O. Groning, C. Emmenegger, L. Nilsson, E. Maillard, L. Diederich and L. Schlapbach, Carbon, 37, (1999) 745. [154] T. Ikunoa, J.T. Ryu, T. Oyamaa, S. Ohkura, Y.G. Baek, S. Honda, M. Katayama, T. Hirao and K. Oura, Vacuum, 66, (2002) 341. [155] J.M.. Kim, W.B. Choi, N.S. Lee and J.E. Jung, Diam. Relat. Mater., 9, (2000) 1184. [156] H. Murakami, M. Hirakawa, C. Tanaka and H. Yamakawa, Appl. Phys. Lett., 76, (2000) 1776. [157] A.N. Obraztov, A.P. Volkov and I. Pavlovsky, Diam. Relat. Mater., 9, (2000) 1190. [158] F. Hoshi, K. Tsugawa, A. Goto, T. Ishikura, S.Yamashita, M. Yumura, T. Hirao, K. Oura and Y. Koga, Diam. Relat. Mater., 10, ( 2001) 254. [159] A. Ilie, A.C. Ferrari, T. Yagi, S.E. Rodil, J. Robertson, E. Barborini and P. Milani, J. Appl. Phys., 90, (2001) 2024. [160] D.N. Davydov, P.A. Sattari, D. AlMawlawi, A. Osika and T.L. Haslett, J. Appl. Phys., 86, (1999) 3983. [161] C. Wang, A. Garcia, D.C. Ingram, M. Lakea and M.E. Kordesch, Electron. Lett., 27, (1991) 1459. [162] N.S. Xu, R.V. Latham and Y. Tzeng, Electron. Lett., 29, (1993) 1596. [163] W. Zhu, C. Bower, G.P. Kochanski and S. Jin, Solid State Electron., 45, (2001) 921. [164] A.N. Obraztsov, I.Y. Pavlovsky, A.P. Volkov, A.S. Petrov, V.I. Petrov, E.V. Rakova, V.V. Roddatis, Diam. Relat. Mater., 8, (1999) 814. [165] J.M. Bonard, J.P. Salvetat, T. Stockli, W.A. de Heer, L. Forro and A. Chatelain, Appl. Phys. Lett., 73, (1998) 918. [166] C. Kimura, Y. Yamamuro, H. Aoki and Takashi Sugino, Diam. Relat. Mater., 16, (2007) 1383. [167] K.B.K. Teo, S.B. Lee, M. Chhowalla, V. Semet, V.T. Binh, O. Groening, M. Castignolles, A. Loiseau, G. Pirio, P. Legagneux, D. Pribat, D.G. Hasko, H. Ahmed, G.A.J. Amaratunga and W.I. Milne, Nanotechnology, 14, (2003) 204. [168] S.F. Ahmed, S. Das, M.K. Mitra and K.K. Chattopadhyay, Ind. Journal of Pure and Applied Physics, 44, (2006) 700. [169] V.L. Merkulov, D.H. Lowndes, Y. Wei and G. Eres, Appl. Phys. Lett., 76, (2000) 3555. [170] S.F. Ahmed, Ph.D. Thesis, Some aspects of nanocrystalline silicon incorporated diamond like carbon and carbon nanostructures, Jadavpur University, Kolkata, India, (2008). [171] S. Gupta, B.L. Weiss, B.R. Weiner and G. Morell, J. Appl. Phys., 89, (2001) 5671.

Synthesis and Electron Field Emission from Different Morphology Carbon Nanofibers 59 [172] A.Y. Cao, X.F. Zhang, X. Xiao, M.Q. Ding, D.M. Zhuang, C.L. Xu, B.Q. Wei, J., Liang and D.H. Wu, Mater. Lett., 51, (2001) 371. [173] L. Xinghui, Z. Changchun, L. Weihua, Z. Fanguang and T. Changhui, Mater. Chem. and Phys., 93, (2005) 473. [174] K.H. Park, J.H. Yim, S. Lee and K.H. Koh, J Vac Sci Technol B, 23, (2005) 776. [175] X.T. Zhou, H.L. Lai, H.Y. Peng, F.C.K. Au, L.S. Liao, N. Wang, I. Bello, C.S. Lee and S.T. Lee, Chem. Phys. Lett., 318, (2000) 58. [176] F.C.K. Au, K.W. Wong, Y.H. Tang, Y.F. Zhang, I. Bello and S.T. Lee, Appl. Phys. Lett., 75, (1999)1700. [177] K. Okano, S. Koizumi, S.R.P. Silva and G.A.J. Amartunga, Nature, 381, (1996) 140. [178] S. Das, S.F. Ahmed, M.K. Mitra and K.K. Chattopadhyay, Appl. Phys. A, 91, (2008) 429. [179] S.F. Ahmed, M.K. Mitra and K.K. Chattopadhyay, Appl. Surf. Sci., 254, (2007) 610. [180] W. Zhu, Vacuum Microelectronics, Wiley, New York, (2001) Chapter 3. [181] A. Modinos, Field, Thermionic, and Secondary Electron Emission Spectroscopy, Plenum, New York, (1984). [182] M. Shiraishi and M. Ata, Carbon, 39, (2001) 1913. [183] M.C. Kan, J.L. Huang, J.C. Sung, K.H. Chen and B.S.Yau, Carbon, 41, (2003) 2839. [184] M.J. Fransen, T.L.V.Rooy and P. Kruit, Appl. Surf. Sci., 146, (1999) 312. [185] Y. Chen, D.T. Shaw and L. Guo, Appl. Phys. Lett., 76, (2000) 2469. [186] L. Nilsson, O. Groening, C. Emmenegger, O. Kuettel, E. Schaller, L. Schlapbach H. Kind, J.M. Bonard and K. Kern, Appl. Phys. Lett., 76, (2000) 2071. [187] S. Frank, P. Poncharal, Z.L. Wang and W.A.D. Heer, Science, 280, (1998) 1744. [188] C. Dekker, W. Henk, C. Postma, T. Teepen, Z. Yao and M. Grifoni, Science, 293, (2001) 5527. [189] A. Bachtold, P. Hadley, T. Nakanishi and C. Dekker, Science, 294, (2001) 1317. [190] R.H. Baughman, A.A. Zakhidov and W.A. de Heer, Science, 297, (2002) 787. [191] M. Yu, O. Lourie, M.J. Drayer, K. Molini, T.F. Kelly and R.S. Ruoff, Science, 287, (2000) 637. [192] B.Q. Wei, R. Vajtal and P.M. Ajayan, Appl Phys Lett., 79, (2001) 1172. [193] W.I. Milne, K.B.K. Teo, G.A.J. Amaratunga, P. Legagneux, L. Gangloffa and J.P. Schnell, J. Mater. Chem., 14, (2004) 933. [194] G.S. Duesberg, A.P. Graham, M. Liebau, R. Seidel, E. Unger and F. Kreupl, Nano Lett., 3, (2003) 257. [195] J. Kong, N.R. Franklin, C. Zhou, M.G. Chaplene, S. Peng and K. Cho, Science, 287, (2000) 622. [196] C. Emmenegger, P. Mauron, P. Sudan, P. Wenger, V. Hermann and R. Gallay, J. Power Sources, 124, (2003) 321. [197] T. Yoshitake, Y. Shimakawa, S. Kuroshima, H. Kimura, T. Ichihashi and Y. Kubo, Physica B, 323, (2002) 124. [198] M. Endo, Y.A. Kim, T. Hayashi, K. Nishimura, T. Matusita and K. Miyashita, Carbon, 39, (2001) 1287. [199] H.S. Sim, S.P. Lau, L.K. Ang, M. Tanemura and K. Yamaguchi, Diam. Relat. Mater., 17, (2008) 1959. [200] Q. Ngo, B.A. Cruden, A.M. Cassell, G. Sims, M. Meyyappan, J. Li and C.Y. Yang, Nano Letters, 4, (2004) 2403.

60

Sk. F. Ahmed and K. K. Chattopadhyay

[201] Y.H. Gao and Y. Bando, Nature, 415, (2002)599. [202] Y.H. Gao, Y. Bando, Z.W. Liu and D. Golberg, Appl. Phys. Lett., 83, (2003) 2913.

In: Nanofibers: Fabrication, Performance, and Applications ISBN: 978-1-60741-947-1 Editor: W. N. Chang ©2009 Nova Science Publishers, Inc.

Chapter 2

PERMEABILITY STUDIES OF ELECTROSPUN CHITIN AND CHITOSAN NANOFIBROUS MEMBRANES Jessica D. Schiffman and Caroline L. Schauer* Drexel University, Philadelphia, PA, USA

1. ABSTRACT Electrospinning has been utilized to fabricate fibrous membranes composed of polymer nanofibers, which have large surface area-to-volume ratios and small pores. Electrospun nanofibrous membranes have potential uses in a variety of industries such as energy, environment, medicine, packaging, and automotive, with specific applications including air filtration, protective clothing, fuel cells, and nanocomposites. Nanofibrous membranes composed of biopolymers have potential uses that harness their inherent biocompatibility. Chitin, the second most abundant, naturally occurring polysaccharide after cellulose, is found in shells of crabs and shrimp. Chitosan, the acid soluble form of chitin, is a non-toxic, biodegradable, biopolymer consisting primarily of β(1→4) linked 2-amino-2-deoxy-β-D-glucopyranose units, and is currently used in tissue engineering, antifouling coatings, separation membranes, stent coatings, enzyme immobilization matrices, and the removal of heavy metals from ground and wastewater. Chitosan is a commercially interesting compound because of its high nitrogen content (6.89%), making it a useful chelating agent for metal ions. Before these chitin or chitosan nanofibrous membranes can be used in the myriad of industries their physical properties, such as permeability, must be known. This chapter focuses on the fabrication and flow cell testing of chitin and chitosan nanofibrous membranes. Additionally, it explores the potential applications of biopolymer and synthetic polymer electrospun membranes.

* Telephone: (215) 895-6797, Fax: (215) 895-6760, E-mail: [email protected]

62

Jessica D. Schiffman and Caroline L. Schauer

2. INTRODUCTION 2.1. Electrospinning Non-woven fibrous membranes, which have large surface area-to-volume ratios can be fabricated utilizing the electrospinning process. A simple schematic of a standard laboratory set-up is given in Figure 1. This is the only production method that produces fibers by utilizing electrical forces as opposed to a mechanical pulling force. As a result of this, submicron fibers are created and deposited in the form of a non-woven membrane.

Figure 1. Diagram displaying the basic components of an electrospinning apparatus including (a) syringe needle loaded with a polymeric solution, which is often advanced at a constant rate by a metering pump, (b) voltage supply, and (c) target onto which the electrospun fibers accumulate.

In a 2007 review article, which discussed the state of electrospun filtration membranes, Barhate and Ramakrishna [1] included a table of enterprises that work in the nanofiber production area. Their table included eight companies in the United States of America, three in Germany, three in Japan, two in South Korea, as well as one in Canada, the Czech Republic, the Republic of Estonia, and Finland. Therefore, there is a widespread invested interest in nanofibrous filtration.

2.2. Select Electrospinning Examples The specific investigations that are discussed in this section are meant to represent (1) the diversity of precursor materials that have been electrospun and (2) the wide range of filtration related applications from said materials. All of the articles discussed in this section have been published within the previous year and utilize synthetic polymers. Electrospun membranes could potentially act as pre-filters to rid solid particulates from a flow so that the down-stream filtration membranes will become less fouled and experience an increased lifecycle between required cleanings or replacements. The functionality of electrospun nylon-6 (dissolved in 75% formic acid) was tested as a pre-filter by attempting to

Permeability Studies of Electrospun Chitin and Chitosan Nanofibrous Membranes

63

pass polystyrene (PS) micro-particles (Ø= 0.5 - 10 µm) through the membrane.[2] It was determined that the membranes became irreversibly fouled. Kim et al. [3] have also electrospun pre-filters from nylon 6/formic acid solutions and investigated their efficiency utilizing polystyrene latex particles with known diameters ranging from 0.02 - 1.0 µm. Membranes with a mean diameter of 100 nm displayed a lower pressure drop in comparison to commercial high efficiency particulate air (HEPA) filters. The collection of aerosol particles by an electrospun polyacrylamide solution containing 2,2′-(bisacrylamino) diethyl disulfide (BAC) was tested by Vetcher et al.[4] Potentially, these filters could be utilized for indoor and outdoor testing of microbes, pollen allergens, and toxins. Liquid filtration applications have also been investigated since electrospun membranes could have higher flux and anti-fouling properties than other ultra-purification membranes. Veleirinho and Lopes-da-Silva [5] have electrospun a solution of poly(ethylene terephthalate) (PET) dissolved in trifluoracetic acid (TFA) : dichloromethane (DCM) (80:20 v/v) to investigated its applicability to apple juice clarification. The clarified juice exhibited physicochemical characteristics comparable when conventional techniques were employed. In addition to the typically randomly accumulated fibrous membranes, hollow fibers have been electrospun at the University of Washington [6, 7] utilizing a multilayer microlayer electrospinning device. Into one microchannel, Srivastava et al. [8] loaded titanium isopropoxide (Ti(OiPr)4) stock solution (ethanol/acetic acid) that was added to poly(vinylpyrrolidone)/ethanol solution. A heavy mineral oil was added via another microchannel and was the core material. From this, hollow core/sheath nanofibers were electrospun. Robust hollow fibers with diameters ranging from 85 to 350 nm were obtained from these networks, whose geometry could be customized according to the end application.

2.3. Chitin and Chitosan The recent articles discussed in section 2.2 utilized synthetic polymers. We have taken a slightly different approach by electrospinning and testing the biopolymers chitin and chitosan. We believe that these biopolymers hold a strong potential for filtration applications because they are renewable resources that have attractive intrinsic properties. Chitin is the second most abundant natural polysaccharide and consists of N-acetyl-β-D-glucosamine chains. The functional polysaccharide chitin, is produced primarily by arthropods and crustaceans as ordered crystalline microfibrils; it also exists as the structural component in the cell walls of fungi and yeast.[9] The distribution of the chitin produced from arthropods annually is displayed in Figure 2.[10] Chitin exists in two main crystalline polymorphic forms: α and β. Alternating sheets of parallel and antiparallel chains tightly pack into an orthorhombic cell, which are obtained from insect cuticles, shrimp and crab shells, fungal and yeast cells, lobster and crab tendons, it is known as α-chitin. [9, 11, 12] A second polymorph, which is far less common, β-chitin, is found in squid pens, in the tubes synthesized by vestimetiferan and pogonophoran worms, Aphrodite chaetae, and in the monocrystalline spines excreted by the diatom. In this form, the chains are arranged parallel within a monoclinic unit cell.[13] A third rare form, γ-chitin, is thought to be a mixture of the α- and β- chitin forms, containing both parallel and antiparallel arrangements.[14, 15]

64

Jessica D. Schiffman and Caroline L. Schauer

Figure 2. Breakdown of the total chitin produced by arthropods (tons per year) annually from three different environments. Data adapted from Cauchie.[10]

Since chitin is a highly insoluble material, chitosan, the N-deacetylated derivative of chitin is often studied. As a result of the increase in available free amine groups, chitosan can be dissolved in aqueous acidic solvents that chitin cannot dissolve in, such as formic acid, acetic acid (AA), and malic acid. Solubility is achieved by protonating the –NH2 function on the C-2 of the D-glucosamine repeat unit, creating a polyelectrolyte in acidic solutions.[9, 16] Both biopolymers have many of the same attributes including antibacterial activity, biocompatibility, chelating capabilities, biodegradability, and adsorption properties.[16] As a result of these properties, chitin and chitosan can be applied to an extensive array of applications in fields ranging from biomedical to environmental. In the remainder of this chapter, we will focus on our current electrospinning of these two biopolymers, as well as initial results concerning their permeability.

3. RESEARCH RESULTS Both chitin [17-20] and chitosan [21-28] fibrous membranes have previously been fabricated utilizing the electrospinning process. In the case of electrospinning chitin and chitosan, the given biopolymers were dissolved in 1,1,1,3,3,3-hexafluoro-2-propanol (HFIP) for at least 72 h and TFA for 24 h, respectively. The polymeric solutions were loaded into (5 mL) Luer-lock syringes (Becton Dickinson & Co, Franklin Lakes, NJ), which were capped with Precision Glide 21-gauge needle (Becton Dickinson & Co, Franklin Lakes, NJ) and placed on a metering pump (Harvard Apparatus, Plymouth Meeting, PA) that was set at a constant advancing speed around 1.2 mL/h. When a voltage of approximately 25 kV was applied by the high voltage supply (Gamma High Voltage Research Inc., Ormond Beach, FL) to the metal needle and the target, a Taylor cone [29] was formed. This is a critical step for the initiation and propagation of fiber formation. Next, the electrostatic force needs to overcome the surface tension force of the Taylor cone so that a thin jet can form and thin out over three stages. They include: jet initiation and extension in a straight line, whipping instability, and jet solidification and fiber collection. In all of our experiments, a copper plate

Permeability Studies of Electrospun Chitin and Chitosan Nanofibrous Membranes

65

wrapped with store bought aluminum foil was utilized as the target. Dry chitin and chitosan fibers were deposited onto the target since most of the solvent utilized to dissolve the biopolymers evaporates over the course of the separation distance (gap between the tip of the needle and the target). The separation distance utilized for the electrospinning of chitin and chitosan was 6.4 cm and 6.0 cm, respectively. Utilizing the aforementioned parameters, chitin and chitosan fibrous membranes were successfully electrospun into nanofibrous membranes comprised of randomly oriented, continuous cylindrical fibers. The practical grade (PG) of chitin from crab shells (SigmaAlrdich, St Louis, MO) that was spun has a degree of deacetylation (DD) of 9% as determined by Fourier transform infrared spectroscopy (FTIR).[19] The DD and molecular weights (MW) of the chitosans that were successfully electrospun [23] are given in Table 1. Table 1. Various chitosan biopolymers, which were purchased from Sigma-Aldrich (St Louis, MO) and successfully electrospun. Their degree of deacetylation (DD) [23] and molecular weight (MW) are given. Chitosan Practial grade (PG) Low molecular weight Medium molecular weight High molecular weight *As determined by FTIR **As supplied by Sigma-Aldrich

DD (%)* 75 74 83 72

MW** 190-375,000 70,000 190-310,000 500-700,000

The average fiber diameter of all membranes electrospun was determined utilizing a Zeiss Supra 50/VP field emission scanning electron microscope (FESEM) by measuring the diameters of fifty random fibers. The term as-spun implies that no additional coatings or alterations were made to the fibrous membranes. As displayed on Table 2, the average fiber diameters of all as-spun chitin and chitosans are within standard deviation from each other. In order for as-spun fibrous membranes to be used in field devices, the chemical stability of the chitosan fibers needed improvement. This was determined as a result of the initial solubility testing conducted on as-spun chitosan membranes. (See Table 2, Solubility: Post 72 h.) The solubility of the chitin and chitosan fibrous membranes were tested [23] utilizing three 15-mm2 petri dishes (Becton Dickinson, Franklin Lakes, NJ), which contained 30 mL of three various solutions: basic (1 M sodium hydrocxide, NaOH), acidic (1 M AA), and aqueous (ultrapure H2O). Two samples of electrospun membrane, (2.54 cm x 1.27 cm) were placed into each solution. After 15 min, if possible, one of the membranes was removed, while the other remained in the solution for 72 h.

66

Jessica D. Schiffman and Caroline L. Schauer

Table 2. Properties of various electrospun membranes including: their average fiber diameter (n=50), the biopolymer that was electrospun and cross-linking method (when applicable), and the fiber membrane stability after fibrous membranes were submerged into 1M solutions of acetic acid (AA), ultrapure aqueous (H2O), and sodium hydroxide (NaOH) for 72 h. All data displayed is for electrospun (PG) practical grade or (MMW) medium molecular weight biopolymers. Electrospun Biopolymer

Stability: Post 72h

Average Fiber

AA

H2O

NaOH

Diameter (nm)

As-spun PG chitin

√

√

√

152

±

70

As-spun PG chitosan

X

X

√

58

±

20

As-spun MMW chitosan X One-step cross-linked MMW chitosan X Two-step cross-linked MMW chitosan √

X

√

77

±

29

√

√

128

±

40

√

√

172

±

75

3.1. Cross-Linking Studies As evident from Table 2, electrospun medium molecular weight (MMW) chitosan membranes did not survive when immersed in aqueous and acetic solutions. Hence, initial investigations regarding two different cross-linking methods were employed. A two-step cross-linking method was first identified, [23] utilizing vapor-phase glutaraldehyde (GA). The first step of this process consists of electrospinning the chitosan solution; step two is exposing the as-spun fibrous membranes overnight to vapor-GA in a vaporization chamber. Shortly after this work, we successfully demonstrated the electrospinning of cross-linked chitosan fibers in one-step, utilizing GA-liquid.[24] In this in-situ method, a small amount of GA liquid is added just prior to the electrospinning of the chitosan solution, and the resultant electrospun fibrous membranes are insoluble in aqueous, basic, and acidic solutions by the completion of the electrospinning session. Both cross-linking methods transform the electrospun chitosan fibrous membranes so that they are stable for at least 72 h in AA, aqueous, and NaOH solutions. Fibrous membranes composed of chitin are stable in the previously mentioned solutions without further processing.

3.2. Lead Ion Testing The PG chitin and MMW two-step cross-linked chitosan fibrous membranes are both chemically stable and have potential to act as filtration devices based on their intrinsic chelation capabilities. Therefore, their capability to bind with lead ions was tested. Solutions containing various amounts of lead acetate were created and PG chitin and two-step crosslinked fibrous membranes of MMW chitosan were subjected to the solutions for 1 h. After this hour, they were removed and rinsed repeatedly with DI water. These samples were next prepared for analysis utilizing a FESEM equipped with energy dispersive x-ray spectroscopy

Permeability Studies of Electrospun Chitin and Chitosan Nanofibrous Membranes

67

(EDS) system. EDS distinctly identified that when fibrous membranes were subjected to solutions with 50 ppm lead, lead was incorporated into the membrane. Chitin and two-step cross-linked MMW chitosan membranes were additionally tested in solutions containing 5 ppm, 500 ppb, and 50 ppb lead. Spectra acquired on these samples indicate that lead has been incorporated into the membranes. However, further work must be done to quantify these amounts, determine binding capacity of the fibrous membranes, and identify the lowest detectable limits of lead for the EDS system.

3.3. Permeability Studies The permeability (k) of two-step cross-linked PG chitosan and as-spun PG chitin membranes were evaluated according to ASTM standard D4491. A constant head test utilizing the flow cell apparatus displayed in Figure 3(c) was employed. A constant head of water was maintained on the electrospun membrane and the quantity of flow was measured over a specific amount of time. In order to calculate the membrane permeability (k), the permittivity (ψ) was first determined according to the following equation: ψ = k/t = q / (∆h * A)

Equation 1

Where: ψ is permittivity (s-1), k is permeability (m/s), t is thickness (m), q is quantity of flow or flow rate (m3/s), ∆h is head lost (m), and A is area of the test specimen (m2). A constant head test over a circular (Ø = 5.08 cm) area was employed. A constant head of water was maintained on the electrospun membrane and the quantity of flow was measured over a specific amount of time. As seen in Figure 3(b), fibrous membranes that were approximately (7.62 cm x 6.35 cm) were electrospun; however the water only flowed through a 5.08 cm diameter circular section of the membranes. To ensure that the membranes did not move during the remaining preparation of the flow cell apparatus or during the experimental run, the membranes were placed between two thin sheets of metal mesh. These background metal meshes were tested independently of the fibrous membranes and it was determined that their permeability is so high that they do not restrict or have a negligible influence on the flow during the testing. The fibrous membranes, sandwiched between the metal mesh was clamped between (5.08 cm diameter) polyvinyl chloride (PVC) pipes that were then clamped together. This area is labeled as “sample holder” in Figure 3(b). After a sample was carefully loaded, de-aired water was added into the discharge pipe until the system was backfilled. The bleed valve was utilized to get rid of any air bubbles that were trapped within the apparatus during this process. Once backfilled, water was added until it reached the overflow outlet, then, the rate of water being added was held constant at a reduced rate. Next, the rotating discharge pipe was moved so that (1) the gauge for measuring head (see Figure 3(c)) displayed a constant height difference (∆h) and (2) there was a constant water runoff from the discharge pipe. At this time, the quantity of flow (q) was recorded for a particular time (s), the readings were averaged, and the permittivity determined according to the equation previously given. This value can then easily be converted to permeability. Multiple samples were utilized and the permittivity of each sample was determined in repetition four times and converted to permeability by utilizing their average thickness

68

Jessica D. Schiffman and Caroline L. Schauer

measurements. It is important to note that this conversion can yield increased amount of standard deviation error. The average permeability of the chitin and two-step cross-linked chitosan membranes was determined to be 0.144 ± 0.013 µm/s and 0.344 ± 0.197 µm/s, respectively. According to ASTM standards, a coefficient of permeability (cm/s) 10-3 > k > 10-5 is the equivalent to sand, dirty sand, or silty sand and has a “low” degree of permeability. The membranes tested fall between this and the permeability category that has a “very low” rating (where 10-5 > k > 10-7); here the membranes fall into the soil rating of silt or silty clay.[30]

Figure 3. Set-up of flow cell apparatus displaying (a) water storage, (b) two-step cross-linked PG chitosan fibrous membrane being loaded into sample holder, and (c) the apparatus utilized to determine permeability of the membranes.

The (5.08 cm diameter) circular area that the water flowed through looks consistent with the rest of the membrane. This indicates that the membranes were both chemically resistant and mechanically strong enough to withstand the forces from the flow cell. Additionally, FESEM micrographs displayed in Figures 4(a) and (c) demonstrates that cylindrical fiber morphology was retained. Both the two-step cross-linked PG chitosan and the PG chitin fibrous membranes display fine cylindrical and continuous fibers. The micrographs were acquired after desiccating the membranes for at least 24 h. The average fiber diameters before and after the flow cell experiment were determined to be within standard deviation of each other before and after use in the flow cells. They were 152

Permeability Studies of Electrospun Chitin and Chitosan Nanofibrous Membranes

69

± 70 nm and 326 ± 171 nm for chitin membranes and 172 ± 75 nm and 253 ± 108 nm for the two-step cross-linked PG chitosan, pre- and post-cell respectively. It is important to note that the biopolymers chitin and chitosan are known to swell in aqueous solutions. Thus, during the flow cell experiments, it can be assumed that the water flowing through the cell is causing the fibers to increase in diameter. These membranes were desiccated prior to the acquisition of the fiber diameter distribution data.

Figure 4. Displays FESEM micrographs and digital images of the fibrous membranes after they were removed from the flow cell. Images (a) and (b) are of a PG chitin membrane, whereas (c) and (d) are a PG chitosan membrane.

The water that has been used throughout the flow cell investigation is unpurified Philadelphia, Pennsylvania tap water. A sample of this water was sent to Robertson Microlit Laboratories (Madison, NJ) to be analyzed. It was determined that the water utilized in this study contains 33 ppm fluorine and less than 1 ppm iron. EDS conducted on the fibrous membranes post-permeability studies were unable to confirm the presence of either of these elements at their aqueous levels in the membranes. Future investigations should include permeability studies that utilize heavy metal ion contaminated waters through the fibrous membranes to determine the detection limits of these fibrous membranes. Additionally, more analysis concerning the uniformity of the fibrous membranes after their use should be evaluated.

4. APPLICATIONS OUTLOOK As devices and their components become miniaturized and contain nano-sized features, higher proportions of atoms are on the surface. This results in new and often enhanced properties such as increased quantum efficiency, surface energy and reactivity, thermal and electrical conductivity, high strength-to-weight ratios, and superparamagnetism.[31, 32] Certainly, by fabricating membranes composed of nanofibers very high surface area-tovolume ratios exist.

70

Jessica D. Schiffman and Caroline L. Schauer

Nanofibrous membranes are appropriate for a wide variety of applications based upon their low weight, high permeability, and small pore sizes. Biopolymer nanofibrous membranes could be appropriate as antimicrobial filters, ultrafiltration membranes, affinity filters, [1] catalytic filters, coalescence filters, ion exchange media, affinity filters, particle filters in vivo, biomedical sutures, [33] filters for metal recovery, [34] as templates, [35-37] and for protective clothing that is both chemically and biologically protective.[38] The porosity of electrospun membranes can be altered [39-41] and thus the properties of these membranes, such as the number of anchoring points for cells, wetting-properties, and degradation rates can vary. Medical textiles, chemical filtration, fuel cell membranes, catalysis, electrochemical cells, and nano-reinforcements would benefit from having a porosity that could be engineered for the needs of the particular application.[42, 43]

5. CONCLUSION Electrospun membranes composed of chitin or chitosan offer many advantages to membranes composed of synthetic polymers. The biopolymers are sustainable and ecoefficient, while also being biocompatible, antibacterial, and biodegradable. These characteristics become even more exciting when the biopolymers are processed into nanofibrous membranes due to the nano-effects that occur. As we start to investigate and understand the permeability properties that electrospun chitin and chitosan membranes have, we foresee these membraneerials will behave as well as synthetic polymers, while having less of an impact on the environment.

REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14]

Barhate, R. S.; Ramakrishna, S. J. Membr. Sci. 2007, 296, 1-8. Aussawasathien, D.; Teerawattananon, C.; Vongachariya, A. J. Membr. Sci. 2008, 315, 11-19. Kim, G. T.; Ahn, Y. C.; Lee, J. K. Korean J. Chem. Eng. 2008, 25, 368-372. Vetcher, A. A.; Gearheart, R.; Morozov, V. N. Polymers for Advanced Technologies 2008, 19, 1276-1285. Veleirinho, B.; Lopes-da-Silva, J. A. Process Biochem. 2009, 44, 353-356. Li, D.; McCann, Jesse T.; Xia, Y. Small 2005, 1, 83-86. McCann, J. T.; Li, D.; Xia, Y. J. Mater. Chem. 2005, 15, 735-738. Srivastava, Y.; Loscertales, I.; Marquez, M.; Thorsen, T. Microfluidics and Nanofluidics 2008, 4, 245-250. Rinaudo, M. Prog. Polym. Sci 2006, 31, 603-632. Cauchie, H. M. Hydrobiologia 2002, 470, 63-96. Schiffman, J. D.; Schauer, C. L. Mater. Sci. Eng., C 2009, 29, 1370-1374. Minke, R.; Blackwell, J. J. Mol. Biol. 1978, 120, 167-181. Gardner, K. H.; Blackwell, J. Biopolymers 1975, 14, 1581-1595. Kurita, K. Prog. Polym. Sci 2001, 26, 1921-1971.

Permeability Studies of Electrospun Chitin and Chitosan Nanofibrous Membranes

71

[15] Lavall, R. L.; Assis, O. B. G.; Campana-Filho, S. P. Bioresour. Technol. 2007, 98, 2465-2472. [16] Kumar, M. N. V. R. React. Funct. Polym. 2000, 46, 1-27. [17] Min, B. M.; Lee, S. W.; Lim, J. N.; You, Y.; Lee, T. S.; Kang, P. H.; Park, W. H. Polymer 2004, 45, 7137-7142. [18] Noh, H. K.; Lee, S. W.; Kim, J. M.; Oh, J. E.; Kim, K. H.; Chung, C. P.; Choi, S. C.; Park, W. H.; Min, B. M. Biomaterials 2006, 27, 3934-3944. [19] Schiffman, J. D.; Stulga, L. A.; Schauer, C. L. Polym. Eng. Sci. 2009, 49, DOI:1002/pen.21434 [20] Schiffman, J. D.; Schauer, C. L. Polymer Reviews 2008, 48, 317-352. [21] Ohkawa, K.; Cha, D.; Kim, H.; Nishida, A.; Yamamoto, H. Macromol. Rapid Commun. 2004, 25, 1600-1605. [22] Ohkawa, K.; Ken-Ichi Minato; Kumagai, G.; Hayashi, S.; Yamamoto, H. Biomacromolecules 2006, 7, 3291-3294. [23] Schiffman, J. D.; Schauer, C. L. Biomacromolecules 2007, 8, 594-601. [24] Schiffman, J. D.; Schauer, C. L. Biomacromolecules 2007, 8, 2665-2667. [25] Sangsanoh, P.; Supaphol, P. Biomacromolecules 2006, 7, 2710-2714. [26] Matsuda, A.; Kagata, G.; Kino, R.; Tanaka, J. J. Nanosci. Nanotechnol. 2007, 7, 852855. [27] Geng, X.; Kwon, O.-H.; Jang, J. Biomaterials 2005, 26, 5427-5432. [28] De Vrieze, S.; Westbroek, P.; Van Camp, T.; Van Langenhove, L. J. Mater. Sci. 2007, 42, 8029-8034. [29] Taylor, G. Proc. R. Soc. London, Ser. A 1964, 280, 383-397. [30] Hoopes, R. J. In The Design and Application of Controlled Low-strength Materials (flowable Fill); Howard AK, Hitch JL, Conshohocken, 1998. ASTM International: 87102. [31] He, J.-H.; Wan, Y.-Q.; Xu, L. Chaos, Solitons and Fractals 2007, 33, 26-37. [32] Bean, C. P.; Livingston, J. D. J. Appl. Phys. 1959, 30, 120S-129S. [33] Zarkoob, S.; Reneker, R. H.; Eby, R. K.; Hudson, S. D.; Erley, D.; Adams, W. W. U.S. Patent 6,110,590. 2000. [34] Ki, C. S.; Gang, E. H.; Um, I. C.; Park, Y. H. J. Membr. Sci. 2007, 302, 20-26. [35] Bognitzki, M.; Hou, H.; Ishaque, M.; Frese, T.; Hellwig, M.; Schwarte, C.; Schaper, A.; Wendorff, J. H.; Greiner, A. Adv. Mater. 2000, 12, 637-640. [36] Caruso, R. A.; Schattka, J. H.; Greiner, A. Adv. Mater. 2001, 13, 1577-1579. [37] Muller, K.; Quinn, J. F.; Johnston, A. P. R.; Becker, M.; Greiner, A.; Caruso, F. Chem. Mater. 2006, 18, 2397-2403. [38] Gibson, P.; Schreuder-Gibson, H.; Rivin, D. Colloids Surf., A 2001, 187-188, 469–481. [39] McCann, J. T.; Marquez, M.; Xia, Y. J. Am. Chem. Soc. 2005, 128, 1436-1437. [40] Casper, C. L.; Stephens, J. S.; Tassi, N. G.; Chase, D. B.; Rabolt, J. F. Macromolecules 2004, 37, 573-578. [41] Bognitzki, M.; Czado, W.; Frese, T.; Schaper, A.; Hellwig, M.; Steinhart, M.; Greiner, A.; Wendorff, J. H. Adv. Mater. 2001, 13, 70-72. [42] Dayal, P.; Kyu, T. J. Appl. Phys. 2006, 100, 043512 043511-043516. [43] Greiner, A.; Wendorff, Joachim H. Angew. Chem. Int. Ed. 2007, 46, 5670-5703.

In: Nanofibers: Fabrication, Performance, and Applications ISBN: 978-1-60741-947-1 Editor: W. N. Chang ©2009 Nova Science Publishers, Inc.

Chapter 3

NOVEL CHITOSAN–CONTAINING MICRO- AND NANOFIBROUS MATERIALS BY ELECTROSPINNING: PREPARATION AND BIOMEDICAL APPLICATION D. Paneva, М. Ignatova, N. Manolova and I. Rashkova Laboratory of Bioactive Polymers, Institute of Polymers, Bulgarian Academy of Sciences, 1113 Sofia, Bulgaria

ABSTRACT At present increased attention is paid on fibrous materials from the natural polymer chitosan because of its numerous beneficial properties (biocompatibility, biodegradability, inherent antibacterial and haemostatic activity). The presence of both hydroxyl and amino groups enables the tailored modification of chitosan into derivatives having targeted properties. The materials containing chitosan or its derivatives are considered as very promising candidates for versatile applications in medicine, pharmacy, food industry, and agriculture. Nowadays the preparation of nanosized fibrous materials is of special interest because of their unique properties, in particular their high surface area-to-volume and aspect ratios. Electrospinning is a cutting edge technique for fabrication of continuous polymer micro- and nanofibers. The basic principles and the effect of the process parameters on the morphology of the electrospun fibers and fibrous materials are briefly discussed in the present Chapter. The first successful attempt to prepare chitosan-containing electrospun materials dates from 2004. This has been achieved by the addition of a non-ionogenic, water-soluble polymer into the spinning solution. The application of this approach for preparation of chitosan-containing fibers is thoroughly discussed in the Chapter. The preparation of neat chitosan nanofibers by electrospinning is outlined as well. The application of suitable chitosan derivatives soluble in water or low toxic organic solvents enables the design of novel non-woven textiles in absence/presence of a non-ionogenic polymer. The preparation of such nontoxic, environmentally friendly materials is detailed. The applied two-step procedures (heat or UV treatment, use of appropriate crosslinking agents) for imparting waterinsolubility to the obtained micro- and nanofibrous materials are described. The main a

e-mail: [email protected].

D. Paneva, М. Ignatova, N. Manolova et al.

74

approaches that have been used for preparation of electrospun materials combining the beneficial properties of chitosan and aliphatic polyesters based on poly(L-lactide): simultaneous electrospinning or electrospinning of the polyester, followed by coating of the non-woven textile with a thin chitosan layer, are summarized. Moreover, the recently developed routes for preparation of chitosan-containing micro- and nanofibers, such as reactive electrospinning, combination of electrospinning and polyelectrolyte complex formation as well as yarns formation, are discussed. The advantages of the one-step imparting of water-insolubility of chitosan fibers by reactive electrospinning and polyelectrolyte complex formation as compared to the two-step procedures are emphasized. Last but not least the potential biomedical application of the obtained microand nanofibers are outlined.

LIST OF ABBREVIATIONS αeq (Me)soln (ne)soln

φp

c* ce AFS C CECh Ch Ch-g-oligo(D,L)LA Ch-g-PLLA CMCh DAS DCM DDA DMF DMPA DMSO DSC GА HFIP HMW LbL LMW Me PAA PAAm PAH PAMPS PBS PEC

equilibrium swelling degree entanglement molecular weight in solution entanglement number of the macromolecules in the solution polymer volume fraction in solution critical chain overlap concentration entanglement concentration applied field strength cardboard N-carboxyethylchitosan chitosan chitosan-graft-oligo(D,L-lactic acid) chitosan-graft-poly(L-lactide) carboxymethylchitosan 4,4’-diazidostilbene-2,2’-disulfonic acid disodium salt dichloromethane deacetylation degree dimethylformamide 2,2-dimethoxy-2-phenylacetophenone dimethylsulphoxide differential scanning calorimetry glutaraldehyde 1,1,1,3,3,3-hexafluoro-2-propanol high-molecular-weight layer-by-layer low-molecular-weight entanglement molecular weight in melt poly(acrylic acid) polyacrylamide poly(allylamine hydrochloride) poly(2-acrylamido-2-methylpropanesulphonic acid) phosphate buffer solution polyelectrolyte complex

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 75 PEG PEG-g-Ch PEG-N,O-Ch PEG-N-Ch PEO PHEMA PLA PLDLA PLGA PLLA PP PVA PVP PЕТ QCh SQ SЕМ TEGDA TFA XRD ТHF ТЕМ

poly(ethylene glycol) poly(ethylene glycol)-graft-chitosan PEG-N,O-chitosan PEG-N-chitosan poly(ethylene oxide) poly(2-hydroxyethyl methacrylate) poly(lactic acid) poly(L-lactide-co-D,L-lactide) poly(lactide-co-glycolide) poly(L-lactide) polypropylene poly(vinyl alcohol) poly(vinyl pyrrolidone) poly(ethylene terephtalate) quaternized chitosan 7-iodo-8-hydroxyquinoline-5-sulphonic acid scanning electron microscopy triethylene glycol diacrylate trifluoroacetic acid X-ray diffraction analysis tetrahydrofuran transmission electron microscopy

1. INTRODUCTION In the recent years much interest is focused on nanofibrous materials. Because of their inherent large specific surface area and their small pore size nanofibrous materials may find a variety of applications, e.g. in military protective clothing and filter applications, in fuel cells, in drug delivery carriers, cosmetics, in nanosensors (thermal, piezoelectric, biochemical and fluorescence optical chemical sensors), or in electronics. Electrospinning is recognized as the most efficient for producing significant in length polymer fibers with diameters in the nanoscale range. Moreover, the transfer of an electrospinning technology from laboratory to industrial scale can be easily achieved. Special attention is given to the possibility of obtaining 3D-scaffolds for cell and tissue engineering, and wound healing dressings as well. Electrospinning of natural polymers, such as collagen, fibrinogen, chitosan and its derivatives, cellulose and its derivatives, is considered as a very promising method for the development of a new generation of fibrous materials for medical applications. It is expected that these materials should have surface structure and topology mimicking those of natural fibrous materials, e.g. of extracellular matrix. Such structure is propitious for the application of electrospun materials as cell and tissue engineering scaffolds. Amongst the natural polymers, chitosan is the most attracting for use in nanofibrous materials. The interest in preparation of chitosan-based polymer materials is due to the beneficial properties of chitosan in terms of its potential application in the biomedical field. Chitosan exhibits inherent antibacterial and haemostatic activity. Combining

D. Paneva, М. Ignatova, N. Manolova et al.

76

these properties with the potentialities of the electrospinning can enable the preparation of large variety of materials, e.g. of new generation wound healing dressings of high efficiency. It is worth to be noted that studies on the potential applications of fibrous materials containing chitosan or its derivatives are still in an early stage of development. In addition, solutions have to be found for obtaining composite materials combining chitosan or its derivatives with the widely used in medical practice aliphatic polyesters such as poly(lactic acid) and poly(εcaprolactone). In the present Chapter, after a brief description of basics of electrospinning, the preparation of nanofibrous materials containing chitosan or chitosan derivatives by electrospinning is surveyed. Special attention is focused on the newest trends in the development of electrospinning such as reactive electrospinning as well as polyelectrolyte complex formation during electrospinning, and on the possibilities of biomedical applications of electrospun materials.

2. ELECTROSPINNING Nanofibers can be prepared by different processing techniques such as: 1) template synthesis [1-3], 2) self-assembly [4,5], 3) phase separation [6,7], 4) drawing [8], 5) meltblowing [9,10] and 6) electrospinning [11-14]. Among them electrospinning stands out as the most promising route for fabrication of fibers having diameters within the micro- and the nanoscale and of length reaching tens of meters and more. Although the electrostatic spinning process has been discovered long time ago, electrospinning has gained much interest only by the end of the 20th century.

900 Number of publications

800 700 600 500 400 300 200 100 0 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 Years Figure 1. Number of scientific publications on electrospinning per year for the period 1999-2008 (source: Scopus®; Elsevier B.V).

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 77 The first patents for obtaining polymer fibers by means of electrostatic spinning belong to J. F. Cooley and W. J. Morton [15,16]. Electrospinning is an attractive technique because of its simplicity and the possibility of its facile transfer from laboratory to the industrial scale. The increase of the number of scientific publications on electrospinning in the period from 1999 to 2008 is an indication of the growing interest towards this method (Figure 1). There are USA and Singapore companies that manufacture electrospun non-woven textile [http://www. spinrati.com]. This Section deals in brief with the electrospinning principle and set-up, as well as with the main process parameters affecting the diameter and the morphology of the electrospun micro- and nanofibers.

2.1. Electrospinning Set-Up and Fundamentals The electrospinning process involves application of a high electric field to a polymer solution or melt. A scheme of an electrospinning set-up is shown in Figure 2. There are three basic components: a high voltage supply, a reservoir with a capillary tip for the spinning solution (or melt) and a metallic collector. The polymer solution is delivered through the capillary by means of an appropriate pump. One electrode lead of a high voltage power supply is immersed into the solution or connected to the capillary tip of the reservoir, and the other is connected to the collector. Applying high voltage (between 10 and 50 kV) on the solution induces electric charges. The mutual charge repulsion creates force acting oppositely to the surface tension. As the applied field strength (AFS) is increased, the hemispherical solution surface at the tip of the capillary deforms into a conical shape (Taylor cone) (Figure 2). When the AFS exceeds a threshold value, the repulsive electrostatic force overcomes the surface tension and the charged jet is ejected from the tip of the Taylor cone.

Syringe High voltage supply

Polymer solution

Jet

Collector

Figure 2. A schematic representation of an electrospinning set-up.

78

D. Paneva, М. Ignatova, N. Manolova et al.

The small jet diameter permits rapid mass exchange and the solvent usually evaporates during its traveling from the capillary to the collector acting as a counter electrode. As a result, charged polymer fiber is deposited on the collector. The fiber charges are gradually neutralized in the environment. The end-product of the process usually consists of randomly deposited fibers (mat) with diameters ranging from nanometers to micrometers. Under certain conditions the charged jet violates its continuity and instead of fibers nano- and/or microparticles of various forms are formed on the collector. This process is called electrospraying and is particularly appropriate for obtaining nano- and/or microparticles. The distance to the counter electrode usually varies between 10 and 25 cm. The substrates used so far for fiber deposition are either the collector of the electrospinning set-up or specific substrates chosen in dependence on the targeted application of the electrospun material. The basic knowledge accumulated in recent years on the jet behavior during electrospinning has been presented in a comprehensive review by Reneker and Yarin [17]. It has been found that jet flight time from the capillary tip to the collector is ca. 0.2 s. Initially the jet only follows a direct path towards the counter electrode. Then it becomes unstable performing a series of bending coils and jet’s radius is continuously enlarged. Investigations with the help of a high-speed digital video camera show that the jet is only one and it moves and bends very quickly (Figure 3).

Figure 3. Stereographic, stroboscopic picture, recorded during electrospinning with a digital video camera, that illustrated the bending paths of the jet (6% solution of poly(ethylene oxide) (PEO) with molar mass 400 000 g/mol, in a mixture of 75% water and 25% ethanol). Reproduced from Reneker and Yarin [17] by permission of Elsevier.

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 79

2.2. Processing Parameters Fiber formation by electrospinning is a complex process affected by a great number of parameters. They are generally divided into: i) solution parameters, ii) process variables, iii) ambient parameters. Solution parameters include polymer type (synthetic or natural, ionogenic or nonionogenic, linear or branched), polymer molar mass, solution characteristics (volatility, dielectric constant, viscosity), and absence/presence of a low-molecular-weight organic or inorganic salt, spinning solution conductivity, surface tension of the solution. Process variables include electric potential at the capillary tip, the gap (distance between the tip and the collector), feeding rate. Ambient parameters include air temperature and humidity. Knowledge and control of these parameters is of great importance for the successful preparation of micro- and nanofibers of polymers with desired morphology. The nature and the molar mass of the polymer hold great significance for the selection of the correct approach to prepare micro- or nanofibers by electrospinning. The successful electrospinning of non-ionogenic synthetic polymers can be performed using various solvents or solvent systems. In the case of ionogenic synthetic and natural polymers the choice of a solvent system is very limited. This imposes the necessity to search for innovative systems for their electrospinning. The solution viscosity is one of the key parameters that affect the formation and morphology of the fibers. Viscosity increase favors the formation of cylindrically shaped fibers of larger and more uniform diameters [18,19]. Vice-versa, at low solution viscosity, fibers having defects along their length are formed [20,21]. With the increase of viscosity the morphology of the obtained fibers is gradually changed: starting from fibers with bead-like defects, through fibers with spindle-like defects to defect-free fibers. SЕМ micrographs of fibers prepared by electrospinning of poly(L-lactide) (PLLA)/poly(ethylene glycol) (PEG) bicomponent solutions at three polymer concentrations are shown in Figure 4 [22]. There are a number of models describing the effect of the spinning solution concentration on the diameters and the morphology of the prepared fibers [18,19,23,24]. It has been propounded that the obtaining of continuous fibers is feasible when polymer chain entanglements occur in a solution at sufficiently high polymer concentration. The macromolecule entanglements allow the formation of an elastic network of long distance order, thus stabilizing the liquid jet and preventing its disintegration into individual drops.

A

B

C

Figure 4. SEM micrographs of PLLA/PEG fibers (weight ratio PLLA/PEG =70/30): (А) fibers with bead-like defects (concentration 5 wt. %); (B) fibers with spindle-like defects (concentration 7 wt. %) and (C) defect-free fibers (concentration 9 wt. %). Reproduced from Spasova et al. [22] by permission of SAGE.

80

D. Paneva, М. Ignatova, N. Manolova et al.

In the case of solutions of low concentration and/or by using polymers of low molar masses, the number of macromolecule entanglements is not sufficient and the jet disperses into nano- or microbeads (electrospraying). Models are developed on the basis of the Huggins equation describing the concentration dependence of viscosity for homogenous solutions of a linear polymer: ηsp(c)=[η]c + kH([η]c)2 +

(1)

where ηsp(c) is the specific viscosity, [η] is the intrinsic viscosity, с is the polymer concentration and кн is the Huggins coefficient. The dimensionless value obtained by the product of the intrinsic viscosity and the concentration ([η]c), is referred to as the Berry number (Be) [25]. This number is a measure of chain overlap in solution. In case of a very dilute solution of a polymer in a good solvent, the polymer molecules are too remote from each other, they rarely come in contact and the Berry number in such case is less than unity. At higher polymer concentration the individual macromolecules interact and entangle; in such case the Berry number is greater than unity. Linear homopolymers of poly(methyl metacrylate) have been used for more detailed study on the effect of the solution viscosity (in a good solvent, DMF, 25°С) on fiber formation during electrospinning [23]. It has been shown that during electrospinning of dilute poly(methyl metacrylate) solutions (at с/с*<1, where c is the concentration, c* is the critical chain overlap concentration) only droplets are obtained. Continuous fiber formation, but still with some defects, has been observed at с/с*∼3.9-4.0. Defect-free nanofibers have been obtained at с/с*>6 for all polymers with narrow molar mass distribution (Mw/Mn ∼ 1.03-1.35), however for polymers having relatively broader molar mass distribution (Mw/Mn ∼ 1.62 and ∼2.12), defect-free fibers are formed at much higher concentrations, viz. с/с*∼9.7 and 10.1, respectively. The authors have also determined the dependence of fiber diameter on concentration [d ∼(с/с*)3.1], as well as of fiber diameter on zero shear solution viscosity [d ∼(ηo)0.71]. These scaling relationships are in agreement with the results obtained by McKee et al. [18,26] in the case of electrospinning of poly(ethyleneterephthalate-co-ethylene isophthalate) solutions in chloroform/DMF (70/30 v/v). A semi-empirical method of predicting the transition from electrospraying to electrospinning in a good solvent at concentrations c>c* has been suggested. The entanglement number of the macromolecules in the solution (ne)soln is defined as the ratio between weight-average molar mass of the polymer ( M W ) and the entanglement molecular weight in solution (Me)soln:

(2) where Me is entanglement molecular weight in melt, φp is polymer volume fraction in solution [19]. This analysis shows that fiber formation is initiated at (ne)soln∼2, while the critical polymer concentration necessary for continuous fiber formation (defect-free fibers)

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 81 corresponds to (ne)soln ≥ 3.5. The validity of this method has been confirmed for polystyrene, poly(lactic acid) (PLA), PEO and poly(vinyl pyrrolidone) (PVP) solutions. It should be pointed out that this method is valid only for polymer solutions in good solvents where polymer-polymer interaction can be neglected. As for systems where it is possible that interaction between polymer chains takes place (based on hydrogen bonding, ion interactions, liquid-liquid or solid-liquid phase separation, hydrophobic interactions), intermolecular interactions stabilize physical entanglements. In such cases the concentration at which defectfree nanofibers are obtained is lower than the one calculated on the basis of the semiempirical method of chain entanglement [24,27-29]. The presence of a low-molecular-weight organic or inorganic salt in the spinning solution affects to a high extent the average diameters of the prepared fibers. On increasing the solution conductivity, the density of the jet charges increases. Stronger elongation forces are imposed to the jet due to the repulsion of the excess charges along the jet length thus resulting in formation of fibers with smaller diameters. For instance the addition of an inorganic salt such as KH2PO4, NaH2PO4, and NaCl to the spinning solutions leads to obtaining defect-free fibers from poly(D,L-lactic acid) with smaller diameters [30]. The addition of 0.8 wt. % of the ionogenic organic compound pyridinium formate leads to an increase in the electrical conductivity of dichloromethane (DCM) spinning solutions and to obtaining defect-free PLLA fibers with smaller average diameters [31]. A substantial decrease of fiber average diameter has been observed during electrospinning of PVP in the presence of 4,4’-diazidostilbene-2,2’-disulfonic acid disodium salt (DAS) [32]. The addition of a lowmolecular-weight salt has been applied for preparing nanofibers by electrospinning of poly[(2-dimethylamino)ethyl methacrylate] aqueous solution [33]. The addition of an organic or inorganic salt to the spinning solution [34] may lead to obtaining micro- and nanofibrous bundles (see also Section 6.2). The surface tension of spinning solutions also affects the morphology of the prepared fiber. It has been shown that defect fibers are obtained from solutions with higher surface tension. As mentioned before, in order to form a spinning jet, the surface tension of the solution must be overcome by the electric voltage. As with solution electrical conductivity, the use of additives, even in low concentrations, may affect the surface tension of polymer solutions. Surfactants may play a decisive role in obtaining defect-free fibers. Nanofibers of polystyrene electrospun from DMF/ТHF mixtures (5-15% w/v) yield defect-free nanofibrous mats only when 0.03-30 mmol/l of the cationic surfactant dodecyl trimethyl ammonium bromide is added to the solution [35]. An important feature in electrospinning is the rapid evaporation of the solvent leading to thinning of the jet. That is why the volatility of the solvent has an impact on the morphology of the obtained fibers. The vapor pressure of the solvent affects the degree and rapidity of its evaporation. Although THF is a good solvent for a number of polymers, and moreover is of high volatility, working with it causes difficulties due to blockage of the capillary tip. In electrospinning of poly(vinyl chloride) from THF, a very broad distribution of fiber diameters is obtained, whereas poly(vinyl chloride) electrospinning from DMF results in a narrow fiber diameter distribution and in an average fiber diameter of about 200 nm. Electrospinning of poly(vinyl chloride) from THF/DMF mixtures leads to decrease of the average fiber diameter with the increase of the DMF amount [36]. Very rapid evaporation of the solvent may therefore impede the formation of fibers with smaller average diameters. Depending on the type of the system, fibers of smaller average diameter can be obtained on increasing the

82

D. Paneva, М. Ignatova, N. Manolova et al.

solvent vapor pressure. For example, it has been reported that when polystyrene solutions in various solvents are electrospun, the average fiber diameter decreases with increasing boiling point of the solvent [37]. The solvent vapor pressure also influences the fiber surface features, such as porosity. When applying solvents with rapid evaporation such as DCM (vapor pressure 46.6 kPa) PLLA fibers with pores along the fiber length are obtained [38]. The process variables affecting fiber diameter and morphology are: (1) applied electric potential, (2) spinning solution feeding rate and (3) the gap (distance between the tip and the collector). Many authors use AFS, which represents the relation between electric potential and the distance between the capillary tip and the collector. AFS is measured in kV/cm. Applying voltage upon the spinning solution induces electric charges. The increase of the applied voltage causes an increase of the electrostatic forces of repulsion between the individual charges acting on the liquid jet [39-41]. This leads to obtaining fibers with smaller diameters. A decrease in the average fiber diameter with the increase of the applied voltage has been observed in electrospinning of numerous polymers: polyacrylonitrile in DMF [42], polystyrene in THF [43], poly(vinyl alcohol) (PVA) in water [44], chitosan/PEO in a dilute acetic acid solution [45], quaternized chitosan (QCh)/PVA in water [28], DNA in ethanol [46]. There are several reports suggesting that applied voltage has no significant effect on the average fiber diameter. This has been observed during electrospinning of hydroxypropyl cellulose in ethanol or propanol [47], of poly(D,L-lactic acid) in chloroform/acetone (2/1 v/v) [48], of PVA in water [49] and of N-carboxyethylchitosan (CECh)/polyacrylamide (PAAm) in water, of poly(2-acrylamido-2-methylpropanesulphonic acid) (PAMPS)/PVA in water and of P(AMPS-co-acrylic acid)/PVA in water [50]. It has also been reported that average polystyrene fiber diameter increases from 0.31 to 1.72 μm with the increase of applied voltage from 5 to 25 kV in electrospinning of a 12.5-22.5 % w/v polystyrene solution in chloroform [51]. For the electrospinning of 17 wt. % aqueous solutions of PVP or of PVPiodine complex, the average fiber diameters increases from 580 to 640 nm for the PVP system and from 150 to 225 nm for the PVP-iodine system, on increasing the AFS from 0.8 to 1.7 kV/cm [32]. Such discrepancy in experimental observations is probably due to differences in the feeding rate of spinning solution, the gap distances, or the concentrations used in the different studies. Another process variable affecting morphology and diameter of the obtained fibers is the gap distance. Decreasing the gap distance shortens jet flight time, thus decreasing the degree of jet elongation as well as the time available for evaporation of the solvent. Baker et al. [51] demonstrate that when electrospinning a polystyrene solution in chloroform (17.5 wt %) on increasing the gap distance between the capillary tip and the collector from 5 to 25 cm (voltage 15 kV) a decrease of the nanofiber average diameter from 1 to 0.66 μm is observed. Hong et al. [52] have studied the effect of the gap distance on the morphology of PVA nanofibers. It has been shown that as the gap distance decreases, drying of the nanofibers is not sufficiently efficient and they are fused when deposited on the collector. Feeding rate of the spinning solution is another parameter with impact on the fiber diameter and morphology. At low solution feeding rate the electrospinning can be intermittent with the Taylor’s cone being depleted. At increase of solution feeding rate either an increase in the fiber diameter [53] or bead-like defects formation is observed. The effect of solution feeding rate on nanofiber morphology has been discussed in a number of publications [5456].

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 83 Ambient parameters (air temperature and humidity) also affect the fiber morphology. The effect of temperature on the morphology of polyuretanecarbamide fibers obtained by electrospinning of their solutions in DMF has been studied. At higher temperature (70°С) fibers of narrower diameter distribution are obtained, compared to diameter distribution of fibers obtained at room temperature [57]. Average diameter of PVP or cellulose acetate nanofibers [58] is the largest at 20°С, and the fibers obtained at 10°С and at 30°С are with smaller average diameters. This particular temperature impact is explained by the authors by two parameters affected by temperature. The first one is evaporation rate of the solvent decreasing exponentially by the decrease of temperature; consequently, it takes a longer time for the jet to solidify. The second parameter is polymer chain rigidity. At higher temperatures the polymer chains have more freedom of movement, leading to a decrease of the solution viscosity, to a higher jet stretching rate and to thinner fibers. At temperature of 10°С the effect of the first parameter probably predominates over the second one due to the exponential variation of the solvent evaporation rate with temperature. At higher temperature of 30°С the effect of the second parameter is predominant due to the exponential viscosity decrease when temperature increases. The majority of reported data on electrospinning of polymer solution are collected from experiments performed in air. Kim et al. [59] have studied the impact of air relative humidity on the diameter of polystyrene electrospun fibers. It has been shown that with the increase of relative humidity from 10% to 70%, the average diameter of polystyrene electrospun fibers increased from 130 to 380 nm. It has been suggested that the formation of thicker fibers at higher relative humidity is probably due to the fact that the electrostatic charges on the surface of polymer solution are easily discharged at higher humidity level, causing reducing of repulsion forces.

2.3. Morphology and Alignment of Electrospun Fibers Nanofibers prepared by electrospinning are usually monolithic with cylindrical shapes (Figure 5 A) [11,60]. However, more often some deviations are observed, such as obtaining ribbon-like fibers (“nanoribbons”) [61-63], fibers with bead-like defects (Figure 5 B) and spindle-like defects (Figure 5 C) [20,21,63]. Beside the described defects, it is possible to obtain also porous fibers (Figure 6) [64-67]. Under a regime of electrospraying, instead of fibers, beads and structures with various shapes non-connected with the fibers can be formed. In low-viscosity spinning solutions polygonal spheres and other mushroom-like structures can be formed [68], as well as “buns” or “cups” [69,70]. Criteria for the complex evaluation of the morphology and alignment of electrospun fibers have been systematized [63]. The main characteristics needed to evaluate the morphology of defect-free fibers comprise: average fiber diameter, minimal and maximal fiber diameter, standard deviation of the diameter and fiber diameter distribution. The main parameters to be used to characterize the occurring defects (bead-like or spindle-like) along the fiber length have been pointed out. The most frequently used electrospinning set-up permits obtaining mats formed by randomly deposited fibers.

D. Paneva, М. Ignatova, N. Manolova et al.

84

А

B

C

Figure 5. A sketch of: (А) a defect-free fiber; (B) fiber with bead-like defects and (C) fiber with spindle-like defects. Reproduced from Spasova et al. [63] by permission of SAGE.

Figure 6. SEM micrographs of porous fibers of PLLA, electrospun from DCM at concentration 7 wt. %. Spasova et al. [66], Electrospun chitosan-coated fibers of poly(L-lactide) and poly(Llactide)/poly(ethylene glycol): preparation and characterization, Macromol. Biosci., 2008, 8, 153-162, Copyright Wiley-VCH Verlag GmbH and Co. KGaA. Reproduced with permission.

When applying collectors of special geometry [12,71-73] aligned fibers may be obtained. A schematic representation of some collectors used so far, are shown in Figure 7. One of the collectors used is a rotating cylindrical collector (Figure 7 А and B) [71,74-76] or a rotating thin disc (Figure 7 C) [77-79]. High effectiveness in fiber deposition is achieved when such collectors are used. The low degree of fiber alignment is a disadvantage of this collector type. In order to achieve good fiber alignment static collectors of parallel electrode type are applied (Figure 7 D and E), as well as static collectors with blades placed in line (Figure 7 F) with conductive strips separated by a gap of different width (larger than several centimeters)

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 85 [63,72,80-82]. Often there are difficulties in getting highly aligned nanofibers over a large area of substantial thickness. Fiber alignment is disturbed on increasing the thickness of the material deposited on the collector. The productivity of electrospinning process can be increased by simultaneous use of multiple spinnerets (multiple nozzles) (Figure 7 G) [83].

A

B

C

D

E

F

G Figure 7. Schematic representation of various electrospinning set-ups to obtain aligned fibers (A-F) and for multiple spinnerets (G). Reprinted from IOP, Teo, W., Ramakrishna S. [12], A review on electrospinning design and nanofibre assemblies, Nanotechnology, 2006, 17, 89-106, with permission from IOP.

Multiple spinnerets have also been used to prepare bicomponent and multicomponent blend nanofibrous mats [84,85]. Reported in literature appliances to electrospinning set-ups for obtaining nanofibrous yarns have been discussed in detail in review articles [12,73].

D. Paneva, М. Ignatova, N. Manolova et al.

86

2.4. Applications Electrospinning offers diverse possibilities for the design of fibrous materials with targeted composition and morphology. It also allows one-step incorporation of additives of different nature into micro- and nanofibers, such as drugs, metallic nanoparticles, carbon materials (carbon nanotubes, fullerenes). New materials that can find application in many different fields may be fabricated using electrospinning. So far electrospinning has been used to obtain new-generation filters for highly efficient gas and liquid filtration [86]; to prepare high-sensitivity nanosensors [87]; in the design of clothing of new generation - extremely light in weight and with desired degree of permeability [88]; to obtain new hybrid fibrous materials combining the useful properties of inorganic nanosized materials and polymer fibers [89]; they are very promising for the development of highly effective heterogenic catalytic systems, also for the needs of optoelectronics; to obtain new materials with application in pharmacy and medicine, for instance as carriers of low-molecular-weight bioactive substances, as cell- and tissue engineering scaffolds, as wound healing dressings [90]. A number of reviews and books have appeared in recent years [90-98], discussing the possibilities of applying the electrospinning process for preparation of new nanomaterials for biomedical applications. In the next Sections the main trends of obtaining nanofibrous materials containing the natural polysaccharide chitosan or its derivatives by electrospinning have been outlined. Results obtained so far on some potential biomedical applications of electrospun fibrous materials are summarized.

3. CHITOSAN – A VERSATILE POLYMER Chitin is a polysaccharide which is the second in abundance after cellulose. It is a basic structural component in the crustaceans’ exoskeleton, in insects’ shells, as well as in the cell walls of some fungi and bacteria [99,100]. Its annual biosynthesis is estimated to approximately 1010 tons [101]. Chitin is a linear polymer with a polymer chain built of poly[β(1→4)-2-acetamido-2-deoxy-D-glucopyranose] (Figure 8A). It is a structural analogue of cellulose however differing from it by the acetamido [-NH(C=O)CH3] group at the C2 carbon atom. Chitosan is obtained by chitin deacetylation and can be regarded as a copolymer built of β(1→4)-bound 2-acetamido-2-deoxy-D-glucopyranose and 2-amino-2-deoxy-Dglucopyranose units (Figure 8B). Chitosan is a basic structural element of some fungi cell walls (Zygomycetes, Aspergillus and Fusarium) [102]. In industrial production of chitosan, chitin deacetylation is usually performed in alkali medium [103]. In recent years biosynthesis methods and the enzymatic deacetylation of chitin are considered as alternative methods of obtaining chitosan. Depending on the chitin type and the hydrolysis conditions chitosan of different molar masses and deacetylation degrees (DDA) are obtained. The molar mass of chitosan may exceed 1000 kDa; the DDA of most of the commercially available products varies from 75% to 95% [103-105].

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 87 OH

NHCOCH3

HO

O

O

NHCOCH3

NHCOCH3

OH

NH2

HO

O

O

HO

OH (A)

OH

O

O O

HO

O

O O

HO

OH

O

NH2

OH

HO

O NHCOCH3

(B) Figure 8. Chain sequence of chitin (A) and of chitosan (B).

It should be noted that chitosan is a natural polymer having behavior of weak polybase (pKa=6.5 [106]) in aqueous solutions, and its solubility highly depends on the medium рН [107]. It is soluble in aqueous medium with рН<6.5 and insoluble in neutral and basic medium. Chitosan is soluble in aqueous solutions of most organic acids (acetic, lactic, hydrochloric, formic, trifluoroacetic (TFA)) [108-110] as well as in some dilute mineral acids (HCl, HBr, HI, HNO3 and HClO4) [111]. Chitosan solubility depends on its DDA. Chitosan is sparingly soluble in organic solvents, however, a significant amount of polar solvent such as ethanol can be added to its solutions in acetic acid [112]. Chitosan combines a number of beneficial biological properties determining the possibilities of its application in medicine, pharmacy, agriculture, food industry, and biotechnology. The properties of chitosan, the possibilities for preparation of a number of its derivatives by chemical and enzymatic modification, as well as some aspects of the application of chitosan and its derivatives in biomedicine and in other fields of practice have been discussed in a number of reviews [101,105,113-116], in e-reference tools [117] and in books [118-122]. Chitosan is non-toxic and biodegradable, it possesses inherent biological activity, it is biocompatible in respect to animal and plant organs, tissues and cells, and it can be chemically or enzymatically modified to derivatives having desired properties. Chitosan degrades under the action of some enzymes such as chitinase, chitosanase, cellulase, glucanase, lipase, some proteases, аmilase, papain, pectinase, etc. [123-126]. In living organisms lysozyme is the main enzyme responsible for chitosan degradation [127]. The enzymatic hydrolysis of chitosan has also been studied in the presence of water-soluble synthetic polymers incorporated in interpenetrating or semi-interpenetrating polymer networks, under the action of enzyme complexes of the beneficial soil microorganisms Trichoderma viride and Bacillus subtilis [128-131]. At present enzymatic hydrolysis is one of the most applied methods for obtaining chitosan oligomers. The possibility of obtaining and applying chitosan oligomers is of special interest because of the complex of valuable properties they possess, namely antitumor effect [132], immunoenhancing effect [133], ability to enhance the plants defense reactions against diseases [134], antifungal [135] and antimicrobial activity [136]. Chitosan ability to degrade in biological medium under the

88

D. Paneva, М. Ignatova, N. Manolova et al.

action of different enzymes produced by microorganisms determines the use of this natural polymer for the design of polymer food packaging materials, which after use can degrade in the environment [137]. The good thermal stability of chitosan enables dry sterilization of chitosan-based materials in the temperature range 110 - 130°С, without causing any essential deterioration of their physico-chemical properties [138]. This feature of chitosan is of great importance for the application of chitosan-based polymer materials in the food industry. Chitosan contains functional groups capable of chelating metal ions; thus it can be effectively used for waste water treatment in respect to purification from metal ions [139]. One of the most beneficial properties of chitosan and of its quaternized or O-carboxylated derivatives is their inherent biocide activity. Chitosan and these derivatives inhibit the growth of a number of pathogenic microorganisms – Gram-positive and Gram-negative bacteria, yeasts and fungi [101,135,140-144]. The antimicrobial activity of chitosan in respect to a certain pathogenic microorganism depends on DDA, on the molar mass of the chitosan and on the medium рН [145]. Several mechanisms have been suggested to explain the antimicrobial activity of chitosan. One of them suggests that the electrostatic interaction occurring between the protonated amino groups of chitosan and the negatively charged moieties (lipopolysaccharides and proteins) of the cell membranes is responsible for the disruption of the integrity of the microorganism cell membrane, loss of cell organelles and cell lysis [146]. Another mechanism is based on the assumption that chitosan oligomer fractions penetrate into the microorganism cells and prevent their growth by impeding the information transfer from DNA to RNA [147]. The biological activity renders chitosan and its derivatives as promising candidates for the design of new wound healing dressings. In addition, chitosan and N,N–dicarboxymethylchitosan show a substantial acceleration effect in terms of wound healing [148,149]. Chitosan-based materials offer favorable conditions for cell adhesion and proliferation, and their structure provides high degree of immune recognition. This natural polysaccharide actively takes part in bone regeneration processes by ensuring cell adhesion [150]. Chitosan has also been studied as a nerve guide, as it is shown to be capable of facilitating adhesion, differentiation and proliferation of nerve cells [151]. Besides these properties chitosan also exhibits haemostatic activity– it accelerates blood coagulation [138,152,153]. In order to obtain chitosan-based materials with improved blood compatibility, an innovative approach of modifying chitosan with various polyacids and their copolymers has been developed [154]. Comparison of the structure of chitosan and heparin (the natural anticoagulant in mammalians) suggests that introducing sulfo- and carboxyl groups in the chitosan structure by means of chemical modification might impart anticoagulant properties to chitosan. It has been found that sulfo-derivatives of chitosan possess anticoagulant activity [155-157]. Carboxymethylation at С-6 is another route of obtaining heparin-like anticoagulants [158] Chitosan and some of its derivatives possess antioxidant activity and thus may contribute to slowing down the progress of a number of chronic diseases related to processes involving free radicals [101]. Due to its ability to decrease blood cholesterol level chitosan is applied as an active nutrition additive for the preparation of dietary foods for people with cardiac diseases [101]. Chitosan possesses three types of reactive groups: primary NH2-groups at С(2) carbon atom and primary and secondary ОН-groups, at С(6) and at С(3) carbon atoms correspondingly. This enables diverse chitosan modification reactions to be carried out aiming at preparation of derivatives having desired properties. Chitosan derivatives have been

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 89 obtained by acylation, N-phtaloylation, tosylation, sulfonation, quaternization [144,159], PEGylation, alkylation, Schiff base formation, O-, N,O- and N-carboxymethylation [160,161], N-carboxyethylation [162], etc. Graft polymerization of various synthetic monomers on chitosan is another possibility of its chemical modification aimed at design of new hybrid materials, combining the properties of the main polysaccharide chain with those of the graft synthetic chains. Different monomers are used for this purpose, such as: acrylonitrile [163], acrylic and methacrylic acid [164,165], vinylpyrrolidone [166], vinylacetate [167,168], acrylamide [169], methylmetacrylate [170,171], methylacrylate [172,173], 2-acrylamido-2-methylpropanesulfonic acid [174], lactic acid [175, 176], etc. Chitosan can be crosslinked to form hydrogels by using various crosslinking agents, e.g. epichlorohydrine [177,178], ethyleneglycoldiglycidyl ether [179,180], glyoxal [181], glutaraldehyde (GА) [182-185], genipin [186, 187], etc. From a technological viewpoint, chitosan solubility in dilute acids determines its processability into films, beads, gels, granules, tablets, microporous scaffolds and microfibers (~50 μm) [113,154,188-192]. The preparation of chitosan-containing nanofibers by means of electrospinning is of particular interest. The next Section treats the research performed up to date on obtaining electrospun fibers. As a polybase, chitosan forms polyelectrolyte complexes (PECs) with various natural and synthetic polyacids such as carboxymethylcellulose [193], alginic acid [194], poly(acrylic acid) (PAA) [195,196], poly(methacrylic acid) [197], PAMPS [198], poly(AMPS-co-AA) [199], hyaluronic acid [200], etc. The possibilities of obtaining nano- and microfibrous materials by means of electrospinning of PEC from chitosan and polyacids are discussed further in the present Chapter. The polyelectrolyte complexes and the materials prepared thereof can find a great variety of applications such as ultrafiltration membranes [201], semipermeable membranes for separation of water-organic mixtures [202], materials for wound healing devices [203], materials with improved blood compatibility [154], for incorporation and controlled release of drugs [204,205], etc. The discussed physicochemical and biological behavior of chitosan determines its great potential as a highly promising polymer that can be used in medicine, pharmacy, agriculture, food industry and biotechnology.

4. CHITOSAN-CONTAINING NANOFIBERS The dissolution of chitosan in aqueous solutions is accompanied by protonation of its amino groups and in aqueous solutions it behaves as a polycation. Chitosan forms solutions having high viscosity, electrical conductivity and surface tension in dilute acetic acid or in dilute or concentrated formic, lactic and propionic acid. These properties of chitosan solutions are often pointed out as the most important ones in terms of the difficulties that accompany the electrospinning of this natural polymer [11,45,95,206-210]. The preparation of continuous defect-free fibers from non-ionogenic polymers requires the polymer concentration to be from 2- to 2.5-fold higher than the entanglement concentration (ce) [24,26].

90

D. Paneva, М. Ignatova, N. Manolova et al.

Figure 9. Plot of specific viscosity (ηsp) vs. concentration for 148 000 g/mol chitosan in 80% acetic acid in water. The entanglement concentration (ce) is 2.9 wt. % and is determined by the change in slope (scaling exponent) on the above log-log plot. Reproduced from Klossner et al. [211] by permission of ACS.

Fibers from aqueous solutions of synthetic ionogenic polymers can be prepared at concentrations 8-fold higher than ce [33]. Klossner et al. [211] discuss the possibility chitosan solutions in 80 % acetic acid to be electrospun and determine ce value (Figure 9). At chitosan concentration equal to ce, the electrospinning leads to the formation of micro- and nanoparticles, and no fibers can be obtained. At chitosan molar mass of 148 000 g/mol, the determined ce value is 2.9 wt. %. It is to be expected that electrospinning should take place at concentrations about 7 wt. %. The viscosity of these solutions, however, can reach very high values (up to 16 000 cP) and at these polymer concentrations the electric field cannot overcome the combined effect of viscosity and solution surface tension. In solutions with chitosan concentration equal to 2×ce, Taylor cone cannot be formed, and at lower concentrations electrospraying takes place and no fibers are formed because of insufficient number of entanglements of the polymer chains. Chitosan has been successfully electrospun in 2004. Two main approaches were applied: a) electrospinning in the presence of an electrospinable water-soluble non-ionogenic polymer (able also to form water-soluble polymer-polymer complexes with chitosan based on hydrogen bonds), PEO and PVА were used as partners [45,206,207]; and b) electrospinning of chitosan without any polymer partner using TFA as a solvent [207]. In this Section these two approaches are discussed in details. The biomedical application of chitosan nanofibrous materials often necessitates their insolubility in body fluids. Two-step strategies for imparting water-insolubility to chitosan-containing non-woven textile are summarized. The known up to date approaches for combination of the beneficial properties of aliphatic and aromatic polyesters and those of chitosan by electrospinning are discussed as well.

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 91

4.1. Electrospinning of Chitosan in the Presence of Water-Soluble Synthetic Polymers As it has been mentioned, the first successful experiments on preparation of chitosancontaining fibers have been performed by electrospinning of aqueous solutions of the natural polymer in dilute acetic acid in the presence of the water-soluble non-ionogenic polymers PEO or PVA [45,206,207]. PEO and PVA have been selected as polymer partners due to several reasons. Both polymers have flexible linear chains and are easily electrospun from aqueous solutions. PEO and PVA can form hydrogen bonds with chitosan [212,213]; thus disrupting the hydrogen bonds between chitosan macrochains. The presence of intermolecular interactions between chitosan and the non-ionogenic polymers based on hydrogen bonds has been evidenced by using diverse methods, such as IR-spectroscopy, differential scanning calorimetry (DSC) and X-ray diffraction analysis (XRD) [45,206,214-217]. In chitosan/PEO mixture PEO acts as a plasticizer facilitating orientation and flow of chitosan macromolecules [45]. This allows the formation of a stable jet and, subsequently, of nanofibers. SEM micrographs of fibers prepared by electrospinning of chitosan in the presence of PEO using 2 wt. % acetic acid as a solvent at different weight ratios between the partners are shown in Figure 10. PEO of molar mass of 800 000 g/mol has been used as a partner in this case. The nanofibers prepared at chitosan/PEO = 0.05 - 0.33 (w/w) and total polymer concentration 5 wt. % are cylindrical in shape with average diameters from 200 to 250 nm. Fibers of smaller diameters are obtained at higher chitosan content. The nanofibers prepared at weight ratio 1/1 and total polymer concentration 3 wt. % have much smaller average diameters (40 nm). However, bead-like defects are observed along the fiber axis. Further increase of chitosan content [chitosan/PEO = 4 (w/w)] does not allow obtaining of fibers; only “tailed” beads are formed. The proposed approach for electrospinning of chitosan in the presence of a nonionogenic partner has been used for preparation of a non-woven textile from this natural polymer. The main characteristics of spinning solutions as well as of the fibers prepared applying this approach are summarized in Table 1. The macromolecular characteristics of chitosan and of the non-ionogenic polymer are presented in the Table since they are of essential importance in terms of the rheological parameters of the spinning solutions, hence of the morphology and the average diameters of the nanofibers. Moreover, for chitosan the method for its obtaining, respectively its molar mass and DDA, affect significantly its behavior in solutions. That is why the suppliers of the corresponding chitosan used in the studies are listed in the Table as well. As seen from Table 1, until now electrospinning of chitosan has been performed using PEO with molar masses ranging from 600 000 to 5 000 000 g/mol. When applying PEO of molar mass below 1 000 000 g/mol and dilute acetic acid solution the fiber formation is possible at weight ratio chitosan/PEO not higher than 1/1. The replacement of part of the dilute acetic acid by dimethylsulphoxide (DMSO) and the decrease of the surface tension by addition of surfactant results in preparation of continuous fibers at chitosan/PEO = 90/10 (w/w) [218]. The use of PEO of molar mass > 5 000 000 g/mol enables the occurrence of sufficient number of chain entanglements by adding a small amount of a non-ionogenic polymer, and continuous fibers can be obtained at chitosan/PEO = 90/10 (w/w) [216]. Solvent mixture of formic acid (HCOOH)/water for electrospinning of chitosan in the presence of PVA has been used [207]. As seen from Тable 1, the used up to date PVA is of

92

D. Paneva, М. Ignatova, N. Manolova et al.

molar mass between 75 000 and 186 000 g/mol. In order to obtain nanofibers enriched in chitosan in this case it is necessary to modulate the main parameters of the spinning solution.

A

B

C

D Figure 10. SEM micrographs of nanofibers electrospun from mixed solutions of chitosan/PEO prepared at AFS 1 kV/cm at weight ratio chitosan/PEO: 0.18; ×1 000 (A), 0.33; ×1 000 (B) (total polymer concentration of 5 wt. %); 1.0; ×5 000 (C), 4.0; ×10 000 (D) (total polymer concentration of 3 wt. %). Reprinted from e-Polymers, Spasova M. et. al., 2004, no. 056 [45], with permission from e-Polymers Foundation.

Table 1. Optimal conditions found for preparation of chitosan-containing nanofibers by electrospinning in the presence of a nonionogenic polymer (the corresponding reference, macromolecular characteristics of chitosan and non-ionogenic polymer as well as the supplier of chitosan are listed in the Table; generally the voltage does not exceed 30 kV, and the gap between the nozzle and the collector is less than 20 cm) Chitosan (molar mass; DDA; supplier)

PEO 600 000 g/mol 80 % Sigma 654 000 g/mol 90 % Qingdao Hisound Biological Engineering (China) 276 000 g/mol 81.7 % Vanson HaloSource (Redmond, WA, USA) 190 000 g/mol 85 % Sigma

85 % Sigma

Non-ionogenic polymer (molar mass)

Solvent system

Chitosan/ Non-ionogenic polymer [w/w]; Total polymer concentration

Mean fiber diameter [nm]

Defects [nm]

Reference

PEO 800 000 g/mol

aq. CH3COOH (2 wt.%)

up to 1/1 3 wt.%

40

Beads-like defects 750/410

[45]

PEO 600 000 g/mol 1 500 000 g/mol 2 300 000 g/mol 4 000 000 g/mol PEO 5 000 000 g/mol

aq. CH3COOH (2 wt.%)

up to 2/1 2–8 wt.%

124±19

-

[206]

Solution formulation Kamterter II, LLC, Lincoln, NE, USA

No data available

-

[223]

PEO 900 000 g/mol

aq. CH3COOH (0.5 M); Triton X-100TM (0.3 wt.%) DMSO (10 %) aq. CH3COOH (3 wt.%) + DMSO CH3COOH/DMSO = 10/1 (w/w)

Nominal concentration of chitosan – 1 wt.% up to 90/10 2.05 wt.%

40

-

[218]

114±19

-

[216]

PEO UHMWPEO > 5 000 000 g/mol

95/5 3 wt. %

Table 1. (Continued) Chitosan (molar mass; DDA; supplier)

PEO 1 400 000 g/mol 87 %; 70 %; 67% Primex Inc. 100 000 g/mol 83 % Sigma 85 % Shen Chiu 90 % Pharmaceutical-grade chitosan was obtained from the Naval Research Laboratory (Washington, DC) 600 000 g/mol; 400 000 g/mol; 148 000 g/mol 75 - 85 % Fluka 1 000 000 g/mol 80 % Primex Inc

Non-ionogenic polymer (molar mass)

Solvent system

Chitosan/ Non-ionogenic polymer [w/w]; Total polymer concentration

Mean fiber diameter [nm]

Defects [nm]

Reference

PEO 900 000 g/mol 300 000 g/mol

aq. CH3COOH

95/5 1.33 wt.% (70 ºC)

80±35

-

[224]

PEO 3 000 000 ~ 5 000 000 g/mol

aq. CH3COOH (1 wt.%)

40/60

ca. 30

Spindle-like defects no data on the sizes

[225]

PEO 600 000 g/mol

aq. CH3COOH (2 wt.%) + CH3OH (4 wt. %)

1/1 3 wt.%

40±9.14

Spindle-like defects no data on the sizes

[226]

PEO 900 000 g/mol

aq. CH3COOH (32 wt.%)

8/9 3.4 wt.%

60±9

-

[211]

PEO 900 000 g/mol

aq. CH3COOH (90 wt.%) + surfactant

3/1 1.6 wt.%

140

-

[227]

Chitosan (molar mass; DDA; supplier)

PVA 210 000 g/mol 78 % 1 300 000 g/mol 77 % Wako Pure Chemical Industries, Ltd., Japan 1 600 000 g/mol 82.5 % Aldrich Low molecular weight 75 – 85 % Aldrich 120 000 g/mol 82.5 % Zhejiang GoldenShell Biochemical Co., Ltd. (Taizhou, China) 78 % Sichuan Biochem-ZX Research Co., Ltd., China 165 000 g/mol 90 % Zhejing Yuhuan Ocean Biochemistry, China

Non-ionogenic polymer (molar mass)

PVA 88 000 g/mol

PVA 124 000 – 186 000 g/mol 87-89% hydrolysed PVA 146 000 – 186 000 g/mol 98-99% hydrolysed PVA 154 000 g/mol 88 % hydrolysed PVA 94 000 g/mol 96 % degree of hydrolysis PVA 80 000 g/mol 98 % degree of hydrolysis

Solvent system

Chitosan/ Non-ionogenic polymer [w/w]; Total polymer concentration

Mean fiber diameter [nm]

Defects [nm]

Reference

HCOOH/H2O

1/1

120

-

[207]

aq. CH3COOH (2 wt.%)

up to 25 % chitosan 6 wt.%

20±5

Spindle like defects no data on the sizes

[219]

aq. CH3COOH (2 wt.%)

5-8 wt.% PVA solution containing 1 wt.% chitosan

160±38

-

[228]

aq. acrylic acid (90 %)

up to 95/5

290

-

[229]

aq. CH3COOH

up to 30/70 7.4 wt.%

125

Spindle-like defects no data on the sizes

[214]

aq. CH3COOH (2 wt.%)

40/60 7%

100±21

Beads-like defects no data on the sizes

[215]

Table 1. (Continued) Chitosan (molar mass; DDA; supplier)

PVA 100 000 g/mol 88 % Zhejiang Golden-Shell Biochemical Co. (Yuhuan, Zhejiang, China) 200 000 g/mol 88 % Zhejiang Golden-Shell Biochemical Co. (Yuhuan, Zhejiang, China) > 10 000 g/mol 100 % water-soluble Hittolife Co. (Kyongki-Do, Korea) Low-viscosity chitosan High-viscosity chitosan Samsung Chitopia Co., Ltd. (Siheung, Korea) PAAm 1 400 000 g/mol (HMW) 80 % Primex 100 000 g/mol (LMW) 70-80 % Sigma

Non-ionogenic polymer (molar mass)

Solvent system

Chitosan/ Non-ionogenic polymer [w/w]; Total polymer concentration

Mean fiber diameter [nm]

Defects [nm]

Reference

260

-

[221]

PVA 170 000 g/mol 88 % degree of hydrolysis

aq. CH3COOH (90 wt.%) + H2O

80/20 7 wt. % chitosan/ acetic acid (90 wt. %) and 10 wt. % PVA aqueous solution

PVA 75 000 g/mol 88 % degree of hydrolysis

aq. CH3COOH (88 wt.%) + hydroxyapatite nanoparticles

90/10 7.3 wt.%

100 - 700

-

[222]

PVA 74 000 g/mol 99.9 % degree of hydrolysis

double distilled water

up to 4/6 12.5 wt.%

~ 200

Beads-like defects

[217]

PVA 44 000 g/mol

HCOOH/H2O AgNO3 or TiO2 nanoparticles

up to 15/85 chitosan was dissolved in formic acid (5 wt %)

290-360

Spindle-like defects no data on the sizes

[220]

aq. CH3COOH (50 wt.%) 25, 40 and 70 ºC

up to 75/25 (25 ºC) up to 90/10 (70 ºC) 1.4 wt.%

50 – 350 depending on chitosan molar mass and spinning temperature

Increasing the temperature the formation of beads-like defects decreases

[230]

PAAm 5 000 000 g/mol

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 97 Until now the studies have aimed at searching for an appropriate solvent that enables the electrospinning of chitosan/PVA system at chitosan/PVA weight ratio higher than 1. The use of dilute acetic acid allows nanofibers from chitosan/PVA to be formed at weight ratio between the polymer partners up to 40/60 [214,215,219]. The preparation of chitosan/PVA fibers at weight ratio 85/15 in HCOOH/H2O solvent has been reported [220]. Using 90 % aqueous solution of acetic acid enables the successful preparation of chitosan/PVA fibers at weight ratio of 80/20 and 90/10 [221,222]. It has been found that after immersion in NaOH aq. solution bath only fibers composed of chitosan/PEO= 90/10 (w/w) preserve their integrity [218]. The electrospinning of chitosan oligomers in the presence of PVA has been reported very recently [217]. Continuous and defect-free fibers with average diameter of ca. 200 nm are obtained using aqueous solutions at total polymer concentration 12.5 wt. % and at weight ratio oligomers/PVA = 40/60. The increase of PVA content enhances the tensile strength of the fibers. Zhou et al. [229] apply acrylic acid for preparation of chitosan/PVA mixed solutions. In this system the electrical conductivity is higher in dilute acrylic acid and at higher chitosan concentration. While continuous defect-free fibers up to chitosan/PVA = 60/40 (w/w) are prepared from 4 % aq. solution of acrylic acid, fibers of higher chitosan content (up to 90/10) are obtained when 90 % acrylic acid is used. Membranes of poly(lactide-co-glycolide) (PLGA), chitosan and PVA nanofibers with average diameter of ca. 300 nm have been obtained by using an electrospinning set-up with two spinnerets, allowing simultaneous electrospinning of a solution containing PLGA and a chitosan/PVA solution (weight ratio 60/40 in 2 wt.% acetic acid) [231].

Figure 11. Fiber diameter (FD, left) and bead density (BD, right) of 1.4 wt. % high molecular weight chitosan/high molecular weight PAAm blend fibers at different air temperature. (Error bars represent standard deviation (n = 60 for FD, and n = 3 for BD), letters indicate significant difference at p < 0.05). Reproduced from Desai et al. [230] by permission of Elsevier.

PAAm is another water-soluble non-ionogenic polymer able to form hydrogen bonds with chitosan. PААm has been applied to facilitate the electrospinning of the chitosan derivative CECh [50]. These studies are discussed in Section 5. Later PAAm has been used for the preparation of chitosan-containing nanofibrous materials [231] from solutions enriched in chitosan [chitosan/PAAm = 95/5 (w/w)]. The effect of the spinning solution temperature (25 ºС, 41 ºС and 70 ºС) on the fibers morphology at different weight ratios

98

D. Paneva, М. Ignatova, N. Manolova et al.

chitosan/PAAm has been evaluated (Figure 11). The temperature increase results in preparation of fibers with significantly smaller number of defects however, with higher average diameters. Phase separation occurs in fibers electrospun from chitosan/PEO = 1/1 (w/w) solutions [206]. While the fibers with larger diameters consist mainly of PEO, those of smaller diameters are enriched in chitosan. For chitosan/PVA pair at chitosan/PVA = 17/83 (w/w), Li et al. have also shown a phase separation between chitosan and the non-ionogenic polymer [219]. This fact has been used to obtain nanoporous fibers by removal of PVA by treatment with NaOH aq. solution. A ТЕМ micrograph of the porous chitosan fiber thus obtained is presented on Figure 12. The fibers consist of chitosan as determined gravimetrically as well as by DSC and IR-spectroscopy. At high chitosan content [chitosan/PEO = 90/10 and 95/5 (w/w)] no phase separation occurs and the fiber composition is almost the same as the feed one, i.e. the polymer partners are homogenously distributed in the fibers [216,230].

Figure 12. TEM of NaOH treated (1 M, 12 h) 17/83 chitosan/PVA bicomponent fibers. Reproduced from Li et al. [219] by permission of Elsevier.

The crystallization of the polymers during electrospinning is hampered since the high rate of transition of the liquid jet into dry fibers impedes crystallites formation [30,232] for example in the case of nanofibers electrospun from chitosan/PVA [214] or chitosan/PEO [45] mixed solutions. Recently, the approach of electrospinning of chitosan in the presence of a non-ionogenic polymer has been applied for preparation of hybrid nanofibers containing hydroxyapatite nanoparticles [221,222], titanium dioxide or silver nitrate [220], as well as silver nanoparticles [233]. The incorporation of bioactive inorganic substances gains significant attention because of the possibility additional beneficial properties to be imparted to the electrospun non-woven textile. For instance, nanosized hydroxyapatite aids the bone cells proliferation; thus becoming a beneficial component in the design of scaffolds, as well as of implants for bone tissue regeneration. The preparation of hybrid nanofibrous materials that

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 99 combine chitosan biodegradability and biocompatibility with hydroxyapatite activity to accelerate the bone cell proliferation is of outstanding importance for the development of a new generation scaffolds for cell and tissue engineering. Ag and TiO2 nanoparticles possess a broad spectrum antimicrobial activity. It is found that hybrid nanoparticles consisting of chitosan and Ag nanoparticles display synergistic effect [234]. Thus, the preparation of hybrid nanosized materials from chitosan and Ag nanoparticles is very promising for the design of novel wound healing dressings [234-237]. The electrospinning of a dispersion containing chitosan/PVA/hydroxyapatite nanoparticles in 88 wt. % acetic acid has been demonstrated [222]. However, the fibers obtained are with broad diameter distribution and their values vary from 100 to 700 nm. Besides, the loading with hydroxyapatite in an amount higher than 10 wt. % is hampered because of nanoparticles agglomeration. The growth of the hydroxyapatite mineral phase can take place onto the electrospun scaffolds. In this case the hydroxyapatite growth and its uniform distribution are supported by the presence of carboxyl groups in the polymer matrix and in the incubation solution. Thus, PAA has been added into the mineralizing bath [221] and better results in respect to the attachment and proliferation of mouse fibroblasts on the surface of CECh/PVA scaffolds as compared to chitosan/PVA scaffolds have been obtained. The preparation of hybrid chitosan/PVA nanofibers, containing AgNO3 or TiO2 nanoparticles has been reported [220]. They use formic acid as a solvent. It is difficult to assess the usefulness of these materials since no data have been provided in terms of crosslinking of at least one of the polymer partners. In addition, no data on weight loss or the fiber morphology after their stay in aqueous solution are presented. The concentrated HCOOH is able to reduce silver ions to Ag nanoparticles [238]. This property of formic acid has been used for preparation of hybrid chitosan/PEO or CECh/PEO nanofibers containing in situ synthesized Ag nanoparticles [233]. At polymer weight ratio of 1/1 cylindrical defect-free fibers with average diameter of 100±29 nm have been obtained with Ag nanoparticles content being 10 wt. % in respect to the total polymer weight. It is worth to be noted that the use of HCOOH as a solvent for electrospinning of chitosan and CECh has an additional beneficial feature. It allows the crosslinking of the polysaccharide under the action of a crosslinking agent to be slowed down, giving opportunity the crosslinking and electrospinning to be performed simultaneously for a long period of time. The obtained nanofibers are waterinsoluble. The one-step approaches for imparting water-insolubility of chitosan-containing nanofibers are discussed in Section 6 of the Chapter. Very recently, the preparation of nanofibers containing surfactant micelles by electrospinning of chitosan/PEO blend solutions has been reported [227]. The used spinning solution consists of chitosan and PEO at weight ratio equal to 3/1 (w/w) and the surfactant concentration is chosen in such a manner so as to exceed the critical micelle one. It is claimed that the micelles can serve as carriers of lipophilic substances, such as lipophilic drugs. Bicomponent fibers consisted of chitosan core and PEO sheath have been prepared [239]. For this purpose two spinnerets, one placed in the other, ending with coaxially positioned needles have been used. The inner spinneret contains chitosan solution, and the outer one – PEO solution. Both solutions form a bicomponent pendant drop at the capillary end, resulting in the formation of a bicomponent Taylor cone. It is claimed that the mixing of the two polymers is limited due to the rapid solvent evaporation.

100

D. Paneva, М. Ignatova, N. Manolova et al.

4.2. Electrospinning of Chitosan Until now a number of studies have been performed on the electrospinning of chitosan alone. As already mentioned, the natural polymer cannot be electrospun alone from dilute or concentrated solutions of formic acid, lactic acid and propionic acid. [207,208,210]. There are data on electrospinning of chitosan from its solutions in concentrated (90 %) acetic acid [70,208]. Here the known up to date on the electrospinning of bare chitosan will be presented. Chitosan (molar mass of 210 000 g/mol) can be electrospun alone using TFA as a solvent [207,240]. The fibers prepared at chitosan concentration equal or less than 6 wt. % have defects along their axis. At a polymer concentration of 7 wt. % the fibers obtained are mainly with average diameter of 490 nm. However, beads-like defects are observed again along their axis. In order to obtain defect-free fibers a mixture of TFA/DCM has been used as a solvent. The optimal chitosan molar mass and chitosan concentrations for preparation of defect-free fibers from chitosan/TFA system depending on chitosan molar mass are summarized in Тable 2. Table 2. Concentration of chitosan above which continuous defect-free fibers are prepared using chitosan solutions in TFA depending on the molar mass of chitosan, according to Ohkawa et al. [240] Chitosan molar mass [g/mol] 210 000 1 310 000 1 580 000 1 800 000

Chitosan concentration [wt.%] 8.00 4.25 3.25 2.00

Avarage fiber diameter [nm] 200±24 103±16 83±11 60±22

TFA is the preferred solvent for electrospinning of bare chitosan in subsequently reported studies [210,241-243]. The use of solvents with lower boiling point and lower surface tension as compared to water facilitates the electrospinning of chitosan. According to Torres-Giner et al. [210] the mixed solvent TFA/CH2Cl2 (70:30 v/v) is the most suitable one for electrospinning of chitosan alone. The tendency of defect-free fibers formation is enhanced increasing chitosan molar mass. The combination of the good physico-mechanical properties and degradability of poly(εcaprolactone) and (co)copolymers of PLA with chitosan antibacterial activity is a promising strategy for obtaining new nanofibrous materials suitable for the design of novel wound healing dressings. The use of TFA as a solvent enables the preparation of composite fibers from chitosan or its derivatives and water-insoluble polyesters such as poly(ethylene terephtalate) (PЕТ) [244], (co)polymers of PLA [245,246] and poly(ε-caprolactone) [247]. Chitosan/PET nanofibers with average diameters from 500 to 800 nm have been electrospun from their mixed solution in TFA [244]. The chitosan content in the materials imparts hydrophilicity and enhances their inhibition activity against pathogenic organisms S. aureus and Klebsiella pneumoniae. Successfully novel composite nanofibrous materials have been prepared by electrospinning of mixed solutions of chitosan and poly(L-lactide-co-D,L-lactide) (PLDLA) in the common solvent TFA/CH2Cl2 [70/30 (v/v)] [246].

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 101

A

B Figure 13. SEM micrographs of mats that have been incubated in S. aureus cell culture (107 cells/mL) for 24 h at 37 °C: PLDLA mat (A), and crosslinked chitosan/PLDLA mat [50/50 (w/w)] (B); magnification: × 2500. Ignatova et al. [246], Electrospun non-woven nanofibrous hybrid mats based on chitosan and PLA for wound-dressing applications, Macromol. Biosci., 2009, 9, 102-111, Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.

Continuous, defect-free and cylindrical chitosan/PLDLA nanofibers with average diameter value of 840 nm are obtained. In order to impart stability of the bicomponent electrospun chitosan/PLDLA nanofibers in aqueous solutions, the protonated by TFA amino groups of chitosan are neutralized under ammonia vapors, and after that the nanofibers are crosslinked with GА vapors. The microbiological screening reveals that the novel chitosan/PLDLA materials are effective to prevent adhesion and to suppress the growth of the Gram-positive bacteria S. aureus and the Gram-negative bacteria E. coli (Figure 13). The fluorinated solvent 1,1,1,3,3,3-hexafluoro-2-propanol (HFIP) is also suitable in terms of electrospinning of chitosan. Composite fibers of diameters 300-500 nm have been prepared from chitosan and poly(ε-caprolactone) by electrospinning of their solution in TFA and HFIP at weight ratios poly(ε-caprolactone)/chitosan equal to 40/60 [247]. Chitosan fibers have been obtained by electrospinning of chitin precursor in HFIP solvent and subsequent deacetylation of the obtained defect-free fibers in a 40 % NaOH aq. solution [248]. No significant changes in the fiber morphology and their average diameters as a result of the deacetylation have been observed.

102

D. Paneva, М. Ignatova, N. Manolova et al.

Chitosan fibers with average diameter of 130 nm have been prepared by electrospinning of chitosan with molar mass of 106 000 g/mol and DDA of 54 % in 90 % acetic acid [70]. The feasibility of electrospinning when using concentrated acetic acid is attributed to the fact that the surface tension of chitosan solution decreases from 54.6 dyn/cm when 10 % acetic acid solution is used, to 31.5 dyn/cm when 90 % acetic acid is used. Successful preparation of defect-free nanofibers from chitosan with DDA 75 – 85 % from its solutions in 90 % acetic acid has been reported [208]. Attempts to electrospin fibers from chitosan with molar mass of 148 000 g/mol and DDA of 80 % applying the approach of Geng et al. [70] and of De Vrieze et al. [208] have been made [211]. The inability to obtain fibers has been explained by the differences in the DDA values of the used chitosan. It is worth to be noted that the three research groups have performed their studies using chitosan purchased by different suppliers. This is an additional indication that the physico-chemical characteristics of chitosan have a remarkable impact on its behavior in terms of the electrospinning process. The chitosan nanofibrous materials electrospun in absence/presence of a non-ionogenic polymer are soluble in body fluids. Thus, the development of suitable and easily feasible approaches to render them water-insoluble is of essential importance. One of the routes for imparting water-insolubility of chitosan-containing nanofibrous materials is their stay in alkaline aqueous medium [218,219,243] or their treatment with ammonia vapors [246] where the protonated amino groups are converted into non-protonated ones and chitosan turns into its insoluble in neutral and alkaline medium form. The possibility of imparting waterinsolubility to chitosan fibers obtained by electrospinning from solutions in TFA by immersion of the mats in NaOH or Na2CO3 aq. solutions has been studied [243]. Neutralizing by 5 М NaOH aq. solution has led to partial preservation of the fiber structure of the nonwoven textile. The use of Na2CO3 instead of NaOH leads to materials that preserve their fibrous structure even after a long (12 weeks) stay in phosphate buffer solution (рН = 7.4) or in distilled water. Crosslinking of chitosan nanofibers under GA vapors has been applied [241] and fibers that preserve their morphology after contact with aq. medium have been obtained. This route for imparting water-insolubility to chitosan fibrous materials is applied both for chitosan fibers [66,221,222,226,231,233,246], and for fibers from chitosan derivatives (particular examples are given in Section 5 of the Chapter). In this case chitosan crosslinking occurs at a second stage after the electrospinning. One-step preparation of chitosan fibers crosslinked with GA has been reported as well, and is discussed in details in Section 6 of the Chapter. Genipin is a very attractive crosslinking agent of chitosan (Figure 6) [249,250]. This is a natural product obtained from geniposide by means of enzyme hydrolysis under the action of β-glucosidase. Geniposide is isolated from the fruits of Genipa americana (South America) and Gardenia jasminoides Ellis (Asia), where its content is in the range 4-6 %. Genipin has found its application as a crosslinking agent of natural polymers having primary amino groups, such as chitosan and proteins (serum albumin, gelatin, fibrinogen) [251]. Because of its natural origin genipin is a preferred agent for obtaining crosslinked natural polymers for the design of new polymer devices for biomedical applications. It has been reported that two reactions that proceed at a different rate are responsible for chitosan crosslinking. The more rapid reaction is a nucleophilic attack on genipin in position С3 (Figure 14) by the primary chitosan amino groups leading to the formation of a genipin heterocyclic compound linked to the glucosamine residue of chitosan. The second, the slower reaction, is a nucleophilic

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 103 substitution of the ester group at С11 (Figure 14), accompanied by release of methanol and formation of amide bond between genipin and chitosan. Dinan et al. [247] have used genipin for crosslinking of electrospun nanofibrous materials from chitosan and poly(ε-caprolactone). For this purpose the electrospun materials has been kept in 1 wt. % of genipin aq. solution for 24 h. 11

COOCH3

6 7

5 8

HOH2C

10

9

4 1

3

O

OH

Figure 14. Genipin formula.

Table 3. Reported data on the biocompatibility of electrospun chitosan-containing nanofibrous non-woven textiles in respect to tissue cells Chitosan-containing electrospun non-woven textile Chitosan/PEO Chitosan/PEO PLGA-chitosan/PVA membranes Chitosan/poly(ε-caprolactone) Chitosan/PVA/HA Chitosan Chitosan/PET Chitosan/PEO

Cell Type

Reference

canine chondrocytes chondrocytes (HTB-94) osteoblasts (MG-63) human embryo skin fibroblasts (hESFs) Schwann cells mouse fibroblasts (L929) Schwann cells NIH 313 fibroblasts osteosarcoma cells (MG-63)

[223] [218] [231] [247] [222] [243] [244] [225]

The combination of the biological activity of chitosan and the high surface-to-volume ratio of the electrospun materials is an excellent prerequisite for preparation of new generation materials that can find application as wound healing dressings, as well as in the cell and tissue engineering [210,220,244,246]. The ability of chitosan-containing nanofibers to serve as drug carriers has also been demonstrated [45]. Owing to the appropriate behavior of chitosan in respect to attachment, proliferation and viability of tissue cells, it is regarded as very suitable for obtaining scaffolds for cell and tissue engineering [252;253]. It is known that chitosan can stimulate proliferation of cells such as chondrocytes, osteoblasts, fibroblasts. This is the reason most of the studies on the biological activity of chitosan-containing nanofibrous materials to be performed using these type of cells (Таble 3). The obtained up to date results are highly encouraging and reveal the great potential of these new materials as scaffolds for cell and tissue engineering. It is worth to be noted that the

104

D. Paneva, М. Ignatova, N. Manolova et al.

obtained results are preliminary ones and thorough research on the possibility chitosan nanofibrous materials to be used for such kind of applications has to be performed.

4.3. Chitosan-Coated Nanofibers As evident from the data presented above, in recent years there is a growing interest towards combining the good physico-mechanical properties and biocompatibility of polyesters such as (co)polymers of lactides, poly(ε-caprolactone) and PET, with chitosan antibacterial activity [231,244-247]. This has been achieved by using an electrospinning setup with two spinnerets for simultaneous electrospinning of a polyester solution and a chitosan/PVA solution [231] or by electrospinning of a common solution of polyester and chitosan using TFA as a solvent or mixture of TFA and CH2Cl2 or HFIP [244-247]. Recently, an original and elegant two-step approach for combining the beneficial properties of polyesters and chitosan has been used [66]. The new strategy applied for preparation of the composite fibrous materials is based on the formation of a thin chitosan coating on electrospun PLLA and PLLA/PEG fibers. It is known that proteins and pathogenic microorganism cells adhere onto PLLA based materials due to the hydrophobic nature of the latter [254]. This fact somewhat limits the potentialities of these materials as implants and wound healing dressings. A widely used approach to diversify their application is their physical or chemical modification with PEG. The polymer products thus obtained are characterized by a higher hydrophilicity which is a prerequisite for decreasing the undesired cells and proteins adhesion. The incorporation of PEG in PLLA fibers during the electrospinning process allows hydrophilicity to the obtained materials to be imparted [22]. It is expected that significantly less blood cells and pathogenic microorganisms would adhere onto PLLA/PEG fibers. These materials are not able to inhibit pathogenic microorganisms’ growth and incorporation of appropriate drugs into them is needed. Based on the knowledge on chitosan inherent bactericidal activity, it has been suggested that the formation of a thin chitosan film on PLLA and PLLA/PEG fibers would lead to obtaining new composite materials [66]. Such materials should combine the physico-mechanical properties of the polyester fibrous materials and the antibacterial and haemostatic activity of chitosan. As mentioned in Section 3 chitosan degrades under the action of enzymes such as lysozyme which is found in the blood serum [255]. Under the action of these enzymes, after a certain period of contact with blood the integrity of the chitosan coating will be disrupted. It has been demonstrated that PLLA and PLLA/PEG electrospun materials are appropriate scaffolds for cell proliferation and tissue regeneration [22,256,257]. That is why it is expected that after the degradation of chitosan coating the polyester-containing mats would serve as a scaffold for regeneration of the injured tissue. Applying the described strategy studies have been performed on the possibility of obtaining hybrid fibers from PLLA, PEG and chitosan as well as on their behavior in contact with blood cells and with the pathogenic microorganism S. aureus [66]. A triple chitosan coating of ca. 20 ± 2 nm thickness has been obtained by immersion of PLLA or PLLA/PEG non-woven textile in 0.05 wt. % chitosan solution. The coating has been crosslinked by treatment with GA vapors. In order to evaluate the interaction of the obtained hybrid materials with blood cells the uncoated and the coated with crosslinked chitosan mats from PLLA and PLLA/PEG have been put in contact with whole human blood

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 105 for an hour. Figure 15 shows SEM micrographs of fibrous materials after contact with whole blood.

