©2001 CRC Press LLC
©2001 CRC Press LLC
P1: SYV/FYX PB035-FM
P2: SYV/ABE March 15, 2001
QC: SYV/ABE 14:5
T1: SYV
...
134 downloads
834 Views
5MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
©2001 CRC Press LLC
©2001 CRC Press LLC
P1: SYV/FYX PB035-FM
P2: SYV/ABE March 15, 2001
QC: SYV/ABE 14:5
T1: SYV
Char Count= 22589
I dedicate my book chapters to the loving memory of my late husband, Ekhiel F. Khait, M.D., and to my children Irene and Alexander. Klementina Khait
I dedicate my contributions in this book to my wife, Virginia McMillan Carr, and my daughters, Rosamond and Louisa. Stephen H. Carr
©2001 CRC Press LLC
v
P1: SYV/FYX PB035-FM
P2: SYV/ABE March 15, 2001
QC: SYV/ABE 14:5
T1: SYV
Char Count= 22589
Contents
Preface Acknowledgments 1.
OVERVIEW OF POLYMER POWDERS STEPHEN H. CARR
References 2.
CONVENTIONAL METHODS OF POWDER PRODUCTION STEPHEN H. CARR
Size Reduction Process Selection for Size Reduction Micromechanics of Pulverization Power Efficiency in Pulverization Processes Foundation for Scaling Laws Other Considerations References 3.
PRINCIPLES OF SOLID-STATE SHEAR PULVERIZATION (S3 P) KLEMENTINA KHAIT
Introduction Review of Strain-Assisted Grinding of Polymers Summary References
©2001 CRC Press LLC
P1: SYV/FYX PB035-FM
P2: SYV/ABE March 15, 2001
4.
QC: SYV/ABE 14:5
T1: SYV
Char Count= 22589
EQUIPMENT FOR THE S3 P PROCESS FOR PLASTICS AND RUBBER MARTIN H. MACK
Machine Functions of the Twin-Screw Extruder Modular Design of Barrel Sections Modular Design of Screw Elements with Conveying, Melting, and Pulverization Functions Conveying Elements Kneading Blocks Heat Transfer Calculations for Barrel and Screws Screw Cooling Process Examples Commingled Plastics from Recycling Sources Pulverization of Cured Rubber Scale-up Considerations for the S3 P Process Summary References 5.
S3 P TECHNOLOGY AND VIRGIN POLYMERS KLEMENTINA KHAIT
Applications for Plastic Powders Characterizing Powders Summary References 6.
S3 P TECHNOLOGY AND POLYMER BLENDS STEPHEN H. CARR
Compatibilization of Polymer Blends The Strategy of Self-Compatibilization The Mechanisms that Underlie the Operation of S3 P References 7.
VALUE-ADDED PRODUCTS MADE FROM RECYCLED PLASTICS VIA S3 P KLEMENTINA KHAIT
State of Plastics Recycling S3 P and Plastics Recycling Physical Properties of Commingled Plastics
©2001 CRC Press LLC
P1: SYV/FYX PB035-FM
P2: SYV/ABE March 15, 2001
QC: SYV/ABE 14:5
T1: SYV
Char Count= 22589
Characterizing Powders Summary References 8.
APPLYING S3 P TECHNOLOGY TO THE RECOVERY OF USED-TIRE RUBBER KLEMENTINA KHAIT
State of Rubber Recycling Characterizing Powders Physical Properties of Tire Rubber Powder/Plastic Composites Emerging Technologies in Tire-Rubber Recycling Summary References Epilogue: Toward the Future Glossary
©2001 CRC Press LLC
P1: SYV/FYX PB035-FM
P2: SYV/ABE March 15, 2001
QC: SYV/ABE 14:5
T1: SYV
Char Count= 22589
Preface
T
book is the first extended look at a new and multifaceted polymer processing technology that has already been discussed in numerous journal articles. Called Solid-State Shear Pulverization (S3 P), this innovative process produces polymeric powders with unique physical properties not found in the output of conventional size-reduction methods such as ambient or cryogenic grinding. This technology, which utilizes a pulverizer based on a modified corotating twin-screw extruder made by the German manufacturer Berstorff, has profound implications for both the creation of new polymer blends and the recycling of plastic and rubber waste. Some principles of S3 P have their origins in work by Percy Bridgman at Harvard University in 1935. Bridgman studied the transformation of solid substances under simultaneous action of high shear and compression. Following up on his research on solids, scientists in Europe, the former Soviet Union, the United States, and Japan studied chemical reactions in polymers induced by the application of mechanical energy. They termed this new field of science “mechanochemistry.” Excellent books in the field of polymer mechanochemistry have been published by Baramboim in the former Soviet Union [1], Simionescu and Oprea in Romania [2], and Porter and Casale in the United States [3]. These authors provide comprehensive reviews of topics including polymer stress reactions, methods for characterization of these reactions induced by stress, and mechanical synthesis of block and graft copolymers due to the formation of a chemically reactive species that they called mechanoradicals. Subsequently, Sohma in Japan reviewed methods for producing mechanoradicals in the solid state and molecular approaches to fracture of polymers [4]. HIS
©2001 CRC Press LLC
P1: SYV/FYX PB035-FM
P2: SYV/ABE March 15, 2001
QC: SYV/ABE 14:5
T1: SYV
Char Count= 22589
Extensive research was conducted by Enikolopyan and co-workers since the early 1970s at the Russian Academy of Sciences, which resulted in the development of Elastic Deformation Grinding (EDG) of polymers. This was a new method of producing powder in a batch mixer or an extruder by first melting a polymer and then grinding it to powder. Despite numerous publications during the past 30 years by the Russian researchers, however, the mechanism of powder formation via EDG is not yet fully understood. These workers eventually came to refer to EDG as “Solid-State Shear Extrusion” (SSSE). I first became interested in EDG (or SSSE) technology in the late 1980s. Initially, I was intrigued by the idea of applying mechanochemistry to the in-situ compatibilization of mixed plastics from the waste stream. Unlike SSSE, where polymers are melted prior to pulverization, I proposed pulverizing mixtures of polymers with the S3 P process, which does not involve melting. By contrast, S3 P maintains polymers in the solid state and avoids the additional heat history that occurs during the SSSE process, which can be detrimental to the physical properties of pulverized materials. The research and development of the S3 P technology at the Polymer Technology Center at Northwestern University has grown significantly since 1990 from the development of a new plastics recycling process to a much broader polymer processing method that allows intimate mixing of polymers with very different viscosities, solid-state dispersion of additives, including pigments, and continuous production of powder with unique shapes and large surface areas. Polymeric powders are of growing importance to plastics processors due to the increased use of plastics in various applications, such as rotational molding, powder coatings, and compounding, which require powder as the feedstock. In more recent years, it has become clear that this process allows for insitu compatibilization of dissimilar polymers by applying mechanical energy to cause chemical reactions. This aspect of the S3 P technology that we describe in this book should interest those individuals who are developing new polymer blends with the use of pre-made compatibilizing agents. In addition, it has been discovered that S3 P efficiently mixes polymer blends with different component viscosities, resulting in the elimination of phase inversion. The S3 P process directly produces blends with matrix and dispersed phase morphology like those obtained after phase inversion during a long melt-mixing process. This phenomenon is of practical importance because a long processing time is required by conventional melt-mixing to produce a stable blend morphology. S3 P is also advantageous for producing thermoplastic or thermoset powder-coating compounds in a one-step process as opposed to a conventional multi-step operation that involves melt extrusion followed by batch grinding. The major capabilities of this new process can be summarized as follows:
r continuous powder production from plastics or rubber feedstocks r blending of immiscible polymers
©2001 CRC Press LLC
P1: SYV/FYX PB035-FM
P2: SYV/ABE March 15, 2001
QC: SYV/ABE 14:5
T1: SYV
Char Count= 22589
r efficient mixing of polymers with unmatched viscosities r environmentally friendly recycling of multicolored, commingled plastics waste
r solid-state dispersion of heat-sensitive additives r engineered plastic/rubber blends
It is hoped that this book will be of use both as a general technical resource and as a class text for undergraduate or graduate courses centered on polymeric powder production, polymer blends, efficient mixing of polymers, and dispersion of additives in plastics. The book should also be of interest to scientists, engineers, and processors, including those involved in the recycling of both preand post-consumer plastic and rubber waste. Because the S3 P process allows economical recovery of mixed plastics without sorting by type or by color, diverse markets should emerge for utilizing S3 P-made powders in creating many new value-added consumer products from recycled waste. KLEMENTINA KHAIT
D
novel methods for processing polymeric materials has been a large and crucial field of opportunity for over a century, and the S3 P process, with which this book is largely concerned, seems likely to become a significant technology in this area. The compelling advantages of the S3 P process include its ability to make fine powders of polymeric materials (plastics as well as elastomers), its ability to create intimate mixtures from very heterogeneous feeds, and its ability to alter physical properties—often making improvements over those of the constituents themselves—in the resulting blends. This latter effect is due, in part, to the mechanochemistry that occurs to the material undergoing S3 P processing. When I first learned of the S3 P process from Dr. Khait about a decade ago, it was immediately clear to me that this was a technology for which I would want to make a contribution, and it has been a pleasure to do so over that period of time. I sincerely wish to express my appreciation to her for letting me become a research collaborator in the advancement of this technology. Furthermore, the intellectual challenges associated with trying to understand its fundamental aspects are substantial—just as a professor would wish! EVELOPING
STEPHEN H. CARR
©2001 CRC Press LLC
P1: SYV/FYX PB035-FM
P2: SYV/ABE March 15, 2001
QC: SYV/ABE 14:5
T1: SYV
Char Count= 22589
References 1. Baramboim, N. K. “Mechanochemistry of Polymers.” Rubber and Plastic Research Association of Great Britain, London: MacLaren & Sons, Ltd. (1964). 2. Simionescu, C. and Oprea, C. V. Mechanochemistry of Polymers, Moscow: Mir (1970). 3. Casale, A. and Porter, R. S. Polymer Stress Reactions, New York: Academic Press (1978). 4. Sohma, J. Progress in Polymer Science, 14:451–586 (1989).
©2001 CRC Press LLC
P1: SYV/FYX PB035-FM
P2: SYV/ABE March 15, 2001
QC: SYV/ABE 14:5
T1: SYV
Char Count= 22589
Acknowledgments
I
was a genuine pleasure for us to collaborate on this book with Martin Mack. We would also like to acknowledge the contributions from Prof. John M. Torkelson and his group (Naomi Furgiuele, Andrew Lebowitz, Manisha Singh, and Albert Davydov) for their basic research on polymer blends made with S3 P. Prof. Mark A. Petrich (formerly of Northwestern University) and his graduate students (Dongchan Ahn and Michelle Dietering) provided significant ESR data and particle shape characterization of S3 P-made powders. Richard Kwarcinski and John N. Rasmussen of our Polymer Technology Center at Northwestern University spent many hours in laboratory testing and processing with our Berstorff machines. Financial support from the Bureau of Energy and Recycling of the Illinois Department of Commerce and Community Affairs helped us to demonstrate the viability of our S3 P technology for plastics and rubber recycling. Credit is also due Northwestern University for providing a platform for carrying out this whole enterprise. The Material Sciences Corporation of Elk Grove Village, Illinois, has provided financial assistance for the continuing development and commercialization of S3 P. Special thanks go to research engineers Matthew A. Darling and Erin G. Riddick. Mr. Darling has dedicated several years to creative efforts in research, development, and scale-up of the S3 P technology and equipment. Mr. Riddick has contributed extensively to the applied research and development of value-added materials from mixed plastics recovered via S3 P, as well as to the scale-up of the technology and equipment. Stephanie V. Spindler and Charles Whitman patiently typed and retyped the manuscript, and Mr. Whitman provided invaluable editorial assistance and several photos and artwork for the book’s cover and color insert. Finally, we want to acknowledge T
©2001 CRC Press LLC
P1: SYV/FYX PB035-FM
P2: SYV/ABE March 15, 2001
QC: SYV/ABE 14:5
T1: SYV
Char Count= 22589
both the invitation from Technomic Publishing Co., Inc., to write this book and the most competent cooperation provided by our editor at Technomic, Susan Farmer. KLEMENTINA KHAIT STEPHEN H. CARR
©2001 CRC Press LLC
P1: FUY pb035-01
February 27, 2001
9:57
Char Count= 0
CHAPTER 1
Overview of Polymer Powders
P
OLYMER
(1) (2) (3) (4)
Non-sintering during storage Low melt viscosity at low shear stresses A wide range between melting and degradation temperatures Good dry-flow characteristics, e.g., low tendency to bridge and a modest value of critical angle of repose
powders are desired for a host of purposes, some of which relate to processing advantages but others of which relate to the nature of the final form. The nearly ubiquitous form of industrial polymers, molding pellets, offers many advantages such as near fluid-like handling characteristics without the risk of dust formation and an ability to be stirred for good mixing or heat transfer. Processing methods, including common extrusion and injection molding, have been optimized over the past decades for the efficient use of molding pellets [1]. Other techniques, however, such as rotational molding and metal coating, cannot be performed well with molding pellets [2]. Here, polymer powders offer the distinct advantage of uniform deposition of the in-process material and its nearly instantaneous fusion rate. Molding pellets lack these characteristics because their discrete, fraction-of-a-gram particles have relatively high thermal masses and, therefore, exhibit a sluggish tendency to stick together when heat is applied. Achieving thin, uniform polymer coatings by depositing polymer solutions works well but requires the management of large amounts of solvent (with its attendant cost penalty). Powders, on the other hand, lack these drawbacks altogether. There is a common set of general requirements for powders [1]. These include the following:
©2001 CRC Press LLC
P1: FUY pb035-01
February 27, 2001
9:57
Char Count= 0
(5) Low moisture sensitivity (6) No significant emission of volatiles during processing (7) Rapid cure after powder fuses, if a thermoset The properties of the material undergoing pulverization relevant to selecting the appropriate size reduction strategy include the following: (1) (2) (3) (4) (5)
Toughness/brittleness/hardness Cohesivity of the particles/particle shape Heat sensitivity Toxicity Potential for producing explosive dusts
The first two of these property groups are specific to the way particles break apart, and the last three deal with unavoidable considerations related to the size-reduction operation itself. A further significant factor is the shape of particles produced during the formation of polymer powders. Equiaxed particles lead to the best behavior in a powder mass, especially in terms of flow characteristics, but elongated particles lead to the most rapid fusion characteristics. Semi-crystalline polymers, especially polyolefins, are usually highly ductile and are, therefore, more difficult to pulverize. During size reduction, these materials acquire high amounts of elastic strain energy and so will fracture according to the different energy release rates for energy accumulation and for particle cleavage. Cross-linking poly(ethylene) via irradiation or chemical means reduces this ductility and makes the material behave better in size reduction processes. Polymers in the form of fine, free-flowing powders can be obtained by a variety of methods, including solution precipitation, isolation of nascent polymer directly from its synthesis reactor, and mechanical size reduction. This latter approach is by far the most commonly practiced. The fundamental considerations in the formation of polymer powders center on the simple idea of subjecting polymers to stresses sufficient to break pieces into successively smaller particles. In general, making powders is an act of size reduction, and the various strategies employed fall into these categories: (1) (2) (3) (4) (5)
Crushing Impacting Cutting Exploding Solution Spraying
©2001 CRC Press LLC
P1: FUY pb035-01
February 27, 2001
9:57
Char Count= 0
Machinery for doing this comes in all imaginable configurations, converting pellets or flakes into powder using cutting action or exploiting crushing forces. Some equipment is capable of varying temperature (especially to create cryogenic conditions) so as to modify the properties of the in-process material. Comminution is the generic term for these processes in which particulates are reduced in size by repetitive grinding via compression or impact. Mechanically ground powder typically has particles ranging from 10 to 40 mesh (2,000 to 420 m) in diameter. Commonly, these methods produce particle size distributions that conform to the log-normal function. Interestingly, there is also a limited amount of particle enlargement that also occurs during pulverization processes. Particle enlargement is termed “granulation” or “cold welding.” The energy consumed in pulverizing materials is exceptionally large, and most of that energy is dissipated in moving material within the equipment, rather than in the direct process of cleaving particles and making new surfaces. The fate of the energy distributes as follows: Machine losses 13 percent Heat dissipated in material 85 percent Size reduction <1 percent Typically, only 30 percent of a material is size-reduced to the desired size in a single pass. Of significant note, and of major concern in this book, is the novel process for producing polymer powders, Solid-State Shear Pulverization (S3 P). S3 P involves processing a feed stream of ground or shredded polymer fed into a modified co-rotating twin-screw extruder with cooling applied along the barrel where heat is ordinarily applied. The in-process material is not allowed to enter the molten state; otherwise, the process would fail to have the desired effect. Accordingly, the discharged material is the desired, free-flowing powder. Feed materials may include any or all of plastics, fibers, rubbers, or any combination of them suitable for S3 P processing. The essential action responsible for the benefit of S3 P lies in the intense shear stresses to which in-process particles become subjected. This process not only performs size reduction very efficiently, but it also produces particles whose surfaces have been modified to be chemically reactive. Such reactivity created by a mechanical process is termed mechanochemistry, and it is well-known to occur in size reduction of many materials. S3 P represents the most controllable size-reduction process ever employed, and it can be put to good advantage for adding value to the material being put through it. S3 P operates successfully on virtually all polymers, as Dr. Khait and her co-workers began learning during the 1990s. Some alterations are occasionally
©2001 CRC Press LLC
P1: FUY pb035-01
February 27, 2001
9:57
Char Count= 0
needed to accommodate the myriad polymers that have been so processed. Furthermore, the processing of the same polymer, in which the feed is a mixture of batches having significantly different viscosities, does successfully produce homogeneous blends with viscosities appropriately intermediate between those of the input batches. Polymer powders produced by S3 P processing are very suitable for fusion coatings (e.g., on steel), and they perform very well as feed for injection and compression molding. S3 P powders perform exceptionally well in rotational molding. The successful ability of S3 P to process blends of polymers has also been explored extensively. Most noteworthy is its ability to cause self-compatibilization to occur in almost all cases. The value of self-compatibilization is that highperforming alloys can be prepared without having intentionally to add another (usually polymeric) substance to achieve good properties. The costly and inefficient tactic of adding a compatibilizer is made unnecessary, because S3 P creates them via of the mechanochemistry resulting from particle fragmentation. These new species are hetero-block copolymers composed of the very same materials at the interface, and only in an amount on the order of a monolayer. No other such hetero-block chains are formed elsewhere in the in-process material. Compatibilizers have their beneficial function as a result of their acting to reduce interfacial energy and to form covalent bridges across interfaces. There are many good examples of these new alloys exhibiting properties superior to those of either of the constituent polymers. The ability of S3 P to process heterogeneous feed materials is very providential, because there is a huge opportunity awaiting a successful process for reutilizing commingled polymeric materials. These include post-consumer waste (ranging from that in curbside collections to construction debris) and industrial recyclate. Studies have been made of polymer blends that model the commonly found constituents in solid waste. It has been shown that S3 P creates valueadded blends of these polymer mixtures, as opposed to what can be obtained from common melt-processed mixtures. The explanation for this is found in the self-compatibilization that S3 P fosters. Used tires represent a huge, untapped depot of valuable hydrocarbon and high-modulus elastomer. In addition, used tires typically can be found in monumental piles that breed vermin and are vulnerable to catching fire from lightning strikes. Fires in tire piles can rarely be extinguished. Another problem is that the reinforcing structures (especially steel wire) built into tires render them very heterogeneous. Similarly, the rubber itself has different compositions and different extents of vulcanization from location to location within a given tire. Although the rubber of a tire cannot be remelted or dissolved, it can be subjected to grinding, but this can produce only a poor-quality powder. By contrast, S3 P can be employed to pulverize tires into powders, and these rubber powders can be very useful in numerous practical applications.
©2001 CRC Press LLC
P1: FUY pb035-01
February 27, 2001
9:57
Char Count= 0
In summary, polymeric powders are used as a feedstock in a wide variety of applications because they provide unique solutions to some processing techniques including, but not limited to, powder coatings, rotational molding, sintering, and compression molding. The usage of polymeric powders will increase with further advances in processing technologies and equipment. REFERENCES 1. Narkis, M. and Rosenzweig, N., eds. Polymer Powder Technology, Chichester, England: John Wiley & Sons, Inc. (1995). 2. Beall, G. L. Rotational Molding, Cincinnati, Ohio: Hanser Gardner Publications (1998).
©2001 CRC Press LLC
P1: FUY pb035-02
March 2, 2001
15:18
Char Count= 0
CHAPTER 2
Conventional Methods of Powder Production
SIZE REDUCTION
C
of pieces of some bulk material into powders is generically known as comminution. In comminution, mechanical work is applied to the in-process material in such a fashion that the pieces progressively become subdivided into ever-smaller particles. This fragmentation requires that cleavage of particles occurs progressively. In general, the equipment used to accomplish size reduction employs one or more of the following actions: ONVERSION
(1) Crushing: This involves the application of stress directly from two opposing solid surfaces. There may be a bed of particles in the space between these surfaces. (2) Impacting: This involves creating a collision between a surface and the particles. In most such processes, high levels of kinetic energy are also exchanged between particles as well. (3) Cutting: This involves the application of stress directly on each particle essentially with the edge of a solid. (4) Exploding: This involves the generation of stress by processes that take place within a particle, including the deposition of energy inside a particle so as to create either high pressures (e.g., from the generation of steam) or simply large thermal gradients. (5) Solution spraying: This involves preparing a solution of the substance in a volatile solvent and then spraying this solution into a drying chamber where the solute in each droplet remains as a dust particle (typically with a smooth, spherical shape).
©2001 CRC Press LLC
P1: FUY pb035-02
March 2, 2001
15:18
Char Count= 0
Over the centuries, specific kinds of equipment have evolved to accomplish these actions, either singly or in combination. These types of equipment can be categorized as follows [1]: (1) Crushers and roller mills: These employ wheels that roll over a flat surface or meshing gears. (2) Jaw crushers: These employ platens that are driven cyclically together, somewhat as in mastication. (3) Gyratory and cone crushers: These involve an eccentric cylinder or a centered cone rotating within a cylindrical chamber. (4) Roll crushers: These employ counter-rotating cylinders set with a gap sufficient to draw particles into it and to cause them to be crushed. (5) Pendulum, table, and bowl-types: These include suspended rollers that can swing against the inside of a cylindrical chamber and do work on the material-in-process there; an air stream is imposed to sweep the fragments into an air classifier and, therefore, prevent the continued size-reduction of fragments as they are formed. (6) Ball and rod mills: These mills involve the in-process material being charged into a cylinder whose axis is substantially inclined from the vertical; balls (or rods) of a material harder than that to be pulverized are added, and then the surrounding cylinder is rotated about its axis. Rotational speed must be limited to the range between that at which the balls or rods will tumble and that at which they will merely roll inside the cylinder. The falling of these balls or rods is what creates the desired crushing action. (7) Mechanical impact mills: These include hammer mills, pin-in-disk mills, and turbine mills. (8) Impact crusher: In these machines, material is fed into a chamber at whose center is a rapidly rotating member. Repeated impingement upon the rotator followed by ricocheting about within the chamber leads to progressive breakup of the particles. (9) Fluid energy mills: These are various devices for accelerating particulate material in ways that cause high-energy impact events with solid surfaces and/or itself. (10) Cutting mills: These machines may involve ballistic impact against knife blades or the forcing of in-process material against such blades. Modifications of these processes may be practiced with certain variations imposed. For example, suspension of the in-process powder in a fluid (e.g., water or oil) is often exploited to eliminate the escape of dust during pulverization. If there is a certain size below which further reduction is undesirable, air entrainment and then recovery in a powder classification step is applied internally
©2001 CRC Press LLC
P1: FUY pb035-02
March 2, 2001
15:18
Char Count= 0
to the process. This prevents further crushing of the particles when they have been reduced to a certain size.
PROCESS SELECTION FOR SIZE REDUCTION The first area of consideration is the inherent nature of the material to be pulverized. The relevant factors include the following: (1) Toughness/brittleness/hardness: This refers to the various properties that make up the mechanical behavior of the material itself. (2) Cohesivity of the particles/particle shape: This refers to the inherent tendency of the particles to stick together. It measures the tendency of the bulk powder to exhibit “structure,” either at the micro- or macroscopic level. (3) Heat sensitivity: This refers to the ability of the material to remain unchanged during processing (when it is unavoidable to dissipate mechanical energy into heat). It may involve chemical damage to the in-process material, or it might include a fundamental change in the behavior of the material due simply to loss of moisture. (4) Toxicity: This refers to any acute or long-term adverse physiological effects to people in the vicinity of the size-reduction process. (5) Potential for producing explosive dusts: This refers to the possibility that dusts are made of combustible particles and are capable of propagating a flame front through them; ignition of such dusts in a confined space usually has explosive consequences. Dusts of non-combustible particles may nevertheless unexpectedly create huge electrostatic charges on surfaces over which they flow. Prasher has categorized these kinds of machinery along lines of the pulverization ranges desired, as shown in Table 2.1 [2].
MICROMECHANICS OF PULVERIZATION The result to be achieved during pulverization is the subdividing of pieces of the inputted material. This requires that cracks be formed within each such piece and that at least one of these cracks be propagated completely across it. The most efficient fragmentation events involve the propagation of a single crack and the least amount of plastic deformation within the piece. The initiation of multiple cracks simply consumes energy unproductively, as only the largest crack will advance the fastest. In addition, cracks that are attended by portions of the piece having undergone plastic deformation will have consumed strain energy that could otherwise have gone to advancing a crack.
©2001 CRC Press LLC
P1: FUY pb035-02
March 2, 2001
15:18
Char Count= 0
TABLE 2.1.
Stage of Reduction Primary crushing Secondary crushing Coarse grinding Fine grinding Very-fine grinding Super-fine grinding
Classification of Comminuted Products by Size. Particle Size (passing 80%, in micrometers)
Specific Energy (kWh/ton)
Compression or impact crusher Compression or impact crusher Rod mill Ball, or vertical spindle mill Tube mill
106 to 105
0–1 to 2
105 to 104
0–1 to 2
104 to 103 103 to 102
2 to 4 5 to 20
102 to 10
20 to 100
Attrition or fluid energy mill
10 to 1
100 to 1000
Typical Machine
c 1987, John Wiley & Sons Limited. Reproduced with permission from Reference [2]. Copyright
It should be kept in mind that stresses are developed within in-process pieces by forces being transmitted between them at their points of contact. This is obviously a statistical process, as neighboring pieces touch each other in an unpredictable way. Furthermore, materials that can deform at a contact point are inclined to nucleate a crack and allow it to advance into the piece. Alternatively, materials that have high fracture (or yield) strengths can support stresses that will ultimately form an internal cavity in a plane parallel with the incident forces. Internal cavitation of this kind is driven by a tensile stress field resolved from the applied compressive forces and will ultimately lead to breakage of the original piece as well. It is key to recognize that strain energy release-rate relationships, such as those advanced by Griffith early in the twentieth century, are pertinent to creating powders by cracking in-process material into ever-smaller pieces [3]. Here, fracture is a strain energy releasing process in which the strain energy released by fracturing must be greater than the energy to make new surface. Of course, there must be a crack propagation mechanism that can allow for this energy interchange to take place, but virtually all materials exhibit this. In addition, there needs to be a place in the piece to be broken where stresses are concentrated to such an extent that it localizes where new surface production will start. Such nucleation sites would most likely occur where there are shape discontinuities or flaws. The size of a flaw, a, determines the amount by which stress is intensified locally, and this represents the basis for the ∝ a −1/2 relationship. It even states that the stress levels needed to break some particle may be mostly a measure of the flaws inherent in the material being pulverized and its surface energy. As a consequence of the important role of inherent flaw sizes, one must conclude that smaller particles are harder to break than larger
©2001 CRC Press LLC
P1: FUY pb035-02
March 2, 2001
15:18
Char Count= 0
ones, due to their diminished likelihood of having flaws of significant (critical) size. The particle size, x, effect is also seen to originate from the following principle. One notes that stored elastic energy goes as x 3 but the work to propagate a crack goes as x 2 . This means that as particles get smaller, the energy stored to drive a crack goes down proportionally. Therefore, there may not be enough energy at the final stage of cracking to complete the process. Particles of highly ductile materials, such as plastics, behave in this way. Small particles convert imposed forces into plane stress conditions, which manifest the most ductility of which the solid is capable. At present, the transition in behavior as particle size diminishes is empirically established. Some plastics show an abrupt change in pulverization behavior, e.g., at 10 m for poly(methyl methacrylate) (PMMA). This size, dc , is dc = 32U f E ∗ /3Y 2
(1)
where Y is a shape factor, U f is the strain energy at fracture point, and E ∗ is elastic modulus. For a propagating crack, dU = dA
(2)
where is the work to make new surface, and A is the area of crack formed. For a crack growing across the midplane of a particle, U f = o a/E ∗
(3)
where o is the far-field stress of applied forces and a is the amount of crack advance that occurs. The fracture surface energy, , is = o a/2E ∗
(4)
Typical values for are glass: 103 –104 ergs/cm2 , plastics: 104 –106 ergs/cm2 , and metals: 106 –108 ergs/cm2 . Independent measurements of interfacial energies in these materials are many orders of magnitude smaller, leading to the usual conclusion that the energy to run a crack through a particle is much more than that required simply to make a new air/solid interface. Quantitatively, this fracturing in an individual particle can be applied to a mass of particles as follows. The quantity of new surface formed when reducing a
©2001 CRC Press LLC
P1: FUY pb035-02
March 2, 2001
15:18
TABLE 2.2.
Char Count= 0
Breakage Distribution Function by Product Size Interval Number.
Size Interval Number
Mass
bi j
Bi j
2 3 4 5 6
1.2 1.5 1.8 0.9
0.20 0.25 0.30 0.15 0.10
1.00 0.80 0.55 0.25 0.10
c 1997, John Wiley & Sons, Inc. Reproduced with permission from Reference [1]. Copyright
particle of size x1 to particles of size x2 is
x13 x23
ks x22 − ks x12 = New Surface
(5)
where x is particle size and ks is a shape factor; the subscripts refer to particles: 2 means larger particle; 1 means smaller particle. In terms of particle size distributions, the breaking process involves particles of size j being broken into smaller particles of size i. The specific breakage rate, S j , is the probability that a particle in classification interval j is broken during the process, and the resulting fragments represent a distribution of sizes, given as Bi j (for the cumulative distribution) or bi j (for the distribution in its differential form). These relationships are illustrated in Table 2.2. In such a process, the net flux of mass (m) converted from larger sizes ( j) into a smaller one (i) is expressed as: j=i−1 dm i S j m j (bi j ) − Si m i = dt j=1
(6)
The first term represents the creation of particles having mass m i , and the second term represents the loss of mass from the size interval in question (m i ) due to its own fragmentation process. It is noted that, in reality, Si is itself a mild function of particle size. This expression is known as the Batch Grinding Equation. This is, therefore, a model in two parameters, b and S.
POWER EFFICIENCY IN PULVERIZATION PROCESSES In general, the mechanical work to do size reduction involves energy supplied to the process, and the fate of that energy is as follows [4]: (1) The making of new surfaces (2) The deformation of the fragments (ubiquitous and pronounced in polymers)
©2001 CRC Press LLC
P1: FUY pb035-02
March 2, 2001
15:18
Char Count= 0
(3) The creation of internal micromechanical damage (4) The movement of particles through the apparatus (5) Frictional losses in the machinery itself Only this first process consumes energy in a fashion that relates to pulverization; the second and third processes represent ways in which work is done on particles, but they are unproductive in terms of reducing particle size. The fourth process consumes a lot of work in just moving particles around; this involves the kinetic energy of transport and is eventually lost to frictional heating within the moving mass of particles. Finally, the fifth process is the unavoidable reality that machines have friction. Empirically, it can be observed that there is a logarithmic (power-law) relationship (Figure 2.1) between specific energy input [Watt-sec/gm] and total specific surface [cm2 /gm] of the powder produced. The relationship seems good over several decades of particle size [5]. Observations over wider ranges of operations show that this is not a straight line; higher exponents are associated with larger particle sizes. The astonishing principle to note here, however, is that energy efficiency of grinding/crushing processes lies in the one percent range for typical crushing equipment!
FIGURE 2.1 The power-law relationship between input energy and the extent of pulverization. (Courtesy of the National Academy of Sciences, Washington, D.C. [5].)
©2001 CRC Press LLC
P1: FUY pb035-02
March 2, 2001
15:18
Char Count= 0
Specifically, the quantitative division of energy between its fates in a pulverizing process is typically machine losses: 13 percent, heat dissipated in material: 85 percent, and size reduction: <1 percent.
FOUNDATION FOR SCALING LAWS Attending this size reduction is an energy term, E. Overall for a process, E = CR
1 1 − x2 x1
(7)
C R is a constant for the process. In differential form, this rate of energy consumption per unit size reduction goes as x −2 . dE 1 = −C R 2 dx x
(8)
As with the usual fracturing process, the energy to drive a crack is orders of magnitude larger than that to simply make a new surface. From the world of practice, Bond found for small particles this energy, E B . Following is the well-known Bond Equation [6]: 10 10 EB = W √ − √ x2 x1
(9)
This relationship is used to calibrate the efficiencies of grinding apparatus, especially that of ball mills. In that context, E B is defined as the “Work Index.” Thus, this is the fundamental basis for the logarithmic (power law) relationship between specific energy input (Watt-sec/gm) and total specific surface (cm2 /gm). There are procedures for rating the efficiencies of such equipment: ASTM 409-71 (1978) and BS 1016 Part 20 (1981). In these tests, particle size is established by determining the size of mesh that will just pass 75 percent of the powder sample. These methods are useful in predicting power consumption associated with achieving different levels of size reduction or the shifting of the feed from one granular material to another. From Equation (9) in differential form, d EB = C B x −3/2 dx
©2001 CRC Press LLC
(10)
P1: FUY pb035-02
March 2, 2001
15:18
Char Count= 0
Or, in general, Equation (10) becomes d EB = C B x −n dx
(11)
where n differs for various pulverizing processes. Various considerations can result in different values of n, such as n = 1 (Kick) or n = 1.5 (Bond). For particles from a distribution of input particle sizes, n = 2 (Rittenger). OTHER CONSIDERATIONS Particle size enlargement can occur inside a pulverization process, and, of course, it opposes the usual purpose of subdividing particles into ever-smaller pieces. It can occur by the operation of several mechanisms. One is attraction, due to van der Waals forces, that develops when the pulverization action forces two contacting particles to reshape themselves so as to create a common, solidsolid interface between them. These larger particles are the result of a process called “granulation.” Another process can be called mechanical welding. It occurs when particles impinge upon each other in such a way that they stick together. This can occur under the action of collision with sufficient kinetic energy to join the particles. It can also happen when there are crushing forces that create sufficient intimate contact and high pressure that brazing occurs. With skilled mixing of many small particles with a few large particles, it is possible to create complete coating of the large particles by the small ones. An extrinsic substance, in either its liquid form or present as an adsorbed layer from the vapor phase, can aid significantly in the aggregation of the smaller particles. It is necessary that such a substance wet the particle surfaces. While powdered rubber is highly desirable because it is then in the optimal form for use in mixing (compounding) processes, the effort to convert gum rubber into powders is expensive [7]. Types and methods of rubber pulverization are: (1) Cryogenic grinding: Use of LN2 or LCO2 for refrigeration, followed by grinding by a selected method (2) Granulation: Fly cutting blades in a chamber (3) Attrition grinding: Kinetic energy impacting
REFERENCES 1. Encyclopedia of Chemical Technology —4th ed., New York: John Wiley & Sons, Inc. 22:279–296 (1997).
©2001 CRC Press LLC
P1: FUY pb035-02
March 2, 2001
15:18
Char Count= 0
2. Prasher, C. L. Crushing and Grinding Process Handbook, Chichester, England: John Wiley & Sons, Inc. (1987). 3. Griffith, A. A. “The Phenomenon of Rupture and Flow in Solids,” Philosophical Transactions of the Royal Society, A221:163–198 (1920). 4. Narkis, M. and Rosenzweig, N., eds. Polymer Powder Technology, Chichester, England: John Wiley & Sons, Inc. (1995). 5. Comminution and Energy Consumption, National Materials Advisory Board Publication NMAB-364, Washington, D.C.: National Academy Press (1981). 6. Bond, F. C. “The Third Theory of Comminution,” Ministry and Engineering Transactions, AIME, 193:484–494 (1952). 7. Evans, C. W. Powdered and Particulate Rubber Technology, London: Applied Science Publishers (1978).
©2001 CRC Press LLC
P1: FUY pb035-03
March 13, 2001
9:56
Char Count= 0
CHAPTER 3
Principles of Solid-State Shear Pulverization (S3 P)
INTRODUCTION
T
growing demand for fine polymeric powders with high bulk density, good flowability, narrow particle size distribution, and a particular shape requires continuous advances in powder technology and equipment. Commercially available polymeric powders are manufactured either by size reduction (comminution) or by solution process. Comminution is the process where materials are reduced in size by repetitive grinding via compression or impact. Mechanically ground powder typically has particles ranging from 10 to 40 mesh (2,000 to 420 microns) in diameter; these particles usually have an irregular shape (depending on the grinding equipment used) and are suitable for dry powder coating applications and moldings. Finer powders ranging from 200 to 635 mesh (75 to 20 microns) in diameter are more suitable for solvent dispersions, bonding agents for textiles, and additives. Particle size is a very important characteristic that affects the choice of a powder for a given application. In general, the smaller the particle size, the higher the price of the powder. Size-reduction equipment is usually not very efficient, is energy-intensive, and is costly to operate because only about 30 percent of a material is ground to a desired size in one pass. The equipment for size reduction is based on four principles: impact, compression, tension, or shear. The choice of equipment is dictated by the nature of the polymer and by the particle size desired. Coarse size reduction is achieved by shattering and then by crushing, where pieces are broken by impact. Fine and ultrafine size reduction is accomplished by pulverizing where particles are formed from tear, shear, abrasion, or attrition. The mechanism of fracture is usually successive, and the material is gradually reduced to a smaller and smaller particle size. HE
©2001 CRC Press LLC
P1: FUY pb035-03
March 13, 2001
9:56
Char Count= 0
Commercially available disintegrating devices for size reduction (grinding) of polymers, such as vibratory or jet mills, are based on two principles: impact (which results in the destruction of brittle materials) and cutting (used on soft, elastomeric substances). These devices are inherently inefficient and costly to operate (frequently their costs are more than 25 percent of total raw material processing) [1,2]. It has been found that at ambient temperature, these devices required up to 5,000 kWh/energy per ton of product, have high wear of knives or blades, and generate considerable frictional heat due to high speed during grinding, especially at the edges of the blades; heat can cause the polymer to stick to the moving surface [2]. The problem of heat can be solved by the use of cryogenic grinding, which employs liquid nitrogen as a cooling agent. Because liquid nitrogen is very expensive, however, the cryogenic process is more costly than ambient grinding.
REVIEW OF STRAIN-ASSISTED GRINDING OF POLYMERS Since the early 1980s, a group of scientists at the Academy of Sciences in Moscow under the direction of the late Enikolopyan conducted basic research on the behavior of various organic substances under the simultaneous action of high pressure and shear [3–5]. This work extended the original research by Nobel Laureate P. W. Bridgman at Harvard University who studied compressibility of solids under high pressure with the apparatus known as the Bridgman anvil [6]. Enikolopyan and his co-workers observed that under certain combinations of torque and temperature, using a modified Banbury mixer, some polymers can be converted into finely divided homogeneous powders without employing any dispersing additives. It was estimated that specific energy consumption was three to 10 times less than that used by existing grinding devices, which may range from 500 to 5,000 kWh per ton of the product [3]. It has also been noted that regardless of processing conditions, the particle size of the powder was dependent only upon the melt index of the polymer that was pulverized. Low-density poly(ethylene) (LDPE) with a melt index (MI) of 0.23 g/10 min and polydispersity of 8.6 was pulverized into a powder that had a particle size six to eight times larger than that of LDPE with an MI of 6.3 g/10 min and polydispersity of 5.9. It was suggested that the explanation for producing differently sized powders from the LDPE samples involved the heterogeneity of their structures, specifically the presence of branched macromolecules in an amorphous region. In the case of polymer mixtures, it has been observed that their deformability depends on the size of the dispersed phase: the larger the particles of that phase, the lower the deformability of the mixture [3]. For example, in polypropylene (PP)/LDPE mixtures, more rapid fracture occurred in the amorphous region of PP. Enikolopyan proposed
©2001 CRC Press LLC
P1: FUY pb035-03
March 13, 2001
9:56
Char Count= 0
that elastic energy accumulated in a solid polymer under high pressure and shear deformation was released with the formation of microcracks. As a result, a powder has been formed [4]. In some experiments, LDPE underwent brittle fracture at a temperature 5–10◦ C below its melting point. Strain-assisted grinding of some polymer mixtures established unusual results; polymers such as high-density poly(ethylene) (HDPE) and PP, which under given conditions were not pulverizable by themselves, were successfully pulverized as a mixture [4]. The compressibility of solids became a basis for the realization of a new principle of grinding polymeric materials under elastic-strain conditions using modified experimental devices including Banbury-type mixers or extruders [7]. This new method was initially named by the Russian researchers “Elastic Deformation Grinding” (EDG). Their study showed that the process was “avalanchelike” and the powder, with various dispersities, was formed simultaneously, not successively. Enikolopyan mentioned that EDG depends on several factors, such as temperature, pressure, and shear, among others. Under the complex stress conditions of compression and shear deformation, some polymers were readily pulverized, while others did not pulverize at all. However, even a small addition (3–5 weight percent) of another polymer pulverizable under those conditions made it possible to pulverize the mixture of two polymers [7]. Enikolopyan and his co-workers observed that polymeric materials that are subjected simultaneously to high pressure and shear undergo a drastic change in both physical and chemical properties [7]. Enikolopyan’s group conducted several experiments on shear deformation of solids with a small Brabender mixer using LDPE, and a powder of small particle size was produced [8]. They observed that intensive grinding occurred at 95–102◦ C, which is close to the temperature of the maximum rate of crystallization of LDPE during cooling of the melt. For this individual polymer, the powder formed at a glass-transition temperature (Tg), but in the case of polymer mixtures, the grinding temperature was a function of their composition [8]. For example, an LDPE/PP mixture at 40 weight percent of PP was converted into powder at Tg of pure LDPE. With higher concentrations of PP, phase inversion took place, and PP instead of LDPE formed the matrix of the material. The Tg value of that powder was 135◦ C, which is closer to that of pure PP [the temperature of the maximum rate of crystallization of PP was 123◦ C, according to differential scanning calorimetry (DSC) data, and 130–136◦ C, according to thermomechanical measurements]. Similar observations have been made for LDPE/PP and LDPE/polystyrene (PS) mixtures. Enikolopyan hypothesized that in the case of LDPE/PS mixture, grinding takes place in two stages starting with the PS portion at 120– 125◦ C followed by the “avalanche-type” grinding of the entire mass [8]. He further assumed that EDG of the polymer mixture starts at the crystallization temperature of the component that forms the matrix of the resultant material.
