Biological and Chemical Degradation of Azo Dyes under Aerobic Conditions ,

Jack T. Spadaro

B.S., Worcester Polytechnic Institute, 1987

A dissertation submitted to the faculty of the

Oregon Graduate Institute of Science & Technology

in partial fulfillment of the

requirements for the degree

Doctor of Philosophy

in

Biochemistry

July 1994 The dissertation "Biological and Chemical Degradation of Azo Dyes under Aerobic Conditions" by Jack T. Spadaro has been examined and approved by the following Examination Committee:

Dr. V. Renganathan Advisor and Associate Professor

Dr. Michael H. Gold Institute Professor

Dr. David R. Boone Professor

Dr. Paul G. Tratnyek Assistant Professor Dedicated to my parents, Richard and Vera,for their depotion to my upbringing and education, and also to all those in the world who strive to make a positive difuence ACKNOWLEDGMENTS

Foremost, I wish to thank my scientific mentors, Dr. V. Renganathan, at the Oregon Graduate Institute, and Dr. William D. Hobey, at Worcester Polytechnic Institute, for their fine and extensive efforts in my training. From OGI, I must also thank: Dr. Michael Gold, for his great interest in and guidance of my work; Dr. James Pankow, for the use of his cryofocusing GC-MS system; Drs. David Boone and Paul Tratnyek for their advice as members of my thesis committee; Lorne Isabelle and Gerry Boehme, for their good natures and their abilities to keep vital instrumentation functioning; Nancy Christie, for her excellent secretarial and management skills, and for her friendship; Drs. Muralikrishna Chivukula, Hiro Wariishi, and Khadar Valli, for sharing so many laboratory techniques and tricks, as well as for their interest in my work. Also of note are the many fine laboratory colleagues and support personnel that I have spent long hours with at OGI, especially Dr. Wenjun Bao, with whom I had the pleasure of sharing a laboratory for more than 5 years, Ciro DiMeglio, a friend through my days at WPI and OGI, and Bruce Godfrey. There have been many teachers throughout my life, especially in school and Scouting, whose understanding, support, and wisdom have meant a great deal to me. They include Mr. Reynard Bums, Mr. Theodore Scalzo, and Mr. David Kling. On the personal front, my sister Nancy, my entire extended family, and many close friends have all contributed greatly to my situation. I also wish to thank Linda Steinle, my companion, for her wonderful support these past four years. TABLE OF CONTENTS

List of Tables viii List of Figures ix Abstract xi

Chapter One - Introduction

I Azo Dyes - General Aspects A. Historical development B. Azo dye synthesis C. Azo dye classes, structures, applications

II. Azo Dyes - Environmental and Health Aspects A. Introduction of dyes to the environment 1. Dye manufacture and application 2. Waste dye treatment 3. Azo dye pollution B. Azo dye metabolism 1. Bacterial metabolism a. Metabolism under anaerobic conditions b. Metabolism under aerobic conditions 2. Mammalian azo dye metabolism

III. Fungal Degradation of Wood and Pollutants A. The fungus Phanerochaete chrysosporium B. Wood C. Physiology of lignin degradation D. Fungal degradation of pollutants

TV. Peroxidases A. Peroxidase history and the catalytic cycle B. Peroxidases and lignin degradation C. Peroxidases and pollutant degradation

V. Chemical Oxidation A. Advanced oxidation processes B. Fenton chemistry C. Mechanisms of chemical oxidation D. Pollutant degradations

VI. Thesis Outline

VII. References Chapter Two - Degradation of azo Dyes by the Lignin-degrading Fungus Phanerochaete chrysosporium

I. Introduction

II. Materials and Methods

III. Results

ZV. Discussion

V. References

Chapter Three - Peroxidase-catalyzed Oxidation of Azo Dyes: Mechanism of Disperse Yellow 3 Degradation

I. Introduction

11. Materials and Methods

III. Results

ZV. Discussion

V. References

Chapter Four - Peroxidase-catalyzed Oxidation of Azo Dyes with Phenylazo Substitutions Generates Benzene

I. Introduction

II. Materials and Methods

DI. Results

IV. Discussion

V. References

Chapter Five - Degradation of Azo Dyes by Hydroxyl Radicals: Evidence for Benzene Generation

I. Introduction 11. Materials and Methods

m. Results

IV. Discussion

V. References

Chapter Six - Conclusion

I. Introduction

TI. Fungal Degradation of Azo Dyes

III. Azo Dye Degradation by Peroxidases A. Decolorization studies B. Proposed mechanisms

IV. Azo Dye Degradation by Chemical Oxidation

V. Future Work

VI. References

Biographical Sketch

vii LIST OF TABLES ms Chapter Two Table I. Mineralization of 14C-labeled azo dyes 51 Table II. Recovery of 14C label 53

Chapter Three Table I. Quantification of NDY3 degradation products 65

Chapter Four Table I. Benzene generated during dye degradation 78 Table II. Benzene and 12NQS produced during degradation of dye I1 under varying conditions 80

Chapter Five Table I. Optimization of Fe(III) and H202 concentrations for dye mineralization 93 Table II. Mineralization of azo dyes by Fe(III)/H202 94

viii LIST OF FIGURES

Chapter One Figure 1. Three common synthetic dye structures Figure 2. Common azo dye syntheses Figure 3. Structures for azo dye classes Figure 4a. Aerobic and anaerobic azo dye metabolism Figure 4b. Biological azo dye degradation using anaerobic and aerobic stages in succession Figure 5. Spruce lignin structure Figure 6. HRP and LiP catalytic cycle Figure 7. MnP catalyic cycle Figure 8. Lip oxidation of 1,2,4,5-tetramethoxybenzene Figure 9a. LIP, HRP, and Mn(III) oxidation of benzenesulfonic acids Figure 9b. LiP oxidation of 2,4,6-trichlorophenol Figure 10a. Hydroxylation of benzene by hydroxyl radical Figure lob. Ring-opening of benzene by Fenton reagent

Chapter Two Figure 1. Mineralization profiles for Solvent Yellow 7 and Disperse Orange 3

Chapter Three Figure 1. Products identified from dye degradation Figure 2. HPK chromatograms of NDY3 degradation products Figure 3. Mass spectra of deuterated acetanilide products Figure 4. Proposed mechanism for NDY3 degradation

Chapter Four Figure 1. Structures of the dyes used in HRP reactions Figure 2. Products from the degradation of dyes I and 11 Figure 3. Proposed mechanism for Solvent Yellow 14 degradation by HRP

Chapter Five Figure 1. HPLC analysis of Disperse Yellow 3 and Disperse Orange 3 degradation products 95 Figure 2. HPLC analysis of acid products from 4-phenylazoaniline 97 Figure 3. Volatiles production by four dyes during the first six hours of dye degradation 98,99 Figure 4. GC-MS analysis of volatile products from N,N-dimethyl-4-phenylazoaniline 100 Figure 5. Proposed mechanism for Fe(III)/H202 degradation of azo dyes 102

Chapter Six Figure 1. Structures of several sulfonated azo dyes tested for fungal degradation Figure 2. 14C-labeled sulfonated azo dyes mineralized by P. ch ysosporiurn Figure 3. Pathway proposed for HRP-catalyzed oxidation of Sudan I Figure 4. Products observed for peroxidatic oxidation of several sulfonated azo dyes Figure 5. Two mechanisms proposed for azo cleavage Figure 6. Proposed redox and hydrolysis reactions in enzymatic azo dye oxidations Figure 7. Photocatalytic degradation of two azo dyes ABSTRACT

Biological and Chemical Degradation of Azo Dyes under Aerobic Conditions

By Jack T. Spadaro

Oregon Graduate Institute of Science & Technology, 1994 Dr. V. Renganathan, Thesis Advisor

Azo dyes represent >SO% of synthetic industrial dyes. Azo dyes are recalcitrant to aerobic bacterial degradation. Reductive cleavage of the azo linkage under anaerobic conditions yields potentially carcinogenic aromatic amines. This thesis examines aerobic azo dye degradation by the white-rot fungus Phanerochaete ch ysosporium, by peroxidases, and by hydroxyl radicals. P. ch ysosporium, a lignin-degrading basidiomycete, extensively mineralized several hydrophobic azo dyes over a 12day period. All dyes were degraded most extensively in ligninolytic cultures. Hydroxyl, acetamido, nitro, and N-alkylamino substituents enhanced dye degradation. Disperse Yellow 3 (DY3, 2-[4'-acetamidophenylazo]-4-methylphenol),a dye mineralized by P. chysosporium, yielded acetanilide as a major metabolite during fungal degradation in cultures that produce lignin and manganese peroxidases (Lip and MnP). Degradation of DY3 by Lip, Mn(III)- malonate (a MnP mimic), and horseradish peroxidase (HRP) was studied. The major products were acetanilide, 4-methyl-1,2-benzoquinone,and dimerized DY3. A mechanism for DY3 degradation is suggested. Either Mn(III) or the H202-oxidized forms of the peroxidases oxidize the phenolic ring of the dye by two electrons, producing an azo-bearing carbonium ion. Hydrolytic azo cleavage forms the quinone product and an acetamidophenyldiazene intermediate. The acetamidophenyldiazene is oxidized by metal or oxygen to produce an acetamidophenyldiazenyl radical, which cleaves homolytically to acetamidophenyl radical and molecular nitrogen. The acetamidophenyl radical abstracts a hydrogen from the surroundings, yielding acetanilide. Consistent with this mechanism, dyes containing phenylazo substitutions were degraded to quinones and benzenes by HRP. Further support for the mechanism was obtained through deuterium labeling studies. Hydroxyl radicals, produced for 24 h by reaction of ferric nitrate and hydrogen peroxide at pH 2.8, degraded large amounts of hydrophobic azo dyes to CO2 and water-soluble compounds. Products included benzene, formed during degradation of phenylazo-substituted dyes, and aliphatic acids. A mechanism resembling that for the peroxidase-catalyzed degradation of azo dyes is proposed for dye degradation by hydroxyl radical.

xii CHAPTER ONE

INTRODUCTION

I. AZO DYES - GENERAL ASPECTS

A. Historical development

Man has been using colorants since prehistoric times, perhaps as long ago as the 25th century B.C (3). The earliest colorants were natural substances prepared from vegetable and animal sources, as well as inorganic pigments taken from the earth (1,3). Some of the earliest known colorants include indigo, Tyrian Purple, and alizarin. Indigo, known from at least 4000 B.C., came from two plant sources, indigo (found in Asia and the Americas) and woad (found in Europe). Ancient Tyrian Purple was an indigoid first extracted by the Tyrians from the sea snail Murex brandaris about 1600 B.C. Alizarin was an anthraquinone dye obtained from the madder plant by ancient Egyptians and Persians as far back as 4000 B.C. (3,14). It was not until the middle of the 19th century that detailed chemical investigations and syntheses of dyestuffs were initiated (1). Among the first examples of synthetic dyestuffs were: Mauveine, an oxidation product of coal tar anilines made in 1856; cyanine, the first polyrnethine dye, also made in 1856; and Fuchsine, a triphenylrnethane dye, synthesized in 1857. In 1858, the discovery of diazotization by P. Griess led to synthesis of the first azo dyes, which contain the azo structure Ar-N=N-Ar (1,4,15). The first commercial azo dyestuffs were Aniline Yellow (4-phenylazoaniline), produced in 1861, and Bismark Brown (a bisazodye), produced in 1863. The ensuing 135 years have seen incorporation of the azo structure into disperse dyes, acid dyes, direct dyes, mordant dyes, metallized azo pigments, bis- and polyazo dyes, and solvent dyes. The azo dyes are now the most important of all commercial dyes, representing well over 50% of dyes used (11). Two other heavily used dye types are anthraquinone dyes and triphenylmethane dyes (Figure 1) (3). Azo Dyes

Orange I

Anthraquinone Dyes x" 0 NHCH2CH2CH2N(CH3)3

Astrazon Blue FGL

Triarylmethine Dyes

Crystal Violet

Basic Heterocyclic Dyes (CH3)2N

ci +

Azure B

Figure 1. Four of the most commonly synthesized dye structures. Azo dyes span the color spectrum from yellow to navy blue, and have excellent extinction coefficients (12). The use of disperse dyes has seen the quickest growth over the past several decades due to the huge increases in the use of synthetic hydrophobic fibers and fabrics, especially polyesters, in the textile industry (10). More than 3000 azo dyes are reportedly in use as industrial colorants and in food stuffs (4,5). Azobenzenes (Ph-N=N-Ph) are all colored and represent the most stable forms of azo compounds (13). They are also generally considered to be of anthropogenic origin (2). But azobenzenes have been shown to be formed from enzymatic and photochemical oxidations of anilines (2,8,127). The related azoxy (R-N(O)=N-R) and hydrazo (R-NH-NH-R) compounds are found with more frequency in nature (2).

B. Azo dye synthesis

The synthesis of most azo dyes involves the diazotization of a primary aromatic amine, and coupling of the resulting diazonium salt to an electron- rich nucleophilic species (Figure 2) (1). Nitrous acid, formed in situ with a mineral acid and sodium nitrite, nitrosates the amine, which then rearranges to yield the diazonium salt (R-N=N+x). A nucleophilic aromatic compound such as phenolate or aniline is then added to the reaction mixture and an azo linkage is formed (12). Azo dyes typically exist in the trans form with ca 120' bond angles about the sp2 hybridized nitrogens (11). The coupling of diazonium to homocylic rings (to form carbocyclic dyes) and to heterocyclic rings (to form heterocylic dyes) occurs at positions with the highest electron density. Carbocyclic dyes predominate. The structures of the dye's aromatic components impart strong effects on the color, brightness, and bleachability of the azo dyes (11). Addition of electron-donating and -withdrawing substituents create longer chains of conjugation in the aromatic system, allowing for shifts in absorbance (11).

C. Azo dye classes, structures, applications

Azo dyes fall into several groups or classes (l,l5). Some of the most common classes are described below. The nomenclature for the dyes is generally based more on their application or properties than on the IUPAC structural names. Comprehensive reviews of azo dye structures and applications are available in references 1 and 11. Structures can be found in Figure 3. The disperse dyes are hydrophobic and practically insoluble in water. They are used to color nonpolar synthetic fibers, such as polyester and cellulose acetate. A common substructure found in these dyes is a nitrodiazobenzene coupled to an N-alkylated aminobenzene. The limited solubility of these dyes can be adjusted through changes in the N-alkyl substituents. Two examples of disperse dyes are Disperse Yellow 3 and Disperse Blue 79. Solvent dyes are even more hydrophobic than disperse dyes and are soluble only in non-aqeous solvents. Solvent dyes contain naphthalene in place of one or both benzene rings attached to the azo linkage, and often have relatively high molecular weights. Solvent dyes find use in petroleum products, waxes, plastics and other highly hydrophobic media. An example is Solvent Yellow 14 (Sudan I). The acid or anionic dyes, which contain sulfonate groups, are water soluble. They are used for dyeing wool under acidic conditions (pH2-6). Anionic dyes that are applied in the presence of complexing metal (mordant) and those that are applied in a premixed metal-dye complex (premetallized) also fall into the acid dye class. The metals commonly used as dye mordant are copper, chromium, and cobalt, and these metals complex the dyes through coordination at hydroxyl groups and azo nitrogens. A common acid dye is Acid Orange 7 (Orange 11). Direct dyes also contain sulfonic acids but are generally of a much larger size than other acid dyes. They are applied to cotton and other cellulosic fibers without the aid of a mordant additive. Electrolyte is used in the dye bath. The direct dyes often contain bisazo type structures (ie. as in benzidine dyes (R-N=N-Ph-Ph-N=N-R)) or bisazostilbene type structures (R- N=N-Ph-CH=CH-Ph-N=N-R). Some dyes also contain urea bridges between two azo dye molecules. An example is Direct Red 23. Under acidic conditions, amino substituents on dyes can be protonated to form ammonium salts. The basic (or cationic) dyes thus formed can be applied to synthetic fibers that contain large numbers of anionic sites. Under neutral or alkaline conditions, these same dyes share many characteristics with disperse dyes. An example is Basic Orange 2 (chrysoidin). -2xa) aw q paqysap saAp oze 30 sassep aw 303 santpnqs a~dums.€ ad!^ Pigment dyes often contain hydroxyl groups ortho to the azo linkage. These dyes are found in plastic resins, printing inks, and paints. The dyes are added to the media in one of three forms: as solid dye; as metal toners, ie. sulfonate or carboxylate salts of calcium or other metals; and as metal chelates where nickel or other transition metals are coordinated to azo linkage nitrogens and ortho hydroxyl groups. These dyes are found in paints, plastics, and printing inks. An example of a metal toner pigment dye is Pigment Red 57 (Lithol Rubine BK).

11. AZO DYES - ENVIRONMENTAL AND HEALTH ASPECTS

A. Introduction of dyes to the environment

1. Dye Manufacture and Application

The United States alone produced more than 127,000 tons of dyes in 1988 (15). Approximately, 3000 azo dyes are in use today, and represent more than 50% of all the dyes manufactured (15). It is clear that the manufacture and use of azo dyes may have a significant impact on the environment. Azo compounds contain many xenobiotic substitutions including azo, sulfonic acid, nitro, chloro and bromo functional groups (3). These groups make the azo dyes highly resistant to chemical, photochemical, and biological degradation, a desireable quality for dye fastness and durability (1,21). But the recalcitrance of the dyes to degradation is undesirable from an environmental standpoint. Several of the dye precursors, such as halogenated, methylated, or nitrated anilines and phenols are listed as priority pollutants (21). The dye manufacturing industries discharge these precursors into the effluent waste stream (16,173). For example, a treated effluent from a dye plant contained at least 40 identifiable dyestuff intermediates and precursors (16). At least one azo dye was also identified. It has been estimated that 10-15% of the dyes used in the textile dyeing industry are released into the waste effluent stream during washing and processing of dyed fibers (3,18,20,21). Of this dye waste, about 20% survives standard wastewater treatment processes and is released to the environment (3). Dyes found in printing inks, paints, and pigments are released to the environment through disposal of printed paper, painted materials made out of wood or metal, and pigmented plastics. Azo dyes are often present in textile mill effluents (19,20). Textile dyeing wastewater from carpet mills in the Coosa River Basin (in Georgia) contributed at least 7 disperse dyes and 7 acid dyes to the river even after passage of the effluent through waste treatment facilities. River water samples taken farther downstream showed only trace levels of dye, but mud samples taken from the river bottom contained at least 7 disperse and 8 acid dyes. The Yamaska River in Quebec, Canada, another site of textile mill activity, also contained disperse dyes that were found in the river water, adsorbed to suspended solids, and in sediments. Dye application also involves dye carriers, including such substances as phenols, phthalates, lignosulfonates, and chlorinated benzenes, as well as detergents (27). These substances also are released to the environment (27).

2. Waste Dye Treatment

Current methods for waste dye removal and/or degradation involve chemical, physical, and biological processes (21,23,27,31). Industrial wastewater treatment generally involves three or four stages: preliminary, primary, secondary, and tertiary. Preliminary treatment includes neutralization and mixing of the waste. Primary treatment involves processes such as screening, sedimentation, flotation, and flocculation, all designed to remove dirt and grit, undissolved chemicals, and fibers. Secondary treatment consists primarily of aerobic and/or anaerobic biological transformation processes, as well as adsorption to coagulants (ie. lime). The tertiary treatment step (also known as the polishing step) involves capturing or destroying chemical and colored wastes that have survived the first three stages. Dye removal or degradation usually occurs in the secondary and tertiary stages (27). Biological degradation of the dye wastes has proven to be quite a challenge (24,27,33,39,118). It has been demonstrated that, overall, azo and other textile dyes are not significantly biodegraded even after 42 days of activated sludge treatment (25,26). The color removal that does occur in these systems is mostly due to adsorption of the dyes to biomass (25). Many microorganisms exhibit strict substrate specificity requirements (24). Nitrites and nitrates, ubiquitous in many activated sludge treatment systems, have been found to drastically inhibit aerobic microbial reduction of the azo linkage (43). Sulfonate groups are effective barriers to biodegradation as they make the dyes less membrane permeable and therefore less available to the intracellular degradation processes (43). Anaerobic bacteria reduce azo dyes to aromatic amines, which are toxic and require further treatment (24,47). These difficulties in synthetic dye biodegradation are understandable since most dyes have been designed to resist oxidative breakdown by biological and physical means (21,25). Currently, efforts are underway to combine anaerobic (reductive) biological treatment with a second stage of aerobic biological treatment (42). Tertiary treatment processes, have received a great deal of study. Many physicochernical methods are available, including adsorption, chemical reduction, chemical oxidation, chemical precipitation (especially for heavy metal wastes), ion pair extraction, electrolysis, and photochemical destruction (21,28,30). Adsorption of dye to a variety of solids appears to be the main method in use today. Dye adsorption to solid supports has been extensively studied (21,22,29). Solid supports tested include cellulosic fibers, activated carbons and charcoals, woods, bentonite clays, chitin, silica gel, manganese oxides, and proteins such as hair and feathers (2932). Activated carbons, with varying pore characteristics, are the most commonly and effectively used adsorbents (21,27). In general, sorbents suffer from several disadvantages (32). Their adsorptive capacity is dependent on the types of dyes, carriers, and inorganics found in the effluent stream. The widely used disperse dyes are generally not retained by the carbon adsorbents, and low molecular weight and/or highly polar chemicals are retained quite poorly (23,27). Activated carbons do not respond well to regeneration, so their lifetime is limited. The adsorbed dyes usually are hard to remove and represent a solid waste problem. Chemical and photochemical oxidation processes have recently been explored. The oxidants commonly used are chlorine, bleach, ozone, chlorine dioxide, hydrogen peroxide, Fenton's reagent, and permanganate (21). Ozonation, chlorination remove color from azo dye wastes, but direct and disperse dyes react slowly (21). Fenton chemistry (iron(II)/hydrogen peroxide) and photochemistry (UV/hydrogen peroxide) appear promising. A drawback to the use of Fenton's reagent is that large amounts of iron hydroxide sludge are formed. The high UV absorbtivity of colored dye solutions is a major drawback to UV-catalyzed hydroxyl radical formation from hydrogen peroxide. Chemical reductants such as sodium hydrosulfite (dithionite) and stannous chloride (SnC12) remove azo dye color (21). However these reductions generate potentially toxic aromatic amines, which must then be dealt with separately.

