Editor-in-Chief Prof. em. Dr. Otto Hutzinger University of Bayreuth c/o Bad Ischl Office Grenzweg 22 5351 Aigen-Vogelhub, Austria E-mail:
[email protected]
Advisory Board Dr. T.A.T. Aboul-Kassim
Prof. Dr. D. Mackay
Department of Civil Construction and Environmental Engineering, College of Engineering, Oregan State University, 202 Apperson Hall, Corvallis, OR 97331, USA
Department of Chemical Engineering and Applied Chemistry University of Toronto Toronto, Ontario, Canada M5S 1A4
Dr. D. Barcelo Environment Chemistry IIQAB-CSIC Jordi Girona, 18 08034 Barcelona, Spain
Prof. Dr. P. Fabian Chair of Bioclimatology and Air Pollution Research Technical University Munich Hohenbacherstraße 22 85354 Freising-Weihenstephan, Germany
Prof. Dr. A.H. Neilson Swedish Environmental Research Institute P.O.Box 21060 10031 Stockholm, Sweden E-mail:
[email protected]
Prof. Dr. J. Paasivirta Department of Chemistry University of Jyväskylä Survontie 9 P.O.Box 35 40351 Jyväskylä, Finland
Dr. H. Fiedler
Prof. Dr. Dr. H. Parlar
Scientific Affairs Office UNEP Chemicals 11–13, chemin des Anémones 1219 Châteleine (GE), Switzerland E-mail:
[email protected]
Institute of Food Technology and Analytical Chemistry Technical University Munich 85350 Freising-Weihenstephan, Germany
Prof. Dr. H. Frank Chair of Environmental Chemistry and Ecotoxicology University of Bayreuth Postfach 10 12 51 95440 Bayreuth, Germany
Department of Veterinary Physiology and Pharmacology College of Veterinary Medicine Texas A & M University College Station, TX 77843-4466, USA E-mail:
[email protected]
Prof. Dr. M. A. K. Khalil
Prof. P.J. Wangersky
Department of Physics Portland State University Science Building II, Room 410 P.O. Box 751 Portland, Oregon 97207-0751, USA E-mail:
[email protected]
University of Victoria Centre for Earth and Ocean Research P.O.Box 1700 Victoria, BC, V8W 3P6, Canada E-mail:
[email protected]
Prof. Dr. S.H. Safe
Preface
Environmental Chemistry is a relatively young science. Interest in this subject, however, is growing very rapidly and, although no agreement has been reached as yet about the exact content and limits of this interdisciplinary discipline, there appears to be increasing interest in seeing environmental topics which are based on chemistry embodied in this subject. One of the first objectives of Environmental Chemistry must be the study of the environment and of natural chemical processes which occur in the environment. A major purpose of this series on Environmental Chemistry, therefore, is to present a reasonably uniform view of various aspects of the chemistry of the environment and chemical reactions occurring in the environment. The industrial activities of man have given a new dimension to Environmental Chemistry. We have now synthesized and described over five million chemical compounds and chemical industry produces about hundred and fifty million tons of synthetic chemicals annually. We ship billions of tons of oil per year and through mining operations and other geophysical modifications, large quantities of inorganic and organic materials are released from their natural deposits. Cities and metropolitan areas of up to 15 million inhabitants produce large quantities of waste in relatively small and confined areas. Much of the chemical products and waste products of modern society are released into the environment either during production, storage, transport, use or ultimate disposal. These released materials participate in natural cycles and reactions and frequently lead to interference and disturbance of natural systems. Environmental Chemistry is concerned with reactions in the environment. It is about distribution and equilibria between environmental compartments. It is about reactions, pathways, thermodynamics and kinetics. An important purpose of this Handbook, is to aid understanding of the basic distribution and chemical reaction processes which occur in the environment. Laws regulating toxic substances in various countries are designed to assess and control risk of chemicals to man and his environment. Science can contribute in two areas to this assessment; firstly in the area of toxicology and secondly in the area of chemical exposure. The available concentration (“environmental exposure concentration”) depends on the fate of chemical compounds in the environment and thus their distribution and reaction behaviour in the environment. One very important contribution of Environmental Chemistry to the above mentioned toxic substances laws is to develop laboratory test methods, or mathematical correlations and models that predict the environ-
VIII
Preface
mental fate of new chemical compounds. The third purpose of this Handbook is to help in the basic understanding and development of such test methods and models. The last explicit purpose of the Handbook is to present, in concise form, the most important properties relating to environmental chemistry and hazard assessment for the most important series of chemical compounds. At the moment three volumes of the Handbook are planned. Volume 1 deals with the natural environment and the biogeochemical cycles therein, including some background information such as energetics and ecology. Volume 2 is concerned with reactions and processes in the environment and deals with physical factors such as transport and adsorption, and chemical, photochemical and biochemical reactions in the environment, as well as some aspects of pharmacokinetics and metabolism within organisms.Volume 3 deals with anthropogenic compounds, their chemical backgrounds, production methods and information about their use, their environmental behaviour, analytical methodology and some important aspects of their toxic effects. The material for volume 1, 2 and 3 was each more than could easily be fitted into a single volume, and for this reason, as well as for the purpose of rapid publication of available manuscripts, all three volumes were divided in the parts A and B. Part A of all three volumes is now being published and the second part of each of these volumes should appear about six months thereafter. Publisher and editor hope to keep materials of the volumes one to three up to date and to extend coverage in the subject areas by publishing further parts in the future. Plans also exist for volumes dealing with different subject matter such as analysis, chemical technology and toxicology, and readers are encouraged to offer suggestions and advice as to future editions of “The Handbook of Environmental Chemistry”. Most chapters in the Handbook are written to a fairly advanced level and should be of interest to the graduate student and practising scientist. I also hope that the subject matter treated will be of interest to people outside chemistry and to scientists in industry as well as government and regulatory bodies. It would be very satisfying for me to see the books used as a basis for developing graduate courses in Environmental Chemistry. Due to the breadth of the subject matter, it was not easy to edit this Handbook. Specialists had to be found in quite different areas of science who were willing to contribute a chapter within the prescribed schedule. It is with great satisfaction that I thank all 52 authors from 8 countries for their understanding and for devoting their time to this effort. Special thanks are due to Dr. F. Boschke of Springer for his advice and discussions throughout all stages of preparation of the Handbook. Mrs. A. Heinrich of Springer has significantly contributed to the technical development of the book through her conscientious and efficient work. Finally I like to thank my family, students and colleagues for being so patient with me during several critical phases of preparation for the Handbook, and to some colleagues and the secretaries for technical help. I consider it a privilege to see my chosen subject grow. My interest in Environmental Chemistry dates back to my early college days in Vienna. I received significant impulses during my postdoctoral period at the University of California and my interest slowly developed during my time with the National Research
Preface
IX
Council of Canada, before I could devote my full time of Environmental Chemistry, here in Amsterdam. I hope this Handbook may help deepen the interest of other scientists in this subject. Amsterdam, May 1980
O. Hutzinger
Twentyone years have now passed since the appearance of the first volumes of the Handbook. Although the basic concept has remained the same changes and adjustments were necessary. Some years ago publishers and editors agreed to expand the Handbook by two new open-end volume series: Air Pollution and Water Pollution. These broad topics could not be fitted easily into the headings of the first three volumes. All five volume series are integrated through the choice of topics and by a system of cross referencing. The outline of the Handbook is thus as follows: 1. 2. 3. 4. 5.
The Natural Environment and the Biogeochemical Cycles, Reaction and Processes, Anthropogenic Compounds, Air Pollution, Water Pollution.
Rapid developments in Environmental Chemistry and the increasing breadth of the subject matter covered made it necessary to establish volume-editors. Each subject is now supervised by specialists in their respective fields. A recent development is the accessibility of all new volumes of the Handbook from 1990 onwards, available via the Springer Homepage http://www.springer. de or http://Link.springer.de/series/hec/ or http://Link.springerny.com/ series/hec/. During the last 5 to 10 years there was a growing tendency to include subject matters of societal relevance into a broad view of Environmental Chemistry. Topics include LCA (Life Cycle Analysis), Environmental Management, Sustainable Development and others.Whilst these topics are of great importance for the development and acceptance of Environmental Chemistry Publishers and Editors have decided to keep the Handbook essentially a source of information on “hard sciences”. With books in press and in preparation we have now well over 40 volumes available.Authors, volume-editors and editor-in-chief are rewarded by the broad acceptance of the “Handbook” in the scientific community. Bayreuth, July 2001
Otto Hutzinger
Contents
Foreword Dušan Gruden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
XIII
Introduction Dušan Gruden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1
Power Units for Transportation Dušan Gruden, Klaus Borgmann, Oswald Hiemesch . . . . . . . . . . . .
15
Means of Transportation and Their Effect on the Environment Hans Peter Lenz, Stefan Prüller, Dušan Gruden . . . . . . . . . . . . . .
107
Legislation for the Reduction of Exhaust Gas Emissions Wolfgang Berg . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
175
Fuels Dušan Gruden . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
255
Subject Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
289
Foreword
Over centuries mankind has pursued technical progress for the benefit of improved prosperity without simultaneously taking appropriate steps to ensure the environmental friendliness of the involved processes. However, in the middle of the 20th century environmental episodes drew attention to the negative impacts on the environment caused by this progress. As a matter of fact, concern about the influence of human activities on the environment is neither a new phenomenon nor a new attribute of modern people but has accompanied human society throughout its existence. What is new, however, is the increasing intensity of man’s efforts to protect his environment as reflected in a multitude of national and international environmental laws enacted all around the globe. Life as a whole, and human existence in particular, are characterized by constant movement and changes. This means that living beings need to be mobile to survive. By developing suitable technical means man has enormously increased his mobility – expressed in terms of speed and distance – when compared with other living beings on our planet. The automobile is one of the inventions that has made a decisive contribution to this mobility and it has become an inseparable part of modern human society. In the second half of the 20th century, the automobile developed from a luxury article and prestige object for a few into a basic commodity for millions of people. It is through this widespread use that negative impacts on the environment have become clearly visible. Therefore, since the late 1960s and early 1970s, automotive development has been accompanied by an ever increasing number of strict legal standards, e.g., about the reduction of exhaust gas pollutants, noise emissions, hazardous substances and waste, as well as about improved recyclability of materials and other aspects. Achievements in improving the ecological characteristics of the automobile are highly impressive: A modern car emits only fractions of the amounts of noise and exhaust gas pollutants produced by its predecessors 30 years ago. Today, 100 modern passenger cars in total emit less of the legally limited exhaust gas constituents than one single car of 1970. The same trend can be found with all the other ecologically relevant automotive features so that the absolute impact of the automobile on our environment is considerably lower today than it was in the past. The development of the automobile is increasingly linked to deliberations about sustainable development.While this term in the recent past was only related to the aspect of ecological consequences for the environment, it comprises
XIV
Foreword
today at least two further essential pillars, namely economic consequences and social responsibility. When discussing sustainability in the context of automotive development, it must be borne in mind that essential technical elements of the automobile – such as safety, power output, torque, fuel consumption, durability, maintenance intervals, and comfort should not be compromised. The modern automobile has achieved outstanding performance and superiority compared to its predecessors in all theses elements and will continue to proceed along this evolutionary development path. This book focuses on ecological aspects related to the development and use of automobiles, leaving many environment-related initiatives towards improvements of the automotive production process out of consideration. It shall, however, be mentioned in this context that also the production of modern cars is not possible without the observance of a wide range of stringent environmental laws. Thus, in order to be allowed to enter the market, a car must not only perform environmental-friendly during its operation but must have been produced to ecological standards as well. Company audits carried out routinely according to EMAS (Eco Management Auditing Scheme) and ISO 14001 show that automotive manufacturers are constantly improving the ecological compatibility of their production processes. The contributions to this book were written by experts, most of whom have been actively involved in the development of modern automobiles and their combustion engines for more than 30 years. They have participated in all phases of the ecological development of the automobile – from the basic attempts to respond to the first exhaust gas emission control requirements in the USA (1966) and Europe (1970) to the cost-intensive efforts towards meeting the comprehensive and highly demanding emission legislations currently existing and further anticipated worldwide. As the 20th century ends and the 21st century begins, these experts have summarized their experience and know-how in this book which bears witness to the successful implementation of ecological considerations into automotive development work. In my capacity as coordinator of the preparatory work for this book I would like to thank my colleagues – Prof. Dr. sc. techn. Hans Peter Lenz and his collaborator, Mr. Stefan Prüller (Dipl.-Ing.) of Technical University of Vienna, Dr. Klaus Borgmann and Mr. Otto Hiemesch (Dipl.-Ing.) of BMW AG and Dr. Wolfgang Berg, Consultant and long-standing collaborator of DaimlerChrysler AG – for their cooperation and valuable contributions. I would like to express particular gratitude to Dr. Ing. h.c. F. Porsche AG for permission to carry out this project. Weissach, June 2003
D. Gruden
The Handbook of Environmental Chemistry Vol. 3, Part P (2003): 1–15 DOI 10.1007/b 10445
The Diversity of Naturally Produced Organohalogens Gordon W. Gribble Department of Chemistry, Dartmouth College, Hanover, NH 03755, USA E-mail:
[email protected]
More than 3700 organohalogen compounds, mainly containing chlorine or bromine but a few with iodine and fluorine, are produced by living organisms or are formed during natural abiogenic processes, such as volcanoes, forest fires, and other geothermal processes. The oceans are the single largest source of biogenic organohalogens, which are biosynthesized by a myriad of seaweeds, sponges, corals, tunicates, bacteria, and other marine life. Terrestrial plants, fungi, lichen, bacteria, insects, some higher animals, and even humans also account for a diverse collection of organohalogens. Keywords. Organohalogen, Organochlorine, Organobromine, Natural halogen
1
Introduction
. . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2
2
Sources and Compounds . . . . . . . . . . . . . . . . . . . . . . .
2
2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 2.10
Marine Plants . . . . . . . . Marine Sponges . . . . . . . Other Marine Animals . . . Marine Bacteria and Fungi . Terrestrial Plants . . . . . . Fungi and Lichen . . . . . . Bacteria . . . . . . . . . . . Insects . . . . . . . . . . . . Higher Animals and Humans Abiogenic Sources . . . . .
3
Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . 13
4
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
. . . . . . . .
. . . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. . . . . . . . . .
. 2 . 3 . 6 . 7 . 7 . 8 . 9 . 10 . 11 . 12
© Springer-Verlag Berlin Heidelberg 2003
2
G. W. Gribble
1 Introduction Thirty years ago some 200 natural organohalogen compounds had been documented (150 organochlorines and 50 organobromines) [1]. Nevertheless, the scientific community generally considered these compounds to be isolation artifacts or rare abnormalities of nature. For example,“present information suggests that organic compounds containing covalently bound halogens are found only infrequently in living organisms” [2]. Unfortunately, even today this myth persists and has entered modern textbooks: “unlike metals, most of these compounds [halogenated hydrocarbons] are man-made and do not occur naturally …” [3]. The striking increase in the number of known natural organohalogens to more than 3700 is partly a consequence of the general revitalization of interest in natural products as a potential source of new medicinal drugs. Furthermore, the relatively recent exploration of the oceans has yielded large numbers of novel organohalogens from marine plants, animals, and bacteria. Much of the success of these explorations is attributed to improved collection methods (SCUBA and remote submersibles for the collection of previously inaccessible marine organisms), selective bioassays for identifying biologically active compounds, powerful multidimensional nuclear magnetic resonance spectroscopy techniques for characterizing sub-milligram quantities of compounds, and new separation and purification techniques (liquid-liquid extraction, high-pressure liquid chromatography). Furthermore, an awareness and appreciation of folk medicine and ethobotany have guided natural product chemists to new medicinal leads. Although most of the biogenic organohalogens discovered over the past thirty years are marine-derived, many other halogenated compounds are found in terrestrial plants, fungi, lichen, bacteria, insects, some higher animals, and humans [4–9]. As of June 2002, the breakdown of natural organohalogens was approximately: organochlorines, 2200; organobromines, 1900; organoiodines, 100; organofluorines, 30 [10].A few hundred of these compounds contain both chlorine and bromine.
2 Sources and Compounds 2.1 Marine Plants
Seaweeds produce an array of both simple and complex organohalogens, presumably for chemical defense. Some simple haloalkanes found in marine algae are shown in Fig. 1. Laboratory cultures of marine phytoplankton produce chloromethane, bromomethane, and iodomethane [11]. The favorite edible seaweed of native Hawaiians is “limu kohu” (Asparagopsis taxiformis), and this delicacy contains more than 100 organohalogens, most of which were previously unknown compounds [12, 13]. Bromoform is the major organohalogen in this seaweed.A selection of others is depicted in Fig. 2.Another red alga, Bonnemaisonia hamifera, contains several brominated heptanones that might be precursors to bromoform formed via a classical “haloform reaction”
The Diversity of Naturally Produced Organohalogens
3
Fig. 1. Some haloalkanes produced by marine algae
Fig. 2. Some organohalogens found in the red alga Asparagopsis taxiformis
[14]. Bromoform may serve as an antifeedant and/or antibacterial agent for the seaweed. A vast number of halogenated terpenes and the related C15 acetogenins are produced by marine plants. Nearly 50 species of the red alga genus Laurencia have yielded hundreds of such compounds; a small selection of recent examples is shown in Figures 3 and 4 [15–22]. Blue-green algae (cyanobacteria) are the source of a large number of halogenated, mainly chlorinated, metabolites [23]. In particular, Lyngbya majuscula is prolific in this regard and some recent examples are shown in Fig. 5 [24–27]. The potent anticancer drug candidate cryptophycin A (1) was isolated from cultures of a Nostoc sp. blue-green alga, and the structurally novel nostocyclophane (2) is produced by Nostoc linckia. A detailed study of the brown alga Cystophora retroflexa reveals the presence of seventeen halogenated phlorethol and fucophlorethol derivatives, one of which is the complex 3 [28] (Fig. 6). Synthetic approaches to cryptophycin are discussed later in this volume. 2.2 Marine Sponges
Sponges also rely heavily on chemicals for their survival, and many of these compounds contain halogen. In some cases, it is evident that bacteria or microalgae associated with the host sponge actually produce the metabolites. Recent exam-
4
Fig. 3. Some Laurencia terpenes
Fig. 4. Some Laurencia C15-acetogenins
G. W. Gribble
The Diversity of Naturally Produced Organohalogens
Fig. 5. Some organohalogens from the blue-green alga Lyngbya majuscula
Fig. 6. Some organohalogens from blue-green and brown algae
5
6
G. W. Gribble
Fig. 7. Some organohalogens from marine sponges
ples of sponge organohalogens include fatty acid derivatives (4) [29], pyrroles (5) [30], indoles (6) [31], phenol derivatives (7) [32], tyrosine derivatives (8) [33], terpenes (9) [34], diphenyl ethers (10) [35], and even dioxins (11) [36]. These fascinating compounds are illustrated in Fig. 7. 2.3 Other Marine Animals
Ascidians (tunicates or sea squirts), nudibranchs (sea slugs), soft corals (gorgonians), bryozoans (moss animals), and acorn worms all produce a dazzling collection of organohalogens. Some recent examples [37–40] are shown in Fig. 8.
The Diversity of Naturally Produced Organohalogens
7
Fig. 8. Some organohalogens from marine animals
2.4 Marine Bacteria and Fungi
A new thrust of natural product research is the study of marine bacteria and fungi.A number of novel organohalogens have been discovered in this endeavor, and recent examples (12–14) [41–43] are shown in Fig. 9. The novel halogenated bipyrroles 15 and 16, which are found in ocean-feeding sea birds [44–46], are most likely produced by marine bacteria. These compounds represent the first case of bioaccumulative natural organohalogens. The related “Q1” (17) has been discovered in a multitude of marine animals and even in the milk of Eskimo women who consume whale blubber [47, 48]. This latter scenario represents the first case of the bioaccumulation of natural organohalogens in humans. 2.5 Terrestrial Plants
By comparison with marine plants, terrestrial plants are relatively devoid of halogenated compounds. However, many notable exceptions do exist. The growth hormone 4-chloroindole-3-acetic acid (18) and its methyl ester are biosynthesized by peas, lentil, vetch, and fava bean (Fig. 10). Bromobenzene has been detected in the volatiles of oakmoss, and the Thai plant Arundo donax contains the weevil repellent 19 [49]. Both chloromethane and bromomethane have several plant sources. Chloromethane is produced by potato tubers [50], and bro-
8
G. W. Gribble
Fig. 9. Some organohalogens from marine bacteria and fungi
momethane, a commercial fumigant and nematicide, is produced by broccoli, cabbage, mustard, pak-choi, radish, turnip, and rapeseed [51]. The global annual production of bromomethane by rapeseed and cabbage is estimated to be 6600 and 400 tons, respectively. The authors conclude that “given the ubiquitous distribution of bromide in soil, methyl bromide production by terrestrial higher plants is likely a large source for atmospheric methyl bromide”. Some recent plant organohalogens (20–22) [52–54] are shown in Fig. 10. The edible Japanese lily (Lilium maximowiczii) produces seven novel chlorophenol fungicides in response to attack by the pathogenic plant fungus Fusarium oxysporum at the site of infection [55]. 2.6 Fungi and Lichen
Fungi and lichen produce a variety of organohalogens, from the simple chloromethane and chloroform to exceedingly complex compounds. The earliest discovered organohalogen compounds are the chlorine-containing fungal metabolites griseofulvin, chloramphenicol, aureomycin, caldariomycin,
The Diversity of Naturally Produced Organohalogens
9
Fig. 10. Some terrestrial plant organohalogens
sporidesmin, ochratoxin A, and others. A study of three species of fungi (Caldariomyces fumago, Mycena metata, and Peniophora pseudopini) revealed that they produce de novo up to 70 µg chloroform L–1 of culture medium per day [56]. The fungus Mollisia ventosa has yielded several calmodulin inhibitors such as KS-504d (23), which contains 70% chlorine by weight [57]. The novel topoisomerase inhibitors topopyrones A (24) and B (25) were isolated from a Phoma sp. fungus [58, 59], and a recent study of the white rot fungus Bjerkandera adusta has yielded bjerkanderol B (26) [60]. Experiments with Na37Cl supplied to the culture revealed incorporation of 37Cl in 26. The slime mold Dictyostelium purpureum produces AB0022A (27), which is the first naturally occurring chlorinated dibenzofuran [61]. These fungal metabolites are listed in Fig. 11. 2.7 Bacteria
Bacteria are amazing chemical factories and the resulting synthetic metabolites often possess astounding structural complexity. More than fifty Streptomyces
10
G. W. Gribble
Fig. 11. Some fungal and lichen organohalogens
species have yielded organohalogen metabolites. The bacterium Amycolatopsis orientalis produces the life-saving glycopeptide antibiotic vancomycin, which has been used for nearly 50 years to treat penicillin-resistant infections [62, 63]. The two chlorine atoms in vancomycin are essential for optimal biological activity. Recent examples of Streptomyces metabolites (28–31) [64–67] are listed in Fig. 12. 2.8 Insects
It is well known that insects use chemicals for both communication (“pheromones”) and defense (“allomones”), but very few of these compounds contain halogen. A notable exception is 2,6-dichlorophenol, the sex pheromone of at least a dozen tick species [68]. The German cockroach utilizes two chlorinated steroids as aggregation pheromones [69]. An extraordinary finding is that chloroform is produced by termites. Six Australian termite species produce chloroform within their mounds up to 1000 times higher than the ambient concentration [70]. The authors conclude that this source may account for as much as 15% of the global chloroform emissions.
The Diversity of Naturally Produced Organohalogens
11
Fig. 12. Some Streptomyces sp. organohalogens
2.9 Higher Animals and Humans
Organohalogens are rare in higher animals. However, several such compounds have been identified. The Ecuadorian frog Epipedobates tricolor has yielded epibatidine (32), and the iodolactone 33 is present in the thyroid gland of dogs. Recently, several halogenated compounds (34–36) were shown to be products of the action of human white blood cell myeloperoxidase-induced halogenation on invading pathogens and in various disease processes [71–73] (Fig. 13). This topic is also the subject of a chapter in this volume. Myeloperoxidase from humans
12
G. W. Gribble
Fig. 13. Some organohalogens from higher animals including humans
converts chlorophenols to chlorinated dioxins and dibenzofurans [74], and thus a human biosynthesis of dioxins is possible. The conversion of predioxins to dioxins in rats has been demonstrated [75]. 2.10 Abiogenic Sources
Natural combustion sources such as biomass fires, volcanoes, and other geothermal processes account for a wide range of organohalogens. The early studies of volcanic gases and the presence of organohalogens discovered therein by Stoiber and Isidorov are well documented [4, 6]. A recent study of the volcanoes Kuju, Satsuma Iwojima, Mt. Etna, and Vulcano has revealed an extraordinarily large array of organohalogens, including 100 organochlorines, 25 organobromines, 5 organofluorines, and 4 organoiodines, most of which are new compounds [76]. This topic is discussed further elsewhere in this volume. Haloalkanes have been found entombed in rocks, minerals, and shales. Thus, when rocks are crushed, for example, during mining operations, small quantities of CH3Cl, CH2Cl2, CHCl3, CCl4, CH3CHCl2, ClCH2CH2Cl, Cl2C = CH2,CH3CH2Br, CF2Cl2, CFCl3, CHF3, chlorobenzene, 1-chloronaphthalene, and other organohalogens are released [77, 78]. For example, 1000 tons of silvinite ore yields 50 g of chloroform. The authors estimate that the potassium salt mining industry alone accounts for the annual liberation of 10,000–15,000 tons of CHCl3 and 100– 150 tons each of CCl4 and CFCl3. Several chlorinated benzoic acids, some chloroalkanes, and other chlorinated aromatics, were found in the meteorites Cold Bokkeveld, Murray, Murchison, and Orgueil [79, 80]. While there is no dispute about the emissions of chloromethane and bromomethane from biomass burning and other natural sources [81, 82], the evidence regarding larger organohalogens, such as dioxins, has been more difficult to obtain and quantify [83]. However, numerous recent studies suggest that the
The Diversity of Naturally Produced Organohalogens
13
dioxins in sediments and clays have originated from natural sources [84, 85], and one such obvious source is biomass burning and subsequent deposition [86, 87]. Moreover, other studies indicate that dioxins are formed in peat and forest soil, presumably via the enzymatic oxidative dimerization of natural chlorophenols [88, 89].
3 Concluding Remarks The incredibly large number of marine and terrestrial organisms that are awaiting exploration for their chemical content virtually guarantees the discovery of numerous new natural organohalogens, many of which will doubtless have significant biological activity. It also seems highly likely that additional mammalian organohalogens will be identified and their role in the biodisinfection process will become understood. The clear and convincing evidence that chlorinated dioxins and dibenzofurans have several natural sources – both abiogenic and biogenic – is one of the most significant and politically important scientific discoveries of our age.
4 References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27.
Siuda JF, DeBernardis JF (1973) Lloydia 36:107 Fowden L (1968) Proc Roy Soc B 171:5 Clark RB (2001) Marine pollution, 5th edn. Oxford University Press, Oxford, p 126 Gribble GW (1996) Prog Chem Org Nat Prod 68:1 Gribble GW (1996) Pure Appl Chem 68:1699 Gribble GW (1998) Acc Chem Res 31:141 Gribble GW (2000) Environ Sci Pollut Res 7:37 Gaus C, Päpke O, Dennison N, Haynes D, Shaw GR, Connell DW, Müller JF (2001) Chemosphere 43:549 Gribble GW (1999) Chem Soc Rev 28:335 Gribble GW unpublished compilation Scarratt MG, Moore RM (1996) Marine Chem 54:263 Moore RE (1977) Acc Chem Res 10 : 40 McConnell O, Fenical W (1977) Phytochemistry 16:367 McConnell OJ, Fenical W (1980) Phytochemistry 19:233 Takahashi Y, Daitoh M, Suzuki M, Abe T, Masuda M (2002) J Nat Prod 65:395 Vairappan CS, Suzuki M, Abe T, Masuda M (2001) Phytochemistry 58:517 Guella G, Pietra F (1998) Chem Eur J 4:1692 Guella G, Pietra F (2000) Helv Chim Acta 83:2946 Vairappan CS, Daitoh M, Suzuki M, Abe T, Masuda M (2001) Phytochemistry 58:291 Iliopoulou D, Vagias C, Harvala C, Roussis V (2002) Phytochemistry 59:111 Guella G, Mancini I. Öztunc A, Pietra F (2000) Helv Chim Acta 83:336 Takahashi Y, Suzuki M, Abe T, Masuda M (1999) Phytochemistry 50:799 Burja AM, Banaigs B,Abou-Mancour E, Burgess JG,Wright PC (2001) Tetrahedron 57:9347 Kan Y, Sakamoto B, Fujita T, Nagai H (2000) J Nat Prod 63:1599 Jiménez JI, Scheuer PJ (2001) J Nat Prod 64:200 Sitachitta N, Márquez BL, Williamson RT, Rossi J, Roberts MA, Gerwich WH, Nguyen V-A, Wills CL (2000) Tetrahedron 56:9103 Luesch H, Yoshida WY, Moore RE, Paul VJ, Mooberry SL (2000) J Nat Prod 63:611
14
G. W. Gribble
28. 29. 30. 31. 32.
Sailler B, Glombitza K-W (1999) Nat Toxins 7 :57 Pham NB, Butler MS, Hooper JNA, Moni RW, Quinn RJ (1999) J Nat Prod 62:1439 Fattorusso E, Taglialatela-Scafati O (2000) Tetrahedron Lett 41:9917 Qureshi A, Faulkner DJ (1999) Nat Prod Lett 13:59 Utkina NK, Fedoreyev SA, Ilyin SG, Antipin MYu (1998) Russ Chem Bull 47:2292; Izves Akad Nauk Ser Khim 2363 Nicholas GM, Newton GL, Fahey RC, Bewley CA (2001) Org Lett 3:1543 Miyaoka M, Shimomura M, Kimura H, Yamada Y (1998) Tetrahedron 54:13467 Bowden BF, Towerzey L, Junk PC (2000) Aust J Chem 53 : 299 Uktina NK, Denisenko VA, Scholokova OV, Virovaya MV, Gerasimenko AV, Popov DYu, Krasokhin VB, Popov AM (2001) J Nat Prod 64:151 Ortega MJ, Zubía E, Ocaña JM, Naranjo S, Salvá J (2000) Tetrahedron 56:3963 Rudi A, Evan T, Aknin M, Kashman Y (2000) J Nat Prod 63:832 Sheu J-H, Sung P-J, Su J-H, Liu H-Y, Duh C-Y, Chiang MY (1999) Tetrahedron 55:14555 Morris BD, Prinsep MR (1999) J Nat Prod 62:688 Hardt IH, Jednsen PR, Fenical W (2000) Tetrahedron Lett 41:2073 Nielsen J, Nielsen PH, Frisvad JC (1999) Phytochemistry 50:263 Renner MK, Jensen PR, Fenical W (1998) J Org Chem 63 : 8346 Rittlemier SA, Simon M, Jarman WM, Elliott JE, Norstrom RJ (1999) Environ Sci Tech 33:26 Gribble GW, Blank DH, Jasinski JP (1999) Chem Comm 2195 Tittlemier SA, Blank DH, Gribble GW, Norstrom RJ (2002) Chemosphere 46:511 Vetter W, Alder L, Kallenborn R, Schlabach M (2000) Environ Pollut 110:401 Jun W, Vetter W, Gribble GW, Schneekloth JS Jr, Blank DH (2002) Angew Chem 41:1740 Miles DH, Tunsuwan K, Cittawong V, Hedin PA, Kokpol U, Ni C-Z, Clardy J (1993) J Nat Prod 56:1590 Harper DB (2000) Nat Prod Rep 17:337 Gan J, Yates SR, Ohr HD, Sims JJ (1998) Geophys Res Lett 25:3595 Chang F-R, Chen C-Y, Wu P-H, Kuo R-Y, Chang Y-C, Wu Y-C (2000) J Nat Prod 63:746 Zhu Y, Yang L, Jia Z-J (1999) J Nat Prod 62:1479 de Hernández ZNJ, Hernández LR, Catalán CAN, Gedris TE, Herz W (1997) Phytochemistry 46:721 Monde K, Satoh H, Nakamura M, Tamura M, Takasugi M (1998) J Nat Prod 61:913 Hoekstra EJ, De Weerd H, De Leer EWB, Brinkman UATh (1999) Environ Sci Technol 33:2543 Nakanishi S, Ando K, Kawamoto I, Yasuzawa T, Sano H, Kase H (1989) J Antibiot 42:1775 Kanai Y, Ishiyama D, Senda H, Iwatani W, Takahashi H, Konno H, Tokumasu S, Kanazawa S (2000) J Antibiot 53:863 Ishiyama D, Kanai Y, Senda H, Iwatani W, Takahashi H, Konno H, Kanazawa S (2000) J Antibiot 53:873 Silk PJ, Aubry C, Lonergan GC, Macaulay JB (2001) Chemosphere 44 : 1603 Sawada T, Aono M, Asakawa S, Ito A, Awano K (2000) J Antibiot 53:959 Williams DH (1996) Nat Prod Rep 13:469 Williams DH, Bardsley B (1999) Angew Chem Int Ed 38:1172 Schimana J, Fiedler H-P, Groth I, Süssmuth R, Beil W, Walker M, Zeeck A (2000) J Antibiot 53:779 Singh SB, Jayasuriya H, Salituro GM, Zink DL, Shafiee A, Heimbuch B, Silverman KC, Lingham RB, Genilloud O, Teran A, Vilella D, Felock P, Hazuda D (2001) J Nat Prod 64:874 Fukuda DS, Mynderse JS, Baker PJ, Berry DM, Boeck LD, Yao RC, Mertz FP, Nakatsukasa WM, Mabe J, Ott J, Counter FT, Ensminger PW,Allen NE,Alborn WE Jr, Hobbs JN Jr (1990) J Antibiot 43:623 Breinholt JB, Gürtler H, Kjaer A, Nielsen SE, Olsen CE (1998) Acta Chem Scand 52 : 1040 Berger RS (1983) J Med Entomol 20:103 Sakama M, Fukami H (1993) Tetrahedron Lett 34:6059 Khalil MAK, Rasmussen RA, French JRJ, Holt JA (1990) J Geophys Res 95:3619 Hazen SL, Heinecke JW (1997) J Clin Invest 99 : 2075
33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71.
The Diversity of Naturally Produced Organohalogens
72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89.
15
Heinecke JW (2000) J Clin Invest 105:1331 Henderson JP, Byun J, Mueller DM, Heinecke JW (2001) Biochemistry 40:2052 Wittsiepe J, Kullmann Y, Schrey P, Selenka F, Wilhelm M (2000) Chemosphere 40:963 Huwe JK, Feil VJ, Zaylskie RG, Tiernan TO (2000) Chemosphere 40:957 Jordan A, Harnisch J, Borchers, R, Le Guern F, Shinohara H (2000) Environ Sci Technol 34:1122 Isidorov VA, Prilepsky EB, Povarov VG (1993) J Ecol Chem 2 – 3 : 201 Isidorov VA, Povarov VG, Prilepsky EB (1993) J Ecol Chem 1 : 19 Nkusi G, Müller G, Schöler HF, Spitthoff B (1996) VM Goldschmidt Conference, March 31–April 4, 1996, Heidelberg Germany, J Conf Abst 1 : 435 Studier MH, Hayatsu R, Anders E (1965) Science 149:1455 Rhew RC, Miller BR, Weiss RF (2000) Nature 403 : 292 Yokouchi Y, Noijiri Y, Barrie LA, Toom-Sauntry D, Machida T, Inuzuka Y, Akimoto H, Li H-J, Fujinuma Y, Aoki S (2000) Nature 403:295 Martínez M, Díaz-Ferrero J, Martí R, Broto-Puig F, Comellas L, Rodríguez-Larena MC (2000) Chemosphere 41:1927 Fiedler H, Lau C, Kjeller L-O, Rappe C (1996) Chemosphere 32:421 Ferrario JB, Byrne CJ, Cleverly DH (2000) Environ Sci Technol 34:4524 Gaus C, Päpke O, Dennison N, Haynes D, Shaw GR, Connell, DW, Müller JF (2001) Chemosphere 43:549 Green NJL, Jones JL, Johnston AE, Jones KC (2001) Environ Sci Technol 35:1974 Silk PJ, Lonergan GC, Arsenault TL, Boyle CD (1997) Chemosphere 35:2865 Hoekstra EJ, De Weerd H, De Leer EWB, Brinkman UATh (1999) Environ Sci Technol 33:2543
The Handbook of Environmental Chemistry Vol. 3, Part T (2003): 15 – 106 DOI 10.1007/b11992HAPTER 1
Power Units for Transportation Dušan Gruden 1 · Klaus Borgmann 2 · Oswald Hiemesch 2 1 2
Dr. Ing. h.c. F. Porsche Aktiengesellschaft, Porschestrasse, 71287 Weissach, Germany E-mail:
[email protected] Bayerische Motoren Werke Aktiengesellschaft, Hufelandstrasse, 80788 München, Germany
For more than 125 years, gasoline and Diesel engines have prevailed as the exclusive drive unit in road transportation. None of the other power units invented to date has been able to make use of the energy content of mineral oil with the piston engine’s same good efficiency. Combustion is the fundamental process by which the chemical energy of fuels is converted into thermal energy and further into mechanical work. If hydrocarbon-containing fuels were completely burnt, the resulting products would be carbon dioxide and water vapor only. Since it is impossible to obtain a 100% complete combustion the exhaust gases always include a great variety of combustion products, the most important are: carbon monoxide, unburnt hydrocarbons, nitrogen oxides and particulate matter. During its 125 years of existence, the Otto (gasoline) engine – as it was called after its inventor – has been developed into a mature combustion engine which is characterized by an excellent efficiency and low pollutant emissions. The properties of the gasoline engine strongly depend on the composition of the air-fuel mixtures and ignition parameters. The influence of the socalled engine design parameters on combustion and exhaust emission is no less important. The emission of many of the exhaust-gas constituents can be influenced and minimized at their place of origin, that is in the combustion chamber by correctly selecting and adapting the relevant engine design and operating parameters. If optimization of engine-internal parameters for further reducing of the exhaust gas emissions are not enough anymore, so-called engine-external measures must be additionally taken. It was found that so-called three-way catalyst reduces the three aforementioned pollutants by clearly more than 90%, provided that a precisely stoichiometric A/F-ratio is used. Thanks to the strict maintenance of a precise stoichiometric air/fuel mixture the three-way catalyst allows very low HC, CO and NOx pollutant emissions to be achieved. However, in this operating range, fuel consumption is 8 to 15% higher (with a resulting higher CO2 emission) than during lean-burn operation. One of the technically most useful solutions to reduce the fuel consumption and CO2 emission of gasoline engines is to make them tolerate lean air/fuel mixtures. The future of the leanburn gasoline engines will almost exclusively depend on the successful development of NOxexhaust-gas after-treatment technologies for lean air/fuel mixtures. Diesel engines are internal combustion units with the highest thermal efficiency. Mixture formation is achieved through high pressure fuel injection. The fuel leads to self-ignition in the highly compressed air of the engine cylinder. The power and torque characteristics of modern Diesel engines are comparable with those of spark ignition (Otto) power units of equal capacity, the fuel consumption however is approx. 20% lower. The Diesel power unit has achieved a high status in transport. The world wide share of Diesel engines in passenger vehicles is now approx. 20%, whereas in freight transport on the roads and by water the share is approaching 100%, diesel being the only cost effective alternative. Increasingly, new methods for injection combustion, exhaust gas recirculation and after treatment (NOx-Cat, Diesel particle filter) are being pursued to meet the ever stricter emission legislations, aimed at limiting the effects on the environment. Ever since its invention, the 4-stroke reciprocating piston engine has been considered as a rather complex thermal unit which should better be replaced by far less complicated designs. © Springer-Verlag Berlin Heidelberg 2003
16
D. Gruden et al.
When summing up all the properties required to smoothly operate cars over wide speed and load ranges and a long lifetime, all alternative concepts have never succeeded in edging the Otto and Diesel engines out of their top positions. Further optimized versions of gasoline and Diesel engines will continue to prevail in the automotive domain in the coming 15 to 20 years. Due to their theoretically high efficiency and low pollutant emissions, fuel cells are among the most promising alternative energy sources of the future. Keywords. Combustion process, Otto engine, Gasoline engine, Diesel engine, Fuel/air mixture, Power output, Fuel consumption, Exhaust gas emission, Carbon monoxide, Unburnt hydrocarbons, Nitrogen oxides, Particulates, Operating parameter, Design parameter, Ignition, Injection, Compression ratio, Combustion chamber, Valve timing, Exhaust gas after-treatment, Catalyst, Particulate filter, Turbo charging, 2-stroke engine, Alternative engine, Fuel cell, Hybrid drive
1
Combustion Fundamentals and Combustion Products (D. Gruden) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
1.1 1.2 1.3 1.4 1.5 1.6
General Issues . . . . . . . . Carbon Monoxide (CO) . . Unburnt Hydrocarbons (HC) Nitrogen Oxides (NOx) . . . Particulate Matter (PM) . . References . . . . . . . . . .
2
The Otto (Gasoline) Engine (D. Gruden)
2.1 2.2 2.3 2.4 2.4.1 2.4.1.1 2.4.1.2 2.4.2 2.4.2.1 2.4.2.2 2.4.3 2.5 2.5.1 2.5.1.1 2.5.1.2 2.5.1.3 2.5.1.4 2.5.2 2.5.2.1 2.5.2.2 2.5.2.3 2.6
General Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . Power Output and Fuel Consumption . . . . . . . . . . . . . . Exhaust Gas Emission . . . . . . . . . . . . . . . . . . . . . . Engine-Internal Measures for Pollutant Reduction . . . . . . . Operating Parameters . . . . . . . . . . . . . . . . . . . . . . Air-Fuel Mixture . . . . . . . . . . . . . . . . . . . . . . . . . Ignition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Design Parameters . . . . . . . . . . . . . . . . . . . . . . . . Combustion Chamber Shape . . . . . . . . . . . . . . . . . . . Compression Ratio . . . . . . . . . . . . . . . . . . . . . . . . Limitation of Pollutant Reduction by Engine-Internal Measures Engine-External Measures for Pollutant Reduction . . . . . . . Fuel-Independent Measures . . . . . . . . . . . . . . . . . . . Secondary Air-Injection . . . . . . . . . . . . . . . . . . . . . EGR (Exhaust-Gas Recirculation) . . . . . . . . . . . . . . . . Portliners . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Thermal Exhaust-Gas After-Treatment . . . . . . . . . . . . . Fuel-Dependent Measures . . . . . . . . . . . . . . . . . . . . Oxidation Catalyst . . . . . . . . . . . . . . . . . . . . . . . . Reduction Catalyst . . . . . . . . . . . . . . . . . . . . . . . . 3-Way Catalyst Plus Oxygen Sensor . . . . . . . . . . . . . . . The Lean-Burn Engine – the Ultimate Target of Otto-Engine Development . . . . . . . . . . . . . . . . . . . . . . . . . . . Problems of Lean-Burn Operation . . . . . . . . . . . . . . . .
2.6.1
. . . . . . . . . . . . . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
18 20 20 21 22 24
. . . . . . . . . . . . . . 25 . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . .
25 27 30 30 30 30 32 33 34 34 34 35 36 36 36 36 37 38 38 38 38
. . 41 . . 41
17
Power Units for Transportation
2.6.2 2.6.3 2.6.3.1 2.6.3.2 2.7
State of the Art . . . . . . . . . . . . . . . . . . . . Exhaust Gas After-Treatment for Lean-Burn Engines DeNOx Catalyst . . . . . . . . . . . . . . . . . . . . NOx Storage Catalysts . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
43 43 44 45 47
3 3.1 3.1.1 3.1.2 3.1.3 3.1.4 3.1.5 3.2 3.2.1 3.2.2 3.2.3 3.2.4 3.3 3.4 3.5 3.5.1 3.5.2 3.6 3.6.1 3.6.2 3.6.3 3.7 3.8
The Diesel Engine (K. Borgmann, O. Hiemesch) . . . . . . General Issues . . . . . . . . . . . . . . . . . . . . . . . . . Formation of the Fuel Mixture, Combustion Process . . . . Power Unit . . . . . . . . . . . . . . . . . . . . . . . . . . Charge Cycle and Turbocharger Technology . . . . . . . . Fuel Injection Systems . . . . . . . . . . . . . . . . . . . . Injector Support and Injection Nozzle . . . . . . . . . . . . Current Status of Modern Diesel Engines and Future Trends Passenger Car Diesel Engines . . . . . . . . . . . . . . . . Utility Vehicle Diesel Engines . . . . . . . . . . . . . . . . Marine Diesels . . . . . . . . . . . . . . . . . . . . . . . . Future Trends in the Use of Diesel Engines . . . . . . . . . Fuel Consumption . . . . . . . . . . . . . . . . . . . . . . Exhaust Emissions . . . . . . . . . . . . . . . . . . . . . . Engine-Internal Measures for Reducing Exhaust Emission . Development of the Combustion Process . . . . . . . . . . Exhaust Gas Recirculation . . . . . . . . . . . . . . . . . . Exhaust Gas After-Treatment . . . . . . . . . . . . . . . . Oxidation Catalyst . . . . . . . . . . . . . . . . . . . . . . DeNOx Catalyst . . . . . . . . . . . . . . . . . . . . . . . . Particle Filter . . . . . . . . . . . . . . . . . . . . . . . . . Exhaust Gas Concepts and Outlook . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . . . . . . . .
48 48 50 53 54 57 59 62 63 67 67 68 71 73 74 75 79 81 81 82 89 88 89
4 4.1 4.2 4.2.1 4.2.2 4.3 4.3.1 4.3.2 4.3.3 4.3.4 4.4 4.5 4.6 4.6.1 4.6.2 4.6.3 4.7
Alternative Propulsion Systems (D. Gruden) . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . Thermal Engine with Discontinuous Combustion . . . . . . Two-Stroke Engine . . . . . . . . . . . . . . . . . . . . . . . Wankel Engine . . . . . . . . . . . . . . . . . . . . . . . . . Thermal Engine with Continuous Combustion . . . . . . . . Gas Turbine . . . . . . . . . . . . . . . . . . . . . . . . . . . Stirling Engine . . . . . . . . . . . . . . . . . . . . . . . . . Steam Engine . . . . . . . . . . . . . . . . . . . . . . . . . . Common Characteristics of Continuous Combustion Engines Electric Motor . . . . . . . . . . . . . . . . . . . . . . . . . . Flywheel Storage System . . . . . . . . . . . . . . . . . . . . Outlook on the Future . . . . . . . . . . . . . . . . . . . . . Hybrid Drive . . . . . . . . . . . . . . . . . . . . . . . . . . Fuel Cell . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Powerplants Using Alternative Fuels . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . .
. . . . . . . . . . . . . . . . .
91 91 91 91 92 93 93 93 94 94 95 96 97 97 99 105 106
18
D. Gruden et al.
1 Combustion Fundamentals and Combustion Products Dušan Gruden 1.1 General Issues
For thousands of years, horse- or ox-drawn carriages were the main means of locomotion to ensure what is called today passenger or public road transportation. The invention of the steam engine in the late 18th century was soon followed by the appearance of the railway train – a means of transportation which offered one essential advantage over the preceding ones: Steam engine-powered trains were much faster than all the previous means of locomotion. This attribute was so attractive that it triggered a people movement from slow individual vehicles to this speedy and more comfortable means of mass transportation. When, at the end of the 19th century, the piston internal combustion engine was invented which was so much smaller and more compact than the big unwieldy steam engine the obvious consequence was to fit it into a horse carriage. In 1886, the first motorized carriage was built in Stuttgart (Fig. 1) which went down in history as one of the first combustion-engine-equipped vehicles. It was
Fig. 1. First passenger car with internal combustion engine built in Stuttgart (1886)
Power Units for Transportation
19
then that the automobile was born. Since, from the very start, motorized automobiles were able to travel at the same speed as – if not faster than – trains they gave rise to another people movement, this time away from mass transportation and back to individual transportation. Despite intensive research and numerous efforts aimed at developing alternative propulsion systems, the piston engine has prevailed as the exclusive powerplant unit in road transportation. For more than 125 years, gasoline and Diesel engines have been the best answers engineers could find to the cheapest and most convenient terrestrial energy source. None of the other power units invented to date has been able to make use of the energy content of mineral oil with the piston engine’s same good efficiency. Gasoline and Diesel oil being regular by-products of oil refining, Otto and Diesel engines have never been mutually exclusive alternative concepts but have always ideally complemented each other in the efficient employment of mineral oil. Combustion is the fundamental process by which the chemical energy of fuels is converted into thermal energy and further into mechanical work needed for locomotion. Combustion in a heat engine consists in the rapid chemical oxidation of HCcontaining fuels. This reaction is accompanied by the release of major amounts of heat and luminous radiation. The released heat energy is then transformed into mechanical work by the reciprocating-piston mechanism. Even though combustion is the basic functional principle of a heat engine, it has not been possible, to date, to define a satisfactory combustion theory which describes the phenomena of combustion in every detail. What we have not got yet is a mathematical method allowing us to precisely calculate all phases of the combustion process taking place in the cylinder of an engine. This lack is due to the fact that combustion is a complicated chemical process characterized by rapidly changing temperatures and pressures and varying concentrations of the reactive substances. The chemical conversions taking place in a combustion engine have little to do with simple chemical reactions. The burning of hydrocarbons triggers chain reactions which are both consecutive and competing with each other. The fuels burnt in the cylinder of a combustion engine are not homogeneous simple hydrocarbons but rather consist of mixtures of hydrocarbons of different structures and highly varying percentages. At the present time, we are far from knowing the whole range of elementary processes going on during combustion. The velocity of the chemical reactions strongly depends on the chemical and physical properties of the reactive substances. The relationship between the reaction velocity (K) and temperature is given by Arrhenius’ law: K = C · e–E/RT where: C constant, E activation energy, R gas constant, and T temperature.
(a)
20
D. Gruden et al.
To simplify matters, combustion can be represented as follows: Fuel (CxHy) + oxygen (air) Æ chain reaction (combustion) Æ CO2 + H2O + CO + HC + NOx + …
(b)
Fuel combustion consists of chain reactions.According to the chain-reaction theory, the initial substances pass through a number of intermediate states before reaching the end-product condition. A chain reaction mainly depends on socalled active centers (free atoms, radicals, peroxides) which do not enter into contact with the initial compounds or intermediate products. If hydrocarbon-containing fuels were completely burnt, the resulting products would be carbon dioxide (CO2) and water steam (H2O). Combustion products also contain excess oxygen (O2) and nitrogen (N2). Since it is impossible to obtain a 100% complete combustion the exhaust gases always include a great variety of other products, too. 1.2 Carbon Monoxide (CO)
Carbon monoxide results from incomplete combustion of the carbons contained in fuel hydrocarbons. Theoretically – in the presence of sufficient oxygen (overstoichiometric, “lean” mixtures) – the carbon monoxide should be completely burnt to non-poisonous CO2 and not be present in the combustion products any longer. However, as CO measurements have shown, the carbon monoxide concentration in the exhaust gas is about 1 vol.-% with stoichiometric mixtures (l=1,0) with small amounts of CO being detectable also if lean mixtures (l>1,0) are used. The percentage of carbon monoxide contained in the exhaust gas strongly depends on the reaction temperature: At high temperatures, permanent counter reactions (CO2 dissociation) take place. Sudden cooling of the combustion gases in the expansion phase “freezes” the balance created at high temperatures thus causing carbon monoxide to be present in the exhaust gas under all operating conditions and A/F ratios. 1.3 Unburnt Hydrocarbons (HC)
Most of the unburnt hydrocarbons an automobile releases into the atmosphere come from the combustion process. The place in the cylinder and the moment at which unburnt hydrocarbons are generated has not yet been precisely determined. They occur even if there is sufficient oxygen for complete combustion, if flame propagation in the combustion chamber is perfect, if there is little residual gas and if there is an efficient distinct charge turbulence. Most scientists believe that the unburnt hydrocarbons result from incomplete flame propagation, causing the flame to be quenched at the cool walls of the combustion chamber (wall quenching). But the theory of flame quenching explains only part of the generation process of unburnt hydrocarbons. A major portion is generated through incomplete fuel combustion caused by residual gases which
Power Units for Transportation
21
Fig. 2. Sources of unburnt hydrocarbons in combustion chamber
strongly dilute the charge or by low cycle temperatures etc. Unburnt hydrocarbons are also created in those cylinder areas where the mixture cannot be reached by the flame, such as the space between the piston top land and the cylinder wall or the piston ring grooves (Fig. 2) [6]. In the expansion and exhaust phases, the unburnt hydrocarbons mix with the products resulting from complete combustion thus continuing their oxidation the intensity of which depends on temperature, the hydrocarbon and oxygen concentrations and the time available. The overall amount of unburnt hydrocarbons in the exhaust gas consists of a multitude of individual hydrocarbons. The exhaust gases of gasoline and Diesel engines contain several hundred hydrocarbon compounds with 1 to 9 (and more) C atoms. Unburnt hydrocarbons include paraffins, olefins, aromatic compounds, acetylene and their isomers, partly oxidized hydrocarbons (aldehydes, ketones, alcohols) as well as organic nitrogen and sulfur compounds. Some of these come unchanged from the fuel whereas others are combustion products. Each individual hydrocarbon compound needs a specific temperature to be generated.Any change of the operating conditions will automatically change the respective compound’s share in the overall amount of hydrocarbons. 1.4 Nitrogen Oxides (NOx)
The atmospheric air used for combustion essentially consists of nitrogen and oxygen molecules. Under normal conditions, it is chemically well balanced and very stable. Under temperatures of several hundred degrees, the two-atom nitrogen and oxygen molecules dissociate into their respective atoms and partly combine to
22
D. Gruden et al.
form nitrogen monoxide (NO). The degree of dissociation depends on the temperature and pressure levels and is accompanied by strong energy consumption. Provided that there is a sufficiently high amount of oxygen, the high cylinder temperatures in a combustion chamber further the partial oxidation of nitrogen from the air forming nitrogen monoxide. The NO concentration in the combustion engine mainly depends on the maximum combustion temperatures, the composition of the air/fuel mixture (A/F ratio) and the reaction time available. It is generally assumed that the combustion process produces NO only and that other nitrogen oxides such as NO2, N2O, N2O3, N2O4 and N2O5 are generated through continued NO oxidation in the expansion and exhaust phases and in the atmosphere. Nitrogen monoxide which has been generated and is then cooled down to ambient temperature will quickly oxidize in the atmospheric air to form NO2. Further atmospheric oxidation of NO2 into N2O4, for example, is considerably slowed down at ambient temperature. Low temperatures and high dilution with air allow nitrogen oxides to continue to exist in the atmosphere for a long time. 1.5 Particulate Matter (PM)
Besides the gaseous CO, HC and NOx emissions, Diesel engines also emit particulate matter (PM). Particulates have been defined as solid matter which is detected by diluting the engine exhaust gases with air, passing them through a filter at a temperature of less than 52°C and weighing the resulting residue. Thus as soot described particulates contained in the exhaust gas is the most obvious form of air pollution caused by combustion engines. The amount of soot measured in the exhaust gas from Diesel engines is a criterion of the quality of both the combustion process and the mixture control. Soot is an inevitable constituent of exhaust gases resulting from the combustion of organic fuels. Its amount and properties, however, depend on how the combustion process goes. In the past, distinction was made between three types of Diesel engine smoke emission: white, blue and black smoke. White smoke is generated if the combustion temperatures are low or if the ignition delay is too long. This kind of smoke occurs after the engine has been started and when the cylinder temperatures are high enough to evaporate but not to self-ignite the fuel. Blue smoke usually occurs when small amounts of lubricating oil penetrate and are burnt in the combustion chamber. Black smoke emitted under higher engine loads almost exclusively consists of carbon and other solid combustion products. The smoke is black if less than 1% of the carbon contained in the fuel is emitted in the form of soot. When analyzing the soot phenomenon, consideration must be given above all to the type of flame used for combustion. In the premixed flame of an gasoline engine, for example, the fuel vapors and the oxygen of the air are closely mixed and in direct contact with each other, so that no soot is generated if the amount of oxygen is sufficiently high (l≥1.0).
Power Units for Transportation
23
The extremely heterogeneous combustion of a Diesel engine is characterized by the simultaneous existence of a mixture of gases, vapors and liquid fuel in the combustion chamber whose concentrations vary continuously. These heterogeneous conditions (diffusion flame) result in incomplete chemical reactions allowing solid particles as well as unburnt or only partly burnt hydrocarbons to occur in the exhaust gas. Soot is generated at the flame front under high pressures and temperatures through various chemical and physical processes. It has not been possible to date, to scientifically determine the mechanisms of soot formation with sufficient precision. There are many hypotheses as to the particulates-forming reactions during Diesel-engine combustion none of which is able to provide a complete description of the processes involved. Quite frequently, polymerization is thought to be the primary source of soot formation in a diffusion flame. Other soot-generating reactions are dehydration, condensation and graphitization. An exemplary soot formation model is shown in Fig. 3 [10]. Particulates mainly consist of soot (black smoke). Soot is elementary carbon resulting from incomplete Diesel combustion. The organic compounds (hydrocarbons) settled down on the soot particles – also known under the designation of SOF (Soluble Organic Fraction) – consist of unburnt, partially cracked or polymerized hydrocarbons coming from the fuel and the lubricating oil. In addition, there are sulfates caused by the burning of the sulfur contained in the fuel. Particles also include residues of lubricants and fuel additives as well as settled-down water. Figure 4 shows the typical particle mixture of a Diesel engine at full load. The results of the particulates analysis suggest that all carbon-containing fuels are susceptible to forming particles.With aromatic compounds, this tendency is greater than with olefins and paraffins.A low hydrocarbon saturation level increases the particle formation trend. This means that the C/H ratio of the fuel is an essential parameter when it comes to evaluating the soot-formation propensity of fuels.
Fig. 3. FVV Project “Soot oxidation model” – soot formation and oxidation in Diesel engines
24
D. Gruden et al.
Fig. 4. Composition of particulates
Fig. 5. Size ranges of different types of particulate matter
The first particles have an almost spherical shape with diameters ranging between 0.002 and 0.01 µm. These particles agglomerate very quickly to form chains. A typical soot particle has a size of about 0.1 to 0.2 µm.With this scatter, Diesel soot is in the same range as numerous other particulates so that it is extremely difficult from a measuring point of view to clearly separate Diesel soot particles and particulates from other sources in the atmosphere (Fig. 5). 1.6 References 1. Woinov AN (1965) Verbrennungsprozesse in schnellaufenden Kolbenmotoren (russ.). Moskau 2. Fristrom RM, Westenberg AA (1965) Flame Structure. McGraw-Hill, New York 3. Bradley JN (1965) Flame and Combustion Phenomena. Methuen & Co Ltd., London
Power Units for Transportation
25
4. Gaydon AG,Wolfhard HG (1970) Flames. Their structure radiation and temperature. Chapman and Hall, London 5. Taylor CF (1985) The Internal Combustion Engine in Theory and Practice. The MIT Press, Cambridge 6. Stone R (1992) Introduction to Combustion Engines. The Macmillan Press, London 7. Warnatz J, Maas U, et al (2001) Verbrennung. Physikalisch-Chemische Grundlagen. Springer, Berlin, Heidelberg, New York 8. Warnatz J (1995) Probleme bei der Simulation von motorischen Verbrennungsprozessen. Symposium Kraftfahrwesen und Kraftfahrzeuge, Stuttgart 9. Polycyclic aromatic hydrocarbons in automotive exhaust emissions and fuels. CONCAWE Report No. 98/55, 1998 10. Pischinger S (1998) Rußbildung und Oxidation im Dieselmotor. FVV-Vorhaben “Rußoxidationsmodell”. FVV Frankfurt 11. Moser FX, Flotho A, et al (1995) Entwicklungsarbeiten an Dieselmotoren für den Nutzfahrzeug- und Industrieeinsatz zur Erfüllung der zukünftigen Emissionsanforderungen. Symposium Kraftfahrwesen und Verbrennungsmotoren, Stuttgart
2 The Otto (Gasoline) Engine Dušan Gruden 2.1 General Issues
When Nikolaus Augustus Otto had his patent registered in 1875, he doubtlessly was unaware of the repercussions his invention was going to have on humanity. During its 125 years of existence, the Otto (gasoline) engine – as it was called after its inventor – has been developed into a mature combustion engine which is characterized by an excellent efficiency. Along with their Diesel counterparts, Otto engines number among the heat engines having the highest combustion efficiency. This has allowed these two reciprocating-piston-engine variants to edge out of the market all other alternative power plant units which have been intensively examined so far as potential substitutes. And everything is pointing to the fact that these two power plant concepts will continue to prevail also in the foreseeable future and far into the 21st century. Of the 750 odd million passenger cars registered world-wide more than 90% are powered by gasoline engines – an indication of the enormous importance this type of propulsion system has had for mankind. The configuration of an Otto engine depends on the fuel type (gasoline) for which it has been laid out.According to the current state of knowledge, gasolines can only be efficiently burnt in a homogenous gasoline/air mixture. That is why, in the Otto engine, the fuel is injected into the intake manifold (or cylinder) in the suction phase already (Fig. 1). The intake and compression strokes (360° C.A.), which account for 50% of the working cycles, provide sufficient time to evaporate the fuel and intensively mix the air and fuel vapors.
26
D. Gruden et al.
Fig. 1. Mixture formation in Otto engine
As a matter of fact, homogeneous air/fuel mixtures need an external ignition source triggered by a spark-plug in order to be able to burn in a controlled regular manner. Following the ignition of the A/F mixture, the flame spreads throughout the combustion chamber at a velocity of 30 to 50 m/s. Gasoline engine combustion is represented by the Otto cycle (Fig. 2), consisting of adiabatic compression (T1–T2), isochoric heat supply (T2–T3), adiabatic expansion (T3–T4) and isochoric heat removal and/or gas exchange (T4–T1). The homogenous A/F mixtures in an Otto engine can be burnt efficiently only in a relatively narrow A/F-mixture range about the stoichiometric ratio (l=approx. 1.0, A/Fª14.5) and require a quantitative engine load control (throttling). With decreasing load, both the amount of fuel and the amount of air must be reduced in order to maintain the A/F ratio at a constant level. This means that the pressure and temperature levels in the combustion chamber at the moment of ignition keep dropping while the engine load diminishes (Fig. 3). In a Diesel engine, the amount of air sucked in and compressed is practically always the same regardless of the engine load. The pressure and temperature
Fig. 2. Thermodynamic cycle (Otto engine)
Power Units for Transportation
27
Fig. 3. P-V diagrams at different loads
levels reached at the end of the compression stroke are high and completely independent of the load which is controlled qualitatively (unthrottled) by reducing the amount of fuel injected. Besides the throttling losses, the unsatisfactory efficiency of the gasoline engine at part load is mainly due to the low pressures and temperature levels during the combustion process. In a Diesel engine, the combustion process takes place always at constantly high energy level. The differences between the partload behaviors of the Otto and Diesel engines are caused, among others, by the differences between their inherent energy potentials at which the combustion processes take place. 2.2 Power Output and Fuel Consumption
The properties of the gasoline engine strongly depend also on the composition of the A/F mixture or the A/F ratio l. Figure 4 illustrates the dependence of the specific work we (mean effective pressure) and the specific fuel consumption be on the A/F ratio (l). In the event of an air deficiency, homogeneous A/F mixtures can always be safely ignited and burnt in what is called the “rich” mixture range (l=0.8–0.9). It is in this range that Otto engines reach their highest mean pressures or power outputs. That was also the reason why the early generations of gasoline engines were exclusively operated on rich A/F ratios over the entire operating range from starting through idling to full-load. These operating conditions made no major demands on engine control. The required amounts of fuel and air were metered in the carburetor; the ignition timing was adjusted via the engine speed by means of a flyweight-controlled regulator in the ignition distributor and via engine load by means of a intake-man-
28
D. Gruden et al.
Fig. 4. Influence of air/fuel-ratio on specific work (power) and fuel consumption
Fig. 5. Air/fuel ratio and ignition timing maps of former gasoline engines
ifold-pressure-controlled vacuum advance unit. Exemplary l- and ignition-timing maps of a former carburetor engine are shown in Fig. 5. The low CO, HC and NOx exhaust emission limits prescribed by environmental legislation as well as the engine manufacturers’ constant efforts to reduce fuel consumption resulted in the development of highly complex electronic A/F-mixture and ignition control and regulation systems for modern gasoline engines (Fig. 6).
Power Units for Transportation
29
Fig. 6. Ignition timing map of modern gasoline engines
Fig. 7. Map of specific fuel consumption
To reach their maximum power output and torque levels, modern Otto engines use slightly enriched A/F ratios under full-load conditions only, yielding as naturally aspirated Otto engines specific power outputs of as high as Pe = 50 – 65 kW/l
(a)
30
D. Gruden et al.
and specific torques of Md = 90 – 105 Nm/l.
(b)
Turbocharged gasoline engines reached specific power outputs of Pe = 65 – 85 kW/l
(c)
and specific torques of Md = 125 – 170 Nm/l.
d)
In the part-load range, advanced Otto engines are operated on stoichiometric A/F ratios (A/F=14.5, l=1.0), in order to create optimum operating conditions for the 3-way catalyst (see Chapter 2.5). The lowest fuel consumption levels realized with modern Otto engines are about bemin=230–240 g/kWh (Fig. 7). 2.3 Exhaust Gas Emission
The explosive increase of the vehicle population in the industrial countries after World War II resulted in a new problem in the big population centers – with awareness starting in Los Angeles, USA: air pollution through exhaust emissions from combustion engines. First, it was the carbon monoxide (CO) and unburnt hydrocarbons (HC) which were rated as being noxious. Shortly thereafter, nitrogen oxides (NOx) were added to this group of pollutants. Since that time, the survival of the gasoline engine has depended and will continue to depend on its ability to comply with all the existing and planned regulations meant to reduce the burden on environment. 2.4 Engine-Internal Measures for Pollutant Reduction
For both Otto and Diesel engines, so-called engine-internal measures are the first choice when it comes to reducing pollutant emission. The emission of many of the exhaust-gas constituents can be influenced and minimized at their place of origin, that is in the engine cylinder or in the combustion chamber, by correctly selecting and adapting the relevant engine design and operating parameters. 2.4.1 Operating Parameters 2.4.1.1 Air-Fuel Mixture
Various investigations of the variables influencing the exhaust emissions of an Otto engine have shown that the amount of individual exhaust-gas constituents mainly depends on the composition of the air-fuel mixture (air/fuel ratio, A/F ratio or l) (Fig. 8).
Power Units for Transportation
31
Fig. 8. Influence of air/fuel ratio on fuel consumption and exhaust gas emission
The A/F ratio influences the composition of the exhaust gas far more strongly than any of the other combustion parameters, because it determines with relatively great precision whether the Otto engine is operated on a rich (l<1.0) stoichiometric (l=1.0) or a lean (l>1.0) mixture. The high levels of CO and unburnt hydrocarbons resulting from rich air/fuel mixtures are due to the fact that the mixture cannot be completely burnt for lack of oxygen. The only way of noticeably reducing these pollutants at their place of origin in the cylinder would be to increase the A/F ratio (mixture enleanment). The lack of air prevents excessive amounts of NOx from being generated even though the maximum combustion temperatures are high. Contrary to the results of corresponding equilibrium calculations, using stoichiometric mixtures (l=1.0) does not completely eliminate the CO contained in the exhaust gas, the residue being about 0.5 to 1.0 vol.%. Due to the reaction kinetics of the CO combustion, the exhaust gas contains a certain amount of CO even when l>1.0. The lowest HC levels are obtained with lean mixtures (lª1.1–1.3) or, in other words, with those A/F ratios at which the highest engine efficiency is reached. The high combustion temperatures and amounts of air required for the oxidation of CO and HC result in a steeply increasing NOx concentration. The maximum NOx level occurs in the same A/F ratio range in which the concentration of unburnt hydrocarbons is lowest.
32
D. Gruden et al.
Further leaning of the mixture deteriorates the combustion conditions: the maximum combustion temperatures and velocities decrease while the combustion time increases. The dropping temperature and deteriorating combustion result in higher HC levels and steeply decreasing NOx concentrations in the exhaust gas. Excessively lean mixtures frequently lead to sluggish combustion and complete misfires. Misfiring and lack of combustion, however, result in extremely high HC concentrations and increased fuel consumption. 2.4.1.2 Ignition
The operating behavior of an Otto engine – that is its power output, torque, fuel consumption and exhaust-gas composition – essentially depends on the ignition parameters, such as the functional characteristics of the spark plug, its location in the combustion chamber, the electrode gap and the ignition point. Not every spark is capable of igniting the A/F mixture. For the ignition to be triggered, the spark must have a certain minimum ignition energy which depends on the physico-chemical properties of the mixture next to the spark plug on the one hand and on the state of the electrodes on the other. Quite obviously, an ignition current of I=80–100 mA, a spark duration of t=1.5 to 2.0 ms and an ignition energy of 50 mJ are sufficient to make gasoline engines run also on lean air/fuel mixtures. It is not useful to further increase the ignition energy beyond the above mentioned values. The position of the spark plug in the combustion chamber influences the octane requirement of the engine, the lean limit of the mixture and the fuel consumption.When optimizing the ignition point (pre-ignition timing) consideration must be given to the power output and torque at WOT (wide open throttle) and to the fuel consumption at part load. Variations of the combustion velocity and
Fig. 9. Influence of ignition timing on fuel consumption and exhaust gas emission, 1-cylinder
engine
Power Units for Transportation
33
temperature caused by the ignition timing also have an influence on the exhaustgas composition (Fig. 9). That is why, for modern gasoline engines, it is essential to maintain the ignition timing stipulated for the instantaneous A/F mixture. The composition of the exhaust gas is not only influenced by the air/fuel mixture and ignition timing but also by all the other operating parameters, such as the temperature of the charge, engine and coolant, the temperature of the exhaust gas, deposits in the combustion chamber, the amount of residual gas etc. 2.4.2 Design Parameters
The influence of the so-called engine design parameters on combustion and exhaust emission is no less important. These design parameters include the cylinder displacement, S/D ratio, valve timing, layout of the intake and exhaust systems, shape of the combustion chamber, its surface/volume ratio and the compression ratio.
Fig. 10. Air/fuel ratio at misfire limit dependent on combustion-chamber shape and com-
pression ratio
34
D. Gruden et al.
2.4.2.1 Combustion Chamber Shape
The shape of the combustion chamber has a decisive influence on behavior of the gasoline engine mainly in the lean-burn range. The combustion chamber shape determines the movements of the charge (turbulences) upon which the combustion process highly depends. There are two possibilities to generate charge turbulences in the cylinder: The first solution consists in designing the intake manifold and intake duct in such a way that the charge is made to swirl or tumble during the intake stroke and that this swirl or tumble is maintained throughout the compression stroke. The second solution consists in shaping the combustion chamber in a way so as to realize squish effects which produce intensive turbulences in the combustion chamber at the end of the compression stroke. Since the intensity of the charge movement induced in the intake manifold and intake duct drops clearly during compression stroke, this solution must be combined with the squish effect produced by the combustion chamber shape. This combination allows optimum combustion conditions to be achieved and is particularly suited for lean air/fuel mixtures. The optimization of the combustion chamber shape is particularly helpful when it comes to shift the lean limit towards higher A/F ratios (so-called “leanburn” engines) (Fig. 10). Particularly good results are obtained when using spherical combustion chambers with two intake and exhaust valves each and a central spark plug. 2.4.2.2 Compression Ratio
Increasing the compression ratio is a generally applied method to improve the efficiency of a gasoline engine. To this solution, however, limits are set by the knock resistance of the fuel used. For decades, the compression ratio was chosen taking into account the power output and engine torque only.After the legislations on exhaust emissions had been introduced it was found that the compression ratio can have a considerable influence on HC and NOx emissions.Today,the compression ratio is chosen with power output, exhaust emissions and fuel consumption in mind. The compression ratio – which is to be chosen in accordance with the cylinder bore and combustion chamber shape – must be high enough to ensure optimum engine operation mainly with lean air/fuel mixtures. The problem of combustion knock at high compression ratios can be solved by providing for an appropriate layout of the combustion chamber (Fig. 11). To account for all those compromises, the compression ratios of modern Otto engines range between e =9.5 and 12. 2.4.3 Limitation of Pollutant Reduction by Engine-Internal Measures
The first measures meant to reduce CO and HC emissions, which started in Europe, followed later by the USA, were paralleled by efforts to lower fuel consumption. Be-
Power Units for Transportation
35
Fig. 11. Influence of combustion-chamber shape and compression ratio
tween R15 – the first emission regulation stipulated in Europe in 1971 – and R1504 which was valid until 1993, the emission limits as well as the fuel consumption of European cars were simultaneously reduced through engine-internal modifications. In doing so, one of the most efficient measures was the leaning of the air/fuel mixtures of gasoline engines from the original A/F ratio of l=0.8–0.9 of the early 1970s to l=1.05–1.15 of the early 1990s. Thus, the last pre-catalyst generation of European gasoline engines had been operated on lean air/fuel mixtures. The following exhaust emission levels could be reached in the ECE test just through engine-internal measures: CO=6.0–8.0 g/km HC=1.0–2.0 g/km NOx=1.5–2.5 g/km But these levels were not sufficient anymore to comply with the more and more severe emission limits. The introduction of extremely stringent exhaust gas limits mainly as far as NOx was concerned put a temporary end to the trend of simultaneously improving both the exhaust emissions and fuel economy. It was not possible any longer to satisfy the legislator’s demands by mere engine-internal improvements. 2.5 Engine-External Measures for Pollutant Reduction
If optimizations of engine-internal parameters for further reduction of the exhaust emissions are no longer sufficient, so-called engine-external measures
36
D. Gruden et al.
must be additionally taken which, in general, do not have any direct influence on fuel consumption. Indirectly, however, fuel economy is influenced as well, thanks to the engine readjustments required to ensure the functional reliability of the exhaust-gas after-treatment accessories. 2.5.1 Fuel-Independent Measures
The first category includes engine-external measures which do not place any special demands on the fuel quality, such as secondary air-injection, exhaust-gas recirculation, portliners and thermal reactors. 2.5.1.1 Secondary Air-Injection
This device provides the exhaust-gas system with fresh air to improve the CO and HC oxidation in the exhaust ducts. It is required for and particularly efficient in the presence of the rich A/F ratios (for cold starting, warming up and acceleration) during which the exhaust gas has a very high chemical energy. This system allows the CO and HC emissions to be lowered by 30 to 50% and by 20 to 40%, respectively, during these phases. To be able to meet extremely low emission limits many of the modern catalyst-equipped Otto engine variants must be fitted with secondary-air injection. During the respective operating phase, the secondary air is injected directly into the exhaust port by means of a secondary-air pump. The latter consumes 1 to 3% of the maximum engine power resulting in an increased fuel consumption. 2.5.1.2 EGR (Exhaust-Gas Recirculation)
Returning part of the exhaust gas into the cylinder – a process called exhaustgas recirculation (EGR) – is a service-proven way to reduce the NOx emission level. The influence of EGR on the combustion process is manifold: it lowers the charge-exchange losses and thus increases the pressure and temperature at the end of the compression stroke. The recirculated gas helps to improve the lean limit by warming up the fresh charge and it influences the flame propagation and thus the HC emission and lean-burn capability of the engine by serving as an inert constituent (residual gas). Therefore, to realize a modern low-NOx Otto engine, it is essential to provide for a precise control of the EGR system. 2.5.1.3 Portliners
The exhaust gas temperature downstream of the exhaust valve should be as high as possible to ensure the secondary reaction of HC and CO mainly if the engine is fitted with a catalytic exhaust-gas after-treatment system. This can be achieved
Power Units for Transportation
37
Fig. 12. Portliner
by heat-insulating tubes – so-called portliners – which prevent the heat of the exhaust gas from being transmitted to the cylinder head (Fig. 12).With rich air/fuel mixtures (cold starting, warming up) the portliners offer advantages only if combined with secondary air-injection. 2.5.1.4 Thermal Exhaust-Gas After-Treatment
As far as carbon monoxide and unburnt hydrocarbons are concerned, thermally well insulated exhaust pipes (thermal reactors) were used allowing the combustion process initiated in the combustion chamber to be continued. In the beginning, this approach – which considered the reactor as being a fully integrated constituent of the exhaust system – was thought to be technically correct and useful. However, to achieve efficient conversion rates in the reactors, temperatures of 700 to 800°C are required and it is essential that these temperature levels be reached also at low engine speeds and loads as well as immediately after cold starting. Consequently, the engine had to be tuned for high exhaust temperatures which resulted in excessive fuel consumption increases. So, the thermal reactor developments were stopped soon. In modern low-pollution engines, however, thermally well insulated exhaust pipes are one of the basic elements of catalytic exhaust after-treatment systems of modern gasoline engines. The engine-external fuel-independent measures for exhaust emission reduction (i.e., secondary air injection, exhaust-gas recirculation, thermal insulation) allow the following emission figures to be reached in the ECE test: CO=4–6 g/km HC=0.5–1.5 g/km NOx=0.5–1.5 g/km
38
D. Gruden et al.
2.5.2 Fuel-Dependent Measures
The above-mentioned solutions for exhaust emission reduction are not sufficient to comply with the stringent emission legislation. In the USA and Japan, catalytic exhaust-gas after-treatment systems have been successfully applied since the introduction of more severe emission limits in 1975. For these systems to be operative, unleaded fuel had to be made generally available, because noble-metal catalysts are sensitive to lead, sulfur and phosphorus and undergo rapid aging if exposed to these substances. 2.5.2.1 Oxidation Catalyst
In the catalytic reactors (catalysts) the oxidation of CO and HC is strongly enhanced by the reaction of the catalyst noble materials used such as platinum, palladium and rhodium. Optimum conversion rates are reached already at exhaust gas temperatures of as low as 200 to 250°C. Lean A/F mixtures offer optimum conditions for the reduction of CO and HC whereas with rich mixtures, a secondaryair pump is needed to inject additional fresh air upstream of the catalyst. Oxidation catalysts do not have any major influence on NOx emissions. 2.5.2.2 Reduction Catalyst
If engine-related measures and EGR do not yield the required low NOx levels, a so-called reduction catalyst must be used. To lower the NOx emission, a low-oxygen atmosphere is required or, in other words, rich air/fuel mixtures must be used. The great amounts of carbon monoxide (CO) contained in rich mixtures make sure that NO is split up into CO2 and N2. 2 CO + 2NO Æ 2 CO2 + N2
(e)
For the oxidation of relatively large volumes of CO and HC during rich A/F mixture operation an additional oxidation catalyst with secondary-air injection is required. Such a catalyst combination, consisting of one reduction and one oxidation catalyst each, is called a dual-bed catalyst. However, this is not a fuel efficient solution as it requires rich air/fuel mixture and permanent secondary-air injection. Therefore, automotive manufactures soon decided to drop this concept. 2.5.2.3 3-Way Catalyst Plus Oxygen Sensor
The legal demand for the drastic reduction of CO, HC and NOx with simultaneous improvement of fuel economy prompted power plant engineers to search for new technologies.After years of intensive engineering work it was found that socalled three-way catalysts including a precisely defined mixture of platinum (Pt),
Power Units for Transportation
39
Fig. 13. Exhaust emission of gasoline engines before and after 3-way catalyst
rhodium (Rh) and/or palladium (Pd) reduce the three aforementioned pollutants by clearly more than 90% provided that a precisely stoichiometric A/F ratio (l=1.0) is used (Fig. 13). By then, a reduction of that magnitude was required to meet the legal standards. When using a three-way catalyst, it must be ensured that the A/F ratio is fixed at l=1.0 throughout the major part of the engine map. This is guaranteed by the oxygen sensor or lambda probe developed for the purpose of determining the stoichiometric air/fuel mixture in the exhaust gas on the one hand and by using efficient electronic mixture control systems on the other (Fig. 14). The pollutant conversion rates of modern three-way catalysts being more than 98%, this technology has firmly established itself because it allows both current and future exhaust-gas standards to be complied with. Today, all Otto engines come with closed-circuit electronic systems for air/fuel mixture control, one or more oxygen sensors and one or more three-way catalysts. Modern gasoline engines use stoichiometric mixtures (l=1.0) over a wide operating range. The mixture is slightly enriched under full-load conditions and during cold-starting only in order to obtain the maximum possible power output from the given engine displacement and make sure that the mixture is correctly ignited and the catalyst protection functions are triggered. Figure 15 shows the A/F ratio map of one of the current gasoline engines with three-way catalyst and with l-control. Depending on the required air mass, measured by means of either an air-flow sensor in the intake manifold or via the intake manifold pressure (using the state equation pV=mRT), the amount of fuel needed for each operating condition is precisely dosed and fed into the engine. The ignition timing is programmed in accordance with the respective engine load, speed and temperature and the required exhaust-gas composition while providing for a safe distance from the knock limit. Thanks to the strict maintenance of a precise stoichiometric air/fuel mixture the three-way catalyst allows very low HC, CO and NOx pollutant emissions to be achieved which meet the current severe emission limits. However, in this oper-
40
D. Gruden et al.
Fig. 14. Gasoline engine with 3-way catalyst and oxygen sensor
Fig. 15. Air/fuel ratio map of gasoline engine with 3-way catalyst
ating range, fuel consumption is 8 to 15% higher (with a resulting higher CO2 emission) than during lean-burn operation if no severe NOx emission limits are to be observed. Gasoline engines reach the best fuel economy figures and lowest CO2 emissions when operated on lean air/fuel mixtures. Therefore, current gasoline engine development is focused on achieving optimum engine efficiency or, in other words, operating the engine at optimum A/F ratios at any operating conditions while meeting all the other engine-relevant demands at the same time.