©2001 CRC Press LLC
P1: FUY pb035-03
March 13, 2001
9:56
Char Count= 0
The researchers stressed that energy consumption was low—40–50 kWh/ton— for a product with a powder particle size of 150–155 microns (100 mesh) [9]. Enikolopyan’s group initially described the principle of EDG as the use of mechanical force to be applied to the substance exposed to pressure and shear. The process resulted in “pumping” elastic energy to form new surfaces of that substance. Because this process can be implemented in modified extruders with a special rotary device at the end, it was sometimes called “extrusion grinding.” By varying the screw speed and the temperature along the zones of the extruder, it was possible to grind thermosets, vulcanized rubber, wood, cellulose-containing products, and others to powders of different particle size and particle size distribution [10]. Despite the success of this novel approach to grinding, the mechanism of EDG has not been studied in detail. Enikolopyan and his co-workers hypothesized that so-called “rheological explosion” took place during EDG; that is, there was an instantaneous destruction of the solid material exposed to high pressure and shear. By varying the temperature, it was possible to affect the extent to which solids deform. They concluded that the fracture occurred at the temperature near melting (crystallization) and that the grinding rate, energy consumption, and particle size depended significantly on the molecular weight and molecular weight distribution of a given polymer. They further postulated that with a decrease of low molecular mass fractions, the size of the particles increases. It is known that a fracture usually originates from local concentration of stress; failure occurs where the concentration of stress is the highest; when a polymer undergoes fracture, the creation of a new surface should occur [1]. Akopyan et al. described experiments in which varying the temperature made it possible to select the conditions under which the deformability of a polymer is minimal and, therefore, the energy of failure is decreased [11]. They designed a continuously operated laboratory device called a “rotary grinder” for practical implementation of the EDG. This equipment had heating elements to allow polymeric materials to reach a state of viscous flow followed by compression, shearing, and solidification upon cooling; destruction of the material resulted in the formation of a finely divided powder. They pulverized LDPE by maintaining the temperature above 120◦ C (which is above the melting point of LDPE of 111◦ C) with subsequent cooling of the powder formed in the grinding zone. Enikolopyan and his co-workers studied the uniaxial deformation of EDGmade powder from LDPE with various melt flow rates (MFR) at the temperature range of 20–90◦ C [12]. They found that the increase of elongation at break near the melting temperature depends on the amount of the low molecular weight fraction in the polymer. Enikolopyan assumed that the character of the rupture of LDPE is determined by its structural inhomogeneity caused by fracture during crystallization. They explained this behavior by noting that, unlike HDPE and PP, LDPE is characterized by a very broad molecular weight distribution, considerable branching, and the presence of a large amount of low
©2001 CRC Press LLC
P1: FUY pb035-03
March 13, 2001
9:56
Char Count= 0
molecular weight fraction in the amorphous phase. These attributes caused the crystallinity of LDPE to be lower than that of linear poly(ethylene) (it is known that the crystallites exhibited more defects). Knunyants and his co-workers investigated the role of crystallization and melting on defect formation [13]. They observed that crystallization coupled with viscoelastic effects caused by shearing strains is a decisive factor in multiple fracture of the polymers. They showed that, in contrast to many polymers including HDPE, LDPE has a significantly higher capacity to accumulate high elastic deformation and exhibits an anomalous decrease in deformability at temperatures close to its melting point of 108◦ C. Enikolopyan’s group also studied the structure and morphology of EDGmade powders derived from several mixtures such as LDPE/PP, HDPE/PP, and LDPE/HDPE utilizing a rotary grinder [14]. The results of X-ray structural analysis and radiothermoluminescence analysis indicate that powders made from PP/LDPE mixtures had enhanced crystallinity, and the amorphous regions of PP showed numerous microcracks. They observed elongated particles with a large number of microchannels and micropores in PP-rich mixtures with more than 40 weight percent of PP. Small quantities of coarse particles with smooth surfaces were present as well. More elongated fibrous particles have also been found in HDPE-rich mixtures. Enikolopyan and his co-workers described the disintegration of polymers into a highly dispersed powder that occurred in a comparatively narrow zone of the rotary grinder or the modified extruder’s transport channels rather than only in the zones of contact between the material and the channel wall [15]. They emphasized that one of the most important factors determining the production capacity of their EDG equipment was the extent of heat removal from the material being processed. Another observation made by Enikolopyan was that the work of rupture decreased markedly at elevated temperatures, which caused the specific energy consumption to be reduced by factors of 3 to 10 when compared with conventional grinding. Both the low noise level and low concentration of dust were also noted. Overall, under a certain combination of pressure, shear, and elevated temperature, polymers have been converted into powder without any dispersing additives; the specific energy consumption was at least 30 percent less than that used by conventional grinding [15]. Enikolopyan’s group also investigated the relationship between the mean particle size of powder, the polymer’s molecular mass distribution, and its MFR. In the case of LDPE, the mean particle size of the powder generally increased with a decrease in MFR. They came to the conclusion that some thermoplastics such as PS, PP, HDPE, polycarbonate (PC), and acrylonitrile butadiene styrene (ABS) are ground with considerably poorer results than LDPE or do not undergo EDG at all. Interestingly, Enikolopyan also observed that during pulverization of the PP/LDPE mixture with a PP content greater that 50 weight percent, a marked increase in the particle size was noted [15]. Therefore, the conclusion was reached
©2001 CRC Press LLC
P1: FUY pb035-03
March 13, 2001
9:56
Char Count= 0
that co-comminution of mixtures was due to a high degree of homogenization during the EDG process, which would not be achievable using conventional mixing equipment. By using the rotary grinder, the PP/HDPE mixture containing 30 weight percent of HDPE was converted into fine powder within the temperature range of 50–140◦ C. Specific energy consumption was minimal at the 30 weight percent of HDPE in the mixture. A similar relationship between the particle size and mixture’s composition was observed for PP/PS and PS/HDPE systems [15]. Friedman et al. and Enikolopyan et al. studied the effect of a controlled level of mechanical energy imposed during EDG on physical properties of polymers due to micro- and macrostresses [16,17]. They showed that the processing of dissimilar polymers under combined high pressure and shear deformation offered a new way of producing materials with specific properties by grafting growing macromolecules on the polymer matrix. Amorphization and change in the crystallinity as well as a significant decrease in the molecular weight of PP have been observed. Mixtures of some of the thermodynamically incompatible polymers under high pressure and shear formed a homogeneous system in the amorphous phase. Manevitch et al. investigated the essential role of dissipation of elastic energy accumulated during deformation [18]. They postulated that immediately after the breakage of polymer chains, the elastic energy stored in the macromolecule during its stretching is dissipated as follows: some is spent on the fragmentation of chains, some leads to the excitation of electrons, and some is transformed into heat. Yerina et al. investigated EDG of isotactic PP with MFR of 1.8 g/10 min pulverized with a modified single-screw rotary grinder [19]. The polymer was heated in the first zone to 100◦ C and to 230◦ C in the second zone to determine the effect of heat history on the degree of powder dispersion. They concluded that the PP was not completely fractured, confirming previous observations that rapid grinding of partially crystalline polymers, including PP, begins at a temperature close to that of the maximum crystallization rate. During crystallization under shear conditions, fractures occurred at the weakest areas, specifically at the interface and at the crystal’s defects. It has been shown that the multiple fracture of polypropylene occurred in two steps [19]. After an oriented melt was formed, it fractured simultaneously with orientational cystallization, resulting in the formation of a finely dispersed powder. Karadzhev et al. conducted further research on strain-assisted grinding, then referred to as elastic-strain powdering (ESP) [20]. They subjected LDPE to a radiation cross-linking at doses up to 1 MGy. Cross-linked LDPE was pulverized in a rotary grinder by preheating it to 150◦ C followed by cooling during compression and shear [20]. Scanning electron microscopy (SEM) of powders made from cross-linked LDPE (60 percent gel fraction) revealed a complicated, highly extended shape of the particles. Unlike non-irradiated LDPE, where the
©2001 CRC Press LLC
P1: FUY pb035-03
March 13, 2001
9:56
Char Count= 0
particle size distribution ranged from ultrafine to 50 microns (270 mesh), the powder made from cross-linked LDPE exhibited wide particle size distribution up to 89 microns (170 mesh). The formation of large particles was due to the fact that stresses created in the polymer under mechanical deformation level off as the network appears. This was equivalent to a decrease in the density of microdefects where the fracture was initiated. Consequently, the fracture resulted in larger particles for irradiated LDPE than for non-irradiated LDPE. Karadzhev concluded that ESP of cross-linked LDPE can follow two mechanisms: it can behave either as a cross-linked elastomer or as a crystalline polymer, depending on the pulverization temperature. Karadzhev and his group also studied a blend of highly-dispersed crosslinked LDPE with non-irradiated LDPE made with ESP under compression and shear in comparison with the same blend that was melt mixed at 150◦ C [21]. The ESP-made sample was obtained at crystallization temperature of LDPE as described by Enikolopyan and colleagues [15]. They observed that the dispersion of particles of cross-linked LDPE in the matrix of non-irradiated LDPE appeared to be similar to that of the cross-linked LDPE. The threedimensional network was retained after the ESP process; the particle shape was dependent upon the radiation dose. The absorbed particles exhibiting smooth surfaces were exposed to the higher radiation dose (1 MGy), in contrast to a more complex dendritic structure with numerous fibrous segments of the sample exposed to the lower radiation dose (400 kGy). Karadzhev noted that a blend of high concentrations of cross-linked LDPE (400 kGy) and non-irradiated LDPE showed some increase in tensile strength and elongation, suggesting strong adhesion between two phases [21]. He concluded that by using highly dispersed cross-linked LDPE with the three-dimensional network retained during ESP, it was possible to prepare blends with a predetermined structure of the dispersed phase. Kuptsov et al. studied X-ray scattering of some polyolefins subjected to ESP, specifically isotactic PP and HDPE [22]. PP powder showed wide particle size distribution that was dependent upon the melt temperature during ESP (it was above the crystallization temperature). PP powder annealed at 165◦ C exhibited two melting peaks at 165◦ C and 182◦ C rather than the single peak observed at 172◦ C for a feedstock PP. It was explained that the 182◦ C peak may be associated with the removal of defects and the growth of some of the crystallites during annealing [22]. The texture of the powder particles showed that the orientation of crystallites differed from that of the same polymer samples obtained by melt extrusion. Subsequent annealing of the PP near the melting point recovered their structure due to the melting of fine crystallites. Unlike PP, no fine powders could be made from HDPE. The HDPE particles looked like short, chaotically entangled fibers. Similar to PP, film of ESP-made HDPE showed a melting peak at 140◦ C, which was 5◦ C higher than that of HDPE pellets (melting peak was observed at 135◦ C).
©2001 CRC Press LLC
P1: FUY pb035-03
March 13, 2001
9:56
Char Count= 0
Khait and Petrich of Northwestern University (Evanston, Illinois) further advanced solid-state pulverization technology both for powder formation and for plastics recycling by using a modified co-rotating twin-screw extruder made by Berstorff Maschinenbau GmbH in Hannover, Germany [23] [Photo 2(a) on the color insert shows Dr. Khait with a Berstorff machine]. That process, also referred to in the scientific literature of the time as Solid-State Shear Extrusion (SSSE), was used to convert ordinarily incompatible commingled (unsorted) multicolor plastics, such as HDPE, LDPE, and PP, into somewhat reactive powders of different particle size for various applications that required powder as a feedstock, such as rotomolding or powder coating. They observed via electron spin resonance (ESR) experiments that the SSSE process causes the rupture of chemical bonds and initiation of free radical reactions, which in turn leads to insitu compatibilization of dissimilar polymers found in the post-consumer waste stream. Although the post-consumer plastics included an array of colors, the SSSE powder had unusually light homogeneous colors. The SSSE powders had hues defined by the predominant color in the particular feedstock. By contrast, the melt-processed multicolor recycled plastics had a streaked or marbleized surface appearance. The ability of the SSSE process to yield homogeneously colored powder allows for products made for re-use to be more attractive and more valuable than melt-processed recycled plastics. Wolfson and Nikol’skii reviewed investigations of elastic strain-assisted grinding of various solid polymers under combined action of shear stress, pressure, and elevated temperature that were conducted since the early 1980s at the Institute of Chemical Physics (at the Institute of Synthetic Polymeric Materials of the Academy of Sciences in Moscow) and the Norplast Research and Production Association in Moscow under the direction of Enikolopyan [10,24]. Extensive studies of the plastic flow of polymeric materials under the abovementioned conditions showed that flow is accompanied by fracture that leads to the formation of fine powders. This phenomenon was initially observed for LDPE of varying molecular mass, and it was suggested that the embrittlement of LDPE at elevated temperatures was related to a rate of melt crystallization, molecular mass, and the inhomogeneity of LDPE. Initial experiments by Chebotarevskii et al. were carried out with a Banbury mixer with a cooled jacket; LDPE powder was made at 15◦ C higher than the melting temperature of LDPE of 140◦ C [25]. The broad particle-size distribution ranging from 100 microns to 50 microns (140 mesh to 35 mesh) was explained by the non-uniform temperature control of LDPE crystallization. In a review paper by Wolfson and Nikol’skii, it was stated that at a critical combination of the above-mentioned parameters, polymers undergo multiple microcracks followed by the formation of finely divided powders [10]. Using modified extruders with windows for observation, they found that pulverization of LDPE takes place at the beginning of the crystallization of the melt and the grinding occurs in one step, without the gradual reduction in particle sizes that occurs with traditional grinding. In addition to polyolefins and
©2001 CRC Press LLC
P1: FUY pb035-03
March 13, 2001
9:56
Char Count= 0
their mixtures, they also obtained fine powders of a copolymer of ethylene with vinylacetate and some polyamides, but they could not pulverize all of the highly crystalline or amorphous polymers. Better temperature control was achieved by Enikolopyan et al. with the Brabender Plastograph; this resulted in LDPE powder with narrow particle size distribution of 25–50 microns (500–270 mesh) [5,26]. Later attempts were made to develop a continuous grinding process using the co-rotating twin-screw Berstorff extruder and the rotary grinder based on a single-screw extruder; in the latter, the gap between the grinding rotor and the cylinder could be varied to obtain narrow particle size distribution [11,27,28]. It was observed through a special window that fine grinding of LDPE occurred during cooling of the melt or after heating of the feedstock material at the optimum temperature; particle size remained unchanged during several passes through the device [11,29]. Based on extensive experimentation by Enikolopyan and his co-workers with grinding of LDPE, it was proposed that semi-crystalline polymers would exhibit a pseudo-brittle behavior during shear deformation at their melting temperature. In contrast with LDPE, crystalline polymers, such as HDPE and isotactic PP, exhibited increase in deformability at melting temperatures, and only coarse, fibrous particles have been produced [29]. They hypothesized that the mechanism of multiple breakage of polymers during shear deformation depends on both the heterogeneity of polymers due to crystallites distributed in an amorphous matrix and on crazing. According to Wolfson, no powder was made from amorphous polymers via elastic strain-assisted grinding at their Tg [24]. A number of publications have been devoted to high-temperature grinding of polymer mixtures including LDPE/PP, LDPE/HDPE, LDPE/PS, LDPE/PC, and others [8,14,19]. It has been noticed that even a small amount of LDPE in the mixture facilitates powder formation. With an increased amount of LDPE, particle size of the resultant powder approaches that of pure LDPE. Enikolopyan hypothesized that the EDG of two or more polymers involving the sequential crystallization of individual components would permit obtaining powders with a complex morphology. Wolfson and Nikol’skii listed vulcanized rubber, synthetic leather, linoleum, and other materials that were successfully ground using EDG [10]. They found that rubber grinding required lower shear stress than that of thermoplastics; additionally, because no melting of rubber occurred, the process was less sensitive to scale-up of the equipment from laboratory-size to production-size. It was noted that the surface area of rubber powder was greater than that of cryogenically ground rubber. Enikolopyan and his group relied on Griffith’s theory of failure to explain the mechanism of extrusion-pulverization. Failure of solids, according to Griffith, occurred spontaneously after maximum tension was achieved due to the defects in polymer crystalline structure, such as microcracks, which weaken the polymer [30]. When solids are in tension, those microcracks grow very fast, causing
©2001 CRC Press LLC
P1: FUY pb035-03
March 13, 2001
9:56
Char Count= 0
fracture. Askadskii emphasized that the mobility of macromolecules played a very important role in deformation processes [31]. He stated that a presence of amorphous regions in any crystalline polymer is usually considered a defect. However, amorphous regions play a positive role due to an increased strain in the regions under deformation; during fracture, macromolecules usually ruptured in the main chain. In addition to the grinding of polymers under shear deformation, the SSSE process can be employed for homogenization of polymer blends during mixing and modification of polymers via so-called “mechanochemistry.” This term is not well known because it was used mainly in the former Soviet Union and Eastern Europe. Initially introduced by Baramboim, mechanochemistry is a branch of polymer chemistry that deals with the effects of the mechanical energy on chemical changes and alterations of polymers [32]. Baramboim concluded that the site of reaction in the stressed state is determined by chain length and polymer structure. Stress concentration occurred at the points on the main chain connecting long branches and at the cross-link points in networks. Baramboim stated that the rupture of the molecular chains of polymers at a site where mechanical forces are being applied takes place when stress exceeds the critical value for the covalent bond between the atoms at that site. He further concluded that during shearing of crystalline polymers, destruction of the crystalline regions occurs. The Romanian researchers Simionescu and Oprea defined “mechanochemistry” more broadly [33]. In their view, this is a new field of science that combines polymer chemistry with other branches of science and engineering including mechanics and physics (especially solid-state physics). They noted that the increase in reactivity of the polymers during rupture can be accounted for by the appearance of free radicals, which they defined as “mechanoactivation.” As a result, new functional groups can be formed as indicated for infrared spectroscopy [34]. In a comprehensive review by Casale and Porter, “mechanochemistry” is used in relation to reactions in polymers that are induced by stress [35]. Data on the effect of shear on polymer behavior are usually reported as a change in molecular weight and viscosity; shear rate is controlled by the design of the equipment used in experiments. They noted that, in specific cases, solid-state processes can be used to alter polymer properties, such as molecular weight and molecular weight distribution, formation of new functional groups, morphology, and others. One effect of applying stress to a polymer is chain rupture, which is then followed by structural, chemical, physical, and rheological changes. Knowledge of the effect of shear as a prime variable is essential in learning solid-state mechanochemistry. Other variables, such as temperature, polymer composition, and residence time, also influence the impact of mechanochemistry on polymers’ behavior. Polymer composition and chain configuration are determining factors for which bonds will be ruptured. Weak bonds can result
©2001 CRC Press LLC
P1: FUY pb035-03
March 13, 2001
9:56
Char Count= 0
from inhomogeneities, adjacent double bonds, or from the presence of the catalyst or chain transfer agents [35]. The generation of macroradicals induced by stress can be measured by ESR, which can detect the kind of radicals, their concentration, and their lifetime. Other evidence of the formation of free radicals can be tested by such methods as the determination of new end groups by infrared spectroscopy, change in molecular weight by gel permeation chromatography (GPC), and change in melt viscosity. Quantitative characterization of branching block copolymers or graft copolymers is difficult because only a small fraction of the chemical bonds are involved in mechanochemical processes [35]. Enikolopyan and Friedman reviewed existing grinding practices emphasizing that conventional grinding consumes a large amount of energy [36]. They noted that during size reduction of materials using various grinding mills, efficiency is very low (about 1 percent) because of the rapid generation of frictional heat, which has to be removed. Cryogenic grinding requires the use of cooling agents, such as liquid nitrogen or Freon® , which are expensive. Moreover, grinding equipment that utilizes knives for cutting produces powders with low surface area, and the knives need frequent sharpening. Numerous laboratory and pilot-scale experiments conducted by Enikolopyan and his co-workers showed that the EDG process involved spontaneous destruction of polymers that are subjected to simultaneous effects of high compression, shear, and torsion. Their hypothesis has been confirmed by measuring molecular mass of varying fractions of the LDPE powder. Prut proposed a mechanism of grinding under elastic-strain conditions for different polymeric materials based on plastic flow instability under complex loading [37]. The energy accumulated in the material under pressure and shear is spent in the formation of a new surface. During grinding under elastic-strain conditions, a highly oriented melt is formed, and a crystallization of this melt proceeds by multimolecular nucleation. At increased temperatures, melting and crystallization affect the formation of defects near Tg . The thermal history of the melt affects the particle size of the powder and the throughput of the process. At a higher melt temperature, the mean particle size increased, the particle size distribution became wider, and the throughput of the grinding process was lower. The size of the powder also depends on the molecular mass of the polymer; for a higher molecular mass, the mean particle size increases. In the case of elastomer grinding, particles undergo a gradual decrease in mean size. Prut postulated at least two concurrent processes: fracture and agglomeration [37]. Which process becomes dominant is determined by the relationship between the initial size of the particles and the length of the grinding zone. During fracture, the formation of relatively small particles takes place, which is unrelated to the viscoelastic flow and is insensitive to the deformation rate. In Prut’s opinion, the formation of small particles appears to be dependent on friction between particles, on friction of the particles against the wall of the
©2001 CRC Press LLC
P1: FUY pb035-03
March 13, 2001
9:56
Char Count= 0
grinding device, and on removal of the small particles from the surface of the crushed crumb. Over the years, a variety of equipment has been developed to carry out strain-assisted grinding of polymers. Chebotarevskii et al. described an intensive, batch-type Banbury mixer in which powder was made by mixing a molten polymeric material followed by cooling to 1–40◦ C below its melting temperature [38]. They postulated that hardening of the polymer is accompanied by fracture, which resulted in the formation of powder with an average particle size of 100–500 microns (140–35 mesh). Specific energy consumption was reported as 60–65 kWh/ton for the pulverization of HDPE, PP, and plasticized PVC. Enikolopyan and his group developed EDG as a continuous process employing single- or twin-screw extruders by compressing a rubber feedstock, including vulcanized rubber, by a force ranging from 0.2 to 0.7 MPa followed by a simultaneous application of a pressure within a range of 0.2 to 50 MPa and a shear force ranging from 0.03 to 5 N/mm2 [39]. The feedstock was heated to a temperature between 80–250◦ C with subsequent cooling down to a temperature ranging from 15◦ to 60◦ C. They established that the heating/cooling cycle can be repetitive, based upon the properties of the feedstock and the desired particle size of powder. Enikolopyan also stated that some additives (e.g., poly[ethylene]) improved the efficiency of the EDG process for obtaining powder of 300–500 microns. Energy consumption during the powder production from vulcanized rubber was reported as 200–400 kWh/ton of the product compared to that of 1,500 kWh/ton during cryogenic grinding. Enikolopyan et al. investigated the EDG process for pulverizing polymers using an apparatus containing a hollow barrel with a rotating screw(s) in which the feedstock is fused to a temperature of 140–150◦ C and conveyed into the second zone of the barrel, where it is then cooled to 20◦ C, pre-crushed, pulverized at a presure of 0.25–0.3 MPa, and discharged. They used a 53-mm diameter twin-screw extruder with 28 L/D and achieved 40–60 kg/hr throughput for the powder containing only 2 percent of a 160-micron product. Further modifications to the pulverization of rubber via the EDG process were accomplished by Mayer and Freist in Germany by arranging a separate cooling unit outside the modified co-rotating Berstorff twin-screw extruder followed by a return of the cooled product to the extruder for a second pass and discharging the rubber powder [40]. They found that the pulverized rubber could be cooled down to 20–80◦ C more efficiently outside the extruder, thus avoiding an agglomeration of fine fractions of the powder. The separate cooling unit was designed as an oscillating, conveying spiral in which discharged powder was cooled by both the surrounding air and water circulated throughout the unit. Khait studied the morphology of used-tire rubber produced on pilot-scale PT-40 (40-mm diameter) and production-scale PT-90 (90-mm diameter) co-rotating twin-screw Berstorff pulverizers. It has been shown by SEM images that the tire rubber powder obtained by pulverization with Berstorff pulverizers
©2001 CRC Press LLC
P1: FUY pb035-03
March 13, 2001
9:56
Char Count= 0
has a unique “cauliflower“ shape with a large surface area as opposed to the flat, angular particles with smaller surface area that are obtained during conventional grinding of tire rubber utilizing hammer mills or other grinding devices [41]. At the Illinois Institute of Technology (IIT; Chicago, Illinois), a group of researchers under the general direction of Arastoopour and Shutov conducted several studies related to pulverization of used-tire rubber, waste polyurethane foam, and other plastics with modified extrusion devices on a small scale. Shutov et al. utilized the Solid-State Shear Extrusion (SSSE) process to pulverize thermosets, cross-linked thermoplastics, and natural polymers with a laboratory-scale counter-rotating conical twin-screw Brabender extruder [42]. They reported that SSSE was carried out under near-ambient temperature, with applied torque of 40 kgm and a pressure up to 150 psi. The resultant powder had a majority of particles in the 100–900 micron range. These researchers found that it was necessary to introduce gas to fluidize the powder at the discharge to prevent agglomeration of the finer particles. Arastoopour studied a process and single-screw apparatus for SSSE pulverization of polymeric materials, which were heated to form a continuous thin film, followed by cooling [43]. In this process, normal and shear forces were applied to polymers to make powder in a single-screw extruder using either a cylindrical or a conical screw. Arastoopour found, however, that the resultant powder must be fluidized upon its discharge to break up agglomerates, resembling the observation by Shutov et al. who used a modified conical twin-screw device [44,45]. At IIT in Chicago, Ivanov continued the development of equipment for the solid-state shear extrusion pulverization of polymers [46]. His apparatus is composed of an elongated hollow barrel with one or two screws having several zones including a high-shear zone for powder formation. Barrel heating and cooling allowed the polymeric material to be heated to a temperature below its melting point and below its decomposition temperature. This apparatus design includes a rotated segment at the end of the screw to prevent agglomeration of fine powder. At the Tennessee Technological University (Cookeville, Tennessee), Shutov described yet another apparatus for shear pulverization of polymeric materials into small particles using a conventional extruder and a special pulverizer head with a rotor that has a conical contact surface and a stationary dish with a corresponding inverted conical contact surface [47]. The polymer is conveyed from an extruder into the gap between the dish and the rotor. Rotation of the rotor generates shear forces within the gap that pulverizes the material. Khait developed non-melting Solid-State Shear Pulverization (S3 P) as a method of making polymeric particulates from polymeric scrap, virgin materials, and mixtures of dissimilar polymers, which are supplied to a co-rotating twin-screw Berstorff pulverizer [48]. The resulting powders are meltprocessable by all existing plastics-fabrication techniques. It has been shown
©2001 CRC Press LLC
P1: FUY pb035-03
March 13, 2001
9:56
Char Count= 0
that, despite the multiplicity of colors in the feedstock material, the powders were of a homogeneous light color. It has also been reported that subjecting polymer mixtures to shear results in chain rupture and consequently leads to in-situ compatibilization of dissimilar polymers via mechanochemistry [48]. Shutov et al. described the SSSE process for pulverization of plastics, rubber, cellulosics, and natural polymers (wood, corn) at both room temperature and elevated temperature (100–180◦ C) to make fine powders of about 50 microns (270 mesh) [49,50]. Energy consumption of the process was estimated as 200– 500 kWh per ton of rubber to produce powder of 100 microns (140 mesh). Enayati et al. presented results on applying SSSE process to recover waste of either low-density flexible, highly-resilient foam or rigid polyurethane foam [51]. They found that particle size depends on the rigidity of the polyurethane matrix; the smallest size of 17–32 microns (635–400 mesh) was obtained from rigid foam while a larger size of 96–170 microns (140–80 mesh) was made from flexible foam. Arastoopour et al. in a related study on recovery of polyurethane foam waste by SSSE revealed that the pulverization process was carried out at 20–100◦ C, which did not alter the chemical structure of the foam [52]. Riahi et al. studied the particle size distribution and shape of used-tire rubber powder made via the SSSE process using a small-scale Brabender extruder [53]. They indicated that the SSSE process utilized low pressure and temperature close to the melting point of the polymer. During the process, a thin film was formed over the screw flights followed by intensive cooling causing disintegration of that film into a powder. They found that, by increasing poly(ethylene) content up to 5 weight percent, powder particles became rounder than tire-rubber particles pulverized alone. Although they postulated that poly(ethylene) acted as a “catalyst” for pulverization of rubber, no evidence for that hypothesis has been reported. Shutov compared the SSSE process for blending LDPE and sawdust with the conventional melt-mixing process [54]. He showed that the Young modulus of SSSE-made samples was higher than that of the melt-mixed materials due to a more uniform distribution of wood powder in the poly(ethylene) matrix. The resulting powder of about 50 microns (270 mesh) was 20 times smaller in size than that of sawdust. Venkatasanthanam et al. demonstrated the utility of the SSSE process to produce fine powder from highly resilient polyurethane foam waste from used car seats [55]. The powder was used as a co-ingredient at 30 percent level with virgin materials to produce semi-rigid polyurethane foam. An improvement in mechanical properties was observed, suggesting that SSSE-made powder functioned as a reinforcing agent. The SEM of the foam showed that most of the powder particles were in the struts of the cells rather than in the cell walls. They proposed that this structure was the main reason for the increase in the properties of foams. Bilgili et al. used a laboratory-scale, modified single-screw conical Brabender extruder to achieve high shear and normal forces in order to pulverize
©2001 CRC Press LLC
P1: FUY pb035-03
March 13, 2001
9:56
Char Count= 0
used-tire rubber via the SSSE process employing multiple passes through the extruder [56]. These investigators observed agglomeration of fine particles due to bridging and insufficient removal of the frictional heat. The authors proposed a mechanism of pulverization of elastomeric materials under high shear and normal forces by a creation of thin film. They believed that excessive shear stresses imposed on elastomer initiated the fracture and the compressive forces hold the fractured thin polymeric film due to friction. The energy stored in the elastomer during deformation reaches a threshold and causes powder to be formed. Patel and Shutov described yet another non-extrusion, mechanical process to make polymeric powders called pressure shear pulverization (PSP) [57]. Like SSSE, this process utilizes the combined action of pressure and shear stresses, but uses a special pulverization head. They obtained 275-micron (50 mesh) powder from virgin LDPE (MFR 55 g/10 min) of irregular shape and high surface area. Powder has also been made from waste cross-linked LDPE foam and mixture of LDPE with paper waste with particle sizes between 319 and 389 microns (from 50 to 45 mesh). No change in the chemical structure or crystallinity of LDPE has been observed based on thermal analysis and DSC of virgin and pulverized LDPE; unlike the SSSE process, MFR of PSP-made LDPE powder remained unchanged. These data were suggestive of the fact that primary bonds did not render any cleavage during the PSP process. In his review of emerging recycling technologies, Scheirs described SSSE as a novel, continuous process for recycling plastic waste utilizing a modified co-rotating twin-screw extruder made by Berstorff in Germany [58,59]. The highest shear forces occur in the solid bed during compaction while frictional heat is removed, causing formation of the powder discharged without a die [60]. This technology has been developed by Khait and her co-workers at the Polymer Technology Center at Northwestern University in Evanston, Illinois, US. This process has been termed “Solid-State Shear Pulverization” (S3 P) to differentiate it from the SSSE high-temperature extrusion-grinding process utilized by Nikol’skii and his co-workers at the Academy of Sciences in Moscow, where polymers are melted prior to pulverization. Subsequently, the term “extruder” was dropped in favor of “pulverizer,” and the machine designation became PT (pulverizer twin)-25. (The figure after the hyphen indicates the screw diameter.) Ahn et al. studied microstructural changes in homopolymers and polymer blends induced by elastic strain pulverization (ESP) [60]. They showed that processing of polymers by ESP caused significant changes to their morphology and microstructure. The ESR data at ambient temperature for LDPE, PP, and their blends suggested a significant increase in free radicals generated after pulverization as a result of the rupture of main-chain carbon-carbon bonds by the use of mechanical forces. Also, a drastic reduction in spherulite size after pulverization was observed. Overall, both ESR and differential scanning calorimetry (DSC) data supported the notion that physical and chemical changes
©2001 CRC Press LLC
P1: FUY pb035-03
March 13, 2001
9:56
Char Count= 0
occurred in the chain structure of PP caused by ESP. DSC and polarized light microscopy were applied to homopolymer PP, blends of HDPE/PP, and blends of HDPE/PS, and it was learned that crystalline structure is drastically altered by solid-state pulverization. The size-reduction technologies developed toward the end of the twentieth century in Russia and the US centered on the concept of fracture in polymers that is induced by elastic strain. According to Kausch, fracture implies the development of a crack growth and propagation to completion of the fracture and is understood as stress-based disintegration [61]. In the absence of flow, chain scission and chain strength determine the mechanical properties of polymers. The mechanical loading and rupture of polymers leads to a chain scission and the formation of free radicals, which in turn causes mechanochemical reactions to occur. The increasing use of polymers in very demanding engineering applications highlights polymer stress-and-strain as a key issue. It is well-known that the stress-and-strain behavior of polymers strongly depends on their homogeneity, temperature, crystalline structure, molecular weight, and molecular weight distribution. A fracture involves the creation of a new surface when a polymer is subjected to the application of external forces. Polymers can undergo fracture in many different ways in accordance with their strain-stress behavior. Glassy polymers are more brittle and fail at relatively low strains. Semi-crystalline polymers break at very high stresses. In general, failure will occur through a combination of molecular fracture and the slippage of molecules for a given polymer depending on its structure and morphology [62]. Nesarikar et al. reported self-compatibilization of polymer blends via mechanochemistry during S3 P process [63]. They proposed that during pulverization, polymer chains ruptured and then recombined, forming block copolymers; if the chain transfer occurs, then graft copolymers are formed. Nuclear magnetic resonance (NMR) data revealed the formation of long chain branching in linear low-density poly(ethylene) (LLDPE), which is equivalent to the presence of graft copolymers in blends. It is proposed that these block or graft copolymers serve as compatibilizing agents during pulverization of ordinarily immiscible polymers. Furgiuele et al. further investigated the utility of the S3 P process for polymer blends and polymeric waste compatibilization [64]. They demonstrated changes in the molecular weight distributions in the PS portion of the PP/PS blend, which suggests that chain scission of PP occurs predominately in the amorphous regions. A decrease in Tg indicated that a more intimate mixing of PS and PP was achieved than in conventional melt-mixing. These data confirmed previous studies by Khait and Petrich of the generation of a free radical population as a result of chain scission during the S3 P process [60]. Khait and Torkelson reviewed S3 P as a novel “green” recycling technology for plastics waste recovery [65]. In S3 P, plastics are processed at temperatures below
©2001 CRC Press LLC
P1: FUY pb035-03
March 13, 2001
9:56
Char Count= 0
their melting points (in the case of crystalline or semicrystalline polymers) or glass transition temperature (in the case of amorphous polymers). The highly crystalline polymer, HDPE, was successfully converted into powder using solidstate shear pulverization with a laboratory-scale PT-25 Berstorff pulverizer at Northwestern University’s Polymer Technology Center. Several ordinarily immiscible polymer mixtures containing HDPE, LLDPE, PP, PS, poly(ethylene terephthalate) (PET), and even polyvinyl chloride (PVC) have been recycled into useful materials. The fact that S3 P processing results in uniform meltprocessable powder containing both PET and PVC is an important advance in recycling technology, because it is known that PET waste contaminated with PVC cannot be melt-processed by conventional methods due to the high melting temperature of PET; in such cases, PVC degrades at temperatures well below the melt temperature of PET. Furgiuele et al. have demonstrated, via characterization of Tg for the polystyrene-rich phase of PS/PP blends, that mixtures processed via the S3 P method achieve much more intimate mixing of components than is normally observed in conventional melt processing alone [66]. This effect, along with the known capacity of S3 P to result in chain scission and to produce reactive free radicals, has the potential for a self-compatibilization of ordinarily dissimilar blends via block copolymer formation at the phase interfaces. Furgiuele et al. have also shown that the S3 P process can achieve efficient mixing of polymer blends in which the component viscosities differ by as much as three orders of magnitude [67,68]. Blends of immiscible polymers (HDPE/PS) or of like polymers (low-viscosity poly[ethylene]/high-viscosity poly[ethylene]) of very different molecular weight often require very long processing times to achieve mixing in the melt or liquid state to undergo phase inversion. In contrast, the same blends subjected to S3 P prior to processing exhibited no phase inversion. Khait and Torkelson described the utility of the S3 P process for mixing like polymers with unmatched viscosity, which is not achievable by conventional melt mixing due to phase separation [69]. They found that when two virgin homopolymer polypropylenes with MFR 6.7 g/10 min and 37.8 g/10 min were mixed via the S3 P process, the resultant materials showed significant increase in elongation at break as compared to mixed samples made by conventional melt processing using a twin-screw extruder. Furgiuele et al. reported an elimination of phase inversion as a result of efficient mixing of polymer blends of extreme viscosity ratio [67]. With melt processing of poly(ethylene)/polystyrene blends of a <0.01 viscosity ratio, phase inversion times ranged from 11 to 34 minutes; however, pulverized blends showed no phase inversion upon melt mixing. Similar results were seen for blends of high and low molecular weight poly(ethylene). Arinstein et al. have continued development of disintegration of nitrile, isoprene, and butadiene-styrene rubber, as well as thermoplastics under intensive stress action such as compression and shear (ISAC & S) [70]. They
©2001 CRC Press LLC
P1: FUY pb035-03
March 13, 2001
9:56
Char Count= 0
developed a family of devices targeting multiple cracking of polymers at high temperature and proposed a theoretical model of ISAC & S.
SUMMARY The mechanochemistry of polymers as a new branch of science was introduced in the 1960s by Baramboim primarily as a result of research at the Academy of Sciences in Moscow. A second comprehensive review of mechanochemistry of polymers was written in 1967 by Simionescu and Oprea in Romania; this work emphasizes the mechanism of mechanochemical destruction and mechanochemical synthesis as well as practical applications. In 1978, Casale and Porter published their detailed review on the polymer stress reactions and the interrelation of mechanochemistry with the physical behavior of polymers. Original research by the Nobel Laureate P. W. Bridgman in 1935 showed sudden dispersion of solids under high pressure between two anvils. This phenomenon was further studied at the Institute of Chemical Physics of the Academy of Sciences in Moscow. Beginning in the early 1980s, Enikolopyan and his co-workers conceptualized and demonstrated a new grinding process for polymeric materials, which is known synonymously in technical literature as strain-assisted grinding, extrusion-grinding, Elastic Deformation Grinding (EDG), elastic strain powderization (ESP), and Solid-State Shear Extrusion (SSSE). The mechanism of the new grinding process is not yet fully understood, but data and observations of Enikolopyan and his group indicate that EDG is a solid-state mechanochemical process that is governed by a combination of pressure, shear strains, and thermoelastic stresses. Their study showed that this new grinding process was “avalanche-like” and the powder was formed simultaneously, unlike conventional grinding in which size reduction occurred in a successive mode. It has been shown that the fracture of polyolefins subjected to elastic strain-assisted grinding under certain processing conditions was dependent on elasticity of the polymer, molecular mass, and crystallinity among other factors. Building on Enikolopyan’s work, researchers at Northwestern University have developed the Solid-State Shear Pulverization (S3 P) process, focusing on polymer powder production, mechanochemical reactions, development of new polymer blends and composites, efficient mixing of polymers of extreme viscosity ratio, and polymer waste recovery. As a result of extensive research and development during the 1990s at Northwestern University in the US and by Berstorff in Germany and the US, the successful transition of S3 P technology and equipment from a pilot scale to a commercial scale was achieved through use of a co-rotating, fully intermeshing twin-screw Berstorff pulverizer. Khait and her co-workers demonstrated the
©2001 CRC Press LLC
P1: FUY pb035-03
March 13, 2001
9:56
Char Count= 0
utility of the S3 P process for producing fine powders from virgin resins, for mixing polymers with unmatched viscosity, and for recovering commingled (unsorted) multi-color plastics by producing homogeneous, light-colored powders with unique shapes and large surface areas. These powders are suitable for a variety of applications including rotational molding, powder coating, extrusion, and injection molding. The S3 P-made powders have a significant free radical population, indicative of rupture of carbon-carbon bonds during S3 P. Torkelson’s research group generated data indicating an in-situ compatibilization of ordinarily dissimilar polymers during the S3 P process, which in turn led to the development of new polymer blends without an addition of pre-made block or graft copolymers. Several research groups in the US are working on solid-state pulverization technology. An investigation of the SSSE process by researchers at the Illinois Institute of Technology under the direction of Arastoopour resulted in the modification of a small-scale, single-screw extruder to make powder from used-tire rubber, polyurethane foam, and other waste feedstocks. At Tennessee Technological University in Cookeville, Tennessee, Shutov and his co-workers produced polymeric powder by a pressure/shear process using a special device. Further development of the S3 P process and equipment is proceeding as a cooperative effort between Northwestern University’s Polymer Technology Center and Berstorff on both a pilot scale and on a commercial scale. Although still under intensive investigation, this process has already shown potential advantages over existing powder production methods involving cryogenic grinding. S3 P also provides for the intimate mixing of dissimilar polymers with very different viscosities and allows self-compatibilization of ordinarily immiscible polymer blends. Beyond this, S3 P has an added economic benefit because it eliminates the expensive step of sorting prior to processing multicolor commingled polymeric waste. When the transition to a commercial scale is complete, this Solid-State Shear Pulverization process is expected to alter significantly the development of new polymer blends and the business of recycling.
REFERENCES Note: Some names of persons (such as Enikolopyan, Yerina, Nepomnyaschy, and Akopyan) have variant spellings in English transliteration. While the chapter text uses a single spelling, the variations are preserved in this reference list. 1. Kinloch, A. J. and Young, R. J. Fracture Behavior of Polymers, New York: Applied Science Publishers (1983). 2. Faulkner, B. P. and Rimmer, H. W. Encyclopedia of Chemical Technology, 21:132–161 (1983). 3. Enikolopian, N. S., Akopian, E. L., and Nikol’skii, V. G. “Some Problems of Strength and
©2001 CRC Press LLC
P1: FUY pb035-03
March 13, 2001
4. 5.
6. 7. 8.