3. Azo Dye Pollution

The introduction of azo dyes into the environment can have several serious consequences. As already noted, reduction of the azo linkage produces aromatic amines, many of which are toxic or mutagenic (47). In environmental systems, azo dye reduction occurs by biological, chemical, or photochemical means. Many bacteria, in both aerobic and anaerobic environments, can perform the reduction of azo dyes (5,128). Bacterial azo linkage reduction systems is discussed in more detail in a following section. Natural chemical redox processes which mediate azo dye reduction in anaerobic sediments, such as those found at the bottom of stagnant or brackish waterways, also appear to be common (115,116). An Fe(II)-Fe(III) redox system is suspected to play a major role in these reactions. Photocatalytic reduction processes are also suspected to cause amine formation from azo dyes. Two sulfonated food dyes were reduced by simulated sunlight to yield the suspected carcinogen 1-aminonaphthalene (128). This process reportedly involved removal of sulfonate and hydroxyl groups from the naphthalene component. Residual dyes, found in dye plant effluents that have already undergone wastewater treatment, also appear to photosensitize the formation of singlet oxygen (123). The singlet oxygen causes increased phenol oxidation and may also inhibit growth of bacteria and other organisms. Azo dyes have detrimental physiological effects on yeast and earthworms. Congo Red, a direct dye of the benzidine type, physically interferes with the proper assembly of cell walls in the yeast Saccharomyces cerevisiae (124). Dyes disposed of in the soil increase mortality of the earthworm Polypheretima elongata during conditions of decreased oxygen tension (125). At sublethal concentrations, the dyes induce morphological changes, such as constrictions and swellings, and cause avoidance behavior on the part of the worm, ie. the worm prefers not to burrow into contaminated soils. Limited bioaccwnulation of dyes in fish has been observed. Small lipophilic dyes (disperse dyes with a molecular weight less than 450) appear to be the most likely candidates for uptake and storage (126). Numerous studies have demonstrated the carcinogenicity of azo dyes and possible structure-activity relationships (47,52-54). These conclusions have been drawn from epidimeological studies of human populations exposed to azo dyes, and from in vivo studies with rats, mice, and hamsters (3). It has also been established that almost all azo dyes with carcinogenic potential are mutagens (53-57). Two "in vivo" mutagenicity tests have been prominently used. The Arnes test uses rat liver microsomal preparations to metabolize, or "activate", the azo dyes, and the resulting products are then applied to cultures of mutant Salmonella typhimurium (55,57). The rate of reversion of the mutant strain back to the native strain is indicative of the promutagenicity of the azo dye and its metabolites. The Ames test can also be run in the absence of the liver preparation, so that the mutagenicity of the azo dyes can be measured without prior biotransformation (58,60). Testing with a mouse embryo cell line (BALB/c-3T3), the second major type of "in vivo" test, yields an even better approximation of the cytotoxicity, mutagenicity, and carcinogenicity of different azo dyes towards mammals (53,54). The aromatic amine reduction products of azo dyes have also been extensively tested by the same methods (51,59-62,128). Most evidence indicates that azo dye reduction to aromatic amines is the primary route to observed azo dye cytotoxicity, mutagenicity, and carcinogenicity (46,128). Sulfonated and polysulfonated dyes and their reduction products generally have little or no toxic, mutagenic, and carcinogenic properties (46,53,54,60,61). Their inability to cross cell membranes, and consequently their lack of intracellular reduction to aromatic amines, has been suggested as the primary reason (43). B. Azo Dye Metabolism

1. Bacterial Metabolism

Bacterial metabolism of azo dyes takes place in both aerobic and anaerobic environments (21,39,40). However, the extent of degradation is quite different under these two conditions.

a. Metabolism under aerobic conditions

The most studied bacterial azo dye degraders have been the Pseudomonads, though degradation has been observed with a number of other bacterial strains (5,24,35,37). The main degradative pathway in the aerobic bacteria appears to be azo linkage reduction followed by aromatic amine metabolism (Figure 4a) (5,128). The azo reductions occur intracellularly and produce primary aromatic amines. The latter are often considered to be toxic, yet the organisms metabolize and detoxify them (40). The aromatic amines are catabolized via known pathways that involve hydroxlations and ring opening reactions (41). The azo reductases generally are induced by azo dyes. Purified azo redudases require NADH or NADPH for activity (5,24,37). Biochemical reduction by reduced flavins has also been observed (43). In order to effectively degrade azo dyes, aerobic bacteria often require long adaptation times (35,40). The structural selectivity of the bacterial azo reductases is generally very high, perhaps even including requirements for the cis or trans isomer of a dye substrate (35,40). And dye solubility, conferred by carboxyl or sulfonate groups, can dictate the rate of dye transport through the cell membrane (43). These factors often limit bacterial dye degradation.

b. Metabolism under anaerobic conditions

More than 30 azo dye degrading anaerobic bacteria have been isolated from soils, sediments, and intestinal microflora (5,44,128). At least 11 strains were found in the human intestine (37,44). Azo linkage reduction is the primary mode for dye degradation by these bacteria. aerobic or anaerobic metabolism

aerobic metabolism 1 only I

CO, or biomass

Figure 4a. Simplified scheme for the metabolism of azo dyes by aerobic and anaerobic organisms. R = H, OH, NH2. Anerobic bacterial azo dye reduction is a widely used method for decolorization of azo dye wastes (21). Generally, it is a non-specific process, and aromatic amines are the final products (39). Apparently, anaerobic bacteria possess very little ability to further catabolize these simple aromatic compounds (Figure 4a) (117,119). Recently, Haug et al. demonstrated azo dye mineralization by a bacterial coculture that passes through successive anaerobic and aerobic phases (Figure 4b) (42). Under anaerobic conditions, one bacterial strain has an efficient and fairly non-selective azo reductase system. It reduces the dyes to aromatic amines which can then be oxidatively degraded by both bacterial strains upon entrance into the aerobic stage. The degradation scheme described appears to combine the full advantages of the aerobic and anaerobic metabolic systems of these bacteria. Yet, the fact remains that the bacteria used in this system are strains that have been selected for their ability to aerobically degrade the specific amine products of this azo cleavage. This bacterial system can't degrade wastes containing a mixture of different dyes.

2. Mammalian Azo Dye Metabolism

Interest in mammalian metabolism of azo dyes first arose in the 1910's in response to the discovery that Orange I orally administered to dogs was reduced to sulphanilic acid. The discovery, in the early 1930fs,of antibacterial action of prontosil, an azo dye, spurred further interest in this area. It was soon found that the active antibiotic sulfanilamide was produced upon azo reduction of the prontosil (2). Azo dye metabolism in mammals and humans appears to be primarily reductive, with formation of aromatic amines (45,128). Reduction of azo linkages occurs in the intestinal microflora, and in the liver microsomes (37,44,46,48). Many azo dyes and aromatic arnines are inducers of the Phase I and Phase 11 carcinogen/drug-metabolizing enzyme systems (47,50). These include cytochromes P-450 and oxidative enzymes (Phase I) and quinone reductases (Phase 11). Early on it was found that water soluble azo dyes pass into the large intestine and are reduced there by intestinal microflora (128). The amines produced are then generally excreted from the body, though some less soluble amines may be absorbed through the intestinal lining. Less soluble dyes that COOH 1

Mordant Yellow 3 I Anaerobl c condl tion Mixed culture

Pseudomonas Mixed culture sp. BN9

Aerobic I condition C02 and biomass

Figure 4b. Simplified scheme showing how anaerobic and aerobic stages can be used in succession to biologically degrade azo dyes. Mixed culture BN6 refers to a bacterial consortium that has been adapted to growth on the 6- amino-2-naphthalenesulfonic acid under aerobic conditions, but can also function to reduce the azo linkage of Mordant Yellow 3 under anaerobic conditions. From Haug, et al. (42). are absorbed into the body before reaching the large intestine are processed in the liver. A major step in their processing is glucuronide conjugation to improve solubility (128). Metabolism of azo dyes by liver microsomes has been extensively studied (48,49,51). Reductive fission of the azo linkage appears to be a result of both enzymic (cytochromes P-450, NADPH-cytochrome c reductase) and non-enzymic (NADPH, NADH, FADH2) processes (2,48,49). The anaerobic transfer of electrons to the azo linkage occurs one electron at a time and is quite sensitive to the presence of oxygen (2). Oxygen can oxidize the transitional intermediates back to the starting azo compound. The aromatic amine products of azo dye reduction, as well as azo dyes that contain aromatic amine substituents, may have adverse biological effects (39,46,47). Aromatic amines are oxidized to aryl-hydroxylamines and - nitrosamines by enzymes in liver rnicrosomes, including cytochromes P-450, flavin monooxygenases, and peroxidases (47). Both of these oxidized arnine products have been shown to electrophilically attack and form adducts with guanine and adenine bases in DNA (47). Reduction in the intestine occurs mainly via intestinal microflora (2,128). At least 11 strains of azo-reducing bacteria have been isolated from human intestine (37,44). All strains require flavin and NAD cofactors for reductase activity. Extracellular azoreductases are active only anaerobically (44). Two aerobically active intracellular azo reductases isolated from Shigella dysenteriae Type 1 are NAD(P)H-requiring flavoproteins (containing FMN) (37). Only water-soluble dyes appear to be substrates for the intestinal bacteria.

111. FUNGAL DEGRADATION OF WOOD AND POLLUTANTS

A. The Fungus Phanerochaete chysosporium

Phanerochaete chrysosporium is a basidiomycetous white-rot fungus that has been isolated from rotting piles of wood chips (120). This filamentous fungus uses its long hyphae to invade the lumen of woody plants cells and non-selectively attacks all components of the wood (120). The fungus secretes oxalic and other organic acids into the surrounding wood to adjust the pH to about 3, a value near the optimum for its extracellular wood degrading system. Oxidative enzymes are secreted to degrade the lignin and hemicellulose that encrust the crystalline cellulose fibrils. Cellulose is the organism's main carbon source. Various cellulases are secreted and they liberate cellobiose and then glucose, the primary energy source of the cell, from the cellulose fibers. The lignin and hemicellulose are degraded cometabolically and cannot serve as a sole source of carbon for the organism. White-rot fungi are thought to have been degrading wood as far back as the Triassic period, at least 180 million years ago, as patterns of fungal wood decay similar to those found today exist in fossil woods from that time (120). Fungi are the major biological decomposers of wood (69).

B. Wood

Wood consists of three main polymers: cellulose; hemicellulose; and lignin (120). Cellulose is a sugar polymer consisting of P-0-1,4-linked glucose (cellobiose) units. It is the most abundant natural and renewable polymer in the biosphere. Hemicelluloses are branched sugar polymers consisting of a mixture of at least six different sugar monomers. The hemicellulose structures vary greatly by plant type. Lignin, the second most abundant natural polymer, is a phenolic free-radical copolymerizate of coniferyl, p- coumaryl, and syringyl alcohol precursors. Plant peroxidases are responsible for the random polymerization of the presursors. Lignin is also the most abundant renewable aromatic material in the biosphere, and many natural products are derived from it. These products include aromatic acids, aldehydes, and alcohols, as well as phenolics, alkyl-aryl ethers, aryl propanoids, and various biphenyls (120). The three important functions of lignin in wood are: 1) to serve as a rigid structural support; 2) to bind the wood fiber cell walls together; and 3) to protect the carbohydrate content of the wood (120). The lignin (up to 36% of the woody cell walls), in conjunction with the hemicellulose, forms a protective barrier around the cellulose. Microbial degradation of both the hemicellulose and the enclosed cellulose is severely curtailed, as lignin, a random, polyphenylpropanoid polymer, strongly resists biological attack (Figure 5). Lignin is slowly biodegraded, and in fact makes up much of the humic material found in soils. The complex structure of lignin dictates that HCOH I 3H

CH30r HOCHZ ~1: 0 I HCOH 1 *&,H $3 I CH30 POCH, HC- -0

HCOH C=O

Figure 5. A schematic diagram of spruce lignin showing the complex, random, polyphenylpropanoid structure. From Adler (143). only extracellular, non-specific, and non-hydrolytic systems can initiate lignin degradation (69,120). The white-rot fungi are the only organisms known that can extensively degrade all of the major polymers in wood (69,74,120). The so-called brown- rot and soft-rot fungi are able to only partially degrade lignin and hemicellulose. Lignin biodegradation by bacteria under anaerobic conditions is apparently non-existent, and degradation by aerobic bacteria is slow and limited. The white rot fungi, such as Phanerochaete chysosporium, fully degrade both the aromatic nuclei and the carbon-bearing side chains of lignin to CO2 and H20. Over the last 20 years, the study of lignin biodegradation has focused on the white-rot fungi, particularly P. chysosporium and its inducible lignin-degrading system (69,120).

C. Physiology of Lignin Degradation

Many studies of lignin degradation have been performed using the highly efficient lignin degrader Phanerochaete ch ysosporium (120). The initial mineralization studies, made using 1%-labeled natural lignins and 1*C-labeled lignin-like synthetic dehydrogenative polymerizates (DHP), showed that lignin degradation occurs as a cometabolic process requiring an energy source (ie. glucose), high oxygen levels, and the proper buffers (69,90,120). These studies also established that lignin degradation can be optimized by varying nutrient levels, culture additives, and growth conditions (90,120). Limiting amounts of nutrient nitrogen, carbon, or sulfur serve to spur the onset of lignin degradation, indicating that lignin degradation is a secondary (idiophasic) metabolic process (69,120). Veratryl alcohol (3,4-dimethoxybenzyl alcohol, VA), a secondary metabolite of P. ch ysosporium, accumulates in lignin-degrading cultures. The presence of VA appears to be necessary for efficient lignin degradation, but the exact physiological role of VA remains unknown (120-122). The metals copper and manganese, both naturally found in wood, also increase the lignin degradation (120). Lignin degradation proceeds at some distance from the fungal hyphae, suggesting that extracellular factors are involved (120). Studies with DHPs indicate that the lignin polymer (600-1000 kd) is oxidatively broken down into fragments of less than 1 kd. Dimeric lignin substructures such as diphenylpropanes and phenylpropane-aryl ethers (mw 200-300) are all degraded by the fungi (91,92). The 14C-labeled analogs of these compounds are converted to 14C02. These lignin model compounds are all degraded under lignin-degrading culture conditions. The fungi are now known to produce oxidative enzymes (specifically extracellular peroxidases) and an extracellular hydrogen peroxide-generating system (69,74,120).

D. Fungal Degradation of Pollutants

The lignin-degrading capability of P. chysosporium was soon recognized for its potential in bioremediation of aromatic priority pollutants (167). The degradation system is extracellular, apparently quite non-specific, and inducible. The fungus grows well in non-sterile, harsh environments, such as in pulp mill primary sludges and under UV irradiation, conditions that restrict the growth of most other fungi and bacteria (3638). The initial studies concentrated on finding chemicals that could serve as indicators of lignin-degrading activity (120). These compounds, including various phenols (ie. 2,6-dimethyl- and 2,6-dimethoxyphenol), were essentially lignin substructures. Several polymeric dyes were also tested. The fungus was successful at degrading or decolorizing these phenols and dyes. Degradation of polymeric dyes can be used to indicate lignin- and pollutant- degrading abilities of fungi. The polymeric dyes Poly B-411, Poly R-481, and Poly Y-606 are decolorized by the white rot fungus P. chysosporium only under lignin degrading culture conditions (64). A later study of other fungal species showed that only fungi with lignin-degrading capability were able to decolorize the dye Poly B-411 (65). P. chrysosporiurn has been shown to extensively decolorize seven triphenylmethane dyes, including Crystal Violet and Malachite Green, and the fluoresein dye Rose Bengal (66,67). Over the last decade, studies have greatly expanded the list of both aromatic and non-aromatic pollutants degraded and/or mineralized by P. ch ysosporium. The list includes polychlorinated biphenyls, chlorinated dioxins, halogenated phenols and anilines, nitroaromatics, chlorinated alkanes, cyanides, polycyclic aromatic hydrocarbons, polyethylenes, and sulphonated anilines and phenols (63,68,96). Veratrylglycerol-P-2,4- dichlorophenylether is a model for humic or lignin bound xenobiotic residues that might arise from herbicide (ie. 2,4-dichlorophenoxyacetic acid) degradation in the environment (92). It was also degraded by the fungus (107). Clearly, the non-specific nature of the lignin degradation system enables the degradation of these pollutants as well. The fungal degradation rates for polycyclic aromatic hydrocarbons (PAHs) have been correlated to fungal decolorization of the dye Poly R-478 (89). The degradation pathways in P. ch ysosporium for chlorophenols, nitrotoluenes, and chlorodibenzodioxins have been elucidated in detail (87,93,94). Extensive roles for peroxidases, nitro reduction systems, methylation systems, and quinone reduction systems are postulated in the catabolic pathways of these compounds (87,93,94). In nearly all cases, the degradation of pollutants and dyes has required culture conditions that induce the lignin degrading system of the fungus. However, some biotransformations and biodegradations have been shown to proceed under non-ligninolytic conditions, as well (95,97-99).

IV. PEROXIDASES

A. Peroxidase History and the Catalytic Cycle

Peroxidases occur ubiquitously in nature (71). They have been found in aerobic bacteria, algae, fungi, plants, and mammals. Peroxidases are also found extensively in hurnic materials and soils (78). The peroxidases were among the first enzymes ever to be isolated, with the first reports of peroxidase activity dating back to at least 1810 (70). Peroxidases are hemoglycoproteins that contain one mole of protoheme IX per mole of protein. The number and size of glycosylating groups can vary. Horseradish peroxidase (HRP) has a molecular weight of approximately 44,000 (70). Various lignin-degrading peroxidases isolated from white-rot fungi vary in molecular weight from 39,000 to 54,000 (120). The catalytic cycle of HRP, first suggested by H. Theorell in 1942, is viewed as the standard for peroxidases (Figure 6)(70,71,75). The peroxidase donates electrons to hydrogen peroxide, which is reduced to water. The oxidized peroxidases then reacquire electrons by oxidizing organic or inorganic substrates. Generally, all peroxidases can oxidize phenols and anilines (70). Figure 6. The standard catalytic cycle of horseradish peroxidase (HRP) and lignin peroxidase (Lip). Native state is at top left, compound I is at top right. Compound I1 is at bottom. AH2 is the substrate molecule that is oxidized during progression from compound I to compound 11, and from compound II back to native. From Dawson (75). The native enzyme is in the ferric (Fe(JII))state. Upon 2-electron oxidation by hydrogen peroxide, the heme assumes the ferryloxy (Fe(IV)=O) porphyrin 7c cation radical form, also known as compound I. One of the oxygens from the hydrogen peroxide is reduced to water, whereas the other oxygen atom is coordinated to the heme iron. Compound I oxidizes one equivalent of substrate by one electron and is converted to compound 11, a ferryl-0x0 species. Compound 3.1 oxidizes another equivalent of substrate, and the enzyme returns to its native ferric state. The oxygen atom is released from the iron as a molecule of water. The net reaction is one molecule of hydrogen peroxide yields two molecules of water and two oxidized substrate molecules. The stoichiometry of the reaction can vary, as compound 11 can oxidize a second time the same molecule oxidized by compound I, or it can oxidize a completely separate substrate molecule (71). Ortiz de Montellano, et al., have proposed that the location of substrate oxidation by HRP compound I is at or near the heme edge, and not near the iron-0x0 group at the heme center (76,77). Peroxidases comprise a large family of related proteins, and their individual roles have been finely tuned to the needs of their host organisms. The two "standard peroxidases to which other peroxidases have been structurally compared are horseradish peroxidase (HRP) and yeast cytochrome C peroxidase (CCP). HRP and turnip peroxidase both share very high amounts of secondary and tertiary structure with CCP, as shown by X-ray crystallography (72). Yet, the amino acid sequences of the three peroxidases share less than 12% identity. Protein and gene alignment studies now strongly suggest, in fact, that the bacterial, plant, and yeast peroxidases have all descended from a common ancestral gene (72,73).