Power Units for Transportation
41
2.6 The Lean-Burn Engine – the Ultimate Target of Otto-Engine Development
One of the technically most useful solutions to reduce the fuel consumption and CO2 emission of gasoline engines is to make them tolerate lean air/fuel mixtures. When compared with mostly stoichiometric engines, the part-load fuel consumption of lean-burn engines can be lowered by 8 to 15%. To reach this target, some basic conditions must be fulfilled beforehand: – Regular undisturbed combustion of externally ignited air/fuel mixtures with a high excess-air coefficient – the so-called lean-burn operation; – Regular undisturbed combustion of lean air/fuel mixtures also during transient operating conditions such as acceleration, deceleration, starting and warming up; – Low pollutant emissions with excess air according to current legislation. To date, the introduction of lean-burn concepts into production has been hampered by the fact that the three-way catalyst is not capable of reducing the NOx emissions during excess-air operation. Carbon monoxide and unburnt hydrocarbons can be reduced via the service-proven noble-metal oxidation catalyst, whereas no safe, durable and production-ready technology is available yet for the required drastic reduction of the NOx emissions generated by lean-burn engines. To date experience has shown that, in order to guarantee the three afore-mentioned prerequisites which are vital for the successful development and market introduction of the so-called lean-burn engine, the engine-internal processes must be precisely controlled. This is mainly true for the air/fuel mixture which must be tuned in such a way that the respective optimum level is maintained for each operating condition. 2.6.1 Problems of Lean-Burn Operation
Gasoline engines show certain combustion irregularities – a phenomenon called “cycle-by-cycle variation” that many generations of engine specialists have had to deal with. Even though this stochastic problem of cyclic variation is characteristic of gasoline engine combustion, it remains almost unnoticed when rich or stoichiometric mixtures are used. This is due to the resulting stronger torque variations and higher emissions outside of the cylinder are not occurring. If the mixture is further leaned beyond the stoichiometric A/F ratio, however, the variations of the working cycles intensify, resulting in more and more irregular crankshaft revolutions thus deteriorating the driving comfort of vehicles powered by lean-burn engines. In Fig. 16, the fuel consumptions as well as the HC and NOx emissions of a standard Otto engine and a lean-burn engine have been plotted versus l at the part-load n=2000 rpm and pme=2 bar. When the engine approaches the leanburn limit, the working-cycle variations and HC emissions start increasing. The
Fig. 16. Air/fuel ratio for gasoline engine
42 D. Gruden et al.
Power Units for Transportation
43
NOx emissions keep dropping as the process temperature decreases. Fuel consumption drops steeply up to l=1.3 and continues to decrease towards a minimum limit value if the mixture is further enleaned. The resulting fuel consumption at that load point is about 10% lower than with l=1.0. If the mixture of a standard production engine is further enleaned beyond l=1.45, fuel consumption increases again due to the distinct slow-down of the combustion and resulting misfires at the lean-burn limit. A “lean-burn” Otto engine should therefore be operated as closely as possible to the lean-burn limit in order to minimize NOx emissions but not too close to it either in order not to be penalized by an excessively high fuel consumption increase. 2.6.2 State of the Art
Over the decades, engineers have tried again and again to design Otto engines for lean-burn operation. Different approaches using homogenous and heterogeneous fuel/air mixtures have been proposed and tried out. In the 1970s, work was concentrated on the so-called stratified-charge engine with the only engine of that type installed in a production car being the Honda CVCC unit (Controlled Vortex Combustion Chamber). However, the classical Otto engine using stoichiometric air/fuel mixtures (l=1.0) has turned out to be a better alternative. It is mainly due to the severe emission limits that none of the lean-burn engines developed to date has been able to assert itself. New efforts for realizing a production-ready lean-burn engine are being made by Toyota, Porsche, Honda, Ford etc. There are two distinct trends in modern lean-burn-engine development: using either homogeneous air/fuel mixtures injected into the intake manifold or heterogeneous air/fuel mixtures by direct injection into combustion chamber offering the possibility of charge stratification. In gasoline engines, operated on homogeneous air/fuel mixtures, the fuel is injected into the manifold upstream of the intake valve. Gasoline engines fed with heterogeneous air/fuel mixtures have direct injection into the cylinder and are called Gasoline Direct Injection or GDI engines. GDI engines can use both homogeneous and heterogeneous mixtures. Both lean-burn variants use lean mixtures in the part-load range only. At full load and high part load, they, too, are operated on stoichiometric or slightly enriched mixtures in order to achieve the maximum possible torque and power output. For reasons of comfort, stoichiometric mixtures are also applied in the nearidling range. Figure 17 shows the characteristic A/F ratios map of a modern direct-injection lean-burn engine. 2.6.3 Exhaust Gas After-Treatment for Lean-Burn Engines
The legislator’s demands in terms of CO, HC and NOx emissions at the present time can only be met with the help of three-way catalysts.
44
D. Gruden et al.
Fig. 17. Air/fuel-ratio map of lean-burn gasoline engine
One of the main problems of lean-burn gasoline engines is the NOx emission which is greatest at l=1.05–1.2. Thanks to the high A/F ratio and low combustion temperatures, however, the NOx raw emissions of lean-burn engines are lower than those of gasoline engines operated on stoichiometric air/fuel mixtures. Due to the strongly oxidizing atmosphere (excess air, 5–9% oxygen content) it is not possible to further lower the NOx levels with conventional exhaust gas after-treatment systems alone. When combined with an engine operated on high A/F ratios, three-way catalysts will at best function as oxidation catalysts converting HC and CO into water vapor and carbon dioxide provided that there is sufficient oxygen. In the presence of lean A/F mixtures, three-way catalysts are not capable of reaching the desired NOx conversion rates because there is not enough CO available to do so on the one hand. On the other, NOx reduction is made impossible by the prevailing low temperature levels and the residual oxygen content of the exhaust gas. 2.6.3.1 DeNOx Catalyst
At the moment, various systems for post-engine NOx reduction are being investigated and developed which also function in the presence of high A/F ratios. Great hopes are placed in the DeNOx catalysts which use so-called zeolites, for example, to reduce NOx with the help of hydrocarbons. There are two potential variants available: – a noble-metal variant which allows nitrogen reduction at relatively low exhaust-gas temperatures of 180 to 200°C. However, this temperature window is very narrow and there is the problem of simultaneous nitrous oxide (N2O) generation; – a noble-metal-free ion-exchange-zeolite variant which seems to be more promising. It has a wide temperature window of about 300 to 600°C which is
Power Units for Transportation
45
better suited to cope with the exhaust gas temperatures of a lean-burn Otto engine. State-of-the-art systems of that kind allow NOx reductions of about 40% to be achieved even with high A/F ratios of l=1.5. Nevertheless, the remaining NOx emissions are up to four times higher than those of an engine running with stoichiometric air/fuel mixture and with an oxygen sensor-controlled three-way catalyst. The resulting demands on engine management are high: It must provide for optimum operating conditions of the exhaust-gas after-treatment system if the DeNOx catalyst is to function smoothly. One approach might be to use a management system which adjust the engine in such a way that an optimum HC/NOx ratio is guaranteed while maintaining the exhaust gas temperature at a level suited for exhaust gas after-treatment. The development of DeNOx catalysts is rapidly progressing but there still are some problems which have to be overcome: durability is not satisfactory yet, the conversion rates are insufficient and the temperature window for efficient conversion still is too narrow. 2.6.3.2 NOx Storage Catalysts
With regard to their comparatively high NOx conversion rates and the three-waycatalyst-type properties at A/F ratios of l=1.0 the NOx storage catalysts are the most promising alternative for the efficient after-treatment of the exhaust gas of combustion engines operated with excess air. With this solution, the nitrogen oxide produced during lean-burn combustion is not reduced but mainly adsorbed in an NOx storage catalyst where it is stored for some time. NOx accumulating catalysts include an additional storage medium consisting of a basic coating (oxide layer made of alkaline or alkaline earth metals) besides the usual noble-metal three-way layer. The resulting properties under stoichiometric conditions are almost the same as those of the three-way catalysts. In lean-burn mode, the storage medium fixes the nitrogen oxides in the form of nitrates. Since the system has a limited storage capacity only, a so-called regeneration must be performed from time to time. This is achieved by enriching the mixture to l£1. In doing so, the nitrates are split up into their constituents and the released nitrogen oxides are converted by the three-way catalyst (noblemetal layer) during stoichiometric or enriched operation. The main reducing agents used are CO, hydrocarbons and hydrogen. CO + NO ´ N2 + CO2
(f)
HC + NO ´ N2 + H2O
(g)
The precise mechanisms of NOx adsorption during lean-burn operation and nitrate disintegration at l£1 have not yet been clearly described. A potential reaction has been illustrated in Fig. 18 using the barium oxide layer (BaO) as a basis.
46
D. Gruden et al.
Fig. 18. Adsorption and nitrate formation of NOx at storage medium barium oxide (BAO)
at l >1.0 and nitrate decomposition at rich mixture
The adsorber is regenerated with the help of a downstream three-way catalyst during stoichiometric or enriched engine operation. For this strategy to function there must be a controlled switching between stoichiometric/rich and lean-burn operation phases. The target is to achieve an optimum balance between lean and stoichiometric/rich phases with maximum fuel economy on the one hand and optimum NOx reduction on the other and with lean-burn operation prevailing. Under optimum conditions and when using a new unaged catalyst combination (adsorber+threeway catalyst), nitrogen oxide reduction of up to 90% can be achieved according to the current state of the art. One of the main problem is the sulfur contained in conventional fuels which hampers the smooth functioning of the NOx storage layer and the catalyst. Slightest amounts of sulfur – which compete with nitrogen for storage capacity in the lean-burn range – drastically reduce the storage efficiency and the thermal stability of the catalyst. In addition, the durability and temperature resistance of the NOx storage catalysts are not sufficient yet for large-scale series introduction. Since none of the approaches for NOx reduction – such as DeNOx catalysts, catalytic converters with additives such as ammonia or urea, selective catalysts, zeolites, adsorbers and so on – have reached production readiness yet, the future of the lean-burn engines will almost exclusively depend on the successful development of NOx exhaust-gas after-treatment technologies for lean air/fuel mixtures.
Power Units for Transportation
47
2.7 References 1. Eberan-Eberhorst R (1969) Abgasforschung zukunftsweisend für den Fahrzeug- Ottomotor. MTZ-30, 9 2. Daniel WA, Wentworth JT (1967) Exhaust Gas Hydrocarbons- Genesis and Exodus. SAEPaper 486 B 3. Gruden D (1973) Veränderungen und Grenzwerte der Abgasemission im 4-Takt-Fahrzeug Ottomotor und Wege zur Verringerung des Abgasgeruches. Dissertation der TH-Wien 4. Obländer K, Nagel A (1984) Übersicht über Maßnahmen zur Minderung von Kfz-Emissionen. Staub-Reinhaltung der Luft 44, 9 5. Gruden D, Markovac U, et al (1979) Schadstoffarme Antriebssysteme – Entwicklungsstand, Wirtschaftlichkeit, Kosten. Berichte 2/80 Umweltbundesamt, Berlin 6. Gruden I (2000) Emission und verbrauchsoptimierte Regelung von homogen mager betriebenen Ottomotoren. Dissertation der TU-Wien 7. Stone R (1992) Introduction to internal combustion engines. The Macmillan Press, London 8. Krämer M, Maly T, et al (1995) Emissionsreduzierung beim mager betriebenen Ottomotor. 16. Wiener Motorensymposium 9. Hohenberg G (1997) Analyse der Gemischbildung und Verbrennung am D.I.-Ottomotor. 18. Wiener Motorensymposium 10. Wurster W, Gruden D (1988) Die Verbrennung im Otto- und Dieselmotor mit direkter Einspritzung. 9. Internationales Wiener Motorensymposium 11. Inoue T, Matsushita S, et al (1993) The Development of a High Fuel Economy and High Performance Four Valve Lean Burn Engine. SAE-Paper 930873 12. Iwamoto Y, Noma K (1997) Development of Gasoline Direct Injection engine. SAE-Paper 970541 13. Grebe UD (2000) Zukunft des Ottomotors – Benzindirekteinspritzung oder Laststeuernde. Variable Ventiltriebe. ÖVK-Wien 14. Moser W (1999) Benzin-Direkteinspritzung – ein Beitrag zur Absenkung der CO2-Emissionen. VDA-Technischer Kongress 15. Noma K, Iwamoto Y, et al (1998) Optimized Gasoline Direct Injection for the European Market. SAE-Paper 980150 16. Baumgarten H, Goerts W, et al (2000) Niedrigstemissionskonzept zur Erfüllung der SULEVEmissionsstandards für Ottomotoren. MTZ 61:10 17. Glück K-H, Göbel U (2000) Die Abgasreinigung der FSI-Motoren von Volkswagen. MTZ 61:6 18. Freidl GK, Piock W, et al (1997) Direkteinspritzung bei Ottomotoren. Aktuelle Trends und zukünftige Strategien. MTZ 58:12 19. Brandt S, Dahle U, et al (1998) Entwicklungsschritte bei NOX-Adsorber Katalysatoren für magerbetriebene Ottomotoren. Stuttgarter Symposium FKFS 20. Domes W, Gerwig W, et al (1985) Die neuen 16-Ventil-Motoren für Scirocco und Golf. ATZ 87:6 21. Göbel U, Kreuzer T, et al (1999) Moderne NOX-Adsorber-Technologien Grundlagen Voraussetzungen, Erfahrungen. VDA-Technischer Kongress 22. Gruden D, et al (1993) Entwicklungstendenzen auf dem Gebiet der Otto-Motoren. Expert Verlag Renningen
48
D. Gruden et al.
3 The Diesel Engine Klaus Borgmann · Oswald Hiemesch 3.1 General Issues
In a thorough process of research and development, Rudolf Diesel set out to develop a thermal engine with a high degree of efficiency, since he regarded the steam engine serving as the principal drive system in the 19th century, with an efficiency of only about 3 percent, as an enormous waste of energy. Applying for a patent for his invention in 1892, and publishing his study on the Theory and Construction of a Rational Thermal Engine in 1893, Diesel successfully laid the foundation for the thermal engine now referred to as the Diesel engine with the highest level of efficiency (currently up to 53 percent) ever achieved. As early as 1897, Diesel’s test engine achieved an efficiency of 26.2 per cent which was a sensation at the time. With this increase in efficiency, by almost a factor of 10, the steam engine was very quickly replaced by the Diesel engine which served initially as a stationary drive system. Thanks to significant progress on fuel injection and turbocharger technologies, the Diesel engine has become used increasingly for transportation purposes in recent decades. In rail-bound and ship transport the Diesel engine has played a dominating role for many years. In utility and commercial road transport (utility vehicles and buses) the Diesel engine is uncontested by any kind of competition. In individual transport (passenger cars) the Diesel engine is looking at an extremely positive future, with registration figures going up steadily, particularly in Europe. With world production of road vehicles continuing to increase (Fig. 1 [1]) the Diesel engine will continue to increase its future market share from the petrol engine. The Diesel engine, like the petrol engine, uses a controlled burning process to convert the chemical energy in the fuel into mechanical energy. Unlike the petrol engine however, combustion of the fuel injected is initiated by self-ignition in the highly compressed air. Most of the energy lost is thermal energy dissipating through the exhaust, into the coolant and by way of radiation. The process of comparison representing the Diesel engine in theory (see Fig. 2) is the Seiliger Process made up of: – – – –
adiabatic compression 1–2, isochoric (Qv) and isobaric (QP) heat input, adiabatic expansion 3–4, and the isochoric heat rejection 4–1.
The p-v diagram shows the useful work L in the area 123¢341. In the T-S diagram this work equals thermal energy Q, again in area 123¢341. The fresh charge air in the cylinder has to be compressed to a high level in order to initiate self-ignition of the injected fuel in the Diesel combustion process (see also Fig. 3 in the Chapter 2.1). The Diesel engine has the highest standard of
Power Units for Transportation
Fig. 1. Forecast of worldwide vehicle production
Fig. 2. Seiliger process
49
50
D. Gruden et al.
Fig. 3. Thermal balance of a Diesel engine
efficiency of all thermal engines because the loss of energy, as shown in Fig. 3, is kept to a minimum. Representing a highly complex energy conversion system, the Diesel engine is made up of the following parts and components: – The fuel/air formation and combustion system consisting of the air and fuel supply units as well as the combustion chamber (or, respectively, the piston combustion chamber in the case of direct fuel injection). – The engine itself is made up of the crankcase, crankshaft, connecting rods, pistons and cylinder head. In the interest of safe and reliable operation, the lubricating system and cooling are incorporated directly within the engine. – Seeking to convert energy with maximum efficiency and reduce emissions affecting the environment to a minimum, the exhaust gas path has been developed into an increasingly complex system serving to cut back noise and exhaust emissions to the lowest possible level. 3.1.1 Formation of the Fuel Mixture, Combustion Process
Proper operation of the Diesel engine requires the fulfillment of various criteria: a) High compression of the air flowing into the charge cylinder in order to reach the self-ignition temperature of the fuel injected. The fuel used for this purpose must be easily ignitable (a factor defined by the cetane number).
Power Units for Transportation
51
Fig. 4. Main phases of mixture formation and combustion in a Diesel engine
b) A sufficient fuel/air mixture formation process for burning the fuel injected in short time during the operating cycle, i.e., with a high standard of efficiency. With this internal fuel/air mixture formation and the combustion process taking a certain period of time, the timing of the process must be coordinated with the position of the piston in the working cylinder. The pressure curve in the operating cylinder following from the combustion process and the position of the crankshaft ultimately determine the energy transmitted to the crankshaft during an working cycle. Engine output, fuel consumption and exhaust emissions thus depend to a large extent on the formation of the fuel/air mixture and the combustion process. With the fuel being injected in a liquid state directly into the hot, highly compressed air, atomizing into minute droplets, igniting where the conditions for self-ignition are most favorable, and finally burning as a fuel/air mixture, we speak of internal heterogeneous fuel/air mixture formation within the diesel engine. This very complex interaction [2] of fuel injection, mixture formation, ignition and combustion depends on local temperatures, pressure conditions, the concentration of substances reacting with one another, velocity factors, etc. physical processes such as flow conditions, atomization, and evaporation are subject to, and affected by, chemical reactions taking place at the same time (see Fig. 4). Since these processes take a certain time, it is easy to understand that the maximum speed a diesel engine is able to reach is limited and lies at a lower level than the running speed of a spark-ignition engine with external fuel/air mixture formation and ignition. The faster the fuel/air mixture formation process, the higher are the engine speeds achievable.
52
D. Gruden et al.
The main factors crucial to the fuel/air mixture formation process are therefore: – the kinetic energy of the injection jet depending primarily on injection pressure; – the controlled movement of air in the operating cylinder, and – the thermal condition (wall temperature) of the combustion chamber and the compressed air affecting the local temperature and, accordingly, the intensity of evaporation of the fuel droplets as well as the fuel film resting on the walls of the combustion chamber. Depending on the mode of fuel injection, either directly into the combustion chamber or into a side or ancillary chamber, we distinguish between – the direct injection Diesel engine and – the chamber Diesel engine. Figure 5 presents the various combustion processes as a function of the air and injection energy required in each case.With the development of high-pressure injection systems, the direct-injection engine, benefiting from its higher degree of efficiency, has taken a leading position in all areas and applications. Running at relatively high speeds (up to 5000 rpm) and with lower noise and exhaust emissions, chamber Diesel processes (see Fig. 6) dominated the passenger car market until the end of the 20th century.
Fig. 5. Combustion process as a function of mixture formation energy
Power Units for Transportation
53
Fig. 6. Diesel chamber combustion process
With drive systems saving fuel and resources in general being widely promoted in the market, the direct-injection Diesel engine has taken on the leading role in all areas. The direct-injection Diesel engine uses the movement of air in the cylinder to form the fuel/air mixture, depending in each case on the cylinder stroke volume (diameter). This controlled air motion, also referred to as swirl action, serves to atomize the fuel injected even with a reduced number of nozzles. The swirl of air in the cylinder results from the specific design and configuration of the intake ports and is used to provide an appropriate level of variability, primarily on multi-valve engines. On large diesels with a bore diameter of approximately 150 mm (5.90˝) or more the combustion process is coordinated through the large number of nozzles and the flatter troughs of the combustion chambers, without any swirl effect (see Fig. 7). 3.1.2 Power Unit
As with every piston engine, the basic design and configuration of a Diesel engine is determined by the – stroke and bore (giving the cylinder capacity), – connecting rod ratio, and – compression ratio. All other design parameters are varied according to the specific function of the engine and the design chosen. The main distinctions when compared to a spark ignition engine are the – – – –
higher compression ratio, higher ignition pressure load, high-pressure injection system with its inherent drive function, and the cold starting system.
Specific measures are taken to increase the compression temperature in the interest of smooth and efficient cold starting and warming-up behavior. On smaller
54
D. Gruden et al.
Fig. 7. Diesel direct-injection process
engines in passenger cars, glow plugs extending directly into the combustion chamber have become the technology of choice. Flame starter units heating up the intake air are used mainly in utility vehicle engines. On large Diesels the starting process is ensured by compressed air moving the pistons until they start to run under their own power. The trend in engine design and configuration on road vehicles points clearly towards a reduction of weight accompanied by an increase in peak pressure. The development and improvement of the materials used is a major factor in this context, with light alloys being used in passenger car engines for the cylinder crankcase to reduce the weight of the engine. New materials are also used for bearings as well as for components subject to substantial thermal loads, such as the turbocharger, etc. The lubrication and cooling system must be specially adapted to the specific running conditions of the Diesel engine in the interest of enhanced reliability. A point worth mentioning is that the lubricant must, where necessary, be modified to meet the more demanding running conditions with a Diesel engine, for example, with Diesels running on heavy oil, and the contamination of the lubricant caused by solids (particles). 3.1.3 Charge Cycle and Turbocharger Technology
The cylinder charge must be replaced after each operating cycle by fresh air following the actual process of combustion. On Diesel engines in road vehicles this change in the cylinder charge follows the four-stroke principle. On large Diesels running at low speeds the two-stroke principle, with unidirectional gas exchange, has proved to be the method of choice since this significantly increases engine output, while keeping the design and configuration of the engine relatively simple and straightforward. Four-stroke diesels are controlled by valves, the valve overlap on a Diesel engine being kept to an absolute minimum by the start-up process and the specific design of the combustion chamber within the pistons themselves. On directinjection Diesels, with a swept volume of up to approximately 1.5 liters/cylinder, the intake port(s) (see Fig. 8) is/are used to generate the desired swirl effect. Here it is preferable to vary the swirl effect by switching on or switching off the intake port as a function of the engine control map.
Power Units for Transportation
55
Fig. 8. Generation of the swirl effect by the specific design of the intake port
On turbocharged Diesel engines [3] the charge cycle is influenced by the specific turbocharger technology used. Turbocharging significantly increases the supply of fresh air to the cylinders, ensuring an appropriate increase in output over a normally aspirated engine with virtually no increase in engine size. This explains why nearly all Diesel engines today are turbocharged, with the added benefit that this technology not only means more power, but advantages in noise management and a reduction in fuel consumption. The standard technology is exhaust gas turbocharging together with an intercooler (Fig. 9). To optimize the turbocharging effect, particular attention is given to the configuration of the inlet manifold/ports (cross-sections, volume, etc.) as well as the exhaust gas ports. Particularly in Diesel engines driving road vehicles, the response of the turbocharger system is crucial to the engine’s transient running behavior. To improve this response, most manufacturers use exhaust gas turbochargers with variable geometry (VNT) (see Fig. 10). This variability ensures not only a more rapid rampup of the turbine when accelerating under transient conditions, but also a higher standard of turbocharger efficiency under full load. This allows better tuning of the engine with high torque at low speeds and greater fuel economy at high speeds. Variable turbocharger also provides advantages in the management and reduction of exhaust gases thanks to more flexible exhaust gas recirculation.
56
D. Gruden et al.
Fig. 9. Exhaust gas turbocharging with intercooling
Fig. 10. Full load curves with variable turbocharger geometry
The following specific turbocharger technologies may provide advantages in certain niche applications: – Two-stage turbocharging to increase engine output and reduce fuel consumption even further. – Turbo-compound charge technology serving to make additional use of exhaust gas energy in a low-pressure turbine ensuring a higher standard of efficiency (used mainly on engines running under stationary conditions). – The comprex principle with fresh air being compressed by pressure waves in the exhaust gas generated by a cell wheel. This method has, however, not succeeded in achieving, let alone exceeding, the benefits of exhaust gas turbocharging.
Power Units for Transportation
57
3.1.4 Fuel Injection Systems
The injection system on a Diesel engine serves a number of purposes: – To dose the amount of fuel injected as a function of the current operating point and conditions (quality control). – To inject fuel at exactly the right time during the operating cycle. – To form an appropriate fuel/air mixture by way of atomization and local distribution of fuel in the combustion chamber. The three most important subsystems within a fuel injection system are as follows: – The low-pressure system – made up of the tank, fuel delivery pump, filter, pipes, etc. – The high-pressure system – building up fuel pressure and conveying this high pressure via the injector into the combustion chamber. – The electronic control unit – determining the right amount and timing of fuel injected as a function of engine running and environmental conditions. It is particularly the high-pressure system, where a wide range of different fuel injection technologies has developed, as a function of specific requirements and applications [4] (Fig. 11), which is of significance with the Diesel engine. These injection systems are as follows: – The inline pump – with a separate pump element for each cylinder, with the injectors connected by a high-pressure pipe. – The distributor pump – with central pressure generation and a distributor system rotating from one engine cylinder to the next. Following the axial principle, these pumps operate at an injection pressure of up to 1,000 bar, however thanks to their operating principle, radial pumps achieve almost twice this injection pressure. – The unit injector – without a connection pipe between the pump and the injector itself, meaning that each engine cylinder forms one unit driven preferably by the engine camshaft. – The insertion pump – usually housed in the cylinder crankcase and driven by the engine camshaft, with a very short pipe to the injector in the cylinder head. – The common rail (CR) system – with a high-pressure pump supplying fuel to a pressure reservoir connected to the injectors by separate pipes. Electrical pulses (transmitted to electromagnetic, or piezo-controlled injectors valves) are able to initiate the injection process with a very high degree of flexibility and accuracy, achieving a positive effect on the fuel/air mixture and combustion process as well as the final treatment of exhaust emissions. The CR system also offers the advantage of pressure being built up in the reservoir more or less independently of engine speed, allowing free choice of injection pressure as a function of engine map management. Particularly on vehicle engines (see Fig. 12), this provides the option to build up sufficient injection pressure for high torque also at low engine speeds.
Fig. 11. Injection systems
58 D. Gruden et al.
Power Units for Transportation
59
Fig. 12. Pressure characteristics of various fuel injection systems
Starting in 1998, CR technology has helped to significantly increase the performance of Diesel engines in passenger vehicles.Additional functions, pre- and post-injection as well as multiple injection provide a significant potential for individually optimizing the injection curve, reducing noise and exhaust emissions from the start. The CR system presented in Fig. 13 allows simple configuration and operation of the high-pressure pump on the engine with lower drive torque peaks and free choice of the number of cylinders. This relatively new technology has a significant potential for further development. Volume control on the high-pressure pumps will help to reduce fuel consumption, piezo-injectors are able to control the injection process more precisely and flexibly. 3.1.5 Injector Support and Injector Nozzle
The injector is the single most important unit crucial to the fuel/air mixture formation. It atomizes and distributes the fuel throughout the combustion chamber. Being fitted directly in the cylinder head, the injector is exposed to substantial pressure and temperature in the combustion chambers. Efficient cooling is therefore required to limit the injector’s operating temperature, and ensure that the fuel will not coke in the injector nozzles so crucial to fuel/air mixture formation. Two types of fuel injectors (Fig. 14) have proved most successful in the market, depending on the specific diesel combustion principle:
60
D. Gruden et al.
Fig. 13. Common rail system. 1 Fuel tank, 2 Filter, 3 Primer pump, 4 High-pressure pump, 5 Pressure-release valve, 6 Pressure sensor, 7 Fuel rail, 8 Injectors, 9 Sensors, 10 Electronic control unit
a
b
Fig. 14 a, b. Different types of injectors. a Throttling pintle nozzle; b Multi hole nozzle
Fig. 15. Electronic engine management system
Power Units for Transportation
61
62
D. Gruden et al.
– The throttle pin nozzle on chamber engines, where the effective cross-section of the injector is determined by the shape of the throttle pin. – The multi-hole injector on direct injection engines. Hole geometry is of utmost significance to the atomization of fuel. By improving the production of injector nozzles and using new technologies (switching over, for example, from drilling to spark eroding or a hydraulic erosion process rounding off the nozzles, etc.), thus making the nozzle smaller (down to approximately 0.1 mm) and increasing injection pressure in the process. The injector nozzle has become the single most important factor in improving the process of fuel/air mixture formation and reducing particulates dramatically. Without electronic engine management (Diesel control) the many functions expected of a Diesel today, as well as the high standard of comfort demanded, would not be possible.As indicated in Fig. 15, the basic functions of conventional mechanical injection systems have been supplemented by a wide range of additional functions [5]. The major benefits of electronic systems are; the high standard of precision in engine management, the option to combine numerous parameters as well as engine and vehicle systems with one another, the freedom in setting the engine to specific operating conditions shown in the engine control map, and the lasting consistency of such electronic systems. In the evolution of modern Diesel engines, the development of these functions and the application of these systems is of great significance. In general, injection systems show an ongoing trend towards higher injection pressure and more flexible injection processes. This must be seen in the light of the greater demands now made of modern power units in terms of both precision and emission control. 3.2 Current Status of Modern Diesel Engines and Future Trends
In the area of transportation Diesel engines are used mainly for – driving passenger cars (personal transport), – driving utility vehicles (trucks and buses), and – driving ships (marine Diesels). The following table presents the current standards in terms of cylinder capacity, specific output, engine running speed, optimum efficiency and the specific power-to-weight ratio (Table 1). Table 1
Cylinder capacity (Single) [Liter] Specific output (kW/liter) Max speed (rpm) Effective degree of efficiency (%) Specific power-to-weight ratio (kg/kW)
Car engines
Utility vehicle engines
Two-stroke marine Diesels
0.5 50 4000 42 1.4
2.0 30 2000 45 approx. 3
up to 1800 3 100 53 approx. 30
Power Units for Transportation
63
While these standard figures reflect the current status of the Diesel engine, they cannot cover the entire range of applications, particularly in the case of large Diesels characterized by the wide range of designs and configurations. Ultimately it is always the customer (user) who decides on the appropriate drive system, taking a wide range of criteria into account such as packaging, output and performance, weight, the cost (investment) involved in purchasing the vehicle, its service life, cost of ownership, etc. Figure 16 presents the most important milestones in the development of the Diesel engine.Assuming a steady increase in worldwide transport in the years to come, the Diesel engine, as shown in Fig. 1, will have an over-proportional share in the market place thanks to its unparalleled fuel economy. Based on some Diesel engines already in existence, the following presents the current status of development of modern engines currently in use today. 3.2.1 Passenger Car Diesel Engines
In a direct comparison with the spark-ignition engine, the Diesel engine used in the passenger car has developed in recent years from a less powerful and rather loud drive technology into a refined and dynamic drive system, thus becoming a genuine alternative equal in its qualities to the spark-ignition/petrol engine. All modern Diesel engines are direct-injection power units with highly developed injection systems (in most cases using common rail technology) and variable turbochargers. Introduced in 1999, the vast proportion of direct-injection Diesel engines are sold in the midrange market sector, however they are now becoming more acceptable in the luxury (with the first “three-liter car” being introduced in the year 2000) and even performance segment of the market. With the share of pre-combustion engines in new car registrations having dropped dramatically, as seen in Fig. 17, reference should be made here to a typical pre-chamber Diesel engine in passenger cars, and to the four-valve series 600 power units built by Mercedes-Benz [6]. Featuring five cylinders displacing a total capacity of 2497 cc, this power unit develops maximum output of 110 kW or 150 bhp at 5000 rpm with the help of turbocharger and intercooler technology. With normal aspiration, this engine series features some interesting technical solutions such as a register resonance charger, intake pressure control, etc. An outstanding representative of swirl-chamber engine technology is the sixcylinder power unit with turbocharger and intercooler [7] introduced by BMW in 1991. Displacing 2500 cc, this engine develops a maximum output of 105 kW or 143 bhp at 4800 rpm. The modified swirl chamber combustion principle with a V-shaped piston crown in conjunction with the turbocharger, exhaust gas recirculation, oxidation catalyst and Digital Diesel Electronics gives this power unit its particular exhaust emission control concept. With its modern configuration and design, this engine also ensures very good acoustics and vibration management. An additional, temperature-controlled capsule reduces noise emissions when starting the engine cold and in the warm-up phase to an even lower level. Direct-injection Diesel technology made its breakthrough into the passenger car market through the introduction of innovative injection systems providing
Fig. 16. Milestones in the development of the Diesel engine
64 D. Gruden et al.
Power Units for Transportation
65
Fig. 17. Share of direct-injection engines in the passenger car Diesel engine segment
an injection pressure of more than 1500 bar and allowing additional functions such as pre-injection. The technologies worth mentioning here are the radial piston pump, the unit injector and, in particular, common rail (see the chapter on injection systems). The pioneers in direct injection Diesel engines for the passenger car were Fiat (1987 in the Fiat Croma) and Audi (1989). Today particularly European car manufacturers have a wide range of directinjection Diesel engines extending from the smallest three-cylinder power units (such as the Smart Diesel engine displacing 0.8 liters and developing maximum output of 30 kW/41 bhp) all the way to the most powerful luxury performance cars (with V8 and V10 Diesel engines built by AUDI, BMW and Mercedes-Benz). Without doubt, the reason for building such Diesel-powered passenger cars is their low fuel consumption, Fig. 18 clearly shows how Diesel cars dominate the market of very low consumption models. As a typical representative of the new generation of direct-injection Diesels, BMW’s four-cylinder diesel [8] displacing 1.95 liters and developing a maximum output of 100 kW/136 bhp at 4200 rpm has all the features of such a future-oriented power unit. Figure 19 presents a longitudinal and cross-sectional view of the engine. Standing out in particular is its integrated intake manifold with intake ports coming in from above (swirl port) as well as the filling port entering the cylinder head from the side, also the radial-piston distributor pump (VP44), with increment cams intentionally slowing down the combustion process in the initial phase. Figure 20 illustrates the full load curves and the fuel consumption control map of BMW’s four-cylinder direct injection engine which is equipped with a variable geometry turbocharger (VNT) giving the specific model involved, the BMW 320d, particularly dynamic and agile performance. It should be noted that leading car manufacturers use mainly common rail direct-injection engines, since it is agreed that this system has the greatest potential for ongoing development.
66
Fig. 18. Passenger car fuel consumption
Fig. 19. Longitudinal- and cross-section
D. Gruden et al.
Power Units for Transportation
67
Fig. 20. Full load operation points and fuel consumption map
3.2.2 Utility Vehicle Diesel Engines
Utility vehicles serve to ensure safe and rational transport of goods and passengers. The range of such vehicles extends from small vans all the way to heavy trucks weighing 40 tonnes and more as used in the construction industry. Diesel engines are, without any doubt, the main type of powertrain for utility vehicles, offering unchallenged economy, environmental compatibility, a long service life and low-cost, easy maintenance. Increasing engine output and reducing fuel consumption were the main objectives in the ongoing evolution of such engines until the 1970s. In the 1980s, the reduction of exhaust emissions became the focal point of development, noise emission standards (see EC regulation 70/157/EWSG) also becoming stricter in the process. Significant progress has been made by modification of the engine and encapsulation of the engine compartment within the vehicle itself. Compared with technologies used in the 1980s, limited components in exhaust emissions, for example, are now down by approximately 80 percent. Engines used by all major manufacturers of utility vehicles are described in detail in the relevant literature [9, 10]. 3.2.3 Marine Diesels
Following a very wide range of different types of engine and engine configurations in the past [11], three types of marine Diesels have been dominating the ship market for more than 30 years: – High-performance marine Diesels running at high speeds (up to approximately 2000 rpm).
68
D. Gruden et al.
– Four-stroke engines with immersion pistons running at medium speeds (approximately 1000 rpm). – Two-stroke Diesels in cross-head configuration running at low speeds (approximately 100 rpm). The demands made of these engines are very good fuel economy, a long running life and a very high standard of reliability. Keeping down-times, due to maintenance, to an absolute minimum, with concepts ranging from preventive maintenance to maintenance only in the event of repair. In terms of exhaust emissions, new ships have been subject, since 1 January 2000, to the IMO regulations with maximum NOx emissions of <5 g/kWh. Another requirement is that marine Diesels must be able to run on low-quality fuel such as heavy oil, which demands the utmost from the lubricants used. Marine drive systems, with four-stroke engines running at high and medium speeds, require a reduction transmission in order to achieve the best propeller speed, while two-stroke engines running at low speeds are able to make do with direct propeller drive. Ultimately, therefore, the user of each vessel must decide which type of engine to use, since this decision depends on a wide range of different criteria. The highly charged two-stroke engine running at low speeds has the highest effective efficiency of all thermal engines with approximately 53 percent. Using the exhaust heat as well as the heat stored in the coolant, such engines are even able to achieve an overall efficiency rating of more than 80 percent. Detailed descriptions of such engines are to be found in the relevant technical literature. 3.2.4 Future Trends in the Use of Diesel Engines
Future trends in the use of Diesel engines include: – Increase in engine output in terms of both overall power and power concentration (specific output), combined with light-weight material technology; – Enhancement of overall economy by the ongoing reduction of fuel consumption; – Fulfillment of future emission standards (noise and exhaust emissions) including recyclability; – Increase in service life and reliability (including greater ease and lower cost of maintenance). The need to increase engine output and reduce fuel consumption while at the same time fulfilling current and future exhaust emission standards places increasing demands on the development engineer. Since specific, individual improvements alone are no longer able to meet these demands, there is a clear need for all-round concepts based in particular on the efficient use of electronics. Promising technologies of the future are the extension of functions by way of greater variability (engine map-specific control) or the interaction of vehicle systems (drive engine, transmission, car stability systems, energy recycled from the brakes, etc.).
Power Units for Transportation
69
Taking modern high-performance car engines as an example [12], we see this emerging area of modern production technology and new, even greater potentials in future. Top models already offer a specific output of 50 kW or 68 bhp/liter, and a new high-end power unit with bi-turbo technology built in small numbers on the basis of a 3.0 liter production engine (61 kW or 83 bhp/liter and 170 Nm or 125 lb-ft/liter) currently represents the cutting edge in this development. This elevates the passenger car Diesel into a performance range so far exclusive to the high-performance spark-ignition engine, at the same time allowing a process of down-sizing with a great potential for reducing fuel consumption. The following methods can be used to realize a further power increase: – Raising the charge pressure increases the air mass in the cylinder. To stay within the maximum permitted engine cylinder pressure limits (currently approx. 160 bar for passenger cars) an adaptation of the compression ratio is necessary (see Fig. 21); – A further increase of injection pressure (see Fig. 22) improves the ignition process and therefore the thermodynamic efficiency; and – Reduction of losses in the low pressure process through careful configuration of the turbocharger (see Fig. 23) offers considerable potential for power increases. A good example for these power increasing solutions can be seen on the impressive race engine [22]. The high-end and racing engine technologies presented above are currently not available for volume production purposes. The concepts presented in following table reveal the significant potential for extra output and torque still provided by appropriate modification of the base engine, with higher ignition pressure, highly flexible injection systems and charge systems with enhanced efficiency (Table 2).
Fig. 21. Potential of higher charge pressure
70
D. Gruden et al.
Fig. 22. Increase in engine output by higher injection pressure
Fig. 23. Comparison of turbocharger configuration – production engine versus racing engine
Table 2
Features
Volume model
High-end
Racing version
Engine capacity (liter) Specific output (kW/liter) Specific torque (Nm/liter) Compression ratio Injection system Injection pressure (bar) Turbocharger technology Max charge pressure (bar)
1.95 51.3 144 18:1 VP44 1750 VNT 2.1
2.925 61 170 16:1 CR 1600 Bi-VNT 2.3
1.95 90 205 15:1 VP44 1900 VNT 2.7
Power Units for Transportation
71
Transmission concepts with “short” setting-off ratios and “soft” converters when starting off support this trend. 3.3 Fuel Consumption
Current passenger car direct-injection Diesel engines have the highest efficiency of all drive concepts. When running at their optimum point, such engines offer specific fuel consumption of approximately 200 g/kWh, equaling an effective degree of efficiency of approximately 42 percent. The comparison of CO2 emissions of spark-ignition and Diesel engines with the same output, as presented in Fig. 24, is based on the direct-injection Diesel engine.As can be seen from this comparison, CO2 emissions with a Diesel engine are 25 percent lower than with a spark-ignition power unit. Presenting further potentials for reducing fuel consumption, Fig. 25 analyzes the energy required by a 1400 kg vehicle developing 100 kW/136 bhp maximum output in the European driving cycle. This analysis shows that optimization of the engine alone only provides part of the future potential. The rest must be ensured by appropriate concepts within the overall engine package, such as down-sizing, start/stop systems, etc. A further option for reducing fuel consumption is to decrease the loss of power. Figure 26 shows how fuel consumption can be reduced noticeably in the mean term by making the transition from pressure-controlled to volume-controlled common rail systems. The examples chosen do present a potential of approximately 5–10 percent provided by carefully optimizing and enhancing subsystems and individual engine units. It is however, becoming, increasingly important to consider the overall system in order to achieve an even higher standard of fuel economy in future with down-sizing, new transmission concepts or even hybrid drive systems with energy recycling allowing further progress.
Fig. 24. Comparison of CO2 emissions – spark ignition versus Diesel engine
72
D. Gruden et al.
Fig. 25. Energy uptake in the MVEG test
Fig. 26. Reduction of fuel consumption by volume-controlled generation of high pressure
Power Units for Transportation
73
By and large, these points apply to all other types and applications of Diesel engines [13]. 3.4 Exhaust Emissions
Combustion engines generate exhaust emissions resulting from the chemical reaction of air and fuel. As described in detail in the chapter “Measures for the reduction of exhaust gas emission”, HC, CO, NOx and particulates emissions are determined in representative test program (in Europe, for example, in the EC driving cycle for passenger cars). Legislation already resolved leading all the way to EU4 requires a drastic reduction in road vehicle emissions. The EU3 emission standards already imposed significantly reduce road traffic emissions in general – with both passenger cars and utility vehicles – leading to the less overall volume of emissions after the year 2000 as in the period prior to 1970, with far fewer cars on the road at the time. Air quality forecasts [15] show that after introduction of the EU3 standard for road traffic there is hardly any reason to expect any infringement of the World Health Organization standards, which therefore will hardly be exceeded any more. Introduction of the EU4 standard halves these limits once again, although the improvements are now relatively small compared with EU3, due to the extremely low absolute ratings already achieved. The principle of heterogeneous fuel/air mixture formation (see the chapter on fuel/air mixture formation) and the lean-burn principle already reduce the level of untreated HC and CO emissions from the Diesel engine to a very low point. The development of emission control technology, therefore, concentrates mainly on NOx and particulates. The generation of nitric oxides during reaction processes within the combustion chamber is favored by the level of temperature and the local supply of oxygen. Reaction processes change in terms of both time and space during combustion, meaning that all modifications serving to reduce temperature (e.g., intercooling) and the local oxygen content (e.g., exhaust gas recirculation) also reduce NOx. The generation of soot particles is mainly attributable to a local lack of oxygen. Reacting quickly and efficiently, hydrogen released from the fuel molecules burns more quickly than the carbon atoms forming larger carbon molecules and leaving the engine as soot particles separated in the exhaust emission test by a defined filter and subsequently analyzed for their properties. A soot particle is made up largely of carbon possibly forming compounds with hydrocarbons (e.g., PAHs), sulfates (resulting from the sulfur content in the fuel), water and metal compounds. The mass of particles is currently determined gravimetrically by weighing the filter samples obtained under partial flow conditions during the exhaust emission test cycle. The soluble organic and inorganic share (including the sulfate content) can then be extracted from the filter samples, the final product remaining on the filter being almost exclusively soot (carbon).
74
D. Gruden et al.
The share of fine dust caused by traffic [16, 21] in Germany is less than 10 per cent of the total volume of dust emissions. Examinations by the German Federal Office of the Environment nevertheless show that Diesel particle emissions have a greater influence on the concentration of dust hovering in the air around the average citizen than, say, emissions generated by power stations. Precisely this is why the German Federal Office of the Environment calls for a reduction in Diesel engine emissions to the same level as with spark-ignition engines. Engine manufacturers are therefore working all-out on the particularly effective Diesel particle filter, seeking to enhance this technology to production level while maintaining the superior fuel economy of a Diesel engine. In the year 2000 a French car maker introduced a Diesel particle filter with a regeneration process supported by a fuel additive on its top model. The main approaches taken in reducing Diesel engine exhaust emissions can be subdivided into specific, individual areas: – modifications within the engine itself, – after-treatment of exhaust gases, – improvement of fuel quality. Modifications within the engine itself and the treatment of exhaust gases are considered in greater detail in the following, while the improvement of fuel quality [13, 17], which also has a very significant potential, is described in Chapter “Fuels”. 3.5 Engine-Internal Measures for Reducing Exhaust Emissions
Optimization of the fuel/air mixture and the combustion process represents a significant step in reducing exhaust emissions. The combustion of a Diesel engine is determined by the following parameters: a) constructive (fixed) parameters such as the compression ratio (e), stroke volume (VH; S; D), the shape of the combustion chamber, the position of the injectors, the injector geometry (hole diameter, number of injector holes), etc., as well as b) variable parameters (adjustable operating parameters), such as the degree of charging (turbocharger level), air motion (swirl), injection pressure, timing of the injection process, injection rate, etc. Some of the variable parameters may be chosen freely and are determined on modern Diesel engines by the electronic engine management as a function of the engine’s operating points. The influence of both fixed and variable parameters on an engine’s output, combustion, noise and emission behavior is very complex, individual parameters often acting against one another (Fig. 27). The process of optimizing an engine and the various parameters therefore increases exponentially with the number of factors involved. New simulation and optimization processes, in some cases using neuronal networks, open up new potentials in this area and provide options so far not available.
Power Units for Transportation
75
Fig. 27. Different trends with exhaust gas components
3.5.1 Development of the Combustion Process
Development of the combustion process in order to minimize untreated emissions while maintaining the greater fuel economy typical of a Diesel represents the most important task and requirement for the future. Effective solutions focus on the injection system as well as air motion and the design/configuration of the combustion chambers, combining to provide the biggest step towards a further reduction in emissions. Diesel engines with subdivided combustion chambers (pre-chamber and swirl-chamber engines) have a high level of kinetic mixture-formation energy in the air flowing into the ancillary chamber. The fuel jet flowing into this hot and turbulent air is processed very quickly, with combustion starting right away. The limited amount of oxygen in the ancillary chamber then slows down the combustion process (which is why this is called the first stage in combustion). The increase in pressure then forces the mixture into the main combustion chamber where combustion in air with a large share of oxygen continues in the second stage – which is why this is referred to as a two-stage combustion process. The ratio between the chamber volume and the main combustion chamber, the geometry of the transition sections and the shape of the combustion area within the pistons have a significant effect on the process of fuel/air mixture formation and combustion. With the combustion process in chamber engines being controlled in this manner, the temperatures are not as high as with a directinjection engine, keeping NOx emissions to a relatively low limit. Now attempts are being made to achieve the same effects of controlled combustion also in the direct-injection engine. In the direct-injection engine with a powerful air swirl, the swirl effect acting on the fuel injected by the injectors into the combustion chamber is even greater in order to avoid any local lack of air (see Fig. 8). With cylinder diameter increasing, the intake swirl effect becomes smaller, since a larger number of injec-
76
D. Gruden et al.
Table 3
Injection pressure Injection starting late Injection hole geometry (Æ smaller) Injection rate
Fuel consumption
NOx
Particles
+ – – o
– + + +
+ – + +
tion holes also means a flatter combustion chamber. This is a concept experts refer to as air-distributed fuel dosage. The injection system is a key factor on the Diesel engine and is therefore also crucial to the engine’s emission behavior. Its functions in providing and dosing the flow of fuel are described in the appropriate chapters of this book. The main parameters crucial to exhaust emissions are injection pressure, injector hole geometry (diameter and length, as well as the hydraulic throughput), the actual timing of fuel injection as well as the injector rate. The following table presents these factors and their general effect in terms of fuel economy, NOx emissions and the formation of particles (Table 3): The table shows how these injection parameters act against one another. High injection pressure (Fig. 28), for example, ensures extremely good atomization of the fuel thanks to the use of smaller injector opening cross-sections and, as a result, low particulates emissions, while at the same time postponing the commencement of fuel injection with a dramatic reduction of NOx emissions. Lower emission of particulates is also conducive to the use of exhaust gas recirculation processes serving, in turn, to reduce NOx emissions. Pre-injection, on the other hand, serving to enhance noise control and management, may be counter-productive with a negative
Fig. 28. Influence of injection pressure on particle NOx emissions
Power Units for Transportation
77
effect on particulates emissions and exhaust gas recirculation. The volume tolerance limits for pre-injection are extremely tight if this technology is to be used without any disadvantages in terms of exhaust emissions.While new injection systems such as common rail technology already provide significant improvements in this respect, great efforts are still required in order to meet the EU4 standards. The gas exchange cycle has a significant impact on the emission of Diesel engines not to be underestimated in practice. Throughout their entire operating range, turbocharged engines have a much higher air surplus than normal-aspiration power units. In combination with intercooling this provides additional options in choosing the right configuration, with a particularly positive effect on exhaust emissions. Figure 29 presents the gas exchange system of a V8 passenger car Diesel [3] consisting of the following components: – – – – – – – – –
intake silencer, hot film air mass sensor, exhaust gas turbocharger compressor, intercooler, air collector, ports and valves, exhaust gas turbocharger turbine with variable geometry, pre-catalyst, underfloor catalyst.
Figure 29 presents also the exhaust gas recirculation system with its influence on the gas exchange cycle. The main parameters to be optimized in the interest of
Fig. 29. Charge cycle of a V8 Diesel engine
78
D. Gruden et al.
Fig. 30. Exhaust gas recirculation and charge pressure control operating areas
Fig. 31. Effect of variable swirl
Power Units for Transportation
79
minimum emissions are charge pressure and the intercooler, which are able, with the right dimensions and configuration as well as air cooling, to reduce the volume of emissions. As is to be seen on Fig. 30, map-specific charge pressure control provides a very wide range of positive effects. An additional option on four-cylinder engines is to appropriately control the swirl level, that is the mixture formation energy on the air side, in order to reduce untreated emissions to a minimum. Figure 31 shows that this can reduce NOx emissions by up to 30 percent under low loads, with the level of particle emissions remaining unchanged. This is done by using a flap closing the supply of air to the cylinder head filling port and, as a result, directing the entire flow of air through the swirl duct into the cylinder. There is still a potential for optimization by the continuous adjustment of this flap using a broader range of engine running conditions. 3.5.2 Exhaust Gas Recirculation
Exhaust gas recirculation (EGR) is the most effective way to reduce NOx on a Diesel engine. As indicated by Fig. 30, EGR reduces the air surplus with a fresh cylinder charge under part load to such an extent that NOx emissions decrease significantly. This effect results from the reduction of local oxygen and the corresponding delay in combustion, avoiding local temperature peaks. Excessive exhaust gas recirculation nevertheless increases HC and CO emissions as well as particulates. Maintaining a specific geometric configuration, the variable turbocharger as presented in Fig. 32 is particularly suitable for influencing the rate of exhaust gas recirculation. EGR cooling, as presented in Fig. 33, also boosts the EGR effect. Particular attention must be given in this context to the dynamic management of exhaust gas recirculation, since the EGR valve must be closed as quickly as possi-
Fig. 32. Influence of turbine blade position on the possible EGR rate
80
D. Gruden et al.
Fig. 33. Effect of EGR cooling
ble when accelerating all-out from low engine loads with a high rate of exhaust gas recirculation in order to avoid any undesired emission of smoke (see Fig. 34). Electronic engine management (Diesel control) and the fast actuators already available today (e.g., electric EGR valves) are able to precisely coordinate and harmonize the various control areas and systems, but involve much more complicated applications and processes. It is fair to say in summary that the injection system will remain the key parameter in the further development of combustion processes. The potentials of flexible common rail systems in terms of their injection curve (pre-injection, multiple injection), as well as variability of the injector itself (optimization of cross-sections), are still far from exhausted [20]. Control systems such as lambda control, as specified in Fig. 35, are required to provide more precise air ratio management, reducing tolerances and discrepancies in production units [17] and, as a result, cutting back the level of untreated emissions quite significantly. Innovative combustion concepts such as homogeneous diesel combustion also show encouraging results in this area.
Fig. 34. Dynamic set-up of exhaust gas recirculation
Power Units for Transportation
81
Fig. 35. Control concept with oxygen sensor
3.6 Exhaust Gas After-Treatment
With emission standards becoming stricter all the time, internal measures within the engine itself are no longer sufficient to fulfill exhaust regulations on the basis of untreated emissions alone. It is therefore fair to assume that new systems for the after-treatment of exhaust gases (see Fig. 36) will be required in specific areas.And since the Diesel engine, on account of its underlying principle, always runs in the lean-burn mode, new catalysts are required in order to reduce NOx emissions. The Diesel particle filter is very effective in ensuring the ongoing, dramatic reduction of particle emissions. The oxidation catalyst already used by all manufacturers of Diesel engines since approximately 1989 will support the various exhaust emission control concepts being introduced in the market. 3.6.1 Oxidation Catalyst
On account of its heterogeneous mixture formation the Diesel engine always runs with an air surplus, meaning that there is always sufficient oxygen in the exhaust gas to convert CO into CO2 and HC compounds into CO2 and H2O in an oxidation catalyst.A pleasant side-effect is that hydrocarbons in soot particles are also burnt in part in this process, reducing the mass of particles accordingly.
82
D. Gruden et al.
Fig. 36. Possibilities of exhaust gas after-treatment
The oxidation catalyst has become the method of choice in the meantime for EU2 and EU3 passenger car emission management concepts. Through their effect they make it possible to retard the injection process and, as a result, to optimize injection for lower NOx emissions. Their particular configuration (application of coatings and layers), in turn, is based on the following criteria: – – – – –
conversion behavior (for low exhaust emission temperatures), response in the warm-up phase, stability against aging, sulfur, etc., suppression of sulfate formation, temperature stability and long-term behavior.
Figure 37 shows the influence of catalyst layers on HC conversion. 3.6.2 DeNOx Catalyst
Since the three-way catalyst with lambda 1 control featured on the spark-ignition engine is not appropriate for the Diesel, the task facing engineers working on the lean-burn engine is to introduce the DeNOx catalyst. Like the DeNOx catalyst on lean-burn spark-ignition engines (see the chapter on spark-ignition engines), this technology also plays a key role in further reducing the emissions generated by a Diesel.
Power Units for Transportation
83
Fig. 37. Influence of catalyst coating
The following processes are being developed in order to reduce NOx through the after-treatment of exhaust emissions, although it is important to note that so far none of these technologies is suitable for general, unrestricted use in the passenger car: – NSCR (non-selective catalyst reaction) catalyst, – SCR (selective catalyst reaction) catalyst, – NOx storage catalyst. The NSCR catalyst uses unburnt hydrocarbons as the reduction agent in the following reaction process: 4 NO + 2 CH2 + O2 Æ 2 N2 + 2 H2O + 2 CO2
(a)
The catalyst uses zeolite coatings doped with copper or precious metals. The disadvantage is that the desired reaction takes place only within specific temperature windows (approximately 200–250°C) not large enough for practical operation of the vehicle. A further point is that results obtained under stationary test conditions which were in part quite satisfactory were not matched in the slightest under real-life driving conditions. With the additional disadvantage of such catalysts not providing appropriate temperature resistance and long-term behavior, the NSCR catalyst is not an attractive solution for the Diesel engine. The SCR process is applied to remove nitric oxides from exhaust gases coming out of large power stations. The reduction agent used for this purpose is ammonia (NH3) reducing the nitric oxides in the exhaust gases to N2. Since ammonia cannot be used for mobile applications, Siemens have developed the so-called SiNOx process using urea as the reduction agent.As shown in Fig. 38, this system requires a hydrolytic catalyst recovering ammonia from urea for the subsequent reduction of NOx. A further requirement is a dosage system providing the necessary supply of ammonia without allowing any NH3 to escape into the environment.
84
D. Gruden et al.
Fig. 38. Selective catalytic NOx reduction (SCR)
SiNOx systems are currently used in small numbers in utility vehicles where there is also a significant potential for further reducing fuel consumption since the SCR process allows the engine to be tuned for maximum fuel economy. The NOx storage catalyst (see Fig. 39) on the Diesel engine follows the same principle as with the lean-burn spark ignition engine. An aggravating factor in the Diesel engine control strategy is that the emission of particles in the short rich-burn phases (see Fig. 40) required for regenerating the storage catalyst also has to be taken into account. This technology would currently appear to be the most appropriate method for reducing NOx in the passenger car provided the fuel is sulfur-free. Even a tiny amount of sulfur such as 10 ppm residual sulfur resulting from the production process, however, requires a complicated desulfatization strategy in order to maintain the long-term function (see Fig. 41) of the storage catalyst. 3.6.3 Particle Filter
Particle filters apply the principle of retaining particles mechanically within the filter matrix. A typical particle filter system is shown in Fig. 42. To avoid any inappropriate increase in exhaust gas counter-pressure (and, accordingly, fuel consumption) in the filter it is essential to ensure regular combustion of particles, enabling the filter to regenerate in the process.
Power Units for Transportation
Fig. 39. Function principle of the NOx storage catalyst
Fig. 40. Sub-stoichiometric diesel operation
Fig. 41. Influence of sulfur on catalyst NOx efficiency
85
Fig. 42. Diesel particle filter system
86 D. Gruden et al.
Power Units for Transportation
87
The exhaust gas temperature of more than 500°C (930°F) generally required for this process can be reduced through the use of fuel additives, such as ferrocene, cerium, etc. Ancillary burners in the emission train are also being tested in particle filters used in city buses or utility vehicles. Large-scale tests already carried out with utility vehicles show that in principle this technology is suitable for practical purposes. Regeneration of the particle filter is a particular problem with the passenger car due to load curves generally remaining at a lower level. Only common rail fuel injection, offering the possibility of subsequent fuel injection, is able to increase emission temperatures to the level required for regeneration of the filter throughout a wide range of engine operation. In combination with suitable fuel additives, this is able to prevent any clogging of the filter even when running under very low loads and in cold weather. The electronic engine management must perform these additional functions dosing the supply of additive from the additive tank. Currently no manufacturers have production experience with systems of this kind proving their safe and effective operation throughout long-running periods. The current concept requires removal of ash deposits every time the car is serviced. The CRT (continuously regenerating trap) system shown in Fig. 43 represents a special type of particle filter combining an oxidation catalyst with a Diesel particle filter in one common process. In its operating principle, the CRT system is based on continuous oxidation of NO into NO2 in the oxidation catalyst, this NO2 then acting as a fuel additive to burn particles separated in the filter at relatively low temperatures of approximately 300°C or 570°F.
Fig. 43. CRT system
88
D. Gruden et al.
The combination of CRT and SCR technology allows simultaneous treatment of exhaust gases for CO, HC, particles, and NOx. 3.7 Exhaust Gas Concepts and Outlook
In the course of the last 20 years, development engineers have made great efforts to fulfill emission standards, as a result, a new vehicle today has only about 10 percent of the exhaust emissions generated by a vehicle built in 1990. Depending on the specific area of use and application (passenger cars, utility vehicles, marine engines, etc.), internal measures within the engine and exhaust gas after-treatment technologies are combined to provide suitable emission management concepts. In the passenger car segment, dominated by pre-chamber Diesel engines in the days of EU1 and EU2, the exhaust gas management concept typical today incorporates exhaust gas turbocharging with an intercooler, exhaust gas recirculation, an oxidation catalyst (underfloor) and electronic Diesel management. Ever since the introduction of EU3 for passenger cars registered as of the year 2000, passenger car engines have been direct-injection power units in virtually all cases. The exhaust gas concept in such advanced Diesel engines incorporates exhaust gas turbocharging with variable turbine geometry, four-valve technology and a variable swirl effect, high-pressure fuel injection with the injector positioned in the middle and common rail technology, electronic Diesel management, and an exhaust gas catalyst near the engine with improved coating. As of the year 2003, Diesel passenger cars in Europe will have to fulfill the EOBD (European On-Board Diagnosis) standard, Fig. 44 presenting the measures required for this purpose and their current state of development.
Fig. 44. Emission concepts for passenger car Diesel engines
Power Units for Transportation
89
Fig. 45. Measures for reducing emissions on large Diesel engines
Diesel engines in utility vehicles have so far relied mainly on improvements of the fuel injection system and modifications in the design of the combustion chambers in order to fulfill the standards required. In this case the trend will continue towards even higher injection pressure and low- or no-swirl combustion processes with multi-jet injectors (with a tight hole cross-section) and flat combustion chambers. Exhaust gas recirculation and the injection rate will also be part of the future exhaust gas concept for utility vehicle engines.After-treatment of exhaust gases [23] by way of combined SCR particle filter systems is required, for example, in order to comply with the strict US 2007 emission standards already proposed today. Marine Diesels, finally, are subject to the ISO 8178 standard applicable to engines not used in road vehicles (as well as construction and rail-bound vehicles, earth-moving machines, etc.). Figure 45 [11] presents possible measures for the reduction of NOx. Since fuel consumption has a very great influence on the overall economy of these engines, hardly any ocean-going ships so far have been fitted with NOx after-treatment systems. 3.8 References 1. Krieger K (1999) Das Einspritzsystem (The Injection System) Special Edition of MTZ – Ten Years of the Audi TDI Engine 2. Hiemesch O, Reibold G (1988) Möglichkeiten zur Begrenzung der Partikelemission bei PKW-Dieselmotoren (Possibilities of Limiting Particle Emissions in Passenger Car Diesel Engines). Paper presented at the Haus der Technik, Essen, Germany, February 1988
90
D. Gruden et al.
3. Steinparzer F, Neuhauser W (2000) Gestaltung des Ladungswechsels zur sicheren Einhaltung. Brüne, H.J der EU3-Emissionsgrenzwerte am Beispiel der neuen BMW Direkteinspritzmotoren (Configuration of the Charge Cycle to Reliably Maintain the EU3 Emission Standards, Taking BMW’s new Direct-Injection Engines as an Example). Aufladetechnische Konferenz 2000, Dresden, Germany 4. Paper published by Robert Bosch “Systeme für die Dieseleinspritzung” (Diesel Injection Systems). 1 987 709 359 5. Ertl C, Hiemesch O (1995) Elektronisches Motormanagement beim Diesel-PKW zur Erfüllung zukünftiger Anforderungen (Electronic Engine Management of the Passenger Car Diesel to Fulfill Future Requirements). International Vienna Engine Symposium 6. Conrad U, Feucht MJ Klingmann R, Krause R (1993) Die neuen Vierventil-Dieselmotoren von Mercedes-Benz. Mercedes-Benz’ New Four-Valve Diesel Engines. MTZ No 7/8 7. Anisits F, Hiemesch O (1991) Der neue BMW Sechszylinder (BMW’s New Six-Cylinder). Kratochwill H, Mundorff F (eds), Special print of MTZ in No 10/11 8. Anisits F, Borgmann K, Kratochwill H, Steinparzer F (1998) Der neue BMW-Vierzylinder Dieselmotor (BMW’s New Four-Cylinder Diesel). Special print of MTZ 9. Schittler M, Heinrich R, Kerschbaum W (2000) Mercedes Benz Baureihe 500 – eine neue V-Motorengeneration für schwere Nutzfahrzeuge (The Mercedes- Benz 500 Series – a New Generation of V-Engines for Heavy Utility Vehicles). ATZ 9 10. Harmeier R, Rieck G (2000) Die Trucknologie Generation TG-A der MAN Nutzfahrzeuge AG (The TG-A Trucknology Generation of MAN Nutzfahrzeuge AG). ATZ 9 11. Mollenhauer K (1997) Handbuch Dieselmotoren (Manual of Diesel Engines). Springer, Berlin, Heidelberg, New York 12. Borgmann K, Steinparzer F, Stütz W (2000) Potenziale und Grenzen moderner PKWHochleistungsdieselmotoren (Potentials and Limits of Modern High-Performance Passenger Car Diesel Engines). The Aachen Colloquium on Vehicle and Engine Technology 13. Zukunftsperspektiven des Verbrennungsmotors (Future Perspectives of the Combustion Engine). Special Issue 60 Years of MTZ / 1999 14. Berger H, Eichlseder H (1998) Das Abgaskonzept des BMW-Vierzylinders (The Exhaust Gas Concept of BMW’s Four-Cylinder Engine). MTZ 15. Metz N, Seiko M (1998) Die Luftqualität in Europa bis zum Jahr 2010 mit und ohne Euro IV-Grenzwerte (Air Quality in Europe up to the Year 2010 with and without the Euro IV Limits). 19th International Vienna Engine Symposium 16. German 1999 Annual Report (page 55). Federal Office of the Environment 17. Anisits F (1991) Der Kraftstoffeinfluss auf die Abgasemission von PKW- et al Wirbelkammermotoren (The Influence of Fuel on the Exhaust Emissions of Passenger Car Turbulence Chamber Engines). MTZ 18. Wieland J,Vent G,Wirbeleit F (2000) Können innermotorische Maßnahmen die aufwendige Abgasnachbehandlung ersetzen? (Can Measures within the Engine Replace Elaborate AfterTreatment of Exhaust Gases?). 21st International Vienna Engine Symposium 19. Allanson R (2000) The Use of the Continuously Regenerating Trap (CRT) on SCRT Systems to Meet Future Emission Legislation. 21st International Vienna Engine Symposium 20. Dürnholz M,Wintrick T, Polach W (2001) Potenzial der Einspritzung zur Reduzierung der Schadstoffemission von Dieselmotoren (Potential of Fuel Injection to Reduce Harmful Emissions of Diesel Engines). 22nd International Vienna Engine Symposium 21. Metz N, Resch G, Schönberger K, Steinparzer F (2000) Size Distribution and Characteristics of Soot Particles from Modern Diesel Engines. MTZ 22. Steinparzer F, Brüne HJ (1998) BMW Diesel in Pole Position. Haus der Technik, Meeting No E-30–315–056–8, March 23. Moser FX, Sams T, Cartellieri W (2001) Der Nutzfahrzeug-Dieselmotor unter Druck (The Utility Vehicle Diesel Engine under Pressure). 22nd International Vienna Engine Symposium
Power Units for Transportation
91
4 Alternative Propulsion Systems Dušan Gruden 4.1 Introduction
Ever since its invention, the 4-stroke reciprocating piston engine has been considered as a rather complex thermal unit which should better be replaced by far less complicated designs. Today, the internal combustion piston engine is looking back at a history of more than 125 years. During that time, there have been many initiatives to substitute piston engines for simpler and more efficient alternative powerplant units. The potential alternatives can be subdivided into two major groups: The first one consists of thermal power units, such as gasoline (Otto) and Diesel engines which use combustion to convert the chemical fuel energy into mechanical work. The best known representatives of this group are – – – – –
the 2-stroke engine, the Wankel engine, the gas turbine, the Stirling engine, and the steam engine.
The second group includes drive units designed to store energy or to directly convert chemical fuel (or mechanical) energy into electric energy, such as – electric batteries, – fly-wheel accumulators, and – fuel cells. As far as certain individual criteria are concerned, most of these drive systems perform better than the four-stroke combustion piston engine, so that, in the past, opinions about their chances as alternative passenger car powerplants were very optimistic (Fig. 1). However, when summing up all the properties required to smoothly operate cars over wide speed and load ranges and a long lifetime, these alternative concepts have never succeeded in edging the Otto and Diesel engines out of their top positions. 4.2 Thermal Engine with Discontinuous Combustion 4.2.1 Two-Stroke Engine
The thermal unit which yields the highest efficiency of more than 53% is the lowspeed two-stroke marine Diesel engine. This great performance combined with the relatively simple design of these cross-flow two-stroke engines has induced
92
D. Gruden et al.
Fig. 1. Previous prognosis about alternative powerplants. Source: Eaton Corp. 1973, Jet Prop.
Lab. 1975
engineers at all times to try and develop a two-stroke engine for automotive application. The last intensive research on the 2-stroke automotive engine concept was carried out by Orbital, an Australian company and ended with the conclusion, in the mid 1990s, that a 2-stroke engine will not be able to meet the demands on modern passenger cars in terms of power output, fuel consumption, NOx emissions and durability. 4.2.2 Wankel Engine
The rotary-piston Wankel engine is the best of the solutions found, when looking for other ways of realizing the Otto cycle. The Wankel concept solves the inherent mechanical problems of the piston engine in an astonishingly simple manner: no reciprocating pistons – and thus no vibration, no valves – and therefore less noise. However, the real challenge in combustion engine development is not in the mechanical configuration but in the combustion process itself. The fact is that
Power Units for Transportation
93
combustion is far easier to influence in a reciprocating-piston engine than in a rotary-piston Wankel engine, whose crescent-shaped combustion chamber with its big surface/volume ratio can be optimized only within rather tight limits. Since the Wankel engine was not capable of solving the urgent problems of exhaust gas pollution, of fuel consumption and durability, it finally disappeared from the gross production market [4]. 4.3 Thermal Engines with Continuous Combustion
With the gas turbine, Stirling engine and steam engine, continuous fuel combustion takes place in an external burner and not in the working medium itself. This configuration strongly facilitates the optimization of the combustion process when compared with a reciprocating piston engine. This is an advantage, which is highly appreciated and comes along with another welcome quality of such thermal engines: their multi-fuel capability. 4.3.1 Gas Turbine
It was in 1950 that a gas turbine was used for the first time to power a (Rover) passenger car. The investigation of gas turbine technology reached its climax in the 1970s. Its major drawbacks are the poor transient behavior and high fuel consumption. High efficiencies can only be achieved if the working medium has a very high temperature when it enters the turbine. With modern turbines, maximum inlet temperatures of Tmax=1100 to 1350 °C are realized.At such temperatures, fuel consumption is 30 to 80% higher than that of a gasoline engine. Under laboratory conditions, the following exhaust emission values are reached: HC = 0.16 g/km CO = 2.10 g/km NOx = 0.25 g/km
(a)
The manufacturing costs of a gas turbine are twice as high as those of a gasoline engine.According to current knowledge, gas turbines cannot be considered as appropriate substitute units for passenger cars. 4.3.2 Stirling Engine
The Stirling engine has been known since 1816. Ever since, it has been intensively investigated for its suitability for various applications. The greatest advantage of this concept is that it is noiseless. It is problematic, however, with regard to the sealing of the working medium – which can be either hydrogen or helium. Engine cooling and engine weight cause problems as well. More than 60% of the heat supplied must be dissipated via radiators. The fuel consumption of a Stirling engine is higher than that of a gasoline or Diesel unit.