9:56
Char Count= 0
Fracture of Polymer Materials,” Makromolecular Chemie, Basel, Switzerland: Huthig & Wepf Verlag, Supplement No. 6, pp. 316–330 (1984). Enikolopian, N. S. “Physical Aspects of Plastic Flow,” Makromolecular Chemie, Basel, Switzerland: Huthig Wepf Verlag, Supplement No. 8, pp. 109–117 (1984). Enikolopyan, N. S., Akopyan, E. L., and Nikol’skii, V. G. “Anomalies in the Strength and Deformational Properties of Amorphous-Crystalline Polymers at High Temperatures,” London: Plenum Publishing Corp., pp. 805–807 (1983). Bridgman, P. W. The Physics of High Pressure, New York: The Macmillan Co. (1931). Enikolopian, N. S. “Some Aspects of Chemistry and Physics of Plastic Flow,” Pure and Applied Chemistry, 57:1707–1711 (1985). Enikolopian, N. S., Fridman, M. L., Karmilov, A. Yu., Vetsheva, A. S., and Fridman, B. M. “Elastic-Deformation Grinding of Mixtures of Thermoplastic Polymers,” London: Plenum Publishing Corp., pp. 834–837 (1988).
9. Enikolopian, N. S. and Fridman, M. L. “Mechanism of Elasto-Deformational Grinding of Polymeric Materials,” London: Plenum Publishing Corp., pp. 817–820 (1987). 10. Wolfson, S. A. and Nikol’skii, V. G. “Strain-Assisted Fracture and Grinding of Solid Polymeric Materials: Powder Technologies,” Polymer Science USSR—Series B, 36:861–874 (1994). 11. Akopyan, E. I., Karmilov, A. Yu, Nikol’skii, V. G., Khachatryan, A. M., and Enikolopyan, N. S. “Elastic Deformational Grinding of Thermoplasts,” London: Plenum Publishing Corp., pp. 971–973 (1987). 12. Yenikolopyan, N. S., Akopyan, Ye. L., Kechekyan, A. S., Nikol’skii, V. G., and Styrikovich, N. M. “High-Temperature Deformation of Low-Density Poly(ethylene): Effect of MolecularMass Distribution and Thermal Prehistory,” Polymer Science USSR, 26:2640–2648 (1984). 13. Knunyants, M. I., Dorfman, I. Ya., Kryuchkov, A. N., Prut, E. V., and Enikolopyan, N. S. “Multiple Failure of Low-Density Poly(ethylene) in Extrusion in the Region of a Phase Transition,” London: Plenum Publishing Corp., pp. 405–407 (1987). 14. Yenikolopyan, N. S., Khachatryan, A. M., Karmilov, A. Yu., Nikol’skii, V. G., Plate, I. V., Fedorova, Ye. A., and Filippov, V. V. “Structure and Morphology of Powdered Polymeric Materials Obtained by Elastic-Deformation Grinding,” Polymer Science USSR, 30:2569–2576 (1988). 15. Yenikolopyan, N. S., Akopyan, Ye. L., Karmilov, A. Yu., Nikol’skii, V. G., and Khachatryan, A. M. “Production of Highly Disperse Powder Materials Based of Thermoplastics and Thermoplastic Blends of Elastic-Deformation Grinding,” Polymer Science USSR, 30:2576–2584 (1988). 16. Friedman, M. L. and Prut, E. V. “Physical and Physiochemical Effects on the Rheological Properties of Polymers During Processing,” Russian Chemical Reviews, 53:186–196 (1984). 17. Enikolopian, N. S., Zarkhin, L. S., and Prut, E. V. “Primary Molecular Products of Mechanical Fracture of Polymers,” Journal of Applied Polymer Science, 30:2291–2295 (1985). 18. Manevitch, L. I., Zarkhin, L. S., and Enikolopian, N. S. “Nonlinear Dynamics and the Problem of Polymer Fracture,” Journal of Applied Polymer Science, 39:2245–2258 (1990). 19. Yerina, N. A., Potapov, V. V., Kompaniets, L. V., Knunyants, M. I., Prut, E. V., and Yenikolopyan, N. S. “Mechanisms of the Elastic-Strain Grinding of Isotactic Polypropylene,” Polymer Science USSR, 32(4):704–710 (1990). 20. Karadzhev, A. C., Gnezdilova, R. B., Gor’kov, D. A., Nikol’skii, V. G., Styrikovich, N. M., Filippov, V. V., Finkel, E. E., and Kechek’yan, A. S. “Elastic-Strain Powdering of Radiation-Cross-Linked Low-Density Poly(ethylene),” Polymer Science USSR, 34:673–677 (1992).
©2001 CRC Press LLC
P1: FUY pb035-03
March 13, 2001
9:56
Char Count= 0
21. Karadzhev, A. C., Gnezdilova, R. B., Gor’kov, D. A., Kechek’yan, A. S., Nikol’skii, V. G., Styrikovich, N. M., Filippov, V. V., and Finkel. E. E. “Structurally Organized Systems Based on Highly Dispersed Cross-Linked Low-Density Poly(ethylene),” Polymer Science USSR, 34:1063–1068 (1992). 22. Kuptsov, S. A., Erina, N. A., Minina, O. D., Prut, E V., and Antipov, E. M. “An X-ray Scattering Study of PO Subjected to Elastic-Strain Powdering,” Polymer Science USSR, 35:307–310 (1993). 23. Khait, K. and Petrich, M. A. Batelle Pacific Northwest Laboratories Report PNL-SA-22193, Vol. 1, Chap. 7 (1993). 24. Wolfson, S. A. and Nikol’skii, V. G. “Powder Extrusion: Fundamentals and Different Applications,” Polymer Engineering and Science, 37:1294–1300 (1997). 25. Chebotarevskii, A. E., Enikolopov, N. S., Nikol’skii, V. G., Miranov, N. A., Parchenkov, G. M., Kotov, J. M., Gabutdinov, M. S., and Gilim’yanov, F. G. USSR Inventor’s Certificate No. 1022735, Byuletin Izobretenii, No. 22, p. 15 (1983). 26. Enikolopyan, N. S. and Fridman, M. K. “Mechanisms of Elasto-Deformational Grinding of Polymeric Materials,” London: Plenum Publishing Corp., pp. 817–820 (1987). 27. Enikolopov, N. S., Wolfson, S. A., Nepomnjaschtschie, A. J., Nikol’skii, W. G., Teleschow, W. A., Filmakowa, L. A., Brinkmann, H., Pantzer, E., and Uhland, E. US Patent 4,607,797 (1986). 28. Enikolopov, N. S., Nikol’skii, V. G., Bratschnikov, E. M., Akopyan, E. I., Nepomnyatschii, A. I., Trubnikov, G. P., and Cherepnina, O. O. USSR Inventor’s Certificate No. 1120587 (1993). 29. Enikolopyan, N. S., Akopyan, E. L., Karmilov, A. Yu., Nikol’skii, V. G., and Khachatryan, A. M. “Highly Dispersed Thermoplastic Powders and Their Mixtures Made by Elastic Deformation Grinding,” Polymer Science USSR, 30:2576–2584 (1988). 30. Griffith, A. A. “The Phenomenon of Rupture and Flow in Solids,” Philosophical Transactions of The Royal Society, A221:163 (1920). 31. Askadskii, A. A. Deformation of Polymers, Moscow: Khimiya (1973). 32. Baramboim, N. K. “Mechanochemistry of Polymers,” Rubber and Plastic Research Association of Great Britain, London: MacLaren & Sons, Ltd. (1964). 33. Simionescu, C. S. and Oprea, C. V. “Mechanochemical Synthesis,” Russian Chemical Review, 57:283–297 (1988). 34. Simionescu, C. and Oprea, C. V. Mechanochemistry of Polymers, Moscow: Mir (1970). 35. Casale, A. and Porter, R. S. Polymer Stress Reactions, New York: Academic Press (1978). 36. Enikolopyan, N. S. and Friedman, M. L. Future of Science, Moscow: Znanie (1986). 37. Prut, E. V. “Plastic Flow Instability and Multiple Fracture (Grinding) of Polymers: A Review,” Polymer Science USSR, 36:493–497 (1994). 38. Chebotarevskii, A. E., Enikolopov, N. S., Nikol’skii, V. G., Mironov, N. A., Parchenkov, G. M., Kotov, J. M., Gabutdinov, M. S., and Gilim’yanov, I. G. USSR Certificate 1022735 (1983). 39. Enikolopov, N. S., Nepomnyaschy, A. I., Filmakova, L. A., Krasnokutsky, V. P., Kurakin, L. I., Akopian, E. L., Markarian, K. A., Negmatov, S. S., Martkarimov, S. K., Polivanov, Y. A., Sherstnev, P. P., and Pavlev, V. B. US Patent 4,607,796 (1986). 40. Mayer, D. and Freist, B. US Patent 5,273,419 (1993). 41. Khait, K. “New Used-Tire Recovery Process for Value-Added Products,” Proceedings of the Rubber Division Meeting of the American Chemical Society, Chicago (1994). 42. Shutov, F., Ivanov, G., and Arastoopour, H. US Patent 5,395,055(1995). 43. Arastoopour, H. US Patent 5,704,555 (1998).
©2001 CRC Press LLC
P1: FUY pb035-03
March 13, 2001
44. 45. 46. 47. 48. 49.
50.
51.
52.
53.
54.
55.
56.
57. 58. 59. 60.
61.
9:56
Char Count= 0
Shutov, F., Ivanov, G., and Arastoopour, H. US Patent 5,397,065 (1995). Shutov, F., Ivanov, G., and Arastoopour, H. US Patent 5,415,354 (1995). Ivanov, G. US Patent 5,743,471 (1998). Shutov, F. US Patent 5,769,335 (1998). Khait, K. US Patent 5,814,673 (1998). Shutov, F., Ivanov, G., and Wolfson, S. A. “New Principles of Thermoplastic Waste Recycling: Solid-State Shear Extrusion (SSSE) Technique,” Proceedings of the Fall Meeting of the American Institute of Chemical Engineers, Miami Beach, Florida (1992). Shutov, F., Ivanov, G., Arastoopour, H., and Wolfson, S. A. “New Principles of Plastic Waste Recycling: Solid-State Shear Extrusion,” Proceedings of the Fall Meeting of the American Chemical Society, Washington, D. C. (1992). Enayati, N., Riahi, A., Li, J., Topac, H., Arastoopour, H., Ivanov, G., and Shutov, F. “Solid-State Shear Extrusion (SSSE) for Pulverization of Flexible and Rigid Polyurethane Foam Wastes,” Proceedings of the Polyurethanes World Congress, Vancouver, pp. 103–106 (1993). Arastoopour, H., Ivanov, G., and Shutov, F. “Recycling of Polyurethane Foam Wastes Using New Pulverization Principle: Solid-State Shear Extrusion (SSSE),” Proceedings of the Second International Congress on Cellular Polymers, Edinburgh (1993). Riahi, A., Li, J., Arastoopour, H., Ivanov, G., and Shutov, F. “Recycling of Plastic Wastes Using Solid-State Shear Extrusion,” Proceedings of the Annual Technical Conference of the Society of Plastics Engineers, ANTEC ’93, New Orleans, pp. 891–892 (1993). Shutov, T. “Recycling of Plastic Wastes Using Solid-State Shear Extrusion,” Proceedings of the Annual Recycling Conference of the Society of Plastics Engineers, ARC ’93, Schaumburg, Illinois, p. 198–206 (1993). Venkatasanthanam, S., Ivanov, G., and Shutov, F. “Polyurethane Foams Reinforced with Recycled Polyurethane Foam Waste Powder,” Proceedings of the Annual Technical Conference of the Society of Plastics Engineers ANTEC ’95, Boston, pp. 3673–3684 (1995). Bilgili, E., Arastoopour, H., and Bernstein, B. “Pulverization under High, Normal and Shear Forces as a First Step Towards Recycling of Rubber,” Proceedings of the Annual Recycling Conference of the Society of Plastics Engineers, ARC ’98, Chicago (1998). Patel, T. and Shutov, F. “Pressure Shear Pulverization (PSP) Process: Properties of Waste and Virgin LDPE” (poster), Annual Recycling Conference, ARC ’98, Chicago (1998). Scheirs, J. Polymer Recycling, New York: John Wiley & Sons (1998). Mack, M. “Twin-Screw Machines Explore Solid-State Extrusion,” Plastics Technology, pp. 75–77 (1993). Ahn, D., Khait, K., and Petrich, M. A. “Microstructural Changes in Homopolymers and Polymer Blends Induced by Elastic Strain Pulverization,” Journal of Applied Polymer Science, 55:1431– 1440 (1995). Kausch, H. H. Polymer Fracture—2nd ed., Berlin: Springer Verlag (1987).
62. Young, R. J. and Lovell, P. A. Introduction to Polymers, London: Chapman and Hall (1991). 63. Nesarikar, A. R., Carr, S. H., Khait, K., and Mirabella, F. M. “Self-Compatibilization of Polymer Blends via Novel Solid-State Shear Extrusion Pulverization,” Journal of Applied Polymer Science, 63:1179–1187 (1997). 64. Furgiuele, N., Khait, K., and Torkelson, J. M. “The Use of Solid-State Shear Pulverization for Polymer Blend and Polymeric Waste Compatibilization,” Proceedings of the Annual Recycling Conference of the Society of Plastics Engineers, ARC ’98, Chicago (1998). 65. Khait, K. and Torkelson, J. M. “Solid-State Shear Pulverization of Plastics: A Green Recycling Process,” Polymer-Plastics Technology and Engineering, 38:445 (1999).
©2001 CRC Press LLC
P1: FUY pb035-03
March 13, 2001
9:56
Char Count= 0
66. Furgiuele, N., Khait, K., and Torkelson, J. M. “Novel Approach for the Compatibilization of Polymer Blends and Polymeric Waste,” Polymeric Materials: Science and Engineering, 70:70–71 (1998). 67. Furgiuele, N., Khait, K., and Torkelson, J. M. “Efficient Mixing of Polymer Blends of Extreme Viscosity Ratio: Elimination of Phase Inversion via Solid-State Shear Pulverization,” Polymer Engineering and Science, 40:1447–1457 (2000). 68. Torkelson, J. M., Khait, K., and Furgiuele, N. “A Novel Process for Efficient Blending of Polymers of Extreme Viscosity Ratio: Solid-State Shear Pulverization,” Proceedings of the Sixth European Symposium on Polymer Blends, Mainz, Germany, May 16–19, 1999. 69. Khait, K. and Torkelson, J. M. “A Novel Polymer Processing Technology: A Solid-State Shear Pulverization (S3 P),” Proceedings of the International Polymer Processing Society Annual Meeting PPS-15,’s Hertogenbosch, Netherlands, May 31–June 4, 1999. 70. Arinstein, A. E., Balyberdin, V. N., Kelly, B. M., Kelly, K. F., and Nikol’skii, V. G. “HighTemperature Shear-Induced Multiple Cracking and Grinding of Polymeric Materials,” Proceedings of the Rubber Division Meeting of the American Chemical Society, Nashville, Tennessee (1998).
©2001 CRC Press LLC
P1: FUY pb035-04
February 27, 2001
10:47
Char Count= 0
CHAPTER 4
Equipment for the S3 P Process for Plastics and Rubber
MACHINE FUNCTIONS OF THE TWIN-SCREW EXTRUDER
C
intermeshing twin-screw extruders have long proved their worth in a wide variety of compounding tasks. Since the mid-1980s, however, a new approach to twin-screw extrusion has emerged in which the main processing occurs while the thermoplastic material stays below its melting point. The term “extrusion” no longer applies because the product at discharge is in a solid powder form and is no longer “extrudable” into a defined shape. This new process takes place under high-shear stress and leads to solid-particle breakup in a continuous operation. Known as S3 P, this process can be applied to numerous polymers as well as to rubber, including vulcanized rubber. This chapter discusses the machine functions of the twin-screw extruder, looking first at its traditional use as a versatile compounding extruder and then at the ways in which these functions can be applied or extended to increase the efficiency of the pulverizer. The principles of this technology challenge the common extrusion theory by which complete melting of the polymer occurs before the main mixing or reaction process starts. During the S3 P process, the highest shear forces in any extruder are formed in the solid pellet bed during its compaction shortly before the melting step. If the resulting frictional heat can be removed efficiently, the high-shear stress applied to the feedstock will fracture the particles, rather than softening the feedstock due to temperature rise and consequent melting. The first experiments in which the S3 P process was discovered and reported were conducted on pilot-sized, co-rotating twin-screw extruders, and the first US patent for the process was granted in 1986 [1]. Meanwhile, results have been published for polyolefin resins and vulcanized rubber [2]. O-ROTATING,
©2001 CRC Press LLC
P1: FUY pb035-04
February 27, 2001
10:47
Char Count= 0
The feed components are typically metered to the feed throat of the extruder in the form of pellets, chips, or flakes to be conveyed into the barrel toward the melting section consisting of kneading blocks. It is important to understand the individual steps that transform a polymer pellet into melt during compounding. In the case of S3 P, however, it is necessary to eliminate melting in order to maintain pulverization. During the melting of LDPE, for example, the following five steps can be distinguished [3]: (1) Pellet deformation by compression in the screw channel (2) Formation of powdery flakes that change into solid, compacted flakes (3) Appearance of a molten film at high-shear points as in the scissoring position of the paddles (4) The presence of a molten film at the barrel wall and at the active side of the paddles “engulfs” the solid regions (5) Melting happens by viscous energy dissipation If one succeeds in removing the energy dissipated during steps 3 and 4, the melting process is interrupted, and powder is produced. The co-rotating, twinscrew extruder, with its tightly intermeshing clearances and modular barrel and screw design, can be set up to perform as a pulverizer. The typical layout of a twin-screw extruder is shown in Figure 4.1. The variable speed motor is
FIGURE 4.1 Berstorff twin-screw extruder ZE-75A-UT for compounding and pelletizing thermoplastics. (Courtesy of Berstorff GmbH.)
©2001 CRC Press LLC
P1: FUY pb035-04
February 27, 2001
10:47
Char Count= 0
connected to the input shaft of a gear box. The gear box has two functions: (1) to reduce the motor speed (for example, 1,750 rpm maximum) to the operating speed of the screws (500 rpm maximum) and (2) to divide the torque and transmit it equally to both screw shafts. Between the process section (barrels and screws) is a connecting piece by which the screws are coupled to the output shafts of the gear box. In the twinscrew pulverizer, the following steps take place: (1) (2) (3) (4) (5)
Feeding of solids in pellet, crumb, or flake form to the first barrel section Conveying of this material to the pulverization section High-shear pulverization section Cooling zone to remove frictional heat Conveying of the powder to the discharge orifice at the screw tips
In contrast to extrusion processes, there is no pressurizing step at the discharge end of the pulverizer. If the powder becomes compressed, it may become agglomerated, and this would defeat the purpose of the process. Steps 3 and 4 can be repeated several times along the pulverizer, depending on the available barrel and screw lengths; these actions affect the rate of powder production and the powder particle size.
MODULAR DESIGN OF BARREL SECTIONS The barrels come in segmented sections of 4–6 L/D and are connected with flanges and guiding pins to the adjacent barrel. It is not practical to manufacture the entire barrel in one piece due to limitations of the machining process. Barrel sections can come with open ports on the top or the side through which ingredients can be added at any stage of the process. If an opening is not used for feeding or venting, it can be closed with a contoured plug. The flanges for the barrel sections can be equipped with wells for temperature probes or pressure indicators. The standard barrel section is made of one piece with nitrated surfaces. For wear-resistant applications, an interchangeable liner of special wear-resistant material can be inserted into the barrel housing. Each barrel section has its own temperature-control zone that activates the electrical resistance heaters (cartridge heaters or band heaters) and controls the flow of cooling fluid. Because pulverization requires a high-shear energy output, the cooling system had to be designed to remove the energy along the barrels. A typical flow schematic of the barrel-cooling system is shown in Figure 4.2, with electrical heater bands and a primary water circulating system. The system pressure is 2–3 bars. The cooling bores are arranged close to the inner barrel surface. If the temperature exceeds a set point, the solenoid valve “e” opens, and water is injected under pressure into the barrel bores. The length of the “open” impulse and the
©2001 CRC Press LLC
P1: FUY pb035-04
February 27, 2001
10:47
Char Count= 0
FIGURE 4.2 Barrel cooling schematic for pulverizers: a) barrel cross-section with peripheral cooling bores and electrical heater bands, b) coolant supply line, c) return line, d) needle valve for fine adjustment, e) solenoid valve, f ) check valve, g) heat exchanger, h) coolant tank, i) coolant pump, k) by-pass valve, l) solenoid valve, and m) chiller unit. (Courtesy of Berstorff GmbH.)
frequency are set by the controller. The check valves “f” ensure that steam from the cooling bores can escape to the return line of the cooling system. For a significant cooling task, the velocity of the cooling media should be at least 2 m/sec, and the temperature increase along the cooling bores should stay below 10◦ C [4]. For the pulverizing process, the barrel temperatures can be set as low as −5◦ C by using a chilled refrigerant and barrel insulation covers to prevent condensation. The cooling efficiency can be improved by using cooled screw shafts. Screw cooling is a common practice for the feed sections in single-screw extruders where the polymer can be prevented from sticking to the hot screw channel along the cold feed sections. In single-screw extruders, after the melting process is
©2001 CRC Press LLC
P1: FUY pb035-04
February 27, 2001
10:47
Char Count= 0
established, screw cooling cannot be used to control the process because it will cause freezing of a layer at the bottom of the screw channel. Twin screws, however, with the intermeshing and self-wiping characteristics, allow for screw cooling over the full length of the process section.
MODULAR DESIGN OF SCREW ELEMENTS WITH CONVEYING, MELTING, AND PULVERIZATION FUNCTIONS The screws consist of two solid shafts that are connected to the gear box by a splined coupling piece. The screw elements are segmented and can be arranged onto the shafts according to the process task. The torque is transmitted from the gear box to the shafts and elements through keyed or splined connections. Especially for a novel process such as S3 P, it was essential to modify the profile in fast turnovers to come up with an optimum. Each of the tested polymers required a more-or-less severe pulverization section to balance the amount of frictional heat with the amount of heat that could be removed at the location of heat generation. Too much local friction would lead to melting, while not enough friction would reduce the efficiency of the pulverizer or would shift the particle size distribution to a more coarse mesh.
CONVEYING ELEMENTS The theoretical conveying capacity of the twin-screw pulverizer is higher than that of any available downstream equipment. Therefore, the feedstock is metered with a separate feeder independently from the screw speed. This means that the screw speed becomes an independent variable for the pulverization process. Higher screw speed does not increase the overall rate but could influence particle size, shear stress, and residence time. The conveying elements as shown in Figure 4.3 are used in the feed section to transport the pellets or chips to the first pulverization section. Between the pulverization sections, conveying elements serve the purpose of tumbling the product and maintaining sufficient heat transfer conditions at the lowest energy input. In the traditional extruder, conveying elements are also used to generate pressure and “extrude” the melt through a die orifice. In the S3 P process, however, powder is discharged at zero pressure to avoid any compaction or agglomeration.
KNEADING BLOCKS Figure 4.4 shows a set of kneading blocks in the area where melting takes place. The displacement angle of kneading paddles is designed either with a
©2001 CRC Press LLC
P1: FUY pb035-04
February 27, 2001
10:47
FIGURE 4.3
Char Count= 0
Double-flighted conveying element. (Courtesy of Berstorff GmbH.)
FIGURE 4.4
©2001 CRC Press LLC
Staggered kneading block. (Courtesy of Berstorff GmbH.)
P1: FUY pb035-04
February 27, 2001
10:47
Char Count= 0
FIGURE 4.5 Pressure profiles (above) and compacted solids in the channels (below) in the radial direction of kneading blocks. (Courtesy of Berstorff GmbH.)
forwarding angle, a reversing angle, or with a neutral staggering angle. Typical compounding can be accomplished over a barrel length of 2 L/D. During the melting process, the dissipation of energy imparted by the interparticle friction is larger than that caused by friction against the barrel wall or by heat transfer through the barrel wall [5]. Due to the compression and expansion of the volumetric cross sections with each screw revolution (illustrated in Figure 4.5), interparticle friction is generated. It has been observed that during the pulverization process, pellets are first fractured, and then powder is formed. Large particles such as pellets are reduced in diameter by (1) Attrition from contact with neighboring pellets (2) Softening and sloughing-off of layers as a consequence of localized interparticle friction (3) Particle breakup followed by extreme deformation under high local-shear stress or elongational stress If the frictional heat during the pulverization process is not sufficiently removed, the polymer will melt (which is undesirable). In the pulverization process, the heat is removed through the screws and barrels, and the powder is formed. The pulverization zone has to be designed carefully to allow for just the right degree of filling. If the kneading blocks are arranged too restrictively, the local energy boost is higher than the cooling power of that zone, and melting occurs. If the
©2001 CRC Press LLC
P1: FUY pb035-04
February 27, 2001
10:47
Char Count= 0
compression forces in the pulverization zone are too weak, the stress forces stay below the critical stress that leads to particle breakup, and the material passes through the pulverizer unchanged.
HEAT TRANSFER CALCULATIONS FOR BARREL AND SCREWS Usually, there is little difficulty in getting heat into a viscous polymer in twin-screw extruders. The installed motor power is sized by a factor of at least 3 over the energy that can be removed through the barrel surface. These are empirical design criteria from the traditional compounding process. The most difficult aspect of the twin-screw machine involves heat removal, a requirement of the pulverization process. From other special compounding tasks such as the dispersion of color pigments in melt, the barrel cooling process has been optimized to allow for low barrel temperatures that will reduce the viscosity of the melt layers at the barrel wall and generate higher shear forces to enhance pigment particle reduction in the viscous melt. Figure 4.6 shows the energy balance for the twin-screw pulverizer. The heat transfer term Q is negative if heat is removed by the barrel cooling system. The cooling capacity can be calculated with the “rules of heat exchanger technology” as in Equation (1): Q = k Ab (Tm − Tcool )
FIGURE 4.6
(1)
Energy balance for the twin-screw pulverizer. (Courtesy of Berstorff GmbH.)
©2001 CRC Press LLC
P1: FUY pb035-04
February 27, 2001
10:47
Char Count= 0
FIGURE 4.7 Heat transfer coefficient between cooling medium and barrel wall in relation to flow velocity. (Courtesy of Berstorff GmbH.)
Ab is the inner barrel surface. K is the heat transfer coefficient and can be calculated with Equation (2): k=
1 1/cm + sb /b + 1/m
(2)
The heat transfer coefficient cm from cooling medium to barrel bores can be taken from Figure 4.7. A typical velocity of the cooling medium in the cooling bores of the barrel is 3 m/sec, with the optimum values for the heat transfer. The term of the heat transfer in the barrel wall “sb /b ” is the heat conduction in solids with the wall thickness “sb ” and the thermal conductivity “b ” of the barrel’s construction material. The most critical term in Equation (2) is “m ,” the heat transfer coefficient from the inner barrel surface to the melt or, in this case, to the powder particles. In the case of melt, equations have been developed where the heat conduction into the molten layer of polymer (that forms after the flight has wiped the barrel wall) is calculated for the period until the next revolution, when the layer
©2001 CRC Press LLC
P1: FUY pb035-04
February 27, 2001
10:47
Char Count= 0
is scraped off again and mixed back into the melt [6]. The situation for the pulverizer is different because loose powder particles are present over most of the channel. Only at high shear rates (over the tips of kneading blocks and in the intermeshing points of the apex) is a sufficient wall-to-material contact made and optimum heat transfer found. It is a great boon for the pulverizing process that in the area of highest stress, the best opportunity for heat transfer occurs because the material layer is thin and in close proximity to cold barrels and cold screw surfaces. An energy balance for the pulverizer can be calculated using the following terms: (1) Heat removed from the system by barrel-cooling water (Equation [3]): Q cool = m W c pw TW
(3)
(2) Heat added to the system via motor energy (Equation [4]): Qm = I U
(4)
(3) Heat absorbed by the product in the form of enthalpy rise (Equations [5] and [6]): Q p = k Ab T
(5)
k = Q p /Ab T
(6)
A typical value for k was found at 400 to 450 W/m2 K. This value is in the same range as is found in classical polymer processes.
SCREW COOLING The surface area for heat transfer can be enlarged drastically if the screws can be cooled as well. In single-screw extruders, screw cooling is used typically in the feed section to prevent pellets from sticking to the screw channel. Farther downstream, after melting of the polymer is achieved, single-screw extruder screws cannot be cooled because the entire screw channel would be closed with a frozen layer, beginning at the bottom of the channel. In the co-rotating intermeshing twin-screw action, however, the adjacent screw strips away any frozen layers from the channel. In the case of the pulverizer, the cooling water can be connected either at the rear of the screws with a special coupling (illustrated in Figure 4.8) or by using rotary joints at the screw tips with hollow shafts. Because there is no pressure at the discharge end of the pulverizer, the latter is more practical. It does not reduce the mechanical strength of the shafts
©2001 CRC Press LLC
P1: FUY pb035-04
February 27, 2001
10:47
Char Count= 0
FIGURE 4.8 Screw cooling arrangement at screw tips: 1) Barrel head flange and bottom discharge for powder, 2) Support flange for bearing assembly, 3) Tie bar connection to barrel flange, 4) Screw tip extension, 5) Bearing plate with seal rings, 6) Axial roller bearings, 7) Rotary joint with coolant inlet and outlet connections, and 8) Screws with hollow shaft. (Courtesy of Berstorff GmbH.)
from the drive end because the shafts remain solid in the feed section. For the proper operation of the rotary joints, radial roller bearings are mounted in the head flange. As it turns out, these bearings are also responsible for guiding the screws and for maintaining accurate clearances between screw flights and barrels. The wear of elements and barrels is minimal. To estimate the cooling effect on the screws, one may use 35% of the barrel cooling power [7].
PROCESS EXAMPLES During the 1990s, the S3 P process was applied to numerous plastic and rubber materials [8]. It was expected that pulverization would occur under
©2001 CRC Press LLC
P1: FUY pb035-04
February 27, 2001
10:47
Char Count= 0
FIGURE 4.9 Particle-size distribution for virgin LDPE powder made with the PT-25 pulverizer. (Courtesy of Matthew A. Darling, Polymer Technology Center at Northwestern University.)
different conditions for different polymers. In the experimental stages of the S3 P process, the term “pulverizability” was created [9]. This term defines what type of powders, if any, can be created from a polymer and under what conditions. If a polymer changes into separate particles (granular, fibrilar, or spherical) at the discharge end of the pulverizer, it is “pulverizable.” If a melt is obtained, however, the polymer cannot be pulverized. The economic aspects of output rate, energy input, and particle size distribution all need to be considered in developing a successful pulverization process. After a polymer has been pulverized, the next step is to determine the powder’s attributes. The particle size of the powder is determined by sieve analysis. Figure 4.9 shows the particle size distribution for LDPE as obtained from a laboratory-scale PT-25 pulverizer. The illustration shows that by adjusting the screw profile, the percent of small particles can be increased significantly. The shape of the powder can vary from a spherical shape in the case of LDPE (Figure 4.10) to a fibrilar shape, which is dominant for HDPE and PP (Figures 11a,b). The fibrilar particles can entangle and are not as useful as the ball-shaped particles. The surface area of an individual particle is typically large due to the rough nature of the fractured polymer surface. Pourability is another criterion applied to powders to determine their processability in mixers, extruders, or rotomolding applications. A important feature allowing the S3 P process to be applied to numerous polymers and rubber sources is the segmented screw design; this permits adjustments according to the material to be pulverized. The increasing variety of applications for the S3 P process is also impacting the viability of traditional melt-mixing processes.
©2001 CRC Press LLC
P1: FUY pb035-04
February 27, 2001
10:47
Char Count= 0
FIGURE 4.10 Optical photograph of virgin LDPE powder made with the PT-25 pulverizer. (Courtesy of Matthew A. Darling, Polymer Technology Center at Northwestern University.)
Polyolefins (LDPE, LLDPE, HDPE, and PP) are the most widely processed polymers for use in the rotomolding market [8]. The use of the pulverization process has been reported first for LDPE [1]. Polystyrene and polystyrene/ polyolefin blends have been converted into powders using S3 P technology [10]. It has also been demonstrated that during the pulverization, efficient mixing of polymer blends of extreme viscosity ratio took place. It was learned at an early stage of the investigation of the S3 P process that multicolored, unsorted commingled plastics or different types of resins resulted in a rather uniform color and consistency when exposed to S3 P [8,11].
COMMINGLED PLASTICS FROM RECYCLING SOURCES The reuse of commingled plastics in an extrusion process normally results in a product that has much lower physical properties than the original polymers in the mix. The main reason is the incompatibility of the individual components. In conventional processes, large isolated polymer pieces pass the extruder without melting or mixing due to large differences in viscosity or melting points. When S3 P is applied to the commingled plastic waste stream, all components are
©2001 CRC Press LLC
P1: FUY pb035-04
February 27, 2001
10:47
Char Count= 0
(a)
(b) FIGURE 4.11 (a) Optical photograph of virgin PP powder made with PT-25 pulverizer. (b) Optical photograph of virgin HDPE powder made with PT-25 pulverizer. (Courtesy of Matthew A. Darling, Polymer Technology Center at Northwestern University.)
©2001 CRC Press LLC
P1: FUY pb035-04
February 27, 2001
10:47
Char Count= 0
reduced to a particle size smaller than that achieved with extrusion. With the resultant powder, further conversion can result in a more homogeneous melt. S3 P can be used to improve the quality of agricultural film that contains LDPE as well as LLDPE to the point where it can be reused for blown-film applications to again serve the original agricultural purpose.
PULVERIZATION OF CURED RUBBER The S3 P technology can be applied to the pulverizing of vulcanized rubber into a fine powder. The pulverized used-tire rubber is then used successfully as a high-grade filler or additive that is found in a variety of recycled products:
r floor mats for farms and horse stables r cast with TPU for tracks on sports arenas and tennis courts r noise-dampening material r mixed with asphalt for road construction r Fine rubber powder used as an abrasion-resistant additive in new tires and rubber products. The typical particle size for rubber powder is less than 140 mesh (100 microns). Figure 4.12 shows a Berstorff production line that converts used tires into rubber powder. The first step is a crushing mill where the rubber is separated from steel cords, and then the rubber pieces of 10 to 20 mm can be fed to the pulverizer. The process of rubber pulverization is different from the S3 P process with thermoplastic materials because high shear stress does not lead
FIGURE 4.12 Turnkey tire-recycling line featuring commercial-size Berstorff PT-90 pulverizer. (Courtesy of Berstorff GmbH.)
©2001 CRC Press LLC
P1: FUY pb035-04
February 27, 2001
10:47
Char Count= 0
to melting of the rubber material. Experience has shown, however, that the process temperature should not exceed 180◦ C in the pulverization sections. Above that temperature, already ground particles become sticky and tend to reagglomerate. At the same time, degradation of the rubber will set in, causing volatiles to escape. A further rise in temperature can lead to smoldering fires in the powder in the presence of oxygen due to the high surface area of the rubber powder. For economic operation of the rubber pulverizer, the average particle size of the rubber should be less than 0.5 mm. A rate of at least 1,000 lb/hr is expected on a pulverizer of the size PT-90A (90-mm diameter). Experiments during the developmental stage of the pulverizer have resulted in 400 to 600 lb/hr rates for the target powder grade. Higher rates could be accomplished at a higher rotor speed, but high product temperatures or uneven particle size would restrict the process and prevent it from meeting the performance targets. Also, the fraction of smaller particles decreased at higher speeds. Figure 4.13 shows the machine layout of a three-stage pulverizer where the mechanical crushing of the rubber stock is carried out inside the barrels. In order to assist the cooling process, an “outboard” cooling device is attached to the barrels; this device continuously conveys, cools, and returns to the extruder a liquid refrigerant [12]. The powder can leave the barrels through bottom openings and enter a spiral-type cooling conveyor that elevates the powder for a gravity-fed re-entry into the next barrel section through a top opening. The powder temperature at the exit is 170◦ C, and at the re-entry, it is around 40◦ C. It was found to be useful to screen out the fine particles of the mix before the reentry, which improved the efficiency of the unit. The barrels and screws could be set up effectively by utilizing the existing building blocks of the twin-screw
FIGURE 4.13 Layout for a three-stage rubber pulverizer. Hot rubber crumbs can be discharged at two downstream locations (10 and 19) for cooling and fed back (13 and 14) for the second pass (II). Berstorff US patent 5,273,419 (1993). (Courtesy of Berstorff GmbH.)
©2001 CRC Press LLC
P1: FUY pb035-04
February 27, 2001
10:47
Char Count= 0
system. The pulverizer produces rubber powder at a rate of about 1,000 lb/hr. The typical specific energy input is 0.4 kWhr/kg. The achieved cooling power from barrels and screws is approximately 0.1 kWhr/kg. SCALE-UP CONSIDERATIONS FOR THE S3 P PROCESS The scale-up from laboratory models to production-size equipment is based on the assumption that the average shear stress will stay constant and that it is critical for the particle breakup, which leads to pulverizing. It is calculated based on Equation 7: = P 9549000/n Ab : P: n: Ab :
(7)
shear stress in N/mm2 power in kW screw speed in rpm barrel surface in mm2
The concentration of the stress forces is over the tight tip clearances of the kneading blocks. The average residence time in these high-shear areas needs to stay constant to avoid overheating and melting. For purposes of scale-up to commercial-size models, the tip speed and radial length of the shear gap will have to remain constant.
SUMMARY When Berstorff’s twin-screw extruder was initially used for size reduction and pulverization, tire recycling was of interest, and the S3 P process was developed from the pilot plant level to a production plant level. In the case of vulcanized rubber, high shear forces could be applied and fine powders obtained. For thermoplastic materials, a limited shear stress could be applied during the pulverization process to prevent melting. The heat transfer capacity between screws and barrels was the limiting factor on standard twin-screw extruder and led to design efforts to optimize heat transfer by redesigning barrels and screws. At the same time, it was discovered that during the pulverization of multicolored commingled plastics, homogeneous colored powders were produced. It was concluded that the explanation for this homogeneity lies in the excellent mixing capability of the pulverizer. The combination of particle size reduction and mixing of individual components led to other discoveries related to polymer processing such as the creation of new polymer blends. Melt processing, which used to be a difficult task for mixing polymers with drastically different
©2001 CRC Press LLC
P1: FUY pb035-04
February 27, 2001
10:47
Char Count= 0
viscosities, is no longer necessary. The non-melting S3 P process can also be used for mixing polymers with additives that are sensitive to the exposure to high temperatures. Furthermore, mixing in the solid state has been found to be most effective and allows for combining polymers that are considered incompatible [11,12]. This permits designing the properties of the materials by blending polymers instead of taking the conventional route to synthesize new polymers.
REFERENCES 1. Enikolopov, N. S., Wolfson, S. A., Nepomnjaschtschie, A. J., Nikol’skii, W. G., Teleschow, W. A., Filmakowa, L. A., Brinkmann, H., Pantzer, E., and Uhland, E. US Patent 4,607,797 (1986). 2. Mack, M. H. “Twin-Screw Machines Explore Solid-State Extrusion,” Plastic Technology, p. 75, (September 1993). 3. Essegir, M. “Melting Mechanisms of Single-Component Polymers in Co-Rotating Twin-Screw Kneading Blocks Through Visual and Microscopic Analysis,” The Annual Technical Conference of the Society of Plastics Engineers, ANTEC ’97, pp. 3684–3689 (1997). 4. Mueller, W. “Beheizung und K¨uhlung,” Kunststoffextrusionstechnik I, Munich: Carl Hanser Verlag (1998). 5. Todd, D., ed. Plastic Compounding, Equipment and Processing, Munich: Carl Hanser Verlag (1989). 6. Martin, G. Verein Deutscher Ingenieure (1986). 7. Schuler, W. “Process Engineering Design of Co-Rotating Twin-Screw Extruders,” unpublished dissertation, p. 133 (1996). 8. Khait, K. US Patent 5,814,673 (1998). 9. Darling, M. “Polyolefin Pulverizability,” senior project in Department of Chemical Engineering, Northwestern University, Evanston, Illinois (1995). 10. Furgiuele, N., Lebovitz, A. H., Khait, K., and Torkelson, J. M. “Efficient Mixing of Polymer Blends of Extreme Viscosity Ratio: Elimination of Phase Inversion vis SSSP,” Polymer Engineering and Science, 40:1447–1457 (2000). 11. Khait, K. and Torkelson, J. M. Proceedings of the International Polymer Processing Society Annual Meeting—PPS-15, Hertogenbosch, Netherlands, May 31–June 4, 1999. 12. Mayer, D. and Dreist, B. US Patent 5,273,419 (1993).
©2001 CRC Press LLC
P1: SYV/FYX pb035-05
P2: SYV/ABE
March 2, 2001
QC: SYV/ABE
15:13
T1: SYV
Char Count= 0
CHAPTER 5
S3 P Technology and Virgin Polymers
P
powders are used in a wide array of applications from decorative and protective coatings to rotational molding of deep-draw hollow parts and serve as processing aids and fillers. The conventional methods, however, for producing these powders from pellets or flakes have several deficiencies, including the batch nature of these methods and their high use of energy. These deficiencies not only make the production of large quantities of powders costly, they also render economically unfeasible the use of certain virgin polymers as feedstock. Furthermore, compounding these polymers with additives after they are made into powders is frequently difficult. OLYMERIC
APPLICATIONS FOR PLASTIC POWDERS Polymeric powders are used in powder coatings, sintering, compression and rotational molding, ram extrusion, and compounding. Powder coatings involve a deposition of polymeric powders on a surface followed by their coalescence to continuous film through applying heat. Coatings on surfaces such as metal and wood can provide protection and decoration. Protective coatings are applied as thicker films (250–2,500 microns) with fluidized-bed or electrostatic spray methods to preheated parts. Decorative coatings are usually applied to cold parts by electrostatic deposition at a lower film thickness (25–75 microns). Decorative grades use finer colored powders than do protective coatings. Powder coatings are 100 percent solids and include both thermosets and thermoplastics. Thermoset powder coatings consist of a resin, a flow promotor, a catalyst, a cross-linking (curing) agent, and a variety of pigments, additives, and fillers. In the presence of a curing agent, these resins polymerize
©2001 CRC Press LLC
P1: SYV/FYX pb035-05
P2: SYV/ABE
March 2, 2001
QC: SYV/ABE
15:13
T1: SYV
Char Count= 0
to a film that is irreversibly cross-linked. Among the parameters affecting flow and uniformity of the coating powder are melt viscosity and reactivity of the resin and curing agent. Particle size, packing of the particles, and curing temperature are also important factors. Polymers used in coatings include epoxy, polyesters, epoxy/polyester hybrids, polyester/urethanes, and acrylics. Manufacturing of powders for coating applications consists of several steps including dry mixing, melt compounding, cooling, primary size reduction, fine grinding, classification, blending with additives, and packaging. The majority of powders are made by the melt-compounding process by which dry-blended ingredients are melted and dispersed in the extruder followed by a grinding operation. Grinding to a desired particle size is coupled with classification of powders, and oversized particles are returned to a grinder for further size reduction [1]. The dominant thermoplastic resins used for powder coatings are poly(vinyl chloride) (PVC), nylons, poly(ethylene) (PE), polypropylene (PP), and fluorocarbons. The fluorocarbons, which include poly(vinylidene fluoride) (PVDF), ethylene-tetrafluoro ethylene, ethylene-chlorotrifluoroethylene, and perfluoroalkoxy, have exceptional chemical resistance and nonstick surfaces [2]. Some thermoplastic polyesters combine impact strength with heat- and chemicalresistance properties and are rotomolded into parts for automotive markets. A newer addition to the family of resins suitable for rotational molding is thermoplastic olefin (polypropylene-ethylene copolymer), which has high stiffness, good impact strength, low warp-, and high environmental stresscrack resistance. Highly viscous polymers, such as ultra-high molecular weight poly(ethylene) (UHMWPE) or poly(tetrafluoroethylene) (PTFE), are processed by ram (screwless) extrusion. Upon melting, these resins become highly viscous and can form thick films that are ideal for protective coatings. As mentioned above, an important property of a polymer used in a powder coating is its melt viscosity. Lower melt viscosity leads to better flow and easier incorporation of various additives such as pigments, dyes, and fillers. In general, thermoplastics are high in molecular weight, have good physical properties, and are difficult to grind to a fine particle size. Thermosets have both low molecular weight and low melt viscosity and are easy to grind. Many thermoplastics have been evaluated for powder coatings but only a few have the desired combination of properties, such as melting temperature, melt viscosity, and thermal stability, to be used in a fusion-coating process. These thermoplastics include polyamides (Nylon 11 and Nylon 12), PE, and PP. Newly developed polyolefin copolymers and functional polyolefins with comonomers, such as maleic anhydride, acrylic or metacrylic acid, and silanes, have dramatically improved the adhesion to metal substrates [3]. Coatings based on PVC are plasticized for improved melt flow. Powders based on PVDF are used as exterior durable and corrosion-resistant coatings.