B. Peroxidases and Lignin Degradation

The breakdown of lignin by white-rot fungi and other organisms relies primarily on the activity of peroxidases (69,74,120). In fact, the action of lignin-degrading peroxidases is often described as "enzymatic combustion" due to the peroxidases' relatively non-selective oxidative activity, and to the kinetically favored, structurally-dependent depolymerization that occurs once the lignin substructures are oxidized by the peroxidases (69). Investigations in the early 1980's on the white-rot fungus P. chysosporium led to the discovery of two peroxidases and a hydrogen peroxide generating system, all found in the extracellular medium. The enzymes lignin peroxidase (Lip) (often called ligninase) and manganese peroxidase (MnP) are induced only under lignin- degrading culture conditions (see sections above). Initially, many studies were made of veratryl alcohol and other lignin substructures (phenyl and phenoxypropanoids), and it was demonstrated that LiP and MnP oxidations yield degradation products similar to those found in whole culture experiments (120). Characterization studies show that the fungal peroxidases Lip and MnP are hemoglycoproteins with molecular masses of 41,000 and 46,000, respectively (74). Both enzymes contain one mole of heme per mole of protein. They both follow catalytic cycles that are in many ways similar to that for HRP (102,103,120) (Figures 6,7). Surprisingly, MnP was found to be a unique type of peroxidase. MnP can use Mn(II) as a substrate for compounds I and II, and it strictly requires Mn(I1) as an electron donor during the transition from compound II to native enzyme (103). Chelated Mn(II) is oxidized to chelated Mn(III), and Mn(IlI) chelate diffuses away from the enzyme and non-selectively oxidizes a substrate molecule. Mn(I1) and Mn(II1) require chelators, such as malonate, pyrophosphate, or the physiologically produced oxalate, to remain stable in solution. LiP is sensitive to hydrogen peroxide (H202), especially in the presence of poor substrates. It can be deactivated when the substrate is unable to quickly reduce the enzyme to the native state (100,101). Under conditions of excess peroxide, Lip compounds Ill and III* are formed. If the enzyme is allowed to remain as compounds 111 and ID* the heme is degraded, resulting in Lip inactivation. Veratryl alcohol rescues the enzyme from inactivation by returning compound III to the native state, apparently by causing release of superoxide from the heme iron (101,121). The role of VA, a fungal secondary metabolite, is a topic of controversy (120-122). VA appears to enhance the degradation of otherwise poor substrates by Lip. Two theories have been put forth to explain the role of VA in these reactions. The first is that VA saves Lip from inactivation by excess H202 (121). Addition of VA to reactions containing poor substrates, such as anisyl alcohol, enhances the oxidation of these substrates even in the presence of hydrogen peroxide levels that would normally lead to ompound I e 14 = 0 [P-f-I

Figure 7. The catalytic cycle for manganese peroxidase (MnP). Mn@) can be oxidized by either compound I, in place of another substrate, or by compound II, where it is required in order to return the enzyme to the native state. From Gold, et al. (74). inactivation of Lip. Veratryl alcohol appears to protect LiP in two ways: by serving as a very good reductant for LiP compound II; and by converting compound III back to native, as detailed above (101,121). The second theory suggests that VA serves as a radical cation mediator, carrying oxidation equivalents away from Lip (122). The veratryl cation radical is proposed to diffuse into the smallest cavities in the lignin polymer matrix, and to oxidize the lignin there. A piece of the lignin that has been oxidatively sheared from the bulk of the matrix can then diffuse back to the LiP or MnP to undergo further oxidation directly by the enzyme. This theory parallels the activity of chelated Mn(III) that can freely diffuse away from MnP. No direct proof for the existence and stability of a veratryl cation radical mediator in the degradation of lignin or model compounds has been obtained. LiP appears to have higher oxidation potentials than peroxidases with similar heme structure, such as MnP and HRP (79-82). The relatively electron-deficient character of the LiP heme has been put forth as one explanation for the relatively high oxidation potential (81). Electrochemical studies established that LiP is able to oxidize aromatic systems with half-wave oxidation potentials (E1/2 vs saturated calomel electrode) of up to 1.49 V, such as 1,3,5-trimethoxy- and 1,2- and l,4-dimethoxybenzenes (79). HRP, MnP, and pyrophosphate-chelated Mn(III), on the other hand, can oxidize only the more electron-rich compounds such as 1,2,4-trimethoxybenzene (El12 of 1.12 V or lower) (80). Lip, MnP (via Mn(III)), and HRP all oxidize the highly electron-rich compound 1,2,4,5-tetramethoxybenzene (El /2 of 0.81 V) to an aryl cation radical, and the product, 2,5-dimethoxybenzoquinoneis formed (Figure 8) (79,80). Further electrochemical studies have shown that LiP can oxidize polyaromatic hydrocarbons (PAH) with relatively high ionization potentials, such as benz[a]anthracene, whereas HRP is limited to oxidizing those PAH with lower ionization potentials, such as benzo[a]pyrene (80,82).

C. Peroxidases and Pollutant Degradation

Horseradish peroxidase and other peroxidases generally have broad substrate specificities (71). This property led to the suggestion that peroxidases be used as an oxidative method for the treatment of chemical wastes (83-85). Figure 8. The formation of an aryl cation radical during Lip, MnP, or HRP oxidation of 1,2,4,5-tetramethoxybenzene. 2,5-dimethoxy-1,4-benzoquinoneis the product. Electron spin resonance (ESR) can detect the aryl cation radical. From Kersten, et al. (79). The enzymes could be used to non-specifically polymerize aqeous chemical wastes into insoluble forms, to catalyze the binding of chemical wastes to solids in an aqueous waste stream, or to immobilize the wastes through oxidative polymerization with humic materials in the soil. The non- selective oxidizing abilities of the fungal lignin-degrading peroxidases suggested usefulness in biotransforming or immobilizing some of the most biologically recalcitrant aromatic pollutants, such as PAH and chlorodibenzodioxins. Nearly all of the pollutants degraded by P. chysosporiurn are substrates for Lip and/or MnP (63). The substrate range of Lip and/or MnP includes most phenol and aniline type compounds. LiP is somewhat sensitive to phenolic compounds (120). MnP, on the other hand, is more similar to HRP in that it easily oxidizes phenolics (120). These compounds may contain several types of substituents, including sulfonic acid, nitro, methyl and methoxyl groups, as well as halogens (63,68). The nitro, sulfonate, and halogen substituents are recognized for their abilities to make these xenobiotic compounds highly recalcitrant to biological degradation. Recent studies show that Lip, MnP, and HRP have the ability to desulfonate, denitrate, or dehalogenate these various substrates (Figures 9af9b) (68,93,94). All of these reactions are proposed to proceed through aryl cation intermediates, where the cation forms at the carbon bearing the xenobiotic, electron-withdrawing substituents. Nucleophilic attack by water at the cation leads to formation of a tetrahedral intermediate, which collapses, expelling the xenobiotic substituent. The peroxidases are performing a unique task that transforms the pollutants into much less recalcitrant species, such as quinones. The very high oxidation potential of LIP enables it to oxidize pollutants previously thought recalcitrant to peroxidase degradation (81). Among these are polycyclic aromatic hydrocarbons, chlorinated dioxins, and coal polymers (82,8648). All of these compounds contain electron-rich aromatic systems, but no easily oxidizable substituents. Azobenzenes have been formed by the reaction of anilines with peroxidases. The first demonstration for the peroxidative formation of azobenzene from aniline was by Daniels and Saunders in 1953 (9). Extensive study of biological conversion of anilines to azobenzenes was made by Bartha and colleagues in the late 1960's and early 1970's (6-8). These studies Figure 9a. Desulfonation mechanisms for 4-hydroxybenzenesulfonic acid and sulfanilic acid during oxidation by Lip, HRP, and Mn(1II)-malonate (a MnP mimic). From Mur alikrishna and Renganathan (68). C I -6- C I ligninose + H2O2 FH+,c 1 e-

ligninase + e- "202 C l.

Figure 9b. Dechlorination mechanism during oxidation of 2,4,6- trichlorophenol by Lip. From Hammel, K.E., and Tardone, P.J. (1988) Biochemistry 27, 6563-6568. demonstrated that chloroanilines, formed in the soil as hydrolytic degradation products of acylanilide herbicides (i.e. propanil, N-(3,4- dichloropheny1)-propionamide), were acted upon by soil-borne peroxidase activities to form azobenzenes. In laboratory experiments, acylanilide herbicides applied to soil at levels of 500 ppm were converted in 2030% yield, whereas in field studies the conversion was -4.5% (7). A later study showed that the purified peroxidase(s) and aniline oxidase(s) associated with the soil- borne fungus Geotrichum candidum were able to perform this activity, in vivo and in vitro (8).

V. CHEMICAL OXIDATION

A. Advanced Oxidation Processes

Advanced oxidation processes (AOPs) generally involve formation of hydroxyl radical (*OH),a highly reactive species. *OHare strong, fast-acting oxidants at standard temperatures and pressures (129). AOPs encompass inorganic, photochemical, electrochemical, and physical methods that catalyze the oxidative degradation of solid and liquid wastes (129). The methods include ozonation, Fenton chemistry, photocatalysis, ultrasonic irradiation, and wet air oxidation (129-131,133-137). These methods are often combined to achieve faster and more complete degradations. Ozonation has been used in conjunction with chlorine, H202, Fenton reagent, and ultraviolet (UV) irradiation. Fenton reagent has been combined with electrochemical oxidation, and UV irradiation. Photoirradiation (usually with W) is combined with porous metal oxide catalysts (Ti02), or as noted above. The use of AOPs is steadily rising as the need to destroy wastes, as opposed to transfer and storage of wastes, assumes importance in industry. AOPs appear to have several benefits over conventional processes. AOPs can generally be performed under conditions of standard temperature and pressure. AOPs also can be performed directly on many aqueous wastes; no concentration of waste to dry forms is necessary. Hydrogen peroxide, a common oxidant used in many AOPs, is a very inexpensive chemical commodity.

B. Fenton Chemistry

The reactions now recognized as Fenton chemistry were first reported by H.J.H Fenton in the late 19th century. Fenton noticed that ferrous salts mixed with hydrogen peroxide in solution oxidized tartaric acid to yield dihydroxy maleic acid (104). Later studies showed that this oxidation occured with other hydroxy acids, sugars, and aromatic compounds (108). Haber and Weiss, in the 1930'~~postulated that the active species in the Fenton reactions is hydroxyl radical (*OH)formed via ferrous iron oxidation by hydrogen peroxide (eq. 1) (104,105). More recent investigations have suggested that ferryloxo ion (FeOH?+ c---> Fe02+) may also be responsible (eq. 2) (105). Kinetically, there appears to be no difference between the two active species (105). Several redox active transition metals can be used to generate hydroxyl radical from hydrogen peroxide, including Cu(I), Ti(m), Cr(II), and Co(II) (104).

Fez+ + H202 ----> F$+ + *OH + OH- (1)

Fe2+ + Hz02 ----> F~OH++ OH- (2)

Recently, hydroxyl radicals have been generated using catalytic amounts of ferric (Fe(lII) salts and high H202 concentrations (106). It is known that Fe(II) is quickly converted to Fe(1II) during Fenton reactions and that the equilibrium "steady-state" concentration of Fe(II) is actually very low (106). After the initial reaction described in eq. I., the following reactions have a large influence on iron speciation:

Reaction 1 is much faster than reaction 4, so the [Fe(III)] is >> [Fe(II)]. The catalytic nature of the ferric iron has made this Fenton-like system more attractive for pollutant degradation than the traditional Fenton chemistry (107). Another major factor is that the Fe(IU) appears to catalyze oxidation effectively without consuming the equivalents of Hz02 that Fe(I1) would use to form Fe(III) (139). The Fenton and Fenton-type reactions are done under acidic pH conditions, generally in the range of pH 2.5-3. In employing the Fe(III)/H202 system to destroy 2,4-dichlorophenoxyacetic acid, anions play a large role in the reaction rate (106). It was found that perchlorate and nitrate are the favored anions, followed by chloride. Sulfate does not allow the reaction to proceed at any appreciable rate. In contrast, reactions with Fe(I1) are very often performed well in solutions of dilute sulfuric acid.

C. Mechanisms of Chemical Oxidation

Because *OHhas an oxidation potential (E, = 2.02 V) second only to that of molecular fluorine, *OHis non-specific in its reactions and reacts with organic compounds at nearly the diffusion controlled rate (108,111,136,138). Numerous reactions are possible (108). Hydrogen atom abstraction by *OH forms alkyl or aryl radicals. *OHalso can remove electrons from phenolates to form phenoxy radicals, and electrons from aryl radicals to form aryl cations. Radicals that are formed can condense with a *OH to form hydroxylated species. Addition of *OHdirectly to unsaturated carbon-carbon bonds also occurs. In benzenes, the intermediate formed is a hydroxycyclohexadienyl radical (Figure 10a) (108,109). This radical undergoes redox reactions to yield bisphenol, phenol, biphenyl, benzene, benzoquinone, dihydrobenzoquinone, and other products (108,109). Phenol can be hydroxylated further to a catechol (140) . A recent report on Fenton reactions proposed that catechol is complexed by Fe(III), and that in the presence of H202 intradiol ring cleavage occurs to form a cis,cis-muconic acid (140) (Figure lob). Carboxylic acids, such as muconic and maleic acids, are also targets for -OH attack (108), and oxidation and decarboxylation continues until the acids are totally degraded to carbon dioxide (C@) (106,108,111,139,141,142). Cyclohexadienyl radical

Figure 10a. Initial steps in benzene oxidation by hydroxyl radical, to form a phenol. The cyclohexadienyl radical that forms can be detected by ESR From Walling (108). Dihydroxybenzyl Radical

Ferric OH / Q0>e+3

Ferric Phenolate

0

Ferric Hydroperoxycatechdabe cis-,cis4 uconic Acid

Figure lob. A proposed mechanism for phenol degradation during oxidation by Fenton reagents. The ring opening to the muconic acid is the critical step in phenol degradation. From Potter and Roth (140). D. Pollutant Degradations

Fenton and Fenton-like chemistry have been extensively studied as a possible method for degrading potentially harmful chemicals such as formaldehyde, perchloroethylene, chlorobenzene, naphthalene, coumarin, chlorophenols, nitrobenzene, toluene, and N,N-dimethylaniline (106,107,109- 114). The Fe(III)/H202 system has recently been reported to mineralize the two herbicides 2,4-di- and 2,4,5-trichlorophenoxyacetic acid (2,4-D, 2,4,5-T), formaldehyde, and methanol (106,139). Ultraviolet irradiation of the reaction is reported to dramatically improve the 2,4-D degradation process (107). Fe(III)-catalysis appeared to lead to herbicide mineralization, though initial transformation of the 2,4-D and 2,4,5-T to chlorophenolics was faster by the Fe(I1) reaction. The only transient intermediates observed during the mineralization of the herbicides were the chlorinated phenol components (106). Methanol appeared to degrade to C02 through formaldehyde and formic acid intermediates (139).

VI. THESIS OUTLINE

In this thesis, the ability of whole cultures of the lignin-degrading fungus Phanerochaete ch ysosporium, and of hydroxyl radicals (*OH)to degrade azo dyes under aerobic conditions was investigated. The fungal and chemical degradation methods are substantially non-specific in their abilities to oxidize organic substrates. In addition, several peroxidases were tested for the ability to degrade azo dyes. In chapter 2, whole cultures of P. chysosporium are shown to extensively mineralize a number of structurally diverse hydrophobic 14C- labeled azo dyes. The mineralization of most of the dyes occurs only under ligninolyic culture conditions. In five azo dyes both aromatic rings are shown to be degraded. In chapter 3, the peroxidases HRP and Lip, along with Mn(II1)-acetate (a MnP mimic), are shown to degrade the dye Disperse Yellow 3 (DY3) (2-(4'- acetamidophenylazo)-Q-methylphenol). DY3 was shown in chapter 2 to be mineralized by P. chysosporium under culture conditions known to produce large quantities of LiP and MnP. The products from DY3 degradation, and from the degradation of two DY3 analogs, include acetanilide and quinones. Product identification and quantitation allows a mechanism for azo dye oxidation to be proposed. The results of deuterium labeling studies also support the proposed mechanism. In chapter 4, peroxidase degradation of several phenylazo substituted dyes is studied. The carcinogen benzene is produced when the dyes are oxidized by HRP, consistent with the oxidation mechanism proposed in chapter 3. The other oxidation products found from the sulfonated and non- sulfonated dyes studied are also consisitent with the proposed mechanism. In chapter 5, -OH, produced by Fe(lIt)/H202, cause extensive mineralization of 14C-labeled azo dyes. Oxidation of azo dyes by *OHis shown to produce benzene from dyes containing phenylazo substitutions, and substituted benzene products from dyes containing substituents on the phenylazo portion. Several hydrophobic dyes are transformed into water soluble products that include aliphatic acids. A mechanism for the -OH- mediated degradation of the azo dyes is proposed based upon product analysis. This mechanism is similar to that proposed for the degradations of azo dyes by peroxidases. 1. Zollinger, H. (1987) In Color Chemistry: Synthesis, Properties, and Applications of Organic Dyes and Pigments, 2nd edit., pp. 1-41,85-148, VCH, Berlin.

2. Miyadera, T. (1975) In The Chemistry of the Hydrazo, Azo, and Azoxy Groups, Part 1, (S. Patai, Ed.), pp. 495-539, John Wiley and Sons, London.

3. Clarke, E.A., and Anliker, R. (1980) In The Handbook of Environmental Chemistry, Vol. 3, Part A, Anthropogenic Compounds, (0.Hutzinger, Ed.), pp. 181-211, Springer-Verlag, Berlin.

4. Catino, S.C., and Farris, R.E. (1978) In The Kirk-Othrner Encyclopedia of Chemical Technology, 3rd Edit., Vol. 3, (Eckroth, D., Ed.), pp. 387-433, John Wiley and Sons, New York.

5. Chung, K.-T., and Stevens, S.E., Jr. (1993) Environ. Toxicol. Chem. 12, 2121- 2132.

6. Bartha, R, and Pramer, D. (1967) Science 156,1617-1618.

7 Bartha, R., and Pramer, D. (1970) Advan. Appl. Microbiol. 13,317-341.

8. Bordelau, L.M., and Bartha, R. (1972) Can. J. Microbiol. 18, 1873-1882.

9. Daniels, D.G.H., and Saunders, B.C. (1953) J. Chem. Soc. 822-826.

10. Dawson, J.F. ( 1984) Rev. Prog. Coloration 14,90-97.

11. Gregory, P. (1992) In Kirk-Othmer Encyclopedia of Chemical Technology, 4thEdit., Vol. 8, (Howe-Grant, M., Ed.), p.548, John Wiley and Sons, New York.

12. Glover, B. (1992) In Kirk-Othmer Encyclopedia of Chemical Technology, 4th Edit., Vol. 8, (Howe-Grant, M., Ed.), pp. 678-679, John Wiley and Sons, New York.

13. Hilgetag, G., and Martini, A. (eds.) (1972) In Preparative Organic Chemistry, p588, John Wiley and Sons, New York.

14. Cofrancesco, A. J. (1992) In Kirk-Othmer Encyclopedia of Chemical Technology, 4th Edit., Vol. 8, (Howe-Grant, M., Ed.), pp784-797, John Wiley and Sons, New York. Chudgar, R.J. (1992) In Kirk-Othmer Encyclopedia of Chemical Technology, 4th Edit., Vol. 3, (Howe-Grant, M., Ed.), pp. 821-875, John Wiley and Sons, New York.

Games, L.M., and Hites, R.A. (1977) Analytical Chem. 49, 1433-1440.

Riefe, A. (1989) In Proceedings of the 41st Southeast Regional ACS Meeting, Winston-Salem, North Carolina, October 9-11.

Vaidya, A.A., and Datye, K.V. (1982) Colourage 14/3-10.

Tincher, W.C., and Robertson, J.R. (1982) Textile Chem. Colorist 14(12), 269- 275.

Maguire, R.J., and Tkacz, RJ. (1991) Water Poll. Res. J. Canada 26(2), 145-161.

Riefe, A. (1992) In Kirk-Other Encyclopedia of Chemical Technology, 4th Edit., Vol. 8, (Howe-Grant, M., Ed.), pp. 753-783.

Yang, Y., Ladisch, C.M., and Ladisch, M.R. (1988) Enzyme Microb. Technol. 10, 632-636.

McKay, G. (1979) Amer. Dyestuff Rep. 68/29-37.

Kulla, H.G., Krieg, R., Zimmerma~,T., and Leisinger, T. (1983) In Curr. Perspect. Microb. Ecol. Proc. Int. Symp., (Klug, M.J., and Reddy, C.A., Eds.), pp 663-667, Am. Soc. Microbiol., Washington, D.C.

Pagga, U., and Brown, D. (1986) Chemosphere 15(4), 479-491.

Shaul, G.M., Holdsworth, T.J., Dempsey, C.R., and Dostal, K.A. (1991) Chemosphere 22(1,2), 107-119.

Park, J., and Shore, J. (1984) J. Soc. Dyers Colour. 100(12)383-399.

Steenken-Richter, I. and Kermer, W.D. (1992) J. Soc. Dyers Colour. 108, 182- 186.

Arvanitoyannis, I., Eleftheriadis, I., and Kavlentis, E. (1987) Chemosphere 16(10-12), 2523-2529.

Elgal, G.M. (1986) Textile Chem. Color. 18(5), 15-20.

Woldman, M.L. (1974) Textile Chem. Color 6(8), 29-33.

McKay, G., Ramprasad, G., and Mowli, P. (1987) Water Res. 21(3), 375-377. Purvis, M.R. (1974) Amer. Dyestuff Rep. 63,19-26.

Nelson, C.R., and Hites, R.A. (1980) Environ. Sci. Technol. 14(9), 1147-1149.

Yatome, C., Matsufuru, H., Taguchi, T., and Ogawa, T. (1993) Appl. Microbiol. Biotechnol. 39, 778-781.

Eaton, D.C., Chang, H.-M., Joyce, T.W., Jeffries, T.W., and Kirk, T.K. (1982) Tappi 65(6), 89-92.