94
D. Gruden et al.
The exhaust emission levels are as follows: CO = 1.20 g/km HC = 0.14 g/km NOx = 0.30 g/km
(b)
The manufacturing costs of a Stirling engine are 5 to 8 times higher than those of an Otto or Diesel unit. The Stirling engine, as it is today, is not being considered as a suitable drive unit for passenger cars. 4.3.3 Steam Engine
In the early 1960s, eight different steam engine concepts were investigated in the USA for their potential as future passenger car drive units [9]. The best known variant was the so-called Leer engine. Low exhaust emissions and noise levels, however, can only be obtained at the expense of high costs, additional accessories such as a burner, heat exchanger, condenser and various control systems, and a poor overall efficiency. Despite repeated and intensive testing, steam engines are not expected to enter the passenger car domain. 4.3.4 Common Characteristics of Continuous Combustion Engines
With all continuous-combustion engine concepts, the heat is generated in the burner and mostly fed to the working medium through the walls. Thus, it is upon the thermal strength of the wall material that the maximum temperature of the working medium depends. The maximum temperature tolerated by current materials is about 1250 to 1350°C. When compared with other thermal engines, reciprocation piston engines (gasoline and Diesel) offer an essential advantage: the heat is directly fed into the working medium (air) in the combustion chamber whose walls are cooled from outside with wall temperature being approx. 250°C. This allows much higher maximum cycle temperatures (higher than 2000°C) to be reached than in other thermal engines, where the temperature of the working medium is limited by the thermal properties of the wall material. The thermal efficiency hth of a heat engine, however, mainly depends on the maximum temperature level of working medium that can be reached (Fig. 2). The higher Tmax, the better is the thermal efficiency. Thus, the alleged drawback of the 4-stroke piston engine – that is the intermittent combustion process – actually turns out to be its greatest benefit. The internal combustion process with its continuous exhaust and refill permits high maximum combustion temperatures of the working medium to be achieved with comparatively low combustion-chamber wall temperatures. Therefore, the reciprocation piston engine with internal combustion will always have a better thermal efficiency and consume less fuel than heat engines with external continuous combustion.
Power Units for Transportation
95
Fig. 2. Peak combustion temperature, thermal efficiency and NOx emission
4.4 Electric Motor
The first electric motor-powered car was built in London, in 1886. At that time, its chances of success were supposed to be far better than those of the cars fitted with gasoline and Diesel engines which had just been invented. The first electric cars yielded a top speed of 106 km/h and had an operating range of about 80 km. By the mid-nineties of the 20th century, various efforts had been made to develop a competitive electric car, but despite intensive research and development, there is not much difference yet today between modern electric cars and their earlier ancestors: recent versions reach top speeds of vmax=120 km/h and their operating range is between 70 and 100 km. The main handicap of electric cars is the low energy density of the batteries used (Fig. 3). From the very beginning, lead-acid (Pb) batteries have been used whose energy density is about 30 Wh/kg. To yield the same amount of energy as a gasoline fuel tank with a capacity of 70 L and a weight of 57 kg, a lead-acid battery of almost 3000 L and 6.5 tons would be required. Other alternative batteries such as sodium-sulfur, nickel-cadmium, lithium-ion and nickel-metal-hybrid units are too heavy and too expensive. Their storage capacities and lives are insufficient and their charging times of 6 to 8 hours are excessively long. Electric cars cannot be considered as zero-pollution cars either, since exhaust emissions do occur during the generation of the electric energy required to operate them (Fig. 4). Therefore, the Californian CARB authority would actually have to revise its plans for the compulsory introduction of electrically powered cars under the name of “Zero-Emission Vehicles” (ZEV) in California [23].
96
D. Gruden et al.
Fig. 3. Fuel versus batteries (energy content corresponding to 70 L of fuel)
Fig. 4. Exhaust emission levels: gasoline and Diesel engines versus electric motor
4.5 Flywheel Storage System
The basic idea is a fascinating one: not to “destroy” – that is transform into heat – the kinetic energy, resulting from braking maneuvers, but to store it instead in an energy accumulator for transformation into different useful types of energy – or, in other words, to recuperate the brake energy. The most suitable devices for energy recuperation are flywheels, which rotate in a vacuum and transform stored mechanical energy into electric power. Systems of that kind are useful for vehicles, such as trains or busses, whose routes are predefined, which run on a lane of their own and whose driving con-
Power Units for Transportation
97
ditions and brake distances are known. For all other cars with random and arbitrary accelerations and decelerations there is no confirmed benefit from such concepts. 4.6 Outlook on the Future
During the last century, alternative drive systems were intensively investigated, but to date, none of those systems has been capable of pushing the 4-stroke piston engine out of its privileged position (Fig. 5). Surprisingly enough, it is not some exotic design but the good old reciprocating piston engine in its well known configuration, which will continue to dominate the market also in the foreseeable future. The reciprocating piston engine has displayed an astonishing potential for improvement and an extraordinary adaptability to all new demands placed upon modern cars. Figure 6 illustrates the forecasts for passenger-car drive-system development in the coming 15 to 20 years. As can be seen, further optimized versions of the gasoline and Diesel engines will continue to prevail in the automotive domain. From today’s point of view, hybrid and fuel-cell powerplants are the only alternative systems expected to gain a foothold in the automotive market. 4.6.1 Hybrid Drive
Frequently, the opinion is held that the best way to solve the problem of fuel consumption and of pollution from automotive emissions would be to fit vehicles with combined drive systems, consisting of electric motors, coupled with small highly efficient internal combustion engines (hybrid drives) (Fig. 7). The fuel economy potential of hybrid vehicles is summarized in the following three points: – Energy regeneration during deceleration (conversion to electrical power); – Electric motor propulsion in low-load conditions and as assistance in acceleration, enabling an internal combustion engine to run within the efficient regions of its operation; – Engine stop during vehicle stops. The hybrid vehicle system also has a high potential for emission control: – Reduced gas volume resulting in lower catalyst deterioration; – Engine stop and propulsion by electric motor to reduce HC emissions especially in lower load ranges; – Power assist by the electric motor during acceleration, allowing the air/fuel mixture to be controlled closely around Lambda=1 in all engine operating conditions by reducing engine load deviations; – Engine quick-start by high motor power avoids unstable combustion during starting.
Fig. 5. Status of alternative drive concepts
98 D. Gruden et al.
Power Units for Transportation
99
Fig. 6. Evolution of and forecast on future powerplant-systems
Fig. 7. Toyota hybrid system (THS) [29]
To date, however, mass production of this type of drive system has been hampered by its excessively high costs. 4.6.2 Fuel Cell
The fuel cell (Fig. 8) was invented as early as in 1839 by the Englishman William Robert Grove. It then was called the “galvanic gas battery”. The chemical reaction between fuel and air produces electricity with no mechanical energy involved. Hydrogen (H2) combines with oxygen (O2) producing water and electric current.
Fig. 8. Structures of a single fuel cell and fuel-cell stack
100 D. Gruden et al.
Power Units for Transportation
101
The core of each fuel cell is an electrolyte, that is a platinum-covered porous plastic foil. While hydrogen is fed to the anode, the platinum acts as a catalyst triggering the ionization of H2.At the cathode-end of the electrolyte foil, the platinum material incites the atmospheric oxygen to absorb electrons. The positive H2 ions and negative O2 ions combine to form water (H2O). Between the cathode and anode, an electric potential of approx. 0.6 to 1.2 V builds up. For this electric energy to be used, several fuel cells must be connected with each other, forming what is called a “stack”.An electric motor for automotive application would need hundreds of individual cells (more than 200). The first mention of fuel cells as potential alternative automotive drive units was made in the early 1960s. It took about 30 years for more intensive research to be finally started. Due to their theoretically high efficiency (Fig. 9) and low pollutant emissions, fuel cells are among the most promising alternative energy sources of the future. Of the various potential fuel cell configurations, only two are currently being investigated in the automotive industry (Fig. 10).
Fig. 9. Efficiency levels of electric-current production
Fig. 10. Fuel cell types and their characteristics
102
D. Gruden et al.
Fig. 11. The fuel cell system
With an operating temperature of approx. 80°C, PEM numbers among the socalled low-temperature fuel cells, whereas SOFC is what is called a high-temperature fuel cell with operating temperatures ranging between 800 and 1000°C. For the fuel cell to function properly, various secondary units are required besides the cathode, anode and electrolytes (Fig. 11). These secondary units are needed to provide the fuel cell with air and fuel and to cool and control it. The efficiency of the fuel cell improves if it is fed with pressurized hydrogen and air (1.5 to 3 bars). Less than 10% of the heat produced by a fuel cell is dissipated through the exhaust gas. For an assumed 40% overall efficiency, approx. 50% of the heat supplied must be dissipated through cooling. The required radiators and cooling power are 4 to 5 times bigger than with an Otto or Diesel engine. The exhaust gas of the fuel cell (water vapor) is cooled down in an exhaust gas condenser. Part of the steam is used to humidify the fuel-cell membrane: These membranes function properly only if moistened. The ideal fuel for fuel-cell operation is hydrogen produced from water by means of solar energy and solar cells (Fig. 12). However, it will probably not be possible to implement the technology required to create such ideal conditions before the middle of this century. Today, hydrogen is produced from mineral oil or natural gas with resulting CO2 emissions. Since the generation, transportation, storage, filling up and onboard storage of hydrogen causes certain problems, alternative fuels such as – methanol, – natural gas,
Power Units for Transportation
103
Fig. 12. Closed circulations, ideal automotive drive system e.g.: hydrogen
– gasoline, – ethanol, are also being examined for their capability to serve as hydrogen sources. From these fuels, hydrogen and CO2 can be obtained on-board in a so-called reformer. The required reforming temperatures range between 500 and 800°C. Reformers reach this temperature in 45 minutes to 2 hours. To coat the fuel cells and reformer, about 50 g of platinum per car will be needed. This is about 10 times the amount required for a three-way catalyst. To properly seal all the components of a fuel cell, about 1 km of gaskets is needed. At the time being, apart from technical problems, the commercial launch of the fuel-cell technology is also hampered by the high manufacturing costs of fuelcell-equipped cars and the enormous expenditure for building up the infrastructure required for fuel supply. Costs for the generation of 1 kW of power by means of fuel cells are said to vary between 5,000 and 30,000 Euro (Fig. 13) compared with 25 to 40 Euro for the same amount of power from a gasoline or Diesel engine. To be able to compete with internal combustion piston engines, the fuel-cell price for 1 kW of power should be lowered to 50 to 80 Euro – a target which might be reached by the end of this decade. Several car manufacturers have announced their intention of launching their first standard fuel-cell-powered cars as of the year 2004. In Germany, more than 200,000 fuel-cell cars are expected to exist in 2010.
Fig. 13. Evolution of fuel-cell costs
104 D. Gruden et al.
Power Units for Transportation
105
Fig. 14. Closed circulations, biomass as an energy source, e.g., biomass
4.6.3 Powerplants Using Alternative Fuels
Parallel to the search for alternative drive systems, engineers are also looking for alternative fuels.All those potential fuels are basically suited for conventional piston engines as well.When these fuels are put on the market, the conventional piston engines will also be adapted to suit their particular properties. Of special importance for future applications are fuels made from so-called regenerative raw materials (biomass) (Fig. 14). Via photosynthesis, solar energy is indirectly employed for the production of fuels: using the CO2 and H2O contained in the atmosphere, biomass is produced through photosynthesis. Biomass is an excellent raw material for the creation of various fuels such as alcohols and methyl esters. Power is then obtained by burning such biological fuels. The resulting products of combustion – that is CO2 and H2O – remain in the closed cycle of biomass production via photosynthesis. This method does not add to the increase of the atmospheric CO2. At the present time, this type of energy production is not yet a useful alternative, either technically or economically. But it is expected by many experts to gain increasing importance in the future.
106
D. Gruden et al.
4.7 References 1. Bertodo R, Smith P (1978) Prime Movers for Future Lift Trucks. SAE-Paper 780752 2. Eaton Corporation, 1973 3. Schorn N, Dürnholz M, et al (1993) Ist der Zweitaktmotor eine Alternative zum Viertaktmotor? VDI-Bericht 1066, Dresden 4. Fuhrmann E (1976) Entwicklungstendenzen beim Bau von Fahrzeugmotoren. Porsche AG 5. Wilson DG (1978) Alternative Automobile Engines. Scientific America 239:1 6. Should we have a new engine? An automobile power systems evaluation, vol I. Jet Propulsion Laboratory. California Institute of Technology, 1975 7. Nacamura H, Motoyama H, et al (1991) Passenger Car Engines for 21st century. SAEPaper 911908 8. Beeck A, Joos F, et al (1999) Fortschrittliche Gasturbinen Technologie. BWK 51, 5/6 9. Amann CA (1987) How shall we power tomorrow’s automobile? Automotive Engine Alternatives. Plenum Press, New York 10. Richey AE (1989) The Automotive Stirling Engine Programme. Final Report. SAE P-230 11. Downs D (1978) The Socially Acceptable Powerplants. Proceedings of Inst Mech Engineers 192 12. Buschmann G, Clemens H, et al (2000) Zero Emission Engine – Der Dampfmotor mit isothermer Expansion. MTZ 61:5 13. Pocrnja A (1982) Energieerzeugung aus Wärme. Erdöl und Kohle 35:5 14. Quissek F (1997) Fahrzeugantriebe der Zukunft – aus Sicht der VW-Forschung. Eurotax Nr. 1 15. Bentley JM, Teagan WP, et al (1997) Global Warming Impacts of Vehicle Propulsion Options. Arthur D. Little 16. Rügen macht e-mobil. VDI-Nachrichten Magazin 9/93 17. Japanese makers exhibit EV-efforts. Automotive News, 14 Dec. 1994 18. Battery price is key hurdle for development of electric. Automotive News, 23 December, 1996 19. Kobe G (1996) Battery of the Future. Automotive Industries 176:9 20. Schwungrad bringt Autos schnell auf Touren. VDI-Nachrichten, Nr 14, 2001 21. Cole AC (1997) Electric and hybrid Vehicles – a report on the SAE-Symposium. Automotive Engineer, September 22. Toyota’s Hybrid ready for California. Automotive Industries, May 1997 23. Calif. Turns ZEV mandate upside down. Automotive news, 29.01.2001 24. Tachtler J, Dietsch T, et al (2000) Fuel Cell Auxiliary Power Unit – Innovation for the Electric Supply of Passenger Cars? SAE-Paper 2000–01–374 25. DaimlerChrysler, Umweltbericht 2000 26. Frank MH, Welter, DL, et al (2000) PEM Fuel Cell System Solutions for Transportation. SAE-Paper 2000–01–0373 27. Zizelman J, Botti J, et al (2000) Solid-oxide fuel cell auxiliary power unit: a paradigm shift in electric supply for transportation. Automotive engineering. September 2000 28. Schon in der zweiten Generation. Automobil Revue, 2000 29. Harada J (2001) Development of a new hybrid vehicle. Auto Technology 1
The Handbook of Environmental Chemistry Vol. 3, Part P (2003): 17–41 DOI 10.1007/b 10456
The Global Cycles of the Naturally-Occurring Monohalomethanes David B. Harper, John T. G. Hamilton Microbial Biochemistry Section, School of Agriculture and Food Science, The Queen’s University of Belfast, Newforge Lane, Belfast, BT9 5PX E-mail:
[email protected]
In terms of atmospheric abundance and environmental significance the most important halogenated compounds formed in nature are undoubtedly the gaseous monohalomethanes chloromethane (CH3Cl), bromomethane (CH3Br) and iodomethane (CH3I). Despite the vast inputs of man-made chlorofluorocarbons into the atmosphere over the last 35 years, CH3Cl, present at an atmospheric concentration of 600 pptv with a total atmospheric burden of 5 Tg (5 million tonnes), is still the most abundant halocarbon in the atmosphere. Together with CH3Br it is responsible for around 27% of halogen-catalysed ozone destruction in the stratosphere. Until seven or eight years ago it was generally accepted that the source of the vast bulk of atmospheric CH3Cl and CH3Br was the oceans and that the sinks for these compounds were overwhelmingly abiotic. Recently, however, it has become apparent that most atmospheric CH3Cl and CH3Br has a terrestrial rather than a marine origin. Moreover, biological sources, formerly regarded as of peripheral significance, have assumed increasing importance in the atmospheric budgets. The belief that abiotic reactions are the only sinks for atmospheric CH3Cl and CH3Br has been displaced by a recognition that microbial degradation in soil may account for a large proportion of the atmospheric burden. In this chapter we review current knowledge concerning the global cycling and atmospheric budgets of CH3Cl, CH3Br and CH3I with particular emphasis on the biological sources and sinks. Keywords. Chloromethane, Bromomethane, Iodomethane, Methyl chloride, Global cycle, Atmosphere, Sources, Sinks
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
1
Introduction
2
Atmospheric Concentrations
3
Abiotic Sinks and Sources . . . . . . . . . . . . . . . . . . . . . . 20
3.1 3.2 3.2.1 3.2.2
Abiotic Sinks . . . . . Abiotic Sources . . . . Natural Sources . . . . Anthropogenic Sources
4
Biological Sources
4.1 4.2
Wood-Rotting Fungi . . . . . . . . . . . . . . . . . . . . . . . . . 24 Macroalgae . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
. . . .
. . . .
. . . .
. . . . . . . . . . . . . . . . . . . . 18
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
20 21 21 23
. . . . . . . . . . . . . . . . . . . . . . . . . . 24
© Springer-Verlag Berlin Heidelberg 2003
18
D.B. Harper and J.T.G. Hamilton
4.3 4.4 4.5
Phytoplankton and Marine Bacteria . . . . . . . . . . . . . . . . . 26 Higher Plants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 Animals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
5
Biological Sinks . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
5.1 5.2 5.3
Soil Microorganisms . . . . . . . . . . . . . . . . . . . . . . . . . 32 Marine Bacteria . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 Higher Plants and Fungi . . . . . . . . . . . . . . . . . . . . . . . 34
6
Atmospheric Budgets
6.1 6.2
Sources Versus Sinks . . . . . . . . . . . . . . . . . . . . . . . . . 35 Budget Modelling Using Isotope Ratios . . . . . . . . . . . . . . . 37
7
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
. . . . . . . . . . . . . . . . . . . . . . . . 35
1 Introduction The recognition that man-made halocarbons, primarily chlorofluorocarbons (CFCs), have caused depletion of ozone in the stratosphere and enhanced the greenhouse effect in the lower atmosphere has stimulated considerable interest in recent years in the global cycles of naturally-occurring organohalogen compounds which may also contribute to stratospheric ozone destruction. The monohalomethanes, chloromethane (CH3Cl), bromomethane (CH3Br) and iodomethane (CH3I), have come under particularly close scrutiny. Atmospheric CH3Cl concentrations at 600 pptv still exceed those of the most abundant CFCs, CFC-11 (CFCl3) and CFC-12 (CF2Cl2), which are present at concentrations of 275 pptv and 530 pptv, respectively [1, 2]. Atmospheric CH3Br, although present at much lower concentrations (ca. 9 pptv) than CH3Cl, is nevertheless of equivalent importance in terms of impact on the ozone layer, as bromine atoms are 60fold more effective than chlorine atoms in destroying stratospheric ozone [2]. Indeed, even today after 35 years of large inputs of CFCs with long residence times into the atmosphere, atmospheric CH3Cl and CH3Br, both of which are shortlived and predominantly naturally produced, account for around 27% of halogen-catalysed ozone destruction in the upper atmosphere [3]. This chapter will review current knowledge concerning the global cycles of CH3Cl, CH3Br and CH3I focusing particularly on the biological sources and sinks.
2 Atmospheric Concentrations Table 1 indicates the environmental concentrations, atmospheric burdens and residence times of the naturally-occurring halomethanes. The absolute concentrations of CH3Cl reported by Khalil and Rasmussen [1] in their analysis of global
19
The Global Cycles of the Naturally-Occurring Monohalomethanes
Table 1. Environmental concentrations, atmospheric burdens and residence times for natu-
rally-occurring monohalomethanes Halomethane
Atmos. conc. (pptv) Mean (range)
Seaweater conc. (ng L–1) Mean (range)
Atmos. burden (Gg)
Atmos. residence time (year)
Ref.
CH3Cl
550 a; 600 b (Marine: 500–620) (Terrest.: 550–1400) 9 (7–20) 2 (0.2–43)
5 (4–6)
5300 b
1.4 b
[1, 4–9, 11]
0.3 (0.05–1.2) 1.2 (0.2–7.5)
150 55
0.7 0.011
[6, 9, 12–18] [19–24]
CH3Br CH3I a b
Estimate based on Ref. [6]. Estimate based on Ref [1].
measurements of CH3Cl conducted over a 16 year period appear around 10 % higher than those reported by most other groups of investigators [4 – 7] suggesting that calibration differences between laboratories have yet to be resolved. Notwithstanding this discrepancy, studies consistently show a latitudinal variation in annually averaged atmospheric concentrations of CH3Cl in the marine boundary layer, with those in the tropics exceeding those at the poles by about 40 pptv [1]. In addition, a marked seasonal cycle is apparent with an amplitude of about 10% in the Northern Hemisphere with concentrations reaching a maximum in spring and falling to a minimum in autumn. Unfortunately, intensive atmospheric sampling has been confined almost entirely to sites representative of the marine boundary layer and only recently have CH3Cl concentrations in air masses associated with land been reported. Such studies indicate substantially higher concentrations at some coastal and continental locations in the tropics. Thus, CH3Cl mixing ratios of 950–1400 pptv CH3Cl were recorded off Okinawa and several other forested tropical islands in the W. Pacific and S.E. Indian Ocean [7–9] and up to 850 pptv were measured at inland sites on various continents [1]. Atmospheric CH3Cl concentrations appear to have changed little over the past century according to analysis of air trapped in polar firn (unconsolidated snow) in Antarctica and Greenland with an increase of no more than 5–10% in the atmospheric burden over this period [10]. Atmospheric CH3Br concentrations are about 60-fold less than those of CH3Cl, but are rather more variable. The differences observed cannot be accounted for purely in terms of seasonal fluctuations. There is a significantly greater abundance of CH3Br in the Northern Hemisphere compared with the Southern Hemisphere with an interhemispheric ratio of 1.2 to 1.4 observed [6, 25]. The presence of additional anthropogenic sources in the Northern Hemisphere and a larger oceanic sink in the Southern Hemisphere may be factors involved in this difference. Attempts to determine the atmospheric history of CH3Br by analysis of air trapped in polar firn have yielded rather perplexing results [10, 26]. Data from four sites in the Antarctic suggest an increase of 20–30% in the atmospheric burden of CH3Br over the last 50 – 100 years. However, samples from Greenland and Devon Island in the Canadian Arctic indicate concentrations of CH3Br near the
20
D.B. Harper and J.T.G. Hamilton
bottom of the firn profile up to 25-fold greater than those at the top.At Devon Island there is an almost exponential increase in CH3Br concentrations with depth. These findings taken at face value imply that atmospheric concentrations in the Northern Hemisphere 50 years ago were as much as 40-fold greater than those in the Southern Hemisphere. In view of the atmospheric residence time of CH3Br this seems highly improbable. Although locally elevated concentrations are conceivable, it does not appear likely that they would be enhanced to the extent observed in the remote areas where sampling was conducted. The most logical explanation is in situ production of CH3Br in the lower firn samples from the Northern Hemisphere probably associated with the presence of organic matter in the firn. Unlike the Antarctic, the Arctic is surrounded by vegetated landmasses, which must provide a strong source of organic matter, both dissolved and particulate, to the Arctic ice caps. Nevertheless, until a feasible mechanism is identified to account for the phenomenon and the probability of the process occurring in Antarctic samples assessed, there must be an element of doubt regarding the past atmospheric history of CH3Br. Atmospheric CH3I concentrations display much greater variability than those of CH3Cl or CH3Br with mixing ratios ranging over two orders of magnitude around a mean of about 2 pptv. Particularly high levels are associated with oceanic regions of high biomass productivity such as the upwelling of cold waters off the Peruvian coast, areas off the coasts of Iceland and South Africa [19] and sites close to macroalgal beds [27].
3 Abiotic Sinks and Sources 3.1 Abiotic Sinks
Table 2 presents the current best estimates of the total global sinks for CH3Cl, CH3Br and CH3I together with global atmospheric inputs of each from known sources. Both biological and abiotic processes are involved in emissions of halomethanes to the atmosphere and in their removal from the atmosphere, but abiotic processes appear on present evidence to dominate source and sink terms. Reaction with OH radical formed by photodissociation is the principal pathway for removal of CH3Cl from the atmosphere and is also probably the most important sink for CH3Br. As atmospheric OH concentrations are substantially higher in the tropics than at the poles, the CH3Cl lifetime for OH attack ranges from 0.8 years in the tropics to 12 years at the poles with a global mean of about 1.4 years. Since atmospheric CH3Cl concentrations are by comparison relatively uniform across latitudes, these differences imply that emissions of CH3Cl to the atmosphere must be very much greater in the tropics than at higher latitudes. On the basis of the OH sink and taking into account transport of CH3Cl to the stratosphere, Khalil and Rasmussen [1] estimated global emissions of CH3Cl to be 3.7 Tg year–1 of which 85% is released between 30°N and 30°S. The only other abiotic sink for CH3Cl likely to be of global significance is reaction with Cl radical in the marine boundary layer which could account for up to 0.4 Tg year–1 [1, 46].
The Global Cycles of the Naturally-Occurring Monohalomethanes
21
Table 2. Estimated total global sinks for atmospheric monohalomethanes and estimated global inputs to the atmosphere from known sources
Halomethane
Estimated global sink (Gg year–1)
Estimated atmospheric inputs from known sources (Gg year–1)
CH3Cl
4500–5500 a [1, 28]
CH3Br
205 b (112–454) [6]
CH3I
1000–2000 [22, 23]
Oceanic, 650 [11]; Biomass burning, 900 [29]; Fungi, 160 [30]; Salt marshes, 170 [31]; Wetlands, 50 [32, 33]; Coal combustion, 105 [34]; Incineration, 45 [34]; Industrial, 10 [34] Total: – 2090 Oceanic, 56 c [6]; Biomass burning, 20 [35]; Fumigation, 47 [36]; Automobiles, 1.5 [37]; Salt marshes, 14 [31]; Wetlands, 5 [32]; Brassica plants, 7 [38]; Fungi, 2 [39] Total: – 152.5 Oceanic, 150–1500 [19, 20, 40–42]; Rice plants, 90 [43]; Biomass burning, 8 [35] Total: – 250–1600
a
Higher estimate based on a soil microbial sink for CH3Cl accounting for a similar proportion of the atmospheric burden to that observed for CH3Br [44, 45]. b Current best estimate. c Gross production disregarding oceanic sink, net input is –21 Gg year–1
For CH3Br the mean atmospheric lifetime with respect to reaction with OH radical is estimated at 1.7 years with a range of 1.5–1.9 years [25]. However, in contrast to CH3Cl the ocean also acts as a significant sink for CH3Br although the loss rate varies by more than two orders of magnitude between polar and equatorial waters [47]. The major abiotic oceanic degradation pathways for CH3Br are hydrolysis and exchange with chloride ion with the rate of the latter reaction exceeding the former by between 5- and 10-fold. The best estimate of the ocean sink for CH3Br is 77 Gg year–1 [6]. The atmospheric lifetime of CH3Br with respect to these chemical losses is estimated at 1.9 years with a range of 1.1–3.9 years [16]. In the case of CH3I, reaction with OH radical is too slow to be of significance as an atmospheric sink. The dominant CH3I removal process is photolysis. Calculations based on the UV absorption spectrum yield an atmospheric lifetime for CH3I with respect to photolytic dissociation ranging from 4 days at the surface to 1.5 days at 10 km assuming a solar zenith angle of 40° [48]. 3.2 Abiotic Sources 3.2.1 Natural Sources
The oceans provide large inputs of all three halomethanes to the atmosphere. Indeed, until seven or eight years ago on the basis of measurements in 1983 [40] indicating supersaturation of surface seawater with these compounds, it was widely
22
D.B. Harper and J.T.G. Hamilton
believed that virtually the entire tropospheric burden of such halomethanes was derived from the oceans. However, in 1996 Moore et al. [5, 49] using data collected in the N.W.Atlantic and Pacific Oceans calculated that release of CH3Cl was considerably less than previously estimated and that the cold waters of higher latitudes beyond 50° are a net sink for CH3Cl.An analysis of measurements made by several groups of investigators between 1983 and 1996 [11] confirmed these conclusions and suggested a net global flux from the oceans of around 650 Gg year–1, no more than 14 % of the estimated total annual global emissions to the atmosphere. Abiotic reactions such as attack of Cl– on biologically formed CH3Br and CH3I [50] or dimethylsulfonium compounds [51] can probably account for the bulk of CH3Cl produced in the oceans [5]. Whilst the oceans also appear to act as a source of atmospheric CH3Br, emissions calculated at 56 Gg year–1 are principally confined to coastal regions and areas associated with upwellings [6, 12] where production is largely biological (see Sect. 4.2 and 4.3). Most of the ocean appears to be a sink for the halocarbon with both abiotic (see Sect. 3.1) and biological (see Sect. 5.2) degradation occurring. The current best estimate of the balance between aquatic production and degradation is a net oceanic CH3Br uptake of around 21 Gg year–1 [6]. Oceanic emissions overwhelmingly dominate the atmospheric budget of CH3I although the input to the atmosphere is poorly constrained.Whilst biological release by macroalgae and phytoplankton contribute to the atmospheric burden (see Sect. 4.2 and 4.3), photochemical processes are probably the main source of CH3I in the marine environment. Moore and Zafiriou [52] suggested that the most likely route is by recombination of methyl radicals produced by photolysis of humic materials and I atoms formed by photochemical oxidation of I– in seawater. Measurements conducted in the Greenland/Norwegian Sea and the tropical Atlantic indicating a correlation between light intensity and the extent of the CH3I saturation anomaly support this hypothesis [53] and imply that biological production is not the most important factor in determining CH3I concentrations in seawater. Further evidence of predominantly photochemical production in seawater is provided by recent work [54] demonstrating a correlation between surface seawater temperature and atmospheric CH3I concentrations at high, middle and low latitudes. An alternative abiotic source of CH3I from methyl cobalamin released by marine bacteria and phytoplankton has been proposed by Manley [55] who showed that this cobalt-containing corrin could react with I– or I2 in seawater to produce CH3I. Biomass burning can lead to the volatilisation of Cl–, Br– and I– as the corresponding halomethanes. Smouldering or low temperature combustion particularly of foliage is especially favourable to such release and can result in approximately 4% of Cl– and Br– in the fuel being converted to the methylated form [35]. As a result, substantial quantities of CH3Cl and CH3Br enter the atmosphere as products of savannah and forest fires and during slash-and-burn agriculture. Global emissions of CH3Cl by biomass burning are estimated at 900 Gg year–1 [29], the largest single source of atmospheric CH3Cl identified to date.Whilst the flux of CH3Br from this source (estimated at 20–30 Gg year–1) is substantially less [35], it nevertheless represents a significant proportion of the annual global input of CH3Br to the atmosphere. Although the conversion of I– to CH3I during
The Global Cycles of the Naturally-Occurring Monohalomethanes
23
Scheme 1. Formation of halomethanes from o-methyoxyphenols in soil [56]
biomass burning appears highly efficient (~ 40%), the low I– content of biomass limits the CH3I flux from this source, which at 7 Gg year–1 is trivial in terms of the global budget [35]. It is perhaps worth noting that, despite the fact that much current biomass burning is directly attributable to man, fires have been an integral part of many plant ecosystems for millions of years and hence a large proportion of emissions from this source must be regarded as natural. A novel abiotic route to halomethanes has recently been proposed by Keppler et al. [56, 57] who demonstrated halomethane production on incubation of soils rich in organic matter under acidic conditions (pH 5) with Fe3+ and halide ion. Release rates of CH3Cl, CH3Br and CH3I from grassland soil of 70, 230 and 2700 ng g dry wt–1 d–1 respectively were observed in the presence of 0.1 M solutions of the corresponding halide ions. The mechanism proposed involves reaction of Fe3+ with o-methoxy-substituted phenols, which are monomeric constituents of many humic substances in soil. Oxidation of such components by Fe3+ to form quinones is assumed to occur almost synchronously with nucleophilic attack on the CH3 moiety of the methoxyl group by halide ion. Estimates of global emissions by this mechanism are not yet available, but will obviously require extensive field measurements under a variety of conditions. Significant fluxes are likely from soils rich in organic matter in highly saline environments. 3.2.2 Anthropogenic Sources
If biomass burning is regarded as a largely natural source of CH3Cl, only a small proportion of the global input of CH3Cl to the atmosphere is anthropogenic in origin. Coal combustion and incineration are estimated to be responsible collectively for no more than 3 % of global atmospheric input, and release of CH3Cl utilized as an intermediate in industry only accounts for 0.2 % [34]. For CH3Br, anthropogenic emissions comprise a significant proportion of atmospheric input, most arising from the use of CH3Br as a fumigant for soils, various perishable and durable commodities and buildings [36]. Automobile emissions can also contain CH3Br, which is released during combustion of ethylene dibromide, an additive in leaded petrol [37]. Anthropogenic emissions of CH3I are insignificant.
24
D.B. Harper and J.T.G. Hamilton
4 Biological Sources 4.1 Wood-Rotting Fungi
Polypore fungi involved in the rotting of wood are amongst the most important biological sources of atmospheric CH3Cl identified to date and may also contribute to the atmospheric CH3Br burden. Biosynthesis of CH3Cl is widespread in the Hymenochaetaceae, a family of over 450 species of white rot fungi [30]. Over half of the 63 species examined from six genera in this family released CH3Cl when grown in the presence of Cl– [58]. The metabolic trait was particularly well expressed in Phellinus and Inonotus, widely distributed genera with characteristic bracket-like perennial fruiting bodies on trees of temperate and tropical forests. The proportion of Cl– volatilised as CH3Cl during fungal growth is not only species-dependent, but also varies with the growth substrate. Release of CH3Cl is usually maximal with cellulose-based substrates [59].Wood, the natural growth substrate for these fungi contains around 60% cellulose. Approximately two thirds of CH3Cl-releasing species are capable of converting more than 10% of Cl– present to CH3Cl with yields rising to as high as 80–90% with some species of Phellinus, for example P. pomaceus, P. ribis and P. occidentalis. The Cl– content of the growth substrate has relatively little effect on the proportion of Cl– converted to CH3Cl. Thus, the percentage of Cl– volatilised by cellulose-grown cultures of P. pomaceus declined only marginally from 90% at a Cl– concentration of 50 mM to 75 % at 0.5 mM [59]. Methylation of Br–, and also I– at concentrations below 1 mM, was almost as efficient as that of Cl–. When all three halide ions were present in the growth medium, the preferred order of attack was I–, Br–, Cl– so that three consecutive waves of halomethane production were observed reflecting sequential formation of CH3I, CH3Br and CH3Cl [59]. By making several assumptions regarding the growth of white rot fungi in their natural habitat, Watling and Harper [30] were able to extrapolate from laboratory data to the approximate magnitude of the global flux of CH3Cl to the atmosphere from wood-rotting fungi. The annual global input to the atmosphere was estimated at 160 Gg year–1 of which 75% originates from tropical and subtropical forest and 86% is attributable to the genus Phellinus. Probably the major area of uncertainty in deriving this estimate is the extent to which vegetative hyphae of white rot fungi ramify beyond the woody substrate with which the fruiting bodies are associated.Watling and Harper [30] made the conservative assumption that the hyphae do not proliferate in non-woody substrates. However, if the hyphae of such fungi extensively penetrate the leaf litter layer and the upper layer of the soil in forest ecosystems, the amount of Cl– available for conversion to CH3Cl would be very much higher than that present in wood which is a relatively low chloride matrix compared with soil and leaf litter. Global emissions would be correspondingly increased perhaps by as much as tenfold [30]. The first field measurements of CH3Cl fluxes from forest floors have recently been reported by Dimmer et al. [33]. Although the site (a conifer plantation in western Ireland) cannot be regarded as globally representative of temperate conifer for-
The Global Cycles of the Naturally-Occurring Monohalomethanes
25
est, either ecologically or perhaps more importantly in terms of soil Cl– content, an extrapolation of the observed emission fluxes led to an estimate of 85 Gg year–1 from such forests worldwide. By comparison Watling and Harper [30] estimated fungal emissions of 40 Gg year–1 from all temperate forest of which coniferous forest comprises about 70%. If the emissions measured by Dimmer et al. [33] from the forest floor are largely of fungal origin, the magnitude of the overall estimate of Watling and Harper [30] for global emissions appears to be of the correct order. The low concentrations of Br– in the terrestrial environment severely restrict CH3Br release by fungi even though methylation of Br– occurs preferentially to that of Cl–. Nevertheless, based largely on the assumptions of Watling and Harper, global emissions of CH3Br by wood-rotting fungi have been estimated at 2 Gg year–1 [39]. However, this calculation is based on very limited data regarding the Br– content of wood and must be viewed as highly tentative. 4.2 Macroalgae
The long-held belief that the oceans were the predominant source of atmospheric halomethanes led to a search for marine biological sources. Production of CH3I by kelp beds off S.W. Ireland was first established by Lovelock [60] in 1975, but it was not until 12 years later that Manley and Dastoor [27] conducted a detailed field and laboratory investigation of halomethane release by Macrocystis pyrifera (the giant kelp) – an important and often dominant primary producer off the Californian coast. On the basis of field measurements they estimated CH3Cl, CH3Br and CH3I emission rates of 160, 9 and 14 ng g wet wt–1 d–1, respectively. Later work [61] extended their observations on CH3I release to other kelp species (e.g. Laminaria farlowii) and a variety of non-kelp macroalgae including brown (e.g. Cystoseira osmundaceae), green (e.g. Ulva sp) and red (e.g. Pterocladia capillacea) algae. Mean production rates for kelp species were approximately eightfold greater than those for non-kelp species. These findings were broadly confirmed by Nightingale et al. [62] in an investigation of macroalgae collected off the west coast of Scotland. Itoh et al. [63] examined halomethane production by 44 species of temperate macroalgae and reported that two species, Paperfusiella kuromo and Sargassum horneri, released particularly large amounts of CH3I (514 and 96 ng g wet wt–1 d–1, respectively), but only trace amounts of CH3Br and no detectable CH3Cl. Macroalgal species from polar and subtropical waters have also been included in recent studies [64–66]. In a study of polar macroalgae from the Antarctic [65], the pattern of monohalomethane emission was similar to that observed by Manley and co-workers [27, 61] for temperate kelps, but the rates of production were 10–50-fold less. A comparative investigation of CH3I release by 29 species of macroalgae derived from various climatic zones [66] indicated that the mean rate of emission by polar species (0.03 ng g wet wt–1 d–1) was approximately half that of subtropical species examined. In light of this study it seems doubtful whether CH3I emissions of 34–20,400 ng g fwt–1 d–1 recorded by Ekdahl et al. [67] for three species of subtropical algae can be regarded as valid. The highest emission rate measured in this study was that of the red alga Falkenbergia hillebrandii. Significantly, emis-
26
D.B. Harper and J.T.G. Hamilton
sion rates reported by the above laboratory for other halocarbons such as trichloroethylene by this species [68] have not been reproduced in subsequent work elsewhere [69]. In recent work, Baker et al. [70] determined release of CH3Cl, CH3Br and CH3I by cultures of five temperate macroalgal species collected on the east coast of England including brown, green and red algae. Production rates of 0.1–3, 0.05–3 and 0.02–2 ng g fwt–1 d–1 were found for CH3Cl, CH3Br and CH3I, respectively. Despite the many uncertainties and widely varying production rates reported by different groups, Baker et al. [70] attempted to estimate global emission rates of these compounds by seaweeds. Production rates by temperate species reported in the literature were averaged and a latitudinal weighting was applied to account for differences in productivity between cold, temperate and warm water species. Total global macrophyte emissions of CH3Cl, CH3Br and CH3I were calculated at 0.14, 0.06 and 0.28 Gg year–1, respectively. Inevitably the errors involved in this approach are large since not only has a very small fraction of globally important macrophytes been surveyed, but free-floating macrophytes such as Sargassum spp which account for around 10% of macrophyte biomass are excluded from consideration.Additionally, the extent of halomethane production by coral reefs is completely unknown although they occupy an area equal to 10% of the coastal buffer zone compared with only 5% that supports macrophytes. Nevertheless, even if emission fluxes were a magnitude higher than that estimated by Baker et al. [70], production by macrophytes would not be significant in terms of the global halomethane budgets. 4.3 Phytoplankton and Marine Bacteria
Investigators had surmised for many years that phytoplankton could be involved in the production of monohalomethanes in the oceans. However, direct evidence of emissions was not obtained until 1995 when Tait and Moore [71] demonstrated release of CH3Cl by laboratory cultures of a wide variety of cold- and warm-water diatoms. Later work not only showed that production of CH3Br and CH3I also occurred, but extended these observations to other classes of phytoplankton including prymnesiophytes, dinoflagellates and microalgae [63, 72–75]. Biomass-normalized release rates exhibited wide variations between classes and species and appeared independent of nutrient status and the presence of bacteria. Biosynthesis was not directly associated with either photosynthesis or respiration, but in general occurred at a maximum rate during the stationary and senescent phases of the growth cycle [72]. This pattern of release led to the suggestion that halomethanes may be produced as a result of autolytic processes rather than as direct products of metabolism [72], but the detection of an Sadenosylmethionine:halide ion methyltransferase in laboratory-grown cultures of Pavlova gyrans argues against this interpretation [63]. Of all the species of phytoplankton examined for halomethane production, the two microalgal prymnesiophytes, Phaeocystis sp and Pavlova gyrans gave the highest release rates (50 ng CH3Cl g wet wt–1 d–1 and 20 ng CH3Br g wet wt–1 d–1 for Phaeocystis sp [72, 73] and 140 ng CH3Br g wet wt–1 d–1 and 170 ng CH3I g wet wt–1 d–1 for P. gyrans) [63]. Nev-
The Global Cycles of the Naturally-Occurring Monohalomethanes
27
ertheless, even in experiments with these species, overall concentrations in laboratory cultures did not exceed 1 nM after 30 days. Simple extrapolation by Moore and co-workers [71, 72] of the results of their laboratory experiments yielded maximal global fluxes of CH3Cl from oceanic phytoplankton of 5–200 Gg year–1, but it was considered unlikely that phytoplankton in the oceans would achieve the production rates observed in the laboratory and, accordingly, these workers revised their estimate downward to 20 Gg year–1, which is less than 4% of the total oceanic flux to the atmosphere. However, similar calculations for CH3Br based on release rates by Phaeocystis, which is a common microalga in coastal waters, suggested that phytoplankton could be a significant source of CH3Br in coastal areas especially in temperate regions [72]. Coastal production of CH3Br of 17 Gg year–1 was estimated at around 95% of the flux from such waters calculated by Lobert et al. [12] on the basis of field measurements. The potential importance of phytoplankton in CH3Br production in coastal areas has been further emphasised by an investigation of a wide selection of classes of phytoplankton [73], which showed that those species that are strictly coastal in distribution are more prolific producers of CH3Br than species more abundant in the open ocean. Extrapolation of the results of this study gave an estimated CH3Br production of 4–78 Gg year–1 in coastal waters and regions of upwelling and 3–47 Gg year–1 for the open ocean. Field measurements in the North Sea confirm a relationship between CH3Br concentrations in seawater and seasonal blooms of phytoplankton dominated by prymnesiophytes such as Phaeocystis [15]. Furthermore, a strong correlation was observed in the open ocean of the north east Atlantic between CH3Br concentrations and those of hexanoyloxyfucoxanthin, a pigment characteristic of prymnesiophytes [15]. Manley and de la Cuèsta [75] attempted estimation of annual global CH3I emissions from the oceans based on the highest cell-normalised rates of production measured in a survey of 15 species of phytoplankton. Annual release was calculated at only 1 Gg year–1 implying that phytoplankton are insignificant contributors to the atmospheric CH3I burden. Several investigators have postulated that marine bacteria may be involved in CH3I production. Thus, weak emissions of CH3I by uncharacterised microbial populations derived from decaying kelp tissue have been recorded during growth on powdered kelp or defined medium [76]. Recently, a survey of a wide variety of terrestrial and marine bacteria indicated that the majority were capable of volatilising I– as CH3I when grown on defined media containing concentrations of I– (0.1 µM) typical of the natural environment [77]. The significance of this biosynthetic trait in the global CH3I cycle has yet to be determined. 4.4 Higher Plants
The release of CH3Cl by a higher plant was first reported in 1982 by Varns [78] who observed emissions of CH3Cl by freshly harvested tubers of the potato (Solanum tuberosum). The phenomenon was not investigated in detail until quite recently when Harper et al. [79] in a survey of 60 potato cultivars showed rates of CH3Cl release ranging from less than 4 to 650 ng g fwt–1 d–1 by tubers assayed
28
D.B. Harper and J.T.G. Hamilton
within 24 h of harvest. Low-level CH3Cl emissions from tubers were detectable six weeks before harvest, but these rose sharply post-harvest reaching a maximum within three or four days of harvest and thereafter diminishing rapidly over the ensuing four weeks. The post-harvest maximum could be delayed and enhanced by storing the tubers at 6°C allowing significant emissions to be observed for up to six months. Presumably, cold storage retards the physiological ageing of the tuber extending the normally transient phase during which CH3Cl release occurs. The observation that whole cells of the halophytic plant Mesembryanthemum crystallinum released small amounts of CH3Cl when incubated in 100 mM KCl [80] prompted a wide-ranging survey of halomethane emissions by the leaves of higher plants by Saini et al. [81]. When leaf discs were floated in 0.1 M solution of KI, CH3I emissions were detected in 75% of 118 herbaceous species examined including 21 halophytes. Release rates for these species, which represented 44 families from 33 orders, ranged over four orders of magnitude from 70 ng g fwt–1 d–1 to 650 µg g fwt–1 d–1. Maximum activity was exhibited by the order Capparales, in particular the Brassicaceae. Various cultivars of the cabbage (Brassica oleracea) displayed the highest release rates, in excess of 300 µg g fwt–1 d–1. Methylation of I– proceeded at a rate 2300-fold greater than that of Cl– and 36-fold that of Br–. Subsequent investigation of the enzymology of the process in B. oleracea revealed that the SAM-dependent enzyme responsible for halide methylation also attacked HS–, SCN– and organic thiols [82, 83]. The enzyme existed as several isoforms some of which displayed a Km with SCN– as substrate of approximately 10–4 less than that with I– and around 10–6 less than that observed with Cl– [84].Attieh et al. [83, 84] therefore concluded that methylation of halide ion is a surrogate activity of an enzyme whose normal physiological function is the detoxification by methylation of organic thiols, SCN– and HS– released on hydrolysis of glucosinolates. The latter group of compounds are thioglycosidic secondary metabolites accumulated by the Brassicaceae which deter attack by herbivores. The extent to which gratuitous methylation of halide ion by this methyltransferase enzyme occurs in plants under normal growing conditions is difficult to assess. The concentration of I– in plant tissue (0.4–16 µM) certainly mitigates against significant CH3I emissions. Interestingly, investigations involving both laboratory and field experiments by Gan et al. [38] have indicated that several Brassica species including broccoli (B. oleracea var. botrytis), cabbage (B. oleracea var. capitata), mustard (B. juncea) and rapeseed (B. napus) release environmentally significant amounts of CH3Br when grown in soil containing natural concentrations of Br– (i.e. 1 mg kg dry wt–1). On the basis of their observations, these authors calculated a global CH3Br flux from agricultural production of rapeseed of approximately 7 Gg year–1 and from that of cabbage of 0.4 Gg year–1, but the total flux of CH3Br from the Brassicaceae may be substantially higher as the family comprises an important component of terrestrial plant biomass both as agriculturally cultivated and wild species. Surprisingly, no mention was made by these authors of concomitant release of CH3Cl. Despite the fact that Cl– is a much poorer substrate (Kcat = 0.005) for the methyltransferase enzyme than Br– (Kcat = 0.07) [82], the normally much higher
The Global Cycles of the Naturally-Occurring Monohalomethanes
29
concentrations of Cl– in plant tissue (5– 70 mM) compared to Br– (2.5–100 µM) might be expected to result in formation of detectable concentrations of CH3Cl. Indeed, measurements in the field by Rhew et al. [85] referred to below indicate significant emissions of CH3Cl by B. juncea. The difficulties regarding extrapolation of results gained from studies in the laboratory to halomethane emissions by plants in the field can to some extent be resolved by directly measuring halomethane release in the field from individual plants or preferably plant communities by enclosure techniques. In recent years, several groups have reported measurements of this nature at a number of sites. Varner et al. [32] showed that release of both CH3Cl and CH3Br occurred at two freshwater peatland sites, one of which was dominated by Sphagnum spp and the other by Sphagnum spp associated with Carex spp and ericaceous shrubs. Fluxes of 10–30 µg m–2 d–1 of CH3Cl and 0.7–2 µg m–2 d–1 of CH3Br were observed towards the end of the growing season.An extrapolation of these rates to wetlands globally yielded emission fluxes of 48 Gg year–1 CH3Cl and 4.6 Gg year–1 CH3Br. Halomethane fluxes from several sites in a rather similar peatland ecosystem on the west coast of Ireland were measured by Dimmer et al. [33] who demonstrated that emissions were strongly correlated with incident light. Global fluxes from peatlands estimated from these data were 5 Gg year–1 CH3Cl, 0.9 Gg year–1 CH3Br and 1.4 Gg year–1 CH3I. Extrapolations to all wetlands globally yielded fluxes of 35 Gg year–1 CH3Cl, 5 Gg year–1 CH3Br and 7 Gg year–1 CH3I. It is not clear in either of the studies whether the flux observed emanates from the plants, the soil or both, and it should be noted that measurements represent net emissions as both production and consumption may occur within the enclosed area. Moreover, it is questionable whether extrapolation from either of these studies to global emissions from wetlands is valid as measurements were restricted in both cases to ecologically similar temperate sites and were collected over short time periods. The relevance of such data to wetland ecosystems in the tropics is debatable. Indeed a study of emissions during the growing season in flooded rice paddies in California revealed a rather different pattern of halomethane release with net fluxes of 35, 10 and 500 µg m–2 d–1 for CH3Cl, CH3Br and CH3I, respectively, suggesting preferential volatilisation of I–1 [43]. Global emissions from rice cultivation worldwide were estimated at 6 Gg y–1 CH3Cl, 1 Gg y–1 CH3Br and 71 Gg y–1 CH3I. Iodomethane emissions of this magnitude would make a significant terrestrial contribution to the atmospheric CH3I budget and are broadly consistent with the results of earlier growth chamber experiments using 125I on volatilisation of I– by rice in Japan which suggested CH3I emissions of around 25 Gg year–1 [86]. Perhaps the most significant finding to date concerning halocarbon emissions from higher plants has been the discovery by Rhew et al. [31] that substantial fluxes of CH3Cl and CH3Br enter the atmosphere from coastal salt marshes. The release of these halomethanes from various vegetation zones in two coastal salt marshes in California was measured using large static flux chambers. Although emissions occurred from all vegetational zones, fluxes were greatest in the middle and upper middle zones where the dominant vegetation included halophytes such as Salicornia spp., Batis maritima and Frankenia grandifolia (Fig. 1). Average daily fluxes of 4–24 mg m–2 d–1 CH3Cl and 0.5–2 mg m–2 d–1 CH3Br were ob-
30
D.B. Harper and J.T.G. Hamilton
Fig. 1. Mean fluxes of CH3Cl and CH3Br from different vegetational zones at Mission Bay saltmarsh, California. Based on data from [31]
served with emissions showing a seasonality corresponding with plant growth. Marked diurnal fluctuations were apparent which were correlated with incident light and air temperature changes. The highest halomethane fluxes were typically associated with high densities of aboveground vegetation and were not inhibited by complete soil inundation nor affected by soil surface temperatures. These observations are consistent with halomethane production by salt marsh vegetation or microflora intimately associated with the plants. Fluxes of CH3Cl and CH3Br displayed a linear correlation with an average molar ratio of emissions of CH3Cl to those of CH3Br of 17:1 regardless of the vegetation zone or the month of the year, pointing to a common mechanism of formation throughout the salt marsh. Assuming their measurements were globally representative of salt marshes, Rhew et al. [31] estimated that salt marshes worldwide may be responsible for release of 170 Gg year–1 CH3Cl (uncertainty range based on spatial variability of emissions 64–440 Gg year–1) and 14 Gg year–1 CH3Br (range 7–29 Gg year–1). The uncertainty range proposed does not encompass uncertainties in estimates of global salt marsh area or in the extent to which Californian salt marshes differ in structure and vegetational composition from other salt marshes. Nevertheless salt marsh ecosystems almost certainly represent the largest terrestrial biological source of CH3Br identified to date and a CH3Cl source comparable to, and possibly exceeding, wood-rotting fungi.Although no indication of the rate of CH3Cl production on a biomass basis is given by Rhew et al. [31], measurements in this laboratory [87] would suggest that the rate of CH3Cl release by whole plants of B. maritima on a fresh wt basis is of the same order as that exhibited by mycelium of the wood-rotting fungus P. pomaceus (Table 3).
31
The Global Cycles of the Naturally-Occurring Monohalomethanes Table 3. Comparison of rates of halomethane production by various organisms
Species
Algae Macrocystis pyrifera (Giant kelp-field observations) Pavlova gyrans (Microalgal prymnesiophyte – laboratory observations) Higher Plants Solanum tuberosum (Freshly harvested potato tubers – range shown by 61 cv) Brassica napus (Rapeseed-field observations) Brassica juncea (Wild mustard-field observations) Batis maritima (Saltwort-whole plant) Fungus Phellinus pomaceus (Mycelium on wood)
Rate of halomethane release (ng g fwt–1 d–1)
Ref.
CH3Cl
CH3Br
CH3I
160
9
14
[27]
ND
140
170
[63]
<4–650
ND
ND
[79]
NR
5
NR
[38]
NR
4
NR
[38]
7.8 ¥ 103
770
290
[87]
20 ¥ 103
ND
ND
[59]
ND not detected, NR not reported.
Subsequent work by Rhew et al. [85] on shrubland ecosystems in southern California demonstrated significant emissions from several plant species, albeit at much lower levels than those displayed by salt marsh species. Thus Artemesia californica (coastal sagebrush) yielded 0.1 mg m–2 d–1 CH3Cl during the dry season whilst Brassica juncea (wild mustard) growing within the sagebrush community had the largest CH3Br release rate of 14 µg m–2 d–1. The emission of CH3Br by B. juncea is consistent with the laboratory-based observations of Gan et al. [38] mentioned above, but the simultaneous emission of much larger quantities of CH3Cl (55 µg m–2 d–1) were also recorded suggesting that the methylation of halide ion by Brassica spp may not be confined to Br– as implied by Gan et al. [38]. The halomethane emissions by plant species in the shrubland ecosystems examined were largely counterbalanced by soil uptake particularly in the wet season so that it is not, as yet, clear whether in overall terms the biome is a net source or sink for these compounds. 4.5 Animals
In light of the considerable quantities of methane released in the digestive system of ruminant animals by anaerobic fermentation of cellulose, Williams et al. [88] measured release of halomethanes by cows in open circuit respiration cham-
32
D.B. Harper and J.T.G. Hamilton
bers. In contrast to methane emissions of approximately 150 g cow–1 d–1, emissions of CH3Cl amounted to only 0.4–1.5 mg cow–1 d–1 and those of CH3Br to 2.6–7 µg cow–1 d–1. On extrapolation to a 1.3 billion cattle population worldwide, fluxes of 0.2–0.7 Gg year–1 CH3Cl and 0.001–0.1 Gg year–1 CH3Br were estimated, inconsequential atmospheric inputs in global budgetary terms.
5 Biological Sinks 5.1 Soil Microorganisms
Soils have been identified as significant sinks for atmospheric CH3Br and preliminary findings indicate that they may be equally, if not more, important in the uptake of atmospheric CH3Cl. Uptake of CH3I by soils is too slow relative to photolysis in the atmosphere for it to be of consequence in the atmospheric CH3I budget. Initial investigations into halomethane uptake by soils at low mixing ratios (ppbv levels) were reported by Shorter et al. [44] and Hines et al. [45] who observed rapid uptake of CH3Br by a variety of soils at CH3Br concentrations ranging from 5 pptv to 14 ppbv. Degradation rates were highest in moist soils of high organic matter content. Uptake of CH3Br appears to be an aerobic bacterial process. Thus, consumption did not occur after autoclaving soil and the temperature response was consistent with biological consumption (i.e. maximum uptake was observed between 25 and 37 °C). Bactericidal antibiotics such as chloramphenicol strongly inhibited CH3Br uptake, but antibiotics such as cycloheximide active against eukaryotic organisms did not affect CH3Br consumption.Activity was abolished by incubation of soil under a nitrogen atmosphere. Temperate forest soils taken from 0 – 3 cm depth showed the maximum rates of uptake, reducing ambient CH3Br concentrations below the detection limit (2 pptv) within a few minutes. Such rapid degradation could result in the rate of removal of CH3Br from the atmosphere being limited by the rate of diffusion of CH3Br into the soil from the atmosphere. However, on the basis of field observations Shorter et al. [44] argued that, whilst soil diffusion limited CH3Br uptake in temperate agricultural, sandy boreal and tropical soils examined, CH3Br consumption was not diffusion limited in temperate forest/woodland/shrubland soils or temperate grassland soils. These authors estimated a global soil sink of 42 Gg year–1 to which temperate forests/woodland/shrubland soils contributed 22 Gg year–1 and temperate grassland soils 10 Gg year–1. A soil sink of this magnitude would account for around 25% of the global annual loss of CH3Br from the atmosphere and would yield an atmospheric lifetime for CH3Br with respect to soil loss of around 3.4 years. Subsequent investigations by these workers [89] suggested that their estimate of CH3Br uptake by agricultural soils was too low and led to a revision of the global soil sink to 47 Gg year–1. Serca et al. [90], on the basis of similar measurements to those of Shorter et al. [44], estimated a much larger soil sink for CH3Br of 143 Gg year–1 which would yield an atmospheric lifetime for CH3Br relative to this sink of around one year.
The Global Cycles of the Naturally-Occurring Monohalomethanes
33
The disparity between these estimates is largely attributable to the much-increased value for uptake by agricultural soils employed by Serca et al. [90] (66 Gg year–1) and the use of a different global inventory of soils by these workers. In contrast, recent work by Rhew et al. [85] implies that CH3Br uptake by soil from at least one biome (shrubland) investigated by Shorter et al. [44] may have been overestimated. Comparatively little investigation of the magnitude of the soil sink for CH3Cl has been conducted. On the basis of soil-atmospheric exchange observations in a Brazilian forest and in arctic grasslands and tundra, Khalil and Rasmussen [28] estimated a provisional global uptake of CH3Cl of 0.5 Tg year–1. However, an authoritative estimate will not be possible until a much broader range of soils from a variety of climatic zones is examined. Compelling evidence for a large sink in the soil for CH3Cl has emerged from microbiological studies. Whilst the possibility of co-metabolism of CH3Cl and CH3Br by several microorganisms has long been recognised [91–95], it is only in recent years that bacteria, both anaerobic [96] and aerobic [97–104], capable of using the compounds as their sole C and energy source have been isolated. Such organisms have been isolated not only from soils where contamination with man-made organohalogens is very probable, that is industrially polluted sites [98–100], sewage sludge [96, 97], and CH3Br-fumigated agricultural soils [101], but also from soils in pristine environments where biological production of CH3Cl might be expected, for example the upper layer of woodland soil, an environment where CH3Cl concentrations are likely to be enhanced by CH3Cl release by wood-rotting fungi [102]. It is significant that soil from an ecosystem similar to this exhibited the highest rate of CH3Br uptake in the studies of Shorter et al. [44]. The lenticels of potato tubers that provide the main points of efflux of CH3Cl from the tubers to the external atmosphere also appear to have a microbial population adapted to use CH3Cl as a C source [79]. McAnulla et al. [104] succeeded in isolating CH3Cl-degrading microorganisms from a variety of pristine environments and concluded that CH3Clutilizing species are ubiquitous in the natural environment. Hence, it seems quite feasible for the microbial soil sink for CH3Cl to be as large a proportion of the atmospheric burden as that constituted by the microbial soil sink for CH3Br. A CH3Cl soil sink of this magnitude would be of the order of 1.6 Tg year–1. As the ability of an organism to use CH3Cl as its sole carbon source is often associated with the ability to either use CH3Br as a sole C source or to co-metabolise it, the CH3Cl-degrading activity of a soil may be a good indicator of its CH3Br-degrading capability. Pursuing this argument further, it is not unreasonable to assume that the uptake of CH3Br by soil microorganisms largely represents a gratuitous activity of a natural microbial population evolved to utilise the much higher atmospheric and environmental concentrations of CH3Cl. 5.2 Marine Bacteria
Stable isotope dilution techniques using 13CH3Br have been employed to determine the extent to which biological degradation contributes to the predominantly abiotic (see Sect. 3.1) removal of CH3Br from surface ocean waters at tempera-
34
D.B. Harper and J.T.G. Hamilton
tures between 21 and 29 °C [47]. Mean loss rates were significantly lower in seawater which had been autoclaved or filtered through a 0.2 µm filter but not a 1.2 µm filter suggesting that bacteria were responsible for some CH3Br removal. The enhancement of CH3Br removal due to such biological losses was measured at 10 – 110 % (mean 41 %) over the chemical losses. The authors argued that biological destruction might well dominate CH3Br removal in high productivity polar waters where low temperatures cause the chemical loss rate to be extremely slow. Goodwin et al. [105] confirmed the presence in seawater of nonmethanotrophic bacteria that could oxidise CH3Br when incubated in natural seawater and would grow on CH3Br in enrichment cultures. In a more recent investigation, Hoeft et al. [106] found CH3Br was degraded by three out of five marine methylotrophic bacterial strains isolated from coastal waters off northeastern USA and the tropical North Atlantic. CH3Br supported biomass production in only one strain; the others appeared to co-metabolize CH3Br during growth on other 1carbon substrates. The authors hypothesized that, in view of the low concentrations of CH3Br in seawater, co-metabolic uptake of CH3Br during bacterial growth on other substrates was much more likely than bacterial growth on CH3Br as the sole carbon and energy source. 5.3 Higher Plants and Fungi
The foliage of a wide variety of higher plants including both deciduous and coniferous tree species rapidly removed CH3Br from the surrounding air when concentrations between 500 pptv and 10 ppmv were present [107]. Loss rates, which were independent of photosynthesis but characteristic of each species, were directly proportional to leaf area and displayed first-order kinetics. Rate measurements at different temperatures suggested that the activation energy of the reaction was 27 kJ mol–1, about a quarter of that associated with the hydrolysis of CH3Br or Cl– exchange with the compound. The authors of this report considered that their observations were consistent with the enzyme-mediated uptake of CH3Br by plant leaves and suggested that the global sink due to plant uptake might be of the same order as that estimated for soil. Wood-rotting fungi of the Hymenochaetaceae which release CH3Cl during growth also utilize the halomethane as a methyl donor in their metabolic processes [108]. The emission of CH3Cl observed represents a net flux, that is the excess of that biosynthesised over that metabolically consumed by the fungus. Many white rot fungi which do not release CH3Cl (e.g. Phanerochaete chrysosporium, Coriolus versicolor, Phlebia radiata) can, however, utilize the compound in the biosynthesis of aromatic compounds such as veratryl alcohol (3,4dimethoxybenzyl alcohol), a key component in the lignin degradation system [109]. Presumably, in these fungi biosynthesis and utilization of CH3Cl is strictly coordinated so that endogenously-synthesised CH3Cl is not released to the atmosphere. Nevertheless, such fungi are capable of utilising exogenous CH3Cl in veratryl alcohol biosynthesis [109–112], which raises the possibility that these fungi might act as a sink for atmospheric CH3Cl, although uptake is likely to be limited by the low concentrations in the atmosphere.
The Global Cycles of the Naturally-Occurring Monohalomethanes
35
6 Atmospheric Budgets 6.1 Sources Versus Sinks
A diagrammatic summary of the fluxes involved in global cycling of CH3Cl, CH3Br and CH3I is shown in Scheme 2 which includes all the significant sources and sinks identified to date for these halomethanes. It should be noted that a given source or sink may not be of equal relevance to the cycling of all halomethanes. Uncertainties in the magnitudes of the global sinks for all of the halomethanes render complete budget closure for any of the compounds unachievable at the present time (Table 2). In the case of CH3I, there are considerable uncertainties in the source terms in addition to the size of the sink so that little progress as regards budgetary reconciliation is possible until a further understanding of the processes responsible for production and removal of atmospheric CH3I is attained. The sources and sinks for CH3Br are better constrained, but nevertheless the best estimate of the global sink exceeds the total estimated input from known sources by approximately 50 Gg year–1 (i.e. approximately 30%). This discrepancy suggests an unknown source of CH3Br which would appear to be terrestrial, as the net oceanic flux of –21 Gg year–1 [6] is fairly tightly constrained. However, even if such a source is identified, until the uncertainty regarding the magnitude of the soil sink is resolved it is unlikely that the budget for atmospheric CH3Br can be accurately quantified. The imbalance between calculated emissions from known sources and the estimated global sink for CH3Cl is substantial, between 2.5 and 3.6 Tg year–1 depending on the strength assigned to the soil sink. This shortfall could result from overestimation of sink terms, in particular the OH sink. Some uncertainty has existed regarding the temperature dependence of the rate constant for reaction of OH with CH3Cl [113]. This uncertainty corresponds to a range of ± 0.9 Tg year–1 around the estimate of 3.7 Tg year–1 for the OH sink by Khalil and Rasmussen [1]. Recent measurements by Herndon et al. [114], whilst approximately halving the uncertainty in the temperature dependence of the rate constant, have resulted in the revision of the rate constant itself upwards by about 5%.A globally-averaged atmospheric lifetime for CH3Cl with respect to loss by reaction with OH of 1.3 year was calculated which would tend to increase the size of the atmospheric sink as compared to that estimated by Khalil and Rasmussen [1] and so accentuate the discrepancy between sources and sinks in the CH3Cl budget. It would therefore appear that the most plausible explanation of the budgetary imbalance is either that a major terrestrial source, almost certainly biological, has yet to be identified or that the strength of a known source has been underestimated. Consistent with the source deconvolution studies of Khalil and Rasmussen [1], all important sources quantified to date are situated predominantly in the tropics. A similar global distribution must necessarily be exhibited by any, as yet unidentified, sources. In light of the finding that the high concentrations of atmospheric CH3Cl observed off tropical islands such as Okinawa appear to be strongly correlated with a-pinene, a plant-derived terpene with an atmospheric lifetime of a
Scheme 2. Sources and sinks involved in the global CH3X cycles (X=Cl, Br or I)
36 D.B. Harper and J.T.G. Hamilton
The Global Cycles of the Naturally-Occurring Monohalomethanes
37
few hours [7], it seems not unlikely that plant communities situated in the tropics, particularly those growing in high salinity environments, may represent a source of CH3Cl of sufficient magnitude to resolve the CH3Cl budget imbalance. This possibility has very recently been given strong support by Yokouchi et al. [115] who have demonstrated high rates of CH3Cl release by trees of the Dipterocarpaceae and certain families of tropical fern. 6.2 Budget Modelling Using Isotope Ratios
One possible approach to constraining source terms in the atmospheric budgets of the halomethanes is to employ the stable isotope techniques which have been applied with some success in compiling a methane budget for the atmosphere [116]. In the case of CH3Cl for example, this would involve using the stable carbon isotope signatures (d13C) for CH3Cl from various sources in conjunction with estimates of their emission fluxes to predict a carbon isotope ratio for atmospheric CH3Cl. Providing a correction is made for any fractionation during loss processes, the value obtained should agree with the measured atmospheric ratio if the budget has been correctly modelled. Existing information on estimated emission fluxes from known sources and the d13C values for each is summarized in Fig. 2 [117]. Data, albeit preliminary and limited, is available on d13C values for emissions from biomass burning, industry, fungi and higher plants, but not for CH3Cl released by the oceans or coal combustion. Nevertheless, it is clear that the distinctive isotope signature of plant-derived CH3Cl should facilitate the use of the isotopic approach in constraining the contribution of higher plants to atmospheric CH3Cl. If the source component of unknown origin in Fig. 2 were to be assigned to higher plants, the effect on the weighted mean calculated for atmospheric CH3Cl would be substantial. Unfortunately, only a few measurements of d13C values for atmospheric CH3Cl at unpolluted sites have been reported. These range from a mean of – 33.5 ‰ for the remote marine atmosphere of the Pacific to – 36.5‰ and –43.5‰ at coastal
Fig. 2. Sources of atmospheric CH3Cl with estimated emission fluxes and mean d13C values
recorded to date [117]
38
D.B. Harper and J.T.G. Hamilton
sites in Japan and New Zealand [118, 119]. No measurements have been conducted at continental sites and d13C values at the marine locations studied showed a wide range. d13C measurements at many more locations will obviously be necessary before a global mean can be computed that can be used in validation of the modelled budget. In addition, a correction will need to be applied to the modelled weighted mean to allow for fractionation during loss processes. No data is currently available on isotopic fractionation during reaction of CH3Cl with OH, but investigations by Miller et al. [120] suggest that fractionation during microbial degradation could be quite substantial. Thus, a microbial soil sink accounting for 20% of atmospheric CH3Cl would produce an overall enrichment of around 13‰ [117].
7 References 1. Khalil MAK, Rasmussen RA (1999) Atmos Environ 33 : 1305 2. Prinn RG, Zander R (1999) In: Ennis CA (ed) Scientific assessment of ozone depletion: 1998. 3. World Meteorological Organisation, Geneva 3. Madronich S, Velders GJM (1999) In: Ennis CA (ed) Scientific assessment of ozone depletion: 1998. World Meteorological Organisation, Geneva 4. Koppmann R, Johnen FJ, Plas-Dülmer C, Rudolph J (1993) J Geophys Res 98:20517 5. Moore RM, Groszko W, Niven SJ (1996) J Geophys Res 101:28529 6. Kurylo MJ, Rodriguez JM (1999) In: Ennis CA (ed) Scientific assessment of ozone depletion: 1998. World Meteorological Organisation, Geneva 7. Yokouchi Y, Noijiri Y, Barrie LA, Toom-Sauntry D, Machida T, Inuzuka Y, Akimoto H, Li H-J, Fujinuma Y, Aoki S (2000) Nature 403:295 8. Li H-J, Yokouchi Y, Akimoto H, Narita Y (2001) Geochem J 35:137 9. Li H-J, Yokouchi Y, Akimoto H (1999) Atmos Environ 33:1881 10. Butler JH, Battle M, Bender ML, Montzka SA, Clarke AD, Saltzman ES, Sucher CM, Severinghaus JP, Elkins JW (1999) Nature 399:749 11. Khalil MAK, Moore RM, Harper DB, Lobert JM, Erikson DJ, Koropalov V, Sturges WT, Keene WC (1999) J Geophys Res 104:8333 12. Lobert JM, Butler JH, Montzka SA, Geller LS, Myers RC, Elkins JW (1995) Science 267 : 1002 13. Khalil MAK, Rasmussen RA, Gunawardena R, (1993) J Geophys Res 98 : 2887 14. Moore RM, Webb M (1996) Geophys Res Lett 23:2951 15. Baker JM, Reeves CE, Nightingale PD, Penkett SA, Gibb SW, Hatton AD (1999) Mar Chem 64:267 16. Yvon-Lewis SA, Butler JH (1997) Geophys Res Lett 24:1227 17. Butler JH (1996) Atmos Environ 30 : i 18. Butler JH, Rodriguez JM (1996) In: Bell CH, Price N, Chakrabarti B (eds) The methyl bromide issue. Wiley & Sons, Chichester, p 27 19. Rasmussen RA, Khalil MAK, Gunawardena R, Hoyt SA (1982) J Geophys Res 87: 3086 20. Reifenhauser W, Heumann KG (1992) Atmos Environ 26A : 2905 21. Scholl C, Heumann KG (1993) Fresenius Z Anal Chem 346:717 22. Chameides WL, Davis DD (1980) J Geophys Res 85:7383 23. Solomon S, Garcia RR, Ravishankara AA (1994) J Geophys Res 99:20491 24. Oram DE, Penkett SA (1994) Atmos Environ 28:1159 25. Singh ON, Fabian P (1999) In: Fabian P, Singh ON (eds) The handbook of environmental chemistry, vol 4, part E. Springer Verlag, Berlin, p 1 26. Sturges WT, McIntyre HP, Penkett SA, Chappellaz J, Barnola J-M, Mulvaney R, Atlas E, Stroud V (2001) J Geophys Res 106 : 1595
The Global Cycles of the Naturally-Occurring Monohalomethanes
27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70.
39
Manley SL, Dastoor MN (1987) Limnol Oceangr 32:709 Khalil MAK, Rasmussen RA (2000) Environ Sci Pollut Res 7 : 79 Lobert JM, Keene WC, Logan JA, Yevich R (1999) J Geophys Res 104 : 8373 Watling R, Harper DB (1998) Mycol Res 102:769 Rhew RC, Miller BR, Weiss RF (2000) Nature 403 : 292 Varner RK, Crill PM, Talbot RW (1999) Geophys Res Lett 26:2433 Dimmer CH, Simmonds PG, Nickless G, Bassford MR (2001) Atmos Environ 35:321 McCulloch A,Aucott ML, Benkovitz CM, Graedel TE, Kleiman G, Midgley PM, Li YF (1999) J Geophys Res 104:8391 Andreae MO, Atlas E, Harris GW, Helas G, de Kock A, Koppmann R, Maenhaut W, Manö S, Pollock WH, Rudolph J, Scarffe D, Schebeske G, Welling M (1996) J Geophys Res 101:14531 Miller M (1996) In: Bell CH, Price N, Chakrabarti B (eds) The methyl bromide issue. Wiley & Sons, Chichester, p 91 Baker JM, Reeves CE, Penkett SA, Cardenas LM, Nightingale PD (1998) Geophys Res Lett 25:2405 Gan J, Yates SR, Ohr HD, Sims JJ (1998) Geophys Res Lett 25:3595 Lee Taylor JM, Holland EA (2000) J Geophys Res 105:8857 Singh HB, Salas LJ, Stiles RE (1983) J Geophys Res 88:3684 Moore RM, Groszko W (1999) J Geophys Res 104:11163 Campos MLAM, Nightingale PD, Jickells TD (1996) Tellus 48B:106 Redeker KR, Wang N-Y, Low JC, McMillan A, Tyler SC, Cicerone RJ (2000) Science 290: 966 Shorter JH, Kolb CE, Crill PM, Kerwin RA, Talbot RW, Hines ME, Harriss RC (1995) Nature 377 : 717 Hines ME, Crill PM, Varner RK, Talbot RW, Shorter JH, Kolb CE, Harriss RC (1998) Appl Environ Microbiol 64:1864 Keene WC, Jacob DJ, Fan S-M (1996) Atmos Environ 30: i King DB, Saltzmann ES (1997) J Geophys Res 102:18715 Roehl CM, Burkholder JB, Moortgat GK, Ravishankara AR, Crutzen PJ (1997) J Geophys Res 102:12819 Tait VK, Moore RM, Tokarczyk R (1994) J Geophys Res 99:7821 Zafiriou OC (1975) J Mar Res 33:75 White RH (1982) J Mar Res 40:529 Moore RM, Zafiriou OC (1994) J Geophys Res 99:16415 Happell JD, Wallace DWR (1996) Geophys Res Lett 23:2105 Yokouchi Y, Nojiri Y, Barrie LA, Toom-Sauntry D, Fujinuma Y (2001) J Geophys Res 106:12661 Manley SL (1994) Mar Chem 46 : 361 Keppler,F, Eiden R, Niedan V, Pracht J, Schöler HF (2000) Nature 403:298 Keppler,F, Eiden R, Niedan V, Pracht J, Schöler HF (2001) Nature 409:382 Harper DB, Kennedy JT, Hamilton JTG (1988) Phytochemistry 27:3147 Harper DB, Kennedy JT (1986) J Gen Microbiol 132:1231 Lovelock JE (1975) Nature 256:193 Manley SL, Goodwin K, North WJ (1992) Limnol Oceanogr 37:1652 Nightingale PD, Malin G, Liss PS (1995) Limnol Oceanogr 40:680 Itoh N, Tsujita M, Ando T, Hisatomi G, Higashi T (1997) 45:67 Schall C, Laturnus F, Heumann KG (1994) Chemosphere 28 : 1315 Laturnus F, Adams FC, Wiencke C (1998) Geophys Res Lett 25 : 773 Giese B, Laturnus F, Adams FC, Wiencke C (1999) Environ Sci Technol 33:2432 Ekdahl A, Pedersen M, Abrahamsson K (1998) Mar Chem 63 : 1 Abrahamsson K, Ekdahl A, Collen J, Pedersen M (1995) Limnol Oceanogr 40:1321 Marshall RA, Hamilton JTG, Dring MJ, Harper DB (2000) Limnol Oceanogr 45:516 Baker JM, Sturges WT, Sugier J, Sunnenberg G, Lovett AA, Reeves CE, Nightingale PD, Penkett SA (2001) Chemosphere – Global Change Sci 3:93
40 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113.