©2001 CRC Press LLC
P1: SYV/FYX pb035-05
P2: SYV/ABE
March 2, 2001
QC: SYV/ABE
15:13
T1: SYV
Char Count= 0
Polymeric powders can also be compression-molded. This process is used mostly for thermosets and involves squeezing heated material between the two halves of a mold situated on a press. Another process utilizing polymeric powders is sintering, in which solid particles are in contact with one another at an elevated temperature. During sintering, an interface between adjacent particles is developed followed by a densification in which the inter-particle cavities are eliminated (melting by sintering is practiced in rotational molding). Rotational molding is a low-pressure process that uses heat and mold rotation to produce hollow parts with complex shapes. Polymeric powder is sintered onto the cavity as the mold rotates; in this no-pressure process, the powder particles are flowing together to produce a solid part. Rotational molding is a polymer-dependent technology. Initially, only vinyl plastisols and polyolefin powders were used [4]. Since the early 1970s, crosslinked and modified poly(ethylenes) were produced for large-tank markets, followed by linear low-density poly(ethylene) (LLDPE). During the 1980s, nylon, PP, and polycarbonate powder were developed for rotomolding applications. Low-, high-, and linear low-density poly(ethylenes), cross-linked highdensity poly(ethylene), and copolymers are the most commonly used resins. Key physical properties include melt flow rate, molecular weight distribution (MWD), and density. Resins with increased melt flow rates have lower tensile strength, low-temperature impact strength, and lower heat-distortion temperature. As MWD narrows, processability and physical properties of materials improve. Properties such as stiffness, heat distortion temperature, and shrinkage usually increase with higher density values. The differences between lowand high-density poly(ethylene) depend on their structure [5]. High-density poly(ethylene) (HDPE) is a linear polymer with few branches that are closely packed (i.e., “high density”). Low-density poly(ethylene) (LDPE) has many long chain branches that prevent the chains from close packing, resulting in lower density. Linear low-density poly(ethylene) (LLDPE) has short chain branches and, therefore, lower density. In the 1990s, some rotationally molded products contained 15–25 percent post-consumer resins. A resin suitable for rotomolding should have a melt flow rate (MFR) between 3 and 20 g/10 min. The MWD of poly(ethylene) should be narrow in order to achieve uniform melting of the powder during the molding cycle. Most resins range in particle size from 7 to 200 mesh (2,000–74 microns), but the standard size is 35 mesh (500 microns). A particle size within this range is required to promote heat transfer from the mold to powder [6]. It has been determined that particles of a size greater than 35 mesh (500 microns) do not melt in the same fashion as finer particles, so the surface of rotomolded parts appears grainy [4]. Much finer particles (below 140 mesh, or 100 microns) have considerably lower bulk density than coarser particles, leading to voids in molded parts. Other undesirable traits of finer powder particles are agglomeration and uneven melting.
©2001 CRC Press LLC
P1: SYV/FYX pb035-05
P2: SYV/ABE
March 2, 2001
QC: SYV/ABE
15:13
T1: SYV
Char Count= 0
The rotational molding process requires powder particles that can fuse together to form a solid mass [6]. Particle size affects the rate at which heat can penetrate to the middle of each particle. Particle size and particle size distribution, shape, and pourability of powder all have a profound impact on the quality of rotationally molded parts and on the efficiency of the process. Particles with a “tail” will prevent the powder from flowing freely and will cause “bridging,” which in turn keeps the powder from filling the corners of the mold. Also, while irregularly shaped particles can trap air and cause voids in the finished part, some degree of particle irregularity is nevertheless needed to achieve free-flow properties. Particle shape is also important for rotational molding because it affects “dry flow” properties. Spherical particles do not sinter well, while fibrous particles cause bridging. The shape of the particles affects density and flow of the powder. Another parameter of importance is bulk density, which depends on the grinding method and on the type of polymer used. The bulk density of the powder depends on the particle size distribution and packing. For spherical particles, the maximum packing fraction is between 0.5 and 0.74 [7]. In general, plastic powders are made by compounding resins with additives and modifiers including pigments, foaming and toughening agents, flow promoters, internal lubricants, antistatic agents, antioxidants, ultraviolet light stabilizers, cross-linking agents, fire retardants, and others. After melt-mixing, these compounds must be ground into a powder by either ambient or cryogenic grinding. Pulverizing plastic pellets into a powder adds to the raw material’s cost. It is known that excessive energy is required when fine powders are produced to a size below 140 mesh (100 microns). Low throughput during grinding is attributed to a large quantity of oversized powder and agglomerated particles. Additionally, during grinding, some resins can lose physical properties and color as a result of overheating. Due to toughness, some thermoplastics are difficult to grind at room temperature. Therefore, grinding is performed at a cryogenic temperature by using pin disc-and-hammer mills. Because of the high cost of nitrogen and other refrigerants and processing difficulties associated with cryogenic grinding, other methods of producing powders continue to be developed. Some involve agitation of resins above the melting point while suspended in nonsolvent or water. Other approaches include mixing with an incompatible polymer to make the thermoplastic more brittle prior to grinding [3]. Further methods involve dissolution and precipitation, but these are undesirable due to the use of solvents. A novel S3 P process can significantly simplify the preparation of powders for coatings or for rotational molding applications by eliminating several steps in the current manufacturing process. This can be achieved by carrying out pulverization and mixing/dispersion of additives in one continuous operation, as shown in the flow chart in Figure 5.1.
©2001 CRC Press LLC
P1: SYV/FYX pb035-05
P2: SYV/ABE
March 2, 2001
QC: SYV/ABE
15:13
T1: SYV
Char Count= 0
FIGURE 5.1 Flow chart for powder coating production by conventional process (left) and with new S3 P technology.
As previously noted, many rotational molding applications require materials with a combination of properties such as toughness, ductility, chemical resistance, and impact strength. Many of these properties are achieved by using resins with high molecular weight, low melt flow rate, low crystallinity, and low density [8]. Typically, powders are made by grinding pellets, although some reactor powders (i.e., polystyrene) are also used. Reactor powders, however, are very fine and do not have the additives required for rotomolding. In addition, these powders have a spherical shape that causes them to roll too easily prior to melting and also results in uneven distribution within the mold, especially for parts with complex forms [6].
CHARACTERIZING POWDERS A comparative study has been conducted by the researchers at Northwestern University in Evanston, Illinois, between the conventional size-reduction techniques of ambient and cryogenic grinding on the one hand and the S3 P process on the other [9]. The polymers studied consisted of two mixtures of polypropylene (PP) and polystyrene (PS) at 80/20 and 20/80 ratios. Prior to blending, two individual virgin polymers—PP homopolymer PP 8020 GU from Equistar Chemical Co. and PS G18 from Amoco Chemical Co.—were analyzed for particle size and shape. Pulverization was carried out using a Berstorff PT-25
©2001 CRC Press LLC
P1: SYV/FYX pb035-05
P2: SYV/ABE
March 2, 2001
QC: SYV/ABE
15:13
TABLE 5.1.
T1: SYV
Char Count= 0
Sieve Analysis of Powder Resulting from Solid-State Shear Pulverization of Virgin PS G18. Size Mesh
Microns
Weight Percent Retained
20 35 50 80 120 140 200 Fines
850 500 300 180 125 105 75 <75
55.9 25.3 11.4 4 1.8 0.6 0.7 0.3
pulverizer with a diameter of 25 mm and a L/D ratio of 26. Cooled barrels allow the removal of frictional heat generated during the S3 P process. Using a K-tron volumetric feeder, a polymer or a polymer blend was fed into the pulverizer. A modular barrel and screw configuration provided flexibility of various designs to be employed during the pulverization. At the end of the pulverizer the powder was discharged and classified. Sieve analysis of S3 P-made powders from virgin PS and PP was performed with a Gilson ultrasonic autosiever using US standard sieves (Tables 5.1, 5.2). These data indicated that particle size for S3 P-made powder from crystalline PP was finer than that of powder made from amorphous PS. The particle shape of powders obtained by the S3 P process is not an easy property to measure or to characterize due to the distribution of shapes within a large number of particles. Scanning electron micrographs at 40 and 250 magnifications of virgin PP and PS are presented in Figures 5.2a,b and Figures 5.3a,b, respectively. Both sets show elongated particles suggestive of shearing during the pulverization process. TABLE 5.2.
Sieve Analysis of Powder Resulting from Solid-State Shear Pulverization of Virgin PP 8020 GU. Size Mesh
Microns
Weight Percent Retained
20 35 50 80 120 140 200 Fines
850 500 300 180 125 105 75 <75
23.9 32.5 25.1 11.1 3.9 1.7 1 0.7
©2001 CRC Press LLC
P1: SYV/FYX pb035-05
P2: SYV/ABE
March 2, 2001
QC: SYV/ABE
15:13
(a)
FIGURE 5.2 PP at 250×.
(b)
(a) SEM image of pulverized virgin PP at 40×. (b) SEM image of pulverized virgin
(a)
FIGURE 5.3 PP at 250×.
T1: SYV
Char Count= 0
(b)
(a) SEM image of pulverized virgin PS at 40×. (b) SEM image of pulverized virgin
©2001 CRC Press LLC
P1: SYV/FYX pb035-05
P2: SYV/ABE
March 2, 2001
QC: SYV/ABE
15:13
TABLE 5.3.
Sieve Analysis of Virgin PP/PS Blends at 80/20 as a Function of the Size-Reduction Method.
S3 P Process Size Mesh Microns 20 35 50 80 120 140 200 Fines
T1: SYV
Char Count= 0
Ambient Grinding
Weight Percent Retained
850 500 300 180 125 105 75 <75
26.6 37.9 21.6 7.3 3.2 1.1 1.3 1
Size Mesh Microns 20 35 50 80 120 140 200 Fines
850 500 300 180 125 105 75 <75
Weight Percent Retained 42.8 29 15.7 7.5 2.5 1.9 0.2 0.4
Cryogenic Grinding Size Mesh Microns 20 35 50 80 120 140 200 Fines
850 500 300 180 125 105 75 <75
Weight Percent Retained 25.9 47.1 15.4 7.4 2.5 1.1 0.5 0.2
The effects of the three size-reduction methods—S3 P, ambient grinding, and cryogenic grinding—on size of powders made from virgin polypropylene/ polystyrene mixture at 80/20 and 20/80 ratios are shown in Tables 5.3 and 5.4, respectively. Sieve analysis of a PP/PS mixture at 80/20 ratio made via the S3 P process showed that ∼38 percent of the powder has a size of 500 m (35 mesh), while the powder made by ambient grinding had 29 percent of particles that size. Sieve analysis of a PP/PS mixture at the same ratio made by cryogenic grinding showed that 47 percent of the powder had a size of 500 m (35 mesh). Particle size analysis of a 20/80 PP/PS mixture in which amorphous PS was a major component, however, showed a different particle size distribution. Specifically, sieve analysis of a PP/PS mixture at 20/80 ratio made via the S3 P process showed that only ∼19 percent of the powder had a size of 500 m (35 mesh), TABLE 5.4.
Sieve Analysis of Virgin PP/PS Blends at 20/80 as a Function of the Size-Reduction Method.
S3 P Process Size Mesh Microns 20 35 50 80 120 140 200 Fines
850 500 300 180 125 105 75 <75
Ambient Grinding
Weight Percent Retained 57 19.3 12.3 6.1 2.9 1.6 0.8 0
©2001 CRC Press LLC
Size Mesh Microns 20 35 50 80 120 140 200 Fines
850 500 300 180 125 105 75 <75
Weight Percent Retained 25.4 38.2 20 9.1 3.6 2.4 1 0.5
Cryogenic Grinding Size Mesh Microns 20 35 50 80 120 140 200 Fines
850 500 300 180 125 105 75 <75
Weight Percent Retained 9.8 40.2 25.9 13.9 4.7 22 1.9 1.4
P1: SYV/FYX pb035-05
P2: SYV/ABE
March 2, 2001
QC: SYV/ABE
15:13
T1: SYV
Char Count= 0
which is half the weight percent for the 80/20 PP/PS mixture. The difference in particle size distribution can be explained by the fact that crystalline polymers are easier to pulverize than amorphous polymers. Particle shape of the powders from virgin PP/PS blends at 80/20 and 20/80 ratios obtained from each of the three methods is shown in Figures 5.4a–f and 5.5a–f, respectively. The combination of pressure and shear during the S3 P process causes a modification of the particle formation mechanism. The S3 P-made particles exhibit an elongated shape while the conventionally ground particles have a sharp, angular shape, especially those made by cryogenic grinding. Powder particles made by ambient grinding have a slightly elongated shape and visually identical surface textures. The effects of S3 P processing on the physical properties of virgin PP, PS, and their binary blends at 80/20 and 20/80 ratios (Tables 5.5–5.8) were also studied [9]. Powders made from PP by the S3 P process were injection-molded directly from coarse (∼200 microns) and fine (∼500 microns) powder into test specimens without prior pelletization, and the results were compared to test specimens injection-molded from as-received pellets of virgin PP (Table 5.5). A significant increase in elongation at break can be observed for both coarse and fine powders as compared to pellets (110 percent vs. 33 percent). Other
(a)
(b)
FIGURE 5.4 (a) SEM image of pulverized virgin PP/PS blend at 80/20 ratio at 40×. (b) SEM image of pulverized virgin PP/PS blend at 80/20 ratio at 250×. (c) SEM image of virgin PP/PS blend at 80/20 ratio after ambient grinding at 40×. (d) SEM image of virgin PP/PS blend at 80/20 ratio after ambient grinding at 250×. (e) SEM image of virgin PP/PS blend at 80/20 ratio after cryogenic grinding at 40×. (f) SEM image of virgin PP/PS blend at 80/20 ratio after cryogenic grinding at 250×.
©2001 CRC Press LLC
P1: SYV/FYX pb035-05
P2: SYV/ABE
March 2, 2001
QC: SYV/ABE
15:13
T1: SYV
Char Count= 0
(c)
(d)
(e)
(f)
FIGURE 5.4
(Continued )
mechanical properties including tensile strength and flexural properties remained unchanged. In the case of PS, the physical properties of injectionmolded samples were similar to those molded from powders regardless of the powder particle size, and they were comparable to the properties of as-received pellets (Table 5.6). Unlike PP, the melt flow rate of S3 P-made PS-powders was higher than that of pellets and varied from 26.4 g/10 min for coarse powder
©2001 CRC Press LLC
P1: SYV/FYX pb035-05
P2: SYV/ABE
March 2, 2001
QC: SYV/ABE
15:13
T1: SYV
Char Count= 0
(a)
(b)
(c)
(d)
FIGURE 5.5 (a) SEM image of pulverized virgin PP/PS blend at 20/80 ratio at 40×. (b) SEM image of pulverized virgin PP/PS blend at 20/80 ratio at 250×. (c) SEM image of virgin PP/PS blend at 20/80 ratio after ambient grinding at 40×. (d) SEM image of virgin PP/PS blend at 20/80 ratio after ambient grinding at 250×. (e) SEM image of virgin PP/PS blend at 20/80 ratio after cryogenic grinding at 40×. (f) SEM image of virgin PP/PS blend at 20/80 ratio after cryogenic grinding at 250×.
©2001 CRC Press LLC
P1: SYV/FYX pb035-05
P2: SYV/ABE
March 2, 2001
QC: SYV/ABE
15:13
T1: SYV
Char Count= 0
(f)
(e)
FIGURE 5.5
(Continued )
to 33.5 g/10 min for fine powder, which suggested that greater chain scission occurred during the pulverization of PS than of PP. The physical properties of virgin 80/20 PP/PS blends that were injectionmolded directly from coarse (∼200 microns) and fine (∼500 microns) powders were compared to the properties of blends repelletized from coarse and fine powders using the conventional single-screw extruder. As can be seen from Table 5.7, physical properties of all four materials were similar with the exception of the melt flow rate, which was higher for coarse powder. The physical properties of the virgin 20/80 PP/PS blend (Table 5.8) injectionmolded from coarse (∼200 microns) and fine (∼500 microns) powders were similar and remained unchanged for powders repelletized by melt extrusion. The melt flow rate of 20/80 PP/PS blend repelletized from fine powder was higher than that of repelletized coarse powder (47.1 g/10 min vs. 36.2 g/10 min, respectively). The physical properties of virgin PP/PS binary blends at 80/20 ratio obtained by the three different size-reduction methods—S3 P processing, cryogenic grinding, and ambient grinding—are shown in Table 5.9. Tensile strength at yield was similar for all three powders, while elongation at break for S3 P-made powder was higher than that for two powders made by conventional grinding. The melt flow rate for S3 P-made powder was slightly higher (24.6 g/10 min) as compared with the ambiently ground PP/PS blend (23.4 g/10 min) or cryogenically ground powder (22.1 g/10 min). In the case of the 20/80 PP/PS blend (Table 5.10), none of the powders exhibited yield during the tensile test.
©2001 CRC Press LLC
b
32.4 25.4 (4,700) (3,690)
110
21 (0.4)
59
73
1,426 48.5 (206,900) (7,030)
20.0
19.1
Molded from fine powder
30.3 25.6 (4,400) (3,710)
120
21 (0.4)
59
73
1,525 50.7 (221,300) (7,380)
20.0
20.1
Molded from pellets (as received)
37.6 (5,450)
33
21 (0.4)
77
75
1,654 54.5 (240,000) (7,900)
20.0
20.0
Source of material: Equistar Chemical Co. Pull rate of 0.2 in/min.
©2001 CRC Press LLC
T1: SYV
a
Molded from coarse powder
Char Count= 0
Virgin PP 8020 GUa
Yield MPa (psi)
QC: SYV/ABE
15:13
Notched Flexural Properties lzod Heat Melt Flow Rate Ultimate Impact Distortion Modulus Strength g/10 min (230◦ C, 2.16 kg) MPa Elong J/m Temperature Hardness MPa MPa Before S3 P After S3 P (psi) % (ft-lb/in) ◦ C 264 psi Shore D (psi) (psi)
Tensile Propertiesb
P2: SYV/ABE
March 2, 2001
Physical Properties of Virgin PP 8020 GU Injection-Molded from S3 P-Made Powder.
P1: SYV/FYX
pb035-05
TABLE 5.5.
Molded from coarse powder
40.5 (5,880)
4
11 (0.2)
72
85
3,364 (488,000)
Molded from fine powder
41.0 (5,960)
4
11 (0.2)
72
83
Molded from pellets (as received)
38.6 (5,600)
4
16 (0.3)
77
81
Virgin PS G-18a
a
Source of material: Amoco Corp. Pull rate of 0.2 in/min. c At 200◦ C, 5 kg load. b
©2001 CRC Press LLC
Melt Flow Rate g/10 min (230◦ C, 2.16 kg) Before S3 P
After S3 P
79.7 (11,560)
16.6
26.4
3,295 (478,000)
74.1 (10,750)
16.6
33.5
3,000 (435,000)
69.6 (9,370)
15.5
18.0c
T1: SYV
Strength MPa (psi)
Char Count= 0
Modulus MPa (psi)
Flexural Properties
QC: SYV/ABE
15:13
Tensile Propertiesb Notched lzod Heat Ultimate Impact Distortion MPa Elong J/m Temperature Hardness ◦ C 264 psi Shore D (psi) % (ft-lb/in)
P2: SYV/ABE
March 2, 2001
Physical Properties of Virgin PS G-18 Injection-Molded from S3 P-Made Powder.
P1: SYV/FYX
pb035-05
TABLE 5.6.
Modulus MPa (psi)
Strength MPa (psi)
Melt Flow Rate (230◦ C, 2.16 kg)
Molded from coarse powder
35.0 26.5 (5,080) (3,840)
18
21 (0.4)
75
74
1,911 (277,200)
61.8 (8,960)
24.6
Molded from fine powder
34.6 26.1 (5,020) (3,780)
21
21 (0.4)
72
73
1,834 (266,100)
62.0 (8,990)
22.2
Repelletized from coarse powder
33.6 25.2 (4,870) (3,660)
23
21 (0.4)
73
74
1,776 (257,700)
57.4 (8,320)
26.1
Repelletized from fine powder
34.1 22.6 (4,940) (3,280)
34
21 (0.4)
75
73
1,792 (260,000)
61.0 (8,850)
21.8
Source of material: polypropylene from Equistar Chemical Co.; polystyrene from Amoco Corp. Pull rate of 0.2 in/min.
©2001 CRC Press LLC
T1: SYV
b
Hardness Shore D
Char Count= 0
a
Ultimate MPa Elong (psi) %
Flexural Properties
Heat Distortion Temperature ◦ C 264 psi
QC: SYV/ABE
Virgin 80/20 PP/PS Blenda
Yield MPa (psi)
Notched lzod Impact J/m (ft-lb/in)
15:13
Tensile Propertiesb
P2: SYV/ABE
March 2, 2001
Physical Properties of Virgin 80/20 PP/PS Blend Injection-Molded from Powder Made via S3 P Process.
P1: SYV/FYX
pb035-05
TABLE 5.7.
b
Hardness Shore D
Modulus MPa (psi)
Strength MPa (psi)
Melt Flow Rate (230◦ C, 2.16 kg)
Molded from coarse powder
32.3 (4,680)
4
11 (0.2)
75
78
2,551 (370,100)
51.6 (7,490)
40.1
Molded from fine powder
31.5 (4,570)
4
16 (0.3)
76
78
2,651 (384,500)
52.7 (7,640)
40.6
Repelletized from coarse powder
31.3 (4,540)
4
11 (0.2)
74
78
2,635 (382,300)
51.0 (7,400)
36.2
Repelletized from fine powder
31.4 (4,550)
4
11 (0.2)
74
77
2,766 (401,200)
51.4 (7,460)
47.1
Source of material: polypropylene from Equistar Chemical Co.; polystyrene from Amoco Corp. Pull rate of 0.2 in/min.
©2001 CRC Press LLC
T1: SYV
a
Elong %
Flexural Properties
Heat Distortion Temperature ◦ C 264 psi
Char Count= 0
Virgin 20/80 PP/PS Blenda
Ultimate MPa (psi)
Notched lzod Impact J/m (ft-lb/in)
QC: SYV/ABE
15:13
Tensile Propertiesb
P2: SYV/ABE
March 2, 2001
Physical Properties of Virgin 20/80 PP/PS Blend Injection-Molded from Powder Made via S3 P Process.
P1: SYV/FYX
pb035-05
TABLE 5.8.
73
1,834 (266,100)
62.0 (8,990)
24.6
71
76
2,011 (291,800)
62.1 (9,010)
23.4
68
74
1,820 (264,000)
60.3 (8,740)
22.1
S3 P Processing
34.6 (5,020)
26.1 (3,780)
21
21 (0.4)
72
Ambient grinding
35.3 (5,125)
34.7 (5,030)
10
21 (0.4)
Cryogenic grinding
35.2 (5,110)
30.8 (4,470)
14
27 (0.5)
Source of material: virgin polypropylene from Equistar Chemical Co.; virgin polystyrene from Amoco Corp. Pull rate of 0.2 in/min.
©2001 CRC Press LLC
Melt Flow Rate g/10 min (230◦ C, 2.16 kg)
T1: SYV
Strength MPa (psi)
Char Count= 0
Modulus MPa (psi)
QC: SYV/ABE
b
Hardness Shore D
Ultimate MPa (psi)
Elong %
Flexural Properties
Heat Distortion Temperature ◦ C 264 psi
Yield MPa (psi)
Virgin 80/20 PP/PS Blenda
a
Notched lzod Impact J/m (ft-lb/in)
15:13
Tensile Propertiesb
P2: SYV/ABE
March 2, 2001
Physical Properties of Virgin 80/20 PP/PS Blend Injection-Molded from Powder as a Function of the Size-Reduction Method.
P1: SYV/FYX
pb035-05
TABLE 5.9.
b
Elong %
Hardness Shore D
Modulus MPa (psi)
Strength MPa (psi)
Melt Flow Rate g/10 min (230◦ C, 2.16 kg)
S3 P Processing
31.5 (4,570)
4
16 (0.3)
76
78
2,651 (384,500)
52.7 (7,640)
40.1
Ambient grinding
33.3 (4,830)
4
16 (0.3)
72
78
2,739 (397,340)
51.0 (7,400)
21.7
Cryogenic grinding
37.0 (5,360)
5
16 (0.3)
73
78
2,645 (383,680)
58.3 (8,450)
25.7
Source of material: virgin polypropylene from Equistar Chemical Co.; virgin polystyrene from Amoco Corp. Pull rate of 0.2 in/min.
©2001 CRC Press LLC
T1: SYV
a
Ultimate MPa (psi)
Flexural Properties
Heat Distortion Temperature ◦ C 264 psi
Char Count= 0
Virgin 20/80 PP/PS Blenda
Notched lzod Impact J/m (ft-lb/in)
QC: SYV/ABE
15:13
Tensile Propertiesb
P2: SYV/ABE
March 2, 2001
Physical Properties of Virgin 20/80 PP/PS Blend Injection-Molded from Powder as a Function of the Size-Reduction Method.
P1: SYV/FYX
pb035-05
TABLE 5.10.
P1: SYV/FYX pb035-05
P2: SYV/ABE
March 2, 2001
QC: SYV/ABE
15:13
T1: SYV
Char Count= 0
As expected, the 20/80 PP/PS blend was more brittle than the blend containing 80 percent by weight of PP. The difference in melt flow rate for the powders made by three different grinding techniques was more pronounced for the 20/80 PP/PS blend: a MFR of 40.1 g/10 min was measured for S3 P-made powder, double that of the MFR value measured for ambiently ground powder. It is believed that the reason for this increase in melt flow is the chain scission caused by a combination of shear and pressure during the pulverization process [10]. To demonstrate the ability of S3 P to produce powder of different particle size and particle size distribution, a wide variety of resins has been converted into powder. These resins include LDPE, LLDPE, HDPE, PP, PS, HIPS, PET, PC, nylons, UHMWPE, acrylonitrile butadiene styrene (ABS), PVDF, and others. It has been shown that virgin LDPE, LLDPE, HDPE, and PP, which are widely used for rotational molding applications, have been successfully processed into 35 mesh (500 microns) powders [10]. Sieve analysis of S3 P-made LLDPE powders vs. the same powder produced by cryogenic grinding is presented in Table 5.11. Particle size distribution is similar for both size-reduction methods, but S3 P powder was made in one step in constrast to cryogenic grinding, which involves multiple steps. In addition, LLDPE powder made via S3 P has a “cauliflower” morphology, indicative of greater surface area than is found in cryoground LLDPE powder (Figures 5.6a,b and 5.7a,b), which has a smaller surface area. Ahn et al. studied microstructural changes in virgin PP and PS induced by solid-state pulverization during powder production [11]. The change in morphology was illustrated by the polarized light optical micrographs of meltcrystallized virgin PP thin films shown in Figure 5.8a,b. A drastic reduction in spherulite size of about 100-fold can be observed in the post-S3 P polypropylene, TABLE 5.11.
Sieve Analysis of Virgin LLDPE Resin as a Function of the Size-Reduction Method.
S3 P Process
Cryogenic Grinding
Mesh
Microns
Weight Percent Retained
35 50 80 120
500 300 180 125
1.40 45.20 30.80 12.60
200 270 450 Fines
75 53 32
7.90 1.80 0.40 0
Size
©2001 CRC Press LLC
Size Mesh
Microns
35 50 80 120 140 200 270
500 300 180 125
Fines
75 53
Weight Percent Retained 0.75 39.25 34.50 11.50 6.75 3.25 2.00 1.00
P1: SYV/FYX pb035-05
P2: SYV/ABE
March 2, 2001
QC: SYV/ABE
15:13
T1: SYV
Char Count= 0
(a)
(b)
FIGURE 5.6 (a) SEM image of pulverized virgin LLDPE at 40×. (b) SEM image of pulverized virgin LLDPE at 250×.
(a)
(b)
FIGURE 5.7 (a) SEM image of virgin LLDPE after cryogenic grinding at 40×. (b) SEM image of virgin LLDPE after cryogenic grinding at 250×.
©2001 CRC Press LLC
P1: SYV/FYX pb035-05
P2: SYV/ABE
March 2, 2001
QC: SYV/ABE
15:13
T1: SYV
Char Count= 0
(a)
(b) FIGURE 5.8 (a) Optical micrograph of melt-crystallized thin films of unpulverized virgin PP under polarized light. (b) Optical micrograph of melt-crystallized thin films of pulverized virgin c 1995, John PP under polarized light. Reprinted with permission from Reference [11]. Copyright Wiley & Sons, Inc.
©2001 CRC Press LLC
P1: SYV/FYX pb035-05
P2: SYV/ABE
March 2, 2001
QC: SYV/ABE
15:13
T1: SYV
Char Count= 0
resembling the effect of adding a nucleating agent [11]. Differential scanning colorimetry (DSC) data of pulverized PP indicated the presence of a small amorphous region from changes in chain structure would create a different type of heterogeneous nucleation site for crystallization [11]. It has been reported in the literature that the comminution of polymers is known to cause the destruction of crystalline structure [12–14]. Nesarikar et al. reported the formation of detectable concentration of long chain branches (LCB) in virgin LLDPE powder due to S3 P processing [15]. Nuclear magnetic resonance (NMR) data indicated that the concentration of LCBs per 1,000 backbone carbon atoms rose from 0.2 to 2 for virgin LLDPE. Furgiuele et al. demonstrated that polymer type, molecular weight, morphology, and shear conditions during pulverization are important factors affecting chain scission [16,17]. Gel permeation chromatography (GPC) data generated for low molecular weight PS under low shear conditions did not show significant reduction in molecular weight (corresponding to significant chain scission, as illustrated in Figure 5.9a). However, when higher molecular weight PS (Mn 110,000 and Mw 205,000) was pulverized, a major reduction in molecular weight was observed, indicating that substantial chain scission took place (Figure 5.9b). Similar effects have resulted from measuring the melt flow rate of virgin PS before and after pulverization. High molecular weight PS showed a seven-fold increase in MFR (from 2.03 g/10 min to 14.2 g/10 min at 230◦ C, 2.16 kg load) as a result of solid-state shear processing [16]. Ganglani et al. studied mechanochemical alteration of several virgin polyolefins during S3 P using the laboratory-scale PT-25 pulverizer [18]. The polyolefins included HDPE (MFR 0.3 g/10 min at 190◦ C), LDPE (MFR 0.25 g/10 min at 190◦ C), and two LLDPEs. One LLDPE-GA copolymer was based on hexene co-monomer (MFR 3.50 g/10 min at 190◦ C) and the other LLDPE-NG copolymer was based on octene co-monomer (MFR 4.5 g/10 min at 190◦ C). Each was processed under low shear conditions (with a 23-mm screw in the 25-mm diameter barrel) and under high shear conditions (with a 25-mm screw in the 25-mm diameter barrel). It has been determined (Figures 5.10a–d) that the low-shear pulverization did not show a significant change in rheology, indicating that few, if any, structural changes occurred. This implies that the stress applied during low-shear S3 P did not exceed the critical shear stress required for chain scission and free radical generation. In contrast, high shear pulverization resulted in a change in viscosity for HDPE and both LLDPE samples. These two polyolefins exhibited lower MFR than unprocessed polymers, and they also showed a higher viscosity in oscillatory shear rheometry at low shear rates. This indicates that under high-shear S3 P, the minimum stress requirement for chain scission is exceeded, allowing for a generation of free radicals. These radicals can potentially migrate and recombine to cause structural changes, such as long chain branching, that increase viscosity.
©2001 CRC Press LLC
P1: SYV/FYX pb035-05
P2: SYV/ABE
March 2, 2001
QC: SYV/ABE
15:13
T1: SYV
Char Count= 0
FIGURE 5.9 (a) GPC chromatogram of virgin PS before S3 P processing at moderately high shear [16]. (b) GPC chromatogram of virgin PS after S3 P processing at moderately high shear [16].
©2001 CRC Press LLC
P1: SYV/FYX pb035-05
P2: SYV/ABE
March 2, 2001
QC: SYV/ABE
15:13
T1: SYV
Char Count= 0
LLDPE-NG showed the greatest increase in apparent viscosity upon highshear S3 P, followed by LLDPE-GA and then HDPE. The explanation of this trend relies on the LLDPE structure. LLDPE has a greater number of short chain branches along the main chain than HDPE and therefore LLDPE is more prone to chain scission, resulting in the creation of more free radicals and branching possibilities than HDPE. The difference in the extent of viscosity change between the two LLDPE samples may be due to heterogeneous catalysts used to make these copolymers, which create chains with random distribution of branches and branch content. Ganglani et al. observed no significant changes either in molecular weight or in the crystallinity of HDPE, LDPE, and the two LLDPE samples [18]. Their
FIGURE 5.10 (a) Log of viscosity vs. log of shear rate for virgin HDPE after S3 P processing. (b) Log of viscosity vs. log of shear rate for virgin LDPE after S3 P processing. (c) Log of viscosity vs. log of shear rate for virgin LLDPE-GA after S3 P processing. (d) Log of viscosity vs. log of shear rate for virgin LLDPE-NG after S3 P processing. Reprinted with permission from Reference c 2000 SPE. [18].
©2001 CRC Press LLC
P1: SYV/FYX pb035-05
P2: SYV/ABE
March 2, 2001
QC: SYV/ABE
15:13
T1: SYV
Char Count= 0
FIGURE 5.10
(Continued )
study indicated that the molecular weight and thermal behavior was unaltered by S3 P under processing conditions studied. It was concluded that high-shear S3 P led to a small but measurable increase in long chain branching as suggested by melt-rheology experiments. The S3 P process has also been shown to be a novel technique to make powdercoating compounds without melting. Powder coating for surface protection is among the most rapidly growing applications for polymeric powders. Standard thermoset polyester and epoxy/polyester hybrid powder coating compounds containing pigments, curing agents, catalysts, and other additives have been made successfully with the laboratory-scale PT-25 pulverizer in one continuous step, resulting in fine powder of 20 mesh (210 microns) in size. These S3 P-made powder coating compounds have been applied to metal surfaces, and the results have been compared with coatings produced by conventional methods. The surface appearance of the S3 P-made compounds and impact strength values were comparable with those made by the traditional melt-extrusion process followed by grinding.
©2001 CRC Press LLC
P1: SYV/FYX pb035-05
P2: SYV/ABE
March 2, 2001
QC: SYV/ABE
15:13
T1: SYV
Char Count= 0
FIGURE 5.10
©2001 CRC Press LLC
(Continued )
P1: SYV/FYX pb035-05
P2: SYV/ABE
March 2, 2001
QC: SYV/ABE
15:13
T1: SYV
Char Count= 0
Although there are several extensive reviews related to mechanochemistry (stress-induced reactions in polymers), deformation and fracture of polymers under the application of mechanical energy are accompanied by very complex phenomena that are still not fully understood [13,14]. According to Porter and Casale, the molecular weight of a polymer is an important parameter in mechanochemistry [19]. With increased molecular weight, the conversion of the mechanical energy applied to a polymer leads to a molecular energy storage that, in turn, causes changes in molecular architecture, chain elongation, bond bending, and eventually the rupture of bonds. The generation of free radicals results in an alteration of molecular weight and its distribution, branching, and cross-linking. Polymers with long chain branching, such as LDPE, are highly prone to stress-induced reactions due to the branch points, which have high stress concentration upon deformation. Depending on shear, an increase in molecular weight causes an increase in viscosity (decrease in MFR). Monitoring the extent of stress-induced reaction in polymers is rather difficult. It was established by several investigators that electron spin resonance (ESR) is a useful experimental technique for studying the reactivity of free radicals produced by fracture or deformation [20]. In the absence of air, the possibility of primary radical detection is enhanced. Most mechanochemical processes reduce molecular weight [13]. In the case of shearing rubber and some polyolefins, however, cross-linking and branching reactions can predominate and thus lead to a molecular weight increase. The temperature of mechanochemical reactions affects bond rupture. Increasing temperature softens the polymer, decreases stress, and reduces the energy input, which in turn reduces rupture and leads to an increase in molecular weight. At lower temperatures, chain motions are reduced, the relaxation processes are slower, and disentangling of chains takes longer. Therefore, the rate and extent of mechanochemical reactions are higher at lower temperatures. It is believed that pulverization involves changes in particle size and shape, rupture of chemical bonds, change in crystallinity, and change in molecular weight as a result of formation of free radicals and modification of properties. The extent of these changes depends on the polymer type and the degree of shear introduced by mechanical energy input, as well as on the processing parameters during pulverization. The relationship between particle size of the virgin LDPE powder and the MFR (molecular weight) has been studied [10]. Sieve analysis of LDPE with two different MFRs is shown in Tables 5.12 and 5.13. These data indicate that the sample of LDPE with a higher MFR (lower Mw) has more fine powder than the LDPE sample with a lower MFR (higher Mw). Since 1998, the S3 P technology has been in transition to a commercial scale at the Polymer Technology Center at Northwestern University. This effort has focused on demonstrating both the process and the pulverization equipment for specific applications using a variety of feedstocks. The first production-scale pulverizer PT-60 (60-mm diameter; length/diameter ratio = 28) has been built by Berstorff, a division of Krauss-Maffei Corporation. This machine is based
©2001 CRC Press LLC
P1: SYV/FYX pb035-05
P2: SYV/ABE
March 2, 2001
QC: SYV/ABE
15:13
TABLE 5.12.
T1: SYV
Char Count= 0
Sieve Analysis of Powder Resulting from S3 P of Virgin LDPE 134.236 (MFR 2.1 g/10 min). S3 P Process Size Mesh
Microns
35 50 80 120 140 200 270 Fines
500 300 180 125 105 75 53
Weight Percent Retained 25.70 25.10 22.90 13.40 7.60 5.50 2.20 0.20
Source of material: Exxon Chemical.
on the Ultra Torque® series in which the torque is transferred by means of spline shafts; this yields an increased output rate as compared with the standard ZE series. The PT-60A pulverizer has a feed system, modular barrel, and screw configuration with open discharge. It has been used to scale-up the S3 P process for LDPE powder production, blends of virgin polymers with unmatched viscosity, mixed-color commingled post-consumer plastics, and rubber powder production. Polymeric powder made with the production-size PT-60A pulverizer has a different particle shape and different dry-flow properties than that of powder produced earlier with a laboratory-scale PT-25 pulverizer. Although particles do have an elongated shape as a result of shearing, the particles made with PT-60A are rounder and do not have “tails.” The differences in shape of virgin LDPE powders can be seen from Figures 5.11a,b and 5.12a,b. TABLE 5.13.
Sieve Analysis of Powder Resulting from S3 P of Virgin LDPE 509.48 (MFR 70 g/10 min). S3 P Process Size Mesh
Microns
35 50 80 120 140 200 270 Fines
500 300 180 125 105 75 53
Source of material: Exxon Chemical.
©2001 CRC Press LLC
Weight Percent Retained 11.20 16.80 27.10 25.60 10.10 8.30 0.80 0.10
P1: SYV/FYX pb035-05
P2: SYV/ABE
March 2, 2001
QC: SYV/ABE
15:13
(a)
T1: SYV
Char Count= 0
(b)
FIGURE 5.11 (a) SEM image of virgin LDPE pulverized with PT-25 at 40×. (b) SEM image of virgin LDPE pulverized with PT-25 at 250×.
(a)
(b)
FIGURE 5.12 (a) SEM image of virgin LDPE pulverized with PT-60 at 40×. (b) SEM image of virgin LDPE pulverized with PT-60 at 250×.
©2001 CRC Press LLC
P1: SYV/FYX pb035-05
P2: SYV/ABE
March 2, 2001
QC: SYV/ABE
15:13
T1: SYV
Char Count= 0
Dry-flow properties depend primarily on the size and shape of the powder particles. The absence of “tails” increases dry flow, which in turn reduces voids. For the LDPE powder of 35 mesh (500 microns), dry flow was measured as 38.9 sec (time required for 100 g of powder to flow through a funnel of standard dimensions in accordance with ASTM 1895-89). It has been reported by McDaid and Crawford that powder dry-flow properties are important during rotational molding because they affect the uniform distribution of the polymer in the mold [21].