Ghosh, D.K., Mandal, A., and Chaudhuri, J. (1992) FEMS Microbiol. Lett. 98, 229-234.

Katayama, A., and Matsurnara, F. (1991) Environ. Sci. Technol. 25, 1329-1333.

Meyer, U. (1981) In Microbial Degradation of Xenobiotics and Recalcitrant Compounds, (T. Leisinger, Cook, A.M., Nuesch, J. and Hutter, R., Eds.), pp. 371-385, Academic Press, London.

Kulla, H.G. (1991) Ibid, pp. 387-399.

Idaka, E., Ogawa, T., and Horitsu, H. (1987) Bull. Environ. Contam. Toxicol. 39,108-113.

Haug, W., Schmidt, A., Norternam, B., Hempel, D.C., Stolz, A., and Knackmuss, H.-J. (1991) Appl. Environ. Microbiol. 57(11), 3144-3149.

Wuhrmann, K., Mechsner, K., and Kappeler, T. (1980) European J. Appl. Microbiol. Biotechnol. 9, 325-338.

Rafii, F., Franklin, W., and Cerniglia, C.E. (1990) Appl. Environ. Microbiol. 56(7), 2146-2151.

Chung, K.-T., Stevens, S.E., Jr. and Cerniglia, C.E. (1992) Crit. Rev. Microbiol. 18,175-190.

Chung, K.-T., and Cerniglia, C.E. (1992) Mutation Res. 277,201-220.

Kadlubar, F.F. (1987) IS1 Atlas of Science: Pharmacol. 1,129-132.

Huang, M.-T., Miwa, G.T., and Lu, A.Y.H. (1979) J. Biol. Chem. 254(10), 3930- 3934.

Fujita, S., and Peisach, J. (1977) Biochem. Biophys. Res. Commun. 78(1), 328- 334. De Long, M.J., Prochaska, H.J., and Talalay, P. (1986) Proc. Nat. Acad. Sci. USA 83,787-791.

Joachim, F., Burrell, and Anderson, J. (1985) Mutation Res. 156, 131-138.

Gregory, P. (1986) Dyes and Pigments 7/45-56.

National Institute of Environmnetal Health Sciences, (1993) Environ. Health. Perspect. 101 (Suppl 1).

Matthews, E.J., Spalding, J.W., and Tennant, R.W. (1993) Environ. Health. Perspect. 101 (Suppl. 2), 347-482.

McCann, J., Choi, E., Yamasaki, E., Ames, B.N. (1975) Proc. Nat. Acad. Sci. USA 72(12), 5135-5139.

Sandhu, I?., and Chipman, J.K. (1990) Mutation Res. 240, 227-236.

McCann, J., and Ames, B.N. (1976) Proc. Nat. Acad. Sci. USA 73(3), 950-954.

Arnes, B.N., Lee, F.D., and Durston, W.E. (1973) Proc. Nat. Acad. Sci. USA 70(3), 782-786.

Krishna, G., Xu, J., and Nath, J. (1986) J. Toxicol. Environ. Health 18, 111-119.

Prival, M.J., Davis, V.M., Pieperl, M.D., and Bell, S.J. (1988) Mutation Res. 206, 247-259.

Chung, K.-T., Fulk, G.E., and Andrews, A.W. (1981) Appl. Environ. Microbiol. 42(4), 641-648.

Chung, K.-T. (1983) Mutation Res. 114,269-281.

Barr, D.P., amd Aust, S.D. (1994) Environ. Sci. Technol. 28(2), 78A-87A.

Glenn, J.K., and Gold, M.H. (1983) Appl. Environ. Microbiol. 45(6), 1741-1747.

Platt, M.W., Hadar, Y., and Chet, I. (1985) Appl. Microbiol. Biotechnol. 21,394- 396.

Bumpus, J.A., and Brock, B.J. (1988) Appl. Environ. Microbiol. 54(5), 1143-1150.

Gogna, E., Vohra, R., and Sharma, P. (1992) Lett. Appl. Microbiol. 14/58-60.

Muralikrishna, C., and Renganathan, V. (1993) Biochem. Biophys. Res. Cornmun. 197(2), 798-804. Kirk, T.K., and Farrell, R.L. (1987) Ann. Rev. Microbiol. 41,465-505.

Paul, K.G. (1986) In Molecular and Physiological Aspects of Plant Peroxidases, (Greppin, H., Penel, C., and Gaspar, T., Eds.), pp. 1-14, University of Geneva, Switzerland.

Dunford, H.B. (1986) In Molecular and Physiological Aspects of Plant Peroxidases, (Greppin, H., Penel, C., and Gaspar, T., Eds.), pp. 15-23, University of Geneva, Switzerland.

Welinder, K.G. (1986) In Molecular and Physiological Aspects of Plant Peroxidases, (Greppin, H., Penel, C., and Gaspar, T., Eds.), pp. 61-70, University of Geneva, Switzerland.

Welinder, K.G. (1991) Biochim. Biophys. Acta 1080,215-220.

Gold, M.H., Wariishi, H. and Valli, K. (1988) ACS Symp. Ser. 389, 127-140.

Dawson, J.H. (1988) Science 240,433-439.

Ator, M.A., and Ortiz de Montellano, P.R. (1987) J. Biol. Chem. 262(4), 1542- 1551.

Ortiz de Montellano, P.R. (1987) Acc. Chem. Res. 20,289-294.

Bollag, J.-M., Chen, C.-M., Sarkar, J.M., and Loll, M.J. (1987) Soil, Biol. Biochem. 19(1), 61-67.

Kersten, P.J., Kalyanaraman, B., Hammel, K.E., Reinhammar, B., and Kirk, T.K. (1990) Biochem. J. 268,475-480.

Popp, J.L., and Kirk, T.K. (1991) Arch. Biochem. Biophys. 288(1), 145-148.

Millis, C.D., Cai, D., Stankovich, M.T., and Tien, M. (1989) Biochemistry 28, 8484-8489.

Hammel, K.E., Kalyanaraman, B., and Kirk, T.K. (1986) J. Biol. Chem. 261(36), 16948-16952.

Alberti, B.N., and Klibanov, A. M. (1981) Biotechnol. Bioeng. Symp. No. 11, 373-379.

Shannon, M.J.R, and Bartha, R. (1988) Appl. Environ. Microbiol. 54(7), 1719- 1723. Nannipieri, P., and Bollag, J.-M. (1991) J. Environ. Qual. 20, 510-517.

Wondrack, L., Szanto, M., and Wood, W.A. (1989) Appl. Biochem. Biotechnol. 20/21,765-780.

Valli, K., Wariishi, H., and M.H. Gold (1992) J. Bacteriol. 174(7), 2131-2137.

Haemrnerli, S.D., Leisola, M.S.A., Sanglard, D., and Fiechter, A. (1986) J. Biol. Chem. 261(15), 6900-6903.

Field, J.A., de Jong, E., Costa, G.F., and de Bont, J.A.M. (1992) Appl. Environ. Microbiol. 58(7), 2219-2226.

Kirk T.K., Schultz, E., Comors, W.J., Lorenz, L.F., and Zeikus, J.G. (1978) Arch. Microbiol. 117,277-285.

Enoki, A., Goldsby, G. P., and Gold, M.H. (1981) Arch. Microbiol. 129,141-145.

Zitzelsberger, W., Ziegler, W., and Wallnofer, P.R. (1987) Chemosphere 16(6), 1137-1142.

Valli, K., and Gold, M.H. (1991) J. Bacteriol. 173(1), 345-352.

Valli, K., Brock, B.J., Joshi, D.K., and Gold, M.H. (1992) Appl. Environ. Microbiol. 58(1), 221-228.

Dhawale, S.W., Dhawale, S.S., and Dean-Ross, D. (1992) Appl. Environ. Microbiol. 58(9), 3000-3006.

Lee, B., Pometto, A.L., III, Fratzke, A., and Bailey, T.B., Jr. (1991) Appl. Environ. Microbiol. 57(3), 678-685.

Stahl, J.D., and Aust, S.D. (1993) Biochem. Biophys. Res. Commun. 192(2), 477- 482.

Yajima, Y., Enoki, A., Mayfield, M.B., and Gold, M.H. (1979) Arch. Microbiol. 123,319-321.

Enoki, A., Yajima, Y., and Gold, M.H. (1981) Phytochemistry 20(7), 1543-1546.

Marquez, L., Wariishi, H., Dunford, H.B., and Gold, M.H. (1988) J. Biol. Chem. 263(22), 10549-10552.

Wariishi, H., and Gold, M.H. (1990) J. Biol. Chem. 265(4), 2070-2077.

Renganathan, V., and Gold, M.H. (1986) Biochemistry 25, 1626-1631. Wariishi, H., Akileswaran, L., and Gold, M.H. (1988) Biochemistry 27, 5365- 5370.

Goldstein, S., Meyerstein, D., and Czapski, G. (1993) Free Radic. Biol. Med. 15, 435-445.

Aust, S.D., Morehouse, L.A., and Thomas, C.E. (1985) J. Free Radic. Biol. Med. 1/3-25,

Pignatello, J. J. (1992) Environ. Sci. Technol. 26,944-951.

Sun, Y., and Pignatello, J.J. (1993) Environ. Sci. Technol. 27, 304-310.

Walling, C. (1975) Acc. Chem. Res. 8, 125-131.

Kunai, A., Hata, S., Ito, S., and Sasaki, K. (1986) J. Am. Chem. Soc. 108,6012- 6016.

Sedlak, D.L., and Andren, A.W. (1991) Environ. Sci. Technol. 25, 777-782.

Leung, S.W., Watts, R.J., and Miller, G.C. (1992) J. Environ. Qual. 21,377-381.

Bowers, A.R., Cho, S.H., and Singh, A. (1992) In Chemical Oxidation: Technologies for the Nineties, Vol. 1, (Eckenfelder, W. W., Bowers, A.R., and Roth, J.A., Eds.), pp. 11-25, Technomic, Lancaster, PA.

Gay, W.A. (1994) In Chemical Oxidation: Technologies for the Nineties, Vol. 2, (Eckenfelder, W.W., Bowers, A.R., and Roth, J.A., Eds.), pp. 44-57, Technomic, Lancaster, PA.

Lipczynska-Kochany, E. (1994) In Chemical Oxidation: Technologies for the Nineties, Vol. 3, (Eckenfelder, W.W., Bowers, A.R., and Roth, J.A., Eds.), pp. 12-27, Technomic, Lancaster, PA.

Weber, E.J., and Wolfe, N.L. (1987) Environ. Toxicol. Chem. 6, 911-919.

Weber, E.J. (1991) Environ. Toxicol. Chem. 10, 609-618.

Brown, D., and Hamburger, B. (1987) Chemosphere 16(7), 1539-1553.

Leisinger, T. and Brunner, W. (1981) In Biotechnology: A Comprehensive Treatise in 8 Volumes, (Rehm, H.-J., and Reed, G., Eds.), Vol. 8., pp. 475-513, VCH, New York.

Berry, D.F.,Francis, A.J., and Bollag, 1.-M. (1987) Microbial. Rev. 51(1), 43-59. Eriksson, K.-E.L., Blanchette, R.A., and Ander, P. (1990) In Microbial and Enzymatic Degradation of Wood and Wood Components, pp. 1-72, 225-332, Springer-Verlag, Berlin.

Valli, K., Wariishi, H., and Gold, M.H. (1990) Biochemistry 29, 8535-8539.

Harvey, P.J., Gilardi, G.F., and Palmer, J.M. (1989) In Enzyme Systems for Lignocellulose Degradation, (Coughlan, M.P., Ed.), pp. 111-120, Elsevier, London.

Tratnyek, P.G., Elovitz, M.S., and Colverson, P. (1994) Environ. Toxicol. Chem. 13/27-33.

Kopecka, M., and Gabriel, M. (1992) Arch, Microbial. 158,115-126.

Ramaswami, V., and Subburam, V. (1992) Bull. Environ. Contami. Toxicol. 48,249-252.

Anliker, R., Moser, P., and Poppinger, D. (1988) Chemosphere 17(8), 1631-1644.

Sundstrom, G., and Octavio Ruzo, L. (1978) In Aquatic Pollutants: Transformation and Biological Effects, (Hutzinger, O., Van Lelyveld, I.H., and Zoeteman, B.C. J., Eds.), pp. 205-222, Pergamon, Oxford.

Brown, M.A., and Devito, S.C. (1993) Crit. Rev. Environ. Sci. Technol. 23(3), 249-324.

Glaze, W.H. (1994) In Chemical Oxidation: Technologies for the Nineties, Vol. 3, (Eckenfelder, W.W., Bowers, A.R., and Roth, J.A., Eds.), pp. 1-11, Technomic, Lancaster, PA.

Kotronarou, A., Mills, G., and Hoffmam, M.R. (1992) Environ. Sci. Technol. 26,1460-1462.

Petrier, C, Micolle, M., Merlin, G., Luche, J.-L., and Reverdy, G. (1992) Environ. Sci. Technol. 26, 1639-1642.

Hao, O.J., and Phull, K.K. (1993) Environ. Sci. Technol. 27, 1650-1658.

Dumdei, B.E., and O'Brien, R.J. (1984) Nature 311, 248-250.

Ho, P.C. (1986) Environ. Sci. Technol. 20,260-267.

Yamamoto, Y., Niki, E., Shiokawa, H., and Kamiya, Y. (1979) J. Org. Chem. 44(13), 2137-2142. Flaherty, K.A., and Huang, C.P. (1994) In Chemical Oxidation: Technologies for the Nineties, Vol. 3, (Eckenfelder, W.W., Bowers, A.R., and Roth, J.A., Eds.), pp. 58-77, Technomic, Lancaster, PA.

Mills, A., Davies, R.H., and Worsley, D. (1993) Chem. Soc. Rev. 22(6), 417-425.

Vanysek, P. (1991) In CRC Handbook of Chemistry and Physics, (Lide, D.R., Ed.), p 8-24, CRC Press, Boca Raton, FL.

Murphy, A.P., Boegli, W.J., Price, M.K., and Moody, C.D. (1989) Environ. Sci. Technol. 23,166-169.

Potter, F.J., and Roth, J.A. (1994) In Chemical Oxidation: Technologies for the Nineties, Vol. 3, (Eckenfelder, W.W., Bowers, A.R., and Roth, J.A., Eds.), pp. 91-111, Technomic, Lancaster, PA.

Izurni, I., Fan, F.-RF., and Bard, A.J. (1981) J.Phys. Chem. 85,218-223.

Hashimoto, K., Kawai, T., and Sakata, T. (1984) J. Phys. Chem. 88,4083-4088.

Adler, E. (1977) Wood Sci. Technol. 11,169-218 CHAPTER TWO

DEGRADATION OF AZO DYES BY THE LIGNIN-DEGRADING FUNGUS PHANEROCHAETE CHRYSOSPORIUM

I. INTRODUCTION

P. ch ysospow'um, the best-studied lignin-degrading fungus, degrades lignin during secondary metabolic (idiophasic) growth, the onset of which is triggered by the depletion of nutrient nitrogen (9,12,13). Ligninolytic cultures of P. chrysosporium produce two extracellular peroxidases, lignin peroxidase and manganese peroxidase, which along with an H202-generating system, are the major extracellular components of its lignin-degrading system (12,13). Ligninolytic cultures of P. chysosporium are capable of mineralizing many persistent aromatic pollutants (2,3,11,22,23). Since azo dyes are relatively recalcitrant to bacterial degradation (15,21,24), the ability of P. chrysosporium to mineralize these compounds was examined. Glenn and Gold first established that ligninolytic cultures of P. ch ysosporium decolorize several polymeric dyes (8). Later, decolorization of azo dyes and triphenylmethane dyes by P. chrysospon'um was demonstrated (4,4,19,20). Decolorization, however, demonstrates only transformation of the chromophoric group of a dye; it does not demonstrate complete degradation of the dye. In this study, radiolabeled substrates have been used to establish that P. chysosporium is capable of mineralizing a wide variety of hydrophobic azo dyes including: 4-phenylazophenol (I), 4-phenylazo-2- methoxyphenol (11), Disperse Yellow 3 (111), 4-phenylazoaniline (IV), N,N- dimethyl-4-phenylazoaniline (V), Disperse Orange 3 (VI)and Solvent Yellow 14 (VII) (Table I). The microbial degradation of these compounds has not been examined previously. All the dyes used in this study, except 4- phenylazophenol (I) and 4-phenylazo-2-methoxyphenol (II), are considered toxic by the Environmental Protection Agency (16). Disperse Yellow 3 is the corresponding acetanilide. Reduction of the nitro group of 4-nitro-N-acetylU- 14c]aniline with stannous chloride (87 ymol) in ethanol-ethyl acetate (1:1, 50 pl) at 750 C, for 2h, generated 4-arnino-~-acet~l[~-~~~]aniline,and this was used in the synthesis of III without further purification (1). Amino azo dyes. These dyes were synthesized by coupling the diazonium salt of an aromatic amine (12-20 ymol) with another appropriate aromatic amine (12-20 pmol) in glacial acetic acid (20-40 p1) for 30 min (17). Radioactivity of diazonium salt or amine when used in these syntheses was 10 pCi. After the coupling step, the pH of the reaction mixture was adjusted to 9 with NaOH, and the precipitated amino azo dye was extracted with ethyl acetate. ~,~-~imeth~l[~-~~~]aniline,required in the synthesis of N,N- dimethyl-4-phenylazo[v-14~]aniline(Va), was prepared by methylating [U- 14c]aniline (0.1 mmol in 0.7 ml of THF) with formaldehyde (37% solution, 50 p1) and sodium borohydride (27.2 mg) under acidic conditions (7). The N,N- dimethylaniline obtained was used in the synthesis of labeled N,N-dimethyl- 4-phenylazoaniline (V) without further purification. Purification of radiolabeled dyes. All the dyes were purified initially by silica gel TLC using hexaneethyl acetate (7:3). All, except for 4- phenylazophenol (I) and 4-phenylazo-2-methoxy phenol (II) were further purified by HPLC using a C-18 reverse phase column (p-Bondapak, Waters Associates) and a water-methanol gradient (20-100% methanol, 10 min, 1 ml/min) as eluant. N,N-Dimethyl-4-phenylazoaniline(V) was contaminated with the corresponding N-monomethyl and free amino dyes. To remove these impurities the reaction mixture was acetylated with acetic anhydride, which acetylates monomethylated and unmethylated amines only. N,N- Dimethyl-4-phenylazoaniline (V) was separated from the acetylated impurities by TLC using hexaneethyl acetate (7:3). All of the purified dyes displayed only a single peak in HPLC analysis, and cochromatographed with standard unlabeled dyes. Syntheses and purification of non-radiolabeled dyes were made on the same scale as the corresponding radioactive dyes; mass spectral characteristics of these non-radiolabeled dyes were identical to those of standard purchased dyes. Specific activities of pure radiolabeled azo dyes ranged from 70 to 200 yCi/mmol. Culture conditions. P. chysosporium OGC 101 (26) was grown from a conidial inoculum (optical density at 650 nm of -10.0, 0.1 rnl) at 370C in stationary culture (20 ml in 250 ml Erlenmeyer flask) as described (6,12). The medium composition was the same as described previously except that 3% glucose, 1% Tween-80,30 pM MnS04 and either 1.2 or 24 rnM ammonium tartrate were used (12). The medium was buffered with 20 mM sodium dimethylsuccinate, pH 4.5. Cultures were incubated under air for 3 days, after which they were purged with 99.9% O2 for 30 min every three days. Mineralization of azo dyes. 14c-labeled azo dye (250 nmol, 2.5 x lo4 -1.2 x lo5 cpm) in ethanol (40 pl) was added to cultures on day 3. During 02 purging, evolved 14C02was trapped in a basic scintillation fluid as previously described (12) and quantitated using a Beckmam LS-3133 liquid scintillation spectrometer. Four flasks were used for each mineralization experiment. Standard deviation was calculated by using Quattro Pro commercial software (Borland International Inc.). Mass balances for the completed experiments were calculated by first determining the amount of radioactivity remaining in the cultures, and then summing that amount and the amount of radioactivity liberated as 14C02,

111. RESULTS

Mineralization of the following specific 14c-labeled azo dyes by P. chrysosporium was studied: 4-phenylazo~-14~]phenol(Ia), 4-[U- 14~]phenylazophenol(Ib), 4-phenylazo-2-methoxy[U-14~]phenol(IIa), 4-[U- 14~]phenylazo-2-methoxyphenol(IIb), a~etamido[u-~~~]phen~lring labeled Disperse Yellow 3 [2-(4'-acetarnido[U-14~]phenylazo)-4-me&ylpheml](111), 4- phenylazo[~-14~]aniline(IV), ~,~-dimeth~l-4-~hen~lazo[~-~~~]aniline (Va), ~,~-dimeth~l-4-[~-~~~]~hen~lazoaniline(Vb), [~-~~C]aminophenyl dng- labeled Disperse Orange 3 [4-(4'-nitrophenylazo)[~-14~]ani1ine](VIa), [U- 14~]nitrophenylring-labeled Disperse Orange 3 [4-(4'-nitro[U- 14~]phenylazo)aniline](VIb), [8-14~]naphthol-labeledSolvent Yellow 14 (1- phenylazo[8-14~]-2-naphthol)(Wa) and [~-~~~]phen~lring-labeled Solvent Yellow 14 (l-[~-~~~]~hen~lazo-2-na~hthol)(VIIb) (Table I). Mineralization of the azo dyes by P. chysosporium was examined under nitrogen-sufficient and nitrogen-limiting culture conditions (Table I). Time courses for the mineralization of 4-phenylazophenol (Ia, Ib) and Disperse Orange 3 (ma, VIb) are shown in Figures la and lb. Only nitrogen- limited, ligninolytic cultures could efficiently mineralize all of the dyes. Table I. Mineralization of Azo Dyes by P. Chrysosporium Mineralization (%) 12 days after subsbate addition Low nitrogen High nitrogen culture culture

Side chain 14c-labeled DHP

Disperse Yellow 3 OH

Disperse Orange 3

W@N-N*NY (Vla)

Went Yellow 14 HO

- -- - * Indiites the location of radioactivity 0" *00 r 0 3 6 9 12 15 18 Days

.-- 0 3 6 9 12 15 18 Days

Figure 1. Mineralization profiles for 1%-labeled dyes under low-nitrogen (1.2 mM ammonium tartrate, closed symbol) and high nitrogen (24 mM ammonium tartrate, open symbol) culture conditions. (A) dphenylazow-

14CIpheno1, Ia (M ,o ) (Solvent Yellow 7); 4-[U-14C]phenylazophenol,Ib (@ ,o ). (B) 4-(4'-nitropheny1azo)-yU-1*C]aniline(Disperse Orange 3), VIa (M,a) ; 4-(4'- nitro yU-14Clphenylazo)-aniline, Vlb (a ,o ) . Table 11. Total recovery of 14C-labeled material through day 15 as percentage of total added to cultures on day 3

Dve LN culture HN culture

IIa IIb

VIa VIb

a - as found in C02, dissolved material, extractable material, and material attached to biomass b - LN - low-nitrogen, ligninolytic cultures; HN - high-nitrogen, non- ligninolytic cultures 4-~hen~lazo~-~~~]~henol(Ia) and 4-phenylazo-2-methoxy[v-14~]phenol (Ib) were mineralized to a lesser but significant degree under nonligninolytic conditions (Table I, Figure 1). The rate of 14c02evolution during mineralization of Ia was maximal in the first three days after substrate addition and decreases significantly after 12 days (Figure 1A). Similar patterns of 14c02release were obtained with most other radioactive dyes. However, 14c02evolution due to Disperse Orange 3 mineralization was linear for nine days before it decreased (Figure 1B). Mass balances for these degradation experiments were determined, as well (Table II).