A

B

C

D

Figure 15. SEM micrographs of PLLA mat (A and B) and PLLA/PEG (C and D) after 1 h contact with whole blood: pristine (A and C) and triple-coated with chitosan (crosslinked) (B and D). Spasova et al. [66], Electrospun chitosan-coated fibers of poly(L-lactide) and poly(L-lactide)/poly(ethylene glycol): preparation and characterization, Macromol. Biosci., 2008, 8, 153-162, Copyright Wiley-VCH Verlag GmbH and Co. KGaA. Reproduced with permission.

PLLA mats show no haemostatic activity. Only single erythrocytes with preserved morphology are observed on the surface of PLLA mats (Figure 15А). Most probably their presence is due to mechanical attachment to the highly porous structure of the mat. High haemostatic activity is displayed by mats with triple chitosan coating. In this case agglutination, deformation and aggregation of the erythrocytes are observed (Figure 15B). Figures 15C and D show SEM micrographs of uncoated and triple-coated with crosslinked chitosan PLLA/PEG fibrous mats after contact with blood. The presence of 30 wt. % PEG in the mats affects the interaction of chitosan coated materials with the blood cells. On the surface of PLLA/PEG mats single erythrocytes can be detected (Figure 15C); the blood cells adhered onto the surface of the coated mats preserve their specific morphology and are not deformed (Figure 15D). The erythrocytes number is less than 5 cells per 1 000 μm2 on triplecoated mat is 55.

D. Paneva, М. Ignatova, N. Manolova et al.

106

A

B

C

D

Figure 16. SEM micrographs of S. aureus cells adhered into PLLA: pristine (A) and triple-coated with chitosan (crosslinked) (B); and onto PLLA/PEG (70/30): pristine (C) and triple-coated with chitosan (crosslinked) (D). Spasova et al. [66], Electrospun chitosan-coated fibers of poly(L-lactide) and poly(Llactide)/poly(ethylene glycol): preparation and characterization, Macromol. Biosci., 2008, 8, 153-162, Copyright Wiley-VCH Verlag GmbH and Co. KGaA. Reproduced with permission.

Compared to a bare PLLA mat, the erythrocyte number on the triple-coated with chitosan mat is approx. twofold higher - 130. This is an indication that PEG reduces the blood cell adhesion. In contact with the pathogenic microorganism S. aureus, a significant number of cells (more than 200 cells/100 μm2) adhere onto the PLLA mat surface (Figure 16). The increased hydrophilicity of PEG-containing fibers leads to some decrease of the number of adhered cells, and chitosan coating on the fibers leads to a substantial decrease of the number of the adhered cells (ca. 5 cells/100 μm2). The inhibition of the pathogenic cells adhesion combined with the haemostatic activity of the chitosan coated PLLA and PLLA/PEG mats render them very promising materials for application as wound healing dressings.

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 107

5. CHITOSAN DERIVATIVES-CONTAINING NANOFIBERS As mentioned above, electrospinning of chitosan is feasible only from a concentrated acid solutions – TFA [207, 210, 241-243] and acetic acid [70,208] or from dilute acid solutions in the presence of a non-ionogenic water-soluble polymer with sufficiently high molar mass [45,206,218,219]. A number of chitosan derivatives are water-soluble at рН ≤ 7. Such are those obtained by carboxymethylation [258] or carboxyethylation [162], sulfonation [259] and quaternization [144]. Solubility in organic solvents is easily attained by acylation of chitosan [260]. The chitosan derivatives obtained by PEGylation [261] are soluble both in water and in organic solvents. The synthesis of chitosan derivatives soluble in water or in easily volatile low-toxic organic solvents may allow the preparation of fibers from acid-free solutions. Thus, the production of nanofibrous materials from chitosan derivatives will be rendered friendly for the environment, and it will offer additional advantages for the biological and biomedical applications of the nanofibers. In this Section the existing up to now studies aimed at preparation of chitosan derivatives - containing micro- and nanofibers will be surveyed. The approaches applied to stabilize chitosan derivatives-containing fibers against dissolving in water and their potential biomedical applications will also be discussed.

5.1. Electrospinning of Chitosan Derivatives in the Presence of Synthetic Polymers QCh derivatives [144], carboxymethylchitosan (CMCh) [160, 258, 262] and CECh [162] (Figure 17) are easily synthesized and purified. Unlike chitosan which is only soluble in acidic medium with рН<6.5 these chitosan derivatives are water-soluble also at рН>6.5. Their biological properties are similar to those of chitosan – they are non-toxic, biocompatible and biodegradable polymers and they can find application in medicine and pharmacy [144,162,263-268]. QCh derivatives have shown higher antibacterial activity, broader spectrum of activity, and higher killing rate as compared to those of chitosan [144,265,266], and thus are potential candidates for design of biologically active wound dressings of new generation that actively take part in the wound healing process. Carboxyalkylated chitosan derivatives attract attention because of their antioxidant and anti-tumor effects. It has been reported that the biodegradability of these water-soluble chitosan derivatives is even higher than that of chitosan [117]. No fibers are obtained in the case of electrospinning of concentrated aqueous solutions of the polyelectrolyte QCh and of the polyampholyte-polyzwitterions CMCh and CECh [28,32,50,82,269,270]. This might be due to the fact that the repulsive forces between ionogenic groups within polymer backbone interfere with the formation of an elastic network of polymer chains of long distance order and impede the formation of continuous fibers. Continuous QCh-containing fibers are formed only when mixed aqueous solutions of QCh and a non-ionogenic water-soluble polymer – PVA or PVP are electrospun [28,32]. It has been shown that the found optimal conditions for electrospinning of QCh/PVA system are: total polymer concentration 8 wt. % and AFS values from 1.5 to 3.5 kV/cm, and for QCh/PVP system - total polymer concentration 20 wt. % and AFS values from 1.6 to 2.8 kV/cm.The average QCh/PVA fiber diameters are in the range of 60 - 200 nm [28]. The effect

D. Paneva, М. Ignatova, N. Manolova et al.

108

of the composition and properties of spinning solutions and of AFS on the morphology and the average diameters of the prepared fibers has been studied. With the increase of the ionogenic polymer QCh content a decrease of the average diameter of the nanofibers, as well as an increase in the number of spindle-like defects is observed. This change in the nanofiber morphology might be explained with the increase in the charge density and with increase of the solution conductivity on increasing the polyelectrolyte content. It has been shown that cylindrically shaped and defect-free nanofibers are formed by electrospinning of mixed aqueous solutions at weight ratio of QCh/PVA = 1/4 (Figure 18А). Spindle-like defects along the fibers are observed at weight ratio of QCh/PVA (w/w) = 2/3, 1/1 and 3/2 (Figure 18B). In all these cases (ne)soln for PVA is significantly lower than 3.5, estimated by the semi-empirical method, proposed by Shenoy et al. [19,24]. Nanofiber feasibility under such conditions is explained with the ability of PVA solutions to undergo physical gelation due to the presence of inter- and intramolecular hydrogen bonding. Similar effect has also been observed by other authors in cases of electrospinning of aqueous PVA solutions [19,62]. At weight ratio of QCh/PVA = 4/1 and higher only non-connected with the fibers “tailed” beads are observed. R4 R R3 + 3_ N I

R1O O

OR2 O R1O

O

R1O O

O

NHR O

NHR

OR2

OR2

R = H(14%), C(O)CH3 (20%); R1 = H (87%), CH3 (13%); R2 = H(12%), CH3 (88%); R3 = CH3 and R4 = C4H9 (66%)

quaternized chitosan (QCh)

OCH2COOH O HO

OH

OH O

O

NHCH2COOH

HO

NH2

N,O-carboxymethylchitosan (CMCh)

HO

HO

O

NHCH2CH2COOH

N-carboxyethylchitosan (CECh)

OCH2COOH O

O

O

OH O

O

NH2

O-carboxymethylchitosan (CMCh)

HO

O

NHCH2COOH

N-carboxymethylchitosan (CMCh)

Figure 17. Chemical structures of quaternized chitosan derivatives, of carboxymethylchitosan and of carboxyethylchitosan.

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 109

A

B Figure 18. SEM micrographs of the nanofibers prepared from QCh/PVA solutions. Weight ratios QCh/PVA = 1/4 (A) and 1/1 (B). Total polymer concentration 8 wt. %, AFS 2.0 kV/cm, magnification: ×5000. Reproduced from Ignatova et al. [28] by permission of Elsevier.

It has been shown that electrospinning of QCh and PVP aqueous solutions at weight ratios of QCh/PVP (w/w) from 1/4 to 4/1 and total polymer concentration of 20 wt. % results in defect-free and cylindrically-shaped fibers with average diameters in the range from 1500 to 2800 nm (Figure 19) [32]. Both in the QCh/PVA system, and in this case, on increasing the ionogenic component (QCh) content, the fiber diameter significantly decreases and the fiber diameter distribution narrows. The observed changes have been attributed to the increased solution conductivity. The AFS changes have different effect on the fiber morphology for the two systems - QCh/PVA and QCh/PVP. In the case of QCh/PVA, a decrease in the average diameters is observed on increasing the AFS. Conversely, in the case of QCh/PVP, the increase in AFS leads to an increase in the average diameter of the nanofibers. The reasons for this difference remain unknown so far. In order to obtain bicomponent fibers from QCh and PLDLA an appropriate solvent system has been proposed, consisting of DMF/DMSO in volume ratio 60/40, which allows the mixed solutions to be obtained and successfully electrospun [246]. The continuous defectfree hybrid QCh/PLDLA nanofibers (average diameter 280 nm, Figure 20A) are randomly deposited when collected onto a stationary target.

110

D. Paneva, М. Ignatova, N. Manolova et al.

A

B

Figure 19. SEM micrographs of the fibers prepared from QCh/PVP solutions. Weight ratios QCh/PVP = 2/3 (A) and 4/1 (B). Total polymer concentration 20 wt. %, AFS 2.2 kV/cm, magnification: ×1000. Reproduced from Ignatova et al. [32] by permission of Elsevier.

Partially aligned defect-free nanofibers from the same polymers are formed by electrospinning of QCh/PLDLA solution onto a rotating collection drum (Figure 20B).

A

B

Figure 20. SEM micrographs of the fibers from QCh/PLDLA = 30/70 w/w, collected onto stationary aluminum plate (A) or onto rotating drum (B). Total polymer concentration 5 wt.-% (DMF/DMSO = 60/40 v/v), AFS 1.4 kV·cm-1 and feeding rate of 1.3 mL·h-1. Ignatova et al. [246], Electrospun nonwoven nanofibrous hybrid mats based on chitosan and PLA for wound-dressing applications, Macromol. Biosci., 2009, 9, 102-111, Copyright Wiley-VCH Verlag GmbH and Co. KGaA. Reproduced with permission.

Attempts to electrospin concentrated aqueous solutions of CECh have proved to be unsuccessful. Formation of micro- and/or nanoparticles is observed depending on the АFS values [50,82]. CECh containing nanofibers can be prepared from its aqueous solutions in the presence of non-ionogenic water-soluble polymers with flexible chains, such as PААm [50] and PVA [50, 82]. CECh/PААm nanofibers obtained at different weight ratios of both polymers, are cylindrically-shaped and with average diameters in the range from 50 to 215 nm [50]. For the CECh/PААm system the composition of the spinning solution has a significant effect on the nanofiber morphology and the average fiber diameter. The decrease of the solution viscosity and the increase of the solution conductivity at higher CECh content favor the formation of nanofibers with smaller average diameter. Continuous defect-free

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 111 nanofibers (Figure 21А) are obtained by electrospinning of solutions with CECh content less than 50%. At CECh content 50% spindle-like defects along the nanofibers are formed (Figure 21С). With the increase of the CECh content the number of defects increases and the average distance between two neighboring defects along the nanofiber is decreased (Figure 21С). The formation of “tailed” beads non-connected with the fibers is observed in the case of nine-fold weight excess of CECh.

A

B

C

Figure 21. Effect of the composition of the spinning solution on the morphology of CECh/PAAm nanofibers. Weight ratio CECh/PAAm=1/4 (A), 1/1 (B) and 4/1 (C). Total polymer concentration 3%, AFS 1.1kV/cm. Reproduced from Mincheva et al. [50] by permission of SAGE.

The effect of the composition of spinning solution and the AFS on the fiber morphology and the average fiber diameter has also been studied in the case of the CECh/PVА system [82]. It has been shown that the average diameters of nanofibers decrease, and the shape of the PVАrichest fibers (weight ratio CECh/PVА = 1/75) changes from ribbon-like (Figure 22А) to cylindrical when the AFS value increases [82]. The ratio of polymer partners substantially affects the shape and the average diameters of CECh/PVA nanofibers [82]. The nanofibers of higher CECh content (weight ratio CECh/PVА = 1/8; 1/4; 1/3; 1/2 and 1/1) are cylindrical (Figure 22В). On increasing the CECh content in spinning solutions (at AFS = 1.6 kV/cm) the average nanofiber diameter decreases – e.g., from 420 nm to 100 nm in the case of CECh/PVA = 1/75 and 1/2, respectively. Both with the CECh/PААm system and in this case, on increasing the CECh content in spinning solutions, an increase of the amount of spindlelike defects is observed, as well as a decrease of the distance between two defects and narrowing of the fiber diameter distribution.

А

B

C

Figure 22. Effect of the composition of the spinning solution on the morphology of CECh/PVA nanofibers. Weight ratio CECh/PVA = 1/75 (A), 1/8 (B) and 1/2 (C); total polymer concentration 9.5%, AFS 1.2 kV/cm. Reproduced from Mincheva et al. [82] by permission of Elsevier.

In the same publication different types of collectors have been used (Figure 23) in order to obtain CECh/PVA nanofibers aligned in one or two directions [82]. Nanofibers, aligned in

D. Paneva, М. Ignatova, N. Manolova et al.

112

one direction, have been obtained by using collector of the А1 type, consisting of two conductive strips. The preparation of transversely aligned nanofibers in two directions has been achieved by the consecutive connection of the conductive strips’ pairs 1а – 1b and 2а – 2b in collectors of А2 and А3 types (Figure 23), separated by an insulating material (polypropylene - PP, quartz – SiO2 or cardboard - C).

Figure 23. Schematic representation of the collectors used for aligned CECh/PVA nanofiber preparation. Reproduced from Mincheva et al. [82] by permission of Elsevier.

It has been shown that the degree of fiber alignment depends on the number and the configuration of the conductive strips-collector type (Figure 24А-C), as well as on the type of the insulating material used (Figure 24D-F) [82].

A) А1 ( d = 300 nm,

θ

= 15º)

B) А2 ( d = 280 nm,

θ

= 12º)

C) А3 ( d = 230 nm,

D) PP ( d = 760 nm,

θ

= 23º)

E) SiO2 ( d = 550 nm,

θ

= 17º)

F) C ( d = 300 nm,

θ

θ

= 10º)

= 15º)

Figure 24. SEM micrographs of aligned CECh/PVA fibers, obtained on the different types of collectors.

θ is the angle at which the individual fiber deviates from the targeted direction, AFS 1.6 kV/cm, duration of electrospinning process 30 min. Reproduced from Mincheva et al. [82] by permission of Elsevier.

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 113 Zhou et al. have followed the preparation of bicomponent CECh/PVA nanofibers from their aqueous solutions [269]. In agreement with the results obtained by Mincheva et al. [82] they have found that the morphology and diameter of nanofibers depend on the weight ratio between the polymer partners. Based on the DSC and XRD analyses it has been assumed that the crystalline microstructure of PVA in the fibers is not well developed. This has been attributed to the stretching of polymer chains in the process of fiber formation. As a result of the rapid solvent evaporation the polymer chains preserve such a strongly extended structure and majority of the chains are in the noncrystalline state. Nanofibers from CMCh have been electrospun with adding of water-soluble polymers – PEO, PAA, PAAm and PVA, to the spinning solution [270]. In the case of CMCh/PEO nanofibers with 30% CMCh can be obtained with average diameter being mostly 300 nm. Their shape is not a cylindrical one, and they merge in the fiber crossover points. Mixing CMCh with either PAAm or PAA enables the formation of nanofibers with a higher CMCh content of 50%, however the fibers contain a considerable number of bead-like defects. The most efficient formation of CMCh/PVA fibers is observed when using PVA, where the CMCh content may reach 80%. It has been found that continuous cylindrical defect-free CMCh/PVА nanofibers are obtained by electrospinning of solutions containing CMCh ≤ 50%. Moreover, the average fiber diameters decrease slightly from 210 to 170 nm on increasing the CMCh content from 20% to 50%.

5.2. Electrospinning of Chitosan Derivatives Chemical modification of chitosan into derivatives that are soluble in a wide variety of organic solvents is an alternative approach to facilitating chitosan electrospinning. Hexanoyl chitosan (Figure 25) is soluble in various common organic solvents and exhibits good blood compatibility [271], antithrombogenic activity and resistance to hydrolysis by lysozyme [260,271,272]; thus could be useful for biomedical applications.

OCO(CH2)4CH3 O O OC

O

N OC

CO

(CH2)4 (CH2)4 (CH2)4 CH3

CH3

Figure 25. Chemical structure of hexanoyl chitosan.

CH3

D. Paneva, М. Ignatova, N. Manolova et al.

114

Continuous defect-free microfibers from hexanoyl chitosan with ribbon-like morphology have been prepared by electrospinning of hexanoyl chitosan solutions in chloroform at polymer concentration 14% w/v (Figure 26A) [260].

A

B

C

Figure 26. SEM micrographs of fibers prepared from: (A) 14% w/v hexanoyl chitosan solution in chloroform, (B) 8% w/v hexanoyl chitosan solution in chloroform without addition of pyridinium formate salt, and (C) with 7.5% w/v pyridinium formate salt addition. AFS 1.0 kV/cm, magnification: ×400. Reproduced from Neamnark et al. [260] by permission of Elsevier.

The electrospun fibers have average diameters in the range of 0.64 - 3.93 μm. It has been found that on increasing the hexanoyl chitosan concentration, the fiber diameters increase and the amount of bead-like defects decrease (Figure 26). The addition of an organic salt, pyridinium formate, to the spinning solution leads to increase of electrical conductivity of the spinning solution, which results in increase of the fiber diameter and and to a smaller number of defects (Figure 26). In addition to using chloroform as a solvent, DCM and THF have been used to electrospin hexanoyl chitosan and hexanoyl chitosan/PLA [273]. The electrospun hexanoyl chitosan fibers are ribbon-like, with smooth surface and large average diameters (0.91 µm and 0.50 µm from chloroform and DCM, respectively from 10% (w/v) hexanoyl chitosan solution). The use of THF solvent leads to low productivity. Fibers from hexanoyl chitosan/PLA obtained at different weight ratios hexanoyl chitosan/PLA in chloroform with the hexanoyl chitosan solution contents of less than or equal to 50% (w/w) are defect-free fibers with rough surface. The diameters of these fibers decrease with the increase of the hexanoyl chitosan content at a constant AFS value (1.06 kV/cm). When electrospinning hexanoyl chitosan/PLA mixed solutions in DCM, with varying the weight ratios between partners from 20/80 to 80/20, fibers with bead-like defects are obtained. The difference observed in the morphology of the hexanoyl chitosan/PLA fibers when applying different solvents in electrospinning has been attributed to the substantially lower viscosities of the mixed hexanoyl chitosan/PLA solutions in DCM than those in chloroform. The average diameter of the fibers obtained from mixture of 10% (w/v) hexanoyl chitosan solution and 24% (w/v) PLA solution in chloroform are in the range of 0.20 - 1.26 µm, while that from mixture of 10% (w/v) hexanoyl chitosan solution and 24% (w/v) PLA solution in DCM are in the range of 0.74 - 1.26 µm. PEG has good solubility in both water and organic solvents and possesses low toxicity, good biocompatibility and biodegradability [274]; and it finds a wide variety of applications in food, cosmetics and pharmaceutical industry [275].

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 115

OH O HO

O

NH CH2CH2(OCH2CH2)mOCH3

PEG-N-Ch OOCR O

O

NHCOR

RCOO

R = (OCH2CH2O)n OCH3

PEG-N,O-Ch Figiure 27. Chemical structures of PEGylated chitosans.

Because of these properties PEG chains have been grafted onto chitosan. PEGylated chitosans - PEG-N-chitosan (PEG-N-Ch) and PEG-N,O-chitosan (PEG-N,O-Ch) (Figure 27) have been synthesized via reductive amination and acylation of chitosan, respectively [261]. The electrospinning of solutions of PEGylated chitosan in distilled water has failed and only beads have been obtained [261]. In electrospinning of a 25% PEG-N,O-Ch solution in DMF fibers intermixed with beads-like defect are formed. To improve both the efficiency of fiber formation and fiber uniformity, it is necessary to add a cosolvent and a nonionic surfactant to the spinning solutions. Thus, the electrospinning of 15% PEG-N,O-Ch from 75/25 (v/v) THF/DMF with 0.5% Triton X-100TM leads to formation of cylindrical continuous defect-free nanofibers with an average diameter of 162 nm. It has been suggested that the improved fiber uniformity is most probably due to the higher number of chain entanglements which is a result of interactions between the chains of the surfactant and the graft PEG chains. In order to improve the mechanical properties of chitosan in its hydrated state with a view to its potential biomedical application, grafting of oligo(D,L-lactic acid) onto chitosan (Ch-goligo(D,L)LA) has been performed by dehydration of chitosan lactate (Figure 28) [176].

D. Paneva, М. Ignatova, N. Manolova et al.

116

OH

OH

O

O O

HO NH3

O- + HO

O

80 0C

O

HO NH

- H 2O H

O

O

n

O

O Ch-g-oligo(D,L)LA Figure 28. Grafting of oligo(D,L-lactic acid) onto chitosan.

Only beads or beaded fibers have been obtained by electrospinning of a solution of this derivative in aqueous acetic acid using conventional electrospinning technique. Fibers have been produced via an electro-wet-spinning technique using a coagulation bath (Figure 29) [176]. The average diameters of the fibers are in the range 0,100 μm - 3 μm. The mats have various pore sizes ranging from 1 μm to less than 30 μm and different porosities up to 80%. The morphology and size of fibers, as well as their porosity depend on the concentration of the Ch-g-oligo(D,L)LA spinning solutions and the solution composition of the coagulation bath. The tensile strength and Young’s modulus of the obtained fibrous materials from Ch-goligo(D,L)LA in hydrated state are much higher compared to those of chitosan (793.4 ± 26.7 kPa and 18.1 ± 2.2 MPa, against 117.2 ± 28.9 kPa and 1.1± 0.2 MPa for Ch-g-oligo(D,L)LA and chitosan, respectively). Pump High voltage supply (1)

Polymer solution Electrode (+)

High voltage supply (2)

Liquid reservoir Electrode (-)

Figure 29. A schematic representation of an electro-wet-spinning set-up.

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 117 It has to be noted that in literature, most probably due to experimental difficulties, data on the mechanical properties of nanofibrous mats are scarce. Determination of such mat characteristics, especially in the hydrated state is of great interest, mainly with a view to their potential biomedical application. Another approach which keeps the chitosan amino groups unchanged is grafting L-lactide oligomers by ring opening polymerization of L-lactide in the presence of methanesulfonic acid which plays a dual solvent and catalyst role [276] (see Figure 30 for chemical structure of chitosan-g-poly(L-lactide) (Ch-g-PLLA).

OR O RO

O

NH2 OH

R = CO CH O CO CH O CO CH CH3 n CH3 CH3 Figure 30. Chemical structure of chitosan-g-poly(L-lactide) (Ch-g-PLLA).

The possibility to control the side chain length by varying the L-lactide/chitosan ratio allows the manipulation of the biodegradation rate and hydrophilicity of the material. Lowmolecular-weight (LMW) and high-molecular-weight (HMW) chitosan have been used however the molar mass of chitosan has not been specified, so it is difficult to discuss the differences observed in the fiber morphology. Microfibers have been prepared in all cases. Defect-free and ribbon-like- microfibers are formed by electrospinning of HMWCh-g-PLLA from its solution in ethyl acetate at polymer concentration 44 wt. % and of LMWCh-g-PLLA from its solution in 2-butanone at polymer concentration 50 wt. %.

5.3. Two-Step Imparting Water-Insolubility of Chitosan DerivativesContaining Nanofibers The possibilities for biomedical application of the bicomponent electrospun mats from water-soluble chitosan derivatives might be significantly enlarged if they are rendered stable in aqueous environment.

D. Paneva, М. Ignatova, N. Manolova et al.

118

A

B

C

D

Figure 31. SEM micrographs of the QCh-containing fibers prepared: from QCh/PVA solutions without (А), and in the presence of 1 wt. % DMPA, 1 wt. % ammonium peroxydisulphate and 10.7 wt. % TEGDA (В), weight ratio QCh/PVA = 2/3, AFS 2,0 kV/cm, polymer concentration 10 wt. %; solvent H2O/DMSO=92/8 (w/w), magnification: ×10 000 (A and B); and from QCh/PVP solutions without (C), and in the presence of 1 wt. % DMPA, 4.5 wt. % TEGDA and 1.5 wt % DAS (D), QCh/PVP = 2/3, AFS 2.2 kV/cm, polymer concentration 20 wt. %; solvent H2O/DMSO=92/8 (w/w), magnification: ×2 000 (C and D). Reproduced from Ignatova et al. [28,32] by permission of Elsevier.

In order to retain the unique nano- and microfibrous structure two-step photo-mediated crosslinking of electrospun QCh/PVA and QCh/PVP mats in the solid state has been performed [28,32]. First, QCh/PVA and QCh/PVP solutions containing photo-crosslnking additives have been electrospun, and then UV irradiation of the fibers in the solid state has been performed. It has been shown that adding the photoinitiator 2,2-dimethoxy-2phenylacetophenone (DMPA), ammonium peroxydisulfate and triethylene glycol diacrylate (TEGDA) as crosslinking agent to the QCh/PVA mixed solution does not hamper the fabrication of fibers (Figure 31 A, B) [28]. The nanofiber diameter decreases (from 217 to 116 nm) and the size distribution of the crosslinked nanofibers is narrower on adding crosslinking additives. The observed effects have been attributed to increased conductivity of the solution (from 2.95 mS/cm to 4.3 mS/cm) when the inorganic salt - ammonium peroxydisulfate is added. Similar effects have been reported for other systems [45,50] and are explained by the higher charge density on the surface of ejected jet during electrospinning, thus imposing higher elongation forces to the jet. Cylindrical defect-free QCh/PVP microfibers have been electrospun at total polymer concentration 20 wt. %, using crosslinking agents DAS and TEGDA, and DMPA as

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 119 photoinitiator (Figure 31 C,D) [32]. In this case, due to the significant increase of the solution viscosity (from 8 200 cP to 9 800 cP) when adding the photo-crosslinking agents, fibers with a higher average diameter are obtained (up to 2 900 nm). Increased solution viscosity leads to an increase in the viscoelastic force which counteracts the Coulomb repulsion force that tries to stretch the charged jet, thus resulting in formation of fibers with higher average diameters and in decrease in the number of defects [20,232]. UV-irradiation for 10 h of the electrospun QCh-containing nano- and microfibrous mats containing crosslinking agents and photoinitiator results in stabilizing of the fibers against disintegration in aqueous medium [28,32]. When put in contact with water and water vapor the fibers keep their morphology and do not dissolve (Figure 32). The equilibrium swelling degree of photo-crosslinked QCh-containing micro- and nanofibers, in distilled water at 23 οC is 100% and 67% in the case of QCh/PVA system and QCh/PVP system, respectively [28,32]. In order to stabilize the electrospun bicomponent QCh/PLDLA [246] and CECh/PVA [269] mats against dissolving in water, a two-step process of crosslinking of fibers has been applied. QCh/PLDLA or CECh/PVA spinning solutions are first electrospun into nanofibers followed by covalent crosslinking with GA vapors.

А

B Figure 32. Effect of water on the morphology of the photo-crosslinked QCh/PVA and QCh/PVP fibers. Photo-crosslinked QCh/PVA mat after contact with water for 6 h (A), weight ratio QCh/PVA = 2/3, total polymer concentration 10 wt. % (H2O/DMSO=92/8 w/w), AFS 2,0 kV/cm, magnification: ×10 000. Photo-crosslinked QCh/PVP mat after contact with water for 6 h (B), weight ratio QCh/PVP = 2/3, total polymer concentration 20 wt. % (H2O/DMSO=92/8 w/w), AFS 2,2 kV/cm, magnification: ×2 000. Reproduced from Ignatova et al. [28,32] by permission of Elsevier.

It has been shown that after soaking in aqueous solutions for 10 h, the hybrid crosslinked QCh/PLDLA nanofibrous mats swell to some extent while maintaining their fibrous morphology and retaining their integrity [246] (Figure 33). The equilibrium swelling degree

120

D. Paneva, М. Ignatova, N. Manolova et al.

(αeq) of crosslinked QCh/PLDLA nanofibers reaches 160% (distilled water, 25 °C). No weight loss has been detected after 24 h stay in water [246].

A

B Figure 33. Effect of water on the morphology of the crosslinked QCh/PLDLA nanofibers (QCh/PLDLA = 30/70 w/w). Non-treated mat crosslinked with GA vapor for 4 h (A), and after contact with water for 10 h (B). Ignatova et al. [246], Electrospun non-woven nanofibrous hybrid mats based on chitosan and PLA for wound-dressing applications, Macromol. Biosci., 2009, 9, 102-111, Copyright Wiley-VCH Verlag GmbH and Co. KGaA. Reproduced with permission.

In order to impart water insolubility to CECh-containing nanofibers with a view to their potential biomedical application, an alternative method of crosslinking CECh/PVA and CECh/PААm nanofibrous materials by heat treatment in solid state below the temperature of softening of the non-ionogenic polymers and below the decomposition temperature of the polyelectrolytes has been reported [50,82]. The conditions for heat treatment have been chosen by taking into account the thermal behavior of the polymer partners. Thus, for example the CECh/PVA mats have been heated at 100°C for 10 h [82], and those prepared from CECh/PААm have been heated at 100, 120 or 150°C for 5 h [50]. SEM analyses of the samples indicate that these temperatures cause no changes in the nanofiber morphology

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 121 [50,82]. The behavior of the nanofibers in contact with water vapor or water is dependent on their composition and on the heating conditions [50,82].

A

B

C

D

Figure 34. SEM micrographs of CECh/PAAm nanofibers. CECh/PAAm=4/1, heat-treated at 120°C for 5 h (A, ×10000) and after subsequent contact with water for 1h (B, ×10000). CECh/PAAm=9/1, heattreated at 120°C for 5 h (C, ×5000) and after subsequent contact with water for 1 h (D, ×5000), AFS 1.2 kV/cm. Reproduced from Mincheva et al. [50] by permission of SAGE.

Unlike the nanofibers from CECh/PAAm=1/1 that have been heat-treated at 120°C for 5 h are partially resistant to water vapor and water, the same treatment of nanofibers enriched in CECh (CECh/PAAm=4/1) leads to a considerable improvement of their resistance to water and water vapor (Figure 34B) [50]. In the case of the system with the highest CECh content (CECh/PAAm=9/1) on contact with water the nanofibers dissolve, while the defects remain unchanged (Figure 34D). This peculiar behavior is attributed to phase-separation during electrospinning at the large CECh excess. For the CECh/PAAm system the crosslinking of the nanofibers is due to interactions between the amino groups and carboxyl groups on CECh. Both with the CECh/PAAm system and in the case of CECh/PVA the resistance of the nanofibers against water depends on the composition of the spinning solution and increases on increasing CECh content [82]. A considerable crosslinking of the mats and retaining the fibrous structure of materials after contact with water for one week has been observed for the heat-treated CECh/PVA nanofibers in weight ratio CECh/PVA = 1/3 (Figure 35D), however the nanofibers with low CECh content (CECh/PVA = 1/8) (Figure 35A) strongly swell after contact with water for one week and the material loses its fibrous structure (Figure 35C).

D. Paneva, М. Ignatova, N. Manolova et al.

122

A

B

C

D

Figure 35. SEM micrographs of nanofibers containing CECh and PVA at weight ratio CECh/PVA = 1/8 (A and B) and 1/3 (C and D) after heating at 100ºC for 10 h (A and C) and after subsequent contact with water for one week (B and D); A) ×10 000, B) ×10 000, C) ×5 000, D) ×2 500; AFS 1.6 kV/cm. Reproduced from Mincheva et al. [82] by permission of Elsevier.

It has been assumed that in the case of CECh/PVA the OH groups from the PVA chains also participate in the crosslinking reactions. The possible reactions with the participation of the both polymers - CECh and PVA, could proceed - amidation reaction (intra-molecular and inter-molecular) between carboxylic and amino groups of CECh or anhydride bond formation between carboxylic groups of CECh, and esterification reaction between the carboxylic groups of CECh and hydroxyl groups of PVA [277]. It has to be noted that the heat-induced crosslinking of the CECh/PAAm and CECh/PVA nanofibers is attained in the temperature range that is suitable for dry sterilization. This fact is important and it paves the way to shorten the production stages of sterile materials for potential biomedical application. Crosslinked water-resistent CMCh/PVA fibrous mats have been prepared by heatinduced crosslinking at 140 °C for 30 min [270]. The results from determination of weight losses and the SEM analyses on fiber morphology show that the crossliking proceeds more considerably for the electrospun mats containing N,O-CMCh (Mv = 405 kDa, DS = 1.14) with longer chain and higher substitution degree (Figure 36). These mats retain their fibrous structure after immersion in water for 1 h in a greater extent than that with much shorter and less substituted O-CMCh (Mv = 89 kDa, DS = 0.36) (Figure 36). It should be noted that the determined weight losses after dipping the mats in water for 3 h are too high – they reach 35.3% for the electrospun mats containing N,O-CMCh (Mv = 405 kDa, DS = 1.14) and 47.2% for those containing O-CMCh (Mv = 89 kDa, DS = 0.36). Although very important, the high weight losses have not been discussed by the authors.

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 123

A

B

C

D

E

F

Figure 36. SEM micrographs of nanofibers from CMCh and PVA at weight ratio CMCh (Mv = 405 kDa, DS = 1.14)/PVA = 50/50 w/w (A and B) and at weight ratio CMCh (Mv = 89 kDa, DS = 0.36)/PVA = 60/40 w/w (C and D) after heating at 140ºC for 30 min (A and C), and after subsequent contact with water for 1 h (B and D), and for 3h (C and F). Reprinted from IOP, Du, J., Hsieh, Y.-L. [270], Nanofibrous membranes from aqueous electrospinning of carboxymethyl chitosan, Nanotechnology, 2008, 19 (12), art. no. 125707, 1-9, with permission from IOP.

5.4. Biomedical Applications of Chitosan Derivatives-Containing Nanofibers As already mentioned in Section 4, the electrospun fibers of chitosan derivatives are the focus of intense study aimed at their biomedical applications, such as wound dressing materials, tissue engineering scaffolds and controlled drug delivery systems. The high specific surface area and small-size pores of electrospun mats are favorable for the adsorption of body fluids and for preventing bacteria penetration and thus provide good conditions for wound healing. It is of interest to incorporate a hydrophilic non-toxic polymer such as PVA [28], PVP [32], poly(ethylene-co-vinyl alcohol) [27], PEO [45,278-280] in the electrospun mats for wound healing applications. To prepare electrospun nanofibrous mats having woundhealing properties two routes have been reported: a) incorporation of a drug (e.g. antibacterial 8-hydroxyquinoline derivatives [45,50], cefazolin [281], itraconazole [282], heparin [280], or silver nanoparticles [283-288]) in the electrospinning solution, and b) electrospinning of polymers with inherent antibacterial and wound-healing properties, such as chitosan [45], QCh [28,32,246], PVP-iodine complex [32,289], sulfonated poly(vinyl phenol) [290], hyaluronic acid [291], collagen [292] and polyurethane [293]. A characteristic of choice of materials designed for wound dressing is its antimicrobial effect. For obtaining the nanofibrous materials with antimicrobial properties three-component spinning solutions containing CECh, PAAm and 7-iodo-8-hydroxyquinoline-5-sulphonic acid (SQ) - a model ionizable drug with broad-spectrum antmicrobial and antimycotic activity

124

D. Paneva, М. Ignatova, N. Manolova et al.

have been used [50]. The presence of SQ has led to a more than two-fold decrease in the diameter of the nanofibers (from 200 to 80 nm), to narrow diameter distribution and to the formation of spindle-like defects but no “tailed” beads. The observed effects have been explained by the increase in the conductivity of the solution on adding a low-molecularweight ionizable compound. In order to estimate the activity of these nanofibrous mats containing a low-molecular-weight compound SQ with known antimicrobial and antimycotic activity, against Gram-negative bacteria (E. coli), Gram-positive bacteria (S. aureus) and the fungus C. albicans, a test by measuring the width of the sterile zone around the nanofibrous mat in Petri dishes with agar medium has been applied [50]. Unlike the blank controls of CECh/PAAm mats around which sterile zones have not appeared, well-defined wide sterile zones around the mats, containing SQ, have been observed. The nanofibrous mats containing high-molecular-weight component QCh with known inherent biocide activity that manifests itself during contact between the bioactive agent and the microorganism, are extremely perspective as wound healing materials. A microbiological test consisting of counting the viable bacterial cells [294] has been applied to assess the antibacterial activity of these mats against Gram-positive bacteria S. aureus and Gramnegative bacteria E. coli [28,32,246]. It has been found that the photo-crosslinked electrospun fibers of PVA, PVP and PLDLA do not inhibit the bacteria growth (Figures 37 and 38).

Log (survivors) cells/mL

6 5 4 control cross-linked PVP mat cross-linked QCh/PVP mat cross-linked QCh/PVP film

3 2 1 0

0

20

40

60

80

100

120

100

120

Exposure time, min

A

Log (survivors) cells/mL

6 5 4 3 control cross-linked PVP mat cross-linked QCh/PVP mat cross-linked QCh/PVP film

2 1 0 0

20

40

60

80

Exposure time, min

B Figure 37. Logarithm plot of the viable bacteria cell number versus the exposure time: for crosslinked PVP mats, for crosslinked QCh/PVP electrospun mats and for crosslinked QCh/PVP films prepared by solvent casting method (A, B). The tests have been carried out against S. aureus (A) and against E. coli (B). Reproduced from Ignatova et al. [32] by permission of Elsevier.

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 125

Log (survivors) cells/mL

5

control PDLLA mat Ch/PDLLA mat Ch/PDLLA film QCh/PDLLA mat QCh/PDLLA film

4

3

2

1

0 0

20

40

60

80

100

120

100

120

Exposure time, min

A

Log (survivors) cells/mL

5

4

3

2

control PDLLA mat Ch/PDLLA mat Ch/PDLLA film QCh/PDLLA mat QCh/PDLLA film

1

0 0

20

40

60

80

Exposure time, min

B Figure 38. Logarithm plot of the viable bacteria cell number versus the exposure time: for PLDLA electrospun mats, for crosslinked QCh/PLDLA electrospun mats, for crosslinked QCh/PLDLA films prepared by solvent casting method, for crosslinked Ch/PLA electrospun mats and for crosslinked Ch/PLA films prepared by solvent casting method (A, B). The tests have been carried out against S. aureus (A) and against E. coli (B). Ignatova et al. [246], Electrospun non-woven nanofibrous hybrid mats based on chitosan and PLA for wound-dressing applications, Macromol. Biosci., 2009, 9, 102111, Copyright Wiley-VCH Verlag GmbH and Co. KGaA. Reproduced with permission.

Unlike them, the crosslinked composite electrospun mats containing QCh or Ch are efficient in inhibiting growth of S. aureus and E. coli (Figures 37 and 38). The electrospun QCh/PVA and QCh/PVP photo-mediated crosslinked mats are found to kill S. aureus faster than E. coli at the same concentration of QCh (3000 µg/ml) (Figure 37) [28,32]. The QCh/PLDLA and Ch/PLDLA nanofibrous mats exhibit higher killing rate against S. aureus and E. coli than the solvent cast films with the same composition (Figure 38) [246]. Similar higher antibacterial efficacy of the electrospun mats has also been observed in other systems [289,290] and has been explained in terms of the high specific surface area of the mats which can result in an increased level of contact between the nanofibrous mats with antibacterial properties and the bacteria suspension, hence in higher bacteria kill rate of the electrospun mats. The obtained results show that the antibacterial activity of crosslinked electrospun QCh- and Ch-containing mats is due to the presence of chitosan or its quaternized derivatives and the effect against Gram-positive bacteria S. aureus and Gram-negative bacteria E. coli is mainly bactericidal. The adhesion of bacteria S. aureus on crosslinked QCh/PLDLA mats and on PLDLA mats has also been studied [246]. For this purpose S. аureus has been cultured for 24 h on the surfaces of electrospun crosslinked mats and after immersion into GA solution in a phosphate

D. Paneva, М. Ignatova, N. Manolova et al.

126

buffer solution (PBS) for cell fixation and subsequent washing with PBS and freeze-drying, the mats have been analyzed by SEM. After 24 h a lot of cells of S. аureus are observed on partially oriented PLDLA nanofibers collected onto rotating drum (Figure 39A). In contrast, no bacteria adhesion is observed on the surface of bicomponent crosslinked QCh/PLDLA mats (Figure 39B).

A

B

Figure 39. SEM micrographs of mats that have been incubated in S. aureus cell culture (107 cells·mL-1) for 24 h at 37 °C. PLDLA mats (A) and crosslinked QCh/PLDLA mat (30/70 w/w) (B), spinning solution in DMF/DMSO = 60/40 (v/v). The mats were collected onto rotating drum. Magnification: × 2500. Ignatova et al. [246], Electrospun non-woven nanofibrous hybrid mats based on chitosan and PLA for wound-dressing applications, Macromol. Biosci., 2009, 9, 102-111, Copyright Wiley-VCH Verlag GmbH and Co. KGaA. Reproduced with permission.

This finding suggests that the tendency for adhesion or proliferation of bacteria has been prevented in great extent on the surfaces of mats containing QCh, which is known with its bactericidal effect. It has been concluded that the combination of adhesion-preventing properties towards pathogenic bacteria S. aureus and high bacteria kill rate render these mats promising candidates for wound healing applications. It is expected that tissue engineering scaffolds based on electrospun nanofibrous mats mimicking the architecture of the extracellular matrix should offer great advantages for tissue engineering. Nanofibrous scaffolds can be not only a substitute or a synthetic substrate for the natural extracellular matrix in the body, but they can provide a three-dimensional environment for better cell adhesion and proliferation stimulating the growing cells to organize into tissue. It has been reported that the fiber diameter may substantially affect the morphology and proliferation of cells grown on the scaffold, whereas nanofibrous scaffolds can facilitate cell attachment and support cell growth [295-297]. Electrospun nanofibrous scaffolds have a high specific surface area favoring protein absorption, thus providing more binding sites to the membrane receptors of the cells. Nanofibrous mats possess exceptionally high porosity which allows the exchange of gases and nutrients to support the growing tissue. The reported studies on the possibilities for use of electrospun nanofibers from chitosan derivatives as scaffolds in tissue engineering are still scarce and however, are directed to seeding the fibrous mats with specific cell types, observing cell attachment and proliferation on them over a period of time, as well as on carrying out cytotoxicity tests.

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 127 In vitro tests using rabbit fibroblasts seeded on chitosan and Ch-g-oligo(D,L)LA nanofibrous scaffolds have shown that the mats exhibit good capability to promote the adhesion and proliferation of fibroblast cells [176]. Moreover, no substantial difference in the size, density or distribution of fibroblasts grown on both materials has been observed. Citotoxicity tests performed using mouse fibroblasts (L929 cell line) on electrospun L-lactide modified chitosan fibers show that the electrospun mats at molar ratio of chitosan/L-lactide = 1/24 are nontoxic to the fibroblasts cells [276]. The potential use of the CECh/PVA electrospun fibrous mats as scaffolds for skin regeneration has been evaluated in vitro using mouse fibroblasts as reference cell lines [269]. Indirect cytotoxicity assessment of the fibrous mats shows that the mats are nontoxic to L929 fibroblast cells and do not release substances harmful to living cells. The L929 cells adhere well on the mat surface and possess a normal morphology. The biocompatibility of hexanoyl chitosan fibrous scaffolds has been assessed in vitro towards human keratinocytes (HaCaT) and human foreskin fibroblasts (HFF) (Figure 40) [298]. Hexanoyl chitosan fibrous scaffolds exhibit higher cell viability than the solvent-cast hexanoyl chitosan films. The electrospun hexanoyl chitosan fibrous scaffolds can support the attachment and the proliferation of both types of cells. In addition, the cells cultured on the hexanoyl chitosan fibrous scaffolds preserve their specific morphology and integrate well with surrounding fibers to form a 3D cellular network. Such matrices might be suitable as tissue engineering scaffolds for skin regeneration.

A

B

Figure 40. SEM micrographs of HaCaT cells that were cultured on hexanoyl chitosan fibrous scaffolds at 24 h (A) and 3 d (B) of cell culturing. Reproduced from Neamnark et al. [298] by permission of Elsevier.

Nanofibrous mats from chitosan derivatives have also shown potential for applications as carriers in controlled drug delivery systems. PLGA/PEG-g-Ch nanofibrous mats loaded with ibuprofen have been obtained by electrospinning aiming at materials suitable for treatment of atrial fibrillation [299]. Two approaches for the incorporation of ibuprofen in the mat have been used: i) dissolving of ibuprofen in the spinning solution where it is electrostatically conjugated to the PEG-g-Ch, and ii) covalent attachment of ibuprofen to the PEG-g-Ch prior to electrospinning. The solubility characteristics of PEG-g-Ch, i.e. solubility in organic solvents and insolubility in water at neutral рН [300] have been considered as an advantage enabling the preparation of tri-component electrospun mats without any need of crosslinking. The electrospun mats are claimed to be mechanically robust and to have capability to conform

128

D. Paneva, М. Ignatova, N. Manolova et al.

to motions and, thus, are compatible with atrial tissue. The performed studies have shown that ibuprofen release rate can be controlled by the presence of PEG-g-Ch and by the route of ibuprofen incorporation [299]. The release of ibuprofen from electrospun PLGA mats due to simple diffusion is very fast (burst release) however it has been slowed down by adding PEGg-Ch. This change in the release rate has been explained by electrostatic interactions between the dissociated carboxyl groups of ibuprofen and the protonated amino groups of chitosan molecules. Covalent attachment of ibuprofen to PEG-g-Ch has led to prolongation of its release more than two weeks.

6. REACTIVE ELECTROSPINNING OF CHITOSAN AND CHITOSAN DERIVATIVES The term “reactive electrospinning” has been recently introduced [301]. It denotes a new approach which combines electrospinning with a reaction that occurs during the spinning process. The reactive electrospinning is particularly attractive in terms of preparation of crosslinked fibrous materials. Such materials can be prepared applying two-step procedures. However, the concept the crosslinking to occur at a one stage during the electrospinning process is a particularly attractive one. The use of the former approach has allowed the preparation of crosslinked hydrogel fibers [301]. The hydrogel materials from crosslinked water-soluble polymers are considered as highly promising candidates for application in the biomedical field for design of wound healing devices, for controlled drug delivery, as scaffolds for tissue engineering, etc. [302]. It is expected that the combination of the properties of certain hydrogels (such as рН- and temperature sensitivity, biocompatibility) with the high surface-to-volume ratio of the nanofibrous materials would be beneficial for the preparation of new materials with improved behavior. There are few data on successful reactive electrospinning of synthetic polymers. In addition, there are patented methods and original reactive electrospinning set-ups for one-step preparation of crosslinked polymer and polymer-nanocomposite nanofibers by in-line mixing of a polymer spinning solution with a solution containing the crosslinking agent [303]. Reactive azides have been added into the spinning solutions and functionalized PET fibers have been obtained [304]. Then the functionalized fibers have been crosslinked by heating, i.e. this is a two-step crosslinking procedure. Photo-crosslinking during the electrospinning process has been performed by electrospinning of poly(methyl methacrylate-co-2-hydroxyethyl acrylate) functionalized with cinnamoyl chloride [305]. Using a more easily feasible method - irradiation during the electrospinning of solution, nanofibers from crosslinked poly(2-hydroxyethyl methacrylate) (PHEMA) have been produced [301]. The solution contains the monomer, thermal initiator, photoinitiator and a crosslinking agent. First oligomers are synthesized by thermally initiated polymerization of HEMA, then electrospinning is performed under irradiation (Figure 41). The studies on the preparation of nanofibers by reactive electrospinning of natural polymers such as hyaluronic acid, alginates, gelatin, collagen, cellulose are much less in number [90,98].

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 129

Figure 41. Schematics of reactive electrospinning of crosslinked PHEMA-based nanofibers and the SEM image of the produced nanofibers. The scale bar in the SEM image is 500 nm. Reproduced from Kim et al. [301] by permission of ACS.

A dual-syringe mixing technique has been used in order to prepare electrospun materials based on hyaluronic acid [291]. It requires a complex system composed of HMW hyaluronic acid and LMW hyaluronic acid derivative modified with 3,3’-dithiobis(propanoic dihydrazide), and PEG diacrylate as an agent able to crosslink the hyaluronic acid derivative. Moreover, as previously discussed in this Chapter for electrospinning of polyelectrolytes, in this case addition of a non-ionogenic polymer (PEO) is necessary. Further in this Section the performed up to date studies on reactive electrospinning aimed at preparation of covalently crosslinked fibers from chitosan or its derivatives are discussed. The ability of chitosan to give physically crosslinked hydrogels by PEC formation with polyacids offers new possibilities for application of the reactive electrospinning targeted to design of new hydrogel nanofibrous materials from the natural polymer. Thus, the gained till now experience and knowledge on preparation of PEC based nanofibers by electrospinning is emphasized in the Section.

6.1. One-Step Preparation of Electrospun Crosslinked Chitosan Nanofibers As discussed in Sections 4 and 5 of the Chapter, data on two-step preparation of crosslinked nanofibers from chitosan and its derivatives are available. Concerning the twostep procedure, the formation of fibers is followed by crosslinking. One-step preparation of crosslinked chitosan nanofibers by electrospinning of its TFA solution containing the crosslinking agent GA has been reported [242]. The lack of data on the electrospun mat morphology after its stay in aqueous medium at different pH values does not allow the degree and the uniformity of the cross-linking to be assessed. TFA is a rather harsh solvent and it is preferable to be replaced by more readily available solvents with milder effect, i.e. less toxic and more suitable for the potential biomedical application of chitosan- and CECh-containing

D. Paneva, М. Ignatova, N. Manolova et al.

130

nanofibrous materials, such as acetic and formic acid. Reactive electrospinning aimed at preparation of hydrogel nanofibers from semi-interpenetrating networks containing chitosan or CECh and HMW PEO has been applied using formic acid as a solvent and GA as a crosslinking agent [233]. Two mechanisms have been proposed for crosslinking of chitosan by GA. One of them is Schiff base formation (Figure 42, left) [182,185]. OH

OH

O

O

O

O

O

HO

O

HO

NH2

NH2

Base

H

O

O

O

H

H

H NH2

O

H

O

H

H

O

HO

O

O

OH

O O

OH

Base

H

O

O

O

HO N

O

H

H

OH O

O O

HO

O

N OH O

HO O

HO

O O

H

O

H

O

O

N

N

O OH

N O HO

O

O OH

H

O

O

H

Figure 42. GA crosslinks chitosan either by a Schiff base imine functionality (left) and/or by Michaeltype adducts with terminal aldehydes (right). Reproduced from Schiffman et al. [242] by permission of ACS.

The other mechanism takes into account the mechanism proposed earlier for crosslinking of proteins [307]. It is assumed that it holds also for the system chitosan/GA [306], in addition to Schiff base formation. It has been proven that GA aqueous solutions contain a significant amount of α,β-unsaturated aldehyde groups, obtained as a result of GA aldol condensation. Thus, it has been suggested that unsaturated bonds adjacent to -CHO groups would rather result in stable against hydrolysis Michael type amino adducts (Figure 42, right). This contributes to the suggestion that the mechanism of chitosan crosslinking with GA is based on imine structures (Schiff bases) formation and in a less extent of Michael type adducts formation [182]. Recent studies [308] on the crosslinking of chitosan with GA has made the assumption that the obtaining of stable in acidic conditions chitosan networks is due to the formation of resonantly stabilized imine groups as a result of reaction between the amino groups of chitosan and the unsaturated double bonds adjacent to the aldehyde groups in the GA oligomers (Figure 43).

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 131 Hydrogel fibers from chitosan or CECh, PEO and GA have been successfully prepared by reactive electrospinning when 85 % formic acid has been used instead of acetic acid as a solvent since at total polymer concentration of 2.0 wt. % HCOOH enables the rapid gel formation to be avoided [233]. Effective electrospinning in this case is feasible only within 90 min after the start of the process; after that the process transforms to electrospraying. The dependence of the dynamic viscosity of these spinning solutions on time (Figure 44) shows that after 90 min the viscosity of the solutions at total polymer concentration of 2.0 wt. % begins to increase significantly. This is an indication that at this polymer concentration gel formation occurs rapidly not allowing effective electrospinning and preparation of a nanofibrous material with a satisfying yield. The decrease of the polymer concentration up to 1.7 wt. % results in retaining of the dynamic viscosity value of the spinning solution as long as 6 h, i.e. to a significant delay of the crosslinking process. This is favorable for the electrospinning since it ensures a much longer operating regime. The hydrogel fibers, prepared at total polymer concentration of 1.7 wt. % have higher average diameters values ( d = 130 nm) than fibers obtained at polymer concentration of 2.0 wt. % ( d = 60 nm). This result can be attributed to fiber formation with the participation mainly of the sol-fraction of the system at polymer concentration of 2.0 wt. %. The latter contains smaller amounts of GA, thus the obtained fibers are loosely crosslinked. Evidence for this is given by studies on the stability of fibrous materials from chitosan/PEO crosslinked with GA

OH O O

HO N CH OHC (CH2)2 C

CH (CH2)2 C CH (CH2)3 CHO CH N OH

O O HO

Figure 43. Schematic representation of the resonantly stabilized imine bonds formation.

132

D. Paneva, М. Ignatova, N. Manolova et al.

Figure 44. Dependence of the dynamic viscosity of a solution containing chitosan, PEO and GA using 85 wt. % HCOOH as a solvent on time, and SEM micrographs of the prepared fibers; total polymer concentration 1.7 or 2.0 wt. %; chitosan/PEO = 1/1 (w/w); molar ratio [aminoglucoside units]/[CHOgroup of GA] = 1/1; 25±0.1 °С. Penchev et al. [233], Electrospun hybrid nanofibers based on chitosan or N-carboxyethylchitosan and silver nanoparticles, Macromol Biosci., 2009, 9, 000-000, on line, DOI: 10.1002/mabi.200900003, Copyright Wiley-VCH Verlag GmbH and Co. KGaA. Reproduced with permission.

against acidic medium. While the average diameter values of fibers prepared at total polymer concentration of 1.7 wt. % is slightly altered after a 24 h stay in aqueous medium at pH 4 (Figure 45А), the fibers obtained at total polymer concentration of 2.0 wt. % swell significantly in this medium, and their diameters increase from 60 nm to 210 nm. The use of 85 wt. % HCOOH as a solvent has enabled the first successful preparation of hydrogel nanofibers from CECh and PEO in the presence of GA as a crosslinking agent. For this purpose spinning solutions at weight ratio CECh/PEO = 1/1; total polymer concentration of 3.4 wt. % and molar ratio [aminoglucoside units of the initial chitosan]/[CHO-group of GA] = 1/1 have been used. Previously, at this ratio water-insoluble pH-sensitive hydrogels from CECh have been prepared [309]. The dynamic viscosity of the spinning solutions is not altered within 24 h, i.e. within this period the gel formation of CECh/PEO/GA system is delayed in a greater extent in the presence of 85 wt. % HCOOH as compared to chitosan/PEO/GA system. Thus, effective electrospinning of CECh/PEO/GA system can be performed within this period of time (Figure 45B).

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 133

A

B

C

D

Figure 45. Nanofibers prepared by reactive electrospinning of chitosan/PEO/GA (А, total polymer concentration 1.7 wt.%) and CECh/PEO/GA (B, total polymer concentration 3.4 wt. %) before and after a 24 h stay at pH 4 (0.3 М СН3СООН) and deionized water, respectively. Penchev et al. [233], Electrospun hybrid nanofibers based on chitosan or N-carboxyethylchitosan and silver nanoparticles, Macromol. Biosci., 2009, 9, 000-000, on line, DOI: 10.1002/mabi.200900003, Copyright Wiley-VCH Verlag GmbH and Co. KGaA. Reproduced with permission.

Additional evidence for the fact that the gel formation is much more delayed in CECh/PEO/GA system has been obtained from the studies performed on the stability of hydrogel fibers after contact with aqueous medium. As seen from Figure 45 chitosan- and CECh-containing fibers swell in contact with acidic medium but do not dissolve. This is an indication that the crosslinking of chitosan under the action of GA during the electrospinning process has effectively proceeded. Unlike chitosan-containing fibers that retain almost unchanged their average diameters values, CECh-containing fibers under these conditions swell in a great extent and coalescence of the individual fibers is observed.

134

D. Paneva, М. Ignatova, N. Manolova et al.

6.2. Electrospun Nanofibers Composed of Polyelectrolyte Complexes Based on Chitosan The formation of water-insoluble PEC from aqueous solutions containing oppositely charged polyelectrolytes is an efficient tool for preparation of novel hydrogel materials [310]. The application of this approach depending on the polymer partners’ nature and medium conditions (polymer concentration, pH value, ionic strength, temperature) allows the preparation of diverse in respect to sizes and morphology materials, such as nanoparticles, micro- and nanostructured gels, and multilayered films. Ionically crosslinked hydrogels prepared in this way are known as complex coacervates, polyion complexes or polyelectrolyte complexes (Figure 46).

polyanion

mixing

polycation

“Coacervate complex” or hydrogel from polyion complex

Figure 46. A schematic representation of the preparation of hydrogels from physical networks based on oppositely charged polyelectrolytes obtained by mixing of their aqueous solutions.

As reported in Section 3 of this Chapter, chitosan behaves as a polycation in aqueous solutions and is able to form complexes with weak and strong polyacids [198,199,311,312]. That is why the possibility of utilizing these properties of chitosan for preparation of nanofibrous materials applying the electrospinning process is highlighted in this Section. The combination of PEC formation and electrospinning has been applied for the first time using a two-step procedure: electrospinning of a polymer followed by formation of a multilayered coating of PEC applying the layer-by-layer technique [313]. After the development of the layer-by-layer (LbL) technique [314] this tool of formation of self-assembled structures has become one of the main approaches applied for preparation of functionalized thin films. It consists in consequent adsorption of oppositely charged polyelectrolytes on different in nature substrates. It is claimed that the thickness of thus prepared multilayer films can be controlled with accuracy within the nanoscale [315]. An exceptional advantage of this technique is the fact that it can be applied by using different types of polyelectrolytes, as well as different in shape, morphology and size substrates [316,317], such as metallic nanotubes [318], short inorganic fibers [319], metal nanoparticles [320], polymer microbeads [321], etc. In the case of using porous substrates the latter are usually obtained by applying phase inversion technique, whereas materials with non-uniform pore size distribution are prepared [322,323]. This structure type has two main disadvantages: a comparatively low porosity degree and a non-controlled pore size distribution [324]. Electrospinning is considered as an alternative

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 135 technology that can be applied for preparation of porous scaffolds with high degree of porosity, better pore size distribution and a high surface-to-volume ratio [17,313,325]. That is why the development of new solutions for combining the both technologies – electrospinning and the LbL technique, is of interest. Electrospun cellulose acetate nanofibrous membrane that has been coated by the LbL technique with a polycation (PAH) and polyanion PAA has been used [326]. Chain segments of the polyelectrolyte partners are shown on Figure 47. CH2 CH

CH2 CH n COOH PAA

CH2

n

NH3+ Cl

Poly(allylamine hydrochloride) (PAH)

Figure 47. Formulae of PAA and PAH.

Since РАА is a weak polyacid (рКа = 4.8, [327]) pH value of the polymer solutions has a substantial impact on the formation of multilayer coatings from PEC. Nanofibers have been prepared from cellulose acetate with a multilayer coating from chitosan/sodium alginate PEC or from chitosan/polystyrene sulfonate PEC [328] Hollow multilayered PEC fibers have been obtained by using polystyrene nanofibers as a template which, after depositing the PEC film, has been removed by selective dissolution (Figure 48) [329,330].

Figure 48. Schematic diagram illustrating the fabrication of hollow multilayer polyelectrolyte nanofiber via LbL coating and removal of template. Reproduced from Ge et al. [330] by permission of Elsevier.

PAH – a weak polybase, and poly(styrene sulfonate) – a strong polyacid, have been used to form PEC [330]. It is assumed that the obtained hollow fibers can find potential application as drug delivery systems, as filters and in the tissue engineering. Successful experiments for reactive electrospinning of solutions containing oppositely charged polyelectrolytes have been performed by using weak polyacids or weak polybases

136

D. Paneva, М. Ignatova, N. Manolova et al.

[331-333]. At certain рН values the ionogenic groups of the weak polyelectrolytes are not in an ionized form and they cannot form PEC with weak or strong oppositely charged polyelectrolytes. A common spinning solution of chitosan and collagen has been electrospun by using a HFIP/TFA solvent system at 90/10 (v/v) [331,333]. The use of this mixed solvent enables the obtaining of nanofibers of average diameter below 500 nm. Undoubtedly these results are interesting despite the lack of any experimental data proving the formation of water-insoluble PEC, e.g. by determination of the fiber stability in the pH range in which the complex exists. The preparation of nanofibers from aqueous solutions containing PAH and PAA (molar ratio PAH/PAA=1/2) is a successful one at pH 1.2 [332] since at this pH value the carboxylic groups of PAA are not ionized and they are not able to form a water-insoluble PEC. The solubility of the fibers in physiological solution imposes the necessity of application of subsequent thermal crosslinking at 140 ºС.

A

B Figure 49. SEM micrographs of nanofibers from PEC chitosan/PAA prepared by electrospinning before (A) and after a 24 h stay at pH 4 in 0.3 М СН3СООН (B); magnification of ×5000. Penchev et al. [334], Novel electrospun nanofibers composed of polyelectrolyte complexes. Macromol. Rapid Commun., 2008, 29, 677-681, Copyright Wiley-VCH Verlag GmbH and Co. KGaA. Reproduced with permission.

A novel approach for preparation of nanofibers from PEC chitosan/PAA has been proposed [334]. Macrophase separation has not been detected and the solution of the weak polybase chitosan and the weak polyacid PAA is homogenous when the mixed solvent H2O/HCOOH (volume ratio of 1/3.4) was used (pH 1). Water-insoluble chitosan/PAA complex is not formed at pH < 3. Below this pH value the predominant part of the PAA

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 137 carboxylic groups are in non-ionized form and the polyacid is not able to participate in PEC formation with the protonated amino groups of chitosan. SEM micrographs of the prepared fibers are shown in Figure 49A. The fibers are cylindrical in shape with average diameter value of 100±40 nm. Water-insoluble PEC chitosan/PAA is formed in narrow pH range from 3 to 6. As seen from the SEM micrograph shown in Figure 49B, the fibers immersed for 24 h at pH 4 have retained their morphology since as known the complex is stable in the pH range from 3 to 6 [311]. The preparation of nanofibers from chitosan and the strong polyacid PAMPS is examined as well when the same solvent system is used [334].

A

B Figure 50. SEM micrographs of nanofibers from PEC chitosan/PAMPS prepared by electrospinning before (A) and after a 24 h stay at pH 4 in 0.3 М СН3СООН (B); magnification ×5000. Penchev et al. [334], Novel electrospun nanofibers composed of polyelectrolyte complexes. Macromol. Rapid Commun., 2008, 29, 677-681, Copyright Wiley-VCH Verlag GmbH and Co. KGaA. Reproduced with permission.

The molar ratio [aminoglucoside units]/[AMPS units] has been selected to be 1/1 since at this ratio maximal amount of the complex chitosan/PAMPS is formed in the pH range from 1 to 6 [198]. The attempts a homogeneous solution from chitosan and PAMPS at this molar ratio to be obtained by mixing of their solutions in 85 % НСООН (volume ratio H2O/HCOOH = 1/6) are unsuccessful since phase separation occurs as a result of the PEC

138

D. Paneva, М. Ignatova, N. Manolova et al.

formation. In the case of systems in which one of the polyelectrolytes is a strong polyacid or a strong polybase, homogeneous solutions may be obtained using a ternary solvent system [335]. The latter consists of water, polar organic solvent and ionizable low-molecular-weight salt at a certain ratio between the components. The role of the organic solvent is to disrupt the hydrophobic interactions since they significantly contribute to the PEC formation process. The role of the low-molecular-weight salt is to shield the charged ionized functional groups of the polyelectrolyte, thus hampering the formation of water-insoluble complexes. A homogeneous solution of chitosan and PAMPS is obtained using 85% HCOOH (volume ratio H2O/HCOOH = 1/6) in the presence of CaCl2. The electrospinning of this solution enables the preparation of fibers cylindrical in shape with diameters of 130±50 nm (Figure 50А). In acidic medium chitosan/PAMPS fibers swell without dissolution (Figure 50B). The stay of the fibrous mat in 0.3 М CH3COOH is accompanied by release of CaCl2 from the fibers in the aqueous medium and to the formation of water-insoluble PEC. It has been demonstrated by gravimetry that the entire amount CaCl2 is released after a 24 h stay of the mat in acidic medium. The determined loss of polymer from the mat is 19 %. Most probably this is due to extraction of the non-complexed polymer partners that are soluble in the aqueous medium at pH 4. The chitosan/PAMPS complex is stable up to pH 8 [198]. The immersion of the mat from PEC chitosan/PAMPS in a buffer solution of pH 9 is accompanied by swelling followed by its fragmentation after 48 h. The fraction of the mat insoluble at pH 9 is soluble in acidic medium. This confirms that the fibers consist of PEC chitosan/PAMPS which disintegrates at pH 9 to a soluble fraction (PAMPS) and an insoluble fraction (chitosan).

Figure 51. Yarns formation from self-assembled fibers during the electrospinning process from chitosan/PAMPS/HCOOH/CaCl2 system using a stationary collector [336].

In the case of the electrospinning of chitosan/PAMPS/CaCl2 system an interesting phenomenon of self-organization of the fibers during the electrospinning has been observed [336]. The initially formed thin fibers grow in height from the negatively charged collector to the positively charged capillary tip accompanied by of an intensive process of self-bundling of the fibers. This self-assembly leads to formation of yarns (Figure 51). This phenomenon has been observed during the electrospinning of non-ionogenic polymers in the presence of low-molecular-weight conductive salts [34]. It has been found that the self-assembling of the

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 139 respective fibers in bundles can occur when the conductivity of the spinning solutions is higher than 400 μS/cm. In the case of chitosan/PAMPS/HCOOH/CaCl2 the electrical conductivity of the spinning solution is 5600 μS/cm. Having in mind that the most recent trends in the field of micro- and nanofibrous yarns formation is the use of self-assembly of conductive nanofibers into yarns [34,337,338] the nanofibers from polyelectrolytes are highly promising in this respect. In addition, owing to their ionogenic nature and the presence of functional groups in their structure, this type of polymers and materials are potential candidates for design of new generation micro- or nanofibrous yarns that can find diverse application in variety of fields: for design of clothing of remarkably low weight and improved barrier properties against humidity and wind; for the design of novel highly effective membranes for bioreactors, as well as for new highly effective filters; for preparation of bioactive wound healing dressings; for diverse technological solutions related to advanced technologies such as design of biosensors; highly effective catalysts, new devices for the optoelectronics, etc. The one-step formation of waterinsoluble hydrogel fibers from PEC in combination with the self-bundling of the fibers is a substantial prerequisite for the successful preparation of new materials with desired properties. Additional advantage is the possibility natural polymers (e.g., chitosan) to be incorporated thus allowing the design of biocompatible and biodegradable polymer materials. Summarily, the reactive electrospinning enables the tailored preparation of hydrogel nanofibers from chitosan. Two approaches can be applied: covalent crosslinking of chitosan or PEC formation between chitosan and weak or strong polyacids. In addition, the formation of micro- and nanofibrous yarns during the reactive electrospinning broadens the possibilities for tailored preparation of novel electrospun materials that have properties prompted by the difference of the morphology and shape as compared to materials prepared by conventional techniques. As evidenced, a very attractive field for comprehensive research on the possibility for preparation of hydrogel fibrous materials by reactive electrospinning has arisen during the last years. It is to be expected that these materials could bring electrospinning closer to the industrial scale.

7. CONCLUSION In the past few years a large number of attempts to prepare chitosan-based nanofibers by electrospinning have been made using different approaches, e.g. use of suitable solvent, or addition of non-ionogenic polymer partner. At present, fibers from chitosan and chitosan derivatives having nanoscale diameters can be easily produced by electrospinning. The obtained nanofibrous materials have proven to be promising for application in diverse fields, and especially in biomedical field, mainly as scaffolds for cell and tissue engineering, and as wound-healing dressings. While progress in the preparation and characterization of the nanofibrous materials from chitosan and chitosan derivatives has been made, there are still efforts to be done to prepare materials with higher mechanical strength and diameter uniformity. Some studies on hybrid nanofibers prepared using a suitable reinforcing polymer partner have already been performed. Much deeper insight on the effect of the process parameters on the fiber morphology is needed. It may be anticipated that along with the conventional mode of electrospinning more attention will be focused on the reactive

140

D. Paneva, М. Ignatova, N. Manolova et al.

electrospinning and on taking advantage of polyelectrolyte complex formation during electrospinning. The potential of electrospinning as a means to create nanofibrous materials based on chitosan is great, and gives reasons for the intensive development of this field of research.

ACKNOWLEDGEMENTS Financial support from the National Science Fund of Bulgaria (Grant DO-02-82/2008) is gratefully acknowledged.

REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24]

Allcock, H., Lampe, F F. Contemporary Polymer Chemistry; Prentice Hall: Engelwood Cliffs, NJ, 1990. Martin, C. R. Chem. Mater. 1996, 8, 1739-1746. Feng, L.; Li, S.; Li, H.; Zhai, J.; Song, Y.; Jiang, L. Angew. Chem. Int. Ed. 2002, 41, 1221-1223. Liu, G. J.; Ding, J. F.; Qiao, L. J.; Guo, A.; Dymov, B. P.; Gleeson, J. T. European J., 1999, 5, 274-2749. Whitesides, G. M.; Grzybowski, B. Science, 2002, 295, 2418-2421. Hongu, T., Phillips, G. O., Eds.; New Fibers; Woodhead Publishing Limited: Cambridge, England, 1997. Ma, P. X.; Zhang, R. J. Biomed. Mat. Res. 1999, 46, 60-72. Ondarcuhu, T.; Joachim, C. Europhys. Lett. 1998, 42, 215-220. Fabbricante, A.; Ward, G.; Fabbricante, T. US Patent 6,114,017, 2000. Torobin, L.; Findlow, R. US Patent 6,183,670, 2001. Huang, Z. M.; Zhang, Y. Z.; Kotaki, M.; Ramakrishna, S. Composites Sci. Techn. 2003, 63, 2223-2253. Teo, W.; Ramakrishna, S. Nanotechnology, 2006, 17, 89-106. Greiner, A.; Wendorff, J. H. Adv. Polym. Sci. 2008, 219, 107–171. Kriegel, C.; Arrechi, A.; Kit, K.; McClements, D. J.; Weiss, J. Crit. Rev. Food Sci. Nutrition, 2008, 48, 775–797. Cooley, J. F. US Patent 692,631, 1902. Morton, W. J. US Patent 705,691, 1902. Reneker, D. H.; Yarin, A. L. Polymer, 2008, 49, 2387-2425. McKee, M.; Elkins, C.; Long, T. Polymer, 2004, 45, 8705-8715. Shenoy, S. L.; Bates, W. D.; Frisch, H. L.; Wnek G. E. Polymer, 2005, 46, 3372-3384. Fong, H.; Chun, I.; Reneker, D. H. Polymer, 1999, 40, 4585-4592. Fong, H.; Reneker, D. H. J. Polym. Sci.: Polym. Phys. Ed., 1999, 37, 3488-3493. Spasova, M.; Stoilova, O.; Manolova, N.; Altankov, G.; Rashkov, I. J. Bioact. Comp. Polym. 2007, 22, 62-76. Gupta, P.; Elkins, C.; Long, T. E.; Wilkes, G. L. Polymer, 2005, 46, 4799-4810. Shenoy, S. L.; Bates, W. D.; Wnek, G. E. Polymer, 2005, 46, 8990-9004.

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 141 [25] Hager, B. L.; Berry, G. C. J. Polym. Sci., Polym. Phys. Ed. 1982, 20, 911-928. [26] McKee, M. G.; Wilkes, G. L.; Colby, R. H.; Long, T. E. Macromolecules 2004, 37, 1760-1767. [27] Kenawy, E. R.; Layman, J. M.; Watkins, J. R.; Bowlin, G. L.; Mattews, J. A.; Simpson, D. G.; Wnek, G. E.Biomaterials, 2003, 24, 907-913. [28] Ignatova, M.; Starbova, K.; Markova, N.; Manolova, N.; Rashkov, I. Carbohydr. Res. 2006, 341, 2098-2107. [29] Mincheva, R.; Manolova, N.; Rashkov, I. Eur. Polym. J. 2007, 43, 2809–2818. [30] Zong, X. H.; Kim, K.; Fang, D. F.; Ran, S. F.; Hsiao, B. S.; Chu, B. Polymer, 2002, 43, 4403-4412. [31] Jun, Z.; Hou, H.; Schaper, A.; Wendorff, J.; Greiner, A. e-Polymers, 2003, no. 009, 1-9. [32] Ignatova, M.; Manolova, N.; Rashkov, I. Eur. Polym. J. 2007, 43, 1609-1623. [33] McKee, M. G.; Hunley, M. T.; Layman, J. M.; Long, T. E. Macromolecules, 2006, 39, 575-583. [34] Wang, X.; Zhang, K.; Zhu, M.; Yu, H.; Zhou, Z.; Chen, Y.; Hsiao, B.S. Polymer, 2008, 49, 2755–2761. [35] Lin, T.; Wang, H. X.; Wang, H. M.; Wang, X. G. Nanotechnology, 2004, 15, 13751381. [36] Lee, K. H.; Kim, H. Y.; La, Y. M.; Lee, D. R.; Sung, N. H. J. Polym. Sci. Part B: Polym. Phys. 2002, 40, 2259-2268. [37] Wannatong, L.; Sirivat, A.; Supaphol, P. Polym. Int. 2004, 53, 1851-1859. [38] Bognitzki, M.; Czado, W.; Frese, T.; Schaper, A.; Hellwig, M.; Steinhart, M.; Greiner, A.; Wendorff, H. Adv. Mater. 2001, 13, 70-72. [39] Shin, Y. M.; Hohman, M. M.; Brenner, M. P.; Rutledge, G. C. Appl. Phys. Letters, 2001, 78, 1149-1151. [40] Shin, Y. M.; Hohman, M. M.; Brenner, M. P.; Rutledge, G. C. Polymer, 2001, 42, 9955-9967. [41] Fridrikh, S. V.; Yu, J. H.; Brenner, M. P.; Rutledge, G. C. Phys. Rev. Lett. 2003, 90, 144502, 1-4. [42] Jalili, R.; Hosseini, S. A.; Morshed, M. Iran. Polym. J. 2005, 14, 1074-1081. [43] Megelski, S.; Stephens, J. S.; Chase, D. B.; Rabolt, J. F. Macromolecules, 2002, 35, 8456-8466. [44] Lee, J. S.; Choi, K. H.; Ghim, H. D.; Kim, S. S.; Chun, D. H.; Kim, H. Y.; Lyoo, W. S. J. Appl. Polym. Sci. 2004, 93, 1638-1646. [45] Spasova, M.; Manolova, N.; Paneva, D.; Rashkov, I. e-Polymers, 2004, no. 056, www.e-polymers.org. [46] Takahashi, T.; Taniguchi, M.; Kawai, T. Japan. J. Appl. Phys. Part 2: Lett. Expr. Lett. 2005, 44, L860-L862. [47] Shukla, S.; Brinley, E.; Cho, H. J.; Seal, S. Polymer, 2005, 46, 12130-12145. [48] Gu, S. Y., Ren, J. Macromol. Mater. Eng., 2005, 290, 1097-1105. [49] Lee, K. H.; Kim, H. Y.; La, Y. M.; Lee, D. R.; Sung N. H. J. Polym. Sci. Part B: Polym. Physics 2002, 40, 2259-2268. [50] Mincheva, R.; Paneva, D.; Manolova, N.; Rashkov, I. J. Bioact. Compat. Polym., 2005, 20, 419-435. [51] Baker, S. C.; Atkins, N.; Gunning, P. A.; Granville, N.; Wilson, K.; Wilson, D.; Southgate, J. Biomaterials, 2006, 27, 3136-3146.

142

D. Paneva, М. Ignatova, N. Manolova et al.

[52] Hong, Y.; Shang, T.; Li, Y.; Wang, L.; Wang, C.; Chen, X.; Jing, X. J. Membr. Sci. 2006, 276, 1-7. [53] Kidoaki, S.; Kwon, K.; Matsuda, T. J. Biomed. Mater. Res. Part B: Apll. Biomater. 2006, 76B(1), 219-229. [54] Buttafoco, L.; Kolkman, N. G.; Engbers-Buijtenhuijs, P.; Poot, A. A.; Dijkstra, P. J.; Vermes, I.; Feijen, J. Biomaterials, 2006, 27, 724-734. [55] Jeun, J. P.; Lim, Y. M.; Nho, Y. C. J. Industr. Eng. Chem. 2005, 11, 573-578. [56] Zhang, C.X.; Yuan, X. Y.; Wu, L. L.; Sheng, J. Eur. Polym. J., 2005, 41, 423-432. [57] Demir, M. M.; Yilgor, I.; Yilgor, E.; Erman, B. Polymer 2002, 43, 3303-3309. [58] De Vrieze, S.; Van Camp, T.; Nelvig, A.; Hagström, B.; Westbroek, P.; De Clerck, K.; J. Mater. Sci. 2009, 44, 1357-1362. [59] Kim, G.-T.; Lee, J.-S.; Shin, J.-H.; Ahn, Y.-Ch.; Hwang, Y.-J.; Shin, H.-S.; Lee, J.-K.; Sung, Ch.-M. Korean J. Chem. Eng., 2005, 22, 783-788. [60] Frenot, A.; Chronakis, I. S. Curr. Opin. Coll. Interface Sci., 2003, 8, 64-75. [61] Koombhongse, S.; Liu, W.; Reneker, D. H. J. Polym. Sci.: Polym. Phys. Ed., 2001, 39, 2598-2606. [62] Koski, A.; Yim, K.; Shivkumar, S. Mater. Lett., 2004, 58, 493-497. [63] Spasova, M.; Mincheva, R.; Paneva, D.; Manolova, N.; Rashkov, I. J. Bioactive Compatible Polym., 2006, 21, 465-479. [64] Bognitzki, M.; Hou, H.; Frese, M. I. T.; Hellwig, M.; Schwarte, C.; Schaper, A.; Wendorff, J. H.; Greiner, A. Adv. Mater., 2000, 12, 637-640. [65] Lee, Y. H.; Lee, J. H.; An, I.-G.; Kim, C.; Lee, D. S.; Lee, Y. K.; Nam, J.-D. Biomaterials, 2005, 26, 3165-3172. [66] Spasova, M.; Paneva, D.; Manolova, N.; Radenkov, Ph.; Rashkov, I. Macromol. Biosci., 2008, 8, 153-162. [67] Spasova, M. Novel Micro- and Nanofibrous Materials, Prepared by Electrospinning from Solutions of: Chitosan/Polyethylene oxide, Poly(L-lactide)/Polyethylene glycol and Polyethylene oxide /Ti(IV)isopropoxide, PhD Thesis, 2007, Technical University/Bulgarian Academy of Sciences, Sofia, Bulgaria. [68] Torres, B. Proc. Natnl. Conf. Undergrad. Res.; University of Kentucky, Lexington, Kentucky, 2001, p 1–5. [69] Liu, C.; Bai, R. J. Membr. Sci., 2005, 267, 68-77. [70] Geng, X.; Kwon, O. H.; Jang, J. Biomaterials, 2005, 26, 5427-5432. [71] Fennessey, S. F.; Farris, R. J. Polymer 2004, 45, 4217- 4225. [72] Li, D.; Wang, Y.; Xia, Y. Adv. Mater. 2004, 16, 361-366. [73] Park, S.; Park, K.; Yoon, H.; Son, J. G.; Min, T.; Kim, G. H. Polym. Int. 2007, 56, 1361-1366. [74] Chew, S.; Wen, J.; Yim, E.; Leong, K. Biomacromolecules, 2005, 6, 2017-2024. [75] Kim, K.; Lee, K.; Khil, M.; Ho, Y.; Kim, H. Fibers Polym. 2004, 5, 122-127. [76] Matthews, J.; Wnek, G.; Simpson, D.; Bowlin, G. Biomacromolecules, 2002, 3, 232238. [77] Theron, A.; Zussman, E.; Yarin, A. Nanotechnology, 2001, 12, 384-390. [78] Inai, R.; Kotaki, M.; Ramakrishna, S. Nanotechnology, 2005, 16, 208-213. [79] Sundaray, B.; Subramanian, V.; Natarajan, T. S.; Xiang, R.-Z.; Chang, C.-C.; Fann, W.S. Appl. Phys. Lett. 2004, 84, 1222-1224. [80] Teo, W.; Ramakrishna, S. Nanotechnology, 2005, 16, 1878-1884.

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 143 [81] [82] [83] [84] [85] [86]

Li, D.; Wang, Y.; Xia, Y. Nano Lett., 2003, 3, 1167-1171. Mincheva, R.; Manolova, N.; Rashkov, I. Eur. Polym. J. 2007, 43, 2809–2818. Zhou, F.-L.; Gong, R.-H.; Porat, I. Polym. Int. 2009, 58, 331-342. Gupta, P.; Wilkes, G. L. Polymer 2003, 44, 6353-6359. Ding, B.; Kimura, E.; Sato, T.; Fujita, S.; Shiratori, S. Polymer 2004, 45, 1895-1902. Gibson, Ph.; Schreuder-Gibson, H.; Rivin, D. Coll. Surf. A: Physicochem. Eng. Aspects, 2001, 187–188, 469–481. [87] Wang, X.; Kim, Y.-G.; Drew, Ch.; Ku, B.-Ch.; Kumar, J.; Samuelson, L.A. Nano Lett. 2004, 4, 331-334. [88] Zhou, F.-L.; Gong R.-H. Polym. Int. 2008, 57, 837-845. [89] Shaffer, M. S. P.; Sandler, J. K. W. In Nanofibers and Nanotechnology in Textiles, Brown, P. J.; Stevens K., Eds.; Woodhead Publishing Limited and CRC Press LLC, 2007; p 194-234. [90] Andrady, A.L. Science and Technology of Polymer Nanofibers; John Wiley and Sons, Inc.: Hoboken, New Jersey, Canada, 2008. [91] Tan, S.; Huang, X; Wu, B. Polym Int., 2007, 56, 1330–1339. [92] Sell, S.; Barnes, C.; Smith, M.; McClure, M.; Madurantakam, P.; Grant, J; McManus, M.; Bowlin, G. Polym Int., 2007, 56, 1349–1360. [93] Greiner, A.; Wendorff J. H. Adv. Polym. Sci., 2008, 219, 107–171. [94] Hunley M. T.; Long, T. E. Polym. Int., 2008, 57, 385–389. [95] Schiffman, J. D.; Schauer, C. L. Polym. Rev., 2008, 48, 317–352. [96] Agarwal, S.; Wendorff, J. H.; Greiner, A. Polymer 2008, 49, 5603-5621. [97] Jian, F.; HaiTao, N.; Tong, L.; XunGai, W. Chinese Sci. Bull. 2008, 53, 2265-2286. [98] Ramakrishna, S.; Fujihara, K.; Teo, W.-E.; Lim, T.-C.; Ma, Z. An Introduction to Electrospinning and Nanofibers; World Scientific Publishing Co. Pte. Ltd.: Ton Tuck Link, Singapore, 2005. [99] Tolaimate, A.; Desbrières, J.; Rhazi, M.; Alagui, M.; Vincendon, M.; Vottero, P. Polymer 2000, 41, 2463-2469. [100] Tharanathan, R. N.; Kittur, F. S. Crit. Rev. Food Sci. Nutr., 2003, 43, 61-87. [101] Harish Prashanth, K.V.; Tharanathan, R.N. Trends Food Sci. Technol. 2007, 18, 117131. [102] Synowiecki, J.; Al-Khateeb, N. Food Chem. 1997, 60, 605-610. [103] Muzzarelli, R. A. A. Chitin; Pergamon Press: Oxford, U.K., 1977. [104] Li, J.; Revol, J. F.; Marchessault, R. H. J. Appl. Polym. Sci. 1997, 65, 373-380. [105] Ravi-Kumar, M. N. V. React. Funct. Polymers 2000, 46, 1–27. [106] Domard, A. Int. J. Biol. Macromol. 1987, 9, 98-104. [107] Thanou, M.; Verhoef, J. C.; Junginger, H. E. Adv. Drug Deliv. Rev., 2001, 52, 117-126. [108] Muzzarelli, R.; Rocchetti, R.; Stanie, V.; Weckx, M.. In Chitin Handbook, Muzzarelli, R.; Peters, M., Eds.; Grottammare: Atee Edizioni, 1977; p 127. [109] Wu, A. C. M.; Bough, W. A. In: Proc. 1st Int. Conf. Chitin/Chitosan, Muzzarelli, R.A.A.; Pariser, E.R., Eds.; MIT-SG, Cambridge, 1978; p 88. [110] Sannan, T.; Kurita, K.; Iwakura, Y. Makromol. Chem. 1976, 177, 3589-3600. [111] Domszy, J. G.; Roberts, G. A. F.; Makromol. Chem. 1985, 186, 1671-1677. [112] Moore, G. K.; Roberts, G. A. F. Int. J. Biol. Macromol. 1980, 2, 115-116. [113] Muzzarelli, R. A. A.; Muzzarelli, C. Adv. Polym. Sci. 2005, 186, 151–209. [114] Prabaharan, M. J. Biomat. Appl. 2008, 23, 5-36.

144

D. Paneva, М. Ignatova, N. Manolova et al.

[115] Agulló, E.; Rodríguez, M. S.; Ramos, V.; Albertengo, L. Macromol. Biosci. 2003, 3, 521-530. [116] Lim, S. H.; Hudson, S. M. J. Macromol. Sci., Part C, Polym. Rev., 2003, C43, 223-269. [117] Stoilova, O.; Paneva, D.; Manolova, N.; Rashkov, I. In Unit 2, Module 3 in Specific Content Issues Polymeric Materials for Food Packaging and Food Preservation; eReference Tools for VET Trainers in Food Industry, http://www.e-fost.org/en/SCI-A-32. [118] Steinbüchel, A.; Rhee, S. K., Eds., Polysaccharides and Polyamides in the Food Industry: Properties, Production, and Patents, Wiley-VCH Verlag GmbH and Co. KGaA: Weinheim, 2005. [119] Uragami, T.; Tokura, S., Eds., Material Science of Chitin and Chitosan., Springer, 2006. [120] Skryabin, K.G., Vikhoreva G.A., Varlamov V.P., Eds., Chitin and Chitosan: Production, Properties and Usage; Nauka, Russian Academy of Sciences: Moscow, 2002 (in Russian). [121] Dumitriu S., Ed., Polymeric Biomaterials, 2nd ed.; Marcel Dekker Inc.: New York., 2001. [122] Dumitriu S., Ed., Polysaccharides: Structural Diversity and Functional Versatility, 2nd ed., CRC press, 2004. [123] Kurakake, M.; Yo-u, S.; Nakagawa, K.; Sugihara, M.; Komaki, T. Current Microbiol., 2000, 40, 6-9. [124] Nogawa, M.; Takahashi, H.; Kashiwagi, A.; Ohshima, K.; Okada, H.; Morikawa, Y. Appl. Envir. Microbiol. 1998, 64, 890-895. [125] Ilyina, A.V.; Tatarinova, N. Yu.; Varlamov, V. P. Process Biochem. 1999, 34, 875-878. [126] Grigolon, L. B.; Azevedo, A.; Santos, R. R.; Franco, T. T. In Chitin Enzymology. Muzzarelli, R. A. A., Ed., Atec: Italy, 2001; p 78-87. [127] Shi, C.; Zhu, Y.; Ran, X.; Wang, M.; Su, Y.; Cheng, T. J. Surg. Res., 2006, 133, 185192. [128] Stoilova, O.; Koseva, N.; Petrova, Ts.; Manolova, N.; Rashkov, I. J. Bioact. Compat. Polym. 2001, 16, 379-392. [129] Paneva, D.; Mincheva, R.; Stoilova, O.; Manolova, N.; Rashkov, I. e-Polymers 2003, no 019. [130] Bonina, P.; Petrova, Ts.; Manolova, N.; Rashkov, I.; Naydenov, M. J. Bioact. Compat. Polym. 2004, 19, 197-208. [131] Bonina, P.; Petrova, Ts.; Manolov, N.; Rashkov, I. e-Polymers 2004, no 062. [132] Suzuki, S. Fragr. J., 1996, 15, 61-68. [133] Tokora, A.; Kobayashi, M.; Tatekawa, N.; Suzuki, K.; Okawa, Y.; Mikami, T. Microbiol. Immunol., 1989, 33, 357-367. [134] Vasyukova, N. I.; Zinov’eva, S. V.; Il’inskaya, L. I.; Perekhod, E. A.; Chalenko, G. I.; Gerasimova, N. G.; Il’ina, A. V.; Varlamov, V. P.; Ozeretskovskaya, O. L. Appl. Biochem. Microbiol., 2001, 37, 103-109. [135] Hirano, S.; Nagao, N. Agric. Biol. Chem., 1989, 53, 3065-3066. [136] Uchida, Y.; Izume, M.; Ohtakara, A. In Chitin and Chitosan Skjak-Braek, G.; Anthonsen, T.; Sandford, P., Eds.,. Elsevier Applied Science: New York, 1989; p 373382. [137] Chen, M.; Yeh, G. H.; Chiang, B. J. Food Process Preserv. 1996, 20, 379-390.

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 145 [138] Rao, S. B.; Sharma, C. P. J. Biomed. Mater. Res. 1997, 34, 21-28. [139] Muzzarelli, R. A. A.; Muzzarelli, B. In Polymeric Biomaterials, 2nd ed.; Dumitriu, S., Ed.; Marcel Dekker: New York, 2002. [140] Yalpani, M.; Johnson, F.; Robinson, L. E. In: Advances in Chitin and Chitosan, Sandford, P. A.; Zikakis, J. P., Eds.; Elsevier Applied Science: London, UK, 1992; p 543. [141] Sudardshau, N. R.; Hoover, D. G.; Knorr D. Food Biotechnol. 1992, 6, 257-272. [142] Li, Z.; Zhuang, X. P.; Liu, X. F.; Guan, Y. L.; De Yao K. Pоlymer 2002, 43, 15411547. [143] Guo, Z.; Chen, R.; Xing, R.; Liu, S.; Yu, H.; Wang, P.; Li, C.; Li, P. Carbohydr. Res. 2006, 341, 351-354. [144] Kim, C. H.; Choi, J. W.; Chun, H. J.; Choi, K. S. Polym. Bull. 1997, 38, 387-393. [145] Liu, X.F.; Guan, Y.L.; Yang, D.Z.; Li, Z.; Yao, K.D. J. Appl. Polym. Sci. 2001, 79, 1324-1335. [146] Hirano, S. Polym. Int. 1999, 48, 732-734. [147] Hadwiger, L. A.; Kendra, D. F.; Fristensky, B. W.; Wagoner, W. In Chitin in Nature and Technology. Plenum Press: New York, 1986, p 209-214. [148] Stone, C. A.; Wright, H.; Clarke, T.; Powell, R.; Devaraj, V. S. Br. J. Plast. Surgery 2000, 53, 601-606. [149] Mi, F. L.; Shyu, S. S.; Wu, Y. B.; Lee, S. T.; Shyong, J. Y.; Huang, R. N. Biomaterials, 2001, 22, 165-173. [150] Wang, X. H.; Zhu, Y.; Feng, Q. L.; Cui, F. Z. J. Bioact. Compat. Polym. 2003, 18, 135146. [151] Suzuki, M.; Ito, S.; Yamaguchi, I.; Takakuda, K.; Kobayashi, H.; Shinomiya, K.; Tanaka, J. J. Neurosci. Res. 2003, 72, 646-659. [152] Hirano, S. In Chitin and Chitosan. Environmental Friendly and Versatile Biomaterials; Stevens, W. F.; Rao, M. S.; Chandrakran-chang, S., Eds.; AIT: Bangkok, 1996, p 22. [153] Wang, Q. Z.; Chen, X. G.; Li, Z. X.; Wang, S.; Liu, C. S.; Meng, X. H.; Liu, C. G.; Lu, Y. H.; Yu, L. J. J. Mater Sci: Mater. Med. 2008, 19, 1371-1377. [154] Paneva, D.; Manolova, N.; Rashkov, I.; Danchev, D. J. Bioact. Compat. Polym. 2005, 20, 133-151. [155] Horton, D.; Just, E. Carbohydr. Res. 1973, 29, 173-180. [156] Whistler, R.; Kosik, M. Arch. Biochem. Biophys. 1971, 142, 106-110. [157] Vongchan, P.; Sajomsang, W.; Subyen, D.; Kongtawelert, P. Carbohydr. Res., 2002, 337, 1239-1242. [158] Tokura, S.; Nishi, N.; Itoyama, K.; Shirai, A.; Nishimura, S.; Azuma, I. In Chitin World, Karnicki, Z.; Wojtas, A.; Bykowski, P., Eds.; Bremershaven: Wirtschaftsverlag, NW, 1994; p 287. [159] Domard, A.; Rinaudo, M.; Terrassin, C. Int. J. Biol. Macromol. 1986, 8, 105-107. [160] Muzzarelli, R. A. A. Carbohydr. Polym. 1988, 8, 1-21. [161] Maresch, G.; Clausen, T.; Lang, G. In Chitin and Chitosan; Skjak-Braek, G.; Anthonsen, T.; Sandford, P., Eds.; Elsevier Appl. Sci.: London, New York, 1988; p 389. [162] Sashiwa, H.; Yamamori, N.; Ichinose, Y.; Sunamoto, J.; Aiba, S. Macromol. Biosci. 2003, 3, 231-233.

146

D. Paneva, М. Ignatova, N. Manolova et al.

[163] Kurita, K. In Application of Chitin and Chitosan; Goosen, M. F. A., Ed.; Technomic Publishing Co.: Lancaster, Basel, 1997; p 297. [164] Shantha, K. L.; Bala, U.; Rao, K. P. Eur. Polym. J. 1995, 31, 377-382. [165] Yazdani-Pedram, M.; Retuert, J.; Quijada, R. Macromol. Chem. Phys. 2000, 201, 923930. [166] Yazdani-Pedram, M.; Retuert, J. J. Appl. Polym. Sci. 1997, 63, 1321-1326. [167] Blair, H. S.; Guthrie, J.; Law, T. K.; Turkington, P. J. Appl. Polym. Sci. 1987, 33, 641656. [168] Nud’ga, L. A.; Petrova, V. A.; Lebedeva, M. F.; Petropavlovskii, G. A. Russ. J. Appl. Chem. 1996, 69, 1058-1063. [169] Berkovich, L. A.; Timofeyeva, G. I.; Tsyurupa, M. P.; Davankov, V. A. Polym. Sci. U.S.S.R., 1980, 22, 2009-2018. [170] Lagos, A.; Reyes, J. J. Polym. Sci. Part A: Polym. Chem. 1988, 26, 985-991. [171] Retuert, J.; Yazdani-Pedram, M. Polym. Bull. 1993, 31, 559-562. [172] Yazdani-Pedram, M.; Lagos, A.; Retuert, J.; Guerrero, R.; Riquelme, P. J. Macromol. Sci. Chem. 1995, A32, 1037-1047. [173] Smirnova, L.A., Semchikov, Yu.D., Tikhobaeva, Ya.G., Pastukhova, N.V., Polymer Science – Series B 2001, 43, 33-36. [174] Najjar A. M. K.; Yunus, W. Z. W.; Ahmad, M. B.; Rahman, M. Z. J. Appl. Polym. Sci., 2000, 77, 2314-2318. [175] Qu, X.; Wirsen, A.; Albertsson, A.C. J. Appl. Polym. Sci. 1999, 74, 3193–3202. [176] Wan, Y.; Cao, X.; Zhang, S.; Wang, S.; Wu, Q. Acta Biomaterialia 2008, 4, 876-886. [177] Wei, Y. C.; Hudson, S. M.; Mayer, J. M.; Kaplan, D. L. J. Polym. Sci.: Part A: Polym. Chem. 1992, 30, 2187-2193. [178] Mayer, J. M.; Smithfield, R. I.; Kaplan, D. L.; Stow, M. US Patent 5,015,293, 1991. [179] Kawamura, Y.; Mitsuhashi, M.; Tanibe, H.; Yoshida, H. Ind. Eng. Chem. Res. 1993, 32, 386-391. [180] Juang, R.-S.; Ju, C.-Y. Ind. Eng. Chem. Res. 1997, 36, 5403-5409. [181] Knaul, J. Z.; Hudson, S. M.; Creber, K. A. M. J. Polym. Sci. Part B: Polym. Phys. 1999, 37, 1079-1094. [182] Roberts, G. A. F.; Taylor, K. E. Macromol. Chem. 1989, 190, 951-960. [183] Rorrer, G.L.; Hsien, T.-Y.; Way, J.D. Ind. Eng. Chem. Res. 1993, 32, 2170-2178. [184] Hsien, T.-Y.; Rorrer, G. L. Ind. Eng. Chem. Res. 1997, 36, 3631-3638. [185] Tual, C.; Espuche, E.; Escoubes, M.; Domard, A. J. Polym. Sci.: Part B: Polym. Phys., 2000, 38, 1521-1529. [186] Mi, F.-L.; Tan, Y.-C.; Liang, H.-F.; Sung H.-W. Biomaterials, 2002, 23, 181-191. [187] Kamiński, K.; Zazakowny, K; Szczubiałka, K; Nowakowska M. Biomacromolecules, 2008, 9, 3127-3132. [188] Remuñán-López, C.; Bodmeier, R. Drug Dev. Ind. Pharm. 1996, 22, 1201-1209. [189] Gupta, K. C.; Ravi-Kumar, M. N. V. Polym. Int. 2000, 49, 141-146. [190] Gupta, K. C.; Ravi-Kumar, M. N. V. Biomaterials 2000, 21, 1115-1119. [191] Koseva, N.; Manolova, N.; Markova, N.; Radoucheva, T.; Rashkov, I. Polym. Bull. 1999, 43, 101-107. [192] Koseva, N.; Stoilova, O.; Manolova, N.; Rashkov, I. J. Bioact. Compat. Polym. 2001, 16, 3-19.

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 147 [193] Mireles, C.; Martino, M.; Bouzas, J.; Torres, J. A. In Advances in Chitin and Chitosan; Brine, C. J., Sandford, P. A., Zikakis, J. P., Eds.; Princeton, 1991; p 506. [194] Cárdenas, A.; Argüelles-Monal, W.; Goycoolea, F.M.; Higuera-Ciapara, I.; Peniche, C. Macromol. Biosci. 2003, 3, 535-539. [195] Takahashi, T.; Takayama, K.; Machida, Y.; Nagai, T. Int. J. Pharm. 1990, 61, 35-41. [196] Peniche-Covas, C; Argüelles-Monal, W. Macromol. Symp. 2001, 168, 103-116. [197] Fong, H.; Chun, I.; Reneker, D. H. Polymer 1999, 40, 4585-4592. [198] Stoilova, O.; Koseva, N.; Manolova, N.; Rashkov, I. Polym. Bull. 1999, 43, 67-73. [199] Paneva, D.; Stoilova, O.; Manolova, N.; Rashkov I. e-Polymers, 2003, no. 033. [200] Kim, S. J.; Shin, S. R.; Lee, K. B.; Park, Y. D; Kim, S. I. J. Appl. Polym. Sci. 2004, 91, 2908- 2913. [201] Michaels, A. S. In Advances in Separations and Purifications, Peny, E. S., Ed., John Wiley and Sons Inc.: New York, 1968. [202] Schwarz, H. H.; Richau, K.; Paul, D. Polym. Bull. 1991, 25, 95-100. [203] Kim, K. Y.; Min, D. S. Trans. Soc. Biomater. 1988, 11, 658-661. [204] Torrado, S.; Prada, P.; de la Torre, P. M.; Torrado, M. Biomaterials, 2004, 25, 917-923. [205] Mi, F. L.; Her, N. L.; Kaun, C. Y.; Wong, T.; Shyu, S. J. Appl. Polym. Sci. 1997, 66, 2495-2502. [206] Duan, B.; Dong, C., Yuan, X.; Yao, K. J. Biomater. Sci. Polymer Edn., 2004, 15, 797– 811. [207] Ohkawa, K.; Cha, D.; Kim, H.; Nishida, A.; Yamamoto, H. Macromol. Rapid Commun., 2004, 25, 1600–1605. [208] De Vrieze, S.; Westbroek, Ph.; Van Camp, T.; Van Langenhove, L. J. Mater. Sci., 2007, 42, 8029-8034. [209] Greiner, A.; Wendorff, J.H. Angew. Chem. Int. Ed., 2007, 46, 5670–5703. [210] Torres-Giner, S.; Ocio, M.J.; Lagaron, J.M. Eng. Life Sci., 2008, 8, 303–314. [211] Klossner, R.R.; Queen, H.A.; Coughlin, A.J.; Krause, W.E. Biomacromolecules, 2008, 9, 2947-2953. [212] Nikolova, A.; Manolova, N.; Rashkov, I. Polym. Bull., 1998, 41, 115-121. [213] Mucha, M. React. Funct. Polymers, 1998, 28, 19-25. [214] Jia Y.-T.; Gong, J.; Gu, X.-H.; Kim, H.-Y.; Dong, J.; Shen X.-Y. Carbohydr. Polym., 2007, 67, 403–409. [215] Zhang, Y.; Huang, X.; Duan, B.; Wu, L.; Li, S.; Yuan, X.; Colloid Polym. Sci. 2007, 285, 855–863. [216] Zhang, Y.Z.; Su, B.; Ramakrishna, S.; Lim C.T. Biomacromolecules, 2008, 9, 136-141. [217] Lee, H.W.; Karim, M.R.; Park, J.H.; Ghim, H.D.; Choi, J.H.; Kim, K.; Deng, Y.; Yeum J.H. J. Appl. Polym. Sci., 2009, 111, 132–140. [218] Bhattarai, N.; Edmondson, D.; Veiseh, O.; Matsen, F.A.; Zhang, M. Biomaterials, 2005, 26, 6176–6184. [219] Li, L.; Hsieh Y.-L. Carbohydr. Res., 2006, 341, 374–381. [220] Son, B.; Yeom, B.-Y.; Song, S.H.; Lee, C.-S.; Hwang, T.S. J. Appl. Polym. Sci., 2009, 111, 2892–2899. [221] Yang, D.; Jin, Y.; Zhou, Y.; Ma, G.; Chen, X.; Lu, F.; Nie, J. Macromol. Biosci., 2008, 8, 239–246. [222] Yang, D.; Jin, Y.; Ma, G.; Chen, X.; Lu, F.; Nie, J. J. Appl. Polym. Sci., 2008, 110, 3328–3335.

148

D. Paneva, М. Ignatova, N. Manolova et al.

[223] Subramanian, A.; Vu, D.; Larsen, G.F.; Lin H.-Y. J. Biomater. Sci. Polymer Edn, 2005, 16, 861–873. [224] Desai, K.; Kit, K.; Li, J.; Zivanovic, S. Biomacromolecules, 2008, 9, 1000-1006. [225] Lou, C.-W.; Lin, J.-H.; Yen, K.-C.; Lu, C.-T.; Lee, C.-Y. Textile Res. J., 2008, 78, 254– 257. [226] Vondran, J. L.; Sun, W.; Schauer, C. L. J. Appl. Polym. Sci., 2008, 109, 968–975. [227] Kriegel, C.; Kit, K. M.; McClements, D. J.; Weiss, J. Polymer, 2009, 50, 189–200. [228] Lin, T.; Fang, J.; Wang, H.; Cheng, T.; Wang, X. Nanotechnology, 2006, 17, 3718– 3723. [229] Zhou, Y.; Yang, D.; Nie, J. J. Appl. Polym. Sci., 2006, 102, 5692–5697. [230] Desai, K.; Kit, K. Polymer, 2008, 49, 4046–4050. [231] Duan, B.; Wu, L.; Li, X.; Yuan, X.; Li., X.; Zhang, Y.; Yao, K. J. Biomater. Sci. Polymer Edn, 2007, 18, 95–115. [232] Deitzel, J.M.; Kleinmeyer, J.; Harris, D.; Beck Tan. N.C. Polymer, 2001, 42, 261–272. [233] Penchev, H.; Paneva, D.; Manolova, N.; Rashkov, I. Macromol. Biosci., 2009, 9, 000000, on line, DOI: 10.1002/mabi.200900003. [234] Sanpui, P., Murugadoss, A.; Prasad, P. V. D.; Ghosh, S. S., Chattopadhyay, A., Int. J. Food Microbiology, 2008, 124, 142-146. [235] Murugadoss, A.; Chattopadhyay, A. Nanotechnology, 2008, 19, 1-9. [236] Lu, S.; Gao, W.; Gu, H.-Y. Burns, 2008, 34, 623-628. [237] Twu, Y.-K.; Chen, Y.-W.; Shih, C.-M. Powder Techn., 2008, 185, 251-257. [238] Huang, C.J.; Lin, J.J.; Shieu F.-S. Jap. J. Appl. Phys., 2005, 44, 6332-6340. [239] Ojha, S. S; Stevens, D. R.; Hoffman, T. J.; Stano, K.; Klossner, R.; Scott M. C.; Krause, W.; Clarke, L. I.; Gorga, R. E. Biomacromolecules, 2008, 9, 2523-2529. [240] Ohkawa, K.; Minato, K.-I.; Kumagai, G.; Hayashi, S.; Yamamoto, H. Biomacromolecules, 2006, 9, 3291-3294. [241] Schiffman, J. D.; Schauer, C. L. Biomacromolecules, 2007, 8, 594-601. [242] Schiffman, J. D.; Schauer, C. L. Biomacromolecules, 2007, 8, 2665-2667. [243] Sangsanoh, P.; Supaphol P. Biomacromolecules, 2006, 7, 2710-2714. [244] Jung, K.-H.; Huh, M.-W.; Meng, W.; Yuan, J.; Hyun S. H.; Bae, J.-S.; Hudson, S. M.; Kang, I.-K. J. Appl. Polym. Sci., 2007, 105, 2816–2823. [245] Xu, J.; Zhang, J.; Gao, W.; Liang, H.; Wang, H.; Li, J. Materials Letters, 2009, 63, 658–660. [246] Ignatova, M.; Manolova, N.; Markova, N.; Rashkov, I. Macromol. Biosci., 2009, 9, 102–111. [247] Dinan, B.; Bhattarai, N.; Li, Z.; Zhang, M. J. Undergrad. Research Bioengineering, pp. 1-5 (www.uweb.engr.washington.edu/education/ pdf/Dinan2007.pdf) [248] Min, B.-M.; Lee, S. W.; Lim, J. N.; You, Y.; Lee, T. S.; Kang, P. H.; Park, W. H. Polymer, 2004, 45, 7137–7142. [249] Mi, F.-L.; Sung, H.-W.; Shyu, S.-S.; Su, C.-C.; Peng, C.-K. Polymer, 2003, 44, 6521– 6530. [250] Mi, F.-L. Biomacromolecules, 2005, 6, 975-987. [251] Butler, M. F.; Ng, Y.-F.; Pudney, P. D. A. J. Polym. Sci. Part A: Polym. Chem., 2003, 41, 3941–3953. [252] Ravi Kumar M. N. V., React. Funct. Polymers, 2000, 46, 1–27.