SUMMARY Plastic powders are used in a wide array of applications, including but not limited to powder coatings, sintering, rotational molding, and compounding with additives. Conventional grinding methods for producing powders based on compression or impact have several disadvantages. The efficiency of comminution equipment decreases as the size of the powder decreases, for it is difficult to apply stress to finer powders. Moreover, during grinding operations, most polymers become overheated and discolored. The limitations and disadvantages of traditional comminution open new opportunities for advanced, cost-effective processes of powder production. The novel technology of S3 P creates powders through shear deformation of the polymeric feedstock while cooling maintains the material in a solid state. This particle formation mechanism is drastically different from the mechanism of grinding, where rotating hammers or vibrating balls impact the polymeric material. The fragmentation is believed to occur through a release of stored elastic strain energy resulting from the high shear deformation of the plastic during pulverization. The researchers at Northwestern University’s Polymer Technology Center have produced a wide variety of plastic powders via the S3 P process on a laboratory scale. Testing of S3 P-made powders has shown that these powders are suitable for direct-melt conversion by all existing plasticsfabrication techniques. By controlling process parameters and screw configuration, the S3 P process can produce powders of particle sizes ranging from coarse (10 mesh/2,000 microns) to very fine (625 mesh/20 microns). The distribution of particle size is affected by both process parameters and inherent properties of the plastics, such as crystallinity, melt flow rate, resin family (thermoset or thermoplastic), and whether virgin or recycled material is treated. The ability to produce fine powders from thermoplastics and thermosets makes the S3 P process very attractive for powder-coating applications. Because polymers are maintained in the solid state in this process, there is no premature curing or catalytic reaction that ordinarily takes place during the preparation of thermoset powder-coating compounds by conventional melt extrusion. The S3 P process is able to produce 35 mesh/500 micron powders in one step for
©2001 CRC Press LLC
P1: SYV/FYX pb035-05
P2: SYV/ABE
March 2, 2001
QC: SYV/ABE
15:13
T1: SYV
Char Count= 0
rotational molding applications. Several polyolefins, such as virgin LDPE, LLDPE, HDPE, and PP, widely used in rotational molding, have been successfully converted into powders with particle size and particle size distribution similar to those of conventionally ground powders. It has also been demonstrated that S3 P is successful in the production of powders from recycled feedstocks. S3 P has a unique advantage in that it produces high-quality powders of leveled (or uniform) color (often light pastels) from multicolor feedstock; that is, it can produce useful powders from feedstocks that are not sorted by color. Besides particle size, particle shape and texture are also important in determining the utility of plastic powder in specific applications. S3 P-made powders have a “cauliflower” morphology, indicative of a large surface area that is favored in many applications. By contrast, powders made by cryogenic grinding yield flat particles with smaller surface areas. Researchers at Northwestern University have worked on the transition of S3 P to a commercial scale since 1998, when a one-of-a-kind, first production-size PT-60 pulverizer built by Berstorff, the German manufacturer, became operational at the University’s Polymer Technology Center. This new technology offers enhanced opportunities for producing polymeric powders of various sizes from both virgin and recycled feedstocks for numerous applications. These include powder-coating compounds, rotational molding, dispersion of additives, and recycling. In addition, S3 P is unique in that it makes possible the in-situ compatibilization of dissimilar polymers in the solid state without an addition of pre-made compatibilizing agents. It has been demonstrated, however, that S3 P is capable of efficient mixing of blends with different component viscosities. The virgin PP blends with unmatched viscosity exhibited exceptionally high elongation at break coupled with improved processability when compared with those blends produced by conventional melt extrusion [22]. Continuing research at Northwestern University has focused on a better understanding of the extent of mechanochemical reactions during pulverization in order to facilitate the commercialization of this breakthrough process. This technology constitutes a new platform for the development of engineered blends as well as the creation of new end-uses for existing materials. REFERENCES 1. Richart, D. S. Encyclopedia of Chemical Technology (4th ed.), New York: John Wiley & Sons, Inc., 6:606–661 (1993). 2. Beall, G. L. Rotational Molding: Design, Materials, Tooling, and Processing, Cincinnati, Ohio: Hanser-Gardner Publications, Inc. (1998). 3. Richart, D. S. “Powder Coating,” Polymer Powder Technology, New York: John Wiley & Sons, Inc. (1995). 4. Dodge, P. T. “Rotational Molding,” Encyclopedia of Polymer Science and Engineering, New York: John Wiley & Sons, Inc., 14:659–670 (1988).
©2001 CRC Press LLC
P1: SYV/FYX pb035-05
P2: SYV/ABE
March 2, 2001
QC: SYV/ABE
15:13
T1: SYV
Char Count= 0
5. Crawford, R. J. “Introduction to Rotational Moulding,” Rotational Moulding of Plastics, New York: John Wiley & Sons, Inc. (1996). 6. Kliene, R. I. “Rotational Moulding of Poly(ethylene),” Rotational Moulding of Plastics, New York: John Wiley & Sons, Inc. (1996). 7. Throne, J. L. “Rotational Molding,” Polymer Powder Technology, New York: John Wiley & Sons, Inc. (1995). 8. Throne, J. L. and Sohn, M. S. “Characterization of Rotational Molding Grade Poly(ethylene) Powders,” Advanced Polymer Technology, 9:181–192 (1989). 9. Khait, K., Henderson, V., Nichols, G. N., and Christ, B. Unpublished data, Northwestern University, Evanston, Illinois (1995). 10. Khait, K. and Torkelson, J. M. “A New Polymeric Powder Technology: Solid-State Shear Pulverization Process,” Proceedings of the Advanced Technologies for Particle Processing— Topical Conference at the AIChE Fall Meeting, November 11–20, 1998, Miami Beach, Florida. 11. Ahn, D., Khait, K., and Petrich, M. A. “Microstructural Changes in Homopolymers and Polymer Blends Induced by Elastic Strain Pulverization,” Journal of Applied Polymer Science, 55:1431– 1440 (1995). 12. Zhang, R., Zheng, H., Lou, X., and Ma, D. “Crystallization Characteristics of Polypropylene and Low Ethylene Content Polypropylene Copolymer with and without Nucleating Agents,” Journal of Applied Polymer Science, 51:51 (1994). 13. Baramboim, N. K. “Mechanochemistry of Polymers,” Rubber and Plastic Research Association of Great Britain, London: MacLaren & Sons, Ltd. (1964). 14. Casale, A. and Porter, R. S. Polymer Stress Reactions, New York: Academic Press (1978). 15. Nesarikar, A. R., Carr, S. H., Khait, K., and Mirabella, F. M. “Self-Compatibilization of Polymer Blends via Novel Solid-State Shear Extrusion Pulverization,” Journal of Applied Polymer Science, 63:1179–1187 (1997). 16. Furgiuele, N., Khait, K., and Torkelson, J. M. “Novel Approach for the Compatibilization of Polymer Blends and Polymeric Waste,” Proceedings of AIChE, November 11–20, 1998, Miami Beach, Florida. 17. Furgiuele, N., Khait, K., and Torkelson, J. M. “SSSP: A New Processing Technology for Recycling Single and Commingled Polymers,” Proceedings of AIChE, March 8–12, 1999, New Orleans, Louisiana. 18. Ganglani, M., Torkelson, J. M., Carr, S. H., and Khait, K. “Trace Levels of Mechanochemical Effects in Pulverized Polyolefins,” submitted to Journal of Applied Polymer Science (2000). 19. Porter, R. S. and Casale, A. “Mechanochemical Reactions,” Encyclopedia of Polymer Science and Engineering—2nd ed., New York: John Wiley & Sons, Inc., pp. 467–484 (1987). 20. Sohma, J., Kawashima, T., Shimada, S., Kashiwabara, H., and Sakaguchl, M. “ESR Studies on the Polymer Radicals Produced by Mechanical Repture,” Proceedings of the 22nd Nobel Symposium, June 20–22, 1972, S¨odergarn, Liding¨o, Sweden. 21. McDaid, J. and Crawford, R. J. “The Grinding of Poly(ethylene) for Use in Rotational Molding,” Rotation, Spring 1997, pp. 27–34. 22. Khati, K. and Torkelson, J. M. “A New Polymer Processing Technology for Polymer Blends with Unmatched Viscosity: Solid-State Shear Pulverization (S3 P),” International Polymer Processing, 15:343–347 (2000).
©2001 CRC Press LLC
P1: FUY pb035-06
March 15, 2001
11:23
Char Count= 0
CHAPTER 6
S3 P Technology and Polymer Blends
P
are blended by many methods, depending upon available equipment, prevailing economics, and what is being combined. Although the resulting materials are alloys, it is exceedingly rare that they exhibit properties with values sufficient to qualify them for demanding applications. In this light, the S3 P process merits special attention. It represents a truly novel approach to the creation of polymer blends because of its heretofore-unattainable versatility and simplicity of operation. Of particular note is its ability to create a wide range of useful materials, especially OLYMERS
(1) Mixtures of the same polymer, but of constituents having different viscosities (2) Mixtures of closely similar polymers, such as ethylene copolymers with only differences in comonomer sequence distribution or only differences in long-chain branching (3) Mixtures of dissimilar, immiscible polymers Ordinarily, such multicomponent systems are understood to combine in such a way as to make solids with properties significantly poorer than those of the constituent polymers. This situation applies to efforts to achieve blends of certain polymers, and it is also encountered in recovering plastics from commingled industrial or post-consumer waste streams. S3 P, on the other hand, does not have this undesirable outcome and leads instead to value-added alloys whose properties often surpass those of the constituent materials [1,2].
©2001 CRC Press LLC
P1: FUY pb035-06
March 15, 2001
11:23
Char Count= 0
COMPATIBILIZATION OF POLYMER BLENDS The vast majority of commercial polymeric materials are obtained by blending component polymers so as to achieve a set of properties desired by the customer and the supplier. Blending, therefore, is a common method of achieving the different grades of polymers to satisfy the requirements of different market segments. Often the polymers that are blended together have essentially the same chemical microstructures. Thus they are miscible with each other, and the final properties are simply an average of those of the constituents. The purpose of mixing different kinds of the same polymer is usually to create a specific “grade” that meets customer needs, such as processability and service performance. The mixing of these is typically done in the melt, usually with a compounding extruder. Similarly, the addition of small amounts of a polymer with a different chemical composition (or even just a different chemical microstructure) is usually accomplished by blending in the melt. If, however, the viscosities and/or the elastic moduli of the component polymers are significantly mismatched, then melt processing disperses the constituents into insufficiently small domains, such that a homogeneous, valueadded material does not result. The strongly exponential dependence of melt viscosity on molecular weight makes the mixing of two masses of the same polymer behave as if it were a two-phase system. Even when the low-viscosity material is the minor constituent, it will become the continuous phase during processing, simply because of the hydrodynamic effect, viz. the fact that it flows the most. Excessive melt processing would be necessary to achieve the phase inversion that is associated with the high-viscosity major component finally becoming the continuous phase. This could involve melt mixing times of 10–30 minutes for poly(ethylene) pairs with a ratio of viscosities on the order of 300 [3]. This means that only one of these “phases” will experience significant deformation during flow and that the other phase will merely undergo translation. None of the droplet formation events necessary for producing ever-thinner domains of one constituent will occur to any appreciable extent. What is usually done, then, is to employ a commercially disadvantageous mixing method—most commonly, solvent mixing. The undesirability of solvent blending stems from the need for extraordinarily large quantities of solvents, high energy needs for drying and solvent recovery, and the necessity of a post-treatment to get the desired final powder particles. The alternative is to use S3 P processing of the material just before meltblending [2]. When mixing two polymers together (as seen in Figure 6.1), the initial charge of material starts to be melt-blended—not much torque develops because the unconsolidated mass does not adhere well to the moving surfaces and because the mass dissipates energy only by the particles moving past each other rather than by experiencing any deformation themselves. As the particles warm in the melting apparatus, they will naturally become more compliant
©2001 CRC Press LLC
P1: FUY pb035-06
March 15, 2001
11:23
Char Count= 0
FIGURE 6.1 The production of a homogeneous melt mass can be followed by torque to drive the mastication process. Here an alloy 92% polystyrene/8% poly(ethylene) is processed from mixtures that had and had not been previously subjected to S3 P pulverization.
and ultimately take on a more viscous, rather than elastic, character. Thus the particles will begin to “soften” and warm further as they dissipate the available mechanical energy and ultimately fuse together. This transition is easily detected as a rapid rise in consumption of mechanical energy [3]. When the material charged into a melt mixer is a powder made by the S3 P process, then the rapid rise in torque indicating the fusion of the entire in-process mass occurs very quickly. Of course, the mixing of polymers with dissimilar chemical microstructures will almost always mean that the resulting blend is a two-phase mixture, because such component polymers are only rarely miscible [4]. Consequently, melt mixing can be employed to make an alloy, but it is extremely difficult to attain a dispersion that will make a useful polymeric material. To achieve a useful blend based on immiscible polymers, it is necessary to strengthen the interfaces between the phases that make up the alloy. If the load-transferring capabilities across these interfaces can be raised to values that match those of the constituent phases, then a useful, multicomponent alloy can be made. The tactic usually employed is to add another substance that spontaneously diffuses to these interfaces and imparts load-transferring capability to them. These substances, called compatibilizers, are usually polymers, and they achieve the
©2001 CRC Press LLC
P1: FUY pb035-06
March 15, 2001
11:23
Char Count= 0
FIGURE 6.2 A schematic drawing showing how linear and graft block copolymer chains will become located with their interblock junction coincident with the interface between two polymeric phases. In this way they serve as a surfactant and provide a covalent bond bridging the two phases. c 1984 SPE. Reprinted with permission from Reference [5].
desired effect by being linear arrays of covalent bonds that bridge across the interface (Figure 6.2). Macromolecules of such compatibilizers are often random, block, or graft copolymers made of the constituents of the immiscible polymer pair (Figure 6.3) being blended [5–7]. Most, though not all, compatibilizers are expensive, and this tactic is inefficient, because much of the compatibilizer resides throughout the alloy, rather than being a monolayer located where it is needed: just at the interface. The addition of a polymer that acts as a surfactant and thereby imparts some measure of compatibilization is discussed by Machado and Lee [6]. There it is shown (Figures 6.4 and 6.5) that poly(methylmethacrylate) (PMMA) is effective in the blends SMA/SAN, SMA/ABS, SAN/PVDF, and ABS/PVDF [SMA/styrene-maleic anhydride network-former, SAN/styrene-acrylonitrile copolymer, ABS/acrylonitrile butadiene styrene impact plastic, and PVDF/ poly(vinylidenefluoride) homopolymer]. Other reviews can be found elsewhere [4,8,9]. Lustiger, Marzinsky, and Mueller reported this strategy of using a copolymer to improve the properties of the immiscible blend, poly(ethylene) and polypropylene [10]. Compatibilizers that operate on the basis of physical/ chemical mechanisms can act as compatibilizers for immiscible polymer pairs, as has been reviewed by Xanthos [11]. In some cases, the compatibilizer is seen to operate as a coupling agent, involving ionic interactions, rather than the more common covalent bonds or van der Waals potentials.
©2001 CRC Press LLC
P1: FUY pb035-06
March 15, 2001
11:23
Char Count= 0
FIGURE 6.3 An example of how much surfactant block copolymer is needed to develop (as measured by tensile strength) compatibilization by alloying an ethylene/propylene block copolymer (Epcar® 847) with varying amounts of polypropylene with poly(ethylene). Reprinted with c 1984 SPE. permission from Reference [5].
The underlying concept that helps us understand how such compatibilizers work involves the way in which they bridge across a polymer/polymer interface. Chemically heterogeneous macromolecules typically locate at the interface between two immiscible polymer phases. Such polymers will function as compatibilizers if portions of them will diffuse into the phases on each
©2001 CRC Press LLC
P1: FUY pb035-06
March 15, 2001
11:23
Char Count= 0
FIGURE 6.4 An example of the development of tensile strength due to the incorporation of a homopolymer (polymethylmethacrylate, PMMA) as a surfactant in alloys of two styrenic copolymers (92% styrene/8% maleic anhydride and 77% styrene/23% acrylonitrile). Again, substantial amounts of surfactant are needed to achieve good levels of compatibilization. Reprinted with permission c 1993 SPE. from Reference [6].
FIGURE 6.5 An example of the development of ductility (>200% strain to failure) in alloys of polyvinylidene fluoride with acrylonitrile/butadiene/styrene (impact plastic) due to the inclusion of a large (10%) amount of surfactant, in this case PMMA. Reprinted with permission from Reference c 1993 SPE. [6].
©2001 CRC Press LLC
P1: FUY pb035-06
March 15, 2001
11:23
Char Count= 0
FIGURE 6.6 A schematic diagram of concentration profiles of polymeric constituents at an interface. Polymer 1 is phase separated with polymer 2, and their concentration drops off precipitously upon crossing the interface and entering the phase of the other. A surfactant polymer (constituent 3) is localized at the interface, with its concentration falling steeply at distances away from the interface. Note, however, that there are finite amounts of all three constituents throughout the interface region. The thinner the interface region, the less surfactant (compatibilizer) is needed to develop good properties with a polymer 1/polymer 2 alloy. Reprinted with permission from Reference [6]. c 1993 SPE.
side of the interface, thereby becoming anchored in place across the interface. Such portions may be side-branches on polymeric chains, or they may be part of main-chain backbones. These portions may have a chemical composition that matches the phase into which they penetrate, or, alternatively, they might have a dissimilar composition that, nevertheless, interacts favorably through secondary bonding with the surrounding chains (Figure 6.6). The continuity of covalent bonds across the interface reduces the interfacial energy, and it provides the improved mechanical connectivity that is associated with compatibilization [5]. THE STRATEGY OF SELF-COMPATIBILIZATION Self-compatibilization means that the desired surfactant chains are formed spontaneously within the system of polymers being blended; it is not necessary
©2001 CRC Press LLC
P1: FUY pb035-06
March 15, 2001
11:23
Char Count= 0
to incorporate additional polymeric substances in order to obtain the desired compatibilization. Typically, tensile strength and ductility of alloys formed in a self-compatibilized alloy obey a weighted average-of-reciprocals relationship: 1 1 1 − 1 = + Pa P1 P2 where Pa is a property of the alloy, P1 and P2 are those properties of the individual components, and 1 is volume fraction of component 1. In some cases, the properties of the alloy exceed those of either component polymer [12]. The polymeric species responsible for this self-compatibilization are generated in situ; they must be new species synthesized by some chemical means. The presence of such self-compatibilizing species, acting in their capacity as surfactants, lead to enhanced dispersion of the component phases. The primary manifestation of this effect is the creation of a dispersion of the minority phase, characterized by a narrower particle size distribution and a smaller mean particle size. Both qualities contribute to improved mechanical properties, as is so often seen. Reactive pairs of polymers, such as Nylon 6 and impact styrenics, can readily undergo self-compatibilization, as reported by Triacca and colleagues (Figure 6.7) [13]. Their additives were miscible in the styrenic phase and were reactive in the nylon phase. The reactive additives were SAN/SMA copolymers of co-monomer composition appropriate for establishing the requisite miscibility. The anhydride functional groups of these copolymers can react with primary amine groups, as occur in the polyamide phases. High-quality dispersions and good physical properties are achieved, although the economic consequences of this strategy make it uncompetitive in the marketplace. George et al. report (see Figure 6.8) that addition of phenol, under proper conditions, can lead to the creation of diblock copolymers at the interfaces in alloys of polypropylene and NBR elastomer [9]. DiLorenzo and Frigione published a helpful review of the work involving self-compatibilization in reactive pairs [14]. Intense mechanical shear fields imposed on in-process polymeric materials offer the opportunity to achieve exceptionally good mixing while simultaneously causing mechanochemistry to occur and thereby accomplish selfcompatibilization—and without any need to add reagents or limit the process to reactive polymer pairs. This is because such methods create compatibilizer macromolecules built solely of those polymer chains that are right at the interface [15]. One method that has been employed recently by Smith and colleagues is cryogenic blending [16]. However, S3 P is a much more versatile method for achieving self-compatibilization and at the same time making a fine dispersion of the constituents [17,18].
©2001 CRC Press LLC
P1: FUY pb035-06
March 15, 2001
11:23
Char Count= 0
FIGURE 6.7 Ductility and impact strength for a 60% Nylon 6/40% rubber-grafted styreneacrylonitrile (SAN-g) alloy. The reactive compatibilizer is 14% maleic anhydride copolymerized (SMA-14) with styrene. Reprinted from Reference [13], Copyright 1991, with permission from Elsevier Science.
©2001 CRC Press LLC
P1: FUY pb035-06
March 15, 2001
11:23
Char Count= 0
FIGURE 6.8 Strength of polypropylene containing 30% of the amine-terminated terpolymer, acrylonitrile/butadiene/methylacrylate, compatibilized with varying amounts of phenolic-modified polypropylene (Ph-PP) or maleic anhydride-modified polypropylene (MA-PP). Reprinted from Reference [9], Copyright 1995, with permission from Elsevier Science.
Furthermore, the creation of self-compatibilizing species, because of their surfactant nature, aids in making beneficial microdispersions that ensure homogeneity; such microdispersions are also responsible for the attractive pastel coloration exhibited by all S3 P-processed alloys. A vast array of polymer blends has been prepared using S3 P, and reference is made to them elsewhere in this book. Recent work includes such interesting systems as a glassy polymer blended with a semicrystalline polymer (Figure 6.9). Here, one sees that S3 P of polystyrene (PS) and polypropylene creates highquality alloys [19].
©2001 CRC Press LLC
P1: FUY pb035-06
March 15, 2001
11:23
Char Count= 0
FIGURE 6.9 Tensile properties of polystyrene blended with varying amounts of low density poly(ethylene) (LDPE); these alloys were repeatedly subjected to S3 P processing for the number of cycles indicated in the inset (from Reference [19]).
Mechanochemical effects, such as a drop in molecular weight, occur when the S3 P process is operated under high-shear conditions. The dispersions resulting are spherical droplets in a matrix of the majority component. If the minority phase is polypropylene, then its crystallization will be altered if it is in very small domains. This effect relates primarily to the circumstance that nucleating heterogeneities will become statistically sparsely distributed in these exceptionally small domains. This effect was not achieved when the PS was alloyed with a different semicrystalline polymer: linear low-density poly(ethylene) (LLDPE). On the contrary, PS blended with low-density poly(ethylene) (LDPE) using S3 P
©2001 CRC Press LLC
P1: FUY pb035-06
March 15, 2001
11:23
Char Count= 0
(a)
(b)
FIGURE 6.10 Scanning electron micrographs of an alloy (15% polystyrene) from Reference [19]. Frame A illustrates the coarse dispersion resulting from using an extruder to blend the constituent polymers; frame B illustrates the finer dispersion achieved with the S3 P process.
produces excellent quality alloys, independent of whether or not the viscosities of these components were matched (Figure 6.10). The resulting alloy is a dispersion of polystyrene spherical domains of size distributed in the range of 2–5 m. A small decrease in molecular weight of the polystyrene was detected (Figure 6.11); similarly, a rise in melt flow rate (MFR) was seen in the majority phase as well (Table 6.1). Blends involving elastomers can be made analogously as with the systems cited above. As seen in Chapter 8, which deals with S3 P of tire stock, proper TABLE 6.1.
Melt Flow Rate Before and After S3 P Processing for Virgin PS and Two PP Samples. Material
Low molecular weight PP pre-S3 P post-S3 P High molecular weight PP pre-S3 P post-S3 P PS pre-S3 P post-S3 P
©2001 CRC Press LLC
Melt Flow Rate (g/10 min) 19.0 29.0 ± 1.4 0.53 ± 0.03 1.02 ± 0.08 2.03 ± 0.05 14.2 ± 0.4
P1: FUY pb035-06
March 15, 2001
11:23
Char Count= 0
FIGURE 6.11 Gel permeation chromatograms (GPC) of polystyrene subjected to S3 processing. It is clear that a molecular weight reduction occurred only for the high molecular weight polystyrene.
©2001 CRC Press LLC
P1: FUY pb035-06
March 15, 2001
11:23
Char Count= 0
selection of equipment design and operating conditions permits low modulus components (elastomeric materials) to successfully undergo pulverization and size reduction. Self-compatibilization of the resulting plastic/elastomer alloys results because of the trace amount of hetero-block copolymers formed at their interfaces. THE MECHANISMS THAT UNDERLIE THE OPERATION OF S3 P The key to understanding why S3 P is fundamentally different in the way it affects the in-process material lies in the fact that it causes individual solid pieces to fracture, while simultaneously causing “cold welding” of others. Mechanistic details of these processes are covered in Chapter 3 of this book. The mechanochemistry that occurs is the heart of what happens during S3 P processing. Chain scission occurs during these particle cleavage events that are associated with the size-reduction process. There then follows a cascade of subsequent chemical events, including the formation of lower molecular weight species, hetero-block and graft species, and stereoregular chains with racimized segments. Macromolecules involving chains from two different polymers are formed as a result of “cold welding” between two particles of dissimilar composition. Such “cold welding” happens when high-shear stresses in the presence of high hydrostatic stress fields urge chains together with sufficient energy to cause exchange of chemical bonds to occur. The hetero-block and graft species are the self-compatibilizer species that lead to value-added alloys made from immiscible polymer pairs. Their presence leads to a lowering of interfacial energies between phases in a two-component dispersion, and this in turn means that the cohesion across such interfaces rises to values approaching that of the bulk. Consequently, these alloys exhibit mechanical behavior (strength and ductility) that is not attenuated as the concentration of a dissimilar polymer rises. Similarly, these new chains also cause the resulting dispersions of the different phases to involve unusually small particles and have an uncommonly narrow particle size distribution. For example, the work of Ganglani et al. on alloys of HDPE, LDPE, and LLDPE formed using the S3 P process employed such techniques as GPC, ESR, NMR, IR, DSC, and melt rheology to show the extent to which broken chains are essential in these processes [20]. Indeed, only a barely detectable quantity of unpaired electrons can be measured, presumably due to the functioning of species of the “stabilizer package” ingredients. As a result, bulk properties of the individual constituents, including mechanical behavior, crystallinity, and melting points, remained unchanged. The difference is seen, most remarkably, in the bulk properties of their alloys, due to self-compatibilization. Data in this study establish that trace amounts of block and graft chains did form between the constituent polyolefins, and one must presume that they are responsible for the beneficial effects seen in the final blends.
©2001 CRC Press LLC
P1: FUY pb035-06
March 15, 2001
11:23
Char Count= 0
S3 P of miscible polymer pairs having very different viscosities works because the phases in the in-process material are entirely in the solid state. Solid phases will support stresses imposed upon them to the point where cracking, normal stress yielding, or shear yielding begin. Any liquid phase present would substantially allow stresses applied by the twin-screw pulverizer to relax and thereby be insufficient to accomplish the desired dispersion [21]. REFERENCES 1. Khait, K. and Carr, S. H. “Solid-State Shear Extrusion Pulverization: A Novel Process for Recycling of Unsorted Post-Consumer Waste,” Proceedings of the Fifth World Congress of Chemical Engineering, New York: American Institute of Chemical Engineers, pp. 76–81 (1996). 2. Khait, K. and Carr, S. H. “Value-Added Materials Made from Recycled Plastics,” Proceedings of the Annual Technical Conference of the Society of Plastics Engineers, ANTEC’ 97, Toronto, pp. 3086–3090 (1997). 3. Furgiuele, N., Khait, K., and Torkelson, J. M. “Mixing Polymer Blends of Unmatched Viscosities Via S3 P,” Polymeric Materials Science Engineering, 81:125–126 (1999). 4. Paul, D. R. and Bucknall, C. B. Polymer Blends—Vol. 1: Formulation, Wiley-Interscience, New York: John Wiley & Sons, Inc. (2000). 5. Barlow, J. W. and Paul, D. R. “Mechanical Compatibilization of Immiscible Blends,” Polymer Engineering and Science, 24:525–534 (1984). 6. Machado, J. M. and Lee, C. S. “Compatibilizing Immiscible Blends with a Mutually Miscible Homopolymer,” Plastics Engineering, October 1993, pp. 33–36. 7. Koning, C., van Duin, M., Pagnoulle, C., and Jerome, R. Progress in Polymer Science, London: Elsevier Science, 23:707–757 (1998). 8. Ajji, A. and Utracki, L. A. “Compatibilization of Polymer Blends,” Progress in Rubber and Plastics Technology, Shrewsbury, UK: Plastics and Rubber Institute, pp.153–188 (1985). 9. George, S., Joseph, R., Thomas, S., and Varughese, K. T. “Blends of Isotactic Polypropylene and Nifrile Rubber: Morphology, Mechanical Properties and Compatibilization,” Polymer, 36:4405–4414 (1995). 10. Lustiger, A., Marzinsky, C. N., and Mueller, R. R. “Spherulite Boundary Strengthening,” Proceedings of the Annual Technical Conference of the Society of Plastics Engineers, ANTEC ’98, Atlanta, pp. 1506–1510 (1998). 11. Xanthos, M. “Interfacial Agents for Multiphase Polymer Systems,” Polymer Engineering and Science, 28:1392–1399 (1988). 12. Carr, S. H. and Khait, K. K. “Toughened Recycled Polypropylene: Blends Produced Via the Solid-State Shear Pulverization Process,” Proceedings of the Society of Plastics Engineers, ANTEC ’98, April 1998, Atlanta, pp. 2939–2941 (1998). 13. Triacca, V. J., Ziaee, S., Barlow, J. W., Keskkula, H., and Paul, D. R. “Reactive Compatibilization of Blends of Nylon 6 and ABS Materials,” Polymer, 32:1401–1413 (1991). 14. DiLorenzo, M. L. and Frigione, M. Journal of Polymer Engineering, 17:429–459 (1997). 15. Furgiuele, N., Lebovitz, A. H., Khait, K., and Torkelson, J. M. “Novel Strategy for Polymer Blend Compatibilization: Solid-State Shear Pulverization,” Macromolecules, 23:225 (2000). 16. Smith, A. P., Spontak, R. J., Ade, H., Smith, S. D., and Koch, C. C. “High-Energy Cryogenic Blending and Compatibilizing of Immiscible Polymers,” Advanced Materials, 11:1277–1281 (1999).
©2001 CRC Press LLC
P1: FUY pb035-06
March 15, 2001
11:23
Char Count= 0
17. Nesarikar, A. R., Khait, K., Mirabella, F., and Carr, S. H. “Self-Compatibilization of Polymer Blends Via Novel Solid-State Shear Extrusion (SSSE) Pulverization,” Journal of Applied Polymer Science, 63:1179–1187 (1997). 18. Khait, K. and Carr, S. H. “Mixed Polyolefin Powders Recycled Via the Solid-State Shear Pulverization Process,” Proceedings of the Society of Plastics Engineers, ANTEC ’98, Atlanta, pp. 2533–2535 (1998). 19. Davydov, A., Khait, K., and Torkelson, J. M. “Effect of Solid State Shear Pulverization on Mechanical Properties and Morphology of Immiscible Polystyrene/Low Density Polyethylene Polymer Blends,” unpublished study. 20. Ganglani, M., Torkelson, J. M., Carr, S. H., and Khait, K. “Trace Level of Mechanochemical Effects in Pulverized Polyolefins,” Journal of Applied Polymer Science (submitted 2000). 21. Furgiuele, N., Khait, K., and Torkelson, J. M. “Novel Approach for the Compatibilization of Polymer Blends and Polymeric Waste,” Proceedings of the Meeting of the American Institute of Chemical Engineers, Fall, 1998.
©2001 CRC Press LLC
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
CHAPTER 7
Value-Added Products Made from Recycled Plastics via S3 P
STATE OF PLASTICS RECYCLING
I
N discussing plastics recycling, it is important to realize that different plastics
are immiscible and, therefore, must be sorted by type and color prior to recycling. Discarded post-consumer plastics waste is the most difficult to recycle because of its variability, contaminants, and presence of different types of single plastics and their blends. One possible re-use of mixed plastics is making lowvalue products, such as flower pots, traffic cones, posts, and fences. The most common approach, however, involves compatibilizing dissimilar plastics by adding pre-made expensive compatibilizing agents with specifically designed functionalities. The new S3 P technology provides a novel means of recycling commingled plastics without sorting and without the use of compatibilizers. The physical properties of S3 P-made plastics allow them to be used as value-added materials in a variety of applications. Despite the increasing quantity of plastics in municipal solid waste (MSW) streams, plastics recycling has been slow to develop. In 1998, plastics contributed about 20 percent of the total volume of materials landfilled. Decreasing landfill space and rapidly rising disposal costs have forced many municipalities to begin curbside recycling of post-consumer waste. The plastic products found in MSW are primarily made by injection-molding, extrusion, blow-molding, and rotational-molding techniques. Each of these processes requires plastics with different melt flow rates (MFRs). For example, high MFR plastics are more suitable for injection-molding, while low MFR materials are used for rotational molding, and plastics with fractional flow MFRs are needed for blow molding. Even for the same type of plastic, such as highdensity poly(ethylene) (HDPE), which is used to make milk jugs, water bottles,
©2001 CRC Press LLC
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
and the base cups for some beverage bottles, the difference in their MFRs would cause melt separation during processing and thus sorting is necessary prior to recycling. It is well known that weak adhesion between phases of ordinarily immiscible polymers results in blends with poor mechanical properties. For example, mixed plastics waste containing PET and HDPE is difficult to process because their incompatibility results in severe die swelling and low melt strength during extrusion [1]. Therefore, compatibilizers, coupling agents, impact modifiers, and reinforcing agents are frequently used to improve mechanical properties of post-consumer plastics for re-use [2]. Most existing processes for recycling plastics waste result in products that are less valuable than the original products. Generally, the physical properties of recycled commingled (unsorted) plastics are inferior to the virgin resins because of chemical incompatibility and the variety of colors in collected waste. Fibers and other additives make recycling of mixed plastics even more difficult and expensive. The term “commingled” plastics is used to define post-consumer (discarded) plastics, which include not only a mixture of several plastics, but also a mixture of different plastic products that have been made using a variety of processing techniques. It is also important to note that post-consumer plastics include products made from the same plastic family but manufactured by different technologies using proprietary additives. These factors make recycling of plastics from unknown sources extremely complicated. The MSW stream consists primarily of discarded packaging, such as containers, bottles, lids, caps, and a variety of plastic bags and wraps. This stream usually comprises polyolefins, such as low-density poly(ethylene) (LDPE) and linear low-density poly(ethylene) (LLDPE), HDPE (translucent and colored), PET (both clear and green), PVC (clear and colored), PP (opaque and colored), and PS (clear and colored). Stumbling blocks to post-consumer recycling come at the start of the whole process—collecting and sorting—rather than in the reprocessing of separated waste streams. Although several automated sorting techniques exist, most sorting of plastic bottles and containers is still done manually and is, therefore, inefficient. S3 P AND PLASTICS RECYCLING Khait et al. developed a new environmentally friendly recycling process for pre- or post-consumer multicolored mixed plastic waste (without sorting) based on the Solid-State Shear Pulverization (S3 P) process [3]. As indicated in Chapter 3, this novel approach to recycling is based on a size-reduction method that was originally conceptualized by the group of Soviet scientists under the
©2001 CRC Press LLC
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
general direction of Enikolopyan at the Institute of Chemical Physics of the Academy of Sciences in Moscow. This process grew out of original work by Bridgman, who studied the deformation of solids under simultaneous action of pressure and shear [4]. It was postulated that powder is formed after a solid polymer accumulated to be released at some threshold value. This process was initially termed Elastic Deformation Grinding (EDG) and subsequently was referred to as high-temperature shear-induced grinding [5]. Since the late 1980s, Khait’s group at Northwestern University’s Polymer Technology Center (PTC) (Evanston, Illinois, US) has conducted research on the S3 P process in close collaboration with Berstorff, a division of KraussMaffei Corporation in Hannover, Germany, and with their US branch in Florence, Kentucky. Berstorff (formerly known as Berstorff Maschinenbau) is a world leader in manufacturing extruders for plastics and rubber processing. Traditionally, co-rotating intermeshing twin-screw extruders have been known as compounding machines, and they perform a wide variety of functions by melting resins with additives and fillers and discharging different shapes such as pellets, sheets, or profiles. In the late 1980s, however, Berstorff added a novel capability to its co-rotating twin-screw extruder that allowed it to convert polymeric materials in a solid state into powders. Mack described “solid-state extrusion” as a previously unthought-of process by which a fused polymer mass was converted into powder with the efficient removal of frictional heat [6]. The high shear forces in the extruder’s solid bed crushed the polymers until the particle size diminished and powder exited the extruder via an open discharge. This process resulted in a joint patent (US) between Hermann Berstorff Maschinenbau GmbH and the Soviet Academy of Sciences in 1986 [7]. The researchers at Northwestern University’s Polymer Technology Center developed a Solid-State Shear Pulverization process with an emphasis on a nomelting approach (unlike prior efforts by the Soviet Academy of Sciences and Berstorff by which polymers were first melted and then formed into powder). A schematic diagram of the co-rotating twin-screw pulverizer with integrated heating and cooling is shown in Figure 7.1. Plastics in chopped or shredded form are fed as mixtures or as individual components and are subjected to compression and shearing forces of differing intensity. This continuous, once-through process converts post- or pre-consumer plastics or virgin pellets into a uniform powder. The powders are formed in the solid state by shear deformation under pressure coupled with a rapid cooling. The degree of shear is controlled by varying the temperature and the process parameters including screw speed, feed rate, and screw design. When the process conditions are optimal, spontaneous fragmentation occurs, producing particles that are smaller than any of the clearances in the pulverizer. The particle-formation mechanism is different from that of conventional grinding and is still under investigation. The size of the products range from coarse to fine powders (below 500 microns). In the S3 P process, plastics are processed at temperatures below their melting points (in the case of
©2001 CRC Press LLC
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
FIGURE 7.1
A schematic diagram of the twin-screw pulverizer.
crystalline or semicrystalline polymers) or at glass-transition temperatures (in the case of amorphous polymers). The post-consumer plastics include a variety of colors, but the products from this process have homogeneous, unusually light colors that are not the conventional dark shades of gray or brown seen in recycled commingled plastics products such as plastic lumber [8,9] (see Photos on the color insert). The homogeneous pastel colors of the pulverized material demonstrate an additional advantage of this process because it is necessary to add only a small amount of expensive pigments or dyes for color correction. Unlike any other process involving resin production in a pellet form, this novel pulverization process generates materials in a powdered form, which is required for several plastics processing operations using powder as a feedstock, such as rotational molding, powder coating, pipe extrusion, compounding, and sintering. The particle size of the produced powder for a given polymer or a mixture of polymers is controlled by changes in the screw design and by the adjustment of process parameters [10]. The formation of powder involves the rupture of chemical bonds; because of the fine powder size, the number of broken bonds is high, and, consequently, the powders are somewhat reactive. Khait and Petrich investigated the reactivity of the powders with electron spin resonance (ESR) [8]. Initial ESR studies of reactive sites of pulverized LDPE suggested that ruptured bonds are formed during pulverization. Figures 7.2a,b show ESR spectra for samples of LDPE before and after pulverization. Prior to pulverizing, the sample was acquired so that its spectrum was enlarged 3.2 times relative to the spectrum of the pulverized sample. The data shown in Figures 7.3a,b indicate that the reactive sites are generated by the mechanical rupture of main-chain carbon bonds during pulverization. The spectrum in Figure 7.3a represents LDPE powder immediately after pulverization, and the spectrum in Figure 7.3b
©2001 CRC Press LLC
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
FIGURE 7.2 (a) ESR spectra of post-consumer LDPE before pulverization. (b) ESR spectra of post-consumer LDPE after pulverization. Reprinted with permission from Reference [3]. c 1994 SPE.
FIGURE 7.3 (a) ESR spectra of pulverized samples, post-consumer LDPE before exposure to air for three days. (b) ESR spectra of pulverized samples, post-consumer LDPE after exposure to c 1994 SPE. air for three days. Reprinted with permission from Reference [3].
©2001 CRC Press LLC
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
(enlarged 25 times) represents the same sample after three days of air exposure at room temperature. The ESR signal, although still observable, is nearly zero after this ambient exposure, indicating that the maximum lifetime expected for these reactive sites is one or two days. None of these spectra represents the total number of broken bonds in the samples and should be normalized by the sample mass to obtain truly quantitative information [8]. It is possible that certain additive molecules present in post-consumer LDPE act as radical scavengers and stabilize the radicals as they are formed. Many commercial dyes contain amines, aromatic, and highly-conjugated groups that are typical of radical scavenging molecules. The drastic increase, however, that is observed in peak intensity after pulverization has shown that free radicals are generated by the rupture of C-C bonds by application of mechanical forces. The possibility of free radicals terminating by heterogeneous combination to form compatibilizing block or graft copolymers in co-processed polymer mixtures establishes the potential of S3 P for commingled plastics recycling. The recovery of post-consumer mixed color polyolefins such as HDPE, LLDPE, and PP of unknown origin was addressed [3]. The single feedstocks and dry blends at various ratios have been processed with a modified Berstorff PT-40A pilot scale co-rotating, intermeshing twin-screw extruder [11]. HDPE/LLDPE mixtures at 40/60 and 60/40 ratios as well as HDPE/PP mixtures at 70/30, 80/20, and 90/10 ratios have been studied because they reflect the proportions of these polyolefins commonly found in the post-consumer waste stream. In spite of the fact that post-consumer plastics used in this study included a multitude of colors, all powders yielded homogeneous, unusually light colors. For example, the PP sample consisted of a mixture of red, yellow, tan, white, and some black flakes that produced a cranberry-colored powder. The LLDPE sample consisted of a mixture of orange, beige, blue, and white flakes that produced a peach-colored powder. The HDPE sample consisted of a mixture of green, blue, yellow, purple, beige, and white flakes that produced an aqua-colored powder. These color shades are possible only with the S3 P process, which avoids melting of the polymeric materials. Although the particle-formation mechanism is not yet fully understood, it is believed that, during the pulverization process, flakes are transformed into powder in certain areas of the pulverizer where shearing (kneading) blocks are located. Depending upon the location of the shearing blocks and processing conditions (RPM, feedrate, etc.) of the pulverization process, it is possible to produce powder of a desired particle size, which can range from coarse material of 10 mesh (2,000 microns) to fine powder of 80 mesh (178 microns) and finer. Both particle size and particle shape are industrially important and can be determined by microscopy and image analysis. Dietering reported shape and surface texture of particles studied with a Hitachi S-510 microscope combined with digital image analysis [12]. It has been noticed that during pulverization aimed at fine powder production, some effects occur that are not present during
©2001 CRC Press LLC
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
the production of coarse powder. These effects include some aggregation and agglomeration. Elongated particles have proportionally greater surface area than the cubiform particles that pass through the same square mesh sieve. A qualitative microscopy yielded a better understanding of the particle-formation mechanism during the S3 P process [12]. Micrographs have been used to observe the behavior of individual particles from fine powders to thin, elongated fluff of varying coarseness. Figures 7.4a,b, 7.5a,b, and 7.6a,b show scanning electron microscopy (SEM) images for PP, HDPE, and LDPE, respectively. In each of these figures, typical particles of each material are shown at various magnifications. The SEM image of PP at low magnification (40×) in Figure 7.4a shows highly elongated particles with smooth surfaces. Other particles (Figure 7.4b) are rolled and twisted by the shearing process (magnification 200×). Figure 7.5a shows HDPE particles at low magnification (40×), which look like smooth, elongated species with numerous outstretched tendrils. Some particles appear to consist of layers (Figure 7.5b at magnification 150×). Figure 7.6a shows several LDPE particles at a magnification of 80×. These particles are significantly smaller than those of HDPE and PP. The particles in Figure 7.6a illustrate rough and laminated surfaces. At the surface edges, some localized deformation was observed, which is typical of shear crazing. At higher magnification (200×), a very thin strand of LDPE is shown (Figure 7.6b).