N.DISCUSSION

All the dyes examined were degraded much more rapidly under nitrogen-limiting conditions than under nitrogen-sufficient conditions (Table I). However, phenolic ring labeled 4-phenylazophenol (Ia) and 4-phenylazo- 2-methoxyphenol (IIa) were also mineralized extensively under nitrogen- sufficient conditions (Table I, Figure la). This observation is extremely interesting and may suggest the involvement of enzyme(s) other than lignin- degrading enzymes in the mineralization of la and IIa. Except for [8- 14~]naphthol-labeledSolvent Yellow 14 (Wa), all the other radiolabeled azo dyes were mineralized to at least 23% under nitrogen-limiting, lignin- degrading conditions (Table I). Aromatic rings substituted with a phenolic or an amino or an acetarnido function, such as in the dyes Ia, IIa, 111, IV,Va and VIa, were degraded to a greater extent than aromatic rings without such substituents. A comparison of mineralization of 4-phenylazo~-14~]phenol (Ia) and 4-phenylazo-2-methoxy~-14~]phenol(IIa) under nitrogen-limiting conditions suggested that a methoxyl group substitution into a phenolic ring enhances aromatic ring degradation. Recently, Paszczynski et al. (19) observed that introduction of a guaiacyl (2-methoxyphenol) substructure into azo dyes enhanced dye decolorization by P. chrysosporium. The present study demonstrates that this substitution also increases the extent of mineralization of these compounds. Disperse Orange 3 (VI) contains two substituted aromatic rings, one with an amino substituent and the other with a nitro substituent. Surprisingly, both aromatic rings were mineralized to a similar extent (Table I, Figure lb). A nitro function might have been expected to retard the mineralization process. However, Valli et al. (23) recently demonstrated that the degradation of 2,4-dinitrotoluene by P. ch ysosporium was initiated by the reduction of a nitro group to the corresponding amine. Similar nitro group reduction reactions may precede the ring-cleavage reactions in the mineralization of Disperse Orange 3 (VI). The phenyl ring of Solvent Yellow 14 (VIIb) was mineralized to approximately the same extent as the corresponding ring in 4-phenylazophenol (Ib) and 4-phenylazo-2- methoxyphenol (1%). However, [&14~]naphthol-labeledSolvent Yellow 14 (VIIa) was mineralized very slowly, probably due to the specific location of the labeled carbon. In a separate experiment, [~-~~~]na~htholitself was mineralized slowly (data not shown). Thus, the azo dye structure probably did not suppress naphthol ring degradation. In summary, this study established that: a) lignin-degrading cultures of P. ch ysosporium were capable of mineralizing a variety of toxic azo dyes; and b) the rate at which P. chrysosporium mineralized aromatic rings of azo dyes was dependent on the nature of ring substituents. Aromatic rings with a hydroxyl, an amino, an acetamido or a nitro substituent were degraded faster than rings without such substituents. V. REFERENCES

1. Bellamy, F. D., and Ou, K. (1984) Tetrahedron Lett. 25,839-842.

2. Bumpus, J. A., Tien, M., Wright, D., and Aust, S.D. (1985) Science 228, 1434-1436.

3. Bumpus, J. A,, and Aust, S.D. (1987) Bioessays 6, 166-170.

4. Bumpus, J. A., and Brock, B.J. (1988) Appl. Environ. Microbiol. 54, 1143- 1150.

5. Cripps,C., Bumpus, J.A., and Aust, S.D. (1990) Appl. Environ. Microbiol. 56, 1114-1118.

6. Enoki, A,, Goldsby, G.P., and Gold, M.H. (1981) Arch. Microbiol. 129, 141-145.

7. Giumanini, A. G., Chiavari, G., Musiani, M.M., and Rossi, P. (1980) Synthesis, 743-746.

8. Glenn, J. K., and Gold, M. H. (1983) Appl. Environ. Microbiol. 45,1741- 1747.

9. Gold, M. H., Wariishi, H., and Valli, K. (1989) In Biocatalysis in Agricultural Biotechnology, ACS Symposium Series no.389, (Whitaker, J.R. and Sonnet, P.E., Eds.),pp. 127-140, American Chemical Society, Washington, D.C.

10. Gold, M. H., Mayfield, M.B., Cheng,T.M., Krishnangura, K., Enoki, A., Shimada, M., and Glenn, J.K. (1982) Arch. Microbiol. 132, 115-122.

11. Hammel, K. E. (1989) Enzyme Microb. Technol. 11, 776-777.

12. Kirk, T. K., Schultz, E., Connors, W.J., Lorenz, L.F., and, and Zeikus, J.G. (1978) Arch. Microbiol. 117,277-285.

13. Kirk, T. K., and Farrell, R.L. (1987) Ann. Rev. Microbiol. 41,465-505.

14. Kratzl, K., and Vierhapper, F. (1971) Monatsh. Chem. 102, 224-232.

15. Kulla, H. G., Klausener, F., Meyer, U., Ludeke, B., and Leisinger, T. (1983) Arch. Microbiol. 135,l-7. Lowry, G. G., and Lowry, R.C. (1988) In Lowry's Handbook of Right to Know and Emergency Planning, pp. 333-382, Lewis Publishers Inc., Chelsea, Michigan.

Maynard, Jr., C. W. (1983) Dye application, Manufacture of Dye Intermediates and Dyes, p. 809-861. In Riegel's Handbook of Industrial Chemistry, (J. A. Kent, ed.), Van Nostrand Reinhold, New York.

McCann, J., and Ames, B.N. (1975) Proc. Natl. Acad. Sci. 73,950-954.

Paszczynski, A., Pasti, M.B., Goszczynski, S., Crawford, D.L., and Crawford, RL. (1991) Enzyme Microb. Technol. 13,378-384.

Paszczynski, A., and Crawford, R.L. (1991) Biochem. Biophys. Res. Comrnun. 178, 1056-1063.

Shaul, G. M., Holdsworth, T.J., Dempsey, C.R, and Dostal, K.A. (1991) Chemosphere, 22, 107-119.

Valli, K., and Gold, M.H. (1991) J. Bacteriol. 173,345-352.

Valli, K., Brock, B.J., Joshi, D., and Gold, M.H. (1992) Appl. Environ. Microbiol. 58,221-228.

Zimrnerman, T., Kulla, H.G., and Leisinger, T. (1982) Eur. J. Biochem. 129,197-203.

Zollinger, H. (1987) In Color Chemistry: Syntheses, Properties and Applications of Organic Dyes and Pigments, pp 92-102, VCH Publishers, New York.

Alic, M. A., Letzring, C., and Gold, M. H. (1987) Appl. Environ. Microbiol. 53,1461-1469. CHAPTER THREE

PEROXIDASE-CATALYZED OXIDATION OF AZO DYES: MECHANISM OF DISPERSE YELLOW 3 DEGRADATION

I. INTRODUCTION

Lignin-degrading cultures of the white-rot basidiomycete Phanerochaete chrysosporium were capable of mineralizing a number of sulfonated and non-sulfonated azo dyes, including Disperse Yellow 3 (DY3), to C02 (14,16). However, the biochemical mechanisms underlying the fungal degradation of azo dyes are not understood. Azo dye degradation by P. chysosporium coincides with production of lignin-degrading peroxidases in the culture medium and recent reports suggest that lignin peroxidase (Lip) and manganese peroxidase (MnP) oxidize and decolorize azo dyes (3,12,13,19). However, the degradation products formed in these reactions have not been identified and the mechanism of azo dye degradation by peroxidases is not known. In this study, degradation of DY3 and two naphthol analogs [1-(4'-acetamidopheny1azo)-2-naphthol (NDY3) (11) and 1-(4'-acetamido-2',6'-dimethylphenylazo)-2-naphthol (NDMDY3) (III)] by Lip, horseradish peroxidase (HRP), and Mn(III)-malonate (a MnP mimic) is examined in detail.

11. MATERIALS AND METHODS

Chemicals. Disperse Yellow 3, 2-naphthol, 3,5-dimethylaniline, 1,2- naphthoquinone, 4-methylcatechol, 2,6-dimethylphenol, 2,6-dimethyl-1,4- benzoquinone, manganese(III) acetate, 1,4-dioxane, 1,4-dioxane-dg (99%D), and D20 (99%D) were purchased from Aldrich Chemical Company. HRP, Tween-80, 4-nitroaniline, and Hz02 were obtained from Sigma Chemical Company. Synthesis of NDY3 (11). 4-Nitroaniline was acetylated with acetic anhydride at room temperature as described (5). The resulting 4- nitroacetanilide was then reduced to 4-aminoacetanilide with stannous chloride in ethano1:ethyl acetate mixture (1:l) at 750 C for 2 h (2). 4- Aminoacetanilide was diazotized with sodium nitrite at 0-50 C, and then reacted with a basic solution of 2-naphthol to obtain the naphthol analog of DY3 (5). The crude NDY3 was purified by silica gel column chromatography with hexane-ethyl acetate solvent mixture as the eluent. Mass spectral analysis supported the proposed structure for NDY3. MS (m/z): 305 (M, loo%), 143 (70.1%),134 (47.8%), 115 (50.1%). Synthesis of NDMDY3 (111). 33-Dimethylaniline was aketylated with acetic anhydride:pyridine (2:1), then nitrated with dilute nitric acid in concentrated sulfuric acid (5). The nitro group was then reduced to the corresponding amino function by reaction with cuprous chloride and sodium borohydride in acetic acid for 2 h (11). The amine was purified by silica gel thin-layer chromatography using hexane-ethyl acetate mixture (3:l) as the solvent system. 4-Amino-3,5-dimethylacetanilidewas then diazotized with nitrous acid at 0-50 C and coupled to 2-naphthol as described (5). The dye was purified by thin-layer chromatography with hexane-ethyl acetate (3:l) as the solvent system. Mass spectral analysis confirmed the structure of NDMDY3. MS (m/z): 333 (M, loo%), 162 (77.3%),143 (47%), 120 (63.5%), 115 (50.4%). Synthesis of 4-acetamido-2,6-dimethylphenol. 2,6-Dimethylphenol was nitrated with nitric acid (1 eq) in acetic acid at 0-50 C for 10 min to obtain 4- nitro-2,6-dimethylphenol. The nitrophenol was reduced with cuprous chloride and sodium borohydride as described (11). The resulting aminophenol was then converted to 4-acetamido-2,6-dimethylphenol by reaction with acetic anhydride. Peroxidase reactions. HRP (type VIa) was purchased from Sigma Chemical Company and was used as received. LiP was purified from the lignin-degrading cultures of P. ch ysospon'urn as described (7,17). Lip (0.125 mg) was dissolved in 2.5 mL of 10 mM sodium dimethylsuccinate, pH 4.5, containing 1%Tween-80 (v/v), 5% dioxane (v/v) and 0.8 polof an azo dye. Hz02 (0.8 pmol) was added in 10 aliquots to this mixture over a two-hour period. The reaction mixture was acidified to pH 2 with dilute HC1, saturated with sodium chloride, and then extracted with ethyl acetate. The extract was evaporated to dryness under a nitrogen stream, redissolved in methanol, and analyzed by HPLC. To identify the quinone products, the quinones were converted to dihydroquinone by dithionite reduction of the reaction mixture. The reduced products were then extracted with ethyl acetate, acetylated with pyridine:acetic anhydride (1:2) and analyzed by gas chromatography-mass spectrometry (GC-MS). HRP reactions were performed in a similar fashion, except that the amount of enzyme used was 0.5 mg and the buffer was 10 mM phosphate, pH 6. Mn(II1)-malonate reactions. A stock solution of Mn(III) acetate (20 mM) in 0.5 M malonate, pH 4.5, was prepared. The dye (0.8 pmol) was dissolved in 2.5 mL of 50 mM malonate buffer, pH 4.5, containing 1% Tween- 80 (v/v) and 5% dioxane (v/v). Mn(III)-malonate (1.6 pmol) was added in 10 aliquots over a 2-h period. The reaction products were isolated as described for the peroxidase reactions. Deuterium incorporation experiments. Lip (0.375 mg) or HRP (1.5 mg) was dissolved in 2.5 mL of an adpropriate buffer containing either 99% deuterated or non-deuterated dioxane (5% v/v). DY3 (I) (0.8 pmol) in dioxane was added in 10 aliquots over a 1-h period. H202 (1.6 pmol) was added in 20 aliquots simutaneously. The products were extracted, separated by HPLC, and the peak corresponding to acetanilide was collected. The solution was evaporated to dryness and the residue was analyzed by GC-MS. Deuterium incorporation experiments with D20 (99%) were performed in a similar fashion except that the buffer was prepared with D20 instead of H20. DY3 and NDY3 metabolites from whole cultures: P. chysosporiurn OGClOl stationary cultures (20 mL in a 250-mL Erlenmeyer flask) containing 1% Tween-80 were prepared as described (16). The cultures were incubated under air at 370 C. On the third day DY3 (1 pmol in 134.5 pl ethanol), or NDY3 (1 pmol in 305 pl ethanol), was added to each culture. The flask was sealed with a manifold and the head spaces of the cultures were purged with oxygen for 20 min. Two days after dye addition, the cultures were extracted with ethyl acetate (40 mL). The ethyl acetate extract was evaporated to dryness and the residue was analyzed by HPLC and GC-MS. HPLC analyses. HPLC analyses were performed using a reverse phase C-18 column (Separations Group, Hesperia, CA) and a water-methanol gradient. Initially the methanol concentration was maintained at 20% for 2 min, then increased to 100% over 10 min and maintained at 100% concentration for 10 min. Solvent flow rate was 1 ml/min. The products were monitored by their absorbance at 254 nm. Dye elution was followed by monitoring their appropriate visible absorption maxima (DY3 (I), 358 nrn; NDY3 (11), 470 nrn; NDMDY3 (III), 490 nm). Retention times for the standards were: acetanilide (V), 9.5 min; 1,2-naphthoquinone (VII), 10.8 min; 2,6- dimethyl-1,4-benzoquinone (IX), 11.7 min; 3,5-dimethylacetanilide (VIII), 12.9 min; DY3 (I), 15 min; NDY3 (II), 15.2 min; and NDMDY3 (III), 16 min. Products were quantified from their corresponding HPLC peak areas by comparison to a standard curve constructed with known concentrations of standard compounds. 1,2-Napthoquinone (VII) used in these analyses was purified by crystallization from diethylether and HPLC analysis indicated a purity of 98+%. GC-MS analyses. GC-MS analyses were performed at 70 eV on a VG Analytical 7070E mass spectrometer fitted with an HP 5790A GC and a fused capillary column (15 m, DB-5, J & W Science). A temperature gradient was used in the GC separations. The initial temperature was 700 C, and it was increased to 3200 C at a rate of 100 C/min. In most cases, the products were identified by comparison of their retention times and mass spectral fragmentation patterns to those of standard compounds. Mass spectra of solids were obtained by fast atom bombardment (FAB-MS) or solid probe techniques. In the FAB-MS method thioglycerol served as the matrix.

111. RESULTS

Three HRP, Lip, and the Mn(III) oxidation products of DY3 (I)were found (Figure 1). Small levels of acetanilide (V) were identified by HPLC analysis (Rt=9.5 min) and by GC-MS analyis. MS (m/z): 135 (M), 93 (M - COCH2) (100% ), 77 (CgHg+),65 (C$I5+). GC-MS analysis of the dithionite reduced and acetylated total reaction products indicated the presence of trace quantities of 4-methyl-1,2-benzoquinone(IV), but this product was not observed by HPLC analysis. MS (m/z): 208 (M) (6.5%), 166 (M - COCH2) (16%), 124 (M - 2COCH2) (100%). The major product (Rt= 16.3 min) had an absorption spectrum (Amax, 360 nm)that strongly resembled that found for DY3 (I). The products UV-visible absorption was completely removed by dithionite, suggesting the presence of an azo linkage. The dithionite reduction products were extracted into ethyl acetate, and derivatized with acetic anhydride. GC-MS analysis of the reduced and acetylated products indicated the presence of the hexa-acetate derivative of dihydroxy-diamino- dimethylbiphenyl, [MS (m/z): 496 (M), 454 (M - COCHz), 412 (M - 2CCHH2), 394 (M - 2COCH2 + HzO), 352 (M - 3COCH2 + H20), 310 (M - 4COCH2 + H20), 292 (M - 4COCHz + 2H20), 268 (M - 5COCHz + H20) (loo%),226 (M - 6COCH2 + H20)] and the tetraacetoxy derivative of 1,4-phenylenediamine [MS (m/z): 276 (M), 234 (M - COCHz), 192 (M - ZCOCHz), 150 (M - XOCHz), 108 (M - 4COCHz)I. Based on all of these analyses, the major product was identified as a DY3 dimer (VI) (Figure 1). FAB-MS analysis of VI further supported this structure assignment. FAB-MS (m/z): 537.6 (M + 1). HPLC analysis of the products formed from NDY3 (II) oxidation by HRP indicated the presence of acetanilide (V, Rt=9.5 min), 1,2-naphthoquinone (W, Rt=10.8 min), and a minor product with a retention time of 14.5 min (Figure 1,Za). GC-MS analysis of the reduced acetylated reaction yielded the di-acetate of 1,2-dihydroxynaphthalene: MS (m/z): 244 (M) (4.5%),202 (M - COCH2) (9.5%), 160 (M - 2COCHz) (100%). The degradation products from Lip oxidation were similar; however, the 14.5 min product was produced in larger amounts (Figure 1,Zb). This product has been tentatively identified by mass spectral analysis as a derivative of NDY3 in which the 2,3- or 2,6-positions of the naphthol is oxidized to the corresponding 2,3- or 2,6-quinone. Dithionite reduction and acetylation of the product yielded the tertra-acetate derivative of 2,3- or 2,6dihydroxynaphthylamine WS (m/z): 343 (M), 301 (M - COCH2), 259 (M - 2COCH2), 217 (M - 3COCH2), 175 (M - 4COCH2) (100%)]. Mn(III)- malonate oxidation of NDY3 provided products similar to HRP, although the product yields were lower than in HRP or Lip oxidations (Figure 1,2c). Acetanilide (V) and 1,2-naphthoquinone (VII) were quantified using HPLC (Table I). HRP oxidation of NDY3 yielded almost equal amounts of acetanilide (V) and naphthoquinone (VII), whereas LiP or Mn(III)-oxidation of NDY3 generated more acetanilide (V) than 1,2-naphthoquinone (VII) (Table I). Peroxidase degradation of NDMDY3 (III) yielded two main products: 1,2-naphthoquinone (VII) and 3,5-dimethylacetanilide (VIII) (Figure 1). HPLC analysis indicated the presence of trace quantities of 2,6-dimethyl-1,4- benzoquinone (IX), and GC-MS analysis of the reduced acetylated product Retention time (min)

Figure 2. HPLC analysis of products formed during NDY3 oxidation by (a) HRP, (b) Lip, and (c) Mn(III)-malonate complex. Peak 1, acetanilide (V, Rt = 9.5 min); Peak 2,1,2-naphthoquinone (VII, Rt = 10.8 min); Peak 3, NDY3 (Rt = 15.2 min). Table I. Quantification of 1,2-naphthoquinone and acetanilide formed, and dye remaining, after oxidation of NDY3 (11) by peroxidases.