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 149 [253] Cho, J. H.; Kim, S.-H.; Park, K. D.; Jung, M. C.; Yang, W. I.; Han, S. W.; Noh, J. Y.; Lee J. W. Biomaterials, 2004, 25, 5743–5751. [254] Harris, L.G.; Mead, L.; Muller-Oberlander, E.; Richards, R.G. Eur. Cell. Mater., 2005, 10, 24. [255] Vårum, K. M.; Myhr, M. M.; Hjerde, R. J. N.; Smidsrød, O. Carbohydr. Res., 1997, 299, 99-101. [256] Ma, P. Materials Today, 2004, 7, 30-40. [257] Bhattarai, S. R.; Bhattarai, N.; Viswanathamurthi, P.; Yi, H. K.; Hwang, P. H.; Kim, H. Y. J. Biomed. Mater. Res., 2006, 78, 247–257. [258] Yu, S. H.; Mi, F. L.; Shyu, S. S.; Tsai, C. H.; Peng, C. K.; Lai, J. Y. J. Membr. Sci., 2006, 276, 68-80. [259] Engibaryan, L. G.; Chernukhina, A. I.; Gabrielyan, G. A.; Gal’braikh, L. S., Fibre Chem. 2005, 37, 285-288. [260] Neamnark, A.; Rujiravanit, R.; Supaphol, P. Carboh. Polym. 2006, 66, 298–305. [261] Du, J.; Hsieh, Y.-L. Cellulose 2007, 14, 543-552, [262] Chen, S. C.; Wu, Y. C.; Mi, F. L.; Lin, Y. H.; Yu, L. C.; Sung, H. W. J. Control. Release 2004, 96, 285-300. [263] Kogan, G.; Skorik, Y.A.; Žitňanová, I.; Križková, L; Ďuračková, Z; Gomes, C. A. R.; Yatluk, Y. G.; Krajčovi, J. Toxicol. Appl. Pharmacol. 2004, 201, 303–310. [264] Orienti, I.; Luppi, B.; Zecchi, V. J. Cosmet. Sci. 1999, 50, 307–313. [265] Jia, Z.; Shen, D.; Xu, W. Carbohydr. Res. 2001, 333, 1-6. [266] Kim, J. Y.; Lee, J. K.; Lee, T. S.; Park, W. H. Int. J. Biol. Macromol. 2003, 32, 23-27. [267] Chen, R. N.; Wang, G. M.; Chen, C. H.; Ho, H. O.; Sheu, M. T. Biomacromolecules 2006, 7, 1058–1064. [268] Chen, X. G.; Wang, Z.; Liu, W. S.; Park, H. J. Biomaterials 2002, 23, 4609–4614. [269] Zhou, Y.; Yang, D.; Chen, X.; Xu, Q.; Lu, F.; Nie, J. Biomacromolecules 2008, 9, 349354. [270] Du, J.; Hsieh, Y.-L. Nanotechnology 2008, 19, art. no. 125707. [271] Lee, K. Y.; Ha, W. S.; Park, W. H. Biomaterials 1995, 16, 1211–1216. [272] Hirano, S.; Noishiki, Y. J. Biomed. Mater. Res., 1985,19, 413–417. [273] Peesan, M.; Rujiravanit, R.; Supaphol, P. J. Biomater. Sci. Polymer Edn., 2006, 17, 547-565. [274] Harris, J. M., Ed. Poly(ethylene glycol) Chemistry: Biotechnical and Biomedical Applications. Plenum Press: New York, 1992. [275] Tirelli, N.; Lutolf, M. P.; Napoli, A.; Hubbell, J. A. Rev. Mol. Biotechnol., 2002, 90, 3– 15. [276] Skotak, M.; Leonov, A. P.; Larsen, G.; Noriega, S.; Subramanian, A. Biomacromolecules 2008, 9, 1902-1908. [277] Kumeta, K.; Nagashima, I.; Matsui, S.; Mizoguchi, K. J. Appl. Polym. Sci. 2003, 90, 2420-2427. [278] Huang, L.; Nagapudi, K.; Apkarian, R.; Chaikof, E. L. J. Biomater. Sci., Polym. Edn. 2001, 12, 979-993. [279] Jin, H. J.; Fridrikh, S.; Rutledge, G. C.; Kaplan, D. Abstr. Pap. Am. Chem. Soc. 2002, 224, 408-409. [280] Casper, C. L.; Yamaguchi, N.; Kiick K. L.; Rabolt, J. F. Biomacromolecules 2005, 6, 1998-2007.

150

D. Paneva, М. Ignatova, N. Manolova et al.

[281] Katti, D. S.; Robinson, K. W.; Ko, F. K.; Laurencin, C. T. J. Biomed. Mater. Res. - Part B Appl. Biomaterials, 2004, 70, 286-296. [282] Verreck, G.; Chun, I.; Rosenblatt, J.; Peeters, J.; Van Dijck, A.; Mensch, J.; Noppe, M.; Brewster, M. E. J. Controlled Release 2003, 92, 349-360. [283] Melaiye, A.; Sun, Z.; Hindi, K.; Milsted, A.; Ely, D.; Reneker, D. H.; Tessier, C. A.; Youngs, W. J. J. Am. Chem. Soc. 2005, 127, 2285-2291. [284] Yu, D.-G.; Teng, M.-Y.; Chou, W.-L.; Yang, M.-C. J. Membr. Sci. 2003, 225, 115-123. [285] Son, W. K.; Youk, J. H.; Lee T. S.; Park, W. H. Macromol. Rapid Commun. 2004, 25, 1632-1637. [286] Xu, X.; Yang, Q.; Wang, Y.; Yu, H.; Chen, X.; Jing, X. Eur. Polym. J. 2006, 42, 20812087. [287] Hong, K. H.; Park, J. L.; Sul, I. H.;. Youk, J. H; Kang, T. J. J. Polym. Sci. Part B: Polym. Physics. 2006, 44, 2468-2474. [288] Hong K. H., Polym. Engineer. Sci. 2007, 47, 43-49. [289] Ignatova, M.; Markova, N.; Manolova, N.; Rashkov, I. J. Biomater. Sci. Polymer Edn., 2008, 19, 373-386. [290] Kenawy, E.-R.; Abdel-Fattah, Y. R. Macromol. Biosci. 2002, 2, 261-266. [291] Ji, Y.; Ghosh, K.; Li, B.; Sokolov, J. C.; Clark R. A. F.; Rafailovich, M. H. Macromol. Biosci. 2006, 6, 811-817. [292] Rho, K. S.; Jeong, L.; Lee, G.; Seo, B.-M.; Park, Y. J.; Hong, S.-D.; Roh, S.; Cho, J. J.; Park, W. H.; Min, B.-M. Biomaterials 2006, 27, 1452-1461. [293] Khil, M. S.; Cha, D. I.; Kim, H. Y.; Kim I. S.; Bhattarai, N. J. Biomed. Mater. Res. 2003, 67B, 675-679. [294] Franklin, T. J.; Snow, G. A. Biochemistry of Antimicrobial Action, Chapman and Hall: London, 1981; pp 58-60. [295] Teixeira, A. I.; Abrams, G. A.; Bertics, P. J.; Murphy, C. J.; Nealey, P F. J. Cell Sci. 2003, 116, 1881–1892. [296] Lee, C. H.; Shin, H. J.; Cho, I. H.; Kang, Y. M.; Kim, I. A.; Park, K. D.; Shin, J. W. Biomaterials 2005, 26, 1261–1270. [297] Li, W. J.; Laurencin, C. T.; Caterson, E. J.; Tuan, R. S.; Ko, F. K. J. Biomed. Mater. Res. 2002, 60, 613–621. [298] Neamnark, A.; Sanchavanakit, N.; Pavasant, P.; Rujiravanit, R.; Supaphol, P. Eur. Polym. J., 2008, 44, 2060-2067. [299] Jiang, H.; Fang, D.; Hsiao, B.; Chu, B.; Chen, W. J. Biomater. Sci., Polym. Ed. 2004, 15, 797–811. [300] Ouchi, T.; Nishizawa, H.; Ohya, Y. Polymer 1998, 39, 5171–5175. [301] Kim, S.H.; Kim, S.; Nair, S.; Moore, E. Macromolecules, 2005, 38, 3719-3723. [302] Hennink, W. E.; van Nostrum, C. F. Adv. Drug Delivery Rev., 2002, 54, 13-36. [303] Xu, X. US Patent WO 2005/026398 A2, 2004. [304] Baker, D. A.; Brown Ph. J. Polym. Preprints, 2003, 44, 118-119. [305] Gupta, R.; Trenor, S. R.; Long, T. E.; Wilkes, G. L. Macromolecules, 2004, 37, 92119218. [306] Muzarelli, R. A. A.; Barontini, G.; Rochetti, R. Biotechnolol. Bioeng., 1975, 18, 14451454. [307] Richards, F. M.; Knowles, J. R. J. Mol. Biol., 1968, 37, 231-233. [308] Monteiro Jr., O. A. C.; Airoldi C. Int. J. Biol. Macromol., 1999, 26, 119-128.

Novel Chitosan–Containing Micro- and Nanofibrous Materials by Electrospinning… 151 [309] Yancheva, E.; Paneva, D.; Maximova, V.; Mespouille, L.; Dubois, Ph.; Manolova, N.; Rashkov, I. Biomacromolecules, 2007, 8, 976-984. [310] Koetz, J.; Kosmella, S. Polyelectrolytes and Nanoparticles. Springer-Verlag Berh Heidelberg 2007. [311] Chavasit, V.; Klenzle-Sterzer, C.; Torres, J.A. Polym. Bull., 1988, 19, 223-230. [312] Dumitriu, S.; Chornet, E. Adv. Drug Deliv. Rev., 1998, 31, 223-246. [313] Wang, X.; Kim, Y.-G.; Drew, Ch.; Ku, B.-Ch.; Kumar, J.; Samuelson, L.A. Nano Lett., 2004, 4, 331-334. [314] Decher, G., Hong, J.D., Schmitt, J. Adv. Materials, 1991, 3, 617-619. [315] Decher, G. Science, 1997, 277, 1232-1237. [316] Crisp, M. T.; Kotov, N. A. Nano Lett., 2003, 3, 173-177. [317] Caruso, F.; Shi, X.; Caruso, R. A.; Susha, A. Adv. Materials, 2001, 13, 740-744. [318] Mayya, K. S.; Gittins, D. I.; Dibaj, A. M.; Caruso, F. Nano Lett., 2001, 1, 727-730. [319] Li B., Shen L., Liu X., Zhang S., Wu C., Liu W. Mater. Sci. Eng. A, 2004, 364: 324332. [320] Gittins, D. I.; Caruso, F. Adv. Mater., 2000, 12, 1947-1949. [321] Caruso, F.; Caruso, R. A.; Mohwald, H. Science, 1998, 282, 1111-1114. [322] Haack, J. M.; Lenk, W.; Lehmann, D.; Lunkwitz, K. J. Membr. Sci., 2001, 184, 233243. [323] Lajimi, R. H.; Abdallah, A. B.; Ferjani, E.; Roudesli, M.S.; Deratani, A. Desalination, 2004, 163,193-202. [324] Wang, X.; Fang, D.; Yoon, K.; Hsiao, B. S.; Chu, B. J. Membr. Sci., 2006, 278, 261268. [325] Wang, X.; Chen, X.; Yoon, K.; Fang, D.; Hsiao, B. S.; Chu, B. Environm. Sci. Techn., 2005, 39, 7684-7691. [326] Ding, B.; Fujimoto, K.; Shiratoroi, S. Thin Solid Films, 2005, 491, 23-28. [327] Zezin, A. B.; Lutsenko, V .V.; Rogacheva, V. B.; Aleksina, O. A.; Kalyuzhnaya, R. I.; Kabanov, V. A.; Kargin, V. A. Polymer Science U.S.S.R., 1972, 14, 857-865. [328] Ritcharoen, W.; Supaphol, P.; Pavasant, P. Eur. Polym J., 2008, 44, 3963-3968. [329] Müller, K.; Quinn, J. F.; Johnston, A. P. R.; Becker, M.; Greiner, A.; Caruso, F. Chem. Mater., 2006, 18, 2397-2403. [330] Ge, L.; Pan, C.; Chen, H.; Wang, X.; Wang, C.; Gu, Z. Colloids Surf. A, 2007, 293, 272-277. [331] Chen, Z.; Mo, X.; Qing, F. Matter. Lett., 2007, 61, 3490-3494. [332] Chunder, A.; Sarkar, S.; Yu, Y.; Zhai, L. Colloids Surf. B: Biointerf., 2007, 58, 172– 179. [333] Chen, Z.; Mo, X.; He, Ch.; Wang, H. Carbohydr. Polym., 2008, 72, 410–418. [334] Penchev, H.; Paneva, D.; Manolova, N.; Rashkov, I. Macromol. Rapid Commun., 2008, 29, 677-681. [335] Michaels, A. S.; Miekka, R. G. J. Chem. Phys., 1961, 65, 1765-1773. [336] Rashkov, I.; Penchev, H unpublished results. [337] Okuzaki, H.; Takahashi, T.; Miyajima, N.; Suzuki, Y.; Kuwabara, T. Macromolecules, 2006, 39, 4276-4278. [338] Okuzaki, H.; Takahashi, T.; Hara, Y.; Yan, H. J. Polym. Sci.: Part B: Polym. Phys., 2008, 46, 305–310.

In: Nanofibers: Fabrication, Performance, and Applications ISBN 978-1-60741-947-1 Editors: W. N. Chang © 2009 Nova Science Publishers, Inc.

Chapter 4

A NOVEL APPROACH FOR ANALYSIS OF PROCESSING PARAMETERS IN ELECTROSPINNING OF NANOFIBERS M. Ziabari1, V. Mottaghitalab1 and A. K. Haghi1,2 1

University of Guilan, Rasht, Iran 2 University of Ottawa, Canada

ABSTRACT The precise control of fiber diameter during electrospinning is very crucial for many applications. A systematic and quantitative study on the effects of processing variables enables us to control the properties of electrospun nanofibers. In this contribution, response surface methodology (RSM) was employed to quantitatively investigate the simultaneous effects of four of the most important parameters, namely solution concentration (C), spinning distance (d), applied voltage (V) and volume flow rate (Q) on mean fiber diameter (MFD) as well as standard deviation of fiber diameter (StdFD) in electrospinning of polyvinyl alcohol (PVA) nanofibers.

Keywords: Electrospinning, Nanofibers, Fiber diameter, Processing variables, Response surface methodology

INTRODUCTION Electrospinning is a novel and efficient method by which fibers with diameters in nanometer scale entitled as nanofibers, can be achieved. In electrospinning process, a strong electric field is applied on a droplet of polymer solution (or melt) held by its surface tension at the tip of a syringe's needle (or a capillary tube). As a result, the pendent drop will become highly electrified and the induced charges are distributed over its surface. Increasing the intensity of electric field, the surface of the liquid drop will be distorted to a conical shape known as the Taylor cone [1]. Once the electric field strength exceeds a threshold value, the repulsive electric force dominates the surface tension of the liquid and a stable jet emerges

154

M. Ziabari, V. Mottaghitalab and A. K. Haghi

from the cone tip. The charged jet is then accelerated toward the target and rapidly thins and dries as a result of elongation and solvent evaporation. As the jet diameter decreases, the surface charge density increases and the resulting high repulsive forces split the jet to smaller jets. This phenomenon may take place several times leading to many small jets. Ultimately, solidification is carried out and fibers are deposited on the surface of the collector as a randomly oriented nonwoven mat [2]-[5]. Figure 1 shows a schematic illustration of electrospinning setup.

Figure 1. Electrospinning setup [6].

Featuring various outstanding properties such as very small fiber diameters, large surface area per mass ratio [3], high porosity along with small pore sizes [7], flexibility, and superior mechanical properties [8], electrospun nanofiber mats have found numerous applications in biomedical (tissue engineering [9]-[11], drug delivery [12], [13], and wound dressing [14], [15]), protective clothing [7], filtration [16], reinforcement in composite materials [8], [17], and micro-electronics (battery [18], supercapacitors [19], transistors [20], sensors [21], and display devices [22]). Morphology of electrospun nanofibers such as fiber diameter, depend on many parameters which are mainly divided into three categories: solution properties (solution viscosity, solution concentration, polymer molecular weight, and surface tension), processing conditions (applied voltage, volume flow rate, spinning distance, and needle diameter), and ambient conditions (temperature, humidity, and atmosphere pressure) [23]. As mentioned earlier, electrospun nanofibers have numerous applications some of which have been commercialized. Most of these applications require nanofibers with desired properties suggesting the importance of the process control. This end may not be achieved unless having a comprehensive outlook of the process and quantitative study of the effects of governing parameters which makes the control of the process possible. In addition, qualitative description of the experimental observations are not adequate to derive general conclusions and either the equations governing behavior of the system must be found or appropriate empirical models need be presented. In Ziabicki's words, “in the language of science ‘to explain’ means to put forward a quantitative model which is consistent with all the known date and capable of predicting new fact” [24]. Employing a model to express the influence of electrospinning parameters will help us obtain a simple and systematic way for presenting the

A Novel Approach for Analysis of Processing Parameters…

155

effects of variables thereby enabling the control of the process. Furthermore, it allows us to predict the results under new combination of parameters. Hence, without conducting any experiments, one can easily estimate features of the product in unknown conditions. That's to say, a model tells us to what extent the output of a system will change if one or more parameters increased or decreased. This is very helpful and leads to a detailed understanding of the process and the effects of parameters. Despite the surge in attention to the electrospinning process, a few investigations have addressed the quantitative study of the effects of the parameters which has hindered the control of the process. Changing the behavior of materials in nano-scale, presence of electric field, branching of the jet, random orientation of fibers, etc. made the analysis of the process extremely complex and difficult that to date there has been no reliable theory capable of describing the phenomenon. Furthermore, the development of an empirical model has also been impeded due to the lack of systematic and characterized experimentations with appropriate designs. Adding to the difficulty is the number of parameters involving in the electrospinning process and the interactions between them which made it almost impossible to investigate the simultaneous effects of all variables. Affecting the characteristics of the final product such as physical, mechanical and electrical properties, fiber diameter is one of the most important structural features in electrospun nanofiber mats. Podgorski et al. [25] indicated that filters made of fibers with smaller diameters have higher filtration efficiencies. This was also proved by the work of Qin et al. [16]. Ding et al. [26] reported that sensitivity of sensors increase with decreasing the mean fiber diameter – due to the higher surface area. In the study on designing polymer batteries consisting of electrospun PVdF fibrous electrolyte by Kim et al. [27], it was demonstrated that lower mean fiber diameter results in a higher electrolyte uptake thereby increased ionic conductivity of the mat. Moroni et al. [28 found fiber diameters of electrospun PEOT/PBT scaffolds influencing on cell seeding, attachment, and proliferation. They also studied the release of dye incorporated in electrospun scaffolds and observed that with increasing fiber diameter, the cumulative release of the dye (methylene blue) decreased. Carbonization and activation conditions as well as the structure and properties of the ultimate carbon fibers are also affected by the diameters of the precursor PAN nanofibers [29]. Consequently, precise control of the electrospun fiber diameter is very crucial. Sukigara et al. [30] employed response surface methodology (RSM) to model mean fiber diameter of electrospun regenerated Bombyx mori silk with electric field and concentration at two spinning distances. They applied a full factorial experimental design at three levels of each parameter leading to nine treatments of factors and used a quadratic polynomial to establish a relationship between mean fiber diameter and the variables. Increasing the concentration at constant electric field resulted in an increase in mean fiber diameter. Different impacts for the electric field were observed depending on solution concentration. Trend of the effects of the two parameters on mean fiber diameter varied with changing the spinning distance which suggests the presence of interaction and coupling between the parameters. Gu et al. [31] and Gu et al. [32] also exploited the RSM for quantitative study of PAN and PDLA respectively. The only difference observed in the procedure was the use of four levels of concentration in the case of PAN. They included the standard deviation of fiber diameter in their investigations by which they were able to provide additional information regarding the morphology of electrospun nanofibers and its variations at different conditions.

156

M. Ziabari, V. Mottaghitalab and A. K. Haghi

Furthermore, they analyzed the significance of the factors in the models in order to understand the level of influence of each parameter. In the case of PAN, voltage as well as its interaction with concentration had no significant effects on both mean and standard deviation of fiber diameter. Hence, they eliminated the terms corresponding to these factors thereby obtained simplified quadratic models according to which mean and standard deviation of fiber diameter increased with polymer concentration. On the contrary, both voltage and its interaction with concentration were found to be significant in the case of PDLA. However, the effect of polymer concentration was more pronounced. Increasing voltage at constant concentration favored thinner fiber formation which gained momentum with increasing concentration. Fibers with more uniform diameters (less standard deviation) were obtained at higher applied voltage or concentration. In the most recent investigation in this field, Yördem et al. [33] utilized RSM to correlate mean and coefficient of variation (CV) of electrospun PAN nanofibers to solution concentration and applied voltage at three different spinning distances. They employed a face-centered central composite design (FCCD) along with a full factorial design at two levels resulting in 13 treatments at each spinning distance. A cubic polynomial was then used to fit the data in each case. As with previous studies, fiber diameter was very sensitive to changes in solution concentration. Voltage effect was more significant at higher concentrations demonstrating the interaction between parameters. Despite high reported R 2 values, the presented models seemed to be inefficient and uncertain. Some terms in the models had very high p-values. For instance, in modeling the mean fiber diameter, p-value as high as 0.975 was calculated for cubic concentration term at spinning distance of 16 cm, where half of the 2 values which were not reported in terms had p-values more than 0.8. This results in low Rpred their study and after calculating by us were found to be almost zero in many cases suggesting the poor prediction ability of their models. As it was mentioned by the previous authors, there are some interactions between electrospinning parameters. In the past studies, however, they only investigated the simultaneous effects of two variables; therefore they were unable to thoroughly capture the interactions which exist between the parameters. For instance, Sukigara et al. [30] and Yördem et al. [33] both agreed that spinning distance has a significant influence on fiber diameter and that this effect varies when solution concentration and/or applied voltage altered. However, they could not describe their findings in terms of quantitative relationships. Hence, the presented models suffer from lack of comprehensiveness. In addition, in every research where modeling of a process is targeted, the obtained models need to be evaluated with a set of test data which were not used in establishing the relationships. Otherwise, the effectiveness of the models will not be guaranteed and there will always be an uncertainty in the prediction of the models in new conditions. Hence, it is possible for a model very efficient in describing experimental data, to present unsatisfactory prediction results. In none of the previous works, however, the presented models were evaluated with a series of test data. Therefore, their models may not generalize well to new data and their prediction ability is obscure. In this contribution for the first time, the simultaneous effects of four electrospinning parameters (solution concentration, spinning distance, applied voltage, and volume flow rate) on mean and standard deviation of polyvinyl alcohol (PVA) fiber diameter were systematically investigated. PVA, the largest volume synthetic water-soluble polymer

A Novel Approach for Analysis of Processing Parameters…

157

produced in the world, is commercially manufactured by the hydrolysis of polyvinyl acetate. The excellent chemical resistance and physical properties of PVA along with non-toxicity and biodegradability have led to its broad industrial applications such as textile sizing, adhesive, paper coating, fibers, and polymerization stabilizers [34], [35]. Several patents reported process for production of ultrahigh tensile strength PVA fibers comparable to Kevlar® [36][38]. PVA has found many applications in biomedical uses as well due to its biocompatibility [39]. For instance, PVA hydrogels were used in regenerating articular cartilages [40], [41], artificial pancreas [42], and drug delivery systems [43], [44]. More recently, PVA nanofibers were electrospun and used as a protein delivery system [45], retardation of enzyme release [45] and wound dressing [46]. The objective of this paper is to use RSM to establish quantitative relationships between electrospinning parameters and mean and standard deviation of fiber diameter as well as to evaluate the effectiveness of the empirical models with a set of test data.

EXPERIMENTAL Solution Preparation and Electrospinning PVA with molecular weight of 72000 g/mol and degree of hydrolysis of >98% was obtained from Merck and used as received. Distilled water as solvent was added to a predetermined amount of PVA powder to obtain 20 ml of solution with desired concentration. The solution was prepared at 80°C and gently stirred for 30 min to expedite the dissolution. After the PVA had completely dissolved, the solution was transferred to a 5 ml syringe and became ready to electrospin. The experiments were carried out on a horizontal electrospinning setup shown schematically in Figure 1. The syringe containing PVA solution was placed on a syringe pump (New Era NE-100) used to dispense the solution at a controlled rate. A high voltage DC power supply (Gamma High Voltage ES-30) was used to generate the electric field needed for electrospinning. The positive electrode of the high voltage supply was attached to the syringe needle via an alligator clip and the grounding electrode was connected to a flat collector wrapped with aluminum foil where electrospun nanofibers were accumulated to form a nonwoven mat. The electrospinning was carried out at room temperature. Subsequently, the aluminum foil was removed from the collector. A small piece of mat was placed on the sample holder and gold sputter-coated (Bal-Tec). Thereafter, the morphology of electrospun PVA fibers was observed by an environmental scanning electron microscope (SEM, Phillips XL-30) under magnification of 10000X. For each specimen, fiber diameter distribution was determined from the SEM micrograph based on 100 measurements of random fibers. A typical SEM micrograph of electrospun nanofiber mat and its corresponding diameter distribution are shown in Figure 2.

158

M. Ziabari, V. Mottaghitalab and A. K. Haghi

(a)

(b)

Figure 2. (a) a typical SEM micrograph of electrospun nanofiber mat, (b) its corresponding diameter distribution.

CHOICE OF PARAMETERS AND RANGE As it was mentioned earlier in this paper, the number of variables which have the potential to alter the electrospinning process is numerous. Hence, investigating all of them in the framework of one single research would almost be impossible. However, some of these parameters can be held constant during experimentation. For instance, performing the experiments in a controlled environmental condition, which is concerned in this study, the ambient parameters (i.e. temperature, air pressure, and humidity) are kept unchanged. Solution viscosity is affected by polymer molecular weight, solution concentration, and temperature. For a particular polymer (constant molecular weight) at a fixed temperature, solution concentration would be the only factor influencing the viscosity. In this circumstance, the effect of viscosity could be determined by the solution concentration. Therefore, there would be no need for viscosity to be considered as a separate parameter. In this regard, solution concentration (C), spinning distance (d), applied voltage (V), and volume flow rate (Q) were selected to be the most influential parameters in electrospinning of PVA nanofibers as for the purpose of this study. The next step is to choose the region of interest – that is the ranges over which these factors are varied. Process knowledge, which is a combination of practical experience and theoretical understanding, is required to fulfill this step. The aim is here to find an appropriate range for each parameter where dry, bead-free, stable, and continuous fibers without breaking up to droplets are obtained. This goal could be achieved by conducting a set of preliminary experiments while having the previous works in mind along with utilizing the reported relationships. The relationship between intrinsic viscosity ( [η ] ) and molecular weight (M) is given by the well-known Mark-Houwink-Sakurada equation as follows: [η ] = KM a

(1)

A Novel Approach for Analysis of Processing Parameters…

159

where K and a are constants for a particular polymer-solvent pair at a given temperature [47]. For the PVA with molecular weight in the range of 69000 g/mol <M<690000 g/mol in water at room temperature, K=6.51 and a=0.628 were found by Tacx et al. [48]. Using these constants in the equation, the intrinsic viscosity for PVA in this study (molecular weight of 72000 g/mol) were calculated to be [η ] =0.73. Polymer chain entanglements in a solution can be expressed in terms of Berry number (B), which is a dimensionless parameter and defined as the product of intrinsic viscosity and polymer concentration ( B = [η ]C ) [49]. At each molecular weight, there is a minimum concentration at which the polymer solution cannot be electrospun. Koski et al. [50] observed that B>5 is required to form stabilized fibrous structures in electrospinning of PVA. On the other hand, they reported the formation of flat fibers at B>9. Therefore, the appropriate range in this case could be found within 50.4 ml/h incurred formation of droplets along with fibers. As a result, 0.2 ml/h≤Q≤0.4 ml/h was chosen as the favorable range of flow rate in this study.

EXPERIMENTAL DESIGN In order to discover some facts about a process or to investigate the behavior of a system, experiments are usually performed by researchers in virtually all fields of inquiry. Experiments often involve several factors and the objective of the experimenter is to determine how these factors affect the output (response) of the system. But how many observations are required for the purpose of a research? Or taking the limiting elements into consideration how is it possible to achieve the most information about the system? These are

160

M. Ziabari, V. Mottaghitalab and A. K. Haghi

the very first questions bearing in mind of the experimenter and cannot be responded in a couple of sentences. In fact, the aim of experimental design is to provide reasonable and scientific answers to such questions. In other words, experimental design is a sequential steps taken to insure that data will be obtained in the most efficient form for the problem being considered and that the analysis will lead to valid statistical inferences [51], [52]. Now, let's consider a process in which several factors affect a response of the system. In this case, a conventional strategy of experimentation, which is extensively used in practice, is the one-factor-at-a-time approach. The major disadvantage of this design is that it fails to consider any possible interaction between the factors – the failure of one factor to produce the same effect on the response at different levels of another factor. As mentioned earlier in the paper, interactions exist between electrospinning parameters making this approach an inappropriate choice for the case of present work. The correct strategy to dealing with several factors is to use a full factorial design in which factors are all varied together, therefore all possible combinations of the levels of the factors are investigated. The advantages of full factorial design are as follows: it is very efficient, makes the most use of the experimental data and takes into account the interactions between factors [51], [52]. It is trivial that in order to draw a line at least two points and for a quadratic curve at least three points are required. Hence, three levels were selected for each parameter in this study so that it would be possible to use quadratic models. These levels were chosen equally spaced. A full factorial experimental design with four factors (solution concentration, spinning distance, applied voltage, and flow rate) each at three levels (34 design) were employed resulting in 81 treatment combinations. This design is shown in Figure 3. C -1

0

1

1

d 0 1 -1

0 -1

0

1 -1

Q

V

Figure 3. 34 full factorial experimental design used in this study.

-1, 0, and 1 are coded variables corresponding to low, intermediate and high levels of each factor respectively. The coded variables (xj) were calculated using Equation from natural variables (ξi). The indices 1 to 4 represent solution concentration, spinning distance, applied voltage, and flow rate respectively. In addition to experimental data, 15 treatments inside the design space were selected as test data and used for evaluation of the models. The natural and coded variables for experimental data (numbers 1-81) as well as test data (numbers 82-96) are listed in Table 6.

A Novel Approach for Analysis of Processing Parameters… xj =

ξ j − [ξ hj + ξ lj ] / 2 [ξ hj − ξ lj ] / 2

161

(2)

RESPONSE SURFACE METHODOLOGY Thus far, the mechanism of some scientific phenomena has been understood well and models depicting the physical behavior of the system have been drawn in the form of mathematical relationships. However, there are numerous processes at the moment which have not sufficiently understood to permit the theoretical approach. Response surface methodology (RSM) is a combination of mathematical and statistical techniques useful for empirical modeling and analysis of such systems. The application of RSM is in situations where several input variables are potentially influence some performance measure or quality characteristic of the process – often called responses. The relationship between the response (y) and k input variables (ξ1,ξ2,...,ξk) could be expressed in terms of mathematical notations as follows: y = f (ξ1 , ξ 2 ,...,ξ k )

(3)

where the true response function f is unknown. It is often convenient to use coded variables (x1,x2,..,xk) instead of natural (input) variables. The response function will then be: y = f ( x1 , x2 ,..., xk )

(4)

Since the form of true response function f is unknown, it must be approximated. Therefore, the successful use of RSM is critically dependent upon the choice of appropriate function to approximate f. Low-order polynomials are widely used as approximating functions. First order (linear) models are unable to capture the interaction between parameters which is a form of curvature in the true response function. A second order (quadratic) model will likely to perform well in these circumstances. In general, the quadratic model is in the form of: k

k

j =1

j =1

k

y = β 0 + ∑ β j x j + ∑ β jj x 2j + ∑∑ β ij xi x j + ε

(5)

i< j j =2

where ε is the error term in the model. The use of polynomials of higher order is also possible but infrequent. The βs are a set of unknown coefficients needed to be estimated. In order to do that, the first step is to make some observations on the system being studied. The model in Equation (5) may now be written in matrix notations as: y = Xβ + ε

(6)

162

M. Ziabari, V. Mottaghitalab and A. K. Haghi

where y is the vector of observations, X is the matrix of levels of the variables, β is the vector of unknown coefficients, and ε is the vector of random errors. Afterwards, method of least squares, which minimizes the sum of squares of errors, is employed to find the estimators of the coefficients ( βˆ ) through: βˆ = ( X ′X ) −1 X ′y

(7)

The fitted model will then be written as: yˆ = Xβˆ

(8)

Finally, response surfaces or contour plots are depicted to help visualize the relationship between the response and the variables and see the influence of the parameters [53], [54]. As you might notice, there is a close connection between RSM and linear regression analysis [55]. In this study RSM was employed to establish empirical relationships between four electrospinning parameters (solution concentration, spinning distance, applied voltage, and flow rate) and two responses (mean fiber diameter and standard deviation of fiber diameter). Coded variables were used to build the models. The choice of three levels for each factor in experimental design allowed us to take the advantage of quadratic models. Afterwards, the significance of terms in each model was investigated by testing hypotheses on individual coefficients and simpler yet more efficient models were obtained by eliminating statistically unimportant terms. Finally, the validity of the models was evaluated using the 15 test data. The analyses were carried out using statistical software Minitab 15.

RESULTS AND DISCUSSION After the unknown coefficients (βs) were estimated by least squares method, the quadratic models for the mean fiber diameter (MFD) and standard deviation of fiber diameter (StdFD) in terms of coded variables are written as: MFD = 282.031 + 34.953x1 + 5.622x2 − 2.113x3 + 9.013x4 − 11.613x12 − 4.304x22 − 15.500x32

(9)

− 0.414x + 12.517x1 x2 + 4.020x1 x3 − 0.162x1 x4 + 20.643x2 x3 + 0.741x2 x4 + 0.877x3 x4 2 4

StdFD = 36.1574 + 4.5788 x1 − 1.5536 x2 + 6.4012 x3 + 1.1531x4 − 2.2937 x12 − 0.1115 x22 − 1.1891x32 + 3.0980 x42

(10)

− 0.2088 x1 x2 + 1.0010 x1 x3 + 2.7978 x1 x4 + 0.1649 x2 x3 − 2.4876 x2 x4 + 1.5182 x3 x4

In the next step, a couple of very important hypothesis-testing procedures were carried out to measure the usefulness of the models presented here. First, the test for significance of the model was performed to determine whether there is a subset of variables which

A Novel Approach for Analysis of Processing Parameters…

163

contributes significantly in representing the response variations. The appropriate hypotheses are: H 0 : β1 = β 2 = L = β k (11) H 1 : β j ≠ 0 for at least one j The F statistics of this test along with the p-values (measure of statistical significance) for both models are shown in Table 1. Table 1. Summary of the results from statistical analysis of the models

MFD StdFD

F

p-value

106.02 42.05

0.000 0.000

R2 95.74% 89.92%

2 Radj

94.84% 87.78%

2 Rpred

93.48% 84.83%

The p-values of the models are very small (almost zero), therefore it is concluded that the null hypothesis is rejected in both cases suggesting that there are some significant terms in 2 2 , and Rpred . R 2 is a each model. There are also included in Table 1, the values of R 2 , Radj measure of the amount of response variation which is explained by the variables and will always increase when a new term is added to the model – regardless of whether the inclusion 2 of the additional term is statistically significant or not. Radj is the R 2 adjusted for the number of terms in the model, therefore it will increase only if the new terms improve the model and 2 decreases if unnecessary terms are added. Rpred implies how well the model predicts the 2 indicate how well the model fits the response for new observations, whereas R 2 and Radj

experimental data. The R 2 values demonstrate that 95.74% of MFD and 89.92% of StdFD 2 values are 94.84% and 87.78% for MFD and StdFD are explained by the variables. The Radj 2 values respectively, which account for the number of terms in the models. Both R 2 and Radj

indicate that the models fit the data very well. The slight difference between the values of R 2 2 2 and Radj suggests that there might be some insignificant terms in the models. Since the Rpred 2 , the models does not appear to be overfit and values are so close to the values of R 2 and Radj

have very good predictive ability. The second testing hypothesis performed in this study was the test on individual coefficients, which would be useful in determining the value of the variables in the models. The hypotheses for testing the significance of any individual coefficient are: H0 : β j = 0 H1 : β j ≠ 0

(12)

Since the model might be more effective with inclusion or perhaps exclusion of one or more variables, by means of this test, we could evaluate the value of each term in the model and eliminate the statistically insignificant terms, thereby obtain more efficient models. The

164

M. Ziabari, V. Mottaghitalab and A. K. Haghi

results of this test for the models of MFD and StdFD are summarized in Table 2 and Table 3 respectively. Table 2. The test on individual coefficients for the model of mean fiber diameter Term Constant C d V Q C2 d2 V2 Q2 Cd CV CQ dV dQ VQ

Coef 282.031 34.953 5.622 -2.113 9.013 -11.613 -4.304 -15.500 -0.414 12.517 4.020 -0.162 20.643 0.741 0.877

T 102.565 31.136 5.008 -1.882 8.028 -5.973 -2.214 -7.972 -0.213 9.104 2.924 -0.118 15.015 0.539 0.638

p-value 0.000 0.000 0.000 0.064 0.000 0.000 0.030 0.000 0.832 0.000 0.005 0.906 0.000 0.592 0.526

Table 3. The test on individual coefficients for the model of standard deviation of fiber diameter Term Constant C D V Q C2 d2 V2 Q2 Cd CV CQ dV dQ VQ

Coef 36.1574 4.5788 -1.5536 6.4012 1.1531 -2.2937 -0.1115 -1.1891 3.0980 -0.2088 1.0010 2.7978 0.1649 -2.4876 1.5182

T 39.381 12.216 -4.145 17.078 3.076 -3.533 -0.172 -1.832 4.772 -0.455 2.180 6.095 0.359 -5.419 3.307

p-value 0.000 0.000 0.000 0.000 0.003 0.001 0.864 0.072 0.000 0.651 0.033 0.000 0.721 0.000 0.002

As depicted, the terms Q 2 , CQ , dQ , and VQ in the model of MFD and d 2 , Cd , and dV in the model of StdFD have very high p-values, therefore they do not contribute significantly in representing the variation of the corresponding response. Eliminating these

A Novel Approach for Analysis of Processing Parameters…

165

terms will enhance the efficiency of the models. Recalculating the unknown coefficients, the new models are then given by: MFD = 281.755 + 34.953 x1 + 5.622 x2 -2.113 x3 + 9.013 x4 − 11.613 x12 − 4.304 x22 − 15.500 x32

(13)

+ 12.517 x1 x2 + 4.020 x1 x3 + 20.643 x2 x3 StdFD = 36.083 + 4.579 x1-1.554 x2 + 6.401x3 + 1.153 x4 − 2.294 x12 − 1.189 x32 + 3.098 x42

(14)

+ 1.001x1 x3 + 2.798 x1 x4 − 2.488 x2 x4 + 1.518 x3 x4

in terms of coded variables and: MFD = 10.3345 + 48.7288C − 22.7420d + 7.9713V + 90.1250Q -2.9033C 2 − 0.1722d 2 − 0.6120V 2 + 1.2517Cd + 0.4020CV + 0.8257 dV

(15)

StdFD = −1.8823 + 7.5590C + 1.1818d + 1.2709V − 300.3410Q -0.5734C 2 − 0.0476V 2 + 309.7999Q 2

(16)

+ 0.1001CV + 13.9892CQ − 4.9752dQ + 3.0364VQ

in terms of natural (uncoded) variables. The results of test for significance as well as R 2 , 2 2 Radj , and Rpred for the new models are given in Table 4. It is obvious that the p-values for the new models are close to zero indicating the existence of some significant terms in each model. Comparing the results of this table with Table 1, the F statistic increased for the new models, indicating the improvement of the models after eliminating the insignificant terms. 2 2 and Rpred increased a great deal for the Despite the slight decrease in R 2 , the values of Radj new models. As it was mentioned earlier in the paper, R 2 will always increase with the number of terms in the model. Therefore, the smaller R 2 values were expected for the new models, due to the fewer terms. However, this does not necessarily mean that the pervious 2 , which provides a more useful tool for models were more efficient. Looking at the tables, Radj comparing the explanatory power of models with different number of terms, increased after eliminating the unnecessary variables. Hence, the new models have the ability to better 2 explain the experimental data. Due to higher Rpred values obtained, the new models also have higher prediction ability. In other words, eliminating the insignificant terms, simpler models were obtained which not only better explain the experimental data, but also are more powerful in predicting new conditions. The test for individual coefficients was performed again for the new models. This time, as it was anticipated, no terms had higher p-value than expected, which need to be eliminated. Here is another advantage of removing unimportant terms. The values of T statistic increased

166

M. Ziabari, V. Mottaghitalab and A. K. Haghi

for the terms already in the models implying that their effects on the response became stronger. Table 4. Summary of the results from statistical analysis of the models after eliminating the insignificant terms

MFD StdFD

F

p-value

R2

155.56 55.61

0.000 0.000

95.69% 89.86%

2 Radj

95.08% 88.25%

2 Rpred

94.18% 86.02%

Now that the relationships have been developed, the test data were used to investigate the prediction ability of the models. Root mean square errors (RMSE) between the calculated responses (Ci) and real responses (Ri) were determined using equation for experimental data as well as test data for the sake of evaluation of both MFD and StdFD models and are listed in Table 5. The models present acceptable RMSE values for test data indicating the ability of the models to generalize well the experimental data to predicting new conditions. Although the values of RMSE for the test data are slightly higher than experimental data, these small discrepancies were expected since it is almost impossible for an empirical model to express the test data as well as experimental data and higher errors are often obtained when new data presented to the models. Hence, the results imply the acceptable prediction ability of the models. n

RMSE =

∑ (C

i

− Ri ) 2

i =1

(17)

n

Table 5. RMSE values of the models for the experimental and test data

MFD StdFD

Experimental data 7.489 2.493

Test data 10.647 2.890

RESPONSE SURFACES FOR MEAN FIBER DIAMETER Solution Concentration Increasing polymer concentration will result in greater polymer chain entanglements. This causes the viscoelastic force to increase enabling the charged jet to withstand a larger electrostatic stretching force leading to a larger diameter of fibers. A monotonic increase in MFD with concentration was observed in this study as shown in Figure 4 (a), (b), and (c) which concurs with the previous observations [23], [29], [56]-[58]. The concentration effect was more pronounced at further spinning distances (Figure 4 (a)). This could be attributed to

A Novel Approach for Analysis of Processing Parameters…

167

the twofold effect of distance which will later be discussed in the paper. At low concentrations, there are more amounts of solvent in the solution and longer distance provides more time not only to stretch the jet in the electric field but also to evaporate the solvent, thereby encouraging thinner fiber formation. At higher concentrations, however, there are extensive polymer chain entanglements resulting in higher viscoelastic forces which tend to resist the electrostatic stretching force. On the other hand, increasing the spinning distance will reduce the electric field strength ( E =

V ) causing the electrostatic force to decrease. As a d

result, increasing MFD with concentration gains more momentum at longer spinning distances. Higher applied voltages also accelerate the concentration impact on MFD (Figure 4 (b)) which may be ascribed to the twofold effect of voltage explained later in the paper. At higher voltages, where the electric field is strong and dominant factor, increasing polymer concentration tend to encourage the effect of voltage on mass flow of polymer. Hence, more solution could be removed from the tip of the needle resulting in further increase in MFD. No combined effect between solution concentration and volume flow rate was observed as depicted in Figure 4 (c). Therefore, concentration had interactions with spinning distance and applied voltage which had been suggested by the existence of terms Cd and CV in the model of MFD. Recall that the term CQ was statistically insignificant and therefore had been removed from the model of MFD.

Spinning Distance The effect of spinning distance on electrospun fiber diameter is twofold. Varying the distance has a direct influence on the jet flight time as well as electric field strength. Longer spinning distance will provide more time for the jet to stretch in the electric field before it is deposited on the collector. Furthermore, solvents will have more time to evaporate. Hence, the fiber diameter will be prone to decrease. On the other hand, increasing the spinning distance, the electric field strength will decrease ( E =

V ) resulting in less acceleration hence d

stretching of the jet which leads to thicker fiber formation. The balance between these two effects will determine the final fiber diameter. Increase in fiber diameter [57], [60], [61] as well as decrease in fiber diameter [29] with increasing spinning distance was reported in the literature. There were also some cases in which spinning distance did not have a significant influence on fiber diameter [56 [62]-[64]. The impact of spinning distance on MFD is illustrated in Figure 4 (a), (d), and (e). As it is depicted in these figures, the effect of spinning distance is not always the same. As mentioned before, there will be more chain entanglements at higher concentrations resulting in an increase in viscoelastic force. Furthermore, the longer the distance, the lower is the electric field strength. Hence, the electrostatic stretching force, which has now become weaker, will be dominated easier by the viscoelastic force. As a result, the increasing effect of spinning distance on fiber diameter will be assisted, rendering higher MFD (Figure 4 (a)). The effect of spinning distance will alter at different applied voltages (Figure 4 (d)). At low voltages, longer spinning distance brought about thinner fiber formation, whereas at high voltages, the effect of spinning distance was totally reversed and fibers with thicker diameters were obtained at further distances. It is supposed that at low

168

M. Ziabari, V. Mottaghitalab and A. K. Haghi

voltages, where the electric field is weak, stretching time becomes the dominant factor. Hence, longer spinning distance, which gives more time to the jet to stretch and thin and to solvent to evaporate, will result in fibers with smaller diameters. At high voltages, however, the electric field strength is high and dominant. Therefore, increasing the distance, which reduces the electric field, causes an increase in fiber diameter. The function of spinning distance was observed to be independent from volume flow rate for MFD (Figure 4 (e)). The interaction of spinning distance with solution concentration and applied voltage demonstrated in Figure 4 (a) and (d), proved the existence of terms Cd and dV in the model of MFD.

Applied Voltage Applied voltage has two major different effects on fiber diameter. On one hand, increasing the applied voltage will increase the electric field strength and larger electrostatic stretching force causes the jet to accelerate more in the electric field, thereby favoring thinner fiber formation. On the other hand, since charge transport is only carried out by the flow of polymer in the electrospinning process [65] and increasing the voltage would introduce more surface charges on the jet, the mass flow rate from the needle tip to the collector will increase, say the solution will be drawn more quickly from the tip of the needle causing fiber diameter to increase. Combination of these two effects will determine the final fiber diameter. Hence, increasing applied voltage may decrease [66]-[68], increase [56], [57], [61] or may not change [23], [29], [62], [69] the fiber diameter. Figure 4 (b), (d), and (f) show the effect of applied voltage on MFD. Increasing the voltage, MFD underwent an increase followed by a decrease. According to the explanation given, at low voltages, where the electric field strength is low, the effect of mass of solution could be dominant. Therefore, fiber diameter increases when the applied voltage rises. However, as the voltage exceeds a limit, the electric field will be high enough to be a determining factor. Hence, fiber diameter decreases as the voltage increases. The effect of voltage on MFD was influenced by solution concentration to some extent (Figure 4 (b)). At high concentrations, the increase of fiber diameter with voltage was more pronounced. This could be attributed to the fact that the effect of mass of solution will be more important for the solutions of higher concentrations. Spinning distance dramatically influenced the way voltage affects fiber diameter (Figure 4 (d)). At a short distance, the electric field is high and dominant factor. Therefore, increasing applied voltage, which strengthens the electric field, resulted in a decrease in fiber diameter. Whereas, at long distances where the electric field wass low, the effect of mass of solution would be determining factor according to which fiber diameter increased with applied voltage. The effect of applied voltage on MFD was found to be independent from volume flow rate. Looking at the figures, it is apparent that there was a huge interaction between applied voltage and spinning distance, a slight interaction between applied voltage and solution concentration and no interaction between applied voltage and volume flow rate which is in agreement with the presence of CV and dV and absence of VQ in the model of MFD.

A Novel Approach for Analysis of Processing Parameters…

169

Volume Flow Rate It was suggested that a minimum value for solution flow rate is required to form the drop of polymer at the tip of the needle for the sake of maintaining a stable Taylor cone [70]. Hence, flow rate could affect the morphology of electrospun nanofibers such as fiber diameter. Increasing the flow rate, more amount of solution is delivered to the tip of the needle enabling the jet to carry the solution away faster. This could bring about an increase in the jet diameter favoring thicker fiber formation. In this study, the MFD slightly increased with volume flow rate (Figure 4 (c), (e), and (f)) which agrees the previous researches [29], [70]-[72]. Flow rate was also found to influence MFD independent from solution concentration, applied voltage, and spinning distance as suggested earlier by the absence of CQ, dQ, and VQ in the model of MFD.

(a)

(b)

(c) Figure 4. (Continued)

170

M. Ziabari, V. Mottaghitalab and A. K. Haghi

(d)

(e)

(f) Figure 4. Response surfaces for mean fiber diameter in terms of: (a) solution concentration and spinning distance, (b) solution concentration and applied voltage, (c) solution concentration and flow rate, (d) spinning distance and applied voltage, (e) spinning distance and flow rate, (f) applied voltage and flow rate.

A Novel Approach for Analysis of Processing Parameters…

171

RESPONSE SURFACES FOR STANDARD DEVIATION OF FIBER DIAMETER Solution Concentration As depicted in Figure 5 (a), (b), and (c), StdFD increased with concentration which concurs with the previous observations [23], [31], [56], [59], [29], [61], [73], [74]. In electrospinning, the elongation flow of the jet causes the coiled macromolecules in the solution to transform into oriented entangled network. Increasing the polymer concentration, the macromolecular chain entanglements increase, prompting a greater difficulty for the jet to stretch and split. This could result in less uniform fibers (higher StdFD). Concentration affected StdFD regardless of spinning distance (Figure 5 (a)), suggesting that there was no interaction between these two parameters (absence of Cd in the model of StdFd). At low applied voltages, the formation of more uniform fibers with decreasing the concentration was facilitated. In agreement with existence of the term CV in the model of StdFd, solution concentration was found to have a slight interaction with applied voltage (Figure 5 (b)). The curvature of the surface in Figure 5 (c) suggested that there was a noticeable interaction between concentration and flow rate and this agrees the presence of the term CQ in the model of StdFD.

Spinning Distance More uniform fibers (lower StdFD) were obtained with increasing the spinning distance as shown in Figure 5 (a), (d), and (e). At longer spinning distance, more time is given to the jet to fly from the tip of the needle to the collector and to solvent to evaporate. Therefore, the processes of stretching the jet and evaporating the solvent will be carried out more gently resulting in more uniform fibers. Our finding is consistent with the trend observed by Zhao et al. [74]. Spinning distance influenced StdFD regardless of solution concentration and applied voltage (Figure 5 (a) and (d)) meaning that no interaction exists between these variables as could be inferred from the model of StdFD. However, the interaction of spinning distance with volume flow rate was noticeable (Figure 5 (e)). The presence of dQ in the model of StdFD proves this observation. The effect of spinning distance was more pronounced at higher flow rates. This could be attributed to the fact that more amount of solution is deliverd to the tip of the needle at higher flow rates; therefore the threads will require more time to dry. If the distance is high enough to provide the sufficient time, uniform fibers will be formed. Decreasing the distance, there will be less time for solvent to evaporate favoring the production of non-uniform fibers (high StdFD).

Applied Voltage StdFD was found to increase with applied voltage (Figure 5 (b), (d), and (f)) as observed in other works [56], [57], [61], [74]. Increasing the applied voltage causes the effect of the

172

M. Ziabari, V. Mottaghitalab and A. K. Haghi

(a)

(b)

(c) Figure 5. (Continued)

A Novel Approach for Analysis of Processing Parameters…

173

(d)

(e)

(f) Figure 5: Response surfaces for standard deviation of fiber diameter in terms of: (a) solution concentration and spinning distance, (b) solution concentration and applied voltage, (c) solution concentration and flow rate, (d) spinning distance and applied voltage, (e) spinning distance and flow rate, (f) applied voltage and flow rate.

174

M. Ziabari, V. Mottaghitalab and A. K. Haghi

electric field on the charged jet to increase. Hence, the flight speed of the jet increases, shortening the time that the jet travels towards the collector. As a result, less time is given to the jet to stretch and thin and to solvent to evaporate. This may result in formation of less uniform fibers (higher StdFD). The effect of applied voltage on StdFD was influenced by solution concentration as depicted in Figure 5 (b), implying the interaction of voltage with concentration which was earlier addressed in the paper by the presence of the corresponding term in the model of StdFD. At low concentrations, the formation of uniform fibers with decreasing the applied voltage was facilitated. No interaction was observed between applied voltage and spinning distance (Figure 5 (d)) as suggested by the absence of the term dV in the model of StdFD. Figure 5 (f) shows a slight interaction of voltage with flow rate, which concurs with the existence of VQ in the model of StdFD.

Volume Flow Rate As demonstrated in Figure 5 (c), (e), and (f), increasing the flow rate, the uniformity of fibers increased (StdFD decreased), reached an optimum value and then decreased (StdFD increased). When the flow rate is low, the amount of solution fed to the tip of the needle is not sufficient, whereas an excess amount of solution is delivered to the tip of the needle at high flow rates. Therefore, unstable jets are formed in the two extremes resulting in the production of non-uniform fibers. Solution concentration, applied voltage, and spinning distance were found to influence the impact of flow rate on StdFD indicating the interaction of flow rate with the other variables as demonstrated by the terms CQ, dQ, and VQ in the model of StdFD. Increasing the solution concentration favored the formation of non-uniform fibers at high flow rates (Figure 5 (c)) which may be the result of the greater difficulty of solution removal due to the increased polymeric material content in the solution. The effect of flow rate on StdFD was more pronounced as the spinning distance decreased (Figure 5 (e)). The shorter the distance, the fewer the time provided to the jet to thin and dry. Therefore, at high flow rates at which more amount of solution is delivered to the tip of the needle, the given time may not suffice, resulting in formation of less uniform fibers. High applied voltage encouraged the increase in StdFD at fast flow rates as depicted in Figure 5 (f).

CONCLUSION For the first time in this paper, the simultaneous effects of four processing variables – solution concentration, applied voltage, spinning distance, and volume flow rate – on MFD and StdFD were investigated quantitatively as well as qualitatively. The appropriate range of parameters where dry, bead-free, and continuous fibers without breaking up to droplets are formed, were selected by referring to the literature along with conducting a series of preliminary experiments. A full factorial experimental design at three levels of each factor (34 design) was carried out besides which 15 treatments inside the design space were selected as test set for evaluating the prediction ability of the models. Nanofibers PVA were then prepared for experimental and test sets through the electrospinning method. After that, MFD and StdFD were determined from SEM micrograph of each sample. RSM was used to

A Novel Approach for Analysis of Processing Parameters…

175

establish quadratic models for MFD and StdFD. The test for significance of the coefficients demonstrated that the terms Q 2, CQ, dQ, and VQ in the model of MFD and d 2, Cd, and dV in the model of StdFD were not of important value in representing the responses. Eliminating these terms, simpler yet more efficient models were obtained which not only explained the experimental data in a better manner, but also had more prediction ability. Afterwards, in order to show the generalization ability of the models for predicting new conditions, the test set was used. Low RMSE of test set for MFD and StdFD were obtained indicating the good prediction ability of the models. Finally, in order to qualitatively study the effects of variables on MFD and StdFD, response surface plots were generated using the relationships at hand. For MFD: 1) Increasing solution concentration, MFD increased rigorously. The effect of concentration was more pronounced at longer spinning distance and also at higher applied voltage. 2) The effect of spinning distance on MFD changed depending on solution concentration and applied voltage. At low applied voltages, MFD decreased as the spinning distance became longer, whereas higher MFD resulted with lengthening the spinning distance when the applied voltage was high. Increasing the solution concentration tended to assist the formation of thicker fibers at longer spinning distance. 3) Rising the applied voltage, MFD was observed to first increase and then decrease. High solution concentrations partly and long spinning distances largely favored the increase of MFD with applied voltage. 4) MFD slightly increased with flow rate. The impact of flow rate on MFD was unrelated to the other variables. For StdFD: [1]

[2] [3] [4]

The higher the solution concentration, the less uniform fibers (higher StdFD) was formed. Low applied voltages facilitated the formation of more uniform fibers (lower StdFD) with decreasing the concentration. The increase of StdFD with concentration gained momentum at high flow rates. Longer spinning distance resulted in more uniform fibers (lower StdFD). The effect of spinning distance was more pronounced at higher flow rates. Rising the applied voltage increased StdFD. Low concentrations facilitated the formation of uniform fibers (high StdFD) with decreasing the applied voltage. Flow rate was found to have a significant impact on uniformity of fibers (StdFD). As flow rate increased, StdFD decreased and then increased. Higher solution concentration, higher applied voltage, and shorter spinning distance encouraged the formation of nonuniform fibers (high StdFD) at fast flow rates.

176

M. Ziabari, V. Mottaghitalab and A. K. Haghi

APPENDIX Table 6. Natural and coded variables for experimental and test data along with corresponding responses Natural Variables No. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39

Coded Variables

Responses

C (%)

d (cm)

V (kV)

Q (ml/h)

x1

x2

x3

x4

MFD (nm)

8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 10 10 10 10 10 10 10 10 10 10 10 10

10 10 10 10 10 10 10 10 10 15 15 15 15 15 15 15 15 15 20 20 20 20 20 20 20 20 20 10 10 10 10 10 10 10 10 10 15 15 15

15 15 15 20 20 20 25 25 25 15 15 15 20 20 20 25 25 25 15 15 15 20 20 20 25 25 25 15 15 15 20 20 20 25 25 25 15 15 15

0.2 0.3 0.4 0.2 0.3 0.4 0.2 0.3 0.4 0.2 0.3 0.4 0.2 0.3 0.4 0.2 0.3 0.4 0.2 0.3 0.4 0.2 0.3 0.4 0.2 0.3 0.4 0.2 0.3 0.4 0.2 0.3 0.4 0.2 0.3 0.4 0.2 0.3 0.4

-1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 0 0 0 0 0 0 0 0 0 0 0 0

-1 -1 -1 -1 -1 -1 -1 -1 -1 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 -1 -1 -1 -1 -1 -1 -1 -1 -1 0 0 0

-1 -1 -1 0 0 0 1 1 1 -1 -1 -1 0 0 0 1 1 1 -1 -1 -1 0 0 0 1 1 1 -1 -1 -1 0 0 0 1 1 1 -1 -1 -1

-1 0 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1

232.62 235.50 252.02 236.84 232.08 249.21 196.05 201.38 215.00 221.10 238.63 242.32 219.76 228.56 242.01 202.62 208.21 213.66 196.63 197.73 206.28 206.69 224.38 242.06 205.25 215.70 231.34 269.91 270.05 291.99 256.11 264.86 278.34 228.21 239.28 238.74 263.67 269.29 277.71

StdFD (nm) 26.60 24.52 25.89 37.30 30.22 34.49 34.76 35.15 39.00 28.88 20.17 21.99 36.19 28.29 28.30 33.22 37.14 34.84 30.69 24.55 22.11 31.56 27.41 26.51 40.32 30.54 32.40 30.35 28.88 33.98 38.54 35.70 49.13 42.33 40.30 46.57 34.16 31.54 29.40

A Novel Approach for Analysis of Processing Parameters… Natural Variables No. 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83

Coded Variables

177

Responses

C (%)

d (cm)

V (kV)

Q (ml/h)

x1

x2

x3

x4

MFD (nm)

10 10 10 10 10 10 10 10 10 10 10 10 10 10 10 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 9 10

15 15 15 15 15 15 20 20 20 20 20 20 20 20 20 10 10 10 10 10 10 10 10 10 15 15 15 15 15 15 15 15 15 20 20 20 20 20 20 20 20 20 20 12.5

20 20 20 25 25 25 15 15 15 20 20 20 25 25 25 15 15 15 20 20 20 25 25 25 15 15 15 20 20 20 25 25 25 15 15 15 20 20 20 25 25 25 15 15

0.2 0.3 0.4 0.2 0.3 0.4 0.2 0.3 0.4 0.2 0.3 0.4 0.2 0.3 0.4 0.2 0.3 0.4 0.2 0.3 0.4 0.2 0.3 0.4 0.2 0.3 0.4 0.2 0.3 0.4 0.2 0.3 0.4 0.2 0.3 0.4 0.2 0.3 0.4 0.2 0.3 0.4 0.3 0.3

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 -0.5 0

0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 -1 -1 -1 -1 -1 -1 -1 -1 -1 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 -0.5

0 0 0 1 1 1 -1 -1 -1 0 0 0 1 1 1 -1 -1 -1 0 0 0 1 1 1 -1 -1 -1 0 0 0 1 1 1 -1 -1 -1 0 0 0 1 1 1 -1 -1

-1 0 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 -1 0 1 0 0

284.20 281.82 282.39 249.42 278.22 286.96 239.45 244.04 251.58 285.67 273.05 280.62 278.10 280.95 306.28 286.23 295.60 293.41 271.20 291.89 295.93 234.13 247.65 247.13 271.93 297.65 296.79 297.94 310.06 312.15 272.24 282.04 288.00 259.63 278.40 279.25 307.42 327.77 337.88 321.78 334.54 342.45 216.53 259.61

StdFD (nm) 38.18 36.27 42.07 40.79 46.15 51.16 27.98 27.43 27.26 35.62 30.74 34.66 40.79 44.58 44.04 27.12 32.91 40.48 34.86 42.78 49.43 39.31 48.60 59.02 33.05 26.75 39.84 38.82 36.84 41.69 39.55 42.35 51.72 34.63 25.35 27.25 42.25 35.71 45.16 46.21 40.68 47.94 24.25 25.67

178

M. Ziabari, V. Mottaghitalab and A. K. Haghi Table 7. (Continued) Natural Variables

No. 84 85 86 87 88 89 90 91 92 93 94 95 96

Coded Variables

Responses

C (%)

d (cm)

V (kV)

Q (ml/h)

x1

x2

x3

x4

MFD (nm)

10 10 9 9 9 10 10 10 9 9 9 10 9

20 20 12.5 20 20 12.5 12.5 20 12.5 12.5 20 12.5 12.5

22.5 15 15 22.5 15 22.5 15 22.5 22.5 15 22.5 22.5 22.5

0.3 0.25 0.3 0.3 0.25 0.3 0.25 0.25 0.3 0.25 0.25 0.25 0.25

0 0 -0.5 -0.5 -0.5 0 0 0 -0.5 -0.5 -0.5 0 -0.5

1 1 -0.5 1 1 -0.5 -0.5 1 -0.5 -0.5 1 -0.5 -0.5

0.5 -1 -1 0.5 -1 0.5 -1 0.5 0.5 -1 0.5 0.5 0.5

0 -0.5 0 0 -0.5 0 -0.5 -0.5 0 -0.5 -0.5 -0.5 -0.5

300.27 235.04 247.57 247.16 212.82 263.70 258.26 272.03 235.75 244.43 252.50 260.71 231.97

StdFD (nm) 35.71 29.64 26.65 31.12 30.26 45.06 26.16 36.28 33.16 24.87 36.01 42.25 32.86

REFERENCES [5] [6] [7]

[8] [9]

[10] [11] [12]

[13]

G.I. Taylor, Electrically Driven Jets, Proceeding of the Royal Society of London, vol. 313, no. 1515, pp. 453-475, 1969. J. Doshi and D.H. Reneker, Electrospinning Process and Applications of Electrospun Fibers, Journal of Electrostatics, vol. 35, 151-160, 1995. H. Fong and D.H. Reneker, Electrospinning and the Formation of Nanofibers, Chapter 6, pp. 225-246, In: D. R. Salem, Structure Formation in Polymeric Fibers, Hanser, Cincinnati, 2001. D. Li and Y. Xia, Electrospinning of Nanofibers: Reinventing the Wheel?, Advanced Materials, vol. 16, no. 14, pp. 1151-1170, 2004. R. Derch, A. Greiner and J.H. Wendorff, Polymer Nanofibers Prepared by Electrospinning, In: J. A. Schwarz, C. I. Contescu and K. Putyera, Dekker Encyclopedia of Nanoscience and Nanotechnology, CRC, New York, 2004. A.K. Haghi and M. Akbari, Trends in Electrospinning of Natural Nanofibers. Physica Status Solidi (a), vol. 204, pp. 1830-1834, 2007. P.W. Gibson, H.L. Schreuder-Gibson and D. Rivin, Electrospun Fiber Mats: Transport Properties, AIChE Journal, vol. 45, no. 1, pp. 190-195, 1999. Z.M. Huang, Y.Z. Zhang, M. Kotaki and S. Ramakrishna, A Review on Polymer Nanofibers by Electrospinning and Their Applications in Nanocomposites, Composites Science and Technology, vol. 63, pp. 2223-2253, 2003. M. Li, M.J. Mondrinos, M.R. Gandhi, F.K. Ko, A.S. Weiss and P.I. Lelkes, Electrospun Protein Fibers as Matrices for Tissue Engineering, Biomaterials, vol. 26, pp. 59996008, 2005.

A Novel Approach for Analysis of Processing Parameters…

179

[14] E.D. Boland, B.D. Coleman, C.P. Barnes, D.G. Simpson, G.E. Wnek and G.L. Bowlin, Electrospinning Polydioxanone for Biomedical Applications, Acta Biomaterialia, vol. 1, pp. 115-123, 2005. [15] J. Lannutti, D. Reneker, T. Ma, D. Tomasko and D. Farson, Electrospinning for Tissue Engineering Scaffolds, Materials Science and Engineering C, vol. 27, pp. 504-509, 2007. [16] J. Zeng, L. Yang, Q. Liang, X. Zhang, H. Guan, C. Xu, X. Chen and X. Jing, Influence of the Drug Compatibility with Polymer Solution on the Release Kinetics of Electrospun Fiber Formulation, Journal of Controlled Release, vol. 105, pp. 43-51, 2005. [17] E.R. Kenawy, G.L. Bowlin, K. Mansfield, J. Layman, D.G. Simpson, E.H. Sanders and G.E. Wnek, Release of Tetracycline Hydrochloride from Electrospun Poly(EthyleneCo-Vinylacetate), Poly(Lactic Acid), and a Blend, Journal of Controlled Release, vol. 81, pp. 57-64, 2002. [18] M.-S. Khil, D.-I. Cha, H.-Y. Kim, I.-S. Kim and N. Bhattarai, Electrospun Nanofibrous Polyurethane Membrane as Wound Dressing, Journal of Biomedical Materials Research - Part B: Applied Biomaterials, vol. 67, pp. 675-679, 2003. [19] B.M. Min, G. Lee, S.H. Kim, Y.S. Nam, T.S. Lee and W.H. Park, Electrospinning of Silk Fibroin Nanofibers and Its Effect on the Adhesion and Spreading of Normal Human Keratinocytes and Fibroblasts in Vitro, Biomaterials, vol. 25, pp. 1289-1297, 2004. [20] X.H. Qin and S.Y. Wang, Filtration Properties of Electrospinning Nanofibers, Journal of Applied Polymer Science, vol. 102, pp. 1285-1290, 2006. [21] J.S. Kim and D.H. Reneker, Polybenzimidazole Nanofiber Produced by Electrospinning, Polymer Engineering and Science, vol. 39, no. 5, pp. 849-854, 1999. [22] S.W. Lee, S.W. Choi, S.M. Jo, B.D. Chin, D.Y. Kim and K.Y. Lee, Electrochemical Properties and Cycle Performance of Electrospun Poly(Vinylidene Fluoride)-Based Fibrous Membrane Electrolytes for Li-Ion Polymer Battery, Journal of Power Sources, vol. 163, pp. 41-46, 2006. [23] C. Kim, Electrochemical Characterization of Electrospun Activated Carbon Nanofibres as an Electrode in Supercapacitors, Journal of Power Sources, vol. 142, pp. 382-388, 2005. [24] N.J. Pinto, A.T. Johnson, A.G. MacDiarmid, C.H. Mueller, N. Theofylaktos, D.C. Robinson and F.A. Miranda, Electrospun Polyaniline/Polyethylene Oxide Nanofiber Field-Effect Transistor, Applied Physics Letters, vol. 83, no. 20, pp. 4244-4246, 2003. [25] D. Aussawasathien, J.-H. Dong and L. Dai, Electrospun Polymer Nanofiber Sensors, Synthetic Metals, vol. 54, pp. 37-40, 2005. [26] S.-Y. Jang, V. Seshadri, M.-S. Khil, A. Kumar, M. Marquez, P.T. Mather and G.A. Sotzing, Welded Electrochromic Conductive Polymer Nanofibers by Electrostatic Spinning, Advanced Materials, vol. 17, pp. 2177-2180, 2005. [27] S.- H. Tan, R. Inai, M. Kotaki and R. Ramakrishna, Systematic Parameter Study for Ultra-Fine Fiber Fabrication via Electrospinning Process, Polymer, vol. 46, pp. 61286134, 2005. [28] A. Ziabicki, Fundamentals of Fiber Formation: The Science of Fiber Spinning and Drawing, Wiley, New York, 1976.

180

M. Ziabari, V. Mottaghitalab and A. K. Haghi

[29] A. Podgóski, A. Bałazy and L. Gradoń, Application of Nanofibers to Improve the Filtration Efficiency of the Most Penetrating Aerosol Particles in Fibrous Filters, Chemical Engineering Science, vol. 61, pp. 6804-6815, 2006. [30] B. Ding, M. Yamazaki and S. Shiratori, Electrospun Fibrous Polyacrylic Acid Membrane-Based Gas Sensors, Sensors and Actuators B, vol. 106, pp. 477-483, 2005. [31] J.R. Kim, S.W. Choi, S.M. Jo, W.S. Lee and B.C. Kim, Electrospun PVdF-Based Fibrous Polymer Electrolytes for Lithium Ion Polymer Batteries, Electrochimica Acta, vol. 50, pp. 69-75, 2004. [32] L. Moroni, R. Licht, J. de Boer, J.R. de Wijn and C.A. van Blitterswijk, Fiber diameter and texture of electrospun PEOT/PBT scaffolds influence human mesenchymal stem cell proliferation and morphologyand the release of incorporated compounds, Biomaterials, vol. 27, pp. 4911-4922, 2006. [33] T. Wang and S. Kumar, Electrospinning of Polyacrylonitrile Nanofibers, Journal of Applied Polymer Science, vol. 102, pp. 1023-1029, 2006. [34] S. Sukigara, M. Gandhi, J. Ayutsede, M. Micklus and F. Ko, Regeneration of Bombyx Mori Silk by Electrospinning – Part 2: Process Optimization and Empirical Modeling using Response Surface Methodology, Polymer, vol. 45, pp. 3701-3708, 2004. [35] S.Y. Gu, J. Ren and G.J. Vancso, Process Optimization and Empirical Modeling for Electrospun Polyacrylonitrile (PAN) Nanofiber Precursor of Carbon Nanofibers, European Polymer Journal, vol. 41, pp. 2559-2568, 2005. [36] S.Y. Gu and J. Ren, Process Optimization and Empirical Modeling for Electrospun Poly(D,L-Lactide) Fibers using Response Surface Methodology, Macromolecular Materials and Engineering, vol. 290, pp. 1097-1105, 2005. [37] O.S. Yördem, M. Papila and Y.Z. Menceloğlu, Effects of electrospinning parameters on polyacrylonitrile nanofiber diameter: An investigation by response surface methodology, Materials and Design, vol. 29, pp. 34-44, 2008. [38] I. Sakurada, Polyvinyl Alcohol Fibers, CRC, New York, 1985. [39] F.L. Marten, Vinyl Alcohol Polymers, In: H. F. Mark, Encyclopedia of Polymer Science and Technology, 3rd Eddition, vol. 8, Wiley, 2004. [40] Y.D. Kwon, S. Kavesh and D.C. Prevorsek (Allied Corporation), Method of Preparing High Strength and Modulus Polyvinyl Alcohol Fibers, U. S. Patent, no. 4,440,711, April 3, 1984. [41] S. Kavesh and D.C. Prevorsek (Allied Corporation), Producing High Tenacity, High Modulus Crystalline Article Such as Fiber or Film, U. S. Patent, no. 4,551,296, November 5, 1985. [42] H. Tanaka, M. Suzuki and F. Uedo (Toray Industries, Inc.), Ultra-High-Tenacity Polyvinyl Alcohol Fibers and Process for Producing Same, U. S. Patent, no. 4,603,083, July 29, 1986. [43] G. Paradossi, F. Cavalieri, E. Chiessim, C. Spagnoli and M.K. Cowman, Poly(Vinyl Alcohol) as Versatile Biomaterial for Potential Biomedical Applications, Journal of Materials Science: Materials in Medicine, vol. 14, pp. 687-691, 2003. [44] G. Zheng-Qiu, X. Jiu-Mei and Z. Xiang-Hong, The Development of Artificial Articular Cartilage – PVA Hydrogel, Bio-Medical Materials and Engineering, vol. 8, pp. 75-81, 1998. [45] M. Oka, K. Ushio, P. Kumar, K. Ikeuchi, S.H. Hyon, T. Nakamura and H. Fujita, Development of Artificial Articular Cartilage, Proceedings of the Institution of

A Novel Approach for Analysis of Processing Parameters…

[46]

[47]

[48] [49]

[50] [51] [52]

[53] [54] [55] [56] [57] [58] [59] [60] [61]

[62]

[63]

[64]

181

Mechanical Engineers – Part H – Journal of Engineering in Medicine, vol. 214, pp. 5968, 2000. K. Burczak, E. Gamian and A. Kochman, Long-Term in Vivo Performance and Biocompatibility of Poly(Vinyl Alcohol) Hydrogel Macrocapsules for Hybrid-type Artificial Pancreas, Biomaterials, vol. 17, pp. 2351, 1996. J.K. Li, N. Wang and X.S. Wu, Poly(Vinyl Alcohol) Nanoparticles Prepared by Freezing-Thawing Process for Protein/Peptide Drug Delivery, Journal of Controlled Release, vol. 56, pp. 117-126, 1998. A.S. Hoffman, Hydrogels for Biomedical Applications, Advanced Drug Delivery Reviews, vol. 43, pp. 3-12, 2002. J. Zeng, A. Aigner, F. Czubayko, T. Kissel, J.H. Wendorff and A. Greiner, Poly(vinyl alcohol) Nanofibers by Electrospinning as a Protein Delivery System and the Retardation of Enzyme Release by Additional Polymer Coatings, Biomacromolecules, vol. 6, pp. 1484-1488, 2005. K.H. Hong, Preparation and Properties of Electrospun Poly(Vinyl Alcohol)/Silver Fiber Web as Wound Dressings, Polymer Engineering and Science, vol. 47, pp. 43-49, 2007. L.H. Sperling, Introduction to Physical Polymer Science, 4th Edition, Wiley, New Jersey, 2006. J.C.J.F. Tacx, H.M. Schoffeleers, A.G.M. Brands and L. Teuwen, Dissolution Behavior and Solution Properties of Polyvinylalcohol as Determined by Viscometry and Light Scattering in DMSO, Ethyleneglycol and Water, Polymer, vol. 41, pp. 947-957, 2000. F.K. Ko, Nanofiber Technology, Chapter 19, In: Y. Gogotsi, Nanomaterials Handbook, CRC Press, Boca Raton, 2006. A. Koski, K. Yim and S. Shivkumar, Effect of Molecular Weight on Fibrous PVA Produced by Electrospinning, Materials Letters, vol. 58, pp. 493-497, 2004. D.C. Montgomery, Design and Analysis of Experiments, 5th Edition, Wiley, New York, 1997. A. Dean and D. Voss, Design and Analysis of Experiments, Springer, New York, 1999. G.E.P. Box and N.R. Draper, Response Surfaces, Mixtures, and Ridge Analyses, Wiley, New Jersey, 2007. K.M. Carley, N.Y. Kamneva and J. Reminga, Response Surface Methodology, CASOS Technical Report, CMU-ISRI-04-136, 2004. S. Weisberg, Applied Linear Regression, 3rd Edition, Wiley, New Jersey, 2005. C. Zhang, X. Yuan, L. Wu, Y. Han and J. Sheng, Study on Morphology of Electrospun Poly(Vinyl Alcohol) Mats, European Polymer Journal, vol. 41, pp. 423-432, 2005. Q. Li, Z. Jia, Y. Yang, L. Wang and Z. Guan, Preparation and Properties of Poly(Vinyl Alcohol) Nanofibers by Electrospinning, Proceedings of IEEE International Conference on Solid Dielectrics, Winchester, U. K., July 8-13, 2007. C. Mit-uppatham, M. Nithitanakul and P. Supaphol, Ultrafine Electrospun Polyamide-6 Fibers: Effect of Solution Conditions on Morphology and Average Fiber Diameter, Macromolecular Chemistry and Physics, vol. 205, pp. 2327-2338, 2004. Y.J. Ryu, H.Y. Kim, K.H. Lee, H.C. Park and D.R. Lee, Transport Properties of Electrospun Nylon 6 Nonwoven Mats, European Polymer Journal, vol. 39, pp. 18831889, 2003. T. Jarusuwannapoom, W. Hongrojjanawiwat, S. Jitjaicham, L. Wannatong, M. Nithitanakul, C. Pattamaprom, P. Koombhongse, R. Rangkupan and P. Supaphol,

182

[65]

[66]

[67] [68]

[69]

[70]

[71]

[72]

[73]

[74]

[75] [76]

[77]

[78]

M. Ziabari, V. Mottaghitalab and A. K. Haghi Effect of Solvents on Electro-Spinnability of Polystyrene Solutions and Morphological Appearance of Resulting Electrospun Polystyrene Fibers, European Polymer Journal, vol. 41, pp. 409-421, 2005. S.C. Baker, N. Atkin, P.A. Gunning, N. Granville, K. Wilson, D. Wilson and J. Southgate, Characterisation of Electrospun Polystyrene Scaffolds for ThreeDimensional in Vitro Biological Studies, Biomaterials, vol. 27, pp. 3136–3146, 2006. S. Sukigara, M. Gandhi, J. Ayutsede, M. Micklus and F. Ko, Regeneration of Bombyx Mori Silk by Electrospinning – Part 1: Processing Parameters and Geometric Properties, Polymer, vol. 44, pp. 5721-5727, 2003. X. Yuan, Y. Zhang, C. Dong and J. Sheng, Morphology of Ultrafine Polysulfone Fibers Prepared by Electrospinning, Polymer International, vol. 53, pp. 1704-1710, 2004. C.S. Ki, D.H. Baek, K.D. Gang, K.H. Lee, I.C. Um and Y.H. Park, Characterization of Gelatin Nanofiber Prepared from Gelatin–Formic Acid Solution, Polymer, vol. 46, pp. 5094-5102, 2005. J.M. Deitzel, J. Kleinmeyer, D. Harris and N.C. Beck Tan, The Effect of Processing Variables on the Morphology of Electrospun Nanofibers and Textiles, Polymer, vol. 42, pp. 261-272, 2001. C.J. Buchko, L.C. Chen, Y. Shen and D.C. Martin, Processing and Microstructural Characterization of Porous Biocompatible Protein Polymer Thin Films, Polymer, vol. 40, pp. 7397-7407, 1999. J.S. Lee, K.H. Choi, H.D. Ghim, S.S. Kim, D.H. Chun, H.Y. Kim and W.S. Lyoo1, Role of Molecular Weight of Atactic Poly(vinyl alcohol) (PVA) in the Structure and Properties of PVA Nanofabric Prepared by Electrospinning, Journal of Applied Polymer Science, vol. 93, pp. 1638-1646, 2004. S.F. Fennessey and R.J. Farris, Fabrication of Aligned and Molecularly Oriented Electrospun Polyacrylonitrile Nanofibers and the Mechanical Behavior of Their Twisted Yarns, Polymer, vol. 45, pp. 4217–4225, 2004. S. Kidoaki, I. K. Kwon and T. Matsuda, Mesoscopic Spatial Designs of Nano- and Microfiber Meshes for Tissue-Engineering Matrix and Scaffold Based on Newly Devised Multilayering and Mixing Electrospinning Techniques, Biomaterials, vol. 26, pp. 37-46, 2005. X. Zong, K. Kim, D. Fang, S. Ran, B.S. Hsiao and B. Chu, Structure and Process Relationship of Electrospun Bioabsorbable Nanofiber Membranes, Polymer, vol. 43, pp. 4403-4412, 2002. D. Li and Y. Xia, Fabrication of Titania Nanofibers by Electrospinning, Nano Letters, vol. 3, no. 4, pp. 555-560, 2003. W.-Z. Jin, H.-W. Duan, Y.-J. Zhang and F.-F. Li, Nonafiber Membrane of EVOHBased Ionomer by Electrospinning, Proceedings of the 1st IEEE International Conference on Nano/Micro Engineered and Molecular Systems, pp. 42-46, Zhuhai, China, January 18-21, 2006. X.M. Mo, C.Y. Xu, M. Kotaki and S. Ramakrishna, Electrospun P(LLA-CL) Nanofiber: A Biomimetic Extracellular Matrix for Smooth Muscle Cell and Endothelial Cell Proliferation, Biomaterials, vol. 25, pp. 1883-1890, 2004. S. Zhao, X. Wu, L. Wang and Y. Huang, Electrospinning of Ethyl–Cyanoethyl Cellulose/Tetrahydrofuran Solutions, Journal of Applied Polymer Science, vol. 91, pp. 242-246, 2004.

In: Nanofibers: Fabrication, Performance, and Applications ISBN 978-1-60741-947-1 Editors: W. N. Chang © 2009 Nova Science Publishers, Inc.

Chapter 5

CARBON NANO-FIBERS AND THEIR APPLICATIONS: DERIVED FROM ELECTROSPINNING AND VAPOR GROWN PROCESSES S. K. Nataraj, B. H. Kim and K. S. Yang1 Carbon Materials Lab, Alan G. MacDiarmid Energy Research Institute (AMERI)/Faculty of Applied Chemical Engineering, Chonnam National University, Gwangju, Korea

ABBREVIATIONS ES = Electrospun VGCFs = Vapor grown carbon nanofibers CNFs = Carbon nanofibers CDI = Capacitive deionization VG = Vapor grown method PAN = Polyacrylonitrile DMF = N,N-Dimethylformamide PS = Polysulfone CVD = Chemical vapor deposition e-spun = Electrospun EDLC = Electrochemical double-layer capacitors ACNFs = Activated carbon nanofibers MWCNT = Multiwalled carbon nanotube LIBs = Lithium-ion batteries HEVs = Hybrid electric vehicles MSB = Magnetic suspension balance IAA = Iron acetylacetonate PI = Polyimide

1 [email protected];[email protected]

184

S. K. Nataraj, B. H. Kim and K. S. Yang PBI = Polybenzimidazoles PVdF = Polyvinylidenefluoride GNFs = Graphitized nanofibers MB = Methylene blue CR = Congo red PEMFCs = Proton exchange membrane fuel cells EDX = Energy dispersive x-ray spectrometry NFM = Nanofiber membrane VOCs = Volatile organic compounds

1. INTRODUCTION Electrospun (ES) and vapor grown carbon nanofibers (VGCFs) are attractive building blocks for functional nanoscale devices. They are promising candidates for various applications, including filtration, protective clothing, polymer batteries and sensors. The continuous progress of nanotechnology in material science has led to the development of nanostructure materials with unique chemical, physical, and thermal properties. Nanofibers possesses significant characteristic that exhibit enormous availability of surface area per unit mass (See Figure 1). Furthermore, their high surface-to-volume ratio renders them attractive as catalyst supports, energy storage devices, as well as in drug delivery and tissue engineering. Since the discovery of carbon nanotubes in 1991 [1] and based on the results obtained from the characterization of these nanostructures [2-4], many other carbon-based nanomaterials have been developed. Of these, carbon nanofibers (CNFs) [5-9], fullerenes [1013], carbon nanohorns [14,15], and nanoporous structures are the subject of extensive experimental and theoretical studies for specific applications [16]. Carbon is a truly remarkable element existing as four allotropes, viz. diamond, graphite, carbynes and fullerenes, each having significant scientific and technological importance. Its most abundant allotrope, graphite, can take many forms with respect to microstructure, amorphous to highly crystalline structure, highly dense with density 2.2 g/cm3 to highly porous with density 0.5 g/cm3 and different shapes. These types of graphites are called synthetic carbons and in technical terms, engineered carbons. Carbon nanofibers are unique in the fact that their whole surface area can be activated. Since carbon nanofibers have a much larger functionalized surface area compared to that of nanotubes, the surface-active groups-to-volume ratio of these materials is much larger than that of the glassy-like surface of the carbon nanotubes. This characteristic, combined with the fact that the number and type of functional groups on the outer surface of the carbon fibers can be well controlled, is expected to allow for the selective immobilization and stabilization of functional biomolecules such as proteins, enzymes, and DNA. Also, the high conductivity of carbon nanofibers seems to be ideal for the electrochemical signal transduction. The oxygen-containing activated sites are ideal for the immobilization and stabilization of biomaterials is an important feature. Industrial applications of these new materials include: polymer and elastomer fillers, commercial hydrogen storage systems, radiowave-absorbing composites, lithium battery electrodes, construction composites, oil additives, gas-distribution layers for fuel cells,

Carbon Nano-Fibers and their Applications

185

absorbents and filters as well as in capacitive deionization (CDI) processes for water treatment. Figure 2, gives the spectrum broad range of application areas of nanofibers [17]. Many applications are being developed for field emission display, electrodes of secondary battery and reinforcement of materials. Among the many future possibilities includes; soft protective vests stronger than Kevlar, bandages that can contract to put pressure on, artificial muscles powered by electricity those expected much lighter than current hydraulics, would make it easier to incorporate electronic sensors and actuators into clothing. All of these possible applications derive from the remarkable properties of carbon nanofiber, the ability to conduct both heat and electricity along with the extreme toughness of the fiber. This chapter presents the various features of carbon nanofibers with elaborated properties description in connection with their different application, produced from electrospinning and vapor grown techniques. Carbon fibers are fibrous carbon materials with carbon content more than 90%. They are transformed from organic matter by 1000-1500oC heat treatment, which are substance with imperfect graphite crystalline structure arranged along the fiber axis. There are two ways to produce carbon fibers: one is from organic precursor fibers and the second once is from gas grown. Among several methods to produce CNFs electrospinning and vapor grown method (VG) are widely used techniques. Carbonaceous materials are found in a various forms such as graphite, diamond, carbon fibers, etc. Especially carbon fibers represent an important class of graphite-related materials from both a scientific and commercial views. CNFs are new class of advanced materials having unusual properties, such as high specific surface area and high rate properties for the energy storage devices, improved electrical conductivity for the electrode materials having increased interconnection density among the fibers, improved mechanical as well as conducting characteristics of the nanocomposite of the ordered carbon fibers, and provides dimensional stability to Li-ion battery by the help of interlocking of the conductive fiber in the devices.

Figure 1. Relationship between surface area and diameter of the fibers.

186

S. K. Nataraj, B. H. Kim and K. S. Yang

Figure 2. Broad spectrum of nanofiber application in various fields.

2. ELECTROSPINNING PROCESS Electrospinning, one of the most established fiber fabrication methods, has attracted much attention due to the ease by which fibers with diameters ranging from 10 nm to 10 μm can be produced from natural or synthetic materials as precursors. It is also a remarkably simple and versatile technique capable of generating continuous fibers directly from a variety of polymers and composite materials. Typically, the diameters of the electrospun fibers can be controlled in nano scale ranges and the fibers can be deposited as nonwoven mats or aligned into uniaxial arrays and further stacked into multilayered architectures. This technique was demonstrated more than 100 years ago and was first patented in 1930s [18]. However, it did not receive much attention till the early 1990s. Electrospun fibers have found widespread use in a broad range of applications owing to their intrinsic large surface areas. Full time efforts have been invested to prepare the nanofiber using electrospinning process in large industrial scale (100 kg/day) in 2002. This particular invention expanded mass production capacity to reach demands from different field [19]. To date, electrospinning has been explored to process nanometer to micrometer diameter solid, porous, hollow or bi-component continuous fibers, microscopic cups, and nanowires [20-22].

Carbon Nano-Fibers and their Applications

187

2.1. Electrospinning of Nanofibers

Figure 3. The schematic sketch of the electrospinning setup having three major parts: a high voltage power supply, a syringe pump and a grounded target (collector drum).

During electrospinning (Figure 3) process, a high voltage is applied to create an electrically charged jet of polymer solution or melt, which dries or solidifies leaving behind the polymer fiber. One electrode is placed into the spinning solution/melt, while and the other is attached to the collector. As the intensity of electric field is increased, the hemispherical surface of the fluid at the tip of the capillary tube elongates to form a conical shape known as the Taylor cone.

Figure 4. Parameters affecting nanofiber formation in electrospinning process.

A droplet forms due to surface tension on the needle tip, when the polymer solution or melt is delivered to it. Application of high electrical voltage gives a positive charge to this droplet. All charges in the droplet have the same polarity (positive charge), therefore an electrostatic repulsive force is generated between the adjacent entities in the droplet. On the other hand, an electrostatic attractive force exists between the positively charged droplet and the grounded target [23]. Besides the electrostatic repulsive force and electrostatic attractive

188

S. K. Nataraj, B. H. Kim and K. S. Yang

force, surface tension exists on the droplet, which holds the droplet on the needle tip, as in Figure 4. When the applied voltage is low, the surface tension is sufficient to counter balance the electrostatic repulsive force and electrostatic attractive force and the droplet hangs on the needle tip. With increasing voltage, both electrostatic repulsive and electrostatic attractive forces increase, which are sufficient to overcome the surface tension. An electrified jet is thus pulled out from needle tip. The jet undergoes bending, looping, spiraling and finally reaches the target. During this travel, the jet is stretched and elongated by the electrostatic repulsive or electrostatic attractive force, resulting in fine fibers. The electrospinning process involves two major stages: (1) Taylor cone formation and jet initiation; (2) Bending instability and elongation of the jet. A conical shape forms when applied electrostatic repulsive force is balanced with or exceeds the surface tension. This conical shape was named as “Taylor cone”, and was first analyzed by Taylor in 1964 [24]. After initiating from the apex of the “Taylor cone”, the electrically charged jet first travels in a straight line for a few centimeters then it undergoes bending and looping due to the jet instability.

3. PARAMETERS AFFECTING ON FIBER FORMATION There are three major parameters that affects the fiber diameter and morphology: (1) Solution properties such as viscosity i.e., polymer concentration, molecular weight, surface tension, electrical conductivity, dielectric constant and solvent evaporation rate; (2) Processing factors such as voltage, distance between the needle tip and the target and solution flow rate; and (3) Environmental factors such as humidity and temperature. Solution viscosity, which is affected by concentration, molecular weight, or the solvent characteristics, has the most significant effect on electrospun fiber diameter and morphology. Polymer molecular weight also determines chain entanglements in solution; therefore, it can affect the fiber diameter and morphology [25-27]. A strong viscosity dependence on nanofiber formation, diameter and morphological variations were obtained due to enhanced solution properties at higher temperatures, higher conductivity and lower surface tension [28]. The effect of concentration on the morphology of electrospun polyacrylonitrile (PAN)-N,NDimethylformamide (DMF) solution reveal that individual beads can be produced when PAN concentration is low (>2 wt %); the beaded fibers were electrospun from a solution having higher PAN concentration (3–6 wt %) and further increasing PAN concentration lead to uniform size nanofibers as displayed in Figure 5. Surface tension tends to minimize the surface area of the jet by forming beads on the fiber. Therefore, bead formation can be minimized by lowering surface tension of the solution. The surface tension can be lowered by using a low surface tension solvent [25] or by the addition of surfactant [29]. For electrospinning the solution having high surface tension, the addition of a small amount of surfactant into spinning solution helps in decreasing surface tension, which in turn lowers the critical voltage for electrospinning thereby, clean and easy fiber formation can be subsequently obtained. The fiber formation relies upon the instability of the spinning liquid droplet that arises due to a competition of the centrifugal force and the Laplace force induced by the surface curvature.

Carbon Nano-Fibers and their Applications

189

Figure 5. SEM micrographs of the electrospun PAN fibers prepared in various concentrations; (a) 0.5, (b) 1.1, (c) 2.1, (d) 3.1, (e) 5.1, (f) 9.6, (g) 13.8, (h) 16.1, (i) 17.5, (j) 19.0, (k) 19.7, and (l) 20.3 wt %. [voltage=22 kV; Q=1 mL/h; distance between tip and the target was kept constant at 10 cm] [28].

This instability triggers the formation of thin liquid jets emerging from the outward driven polymer solution, yielding solid nanofibers after evaporation of the solvent. The critical field for the onset of fiber formation is predicted to be proportional to the square root of the surface tension. This finding is roughly in agreement with experimental findings. The observation is that a variation of surface free energy keeping the other controlling parameters constant did not significantly affects the fiber diameter and morphology. Surface tension tends to minimize the surface area of the jet by forming beads on the fiber. Therefore, bead formation can be minimized by lowering surface tension of the solution. The surface tension can be lowered by using a low surface tension solvent [25] or by the addition of surfactant [29]. For electrospinning the solution having high surface tension, the addition of a small amount of surfactant into spinning solution helps in decreasing surface tension, which in turn lowers the critical voltage for electrospinning thereby, clean and easy fiber formation can be subsequently obtained. The fiber formation relies upon the instability of the spinning liquid droplet that arises due to a competition of the centrifugal force and the Laplace force induced by the surface curvature. This instability triggers the formation of thin liquid jets emerging from the outward driven polymer solution, yielding solid nanofibers after evaporation of the solvent. The critical field for the onset of fiber formation is predicted to be proportional to the square root of the surface tension. This finding is roughly in agreement with experimental findings. The observation is that a variation of surface free energy keeping the other controlling parameters constant did not significantly affects the fiber diameter and morphology. One of the key parameter is the conductivity of the solution. An increase of the conductivity is expected to increase the surface charge density, which in turn should lead to a

190

S. K. Nataraj, B. H. Kim and K. S. Yang

stabilization of the whipping modes giving rise to fiber extension. Further more, bead formation arising from other modes of instability should be suppressed. Electrospinning involves stretching of the solution caused by repulsion of the charges at its surface. Thus, if the conductivity of the solution is increased, more charges can be carried by the electrospinning jet. When a small amount of salt or polyelectrolyte is added to the solution, the increased charges carried by the solution will increase the stretching of the solution. The increase in the stretching of the solution will also tend to yield fibers of smaller diameter [30]. Since the presence of ions increases the conductivity of the solution, the critical voltage for electrospinning to occur is also reduced. Another effect of the increased charges is that it results in a greater bending instability [31,32]. As a result the deposition area of the fibers is increased. This will also favor the formation of finer fibers since, the jet path is now increased. The temperature plays major role on the cone/jet/fiber morphologies during electrospinning of solutions by varying solution properties and processing variables. Electrospinning at various temperatures and the effect on cone/jet/fiber morphologies as well as birefringence and crystallinity of the collected PAN fibers were well recorded [33]. These results suggested that PAN nanofibers with a diameter of 65 nm are readily obtained by electrospinning at 89°C compared to electrospinning at room temperature. Generally, a solution with a greater dielectric property reduces the beads formation and the diameter of the resultant electrospun fiber. Thus, dielectric constant of a solvent has a significant influence on electrospinning. The bending instability of the electrospinning jet also increased with higher dielectric constant. This may also facilitate the reduction of the fiber diameter due to the increased jet path, covering increased deposition area of the fibers [3436]. However, if a solvent of a higher dielectric constant is added to a solution to improve the electrospinnability of the solution, the interaction between the mixtures such as the solubility of the polymer will also have an impact on the morphology of the resultant fibers. It has been shown that electrospun polysulfone (PS) fibers can be produced by using various solvent systems, and among these, DMF was found to be the most favorable solvent for producing uniform round fibers with smooth surfaces due to its high boiling point, solution conductivity and high dielectric constant compared to other solvents [37,38]. In a typical electrospinning process, fibers are drawn from a solution or melt through a blunt needle by electrostatic forces. Electrospun nanofibers are most commonly collected as randomly oriented or parallel-aligned mats. Randomly oriented fiber mats result when a simple static collecting surface is used, and parallel-aligned mats have been collected by several methods. Generally, it was noticed with the results obtained from SEM analysis clearly indicated that there was no correlation between the needle diameter used and the average nanofiber diameter obtained in the solution electrospinning process, but a broader range of nanofibers diameters was obtained with smaller needle diameter [39]. Figure 6, gives the graphical presentation of the relation between needle diameter and the average diameters of the fibers prepared through them. The internal diameter of the needle or the pipette orifice has a certain effect on the electrospinning process. A smaller internal diameter found to reduce the clogging as well as the amount of beads on the electrospun fibers. The reduction in the clogging could be due to less exposure of the solution to the atmosphere during electrospinning. Decrease in the internal diameter of the orifice was also found to cause a reduction in the diameter of the electrospun fibers.

Carbon Nano-Fibers and their Applications

191

Figure 6. Average nanofiber diameter with varying sizes of needle diameter [39].

When the size of the droplet at the tip of the orifice is decreased, such as in the case of a smaller internal diameter of the orifice, the surface tension of the droplet increases. However, if the diameter of the orifice is too small, it may not be possible to extrude a droplet of solution at the tip of the orifice [30,41-43]. The typical path of the jet is a straight segment followed by a coil of increasing diameter. After several turns were formed, a new electrical bending instability formed a smaller coil on a turn of the larger coil. The turns of the smaller coil transformed into an even smaller coil and so forth until the elongation stopped, usually by solidification of the thin jet. There are three turns, the perimeter of each turn of the coils associated with each bending instability grew monotonically as shown in Figure 7. Along with these parameters humidity also play important roles in some cases. Humidity of the environment will also determine the rate of evaporation of the solvent in the solution. At very low humidity, a volatile solvent may dries very rapidly. The evaporation of the solvent may faster than the removal of the solvent from the tip of the needle. As a result the process may only be carried out for a few minutes before the needle tip is blocked with dried mass. It has also been observed that the high humidity can help the discharge of the electrospun fiber [44].

4. THERMO-OXIDATIVE STABILIZATION Carbon nanofiber is a long, thin strand of material of about 5-1000 nm diameter composed mostly of carbon atoms that are bonded together in microscopic crystals and are aligned parallel to the long axis of the fiber. The crystal alignment makes the fiber incredibly strong for its size. Here this section is explained using PAN as a nanofiber precursor.

192

S. K. Nataraj, B. H. Kim and K. S. Yang

Figure 7. An optical image of the cone on a pendant drop as the tip of the cone approached the critical shape at which a jet was ejected (left). A diagram that shows the prototypical instantaneous position of the path of an electrospinning jet that contained three successive electrical bending instabilities. The straight segment transformed into a three-dimensional coil. The jet path continued and transformed to a smaller three-dimensional coil, with an axis that followed the curved path extrapolated from the first coil. The second spiral eventually transformed to an even smaller spiral and so forth until the jet solidified, by evaporation of the solvent. At least four successive bending instabilities were observed in some experiments [42].

The preparation of CNFs from PAN precursor involves three main stages; electrospinning, oxidative stabilization and carbonization. An adequate stabilization is necessary and important in obtaining good quality CNFs. The stabilization involves heating of the PAN fibers in an oxygen-containing atmosphere to further orient and then cross-link the molecules, such that they can survive higher temperature pyrolysis without decomposing. However, the chemistry of stabilization process is complex, but consists of cyclization of nitrile groups (C≡N) and cross-linking of the chain molecules in the form of –C=N–C=N–. Thus, stabilization involves the transformation of thermoplastic chain molecule into a nonmeltable ladder polymer by cyclization, dehydrogenation and oxidative reactions. This process prevents the melting during subsequent carbonization.

4.1. Structural and Morphological Changes During Stabilization The PAN homopolymer contains highly polar nitrile groups, hindering the alignment of the molecular chains during spinning. Therefore, a copolymer of PAN is used. The comonomer content generally ranges from 2% to 15%; typical comonomers are acrylic acid, methacrylic acid, and methacrylate. The use of comonomers partially disrupts the nitrilenitrile interactions, allowing for better chain alignment. Typical carbon yield from PANbased precursors is 50-60%. Figure 8, shows the surface structures of as-spun PAN nanofibers. The formation of a rough surface is ascribed to the vapor pressure of the solvent.

Carbon Nano-Fibers and their Applications

193

Diameter of the as spun nanofibers is log-normal distribution. When PAN nanofibers were heated in the presence of oxygen, the resulting ladder structures formed in the polymers caused the initial changes in color of nanofibers webs. As shown in Figure 9, oxidative stabilization in air is accompanied by a change in color from white to reddish-brown, tan in the end. Furthermore, when the as-stabilized fibers were carbonized at elevated temperature, up to 450°C in nitrogen gas, a dark brown nanofibers web was obtained. The structural changes during the oxidative thermal stabilization process are shown schematically in Figure 10.

Figure 8. FE-SEM micrographs of as spun PAN nanofibers.

Figure 9. Photographs of nanofibers webs: (A) electrospun as-spun fibers web (white), (B) as-stabilized fibers web (buff) (250°C, 30 min in air), (C) as-stabilized fibers web (yellow) (280°C, 30 min in air), (D) as stabilized fibers web (reddish-brown) (300°C, 30 min in air), (E) as stabilized fibers web (brown) (300°C, 120 min in air), (F) as-stabilized fibers web (tan) (310°C, 120 min in air), (G) ascarbonized fibers web (dark brown) (450°C, 120 min in argon gas) [45].

194

S. K. Nataraj, B. H. Kim and K. S. Yang

A FT-IR spectrum of PAN nanofibers after different heating rates and holding times is shown in Figure 11. The vibrations characteristic of PAN structure are those of CN nitrile group at approximately 2243–2241 cm−1, and the bands in the regions 2931–2870, 1460– 1450, 1380–1350 and 1270–1220 cm−1 which are assigned to the aliphatic CH group vibrations of different modes in CH and CH2 [46]. As the heating temperature increased, the most prominent structural changes were the decrease in the intensities of the 2243–2241 cm−1, attributed to C≡N band and the decrease of those for aliphatic C–H, respectively, concomitant with the advent and increase of a shoulderlike peak at 1700 cm−1 due to cyclic C=O, the band at 1590 cm−1 due to C=N, C=C, N=H mixed and the band at 810 cm−1 is due to C=C=H) [47]. These spectroscopic results show that some chemical processes occurred during the stages of oxidative stabilization [48]. Firstly, reaction of nitriles results in conjugated C=N containing structures which result from intramolecular cyclization or intermolecular crosslinking. Secondly, the generation of conjugated C=C structures results from dehydrogenation or from imine–enamine tautomerization and subsequent isomerization [49]. At last oxidative temperature treatment gives rise to carbonyl groups on the fiber matrix.

Figure 10. Schematic representation of structural changes in PAN nanofibers during thermal-oxidative stabilization process.

Carbon Nano-Fibers and their Applications

195

Figure 11. FT-IR spectra of PAN nanofibers at different heating temperature with different holding times.

5. CARBONIZATION OF THE ELECTROSPUN NANOFIBERS The second step involves carbonizing heat treatment of the stabilized PAN fibers to remove non-carbon elements in the form of different gases like H2O, NH3, CO, HCN, CO2 and N2. Carbonization is generally carried out in an inert atmosphere [50-52]. The stabilized fibers are carbonized in nitrogen in the 1000oC to 1700oC range. Various gases evolve during pyrolysis of PAN. Stretching during stabilization minimizes the need for stress during carbonization and graphitization. During the carbonization process, carbon content increases to above 90% and a three-dimensional, near-amorphous carbon structure with microcrystal forms. These fibers can be further heat-treated between 2000oC and 3000oC in an inert environment for graphitization. In the carbonization process (Figure 12), the heating rates in the two zones are crucial. The first is up to about 600oC, requires a low heating rate of less than 5oC/min, so as to make slow mass transfer. A faster mass transfer at higher heating rates may cause surface irregularities in the form of pores due to the diffusion of evolving gases primarily. This zone

196

S. K. Nataraj, B. H. Kim and K. S. Yang

is very crucial because it involves most of the chemical reaction and the evolution of volatile by-products. In this step, liberation of moisture is due to the crosslinking of ladder polymer chains through the oxygen containing groups. In second zone, heat treatment between 6001300oC, is applied at higher heating rates because of the reduced possibility of damage due to exothermic reactions or the evolution of by-products. During the early stages of carbonization in the temperature range of 400-600oC, the hydroxyl groups present in the oxidized PAN fibers start crosslinking condensation reactions, which makeup in reorganization and coalescence of the cyclized structures. This crosslinking probably fixes the structure of the polymer, which the remaining linear segments either become cyclized or undergo chain scission evolving the gaseous byproducts. These cyclized structures further undergo dehydrogenation and instigate to link up in the lateral direction, producing graphite like structure consisting of three hexagons in the lateral directions and bounded by nitrogen atoms. To analyze the microstructure of the CNFs, Raman spectroscopy can be used because it is very sensitive to subtle variations in the structure of carbon-based materials. It is predicted that condensation reactions in which the carbon atoms in one cyclized sequence fit into spaces left by the nitrogen atoms in an adjacent cyclized sequence.

Figure 12. Formation of intermediate ladder structure via intermolecular crosslinking of stabilized PAN nanofibers through oxygen containing groups during high heat treatment process, finally yielding carbonized structure.