PHYSICAL PROPERTIES OF COMMINGLED PLASTICS The physical properties and melt flow rates of S3 P-made binary and tertiary mixes of post-consumer plastics (Table 7.1) were initially reported by Khait and Petrich [8]. These properties were measured in accordance with the ASTM tests of tensile strength and elongation (D638), Notched Izod Impact strength (D256), hardness shore D (D2240), heat distortion temperature (D648), and melt flow rate (MFR) (D1238). As can be seen from these data, both HDPE/LLDPE blends had reasonable tensile strength and Notched Izod Impact. Furthermore, three-component mixtures of ordinarily immiscible polyolefins, such as HDPE, LLDPE, and PP made via the S3 P process, have shown acceptable physical properties for re-use by injection molding or extrusion. It has also been demonstrated that, analogous to mixtures of HDPE with either LLDPE or PP, light-colored homogeneous powders can be obtained from individual polyolefin recycled feedstocks [11]. Injection-molded test specimens had smooth and shiny surfaces and did not delaminate upon breaking. Key physical properties of the above-mentioned post-consumer polyolefins are shown in Table 7.2. Pulverized PP had the highest values of Notched Izod Impact strength, tensile strength, and elongation at break as compared to those of HDPE or LLDPE.
©2001 CRC Press LLC
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
(a)
(b) FIGURE 7.4 (a) SEM image of pulverized post-consumer PP at 40×. (b) SEM image of pulverc 1995 SPE). ized post-consumer PP at 200× (Reprinted with permission from Reference [11].
©2001 CRC Press LLC
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
Other blends containing recycled HDPE and PET at 40/60, 50/50, and 60/40 ratios have been studied [13]. (Non-S3 P processes require HDPE and PET to be separated from one another in order to be recovered.) In addition, threecomponent mixtures containing dissimilar recycled plastics, such as HDPE, PET, and LLDPE at various ratios, have also been successfully made via the S3 P process (Table 7.3). Binary HDPE/PET blend containing 60-weight-percent PET had the best tensile strength and Notched Izod Impact of the three ratios tested. It should be noted that all HDPE/PET blends exhibited yield during tensile tests. Tertiary blends of HDPE, LLDPE, and PET had similar properties, except for the 30/30/40 blend, which exhibited much higher HDT, as anticipated for the blend with the highest percentage of PET (40 percent by weight vs. 30 percent in two other blends). More dramatic results have been obtained with commingled (unsorted) multicolored post-consumer waste containing mixtures of HDPE, LDPE, PP with PET, PS, and even PVC [14,15]. It is well established by the recycling industry
(a)
FIGURE 7.5 (a) SEM image of pulverized post-consumer HDPE at 40× (Reprinted with permisc 1995 SPE). (b) SEM image of pulverized post-consumer HDPE at sion from Reference [11]. 150×.
©2001 CRC Press LLC
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
(b) FIGURE 7.5
(Continued )
that cross contamination of PET and PVC makes their recovery impossible because PVC would degrade prior to melting of PET. Using solid-state (no melting) pulverization technology, it was possible to make materials from combinations of six unsorted plastics, including PET and PVC, commonly found in the discarded waste stream. Injection-molded test specimens had an unusually uniform pastel color and glossy surface appearance similar to samples molded from S3 P-made powders derived from single post-consumer plastics. The physical properties of mixtures of unsorted, multicolored commingled plastics made by the solid-state pulverization process are given in Table 7.4. Although the Notched Izod Impact strength of S3 P-made materials from multicomponent post-consumer waste stream was low (0.2–0.3 ft-lb/in), the ultimate tensile strength was similar to that of recycled HDPE and varied between 2,210 and 2,610 psi, depending on composition; mixture of polyolefins with PET even exhibited yield strength of 2,710 psi. A sieve analysis (made with the Gilson ultrasonic autosiever) of a 100-percent recycled post-consumer four-component mixture of HDPE/LDPE/PP/PVC at
©2001 CRC Press LLC
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
(a)
(b) FIGURE 7.6 (a) SEM image of pulverized post-consumer LDPE at 80× (Reprinted with c 1995 SPE). (b) SEM image of pulverized post-consumer permission from Reference [11]. LDPE at 150×.
©2001 CRC Press LLC
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
TABLE 7.1.
T1: SYV
Char Count= 0
Key Physical Properties of Injection-Molded Post-Consumer Polyolefin Blends Pulverized by the S3 P Process.
Tensile Properties Ultimate Feedstock/Ratio MPa (psi)
a
Elong %
Notched Izod Heat Impact Distortion Melt Flow J/m Hardness Temperature Rate ◦ (ft-lb/in) Shore D C at 66 psi g/10 mina
HDPE/LLDPE 40/60
17.6 (2,550)
15
32 (0.6)
59
45
37.4
HDPE/LLDPE 60/40
19.7 (2,860)
13
32 (0.6)
63
57
20.3
HDPE/LLDPE/PP 60/30/10
20.5 (2,970)
9
21 (0.4)
64
58
59.7
At 190◦ C, 2.16 kg load.
55/30/10/5 ratio showed that just less than 50 percent of the powder has a size of 300 and 180 m (Table 7.5). This blend was pulverized at high shear conditions, which was accomplished by using a 25-mm screw in the 25-mm barrel. Scanning electron microscopy of S3 P-made powders has revealed interesting characteristics of the particle shape. While flakes or chips of as-received recycled feedstock have sharp, angular surfaces resulting from the conventional grinding process, all pulverized materials have smooth surfaces, indicative of a solid-shearing process. Fine powder particles have a unique, elongated shape (Figures 7.7a,b), whereas larger fluff consists of fibrous, easily peeled particles. Tables 7.6, 7.7, and 7.8 show the dependence of the particle size of the S3 P powder on the composition of the feedstock. A sieve analysis of a 100-percent recycled post-consumer five-component mixture of HDPE/LLDPE/PP/PET/ TABLE 7.2.
Key Physical Properties of Injection-Molded Post-Consumer HDPE, LLDPE, and PP Pulverized by the S3 P Process. Tensile Properties
Feedstock
Yield Ultimate MPa (psi) MPa (psi)
HDPE LLDPE PP
a
19.8 (2,870)
Elong % 7.5
Notched Heat Izod Distortion Melt Flow Impact J/m Hardness Temperature Rate (ft-lb/in) Shore D ◦ C at 66 psi g/10 mina 26 (0.5)
63
60
49.6
14.2 (2,060)
60
37 (0.7)
57
50
57.3
36.2 4,730
100
42 (0.8)
72
93
9.5
At 190◦ C, 2.16 kg load.
©2001 CRC Press LLC
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
Physical Properties of Injection-Molded Post-Consumer HDPE/PET and HDPE/LLDPE/PET Blends Pulverized Via the S3 P Process.
TABLE 7.3.
Feedstock/Ratio
Notched Izod Heat Tensile Properties Impact Distortion Yield Ultimate Elong J/m Hardness Temperature MPa (psi) MPa (psi) % (ft-lb/in) Shore D ◦ C at 66 psi
HDPE/PET 40/60
28.3 (4,100)
5
37 (0.7)
70
68
HDPE/PET 50/50
24.8 (3,600)
4
26 (0.5)
69
70
HDPE/PET 60/40
19.2 (2,790)
3
10 (0.2)
68
70
HDPE/LLDPE/PET 60/10/30
20.8 (3,020)
5
10 (0.2)
68
67
HDPE/LLDPE/PET 40/30/30
16.9 (2,450)
5
10 (0.2)
65
63
HDPE/LLDPE/PET 30/30/40
17.5 (2,540)
5
16 (0.3)
65
144
PVC at 40/30/5/20/5 ratio made at high shear conditions showed that 35 percent of the powder has a size of 125-m or finer (Table 7.6). The shape of the powder was similar to that of the above-mentioned fourcomponent blend of HDPE/LLDPE/PP/PVC. A sieve analysis of a 100-percent recycled post-consumer, five-component mixture of HDPE/LLDPE/PP/PS/ PVC at 50/30/10/5/5 ratio showed that the majority of the powder has a size of 300 m followed by a size of 500 m and 180 m (Table 7.7). TABLE 7.4.
Physical Properties of Injection-Molded Unsorted Post-Consumer Plastics Mixtures after S3 P Processing. Notched Heat Izod Distortion Impact J/m Hardness Temperature Yield Ultimate Elong (ft-lb/in) Shore D ◦ C at 66 psi MPa(psi) MPa(psi) % Tensile Properties
Feedstock/Ratio HDPE/LDPE/PP/PET 40/30/10/20
18.7 (2,710)
10
16 (0.3)
65
59
HDPE/LDPE/PP/PET/PS 40/30/5/20/5
18.0 (2,610)
8
11 (0.2)
64
64
HDPE/LDPE/PP/PVC 55/30/10/5
16.5 (2,390)
6
11 (0.2)
63
65
HDPE/LDPE/PP/PET/PVC 40/30/5/20/5
16.1 (2,290)
9
11 (0.2)
65
65
HDPE/LDPE/PP/PET/PS/PVC 40/30/5/15/5/5
15.2 (2,210)
11 (0.2)
65
61
©2001 CRC Press LLC
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
Sieve Analysis of 100-Percent Post-Consumer Recycled Mixture of HDPE/LLDPE/PP/PVC at 55/30/10/5 Ratio (High-Shear S3 P at 0◦ C).
TABLE 7.5.
Size Mesh
Microns
35 50 80 120 200 270 450 Fines
500 300 180 125 75 53 35
Weight Percent Retained 41.5 26.9 18.2 7 4.7 1.3 0.4 0
The shape of the powder particles was typical of fractured surface resulting from shear deformation (Figures 7.8a,b). A sieve analysis of a 100-percent recycled post-consumer six-component mixture of HDPE/LLDPE/PP/PET/PS/PVC at 40/30/15/5/5/5 ratio showed that the majority of the powder has a size of 300 m and finer (Table 7.8). The fine powder particles had unique, elongated shapes suggestive of a solidshearing process (Figures 7.9a,b), while larger particles had a fibrous structure.
(a)
(b)
FIGURE 7.7 (a) SEM image of pulverized post-consumer mixture of HDPE/LLDPE/PP/PVC at 55/30/10/5 ratio at 40×. (b) SEM image of pulverized post-consumer mixture of HDPE/LLDPE/ PP/PVC at 55/30/10/5 ratio at 250×.
©2001 CRC Press LLC
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
Sieve Analysis of 100-Percent Post-Consumer Recycled Mixture of HDPE/LLDPE/PP/PET/PVC at 40/30/5/20/5 Ratio (High-Shear S3 P at 0◦ C).
TABLE 7.6.
Size Mesh
Microns
35 50 80 120 200 270 450 Fines
500 300 180 125 75 53 35
Weight Percent Retained 15.6 22.7 27.3 11.4 12.7 5.8 3.8 0.7
Sieve Analysis of 100-Percent Post-Consumer Recycled Mixture of HDPE/LLDPE/PP/PS/PVC at 50/30/10/5/5 Ratio (High-Shear S3 P at 0◦ C).
TABLE 7.7.
Size Mesh
Microns
35 50 80 120 200 270 450 Fines
500 300 180 125 75 53 35
Weight Percent Retained 28.3 36.4 21 6.1 4.7 1.9 1.5 0.1
Sieve Analysis of 100-Percent Post-Consumer Recycled Mixture of HDPE/LLDPE/PP/PET/PS/PVC at 40/30/15/5/5/5 Ratio (High Shear S3 P at 0◦ C).
TABLE 7.8.
Size Mesh
Microns
35 50 80 120 140 200 270 Fines
500 300 180 125 105 75 53
©2001 CRC Press LLC
Weight Percent Retained 18.3 44.6 19.8 8.7 4.5 2.8 0.4 0.9
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
(a)
(b)
FIGURE 7.8 (a) SEM image of pulverized post-consumer mixture of HDPE/LLDPE/PP/PS/PVC at 50/30/10/15/5 ratio at 40×. (b) SEM image of pulverized post-consumer mixture of HDPE/ LLDPE/PP/PS/PVC at 50/30/10/15/5 ratio at 250×.
(a)
(b)
FIGURE 7.9 (a) SEM image of pulverized post-consumer mixture of HDPE/LLDPE/ PP/PET/PS/PVC at 40/30/5/15/5/5 ratio at 40×. (b) SEM image of pulverized post-consumer mixture of HDPE/LLDPE/PP/PET/PS/PVC at 40/30/5/15/5/5 ratio at 250×.
©2001 CRC Press LLC
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
Because it had already been proven that the use of the Berstorff pulverization equipment during the non-melting S3 P process does not require separation of unlike polymers, it was possible to recycle commingled plastic waste contaminated with printed paper or plastic labels, glue, and pull-rings from aluminum cans [3,8,15] [for example, see Photos 1(a–c) on the color insert]. Solid-state powder made from this unsorted plastic waste had a homogeneous, light color and was directly processed into small industrial parts by injection molding (without prior pelletization); this process avoids additional heat history and property degradation. The researchers at Northwestern University’s PTC have also discovered that the solid-state pulverization process allows for both intimate mixing of like polymers and for mixtures of unlike polymers with unmatched viscosity [13]. It is known that even if two resins of the same family have different melt flow rates, they cannot be co-processed due to phase separation in the melt. The MFRs of multicolored flakes derived from a commingled post-consumer plastic waste stream of unknown origin were studied. It was determined that the flakes exhibited a wide range of values. Three post-consumer multicolored polyolefins—HDPE, LLDPE, and PP—have been tested, as they represent the largest volume of discarded products in the MSW. It is believed that the HDPE flakes consisted of a mixture of granulated milk and water bottles made from HDPE homopolymer and flakes from household chemical bottles made from HDPE copolymer. Both types of HDPE had a fractional melt flow (below 1 g/10 min). It was also assumed that the PP flakes consisted of a mixture of homopolymer PP used to make closures and caps, as well as of copolymer PP used to make battery cases, lids, housewares, toolboxes, and other consumer goods. Each of these multicolored feedstocks was intentionally separated by color to determine melt flow rates for these single-colored flakes (Tables 7.9–7.11). For comparison, melt flow rates of S3 P-made powders derived from these flakes are given in Table 7.12. TABLE 7.9.
Melt Flow Rate of Single-Colored Flakes of HDPE Feedstock Prior to Pulverization. Flake Color
Melt Flow g/10 min
Blue Cream Green Opaque Purple Translucent Off-white #1 Off-white #2 White
29.9 26.5 16.1 12.0 27.8 22.6 18.5 56.3 15.9
©2001 CRC Press LLC
Conditions kg Load
◦ C,
150, 2.16 150, 2.16 150, 2.16 190, 2.16 150, 2.16 150, 2.16 150, 2.16 190, 2.16 190, 2.16
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
TABLE 7.10.
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
Melt Flow Rate of Single-Colored Flakes of LLDPE Feedstock Prior to Pulverization. Flake Color
Melt Flow g/10 min
Orange Blue Opaque white Translucent white Cream
56.1 60.9 55.0 55.9 54.9
Conditions kg Load
◦ C,
150, 2.16 150, 2.16 150, 2.16 150, 2.16 150, 2.16
As can be seen from Table 7.9, the melt flow rates of opaque, off-white #2, and white HDPE flakes were measured at 190◦ C, while other colored flakes were measured at 150◦ C. It is believed that these flakes represent a copolymer presence in post-consumer HDPE feedstock. The melt flow rate of individual colored flakes varied widely from 16 to 29 g/10 min (at 150◦ C) and from 12 to 56 g/10 min (at 190◦ C). The melt flow rate of PP flakes (Table 7.11) also varied from 18 g/10 min for black color to 27 g/10 min for clear flakes. The melt flow rate of LLDPE feedstock (Table 7.10) of 55 to 61 g/10 min did not vary as much as that of either HDPE or PP. In spite of variations in melt flow values for individual colored flakes of HDPE feedstock, the resultant powder had a melt flow rate of 49.6 g/10 min (Table 7.12). A similar trend has been observed for S3 P-made PP powder, which had a melt flow rate of 26.4 g/10 min. It is believed that this phenomenon can be explained by intimate mixing that occurred during the pulverization process. It has been reported that powders derived from multicolored post-consumer polyolefin feedstock have been directly injection-molded (without prior pelletization) into small industrial parts of complex geometry with good surface appearance [13]. Many more mixtures of dissimilar plastics from the MSW have been recycled via the S3 P process. This was not possible by standard melt extrusion due to
TABLE 7.11.
Melt Flow Rate of Single-Colored Flakes of PP Feedstock Prior to Pulverization. Flake Color
Melt Flow g/10 min
Black Blue Clear Opaque Red White Yellow
18.8 22.3 27.5 25.9 20.6 25.0 24.3
©2001 CRC Press LLC
Conditions kg Load
◦ C,
230, 2.16 230, 2.16 230, 2.16 230, 2.16 230, 2.16 230, 2.16 230, 2.16
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
TABLE 7.12.
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
Melt Flow Rate of Multicolored, Unsorted Post-Consumer HDPE, LLDPE, and PP after Pulverization. Feedstock
Melt Flow g/10 min
HDPE LLDPE PP
49.6 57.3 26.4
Conditions kg Load
◦ C,
190, 2.16 190, 2.16 230, 2.16
their unmatched viscosity and their phase separation. The best-known example involves a waste stream containing HDPE bottles with PP caps [see Photos 2(b–c) on the color insert]. If the waste is contaminated during HDPE sorting by more than 10 weight percent of PP, the mixture cannot be melt-processed into useful products [16]. It has been learned, however, that despite a known incompatibility between HDPE and PP when PP content exceeds seven weight percent when processed by conventional melt mixing, HDPE with 30 weight percent PP has been successfully converted into powder via the S3 P process [17]. A comparison of the properties of S3 P-made powder of HDPE and PP is presented in Table 7.13. The tensile strength of the 80/20 HDPE/PP blend approached that of HDPE. In spite of the unmatched viscosity of HDPE and PP as measured by MFR for HDPE (52 g/10 min) and PP (9.5 g/10 min), the resultant 80/20 blend was homogeneous and exhibited MFR of 42 g/10 min. Physical properties of the 70/30 HDPE/PP blend were similar to those of the 80/20 blend with the exception of the Notched Izod Impact, which was lower (0.2 vs. 0.4 ft-lb/in respectively). Using S3 P technology, it was possible to incorporate even 30 weight percent of PP into HDPE with only partial reduction of impact strength.
CHARACTERIZING POWDERS In many powder-processing operations, particle size and particle size distribution play a key role in determining the application for the powder. The description of the shapes of irregularly shaped particles is very challenging because no two particles have exactly the same shape. Understanding and interpreting the morphology of powders exhibited by various polymers and their mixtures is important because it has a direct impact on performance of powders during fabrication of products. 1A sieve analysis of 100-percent recycled post-consumer HDPE showed that just under 50 percent of the powder has a size of 125 m or smaller (Table 7.14). Fine powders have a unique elongated shape, indicative of a solid-shearing process (Figures 7.10a,b), while larger fluff consists of fibrous, easily peeled particles.
©2001 CRC Press LLC
Hardness Shore D
Modulus MPa (psi)
Stength MPa (psi)
58
62
1,050 (153,000)
36.2 (5,250)
52.0
883 (128,000)
25.0 (3,630)
43.0
Flexural Properties
Melt Flow Rate g/10 min, 190◦ C, 21.6 kg
MPa (psi)
Elong %
HDPE
22.3 (3,230)
16
27 (0.5)
HDPE/PP 90/10
19.8 (2,870)
12
21 (0.4)
HDPE/PP 80/20
20.1 (2,920)
9
21 (0.4)
76
63
862 (125,000)
26.0 (3,770)
42.0
HDPE/PP 70/30
19.6 (2,840)
8
11 (0.2)
78
62
958 (139,000)
27 (3,880)
32.5
375
32 (0.6)
92
71
1,710 (248,000)
64.3 (9,330)
9.5
PP
32.6 (4,710)
©2001 CRC Press LLC
T1: SYV
Yield MPa (psi)
Heat Distortion Temperature ◦ C 66 psi
Char Count= 0
Feedstock/ Ratio
Ultimatea
Notched Izod Impact J/m (ft-lb/in)
QC: SYV/ABE
11:27
Tensile Properties
P2: SYV/ABE
March 15, 2001
Physical Properties of Injection-Molded Post-Consumer HDPE, PP, and HDPE/PP Blends Made Via S3 P Process and Comparison to Properties of Post-Consumer HDPE and PP.
P1: SYV/FYX
pb035-07
TABLE 7.13.
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
TABLE 7.14.
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
Sieve Analysis of 100-Percent Post-Consumer Recycled HDPE Powder after Pulverization (High-Shear S3 P at 0◦ ). Size Mesh
Microns
35 50 80 120 200 270 450 Fines
500 300 180 125 75 53 32
Weight Percent Retained 15.3 15.5 23.3 11.6 17.6 6.7 6.5 3.6
Using the same high shear and temperature conditions, the sieve analysis of S3 P-processed 100-percent recycled PP showed that less than three percent of the powder has a size of 125 m or smaller (Table 7.15). Fine PP powder (Figures 7.11a,b) has a shape similar to that of HDPE powder. A mixture of post-consumer recycled HDPE/PP at a 70/30 ratio processed via S3 P showed a reduction in the amount of the most coarse powder (30 mesh or >500 m) to 7.7 percent (Table 7.16) as compared to 15.3 percent (Table 7.14)
(a)
(b)
FIGURE 7.10 (a) SEM image of pulverized post-consumer HDPE at 40×. (b) SEM image of pulverized post-consumer HDPE at 250×.
©2001 CRC Press LLC
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
TABLE 7.15.
T1: SYV
Char Count= 0
Sieve Analysis of 100-Percent Post-Consumer Recycled PP Powder after Pulverization (High-Shear S3 P at 0◦ ). Size Mesh
Microns
35 50 80 120 200 270 450 Fines
500 300 180 125 75 53 35
Weight Percent Retained 41.5 26.9 18.2 7.0 4.7 1.3 0.4 0
of that powder size for HDPE alone. The fact that there was a reduction in the amount of the most coarse powder upon addition of PP to HDPE is not unexpected because PP, which has a lower crystallinity than HDPE, is much more pulverizable than HDPE [18]. This is evidenced by the fact that under less severe conditions of shear in the pulverizer, it is impossible to make a substantial quantity of powder from HDPE, although it is possible to produce a relatively fine powder from PP. Low shear conditions were accomplished by using a 23-mm diameter screw in the 25-mm diameter barrel. Table 7.17 shows the sieve analysis of 100-percent post-consumer recycled PP at relatively
(a)
(b)
FIGURE 7.11 (a) SEM image of pulverized post-consumer PP at 40x. (b) SEM image of pulverized post-consumer PP at 250x.
©2001 CRC Press LLC
Paul Lane Charles Whitman
(a)
(b)
(c)
Photo 1 (a) PTC lab manager John Rasmussen granulating post-consumer bottles with caps and printed labels without sorting; (b) resultant flake feedstock (2x) from granulation showing printed label pieces; (c) S3P-made uniform powder from feedstock (b) without added compatibilizer (unique because it is not possible to produce such powder by conventional processing). ©2001 CRC Press LLC ©2001 CRC Press LLC
Paul Lane
(a)
(c)
(d)
(e)
Charles Whitman
(b)
Photo 2 (a) One of the authors, Dr. Khait, with laboratory-scale Berstorff pulverizer that has pink powder at the discharge end; (b) flake feedstock of post-consumer HDPE/PP blend (90/10 ratio); (c) injection-molded test bar made by conventional processing, showing striations (uneven colors); (d) unique S3P-made uniform powder from feedstock (b) without added compatibilizer; (e) injection-molded test bar made from S3P powder (d) without pelletization, showing uniform color. ©2001 CRC Press LLC
(a)
(e)
(c)
(f)
Charles Whitman
(d)
(b)
(g) Photo 3 (a) Feedstock from unsorted post-industrial scrap (auto taillamps) containing polycarbonate, PMMA, and contaminants; (b) unique S3P-made uniform powder from feedstock (a) without added compatibilizer; (c) injection-molded test bar (with translucence) made from S3P powder (b) without pelletization; (d) mixed plastic feedstock from post-consumer wire and cable jacketing; (e) unique S3P-made uniform powder from feedstock (d) without added compatibilizer; (f) injectionmolded test bar made from S3P powder (e) without pelletization, showing uniform color; (g) further examples of discarded post-consumer items successfully processed via S3P [CDs, bubble wrap, floppy disks (with labels and sleeves), foam packaging, metallized printed film, cable jacketing, and phone cases]. ©2001 CRC Press LLC
©2001 CRC Press LLC
(a)
(b)
(c)
(d)
(e)
(f)
(g)
Charles Whitman
(h)
(i) Photo 4 (a-e) Post-consumer HDPE, LDPE, PP, PS, and PVC flakes; (f) pre-blended, post-consumer HDPE/LDPE/PP/PS/PVC flake feedstock (15/68/13/2/2 ratio); (g) unique S3P-made uniform powder from a pre-blended mixture (f) without added compatibilizer; (h) injection-molded test bar made from S3P powder (g) without pelletization, showing uniform color; (i) several test bars (h) subjected to tensile testing showing exceptionally high elongation at break. ©2001 CRC Press LLC ©2001 CRC Press LLC
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
TABLE 7.16.
T1: SYV
Char Count= 0
Sieve Analysis of 100-Percent Post-Consumer Recycled 70/30 HDPE/PP Blend (High-Shear S3 P at 0◦ ). Size Mesh
Microns
35 50 80 120 200 270 450 Fines
500 300 180 125 75 53 32
Weight Percent Retained 7.7 18.4 27.2 12.8 14.7 7.1 6.9 5.1
low shear where less than three percent of the powder has a size of 120 mesh (125 m) or finer. SEM of HDPE/PP powder (Figures 7.12a,b) revealed elongated shape of particles due to the high shear that occurred during the S3 P process. It has been demonstrated that although substantial tunability of powder size can be achieved by controlling the S3 P processing conditions, the conditions needed for particular powder characteristics are strongly polymer-dependent [18]. Material properties such as hardness, toughness, cohesivity, crystallinity, and others affect product particle size. Nevertheless, with appropriate process optimization, it is possible to make powder from 100-percent recycled feedstock that meets requirements for protective and decorative coatings (170 m or finer), rotational molding (500 m), extrusion, and direct injection molding (2,000 m) without prior pelletization. Further examples of intimate mixing during the S3 P process are addressed by Khait and Carr by studying copulverization of recycled LLDPE/PP mixtures derived from the post-consumer waste stream [17]. Properties of individual TABLE 7.17.
Sieve Analysis of 100-Percent Post-Consumer Recycled PP (Relatively Low-Shear S3 P at 0◦ ). Size Mesh
Microns
35 50 80 120 200 270 450 Fines
500 300 180 125 75 53 35
©2001 CRC Press LLC
Weight Percent Retained 75.67 14.26 6.38 2.01 0.84 0.67 0.17 0
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
(a)
(b)
FIGURE 7.12 (a) SEM image of pulverized post-consumer mixture of HDPE/PP at 70/30 ratio at 40× [10]. (b) SEM image of pulverized post-consumer mixture of HDPE/PP at 70/30 ratio at 250× [10].
LLDPE and PP powders are shown for comparison (Table 7.18). Surprisingly, ultimate strength and flexural properties of 80/20 and 70/30 LLDPE/PP blends were very close and approached those of LLDPE. Once again, it was possible to co-process LLDPE (MFR 111 g/10 min) with PP (MFR 9.5 g/10 min) with extreme viscosity ratios and produce homogeneous materials suitable for injection-molding applications. In addition to processing unsorted, multicolored post-consumer polyolefins, it has been shown that engineering resins and their mixtures as well as feedstocks contaminated with paint or protective coatings have been recovered with the S3 P process using the Berstorff PT-25 pulverizer [19]. For example, housings from discarded telephones and business machines are made from various grades of fire-retardant acrylonitrile butadiene styrene (ABS) or high-impact polystyrene (HIPS), which are incompatible with one another. It has been shown, however, that both recycled individual feedstocks and their 75/25 blend have been successfully converted into uniform powders via the S3 P process (Table 7.19). As can be seen from Table 7.19, S3 P-made ABS and HIPS had respectable mechanical properties including Notched Izod Impact strength of 1.1 ft-lb/in; their 75/25 mixture exhibited high tensile strength and flexural properties approaching those of ABS. Despite unmatched viscosity of individual components, their mixture was homogeneous and exhibited a melt flow rate of 5.1 g/10 min. Injection-molded parts had good gloss and surface appearance.
©2001 CRC Press LLC
LLDPE
Hardness Shore D
Modulus MPa (psi)
Strength MPa (psi)
13.4 (1,940)
190
37 (0.7)
49
55
540 (78,300)
21.8 (3,160)
Elong %
Melt Flow Rate g/10 min, 190◦ C, 21.6 kg 111
12.9 (1,870)
12
21 (0.4)
58
58
490 (71,000)
16.5 (2,400)
99.0
LLDPE/PP 70/30
12.9 (1,870)
8
32 (0.2)
59
62
510 (74,000)
18.5 (2,690)
70.0
375
32 (0.6)
92
71
1,710 (248,000)
64.3 (9,390)
32.5 (4,710)
Pull rate of 0.2 in/min. At 230◦ C, 2.16 kg load, MFR = 45.0.
©2001 CRC Press LLC
9.5b
T1: SYV
b
Ultimate MPa (psi)
LLDPE/PP 80/20
PP a
Yield MPa (psi)
Char Count= 0
Feedstock/ Ratio
Flexural Properties
QC: SYV/ABE
Heat Distortion Temperature ◦ C at 66 psi
11:27
Notched Izod Impact J/m (ft-lb/in)
Tensile Propertiesa
P2: SYV/ABE
March 15, 2001
Physical Properties of Injection-Molded Post-Consumer LLDPE, PP, and LLDPE/PP Blends Made Via the S3 P Process.
P1: SYV/FYX
pb035-07
TABLE 7.18.
Elong %
Hardness Shore D
Modulus MPa (psi)
Strength MPa (psi)
Flexural Properties
Melt Flow Ratea g/10 min
ABS
47.9 (6,950)
13
180 (3.4)
76
77
2,660 (386,000)
81.4 (11,800)
3.0
HIPS
26.9 (3,900)
24
69 (1.3)
76
81
2,170 (314,000)
50.0 (7,250)
8.6
ABS/HIPS 75/25
40.9 (5,940)
17
58 (1.1)
74
80
2,540 (368,000)
73.6 (10,670)
5.1
At 200◦ C, 5 kg load.
©2001 CRC Press LLC
T1: SYV
a
Yield MPa (psi)
Heat Distortion Temperature ◦ C at 264 psi
Char Count= 0
Feedstock/ Ratio
Notched Izod Impact J/m (ft-lb/in)
QC: SYV/ABE
11:27
Tensile Properties
P2: SYV/ABE
March 15, 2001
Physical Properties of Injection-Molded Post-Consumer HIPS, ABS, and 75/25 ABS/HIPS Mixture Made Via the S3 P Process.
P1: SYV/FYX
pb035-07
TABLE 7.19.
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
The recovery of metallized printed compact discs (CDs) made from polycarbonate (PC) was investigated [20]. Injection-molded samples of pre-consumer (industrial waste) CDs and post-consumer CDs from S3 P-made powders were uniform and had high gloss (Table 7.20). Both materials exhibited similar mechanical and thermal properties except for the Notched Izod Impact, which was higher for pre-consumer material, as expected due to the absence of a metallized layer (1.8 vs. 0.54 ft-lb/in, respectively). Using S3 P technology, it was possible to recycle CDs contaminated with paper labels, plastic envelopes, and fabric liners into homogeneous powders, which were directly processable into parts by injection-molding (without prior pelletization) [Photo 3(g) on the color insert includes CDs among other items. Photos 3(d–f) show processing of wire and cable jacketing]. It has been reported that rigid multicolored flakes derived from PVC waste bottles were converted into homogeneous light-colored powders via the solidstate pulverization process (Table 7.21) [20]. S3 P-made PVC powder exhibited high tensile strength and flexural modulus comparable to those of virgin PVC. The reprocessing of recycled plastics reconstituted via the S3 P technology has been investigated [20]. Each processing round included the production of powder followed by injection-molding (directly from powder) and grinding into feedstock for subsequent pulverization. The physical properties of post-consumer multicolored PP of unknown origin after four rounds of S3 P processing are shown in Table 7.22. After four cycles, the S3 P-made samples retained both high tensile strength at yield and high flexural properties. Elongation at break was slightly reduced (from 700 percent to 543 percent), but Notched Izod Impact strength remained unchanged. For comparison, the same post-consumer PP feedstock has been injectionmolded from regrind and then reground for the next molding cycle (Table 7.23). In contrast with the S3 P-made samples, injection-molded regrind exhibited a loss in tensile strength from 2,637 psi to 2,488 psi after three molding cycles; elongation at break dropped from 663 percent to 374 percent. Notched Izod Impact strength of samples after three cycles, however, was unaffected [21]. Khait and Torkelson demonstrated that crystalline polymers such as HDPE and PP, which are traditionally difficult to make into fine powders, have been successfully pulverized via the non-melting S3 P process using the PT-25 pulverizer [19]. In fact, Enikolopyan and his co-workers indicated that it was impossible to pulverize HDPE or PP into powder using their elastic strainassisted, high-temperature grinding approach [22]. Khait and Torkelson reported no deterioration in tensile strength, elongation, Notched Izod Impact strength, or flexural modulus and strength for parts injection-molded from pulverized recycled HDPE, LLDPE, and PP as compared to parts molded from recycled homopolymers (Table 7.24). In the case of LLDPE, there was a significant improvement in elongation by a factor of 5–6 for the pulverized material as compared to the as-received flakes. This is particularly important in view
©2001 CRC Press LLC
57.3 (8,310)
43.8 (6,360)
55
Post-consumer CD
56.8 (8,240)
47.6 (6,910)
21
a
At 300◦ C, 1.2 kg load.
©2001 CRC Press LLC
Elong %
96 (1.8)
80
28 (0.5)
76
Melt Flow Ratea g/10 min
Modulus MPa (psi)
Strength MPa (psi)
119
2,300 (334,000)
84.1 (12,200)
69.8
118
2,240 (325,000)
95.1 (13,800)
73.4
T1: SYV
Pre-consumer CD
Feedstock
Hardness Shore D
Flexural Properties
Char Count= 0
Ultimate MPa (psi)
Heat Distortion Temperature at ◦ C 66 psi
QC: SYV/ABE
11:27
Yield MPa (psi)
Notched Izod Impact J/m (ft-lb/in)
P2: SYV/ABE
March 15, 2001
Tensile Properties
P1: SYV/FYX
pb035-07
Physical Properties of Injection-Molded 100-Percent Recycled CDs Made via the S3 P Process.
TABLE 7.20.
a
At 190◦ C, 2.16 kg load.
©2001 CRC Press LLC
50.3 (7,300)
Elong % 11
620 (11.6)
Hardness Shore D 80
63
Flexural Properties Modulus MPa (psi)
Strength MPa (psi)
2,592 (375,900)
78 (11,455)
Melt Flow Ratea g/10 min 4.8
T1: SYV
Rigid PVC
Ultimate MPa (psi)
Heat Distortion Temperature at ◦ C 66 psi
Char Count= 0
Feedstock
Yield MPa (psi)
Notched Izod Impact J/m (ft-lb/in)
QC: SYV/ABE
11:27
Tensile Properties
P2: SYV/ABE
March 15, 2001
Key Physical Properties of Injection-Molded 100-Percent Post-Consumer PVC after S3 P Processing.
P1: SYV/FYX
pb035-07
TABLE 7.21.
Melt Flow Rateb g/10 min
50
67
41.1
33.0 (4,790)
50
68
42.4
915 (133,000)
33.4 (4,840)
50
67
51.6
895 (130,000)
32.1 (4,650)
53.5
68
51.2
PP Cycle 1 S3 P
26.4 (3,840)
21.1 (3,060)
701
31 (0.6)
772 (112,000)
31.5 (4,570)
PP Cycle 2 S3 P
27.2 (3,950)
20.0 (2,910)
765
33 (0.6)
871 (126,000)
PP Cycle 3 S3 P
27.6 (4,010)
18.4 (2,670)
741
32 (0.6)
PP Cycle 4 S3 P
27.9 (4,050)
17.4 (2,530)
543
25 (0.5)
Pull rate of 2 in/min. At 230◦ C, 2.16 kg load.
©2001 CRC Press LLC
Elong %
Heat Distortion Temperature ◦ C at 264 psi
Hardness Shore D
Material/ Cycle
Ultimate MPa (psi)
T1: SYV
Strength MPa (psi)
Yield MPa (psi)
Char Count= 0
b
Modulus MPa (psi)
Flexural Properties
QC: SYV/ABE
a
Notched Izod Impact J/m (ft-lb/in)
11:27
Tensile Propertiesa
P2: SYV/ABE
March 15, 2001
Key Physical Properties of Injection-Molded 100-Percent Post-Consumer Recycled PP after Four Cycles Utilizing S3 P Process.
P1: SYV/FYX
pb035-07
TABLE 7.22.
Melt Flow Rateb g/10 min
47
68
31.2
32.0 (4,650)
49
67
32.1
879 (128,000)
33.5 (4,860)
53
70
33.5
923 (134,000)
34.0 (4,930)
52.5
69
35.1
26.7 (3,880)
18.1 (2,640)
663
31 (0.6)
744 (108,000)
29.4 (4,270)
Cycle 1 Molding
27.4 (3,980)
17.9 (2,610)
499
28 (0.5)
871 (126,000)
Cycle 2 Molding
27.9 (4,050)
18.1 (2,630)
469
28 (0.5)
Cycle 3 Molding
27.5 (3,990)
17.1 (2,490)
374
28 (0.5)
Pull rate of 2 in/min. At 230◦ C, 2.16 kg load. Regrind prior to further molding.
©2001 CRC Press LLC
T1: SYV
Hardness Shore D
PP as receivedc
Elong %
Heat Distortion Temperature ◦ C at 264 psi
Char Count= 0
c
Strength MPa (psi)
Ultimate MPa (psi)
QC: SYV/ABE
b
Modulus MPa (psi)
Flexural Properties
Yield MPa (psi)
Material/ Cycle
a
Notched Izod Impact J/m (ft-lb/in)
11:27
Tensile Propertiesa
P2: SYV/ABE
March 15, 2001
Key Physical Properties of Injection-Molded Post-Consumer PP Regrind after Three Molding Cycles.
P1: SYV/FYX
pb035-07
TABLE 7.23.
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
TABLE 7.24.
T1: SYV
Char Count= 0
Key Properties of Post-Consumer Polyolefins—Effect of S3 P Process. Tensile Properties Yield MPa (psi)
Ultimate MPa (psi)
HDPE—not processed by S3 P
—
22.3 (3,230)
12
27 (0.5)
—
—
HDPE—processed by S3 P
—
22.0 (3,190)
16
27 (0.5)
1,050 (152,000)
36.2 (5,250)
13.6 (1,970)
—
33a
42 (0.8)
—
—
LLDPE—processed 13.4 by S3 P (1,940)
—
190a
37 (0.7)
540 (78,300)
21.8 (3,160)
PP—not processed 33.3 (4,830) by S3 P
—
330a
37 (0.7)
1,900 (276,000)
59.3 (8,600)
PP—processed by S3 P
—
375a
32 (0.6)
1,710 (248,000)
64.3 (9,330)
Feedstock
LLDPE—not processed by S3 P
a
Notched Flexural Properties Izod Impact Modulus Strength J/m MPa MPa (ft-lb/in) (psi) (psi)
32.5 (4,710)
Elong %
Pull rate of 0.2 in/min.
of the fact that elongation properties are often found to decrease in recycled materials. In Table 7.25, key properties of blends of post-consumer plastics are reported in comparison to properties of virgin polyolefins listed at the bottom of the table. The S3 P process allowed up to 30 weight-percent incorporation of PP and HDPE while maintaining reasonable, although not necessarily optimal, mechanical properties. Similar observations can be made regarding copulverized blends of recycled LLDPE and PP. Another blend (40/60 HDPE/LLDPE) pulverized via the S3 P process had an elongation at break of 130 percent, whereas material processed without pulverization had an elongation of only 17 percent, a difference of a factor of 7–8. The copulverized blend exhibited a yield tensile strength comparable to the ultimate tensile strength (no yield exhibited) for the material that was not copulverized [18]. Finally, four-, five-, and six-component blends of recycled plastics can be made via S3 P processing, which have key mechanical properties that are highly competitive with the properties of virgin HDPE and LLDPE. The percent elongation in these blends averages to about 40 percent of that for the virgin HDPE material, and the Notched Izod Impact strength is typically about half that of virgin PP. Thus, these multicomponent blends could be useful in applications where flexural properties and ultimate tensile strength need to be comparable to virgin polyolefin properties, but where
©2001 CRC Press LLC
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
Key Properties of Blends of Post-Consumer Plastics after S3 P Processing, with Comparison to Properties of Virgin Polyolefins.
TABLE 7.25.