NDY3 Enzvme 1.2-Na~hthoquinone Acetanilide remaining

HRP 0.143 pmol 0.148 pmol 0.296 pmol

Lip 0.072 pmol 0.262 pmol 0.325 pmol

Mn(1II)- 0.027 pmol 0.068 pmol 0.732 pmol malonate indicated 2,6-dimethyl-1,4-dihydroquinone. MS (m/ z): 222 (M) (7%), 180 (M - COCH2) (17%), 138 (M - 2COCH2) (100%). In a separate experiment, 4- acetamido-2,6-dimethylphenol,a possible degradation product of NDMDY3, was oxidized to 2,6-dimethyl-1,4-benzoquinone(IX) by HRP and Lip. Deuterium incorporation experiments. These experiments were performed in the absence of Tween-80. Under these conditions, only DY3 (I) had limited solubility; NDY3 (11) and NDMDY3 (111) were totally insoluble in the reaction medium. For this reason, deuterium incorporation experiments were performed only with DY3. We investigated incorporation of deuterium from D20 and dioxane-d8 into acetanilide, formed from DY3 reaction with LiP and HRP. The mass spectrum of standard acetanilide showed prominent peaks corresponding to mass units 135 (M) (22%), 93 (M-COCH2) (loo%),77 (C,~H~+)(5.5%) and 65 (C5~5+)(15%) (Figure 3a). If deuterium was incorporated into the aromatic ring, the masses of these peaks would be increased by one mass unit and the percentage of deuterium incorporation could be determined from the ratio of the intensities of mass peaks 135 and 136. LiP and HRP provided similar results. No incorporation from D20 was observed (Figure 3b). However, in the peroxidase reactions containing H20 and 1,4-dioxane-d8 (99%D, 5% v/v), approximately 67% and 55% deuterium incorporation was observed for the LiP and HRP reactions, respectively (Figure 3c). In the presence of Tween-80, deuterium incorporation into acetanilide from 1,4-dioxane-dg was not observed. DY3 metabolites from whole cultures. The DY3- and NDY3-added cultures were totally decolorized after 48 h incubation. HPLC analysis of the ethyl acetate extracts of these cultures indicated an intense HPLC peak corresponding to acetanilide. DY3 dimer (VI), the major DY3 oxidation product of peroxidase reactions, did not form in appreciable amounts. The yield of acetanilide from whole culture degradation of DY3 (I) and NDY3 (II) was estimated to be 25% and 21% respectively. 77

0 IIIIIi,I,. ,I1. . I, 50 60 70 80 90 100 110 120 130 140

Figwe 3. Mass spectra of (a) standard acetanilide, (b) acetanilide isolated from a Lip reaction with DY3 containing D20 (99%D), and (c) acetanilide isolated from a LIP reaction containing l,4-dioxane-d8 (99%D, 5% v/v). IV. DISCUSSION

Very little is known about either the nature of degradation products formed during azo dye degradation by peroxidases or the mechanism by which peroxidases oxidize dye molecules. In this study, peroxidase-catalyzed oxidations of the azo dye DY3 (I) and its analogs were examined. DY3 (I) is one of the azo dyes extensively mineralized by lignin-degrading cultures of P. chrysosporium (16). Acetanilide (V), trace quantities of 4-methyl-1,2-benzoquinone(IV), and a dimer of DY3 (VI) have been identified as the degradation products formed from peroxidase-dependent oxidation of DY3 (I) (Figure 1). The DY3 dimer (VI) appears to be the major product formed in this reaction. Since the yields of acetanilide (V) and 4-methyl-1,2-benzoquinone(IV) were low, peroxidase degradation of the naphthol analog of DY3 (NDY3,II) was investigated. Oxidation of NDY3 (II) by Lip, HRP, or Mn(III)-malonate complex yields 1,2-naphthoquinone (VII) and acetanilide (V) as the major products (Figure 1,2). Yields of acetanilide (V) and 1,2-naphthoquinone (VII) produced from NDY3 (11) oxidation are provided in Table I. In a peroxidase reaction, Hz02 initially oxidizes the ferric enzyme by two electrons to produce compound I, a ferryl-oxo species with a n-cation radical character (4,15). Donation of two electrons from a substrate such as a phenol reduces compound I first to compound 11, and then to the native enzyme. MnP is different from other peroxidases in that Mn(I1) is its preferred substrate, and it oxidizes Mn(II) to Mn(III). The enzyme-generated Mn(III) is complexed by either dicarboxylic acid (such as malonate) or an a- hydroxy acid, and this complex diffuses from the enzyme and oxidizes substrate (6,lB). Chemically prepared Mn(II1)-malonate and Mn(III)-lactate have been shown to mimic MnP reactions (6). Figure 4 presents a mechanism for NDY3 degradation that explains the formation of the products. According to this proposal, the H202-oxidized forms of a peroxidase or Mn(II1)-malonate complex successively oxidize the naphthol ring of NDY3 by two electrons to produce a carbonium ion on the C- 1 carbon of the naphthol ring. Next, a nucleophilic attack by water generates an unstable tetrahedral intermediate (X), which breaks down to produce 12- naphthoquinone (VII) and 4-acetamidophenyldiazene (XI) (Pathway a, Figure 4). Previous studies have shown that phenyldiazenes are unstable and that - NHCOCHB

pathway b - pathway a - ionic radical intermediate / \ intermediates

Figure 4. Two possible mechanisms for the degradation of the azo dye NDY3 (II) by peroxidases. The reaction appears to follow pathway a, containing radical intermediates, and not pathway b, containing an ionic intermediate. oxygen readily oxidizes them by one electron to generate the corresponding phenyldiazene radicals (8,lO). The latter is also unstable and readily loses the azo linkage as nitrogen via homolytic bond breakage reactions to yield a phenyl radical (1,8,10). The latter abstracts a hydrogen radical from its surroundings to yield a stable aromatic compound. A similar mechanism might explain the formation of acetanilide (V) from 4- acetamidophenyldiazene (XI) (Figure 4). Peroxidases can also oxidize phenyldiazenes, however, the reaction products appear to be similar to that of oxygen reaction. For example, HRP transforms unsubstituted phenyldiazene to benzene (1). Thus, peroxidases might also be catalyzing the conversion of 4-acetamidophenyldiazene to acetanilide. It is possible that two of the phenyl radicals could couple to yield diacetamidobiphenyl (XIV), but there is no evidence from this study for the formation of biphenyl-type products in the degradation of DY3, NDY3, or NDMDY3. Reduction appears to be the preferred pathway for phenyl radical degradation, because acetanilide (V) and 3,5-dimethylacetanilide (VIII) are the only major products identified from the acetamidophenyl ring portion of DY3 (I), NDY3 (II), and NDMDY3 (III). Trace quantities of 2,6-dimethyl-1,4-benzoquinone(IX) are formed from NDMDY3 degradation (LU) (Figure 1). It is possible that small amounts of 3,5- dimethylacetamidophenyl radical are scavenged by oxygen to produce the corresponding hydroperoxide (9). The latter may decompose to produce 4- acetamido-33-dimethylphenoxyradical, which may be further oxidized by the peroxidase to yield 2,6-dimethyl-l,4-benzoquinone(IX). Our observation that peroxidases oxidize 4-acetamido-3,5-dimethylphenolto IX further supports this proposal. Nevertheless, 4-acetamidophenyl radical scavenging by oxygen is not a preferred pathway for acetamidophenyl radical degradation. Our observations are in agreement with previous findings that phenyl radicals react sluggishly with molecular oxygen (20). Formation of 1,2-naphthoquinone (VII) and acetanilide (V) can be explained through yet another pathway, which does not involve phenyldiazene as an intermediate (Pathway b, Figure 4). In this route, the tetrahedral intermediate (X) breaks down in a single step to produce 1,2- naphthoquinone (VII), molecular nitrogen, and the phenyl carbanion of acetanilide. The latter could abstract a proton from water to yield acetanilide (V). To identify the correct mechanism, deuterium incorporation experiments were performed with DY3. If pathway b is operative, deuterium incorporation from D20 into the aromatic ring of acetanilide (V) is expected. However, this was not observed (Figure 3b), suggesting that the degradation of DY3 and its analogs does not proceed through pathway b. Lack of incorporation from water further suggests that a 4-acetamidophenyl carbanion, which could be formed via a one-electron reduction of the phenyl radical, also is not an intermediate in these degradations. Therefore, acetamidophenyl radical must be abstracting a hydrogen radical from its environment to form acetanilide. The potential hydrogen radical donors in the reaction medium include Tween-80, l,4-dioxane, enzyme, and possibly H202. In the presence of 1,4-dioxane-d8 (99%D, 5% v/v), and in the absence of Tween-80, 67% deuterium incorporation into acetanilide was observed in the LiP reaction (Figure 3c), and a 55% incorporation was found in the HRP reaction. However, in the presence of Tween-80 (1%v/v), deuterium incorporation from 1,4-dioxane-dg (5% v/v) was not observed, suggesting that Tween-80 competes with dioxane for the reduction of acetamidophenyl radical (XIII). These results support our proposal that a phenyl radical is an intermediate in azo dye degradation, and that this radical abstracts a hydrogen radical from its surroundings to produce a stable aromatic compound (20). DY3 and NDY3 degradations by whole cultures of P. chysosporium produced acetanilide (V)in 20-25% yield. DY3 dimer (VI) does not appear to be formed from DY3 degradation. This is either because dimerization is not favored in whole cultures or that the DY3 dimer, a bis-phenolic compound, is further oxidized by peroxidases. In any case, identification of acetanilide from in vivo experiments suggests strongly that peroxidases are involved in the metabolism of azo dyes by P. ch ysosporium. In summary, we have described the first mechanism for the peroxidase-catalyzed degradation of azo dyes which explains the formation of all the products. Salient features of this mechanistic proposal are the release of azo linkage as molecular nitrogen, and the production of phenyldiazene and phenyl radicals as intermediates. Mammals produce a number of peroxidases including myeloperoxidase, lactoperoxidase, and thyroid peroxidase. The proposed mechanism for azo dye degradation by fungal and plant peroxidases might also be applicable to mammalian peroxidases. V. REFERENCES

1 Ator, M. A., and Ortiz de Montellano, P. R. (1987) J. Biol. Chem. 262, 1542-1551.

2. Bellamy, F. D., and Ou, K. (1984) Tetrahedron Lett. 839-842.

3. Cripps, C., Bumpus, J. A., and Aust, S. D. (1990) Appl. Environ. Microbiol. 56,1114-1118.

4. Dunford, B. (1982) Adv. Inorg. Biochem. 4,4-68.

5. Furnis, B. S., Hamaford, A. J., Smith, P. W. G., and Tatchell, A. R (1989) Vogel's Textbook of Practical Organic Chemistry, pp 919-920 and 946-953, John Wiley and Sons, Inc., New York.

6. Glenn, J. K., Akileswaran, L., and Gold, M. H. (1986) Arch. Biochem. Biophys. 251,688-696.

7. Gold, M. H., Kuwahara, M., Chiu, A. A., and Glenn, J. K. (1984) Arch. Biochem. Biophys. 234,353-362.

8. Huang, P.-K. C., and Kosower, E. M. (1968) J. Am. Chem. Soc. 90,2367- 2376.

9. Ingold, K. (1969) Accounts Chem. Res. 2,l-9.

10. Kosower, E. M., Huang, C. P.-K., and Tsuji, T. (1969) J. Am. Chem. Soc. 91,2325-2329.

11. Ono, A., Hiroi, M., and Shimazaki, K. (1984) Chem. Ind. (London), 75.

12. Pasti-Grigsby, M. B., Paszczynski, A., Goszczynski, S., Crawford, D. L., and Crawford, R. L. (1992) Appl. Environ. Microbiol. 58,3605-3613.

13. Paszczynski, A., and Crawford, R. L. (1991) Biochem. Biophys. Res. Commun. 178,1056-1063.

14. Paszczynski, A., Pasti-Grigsby, M. B., Goszczynski, S., Crawford, R L. and Crawford, D. L. (1992) Appl. Environ. Microbiol. 58,3598-3604.

15. Renganathan, V., and Gold, M. H. (1986) Biochemistry 25,1626-1631.

16. Spadaro, J.T.,Gold, M. H., and Renganathan, V. (1992) Appl. Environ. Microbiol. 58,2397-2401. 17. Wariishi, H., and Gold, M. H. (1990) J. Biol. Chem. 265, 2070-2077.

18. Wariishi, H., Valli, K., and Gold, M. H. (1992) J. Biol. Chem. 267, 23688- 23695.

19. Pasczynski, A., Pasti, M. B., Gosczynski, S., Crawford, D. L., and Crawford, R. L. (1991) Enzyme Microb. Technol. 13,378-384.

20. Russell, G. A., and Bridger, R. F. (1963) J. Am. Chem. Soc. 85,3765-3766. CHAPTER FOUR

PEROXIDASE-CATALYZED OXIDATION OF AZO DYES WITH PHENYLAZO SUBSTITUTIONS GENERATES BENZENE

I. INTRODUCTION

Decolorization of azo dyes by lignin peroxidase (Lip), manganese peroxidase (MnP), and bacterial peroxidases has been extensively documented (8-12). This laboratory recently proposed a mechanism for the breakdown of the azo dye Disperse Yellow 3 (DY3) (2-(4'-acetamidopheny1azo)-4- methylphenol) and a naphthol analog (NDY3) by horseradish peroxidase (HRP), Lip, and Mn(III)-acetate (a MnP mimic) (13, chapter 3). The peroxidase oxidizes the azo dye at an aromatic hydroxyl substituent to form an aryl radical, which is then oxidized to an aryl cation located at the azo-bearing carbon. Attack by water, and subsequent hydrolysis, then occurs. A 4- acetamidophenyldiazene intermediate, expelled during hydrolysis, oxidatively degrades to acetanilide via phenyldiazenyl and phenyl radicals. Deuterium is incorporated into acetanilide from 1,4-dioxane-dg, but not from D20. In this chapter, evidence is presented for the extensive oxidation of seven azo dyes, including one heterocyclic dye, three sulfonated dyes, and three hydrophobic dyes, by HRP. Three of the dyes tested, Solvent Yellow 14, N,N-dimethyl-4-phenylazoaniline,and chrysoidin, are recognized as carcinogens (4-733). All of these azo dyes contain phenylazo groups, and all produce the carcinogenic compound benzene upon oxidation. A mechanism for the formation of benzene during HRP catalyzed oxidation of azo dyes is proposed. 11. MATERIALS AND METHODS

Chemicals. Solvent Yellow 14 (I), 5-hydroxy-3-methyl-4- phenylazopyrazole (ID), chrysoidin (V), Naphthol Blue Black (VI), and Direct Red 23 (VII) were all purchased from Aldrich (Milwaukee, WI) (Figure 1). I was recrystallized from ethyl acetate/hexane to give pure material. 4- hydroxy-3-phenylazonaphthalenesulfonic acid (11) was synthesized by diazotizing aniline with nitrous acid at 0-50 C and then coupling it to 4- hydroxynaphthalenesulfonic acid, as described (1,14). N,N-dimethyl-4- phenylazoaniline (IV) was prepared as described previously (15). II, VI, and VII were purified by silica gel column chromatography with hexane/ethyl acetate/methanol, and V with hexane/ethyl acetate, as eluents, respectively. Negative ion fast atom bombardment mass spectroscopy of II in glycerol gave a parent ion of m/z 327. III was used as received. Enzyme reactions. Horseradish peroxidase (HRP) (type Vla), glucose, and bovine serum albumin were used as received from Sigma, St. Louis, MO. Reactions contained HRP (0.5 mg), an azo dye (1.25 pmol), and 1,4-dioxane (5% (v/v)) (J.T. Baker, Phillipsburg, N.J.) dissolved in 2 rnl of phosphate buffer (lOmM, pH 6). The reactions of hydrophobic dyes I, III, IV, and V contained Tween 80 (1% (v/v)) (Sigma, St. Louis, MO) to increase dye solubility. Hydrogen peroxide (H202) (1.25 pmol) (Sigma, St. Louis, MO) was added in ten aliquots of 20 yl over a 1h period using a gas tight syringe (Hamilton, Reno, NV). Because VI and VII each contained two HRP reaction sites, 2.5 pmol H202 was added, in ten 20 pl aliquots, to the reactions of these two dyes. For deuterium labeling experiments, l,4-dioxane-ds (5% (v/v)) (Aldrich, Milwaukee, WI) was used in place of 1,4-dioxane. Duplicate reactions were performed with vigorous stirring in sealed 14 ml vials capped with Teflon-Silicone septa (Pierce, Rockford, IL), at room temperature. Product analysis. For head space analysis, 10 ml of the gas was analyzed by a gas chromatograph (GC) (HI? 5790A, Hewlett Packard Co.) fitted to a mass spectrometer (Finnigan 4000, Finnigan Associates), as described (16). The GC column (Econocap Carbowax ,Alltech Associates, Deerfield, IL) was initially at -500 C, to cryofocus volatile organic compounds. The injector was maintained at 2000 C, and the column temperature was increased to 2700 C at a rate of 150 C/min. Electron impact mass spectra (70 eV) were collected for compounds eluting from the GC column. Benzene quantification was made by isotope VII

Figure 1. The seven dyes (I-VII) used in the HRP reactions. All contain one phenylazo substitution. 11, VI, and VII are sulfonated, water-soluble dyes. I, III, IV, and V are hydrophobic dyes. dilution. The internal standard benzene-d6 (125 nmol in 17.5 pl MeOH) (Aldrich, Milwaukee, WI) was added to the sealed final reaction mixture via a gas-tight syringe and allowed to stir vigorously, and equilibrate, for 15 min. Head-space analysis was performed as above, and comparison of the integrated mass spectral ion signal areas for benzene (m/z 78) and benzene-& (m/z 84) afforded a direct molar ratio of the two benzene speaes. Monodeuterated benzene production (m/z 79) was assessed by directly comparing the areas of the signals for m/z 78 and m/z 79. Identification and quantitation of non-volatile products, in comparison with pure standards, was made by HPLC (Waters Associates). Hydrophobic dye reactions were saturated with NaC1, adjusted to pH 2.5 with dilute HC1, and extracted with ethyl acetate. The extracts were gently evaporated to dryness with nitrogen. Hydrophobic products (injected in methanol (J.T. Baker, N.J.)) were analyzed by using a C-18 reverse phase column (Separations Group, Hesperia, CA) eluted with 20% methanol for 2.5 min, followed by a gradient to 100% methanol over 10 min (13). Water-soluble dye reactions were made to 0.1 M in phosphate (pH 7) through buffer addition and were analysed on the C-18 reverse-phase column by using an ion-pair chromatography technique. The column was eluted with phosphate buffer (0.1 M, pH 7) for 5 min, followed by a 10-min gradient to 100% 1:l MeOH:H20 (25). Flow rates were 1 ml/min and compound elution was monitored at 254 nm.

111. RESULTS

All of the azo dyes studied here were extensively oxidized by HRP (Table I). Head space analysis of these reactions indicated the presence of benzene in yields ranging from 3.7% to 11.8%. The products formed from dyes I and 11 were analyzed in detail (Figure 2). The major products identified from reaction of I were 1,2- naphthoquinone (12NQ) (7.6%) and benzene (3.7%). Reaction of 11 yielded 1,2- naphthoquinonesulfonate (12NQS) (5.3%)' benzene (9.8%), and another major product, which was not able to be identified. Nitrobenzene (1.7%) and acetanilide (5.0%) were identified as degradation products of VI and VI. Table I. Benzene yields determined for the HRP oxidation of dyes I - VII. c

Substrate Benzene vield (%I

a - reaction contained 1%Tween 80 (v/v) and 5% 1,4-dioxane (v/v) b - reaction contained 5% 1,4-dioxane (v/v) c - yields are expressed as molar percentage of dye substrate added to reaction Figure 2. Products identified after the HRP-catalyzed oxidation of dyes I and II. Table 11. Product yields, and dye remaining, after HRP oxidation of dye 11 in the presence of different reaction additives. a

Additive Benzene Yield (%) 1.2NOS vield (%)

glucose 7.5 2% (w/v) bovine serum albumin 3.1 2.5% (w/v)

a - 1,2NQS - 1,2-naphthoquinone-4-sulfonicacid respectively. Reaction products other than benzene were not determined for III, TV, and V. When 1,4-dioxane-d8 (5% v/v) was substituted for 1,4-dioxane (5% v/v) in the oxidation reaction of II, ~70%of the benzene product was monodeuterated. MS (m/z): 80 (8%), 79 (loo%), 78 (41%), 77 (10.2%), 63 (3%), 51 (24%),50 (18%). The reaction of dye 11 could be performed in the absence of a 1,4- dioxane due to the dye's aqueous solubility. 1,4-Dioxane is not found physiologically. To test whether benzene could be formed in the presence of physiological molecules, the biomolecule glucose (2% w/v) and the protein bovine serum albumin (BSA) (2.5% w/v) were added to reactions in place of l,4-dioxane. Benzene and 1,2NQS were produced in both experiments (Table II). 1,2NQS, but not benzene, was produced in the absence of additives.