D.B. Harper and J.T.G. Hamilton Tait VK, Moore RM (1995) Limnol Oceanogr 40:189 Scarratt MG, Moore RM (1996) Mar Chem 54: X263 Sæmundsdóttir S, Matrai PA (1998) Limnol Oceanogr 43:81 Scarratt MG, Moore RM (1999) Limnol Oceanogr 44:703 Manley SL, de la Cuèsta JL (1997) Limnol Oceanogr 42:142 Manley SL, Dastoor MN (1988) Mar Biol 98:477 Amachi S, Kamagata Y, Kanagawa T, Muramatsu Y (2001) Appl Environ Microbiol 67: 2718 Varns JL (1982) Am Potato J 59:593 Harper DB, Harvey BMR, Jeffers MR, Kennedy JT (1999) New Phytol 142:5 Wuosmaa AM, Hager LP (1990) Science 249:160 Saini HS, Attieh JM, Hanson AD (1995) Plant Cell Environ 18:1027 Attieh JM, Hanson AD, Saini HS (1995) J Biol Chem 270:9250 Attieh J, Kleppinger-Sparace KF, Nunes C, Sparace SA, Saini HS (2000) Plant Cell Environ 23:165 Attieh J, Sparace SA, Saini HS (2000) Arch Biochem Biophys 380:257 Rhew RC, Miller BR, Vollmer MK, Weiss RF (2001) J Geophys Res 106:20875 Muramatsu Y, Yoshida S (1995) Atmos Environ 29:21 Harper DB, unpublished observations Williams J, Wang N-Y, Cicerone RJ, Yagi K, Kurihara M, Terada F (1999) Global Biogeochem Cycles 13 : 485 Varner RK, Crill PM, Talbot RW, Shorter JH (1999) Geophys Res Lett 26:727 Serca D, Guenther A, Klinger L, Helmig D, Hereid D, Zimmerman P (1998) Atmos Environ 32:1581 Stirling DI, Dalton A (1979) FEMS Microbiol Lett 5:315 Keuning S, Janssen DB, Witholt B (1985) J Bacteriol 163:635 Rasche ME, Hicks RE, Hyman MR, Arp DJ (1990) J Bacteriol 172:5368 Oremland RS, Miller LG, Culbertson CW, Connell TL, Jahnke L (1994) Appl Environ Microbiol 60:3640 Han J-I, Semrau JD (2000) FEMS Microbiol Lett 187:77 Traunecker J, Preuß A, Diekert G (1991) Arch Microbiol 156:416 Hartmans S, Schmuckle A, Cook AM, Leisinger T (1986) J Gen Microbiol 132:1139 Doronina NV, Sokolov AP, Trotsenko YA (1996) FEMS Microbiol Lett 142:179 Doronina NV, Trotsenko YA (1997) Microbiologiya 66:70 Vannelli T, Studer A, Kertesz M, Leisinger T (1998) Appl Environ Microbiol 64:1933 Connell Hancock TL, Costello AM, Lidstrom ME, Oremland RS (1998) Appl Environ Microbiol 64:2899 Coulter C, Hamilton JTG, McRoberts WC, Kulakov L, Larkin MJ, Harper DB (1999) Appl Environ Microbiol 65:4301 McDonald IR, Doronina NV, Trotsenko YA, McAnulla C, Murrell JC (2001) Int J Syst Evol Micr 51: 119 McAnulla C, McDonald IR, Murrell JC (2001) FEMS Microbiol Lett 201:151 Goodwin KD, Schaefer JK, Oremland RS (1998) Appl Environ Microbiol 64:4629 Hoeft SE, Rogers DR, Visscher PT (2000) Aquat Microb Ecol 21:231 Jeffers PM, Wolfe NL, Nzengung, V (1998) Geophys Res Lett 25 : 43 Harper DB, Hamilton JTG, Kennedy JT, McNally KJ (1989) Appl Environ Microbiol 55:1981 Harper DB, Buswell JA, Kennedy JT, Hamilton JTG (1990) Appl Environ Microbiol 56: 3450 Coulter C, Hamilton JTG, Harper DB (1993) Appl Environ Microbiol 59:1461 Harper DB, Buswell JA, Kennedy JT (1991) J Gen Microbiol 137:2867 Harper DB, McRoberts WC, Kennedy JT (1996) Appl Environ Microbiol 62:3366 DeMore WB, Sander SP, Golden DM, Hampson RF, Kurylo MJ, Howard CJ, Ravishankara AR, Kolb CE, Molina MJ (1997) Chemical kinetics and photochemical data for use in stratospheric modeling, evaluation 12. NASA, JPL, Pasadena
The Global Cycles of the Naturally-Occurring Monohalomethanes
41
114. Herndon SC, Gierczak T, Talukdar RK, Ravishankara AR (2001) Phys Chem Chem Phys 3:4529 115. Yokouchi Y, Ikeda M, Inuzuka Y, Yukawa T (2002) Nature 416 : 163 116. Cantrell CA, Shetter RE, McDaniel AH, Calvert JG, Davidson JA, Lowe DC, Tyler SC, Cicerone RJ, Greenberg JP (1990) J Geophys Res 95 : 22455 117. Harper DB, Kalin RM, Hamilton JTG, Lamb C (2001) Environ Sci Technol 35:3616 118. Rudolph J, Lowe DC, Martin RJ, Clarkson TS (1997) Geophys Res Lett 24:659 119. Tsunogai U, Yoshida N, Gamo T (1999) J Geophys Res 104 : 16033 120. Miller LG, Kalin RM, McCauley SE, Hamilton JTG, Harper DB, Millet DB, Oremland RS, Goldstein AH (2001) Proc Natl Acad Sci USA 98:5833
The Handbook of Environmental Chemistry Vol. 3, Part P (2003): 63–84 DOI 10.1007/b 10446
Abiotic Formation of Organohalogens During Early Diagenetic Processes Heinz F. Schöler, Frank Keppler Institute of Environmental Geochemistry, University of Heidelberg, 69120 Heidelberg, Germany. E-mail:
[email protected]
To date more than 3650 organohalogen compounds are known to be naturally produced by biogeochemical processes. The current understanding of the abiotic formation of organohalogens during early diagenetic processes in soils and sediments are reviewed here. Next to volatile alkyl halides and polar organohalogens such as haloacetates there is evidence that even semivolatile organohalogens (e.g. polychlorinated dibenzodioxins) and halogenated humic substances are naturally formed by geochemical processes. Keywords. Abiotic processes, Early diagenesis, Natural halogenation
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
1
Introduction
1.1 1.2 1.3 1.4
Volatile Organohalogens (VHOCs) . . . . . . . . . . . . . . . . . Semi-Volatile Organohalogens (e.g. PCDDs and PCDFs) . . . . . . Polar Organohalogens (e. g. Haloacetates) . . . . . . . . . . . . . . High Molecular Polymeric Organohalogens (Halogenated Humus)
2
Prerequisites for Abiotic Early Diagenetic Processes in the Terrestrial Environment . . . . . . . . . . . . . . . . . . . . . 71
2.1 2.2 2.3
Inorganic Halide . . . . . . . . . . . . . . . . . . . . . . . . . . . 72 Redox-Sensitive Elements (e. g. Fe) . . . . . . . . . . . . . . . . . 72 Organic Matter . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
3
Abiotic Formation of Organohalogens in the Terrestrial Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
4
Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
5
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
65 66 67 69
1 Introduction The topic “natural organohalogens in soils and sediments” has been accessed from various sides: natural product chemists looked for new antibiotics from microorganisms and fungi, biologists have elucidated biochemical metabolic pathways of soil organisms, water chemists have found unexpectedly elevated con© Springer-Verlag Berlin Heidelberg 2003
64
H.F. Schöler and F. Keppler
Fig. 1. Scheme of the abiotic formation of organohalogens in the terrestrial environment
centrations of chlorinated solvents in groundwater and environmental scientists have applied a new analytical technique (so-called AOX parameter = adsorbable organic halogen) for depth profiling in soils and sediments. Now to the flip side of the coin: the driving force for this research has been the need to understand the impact of man-made organohalogens, for example chloropesticides, which are directly or indirectly distributed into the environment and, surprisingly, environmental chemists have found naturally produced organics which proved to be identical or similar to the man-made ones. The current knowledge on natural organohalogens in soils and sediments will be reviewed with special emphasis on organohalogens formed by abiotic processes. This might be the starting point for deeper insights into biotic and abiotic soil processes such as immobilization of pesticides, natural attenuation and remediation of polluted sites. The abiotic formation of halocarbons during diagenesis processes can be structured in three branches (see Fig. 1). Biomass burning means radical chemistry of organic material in the presence of halides at elevated temperatures resulting in methyl halides [1–3]. Volcanoes produce a whole bunch of volatile organohalogens including fluoro compounds via radical chemistry starting from methane, ethene and ethyne in the presence of halides on very hot mineral surfaces [4–5]. Early diagenetic processes in soils and sediments comprise radical chemistry of organic material in the presence of halides at ambient temperatures driven by redox-sensitive elements such as iron [6 – 12]. This review deals with early diagenetic processes in the terrestrial environment at ambient temperatures which produce organohalides and comprises a detailed discussion of four classes of organohalides for which evidence has arisen in recent years that they are naturally produced in this compartment. To date, most of these organohalogens are claimed to be formed by biotic processes, but there are recent studies concerning a major contribution from abiotic processes: – Volatile organohalogens (VHOCs), – Semi-volatile organohalogens (e. g. PCDDs, PCDFs),
Abiotic Formation of Organohalogens During Early Diagenetic Processes
65
– Polar organohalogens (e. g. haloacetates), – High molecular polymeric organohalogens (halogenated humus). 1.1 Volatile Organohalogens (VHOCs)
Relevant data about the quantity and variety of low molecular organohalides in the terrestrial environment are limited [13, 14]. This compartment receives significant fluxes of inorganic halide via the deposition of sea-salt aerosols and of combustion processes, and in addition from weathering processes of rocks. The lower concentrations of halides in soil point to the minor importance of natural halogenation as a source of volatile organohalogens. However, there are now a number of studies which demonstrate that natural halogenation is leading to VHOCs. Terrestrial biota, such as fungi, plants, animals and insects, are now known to be de novo producers of volatile organohalogens. There are also abiotic sources of natural VHOCs observed in the terrestrial environment. Methyl chloride (CH3Cl) was found to be produced by fungi and higher plants [15–16]. High emissions of methyl halides have been observed over peatlands, salt marshes and rice paddies [17–19]. A natural formation of chloroform in forest soils has been reported by Laturnus et al. [20] and Hoekstra et al. [21 – 22]. They assume that chloroform is formed by exo-enzymes present in soil such as chloroperoxidase [22 – 24]. An entirely new source of CH3Cl, CH3Br, CH3I and other methyl halides from the terrestrial environment has recently been suggested by Keppler et al. [6]: methyl halides are formed during degradation of organic matter by an oxidant (e.g. Fe(III)) in the presence of halide ions (a possible reaction mechanism is discussed in later). This abiotic process could generate large amounts of volatile methyl halides, but as yet there are no estimates available for this source from the soil. In addition to the monohalogenated methanes, long-chain alkyl halides such as C2H5Cl, C3H7Cl and C4H9Cl are also formed. The corresponding alkyl bromides or alkyl iodides are also produced when bromide or iodide are applied as the halide source. Very recently, vinyl chloride has joined the growing list of some 2000 organochlorine compounds that are produced by natural biogeochemical processes [8, 10, 13]. In this case the formation of vinyl chloride and other organohalogens was observed in soil samples and in commercially available humic acid. The soil was observed to generate vinyl chloride at a rate of up to 120 pg g–1 soil h–1, as well as measurable amounts of C1 –C3 monochlorinated alkanes. This process could be modelled by the reaction of catechol with KCl and Fe2(SO4)3, which resulted in pg quantities of vinyl chloride. Moreover, the natural formation of vinyl chloride could be verified in soil air of salt marshes from Northern Germany. Another novel “natural organic halogen compound” is chloroethyne [8]. This highly reactive unsaturated compound has also been found to be produced during oxidative degradation of humic substances in the presence of chloride. It is assumed that bromoethyne and iodoethyne are also produced when bromide or iodide are applied as the halide source. Furthermore, it is not clear if polyhalo-
66
H.F. Schöler and F. Keppler
Table 1. Volatile halogenated organic compounds observed and presumed to be formed by abiotic processes
VHOC
Reference
Chloromethane Bromomethane Iodomethane Chloroethane Bromoethane Iodoethane 1-Chloropropane 1-Bromopropane 1-Iodopropane 2-Chloropropane 2-Bromopropane 2-Iodopropane 1-Chlorobutane 1-Bromobutane 1-Iodobutane 2-Iodobutane Chloroethene Bromoethene Iodoethene Chloroethyne Bromoethyne Iodoethyne Dichloromethane Dibromomethane Diiodomethane Trichloromethane Tribromomethane Triiodomethane Tetrachloromethane
[6, 10] [6] [6] [6, 10] presumed [6, 12] [10] presumed [6, 12] [10] presumed [12] presumed presumed [6, 12] [12] [8, 10] presumed presumed [8] presumed presumed presumed presumed presumed presumed presumed presumed presumed
genated C1-organohalogens such as dichloromethane or tetrachloromethane can be formed during decomposition of dead organic matter in soil. Sometimes it is a hard task – and makes no sense – to differentiate between biotic and solely abiotic processes in soil. Sometimes both processes are linked to each other in producing VHOCs. When geochemical processes are involved, the VHOC formation does not seem to be purposeful or obvious. In our opinion their formation is merely a caprice of nature, because halides are present when highly reactive organic compounds are produced in the course of soil processes. 1.2 Semi-Volatile Organohalogens (e. g. PCDDs and PCDFs)
Natural sources of polychlorinated dibenzodioxins (PCDDs) and polychlorinated dibenzofurans (PCDFs) are well established by pre-industrial samples of soil and herbage from a controlled long-term agricultural experiment at Rothamsted Ex-
Abiotic Formation of Organohalogens During Early Diagenetic Processes
67
perimental Station (England) [25 – 26]. The PCDD/PCDF content of the soil before 1880 is very low and the associated congener profile is completely different to all known anthropogenic sources. Studies on dated lake sediments from Germany revealed significant PCDD/F contents in sediment layers believed to pre-date the large-scale industrial production of chlorinated organics [27]. Kjeller and Rappe [28] also found PCDD/F in a 110 year-old sediment core from the Baltic Sea in small but significant levels around the year 1882 at the end of the 19th century. There is evidence that biogenic processes such as composting and humification within peat bogs may lead to increased PCDD/F concentrations [29–32]. Silk et al. [33] applied 36Cl– during their incorporation experiments with peat material and found most of the activity in the NaOH-extractable fraction. Enzymatically-mediated PCDD/F formation has been reported by various research groups [33 – 39]. Recently, Hoekstra et al. [40] reported a natural production of 37Cl-enriched PCDD/F in a forest soil one year after they had applied a Na37Cl solution onto the soil. This gives convincing evidence for an in situ PCDD/F formation at ambient temperatures and is ascribed to naturally formed chlorophenols as precursors. An additional natural source of PCDD/F is combustion of organic material at temperatures above 200 °C [41]. Within the temperature range of 200–650°C PCDD/F formation proceeds via chlorination of the corresponding non-chlorinated furan and dioxin which have been formed at higher temperatures. Thus, for the PCDD/F formation within combustion processes (oxidation at elevated temperatures) only a few ingredients are necessary: chloride, a redox catalyst (such as Fe or Cu), an appropriate supply of carbon (e.g. soot, graphite, etc.), oxygen and a highly reactive surface [42–43]. Recent findings of elevated dioxin concentrations in clays mined from deposits reported to be millions of years old (up to 40 million years) in disparate regions in the United States [44] and in Germany [45] with distinct unprecedented isomer patterns point to a common natural geologic mechanism to account for their origin. Indeed, there are similarities of the isomeric patterns pointing to biogenic sources [33, 41] as Rappe et al. [46–47] stated. To date, no definitive experimental evidence has been brought forward either to account for the presence of the dioxins from known anthropogenic sources or to explain the selective chemical synthesis of PCDDs under the conditions inherent to the clay formation of some 40 million years ago. On summarizing these data, PCDD/F formation prior to the mass production of man-made organohalides can be attributed to biogenic sources (enzymatically-mediated formation from appropriate educts) and to trace fire chemistry (starting from organic material and halides being mediated by redox-sensitive heavy metals such as Cu and Fe). 1.3 Polar Organohalogens (e. g. Haloacetates)
Trichloroacetic acid (TCA) is widespread in precipitations in the Northern and Southern Hemisphere and, despite large emissions of possible anthropogenic
68
H.F. Schöler and F. Keppler
precursors in the Northern Hemisphere [48], TCA concentrations in snow are not significantly different in arctic and subarctic regions than in Antarctica. A firn core from Antarctica with snow accumulated from the past 200 years exhibits haloacetic acids even in pre-industrial layers [49]. Moreover, TCA is also present in glacier ice of pre-industrial origin from Northern Sweden [50] and from Monte Rosa (Switzerland) [51], dated around 1900, before the mass production of chlorinated solvents, therefore indicating that natural sources must also exist. By mass balance calculations for TCA, which deliver a rough estimation of its fluxes in the environment, Hoekstra et al. [52–53] and Schöler [54] came to the conclusion that there must be an additional natural source of TCA in soil. To date, only one known biotic pathway via peroxidases seems to be important for the natural formation of TCA in soil [55]. The chloroperoxidase (CPO) from the fungus Caldariomyces fumago is able to produce reactive chlorine species. Starting from the educts CPO, chloride, and hydrogen peroxide, hypochlorous acid is produced which chlorinates organic material in a non-specific way. The occurrence of chloroperoxidase is ubiquitous in organisms and plants.Asplund et al. [56] and Laturnus et al. [57] observed a CPO-like activity in their soil extracts. Ballschmiter and co-workers [58 – 60] reported the CPO-mediated trichloromethane (TCM) formation from organic compounds (such as acetone, propionic acid and citric acid). Hoekstra et al. [61 – 62] demonstrated the CPO-mediated formation of TCM and TCA from humic acids. Haiber et al. [51] have carried out similar experiments, incubating humic acid and a range of naturally occurring carboxylic acids (acetic, malic, lactic, citric) with chloride ions and hydrogen peroxide, with and without CPO and confirmed that haloacetates are indeed formed. They also reported the surprising TCA formation even without the addition of the enzyme chloroperoxidase; the effect was most profound with humic acid. In addition, they also found that TCA was rapidly bound, either physically or chemically, to humic acid suggesting that, under some circumstances, an inability to detect TCA may not be evidence of its non-formation. Hoekstra et al. [63] showed that 37Cl-enriched TCM is formed in forest soil (Douglas fir), which was spiked in situ with an aqueous solution of Na37Cl (also see above). 37Cl-enriched TCA might also be produced as a coupled product, but was not measured. Frank et al. (64] and Hoekstra et al. [63, 65–66] found that in coniferous forests TCM concentrations in soil air are significantly enriched compared to ambient air pointing to proceeding chlorination processes in soil. From a chemical point of view an abiotic halogenation might be conceivable. Hydrogen peroxide can oxidise chloride to hypochlorous acid as can be derived from the standard oxidation/reduction potentials of hydrogen peroxide and chloride: E° (H2O2) = 1.776 V
E° (Cl–) = 1.36 V.
(a)
Hoekstra et al. [61 – 62] and Haiber et al. [51] have observed a TCA formation in laboratory experiments in the absence of chloroperoxidase. Further in vitro studies by Fahimi et al. [9, 11] showed that the addition of Fe(III) to the reaction mixture of humic acid, hydrogen peroxide and chloride enhanced the production of TCA. They proposed that Fe(III) was reduced by the humic substances and that
Abiotic Formation of Organohalogens During Early Diagenetic Processes
69
the Fe(II) formed started the Fenton reaction, producing highly reactive hydroxyl radicals (see below). From studies on the anthropogenic use of hypochlorous acid in disinfection, bleaching and detoxification processes TCM and TCA were found to be the main chlorinated products of low molecular weight [67]. They are formed by chlorination of the aromatic rings of the humic structures via the haloform reaction. The TCM/TCA ratio is dependent on the structural elements of the humic material: at pH below 7 the TCA formation is favoured and vice versa at pH >7. Plümacher [68] and Hoekstra et al. [65] showed a strong positive correlation between TCM and TCA. This correlation may indicate that TCA can be decarboxylated to TCM in soil or that both compounds are products from the same educt. But, the chemical decarboxylation of TCA resulting in TCM under soil conditions is expected to be very slow. The high TCA levels (300–1000 ng L–1) found in bog waters (containing high levels of dissolved organic carbon, including humic materials) were unexpected, but pointed to a net production of TCA in such media, possibly associated with their relatively acidic pH (pH 4) [51]. Chlorobenzoic acids have also been found in bog water, with the concentration of 2,4-dichlorobenzoic acid being correlated with that of TCA, suggesting a common natural source rather than one from degradation of polychlorinated biphenyls [69]. 1.4 High Molecular Polymeric Organohalogens (Halogenated Humus)
While the halide distribution in soils and sediments has extensively been investigated, to date the situation for organically bound halogens is less defined and there is an urgent need for further improvement [70–72]. In her overview, Öberg [73] stated that the chlorine content of soil organic material is similar to that of phosphorus, which means that organic matter takes up halides from the percolating water and stores it, organically bound, to appropriate structural units of the humic acids. The halides might be released again when the organic matter is subjected to changing redox conditions, providing nucleophiles for substitution reactions or even a complete mineralisation of the halogenated compound. By applying AOX measurements (a standard analytical parameter to characterise the complex high molecular weight non-volatile halogenated organic matter) to groundwater samples, Asplund et al. [74] found AOX concentrations of 230–370 µg g–1 organic matter. The ages of three different Swedish groundwaters were estimated from 14C measurements of the fulvic acid content to be between 1300 and 5200 years old; hence, these were formed well before the industrial use of chlorine and chlorinated compounds. The presence of AOX (1–80 mg L–1) in all but one of the groundwater samples drawn from 145 wells in a national survey in Denmark points to the ubiquitous presence of organohalogen compounds [75–76]. AOX measurements on a raised bog in a remote region of Sweden suggest the presence of amounts of AOX approximately 300 times larger than expected from annual deposition of organochlorine pollutants [74]. In their pioneering work entitled ‘Organohalogen compounds in aquatic sediments: anthropogenic and biogenic’, Müller and Schmitz [77] published results
70
H.F. Schöler and F. Keppler
from a dated sediment core of Lake Constance which showed elevated organohalogen contents well before 1880 when the production of man-made organohalogens started. Related studies of sediments from different origins, of peat, and of a soil humic acid all reveal the presence of significant amounts of organohalogen compounds believed to be naturally formed [78 – 81]. Even in lignites (Tertiary) and bituminous coal (Carboniferous) considerable concentrations of high molecular organohalides up to 200 mg kg–1 have been found [80]. An investigation of 26 soil samples from ten countries by Asplund and Grimvall [70] showed AOX/Corg ratios of 0.2–2.8 g Cl kg–1 Corg with a mean value of 1 g Cl kg–1 Corg. In some soils the content of organically bound halogen was even higher than the halide concentration. Special attention has been given to coniferous forest soils in which the organohalogen concentrations (µg Cl g–1 soil) decrease with increasing depth, pointing to a degradation of organic matter with depth, although the ratio mg Clorg/g Corg increases with depth [82]. Studies on leaf litter decomposition showed a net production of organically bound halides. The pool of organically bound halogens (mainly chlorine) in the soil of a spruce forest floor was estimated to be 630 kg ha–1, with an accumulation of 0.35 kg ha–1 yr–1 from litterfall and loss by leaching of 0.63 kg ha–1 yr–1 [83–85]. When introducing a contribution from throughfall of 0.38 kg ha–1 yr–1 into the balance, the overall net formation within forest soil will be 0.36 kg ha–1 yr–1. Recently, Myneni [86] demonstrated the natural formation of chlorinated hydrocarbons during early diagenetic processes of plant material. While chlorine in plants predominantly occurs as chloride, during plant decay chloride is bound to aliphatic and aromatic structures embedded as high molecular organochlorine compounds. By forming a budget of carbon turnover from plant to humic material with a yearly rate of 0.4 ¥109 t which is in situ halogenated with a ratio of mg Clorg/g Corg , the yearly net formation of organohalides accounts for 0.4 ¥ 106 t. The global inventory of soil humic acids is estimated to be around 1–1.5 ¥ 1012 t, which means that today more than 109 t of organically bound halogens are present in our soils [87–88]. Keppler and Biester [89] recently reported that organochlorines are well preserved in most parts of the peat and dechlorination processes seem to be of minor importance. This would also fit with the observation that even in lignites (Tertiary) and bituminous coal (Carboniferous) considerable organohalide concentrations up to 200 mg kg–1 have been found [80]. Based on results from Canadian and European peat bogs [33, 80, 89], the average concentration of organic chlorine in peatlands is 300–1100 mg kg–1. On the basis of total carbon stored in the Earth’s peatlands [90–91], between 0.28 and 1.0 ¥ 109 t of organically bound chlorine have been accumulated in the terrestrial ecosystem during the postglacial period (~ 10,000 years). Research has been undertaken to elucidate the structural units of these complex organic materials present in soil. Two chemical degradation methods were applied: pyrolysis and subsequent GC-MS analysis, or chemical oxidation followed by fragment analysis by GC-MS [92 – 95]. The identified fragments re-
Abiotic Formation of Organohalogens During Early Diagenetic Processes
71
sembled both those produced during wood bleaching and those lignin-derived chlorinated metabolites which fungi synthesize during wood rotting. To date, the mechanisms by which organohalides are formed in soil are poorly understood, although chloroperoxidase-type enzymes, capable of chlorinating fulvic and humic acid, are present in soils [56–57]. In addition to the characterisation of products formed during the chlorination of humic materials [72, 76, 96–97], chlorination studies were carried out with resorcinol which has been chosen as a model compound for humic acids [67]. By using isotopic labelling the reaction pathway from the educt resorcinol to the products trichloromethane and trichloroacetic acid could be established. These results suggest that a considerable agreement exists between natural chlorination processes in soil and water disinfection with molecular chlorine.
2 Prerequisites for Abiotic Early Diagenetic Processes in the Terrestrial Environment Soil contains a complex and highly dynamic mixture of minerals and organic material stemming from the decay of dead plants, animals and microorganisms. A multitude of chemical and biochemical reactions leads to products that are partly released into soil air or dissolved, solubilised or suspended into the percolating soil water, leaching through the soil into groundwater. Sediments of lakes and rivers are deposited particulate matter from the supernatant water. Sediments contain minerals, organic material and living organisms, partly dissolved in pore water, partly adsorbed on particle surfaces. By a manifold of biogeochemical processes the organic matter is slowly degraded and oxygen is depleted and nitrate and sulfate are reduced. Further anaerobic reductive transformations of organic material lead to reduced heavy metals (Fe, Mn) and finally to methane. Sediments are appropriate archives to reveal the history of their constituents during the time of sedimentation. One can say that sediment is the mind of a riverine system and for heavy metals it is mostly an open book, but for organic constituents it is sometimes a labyrinth. To disentangle the multitude of transformation processes taking place in soil and sediment all at the same time we have to look for the prerequisites of possible halogenating processes. What are the main ingredients of soil and which of them might be the starting material for naturally produced organohalides? We need inorganic chloride for the purposeful formation of organohalides – that is a matter of course. In addition, a highly reactive organic molecule that might be attacked by chloride or – vice versa – an unreactive organic molecule that might react with the highly reactive hypochlorous acid is essential. For the formation of hypochlorous acid an electron transfer from chloride to an electron acceptor is necessary; this points to redox-sensitive elements which are ubiquitous in the terrestrial environment (e.g. Fe, Mn). In the following sections we deal with: – Inorganic Halide, – Redox-Sensitive Elements (e.g. Fe), – Organic Matter.
72
H.F. Schöler and F. Keppler
2.1 Inorganic Halide
By weathering processes of primary rocks and minerals large amounts of halides have been liberated, then dissolved in water and transported via rivers to the oceans (a final sink). One litre of ocean water contains 0.5 mol chloride, 1 mmol bromide and 1 µmol iodide; the ratio of Cl– :Br– :I– is 500,000:1000:1. Part of the halides is mobilized by sea-spray or as organohalides and transported to the terrestrial environment. The atmospheric deposition of halides is dependent on the distance from the ocean and on the amount of precipitation. In addition the primary rocks and evapotranspiration contribute to the halide content of soils [98]. The mean chloride content of soils in humid climates is in the range 100–300 mg kg–1; the corresponding bromide and iodide contents are 5–50 mg kg–1 and 3–30 mg kg–1, respectively. The molar ratio of Cl– :Br– :I– in soil (~100:5:1) is distinctly shifted relative to those of ocean water, especially iodide which is enriched by a factor of 5000. 2.2 Redox-Sensitive Elements (e. g. Fe)
Next to aluminum, iron is the most abundant metal in the continental crust with a mean content of 4.2%. In primary minerals such as biotite, olivin and magnetite, iron is mostly in the +2 oxidation state and through weathering processes it is liberated, oxidised by oxygen in the presence of water and immobilised after polymerisation as Fe(III)-oxide-hydroxide. Through mineralisation processes the minerals ferrihydrite, goethite and hematite, which are responsible for the brown or red colour of soils, are formed. The Fe content of soils is in the range 0.5–5% [98].As very stable weathering products, Fe(III)-oxides reside in soil under aerobic conditions.When microbial oxidation of organic material takes place, Fe(III)-oxide serves as an electron acceptor and is reductively dissolved [99]. Diffusion to aerobic environments leads again to precipitation of Fe(III)-oxide-hydroxide. Reductive dissolution and oxidative precipitation are parts of the Fe redox-cycling under changing redox conditions [100]. Fe(III) is stable in aqueous soil solutions in oxic environments at pH values <3.5, while at pH values >4 Fe(III) concentrations are extremely low and limited to soluble organic Fe(III) complexes. This means that Fe availability in soil is determined by the interaction of poorly crystalline Fe minerals and soluble organic complexants among which are humic acids [101], microbially produced siderophores [102] and root exudates [103]. 2.3 Organic Matter
The organic matter of soil comprises the detritus from dead plants, animals, microorganisms and their transformation products. The content of organic matter varies in the different soil horizons from 100% in the O-horizon to low percentages (1–4%) in the Ah-horizon of forest or agricultural soils [98]. Plants are the
Abiotic Formation of Organohalogens During Early Diagenetic Processes
73
main supplier of the educts of organic matter. Their chemical constituents are cellulose, hemicellulose, lignine and protein (more than 95%). These compounds are the main starting material for the microbial transformation reactions and for humification. The remaining 5% consists of phenols, sugars, amino acids and peptides [104]. If this debris comes into contact with the mineral phase of the soil a prompt degradation (biotic and abiotic) takes place, which liberates CO2. The remains, for example lignine, are only slowly degraded. During the lignine depolymerisation by fungi a multitude of chemical reactions leads to humic material: – ether cleavage of methoxy groups, – hydroxylation of aromatic rings, – oxidation of alkyl chains to aldehydes or carboxylic acids. Aldehydes and organic acids might be either enzymatically decarboxylated to polyphenols or microbially oxidized and mineralised to CO2 [104].A further entry to this class of compounds starts with cellulose degradation via the biochemical turnover within microorganisms. While in anaerobic environments (water-logged as peat bogs and sediments) the lignine route prevails, under aerobic conditions in agricultural and forested areas the cellulose route is the most important. Polyphenols and the corresponding quinones are the central building blocks for the formation of refractory organic compounds in soil and can serve as redox shuttles. These highly reactive compounds are able to form covalent bonds with nucleophiles. The incomplete oxidation of polyphenols leads to radicals that initiate polymerisation with appropriate organic substrates. Both nucleophilic addition and radical polymerisation increase the molecular size of humic material and as consequence the degradation rate decreases. Humic acid material is more recalcitrant to microbial degradation, residing with only marginal changes in soil. These chaotic processes starting from varying educts result in humic compounds whose chemical composition is so diverse that only very few chemical molecules are alike. The building blocks consist of fission products from lignine, polysaccharides and proteins. These molecular units are linked to each other by -O-, -NH-, -N=, -CH2-, -CO- or -S- bridging. To get an impression of the complexity of humic material in soil see Fig. 2. The properties of humic material are mostly determined by polar functional groups, for example carboxyl, phenolic and, to an lesser extent, alcoholic moieties. The more non-polar aromatic structures stem from transformed lignine material (for detailed information see Refs. [98, 104–108]). In addition, it is well known that in soils a couple of monocyclic aromatic diols (with hydroxy groups in the o-position) produced by microorganisms can be detected though their facile microbial degradation [109]. This phenomenon can be explained through steady-state concentrations: synthesis and degradation occur with similar rates. The so-called ‘siderophores’ (i. e. Fe carriers) are a class of compounds that enable microorganisms and plants to dissolve and complex Fe(III) and to make it available for the biochemical turnover within the cell [101–103, 110–111]. Next to diols, microorganisms and plants apply hydroxamic acids for the dissolution of Fe and Mn minerals [112–113]. The distribution pro-
74
H.F. Schöler and F. Keppler
Fig. 2. Proposed structure of humic material (modified from Scheffer and Schachtschnabel
[98])
files could be unravelled by using 14C-labelled diols during degradation tests: partially 14C-labelled diols were mineralised to 14CO2, they were partly integrated into the biomass and they also partly remained as bound residue in the refractory organic substances in soil [114–116]. The integration might be caused by the enzymes phenolase and peroxidase; in addition, there are speculations about participation from Fe- and Mn-oxides during this process [117]. Catechols (o-dihydroxy aromatic compounds) such as catechol itself or 2,3-dihydroxybenzoic acid are easily oxidized by Fe(III) to the corresponding oquinone at pH values 1–2 [118]. The kinetics of this reaction were elucidated by Xu and Jordan [119–120] and suggest the reaction mechanism shown in Fig. 3. Intially, the colour of the reaction mixture turns abruptly to blue due to formation of an intermediate chelate complex. During the proceeding reaction the blue complex is oxidized to a semiquinone by a second Fe(III); the colour then changes to yellow indicating a second electron transfer and formation of quinone and Fe(II). Under oxic conditions in soils Fe(III)-oxides can be reductively dissolved by these catechols. There is no need for a catalytic participation from microorganisms or enzymes during this process. Fe(III)-oxide acts as oxidant and the organic compound is oxidized to quinone-like structures or to highly reactive radicals while Fe(III) is reduced to Fe(II) [111, 121–122]. The radical intermediates induce a multitude of side-reactions: dimerization, oligomerisation and polymerisation. Humic acids and phenolic moieties (e.g. catechol, guaiacol, 2,3-dihydoxybenzoic acid) which can be seen as building blocks of humic material in soil were applied for the experiments.As a reaction mechanism, the following sequence seems likely: adsorption of the phenolic moieties onto the mineral surface, electron transfer from the adsorbed organic compound and desorption of
Abiotic Formation of Organohalogens During Early Diagenetic Processes
75
Fig. 3. Reaction of 2,3-dihydroxybenzoic acid and Fe(III) leading to a blue coloured interme-
diate and finally to a quinone and Fe(II)
the products. The Fe(II) production rate is dependent on the following parameters: type of organic compound, pH value and crystallisation grade of the Fe(III)oxide [111, 123–125]. The amorphous ferrihydrite is more rapidly dissolved than the well-crystallised goethite or hematite, and the Fe(III) reduction rate increases with decreasing pH values.
3 Abiotic Formation of Organohalogens in the Terrestrial Environment The previous sections have described the biogeochemical turnover of carbon and halide in the terrestrial environment. Halogenation processes taking place in this compartment are mostly ascribed to the omnipresence of biota. Natural abiotic halogenation reactions are also known but have scarcely been investigated. Thermodynamic considerations show it is possible that halide ions may form organohalide compounds naturally by purely chemical processes which are known to occur in vitro. There are reports that chloromethane may arise from bromomethane or iodomethane by a simple nucleophilic substitution reactions involving chloride ions [126–130]. However, recent findings from Coulter et al. [131] pointed out that this exchange reaction might be enzyme-mediated. MeBr + Cl– Æ MeCl + Br–
(1)
Highly reactive compounds (e. g. epoxy compounds or quinones) are prerequisites for the following reaction types: the first example is the synthesis of chlorogentisylol. The addition of chloride to the educt epoxidione leads to a 1,2-chloro-
76
H.F. Schöler and F. Keppler
Fig. 4. Formation of chlorogentisylol by nucleophilic addition of chloride to a highly reactive epoxy compound [132]
hydrin that is then reduced to chlorogentisylol. This reaction scheme was verified by Nabeta et al. [132] (Fig. 4). Quinones show chemical properties of a b-unsaturated ketone that can be attacked by Cl– at the 4-position. The nucleophilic addition of chloride to p-quinone leads to an intermediate that results in chlorohydroquinone after re-aromatisation [133] (see Fig. 5). This reaction type, repeated four times, was formerly applied in industry for the production of tetrachloro-p-quinone (p-chloranil) as a technical product. p-Chloranil was used as a herbicide until it was discovered that this compound was heavily contaminated with PCDDs and its production was ceased immediately. Similar reactions might occur in the terrestrial environment that could be responsible for small PCDD concentrations in ancient sediment layers and archived soil samples. A very similar reaction between humic acid and iodide was investigated by Rädlinger and Heumann [134–135]. They stirred an aqueous solution of humic acid and iodide for one hour and analysed the reaction products by size-exclusion chromatography and ICP-MS and found that iodide was chemically bound to the humic acid backbone, especially within the high molecular fraction. The reaction type is conceivable if quinonic moieties are an integral part of the humic acid structure. Recently, a new abiotic halogenation reaction was reported by Keppler et al. [6], which forms alkyl halides in the aerobic layer of soil. The thermodynamically labile organic matter is oxidized and the redox partner Fe(III) is reduced to
Fig. 5. Formation of chlorohydroquinone by nucleophilic addition of chloride to p-quinone
used for the industrial production of tetrachloro-p-quinone
Abiotic Formation of Organohalogens During Early Diagenetic Processes
77
Fig. 6. Model for alkyl halide formation by the reaction of Fe(III) and organic matter in the presence of halide ions
Fe(II). Phenolic moieties of the natural organic matter containing alkoxy groups might be oxidized while Fe(III) is reduced. During this process halides (Cl, Br, I) present in soils are alkylated, and the alkyl halides (methyl, ethyl, propyl and butyl halides) formed represent degradation products of oxidized organic matter (Fig. 6). As organic matter in soil displays a highly complex polymeric structure, it is difficult to describe chemical reactions taking place in soil. To reduce this complexity, small molecules – so-called model compounds – are used to represent the structural elements or redox features of the organic matter. Widely accepted model compounds for aromatic structures are catechol, hydroquinone, resorcinol, guaiacol and 2,3-dihydroxybenzoic acid (see Fig. 7).
Fig. 7. Monomeric structural units of soil humic matter
78
H.F. Schöler and F. Keppler
Fig. 8. Oxidation reaction of guaiacol with ferrihydrite producing methyl halides
Fig. 9. Reaction scheme for the guaiacol oxidation with Fe(III)
One of these natural monomeric constituents, guaiacol, was used as a methylgroup donor for the oxidation reaction with dissolved Fe(III) or with the mineral ferrihydrite (5 Fe2O3 · 9 H2O) and halides. Methyl halides, Fe(II) and o-quinone have been identified as reaction products (Fig. 8). It is assumed that methyl halides are produced in an almost synchronous reaction scheme: (1) the oxidation of guaiacol by ferrihydrite and (2) nucleophilic substitution of the methyl group by halide (Fig. 9). Keppler et al. [10] recently described a natural formation of the highly reactive chlorinated compound vinyl chloride (VC) in soil. In this case, they consider catechol as a model for the redox-sensitive functional aromatic groups of soil organic matter and the corresponding o-quinone (Fig. 2) as the intermediate precursors for vinyl chloride. Catechol also plays a key role within the biochemical degradation pathway of aromatic compounds. Previous laboratory experiments with catechol have shown that it can be oxidized by Fe(III) producing CO2 [7] and, if halides are added, alkyl halides [12] (Fig. 10). The CH3Cl/VC ratio was about eight. There was no VC or CH3Cl formation when Fe(III) was absent. Moreover, no VC production was observed by using H2O2 , another naturally occurring oxidant.When both oxidants Fe(III) and H2O2 were applied the VC production increased significantly, probably caused by the Fenton reaction by which H2O2 and Fe(II) generate hydroxyl radicals. The prerequisite Fe(II) is provided by the reaction of catechol with Fe(III). OH radicals are powerful oxidants and could be responsible for the augmented formation of VC and CH3Cl. OH radicals are also the topic of a very recent paper of Fahimi et al. [9, 11] which deals with the abiotic formation of haloacetates from soil, commercially
Abiotic Formation of Organohalogens During Early Diagenetic Processes
79
Fig. 10. Reaction scheme for the catechol oxidation with Fe(III) producing vinyl chloride
Fig. 11. The formation of DCA and TCA from humic acid increases dose-dependently with the concentration of H2O2 and Fe2(SO4)3
available humic acid and phenolic model substances. It is shown that haloacetates are formed from humic material with a linear relationship between the humic acid used and haloacetates formed (see Fig. 11). More dichloroacetate (DCA) than trichloroacetate (TCA) is formed. The addition of Fe(II), Fe(III) and H2O2 leads to an increased yield. Furthermore, the relationship between structure and reactivity of phenolic substances, which can be considered as monomeric units of humic substances, has been examined. Ethoxyphenol with built-in ethyl groups forms large amounts of DCA and TCA. With other phenolic substances a cleavage of the aromatic ring was observed.
80
H.F. Schöler and F. Keppler
Fig. 12. Proposed reaction scheme for formation of trichloroacetic acid and trichloromethane
[53]
These investigations clearly indicate that haloacetates are formed abiotically from humic acid and soils. Hydroxyl radicals and chloride form an equilibrium system with the hypochlorous acid anion and the chloride anion [136]. The latter anions react with each other forming chlorine that induces the so-called swimming pool chemistry. An alternative explanation for the observed effect of the addition of Fe(III) could be the formation of a humic acid/Fe3+-complex that acts similar as a heme group in the CPO-mediated production of hypochlorous acid [53]. From studies on the anthropogenic use of hypochlorous acid, TCM and TCA were found to be the main chlorinated products of low molecular weight [67]. They are formed by chlorination of the aromatic rings of the humic structures via the haloform reaction as presented in Fig. 12. Resorcinol is a common structural element of humic material. Most of the reported reaction schemes for the abiotic halogenation in the terrestrial environment are linked to radical chemistry by two essential redox-sensitive constituents of soil: iron and organic matter. Perhaps oxygen is a third reaction partner. It seems to be by chance that halides are involved in these reactions. Halides are soil constituents and interfere with these soil processes by their mere presence.
4 Outlook The discovery of abiotic halogenating processes at ambient temperatures – socalled early diagenetic processes – in the terrestrial environment has astonished the scientific community. But, we have only lifted a tip of the blanket. The extent
Abiotic Formation of Organohalogens During Early Diagenetic Processes
81
of these processes is to date far from being realistically estimated. One has to bear in mind. – That reduction of insoluble Fe(III)-oxides is one of the most significant geochemical processes that takes place in the sedimentary environment. – Worldwide, 1500–2200 Gt of organic carbon is stored as humic matter. – There is a sufficient supply of halide from rock weathering and via precipitation. In addition there is a big gap in the Earth’s budget between the globally produced amount of methyl chloride and methyl bromide and their actual atmospheric concentrations. This gap might be plugged by biotic and abiotic sources from the terrestrial environment. Acknowledgement. The authors express their gratitude to I. Fahimi, G. Kilian and K.Wolkenstein for reviewing the manuscript.
5 References 1. Crutzen PJ, Andreae MO (1990) Science 259:1669 2. Andreae MO (1993) In: Oremland RS (ed) Biogeochemistry of global change: Radiatively active trace gases. Chapman and Hall, New York, p 113 3. Andreae MO, Atlas E, Harris GW, Helas G, Kock A, Koppmann R, Maenhout W, Mano S, Pollock WH, Rudolph J, Scharffe D, Schebeske G, Welling M (1996) J Geophys Res 101: 23603 4. Jordan A, Harnisch J, Borchers R, Le Guern F, Shinohara H (2000) Environ Sci Technol 34:1122 5. Harnisch J, Frische M, Borchers R, Eisenhauer A, Jordan A (2000) Geophys Res Lett 27:1883 6. Keppler F, Eiden R, Niedan V, Pracht J, Schöler HF (2000) Nature 403:298 7. Pracht J, Boenigk J, Isenbeck-Schröter M, Keppler F, Schöler HF (2001) Chemosphere 44:613 8. Keppler F (2001) Proceedings 2nd international conference on ‘naturally produced organohalogens’, Heidelberg 9. Fahimi IJ, Keppler F, Seiss M, Schöler HF (2001) Poster presented at the 2nd international conference on ‘naturally produced organohalogens’, Heidelberg 10. Keppler F, Borchers R, Pracht J, Rheinberger S, Schöler HF (2002) Environ Sci Technol 36:2484 11. Fahimi IJ, Keppler F, Schöler HF (2002) Chemosphere (in press) 12. Keppler F, Borchers R, Elsner P, Fahimi I, Pracht J, Schöler HF (2002) Chemosphere (in press) 13. Gribble GW (2002) Chemosphere (in press) 14. Winterton N (2000) Green Chemistry 2:173 15. Harper DB (1985) Nature 315 : 55 16. Yokouchi Y, Ikeda M, Inuzuka Y, Yukawa T (2002) Nature 416 : 163 – 165 17. Dimmer CH, Simmonds PG, Nickless G, Bassford MR (2001) Atmos Environ 35:321 18. Redeker KR,Wang NY, Low JC, McMillan A, Tyler SC, Cicerone RJ (2000) Science 290:966 19. Rhew RC, Miller BR, Weiss RF (2000) Nature 403 : 292 20. Laturnus F, Lauritsen FR, Gron C (2000) Water Resour Res 36:2999 21. Hoekstra EJ, Duyzer JH, Leer de EWB, Brinkman UAT (2000) Atmos Environ 35:61 22. Hoekstra EJ, Verhagen FJM, Field JA, De Leer de EWB, Brinkman UAT (1998) Phytochemistry 49:91
82 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68.
H.F. Schöler and F. Keppler Brush RC, Rice GW (1994) J Chem Educ 71:293 Reckhow DA, Singer PC, Malcolm RL (1990) Environ Sci Technol 24:1655 Kjeller LO, Jones KC, Johnston AE, Rappe C (1996) Environ Sci Technol 30:1398 Alcock RE, McLachlan MS, Johnston AE, Jones KC (1998) Environ Sci Technol 32:1580 Jüttner I, Henkelmann B, Schramm K-W, Steinberg CEW,Winkler R, Kettrup A (1997) Environ Sci Technol 31:806 Kjeller LO, Rappe C (1995) Environ Sci Technol 29:346 Öberg LG, Wagman N, Andersson R, Rappe C (1992) Organohalogen Compd 11:297 Öberg LG, Wagman N, Andersson R, Rappe C (1996) Chem Abstr 124:126,031h Sievers S, Schacht U (1996) Chem Abstr 124:126,032j Fiedler H (1996) Chem Abstr 124:126,033k Silk PJ, Lonergan GC, Arsenault TL, Boyle CD (1997) Chemosphere 35:2865 Svenson A, Kjeller LO, Rappe C (1989) Environ Sci Technol 23:900 Svenson A, Kjeller LO, Rappe C (1989) Chemosphere 19:585 Hemming J, Holmbom B, Larsson B (1991) Org Micropollut Aquatic Environ 21:523 Öberg LG, Rappe C (1992) Chemosphere 25:49 Vikelsøe J, Lassen P, Johansen E, Carlsen L (1994) Organohalogen Compd 20 : 351 Morimoto K, Tatsumi K (1997) Chemosphere 34 : 1277 Hoekstra EJ,Weerd de H, Leer de EWB, Brinkman UAT (1999) Environ Sci Technol 33:2543 Wikström E, Tysklind M, Marklund S (1999) Environ Sci Technol 33:4263 Hisham MWM, Benson SW (1995) J Phys Chem 99:6194 Addink R, Altwicker ER (1998) Environ Eng Sci 15 : 19 Ferrario JB, Byrne CJ, Cleverly DH (2000) Environ Sci Technol 34:4524 Jobst H, Aldag R (2000) Umwelt Chem Okotox 1:2 Rappe C, Bergek S,Andersson R, Cooper K, Fiedler H, Bopp R, Howell F, Bonner M (1999) Organohalogen Compd 43:111 Rappe C, Bergek S, Fiedler H, Cooper K (1998) Chemosphere 36 : 2705 McCulloch A (2002) Chemosphere 47:667 Sydow von L, Nielsen AT, Grimvall A, Boren H (2000) Environ Sci Technol 34:239 Sydow von L, Boren H, Grimvall A (1999) Chemosphere 39 :2479 Haiber G, Jacob G, Niedan V, Nkusi G, Schöler HF (1996) Chemosphere 33 : 839 Hoekstra EJ, Leer de EWB, Brinkman UAT (1999) Chemosphere 38:551 Hoekstra EJ (2002) Chemosphere (in press) Schöler HF (1998) Fluxes of trichloroacetic acid between atmosphere, biota, soil, and groundwater. Report to Eurochlor, Brussels Neidleman SL, Geigert J (1986) Biohalogenation: principles, basic roles and applications, Ellis Horwood, Chichester, England Asplund G, Christiansen JV, Grimvall A (1993) Soil Biol BioChem 25:41 Laturnus F, Mehrtens G, Grön C (1995) Chemosphere 31:3709 Walter B, Ballschmiter K (1991) Fresenius J Anal Chem 341:564 Walter B, Ballschmiter K (1992) Fresenius J Anal Chem 342:827 Urhahn T, Ballschmiter K (1998) Chemosphere 37:1017 Hoekstra EJ, Leer de EWB (1993) In: Arendt F, Annokklee GJ, Bosman R, Brink van den WJ (eds) Contaminated soil ‘93. Kluwer, Dordrecht, p 215 Hoekstra EJ, Lassen P, Leeuwen van JGE, Leer de EWB, Carlsen L (1995) In: Grimvall A, Leer de EWB (eds) Naturally-produced organohalogens. Kluwer, Dordrecht, p 149 Hoekstra EJ, De Leer de EWB, Brinkman UAT (1998) Environ Sci Technol 32:3724 Frank H, Frank W, Thiel D (1989) Atmos Environ 23 : 1333 – 1335 Hoekstra EJ, De Leer de EWB, Brinkman UAT (1999) Chemosphere 38 : 2875 Hoekstra EJ, Duyzer JH, De Leer de EWB, Brinkman UAT (2001) Atmos Environ 35:61 Boyce SD, Hornig JF (1983) Environ Sci Technol 17:202 Plümacher J (1995) Untersuchung verschiedener Kiefernstandorte in Berlin und Umgebung hinsichtlich ihrer Immissionsbelastung durch leichtflüchtige Chlorkohlenwasserstoffe (LCKW) und Trichloressigsäure. PhD Thesis, Schriftenreihe des Fachbereichs Umwelt und Gesellschaft no. 95, Technische Universität Berlin, Berlin
Abiotic Formation of Organohalogens During Early Diagenetic Processes
83
69. Niedan V, Schöler HF (1997) Chemosphere 35 : 1233 70. Asplund G, Grimvall A (1992) Environ Sci Technol 25:1346 71. Asplund G (1995) In: Grimvall A, Leer de EWB (eds) Naturally-produced organohalogens. Kluwer, Dordrecht, p 35 72. Grimvall A (1995) In: Grimvall A, Leer de EWB (eds) Naturally-produced organohalogens. Kluwer, Dordrecht, p 3 73. Öberg G (1998) Acta Hydrochim Hydrobiol 26:137 74. Asplund G, Grimvall A, Pettersson C (1989) Sci Total Environ 81/82:239 75. Grön C (1991) In: Allard B, Boren H, Grimvall A (eds) Humic substances in the aquatic and terrestrial environment. Springer, Berlin Heidelberg New York, 33:485 76. Grön C (1995) In: Grimvall A, Leer de EWB (eds) Naturally-produced organohalogens. Kluwer, Dordrecht, p 49 77. Müller G, Schmitz W (1985) Chem Ztg 109:415 78. Paasivirta J, Knuutinen J, Maatela P, Paukku R, Soikkeli J, Sarkka J (1988) Chemosphere 17:137 79. Grön C, Rabenlange B (1992) Sci Total Environ 113:281 80. Müller G, Nkusi G, Schöler HF (1996) J Prakt Chem/Chem Ztg 338 : 23 81. van Metre PC, Callender E, Fuller CC (1997) Environ Sci Technol 31:2339 82. Hjelm O, Johansson M-B, Öberg-Asplund G (1995) Chemosphere 30 : 2353 83. Öberg G, Nordlund E, Berg B (1996) Can J For Res 26:1040 84. Öberg G, Grön C (1998) Environ Sci Technol 32:1573 85. Öberg G, Johansen C, Grön C (1998) Chemosphere 36:1689 86. Myneni SCB (2002) Science 295:1039 87. Hays MHB, MacCarthy P, Malcolm RI, Swift RW (1989) Humic substances. Search for structure. Wiley-Interscience, New York 88. Gribble GW (1996) Prog Chem Org Nat Prod 68:1 89. Keppler F, Biester H (2002) Chemosphere (in press) 90. Post WM, Emanuel WR, Zinke PJ, Stangenberger AL (1982) Nature 298:156–159 91. Post WM (2002) In: Lal R (ed) Encyclopedia of soil science. Dekker, New York (in press) 92. de Lijser HJP, Erkelens C, Knol A, Pool W, Leer de EWB (1991) In: Allard B, Boren H, Grimvall A (eds) Humic substances in the aquatic and terrestrial environment. Springer, Berlin Heidelberg New York, 33:485 93. Dahlman O, Reimann A, Ljungquist P, Mörck R, Johansson C, Boren H, Grimvall A (1994) Water Sci Technol 29:81 94. Dahlman O, Mörck R, Ljunguist P, Reimann A, Johansson C, Boren H, Grimvall A (1993) Environ Sci Technol 27:1616 95. Flodin C, Johansson E, Boren H, Grimvall A, Dahlman O, Mörck R (1997) Environ Sci Technol 31:2464 96. Asplund G, Boren H, Carlsson U, Grimvall A (1991) In: Allard B, Boren H, Grimvall A (eds) Humic Substances in the Aquatic and Terrestrial Environment. Springer, Berlin Heidelberg New York, 33:475 97. Grimvall A, Laniewski K, Boren H, Jonsson S, Kaugare S (1994) Toxicol Environ Chem 46:183 98. Scheffer F, Schachtschabel P (1998) Lehrbuch der Bodenkunde. Ferdinand Enke, Stuttgart 99. Lovley DR (1991) Microbiol Rev 55:259 100. Stumm W, Morgan JJ (1981) Aquatic chemistry. Wiley, New York 101. Khan TR, Cooper LH, Skippen GB (1984) Org GeoChem 7 :261 102. Goodell B, Jellison J, Liu L, Daniel G, Paszczynski A, Fekete F, Krishnamurthy S, Jun L, Xu G (1997) J BioTechnol 53 : 133 103. Römheld V, Marschner H (1983) Plant Physiol 71:949 104. Haider K (1996) Biochemie des Bodens. Enke, Stuttgart 105. Aiken GR, McKnight DM,Wershaw RL, MacCarthy P (1985) Humic substances in soil, sediment, and water. Wiley, New York 106. Schnitzer M, Khan SU (1974) Humic Substances in the Environment. Dekker, New York
84 107. 108. 109. 110. 111. 112. 113. 114. 115. 116. 117. 118. 119. 120. 121. 122. 123. 124. 125. 126. 127. 128. 129. 130. 131. 132. 133. 134. 135. 136.
H.F. Schöler and F. Keppler: Abiotic Formation of Organohalogens Schnitzer M, Khan SU (1978) Soil organic matter. Elsevier, Amsterdam, New York Schulten HR, Schnitzer M (1997) Soil Science 162:115 Whitehead DC, Dibb H, Hartley RD (1983) Soil Biol Biochem 15:133 Fritsche W (1998) Umweltmikrobiologie – Grundlagen und Anwendungen. Fischer, Stuttgart LaKind JS, Stone AT (1989) Geochim Cosmochim Acta 53:961 Waid JS (1975) In: Paul EA, McLaren AD (eds) Soil biochemistry. Dekker, New York, 4:65–101 Marschner H (1986) The mineral nutrition of higher plants. Academic Press, London Haider K, Martin JP (1975) Soil Sci Soc Am J 39:657 Martin JP, Haider K, Linhares LF (1979) Soil Sci Soc Am J 43:100 Haider K (1992) In: Stotzky G, Bollag J-M (eds) Soil biochemistry. Dekker, New York, 7:55 Lehmann RG, Cheng HH, Harsh JB (1987) Soil Sci Soc Am J 51:352 Mentasti E, Pelizetti E, Saini G (1973) J Chem Soc 2609 Xu J, Jordan RB (1988) Inorg Chem 27 : 1502 Xu J, Jordan RB (1988) Inorg Chem 27 : 4563 Mazellier P, Bolte M (1997) Chemosphere 35:2181 Desjardins S, Landry JA, Farant JP (1999) Clay J Soil Contam 8:175 Waite TD, Morel FMM (1984) Environ Sci Technol 18:860 Suter D, Banwart S, Stumm W (1991) Langmuir 7:809 Deng Y (1992) Formation and dissolution of aquatic iron (iii)(hydro)oxides – Implications for redox cycling of iron in natural waters. PhD Thesis, Swiss Federal Institute of Technology, Zürich Moelwyn-Hughes EA (1938) Proc R Soc London Ser A 164:295 Zafiriou OC (1975) J Mar Res 33:75 Elliott S, Rowland FS (1993) Geophys Res Lett 20:1043 Elliott S, Rowland FS (1995) J Atmos Chem 20:229 Jeffers PM, Wolfe NL (1996) Geophys Res Lett 23 : 1773 Coulter C, Hamilton JTG, McRoberts WC, Kulakov L, Larkin MJ, Harper DB (1999) Appl Environ Microbiol 65:4301 Nabeta K, Ichihara A, Sakamura S (1975) Agric Biol Chem 39:409 Beyer H, Walter W (1991) Lehrbuch der Organischen Chemie. Hirzel, Stuttgart Rädlinger G and Heumann KG (1997) Fresenius J Anal Chem 359:430 Rädlinger G, Heumann KG (2000) Environ Sci Technol 34:3932 Herrmann H, Ervens B, Nowacki P, Wolke R, Zellner R (1999) Chemosphere 38 : 1223
The Handbook of Environmental Chemistry Vol. 3, Part P (2003): 85–101 DOI 10.1007/b 10449
Marine Sources of Volatile Organohalogens Robert M. Moore Department of Oceanography, Dalhousie University, Halifax, Nova Scotia, Canada E-mail:
[email protected]
The ocean is a massive reservoir of the halogens, elements which play important roles in the atmosphere, most notably in their catalysis of ozone decomposition. While ionic halides are emitted to the atmosphere as sea salt particles, it is the much smaller fluxes of halogenated organic gases that can provide a pathway for halogens to reach the upper atmosphere This chapter reviews what is known about the processes, mainly biological, that effect the conversion of seawater halides to volatile halogenated gases. Particular attention is given to the methyl halides, which have become the best-studied group of compounds, in part as a result of studies focussed on natural sources of methyl bromide. Since the ultimate contribution of marine processes to atmospheric halogens depends on the net fluxes to or from the atmosphere, a glimpse is provided of marine sinks: those processes that are consuming halogenated gases in the ocean. Keywords. Volatile organohalogens, Marine production
1
Introduction
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
2
Source Types
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
2.1 2.2 2.3
Macroalgal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 87 Microalgal Sources . . . . . . . . . . . . . . . . . . . . . . . . . . 88 Chemical Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
3
Determination of Fluxes . . . . . . . . . . . . . . . . . . . . . . . 92
4
Methyl Halides . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
4.1 4.2 4.3
Methyl Chloride . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93 Methyl Bromide . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94 Methyl Iodide . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
5
Polychlorinated Compounds . . . . . . . . . . . . . . . . . . . . . 96
5.1 5.2 5.3
Chloroform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 Trichloroethylene and Tetrachloroethylene . . . . . . . . . . . . . 97 Dichloromethane . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
6
Sinks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 98
7
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 © Springer-Verlag Berlin Heidelberg 2003
86
R.M. Moore
1 Introduction Of the vast array of naturally-produced halogenated compounds, this chapter will focus on a small subset comprising those for which there is evidence of significant marine sources, and which are sufficiently volatile to be emitted to the atmosphere. A major reason for interest in these compounds stems from the now wellknown involvement of halogen chemistry in the control of stratospheric ozone levels. The role of naturally-produced methyl chloride as a source of stratospheric chlorine was first reported by Lovelock in 1975 [1]. With co-workers he had already suggested [2] that methyl iodide is produced in the sea, and that marine algae were a source. Bromine and iodine also subsequently received attention as potential contributors to stratospheric ozone loss [3–5]. In this context the importance of organohalogen compounds lies in their ability to transfer halogen from the Earth’s surface reservoirs, of which the ocean is the most important, to the lower and upper atmosphere. For transport to the upper atmosphere, it is essential that the halogen should be in a form that has low aqueous solubility so that it can avoid rapid return to the surface in precipitation. It is for this reason that organohalogens are much more important to upper atmosphere chemistry than the vastly greater quantity of particulate sea salt that enters the atmosphere, this material being quickly washed from the atmosphere. The proportion of a compound that mixes into the stratosphere increases with its tropospheric lifetime. In the case of methyl chloride, for example, which has an atmospheric lifetime of ~1.4 years, about 5% will mix into the stratosphere where the halogen can enter catalytic ozone-destroying reactions. However, it has been shown that even very short-lived compounds can be locally lofted to the lower stratosphere by intense convection in the atmosphere [6]; this is relevant to possible involvement of iodinated and polybrominated compounds in stratospheric ozone regulation [4]. Furthermore, Dvortsov et al. [5] have recently re-evaluated the significance of shortlived bromine compounds, particularly bromoform, to the supply of reactive bromine to the lower stratosphere, a region in which current models have had difficulty in accounting for observed ozone loss. After it was recognised that synergistic interactions between chlorine and bromine were important [7] and could account for ~ 20 % of polar stratospheric ozone depletion [8], the problem of apportioning methyl bromide releases to natural and anthropogenic sources stimulated much research into its natural sources in both terrestrial and marine environments [9 – 16]. In parallel, studies have been made of the marine natural sources of atmospheric chlorine and iodine, namely methyl chloride [17, 18] and methyl iodide [19–22]. Halogenated compounds are significant to tropospheric as well as stratospheric chemistry. Natural organobromine compounds have been studied in the Arctic as possible sources of bromine atoms that could initiate catalytic reactions leading to the sudden loss of tropospheric ozone that has been observed in springtime [23], though subsequently the bromine source has been considered to be inorganic sea salt bromide [24]. Efforts have been made to quantify the total emission of reactive chlorine to the atmosphere and to map the emissions on a global grid [25]. Since there are
Marine Sources of Volatile Organohalogens
87
relatively good estimates of the atmospheric lifetimes of these compounds, together with sometimes uncertain measurements of their atmospheric abundance, it has been possible to calculate the size of the annual atmospheric sinks. Then, from best estimates of industrial production and release, it has also been possible to estimate the natural sources such that a balance exists between sources and sinks. For some compounds (e. g. chloroform) this points to important natural sources among which the ocean may be numbered. Such exercises are, however, subject to many uncertainties, and it will be shown that apparent imbalances between sources and sinks are not always matched by good evidence for marine production. This chapter will look at possible marine sources of volatile organohalogen compounds, and then show how estimates are made of the global source strength to the atmosphere. While its primary aim is to focus on sources, it will be shown that marine sinks commonly coexist with sources, so what is known of these sinks will be briefly reviewed. Some of the best studied compounds, individually or as groups, will then be discussed.
2 Source Types 2.1 Macroalgal
Macroalgal sources of volatile, as well as non-volatile, compounds are the best understood. This is probably largely due to the greater ease with which significant quantities of individual species of macroalga, in contrast with microalga, may be harvested from the ocean and studied. This facilitates the measurement of release rates of halogenated compounds, and also the biochemical studies that are needed to identify the production mechanisms. A 1979 review [26] of halogenated compounds present in marine red algae of the family Bonnemaisoniaceae revealed an extremely wide variety of halogenated metabolites present within these macrophytes including, among the C1 compounds, methyl iodide, chloroiodomethane, trihalomethanes including CHBr3 , CHCl3 , CHBr2Cl, CHBr2I, CHBrCl2 , and fully-halogenated species, CBr4 and CCl4 . The release into seawater of these and similar compounds by various red, brown and green algae was studied by Gschwend et al. [27]. Their findings included release rates of the brominated species CHBr3 , CHBr2Cl and CH2Br2 , which were not only high, but also large in proportion to the concentrations within tissues, suggesting either rapid transport from the tissue, or synthesis near the plant surface. They reported the presence of a number of alkyl iodides, including ethyl iodide, isopropyl iodide, 1-iodopropane, 1-iodobutane, 1iodopentane, within the tissues of some macrophyte species. Also identified at lower abundances were the dihalomethanes, chloroiodomethane, bromoiodomethane, and diiodomethane. The work provided evidence for seasonal variations in release of polybromomethanes. Based on estimates of global algal biomass, calculations were made of the potential total organobromine and organoiodine releases to the atmosphere; these suggested an important
88
R.M. Moore
bromine source, but a minor source of iodinated compounds. Recent work [28] has added the compounds 1,2-dibromoethylene and tribromoethylene to the list of macrophyte products; they were released by the Falkenbergia stages of the red alga, Asparagopsis taxiformis. From field studies, there are reports of elevated levels of various organohalogens, including bromoform and methyl iodide in coastal waters and in seaweed beds [1, 29, 30]. Nightingale et al. [31] have measured halocarbon release rates from 11 common species of macroalgae in laboratory experiments, and reported broad agreement with in situ measurements in a rock pool. Among the compounds showing significant release rates was CHCl3 , which they proposed was a product of chloroperoxidase activity. A subsequent rock pool study by Baker et al. [32] provided confirmation of natural production of 1,2-dibromoethylene, and also the first measurements showing production of dichloromethane. Estimates of the global contributions of volatile organohalogens made by macroalgae are for most compounds insignificant [19, 32], the primary reason being that the coastal zone which they occupy accounts for a small area, equivalent to only 0.3 % of the global ocean. 2.2 Microalgal Sources
Early observations that macrophytes are frequently producers of halogenated compounds led to an interest in whether microalgae might also prove to be a source. Such organisms are potentially very much more important than macrophytes since they differ by having an ocean-wide distribution and could consequently make a much larger contribution to ocean-atmosphere fluxes. Furthermore, early reports [33–35] of elevated concentrations or supersaturations of various halogenated compounds in offshore waters suggested a non-macrophyte source. Sturges et al. [36] demonstrated that Arctic ice algae (mainly pennate diatoms) were a source of bromoform. Laboratory studies of unialgal phytoplankton cultures have been of value in determining whether selected species phytoplankton can produce organohalogens. It should be noted that, in many of these studies, both methyl halides and trihalomethanes have not been measured; this reflects both analytical limitations and, in many cases, a focus on a particular compound, for example methyl bromide or bromoform. Also, it should be noted that cold-water diatoms have received more attention than warm-water species on account of early interest in sources of Arctic bromine [23] that might lead to surface ozone loss. A survey of ten species of warm- and cold-water phytoplankton by Tokarczyk and Moore [37] revealed two species that produced organohalogens from the suite measured that comprised CH2Br2 , CHBr3 , CHBr2Cl and CH2ClI. Each of the cold-water marine diatoms, Nitzschia (CCMP 580) and Porosira glacialis produced the four measured organohalogens, while two other Nitzschia species produced none. A comparison was made of the release rates of bromoform per unit biomass of microalgae and macrophytes using results reported by Manley et al. [30]. This showed that, although the phytoplankton releases were 10 – 100 times lower, these microorganisms had the potential to be an important source of
Marine Sources of Volatile Organohalogens
89
volatile organohalogens because they occupy an ocean area about 200 times greater than that occupied by macrophytes. A follow-up laboratory study [38] which included the same halocarbon-producing Arctic diatoms showed that Nitzschia sp. CCMP 580 and Nitzschia arctica produced CHBr3 , CH2Br2 , CH2I2 and CH2ClI (Nitzschia seriata produced none of these compounds). Porosira glacialis produced the same compounds except for CH2I2 and CH2CII. Two species (Nitzschia sp. CCMP 580 and Navicula sp. CCMP 545) were then grown in bulk and tested for haloperoxidases. The Nitzschia was found to contain a bromoperoxidase, probably heme-containing, while the Navicula contained an iodoperoxidase which was not characterized. Separate studies have shown marine phytoplankton to be sources of methyl halides. Methyl chloride production was demonstrated in all of seven species of cultured phytoplankton, which included both cold-water (diatoms: Porosira glacialis, Nitzschia seriata, Nitzschia sp. CCMP 580), and warm-water species (diatoms: Odontella mobiliensis, Phaeodactylum tricornutum, Thalassiosira weissflogii; prymnesiophyte: Isochrysis galbana) [39].A subsequent study [12] showed that three species of phytoplankton grown in culture (Phaeodactylum tricornutum, Thalassiosira weissflogii and Phaeocystis sp.) all produced methyl bromide as well as methyl chloride, with the Phaeocystis giving the highest production rate for both compounds. It was found that the production was generally unaffected by limitation of either carbon or nitrogen, by the presence or absence of bacteria, by darkness, or by poisoning with sodium azide. It was concluded that production of these two compounds might be the result of an autolytic process rather than a direct product of cell metabolism. An observed lack of influence of light on the methyl chloride and bromide production is in marked contrast with studies of the production of bromoform by either microalgae [38, 40, 41] and is consistent with distinctly different production mechanisms for the two classes of compound [42].A further study [14] of a wider selection of phytoplankton comprising nine species, illustrated the widespread production of methyl chloride (all species produced this compound) and methyl bromide (produced by all but two). There was a rather constant ratio (average 7.4, CH3Cl:CH3Br) in the production rates of these two methyl halides, and an attempted global extrapolation using the rates from the most productive species (Phaeocystis) suggested that they could account for 40 – 130 % of the estimated ocean-to-atmosphere flux of CH3Cl [18] and 30% of the global CH3Br estimated by Lobert et al. [10]. However, since these estimates are upper limits based on Phaeocystis production, the work suggested that to account for reported ocean production of methyl halides, other marine sources might exist, possibly zooplankton or fungi. A study of CH3Br production in cultured marine phytoplankton [43] showed that 13 of 19 species tested produced methyl bromide, with producers amongst all of the six taxonomic classes represented in the experiments. No trend with respect to temperature was apparent in CH3Br concentrations normalised to chlorophyll for these species that were grown at temperatures of 4–22°C. It was observed that CH3Br increased most rapidly in the stationary and senescent phases of culture growth. Bacterial growth that occurred in most of the cultures may play a role in CH3Br production or release from cells, and it is also possible that it could be responsible for some removal of methyl bromide so that the mea-
90
R.M. Moore
sured release rates represent minimum values. It appeared that production rates were of the right order of magnitude to account for estimated global ocean production [10]. It was significant that again the ubiquitous, bloom-forming prymnesiophyte, Phaeocystis, showed the highest rate of production normalised to chlorophyll a. A survey [20] of methyl iodide production by 15 species of cultured marine phytoplankton revealed production in five. Bacteria were present in most of the cultures, but the CH3I production in three diatom cultures was probably attributable to the algae rather than bacteria, since its rate slowed with the algal growth while bacterial numbers increased. Based on these measurements, it was estimated that phytoplankton could make an insignificant contribution to the ocean production of the gas. A study of the red microalga Porphyridium purpureum [44] showed CH3I production about 40 times faster than that found by Manley and de la Cuesta [20] for Porosira glacialis. It is noteworthy that Porphyridium purpureum also produced CHCl3 (though not CH2Cl2). However, its restricted distribution makes it unlikely to be a globally important producer of either compound. Although experiments of the kind described above have been of value in demonstrating the ability of phytoplankton to produce various volatile organohalogens, there are nevertheless major limitations associated with these laboratory studies. First, the algae that grow easily in the laboratory are not representative of what is abundant in the open ocean, indeed relatively few phytoplankton have been successfully grown in the laboratory. Second, the growth conditions that are appropriate to laboratory experiments are very unlike natural conditions. To ensure the presence of sufficient biomass, the culture is grown with high concentrations of nutrients. Because the aim is to pinpoint the production of organohalogens, the laboratory cultures are, whenever possible, monocultures grown in the absence of bacteria. In contrast, marine phytoplankton grow in a complex community of algae, bacteria and grazing organisms. While such a system is difficult to study and quantify, field measurements are essential to our understanding of marine organohalogen production. Field incubations [45] of microalgal communities inhabiting the lower surface of sea ice have shown that CHBr3 , CH2Br2 , and CHBr2Cl are being actively produced, the CHBr3 at a rate 30–100 times greater than the other two compounds. The diatom communities were dominated by species of Nitzschia and Navicula. It has been shown by experiments with metabolic inhibitors [36] that ice algae rather than bacteria are responsible for bromoform production. The iodinated C1-C4 hydrocarbons, methyl, ethyl, 1- and 2-propyl iodides have been found [46] associated with Antarctic ice algae. In that study, brominated compounds were not elevated in comparison with the underlying water. Marine microalgal production of a volatile compound can lead to a characteristic vertical concentration profile in the water column. This shows a concentration maximum within the upper 100 m or so. Lower concentrations at shallower depths are primarily due to loss to the atmosphere, while the lower concentrations at greater depths are due to a combination of declining production with depth, downward mixing of the compound, and necessarily a loss process. Such distributions are commonplace for methyl chloride, bromide and
Marine Sources of Volatile Organohalogens
91
iodide, all of which are likely to have lifetimes of days to months in the upper water column. In spite of the growing body of evidence for microalgal production of methyl halides, little is known about the production mechanism or about the reason for their production. Itoh et al. [47], in a study of methyl bromide and iodide production by both macro- and microalgae, identified a relatively prolific source from the marine phytoplankton species, Pavlova gyrans. Identified in this organism was a halide ion methyl transferase which catalysed the methylation of bromide and iodide ions by S-adenosyl-L-methionine (SAM). This enzyme was inactive to chloride ions, and in this respect it differs from a methyltransferase that had been detected [48] in a marine red macroalga (Endocladia muricata) which catalysed methylation of chloride, bromide and iodide by SAM. 2.3 Chemical Sources
The most obvious of the possible chemical sources is, perhaps, the interconversion of halocarbons, for example the conversion of methyl iodide or bromide to methyl chloride by nucleophilic substitution [49].While this does not assist in the search for primary mechanisms for methyl halide production, it is of some consequence for marine methyl chloride. It has been estimated [22] from a Pacific Ocean data set that a CH3I source could account for about 15% of the CH3Cl flux to the atmosphere. The contribution of the CH3Br reaction may be estimated from data given by Butler and Rodriguez [50] on the total loss of CH3Br in the ocean and the proportion of it lost by reaction with chloride ion. This yields 1.5 ¥ 109 moles year–1, or about 20 % of the CH3Cl flux to the atmosphere. Since abundant sources of bromoform and dibromomethane have been identified in macro and microalgae, it may be presumed that there is some progressive conversion of these compounds through chlorobromo compounds, finally yielding chloroform and dichloromethane. Experiments done by Geen [51] indicated a very slow reaction of CHBr3 with seawater, such that its half-life in seawater was estimated as 18.5 years at 25 °C, the temperature of the experiments, and therefore longer at typical ocean water temperatures. An interesting photochemically-mediated conversion has been reported for diiodomethane [52] in seawater to yield chloroiodomethane. This probably occurs with the mediation of an iodinated cation that then reacts with a chloride ion. Field measurements support the occurrence of this reaction [35]. It has been proposed that methyl halides should be formed by a nucleophilic substitution reaction between dimethyl sulfonium propionate (DMSP) and halide ions in seawater [53]. DMSP is an algal osmolyte present in many species of prymnesiophytes, dinoflagellates and chrysophytes; its concentration within cell fluids may reach 0.2 – 0.4 moles L–1 [54]. Dissolved DMSP is found in ocean waters at concentration around 10 nM and, in coastal waters, it has been measured at levels up to 200 nM [55]. An environmentally important decomposition pathway is cleavage into dimethyl sulfide and acrylic acid. It has been shown that reaction with bromide ion yields CH3Br [53] and with iodide yields CH3I [56]. A
92
R.M. Moore
kinetic study [57] of the reactions between DMSP and Cl–, Br– and I–, led to the conclusion that, unless catalysed, the rates of these reactions are too slow to result in significant production of the corresponding methyl halides in seawater. Moore and Zafiriou [58] reported photochemical production of CH3I in laboratory experiments with seawater. It was proposed that the mechanism involved combination of methyl radicals originating from photolysis of humic substances and iodine atoms formed by reaction of iodide with photochemically-produced oxidants. This is not expected to be an efficient process, since the main fate of methyl radicals in water would be reaction with the vastly more abundant oxygen that is normally present, particularly in sunlit surface waters. No field experiments designed to investigate the importance of this process have been reported. A complicating factor in investigating this process in the field is that methyl iodide, according to many studies, does have a marine biological source being produced by various species of seaweeds and phytoplankton. Therefore, experiments on photochemical production must ensure that biological production is either prevented or otherwise accounted for. Happell and Wallace [59] have argued that their field measurements of methyl iodide do support a photochemical source; light intensity was found to explain significant variance in methyl iodide saturation anomalies that they measured in the Greenland and Norwegian Seas. It could be questioned whether light intensity played a role through biology rather than direct photochemistry. There are a number of interconversions between organohalogens in addition to those mentioned above. For example, it has been suggested by Tanhua et al. [60] that when CCl4 is reduced in waters that are anoxic or suboxic, CHCl3 is an intermediate. In their Black Sea study, the authors acknowledge the lesser possibility that the CHCl3 feature referred to could be algal in origin. Dehydrohalogenation of 1,1,2,2-tetrachloroethane and of pentachloroethane has been shown to rapidly produce trichloroethylene and tetrachloroethylene respectively in laboratory experiments [61]. Whether this is significant in the oceans is unknown.