Feedstock
a
3
S P
Tensile Properties Notched Izod Impact Ultimate J/m MPa Elong (ft-lb/in) (psi) %
Modulus MPa (psi)
Strength MPa (psi)
Flexural Properties
HDPE/PP 90/10
Yes
19.8 (2,870)
12
21 (0.4)
883 (128,000)
25.0 (3,630)
HDPE/PP 80/20
Yes
20.1 (2,920)
9
21 (0.4)
862 (125,000)
26.0 (3,770)
HDPE/PP 70/30
Yes
19.6 (2,840)
8
11 (0.2)
958 (139,000)
26.8 (3,880)
LLDPE/PP 80/20
Yes
12.9 (1870)
12
21 (0.4)
490 (710,00)
17.0 (2,470)
LLDPE/PP 70/30
Yes
12.9 (1,870)
8
32 (0.6)
510 (74,000)
19.0 (2,760)
HDPE/LLDPE 60/40
No
19.6 (2,840)
13
21 (0.4)
807 (117,000)
27.4 (3,970)
HDPE/LLDPE 60/40
Yes
18.1 (2,640)
22
21 (0.4)
758 (110,000)
29.0 (4,210)
HDPE/LLDPE 40/60
No
17.4 (2,520)
17
27 (0.5)
738 (107,000)
25.0 (3,630)
HDPE/LLDPE 40/60
Yes
16.4 (2,380) (yield)
130
32 (0.6)
662 (96,000)
25.0 (3,630)
HDPE/LLDPE/PP 60/30/10
No
18.5 (2,680)
9
21 (0.4)
793 (115,000)
27.2 (3,950)
HDPE/LLDPE/PP 60/30/10
Yes
19.2 (2,780)
13
21 (0.4)
972 (141,000)
32.0 (4,640)
HDPE/LLDPE/PP/PVC 55/30/10/5
Yes
16.5 (2,390)
13
11 (0.2)
800 (116,000)
23.0 (3,340)
HDPE/LLDPE/PP/PET/PVC 40/30/5/20/5
Yes
16.1 (2,340)
9
11 (0.2)
980 (142,000)
25.0 (3,630)
HDPE/LLDPE/PP/PS/PVC 50/30/10/5/5
Yes
15.6 (2,260)
13
16 (0.3)
807 (117,000)
22.0 (3,190)
HDPE/LLDPE/PP/PET/PS/PVC 40/30/5/15/5/5
Yes
13.7 (1,990)
6
5 (0.1)
910 (132,000)
22.0 (3,190)
HDPE/PET/PP/PS 50/30/10/10
Yes
30.8 (4,470)
10
32 (0.6)
1,489 (216,000)
46.0 (6,670)
HDPE Q LB5602 copolymer virgina
No
20.1 (2,920)
24
380 (7.1)
1,117 (162,000)
36.0 (5,220)
LLDPE Q GA601 virgina
No
16.5 (2,390)
480
—
300 (43,500)
13.0 (1,880)
PP Q 8020 virgina
No
28.2 (4,090)
38
21 (0.4)
1,427 (207,000)
48.0 (6,960)
Source of material: Equistar Chemical Co. c 1999, Marcel Dekker, Inc. Reproduced with permission from Reference [18]. Copyright
©2001 CRC Press LLC
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
Key Physical Properties of S3 P-Made Powders Made from Recycled PP and Blends with Polyolefin Affinity Plastomer PT 1407 (POP). TABLE 7.26.
Tensile Propertiesa Feedstock/ Ratio Recycled PP 100%
a b
Notched Izod Heat Melt Flow Impact Distortion Rateb J/m Temperature Hardness (ft-lb/in) ◦ C at 264 psi Shore D g/10 min
Yield MPa (psi)
Elong %
32.4 (4,700)
110
21 (0.4)
73
59
45.0
95/5 PP/POP
33.7 (4,890)
80
37 (0.7)
75
48
25.8
90/10 PP/POP
31.0 (4,500)
190
37 (0.7)
72
47
28.0
80/20 PP/POP
28.9 (4,190)
150
42 (0.8)
65
47
20.9
Pull rate of 2 in/min. At 230◦ C, 2.16 kg load.
some loss in impact strength and elongation are acceptable. Especially important for four of the multicomponent blends is the presence of 5 weight percent PVC in the blends, including two instances of 15 and 20 weight percent PET. This represents an important advance in recycling technology as it is known that PET waste contaminated with PVC cannot be melt processed by conventional methods due to the degradation of PVC at temperatures well below the melt temperature of PET. Khait and Carr investigated toughening of recycled PP by an addition of a small amount of polyolefin plastomer (POP). POPs are ethylene alpha-olefin polymers containing various amounts of co-monomer, providing a range of elastomeric properties [23]. Affinity PT 1407 from Dow Chemical Co. was used in blends containing 5.10 and 20 weight percent of POP. Powders obtained via S3 P were injection-molded, and physical properties of recycled PP and these blends are listed in Table 7.26. An addition of POP improved the Notched Izod Impact strength of recycled PP. Blends containing five and 10 weight percent of POP exhibited 40 percent improvement of the notched impact. Elongation at break for blends containing 10 and 20 weight percent of POP also increased, while tensile strength at yield remained unchanged. Processability of recycled PP/POP blends decreased compared to that of PP as expected, due to the presence of the elastomeric component. Furgiuele et al. showed that both high and low melt flow rates of recycled PP undergo significant reduction in molecular weight as evidenced by an increase in melt flow rate (Table 7.27) [24]. Due to a modular screw design, it is possible to employ different degrees of shear, depending on the quantity and location of the shearing (kneading) blocks. Each polymer was processed under low shear conditions (a 23-mm diameter screw in the 25-mm barrel) and under high
©2001 CRC Press LLC
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
TABLE 7.27.
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
Effect of Shear During Pulverization on Melt Flow Rate of Recycled PP.
Material
Melt Flow Rate (g/10 min) 230◦ C, 2.16 kg load
Low Melt Flow Rate
Pre-S3 P Post-S3 P (low shear) Post-S3 P (high shear)
6.6 6.9 12.0
High Melt Flow Rate
Pre-S3 P Post-S3 P (low shear) Post-S3 P (high shear)
37.0 39.0 67.0
shear conditions (a 25-mm diameter screw in the 25-mm barrel). Altering the design of the pulverizer can affect the magnitude of change of a material’s molecular weight distribution—at low shear conditions, PP undergoes little change in molecular weight, while at high shear conditions, the melt flow rate increases by a factor of two. In another study, Furgiuele et al. showed a major reduction in molecular weight of PS and hence substantial chain scission from S3 P [25]. Therefore, with high shear, PP and PS can be pulverized to yield chain scission and to generate free radicals sufficient to compatibilize a PP/PS blend. It has been demonstrated, via characterization of glass transition temperatures for the polystyrene-rich phase in PS/PP blends, that mixtures processed via the S3 P method achieve much more intimate mixing of components than is observed in conventional melt processing [26]. This effect, along with the known capacity of S3 P to cause chain scission and to produce free radicals, has the potential to lead to an in-situ compatibilization of the dissimilar polymers via block copolymer formation at the interface. Another example of the utility of S3 P technology for mixed discarded plastics waste derived from consumer goods consisting of PP and PS blend at 95/5 ratio is shown in Table 7.28. (Because PP and PS are incompatible, they cannot be recycled as a blend by a conventional technology and must be separated into two individual streams [20].) Processability of S3 P-made material from PP/PS blend at 95/5 ratio made at high shear is significantly higher than that of the material made at low shear and by conventional melt mixing using a twin-screw extruder (29 g/10 min vs. 16.2 g/10 min, respectively). This decrease in MFR is expected, due to the rupture of chemical bonds during the S3 P process and a reduction of molecular weight. Elongation at break of the blend made at high shear increased by a factor of 2 as compared with that of melt-processed material, while other physical properties were similar.
©2001 CRC Press LLC
132
27 (0.5)
59
18.0 (2,610)
165
32 (0.6)
18.6 (2,700)
98
32 (0.6)
PP/PS 95/5 (low shear)
32.6 (4,720)
17.9 (2,590)
PP/PS 95/5 (high shear)
31.1 (4,504)
PP 95/5—melt mixed with TSEc
32.1 (4,660)
a
Pull rate of 2 in/min. MFR at 230◦ C, 2.16 kg. c Twin-screw extruder. Source of material: Butler-MacDonald Corp. b
©2001 CRC Press LLC
Strength MPa (psi)
Melt Flow Rateb g/10 min
71
1,250 (181,000)
41.1 (5,970)
15.3
58
71
1,190 (172,000)
40.5 (5,880)
29.0
60
72
1,240 (180,000)
42.1 (6,100)
16.2
T1: SYV
Hardness Shore D
Elong %
Modulus MPa (psi)
Char Count= 0
Heat Distortion Temperature ◦ C at 264 psi
Ultimate MPa (psi)
Feedstock/Ratio
Flexural Properties
Notched Izod Impact J/m (ft-lb/in)
Yield MPa (psi)
QC: SYV/ABE
11:27
Tensile Propertiesa
P2: SYV/ABE
March 15, 2001
Physical Properties of Injection-Molded Recycled PP/PS Mixture at 95/5 Ratio Made Via the S3 P Process at Low and High Shear.
P1: SYV/FYX
pb035-07
TABLE 7.28.
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
Additional examples demonstrated by the researchers at Northwestern University’s PTC concerned mixed plastics recovery via the S3 P process containing ABS and medium-impact polystyrene (MIPS); these are illustrated in Table 7.29 [19]. Once again, despite a known incompatibility between ABS and MIPS, a 5/95 ABS/MIPS blend made via S3 P processing exhibited properties close to those of MIPS with a slight reduction of Notched Izod Impact. Similar trends have been observed for a reversed ratio of 95/5 ABS/MIPS blend made via the shear pulverization process. As Table 7.30 shows, the physical properties of recycled polycarbonate (PC) contaminated with unknown amounts of PP and nylon, as well as recycled PC with silicone coating, have also been studied [19]. Although PC is known to be incompatible with both PP and nylon, the resultant material processed via S3 P has tensile and flexural properties similar to those of virgin PC [see Photos 3(a–c) on the color insert]. It should be noted that the MFR of S3 P-made PC powder derived from PC contaminated with PP and nylon increased four-fold compared with the MFR of the feedstock (108 g/10 min vs. 25.7 g/10 min, respectively), which is indicative of chain scission during the pulverization process. It has also been shown that painted car parts made from thermoplastic vulcanizate (TPV) can be reconstituted without paint removal by the S3 P process into powder for creating valuable materials for re-use [20]. Injection-molded test specimens molded directly from S3 P powder had a glossy appearance and did not delaminate upon breaking. As can be seen from Table 7.31, elongation at break of S3 P-made material was significantly higher than that of the meltmixed sample processed with conventional twin-screw extruder (380 percent vs. 116 percent, respectively). A combination of high Notched Izod Impact of 6.5 ft-lb/in and exceptionally high elongation at break of 380 percent is suggestive of numerous applications for the S3 P powder derived from painted TPV parts, including under-the-hood parts. The unique ability of the S3 P technology to convert complex, unsorted mixtures of all six plastics found in the waste stream into a uniform powder for further re-use has been demonstrated [20]. This is a significant advance in recycling inasmuch as plastics are a rapidly growing segment of the MSW. According to the 1998 Report of the Environmental Protection Agency (EPA), the following plastics are used in durable or non-durable goods: PET, HDPE, PVC, LDPE/LLDPE, PP, PS, and others [27]. Khait et al. described several materials that were successfully processed via S3 P using the waste compositions and ratios adapted from the EPA report [20]. Three mixtures contained either five or six components, while one mixture did not contain HDPE and PET because both HDPE and PET are being sorted and successfully recycled. Although feedstocks obtained from recyclers had an array of bright colors,
©2001 CRC Press LLC
Ultimate MPa (psi)
30.9 (4,480)
27.0 (3,920)
27
52 (1.0)
78
42.4 (6,150)
13
180 (3.3)
26.7 (3,900)
14
42.4 (6,150)
12
ABS ABS/MIPS 5/95
31.6 (4,580)
ABS/MIPS 95/5 a b
Pull rate of 2 in/min. MFR at 230◦ C, 5.0 kg.
©2001 CRC Press LLC
Modulus MPa (psi)
Strength MPa (psi)
81
2,570 (373,000)
78
78
32 (0.6)
79
120 (2.3)
78
Melt Flow Rateb g/10 min Before S3 P
After S3 P
50.3 (7,300)
8.2
10.8
2,310 (335,000)
67.0 (9,710)
3.1
3.2
79
2,520 (366,000)
50.4 (7,310)
8.3
10.3
79
2,270 (329,000)
66.3 (9,610)
3.6
3.2
T1: SYV
MIPS
Hardness Shore D
Yield MPa (psi)
Elong %
Flexural Properties
Heat Distortion Temperature ◦ C at 264 psi
Char Count= 0
Feedstock/Ratio
Notched lzod Impact J/m (ft-lb/in)
QC: SYV/ABE
11:27
Tensile Propertiesa
P2: SYV/ABE
March 15, 2001
Physical Properties of Injection-Molded Recycled ABS, MIPS, and Their Blends Made Via the S3 P Process.
P1: SYV/FYX
pb035-07
TABLE 7.29.
PC with PP and nylon
6.0 (8,700)
56.0 (8,124)
106
300 (5.6)
124
PC with silicone coating
(60.8) (8,810)
(53.7) (7,780)
91
110 (2.0)
125
a
Modulus MPa (psi)
Strength MPa (psi)
82
2,250 (326,000)
84
2,290 (332,000)
Melt Flow Rateb g/10 min Before S3 P
After S3 P
94.5 (13,700)
25.7
108.2
94.5 (13,700)
14.2
15.3
Pull rate of 2 in/min. MFR at 300◦ C, 1.2 kg. Sources of material: 2 = PC with PP and nylon from Butler-MacDonald Corp. of Indianapolis, Indiana; 3 = PC with Si coating from H. Sattler Plastics Co., Chicago, Illinois.
b
©2001 CRC Press LLC
T1: SYV
Ultimate MPa (psi)
Char Count= 0
Hardness Shore D
Feedstock/Ratio
Yield MPa (psi)
Elong %
Flexural Properties
Heat Distortion Temperature ◦ C at 264 psi
QC: SYV/ABE
Notched lzod Impact J/m (ft-lb/in)
11:27
Tensile Propertiesa
P2: SYV/ABE
March 15, 2001
Physical Properties of Injection-Molded Recycled Polycarbonate (PC) Contaminated with PP and Nylon and PC with Silicone Coating Processed via S3 P.
P1: SYV/FYX
pb035-07
TABLE 7.30.
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
TABLE 7.31.
T1: SYV
Char Count= 0
Physical Properties of TPV Painted Parts from Automotive Waste after S3 P Processing Made with PT-25 Pulverizer. Tensile Propertiesa
Feedstock
Yield MPa (psi)
Ultimate MPa Elong (psi) %
Notched Izod Impact J/m (ft-lb/in)
Flexural Properties Modulus Strength MPa MPa (psi) (psi)
Melt Flow Rateb g/10 min
Painted TPV (high shear)
11.8 (1,710)
12.3 (1,790)
380
350 (6.5)
393 (57,000)
14.7 (2,130)
10.2
Painted TPV (Melt Mixed with TSEc )
13.1 (1,910)
10.7 (1,551)
116
330 (6.1)
475 (68,900)
15.2 (2,200)
9.5
a
Pull rate of 2 in/min. MFR at 230◦ C, 2.16 kg. c Twin-screw extruder. b
including pink, peach, white, gray, yellow, green, red, blue, orange, purple, brown, and even black, the resultant powders after S3 P processing have uniform, pastel colors. This phenomenon is particularly important because it allows recycling of single-colored plastics without sorting by color through intimate solid-state mixing during the pulverization process. Key physical properties of S3 P-made materials derived from multi-component (unsorted) plastics waste are given in Table 7.32. One mixture consisting of HDPE/LDPE/PP/PS/PVS at a 15/68/13/2/2 ratio even exhibited a yield strength of 2,250 psi during the tensile test, very high elongation at break of 510 percent, and high Notched Izod Impact strength of 4.2 ft-lb/in [see Photos 4(a–i) on the color insert]. The MFR data suggested that S3 P-made materials derived from unsorted multicolored post-consumer plastics waste could be successfully re-used in film or extrusion applications. Additionally, every ingredient of these five- and six-component mixtures (with the exception of PVC) has been processed with the PT-25 pulverizer (Table 7.33). Soft PVC flakes could not be successfully pulverized due to the presence of an unknown amount of plasticizer. It is noteworthy that HDPE and PP powders had very high elongation at break (470 percent and 700 percent respectively). Additionally, HDPE powder exhibited a yield value of 3,340 psi, which is comparable to that of virgin HDPE. By contrast, LDPE powder had an exceptionally high Notched Izod Impact strength of 11.7 ft-lb/in but relatively lower elongation of 130 percent;. Both HDPE and LDPE powder had MFR values of 0.8 g/10 min and 0.4 g/10 min, which would be suitable for blow-molding applications. Due to the intense shearing forces inherent in S3 P, the particles produced have an elongated morphology, which can be seen by scanning electron microscopy (SEM). All micrographs of the powders were taken using a Hitachi S-510 microscope. With SEM, differences in the morphology of the
©2001 CRC Press LLC
Melt Flow Ratec g/10 min
18.6 (2,710)
510
220 (4.2)
42
59
500 (72,800)
17.1 (2,480)
1.1
LDPE/PP/PS/PVC 50/19/18/13
—
21.6 (3,140)
15
21 (0.4)
64
65
942 (137,000)
30.4 (4,410)
1.9
HDPE/LDPE/PP/PET/PS/PVC 40/30/5/15/5/5
—
18.5 (2,690)
10
16 (0.3)
47
65
834 (121,000)
25.0 (3,620)
2.0
HDPE/LDPE/PP/PET/PS/PVC 22/34/13/10/12/8
—
18.9 (2,740)
9
16 (0.3)
53
65
942 (137,000)
27.4 (3,980)
2.6
Sources of material: HDPE = Eaglebrook plastics; LDPE = Maine Plastics; PP = St. Joseph; PS = Maine Plastics; PET = Groot recycling; PVC = Clearvue, Inc. Pull rate of 2 in/min. At 190◦ C, 2.16 kg load.
©2001 CRC Press LLC
T1: SYV
Strength MPa (psi)
Char Count= 0
c
Modulus MPa (psi)
QC: SYV/ABE
15.5 (2,250)
HDPE/LDPE/PP/PS/PVC 15/68/13/2/2
b
Hardness Shore D
Ultimate MPa (psi)
Feedstocka /Ratio
a
Heat Distortion Temperature ◦ C at 264 psi
Yield MPa (psi)
Elong %
Flexural Properties
Notched Izod Impact J/m (ft-lb/in)
11:27
Tensile Propertiesb
P2: SYV/ABE
March 15, 2001
Physical Properties of Injection-Molded Multi-Component Post-Consumer Plastics from the MSW Stream Made by the S3 P Process.
P1: SYV/FYX
pb035-07
TABLE 7.32.
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
Physical Properties of Individual Components of Post-Consumer Mixed-Color Plastics from the MSW Stream after S3 P Processing with a PT-25 Pulverizer.
TABLE 7.33.
Tensile Propertiesa
c
Strength MPa (psi)
Melt Flow Rate g/10 min
Ultimate MPa (psi)
HDPE
23.0 (3,340)
16.8 (2,440)
470
170 (3.2)
717 (104,000)
21.9 (3,180)
0.8b
LDPE
13.9 (2,010)
15.9 (2,310)
130
630 (11.7)
308 (44,700)
13.2 (1,920)
0.4b
PP
26.5 (3,840)
21.1 (3,060)
700
32 (0.6)
772 (112,000)
31.5 (4,570)
41.1c
40.1 (5,810)
5
11 (0.2)
3,120 (453,000)
55.6 (8,060)
38.1c
PS
b
Modulus MPa (psi)
Flexural properties
Yield MPa (psi)
Feedstock
a
Notched Izod Impact J/m (ft-lb/in)
Elong %
Pull rate of 2 in/min. At 190◦ C, 2.16 kg. At 230◦ C, 2.16 kg.
HDPE/LDPE/PP/PS/PVC 15/68/13/2/2 blend produced on the laboratory-scale PT-25 pulverizer and the production-scale PT-60 pulverizer are also visible (Figures 7.13a,b and 7.14a,b). Since the late 1990s, the S3 P process has been in transition from a laboratory scale to a commercial scale using a first-of-its-kind production-size PT-60 Berstorff pulverizer at the Polymer Technology Center at Northwestern University. Scale-up efforts have included the processing of a five-component multicolored blend derived from the film portion of the post-consumer waste stream. This mixture consisted of HDPE/LDPE/PP/PS/PVC at a 15/68/13/2/2 ratio, a composition adapted from the 1998 EPA report on characterization of the MSW in the United states [27]. A 1,000-pound batch was pre-blended by Major Prime Plastics in Villa Park, Illinois using a commercial blender. Physical properties of this post-consumer mixture made with the PT-60 pulverizer are presented in Table 7.34. For comparison, the same mixture was processed with a laboratory-scale PT-25 Berstorff pulverizer and with a conventional co-rotating twin-screw Berstorff extruder ZE-25 using melt-mixing. The most striking property achieved with commercial-scale pulverization equipment was a Notched Izod Impact of 10.4 ft-lb/in as compared with that of 6.8 ft-lb/in measured on material made with the laboratory-scale PT-25 pulverizer. Melt-mixed material exhibited a Notched Izod Impact of 4.6 ft-lb/in, significantly lower than that for pulverized material. Specific energy consumption values for different polymers and their blends will be generated using the production-scale PT-60 pulverizer.
©2001 CRC Press LLC
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
(a)
T1: SYV
Char Count= 0
(b)
FIGURE 7.13 (a) SEM image of post-consumer mixture of HDPE/LDPE/PP/PS/PVC at 15/68/13/2/2 ratio pulverized with PT-25 at 40×. (b) SEM image of post-consumer mixture of HDPE/LDPE/PP/PS/PVC at 15/68/13/2/2 ratio pulverized with PT-25 at 250×.
(a)
(b)
FIGURE 7.14 (a) SEM image of post-consumer mixture of HDPE/LDPE/PP/PS/PVC at 15/68/13/2/2 ratio pulverized with PT-60 at 40×. (b) SEM image of post-consumer mixture of HDPE/LDPE/PP/PS/PVC at 15/68/13/2/2 ratio pulverized with PT-60 at 250×.
©2001 CRC Press LLC
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
Physical-Property Comparison of Post-Consumer Mixture of HDPE/LDPE/PP/PS/PVCa at 15/68/13/2/2 Ratio Made with Different Processing Equipment.
TABLE 7.34.
Tensile Propertiesb
a b c
Notched Izod Impact J/m (ft-lb/in)
Modulus MPa (psi)
Strength MPa (psi)
Melt Flow Ratec g/10 min
Flexural Properties
Processing Equipment
Ultimate MPa (psi)
S3 P with PT-60 Pulverizer
16.8 (2,430)
144
560 (10.4)
461 (66,900)
15.6 (2,260)
0.9
S3 P with PT-25 Pulverizer
17.0 (2,460)
220
360 (6.8)
467 (67,700)
17.2 (2,490)
0.9
Melt Mixed with Twin-screw Extruder
14.7 (2,130)
157
250 (4.6)
432 (62,700)
16.1 (2,330)
1.8
Elong %
1000 lb pre-blended batch. Pull rate of 2 in/min. At 190◦ C, 2.16 kg.
SUMMARY Plastics have a significant impact on the environment, starting with raw material processing, continuing through parts manufacture, and ending with either post-consumer recycling or disposal. Management of MSW continues to be a high priority world-wide. In many areas, disposal of plastic waste into landfills has become prohibitive due to high costs, legislative pressure, and growing environmental concerns. Incentives for using recycled plastics include new laws, lower cost, consumer demand, and reduction in landfill availability. Until the late 1980s, plastic recycling was slow to develop due to the variability of supply, contamination of materials, multiplicity of colors, use of expensive automated sorting equipment, and underdeveloped end-use applications. Other complications include a variety of additives, fillers, modifiers, and the wide range of melt flow rates used in plastics production, as well as the immiscibility of plastics. At the end of the 20th century, only the two thermoplastics—PET and HDPE—were being recycled successfully in the United States. A steadily increasing demand for recycling has resulted in the development of new technologies and equipment to recover plastics. Post-consumer plastics have to be collected, sorted, granulated, cleaned, pelletized, and packaged for processing as the secondary feedstock by injection-molding or extrusion. For plastics recyclers, the development of appropriate markets for reclaimed plastics is one of their most difficult tasks. Key factors in the success of recycled materials are their consistent quality and price advantage relative to those of the virgin resins. It is clear that state and federal legislation as well as the concern for a green environment will result in a sharp growth in recycling in the 21st century.
©2001 CRC Press LLC
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
The pollution-free Solid-State Shear Pulverization (S3 P) process is an advanced technology for recycling plastics that has been under development at the Polymer Technology Center (PTC) at Northwestern University in Evanston, Illinois, since the late 1980s. This process creates new plastic materials by pulverization of mixtures of dissimilar plastics without melting, producing powders that can be used to make new products. The S3 P process has great potential for recycling commingled plastics without sorting and for improving their physical properties when compared with existing processes. This technology can convert either single or commingled (unsorted) multicolored plastic derived from preor post-consumer waste into high-quality powders of leveled color to be used in virtually any conventional fabrication technique. The homogenization of color via the S3 P process could alleviate the cost associated with sorting both singleand multi-component plastics by color. Therefore, the S3 P process is capable of addressing both the deleterious effect of recycling on the properties of thermoplastics and the aesthetic issues related to the efficient re-use of thermoplastics. The results obtained at the PTC confirm that S3 P-made materials exhibit mechanical and physical properties higher than those processed by conventional melt mixing. Moreover, the S3 P process can be used to mix like polymers with unmatched viscosity, something not possible through conventional melt mixing. The commercialization of the S3 P process for powder production and for mixed plastics recovery began in the late 1990s. In 1997, Material Sciences Corporation of Elk Grove Village, Illinois, entered into an agreement with Northwestern University to commercialize and continue the development of S3 P technology for creating new polymeric materials from virgin and recycled feedstocks. The first Berstorff production-scale pulverization line became operational at the PTC in 1998. Trials continue to be conducted to establish the costs and capabilities of the S3 P process for recycling dissimilar plastics without sorting. Ongoing studies seek to understand relationships between the processing parameters, the particle size, and the particle size distribution of the resultant powders, as well as to understand the mechanochemistry that permits in-situ compatibilization of ordinarily immiscible polymers. The continued advances in the S3 P process offer an opportunity to facilitate plastic recycling by creating value-added materials made from discarded plastics for a wide variety of applications and markets.
REFERENCES Note: Some names of people (such as Enikolopyan, Yerina, Nepomnyaschy, and Akopyan) have variant spellings in English transliteration. While the chapter text uses a single spelling, the variations are preserved in this reference list. 1. Chen, I. M. and Shiah, C. M. Proceeding of the Annual Technical Conference of the Society of Plastics Engineers, ANTEC ’89, New York, pp. 1802–1806 (1989).
©2001 CRC Press LLC
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
2. Xanthos, M. Conference Proceedings of Compalloy, Jersey City, New Jersey (1992). 3. Khait, K. “Novel Elastic-Deformation Grinding Process for Commingled Plastic Waste Recovery,” Proceedings of the Annual Technical Conference of the Society of Plastics Engineers, ANTEC ’94, San Francisco, p. 3006 (1994). 4. Bridgman, P. W. The Physics of High Pressure, New York: The Macmillan Co. (1931). 5. Wolfson, S. A., Khait, K., and Dienst, M. “Extrusion-Pulverizing Process is a Low-Cost Route to Quality Resin Powders,” Modern Plastics, pp. 63–64 (1994). 6. Mack, M. “Twin-Screw Machines Explore Solid-State Extrusion,” Plastic Technology, pp. 75–77 (1993). 7. Enikolopov, N. S., Nepomnyaschy, A. I., Filmakova, L. A., Krasnokutsky, V. P., Kurakin, L. I., Akopian, E. L., Markarian, K. A., Negmatov, S. S., Martkarimov, S. K, Polivanov, Y. A., Sherstnev, P. P., and Pavlev, V. B. US Patent 4,607,797 (1986). 8. Khait, K. and Petrich, M. A. “Novel Pulverization Process for Commingled Plastic Waste Recovery,” Proceedings of the Pack Allimentaire Technical Conference ’93, Chicago (1993). 9. Ehrig, R. J., ed. Plastics Recycling: Products and Processes. Munich: Carl Hanser Verlag (1992). 10. Khait, K. and Carr, S. H. “Mixed Polyolefin Powders Recycled Via Solid-State Shear Pulverization (S3 P) Process,” Proceedings of the Annual Technical Conference of the Society of Plastics Engineers, ANTEC ’98, Atlanta, pp. 2533–2537 (1998). 11. Khait, K. “Recycling of Unsorted Plastic Waste by New SSSE Pulverization Process,” Proceedings of the Annual Technical Conference of the Society of Plastics Engineers, ANTEC ’95, Boston, pp. 2066–2070 (1995). 12. Dietering, M. L. “Particle Characterization of Recycled Polymeric Powders,” Master’s thesis, Northwestern University, Evanston, Illinois (1994). 13. Khait, K. “Advanced Reclamation Technology for Mixed Polymeric Waste,” Proceedings of the Annual Technical Conference of the Society of Plastics Engineers, ANTEC ’96, Indianapolis, Indiana, pp. 3120–3124 (1996). 14. Ogando, J. “Solid-State Extrusion Technique Tackles Commingled PCR,” Plastics Technology, p. 37 (1994). 15. Khait, K. “Advanced Elastic-Deformation Grinding Technology for Plastic Recycling,” Proceedings of the Annual Recycling Conference of the Society of Plastics Engineers, Schaumburg, Illinois, p. 72 (1994). 16. Wendorf, M. A. “Blends of Polypropylene with Recycled HDPE,” Proceedings of the Regional Technical Conference of the Society of Plastics Engineers, RETEC ’93, Schaumburg, Illinois (1993). 17. Khait, K. and Carr, S. H. “Value-Added Materials Made from Recycled Plastics,” Proceedings of the Annual Technical Conference of the Society of Plastics Engineers, ANTEC ’97, Toronto, pp. 3086–3090 (1997). 18. Khait, K. and Torkelson, J. M. “Solid-State Shear Pulverization of Plastics: A Green Recycling Process,” Polymer-Plastics Technology and Engineering, 38:445–457 (1999). 19. Khait, K. and Torkelson, J. M. “Environmentally-Benign Recycling Process for Commingled Plastics,” Proceedings of the Annual Recycling Conference of the Society of Plastics Engineers, ARC ’98, Chicago (1998). 20. Riddick, E. G., Khait, K., and Torkelson, J. M. “New Environmentally-Friendly Recycling Technology for Mixed Multicolor Plastics,” Proceedings of the Annual Recycling Conference of the Society of Plastics Engineers, ARC ’99, Detroit (1999). 21. Khait, K. and Carr, S. H. “Mechanochemistry Effects in Recycled PP and its Blends During
©2001 CRC Press LLC
P1: SYV/FYX pb035-07
P2: SYV/ABE
March 15, 2001
22.
23.
24.
25.
26. 27.
QC: SYV/ABE 11:27
T1: SYV
Char Count= 0
Solid-State Shear Pulverization (S3 P),” Proceedings of the Annual Technical Conference of the Society of Plastics Engineers, ANTEC ’99, New York, pp. 3246–3250 (1999). Enikolopyan, N. S., Akopyan, Ye. L., Karnuilovl, A. Yu., Nikol’skii, V. G., and Khacharyan, A. M. “Production of Highly Disperse Powder Materials Based on Thermoplastics amd Thermoplastic Blends of Elastic-Deformation Grinding,” Polymer Science USSR, 30:2576 (1988). Khait, K. and Carr, S. H. “Mixed Polyolefin Powders Recycled Via Solid-State Shear Pulverization (S3 P) Process,” Proceedings of the Annual Technical Conference of the Society of Plastics Engineers, ANTEC 98, Atlanta, Georgia, pp. 2939–2941 (1998). Furgiuele, N., Khait, K., and Torkelson, J. M. “The Use of Solid-State Shear Pulverization for Polymer Blend and Polymeric Waste Compatibilization,” Proceedings of the Annual Recycling Conference of the Society of Plastics Engineers, ARC ’98, Chicago (1998). Furgiuele, N., Khait, K., and Torkelson, J. M. “Novel Approach for the Compatibilization of Polymer Blends and Polymeric Waste,” Polymeric Materials Science Engineering, 79:70 (1998). Furgiuele, N., Lebovitz, A. H., Khait, K., and Torkelson, J. M. “Novel Strategy for Polymer Blend Compatabilization: Solid-State Shear Pulverization,” Macromolecules, 33:225 (2000). United States Environmental Protection Agency Report. “Characterization of Municipal Solid Waste in the US: 1997 Update,” EPA 530-R-98-007 (1998).
©2001 CRC Press LLC
P1: FUY pb035-08
March 1, 2001
20:48
Char Count= 0
CHAPTER 8
Applying S3 P Technology to the Recovery of Used-Tire Rubber
STATE OF RUBBER RECYCLING
A
large quantity of scrap tires—about 270 million—is generated annually in the US. Most of these are disposed of in landfills, but some are re-used as ground rubber in such civil engineering applications as supplemental fuel in cement kilns or as tire-derived fuel as well as in rubberized asphalt, athletic surfaces, playgrounds, mats, and a variety of consumer products. At present, scrap tires are converted into so-called “crumb rubber” by either ambient or cryogenic grinding. Because of the high cost of liquid nitrogen used as a refrigerant in the cryogenic method, size reduction at an ambient temperature is used more often for coarse powder production. Tires are shredded into chips about 0.75-in long. This is followed by a magnetic separation of steel and removal of textile cord. The rubber chips are then reduced to rougher, smaller pieces by different milling devices in a series of screening and re-grinding operations to achieve the desired particle size. For the traditional rubber “reclaim,” crumb rubber is mixed with water, oil, and chemicals and is then heated under pressure. During this process, the carbon-sulfur bonds are ruptured, and the rubber becomes mostly devulcanized, so it is then capable of being shaped into slabs. These slabs are used by tire manufacturers as an alternative to virgin rubber for re-use in new tires or as an ingredient in other rubber products. Because reclaimed rubber has reduced elasticity, it is currently used for only about five percent of all new-tire production. Rader and Lemieux defined the terms recovery (re-use) and recycling as they are currently used by the processors [1]. Recovery (re-use) indicates the creation of something valuable from discarded post-consumer material, which otherwise would be disposed of, usually to a landfill. Recycling suggests reconstituting a
©2001 CRC Press LLC
P1: FUY pb035-08
March 1, 2001
20:48
Char Count= 0
material, either by more finely dividing or reshaping it or by chemically changing it into a different feedstock. Process scrap from plastic and rubber manufacturing is generally called regrind or industrial waste, and it is historically utilized at various levels by re-introducing regrind to a virgin feedstock. The tire rubber is usually composed of 40–50 percent rubber (styrene-butadiene rubber, natural rubber, and butyl rubber), 25–40 percent carbon black, and 10–15 percent low molecular weight additives (the exact composition depends on the specific type of tire). Ambiently ground tire rubber is usually produced with a cracker mill, it is supplied at 10 to 40 mesh (2,000 to 420 microns), and it is the least expensive recycled rubber on the market. Cryogenically ground rubber is available from 40 mesh size (420 microns) and finer. The cost of this material is much higher than that of ambiently ground rubber due to the cost of liquid nitrogen. However, finer rubber can be added as an ingredient in the rubber compound at higher levels than those for coarser rubber without any loss in processability and physical properties. Another technology for recycling used-tire rubber involves surface modification. Stark and Leighton modified vulcanized scrap rubber by an addition of a small amount of a liquid unsaturated curable polymer in combination with a curing agent [2]. The resultant dry mixture can be used as an additive to rubber compounds in high concentrations. Bauman developed a surface modification technology for scrap tire rubber using a reactive gas atmosphere, which causes a permanent chemical change to the rubber particles [3]. Bauman observed that surface-modified rubber has enhanced adhesion to plastics in rubber/plastic composites. Oliphant and Baker showed the use of cryogenically ground tire rubber as a filler in polyolefin blends [4]. They found that tire rubber particles precoated with ethylene-acrylic acid copolymer can be added to LLDPE or PP without diminishing the mechanical properties of rubber/plastic composites. Rouse reported the development of ambiently produced, high-surface area fine rubber powder of 80 mesh (178 microns) by wet grinding; even finer powders—known as ultra-fine—were also produced [5]. These highly resilient rubber powders can be used in many component parts of the tires as reinforcing fillers and processing aids. Rouse stated that 80 mesh tire rubber powder behaved more like a reinforcing carbon black than an inert filler due to the enhanced surface morphology of the rubber particles. Since about 1980, certain technological and commercial successes have allowed for increased recovery of rubber products. Better-known examples include the incineration of tires. Nevertheless, the recycling of rubber tires has been slow to develop due to differences in chemistry, morphology, and fabrication methods used in the production of tires [6]. Approximately 95 percent of the rubber fabricated in the US is thermoset with a three-dimensional network, so the cleavage of primary chemical bonds that is required to break this network
©2001 CRC Press LLC
P1: FUY pb035-08
March 1, 2001
20:48
Char Count= 0
becomes a major issue. Therefore, some used rubber parts have been recycled by chemical treatment to produce reclaim as a component for compounding operations. Rubber recovery is made more difficult by the fact that pneumatic tires, belts, and hoses consist of more than one rubber compound; they can also include steel wire, textile cord, or plastic fiber. The recycling of these products requires separating the steel and fibers from the rubber by multiple processing steps and the use of expensive equipment. Consequently, huge stockpiles of used tires tend to accumulate in the US, raising both health and environmental issues. Some of this stockpiled material is used as tire-derived fuel due to its high heat value by grinding tires into powder for re-use in rubberized asphalt and by shredding and re-use in automobile paddings. In the mid-1980s, an unconventional used-tire recovery process was developed as a joint effort between the Academy of Sciences in Moscow, Russia, and Berstorff Maschinenbau GmbH in Hannover, Germany; the latter subsequently became simply Berstorff GmbH [7-9]. The process utilized Berstorff’s modified co-rotating, intermeshing twin-screw extruder. The first commercial-size PT-90A (90-mm diameter) pulverizer for the pulverization of used-tire rubber has been in operation in Germany since 1991, as a part of a complete turn-key used-tire recycling line. This equipment uses a patented cooling system [10]. After steel and textile cords are removed, chips of coarsely shredded rubber about 0.25-in long are fed into a hopper of the pulverizer where they are subjected to compression and shear with rapid removal of frictional heat. The powder is discharged without a die and is classified prior to packaging. Rubber powder of about 60–80 mesh size (250–178 microns) is made in one pass through the extruder at a throughput of about 800 lb per hour. If finer powder is needed, the cooled, oversized rubber powder is returned to the pulverizer for a second pass. There is an increasing need for fine scrap tire rubber powder of 80 mesh (178 microns) and finer in order to create parts with smoother surfaces. Finer powders also improve the physical properties of rubber compounds and allow for faster mixing times when rubber powder is used as a partial substitute for virgin rubber. Few techniques, however, have been found for producing fine tire rubber powder economically.
CHARACTERIZING POWDERS At the Polymer Technology Center at Northwestern University, Khait and her co-workers used the S3 P process to develop used-tire rubber/recycled plastic composites combining tread rubber (TR) and plastics derived from postconsumer waste streams [11]. This work was carried out with the Berstorff pilot-scale PT-40A pulverizer (Charlotte, North Carolina) and the commercialscale PT-90A pulverizer (Germany).
©2001 CRC Press LLC
P1: FUY pb035-08
March 1, 2001
20:48
Char Count= 0
TABLE 8.1.
Sieve Analysis of Tread Rubber (TR) Powder. Mesh Size (microns)
Method of Size Reduction S3 P
Weight percent retained Weight percent retained
Ambient grinding
35 (500)
60 (250)
80 (178)
100 (150)
140 (104)
200 (74)
230 (63)
10.3
53.6
21.7
4.26
7.87
3.16
0.48
66.2
18.9
8.8
3
3.6
2
0.5
Particle size and particle size distribution of TR-40 of 40 mesh (600 microns) powder made via S3 P process and by ambient grinding for comparison are given in Table 8.1. Sieve analysis showed that powder made from TR using S3 P technology had a different particle size distribution than that of powder made by ambient grinding. The S3 P-made powder was finer and had a majority of particles of 60 mesh (250 microns), while only 18.9 percent of the particles produced by the conventional ambient process were of that size. Ambient-temperature acetone extraction data for TR and whole tire product (WTP) was reported [12]. The data in Table 8.2 suggest that partial devulcanization of the rubber took place during the S3 P process. Both TR and WTP samples had higher acetone values compared to both ambiently ground tread rubber (TR-40) of 40 mesh (420 microns) and whole tire product (WTP-30) of 30 mesh (600 microns). TR and WTP rubber were obtained from Baker Rubber, Inc., South Bend, Indiana. Table 8.3 summarizes additional elevated-temperature extraction data both for the tire rubber powder made by S3 P and for TR and WTP powder of 60–80 mesh (250–178 microns) made by conventional grinding at an ambient temperature [12]. The data in Table 8.3 show that the S3 P process produces rubber powder with greater elevated-temperature solubility than that of conventionally ground TABLE 8.2.
Ambient-Temperature Acetone Extraction Data of Tire Rubber Powder Made by Different Grinding Methods.
Source of Waste Rubber
Method of Size Reduction
Percent Soluble
Tread rubber (TR-40) Tread rubber (TR-40) Whole tire (WTP-30) Whole tire (WTP-30)
Ambient grinding S3 P Ambient grinding S3 P
11.56 14.60 11.60 12.60
©2001 CRC Press LLC
P1: FUY pb035-08
March 1, 2001
TABLE 8.3.
20:48
Char Count= 0
Elevated-Temperature Solubility of Tire Rubber Powder.
Source of Waste Rubber Tread rubber (TR-40) Tread rubber (TR-40) Whole tire (WTP-30) Whole tire (WTP-30)
Method of Size Reduction
Percent Soluble
Ambient grinding S3 P Ambient grinding S3 P
61.2 63.5 55.9 69.5
rubber. This finding was consistent with the ambient-temperature extraction test using acetone. The surface texture and unique particle shape of S3 P-made tire rubber powder was reported by Khait [11]. A Hitachi microscope model S-570 was used for observation of particle shape. A series of scanning electron micrographs (SEMs) at 40× and 250× magnification of WTP and TR rubber after pulverization, as well as micrographs of the same rubber feedstock made by ambient and cryogenic grinding, are shown in Figures 8.1a,b through 8.5a,b. The SEMs of WTP (Figure 8.4a,b) and TR (Figures 8.2a,b and 8.5a,b) rubber produced by conventional grinding showed irregular rod-like or diamond-like chunks. Particles of WTR and TR rubber powder after S3 P had a “cauliflowerlike,” open morphology with large surface area. Some particles of WTP rubber contained embedded fibers. In contrast, the conventionally ground rubber had
(a)
(b)
FIGURE 8.1 (a) SEM of used tread-tire rubber powder (TR) pulverized with PT-90 at 40×. (b) SEM of used tread-tire rubber powder (TR) pulverized with PT-90 at 250×.
©2001 CRC Press LLC
P1: FUY pb035-08
March 1, 2001
20:48
Char Count= 0
(a)
(b)
FIGURE 8.2 (a) SEM of used tread-tire rubber powder (TR) after ambient grinding at 40×. (b) SEM of used tread-tire rubber powder (TR) after ambient grinding at 250×.
(a)
(b)
FIGURE 8.3 (a) SEM of used whole-tire product (WTP) pulverized with PT-25 at 40×. (b) SEM of used whole-tire product (WTP) pulverized with PT-25 at 250×.