IV. DISCUSSION

The products isolated from the reactions of I and II enable us to propose a mechanism for the degradation of I as follows (Figure 3). First, the H202- oxidized forms of HRP oxidize the naphthol ring of I by two electrons to yield a carbocation at carbon C-1. Water nucleophilically attacks the cation to produce a tetrahedral intermediate (Vm) that collapses, forming phenyldiazene (IX) and 1,2-naphthoquinone (X). A one-electron oxidation of phenyldiazene (IX) by molecular oxygen, metals, or HRP produces highly unstable phenyldiazene radical (XI) which then homolytically decomposes to phenyl radical (XII) with the liberation of molecular nitrogen (18-22). The phenyl radical (XU)then abstracts a hydrogen from an organic species in its surroundings to yield benzene (MII) (23). Incorporation of deuterium (>70%)from l,4-dioxane-dg into the benzene product during oxidation of 11 supports a pathway involving a phenyl radical intermediate (23). Production of nitrobenzene and acetanilide during oxidation of VI and VII, respectively, also supports this mechanism. The amounts of acetanilide (5.0%) and benzene (4.3%) found after degradation of VII are nearly equal, consistent with VII being constructed from analogous halves.

Several studies have postulated that phenyldiazene, phenyldiazene radical, and phenyl radical are the three key succesive intermediates involved in benzene and nitrogen production When phenylhydrazine (Ph-NH-NH2) and phenyldiimides (Ph-N=N-C02R) are oxidized by metals and heme enzymes benzene and nitrogen are produced (18-22). The authors of those same studies have postulated that phenyldiazene, phenyldiazene radical, and phenyl radical are the three successive intermediates leading to those products. Expulsion of the phenyldiazene (IX) leaving group from the tetrahedral intermediate (W)is analogous to chloride, sulfite, and methoxide expulsion in the degradation of chlorophenols, hydroxybenzenesulfonic acids, and tetramethoxybenzene by peroxidases (24- 26). Benzene was also produced when glucose or bovine serum albumin were present in the reaction of II (Table I,). In the absence of additives no benzene was produced. These results indicate that common physiological biochemicals and biomolecules can reduce the phenyl radical. Low yields of the benzene and quinone products may be attributable to the high reactivity of phenyl radicals toward other radical species, quinones, hemes, and numerous other compounds (21-23). Formation of unextractable coupled products, such as quinone-radical or heme-radical adducts, may consume large quantities of the phenyl radicals and naphthoquinones that are formed. Solvent Yellow 14 (I), N,N-dimethyl-4-phenylazoaniline(IV), and chrysoidin (V) are carcinogens, thought to be due largely to reduction of the azo linkage to reactive aromatic amines (2,4-733). An intriguing possibility is that the carcinogenicity of these dyes may also be related to the formation of phenyl radicals or benzenes during physiological oxidation by peroxidases. A previous study on the HRP-catalyzed degradation of I showed that the phenylazo portion of the dye yields a product that can bind to DNA (27). The products of HRP oxidation of I and I1 are quite similar to products formed during previous studies on analogous compounds. Recent work in this laboratory with a Fenton-like chemical oxidation system (Fe(III)-H202) yields benzene as a product from I (16,chapter 5). Ozonolysis of I yields benzene and 1,2-naphthoquinone as the major products (29). Photocatalytic degradations of I, 4-arylazonaphthols, and 3-(2'-methoxypheny1azo)-4- hydroxynaphthalenesulfonic acid (an analog of II) yield quinones and benzenes analogous to the products found in this study (30-32). Taken together, these data suggest that similar degradation mechanisms may exist for azo compounds undergoing chemical, photochemical, or biological oxidation. Formation of benzene, a volatile product, during oxidations of these dyes may explain the low mass balances observed for phenylazo substituted dyes during mineralization studies with P chrysosporiurn cultures (15,chapter 2). Under the ligninolytic culture conditions used in those experiments, there exist high levels of the peroxidases LiP and MnP, and benzene generation would be expected during dye degradation (37,39). The radiolabeled benzene probably escaped the culture vessel during purging with oxygen since benzene is not trapped by the toluene-ethanolamine C@ trap that was used; a trapping solvent such as 2-methoxyethanol is required (16,40,chapter 5). HRP has been viewed as a model for the large families of plant, bacterial, fungal, and mammalian extrahepatic peroxidases which appear to be closely related structurally and functionally (17,36-38). The use of peroxidases and peroxidase-producing organisms to treat azo dye wastes has been proposed (3). Azo compounds are also being investigated medically for the treatment of cancers and human immunodeficiency virus (HIN)(34,35). The mechanism presented here may help address the products expected from degradation of azo dyes in natural environments, remedial waste treatment systems, and mammalian metabolism. V. REFERENCES

1. Zollinger, H. (1987) In Syntheses, Properties, and Applications of Organic Dyes and Pigments, VCH, New York.

2. Chung, K.-T., and Cerniglia, C.E. (1992) Mutation Res. 277,201-220.

3. Chung, K.-T., and Stevens, S.E., Jr. (1993) Environ. Toxicol. Chem. 12, 2121-2132.

4. Kadlubar, F.F. (1987) In IS1 Atlas of Science: Pharmacol. 1, 129-132.

5. Chung, K.-T. (1983) Mutation Res. 114,269-281.

6. National Institute of Environmental Health Sciences (1993) In Environ. Health Perspect. 101 (Suppl. 1).

7. Matthews, E.J., Spalding, J.W., and Temant, R.W. (1993) Environ. Health Persped. 101 (Suppl. 2), 347-482.

8. Pasti-Grigsby, M.B., Paszczynski, A., Goszczynski, S., Crawford, D.L., and Crawford, R.L. (1992) Appl. Environ. Microbiol. 58(11), 3605-3613.

9. Paszczynski, A., Pasti, M.B., Goszczynski, S., Crawford, D.L., and Crawford, RL. (1991) Enzyme Microb. Technol. 13,378-384.

10. Cao, W., Mahadevan, B., Crawford, D.L., and Crawford, R.L. (1993) Enzyme Microbiol. Technol. 15, 810-817.

11. Cripps, C., Bumpus, J.A., and Aust, S.D. (1990) Appl. Environ. Microbiol. 56(4), 1114-1118.

12. Ollikka, P., Alhonrnaki, K., Leppanen, V.-M., Glumoff, T., Raijola, T., and Suorninen, I. (1993) Appl. Environ. Microbiol. 59(12), 4010-4016.

13. Spadaro, J.T., and Renganathan, V. (1994) Arch. Biochem. Biophys. (in press).

14. Furnis, B.S., Hamaford, A.J., Smith, P.W.G., and Tatchell, A.R. (1989) In Vogel's Textbook of Organic Chemistry, pp. 950-952, JohnWiley and Sons, New York.

15. Spadaro, J.T., Gold, M.H., and Renganathan, V. (1992) Appl. Environ Microbiol. 58(8), 2397-2401. Spadaro, J.T., Isabelle, L., and Renganathan, V. (1994) Environ. Sci. Technol. (in press).

Dunford, H.B. (1986) In Molecular and Physiological Aspects of Plant Peroxidases, (Greppin, H., Penel, C., and Gaspar, Th., Eds.), pp. 15-23, University of Geneva, Switzerland.

Hill, H.A.O., and Thornalley, P. J. (1981) FEBS Lett. 125(2), 235-238.

Nicholson, J., and Cohen, S.G. (1966) J. Am. Chem. Soc. 88(10), 2247- 2252.

Goldberg, B., Stern, A., and Peisach, J. (1976) J. Biol. Chem. 251(10), 3045- 3051.

Augusto, O., Kunze, K.L., Ortiz de Montellano, P.R. (1982) J. Biol. Chem. 257(1I), 6231-6241.

Huang, P.C., and Kosower, E.M. (1968) J. Am. Chem. Soc. 90(9), 2367- 2376.

Russell, G.A., and Bridger, R.F. (1963) J. Am. Chem. Soc. 85,3765-3766.

Hammel, K.E., and Tardone, P.J. (1988) Biochemistry 27,6563-6568.

Mur alikrishna, C., and Renganathan, V. (1993) Biochem. Biophys. Res.Cornmun. 197(2), 798-804.

Kersten, P.J., Kalyanaraman, B., Hammel, K.E., Reinhammar, B., and Kirk, T.K.(1990) Biochem. J. 268,475-480.

Stiborova, M., Afsaw, B., and Anzenbacher, P. (1988) FEBS Lett. 232(2), 387-390.

Goszczynski, S., Paszczynski, A., Pasti-Grigsby, M.B., Crawford, RL., and Crawford, D.L. (1994) J. Bacteriol. 176(5), 1339-1347.

Matsui, M., Shibata, K., and Takase, Y. (1984) Dyes and Pigments 5,321- 328.

Haag, W.R.,and Mill, T. (1987) Environ. Toxicol. Chem. 6,359-369.

Albini, A., Fasani, E., Pietra, S., and Sulpizio, A. (1984) J. Chem Soc., Perkin Trans 11 1689-1692. Griffiths, J., and Hawkins, C. (1977) J. Chem Soc., Perkins Trans. I1 747- 752.

Brown, M.A., and DeVito, S.C. (1993) Crit. Rev. Environ. Sci. Technol. 23(3), 249-324.

De Long, M. J., Prochaska, H. J., and Talalay, P. (1986) Proc. Natl. Acad. Sci. 83,787-791.

Gruszecka-Kowalik, E., Haugwitz, R.D., and Zalkow, L.H. (1992) Biochem. Biophys. Res. Commun. 187(3), 1409-1417.

Welinder, K.G. (1991) Biochim. Biophys. Acta 1080,215-220.

Gold, M.H., Wariishi, H., and Valli, K. (1988) ACS Symp. Ser. 389, 127- 140.

Huang, J., and Dunford, H.B. (1991) Arch. Biochem. Biophys. 287(2), 257-262.

Kirk, T.K. and Farrell, R.L. (1987) Ann. Rev. Microbial. 41, 465-505.

Abbott, C.K., Sorenson, D.L., and Sims, R.C. (1992) Environ. Toxicol. Chem. 11, 181-185. CHAPTER FIVE

DEGRADATION OF AZO DYES BY HYDROXYL RADICALS: EVIDENCE FOR BENZENE GENERATION

I. INTRODUCTION

Advanced oxidation processes (AOPs) use hydroxyl radical (*OH)to degrade organic compounds (1). Recent research has focused on applying AOPs to the detoxification and destruction of persistent hazardous wastes (1- 3). Chemical and photochemical reactions are used for *OHgeneration, usually in combination with hydrogen peroxide or ozone (1,2). In Fenton chemistry, a well known process for *OHgeneration, a ferrous or ferric salts are reacted with Hz02 to produce *OH,as shown in equations 1-3 (6).

In the photochemical reaction, *OHis produced via UV-H202, UV-Ti02, or UV-03 reaction (7-9). Fe*f -H202treatment has been suggested as an alternative method for removing dyes from industrial effluent (10-12); however, a detailed study of this process has not been performed. This study, using 14C-labeled dyes, demonstrates that *OH, generated via the reaction of ~e3+and H202 at pH 2.8, degrades hydrophobic azo dyes to C02. Evidence for the generation of benzene arising from *OHradical degradation of azo dyes with phenylazo substitutions is also presented. A probable mechanism for benzene 89 generation is proposed. Among the dyes used in this study, 4- phenylazoaniline is a mutagen, whereas, N,N-dimethyl-4-phenylazoaniline, Disperse Yellow 3, and Solvent Yellow 14, are carcinogens (4,5,13).

II. MATERIALS AND ME'IHODS

Chemicals. 14C-Labeled azo dyes were synthesized and purified as described by Spadaro et al. (14, chapter 2) (Table il). 4-Phenylazophenol (I) and 4-phenylazoaniline (ill) were obtained from Fluka, Ronkonkoma, New York. Disperse Yellow 3 (il), Disperse Orange 3 (V), and Solvent Yellow 14 (VI) were purchased from Aldrich Chemical Company, Milwaukee, Wisconsin. N,N- dimethyl-4-phenylazoaniline (IV) was synthesized as previously described (14). Ferric nitrate (Fe(N03h . 9H20) and H202 were purchased from Sigma Chemical Company, St. Louis, Missouri. Degradation of 14C-Iabeled azo dyes. A radiolabeled dye (50,000-72,000 cpm, 50 llM) in ethanol (250 llL) was added to deionized water (25 mL) containing Tween-80 (0.1% v Iv), ferric nitrate (2 mM), and nitric acid (1 mM). Tween-80, a nonionic detergent, was needed to solubilize the dyes into solution. The reaction was started with the addition of H202 (75 mM). The reaction flask was sealed with a gas exchange manifold and shaken on a rotary shaker (200 rpm) in the dark. 14C02 released due to 14C-labeled dye degradation was trapped with NaOH (1 N, 0.6 mL) kept in a polycarbonate vial attached to the manifold. After 10 h, shaking was stopped; the head space was purged with air for 15 min and the purged air was passed through 1 mL of 1 N NaOH to capture untrapped 14C02. After replacing the 14C02 trap with a fresh one, H202 (75 mM) was once again added to the reaction mixture, and the manifold resealed. The reaction continued for an additional 14 h, after which released 14C02 released was again collected. The amount of radioactivity in the 14C02 traps was determined using a liquid scintillation counter (Model LS-3133P, Beckman Instruments Inc., Fullerton, CA). Ecolite (ICN Biomedical) was used as the scintillation fluid. The sum of 14C02 released after 10 hand 24 h represented the total amount of dye degraded to CO2 in 24 h. Four replicate samples were used for each azo dye degradation experiment. 90

The quantity of water-soluble and methylene chloride-soluble products formed after 24 h of reaction was determined. The radioactive material remaining in the reaction mixture was determined by pooling 1 mL aliquots from pairs of reaction mixtures, then extracting the 2 mL portions with 2 mL methylene chloride. The radioactivity remaining in the aqueous phase, presumably due to water-soluble products, was quantitated. Radioactive material in the methylene chloride extract was estimated by subtracting the radioactive counts in the aqueous phase after extraction from the radioactive counts found before extraction. Determination of total volatile compounds. 14C-labeled azo dyes were degraed under reaction conditions similar to those described above except that release of volatile organics was determined only at 1.5,3,4.5, and 6 h. Simultaneously, the amount of 14C02 released was also determined, as described above. Volatile organic components were trapped by purging the head space with air and bubbling the purged air through a 2-methoxyethanol: toluene mixture (1:1, vIv) containing scintillation fluors (20). The amount of radioactivity trapped was determined by liquid scintillation counting. Identification of volatile organics. The nature of volatile organics formed from the degradation of 4-phenylazophenol (I), 4-phenylazoaniline (In), N,N-dimethyl-4-phenylazoaniline (IV), and Solvent Yellow 14 (VI) were determined. Reactions contained dye (50 ~M, added in 125 ~l methanol), Fe3+ (2 mM), nitric acid (1 mM) and H202 (75 mM), and were performed in the dark at room temperature for 1.5 to 4 h. In addition, 4-phenylazophenol (I) and Solvent Yellow 14 (VI) reactions contained 0.1% Tween-BO. The total reaction volume was 25 mL and reactions were conducted in shaken (200 rpm) 40 mL vials fitted with Teflon-Silicone septa (Pierce, Rockford, IL). The gas phase of the reaction mixture was analyzed by a gas chromatograph (HP 5790A, Hewlett Packard Company) coupled to a mass spectrometer (Finnigan 4000, Finnigan Associates). The GC column (Econocap Carbowax, Alltech Associates, Deerfield, IL) was maintained at an initial temperature of -500C and the injector was at 2000C. The column temperature was increased from - 500 to 2700C at the rate of 150C per min during analysis. HPLC and GC-MS analyses. HPLC analyses for hydrophobic compounds were performed using a C-1Breverse phase column (Separations Group, Hesperia, CA) and a gradient eluent consisting of water and methanol. The methanol concentration was maintained at 20% for 2 min; then it was 91

increased to 100% in 10 min and maintained at 100% concentration for an additional 10 min. Compound elution was monitored at 254 nm. GC-MS analyses were performed at 70 eV on a VG Analytical 7070E mass spectrometer fitted with an HP 5790AGC and a fused capillary column (15 m, DB-5,J& W Science). A temperature gradient was used in the GC separations. The initial temperature was 70°C, and it was increased to 3200C at a rate of 100C/min. Determination of acid products. 4-Phenylazoaniline (III), radiolabeled III (62,500cpm) and N,N-dimethyl-4-phenylazoaniline (IV) (150 11M,added in 30 ilL methanol) were reacted with Fe3+(2 mM), nitric acid (1 mM), and H202 (150 mM) in a total volume of 2 mL. The reactions, performed as above, were stopped after 6 hours by addition of NaOH (to pH 10). The reaction mixture was centrifuged to remove precipitates. The supernatant was acidified to pH 2, by addition of HCl, and extracted with ether. The extract was then gently evaporated, and the residue dissolved in 0.1% phosphoric acid. The residues were analysed by HPLC using a Supelcogel C-610H column (Supelco, Bellafonte, PA) with 0.1% phosphoric acid as eluent at a 0.4 mLI min flow rate. Compound elution was monitored at 210 nm. Products were identified by comparison to elution of standards, and peaks from the degradation of radiolabeled dye III were collected and assayed for radioactivity by liquid scintillation (0.5 mL eluent: 6 mL Ecolite scintillation fluid). Acid products were also analyzed by GC-MS after derivatization with either 2,4'-dibromoacetophenone (2,4'-DBA) or trimethylsilane (TMS). These reactions were performed on a 20 mL scale. For 2,4'-DBAderivatization, the pH was adjusted to 6.5 by NaOH addition to the completed reaction, and 2,4'- DBA was added to the reaction mixture (3mM final concentration). Then 20 mL of methanol was added and the mixture was refluxed for 30 min. Ethyl acetate extraction followed. The extract was evaporated, and the residues analyzed (19). For TMS derivatization, the reaction was acidified to pH 2 with HCl, saturated with NaCl, and extracted with diethyl ether. The extract was evaporated, and the residues heated (800C) with 150ilL bis(trimethylsilyl)- trifluoroacetamide (BSTFA):pyridine (2:1) for 5 min. 92

III. RESULTS

In this report, degradation of specifically 14C-labeled azo dyes by .OH, generated via the reaction between ferric nitrate and H202 at pH 2.8, was studied. The optimal concentrations of ferric nitrate and H202 required for dye degradation were identified by studying the effect of Fe3+ and H202 concentrations on the degradation of 4-phenylazo[U-14C]phenol (Ia) to 14C02 (Table I). Fe3+ at 2 mM and H202 at 75 mM concentrations provided the highest levels of mineralization. These concentrations of Fe3+ and H202 were subsequently used in all the dye degradation reactions. Since many of the azo dyes investigated are water-insoluble, Tween-80 was included to solubilize them. Preliminary studies on the effect of Tween-80 on the degradation of Ia to CO2 suggested that Tween-80 does not affect dye degradation significantly (data not shown). All the dyes examined were totally decolorized in 24 h indicating the loss of their chromophoric group (data not shown). Degradation of dyes to CO2 ranged from 8-30% (Table ll). In general, high levels of 14C02 evolution were observed with phenolic or aniline ring labeled azo dyes such as la, ill, and IVa (Table ll). About 30-62% of the aromatic ring appears to have been degraded to water-soluble products. The initial pH of the reaction was 2.8, and the final pH was 2.3. Degradation products. When analyzed by HPLC after 4 h, the Disperse Yellow 3 (ill) degradation reaction yielded two products (~= 3 and 13.85 min, peaks 1 and 2) (Figure 1a). The mass spectrum of the product with a retention time of 13.85 min corresponded to that of acetanilide. MS (m/z): 135 (46.6%); 93 (100%); 77 (4.2%). Acetanilide yield was estimated to be approximately 44.7 mole percent. HPLC analysis of the Disperse Orange 3 (V) reaction mixture also showed the presence of two products (~= 3 and 16 min, peaks 1 and 2) (Figure 1b). The product with a retention time of 16 min corresponded to nitrobenzene; GC-MS analysis of the methylene chloride extract of the reaction mixture confirmed the presence of nitrobenzene. MS (m/ z): 123 (61.9%); 77 (100%); 65 (11.6%). Nitrobenzene yield was estimated as 2.9 mole percent. HPLC analyses of the total reaction products from N,N-dimethyl-4- phenylazoaniline (IV), and Solvent Yellow 14 (VI) indicated the presence of only one peak with a retention time of 2.8 min (data not shown). Degradation products with HPLC retention times of 2-3 min appeared to be 93

Table 1. Effect of Fe3+ and H202 variation on the mineralization of 4- phenylazo[U-14C]phenol (Ia)a

Rea.ctfint Dye mineralized

2mM Fe3+ 29.2% 4mM Fe3+ 31.0% 6mM Fe3+ 30.9%

25mM H202 21.0% 50mM H202 25.0% 75mM H202 32.2% a - Optimization experiments contained 100 mM 4-phenylazo[U-14C]phenol, 25-75 mM H202, 2-6 mM ferric nitrate, and 1 mM nitric acid. Experiments were performed in triplicate, and the values reported are mean values. 14C02 was assayed after 9 h. Table ll. Degradation of azo dyes catalyzed by the reaction of Fe3+ and H202

Water-soluble CH2CI2 soluble Mass balance 14C-Substratea 14C02 (%) b compounds (%) I compounds (%) I determined (%)

4-Phenylazophenol 29.9 I 41.6 I 19.2 I 90.7 @-N=N-@-OH (Ia) I I 18.4 I 30.5 I 22.7 I 71.6 @-N=N-@-OH (Ib)

Disperse Yellow 3 OH

(II) I 15.8 I 43.4 I 36.3 I 95.5

(VIa) I 7.7 I 62.5 I 26.5 I 96.7 @-'='t HO

@-,=" (VIb) I 16.4 I 29.1 I 17.0 I 62.5

a - Asterisks (*)indicate the location of 14C-Iabel. \D b - The first standard deviations for 14C02 values ranged from -0.1% to -1.9%. ~ I I I I I 0 5 10 15 20 25 Retention time (min)

t 0 5 10 15 20 25 Retention time (min)