3 Determination of Fluxes As pointed out above, much of the interest in marine volatile organohalogens has come from the field of atmospheric chemistry. From that viewpoint what is required is quantification of the net flux of the compound from ocean to atmosphere, preferably with information on both regional and seasonal variations. The way in which the flux estimates are made is using an equation of the form, Flux = K DC, where DC represents the concentration difference across a stagnant microlayer at the ocean surface, and K is an exchange coefficient [62]. DC is given by the difference between the concentration of the compound that would be at equilibrium with the atmosphere (i.e. the concentration assumed to exist at the top of the microlayer) and the measured concentration in the top few meters of the water column which is presumed to be well mixed (this therefore represents the concentration at the base of the microlayer). The equilibrium concentration of the compound is commonly determined from its concentration in the atmosphere above the water, and its solubility in seawater. Much effort continues to be
Marine Sources of Volatile Organohalogens
93
devoted to satisfactory estimation of the exchange velocity (K), which is a function of a number of factors amongst which wind speed and temperature are important. At present it is likely that uncertainties in K alone could result in an uncertainty in flux by as much as a factor of two. Even larger contributors to the uncertainty in net fluxes of any specific compound are the usually woefully inadequate spatial and temporal sampling of the concentrations. In the case of some compounds that have a transitory existence in the atmosphere (such as methyl iodide), the simplification can sometimes be made that the concentration corresponding to equilibrium with the atmosphere is zero in comparison with the measured concentration (Cw) in the upper water column. The local flux is then the product of the exchange velocity and Cw. Where the focus of marine organohalogen studies is determining the flux to the atmosphere, only surface ocean measurements are needed. If the total ocean production is required, information on the concentrations of the compound throughout the water column is needed as well as knowledge of the rates of all of the loss processes.A summary will be given of our state of knowledge of fluxes of some compounds that are of particular interest from the viewpoint of atmospheric chemistry.
4 Methyl Halides 4.1 Methyl Chloride
Methyl chloride has been of special interest amongst naturally-produced halocarbons, as it is the largest natural contributor to atmospheric chlorine. Its atmospheric abundance of about 550 ppt and its lifetime of 1.4 years with respect to loss by reaction with OH radical alone require an annual supply of around 3.7 ¥ 106 t year–1 [63], although, as pointed out by Harper [64], the existence of various other sinks, particularly soils, would increase the required source.While an early report based on very limited data indicated that the ocean was the major source [65], it became apparent that the ocean is not supersaturated everywhere in methyl chloride [18], and Moore [66] revised the estimate of the ocean source to about 9 – 11 % of the total atmospheric supply. Vertical profiles (Fig. 1) of its concentration in the water column show near surface maxima consistent with in situ production. Decreasing concentrations deeper in the water column imply in situ removal. Although laboratory experiments have shown CH3Cl to have a phytoplankton source, no simple relationship exists between its concentration and the most basic indicator of phytoplankton biomass, the concentration of chlorophyll a [18]. This observation, frequent also for CH3Br and CH3I, is consistent with the fact that production rates are dependent on the species of phytoplankton, as seen from laboratory studies, and very likely also dependent on other factors such as the growth stage and level of environmental stress experienced by the producers. As it is difficult to obtain a detailed taxonomic description of water samples, it will become more common for a wider range of photosynthetic pigments to be measured from
94
R.M. Moore
Fig. 1. Typical vertical profiles of methyl chloride (upper left) and methyl bromide (upper right) concentrations showing well-defined subsurface maxima (Pacific Ocean, 13°41¢N 160°W, October 1995); and a comparison of profiles of anthropogenic CFC-11 (lower left) and dichloromethane (lower right), from the Labrador Sea (59°4¢N 49°57¢W, June 2000)
which information on the broad classes of algae present can be determined. The surprising contrast between undersaturations observed in productive cold waters to the north of the Gulf Stream and supersaturations in the blue oligotrophic waters to the south may be related also to the vigour of in situ removal mechanisms, but work on this topic, as indicated below, is in its early stages. 4.2 Methyl Bromide
Concern about the possible adverse effects on the ozone layer of increasing anthropogenic releases of fumigant methyl bromide led to a proliferation of re-
Marine Sources of Volatile Organohalogens
95
search on all aspects of its environmental chemistry, sources and sinks.Although our knowledge of these is far from complete, it is probable that this has become the single best-understood naturally-produced organohalogen. The belief that the ocean was a large net source of methyl bromide to the atmosphere [65] has been replaced by a recognition that the ocean, though a massive producer, represents a net sink of 11– 20 Gg year–1 with respect to the atmosphere [67]. That is, while methyl bromide is being produced within the oceans, simultaneous loss processes (outlined below) cause its concentration in some areas, particularly where water temperatures are high, to be below that representing equilibrium with the atmosphere; this undersaturation causes an influx from the atmosphere. It has been shown above that there is evidence for both macroalgal and phytoplankton sources of methyl bromide in the ocean. Few consistent patterns have emerged for the emissions of the gas to the atmosphere. Some studies have reported higher concentrations in coastal waters than in the open ocean [10], while others have found oceanic concentrations locally as high as coastal sources [67]. Temperature does exert a major influence on concentrations and saturation levels, for example through the rapid reaction with Cl– [68] at high temperatures that contribute to a rapid chemical turnover (8– 42% d–1 at 20 – 30 °C [50]) of CH3Br and tends to lead to undersaturation. The composition of the algal community is expected to be a major factor controlling CH3Br concentrations. Baker et al. [16] in a North Sea study provided strong evidence that methyl bromide production was related to the growth of the Prymnesiophyte, Phaeocystis. In a separate, non-seasonal study in the NE Atlantic, they found good correlation between CH3Br concentrations in the water and the pigment 19¢-hexanoyloxyfucoxanthin that is associated with Prymnesiophytes. Where there is evidence that organohalogen production in the oceans is a biological process, it may be expected that the rate of production will vary seasonally. Methyl bromide is one of only a few compounds for which evidence for seasonality has been reported; usually data sets are far too sparse to allow such an effect to be confirmed. The yearlong North Sea study by Baker et al. [16] showed that CH3Br was supersaturated in the North Sea for a three-month period, and that its concentrations showed a similar seasonal trend to chlorophyll a. 4.3 Methyl Iodide
Methyl iodide has been of interest as a carrier of this biologically essential element through the atmosphere to the terrestrial environment.Also, it has received attention in connection with the chemistry of iodine in the atmosphere. It should be noted that though methyl iodide is the most widely studied of the marine-produced volatile iodinated compounds, it is not the only one; others include CH2I2 , CH2ClI, C2H5I and C3H7I. Its atmospheric lifetime is so short that its atmospheric concentration is commonly very low, ~1–2.5 ppt. A consequence is that the surface ocean is almost invariably supersaturated unlike the case for CH3Cl and CH3Br, though undersaturation has been reported in the Greenland-Norwegian Sea area in the month of November [59].A recent study [22] of methyl iodide using data from Atlantic and Pacific waters arrived at an estimate for the ocean to
96
R.M. Moore
atmosphere flux of 0.9–2.5 ¥ 109 mol year–1. Surface concentrations and vertical profiles were consistent with an average production rate of 0.5 pmol L–1 d–1. In a study made north of the Azores during summer, a number of water samples were incubated on board ship and CH3I measured with time; four such measurements gave an average production rate of 2.4 pmol L–1 d–1[69].
5 Polychlorinated Compounds Whereas the study of the major polybrominated compounds, CHBr3 and CH2Br2 , in the ocean has been simplified by the fact that they do not have significant anthropogenic sources, the same is not true of the equivalent chlorinated compounds. In some cases there is undisputed evidence that a compound does have some production within the ocean, for example, chloroform, but in many cases much remains to be done in determining the true magnitude of marine production as well as the source mechanisms. 5.1 Chloroform
From the observed concentration of chloroform in the atmosphere and from its chemical loss rate within the atmosphere it becomes apparent that anthropogenic sources account for only about 15 % of the supply [25]. The sole estimate of the ocean to atmosphere flux of chloroform, based on the analysis of only ten samples of ocean water [70] is 0.35 Tg year–1, which would account for about 76% of the atmospheric sink. When combined with an estimated soil emission of 0.2 Tg year–1 and the anthropogenic release of 0.07 Tg year–1, the total emissions are about 37 % higher than needed to balance the atmospheric sink (0.46 g year–1). Keene et al. [25] point out that such a large ocean source appears inconsistent with observations that the gas is twice as abundant in the Northern Hemisphere as in the Southern. It seems likely that an excessive exchange velocity has been used to calculate the ocean to atmosphere fluxes [70], so that ocean emissions could be lower by a factor of two. While a macroalgal source of chloroform has been established [31, 32], its global contribution has been estimated [32] to be insignificant. At present there is only one report of chloroform production by a marine microalga [44], a species that has a restricted distribution in the oceans, being limited primarily to the intertidal and subtidal zones. A further difficulty that arises in ascertaining the contribution of marine-produced chloroform is the fact that it appears to have a long residence time in the ocean giving a high background concentration throughout the water column. This makes it relatively difficult to identify local production and, as discussed in relation to trichloroethylene and tetrachloroethylene below, could also account for a portion of the ocean supersaturation. Thus, while there is strong evidence that CHCl3 is largely natural, the magnitude and nature of its marine source are highly uncertain.
Marine Sources of Volatile Organohalogens
97
5.2 Trichloroethylene and Tetrachloroethylene
There has recently been much interest in the possibility that chlorine atom chemistry is important over the ocean [71]. This has led to some particular interest in sources of atmospheric tetrachloroethylene as it can provide information on the relevance of chlorine atom-initiated oxidation reactions. The reason is that there must be consistency between the observed concentrations of the gas in the atmosphere, its rate of chemical destruction and its rate of supply. The chemical loss rate would be particularly sensitive to Cl atom concentration because the gas reacts at a rate ~300 times faster with Cl than with the OH radical [72]. There are reports that tetrachloroethylene and also trichloroethylene are supersaturated in ocean waters [73, 74], and therefore the ocean represents a source of these compounds to the atmosphere. The Reactive Chlorine Emissions Inventory has put the oceanic source at 0.019 Tg year–1, equivalent to about 4% of estimated industrial emissions. There is a mismatch between total sources and sinks of 0.13 Tg year–1, with sinks exceeding known sources. Natural production of both C2HCl3 and C2Cl4 has been reported [75, 76] in laboratory cultures of both macrophytes and microalgae, with additional evidence from measurements in a tidal pool [77] and, for trichloroethylene, from measurements in the open ocean [78]. In contrast, Marshall et al. [61] did not detect production of either compound in cultures of the same macrophyte, Falkenbergia.An independent laboratory study [44] of the red microalga, Porphyridium purpureum, found no release of either compound. In their study of halocarbon release by macrophytes, two rock pool studies [31, 32] did not find evidence for production of C2HCl3 and C2Cl4. It should be noted that the absence of production of a compound in a particular study cannot prove that it cannot be produced. Marshall et al. [61] pointed out the possibility of production of both C2HCl3 and C2Cl4 through dehydrohalogenation reactions. Moore [74] has found that both C2HCl3 and C2Cl4 have rather unusual oceanic distributions with their concentrations increasing with depth. Such distributions are different from the methyl halides, which have their highest concentration near the surface where they are produced. In part the difference apparently stems from a much longer lifetime of both C2HCl3 and C2Cl4 in ocean waters. But recent measurements of C2HCl3 in the Labrador Sea show similarities between its distribution and that of trichlorofluoromethane (CFC11), which is entirely anthropogenic and is invading the deep ocean from the atmosphere. Their short atmospheric lifetimes cause C2HCl3 and C2Cl4 to have a strong seasonal variation of their concentrations in high northern latitudes, with concentrations reaching their maxima during winter. Subsurface and deep waters have initially acquired their temperature and atmospheric gas composition through equilibration, or partial equilibration, with the winter atmosphere. It appears that atmospherically-derived C2HCl3 and C2Cl4 , as well as man-made Freons may be preserved within these waters. Subsequent mixing with surface waters at lower latitudes or in summer months could result in supersaturations of these compounds at the surface. Thus, the observation that supersaturations exist should not lead to the presumption that a compound has been formed within the ocean. If much more ex-
98
R.M. Moore
tensive data sets were available, including wintertime distributions, there should be no danger of a compound being incorrectly assigned an oceanic source, for the compensating uptake from the atmosphere would be apparent from the winter measurements. 5.3 Dichloromethane
This compound is relatively abundant in ocean waters and there are plausible mechanisms whereby it could be produced in situ. These include reactions of CH2Br2 , CH2I2 and CH2ClI with chloride ion. The possibility exists also for its direct production by mechanisms analogous to those that yield CH2Br2 and CH2I2 . There is one recent report of its enrichment in a rock pool [32]. Dichloromethane shows a remarkable degree of similarity in North Atlantic concentration profiles with those of CFC11 (Fig. 1). This would be consistent with it having an atmospheric source, though it is possible that it is also related to a slow degradation of CH2Cl2 in seawater. Therefore, the ocean inventory of CH2Cl2 may have significantly, if not entirely, been derived from anthropogenic emissions. In the presence of a substantial background of manmade CH2Cl2 it could be very difficult to identify true marine-derived material.
6 Sinks Although the focus of this chapter is marine sources of organohalogens, it is commonly the case that we wish to know the net flux of a compound between ocean and atmosphere, or more specifically, we wish to understand the processes that account for the level of saturation of that compound in any water body. This means that we must be concerned with the existence of sinks as well as sources. A good example is provided by oceanic methyl bromide that, while having major marine production, has a net flux from atmosphere to ocean. In some cases, simple chemical reactions such as hydrolysis and nucleophilic substitution can represent important sinks, as in the case of methyl bromide [68]. Chemical loss rates in warm waters of the Caribbean and Pacific have been reported [79] as 0.3 and 0.25 day–1, respectively. Were these the only sinks, it would be expected that major emissions of algal-produced organohalogens would occur in cold, highly productive seas, but in contrast it is found that methyl bromide and chloride are commonly undersaturated in these waters [80, 81]. For methyl bromide an apparently biological sink has been identified in both coastal and ocean waters with the loss appearing from filtration studies to be the result of microbial processes [13, 82, 83]. Tokarczyk and Saltzman [79] report that in the relatively low temperature waters of the N. Atlantic, biological loss accounted for ~ 35 % of the in situ losses of methyl bromide. Recent work has provided evidence for a seasonal cycle in the removal of methyl chloride from seawater, again with an inferred microbial mechanism. Biological processes have been shown to be responsible for losses of CH2Br2 and CHBr3 in laboratory experiments with Asparagopsis [28]. The CH2Br2 loss
Marine Sources of Volatile Organohalogens
99
appeared to be attributable to organisms that occurred on the macrophyte, but which were released to the medium on prolonged incubation (since CH2Br2 loss was seen to occur whether or not the macrophyte was removed from the medium, but loss would not occur in a spiked seawater control). Bromoform removal occurred effectively only when the alga was present and was attributed to physical absorption or biochemical conversion within the algal biomass. Goodwin et al. [82, 83] have demonstrated that bacteria are able to oxidize CH2Br2 in seawater. There are a number of reports of halocarbon degradation in low oxygen waters; several refer to CCl4 [84, 85, 60], but there is evidence also for reduction of chloroform [60] in the Black Sea. Since low oxygen conditions are not common in the open ocean, these processes are probably not significant on a global scale, but they are relevant in environments such as the Black Sea and Baltic Sea, and also in tracer studies which are based on the principle of conservative behaviour of the tracer [84, 85].
7 References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27.
Lovelock JE (1975) Nature 256:193 Lovelock JE, Maggs RJ, Wade RJ (1973) Nature 241:194 Wofsy SC, McElroy MB, Yung YL (1975) Geophys Res Lett 21 : 215 Solomon S, Garcia RR, Ravishankara AR (1994) J Geophys Res 99:20,491 Dvortsov VL, Geller MA, Solomon S, Schauffler SM,Atlas EL, Blake DR (1999) Geophys Res Lett 26:1699 Kritz MA, Rozner SW, Kelly KK, Loewenstein M, Chan KR (1993) J Geophys Res 98 : 8725 McElroy MB, Salawitch RJ, Wofsy SC, Logan JA (1986) Nature 321:759 Anderson JG, Bruhne WH, Lloyd SA, Toohey DW, Sander SP, Starr WL, Loewenstein M, Podolske JR (1989) J Geophys Res 94:11480 Butler JH (1994) Geophys Res Lett 21:185 Lobert J, Butler JH, Montzka SA, Geller LS, Myers RC, Elkins JW (1995) Science, 267 : 1002 Yvon S, Butler JH (1996) Geophys Res Lett 23:53 Scarratt MG, Moore RM (1996) Mar Chem 54:263 King DB, Saltzman ES (1997) J Geophys Res 102:18,715 Scarratt MG, Moore RM (1998) Mar Chem 59:311 Groszko W, Moore RM (1998) J Geophys Res 103:16,737 Baker JM, Reeves CE, Nightingale PD, Penkett SA, Gibb SW, Hatton AD (1999) Mar Chem 64:267 Tait VK, Moore RM (1995) Limnol Oceanogr 40:189 Moore RM, Groszko W, Niven S (1996) J Geophys Res 101:28,529 Manley SL, Dastoor MN (1988) Mar Biol 98:477 Manley SL, de la Cuesta J (1997) Limnol Oceanogr 42:142 Reifenhäuser W, Heumann KG (1992) Atmos Environ 26A : 2905 Moore RM, Groszko W (1999) J Geophys Res 104:11,163 Barrie LA, Bottenheim JW, Schnell RC, Crutzen PJ, Rasmussen RA (1988) Nature 334:138 McConnell JC, Henderson GS, Barrie L, Bottenheim J, Niki H, Langford CH, Templeton EMJ (1992) Nature 355:150 Keene WC et al. (1999) J Geophys Res 104 : 8429 McConnell OJ, Fenical W (1979) In: Hoppe AH, Tanaka Y (eds) Marine algae in pharmaceutical science. Walter de Gruyter, Berlin, p 403 Gschwend PH, MacFarlane JK, Newman KA (1985) Science 227:1033
100 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45.
46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61 62. 63. 64. 65. 66.
R.M. Moore
Marshall RA, Harper DB, McRoberts WC, Dring MJ (1999) Limnol Oceanogr 44:1348 Dyrssen D, Fogelqvist E (1981) Oceanologica Acta 4:313 Manley SL, Goodwin K, North WJ (1992) Limnol Oceanogr 37:1652 Nightingale, PD, Malin G, Liss PS (1995) Limnol Oceanogr 40:680 Baker JM, Sturges WT, Sugier J, Sunnenberg G, Lovett AA, Reeves CE, Nightingale PD, Penkett SA (2001) Chemosphere Global Change Sci 3 : 93 Krysell M (1991) Mar Chem 33:187 Klick S, Abrahamsson K (1992) J Geophys Res 97:12,683 Moore RM, Tokarczyk R (1993) Global Biogeochem Cycles 7 : 195 Sturges WT, Cota GF, Buckley PT (1992) Nature 358:660 Tokarczyk R, Moore RM (1994) Geophys Res Lett 21 : 285 Moore RM, Webb M, Tokarczyk R, Wever R (1996) J Geophys Res 101:20,899 Tait VK, Moore RM (1995) Limnol Oceanogr 40:189 Sundström J, Collén J, Abrahamsson K, Pedersén M (1996) Phytochemistry 42:1527 Pedersén M, Collén J, Abrahamsson K, Ekdahl A (1996) Sci Mat 60:257 Urhahn T, Ballschmiter K (1998) Chemosphere 37:1017 Saemundsdottir S, Matrai PA (1998) Limnol Oceanogr 43:81 Scarratt MG, Moore RM (1999) Limnol Oceanogr 44:703 Moore RM, Tokarczyk R, Tait V, Poulin M, Geen C (1995) Marine phytoplankton as a natural source of volatile organohalogens. In: Grimvall A, de Leer EWB (eds) Naturally produced organohalogens. Kluwer Academic Publishers, Dordrecht, The Netherlands. Fogelqvist E, Tanua T (1995) Iodinated C1-C4 hydrocarbons released from ice algae in Antarctica. In: Grimvall A, de Leer EWB (eds) Naturally produced organohalogens. Kluwer Academic Publishers, Dordrecht, The Netherlands Itoh N, Tsujita M, Ando T, Hisatomi G, Higashi T (1997) Phytochemistry 45:67 Wuosmaa AM, Hager LP (1990) Science 249:160 Zafiriou OC (1975) J Mar Res 33:75 Butler JH, Rodriguez JM (1996) Methyl bromide in the atmosphere. In: Bell CH, Price N, Chakrabarti B (eds) The methyl bromide issue. Wiley & Sons, p 27 Geen CE (1992) Selected marine sources and sinks of bromoform and other low molecular weight organobromines. MSc thesis, Dalhousie University, Halifax, Canada Class Th, Ballschmiter K (1987) Fresenius Z Anal Chem 327:40 White RH (1982) J Mar Res 40:529 Keller MD, Bellows WK, Guillard RRL (1989) Dimethyl sulfide production in marine phytoplankton. In: Saltzman ES, Cooper WJ (eds) Biogenic sulfur in the environment. ACS Symposium Series No. 393, Washington DC, p 101 Turner SM, Malin G, Liss PS (1989) Dimethyl sulphide and dimethylsulphoniopropionate in European coastal and shelf waters. In: Saltzmann ES, Cooper WJ (eds) Biogenic sulfur in the environment. ACS Symposium Series No. 393, Washington DC, p 183 Brinckman FE, Olson GJ, Thayer JS (1984) Biological mediation of marine metal cycles: the case of methyl iodide. In: Sigleo AC, Hattori A (eds) Marine and estuarine chemistry. Lewis Publishers, Chelsea Michigan, p 227 Hu Z, Moore RM (1996) Mar Chem 52 : 147 Moore RM, Zafiriou (1994) J Geophys Res 99:16:415 Happell JD, Wallace DWR (1996) Geophys Res Lett 23:2105 Tanhua T, Fogelqvist E, Bastürk Ö (1996) Mar Chem 54 : 159 Marshall RA, Hamilton JTG, Dring MJ, Harper DB (2000) Limnol Oceanogr 45:516 Liss PS, Merlivat L (1986) Air-sea gas exchange rates: Introduction and synthesis. In: BuatMenard P (ed) The role of air-sea exchange in geochemical cycling. D. Reidel, Norwell, Mass, p 113 Khalil MAK, Rasmussen RA (1999) Atmos Environ 33 : 1305 Harper DB (2000) Nat Prod Rep 17:337 Singh HB, Salas LJ, Stiles RE (1983) J Geophys Res 88:3675 Moore RM (2000) Chemosphere: Global Change Sci 2 : 95
Marine Sources of Volatile Organohalogens
101
67. King DB, Butler JH, Montzka SA, Yvon-Lewis AS, Elkins JW (2000) J Geophys Res 105:19,763 68. Elliott S, Rowland FS (1993) Geophys Res Lett 20:1043 69. Hughes C (2001) Oceanic methyl iodide: Production rates, relationship with photosynthetic pigments and a biological loss process. MSc thesis, Dalhousie University 70. Khalil MAK, Rasmussen RA, Hoyt SD (1983) Tellus 35B : 266 71. Finlayson-Pitts BJ (1993) Res Chem Intermed 19 : 235 72. Nicovich JM, Wang S, McKee ML, Wine PH (1996) J Phys Chem 100:680 73. Khalil MAK et al. (1999) J Geophys Res 104 : 8333 74. Moore RM (2001) J Geophys Res 106 : 27,135 75. Abrahamsson K, Ekdah A, Collén J, Fahlstrom E, Pedersén M (1995) The natural formation of trichloroethylene and perchloroethylene in sea water. In: Grimvall A, de Leer EWB (eds) Naturally produced organohalogens. Kluwer Academic Publishers, Dordrecht, The Netherlands 76. Abrahamsson K, Ekdah A, Collén J, Pedersén M (1995) Limnol Oceanogr 40:1321 77. Ekdahl A, Pedersén M, Abrahamsson K (1998) Mar Chem 63:1 78. Abrahamsson K, Pedersén M (2000) Limnol Oceanogr 45:520 79. Tokarczyk R, Saltzman ES (2001) J Geophys Res 106 : 9843 80. Moore RM, Webb M (1996) Geophys Res Lett 23:2951 81. Lobert JM,Yvon-Lewis SA, Butler JH Montzka SA Myers RC (1997) Geophys Res Lett 24:171 82. Goodwin KD, Lidstrom ME, Oremland RS (1997) Environ Sci Technol 31:3188 83. Goodwin KD, Shaefer JK, Oremland RS (1998) Applied Environ Microbiol 64:4629 84. Krysell M, Fogelqvist E, Tanhua T (1994) Geophys Res Lett 21 : 2511 85. Wallace DWR, Beining P, Putzka A (1994) J Geophys Res 99:7803
The Handbook of Environmental Chemistry Vol. 3, Part P (2003): 103–119 DOI 10.1007/b 10451
An Update on Organohalogen Metabolites Produced by Basidiomycetes Jim A. Field 1, Joannes B.P.A. Wijnberg 2 1 2
Department of Chemical and Environmental Engineering, University of Arizona, P.O. Box 210011, Tucson, Arizona 85721-011, USA. E-mail:
[email protected] Laboratory of Organic Chemistry, Wageningen University, Dreijenplein 8, 6703 HB, Wageningen, The Netherlands
Basidiomycetes are an ecologically important group of higher fungi known for their widespread capacity to produce organohalogen metabolites. To date, 100 different organohalogen metabolites (mostly chlorinated) have been identified from strains in 70 genera of Basidiomycetes. This manuscript provides an update of newly discovered chlorinated metabolites described since 1997.Additionally, the biosynthesis, physiological role, environmental fate and significance of Basidiomycete organohalogen metabolites are reviewed. Novel metabolites include chlorinated methoxybenzene azoxyformamide, pterulones (chlorinated 1-benzoxepins), chlorinated anisylpropanoids, tri- and tetrachlorinated phenols. Chlorinated p-anisyl metabolites (CAM) are the most ubiquitous and ecologically significant natural organohalogens produced by higher fungi. Evidence is presented indicating their synthesis from phenylalanine via the phenylpropanoid pathway. They are estimated to be produced at a rate of 300 g ha–1 year–1 in European forests inhabited by the common occurring mushroom, Hypholoma fasciculare. Organohalogen metabolites have several purposeful physiological functions ranging from antibiotic properties, metabolites involved in lignin degradation and synthons for biosynthesis. Keywords. Basidiomycete, Fungi, Biohalogenation, Organohalogen, Chlorinated metabolites, Halometabolites, Secondary metabolites
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
1
Introduction
2
Most Common Organohalogen Metabolites
2.1 2.2 2.3 2.4
Chlorinated Anisyl Metabolites . . Drosophilin A Metabolites . . . . . Chlorinated Strobilurins . . . . . . Chloromethane . . . . . . . . . . .
3
New Organohalogen Metabolites
3.1 3.2 3.3
Novel Metabolites Related To Previously Described Structures . . 107 Novel Metabolite Types . . . . . . . . . . . . . . . . . . . . . . . . 108 Chloroform . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
4
Biosynthesis
4.1 4.2
Halogenation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 Precursors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 110
. . . .
. . . .
. . . .
. . . .
. . . . . . . . . . . . 105 . . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
105 105 105 106
. . . . . . . . . . . . . . . . . . 106
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109
© Springer-Verlag Berlin Heidelberg 2003
104
J. A. Field and J.B.P.A. Wijnberg
5
Biotransformation Products . . . . . . . . . . . . . . . . . . . . . 111
5.1 5.2 5.3
Oxidative Enzymes . . . . . . . . . . . . . . . . . . . . . . . . . . 111 Biotransformations by Aerobic Bacteria . . . . . . . . . . . . . . . 112 Biotransformations by Anaerobic Bacteria . . . . . . . . . . . . . 113
6
Physiological Roles . . . . . . . . . . . . . . . . . . . . . . . . . . 114
6.1 6.2 6.3
Antibiotics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 114 Roles of Chlorinated Metabolites in Lignin Biodegradation . . . . 114 Synthons for Biosynthesis . . . . . . . . . . . . . . . . . . . . . . 116
7
Environmental Significance . . . . . . . . . . . . . . . . . . . . . 116
8
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
1 Introduction Basidiomycetes are higher fungi that produce macroscopic fruiting bodies such as mushrooms and brackets. They play important ecological roles in the recycling of nutrients, decomposing plant debris [1] and the symbiosis with living plant roots, mycorrhiza [2]. The predominant organisms responsible for the degradation of wood, white rot and brown rot fungi belong for the most part to Basidiomycetes [3].White rot fungi are unique in their ability to attack the natural aromatic polymer lignin in wood with extracellular oxidative enzymes [4–5]. Basidiomycetes constitute a major portion of the living biomass produced in the litter layer of a forest, accounting for the production of approximately one ton of mycelial biomass (dry weight) per hectare per year in a typical temperate hardwood forest [6]. Basidiomycetes are also important producers of organohalogen compounds. Due to their widespread occurrence and large scale of biomass production, the biosynthesis of organohalogens by this group of organisms constitutes a very significant source of organohalogens. The first report of an organohalogen metabolite produced by a mushroom species occurred in 1952 with the detection of drosophilin A (tetrachloro-4-methoxyphenol) in Psathyrella (syn. Drosophila) subatrata [7–8]. Since then, numerous observations have been reported indicating a widespread production of diverse organohalogen compounds by Basidiomycetes. Two previous review articles have inventoried the halogenated metabolites from Basidiomycetes up to the beginning of 1997 [9–10]. At that time, eighty-one halogenated metabolites were reported belonging to halogenated aliphatics or halogenated aromatics. Most of the metabolites were chlorinated, with only a few reports of iodinated and brominated metabolites. The review articles also indicated an immense biodiversity of organohalogen production among Basidiomyctes with the capacity having been observed among sixty-eight different genera belonging to 20 different Basidiomycete families. The biodiversity was further confirmed by a screening of 191 Basidiomycete strains
An Update on Organohalogen Metabolites Produced by Basidiomycetes
105
for the bulk organohalogen parameter, adsorbable organic halogens (AOX), which revealed that 50 % of the strains tested produced AOX [11].
2 Most Common Organohalogen Metabolites 2.1 Chlorinated Anisyl Metabolites
Based on the results of the previous review articles [9–10], several families of common occurring metabolites can be recognized. Chlorinated anisyl metabolites (CAM) are clearly produced in the highest amounts by the largest diversity of Basidiomycetous genera. The most common occurring constituents are 3chloro- and 3,5-dichloro-p-anisyl alcohols and aldehydes (compounds 1–4 depicted in Fig. 1). At least one of these four compounds is produced by strains within seventeen genera of Basidiomycetes. One genus has been added to the list since the last review, due to the recent report of 2 and 4 in Psathyrella scobinacea [12]. Chlorinated p-anisic acids (compounds 5 and 6 in Fig. 1) are also produced by four of these genera, of which the production of 6 by Stropharia squamosa was overlooked in the last review [13] and the reference of 6 in Hypholoma fasciculare has been published [14]. The specific yield of the CAM is generally high among most of the producers. In the common occurring sulfur tuft, Hypholoma fasciculare, CAM production is equivalent to 3–4.4% of the mycelium dry weight [14].A concentration of 108 mg L–1 CAM compounds in the culture broth was reported for Hypholoma elongatum [15]. 2.2 Drosophilin A Metabolites
The next most important organohalogen metabolites are the drosophilin A metabolites (compounds 7 and 8 depicted in Fig. 1). They are produced by fungi from eleven genera, including reports of drososphilin A methyl ether (8) production by Marasmius androsaceus and Polyporus porrectus which were overlooked [10, 16–17]. Also the reference to the production of compound 8 by Hypholoma fasciculare has been published [14]. Generally, Basidiomycetes produce low concentrations of drosophilin A metabolites, with the exception of high concentrations in Phellinus fastuosus [18–19]. Compounds with related structures having fewer chlorine atoms have also been noted such as 2-chloro-1,4dimethoxybenzene, 2,6-dichloro-1,4-dimethoxybenzene, and 2,6-dichloro-4methoxyphenol [10]. The latter compound occurs at high concentrations (250 mg kg–1 fresh weight) in the mushrooms of Russula subnigricans [20]. 2.3 Chlorinated Strobilurins
Strobilurin B (9) and the closely related oudemansin B (10) are produced by fungi from three genera of Basidiomycetes, including commonly occurring Mycena.
106
J. A. Field and J.B.P.A. Wijnberg
Fig. 1. Common occurring chlorinated metabolites produced by Basidiomycetes. Chlorinated anisyl metabolites (CAM): 3-chloro-p-anisyl alcohol (1), 3,5-dichloro-p-anisyl alcohol (2), 3chloro-p-anisaldehyde (3), 3,5-dichloro-p-anisaldehyde (4), 3-chloro-p-anisic acid (5), and 3,5dichloro-p-anisic acid (6). Drosophilin A metabolites: drosophilin A (7) and drosophilin A methyl ether (8). Chlorinated strobilurins: strobilurin B (9) and oudemansin B (10)
These compounds along with their non-chlorinated counterparts are powerful respiratory inhibitors binding to mitochondrial cytochrome bc1 complex. They have very specific toxicity against fungi and are non-toxic towards mammals and bacteria; thus, they have become lead compounds for a new generation of agricultural fungicides [21–22]. 2.4 Chloromethane
Basidiomycetes from eight different genera produce chloromethane (11) [10, 23]. Most of these are polypores within the order Aphyllophorales. Five genera are white rot fungi within the family Hymenochaetaceae. From a screening of sixty-three species within Hymenochaetacea, 55% were shown to emit chloromethane and the biosynthetic capacity was greatest within the genera Phellinus and Inonotus [23]. The commercially important cultivated mushroom, A. bisporus, is the only fungus in the order Agaricales (mushrooms with gills) known to produce chloromethane.
3 New Organohalogen Metabolites Since the last review article, nineteen new de novo chlorinated metabolites from Basidiomycetes have been reported, bringing the total number of known organohalogen metabolites from this group of fungi up to one hundred. Several
An Update on Organohalogen Metabolites Produced by Basidiomycetes
107
of the novel metabolites are related in structure to previously described metabolites and others represent new types of metabolites. In addition, better evidence for the production of a suspected organohalogen metabolite (chloroform) has been obtained.Among the Basidiomycetes producing these novel organohalogen metabolites, only two new genera have been identified, Pterula [24] and Lycoperdon [25], bringing the total number of known organohalogen producing genera up to seventy. 3.1 Novel Metabolites Related To Previously Described Structures
Some of the novel organohalogen metabolites represent a small variation to a previously described group of organohalogen metabolites. Such variations have been found for chlorinated amino acids, chlorinated anthraquinones, chlorinated orsenillate sesquiterpenes, chlorinated anisylpropanoids and chlorophenol methyl ethers. One new chlorinated orsenillate sesquiterpene, 6¢-chloro-10a-hydroxymelleolide (structure 12 in Fig. 2) was isolated from Armillaria novae-ze-
Fig. 2. Novel chlorinated metabolites from Basidiomycetes, similar in structure to previously-
known metabolites. 6¢-chloro-10a-hydroxymelleolide (12), 7-chloroemodin (13), 5,7dichloroemodin (14), 5,7-dichloroendocrocin (15), 2-amino-5-chloro-5-hexenoic acid (16), trametol (17), bjerkanderol B (18), 1-(3¢-chloro-4¢-methoxyphenyl)-3-hydroxy-1-propanone (19), tetrachloropyrocatechol (20), tetrachloropyrocatechol methyl ether (21), 2,4,6trichloro-3-methoxyphenol (22), 2,3,5-trichloro-4,6-dimethoxyphenol (23), and 2,4,5trichloro-3,6-dimethoxyphenol (24)
108
J. A. Field and J.B.P.A. Wijnberg
landiae mycelium [26]. Previously, eight other chlorinated orsenillate sesquiterpenes were reported from many species of Armillaria and Clitocybe elegans [10]. Three new chlorinated anthraquinone metabolites, 7-chloroemodin, 5,7dichloroemodin and 5,7-dichloroendocrocin (structures 13–15 in Fig. 2), were isolated from Dermocybe sanguinae fruiting bodies [27]. Two other chlorinated anthraquinones were described before from many species within Dermocybe and the very closely related Cortinarius [10]. One new chlorinated amino acid, 2amino-5-chloro-5-hexenoic acid (structure 16 in Fig. 2) was detected at trace quantities in the fruiting bodies of Amanita miculifera [28]. Previously, five other chlorinated amino acids have also been isolated from various Amanita fruiting bodies [10]. A family of chlorinated p-anisylpropanoids has been identified from three genera of white rot fungi. Trametol (structure 17 in Fig. 2) was first reported from Trametes sp. [29]. Since then, bjerkanderol B (18) has been identified in various Bjerkandera [30–32] as well as Hypholoma strains [31].Additionally, 1-(3¢-chloro4¢-methoxyphenyl)-3-hydroxy-1-propanone (19) was produced by Bjerkandera fumosa [30]. Previously, the chlorophenol methyl ethers, 3,5,6-trichloro-1,2,4-trimethoxybenzene, and 4,5,6-trichloro-1,2,3-trimethoxybenzene were identified in Phellinus fastuosus [33]. In the last few years, several new types of tri- and tetrachlorinated phenols have been identified as natural products in Basidiomycetes. Mycena sp. strain 87202 produced tetrachloropyrocatechol and its methyl ether (structures 20 and 21 in Fig. 2) [34]. Hypholoma elongatum was shown to produce 2,4,6-trichloro-3-methoxyphenol, 2,3,5-trichloro-4,6-dimethoxyphenol, and 2,4,5-trichloro-3,6-dimethoxyphenol [35] (structures 22–24 in Fig. 2). 3.2 Novel Metabolite Types
A novel chlorinated benzene azoxyformamide metabolite, 3,5-dichloro-4methoxybenzene-1-ONN-azoxyformamide (structure 25 in Fig. 3) was isolated from the wood degrading puff ball, Lycoperdon pyriforme [25]. Several bioactive chlorinated compounds with 1-benzoxepin skeletons referred to as pterulones have been isolated from two species of Basidiomycetes. Pterulone and (E,Z)pterulinic acid (compounds 26 and 31, respectively, in Fig. 3) were isolated from mycelium and culture broth of the wood and litter-degrading coral fungus, Pterula sp. strain 82168 [24, 36]. Pterulone B (30) was isolated only when the fungus was grown on beech wood [37]. Similar compounds were present in the latex of the mushroom stipes of the forest litter degrading agaric Mycena galopus. These include, 6-hydroxypterulone (27), [(3E,Z)-3-(chloromethylene)-2,3-dihydro-1-benzoxepin-7-yl]-methanol (28), and (3E,Z)-3-(chloromethylene)-2,3-dihydro-1-benzoxepin-7-carbaldehyde (29) [38 – 39]. Pterulones 26 and 31 were tested and found to be effective inhibitors of eukaryotic respiration, targeting NADH:ubiquinone oxidoreductase. In accordance, pterulones have potent antifungal activities [24, 39]. Lastly, veratryl chloride (3,4-dimethoxybenzyl chloride) was identified in cultures of Bjerkandera [40], but it has also been shown to be an artifact of a low pH
An Update on Organohalogen Metabolites Produced by Basidiomycetes
109
Fig. 3. Novel groups of organohalogen metabolites from Basidiomycetes: 3,5-dichloro-4methoxybenzene-1-ONN-azoxyformamide (25), pterulone (26), 6-hydroxypterulone (27), [(3E,Z)-(3-chloromethylene)-2,3-dihydro-1-benzoxepin-7-yl]methanol (28), (3E,Z)-(3chloromethylene)-2,3-dihydro-1-benzoxepin-7-carbaldehyde (29), pterulone B (30), and (E,Z)-pterulic acid (31)
reaction between veratryl alcohol and chloride [Silk, personal communication]; thus, its designation as a natural product is not confirmed. 3.3 Chloroform
Chloroform (32) was previously detected while analyzing for volatile components released from the pressed juice of the mushroom, Cantharellus cibarius [41]. However, the study did not clarify if the chloroform resulted from natural production. To prove natural production of chloroform among Basiomycetes, careful experiments were conducted with fungal liquid cultures incubated in a confined atmosphere in parallel with an uninoculated medium controls. The Basidiomycetes Peniophora pseudopini and Mycena metata reliably produced chloroform in all replicates and production levels were in large excess (5– 20 times) of the background levels in uninoculated medium blanks [42]. Incidental production occurred in Agaricus arvensis, Bjerkandera sp. strain BOS55 and Phellinus pini.
4 Biosynthesis 4.1 Halogenation
Little is known about the actual mechanisms of biohalogenation during the formation of most organohalogens produced by Basidiomycetes. Several attempts have been made to demonstrate chloroperoxidase activity in Basidiomycetes while they were actively producing the CAM. The common chloroperoxidase substrate, phenol red, was utilized with whole cells, extracellular fluids and intra-
110
J. A. Field and J.B.P.A. Wijnberg
cellular cell free extracts; however, no activity was found in numerous strains tested [Terrón et al., unpublished results]. Consequently, alternative methods of halogenation need to be considered aside from haloperoxidases. Recently, evidence has been reported in bacteria for novel O2 and FADH2-dependent halogenating enzymes with specific activities [43]. In the case of halomethanes, more is known. Membrane-bound S-adenosylmethionone (SAM) methyl transferases have been partially purified from Phellinus pomaceus and shown to have activity transferring methyl group of SAM to halide ions (Cl–, Br–, and I–) [23].A novel indirect mechanism of halomethane formation is through chemical oxidation of methoxyphenols by Fe3+, which promotes a nucleophilic attack of the methoxy group by a halide ion [44]. Given the ubiquitous capacity of wood rotting fungi to oxidize iron and produce methoxylated aromatic metabolites, the indirect chemical mechanism of halomethane formation may be important in some cases. 4.2 Precursors
The most ecologically significant metabolites, CAM, are biosynthesized via the phenylpropanoid pathway with phenylalanine as the major precursor. Production of CAM was stimulated by addition of phenylalanine as well as intermediates of the phenylpropanoid pathway (cinnamate, benzoate). Various types of labeling experiments utilizing either 2H, 13C, or fluoro-labeled phenylalanine and/or other phenylpropanoid intermediates have been utilized to conclusively demonstrate their involvement as precursors to CAM in Bjerkandera [31, 45–46]. Phenyl and phenolic intermediates were found to increase CAM production more rapidly than anisyl intermediates, suggesting that chlorination occurs prior to methylation [45]. The production of CAM is greater on natural lignocellulosic substrates compared to defined laboratory medium because phenolic products released from lignin degradation supplement the supply of precursors [47]. The proposed route of CAM biosynthesis by Bjerkandera is shown in Fig. 4. The figure illustrates the conversion of phenylalanine to cinnamic acid by phenylalanine ammonia lyase (PAL) and the conversion of cinnamic acid to benzoic acid through Claisen cleavage [48]. Benzoate is hydroxylated forming 4-hydroxybenzoic acid, which is postulated to be the direct substrate of the chlorinating system, forming 3,5-dichloro-4-hydroxybenzoic acid. Methylation with either SAM- or CH3Cl-dependent methyltransferases [23], results in the formation of 3,5-dichloro-p-anisic acid. Reduction by aryl alcohol dehydrogenase (AAD) activity is responsible for the formation of 3,5-dichloro-p-anisaldehyde and 3,5dichloro-p-anisyl alcohol [49]. The formation of chlorinated phenylpropanoid metabolites such as trametol and bjerkanderol B results from a stereoselective adduct formation between a C7 unit (3-chloro- or 3,5-dichloro-p-anisaldehyde) and a C2 unit derived from pyruvate [31]. This type of adduct formation with aldehydes, known as C2-homologation, was first described for Baker’s yeast (Saccharomyces cerevisiae) [50]. The enzyme responsible for C2-homologation is a pyruvate decarboxylase (PDC) [51]. The formation of bjerkanderol B from 3,5-dichloro-p-anisaldehyde is shown in Fig. 4. Exceptional enantioselective reductase activity has been observed in the
An Update on Organohalogen Metabolites Produced by Basidiomycetes
111
Fig. 4. Proposed route of 3,5-dichloro-p-anisyl alcohol biosynthesis by Bjerkandera spp. See
text for definitions of abbreviations and description of pathway
fungus, Merulius tremellosus [52], which could account for the conversion of C2homologation products to bjerkanderol B. Similar reductase activity was also observed in Bjerkandera sp. BOS55 [Hage, unpublished results].
5 Biotransformation Products Organohalogen metabolites produced by basidiomycetes are subject to biotransformations once they are released into the environment. These metabolites have been shown to be biotransformed by oxidative enzymes of ligninolytic fungi, as well as by aerobic and anaerobic bacteria. 5.1 Oxidative Enzymes
Lignin peroxidase (LiP) from the white rot fungi Phanerochaete chrysosporium and Bjerkandera sp. BOS55 as well as a novel manganese peroxidase (with Mn-
112
J. A. Field and J.B.P.A. Wijnberg
independent activity from Bjerkandera), were shown to oxidize 2-chloro-1,4dimethoxybenzene [53 – 55]. 2-Chloro-1,4-dimethoxybenzene is a natural product described in Bjerkandera sp. and Lepsita nuda [10]. The first product of the oxidation by the peroxidases was 2-chloro-1,4-benzoquinone and methanol.Additionally, several dimers, probably resulting from coupling of the radical cation of 2-chloro-1,4-dimethoxybenzene, were detected with mass spectrometry, but their exact structures could not be established [54] (Fig. 5A). Manganese-independent peroxidase (MIP) from Phellinus fastuosus was shown to oxidize drosophilin A to two dimers each with an elemental analysis of C14H6O4Cl8 and an accurate mass of 517.7774 Da [56]. The proposed structures of these dimers are shown in Fig. 5B. The proposed structures assume coupling of a phenoxy radical with positions 2 or 4. 5.2 Biotransformations by Aerobic Bacteria
Research was conducted on the biodegradation of 3,5-dichloro-p-anisyl alcohol by indigenous microflora in forest soils. Under aerobic conditions, 200 mg L–1 of the metabolite was readily mineralized by microflora from three forest soils (beech, oak, and pine forests) and by microflora in activated sludge [Verhagen et al., unpublished results]. The organically-bound chlorine was completely recovered as inorganic chloride after three weeks of incubation. Enrichment cultures were made utilizing the oak-forest soil inoculum. The resulting bacterial isolates
Fig. 5A, B. Dimeric biotransformation products formed by oxidative enzymes of ligninolytic fungi. A Formation of 2-chloro-1,4-benzoquinone, dichloro-tetramethoxybiphenyl, and 2chloro-(chloro-dimethoxybenzene)-benzoquinone from 2-chloro-1,4-dimethoxybenzene by lignin peroxidase (LiP) [54]. B Formation of octachlorodimers from the oxidation of drosophilin A by manganese-independent peroxidase (MIP) [56]
An Update on Organohalogen Metabolites Produced by Basidiomycetes
113
were all found to be Burkholderia cepacia [Verhagen et al., unpublished results]. B. cepacia strains are also known from the literature as outstanding degraders of anthropogenic organohalogens. The natural production of high concentrations of CAM in forest soils by fungal colonies will create a local selection pressure for organohalogen-degrading bacteria. This hypothesis might account for the recovery of 3-chlorobenzoate and 2,4-dichlorophenoxyacetate-degrading bacteria in pristine environments [57]. 5.3 Biotransformations by Anaerobic Bacteria
Research was also conducted to evaluate the biotransformation of chlorinated aromatic metabolites under anaerobic conditions, since the highest organohalogen-producing fungus, Hypholoma elongatum, grows in wetlands in sphagnum moss. The main metabolite of this fungus, 3,5-dichloro-p-anisyl alcohol, was incubated with anaerobic sludge [58]. The metabolite was rapidly demethylated, yielding initially 3,5-dichloro-4-hydroxybenzyl alcohol, which was further oxidized to 3,5-dichloro-4-hydroxybenzoic acid. Some decarboxylation of the benzoic acid intermediate occurred yielding 2,6-dichlorophenol. Also, during these incubations a tetrachlorinated adduct intermediate was detected, which was identified as bis(3,5-dichloro-4-hydroxyphenyl)methane. This adduct was found to be produced by an abiotic coupling of 3,5-dichloro-4-hydroxybenzyl alcohol, catalyzed by components in autoclaved sludge [58].An overall pathway of the biotransformation of the major fungal metabolite, 3,5-dichloro-p-anisyl alcohol under anaerobic conditions is shown in Fig. 6. The main trend is that anaerobic biotransformation of fungal chlorinated aromatic metabolites leads to the formation of chlorophenols.
Fig. 6. Pathway of the biotransformation of the major fungal metabolite, 3,5-dichloro-p-anisyl
alcohol under anaerobic conditions [58]. See text for description of pathway
114
J. A. Field and J.B.P.A. Wijnberg
6 Physiological Roles Basidiomycetes produce significant amounts of chlorinated metabolites. The question that remains is why Basidiomycetes produce these organohalogens. Their biosynthesis does not seem to be a biological accident since some Basidiomycetes (Hypholoma) produce organically-bound halogens in quantities equivalent to up to 3% of their biomass dry weight [11]. To produce such large quantities of organically-bound halogens, an enormous amount of metabolic energy must be invested. The scope of the production is a first indication that their synthesis is purposeful. Indeed, several physiological functions have been proposed for the organohalogen metabolites. These physiological functions vary from antibiotic properties, metabolites involved in lignin degradation, and synthons for biosynthesis. 6.1 Antibiotics
The first physiological function attributed to organohalogens is their antibiotic properties. The antibiotic properties of drosophilin A were described along with the first characterization of the metabolite [8]. The antibiotic properties of chlorinated metabolites from Basidiomycetes have been previously reviewed [9]. Further evidence of antifungal activity of CAM compounds at 50–100 mg L–1 has been reported [13]. New chlorinated compounds with antibiotic activity include the pterulones [24, 37, 39] and tetrachlorinated catechols [34]. Clearly the antibiotic properties of these metabolites serve to protect the fungal colony from competing microorganisms. There are several examples of increased chlorinated metabolite production in response to antagonizing (invading) fungal cultures. Chlorinated orsenillate sesquiterpene production by Armillaria ostoyae was greatly enhanced in the presence of an antagonizing Basidiomycete culture (Gloeophyllum abietinum). The inducing agents were shown to be low molecular weight metabolites less than 3000 Da [59]. Likewise, drosophilin A methyl ether production by Phellinus fastuosus could be induced by metabolites in the extracellular fluid of Phlebia radiata [19]. 6.2 Roles of Chlorinated Metabolites in Lignin Biodegradation
Apart from their antibiotic properties, chlorinated metabolites are also involved in fungal metabolism. Most Basidiomycetes that produce CAM are white rot fungi, and CAM alcohols are important metabolites in the ligninolytic system of white rot fungi [60]. Both 3-chloro- and 3,5-dichloro-p-anisyl alcohol are substrates for extracellular hydrogen peroxide production. Hydrogen peroxide production is important for ligninolytic degrading enzymes, since many are peroxidases. The extracellular enzymes responsible for the oxidation of aryl alcohols to aryl aldehydes at the expense of O2 reduction to H2O2 are aryl alcohol oxidases (AAO) [60]. Almost all fungi that synthesize CAM produce AAO. The CAM alcohols are much better substrates for AAO compared to their structurally similar
An Update on Organohalogen Metabolites Produced by Basidiomycetes
115
non-chlorinated secondary metabolite counterparts, veratryl alcohol (3,4dimethoxybenzyl alcohol) and p-anisyl alcohol. Aryl aldehydes produced from the extracellular oxidation are recycled to the corresponding alcohols by intracellular NADPH-dependent aryl alcohol dehydrogenases (AAD), which generates a physiological cycle for continuous H2O2 production [49, 60–61], as shown in Fig. 7 A. Both veratryl alcohol and p-anisyl alcohol are substrates of lignin peroxidase; whereas, the chloro group of CAM alcohols renders the aryl alcohol resistant to lignin peroxidase, conserving the molecule for the H2O2-producing redox cycle [60] The fungal metabolite 2-chloro-1,4-dimethoxybenzene (2CDMB) was found to act as a redox mediator for lignin peroxidase in the same fashion as veratryl alcohol. Veratryl alcohol is a common metabolite of white rot fungi associated with lignin peroxidase production. The presence of sub-stoichiometric concentrations of 2CDMB enabled lignin peroxidase to oxidize various substrates which otherwise are not oxidized very well or indeed at all, replacing the function of veratryl alcohol [53 – 54, 62]. The substrates oxidized in the presence of either ver-
Fig. 7A, B. Involvement of de novo chlorinated aromatic metabolites of white rot fungi in lignin degradation. A Cycles of extracellular oxidation and intracellular reduction of chlorinated panisyl metabolites to support continuous H2O2 production [60]. B Role of 2-chloro-1,4dimethoxybenzene as redox mediator for lignin peroxidase [53–54, 62]. See text for definitions of abbreviations and description of pathways
116
J. A. Field and J.B.P.A. Wijnberg
atryl alcohol or 2CDMB were azo dyes, polymeric dyes, oxalate, p-anisyl alcohol, and 4-methoxymandelic acid. The evidence indicates that 2CDMB is oxidized to a radical cation (2CDMB• +), which oxidizes the terminal substrate as illustrated in Fig. 7B by the example of 4-methoxymandelic acid, which is not a direct a substrate of lignin peroxidase. The occurrence of a radical cation was demonstrated with an electro paramagnetic resonance spectrometer [54]. Additionally, the presence of terminal substrates decreased the conversion of 2CDMB by lignin peroxidase, which would be expected if the intermediate oxidation product (radical cation) was reduced back to 2CDMB upon reaction with the terminal substrate [62]. 6.3 Synthons for Biosynthesis
Organic chemists utilize chloro groups for the direct synthesis of organic compounds. Therefore, long ago it was speculated that nature utilizes halogenated intermediates to assist in the biosynthesis of non-halogenated metabolites [63]. Chloromethane serves this function in white rot fungi and is the only known example of a halogenated synthon in biological systems. Chloromethane was shown to be readily utilized as a methyl donor for the biosynthesis of veratryl alcohol. Deuterium-labeled chloromethane was incorporated as methoxy groups in veratryl alcohol an important secondary metabolite of white rot fungi [23, 64]. Surprisingly, chloromethane is not only used as a methyl donor by chloromethaneemitting white rot fungi, such as Phellinus pomaceus, but also by Phanerochaete chrysosporium and Trametes versicolor, which are non-chloromethane-emitting fungi. The results suggest that the latter fungi have a tightly coupled methylating system in which chloromethane does not accumulate; however, this hypothesis still needs to be proven.
7 Environmental Significance Basidiomycetes are an ecologically significant group of organisms, which account for a large proportion of the living biomass in plant litter [6]. Therefore, production of organohalogens by this class of organisms may be of global significance. There is good evidence that many of the organohalogen metabolites of Basidiomycetes are produced while growing on natural substrates. Chloromethane production by Phellinus spp. is favored when grown on cellulose [23]. CAM was also shown to be produced by strains of Bjerkandera and Hypholoma when grown on natural wood or forest litter [14, 65–66]. Petrulones were produced by Pterula sp. strain 82168 when grown on beech leaves or wood shavings [37]. Organohalogen metabolites based on adsorabable organic halide (AOX) or total organic halide (TOX) measurements have been detected on natural wood and litter substrates inoculated with a variety of Basidiomycete strains [11, 66]. Utilizing ergosterol content to estimate mycelium biomass, specific production of AOX by Hypholoma fasciculare is 0.88, 0.92, and 1.28% of mycelium dry weight on pine wood, beech wood, and forest litter, respectively. The production of CAM has
An Update on Organohalogen Metabolites Produced by Basidiomycetes
117
been confirmed in the natural environment in field samples of substrata adjacent to mushroom colonies of Bjerkandera, Hypholoma, Pholiota, Stropharia, and Lepsita [65, 67]. Hypholoma had the highest field concentrations of CAM, averaging 75 mg kg–1 dry weight. The annual global emissions of chloromethane by white rot fungi from Hymenochaetaceae was estimated to be 160,000 t year–1 [23], which is the same order of magnitude as estimates based on emissions from temperate conifer forest soils. The fungal contribution is significant compared to the total global emission of 4,000,000 t year–1 of which industrial emissions only accounts for 10,000 t year–1. CAM are the most ubiquitous occurring organohalogens in terrestrial environments. The annual production of CAM per hectare of forest has been estimated based on specific AOX production per unit weight of mycelium [11, 14] together with mycelium production in selected forests in The Netherlands. The mycelium weight was based on field data of mushroom yields and the relation between mushroom and total mycelium weights. Alternatively, estimates of the total Basidiomycete mycelium production per hectare was multiplied by the relative frequency of the mushroom species. In Dutch forests, Hypholoma fasciculare are the mushrooms with the highest frequency (1.3 % of all sightings), and they also have the highest specific yield of AOX. The production of AOX would thus be largely attributable to this species. Estimates made in the two different ways for Hypholoma fasciculare indicated a production of AOX of 90–110 g ha–1 year–1. Since almost all of the AOX is attributable to CAM compounds, 3,5-dichloro-panisyl alcohol and the corresponding aldehyde, this amounts to a CAM production of approximately 300 g ha–1 year–1.
8 References 1. Boddy L, Watkinson SC (1995) Can J Bot 73 : S1377 2. Buscot F, Munch JC, Charcosset JY, Gardes M, Nehls U, Hampp R (2000) FEMS Microbiol Rev 24:601 3. Eriksson KE, Blanchette RA,Ander P (1990) Microbial and enzymatic degradation of wood and wood components. Springer-Verlag, Berlin 4. Kirk TK, Farrell RL (1987) Annu Rev Microbiol 41:465 5. De Jong, E, Field JA, de Bont JAM (1994) FEMS Microbiol Rev 13:153 6. Swift MJ (1982). Basidiomycetes as components of forest ecosystems. In: Frankland JC, Hedger JN, Swift MJ (eds) Decomposer Basidiomycetes: Their biology and ecology. Cambridge Univ. Press, Cambridge, p 307 7. Anchel M (1952) J Am Chem Soc 74:2943 8. Kavanagh F, Hervey A, Robbins WJ (1952) Proc Nat Acad Sci USA 38:555 9. Field JA, Verhagen F, de Jong E (1995) Natural organohalogen production by basidiomycetes. Trends Biotechnol 13:451 10. De Jong E, Field JA (1997) Annu Rev Microbiol 51:375 11. Verhagen FJM, Swarts HJ, Kuyper TW, Wijnberg JPBA, Field JA (1996) Appl Microbiol Biotechnol 45:710 12. Taha, AA (2000) Phytochemistry 55:921 13. Thines E, Daussmann T, Semar M, Sterner O, Anke H (1995) Z Naturforsch C 50:813 14. Verhagen, FJM, van Assema FBJ, Boekema BKHL, Swarts HJ,Wijnberg JBPA, Field JA (1998) FEMS Microbiol Lett 158:167
118
J. A. Field and J.B.P.A. Wijnberg
15. Swarts HJ, Teunissen PJM, Verhagen FJM, Field JA, Wijnberg JBPA (1997) Mycol Res 101:372 16. Zhu S, Li S, Liu L, Li G (1988) Zhongcaoyao 19:338 17. Pandit SS; Kulkarni AB (1976) Indian J Chem B 14:230 18. Singh P, Rangaswami S (1966) Tetrahedron Lett 11:1229 19. Teunissen PJM, Swarts HJ, Field JA (1997) Appl Microbiol Biotechnol 47:695 20. Takahashi A, Agatsuma T, Matsuda M, Ohta T, Nunozawa T (1992) Chem Pharm Bull. 40:3185 21. Anke T (1997) Strobilurins. In: Anke T (ed), Fungal biotechnology. Chapman Hall, London, p 206 22. Anke T (1995) Can J Bot 73 : S940 23. Harper DB (2000) Nat Prod Rep 17:337 24. Engler M, Anke T, Sterner O, Brandt U (1997) J Antibiot 50:325 25. Koepcke B, Mayer A, Anke H, Sterner O (1999) Nat Prod Lett 13 : 41 26. Cremin P, Guiry PJ,Wolfender JL, Hostettmann K, Donnelly DMX (2000) J Chem Soc Perkin Trans 1 2000:2325 27. Räisänen R, Bjork H, Hynninen PH (2000) Z Naturforsch C 55:195 28. Hatanaka S, Niimura Y, Takishima K, Sugiyama J (1998) Phytochemistry 49:573 29. Brambilla U, Nasini G, Depava OV (1995) J Nat Prod 58:1251 30. Swarts HJ, Verhagen FJM, Field JA, Wijnberg JBPA (1998) J Nat Prod 61 : 1110 31. Silk PJ, Aubry C, Lonergan GC, Macaulay JB (2000) Chemosphere 44 : 1603 32. Levy LM, Cabrera GM, Wright JE, Seldes AM (2000) Molecules 5:354 33. Jain SC, Kumar R, Bharadvaja A, Parmar VS, Errington W (1996) Novel products of some wood rotting fungi. Abstr 20th IUPAC Symp Chem Nat Prod. Chicago, Sept 15th–20th, p SE-35 34. Daferner M, Anke T, Hellwig V, Steglich W, Sterner O (1998) J Antibiot 51 : 816 35. Swarts HJ, Verhagen FJM, Field JA, Wijnberg JBPA (1998) Phytochemistry 49 : 203 36. Engler M, Anke T, Sterner O (1997) J Antibiot 50:330 37. Engler M, Anke T, Sterner O (1998) Z Naturforsch C 53 : 318 38. Wijnberg JBPA, van Veldhuizen A, Swarts HJ, Frankland JC, Field JA (1999) Tetrahedron Lett 40:5767 39. Gruijters BWT, Wijnberg JBPA, Weijers C (2001) J Nat Prod (in press) 40. Swarts HJ, Mester T,Verhagen FJM, Field JA,Wijnberg JBPA (1997) Phytochemistry 46:1011 41. Pyysalo H (1976) Acta Chem Scand B 30:235 42. Hoekstra EJ,Verhagen FJM, Field JA, de Leer EWB, Brinkman UAT (1998) Phytochemistry 49:91 43. Van Pee KH (2001) Arch Microbiol 175:250 44. Keppler F, Eiden R, Niedan V, Pracht J, Schöler HF (2000) Nature 403:298 45. Mester T, Swarts HJ, Romero S, de Bont JAM, Field JA (1997) Appl Environ Microbiol 63:1987 46. Lauritsen FR, Lunding A (1998) Enzyme Microb Technol 22:459 47. Mester, T, Sierra-Alvarez R, Field JA (1998) Holzforschung 52:351 48. Jensen Jr KA, Evans KMC, Kirk TK, Hammel KE (1994) Appl Environ Microbiol 60:709 49. Hage A, Schoemaker HE, Field JA (1999) Appl Microbiol Biotechnol 52:834 50. Ohta H, Ozaki K, Konishi J, Tsuchihashi G (1986) Agric Biol Chem 50:1261 51. Crout DHG, Hutchinson DW, Miyagoshi M (1991) J Chem Soc Perkin Trans 1991:1329 52. Hage A, Petra DGI, Field JA, Schipper D, Wijnberg JBPA, Kamer PCJ, Reek JNH, van Leeuwen PWNM, Wever R, Schoemaker HE (2001) Tetrahedron Asymmetry 12:1025 53. Teunissen, PJM, Field JA (1998) Appl Environ Microbiol 64:830 54. Teunissen, PJM, Sheng D, Bhasker Reddy V, Moënne-Loccoz P, Field JA Gold MH (1998) Arch Biochem Biophys 360 : 233 55. Mester T, Field JA (1998) J Biol Chem 273:15412 56. Teunissen PJM (1999) The role of natural hydroquinone metabolites in ligninolytic fungi. PhD dissertation, Wageningen University 57. Fulthorpe RR, Rhodes AN, Tiedje JM (1996) Appl Environ Microbiol 62:1159
An Update on Organohalogen Metabolites Produced by Basidiomycetes
58. 59. 60. 61. 62. 63. 64. 65.
119
Verhagen FJM, Swarts HJ,Wijnberg JPBA, Field JA (1998) Appl Environ Microbiol 64:3225 Sonnenbichler J, Dietrich J, Peipp H (1994) Biol Chem Hoppe Seyler 375 : 71 De Jong E, Cazemier AE, Field JA, de Bont JAM (1994) Appl Environ Microbiol 60:271 Guillen F, Evans CS (1994) Appl Environ Microbiol 60:2811 Teunissen PJM, Field JA (1998) FEBS Lett 439:219 Neidleman SL (1975) Crit Rev Microbiol 5:333 Harper DB, Buswell JA, Kennedy JT, Hamilton JTG (1990) Appl Environ Microbiol 56:3450 De Jong E, Field JA, Spinnler HE, Wijnberg JBPA, de Bont JAM (1994) Appl Environ Microbiol 60 : 264 66. Oberg G, Brunberg H, Hjelm O (1997) Soil Biol Biochem 29:191 67. Hjelm O, Boren H, Oberg G (1996) Chemosphere 32:1719
The Handbook of Environmental Chemistry Vol. 3, Part P (2003): 121–139 DOI 10.1007/b 10450
Volcanic Formation of Halogenated Organic Compounds Armin Jordan Max-Planck-Institute for Biogeochemistry, P.O. Box 100164, 07701 Jena, Germany E-mail:
[email protected] Formerly at Max-Planck-Institute for Aeronomy, 37191 Katlenburg-Lindau, Germany
Volcanoes are very important halogen emitters. Currently observed levels of halogens in the ocean are the result of continuous volcanic degassing during Earth’s history.Although the halogens are emitted almost exclusively as inorganic halides, volcanic production of organohalogens has also been observed in several studies. The range of compounds detected includes chlorofluorocarbons, methyl halides, short-chain halohydrocarbons as well as halogenated aromatics and heterocycles. The presence of halogenated hydrocarbons was a matter of controversy as it is in disagreement with thermodynamic considerations. Organic compounds detected in volcanic gases point to a non-equilibrium reaction mechanism including the thermal cracking of non-magmatic methane and subsequent halogenation reactions. Emission rate estimates for halocarbons result in very low global fluxes that are unlikely to have any significance for the global atmospheric budgets of relevant halogenated trace gases. Keywords. Volcanoes, Organohalogens, Reaction mechanism, Chlorofluorocarbons, Methyl
chloride
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
1
Introduction
2
Organohalogens in Volcanic Gases
2.1 2.2
Qualitative Studies . . . . . . . . . . . . . . . . . . . . . . . . . . 123 Quantitative Data . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
3
Formation Mechanism . . . . . . . . . . . . . . . . . . . . . . . . 129
3.1 3.2 3.3 3.3.1 3.3.2 3.3.3
Thermodynamic Considerations . . . . . . . . . . . . . . . . . . Occurrence of Hydrocarbons Above Magmatic Equilibrium Levels Relative Abundance of Fluorine, Chlorine, Bromine and Iodine . . Halogen Concentrations in Magmas and Volcanic Gases . . . . . . Degassing Properties . . . . . . . . . . . . . . . . . . . . . . . . . Chemical Reactivity . . . . . . . . . . . . . . . . . . . . . . . . .
4
Environmental Impact of Volcanic Organohalogen Production and Global Emissions . . . . . . . . . . . . . . . . . . . . . . . . 135
5
Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
6
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
. . . . . . . . . . . . . . . . . 123
129 129 133 133 134 134
© Springer-Verlag Berlin Heidelberg 2003
122
A. Jordan
1 Introduction The degassing process of the Earth’s interior through volcanoes has generated the Earth’s atmosphere and hydrosphere during the past 4.6 billion years. Namely water vapour and CO2 from magmatic gases have thus created the basis for the development of the whole organic life on Earth. Other compounds emitted in large quantities with volcanic gases include halogens. The levels of chlorine and bromine found in the oceans, which are the largest exospheric reservoir of these elements, have accumulated as a result of continuous extraction from the Earth’s mantle via volcanic degassing [1 – 3]. The biggest part of atmospheric hydrogen chloride and hydrogen bromide now originates from sea salt volatilisation, but still some 0.5 – 10 % of the total atmospheric HCl source strength are emitted directly from volcanoes. Inorganic fluorine emissions from volcanoes are among the biggest source for fluorine emitted into the atmosphere [4]. Table 1 compares volcanic emission rates of hydrogen halides with other major global sources. This vast halogen source coupled with temperatures of several hundreds to 1000 °C make volcanoes interesting chemical reactors for halogen chemistry. However, the bulk species of halogen emissions are inorganic compounds such as hydrogen halides or alkaline halides. Compared to inorganic halogen emissions, published data on organohalogens emitted by volcanoes are very scarce. This chapter summarizes the current knowledge on volcanogenic halocarbons. In the beginning an overview of qualitative and quantitative published data is given. The potential formation mechanism is discussed with respect to contradictory results from theoretical considerations and measured data. Finally, the environmental significance of volcanic organohalogen emissions is assessed. Table 1. Halogen fluxes to the atmosphere a
Volcanic emission Global emission of all sources Biggest source for atmospheric input
a b
HF
HCl
HBr
HI
0.06–6 [5, 7, 8] 0.4–5 [4]
0.4–11 [5–9] <100–1200 [13, 14] seasalt volatilisation
0.001–0.05 [5, 10 b, 11] 1 [8]
0.00003–0.001 [5, 12] b 0.4–0.5 [14]
seasalt volatilisation
photochemical conversion of marine iodinated compounds
volcanoes, dust
Tg year –1 = 1012 g year –1. Volcanic emission rates estimated for HBr and HI are based on the volcanic HCl flux [5] multiplied by a molar ratio of HCl/HBr = 273 [10] and Cl/I = 1200 [12], respectively.