©2001 CRC Press LLC
P1: FUY pb035-08
March 1, 2001
20:48
(a)
Char Count= 0
(b)
FIGURE 8.4 (a) SEM of used whole-tire product (WTP) after ambient grinding at 40×. (b) SEM of whole-tire product (WTP) after ambient grinding at 250×.
(a)
(b)
FIGURE 8.5 (a) SEM of used tread-tire rubber (TR) after cryogenic grinding at 40×. (b) SEM of used tread-tire rubber (TR) after cryogenic grinding at 250×.
©2001 CRC Press LLC
P1: FUY pb035-08
March 1, 2001
20:48
Char Count= 0
particles with a flat, smooth fracture and a relatively small surface area. It is well-known in the rubber industry that rubber powder having a large surface area is more desirable in compounding operations because it allows for a shorter mixing cycle and improved surface appearance of rubber products. In powder technology, physical characteristics such as size, shape, and surface texture are very important in determining particle behavior. Each method of size reduction produces particles with a unique and characteristic surface texture. Using SEM and digital images, Deitering studied the effect of grinding techniques on the morphology of rubber particles produced by four different methods: S3 P process, cryogenic, wet grinding, and dry (ambient) grinding [13]. Each grinding technique was shown to produce particles with unique morphological features. In each of the four grinding methods, the tires were first debeaded and then reduced in size using granulators, producing chunks of tire rubber between 50 and 150 mm in size. The textile and steel cords were removed (and recycled), and the remaining material was further reduced to chips about 0.25-in long. The cryogenically ground rubber powder made from discarded tires was of 80 mesh (178 microns). In this process, tire chunks were cooled to approximately −186◦ C (−320◦ F) by liquid nitrogen to embrittle them prior to fracture. The frozen pieces were then impacted in a high-velocity mill to make crumb rubber. Dry and wet grinding processes were carried out at a temperature above the freezing point of tire rubber. The dry (ambient) grinding process tears the rubber pieces through the use of vigorous cutting paths, producing powders of 40 mesh (420 microns); powder was provided for this study by Baker Rubber, Inc., of South Bend, Indiana. Wet ground tire powder of 80 mesh (178 microns) was provided by Rouse Industries in Vicksburg, Mississippi. S3 P-made powder of 80 mesh (178 microns) was made with the commercial-scale twin-screw PT-90A Berstorff pulverizer in Germany. Two typical examples of rubber particles produced by cryogenic grinding are shown in Figure 8.6. The appearance of cryogenically ground rubber is different from that of ambiently ground rubber. The cryogenically ground rubber powder has flat particles and sharp corners almost like crushed stone. In this process, particles are formed by brittle fracture, resulting in the formation of sharp edges and relatively featureless surfaces. These two images show a series of ridges on the particle surface. This characteristic feature is called a twist hackle and is formed by lateral breakthrough of parallel crack surfaces. Wet ground tires particles appear to be made up of clusters of smaller particles, as shown in Figure 8.7. These particles appear very rough because of the clustering effect, yet the subunits have relatively smooth surfaces. Figure 8.8 shows two rubber particles produced by dry (ambient) grinding. The rubber ground at ambient temperature looks like a sponge. The surface texture shown is typical of rubber powders produced by this method and is characteristic of rubber tear. The morphology suggests that the process involves first stretching and then rupture of the rubber particles.
©2001 CRC Press LLC
P1: FUY pb035-08
March 1, 2001
20:48
Char Count= 0
(a)
(b) FIGURE 8.6 SEM images of two different particles of ground tire rubber after cryogenic grinding (300× magnification, 4.7 cm–100 m).
The failed rubber then relaxes into folds, producing small-sized texture features, approximately 5–10 mm in size. Dry-ground particles, similar to wetground ones, appear to be made up of clusters of smaller subunits. In this case, however, the subunits are larger in size and possess textural features.
©2001 CRC Press LLC
P1: FUY pb035-08
March 1, 2001
20:48
Char Count= 0
(a)
(b) FIGURE 8.7 SEM images of two different particles of ground tire rubber after wet grinding (300× magnification, 4.7 cm–100 m).
The SEM images in Figure 8.9 are typical examples of rubber particles produced by the S3 P method. The particles have highly developed surfaces with numerous protruding sections. The combination of high pressure and high shear strains during pulverization generate prominent folds and knobuals.
©2001 CRC Press LLC
P1: FUY pb035-08
March 1, 2001
20:48
Char Count= 0
(a)
(b) FIGURE 8.8 SEM images of two different particles of ground tire rubber after dry (ambient) grinding (300× magnification, 4.7 cm–100 m).
As Figures 8.6 through 8.9 show, each grinding technique produces powders with different morphologies. Most notably, the S3 P powders show a unique surface texture that would be beneficial for re-use of tire rubber powder in mixing operations with virgin rubber and appropriate additives.
©2001 CRC Press LLC
P1: FUY pb035-08
March 1, 2001
20:48
Char Count= 0
(a)
(b)
FIGURE 8.9 SEM images of two different particles of ground tire rubber pulverized with PT-90 (300× magnification, 4.7 cm–100 m).
It was reported that the S3 P-made powder derived from WTP and TR usedtire rubber was used as a partial replacement of virgin rubber in the softtread rubber formulation (Table 8.4) [12]. The “soft” tread grade compounds were prepared and tested by Midwest Custom Mixing, Inc., of North Miami, Oklahoma.
©2001 CRC Press LLC
P1: FUY pb035-08
March 1, 2001
TABLE 8.4.
20:48
Char Count= 0
Test Results of Soft Tread Grade Compounds Containing 5 Percent by Weight of Pulverized Used-Tire Rubber. Properties
Tensile strength MPa (psi) Elongation % 100% Modulus (psi) Hardness Shore A Tear Strength die C MPa (psi)
A
B
C
20.3 (2,950) 820 0.73 (106) 52
15.2 (2,210) 750 0.72 (105) 52
14.3 (2,080) 740 0.73 (106) 53
1.74 (253)
1.65 (240)
1.68 (243)
A = control (soft tread grade compound). B = control +5% by weight of pulverized tread rubber (TR). C = control +5% by weight of pulverized whole tire product (WTP).
An addition of five weight percent of pulverized tire rubber to the soft tread compound did not affect the 100 percent modulus (ASTM D 412), hardness shore A, or tear strength (die C) properties of the resultant material. However, the tensile strength of the test specimens containing pulverized tire rubber decreased from 2,947 psi for the control compound (made from virgin rubber) to 2,210 psi for the sample made with five weight percent of TR powder and to 2,080 psi for the sample made with five weight percent of WTP powder. Elongation at break also decreased from 820 percent for the control compound to 750 percent for the sample made with TR powder and to 740 percent for the sample made with WTP powder. As Table 8.4 shows, the tear strength of the compound containing TR was only slightly lower than that for the control compound (240 psi vs. 253 psi, respectively). A similar result was observed for the compound containing WTP rubber (243 psi vs. 253 psi, respectively). It has been shown that used-tire rubber made via S3 P also has a potential for use in rubberized asphalt [12]. The absolute viscosity of rubberized asphalt containing five weight percent S3 P-made powder was measured at 1,990 poise at 140◦ F, which was about three times higher than for asphalt cement (791 poise). The mixing of S3 P-made tire rubber powder with asphalt was performed by Koch Materials Co. of St. Paul, Minnesota. Their data suggest that significant reaction occurred when the pulverized tire rubber was added to the asphalt cement [12]. In addition to permitting a continuous conversion of used-tire rubber into powder of the desired particle size and particle size distribution, the S3 P process is more energy-efficient than existing grinding techniques. According to Lipp, the specific energy consumption during the S3 P process ranges between 0.3 and 0.9 kWh/kg of the product, depending upon the type of feedstock rubber used
©2001 CRC Press LLC
P1: FUY pb035-08
March 1, 2001
20:48
Char Count= 0
and particle size desired [14]. This continuous pulverization process represents a significant breakthrough in waste tire recovery, because it makes tire recycling easier and permits greater end-use of reconstituted fine tire rubber in valueadded products.
PHYSICAL PROPERTIES OF TIRE RUBBER POWDER/PLASTIC COMPOSITES The developmental composite materials combining both polymeric waste streams, such as used-tire rubber and post-consumer plastics via the S3 P process, have been described [15]. It has been demonstrated that physical properties of composites could be tailored to end-uses by varying the ratio between the rubber and plastic components of the mixtures. The properties of composites were also influenced by processing conditions, such as temperature and time during melt blending, that followed pulverization. The plastics component of the composites consisted of HDPE or LLDPE powder, derived from a post-consumer waste stream of unknown origin. The rubber component of the developmental composites included pulverized tread rubber (PTR) made with the pilot-scale PT-40A Berstorff pulverizer. HDPE or LLDPE S3 P-made powders have been melt-mixed with PTR using the small Haake Rheocord System 40 mixer. Two samples of conventionally ground tread rubber (GTR) such as GTR-30 and GTR-40 from Baker Rubber, Inc. of South Bend, Indiana, have been used as controls. Two composites at 60/40 and 40/60 rubber/plastic ratios were meltmixed at 180◦ C, 80 rpm for 15 minutes followed by compression molding at 180◦ C at seven tons of pressure. Key physical properties of die cut test specimens containing tire rubber and HDPE or LLDPE powders are presented in Tables 8.5 and 8.6, respectively. As shown in Tables 8.5 and 8.6, at a 60/40 ratio, both PTR/HDPE-P and PTR/LLDPE-P composites exhibited the best combination of tensile properties and tear strength as compared to those with composites utilizing conventionally ground GTR-30 or GTR-40 rubber. At a 40/60 ratio, however, composites made with GTR-40 with either HDPE-P or LLDPE-P had slightly higher tensile strength than the two other mixes. The physical properties of developmental composites made with as-received HDPE flakes instead of powder were reported [16]. Lower tensile strength and tear strength values were due to a low surface area of flat, angular flakes (Table 8.7). It has also been shown that the physical properties of rubber/plastic composites were significantly higher, as expected, when virgin HDPE was used in place of recycled feedstock in 60/40 rubber/plastic mixtures (Table 8.7). The properties of three-component developmental rubber/plastic composites containing ethylene vinyl acetate (EVA) copolymer UE-645 (Quantum Chemical) with 28 percent vinyl acetate as a binder were described [16]. The
©2001 CRC Press LLC
P1: FUY pb035-08
March 1, 2001
TABLE 8.5.
20:48
Char Count= 0
Physical Properties of Developmental Composites Based on Scrap Tire Rubber and Post-Consumer HDPE. Tensile Properties
a b c
Ratio/Composition
Ultimate MPa (psi)
60/40 PTRa /HDPE-Pb 60/40 GTRc -30/HDPE-P 60/40 GTR-40/HDPE-P 40/60 PTR/HDPE-P 40/60 GTR-30/HDPE-P 40/60 GTR-40/HDPE-P
5.1 (750) 4.2 (620) 5.2 (760) 7.5 (1,100) 8.3 (1,200) 8.8 (1,300)
Elong % 10 8 11 5 5.7 6
Tear Strength kN/m (lb/in) 29.8 (170) 24.5 (140) 26.3 (150) 42.7 (244) 39.6 (226) 42.2 (241)
Hardness Shore D 45 46 43 54 54 53
PTR = pulverized tire rubber. P = pulverized plastic. GTR = ground tire rubber.
TABLE 8.6.
Physical Properties of Developmental Composites Based on Scrap Tire Rubber and Post-Consumer LLDPE. Tensile Properties
Ratio/Composition
Ultimate MPa (psi)
60/40 PTRa /LLDPE-Pb 60/40 GTR-30/LLDPE-P 60/40 GTR-40/LLDPE-P 40/60 PTR/LLDPE-P 40/60 GTRc -30/LLDPE-P 40/60 GTR-40/LLDPE-P
4.3 (630) 3.6 (530) 4.1 (600) 7.1 (1,000) 6.8 (990) 9.2 (1,300)
PTR = pulverized tire rubber. P = pulverized plastic. c GTR = ground tire rubber. a b
©2001 CRC Press LLC
Elong % 19 15 15 11 13 6
Tear Strength kN/m (lb/in) 23.6 (135) 22.8 (130) 22.8 (130) 30.8 (176) 30.5 (197) 41.3 (236)
Hardness Shore D 37 35 36 46 46 54
P1: FUY pb035-08
March 1, 2001
20:48
Char Count= 0
TABLE 8.7. Physical Properties of Tire Powder/HDPE Mixtures at 60/40 Ratio Using S3 P-made HDPE Powder and Comparison with Recycled HDPE Flakes and Virgin HDPE Pellets.
Tensile Properties
Composition
Ultimate MPa (psi)
PTRa /HDPE-Pb
Elong %
5.2 (760) 4.6 (670) 7.7 (1,120)
PTR/HDPE-Fc PTR/HDPE-Vd
12 8 37
Tear Strength kN/m (lb/in)
Hardness Shore D
31.0 (177) 28.9 (165) 55.3 (316)
45 43 49
PTR = pulverized tire rubber. P = powder. F = flakes. d V = virgin pellets. a b c
physical properties of PTR/HDPE-F blends with 10 weight percent of EVA were compared with those containing ambiently ground GTR-40 rubber from Baker Rubber with post-consumer HDPE in a flake form (Table 8.8). As shown in Table 8.8, the tear strength of PTR/HDPE-F/EVA composites was higher than that of an analogous mixture containing ambiently ground rubber due to the larger surface area of S3 P-made rubber powders. Sieve analysis results for American-made pulverized tread tire rubber (APTR) made with the commercial-scale PT-90A Berstorff pulverizer in Germany are shown in Table 8.9. The majority of powder (38.2 percent by weight) was of 60 mesh followed by 80 mesh (20.5 percent by weight) and 140 mesh (11.8 percent by weight); a small amount of coarser and finer powder was also present. Overall, APTR Physical Properties of 50/40/10 Tire Rubber/HDPE-F/EVA UE-645 Composites Using S3 P-made Rubber Powder and Comparison with Ambiently Ground Rubber.
TABLE 8.8.
Tensile Properties
Rubber Type PTRa GTR-40b a b
Ultimate MPa (psi) 6.8 (990) 6.8 (990)
PTR = Pulverized Tire Rubber. GTR = Ground Tire Rubber.
©2001 CRC Press LLC
Elong % 28 20
Tear Strength kN/m (lb/in) 42.4 (242) 39.2 (224)
Hardness Shore D 41 45
P1: FUY pb035-08
March 1, 2001
TABLE 8.9.
20:48
Char Count= 0
Sieve Analysis for American-Made Pulverized Tread Rubber (APTR). Size Mesh
Microns
Weight Percent Retained
35 60 80 100 140 200 230 325
500 250 178 150 104 74 63 45
7.5 36.7 19.7 6.3 11.3 8.5 1.9 7.9
powder made with the commercial-scale PT-90 pulverizer was finer than that of PTR made with the PT-40A pilot scale Berstorff twin-screw pulverizer at their US facility. Table 8.10 shows the physical properties of developmental rubber/plastic composites utilizing tread rubber pulverized with the commercial-scale PT-90A Berstorff pulverizer and recycled polyolefins with EVA UE-645 (28 percent vinyl acetate) as a binder. Post-consumer polyolefins used in those composites include LLDPE, HDPE, and LDPE. These developmental composites have been made by two different techniques. The first involved melt-mixing of all three components in a powder form Physical Properties of Developmental Composites Based on Tire Rubber, Post-Consumer Polyolefins, and a Binder at 50/40/10 Ratio.
TABLE 8.10.
Tensile Strength
a b
Ingredients
Ultimate MPa (psi)
APTRa /LLDPE-Pb /EVA UE645 APTR/LLDPE-P/EVA UE645-P APTR/HDPE-P/EVA UE645 APTR/HDPE-P/EVA UE645-P APTR/LDPE-P/EVA UE645 APTR/LDPE-P/EVA UE645-P
5.9 (860) 6.1 (890) 8.8 (1,280) 9.0 (1,310) 5.8 (850) 6.7 (980)
APTR = American-made pulverized tread rubber. P = pulverized.
©2001 CRC Press LLC
Elong % 26 34 33 28 100 168
Tear Strength kN/m (lb/in) 41.7 (238) 42.5 (243) 69.1 (395) 67.4 (385) 51.5 (294) 53.7 (307)
Hardness Shore D 40 40 50 50 34 35
P1: FUY pb035-08
March 1, 2001
20:48
Char Count= 0
in a small Brabender mixer with cam blades (designated as APTR/polyolefin-P), and the second involved co-pulverization of rubber with polyolefins (designated as APTR/polyolefin-P/EVA-P) using the PT-25 Berstorff pulverizer. All pulverized samples were melt mixed at 180◦ C at 80 rpm for five minutes followed by compression molding at 220◦ C. As can be seen from Table 8.10, there is a consistent increase in both tensile strength and tear strength of the composites made by the second technique, which utilized solid-state co-pulverization of tire rubber with recycled plastics. These results suggest enhanced compatibility of the blend components due to intimate mixing during the S3 P process. In addition to used-tire rubber, other waste rubber products, including automotive hoses and seals made from nitrile rubber and ethylene-propylene-diene (EPDM) rubber, have been successfully converted into powders via S3 P (Figure 8.10). Particle size and particle size distribution were similar to that of pulverized used-tire rubber discussed earlier in this chapter.
FIGURE 8.10 Waste rubber tubes and hoses, S3 P-made powders, and a mat produced by compression molding.
©2001 CRC Press LLC
P1: FUY pb035-08
March 1, 2001
20:48
Char Count= 0
EMERGING TECHNOLOGIES IN TIRE-RUBBER RECYCLING Subsequent advances in cryogenic grinding allowed production of highvalue fine crumb as compared to older ambient grinding techniques. Kohler and O’Neill reported that the finer the crumb, the larger the amount that can be re-used without causing the new product’s properties to deteriorate [17]. If the surface of crumb rubber is modified, however, an even larger percentage of scrap can be incorporated into the compound. They also reported that the crumb size tested in natural rubber and styrene-butadiene rubber compounds of 10 to 80 mesh (2,000 to 178 microns) had little effect on tensile strength, but elongation and Mooney viscosity improved with finer crumb. They noted, however, that only 40 mesh (420 microns) rubber can be obtained economically via cryogenic grinding. Kohler and O’Neill found that crumb produced at ambient temperature has a rough surface, while cryogenically ground crumb had flat planes and straight line fractures. They concluded that for devulcanizing, both types of crumb rubber behaved similarly, and thus the choice of grinding method would depend primarily on the economics of the process. Several new technologies have emerged for surface modification of rubber. Some involve the devulcanization of rubber while others coat the rubber surface or oxidize it to increase adhesion to virgin rubber. While the production of surface-modified crumb is greater than that of unmodified crumb, it is still limited. The so-called “De-Vulc process” claims that rubber devulcanized by this method can be revulcanized without additional vulcanizing chemicals [18]. In yet another process, the crumb is claimed to be devulcanized by adding between two and six parts of a reactant called “De-Linc” by using a two-roll mill or an internal mixer [19]. The devulcanized rubber can then be blended with virgin rubber followed by molding, extrusion, or calendering. Wolfson and Nikol’skii reviewed elastic-strain-assisted, high-temperature grinding of vulcanized rubber [20]. Unlike thermoplastics, rubber did not exhibit phase transitions at an elevated temperature; other features, such as non-uniform density of crosslinks and the presence of fillers, could thus affect the grinding process. This study indicated that with a temperature increase, the specific energy consumption decreases because lower shear stress was required to break rubber particles. It was found that the specific area of rubber powder made by high-temperature, elastic-strain-assisted grinding was significantly higher than that of cryogenically ground rubber. Wolfson and Nikol’skii also stated that the elastic-strain grinding of rubber is realized with specially designed extruders (screw diameter up to 120 mm) with a throughput of 1,200 kg/hr. They estimated that a specific energy consumption was 160 kWh/ton of powder with an average particle size of 20–50 mesh (300–800 microns). According to them, a special device developed by Rospolymer in Moscow performed repeated elastic-deformation grinding of separated coarse rubber that yielded three fractions of powder: 250, 250–500, and 500–1,000 microns (60, 60–35, and 35–18 mesh). The equipment for tire-grinding based on a single-screw
©2001 CRC Press LLC
P1: FUY pb035-08
March 1, 2001
20:48
Char Count= 0
extruder with a throughput from 40 kg/h to 250 kg/h was manufactured by the Stimul Co. of Moscow. Wolfson and Nikol’skii have indicated that several commercial extrusion operations using modified single-screw devices to recover used-tire rubber are located in Russia, Uzbekistan, Ukraine, China, Bulgaria, and the US. Dorfman et al. studied the mechanism of shear stress-assisted grinding and concluded that, during deformation, rubber particles break in a step-wise manner developing various cracks [21]. Grinding of denser cross-linked rubber resulted in smaller particle size powder with a bimodal particle size distribution. These researchers also found that an addition of 5–15 weight percent of mineral fillers assisted in both producing smaller particle size powder and reducing specific energy consumption. Because mineral fillers are abrasive, however, they caused the increased wear of the screw elements. Comminution of tire rubber with thermoplastics was reported by Shutov et al. [22]. It was shown that an addition of 20–25 weight-percent of LDPE decreased the size of rubber powder and changed the shape and surface properties of the particles. The throughput of the process, however, was lower due to a greater heat generation in the grinding zone. Other devices using shear-assisted grinding such as a disk extruder and ozoneinitiated mechanical destructors have appeared. A mechanical destructor was invented by Danschikov et al. in which the recycled reinforced rubber is placed in the chamber containing ozone at about 0.5 percent under pressure of at least 0.5 kg/cm2 [23]. The deformation process allows the rubber to be broken at ambient temperature in less than 15 minutes. This process causes separation of metal and plastic cord from tire rubber with minimal energy consumption. The method is based on cleavage of carbon-carbon bonds in the presence of ozone. Kelly et al. reported another method of re-use of thermoset rubber (which is a variation of pyrolysis), based on high shear that is achieved in the rotor-disperser called “Milst,” which is operational in the US. Models of a later version called “Rokel” are located in Russia, Finland, and the US [24]. Arinstein et al. reported data on multiple disintegration of solids under intensive stress action, such as compression and shear (ISAC & S) using a family of devices generically named “Rotor Dispersers” [25]. This disperser consists of an immobilized cylinder that is matched with an inner coaxial cylindrical rotor connected to a drive. In this apparatus, a ring-shaped layer of a polymer is formed first. The layer is then exposed to compression and heat, and the formation of powder follows. These researchers claimed that the resultant powder is porous with a specific area in the range of 0.5–10 m2 /g. A theoretical model of the ISAC & S has been proposed, based on the assumption that under the combination of high temperature, compression, and shear, the polymer accumulates micro-defects; when the concentration of defects reaches some critical value, the powder is formed. For this powder, the particle shape resembles
©2001 CRC Press LLC
P1: FUY pb035-08
March 1, 2001
20:48
Char Count= 0
a sponge-like structure, unlike the smooth fractured surface of cryogenically ground rubber. Because recycling of used tires presents a tremendous problem in waste disposal worldwide, much attention has been focused on the development of new technologies to recover tire rubber for re-use. A novel patented process for the devulcanization of waste rubber has been invented by Isaev, based on the use of ultrasonic waves combined with pressure and heat [26]. The devulcanized rubber can be revulcanized and re-used in various rubber compounds as a virgin rubber substitute. Luo and Isaev reported rubber/plastic blends based on ultrasonically devulcanized ground tire rubber and polypropylene, where best results were achieved with a maleic-anhydride grafted PP compatibilizer and a phenolic-resin cure system [27]. They found that Young’s modulus and the tensile strength of dynamically revulcanized tire rubber/PP blends improved over the same properties for modified PP; the rheological behavior of blends, however, did not depend on the type of PP used. McKirahan et al. investigated thermoplastic composites of recycled HDPE and used-tire rubber [28]. They found that an addition of recycled rubber decreased tensile strength, elongation, and hardness of recycled HDPE; the ductility of composites was improved when finer tire powder was used. Li and Liu studied rheological properties of these composites [29]. According to them, the viscosity of the higher rubber containing composites decreased at a high shear rate; this was attributed to an increased interaction between recycled HDPE and the tire rubber powder. Gu and Min studied the effect of chemically modified ground tire rubber on properties of blends with polyamides [30]. Their research shows that the oxidation of the rubber surface by plasma positively affected mechanical properties of rubber/polyamide blends due to better adhesion between phases. In another study, Kohler reported improved cryogenic grinding with reduced consumption of liquid nitrogen [31]. This development is significant because nitrogen consumption accounts for 30–50% of the cost of fine grinding. Kohler noted that Praxair, Inc. in Danbury, Connecticut, has used the proprietary Very Fine Grinding SystemTM (VFGS) with significant modifications to both mill design and processing parameters. According to Kohler, more than 60 percent of VFGS production is finer than 80 mesh (178 microns) at a rate of more than 4,000 lb/hr (1,820 kg/hr) and a nitrogen consumption of about 1.5 lb of nitrogen per pound of rubber feed. Another new developmental system, the Ultra Fine Grinding SystemTM , uses increased impact velocities of about three times those of conventional mills and is expected to produce more than 2,000 pounds (900 kg) of at least 80 percent minus 80 mesh (177 microns) and as much as 90 percent minus 100 mesh (140 microns) an hour using only 0.75 pounds of nitrogen per pound of feed.
©2001 CRC Press LLC
P1: FUY pb035-08
March 1, 2001
20:48
Char Count= 0
McDonel et al. discussed a proprietary ground rubber additive (GRA) that is used for the enhancement of ground tire rubber crumb [32]. The GRA is blended with rubber crumb in a Banbury-type mixer, followed by milling and sheeting as a slab. This slab is then added to the tire rubber compound at about 10 percent by weight with little loss in physical properties. Arastoopour et al. described a process for recycling rubber, elastomers, and thermoset materials using a small-scale, conical counter-rotating screw apparatus in which the material is pulverized without change in the chemical composition followed by compression molding [33]. Goncharuk et al. reported the mechanical properties of rubber-filled plastics based on LDPE and rubber powders prepared from used tires by various techniques [34]. This group learned that the tensile strength and elongation at break of those composites increased for the powder with larger surface area. Particle shape of rubber crumb did not affect mechanical properties of rubber-filled plastics. SUMMARY Improved recycling methods for used-tire rubber are needed if the environment is to be protected. Huge quantities of used tires and other rubber scrap are still being landfilled. In the US at the end of the 1990s, an estimated 270 million tires a year were discarded in addition to about 400 million used tires that were stockpiled. It is also estimated that about 350 million pounds of rubber are accumulated from post-industrial sources including trim, defective parts, and runners. Additional waste accrues from post-consumer discards of various rubber products. Although much effort has been directed toward rubber recycling and the development of new uses for scrap tires, there is still a lack of economically viable methods of waste rubber recovery. At the close of the 20th century, the major techniques of producing recycled rubber were limited either to reclamation or to ambient or cryogenic grinding. However, only relatively coarse rubber of 10 to 40 mesh (2,000 to 420 microns) can be made using inexpensive grinding methods. Consequently, recycled used-tire rubber is limited primarily to re-use as a filler. In contrast, the novel S3 P process permits continuous, one-step production of tire rubber powder finer than 40 mesh (420 microns) with large surface area and a unique “cauliflower” morphology that is advantageous for further mixing with virgin rubber, additives, or plastic during the compounding operation. The results obtained at the Polymer Technology Center at Northwestern University suggest that tire rubber powder/plastic composites made via the S3 P process exhibit mechanical properties and processability that are suitable for making a variety of consumer goods such as mats and hoses. This creation of value-added products from recycled waste rubber can relieve the demand on ever-diminishing landfill capacity, representing a substantial saving of both money and energy.
©2001 CRC Press LLC
P1: FUY pb035-08
March 1, 2001
20:48
Char Count= 0
REFERENCES Note: Some names of people (such as Enikolopyan, Yerina, Nepomnyaschy, and Akopyan) have variant spellings in English translation. While the chapter text uses a single spelling, the variations are preserved in this reference list. 1. Rader, C. P. and Lemieux, M. A. “The Recycle of Plastics and Rubber—a Contrast,” Rubber World, 216(2):24–30 (1997). 2. Stark Jr., F. J. and Leighton, A. “The Conversion of Scrap Rubber into a New Compounding Tool,” Rubber World, 188(5):36–51 (1983). 3. Bauman, B. D. “Surface-Modified Rubber Particles in Polyurethane,” Rubber Plastic News, 23 (July 4, 1994):35–36. 4. Oliphant, K. and Baker, W. E. “The Use of Cryogenically Ground Rubber Tires as a Filler in Polyolefin Blends,” Polymer Engineering and Science, 33(3):166–174 (1993). 5. Rouse, M. W. “Development and Application of Superfine Tire Powders for Rubber Compounding,” Rubber World, 206(3):25–40 (1992). 6. United States Environmental Protection Agency Report, “Markets for Scrap Tires,” EPA/530SW-90-074A (1991). 7. Enikolopov, N. S., Wolfson, S. A., Nepomnjaschtschie, A. J., Nikol’skii, W. G., Teleschow, W. A., Filmakowa, L. A., Brinkmann, H., Pantzer, E., and Uhland, E. US Patent 4,607,797 (1986). 8. Capelle, G. Proceedings of the Rubber Division Meeting of the American Chemical Society, Detroit (1991). 9. Enikolopov, N. S., Nepomnyaschy, A. I., Filmakova, L. A., Krasnokutsky, V. P., Kurakin, L. I., Akopian, E. L., Markarian, K. A., Negmatov, S. S., Martkarimov, S. K., Polivanov, Y. A., Sherstnev, P. P., and Pavlev, V. B. US Patent 4,607,796 (1986). 10. Mayer, D. and Freist, B. US Patent 5,273,419 (1993). 11. Khait, K. “New Used-Tire Recovery Process for Value-Added Products,” Proceedings of the Rubber Division Meeting of the American Chemical Society, Chicago (1994). 12. Khait, K. Grant Report TM-50: “Product Development for Scrap Tire Rubber,” Department of Commerce and Community Affairs, Bureau of Energy and Recycling, Springfield, Illinois (1996). 13. Dietering, M. L. “Particle Characterization of Recycled Polymeric Powders,” Master’s thesis, Northwestern University (1994). 14. Lipp, R. Proceedings of Recycle Exposition ’91—Fourth International Forum, Davos, Switzerland (1991). 15. Khait, K. “Application Development for Used-Tire Rubber Recovered by a Novel Solid-State Shear Extrusion Process,” Proceedings of the Rubber Division Meeting of the American Chemical Society, Cleveland, Ohio (1995). 16. Khait, K. “New Solid-State Shear Extrusion Process for Used-Tire Rubber Recovery,” Proceedings of the Rubber Division Meeting of the American Chemical Society, Louisville, Kentucky (1996). 17. Kohler, R. and O’Neill, J. “New Technology for the Devulcanization of Sulfur-Cured Scrap Elastomers,” Rubber World, 216(2):32–36 (1997). 18. Hon, K. K. and Siesseger, F. “Devulcanization of Sulfur-Cross-Linked Rubber by the DeLinkTM Recycling Process,” Proceedings of the Rubber Division Meeting of the American Chemical Society, Cleveland, Ohio (1995).
©2001 CRC Press LLC
P1: FUY pb035-08
March 1, 2001
20:48
Char Count= 0
19. Sekhar, T. S. and Kormer, V. A. European Patent Application EP 0690091 (1995). 20. Wolfson, S. A. and Nikol’skii, V. G. “Strain-Assisted Fracture and Grinding of Solid Polymeric Materials: Powder Technologies,” Polymer Science USSR—Series B, 36:861–874 (1994). 21. Dorfman, I. Ya., Kryrchuv, A. N., Prut, E.V., and Enikolopyan, N. S. “Plastic Flow Instability in the Solid Phase Polymer Extrusion Process,” Dokladi Academii Nauk SSSR, 278:141 (1984). 22. Shutov, F., Ivanov, G., Arastoopour, H., and Wolfson, S. A. “New Principle of Plastic Waste Recycling: Solid-State Shear Extrusion,” Proceedings of the Fall Meeting of the American Chemical Society, 67:112 (1992). 23. Danschikov, E. V., Luchnik, I. N., Ryazonov, A. V., and Chinko, S. V. US Patent 5,492,657 (1996). 24. Kelly, K. F., Nikol’skii, V. G., Bolyberdin, V. N., Benham, N., Morris, I., and Kelly, B. M. “Improved Method for Re-utilizing Rubber Materials from Factory Scrap and their Consequent Remolding Characteristics,” Proceedings of the Rubber Division Meeting of the American Chemical Society, Nashville, Tennessee (1997). 25. Arinstein, A. E., Balyberdine, V. N., Kelly, B. M., Kelly, K.F., and Nikol’skii, V. G. “HighTemperature Shear-Induced Multiple Cracking and Grinding of Polymeric Materials,” Proceedings of the Rubber Division Meeting of the American Chemical Society, Nashville, Tennessee (1998). 26. Isaev, A. I. US Patent 5,258,413 (1993). 27. Luo, T. and Isaev, A. I. “Rubber/Plastic Blends Based on Devulcanized Ground Tire Rubber,” Journal of Elastomers and Plastics, 30:133 (1998). 28. McKirahan, J., Liu, P., and Brillhart, M. “Thermoplastic Composites of Recycled HDPE and Recycled Tire Rubber Particles,” Proceedings of the Annual Technical Conference of the Society of Plastics Engineers, ANTEC ’96, p. 310 (1996). 29. Li, Y. and Liu, P. “Rheological Behavior of Composites of Recycled HDPE and Recycled Tire Rubber Particles,” Proceedings of the Annual Technical Conference of the Society of Plastics Engineers, ANTEC ’97, p. 1082 (1997). 30. Gu, C. J. and Min, K. “ Polymides Reinforced with Ground Tire Particles,” Proceedings of the Annual Technical Conference of the Society of Plastics Engineers, ANTEC ’97, p. 3146 (1997). 31. Kohler, R. “Advances in Cryogenic Fine Grinding,” Proceedings of the Rubber Division Meeting of the American Chemical Society, Cleveland, Ohio (1997). 32. McDonel, T., Fusco, J., and Wheeler, M. “Ground Rubber Additive,” Proceedings of the Rubber Division Meeting of the American Chemical Society, Cleveland, Ohio (1997). 33. Arastoopour, H., Schoekl, D. A., Bernstein, B., and Bilgili, E. US Patent 5,904,885 (1999). 34. Goncharuk, G. P., Knunyants, M. I., Kyuchkov, A. N., and Obolonkova, E. S. “Effect of the Specific Area and the Shape of Rubber Crumb on the Mechanical Properties of Rubber-Filled Plastics,” Polymeric Science USSR—Series B, 40(5–6):166–169 (1998).
©2001 CRC Press LLC
P1: SYV/FYX pb035-ep
P2: SYV/ABE
February 27, 2001
QC: SYV/ABE 12:11
T1: SYV
Char Count= 0
Epilogue: Toward the Future
the S3 P process was developed as a new size-reduction method for producing polymeric powders. Over several years, however, as knowledge of the process has broadened, it has been discovered that S3 P offers surprising possibilities for becoming a new solid-state polymer processing technology. In addition to the powder-production capability, it has been demonstrated that the process allows for blending incompatible polymers via mechanochemistry, mixing polymers of unmatched viscosity, incorporating and homogenizing additives, and recycling unsorted polymeric waste and rubber recovery (including tire rubber). All of these processes can occur in the melt, but they are deficient or may require compatibilizers. With S3 P, these processes occur in the solid state, resulting in powders with properties superior to those of melt-processed materials. The use of polymeric powders has increased rapidly over recent years due to their unique combination of processability and performance. Their use is also associated with the rapid growth of the dynamic field of rotational molding, which utilizes powder as a feedstock. Powders made with the S3 P technology would enhance the manufacturing of parts due to their unique morphology and surface characteristics. Because S3 P deals with the behavior of solids, sometimes referred to as particulates or powders, it is important to study the relationship between powder properties, such as particle size, shape, and bulk density and their motions during processing. Further research toward a better understanding of the mechanochemistry that occurs during the S3 P process would open new avenues for modifying physical properties of existing polymers. These polymers with tailored properties would be made more quickly and more economically in comparison with the traditional synthesis of new polymers. Continued modifications to pulverization equipment
O
RIGINALLY,
©2001 CRC Press LLC
P1: SYV/FYX pb035-ep
P2: SYV/ABE
February 27, 2001
QC: SYV/ABE 12:11
T1: SYV
Char Count= 0
would allow for a one-step production of powder from difficult-to-grind resins such as ultra-high molecular weight poly(ethylene), polyvinylidene fluoride, and other specialty polymers. Future research efforts might involve modeling of the pulverization process to assist in establishing optimal parameters and quantitative relationships between equipment design and the powder formation mechanism. Modeling of the S3 P process would require a multidisciplinary team of physicists, mathematicians, chemists, and engineers using statistical design of experiments. Computerized data monitoring, collection, and retrieval would assist in the optimization of the S3 P process by providing modern techniques of data analysis. Continued basic and applied research will involve further studies on the polymers’ behavior and on the mixing of polymer blends with unmatched viscosity with the S3 P process. Efficient mixing in the solid state will permit the development of new polymer blends with stable morphologies while eliminating phase inversion associated with conventional melt-mixing techniques. The S3 P process opens an opportunity to disperse various additives and pigments in plastics in one step in the solid-state without melting, thereby preserving physical properties of the resins as compared with conventional melt extrusion and eliminating additional operations associated with grinding pellets into a powder. Conventional compounding equipment is being used extensively for melt blending, reactive processing, and mixing with additives. All of these processes deal with melting, flow, reactions, and dispersive and distributive mixing. By contrast, the pulverization process deals with the behavior of solids, associated with flow patterns, compaction, agglomeration, heat transfer, and others. Future research would include better understanding of the above-mentioned phenomena and provide advanced knowledge for modifications of existing pulverization equipment. The S3 P process will make a valuable contribution to eco-efficiency of consumer waste recovery by conversion of unsorted discarded plastics and rubber into value-added materials. Recycling of plastics and rubber (including used tires) will continue to be an integral part of waste management in the new millennium to minimize the impact of discarded polymeric waste on the global environment. The recovery and recycling of post-consumer plastic and rubber waste will continue due to the increased cost of fossil fuels, public support for protecting the environment, and government regulations toward an increase of recovering value from polymeric materials. Disposal of post-consumer plastic waste in landfills has become a major concern because of high cost, governmental requirements, and public pressure for a greener environment. The growing consumption of plastics per capita in industrialized countries is about five times that in non-industrialized ones, primarily due to the use of plastics in disposable products. Rapidly developing countries like China, India, and Taiwan, however, will also contribute to this explosion in plastics use and thus inevitably to associated disposal problems.
©2001 CRC Press LLC
P1: SYV/FYX pb035-ep
P2: SYV/ABE
February 27, 2001
QC: SYV/ABE 12:11
T1: SYV
Char Count= 0
Two obstacles most frequently encountered in recycling plastics are their mutual incompatibility and the presence of contaminants. A major barrier to recovering post-consumer waste is the high cost of sorting. Although a variety of recycling technologies exist, they all are both costly and inefficient, and thus their usefulness is limited. The non-conventional S3 P process holds dramatic potential for the recycling of unsorted post-consumer waste into value-added materials. As an environmentally friendly technology, S3 P can handle multicolored feedstocks having contaminants such as labels, ink, aluminum foil, and adhesives, among others. In coming years, S3 P-reconstituted plastics should be used increasingly in the manufacture of cars, business machines, photocopying equipment, and numerous other such fields. Similar issues arise with regard to the disposal of scrap tires in the more industrialized societies. In Germany, regulations are already in place that require tire makers to reacquire used tires and to convert them into other automobile parts, to grind them for re-use in rubber goods, or to use them in making rubberized asphalt. For the last two applications, however, it is desirable to produce fine grades of rubber powder that are difficult to make by existing grinding methods. The S3 P process is ideally suited for recyling used-tire rubber because the morphology of the powder it produces is advantageous both for further compounding with virgin rubber in new tire production and for mixing with plastics to make molded or extruded goods. The S3 P process can thus facilitate tire recycling by increasing end-uses of reconstituted tire rubber in value-added products. In the future, one should be able to write another book entitled The Science and Technology of the Solid-State Shear Pulverization Process via Mechanochemistry after greater knowledge has been obtained regarding this complex process. I believe that further exploration of the unique solid-state pulverization process can open new scientific and technological frontiers in predicting the behavior of polymers and can also lead to the creation of new materials with enhanced properties and performance.
KLEMENTINA KHAIT
©2001 CRC Press LLC
P1: SYV/FYX PB035-glo
P2: SYV/ABE March 2, 2001
QC: SYV/ABE 14:36
T1: SYV
Char Count= 0
Glossary
ABS: Acrylonitrile-butadiene-styrene terpolymer APTR: American pulverized tire rubber DSC: Differential scanning calorimetry EDG: Elastic deformation grinding EPDM: Ethylene-propylene-diene monomer ESP: Elastic strain powdering ESR: Electron spin resonance EVA: Ethylene-vinyl acetate copolymer GPC: Gel permeation chromatography HDPE: High-density poly(ethylene) HIPS: High-impact polystyrene IR: Infrared LDPE: Low-density poly(ethylene) LLDPE: Linear low-density poly(ethylene) MA-PP: Maleic anhydride-modified polypropylene MFR: Melt flow rate MI: Melt index (early equivalent of MFR) MIPS: Medium-impact polystyrene MSC: Material Sciences Corporation MSW: Municipal solid waste NBR: Nitrile butadiene rubber NMR: Nuclear magnetic resonance NU: Northwestern University PBT: Poly(butylene terephthalate) PC: Polycarbonate PET: Poly(ethylene terephthalate)
©2001 CRC Press LLC
P1: SYV/FYX PB035-glo
P2: SYV/ABE March 2, 2001
QC: SYV/ABE 14:36
T1: SYV
Char Count= 0
Ph-PP: Phenolic-modified polypropylene PMMA: Polymethylmethacrylate POP: Polyolefin plastomer PP: Polypropylene PS: Polystyrene PSP: Pressure shear pulverization PTC: Polymer Techology Center PTFE: Poly(tetrafluoroethylene) PTR: Pulverized tire rubber PVC: Poly(vinyl chloride) PVDF: Poly(vinylidene fluoride) SAN: Styrene acrylonitrile copolymer SEM: Scanning electron microscopy SMA: Styrene maleic anhydride copolymer SSSE: Solid-state shear extrusion SSSP (S3 P): Solid-state shear pulverization TR: Tread rubber UHMWPE: Ultra-high molecular weight poly(ethylene) WTP: Whole tire rubber
©2001 CRC Press LLC