Carbon Nano-Fibers and their Applications

197

5.1. Electrical Conductivity of the E-Spun Carbon Nanofibers Although the process of charge conduction in small diameter nanofibers is not clearly understood, it has been suggested that unusual current conduction properties arise when the size of the fibers is reduced below a certain critical thickness. Hence, it is expected that new contacts may behave more favorably when they are confined to operate in such reduced dimensional regimes of nanoscale range. Stabilized nanofibers and carbonization at 600oC did not generate a conductance high enough to be detected. At temperatures higher than 600oC, the conductivity increased sharply with an increasing carbonization temperature. The electrical conductivities of the electrospun carbon nanofiber in the form of web are shown in Figure 13, for PAN (1.95 S/cm), polyimide (PI), polybenzimidazoles (PBI) and pitch (52.03S/cm) carbonized at 1000 oC, for 1 h in N2 atmosphere. Carbonized pure PAN nanofibers have in general electrical conductivity of a few S/cm, much larger than that of polymer nanofibers without carbonization. In comparison with the polymer based electrospun carbon fibers, pitch based carbon fiber web shows much higher electrical conductivity due to its morphological structure derived from molecular structure of the precursor pitch as shown in Figure 14(a) & (b). We measured the electrical conductivity parallel to the winding direction (or fiber axis) by the four-point probe method.

Figure 13. SEM micrograph of PAN CNFs and electrical conductivity of carbon nanofibers produced from different polymer precursors, carbonized at 1000oC for 1 h.

Advantage of the carbonized PAN nanofiber is that the nanofiber surface can be modified and functionalized by activation process under different ambient conditions. Rayon and PAN based carbon fibers are produced with higher-cost chemical intermediaries such as polymers. Recent year, attentions has been diverted to on scalable technology that enables mass production of high-performance isotropic pitch carbon fibers from lower-cost feedstocks derived from petroleum or coal tar as well as naphthalene sources (see Figure 14(c)). Researchers are devoting their efforts to develop both, materials and process technology to custom-engineer pitch carbon fiber at relatively low cost and in large quantities. Mesophase pitch turns from different sources into partly hydrogenated fused ring aromatic hydrocarbon through hydrogenation reaction. Due to the change of carbon atom orbit and the molecular structure, the π-electron conjugation system is reduced and the steric

198

S. K. Nataraj, B. H. Kim and K. S. Yang

hindrance increases. The interaction between molecules is weakened, resulting in the formation of pitch with higher molecular weight, low melt temperature and lower viscosity. Traditional production technologies of pitch-based CFs are melt spinning (M-spun) and meltblown spinning, producing typical fibers with diameters of above 10 μm by extruding melt through a spinneret, and subsequent drawing to be reduced diameter. The orientations of precursor pitch along the fiber are induced in the spinning process, where the aromatic planes in the pitch components could be aligned by shearing in the nozzle, and by stretching the spinning line. Isotropic pitch precursor (IPP) can be processed to carbon fiber web with diameter ranging from 2 to 3 μm through electrospinning [31]. The CFs can be activated to ultra-fine carbon fiber morphology having good flexibility for easy compression process. Figure 15, shows the comparative morphological structures of pitch carbon fibers produced from electrospinning and melt spinning processes.

6. VAPOR GROWN CARBON NANOFIBERS (VGCFS) Carbon nanofibers (CNFs) are nanoscale cylinders of graphitic carbon with a high aspect ratio. A CNF generally can be composed of two phases of carbon. The catalytic phase is the one, which forms and propagates by the catalytic action of the metal seed particle, using this the CNF grows. This phase tends to be composed of well ordered, graphite-like planes of carbon atoms, although with some curvature as required to form a cylindrical material rather than a sheet. The deposited phase is formed by chemical vapor deposition (CVD) of carbon on top of the catalytic phase.

Figure 14. Chemical structure of unit molecules derived from (a) petroleum, (b) coal tar, (c) naphthalene. (♦: active site).

Carbon Nano-Fibers and their Applications

199

Figure 15. SEM micrographs of comparative morphological structures of pitch carbon fibers produced from electrospinning (left column) and melt spinning processes (right column).

VGCFs are characterized by the highly preferred orientation of graphitic basal planes parallel to the fiber axis with annual ring texture in the cross-section, which give high mechanical strength, excellent electrical conductivity and high graphitizability. There are two methods to grow carbon fibers developed. (1) Seeding catalysts on a substrate and (2) Fluidizing catalysts in a space. Carbon fibers may be grown on several types of substrates including carbon, silicon, quartz etc., and from many hydrocarbon precursors namely; acetylene, benzene natural gas etc. But in all cases, growth is favored in hydrogen atmosphere. Figure 16, shows the apparatus used by Endo to produce carbon fibers over fluidizing catalysts methods [53-55]. Vapor grown carbon nanofibers are produced by catalytic dehydrogenation of hydrocarbons (e.g. methane, benzene or naphthalene) at high temperatures in a flow system. Small catalytic transition metal particles (e.g. iron, ~10 nm diameter) are formed by spraying a solution of the metal salt as a mist into a heated stream composed of hydrogen, hydrocarbon and H2S vapors [56,57]. Metal particles serve as the nucleation sites for each fiber. Initially formed carbon filament with a diameter directly proportional to the metal particle diameter, i.e., 10–15 nm, grows from the catalyst [58–60]. The initial metal particle size must be very small (<15 nm) because a considerable decrease in activity is observed when the metal particle diameter exceeds 15 nm [61]. Thus, it is very important to retain the small catalyst particle size in the flowing gas and to avoid particle coagulation to larger ineffective diameters [62]. Metal particles begin extruding long slender hollow filaments of graphitic carbon at temperatures above 900oC. At 1050–1300ｏC the filaments graphitic order is high. These fibers grow as rapidly as 1 mm/min and may lengthen for several minutes until the iron

200

S. K. Nataraj, B. H. Kim and K. S. Yang

particle is deactivated or until they leave the reaction zone [63,64]. An outer layer of carbon is deposited on the filament by CVD, if the filaments remain in the reaction zone.

Figure 16. Schematics reactors for the production of vapor grown carbon nanofibers in floating reactant methods [53].

6.1. Growth Mechamism There are two type of growth mechanism for the production of vapor grown carbon nanofibers. There are two general CNF growth mechanisms; first is tip growth mechanism and second is root growth mechanism for vapor grown carbon nanofiber preparation as shown in Figure 17(a) and (b). Stepwise growth of carbon nanofibers in vapor deposition can be pointed out as following, taking iron nanoparticle as example; (1) An inducing stage occurs for the iron particle dissolving carbon to form γ-Fe; (2) Carbon atoms dissolve continuously in to γ-Fe to

Carbon Nano-Fibers and their Applications

201

form Fe-C liquid, and then precipitate on the other side to form VGCF, competing against diffusion toward the center to form Fe-C liquid; (3) Fe3 C forms on the surface when the whole catalyst particle changes into Fe-C liquid; (4) A balance is built up between the formation and reduction of Fe3C and VGCF grow continuously, and (5) If the balance is broken, pyrocarbon deposites, which ends up the growth of the carbon fiber.

Figure 17(a). Schematical representation of tip growth mechanism of CNF following gas-liquid-solid phase using catalyst in vapor grown process.

As stated, CNFs are produced as a result of decomposition of hydrocarbons on the surface of transition metal nanoparticles. During this reaction, the hydrocarbon first adsorbs and then decomposes on the surface of the metal particle. The resulting carbon atoms then dissolve in and diffuse through the metal particle. The precipitation of carbon from the saturated metal particle leads to the formation of CNF [65,66]. The rationale for choosing these metals as catalyst for CVD growth of CNF lies in the phase diagrams for the metals and carbon. At high temperatures, carbon has finite solubility in these metals, which leads to the formation of metal-carbon solutions and, therefore, the aforementioned growth mechanism. In the majority of the cases, the catalyst particle is carried away from the surface of the support, and there is sufficient evidence to suggest that the CNF adhere very strongly to the support.

202

S. K. Nataraj, B. H. Kim and K. S. Yang

Figure 17(b). Schematical representation of root growth mechanism of carbon nanofiber using liquid as hydrocarbon material.

Figure 18. Schematics of growth of graphitic filament as more carbon is supplied by gas phase and tends to precipitation followed lengthening with time, finally yielding concentric cones of graphitic planes.

Carbon Nano-Fibers and their Applications

203

The growth of such nanofibers involves the dissolution and diffusion of carbon at the exposed surface of the catalyst particles and precipitation at the metal/support interface. In the case of low metal support interaction, the catalyst particle is carried on top of the growing filament (Figure 18) [67,68]. This “lengthening” stage stops when the catalyst particle becomes covered with a carbon layer; thereafter, thickening occurs by CVD of pyrolytic carbon, which grows in concentric layers [69,70]. In recent findings [71,72], the nonsolid drop consists of an iron-carbon eutectic, the particle size influencing the melting point of the particles have been proposed. In this growth process, the role of hydrogen is to avoid the poisoning of the metallic particle by carbon deposition. In a recent investigation on the role of sulfur in the VGCF production [73], suggested that the liquefaction of the particle increases the rate of filament nucleation. Although Tibbetts’ work was relative to a continuous process with the production of short filaments, it supports the model of Gadelle. In this model, the metallic particle is located at the tip of the fibers. Very recently a second model has been proposed [74] for the production of long VGCF, in which the liquid drop consists of a carbon material: coronene (C24H12, melting point 438°C). ln this case, the role of the metallic particle is to initiate the growth process by a discharge of hydrogen that produces a molten drop of coronene. In this model the metallic particle remains on the support.

7. COMPARISION BETWEEN E-SPUN AND VAPOR GROWN FIBERS Electrospinning offers advantages like, control over morphology, porosity and composition using simple equipment. Variety of fibers can be produced in the process. Nanofibers with of 40-2000 nm can be produced by selecting suitable combination of polymer and solvent precursor system. Morphological structure of the nanofiber can be tailored to reach requirements in various fields like filtration products, biomedical applications, tissue engineering to produce artificial blood vessels, non-woven fabrics, fuel cells, nanofiber cloths/mats for electrode support etc. Advantages of electrospun material are intrinsic to the process. Electrospinning is a noninvasive method of fiber production. The process can run at ambient temperature and the process of solidification does not involve any coagulation chemistry, simply the removal of solvent from the system. Electrospun fiber has three potential advantages depending on the application. First, having small diameters the fibers possess very high surface area to volume ratio making them ideal for any application that depends on a large surface area, (e.g. catalysts, bioreactor substrates, active filtration). Second is by virtue of the mechanism of fiber formation, where the fiber is stretched by a whipping motion while in flight giving a very high draw ratio. This useful in energy storage applications, potentially can result in a high degree of molecular alignment and minimize defects in the bulk material which would allow the fiber to attain a tensile strength closer to the theoretical strength, i.e., perfect for any high strength fiber application in nanocomposites. Finally, the size and morphology that typical electrospun fiber forms with nano scale fiber diameter and pores between fibers in a mesh, makes it ideal for certain specialist applications, for e.g. fiber size similar to collagen matrix in biological tissue so ideal for tissue scaffolds or nano pores in mesh are ideal for ultra-filtration.

204

S. K. Nataraj, B. H. Kim and K. S. Yang

In VG process, growth precedes one molecule at a time and one layer at a time. The process is slow; it takes more time to grow a small amount of CNFs. The advantage of vapour growth is that very pure crystals can be grown by this method, while the disadvantage is that it is slow. The catalyst used for the growth of carbon fibers in vapor grown methods gives a greater degree of control and is more feasible for scale-up. However, the metal catalysts remain within the deposited products and must be removed using caustic reagents. Continuous fiber pre-forms are treated with a catalyst or catalyst precursor and processed to yield VGCF produced in-situ resulting a highly entangled mass of VGCF infused with the continuous fiber pre-form. VGCFs are used in various applications including; 1) Automotive fuel loading systems, mirror housings, electrostatically spray painted automotive fenders, finish for painting, interior parts, bumpers, long-lasting tires [75], adds electrical conductivity properties to fiberglass reinforced plastics; 2) Aerospace composite structures, thermal management, EMI/RFI shielding; 3) Environmental, chemical waste treatment and purification; 4) Industrial silicon wafer production, lithium, lithium-ion, lithium polymer batteries, alkaline batteries, fuel cells, disk drive components, dust free handling trays for use in clean rooms, rubber reinforcement, polymer reinforcement, crack mitigation for concrete, use in synthetic wood and so on.

8. APPLICATIONS OF CARBON NANOFIBERS Worldwide production of carbon fibers has increased from 15 million kg in 1997 to 20 million kg in 2002 and it has reached to more than 30 million kg by 2008. Carbon fiber costs have come down significantly since the 1980s, and the PAN and pitch-based fiber production technology appears to have matured. Carbon fibers are used in aerospace, aircraft, nuclear, sporting goods, biomedical, and high-end automotives. The high strength, high modulus and low density of carbon fibers make them suitable for aerospace and sporting-goods applications. Carbon fibers are also used for chemical protective clothing, electromagnetic shielding and as non-woven fire retardant. Table 1, gives the over view of the properties of CNFs and their particular usage in different fields. Table 1. Characteristics and applications of CNFs produced from electrospinning Properties required

Application field

1

Mechanical properties

Aerospace, road and marine transport, sporting goods

2

High performance and high temperature

Missiles, aircraft brakes, aerospace antenna and optical instruments etc

3

Good vibration damping, strength, and toughness

Audio equipment, loudspeakers for Hi-fi equipment, pick-up arms and robot arms

4

Electrical conductivity

Automobile hoods, novel tooling, casings and electronic equipments

5

Biological inertness and x-ray permeability

Medical implants, in prostheses, surgery and x-ray equipment

Carbon Nano-Fibers and their Applications

205

Properties required

Application field

6

Fatigue resistance, self-lubrication, high damping

Textile machinery, general engineering

7

Chemical inertness, high corrosion resistance

Nuclear plants, valves, seals, and pump components in process plants

8

Electromagnetic properties

Large generator retaining rings, radiological equipment

8.1. CNFS IN ELECTRONIC AND ELECTROCHEMICAL DEVICES Because of their unique tailor made properties, nanofibers attract much attention as these bear many potential applications in different fields of commercial interest. Their unusual structural and electronic properties make carbon nanostructures applicable in electrochemical double-layer capacitors (EDLC), batteries, catalyst supports, energy storage, fuel cells and different electronic devices. Many attempts have been made to produce PAN CNFs with large surface area, high power energy and long lifecycles to apply in supercapacitors that are intensively investigated as the backup for energy storage systems. Their double-layer capacitance strongly depends on the type and form of electrode materials. Various forms and textures of porous carbons have been examined as possible electrode materials for supercapacitors. In view of their scientific and technological importance, at more elevated temperature the channel between carbon basal planes in PAN will be filled up resulting in low surface area. After CNFs were treated in CO2 atmosphere at 880°C for several minutes, the surface of CNFs get etched and the new pore formed causing an improved surface area again [75,76], which is desired for electrochemical applications. Carbon materials have been extensively studied as electrodes for energy storage devices since they present a very interesting electrochemical behavior. These materials may have high electrical conductivity and can both donate and accept electrons. Carbon electrodes have a good polarizability and their properties are tunable depending on the porosity, thermal treatment, microtexture, hybridization, content of heteroatoms, etc. Moreover, they are chemically stable in most solvents and present relatively low cost and easy processability. One important application is in rechargeable batteries in which carbon materials are used as a lithium reservoir at the negative electrode. Additionally, the storage of energy in supercapacitors, based on the electrical double layer and pseudocapacitance of carbon materials have also been deeply analyzed with very promising results [77–79]. Carbon nanofibers have shown considerable promise when compared to similar devices using conventional activated carbon powders. The most noticeable advantage in using these materials in electrochemical capacitors is that the CNFs offers very fast frequency response in a device. Single cell devices fabricated from carbon nanotubes and nanofibers have displayed “knee” frequencies of 100 and 50 Hz, respectively, which makes these materials among the fastest carbons every tested for this application. Various nanocomposite carbon nanofibers have been developed by electrospinning process to enhance the electrochemical properties of the materials [80,81]. Recently, a solution method was adopted to prepare porous and smaller-sized fibrous carbon in the form of thin webs using zinc chloride skipping activation process, as shown in Figure 19. Specific

206

S. K. Nataraj, B. H. Kim and K. S. Yang

surface areas obtained for samples prepared from PAN containing 1, 3, and 5 wt. % zinc chloride were respectively, 310, 420, and 550 m2/g. Higher surface areas of the fibrous materials exhibited higher specific capacitances. The CNFs webs with 5 wt. % zinc chloride exhibited the largest specific surface area (550 m2/g–1), highest capacitance (140 F/g–1) with a good rate capability because of their smaller fiber diameters and large specific surface area. Performances of the hybrid supercapacitor with electrodes of carbon/RuO2 were investigated [82] in 6M of aqueous KOH solution. Electrical conductivity of PAN-based activated carbon nanofibers (ACNFs) increased from 0.42 S/cm to 0.98 S/cm by the dispersion of 3 wt. % of multiwalled carbon nanotube (MWCNT). The capacitances of pristine PAN, MWCNT/PAN, RuO2/PAN, and RuO2/MWCNT/PAN-based ACNFs were respectively, 140, 180, 390 and 530 F/g. The capacitance as increased by 4 times by composition with 3 wt. % of MWCNT and deposition of 20 wt. % of RuO2, even though specific surface area was reduced by deposition of RuO2 down to 1/3 of the original value.

Figure 19. a) Low-resolution FE-SEM image showing a highly thin (ca. 55lm) web prepared by electrospinning 5 g PAN. b) FE-SEM image exhibiting a round, individual, organic nanofiber with a diameter of ca. 200 nm. (c–e) FE-SEM images of the electrospun organic nanofibers from PAN solutions containing different weight percents of zinc chloride, c) 1, d) 3, and e) 5 wt%, and their corresponded diameter distributions [82].

PAN based CNFs electrodes exhibiting a large accessible surface area (derived from the nanometer-sized fiber diameter), high carbon purity (without binder), relatively high electrical conductivity, structural integrity, thin web macromorphology, a large reversible capacity (ca. 450 mAh g–1), and a relatively linearly inclined voltage profile, were fabricated using PAN solution as shown in Figure 20. It is envisaged that these characteristics of this novel carbon material will make it an ideal candidate for the anode material of high-power lithium-ion batteries (where a high current is critically needed), owing to the highly reduced lithium-ion diffusion path within the active material [83]. Since, sp2-based carbon materials were adopted as the anode material in commercialized lithium-ion batteries (LIBs) to solve a safety problem, various types of carbon textures have been examined in order to improve the specific energy and power density of batteries. Both the capacity and cyclic performance of

Carbon Nano-Fibers and their Applications

207

LIBs strongly depend on the carbon texture, owing to the very different lithium ion (Li+) insertion and removal mechanisms available. More recently, the development of high power LIBs has been requested to meet the growing demand for use in lighter, electronic portable devices and hybrid electric vehicles (HEVs). Thus, intense studies have focused on nanometer-sized carbon materials as promising materials for storing Li+ because they show an increased capacity as an active material and also improved cyclic characteristics as an additive to the anode material.

Figure 20. (a) Charge–discharge curves for nanofiber webs thermally treated at 700, 1000 and 2800 °C (current density=30 mA g–1, second cycle). (b) Some of commercially available carbon nanofiber composite anode materials [83].

8.2. CNFs in Energy and Hydrogen Storage Systems Porous carbons have also been extensively studied for energy storage application, giving higher values for storage and much better reproducibility. Currently, four methods are being considered for hydrogen storage in commercial applications; pressurized gas storage, liquefied hydrogen, selected metal hydrides and refrigerated super activated carbon. Pressurized gas offers the advantage of being simple however, in mobile applications the large volume coupled with the small capacity will limit its practicality. Liquefied hydrogen is

208

S. K. Nataraj, B. H. Kim and K. S. Yang

expensive since it requires constant refrigeration and loss of the gas by evaporation is inevitable. While the latter two approaches may offer benefits over the other technologies with regard to safety aspects, they do however have their own set of drawbacks. Metal hydrides are heavy, expensive, release heat during the hydrogen absorption process [84] and require the use of about one third of the storage energy during the release of the hydrogen fuel. Hydrogen has been considered as an ideal energy medium for replacing fossil fuel to mitigate the global environmental issues. A challenge we have to contend with is how to transport the hydrogen safely and efficiently. Recent claims that some new carbon materials, especially CNFs can store a large amount of hydrogen, because of their extremely high surface area, active carbon constitute without any doubt the preferred adsorbent in many processes. Carbon molecular sieves have been known for several decades and present an alternative choice for many commercial gas separation processes. These structures are produced from a variety of carbonaceous solids of different origins and precursors. Activated CNFs possess a wide pore size distribution, where the fraction of micro and nanopores are rather small. While these materials are very effective for the adsorption of a variety of molecules, one has to consider that the interaction between the adsorbent and the adsorbate is only of a physical nature and as a consequences, the retention of gases is only achieved at extremely low temperatures. The use of ACNFs for energy storage at high temperatures has been found to be ineffective since the solid takes up storage value without appearing to add any substantial benefits to the overall capacity [85-87]. Recent developments resulted in obtaining ordered porous carbons with tailored pore size and the role of porosity has been discussed as well as the importance of the dangling carbon atoms in the hydrogen storage. In general, the higher the narrow porosity, the higher the hydrogen uptake and the uptake decreases with decrease in the number of dangling atoms [88-96]. There was an attempt to [97] electrospun the PAN nanofibers with or without iron(III) acctylacetonate to induce catalytic graphitization in the range of 900-1500°C to obtain ultrafine carbon fibers having diameter range of about 90 to 300 nm. Hydrogen storage capacities of as produced CNFs were evaluated by a gravimetric method using magnetic suspension balance (MSB) at room temperature and at 100 bars pressure. The CNFs showed hydrogen storage capacities of 0.16-0.50 wt. % that increased with increasing of carbonization temperature, but that of CNF at 1500°C had the lowest. Hydrogen storage capacities of GNFs with low surface areas of 100-250 m2/g were 0.14 - 1.01 wt. %, others reported [98,99] the effects of iron(III) acetylacetonate (IAA) on carbonization behavior of electrospun polyimide and PAN nanofibers resulting in GNFs. Considering the advantages of electrospun CNFs, there was an interesting attempt on investigating practical ways to find the hydrogen storage efficiencies using chemical agents such as zinc chloride and potassium hydroxide [100]. Electrospun CNFs have been modified to get high specific surface area and pore volume by activation for improved hydrogen storage. Comparing the pore structure of KOH activated CNFs and ZnCl2 activated CNFs, it is believed that KOH activation is much more effective to increase specific surface area and total pore volume than ZnCl2 activation. Even though the capacity of hydrogen adsorption was increased with increasing specific surface area and pore volume, it was found that in PAN-based electrospun activated CNFs, the most effective factor which can attract the hydrogen adsorption capacity positively is mainly the pore volume in the range of pore width i.e., from 0.6 to 0.7 nm.

Carbon Nano-Fibers and their Applications

209

Figure 21. FE-SEM images of activated carbon fibers and adsorption equilibrium isotherms of benzene on PAN carbon nanofibers and A–10 [102].

210

S. K. Nataraj, B. H. Kim and K. S. Yang

In addition, carbonization of electrospun polyvinylidenefluoride(PVdF) nanofibers with an average fiber diameter of 200–300 nm produced microporous carbon nanofibers with much higher surface areas than PAN-based graphitized nanofibers (GNFs). The hydrogen storage capacity of the PVdF carbon fiber was much lower than that of PAN-based GNFs [101]. Carbonized and activated electrospun PAN fibers were also tested other than hydrogen storage in some studies [102] such as, the adsorption equilibrium, thermal desorption and kinetics of organic compound using PAN CNFs to assess their possibility as alternative adsorbent to commercial ACNFs, activated commercial carbon nanofibers (A–10) (see Figure 21). Comparative analysis [103] on energy distribution functions provided significant information on the energetic and structural heterogeneities of CNFs. Furthermore, an investigation of adsorption equilibrium and kinetics of methylene blue (MB) and congo red (CR) (include characteristics of MB and CR) revealed that adsorption capacity and kinetics of MB were much higher and faster than that of CR for a given sample.

8.3. CNFs in Fuel Cell Applications Proton exchange membrane fuel cells (PEMFCs) are promising energy conversion systems, while further improvement of electrochemical performance as well as a considerable decrease in noble metal loading is essential for their commercialization. Applications of carbon nanofibers, instead of state-of-the-art carbon black, may realize such improvements by nano-structuring electrode catalysts. In general, carbon black with a satisfactory electronic conductivity having large specific surface area, suitable for catalyst dispersion, while carbon nanofibers possess an excellent electronic conductivity along with the fibers. Triethylamine catalyst was used to increase the molecular weight of poly(amic acid). In electrospinning, diameter of PAA nanofibers was controlled by molecular weight of poly(amic acid) and concentration of poly(amic acid)/N-N dimethylformamide solution. The diameters of polyimide based carbon nanofibers reached to a minimum value of around 100 nm. The conductivity of the carbon nanofiber web increased with the decreasing of fiber diameters, while the mechanical strength of polyimide and carbon nanofiber web was diameter independent. Further more, Pt was successfully impregnated in carbon nanofiber. Reduction accompanied with carbonization in Ar gas at up to 1000oC produced small and well-dispersed Pt nanoparticles (3-5 nm). Our earlier paper [104] reports the novel results regarding the effects of electrospun carbon nanofibers (e-CNF) as a catalyst support by comparison with the commercial Vulcan XC-72R (denoted as XC-72R) as granular particles. The e-CNF was synthesized by stabilizing and carbonizing the electrospun PAN–based fibers. The e-CNF showed an average diameter of 250 nm with a rough surface and was partially aligned along the winding direction of the drum winder. The characteristic morphology was fundamentally dependant on the shape of the carbon materials. The average pore size of the e-CNF was 2.36 nm, while that of the XC-72R was 10.92 nm as shown in Figure 22. The morphology of e-CNF was developed by shallow pores with rough surfaces due to the effects of electrospinning and carbonization, while that of the XC-72R was largely developed by mesopores rather than micropores due to the granular shape. Compared to XC-72R, the performance of the MEA prepared by e-CNF was excellent, owing to the characteristics of the morphology and the

Carbon Nano-Fibers and their Applications

211

enhanced electrical conductivity. The Pt utilization of Pt/e-CNF was 69%, whilst that of Pt/XC-72R was 35% as shown in Figure 23.

Figure 22. SEM micrographs of fine dispersed Pt coating on carbon (Vulcan XC) and Pt dispersed CNFs as well as their pore size distribution pattern. [104]

212

S. K. Nataraj, B. H. Kim and K. S. Yang

Figure 23. PEMFC unit cell used for the analysis of Pt/C nanofibers and polarization curves resulted.

8.4. E-Spun CNFs in Photocatalytic Applications For air and liquid purification, immobilized photo-catalysts on support materials are usually employed. Generally, surface area and activity are reduced by the immobilization of photo-catalysts. Therefore, support materials with high surface areas have been applied to immobilize photo-catalysts. Carbon nanofibers and activated carbon has commonly been used as a support material. Activated carbon has a high surface area, which is closely related to the enhancement of adsorption and photo-catalytic activities. Another problem is insufficient adhesion between photo-catalysts and support materials. To achieve good adhesion, complicated coating processes have been introduced. Therefore, it is necessary to search for simpler and more convenient immobilization methods. Recently, we have investigated the use of activated e-spun CNFs as support materials of TiO2 and SiO2 using electrospinning process. Further, electrospun composite nanofibers were stabilized and carbonized in nitrogen to get composite CNFs. The SEM images (see Figure 24) showed that CNFs coated with TiO2/SiO2 appeared to have some deficient homogeneity in dispersion. The energy dispersive x-ray spectrometry (EDX) of the (TiO2/SiO2)/CNFs was used to investigate content elements. The elemental analysis reveals the presence of TiO2/SiO2 on the CNFs surface, the elemental ratio of C, Ti and Si are 89, 7, and 4% respectively. However, after post-oxidation, the total carbon and nitrogen contents were significantly reduced and the BET surface areas of the composite fibers were increased from 254 to 272 for (TiO2/SiO2)/CNFs and that of from 329 to 368 for TiO2/CNFs as shown in Table 2. The crystallite size of anatase phase was 20 nm and fraction of one was 80%. The reduction in surface area of composite CNFs with silica was due to a combinational effect of high temperature and deposited TiO2. Crystallization process of TiO2, either from amorphous to anatase phase or from anatase to rutile phase, was strongly delayed when silica was incorporated into the titania matrix. Finally, nanoscaled TiO2 particles on a SiO2/CNFs surface formed resulting in a nano-structured composite TiO2/SiO2-carbon photo-catalyst, used for the degradation of methylene blue dye analogous to oxidative wastewater treatment. Figure 25, shows the efficiency of composite catalyst in degrading the dye about 96 to 99 % by removing color within 3 hours, which was superior by three folds that of removal achieved with the simple TiO2 photo-catalyst.

Carbon Nano-Fibers and their Applications

213

Figure 24. The SEM images EDX of CNFs coated with TiO2/SiO2 nanoparticles.

Table 2. Crystal properties and surface area of different composites samples of (TiO2/SiO2)/CNFs and TiO2/CNFs Fraction (%) Sample

Crystal size (nm)

Heat temp

BET SA (m2/g)

Rutil

Anatase

Rutil

Anatase

900 oC

20

80

29

20

254

800 oC

8

62

22

14

272

900 oC

70

30

40

36

329

800 oC

58

42

30

25

368

(TiO2/SiO2)/CNFs

TiO2/CNFs

8.5. CNFs in Composite Applications One of the most important applications of traditional (micro-size) fibers, especially engineering fibers such as carbon, glass, and Kevlar fibers, is to be used as reinforcements in composite developments. With these reinforcements, composite materials will provide superior structural properties such as high modulus and strength to weight ratios that are generally not achievable by other monolithic materials.

214

S. K. Nataraj, B. H. Kim and K. S. Yang

Figure 25. Results of photo-catalytic degradation of methylene blue dye using and CNFs, TiO2/CNFs and (TiO2/SiO2)/CNFs as well as actual photograph of degraded dye using (TiO2/SiO2)/CNFs photocatalyst.

However, nanofibers will also eventually find important applications in making nanocomposites because they have even better mechanical properties than micro-fibers of the same materials and hence, superior structural properties of nanocomposites can be anticipated [105]. For instance, if there is a difference in refractive indices between fiber and matrix, the resulting composite becomes opaque or nontransparent due to light scattering. This limitation can be circumvented when fiber diameters become significantly smaller than the wavelength of visible light. These nanofibers provide a higher ratio of surface area to mass than carbon fibers ordinarily used in composites. Carbon nanofibers can be useful in filters, as a support for catalysts in high temperature reactions, in composites to improve mechanical properties or for thermal management in semiconductor devices [106]. Short fibers can offer advantages of economy and ease of processing. When the fibers are not long, the equal strain condition no longer holds under axial loading, since the stress in the fibers tends to fall off towards their ends (see Figure 26). This means that the average stress in the matrix must be higher for the long fiber case. The lower stress in the fiber and correspondingly higher average stress in the matrix (compared with the long fiber case) will depress both the stiffness and strength of the composite, since the matrix is weak and less stiff than the fibbers. Therefore, interest in quantifying the change in stress distribution as the fibers are shortened. Several models are in common use, ranging from fairly simple analytical methods to complex numerical packages.

Carbon Nano-Fibers and their Applications

215

Figure 26. Schematical representation of axial holdings of nanofibers in composite materials.

The simplest is the so-called "shear lag" model. This is based on the assumption that all of the load transfer from matrix to fiber occurs via shear stresses acting on the cylindrical interface between the two constituents. The build-up of tensile stress in the fiber is related to these shear stresses by applying a force balance to an incremental section of the fiber. Carbon-fiber-reinforced composites are made up of polymer, metal, ceramic, and carbon matrices. Although the composite materials do not yield the same mechanical properties as the fibers alone, the matrix adds other properties to the composite for specific applications and holds the fiber together such as automotive parts, high strength gas storage tanks, high performance composite sheets (see Figure 27). As compared to most matrices, carbon fiber's coefficient of thermal expansion is typically two orders of magnitude lower, therefore it can improve the dimensional stability of the composite. The early development of carbon fibers was prompted for defense and space applications. Now, use of carbon fiber in the civilian aerospace industry is rapidly increasing. For example, in a Boeing 767, carbon-fiber composites made up 3% of the total materials. This increased to 7% for the latest Boeing 777 model, while the Boeing 7E7, to roll off the assembly line in 2008, will have about 50 wt.% composite materials and will use 20% less fuel comsumption than airliners of the same size. Today, most carbon nanofiber still finds its way into aerospace applications, such as the new A380 and 787 Dreamliner and carbon nanofiber use is increasing in industrial applications as well [107-110]. Today most of the satellites in orbit use monopropellant hydrazine as a propulsion subsystem for orbit correction and positioning operations. The thrust is obtained by catalytic decomposition of the monopropellant on a highly loaded catalyst containing about 30 to 40 % of iridium supported on alumina [111]. Anhydrous hydrazine decomposition initially leads to nitrogen and ammonia formations and then ammonia is further decomposed into nitrogen and hydrogen. The first decomposition is a structure insensitive reaction, whereas ammonia decomposition is a structure sensitive reaction, where the large iridium particles are more active than the small ones. It is of interest to find a new catalytic system, which can provide a high dispersion of the metal along with a macro and mesoporous network in order to increase the active sites accessibility. Support with higher thermal conductivity is also needed in order to decrease hot spot formation. Carbon nanofiber composite impregnated with 30 wt. % of

216

S. K. Nataraj, B. H. Kim and K. S. Yang

iridium was successfully tested in the catalytic decomposition reaction of hydrazine in a micro-pilot plant under laboratory conditions as shown in SEM picture in Figure 28.

Figure 27. Nanofiber reinforced composite materials to improve the matrix dependent properties of automobile parts, gas storage tanks and of carbon/epoxy parts.

8.6. Nannofibers as Filtration Membranes Indoor air contaminants, such as volatile organic compounds (VOCs), microorganisms, allergens, and other pollutants (e.g., tobacco smoke) pose serious health and productivityrelated problems for occupants of indoor spaces. These toxic compounds are complex mixtures of particles, 90% of which are smaller than 1 μm in diameter. These particles have hundreds of chemicals adsorbed onto their surfaces, including many known or suspected mutagens and carcinogens. Gaseous pollutants contain many irritants, toxic chemicals and nitrogen oxides, which are ozone precursors, and can have negative environmental impacts. The minute size and the abundance of these toxins give them a greater opportunity to enter our bodies via air and water. As a result, the filter industry is looking for new filter media that can create effective barriers for particles smaller than 3 μm and adsorb pollutant gases. Recently, PAN-based carbon hollow carbon fiber membranes [112] were development and successfully tested for their liquid separation performances. Various factors influencing electrospun nonwoven fibrous membrane structure and transport properties were discussed. PAN membranes were pyrolyzed at temperature ranging from 500oC to 800oC for 30 minutes of thermal soak time. Pyrolysis temperature was found to significantly change the structure and properties of nanofibrous carbon membrane. Electrospun layers exhibit minimal impedance to moisture vapor diffusion required for evaporative cooling. Effects of membrane distortion on the transport behavior of the structure might be significant. Experiments and

Carbon Nano-Fibers and their Applications

217

theoretical calculations showed electrospun fiber mats are extremely efficient at trapping airborne particles. High filtration efficiency is a direct result of the submicron-size fibers generated by electrospinning process [113].

Figure 28. SEM Photograph and TEM image of a nanofiber covered by iridium particles and nanofiber composite rocket ablative materials (NRAMs).

8.7. Nanofiber Clothings Carbon nanofiber can now be spun like yarn to make an amazing variety of new inventions possible, including lightweight bulletproof uniforms (see Figure 30). Carbon fiber and carbon fiber cloth consist of bulk, chopped fibers, continuous strands or woven cloth forms of carbon or graphite. Carbon and graphite are used in reinforcing composites as well as other specialized electrical and thermal applications. Carbon fiber is made by charring synthetic polymer fibers made of PAN, by using an oxidation or thermal process. Carbon monofilament is composed of many long thin sheets of carbon molecules. The filaments are then spun into thread and woven into cloth. Carbon fiber and carbon fiber cloth is used in woven cloth including cloths and tapes for marine and composite specialty fabrics combined with polyester resins. Earlier Kevlar was well-known brand of carbon fiber fabric those are still using in bulletproof vests and other ballistic products.

218

S. K. Nataraj, B. H. Kim and K. S. Yang

Figure 30. Some of commercially available nanofiber yarns used different protecting clothings [114].

Carbon fiber and carbon fiber cloth in combination with a resin is used to create a layer of water-proofing over the wood used in boat building. Carbon fiber cloth is also used to make surfacing veils and mats, automotive parts, and sealing tapes. Other non-polymer matrix materials such as carbon fiber-reinforced graphite are used in high temperature applications, including the nose cone and leading wing edges of the space shuttle. Among the many future possibilities includes; soft protective vests stronger than Kevlar, bandages that can contract to put pressure on, artificial muscles powered by electricity those expected much lighter than current hydraulics, would make it easier to incorporate electronic sensors and actuators into clothing. All of these possible applications derive from the remarkable properties of carbon nanofiber, the ability to conduct both heat and electricity along with the extreme toughness of the fiber. The nanofibers versatile structure allows it to be used for a variety of tasks in and around the body. Although often seen especially in cancer related incidents, the CNFs are often used as a vessel for transporting drugs into the body. The nanofibers allow for the drug dosage to hopefully be lowered by localizing its distribution, as well as significantly cut costs to pharmaceutical companies and their consumers. The nanocarbons commonly carries the drug one of two ways: the drug can be attached to the side or trailed behind or the drug can actually be placed inside the nanotube. Both of these methods are effective for the delivery and distribution of drugs inside of the body.

Carbon Nano-Fibers and their Applications

219

9. RECYCLING OF CARBON MATERIALS While carbon fibers continue to gain public attention through the marketing efforts of aircraft companies, such as Airbus and Boeing, as well as Formula 1 racing teams and original equipment manufacturers (OEMs) of high-end consumer goods. As a result, there is a current shortage, which is being remedied by the world's fiber manufacturers, who are building more production lines to meet the new demand. The current world capacity is said to be on the order of 25,000 metric tones (about 55 million lb) per annum. As the new capacity comes on line over the next few years, worldwide capacity could reach 35,000 metric tones (about 77 million lb) per annum. It could reasonably be assumed that most of this carbon fiber will find its way into composites of one form or another and, if one were to take the waste figures being bandied about - anything up to 40 percent - the math is easy. If all the waste were recycled and reintroduced to the market, it could go a long way toward alleviating shortages. Electrospun fiber coatings produce exceptionally light weight multifunctional membranes for protective clothing applications, which exhibit high breathability, elasticity and filtration efficiency. The high specific surface area of ultra fine fibers can be used as high performance filters, scaffolds in tissue engineering, sensors and nevertheless composites applications. High porosity, interconnectivity, microscale interstitial space, and a large surface-to-volume ratio mean that nonwoven electrospun nanofiber meshes are the excellent materials for membrane preparation, especially in biotechnology and environmental engineering applications. Methods exist today by which carbon fibers and prepregs can be recycled, and the resulting recyclate retains up to 90 percent of the fibers' mechanical properties. In the case of recycled carbon fiber composites, this is far from accuracy. In some cases, the method enhances the electrical properties of the recyclate because the carbon recyclate can deliver performance near to or superior to virgin material. All that remains is to create demand for recycled fiber by packaging it in a form useful reproduct to end-users. To make such recycled products viable, of course, there must be an infrastructure established for collecting and identifying the waste and the end-user must have confidence in product quality. This has been done successfully in the plastics industry - it was not long ago that some form of identification was required on plastic parts to facilitate collection and segregation. Most composites manufacturers are already engaged in waste management procedures, and could be counted on to recycle waste materials on the shop floor when collection and processing links are in place.

10. FUTURE PROSPECTS Electrospun and vapor grown nanofiber materials are finding their application in the worldwide market covering number of areas. The nanotechnologies would certainly offer a great chance to gain novel physical and chemical properties of the fiber-based materials. New classes of conductive PAN-based carbon-clay, nano-composites, etc., are useful for applications in diverse electrochemical devices, such as lithium batteries, sensors, electrocatalysts, conducting nanowires and as atomic force microscopy tips. Nanofiber applications for ballistic and chem-bio-protection are being actively investigated and the

220

S. K. Nataraj, B. H. Kim and K. S. Yang

future research on polymer-metal oxide in nanoscale composite structures are expected to be of great utility in a number of applications, including sensors, photovoltaic cells and as catalytic surfaces. High porosity, interconnectivity, microscale interstitial space, and a large surface-to-volume ratio mean that nonwoven electrospun nanofiber meshes are the excellent materials for membrane preparation, especially in biotechnology and environmental engineering applications. Today, a number of fibers are available with tensile strength ten times stronger than metal fibers value. Although carbon-fiber processing from PAN and pitch appears to have matured, PAN/CNFs composite fibers have been processed, exhibiting improved tensile modulus and strength and reduced thermal shrinkage when compared to the control PAN fibers. These PAN/CNF composite fibers are good candidates for the development of next generation carbon fibers with improved tensile strength and modulus while retaining compressive strength. Carbon-fiber tensile strength increases with decreasing diameter. Therefore, it is expected that the small-diameter electrospun PAN fibers (~100 nm) may exhibit significantly higher tensile strength values than achieved so far. There is a very large quantity of academic research being done on the product of a wide range of vapor grown and electrospun materials. These two techniques of fiber preparation are more likely to become a valuable technique for producing fibers with either better performance or with functionality normally difficult/impossible to obtain or tailor using traditional methods, bringing growth in new markets or in existing high performance materials and textile markets.

REFERENCES [1] [2] [3] [4] [5] [6]

[7] [8] [9] [10] [11] [12] [13]

S. Iijima, Nature 354, 1991, 56. T.W. Odom, J.L. Huang, P. Kim, C.M. Lieber, J. Phys. Chem. B 104,2000, 2794. C.N. Rao, B.C. Satishkumar, A. Govindaraj, M. Nath, Chem. Phys. Chem. 2, 2001, 78. R.H. Baughman, A. Zakhidov, W.A. De Heer, Science 297, 2002, 787. N.M. Rodriguez, M.S. Kim, R.T.K. Baker, J. Phys. Chem. 98, 1994, 13108. T.E. McKnight, A.V. Melechko, M.A. Guillorn, V.I. Merkulov, M.J. Doktycz, C.T. Culbertson, S.C. Jacobson, D.H. Lowndes, M.L. Simpson, J. Phys. Chem. B 107, 2003, 10722. T.E. McKnight, A.V. Melechko, D.K. Hensley, D.G.J. Mann, G.D. Griffin, M.L. Simpson, Nano Lett. 4, 2004, 1213. S.E. Baker, K.Y. Tse, E. Hindin, B.M. Nichols, R. Hamers, J. Chem. Mater. 17, 2005, 4971. M. Rzepka, E. Bauer, G. Reichenauer, T. Schliermann, B. Bernhardt, K. Bohmhammel, E. Henneberg, U. Knoll, H.E. Maneck, W. Braue, J. Phys. Chem. B 109, 2005, 14979. V.G. Gavalas, N.A. Chaniotakis, Anal. Chim. Acta 409, 2000, 131. Z. Chen, R.B. King, Chem. Rev. 105, 2005, 3613. M. Capaccio, V.G. Gavalas, M.S. Meier, J.E. Anthony, L.G. Bachas, Bioconjugate Chem. 16, 2005, 241. C.H. Sun, G.Q. Lu, H.M. Cheng, J. Phys. Chem. B 110, 2006, 218.

Carbon Nano-Fibers and their Applications

221

[14] E. Bekyarova, K. Kaneko, M. Yudasaka, K. Murata, D. Kasuya, S. Iijima, Adv. Mater. 14, 2002, 973. [15] J. Fan, M. Yudasaka, J. Miyawaki, K. Ajima, K. Murata, S. Iijima, J. Phys. Chem. B 110, 2006, 1587. [16] S. Sotiropoulou, V. Gavalas, V. Vamvakaki, N.A. Chaniotakis, Biosens. Bioelectron. 18, 2003, 211. [17] [18] W. J. Morton, U.S. Patent, No. 705, 1902, 691. [19] Y.M. KIM, NanoTechnics, patent 10-2004-0088578, 2002. [20] D. Li and Y. Xia, Nano Letters, 4, 2004, 933. [21] M. Bognitzki, M. Czado, T. Frese, A. Schaper, M. Hellwig, M. Steinhart, A. Greiner, and J. H. Wendorff, Advanced Materials, 13, 2001, 70. [22] J. H. Yu, S. V. Fridrikh, and G. C. Rutledge, Advanced Materials, 16, 2004, 1562. [23] D. Li, Y. Wang, and Y. Xia, Advanced Materials, 16, 2004, 361. [24] G. Taylor, Proceedings of the Royal Society of London, Series A, 280, 1964, 383. [25] C. Huang, S. Chen, C. Lai, D. H. Reneker, H. Qiu, Y. Ye, and H. Hou, Nanotechnology, 17, 2006, 1558. [26] M. G. Mckee, G. L. Wilkes, R. H. Colby, and T. E. Long, Macromolecules, 37, 2004, 1760. [27] C. M. Uppatham, M. Nithitanakul, and P. Supaphol, Macromolecular Chemistry and Physics, 205, 2004, 2327. [28] T. Wang, S. Kumar Electrospinning of polyacrylonitrile nanofibers. J. Appl. Poly. Sci. 102, 2006, 1023–1029. [29] L. Yao, T. W. Haas, A. G. Elie, G. L. Bowlin, D. G. Simpson, G. E. Wnek, Chem. Mater., 15, 2003, 1860. [30] R.W. Moncrieff, A Halsted. Man-Made fibers. Press Book: John Wiley and Sons Inc 1975, 797-799. [31] X. Yang, C. Shao, R. Liu, H. Mu, Thin Solid Films 478, 2005, 228. [32] Y.M Shin, M.M. Hohman, M.P. Brenner, G.C. Rutledge, Polymer 42, 2001, 09955. [33] C. Wang S.H. Chien, C.H. Hsu, Y.C. Wang, C.T. Wang, H.A. Lu, Macromolecules 40, 2007, 7973. [34] S. Megelski, J.S. Stephens, D.B. Chase, J.F. Rabolt, Macromolecules 35(22), 2002, 8456. [35] K.H. Lee, H.Y. Kim, H.J. Bang, Y.H. Jung, S.G. Lee, Polymer 44, 2003, 4029. [36] L. Wannatong, A. Sirivat, P. Supaphol, Polym. Int., 53, 2004, 1851. [37] T. Jarusuwannapoom, W. Hongroijanawiwat, S. Jitjaicham, L. Wannatong, M. Nithitanakul, C. Pattamaprom, Eur. Polym. J. 41(3), 2005, 409. [38] C. Pattamaprom, W. Hongrojjanawiwat, R. Rangkupan, Macromol. Mater. Eng. 291, 2006, 840. [39] J. Macossay, A. Marruffo, R. Rincon, T.Eubanks and A. Kuang; Polym. Adv. Technol. 18, 2007, 180. [40] W. Zhang, E. Yan, Z. Huang, C. Wang, Y. Xin, Q. Zhao, Y. Tong; European Polymer. Journal 43, 2007, 802. [41] X.M. Moa, C.Y. Xub, M. Kotakib, S. Ramakrishna, Biomaterials 25, 2004, 1883. [42] D.H. Reneker, A.L. Yarin, H. Fong, S. Koombhongse, J. App. Phy. 87, 2000, 4531.

 

222

S. K. Nataraj, B. H. Kim and K. S. Yang

[43] J.M. Deitzel, W. Kosik, S.H. McKnight, N.C.B. Tan, J.M. DeSimone, S. Crette, Polymer, 43, 2002, 1025-1029. [44] D. Li and Y. Xia, Adv. Mater. 16, 2004, 1151. [45] W. Zhang, Y. Wang, C.-F. Sun, J. Polym. Res. 14, 2007 467. [46] P. Bajaj, D.K. Paliwal, A.K. Gupta, J. Appl. Polym. Sci. 49, 1993, 823. [47] K, Hideto, T, Kohji, Polym. J. 29, 1997, 557. [48] W.X. Zhang, J. Liu, G. Wu, Carbon 41, 2003, 2805. [49] S. Dalton, F. Heatley, P.M. Budd, Polymer 40, 1999, 5531. [50] P.J. Walsh, Composites-ASM Handbook. Ohio: 2001, 49–53. [51] R.C. Bansal, J.B. Donnet; Polymer Reactions. Oxford: Pergamon Press; 1990, 501–520. [52] D.D. Edie, Carbon 36, 1998, 345–362. [53] M. Endo, CHEMTECT, 1988, 568. [54] M. Endo, T. Koyama, and Y. Hishiyama, Jpn. J. Apl. Phys., 15, 1976, 2073. [55] Oberlin, M. Endo, T. Koyama, Carbon, 14, 1976, 133. [56] G.G Tibbetts, C.P. Beetz Jr., J Phys D: Appl. Phys., 20,1987, 292. [57] G.G. Tibbetts, M. Endo, C.P. Beetz Jr., SAMPE J. 22, 1986, 30. [58] M.S. Dresselhaus, G. Dresshaus, K. Sugihara, I.L. Spain, H.A Goldberg, Graphite fibers and filaments. New York: Springer-Verlag, 1988. [59] G.G. Tibbetts, G.L. Doll, D.W. Gorkiewicz, J.J. Moleski, T.A. Perry, C.J. Dasch, Carbon 31, 1993, 1037. [60] G.G. Tibbetts, J Cryst Growth, 66, 1984, 632. [61] F. Benissad, P. Gadelle, M. Coulon, L. Bonnetain, Carbon 26, 1988, 425. [62] Madronero, E. Ariza, M.W.V. Brandl, C. Barda, J. Mater. Sci., 31, 996, 6189. [63] G. Zheng, K. Kouda, H. Sano, Y. Uchiyama, Y. Shi, H. Quan, Carbon 42, 2004, 635. [64] H. Dai, M.S. Dresselhaus, J. Dresselhaus, P. Avouris, Carbon Nanotubes; Springer: Berlin, 2001. [65] G.G. Tibbetts, Carbon 27, 1989, 745. [66] G.G. Tibbetts, Figueiredo, R.T.K. Baker, K.J. Hiittinger, Eds.), Kluwer, Dordrecht, 1990, 87. [67] L.S. Lobo, D.L. Trimm, J.L. Figueiredo, in Proc. Int. Congress On Cafulysis, (J.W. Hightower, Ed)vol. 2, , North-Holland, Amsterdam, 1972, 1125. [68] R.T.K. Baker, M.A. Barber, P.S. Harris, F.S. Featcs, R.J. Waite, I. Carul, 26, 1972, 51. [69] F. Benad, P. Gadelle, M. Coulon, L. Bonnetain, Curbon 26, 1988, 425. [70] F. Benissad, P. Gadelle, Carbon 31, 1993, 21. [71] F. Benad, P. Gadelle, M. Coulon, L. Bonnetain, Carbon 26, 1988, 425. [72] F. Benissad, P. Gadelle, Carbon 31, 1993, 21. [73] G.G. Tibbetts, CA. Bernardo, D.W. Gorkiewic, R.L. Alig, Carbon, 32, 1994, 569. [74] Madrohero, J. Muter. ki. 30, 1995, 2061. [75] T.H. Ko, H.Y. Ting, C.H. Lin, J. Appl. Poly. Sci., 35, 1988, 631. [76] T.H. Ko, J. Appl. Poly. Sci., 43, 1991, 589. [77] E. Frackowiak, F. Beguin, Carbon, 39,2001, 937. [78] B.E. Conway, Electrochemical supercapacitors: scientific fundamentals and technological applications. New York: Kluwer Academic/Plenum Publishers, 1999. [79] M.J. Bleda-Martınez, J.A. Macia-Agullo, D. Lozano-Castello, E. Morallon, D. CazorlaAmoros, A. Linares-Solano, Carbon 43, 2005, 2677.

Carbon Nano-Fibers and their Applications

223

[80] Y.W. Ju, G.R. Choi, H.R. Jung, C. Kim, K.S. Yang, W.J. Lee; J. Electrochem. Soc., 154, 2007, A192. [81] C. Kim, B.T.N. Ngoc, K.S. Yang, M. Kojima, Y.A. Kim, Y.J. Kim, M. Endo, S.C. Yang. Adv. Mater., 19, 2007, 2341. [82] C. Kim, Y.H. Joo, J.H. Kim, W.J. Lee, K.S. Yang, WSEAS Transactions on Systems, 5, 2006, 1971. [83] C. Kim, K.S. Yang, M. Kojima, K. Yoshida, Y.J. Kim, Y.A. Kim, M. Endo, Adv. Funct. Mater. 16, 2006, 2393–2397 [84] F.E. Lynch, J Less Common Metals, 172, 1991, 943. [85] J. S. Matton, H.B. Mark Jr; Activated Carbons, Dekker, New York. [86] J.S. Noh, R.K. Agarwal, J.A Schwarz, Int. J. Hyroden Energy, 12, 1987, 693. [87] B.R. Puri, P.L. Walker Jr., Chemistry and Physics of Carbon; Dekker, New York, 6, 191. [88] K. Jurewicz, E. Frackowiak, F. Beguin, Electrochem Solid State Lett, 4, 2001, A27. [89] E. Frackowiak, F. Beguin, Carbon 40, 2002, 1775. [90] F. Beguin, M. Friebe, K. Jurewicz, C. Vix-Guterl, J. Dentzer, E. Frackowiak, Carbon, 44, 2006, 2392. [91] F. Beguin, K. Kierzek, M. Friebe, A. Jankowska, J. Machnikowski, K. Jurewicz, Electrochim Acta, 51, 2006, 2161. [92] K. Jurewicz, E. Frackowiak, F. Beguin, Fuel Process Technol., 77, 2002, 415. [93] K. Jurewicz, E. Frackowiak, F. Beguin, Appl. Phys. A–Mater. Sci. Process, 78, 2004, 981. [94] C. Vix-Guterl, E. Frackowiak, K. Jurewicz, M. Friebe, J. Parmentier, F. Beguin, Carbon, 43, 2005, 1293. [95] M.M. Hernandez, M.E. Rincon, J. Solid State Electrochem, 9, 2005, 646. [96] C.Z. Wu, X. Zhu, L.L. Ye, C.Z. Ouyang, S.Q. Hu, L.Y. Lei, Inorganic Chem., 45, 2006, 8543. [97] D.K. Kim, S.H. Park, B.C. Kim, B.D. Chin, S.M. Jo, D.Y. Kim, Macromol. Res., 13, 2005, 521. [98] G.S. Chung, S.M. Jo, B.C. Kim, J. Appl .Polym. Sci., 97, 2005, 165. [99] S.H. Park, S.M. Jo, D.Y. Kim, W.S. Lee, B.C. Kim, Syn. Met., 150, 2005, 265. [100] J.S. Im, S.J. Park, T.J. Kim, Y.H. Kim, Y.S. Lee, J. Coll. Inter. Sci., 318, 2008, 42. [101] C. Kim, S.H. Park, W.J. Lee, K.S. Yang, Electrochimica Acta, 50, 2004, 877. [102] W.G. Shim, C. Kim, J.W. Lee, J.J. Yun, Y.I. Jeong, H. Moon, K.S. Yang, J. Appl. Poly. Sci., 102,2006, 2454. [103] C.U. Pittman Jr, W. Jiang, Z.R. Yue, C.A. Leon, Carbon, 37,1999, 85. [104] J. H. Park, Y.W. Ju, S. H. Park, H.R. Jung, K. S. Yang, W. J. Lee, Journal of Applied Electrochemistry, in press, December 2008. [105] S. Chand, J. Mater .Sci., 35, 2000, 1303. [106] Z.M. Huang, Y.Z. Zhang, M. Kotaki, Comp. Sci. Tech., 63, 2003, 2223. [107] G.G. Chase, D.H. Reneker, Fluid - Part Sep J, 16, 2004, 105. [108] K. Yoon, K. Kim, X. Wang, D. Fang, B. Chu, Polymer, 47, 2006, 2434. [109] D.A. Musale, A. Kumar, J. Appl. Poly. Sci., 77, 2000, 1782. [110] M.V. Chandorilar, P.C. Bhavsar, Indian J. Tech., 21, 1983, 124. [111] V. Ricardo, B.-N. Demétrio, L. M.-Jacques, P.-H. Cuong, Applied catalysis. A, General, 279, 2005, 35.

224

S. K. Nataraj, B. H. Kim and K. S. Yang

[112] S.M. Saufi, A.F. Ismail, Membr Sci .and Tech., 2002, 843. [113] P. Gibson, H. Schreuder-Gibson, D. Rivin, Coll Surf A: Physico chem.. Eng. Aspects, 187-188, 2001, 469. [114] H. Catchpole, Knitting in nanometers: Friday, 19 November 2004: http://www.abc.net.au/science/news/stories/s1244935.htm

In: Nanofibers: Fabrication, Performance, and Applications ISBN: 978-1-60741-947-1 Editor: W. N. Chang ©2009 Nova Science Publishers, Inc.

Chapter 6

CARBON NANOFIBERS AS SENSORS Sharlene A. Lewis and Charles M. Lukehart Department of Chemistry, Vanderbilt University Nashville, Tennessee, USA

ABSTRACT Carbon nanofibers (CNFs) are platelet or conical (herringbone) carbon nanostructures consisting of nested cup-shaped or platelet graphene sheets stacked along the long fiber axis. CNFs typically have diameters on the nanometer scale and lengths on the micrometer scale and possess attractive properties, such as large surface area, high electrical conductivity, and good mechanical strength and thermal stability. As-prepared CNFs have surfaces along the long axis that terminate in C(sp2)-H edge sites that are suitable for chemical functionalization. The surface charge, wettability, dispersibility, and chemical reactivity of CNFs can be altered through chemical and physical modifications of these CNF surface sites. CNFs or surface-functionalized CNFs have been used as sensor media either as pristine nanofibers or as CNF-based composites. Large changes in electrical properties, such as electrical resistance, are observed depending on the presence or absence of gaseous analytes that adsorb, bind, or electrochemically react with the CNF component. CNFs functionalized with biomolecules, such as enzymes or DNA oligomers, have been used as biosensors to detect complexation of specific proteins or complementary DNA oligomers. CNFs also act as mechanical sensors. Deflection along the vertical axis of CNFs by acoustic fields results in the generation of electrical fields that can be detected. This chapter summarizes diverse applications in which CNFs are used as sensor media.

INTRODUCTION Though somewhat overshadowed by the popularity and utility of carbon nanotubes, carbon nanofibers (CNFs) are proving to be useful and versatile materials both as economical alternatives to carbon nanotubes as well as uniquely valuable materials. CNFs are electrically and thermally conductive (electrical conductivity of herringbone CNFs as high as 2.4 x 104 S m-1), [1-3] possess high Young’s modulus and tensile strength (modulus of a single

226

Sharlene A. Lewis and Charles M. Lukehart

herringbone CNF is over 600 GPa), [4] have high surface areas, are easily surfacefunctionalized, and are biocompatible. Carbon nanofibers (CNFs) consist of nested canted (herringbone) or platelet graphene sheets stacked along the long nanofiber axis. CNFs having average diameters of 25-200 nm and lengths on the micrometer scale readily grow as single fibers from surfaces of metal alloy growth catalyst nanoparticles by reformative decomposition of hydrocarbon gases. Examples of CNF formation as powders or arrays with or without surface modification are known. [5-9]

SYNTHESIS OF CARBON NANOFIBERS Carbon nanofibers are synthesized by catalytic chemical vapor deposition (C-CVD) or by catalytic plasma-enhanced chemical vapor deposition (C-PECVD) methods. [10-13] The CCVD process entails adsorption and decomposition of carbon-containing gases, such as ethene, carbon monoxide or methane, on metal catalyst particles. Carbon atoms diffuse through catalyst particles and precipitate on specific facets of the catalyst to grow a single carbon nanofiber. Growth catalyst particle orientation determines graphene plane alignment as well as the degree of growth defects. Nanofiber diameter is approximately equal to that of the catalyst particle. Iron, copper and nickel metals or mixed metal alloys are effective growth catalysts and are most commonly used. Nanofiber growth occurs from substrate-supported growth catalyst particles (supported catalyst) or from growth catalyst particles lifted from a substrate surface by the growing nanofiber (floating catalyst). Use of nanofiber growth catalysts enables CNF formation at moderate temperatures (between 400 and 1000 °C) and at low cost. Nanofiber morphology is controlled by the choice of growth catalyst composition, carbonaceous gas, and growth temperature. [5, 8, 9] In C-PECVD, electron impact is used to activate carbonaceous gas molecules and reduce the activation energy of carbon deposition. C-PECVD is the only technique that enables deterministic growth of CNFs in which the location, alignment, size, shape and structure of each CNF can be controlled. Nonequilibrium plasmas (glow discharge) are typically used. By using C-PECVD, vertically aligned carbon nanofiber (VACNF) arrays can be produced. [5]

SURFACE MODIFICATION OF CARBON NANOFIBERS CNF surfaces along the long fiber axis terminate at carbon (sp2) edge sites that are usually passivated by hydrogen atoms. As-prepared CNFs exhibit poor dispersibility and high chemical inertness, so it is difficult to incorporate these materials into devices or composites. The surface charge, wettability, dispersibility, and chemical reactivity of CNFs, as well as the extent of CNF/matrix interfacial binding within composites, can be altered through chemical or physical modification of CNF surfaces.[5-7] Surface modification significantly enhances the chemical reactivity of CNFs without degradation of the CNF backbone and yields surfacederivatized CNFs with a high surface areal density of functional groups. Physical modification of CNF surfaces usually involves deposition of thin films of inorganic or polymer coatings onto the surface of the CNF either during or after synthesis.

Carbon Nanofibers as Sensors

227

Dielectrics, metals, or polymers have been used as coatings. Deposition techniques include atomic layer deposition (ALD), electro- and electroless plating, or CVD. [14-16] CNF edge sites are especially suitable for chemical functionalization. Surface derivatization usually begins with an oxidation step that also removes impurities, such as residual catalyst particles and amorphous carbon, and generates oxygen-containing surface functional groups, such as carboxylic acid, OH groups, or related species. Oxidation is usually accomplished by wet chemical etching in oxidizing acids or by plasma treatment. Acetic acid, oxygen, or carbon dioxide plasmas have been used. Irradiation by UV light initiates reaction of surface C-H groups with alkenes resulting in covalent surface functionalization .[15,17] Surface-oxidized CNFs containing edge sites derivatized with carboxylic acid groups can be further functionalized using carboxyl group coupling chemistry. In one approach, Li et al. convert surface carboxylic acid groups to acid chloride groups and then covalently attach several diamines or triamines to CNF surface sites through amide condensation. More recently, atom-transfer radical polymerization radical initiators, (4-hydroxymethyl)benzyl-2bromopropionate or 2-hydroxyethyl-2′-bromopropionate, have been attached to CNF surface sites to form CNF-polymer brushes with controlled dispersibilities and surface reactivity. [6,7]

CARBON NANOFIBER SENSORS Important characteristics for sensing materials include high sensitivity and selectivity for analytes of interest, good stability, low detection limit, fast response, and ease of fabrication. Carbon nanomaterials have been widely applied as sensing media, because of their high electrical conductivity, biocompatibility, large surface areas and surface reactivity. Although carbon nanotubes (CNTs) have been extensively studied for applications involving electrochemical/biosensing, several challenges complicate use of CNTs as sensors. Surface functionalization of single-walled CNTs (SWNTs) is commonly limited to carbon sites located near the nanotube open ends. This limits the degree of functionalization that can be achieved. In addition, the conductivity of CNTs is dependent on the chirality, diameter, and degree of bundling of CNTs. At present, it is difficult to prepare one specific type of CNT in high purity. CNFs can be produced on the bulk scale in high purity and possess a long-axis surface that can be chemically functionalized. The choice of functional group can be controlled and a wide range of biomolecules, such as proteins and DNA oligomers, have been successfully immobilized on CNF surfaces. CNF sensor media usually exhibit higher sensitivity than media based on other carbon materials.[18,19]

GAS SENSORS Sensing gas molecules is important in environmental monitoring of medical and agricultural operations, industrial processes, space missions, and for control of air quality in green houses, mines, and houses. [20-22] Current gas sensors commonly use carbon-

228

Sharlene A. Lewis and Charles M. Lukehart

black/polymer composites, semiconducting metal oxides, or organic materials as sensing media. The effectiveness of these materials as sensor media is limited by high operating temperatures, poor sensitivity, and/or high resistivity.[23] Carbon nanofibers, surface-derivatized CNFs, and CNF-polymer composites have all been successfully used in gas sensing applications usually through detection of change in electrical resistance (see Table 1). Analyte adsorption to CNF surfaces or swelling of CNF/polymer composites occur usually with measurable increase in electrical resistance (R). Sensor response to analyte gases is frequently calculated by the equation, (Rgas – Rair)/Rair, and reported either as a ratio or as a percentage increase in resistance upon exposure to analyte. Analyte absorption by composite media can also produce nanofiber/adsorbate charge-transfer interactions. Increasing the volume of a sensing material due to analyte swelling inhibits electron conduction and increases electrical resistance. Analyte adsorption leading to CNF/analyte charge transfer processes generates local fields that alter electrical conductivity. [20] Table 1. Carbon Nanofiber (CNF) Sensor Media and Analyte(s) for Gas Sensinga Carbon Nanofiber Sensor Medium CNF/Si (001) CNF/Kapton foil CNF-poly(butyl) acrylate & CNFpoly(acrylic) acid polymer brushes CNF/glass powder CNF/polystyrene paste & CNF/carbon black/polystyrene paste CNF/polypyrrole coaxial nanocables a

Analyte(s) methane He, Ar, air, N2, N2O, O2, C2H2 Acetone, CHCl3, MeOH, toluene, THF, TEA, DMF, NH3 NH3, H2S, toluene THF, benzene, toluene, cyclohexane, hexane, CCl4, ethyl acetate, Et2O, acetone, MeOH, EtOH, propanol, isopropanol NH3, HCl

Reference 21 24 25 26 23, 27

28

Abbreviations: tetrahydrofuran (THF); triethylamine (TEA); dimethylformamide (DMF); diethyl ether (Et2O).

Roy et al. reported use of carbon nanofiber or nanotube films grown on Si(001) wafers as sensor media for detection of various concentrations of methane at room temperature. [21] Electrical contacts are attached at the two ends of the CNF film When a CNF film is exposed to methane, an increase in electrical resistivity is measured. It is postulated that methane reacts with CNF surfaces by addition of methyl and hydrogen radicals to C=C π bonds increasing the carbon hybridization state from sp2 to sp3 and decreasing electrical conductivity. Maximum sensor response is observed at a methane concentration of 4000 ppm. At higher concentrations of methane, sensor response decreases due to a decrease in the number of free π bonds on the CNF surface. A maximum sensor response of 62% is reached after ca. 50 s of exposure. Response decreases as the temperature is increased from 300 to 473 K due to increasing conductivity and possible desorption of methane at elevated temperatures. Senor response decreases over repeated cycles of gas exposure. An ionization gas sensor consisting of a CNF film grown on a Kapton plastic foil cathode with a Cu plate anode has been used by Sim et al. to measure the breakdown voltage of He, Ar, air, N2, N2O, O2, or C2H2 gases. [24] The CNF surface number density is 7 x 107 nanofibers/cm2. Analyte gases are indentified by their characteristic breakdown voltages at analyte gas pressures ranging from 20-300 Pa. These sensors operate at low voltage and are very robust lasting for at least a month without incurring noticeable degradation.

Carbon Nanofibers as Sensors

229

CNF-poly(n-butyl acrylate) (CNF-PBA), and CNF-poly(acrylic acid) (CNF-PAA) polymer brushes deposited onto a Pt-wire interdigitated array (IDA) act as gas sensors for ammonia, acetone, chloroform, dimethylsulfoxide, methanol, tetrahydrofuran, toluene, or triethylamine vapor (see Li et al.). [25] As-prepared CNFs and CNFs surface-derivatized with (3,4’-oxydianilinyl)amido groups were tested as controls. Electrical resistance increases upon exposure to analyte vapor. Sensor response ranges over five orders of magnitude with asprepared CNFs having the smallest response and CNF-PAA/ammonia having the greatest response. Sensor response increases with expected enhancement of CNF or CNF-polymer brush/vapor interactions. Sensor response curves of CNF-ODA material are gradual and asymmetric, while those of as-prepared CNFs are more sharp and symmetric. This trend is consistent with slower vapor adsorption/desorption kinetics for a functionalized nanofiber that interacts strongly with analyte vapor and faster adsorption/desorption kinetics for nonfunctionalized surfaces that bind more weakly with analyte gases. Enhanced sensor responses and improved analyte chemoselectivity are observed with sensors made from CNF-PBA or CNF-PAA. Maximum sensor response of CNF-PBA media is ten times greater than that observed for CNF-ODA nanofibers. CNF-PAA sensors exhibit responses and analyte selectivity comparable to those of CNF-PBA for sensing of organic vapors. However, CNF-PAA sensor response to ammonia vapor is about 103 times greater than that observed for CNF-PAA or CNF-PBA sensing of organic vapors. Strong hydrogen bonding between the carboxylic acid groups of PAA polymer chains and ammonia analyte molecules greatly enhances detection. Wang et al. report a CNF-based, thin-film sensor prepared by mixing worm-like CNFs (WCNFs) and glass dust in a 3:2 proportion with terpineol, for the detection of ammonia, H2S, and toluene vapors. [26] Sensor circuit resistance increases upon exposure to vapor. The sensitivity, response, reproducibility and restoration characteristics of the sensor were determined at atmospheric pressure and at room temperature. Ammonia is detected at low concentrations (0.175-0.35 mg/m3) with a response time of 0.05 s, a restoration time of 1 min, and with good reproducibility. At vapor concentration of 0.35 mg/m3, sensor sensitivity to ammonia is 12 times greater than that measured for H2S and ca 30 times greater than that measured for toluene vapor. Gas sensors based on CNF/polystyrene or CNF/carbon black(CB)/polystyrene(PS) composites have been studied by Zhang et al. [23,27] Using CNFs as an additive enhances the electrical conductivity of polystyrene sensor media. CNF/PS sensors exhibit good gas sensitivity and stability in the presence of analyte vapors having similar polarity or solubility parameters as the matrix. The sensitivity of these composite sensors increases with increasing CNF content or temperature. Maximum sensor response ranges over four orders of magnitude with detection of THF vapor having the greatest response and detection of isopropanol having the weakest response. CB/polymer composites exhibit good sensitivity and fast response times when used as gas sensing media but the gas sensitivity of these materials is often times not very stable. However, CNF/CB/PS composites have greatly enhanced conductivity and stability as gas sensor media. Gas detection sensitivity of these composites is affected by the CNF/CB mass ratio, the total weight percent of the hybrid filler material, and the thickness of the composite film. The CNF/CB/PS composite exhibits higher gas sensitivity than a CB/PS composite sensor for those gas vapors, such as benzene, chloroform and acetone, which have similar polarities to or solubilities of the PS matrix. CNF/CB/PS composite sensors exhibit poor

230

Sharlene A. Lewis and Charles M. Lukehart

sensitivity to methanol or ethanol vapor due to their high polarities and hydrogen-bonding properties. Toxic gas sensing of ammonia or HCl vapors by coaxial polypyrrole-coated CNF (PPy/CNF) nanocables has been demonstrated (see Jang et al.). [28] Electrical conductivity of PPy/CNF compositse is greater than that of pristine CNFs. In the presence of ammonia or HCl vapors, the electrical resistance of PPy/CNF sensor media increases over 3-fold (for NH3) or 6-fold (for HCl) with analyte vapor pressures ranging from 10-50 ppm. For either analyte, the detection limit is 10 ppm. Sensor response is reversible and reproducible and is affected by the thickness of the polypyrrole layer (maximum response at a PPy coating thickness of 22 nm).

BIOSENSORS CNFs are attractive media for biosensing applications due to their biocompatibility, high electrical conductivity, high surface area, and ease of surface modification (see Table 2). Functional biomolecules, such as proteins, enzymes, and DNA oligomers, have been successfully attached to CNF surface sites and used for analyte sensing or sequestration. Table 2. Carbon Nanofiber (CNF) Sensor Media and Analyte(s) for Bio-Sensinga Carbon Nanofiber Sensor Medium VACNF

a

VACNF-NH2 functionalized VACNF-biotinylated VACNF-undecylenic acid/cytochrome c CNF/GOx Silica/oxidized CNF/poly(L-lysine)/Dm. AChE & with silica shield CNF/mineral oil CNF on graphite powder/mineral oil CNF/mineral oil/GOx/silicone oil CNF/GOx/GCE CNF-(CA125)(thionine)/GCE CNF-chitosan/GCE Poly(thionine)-CNF/alcohol oxidase Oxidized CNF/GCE CNF/alcohol dehydrogenase/GCE CNF-FeTMPyP/alcohol oxidase/GCE/PVA CNF/Nafion/GCE CNF/GCE CNF-hemoglobin/Nafion/GCE

Analyte(s) Neuroelectrical activity in hippocampal tissue Complementary DNA oligomer Avidin complexation ABTS glucose acetylcholinesterase

Reference 29 30 17 31 18 32, 33

Dopamine, ascorbic acid, uric acid NADH Glucose, calf thymus DNA glucose HRP-labeled CA125 antibody K562 cells EtOH H2O2, NADH EtOH EtOH H2O2 H2O2, NADH H2O2

34 35 37 38 39 40 41 42, 43 43 44 45 19, 46 47

CNF or oxidized CNF/GCE or Au

NADH

48

Abbreviations: vertically aligned carbon nanofiber array (VACNF); 2,2’-azino-di-[3ehtylbenzthiazoline sulphonate] (ABTS); glucose oxidase (GOx); Drosophila melanogaster acetylcholinesterase (Dm. AChE); dihydronicotinamide adenine dinucleotide (NADH); glassy carbon electrode (GCE); carcinoma antigen 125 (CA125); horseradish peroxidase (HRP); Fe(III) meso-tetrakis(N-methylpyridinum-4-yl)porphyrin (FeTMPyP); poly(vinyl) alcohol (PVA)

Carbon Nanofibers as Sensors

231

Vertically aligned carbon nanofibers (VACNF) ultramicroelectrode arrays (MEAs) have been used to stimulate and measure neuroelectrical activity of hippocampal (brain) tissue (see Yu et. al.). [29] VACNF MEAs perform a standard gamut of electrophysiological tests normally conducted with conventional MEAs made from metal but with a lower noise level. VACNF MEAs stimulate sample tissue without incurring collateral damage to either the tissue sample or the electrode. These arrays can be functionalized with proteins to enhance neuronal interfacing and can be used for direct neurochemical sensing through cyclic voltammetry or amperometry. Surface-derivatized VACNFs have also been used for various biosensing applications by the Hamers group. [17, 30, 31] VACNFs surface-modified with reactive amino groups using two chemical methods have been coupled to DNA oligomers for complementary DNA complexation. [17, 30] Both methods result in VACNFs having a similar number density of DNA molecules attached to their surfaces. The number density of DNA hybridization of a VACNF surface is seven to eight times greater than that achieved for planar surfaces, such as glassy carbon of similar geometric dimensions. A significant fraction of both sidewall and end sites of VACNFs are functionalized, so that much of the available surface area is accessible to DNA hybridization. However, the available surface area is less than expected due to nanofiber bundling. DNA-functionalized VACNFs exhibit good selectivity, chemical stability, and contain a large number of biologically accessible binding sites. Electrochemical reduction of VACNF arrays surface-modified with nitro-phenyl groups gives addressable domains containing primary amino groups that can be covalently coupled to DNA oligomers. Following exposure to fluorescein-labeled complementary DNA oligomer strands, fluorescence images confirm that only those nanofibers that have been electrochemically reduced exhibit fluorescence. These results demonstrate that electrically addressable nanofibers within VACNF arrays can be selectively functionalized with biomolecules, such as DNA oligomers. Likewise, avidin binds to biotinylated-VACNFs with a number density seven times greater than that observed for biotinylated planar substrates. [17] Such functionalized arrays might prove useful in bioagent sensing or sequestering applications. VACNFs surface-functionalized with undecylenic acid groups bind cytochrome c which is a redox active metalloprotein widely used in protein-modified electrode applications. [31] The electrochemical redox activity of cytochrome c-functionalized VACNFs toward oxidation of ABTS (2,2’-azino-di-[3-ethylbenzthiazoline sulphonate]) in the presence of H2O2 is ca. 10 times greater than that measured with cytochrome c-modified-glassy carbon or gold surfaces. This enhanced activity is expected due to the high surface area of carbon nanofibers, although high capacitive currents give a reduction in signal-to-noise ratio. Several CNF enzyme-based biosensors have been developed by the Chaniotakis group. [10, 24, 25] Vamvakaki et al. fabricated glucose biosensors using CNFs formed at different temperatures and with different surface treatments, and compared the performance of these sensors to those fabricated with SWNTs or graphite powder. [18] Adsorption of glucose oxidase occurs more efficiently on the CNF materials (16-25 units/g) than with CNTs (5.5 units/g). Glucose sensing is determined by measuring the cyclic voltammetric anodic current response derived from H2O2 oxidation. CNF-based sensors exhibit similar electrical resistivity but different chemical properties when formed with differing relative amounts of acidic or basic surface groups. CNF-based sensors display the highest activity for glucose oxidation after 100 hours of operation, while carbon nanotube-based sensors exhibit the

232

Sharlene A. Lewis and Charles M. Lukehart

lowest prolonged activity. Biosensors made from CNFs having the highest surface area and capacity of enzyme binding exhibit excellent reproducibility over a range of glucose concentrations. Amperometric acetylcholinesterase biosensors have been prepared using porous silica media and CNF additive. Poly(L-lysine)-templated sol-gel silica mineralization in the presence of acetylcholinesterase (AChE) and in the presence or absence of oxidized CNFs gives AChE/silica or AChE/CNF/silica composites (see Hatzimarinaki et al.). [32] Immobilization of enzymes within porous silica occurs without any deleterious effects on enzyme activity. By incorporating CNFs into AChE/silica composites, enhanced electrical conductivity of the ceramic phase permits more sensitive amperometric detection of AChE enzyme activity. Upon exposure to acetylthiocholine, AChE produces thiocholine and acetic acid. Oxidation of thiocholine on an electrode surface results in an increase in anodic current. Silica/CNF/AChE biosensors show a linear response to acetylcholine concentrations ranging from 0.04 to 0.4 mM. Sensor response time is between 1 and 3 minutes with a sensitivity of 8.2 µA/mM. Silica/AChE biosensors lacking CNF additive exhibit a linear response to acetylcholine analyte concentrations ranging from 0.04 to 0.08 mM, a longer response time and a much lower sensitivity of 2.5 µA/mM and for the same amount of enzyme. By first immobilizing the AChE enzyme on the surfaces of oxidized CNFs followed by poly(Llysine)-templated silica encapsulation, Vamvakaki et al. [33] found that the ceramic coating protects the enzyme from thermal denaturation and protease degradation. Electrochemical detection of several small-molecule analytes of biological importance using various CNF composite paste electrode media have been reported by the You group. [34, 35, 36] Pristine carbon nanofiber-modified carbon-paste electrodes (CNF-CPE) prepared by mixing polyacrylonitrile (PAN)-derived CNFs into mineral oil permit simultaneous determination of dopamine, ascorbic acid and uric acid. These sensors exhibit low detection limits (0.04 µM, 2 µM, and, 0.2 µM, respectively), good selectivity, and well-separated differential pulse voltammetric peaks for each of the analytes, see Liu et al. [34] Liu et al. report amperometric detection of dihydronicotinamide adenine dinucleotide (NADH) using a similar modified carbon paste electrode. [35] A suspension of pristine CNFs in water is cast onto the surface of a graphite carbon powder paste electrode without using a mediator. Differential pulse voltammetry enables simultaneous detection of NADH and ascorbic acid, which is a common interferent in NADH analysis. CNF-modified carbon-paste electrodes loaded with Pd nanoparticles have been used in simultaneous electrochemical detection of dopamine, uric acid and ascorbic acid, see Huang et al. [36] CNFs decorated with Pd nanoparticles are cast onto the surface of graphite powder/mineral oil paste. This electrode exhibits high electrocatalytic activity towards analyte oxidation by enhancing peak currents and decreasing oxidation overpotentials compared to solely carbon powder/mineral oil paste electrode media. Analyte detection limits are 0.2 µM, 0.7 µM, and 15 µM, respectively. Dopamine can be detected in injectable medicine form, and uric acic can be detected in diluted urine samples. Pruneanu et al. report detection of glucose or calf thymus DNA oxidation using a CNF/mineral oil/GOx/silicon oil or CNF/mineral oil electrode media. [37] A guanine oxidation peak is observed when single stranded DNA adsorbs to the surface of the electrode. No such peak is observed when double stranded DNA is used, as the guanine molecules are hidden within the DNA double helix and are not free to interact with the electrode surface.

Carbon Nanofibers as Sensors

233

For glucose detection, ferrocenecarboxylic acid is present as a mediator additive within the CNF paste. This sensor shows a linear response to glucose for concentrations ranging from 1.7 to 7mM and reaches 95 percent of the maximum steady-state current after approximately 30 s. Several CNF-modified glassy carbon electrodes have been developed by the Ju group for biosensing [38-44]. A CNF-modified glassy carbon electrode (GCE) can be used for amperometric sensing of glucose oxidation. [38] This sensor is prepared by casting a solution of CNFs and glucose oxidase (GOx) onto the surface of a GCE. Nafion is used to stabilize sensor components within a membrane and to improve the anti-interferent properties of the sensor. When GOx oxidizes glucose dissolved oxygen is consumed, electrochemical detection of oxygen levels can be used to monitor glucose oxidase activity. Oxidized CNFs exhibit excellent electrocatalytic activity towards electroreduction of dissolved oxygen. This sensor shows a fast response time, good stability, and good fabrication reproducibility with linear response to glucose concentrations ranging from 10 to 350 µM, a detection limit of 2.5 µM, and a sensitivity of 36.3 nAcm-2 µM-1. Presence of uric acid or ascorbic acid did not interfere. A CNF-based amperometric immunosensor has been developed for separation-free immunoassay of cancer antigen 125 (CA125). [39] This sensor is fabricated by covalent attachment of CA125 and thionine (as mediator) onto an oxidized CNF modified glassy carbon electrode. Exposure to CA125 antibody labeled with horseradish peroxidase (HRP) initiates catalytic oxidation of thionine by H2O2 following antigen/antibody binding. Immunosensing occurs with satisfactory stability, good precision, high sensitivity (detection limit of 1.8 units/mL) and satisfactory reproducibility of electrode fabrication. Immobilization of antigens onto surface-modified CNFs might serve as a general method of amperometric immunosensor fabrication for the detection of other antigens. Immobilization and cytosensing of K562 cells has been achieved using CNF-chitosan nanocomposites. [40] Oxidized CNFs react with the reactive amino and hydroxyl functional groups of chitosan (CS) to form a soluble CNF-doped CS colloidal solution. This material is biocompatible, conductive, and possesses cell immobilization properties. When a portion of a CNF-doped CS colloidal solution is deposited onto a glassy carbon electrode, electrochemical impedance spectroscopy can be used to detect cell adhesion. As cell coverage on the electrode increases, the charge-transfer impedance value increases. Impedance cell sensing occurs with good precision, sensitivity (detection limit of 1000 cells/mL), and fabrication reproducibility. Amperometric sensor detection of ethanol by means of electrochemical polymerization of a thionine-CNF nanocomposite solution in the presence of alcohol oxidase (AOD) has been demonstrated. [41] Formation of a biocompatible poly(thionine)-CNF/AOD composite film on a glassy carbon electrode surface permits the monitoring of the electrocatalytic reduction of dissolved oxygen and, therefore, of oxygen consumption at the poly(thionine)-CNF/AOD film. Incorporation of CNFs results in a polymer film with excellent catalytic activity and provides for a sensitive ethanol biosensor with a low detection limit of 1.7 µM, good stability, and fast response time. Electropolymerized films might serve as useful constructs for biosensor fabrication. Deposition of oxidized CNFs onto the surface of a glassy carbon electrode results in a porous membrane that serves as an electrode in hydrogen peroxide detection. [42] The sensor exhibits good stability, a detection limit of 0.15 µM, and fabrication reproducibility. A low overpotential for hydrogen peroxide detection is observed which is indicative of good

234

Sharlene A. Lewis and Charles M. Lukehart

selectivity and sensitivity. This electrode can also be used for flow injection detection of hydrogen peroxide which is useful for analyte monitoring in continuous processes. Oxidized CNFs have been used as amperometric transducers in biosensors for the detection of NADH or ethanol. [43] Alcohol dehydrogenase (ADH), immobilized on an oxidized CNF electrode surface, serves as an amperometric sensor for presence of ethanol and also exhibits very efficient electrocatalytic sensing of NADH oxidation (detection limit of 0.15 µM) at low overpotentials. This feature is attributed to the large numbers of oxygencontaining groups formed during acid oxidation of CNFs. Ethanol biosensing occurs with fast response time (ca. 10 s), low detection limit (3.0 µM), and good fabrication reproducibility. Ethanol detection has also been achieved using a water-soluble CNF/porphyrin nanocomposite material (CNF-FeTMPyP/GCE/AOx/PVA) as electrode media, where FeTMPyP is Fe(III) meso-tetrakis(N-methylpyridinum-4-yl)porphyrin, AOx is alcohol oxidase, and PVA is poly(vinyl) alcohol. [44] Ethanol oxidation is measured indirectly by monitoring oxygen consumption. This sensor possesses an attractive combination of properties due to the high electrical conductivity of CNFs and the catalytic activity of the porphyrin component towards reduction of dissolved oxygen. The sensor exhibits high sensitivity (detection limit of 1.2 µM; more sensitive than AOx/chitosan sensors) to ethanol detection at low potentials and with rapid response time (<10s; faster than observed with AOx/gelatin electrodes). Several carbon nanomaterials (carbon nanotubes, fullerenes and carbon nanofibers) have been studied as carbon nanomaterial/Nafion/GCE composite biosensors and their suitability as transducers or electron-transfer mediators ([Fe(CN)6]4-/([Fe(CN)6]3- redox couple) by Vamvakaki et al. [45] Even though all the carbon nanomaterials are highly conductive, specific characteristics, such as efficiency of hydrogen peroxide oxidation, heat stability and Faradaic and non-Faradaic current, are different for each material. Consequently, each material may be better suited for specific sensing applications. Carbon nanotube transducers have a very low overpotential for the oxidation of peroxide, although they have a significantly high background current. Carbon nanofibers transducers exhibit high thermal stability and are more suited for highly stable and sensitive biosensor systems. Performance of glassy carbon electrodes modified with platelet or herringbone CNFs or tubular (MWNT-type) microstructures in the amperometric detection of hydrogen peroxide in solution has been compared by Li et al. [19] All three types of CNFs exhibit electrocatalytic activity towards hydrogen peroxide oxidation. Platelet CNF/GCE sensors show the highest sensitivity (3.20 µA/mM H2O2 versus 1.31 µA/mM H2O2 for herringbone CNFs and only 0.74 µA/mM H2O2 for MWNTs) and electrocatalytic activity due to the large surface area, large number of active edge sites, and more numerous favorable sites for electron transfer that is characteristic of this type of microstructure. Herringbone CNFs/GC sensors perform over the largest linear range of analyte concentration (0.18-3.13 mM). Platelet and herringbone CNFs are considered to be most promising materials for biosensor applications. Electrochemical detection of β-nicotinamide adenine dinucleotide (NADH) at the surface of glassy carbon electrodes modified with carbon nanofibers has been reported by Arvinte et al. [46] NADH and its oxidized form are important coenzymes for over 300 dehydrogenase enzymes and biomarker system components. Oxidation of NADH at bare electrodes occurs at high overpotentials and is plagued by problems, such as low sensitivity and the fouling of the electrode surface. CNF-modified electrodes exhibit better electrochemical reactivity towards oxidation of NADH than do unmodified glassy carbon electrodes. This electrode performs

Carbon Nanofibers as Sensors

235

with linear response over an analyte concentration range of 0.0 to 1.0 mM NADH, and analyte concentrations as low as 0.1 mM can be accurately detected. A CNF/hemoglobin sensor for hydrogen peroxide detection that operates under reagentfree and mediator-free conditions has been reported by Lu et al. [47] The working electrode (Hb–Nafion–CNF/GCE) is prepared by casting a suspension of CNFs, Nafion and hemoglobin onto a glassy carbon electrode. Immobilized hemoglobin peptide retains its secondary structure and facilitates direct electron transfer and good bioelectrocatalytic activity towards hydrogen peroxide detection. This sensor exhibits good selectivity and stability, high sensitivity (394 mA/cm2M), low detection limit (0.1 µM), fast response and good fabrication reproducibility. Electrochemical detection of β-nicotinamide adenine dinucleotide (NADH) has been compared using as-prepared or oxidized carbon microparticles and herringbone CNFs as sensor media on either gold or glassy carbon electrodes (see Perez et al). [48] Best results are obtained when using a glassy carbon electrode and untreated CNFs. For such an electrode, peak current (voltage) for NADH detection for CNF or carbon microparticle electrode media are 84.7 µA (+0.352 V) and 41.0 µA (+0.538 V), respectively. CNF-modified electrodes exhibit lower overpotentials and higher oxidation currents for detection of NADH compared to carbon microparticle electrodes.

MECHANICAL SENSORS Although the development of CNF-based gas sensors and biosensors is currently a major thrust in CNF-based sensor research, CNFs have been used recently in other sensing applications. Carbon nanofibers forests have been used as an acoustic sensor to detect nanodeflections generated by an airborne ultrasonic acoustic field. [49] Heterodyne interferometry is used to measure deflections in CNFs, when the sensor medium is exposed to acoustic fields having frequencies of 1 MHz, 500, or 250 kHz. Nanofiber deflection differs for each frequency. This technique, however, is not suitable for the detection of acoustic fields of lower audible frequencies due to the inherent stiffness of CNFs. Park et al. measured load, micro-damage sensitivity and stress-transferance effects of carbon nanotube (CNT), CNF, or carbon black (CB)/epoxy composites using wettability tests, electrical resistance, and acoustic emission measurements. [50] Minimum loadings of CNT, CNF and CB required for the composite to be electrically conductive were 0.5 vol% and above 0.5 and 2.0 vol%, respectively. Micro-damage sensitivities of 2 vol% CNT, 2 vol% CNF and 7 vol% CB/epoxy composites are detected simultaneously from electrical resistance and acoustic emission (AE) measurements. Electrical resistance of the CNF/epoxy composites rapidly increases following nanofiber fracture. Micro-damage sensitivities measured by AE of CNT and CNF/epoxy nanocomposites are greater than that of CB/epoxy composites. CNT/epoxy nanocomposites exhibit the highest sensitivity to fiber fracture, matrix deformation, and fiber tension sensing. CB/epoxy nanocomposites are the least sensitive sensor materials. The extent of micro-damage is similar for CNF and CNT/epoxy nanocomposites at similar loadings of carbon nanomaterial additives. Fabrication of nanoscale bio-switches for detecting chemically induced cleavage of chemical bonds using CNF nanowires has been reported by Li et al. [51] Dielectrophoretic

236

Sharlene A. Lewis and Charles M. Lukehart

manipulations form single-nanowire bridges across gold microelectrode junctions. Nanowires are kept in place by strong coupling interactions between biotinylated-CNF nanowires and avidin molecules. The biotin used contains a disulfide spacer functional group. Chemical reduction of the S-S bond cleaves nanowires from the gold microelectrode with subsequent nanowire removal by fluid flow. Carbon nanowire loss is detected in real time by monitoring changes in AC electrical response due to loss of highly conductive carbon nanowires and capacitive coupling across the nanowire bridges. This detection technique could be extended to other biotinylated macromolecules and to the detection of chemical bond cleavage other than cleavage of disulfide S-S bonds.

CONCLUSION Given the current high level of interest in developing more sensitive and analyte-specific chemical or bio-agent detection methods, carbon nanofiber materials will continue to make a contribution to solving these needs. CNF materials under development for sensor applications include tubular CNF arrays as real-time gas sensors, [52] CNF arrays containing polymerimmobilized Ag/AgCl core-shell nanoparticles as potential electrophysiological sensors, [53] and hierarchical CNF/CNT structures functionalized with metal alloy nanoparticles for chemical sensing. [54] Due to favorable chemical and physical properties and convenient methods of effecting surface derivatization, carbon nanofibers will continue to find applications as sensing media. The significant number of papers already published in this area, especially within the last decade, indicates that many in the scientific community are beginning to appreciate the potential use of CNF materials as sensors. It is expected that research in this area will remain active for quite sometime.

ACKNOWLEDGEMENTS Financial support from the College of Arts and Science of Vanderbilt University is gratefully acknowledged.

REFERENCES [1] [2] [3] [4]

Ngo, Q.; Cassell, A. M.; Austin, A. J.; Li, J.; Krishnan, S.; Meyyappan, M.; Yang, C. Y. IEEE Electron Device Lett. 2006, 27, 221-224. Kimura, T.; Koizumi, H.; Kinoshita, H.; Ichikawa, T. Nucl. Instrum. Methods Phys. Res., Sect. B 2005, 236, 474-481. Kikuchi, I; Maruyama, I.; Fujii, T.; Kobayashi, T.; Kim, S. H.; Kim, D. U.; Lee, S. K. Thin Solid Films 2006, 503, 219-223. Jacobsen, R. L.; Tritt, T. M.; Guth, J. R.; Ehrlich, A. C.; Gillespie, D. J. Carbon 1995, 33, 1217-1221.

Carbon Nanofibers as Sensors [5]

[6] [7] [8] [9] [10] [11] [12] [13] [14] [15]

[16]

[17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33]

237

Melechko, A. V.; Merkulov, V. I.; McKnight, T. E.; Guillorn, M. A.; Klein, K. L.; Lowndes, D. H.; Simpson, M. L. J. Appl. Phys. 2005, 97, 041301/041301041301/041339. Li, J.; Vergne, M. J.; Mowles, E. D.; Zhong, W.-H.; Hercules, D. M.; Lukehart, C. M. Carbon 2005, 43, 2883-2893. Li, L.; Lukehart, C. M. Chem. Mater. 2006, 18, 94-99. Rodriguez, N. M.; Kim, M.-S.; Baker, R. T. K. J. Phys. Chem. 1994, 98, 13108-13111. Rodriguez, N. M.; Chambers, A.; Baker, R. T. K. Langmuir 1995, 11, 3862-3866. Rodriguez, N. M. J. Mater. Res. 1993, 8, 3233-3250. Carneiro, O. C.; Rodriguez, N. M.; Baker, R. T. K. Carbon, 2005, 43, 2389-2396. Rodriguez, N. M.; Kim, M. S.; Fortin, F.; Mochida, I.; Baker, R. T. K. Appl. Catal. A: General 1997, 148, 265-282. Chambers, A.; Rodriguez, N. M.; Baker, R. T. K. J. Mater. Res. 1996, 11, 430-438. Kemell, M.; Pore, V.; Tupala, J.; Ritala, M.; Leskelae, M. Chem. Mater. 2007, 19, 18161820. Klein, K. L.; Melechko, A. V.; McKnight, T. E.; Retterer, S. T.; Rack, P. D.; Fowlkes, J. D.; Joy, D. C.; Simpson, M. L. J. Appl. Phys. 2008, 103, 061301/061301061301/061326. Melechko, A. V.; McKnight, T. E.; Guillorn, M. A.; Austin, D. W.; Ilic, B.; Merkulov, V. I.; Doktycz, M. J.; Lowndes, D. H.; Simpson, M. L. J. Vac. Sci. Technol. B 2002, 20, 2730-2733. Baker, S. E.; Tse, K.-Y.; Hindin, E.; Nichols, B. M.; Clare, T. L.; Hamers, R. J. Chem. Mater. 2005, 17, 4971-4978. Vamvakaki, V.; Tsagaraki, K.; Chaniotakis, N. Anal. Chem. 2006, 78, 5538-5542. Li, Z.; Cui, X.; Zheng, J.; Wang, Q.; Lin, Y. Anal. Chim. Acta 2007, 597, 238-244. Bhoga, S. S.; Singh, K. Ionics 2007, 13, 417-427. Roy, R. K.; Chowdhury, M. P.; Pal, A. K. Vacuum 2005, 77, 223-229. Comini, E.; Faglia, G.; Sberveglieri, G.; Pan, Z.; Wang, Z. L. Appl. Phys. Lett. 2002, 81, 1869-1871. Zhang, B.; Fu, R.; Zhang, M.; Dong, X.; Zhao, B.; Wang, L.; Pittman, C. U. Composites, Part A 2006, 37, 1884-1889. Sim, H. S.; Lau, S. P.; Ang, L. K.; Tanemura, M.; Yamaguchi, K. Diamond Relat. Mater. 2008, 17, 1959-1962. Li, L.; Li, J.; Lukehart, C. M. Sens. Actuators B 2008, B130, 783-788. Wang, J.-Z.; Chen, X.; Li, M.-Q.; Liu, J.-H. Chin. J. Chem. 2008, 26, 649-654. Zhang, B.; Fu, R.; Zhang, M.; Dong, X.; Wang, L.; Pittman, C. U. Mater. Res. Bull. 2006, 41, 553-562. Jang, J.; Bae, J. Sens. Actuators, B 2007, B122, 7-13. Yu, Z.; McKnight, T. E.; Ericson, M. N.; Melechko, A. V.; Simpson, M. L.; Morrison, B., III. Nano Lett. 2007, 7, 2188-2195. Lee, C.-S.; Baker, S. E.; Marcus, M. S.; Yang, W.; Eriksson, M. A.; Hamers, R. J. Nano Lett. 2004, 4, 1713-1716. Baker, S. E.; Colavita, P. E.; Tse, K.-Y.; Hamers, R. J. Chem. Mater. 2006, 18, 44154422. Hatzimarinaki, M.; Vamvakaki, V.; Chaniotakis, N. J. Mater. Chem. 2009, 19, 428-433. Vamvakaki, V.; Hatzimarinaki, M.; Chaniotakis, N. Anal. Chem. 2008, 80, 5970-5975.

238 [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54]

Sharlene A. Lewis and Charles M. Lukehart Liu, Y.; Huang, J.; Hou, H.; You, T. Electrochem. Commun. 2008, 10, 1431-1434. Liu, Y.; Hou, H.; You, T. Electroanalysis 2008, 20, 1708-1713. Huang, J.; Liu, Y.; Hou, H.; You, T. Biosens. Bioelectron. 2008, 24, 632-637. Pruneanu, S.; Ali, Z.; Watson, G.; Hu, S.-Q.; Lupu, D.; Biris, A. R.; Olenic, L.; Mihailescu, G. Part. Sci. Technol. 2006, 24, 311-320. Wu, L.; Zhang, X.; Ju, H. Biosens. Bioelectron. 2007, 23, 479-484. Wu, L.; Yan, F.; Ju, H. J. Immunol. Methods 2007, 322, 12-19. Hao, C.; Ding, L.; Zhang, X.; Ju, H. Anal. Chem. 2007, 79, 4442-4447. Wu, L.; McIntosh, M.; Zhang, X.; Ju, H. Talanta 2007, 74, 387-392. Wu, L.; Zhang, X.; Ju, H. Analyst 2007, 132, 406-408. Wu, L.; Zhang, X.; Ju, H. Anal. Chem. 2007, 79, 453-458. Wu, L.; Lei, J.; Zhang, X.; Ju, H. Biosens. Bioelectron. 2008, 24, 644-649. Vamvakaki, V.; Chaniotakis, N. A. Sens. Actuators B 2007, B126, 193-197. Arvinte, A.; Valentini, F.; Radoi, A.; Arduini, F.; Tamburri, E.; Rotariu, L.; Palleschi, G.; Bala, C. Electroanalysis 2007, 19, 1455-1459. Lu, X.; Zhou, J.; Lu, W.; Liu, Q.; Li, J. Biosens. Bioelectron. 2008, 23, 1236-1243. Perez, B.; del Valle, M.; Alegret, S.; Merkoci, A. Talanta 2007, 74, 398-404. Koukoulas, T.; Theobald, P. D.; Zeqiri, B.; Bu, I. Y.; Milne, W. I. Opt. Lett. 2008, 33, 947-949. Park, J.-M.; Kim, D.-S.; Kim, S.-J.; Kim, P.-G.; Yoon, D.-J.; DeVries, K. L. Composites Part B 2007, 38B, 847-861. Li, B.; Shang, L.; Marcus, M. S.; Clare, T. L.; Perkins, E.; Hamers, R. J. Small 2008, 4, 795-801. Hsieh, C. T.; Chen, J. M.; Kuo, R. R.; Huang, Y. H. Appl. Phys. Lett. 2004, 84, 11861188. Watts, P. C. P.; Henley, S. J.; Mendoza, E.; Silva, S. R. P.; Irvine, J. K.; McAdams, E. T. Nanotechnology 2007, 18, 205502. Zhao, X.; Lee, U. J.; Lee, K. H. J. Phys. Chem. C 2008, 112, 9539-9543.

In: Nanofibers: Fabrication, Performance, and Applications ISBN 978-1-60741-947-1 Editors: W. N. Chang © 2009 Nova Science Publishers, Inc.

Chapter 7

PROCESSING-STRUCTURE RELATIONSHIPS OF ELECTROSPUN NANOFIBERS Xiangwu Zhang1 Fiber and Polymer Science Program, Department of Textile Engineering, Chemistry and Science, North Carolina State University, Raleigh, NC, USA

ABSTRACT Electrospinning is a simple and versatile method for producing nanofibers from various materials including polymers, composites, carbons, ceramics, and metals. One unique and important aspect of electrospinning is its ability to manipulate the structures of nanofibers through careful control of processing parameters, including: i) intrinsic properties of the spinning solution such as rheological behavior, conductivity, surface tension, polymer molecular weight, and solution concentration; and ii) operational conditions such as voltage, solution flow rate, nozzle diameter, spinneret-collector distance, spinneret configuration, and motion of the collector. This chapter addresses the fundamental relationships between processing and structures of electrospun nanofibers and the utilization of such first-principle knowledge to achieve nanofibers with desirable structures. Nanofiber structures that are covered include fiber diameter, primary pore structure, secondary pore structure, and other secondary pore structures. The focus of this chapter is on polymer nanofibers, but electrospun fibers of other materials, such as composites, carbons, ceramics, and metals, are also discussed.

1.

INTRODUCTION

Electrospinning is a simple, non-mechanical technique that is gaining attention due to its capability and feasibility to generate large quantities of nanofibers with well-defined structures at relatively low cost.[1-4] Figure 1 shows a basic setup for electrospinning. A high voltage is applied between a precursor solution contained in a syringe and a metallic collector.

1

Correspondence to: Xiangwu Zhang (Email: [email protected]).

240

Xiangwu Zhang

Figure 1. Schematic (A) and photograph (B) of a basic electrospinning setup.

When the voltage reaches a critical value, electrostatic forces overcome the surface tension of the solution and eject a liquid jet from the nozzle of the spinneret (metallic needle) by creating a so-called Taylor Cone. The electrically-charged jet then undergoes a stretching-andwhipping process resulting in the formation of a long, thin thread. This stretching-andwhipping process is accompanied by the rapid evaporation of the solvent that reduces the jet diameter from hundreds of micrometers to as small as tens of nanometers. The dry fibers are accumulated on the surface of the collector forming a nonwoven porous mat of nanofibers. The process can be adjusted to control the fiber diameter by varying the solution properties and operational conditions.[5] Typical diameter of electrospun fibers ranges from 20 nm to 1 μm,[1] however, the smallest diameter reported is 1.6 nm by Huang et al.[6] for Nylon 4,6 nanofibers. By modifying the electrospinning setup, nanofibers with core-sheath, hollow, or porous structures can be prepared.[1, 3] Historically, electrospinning was largely limited to the fabrication of nanofibers from organic polymers. However, electrospun nanofibers of composite, carbon, ceramic, and metal materials can also been fabricated.[7-9] The electrospinning setup shown in Figure 1 is the most basic configuration. Similar setups have been used by many research groups to conduct fundamental electrospinning research. However, the process of generating nanofibers using this basic setup is relatively slow. Hence, efforts have been made to develop different electrospinning approaches that can product nanofibers at high speeds. For example, the Department of Nonwovens at the Technical University of Liberec (TUL) developed and patented a NanospiderTM technology, which was exclusively licensed to Elmarco for the production and sale of NanospiderTM machines. Instead of creating a Taylor Cone from a needle tip, NanospiderTM uses electrostatic force to generate nanofibers from a thin solution film on a rotating drum, and as a result, NanospiderTM can produce large quantities of nanofibers on an industrial scale. In addition to NanospiderTM, other techniques, such as multiple-needle systems, have also been developed to electrospin nanofibers at large scales. Compared with most nanotechnologies, electrospinning is a relatively low-cost process; and hence it has been used in many industries. For example, Freudenberg Nonwovens has used

Processing-Structure Relationships of Electrospun Nanofibers

241

electrospinning to product low-cost air filters, acoustic nonwovens, and wound pads for over 20 years. Donaldson Company, Inc. has commercialized low-cost electrospun Ultra-Web® filters for dust collection and Spider-Web® filters for gas turbine air filtration. In addition, eSpin Technologies, Inc. is fabricating electrospun carbon nanofibers for thermal insulation and lightweight structural composites. The unique and interesting features associated with electrospun nanofibers make them excellent materials for many fascinating applications, such as gas separation, water purification, catalyst support, sensor, medicine, and energy storage and conversion, among others. The requirements on the structure and properties of nanofibers are different for many of these applications, and hence it is important to understand the fundamental processingstructure relationships of electrospun nanofibers, which can guide the design and fabrication of nanofibers with desired structures and enhanced properties. This chapter provides an overview of recent advances in the fundamental relationships between processing and structures of electrospun nanofibers and the utilization of such first-principle knowledge to achieve nanofibers with controlled structures and desired functions. All the nanofiber images presented in this chapter are obtained in the author’s laboratory in last two years.

2. PROCESSING-STRUCTURE RELATIONSHIPS OF POLYMER NANOFIBERS Polymers are the most common materials used in electrospinning. Synthetic polymers that have been electrospun into nanofibers include polyacrylonitrile,[10-14] polystyrene,[1517] polyamide,[18-21] polyimide,[22] polyacrylic acid,[23] polycarbonate,[24] polyetherimide,[25] polyethylene oxide,[26] polyurethane,[27] polybenzimidazol,[28, 29] polyvinyl chloride,[27] polyvinyl alcohol,[30, 31] polyvinyl carbazole,[24] polyvinyl pyrrolidone,[32] polyvinylidene fluoride,[33, 34] polyethylene terephthalate,[35] polytrimethylene terephthalate,[36] polyvinylidene fluoride-co-hexafluoropropylene,[37] and polystyrene-co-butadiene-co-styrene,[38] etc. In addition, many biodegradable synthetic polymers have also been electrospun into nanofibers, such as degradable polydioxanone,[39] polyglycolide,[40] poly(L-lactic acid),[40] poly(ε-caprolactone),[41, 42] poly(L-lactic acidco-ε-caprolactone),[43] poly(D,L-lactide-co-glycolide),[44] poly(L-lactide-co-glycolide),[45] and poly(L-lactic-co-glycolic acid),[40] etc. In addition to synthetic polymers, natural polymers can be electrospun into nanofibers and most of these electrospinnable natural polymers are proteins and polysaccharides. Proteins that have been electrospun include but not limit to collagen,[46, 47] gelation,[48, 49] fibrinogen,[50] and silk fibroin.[51-53] Examples of polysaccharides that have been electrospun are cellulose,[54, 55] chitosan,[56, 57] and hyaluronan.[58] One focus of current electrospinning research is on the control of the fiber diameter, primary pore structure, and secondary pore structure of electrospun polymer nanofibers through selectively adjusting various processing parameters. The structure control of nanofibers with other secondary structures, such as core-sheath and hollow fibers, are also being studied.