Volcanic Formation of Halogenated Organic Compounds
123
2 Organohalogens in Volcanic Gases 2.1 Qualitative Studies
The pioneering work that gave evidence of volcanogenic halogenated organic compounds was done in 1971 by Stoiber et al. [15]. They investigated whole air samples of fumarolic gases taken at Santiaguito Volcano (Guatemala). In these gas mixtures there were two chlorohydrocarbons, two fluorocarbons and five chlorofluorocarbons among 42 identified organic compounds (see Tables 2 and 3). This finding came a few years before chlorofluorocarbons were identified as destroyers of stratospheric ozone [16], and it took some time until additional studies were undertaken by atmospheric chemists. Rasmussen et al. [17, 18] could not detect any chlorofluorocarbons in gases of Kilauea and Mauna Loa volcanoes (Hawaii). Yet, they reported elevated levels of CH3Cl in ambient air samples collected in the vicinity of the Kalapana-Kilauea eruption 1977 and CH3Cl, CH3Br and CH3I at Mauna Loa [19]. Methyl chloride concentrations that were ten times above the ambient background level were also detected in the tropospheric [20] and the stratospheric plume [21] of the eruptions of Mt. St. Helens during March to May 1980. Again, the concentrations of CFCl3 and CF2Cl2 levels were not elevated compared to ambient background levels. Thermodesorption experiments undertaken with ash emitted during the big eruption of Mt. St. Helens on 18th May 1980 released considerable amounts of CH3Cl, CH3Br and CH3I and “other halocarbons of higher molecular weight” [18] as well as polychlorinated biphenyls and chlorobenzoic acids [22]. The presence of halogenated hydrocarbons on pyroclastic material has also been reported for volcanic ash samples from Kamchatka, the Kurile Islands and Indonesia [23]. Evidence for chlorinated methanes, chlorofluorocarbons and tetrachloroethene in volcanic gas emanations from Kamchatka and the Kurile Islands has been collected by Isidorov et al. [24] (and publications in Russian, cited therein). High concentrations of CFCl3 , CF2Cl2 , chloroform and carbon tetrachloride were quantitatively determined (see next paragraph). Wahrenberger [25] also investigated volcanic gases from the Kurile volcano island Kudriavy and reported the presence of fluoroethene in a 300°C fumarole and the detection of C2H3ClF2 and C3H6F2 in a 900 °C fumarole. At Vulcano Island (Italy) he detected chloroethene in gases from one fumarole and CH2Cl2 , C2H3Cl3 , C2H2Cl4 and two C2Cl4F2 isomers. Fumarolic gas samples collected at Vulcano were also investigated by Jordan et al. [26] together with volcanic gases taken at Mt. Etna and at the Japanese volcanoes Kuju and Satsuma-Iwojima. A multitude of more than 100 halogenated compounds was identified in these volcanic gases including saturated and unsaturated aliphatic halohydrocarbons, as well as aromatic and heteroaromatic compounds, most of them at very low concentrations (5 – 50 ppt). The most abundant halogenated hydrocarbon was CH3Cl; other major organohalogens were CH3Br and CH3I, the higher chlorinated methanes CH2Cl2 , CHCl3 and CCl4 , as well as C2HCl, C2H3Cl and chlorobenzene. Only a few fluorohydrocarbons were identified at very low levels (< 10 ppt) in gas samples from
124
A. Jordan
Table 2. Chlorinated organic compounds detected in volcanic gases (number of isomers in
parentheses) Methyl chloride Dichloromethane Trichloromethane Tetrachloromethane Chlorodifluoromethane Dichlorofluoromethane Trichlorofluoromethane Dichlorodifluoromethane Chlorobromomethane Dichlorobromomethane Trichlorobromomethane Chlorodibromomethane Chloroiodomethane Chloroethane Dichloroethane (2) Trichloroethane Tetrachloroethane Bromochloroethane Chlorodifluoroethane Trichlorotrifluoroethane Tetrachlorodifluoroethane (2) Chloroethene Dichloroethene (3) Trichloroethene Tetrachloroethene Chlorotrifluoroethene Bromochloroethene Bromodichloroethene (2) Bromotrichloroethene Chloroethyne Dichloroethyne Chloropropane Chloropropene (4) Dichloropropene (3) Trichloropropene (2) Tetrachloropropene (2) Bromochloropropene (3) Bromodichloropropene Chloropropyne Dichloropropyne (2) Trichloropropyne Tetrachloropropyne Tetrachloropropadiene Bromochloropropyne Bromodichloropropyne Chlorobutane (4) Dichlorobutane Trichlorobutane Tetrachlorobutane Chlorobutene Dichlorobutene (2)
CH3Cl CH2Cl2 CHCl3 CCl4 CHF2Cl CHFCl2 CFCl3 CF2Cl2 CH2ClBr CHCl2Br CCl3Br CHClBr 2 CH2ClI C2H5Cl C2H4Cl2 C2H3Cl3 C2H2Cl4 C2H4ClBr C2H3F2Cl C2F3Cl3 C2F2Cl4 C2H3Cl C2H2Cl2 C2HCl3 C2Cl4 C2F3Cl C2H2ClBr C2HCl2Br C2Cl3Br C2HCl C2Cl2 C3H7Cl C3H5Cl C3H4Cl2 C3H3Cl3 C3H2Cl4 C3H4ClBr C3H3Cl2Br C3H3Cl C3H2Cl2 C3HCl3 C3Cl4 C3Cl4 C3H2ClBr C3HCl2Br C4H9Cl C4H8Cl2 C4H7Cl3 C4H6Cl4 C4H7Cl C4H6Cl2
[15, 17–19, 26, 27] [24–26] [25, 26] [25, 26] [15] [15] [15, 24, 26] [24] [26] [26] [26] [26] [26] [26] [26] [25] [25] [26] [25] [15] [25] [25, 26] [26] [15, 25, 26] [24, 26] [15] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26]
125
Volcanic Formation of Halogenated Organic Compounds Table 2 (continued)
Trichlorobutene Chlorobromobutene Chlorobutyne Dichlorobutadiene (2) Dichlorobutyne (2) Chlorobutenyne (2) Dichlorobutenyne Trichlorobutenyne Tetrachlorobutadiene Chlorocyclopentene (3) Dichlorocyclopentene (2) Trichlorocyclopentene Chlorohexane Chlorocyclohexane (2) Chlorobenzene Fluorochlorobenzene Dichlorobenzene (3) Trichlorobenzene (2) Chlorotoluene (3) Chloroethylbenzene Chlorostyrene (3) Chlorofuran (2) Dichlorofuran (2) Chlorothiophene (2) Dichlorothiophene (3) Trichlorothiophene Chloroselenophene (3)
C4H5Cl3 C4H6ClBr C4H5Cl C4H4Cl2 C4H4Cl2 C4H3Cl C4H2Cl2 C4HCl3 C4H2Cl4 C5H7Cl C5H6Cl2 C5H5Cl3 C6H13Cl C6H11Cl C6H5Cl C6H4FCl C6H4Cl2 C6H3Cl3 C7H7Cl C8H9Cl C8H7Cl C4H3OCl C4H2OCl2 C4H3SCl C4H2SCl2 C4HSCl3 C4H3SeCl
[26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26]
Vulcano. With the exception of CFCl3, chlorofluorocarbon concentrations were generally present at concentrations below ambient levels, but did show the same relative abundances of single chlorofluorocarbons (CFCl3/CF2Cl2/C2F2Cl4/ C2F3Cl3) as in background air, which points to their atmospheric origin. However, CFCl3 concentrations were slightly enhanced above atmospheric levels in most samples. Several brominated and a few iodinated compounds were also observed in this study as well as molecules containing chlorine together with any other halogen. In addition to fumarole samples, volcanic gas could be collected directly from residual degassing lava at the point were it poured out of a fissure in the South East crater of Mt. Etna (Italy). The biggest variety of compounds was found in these lava gases. Some of the compounds appeared only in some of the samples taken one after the other at the same location and there was a big variability in the concentration levels. The concentrations of chlorofluorocarbons in all samples of the South East crater (fumaroles and lava gases) were below ambient, but did show the same relative abundance of single chlorofluorocarbons (CFCl3/CF2Cl2/C2F2Cl4/C2F3Cl3) as in background air, which points to their atmospheric origin. In contrast to this, fumarolic samples from the Japanese volcanoes, as well as Vulcano and the North East crater of Mt. Etna did show CFCl3 concentrations that were slightly above background.
126
A. Jordan
Table 3. Fluorinated, brominated and iodinated organic compounds detected in volcanic gases
(number of isomers in parentheses) Fluorinated compounds Chlorodifluoromethane Dichlorofluoromethane Trichlorofluoromethane Dichlorodifluoromethane Tetrafluoroethene Chlorodifluoroethane Trichlorotrifluoroethane Tetrachlorodifluoroethane (2) Trifluoropropene Hexafluoropropene Fluorobenzene Tetrafluorobenzene Fluorochlorobenzene
CHF2Cl CHFCl2 CFCl3 CF2Cl2 C3H3F3 C2H3F2Cl C2F3Cl3 C2F2Cl4 C3H3F3 C3F6 C6H5F C6H2F4 C6H4FCl
[15] [15] [15, 24, 26] [24] [15] [25] [15] [25] [26] [15] [26] [26] [26]
Iodinated compounds Methyl iodide Chloroiodomethane Iodoethane Iodoethene
CH3I CH2ClI C2H5I C2H3I
[18, 19, 26] [26] [26] [26]
Brominated compounds Methyl bromide Dibromomethane Tribromomethane Chlorobromomethane Dichlorobromomethane Trichlorobromomethane Chlorodibromomethane Bromoethyne Bromoethene Bromoethane Bromochloroethane Bromochloroethene Bromodichloroethene (2) Bromotrichloroethene Bromopropene (3) Bromochloropropene (3) Bromodichloropropene Bromochloropropyne Bromodichloropropyne Bromobutane Bromobutadiene Chlorobromobutene Bromobutyne Bromobenzene Bromofuran (2)
CH3Br CH2Br2 CHBr3 CH2ClBr CHCl2Br CCl3Br CHClBr2 C2HBr C2H3Br C2H5Br C2H4ClBr C2H2ClBr C2HCl2Br C2Cl3Br C3H5Br C3H4ClBr C3H3Cl2Br C3H2ClBr C3HCl2Br C4H9Br C4H5Br C4H6ClBr C4H5Br C6H5Br C4H3OBr
[18, 19, 26, 27] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26] [26]
Volcanic Formation of Halogenated Organic Compounds
127
Very recently, fumarolic gases collected at subduction zone volcanoes of Nicaragua were investigated and again showed CH3Cl and CH3Br as dominant halogenated hydrocarbons [27]. A complete list of halogenated compounds reported so far is shown in Tables 2 and 3. The range of organohalogens reported in the different studies is consistent, but also show remarkable differences. On the one hand, these differences may have geologic causes such as different chemistries of magmas from different tectonic settings or of volcanic gases during different stages of volcanic activity. However, it should be noted that the variety of organohalogens reported in each study is also determined by the focus of interest of the respective study. Moreover, it is limited by the respective analytical methods. Different sampling procedures and analytical approaches in the individual studies make the comparison of different results problematic. For example, the use of adsorption tubes results in the breakthrough of very volatile compounds and may result in a chemical modification of adsorbed compounds on the traps (e. g. halogen exchange reactions of iodinated or brominated species). Likewise, chromatographic parameters and detector settings limit the range of compounds that are detected. In contrast to the other methyl halides, methyl fluoride has never been reported, probably because of its low molecular weight which is identical to H2S. Low volatile compounds such as polyhalogenated dibenzofurans and dibenzodioxins were also excluded from detection with the respective methods of each of the investigations on organohalogens in volcanic gases. The great number of compounds identified by Jordan et al. [26] was made possible by using a very sensitive GC-ion trap mass spectrometer in combination with a large mass spectra library. Many of these compounds were probably also present in samples of other studies, but below the detection limits of the respective analytical method. 2.2 Quantitative Data
Only a few of the compounds listed in Tables 2 and 3 have been quantified. These concentration data are depicted in Tables 4 and 5. Rasmussen et al. [17] reported levels of 20 – 200 ppb of CH3Cl at Kalapana-Kilauea 1977 and mean concentrations of 128 ppb CH3Cl, 218 ppb CH3Br and 109 ppb CH3I at Mauna Loa ranging up to the low ppm level [19]. Isidorov et al. [23, 24] have published concentrations of CFCl3 , CF2Cl2, CHCl3 and CCl4 in fumarolic gases and in air sampled inside the crater of three volcanoes from the Kurile Islands. Chlorofluorocarbon concentrations were enriched in all samples relative to background levels by a factor of up to 500. The concentrations ranges were 0.5 – 80 ppb for CFCl3 and 1–160 ppb for CF2Cl2 . These figures demonstrate the high variability in halocarbon concentrations detected at different vents and points of time within the same volcano. Data for CHCl3 and CCl4 concentrations were limited to fewer samples and generally ranged from 1–10 ppb, except for one sample with an exceptionally high chloroform value of 820 ppb in volcanic gas from Mutnovskii volcano. Besides this very high value, these CHCl3 and CCl4 concentrations are of the same
128
A. Jordan
Table 4. Concentrations of organohalogens in volcanic gases
Mauna Loa (Hawaii) [19] CFCl3 CF2Cl2 CH3Cl CH2Cl2 CHCl3 CCl4 CH3Br CH3I
128
Kilauea (Hawaii) [17]
Mendeleev (Kurile Islands) [23]
Golovnin (Kurile Islands) [23]
Mutnovskii (Kurile Islands) [23]
0.5–78.9 0.84/140
3.4–80 1.1–160
75.5 18.5
Kudriavy (Kurile Islands) [25]
20–200 70 41.4 3.8
820.8 19.1
218 109
order of magnitude as those reported for Kuju, Satsuma Iwojima, Vulcano and Mt. Etna [26]. The maximum CH3Cl concentrations found in the latter study also correspond to those detected in the gases of Mauna Loa, but CH3Br and CH3I levels differed by one order of magnitude. Generally, chlorohydrocarbons were the most abundant compounds, which is in line with hydrogen chloride being the most abundant hydrogen halide in the gases. However, in all gas samples from Kuju, Satsuma Iwojima,Vulcano and the North East crater of Mt. Etna, CH3I levels were above those of CH3Br. This is remarkable as it is in contrast to the relative abundance of the respective halogens in magmas. Concentrations of single halogenated hydrocarbons varied considerably on a very short timescale. In a series of samples taken one after Table 5. Concentrations of organohalogens in volcanic gases (data from [26])
CFCl3 CF2Cl2 CH3Cl CH2Cl2 CHCl3 CCl4 CH3Br CH2Br2 CHBr3 CH3I C2HCl C2H3Cl C2HCl3 C2Cl4 C6H5Cl a
Kuju (Japan)
Satsuma Iwojima (Japan)
Vulcano (Italy)
Mt. Etna (Italy)
Etna lava gas
0.63 0.43 a 41 0.04 a 0.01 a 0.10 a 0.37 <0.005 <0.005 1.4 0.04 0.03 0.07 0.04 27
0.36 0.49 a 120 0.05 a 0.04 0.11 a 4.3 <0.005 <0.005 8.9 <0.01 0.03 0.04 0.02 0.92
0.97–0.98 0.4–0.57 a 24–25 0.12–0.21 0.06–0.13 0.09 a –0.40 0.05–1.5 <0.005 <0.005 0.16–1.8 0.19–1.1 0.06–0.25 0.23–0.40 0.10–0.38 0.73–2.0
0.16 a –0.45 0.21–0.58a 0.61 a –84 0.01 a –1.8 0.03–5.5 0.1 a –9.2 0.01 a –3.7 <0.005–0.008 a <0.005–0.04 0.02–1.9 0.01–49 0.04–15 0.02–1.8 0.01 a –0.16 0.06–30
0.03–0.09 a 0.11–0.28a 0.95–74 0.04 a –12 0.05–3.7 0.04 a –6.1 0.16–3.7 <0.005–0.1 <0.005–0.007 a 0.11–1.4 <0.01–5.8 0.01–0.68 0.02–2.2 0.02–1.5 0.01–5.3
Indicates concentrations at or below atmospheric background levels.
Volcanic Formation of Halogenated Organic Compounds
129
the other at the same spot in the South East crater of Mt. Etna, the concentrations spanned a range of up to two orders of magnitude. The concentrations of chlorofluorocarbons in all lava gas samples were below ambient, but did show the same relative abundances of the single chlorofluorocarbon compounds (CFCl3/CF2Cl2/ C2F2Cl4/C2F3Cl3) as in background air, which points to their atmospheric origin. There is agreement in the various studies that methanes are the predominant molecules. But in some lava gas samples, chloroethyne, chloroethene and chlorobenzene were reaching comparable concentration levels as methyl chloride.
3 Formation Mechanism 3.1 Thermodynamic Considerations
There are explanations for the scarcity of data on organic halogenated carbons. The most obvious cause is the difficulty and danger to access samples. Yet, there are two more important reasons that have kept the interest of volcanologists from this issue. On the one hand, all samples taken by Stoiber et al. [15] and Isidorov et al. [24] were diluted with ambient air.As geologists are mainly interested in gas samples that contain only magmatic volatiles (less than 1% air), this air admixture makes them “contaminated” and thus less useful for their purposes [28, 29]. The second and even more important point is the fact that the occurrence of organic carbon including organohalogens under magmatic conditions contradicts thermodynamic theory. Computer programs are used to calculate the partial pressures of selected compounds for different temperature and oxidation regimes typically present in volcanoes based on thermodynamic equilibrium data [30]. It has been shown in several studies that the formation of halogenated organic compounds at detectable levels can be excluded under magmatic conditions [5, 25, 31]. To illustrate the inconsistency of measured data and theoretically derived values, the concentration data for selected halogenated organic compounds are compared with the calculated fugacities they are expected to have at thermodynamic equilibrium in Table 6. There is a discrepancy of many orders of magnitude for all compounds. This contradiction led to doubts as to whether experimental data was rather produced by analytical artefacts and contamination during sampling [32] or were caused by biomass burning [17, 31]. 3.2 Occurrence of Hydrocarbons Above Magmatic Equilibrium Levels
In the few studies dealing with volcanogenic organohalogens, the two viewpoints remained incompatible for a long time. However, a similar inconsistency has been stated for non-halogenated hydrocarbons in volcanic gases. Numerous studies have detected methane and other hydrocarbons in volcanic gas samples at levels that exceed calculated equilibrium concentrations by orders of magni-
130
A. Jordan
Table 6. Concentrations of selected organohalogens in volcanic gases: measured data and theoretically derived partial pressures (concentration data from Refs. [15, 19, 24–26]; calculated data from Refs. [5, 25]
CH3Cl CH2Cl2 CHCl3 CCl4 CFCl3 CF2Cl2 C2F4 C2Cl4
Measured data
Calculated for thermodynamic equilibrium
Relative discrepancy
10–9 –10–6 10–11 –10–8 10–11 –10–7 10–11 –10–8 10–10 –10–7 10–9 –10–7 >10–11 10–11 –10–9
10–22 –10–14 10–24 –10–21 10–30 –10–28 10–28 10–34 10–34 10–48 10–43
= 105 = 1010 = 1017 = 1017 = 1024 = 1025 = 1037 = 1032
tude [33 – 37]. It is generally assumed that these hydrocarbons are not of deep origin from the Earth’s mantle. They are generated at shallow depth from thermal decomposition of organic matter in the basement sedimentary rocks that surround the volcanoes. These thermogenic processes are enhanced by the above average geothermal gradient observed at active volcanic regions [38 – 40]. It is common that even high-temperature volcanic gases are mixtures of three components: one containing deep magmatic volatiles, one of atmospheric origin and a third of shallow origin that can contain hydrocarbon gases from hydrothermal systems [29, 41 – 43]. When these hydrothermal fluids get in contact with the hot magma column they cause a disequilibrium in the high-temperature volcanic gas samples. This mixture may not fully re-equilibrate again due to the slow oxidation reaction rates of hydrocarbons even at high temperatures [41, 43]. The most abundant hydrocarbon in these hydrothermal gases is methane. Smaller amounts of alkanes, alkenes and benzene have also been detected and their formation described in equilibrium under hydrothermal conditions [44]. Various hydrocarbon products have been detected in fumaroles of different volcanoes [23, 41, 45], and Capaccioni et al. [41] have described the different processes leading to the products observed. The reactions involved include thermal and catalytic cracking that produce light alkenes and aromatics. An experiment performed by Bondarev and Porshnev [46] demonstrated the formation of higher hydrocarbons when blowing methane and water vapour over red-hot lava. The dominant reaction products they observed were unsaturated hydrocarbons of low molecular weight, benzene and polyaromatic compounds as well as oxidised hydrocarbons. These groups of compounds were also detected by Stoiber et al. [15] and Isidorov et al. [23], and the main hydrocarbon they detected in volcanic gases was methane. Bondarev and Porshnev [46] studied lava at a temperature of 1030 °C, and this experiment therefore corresponds to the industrial process for ethyne production based on the pyrolysis of methane. The reaction by-products of this process are well known and consist mainly of ethene and smaller amounts of mixed alkene-ynes, alkynes and aromatics such as styrene
Fig. 1. Reaction sequence forming organohalogen compounds in volcanic gases
Volcanic Formation of Halogenated Organic Compounds
131
132
A. Jordan
[47]. The hydrocarbon fraction found in lava gases from Mt. Etna corresponded to this composition including all C3-C6 alkenes, benzene, toluene, styrene and heteroaromatic compounds, as well as small amounts of molecules such as buteneyne, butadiyne, styrene or phenylethyne, which have also been described as byproducts of the industrial ethyne production [47]. As the reaction mixture is rapidly quenched when it escapes the volcanic vents, even unstable products are preserved. The sequence of reactions forming organohalogens can be deduced from the reaction products observed in the studies reporting halocarbons and especially those detected in the lava gases from Mt. Etna. The overall reaction process can be described in several stages that all occur in parallel (Fig. 1). The starting point is the pyrolysis of hydrothermal methane or other hydrocarbons when they become in contact with the magmatic column. Pyrolysis products formed are mainly ethyne and ethene and minor amounts of other unsaturated hydrocarbons. Ethyne is very reactive and can form several further products. It condenses to give aromatic compounds as benzene, toluene and styrene, forms heterocyclic compounds such as furan by reaction with oxygen, thiophene by reaction with FeS2 or H2S, and selenophene by reaction with selenium [48]. All of these precursors, primary reaction products and further reaction products are in a chemical environment highly favouring halogenation reactions. These conditions include temperatures of several hundred degrees, high concentrations of hydrogen halides and the presence of catalytically active surfaces. The last of these are generated by halogenide sublimates of trace elements on juvenile ash particles. The predominant halogenated products are halogenated methanes and ethenes. Reaction of methane and HCl over Fe2O3 at 500 °C leads to the synthesis of chlorinated methanes, and halogenated alkenes are formed by addition of hydrogen halides to alkynes. Halogenated methanes with different halogens result, for example, from the reaction of methane with HBr over AlCl3 . As long as the reaction mixture has not left the lava, all of these products are not stable and can decline or react further. Pyrolysis of chloroform and CCl4 at 500°C leads to C2Cl4 and chloroethyne, respectively [48]. A variety of products similar to the multitude of unsaturated short-chain chlorocarbons detected at very low levels in lava gas samples from Mt. Etna has been described as reaction products from the thermolysis of trichloroethene and tetrachloroethene. Chloroethyne and ethyne then form chlorobenzene and other chlorinated aromatics. Further reactions include halogen exchange reactions. The formation path for chlorofluorocarbons is very likely a rehalogenation of CCl4 with HF. This reaction is facilitated by the presence of catalytically active metal halide compounds in the pyroclastic material. At Mt. Etna, Fe, Cu, Mn, Cd and especially high amounts of alkaline halides have been reported that have such catalytic properties [49]. This chemical reaction mechanism accounts for several features observed in volcanic gas mixtures: – the predominant abundance of halogenated methanes and ethenes points to methane and ethyne as the precursor compounds
Volcanic Formation of Halogenated Organic Compounds
133
– the great variety of different unstable short-chain organohalogens indicates a non-selective cracking mechanism with subsequent quenching that produces many thermodynamically unstable reaction products – the great variability in concentrations at one sampling location is also an indication of a non-equilibrium process. It seems that the cause for conflicting results from theoretical considerations and field sample analysis as shown in Table 6 was the implicit assumption that the volcanic gases studied were in thermodynamic equilibrium. This assumption is not appropriate for volcanic gas mixtures with portions of non-magmatic fluids that are added at shallow depth. An additional path of volcanic emission of halocarbons is the release of CF4 , SF6 , NF3 and chlorofluorocarbons trapped in fluid inclusions of igneous rocks on melting [50]. However, the resulting flux can be expected to be extremely low and neither CF4 nor SF6 has been detected in volcanic gases [51]. 3.3 Relative Abundance of Fluorine, Chlorine, Bromine and Iodine
The relative distribution of the single halogens within halogenated hydrocarbons clearly shows that chlorine is the most abundant halogen. This seems reasonable, as chlorine is the most abundant halogen in volcanic gases. However, the relative abundances of fluorinated, brominated and iodinated organic compounds cannot be explained simply by the halogen content of the magmatic gas (Table 7). Factors that also determine the formation rate of the different halocarbons include the degassing properties of the respective halogen and its reactivity under the temperature and chemical conditions present in the volcanic environment. 3.3.1 Halogen Concentrations in Magmas and Volcanic Gases
Chlorine contents of 1700–2100 ppm and Cl/Br ratios of about 250 have been reported for pre-eruptive magmatic melts [53, 54]. Most basaltic rocks contain 50–640 ppm wt F and 15–2000 ppm wt Cl [53], and the Cl/Br ratio in basalts has been shown to be fairly constant at 400 ± 100 for chlorine levels of 20 – 1000 ppm and bromine contents of 60 – 1300 ppb [55]. This corresponds to the exospheric chlorine/bromine ratio. Iodine concentrations are generally low in magmatic rocks with mass concentrations of 4–9 ppb [12]. One factor that can influence the halogen contents of volcanic gases and the relative halogen distribution are admixtures of brines generated from seawater or other hydrothermal waters [25]. As iodine has just a very small magmatic abundance compared to its contents in the Earth’s crust [12], such admixtures are likely to account for a substantial part of this element in volcanic gases. At Vulcano, changing bromine/chlorine ratios have been explained by different contributions of deep and shallow components feeding the fumarolic fluids [25]. The tectonic setting determines the halogen contents as andesitic magmas from convergent plate volcanoes generally have higher Cl and Br contents than basalts from divergent-plate or hot-spot volcanoes
134
A. Jordan
Table 7. Parameters determining the abundance of organohalogens in magmatic gases
Concentration in magmatic rocks (ppm)
Fluid/melt partitioning coefficients [10, 54] Hydrogen halide concentration in volcanic gases (vol. %) Molar ratio chlorine/ other halogen Bond dissociation energy (kJ mol–1) [66]
F
Cl
Br
I
basalts: 50–300 andesites: 240–640 [53] rhyolites: 400–640 [53] 1
basalts: 50–200 andesites: 1000–1800 [53] rhyolites: 600–1900 [53] 8–10
50 ppb– 2500 ppb [55]
4–11 ppb [12, 55]
3.7–17.5
104
0.003–0.7 [29, 49, 53, 56]
0.02–6 [29, 49, 53, 56]
no data
no data
75–2500 [61, 62] 366.3
10,000 [61] 298.4
2–100 [57– 60] 569.9
431.6
[29]. The relative halogen content of single volcanoes is also not constant, but changes with different stages of volcanic activity [60, 63]. A compilation of published data of HCl and HF in volcanic gases is summarized in Table 7 together with relative molar ratios of the halogens. There are very little data on bromine in volcanic gases and none about iodine. 3.3.2 Degassing Properties
The availability of the reactive halogens in the magmatic melt is determined by the degassing properties of the respective halogen and the presence of trace elements that form volatile halogenide complexes. Degassing is a rather complex dynamic process involving fluid-magma interaction, shallow depth crystallization and eruption dynamics. A major difference between fluorine and the other halogens is the solubility of HF in aluminosilicate melts, which is significantly higher compared to those of HCl or HBr. Hence, fluorine degassing is very low and it is strongly concentrated in magmas relative to any coexisting vapour phase [54, 64, 65]. Mean fluorine concentrations of erupted magma have been found to be very close to the pre-eruptive value [54] or at 70–90% of its initial fluorine content [58]. HCl, HBr and HI exhibit an opposite behaviour and are released to more than 90% from the melts [10]. Cl/Br systematics indicate that Br behaviour is similar to Cl in most magmatic processes [54]. 3.3.3 Chemical Reactivity
The chemical bond energies of the hydrogen halides drastically increase from HI to HF [66]. Iodine forms radicals most easily, whereas the HF molecule is ex-
Volcanic Formation of Halogenated Organic Compounds
135
tremely stable and no direct fluorination of saturated hydrocarbons will occur even at 1000°C. This explains the small amounts of fluorinated compounds as compared to the relative abundance of fluorine compared to its homologues. The formation of organohalogens is dependent not only on the available precursor compounds methane and halogen halide, but it is also greatly influenced by the presence of catalytically active transition metal compounds. Therefore, differences in the lava chemistry may also cause large differences in the occurrence of halocarbons in volcanic gases. Isidorov [24] supposed that a likely reason for the absence of detectable chlorofluorocarbons and other fluorinated compounds in volcanic gas samples from Hawaii [17, 19] was the low fluorine levels present in this type of volcano compared to volcanoes in Central America and Kamchatka. However, Mt. Etna has been stated to be the biggest point source globally for HF and HCl [57], and HCl/HF ratios reported for Mt. Etna [57, 58] are similar to those from Santiaguito volcano and Nicaraguan volcanoes [59, 67]. Yet, in most of the samples taken at Mt. Etna, no chlorofluorocarbons were present. Although the Antarctic volcano Mt. Erebus is also an important halogen emitter [60], no CF3-containing compounds were detected by Penkett et al. [52]. Therefore, it is questionable whether most of the Earth’s volcanoes are likely to emit chlorofluorocarbons at concentrations of 100 ppb as suggested by Isidorov [23].
4 Environmental Impact of Volcanic Organohalogen Production and Global Emissions Several of the compounds listed in Tables 2 and 3 have known toxic properties. Some examples include carbon tetrachloride, vinyl chloride, chloroethyne or methyl iodide. Although the existence of polychlorinated dibenzodioxins and dibenzofurans (PCDD/F) has not yet been detected because of inadequate analytical methods, their formation seems quite probable in view of the existence of furan as well as halogenated benzenes. However, all of these toxic organohalogens have been detected at low trace levels. They get even further diluted in the volcanic plume so that the risk of intoxication from these compounds is negligible compared to other hazards of volcanic activity. Several toxic inorganic gases are present at far higher levels such as hydrogen sulfide, carbon disulfide, arsine and especially hydrogen fluoride. The biggest danger associated with volcanic gas arises from extremely hot, sudden volcanic gas flows of CO2 and SO2 dominated gases that regularly lead to victims among the nearby settling people [68]. Large volcanic eruptions not only have devastating consequences on a regional scale, but they also have an impact on the global climate [69–71]. Sulfuric acid aerosol generated in the stratosphere from volcanic SO2 , scatters back the solar radiation to space and thus produces a cooling effect. These sulfuric acid particles also form polar stratospheric clouds (PSC) that play a crucial role in the heterogeneous processes of chlorine radical activation that lead to stratospheric ozone depletion. However, the injection of inorganic halogens to the stratosphere in major eruptions does not significantly contribute to the ozone-destroying processes
136
A. Jordan
[73], which are mainly caused by chlorofluorocarbons [74] and other chlorocarbons. The detection of chlorofluorocarbons in volcanic gases gave rise to the hypothesis that this could be a significant natural source for these compounds [23, 75]. Isidorov [23] proposed that the volcanic source could have caused a natural background concentration of 17 pptv CFCl3 referring to ambiguous data from ice core air analysis [76]. Volcanic emissions could also be contributors to the missing source in the global budgets for methyl halides. These are still imbalanced with respect to the source strengths of the known sources [77]. The important question was therefore to establish the actual quantities of halocarbons emitted by volcanoes to evaluate the significance of these fluxes compared to the global halocarbon budgets. It has been suggested that at any given time 50 of the 550 active volcanoes are emitting large plumes and 50 others are emitting small ones [78]. Up to now there are concentration data for organohalogens in gases for only nine volcanoes, which makes it obvious that the current data set can just provide a guess of quantitative emissions. The formation mechanism for volcanogenic organohalogens discussed in this chapter implies the co-action of magmatic gas with non-magmatic hydrocarbons. Thus, their volcanic fluxes cannot be directly linked to the emission of volcanic fume that makes it even more difficult to assess. To get a very rough idea of the order of magnitude, organohalogen emissions are estimated to be linked to inorganic halogen fluxes. There are two general approaches that have been used to estimate global halogen fluxes of volcanic gas emissions. One is the petrological method, which interprets the halogen contents of minerals crystallized during different eruption stages; the other method is based on monitored concentration data of single volcanoes to sum up these fluxes and scale them to global emissions. The petrological approach uses differences in halogen concentrations of pre-eruptive magma (represented by crystal-free compositions which are assumed to represent the non-degassed melt composition) and fully degassed lava which is multiplied by the total mass of ejected lava to calculate the amount of gas released during single eruptions. This method has shown to produce comparable results to emission rate estimates for hydrogen halide emissions obtained with other methods for selected volcanic eruptions [79, 80]. Cadle [7] made a simplified calculation to estimate the global HCl and HF fluxes by multiplying the mean annual volume of lava emitted over 400 years (1 km3 year–1) times the adopted mass fraction of volatiles released from the melt (2.5%) times an average concentration of HCl and HF in these gases (1% and 0.05%, respectively).Adjusting his input data according to more recent data for lava production (4 km3 year–1) shifts his estimate of 0.78 Tg HCl to 3 Tg year–1, which is identical with estimates for oceanic influx calculation performed by Schilling et al. [81] and Jambon et al. [9]. However, this simplified approach suffers from a systematic drawback. Subduction zone volcanoes account for 80% of the historical eruptions, but only of 10% of the mass of the global lava production, whereas the rift zone volcanoes produce the bulk amount of ejected lava [82], but have a lower fraction of volatiles that is of different chemical composition [29]. The second method to assess the inorganic halogen fluxes from volcanoes was used by Symonds et al. [5]. It links the volcanic hydrogen halide fluxes to volcanic
Volcanic Formation of Halogenated Organic Compounds
137
SO2 emissions for which a reliable global inventory has been compiled (2.0 ¥ 1011 mol year–1 = 12 Tg year–1) [83]. Typical molar ratios of HCl/SO2 and HF/SO2 in volcanic gases of 1–10 and 1–100, respectively, are assumed to calculate annual global emissions of 0.4–11 Tg HCl and 0.06–6 Tg HF. This confirms the estimate made by Cadle. However, the assumption of a proportionality between SO2 and halogen emissions also constitutes a big simplification as shown by sulfur/halogen ratios reported in the literature [29], and the S/Cl ratio shows a dependence on the volcanic activity [60, 63]. Using this approach, Symonds et al. [5] also calculated halocarbon emissions. On the basis of the organohalogen fugacities calculated for the thermodynamic equilibrium under magmatic conditions, an annual flow of just 1 g of CH3Cl was derived and an even lower amount for chlorofluorocarbons. Wahrenberger [25] came up with a source strength of about 330 t year–1 for CH2Cl2 and 10 t year–1 for CFCl3 calculated in the same way, but taking measured concentration data, whereas Jordan et al. [26] expected upper limits of 15 t year–1 for CH3Cl, the most abundant halogenated hydrocarbon, and less than 0.3 t year–1 for CFCl3. Both flux estimates are distinctly below anthropogenic chlorofluorocarbon emissions and have no significance compared to other natural or anthropogenic sources for other halocarbons. In particular, such a small CFCl3 source falls short of producing an atmospheric background concentration of 17 pptv that was previously suggested [23]. Recent investigations of polar firn ice have unambiguously confirmed the absence of any pre-industrial background of chlorofluorocarbons [84].
5 Summary Although evidence has been given that proves the existence of halogenated hydrocarbons in volcanic gases, this volcanic production of organohalogens has been controversial for a long time. This dispute appears to have been caused mainly by a different understanding of the term of “volcanic gases”. Halogenated organic compounds are not expected to be produced in thermodynamic equilibrium under the temperature, pressure and chemical conditions of the mantle. However, it is very common that gases escaping at fumarolic vents constitute mixtures of a magmatic component with hydrothermal fluids containing hydrocarbons. This interaction of non-magmatic hydrocarbons with hot magmatic gases is likely to form volcanogenic halocarbons. Although emission rates are likely to be too low to have any significance for the global atmospheric budgets of relevant halogenated trace gases, the reaction products resulting from this process show a fascinating variety. Acknowledgement. I wish to thank Jochen Harnisch and Francois LeGuern for having inspired me to work in this exciting field of research. I also thank Jochen Harnisch, Frank Keppler and Matthias Frische for their helpful comments during the preparation of this manuscript.
138
A. Jordan
6 References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44.
Rubey WW (1951) Bull Geol Soc Am 62:1111 Warneck P (1988) Chemistry of the natural atmosphere. Academic Press, San Diego, p 602 Tajika E (1998) Geophys Res Lett 25 : 3991 Harnisch J (1999) Reactive fluorine compounds. In: Fabian P, Singh ON (eds) Reactive Halogen compounds in the atmosphere. Springer, Berlin Heidelberg New York Symonds RB, Rose WI, Reed MH (1988) Nature 337:415 Stoiber RE, Williams SN, Huebert B (1987) J Volcanol Geotherm Res 33:1 Cadle RD (1975) J Geophys Res 80:1650 Cadle RD (1980) Rev Geophys Space Phys 18:746 Jambon A, Déruelle B, Dreibus G, Pineau F (1995) Chem Geol 126:101 Bureau H, Keppler H, Métrich N (2000) Earth Planet Sci Lett 183:51 Anderson AT (1974) Geol Soc Amer Bull 85:1485 Muramatsu Y, Wedepohl KH (1998) Chem Geol 147:201 Graedel TE, Keene WC (1995) Glob Biogeochem Cycl 9:47 Cicerone RJ (1981) Rev Geophys Space Phys 19 : 123 Stoiber RE, Leggett DC, Jenkins TF, Murrmann RP, Rose WI (1971) Geol Soc Amer Bull 82:2299 Molina MJ, Rowland FS (1974) Nature 249:810 Rasmussen RA, Rasmussen LE, Khalil MAK, Dalluge RW (1980) J Geophys Res 85 : 7350 Rasmussen RA, Khalil MAK, Dalluge RW, Penkett SA, Jones B (1982) Science 215 : 667 Rasmussen RA (1978) cited in Ref. [7] Hobbs PV, Tuell JP, Hegg DA, Radke LF, Eltgroth MW (1982) J Geophys Res 87:11062 Inn ECY, Vedder JF, Condon EP (1981) Science 211:821 Pereira WE, Rostad CE, Taylor HE (1980) Geophys Res Lett 7:953 Isidorov VA (1990) Organic chemistry of the Earth’s atmosphere. Springer, Berlin Heidelberg New York, p 98 Isidorov VA, Zenkevich IG, Ioffe BV (1990) J Atmos Chem 10 : 329 Wahrenberger CM (1997) Some aspects of the chemistry of volcanic gases, PhD thesis, ETH Zuerich Jordan A, Harnisch J, Borchers R, Le Guern F, Shinohara H (2000) Environ Sci Technol 34:1122 Frische M (2001) personal communication Gerlach TM (1982) Bull Volcanol 45:235 Symonds RB, Rose WI, Bluth GJS, Gerlach TM (1994) Rev Mineral 30:1 Symonds RB, Reed MH (1993) Am J Sci 293:758 Gerlach TM (1980) J Volcanol Geotherm Res 7 : 295 Gaffney JS (1995) Environ Sci Technol 29:8A Giggenbach WF (1996) Geochim Cosmochim Acta 61:3763 Chiodini G, Cioni R, Marini L (1993) Appl Geochem 8:357 Giammanco S, Inguaggiato S, Valenza M (1998) J Volcanol Geotherm Res 81:297 Matsuo S, Ossaka J, Hirabayashi J, Ozawa T, Kimishima K (1982) Bull Volcanol 45:261 D’Alessandro W, De Gregorio S, Dongarra G, Gurrieri G, Parello F, Parisi B (1997) J Volcanol Geotherm Res 78 : 65 Kiyosu Y, Asada N (1995) Geochem J 29 : 231 Darling WG (1998) Appl Geochem 13:815 Welhan JA (1988) Chem Geol 71 : 183 Capaccioni B, Martini M, Mangani F (1995) Bull Volcanol 56:593 Shinohara H, Giggenbach WF, Kazahaya K, Hedenquist JW (1993) Geochem J 27:271 Taran YA, Connor CB, Shapar VN, Ovsyannikov AA, Bilichenko AA (1997) Bull Volcanol 58:441 Capaccioni B, Mangani F (2001) Earth Planet Sci Lett 188:543
Volcanic Formation of Halogenated Organic Compounds
139
45. Mangani F, Capiello A, Capaccioni B, Martini M (1991) Chromatographia 32:441 46. Bondarev VB, NV Porshnev NV (1980) cited in Ref. [23] 47. Elvers B, Ullmann F (eds) (1985) Ullmann’s encyclopedia of industrial chemistry, 5th edn. Verlag Chemie Weinheim, vol A1, p 105 48. Beilstein FK (1918) Handbook of organic chemistry, and supplements.Verlag Beilstein Information, Frankfurt 49. Le Guern F (1988) Ecoulements gazeux reactifs a hautes temperatures, PhD thesis Univ Paris VII 50. Harnisch J, Frische M, Borchers R, Eisenhauer A, Jordan A (2000) Geophys Res Lett 27:1883 51. Harnisch J, Eisenhauer A (1998) Geophys Res Lett 25:2401 52. Penkett SA, Prosser JD, Rasmussen RA, Khalil MAK (1981) J Geophys Res 86 :5172 53. Giggenbach WF (1996) Chemical composition of volcanic gases. In: Scarpa R, Tilling RI (eds) Monitoring and mitigation of volcano hazards. Springer, Berlin Heidelberg New York 54. Villemant B, Boudon G (1999) Earth Planet Sci Lett 168 : 271 55. Jambon A (1994) Rev Mineral 30:479 56. Gerlach TM, Graeber EJ (1985) Nature 313 : 273 57. Francis P, Burton MR, Oppenheimer C (1998) Nature 396:567 58. Pennisi M, Le Cloarec MF (1998) J Geophys Res 103:5061 59. Lazrus AL, Cadle RD, Gandrud BW, Greenberg JP, Huebert BJ, Rose WI (1979) J Geophys Res 84:7869 60. Zreda-Gostynska G, Kyle PR (1993) Geophys Res Lett 20 : 1959 61. Sugiura T, Mizutani Y, Oana S (1963) J Earth Sci Nagoya Univ 11:272 62. Kraft K, Chaigneau M (1976) C R Acad Sci Ser D 282:341 63. Rose WI, Heiken G, Wohletz K, Eppler D, Barr S, Miller T, Chuan RL, Symonds RB (1988) J Geophys Res 93:4485 64. Rowe EC, Schilling JG (1979) Nature 279:33 65. Carroll MR, Webster JD (1994) Rev Mineral 30:231 66. Lide DR (1999) CRC handbook of chemistry and physics 79th edn. CRC Press, Boca Raton, FL 67. Stoiber RE, Williams SN, Huebert BJ (1986) J Geophys Res 91:12215 68. Le Guern F, Tazieff H, Faivre-Pierret R (1982) Bull Volcanol 45:153 69. Robock A (2000) Rev Geophys 38 : 191 70. Lindzen RS, Giannitsis C (1998) J Geophys Res 103:5929 71. White DE, White JMC, Steig EJ, Barlow LK (1997) J Geophys Res 102:19683 72. McCormick MP, Thomason LW, Trepte CR (1995) Nature 373:399 73. Tie XX, Brasseur G (1995) Geophys Res Lett 22:3035 74. Russel JM, Luo MZ, Cicerone RJ, Deaver LE (1996) Nature 379:526 75. Gribble GW (1995) Environ Sci Technol 29:8A 76. Khalil MAK, Rasmussen RA (1982) Chemosphere 11:877 77. Butler JH (2000) Nature 403 : 260 78. Stoiber RE, Jepsen A (1973) Science 182 : 577 79. Thordarson T, Self S, Oskarsson N, Hulsebosch T (1996) Bull Volcanol 58:205 80. Edmonds M, Pyle D, Oppenheimer C (2001) Earth Planet Sci Lett 186:159 81. Schilling JG, Unni CK, Bender ML (1978) Nature 273:631 82. Tilling RI (1996) Hazards and climatic impact of subduction-zone volcanism. In: Bebout GE, Scholl DW, Kirby SH, Platt JP (eds) Subduction top to bottom. American Geophysical Union, Washington, p 331 83. Andres R J, Kasgnoc AD (1998) J Geophys Res 103:25251 84. Butler JH, Battle M, Bender ML, Montzka SA, Clarke AD, Saltzman ES, Sucher CM, Severinghaus JP, Elkins JW (1999) Nature 399:749
The Handbook of Environmental Chemistry Vol. 3, Part P (2003): 141–169 DOI 10.1007/b 10454
Fluorinated Natural Products: Occurrence and Biosynthesis David B. Harper 1, David O’Hagan 2, Cormac D. Murphy 2 1
2
Microbial Biochemistry Section, School of Agriculture and Food Science, The Queen’s University of Belfast, Newforge Lane, Belfast, BT9 5PX, UK E-mail:
[email protected] University of St Andrews, School of Chemistry, Biomolecular Sciences, North Haugh, St Andrews, Fife, KY16 9ST, UK
Despite the abundance of fluorine in the Earth’s crust naturally occurring organofluorine compounds are comparatively rare in nature with those isolated to date totalling no more than about a dozen confined to a few plants and two species of bacteria. Notwithstanding considerable interest and a variety of speculative suggestions, the mechanism of biological C-F bond formation is still unknown, although significant progress has been made in elucidating the pathway by which biosynthesis of fluoroacetate and 4-fluorothreonine occurs in the bacterium Streptomyces cattleya. In this chapter we review the nature and distribution of organofluorine compounds formed biologically and discuss the progress made in our understanding of C-F bond biosynthesis over the 60 years since fluoroacetate was first identified as a natural product. Keywords. Organofluorine, Fluoroacetate, Natural products, Biosynthesis, Plants, Bacteria,
Dichapetalum, Streptomyces
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
1
Introduction
2
Fluorinated Natural Products in Plants . . . . . . . . . . . . . . . 143
2.1 2.2 2.3 2.4
Fluoroacetate . . . Fluorocitrate . . . Fluoroacetone . . . w-Fluorofatty Acids
3
Fluorinated Natural Products from Microorganisms
3.1 3.2
Nucleocidin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 Fluoroacetate and 4-Fluorothreonine . . . . . . . . . . . . . . . . 150
4
Biosynthesis of Fluorinated Natural Products . . . . . . . . . . . 151
4.1 4.1.1 4.1.2 4.1.3 4.1.4 4.2
Plants . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ethylene or Acetate as Precursors . . . . . . . . . . . . . The Haloperoxidase Reaction . . . . . . . . . . . . . . . Fluorophosphate as an Intermediate . . . . . . . . . . . . Pyridoxal Phosphate Catalysed Incorporation of Fluoride Microorganisms . . . . . . . . . . . . . . . . . . . . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
143 145 146 147
. . . . . . . 149
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
152 152 153 154 155 156
© Springer-Verlag Berlin Heidelberg 2003
142 4.2.1 4.2.2 4.2.3 4.2.4 4.2.5
D.B. Harper et al.
4.2.6 4.2.7
Incorporation of Glycolate . . . . . . . . . . . . . Incorporation of Glycine, Serine and Pyruvate . . Incorporation of Glycerol and Glucose . . . . . . Stereochemistry of Fluorination . . . . . . . . . . Fluoroacetaldehyde: the Common Precursor of the Fluorometabolites in S. cattleya . . . . . . . . . . Defluorination in S. cattleya . . . . . . . . . . . . Enzymic Synthesis of Glycosyl Fluorides . . . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
. . . .
157 158 160 162
5
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
. . . . . . . . . 163 . . . . . . . . . 164 . . . . . . . . . 165
1 Introduction Of all the halogens, fluorine is the most abundant with concentrations in the Earth’s crust in the range 270–740 ppm compared to 10–180 ppm for those of chlorine [1, 2]. However, the bulk of the element is bound in insoluble form with the consequence that seawater contains only 1.3 ppm fluoride in contrast to 1900 ppm chloride. The relative insolubility of fluorine-containing minerals renders the halogen largely unavailable to living organisms, which is one of the principal reasons for the dearth of organofluorine compounds in the biosphere compared with the multitude of natural products containing the other halogens. However, another equally important factor constraining the participation of fluorine in biochemical processes is the high heat of hydration of the fluoride ion [3]. In water, fluoride is thus heavily hydrated which ensures that it is only weakly nucleophilic and restricts the role that it can play in displacement reactions. The greater heat of hydration of the fluoride ion relative to those of other halide ions is mainly responsible for the high redox potential necessary to generate F+ from F– compared to that required to generate the other halonium ions from their respective halide ions [4]. This difference precludes incorporation of fluorine into natural products by the haloperoxidase reaction, which had been considered until recently the major route by which organohalogen compounds are formed in nature [5]. Despite the low concentrations of available fluorine in soil, some marine and terrestrial organisms can accumulate significant quantities of inorganic fluoride. Thus, the marine sponge Halichondria moorei can contain as much as 10% fluorine on a dry wt basis as potassium fluorosilicate [6].Amongst terrestrial plants the genus Camellia, which includes the tea plant (Thea sinensis syn Camellia sinensis), is noted for its ability to selectively concentrate fluoride from the soil [7, 8]. Fluoride concentrations of 70–180 mg kg–1 dry wt have been reported in commercial tea and levels of up to 3000 mg kg–1 dry wt have been recorded in older leaves of ornamental Camellia species. Some plant species growing in areas of fluoride rich bedrock or on sites contaminated with fluorspar (CaF2) mining waste can accumulate up to 10,000 mg kg–1 dry wt fluoride [9, 10]. However, organically-bound fluorine has been identified in only a few genera of tropical and subtropical plants and in only two species of microorganism. Not a single organofluorine compound has been isolated from any marine organism. More-
Fluorinated Natural Products: Occurrence and Biosynthesis
143
over, to date there is no evidence of de novo synthesis of the C-F bond in animals or insects. Although caterpillars of the moth Sindrus albimaculatus can contain fluoroacetate at concentrations that render them highly toxic to predators, the compound is derived from their host plant Dichapetalum cymosum [11]. The moth Nygmia pseudoconspersa which feeds on the tea plant can concentrate fluorine to 5000 mg kg–1 dry wt in the wings and 1100 mg kg–1 dry wt in the cocoon [12]. Whether any of this fluorine is in organic form is not known but the toxicity associated with the scales of the insect suggests that further investigation might be worthwhile. In this chapter we consider the nature and distribution of the few organofluorine compounds formed biologically and review the progress made in our understanding of the mechanism of C-F bond synthesis in nature over the 60 years since fluoroacetate was first identified as a natural product.
2 Fluorinated Natural Products in Plants 2.1 Fluoroacetate
Fluoroacetate (1) was the first fluorine-containing natural product discovered and is by far the most widespread. It was initially isolated in South Africa by Marais [13, 14] in 1943 as the toxic principle in the low growing shrubby plant Dichapetalum cymosum. The inhabitants of the Transvaal had long recognised this plant as a hazard to livestock and had accordingly named it gifblaar (poison leaf) [15]. The young leaves of the plant in early spring are particularly toxic and can contain up to 2500 mg kg–1 dry wt. [16, 17]. Many other species of the genus Dichapetalum are now known to produce fluoroacetate. The highest concentrations reported to date are those found in D. braunii from Tanzania in which 8000 mg kg–1 on a dry wt were recorded in the young leaves and seeds [18]. Fluoroacetate has also been identified in the leaves of D. heudelotti [19], D. stuhlmannii [18, 20] and D. toxicarium [21, 22], a West African species which accumulates w-fluorinated lipids in the seeds (see Sect. 2.4). Although not chemically confirmed, the presence of fluoroacetate is inferred on the basis of toxicological evidence in nine other Dichapetalum spp [22]. The toxicity of all parts of the West African plant Spondianthus preussi of the Euphorbiaceae has also been ascribed to the presence of a mixture of oxalate and fluoroacetate although the chemical characterisation of fluoroacetate reported is far from conclusive [23]. In Australia fluoroacetate is produced by around forty plant species from three genera of the Leguminosae [24, 25]. In some areas a single species can constitute 80% of the understorey vegetation. Thirty-three species belonging to the genera Gastrolobium and Oxylobium are confined to the southwest corner of Western Australia. Oxylobium parviflorum (box poison) and Gastrolobium bilobum (heart leaf poison) are amongst the most toxic containing up to 2600 mg kg–1 dry wt of fluoroacetate in their leaves and up to 6500 mg kg–1 in their seeds [24, 26]. Recently, a reclassification of these genera has been conducted employing the presence or absence of fluoroacetate as one of the taxonomic criteria. Many of the
144
D.B. Harper et al.
non-toxic Gastrolobium spp were placed in the genus Nemcia and the toxic Oxylobium spp transferred to Gastrolobium [27]. However, the recent identification of fluoroacetate in Nemcia spathulata (formerly G. spathulatum) has cast doubt on the reliability of using the presence of this natural product as an indication of genetic relationships within these species [25]. Three species accumulating fluoroacetate occur in northern and central Australia but are much less toxic than those from the southwest of the country. Leaves of G. grandiflorum (wall-flower poison) can contain up to 185 mg kg–1 dry wt, and seedpods of G. brevipes up to 300 mg kg–1 dry wt [28, 29]. Fluoroacetate is also frequently present in the leaves and seeds of Acacia georginae (gidyea) a small tree found in Queensland [30 – 32]. Levels of 400 mg kg–1 dry wt have been reported in seeds, but amounts vary considerably; samples from some geographical locations show no detectable organic fluorine suggesting significant genetic variability [20].Although fluoroacetate-producing plants are in general confined to Australia and Africa, a Brazilian species Palicourea marcgravii of the Rosaceae can accumulate concentrations of 5000 mg kg–1 dry wt in seeds and flower stalks [20, 33]. Fluoroacetate has also been reported, albeit at comparatively low concentrations (10 mg kg–1 dry wt), in guar gum from the legume Cyamopsis tetragonolobus found on the Indian subcontinent [34].
A study of the literature reveals large discrepancies in the fluoroacetate concentrations reported by different investigators for a particular species. At least some of this variation is associated with changes in fluoroacetate content with the season and age of the plant. In general, young shoots and leaves possess high levels of the toxin in spring but, as the season progresses and the plant matures, concentrations fall, sometimes below the detection threshold [17, 18, 21]. Fluoroacetate content also varies markedly between different plant organs with ephemeral tissues such as flowers, seeds and young leaves containing the greatest concentrations [20]. This localization is consistent with the chemically-mediated defence strategies displayed by many plant species against herbivores [24]. Geographical location can sometimes be an important factor in determining fluoroacetate concentrations in a species but this does not seem to be a reflection of differences in the fluorine status of the soil as most plants accumulating fluoroacetate are capable of concentrating the low levels of fluoride normally present [24]. Instead, genetic differences in plant populations in different areas appear a more likely explanation of the wide range of fluoroacetate concentrations observed within a single species [20, 35]. When grown in a high fluoride environment, some widely cultivated crops and forage plants such as soya bean (Glycine max) and crested wheat grass (Agropyron cristatus) can biosynthesise fluoroacetate (and fluorocitrate) albeit in trace
Fluorinated Natural Products: Occurrence and Biosynthesis
145
non-toxic quantities [36–38]. Experiments with cell cultures of soya bean grown in the presence of 1 mM fluoride show fluoroacetate concentrations of 4 mg kg–1 dry wt [39]. In a survey of plant material collected from areas in Finland with bedrock rich in fluoride, fluoroacetate was present above the detection limit of 0.005 mg kg–1 dry wt in 76 % of over 100 samples examined. The overall mean concentration of fluoroacetate was 0.034 mg kg–1 dry wt with a maximum concentration of 0.21 mg kg–1 dry wt [40]. These workers also reported the analysis of 21 samples of commercial tea which displayed fluoroacetate concentrations ranging from 0.06 to 0.48 mg kg–1 dry wt (mean 0.19 mg kg–1 dry wt). The ability of the tea plant (Camellia sinensis) to selectively concentrate fluoride from the soil was mentioned in Sect. 1. Thus, it appears that many plants have the potential to biosynthesise fluoroacetate in a high fluoride environment. The plant species producing toxic levels in nature have simply refined and enhanced this ability as a chemical defence. Callus tissue culture potentially offers an accessible and convenient system in which to study the biological fluorination process and several attempts have been made to establish cultures of fluoroacetate producing plants. Callus cultures of A. georginae were generated in 1970 from stem sections [41, 42] and independently in 1983 from leaf discs [43]. Unfortunately, no quantification of fluoroacetate production was conducted in either case. Work in South Africa led to the establishment of a tissue culture from the immature fruits of D. cymosum [44]. Under optimum conditions in the presence of 6 mM fluoride the culture initially yielded fluoroacetate at a concentration of 1200 mg kg–1 dry wt. Disappointingly, this production rate was not maintained on continued subculture and yields eventually stabilised at only 150 mg kg–1 dry wt [45]. 2.2 Fluorocitrate
Fluoroacetate can undergo further metabolism in vivo to form fluorocitrate (2) and it is this transformation for which Peters et al. [46] coined the term ‘lethal synthesis’ that is responsible for the toxicity of fluoroacetate. Many fluoroacetateaccumulating plants including soya bean (Glycine max) [38] and alfalfa (Medicago sativa ssp sativa) [36] thus contain trace amounts of fluorocitrate. Low concentrations are even present in commercial tea (up to 30 mg kg–1 dry wt) and oatmeal (up to 60 mg kg–1 dry wt) albeit well below the toxic threshold [39]. Fluorocitrate is formed in vivo by activation of fluoroacetate to fluoroacetyl CoA, which then acts as a substrate for the citric acid cycle enzyme, citrate synthase. In this reaction, fluoroacetyl CoA is condensed with oxaloacetate to generate the single stereoisomer (2R, 3R)-2-fluorocitrate [47]. The stereospecific nature of this reaction is attributable to the removal solely of the 2-pro-S hydrogen of fluoroacetyl CoA. A detailed computational analysis [48] revealed that abstraction of the 2-pro-S hydrogen generates a lower energy enzyme-bound E-enol intermediate 4, whereas abstraction of the 2-pro-R hydrogen leads towards a higher energy enzyme-bound Z-enol (Scheme 1). The energy difference between these intermediates (~4.3 kcal mol–1) is sufficient to account for the exclusive partitioning towards 2-pro-S hydrogen abstraction. It is interesting to note that of the
146
D.B. Harper et al.
four possible stereoisomers of fluoroacetate, the biologically formed isomer is the only one that is toxic [47]. This compound is dehydrated by aconitase, the enzyme that succeeds citrate synthase in the citric acid cycle. Although the enzyme reaction follows a similar course to that with citrate as substrate, the enzyme intermediate formed is a reactive allyl fluoride which undergoes an SN2¢ substitution by hydroxide to generate 4-hydroxy-trans-aconitate, a potent inhibitor of aconitase. The X-ray structure of 4-hydroxy-trans-aconitate bound to citrate synthase has been solved and supports this hypothesis [49].
Scheme 1. Stereospecific condensation of fluoroacetyl CoA with oxaloacetate to form fluoroc-
itrate
2.3 Fluoroacetone
The identification of fluoroacetone (3) in homogenates of the fluoroacetate accumulating plant Acacia georginae has been reported by Peters and Shorthouse [50, 51]. Incubation of homogenates with fluoride,ATP and pyruvate led to a significant decrease (up to 34%) in total fluorine. To investigate this loss, the volatiles released during incubation were passed through a solution of 2,4-dinitrophenylhydrazine yielding a 2,4-dinitrophenylhydrazone derivative with an identical retention time on paper chromatography to that of fluoroacetone. The fluorine recovered in this form amounted to approximately 13% of that lost from homogenates. Fluoroacetone had previously been identified in rat liver perfused with fluoroacetate [52] presumably formed by condensation of fluoroacetyl CoA with malonyl-ACP generating 4-fluoroacetoacetyl CoA, followed by hydrolysis and decarboxylation. A similar process may be operating in homogenates of A. georginae. However, as Peters and Shorthouse themselves point out [51], their derivatisation technique could not distinguish between fluoroacetone and fluoroacetaldehyde. With a role for fluoroacetaldehyde now emerging in the biosynthesis of fluoroacetate in S. cattleya (see Sect. 4.2.5), it is conceivable that the hydrazone derivative isolated was that of fluoroacetaldehyde.
Fluorinated Natural Products: Occurrence and Biosynthesis
147
2.4 w -Fluorofatty Acids
The seeds of the W. African shrub Dichapetalum toxicarium are highly toxic and were reportedly used at one time by witchdoctors to produce paralysis, loss of sensation in the limbs and ultimately death [53]. Ingestion by browsing animals frequently causes livestock losses and the ground seeds are widely employed as a rodenticide explaining the local name of the plant, ratsbane.Although, as mentioned in Sect. 2.1, the young leaves of the plant can accumulate fluoroacetate, this compound is not the toxic principle in the seeds, which nevertheless can contain up to 1800 µg g–1 dry wt of organic fluorine [20]. In 1959, Peters and co-workers [54–56] succeeded in isolating the principal fluorinated compound in the seed oil which accounted for around 80% of the organic fluorine present and identified it as w-fluorooleic acid (C18 : 1F) (8). Small amounts of w-fluoropalmitic acid (C16 : 0F) were also purified from the oil. A recent re-examination of the lipids of the seed oil [57] by gas chromatography/mass spectrometry (GC/MS) has established the presence of an additional five fluorinated acids, w-fluoropalmitoleic (C16 : 1F), w-fluorostearic (C18 : 0F), w-fluorolinoleic (C18 : 2F), w-fluoroarachidic (C20 : 0F) and w-fluoroeicosenoic (C20 : 1F) acids. In this latter study, w-fluorinated compounds represented 12.9% of the total fatty acid content of the seed oil. The fluorine in this fraction was mainly distributed between C18 : 1F (74%), C18 : 0F (16%), C18 : 2F (6%) and C16 : 0F (4%) with other fluorinated fatty acids present only in trace amounts. GC/MS analysis of the picolinyl esters of the unsaturated acids has allowed the position of the double bonds to be located [58]. The C16 : 1F and C20 : 1F acids were each shown to occur as two isomers with unsaturation at the 7- or 9-positions and the 9- or 11-positions, respectively. Threo-18-fluoro-9,10-dihydroxystearic acid (7) has also been identified in small amounts (1 % of the organic fluorine) in seed oil of D. toxicarium by Harper et al. [59] who have proposed that this compound is derived from the C18 : 1F acid via the 9,10-epoxide 6. Although this compound has not been isolated, strong evidence for its presence in seed oil was obtained using BF3-methanol transesterification-silylation techniques followed by GC/MS in the single-ion monitoring mode [57]. The chain lengths and degree of unsaturation of the w-fluorofatty acids present in seed oil paralleled those of their non-fluorinated analogues, but the former were 5–10-fold less abundant [57, 59]. This close correspondence can be explained if the fatty acid synthase in this plant has sufficiently broad substrate specificity to utilize fluoroacetyl CoA in addition to acetyl CoA for condensation with malonyl acyl carrier protein (malonyl-ACP), the initial stage in fatty acid synthesis (Scheme 2). The observation that fluorine is confined to the terminal position in fluorofatty acids and D. toxicarium indicates enzymic constraints on the use of fluorinated analogues at subsequent stages of the biosynthetic process. Either fluoromalonyl CoA cannot be synthesized by acetyl CoA carboxylase or chain elongation involving the incorporation of fluoromalonyl-ACP instead of malonyl-ACP does not occur. Clearly both stearoyl-ACP and its w-fluorinated analogue can act as substrates for stearoyl desaturase leading to the formation of w-fluorooleyl-ACP (5), which can be converted to w-fluorooleic acid-containing triglycerides or further processed to other w-fluorosubstituted lipids (Scheme 2).
148
D.B. Harper et al.
Scheme 2. Postulated biosynthetic route to w-fluorofatty acids from fluoroacetate in D. toxi-
carium
A study by Vickery and Vickery [21] of the distribution of fluoroacetate in the D. toxicarium plant at different stages in its life cycle suggests that the fluoroacetate used in w-fluoro fatty acid biosynthesis in the seed is not formed in situ. At no time during maturation of the seed can significant amounts of fluoroacetate be detected in the seed itself. Instead, fluoroacetate appears to be biosynthesised in large quantities in the young leaves and subsequently stored in small leaves adnate to the inflorescence. After fertilization, it is transported to the developing embryo where, as fluoroacetyl CoA, it is incorporated into long chain fatty acids.
Fluorinated Natural Products: Occurrence and Biosynthesis
149
3 Fluorinated Natural Products from Microorganisms 3.1 Nucleocidin
Nucleocidin (9) was the first organofluorine compound to be isolated from a microbial source. This broad-spectrum antibiotic was purified in 1957 from the fermentation broth of an actinomycete, Streptomyces calvus, isolated from an Indian soil sample [60]. Unfortunately, the toxicity of nucleocidin precluded its clinical use [61, 62]. The compound was initially correctly identified as an adenine glycoside esterified with sulfamic acid [63] but, surprisingly, twelve years elapsed after its discovery before it was appreciated that the compound also contained a fluorine atom. The doubling of certain signals in the 1H NMR spectrum attributed in the original work to hindered rotation was in fact due to coupling with fluorine and a revised structure, 4¢-fluoro-5¢-O-sulfamoyladenosine was proposed [64, 65]. This was finally confirmed by total synthesis in 1976 [66]. The source of fluorine in the fermentation broth from which nucleocidin was initially isolated is still not clear; however, the low yield of nucleocidin (2–5 mg L–1) implies that only micromolar levels of fluoride contamination in mineral salts or the tap water used in medium preparation would have supplied the requirement. Conceivably, the availability of fluorine may have limited nucleocidin production and the yield could have been substantially increased if the medium had been supplemented with fluoride. The site of fluorine substitution in nucleocidin at the 4-position of the ribose component renders it highly unlikely that the introduction of fluorine biosynthetically occurs via a fluoroacyl moiety and signifies that the mechanism of fluorination in S. calvus is almost certainly radically different from that in other organofluorine-producing organisms (see Sect. 4.2.7). Thus, it is especially frustrating that repeated attempts in recent years to re-isolate nucleocidin from S. calvus cultures grown from freeze-dried organisms deposited in several culture collections have met with no success, suggesting the total or partial loss of the gene coding for nucleocidin biosynthesis [67, 68]. The loss of secondary metabolic pathways during preservation of organisms can sometimes occur if essential biochemical steps are plasmidencoded. However, it is encouraging to note that a compound with physical and spectral characteristics identical to those described by Shuman et al. [65] for chemically synthesized 4¢-defluoronucleocidin has been isolated in good yield from the fermentation broth of a Streptomyces strain obtained from an Indonesian soil sample [69]. Hence, there is reason to hope that intensive screening of tropical soil samples for Streptomyces spp may eventually lead to rediscovery of strains able to biosynthesise this fascinating organofluorine compound.
150
D.B. Harper et al.
3.2 Fluoroacetate and 4-Fluorothreonine
In 1986, whilst attempting to maximise production of the b-lactam antibiotic thienamycin in fermentation broths of Streptomyces cattleya, Sanada et al. [70] observed that, when grown on a complex medium containing soya bean casein, the actinomycete biosynthesised 4-fluorothreonine (10) and fluoroacetate (1). Subsequent investigations demonstrated that the soya bean casein used contained 0.7% fluorine as inorganic fluoride. When casein was excluded from the medium, fluorometabolite production did not occur but biosynthesis could be restored by supplementation of the medium with 2 mM fluoride. Supplementation of the medium with other halide ions failed to result in the biosynthesis of the corresponding halometabolites. 4-Fluorothreonine isolated from cultures of S. cattleya was an optically active single stereoisomer which when assayed against a range of bacteria behaved as an antimetabolite of L-threonine suggesting an analogous stereochemistry. This has now been confirmed by asymmetric synthesis [71], which has identified a (2S, 3S)-configuration for natural 4-fluorothreonine. In investigations with batch cultures of S. cattleya grown on complex medium containing fluoride, Sanada et al. [70] observed production of 2–3 mM fluoroacetate and 1 mM 4-fluorothreonine in the culture supernatant. Using 19F NMR, Reid et al. [72] monitored the biosynthesis of fluorometabolites by S. cattleya cultured on a chemically defined medium in the presence of 2 mM fluoride. Yields of 1.2 mM fluoroacetate and 0.5 mM 4-fluorothreonine were observed after 28 d incubation. Both compounds were clearly secondary metabolites, their production confined to the idiophase (Fig. 1). Cell suspensions from batch cultures harvested at the growth maximum of 4 d were not capable of fluoride uptake or fluorometabolite biosynthesis, but by 6 d had developed an efficient fluoride uptake system and produced the two fluorometabolites in almost equal proportions. As the harvest age increased, the proportion of fluo-
Fig. 1. Production of fluorometabolites during growth of S. cattleya on defined medium in presence of 2 mM fluoride. , growth; , fluoride; , total fluorometabolites; open squares, fluoroacetate, open circles, 4-fluorothreonine [72]
Fluorinated Natural Products: Occurrence and Biosynthesis
151
roacetate to 4-fluorothreonine formed by cell suspensions rose progressively so that, after 16 d, cells yielded the compounds in a ratio of 3:1 [72]. Sanada et al. [70] reported that resting cells in the absence of fluoride were capable of synthesizing 4-fluorothreonine when supplied with fluoroacetate and conversely could produce fluoroacetate when supplied with 4-fluorothreonine. However, Reid et al. [72] showed that the extent of this interconversion was comparatively minor and not due to the direct conversion of each compound to the other but was instead attributable to cell suspensions possessing a small but significant defluorinating capability (see Sect. 4.2.6). Fluoride ion released is reincorporated into the organic form by the fluorinating system, thus leading to the apparent net transformation of a small proportion of each fluorometabolite into the other. Defluorination and de novo formation of fluorometabolites was also observed when cell suspensions were incubated with other fluorinated biochemical intermediates such as 4-fluoroglutamate, D,L-fluorosuccinate, fluorofumarate and 3fluoroalanine (see Sect. 4.2.6). Both microorganisms elaborating fluorinated secondary metabolites isolated to date have been Streptomyces spp. Moreover, in both cases the discovery of these exotic natural products was made quite fortuitously after extraneous contamination of the culture medium with fluoride. It is possible that the paucity of fluorinated natural products of microbial origin isolated to date may simply reflect the fact that fluoride is not a normal component of standard microbiological culture media. A systematic survey of species of Streptomyces and other actinonomycete genera when cultured in the presence of fluoride could reveal other fluorinated secondary metabolites. A pilot survey conducted by Reid [73] involved screening twenty-five Streptomyces spp. for their ability to take up fluoride from a culture medium containing 2 mM fluoride. Over half the species tested exhibited significant uptake and eleven showed uptake of 10–15% of the fluoride present in the medium. Supernatants from cultures of the latter species were examined by 19F NMR spectroscopy, but no fluorinated metabolites were observed at concentrations above the detection limit of 0.05 mM. However, this does not rule out the possibility of fluorometabolite biosynthesis by these species as culture conditions and media composition are critical in secondary metabolite biosynthesis, and the detection threshold of the 19F NMR technique was relatively high. A survey of a much wider range of actinomycete species under a variety of culture conditions using more sensitive analytical techniques might well meet with success.
4 Biosynthesis of Fluorinated Natural Products Much of the early investigation of the biosynthesis of organofluorine compounds was conducted on plants, since it is only in the last 15 years that the discovery of a microorganism capable of fluoroacetate biosynthesis has provided a more reliable and convenient system in which to study biological fluorination. Several possible mechanisms for C-F bond biosynthesis in plants have been proposed with Peters and co-workers in particular making a valuable contribution in the late 1960s, and Mead and Segal advancing an interesting hypothesis in 1972. The speculations of these and other workers are reviewed in Sect. 4.1. More recent
152
D.B. Harper et al.
work has principally focused on microbial biosynthesis of organofluorine compounds by S. cattleya and progress in this area is summarized in Sect. 4.2. Unfortunately, it is not possible to include in this section a discussion of the very recent report by O’Hagan et al. of the discovery of cell-free fluorinating activity in extracts of S. cattleya [73]. 4.1 Plants 4.1.1 Ethylene or Acetate as Precursors
Peters and Shorthouse [50, 51, 74] demonstrated that homogenates of A. georginae on incubation with ATP, pyruvate and 1 mM fluoride were able to volatilise around 30% of fluoride present. Although, as indicated in Sect. 2.3, a small proportion of the volatile fraction was tentatively identified as fluoroacetone, the remainder was not characterized. Since homogenates also released ethylene, Peters [75] suggested that volatile fluorocarbons derived from ethylene such as vinyl fluoride or fluoroethane could be intermediates in fluoroacetate biosynthesis (Scheme 3). A pathway via ethylene oxide and fluoroethanol is also conceivable as ethylene oxide is reported to be an important product of ethylene metabolism [76, 77]. Unfortunately, the significance placed by Peters on these volatilisation losses is called into question by his observation that many common non-fluoroacetate accumulating species such as Pisum sativum and Poa annua can also volatilise up to 50% of inorganic fluoride present [50]. It therefore becomes necessary to postulate that only fluoroacetate accumulating species possess the enzyme complement required for conversion of volatile fluoro compounds to fluoroacetate. Moreover, the widespread manifestation amongst plants incapable of fluoroacetate biosynthesis of an ability to volatilise fluorine as fluorocarbons should lead to the appearance of such compounds as trace gases in the atmosphere. Despite many years of intense monitoring of atmospheric trace gases, no evidence for biological production of compounds of this type has ever
Scheme 3. Possible routes to fluoroacetate from ethylene [75]
Fluorinated Natural Products: Occurrence and Biosynthesis
153
been obtained. In light of these considerations, the synthesis of fluoroacetate via the pathway suggested by Peters seems implausible. Direct attack of fluoride ion on acetate, a mechanistically unlikely route to fluoroacetate, has also been ruled out by the observation of Eloff and Grobbelaar [78] that leaves of D. cymosum incubated with [2-14C]-acetate displayed no incorporation of label into fluoroacetate. 4.1.2 The Haloperoxidase Reaction
The classical enzymic mechanism for insertion of the halogens, chlorine, bromine and iodine into biological substrates is the haloperoxidase reaction [5]. Reduction of hydrogen peroxide [E° = +1.71 V] enables oxidation of chloride [E° = –1.36 V], bromide [E° = –1.07 V] or iodide [E° = –0.54 V] and active incorporation of halogen into a substrate such as an alkene to form a halonium ion (Scheme 4). Passive incorporation of a nucleophile (normally OH– in dilute aqueous solution) then occurs leading to the formation of a halohydrin. In the presence of high concentrations of halide ion, the nucleophile can be a halide ion leading to the formation of a vicinal dihalide [79, 80]. Vickery et al. [81] speculated that fluoroacetate could be formed in nature by direct activation of fluoride by a fluoroperoxidase. This hypothesis arose from the detection of small quantities of fluoroacetate (in addition to the major product chloroacetate) on incubation of malonate abiotically with sodium hypochlorite and fluoride. However, the redox potential for oxidation of fluoride [E° = –3.06 V] renders it thermodynamically impossible for activation of fluoride to occur by reduction of H2O2. If the experimental observations of Vickery et al. [81] are correct, the most feasible mechanism is probably incorporation of fluoride by displacement of chloride from chloromalonic acid and decarboxylation of fluoromalonate to form fluoroacetate. As a biomimetic model for fluoroacetate biosynthesis, this reaction suffers from the impediment of necessarily involving the simultaneous synthesis of large quantities of chloroacetate. Since Peters et al. [76] and other workers have had no success in detecting chloroacetate in plants accumulating fluoroacetate, this pathway for C-F bond synthesis in nature seems unlikely. An alternative mechanism involving haloperoxidase attack was advanced by Neidelman and Geigert [5] based on their observation that passive incorporation of F– can occur during haloperoxidase attack on an alkene in the presence of oxidisable halide ion, H2O2 and fluoride as indicated in Scheme 4 [79]. These work-
Scheme 4. Incorporation of halogen via the haloperoxidase reaction [5]
154
D.B. Harper et al.
Scheme 5. Postulated biosynthesis of fluoroacetate by haloperoxidase attack on w-unsaturated
fatty acid [5]
ers surmised that an w-unsaturated fatty acid could be the substrate for an iodoperoxidase which in the presence of both fluoride and iodide ions would lead to the formation of a fluoroiodo derivative 11 (Scheme 5). Enzymic reduction and deiodination of this compound would yield an w-fluoro-fatty acid which could undergo b-oxidation to fluoroacetate. This novel idea was clearly prompted by the presence of w-fluorofatty acids in D. toxicarium (see Sect. 2.4). However, no experimental data has been adduced in support of the hypothesis and indeed all the evidence to date indicates that w-fluorofatty acids are synthesized from fluoroacetate not vice versa. Furthermore, intensive GC/MS study of the lipids of D. toxicarium has failed to reveal the presence of iodo or fluoroiodo derivatives [57]. 4.1.3 Fluorophosphate as an Intermediate
The possibility that fluorophosphate (12) could be an intermediate in fluoroacetate biosynthesis was raised by Peters [83]. In studies on extracts from pig heart in 1957, Ochoa et al. [84, 85] showed that pyruvate kinase, which catalyses the conversion of phosphoenolpryuvate to pyruvate in glycolysis, also exhibited a ‘fluorokinase’ activity in that it was capable of mediating CO2-dependent phosphorylation of fluoride by ATP (Scheme 6). Although Peters could not confirm fluorophosphate formation in A. georginae, he detected some metabolism of the compound in homogenates from the plant [86]. However, there is no other evi-
Scheme 6. Formation of fluorophosphate by fluorokinase activity of pyruvate kinase [84, 85]
Fluorinated Natural Products: Occurrence and Biosynthesis
155
dence that fluorophosphate might play a role in C-F bond formation in nature even though binding of fluoride in the form of fluorophosphate would be an elegant solution to the problem of transporting the heavily hydrated fluoride ion within the cell. 4.1.4 Pyridoxal Phosphate Catalysed Incorporation of Fluoride
An attractive mechanism for fluoroacetate biosynthesis was proposed in 1972 by Mead and Segal [87] based on the widespread occurrence in plants of a variety of b-substituted alanines. Biosynthesis of these compounds proceeds through nucleophilic attack at C-3 of a pyridoxal phosphate enamine adduct 13 formed by elimination of the b-substituent from serine or cysteine (Scheme 7). At least seven different b-substituted alanines occur in Acacia spp suggesting that a relatively non-specific synthase allows attack by a number of different nucleophiles. Mead and Segal argued that enzyme specificity could be sufficiently relaxed to permit attack of F– with formation of pyridoxamine phosphate-bound fluoropyruvate (Scheme 8). Hydrolysis and oxidative decarboxylation would yield fluoroacetate. Alternatively, initial decarboxylation and hydrolysis to fluoroacetaldehyde and subsequent oxidation to fluoroacetic acid could occur. The feasibility of fluoropyruvate as an intermediate in fluoroacetate biosynthesis is demonstrated by experiments with D. cymosum tissue cultures, which show that intact cells can mediate rapid oxidative decarboxylation of fluoropyruvate to fluoroacetate [88]. To obtain further evidence in support of their hypothesis Mead and Segal [89] investigated acetone powders of A. georginae which were shown to possess an
Scheme 7. Biosynthesis of b-substituted alanines in plants via a pyridoxal phosphate enamine adduct derived from serine or cysteine [87]
156
D.B. Harper et al.
a,b-eliminase catalysing the pyridoxal-dependent formation of pyruvate from cysteine presumably via the enamine adduct in Scheme 7. This transformation was completely inhibited by the addition of cyanide and an alternative product b-cyanoalanine was formed. As b-cyanoalanine is not reported to be a natural product of A. georginae, the synthesis of the compound in vitro indicates a lack of substrate specificity by the enzyme in this species allowing the enamine intermediate to undergo nucleophilic attack by CN–. Disappointingly, the formation of b-fluoroalanine on addition of fluoride to extracts was not observed nor was fluoropyruvate synthesis or fluoride uptake apparent with cysteine or serine as substrate. However, the investigators did not take into account the relative instability of b-fluoroalanine above pH 7.0 [90]. Since the assays were conducted at pH 8.5, b-fluoroalanine would have rapidly decomposed with elimination of fluoride.Also, any fluoropyruvate synthesised would be readily defluorinated by the pyruvate dehydrogenase complex in cell free extracts of plants [91, 92]. Thus, any organofluorine compounds synthesised in these experiments may have been degraded as soon as they were formed so the negative findings, although discouraging, do not entirely refute the hypothesis. Furthermore, as plant material grown from A. georginae seed derived from plants devoid of fluoroacetate was used in these experiments, genetic variability in the specificity of enzymes involved in bsubstituted alanine biosynthesis within populations of this species could be responsible for the failure to observe C-F bond formation. 4.2 Microorganisms
Since recent attempts to re-isolate nucleocidin from cultures of S. calvus have been unsuccessful (Sect. 3.1), the study of fluorometabolite biosynthesis in microorganisms has been confined to S. cattleya. Efforts have been directed toward
Scheme 8. Possible mechanism for biosynthesis of fluoroacetate via a pyridoxal phosphate
enamine adduct derived from serine or cysteine [87]
Fluorinated Natural Products: Occurrence and Biosynthesis
157
delineating the biosynthetic pathway by incubating resting cells of S. cattleya with isotopically-labelled possible precursors and determining the incorporation of isotope into the fluorometabolites using GC/MS and 19F NMR techniques. 4.2.1 Incorporation of Glycolate
Enzymes that mediate defluorination of fluoroacetate yielding glycolate have been isolated from several organisms [93–96]. The mechanism initially proposed for these haloacetate halidohydrolases involved attack by a thiol group at the active site of the enzyme on the substrate displacing fluoride and forming a thioether, which was subsequently hydrolysed yielding glycolate (Scheme 9). Goldman and Milne [94] suggested that an enzyme reaction of this nature might be reversible and could result in the biosynthesis of fluoroacetate from glycolate and fluoride. To examine this possibility fluoride and glycolate were incubated in H218O with a haloacetate halidohydrolase from a Pseudomonas sp. If the reaction is reversible 18O should be incorporated into glycolate. However, no significant label was detected by mass spectrometry in glycolate re-isolated after incubation with the enzyme indicating that the reaction is irreversible. Subsequently, it has been shown that the mechanism of a haloacetate halidohydrolase isolated from a Moraxella sp. involves nucleophilic attack by an aspartate carboxyl at the active site displacing fluoride followed by base catalysed hydrolysis of the ester intermediate [97]. If this mechanism is universal amongst such enzymes it is unlikely that defluorination is reversible. Glycolate was investigated by both Reid et al. [72] and Tamura et al. [98] as a possible precursor of fluoroacetate in S. cattleya.Whilst Reid et al. [72] observed higher levels of incorporation from [U-14C]-glycolate into fluoroacetate than from other radiolabelled compounds such as [U-14C]-glycerol and [U-14C]glycine, Tamura et al. [98] reported that [2-14C]-glycolate was a comparatively poor precursor. This inconsistency was resolved by Hamilton et al. [99] who demonstrated that it arose as a result of the different experimental conditions employed by the two groups of workers. Reid et al. [72] conducted cell incubations over 96 h, whilst Tamura et al. [98] used a much shorter period of 5–12 h. As, compared with other possible fluorometabolite precursors examined, glycolate is only sluggishly utilized as a carbon source by cells of S. cattleya, the com-
Scheme 9. Possible mechanism of defluorination of fluoroacetate by haloacetate halidohydro-
lase from Pseudomonas sp [94]
158
D.B. Harper et al.
pound remained available for incorporation into fluoroacetate over the entire 96 h incubation period employed by Reid et al. [72] and thereby yielded higher levels of incorporation than other more effective precursors which were rapidly metabolised by cells. Hamilton et al. [99] showed that under conditions in which initial concentrations of all precursors and incubation times are such that compounds are present throughout the incubation period, glycolate is a relatively poor precursor of fluoroacetate as reported by Tamura et al. [98]. 4.2.2 Incorporation of Glycine, Serine and Pyruvate
High levels of double isotope labelling were detected by GC/MS in fluoroacetate and C-3 and C-4 of 4-fluorothreonine when [2-13C]-glycine was incubated with resting cell suspensions of S. cattleya [99, 100]. The 19F {1H} NMR spectrum of the supernatant (Fig. 2) showed signals of the unlabelled compounds at –214.6 ppm (fluoroacetate) and –229.15 ppm (4-fluorothreonine) with a prominent doublet of doublets associated with each, indicating the presence of F-13C-13C, F-12C-13C and F-13C-12C combinations. Interestingly, the labelling patterns of the two fluorometabolites are strikingly similar in terms of magnitude and regiochemistry suggesting that C-1 and C-2 of fluoroacetate and C-3 and C-4 of 4-fluorothreo-
Fig. 2. 19F {H} NMR spectrum of fluoroacetate and 4-fluorothreonine in the supernatant of S. cattleya resting cultures after incubation with [2-13C]-glycine. Insets show expansions of the signals [99]
Fluorinated Natural Products: Occurrence and Biosynthesis
159
nine originate from a common biosynthetic precursor and that there is a single fluorinating enzyme in S. cattleya. Negligible incorporation of label was detected on incubation with [1-13C]-glycine, indicating that there must be recombination of C-2 of glycine prior to incorporation into the fluorometabolites. Such a recombination occurs in the formation of serine from glycine where the enzyme serine hydroxymethyltransferase catalyses the cleavage of glycine and the condensation of C-2 and N5,N10-methylenetetrahydrofolate with another molecule of glycine yielding serine. When [3-13C]-serine was incubated with resting cells, single label was incorporated into C-4 of 4-fluorothreonine and C-2 of fluoroacetate consistent with serine acting as an intermediate in fluorometabolite biosynthesis. Pyruvate, which can be formed from serine by the action of serine dehydratase, was also an effective precursor of the fluorometabolites [99, 100]. Incubation of various 13C-labelled pyruvates with cells demonstrated that C-2 and C3 of this precursor were incorporated into C-1 and C-2 of fluoroacetate and C-3 and C-4 of 4-fluorothreonine. No isotope incorporation from [1-13C]-pyruvate into these positions was observed implying that C-1 is lost during the biosynthesis of fluorometabolites from the precursor. The pathway by which incorporation of glycine, serine and pyruvate into fluorometabolites occurs is summarized in Scheme 10. Incubation of resting cells with [2-13C]-acetate [100] resulted in double labelling of the F-C-C-portion of the fluorometabolites consistent with entry of acetate as acetyl CoA into the citric acid cycle and subsequent processing through the cycle to generate oxaloacetate. This experiment established that pyruvate is not converted to acetyl CoA before incorporation into fluorometabolites. The labelling patterns obtained with the various 13C-enriched pyruvates is compatible
Scheme 10. Incorporation of [2-13C]-glycine into fluorometabolites
160
D.B. Harper et al.
with the metabolism of this precursor via the glycolytic pathway and subsequent experiments with other possible precursors have confirmed this as the likely route to the fluorometabolites (Section 4.2.3). 4.2.3 Incorporation of Glycerol and Glucose
Tamura et al. [98] observed high incorporation from [2-13C]-glycerol into C-1 of fluoroacetate and also incorporation of 14C from C-2 and C-3 of radiolabelled bhydroxypyruvate into C-1 and C-2 of fluoroacetate. These workers suggested that fluoroacetate was synthesised via b-hydroxypyruvate and b-fluoropyruvate although no evidence for the presence of the latter compound was obtained. To explore further the biosynthetic pathway from glycerol, Neischalk et al. [100] prepared enantiomerically labelled (2R)-[1-2H2]- and (2S)-[1-2H2]-glycerols and incubated them with cell suspensions of S. cattleya. Figure 3 shows the 19F {1H} NMR spectrum of the culture supernatant from the experiment conducted with (2R)-[1-2H2]-glycerol. The signals from both unlabelled fluoroacetate and 4-fluorothreonine have associated with them signals shifted upfield by 0.6 ppm and 1.2 ppm resulting from single and double deuterium labelling of the fluoromethyl group, respectively. No deuterium incorporation into fluorometabolites from (2S)-[1-2H2]-glycerol was detected. Thus, only the pro-R carbon of glycerol is eventually fluorinated whilst the pro-S hydroxymethyl group must be cleaved during fluorometabolite biosynthesis. Significantly, it is the pro-R carbon of glycerol that is phosphorylated by glycerol kinase to form sn-glycerol-3phosphate, the first step in glycerol catabolism by the glycolytic pathway. Thus, the carbon atom activated by phosphorylation is also the one that is fluorinated downstream. The retention of both deuterium atoms from (2R)-[1-2H2]-glycerol in the fluoromethyl groups implies that there is no oxidation at this centre and that regardless of the mechanism of fluorination there is formal replacement of phosphate with fluorine. In light of these findings, Nieschalk et al. [101] proposed that fluorination probably involved nucleophilic substitution by fluoride at the phosphorylated (or enzymically activated) carbon atom of sn-glycerol-3phosphate or a metabolite of the latter on the glycolytic pathway to pyruvate (Scheme 11).
Scheme 11. Incorporation of (2R)-[1-2H2]-glycerol into fluorometabolites via [3-2H2]-glycerol-
3-phosphate
Fluorinated Natural Products: Occurrence and Biosynthesis
161
19F
{H} NMR spectrum of fluoroacetate and 4-fluorothreonine in the supernatant of S. cattleya resting cultures after incubation with (2R)-[1-2H2]-glycerol. Insets show expansions of the signals [101] Fig. 3.
Studies with various isotopically-labelled glucoses provided additional evidence for the involvement of glycolytic intermediates in fluorometabolite biosynthesis [102]. Isotopic label was incorporated solely into C-2 of fluoroacetate and C-4 of 4-fluorothreonine from both [1-13C]– and [6-13C]-glucose consistent with metabolism of these compounds through the glycolytic pathway to a C-3 intermediate. However, incorporation into the fluorometabolites from [6-13C]-glucose (30%) was twice that from [1-13C]-glucose (ca. 15%). This difference is best explained by metabolism of a proportion of glucose through the pentose phosphate pathway where label from [1-13C]-glucose is lost as 13CO2 in the conversion of 6-phosphogluconate to ribulose-5-phosphate (Scheme 12); label from [6-13C]glucose is not lost in this way and can be incorporated via glyceraldehyde-3-phosphate. In an attempt to ascertain the identity of the intermediate in the glycolytic pathway metabolically most closely related to the carbon substrate for fluorination, Murphy [102] investigated the effect of various unlabelled intermediates in the pathway between glycerol and pyruvate on the incorporation of [2-13C]-glycerol and [2-13C]-pyruvate into fluorometabolites. The greatest dilution of label was recorded on the presence of D-glyceraldehyde and dihydroxyacetone phosphate and their phosphorylated derivatives. It was tentatively concluded that one of these triose phosphates is closely related to the substrate for fluorination. However, neither compound yielded an organofluorine product when incubated with fluoride and cell-free extracts of S. cattleya.
162
D.B. Harper et al.
Scheme 12. Fate of isotopic label in [1-13C]-and [6-13C]-glucose during metabolism to fluo-
rometabolites
4.2.4 Stereochemistry of Fluorination
Findings to date suggest that the formation of the C-F bond in S. cattleya involves formal replacement of a phosphate group by a fluorine atom. To gain a mechanistic insight into this process, the stereochemistry of the conversion of a phosphorylated glycolytic intermediate to fluoroacetate in S. cattleya has been examined [103].When [2H4]-succinate was incubated with resting cells of S. cattleya the fluoroacetate biosynthesised was labelled with a single deuterium in the fluoromethyl group (19% incorporation). Since enzyme reactions are usually stereospecific, the chiral [2-2H1]-fluoroacetate so formed is likely to be enantiomerically enriched. Succinate is known to be metabolised by the citric acid cycle through fumarate, malate and oxaloacetate which can enter the glycolytic pathway via phosphoenolpyruvate.As the stereochemistry of these enzymic reactions is well established, it can readily be deduced that the residual deuterium atom from [2H4]-succinate will occupy the (R)-configuration in 3-[3-2H1]-phosphoglycerate (Scheme 13). If the enantiomeric composition of [2-2H1]-fluoroacetate formed after incubation of [2H4]-succinate with S. cattleya can be determined, the
Fluorinated Natural Products: Occurrence and Biosynthesis
163
stereochemical relationship (inversion or retention of configuration) between the phosphorylated intermediates in glycolysis and fluoroacetate should be apparent. Accordingly, fluoroacetate was isolated from such an experiment, derivatised and assayed for chirality by 2H-NMR in a chiral liquid crystalline medium, a powerful new technique for enantiomeric discrimination [104, 105]. The compound was shown to have a significant enantiomeric excess (25 % ee) of the (R)-enantiomer. The comparatively low enantiomeric excess was probably attributable to some racemisation occurring as succinate is metabolised through oxaloacetate, a compound which readily enolises at the stereogenic centre bearing the deuterium atom. The predominance of the (R)-enantiomer in the recovered [2-2H1]fluoroacetate suggests that there is overall retention of configuration by the phosphate-bearing carbon atom of 3-phosphoglycerate on conversion to the fluoromethyl group. Clearly, this study does not support a mechanism involving direct displacement of a phosphate group by fluoride ion (SN2 reaction) as this would result in an inversion of configuration. Perhaps biological fluorination involves two stereochemical inversions in the course of conversion of phosphorylated intermediate to fluoroacetate.
Scheme 13. Metabolic route from [2H4]-succinate to chiral fluoroacetate
4.2.5 Fluoroacetaldehyde: the Common Precursor of the Fluorometabolites in S. cattleya
Studies with isotopically-labelled precursors have established that C-1 and C-2 of fluoroacetate has the same biosynthetic origin as C-3 and C-4 of 4-fluorothreonine in S. cattleya (Sect. 4.2.2). Since neither fluorometabolite is derived by metabolism of the other [72, 106] they must be derived from a common fluorinated precursor. Recently, this intermediate was identified as fluoroacetaldehyde [107]. When [1-2H1]-fluoroacetaldehyde was incubated with cell suspensions of S. cattleya there was substantial deuterium incorporation into C-3 of
164
D.B. Harper et al.
Scheme 14. Biotransformation of fluoroacetaldehyde to fluoroacetate and 4-fluorothreonine catalysed by cell suspensions of S. cattleya
4-fluorothreonine. Furthermore, cell-free extracts of the bacterium readily oxidised fluoroacetaldehyde in the presence of NAD+ (Scheme 14). Interestingly Sanada et al. [70], the group who first reported fluorometabolite production by S. cattleya in 1986, speculated at that time that fluoroacetaldehyde was the precursor of fluoroacetate and 4-fluorothreonine. The aldehyde dehydrogenase responsible for oxidation of fluoroacetaldehyde is the first enzyme directly involved in fluorometabolite biosynthesis in S. cattleya to be isolated and characterized [108]. In batch culture the enzyme is induced in the late exponential phase and enzyme synthesis continues into the stationary phase consistent with a role in secondary metabolism. Fluoroacetaldehyde dehydrogenase appears similar to other aldehyde dehydrogenases in many respects; it is a tetramer (mol wt 200,000), has a pH optimum of 9 and is sensitive to thiol-blocking agents. Both fluoroacetaldehyde and glycolaldehyde are good substrates for the enzyme, but acetaldehyde, propionaldehyde and glyceraldehyde are oxidised relatively slowly. Since enzymes involved in secondary metabolism usually originate by duplication and mutation of enzymes of primary metabolism it is conceivable that fluoroacetaldehyde dehydrogenase evolved from an aldehyde dehydrogenase involved in primary metabolism with glycolaldehyde as its natural substrate. If the genes coding for fluorometabolite biosynthesis are situated in a single cluster, it may be possible to locate and clone the cluster using the gene sequence of fluoroacetaldehyde dehydrogenase as a probe. The mechanism of 4-fluorothreonine biosynthesis from fluoroacetaldehyde is not clear. Sanada et al. [70] speculated that the amino acid may arise by condensation of fluoroacetaldehyde with glycine, but experiments using various isotopically-labelled glycines as precursors have demonstrated that glycine is not directly incorporated into C-1 and C-2 of 4-fluorothreonine [99]. Work currently in progress aims to elucidate the origin of these carbon atoms. 4.2.6 Defluorination in S. cattleya
Since fluoroacetate is toxic due to lethal synthesis of fluorocitrate in vivo (Sect. 2.2), S. cattleya must possess some unusual metabolic adaptations to resist the effects of the compound. Specific enzymes that convert fluoroacetate to glycolate have been isolated from some bacteria (Sect. 4.2.1), but there is no evidence that S. cattleya produces such an enzyme. Fluorocitrate does not appear to inhibit the aconitase of S. cattleya; indeed fluoride was released when cell suspensions of S. cattleya were incubated with several organofluorine substrates including
Fluorinated Natural Products: Occurrence and Biosynthesis
165
4-fluoroglutamate, DL-fluorosuccinate, fluorofumarate and fluorocitrate [72]. Such defluorination can possibly be explained if S. cattleya cells are capable of metabolising such compounds via the citric acid cycle to fluorofumarate. The action of fumarase on this compound is known to yield fluoromalate, which undergoes spontaneous non-enzymic elimination of fluoride to form oxaloacetate [109]. Murphy [102] investigated defluorination of 4-fluoroglutamate in cell-free extracts of S. cattleya and concluded that release of fluoride was not associated with the deamination of 4-fluoroglutamate by glutamate oxidase. Deaminating and defluorinating activities in cell-free extracts were separated by anion exchange chromatography confirming that the organism possesses an unusual and specific defluorinating enzyme. 4-Fluoroglutamate may arise in S. cattleya by metabolism of fluoroacetate as fluoroacetyl CoA through the citric cycle to 4-fluoroa-ketoglutarate; one fate of this may be conversion by glutamate dehydrogenase to 4-fluoroglutamate (Scheme 15). This fluoro-amino acid could potentially act as an analogue of glutamate and inhibit protein synthesis or become incorporated into proteins and interfere in their functioning. In such circumstances, an efficient mechanism for degrading 4-fluoroglutamate would be essential.
Scheme 15. Possible pathway for formation of 4-fluoroglutamate from fluoroacetate in S. catt-
leya
4.2.7 Enzymic Synthesis of Glycosyl Fluorides
The enzymic formation of C-F bonds in active site mutants of transglucosidase enzymes from Agrobacterium sp and Cellulomonas fimi have recently been reported [110, 111]. In each case the original glutamate residue of the native enzyme was replaced using site-directed mutagenesis with glycine, alanine or serine. Removal of the catalytic nucleophilic carboxylate groups of the glutamate
166
D.B. Harper et al.
renders the normal double displacement mechanism of the enzyme inoperative and glycoside-cleaving activity is lost. However, when these mutant enzymes were assayed with a 2,4-dinitrophenyl-b-glycoside substrate in the presence of high concentrations of fluoride (2 M), substantial glycoside bond-cleaving activity was restored (Scheme 16). Fluoride apparently acts in the place of the missing catalytic nucleophile and transglycosylation occurs via transitory formation of a glycosyl fluoride. Wild type enzymes did not catalyse the formation of glycosyl fluorides.Withers and co-workers [110, 111] argued that the amino acid replacements in the mutant enzyme allowed space for F– to approach the reactive anomeric centre of the sugar substrate and, in the case of the serine mutant, assisted presentation of the fluorine nucleophile to the reaction site by hydrogen bonding between F– and the hydroxyl hydrogen. These mutant enzymes will also catalyse nucleophilic halogenation with Cl– and Br–, but the reactivity is in the order F– >Cl– >Br– , that is the opposite to that predicted by halide nucleophilicity in water. However, this order is observed in organic solvents so it is quite likely that desolvation of F– occurs at the active site. The investigators draw attention to the fact that nucleocidin (Sect. 3.1) is also a glycosyl fluoride.Although C-F bond formation in their study was mediated by a genetically engineered protein it is conceivable that nucleocidin in S. calvus and possibly even fluoroacetate in S. cattleya are biosynthesised by an analogous mechanism involving desolvation of fluoride anion and hydrogen bond formation.
Scheme 16. Mechanism of fluoride-catalysed transglylcosylation by mutant enzyme from Cel-
lulomonas fimi involving transient glycosyl fluoride formation. DNP – 2,4-dinitrophenyl, R¢OH – glycoside acceptor [111].
Fluorinated Natural Products: Occurrence and Biosynthesis
167
5 References 1. Paul EA, Huang PM (1980) In: Hutzinger O (ed) Handbook of environmental chemistry, vol 1, part A. Springer Verlag, Berlin, p 69 2. Bowen HJM (1966) Trace elements in biochemistry. Academic Press, London 3. Smart BE (1986) In: Liebman JL, Greenberg A (eds) Molecular structure and energetics, vol 3. VCH, Deerfield Park, Florida, p 141 4. Sharpe AG (1967) In: Gutmann V (ed) Halogen chemistry, vol 1. Academic Press, p 12 5. Neidleman SL, Geigert J (1986) Biohalogenation: Principles, basic roles and applications. Ellis Horwood Ltd, Chichester 6. Gregson RP, Baldo BA, Thomas PG, Quinn RJ, Bergquist PR, Stephens JF, Horne AR (1979) Science 206:1108 7. Zimmerman PW, Hitchcock AE, Gwirtsman J (1957) Contr Boyce Thompson Inst Pl Res 19:49 8. Venkateswarlu P, Armstrong WD, Singer L (1965) Plant Physiol 40:255 9. Cooke JA, Johnson MS, Davison AW, Bradshaw AD (1976) Environ Pollut 11:9 10. Andrews SM, Cooke JA, Johnson MS (1989) Environ Pollut 60:165 11. Meyer M, O’Hagan D (1992) Chem Br 28 : 785 12. Matuura S, Kokubu N, Watanabe S, Samesima Y (1955) Mem Fac Sci Kyushu Univ Ser C 2:75 13. Marais JSC (1943) Onderstepoort J Vet Sci Anim Ind 18:203 14. Marais JSC (1944) Onderstepoort J Vet Sci Anim Ind 20:67 15. Pattison FLM (1959) Toxic aliphatic fluorine compounds. Elsevier, Amsterdam 16. Von Sydow B (1969) Flora Jenu A 160:196 17. Tannock J (1975) Rhod J Agric Res 13:67 18. O’Hagan D, Perry R, Lock JM, Meyer JJM, Dasaradhi L, Hamilton JTG, Harper DB (1993) Phytochemistry 33:1043 19. Vickery B, Vickery ML, Ashu JT (1973) Phytochemistry 12: 145 20. Hall RJ (1972) New Phytol 71:855 21. Vickery B, Vickery ML (1972) Phytochemistry 11 : 1905 22. Vickery B, Vickery ML (1973) Vet Bull 43 : 537 23. Kamgue RT, Sylla O, Pousset JL, Laurens A, Brunet JC, Sere A (1979) Plantes Med Phytotherapie 13:252 24. Twigg LE, King DR (1991) Oikos 61:412 25. Twigg LE, King DR, Bowen LH, Wright GR, Eason CT (1996) Aus J Bot 44 : 411 26. Aplin TEH (1971) Poison plants of Western Australia: The toxic species of Gastrolobium and Oxylobium, West Aust Dept Bull No 3772, p 1 27. Crisp MD, Weston PH (1994) Cladistics and legume systematics with an analysis of the Bossiaeae, Brongniartieae and Mirbelieae. In: Stirton CH (ed) Advances in legume systematics Part 3. Royal Botanic Gardens, Kew, p 65 28. McEwan T (1964) Nature 202:827 29. Twigg LE, Wright GR, Potts MD (1999) Aus J Bot 47 : 877 30. Oelrichs PB, McEwan T (1961) Nature 190:808 31. Oelrichs PB, McEwan T (1962) Qd J Agric Sci 19:1 32. Murray LR, Woolley DR (1968) Aus J Soil Res 6:203 33. De Oliveira MM (1963) Experientia 19 :586 34. Vartiainen T, Gynther J (1984) Fd Chem Toxic 22 : 307 35. Peters RA (1972) In: Elliott K, Birch J (eds) Carbon-fluorine compounds: Chemistry, biochemistry and biological activities. Ciba Foundation Symposium, Associated Scientific Publishers, Amsterdam, p 118 36. Lovelace CJ, Miller GW, Welkie GW (1968) Atmos Environ 2 : 187 37. Cheng JY, Yu MH, Miller GW, Welkie GW (1968) Environ Sci Technol 2 : 367 38. Yu MH, Miller GW (1970) Environ Sci Technol 4:492
168
D.B. Harper et al.
39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49.
Peters RA, Shorthouse M (1972) Phytochemistry 11:1337 Vartiainen T, Kauranen P (1984) Anal Chim Acta 157:91 Preuss PW, Colavito L, Weinstein LH (1970) Experientia 26:1059 Preuss PW, Birkhahn R, Bergmann ED (1970) Isr J Bot 19:609 Bennett LW, Miller GW, Yu MH, Lynn RI (1983) Fluoride 16:111 Grobbelaar N, Meyer JJM (1989) J Plant Physiol 135:550 Meyer JMM, O’Hagan D, unpublished results Peters RA, Wakelin RW, Buffa P, Thomas LC (1953) Proc Roy Soc B 40:497 Kun E, Dummel RJ (1969) Methods Enzymol 13:632 O’Hagan D, Rzepa HS (1994) J Chem Soc Chem Commun 2029 Lauble H, Kennedy MC, Emptage MH, Beinert H, Stout CD (1996) Proc Natl Acad Sci USA 13:699 Peters RA, Shorthouse M (1967) Nature 216:80 Peters RA, Shorthouse M (1971) Nature 231:123 Miura K, Otsuka S, Honda K (1956) Bull Agric Chem Soc Jpn 20:219 Renner W (1904) Brit Med J 1:1314 Peters RA, Hall RJ (1959) Biochem Pharmacol 2:25 Peters RA, Hall PJ, Ward PFV, Sheppard N (1960) Biochem J 77:17 Ward PFV, Hall RJ, Peters RA (1964) Nature 201: 611 Hamilton JTG, Harper DB (1997) Phytochemistry 44:1129 Christie WW, Hamilton JTG, Harper DB (1998) Chem Phys Lipids 97 : 41 Harper DB, Hamilton JTG, O’Hagan D (1990) Tetrahedron Lett 31:1776 Thomas SO, Singleton VL, Lowery JA, Sharpe RW, Pruess LM, Porter JN, Mowat JH, and Bohonos N (1957) Antibiotics Ann 1956–7:716 Hewitt RI, Gumble AR, Tayler LH, Wallace WS (1957) Antibiotics Ann 1956–57:722 Thomas SO, Lowery JA, Singleton VLR (1959) United States Patent, Office No. 2, 914525, November 24, 1959; Chem Abstr 54 : 5024 Waller CW, Patrick JB, Fulmor W, Meyer WE (1957) J Am Chem Soc 79:1011 Morton GO, Lancaster JE, Van Lear GE, Fulmor W, Meyer WE (1969) J Am Chem Soc 91:1535 Shuman DA, Robins RK, Robins MJ (1969) J Am Chem Soc 91:3391 Jenkins ID, Verheyden JPH, Moffatt JG (1976) J Am Chem Soc 98 : 3346 Maguire AR, Meng W-d, Roberts SM, Willetts AJ (1993) J Chem Soc Perkin Trans 1:1795 Harper DB, unpublished results Wall W, personal communication (available on request to DBH) Sanada M, Miyano T, Iwadare S, Williamson JM, Arison BH, Smith JL, Douglas AW, Liesch JM, Inamine E (1986) J Antibiotics 39:259 Amin MR, Harper DB, Moloney JM, Murphy CD, Howard JAK, O’Hagan D (1997) Chem Commun 1997:1471 Reid KA, Hamilton JTG, Bowden RD, O’Hagan D, Dasaradhi L, Amin MR, Harper DB (1995) Microbiology 141:1385 Reid KA (1994) PhD thesis, The Queen’s University of Belfast; O’Hagan D, Schaffrath C, Cobb SL, Hamilton JTG, Murphy CD (2002) Nature 416:279 Peters RA, Shorthouse M (1967) Life Sci 6:1505 Peters RA (1973) Fluoride 6:189 Jerie PH, Hall MA (1978) Proc R Soc Lond B 200:87 Dodds JH, Musa SK, Jerie PH, Hall MA (1979) Plant Sci Lett 17:109 Eloff J. N. Grobbelaar N (1972) Joernaal van die Suid–Afrikaanse Chemiese Institute 25:109 Neidleman SL, Geigert J (1983) Trends Biotechnol 1 :21 Neidleman SL, Geigert J (1985) Ann Proc Phytochem Soc Eur 26:267 Vickery B, Vickery ML, Kabeira F (1979) Experientia 35:299 Peters RA, Murray LR, Shorthouse M (1965) Biochem J 95:724 Peters RA (1967) Rec Chem Progress 28 : 197 Flavin M, Castro-Mendoza H, Ochoa S (1957) J Biol Chem 229:981
50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84.
Fluorinated Natural Products: Occurrence and Biosynthesis
85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111.
169
Tietz A, Ochoa S (1958) Arch Biochem Biophys 78:477 Peters RA, Shorthouse M (1964) Biochem J 93:20P Mead RJ, Segal W (1972) Aust J Biol Sci 25:327 Meyer JJM, O’Hagan D (1992) Phytochemistry 31:499 Mead RJ, Segal W (1973) Phytochemistry 12:1977 Kun E (1972) In: Elliot K, Birch J (eds) Carbon-fluorine compounds: Chemistry, biochemistry and biological activities. Ciba Foundation Symposium, Associated Scientific Publishers, Amsterdam, p 158 Leung LS, Frey PA (1978) Biochem Biophys Acta 81:274 Gish G, Smyth T, Kluger R(1988) J Am Chem Soc 110:6230 Goldman P (1965) J Biol Chem 240:3434 Goldman P, Milne GWA (1966) J Biol Chem 241:5557 Walker JRL, Lien BC (1981) Soil Biol Biochem 13 : 231 Kawasaki H, Miyashi K, Tonomura, K (1981) Agri Biol Chem 45:543 Lui J-Q, Kurihara JT, Ichiyama S, Miyagi M, Tsunasawa S, Kawasaki H, Soda K, Esaki N (1998) J Biol Chem 273:30897 Tamura T, Wada M, Esaki N, Soda K (1995) J Bacteriol 177:2265 Hamilton JTG, Amin MR, O’Hagan D, Harper DB (1997) Chem Commun 1471 Hamilton JTG, Murphy CD, Amin MR, O’Hagan D, Harper DB (1998) J Chem Soc Perkin Trans 1 759 Neischalk J, Hamilton JTG, Murphy CD, Harper DB, O’Hagan D (1997) Chem Commun 799 Murphy CD (1998) PhD Thesis, The Queen’s University of Belfast O’Hagan D, Goss RM, Courtieu J, Meddour A, unpublished results Meddour A, Canlet C, Blanco L, Courtieu J (1999) Angew Chem Int Ed 38:2391 Lesot P, Merlet D, Loewenstein A, Courtieu J (1998) Tetrahedron Asymmetry 9:1871 Reid KA, Bowden RD, Harper DB (1995) In: Grimvall A, deLeer EWB (eds) Naturally produced organohalogens. Kluwer Academic Publishers, Dordrecht, p 269 Moss SJ, Murphy CD, Hamilton JTG, McRoberts WC, O’Hagan D, Schaffrath C, Harper DB (2000) Chem Commun 2281 Murphy CD, Moss SJ, O’Hagan D (2001) Appl Environ Microbiol (in press) Harper DB, Blakley ER (1971) Can J Microbiol 17:645 Nashiru O, Zechel DL, Stoll D, Mohammadzadeh T,Warren RAJ,Withers SG (2001) Angew Chem Int Ed 40:417 Zechel DL, Reid SP, Nashiru O, Mayer C, Stoll D, Jakeman DL, Warren RAJ, Withers SG (2001) J Am Chem Soc 123:4350
The Handbook of Environmental Chemistry Vol. 3, Part P (2003): 171–199 DOI 10.1007/b 10457
Enzymology and Molecular Genetics of Biological Halogenation Karl-Heinz van Pée, Susanne Zehner Institut für Biochemie, TU Dresden, 01062 Dresden, Germany E-mail:
[email protected]
Heme-containing haloperoxidases were the first halogenating enzymes detected. Later vanadium-containing haloperoxidases were isolated from eukaryotic sources and perhydrolases could be isolated from bacteria. However, elucidation of the three-dimensional structures and reaction mechanism of haloperoxidases and perhydrolases showed that they all produce more or less directly hypohalous acids leading to halogenation reactions without substrate specificity and regioselectivity. Since these properties have to be expected from halogenating enzymes involved in the biosyntheses of more complex halometabolites, it became clear that other halogenating enzymes must exist. Cloning and characterization of the biosynthetic gene clusters of many halometabolites has revealed that a new type of halogenating enzyme, the FADH2dependent halogenases, seem to be a major player in biological halogenation. However, there is evidence that other types of halogenating enzymes and mechanism, not yet detected, must exist. Keywords. Halogenase, Haloperoxidase, Perhydrolase, Chlorination, Bromination
1
The Detection of Halogenating Enzymes . . . . . . . . . . . . . . 172
2
Biochemistry of Halogenating Enzymes
2.1 2.1.1 2.1.2 2.1.3 2.1.4 2.1.5
Haloperoxidases . . . . . . . . . . . . Heme-Containing Haloperoxidases . . Vanadium-Containing Haloperoxidases Perhydrolases . . . . . . . . . . . . . . FADH2-Dependent Halogenases . . . . Methyl Halide Transferases . . . . . . .
3
Molecular Genetics and Reaction Mechanisms . . . . . . . . . . . 176
3.1 3.2 3.3 3.4 3.5
Heme-Containing Haloperoxidases . . Vanadium-Containing Haloperoxidases Perhydrolases . . . . . . . . . . . . . . FADH2-Dependent Halogenases . . . . Fluorinating Enzymes . . . . . . . . .
4
Functions of Halogenating Enzymes
4.1 4.2 4.3
Haloperoxidases . . . . . . . . . . . . . . . . . . . . . . . . . . . 186 Perhydrolases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188 FADH2-Dependent Halogenases . . . . . . . . . . . . . . . . . . . 188
. . . . . . . . . . . . . . 172 . . . . . .
. . . . .
. . . . . .
. . . . .
. . . . . .
. . . . .
. . . . . .
. . . . .
. . . . . .
. . . . .
. . . . . .
. . . . .
. . . . . .
. . . . .
. . . . . .
. . . . .
. . . . . .
. . . . .
. . . . . .
. . . . .
. . . . . .
. . . . .
. . . . . .
. . . . .
. . . . . .
. . . . .
. . . . . .
. . . . .
. . . . . .
. . . . .
172 172 174 175 175 175
176 178 179 180 186
. . . . . . . . . . . . . . . . 186
© Springer-Verlag Berlin Heidelberg 2003
172
K.-H. van Pée and S. Zehner
4.3.1 4.3.1.1 4.3.1.2 4.3.1.3 4.3.1.4 4.3.1.5 4.3.1.6 4.4
Biosynthesis of Halometabolites . . . . . . . . . . 7-Chlorotetracycline Biosynthesis . . . . . . . . . Pyrrolnitrin Biosynthesis . . . . . . . . . . . . . Pyoluteorin Biosynthesis . . . . . . . . . . . . . . Biosynthesis of the Vancomycin Group Antibiotics Chloramphenicol Biosynthesis . . . . . . . . . . . Barbamide Biosynthesis . . . . . . . . . . . . . . Methyl Halide Transferases . . . . . . . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
. . . . . . . .
188 188 190 190 191 192 194 195
5
Applications of Halogenating Enzymes and their Genes . . . . . . 195
6
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
1 The Detection of Halogenating Enzymes Although the existence of halometabolites has been known since 1896 [1], enzymatic halogenating activity was detected as late as 1959 [2]. During the investigation of caldariomycin biosynthesis, an antibiotic produced by the fungus Caldariomyces fumago, the enzymatic chlorination of b-ketoadipic acid to d-chlorolevulinic acid (Fig. 1) was detected in cell-free extracts of Caldariomyces fumago [2]. Fractionation and partial purification of the enzyme revealed that it required hydrogen peroxide and chloride ions for catalysis of carbon-halogen bond formation and was thus named chloroperoxidase [3]. In addition to chloroperoxidases that catalyse the iodination, bromination and chlorination of organic substrates, bromoperoxidases, which catalyse iodination and bromination reactions, and iodoperoxidases, catalysing only the iodination of organic compounds, belong to the haloperoxidases [4, 5]. However, these haloperoxidases cannot be involved in the formation of fluorometabolites, since oxidation of fluoride ions by hydrogen peroxide is impossible [4].
2 Biochemistry of Halogenating Enzymes 2.1 Haloperoxidases 2.1.1 Heme-Containing Haloperoxidases
Early purification procedures for chloroperoxidase from Caldariomyces fumago used mycelium from which a soluble extract was obtained.When this extract was assayed for chlorinating activity using radioactive chloride, neither hydrogen peroxide nor any other co-substrates were required [6]. However, fractionation of the crude extract revealed that it contained two enzymes that were required for chlorination of b-ketoadipic acid, an intermediate of caldariomycin biosynthesis [6]. One enzyme was identified as a glucose oxidase producing hydrogen
Enzymology and Molecular Genetics of Biological Halogenation
173
Fig. 1. Chlorination of b-ketoadipic acid to d-chlorolevulinic acid by chloroperoxidase from the
fungus Caldariomyces fumago [2]
peroxide that was required by the second enzyme, the chloroperoxidase. Later it was realized that chloroperoxidase is an extracelluar enzyme and thus the enzyme could be purified from the filtrate obtained after removal of the mycelium. Purification and crystallisation of the enzyme showed that it was a glycosylated monomeric heme enzyme with protoporphyrin IX as the prosthetic group with a molecular weight of 42,000 [7]. The group of L. P. Hager developed a spectrophotometric assay for the detection and quantification of haloperoxidase activity. This assay was based on the chlorination and bromination of the synthetic compound monochlorodimedone (Fig. 2). The enol form of monochlorodimedone has a much higher extinction coefficient at 290 nm than the keto form, and since the product of the halogenation reaction, dihalodimedone, only exists as the keto form, halogenation of monochlorodimedone can be determined by a decrease in the absorbance maximum at 290 or 278 nm where the extinction coefficient is independent of the pH [8]. Since monochlorodimedone is not a natural compound, this assay has the severe draw back that it selects for halogenating enzymes with either very broad substrate specificities or no substrate specificity at all, and the additional use of hydrogen peroxide selects for peroxidase-type enzymes. This lack of specificity was thought to be compensated by the use of bromide instead of chloride as the halide ion, since it is less difficult to oxidize than chloride [9]. Iodination activity can easily be detected by the oxidation of iodide ions to iodine or the enzymatic formation of monoiodo- and diiodotyrosines [8]. Using these assays, the monochlorodimedone and the iodination assays, a large number of bromo- and a few iodo- and chloroperoxidases could be isolated and characterized from eukaryotic organisms [5] and the
Fig. 2. Chemical structure of 2-chloro-1,3-cyclopentanedione, a late intermediate of caldari-
omycin biosynthesis and enzymatic halogenation of the synthetic compound monochlorodimedone [8]
174
K.-H. van Pée and S. Zehner
haloperoxidase activities of well-known peroxidases, such as horseradish peroxidase (iodoperoxidase), lactoperoxidase (bromoperoxidase) and myeloperoxidase (chloroperoxidase) could be established [4, 5].As late as 1984, the first bromoperoxidase from a prokaryotic organism was detected and subsequently isolated [10, 11].All haloperoxidases detected until 1983 were hemoproteins containing non-covalently bound ferriprotoporphyrin IX as the prosthetic group, which seemed to be covalently attached through ester linkages in lactoperoxidase [12]. An exception is myeloperoxidase that contains a covalently bound derivative of ferriprotoporphyrin IX [13]. Interestingly, a halogenating enzyme that showed chloroperoxidase activity was isolated from the marine worm Notomastus lobatus [14]. This enzyme is composed of two dissociable protein moieties, a heme protein and a flavoprotein. Neither the flavoprotein nor the heme protein alone has chloroperoxidase activity. This suggests that this enzyme might not be a heme-containing chloroperoxidase. However, since the genes of this enzyme system have not been cloned and sequenced, it is not yet clear to which type of halogenase this enzyme actually belongs. It is possible that this enzyme belongs to a novel type of halogenases, the FADH2-dependent halogenases, an assumption that is supported by the fact that this enzyme, unlike haloperoxidases, does not chlorinate monochlorodimedone, but accepts a number of different phenolic compounds as substrates. The molecular weights of heme-containing haloperoxidases range from around 40,000 for horseradish peroxidase [7] to 160,000 for bromoperoxidase from Pseudomonas aureofaciens [5]. The subunit composition varies from monomeric enzymes such as horseradish peroxidase to heterotrimers in the case of myeloperoxidase [5], and the specific activities with monochlorodimedone as the substrate vary over an enormous range from about 0.8 units mg–1 for bromoperoxidase from Pseudomonas aureofaciens to about 56,000 units mg–1 for chloroperoxidase from Penicillus capitatus [5]. 2.1.2 Vanadium-Containing Haloperoxidases
In 1983, Vilter [15] reported the isolation of a bromoperoxidase from the brown alga Ascophyllum nodosum, which did not show an absorption spectrum of hemoproteins. Detailed investigations revealed that this haloperoxidase did not contain a heme group, but required vanadium(V) for brominating activity [16, 17]. In the following years, similar vanadium-containing haloperoxidases were isolated from other algae [18–21], lichen [22] and fungi [23]. In contrast to hemecontaining bromoperoxidases that have also been isolated from bacteria [24], vanadium-dependent haloperoxidases have so far not been detected in bacteria. The molecular weights of vanadium haloperoxidases were found to be between 67,000 for monomeric enzymes such as chloroperoxidase from Curvularia inaequalis [25] and 790,000 for the Corallina pilulifera bromoperoxidase which consists of 12 subunits of identical molecular weights [26]. Specific activities for the oxidation of bromide are between 127 units mg–1 for bromoperoxidase from Ascophyllum nodosum and 1730 units mg–1 for bromoperoxidase from Macrocystis pyrifera [27].
Enzymology and Molecular Genetics of Biological Halogenation
175
2.1.3 Perhydrolases
However, a different non-heme type of halogenating enzyme requiring hydrogen peroxide for halogenating activity, was isolated from bacteria.Although this type of halogenating enzymes was detected by its “haloperoxidase activity” using monochlorodimedone as the organic substrate [28], they are no real peroxidases [29]. These enzymes contain neither a heme group nor any metal ions [30] and should be classified as perhydrolases. They are a very homogeneous type of halogenating enzymes. They consist of identical subunits with a molecular weight of about 31,000 and the native enzymes are either dimers or trimers [24]. The specific activities for bromination of monochlorodimedone are 0.5–45 units mg–1 [31]. All the abovementioned haloperoxidases or perhydrolases have a lack of substrate specificity and regioselectivity in common [32]. 2.1.4 FADH2-Dependent Halogenases
In 1997, the detection of two halogenases which belong to a novel type of halogenating enzymes was described [33]. These two halogenases were identified in a Pseudomonas fluorescens strain producing the antifungal antibiotic pyrrolnitrin [34]. Chlorinating activity could only be detected in crude extracts, when NADH and the natural substrates, tryptophan and monodechloroaminopyrrolnitrin, respectively, were added [33]. However, during early stages of the purification of the enzymes, it was realized that FAD was also required [29]. Further purification revealed that not only NADH and FAD were necessary for halogenating activity, but that a second enzyme, which was identified as a flavin reductase, was also required. This flavin reductase was only necessary for the production of FADH2 from FAD and NADH, which was then used by the halogenase. While FMN was accepted by the flavin reductase, FMNH2 was not accepted by the halogenase. No specificity between the flavin reductase and the halogenases is required, since the flavin reductase present in the Pseudomonas fluorescens strain can be substituted by other flavin reductases from different bacterial sources [35, 36]. Tryptophan 7- and monodechloroaminopyrrolnitrin 3-halogenase consist of identical subunits with molecular weights of 61,000 and 64,000, respectively; however, it is not yet known whether these enzymes are monomers or dimers. Metal analysis showed that monodechloroaminopyrrolnitrin 3-halogenase contains copper as a metal ion [Wage, unpublished results], whereas only calcium was detected in tryptophan 7-halogenase [Unversucht, unpublished results]. 2.1.5 Methyl Halide Transferases
A very different type of enzyme responsible for the formation of methyl halides are the methyl halide transferases [37]. These enzymes are actually not halo-
176
K.-H. van Pée and S. Zehner
genating, but rather methylating enzymes [38]. They catalyse the synthesis of methyl halides from S-adenosine-L-methionine and halide ions (chloride, bromide, iodide). Methyl halide transferases have been detected in fungi [39], marine algae [40] and halophilic plants [41].
3 Molecular Genetics and Reaction Mechanisms 3.1 Heme-Containing Haloperoxidases
The primary structures of heme-containing haloperoxidases are not very homogeneous. The gene of chloroperoxidase from Caldariomyces fumago, which is the best-known and investigated haloperoxidase, has been cloned and sequenced [42]. The protein sequence shows similarities in protein database searches only to a putative peroxidase from sterigmatocystin biosynthesis from Emericella nidulans [43] and with a putative chloroperoxidase from Agaricus bisporus [44]. The identities are 31 % and 33 %, respectively. They share the Cys29 , the hemebinding residue, which was deduced from the crystal structure of Caldariomyces fumago chloroperoxidase [45]. They do not show similarities in primary structures to other heme-containing peroxidases, whereas human myeloperoxidase [46], bovine lactoperoxidase [47], human thyroid peroxidase [48] and human eosinophil peroxidase [49] share identities with each other in the range of 43 – 70 %. Neither these enzymes nor chloroperoxidase from Caldariomyces fumago show any sequence similarities to horseradish peroxidase and cytochrome P450 monooxygenases. Investigations into the reaction mechanism of heme-containing haloperoxidases were based on work by Chance and George on the roles of compound I and compound II in the mechanism of peroxidase catalysis [50, 51]. This mechanism invariably involves oxidation of the haloperoxidase by peroxide resulting in formation of compound I. The mechanism of compound I formation was based on the crystal structure of yeast cytochrome c peroxidase [52]. In contrast to the normal peroxidase reaction, in which compound I oxidizes the peroxidase substrate, compound I of haloperoxidases is the oxidant of the halide ions leading to a transient chlorinating intermediate formed by addition of chloride to the ferryl oxygen atom of haloperoxidase compound I [53]. This intermediate is very unstable at low pH values and will thus decompose very rapidly resulting in the formation of hypohalous acids that act as the actual halogenating agents (Fig. 3). According to this mechanism, which is based on biomimetic studies and the three-dimensional structure of chloroperoxidase from Caldariomyces fumago [45, 54], only substrates that can reach the active site very rapidly and fit exactly into the active site would be halogenated in a directly enzyme-catalysed reaction. All other substrates will be halogenated or oxidized in a non-enzymatic reaction by free hypohalous acid. This explains why heme-containing haloperoxidases seem to lack substrate specificity and regioselectivity.
Enzymology and Molecular Genetics of Biological Halogenation
177
Fig. 3. Proposed reaction mechanism of the heme-containing chloroperoxidase from Caldar-
iomyces fumago [45, 54]. Fe is the heme ion; the porphyrin ring system is not shown
178
K.-H. van Pée and S. Zehner
3.2 Vanadium-Containing Haloperoxidases
The vanadium-containing haloperoxidases share more sequence similarities than the heme-containing haloperoxidases; however, these similarities depend on the origin of the enzymes. The amino acid sequences of the vanadium chloroperoxidases from the fungi Curvularia inaequalis [55], Embellisia didymospora [56] and Drechslera biseptata [57], show sequence identities to each other in the range of 65 – 95 %. The vanadium bromoperoxidase from the brown alga Ascophyllum nodosum [58, 59] shares 43–45% identity with the two vanadium bromoperoxidases from the red alga Corallina pilulifera [60], whereas the bromoperoxidases from the two brown algae A. nodosum and Fucus distichus [61] have a sequence identity of 86%. The sequences from the red alga Corallina pilulifera and those from the brown alga Fucus distichus share 32 – 34% identity [62]. The similarities between algal bromoperoxidases and fungal chloroperoxidases are very low. They show only three stretches of high similarity in the regions providing the metal ion-binding site and the catalytic region located at the C-terminus [62, 63]. The vanadium-binding motif and the active site were first elucidated from the X-ray structure of the vanadium-containing chloroperoxidase from the fungus Curvularia inaequalis. It spans a length of approximately 150 amino acid residues. This motif contains the metal-coordinating histidine residue His496 . To stabilize the structure, various hydrogen bonds to the oxygen atoms of vanadate are formed from Lys353 ,Arg360 , Ser402 , Gly403 and Arg490 .An internal His404 is proposed to act as an acidic/basic group in the catalytic mechanism [63]. In the reaction of vanadium-containing haloperoxidases, the first step in the reaction sequence is the binding of the peroxide to the vanadate to form a peroxovanadate intermediate – the structure of which has not yet been elucidated. Binding of the halide ion is suggested to be supported by protonation of a histidine residue at the active site. The halide ion is not bound to the vanadium, but to an oxygen atom derived from the peroxide, and breaking of the oxygen-oxygen bond results in formation of hypohalous acids (Fig. 4) [27, 59, 62–66].Again, as in the case of the heme-containing haloperoxidases, the formation of free hypohalous acids results in non-enzymatic halogenation without substrate specificity and regioselectivity. There are striking similarities in the architecture of these enzymes with phosphatases and apo-chloroperoxidase from Curvularia inaequalis was shown to have phosphatase activity [57, 67].
Fig. 4. Proposed reaction mechanism of vanadium-containing chloroperoxidase from Curvu-
laria inaequalis [68]. E–V is the enzyme-bound catalytic vanadium ion
Enzymology and Molecular Genetics of Biological Halogenation
179
Similarities to bacterial non-heme haloperoxidases/perhydrolases occur only in a short region (residues 259–280) [68]. This region also has sequence similarity with other a/b-hydrolase fold enzymes. However, the crystal structure of vanadium-containing chloroperoxidase from Curvularia inaequalis lacks structural similarity to the characteristic a/b-hydrolase fold enzymes; the secondary structure of this protein is mainly a-helical [63]. 3.3 Perhydrolases
The sequences of this group of enzymes are very similar and form a homogeneous group.Within this group, the similarities are in the range 37–68%. The primary structures from all these enzymes show significant similarities with hydrolases and esterases, for instance 25% to the esterases EstF from Pseudomonas fluorescens [69] and EstP from Pseudomonas putida MR-2068 [70]. Furthermore, they have similarities to different hydrolysing enzymes such as carboxyesterases, lactonohydrolases and thioesterases. Since the bacterial non-heme halogenating enzymes that were isolated using their apparent haloperoxidase activity did not contain any metal ions, it was difficult to imagine how they could catalyse a peroxidase-type reaction. A catalytic mechanism involving the oxidation of a methionine residue by hydrogen peroxide [71] had to be abandoned, since this methionine residue was not conserved in these enzymes. Elucidation of the three-dimensional structure showed that they contain a catalytic triad consisting of a serine, histidine and aspartate residue, normally found in a/b-hydrolases [72]. Biochemical investigations revealed that these halogenating enzymes produce a low molecular weight organic compound in the presence of hydrogen peroxide and acetate buffer that could be separated from the enzyme by ultrafiltration. This small organic compound could then be used to achieve the bromination of monochlorodimedone, when added to a solution containing monochlorodimedone and bromide ions [73]. Obviously, these enzymes function as perhydrolases. The first step in the reaction sequence is the formation of an ester linkage between the active-site serine residue and acetic acid. This ester is cleaved by hydrogen peroxide in a perhydrolase reaction resulting in the formation of peracetic acid. The enzymaticallyformed peracetic acid will then oxidize the halide ions resulting in the formation of hypohalous acid. Thus, as in the case of heme- and vanadium-containing haloperoxidases, hypohalous acids are produced which act as the halogenating agents (Fig. 5) [31]. However, the bacterial perhydrolases do not produce hypohalous acid enzymatically, but only the peracid is produced in the enzyme-catalysed reaction. Perpropionic and perbutyric acid can also be produced by these perhydrolases [73] and the peracids of longer chain fatty acids can be produced by lipases [74]. As a consequence of the aforementioned mechanism, halogenation accomplished with these perhydrolases also proceeds without substrate specificity and regioselectivity.
180
K.-H. van Pée and S. Zehner
Fig. 5. Proposed reaction mechanism for halogenation reactions initiated by perhydrolases
[31, 73]
3.4 FADH2-Dependent Halogenases
The two halogenases involved in pyrrolnitrin biosynthesis from Pseudomonas fluorescens belong to a novel group of halogenating enzymes [33]. Their sequences show no similarity to the formerly mentioned halogenating enzymes. The halogenating activity in vivo could be demonstrated by mutational analysis of the encoding genes [34]. Interestingly, these two enzymes from the same organism and even the same secondary metabolite biosynthesis, are very distinct in their primary structure. The similarity between these sequences is below 25%. They share only two highly conserved motifs, the supposed FADH2-binding site (GXGXXG) and a motif containing two tryptophan residues (WXWXIP) (Figs. 6 and 7). Similar genes were detected during the study of halometabolite biosyntheses in other microorganisms. All of these putative halogenases contain both conserved motifs. The sequences show either similarity to the first or second halogenase from pyrrolnitrin biosynthesis. These similarities in the primary structures correspond to the substrate specificities. Thus, a classification of FADH2-dependent halogenases into at least two subgroups is suggested; the first acts on indole or tryptophan derivatives and the second halogenates phenyl or pyrrole derivatives. Five different genes from pyrrolnitrin-producing strains belong to the subgroup of indole-halogenating enzymes, the tryptophan 7-halogenases from Pseudomonas fluorescens CHA0, P. pyrrocinia, P. aureofaciens, Burkholderia cepacia, and Myxococcus fulvus [75] with identities in their protein sequences between 49–95% (Table 1). Genetic investigations led to the detection of a novel putative halogenase gene from the thienodolin producer Streptomyces albogriseolus. The amino acid se-
181
Enzymology and Molecular Genetics of Biological Halogenation
Table 1. Identities of the subgroup I (indole/tryptophan derivatives-halogenating enzymes) of FADH2-dependent halogenases with tryptophan 7-halogenase from Pseudomonas fluorescens BL915
Enzyme
Strain
Metabolite
Identity (%)
Reference
PrnA-P PrnA-A PrnA-F PrnA-B PrnA-M Thal
Pseudomonas pyrrocinia Pseudomonas aureofaciens Pseudomonas fluorescens CHA0 Burkholderia cepacia Myxococcus fulvus Sreptomyces albogriseolus
pyrrolnitrin pyrrolnitrin pyrrolnitrin pyrrolnitrin pyrrolnitrin thienodolin
95 95 94 90 49 58
Cml
Streptomyces venezuelae
chloramphenicol
25
[75] [75] [75] [75] [75] [Schmid, unpublished results] [85], accession no. AY026946 a
a
http://www.ncbi.nlm.nih.gov/Entrez/.
Fig. 6. Alignment of FADH2-dependent halogenases proposed to catalyse the halogenation of indole derivatives showing the two conserved motifs. PrnA-P PrnA from Pseudomonas pyrrocinia, PrnA-B PrnA from Burkholderia cepacia, PrnA-A PrnA from Pseudomonas aureofaciens, PrnA-F PrnA from Pseudomonas fluorescens BL915, PrnA-M PrnA from Myxococcus fulvus, Thal tryptophan 6-halogenase from Streptomyces albogriseolus
quence shows significant similarities and fits perfectly into the group of indolehalogenating enzymes (Table 1). The sequence contains the conserved motifs of FADH2-dependent halogenases (Fig. 6). This potential tryptophan 6-halogenase from the thienodolin producer Streptomyces albogriseolus is the first halogenase from an actinomycete that is supposed to halogenate an indole derivative [Schmid, unpublished results]. A gene for another potential tryptophan 7-halogenase was detected by hybridisation and PCR in the rebeccamycin producer Saccharothrix aerocoligenes [76].
182
K.-H. van Pée and S. Zehner
The group of supposed phenyl- or pyrrole-halogenating enzymes is less homogeneous. The halogenases from the pyrrolnitrin biosynthetic pathways (PrnCs) of five different species share significant similarities with proteins from other halometabolite biosynthetic gene clusters. Until now, this subgroup includes 15 published protein sequences of putative halogenases, presumably acting on amino acid- or polyketide-derived substrates, or other aromatic compounds. These halogenase genes were detected in the biosynthetic gene clusters of the peptide antibiotics chloroeremomycin [77], balhimycin [78], complestatin [79] and the anabaenopeptilides [80]. Additional halogenase genes were found in the clusters of the polyketide-derived antibiotics 7-chlorotetracycline [81] and pyoluteorin [82] and in the avilamycin A [83] and pyrrolnitrin biosynthetic gene clusters [75]. Two such genes were also detected in the pentachloropseudilin producer Actinoplanes sp. [Wynands, unpublished results] and one in the xanthomonadin producer Xanthomonas oryzae [84]. The identities of the genes of this subgroup are 25–95% (Table 2). All these sequences fit into this subgroup and show the typical conserved motifs of FADH2-dependent halogenases (Fig. 7). Table 2. Identities of subgroup II (phenyl or pyrrole derivatives-halogenating enzymes) of
FADH2-dependent halogenases with monodechloroaminopyrrolnitrin 3-halogenase from Pseudomonas fluorescens BL915 Enzyme
Strain
Metabolite
Identity (%)
Reference
PrnC-P PrnC-B PrnC-F Cts4
Pseudomonas pyrrocinia Burkholderia cepacia Myxococcus fulvus Streptomyces aureofaciens
95 94 80 42
[75] [75] [75] [81]
HalA
Actinoplanes sp.
pyrrolnitrin pyrrolnitrin pyrrolnitrin 7-Cl-tetracycline pentachloropseudilin
44
HalB
Actinoplanes sp.
pentachloropseudilin
44
PltM PltA BhaA ORF10
Pseudomonas fluorescens Pf-5 Pseudomonas fluorescens Pf-5 Amycolatopsis mediterranei Amycolatopsis orientalis
37 24 27 25
ComH AviH ORF11
Streptomyces lavendulae Streptomyces viridochromogenes Xanthomonas oryzae
pyoluteorin pyoluteorin balhimycin chloroeremomycin complestatin avilamycin A xanthomonadin
[Wynands, unpublished results] [Wynands, unpublished results] [82] [82] [78] [77]
AdpC
Anabena sp. 90
21
Cml
Streptomyces venezuelae
anabaenopeptilide chloramphenicol
a
http://www.ncbi.nlm.nih.gov/Entrez/.
25 22 25
25
[79] [84] [83], accession no. AAG38844 a [80] [85], accession no. AY026946 a
Enzymology and Molecular Genetics of Biological Halogenation
183
Fig. 7. Alignment of FADH2-dependent halogenases proposed to catalyse the halogenation of
phenyl and pyrrole derivatives. PrnC-F PrnC from Pseudomonas fluorescens Bl915, PrnC-M PrnC from Myxococcus fulvus, Cts4 halogenase from Streptomyces aureofaciens, HalB halogenase B from Actinoplanes sp., HalA halogenase A from Actinoplanes sp., PltM halogenase M from Pseudomonas fluorescens Pf-5, ere-ORF10 halogenase from Amycolatopsis orientalis, BhaA halogenase from Amycolatopsis mediterranei, ComH halogenase from Streptomyces lavendulae, AviH halogenase from Streptomyces viridochromogenes, xan-ORF11 halogenase from Xanthomonas oryzae, PltA halogenase A from Pseudomonas fluorescens Pf-5, AdpC halogenaase from Anabena sp. 90
A similar gene from the chloramphenicol producer Streptomyces venezuelae was recently detected by Piraee and Vining [85]. The gene product is supposed to be involved in halogenation of the aliphatic side chain of chloramphenicol. It shows similarities to enzymes of both subgroups (25%). The corresponding enzyme may belong to a further subgroup of FADH2-dependent halogenases, the alkyl halidases. Further genes with significant similarities to genes of FADH2-dependent halogenases, however, without any information about a relationship to halometabolite biosyntheses, have recently been detected in a number of different organisms. Seven genes were detected in the cyanobacterium Caulobacter crescentus [86], one during sequencing of the Streptomyces coelicolor genome [87], one in Mus
184
K.-H. van Pée and S. Zehner
Fig. 8. Unrooted phylogenetic tree of FADH2-dependent halogenases from bacteria
musculus [88] and one in the African crawn frog Xenopus laevis [89]. Figure 8 shows an unrooted phylogenetic tree of FADH2-dependent halogenases. Biochemical investigations concerning the novel type of halogenating enzymes, the FADH2-dependent halogenases, are just in their infancies. Until now, only the two halogenases, involved in the biosynthesis of the antifungal antibiotic pyrrolnitrin [34], tryptophan 7- and monodechloroaminopyrrolnitrin 3halogenase, have been investigated in some detail [35]. They do not contain any cofactor or prosthetic group with an absorbance in the visible region. Based on
Enzymology and Molecular Genetics of Biological Halogenation
185
Fig. 9. Proposed reaction mechanism of FADH2-dependent halogenases [35, 36]. As an example, the chlorination of monodechloroaminopyrrolnitrin to aminopyrrolnitrin is shown. The reduction of FAD is catalysed by a flavin reductase. FADH2 is suggested to bind to the halogenase before reacting with oxygen. Halogenase - - - FADHOOH is the halogenase containing the flavin hydroperoxide. The actual halogenation and the dehydration step are suggested to be both catalysed by the halogenase
the requirement for FADH2 , oxygen and halide ions, a mechanism for the halogenation reaction catalysed by these enzymes is suggested that involves activation of oxygen by FADH2 , followed by reaction of the FADH2-activated oxygen with a double bond of the organic substrate. This could lead to formation of an epoxide or an open structure with subsequent incorporation of the halide ion as a nucleophile. The halohydrin thus formed would then be dehydrated to the halogenated end-product (Fig. 9). Theoretically, such a mechanism would not only allow iodination, bromination and chlorination of specific organic substrates, but also fluorination. However, so far neither iodination nor fluorination of tryptophan or monodechloroaminopyrrolnitrin has been reported. This suggests that these enzymes not only show substrate specificity for the organic substrate, but also for the halide ion and that iodide is probably to large. Fluoride is usually hydrated in aqueous media and is therefore only a very weak nucleophile. Investigations of the substrate specificity of tryptophan 7-halogenase have shown that this enzyme accepts a number of tryptophan, indole and phenyl pyrrole derivatives in addition to its natural substrate tryptophan; however, with
186
K.-H. van Pée and S. Zehner
Fig. 10. Proposed reaction mechanism for fluorination reactions catalysed by mutant a-glucosidases [92]
any other substrate than tryptophan, the reaction proceeds with a relaxed regioselectivity. While tryptophan is exclusively chlorinated at the 7-position, other tryptophan or indole derivatives are chlorinated at the 2- or/and 3-positions [90]. 3.5 Fluorinating Enzymes
To date, only two enzymes have been reported that catalyse the incorporation of fluoride ions into an organic substrate [91, 92]. In contrast to the other halogenating enzymes described, these two enzymes are not natural halogenases. They are active site mutants of two retaining glucosidases, Agrobacterium sp. b-glucosidase and Cellulomonas fimi b-mannosidase. They can catalyse the formation of carbon-fluorine bonds with nucleophilic fluoride. In both mutants, the catalytic glutamate nucleophile was replaced with alanine, glycine or serine, respectively. Fluoride, which is probably desolvated in the active site, is obviously acting in the place of the missing catalytic nucleophile (Fig. 10). The two mutant enzymes can also catalyse the nucleophilic halogenation of 2,4-dinitrophenyl b-glycoside with chloride and bromide. A novel fluorinating enzyme was identified in the fluoroacetate- and 4-fluorothreonine-producing bacterium Streptomyces cattleya; however, no data on the enzyme are yet available [O’Hagan, personal communication].
4 Functions of Halogenating Enzymes 4.1 Haloperoxidases
Chloroperoxidase from Caldariomyces fumago was isolated under the assumption that it was involved in caldariomycin biosynthesis. This was supported by the ability of the enzyme to catalyse the chlorination of b-ketoadipic acid to
Enzymology and Molecular Genetics of Biological Halogenation
187
d-chlorolevulinic acid, a postulated intermediate of caldariomycin biosynthesis [6, 93], and also the chlorination of 1,3-cyclopentanedione and 2-chloro-1,3cyclopentanedione, possible late intermediates of caldariomycin biosynthesis [94, 95]. Interestingly, chloroperoxidase is an extracellular enzyme, while caldariomycin biosynthesis proceeds intracellularly. These early investigations have already shown that chloroperoxidase from Caldariomyces fumago catalyses the chlorination of a number of structurally different substrates and obviously has a very low substrate specificity, if any at all. The lack of substrate specificity of this and other heme- and vanadium-containing haloperoxidases, was later confirmed and can be explained by the reaction mechanisms of these enzymes [96]. Additionally, the involvement of haloperoxidases in the biosynthesis of halometabolites has never been proven for any chloro- or bromoperoxidase. On the contrary, it could be shown that a heme-containing bromoperoxidase, which was initially thought to be involved in chloramphenicol biosynthesis, was not involved in the formation of this antibiotic [97]. The only haloperoxidase which seems to be involved in the specific formation of a halometabolite is thyroid peroxidase, which seems to be involved in iodination of tyrosine residues contained in thyroglobulin [98]. From comparisons of the iodination reactions catalysed by horseradish peroxidase, lactoperoxidase and thyroid peroxidase, it was concluded that, while horseradish and lactoperoxidase liberate HOI, thyroid peroxidase catalysis iodination of tyrosine residues through enzyme-bound hypoiodous acid [99]. The biological function of the heme-containing haloperoxidases myeloperoxidase, eosinophil peroxidase and lactoperoxidase is in the defence against infections [100]. Myeloperoxidase, released from activated neutrophils, was reported to be employed to damage apolipoprotein A-1 in the presence of hydrogen peroxide and chloride [101]. The myeloperoxidase/hydrogen peroxide/chloride system was also reported to chlorinate NADPH that might also have a microbiocidal effect, however, this could also be deleterious to the host cell [102]. Eosinophil peroxidase may play a role in chronic parasite infections as a major risk factor for cancer development in many underdeveloped countries. Eosinophil peroxidase is produced at higher levels in the case of parasite infections and in many forms of cancer. Plasma levels of bromide can be used by eosinophil peroxidase to produce hypobromous acid, which can lead to bromination of free nucleosides and double-stranded DNA possibly leading to mutations [103]. Concerning the biological function of vanadium-containing haloperoxidases, it is clear that these enzymes are not directly involved in the specific biosynthesis of halometabolites. However, it can be speculated that the hypohalous acids produced by these enzymes can be used for a number of different purposes, such as in the defence against infection or for the degradation of lignin [104]. From the sporophytes Laminaria digitata and L. saccharina, it was reported that they seem to require extracellular haloperoxidase-mediated oxidation of iodide for the uptake of iodine [105]. However, it could quite well be that the biological function of these enzymes is not related to their halogenating activity.
188
K.-H. van Pée and S. Zehner
4.2 Perhydrolases
The perhydrolases, which were originally isolated as metal- and cofactor-free haloperoxidases using the monochlorodimedone assay, show halogenating activity only under conditions that are very unlikely to exist in vivo. In a genedisruption experiment, it could be shown that the perhydrolase from a pyrrolnitrin-producing Pseudomonas fluorescens strain is not involved in halometabolite biosynthesis [106]. Sequence analyses of all the biosynthetic gene clusters for halometabolites, cloned so far, have also shown that they do not contain haloperoxidase or perhydrolase genes [36, 81] with the exception of the biosynthetic gene clusters for the vancomycin-type antibiotics chloroeremomycin [77] and balhimycin [78]. These gene clusters contain a gene with similarity to perhydrolase genes; however, they also contain the gene of an FADH2-dependent halogenase. 4.3 FADH2-Dependent Halogenases
In contrast to haloperoxidases and perhydrolases, the involvement of FADH2dependent halogenases in halometabolite biosynthesis has been established unequivocally [36]. 4.3.1 Biosynthesis of Halometabolites
Although halogenating enzymes have usually been isolated from organisms producing halometabolites, and it was therefore assumed that the isolated halogenating enzymes were involved in halometabolite formation, there was no real proof for this assumption. The haloperoxidases isolated showed very broad substrate specificity, or rather lacked any such, and in vitro experiments on the halogenation of potential intermediates in the biosynthetic pathways were only positive as long as these substrates were also susceptible to electrophilic chemical halogenation. One problem was that in many cases in which the halometabolite produced by the organism was a chlorometabolite, only bromoperoxidases could be detected [107–109]. In these cases, it was argued that either the correct substrate was not used or that the enzymes could only show chlorinating activity in vivo [6, 110]. This situation changed with the use of molecular genetic methods in the research on biological halogenation. 4.3.1.1 7-Chlorotetracycline Biosynthesis
Several perhydrolases with brominating and chlorinating activity [111–113] and their corresponding genes [114–116] have been isolated from 7-chlorotetracycline-producing Streptomyces aureofaciens strains. Although biosynthetic studies showed that 4-ketoanhydrotetracycline must be the substrate for the halo-
Enzymology and Molecular Genetics of Biological Halogenation
189
Fig. 11. Pathway for the biosynthesis of 7-chlorotetracycline [117]
genating enzyme, and that this enzyme must have substrate specificity, since after the subsequent transamination step (Fig. 11), halogenation is not possible anymore [117], and must also have regioselectivity, no investigations into the involvement of the isolated perhydrolases in 7-chlorotetracycline biosynthesis using 4-ketoanhydrotetracycline as a substrate could be performed, since this compound was not available. Dairi et al. [81] cloned the biosynthetic gene cluster for 7-chlorotetracycline biosynthesis from Streptomyces aureofaciens and the halogenase gene was identified by mutation and complementation experiments. The amino acid sequence derived from this gene had no similarity to any known halogenating enzyme nor did the sequence, at that time, give any hint on the type of enzyme this halogenase could belong to. This was partially due to the fact that, because of a sequencing error, 100 amino acids were missing at the amino terminal end of the amino acid sequence originally published [81, 82, 118]. This was the first proof for the existence of halogenases other than haloperoxidases and perhydrolases.
190
K.-H. van Pée and S. Zehner
4.3.1.2 Pyrrolnitrin Biosynthesis
During investigations into the biosynthesis of the antifungal antibiotic pyrrolnitrin that is produced by many Pseudomonas [119] and a few Myxococcus strains [120], a perhydrolase had been isolated and characterized. Cloning of the gene and its use in a gene-disruption experiment revealed that this perhydrolase was not involved in pyrrolnitrin biosynthesis [106]. Hammer et al. [118] succeeded in isolating the pyrrolnitrin biosynthetic gene cluster, which consists of four genes. It was already known by then, that pyrrolnitrin biosynthesis requires two halogenation steps [119], but it was not known, whether they were catalysed by a single enzyme or by two different halogenases. Preparation of mutants blocked in every single step and feeding experiments allowed the identification of the genes [34] showing that the cluster contained two novel halogenase genes with no obvious sequence similarity to each other [118]. The first step in pyrrolnitrin biosynthesis is the regioselective chlorination of tryptophan by tryptophan 7-halogenase. The second step is a ring rearrangement leading to the phenyl pyrrole structure, and the third step is the regioselective chlorination of monodechloroaminopyrrolnitrin in the 3-position of the pyrrole ring by monodechloroaminopyrrolnitrin 3-halogenase. The last step is the oxidation of the amino group of aminopyrrolnitrin to the nitro group of pyrrolnitrin (Fig. 12) [34]. Thus the substrates for the two halogenases are structurally rather different and the two enzymes cannot substitute each other in vivo [118], although tryptophan 7-halogenase is able to chlorinate monodechloroaminopyrrolnitrin in vitro, however, without regioselectivity [90], whereas monodechloroaminopyrrolnitrin 3-halogenase is not able to chlorinate tryptophan [Wage, unpublished result]. The biosynthetic gene cluster does not contain the gene of the flavin reductase, which is necessary for the formation of FADH2 from NADH and FAD, required as a cofactor by both halogenases [36, 118]. 4.3.1.3 Pyoluteorin Biosynthesis
Pyoluteorin is also an antibiotic containing a pyrrole and a phenyl ring, but it is only dichlorinated in the pyrrole moiety (Fig. 13). Nowak-Thompson et al. [82] isolated and sequenced the biosynthetic gene cluster. They detected three potential halogenase genes (pltA, pltD and pltM) with similarity to the halogenase gene from 7-chlorotetracycline biosynthesis and to the monodechloroaminopyrrolnitrin 3-halogenase gene, but without similarity to the tryptophan 7halogenase gene. However, one of these three enzymes (PltD) lacks the FADH2binding site and can therefore not function as an FADH2-dependent halogenase. Since pyoluteorin contains only two chlorine atoms, two halogenases should be sufficient for pyoluteorin synthesis. However, the fact that PltD is required for pyoluteorin biosynthesis suggests that this enzyme is not a halogenase, but has a different function in pyoluteorin biosynthesis. The phenyl ring of pyoluteorin is synthesized by polyketide synthases [121], whereas the pyrrole ring is derived from proline [82], but it is not known,
Enzymology and Molecular Genetics of Biological Halogenation
191
Fig. 12. Pathway for the biosynthesis of pyrrolnitrin [34, 90]
how and when the two ring systems are connected and whether chlorination occurs before or after the connection of the two aromatic rings. Preliminary investigations into the in vitro activity of PltM have shown that it is actually an FADH2-dependent halogenase that can chlorinate pyrrole and also indole, although the activity is rather low which is not astonishing, since both compounds are not likely to be natural substrates for this enzyme [Falke, unpublished results]. 4.3.1.4 Biosynthesis of the Vancomycin Group Antibiotics
The cyclic peptide antibiotics of the vancomycin group, such as vancomycin, teicoplanin, balhimycin, chloroeremomycin and complestatin, contain mono- or dichlorinated phenolic rings, derived from the amino acid tyrosine [122]. In the last few years, the biosynthetic gene clusters of chloroeremomycin [77], balhimycin [78] and complestatin [79] have been cloned and sequenced. While in
192
K.-H. van Pée and S. Zehner
Fig. 13. Hypothetical pathway for the biosynthesis of pyoluteorin [82, 121]
the chloroeremomycin and balhimycin gene clusters a perhydrolase and a gene for an FADH2-dependent halogenase were found, the complestatin biosynthetic cluster contains only the gene for an FADH2-dependent halogenase. The results obtained by Puk et al. [123] show that the FADH2-dependent halogenase is responsible for the incorporation of the chloride atoms into balhimycin and most likely into chloroeremomycin too, since the balhimycin and chloroeremomycin biosynthetic cluster are very homologous. Although it is not yet known when chlorination occurs during the biosynthetic pathway, it was shown that free tyrosine is not the substrate [Böhm, unpublished results]. However, whether the peptidyl carrier protein-tethered tyrosine or b-hydroxytyrosine, free b-hydroxytyrosine or a longer peptide chain containing a b-hydroxytyrosine residue is the substrate for the halogenase, is not yet known. The results by Puk et al. [123] suggest that the perhydrolase acts as a thioesterase and catalyses the release of b-hydroxytyrosine, and that halogenation occurs after incorporation of b-hydroxytyrosine into the peptide chain. Complestatin does not contain a halogenated b-hydroxytyrosine residue and in the biosynthetic gene cluster no peptide synthetase domain and no perhydrolase gene for the formation and release of bhydroxytyrosine were found. Thus, while the substrates for the FADH2-dependent halogenases from chloroeromymycin and balhimycin biosynthesis are probably identical, the substrate for the FADH2-dependent complestatin halogenase must be different. 4.3.1.5 Chloramphenicol Biosynthesis
In contrast to the halometabolites discussed above, chloramphenicol which is produced by Streptomyces venezuelae and S. phaeochromogenes is not chlorinated
Enzymology and Molecular Genetics of Biological Halogenation
193
Fig. 14. Hypothetical pathway for chloramphenicol biosynthesis [124, 125]. R is either H or an acetyl moiety
at an aromatic ring, but contains a dichloroacetyl residue. As with most other halometabolites, it is not known exactly when halogenation occurs during biosynthesis. It is not clear whether chlorination occurs before or after acetylation and thus the substrate for the halogenating enzyme is not known. It was shown that acetic acid is not the substrate and Vats et al. [124] suggested that acetoacetyl-CoA could be the substrate (Fig. 14) [125]. Piraee and Vining [85] recently isolated the chloramphenicol biosynthetic gene cluster and demonstrated that the cluster contains a gene with similarity to an FADH2-dependent halogenase gene.According to the mechanism suggested for these enzymes [33, 36] they require a double bond for the incorporation of a halide ion. Thus, the enol form of acetoacetyl-CoA could be a potential substrate for the halogenase involved in chloramphenicol biosynthesis (Fig. 15).