242

Xiangwu Zhang

2.1. Relationship Between Processing and Fiber Diameter One of the most important structure parameters related with electrospun nanofibers is the fiber diameter. Typical diameters of electrospun fibers range from 10 nm to 1 μm,[1] however, the smallest diameter reported is 1.6 nm by Huang et al.[6] for Nylon 4,6 nanofibers. The diameter of electrospun nanofibers are dependent on a number of processing parameters: i) intrinsic properties of the solution such as rheological behavior, conductivity, surface tension, polymer molecular weight, and solution concentration; and ii) operational conditions such as voltage, solution flow rate, nozzle diameter, spinneret-collector distance, spinneret configuration, and motion of the collector.[3, 59-62] However, not all these parameters are fundamental nor are all independent.[61] For example, solution rheology is a function of both polymer molecular weight and solution concentration. In addition, voltage, solution flow rate, and nozzle diameter are interrelated, as are spinneret-collector distance and motion of the collector.

2.1.1. Solution Properties In fiber diameter controlling, the most challenging is the preparation of a suitable polymer solution for electrospinning because the system requires a balance of forces that are controlled by many solution properties such as rheological behavior, conductivity, and surface tension.[7, 60, 63] 2.1.1.1. Rheological Behavior, Conductivity, and Surface Tension Solution rheology has a significant influence on the diameter of electrospun nanofibers;[60, 62, 64] for example, McKee at al.[64] studied electrospun fiber formation of linear and branched polyester and found that fiber diameter (D) increases with the zero shear rate viscosity (η0) of the solution by

D = 0.05η0 0.8

(1)

In addition to viscosity, it is also found that viscoelasticity particularly relaxation time, plays a significant role in electrospinnability of a system and the diameter of the resultant nanofibers.[65-67] The morphology of electrospun nanofibers is also influenced by the net charge density carried by the moving jet, which is determined from the electric current and the mass of dry polymer collected on the metal collector.[60] The electrostatic repulsion from the excess charge in the liquid jet always tries to increase the surface area, which opposes the formation of spheres and results in thin fibers. Net volume charge density is described by[60] t

Net volume charge density = cρ ∫ Idt / m 0

(2)

where I is the jet current, t the collecting time, c the polymer concentration, ρ the solution density, and m the mass of dry polymer. Generally, the net volume charge density is proportional to the conductivity of the polymer solution.[60] Therefore, the solution

Processing-Structure Relationships of Electrospun Nanofibers

243

conductivity can be used as one of the determining parameters to control the diameter of electrospun fibers since it is easy to measure and adjust the solution conductivity. In addition to rheological behavior and conductivity, the solution surface tension also plays an important role in the nanofiber formation, and the final morphology and diameter of electrospun nanofibers strongly depend on the interplay between surface tension, electrostatic repulsion (conductivity), and viscoelastic force (rheological behavior).[68] Typically, when the solution tension is too high, beads tend to be formed during electrospinning. However, the formation of beams can be avoided by achieving a balance among surface tension, electrostatic repulsion, and viscoelastic force.[60] The following are some examples on how to control the fiber diameter by selectively adjusting the solution viscosity, conductivity, and surface tension using different approaches.

2.1.1.2. Polymer Concentration There are many approaches to adjust the properties of electrospinning solutions. For example, the concentration of the polymer solution has direct impact on the solution viscosity, and hence the diameter of the resultant nanofibers. Figure 2 shows SEM images of Nylon 66 nanofibers electrospun from solutions with different Nylon concentrations. All three samples exhibit long and straight fibrous morphology. However, with increase in polymer concentration, the fiber diameter increases due to the increased solution viscosity. 2.1.1.3. Nanoparticle Filler Adding nanoparticle fillers into the solution can also adjust the viscosity, and hence the electrospinnability and the diameter of resultant nanofibers. Generally, adding filler particles increases the viscosity of electrospinning solutions, but the extent differs from system to system and decreases with increasing shear rate. For many composite systems, the viscosity of dispersions can be predicted using the equation

η = 1+ 2.5φ ηs

(3)

where η is the viscosity of the composite, ηs the viscosity of the suspending medium, and φ the filler volume fraction.[69] Therefore, adding filler particles favors nanofiber formation by increasing the viscosity. However, when φ reaches a critical value, filler particles form a continuous network, which obstructs the flow of the dispersions[69] and is detrimental to the nanofiber formation. We have studied the influence of silica nanoparticles on the morphology of electrospun polyacrylonitrile (PAN) nanofibers, and Figure 3 shows SEM images of pure PAN and silica/PAN nanofibers with various silica contents. It is seen that, when the silica content is high (e.g., 2 wt.%), irregularities and nonuniformities (such as beaded segments) appear. In addition, the silica content has influence on the fiber diameter distribution. For example, the diameter distributions of pure PAN nanofibers and composite nanofibers with low silica content (i.e., 1 wt.%) are relatively narrow, but composite nanofibers with high silica content (i.e., 2 wt.%) have wide fiber diameter distribution.

244

Xiangwu Zhang

Figure 2. Relationship between polymer concentration and fiber diameter for electrospun nylon 66 nanofibers. Nylon concentration: (A) 15, (B) 17.5, and (C) 20 % wt/vol.

Processing-Structure Relationships of Electrospun Nanofibers

245

Figure 3. Relationship between silica nanoparticle filler content and fiber diameter for electrospun PAN nanofibers. Filler content: (A) 0 (pure PAN), (B) 1, and (C) 2 wt %.

246

Xiangwu Zhang

If the nanoparticle fillers are made of conducting materials, the solution conductivity will also be changed. Conducting nanoparticles can increase the electronic conductivity and hence increase the spinning current carried by the moving jet. Furthermore, the electronic conductivity also has direct influence on the net charge density that can be built in the liquid jet, resulting in changed fiber diameter.

2.1.1.4. Salt Additive In addition to nanoparticle fillers, adding salts is another means to adjust the solution properties. Figure 4 shows SEM images of PAN/ZnCl2 composite nanofibers electrospun from 7 wt % PAN solutions with different ZnCl2 concentrations (0, 5, 10, and 15 wt. %). It is seen that all the fibers are relatively uniform and randomly oriented, forming interwoven networks. The average diameter of composite nanofibers gradually decreases from 290 to 250, 180, and 130 nm when ZnCl2 concentration increases from 0 to 5, 10, and 15 wt %, respectively. The viscosity, surface tension, and conductivity of the PAN solutions are shown in Figure 5A. With increase in ZnCl2 concentration, the solution viscosity increases (e.g., an increase of 31 % when ZnCl2 concentration increases from 0 to 15 wt %), which can be explained by the formation of interactions among PAN molecules, ZnCl2 salt, and DMF solvent. For example, the Zn2+ ions can form complexes with the oxygens in DMF and/or the nitrogens in PAN, and reduce the mobility of PAN and DMF molecules. As shown in Figure 5A, with increasing ZnCl2 concentration, there is also a slight increase in surface tension (e.g., an increase of 6 % when ZnCl2 concentration increases from 0 to 15 wt %) due to the interactions among PAN, ZnCl2, and DMF. Figure 5A also shows a dramatic influence of ZnCl2 on the solution conductivity, showing an increase from 58.7 µs/cm for pure PAN to 330.2 µs/cm for the 15 wt % ZnCl2 solution since more free ions are available at higher ZnCl2 concentration. This distinct change (an increase of 460 %) in solution conductivity is significantly greater than those of viscosity (31 %) and surface tension (6 %). Therefore, the decrease of nanofiber diameter with increasing ZnCl2 concentration should be attributed predominantly to the increased conductivity of the polymer solutions. The increased conductivity of the polymer solution can cause an increase in net charge density, and hence a larger force of electrostatic repulsion. The massive increase in electrostatic repulsion, compared to the small changes of surface tension and viscosity, tends to increase the surface area, which opposes the formation of spheres and leads to a decrease in fiber diameters. Depending on the electrospinning system studied, the impact of salt additives on the solution properties can be more complicated. For example, iron acetylacetonate (AAI) salt affects the solution properties and hence the fiber diameters differently from ZnCl2. Figure 5B shows the effect of AAI concentration on the viscosity, conductivity, and surface tension of PAN solutions. It is seen that the solution viscosity first increases with increase in AAI concentration, but it decreases after the salt concentration exceeds 14 μmol/g. Like ZnCl2, the change in solution viscosity caused by AAI is also a result of the formation of PAN-salt and DMF-salt complexes in the solution. After the addition of AAI, the ion Fe3+ can form complexes with the oxygens in DMF and/or the nitrogens in PAN, and reduces the mobility of PAN and DMF molecules. This leads to increased solution viscosity, especially at low salt concentrations. However, the formation of DMF–salt complexes reduces the salvation power of DMF for PAN due to a lack of DMF molecules available for interaction with PAN chains.

Processing-Structure Relationships of Electrospun Nanofibers

247

Figure 4. Relationship between ZnCl2 salt concentration and fiber diameter for electrospun PAN nanofibers. ZnCl2 salt concentration: (A) 0 (pure PAN), (B) 5, (C) 10, and (D) 15 wt %.

This causes a decrease in solution viscosity at high AAI concentrations. Similar to viscosity, surface tension also increases with increase in AAI concentration and reaches a maximum when the AAI concentration is 28 μmol/g. Figure 5B also shows that, with increase in AAI concentration, solution conductivity increases, and then remains relatively constant with salt concentrations beyond 14 μmol/g. The conductivity of salt solutions is governed by charge, electrochemical mobility, and concentration of ions. When the AAI concentration is lower than 14 μmol/g, the observed increase in solution conductivity is simply due to more ions being introduced into the system. The electrochemical mobility of ions decreases with increase in salt concentration due to the formation of salt-DMF and salt-PAN complexes. This balances the increased ion amount at high AAI concentrations, so the solution conductivity remains constant when the AAI concentration is higher than 14 μmol/g. Figure 6 shows SEM images of nanofibers electrospun from pure PAN solution and AAI (57 μmol/g)-added PAN solution. Unlike ZnCl2, which causes a large conductivity change (460 %), the conductivity increase (25 %) caused by AAI is not significantly enough to change the fiber diameters. As a result, nanofiber electrospun from solutions with and without AAI (57 μmol/g) have similar diameters since both solution have similar viscosities and surface tensions.

248

Xiangwu Zhang

Figure 5. Effect of salt concentration on the viscosity, conductivity, and surface tension of PAN electrospinning solutions with different salt additives. Salt additive: (A) ZnCl2, and (B) AAI.

Processing-Structure Relationships of Electrospun Nanofibers

249

Figure 6. Relationship between AAI salt concentration and fiber diameter for electrospun PAN nanofibers. AAI salt concentration: (A) 0 (pure PAN), and (B) 57 μmol/g.

2.1.1.5.Mixed Solvent Mixing solvents with different surface tensions is a simple way to adjust the properties, especially surface tension, of electrospinning solutions. For example, Liu et al.[55] studied electrospinning of cellulose acetate. They found that neither acetone (surface tension: 23.7 dyne/cm) nor dimethylacetamide (surface tension: 32.4 dyne/cm) allows formation of continuous fibers, but a 2:1 acetone:dimethylacetamide solvent mixture, with a surface tension around 26 dyne/cm, enables cellulose acetate to be continuously electrospun into nonwoven mats with fiber diameters ranging from 100 nm to ~1 µm. Fong et al.[60] added ethanol to a polyethylene oxide (PEO)/water solution and produced smooth fibers without the formation of beads. 2.1.2.Operational Conditions In addition to solution properties, the electrospinning operational conditions, such as voltage, solution flow rate, nozzle diameter, spinneret-collector distance, spinneret configuration, and motion of the collector, have significant influence on the diameter of electrospun nanofibers. 2.1.2.1.Voltage One of the crucial parameters is the voltage between the spinneret and collector plate. A higher voltage indicates larger electric field strength. In general, a voltage higher than 6 kV is able to cause the electrospinning solution drop at the tip of the needle to distort into the shape of a Taylor Cone during the initiation of nanofiber electrospinning.[70] At the same time, a higher voltage often leads to greater stretching of the solution due to the greater columbic forces in the jet as well as the stronger electric field. As a result, the fiber diameter may become smaller. In addition, when a solution of lower viscosity is used, a higher voltage favors the formation of secondary jets, which in turn can also causes reduced diameter. However, when the voltage is too high, the flight time for the polymer jet will be reduced, which may cause larger fiber diameters due to the reduced time for the fibers to stretch and elongate.

250

Xiangwu Zhang

Figure 7. Relationship between operational condition and fiber diameter for electrospun PAN nanofibers. Operational condition: (A) voltage, (B) spinneret-collector distance, and (C) feed rate.

In addition, higher voltage also leads to larger mass throughput of the jet, and may further increase the fiber diameter. For example, Figure 7A shows the effect of voltage on the electrospun PAN nanofibers. It is seen that, the fiber diameter increases with increase in voltage. In this particular example, the increase in fiber diameter is mainly caused by the reduced flight time and increased mass throughput of the spinning jet at higher electrostatic force. Similar voltage-diameter relationship has been found in the electrospinning of

Processing-Structure Relationships of Electrospun Nanofibers

251

polyurethane fibers by Demir et al.,[71] polyamide-6 fibers by Mit-Uppatham et al.,[19] and poly(ethylene oxide) fibers by Deitzel et al.[72]

2.1.2.2. Spinneret-Collector Distance The distance between the spinneret and collector hasve a direct influence on both the flight time and electric field strength. During the nanofiber electrospinning, the polymer solution jet must have enough time for most of the solvents to be evaporated. In general, a smaller spinneret-collector distance leads to greater jet stretching and elongation by increasing the electric field strength, which in turn results in smaller fiber diameter. When adjusting spinneret-collector distance, it is very important to ensure that the electrospinning solution jet has enough flight time for the solvents to evaporate. As shown in Figure 7B, when the spinneret-collector distance is too small (10 cm), the jet may not have enough time to dry, leading to a non-uniform fiber mat with spindles. When the distance becomes longer (15 cm), the jet has enough time to dry, and hence relative uniform nanofibers are obtained although the average diameter increase from 528 to 703 nm due to the reduced electric field strength. On the other hand, it is also possible that, when the spinneret-collector distance is reduced, the jet will have a shorter distance to travel before reaching the collector, which may increase the fiber diameter.[73] 2.1.2.3.Solution Flow Rate Changing the solution flow rate is another means to control the nanofiber diameter. When the flow rate is increased, there is a corresponding increase in the fiber diameter simply because a greater volume of solution is drawn away from the spinneret. Figure 7C shows the effect of flow rate on electrospun nanofibers. When the flow rate is increased, the average fiber diameter increases from 700 to 1000 nm since a greater volume of solution is drawn away from the spinneret. It is also seen that fibers electrospun at the lower feed rate (0.50 mL/h) are relatively smooth, but those spun at the high (0.65 mL/h) feed rate are not uniform and spindles are formed. The formation of spindles at high feed rate may be caused by the reduced charge density in the electrospinning jet. A greater volume of solution is drawn at higher feed rate, so an increase in charges is required to get steady jet. As a result, at high feed rate (0.65 mL/h), the voltage may not be high enough to uniformly stretch the jet, and spindles are formed although the electrostatic repulsion within in the jet is sufficient to continue the spinning. Therefore, in order to obtain uniform nanofibers at high flow rates, higher voltages are often needed. 2.1.2.4.Spinneret Diameter Spinneret diameter is also an important operational parameter in controlling the fiber diameter. A decrease in spinneret diameter typically leads to smaller fiber diameters and fewer beads. However, if the spinneret diameter is too small, it may not be possible to extrude the electrospinning solution at the spinneret and hence nanofiber cannot be formed.

252

Xiangwu Zhang

2.1.2.5.Other Operational Conditions In addition to the operational conditions described above, many other processing parameters, such as temperature, humidity, pressure, and atmosphere type, can be used to control the diameter of electrospun nanofibers. For example, increasing temperature can reduce the viscosity of the electrospinning solution, and hence leads to smaller fiber diameter.[19] Increasing humidity can cause water condensation on the fiber surface and influence the surface morphology by improving non-uniformity or even creating surface pore structures.[15, 24, 74] Reducing pressure can increase the tendency of the solution to flow out of the needle, but may cause unstable jet initiation.[4] A very low pressure may also cause direct discharge of the electrical charges and make the electrospinning impossible.[4] Changing air to other gas with higher breakdown voltage such as Freon-2 can increase the fiber diameter.[75]

2.2. Relationship between Processing and Primary Pore Structure In addition to fiber diameter, pore structure (such as pore size and porosity) has significant influence on the performance of electrospun nanofibers. During the electrospinning, nanofibers are typically collected on a grounded collector as a randomlyoriented, nonwoven porous mat, in which the interwoven fibers form primary pores (Figure 8). In addition to primary pores, secondary pores, which are formed inside or on the surface of nanofibers (Figure 8), can also be obtained by using appropriate pore generators. Here, we discuss the processing-primary pore structure relationship, and the formation of secondary pore structure will be addressed in next sSection 2.3.

2.2.1.Pore Size Electrospun nanofiber mats are typically collected by depositing dry fibers on the surface of a grounded collector, and the bending instability associated with the spinning jet help achieve the randomly-oriented nanofiber assembly. By appropriate control of processing parameters, nonwoven nanofiber mats with pore sizes ranging from tens of nanometers to tens of micrometers can be obtained.[21, 37] According to the statistical geometry of electrospun nanofiber mats, the pore size is dominantly determined by the fiber diameter, but other fiber variables (such as porosity, areal density, and fiber density) also have influence.[76] During electrospinning, the pore size can be reduced by decreasing the distance between the spinneret and the collector, increasing the time of nanofiber collection, increasing polymer concentration, and decreasing applied voltage.[77] Figure 9 shows that the size of primary pores of electrospun PAN nanofibers decreases significantly when the spinneret-collector distance decreases from 15 to 3 cm. In most electrospinning setups, the collector plate is made of a conductive material such as aluminum foil, which is electrically grounded. When a nonconductive material is used as the collector, charges on the electrospinning jet can quickly accumulate on the collector which hinders the further deposition of nanofibers.[55, 73] In addition, due to the repulsive forces of the accumulated charges, nanofibers that are collected on the nonconductive material usually have a lower packing density than those collected on a conducting surface.

Processing-Structure Relationships of Electrospun Nanofibers

253

Figure 8. Primary and secondary pores of electrospun nanofibers.

Figure 9. Relationship between spinneret-collector distance and primary pore structure for electrospun PAN nanofibers. Spinneret-collector distance: (A) 15 cm, and (B) 3 cm.

In some cases, the repulsive forces of the accumulated charges on the deposited nanofibers may even cause the formation of large-pore structures (such as dimples).[73]

2.2.2. Porosity In addition to pore size, the porosity has influence on the properties of the electrospun nanofiber assembly. In general, the porosity of a porous structure is always closely related to other structure parameters (such as fiber diameter and pore size in electrospun nanofiber mats). For example, as shown in Figure 9, when the pore size is reduced by the decreasing spinneret-collector distance, the porosity also decreases. Therefore, the porosity of electrospun nanofiber mats can be manipulated by using the similar approaches to those in the control of fiber diameter and primary pore size.

2.3. Relationship between Processing and Secondary Pore Structure In addition to primary pore structure, the secondary pore structure has a significant influence on the properties and applications of electrospun nanofibers. Electrospun nanofibers usually exhibit a solid interior and smooth surface.[3] However, nanofibers with secondary pores have larger surface areas and they can be obtained by using appropriate pore generators.

254

Xiangwu Zhang

2.3.1.Polymer Pore Generator One of the most commonly used methods of fabricating porous nanofibers is the electrospinning of polymer blend fibers, followed by the selectively removal of one of the polymer components.[1, 3, 24] For example, Wendorff and coworkers[78] have prepared porous nanofibers of polylactic acid (PLA) and polyvinylpyrrolidone (PVP) by electrospinning PLA/PVP blend solution, followed by selective removal of one of the polymers. The advantage of this method is that it is relatively easy to obtain nanofibers from polymer blends. However, because of the slow solubilization kinetics of most polymers, this porous nanofiber generation method is time-consuming and it is difficult to completely remove the unwanted polymer component. In addition, it is challenging to control the pore size when a soluble polymer is used as the pore generator. 2.3.2.Nanoparticle Pore Generator In addition to polymers, inorganic nanoparticles, can be used as the pore generators to create porous nanofibers. The advantage of using inorganic nanoparticles as the pore generator is that the pore size and shape of the resultant nanofibers can be precisely controlled by selecting nanoparticles with desired size and shape. Figure 10 compare the structures of solid PAN nanofibers and porous PAN nanofibers, which were prepared by the electrospinning of SiO2/PAN nanofibers, followed by the selectively removal of SiO2 nanoparticles using hydrofluoric (HF) to generate pores. It is seen that the surface of PAN nanofibers without pore generator is relatively smooth. However, a number of cavities and valleys can be seen on nanofibers using SiO2 nanoparticles are a pore generator. 2.3.3. Salt Pore Generator Salts can also be used to generate open pores of electrospun nanofibers. Figure 11 shows open pores generated by the removal of LiCl salt from a LiCl-added Nafion/polyaniline blend using liquid water. In addition to salts, acids or bases can also be used as pore generators depending on the polymer type. We used Fourier Transform Infrared (FTIR) spectroscopy to investigate the pore generation mechanism, and found that the pore generation is related to the breaking of intermolecular interaction between polymer chains by the salt. The salt breaks the intermolecular interactions among polymer chains by forming complexes with them. The breaking of intermolecular interactions reduces the solution viscosity and benefits the electrospinning of thin nanofibers, which in turn results in a higher degree of molecular orientation in electrospun nanofibers. After electrospinning, the removal of salt causes the recovery of intermolecular interactions among polymer chains, and the rearrangement of these polymer chains results in the formation of open pores. One advantage of this method is that the fast solubilization kinetics of salts significantly increases the pore generation speed and reduces the cost for fabricating porous nanofibers with controlled structures. In addition, the pore size and porosity can be easily modified by selectively adjusting the type and concentration of salts. Typically, with increase in salt concentration, pore size and porosity increase due to the increased interaction sites between salts and polymer chains.

Processing-Structure Relationships of Electrospun Nanofibers

255

Figure 10. SEM images of (A) PAN nanofibers, and (B) porous PAN nanofibers, in which the secondary pores were generated by the removal of SiO2 nanoparticles after the electrospinning of SiO2/PAN composites.

Figure 11. Open pores generated from Nafion-polyaniline by the removal of LiCl salt.

The salt removal process (such as solvent type and temperature) also has influence on the generation and structures of secondary pores. Selecting a solvent that has better ability to dissolve salts results in larger pore size and higher porosity due to the enhanced salt solubilization kinetics. Increasing the process temperature is another means to get larger pore size and higher porosity because the removal of salts is faster at higher temperatures.

2.4. Relationship between Processing and Other Secondary Structure In addition to porous fibers, nanofibers with other secondary structures can also be synthesized by using appropriate processing parameters or design of spinnerets. Figure 12 shows schematics of nanofibers with four different secondary structures: A) core-sheath; B) hollow; C) core-sheath, porous; and D) hollow, porous nanofibers.

2.4.1.Core-Sheath Structure Core/sheath nanofibers (Figure 12A) can be obtained by co-electrospinning two different polymer solutions through a spinneret comprising two coaxial needles (Figure 13).

256

Xiangwu Zhang

Figure 12. Schematics of electrospun nanofibers with various secondary structures: (A) core-sheath; (B) hollow; (C) core-sheath, porous; and (D) hollow, porous nanofibers.

Figure 13. Schematic of a coaxial spinneret system.

For example, Sun et al.[79] used this coaxial spinneret system to co-electrospin PEOPEO, polysulfone (PSU)-PEO, polydodecylthiophene (PDT)-PEO, Pd metal-poly(L-lactide) (PLA) core-sheath nanofibers.

2.4.2.Hollow Structure Compared with core-sheath fibers, nanofibers with hollow interiors can have more surface area (Figure 12B). Xia and coworkers [1, 3, 80] used the coaxial spinneret system to co-electrospin mineral oil-TiO2/polyvinylpyrrolidone (PVP) core-sheath nanofibers, and the

Processing-Structure Relationships of Electrospun Nanofibers

257

selective removal of the oil phase (core) by solvent extraction resulted in the formation of hollow nanofibers consisting of TiO2/PVP composite walls. The oil core can also be removed by heat-treating the core-sheath nanofibers at elevated temperatures.[3]

2.4.3.Other Secondary Structures The above four secondary structures can be combined to create others, such as coresheath, porous (Figure 12C) and hollow, porous (Figure 12D) nanofibers. Core-sheath, porous nanofibers have solid core and porous sheath, while hollow, porous nanofibers have hollow interior and porous sheath. In addition to nanofibers mentioned above, other types of nanofibers, including ribbonlike,[81, 82] branched,[81] helical[83] structures, can also be obtained by appropriate control of the electrospinning process.

3. PROCESSING-STRUCTURE RELATIONSHIPS OF POLYMER COMPOSITE NANOFIBERS There are many approaches to prepare polymer composite nanofibers, and the following are some examples.

3.1. Nanoparticle-Added Composite Nanofibers The most straightforward method to fabricate polymer composite nanofibers is to incorporate nanoparticles (e.g., titania, carbon black, silver, and iron oxides)[27-30] and nanotubes [e.g., carbon nanotubes (CNT)][31-33] into electrospun polymers by directly adding these nanoparticles to the polymer solution before electrospinning. One challenge of this method is to obtain a homogenous distribution of nanoparticles in nanofibers. Figure 14 shows TEM images of SiO2/PAN composite nanofibers, in which SiO2 nanoparticles aggregate. The formation of aggregates can be avoided by adding surfactant to the electrospinning solution or surface modifying the nanoparticles.

3.2. Metal-Deposited Composite Nanofibers Composite nanofibers can also be made by adding nanoparticles into or onto polymer nanofibers after electrospinning. Figure 15 show the preparation of Cu- and Pt-loaded composite nanofibers using an electrodeposition method, i.e., depositing particles from metal salt solutions at a constant potential. Electrodeposition of metal particles can also be carried out by applying successive voltammetric cycles. The size of deposited particles can be controlled by selectively adjusting the electrodeposition conditions, such as voltage, time, cycle rate, and cycle number. For example, by reducing the electrodeposition time, smaller Pt nanoparticles can be obtained on nanofiber surface (Figure 16). In addition to particle size, the shape of deposited metal particles can also be controlled by selectively adjusting the deposition conditions, such as current, potential, and salt concentration, etc. Figure 17 shows different structures obtained for Cu depositions, including smooth particles, rough particles,

258

Xiangwu Zhang

pea-shape coating, and thin-film coating. In addition to electrodeposition, chemical deposition can be used to create metal-deposited composite nanofibers.

Figure 14.TEM images of SiO2/PAN composite nanofibers.

Figure 15. Electrodeposition of nanoparticles on nanofiber surfaces. (A) Cu particles, and (B) Pt particles.

Processing-Structure Relationships of Electrospun Nanofibers

259

Figure 16. TEM image of a Pt nanoparticle-deposited nanofiber.

Figure 17. SEM images of Cu-deposited nanofibers with different Cu deposition structures: (A) smooth particles, (B) rough particles, (C) pea-shape coating, and (D) thin-film coating.

260

Xiangwu Zhang

3.3. Salt-Added Composite Nanofibers Another unique means to create composite nanofibers is to electrospin fibers from saltadded polymer solutions. Figure 18 shows TEM images of ZnCl2-added PAN nanofibers. Reducing the salt into metal nanoparticles or converting the salt into metal oxides can lead to the formation of nanoparticle-filled polymer nanofibers. In addition, composite nanofibers with unique structures can also be obtained by leaching the dissolved salt onto the fiber surface to form salt layers, particles, or plates. Figures 19A and B show ZnCl2/PAN composite nanofibers with different surface ZnCl2 structures, i.e., layers and plates, respectively.

4. PROCESSING-STRUCTURE RELATIONSHIPS OF CARBON NANOFIBERS Carbon nanofibers can be prepared by carbonization of fibers electrospun from different types of precursor polymers, such as PAN, polyimide (PI), and pitch, etc.[34-37] Solid, porous, and composite carbon nanofibers can also be obtained by using appropriate precursor composition and spinning condition.

4.1. Solid Carbon Nanofibers Figure 20 shows solid carbon nanofibers, which were prepared by the electrospinning of PAN nanofibers, following by carbonization. Carbonization conditions (such as carbonization temperature and heating rate) can be adjusted to control the carbon crystalline structure and hence the properties of carbon nanofibers.[34-36] Typically, with increase in carbonization temperature, carbon nanofibers present more ordered carbon structure. When the carbonization goes beyond 1,800 oC, carbon nanofibers can be converted to graphite nanofibers, which have higher crystallinity and larger modulus.

4.2. Porous Carbon Nanofibers Porous carbon nanofibers can be obtained by combining the carbonization and pore generation processes. Pore generators are typically added directly into the electrospinning solutions, but the pore generation step (i.e., the removal of the pore generator) can be carried out before, during, or after the carbonization of nanofibers. Figure 21 shows porous carbon nanofibers, which were prepared by electrospinning a PLA/PAN solution, followed by the carbonization. During the carbonization of PAN, PLA was removed to generate pores because PLA degrades at the carbonization temperature of PAN. The carbonization condition, such as temperature and atmosphere, has influence on the pore structure of carbon nanofibers. For example, From Figure 21, it is seen that carbon nanofibers carbonized at 1,000 oC in nitrogen have better-defined pore structure than those carbonized at 800 oC in argon.

Processing-Structure Relationships of Electrospun Nanofibers

261

Figure 22 shows porous carbon nanofibers, which were made by electrospinning and carbonizing SiO2/PAN precursor solutions. The pores were generated by the removal of SiO2 nanoparticles using HF acid after the carbonization of PAN.

Figure 18.TEM images of ZnCl2/PAN composite nanofibers.

Figure 19.TEM images of ZnCl2/PAN composite nanofibers, with different ZnCl2 structures: (A) layers, and (B) salt plates.

Figure 20.SEM images of PAN-based carbon nanofibers.

262

Xiangwu Zhang

Figure 21.Relationship between carbonization condition and structure for porous carbon nanofibers prepared by carbonizing electrospun PLA/PAN nanofibers. Carbonization conduction: (A) 800 oC in argon, and (B) 1,000 oC in nitrogen.

Figure 22.Relationship between carbonization condition and structure for porous carbon nanofibers prepared by carbonizing electrospun SiO2/PAN nanofibers, followed by the removal of SiO2 nanoparticles using HF. Carbonization conduction: (A) 700 oC in nitrogen, and (B) 1,000 oC in nitrogen.

From Figure 22, it is seen that nanofibers carbonized at 1,000 oC have homogenous pore distribution, but those carbonized at 700 oC have irregular pore distribution and the pore size is relatively large. Porous carbon nanofibers can also be prepared by using salts as the pore generator. The salt can be removed before or after the carbonization to generate pores. In salt-induced porous nanofibers, the formation of pores is caused by the rearrangement of polymer chains during the removal of salts. As a result, salts can generate smaller pores than polymer and nanoparticle pore generators. Figure 23 shows porous carbon nanofibers prepared by electrospinning and carbonizing ZnCl2/PAN nanofibers, followed by the removal of Zn residual using HCl solution and liquid water. Due to the small pore size, pores generated in these nanofibers actually cannot be seen in the SEM image.

Processing-Structure Relationships of Electrospun Nanofibers

263

Figure 23. SEM images of porous carbon nanofibers prepared by carbonizing electrospun ZnCl2/PAN nanofibers, followed by the removal of Zn residual using HCl solution and DI water.

However, Brunauer-Emmett-Teller (BET) nitrogen adsorption measurements show that carbon nanofibers shown in Figure 23 have a large amount of nanopores, lead to a large nanofiber surface area > 300 m2/g. The pore size and porosity of salt-generated porous carbon nanofibers can be easily modified by selectively adjusting the type and concentration of salts. Typically, higher salt concentration leads to larger pore size and greater porosity.

4.3. Composite Carbon Nanofibers Composite carbon nanofibers can be prepared by carbonizing electrospun polymer composite fibers.[33,38] For these carbon composite nanofibers, the carbonization process of the polymer matrix may change the nanoparticle distribution due to fiber shrinkage. An extreme example is the preparation of single-walled carbon nanotube (SWNT)-filled carbon nanofibers via electrospinning and carbonization. Ko et al.[33] homogenously dispersed SWNT into PAN/dimethylformamide solution, which wasere electrospun and heat-treated to form SWNT-filled carbon nanofibers. TEM observations confirm that most SWNTs maintain their straight shape after carbonization and are parallel to the axis of the carbon fiber without agglomeration. However, a small amount of SWNTs protrude out of the carbon fiber as a direct result of the fiber shrinkage during heat treatment. Metal oxide-filled carbon nanofibers can be prepared by electrospinning salt-added polymer nanofibers, followed by converting the salt to metal oxides during or after the carbonization. Figure 24 shows a Mn oxide-filled composite carbon nanofiber, which was prepared by carbonizing MnCl2/PAN nanofibers. During the carbonization, the Mn salt was converted to Mn oxide particles. In addition to metal oxides, metal particles can also be obtained in carbon nanofibers by reducing the salt additives.

5. PROCESSING-STRUCTURE RELATIONSHIPS OF CERAMIC NANOFIBERS Ceramic nanofibers can be synthesized by electrospinning their precursor solutions, followed by calcination. Ceramic nanofibers synthesized by electrospinning were reported

264

Xiangwu Zhang

first in 2002;[7, 63] since then, more than twenty ceramic systems (e.g., SiO2, Al2O3, and ZrO2)[1], [84-92] have been fabricated as nanofibers. These fibers can be amorphous, polycrystalline, dense, porous, or hollow.[63]

Figure 24. TEM image of a Mn oxide-filled carbon nanofiber prepared by carbonizing electrospun MnCl2/PAN nanofiber.

Figure 25. SEM image of ZrO2 nanofibers.

Figure 25 shows ZrO2 nanofibers, which were electrospun from zirconium oxychloride (ZrOCl2) precursor. During the electrospinning, polyvinyl alcohol (PVA) was used to adjust the precursor viscosity and improve the electrospinnability. The electrospun ZrOCl2/PVA nanofibers were heat-treated to form ZrO2 nanofibers (PVA was pyrolyzed during the heattreatment).

Processing-Structure Relationships of Electrospun Nanofibers

265

6. PROCESSING-STRUCTURE RELATIONSHIPS OF METAL NANOFIBERS Metal nanofibers can also be produced using electrospinning. Bognitzki et al.[93] fabricated Cu nanofibers by the electrospinning of copper nitrate–polyvinylbutyral (PVB) solutions to corresponding composite fibers, followed by heat treatment in air (conversion of copper nitrate to copper oxide and degradation of PVB), followed by heat treatment in a hydrogen atmosphere (conversion of CuO to Cu). In addition, Fe, Co, and Ni nanofibers have also been prepared by the combination of electrospinning and heat treatment.[94]

CONCLUSION Electrospinning has become an important technology to produce nanofibers from a rich variety of polymers, composites, carbons, ceramics, and metals. The diameter of electrospun polymer nanofibers can be controlled by selectively adjusting the solution properties and operational conditions. The primary pore structure of electrospun polymer nanofibers can be manipulated by applying the strategy similar to that used in fiber diameter control. However, the secondary pore structure of electrospun polymer nanofibers is typically obtained by using different pore generators, such as polymers, nanoparticles, and salts. In addition to secondary pores, other secondary structures, such as core-sheath and hollow configurations, can also be created in electrospun polymer nanofibers by using appropriate processing parameters or design of spinnerets. Electrospun nanofibers of composites, carbons, ceramics, and metals are often obtained from polymer-containing precursors, and hence the processing-structure relationships of polymer nanofibers can be used after careful consideration of the unique characteristics of these materials.

ACKNOWLEDGEMENT The author thanks the contributions from Liwen Ji, Zhan Lin, Yingfang Yao, Shuli Li, Quan Shi, Narendiran Vitchuli, Jinmei Du, Andrew James Medford, and Samantha Shintay.

REFERENCES [1] [2]

[3]

McCann, J. T., Li, D. and Xia, Y. Electrospinning of Nanofibers with Core-Sheath, Hollow, or Porous Structures. J. Mater. Chem. 15, 735-738 (2005). Norris, I. D., Shaker, M. M., Ko, F. K. and MacDiarmid, A. G. Electrostatic Fabrication of Ultrafine Conducting Fibers: Polyaniline/Polyethylene Oxide Blends. Synth. Met. 114, 109-114 (2004). Li, D. and Xia, Y. Electrospinning of Nanofibers: Reinventing the Wheel. Adv. Mater. 16, 1151-1170 (2004).

266 [4]

[5]

[6] [7]

[8] [9]

[10] [11]

[12] [13]

[14] [15]

[16]

[17] [18] [19]

[20]

[21]

Xiangwu Zhang Ramakrishna, S., Fujihara, K., Teo, W. E., KLim, T. C. and Ma, Z. An Introduction to Electrospinning and Nanofibers, World Scientific Publishing Co. Pte. Ltd., Singapore. (2005). Norris, I. D., Shaker, M. M., Ko, F. K. and MacDiarmid, A. G. Electrostatic Fabrication of Ultrafine Conducting Fibers: Polyaniline/Polyethylene Oxide Blends. Synth. Met. 114, 109 (2004). Huang, C. et al. Electrospun Polymer Nanofibres with Small Diamters. Nanotech. 17, 1558-1563 (2006). Dai, H., Gong, J., Kim, H. and Lee, D. A Novel Method for Preparing Ultra-Fine Alumina-Borate Oxide Fibres via an Electrospinning Technique. Nanotechnology 13, 674 (2002). Kanehata, M., Ding, B. and Shiratori, S. Nanoporous Ultra-High Specific Surface Inorganic Fibres. Nanotech. 18, 315602 (2007). Kim, C. et al. Fabrication of Electrispinning-Derived Carbon Nanofiber Webs for the Anode Material of Lithium-Ion Secondary Batteries. Adv. Funct. Mater. 16, 2393-2397 (2006). Wang, C. et al. Electrospinning of Polyacrylonitrile Solutions at Elevated Temperatures. Macromolecules 40, 7973 -7983 (2007). Fennessey, S. F. and Farris, R. J. Fabrication of Aligned and Molecularly Oriented Electrospun Polyacrylonitrile Nanofibers and the Mechanical Behavior of their Twisted Yarns. Polymer 45, 4217-4225 (2004). Lee, H. K., Jeong, E. H., Baek, C. K. and Youk, J. H. One-Step Preparation of Ultrafine Poly(acrylonitrile) Fibers Containing Silver Nanopartciles. (2005). Phadke, M. A., Musale, D. A., Kulkarni, S. S. and Karode, S. K. Poly(Acrylonitrile) Ultrafiltration Membranes. I. Polymer-Salt-Solvent Interactions. J. Polym. Sci.: Polym. Phys. 43, 2061-2073 (2005). Samatham, R. et al. Electrospun Nanoscale Polyacrylonitrile Artificial Muscle. Smart Mater. Struct. 15, N152-N156 (2006). Casper, C. L., Stephens, J. S., N.G., T., Chase, D. B. and Rabolt, J. F. Controlling Surface Morphology of Electrospun Polystyrene Fibers: Effect of Humidity and Molecular Weight in the Electrospinning Process. Macromolecules 37, 573-578 (2004). Jarusuwannapoom, T. et al. Effect of Solvents on Electro-Spinnability of Polystyrene Solutions and Morphological Appearance of Resulting Electrospun Polystyrene Fibers. Euro. Polym. J. 41, 409-421 (2005). Lee, K. H., Kim, H. Y., Bang, H. J., Jung, Y. H. and Lee, S. G. The Change of Bead Morphology Formed on Electrospun Polystyrene Fibers. Polymer 44, 4029 (2003). Bergshoef, M. M. and Vancso, G. J. Transparent Nanocomposites with Ultrathin Electrospun Nylon-4,6 Fiber Reinforcement. Adv. Mater. 11, 1362-1365 (1999). Mit-uppatham, C., Nithitanakul, M. and Supaphol, P. Ultratine Electrospun Polyamide6 Fibers: Effect of Solution Conditions on Morphology and Average Fiber Diameter. Macromol. Chem. Phys. 205, 2327-2338 (2004). Stephens, J. S., Chase, D. B. and Rabolt, J. F. Effect of the Electrospinning Process on Polymer Crystallization Chain Conformation in Nylon-6 and Nylon-12. Macromolecules 37, 877-881 (2004). Ryu, Y. J., Kim, H. Y., Lee, K. H., Park, H. C. and Lee, D. R. Transport Properties of Electrospun Nylon 6 Nonwoven Mats. European Polym. J. 39, 1883 (2003).

Processing-Structure Relationships of Electrospun Nanofibers

267

[22] Chung, G. S., Jo, S. M. and Kim, B. C. Properties of carbon Nanofibers Prepared from Electrospun Polyimide. J. Appl. Polym. Sci. 97, 165-170 (2005). [23] Ding, B., Yamazaki, M. and Shiratori, S. Electrospun Fibrous Polyacrylic Acid Membrane-Based Gas Sensors. Sensor Actuat B-Chem. 106, 477-483 (2004). [24] Bognitzki, M. et al. Nanostructured Fibers via Electrospinning. Adv. Mater. 13, 70-72 (2001). [25] Choi, S. S., Lee, S. G., Joo, C. W., Im, S. S. and Kim, S. H. Formation of Interfiber Bonding in Electrospun Poly(Etherimide) Nanofiber Web. J. Mater. Sci. 39, 1511-1513 (2004). [26] Deitzel, J. M., Kleinmeyer, J. D., Hirvonen, J. K. and Tan, N. C. B. Controlled Deposition of Electrospun Poly(Ethylene Oxide) Fibers. Polymer 42, 8163-8170 (2001). [27] Lee, K. H., Kim, H. Y., Ryu, Y. J., Kim, K. W. and Choi, S. W. Mechanical Behavior of Electrospun Fiber Mats of Poly(vinyl chloride)/Polyurethane Polyblends. J. Polym. Sci. Pol. Phys. 41, 1256-1262 (2003). [28] Kim, J. S. and Reneker, D. H. Polybenzimidazole Nanofiber Produced by Electrospinning. Polym. Engr. Sci. 39, 849 - 854 (2004). [29] Kim, C., Kim, Y. J. and Kim, Y. A. Fabrication and Structural Characterization Of Electro-Spun Polybenzimidazol-Derived Carbon Nanofiber by Graphitization. Solid State Commun. 132, 567-571 (2004). [30] Koski, A., Yim, K. and Shivkumar, S. Effect of Molecular Weight on Fibrous PVA Produced by Electrospinning. Mater. Lett. 58, 493-497 (2004). [31] Bai, J. et al. A Simple and Effective Route for the Prepartion of Poly(Vinylalcohol) (PVA) Nanofibers Containing Gold Nanopartciles by Electrospinning Method. Solid State Comm. 141, 292-295 (2007). [32] Yang, Q. B. et al. Influence of Solvents on the Formation of Ultrathin Uniform Poly(vinyl pyrrolidone) Nanofibers with Electrospinning. J. Polym. Sci. Pol. Phys. 42, 3721-3726 (2004). [33] Choi, S. S. et al. Electrospun PVDF Nanofiber Web as Polymer Electrolyte or Separator. Electrochim. Acta 50, 338-342 (2004). [34] Choi, S. W., Jo, S. M., Lee, W. S. and Kim, Y. R. An Electrospun Poly(vinylidene fluoride) Nanofibrous Membrane and Its Battery Applications. Adv. Mater. (2003). [35] Ma, Z. W., Kotaki, M., Yong, T., He, W. and Ramakrishna, S. Surface Engineering of Electrospun Polyethylene Terephthalate (PET) Nanofibers Towards Development of A New Material For Blood Vessel Engineering. Biomater. 26, 2527-2536 (2005). [36] Khil, M. S., Kim, H. Y., Kim, M. S., Park, S. Y. and Lee, D. R. Nanofibrous Mats of Poly(Trimethylene Terephthalate) via Electrospinning. Polymer 45, 295-301 (2004). [37] Kim, J. R., Choi, S. W., Jo, S. M., Lee, W. S. and Kim, B. C. Characterization and Properties of P(VdF-HFP)-Based Fibrous Polymer Electrolyte Membrane Prepared by Electrospinng. J. Electrochem. Soc. 152, A295 (2005). [38] Fong, H. and Reneker, D. H. Elastomeric Nanofibers of Styrene-Butadiene-Styrene Triblock Copolymer. J. Polym. Sci.: Part B: Polym. Phys. 37, 3488 (1999). [39] Boland, E. D. et al. Electrospinning Polydioxanone for Biomedical Applications. Acta. Biomater. l, 115-123 (2005).

268

Xiangwu Zhang

[40] You, Y., Min, B. M., Lee, S. J., Lee, T. S. and Park, W. H. In Vitro Degradation Behavior of Electrospun Polyglycolide, Polylactide, and Poly(lactide-coglycolide). J. Appl. Polym. Sci. 95, 193-200 (2005). [41] Yoshimoto, H., Shin, Y. M., Terai, H. and Vacanti, J. P. A Biodegradable Nanofiber Scaffold by Electrospinning and its Potential For Bone Tissue Engineering. Biomater. 24, 2077-2082 (2003). [42] Ohkawa, K., Kim, H. Y. and Lee, K. H. Biodegradation of Electrospun Poly(ecaprolactone) Non-woven Fabrics by Pure-Cultured Soil Filamentous Fungi. J. Polym. Environ. 12, 211-218 (2004). [43] Mo, X. M., Xu, C. Y., Kotaki, M. and Ramakrishna, S. Electrospun P(LLA-CL) Nanofiber: A Biomimetic Extracellular Matrix for Smooth Muscle Cell and Endothelial Cell Proliferation. Biomater. 25, 1883-1890 (2004). [44] Li, W. J., Laurencin, C. T., Caterson, E. J., Tuan, R. S. and Ko, F. K. Electrospun Nanofibrous Structure: A Novel Scaffold for Tissue Engineering. J. Biomed. Mater. Res. 60, 613-621 (2002). [45] Bini, T. B. et al. Electrospun Poly(L-Lactide-co-Glycolide) Biodegradable Polymer Nanofibre Tubes for Peripheral Nerve Regeneration. Nanotechnol. 15, 1459-1464 (2004). [46] Matthews, J. A., Wnek, G. E., Simpson, D. G. and Bowlin, G. L. Electrospinning of Collagen Nanofibers. Biomacromolecules 3, 232-238 (2002). [47] Shields, K. J., Beckman, M. J., Bowlin, G. L. and Wayne, J. S. Mechanical Properties and Cellular Proliferation of Electrospun Collagen Type II. Tissue Eng. 10, 1510-1517 (2004). [48] Huang, Z. M., Zhang, Y. Z., Ramakrishna, S. and Lim, C. T. Electrospinning and Mechanical Characterization of Gelatin Nanofibers. Polymer 45, 5361-5368 (2004). [49] Zhang, Y., Ouyang, H., Lim, C. T., Ramakrishna, S. and Huang, Z. M. Electrospinning of Gelatin Fibers and Gelatin/PCL Composite Fibrous Scaffolds. J. Biomed. Mater. Res. Part B. 72B, 156-165 (2005). [50] Wnek, G. E., Carr, M. E., Simpson, D. G. and Bowlin, G. L. Electrospinning of Nanofiber Fibrinogen Structures. Nano Lett. 3, 213-216 (2003). [51] Jin, H. J., Fridrikh, S. V., Rutledge, G. C. and Kaplan, D. I. Electrospinning Bombyx Mori Silk with Poly(Ethylene Oxide). Biomacromolecules 3, 1233-1239 (2002). [52] Ohgo, K., Zhao, C. H., Kobayashi, M. and Asakura, T. Preparation of Non-Woven Nanofibers of Bombyx Mori Silk, Samia Cynthia Ricini Silk and Recombinant Hybrid Silk with Electrospinning Method. Polymer 44, 841-846 (2003). [53] Min, B. M. et al. Electrospinning of Silk Fibroin Nanofibers and its Effect on the Adhesion and Spreading of Normal Human Keratinocytes and Fibroblasts in Vitro. Biomater. 25, 1289-1297 (2004). [54] Son, W. K., Youk, J. H., Lee, T. S. and Park, W. H. Electrospinning of Ultrafme Cellulose Acetate Fibers: Studies of a New Solvent System and Deacetylation of Ultrafme Cellulose Acetate Fibers. J. Polym. Sci. Pt. B-Polym. Phys. 42, 5-11 (2004). [55] Liu, H. and Hsieh, Y.-L. Ultrafine Fibrous Cellulose Membranes from Electrospinning of Cellulose Acetate. J. Polym. Sci. Part B: Polym. Phys. 40, 2119-2129 (2002). [56] Ohkawa, K., Cha, K., Kim, H. Y., Nishida, A. and Yamamoto, H. Electrospinning of Chitosan. Macromol. Rapid Comm. 25, 1600-1605 (2004).

Processing-Structure Relationships of Electrospun Nanofibers

269

[57] Duan, B., Dong, C. H., Yuan, X. Y. and Yao, K. D. Electrospinning of Chitosan Solutions in Acetic Acid with Polyethylene Oxide. J. Biomat. Sci.-Polym. E. 15, 797811 (2004). [58] Um, I. C., Fang, D. F., Hsiao, B. S., Okamoto, A. and Chu, B. Electro-Spinning and Electro-Blowing of Hyaluronic Acid. Biomacromolecules 5, 1428-1436 (2004). [59] Reneker, D. H. and Chun, I. Nanometre Diameter Fibres of Polymer, Produced by Electrospinning. Nanotechnol. 7, 216-223 (1996). [60] Fong, H., Chun, I. and Reneker, D. H. Beaded Nanofibers Formed During Electrospinning. Polymer 40, 4585-4592 (1999). [61] Shenoy, S. L., Bates, W. D., Frisch, H. L. and Wnek, G. E. Role of Chain Entanglements on Fiber Formation During Electrospinning of Polymer Solutions: Good Solvent, Non-Specific Polymer-Polymer Interaction Limit. Polymer 46, 3372-3384 (2005). [62] Gupta, P., Elkins, C., Long, T. E. and Wilkes, G. L. Electrospinning of Linear Homopolymers of Poly(methyl methacrylate): Exploring Relationships Between Fiber Formation, Viscosity, Molecular Weight and Concentration in a Good Solvent. Polymer 46, 4799-4810 (2005). [63] Sigmund, W. et al. Processing and Structure Relationships in Electrospinning of Ceramic Fiber Systems. J. Am. Ceram. Soc. 89, 395-407 (2006). [64] McKee, M. G., Wilkes, G. L., Colby, R. H. and Long, T. E. Correlations of Solution Rheology with Electrospun Fiber Formation of Linear and Branched Polyesters. Macromolecules 37, 1760-1767 (2004). [65] Yu, J. H., Fridrikh, S. V. and Rutledge, G. C. The Role of Elasticity in the Formation of Electrospun Fibers. Polymer 47, 4789-4797 (2006). [66] Talwar, S. Hydrophobically Modified Associative Polymers: Solution Rheology and Application in Electrospinning. PhD Thesis NCSU (2008). [67] Talwar, S., Hinestroza, J. P., Pourdeyhimi, B. and Khan, S. A. Electrospun Nanofibers of Associative Polymer-Surfactant Systems. ACS Nano (submitted) (2008). [68] Fridrikh, S. V., Yu, J. H., Brenner, M. P. and Rutledge, G. C. Controlling the fiber diameter during electrospinning. Phys. Rev. Lett. 90, 144502/1-144502/4 (2003). [69] Zhang, X., Pan, Y., Cheng, J. and Yi, X. The Influence of Low-Melting-Point Alloy on the Rheological Properties of a Polystyrene Melt. J. Mater. Sc. 35, 4573-4581 (2000). [70] Taylor, G. Disintegration of Water Drops in an Electric Field. Proc. R. Soc. Lond. A. 280, 918-924 (1964). [71] Demir, M. M., Yilgor, I., Yilgor, E. and Erman, B. Electrospinning of Polyurethane Fibers. Polymer 43, 3303-3309 (2002). [72] Deitzel, J. M., Kleinmeyer, J., Harris, D. and Beck Tan, N. C. The Effect of Processing Variables on the Morphology of Electrospun Nanofibers and Textiles. Polymer 42, 261 (2001). [73] Ramakrishna, S., Fujihara, K., Teo, W. E., Lim, T. C. and Ma, Z. An Introduction to Electrospinning and Nanofibers, World Scientific Publishing Co. Pte. Ltd. (2005). [74] Megelski, S., Stephens, J. S., Chase, D. B. and Rabolt, J. F. Micro- and Nanostructured Surface Morphology on Electrospun Polymer Fibers. Macromolecules 35, 8456-8466 (2002). [75] Baumgarten, P. K. Electrostatic Spinning of Acrylic Microflbers. J. Colloid Interf. Sci. 36, 75-79 (1971).

270

Xiangwu Zhang

[76] Eichhorn, S. J. and Sampson, W. W. Statistical Geometry of Pores and Statistics of Porous Nanofibrous Assemblies. J. R. Soc. Interface 2, 309 (2005). [77] Hong, Y. L. et al. Synthesis Using Electrospinning and Stabilization of Single Layer Macroporous Films and Fibrous Networks of Poly(Vinyl Alcohol). J. Membr. Sci. 276, 1 (2006). [78] Bognitzki, M., Frese, T., Steinhart, M., Greiner, A. and Wendorff, J. H. Preparation of Fibers with Nanostructured Morphologies: Electrospinning of Polymer Blends. Polym. Engr. Sci. 41, 982 (2001). [79] Sun, Z., Zussman, E., Yarin, A. L., Wendorff, J. H. and Greiner, A. Compound CoreShell Polymer Nanofibers by Co-Electrospinning. Adv. Mater. 15, 1929 (2003). [80] Li, D. and Xia, Y. Direct Fabrication of Composite and Ceramic Hollow Nanofibers by Electrospinning. Nano Lett. 4, 933 (2004). [81] Koombhongse, S., Liu, W. and Reneker, D. H. Flat Polymer Ribbons and Other Shapes by Electrospinning. J. Polym. Sci., Part B. Polym. Phys, 39, 2598-2606 (2001). [82] Huang, L. et al. Production of Synthetic Elastin-Mimetic Nano-Fibers and its Application as Biomaterials. Abstr. Pap. Am. Chem. Soc. 221, U447-U447 (2001). [83] Kessick, R. and Tepper, G. Microscale Polymeric Helical Structures Produced by Electrospinning. Appl. Phys. Lett. 84, 4807-4809 (2004). [84] Zhang, G. et al. Electrospun Nanofibers for Potential Space-Based Applications. Mat. Sci. Eng. B-Solid 116, 353 (2005). [85] Choi, S. S., Lee, S. G., Im, S. S., Kim, S. H. and Joo, Y. L. Silica Nanofibers from Electrospinning/Sol-Gel Process. J. Mater. Sci. Lett. 22, 891 (2003). [86] Larsen, G., Velarde-Ortiz, R., Minchow, K., Barrero, A. and Losertales, I. G. A Method for Making Inorganic and Hybrid (Organic/Inorganic) Fibers and Vesicles with Diameters in the Submicormeter and Micrometer Range via Sol-Gel Chemistry and Electrically Forced Liquid Jets. J. Am. Chem. Soc. 125, 1154 (2003). [87] Yuh, J., Nino, J. C. and Sigmund, W. Synthesis of Barium Titanate (BaTiO3) Nanofibers via Electrospinning. Mater. Lett. 59, 3645 (2005). [88] Wang, M., Singh, H., Hatton, T. A. and Rutledge, G. C. Field-Responsive Superparamagnetic Compsoite Nanofibers by Electrospinning. Polymer 45, 5505-5514 (2004). [89] Viswanthamurthi, P. et al. GeO2 Fibers: Preparation, Morphology and Photoluminescence Property. J. Ceram. Phys. 121, 441 (2004). [90] Madhugiri, S., Sun, B., Smirniotis, P. G., J.P., F. and Balkus, K. J. Electrospun Mesoporous Titanium Dioxide Fibers. Micropor. Mesopor. Mater. 69, 77 (2004). [91] Shao, C. L., Guan, H. Y., Liu, Y. C., Yu, R. and Yang, X. H. A Novel Method for Making ZrO2 Nanofibers via an Electrospinning Technique. J. Cryst. Growth. 267, 380 (2004). [92] Jing, N., Wang, M. and Kameoka, J. Fabrication of Ultrathin ZrO2 Nanofiber by Electrospinning. J. Photopolym. Sci. Tech. 18, 503 (2005). [93] Bognitzki, M. et al. Preparation of Sub-micrometer Copper Fibers via Electrospinning. Adv. Mater. 18, 2384-2386 (2006). [94] Wu, H., Zhang, R., Liu, X. X., Lin, D. D. and Pan, W. Electrospinning of Fe, Co, and Ni nanofibers: Synthesis, Assembly, and Magnetic Properties. Chem. Mater. 19, 35063511 (2007).

In: Nanofibers: Fabrication, Performance, and Applications ISBN 978-1-60741-947-1 Editors: W. N. Chang © 2009 Nova Science Publishers, Inc.

Chapter 8

GLYCOSYLATED NANOFIBERS FOR PROTEIN ADSORPTION AND RECOGNITION Ai-Fu Che, Ling-Shu Wan and Zhi-Kang Xu∗ Institute of Polymer Science, and Key Laboratory of Macromolecular Synthesis and Functionalization (Ministry of Education), Department of Polymer Science and Engineering, Zhejiang University, Hangzhou, P.R. China

ABSTRACT As a kind of biomacromolecules, carbohydrates are found on the external surface of cell membranes in the forms of glycoproteins, glycolipids and polysaccharides. They play essential roles in biological processes such as cell adhesion, blood coagulation, viral infection, immune response and apoptosis. Numerous biological phenomena are based on the carbohydrate-protein interaction. In nature, carbohydrates always interact with specific proteins through multivalent interaction, namely “cluster glycoside effect”. A great number of glycopolymers have been designed and synthesized to mimic the multivalent functions of natural glycoconjugates. It is expected that nanostructured materials with morphology similar to the native extracellular matrix will be more interesting for the mimicking of “cluster glycoside effect”. Therefore, a series of polyacrylonitrile-based nanofibers with glycosylated surfaces were studied in our laboratory. Two protocols were used to fabricate these glycosylated nanofibers. One is the synthesis of glycopolymers followed by electrospinning and the other is the surface modification of polyacrylonitrile nanofibers having reactive groups. We found that the morphology of the glycosylated nanofibers could be modulated by the characteristics of glycopolymers and the parameters of electrospinning and surface modification. These glycosylated nanofibers were studied for protein adsorption and recognition. Concanavalin A (Con A), peanut agglutinin (PNA) and bovine serum albumin (BSA) were used for comparison. Water contact angle measurement confirms that the glycosylated nanofiber surface is hydrophilic which facilitates the resistance to the nonspecific adsorption of proteins. Because of the specific interaction of Con A and glucose residues, nanofibers with glucose groups have strong affinity with Con A, but present no binding with PNA or BSA. By contrast, those with galactose groups can ∗ Corresponding author. Email: [email protected]. Fax: + 86 571 8795 1773.

272

Ai-Fu Che, Ling-Shu Wan and Zhi-Kang Xu selectively recognize PNA instead of Con A or BSA. The results suggest that the glycosylated nanofibers possess the capability to recognize the corresponding protein, which is strongly dependent on the specific carbohydrate-protein interaction. Furthermore, the adsorbed surface can be regenerated by incubation with high concentration of sugar solutions and then be reused. As a consequence, it is believed that the glycosylated nanofibers have potential applications in separation and purification of proteins.

Keywords: Polyacrylonitrile; glycosylated nanofiber; glycosylation; electrospinning, glycopolymer; surface modification; protein adsorption, specific recognition

1. INTRODUCTION Electrospinning is a simple and efficient approach to produce nanometer- to micrometerscale fibers and has attracted much attention in recent years. Generally, the process consists of three stages: (1) the jet is initiated, and forms a Taylor cone at the tip of the syringe and extends along a straight line; (2) the bending instability as proposed by Reneker (Reneker et al. 2000) occurs and further elongates in the jet path; (3) the jet is solidified into fibers on the collector. To better understand the electrospinning process and mechanisms, several theoretical models and a series of experimental confirmations were proposed by Hohman et al. (Hohman et al. 2001a; Hohman et al. 2001b) and Yarin et al. (Yarin et al. 2001). Up to now, more than 100 different polymers have been electrospun into ultrafine fibers with diameter ranging from tens of nanometers to a few micrometers (Li and Xia 2004). Because the electrospinning technique can be used to prepare materials with large surface area-tovolume ratios and high porosity, it has been considered as an effective method for specific functionality (Huang et al. 2006a), cell culture (Chua et al. 2005; Wang et al. 2005), enzyme immobilization (Huang et al. 2006b; Wang et al. 2006c), and affinity separation (Ma et al. 2005). In nature, carbohydrate always interacts with specific protein through multivalent interactions, namely two or more carbohydrates with a multimeric protein, which is accepted as “cluster glycoside effect” (Lee and Lee 1995; Lundquist and Toone 2002). In the past few decades, a great number of glycopolymers have been designed and synthesized to mimic the multivalent functions of natural glycoconjugates on the cell surface (Ladmiral et al. 2004; Okada 2001). However, it is well known that the random clustering of carbohydrates sometimes reduces the strength of affinity with guest proteins, which is ascribed to the steric hindrance of the multiple sugar branches. A surface tethered with sugar moieties (glycosylated surface) can effectively facilitate the specific adsorption of protein without much steric hindrance (Park and Shin 2002; Smith et al. 2003). It can be envisaged that electrospinning is a facile method to fabricate glycosylated nano-scale fibers that exhibit very high sugar content to the volume ratio and relatively low steric hindrance for protein binding. Polyacrylonitrile (PAN) is a kind of polymers with excellent physicochemical performances and film/fiber-forming properties. It has been successfully applied as a membrane material in the fields of ultrafiltration, dialysis and pervaporation (Godjevargova et al. 2000; Lin et al. 2004; Ray et al. 1999). PAN-based glycosylated nanofibers may be superior materials for

Glycosylated Nanofibers for Protein Adsorption and Recognition

273

studying the carbohydrate-protein interaction, where the PAN part contributes to the fiber formation and the carbohydrate residues serve as the functionality. Herein, two methods being used to fabricate glycosylated nanofibers are summarized, i.e., the synthesis of glycopolymers followed by electrospinning, and the surface modification of PAN-based nanofibers. Then the specific interactions between glycosylated nanofibers with different carbohydrate residues and the proteins are present. These glycosylated nanofibers may be applied in the fields of separation and purification.

2. FABRICATION OF GLYCOSYLATED NANOFIBERS FROM GLYCOPOLYMERS Two kinds of glycopolymers were synthesized for comparison in our work. One has cyclic glucose residues such as poly[acrylonitrile-co-(α-allyl glucoside)] (PANCAG) and poly{acrylonitrile-co-[2-(2-,3-,4-,6-tetra-O-acetyl-β-D-glucosyloxy) ethyl methacrylate]} (PANCAcGEMA). The other bears linear glucose residues such as poly[acrylonitrile-co-(Dgluconamidoethyl methacrylate)] (PANCGAMA).

2.1. Synthesis and Characterization of the Glycopolymers Because acrylonitrile (AN) can easily copolymerize with a variety of vinyl comonomers, the copolymerizations of AN with various comonomers such as maleic acid (Che et al. 2005; Nie et al. 2003; Nie et al. 2004a; Nie et al. 2004b; Nie et al. 2004c; Ye et al. 2005; Ye et al. 2006a; Ye et al. 2006b), N-vinyl-2-pyrrolidone (Wan et al. 2006a; Wan et al. 2005a; Wan et al. 2005b; Wan et al. 2006b), 2-hydroxyethyl methacrylate (Huang et al. 2006a; Huang et al. 2005a; Huang et al. 2005b; Huang et al. 2008), and acrylic acid (Che et al. 2008a; Che et al. 2008b; Che et al. 2006; Wang et al. 2007a; Wang et al. 2007b; Wang et al. 2006a; Wang et al. 2006b; Wang et al. 2006c) have been performed using free radical polymerization method in our laboratory. In this work, PAN-based glycopolymers containing glycomonomers such as α-allyl glucoside (AG), 2-(2-,3-,4-,6-tetra-O-acetyl-β-D-glucosyloxy) ethyl methacrylate (AcGEMA) and 2-gluconamidoethyl methacrylate (GAMA) were studied. AG was synthesized with the method reported by Talley et al. (Figure 1a) (Talley et al. 1945). It was incorporated into PAN by water phase precipitation copolymerization in our previous work (Kou et al. 2003; Xu et al. 2003). The structure of PANCAG was confirmed by 1 H NMR (Figure 2a). Effects of initiator concentration, reaction time and temperature, and AG concentration in feed on the copolymerization were discussed. Typical results are presented in Figure 3. It can be seen that the copolymerization of AN with AG behaved as a common free radical chain polymerization. At a constant 20 wt.% total monomer concentration, the yield (overall monomer conversion) increases rapidly with the initiator concentration and reaction time as well as reaction temperature at first and then levels off (Figure 3a-c). Increasing the total monomer concentration raises the yield remarkably at a constant initiator/monomer ratio (1/500) at 60 °C. On the other hand, the molecular weight of the resultant copolymer decreases with the initiator concentration and reaction temperature while it increases slightly with reaction time and total monomer concentration.

274

Ai-Fu Che, Ling-Shu Wan and Zhi-Kang Xu

Figure 1. Schemes for synthesis of glycomonomers: (a) AG; (b) AcGEMA; and (C) GAMA.

Figure 2. Chemical structures and typical 1H NMR spectra of PAN-based glycopolymers: (a) PANCAG; (b) PANCAcGEMA; and (C) PANCGAMA (Yang et al. 2006). Copyright © (2006). Reprinted with permission of John Wiley and Sons, Inc.

Glycosylated Nanofibers for Protein Adsorption and Recognition

275

Figure 3. Typical results for the water phase precipitation copolymerization of PANCAG. Effects of (a) initiator concentration ([AN]/[AG] mole ratio = 85/15, [Total monomers] = 0.20 g/mL, 60 °C, 6 h); (b) reaction time ([I]/[M] = 1/500, [AN]/[AG] = 85/15, [Total monomers] = 0.20 g/mL, 60 °C); (c) reaction temperature ([I]/[M] = 1/500, [AN]/[AG] = 85/15, [Total monomers] = 0.20 g/mL, 6 h); (d) AG concentration in feed, [Total monomers] = 0.20 g/mL, 60 °C, 6 h) (Xu et al. 2003). Copyright (2003) American Chemical Society

From these results, the optimal condition for PANCAG copolymerization in water has been obtained as follows: total monomer concentration of 20 g/100 mL water; initiator/monomer ratio of 1/500, reaction temperature of 60 °C and reaction time of 6 h. With increasing AG content, the water contact angle of the copolymer films decreases from 68.5° to 29.7° due to the contribution of AG moieties, as shown in Figure 4. The adsorption behavior of bovine serum albumin (BSA) and the adhesion of macrophage on the film surface were also measured (Figure 5). It can be seen that higher BSA concentration leads to a larger amount of BSA adsorbed on the copolymer films, and the adsorbed amount decreases significantly with the increase of AG content in the copolymer. The decrease of BSA adsorption can be mainly ascribed to the improvement of the hydrophilicity by the carbohydrate moieties on the copolymer film surface. Besides, the number of macrophages adhered on the PANCAG film also decreases in comparison with that on the PAN film, which implies better biocompatibility of the copolymer. However, it seems that the macrophage adhesion number increases again when the AG content in the copolymer exceeds around 20 wt. %. This may be due to the cluster effect of carbohydrate moieties.

276

Ai-Fu Che, Ling-Shu Wan and Zhi-Kang Xu

Figure 4. Relationship between the contact angle and AG content of PANCAG films at 25 °C (Xu et al. 2003). Copyright (2003) American Chemical Society.

Figure 5. (a) Adsorption of BSA and (b) relative macrophage number onto the PANCAG films at 25 °C (Xu et al. 2003). Copyright (2003) American Chemical Society.

Glycosylated Nanofibers for Protein Adsorption and Recognition

277

These results reveal that both the hydrophilicity and biocompatibility of PAN have been improved by the copolymerization of AN with vinyl carbohydrates, which facilitate the resistance to protein adsorption and cell adhesion. AcGEMA was synthesized using boron trifluoride diethyl etherate (BF3⋅Et2O) as the catalyst according to the literature (Craig Fleming et al. 2005), as shown in Figure 1b. Compared with other traditional methods, it can avoid the usage of the toxic metallic catalyst and ensure the specificity of configuration. PANCAcGEMA was then synthesized with solution polymerization method using AIBN as the initiator (Figure 2b). Typical results are shown in Table 1. The difference between water phase precipitation polymerization and solution polymerization was discussed in our previous work (Xu et al. 2003). The hydrophilicity of PANCAcGEMA film was characterized with water contact angle measurement. The film after a deprotection process in methanol sodium/methanol solution (PANCGEMA film) and poly(acrylonitrile-co-hydroxyethyl metharylate) (PANCHEMA) film were also prepared for comparison. The results in Figure 6 show that the initial contact angle of PANCGEMA film is a little lower, while those of other films present no significant differences. However, the contact angle of PANCGEMA film decreases more quickly with time in comparison with that of PANCAcGEMA film, which indicates that PANCGEMA film surface is relatively hydrophilic due to the exposure of glucose residues after the deprotection of acetyl groups. Table 1. Effect of the Monomer Feed Ratios on the Synthesis of PANCAcGEMA Copolymer

Samples

AN/AcGEMA

AcGEMA content (mol.%)

Yield (%)

[η] (dL/g)

Mη (10-4, g/mol)

PAN PANCAcGEMA2 PANCAcGEMA5 PANCAcGEMA10 PANCAcGEMA15

100:0 98:2 95:5 90:10 85:15

0 2.3 6.1 17.2 30.8

95.7 91.8 89.3 87.1 83.3

2.48 2.01 1.62 1.29 1.01

15.2 11.5 8.7 6.4 4.7

In addition, to study the effect of carbohydrate structures on the carbohydrate-protein interaction, glycommonomer (GAMA) bearing linear glucose residues was also synthesized (Figure 1c) (Yang et al. 2006). Subsequently, PANCGAMA was synthesized using water phase precipitation copolymerization method. The typical structure was demonstrated by 1H NMR (Figure 2c). Effects of initiator concentration, total monomer concentration, monomer feed ratio and reaction time on the copolymerization have been discussed, as shown in Figure 7. The hydrophilicity and biocompatibility of PANCGAMA film were also evaluated by water contact angle measurement and platelet adhesion, respectively. The results indicate that with the increase of GAMA content in the copolymer, the water contact angle decreases obviously and the platelet adhesion is strongly resisted, which suggest that the hydrophilicity and biocompatibility has been highly improved, as shown in Figure 8.

278

Ai-Fu Che, Ling-Shu Wan and Zhi-Kang Xu

Figure 6. Water contact angle measurement of different films with time.

Therefore, it is believed that the incorporation of GAMA is also a feasible method to improve the properties of PAN.

Figure 7. Typical results for the water phase precipitation copolymerization of PANCGAMA. Effects of (a) initiator concentration ([AN]/[GAMA] = 90/10, [Total monomer concentration] = 2 M, 60 oC, 6 h); (b) total monomer concentration ([I]/[M] = 1/500, [AN]/[GAMA] = 90/10, 60 oC, 6 h); (c) monomer feed ration ([I]/[M] = 1/500, [Total monomer concentration] = 2 M, 60 oC, 6 h); (d) reaction time ([I]/[M] = 1/500, [AN]/[GAMA] = 90/10, [Total monomer concentration] = 2 M, 60 oC).

Glycosylated Nanofibers for Protein Adsorption and Recognition

279

Figure 8. (A) Water contact angle measurement and (B) blood platelet adhesion on (a) PAN film and PANCGAMA films with GAMA feed ratios of (b) 5 mol.%; (c) 10 mol.%; and (d) 15 mol.%.

2.2. Fabrication of Glycosylated Nanofibers PAN-based glycopolymers can be fabricated into nanofibers by electrospinning, which have been described as “sugar sticks”. In general, the morphology of the nanofibers is strongly influenced by three kinds of parameters including (a) fluid properties such as type of polymer, molecular weight and molecular weight distribution, solution concentration and surface tension of the solvent (Lee et al. 2003; Shenoy et al. 2005; Sukigara et al. 2003; Wang and Kumar 2006; Zhao et al. 2005); (b) processing parameters such as feeding rate of polymer solution, applied electrostatic voltage, and tip-to-collector distance (Deitzel et al.

280

Ai-Fu Che, Ling-Shu Wan and Zhi-Kang Xu

2001; Sukigara et al. 2003; Supaphol et al. 2005), and (c) ambient conditions such as temperature and humidity (Demir et al. 2002). Figure 9 shows the SEM images of PANCAG, PANCAcGEMA and PANCGAMA nanofibers that were fabricated with a variety of electrospinning parameters in Table 2. The sugar contents in the copolymers were calculated from 1H NMR spectra.

Figure 9. Typical FESEM photographs of glycosylated nanofibers with a variety of electrospinning parameters (See Table 2): (A) PANCAG; (B) PANCAcGEMA; (C) PANCGAMA (Yang et al. 2006). Copyright © (2006). Reprinted with permission of John Wiley and Sons, Inc.

Glycosylated Nanofibers for Protein Adsorption and Recognition

281

Table 2. Effect of Electrospinning Parameters on the Morphology of Different Glycosylated Nanofibers (Yang et al. 2006). Copyright © (2006). Reprinted with permission of John Wiley and Sons, Inc

Samples

Molecular Weight (10-4, g/mol)

Sugar Content (mol.%)

PANCAG

13.7

1.37

PANCAcGEMA

8.7

6.1

11.5

2.3

11.9

5.53

10.5

11.92

PANCGAMA

Feeding Rate (mL/h)

12 12 16 16 16 16 12 12 8 12 12 12 16 16 12 16 16

Solution Concentration (wt.%) 4 6 6 6 6 6 12 15 15 15 15 15 4 6 6 6 6

16

6

Voltage (kV)

Fiber Diameter (nm)

Figures

0.3 0.3 0.3 0.3 0.6 1.0 0.6 0.6 0.6 0.6 1.0 0.6 0.3 0.3 0.3 0.3 0.5

Tip-toCollector Distance (cm) 15 15 15 20 15 15 15 15 15 20 15 15 15 15 15 20 15

95 ± 11 240 ± 18 219 ± 16 142 ± 14 154 ± 16 323 ± 21 162 ± 19 370 ± 26 445 ± 23 334 ± 17 410 ± 22 494 ± 25 Beads 89 ± 24 86 ± 21 82 ± 12 96 ± 10

9a 9b 9c 9d 9e 9f 9a’ 9b’ 9d’ 9d’ 9e’ 9f’ 9a’’ 9b’’ 9c’’ 9d’’ 9e’’

0.3

15

79 ± 12

9f’’

By adjusting the electrospinning parameters, sugar sticks with different morphologies were obtained. It is found that polymer solution concentration has great influence on the morphology of the fibers. With the increase of solution concentration, the fiber diameter increases obviously, e.g., for PANCAG, the fiber diameter is doubled with the increase in concentration from 4 to 6 wt. % (Figure 9a and b). Similar trend can also be found for PANCAcGEMA. However, for PANCGAMA under the same electrospinning conditions, when the solution concentration is low (4 wt. %, Figure 9a’’), the charged jet of polymer solution is unstable and likely to break up and form droplets, which could be ascribed to the relatively low molecular weight and high sugar content of PANCGAMA compared with PANCAG. Electrospinning voltage is another crucial factor for the morphology of glycosylated fibers. High voltage leads to an intense electric field and subsequently brings violent bending instability to the electrical jet resulting in a highly charged jet. For example, the fiber diameter of PANCAG decreases from 240 ± 18 to 219 ± 16 nm with the increase of electrospinning voltage (from 12 to 16 kV, Figure 9b and c) keeping other conditions constant, and that of PANCGAMA varies from 126 ± 9 to 79 ± 12 nm (Figure 9b’’ and c’’). However, we also observed that the fiber diameter started fluctuating at high voltage, which could be ascribed to the asymmetric dehiscence of the polymer solution.

282

Ai-Fu Che, Ling-Shu Wan and Zhi-Kang Xu

Besides, the tip-to-collector distance also has a great effect on the fabrication of glycosylated fibers. With increasing the tip-to-collector, the diameters of fibers tend to be thinner (Figure 9c and d). The feeding rate was also studied in our work. At a lower feeding rate, the fibers were a little thinner, but with a relatively broad diameter distribution; when the feeding rate increases, the fibers become thicker and uniform. This phenomenon primarily depends on the volume charge density of the jet, which may directly affect the morphology of the fibers. At a lower feeding rate, the volume charge density is higher so that the instability occurrs more easily and a broad diameter distribution is obtained. At a higher feeding rate, the positive and negative charges have no time to be separated completely and the jet fluid still carries residue charges making the volume charge density decrease and thus the diameter becomes thicker. Take PANCAG as an example, the diameter of glycosylated fibers becomes thicker from 142 ± 14 to 323 ± 21 nm with the increase of the feeding rate (from 0.3 to 1.0 mL/h, Figure 9d, e and f). In addition, the effect of sugar content on the fabrication of glycosylated fibers has also been discussed. Take PANCAcGEMAs with sugar contents of 2.3 and 6.1 mol. % as examples, under the same electrospinning condition, the diameters of PANCAcGEMA fibers are 494 ± 25 and 370 ± 26 nm, respectively, which can be ascribed to the stronger polarity of the copolymer with the increase of sugar content. Briefly, several parameters such as the solution concentration, electrospinning voltage, tip-to-collector distance, feeding rate, and sugar content in the copolymer have been confirmed to have great influences on the morphology of the glycosylated fibers in our work. Some rules on the basis of the experimental results are summarized in Table 3. It should be noted that these parameters are not independent of each other. To achieve a controllable fabrication of uniform and smooth glycosylated fibers, several parameters must be adjusted at the same time. Table 3. Summary of Different Electrospinning Parameters on the Morphology of PANBased Glycosylated Nanofibers Electrospinning Parameters Polymer concentration↑ Applied voltage↑ Feeding rate↑ Tip-to-collector distance↑ Sugar content↑

Effects of the Morphology of Nanofibers Fiber Diameter↑, bead defects at low concentration Fiber Diameter↓(broad diameter distribution) Fiber Diameter↑ Fiber Diameter↓ Fiber Diameter↓

3. GLYCOSYLATION OF NANOFIBERS THROUGH SURFACE MODIFICATION Surface modification is one of the commonly used methods for achieving the functionalization, which includes surface coating, plasma treatment, chemical reaction and UV radiation graft polymerization. In our work, glycosylated nanofibers have been constructed with carbohydrates such as chitosan and glucose on the electrospun nanofibers

Glycosylated Nanofibers for Protein Adsorption and Recognition

283

through chemical reactions to mimic the carbohydrate-protein interaction on the surface of cell membrane.

3.1. Preparation of PAN-Based Nanofibers As mentioned above, copolymerization is a superior method to synthesize PAN-based copolymers because it can greatly improve the hydrophilicity and biocompatibility of PAN materials and promote its further applications in biological field. Furthermore, it is facile to incorporate reactive groups into PAN for further coupling. Herein, we attempt to apply PANbased copolymers including poly(acrylonitrile-co-acrylic acid) (PANCAA) and PANCHEMA in the field of glycosylation. PANCAA and PANCHEMA were synthesized using solution polymerization and water phase precipitation polymerization, respectively (Che et al. 2008a; Che et al. 2008b; Huang et al. 2005a; Huang et al. 2005b). The copolymers with reactive groups varying from 0 to 20 mol. % can be obtained through optimiztion of the experimental conditions. Subsequently, PANCAA with AA of 15.3 mol.% and PANCHEMA with HEMA of 11.3 mol. % were selected for preparation of PAN-based nanofibers. Water contact angle results reveal that with the incorporation of the AA and HEMA, the hydrophilicity of PAN is relatively enhanced. Take PANCHEMA as an example, the water contact angle of PANCHEMA with 17 mol.% HEMA decreases to 53° in comparison with that of PAN (ca. 64°). Blood platelet adhesion and cell adhesion on the copolymer film are resisted, which indicates the improvement of hydrophilicity and biocompatibility of the copolymer by the introduction of carboxylic groups and hydroxyl groups. Typically, the number of cell adhered on PANCHEMA film with ~15 mol. % HEMA is half of that on PAN film (~700 cell/mm2).

3.2. Preparation of PAN-Based Nanofibers Several electrospinning parameters have been discussed for the preparation of PANbased nanofibers. It is found that concentration of polymer solution, feeding rate and electric field strength play essential roles in modulating the morphology of the nanofibers. Take PANCAA as an example, we studied the effects of electrospinning parameters on the fabrication of PAN-based electrospun nanofibers. As can be seen from Figure 10A, when the concentration of PANCAA is lower (2 wt. %), a composite of thinner nanofibers and bigger beads is formed. With increasing the concentration, the shape of the beads changes from spherical to spindle-like, and continuous and smooth nanofibers are obtained at 4 wt. % or above and the diameter of the fibers increases obviously. This is mainly attributed to the competition between the surface tension and viscoelastic force of the polymer solution. To clarify the effect of polymer solution concentration, viscosity has been measured. It is found that when the concentration is lower, the viscosity is also lower, as shown in Figure 10B, bead defects are thus formed because of the low molecular chain entanglement of PANCAA solution that is insufficient to prevent the breakup of the charged jet. With the increase in the concentration, the viscosity increases substantially so that the electrically driven jet cannot break up into small droplets, and thus uniform fibers are formed.

284

Ai-Fu Che, Ling-Shu Wan and Zhi-Kang Xu

Figure 10. (A) Typical FESEM photographs of nanofibers electrospun from PANCAA solutions with different concentrations: (a) 2 wt.%; (b) 3 wt.%; (c) 4 wt.%; (d) 5 wt.%; (e) 6 wt.%; and (B) the corresponding diameters of fibers and solution viscosities (Che et al. 2008). Copyright (2008), with permission from CSIRO

The feeding rate and the electric field strength also have influences on the morphology of the fibers. At low feeding rate and high electric field strength (a ratio of the electrospinning voltage to the tip-to-collector), thinner fibers are easily deposited onto the collector due to the bending instability caused by the high charged density of the polymer solution at the tip. The rules of electrospinning parameters in the fabrication of PANCAA nanofibers are similar to those listed in Table 3. To prepare uniform, smooth and ultrathin PANCAA nanofibers for further carbohydrate modification, the conditions are optimized to be solution concentration of 5 wt. %, feeding rate of 1.0 mL/h, electric field strength of 12.0 kV/15 cm at 28 oC with the humidity ca. 35 RH%. In this case, PANCAA nanofibers with average diameter of 190 ± 23 nm were obtained.

Glycosylated Nanofibers for Protein Adsorption and Recognition

285

Figure 11. Typical FESEM images of PANCHEMA nanofibers with various concentrations: (a) 1 wt.%; (b) 2 wt.%; (c) 3 wt.%; (d) 4 wt.%.

In addition, several parameters for the preparation of PANCHEMA nanofibers were also discussed. It is found that the morphology of the PAN-based nanofibers is strongly dependent on the concentration of PANCHEMA polymer solution. As shown in Figure 11, with increasing the concentration, the morphology changes correspondingly from discontinuous fibers to spindle-like fibers with bead defects and to continuous and uniform nanofibers, which is similar to that of PANCAA nanofibers. Finally, PANCHEMA nanofibers with of 198 ± 28 nm could be acquired under conditions of solution concentration of 4 wt. %, feeding rate of 1.0 mL/h, electric field strength of 12.0 kV/15 cm at room temperature with the humidity 40-50 RH%.

3.3. Surface Modification of PAN-Based Nanofibers Because of the excellent mechanical and physicochemical properties of PAN-based copolymers and the large area-to-volume ratio of electrospun nanofibers, surface modification of PAN-based nanofibers was performed to achieve carbohydrate immobilization in our work. According to the reactive groups of PAN-based copolymers, various chemical reactions have been designed to achieve the carbohydrate immobilization.

286

Ai-Fu Che, Ling-Shu Wan and Zhi-Kang Xu

PANCAA nanofibers that possess a large amount of carboxylic groups on the surface are used as the matrices for chitosan immobilization. The glycosylated surface has been fabricated through a coupling reaction between the carboxylic groups of PANCAA and the primary amino groups of chitosan using carbodiimide hydrochloride/N-hydroxysuccinimide (EDC/NHS) as the coupling agent. The results show that the coupling degree of chitosan (CDC) is strongly dependent on the reaction conditions, e.g. EDC concentration and reaction time. With increasing EDC concentration or extending reaction time, the CDC is highly increased, as listed in Table 4. The corresponding chemical structures of the nanofibers before and after the coupling were confirmed with attenuated total reflectance Fourier transform infrared spectroscopy (FT-IR/ATR), X-ray photoelectron spectroscopy (XPS). It is found that two new striking characteristic peaks appear at 1566 cm-1 and 1621 cm-1 in Figure 12A, which are caused by N-H bending vibration of -NH2 groups and that of -CONH- groups after chitosan coupling, respectively. In addition, the intensities of C 1s and N 1s become weaker, whereas the intensity of O 1s gradually increases compared with the native PANCAA nanofibers (Figure 12B), which also confirms the occurrence of the coupling reaction. Table 4. The Coupling Degrees of Chitosan (CDCs) on PANCAA Nanofibers (Che et al. 2008). Copyright (2008) American Chemical Society Samplea

Reaction Conditions

PANCAA-CS1 PANCAA-CS2 PANCAA-CS3

Concentration of chitosan (mg/mL) 5 10 10

Reaction Time (h) 20 20 44

CDC (wt %)

Contact Angles (o)

6.62 ± 0.98 10.8 ± 1.21 20.7 ± 1.36

65.6 ± 1.7 64.5 ± 1.8 63.8 ± 1.5

a

: PANCAA-CS1, 2, 3 represent the chitosan-modified PANCAA nanofibers with different CDCs.

The morphological changes of the nanofibers before and after the coupling were evaluated with field-emission scanning electron microscopy (FESEM), as shown in Figure 13. Compared with the native PANCAA sample, the modified nanofibers become a bit looser. However, the diameters do not change much with single fiber remaining clearly. The results indicate that the electrospun PANCAA nanofiber is a superior matrix for chitosan modification. PANCHEMA nanofibers with a large amount of hydroxyl groups are used as the matrices for glucose immobilization through the chemical reaction between the hydroxyl groups and glucose pentaacetate using BF3⋅Et2O as the catalyst followed by the deprotection of acetyl groups as shown in Figure 14 (PAH-g-GLC). Quartz crystal microbalance (QCM) was used to monitor the glycosylation process on the surface of PANCHEMA thin film deposited on the quartz crystal under the same condition with that of the PANCHEMA nanofibers.

Glycosylated Nanofibers for Protein Adsorption and Recognition

287

Figure 12. (A) FT-IR/ATR spectra and (B) XPS spectra of the PANCAA nanofibers before and after chitosan coupling: (a) PANCAA; (b) PANCAA-CS1; (c) PANCAA-CS2; (d) PANCAA-CS3; (e) chitosan (Che et al. 2008). Copyright (2008) American Chemical Society.

288

Ai-Fu Che, Ling-Shu Wan and Zhi-Kang Xu

Figure 13. Typical FESEM images of the PANCAA nanofibers before and after chitosan coupling: (a) PANCAA; (b) PANCAA-CS1; (c) PANCAA-CS2; (d) PANCAA-CS3 (Che et al. 2008). Copyright (2008) American Chemical Society

The immobilization degree of glucose on the PANCHEMA film surface is estimated to be 1.75 ± 0.09 μg/cm2 according to the frequency changes according to Sauerbrey’s equation (Sauerbrey 1959). FESEM photographs show that compared with the native PANCHEMA nanofibers, the glycosylated nanofibers present a slight adhesion between each other after glycosylation, which may be attributed to long-time immersion in the organic solvent during the reaction process and the strong interaction among the glucose residues, as presented in Figure 15.

Figure 14. Schematic representation for surface modification of PANCHEMA with glucose.

Glycosylated Nanofibers for Protein Adsorption and Recognition

289

Figure 15. FESEM photographs of the PANCHEMA nanofibers before and after glucose immobilization: (a) PANCHEMA; (b) PAH-g-GLC.

4. PROTEIN ADSORPTION AND RECOGNITION BY THE GLYCOSYLATED NANOFIBERS 4.1. Interactions between Glycosylated Nanofibers and the Specific Protein In our work, we have prepared five kinds of glycosylated nanofibers using the methods in sections 2 and 3, which are enriched with different carbohydrate residues on the surface of nanofibers, as summarized in Table 5. According to literatures, different carbohydrates can interact with the specific protein (Goldstein et al. 1965; Poretz and Goldstein 1970). Herein, we focus on the study of the protein with the carbohydrate residues on the glycosylated nanofibers. Lectin is one kind of glycoproteins from many sources such as plants, animals and microorganisms, which are involved in diverse biological processes such as cell interaction in immune system, adhesion of infectious agents to host cells, blood-clotting cascade, and recruitment of leukocytes to inflammatory sites. Because lectins can selectively and reversibly bind with specific mono- and oligosaccharides, they are often selected as the probes to study carbohydrate-protein interaction. Among them, concanavalin A (Con A) is one of the most commonly used lectins, which shows strong affinity to D-mannose and Dglucose as well as N-acetyl-D-glucosamine (Goldstein et al. 1965; Hong et al. 2001; Lee and Lee 1995; Li et al. 2000; Morimoto et al. 2001). PANCAG nanofibers was dipped in Con A/BSA mixture solution for 30 min at 25 oC and then isolated by centrifugation. The relative concentrations of Con A and BSA in the feed and resultant solutions were determined by electrophoresis and the typical results are shown in Table 6. It is found that PANCAG nanofibers show strong binding to Con A that the concentration of Con A decreases from 60.30 to 34.47 mol. % with very low sugar content (1.37 mol. %) in the nanofibers. This is ascribed to the specific interaction between Con A and glucose residues on PANCAG nanofibers. The concentration of Con A remains almost the same when the sugar content increases to 3.37 mol.% and then, a slight rise in sugar content from 3.37 to 3.91 mol. % results in the complete adsorption of Con A. This may be ascribed to the “cluster glucoside effect”, i.e., when the sugar density exceeds the critical value, the affinity between the glycoligand and the protein increases tremendously. Further increase in sugar content (5.22 mol. %) for PANCAG gives the same result in the recognition of Con A from the protein mixture.

290

Ai-Fu Che, Ling-Shu Wan and Zhi-Kang Xu

Table 5. Interactions between Proteins and Different Carbohydrate Residues on the Glycosylated Nanofibers

1 2 3 3’ 4 5

Glycosylated Nanofibers PANCAG PANCGEMA PAH-g-GLC PAH-g-GAL PANCGAMA PANCAA-gCS

Carbohydrate Residues

Proteins

Kinds of Interaction

α-Glucose β-Glucose β-Glucose β-Galactose Linear glucose Chitosan

Con A Con A Con A PNA None Con A

Specific recognition Specific recognition Specific recognition Specific recognition --Electrostatic interaction and specific recognition

Table 6. Effects of Sugar Moiety Species and Contents on the Protein Isolation Properties (Yang et al. 2006). Copyright © (2006). Reprinted with permission of John Wiley and Sons, Inc Sample PANCAG5 PANCAG10 PANCAG15 PANCAG20 PANCGAMA5 PANCGAMA10 PANCGAMA15

Fiber Diameter (nm) 95±11 80±14 88±14 98±16 82±14 79±12 75±28

Sugar Content (mol.%) 1.37 3.37 3.91 5.22 5.53 11.92 16.85

Con A (mol.%)

BSA (mol.%)

34.47 33.69 0 0 55.99 45.92 42.51

65.53 66.31 100 100 44.01 54.08 57.49

Figure 16. FESEM images of PANCAcGEMA nanofibers before and after deprotection: (a) PANCAcGEMA5; (b) PANCGEMA5; (c) PANCAcGEMA15; and (d) PANCGEMA15.

Glycosylated Nanofibers for Protein Adsorption and Recognition

291

Figure 17. CLSM photographs of PANCHEMA nanofibrous sugar sticks after protein adsorption: (a) PANCGEMA2, Con A; (b) PANCGEMA5, Con A; (c) PANCGEMA10, Con A; (d) PANCGEMA15, Con A; (e) PANCAcGEMA5, Con A; (f) PANCGEMA10, PNA.

On the contrary, the PANCGAMA nanofibers show very low or even no affinity to Con A. It can be explained by the fact that only the pyranose ring formed participates in the interaction with Con A (Goldstein et al. 1965). However, a slight decrease in Con A concentration is still found in the case of PANCGAMA, which could be ascribed to the nonspecific adsorption of Con A on the PANCGAMA nanofibers. PANCGEMA nanofibers with a large amount of β-glucose residues on the surface were obtained by the deprotection of acetyl groups from PANCAcGEMA nanofibers. It seems that the deprotected PANCGEMA nanofibers (Figure 16b) appears a bit rougher and adhere with each other compared with the PANCAcGEMA nanofibers. When the sugar content of the glycosylated copolymers increases to 30.8 mol. %, the adhesion becomes more obvious, as shown in Figure 16d. This result may be attributed to the strong interaction of glucose residues on the nanofibers surface, which is similar to that of PAH-g-GLC nanofibers in section 3.2. FL-Con A (fluorescein FITC labeled Con A) was also used to evaluate the affinity interaction between Con A and glucose residues on PANCGEMA nanofibers. The confocal laser scanning microscopy (CLSM) photographs show that there is weak interaction between Con A and glucose residues at low sugar content (2.3 mol. %, Figure 17a). With increasing the sugar content from 6.1 to 17.2 mol. %, more adsorbed amount of FL-Con A indicate the occurrence of strong interaction. Further increasing the sugar content to 30.8 mol. %, the interaction become stronger correspondingly (Figure 17d). In addition, PANCAcGEMA nanofibers are also evaluated for comparison. It is found that the protein adsorbed on the surface can be negligible, which implies that glucose residues play essential roles in the specific interaction but pentaacetate glucose residues do not work. Moreover, fluorescein labeled peanut agglutinin (FL-PNA) that can selectively recognize galactose residues is used as a reference fluorescent glycoprotein to evaluate the specific recognition capability of PANCGEMA nanofibers. The results show that no interaction occurs between

292

Ai-Fu Che, Ling-Shu Wan and Zhi-Kang Xu

PNA and the glucose residues, which provides good evidence for the specificity of PANCGEMA nanofibers towards Con A. Similar with PANCGEMA nanofibers, PAH-g-GLC nanofibers with β-glucose residues on the surface of PANCHEMA nanofibers are prepared for the study of protein recognition. Besides, PAH-g-GAL nanofibers with β-galactose residues are also fabricated for comparison using the same methods with PAH-g-GLC nanofibers in section 3.2. PANCHEMA, PAH-gGLC and PAH-g-GAL (PANCHEMA nanofiber modified with galacose residues) nanofibers that were prepared under the same condition are applied for FL-Con A and FL-PNA adsorption, respectively. The corresponding relative fluorescent intensities on the surface measured by CLSM are illustrated in Figure 18A. It is found that PANCHEMA nanofibers present no affinity with neither of Con A or PNA due to the improvement of the hydrophilicity and biocompatibility after the incorporation of HEMA unit. PAH-g-GLC nanofibers show strong affinity with Con A but nearly no affinity with PNA, while PAH-gGAL nanofibers give relatively strong affinity with PNA but not Con A. The results suggest that the glycosylated PANCHEMA nanofibers have highly selective protein recognition capability, which is strongly dependent on the kinds of carbohydrate residues on the surface.

Figure 18. Relative fluorescent intensity of PANCHEMA and the glycosylated PANCHEMA nanofibers for protein recognition.

Generally, chitosan is a kind of cationic polysaccharide that processes several prominent characteristics like hydrophilicity, biocompatibility, biodegradability and anti-bacterial property. Chitosan modified PANCAA nanofibers (PANCAA-g-CS) were also studied for protein adsorption as schematically shown in Figure 19. Fluorescent images and relative fluorescence intensities by CLSM were measured to evaluate FL-Con A adsorption on PANCAA-g-CS nanofibers. From the high-resolution images in Figure 20, it is found that PANCAA-g-CS nanofibers exhibit green color when excited at 488 nm, which confirms the affinity interaction between chitosan and Con A.

Glycosylated Nanofibers for Protein Adsorption and Recognition

293

Figure 19. Schematic representation for the preparation and protein adsorption of PANCAA-g-CS nanofibers (Che et al. 2008). Copyright (2008) American Chemical Society.

Figure 20. CLSM images of PANCAA-g-CS nanofibers with FL-Con A adsorption at pH 5.3: (a) PANCAA; (b) PANCAA-CS1; (c) PANCAA-CS2; and (d) PANCAA-CS3 (Che et al. 2008). Copyright (2008) American Chemical Society

294

Ai-Fu Che, Ling-Shu Wan and Zhi-Kang Xu

It should be noted that the slight adsorption of Con A on PANCAA nanofibers (Figure 20a) could be attributable to the nonspecific interaction. With the increase of CDC, the adsorbed amount increases correspondingly because of the stronger interaction between chitosan and Con A, which can be demonstrated quantitatively through the relative fluorescence intensity measurement, as illustrated in Figure 21. As we know, the pKa value of chitosan is 6.3-7.0 (Claesson and Ninham 1992), and the isoelectric point of Con A is about 5.0. It means that at pH 5.3, chitosan is positively charged and can interact with negatively charged Con A through strong electrostatic interaction. On the other hand, the electrostatic interaction is much weakened at pH 7.5, thus in this case the specific recognition interaction between N-acetyl-D-glucosamine groups and Con A plays the major role, as confirmed in the literatures (Banerjee and Chen 2007; Farina and Wilkins 1980). Therefore, it is believed that the electrostatic interaction is crucial at pH 5.3, while the specific recognition is dominant at pH 7.5.

Figure 21. Comparison of the mean fluorescence intensities after FL-Con A adsorption on PANCAA-gCS nanofibers at different pHs (Che et al. 2008). Copyright (2008) American Chemical Society

4.2. Applications of The Glycosylated Nanofibers in Protein Separation Different glycosylated surfaces have been constructed in our work. The selective affinity interactions of the glycosylated nanofibers with the specific protein have been successfully demonstrated, as presented in Table 5. What is more, we hope to apply these glycosylated nanofibers in the separation and purification of proteins based on the specific interaction. Take PAH-g-GLC nanofibers as an example, applications of glycosylated nanofibers are preliminarily studied.

Glycosylated Nanofibers for Protein Adsorption and Recognition

295

Figure 22. Con A and BSA adsorption on PANCHEMA and glycosylated nanofibers at 25 oC.

The adsorption of Con A with concentrations ranging from 0.0125 to 0.2 mg/mL on PANCHEMA-based nanofibers was measured by ultraviolet spectrophotometer, as shown in Figure 22. It is found that a large amount of Con A is adsorbed on PAH-g-GLC nanofibers and the adsorbed amount increases with the Con A concentration, while almost no adsorption occurs on PANCHEMA and PAH-g-PAG nanofibers. In addition, BSA is employed as a reference protein to evaluate the nonspecific adsorption on the glycosylated surface. The little adsorbed amount on PAH-g-GLC nanofibers implies that the nonspecific adsorption of Con A can be negligible. That is to say, Con A adsorption on PAH-g-GLC nanofibers is completely attributed to the specific interaction. To further estimate the association saturation constant K a , the affinity binding of Con A to the glycosylated PAH-g-GLC nanofibers was measured by saturation analysis based on Langmuir adsorption equation (Banerjee and Chen 2007; Farina and Wilkins 1980) as follows:

C C 1 1 = + ⋅ q ∗ qmax qmax K a Wherein, q ∗ is the amount of adsorbed Con A, mg/g-fiber; qmax is the maximum amount of adsorbed Con A at saturation, mg/g-fiber; C is Con A concentration in the buffer solution, M; and K a is the association saturation constant, M-1, which is obtained by the ratio of slope to intercept according to the Scatchard plot of

C ∼C . q*

296

Ai-Fu Che, Ling-Shu Wan and Zhi-Kang Xu

Figure 23. (a) Langmuir isothern curve and (b) the corresponding Scatchard plot.

As shown in Figure 23b, the Scatchard plot gives a linear plot, indicating that the Langmuir adsorption model adequately describes Con A adsorption on PAH-g-GLC nanofibers. Consequently, K a was calculated as 2.20 × 105 M-1 with the ratio of slope to intercept based on the plot, which is similar to the results as reported in the literatures (Ebara and Okahata 1994; Huang et al. 2007; Yu et al. 2007). The value of K a confirms the presence of strong multivalent interactions (cluster glycoside effect) between Con A and glucose residues on PAH-g-GLC nanofibers, suggesting the successful construction of a multivalent glycosylated surface. As mentioned above, the glycosylated surface exhibited strong multivalent interactions with Con A as a result of “glucose cluster effect”. Generally, an ideal glycosylated surface to mimic carbohydrate functions should be reproducible for reuse. Herein, high concentration of D-glucose (1 M) was used to remove the adsorbed Con A on the PAH-g-GLC nanofibers. It is

Glycosylated Nanofibers for Protein Adsorption and Recognition

297

found that the desorbed PAH-g-GLC nanofibers can rebind Con A and the adsorbed amount is almost the same to that of the first use, which suggests that the glycosylated surface can be completely restored. After repeated use for several times, the PAH-g-GLC nanofibers still have the adsorption capability and remain the morphology without any changes, as shown in Figure 24. The complete regeneration of the glycosylated surface encourages its potential applications in many fields, especially in protein separation and purification. Furthermore, dynamic filtration experiments are undergoing in our laboratory.

Figure 24. (a) The repeatability of PAH-g-GLC nanofibers for Con A adsorption and (b) the FESEM images of PAH-g-GLC nanofibers after repeated use for several times.

5. CONCLUSIONS Two approaches have been used to fabricate the glycosylated nanofibers in our laboratory, i.e. synthesis of glycopolymers followed by electrospinning and surface modification of PAN-based nanofibers. Different electrospinning parameters such as concentration of polymer solution, feeding rate, electrical voltage, tip-to-collector distance and sugar content have been explored to modulate the morphology of the nanofibers.

298

Ai-Fu Che, Ling-Shu Wan and Zhi-Kang Xu

Eventually, five kinds of glycosylated nanofibers with different carbohydrate residues were successfully prepared. It is demonstrated that the hydrophilicity and biocompatibility are greatly improved due to the incorporation of hydrophilic carbohydrate residues, as confirmed by the water contact angle measurement, blood platelet adhesion and cell adhesion. Con A, PNA and BSA were used to evaluate the protein adsorption and recognition capability of these glycosylated nanofibers on the basis of the specific interaction between the carbohydrates and the corresponding protein. The results show that the glycosylated nanofibers with glucose residues such as PANCAG, PANCGEMA, and PAH-g-GLC can bind Con A but present no affinity with PNA or BSA, while the glycosylated nanofibers with galactose residues, e.g., PAH-g-GAL, can selectively recognize PNA, which confirms the specific recognition capability of these glycosylated nanofibers. By contrary, PANCGAMA nanofibers with linear glucose residues show almost no affinity with the proteins, which implies that cyclic carbohydrate is required for the protein recognition. In addition, the nanofibers modified with chitosan have been prepared for protein adsorption. It is found that PANCAA-g-CS nanofibers show strong affinity with Con A due to the electrostatic interaction and the specific recognition. Furthermore, the association saturation constant K a was measured to based on Langmuir adsorption model, e.g., K a of PAH-g-GLC nanofibers with Con A was measured to be 2.20 × 105 M-1, which indicates the presence of multivalent interaction (“cluster glycoside effect”). Last but not least, applications of the glycosylated nanofibers in separation and purification of proteins have been preliminarily studied and further experiments will be carried out in our laboratory.

ACKNOWLEDGMENTS Financial support from the National Natural Science Foundation of China (grant no. 20774080) and the National Natural Science Foundation of China for Distinguished Young Scholars (to Prof. Z. -K. Xu, grant no. 50625309) is gratefully acknowledged. The authors thank Dr. Qian Yang, Jun-Jie Li, and Li-Li Wu very much for their contribution.

REFERENCES Banerjee SS, Chen D-H (2007) Glucose-grafted gum arabic modified magnetic nanoparticles: Preparation and specific interaction with concanavalin A. Chem. Mater. 19:3667-3672. Che AF, Liu ZM, Huang XJ, Wang ZG, Xu ZK (2008a) Chitosan-modified poly(acrylonitrile-co-acrylic acid) nanofibrous membranes for the immobilization of concanavalin A. Biomacromolecules 9:3397-3403. Che AF, Huang XJ, Wang ZG, Xu ZK (2008b) Preparation and surface modification of poly(acrylonitrile-co-acrylic acid) electrospun nanofibrous membranes. Aust. J. Chem. 61:446-454. Che AF, Nie FQ, Huang XD, Xu ZK, Yao K (2005) Acrylonitrile-based copolymer membranes containing reactive groups: Surface modification by the immobilization of biomacromolecules. Polymer 46:11060-11065.

Glycosylated Nanofibers for Protein Adsorption and Recognition

299

Che AF, Yang YF, Wan LS, Wu J, Xu ZK (2006) Molecular imprinting fibrous membranes of poly(acrylonitrile-co-acrylic acid) prepared by electrospinning. Chem. Res. Chinese Univ. 22:390-393. Chua KN, Lim WS, Zhang P, Lu H, Wen J, Ramakrishna S, Leong KW, Mao HQ (2005) Stable immobilization of rat hepatocyte spheroids on galactosylated nanofiber scaffold. Biomaterials 26:2537-2547. Claesson PM, Ninham BW (1992) pH-Dependent interactions between adsorbed chitosan layers. Langmuir 8:1406-1412. Craig Fleming, Maldjian A, Costa DD, Rullay AK, Haddleton DM, John JS, Penny P, Noble RC, Cameron NR, Davis BG (2005) A carbohydrate-antioxidant hybrid polymer reduces oxidative damage in spermatozoa and enhances fertility. Nat. Chem. Bio. 1:270-274. Deitzel JM, Kleinmeyer J, Harris D, Tan NCB (2001) The effect of processing variables on the morphology of electrospun nanofibers and textiles. Polymer 42:261-272. Demir MM, Yilgor I, Yilgor E, Erman B (2002) Electrospinning of polyurethane fibers. Polymer 43:3303-3309. Ebara Y, Okahata Y (1994) A kinetic study of concanavalin A binding to glycolipid monolayers by using a quartz-crystal microbalance. J. Am. Chem. Soc. 116:11209-11212. Farina RD, Wilkins RG (1980) Kinetics of Interaction of Some α- and β-D-Monosaccharides with Concanavalin A. Biochim. Biophys. Acta 631:428-438. Godjevargova T, Konsulov V, Dimov A, Vasileva N (2000) Behavior of glucose oxidase immobilized on ultrafiltration membranes obtained by copolymerizing acrylonitrile and N-vinylimidazol. J. Membr. Sci. 172:279-285. Goldstein IJ, Hollerman CE, Smith EE (1965) Protein-carbohydrate interaction. Ⅱ. Inhibition studies on the interaction of concanavalin A with polysaccharides. Biochemisty 4:876883. Hohman MM, Shin M, Rutledge G, Brenner MP (2001a) Electrospinning and electrically forced jets. I. Stability theory. Phys. Fluids 13:2201-2220. Hohman MM, Shin M, Rutledge G, Brenner MP (2001b) Electrospinning and electrically forced jets. II. Applications. Phys. Fluids 13:2221-2236. Hong MF, Cassely A, Mechref Y, Novotny MV (2001) Sugar-lectin interactions investigated through affinity capillary electrophoresis. J. Chromatogr. B 752:207-216. Huang MC, Shen ZH, Zhang YL, Zeng XQ, Wang PG (2007) Alkanethiol containing glycopolymers: A tool for the detection of lectin binding. Bioorg. Med. Chem. Lett. 17:5379-5383. Huang XJ, Xu ZK, Huang XD, Wang ZG, Yao K (2006a) Biomimetic surface modification on polyacrylonitrile-based asymmetric membranes via direct formation of phospholipid moieties. Polymer 47:3141-3149. Huang XJ, Xu ZK, Wan LS, Innocent C, Seta P (2006b) Electrospun nanofibers modified with phospholipid moieties for enzyme immobilization. Macromol. Rapid Commun. 27:1341-1345. Huang XJ, Xu ZK, Wan LS, Wang ZG, Wang JL (2005a) Novel acrylonitrile-based copolymers containing phospholipid moieties: Synthesis and characterization. Macromol. Biosci. 5:322-330.

300

Ai-Fu Che, Ling-Shu Wan and Zhi-Kang Xu

Huang XJ, Xu ZK, Wan LS, Wang ZG, Wang JL (2005b) Surface modification of polyacrylonitrile-based membranes by chemical reactions to generate phospholipid moieties. Langmuir 21:2941-2947. Huang XJ, Yu AG, Xu ZK (2008) Covalent immobilization of lipase from Candida rugosa onto poly(acrylonitrile-co-2-hydroxyethyl methacrylate) electrospun fibrous membranes for potential bioreactor application. Bioresource Technol. 99:5459-5465. Kou RQ, Qu C, Xu ZK, Xu YY, Yao K (2003) Reducing nonselective protein adsorption and cell adhesion on polyacrylonitrile films by copolymerization of acrylonitrile with α-allyl glucoside. Chinese J. Polym. Sci. 21:373-375. Ladmiral V, Melia E, Haddleton DM (2004) Synthetic glycopolymers: an overview. Eur. Polym. J. 40:431-449. Lee KH, Kim HY, Bang HJ, Jung YH, Lee SG (2003) The change of bead morphology formed on electrospun polystyrene fibers. Polymer 44:4029-4034. Lee YC, Lee RT (1995) Carbohydrate-protein interactions: Basis of glycobiology. Acc. Chem. Res. 28:321-327. Li D, Xia Y (2004) Electropinning of nanofibers: Reinventing the wheel? Adv. Mate.r 16:1151-1170. Li XB, Tushima Y, Morimoto M, Saimoto H, Okamoto Y, Minami S, Shigemasa Y (2000) Biological activity of chitosan-sugar hybrids: Specific interaction with lectin. Polym. Adv. Technol. 11:176-179. Lin WC, Liu TY, Yang MC (2004) Hemocompatibility of polyacrylonitrile dialysis membrane immobilized with chitosan and heparin conjugate. Biomaterials 25:1947-1957. Lundquist JJ, Toone EJ (2002) The cluster glycoside effect. Chem. Rev. 102:555-578. Ma ZW, Kotaki M, Ramakrishna S (2005) Electrospun cellulose nanofiber as affinity membrane. J. Membr. Sci. 265:115-123. Morimoto M, Saimoto H, Usui H, Okamoto Y, Minami S, Shigemasa Y (2001) Biological activities of carbohydrate-branched chitosan derivatives. Biomacromolecules 2:11331136. Nie FQ, Xu ZK, Huang XJ, Ye P, Wu J (2003) Acrylonitrile-based copolymer membranes containing reactive groups: Surface modification by the immobilization of poly(ethylene glycol) for improving antifouling property and biocompatibility. Langmuir 19:98899895. Nie FQ, Xu ZK, Wan LS, Ye P, Wu J (2004a) Acrylonitrile-based copolymers containing reactive groups: synthesis and preparation of ultrafiltration membranes. J. Membr. Sci. 230:1-11. Nie FQ, Xu ZK, Yang Q, Wu J, Wan LS (2004b) Surface modification of poly(acrylonitrileco-maleic acid) membranes by the immobilization of poly(ethylene glycol). J. Membr. Sci. 235:147-155. Nie FQ, Xu ZK, Ye P, Wu J, Seta P (2004c) Acrylonitrile-based copolymer membranes containing reactive groups: effects of surface-immobilized poly(ethylene glycol)s on antifouling properties and blood compatibility. Polymer 45:399-407. Okada M (2001) Molecular design and syntheses of glycopolymers. Prog. Polym. Sci. 26:67104. Park SJ, Shin IJ (2002) Fabrication of carbohydrate chips for studying protein-carbohydrate interactions. Angew Chem. Int. Ed. 41:3180-3182.

Glycosylated Nanofibers for Protein Adsorption and Recognition

301

Poretz RD, Goldstein IJ (1970) An examination of the topography of the saccharide binding sites of concanavalin A and of the forces involved in complexation. Biochemistry 9:28902896. Ray SK, Sawant SB, Joshi JB, Pangarkar VG (1999) Methanol selective membranes for separation of methanol-ethylene glycol mixtures by pervaporation. J. Membr. Sci. 154:113. Reneker DH, Yarin AL, Fong H, Koombhongse S (2000) Bending instability of electrically charged liquid jets of polymer solutions in electrospinning. J. Appl. Phys. 87:4531-4547. Sauerbrey GZ (1959) Use of quartz vibration for weighing thin films on a microbalance. Z. Phys. 155:206–212. Shenoy SL, Bates WD, Frisch HL, Wnek GE (2005) Role of chain entanglements on fiber formation during electrospinning of polymer solutions: good solvent, non-specific polymer-polymer interaction limit. Polymer 46:3372-3384. Smith EA, Thomas WD, Kiessling LL, Corn RM (2003) Surface plasmon resonance imaging studies of protein-carbohydrate interactions. J. Am. Chem. Soc. 125:6140-6148. Sukigara S, Gandhi M, Ayutsede J, Micklus M, Ko F (2003) Regeneration of Bombyx mori silk by electrospinning-part Ⅰ: Processing parameters and geometric properties. Polymer. 44:5721-5727. Supaphol P, Mit-Uppatham C, Nithitanakul M (2005) Ultrafine electrospun polyamide-6 fibers: Effect of emitting electrode polarity on morphology and average fiber diameter. J. Polym. Sci, Part B: Polym Phys 43:3699-3712. Talley EA, Vale MD, Yanovsky E (1945) Allyl ethers of carbohydrates. Ⅲ. Ethers of glucose and galactose. J. Am. Chem. Soc. 67:2037. Wan LS, Xu ZK, Huang XJ, Che AF, Wang ZG (2006a) A novel process for the posttreatment of polyacrylonitrile-based membranes: Performance improvement and possible mechanism. J. Membr. Sci. 277:157-164. Wan LS, Xu ZK, Huang XJ, Wang ZG, Wang HL (2005a) Copolymerization of acrylonitrile with N-vinyl-2-pyrrolidone to improve the hemocompatibility of poly acrylonitrile. Polymer 46:7715-7723. Wan LS, Xu ZK, Huang XJ, Wang ZG, Ye P (2005b) Hemocompatibility of poly(acrylonitrile-co-N-vinyl-2-pyrrolidone)s: Swelling behavior and water states. Macromol. Biosci .5:229-236. Wan LS, Xu ZK, Jiang HL (2006b) Fibrous membranes electrospinning from acrylonitrilebased polymers: Specific absorption behaviors and states of water. Macromol. Biosci. 6:364-372. Wang T, Kumar S (2006) Electrospinning of polyacrylonitrile nanofibers. J. Appl. Polym. Sci. 102:1023-1029. Wang YK, Yong T, Ramakrishna S (2005) Nanofibres and their influence on cells for tissue regeneration. Aust .J. Chem. 58:704-712. Wang ZG, Ke BB, Xu ZK (2007a) Covalent immobilization of redox enzyme on electrospun nonwoven poly(acrylonitrile-co-acrylic acid) nanofiber mesh filled with carbon nanotubes: A comprehensive study. Biotechnol. Bioeng. 97:708-720. Wang ZG, Wan LS, Xu ZK (2007b) Surface engineerings of polyacrylonitrile-based asymmetric membranes towards biomedical applications: An overview. J. Membr. Sci. 304:8-23.

302

Ai-Fu Che, Ling-Shu Wan and Zhi-Kang Xu

Wang ZG, Wang JQ, Xu ZK (2006a) Immobilization of lipase from Candida rugosa on electrospun polysulfone nanofibrous membranes by adsorption. J. Mol. Cat. B: Enzymatic 42:45-51. Wang ZG, Xu ZK, Wan LS (2006b) Modulation the morphologies and performance of polyacrylonitrile-based asymmetric membranes containing reactive groups: Effect of non-solvents in the dope solution. J. Membr. Sci. 278:447-456. Wang ZG, Xu ZK, Wan LS, Wu J, Innocent C, Seta P (2006c) Nanofibrous membranes containing carbon nanotubes: Electrospun for redox enzyme immobilization. Macromol. Rapid Commun. 27:516-521. Xu ZK, Kou RQ, Liu ZM, Nie FQ, Xu YY (2003) Incorporating alpha-allyl glucoside into polyacrylonitrile by water-phase precipitation copolymerization to reduce protein adsorption and cell adhesion. Macromolecules 36:2441-2447. Yang Q, Wu J, Li JJ, Hu MX, Xu ZK (2006) Nanofibrous sugar sticks electrospun from glycopolymers for protein separation via molecular recognition. Macromol. Rapid Commun. 27:1942-1948. Yarin AL, Koombhongse S, Reneker DH (2001) Bending instability in electrospinning of nanofibers. J. Appl. Phys. 89:3018-3026. Ye P, Xu ZK, Che AF, Wu J, Seta P (2005) Chitosan-tethered poly(acrylonitrile-co-maleic acid) hollow fiber membrane for lipase immobilization. Biomaterials 26:6394-6403. Ye P, Xu ZK, Wu J, Innocent C, Seta P (2006a) Nanofibrous membranes containing reactive groups: Electrospinning from poly(acrylonitrile-co-maleic acid) for lipase immobilization. Macromolecules 39:1041-1045. Ye P, Xu ZK, Wu J, Innocent C, Seta P (2006b) Nanofibrous poly(acrylonitrile-co-maleic acid) membranes functionalized with gelatin and chitosan for lipase immobilization. Biomaterials 27:4169-4176. Yu L, Huang M, Wang PG, Zeng X (2007) Cross-linked surface-grafted glycopolymer for multivalent recognition of lectin. Anal. Chem. 79:8979-8986. Zhao YY, Yang QB, Lu XF, Wang C, Wei Y (2005) Study on correlation of morphology of electrospun products of polyacrylamide with ultrahigh molecular weight. J. Polym. Sci,. Part B: Polym. Phys. 43:2190-2195.

In: Nanofibers: Fabrication, Performance, and Applications ISBN 978-1-60741-947-1 Editors: W. N. Chang © 2009 Nova Science Publishers, Inc.

Chapter 9

PORPHYRINATED POLYMER NANOFIBERS BY ELECTROSPINNING Yuan-Yuan Lv1, Jian Wu1, Zhen-Mei Liu2, and Zhi-Kang Xu2* 1

2

Department of Chemistry, Zhejiang University, Hangzhou, China Institute of Polymer Science, and Key Laboratory of Macromolecular Synthesis and Functionalization, Department of Polymer Science and Engineering, Zhejiang University, Hangzhou, China

ABSTRACT Electrospinning has been suggested as a useful method to prepare non-woven fabrics of sub-micron or nano-scale fibers, which have high porosity and large surface area-tovolume ratio, small pore size between the depositing fibers of the electrospun mats, and vast possibility for surface functionalization. These characteristics make the non-woven fabrics attractive for many applications, such as functional membranes, photocatalysts, biosensors, and nanoelectronics. On the other hand, porphyrins play important roles in biological processes and much attention has been paid to design and synthesize porphyrin-functionalized polymers for potential applications including molecular recognition or molecular imprinting, sensors, interactions with biological systems, and enzyme mimics for catalysis. Combining the merits of electrospinning with the bioinspired applications of porphyrinated polymers may generate functionalized nanofibers for more multiple purposes. Following this idea, various porphyrinated polymers, which include polyacrylonitrile, polyimide and polypeptide, were synthesized by either physical blending or chemical copolymerization. They were fabricated into nanofibrous membranes by electrospinning process. We found these porphyrinated polymer nanofibers not only preserved their nature characteristics but also endowed new spectroscopy properties of porphyrins. On the other hand, it is well known that porphyrins have been used as red emitting materials that have reasonable fluorescence efficiency and good thermal stability. Based on this, fluorescent microspheres or nanofibers with different diameters were prepared from the porphyrinated polymers by changing the parameters for electrospinning process, such as solution concentration and molecular weight of the polymers. Confocal laser scanning microscopy (CLSM) showed

* Corresponding author. E-mail: [email protected]; fax: + 86 571 8795 1773.

304

Yuan-Yuan Lv, Jian Wu, Zhen-Mei Liu et al. that red light emitted uniformly through out the nanofibers in spite of fiber morphologies. Besides being used as emitting materials, we expect that these luminescent nanofibers may be latent materials applied in many areas such as catalysis, molecular imprinting, biosensors, and light/energy conversion.

Keywords: electrospinning; fluorescence; nanofiber; porphyrin; copolymers

INTRODUCTION Porphyrins are a class of naturally occurring macrocyclic compounds, which play a very important role in the metabolism of living organisms. The porphyrin molecule contains four pyrrole rings linked via methine bridges (Figure. 1). During the last decades, much attention has been paid to design and synthesize porphyrin-functionalized polymers for potential applications in solar energy conversion [1], electron and energy transfer [2], non-linear optics [3], and photodynamic therapy for solid tumors [4]. Great efforts have been devoted to the design and synthesis of porphyrin polymers over the past decades. Up to now, several kinds of porphyrin polymers have been reported. Electrospinning is a versatile technique to prepare nanofibers with the notable advantages. When the diameters of polymer fiber decreases from micrometers (e.g. 10-100 μm) to submicrons or nanometers (e.g. 10×10-3-100×10-3μm), there appear several amazing characteristics such as very large surface area to volume ratio (this ratio for a nanofiber can be as large as 103 times of that of a microfiber), flexibility in surface functionalities, and superior mechanical performance (e.g. stiffness and tensile strength) compared with any other known form of the material [5]. These outstanding properties make the polymer nanofibers optimal candidates for many important applications. Combining the merits of electrospinning with the bioinspired applications of porphyrin polymers may generate functionalized nanofibers for multiple purposes such as molecular recognition or molecular imprinting, sensors, light-emitting and energy/electron transfer materials, interactions with biological systems, and enzyme mimics for catalysis. Herein, we will try to review some recent progresses in electrospun nanofibrous materials from porphyrinated polymers.

Figure 1. Schematic representative of porphyrin macrocyclic system.

Porphyrinated Polymer Nanofibers by Electrospinning

305

1. PORPHYRIN-FILLED POLY(ACRYLONITRILE-CO-ACRYLIC ACID) (PANCAA) NANOFIBERS The unique characteristics of electrospun nanofibers make them much superior over other kinds of carrier materials for enzyme immobilization. The large surface-to-volume ratio provides large amount of immobilized enzyme, and the high porosity can enhance the enzyme activity through decreasing the diffusion resistance to substrates and products. However, for the immobilization of catalase, a redoxase that requires electron exchange during catalysis, electron transfer pathway is needed. It is well-known that porphyrins are some of the most ideal electron mediators because of their high highest occupied molecular orbital (HOMO) and low lowest unoccupied molecular orbital (LUMO). Here, three kinds of porphyrin-filled nanofibers containing reactive groups were prepared by electrospinning the mixtures of 5,10,15,20-tetraphenylporphyrin (TPP) and its metalloderivatives (ZnTPP and CuTPP) with poly(acrylonitrile-co-acrylic acid) (PANCAA). Catalase was then covalently immobilized onto these porphyrin-filled nanofibers to study the interaction between the porphyrins and the immobilized catalase. Nanofibers with uniform diameter around 150 nm were fabricated for each sample under optimized electrospinning conditions (Figure. 2). Compared with PANCAA nanofibers, both the TPP and the metalloporphyrin-filled ones exhibit rougher surfaces and these electrospun nanofibrous membranes with different porphyrins display distinct colors (Figure. 3).

Figure 2 FESEM (×50,000) images of nanofibers electrospun from (A) PANCAA; (B) PANCAA/TPP; (C) PANCAA/ZnTPP and (D) PANCAA/CuTPP.

306

Yuan-Yuan Lv, Jian Wu, Zhen-Mei Liu et al.

Figure 3 Images of the nanofibrous membranes prepared from PANCAA and its porphyrinated derivatives. (1) PANCAA, (2) PANCAA/TPP, (3) PANCAA/ZnTPP, (4) PANCAA/CuTPP.

Table 1. Bound enzyme, specific activity as well as activity retention of catalase immobilized on various carriers. Information of free catalase is also listed as a comparison Sample

Nanofiber diameter

Bound enzyme

Specific activity

(mg/g)

(Units)

(nm)

Activity retention (%)

free catalase

—

—

1532 ± 25

—

PANCAA

134 ± 19

11.54 ± 0.63

436 ± 23

28.4

PANCAA/TPP

163 ± 22

12.57 ± 0.76

565 ± 37

36.9

PANCAA/ZnTPP

121 ± 32

11.56 ± 0.59

839 ± 42

54.8

PANCAA/CuTPP

150 ± 30

11.93 ± 0.80

558 ± 21

36.4

Catalase was covalently immobilized onto the nanofibers as described previously.[6]. The amounts of bound catalase and the activity are listed in Table 1. It can be seen that for the four nanofibrous carriers, the amounts of bound catalase are almost the same. However, catalase immobilized on the porphyrin-filled nanofibers possesses much higher activity, the enhancement in enzymatic activity depends mainly on the types of porphyrins. For example,

Porphyrinated Polymer Nanofibers by Electrospinning

307

the activity retention of catalase on nanofibers containing TPP and CuTPP increases from 28.4 to 36.9 and 36.4%, respectively, when comparing with that on the PANCAA nanofibers. Interestingly, ZnTPP-filled nanofibers present encouraging results in that the activity retention of catalase on these nanofibers is improved up to 54.8%. Additional experimental results confirm that this improvement is caused by the porphyrin, not by others such as free metal ions (Zn2+ or Cu2+) (Figure 4).

Figure 4 Effect of metal ions on the catalase activity. Free catalase (0.1 mL, 0.01 mg/mL) and metal ions (0.1 mL, 2mM) were injected into hydrogen peroxide solution (3 mL, 9.7mM) and the decrease in the absorbance at 240 nm was recorded. Results indicate that metal ions show no influence on the catalase activity.

Figure 5 QCM measurement of catalase adsorption onto the PANCAA films with or without porphyrin. The thin PANCAA films were spin-coated on the sensor surface prior to QCM investigation.

308

Yuan-Yuan Lv, Jian Wu, Zhen-Mei Liu et al.

Figure 6 Fluorescence spectra of porphyrin and catalase in water/DMSO mixture (50:50 volume ratio). Excitation wavelength is 420 nm.

Porphyrinated Polymer Nanofibers by Electrospinning

309

To understand the mechanism of improved catalase activity by the incorporation with porphyrins, quartz crystal microbalance (QCM) and fluorescence spectroscopy were adopted. In QCM measurement, the resonant frequency of crystal depends on the total oscillation mass. An increase in mass adsorbs to the quartz surface causes the frequency to decrease. Therefore, protein (here catalase) adsorption can be monitored in real time by QCM. To obtain that, the films of enzyme carrier were firstly spin-coated onto QCM sensor surface. The film thickness was ca. 25 nm as calculated from Sauerbrey equation [7]. Figure 5 shows the representative QCM curves from 5 parallel measurements. From these protein adsorption curves, it can be deduced that the films with porphyrin adsorb more catalase and at a quicker rate than PANCAA film, implying that the films with porphyrin have higher affinity for catalase. Furthermore, compared with TPP-filled film, those with metalloporphyrin show stronger affinity for catalase, which is reasonable because the metal center may play an important role in interacting with catalase. This interaction with porphyrin in turn changes the structure and the activity of catalase. The improvement of catalase activity by porphyrin, especially by ZnTPP in our cases, may be attributed to several causes, which are still complicated to fully understand for us. However, it is well known that porphyrins have high affinity for heme-containing proteins owing to the analogical structure, which was confirmed by our QCM results as well as the results of interactions between porphyrin and peroxidase, hemoglobin or myoglobin reported by other research groups [8-10]. Since catalase containing hemes, this affinity makes the tight interacting between the porphyrin and catalase possible, which in turn facilitates the electron transfer. This was evidenced by the fluorescence spectroscopy curves. As is shown in Figure 6, the emission bands at 480~510 nm, 600 nm and 650 nm (excited at 420 nm) are ascribed to the protoporphyrin core in catalase, ZnTPP and TPP, respectively. The fluorescence of CuTPP is fully quenched because of the paramagnetic metal center. The intensity of emission peaks at 480~510 nm, which are induced by catalase, decreases with the concentration of porphyrin when keeping the concentration of catalase constant. The quenching of the catalase singlet state is entirely due to singlet-singlet electron transfer to the porphyrins, whose first excited singlet state lies at lower energy [11]. Moreover, ZnTPP with stronger ability to facilitate the electron transfer induces larger activity increment. Therefore, the building of electron transfer pathway between catalase and its matrix might be mainly responsible for the improvement of activity. Besides, it should be noted that CuTPP slightly increases the catalase activity although the electron transfer between CuTPP and catalase is as remarkable as that for ZnTPP. It means the properties (e.g. the paramagnetic property of copper) of metal center of porphyrin should also be taken into account. This type of porphyrin-filled nanofibers could be applied to biosensor.

2. VINYL PORPHYRIN WITH ACRYLONITRILE COPOLYMER NANOFIBERS Besides physical blending, covalent binding is another method to introduce porphyrin into polymer in which porphyrin acts as core, main chain unit or pendant. Herein, we describe the synthesis of copolymers of vinyl porphyrins with acrylonitrile, and the electrospinning of the resultant copolymers into nanofibers [12]. Solution copolymerization of acrylonitrile(AN)

310

Yuan-Yuan Lv, Jian Wu, Zhen-Mei Liu et al.

with porphyrins was performed at 60 oC in DMSO initiated with AIBN, the schematic representative is shown in Figure 7. Nanofibers were successfully prepared from the resultant porphyrinated polymers by electrospinning, the FESEM images of which are shown in Figure 8. During the electrospinning process, many factors can influence the diameters and the morphologies of nanofibers. Among them, polymer concentration is one of the most important factors for a certain polymer solution. In our cases, solutions with 5 wt.-% porphyrinated polymer in DMF resulted in microspheres with diameters between 0.5 and 2 μm (Figure 8 (A)). By increasing the concentration of solutions from 5 wt.-% to 15 wt.-%, uniform nanofibers with diameter around 330 ± 34 nm could be fabricated as indicated in Figure 8 (C). In this way, fibers with different diameters could be feasibly prepared to meet the requirements for various purposes.

Figure 7 Schematic representation for the synthesis of of vinyl porphyrins with acrylonitrile (Wan et al. 2006). Copyright © (2006). Reprinted with permission of John Wiley and Sons, Inc.

It should be noted that the nanofibers containing metalloporphyrins could also be prepared (Figure 8 (E) and (G)). Since the center metal is very important for the functions of porphyrins, for example, aluminium porphyrin is useful as an initiator of polymerization [13], method introduced in this part provides a new technique to fabricate multifunctional nanofibers containing porphyrin complexes with other metals such as Fe, Mn, Ni, Co, Cu, and Al. On the other hand, versatile monomers (e.g. styrene) are easy to copolymerize with vinyl porphyrin [14], which could also extend the applications of the porphyrinated nanofibers. Following this work, terpolymer composed of acrylonitrile (AN), acrylic acid (AA), and vinyl porphyrin (PP) was synthesized in our lab [15], according to the method reported in previous work [12] (Figure 9). Nanofibers can be conveniently fabricated from the terpolymer via the electrospinning technique filled with or without carbon nanotubes (CNTs) (Figure 10). After that, catalase, a kind of redox enzyme, was chemically immobilized on the porphyrinated nanofibers. The activity and stabilities of the immobilized catalases with the porphyrinated nanfibers as matrix were studied.

Porphyrinated Polymer Nanofibers by Electrospinning

311

Figure 8 FESEM micrographs of vinyl porphyrin with acrylonitrile copolymer (A) microspheres electrospun from a 5 wt.-% copolymer solution and (C) nanofibers electrospun from a 15 wt.-% copolymer solution; and fluorescence microscope images (×100) of those electrospun from (B) 5 wt.% and (D) 15 wt.% copolymer solutions. Here the copolymer contains MATPP. Micrographs (E) to (H) corresponds to the copolymer containing ZnMATPP. Polymer solutions were directly electrospun to a small cover glass for fluorescence microscope observation. Wavelength of excitation is 488 nm. Electrospinning was performed for (B) 20 min, (D) 1 min, (F) 5 min and (H) 1 min for the fluorescence microscope observations (Wan et al. 2006). Copyright © (2006). Reprinted with permission of John Wiley and Sons, Inc.

312

Yuan-Yuan Lv, Jian Wu, Zhen-Mei Liu et al.

Figure 9. Schematic representation for the synthesis of porphyrin-containing terpolymers and enzyme binding onto the nanofiber surfaces (Wan et al. 2007). Copyright (2007) American Chemical Society.

Figure 10. FESEM images of the electrospun nanofibers: (A) PAN (diameter 154 ± 30 nm); (B) PANAACoPP (diameter 180 ± 30 nm); (C) PAN/CNT (diameter 180 ± 34 nm); (D) PANAACoPP/CNT (diameter 165 ± 37 nm). (×50,000) (Wan et al. 2007). Copyright (2007) American Chemical Society.

Porphyrinated Polymer Nanofibers by Electrospinning

313

Table 2. Activity and kinetic parameters for the free and immobilized catalases on various electrospun matrix (Wan et al.). Copyright (2007) American Chemical Society Sample

Carboxyl content (mmol/g)

Bound enzyme (mg/g fibers)

free catalase

—

—

PAN

0.18 ± 0.03

PAN/CNT

Specific activity (Units)

Activity retention (%)

Km

Vmax

(mM)

(Units)

2491.03 ± 50.21

—

34.1

12124.7

24.45 ± 2.33

807.63 ± 107.82

32.4

82.3

9077.4

0.25 ± 0.04

29.81 ± 3.76

1127.91 ± 103.17

45.3

65.1

10063.9

PANAACoPP

0.14 ± 0.04

18.93 ± 4.03

979.46 ± 120.24

39.3

73.9

9641.0

PANAACoPP/ CNT

0.20 ± 0.06

22.81 ± 4.22

1207.39 ± 134.66

48.5

64.8

11097.4

Data listed in Table 2 reveal that the introduction of porphyrin pendants into polymer chains benefits the activity of the immobilized catalase. For example, the activity retention of catalases on PANAACoPP nanofibers increase to 39.3% from 32.4%, compared with that on the PAN nanofibers. Furthermore, the activity retention of catalase on PANAACoPP/CNT nanofibers, 48.5%, is larger than that on PAN/CNT, 45.3%. Although the activity increment induced by the porphyrin pendants is not very large, the result is still encouraging, which confirms our principle idea, i.e. porphyrin moieties have positive effects on the immobilized enzyme. It has been demonstrated in our previous work [16] that, as a matrix for enzyme immobilization, electrospun nanofibers can not only offer high enzyme loading due to the large surface area to volume ratio, but enhance the enzyme activity because of the high porosity which can decrease the diffusion limitation during the catalytic reaction. Moreover, in this work, extra improvement on the activity of immobilized catalase due to both CNTs and porphyrin pendants is achieved. This increment should be first attributed to the facilitation of electron transfer between the matrix and immobilized catalases. It has been proposed that the enhanced electronic conductivity induced by the unique electronic properties of CNTs is in favor of the immobilized glucose oxidase [17]. Also, porphyrin has been widely applied in light emitting or energy transfer materials and the possibility of electron/energy transfer of porphyrin has been proved [18, 19]. CNT is generally recognized as an electron acceptor while porphyrin is an electron donor [20], thus the composites of CNTs with porphyrin have been studied by some researchers and the unique interactions have been confirmed [21, 22]. Since catalase is a kind of redox enzyme that requires electron exchange during biocatalysis, the facilitation of electron transfer will unambiguously increase the enzyme activity. The second factor for the increase of catalase activity might be the steric effects of the porphyrinated polymers. The steric and electronic properties of the support, which can be greatly influenced in our case by the porphyrin pendants on the nanofiber surface, might be very important for supported catalysis [23, 24]. Zhou et al. suggested that certain tetraphenylphorphyrins could bind to complementary protein surfaces with considerable selectivity in which the hydrophobic core of the porphyrin primarily contributed to the binding affinity [25]. The porphyrin pendants also show remarkable steric effects.

314

Yuan-Yuan Lv, Jian Wu, Zhen-Mei Liu et al.

In fact, the improvement of enzyme retention activity can be attributed to several reasons. Facilitation of electron transfer process might also be one of such, but it is not clarified till now because of the difficulty in monitoring or observing the process directly except in the case of biosensor. Up to now, experimental evidences for electron transfer were mostly collected from biosensor researches. Therefore, the above-mentioned experimental results can only preliminarily support the facts that the introduction of both CNTs and porphyrin pendants are in favor of the immobilized catalases in terms of improving the activity and stabilities of the immobilized catalases.

3. PORPHYRINATED POLYIMIDE NANOFIBERS Polyimides constitute an important class of polymers due to their superior thermal and chemical resistance as well as mechanical properties that can find applications in various fields [26, 27]. Polyimide nanofibers based on pyromellitic dianhydride (PMDA) and 4,4’oxidianiline (ODA) (PMDA-ODA) with ultra-low dielectric constant were prepared by electrospinning the precursor of polyimide, a poly(amic acid) (PAA) solution, with subsequent thermal imidization [28] and the obtained nanofibers has many attractive applications [29-31]. Copolyimides (CPIs) containing 5,15-Bis(4-aminophenyl)-10,20-diphenylporphyrin (trans-DATPP) and 5,10-bis (4-aminophenyl)-15,20-diphenylporphyrin (cis-DATPP moieties could be synthesized by a two-step reaction procedure similar to the synthesis of common polyimide. Figure 11 illustrates the synthesis of the target polymers [32]. Nanofibers were successfully prepared from the resultant copoly(amic acid)s (CPAAs) by electrospinning a 15 wt.% solution in DMAc. To prepare polyimide nanofibers with high temperature resistance, imidization of the as-spun CPAA nanofibers was performed by heating step by step with N2 protection at 80 °C for 0.5 hour, 160 °C for 1 hour, and 250 °C for 4 hours. CPAA nanofibers were converted into CPI nanofibers through the imidization process. Figure 12 represents the thermogravimetric (TG) curves of the CPI nanofibers. It reveals that the four CPI nanofibers display good thermal stabilities up to approximately 500 °C, indicating the electrospun nanofibers preserve the thermal-resistant nature of aromatic polyimide materials. The CPI nanofibers containing cis-DATPP are slightly more stable than that containing trans-DATPP at the same level of porphyrin content. The disappearance of the thermolysis temperatures at around 380 °C for the starting DATPPs (trans-DATPP and cis-DATPP) indicates that the thermal stability of porphyrin units in the CPIs is improved through imidization. FESEM micrographs of the CPAA and CPI nanofibers (Figure 13) show that the fiber diameter increases with the contents of DATPP. The average diameters of CPAA nanofibers range between 150-260 nm while those of CPI nanofibers are 140-230 nm. In other words, the nanofibers shrink obviously after the imidization process in spite that the CPI nanofibers are still uniform and continuous. Figure 14 shows the fluorescence intensities of the copoly(amic acid)s (CPAA) and copolyimides (CPI) nanofibers recorded by laser scanning confocal microscope (LSCM). The corresponding dense films prepared by spinning coating were also investigated as controls.

Porphyrinated Polymer Nanofibers by Electrospinning

Figure 11 Schematic representation for the synthesis porphyrinated polyimide (Lv et al. 2008). Copyright (2008) American Chemical Society.

315

316

Yuan-Yuan Lv, Jian Wu, Zhen-Mei Liu et al.

Figure 12 Thermogravimetry curves of CPI nanofibers (Lv et al. 2008). Copyright (2008) American Chemical Society.

There is a decrease of fluorescence intensity after imidization for both nanofibers and dense films. It indicates that fluorescence quenching for the porphyrin took place when imidized. This could be attributed to the electron transfer from a porphyrin donor to a diimide acceptor group, a five-atom ring, which forms after imidization. Because the aromatic polyimides contain an alternating sequence of electron-rich donor and electron-deficient acceptor subunits, intramolecular charge-transfer complexes could be formed [33-35]. It is well known that porphyrin is an electron-rich moiety. When porphyrin is incorporated into polyimide chains, charge-transfer complexes may form which induce some extraordinary photophysical porperties [36]. Furthermore, there is another interesting phenomenon as observed from Figure 14. The fluorescence intensities of the dense films are significantly weaker than the nanofibers. This result suggests a superior luminescence characteristic of the polyimide nanofibers. It should be noted that these luminescent nanofibers might be applied in many areas. For example, they could be useful in the detection of some metal ions such as mercury because fluorescence quenching could be visually detected when the porphyrin units coordinate with Hg(II). For the monitoring of some corrosive gaseous pollution such as HCl and SO2 produced by chemical industry, this type of luminescent nanofibers with high temperature resistance and stable characters is of special interest.

Porphyrinated Polymer Nanofibers by Electrospinning

317

Figure 13 FESEM micrographs of (A) CPAA-1 (Φ=156~180 nm), (B) CPI-1 (Φ=141~165 nm), (C) CPAA-2 (Φ=221～245 nm), (D) CPI-2 (Φ=201～225 nm), (E) CPAA-3 (Φ=211～235nm), (F) CPI-3 (Φ=186～210 nm), (G) CPAA-4 (Φ=231～255 nm) and (H) CPI-4 (Φ=206～230 nm) (Lv et al. 2008). Copyright (2008) American Chemical Society.

318

Yuan-Yuan Lv, Jian Wu, Zhen-Mei Liu et al.

Figure 14 Fluorescence intensity of the CPAA and CPI nanofibers (A) and dense films (B) investigated by LSCM. The wavelength of excitation is 488 nm and the scope of fluorescence signal for investigation ranges from 650 to 700 nm (Lv et al. 2008). Copyright (2008) American Chemical Society.

Porphyrinated Polymer Nanofibers by Electrospinning

319

4. PORPHYRINATED POLY(γ-STEARYL α L-GLUTAMATE)S NANOFIBERS Poly(amino acid)s and their derivatives are attractive polymers with potential applications in various fields such as tissue engineering and pharmaceutical materials [37, 38]. Ringopening polymerization (ROP) of α-amino acid N-carboxyanhydrides (NCA) is the most economical and convenient way to synthesize poly(amino acid)s with high molecular weight. NCA polymerization can be initiated with different nucleophiles and bases. Among them primary amines are good initiators due to their nucleophilicity and the narrow molecular weigh distribution of the synthesized poly(amino acid)s. Furthermore, the the initiator acts as the end group of synthesized poly(amino acid)s through the formation of amide bond at the initiation step. In our previous study [39], meso-tetrakis(4-aminophenyl) porphyrin (TAPP) was used to initiate the ROP of γ-stearyl α L-glutamate N-carboxyanhydride (SLGNCA). In that case, the ROP of SLGNCA was based on the nucleophilic attack from the amino group of porphyrin on the C(5) carbonyl group of SLGNCA, yielding four-arm branched poly(γ-stearyl α L-glutamate) (PSLG) with the porphyrin as the core. Following this work, we synthesized a series of novel porphyrinated PSLGs initiated by 5-(4-aminophenyl)- 10,15,25triphenylporphyrin (APTPP) and corresponding metalloporphyrins. It has been proved that porphyrin substituted with four amino groups is an efficient initiator for the ROP of SLGNCA in our previous work [39], while the intrinsic viscosity of the synthesized PSLG is relative low, which is a limitation for the preparation of well-defined fibers by electrospinning. Deming et al. demonstrated that metal-amine complexes (mixtures of N,N’-bis(3,5-di-nitryl-salicylidene)methyltris(2-aminoethyl)aminonium chloride and metal acetates; metal ion = Ni(II), Cu(II), Pd(II), Co(II)) could greatly control over the molecular weight of poly(amino acid)s than conventional primary amines [40]. They proposed that through the coordination between the amino group and the metal ion, the nucleophilicity of the amino group could be tuned to eliminate the side reaction, resulting in synthesized poly(amino acid)s with relatively high molecular weight. In the case of metallo-APTPP and metallo-TAPP, the coordination between the metal ion and the amino group in metal APTPP (MAPTPP) is much stronger than that between the metal ion and the four amino groups in metal TAPP (MTAPP). It can be expected that the nucleophilicity of MAPTPP can be tuned more obviously than that of MTAPP. MAPTPPs with Zinc(II) and Cobalt(II) as metal ion were used to initiate the polymerization of SLGNCA for the synthesis of porphyrinated PSLGs, APTPP was also studied for comparison (Figure 15). PSLGs porphyrinated by APTPP, ZnAPTPP, and CoAPTPP at SLGNCA/initiator ratio of 100 were designed as APTPP100-PSLG, ZnAPTPP100-PSLG and CoAPTPP100-PSLG respectively. CoAPTPP150-PSLG was designation for the PSLG porphyrinated by CoAPTPP at SLGNCA/initiator ratio of 150. These PSLGs were tried to fabricate into microfibers by electrospinning. Unfortunately, for the APTPP100-PSLG, even with high polymer concentration of 26 wt.-%, a small fraction of beads are observed along the fiber axis although continuous fibers can be obtained. For the ZnAPTPP100-PSLG at the same concentration, more beaded fibers present (Figure 16 (A) and (C)). The relatively low viscosities of APTPP100-PSLG and ZnAPTPP100-PSLG solution are the main reason for the formation of beaded fibers[41].

320

Yuan-Yuan Lv, Jian Wu, Zhen-Mei Liu et al.

Figure 15. Schematic representation for the synthesis of the porphyrinated PSLGs.

Figure 16. FESEM (left) and CLSM (right) (×63) images for the electrospun mats from the solutions of APTPP100-PSLG (a, b) and ZnAPTPP100-PSLG (c, d) in chloroform with concentration of 26 wt.-%. The excitation wavelength of CLSM is 488 nm.

From the data of porphyrinated PSLGs listed in Table 3, PSLGs initiated by CoAPTPP have much higher intrinsic viscosities than those initiated by APTPP100-PSLG and ZnAPTPP100-PSLG. This makes them promising to be electrospun into well-defined fibers. The results are shown in Figure 17. Continuous and homogenous fibers are observed for mat fabricated from CoAPTPP100-PSLG (Figure 17 (A)) with a solution concentration of 26 wt.% (2.56 ± 0.89 μm). At the same polymeric solution concentration, uniform fibers are obtained for the CoAPTPP150-PSLG with smaller diameters (1.26 ± 0.75 μm) (Figure 17 (C)).

Porphyrinated Polymer Nanofibers by Electrospinning

321

Table 3. Effects of the structure of porphyrin initiators on the synthesis of porphyrinated PSLGs PSLG sample

SLGNCA/porphyrin (mol/mol)

Yield (wt.-%)

Intrinsic viscosity (dL/g)

APTPP100-PSLG ZnAPTPP100-PSLG CoAPTPP100-PSLG CoAPTPP150-PSLG

100:1 100:1 100:1 150:1

78.2 71.5 73.5 72.7

0.100 0.105 0.181 0.283

Decreasing the concentration of CoAPTPP150-PSLG solution to 15 wt.% leads to the formation of dish-shaped particles with diameters of 6.52 ± 2.72 μm (Figure 17 (E)). These results demonstrate that different morphologies of electrospun mats from the porphyrinated PSLGs can be prepared by changing the solution parameters, such as concentration and viscosity, which may meet various acquirements for applications. Since both NCA monomers and metal ion in metalloporphyrin can be modified, it is possible to prepare diverse poly(amino acid)s and to greatly extend the application of the porphyrinated poly(amino acid)s.

5. OPTICAL PROPERTIES OF PERPARED PORPHYRINATED NANOFIBERS The optical features showed by porphyrins and related compounds make these molecules particularly appealing for some optical purposes. Porphyrins have been used as red emitting materials which have reasonable fluorescence efficiency and good thermal stability. All the porphyrinated nanofibers described above have unique fluorescent properties. The luminescence properties of these porphyrinated nanofibers were characterized by fluorescence microscope or confocal laser scanning microscopy (CLSM). As have been showed in Figure 8 (B), (D), (E) and (F), Figure 13 (I) and (J), Figure 16 (B) and (D), and Figure 17 (B), (D) and (F)), red light was emitted uniformly from the electrospun mat regardless of their morphologies. It reveals the homogenous distribution of porphyrin on the fibers or microparticles without aggregation.

6. CONCLUSIONS Electrospinning is a simple, convenient, and versatile technique for generating fibers with diameters ranging from several micrometers to tens of nanometers. The collected fibrous membranes (or mats) have many exciting characteristics such as very high surface area to volume ratio, good mechanical strength, excellent flexibility, and large porosity.

322

Yuan-Yuan Lv, Jian Wu, Zhen-Mei Liu et al.

Figure 17 FESEM (left) and CLSM (right) (×63) images for the electrospun mats from the solutions of CoAPTPP100-PSLG (a, b), CoAPTPP150-PSLG (c, d) in chloroform with a concentration of 26 wt.-%, and CoAPTPP150-PSLG (e, f) in chloroform with a concentration of 15 wt.-%, respectively. The excitation wavelength of CLSM is 488 nm.

Porphyrinated Polymer Nanofibers by Electrospinning

323

Therefore, the preparation and utilization of electrospun nanofibers has become a subject of research interest with worldwide participation in both industrial and academic laboratories. Outstanding improvements have been achieved in the most recent years. These advances in electrospinning have made it possible to readily fabricate fibers with a great variety of morphologies and compositions. In this review, several kinds of porphyrinated polymers having good fiber-forming property were synthesized and the resultant copolymers were electrospun into luminescent nanofibers by electrospinning technique. It should be noted that besides free base porphyrin, metalloporphyrinated nanofibers could also be prepared. The open site of the porphyrin makes it possible to coordinate with almost all metals present in the Periodic Table, which greatly extend the applications of the porphyrinated nanofibers since the center metal is very important for the functions of porphyrins. On the other hand, as the organic chemistry of porphyrins is well developed, a wide range of different substituents can be introduced at their peripheral positions to tailor the structure of porphyrin, which in turn remarkably extends the variety of porphyrin-derived copolymers. Besides used for catalase immobilization, it is speculated that these luminescent nanofibers may be latent supports of porphyrins for various purposes such as catalysis, molecular imprinting, sensors, and light/energy conversion.

ACKNOWLEDGEMENTS Financial support from the National Natural Science Foundation of China for Distinguished Young Scholars (to Prof. Z. -K. Xu, grant no. 50625309) is gratefully acknowledged. The authors thank Dr. Zhen-Gang Wang, Dr. Ling-Shu Wan, Bei-Bei Ke and Lin-Jun Shao very much for their contribution.

REFERENCES [1]

[2]

[3] [4]

[5]

Wennerstrom, O.; Ericsson, H.; Raston, I.; Svensson, S. and Pimlott, W. (1989). Mesotetra(meso-tetraphenylporphyrinyl)porphyrin, a macrocycle with five covalently linked porphyrin units. Tetrahedron Lett., 30, 1129-1132. Ohkita, H.; Ogi, T.; Kinoshita, R.; Ito, S. and Yamamoto, M. (2002). Photoinduced electron transfer in nanostructures of ultrathin polyimide films containing porphyrin moieties. Polymer, 43, 3571-3577. Jeong, S. W. and Lim, H. S. (2001). Study on the preparation of aromatic polyimide Langmuir-Blodgett films. Synthetic Met., 123, 183-187. Mineo, P.; Scamporrino, E. and Vitalini, D. (2002). Synthesis and characterization of uncharged water-soluble star polymers containing a porphyrin Core. Macromol. Rapid Commun., 23, 681-687. Huang, Z. M.; Zhang, Y. Z.; Kotaki, M. and Ramakrishna, S. (2003). A review on polymer nanofibers by electrospinning and their applications in nanocomposites. Compos. Sci. Technol., 63, 2223-2253.

324 [6]

[7] [8] [9] [10] [11]

[12] [13]

[14]

[15]

[16]

[17]

[18]

[19]

[20]

[21] [22]

Yuan-Yuan Lv, Jian Wu, Zhen-Mei Liu et al. Wang, Z. G.; Xu, Z. K.; Wan, L. S.; Wu, J.; Innocent, C. and Seta, P. (2006). Nanofibrous membranes containing carbon nanotubes: electrospun for redox enzyme immobilization. Macromol. Rapid Commun., 27, 516-521. Sauerbrey, G. (1959). Use of quartz vibration for weighing thin films on a microbalance. J. Physik, 155, 206-212. Sil, S. and Chakraborti, A. S. (2002). Hematoporphyrin interacts with myoglobin and alters its functions. Mol. Cell. Biochem., 237, 103-110. Chakraborti, A. S. (2003). Interaction of porphyrins with heme proteins-a brief review. Mol. Cell. Biochem., 253, 49-54. Sil, S. and Chakraborti, A. S. (2005). Binding of porphyrin to horseradish peroxidase: effects on structure and function. Int. J. Biol. Macromol., 36, 16-22. Kuciauskas, D.; Lin, S.; Seely, G. R.; Moore, A. L.; Moore, T. A. and Gust, D. (1996). Energy and photoinduced electron transfer in porphyrin-fullerene dyads. J. Phys. Chem., 100, 15926-15932. Wan, L. S.; Wu, J. and Xu, Z. K. (2006). Porphyrinated nanofibers via copolymerization and electrospinning. Macromol. Rapid Commun., 27, 1533-1538. Chisholm, M. H.; Navarro-Llobet, D. and Simonsick, W. J. (2001). A comparative study in the ring-opening polymerization of lactides and propylene oxide. Macromolecules, 34, 8851-8857. Aramata, K.; Kajiwara, A.; Hashidzume, A.; Morishima, Y. and Kamachi, M. (1998). Radical polymerization of 5-(4’-acrylamidophenyl)-10,15,20-triphenylporphine and 5(4’-methacrylamidophenyl)-10,15,20-triphenylporphine. Polym. J., 30, 702-707. Wan, L. S.; Ke, B. B.; Wu, J. and Xu, Z. K. (2007). Catalase immobilization on electrospun nanofibers: effects of porphyrin pendants and carbon nanotubes. J. Phys. Chem. C, 111, 14091-14097. Ye, P.; Xu, Z. K.; Wu, J.; Innocent, C. and Seta, P. (2006). Nanofibrous Membranes containing reactive groups: electrospinning from poly(acrylonitrile-co-maleic acid) for lipase immobilization. Macromolecules, 39, 1041-1045. Tsai, Y. C.; Li, S. C. and Chen, J. M. (2005). Cast thin film biosensor design based on a nafion backbone, a multi-walled carbon nanotube conduit, and a glucose oxidase function. Langmuir, 21, 3653-3658. Li, B. S.; Li, J.; Fu, Y. Q. and Bo, Z. S. (2004). Porphyrins with four monodisperse oligofluorene arms as efficient red light-emitting materials. J. Am. Chem. Soc., 126, 3430-3431. Luo, L. Y.; Lo, C. F.; Lin, C. Y.; Chang, I. J. and Diau, E. W. G. (2006). Femtosecond fluorescence dynamics of porphyrin in solution and solid films: The effects of aggregation and interfacial electron transfer between porphyrin and TiO2. J. Phys. Chem. B, 110, 410-419. Guldi, D. M.; Taieb, H.; Rahman, G. M. A.; Tagmatarchis, N. and Prato, M. (2005). Novel photoactive single-walled carbon nanotube-porphyrin polymer wraps: efficient and long-lived intracomplex charge separation. Adv. Mater., 17, 871-875. Guldi, D. M.; Rahman, G. M. A.; Zerbetto, F. and Prato, M. (2005). Carbon nanotubes in electron donor-acceptor nanocomposites. Acc. Chem. Res., 38, 871-878. Satake, A.; Miyajima, Y. and Kobuke, Y. (2005). Porphyrin-carbon nanotube composites formed by noncovalent polymer wrapping. Chem. Mater., 17, 716-724.

Porphyrinated Polymer Nanofibers by Electrospinning

325

[23] Yu, X. Q.; Huang, J. S.; Yu, W. Y. and Che, C. M. (2000). Polymer-supported ruthenium porphyrins: versatile and robust epoxidation catalysts with unusual selectivity. J. Am. Chem. Soc., 122, 5337-5342. [24] Akgol, S.; Kacar, Y.; Ozkara, S.; Yavuz, H.; Denizli, A. and Arica, M. Y. (2001). Immobilization of catalase via adsorption onto L-histidine grafted functional pHEMA based membrane. J. Mol. Catal. B-Enzym., 15, 197-206. [25] Zhou, H. C.; Baldini, L.; Hong, J.; Wilson, A. J. and Hamilton, A. D. (2006). Pattern recognition of proteins based on array of functionalized porphyrins. J. Am. Chem. Soc., 128, 2421-2425. [26] Zhang, Q.; Wu, D. Z.; Qi, S. L.; Wu, Z. P.; Yang, X. P. and Jin, R. G. (2007). Preparation of ultra-fine polyimide fibers containing silver nanoparticles via in situ technique. Mater. Lett., 61, 4027-4030. [27] Huang, C. B.; Wang, S. Q.; Zhang, H. A.; Li, T. T.; Chen, S. L.; Lai, C. L. and Hou, H. Q. (2006). High strength electrospun polymer nanofibers made from BPDA–PDA polyimide. Eur. Polym. J., 42, 1099-1104. [28] Liu, J. Y.; Min, Y.; Chen, J. Y.; Zhou, H. W. and Wang, C. (2007). Preparation of the ultra-low dielectric constant polyimide fiber membranes enabled by electrospinning. Macromol. Rapid Commun., 28, 215-219. [29] Kim, C.; Choi, Y. O.; Lee, W. J. and Yang, K. S. (2004). Supercapacitor performances of activated carbon fiber webs prepared by electrospinning of PMDA-ODA poly(amic acid) solutions. Electrochim. Acta., 50, 883-887. [30] Yang, K. S.; Edie, Dan D.; Lim, D. Y.; Kim, Y. M. and Choi, Y. O. (2003). Preparation of carbon fiber web from electrostatic spinning of PMDA-ODA poly(amic acid) solution. Carbon, 41, 2039-2046. [31] Xuyen, N. T.; Ra, E. J.; Geng, H. Z.; Kim, K. K.; An, K. H. and Lee, Y. H. (2007). Enhancement of conductivity by diameter control of polyimide-based electrospun carbon nanofibers. J. Phys. Chem. B, 111, 11350-11353. [32] Lv, Y. Y.; Wu, J.; Wan, L. S. and Xu, Z. K. (2008). Novel porphyrinated polyimide nanofibers by electrospinning. J. Phys. Chem. C, 112, 10609–10615. [33] O’Mahoney, C. A.; Williams, D. J.; Colquhoun, H. M.; Mayo, R.; Young, S. M.; Askari, A.; Kendrick, J. and Robson, E. (1991). Single-crystal x-ray studies of aromatic oligomers: conformation and packing in isomeric pyromellitimide-ether-sulfones. Macromolecules, 24, 6527-6530. [34] Dinan, F. J.; Schwartz, W. T.; Wolfe, R. A.; Hojnicki, D. S.; Clair, T. St. and Pratt, J. R. (1992). Solid-state 13C-NMR spectral evidence for charge transfer complex formation in aromatic diimides and dianhydrides. J. Polym. Sci., Polym. Chem., 30, 111-118. [35] Kawakami, H.; Anzai, J. and Nagaoka, S. (1995). Gas transport properties of soluble aromatic polyimides with sulfone diamine moieties. J. Appl. Polym. Sci., 57, 789-795. [36] Peng, Z. H.; Bao, Z. A. and Yu, L. P. (1994). Large photorefractivity in an exceptionally thermostable multifunctional polyimide. J. Am. Chem. Soc., 116, 60036004. [37] Deming, T. J. (2002). Methodologies for preparation of synthetic block copolypeptides: materials with future promise in drug delivery. Adv. Drug Deliver Rev., 54, 1145-1155. [38] Deming, T., J. (2007). Synthetic polypeptides for biomedical applications. Prog. Polym. Sci., 32, 858-875.

326

Yuan-Yuan Lv, Jian Wu, Zhen-Mei Liu et al.

[39] Wu, J. and Hu, X. P. (2002). Polymerization of γ-stearyl L-glutamate Ncarboxyanhydride with multifunctional porphyrin initiator. Polym. Adv. Technol., 13, 201-204. [40] Deming, T. J. (1997). Transition metal-amine initiators for preparation of well-defined poly(γ-benzyl L-glutamate). J. Am. Chem. Soc., 119, 2759-2760. [41] Deitzel, J. M.; Kleinmeyer, J.; Harris, D. and Beck, Tan N. C. (2001). The effect of processing variables on the morphology of electrospun nanofibers and textiles. Polymer, 42, 261-272.

In: Nanofibers: Fabrication, Performance, and Applications ISBN 978-1-60741-947-1 Editors: W. N. Chang © 2009 Nova Science Publishers, Inc.

Chapter 10

A NANOFIBRILLAR PROSTHETIC MODIFIED WITH FIBROBLAST GROWTH FACTOR-2 FOR SPINAL CORD REPAIR Sally Meiners11, Suzan L. Harris1, Roberto Delgado-Rivera1,2, Ijaz Ahmed1, Ashwin N. Babu1, Ripal P. Patel1 and David P. Crockett3 1

Department of Pharmacology, UMDNJ-Robert Wood Johnson Medical School, Piscataway, NJ, USA 2 Department of Chemistry and Chemical Biology, Rutgers University, Piscataway, NJ, USA 3 Department of Neuroscience and Cell Biology, UMDNJ-Robert Wood Johnson Medical School, Piscataway, NJ, USA

ABSTRACT Thousands of new cases of spinal cord injury occur each year in the USA alone. However, despite recent advances, there is at present no cure for the resulting paraplegia or quadriplegia. This chapter evaluates a spinal cord prosthetic (SCP) developed in our laboratoy that is comprised of longitudinally bundled strips of nanofibers whose surfaces have been modifed with fibroblast growth factor-2 (FGF-2). The SCP is designed to be a prefabricated implant that can be grafted into the lesion site not only to provide structural but also to provide chemical cues that permit regenerating axons to cross the lesion site. For a comparative study, two separate SCPs were produced with one containing unmodified nanofibers and the other containing FGF-2-modified nanofibers. Both SCPs correctly guided regenerating axons across the injury gap created by an over-hemisection to the adult rat thoracic spinal cord and encouraged revascularization of the injury site. Neither SCP initiated glial scarring when implanted into the injured rat spinal cord. However, devices that incorporated nanofibers modified with FGF-2 encouraged more axonal regrowth and significantly better functional recovery than did devices that

1 Proofs and correspondence to: Dr. Sally Meiners Department of Pharmacology UMDNJ-Robert Wood Johnson Medical School 675 Hoes Lane, Piscataway, NJ 08854 (732) 235-2890; FAX: (732) 235-4073, E-mail: [email protected].

328

Sally Meiners, Suzan L. Harris, Roberto Delgado-Rivera et al. incorporated unmodified nanofibers as assessed using the Basso, Beattie, Bresnahan (BBB) locomotor rating scale. As such, the FGF-2-modified SCP provides a multifaceted approach to spinal cord repair.

INTRODUCTION This work explores the utility of electrospun nonwoven polyamide nanofibers (average fiber diameter of ~180 nm) as components of implantable prosthetics for the regeneration of axons in the lesioned spinal cord. Electrospun nanofibers were chosen because they are structurally mimetic of the basement membrane, an ultra-dense form of the extracellular matrix [Ahmed et al., 2006]. This property has created great interest in the use of nanofibers for tissue engineering applications, yet their actual use for in vivo purposes is still in its early stages. The polyamide nanofibers considered in this chapter have several important advantages for regenerative medicine strategies. As we have discussed previously [Meiners et al., 2007], they promote neuronal attachment and neurite generation [Ahmed et al., 2006] and produce no apparent cytotoxicity in vitro [Ahmed et al., 2006] or in vivo [Meiners et al., 2007]. They do not rapidly degrade, and as such, their structural integrity is maintained for a number of weeks (at least 3-5) in vivo [Meiners et al., 2007]. This is highly relevant in that a scaffold to be used for repair of spinal cord injuries must be present in the damaged spinal cord for a sufficient period of time to allow optimal axonal regeneration across the injury site. This process generally requires weeks to months depending on the position and extent of the injury. Furthermore, scaffolds composed of materials that rapidly degrade in the body, such as polylactate, can release acidic monomers as they break down, with negative effects on remodeling tissue [Cordewener et al., 2004]. An additional attractive attribute of nonwoven polyamide fabrics is that they are strong yet exceedingly flexible [Moeschel et al., 2002], allowing them to adapt to a variety of sizes and shapes of lesion gaps. Finally, and perhaps most importantly, recent work from our laboratory demonstrates the promise of polyamide nanofibers in strategies for repair of the injured spinal cord [Meiners et al., 2007]. The work described in the Meiners et al. [2007] study utilized a nanofibrillar fabric comprised of randomly oriented electrospun polyamide nanofibers, itself randomly folded upon implantation into the lesioned spinal cord. This implant was associated with several promising attributes, including promotion of modest axonal regrowth with concurrent reduction of cavitation, infiltration of inflammatory cells, and glial scarring, yet it failed to orient the growth of regenerating neuronal processes. Different methodologies were considered to overcome this problem. Because neurites extend along the axis of the nanofibers [Yang et al., 2005], nanofibers with a parallel orientation would appear to be ideal for encouraging targeted axonal regeneration in the damaged spinal cord. However, aligned nanofibers exhibit considerable rigidity in comparison to randomly oriented nanofibers [Lee et al., 2005], and thus implants of aligned nanofibers may cause further damage to the delicate spinal cord tissue. As an alternative, we developed a spinal cord prosthetic (SCP) that incorporates narrow strips of randomly oriented nanofibers that are longitudinally bundled to provide appropriate geometric cues for axonal regrowth. (See Figure 1.) The advantage of such an SCP is manifest in cases involving large injuries with extensive tissue loss. While the design of this

A Nanofibrillar Prosthetic Modified with Fibroblast Growth Factor-2…

329

SCP provided benefits for axonal regeneration, a large body of work suggests that the nanofibrillar SCP alone will be unlikely to promote optimal axonal regeneration and functional recovery following spinal cord injury in the absence of growth promoting chemical cues. As an example, axonal regrowth on the randomly folded nanofibrillar implant described in Meiners et al. [2007] was enhanced in the presence of a neurite outgrowth promoting peptide identified by our research group [Meiners et al., 2001; Ahmed et al., 2006] and derived from the amino acid sequence of the extracellular matrix molecule tenascin-C. Fibroblast growth factor-2 (FGF-2) is another growth promoting chemical cue that is perhaps an even more promising candidate for inclusion in strategies to repair spinal cord injury than the tenascin-C-derived peptide. In addition to stimulating neurite outgrowth, FGF2 guides extending neuronal processes, promotes neuronal survival, and encourages angiogenesis [Teng et al., 1999; Webber et al., 2005; Gill et al., 2006; Heinzman et al., 2008]. The potential of FGF-2 in spinal cord injury repair is supported by ample experimental evidence. In one study, FGF-2 was expressed by adenoviral injection within glia in the dorsal spinal cord, resulting in regeneration of crushed axons within the dorsal root and significant recovery of thermal sensory function [Romero et al., 2001]. In another study, continuous intrathecal administration of soluble FGF-2 to the spinal cord following contusion injury significantly improved recovery of hindlimb function [Rabchevsky, 2000]. Our own data, presented in this chapter, demonstrate that an SCP covalently modified with FGF-2 can increase axonal regeneration and hindlimb motor functional recovery in the over-hemisected rat thoracic spinal cord in comparison to an unmodified SCP. An additional point can be made about the significant practical advantage of utilizing immobilized FGF-2 as opposed to soluble FGF-2. FGF-2-modified nanofibers can be stored in their dry state for 6 months at 4o C with retention of significant biological activity (Nur-EKamal et al., 2008). In contrast, soluble FGF-2 is notoriously unstable, with an approximate half-life in tissue culture media of 6-8 hours (Caldwell et al., 2004). The stability of bound FGF-2 is a promising attribute that could lend itself to a grafting material designed to be prefabricated in bulk, trimmed to fit into a variety of defects, and ready for use as the need arises, all major advantages for its proposed future use in the clinic. The design of the FGF-2modified SCP suggests that it would also be of benefit for applications involving repair of damage within the peripheral nervous system (PNS), as will be discussed below.

MATERIALS AND METHODS Polyamide Nanofibers. Randomly oriented polyamide (proprietary composition) nanofibers were electrospun onto aluminum foil by Donaldson Co., Inc. (Minneapolis, MN) from a blend of two polymers [(C28O4N4H47)n and (C27O4.4N4H50)n]. The nanofibrillar mat was cross-linked in the presence of an acid catalyst. The nanofibers were carefully peeled off the foil and were left unmodified or were covalently modified with a proprietary polyamine polymer by Surmodics, Inc. (Eden Prairie, MN) to provide functional groups for attachment of bioactive molecules. Nanofibers and amine-modified nanofibers were then exposed to UV light for 15-30 min for sterilization. The amine-modified nanofibers were in turn covalently modified with FGF-2 under sterile conditions as we have described previously (Nur-E-Kamal et al., 2008).

330

Sally Meiners, Suzan L. Harris, Roberto Delgado-Rivera et al.

Manufacture of SCPs. Strips of the unmodified or FGF-2-modified nanofibrillar fabric were cut using a Stoelting (Wood Dale, IL) tissue slicer to a width of 0.5 mm to allow ample room for axonal fasciculation and to a length of 1 cm. The strips were dipped one by one into SeaPrep® agarose (Cambrex Bioscience Rockland, Inc., Rockland, ME) (2% in dd H2O) and stacked next to and on top of each other to result in a device ~2 mm high and 3 mm wide when hydrated. The device was chilled at 4oC, cut into 2 mm lengths, and kept in a sterile, moist environment until implanted into the lesioned rat spinal cord as described below. SeaPrep® agarose has a gelling temperature of 8-17oC, and remains semi-liquid at room temperature. Thus it provides a liquid or semi-liquid medium at body temperature to allow for ready infiltration of cells and fasciculation of axons in between the nanofiber layers. Surgical Procedure and Postoperative Care. All animal procedures were performed in strict accordance with institutional guidelines and approved animal protocols (Institutional Animal Care and Use Committee). The animal subjects were adult female Sprague Dawley rats (250-260 g) (Hilltop Laboratories Animals, Inc., Scottsdale, PA). The rats were divided into three groups of five rats each. The first group received a multi-layered implant that incorporated unmodified nanofibers, while the second group received a multi-layered implant that incorporated FGF-2-modified nanofibers. The third group received an injury but no implant (injury only control group). The implantation procedure was done as follows. Rats were anesthetized using isoflurane administered in a fume hood followed by ketamine/xylazine (75 mg/kg + 10 mg/kg, intraperitoneal (IP)). They were also given buprenorphine (0.05 mg/kg, delivered subcutaneously) before surgery for post-operative pain. In addition, bupivocaine was used as a topical anesthetic. Using aseptic rodent surgery techniques, an incision was made over the thoracic region of the spinal cord. A double laminectomy was performed at thoracic level 8-9. Irridectomy scissors were used to make a transverse cut 2 mm deep to the dorsal columns of the spinal cord. Another transverse cut was made 2 mm caudal to the first, and the tissue in between was removed with scissors to a depth of 2 mm. After creating the lesion, a sterile SCP (unmodified or modified with FGF-2) made on the prior day was placed in the cavity with the 2 mm long axis longitudinal to the spinal cord. Control animals were subjected to the identical surgery but received no SCP. The musculature was sutured with silk, and the skin was closed with stainless steel surgical clips. Immediately post-surgery, rats were placed in a cage set on top of a heating pad set to 37°C until consciousness was regained. (Ambient cage temperature was no more than 85°F.) They were provided with means to move away from the heating pad once they were awake. Animals were monitored at least every 15 minutes until anesthetic recovery. They were then monitored 3 times daily during the first 4-7 days post-surgery for any signs of pain or infections and 1-2 times daily on weekdays and at least once a day on weekends thereafter so that any health problems could be observed as soon as possible. The rats were housed two to a cage and were provided with red transparent plastic tubes (3 inches in diameter and 6 inches long) for nesting once they regained enough rear hindlimb mobility to allow maneuvering in and out of the tube. This generally did not occur for at least 7 days post surgery. Sterile lactated Ringer’s solution was administered subcutaneously post-surgery 3 times daily for a total of 15 ml per day until the animals were observed to be drinking on their own. Apple slices (sometimes dipped in peanut butter) were also provided to stimulate the appetite and to provide extra hydration for 4-5 days. In addition, the rats were given analgesics (buprenorphine, 0.05 mg/kg delivered subcutaneously) and antibiotics (enrofloxacin (Baytil),

A Nanofibrillar Prosthetic Modified with Fibroblast Growth Factor-2…

331

5 mg/kg delivered subcutaneously) immediately after surgery. Buprenorphine was given 2-3 times daily for 1-2 days, and enrofloxacin was given 2 times daily for at least 3 days. Evacuation of bladders was done 3 times daily until autonomic bladder function was restored (usually 4-7 days). Surgical clips were removed after 2 weeks. Functional Assessment. Recovery of hindlimb locomotor function was assessed on a daily basis for 3 weeks using the 21 point Basso, Beattie, Bresnahan (BBB) locomotor rating scale to evaluate open field locomotion (Basso et al., 1995). The BBB scale has the advantage of providing an assessment tool spanning complete paralysis to normal hindlimb locomotion by rating a wide variety of parameters observed during recovery of function. Although this scale was designed for assessment of open-field locomotion recovery after contusion injuries, it has been successfully applied to the assessment of over-hemisection injuries as well (Deumens et al., 2006). Immunohistochemistry. Following postoperative periods of 3 weeks, animals were reanesthetized using isoflurane followed by ketamine/xylazine (75 mg/kg + 10 mg/kg, IP) and perfused transcardially first with 0.9% NaCl in 0.1 M sodium phosphate buffer containing 50 U/ml heparin (Sigma Chemical Co., St. Louis, MO) and then with 0.1 M sodium phosphate buffer containing 4% paraformaldehyde fixative. Spinal cords were removed, and sagittal sections were cut on a cryostat to a thickness of 20 µm. Sections were mounted on plus (+) gold glass slides (Fisher Scientific, Inc., Hampton, NH). They were then immunolabeled for axons with a mouse monoclonal antibody against neurofilament-M (Millipore Corporation, Billerica, MA) (1:500 dilution overnight at room temperature) and a CY3-conjugated goat anti-mouse secondary antibody (1:500 dilution for 1 hour at room temperature). Some sections were double labeled as described with the antibody against neurofilament-M followed by a rabbit polyclonal antibody against calcitonin gene-related peptide (CGRP) (Millipore (formerly Chemicon), Billerica, MA) (1:5,000 dilution overnight at room temperature) and a donkey anti-rabbit secondary antibody (1:300 dilution for 1 hour at room temperature). Other sections were stained with a rabbit polyclonal antibody against collagen IV (Research Diagnostics Inc., Division of Fitzgerald Industries, Concord, MA) (1:500 dilution overnight at room temperature) and a CY2-conjugated donkey anti-rabbit secondary antibody (1:300 dilution for 1 hour at room temperature) to detect blood vessels and Schwann cells; or with a rabbit polyclonal antibody against glial fibrillary acidic protein (GFAP) (Dako Inc., Carpinteria, CA) (1:500 dilution overnight at room temperature) and a CY2-conjugated donkey anti-rabbit secondary antibody (1:300 dilution for 1 hour at room temperature) to detect astrocytes. At least 15-20 sections per spinal cord were examined. Both secondary antibodies were from Jackson Laboratories (West Grove, PA), and all antibodies were diluted in phosphate buffered saline (PBS) containing 0.3% Triton X-100. Images of the labeled spinal cord sections were captured using a Zeiss Axioplan microscope (Carl Zeiss, Inc., Maplegrove, MN) equipped with an epi-fluorescence illuminator and axiovision software.

RESULTS AND DISCUSSION Nanofibrillar SCPs. Electrospun polyamide nanofibers used for the manufacture of the SCP are shown in Figure 1 (top).

332

Sally Meiners, Suzan L. Harris, Roberto Delgado-Rivera et al.

Figure 1. Manufacture of the SCP. (Top) Nonwoven fabric of electrospun nanofibrillar fabric used to manufacture the SCP. Scale bar, 5 μm. (Middle) Schematic diagram of the SCP within the spinal cord. (Bottom) Epi-fluorescence image showing a cross section through an unmodified SCP prior to implantation into the spinal cord. The nanofibrillar strips can be visualized due to their low level of autofluorescence. Note the conduits resulting from the stacking of the strips next to and on top of each other. Scale bar, 100 μm.

The nonwoven nanofibrillar fabric was ~ 2 μm thick and comprised of nanofibers with a median diameter of 180 nm. The nanofibers were interspersed with pores with diameters ranging from 100-800 nm (Schindler et al., 2005). A schematic diagram of the SCP is illustrated in Figure 1 (middle). The SCP was comprised of a bundle of nanofibrillar strips (unmodified or FGF-2-modified) 0.5 mm wide x 2 mm long bundled together with low gelling temperature agarose. Each device was approximately 6 nanofibrillar strips wide and 45 nanofibrillar strips high and was implanted into the over-hemisected spinal cord with the 2 mm long axis longitudinal to the spinal cord. A cross section through an SCP made on the day prior to implantation and stored at 4oC until use (Figure 1, bottom) revealed that the nanofibrillar strips within the device resembled a stack of bricks laid in a staggered pattern, providing both surface area for cellular growth and agarose filled conduits in between the nanofibrillar layers. Axonal Regrowth on Unmodified and FGF-2-Modified SCPs. Immunohistochemical techniques were employed utilizing epi-fluorescence microscopy and an antibody against neurofilament-M, a general axonal marker, to visualize axons 3 weeks after injury (Figure 2).

A Nanofibrillar Prosthetic Modified with Fibroblast Growth Factor-2…

333

The nanofibrillar strips within the implanted SCPs were detectable as a consequence of their low level of autofluorescence. Some axons were observed within SCPs that incorporated strips of the unmodified polyamide nanofibrillar fabric (Figure 2, top). The axons were observed both associated with the nanofibrillar surface (left-hand arrow) and in between the nanofibrillar layers (right-hand arrow), although it is possible that the latter were growing on nanofibers out of the focal plane of the section or on cells such as Schwann cells (see below) that were themselves associated with nanofibers. More, and apparently longer, neurofilament-M-labeled axons were observed within SCPs that incorporated nanofibrillar strips covalently modified with FGF-2 in comparison to SCPs that incorporated unmodified nanofibrillar strips (Figure 2, middle). The axons grew with the correct longitudinal orientation within both types of SCP, while agarose alone provided no guidance to the extending processes (data not shown). No axons were observed within the SCPs immediately after injury, indicating that the observed axons were regenerating as opposed to axons spared in the lesion process (data not shown). Some axons were observed in the lesion site in injury only control animals, but these were for the most part short, fragmented, and randomly oriented (Figure 2, bottom).

Figure 2. Axonal elongation onto SCPs in the injured spinal cord. Rostral is to the left. Immunostaining was performed using an antibody against neurofilament-M. Neurofilament M-labeled axons were detected using epi-fluorescence microscopy. Nanofiber strips are faintly visible due to autofluorescence. The lesion edge is marked with asterisks. (Top) Axons were observed within the SCP containing bundled strips of unmodified nanofibers 3 weeks after injury. An axon apparently growing on the surface of a nanofibrillar strip is marked with an arrow pointing to the left, whereas an axon growing in between nanofibrillar strips is marked with an arrow pointing to the right. (Middle) More axons were seen within the FGF-2-modified prosthetic. A large degree of fasciculation was observed. (Bottom) Some axons were observed in the lesion site in injury only controls, but they grew with no particular orientation.

334

Sally Meiners, Suzan L. Harris, Roberto Delgado-Rivera et al.

Figure 3. Morphology of regenerating axons. Immunostaining was performed using an antibody against neurofilament-M. Axons were detected using epi-fluorescence microscopy. (Top) Axons within unmodified SCPs demonstrated relatively little branching. (Middle) Axons within FGF-2-modifed SCPs had a more complex morphology, with abundant branching. The boxed area is enlarged (bottom) to highlight the ramified structure.

Inspection of axons at higher magnification revealed fasciculated neuronal processes within both the unmodified (Figure 3, top) and FGF-2-modified (Figure 3, middle, bottom) SCPs. However, axons within the FGF-2-modified devices demonstrated a more complicated morphology, with an increase in axon branching. In another study, Szbenyi et al. [2001)] demonstrated that local application of FGF-2 (presented as FGF-2-coated polystyrene microspheres that were first covalently coupled to heparin) promoted branching in vitro of

A Nanofibrillar Prosthetic Modified with Fibroblast Growth Factor-2…

335

embryonic sensory motor cortical neuronal axons. Distal regions of the axon, in particular the growth cone, were more likely to respond to the branch-inducing effects of the growth factor, suggesting that FGF-2 in target regions of the developing cortex may influence target recognition and synapse formation through arborization of interstitial collaterals. In a similar fashion, FGF-2 immobilized to nanofibers in the injured spinal cord may enhance neuronal plasticity and synapse formation may encourage ramification of regenerating axons. Some sections were double labeled using antibodies against neurofilament-M (red) as well as CGRP (green) (Figure 4). CGRP is a marker for primary peptidergic sensory neurons [Xu et al., 1999; Jones et al., 2001], although it can also be expressed in motor neurons [Ramer et al., 2003]. Evidence for synapse formation was observed in the form of swellings (arrows), or boutons, along the length of the CGRP-labeled processes in close proximity to the neurofilament-M-labeled axons. These “boutons en passant” were seen within both unmodified and FGF-2-modified SCPs. Examination of the sections using rhodamine (CY3) and fluorescein (CY2) optics revealed that many of the CGRP-labeled axons did not appear to be labeled with the antibody against neurofilament-M (data not shown). This observation is in agreement with that of Petruska et al. [2002] that some, but not all, CGRP-positive dorsal root ganglion neurons express neurofilament-M.

Figure 4. Boutons en passant. Immunostaining was performed using antibodies against neurofilamentM (red) and CGRP (green). Axons were detected using epi-fluorescence microscopy. CGRP-labeled axons within unmodified (top) and FGF-2-modified (bottom) SCPs demonstrated localized swellings (arrows) along their length that were characteristic of boutons en passant. The nanofibrillar layers in the bottom panel autofluoresced green. Scale bar, 25 μm.

336

Sally Meiners, Suzan L. Harris, Roberto Delgado-Rivera et al.

Recovery of Hindlimb Motor Function. Results from the axon labeling studies were most encouraging in regard to the regenerative performance of the FGF-2-modified SCP. However, given the difficulties inherent in quantifying axonal regrowth by immunohistochemical methods, functional assays provide the best means for evaluation of an intervention to facilitate spinal cord repair. Indeed, even data indicative of boutons en passant (Figure 4), or data from tract tracing experiments demonstrating the growth of axons off the implant and into the caudal spinal cord [Teng et al., 2002], do not provide information about restoration of functional connections. Therefore, BBB locomotor scores were determined for rats that received an unmodified SCP, an FGF-2-modified SCP, or an identical over-hemisection lesion with no implant. BBB scores were significantly improved as early as 3-4 days following injury for rats that received an FGF-2-modified SCP in comparison to an unmodified SCP (Figure 5). Both groups continued to improve on the BBB rating scale throughout the 21 days of the experiment, but the FGF-2-modified SCP group showed significant improvements over the unmodified SCP group at the majority of the time points. Because the enhanced impact of the growth factor-modified device was first noted at such an early time point, before axonal regeneration could reasonably be expected to occur, it is likely that the FGF-2-modified SCP had an anti-lesion and/or a neuroprotective effect in addition to or instead of an axonal regeneration effect.

Figure 5. Hindlimb functional assessment with the BBB locomotor rating scale. Curves represent the mean +/- the standard error of the mean (n = 5 rats per group). Animals were tested daily for 3 weeks. Rats that received an FGF-2-modified SCP (white circles) showed enhanced functional recovery in comparison to rats that received an unmodified SCP (black circles). Single asterisks denote significant differences between the two groups at each time point with p < 0.05; double asterisks denote significant differences with p < 0.01 (Student’s T-test).

A Nanofibrillar Prosthetic Modified with Fibroblast Growth Factor-2…

337

Figure 6. Evidence for buildup of pressure within the SCP. Rostral is to the left. Immunostaining was performed using an antibody against neurofilament-M. Axons within an unmodified SCP were detected using epi-fluorescence microscopy. A large cavity suggestive of a fluid-filled cyst was observed at the caudal end of the device.

On the other hand, the degree of functional recovery afforded by the unmodified SCP did not vary significantly from the extent of spontaneous functional recovery observed for injury only control animals (data not shown). The spontaneous recovery for the latter was probably due to lingering function of the central pattern generator for locomotion present in rats [Babu and Namasivayam, 2008]. This result was surprising given clear evidence that the nanofibrillar conduits provided guidance for regenerating neuronal processes (Figure 2) and in turn suggests that some other property of the implant impeded optimal functional recovery in spite of the oriented regrowth. Inspection of the entire lesion site revealed evidence of a fluid-filled cyst at the caudal end of the implant. Cyst formation occurred to varying degrees and was observed for both unmodified (Figure 6) and FGF-2-modified SCPs. Importantly, somewhat more regrowth within the unmodified SCP is visible in the section shown in Figure 6 in comparison to the section shown in Figure 2, again underscoring the vagaries of immunohistochemical assessment of axonal regeneration in comparison to behavioral assessment of functional recovery. The probable cause of the cyst formation was an impediment to the flow of cerebral spinal fluid (CSF) due to the viscosity of the agarose. Therefore, it seems likely that employing a more porous medium to separate the nanofibrillar layers would result in an SCP that allows better axonal regrowth at and beyond the distal end of the implant, with additional behavioral benefits. We previously discussed a multi-layer nanofibrillar fabric engineered by Donaldson Co., Inc. for possible application in spinal cord injury strategies [Meiners et al., 2007]. This fabric incorporated individual nanofibrillar layers that were separated by beads. While this design would be expected to aid in the free flow of CSF, the nanofibrillar layers were only separated

338

Sally Meiners, Suzan L. Harris, Roberto Delgado-Rivera et al.

from each other by about 10 μm. On the other hand, a separation of 50-200 μm is optimal to allow sufficient room for cellular growth and revascularization [Stokols et al., 2004; Yu et al., 2005], necessitating larger beads or a different kind of spacer. Other possibilities for separation of nanofibrillar layers with requisite porosity and spacing include templated agarose scaffolds containing linear channels [Stokols et al., 2006] or cross-linked hydrogels [Pike et al., 2006]. Revascularization and Infiltration of Schwann Cells. Immunostaining with an antibody against collagen IV was done to evaluate ingrowth of blood vessels and infiltration of Schwann cells, both of which express this basement membrane protein following spinal cord injury [Klapka and Werner Muller, 2006; Buss et al., 2007]. An abundance of collagen IVpositive blood vessels (BV) was observed within both unmodified (data not shown) and FGF2-modified SCPs (Figure 7). This is important because revascularization is vital and necessary for the ultimate efficacy of the SCP. Blood vessels supply nutrients to regenerating tissues and precede axons in successful models of regeneration [Loy et al., 2002].

Figure 7. Re-vascularization within an FGF-2-modified SCP. Rostral is to the left. Immunostaining was performed using an antibody against collagen IV. Blood vessels (BV) and Schwann cells (SC) were detected within an FGF-2-modified SCP using epi-fluorescence microscopy. Blood vessels and Schwann cells appeared to grow both on the surface of the nanofibrillar layers (arrows pointing to the right) and in the agarose in between the layers (arrows pointing to the left).

A Nanofibrillar Prosthetic Modified with Fibroblast Growth Factor-2…

339

Collagen IV-positive cells with the characteristic bipolar morphology of Schwann cells (SC) were also observed within unmodified (not shown) and FGF-2-modified SCPs (Figure 7). These cells could be differentiated from blood vessels by their smaller size as well as by their morphology. Thus it is likely that the Schwann cells, in addition to the engineered components of the SCP, encouraged axonal regrowth. Indeed, as observed for cellular relationships within peripheral nerve implants [Kim et al., 2008], axons were generally observed in close proximity to the Schwann cells (data not shown). Finally, as seen for axons (Figure 2), blood vessels and Schwann cells appeared to be associated both with the nanofibrillar surface (right-hand arrows) and the agarose component of the SCP (left-hand arrows). Since FGF-2 is an angiogenic factor in the spinal cord [Hayashi et al., 1999] and furthermore promotes proliferation of Schwann cells [Shen et al., 2008], incorporation of FGF-2 into the agarose (or other spacer constituent employed) as well as onto the nanofibers may prove beneficial for the repair process. Glial Scarring. Astrocytes, generally regarded as permissive for neuronal growth during development of the brain and spinal cord, can become reactive following brain or spinal cord injury and contribute to scar formation and regenerative failure [McLeon et al., 1991; Davies et al., 2004]. GFAP immunostaining followed by epi-fluorescence microscopy was therefore performed to visualize both reactive and non-reactive astrocytes. Reactive astrocytes, characterized by up-regulated GFAP and a hypertrophied morphology, were observed for injury only controls at the lesion edge (Figure 8, right), whereas only non-reactive astrocytes were observed at the lesion edge for animals that received an unmodified SCP (Figure 8, left) or an FGF-2-modified SCP (Figure 8, middle). However, non-reactive astrocytes were not observed within the device itself. Why this should be is unclear since astrocytes readily grow on unmodified and FGF-2-modified nanofibrillar matrices in vitro (Nur-E-Kamal et al., 2008) and on nanofibers removed from the lesioned spinal cord 3 weeks after implantation (data not shown). While lack of scarring is a crucial feature for a successful implant for tissue engineering in the central nervous system (CNS), the presence of non-reactive astrocytes would almost assuredly encourage more robust axonal regeneration and recovery of function than that currently observed. In support of this premise, astrocytes generated from glial-restricted precursor cells promoted vigorous axonal regeneration into and beyond the lesion site when transplanted into dorsal column spinal cord injuries in adult rats [Davies et al., 2006]. Significant improvements in locomotor functional recovery as assessed by grid-walk analysis were also documented. Importantly, astrocytes grown on FGF-2-modified nanofibers become more supportive of neuronal growth when co-cultured with an overlay of neurons in comparison to their counterparts grown on more traditional glass or plastic surfaces (unpublished data). These results suggest that polyamide nanofibers may provide an ideal scaffolding material for the implantation of exogenous non-reactive astrocytes or astrocytic stem cells into the damaged spinal cord. Like astrocytes, infiltrating (Figure 7) and transplanted Schwann cells also promote axonal growth from spinal cord neurons and are generally regarded as permissive for the regeneration process [Bunge, 2008]. However, an advantage of astrocytes over Schwann cells is that they are endogenous to the CNS, and an implant containing astrocytes may create less of a demarcation between “self” and implant, in particular in cases where the implant does not provoke scarring in the distal tissue.

340

Sally Meiners, Suzan L. Harris, Roberto Delgado-Rivera et al.

Figure 8. Exclusion of reactive astrocytes by SCPs. Rostral is to the left. Immunostaining was performed using an antibody against GFAP 3 weeks after injury. GFAP-labeled astrocytes were detected using epi-fluorescence microscopy. The lesion edge is marked with asterisks. Non-reactive astrocytes were observed within the spinal cord tissue distal to unmodified (left) and FGF-2-modified (middle) SCPs but were largely absent from the SCPs themselves. No glial scarring was observed. In contrast, reactive astrocytes were observed migrating from the lesion edge into the lesion center in injury only controls (right).

Hence axons may be more prone to grow through an implant and to exit at the distal end into the spinal cord tissue in the case of an implant containing astrocytes as opposed to one pre-loaded with Schwann cells. Other Applications. During the course of our investigations, a similar implant design for PNS regeneration was reported by Kim et al. [2008]. In this study, nanofibrillar strips comprised of randomly oriented or aligned poly (acrylonitrile-co-methylacrylate) nanofibers prepared by electrospinning were stacked inside a polysulfonene nerve conduit. No agarose or other spacer was necessary because the nanofibrillar strips were constrained within the nerve by the conduit. The nanofibrillar filled conduits were used to bridge a 17 mm long gap in the transected tibial nerve of adult rats. Significantly more axons regenerated through the conduits that incorporated strips of aligned nanofibers in comparison to conduits that incorporated strips of randomly oriented nanofibers, with enhanced behavioral results. The addition of growth factors to the nanofibers was not evaluated in this study. It is likely that the FGF-2-modified SCP described in this chapter could be modified to encourage PNS as well as CNS regeneration, and further, that the mechanical stiffness of aligned polyamide nanofibers would not be a limitation in large peripheral nerves such as tibial and sciatic nerves but would in fact improve the regenerative capacity of the SCP.

CONCLUSION An SCP comprised of longitudinally bundled strips of an electrospun, randomly oriented polyamide nanofibrillar fabric that were separated with agarose encouraged guided axonal regrowth through the implant. This property and concomitant behavioral outcomes were improved by the addition of FGF-2. Both types of SCP were permeated by blood vessels and Schwann cells, and neither type promoted glial scarring. On the other hand, quiescent astrocytes failed to populate the SCPs, and cystic cavities were frequently observed at the caudal end of the implant. In conclusion, the FGF-2-modified nanofibrillar SCP is a

A Nanofibrillar Prosthetic Modified with Fibroblast Growth Factor-2…

341

promising prototype for inclusion in strategies to repair spinal cord injury, but further improvements are warranted.

ACKNOWLEDGEMENTS This work was supported by New Jersey Commission on Spinal Cord Research Grant 06A-007-SCR1 to S.M. R.D.-R. gratefully acknowledges the Rutgers-National Science Foundation Integrative Graduate Education and Research Traineeship (IGERT) program on Integratively Engineered Biointerfaces.

REFERENCES Ahmed, I., Liu, H.-Y., Mamiya, P.C., Ponery, A.S., Babu, A.N., Weik, T., Schindler, M., and Meiners, S. (2006). Three-dimensional nanofibrillar surfaces covalently modified with tenascin-C-derived peptides enhance neuronal growth in vitro. J. Biomed. Mater. Res. A, 76, 851-860. Babu, R.S. and Namasivayam, A. (2008) Recovery of bipedal locomotion in bonnet macaques after spinal cord injury: footprint analysis. Synapse, 62, 432-447. Basso, D.M., Beattie, M.S., and Bresnahan, J.C. (1995). A sensitive and reliable locomotor rating scale for open field testing in rats. J. Neurotrauma, 12, 1-21. Bunge, M.B. (2008) Novel combination strategies to repair the injured mammalian spinal cord. J. Spinal Cord Med., 31, 262-269. Buss, A., Pech, K., Kakulas, B.A., Martin, D., Schoenen, J., Noth, J., and Brook, G.A. (2007) Growth-modulating molecules are associated with invading Schwann cells and not astrocytes in human traumatic spinal cord injury. Brain, 130, 940-953. Cordewener, F.W., van Geffen, M.F., Joziasse, C.A.P.S.J.P., Bos, R.R.M., Rozema, F.R., and Pennings, A.J. (2000). Cytotoxicity of poly(96L/4D-lactide): the influence of degradation and sterilization. Biomaterials, 21, 2433-2442. Davies, J.E., Huang, C., Proschel, C., Noble, M., Mayer-Proschel, M., and Davies, S.J. (2006) Astrocytes derived from glial-restricted precursors promote spinal cord repair. J. Biol., 5, article 7. Davies, J.E., Tang, X., Denning, J.W., Archibald, S.J., and Davies, S.J. (2004) Decorin suppresses neurocan, brevican, phosphacan and NG2 expression and promotes axon growth across adult rat spinal cord injuries. Eur. J. Neurosci., 19, 1226-1242. Deumens, R., Koopmans, G.C., Honig, W.M., Maquet, V., Jerome, R., Steinbusch, H.W., and Joosten, E.A. (2006). Limitations in transplantation of astroglia-biomatrix bridges to stimulate corticospinal axon regrowth across large lesion gaps. Neurosci. Lett., 400, 208212. Gill, J.C. and Tsai, P.-S. (2006). Expression of a dominant negative FGF receptor in developing GNRH1 neurons disrupts axon outgrowth and targeting to the median eminence. Biol. Repro., 74, 463-472.

342

Sally Meiners, Suzan L. Harris, Roberto Delgado-Rivera et al.

Hayashi, T., Sakurai, M., Abe, K., Sadahiro, M., Tabayashi, K., and Itoyama, I. (1999) Expression of angiogenic factors in rabbit spinal cord after transient ischaemia. Neuropathol. Appl. Neurobiol., 25, 63-71. Heinzman, J.N., Brower, S.L., and Bush, J.E. (2008). Comparison of angiogenesis-related factor expression in primary tumors under normal and hypoxic growth conditions. Cancer Cell Int., e-pub ahead of print. Jones, L.L., Oudega, M., Bunge, M.B., and Tusynski MH (2001) Neurotrophic factors, cellular bridges and gene therapy for spinal cord injury. J. Physiol., 533, 83-89. Kim, Y.-t., Haftel, V.K., Kumar, S., and Bellamkonda, R.V. (2008) The role of aligned polymer fiber-based constructs in the bridging of long peripheral nerve gaps. Biomaterials, 29, 3117-3127. Lee, C.H., Shin, H.J., Cho, I.H., Kang, Y.-M., Kim, I.A., Park, K.-D., Shin, J.-W. (2005). Nanofiber alignment and direction of mechanical strain affect the ECM production of human ACL fibroblast. Biomaterials, 26, 1261-1270. Loy, D.N., Crawford, C.H., Darnall, J.B., Burke, D.A., Onifer, S.M., and Whittemore, S.R. (2002) Temporal progression of angiogenesis and basal lamina deposition after contusive spinal cord injury in the adult rat. J. Comp. Neurol., 445, 308-324. Meiners, S., Ahmed, I., Ponery, A.S., Amor, N., Harris, S.L., Ayres, V., Fan, Y., Chen, Q., Delgado-Rivera, R., and Babu, A.N. (2007). Engineering electrospun nanofibers for spinal cord repair: A discussion. Polymer Internat., 56, 1340-1348. Klapka, N. and Werner Muller, H. (2006) Collagen matrix in spinal cord injury. J. Neurotrauma, 23, 422-435. McKeon, R.J., Schreiber, R.C., Rudge, J.S., and Silver, J. (1991) Reduction of neurite outgrowth in a model of glial scarring following CNS injury is correlated with the expression of inhibitory molecules on reactive astrocytes. J. Neurosci., 11, 3398-3411. Moeschel, K., Nouaimi, M., Steinbrenner, C., and Bisswanger, H. (2002). Immobilization of thermolysin to polyamide nonwoven materials. Biotech. Bioengineer., 82, 190-199. Nur-E-Kamal, A., Ahmed, I., Kamal, J., Babu, A.N., Schindler, M., and Meiners, S. (2008). Covalently attached FGF-2 to three-dimensional polyamide nanofibrillar surfaces demonstrates enhanced biological stability and activity. Mol. Cell. Biochem., 309,157– 166. Petruska, J.C., Napaporn, J., Johnson, R.D., and Cooper, B.Y. (2002) Chemical responsiveness and histochemical phenotype of electrophysiologically classified cells of the adult rat dorsal root ganglion. Neuroscience, 115, 15-30. Pike, D.B., Cai, S., Pomraning, P.R., Firpo, M.A., Fisher, R.J., Shu, X.Z., Prestwich, G.D., and Peattie, R.A. (2006) Heparin-regulated release of growth factors in vitro and angiogenic response in vivo to implanted hyaluronan hydrogels containing VEGF and bFGF. Biomaterials, 27, 5242-5251. Rabchevsky, A.G., Fugaccia, I., Turner, A.F., Blades, D.A., Mattson, M.P., and Scheff, S.W. (2000). Basic fibroblast growth factor (bFGF) enhances functional recovery following severe spinal cord injury to the rat. Exp. Neurol., 164, 280-291. Ramer, M.S., Bradbury, E.J., Michael, G.J., Lever, I.J., and McMahon, S.B. (2003) Glial cell line-derived neurotrophic factor increases calcitonin gene-related peptide immunoreactivity in sensory and motoneurons in vivo. Eur. J. Neurosci., 18, 2713-2721.

A Nanofibrillar Prosthetic Modified with Fibroblast Growth Factor-2…

343

Romero, M.I., Rangappa, N., Garry, M.G., and Smith, G.M. (2001). Functional regeneration of chronically injured sensory afferents into adult spinal cord after neurotrophin gene therapy. J. Neurosci., 21, 8408-8416. Schindler, M., Ahmed, I., Kamal, J., Nur-e-Kamal, A., Grafe, T.H., Chung, H.Y., and Meiners, S. (2005). Three dimensional nanofibrillar surfaces promote in vivo-like organization and morphogenesis for cells in culture. Biomaterials, 26, 5624-5631. Shen, B., Pei, F.X., Duan, H., Chen, J., and Mu, J.X. (2008) Preparation and in vitro activity of controlled release microspheres incorporating bFGF. Chin. J. Traumatol.,11, 22-27. Stokols, S., Sakamoto, J., Breckon, C., Holt, T., Weiss, J., and Tuszynski, M.H. (2006) Templated agarose scaffolds support linear axonal regeneration. Tissue Eng., 12, 27772787. Stokols, S. and Tuszynski, M.H. (2004) The fabrication and characterization of linearly oriented nerve guidance scaffolds for spinal cord injury. Biomaterials, 25, 5839-5846. Szebenyi, G., dent, E.W., Callaway, J.L., Seys, C., Lueth, H., and Kalil, K. (2001) Figroblast growth factor-2 promotes axon branching of cortical neurons by influencing morphology and behavior of the primary growth cone. J. Neurosci., 21, 3932-3941. Teng, Y.D., Mocchetti, I., Taveira-DaSilva, A.M., Gillis, R.A., and Wrathall, J.R. (1999). Basic fibroblast growth factor increases long-term survival of spinal motor neurons and improves respiratory function after experimental spinal cord injury. J. Neurosci., 19, 7037-7047. Teng, Y.D., Lavik, E.B., Qu, X., Park, K.I., Ourednik, J., Zurakowski, D., Langer, R., and Snyder, E.Y. (2002) Functional recovery following traumatic spinal cord injury mediated by a unique polymer scaffold seeded with neural stem cells. Proc. Natl. Acad. Sci. U.S.A., 99, 3024-3029. Webber, C.A., Chen, Y.Y., Hehr, C.L., Johnston, J., and McFarlane S (2005). Multiple signaling pathways regulate FGF-2-induced retinal ganglion cell neurite extension and growth cone guidance. Mol. Cell. Neurosci., 30, 37-47. Xu, X.M., Zhang, S.X., Li, H., Aebischer, P., and Bunge, M.B. (1999) Regrowth of axons into the distal spinal cord through a Schwann-cell-seeded mini-channel implanted into hemisected adult rat spinal cord. Eur. J. Neurosci. 11, 1723-1740. Yang, F., Murugan, R., Wang, S., and Ramakrishna, S. (2005). Electrospinning of nano/micro scale poly(L-lactic acid) aligned fibers and their potential in neural tissue engineering. Biomaterials, 26, 2603-2610. Yu, T.T. and Shoicet, M.S. (2005) Guided cell adhesion and outgrowth in peptide-modified channels for neural tissue engineering. Biomaterials, 26, 1507-1514.

Reviewed by Dr. Bonnie Firestein-Miller, Department of Cell Biology and Neuroscience, Rutgers University, Piscataway, New Jersey, USA.

In: Nanofibers: Fabrication, Performance, and Applications ISBN 978-1-60741-947-1 Editors: W. N. Chang © 2009 Nova Science Publishers, Inc.

Chapter 11

FABRICATION, PERFORMANCE, AND BIOMEDICAL APPLICATION OF COLLAGEN-, GELATIN- OR KERATIN-CONTAINING PHBV NANOFIBERS Inn-Kyu Kang1, Zhi-Cai Xing1, Jiang Yuan1, Oh Hyeong Kwon2, Jung Chul Kim3, and Yoshihiro Ito4 1

Department of Polymer Science, Kyungpook National University, Daegu 702-701, Republic of Korea 2 Department of Polymer Science and Engineering, Kumoh National Institute of Technology, Gyeongbuk 730-701, Republic of Korea 3 Department of Immunology, Kyungpook National University, Daegu 700-422, Republic of Korea 4 Nano Medical Engineering Laboratory, RIKEN Advanced Science Institute, 2-1 Hirosawa, Wako, Saitama, 351-0198, Japan

ABSTRACT Electrospinning has recently emerged as a leading technique for the formation of nanofibrous structures made of synthetic and natural extracellular matrix components. In this chapter, nanofibrous scaffolds were obtained by electrospinning a combination of poly(3-hydroxybutyrate-co-3-hydroxyvalerate)(PHBV) and cell attachment factor such as type-I collagen, gelatin and keratin in 1,1,1,3,3,3-hexafluoro-2-isopropanol (HIFP). The resulting fibers ranged from 300 to 800 nm in diameter. Their surfaces were characterized by attenuated total reflection Fourier transform infrared spectroscopy (ATR-FTIR), electron spectroscopy for chemical analysis (ESCA) and atomic force microscopy (AFM). The PHBV and protein components such as collagen and keratin were biodegraded by PHB depolymerase, type-I collagenase and trypsin solution, respectively. The results of cell adhesion experiment showed that NIH 3T3 cells more adhered to the PHBV/protein nanofibrous mats than to the PHBV nanofibrous one. It was also found, 1

Correspondence: Inn-Kyu Kang, Prof. Department of Polymer Science, Kyungpook National University, Daegu 702-701, Republic of Korea. Tel.: +82-53-950-5629. Fax: +82-53-950-6623. E-mail address: [email protected].

346

Inn-Kyu Kang, Zhi-Cai Xing, Jiang Yuan et al. from a BrdU assay, that the PHBV/protein nanofibrous mats could accelerate the proliferation of fibroblast cells more effectively than the PHBV nanofibrous mats, suggesting the good scaffolds for tissue engineering.

Keywords: Electrospinning, nanofibers, gelatin, PHBV, nanofibrous scaffold, tissue engineering, collagen, keratin

1. INTRODUCTION Restoration of organ function utilizing tissue engineering requires the use of a threedimensional scaffold. The scaffold should be mechanically stable and capable of functioning biologically in the implant site [1]. Recently, a variety of techniques have been developed for nano- or submicron fiber fabrication, including phase separation [2,3], electrospinning [4], and self-assembly [5]. Among them, electrospinning provides a straightforward way to fabricate fibrous scaffolds with nano-sized diameter fibers, which mimic the structure of natural extracellular matrix (ECM) [6]. To date, electrospinning has been used for the fabrication of ultra-thin fibrous scaffolds from biodegradable synthetic polymers, such as poly(lactic acid) (PLA) [7,8], poly(glycolic acid) (PGA) [9], poly(lactide-co-glycolide) (PLGA) [4,10], and poly(ε-caprolactone) (PCL) [11]. Poly (3-hydroxybutytric acid-co-3hydroxyvaleric acid) (PHBV) is a well known biodegradable, biocompatible, non-toxic and thermoplastic polyester produced by bacteria [12]. PHBV is being developed and commercialized as an ideal substitute for non-biodegradable polymeric materials in commodity applications because of its biodegradability and easy processability. Very recently, some proteins, which include silk [13, 14], fibrinogen [15], collagen [16], and gelatin [17, 18] have been successfully electrospun. On the other hand, keratin is a chief component found in hair, skin, fur, wool, horns, and feathers. Reinforced with calcium salts, it is also found in hooves, nails, claws and beaks [19]. Keratin can be used in a variety of biomedical applications due to its biocompatibility and biodegradability. The three dimensional product of keratin can be used as a cross-linked implantable biomaterial for soft and hard tissue replacement. Keratin is extremely insoluble so that the primary task is to enhance it’s solubility in water or organic solvent. Till now various attempts have been made to obtain water soluble keratin with 2-mercaptonethanol or mercaptoacetic acid after the reduction of disulphide bonds [20, 21]. Yamauchi et al. has extracted keratin from wool and explored its interaction with cell [22, 23]. Schrooyen et al. also extracted keratin from feathers, and partially modified it [24, 25]. Compared to synthetic polymers, natural biopolymers have a good biocompatibility. However, in general, their processability is rather poor [26]. The electrospinning process provides a promising means for creating a tissue engineered scaffold because it can produce composite nanofibers consisting of biodegradable polymer and biological proteins such as collagen and keratin. The obtained composite nanofibrous structure has two unique features that make it well suited for tissue engineering application [27]. First, most natural extracellular matrices (ECM) are composed of randomly oriented collagens of nanometerscale diameters. The morphology and architecture of the electrospun structure is similar to that of the natural ECM. Second, the electrospun fibrous scaffold has a highly porous structure. A highly porous structure is desirable to allow cell seeding or migration.

Fabrication, Performance, and Biomedical Application…

347

Throughout the material and pore size plays a critical role in both cell attachment and the exchange of nutrient and metabolic waste. In this chapter, PHBV was blended with collagen, gelatin or keratin, and electrospun to produce PHBV/collagen, PHBV/gelatin and PHBV/keratin composite fibrous scaffold. In case of keratin, thiol groups were chemically coupled with iodoacetic acid (IAA) to prevent them from oxidation, giving high solubility for organic solvent. The characteristics of nanofibrous scaffold were examined using ATR-FTIR spectroscopy, electron spectroscopy for chemical analysis (ESCA), and atomic force microscopy (AFM). The biodegradation of the nanofiber mats were evaluated using PHB depolymerase, collagenase and trypsin. The behavior of the fibroblasts on nanofibrous mats was also investigated.

2. ELECTROSPINNING Formhals, in 1934 [28], patented a process whereby an experimental setup protocol was outlined for the production of polymer filaments by using electrostatic force. The process is referred to “electrospinning” when it is used to spin fibers. In other words, electrospinning is a process that forms nanofibers through an electrically charged jet of polymer solution of polymer melt. In order to perform electrospinning, the polymer must be in a liquid form, either as a molten polymer or as a polymer solution. The liquid polymer solution is passed through the electrospinning system to form nanofibers. A basic electrospinning system usually consists of three major components: a high voltage power supply, a spinneret (e.g. a pipette tip/syringe) and a ground collecting. When a charged polymer solution is fed through the spinneret under an external electric field, a suspended conical droplet is formed, whereby the surface tension of the droplet is in equilibrium with the electric field. When the applied electric field is strong enough to overcome the surface tension, a tiny jet is ejected from the surface of the droplet and drawn toward the collecting plate. During the jet propagation toward the collecting plate, the solvent in the jet stream gradually evaporates. The resulting product consists of a non-woven fibrous scaffold with a large surface area-to-volume ratio and a small pore size [29]. The electrospun fiber properties and morphology depend on many parameters, such as solution parameters (viscosity, surface tension, conductivity etc.), processing conditions (voltage, feed rate, temperature etc.) and ambient parameters (humidity, type of atmosphere, pressure etc.).

2.1. Fabrication of Collagen-Containing PHBV Nanofibers Poly (3-hydroxybutyrate-co-3-hydroxyvalerate) (PHBV) (0.7g) and collagen type І (0.3g) were blended at 2 wt% in HFIP under mechanical stirring. The blend solution was delivered to a metal needle (20 G) connected to a high voltage power supply (Chungpa EMT, Seoul, Korea). Upon applying a high voltage, a fluid jet was ejected from the needle. As the jet accelerated towards a grounded collector, the solvent evaporated and a charged polymer fiber was deposited on the collector in the form of a nonafibrous web.

348

Inn-Kyu Kang, Zhi-Cai Xing, Jiang Yuan et al.

2.2. Fabrication of Gelatin-Containing PHBV Nanofibers A transparent polymer solution was prepared by dissolving gelatin and PHBV in TFE with sufficient stirring at room temperature. To examine the effect of gelatin content on fiber morphology, a 6 wt% polymer solution was prepared using different ratios of PHBV and gelatin. In order to examine the effect of polymer concentration on fiber morphology, a PHBV/gelatin (50/50) blend was dissolved using TFE to prepare several solutions with concentrations ranging from 2 to 8 wt%. The polymer solutions were electrospun with 2.0 ml/h of a mass flow rate, 7kV of voltage, and 12cm of the distance between the tip and the collector. After electrospinning, all nanofibrous scaffolds were placed in a vacuum drying oven at room temperature for several days of drying treatment.

2.3. Fabrication of Keratin-Containing PHBV Nanofibers The process to prepare m-keratin includes two steps: extraction and chemical modification. Keratin (12 g) was mixed with urea (250 g), SDS (20 g), 2-mercaptoethanol (50 ml) and water (600 ml) in a 1000 ml round-bottom flask. The pH of the mixed solution was adjusted to 9 using 1 M of NaOH and kept stirring for 12 h at 60ºC. The resulting mixture was filtrated through a glass filter. Subsequently, the filtrate was dialyzed against deionized water containing 0.1 wt % 2-mercaptoethanol to afford a colorless and clear solution. The dialysate was replaced every 12 h and dialysis was stopped after 48 h. The concentration of sulphydryl groups in this solution was measured using a DTNB [30]. By treating the keratin with 2-mercaptoethanol, most of disulfide bonds in the keratin are broken and changed into sulphydryl (SH) groups. The SH groups are very easily oxidized into disulfide bonds again. Therefore, in order to increase water solubility of keratin, it is necessary to protect SH groups from oxidation. The most popular compound used for the protection of SH groups is iodoacetic acid (IAA) [31, 32]. The molar ratio of IAA/SH was set at 5 to give S-(carboxymethyl) keratin. The calculated amount of IAA was mixed with the keratin solution, and then stirred for 1 h at room temperature. After reaction, the amount of the SH group was measured again to evaluate the modification degree. The pH of the mkeratin solution was adjusted to 7 using 1 M of NaOH and dialyzed thoroughly for 3 days. Finally, this solution was lyophilized to obtain modified keratin (m-keratin). The transparent polymer solution for electrospinning was obtained by dissolving mkeratin and PHBV in HFIP with sufficient stirring at room temperature. In order to examine the effect of m-keratin content on fiber morphology, the polymer solution was prepared using the different ratios of PHBV and m-keratin. The blended solution was delivered to a metal needle (18 G) connected to a high voltage power supply. In this study, the typical parameters of electrospinning were as follows. The voltage was 10 KV and the distance between the spinneret and the drum was 15 cm. The feed rate was 1 ml/h for 6 wt % PHBV/m-keratin in HFIP. m-Keratin nanofiber mats need to be cross-linked to reduce their solubility in water. The electrospun keratin nanofibrous mat was cross-linked by treating it with glutaraldehyde vapor and saturated with a 25% glutaraldehyde aqueous solution at room temperature for various time periods. This was followed by treatment with 0.1 M glycine aqueous solution to block unreacted aldehyde groups.