194
K.-H. van Pée and S. Zehner
Fig. 15. Proposed chlorination of acetoacetyl-CoA during chloramphenicol biosynthesis. For details see legend to Fig. 9
4.3.1.6 Barbamide Biosynthesis
Barbamide is a halometabolite produced by the cyanobacterium Lyngbya majuscula and is also chlorinated in the aliphatic part of the molecule (Fig. 16) [126]. Feeding experiments with stable isotopes provided clear evidence that barbamide biosynthesis involves chlorination of the unactivated methyl group of leucine [127]. Since there was obviously no loss of deuterium from C-3 or C-4 of leucine, when deuterated L-[2H10]leucine was fed to cultures of Lyngbya majuscula, no formation of a double bond could occur and therefore formation of an intermediate epoxide with subsequent halohydrin formation would not be possible during the synthesis of the trichloromethyl group of barbamide. Thus, halogenation during barbamide biosynthesis could not occur through an electrophilic or nucleophilic addition of a chlorine species indicating the possibility of a radical halogenation mechanism.
Enzymology and Molecular Genetics of Biological Halogenation
195
Fig. 16. Chemical structure of barbamide [126]
4.4 Methyl Halide Transferases
Methyl halide transferases are involved in the production of methyl chloride by fungi of the hymenochaetaceae family [38]. Methyl chloride was shown to act as a methyl donor in the biosynthesis of the methyl esters of benzoic acid and in the methylation of phenol and butyrate by Phellinus pomaceus [128], and also in the formation of veratryl alcohol during lignin degradation by several lignin degrading fungi [129].
5 Applications of Halogenating Enzymes and their Genes Due the toxicity of many halogenated organic compounds, there is a demand for regioselective halogenation reactions which would allow the synthesis of halocompounds without the formation of unwanted by-products. However, chemical halogenation, which very often uses the free halogens or very reactive intermediate halocompounds, often produces halogenated by-products that cause severe problems and damage to the environment. Thus, biological halogenation was thought to be a possible solution to part of this problem. However, the results obtained using haloperoxidases and perhydrolases in the synthesis of halocompounds were rather disappointing. Although many attempts were made to use these enzymes in organic synthesis [5, 96], there were no substantial differences and improvements concerning the product formation [130]. The novel FADH2-dependent halogenases seem to have the potential to change this situation, since they can catalyse regioselective halogenation reactions. One disadvantage of these enzymes is the requirement of cofactors; however, this could be overcome when the genes of these halogenases are used in combinatorial biosynthesis, where they can be of a very high potential value for the production of novel halogenated metabolites with new biological activities and properties. Acknowledgement. We wish to thank Thomas Böhm, David O’Hagan, Corina Schmid, Tobias
Wage, Ina Wynads and Susanne Unversucht for allowing us to incorporate some of their unpublished results.
196
K.-H. van Pée and S. Zehner
6 References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43.
Drechsel E (1896) Z Biol 33 : 85 Shaw PD, Hager LP (1959) J Am Chem Soc 81:1012 Shaw PD, Hager LP (1960) J Biol Chem 236:1626 Morrison M, Schonbaum GR (1976) Ann Rev Biochem 45:861 Neidleman SL, Geigert J (1986) Biohalogenation: principles, basic roles and applications. Ellis Horwood, Chichester Shaw PD, Beckwith JR, Hager LP (1959) J Biol Chem 234:2560 Morris DR, Hager LP (1966) J Biol Chem 241:1763 Hager LP, Morris DR, Brown FS, Eberwein H (1966) J Biol Chem 141 : 1769 Hewson WD, Hager LP (1980) J Phycol 16:340 van Pée K-H, Lingens F (1984) FEBS Lett 173:5 van Pée K-H, Lingens F (1985) J Gen Microbiol 131:1911 Rae TD, Goff HM (1996) J Am Chem Soc 118:2103 Fiedler TJ, Davey CA, Fenna RE (2000) J Biol Chem 275:11964 Chen YP, Lincoln DE, Woodin, SA, Lovell CR (1991) J Biol Chem 266:23909 Vilter H (1983) Bot Mar 26 :451 Vilter H (1984) Phytochemistry 23:1387 De Boer E, Tromp MGM, Plat H, Krenn GE, Wever R (1986) Biochim Biophys Acta 872 : 104 Itoh N. Izumi Y, Yamada H (1986) J Biol Chem 261:5194 Krenn BE, Izumi Y, Yamada H, Wever R (1989) Biochim Biophys Acta 998:63 Sheffiled DJ, Harry TR, Smith AJ, Rogers LJ (1993) Phytochemistry 32 : 21 Almeida M, Filipe S, Humanes M, Maia MF, Mel R, Severino N, da Silva JAL, Frausto da Silva JJR, Wever R (2001) Phytochemistry 57 : 633 Plat H, Krenn BE, Wever R (1987) Biochem J 248 : 277 Liu T-N, M’Timkulu T, Geigert J, Wolf B, Neidleman SL, Silva D, Hunter-Cevera JC (1987) Biochem Biophys Res Commun 142:329 van Pée K-H (1990) Biocatalysis 4:1 van Schijndel JWPM, Vollenbroek EGM, Wever R (1993) Biochim Biophys Acta 1161:249 Itoh N, Izumi, Y, Yamada H (1986) J Biol Chem 261:5194 Soedjak HS, Butler A (1990) Biochemistry 29 : 7974 Wiesner W, van Pée K-H, Lingens F (1986) FEBS Lett 209:321 van Pée K-H, Keller S, Wage T, Wynands I, Schnerr H, Zehner S (2000) Biol Chem 381:1 Wiesner W, van Pée K-H, Lingens F (1988) J Biol Chem 263:13725 Hofmann B, Tölzer S, Pelletier I, Altenbuchner J, van Pée K-H, Hecht HJ (1998) J Mol Biol 279:889 van Pée K-H (1996) Annu Rev Microbiol 50:375 Hohaus K, Altmann A, Burd W, Fischer I, Hammer PE, Hill DS, Ligon JM, van Pée K-H (1997) Angew Chem Int Ed 36:2012 Kirner S, Hammer PE, Hill DS, Altmann A, Fischer I, Weislo LJ, Lanahan M, van Pée K-H, Ligon JM (1998) J Bacteriol 180 : 1039 Keller S, Wage T, Hohaus K, Hölzer M, Eichhorn E, van Pée K-H (2000) Angew Chem Int Ed 39:2300 van Pée K-H (2001) Arch Microbiol 175:250 Harper DB (1985) Nature 315:55 Harper DB, Kennedy JT, Hamilton JTG (1988) Phytochemistry 27:3147 Harper DB, Hamilton JTG (1988) J Gen Microbiol 134:2831 Wuosmaa AM, Hager LP (1990) Science 249:160 Ni X, Hager LP (1998) Proc Natl Acad Sci USA 95:12866 Nuell MJ, Fang G-H, Axley MJ, Kenigsberg P, Hager LP (1988) J Bacteriol 170:100 Brown DW,Yu JH, Kelkar HS, Fernandes M, Nesbitt TC, Keller NP, Adams TH, Leonard TJ (1996) Proc Natl Acad Sci USA 93:1418
Enzymology and Molecular Genetics of Biological Halogenation
197
44. Morales-Almora P, Thurston CF Molecular analysis of the cellulolytic genes in Agaricus bisporus accession AJ293759 45. Sundaramoorthy M, Terner J, Poulos TL (1998) Chem Biol 5 : 461 46. Morishita K, Kubota N, Asano S, Kaziro Y, Nagata S (1987) J Biol Chem 262:3844 47. Dull TJ, Uyeda C, Strosberg AD, Nedwin G, Seilhamer JJ (1990) DNA Cell Biol 9 : 499 48. Kimura S, Kotani T, McBride OW, Umeki K, Hirai K, Nakayama T, Ohtaki S (1987) Proc Natl Acad Sci USA 84:5555 49. Sakamaki K, Tomonaga M, Tsukui K, Nagata S (1989) J Biol Chem 264:16828 50. Chance B (1949) Arch Biochem Biphys 22:224 51. George P (1953) Biochem J 54 : 267 52. Poulos TL, Kraut J (1980) J Biol Chem 255:8199 53. Lambeir A-M, Dunford HB (1983) J Biol Chem 258:13558 54. Wagenknecht H-A, Woggon W-D (1997) Chem Biol 4 :367 55. Simons BH, Barnett P,Vollenbroek EG, Dekker HL, Muijsers AO, Messerschmidt A, Wever R (1995) Eur J Biochem 229:566 56. Barnett P, Hemrika W, Dekker HL, Muijsers AO, Renirie R, Wever R (1998) J Biol Chem 273:23381 57. Hemrika W, Renirie R, Dekker HL, Barnett P, Wever R (1997) Proc Natl Acad Sci USA 94:2145 58. Vilter H (1995) Met Ions Biol Syst 31:325 59. Weyand M, Hecht HJ, Kieß M, Liaud M-F,Vilter H, Schomburg D (1999) J Mol Biol 293:595 60. Shimonishi M, Kuwamoto S, Inoue H, Wever R, Ohshiro T, Izumi Y, Tanabe T (1998) FEBS Lett 428:105 61. Vreeland V, Ng K, Epstein L (1998) accession AAC35279 62. Isupov MN, Dalby AR, Brindley AA, Izum Y, Tanabe T, Murshudov BN, Littlechild JA (2000) J Mol Biol 299:1035 63. Messerschmidt A, Wever R (1996) Proc Natl Acad Sci USA 93:392 64. De Boer E, Wever R (1988) J Biol Chem 263 : 12326 65. van Schijndel JWPM, Barnett P, Roelse J,Vollenbroek EGM,Wever R (1994) Eur J Biochem 225:151 66. Davey CA, Fennan RE (1996) Biochemistry 35:10967 67. Renirie R, Hemrika W, Piersma S, Wever R (2000) Biochemistry 39:1133 68. Wever R, Hemrika W (1998) In: Nriagu JO (ed) Vanadium in the environment. Part I: Chemistry and biochemistry. VCH 69. Choi KD, Jeohn GH, Rhee JS, Yoo OJ (1990) Agric Biol Chem 54:2039 70. Ozaki E, Sakimae A, Numazawa R (1995) Biosci Biotechnol Biochem 59:1204 71. Haag T, Lingens F, van Pée K-H (1991) Angew Chem Int Ed 30:1487 72. Hecht HJ, Sobek H, Haag, T, Pfeifer O (1994) Nature Struct Biol 1:532 73. Picard M, Gross J, Lübbert E, Tölzer S, Krauss S, van Pée K-H, Berkessel A (1997) Angew Chem Int Ed 36:1196 74. Kirk O, Conrad LS (1999) Angew Chem Int Ed 38:1196 75. Hammer PE, Burd W, Hill DS, Ligon JM, van Pée K-H (1999) FEMS Microbiol Lett 180:39 76. Zehner S, Wynands I, Schmid C, Tölzer S, Kuschel M, Barth G, van Pée K-H (2000) 100th General Meeting of the American Society for Microbiology, Los Angeles, USA 77. van Wageningen AMA, Kirkpatrick PN,Williams DH, Harris BR, Kershaw JK, Lennard NJ, Jones M, Jones SJM, Solenberg PJ (1998) Chem Biol 5:155 78. Pelzer S, Süßmuth R, Heckmann D, Recktenwald J, Huber P, Jung G, Wohlleben W (1999) Antimicrob Agents Chemother 43 : 1565 79. Chiu H-T, Hubbard BK, Shah AN, Eide J, Fredenburg RA, Walsh CT Khosla C (2001) Proc Natl Acad Sci USA 98:8548 80. Rouhiainen L, Paulin L, Suomalainen S, Hyytiäinen H, Buikema W, Haselkorn R, Sivonen K (2000) Mol Microbiol 37:156 81. Dairi T, Nakano K, Aisaka K, Katsumata R, Hasegawa M (1995) Biosci Biotech Biochem 59:1099 82. Nowak-Thompson B, Chaney N, Wing JS, Gould SJ, Loper JE (1999) J Bacteriol 181 : 2166
198
K.-H. van Pée and S. Zehner
83. Weitnauer G, Mühlenweg A, Trefzer A, Hoffmeister D, Süßmuth RD, Jung, G, Welzel K, Vente A, Girreser U, Bechthold A (2001) Chem Biol 8:569 84. Goel AK, Rajagopal L, Nagesh N, Sonti RV (2001) Xanthomonadin biosynthetic gene cluster from Xanthomonas oryzae pv. oryzae accession AY010120, protein AAG38844.2 85. Piraee M,Vining LC (2001) Streptomyces venezuelae halogenase accession AY026946 86. Nierman WC, FeldblyumTV, Laub MT, Paulsen IT, Nelson KE, Eisen J, Heidelberg JF, Alley, RK, Ohta N, Maddock JR, Potocka I, Nelson WC, Newton A, StephensC, Phadke ND, Ely B, DeBoy RT, Dodson RJ, Durkin AS, Gwinn ML, Haft DH, Kolonay JF, Smit J, Craven M, Khouri H, Shetty J, Berry K, Utterback T, Tran K, Wolf A, Vamathevan J, Ermolaeva M, White O, Salzberg SL, Venter JC, Shapiro L, Fraser CM (2001) Proc Natl Acad Sci USA 98:4136 87. Redenbach M, Kieser HM, Denapaite D, Eichner A, Cullum J, Kinashi H, Hopwood DA (1996) Mol Microbiol 21:77 88. Marra M, Hillier L, Kucaba T, Martin J, Beck C, Wylie T, Underwood K, Steptoe M, Theising B,Allen M, Bowers Y, Person B, Swaller T, Gibbons M, Pape D, Harvey N, Schurk R, Ritter E, Kohn S, Shin T, Jackson Y, Cardenas M, McCann R,Waterston R,Wilson R (1999) The WashU-NCI Mouse EST project accession AI426004 89. Clifton S, Johnson SL, Blumberg B, Song J, Hillier L, Pape D, Martin J,Wylie, T, Underwood K, Theising B, Bowers Y, Person B, Gibbons M, Harvey N, Ritter E, JacksonY, McCann R, Waterston R, Wilson R (1999) Xenopus laevis gastrula non normalized (1999) WashU Xenopus EST project accession BF614668 90. Hölzer M, Burd W, Reißig H-U, van Pée K-H (2001) Adv Synth Catal 343:591 91. Zechel DL, Stoll D, Mohammadzadeh T, Warren RAJ, Withers SG (2001) Angew Chem Int Ed 40:417 92. Zechel DL, Reid SP, Nashiru O, Mayer C, Stoll D, Jakeman DL, Warren RAJ, Withers SG (2001) J Am Chem Soc 123:4350 93. Shaw PD, Hager LP (1959) J Biol Chem 234:1565 94. Beckwith JR, Clark R, Hager LP (1963) J Biol Chem 238:3086 95. Beckwith JR, Hager LP (1963) J Biol Chem 238:3091 96. Franssen MCR (1994) Biocatalysis 10:87 97. Facey SJ, Groß F, Vining LC, Yang K, van Pée K-H (1996) Microbiology 142:657 98. Taurog A (1970) Recent Prog Horm Res 26:189 99. Sun W, Dunford HB (1993) Biochemistry 32:1324 100. Dunford HB (1999) Heme peroxidases. Wiley-VCH, New York 101. Bergt C, Marsche G, Panzenboeck U, Heinecke JW, Malle E, Sattler W (2001) Eur J Biochem 268:3523 102. Auchère F, Bertho G, Artaud I, Girault JP, Capeillère-Blandin C (2001) Eur J Biochem 268:2889 103. Shen Z, Mitra SN, Wu W, Chen Y, Yang Y, Qin, J, Hazen SL (2001) Biochemistry 40:2041 104. Wever R, Barnett P, Hemrika W (1997) Structure and physiological function of vanadium chloroperoxidase. In: Winkelmann G, Carrano CJ (eds) Transition metals in microbial metabolism. Harwood Academic Publishers, UK, p 415 105. Küpper FC, Schweigert N, Ar Gall E, Legendre J-M, Vilter H, Kloareg B (1998) Planta 207:163 106. Kirner S, Krauss S, Sury G, Lam ST, Ligon JM, van Pée K-H (1996) Microbiology 142:2129 107. van Pée K-H, Lingens F (1985) J Bacteriol 161 : 1171 108. Plat H, Krenn BE, Wever R (1987) Biochem J 248 : 277 109. Knoch M, van Pée K-H, Vining LC, Lingens F (1989) J Gen Microbiol 135:2493 110. van Pée K-H (1990) Biotech Adv 8: 185 111. van Pée K-H, Sury G, Lingens F (1987) Biol Chem Hoppe-Seyler 368:1225 112. Krenn BE, Plat H, Wever R (1988) Biochim Biophys Acta 952:255 113. Weng M, Pfeifer O, Krauss S, Lingens F, van Pée K-H (1991) J Gen Microbiol 137:2539 114. van Pée K-H (1988) J Bacteriol 170:5890 115. Pfeifer O, Pelletier I, Altenbuchner J, van Pée K-H (1992) J Gen Microbiol 138:1123 116. Pelletier I, Pfeifer O, Altenbuchner J, van Pée K-H (1994) Microbiology 140:509
Enzymology and Molecular Genetics of Biological Halogenation
199
117. McCormick JRD (1967) Tetracyclines. In: Gottlieb D, Shaw PD (eds) Antibiotics, Vol IIBiosynthesis. Springer Berlin, Heidelberg. New York 118. Hammer PE, Hill DS, Lam ST, van Pée K-H, Ligon JM (1997) Appl Environ Microbiol 63:2147 119. van Pée K-H, Ligon JM (2000) Nat Prod Rep 17:157 120. Gert K, Trowitsch W, Wray V, Höfle G, Irschik H, Reichenbach H (1982) J Antibiotics 35:1101 121. Nowak-Thompson B, Gould SJ, Loper JE (1997) Gene 204 : 17 122. Gribble GW (1996) In: Herz W, Kirby GW, Moore RE, Steglich W, Tamm Ch (eds) Progress in the chemistry of organic natural products – Naturally occurring organohalogen compounds – A comprehensive survey. Springer, Wien New York, p 261 123. Puk O, Huber P, Bischoff D, Recktenwald J, Jung G, Süßmuth R, van Pée K-H Wohlleben W Pelzer S (2002) Chem Biol 9 : 225 124. Vats S, Stuttard C, Vining LC (1987) J Bacteriol 169:3809 125. Groß F, Lewis EA, Piraee M, van Pée K-H, Vining LC, White RL (2001) Med Chem Lett 12:283 126. Orjala J, Gerwick WH (1996) J Nat Prod 59 : 427 127. Sitachitta N, Márquez BL, Williamson RT, Rossi J, Roberts MA, Gerwick WH, Hguyen VA, Willis CL (2000) Tetrahedron 56:9103 128. McNally KJ, Harper DB (1991) J Gen Microbiol 137:1029 129. Harper DB, Buswell JA, Kennedey JT, Hamilton JTG (1990) Appl Environ Microbiol 56:3450 130. van Pée, K-H (1995) Halogenation. In: Drauz K, Waldmann H (eds) Enzyme catalysis in organic synthesis, Vol II. VCH, Weinheim, p 783
The Handbook of Environmental Chemistry Vol. 3, Part P (2003): 201–214 DOI 10.1007/b 10448
Myeloperoxidase and Eosinophil Peroxidase: Phagocyte Enzymes for Halogenation in Humans Jeffrey P. Henderson 1, Jay W. Heinecke 2 1 2
Department of Medicine, Washington University School of Medicine, St. Louis, MO 63110, USA Department of Medicine, University of Washington, Seattle, WA 98195, USA E-mail:
[email protected]
Oxidative damage to biomolecules has been implicated in tissue damage during acute and chronic inflammation. One potential mechanism involves reactive oxygen species produced by peroxidases of professional phagocytes – neutrophils, monocytes, macrophages, and eosinophils. We have shown that activated phagocytes employ myeloperoxidase and eosinophil peroxidase to halogenate proteins, lipids, and nucleobases in vitro. The reaction pathways involve hypohalous acids, molecular halides, and interhalogen compounds.We have used sensitive and specific mass spectrometric methods to demonstrate that certain of these halogenated products are present in human inflammatory tissue. These observations indicate that myeloperoxidase and eosinophil peroxidase promote biohalogenation reactions in vitro and in vivo. We therefore propose that reactive species produced by peroxidases halogenate amino acids, proteins, nucleotide precursors, RNA, and DNA. Collectively, our observations indicate a novel mechanism for tissue damage by activated phagocytic white blood cells during inflammation. This process might alter proteins and genes, enabling phagocyte peroxidases to produce cytotoxic or even tumorigenic changes in inflamed tissue. Keywords. Hypochlorous acid, Hypobromous acid, 3-Chlorotyrosine, 5-Chlorouracil, 5-Bromodeoxycytidine
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 202
1
Introduction
2
The Phagocyte Peroxidases
3
Tyrosine Halogenation in Humans
4
Nucleic Acid Halogenation in Humans
5
Phagocyte Peroxidases as a Source of Interhalogen Compounds
6
Chemical Mechanisms of Bromination by Phagocytes . . . . . . . 211
7
Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
8
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
. . . . . . . . . . . . . . . . . . . . . 203 . . . . . . . . . . . . . . . . . 204 . . . . . . . . . . . . . . . 206 . 209
© Springer-Verlag Berlin Heidelberg 2003
202
J. P. Hendersen and J. W. Heinecke
1 Introduction Reactive oxygen species derived from the NADPH oxidase of phagocytic white blood cells play a central role in immune defense against invading pathogens [1, 2]. The use of potentially lethal oxidants by professional phagocytes – neutrophils, monocytes, macrophages, and eosinophils – suggests that generation of superoxide and other oxidants must be carefully controlled. Otherwise, such reactive species might damage tissue at sites of inflammation. Indeed, many lines of evidence implicate phagocyte oxidants in the pathogenesis of diseases ranging from atherosclerosis to ischemia-reperfusion injury to cancer [3 – 15]. Neutrophils form the first line of defense against invading bacteria and fungi. Circulating neutrophils are quiescent, but they are quickly and synchronously activated by a wide array of microbial and non-microbial agonists. One important agonist in vivo is a component of opsonized bacteria that binds to receptors on the neutrophil membrane [16, 17]. Binding leads to activation of NADPH oxidase and triggers the phagocytosis of the particulate bacterium. The NADPH oxidase reduces molecular oxygen to superoxide (O2∑ –) by using NADPH, but not NADH, as a co-factor [2]. NADPH + 2O2 Æ 2O2∑ – + NADP+ + H+
(1)
The key role of the NADPH oxidase system in host defenses against microbial pathogens is illustrated by chronic granulomatous disease. In this genetic disorder, defects in specific components of the oxidase impair superoxide production, and recurrent bacterial and fungal infections result [18]. Thus, superoxide and its dismutation product, hydrogen peroxide (H2O2), are critical for killing pathogenic organisms. Neutrophils, monocytes, and certain populations of macrophages also express the heme enzyme myeloperoxidase. Myeloperoxidase uses the oxidizing equivalents of H2O2 to oxidize chloride to the chlorinating agent hypochlorous acid (HOCl) [19, 20]. Cl– + H2O2 + H+ Æ HOCl + H2O
(2)
Superoxide and myeloperoxidase are secreted into a subcellular compartment called the phagolysosome, where bacteria become bathed in high local concentrations of oxidants [21]. Neutrophil granules containing an array of microbicidal proteins also are secreted into the phagolysosome, thereby enhancing its toxic environment [22]. Moreover, phagocytosis stimulates neutrophils to secrete superoxide, H2O2 , and myeloperoxidase into the extracellular milieu, as do soluble agonists that trigger oxidant production [23]. Therefore, phagocytes can generate reactive oxygen species that kill both phagocytosed and extracellular microbes. Eosinophils, which play a role in immediate hypersensitivity and host defense against parasites, have been implicated in tissue damage during asthma and parasitic infections [24, 25]. Their histochemical hallmark is cytoplasmic granules that stain with eosin. When these cells are activated, these granules release spe-
Myeloperoxidase and Eosinophil Peroxidase: Phagocyte Enzymes for Halogenation in Humans
203
cialized proteins that are key to the eosinophilic cytotoxic armamentarium. Eosinophil peroxidase, a major component of the granules, uses H2O2 derived from the NADPH oxidase to convert bromide to the brominating agent hypobromous acid (HOBr) [26–28]. The diffusible hypohalous acids produced by myeloperoxidase and eosinophil peroxidase are significantly more cytotoxic than H2O2 . Moreover, they leave a distinctive chemical signature by halogenating proteins and nucleic acids into stable end-products that have been identified by in vitro studies [29–34]. Phagocytic cells thus possess an arsenal of diverse proteins capable of killing cells by multiple mechanisms such as oxidant generation, proteolytic digestion, and disruption of the cell membrane. The actions of phagocytic peroxidases, particularly myeloperoxidase, have been of increasing interest because of their involvement in defense against microorganisms. Tissue damage by myeloperoxidase has also been implicated in the genesis of cancer, atherosclerosis, multiple sclerosis, and Alzheimer’s disease [35–44]. Recent studies suggest that eosinophil peroxidase may oxidize biomolecules during asthma and other conditions associated with an eosinophilic tissue infiltrate [45, 46]. This chapter discusses halogenation by phagocyte enzymes and its relationships to host defense mechanisms and human disease.
2 The Phagocyte Peroxidases The gene family of mammalian peroxidases includes myeloperoxidase, eosinophil peroxidase, lactoperoxidase, which is present in exocrine secretions, and thyroid peroxidase, which is involved in the iodination and aromatic coupling reactions of thyroid hormone synthesis [47, 48]. These peroxidases are significantly homologous with each other, as well as with the fungal chloroperoxidase and the plant horseradish peroxidase [49]. Myeloperoxidase is an exceptionally cationic protein [50]. It is highly abundant in the azurophilic granules of phagocytic monocytes and neutrophils, reaching up to 5% of neutrophil protein by weight [51, 52]. Myeloperoxidase was first purified in 1941 by Agner [51], who named it verdoperoxidase because of its striking green color. This color, responsible for the green hue of pus and infected skin lesions, is attributable to the unique heme group at the enzyme’s active site. The active site of native myeloperoxidase contains a ferric heme, which is oxidized by H2O2 in a two-electron reaction to form compound I (Fig. 1) [53–55]. The latter is a short-lived intermediate composed of oxyferryl iron (Fe(IV) = O) coordinated to a p-cation radical porphyrin. Myeloperoxidase compound I has a reduction potential of 1.1 V, which enables it to convert chloride to HOCl by another two-electron oxidation reaction that reforms the native ferric heme. Compound I can also oxidize other halides (bromide, iodide) and the pseudohalide thiocyanate [56]. Other compounds, such as nitrite and aromatic compounds such as the amino acid L-tyrosine, undergo a one-electron oxidation reaction that yields a radical intermediate derived from the reducing substrate and converts compound I to compound II [57, 58]. At plasma concentrations of these substrates, chloride is widely considered to be the favored substrate.
204
J. P. Hendersen and J. W. Heinecke
Fig. 1. Reactions of peroxidases with H2O2 and reducing substrates. Peroxidases react with
H2O2 to form compound I, a short-lived intermediate composed of oxyferryl iron (Fe(IV) = O) coordinated to a p-cation radical porphyrin. Compound I can convert halides to hypohalous acids via a two-electron oxidation reaction, regenerating native enzyme. Compound I can also react with one-electron donors (such as nitrite and L-tyrosine) to generate a radical intermediate and compound II. Compound II is then reduced to native enzyme with the concomitant one-electron oxidation of another reducing substrate
Like myeloperoxidase, eosinophil peroxidase also undergoes peroxidase cycling, but it favors substrates with lower oxidation potentials than chloride. One such substrate is bromide (Br –), which is present at 20–100 µM in plasma [59, 60]. Br – + H2O2 + H+ Æ HOBr + H2O
(3)
Remarkably, at plasma concentrations of chloride (ª 100 mM) and bromide (ª 100 µM), eosinophil peroxide exclusively converts bromide ions to hypobromous acid [26–28].
3 Tyrosine Halogenation in Humans Hypochlorous acid generated by myeloperoxidase can oxygenate several amino acids (e.g., methionine, cysteine) and deaminate several others (e.g., lysine, taurine) [61–64]. It can also oxygenate and/or fragment the ring of aromatic amino acids (e.g., tryptophan) [65]. The products are not specific to hypochlorous acid because they can also be produced by certain non-halogenating oxidants. However, in vitro studies suggest that only the myeloperoxidase system – and not any other oxidant system proposed to be active in humans – chlorinates tyrosine to 3-chlorotyrosine (Fig. 2) [5, 33]. The presence of 3-chlorotyrosine in biological material therefore indicates that myeloperoxidase was responsible for the oxidative damage. 3-Chlorotyrosine is generated by activated neutrophils exposed to free L-tyrosine and has been found in human atherosclerotic lesions, plasma proteins
Myeloperoxidase and Eosinophil Peroxidase: Phagocyte Enzymes for Halogenation in Humans
205
Fig. 2. Halogenation of tyrosine by myeloperoxidase and eosinophil peroxidase
from patients undergoing dialysis, sputum proteins of patients with cystic fibrosis, and in lung washings from patients with acute respiratory distress syndrome [5, 33, 66–68]. Oxidant stress has been implicated in these conditions, and the presence of elevated levels of chlorinated tyrosine indicates that halogenating oxidants generated by myeloperoxidase can damage biomolecules in vivo. 3-Chlorotyrosine levels rise dramatically in mouse models of acute infection and inflammation, whereas genetically altered mice that lack functional myeloperoxidase fail to make this unique marker under the same conditions [69, 70]. This finding strongly supports the hypothesis that myeloperoxidase is the major source of chlorinating oxidants in vivo. HOCl generated by myeloperoxidase might not be the intermediate that chlorinates tyrosine, however, because generation of free 3-chlorotyrosine by the complete myeloperoxidase-H2O2-chloride system or by HOCl itself is optimal at acidic pH [33]. Because the pKa for HOCl/ClO– is ª 7.4, the chlorinating intermediate might be Cl2 rather than HOCl. Cl2 is in equilibrium with HOCl via a reaction that requires H+ and chloride. HOCl + Cl– + H+ a Cl2 + H2O
(4)
To investigate this possibility, we determined whether tyrosine chlorination requires chloride. HOCl failed to chlorinate free tyrosine in the absence of this halide. The reaction also was optimal under acidic conditions, which was consistent with a requirement for H+. Finally, at neutral pH and in the absence of chloride, molecular chlorine readily generated 3-chlorotyrosine. These results strongly suggest that Cl2 , rather than HOCl, is the chlorinating intermediate with which myeloperoxidase oxidizes free tyrosine. In vitro studies have also implicated Cl2 in the chlorination of lipids and nucleobases [29, 71].
206
J. P. Hendersen and J. W. Heinecke
Results from mass spectrometric analysis of head space gas derived from the myeloperoxidase-H2O2-chloride system are consistent with this interpretation. The system produced a gas with the expected mass-to-charge ratio and isotope pattern of Cl2 [33] Taken together, these observations suggest that the myeloperoxidase system of human neutrophils generates Cl2 , a potent halogenating agent that would be expected to more readily permeate lipid membranes than HOCl. Eosinophil peroxidase and, to a variable degree, myeloperoxidase can generate brominating oxidants [26 – 28, 30, 34]. When compared with HOCl, peroxidase-derived brominating agents have a greater propensity for halogenating aromatic compounds under physiologically plausible conditions. Accordingly, mammalian peroxidases generate significant yields of 3-bromotyrosine at physiological concentrations of chloride and bromide (Fig. 2) [34, 70]. Mass spectrometric studies have detected a rise in 3-bromotyrosine levels in mice during acute infection and inflammation and in asthmatic humans after exposure to allergens [45, 70]. 3-Bromotyrosine production is impaired in mice deficient in either myeloperoxidase or eosinophil peroxidase, indicating that both pathways might generate brominating intermediates in vivo [70, 72].
4 Nucleic Acid Halogenation in Humans In addition to oxidizing amino acids, phagocyte products can also oxidize nucleic acids. This ability might help explain why chronic inflammation caused by bacteria, parasites, viruses, foreign bodies, environmental exposures, anatomical malformations, or unknown agents has been linked to human cancer [9, 73–75]. Infections associated with specific cancers include schistosomiasis (bladder cancer), liver fluke infection (bile duct carcinoma), chronic viral hepatitis (liver cancer), and Helicobacter pylori infection (gastric cancer). Non-infectious inflammatory diseases with cancer associations include ulcerative colitis (colon cancer), reflux esophagitis (esophageal adenocarcinoma), and pulmonary asbestosis (mesothelioma). Moreover, the ability of cigarette smoke to provoke an inflammatory response in the lungs might contribute to the increased risk of lung cancer among smokers. The great variety of inflammatory conditions that predispose humans to cancer has intensified interest in identifying features of the inflammatory response that contribute to mutagenesis and cancer development. Recent genetic epidemiological studies have found a relationship between polymorphisms in the promoter of the myeloperoxidase gene and the risk for promyelocytic leukemia, lung, and laryngeal cancers [35–40]. This observation, along with the strong association between specific cancers and infection by parasites such as Schistosoma haematobium, Chlonorchis sinensis, and Opisthorchis viverrini, also raise the question of whether eosinophil peroxidase exerts similarly mutagenic effects during inflammation. To determine whether the halogenating agents produced by mammalian myeloperoxidase and eosinophil peroxidase might be mutagenic and potentially carcinogenic, we determined whether these agents react with deoxyribonucleosides. We found that phagocyte peroxidase systems modify deoxycytidine to form the aromatic electrophilic substitution products 5-chlorodeoxycytidine and
Myeloperoxidase and Eosinophil Peroxidase: Phagocyte Enzymes for Halogenation in Humans
207
Fig. 3. Halogenation of pyrimidines by myeloperoxidase (MPO) and eosinophil peroxidase
(EPO)
5-bromodeoxycytidine (Fig. 3) [29 – 32]. Both activated neutrophils and the myeloperoxidase system generated 5-chlorodeoxycytidine.Activated eosinophils, activated neutrophils, the eosinophil peroxidase system, and the myeloperoxidase system generated 5-bromodeoxycytidine in the presence of physiologically plausible bromide concentrations. Thus, halogenation of nucleobases might be one potential mechanism for cytotoxicity and mutagenesis during inflammation. Unlike oxidized bases such as 8-hydroxydeoxyguanosine, halogenated pyrimidines are base analog mutagens that can be taken up from the extracellular space and erroneously used for DNA synthesis [76, 77]. The best-known base analog mutagen is 5-bromodeoxyuridine (BrdU), a thymidine analog. 5-Chloro, 5bromo-, and 5-iododeoxyuridine are thymidine analogs because the halogen group mimics the 5-methyl group of the thymine ring. 5-Fluorouracil is an analog of the RNA base uracil because of fluorine’s smaller atomic radius. Thymine normally forms a base pair with adenine in double-stranded DNA. However, the electron-withdrawing effect of 5-halopyrimidines increases the likelihood that a tautomer and/or anion will form [78, 79].As a result, thymidine analogs can base pair with guanine, and incorporation of these analogs leads to transition mutations (T to C, C to T, G to A, A to G). Dividing fibroblasts exposed to the product of deoxycytidine oxidation by eosinophil peroxidase incorporated BrdU into their nuclear DNA [31]. This observation is consistent with the mutagenic scheme offered in Fig. 4. The model suggests that mammalian peroxidases convert deoxycytidine to 5-halogenated deoxycytidine. The latter is transported into cells, where it is deaminated and phosphorylated. The resulting nucleotide can then be incorporated into DNA.
208
J. P. Hendersen and J. W. Heinecke
Fig. 4. A model for mutagenesis involving incorporation of halogenated nucleotides during DNA synthesis. Most studies of mutagenesis have focused on direct oxidative damage to DNA, but mutations could also arise if nucleobase analogs were incorporated into DNA in place of normal nucleobases, as occurs when proliferating cells are incubated with 5-chlorouracil, 5bromouracil, or 5-bromodeoxyuridine. We propose that reactive species produced by peroxidases halogenate nucleotides and nucleotide precursors that can be subsequently incorporated into DNA. Our observations suggest a novel mechanism for nucleotide precursor mutagenesis. This process might alter genes, enabling pyrimidine halogenation by myeloperoxidase (MPO) or eosinophil peroxidase (EPO) to produce carcinogenic changes in inflamed tissue
This model is unique for two reasons. First, it implicates endogenous halogenating agents in mutagenesis. Second, it places the critical oxidative event in the cytoplasm or outside the cell rather than directly in chromosomal DNA. Indeed, bacteria exposed to the myeloperoxidase chlorinating system accumulate 5-chlorodeoxycytidine in their RNA but not their DNA [29]. The surprising conclusion is that that RNA – rather than DNA – is the major target for oxidative damage by halogenating intermediates in bacteria and perhaps mammalian cells. Its susceptibility to damage might result from its single-stranded structure or location in the cell. Because cells can halogenate and deaminate cytosine and incorporate the resulting halogenated uracil into DNA, uracil itself might serve as a substrate for peroxidase-induced mutagenesis (Fig. 4). In vitro experiments have shown that phagocyte peroxidases can halogenate both deoxyuridine and uracil, the latter in near-quantitative yield [32]. These enzymes might therefore act as mutagenic alternatives to thymidylate synthase, converting the RNA base uracil to an analog of a DNA base by adding a bulky group (in this case chlorine or bromine) to the 5 position of the pyrimidine ring. Interestingly, halogenated deoxyuridines can act as irreversible, mechanism-based inhibitors of thymidylate synthase, depleting cells of thymidine and facilitating their own incorporation as thymidine analogs into DNA [77].
Myeloperoxidase and Eosinophil Peroxidase: Phagocyte Enzymes for Halogenation in Humans
209
We recently used mass spectrometry to demonstrate that free 5-chlorouracil and 5-bromouracil are detectable in human tissue obtained from sites of inflammation. These molecules represent initial products of endogenous halogenation by myeloperoxidase and eosinophil peroxidase (Fig. 4) and can also be derived by deaminating halogenated cytosine. We identified them by molecular mass, chromatographic retention time, and the unique isotope patterns characteristic of chlorinated and brominated compounds. We used isotope-labeled uracil to verify that the products could not be attributed to artifactual halogenation during the analytical workup. Plasma and urine from healthy volunteers used as controls contained no detectable 5-chlorouracil or 5-bromouracil. These observations provide compelling evidence that the inflammatory response in humans involves pyrimidine chlorination and bromination. Collectively, our results demonstrate that halogenation of pyrimidines is a physiological process and not a reaction confined to the chemist’s bench.
5 Phagocyte Peroxidases as a Source of Interhalogen Compounds Although oxidation of chloride by myeloperoxidase is well documented, both human neutrophils and eosinophils preferentially brominate uracil and deoxycytidine in the presence of physiological chloride and bromide concentrations [30]. We resolved this seemingly contradictory observation by showing that reagent HOCl and bromide generate brominating intermediates at plasma concentrations of halide ion. Taurine, a potent HOCl scavenger, inhibited this pathway but did not affect bromination by HOBr. It also inhibited bromination of nucleobases and nucleosides by both the myeloperoxidase system and activated neutrophils. These experiments suggest that oxidation of bromide by myeloperoxidase-derived HOCl might be a significant mechanism for generating brominating agents in vivo. Bromide and HOCl might react to form interhalogen compounds, which are combinations of different halogens (XXn¢). Both binary (BrCl, IBr, and ICl) and ternary (ICl3) interhalogens have been characterized. One pathway for their formation requires hypohalous acid (HOX) and halide ion (X¢ –). HOX + H+ + X¢ – a XX¢ + H2O
(5)
HOCl then reacts with bromide to form the interhalogen gas BrCl, itself a potent brominating agent [80]. HOCl + Br– + H+ Æ BrCl a HOBr + Cl– + H+
(6)
The redox potentials of chloride and bromide make this reaction essentially irreversible.Anions of interhalogens and polyhalides are also known; they include Cl3–, Br–3, Br2Cl–, and BrCl2– [81]. Interhalogens are extremely corrosive species that attack a wide range of compounds. To determine whether BrCl might play a role in bromination by myeloperoxidase, we sparged a reaction mixture containing the HOCl-Br– or the
210
J. P. Hendersen and J. W. Heinecke
Fig. 5 A – C. Mass spectrometric detection of 1-bromo-2-chlorocyclohexane (A) produced by HOCl-bromide (B) and the myeloperoxidase-H2O2-chloride-bromide system (C). (Figure reproduced with permission from the Journal of Biological Chemistry [30])
myeloperoxidase-H2O2-Cl–-Br– system with nitrogen gas that was subsequently passed through cyclohexene [30]. Mass spectrometric analysis of the resulting solution detected an ion with the expected mass-to-charge ratio, GC retention time, and isotopic abundance of 1-bromo-2-chlorocyclohexane (Fig. 5). Because cyclohexene is an aprotic, non-polar solvent that should not contain halide ions under these conditions, its bromination cannot involve a backsided nucleophilic attack of a bromonium ion intermediate by free Cl–. Instead, bromination is likely to involve attack on the double bond by [Br+ – Cl–] derived from molecular BrCl. Detecting 1-bromo-2-chlorocyclohexane therefore provides strong evidence that HOCl and myeloperoxidase generate the interhalogen gas BrCl. Although interhalogen compounds have not previously been invoked in living organisms, our observations strongly suggest that myeloperoxidase and, by extension, human neutrophils generate the interhalogen gas bromine chloride. Thus, production of molecular halides and interhalogen compounds by phagocyte peroxidases might be a physiologically relevant pathway for generating oxidants (Fig. 6). To determine if myeloperoxidase helps generate brominating agents in vivo, we determined levels of 3-bromotyrosine (an established marker of amino acid bromination by eosinophil peroxidase) in inflammatory fluid from normal mice and mice lacking functional myeloperoxidase by mass spectrometry [70]. The myeloperoxidase-deficient mice produced 60 % less 3-bromotyrosine than the normal mice, suggesting that myeloperoxidase provides one pathway for bromination during inflammation.
Myeloperoxidase and Eosinophil Peroxidase: Phagocyte Enzymes for Halogenation in Humans
211
Fig. 6. Generation of molecular halides and interhalogen compounds by HOCl and HOBr. Ul-
traviolet absorption spectrum of solutions following addition of 100 µM HOCl to 0.1 N HCl, 100 µM NaBr; 0.1 N HCl, 10 mM NaBr; or 0.1 N HBr, upper panel. The observed absorption maxima at 232, 245, 269 nm are consistent with formation of the interhalogen species BrCl2–, Br2Cl–, and Br3– respectively. These results support the conclusion that eosinophil peroxidase (EPO) and myeloperoxidase (MPO) generate interhalogens at physiological halide concentrations, lower panel
6 Chemical Mechanisms of Bromination by Phagocytes Eosinophil peroxidase uses the peroxidase catalytic cycle to oxidize bromide to HOBr [26–28]. HOCl also reacts with bromide to generate brominating intermediates that are in equilibrium with HOBr and BrCl [30]. Under physiologically plausible conditions, HOBr and other brominating agents halogenate aromatic compounds more effectively than HOCl. Indeed, eosinophil peroxidase readily halogenates L-tyrosine and pyrimidines in quantitative yields [31, 32, 34]. HOBr, the initial brominating oxidant released by eosinophil peroxidase, also rapidly reacts with primary amines to form bromamines. Rather than being inert products, bromamines are equally or more reactive than HOBr toward biomolecules. This suggests that microenvironments high in primary amines might concentrate reactive bromine in vivo.
212
J. P. Hendersen and J. W. Heinecke
Although HOBr has a lower oxidation potential than HOCl, it may be more reactive in some instances because it can produce higher order molecular halogens (Fig. 6) [81]. HOBr + nX– + H+ a BrXn
(7)
Intermediates such as Br2 and BrCl are much more effective at brominating aromatic compounds than is HOBr itself [82]. Moreover, Br2 and BrCl are in equilibrium with HOBr by reactions that require bromide and chloride [80]. HOBr + Br– + H+ Æ Br2 + H2O
(8)
HOBr + Cl– + H+ Æ BrCl + H2O
(9)
It is thus possible that stronger oxidants such as Br2 or BrCl derived from HOBr (or HOCl) in the presence of physiological concentrations of chloride and bromide are involved in brominating tyrosine and pyrimidines.
7 Conclusions The demonstration of links between phagocytes and oxidative damage to amino acids and nucleic acids implicates myeloperoxidase and eosinophil peroxidase in host defense mechanisms and tissue injury. The detection of the enzymes and their products in inflammatory lesions strongly supports the hypothesis that systems that produce hypohalous acids, molecular halides, and interhalogen compounds might be pivotal in the halogenation of biomolecules in vivo. These observations raise the possibility that inhibitors of peroxidases might retard or prevent the development of oxidative tissue damage inflicted by activated phagocytes, offering a new approach to cancer prevention and treatment.
8 References 1. Babior BM (1978) N Engl J Med 298 : 659 2. Babior BM (1999) Blood 93:1464 3. Leeuwenburgh C, Rasmussen JE, Hsu FF, Mueller DM, Pennathur S, Heinecke JW (1997) J Biol Chem 272 : 3520 4. Leeuwenburgh C, Hardy MM, Hazen SL,Wagner P, Oh-ishi S, Steinbrecher UP, Heinecke JW (1997) J Biol Chem 272:1433 5. Hazen SL, Heinecke JW (1997) J Clin Invest 99 : 2075 6. Heller JI, Crowley JR, Hazen SL, Salvay DM, Wagner P, Pennathur S, Heinecke JW (2000) J Biol Chem 275 : 9957 7. Hazen S, Gaut J, Crowley J, Hsu F, JW H (2000) Biochem J 352 : 693 8. Kilgore K, Lucchesi B (2000) Am Heart J 139:32 9. Weitzman SA, Gordon LI (1990) Blood 76:655 10. Marnett LJ (2000) Carcinogenesis 21:361 11. Beckman KB, Ames BN (1997) J Biol Chem 272:19633
Myeloperoxidase and Eosinophil Peroxidase: Phagocyte Enzymes for Halogenation in Humans
213
12. Dizdaroglu M, Olinski R, Doroshow JH, Akman SA (1993) Cancer Res 53:1269 13. Chaudhary AK, Nokubo M, Reddy GR,Yeola SN, Morrow JD, Blair IA, Marnett LJ (1994) Science 265:1580 14. Wink DA, Kasprzak KS, Maragos CM, Elespuru RK, Misra M, Dunams TM, Cebula TA, Koch WH, Andrews AW, Allen JS, Keefer LK (1991) Science 254:1001 15. Byun J, Henderson JP, Mueller DM, Heinecke JW (1999) Biochemistry 38:2590 16. McKenzie S, Schreiber A (1998) Curr Opin Hematol 5:16 17. Sengelov H (1995) Crit Rev Immunol 15:107 18. Lekstrom-Himes J, Gallin J (2000) New Engl J Med 343:1703 19. Harrison JE, Schultz J (1976) J Biol Chem 251:1371 20. Weiss SJ, Klein R, Slivka A, Wei M (1982) J Clin Invest 70:598 21. Hampton M, Kettle A, Winterbourn C (1998) Blood 92 : 3007 22. Dell’Angelica E, Mullins C, Caplan S, Bonifacino J (2000) FASEB J 14:1265 23. Klebanoff S (1999) Proc Assoc Am Physicians 111:383 24. Epstein FH (1998) N Engl J Med 338:1592 25. Gleich G, Adolphson C, Leiferman K (1993) Annu Rev Med 44:85 26. Weiss SJ, Test ST, Eckmann CM, Ross D, Regiani S (1986) Science 234:200 27. Mayeno AN, Curran AJ, Roberts RL, Foote CS (1989) J Biol Chem 264 : 5660 28. Thomas EL, Bozeman PM, Jefferson MM, King CC (1995) J Biol Chem 270 : 2906 29. Henderson JP, Byun J, Heinecke JW (1999) J Biol Chem 274 : 33440 30. Henderson JP, Byun J, Williams MV, Mueller DM, McCormick ML, Heinecke JW (2001) J Biol Chem 276 : 7867 31. Henderson JP, Byun J, Williams MV, McCormick ML, Parks WC, Ridnour LA, Heinecke JW (2001) Proc Natl Acad Sci USA 98:1631 32. Henderson JP, Byun J, Mueller DM, Heinecke JW (2001) Biochemistry 40:2052 33. Hazen SL, Hsu FF, Mueller DM, Crowley JR, Heinecke JW (1996) J Clin Invest 98 : 1283 34. Wu W, Chen Y, d’Avignon A, Hazen SL (1999) Biochemistry 38:3538 35. Piedrafita FJ, Molander RB,Vansant G, Orlova EA, Pfahl M, Reynolds WF (1996) J Biol Chem 271:14412 36. Reynolds WF, Chang E, Douer D, Ball ED, Kanda V (1997) Blood 90:2730 37. London SJ, Lehman TA, Taylor JA (1997) Cancer Res 57:5001 38. Cascorbi I, Henning S, Brockmoller J, Gephart J, Meisel C, Muller JM, Loddenkemper R, Roots I (2000) Cancer Res 60:644 39. Le Marchand L, Seifried A, Lum A, Wilkens LR (2000) Cancer Epidemiol Biomarkers Prev 9:181 40. Kantarci O, Lesnick T, Yang P, Meyer R, Hebrink D, McMurray C, Weinshenker B (2002) Mayo Clin Proc 77:17 41. Gaut J, Heinecke J (2001) Trends Cardiovasc Med 11:103 42. Nauseef W (2001) Int J Hematol 74:125 43. Reynolds W, Rhees J, Maciejewski D, Paladino T, Sieburg H, Maki R, Masliah E (1999) Exp Neurol 155:31 44. Reynolds W, Hiltunen M, Pirskanen M, Mannermaa A, Helisalmi S, Lehtovirta M,Alafuzoff I, Soininen H (2000) Neurology 55:1284 45. Wu W, Samoszuk MK, Comhair SAA, Thomassen MJ, Farver CF, Dweik RA, Kavuru MS, Erzurum SC, Hazen SL (2000) J Clin Invest 105:1455 46. MacPherson J, Comhair S, Erzurum S, Klein D, Lipscomb M, Kavuru M, Samoszuk M, Hazen S (2001) J Immunol 166:5763 47. O’Brien P (2000) Chem Biol Interact 129 : 113 48. Taurog A (1999) Biochimie 81:557 49. Ortiz de Montellano P (1992) Annu Rev Pharmacol Toxicol 32 50. Andersen MR, Atkin CL, Eyre HJ (1982) Arch Biochem Biophys 214:273 51. Agner K (1941) Acta Physiol Scand 2:1 52. Schultz J, Kaminker K (1962) Arch Biochem Biophys 96 : 465 53. Kettle AJ, Winterbourn CC (1997) Redox Rep 3:3 54. Marquez L, Huang J, Dunford H (1994) Biochemistry 33:1447
214
J. P. Hendersen and J. W. Heinecke: Myeloperoxidase and Eosinophil Peroxidase
55. Kooter IM, Moguilevsky N, Bollen A, van der Veen LA, Otto C, Dekker HL, Wever R (1999) J Biol Chem 274:26794 56. van Dalen CJ, Whitehouse MW, Winterbourn CC, Kettle AJ (1997) Biochem J 327:487 57. van Dalen C, Winterbourn C, Senthilmohan R, Kettle A (2000) J Biol Chem 275 : 11638 58. Marquez L, Dunford H (1995) J Biol Chem 270:30434 59. Holzbecher J, Ryan DE (1980) Clin Biochem 13 : 277 60. Ramsey PG, Martin T, Chi E, Klebanoff SJ (1982) J Immunol 128:415 61. Carr AC, Winterbourne CC (1997) Biochem J 327:275 62. Tsan M, Denison R (1981) Inflammation 5:379 63. Hazell LJ, van den Berg JJM, Stocker R (1994) Biochem J 302:297 64. Cunningham C, Tipton K, Dixon H (1998) Biochem J 330:939 65. Ximenes V, Catalani L, Campa A (2001) Biochem Biophys Res Commun 287:130 66. Himmelfarb J, McMenamin M, Loseto G, Heinecke J (2001) Free Rad Biol Med 31:1163 67. van der Vliet A, Nguyen MN, Shigenaga MK, Eiserich JP, Marelich GP, Cross CE (2000) Am J Physiol Lung Cell Mol Physiol 279:L537 68. Lamb N, Gutteridge J, Baker C, Evans T, Quinlan G (1999) Crit Care Med 27 : 1738 69. Brennan M, Anderson M, Shih D, Qu X, Wang X, Mehta A, Lim L, Shi W, Hazen S, Jacob J, Crowley J, Heinecke J, Lusis A (2001) J Clin Invest 107:419 70. Gaut JP, Yeh GC, Tran HD, Byun J, Henderson JP, Richter GM, Brennan ML, Lusis AJ, Belaaouaj A, Hotchkiss RS, Heinecke JW (2001) Proc Natl Acad Sci USA 98:11961 71. Hazen SL, Hsu FF, Duffin K, Heinecke JW (1996) J Biol Chem 271:23080 72. Denzler K, Borchers M, Crosby J, Cieslewicz G, Hines E, Justice J, Cormier S, Lindenberger K, Song W, Wu W, Hazen S, Gleich G, Lee J, Lee N (2001) J Immunol 167:1672 73. Wiseman H, Halliwell B (1996) Biochem J 313 : 17 74. Ohshima H, Bartsch H (1994) Mutat Res 305:253 75. Rosin MP, Anwar WA, Ward AJ (1994) Cancer Res 54:1929s 76. Morris SM (1991) Mutat Res 258:161 77. Morris SM (1993) Mutat Res 297:39 78. Sowers LC, Goodman MF, Eritja R, Kaplan B, Fazakerley GV (1989) J Mol Biol 205:437 79. Yu H, Eritja R, Bloom LB, Goodman MF (1993) J Biol Chem 268:15935 80. Kumar K, Margerum DW (1987) Inorg Chem 26 : 2706 81. Wang TX, Kelley MD, Cooper JN, Beckwith RC, Margerum DW (1994) Inorg Chem 33:5872 82. de la Mare PBD (1976) Electrophilic halogenation. Cambridge University Press, Cambridge
Subject Index
A ACEA (Association des Constructeurs Européens d’Automobile) 229, 246, 262, 267, 269 Acetylene 21 Acid rain 9 Additives 23, 46, 87, 270, 271, 274, 282 Agriculture 131, 134 Air pollution 9, 30, 110, 176, 182–184, 186, 203 Air pollution control act 183, 186 Air quality 73, 110, 116, 130, 177, 179, 183, 185, 186, 191, 192, 195, 203 Air-fuel mixture 22, 26–28, 30, 35, 39, 51, 74, 257 Air-Fuel Ratio (A/F) 20, 25, 27, 39, 44, 209, 263 Air-Toxic Components 266 Alcoholes 21, 105, 179, 279, 281, 282 Aldehydes 21, 179 Alkaline 45 Alternative fuels 11, 105, 276, 277, 284, 285 Alternative propulsion systems 19, 25, 91, 101, 283, 287 Aluminium (Al) 166, 167, 170 AMA (Automobile Manufacturer Association) 222 Ammonia (NH3) 46, 83 Anthropogenic greenhouse effect 118, 120 Anthropogenic sources 114, 120, 124, 129, 131, 134, 227 Aromatics 21, 23, 211, 261–264, 267, 269, 270, 274 Arrhenius 19 Automobile 6, 7, 9, 10, 284 Auto-Oil Programm 262, 266 B Baiersbronner Programm 203 Barium oxide (BaO) 45 Benzene 113, 266, 267, 269, 270 Bio-fuels 280–283
Biomass 105, 124, 277, 279, 280–282 Boiling curve 257, 258 Butane 277–279 C C/H ratio 23, 270, 279, 287 Cadmium (Cd) 165, 170 CAFE (Corporate Average Fuel Economy) 231, 232 California Air Resources Board (CARB) 95, 177, 178, 185, 193, 224, 236, 283 California test 209 Carbon (C) 22, 73, 281 Carbon canister 223 Carbon dioxide (CO2) 9, 13, 20, 38, 40, 41, 71, 81, 105, 113, 115, 116, 118, 121, 146, 151, 220, 227, 229, 233–235, 250, 258, 262, 266, 267, 271, 272, 275, 277, 279, 281–285 Carbon monoxide (CO) 9, 13, 20, 28, 30, 31, 35, 38, 41, 44, 73, 79, 81, 88, 113, 116, 126–128, 140, 141, 146, 147, 179, 184, 186, 187, 192, 193, 195, 196, 198–202, 205, 209, 214, 216, 219, 224, 239, 245, 263, 267, 274, 278, 282 Carburetor 27, 221–223, 268, 269 Catalyst 36, 37, 38, 88, 144, 184, 191, 203–205, 218, 236, 245, 273, 282 Cerium (Ce) 87 Cetan index 260 Cetane number 50, 258, 260, 264, 266, 269, 274, 282, 284, 287 Chamber Diesel engine 52 Charge turbulence 20, 34 Chlorofluorocarbons (CFCs) 124 Chromium VI (CrVI) 165, 170 Clean air act 113, 183, 184, 185, 195–197, 262 Clean fuels 179, 185, 270, 275, 287 Climate change 9, 233, 266 Closed crankcase ventilation 202 CO2 equivalent 119
290 Combustion 18, 20, 22, 27, 33, 50, 51, 74, 75, 91, 93, 94, 260, 275, 282 Combustion chamber 20, 22, 25, 30, 32–34, 37, 43, 50, 52, 54, 74, 75, 89, 93, 94, 260, 270 Combustion engine 22, 30, 268 Combustion products 20, 261, 277, 281 Commercial vehicle 110, 121, 137, 138, 147, 151, 152, 159, 160, 164 Common Rail 57, 59, 63, 65, 77, 80, 87, 88 Compression ratio (CR) 33, 34, 53, 69, 74, 258, 259 Compression stroke 25, 27, 34 Conformity of production (COP) 177, 178, 180 Continuously regenerating trap (CRT) 87, 88 CRC (Coordinating Research Council) 222 CVS (constant volume sampler) 211, 214, 219 Cylinder 20, 21, 30, 31, 33, 36, 51, 52, 79 D DeNOX catalyst 44–46, 82, 272 Deposition 115 Deposits 33, 115, 270, 271 Design parameters 33 Diesel engines 19, 21–23, 26, 27, 30, 48, 54, 103, 137, 140, 145, 180, 184, 185, 192, 195–197, 199, 200, 205, 211, 214, 229, 233, 236, 238, 239, 241, 249, 256–258, 260, 261, 266, 271, 273, 274, 277, 282, 284 Diesel fuel 19, 241, 244, 256, 258, 260, 266, 269, 271, 275, 276, 283, 284 Diffusion flame 23 Dimethyl Ether (DME) 279, 280 Dinitrogen monoxide (N2O) 118, 125 Direct injection 43, 50, 63, 75, 269, 270, 272 Direct injection Diesel 52–54, 63, 65, 71, 200, 239, 244, 266, 272 Distributor pump 57, 65 DOE (Department of Energy) 232, 283 DOT (Department of Transportation) 232 Driving cycle 209, 211, 214, 217, 218, 230 Dual-bed catalyst 38 Dumping 164, 165, 167, 171 E ECE – Regulation 202–204, 219, 245, 250 ECE test 35, 37, 73, 219 EEC – Commission 188, 201–205, 245, 269 Electric batteries 91, 95, 180 Electric energy 95, 101 Electric motor 95, 97, 101 Electric vehicle 95, 180, 182, 234 Emission 9, 74, 115, 138, 176, 183, 188
Subject Index Emission regulations 35, 88, 89, 184, 185, 187, 188, 190, 191 Emission standards 35, 38, 141, 191, 195, 198, 201–204, 209, 211, 218, 224, 238, 239, 245 Emission testing 183, 192, 209, 214, 216, 217 End of life vehicles (ELV) 163, 165, 166 Energy 3, 4, 9, 48, 50, 95, 115, 164, 167–169, 257, 274–276, 280, 281, 283, 284 Engine 3, 9, 160, 179, 185, 284, 285 Engine control 27, 62, 80, 271 Engine external measures 35, 36 Engine internal measures 30, 35, 41, 74, 88, 239 Environment 3, 8, 138 Environmental audits 171 Environmental friendly vehicle 284 Environmental management and auditing system (EMAS) 171 Environmental protection 9, 185 Environmental protection agency (EPA) 178, 184, 185, 192, 222, 230, 239, 268, 274 Ethane 277, 278 Ethanol 103, 277, 281–283, 285 EU3 – Emission Standard 73, 82, 88, 144, 151, 199, 219, 236, 245, 246, 267 EU4 – Emission Standard 73, 77, 146, 152, 178, 183, 199, 206, 245, 246, 267, 272 European stationary cycle (ESC) 245, 246 European transient cycle (ETC) 245 EUROPIA (European Petroleum Industry) 262 Evaporative emission 52, 117, 177, 179, 184, 220–222, 224, 227, 237, 238, 250, 258, 264, 270 Exhaust after-treatment 9, 36–38, 43–46, 74, 81, 83, 88, 89, 191, 197, 239, 263, 268, 271, 274, 284 Exhaust gas 8, 9, 20–23, 31–33, 116, 175, 179, 204, 217, 250, 264, 270 Exhaust gas emission 30, 33, 34, 50, 51, 67, 68, 73, 74, 93, 94, 211, 214, 227, 230, 238, 261, 262, 273 Exhaust pipe 37, 160 Exhaust port 36 Exhaust-gas recirculation (EGR) 36, 38, 55, 63, 77, 79, 88, 89, 236, 239, 245, 270 F Federal standards 192, 195, 222, 223, 239 Federal test procedures 209, 224 Ferocene 87 FID (flame ionization detection) 211, 214 Flame 20, 22
291
Subject Index
Flexible fuel vehicle 185, 282 Fluff 166–168 Fly-wheel accumulators 91, 96 Forest decay 9 Formaledehyde (HCHO) 179, 197, 224, 266, 267, 279 Fossil fuels 266, 276, 277, 280 Four stroke 54, 68, 91, 94, 97 Fuel 19, 20, 22, 23, 25, 74, 177, 179, 185, 186, 204, 250, 256–258, 261, 266–268, 274, 275, 284, 285 Fuel cells 91, 97, 101–103, 234, 283, 287 Fuel consumption 9, 11, 12, 27, 28, 30, 32, 34–36, 40, 41, 43, 51, 55, 65, 67, 68, 71, 93, 141, 220, 227, 229, 230, 234, 239, 241, 258, 259, 266, 267, 271, 275, 277, 284 Fuel injection 48, 57, 59 Fuel vapors 22, 25 G Gas guzzler 232 Gas to liquid (GTL) 279 Gas turbine 91, 93 Gasoline 19, 25, 103, 185, 203, 256–258, 269, 273, 275, 276, 279, 282–284, 287 Gasoline/air mixture 25 Gasoline direct injection (GDI) engines 43 Gasoline engine 19, 21, 22, 25, 27, 34, 40, 41, 103, 141, 145, 185, 195, 198, 199, 204, 205, 214, 217–219, 236, 256, 257, 261, 266, 270, 279 Global warming effect 227, 229, 234 Glow plugs 54 Greenhouse effect 9, 113, 115, 116, 118, 120, 152, 266 Greenhouse gases 11, 118, 120, 145, 146, 152, 227, 233, 282 GRPA (Group de Rapporteurs sur la Pollution de l’Air) 202, 204 H Halogenated hydrocarbons 115, 118, 124 Heavy duty vehicles (HDV) 238, 239, 241, 244–246, 249 Heavy methals 165 Helium (He) 93 Heterogenous air fuel mixtures 43, 51, 73, 81, 257 Highway cycle 230 Homogenous air/fuel mixtures 43 Homogenous diesel combustion 80 Hybrid drive 71, 97, 234 Hydrocarbons (HC) 9, 13, 19–21, 23, 28, 35, 36, 73, 79, 81, 88, 97, 113, 116, 117, 140, 141, 147, 152, 177, 184, 186, 187, 192, 193,
196, 198–200, 205, 209, 211, 214, 219, 223, 224, 239, 245, 261, 263, 264, 267, 271, 274, 278 Hydrogen (H2) 93, 99, 101, 102, 234, 277, 281, 283, 285, 287 I Ignition 32 Ignition delay 22, 260 Ignition timing 27, 32, 33, 39, 198 Immission 114, 115, 128, 130, 132, 137, 158 Injector nozzle 59, 62, 74, 75, 271, 274 Injection pressure 52 Injection system 75, 76, 80, 89 Inlet manifold 55 Inline pump 57 Inspection and maintenance (I/M) 178 Intake manifold 27, 33, 34, 39, 43, 65 Intake ports 53, 65 Intake stroke 25, 34 Internal combusiton engine 18, 91, 256, 268, 278, 280, 282–284, 287 In-use testing 178 Iron (Fe) 163, 166, 167, 170 ISO 14001 171 K Ketones 21, 179 Knock limit 39 Knocking combustion 34, 258, 259, 270 Kyoto 11, 227, 229, 234 L LA4-cycle 191, 211, 217 Lambda 27, 28, 30, 80, 82, 130 Lambda probe 39 LDT (light duty trucks) 232 LDV (light duty vehicles) 232 Lead (Pb) 38, 113, 165, 166, 170, 186, 204, 218, 273 Lead-acid batteries 95 Lean burn 34, 36, 41, 43–46, 73, 81, 82, 84, 272 Lean limit 32, 34, 36, 41 Lean mixture 20, 31, 32, 34, 35, 38, 40, 41, 43, 46, 205 Leer engine 94 Life cycle assessment (LCA) 171, 274, 275 Limites exhaust gas components 113 Liquid petroleum gas (LPG) 277, 279, 285 Low emisison vehicle (LEV) 179, 180, 193, 194, 224, 244 Lubricating oil 22, 23, 54
292 M Marine diesels 67, 68, 89, 91 Mean effective pressure (pme) 27 Mercury (Hg) 165, 170 Methane (CH4) 115, 118, 121, 146, 179, 192, 277, 278 Methanol (CH3OH) 102, 179, 185, 277, 279, 283, 285, 287 Methyl bromide (CH3Br) 124 Methyl chloride (CH3Cl) 124 Methyl esters 105 Mineral oil 19, 102, 256, 257, 276, 283 Misfiring of combustion 32 Mixture control 22, 39 Mobility 5, 6, 9, 110 Motor vehicle emissions group (MVEG) 249 Multi Point Injection (MPI) 269 N National ambient air quality standards (NAAQS) 178 Natural gas (LNG, CNG) 102, 277–279, 283, 285 Natural sources 114, 120, 121, 125, 134, 227 Naturaly aspirated engines 29 Nature 1, 3–5 NDIR (Non dispersiv infra red) 209, 211, 214 NEDC (New European Driving cycle) 217, 219, 220 Nicolaus Augustus Otto 25 Nitrates 45 Nitrogen (N2) 20–22, 38, 46, 114 Nitrogen dioxide (NO2) 22, 38, 116, 117, 130, 131 Nitrogen monoxide (NO) 22, 38, 73, 116, 117, 130, 131 Nitrogen oxides (NOX) 9, 13, 20, 21, 28, 30, 31, 35, 36, 44, 68, 73, 75, 76, 88, 113, 117, 125, 130–132, 140, 141, 146, 147, 152, 177, 179, 180, 184, 187, 192, 193, 196–200, 205, 211, 214, 224, 229, 233, 239, 244, 245, 261, 263, 264, 267, 274, 283 Nitrous oxide (N2O) 22, 44, 115, 125, 126, 146 Noble metal catalyst 38, 41, 44, 45, 271 Noise 8, 9, 50, 55, 63, 67, 68, 74, 158–160 Non methane hydrocarbons (NMHC) 129, 130, 146, 239 Non methane organic gases (NMOG) 118, 179, 180, 194, 197, 224 Non-ferrous metals 166, 167 Non-selective catalyst reaction (NSCR) 83 NOX storage catalyst 45, 46, 83, 84, 272
Subject Index O Octane number (ON) 258, 259, 269, 274, 282, 283, 287 Octane requirement 32, 259, 284 Off-road vehicles 127 Olefins 21, 23, 211, 261, 270 On board diagnostic (OBD) 88, 178, 184, 220, 224, 236–238, 246, 249, 250, 273 Operating parameters 30, 74 Organic fuels 22 Organis compounds 177, 179 Otto cycle 26, 92 Otto engine 19, 25, 27, 30, 34, 41, 141, 218, 257, 258, 264, 272, 277, 284 Oxidation catalyst 38, 41, 44, 63, 81, 82, 87, 88, 245, 272 Oxigen (O2) sensor 39, 45, 81, 196, 236, 262, 271, 273, 279 Oxygen (O2) 2, 20–22, 31, 99, 101, 261, 273, 282 Ozone (O3) 117, 118, 144, 151, 177, 179 Ozone formation potential (OFP) 117, 118, 144, 151, 152, 179, 266 Ozone hole 9 Ozone layer 115, 116 Ozone smog 9 Ozone-precursor 118, 179 P Palladium (Pd) 38, 39 Paraffins 21, 23, 211, 261, 283 Partial zero emission vehicles (PZEV) 180, 182 Particle filter 81, 84, 87, 89, 197, 214, 241, 245, 272 Particulate matter (PM) 13, 22–24, 73, 76, 79, 88, 117, 134, 147, 180, 184, 186, 192–194, 196, 199, 200, 205, 214, 224, 233, 239, 241, 244, 250, 264, 267, 271, 273, 274, 282 Passenger cars 6, 9, 25, 48, 88, 110, 113, 121, 137, 138, 140, 144, 145, 158, 164, 230, 233, 236, 239 Petrol engine 48, 63 Phosphorus (P) 38 Photooxidants 113, 195 Photosynthesis 105, 116, 281 Piston 18, 21, 50, 51, 53, 54, 91, 92 Plastic 164, 166–168, 170 Platinum (Pt) 38, 101, 103 Pollutants 9, 116, 261, 266, 268, 274, 277, 284 Polyciclic aromatic hydrocarbons (PAH) 266, 267, 269, 270 Portliners 36, 37
Subject Index
Power output 27, 29, 32, 34, 36, 39, 43, 62, 65, 67–69, 74, 259, 278 Power units 15, 25, 53, 91 Pre-catalyst 77 Pre-chamber engines 63, 75, 200, 266 Precious metals 83 Pre-injection 65, 76, 77, 80 Propane 277–279 Propulsion system 25, 91 R Rapseed methyl ester (RME) 282 Reciprocting piston engine 25, 91, 93, 94, 97 Recycling rate 165, 169 Recycling 125, 163, 164, 166–171, 281 Reduction catalyst 38, 241 Reformer 103, 283 Reformulated gasoline 262, 282 Regulated emissions 184, 204, 262, 270, 271 Renewable energy 234, 280–282 Residual gases 20, 36 Rhodium (Rh) 38 Rich mixture 31, 38 Road traffic 7, 8, 10, 12, 110, 112, 113, 121, 127, 128, 134, 137, 158, 284 Road transport 48 Rotary-piston engine 92, 93 Rudolf Diesel 48 Running losses 224 S Saturated hydrocarbons 261, 267, 270, 283 Scrap 163, 166, 167 Secondary air injection 36–38 Seiliger Process 48 Selective catalyst reaction (SCR) 83, 84, 88, 89, 241 Self ignition 22, 48, 50, 51 257, 264 Seven (7)-mode cycle 192, 211, 216 SHED (Sealed Housing for Evaporative Emission Determination) 222–224 Shredding 166–168 SiNOX 83, 84 Smog 176, 177, 211 Smoke 22, 80, 176, 186, 200, 214, 241, 271 Society of automotive engineers (SAE) 230 Solvents 129 Soot 22, 23, 73, 81, 137, 186, 246, 271, 279 Spark ignited engine 53, 63, 69, 71, 74, 82, 257 Specific fuel consumpiton 27 Specific work (we) 27, 259 Squish 34
293 Steam engine 18, 48, 91, 93, 94 Stirling engine 91, 93 Stoichiometric mixtures 20, 26, 30, 31, 39, 41, 43–45 Stratified charge engine 43 Sulfates 23, 73, 82, 273 Sulfur 21, 23, 38, 46, 73, 84, 113, 239, 241, 244, 267, 269, 272, 274, 282–284, 287 Sulfur hexafluoride 118 Summer smog 112 Super ultra low emission vehicle (SULEV) 180, 194, 224, 227 Sustainable development 171 Swirl 34, 53, 63, 65, 74, 79, 88 Swirl chamber 75 T 10-mode test 200 Technical guideline for air quality 129, 134, 137 Telematic 12 Test procedure 199–201 Thermal efficiency 94 Thermal reactor 36, 37 Three-way catalyst 38, 39, 41, 43–45, 82, 103, 114, 126, 141, 184, 196, 199, 217, 262, 263, 273, 279 Torque 30, 32, 34, 41, 43, 69 Toxicity 116 Trace gases 113, 115, 118 Trace substances 115, 116 Traffic 6–9, 11, 12, 115, 160, 217, 268, 275 Traffic management 11 Transport 5, 7, 115, 227, 233, 283 Transportation 5, 11, 12, 15, 18, 19, 48, 284, 285 Trucks 6, 67 Tumble 34 Turbocharged engines 30, 55, 56, 63, 69, 77, 88 Two-stroke 54, 68, 91, 92, 130 U Ultra low emission vehicle (ULEV) 179, 180, 193, 194 UBA (Umweltbundesamt, German Environment Agency) 203 Unburnt hydrocarbons (HC) 20, 21, 23, 30, 31, 41, 83, 113, 282 Underfloor catalyst 77 UNECE 130, 131 Unit injector 57 Unregulated exhaust gas components 113, 250, 266, 270 Urea 46, 83
294 UTAC (Union Techniques de l’Automobile du Motorcycle et du Cycle) 218 Utility vehicle 67, 87, 89 V Valve timing 33 Vapor pressure 264 VDA ( Association of German Automobil Makers) 146, 201, 202, 217, 218, 229 VDI (Association of German Engineers) 201, 218 Vegetable oil 277, 280–282 Volatile organic gases 118, 130 W Wankel engine 91–93 Waste 164, 167, 168, 171
Subject Index Water steam (H2O) 20, 281 Water vapor (H2O) 44, 81, 99, 101, 102, 105, 114, 115, 120, 281, 283 Well to wheel analysis 274 WHO (World Health Organisation) 129, 134 Working cylce 25, 41, 51 Working medium 94 World Wide Fuel Charter (WWFC) 269–271, 284 Z Zeolites 45, 46, 83 Zero-emission vehicle (ZEV) 95, 179, 180, 182, 185, 193, 194, 206, 224, 227, 250