Fabrication, Performance, and Biomedical Application…

349

3. METHODS 3.1. Surface Characterization The ATR-FTIR spectra of the nanofibrous scaffolds were obtained using a FT-IR spectrometer (Jasco-620, Tokyo, Japan). The electrospun scaffolds were analyzed using electron spectroscopy for chemical analysis (ESCA, ESCA LAB VIG microtech, Mt 500/1 etc, East Grin, UK) equipped with Mg Kα at 1253.6 eV and 150 W power at the anode. A survey scan spectrum was taken and the surface elemental compositions relative to carbon were calculated from peak heights with a correction for atomic sensitivity. For analysis of the morphology of the electrospun fibers, the samples were sputter-coated with gold and examined using a field-emission scanning electron microscope (FE-SEM, Hitachi S-4300, Japan). The diameter of the electrospun nanofibers was measured using an image analyzer (TDISE V3.1.73). Topographic images of nanofibrous scaffold (1x1 cm2) were examined using an atomic force microscope (AFM, a Nanoscope Ⅲa controller, Digital Instruments, Santa Barbara, CA, USA) combined with an optical microscope. Tapping mode of AFM was employed to observe the nanofiber. In the tapping mode, the sample oscillates with a high frequency close to its resonant frequency.

3.2. In Vitro Biodegradation The nanofibrous scaffold was cut into rectangles (20 x 20 x 0.05 mm) for in vitro biodegradation test. Each specimen was placed in a test tube containing 10ml of phosphatebuffered saline (PBS, Gibco, pH 7.0) and incubated for requisited time at 37℃. After incubation, the samples were washed and lyophilized for 24 h. To measure the enzymatic degradation of nanofibrous scaffolds, the collagen-containing PHBV nanofibers and gelatin-containing PHBV nanofibers were placed in a PBS containing collagenase type І (10 mg/ml, Sigma) or Pseudomonas stutzeri BM190 depolymerase (0.1 mg/ml). In case of keratin-containing PHBV nanofibers, they were placed in PBS containing collagen type 1 or trypsin (10 mg/ml). After a requisite incubation time, the samples were taken out from the enzyme solution, washed with distilled water and lyophilized for 24 h. Morphological change of nanofibrous scaffolds were examined using a field emission scanning electron microscope (FE-SEM, S-4300, Japan).

3.3. Cell Culture and Biological Assay To examine the interaction of nanofibrous scaffolds with cells, the circular nanofibrous scaffolds were fitted in a 24-well culture dish and subsequently immersed in a DMEM medium containing 10% fetal bovine serum (FBS) (Gibco, Japan) and 1% penicillin Gstreptomycin (Gibco, Japan). One ml of NIH 3T3 cell solution (5 × 104 cells/cm2) was added to the sample sheet and incubated in a humidified atmosphere of 5% CO2 at 37℃ for 4h to see the early cell adhesion activity on the nanofibers. After incubation, the supernatant was removed, washed twice with a PBS, and fixed in a 2.5% glutardialdehyde aqueous solution

350

Inn-Kyu Kang, Zhi-Cai Xing, Jiang Yuan et al.

for 20 min. The sample sheet was then dehydrated, dried in a critical point drier, and finally sputter-coated with gold. The surface morphology of the samples was then observed with an FE-SEM. The proliferation of NIH 3T3 cells seeded on the nanofibrous scaffolds were determined using a colorimetric immunoassay for the quantification of cell proliferation, based on the measurement of 5-bromo-2'-deoxyuridine (BrdU) incorporation during DNA synthesis[33, 34]. The BrdU ELISA (EL × 800, Roche Molecular Biochemicals, USA) was performed according to the manufacturer,s instruction. Briefly, after cell culture for 48 h, BrdU labeling solution was added to each well and allowed to incorporate into the cells in a CO2-incubator at 37oC for a further 20h. Subsequently the supernatant in each well was removed by pipetting and washed with PBS two times. The cells were treated with 0.25% trypsin-EDTA (Gibco, Japan) and harvested by centrifugating the cell solution at 1000 rpm for 15 min. The harvested cells were mixed with FixDenat solution to fix the cells and denature the DNA and incubated for 30 min. Subsequently the diluted anti-BrdU-peroxidase (dilution ration = 1: 100) was added and kept at 20oC for 120min. After removing the unbound antibody conjugate, 100μl of substrate solution were added for 20 min and the reaction stopped by adding 1 M H2SO4 solution. The solution was transferred to a 96-well plate and measured within 5 min at 450 nm with a reference wavelength at 690 nm using an ELISA plate reader. The blank corresponded to 100 μl of culture medium with or without BrdU.

3.4. Wound Healing Test and Histological Examination Wound healing test was carried on Wistar rat model. Wistar rats (180-200g) were anesthetized by intramuscular injection of Ketamine and Xylaxinl. The skin of the animal was shaved and disinfected in 70% ethanol. After disinfection, the surgical site was prepared for aseptic surgery. A full thichness skin wound of 1 cm in diameter was prepared by excising the dorsum of the Wistar rat. The excised wound was covered with an equal size of electrospun nanofibrous mats (PHBV and PHBV/Col, PHBV and PHBV/keratin). The wound, which was treated with cotton gauze, was used as control. The area of wound was measured at 4, 7 and 9 days. The percentage of wound healing is defined as B/A ×100%, where A is the initial wound area and B is the wound area after a fixed time interval. At the 9th postoperative day, the Wistar rats were killed. A fixation in 10% formaldehyde was immediately carried out after macroscopic observation of wound status. A skin wound tissue was cut from the central regions of the wound in a strip of 0.5 cm by 2.5 cm, embedded in paraffin wax, sectioned (4 µm) and stained with hematoxylin and eosin (H and E) staining. The results of healing effects were histologically investigated.

3.5. Statistical Analysis Results are displayed as the mean ± standard deviation. Statistical differences were determined by student's two-tailed t test. Scheffe,s method was used for multiple comparison tests at level of 95%.

Fabrication, Performance, and Biomedical Application…

351

4. RESULTS 4.1. Morphology of Electrospun Nanofibrous Scaffolds Figure 1 shows the SEM images of electrospun nanofibrous scaffolds obtained from different processing parameters. The image showed continuous fiber morphologies and did not contain beads independent of the kind of polymer. The fiber diameter of PHBV was in the range of 300 to 600 nm and the diameter was decreased by the incorporation of collagen into PHBV, probably due to the increment of polarity. Wagner et al. [35] have fabricated elecrospun polyetherurethaneurea (PEUU)/collagen scaffolds by combining PEUU with type I collagen at various ratio. In their result, SEM images revealed continuous fiber morphologies spun at all ratios examined.

Figure 1. SEM micrographs of nanofibrous scaffolds. (a) PHBV, (b) PHBV-Col, (c) collagen.

To evaluate the effect of the dope composition and concentration, electrospinning was performed with various composite ratios (PHBV/gelatin at ratio of 30/70, 50/50, 70/30) and different polymer concentrations (2, 4, 6, 8 wt%). The morphological structures of electrospun PHBV/gelatin nanofibers in different mixing ratio are shown in Figure 2. As shown in the SEM photographs of Figure 2, a very uniform and finest nanofiber could be obtained at different mixing ratios. To consider the mechanical properties and biocompatibility of nanofibrous scaffold, 50/50 mixing ratio of dope solution was fixed. Figure 3 shows PHBV/gelatin nanofiber depending on the concentration of 2 ~ 8 wt% at 50/50 mixing ratio. At a concentration of 4 and 6wt%, continuous nanofibers without beads could be obtained. As the content of gelatin and the concentration of solutions increased, the diameter of nanofibers was getting thicker and the distribution of fibers was getting broader. In this study, a 6 wt% (50/50) concentration of polymer solution was fixed throughout further experiments. Image analysis of PHBV/gelatin nanofibers (6 wt%, 50/50) revealed that their diameters ranged from 400nm to 1μm, as shown in Figure 3. A series of ratios of PHBV and m-keratin (10/0, 7/3, 3/7, 0/10) were dissolved and electrospun. When electrospun m-keratin (10/0) using HFIP, the resulting fibers contained many beads. This is due to broad molecular weight distribution and low dissolvability of mkeratin (Figure 4a). Fibers that had good image were obtained when co-electrospinning mkeratin with PHBV. Figures 4b, c and d showed the SEM images of the electrospun PHBV, PHBV/m-keratin (7/3), and PHBV/m-keratin (3/7) nanofibrous mats that were obtained under optimum conditions.

352

Inn-Kyu Kang, Zhi-Cai Xing, Jiang Yuan et al.

Figure 2. SEM micrographs of electrospun PHBV/gelatin nanofibrous mats using TFE solutions as the solvent (6 wt%). (a) 30/70, (b) 50/50, (c) 70/30.

Figure 3. SEM micrographs of electrospun fibers of PHBV/gelatin (50/50) using TFE solutions as a function of concentration. (a) 2 wt%, (b) 4 wt%, (c) 6 wt%, (d) 8 wt%.

The diameters of PHBV, PHBV/m-keratin (7/3) and PHBV/m-keratin (3/7) fiber were about 815±98, 720±124 and 487±161nm, respectively. These data were calculated by Image J 1.38 software (http://rsb.info.nih.gov/ij/download.html). It is well known that polar polymer has a higher conductivity than non-polar polymer. Keratin is a polar biopolymer that contains many polar groups such as amide and carboxyl groups. When PHBV is blended with keratin and electrospun, the conductivity of the blend solution will increase, thus leading smaller diameter of PHBV fibers.

Fabrication, Performance, and Biomedical Application…

353

Figure 4. SEM micrographs of nanofibrous mat: (a) 6% m-keratin in HFIP, (b) 6% PHBV in HFIP, (c) PHBV/m-keratin=7/3, 6% in HFIP, and (d) PHBV/m-keratin=3/7, 6% in HFIP.

4.2. Characterization of Nanofibrous Scaffolds Figure 5 shows the ATR-FTIR spectra of nanofibrous scaffolds. The ATR-FTIR spectrum of PHBV-Col (Figure 5 (a)) showed absorptions at 1656 and 1538 cm-1 based on the conformation of amide І and П of collagen, respectively. In Figure 5 (b), the absorption at 1733 cm-1 is attributed to the ester groups of PHBV. Changes in the chemical structure of nanofibrous scaffolds were investigated using ESCA. Figure 6 shows the ESCA survey scans of the nanofibrous scaffold surfaces. The collagen nanofiber scaffold (Figure 6 (c)) showed three peaks corresponding to C1s (binding energy, 285eV), N1s (binding energy, 400eV) and O1s (binding energy, 532eV), while the PHBV nanofiber scaffold (Figure 6 (a)) showed two peaks corresponding to C1s and O1s. The chemical composition of the nanofibrous scaffolds, calculated from the ESCA survey scan spectra, are shown in Table І. The oxygen content (36.8%) of the PHBV nanofiber surface decreased to 32.0% due to the incorporation of collagen (PHBV-Col), while the nitrogen content increased to 6.4%, indicating the presence of collagen on the surface. In addition, we have calculated the data from Table І further. If collagen diffused into PHBVcollagen nanofiber homogenously, the theoretical N value should be 5.73% according to the 30% collagen content in the mixed fiber. This value was lower than the determined value of 6.4%, which meant the collagen was distributed uniformly and concentrated on the surface.

354

Inn-Kyu Kang, Zhi-Cai Xing, Jiang Yuan et al.

Figure 5. Attenuated total reflection-Fourier transform infrared spectra of nanofibrous scaffolds: (a) collagen, (b) PHBV-Col.

In order to study the surface morphology of the PHBV and PHBV-Col nanofibers, an atomic force microscope (AFM) image was studied using a tapping mode and it was expressed in the form of phase image as shown in Figure 7. As the results, the PHBV nanofiber surface (a) showed a relatively homogeneous color pattern while PHBV/Col (b) showed a heterogeneous color pattern, thus showing the presence of collagen in PHBV.

Figure 6. ESCA survey scan spectra of (a) PHBV, (b) PHBV-Col, and (c) collagen.

Fabrication, Performance, and Biomedical Application…

355

Table І. Chemical composition of the nanofibrous scaffolds calculated from the ESCA survey scan spectra Atomic percent (%) Sample

C

O

N

PHBV

62.9

36.8

0.3

PHBV-Col

61.6

32.0

6.4

Collagen

57.3

23.6

19.1

Figure 7. AFM phase images of nanofibrous scaffold surfaces: (a) PHBV, (b) PHBV-Col.

Figure 8 shows the ATR-FTIR spectra of nanofibrous scaffolds. The ATR-FTIR spectrum of gelatin (a) showed two characteristic peaks at around 1650 and 1540 cm-1, based on amide Ⅰ and Ⅱ as expected. The spectrum of PHBV/gelatin (Figure 8(b)) also showed absorption peaks at around 1650 and 1540 cm-1, based on amide І and П of gelatin, corresponding to the stretching vibration of C=O bond, and coupling of bending of N-H bond and stretching of C-N bond, respectively. The absorption at 1733 cm-1 is attributed to the ester groups of PHBV (Figure 8 (c)). Changes in the chemical structure of nanofibrous scaffolds were investigated using ESCA. Figure 9 shows ESCA survey scans of the nanofibrous scaffold surfaces. As expected, the co-electrospun PHBV/gelatin scaffold shows three peaks corresponding to C1s (binding energy, 285eV), N1s (binding energy, 400eV), and O1s (binding energy, 532eV). The chemical compositions of the nanofibrous scaffolds calculated from the ESCA survey scan spectra are shown in Table П. The oxygen content (32%) of the PHBV nanofiber surface was decreased by the incorporation of gelatin (PHBV/gelatin, 25%). On the other hand, nitrogen (7%) was found on the PHBV/gelatin scaffold surface. To study the surface morphologies of PHBV and PHBV/gelatin nanofiber surfaces, AFM image was examined using a tapping mode and expressed as phase images. On the PHBV nanofiber surface, a relative homogeneous pattern was observed as shown in Figure 10(a). On the PHBV/gelatin nanofiber surface (Figure 10(b)), a phase-separated structure appeared,

Inn-Kyu Kang, Zhi-Cai Xing, Jiang Yuan et al.

356

showing the distribution of gelatin on the PHBV matrix well. The phase-separated structures are probably attributed to the globular structure and hydrophilicity of gelatin.

Figure 8. Attenuated total reflection-Fourier transform infrared spectra of a gelatin (a), PHBV/gelatin (b) and PHBV (c) nanofiber mats.

Table II. Chemical composition of a gelatin, PHBV and PHBV/gelatin (5:5 wt %) nanofiber mats calculated from ESCA survey scan spectra Atomic percent (%) Substrate Gelatin

C 44

O 46

N 10

PHBV

68

32

0

PHBV/gel

68

25

7

The water contact angle of PHBV film, PHBV nanofiber mats, as determined by the sessile drop method, was 108º and 110º, respectively [36]. It was indicated that PHBV had a high hydrophobic nature. As expected, PHBV/m-keratin mats showed much better wettability than PHBV due to the introduction of m-keratin. The water drops on the PHBV/m-keratin disappeared after a few seconds. This indicated that the PHBV/m-keratin nanofibers had much better hydrophilicity than the PHBV mats. This property will be very helpful for cell adhesion. The ATR-FTIR spectra of the PHBV fibers (a), m-keratin fibers (b) and PHBV/mkeratin (c) are shown in Figure 11. The strong absorption peak at 1725 cm-1 is attributed to the ester groups of PHBV (Figure 11a). Figure 11b illustrates that common bands of mkeratin appeared at approximately 1650 cm-1 (amide I) and 1545 cm-1 (amide II), corresponding to the stretching vibration of C=O bond (amide I), the coupling of bending of

Fabrication, Performance, and Biomedical Application…

357

N–H bond (amide II) and the stretching of C–N bond, respectively. The amide I bond absorption at 1650 cm-1, that appeared in the PHBV/m-keratin spectrum (Figure 11c), was obviously due to the presence of keratin. When blended and electrospun with PHBV, the peak of amide I did not shift, indicating that the conformation of m-keratin was maintained.

Figure 9. ESCA survey scan spectra of (a) gelatin nanofiber, (b) PHBV nanofiber, and (c) PHBV/gelatin nanofiber.

Figure 10. AFM images represented by phase mode: (a) PHBV fiber, (b) PHBV/gelatin fiber.

358

Inn-Kyu Kang, Zhi-Cai Xing, Jiang Yuan et al.

Figure 11. Attenuated total reflection-Fourier transform infrared spectra of (a) PHBV, (b) m-keratin, and (c) PHBV/m-keratin=7:3.

Changes in the chemical structure of nanofibers were investigated with the ESCA (Figure 12). Compared with PHBV, the newly-appeared peaks at 400 eV in the PHBV/m-keratin obviously could be attributed to nitrogen of keratin. The sulfur peak in PHBV/m-keratin was very weak due to an extremely insufficient amount of content.

Figure 12. Electron spectroscopy for chemical analysis (ESCA) survey scan spectra of nanofiber mat (a PHBV, b m-keratin, c PHBV/m-keratin=7/3).

Fabrication, Performance, and Biomedical Application…

359

4.3. In Vitro Biodegradation Figure 13 illustrates the morphological changes of nanofibrous scaffold surfaces before (a, b, c) and after (d, e, f) incubation with a collagenase Type I aqueous solution. For the results, the surface morphology of the PHBV (a) and PHBV-Col (b) nanofiber scaffolds did not change after they were dipped in a PBS solution. The surface of the collagen (c) nanofiber, however, lost its original fiber morphology, probably due to swelling caused by water. On the other hand, the collagen nanofiber scaffold was partially biodegraded by the treatment of a collagenase Type I aqueous solution, as shown in Figure 13 (f). In the case of PHBV containing 30wt% collagen (PHBV-Col), the fibers preserved their original morphology within 12 h after enzyme treatment. Figure 13 shows the surface morphology of the nanofibrous scaffolds before and after incubation in a PHB depolymerase aqueous solution for 12 h. The surfaces of both PHBV and PHBV-Col (d and e) fibers were severely eroded by the treatment of PHB depolymerase, while the collagen (f) fibers showed very little erosion.

Figure 13. SEM images of nanofibrous scaffolds incubated for 12h in PBS solution (a,b,c) and in collagenase Type І solution (d,e,f): a, d = PHBV; b, e = PHBV-Col; c,f = collagen.

Figure 14 illustrates the morphological changes on nanofibrous scaffold surfaces after incubation in PBS with or without depolymerase (Pseudomonas stutzeri BM190). Before the depolymerase treatment, PHBV/gelatin (Figure 14(a)) exhibited a preserved nanofibrous structure. After 4~6 h incubation in depolymerase solution, the PHBV nanofiber showed very high morphological changes (Figure 14(c, d)). After collagenase treatment, as shown in Figure 15, the PHBV/gelatin fibers broke down and partially adhered to each other after 24 and 48 h incubation time.

360

Inn-Kyu Kang, Zhi-Cai Xing, Jiang Yuan et al.

Figure 14. Biodegradation of PHBV/gelatin (50/50) nanofibrous scaffold by PHB depolymerase solution as a function of incubation time. (a) 0 h, (b) 1h, (c) 4 h, (d) 6 h.

Figure 15. Biodegradation of PHBV/gelatin nanofibrous scaffold by collagenase solution as a function of incubation time. (a) 0 h, (b) 12 h, (c) 24 h, (d) 48 h.

Fabrication, Performance, and Biomedical Application…

361

The morphology of the mats degraded in vitro was examined with the FE-SEM. Figure 16 illustrates the morphological changes of the electrospun PHBV/m-keratin mats during in vitro degradation. After 12 h of degradation in the PHB depolymerase aqueous sollution, a large part of the nanofibers disappeared; only chunks of degraded materials remained (Figures 16 c and d). The fiber-binding points were broken, and the mat was found to be rather brittle and powdered. Fibrous morphology changed slightly when the mat was subjected to degradation in a trypsin solution for 12 h. Even if the fibers were incubated in a trypsin solution for 24 h, the degradation was also not very significant and some fibers began to break down (Figure 16 e and f). It can be seen that the fibers had been severely eroded by the treatment of PHB depolymerase, while little erosion was found in the trypsin solution. These results indicated that PHB depolymerase was more sensitive to PHBV than to keratin. The enzymatic degradation of PHBV films by PHB depolymerase has been reported [37]. It is commonly accepted that m-keratin could be degraded in a trypsin solution [22].

Figure 16. SEM images of the nanofibrous scaffolds incubated in the PHB depolymerase solution and the trypsin solution. (a) PHBV original, (b) PHBV/m-keratin=7/3 fiber original, (c) PHBV fibers incubated for 12h after PHB depolymerase aqueous solution, (d) PHBV/m-keratin=7/3 fibers incubated for 12h after the PHB depolymerase aqueous solution, (e) PHBV fibers incubated for 24h in trypsin aqueous solution, (f) PHBV/m-keratin=7/3 fibers incubated for 24h in trypsin aqueous solution.

Inn-Kyu Kang, Zhi-Cai Xing, Jiang Yuan et al.

362

4.4. Cell-Scaffold Interaction Collagen has an Arg-Gly-Asp sequence that can be recognized by one of the receptors of cell membranes [38, 39]. Figure 17 shows the SEM images of the NIH 3T3 fibroblasts that adhered to the nanofibrous scaffolds when cultured in a Dulbecco’s modified eagle medium containing 10% fetal bovine serum for 4 h. The cells are more adhered on collagen-containing PHBV (PHBV-Col) than the PHBV scaffold. In the case of the collagen nanofibrous scaffold, the cells spread largely and formed clusters with a stratified appearance to the cell layers.

Figure 17. SEM micrographs of NIH 3T3 cells cultured for 4h on electrospun nanofibrous scaffolds. (a) PHBV, (b) PHBV-Col, (c) collagen.

Figure 18 shows the proliferation of cells on the nanofibrous scaffolds when cultured in a DMEM with 10% serum for 20 h. Cell proliferation on the PHBV nanofibrous scaffold was significantly accelerated by the introduction of collagen Type I (PHBV-Col) (P < 0.01). Figure 19 shows the cell viability of fibroblasts cultured for 5 days on the nanofibrous scaffolds. The cell viability on the PHBV-Col was higher than that on the control PHBV (P < 0.05), while it was lower than that on the collagen (P < 0.05). It is considered that the high cell proliferation on the collagen and collagen-containing PHBV (Figure19) may probably be due to the high cell adhesion on the same scaffold (Figure17).

3.0

Absorbance (A450nm-A690nm)

2.5

2.0

1.5

1.0

0.5

0.0

PHBV

P H B V -C ol

collagen

Figure 18. Proliferation of NIH 3T3 cells cultured for 68 h. Data are expressed as mean ± SD (n = 5) for the specific absorbance.

Fabrication, Performance, and Biomedical Application…

363

0.7

Cell viability Absorbance (A570nm)

0.6 0.5 0.4 0.3 0.2 0.1 0.0

PHBV

PHBV-Col

collagen

Figure 19. MTT assay, formazan absorbance expressed as a measure of cell viability from NIH 3T3 cells cultured on to nanofibrous scaffolds for 3 days.

As shown in Figure 20, NIH 3T3 fibroblasts highly spreaded on to the PHBV (Figure 20 c) and PHBV/gelatin scaffold (Figure 20d) compared with culture dish (Figure 20(a)) and PHBV film (Figure 20(b)) after 4 h incubation. NIH 3T3 fibroblasts cultured on PHBV and PHBV/gelatin nanofibers for 1 day are shown in Figure 21. As the results cells adhered to the PHBV/gelatin scaffold (Figure 21(b)) and quickly formed monolayer than that of the PHBV nanofibrous scaffold (Figure 21(a)), showing a good tissue compatibility of the PHBV/gelatin nanofibrous scaffold.

Figure 20. Adhesion of fibroblasts on culture dish (a), PHBV film (b), PHBV (c) and PHBV/gelatin nanofibrous scaffold (d) for 4 hour incubation.

364

Inn-Kyu Kang, Zhi-Cai Xing, Jiang Yuan et al.

Figure 21. Morphology of fibroblasts on PHBV (a) and PHBV/gelatin nanofibrous scaffold (b) for 1 day incubation.

Figure 22 shows cross section images of PHBV/gelatin scaffold after fibroblasts culture. After 3 days incubation, a monolayer morphology was found. After 6 days incubation, a thicker layer was formed, indicating that the composite scaffold allowed cellular penetration or infiltration into the inside of PHBV/gelatin composite fibrous structure.

Figure 22. Cross section image of PHBV/gelatin nanofibrous scaffold after fibroblasts cultured for 3 days (a) and 6 days (b) incubation.

Comparable results for differences in proliferation behavior, expressed as the amount of newly synthesized DNA, are shown in Figure 23. The DNA synthesis of NIH 3T3 cells cultured on the matrices for 24 h revealed that cells could proliferate on all matrices. Compared to PHBV film, fibroblasts proliferation on PHBV nanofibrous scaffold was significantly higher (p < 0.01). In addition, the cell proliferation on PHBV/gelatin nanofibrous scaffold was higher than on the PHBV nanofibrous scaffold (p < 0.05). An MTT assay was used to measure relative cell viability. Figure 24 shows the cell viability of fibroblasts seeded on the nanofibrous scaffolds for different incubation times. Metabolically active mitochondrial dehydrogenases convert the tetrazolium salt MTT to insoluble purple formazon crystals of which the amount is proportional to cell viability[24]. The value of cell viability (Figure 24) means the maximum absorption of MTT formazan (570 nm) collected from the cells seeded on nanofiber scaffold.

Fabrication, Performance, and Biomedical Application…

365

After 3~14 days incubation, the cell viabilities on both PHBV and PHBV/gelatin nanofibrous scaffolds exhibited a similar tendency. After 7 days incubation, compared to PHBV film, the cell viability of fibroblasts cultured on PHBV and PHBV/gelatin nanofibrous scaffolds was significantly enhanced (p< 0.01).

Figure 23. Assessment of NIH 3T3 cell proliferation using the BrdU ELISA. Data are expressed as mean ± SD (n = 5) of the specific absorbance.

Figure 24. MTT assay, Formazan absorbance expressed as a measure of cell viability from NIH 3T3 cells cultured onto nanofibrous scaffolds for 5days. Data are expressed as mean ± SD (n = 5) of the specific absorbance.

366

Inn-Kyu Kang, Zhi-Cai Xing, Jiang Yuan et al.

Ideal tissue engineering scaffold material must support cellular attachment and growth. To evaluate cellular behavior on electrospun fibers, fibroblasts were seeded and cultivated. Figure 25 shows the SEM images of cells that adhered to the nanofibrous mats when cultured in a Dulbecco’s modified eagle medium (DMEM) containing 10% fetal bovine serum for 4 h. As expected, on the surface of all electrospun fibers, more cells were attached and showed much more spread morphology than that of the PHBV film at an earlier culture stage (4 h). In spite of hydrophobic nature of PHBV, cells more adhered onto PHBV fibers than PHBV film, due to their three-dimensional structures. These fibroblasts interacted and integrated well with the surrounding fibers. The cells that were barely adhered to the PHBV film had more of a round shape. SEM micrographs showed that the development of cells growth was guided by fibrous architecture. Cells grew in the direction of the fiber orientation, and then formed a three-dimensional and multicellular network, according to the architecture of the nanofibrous structure.

Figure 25. SEM micrographs of NIH 3T3 cells cultured for 4 h on electrospun nanofibrous scaffolds: (a) m-keratin fibers, (b) PHBV film, (c) PHBV fibers and (d) PHBV/m-keratin=7/3 fibers.

Figure 26 shows the proliferation of cells on the nanofibrous mats when cultured in DMEM medium with 10% serum for 20 h. m-Keratin fibers exhibited the highest cell proliferation when compared to the PHBV or PHBV/m-keratin fibers. Cell proliferation on mkeratin fibers was significantly different from that on PHBV fibers and PHBV/m-keratin fibers. However, there is no significance difference between PHBV/m-keratin and PHBV control due to low content of m-keratin. It is reasonably concluded that proliferation of NIH3T3 cells was accelerated by m-keratin proteins.

Fabrication, Performance, and Biomedical Application…

367

* p=0.0003

2.0 1.8

Cell proliferation Absorbance(A490nm-A690nm)

1.6 1.4 1.2 1.0 0.8 0.6 0.4 0.2 0.0

PHBV

PHBV/m-keratin=7/3

m-keratin

Figure 26. The proliferation of NIH 3T3 cells cultured for 20 h. (Data are expressed as means ± SD (n=6) for the specific absorbance, * p < 0.01, values are significantly different from those of the previous group). Cytokeratin 8

α-Smooth muscle actin

PHBV

PHBV-Col

Figure 27. Colocalization of extracellular materials (ECM) produced from PHBV matrices cocultured with ORS and DS cells. The production of ECM was examined using immunohistochemical staining of human cytokeratin 8 and α-SMA.

4.5. Wound Healing Test and Histological Examination Figure 27 shows the biological expression of cells cultured on the PHBV matrices. MAbs of human cytokeratin 8 and α-SMA were colocalized to identify extracellular materials

368

Inn-Kyu Kang, Zhi-Cai Xing, Jiang Yuan et al.

(ECM) production in matrices cocultured for 5 days. Stronger signals of α-SMA fibers were produced in PHBV-Col than in PHBV, while both matrices showed similar expression of cytokeratin 8 on round cytoplasm of outer root sheath (ORS) cells. This result was correlated with faster cell attachment and growth of cells in PHBV-Col over PHBV (Figure 20, 21 and 23). Wound healing effects of cocultured matrices are shown in Figure 28. Until the second day after grafting with matrices, all experimental sets showed similar healing progress. However, on the fourth day, PHBV-Col set produced better wound closure than PHBV and the control set did. day 0

day 2

day 4

day 8

Control

PHBV

PHBV-Col

Figure 28. Wound healing progress of PHBV matrices cocultured with ORS and DS cells or matrices alone.

Figure 29 shows the change in wound areas at different healing times using PHBV, PHBV/keratin, and gauze (negative control). At 4 days, there was reduction in wound area for PHBV and PHBV/keratin mats, whereas no reduction was observed for the control. At 7 days, the wound areas decreased gradually and reached about 13% when PHBV/keratin dressing was used. After 9 days, in case of PHBV/keratin mats, the wound was nearly closed completely. Electrospun PHBV/keratin mats was found to be better than PHBV and gauze in promotion wound healing. It indicated that keratin has an ability to accelerate wound healing.

369

Fabrication, Performance, and Biomedical Application…

110 100 90

Wound size %

80 70 60 50 40 30

Gauze(control) PHBV PHBV+Keratin

20 10 0 -10 0

2

4

6

8

10

Post-treatment time (d)

Figure 29. Wound healing test of gauze control (■), PHBV mats (●), and PHBV/keratin=7/3 (▲).

Figure 30 exhibits the histological results of wound after 9 days of covering with gauze, PHBV mats, and PHBV/keratin mats, respectively. In the 9 days, the wound surface of gause control was covered partially by fibrinous tissue debris. For PHBV group, most of the surface tissue debris disappeared, and there was prominent proliferation of fibroblasts. Some inflammatory cells also remained. For PHBV/keratin group, there were no apparent inflammatory cells. Connective tissue was densely formed and wounds were completely covered with new epithelium. It indicated that the introduction of keratin could promote wound healing.

Figure 30. Histology of H-E stained wound after 9-day treatment with gauze (a), PHBV mats (b), and PHBV/keratin mats=7/3 (c). (Magnification=100).

370

Inn-Kyu Kang, Zhi-Cai Xing, Jiang Yuan et al.

5. DISCUSSION Unlike a conventional fiber fabrication process, electrospinning provides a straightforward way to fabricate fibrous scaffolds with fiber diameters in the tens of nanometers [11]. In the present study, we have studied cell behavior on nanofibrous scaffolds which was brought about by electrospinning using biodegradable PHBV and collagen Type Ι. Recently, a polymer blend such as poly(ε-caprolactone)/gelatin [26] and chitosan/gelatin [40] have been successfully electrospun. Wagner et al. [35] have fabricated elecrospun polyetherurethaneurea (PEUU)/collagen scaffolds by combining PEUU with Type I collagen at various ratios. In their results, the SEM images revealed continuous fiber morphology at all of the examined ratios. The diameter of co-electrospun fibers was in the range between 100 and 900 nm. The results which were obtained from the cell experiments showed that cells adhered well to the fibers with diameter smaller than the size of the cells [41]. These findings has led to a necessity of mimicking nanoscale structures that are found in natural ECM when creating a polymeric synthetic ECM scaffold for tissue engineering [42]. In order to illustrate the presence of the protein (collagen, gelatin and keratin) in the coelectrospun scaffolds, a PHBV-protein scaffold was subjected to ATR-FTIR measurements. The resulting spectrum yielded peaks that were characteristic of protein, as reported previously by other researchers [43, 44]. Figure 5, 8 and 11 illustrated common bands of protein appeared at approximately 1650 cm-1 (amide I) and 1535 cm-1 (amide II), corresponding to the stretching vibration of C=O bond, and coupling of bending of N–H bond and stretching of C–N bond, respectively. The amide I band, at 1650 cm-1, was attributable to both a random coil and α-helix conformation. When blended and electrospun with PHBV, the peak of amide I did not shift, which indicated that the conformation of protein is kept. The enzymatic degradation of PHBV films by PHB depolymerase has been reported previously [37]. Park et al. [45] have carried out biodegradation tests of PHBV in the form of fibrous structures or film, by using a simulated municipal solid waste aerobic composting method. They reported that the degradation of the PHBV non-woven structures was faster than the PHBV film. In this study, the surfaces of the PHBV and PHBV-protein fibers were severely eroded by the PHB depolymerase treatment. On the other hand, the PHBV-protein fibers were partially biodegraded by the treatment of collagenase or trypsin solution and their cylindrical morphology was preserved. Recently, a great deal of research has focused on the influence of scaffold microarchitecture on cell behavior [11, 46, 47]. Shin et al. [48] assessed the interaction of human fibroblasts with electrospun nanofibrous scaffolds and reported that the nanofiber structures provided an environment for rapid proliferation. In our research, the PHBV-protein nanofibrous scaffold accelerated the adhesion and growth of NIH 3T3 cells as compared to the PHBV nanofibrous scaffold. Pores in a tissue-engineered scaffold make up the space in which cells reside. Pore properties, such as porosity, dimension, and volume, are parameters directly related to the success of a scaffold. High porosity provides more structural space for cell accommodation and makes the exchange of nutrient and metabolic waste between a scaffold and environment more efficient. These characteristics are fundamental criteria for a successful tissueengineered scaffold [49, 50].

Fabrication, Performance, and Biomedical Application…

371

Nanofibrous scaffolds are very porous, but the pores formed are much smaller than the normal cell size of a few to tens of micrometers, which would inhibit cell migration. As such, the phenomenon of cell ingrowth into a nanofiber structure has been doubted by many researchers and which leading to some reasonable explanations [4, 51]. Pores in an electrospun structure are formed by differently oriented fibers lying loosely upon each other. When cells perform amoeboid movement to migrate through the pores, they can push the surrounding fibers aside to expand the hole as the small fibers offer little resistance to cell movement. This dynamic architecture provides cells with an opportunity to optimally adjust the pore diameter and grow into the scaffold even though some pores are relatively small. Therefore, some pores with smaller diameters in this structure may not hinder cell migration. Such a hypothesis of dynamic cell–scaffold interaction needs to be further investigated.

6. CONCLUSION Both PHBV and protein (collagen, gelatin and keratin) were dissolved in hexafluoro-2propanol and the polymer blend solution was electrospun to produce composite nanofibrous scaffolds. The fiber diameter could be controlled within a range of 300 ~ 800 nm. The presence of protein in the nanofibrous scaffold was confirmed by using ESCA and ATRFTIR. From the in vitro experiments, it was determined that the NIH 3T3 cells showed significant adherence and proliferation on the PHBV/protein nanofibrous scaffold, when compared with the PHBV control. It is concluded that the PHBV/protein composite nanofiber mats would be good candidates for biomedical applications, such as wound dressing and scaffolds for tissue engineering.

ACKNOWLEDGEMENT This work was supported by the Korea Ministry of Education, Science and Technology (2009-0073282).

REFERENCES [1] [2] [3]

[4]

Thomson, R. C., Yaszemski M. J. and Mikes, A. G. (1997). Polymer scaffold processing: Principles of tissue engineering. TX, USA: R. G. Landes Co. Ma, P. X. and Zhang R. (1999). Synthetic nano-scale fibrous extracellular matrix. J Biomed Mater Res, 46, 60-72. Yang, F., Murugan, R., Ramakrishina, S., Wang, X., Ma, Y. W. and Wang, S. (2004). Fabrication of nano-structured porous PLLA scaffold intended for nerve tissue engineering. Biomaterials, 25, 1891-1990. Li, W. J., Laurencin, C., Caterson, E., Tuan R. and Ko, F. (2002). Electrospun nanofibrous structure: A novel scaffold for tissue engineering. J Biomed Mater Res, 60, 613-621.

372 [5] [6]

[7]

[8] [9]

[10]

[11]

[12]

[13]

[14]

[15] [16] [17] [18]

[19] [20]

Inn-Kyu Kang, Zhi-Cai Xing, Jiang Yuan et al. Hartgerink, J. D., Beniash E. and Stupp, S. I. (2001). Self-assembly and mineralization of peptide-amphiphile nanofibers. Science, 294, 1684-1688. Yang, F., Xu, C. Y., Kotaki, M., Wang S. and Ramakrishna S. (2004). Characterization of neural stem cells on electrospun poly (L-lactic acid) nanofibrous scaffold. J Biomater Sci Polmer Edn, 15, 1483-1497. Kim, K.S., Yu, M., Zong, X., Chiu, J., Fang, D., Seo, Y. S., Hsiao, B. S., Chu, B. and Hadjiargyrou M. (2003). Control of degradation rate and hydrophilicity in electrospun non-woven poly ( , -lactide) nanofiber scaffolds for biomedical applications. Biomaterials, 24, 4977-4985. Zeng, J., Chen, X., Xu, X., Liang, Q., Bian, X., Yang, L. and Jing, X. (2003). Ultrafine fibers electrospun from biodegradable polymers. J Appl Polym Sci, 89, 1085-1092. Boland, E., Wnek, G., Simpson, D., Pawlowski K. and Bonlin, G. J. (2001). Tailoring tissue engineering scaffolds using electrostatic processing techniques: a study of poly (glycolic acid) electrospinning. Macromol Sci Part A Pure and Applied Chem, 38, 1231-1243. Katti, D., Robinson, K., Ko, F. and Laurencin, C. J. (2004). Bioresorbable nanofiberbased systems for wound healing and drug delivery: Optimization of fabrication parameters. Biomed Mater Res Part B: Appli biomater, 70, 286-296. Yoshimoto, H., Shin, Y. M., Teral, H. and Vacanti, J. P. (2003). A biodegradable nanofiber scaffold by electrospinning and its potential for bone tissue engineering. Biomaterials, 24, 2077-2082. Williams, S. F., Martin, D. P., Horowitz, D. M. and Peoples, O. P. (1999). PHA applications: addressing the price performance issue I. Tissue engineering. Int J Biol Macromol, 25, 111-121. Sukigara, S., Gandhi, M., Ayutsede, J., Micklus, M. and Ko, F. (2003). Regeneration of Bombyx mori silk by electrospinning - part 1: processing parameters and geometric properties. Polymer, 44, 5721-5727. Ohgo, K., Zhao, C., Kobayashi, M. and Asakura, T. (2003). Preparation of non-woven nanofibers of Bombyx mori silk, Samia cynthia ricini silk and recombinant hybrid silk with electrospinning method. Polymer, 44, 841-846. Wnek, G., Carr, M., Simpson, D. and Bonlin, G. (2003). Electrospinning of Nanofiber Fibrinogen Structures. Nano Lett, 3, 213-216. Matthews, J., Wnek, G., Simpson, D. and Bowlin, G. (2002). Electrospinning of collagen nanofibers. Biomacromolecules, 3, 232-238. Hung, Z. M., Zhang, Y. Z., Ramakrishna, S. and Lim, C. T. (2004). Electrospinning and mechanical characterization of gelatin nanofibers. Polymer, 45, 5361-5368. Chang, S. K., Baek, D. H., Gang, K. D., Lee, K. H., Um, I. C. and Parl, Y. H. (2005). Characterization of gelatin nanofiber prepared from gelatin-formic acid solution. Polymer, 46, 5094-5102. MacGregor, E. A. and Greenwood, C. T. (1980). Polymers in Nature, John Wiley and Sons press: New York. Jones, C.B. and Mecham, D.K. (1943). The dispersion of keratins. II. Studies on the digestion of keratins by reduction in neutral solutions of protein denaturants. Arch Biochem., 3, 193-202.

Fabrication, Performance, and Biomedical Application…

373

[21] Kuzuhara, A. and Hori, T. (2003). Reduction mechanism of tioglycolic acid on keratin fibers using microspectrophotometry and FT-Raman spectroscopy. Polymer, 44, 79637970. [22] Yamauchi, K., Yamauchi, A., Kusunoki, T., Kohda, A. and Konishi, Y. (1996). Preparation of stable aqueous solution of keratins, and physiochemical and biodegradational properties of films. J. Biomed. Mater. Res., 31, 439-444. [23] Tachibana, A., Kaneko, S., Tanabe, T. and Yamauchi, K. (2005). Rapid fabrication of keratin-hydroxyapatite hybrid sponges toward osteoblast cultivation and differentiation. Biomaterials, 26, 297-302. [24] Schrooyen, Peter M. M., Dijkstra, P. J., Oberthuer, R. C., Bantjes, A. and Feijen, J. (2000). Partially Carboxymethylated Feather Keratins. 1. Properties in Aqueous Systems. J. Agric Food Chem., 48, 4326-4334. [25] Schrooyen, Peter M. M., Dijkstra, P. J., Oberthuer, R. C., Bantjes, A. and Feijen, J. (2001). Partially Carboxymethylated Feather Keratins. 2. Thermal and Mechanical Properties of Films. J. Agric Food Chem., 49, 221-230. [26] Zhang, Y. Z., Ouyang, H. W., Lim, C. T., Ramakrishna S. and Huang, Z. M. (2005). Electrospinning of gelatin and gelatin /PCL composite fibrous scaffolds. J Biomed Mater Res Part B: Appl Biomater, 72, 156-165. [27] Kwon, O. H., Jeon, H. A., Kang, I. –K., Kim, Y. J. and Ito, Y. (2009). Bottom-Up Nanofabrication; Supramolecules, Self-Assemblies, and Organized Films, Volume 6, Stevenson Ranch, CA, K. Ariga and H.S. Nawla (Ed.), American Scientific Publisher. PP. 1-19. [28] Formhals, A. (1934). Process and apparatus for preparing artificial threads. US patent, 975504. [29] Liang, D., Hsiao, B.S. and Chu, B. (2007). Functional electrospun nanofibrous scaffolds for biomedical applications. Adv. Drug Delivery Rev., 59, 1392-1412. [30] Ellman, G. L. (1958). A colorimetric method for determining low concentrations of mercaptans. Arch. Biochem. Biophys., 74, 443-450. [31] Harrap, B. S. and Woods, E. F. (1964). Soluble derivatives of feather keratin: 1. Isolation, fractionation, and amino acid composition. Biochem. J., 92, 8-18. [32] Harrap, B. S. and Woods, E. F. (1964). Soluble derivatives of feather keratin: 2. Molecular weight and conformation. Biochem. J., 92, 19-26. [33] Maghni, K., Nicolescu, O. M. and Martin, J. G. (1999). Suitability of cell metabolic colorimetric assays for assessment of CD+ T cell proliferation: comparison to 5-bromo2-deoxyuridine (BrdU) ELISA, J Immunoll Methods, 223, 185-194. [34] Cui, Y. L., Qi, A. D., Liu, W. G., Wang, X. H., Wang, H. Ma, D. M. and Yao, K. D. (2003). Biomimetic surface modification of poly (L-lactic acid) with chitosan and its effects on articular chondrocytes in vitro. Biomaterials, 24, 3859-3868. [35] Stankus, J.J., Guan, J. J. and Wagner, W. R. (2004). Fabrication of biodegradable elastomeric scaffolds with sub-micron morphologies. J Biomed Mater Res Part A, 70, 603-614. [36] Gassner, F. and Owen, A. J. (1994). Physical-Properties of Poly(Beta-Hydroxybutyrate) Poly(Epsilon-Caprolactone) Blends. Polymer, 35, 2233-2236. [37] Kang, I. K., Choi, S. H., Shin, D. S. and Yoon, S. C. (2001). Modification of polyhydroxyalkanoate films and their interaction with human fibroblasts. Int. J. Biol. Macromol., 28, 205-212.

374

Inn-Kyu Kang, Zhi-Cai Xing, Jiang Yuan et al.

[38] Nicol, A, Gowda, D. C. and Urry, D. W. (1992). Cell adhesion and growth on synthetic elastometric matrices containing Arg-Gly-Asp-Ser. J Biomed Mater Res., 26, 393-413. [39] Singer, II., Kawka, D. W., Scott, S., Mumford, R. A. and Lark, M. W. (1987). The fibronectin cell attachment sequence Arg-Gly-Asp-Ser promotes focal contact formation during early fibroblast attachment and spreading. J Cell Biology., 104, 573584. [40] Xia. W. Y., Liu, W., Cui, L., Liu, Y. C., Zhong, W., Liu, D. L., Wu, J. J., Chua, K. and Cao, Y.L. (2004). Tissue engineering of cartilage with the use of chitosan-gelatin complex scaffolds. J Biomed Mater Res Part B Appl Biomater, 71, 373-380. [41] Laurencin, C. T., Ambrosio, A. M., Borden, M. D. and Cooper, JA. Jr. (1999). Tissue engineering: orthopedic applications. Annu Rev Biomed Eng, 1, 19-46. [42] Xu, C. Y., Inai, R., Kotaki, M. and Ramakrishna, S. (2004). Aligned biodegradable nanofibrous structure: a potential scaffold for blood vessel engineering. Biomaterials, 25, 877-886. [43] Sachlos, E., Reis, N., Ainsley, C., Derby, B. and Czernuszka, J. T. (2003). Novel collagen scaffolds with predefined internal morphology made by solid freeform fabrication. Biomaterials, 24, 1487-1497. [44] Camacho, N. P., West, P., Torzilli, P. A. and Mendelsohn, R. (2001). FTIR spectroscopic imaging of collagen and proteoglycan in bovine cartilage. Biopolymers, 62, 1-8. [45] Choi, J. S., Lee, S. W., Jeong, L., Bae, S. H., Min, B. C, Youk, J. H. and Park, W. H. (2004). Effect of organosoluble salts on the nanofibrous structure of electrospun poly (3-hydroxybutyrate-co-3-hydroxyvalerate). Int J Biol Macromol, 34, 249-256. [46] Li, M. Y., Mondrinos, M. J., Gandhi, M. R., Ko, F. K., Weiss, A. S. and Lelkes, P. I. (2005). Electrospun protein fibers as matrices for tissue engineering. Biomaterials, 26, 5999-6008. [47] Riboldi, S. A., Sampaolesi, M., Neuenschwander, P., Cossu, G. and Mantero, S. (2005). Electrospun degradable polyesterurethane membranes: potential scaffolds for skeletal muscle tissue engineering. Biomaterials, 26, 4606-4015. [48] Lee, C. H., Shin, H. J, Cho, I. H., Kang, Y. M., Kim, I. A., Park, K. D. and Shin, J. W. (2005). Nanofiber alignment and direction of mechanical strain affect the ECM production of human ACL fibroblast. Biomaterials, 26, 1261-1270. [49] Ma, P. X. and Choi, J. W. (2001). Biodegradable polymer scaffolds with welldefined interconnected spherical pore network. Tissue Eng, 7, 23–33. [50] Bhattarai, S. R., Bhattarai, N., Yi, H. K, Hwang, P. H., Cha, D. I. and Kim, H.Y. (2004). Novel biodegradable electrospun membrane: scaffold for tissue engineering. Biomaterials, 25, 2595-2602. [51] Zhang, Y. Z., Venugopal, J., Huang, Z. M., Lim, C. T. and Ramakrishna, S. (2005). Characterization of the surface biocompatibility of the electrospun PCL-collagen nanofibers using fibroblasts. Biomacromolecules, 6, 2583-2589.

In: Nanofibers: Fabrication, Performance, and Applications ISBN 978-1-60741-947-1 Editors: W. N. Chang © 2009 Nova Science Publishers, Inc.

Chapter 12

FABRICATION AND CHARACTERIZATION OF POLYPROPYLENE FIBER REINFORCED BY CARBON NANOFIBER Yuanxin Zhou and Shaik Jeelani Tuskegee University’s Center for Advanced Material (T-CAM), Tuskegee, AL, USA

ABSTRACT In this study, vapor grown carbon nanofiber (CNF) has been used to improve thermal and mechanical properties of polypropylene. The CNFs were first dispersed over the polypropylene particles using sonication coating method, and then extruded into filaments with a single screw extruder. The thermal properties of neat and nanophased polypropylene were characterized by TGA and DSC. TGA thermograms showed that the nanoparticle-infused systems are more thermally stable, and DSC results indicated that CNFs have no effect on melting temperature. Tensile tests were performed on the single filament at a strain rate range from 0.02/min to 2/min. Results indicate that both neat and nanophased polypropylene were strain rate-strengthening material. The tensile modulus and yield strength both increased with increasing strain rate. Experiment results also show that infusing polypropylene with nanofibers increases tensile modulus and yield strength, but decreases ductility. Finally, based on the tensile test results, a nonlinear constitutive equation was developed to describe strain rate-sensitive behavior of neat and nanophased polypropylene.

1. INTRODUCTION Polypropylene is a semi-crystalline engineering thermoplastic and is known for its balance of strength, modulus and chemical resistance. It have many potential applications in automobiles, appliances and other commercial products in which creep resistance, stiffness and some toughness are demanded in addition to weight and cost savings. The incorporation of inorganic particulate fillers has been proved to be an effective way of improving the

376

Yuanxin Zhou and Shaik Jeelani

mechanical properties, and in particular the toughness, of polypropylene. For example, Stamhuis [1] has shown that talc filler can significantly increase the impact resistance of polypropylene if it is physically blended with either an SBS or an EPDM elastomer. R. S. Hadal and R. D. K. Misra [2] have reported that modulus of polypropylene has been improved by talc and wollastonite. Radhakrishnan and Saujanya [3] have reported that the needle-shaped CaSO4 filler can improve the properties of polypropylene with its high aspect ratio. However, the typical filler content needed for significant enhancement of these properties can be as high as 10-20% by volume. At such high particle volume fractions, the processing of the material often becomes difficult, and since the inorganic filler has a higher density than the base polymer, the density of the filled polymer is also increased. The advantages of polymers, i.e. their ease of processing and lightweight, get therefore lost, and which limits various applications of polypropylene composite. To overcome this drawback, a composite with improved properties and lower particle concentration is highly desired. With regard to this, the newly developed nanocomposites would be competitive candidates. Nanoparticle filled polymers are attracting considerable attention since they can produce property enhancement that are sometimes even higher than the conventional filled polymers at volume fractions in the range of 1 to 5%. M. Z. Rong et al have reported tensile performance improvement of low nano-particle filled polypropylene composite[4]. C. L Wu et al have investigated tensile behavior of nano-phased polypropylene at different strain rate. The strength of nano-phased composite are increased with either increasing strain rate or increasing volume fraction of nano-particle[5]. Several different types of these polymer-based nanocomposites are now becoming commercially available and finding applications ranging from barrier materials to automotive body components [6]. Vapor-grown carbon nanofibers (CNFs), due to their high tensile strength, modulus, and relatively low cost, are drawing significant attention for their potential applications in nanoscale polymer reinforcement. They are synthesized in the presence of a catalyst from pyrolysis of hydrocarbons or carbon monoxide in the gaseous state [7-8]. Vapor-grown CNFs are different from other types of nanofibers, including polyacylonitrile and mesophase pitchbased carbon fiber, in their method of production, physical properties, and structure. Thermoplastics, such as polypropylene [9-10], polycarbonate [11], nylon [12]; thermoset, such as epoxy [13]; and thermoplastic elastomers, such as butadiene-styrene diblock copolymer [14], have been reinforced with carbon nanofibers. In practice, however, many reports indicate that CNF-reinforced composites are weaker or only slightly stronger than neat polymers [15]. The most common explanation is that CNFs were randomly distributed in the composite. As a result, alignment control has become a fundamental challenge in developing high-performance CNF-reinforced nanocomposites. Research on alignment of various carbonaceous materials, such as carbon black (CB), carbon nanofiber (CNF), and carbon nanotubes (CNTs), has received considerable attention. CNTs have a strong structural anisotropy, and the alignment of CNT slang in the direction of an applied magnetic field was attempted by using this strong anisotropic diamagnetic property [16]. Another way to align carbonaceous materials is by using an electric field. Yamamoto et al. have reported electrophoresis of CNTs in IPA [17]. The CNTs in IPA moved toward the cathode under a dc electric field. Chen et al. and Kumar et al. have achieved aligned structures of CNTs between electrodes, with 25 um and 3 um gaps, respectively, under an electric field by completely removing the solvents [18-19]. Compared to the above methods, extrusion and spinning are easy ways to control the nanofiber direction in liquid polymers for

Fabrication and Characterization of Polypropylene Fiber…

377

structural nanocomposites. During the process, CNTs or CNFs will turn to the flow direction. Mahfuz et al. [20] reported 100% enhancement in both strength and stiffness in extruded CNT/nylon 6, as compared to neat nylon 6. Kumar et al [21] fabricate CNF/PP fiber by using conventional melt spinning equipment, and modulus and compressive strength of polypropylene increased by 50% and 100%. Like many other polymer materials, the mechanical responses (deformation, strength, and failure) of polypropylene and its composites, depend on the rate of deformation [22]. It is necessary to know the mechanical behavior of a nanocomposite at different strain rates if a component made of the nanocomposite undergoes different loading speeds. However, strain rate sensitivity of polypropylene has not been well studied. In this study, aligned CNFreinforced polypropylene (PP) filaments were fabricated by extrusion processing. Single filament tests were performed at different strain rates to evaluate mechanical performance. Thermogravimetric Analysis (TGA) and Differential Scanning Calorimetry (DSC) were used to evaluate thermal performances. Microscopic approaches were used to investigate the fracture behavior of material. Based on tensile stress-strain curves, a nonlinear constitutive equation was developed to describe the mechanical behavior of neat and nanophased polypropylene at different strain rates.

2. MATERIALS PROCESSING The PR-24 carbon nanofibers were obtained from Applied Science, Inc. in Cedarville, OH. The fiber diameters ranged from 60 to 200 nm and the fiber lengths ranged from 30 to 100 μm . Figure 1A and 1B show the pictures of as-received carbon nanofibers at different magnifications. High specific surface area and cotton-like entanglement cause the formation of agglomerates. Agglomerates of CNFs, called nanoropes [23], are difficult to separate and infiltrate with matrix. Polypropylene powders were supplied by ChemPoint. The diameters of the powder ranged from 5um to 20um (as shown in Figure 1C and 1D). Although there are various techniques such as melt mixing, solution mixing, shear mixing, magnetic mixing and mechanical stirring to infuse nanoparticles or nanoscale fillers into a liquid polymer, acoustic cavitation is considered as one of the most efficient ways to disperse nanoparticles into the virgin materials. In this case, the application of alternating acoustic pressure above the cavitation threshold creates numerous cavities in the liquid. Some of these cavities oscillate at a frequency of the applied field (usually 20 kHz) whereas the gas content inside these cavities remains constant. However, some other cavities grow intensely under tensile stresses whereas yet another portion of these cavities that are not completely filled with gas start to collapse under the compression stresses of the sound wave. In the latter case, the collapsing cavity generates tiny particles of debris and the energy of the collapsed one is transformed into pressure pulses. It is noteworthy that the formation of the debris further facilitates the development of cavitation. It is assumed that acoustic cavitation in liquids develops according to a chain reaction. Therefore, individual cavities on real nuclei develop so rapidly that within a few microseconds an active cavitation region is created close to the source of the ultrasound probe. The development of cavitation processes in the ultrasonically processed melt creates favorable conditions for the intensification of various physicochemical processes. Acoustic cavitation accelerates heat and mass transfer processes

378

Yuanxin Zhou and Shaik Jeelani

such as diffusion, wetting, dissolution, dispersion, and emulsification. Recently it has been reported that there is a clear acceleration of polymer reaction under ultrasound in both catalyzed and uncatalyzed reactions.

A

B

C

D

Figure 1. SEM pictures of agglomerated CNF (A and B), PP powders (C and D).

In the present investigation, sonication coating method was employed to disperse individual CNF on surface of polypropylene particles(as shown in Figure 2). 0.5g Carbon nanofibers and 100g acetone were carefully weighed, and mixed together in a suitable beaker. The mixing was carried out through a high intensity ultrasonic irradiation (Ti-horn, 20 kHz Sonics Vibra Cell, Sonics Mandmaterials, Inc, USA) for half an hour with pulse mode (50sec. on/ 25sec. off). To avoid a temperature rise during the sonication process, external cooling was employed by submerging the beaker containing the mixture in an ice-bath. Once the irradiation was completed, 99g polypropylene powder were added to the mixture then mixed through the same high intensity ultrasonic irradiation for 2 hours. After sonication, the mixture was put in a vacuum oven at 100oC for 24 hours to remove any trace of acetone.

Fabrication and Characterization of Polypropylene Fiber…

379

Figure 2. Sonication coating method.

SEM pictures were used to evaluate dispersion of CNF in polypropylene. After sonication coating process, agglomerated CNF particles did not show in the mixture, and individual CNFs was uniformly coated on the surface of polypropylene particles (Figure 3). Compared with original polypropylene particles, the surface of polypropylene after sonication coating is not smooth anymore. Some CNFs have merged into polypropylene surface and some CNFs randomly distributed in mixture.

A

B

Figure 3. SEM pictures of PP+CNF mixture.

The polypropylene and CNF mixture was then extruded into filaments using a Wayne Yellow Label Table Top Extruder. The extruder has a 19 mm diameter screw encased inside the barrel that is driven by a 2 HP motor via a toothed timing belt for smooth speed reversal. Five thermostatically controlled heating zones were used to melt the mixture prior to extrusion; three were located in the barrel zone and two were located in the die zone. The

380

Yuanxin Zhou and Shaik Jeelani

heaters inside the barrel zone were placed at the hopper side, center, and die zone side of the barrel and set at a temperature of 2430 C, 2600 C, and 2770 C, respectively. The purpose of these three heaters was to maintain a gradually increasing temperature in the molten mass. First, the nanophased, dry-mixed powder was poured through the feed hopper into the barrel zone of the extruder. As soon as it reached the barrel zone, the PP started melting due to the high barrel temperature. The outer surface of the screw is designed to closely fit the barrel’s inner surface; as a result, the melted PP was trapped inside the screw segment. As the extruder screw rotated, the PP mixed with the carbon nanofiber and reached the screw end, which is next to the die zone. The liquid polypropylene containing CNFs then entered the die zone, which consists of a circular plate, 10 cm-long steel tubing with a 4 mm inner diameter, and the die itself. The two heaters at this zone were set at a temperature of 2770 C to keep the flowing mass at a constant temperature. One of the heaters was placed after the circular plate and the other one was immediately before the die. The circular plate was 15 mm in diameter and contained about 20 orifices, which were each 2 mm in diameter. As the bulk materials passed through the plate, they were disintegrated into several branches and then recombined, allowing further mixing of the CNFs and PP. The CNFs mixed with liquid PP passed through the 10 cm-long steel tube and arrived close to the die before passing through a carefully designed die (Figure 4). The composite filaments were continuously extruded at a screw speed of 8 rpm and a feed rate of approximately 60 g/h. Filaments were allowed to travel about 3 to 4 m in air (temperature 20O C) from the die outlet to a water trough, godet machine-1, heater, godet machine-2, set of tension rollers, and winder guide rollers, and then they were wound on a spool using a Wayne Desktop Filament Winder at a winding speed of 70 rpm (Figure 5).

Figure 4. (a) Circular plate with multiple holes, (b) Side view of the extruder die.

Fabrication and Characterization of Polypropylene Fiber…

381

3. EXPERIMENTAL RESULTS AND DISCUSSION Thermal properties The thermal stability of the neat and nanophased polypropylene were evaluated by Thermogravimetry Analysis (TGA). Tests were conducted with a TA Instruments TGA2950 at a heat rate of 10° C/min from ambient to 600° C under a nitrogen gas atmosphere. The realtime characteristic curves were generated by a data acquisition system, Universal Analysis 2000-TA Instruments, Inc. Figure 6 shows the TGA results of neat and nanophased polypropylene. In this study, 10% of the total weight loss was considered to be the initial decomposition of materials. Figure 6 also shows that neat polypropylene began to decompose at 394° C and the CNF/PP nanocomposite began to decompose at 430° C. The derivative peaks of weight v. temperature curves show the decomposition temperature. The decomposition temperatures were 444° C and 474° C for neat polypropylene and nanophased polypropylene, respectively (Figure 6). This clearly demonstrates that the CNF/PP system is thermally more stable than the corresponding neat polypropylene system. The Differential Scanning Calorimetric (DSC) tests were carried out using a modulated DSC apparatus (TA, Q50) from TA Instruments Inc. The samples were cut into small pieces with weights of about 5 to 20 mg, sealed in aluminum crucibles, and placed inside the apparatus. Experiments were carried out under a nitrogen gas atmosphere at a heating rate of 100 C/min. Figure 7 shows the DSC curves of neat and nanophased PP. Melt temperatures of neat and nanophased polypropylene were 164.7° C and 167.7° C, respectively. The enhanced thermal stability and melting temperature of the nanoinfused filaments may be attributed to the adhesion between the nanofibers and the matrix, which stabilizes the composite against thermal decomposition and melting.

Figure 5. Schematic diagram for the manufacturing of CNF reinforced polypropylene.

382

Yuanxin Zhou and Shaik Jeelani 200

3 o

444.8 C o

463.7 C 150

Weight (%)

1

100

0

o

o

430 C

394 C

-1 50

Deriv. of Weight (%/ oC)

2

Neat PP -2

0.5% CNF/PP

0

-3 200

300

400

Temperature ( o C)

500

600

Figure 6. TGA results of neat and nanophased polypropylene filaments. 0

o

Heat Flow (mW)

164.5 C -4

-8

Neat PP CNF/PP

o

167.7 C -12 140

150

160

Temperature ( o C)

170

180

Figure 7. DSC results of neat and nanophased polypropylene filaments.

Mechanical Properties According to the ASTM Standard D3379-75, tensile tests for single filaments were conducted on a Zwick Roell Tensile tester equipped with a 20 N Load Cell. The tests were performed in displacement control mode and strain rates ranged from 0.02 min-1 to 2 min-1. At

Fabrication and Characterization of Polypropylene Fiber…

383

least ten specimens were tested in each category. The stress-strain curves were generated by Zwick testXpert software, a data acquisition system. 400

Polypropylene (PP)

Stress (MPa)

300

200

Strain Rate (1/min) 100

0.02 0.2 2 0 0

50

100

150

200

250

40

50

Strain (%) A 500

0.5wt% CNF Filled Polypropylene

Stress (MPa)

400

300

200

Strain Rate (1/min) 0.02 100

0.2 2

0 0

10

20

30

Strain (%) B Figure 8. Stress strain curves of neat(A) and nanophased (B) polypropylene.

384

Yuanxin Zhou and Shaik Jeelani

Tensile stress-strain curves of the neat polypropylene and nanophased polypropylene at different strain rates are shown in Figure 8. The stress-strain relationships of polypropylene and its nanocomposites were nonlinear even at strains lower than the yield strain. After yielding, the stress increased steadily with strain until the materials fractured. See Table 1 for the tensile properties of each material at different test conditions. Figure 8 shows the effect of strain rate on stress-strain curves. With increased strain rates, the modulus, yield strength, and tensile strength all increased for each material. However, the failure strain decreased, indicating that increased strain rates reduce the ductility of materials. Table 1. Mechanical properties of polypropylene and its nanocomposite

Material

Neat PP 0.5 wt% CNF/PP

Strain Rate (1/min)

E (GPa)

0.02 0.2 2 0.02 0.2 2

1.83 3.01 4.39 3.60 4.37 4.80

Hardening Modulus (MPa) 88.4 89.0 88.6 377 370 387

σY (MPa) 78.6 121 188 266 304 324

Failure Strain (%) 229 209 164 51.6 38.2 26.3

500

Nanophased PP

Stress (MPa)

400

300

Neat PP 200

100

Strain Rate is 0.02/min 0 0

50

100

150

200

Strain (%) Figure 9. Tensile stress-strain curves of neat and nanophased polypropylene.

250

UTS (MPa) 256 290 315 426 432 440

385

Fabrication and Characterization of Polypropylene Fiber…

Figure 9 shows stress strain curves of neat and nanophased polypropylene at the same strain rate. From Table 1 and Figure 9, it can be concluded that filling CNFs into polypropylene can significantly increase polypropylene’s mechanical performances. At the strain rate 0.02 min-1, 154% enhancement in tensile modulus and 69.5% enhancement in ultimate tensile strength were observed in the CNF/PP system, as compared to neat polypropylene. However, filling the polypropylene with nanofiber decreased the ductility of the material by 77%.

STRAIN RATE SENSITIVITY OF YIELD STRENGTH As shown in Figure 8 and Table 1, polypropylene and polypropylene matrix nanocomposites are strain rate-sensitive materials. Figures 10-12 show the variation of tensile strength, yield strength, and failure strain with ln ε& for two kinds of materials. The relationships between

σ b and ln ε& , σ Y and ln ε& , and ε b and ln ε& are represented by

single straight lines, the slopes of which give information about the strain rate sensitivities. The following relationships were found to fit the relationship between yield strength, tensile strength, and strain rate:

⎛

σ b = σ b 0 ⎜⎜1 + λ1 ln ⎝

⎛

ε& ⎞ ⎟ ε&0 ⎟⎠

σ Y = σ Y 0 ⎜⎜1 + λ 2 ln ⎝

⎛

ε b = ε b 0 ⎜⎜1 + λ3 ln ⎝

ε& ⎞ ⎟ ε&0 ⎟⎠

ε& ⎞ ⎟ ε&0 ⎟⎠

(1)

(2)

(3)

σ b 0 , σ Y 0 , ε b 0 , and ε&0 are reference tensile strength, reference yield strength, reference failure strain, and reference strain rate, respectively. Three other parameters, λ 1,2,3,

where

appearing in Equation (1)-(3), are defined as strain rate coefficients. Mathematically, they are defined as:

λ1, 2,3 =

∂ (σ b ,σ y , ε b ) ∂ ln ε&

(4)

The values of the strain rate coefficients are calculated from the experiment results using the least square method (Table 2). According to the Eyring equation, the activation volume of the material can be obtained from the strain rate-strengthening coefficient as follows:

386

Yuanxin Zhou and Shaik Jeelani

V =

RT

(5)

λ2

where V is activation volume, R is gas constant, and T is the temperature. The activation volume of the neat polypropylene and its nanocomposite were calculated using the yield strength data in Table 1. This table shows that the activation volumes were decreased about 31%. The higher yield strength and lower activation volumes of the nanocomposite are attributed to the restricted segmental motions in the CNF/PP interfaces. Table 2. Constitutive parameters of polypropylene and its nanocomposite Material PP CNF/PP

λ3

λ1

λ2

(MPa) 10.6 9.12

(MPa) 23.8 34.3

V ( nm)

0.14 0.056

3

η (MPa/min) 2

0.2

0.02

0.200 0.138

70 80

550 950

4000 12000

-1

0

400

Yield Strength (MPa)

300

CNF/PP

200

100

Neat PP

0 -4

-3

-2

.

ln ε

Figure 10. Effect of strain rate on yield strength of neat and nanophased PP.

1

387

Fabrication and Characterization of Polypropylene Fiber… 500

Tensile Strength (MPa)

CNF/PP

400

300

Neat PP

200 -4

-3

-2

.

ln ε

-1

0

1

Figure 11. Effect of strain rate on tensile strength of neat and nanophased PP. 300

Failure Strain (%)

Neat PP 200

100

CNF/PP 0 -4

-3

-2

.

ln ε

-1

Figure 12. Effect of strain rate on failure strain of neat and nanophased PP.

0

1

388

Yuanxin Zhou and Shaik Jeelani

Fracture Surface Figures 13a-d show the fracture surfaces of the neat and nanophased polypropylene at a strain rate of 2 min-1. Similar results were observed at other strain rates. For neat polypropylene, axial symmetric fracture surfaces indicate that initial failure occurred at the center of the filament (Figs. 13a and 13b). The surface was relatively smooth with many terraced markings. For nanophased polypropylene, the failure of the filament was initiated at the edge of the sample. The side surface of the specimen shows craze normal to the tensile direction (Figure 13c). Craze widens and joins together to form a crack. The fracture surface of the nanophased polypropylene becomes rougher, containing some large matrix fibrils (Figure 13d).

A

B

C

D

Figure 13. Fracture of neat and nanophased PP (A: neat PP; B: neat PP; C: CNF/PP; D: CNF/PP).

Figures 14a-d show the tensile fractured surfaces of neat polypropylene and the nanocomposite at the high magnification. The micrographs indicate that many unbroken ligaments deform plastically as decohesion of fibrils and microfibrils takes place. These ligaments are fibrils that were separated from the surrounding constrained matrix during deformation. The CNFs were dispersed in the matrix uniformly and aligned in the extrusion direction (Figure 14c), and all CNFs parallel to the tensile direction and polypropylene hold the CNFs well (Figure 14d).

389

Fabrication and Characterization of Polypropylene Fiber…

A

B

C

D

Figure 14. Fracture surface of neat and nanophased polypropylene.(A: neat PP; B: neat PP; C: CNF/PP; D: CNF/PP).

CONSTITUTIVE EQUATION In this study, a standard linear solid model was used to describe the strain rate-sensitive behavior of PP and its nanocomposite (Figure 15). The total strain is assumed to be the sum of elastic strain and inelastic strain, so that the strain rate can be written as

ε& = ε&1 + ε&2

(6)

The elastic strain rate is assumed to be path-independent, such that

ε&1 =

1 dσ E1 dt

(7)

where E1 is elastic modulus of the material, The inelastic strain rate is controlled by dashpot,

ε&2 =

σ2 η

(8)

390

Yuanxin Zhou and Shaik Jeelani Spring 2 E2 Spring 1

σ

σ Dashpot

E1

η Figure 15. Standard linear solid model.

Considering

E1ε 1 = E 2 ε 2 + ηε&2

(9)

Equation (6) can be expressed as

⎛

ε& =

σ& E

+

σ − E 2 ⎜⎜ ε − ⎝

η

σ ⎞

⎟ E1 ⎟⎠

(10)

During the constant strain rate tensile test, stress-strain relationship can be expressed as:

σ=

E1 E 2 E12 ε&η ε+ E1 + E 2 (E1 + E 2 )2

⎡ ⎛ E1 + E 2 ε ⎞⎤ ⎟⎟⎥ ⎢1 − exp⎜⎜ − & η ε ⎝ ⎠⎦ ⎣

(11)

where E1 and E2 are the modulus of the spring in front and the spring in the Kelvin element. The quantity η is the viscosity coefficient with the dimension of stress x time. According to Equation (11), the initial slope of stress-strain curves can be expressed as:

dσ = E1 dε

(12)

At the large deformation, the slope can be expressed as:

EE dσ = 1 2 dε E1 + E 2

(13)

The values of E1 and E2 can be obtained from the tensile modulus and the hardening modulus of material. After that, the quantity η can be simulated from experimental stress-

Fabrication and Characterization of Polypropylene Fiber…

391

strain curves by using the least square method (Table 2). Figure 16 shows the relationship between η and the strain rate. The values of η decreased with increasing strain rate and infusing nanoparticles increased the viscosity of the polypropylene matrix. For neat PP:

η = 130ε& −0.879 (MPa × min ) For PP with 0.5 wt% CNF:

η = 168ε& −1.09 (MPa × min ) Substituted for parameters in Equation (11), the simulated stress-strain plots fit the experiment data very well (Figure 17).

4. CONCLUSIONS Tensile tests of polypropylene and vapor-grown carbon reinforced polypropylene have been carried out at different strain rate. Based on the experiment’s results, we reached the following conclusions: 1. Neat and nanophased polypropylene are strain rate-sensitive materials. The yield strength, modulus, and tensile strength all increase with increasing strain rate. 2. Tensile results show that filling the polypropylene with CNFs increases the modulus, yield strength, hardening modulus, and tensile strength, but decreases failure strain. 15000

(1/min)

10000

η

CNF/PP

5000

PP 0 0.01

0.10

1.00

Strain Rate (1/min) Figure 16. Effect of strain rate on viscosity coefficient η .

10.00

392

Yuanxin Zhou and Shaik Jeelani 500

Material

Strain Rate (1/min) Neat PP

0.02

Neat PP

0.2

Neat PP

2

CNF/PP

0.02

CNF/PP

0.2

CNF/PP

2

Stress (MPa)

400

300

200

100

0 0

50

100

150

200

250

Strain (%) Figure 17. Comparision between simulated results and experimental date.

3. TGA and DSC results show that nanoparticles can increase decomposition temperature and melt temperature. 4. A nonlinear constitutive model has been established to describe the strain ratesensitive behavior of the polypropylene and its nanocomposites. The parameters in this model are the modulus E, hardening modulus E1, and viscosity coefficient η . The simulated results show that values of

η decreased with increasing strain rate,

and infusing nanoparticles can increase the viscosity of the nylon matrix.

ACKNOWLEDGEMENTS The authors would like to gratefully acknowledge the support of the Air Force Minority Leaders Nanocomposites Research and Education Program and National Science Fundation.

REFERENCES [1] [2] [3] [4]

J. E. Stamhuis, Polym. Comp., 5, (1984) 202. R. S. Hadal and R. D. K. Misra, Material Science and Engineering A, 374 (2004) 374 S. Radhakrisnan and C. Saujanya, J. Mater. Sci., 33, (1998) 1069. M. Z. Rong, M. Q. Zhang, Y. X. Zheng, H. M. Zeng and K. Friedrich, Polymer., 42, (2001) 3001.

Fabrication and Characterization of Polypropylene Fiber… [5] [6] [7] [8] [9]

[10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21]

[22] [23]

393

C. L Wu, M. Q. Zhang, M. Z. Rong and K. Friedrich, Composite Science and Technology., 62, (2002) 1327. A. Oya, Y. Kurokawa and H. Yasuda, Journal of Materials Science., 35, (2000) 1405. M Ward, Mechanical Properties of Solid Polymers, 2nd ed. John Wiley and Sons, New York (1983). H. Miyagawa, M. J. Rich and L. T. Drzal, Journal of Polymer Science Part B: Polymer Physis, 42 (2004): pp4384-4390. Lake, M.L. and J.M. Ting. 1999. Vapor-grown carbon fiber composites In: T.D. Burchell, Editors, Carbon Materials for Advanced Technologies, Pergamon Press, Oxford, UK: 139-167. Van Hattum, F.W.J., C.A. Bernardo, J.C. Finegan, G.G. Tibbetts, R.L. Alig, and M.L. Lake. 1999. Polymer Composites 20 (5): 683–688. Carneiro, O.S., J.A. Covas, C.A. Bernardo, G. Caldiera, F.W.J. Van Hattum, and J.M. Ting et al. 1998. Composites Science and Technology 58 3-4: 401–407. R.T. Pogue, J. Ye, D.A. Klosterman, A.S. Glass and R.P. Chartoff. 1998. Composites, Part A. 29A 9-10: 1273–1281. R.D. Patton, C.U. Pittman, Jr., L. Wang and J.R. Hill. Composites, Part A. Applied Science and Manufacturing 30A 9 (1999) 1081–1091. V. Chellappa, Z.W. Chiou and B.Z. Jang. 1994. Electrical behavior of carbon whisker reinforced elastomer matrix composites. Proc 26th SAMPE Tech Conf. 12-18. P. Farhana, Y.X. Zhou, V. Rangari, and S. Jeelani, Materials Science and Engineering A, 405(1-2) (2005), 246-253. B.W. Smith, Z. Benes Z, D.E. Luzzi, J.E. Fischer, D. AS. Walters, M.J. Casavant, et.al. Appl Phys lett 2000; 77:663-5 Yamamoto, S. Akita, U. Nakayama. Jpn J Appl Phys, 1996. 35:917-8 X.Q. Chen, T. Saito, H. Yamasa, K. Appl Phys Lett 2001. 78(23). 3714-6 M.S. Kumar, T.H. Kim, S.H. Lee, S.M. Song, J.W. Yang, K.S. Nahm et al, Chem Phys Lett 2004. 383:235-9 Mahfuz, H., Adnan, A., Rangari, V., Hasan, M., Jeelani, S., Wright, W. J., and Deteresa, S. J., Applied Physics Letters, 88 (2006), 083119. Satish Kumar, Harit Doshi, Mohan Srinivasarao, Jung O. Park and David A. Schiraldi, Fibers from polypropylene/nano carbon fiber composites Polymer, Volume 43, Issue 5, March 2002, Pages 1701-1703. Yuanxin Zhou, P. K. Mallick, Polymer Engineering and Science: Dec. 2002 244912460 Gojny F.H., Wichmann M.H.G., Fiedler B. and Schulte K., Influence of different carbon nanotubes on the mechanical properties of epoxy matrix composites – A comparative study, Composites Science and Technology, Volume 65, Issues 15-16, 2005, Pages 2300-2313

In: Nanofibers: Fabrication, Performance, and Applications ISBN 978-1-60741-947-1 Editors: W. N. Chang © 2009 Nova Science Publishers, Inc.

Chapter 13

THERMAL ANALYSIS OF CARBON NANOTUBES INCORPORATED POLYURETHANES NANOCOMPOSITES Shahrul Azam Abdullah and Lars Frormann1 University of Applied Science Zwickau, Institute for Production Technology, Zwickau, Germany

ABSTRACT Carbon nanotubes (CNTs) have a number of outstanding mechanical and physical properties which make them attractive as reinforcement in polymer matrix. CNTs reinforced polyurethane nanocomposites provide the possibility to tailor the material strength, stiffness and thermal behavior of polyurethane. Multi-walled carbon nanotubes (MWNTs) filled polyurethane composites were prepared by mixing and injection molding and its thermal characteristics were investigated. The analysis of the influences of MWNTs particles and composites mixing methods were done using dynamic mechanical analysis (DMA), differential scanning calorimetry (DSC) and thermogravimetry analysis (TGA). The storage moduli of the composites increased with increasing MWNTs loading which indicate a good matrix/filler adhesion and increased the stiffness of the composites. However, the increase in processing speed has decrease the storage modulus. Addition of MWNTs filler also broadened and lowered the peak of tan δ denotes that the polyurethane composite became more elastic. Thermal stability of the polyurethane was improved with MWNTs loading which is associated to high thermal stability of CNTs. A very high processing speed reduced the composites thermal stability making it easier to degrade. DSC analysis indicated that the inclusion of MWNTs increased the melting temperature and act as restriction sites for the polyurethanes soft segments. The present chapter revealed the potential of MWNTs as agent for better thermal properties of polyurethane nanocomposites and their properties depend strongly on the dispersion and distribution of nanotubes in polyurethane matrix. 1

Corresponding author: Prof. Dr-Ing. Lars Frormann Tel.: +49 375 536 1721, Fax: +49 375 536 1713, E-mail address: [email protected].

396

Shahrul Azam Abdullah and Lars Frormann

Keywords: polyurethane; nanocomposite; carbon nanotubes; extrusion; mechanical properties; thermal properties

1. INTRODUCTION Carbon nanotubes (CNTs) which were discovered by Ijima in 1991 [1] can be thought of as a seamless, hollow cylinder, with carbon atoms in a regular hexagonal arrangement [2,3]. The CNTs dimension were typically few nanometers in diameters and several hundred nanometers in length which result in a very high aspect ratio, as high as 100-1000 [3-7]. Along with their size and high aspect ratio CNTs also posses extraordinary mechanical properties and other distinct properties such as outstanding thermal and electrical properties which make CNTs excellent choice to be used as reinforcing agent in polymer composites [27]. Polymer nanocomposites based on CNTs have attracted a tremendous amount of attention during the past years from the research and industrial communities [8,9,10] The composites were reported to have high strength and stiffness as well as good electrical conductivity at relatively low CNTs loading [9]. CNTs were used as fillers in this study in order to enhance polyurethane properties especially mechanical and thermal conductivity. Thermoplastic polyurethanes are segmented copolymers comprising of a repeating structure of soft and hard segments which demonstrate both thermoplastic and elastomeric properties and are known for having a high resilience, tough and excellent resistance to abrasion [11-13] Furthermore they are preferred to be used in many application because can be processed using equipment such as extruder and injection molding machines. However, the low stiffness and poor resistance to heat limit application of pure polyurethane [4,5]. Incorporation of filler into polyurethane matrices such as CNTs in this study, give the prospect for tailoring the material properties to reach their technological potential [15-28]. The challenge in using CNTs as polymer reinforcing agent is the aggregation and agglomeration problem due to van der Waals interactions between nanotubes should be properly addressed in preparing a homogeneously distributed CNT/polymer nancomposites [3-5,7]. A uniform distribution of nanoparticles within polymer matrix and a good nanoparticle-polymer adhesion without destroying their integrity or reducing their aspect ratio are the key issues in producing CNT/polymer nanocomposites with outstanding properties [2-5,7, 29]. The dispersion of the nanotubes in polymer matrices can be done using techniques such as melt processing, solution processing, or in-situ polymerization. Melt mixing methods of CNTs/polymer nanocomposite is favored for dispersion of filler into polymer matrices because of the speed, simplicity and availability in plastic industrial processes [2,4,5]. Application of appropriate time, temperature and shear could diminish the affinity for aggregation and agglomeration of CNTs [3,4,7]. Therefore, this chapter is concentrated on preparation CNTs incorporated polyurethane nanocomposites and to analyze the mixing characteristics and other issues related to the properties of the nanocomposites at various loading and processing speed. The ThermoHaake minilab extruder was used to mix the MWNTs and polyurethane since the small scale mixer is more preferable to scrutinize and evaluate the influence of nanotubes and mixing methods

Thermal Analysis of Carbon Nanotubes Incorporated Polyurethanes Nanocomposites 397 on the properties of polymer composites. This chapter discussed the dynamic mechanical analysis, thermogravimetric and differential scanning calorimetry analyses of CNT filled polyurethane composites.

2. EXPERIMENTAL 2.1. Materials In this study, Thermoplastic polyurethane elastomers (TPU) Elastollan supplied from Elastogran GmbH, Germany was used as the matrix. This polymer is formed from the interreaction polyols (long-chain diols), diisocyanates, and short-chain diols. This polymer has specific heat of 1.5-1.8 J/g K (room temperature) and heat of combustion between 2600031000 J/g. As for fillers, multi-walled carbon nanotubes (MWNT) provided by FutureCarbon GmbH, Germany were used. The average dimensions for these MWNTs are 15 nm and purity >98% as synthesized.

2.2. Sample Preparation The polyurethane pellets were ground to powder form before compounding for better mixing. Then, to reduce the moisture effect, the polymer powder was properly dried in a vacuum oven for at least 2 hours at 80°C. Before mixing, the nanotubes were dried in vacuum as well for 3 hours at temperature around 200°C. After drying, both the polymers and MWNTs were manually mixed. After that, the premixed of polyurethane and MWNTs were fed into ThermoHaake minilab extruder for further compounding. The temperature used is 210°C with different rotating speeds to analyze the influences of the nanotoubes and screw rotation on the properties of the polymer composites. The polyurethane composites then were injected to bone and round shape samples by using a laboratory injection molding machine (All-rounder 320C 600-250, Arburg, Germany) at temperature of 210°C. In this chapter pure polyurethane is labeled as PU whereas the composites were labeled as PUXXCNTYYY. XX refers to the filler loading in wt% and YY refers to extruder screw rotation is rpm used in the process.

2.3. Characterization The conventional differential scanning calorimeter DSC-2920 (TA Instruments, Alzenau, Germany) was used to investigate the differential scanning calorimetry (DSC) measurements. The measurements were performed on nanocomposite samples with weight of around 8-10 mg at temperature range of -50 – 250°C and 10°C/min heating rate. The accuracy of the measurements is 3%. TAInstrument DMA 2980 at the frequency of 10 Hz was used to measure the dynamic mechanical properties (DMA). The scanning rate used is 10°C/min with the temperature range between -50 to 100°C while a constant nitrogen flow was used to purge the instrument. The information on mechanical behavior of the polyurethane composites at

398

Shahrul Azam Abdullah and Lars Frormann

the molecular level especially storage modulus, loss modulus and damping coefficient as well as various transition such as the glass transition can be analyzed by DMA. Moreover, thermogravimetric analyses (TGA) were done using the TA Instruments, Thermal Analysis and Rheology TGA-2950 (TA Instruments, Alzenau,Germany). The CNTs/polyurethane nanocomposite samples were heated from 0°C to 800°C using a constant heating rate of 10°C/min. A constant nitrogen flow was also used to purge the instrument.

3. RESULT AND DISCUSSION Differential Scanning Calorimetry (DSC) Figure 1, 2 and 3 shows the differential scanning calorimetry (DSC) results of CNTs/polyurethane nanocomposites. The DSC used to analyze the effect of MWNTs concentration on thermal properties by the difference in heat flow to/from the samples due to their thermal reactions. It is clear from figures that the materials show an amorphous behavior and the material melting were shown to be at <200°C. The rise in heat flow at temperature around -40 to 25°C responded to the glass transition of the material where the energy is absorbed. As seen in the figures, the presence of MWNTs lowered and broadened the peaks in the DSC curves of as well as the temperature at which the melting occurs were slightly modified. The crystallization temperature and the enthalpy of crystallization were increased but the crystallinity decreased as the amount of MWNTs increased which suggest the MWNTs particles nucleate the crystallization process and the strong filler/matrix interaction. Increasing the screw speed further lowered, broadened and shifts the curves further.

Dynamic Mechanical Analysis (DMA) The incorporation of CNTs into polyurethane matrix is anticipated to improve mechanical reinforcement of the composites. The elasticity of the materials can be measured by the value of storage modulus. The storage modulus against temperature of injection molded MWNTs/polyurethane composites measured at 10 Hz is depicted in Figure 4, 5 and 6. These figures show a sudden drop in the storage modulus value when the temperature was increased from below to above glass transition temperature (Tg) with respect to the polymer transformation from glassy to rubbery phase. The result shows an increment in storage modulus especially at lower Tg with addition of MWNTs. This also manifests the stiffness of the nanocomposites increased significantly with the presence of MWNTs which also suggest a good adhesion between MWNTs and the polyurethane matrix. High aspect ratio and smaller diameter of MWNTs as compared to other fillers have some consequence for effective load transfer and the reinforcement of the polyurethane. The storage modulus of the nanocomposites decreased sharply when the temperature was increased to the melting point (Tm) and move toward to that of pure polyurethane implying the storage modulus was dominated by the matrix intrinsic modulus [30].

Thermal Analysis of Carbon Nanotubes Incorporated Polyurethanes Nanocomposites 399

Figure 1. DSC thermograms of MWNTs/polyurethane nanocomposites for 1 wt% and 5 wt% at 100 rpm.

Figure 2. DSC thermograms of polyurethane nanocomposites with 5 wt% MWNTs processed at 100 and 150 rpm.

400

Shahrul Azam Abdullah and Lars Frormann

Figure 3. DSC thermograms of polyurethane nanocomposites with 1 wt% MWNTs processed at 100, 200 and 230 rpm.

Figure 4. Storage modulus vs. temperature for unfilled polyurethane and different concentration of MWNTs filled polyurethane composites at 100 rpm processing speed.

Thermal Analysis of Carbon Nanotubes Incorporated Polyurethanes Nanocomposites 401

Figure 5. Storage modulus vs. temperature for unfilled polyurethane and polyurethane composites with 5 wt% MWNTs processed at 100 and 150 rpm.

Figure 6. Storage modulus vs. temperature for unfilled polyurethane and polyurethane composites with 1 wt% MWNTs processed at 100, 200 and 230 rpm.

402

Shahrul Azam Abdullah and Lars Frormann

Furthermore, the polyurethane matrix also has the ability to sustain a high modulus value at high temperatures with the incorporation of MWNTs. At around 20°C the addition of 1 and 5 wt% results in about 5 and 20% increase in storage modulus for nanocomposites, respectively. The improvement effect of storage modulus with introduction of CNTs in the polymer was also reported in other paper [31]. Nonetheless, increasing the screw decreased the storage modulus for the nanocomposites with the same carbon fractions (1 wt%). A higher processing possibly disrupts the filler-polymer interaction and enables the movement of polymer chains. Loss modulus is measured as the ability of materials to dissipate the energy as heat during deformation and is used for assessing the viscous properties. Loss modulus plots for unfilled polyurethane and MWNTs/polyurethane are presented in Figure 7, 8 and 9. These figures show loss modulus curves with sudden peak where the materials were changed from energy elastic state to entropy elastic state. The result shows all nanocomposites formulation display loss modulus values higher than that of pristine polyurethane. Further addition of more MWNTs loading increased the value of loss modulus especially at temperature below 40°C as shown in Figure 7 when filler content was increased from 1 wt% to 5 wt%. The phase transition region (peak) also shifted to higher temperature with the addition of MWNTs. Higher glass transition temperatures suggest that the composites have more restricted molecules especially at area around MWNTs particles. However, increasing the processing speed decreased the values of loss modulus as shown in Figure 9 for 1 wt% MWNT. The tan δ or damping factor is the ratio of loss to storage modulus. Figure 10, 11 and 12 represents the tan δ as a function of temperature curves for unfilled polyurethane and MWNTs/polyurethane nanocomposites containing various volume fractions and processing speed. The pure polyurethane and the polyurethane nanocomposites exhibit increase in the tan δ with temperature and reached a peak where maximum heat dissipation occurs. The glass transition temperature can be determined using the peak in the tan δ versus temperature curve. It should be noted that all formulation shows that the glass transition temperature occurs over a range of temperature. The peak for pure polyurethane tan δ curve was found to be at around -31.93°C in this study. The tan δ curve changes with addition of MWNTs consequently change the Tg of the materials. The inclusion of MWNTs into the polyurethane matrix has lowered, broadened and flattened the peak position of tan δ. The peak of tan δ curve was also slightly shifted to a higher temperature with the presence of MWNTs which manifested the increases in the glass transition temperature (Tg). This reflects the effects of filler/matrix adhesion to the relaxation of the polymer chains where less energy was absorbed by the materials. The mobility and relaxation of the polymer segment were restricted with the presence of MWNTs particles and reduced the peak intensity as well as slightly increased the Tg [4,32]. The amount of immobile material is very high with the incorporation of MWNTs at relatively low loading because of their diameter and high surface/volume ratio [4]. A lower tan δ value indicates that the material is more elastic. The decrease of tan δ peaks was also associated with fewer amounts of materials participating in the transition [12]. PU01CNT100, PU01CNT200, and PCT230 show a similar shape and tan δ curves insinuating the similarity in filler/matrix interaction and mobility of the particles.

Thermal Analysis of Carbon Nanotubes Incorporated Polyurethanes Nanocomposites 403

Figure 7. Loss modulus vs. temperature for unfilled polyurethane and different concentration of MWNTs filled polyurethane composites at 100 rpm processing speed.

Figure 8. Loss modulus vs. temperature for unfilled polyurethane and polyurethane composites with 5 wt% MWNTs processed at 100 and 150 rpm.

404

Shahrul Azam Abdullah and Lars Frormann

Figure 9. Loss modulus vs. temperature for unfilled polyurethane and polyurethane composites with 1 wt% MWNTs processed at 100, 200 and 230 rpm.

Figure 10. Temperature dependence of tan δ for unfilled polyurethane and different concentration of MWNTs filled polyurethane composites at 100 rpm processing speed.

Thermal Analysis of Carbon Nanotubes Incorporated Polyurethanes Nanocomposites 405

Figure 11. Temperature dependence of tan δ for unfilled polyurethane and polyurethane composites with 5 wt% MWNTs processed at 100 and 150 rpm.

Figure 12. Temperature dependence of tan δ for unfilled polyurethane and polyurethane composites with 1 wt% MWNTs processed at 100, 200 and 230 rpm.

406

Shahrul Azam Abdullah and Lars Frormann

Figure 13. Storage and loss modulus as well as tan δ curves polyurethane composite with 5 wt MWNTs at (a) 150 rpm and (b) 100 rpm.

The results shown propose the improvement in mechanical properties of polyurethane nanocomposites with the addition of CNTs. Figure13 shows the storage and loss modulus as well as tan δ for polyurethane with 5 wt% MWNTs.

Thermogravimetry Analysis Figure 14, 15 and 16 presents the analysis for thermal degradation of the pure polyurethane and MWNTs/polyurethanes naocomposites samples. The pure polyurethane is observed to have two steps thermal degradation where the first decomposition step starts at temperature about 295°C meanwhile the second step started at temperature around 460°C. From the figure it can be seen that the pure polyurethane has the highest weight loss as a function of temperature and the TGA curves shows excellent agreement for the composites where the onset of the decomposition temperature were slightly increased when the MWNTs were introduced into the polymer matrix. The full decomposition temperature was shifted to higher temperature as well when the MWNTs particles are present which illustrate higher thermal stability. The increase in thermal degradation is related to the high thermal stability of MWNTs where the applied were mostly absorbed by the MWNTs particles and also the restriction effect of MWNTs on the polymer chains [6]. Higher screw rotation also marginally increased the degradation temperature of the nanocomposite because better distribution of MWNTs particles in the polyurethane matrix helps improve the thermal stability of the materials [32]. The results show that the thermal degradataion and thermal stability of polyurethane can be improved by addition of MWNTs and this result are in consistent other report previously published other matrices [33,34].

Thermal Analysis of Carbon Nanotubes Incorporated Polyurethanes Nanocomposites 407

Figure 14. Thermogravimetry result of unfilled polyurethane and different concentration of MWNTs filled polyurethane composites at 100 rpm processing speed as a function of temperature.

Figure 15. Thermogravimetry result of unfilled polyurethane and polyurethane composites with 5 wt% MWNTs processed at 100 and 150 rpm as a function of temperature.

408

Shahrul Azam Abdullah and Lars Frormann

Figure 16. Thermogravimetry result of for unfilled polyurethane and polyurethane composites with 1 wt% MWNTs processed at 100, 200 and 230 rpm as a function of temperature.

CONCLUSION MWNTs filled polyurethane composites were prepared by mixing and injection molding and the presence of MWNTs in polyurethane matrix display an improvement in their thermal characteristics. DSC analysis shows an amorphous where the glass transition region is between -40 to 25°C. The addition of MWNTs has lowered and broadened the peaks in the DSC curves of as well as the temperature at which the melting occurs were slightly modified and the curves further lowered, broadened and shifted further when the processing speed were increased. The addition of MWNTs results in higher storage moduli especially at lower Tg and better stiffness of the polyurethane composites reflects a good adhesion between MWNTs and the polyurethane matrix. Nonetheless, increasing the screw decreased the storage modulus for the nanocomposites. All nanocomposites formulation exhibit loss modulus values higher than that of pristine polyurethane. The incorporation of MWNTs into polyurethane also shifted the loss modulus peak to higher temperature. The tan δ peak of polyurethane composites were also broadened and lowered with the presence MWNTs which manifest the increases in glass transition temperature (Tg) as a result from the restriction of polymer segment mobility by the fillers. The high thermal stability of MWNTs helps to increase the thermal stability of polyurethane. Better distribution of MWNTs results from higher speed of the extruder screw also marginally increased the degradation temperature of the polymer nanocomposites.

Thermal Analysis of Carbon Nanotubes Incorporated Polyurethanes Nanocomposites 409

ACKNOWLEDGEMENTS This research is supported by a scholarship from Ministry of Higher Education and Universiti Teknologi MARA (UiTM), Malaysia under Skim Latihan Akademik Bumiputera (SLAB).

REFERENCES [1] [2] [3] [4] [5] [6] [7]

[8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22]

Ijima, S. Nature 1991, 354,56. Breuer, O.; Sundararaj, U. Polymer Composites 2004, 25, 630. Lin, B.; Sundararaj, U.; Pötschke, P. Macromol. Mat. Eng. 2006, 291, 227. Pötschke, P.; Bhattachryya, A. R.; Janke, A.; Goering, H. Composites Interfaces 2003, 10, 389. Pötschke, P.; Bhattacharyya, A. R.; Janke, A. European Polymer Journal 2004, 40, 137. Jose, M. V.; Dean, D.; Tyner, J.; Price, G.; Nyairo, E. J. Appl. Polym. Sci. 2007, 103, 3844. Pötschke, P.; Bhattachryya, A. R.; Janke, A.; Pegel, S.; Leonhardt, A., Täschner, C.; Ritschel, M.; Roth, S.; Hornbostel, B.; Cech, J. Fullerenes, Nanotubes, and Carbon nanostructures 2005, 13, 211 Pötschke,P.; Fornes, T. D.; Paul, D. R. Polymer 2002, 43, 3247. Jose, Moncy V.; Dean, D.; Tyner, J.; Price, G.; Nyairo, E. Journal of Applied Polymer Science 2007, 103, 3844. Subramoney, S. Adv. Mater. 1998,10(15), 1157. Kim, B. K.; Lee, S. Y.; Xu, M. Polymer 1996, 37, 5781. Buffa, F.; Abraham, G. A.; Grady, B. P.; Resasco, D. J. Polym. Sci: Part B: Polym Physics 2007, 45, 490. Gall, K.; Dunn, M. L.; Liu, Y.; Finch, D.; Lake, M.; Munshi, N. A. Acta Materialia 2002, 50, 5115. Liu, Y.; Gall, K.; Dunn, M. L.; McCluskey, P. Mechanics of Materials 2004, 36, 929. Li, F.; Qi, L.; Yang, J.; Xu, M.; Luo, X.; Ma, D. J.Appl. Polym. Sci. 2000, 75, 68. Chang, J-H.; An, Y. U.; J Polym Sci: Part B: Polym. Physics 2002, 40, 670. Koerner, H.; Price, G.; Pearce, N. A.; Alexander, M.; Vaia, R. A. Nat Mater 2004, 3, 115. Gall, K.; Dunn, M. L.; Liu, Y. Applied Physics Letters 2004, 85, 290. Koerner, H.; Liu, W.; Alexander, M.; Mirau, P.; Dowty, H.; Vaia, R. A. Polymer 2005, 46, 4405. Chen, Y.; Zhou, S.; Yang, H.; Wu, L. J. Appl. Polym. Sci. 2005, 95, 1032. Kuan, H.; Ma, C. M.; Chang, W.; Yuen, S. M.; Wu, H.; Lee, T. Composites Science and Technology. 2005, 65, 1703. Chou, C.; Hsu, S.; Chang, H.; Tseng, S.; Lin, H. Polymer Degradation and Stability 2006, 91, 1017.

410

Shahrul Azam Abdullah and Lars Frormann

[23] Buckley, P. R.; McKinley, G. H.; Wilson, T. S.; Small, IV, W.; Benett, W. J.; Bearinger, J. P.; McElfresh, M. W.; Maitland, D. J. IEEE Transactions on Biomedical Engineering 2006, 53, 2075. [24] Paik, I. H.; Goo, N. S.; Jung, Y. C.; Cho, J. W. Smart Mater. Struct. 2006,15, 1476. [25] Dan, C. H.; Lee, M. H.; Kim, Y. D.; Min, B. H.; Kim, J. Polymer 2006, 47, 6718. [26] Xia, H.; Song, M.; Zhang, Z.; Richardson, M. J. Appl. Polym. Sci. 2007, 103, 2992. [27] Rodríguez, J. G.; Carreira, P.; Garcia-Diez, A.; Hui, D.; Artiaga, A.; Liz-Marzán, L. M. Journal of Thermal Analysis and Calorimetry 2007, 87, 45. [28] Razzaq, M. Y.; Anhalt, M.; Frormann, L.; Weidenfeller, B. Materials Science and Engineering A 2007, 444, 227. [29] Senneth, M.; Welsh, E.; Li, W. Z.; Wen, J. G.; Ren, Z. F. Appl. Phys. A 2003, 76, 111. [30] Frormann, L.; Iqbal, A.; Abdullah, S. A. Journal of Applied Polymer Science 2008, 107, 2695. [31] Leelapornpisit, W.; Ton-That, M.; Perrin-Sarazin, F.; Cole, K. C.; Denault, J.; Simard, B. J Polym Sci: Part B: Polym. Physics 2005, 43, 2445. [32] Choi, Y. J.; Hwang, S.; Hong, Y. S.; Kim, J.; Ok, C.; Huh, W.; Lee, S. Polymer Bulletin 2005, 53, 393. [33] Yang, S.; Taha, T. J.; Serrato, D. V.; Hernandez, K.; Lozano, K. Compos B 2007, 38, 228. [34] Zhu, J.; Peng, H.; Rodriguez-Macias, F.; Margrave, J. L.; Khabashesku,V. N.; Imam, A. M.; Lozano, K.; Barrera, E. V. Adv. Funct. Mater. 2004, 14, 643.

In: Nanofibers: Fabrication, Performance, and Applications ISBN 978-1-60741-947-1 Editors: W. N. Chang © 2009 Nova Science Publishers, Inc.

Chapter 14

CARBON NANOTUBES INCORPORATED POLYURETHANES NANOCOMPOSITES FOR THERMAL AND ELECTRICAL CONDUCTIVE APPLICATIONS Shahrul Azam Abdullah and Lars Frormann1 University of Applied Science Zwickau, Institute for Production Technology, Zwickau, Germany

ABSTRACT Polyurethane composites filled with multi-walled carbon nanotubes (MWNTs) were prepared by mixing and injection molding and its thermal as well as electrical conductivity characteristics were investigated. The influences of MWNTs addition and mixing methods on thermal and electrical conductivity of MWNT/polyurethane nanocomposites were investigated using a high resistance meter and thermal conductivity analyzer. The electrical resistivity of MWNTs incorporated polyurethanes were decreased in relation to filler concentration which is attributed by the formation of a conductive path made up from MWNTs particle. Increasing the processing speed will further decrease the resistivity because the dispersion of CNTs in polymer is improved. Higher processing speed samples shows resistivity values closer to the theoretical value because of better dispersion of CNTs in polyurethane and more conductive pathway were formed. The result shows that the addition of MWNTs fillers improved the thermal conductivity of the polyurethane composites. Higher filler concentration and higher shear rate results in better thermal conductivity because better formation of thermally conductive networks along polymer matrix to ensure the thermal was conducted through the matrix and the network along the polymer composites. The theoretical thermal conductivity comparisons approximately agree with the experimental measurements for the composites studied. The present study revealed the potential of MWNTs as agent for better thermal and electrical conductivities of polyurethane nanocomposites and their properties depend strongly on the dispersion and distribution of nanotubes in polyurethane matrix.

1 Corresponding author: Prof. Dr-Ing. Lars Frormann, Tel.: +49 375 536 1721, Fax: +49 375 536 1713, E-mail address: [email protected].

412

Shahrul Azam Abdullah and Lars Frormann

Keywords: polyurethane; nanocomposite; carbon nanotubes; extrusion; electrical resistivity; thermal properties

1. INTRODUCTION Polyurethanes are polymers which demonstrate both thermoplastic and elastomeric properties due to their general repeat unit structure of hard and soft segments. Because of the unique hard and soft segment structure, polyurethanes possess a high resilience, tough, and excellent resistance to abrasion [1]. Their melt processable advantages with equipment such as extruder and injection molding machine also make them preferred to be used many industries. Reinforced polyurethane composites provide the possibility for tailoring the material properties especially their thermal and electrical conductivities to reach their technological potential. Carbon nanotubes (CNTs) have shown to be fascinating fillers since they encompass excellent thermal and electrical properties and the polymer conductivities can be improved at relatively low CNTs content. Incorporation of fillers into polymer matrices is one of the most favored methods to improve the properties of the polymer and overcome the low stiffness, poor electrical conductivity and resistance against heat [2-17]. CNTs were used as fillers in this study in order to enhance the thermal and electrical properties of polyurethane which are vital for conductive applications. CNTs are unique nanostructured cylindrical molecules which encompass of hexagonal network of carbon atoms rolled up to form seamless, hollow cylinder [18,19]. Furthermore, CNTs posses very high aspect ratio, as high as 100-1000 which makes them a great choice to be used as reinforcing agent especially in polymer composites [19-22]. In addition, their extraordinary mechanical properties, thermal conductivity and electrical stability are also particularly desired for nanocomposites preparation [18-23]. The dispersion and distribution of the CNTs within the matrices while maintaining their integrity and aspect ratio plays important role in determining the efficiency of CNTs as reinforcing agents in polymer composites [18-21,23,24]. Because of speed, simplicity, and plastic industrial friendly, melt mixing of CNTs into polymer matrix is the most preferred methods for composite formation [18,20,21]. Nonetheless, the major challenge in preparing homogeneously distributed CNT/polymer nanocomposites is the aggregation and agglomeration problem due to high Van der Waals interactions between nanotubes [19-21,23]. The penchant of the CNTs particles to form aggregates could be minimized by the application of appropriate time, temperature and shear during melt mixing process [19,20,23]. The preparation and characterization of polyurethane incorporated with multi-walled carbon nanotubes (MWNTs) is focused in this chapter with the concentration on the issue related to the thermal and electrical conductivity for MWNTs/polyurethane nanocomposites at various loading and processing speed. The aspiration of this research effort is to understand mixing characteristics and the conductivity properties of polyurethane and MWNTs nanocomposites. With the regard to MWNTs processing, a small scale mixer is more preferable to study the controlling factors for melt mixing and evaluating the nanocomposites in comparison to a big one [19]. The ThermoHaake minilab extruder was used to mix the MWNTs and polymer in this study. Experimentally obtained electrical and thermal properties

Carbon Nanotubes Incorporated Polyurethanes Nanocomposites…

413

of the extrudate were used to relate the influences of nanotubes and processing speed with the conductivity of the MWNTs filled polymer. This chapter deals with electrical resistivity and thermal conductivity of CNT filled polyurethane composites.

EXPERIMENTAL 2.1. Materials For this study, thermoplastic polyurethane (TPU) Elastolan formed from the interreaction polyols (long-chain diols), diisocyanates, and short-chain diols supplied from Elastogran GmbH, Germany was used as the matrix. This polymer matrix has the thermal conductivity range of 0.19-0.25 W/mK and the specific heat of 1.5-1.8 J/g K (room temperature). As for the fillers, multi-walled carbon nanotubes (MWNTs) with average dimension of 15nm and purity >98% as synthesized provided by FutureCarbon GmbH, Germany were used.

2.2. Sample Preparation The polymer pellets were ground to powder form for better mixing. Then, the materials were properly dried before mixing where the TPU was dried in a vacuum oven at for at least 2 hours at 80°C. The MWNTs were dried as well for 3 hours at around 200°C prior mixing. The mixing was performed in two steps. The TPU and MWNTs were first mixed manually before being fed into the extruder for further compounding. ThermoHaake minilab extruder were used for compounding the MWNTs/PU nanocomposites at temperature of 210°C with different rotating speeds to analyze the effect of the nanotoubes and processing screw speed on the conductivity properties of the polymer composites. A laboratory injection molding machine (All-rounder 320C 600-250, Arburg, Germany) at 210°C then used to prepare the round shape samples with average diameter of 50 mm and 2 mm thickness for the thermal and electrical conductivity. The pristine polyurethane in this article is labeled as PU while the MWNTs/polyurethane nanocomposites were denoted as PUXXCNTYY where XX refers to the filler loading and YY refers to extruder screw speed used in the process.

2.3. Characterization Electrical resistivity was measured directly on the round samples using a high resistance meter (Model 4339B, Agilent Technologies, Böblingen, Germany) with output voltage 100 mV. This resistance meter is designed for measuring high resistance and related parameters of insulation materials where the volume resistivities of all the samples were measured in the normal direction across the thickness. The instrument was calibrated as recommended by the manufacturer. The volume resistivity results were also measured as a function of temperature. The thermal conductivity measurements of the composites were performed using Thermal Conductivity Analyzer (TCA-200LT-A, Netzsch, Germany) with the guarded heat

414

Shahrul Azam Abdullah and Lars Frormann

flow meter method and was measured as temperature dependence. The injection molded discs were placed between to heated plates controlled at different temperature where the heat flow from hot to cold surface and then the thermal conductivity data were collected when the samples were introduced to a temperature range of 25 to 80°C with the accuracy of 4%. The measurements were also done in normal direction across the sample thickness and coupling agent was applied to reduce the thermal resistance of the samples and adjacent plate interfaces.

3. THEORETICAL CONSIDERATION Selected Theoretical Models for Electrical Conductivity/Resistivity Many useful models have been developed to explain the experimental results and to predict new applications of composite materials. Nielsen model [25], where

(1) (2) (3)

McCullough model [26],

(4) Ondracek model [27],

(5)

A volume resistivity model for electrically conductive composite was proposed by Weber and Kamal [28].

(6) The parameter ρc and ρf are the resistivity of the composite and fiber respectively while d, dc, l, and θ denote the diameter of the fiber, diameter of circle of contact, fiber average length and fiber orientation angle. ∅p reflects the volume fraction of fibers involved in conductive pathway formation whereas X is depending on fibers contact numbers, m and can be calculate by,

415

Carbon Nanotubes Incorporated Polyurethanes Nanocomposites…

(7) The conductivity of the matrix is ignored in this model because it is assumed to be very low and has no influence on the conductivity of the composite. Taipalus and Friedrich [29] proposed a new model to predict the conductivity of composites with the influence of the conductivity of the matrix is taken into consideration.

(8) Where σc, σm, and σf represent the conductivity of the composite, matrix and fibers respectively. Selected theoretical models for thermal conductivity A number of researchers have proposed theoretical model for thermal conductivity of composite. Bruggeman model [30], (9) Böttcher model [31],

(10) deLoor model (φ<0.2) [32],

(11) van Beek model (φ<0.2) [33],

(12) Agari and Uno [34] proposed thermal conductivity model for particle filled composite. (13) kc, km and kf represent the thermal conductivity of the composite, matrix and fillers respectively while ∅ are the volume fraction of fillers. C1 measure the influence of the particles on crystallinity while C2 describe the easiness to form conductive chains of fillers within the matrix.

416

Shahrul Azam Abdullah and Lars Frormann

Though, since nanotubes have a very high aspect ratio and most the equation ignored the interaction amongst the filler as well as kf of nantoubes is much larger as compared to km of the polymer matrix, these thermal conductivity model above is not suitable for CNTs filled polyurethanes. A new thermal conductivity model for nanotubes composites based on an effective medium approach (EMA) has been proposed by Nan et al. [35] where the interface effect on the thermal conductivity of the composite were also taken into account.

(14)

kc, km and kf are thermal conductivity of composite, matrix, and fillers respectively. L and ∅ are the length and volume fraction of the CNTs used whilst Ri is interface thermal resistance or Kapitza resistance [36,37]. Huxtable et al. [36] reported that the interface thermal resistance across the CNT and matrix is about about 8.0 x 10-8 m2/KW. Cos2θ is considered to be 1/3 in this study as the nanotubes are randomly dispersed in the polymer matrix [36,37].

Result and Discussion The aim of adding MWNTs to TPU was mainly to obtain composites with certain electrical and thermal conductivity for better conductive applications. The electrical properties of nanocomposites were influenced by the condition of microstructures (dispersion state, filler geometry, and filler–filler interaction) [38]. Additions of MWNTs generally decrease the electrical resistivity as shown in Figure 1. The formation of conductive path of MWNTs fillers through the matrix leads the decrease in the bulk resistivity [18,38]. The electrical resistivity result shows a reduced in value with addition of even small amount MWNTs due to the shape and higher aspect ratio of MWNTs which make the particle/particle and particle/matrix interaction easier [18,38]. Similar behavior were also reported by other researchers [23,39]. The reduced electrical resistivity of 1 wt% CNT filled polyurethane as a function of screw speed is plotted in Figure2. Figure3 shows the electrical resistivity of pure PU and 5wt% CNT/PU nanocomposite processed at 100 and 150 rpm. Both figures show the decreased in resistivity when the screw speed increased for composites with same MWNTs loadings. The MWNTs aggregates were not properly dispersed at lower processing speeds which make the composites lack the conductivity network through the matrix. At higher processing speed, the nancomposites samples processed shows a lower resistivity signifying the MWNT particles were better distributed and dispersed to form the desired conductivity path. From Table 1, the theoretical is observed to have a slightly lower resistivity as compared to experimental values for all samples because the filler were assumed to be homogeneously dispersed whereas in experimental some of the nanotubes are still in bundle and less conductive pathway were formed.

Carbon Nanotubes Incorporated Polyurethanes Nanocomposites…

Figure 1. Electrical resistivity value for pure polyurethane and MWNTs filled polyurethane nanocomposites for 1 wt% and 5 wt% at 100 rpm.

Figure 2. Effect of extruder processing speed on the electrical resistivity of 1 wt% MWNTs filled polyurethane nanocomposites.

417

418

Shahrul Azam Abdullah and Lars Frormann

Figure 3. Electrical resistivity of pure polyurethane and 5 wt% MWNTs/polyurethane nanocomposite processed at 100 and 150 rpm.

Table 1. Comparison of experimental and theoretical results for the electrical resistivity of unfilled polyurethane, 1 wt% and 5 wt% MWNTs/polyurethane nanocomposites at different processing speed

Experimental

PU PU01CNT100 PU01CNT200 PU01CNT230 PU05CNT100 PU05CNT150 Weber and Kamal (eq. 6)

Theoretical Taipalus and Friedrich (eq. 7)

CNTs content (wt%) 0 1 1 1 5 5 1 5 1 5

Electrical resistivity (Ωcm) 3.1789 x 1011 1.6376 x 1010 1.4237 x 1010 8. 2959 x 109 1.5642 x 1010 1.5100 x 1010 4.8752 x 109 1.52778 x 109 4.8016 x 109 1.5205 x 109

The resistivity value for samples processed at higher shear rate shows resistivity values closer to the theoretical because of better dispersion of CNTs in polyurethane and more conductive pathway were formed.

Carbon Nanotubes Incorporated Polyurethanes Nanocomposites…

419

Figure 4. Thermal conductivity vs. temperature for unfilled polyurethane and different concentration of MWNTs filled polyurethane composites at 100 rpm processing speed.

The formation of thermally conductive networks along polymer matrix is very important in preparing materials with enhanced thermal conductivity. The thermal conductivity for MWNTs filled PU nanocomposites plotted as a function of temperature was illustrates in Figure 5. In general, the addition of nanotubes increased the thermal conductivity of the polyurethane composites. The thermal conductivity value for unfilled polyurethane investigated is 0.1424 W/mK at 45°C. Figure 5 and 6 shows at 45°C, the addition of 1 wt% increased the thermal conductivity of the materials by 31% to 0.18663 W/mK for PU01CNT100, while addition of 5 wt% further increased the conductivity to 0.21272 W/mK for PU05CNT100. The conductivity increment for the material can be seen at all temperatures. In this study the highest thermal conductivity for 1 wt% MWNTs loading at all temperatures was PU01CNT200 samples processed at 200 rpm. The high thermal conductivity of fillers up to 3000 W/mk for MWNTs has contributed to the increase in thermal conductivity of the composites [35]. The heat was conducted by two means, through interconnected MWNTs structures and also through filler and matrix interfaces along the polymers composite. Since the interconnected MWNTs conduct the heat better than filler/matrix interface, the formation of the conductive filler network in the matrix is preferred. Higher fillers concentration shows better thermal conductivity since the fillers are better connected in the polymer matrix which results in better thermal conductivity. Kashiwage et. al. [40] also reports similar result where thermal conductivity increases with increase in the amount of CNTs. The thermal conductivity of the composites was also slightly influenced by screw processing speed. Lower processing speed was unable to initiate a proper dispersion to enable the heat conductivity.

420

Shahrul Azam Abdullah and Lars Frormann

Figure 5. Thermal conductivity of unfilled polyurethane and 1 wt% MWNTs/polyurethane nanocomposites processed at 100, 200 and 230 rpm at various temperatures.

Figure 6. Thermal conductivity of unfilled polyurethane and 5 wt% MWNTs/polyurethane nanocomposites processed at 100 and 150 rpm at various temperatures.

Carbon Nanotubes Incorporated Polyurethanes Nanocomposites…

421

Table 2. Comparison of experimental and theoretical results for the thermal resistivity of unfilled polyurethane, 1 wt% and 5 wt% MWNTs/polyurethane nanocomposites at different processing speed

Experimental

PU PU01CNT100 PU01CNT200 PU05CNT100 PU05CNT150

Theoretical

Nan et al. (eq. 14)

CNTs content (wt%) 0 1 1 5 5 1 5

Thermal conductivity (W/mK) 0.14242 0.18663 0.22576 0.21272 0.23632 0.2060 – 0.2660 0.2699 – 0.3299

A higher processing speed will disperse and steer the filler to form conductive network for improved conductivity. Using Nan et al. model, the theoretical value for thermal conductivity for 1 wt% CNTs filled polyurethanes is found to be in the range from 0.19024 to 0.25024 W/mK. Table 2 shows comparison of experimental and theoretical result for the thermal conductivity of MWNTs/polyurethane nanocomposites. These theoretical thermal conductivity ranges approximately agree with the experimental measurements for the composites studied.

CONCLUSIONS Carbon nanotubes (CNTs) filled polyurethane composites were prepared by mixing and injection molding. The additions of CNTs into polyurethane matrix display an improvement in electrical and thermal conductivities. The electrical resistivity was generally decreased with the additions of CNTs filler. The resistivity was reduced even when small amount of MWNTs were added due to the influence of high aspect ratio and the condition of microstructures which alleviate the formation of conductive path of MWNTs fillers through polyurethane matrix. Increasing the processing speed decreased the electrical conductivity signifying the MWNTs were better distributed and dispersed to form the desired conductivity path. The theoretical is observed to have a slightly lower resistivity as compared to experimental values for all samples where the result for samples processed at higher shear rate shows resistivity values closer to the theoretical. The addition of nanotubes increased the thermal conductivity of the polyurethane composites at all temperatures investigated because of the high thermal conductivity of MWNTs fillers. The heat was conducted through interconnected MWNTs structures and also through filler and matrix interfaces along the polymers composite. Higher fillers concentration shows better thermal conductivity since the fillers are better connected in the polymer matrix. The thermal conductivity of the composites was also slightly influenced by screw processing speed. A higher processing speed will disperse and steer the filler to form better conductive network for improved conductivity. The theoretical value for thermal conductivity ranges approximately agree with the experimental measurements for the composites studied.

422

Shahrul Azam Abdullah and Lars Frormann

ACKNOWLEDGEMENTS This research is supported by a scholarship from Ministry of Higher Education and Universiti Teknologi MARA (UiTM), Malaysia under Skim Latihan Akademik Bumiputera (SLAB).

REFERENCES [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12]

[13] [14] [15] [16] [17] [18] [19] [20] [21] [22]

Buffa, F.; Abraham, G. A.; Grady, B. P.; Resasco, D. J. Polym. Sci: Part B: Polym Physics 2007, 45, 490. Gall, K.; Dunn, M. L.; Liu, Y.; Finch, D.; Lake, M.; Munshi, N. A. Acta Materialia 2002, 50, 5115. Liu, Y.; Gall, K.; Dunn, M. L.; McCluskey, P. Mechanics of Materials 2004, 36, 929. Li, F.; Qi, L.; Yang, J.; Xu, M.; Luo, X.; Ma, D. J Appl Polym Sci 2000, 75, 68. Chang, J-H.; An, Y. U.; J Polym Sci: Part B: Polym. Physics 2002, 40, 670. Koerner, H.; Price, G.; Pearce, N. A.; Alexander, M.; Vaia, R. A. Nat Mater 2004, 3, 115. Gall, K.; Dunn, M. L.; Liu, Y. Applied Physics Letters 2004, 85, 290. Koerner, H.; Liu, W.; Alexander, M.; Mirau, P.; Dowty, H.; Vaia, R. A. Polymer 2005, 46, 4405. Chen, Y.; Zhou, S.; Yang, H.; Wu, L. J. Appl .Polym. Sci .2005, 95, 1032. Kuan, H.; Ma, C. M.; Chang, W.; Yuen, S. M.; Wu, H.; Lee, T. Composites Science and Technology 2005, 65, 1703. Chou, C.; Hsu, S.; Chang, H.; Tseng, S.; Lin, H. Polymer Degradation and Stability 2006, 91, 1017. Buckley, P. R.; McKinley, G. H.; Wilson, T. S.; Small, IV, W.; Benett, W. J.; Bearinger, J. P.; McElfresh, M. W.; Maitland, D. J. IEEE Transactions on Biomedical Engineering 2006, 53, 2075. Paik, I. H.; Goo, N. S.; Jung, Y. C.; Cho, J. W. Smart Mater Struct. 2006,15, 1476. Dan, C. H.; Lee, M. H.; Kim, Y. D.; Min, B. H.; Kim, J. Polymer 2006, 47, 6718. Xia, H.; Song, M.; Zhang, Z.; Richardson, M. J. Appl. Polym .Sci. 2007, 103, 2992. Rodríguez, J. G.; Carreira, P.; Garcia-Diez, A.; Hui, D.; Artiaga, A.; Liz-Marzán, L. M. Journal of Thermal Analysis and Calorimetry 2007, 87, 45. Razzaq, M. Y.; Anhalt, M.; Frormann, L.; Weidenfeller, B. Materials Science and Engineering A 2007, 444, 227. Breuer, O.; Sundararaj, U. Polymer Composites 2004, 25, 630. Lin, B.; Sundararaj, U.; Pötschke, P. Macromol. Mat. Eng. 2006, 291, 227. Pötschke, P.; Bhattachryya, A. R.; Janke, A.; Goering, H. Composites Interfaces 2003, 10, 389. Pötschke, P.; Bhattacharyya, A. R.; Janke, A. European Polymer Journal 2004, 40, 137. Jose, M. V.; Dean, D.; Tyner, J.; Price, G.; Nyairo, E. J. Appl. Polym. Sci. 2007, 103, 3844.

Carbon Nanotubes Incorporated Polyurethanes Nanocomposites…

423

[23] Pötschke, P.; Bhattachryya, A. R.; Janke, A.; Pegel, S.; Leonhardt, A., Täschner, C.; Ritschel, M.; Roth, S.; Hornbostel, B.; Cech, J. Fullerenes, Nanotubes, and Carbon nanostructures 2005, 13, 211. [24] Senneth, M.; Welsh, E.; Li, W. Z.; Wen, J. G.; Ren, Z. F. Appl. Phys. A 2003, 76, 111. [25] Nielsen, L.E. Ind. Eng. Chem. Fundam. 1974,13,17. [26] McCullough, R. L. Compos. Sci. Tecnol. 1985,22,3. [27] Ondracek, G. Rev. Powder Metall Phys. Ceram. 1987,3, 1205. [28] Weber, M.; Kamal, M. R. Polym. Comp. 1997, 18, 711. [29] Taipalus, R., Friedrich, K. Int Conf on Composite Materials ICCM-12,1999, Paris, France, 359. [30] Bruggeman, D. Ann. Phys. 1935, 24, 636. [31] Böttcher, C. Theory of Electric Polarisation; Elsevier: Amsterdam,The Netherlands, 1952. [32] deLoor, G. PhD Thesis, University of Leiden, NL: Leiden, 1956. [33] van Beek, L. Prog. Dielect. 1967, 7, 69. [34] Agari, Y.; Uno, A. J. Appl. Polym. Sci. 1986, 32, 5705. [35] Nan, C.; Liu, G.; Lin, Y.; Li, M. Applied Physics Letters 2004, 85, 3549. [36] Huxtable, S.T.; Cahill, D. G.; Shenogin, S.; Xue, L.; Ozisik, R.; Barone, P.; Usrey, M.; Strano, M. S.; Siddons, G.; Shim, M.; Keblinski, P. Nat. Mater. 2003, 2, 731. [37] Sivakumar, R.; Guo, S.; Nishimura, T.; Kagawa, Y. Scripta Materilia 2007, 56, 265. [38] Song, Y. S.; Youn, J. R. Carbon 2005, 43, 1378. [39] Andrews, R.; Jacques, D.; Minot, M.; Rantell, T. Macromol Mater Eng .2002, 287, 395 [40] Kashiwagi, T.; Grulke, E.; Hilding, J.; Groth, K.; Harris, R.; Butler, K.; Shields, J.; Kharchenko, S.; Douglas, J. Polymer 2004, 45, 4227.

INDEX A absorbents, ix, 185 absorption, 126, 208, 228, 303, 355, 357, 358, 366 academic, 220, 325 acceleration, 88, 167, 382 acceptor, 315, 318, 326 accessibility, 215 accidental, 7 accommodation, 372 accuracy, 134, 219, 401, 418 acetate, 66, 83, 117, 135, 157, 228, 249 acetic acid, 63, 64, 66, 82, 87, 89, 90, 91, 96, 97, 99, 100, 102, 107, 116, 131, 232 acetone, 82, 228, 229, 249, 382 acetylcholine, 232 acetylcholinesterase, 230, 232 acetylene, 9, 14, 19, 199 acidic, 64, 65, 66, 107, 130, 132, 133, 138, 231, 330, 333 ACL, 344, 377 acoustic, x, 225, 235, 241, 381 acoustic emission (AE), 235 acrylate, 128, 228, 229 acrylic acid, 74, 82, 89, 95, 97, 192, 229, 275, 285, 300, 301, 303, 307, 312 acrylonitrile, 89, 266, 275, 279, 285, 300, 301, 302, 303, 304, 307, 311, 312, 313, 326, 342 ACS, 90, 129, 130, 270 activated carbon, 205, 206, 207, 209, 212, 327 activation, 12, 14, 51, 155, 197, 205, 208, 226, 389, 390 activation energy, 226 activation volumes, 390 active site, 20, 198, 215 actuators, ix, 185, 218 acylation, 89, 107, 115

additives, ix, 81, 86, 118, 184, 235, 246, 248, 263 adducts, 130 adenine, 230, 232, 234, 235 ADH, 234 adhesion, x, xii, 12, 88, 101, 104, 106, 125, 126, 127, 212, 233, 273, 277, 279, 281, 285, 290, 291, 293, 300, 302, 304, 345, 347, 351, 358, 364, 372, 376, 385, 399, 400, 402, 406, 412 adhesion force, 12 adsorption, x, 20, 64, 123, 134, 208, 209, 210, 212, 226, 228, 229, 263, 273, 274, 277, 279, 291, 293, 294, 295, 296, 297, 298, 299, 300, 302, 303, 304, 309, 311, 327 adult, xi, 329, 332, 341, 342, 343, 344, 345 aerobic, 372 aerosol, 63 aerospace, 204, 215 Ag, 99, 236 agar, 124 agent, vii, xiii, 61, 99, 102, 118, 124, 128, 129, 132, 236, 288, 399, 400, 415, 416, 418 agents, viii, 73, 89, 118, 119, 208, 291, 416 agglutination, 105 aggregates, 257, 416, 420 aggregation, 105, 323, 326, 400, 416 agricultural, 227 agriculture, viii, 73, 87, 89 AIBN, 279, 312 aid, 26, 339 air, vii, 61, 63, 67, 79, 83, 97, 98, 158, 193, 212, 216, 227, 228, 241, 252, 265, 384 Air Force, 396 air quality, 227 airborne particles, 217 alcohol, 75, 82, 123, 181, 182, 230, 233, 234 alcohol oxidase, 230, 233, 234 aldehydes, 130

426

Index

alkali, 86 alkaline, 102, 204 alkenes, 227 alkylation, 89 allergens, 63, 216 alloys, 31, 226 allylamine, 74, 135 alpha, 304 alternative, 86, 113, 120, 134, 208, 210, 330 alternatives, 225 alters, 326 aluminum, 8, 21, 65, 110, 157, 252, 312, 331, 385 amide, 103, 227, 321, 354, 355, 357, 358, 372 amine, 64, 321, 328, 331 amines, 321 amino, vii, viii, 61, 73, 86, 88, 89, 101, 102, 117, 121, 122, 128, 130, 137, 231, 233, 288, 321, 323, 331, 375 amino acid, 321, 323, 331, 375 amino groups, viii, 73, 88, 89, 101, 102, 117, 121, 122, 128, 130, 137, 231, 288, 321 ammonia, 9, 14, 101, 102, 215, 229, 230 ammonium, 9, 81, 118 amoeboid, 373 amorphous, ix, 12, 13, 25, 184, 195, 212, 227, 264, 402, 412 amorphous carbon, 12, 13, 25, 195, 227 amplitude, 22 analgesics, 332 analog, 6 anatase, 212 angiogenesis, 331, 344 angiogenic, 341, 344 animals, 291, 332, 333, 335, 338, 339, 341 anisotropy, 380 annealing, 9, 17, 25 anode, 8, 33, 35, 36, 37, 38, 39, 41, 44, 52, 206, 207, 228, 266, 351 antenna, 204 antibacterial, viii, 64, 70, 73, 75, 100, 104, 107, 123, 124, 125, 294 antibacterial properties, 125 antibiotics, 332 antibody, 230, 233, 333, 334, 335, 336, 337, 339, 340, 342, 352 anticoagulants, 88 antigen, 230, 233 antioxidant, 88, 107, 301 anti-tumor, 87, 107 apoptosis, x, 273 appetite, 332 application, viii, ix, 8, 26, 54, 63, 70, 73, 75, 77, 78, 86, 87, 90, 102, 103, 104, 106, 107, 115, 117,

120, 122, 128, 129, 134, 135, 136, 139, 161, 185, 186, 203, 205, 207, 219, 302, 323, 336, 339, 348, 381, 400, 416 aqueous solution, 68, 81, 82, 87, 89, 90, 91, 96, 97, 99, 101, 107, 108, 109, 110, 113, 119, 130, 134, 136, 350, 351, 361, 363, 375 arc plasma, 17 argon, 8, 26, 31, 42, 193, 260, 262 aromatic polyimide, 316, 318, 325, 327 arthropods, 63, 64 articular cartilage, 157 artificial organs, vii ascorbic, 230, 232, 233 ascorbic acid, 230, 232, 233 aseptic, 332, 352 Asia, 102 aspect ratio, viii, 4, 22, 25, 41, 42, 49, 51, 52, 73, 198, 380, 400, 402, 416, 420, 425 aspiration, 416 assessment, 127, 333, 338, 339, 376 assumptions, 23 ASTM, 67, 71, 386 astrocytes, 333, 341, 342, 343, 344 atmosphere, 154, 190, 192, 195, 197, 199, 205, 252, 260, 265, 349, 351, 385 atmospheric pressure, 229 atomic force, xii, 40, 219, 347, 349, 351, 356 atomic force microscope, 40, 351, 356 atomic force microscopy (AFM), xii, 40, 45, 219, 347, 349, 351, 356, 357, 359 atomic orbitals, 1 atoms, 2, 3, 6, 10, 12, 16, 21, 24, 69, 88, 191, 196, 198, 200, 201, 208, 226, 400, 416 atrial fibrillation, 127 attachment, xii, 14, 99, 103, 105, 126, 127, 155, 233, 330, 331, 347, 349, 368, 370, 376 attenuated total reflectance, 288 autofluorescence, 334, 335 automobiles, 379 availability, ix, 184, 400 axon, 335, 336, 338, 343, 345 axonal, xi, 329, 330, 331, 332, 334, 338, 339, 341, 342, 345 axons, xi, 329, 330, 331, 332, 333, 334, 335, 336, 337, 338, 340, 341, 342, 345

B BAC, 63 Bacillus subtilis, 87 backscattering, 52 bacteria, 86, 88, 101, 123, 124, 125, 126, 348 bacterial, 124, 294 bacterial cells, 124

427

Index band gap, 26, 30 barrier, 22, 23, 24, 30, 35, 50, 139, 380 barriers, 216 basal lamina, 344 basement membrane, 330, 340 batteries, viii, 155, 183, 184, 204, 205, 206, 219 battery, ix, 52, 154, 184, 185 BBB, xii, 330, 333, 338 beams, 243 behavior, x, xii, 4, 25, 78, 87, 89, 91, 102, 103, 104, 120, 121, 128, 154, 155, 159, 161, 205, 208, 216, 239, 242, 243, 277, 303, 345, 349, 366, 368, 372, 379, 380, 381, 393, 396, 397, 399, 401, 402, 420 behavioral assessment, 339 benchmarks, 7 bending, 25, 78, 188, 190, 191, 192, 252, 274, 283, 286, 288, 357, 358, 372 benefits, 208, 254, 315, 331, 339 benzene, 2, 199, 209, 228, 229 Bessel, 53 bicarbonate, 9 binding, xi, 1, 66, 126, 226, 231, 232, 233, 273, 274, 291, 297, 301, 302, 311, 314, 315, 355, 357, 363 binding energy, 1, 355, 357 biocatalysis, 315 biocide activity, 88, 124 biocompatibility, vii, viii, 61, 64, 73, 99, 103, 104, 114, 127, 128, 157, 227, 230, 277, 279, 285, 294, 300, 302, 348, 353, 377 biocompatible, 70, 87, 107, 139, 226, 233, 348 biodegradability, viii, 64, 73, 99, 107, 114, 157, 294, 348 biodegradable, vii, 61, 70, 87, 107, 139, 241, 348, 372, 374, 376, 377 biodegradation, 117, 349, 351, 372 biological activity, 87, 88, 103, 331 biological behavior, 89 biological processes, x, xi, 273, 291, 305 biological stability, 344 biological systems, xi, 305, 306 biomacromolecules, x, 273, 300 biomarker, 234 biomaterial, 348 biomaterials, ix, 184 biomedical applications, 76, 86, 102, 107, 113, 123, 203, 303, 327, 348, 373, 374, 375 biomolecules, ix, x, 184, 225, 227, 230, 231 biopolymers, vii, 61, 63, 64, 65, 66, 68, 70, 348, 354, 376 bioreactor, 139, 203, 302 biosensors, x, xi, 139, 225, 231, 232, 234, 235, 305 biosynthesis, 86 biotechnology, 87, 89, 219, 220

biotin, 236 bipedal, 343 bipolar, 341 birefringence, 190 bladder, 333 blood, vii, x, 88, 89, 104, 105, 106, 113, 203, 273, 281, 291, 300, 302, 333, 340, 341, 342, 376 blood vessels, vii, 203, 333, 340, 341, 342 body fluid, 90, 102, 123 body temperature, 332 Boeing, 215, 219 boiling, 82, 100, 190 Boltzmann constant, 46 bonding, 2, 81, 108, 229, 230 bonds, 2, 4, 6, 52, 91, 130, 131, 228, 236, 348, 350 boutons, 337, 338 bovine, x, 273, 277, 351, 364, 368, 376 brain, 231, 341 branching, 155, 336, 345 BrdU, xii, 348, 352, 367, 376 breakdown, 228, 252 broad spectrum, 99 bubbles, 67 buffer, 74, 102, 126, 138, 297, 333 building blocks, viii, 52, 184 Bulgaria, 73, 140, 142 bulk materials, 384 bundling, 138, 139, 227, 231 burning, 1 butadiene, 241, 380 butadiene-styrene, 380 by-products, 196

C calcitonin, 333, 344 calcium, 348 calf, 230, 232 CAM, 379 Canada, 62, 143, 153 cancer, 218, 233 Candida, 302, 303 candidates, viii, 73, 88, 107, 126, 128, 139, 184, 220, 306, 373, 380 capacitance, 205, 206 capillary, 77, 78, 79, 81, 82, 99, 138, 153, 187, 301 caprolactone, 76, 100, 101, 103, 104, 241, 348, 372 carbazole, 241 carbohydrate, x, 273, 274, 275, 277, 279, 285, 286, 287, 291, 294, 298, 300, 301, 302, 303 carbon atoms, 2, 6, 10, 12, 16, 21, 88, 191, 196, 198, 201, 208, 400, 416 carbon dioxide, 227 carbon film, 26

428

Index

carbon materials, ix, 1, 7, 86, 185, 205, 206, 208, 210, 227 carbon molecule, 217 carbon monoxide, 226, 380 carbon nanotubes (CNTs), ix, xii, xiii, 2, 4, 6, 7, 8, 11, 12, 13, 14, 15, 16, 17, 18, 24, 25, 26, 30, 31, 43, 45, 51, 52, 53, 86, 184, 205, 225, 227, 231, 234, 257, 303, 304, 312, 315, 316, 326, 380, 397, 399, 400, 401, 402, 406, 410, 415, 416, 417, 420, 422, 423, 425 carbonates, 9 carbon-fiber, 215, 220 carbonization, 192, 195, 197, 208, 210, 260, 261, 262, 263 carbonyl groups, 194 carboxyl, 227 carboxyl groups, 88, 99, 121, 128, 354 carboxylic, 122, 136, 137, 227, 229, 285, 288 carboxylic groups, 122, 136, 137, 285, 288 carboxymethylcellulose, 89 carcinogens, 216 carcinoma, 230 cardboard, 74, 112 carrier, 8, 307, 311 cartilage, 376 cast, 125, 127, 232 casting, 124, 125, 233, 235 catalase, 307, 308, 309, 310, 311, 312, 315, 325, 326, 327 catalysis, xi, 1, 12, 70, 223, 305, 306, 307, 315, 325 catalyst, ix, 6, 8, 10, 11, 12, 13, 14, 15, 16, 17, 20, 21, 25, 28, 29, 30, 31, 33, 39, 40, 42, 45, 117, 184, 199, 201, 203, 204, 205, 210, 212, 214, 215, 226, 227, 241, 279, 288, 331, 380 catalytic activity, 233, 234 catalytic C, 7, 9 catalytic system, 86, 215 cathode, 17, 32, 33, 52, 228, 380 cavitation, 330, 381 cavities, 17, 254, 342, 381 C-C, 226 CDC, 288, 296 cefazolin, 123 cell adhesion, x, xii, 88, 106, 126, 233, 273, 279, 285, 300, 302, 304, 345, 347, 351, 358, 364 cell culture, 101, 126, 274, 352 cell growth, 126 cell line, 127, 344 cell membranes, x, 70, 88, 273, 364 cell organelles, 88 cell surface, 274 cellulose, vii, 61, 75, 83, 86, 128, 135, 149, 182, 241, 249, 269, 302

central nervous system, 341 central pattern generator, 339 ceramic, 215, 232, 240, 264 ceramics, x, 239, 265 CH3COOH, 93, 94, 95, 96, 138 CH4, 12 chain molecules, 192 chain rigidity, 83 chain scission, 196 channels, 340, 345 charge density, 108, 118, 154, 189, 242, 246, 251, 284 charring, 217 chemical agents, 208 chemical bonds, 1, 235 chemical deposition, 258 chemical etching, 9, 227 chemical industry, 318 chemical properties, 1, 88, 219, 231 chemical reactions, 285, 287, 301 chemical reactivity, x, 225, 226 chemical sensing, 236 chemical stability, 25, 52, 65, 231 chemical structures, 288 chemical vapor deposition, 9, 12, 14, 16, 26, 30, 31, 198, 226 chemical vapour, 8, 9 chemical vapour deposition, 8, 9 chemicals, 216 China, 93, 95, 96, 182, 273, 300, 305, 325 chiral, 4 chirality, 4, 227 chitin, 61, 63, 64, 65, 66, 67, 68, 69, 70, 86, 87, 101, 143, 144, 145, 146, 147 chloride, 67, 81, 128, 205, 206, 208, 227, 241, 267, 321 chloroform, 80, 82, 114, 229, 322, 324 cholesterol, 88 chondrocytes, 103, 376 chronic disease, 88 chronic diseases, 88 Cincinnati, 178 cis, 316 civilian, 215 classes, 219 classical, 4, 24 clay, 67 cleavage, 235 closure, 370 CLSM, xi, 293, 294, 295, 305, 322, 323, 324 clustering, 274 clusters, 2, 10, 11, 25, 364 C-N, 357

Index CNS, 341, 342, 344 CO2, 195, 205, 351, 352 coagulation, x, 88, 116, 199, 203, 273 coal, 197, 198 coal tar, 197, 198 coatings, vii, 61, 65, 135, 219, 226 cobalt, 21 coefficient of variation, 156 coil, 191, 192, 372 collagen, xii, 75, 123, 128, 136, 203, 241, 333, 340, 347, 348, 349, 351, 353, 355, 356, 361, 364, 372, 373, 374, 376, 377 collateral, 231 collateral damage, 231 colloids, 9 colors, 307 combined effect, 36, 51, 90, 167 combustion, 401 commercialization, 7, 210 commodity, 348 communities, 400 community, 6, 236 compatibility, 88, 89, 113, 302, 365 competition, 188, 189, 285 complementary DNA, x, 225, 231 components, xii, 9, 52, 62, 69, 77, 138, 198, 204, 205, 233, 234, 254, 330, 341, 347, 349, 380 composites, ix, x, xii, xiii, 8, 184, 213, 214, 215, 217, 219, 225, 226, 228, 229, 232, 235, 239, 241, 255, 265, 315, 326, 380, 381, 397, 399, 400, 401, 402, 404, 405, 406, 407, 408, 409, 410, 411, 412, 415, 416, 417, 419, 420, 423, 425 composition, 9, 14, 86, 98, 108, 110, 111, 116, 121, 125, 203, 206, 226, 260, 331, 353, 355, 357, 358, 375 composting, 372 compounds, 180, 216, 306, 323 compressive strength, 220, 381 condensation, 8, 130, 196, 227, 252 conductance, 52, 197 conduction, 30, 51, 52, 197, 228, 262 conductive, xiii, 4, 25, 52, 84, 112, 138, 139, 185, 219, 225, 233, 234, 235, 236, 252, 415, 416, 418, 419, 420, 422, 423, 425 conductivity, ix, x, xiii, 4, 7, 25, 52, 69, 79, 81, 89, 97, 108, 109, 110, 114, 118, 124, 139, 155, 184, 185, 188, 189, 190, 197, 199, 204, 205, 206, 210, 211, 215, 225, 227, 228, 229, 230, 232, 234, 239, 242, 243, 246, 247, 248, 315, 327, 349, 354, 400, 415, 416, 417, 419, 420, 423, 424, 425 conductor, 4 confidence, 219 configuration, x, 33, 112, 239, 240, 242, 249, 279

429

confinement, 2, 4 Congo red, 184 conjugation, 197 connective tissue, 371 consciousness, 332 constant rate, 62 constituent materials, 21 constraints, 3 construction, ix, 4, 5, 184, 298 consumer goods, 219 consumers, 218 consumption, 233, 234 contaminants, 216 continuity, 78 control, viii, x, 1, 7, 9, 11, 16, 20, 31, 79, 117, 153, 154, 155, 203, 204, 220, 227, 239, 240, 241, 243, 251, 252, 253, 254, 257, 260, 265, 321, 327, 332, 335, 339, 352, 364, 368, 370, 371, 373, 380, 386 control group, 332 contusion, 331, 333 conversion, vii, xi, 53, 67, 210, 241, 265, 275, 306, 325 cooling, 16, 52, 216, 382 Copenhagen, 57 copolymer, 86, 192, 275, 277, 279, 284, 285, 300, 302, 313, 380 copolymerization, xi, 275, 277, 279, 280, 285, 302, 304, 305, 311, 326 copolymers, 88, 100, 282, 285, 287, 293, 301, 302, 306, 311, 325, 400 copper, 6, 9, 19, 20, 64, 226, 265, 311 copper oxide, 265 core-shell, 236 correction factors, 36 correlation, 24, 190, 304 corrosion, 205 corrosive, 318 cortex, 337 cortical neurons, 345 corticospinal, 343 cosmetics, 75, 114 cost saving, 379 costs, 7, 204, 218 cotton, 352, 381 Coulomb, 119 coupling, 155, 227, 236, 285, 288, 289, 290, 357, 358, 372, 418 covalent, 1, 6, 52, 119, 127, 139, 227, 233, 311 covalent bond, 1, 52 covering, 190, 219, 371 CPI, 316, 318, 319, 320 crab, 63, 65 crack, 7, 204, 392

430

Index

CRC, 53, 143, 144, 178, 180, 181 creep, 379 critical value, 240, 243, 291 crosslinking, viii, 73, 89, 99, 102, 118, 119, 120, 121, 122, 127, 128, 129, 130, 131, 132, 133, 136, 139, 194, 196 cross-linking, 66, 129, 192 crosslinking reactions, 122 crustaceans, 63, 86 crystalline, ix, 1, 7, 31, 52, 63, 113, 184, 185, 260, 379 crystallinity, 8, 28, 52, 190, 260, 402, 419 crystallites, 98 crystallization, 98, 402 crystals, 191, 204, 366 CSF, 339 cues, xi, 329, 330 cultivation, 375 culture, 331, 345, 351, 352, 365, 366, 368 culture media, 331 CVD, 7, 8, 9, 12, 15, 26, 28, 32, 39, 43, 45, 183, 198, 200, 201, 203, 226, 227 cycles, 228, 257 cyclic voltammetry, 231 cyclohexane, 228 cyclotron, 9 cyst, 339 cytochrome, 230, 231 cytoplasm, 370 cytotoxicity, 126, 127, 330 Czech Republic, 62

D damping, 204, 205, 402, 406 deacetylation degree, 74, 86 decomposition, 7, 9, 12, 13, 14, 15, 19, 31, 120, 201, 215, 226, 385, 396, 410 decomposition temperature, 120, 385, 396, 410 defects, 7, 9, 21, 28, 30, 42, 51, 52, 79, 80, 82, 83, 84, 91, 93, 94, 95, 96, 98, 100, 108, 111, 113, 114, 119, 121, 124, 203, 226, 284, 285, 287, 331 defense, 87, 215 deformation, 105, 235, 381, 392, 394, 406 degradation, 70, 87, 104, 212, 214, 226, 228, 232, 265, 343, 351, 363, 372, 374, 410, 412 degradation rate, 70, 374 degrading, 212 dehiscence, 283 dehydration, 19, 115 dehydrogenase, 230, 234 dehydrogenases, 366 dehydrogenation, 192, 194, 196, 199 delivery, vii, 157, 218

denaturation, 232 density, ix, 23, 25, 26, 28, 29, 30, 32, 36, 38, 39, 41, 42, 43, 44, 45, 46, 49, 50, 51, 81, 97, 127, 184, 185, 204, 206, 207, 226, 228, 231, 242, 252, 284, 286, 291, 380 deposition, 8, 9, 12, 14, 16, 21, 26, 30, 31, 32, 39, 42, 45, 78, 84, 183, 190, 198, 200, 203, 206, 226, 252, 257, 259, 344 derivatives, viii, 73, 75, 76, 86, 87, 88, 100, 102, 107, 108, 113, 117, 123, 125, 126, 127, 129, 139, 302, 308, 321, 375 desorption, 210, 228, 229 detection, 69, 227, 228, 229, 230, 232, 233, 234, 235, 236, 301, 318 deviation, 156, 157 dialysis, 274, 302, 350 diamines, 227 diamond, ix, 1, 2, 3, 5, 26, 38, 58, 184, 185 diamond films, 38 dianhydrides, 327 dielectric constant, 26, 79, 188, 190, 316, 327 dietary, 88 differential scanning, xii, 74, 91, 399, 401, 402 differential scanning calorimeter, 401 differential scanning calorimetry (DSC), xii, 74, 91, 98, 113, 379, 381, 385, 386, 396, 399, 401, 402, 403, 404, 412 differentiation, 88, 375 diffusion, 10, 12, 15, 21, 32, 33, 128, 195, 201, 203, 206, 216, 307, 315, 382 diffusion process, 21 digestion, 375 digital images, 69 diisocyanates, 401, 417 dimethylformamide, 74, 210, 228, 263 dimethylsulfoxide, 229 dimethylsulphoxide, 74, 91 diodes, 52 dipole, 25 discs, 418 diseases, 87, 88 disinfection, 352 dislocations, 52 dispersion, xiii, 99, 206, 210, 212, 215, 375, 382, 383, 399, 400, 415, 416, 420, 422, 423 displacement, 386 dissociation, 14 dissolved oxygen, 233, 234 distilled water, 96, 102, 115, 119, 120, 351 distortions, 6 distribution, ix, xiii, 8, 14, 63, 68, 80, 81, 83, 99, 109, 111, 118, 124, 127, 134, 157, 158, 184, 193, 208, 210, 211, 214, 218, 243, 257, 262, 263, 281,

Index 284, 321, 323, 353, 358, 399, 400, 410, 412, 415, 416 distribution function, 210 disulfide, 63, 236, 350 disulfide bonds, 350 diversity, 62 DMF, 74, 80, 81, 82, 83, 109, 110, 115, 126, 183, 188, 190, 228, 246, 247, 312 DNA, ix, x, 82, 88, 184, 225, 227, 230, 231, 232, 352, 366 donkey, 333 donor, 315, 318, 326 dopamine, 232 doped, 26, 233 dosage, 218 double bonds, 130 double helix, 232 download, 354 dressing material, 123 dressings, 75, 86, 88, 99, 100, 103, 104, 106, 107, 139 drinking, 332 Drosophila, 230 drug carriers, 103 drug delivery, ix, 75, 123, 127, 128, 135, 154, 157, 184, 327, 374 drug delivery systems, 123, 127, 135, 157 drugs, 86, 89, 99, 104, 218 drying, 9, 82, 126, 350, 401 DSC, xii, 74, 91, 98, 113, 379, 381, 385, 386, 396, 399, 401, 402, 403, 404, 412 ductility, xii, 379, 388, 389 duration, 20, 112 dust, 204, 229, 241 Dynamic Mechanical Analysis (DMA), 402 dynamic viscosity, 131, 132

E E. coli, 101, 124, 125 ECM, 344, 348, 369, 370, 372, 377 education, 343, 396, 413, 426 elasticity, 219, 402 elastomers, 380, 401 electric arc, 8 electric charge, 77, 82 electric current, 242 electric field, vii, 14, 22, 24, 25, 26, 28, 29, 30, 35, 36, 44, 51, 53, 77, 90, 153, 155, 157, 167, 168, 174, 187, 249, 251, 285, 286, 287, 349, 380 electric potential, 79, 82 electrical conductivity, ix, xiii, 4, 25, 69, 81, 89, 97, 114, 139, 185, 188, 197, 199, 204, 205, 206, 211,

431

225, 227, 228, 229, 230, 232, 234, 400, 415, 416, 417, 425 electrical fields, x, 225 electrical power, vii, 53 electrical properties, x, 155, 219, 225, 400, 416, 420 electrical resistance, x, 225, 228, 230, 235 electricity, ix, 6, 185, 218 electroanalysis, 238 electrochemical deposition, 21 electrochemical detection, 232, 233 electrochemical impedance, 233 electrodeposition, 257 electrodes, ix, 8, 32, 52, 184, 205, 206, 232, 233, 234, 235, 380 electrolyte, 155 electromagnetic, 204 electron, xii, 2, 6, 9, 22, 23, 25, 26, 28, 30, 33, 43, 45, 52, 65, 157, 197, 226, 228, 234, 235, 306, 307, 311, 315, 316, 318, 325, 326, 347, 349, 351 electron cyclotron resonance, 9 electron microscopy, 10, 52, 75, 288 electrons, 1, 4, 22, 23, 24, 26, 30, 36, 42, 46, 50, 205 electron-transfer, 234 electrophoresis, 291, 301, 380 electroreduction, 233 electrostatic force, 16, 17, 64, 77, 82, 167, 190, 240, 250, 349 electrostatic interactions, 128 ELISA, 352, 367, 376 elongation, 16, 81, 82, 118, 154, 171, 188, 191, 251, 335 embryo, 103 emission, vii, ix, 8, 9, 22, 23, 24, 25, 26, 28, 29, 30, 31, 33, 35, 36, 37, 38, 39, 41, 42, 44, 46, 49, 50, 51, 52, 65, 185, 235, 288, 311, 351 emitters, 25, 30, 52 emulsification, 382 encapsulation, 232 end-users, 219 energy, vii, ix, xi, 1, 2, 20, 22, 30, 50, 52, 61, 66, 184, 185, 203, 205, 206, 207, 208, 210, 212, 241, 306, 311, 315, 325, 355, 357, 381, 402, 406 energy transfer, 306, 315 England, 140 entanglement, 74, 80, 89, 90, 285, 381 entanglements, 79, 80, 81, 90, 91, 115, 159, 166, 167, 171, 188, 303 entropy, 406 environment, vii, 16, 17, 20, 61, 70, 78, 88, 107, 117, 126, 191, 195, 332, 372 environmental impact, 1, 216 environmental issues, 208 enzymatic, 86, 87, 308, 351, 363, 372

432

Index

enzymatic activity, 308 enzyme immobilization, vii, 61, 274, 301, 304, 307, 315, 326 enzymes, ix, x, 87, 104, 184, 225, 230, 232, 234 Epi, 334 epithelium, 371 epoxy, 25, 216, 235, 380, 397 equality, 12 equilibrium, 74, 119, 209, 210, 349 erosion, 21, 361, 363 erythrocyte, 106 erythrocytes, 105 ester, 103, 355, 357, 358 esterification, 122 estimators, 162 Estonia, 62 etching, 9, 21 ethanol, 63, 78, 82, 87, 230, 233, 234, 249, 279, 352 ethanol detection, 234 ethers, 303 ethyl acetate, 117, 228 ethylene, 63, 75, 78, 79, 80, 84, 100, 105, 106, 123, 149, 251, 302, 303 ethylene glycol, 75, 79, 84, 105, 106, 149, 302, 303 ethylene oxide, 75, 78, 251 Euro, 266 evaporation, 9, 81, 82, 83, 99, 113, 154, 159, 188, 189, 191, 192, 208, 240 EVOH, 182 evolution, 196 excitation, 313, 320, 322, 324 exclusion, 163 exoskeleton, 86 experimental condition, 159, 285 experimental design, 155, 160, 162, 174 exposure, 124, 125, 190, 228, 229, 231, 232, 233, 279 extracellular matrix, x, xii, 75, 126, 273, 330, 331, 347, 348, 373 extraction, 138, 257, 350 extrusion, 380, 381, 383, 392, 400, 416

F fabric, 217, 330, 332, 334, 335, 339, 342 factorial, 155, 156, 160, 174 failure, 160, 341, 381, 388, 389, 391, 392, 395 fasciculation, 332, 335 feedback, 16 feeding, 79, 82, 110, 281, 284, 285, 286, 287, 299 feedstock, 12, 15 Fermi level, 23, 26, 50 Fermi-Dirac, 23 fertility, 301

fetal, 351, 364, 368 FGF-2, xi, 329, 331, 332, 334, 335, 336, 337, 338, 339, 340, 341, 342, 344, 345 fiber membranes, 216, 327 fibrils, 392 fibrinogen, 75, 102, 241, 348 fibroblast, xi, xii, 127, 329, 344, 345, 348, 376, 377 fibroblast growth factor, xi, 329, 344, 345 fibroblasts, 99, 103, 127, 349, 364, 365, 366, 368, 371, 372, 376, 377 fibroin, 179, 269 fibronectin, 376 field-emission, 35, 288, 351 filament, xii, 6, 9, 32, 199, 202, 203, 379, 381, 392 filled polymers, 380 filler particles, 243 fillers, ix, xiii, 184, 243, 246, 379, 381, 400, 401, 402, 412, 415, 416, 417, 419, 420, 423, 425 film, vii, 10, 11, 12, 22, 25, 26, 27, 31, 33, 35, 36, 37, 38, 39, 40, 42, 45, 52, 104, 135, 228, 229, 233, 240, 258, 259, 274, 277, 279, 281, 285, 288, 290, 311, 326, 358, 365, 366, 367, 368, 372 film thickness, 11, 26, 40, 311 films, 9, 16, 22, 25, 26, 28, 29, 32, 33, 36, 38, 40, 41, 43, 44, 45, 46, 49, 89, 124, 125, 127, 134, 226, 228, 233, 277, 278, 279, 280, 281, 302, 303, 309, 311, 316, 318, 320, 325, 326, 363, 372, 375, 376 filters, ix, 62, 63, 69, 86, 135, 139, 155, 185, 214, 219, 241 filtration, vii, viii, 61, 62, 63, 66, 70, 86, 154, 155, 184, 203, 217, 219, 241, 299 Finland, 62 fire, 204 FITC, 293 fixation, 126, 352 flat panel displays, 52 flexibility, 154, 198, 306, 323 flight, 78, 82, 167, 174, 203, 249, 250, 251 floating, 25, 200, 226 flow, vii, viii, x, 8, 15, 17, 32, 45, 61, 62, 67, 68, 69, 91, 153, 154, 156, 158, 159, 160, 162, 167, 168, 169, 170, 171, 173, 174, 175, 188, 199, 234, 236, 239, 242, 243, 249, 251, 252, 339, 350, 381, 401, 402, 418 flow rate, viii, x, 8, 67, 153, 154, 156, 158, 159, 160, 162, 167, 168, 169, 170, 171, 173, 174, 175, 188, 239, 242, 249, 251, 350 fluctuations, 17 fluid, 12, 187, 236, 281, 284, 339, 349 fluorescence, xi, 75, 231, 294, 296, 305, 306, 311, 313, 316, 318, 320, 323, 326, 333, 334, 335, 336, 337, 339, 340, 341, 342 fluoride, 179, 241, 267

Index fluorinated, 101 fluorine, 69 food, viii, 73, 87, 89, 114 food industry, viii, 73, 87, 89 forests, 15, 235 formaldehyde, 352 fossil fuel, 208 fouling, 63, 234, 302 Fourier transform infrared spectroscopy, xii, 65, 288, 347 fractionation, 375 fracture, 6, 235, 381, 392 fragmentation, 138 France, 54, 427 free energy, 189 free radical, 88, 275 free radicals, 88 freedom, 83 fruits, 102 FTIR, xii, 65, 254, 347, 349, 351, 355, 357, 358, 372, 373, 376 FT-IR, 194, 195, 288, 289, 351 FTIR spectroscopy, 349 fuel, vii, ix, 52, 61, 70, 75, 184, 203, 204, 205, 208, 210, 215 fuel cell, vii, ix, 52, 61, 70, 75, 184, 203, 204, 205, 210 fullerene, 2, 7, 326 fullerenes, ix, 2, 3, 4, 7, 53, 86, 184, 234, 413, 427 functionalization, x, xi, 225, 227, 284, 305 fungal, 63 fungi, 63, 86, 88 fungus, 124 Fusarium, 86

G G-6, 103 gallium, 52 gamma, 64, 157 ganglion, 337, 344, 345 gas, ix, 8, 10, 12, 14, 15, 17, 20, 26, 28, 31, 32, 33, 39, 42, 45, 52, 86, 184, 185, 193, 199, 201, 202, 207, 208, 210, 215, 216, 226, 227, 228, 229, 230, 235, 236, 241, 252, 381, 385, 390 gas phase, 14, 202 gas sensors, 227, 229, 235, 236 gas separation, 208, 241 gas turbine, 241 gases, 8, 20, 126, 195, 208, 216, 226, 228, 229 gauge, 52, 64, 67 GCE, 230, 233, 234, 235 gel, 131, 132, 133, 358 gel formation, 131, 132, 133

433

gelatin, xii, 102, 128, 234, 304, 347, 348, 349, 350, 351, 353, 354, 357, 358, 359, 361, 362, 365, 366, 367, 372, 373, 374, 375, 376 gelation, 108, 241 gene, vii, 333, 344, 345 gene therapy, 344, 345 generalization, 175 generation, x, 75, 86, 99, 103, 107, 139, 194, 225, 254, 255, 260, 330 generators, 252, 253, 254, 260, 262, 265 Germany, 62, 399, 401, 415, 417 glass, 213, 228, 229, 313, 333, 341, 350, 402, 406, 412 glass transition, 402, 406, 412 glass transition temperature, 402, 406, 412 glasses, 26 glia, 331 glial, xi, 329, 330, 333, 341, 342, 343, 344 Glial, 341, 344 glial fibrillary acidic protein (GFAP), 333, 341, 342 glial scar, xi, 329, 330, 342, 344 glow discharge, 16, 226 glucose, x, 230, 231, 232, 233, 273, 275, 279, 284, 288, 290, 291, 292, 293, 294, 298, 300, 301, 303, 315, 326 glucose oxidase, 230, 231, 233, 301, 315, 326 glucoside, 275, 291, 302, 304 glutamate, 321, 328 glutaraldehyde, 66, 74, 89, 350 glycine, 350 glycoconjugates, x, 273, 274 glycol, 75, 118, 142 glycolipids, x, 273 glycopolymers, x, 273, 274, 275, 276, 281, 299, 301, 302, 304 glycoprotein, 293 glycoproteins, x, 273, 291 glycoside, x, 273, 274, 298, 300, 302 glycosylated, x, 273, 274, 275, 282, 283, 284, 288, 290, 291, 293, 294, 296, 297, 298, 299 glycosylation, 274, 285, 288, 290 gold, 157, 231, 235, 236, 333, 351, 352 grafting, 115, 117, 331, 370 grain, 20, 21, 26 grain boundaries, 21, 26 gram-negative, 88, 101, 124, 125 gram-positive, 88, 101, 124, 125 granules, 89 graph, 28, 36, 41, 44, 47, 48 graphene sheet, ix, 2, 4, 6, 225, 226 graphite, ix, 1, 2, 3, 4, 5, 6, 8, 10, 28, 30, 184, 185, 196, 198, 217, 218, 230, 231, 232, 260 gravity, 17

434

Index

grounding, 157 growth factor, xi, 329, 331, 337, 338, 342, 344, 345 growth factors, 342, 344 growth mechanism, 11, 12, 13, 18, 19, 21, 54, 200, 201, 202 growth modes, 15 growth rate, 10, 11, 14, 15 growth temperature, 12, 14, 28, 29, 226 guanine, 232 guidance, 335, 339, 345 guidelines, 332

H H2, 352 half-life, 331 handling, 204 hardening, 394, 395, 396 healing, 76, 88, 123, 139, 352, 370, 371 health, 216, 332 health problems, 332 heat, viii, ix, 5, 32, 73, 120, 121, 122, 185, 195, 196, 208, 218, 234, 257, 263, 264, 265, 381, 385, 400, 401, 402, 406, 416, 417, 423, 425 heat conductivity, 423 heating, 9, 12, 121, 122, 123, 128, 192, 194, 195, 260, 316, 332, 383, 385, 401 heating rate, 194, 195, 260, 385, 401 heavy metal, vii, 61, 69 heavy metals, vii, 61 height, 21, 23, 25, 30, 35, 51, 67, 138 Heisenberg, 23 helicity, 4 helium, 8 helix, 232, 372 heme, 311, 326 hemisphere, 2, 25 hemocompatibility, 303 hemoglobin, 230, 235, 311 Heparin, 344 hepatocyte, 301 heterogeneous, 1, 22, 356 heterogeneous catalysis, 1 hexafluoropropylene, 241 hexane, 228 HFP, 268 high pressure, 2 high temperature, 4, 8, 15, 16, 50, 199, 201, 204, 208, 212, 214, 218, 316, 318, 406 high-speed, 78 hippocampal, 230, 231 histidine, 327 histochemical, 344 histological, 371

Holland, 222 HOMO, 307 homogeneity, 212 homogenous, 80, 136, 257, 262, 322, 323 homopolymers, 80 host, 4, 291 HRP, 230, 233 HRTEM, 43, 45 human, 103, 104, 127, 180, 343, 344, 369, 372, 376, 377 humidity, 79, 83, 139, 154, 158, 188, 191, 252, 282, 286, 287, 349 hybrid, 2, 16, 86, 89, 98, 101, 104, 109, 110, 119, 120, 125, 126, 132, 133, 139, 206, 207, 229, 301, 374, 375 hybridization, 1, 2, 205, 228, 231 hybrids, 302 hydration, 332 hydrazine, 215 hydrides, 207 hydro, x, 6, 9, 31, 123, 199, 201, 273, 279, 300, 380 hydrocarbons, 6, 9, 31, 197, 199, 201, 202, 226, 380 hydrogels, 89, 128, 129, 132, 134, 157, 340, 344 hydrogen, ix, 9, 81, 90, 91, 97, 108, 184, 199, 203, 207, 208, 210, 215, 226, 228, 229, 230, 233, 234, 235, 265, 309 hydrogen atoms, 226 hydrogen bonds, 90, 91, 97 hydrogen peroxide, 233, 234, 235, 309 hydrogenation, 197 hydrolysis, 86, 87, 95, 96, 102, 113, 130, 157 hydrophilic, x, 123, 273, 279, 300 hydrophilicity, 100, 104, 106, 117, 277, 279, 285, 294, 300, 358, 374 hydrophobic, 81, 104, 138, 315, 358, 368 hydrophobic interactions, 81, 138 hydroxide, 66, 208 hydroxyapatite, 96, 98, 375 hydroxyl, viii, 73, 122, 196, 233, 285, 288 hydroxyl groups, 122, 196, 285, 288 hydroxypropyl, 82 hydroxypropyl cellulose, 82 hypothesis, 162, 163, 373 hypoxic, 344

I ibuprofen, 127 IDA, 229 identification, 219 images, 6, 10, 11, 40, 69, 206, 209, 212, 213, 231, 241, 243, 246, 247, 255, 257, 258, 259, 260, 261, 263, 282, 287, 290, 292, 294, 295, 299, 307, 312,

Index 313, 314, 322, 324, 351, 353, 357, 359, 361, 363, 364, 366, 368, 372 imaging, 12, 303, 376 imidization, 316, 318 immersion, 97, 102, 104, 122, 125, 138, 290 immobilization, ix, 184, 212, 233, 287, 288, 290, 291, 300, 301, 302, 303, 304, 307, 325, 326 immune response, x, 273 immune system, 291 immunohistochemical, 338, 339, 369 immunoreactivity, 344 impedance spectroscopy, 233 implants, 98, 104, 204, 330, 341 imprinting, xi, 301, 305, 306, 325 impurities, 21, 227 in situ, 12, 99, 161, 327 in vitro, 127, 330, 336, 341, 343, 344, 345, 351, 363, 373, 376 in vivo, 69, 330, 344, 345 inclusion, xiii, 163, 331, 343, 399, 406 incubation, xi, 99, 274, 351, 361, 362, 365, 366 incubation time, 351, 361, 362, 366 India, 1, 58 Indian, 223 indication, 77, 102, 106, 131, 133 indices, 4, 160, 214 indium, 26 indium tin oxide (ITO), 26 industrial, 52, 75, 77, 86, 139, 157, 186, 215, 227, 240, 325, 400, 416 industrial application, 157, 215 industrial production, 86 industry, viii, 7, 73, 88, 89, 114, 215, 216, 219, 318 inelastic, 393 inert, 8, 195 inertness, 204, 205, 226 infection, x, 273, 332 infectious, 291 infiltration, 340 infinite, 4, 5 inflammatory, 291, 330, 371 inflammatory cells, 330, 371 infrared, xii, 65, 288, 347, 356, 358, 360 infrared spectroscopy, xii, 65, 288, 347 infrastructure, 219 inhibition, 100, 106 inhibitory, 344 inhomogeneity, 26 initiation, 64, 188, 249, 252, 321 injection, xii, xiii, 234, 331, 352, 399, 400, 401, 402, 412, 415, 416, 417, 418, 425 injuries, 330, 333, 341, 343

435

injury, 329, 330, 331, 332, 334, 335, 338, 339, 340, 341, 342, 343, 344, 345 inorganic, 79, 81, 86, 98, 118, 134, 226, 254, 379 inorganic filler, 380 insects, 86 insertion, 207 insight, 139 inspection, 336, 339 instabilities, 192 instability, 16, 64, 188, 189, 190, 191, 252, 274, 283, 284, 286, 303, 304 instruction, 352 instruments, 204 insulation, vii, 241, 417 integration, vii, 53 integrity, 25, 88, 97, 104, 119, 206, 330, 400, 416 interface, 12, 16, 17, 22, 52, 203, 215, 420, 423 intermediaries, 197 intermolecular, 81, 91, 194, 196, 254 intermolecular interactions, 81, 91, 254 interstitial, 219, 220, 337 interval, 352 intervention, 338 intramuscular, 352 intramuscular injection, 352 intraperitoneal, 332 intrinsic, x, 63, 66, 80, 158, 159, 186, 203, 239, 242, 321, 322, 402 intrinsic viscosity, 80, 158, 159, 321 invasive, 203 inventions, 217 iodine, 82, 123 ion beam, 9 ion bombardment, 20, 30 ionic, 134, 155 ionization, 52, 228 ionomer, 182 ions, vii, 16, 61, 66, 88, 99, 190, 246, 247, 309, 318 IOP, 85, 123 IR spectra, 195 Iran, 141, 153 iridium, 215, 217 iron, 7, 30, 69, 199, 200, 203, 208, 246, 257 irradiation, 20, 118, 119, 128, 382 IR-spectroscopy, 91, 98 ischaemia, 344 Islam, 53 isoelectric point, 296 isomerization, 194 isotherms, 209 isotropic, 197 Italy, 144 ITEP, 57

436

Index

J Japan, 54, 62, 95, 141, 347, 351, 352 Jun, 141, 300, 325 Jung, 58, 148, 149, 221, 223, 266, 302, 347, 397, 414, 426

K Kentucky, 142 keratin, xii, 347, 348, 349, 350, 351, 352, 353, 354, 355, 358, 360, 363, 368, 370, 371, 372, 373, 375 keratinocytes, 127 ketamine, 332, 333 kinetic parameters, 315 kinetics, 210, 229, 254, 255 KOH, 206, 208 Korea, 1, 96, 183, 347, 349, 373 Korean, 70, 142

L LAB, 351 labeling, 338, 352 lactic acid, 74, 75, 76, 81, 82, 89, 100, 115, 116, 241, 345, 348, 374, 376 lamina, 344 laminectomy, 332 Langmuir, 55, 56, 237, 297, 298, 300, 301, 302, 325, 326 Langmuir-Blodgett, 325 language, 6, 154 large-scale, vii, 53 laser, xi, 8, 293, 305, 316, 323 laser ablation, 8 latex, 63 lattice, 4, 43 leaching, 260 lectin, 301, 302, 304 leukocytes, 291 liberation, 196 lifecycle, 62 light scattering, 214 limitation, 214, 315, 321, 342 linear, 21, 24, 79, 80, 86, 91, 161, 162, 196, 232, 233, 234, 235, 242, 275, 279, 298, 300, 306, 340, 345, 393, 394 linear regression, 162 lines of force, 37, 38 linkage, 16 links, 219 lipase, 87, 302, 303, 304, 326 lipophilic, 99 lipopolysaccharides, 88

liquefaction, 203 liquid nitrogen, 33 liquid phase, 81 liquid water, 254, 262 liquids, 381 lithium, ix, 184, 204, 205, 206, 219 lithium, 180, 183, 266 L-lactide, viii, 74, 75, 79, 84, 100, 105, 106, 117, 127, 142, 241, 256 LMW, 74, 96, 117, 129 loading, xii, 99, 204, 210, 214, 315, 381, 399, 400, 401, 406, 416, 417, 423 location, 226 locomotion, 333, 339, 343 London, 53, 54, 57, 71, 145, 150, 178, 221 long distance, 79, 107, 168 low molecular weight, 283 low temperatures, 9, 28, 208 low-temperature, 16 lubrication, 205 luminescence, 318, 323 LUMO, 307 lying, 36, 42, 46, 373 lysine, 230, 232 lysis, 88 lysozyme, 87, 104, 113

M machinery, 205 machines, 240, 400 macromolecules, 74, 80, 91, 171, 236 macrophage, 277, 278 magnetic, 2, 208, 300, 380, 381 magnetic field, 380 magnetron sputtering, 39 Malaysia, 413, 426 malic, 64 management, 204, 214, 219 manipulation, 8, 117 manufacturer, 352, 417 manufacturing, 385 market, 219 marketing, 219 markets, 220 mass spectrometry, 8 mass transfer, 195, 381 mass transfer process, 381 matrices, 178 matrix, x, xii, xiii, 26, 75, 99, 126, 161, 162, 194, 203, 212, 214, 215, 216, 218, 226, 229, 235, 263, 273, 288, 311, 312, 315, 330, 331, 344, 347, 348, 358, 373, 381, 385, 389, 392, 395, 396, 397, 399,

Index 400, 401, 402, 406, 410, 412, 415, 416, 417, 419, 420, 423, 425 measurement, x, 13, 52, 273, 279, 280, 281, 296, 300, 309, 311, 352 mechanical behavior, 381, 401 mechanical performances, 389 mechanical properties, xii, 6, 7, 52, 100, 104, 115, 117, 154, 214, 215, 219, 316, 353, 379, 380, 397, 400, 401, 410, 416 media, x, 1, 69, 216, 225, 227, 228, 229, 230, 232, 234, 235, 236, 331 median, 334, 343 mediators, 234, 307 medicine, vii, viii, 61, 73, 86, 87, 89, 107, 232, 241, 330 melt, 13, 74, 76, 77, 80, 153, 187, 190, 198, 199, 349, 381, 383, 396, 400, 416 melting, xii, xiii, 12, 192, 203, 379, 384, 385, 399, 402, 412 melting temperature, xii, xiii, 12, 379, 385, 399 membrane permeability, 67 membranes, vii, xi, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 89, 103, 123, 139, 216, 219, 300, 301, 302, 303, 304, 305, 323, 326, 377 mercaptans, 375 Merck, 157 mercury, 318 mesenchymal stem cell, 180 mesophase, 197 metabolic, 349, 372, 376 metabolism, 306 metal ions, vii, 61, 88, 309, 318 metal nanoparticles, 134, 260 metal oxides, 228, 260, 263 metal recovery, 69 metalloporphyrins, 312, 321 metals, vii, x, 1, 7, 8, 9, 55, 61, 201, 226, 227, 239, 265, 312, 325 methacrylic acid, 89, 192 methane, 9, 26, 28, 199, 226, 228 methanol, 103, 229, 230, 279, 303 methyl methacrylate, 128, 269 methylene, 155, 210, 212, 214 metric, 8, 219 micelles, 99 microbes, 63 microelectrode, 236 micrometer, ix, 7, 33, 43, 45, 186, 225, 226, 271, 274 microorganism, 88, 104, 106, 124 microorganisms, 87, 88, 104, 216, 291 microparticles, 78, 235, 323 microscope, 6, 20, 52, 65, 157, 313, 316, 323, 333, 351

437

microscopy, xi, xii, 10, 52, 75, 219, 288, 293, 305, 323, 334, 335, 336, 337, 339, 340, 341, 342, 347, 349 microspheres, xi, 305, 312, 313, 336, 345 microstructure, ix, 7, 25, 113, 184, 196, 234 microstructures, 234, 420, 425 microwave, 9, 26, 30 microwave heating, 9 migration, 349, 373 military, 75 mimicking, x, 75, 126, 273, 372 mineralization, 232, 374 mines, 227 miniaturization, 6 Ministry of Education, 273, 373 mirror, 4, 204 missions, 227 mitochondrial, 366 mixing, xii, xiii, 2, 99, 128, 129, 134, 137, 229, 232, 353, 381, 382, 384, 399, 400, 401, 412, 415, 416, 417, 425 MMW, 66 mobility, 6, 246, 247, 332, 406, 412 modeling, 52, 156, 161 models, 4, 25, 79, 154, 156, 157, 160, 161, 162, 163, 165, 166, 174, 214, 274, 340, 418, 419 modulus, xii, 5, 7, 116, 204, 213, 220, 225, 260, 379, 380, 381, 388, 389, 393, 394, 395, 396, 399, 402, 404, 405, 406, 407, 408, 410, 412 moieties, 88, 274, 277, 301, 315, 316, 325, 327 moisture, 196, 216, 401 molar ratio, 127, 132, 136, 137, 350 mole, 277 molecular orientation, 254 molecular structure, 197 molecular weight, x, xi, 65, 66, 74, 80, 95, 97, 154, 157, 158, 159, 188, 198, 210, 239, 242, 275, 281, 283, 304, 305, 321, 353 molecular weight distribution, 281, 353 molecules, 2, 12, 14, 15, 80, 128, 192, 198, 208, 217, 226, 227, 229, 231, 232, 236, 246, 323, 331, 343, 344, 406, 416 molybdenum, 9, 32 momentum, 23, 156, 167, 175 monoclonal, 333 monoclonal antibody, 333 monolayer, 365, 366 monolayers, 301 monomer, 128, 275, 277, 279, 280 monomers, 89, 277, 312, 323, 330 monosaccharides, 301 morphogenesis, 345

438

Index

morphological, 19, 188, 197, 198, 199, 288, 353, 361, 363 Moscow, 57, 144 motion, x, 3, 203, 239, 242, 249 motor function, 331 motor neurons, 337, 345 mouse, 99, 103, 127, 333 movement, 83, 373, 406 multilayer films, 134 multiwalled carbon nanotubes, 11, 43, 45 municipal solid waste, 372 muscle, 377 muscle tissue, 377 muscles, ix, 185, 218 myoglobin, 311, 326

N N,N-Dimethylformamide, 183, 188 N-acety, 63, 291, 296 NaCl, 81, 333 NADH, 230, 232, 234, 235 nafion, 230, 233, 234, 235, 254, 255, 326 nanocatalyst, 19 nanocomposites, 61, 178, 203, 214, 233, 235, 266, 325, 326, 380, 381, 388, 389, 396, 399, 400, 402, 403, 404, 406, 410, 412, 415, 416, 417, 420, 421, 422, 423, 424, 425 nanocrystalline, 26, 36, 58 nanocrystals, 4, 20 nanoelectronics, xi, 52, 305 nanofibrous membranes, vii, xi, 61, 65, 69, 70, 300, 303, 305, 307, 308 nanohorns, ix, 184 nanomaterials, ix, 51, 86, 184, 227, 234 nanometer, ix, 2, 6, 13, 25, 153, 186, 206, 225, 274, 348 nanometer scale, ix, 153, 225 nanometers, vii, 78, 224, 240, 252, 274, 306, 323, 372, 400 nanoparticles, 12, 15, 17, 19, 20, 86, 90, 96, 98, 110, 123, 132, 133, 134, 201, 210, 213, 226, 232, 236, 243, 246, 254, 255, 257, 258, 260, 261, 262, 265, 300, 327, 381, 395, 396, 400 nanoribbons, 83 nanorods, 36 nanoscale structures, 372 nanostructured materials, x, 273 nanostructures, vii, ix, 2, 10, 26, 31, 53, 58, 184, 205, 225, 325, 413, 427 nanotechnologies, 219, 240 nanotechnology, viii, 184 nanotube, 2, 3, 4, 5, 6, 9, 12, 13, 25, 26, 39, 45, 52, 183, 206, 218, 227, 228, 231, 234, 235, 263, 326

nanotube films, 228 nanotubes, ix, xii, xiii, 2, 3, 4, 5, 6, 8, 12, 14, 16, 17, 18, 21, 24, 25, 26, 30, 31, 33, 43, 45, 51, 52, 53, 54, 86, 134, 184, 205, 225, 227, 234, 257, 303, 304, 312, 326, 380, 397, 399, 400, 401, 415, 416, 417, 420, 423, 425 nanowires, 37, 186, 219, 235, 236 naphthalene, 197, 198, 199 National Science Foundation, 343 natural, viii, x, xii, 63, 73, 75, 79, 86, 87, 88, 89, 91, 100, 102, 126, 128, 129, 139, 160, 161, 165, 186, 199, 241, 273, 274, 347, 348, 372 natural gas, 199 natural polymers, 75, 79, 102, 128, 139, 241 NCA, 321, 323 NEA, 25 needles, 99, 255 neodymium, 8 nerve, 88, 341, 342, 344, 345, 373 nerve cells, 88 nerve gaps, 344 nerves, 342 nesting, 332 Netherlands, 427 network, xiii, 79, 107, 127, 171, 215, 243, 368, 377, 415, 416, 420, 423, 425 neural stem cells, 345, 374 neural tissue, 345 neurofilament, 333, 334, 335, 336, 337, 339 neuronal plasticity, 337 neuronal survival, 331 neurons, 337, 341, 343 neuroprotective, 338 neurotrophic, 344 New Jersey, 143, 181, 343, 345 New York, 54, 56, 57, 59, 144, 145, 147, 149, 178, 179, 180, 181, 222, 223, 375, 397 next generation, 220 NHS, 288 nickel (Ni), 8, 9, 10, 11, 12, 16, 28, 29, 30, 31, 33, 39, 40, 42, 43, 45, 145, 226, 265, 271, 312, 321 nicotinamide, 234, 235 Nielsen, 56, 418, 427 NIH, xii, 103, 347, 351, 352, 364, 365, 366, 367, 368, 369, 372, 373 nitrate, 9, 19, 98, 265 nitridation, 30 nitrogen, vii, 14, 30, 33, 61, 193, 195, 196, 212, 215, 216, 260, 262, 263, 355, 357, 360, 385, 401 nitrogen gas, 193, 385 nitrogen oxides, 216 NMR, 275, 276, 279, 282, 327 N-N, 210

439

Index non-invasive, 203 nonionic, 115 non-linear optics, 306 non-metals, 7 nontoxic, 127 non-uniform, 134, 171, 174, 175, 251, 252 non-uniformity, 252 normal distribution, 193 North Carolina, 239 nose, 218 n-type, 9, 26 nuclear, 204 nucleation, 10, 199, 203 nuclei, 381 nucleophiles, 321 nucleophilicity, 321 null hypothesis, 163 nutrient, 349, 372 nutrients, 126, 340 nutrition, 88 nylon, 62, 244, 380, 381, 396

O observations, 21, 82, 154, 159, 161, 162, 163, 166, 171, 263, 313 OCs, 216 Ohio, 222 oil, ix, 63, 65, 157, 184, 228, 230, 232, 252, 256, 331 oligomers, x, 87, 88, 97, 117, 128, 130, 225, 227, 230, 231, 327 oligosaccharides, 291 one dimension, 13, 33 open field test, 343 open-field, 333 optical, 2, 4, 75, 192, 204, 323, 351 optical chemical sensors, 75 optical properties, 2 optics, 306, 337 optoelectronics, 86, 139 orbit, 197, 215 organ, 348 organic, viii, ix, 73, 79, 81, 87, 89, 107, 113, 114, 127, 138, 184, 185, 206, 210, 216, 228, 229, 240, 290, 325, 348, 349 organic compounds, 184, 216 organic matter, ix, 185 organic polymers, 240 organic solvent, viii, 73, 87, 107, 113, 114, 127, 138, 290, 348, 349 organic solvents, viii, 73, 87, 107, 113, 114, 127 orientation, 10, 16, 26, 28, 91, 155, 199, 226, 330, 335, 368, 418 orthorhombic, 63

oscillation, 311 osteoblasts, 103 osteosarcoma, 103 oxidation, 212, 217, 227, 231, 232, 233, 234, 235, 349, 350 oxidative, 192, 193, 194, 212, 301 oxidative damage, 301 oxidative reaction, 192 oxide, 19, 26, 31, 42, 75, 78, 142, 220, 241, 249, 251, 263, 264, 265, 326 oxides, 7, 257, 263 oxygen, ix, 184, 192, 193, 196, 227, 233, 234, 355, 357 oxygen consumption, 233, 234 ozone, 216

P PAA, 74, 89, 99, 113, 135, 136, 210, 229, 316 packaging, vii, 52, 61, 88, 219 pain, 332 palladium, 16 pancreas, 157 paralysis, 333 paramagnetic, 311 parameter, 82, 83, 155, 156, 158, 159, 160, 189, 251, 418 Paris, 54, 427 particles, xii, 8, 12, 20, 25, 33, 34, 63, 199, 203, 210, 212, 215, 216, 217, 226, 227, 243, 257, 258, 259, 260, 263, 323, 379, 381, 382, 383, 399, 402, 406, 410, 416, 419, 420 patents, 77, 157 pathogenic, 88, 100, 104, 106, 126 pathways, 345 patterning, 21 PBA, 229 PBI, 184, 197 PBT, 155, 180 PEMFC, 212 penicillin, 351 Pennsylvania, 69 peptide, 235, 331, 333, 344, 345, 374 Peptide, 181 peptides, 343 periodic, 1 periodic table, 1, 325 periodicity, 3 Peripheral, 268 peripheral nerve, 341, 342, 344 peripheral nervous system, 331 permeability, vii, 61, 64, 67, 68, 69, 70, 86, 204 permit, xi, 161, 232, 329 permittivity, 67

440

Index

peroxide, 233, 234, 235 perturbations, 4 petroleum, 197, 198 PGA, 348 pH, 129, 132, 133, 134, 135, 136, 137, 295, 296, 301, 350, 351 pH values, 129 pharmaceutical, 114, 218, 321 pharmaceutical companies, 218 pharmaceutical industry, 114 phase diagram, 201 phase inversion, 134 PHB, xii, 347, 349, 361, 362, 363, 372 phenol, 123 phenotype, 344 Philadelphia, 61, 69 phosphate, 74, 102, 125, 333, 351 photocatalysts, xi, 305 photocatalytic, 212 photographs, 282, 286, 290, 291, 293, 353 photoinduced electron transfer, 326 photoluminescence, 271 photovoltaic cells, 220 physical properties, vii, xii, 61, 157, 236, 380, 399 physicochemical, 89, 274, 287, 381 physico-chemical characteristics, 63, 102 physicochemical properties, 287 physico-chemical properties, 88 physiological, 136 piezoelectric, 75 pitch, 7, 33, 197, 199, 204, 220, 260, 380 planar, 1, 2, 231 plants, 87, 205, 291 plasma, 8, 9, 14, 15, 16, 17, 26, 30, 31, 32, 226, 227, 284 plastic, 52, 219, 228, 332, 341, 400, 416 plasticity, 337 plasticizer, 91 plastics, 204, 219 platelet, ix, 225, 226, 234, 279, 281, 285, 300 play, x, xi, 26, 28, 40, 81, 191, 273, 285, 293, 305, 306, 311 PLGA, 75, 97, 103, 127, 348 PLLA, 74, 75, 79, 81, 82, 84, 104, 105, 106, 117, 373 PMDA, 316, 327 PNA, x, 273, 292, 293, 294, 300 poisoning, 203 polar groups, 354 polarity, 187, 229, 284, 303, 353 polarizability, 205 polarization, 212 pollen, 63

pollutant, 216 pollutants, 216 pollution, 318 poly(2-hydroxyethyl methacrylate), 75, 128 poly(amic acid), 210, 316, 327 poly(ethylene terephthalate)(PET), 63, 100, 103, 104, 128, 268 poly(L-lactide), viii, 74, 75, 79, 84, 105, 106, 117, 256 poly(vinyl chloride) (PVC), 67, 81 poly(vinylpyrrolidone), 63 polyacrylamide, 63, 74, 82, 304 polyamide, 241, 251, 303, 330, 331, 333, 335, 341, 342, 344 polyamine, 331 polyaniline, 254, 255 polycarbonate, 241, 380 polycrystalline, 264 polyelectrolytes, 120, 129, 134, 135, 138, 139 polyester, viii, 74, 104, 217, 242, 348 polyesters, viii, 74, 76, 90, 100, 104 polyethylene, 241, 249 polyethylene terephthalate, 241 polyimide, xi, 197, 208, 210, 241, 260, 305, 316, 317, 318, 325, 327 polyimide film, 325 polymer blends, 254 polymer chains, 81, 83, 90, 107, 113, 196, 229, 254, 262, 315, 406, 410 polymer composites, xiii, 228, 229, 400, 401, 415, 416, 417 polymer film, 233 polymer materials, 75, 88, 139, 381 polymer matrix, xii, xiii, 99, 218, 263, 399, 400, 410, 415, 416, 417, 420, 423, 425 polymer molecule, 80 polymer nanocomposites, 400, 412, 416 polymer networks, 87 polymer solutions, 81, 135, 246, 255, 260, 303, 350 polymer-based, 7, 380 polymeric materials, 348 polymerization, vii, 89, 117, 128, 157, 233, 275, 279, 284, 285, 312, 321, 326, 400 polymers, x, xi, 62, 63, 70, 75, 79, 80, 81, 82, 87, 89, 90, 91, 98, 99, 100, 102, 104, 107, 110, 113, 120, 122, 123, 128, 138, 139, 141, 186, 193, 197, 227, 239, 240, 241, 254, 257, 260, 265, 274, 303, 305, 306, 312, 315, 316, 321, 325, 331, 348, 374, 380, 401, 416, 423, 425 polynomial, 155, 156 polynomials, 161 polypeptide, xi, 305 polypeptides, 327

441

Index polypropylene, xii, 75, 112, 379, 380, 381, 382, 383, 384, 385, 386, 387, 388, 389, 390, 392, 393, 395, 396, 397 polysaccharide, vii, 61, 63, 86, 88, 89, 99, 294 polysaccharides, x, 144, 241, 273, 301 polystyrene, 63, 81, 82, 83, 135, 228, 229, 241, 302, 336 polystyrene latex, 63 polyurethane, xii, xiii, 123, 241, 251, 301, 399, 400, 401, 402, 403, 404, 405, 406, 407, 408, 409, 410, 411, 412, 415, 416, 417, 420, 421, 422, 423, 424, 425 polyurethanes, xiii, 399, 400, 410, 415, 416, 420, 425 polyvinyl acetate, 157 polyvinyl alcohol, viii, 153, 156, 241, 264 polyvinyl chloride, 67, 241 polyvinylpyrrolidone, 254, 256 pore, x, xi, 21, 69, 75, 116, 134, 154, 205, 208, 210, 211, 239, 241, 252, 253, 254, 255, 260, 262, 263, 265, 305, 349, 373, 377 pores, vii, 21, 61, 82, 123, 195, 203, 210, 252, 253, 254, 255, 260, 261, 262, 265, 334, 373 porosity, xi, 1, 70, 82, 116, 126, 134, 154, 203, 205, 208, 219, 220, 252, 253, 254, 255, 263, 274, 305, 307, 315, 323, 340, 372 porous, ix, 83, 84, 98, 105, 134, 184, 186, 205, 208, 232, 233, 240, 252, 253, 254, 255, 256, 257, 260, 261, 262, 263, 264, 339, 348, 373 porphyrins, xi, 305, 307, 308, 311, 312, 323, 325, 326, 327 positive feedback, 16 postoperative, 333, 352 potassium, 208 powder, 8, 157, 228, 230, 231, 232, 381, 382, 384, 401, 417 powders, 205, 226, 381, 382 power, vii, 14, 39, 45, 53, 77, 157, 165, 187, 205, 206, 246, 349, 350, 351 precipitation, 10, 12, 16, 17, 201, 202, 203, 275, 277, 279, 280, 285, 304 precursor cells, 341 prediction, 156, 165, 166, 174 press, 144, 223, 375 pressure, ix, 2, 8, 15, 31, 32, 33, 39, 42, 43, 45, 52, 63, 81, 154, 158, 185, 192, 208, 218, 229, 252, 339, 349, 381 primary tumor, 344 pristine, x, 105, 106, 206, 225, 230, 232, 406, 412, 417 probability, 23 probable cause, 339 probe, 20, 37, 197, 381

process control, 154 processing variables, viii, 153, 174, 190, 301, 328 production, 6, 9, 31, 52, 62, 86, 107, 122, 157, 171, 174, 186, 197, 198, 200, 203, 204, 219, 240, 344, 349, 369, 370, 377, 380 production technology, 204 productivity, 85, 114, 216 program, 343 proliferation, xii, 88, 98, 103, 104, 126, 127, 155, 180, 341, 348, 352, 364, 366, 367, 368, 369, 371, 372, 373, 376 propagation, 7, 64, 349 property, 9, 22, 26, 28, 46, 51, 53, 99, 190, 294, 302, 311, 325, 330, 339, 342, 358, 380 propionic acid, 89, 100 propulsion, 215 propylene, 326 prostheses, 204 prosthetics, 330 proteases, 87 protection, 21, 219, 316, 350 protective clothing, vii, viii, 61, 69, 75, 154, 184, 204, 219 protein, x, xii, 126, 157, 231, 273, 274, 279, 291, 293, 294, 295, 296, 297, 299, 300, 302, 303, 304, 311, 315, 333, 340, 347, 372, 373, 375, 377 protein binding, 274 proteins, ix, x, 88, 102, 104, 130, 184, 225, 227, 230, 231, 241, 273, 274, 275, 296, 300, 311, 326, 327, 348, 368 protocols, x, 273, 332 prototype, 343 pseudomonas, 351, 361 public, 219 pulse, 232, 382 pulses, 381 purification, xi, 8, 63, 88, 204, 212, 241, 274, 275, 296, 299, 300 PVA, viii, 75, 82, 91, 95, 96, 97, 98, 99, 103, 104, 107, 109, 110, 111, 112, 113, 118, 119, 120, 121, 122, 123, 124, 125, 127, 153, 156, 157, 158, 159, 174, 180, 181, 182, 230, 234, 264, 267 PVP, 75, 81, 82, 83, 107, 109, 110, 118, 119, 123, 124, 125, 254, 256 pyrolysis, 192, 195, 380 pyrolytic graphite, 7 pyromellitic dianhydride, 316 pyrrole, 306

Q quadratic curve, 160 quantum, 4, 22, 50, 69 quantum chemistry, 2

442

Index

quantum confinement, 3 quartz, 8, 9, 21, 31, 39, 42, 112, 199, 288, 301, 303, 311, 326

R radiation, 284 radical polymerization, 227, 275 radio, 9 radiological, 205 radius, 4, 24, 25, 51, 52, 78 Raman spectroscopy, 196, 375 random, 10, 25, 26, 28, 30, 65, 155, 157, 162, 274, 372 random errors, 162 range, xii, 6, 8, 25, 36, 42, 44, 46, 51, 62, 75, 88, 102, 107, 109, 110, 114, 116, 122, 136, 137, 158, 159, 174, 185, 186, 190, 195, 196, 197, 208, 220, 227, 232, 234, 235, 242, 316, 325, 353, 372, 373, 379, 380, 401, 406, 417, 418, 425 rating scale, xii, 330, 333, 338, 343 reactant, 200 reaction temperature, 275, 277 reaction time, 275, 277, 279, 280, 288 reaction zone, 200 reactive groups, x, 88, 273, 285, 287, 300, 302, 304, 307, 326 reactivity, x, 69, 225, 226, 227, 234 reagent, 235 reagents, 204 real time, 236, 311 receptors, 126, 364 recognition, x, xi, 88, 273, 274, 291, 292, 293, 294, 296, 300, 304, 305, 306, 327, 337 reconstruction, 20 recovery, xii, 1, 254, 329, 331, 332, 333, 338, 339, 341, 344, 345 red light, xi, 306, 323, 326 redox, 231, 234, 303, 304, 312, 315, 326 reflection, xii, 347, 356, 358, 360 refractive indices, 214 refrigeration, 208 regeneration, 88, 98, 104, 127, 299, 303, 330, 331, 338, 339, 340, 341, 342, 345 regenerative capacity, 342 regenerative medicine, 330 regression, 162 regression analysis, 162 regrowth, xi, 329, 330, 338, 339, 341, 342, 343 regular, 19, 400 reinforcement, ix, xii, 154, 185, 204, 380, 399, 402 relationship, 24, 155, 158, 161, 162, 250, 252, 389, 394, 395

relationships, x, 80, 156, 157, 158, 161, 162, 166, 175, 239, 241, 265, 341, 388, 389 relaxation, 242, 406 relaxation time, 242 remodeling, 330 renewable resource, 63 repair, xii, 330, 331, 338, 341, 343, 344 repeatability, 299 reservoir, 77, 205 residues, xi, 273, 275, 279, 290, 291, 293, 294, 298, 300 resilience, 400, 416 resin, 218 resins, 217 resistance, x, xiii, 1, 113, 121, 157, 205, 225, 228, 229, 230, 235, 273, 279, 307, 316, 318, 373, 379, 400, 415, 416, 417, 418, 420 resistive, 7 resistivity, xiii, 228, 231, 415, 416, 417, 418, 420, 421, 422, 425 resolution, 206, 294 resources, 63 respiratory, 345 response surface methodology (RSM), viii, 153, 155 response time, 229, 232, 233, 234 responsiveness, 344 retardation, 157 retention, 208, 308, 309, 315, 316, 331 revascularization, xi, 329, 340 rheology, 242 Rho, 150 rigidity, 83, 330 rings, 2, 205, 306 RNA, 88 rodent, 332 rolling, 4, 5 room temperature, 12, 20, 31, 42, 83, 157, 159, 190, 208, 228, 229, 287, 332, 333, 350, 401, 417 ROP, 321 roughness, 12 Royal Society, 178, 221 rubber, 204 runoff, 67 Russian, 6, 54, 144 Russian Academy of Sciences, 144 ruthenium, 327 rutile, 212

S safety, 206, 208 saline, 333, 351 salt, 74, 79, 81, 114, 118, 138, 190, 199, 246, 247, 248, 249, 254, 255, 257, 260, 261, 262, 263, 366

Index salts, 138, 246, 254, 255, 262, 263, 265, 348, 376 sample, 13, 15, 32, 33, 35, 36, 37, 38, 39, 41, 43, 44, 67, 68, 69, 157, 174, 210, 231, 288, 307, 323, 351, 392, 418 sand, 67 saturation, 12, 297, 300 scaffold, 104, 126, 301, 330, 345, 348, 349, 351, 353, 355, 357, 361, 362, 364, 365, 366, 368, 372, 373, 374, 376, 377 scaffolding, 341 scaffolds, xii, 75, 86, 89, 98, 103, 104, 123, 126, 127, 128, 135, 139, 155, 180, 203, 219, 330, 340, 345, 347, 348, 350, 351, 352, 353, 355, 356, 357, 361, 363, 364, 365, 366, 367, 368, 372, 373, 374, 375, 376, 377 scalable, 197 scaling, 80, 90 scaling relations, 80 scaling relationships, 80 scanning electron microscopy, 10, 75, 288 scattering, 214 schema, 255 Schiff base, 89, 130 scholarship, 413, 426 Schwann cells, 103, 333, 335, 340, 341, 342, 343 scientific community, 6, 236 scores, 338 SCP, xi, 329, 330, 331, 332, 333, 334, 335, 338, 339, 340, 341, 342 SDS, 350 seals, 205 search, 79, 212 searching, 97 seed, 16, 21, 198 seeding, 126, 155, 349 seeds, 17 segregation, 219 selecting, 203, 254 selectivity, 227, 229, 231, 232, 234, 235, 315, 327 self, 374, 375 self-assembling, 138 self-assembly, 76, 138, 139, 348 self-organization, 138 SEM micrographs, 33, 79, 84, 91, 92, 101, 105, 106, 109, 110, 112, 114, 118, 121, 122, 123, 126, 127, 132, 136, 137, 189, 193, 199, 211, 353, 354, 355, 364, 368 semiconductor, 214 semi-permeable membrane, 89 sensing, 52, 227, 228, 229, 230, 231, 233, 234, 235, 236 sensitivity, vii, 53, 86, 128, 155, 227, 228, 229, 232, 233, 234, 235, 351, 381

443

sensors, viii, ix, x, xi, 8, 52, 75, 154, 155, 184, 185, 218, 219, 225, 227, 228, 229, 231, 232, 234, 235, 236, 305, 306, 325 sentences, 160 separation, vii, xi, 35, 36, 38, 39, 41, 44, 61, 65, 76, 81, 89, 98, 121, 136, 137, 208, 216, 233, 241, 274, 275, 296, 299, 300, 303, 304, 326, 340, 348 series, x, 57, 78, 156, 159, 174, 273, 274, 321, 353 serum, x, 102, 104, 273, 277, 351, 364, 368 serum albumin, x, 102, 273, 277 shape, 2, 15, 20, 21, 37, 51, 77, 91, 111, 113, 134, 137, 138, 139, 153, 187, 188, 192, 210, 226, 249, 254, 257, 259, 263, 285, 368, 401, 406, 417, 420 shear, xiii, 80, 215, 242, 243, 381, 400, 415, 416, 422, 425 Shell, 95, 96, 271 shortage, 219 shortages, 219 shrimp, vii, 61, 63 signal transduction, ix, 184 signaling, 345 signaling pathway, 345 signals, 52, 370 signal-to-noise ratio, 231 signs, 332 silica, 212, 230, 232, 243, 245 silicon, 6, 9, 14, 28, 37, 58, 199, 204, 232 silk, 155, 241, 303, 332, 348, 374 silver, 98, 123, 132, 133, 257, 327 similarity, 406 Singapore, 77, 143, 266 SiO2, 112, 212, 213, 214, 254, 255, 257, 258, 261, 262, 264 sites, ix, xiii, 10, 20, 25, 28, 30, 126, 184, 199, 215, 225, 226, 227, 230, 231, 234, 254, 291, 302, 399 skeletal muscle, 377 skin, 103, 127, 332, 348, 352 SMA, 369 SO2, 318 sodium, 65, 66, 135, 279, 333 sodium hydroxide, 66 software, 162, 333, 354, 387 soil, 67, 87 solar, 306 solar energy, 306 sol-gel, 9, 232 solid phase, 201 solid state, 12, 118, 120 solid tumors, 306 solid waste, 372 solidification, 64, 154, 191, 203 solubility, 12, 65, 87, 89, 114, 127, 136, 190, 201, 229, 348, 349, 350

444

Index

soot, 8 South America, 102 South Korea, 62 space shuttle, 218 spacers, 38 Spain, 222 spatial, 16, 17 species, 6, 14, 15, 227 specific adsorption, 274 specific heat, 401, 417 specific surface, 75, 123, 125, 126, 185, 206, 208, 210, 219, 381 specificity, 279, 294 spectroscopy, xi, xii, 66, 233, 254, 305, 311, 347, 349, 351, 360 spectrum, 99, 107, 123, 185, 186, 194, 351, 355, 357, 359, 372 speed, xii, xiii, 64, 78, 174, 254, 383, 384, 399, 400, 402, 404, 406, 407, 408, 411, 412, 415, 416, 417, 420, 421, 422, 423, 425 spermatozoa, 301 spheres, 83, 242, 246 spin, 159, 309, 311, 349 spinal cord, xi, 329, 330, 331, 332, 333, 334, 335, 337, 338, 339, 340, 341, 342, 343, 344, 345 spinal cord injury, xi, 329, 331, 339, 340, 341, 343, 344, 345 spindle, 79, 83, 84, 108, 111, 124, 285, 287 spines, 63 sponges, 375 spontaneous recovery, 339 sputtering, 9, 20, 21, 31, 39, 42, 45 St. Louis, 333 stability, ix, xi, xii, 25, 30, 52, 65, 66, 88, 101, 131, 133, 136, 185, 215, 225, 227, 229, 231, 233, 234, 235, 305, 316, 323, 331, 344, 385, 399, 410, 412, 416 stabilization, ix, 184, 190, 192, 193, 194, 195 stabilize, 81, 107, 119, 233 stabilizers, 157 stages, 64, 122, 188, 192, 194, 196, 274, 330 stainless steel, 32, 33, 37, 332 standard deviation, viii, 65, 67, 68, 83, 97, 153, 155, 156, 162, 164, 173, 352 standard error, 338 standards, 67 star polymers, 325 statistical analysis, 163, 166 statistical inference, 160 statistics, 23, 163 steel, 32, 33, 37, 332, 384 stem cells, 341, 345, 374 stent, vii, 61

steric, 197, 274, 315 sterile, 122, 124, 331, 332 sterilization, 88, 122, 331, 343 stiffness, xii, 214, 235, 306, 342, 379, 381, 399, 400, 402, 412, 416 stimulus, 7 stock, 63 storage, vii, ix, xii, 52, 68, 184, 185, 203, 205, 207, 208, 210, 215, 216, 241, 399, 402, 406, 410, 412 strain, xii, 2, 214, 344, 377, 379, 380, 381, 386, 387, 388, 389, 390, 391, 392, 393, 394, 395, 396 strains, 388 strategies, 90, 330, 331, 339, 343 strength, ix, xii, 2, 5, 6, 7, 52, 69, 71, 74, 77, 97, 116, 134, 139, 153, 157, 167, 168, 199, 203, 204, 210, 213, 214, 215, 220, 225, 249, 251, 274, 285, 286, 287, 306, 323, 327, 379, 380, 381, 388, 389, 390, 391, 395, 399, 400 stress, 12, 16, 17, 195, 214, 215, 235, 381, 387, 388, 389, 394, 395 stress-strain curves, 381, 387, 388, 394, 395 stretching, 83, 113, 166, 167, 168, 171, 190, 198, 240, 249, 251, 357, 358, 372 strong interaction, 290, 293 structural changes, 193, 194 structural defects, 30, 51 structuring, 210 styrene, 135, 241, 312 substances, 86, 98, 99, 127 substitution, 103, 122 substrates, 4, 8, 12, 20, 26, 31, 33, 39, 42, 45, 52, 78, 134, 199, 203, 231, 307 Sudan, 59 sugar, xi, 274, 281, 283, 284, 291, 293, 299, 302, 304 sulfur, 16, 203, 360 supercapacitor, 206 supernatant, 351, 352 suppliers, 91, 102 supply, 9, 13, 14, 32, 45, 62, 64, 77, 157, 159, 187, 340, 349, 350 surface area, vii, viii, ix, xi, 61, 62, 69, 73, 154, 155, 184, 185, 186, 188, 189, 203, 205, 206, 208, 210, 212, 213, 214, 225, 226, 227, 230, 231, 232, 234, 242, 246, 253, 256, 263, 274, 305, 306, 315, 323, 334, 349 surface chemistry, 1 surface diffusion, 21 surface energy, 20, 69 surface modification, x, 226, 230, 273, 274, 275, 287, 290, 299, 300, 301, 376 surface structure, 75, 192

Index surface tension, x, 64, 77, 79, 81, 89, 90, 91, 100, 102, 153, 154, 187, 188, 189, 191, 239, 240, 242, 243, 246, 247, 248, 249, 281, 285, 349 surface treatment, 9, 231 surfactant, 81, 91, 94, 99, 115, 188, 189, 257 surgery, 145, 204, 332, 352 surgical, 332, 352 survival, 331, 345 swelling, 74, 119, 138, 228, 361 SWNTs, 2, 8, 227, 231, 263 symmetry, 4 synapse, 337 synergistic, 99 synergistic effect, 99 synthesis, vii, x, 7, 8, 11, 14, 32, 33, 39, 43, 45, 54, 76, 107, 226, 273, 275, 276, 299, 302, 306, 311, 312, 314, 316, 317, 321, 322, 323, 352, 366 synthetic polymers, 62, 63, 70, 79, 87, 128, 241, 348 systems, ix, xii, 9, 16, 25, 79, 81, 109, 118, 123, 125, 127, 135, 138, 157, 161, 184, 190, 204, 205, 210, 234, 240, 243, 264, 374, 379

T T cell, 376 ta-C film, 25 talc, 380 tanks, 215, 216 technology, 75, 135, 197, 240, 265 TEM, 16, 18, 19, 21, 26, 27, 98, 217, 257, 258, 259, 260, 261, 263, 264 temperature dependence, 51, 418 tenascin, 331, 343 tendons, 63 Tennessee, 225 tensile, xii, 5, 16, 17, 97, 116, 157, 203, 215, 220, 225, 306, 379, 380, 381, 386, 388, 389, 391, 392, 394, 395 tensile strength, 5, 97, 116, 157, 203, 220, 225, 306, 380, 388, 389, 391, 395 tensile stress, 16, 17, 215, 381 tension, 77, 81, 153, 188, 189, 235, 243, 246, 249, 349, 384 test data, 156, 157, 160, 162, 166, 176 tetrahydrofuran, 75, 228, 229 textile, viii, 74, 77, 90, 91, 98, 102, 103, 104, 157, 220 textiles, viii, 70, 73, 103, 301, 328 TFE, 350, 354 TGA, xii, 379, 381, 385, 386, 396, 399, 402, 410 therapy, 306 Thermal Conductivity, 6, 417 thermal decomposition, 6, 19, 385 thermal degradation, 410

445

thermal denaturation, 232 thermal expansion, 215 thermal properties, viii, xii, xiii, 184, 379, 399, 400, 402, 416 thermal resistance, 418, 420 thermal stability, ix, xi, xii, 88, 225, 234, 305, 316, 323, 385, 399, 410, 412 thermal treatment, 205 thermodynamic equilibrium, 23 thermodynamics, 7 thermograms, xii, 379, 403, 404 thermogravimetric, 316, 401, 402 thermogravimetry, xii, 399 thermolysis, 316 thermoplastic, 192, 348, 379, 380, 400, 416, 417 thermoplastic polyurethane, 417 thin film, 22, 31, 32, 33, 35, 39, 42, 43, 45, 46, 134, 226, 288, 303, 326 thoracic, xi, 329, 331, 332 three-dimensional, 126, 192, 195, 344, 348, 368 threshold, 9, 25, 26, 28, 29, 42, 44, 46, 49, 77, 153, 381 thymus, 230, 232 ticks, 283, 293, 304 time periods, 350 timing, 383 tin, 26 tin oxide, 26 TiO2, 96, 99, 212, 213, 214, 256, 326 tissue, vii, ix, xii, 61, 75, 86, 98, 103, 104, 123, 126, 127, 128, 135, 139, 154, 184, 203, 219, 230, 231, 303, 321, 330, 331, 332, 341, 342, 345, 348, 352, 365, 368, 371, 372, 373, 374, 377 tissue engineering, vii, ix, xii, 61, 75, 86, 99, 103, 123, 126, 127, 128, 135, 139, 154, 178, 179, 184, 203, 219, 268, 321, 330, 341, 345, 348, 368, 372, 373, 374, 377 titania, 212, 257 titanium, 63, 98, 271 titanium dioxide, 98 titanium isopropoxide, 63 tobacco smoke, 216 Tokyo, 351 toluene, 228, 229 topical anesthetic, 332 topology, 75 total energy, 2 toughness, ix, 185, 204, 218, 379 toxic, vii, viii, 61, 73, 87, 107, 123, 129, 216, 279, 348 toxicity, 114, 157 toxins, 63, 216 trading, 2

446

Index

traffic, 52 transfer, 75, 77, 88, 195, 215, 227, 228, 233, 234, 235, 306, 307, 311, 315, 316, 318, 325, 326, 327, 381, 402 transformation, 192, 402 transistors, 52, 154 transition, 8, 9, 80, 98, 199, 201, 402, 406 transition metal, 8, 9, 199, 201 transition temperature, 402, 406, 412 translation, 4 transmission, 6, 43, 45, 46, 50, 52, 75 transmission electron microscopy, 75 transparent, 332, 350 transplantation, 343 transport, 4, 15, 52, 168, 204, 208, 216, 327 travel, 188, 251, 384 Trichoderma viride, 87 trifluoroacetic acid, 75 triggers, 189 trypsin, xii, 347, 349, 351, 352, 363, 372 T-test, 338 tubular, 2, 4, 9, 234, 236 tumor, 107 tumors, 306, 344 tunneling, 22, 25, 50 two-dimensional, 4, 6

U ultra-fine, 198, 327 ultrasound, 381 ultra-thin, 348 ultraviolet (UV), viii, 73, 118, 119, 227, 284, 297, 331 uncertainty, 23, 156 uniform, 16, 17, 23, 79, 99, 134, 156, 171, 174, 175, 188, 190, 246, 251, 284, 285, 286, 287, 307, 312, 316, 322, 353, 400 Union Carbide, 7 United States, 62 urea, 350 uric acid, 230, 232, 233 urine, 232 UV irradiation, 118, 119 UV light, 227, 331 UV radiation, 284

V vacuum, 22, 26, 30, 33, 50, 52, 350, 382, 401, 417 valence, 1 validity, 81, 162 values, xiii, 23, 25, 33, 35, 36, 39, 43, 44, 90, 99, 102, 107, 110, 129, 131, 132, 133, 136, 156, 163,

164, 165, 166, 207, 220, 369, 389, 394, 396, 406, 412, 415, 420, 422, 425 Van der Waals, 2, 400, 416 vapor, viii, ix, xii, 6, 8, 9, 12, 14, 16, 26, 30, 31, 54, 66, 81, 119, 120, 121, 183, 184, 185, 192, 198, 200, 201, 204, 216, 219, 220, 226, 229, 230, 350, 379, 395 vapor grown carbon nanofibers, viii, 184, 200 variables, 79, 82, 153, 155, 156, 158, 160, 161, 162, 163, 165, 171, 174, 175, 176, 252 variation, 16, 36, 42, 44, 46, 49, 50, 83, 163, 164, 189, 389 vascularization, 340 vector, 4, 162 VEGF, 344 vehicles, 183, 207 velocity, 15 vessels, 340, 341 vibration, 204, 288, 303, 326, 357, 358, 372 vinyl chloride, 81, 267 vinylidene fluoride, 267 violent, 283 viral infection, x, 273 viscosity, 79, 80, 83, 89, 90, 96, 110, 119, 131, 154, 158, 188, 198, 242, 243, 246, 247, 248, 249, 252, 254, 264, 285, 323, 339, 349, 394, 395, 396 visible, 214, 335, 339 visualization, 7 volatility, 79, 81 voltammetric, 231, 232, 257

W waste management, 219 waste treatment, 204 waste water, vii, 61, 88, 212 wastewater treatment, 212 water vapor, 119, 121 water-soluble, viii, 73, 87, 90, 91, 96, 97, 107, 110, 113, 117, 128, 156, 234, 325 water-soluble polymers, 110, 113, 128 wave vector, 3 weak interaction, 293 web, 193, 197, 198, 206, 210, 327, 349 weight loss, 99, 120, 122, 385, 410 weight ratio, 69, 79, 91, 92, 97, 99, 101, 108, 109, 110, 111, 113, 114, 118, 119, 121, 122, 123, 132, 213 wettability, x, 225, 226, 235, 358 wetting, 16, 70, 382 wind, 139 wires, 2, 4 Wistar rats, 352 wood, 204, 218

447

Index wool, 348 workers, 7 worms, 63, 229 wound healing, 75, 86, 88, 89, 99, 100, 103, 104, 106, 107, 123, 124, 126, 128, 139, 352, 370, 371, 374

X X-ray diffraction, 75, 91, 113 X-ray photoelectron spectroscopy (XPS), 288, 289

Y yarn, 217 yeast, 63 yield, xii, 8, 31, 67, 81, 131, 190, 192, 204, 215, 275, 379, 388, 389, 390, 395 yttrium, 8

Z zinc (Zn), 205, 206, 208, 262, 263, 321 zirconium, 264 Zn, 262, 263