Figure 1. HPLC analyses of (a) Disperse Yellow 3 (II), and (b) Disperse Orange 3 (V) after 4-h reaction with Fe(III)/H202. Figure la. Peak 1, unknown (Rt = 3 rnin); peak 2, acetanilide (Rt = 13.8 min); peak 3, Disperse Yellow 3 (Rt = 18.7 min). Figure lb. Peak 1, unknown (Rt = 3 min); peak 2, nitrobenzene (Rt = 16 rnin). Disperse Orange 3 has a retention time of 18 min. due to water-soluble compounds. Hydrophobic compounds bind strongly to the C-18 HPLC column and consequently, display longer retention times. However, hydrophilic compounds (ie. aliphatic carboxylic acids) bind weakly to this column and thus exhibit shorter retention times. I-IPLC analysis of water soluble products indicated that 4- phenylazoaniline (111) and N,N-dimethyl-4-phenylazoaniline(IV) both produce formic (14.6 min, peak 4) and maleic (9.7 min, peak 2) acids (Figure 2). 4-~henylazo[~-14CIaniline(III) degradation leads to radiolabeled products that coelute with the formic and maleic acid peaks (Figure 2). The identities of the products eluting at 12.9 min and 6.6 min (peaks 1 and 3) are not known. GC- MS analysis of the 2,4'-DBA derivatives of products from degradation of IV indicated the presence of the p-bromophenacyl ester of formic acid (2-formyl- 4'-bromoacetophenone): MS (m/z): 244 (3.1%); 242 (3.1%); 185 (93.2%); 183 (100%); 157 (37.7%); 155(40.1%); 131 (1.9%); 129 (1.5%); 104 (5.6%); 76 (53.7%); 75 (53.1%). GC-MS analysis of the TMS derivatives of products from degradation of JY indicated the presence of the bis-TMS ether of maleic acid: MS (m/z): 260 (1.1%); 245 (24.7%); 215 (2.2%); 170 (4.3%); 148 (16.1%); 147 (100%); 133 (4.3%); 115 (5.4%); 83 (3.2%); 75 (11.8%); 73 (59.1%). Identification of benzene. At the end of each dye degradation experiment, the mass balance was calculated through addition of the radioactive counts found in the 14C02, methylene-chloride , and aqueous fractions (Table 11). This value was in the range of 88-97% for radiolabeled dyes Ia, 11, III, ma, V, and VIa. However, for dyes Ib, IVb, and VIb, in which the phenyl rings were 14C-labeled, the mass balance values were only 72%, 68% and 62% respectively (Table 11). In these experiments, the gas phase was analyzed only for 14C02. If the gas phase contained other volatile organic components, they probably were not captured by the 14C02 traps. In a separate 6 h experiment, evolution of C02 and volatile organics from the degradation of 14C-labeled dyes was determined simultaneously. All the phenyl-ring 14C- labeled azo dyes, 4-phenylazophenol (Ib), N,N-dimethyl-4-phenylazoaniline (IVb) and Solvent Yellow 14 (VIb), produced radioactive volatile organics, whereas, the aniline-ring 14C-labeled N,N-dimethyl-4-phenylazoanilinepa) and the phenolic ring 14C-labeled 4-phenylazophenol (Ia) did not (Figure 3). The amount of volatile products formed from Ib, IVb, and VIb degradation in the first 6 h was 8%, 20%, and 26%, respectively. Mass balance for phenyl-ring 0 5 10 15 20 Retention time (min)

Figure 2. HPLC analysis of water soluble products of phenylazoaniline (III) (--) and 4-phenylazo[~-l~C]aniline(111) (--) after 6 hours of degradation by Fe(III)/H202. Peak 1, unknown (Rt = 6.6 min); peak 2, maleic acid (Rt = 9.7 min); peak 3, unknown (13 rnin); peak 4, formic acid (Rt = 14.7 rnin). - Color --t-- 14~-~olatiles Q -14~0~

0 40-

0 30-

10-

0 1 2 3 4 5 6 Time (hr)

- Color -14~-~olatiles -14~0~

$ 20-

0 1 2 3 4 5 6 Time (hr)

Figure 3. Fe(III)/H202 degradation of (a) phenyl ring-labeled phenylazophenol (Ib), (b) phenyl ring-labeled Solvent Yellow 14 (VIb). Decolorization of I and VI was monitored at 350 and 482.5 nm, respectively. C - Color --t-- 14~-~otatiies -i4w2

0 1 2 3 4 5 6 Time (hr)

- Color -14~-~01atiies -"b2

0 1 2 3 4 5 6 Time (hr)

Figure 3. Fe(III)/H202 degradation of (c) phenyl ring-labeled N,N-dimethyl-4- phenylazoaniline (IVb), and (d) aniline ring-labeled N,N-dimethyl-4- phenylazoaniline (IVa). Decolorization of IV was monitored at 415 nm. Scan

Figure 4. GC-MS analysis of volatile organics. (a) elution profile for N,N- dime thyl-4-phenylazoaniline (IV) degradation products after Fe(III)/Hz02 reaction; and (b) mass spectrum at scan 610. Peak at scan 502 was an unknown contaminant also present in control reactions. labeled dyes Ib, IVb, and VIb determined after 6 h reaction was 94%, 86%, and 94% respectively. GC-MS analysis of the gas phase of the reaction mixture from N,N- dimethylphenylazoaniline (W)indicated the presence of only a single volatile product and the mass spectrum of this compound corresponded to that of benzene (Figure 4). Similar GC-MS analyses of the volatile products from phenylazophenol (I) and Solvent Yellow 14 (VI) degradation also indicated the presence of only benzene. Disperse Yellow 3 (11) and Disperse Orange 3 (V), two azo dyes with substituents on the phenyl azo portion, did not generate volatile organic products.

IV. DISCUSSION

This study demonstrates that ~e3+-~202treatment, at room temperature, degrades a substantial portion of hydrophobic azo dyes to water- soluble products and C02. There is definitive evidence for the generation of benzene when azo dyes with a phenylazo substitution are degraded with Fe3+- H202. These results suggest that the lower mass balances observed for these dyes in the mineralization experiments (Table 11) were because the volatile organic compounds were not captured. Figure 5 presents a probable mechanism for benzene formation in the -OH mediated degradation of azo dyes. According to this proposal, *OHadds to the carbon (C-4), in a hydroxy or amine-substituted ring, that bears the azo linkage. The resulting *OHadduct breaks down to produce phenyldiazene and a phenoxy radical. Phenyldiazene is extremely unstable; -OH, molecular oxygen, or Fe(IlI) can readily oxidize it by one electron to yield a phenyldiazene radical (15) (Figure 5). The latter intermediate is also unstable and cleaves homolytically to generate a phenyl radical and molecular nitrogen (15). The phenyl radical might abstract a hydrogen radical from *02H,Tween-80, or dye-degradation products to produce benzene. Phenyl radical is not likely to be scavenged by molecular oxygen since it reacts sluggishly with oxygen (16). As for the fate of the phenoxy radical, it might further react with *OHand oxygen leading to the aromatic ring degradation. Phenyldiazene A R

Aromatic Ring

/ f Aromatic Ring Degradation

Figure 5. A proposed mechanism for benzene generation during degradation of phenylazo-substituted dyes by hydroxyl radicals formed via the Fenton-like reagents Fe(TII)/H2@. The mechanism proposed for benzene formation also accounts for the observed nitrobenzene and acetanilide production from Disperse Orange 3 and Disperse Yellow 3, respectively. Surprisingly, azo dye degradation by -OH resembles peroxidase-catalyzed degradation of azo dyes described recently by this laboratory (17). Previous work on photocatalysis of azo dyes by Albini, et al., has shown that benzene is formed from the N,N-diethyl analog of N,and that toluene is formed from the @-methylanalog of VI (21). Benzene is a priority pollutant; it causes leukemia in humans and reproductive impairment in fish (18). Several water-soluble commercial azo dyes such as Naphthol Blue Black, Azogeranine, Chromotrope 2R and Direct Orange 2G have a phenylazo-substitution as part of the dye structure, and are likely to generate benzene when degraded with *OH. Textile industry effluent usually contains a mixture of dyes. This study cautions that the structure and concentration of individual azo dyes should be known before the effluent is treated using a process which relies on the hydroxyl radical chemistry. * VI. REFERENCES

1. Glaze, W.H. (1994) In Chemical Oxidations, Technologies for the Nineties, Vol. 3, (Eckenfelder, W. W., Bowers, A. R., and Roth, J. A., Eds.), pp. 1-ll,Technomic, Lancaster, Pennsylvania.

2. Eckenfelder, W.W. (1992) In Chemical Oxidations, Technologies for the Nineties, (Eckenfelder, W. W., Bowers, A. R., and Roth, J. A., Eds.), pp. 1-10, Technomic, Lancaster, Pennsylvania.

3. Lipczynska-Kochany, E. (1994) In Chemical Oxidations, Technologies for the Nineties, Vol. 3, (Eckenfelder, W. W., Bowers, A. R., and Roth, J. A., Eds.), pp. 12-27,Technomic, Lancaster, Pennsylvania.

4. Zollinger, H. (1987) Color Chemistry - Syntheses, Properties and Applications of Organic Dyes and Pigments, VCH, New York, pp 92-102.

5. Chung, K-T., and Cerniglia, C. E. (1992) Mutat. Res. 277,201-220.

6. Walling, C. (1975) Acc. Chem. Res. 8,125-131.

7. Al-Ekabi, H., and Serpone, N. (1988) J. Phys. Chem. 92,5726-5731.

8. Froelich, E. M. (1992) In Chemical Oxidations, Technologies for the Nineties, (Eckenfelder, W. W., Bowers, A. R., and Roth, J. A., Eds.), Technomic, Lancaster, Pennsylvania, pp 104-113.

9. Peyton, G. R., and Glaze, W. H. (1988) Environ. Sci. Technol. 22,761- 767.

10. Kuo, W. G. (1992) Water Res. 26,881-886.

11. Bigda, R. J., and Elizardo, K. P. (1992) Abstracts of Papers, 204th ACS National Meeting, Washington, DC, American Chemical Society, Washington, DC, Vol. 32, No. 2, pp 259-262.

12. Moore, S. B., and Antenucci, A. (1993) Abstracts of Papers, ACS national Meeting, Denver, Colorado;,American Chemical Society, Washington, DC, Vol. pp 465-466.

13. National Institutes of Environmental Health (1993) Environ. Health Perspect. 101 (Suppl. ), 121-123.

14. Spadaro, J. T., Gold, M. H., and Renganathan, V. (1992) Appl. Environ. Microbiol. 58,2397-2401. 15. Huang, P-K. C., and Kosower, E. M. (1968) J. Am. Chem. Soc. 90,2367- 2376.

16. Russell, G. A., and Bridger, R. F. (1963) J. Am. Chem. Soc. 85,3765-3766.

17. Spadaro, J. T, and Renganathan, V. (1994) Arch. Biochem. Biophys. (in press).

18. Sittig, M. (1980) Priority Toxic Pollutants - Health Impacts and Allowable Limits, Noyes Data Corporation, Park Ridge, NJ, pp 87-90.

19. Umeh, E.O. (1971) J. Chromatogr. 56/29-36.

20. Abbott, C.K., Sorensen, D.L., and Sims, R.C. (1992) Environ. Toxicol. Chem. 11,181-185.

21. Albini, A., Fasani, E., Pietra, S., and Sulpizio, A. (1984) J. Chem. Soc. Perkin Trans. I1 1689-1692. CHAPTER SIX

CONCLUSION

I. INTRODUCTION

The studies described in chapters two, three, four, and five demonstrate that Phanerochaete chrysosporiurn, peroxidases, and Fe(III)/H202 extensively degrade many hydrophobic azo dyes under aerobic conditions. Proposed mechanisms for the chemical and biological oxidation of azo dyes have also been presented. Previous studies had concluded that biological degradation of azo dyes under aerobic conditions was not feasible (25,26). Bacterial or chemical reduction under anaerobic conditions were viewed as the most likely methods for dye degradation (25). But, reductive degradation of the dyes produces potentially toxic aromatic amines (33). In this chapter, our findings are analyzed with respect to the work, over the past five years, of other laboratories in the field of aerobic azo dye degradation.

11. FUNGAL DEGRADATION OF AZO DYES

There has been a renewed interest in aerobic azo dye degradation by fungi and bacteria during the last five years (4). Initially, Cripps, et a1 (2), demonstrated that ligninolytic cultures of P. ch ysosporium decolorize three sulfonated azo dyes. Their study also found that non-ligninolytic cultures degrade a small portion of the azo dyes. All of these dyes were adsorbed to the fungal biomass, and this complicated the measurement of the true rate of dye decolorization and degradation. Furthermore, no products from the dye degradation were identified by the authors. More recently, Dr. Ronald L. Crawford and coworkers (at the University of Idaho, Moscow, ID) demonstrated that ligninolytic cultures of P. chrysosporium are capable of degrading 16 sulfonated commercial dyes, as well as up to 18 other synthesized structural variants of the sulfonated azo dyes (5,6,10). One study tested degradation 22 different dye structures. Many of the dyes contain a 4-(4'-R-pheny1azo)-benzenesulfonic acid structures (R = hydroxy or amino), with various substituents at carbons 2', 3', 5', and 6' (Figure 1). Excepting dyes 8 and 15, neither the substituent pattern nor the substituent identity appear to prevent sigruficant dye decolorization by the cultures. The actinomycete bacteria of the genus Streptomyces are known to partially degrade lignin, and they secrete extracellular peroxidases during the process (6,2032). Of the 22 dye structures tested in one study, 7 dyes (1- 4,9,19,32) were degraded by five different Streptomycete strains (Figure 1) (6). All the degraded dyes but dye 32 have hydroxyls para to the azo linkage and alkyl or alkoxy groups at positions meta to the azo linkage. In 1992, our laboratory and that Dr. Crawford independently established that ligninolytic cultures of P. chysosporium mineralize a nwnber of sulfonated and non-sulfonated azo dyes (1, chapter 2,7) (Figure 2). The extent of sulfonated dye mineralization was on the order of 19-35%, similar to that found for the hydrophobic dyes in our study (1, chapter 2). All of the sulfonated dyes were labeled only in the ring bearing the sulfonic acid substituent. The ligninolytic actinomycete Streptomyces chromofuscus mineralized only 3 of the sulfonated dyes (dyes 3,1932) to a minimum extent (less than 3.6%). Dyes 3 and 19 contain guaiacyl-type (0-methoxyphenol) substituent patterns. In both mineralization studies, dyes with guaiacyl-type substituent patterns were degraded to a greater extent than other dyes tested &chapter 2,7). This may reflect the fact that the lignin degrading system of the fungi and bacteria were developed to degrade lignin substructures.

A. Decolorization Studies

In 1988, Stiborova, et al. showed that horseradish peroxidase (HRP) degrades the carcinogenic azo dye Sudan I (9). A putative benzenediazonium ion, arising from the phenylazo portion of the Sudan I, appeared to bind to Figure 1. Structures of sulfonated dyes tested for decolorization by cultures of Phanerochaete chrysosporium and Streptomyces chromofuscus All, and by the peroxidase HRP and crude peroxidase-containing enzyme preparations from cultures of P. ch ysosporium. and S. chromojiiscus All. From Pasti Grigsby, et al. (6). Figure 2. Structures of radiolabeled sulfonated azo dyes tested for mineralization by ligninolytic cultures of Phanerochaete ch ysosporium and Streptomyces chromofuscus. The radiolabel was uniform in the rings containing the sulfonate function. From Paszczynski, et al. (7). DNA (Figure 3). This result purportedly explains the carcinogenicity of Sudan I. The only two products identified were hydroxylated analogs of the starting dye. The benzenediazoniurn ion was trapped by adding l-phenyl-3- methyl-5-pyrazolone (PMP), an azo coupling component, to the completed reaction. A new compound, formed in the mixture, matched the authentic dye 1-phenyl-3-methyl-4-phenylazo-5-pyrazolonein TLC mobility and in UV- visible absorbance, indicating that a benzenediazonium ion had indeed coupled with PMP. The more recent studies on azo dye degradation have focused on peroxidases from two ligninolytic organisms: the fungus Phanerochaete ch ysosporium; and bacteria of the genus Streptomyces. Cripps, et al. were the first to report decolorization of azo dyes by lignin peroxidase (Lip) from P. chysosporium, but no products were identified (2). The Crawford group established that the peroxidases isolated from P. chrysosporium are able to degrade a diverse group of sulfonated azo dyes (5,6,10). One study examined 22 different azo dye structures and their oxidation rates by LiP and manganese peroxidase (MnP) (6) (Figure 1). Lil? efficiently oxidized dyes 6,8,10,11,16-18/30, and 31. However, dyes containing phenolic groups para to the azo linkage were not degraded well by Lip. Dye 4 was not degraded by Lip at all. When two fluorine atoms were placed in the positions ortho to the hydroxyl groups, ie. azo dye 11, LiP oxidation improved dramatically. In fact, dye 11 was by far the best Lip substrate in comparison to all the dyes tested. MnP efficiently degraded dyes 1-4,9,10,19, and 32. But dye 11 could not be degraded by MnP. The two peroxidases appeared to complement each other in substrate specificity. The same studies also extensively addressed degradation of the azo dyes by HRP and by bacterial peroxidases isolated from Streptomyces spp. (5,6). HRP and the bacterial peroxidase from S. chrornofuscus generally show a pattern of substrate specificity similar to that for MnP, though the azo dye oxidation rates vary considerably. The specificity of the bacterial peroxidases appears to limit efficient degradation to those azo dyes containing guaiacyl- or syringyl-type substituent patterns. The peroxidases of bacterial spp. representing five other genera have also been tested for their ability to oxidize azo dyes (11). The peroxidases produced by a FEavobacterium sp and by a Frankia sp. both are able to degrade the model dye 4-(3',5'-dimethyl-4'- hydroxypheny1azo)-benzenesulfonic acid (Figure 1, dye 1) (11). DNA R NA prote~n

Figure 3. The oxidative cleavage of the dye Sudan I (Solvent Yellow 14) by HRP, as envisioned by Stiborova, et al. (9). The hydroxylated products 2 and 3 were observed, and the presence of benzenediazoniurn ion was indicated. As noted in chapter 1, the role of verartyl alcohol (VA) in peroxidase reactions is a matter of much debate. One theory is that VA protects the enzyme from inactivation by serving as a good substrate for LiP compound 11. The second theory suggests that VA cation radicals, formed during LiP oxidation of VA, function as mediators in the oxidation of compounds which are poor or non-substrates for LiP (21,22). Two recent studies have suggested that veratryl alcohol is required for the efficient oxidation of azo dyes by Lip, and that azo dyes are unsuitable substrates for regeneration of native Lip from Lip compound 11 (6,lO). The dyes tested are either phenolic or hydroxypyrazol type dyes. In another study, an aniline dye is effectively decolorized by Lip in the absence of veratryl alcohol (5). In all of these experiments, the concentration of Hz02 is routinely 5-10 times the concentration of starting substrate, and the amount of Lip used is very small. Gold and coworkers have shown that LiP is inactivated in the presence of poor substrates and excess peroxide (21). Thus, the function of VA may be to protect Lip from H202-induced inactivation. In our studies on the LiP oxidation of the naphthol analog of Disperse Yellow 3 (NDY3) (chapter 3) we avoided high concentrations of H202. NDY3 degradation did not appear to require veratryl alcohol. More recently, Ollikka, et al. reported the extensive decolorization of the azo dyes Congo Red and Methyl Orange (both aromatic amine dyes) by crude Lip, and varying decolorization by three purified Lip isozymes (12). These authors also demonstrated that 2mM veratryl alcohol increased Methyl Orange decolorization by pure LiP isozymes, but not by a crude, dialyzed LiP fraction from culture medium. However, veratryl alcohol addition did not change Congo Red decolorization, which remained at approximately 50%. In these studies, the H202 concentration was 13-19 times the concentration of the dye substrate. Again, excess I3202 may be responsible for LiP inactivation. The crude Lip fraction may have contained catalases or other H202- consuming enzymes which lowered the effective H202. concentration.

B. Proposed Mechanisms

In chapters 3 and 5, mechanisms for the oxidative breakdown of several azo dyes were reported. The mechanisms rest upon the formation of a phenyldiazene and a quinone as the initial dye cleavage products. The phenyldiazene is oxidatively degraded, and passes through both phenyldiazenyl and phenyl radicals as intermediates to a final benzene product. The HRP catalyzed degradation of the carcinogenic azo dye Sudan I (Solvent Yellow 14), as described by Stiborova, et al., purportedly involves cleavage of the dye at the azo bearing carbon on the naphthol component, to yield a benzenediazonium ion and a 1,2-dihydroxynaphthalene (9) (Figure 3). The benzenediazonium ion was only indirectly identified, and no product corresponding to the naphthol portion of the molecule was identified. Recently, Goszczynski, et al. have performed product analysis on the Lip, MnP, and S. chromofuscus peroxidase degradations of two azo dyes, and have proposed a bifurcating mechanistic pathway for the dye degradation reactions (Figure 4). One dye is a sulfonate, the other a sulfonamide. The sulfonate dye yielded the following products: a benzoquinone; 4- hydroxybenzenesulfonic acid; sulfanilic acid; nitrosobenzenesulfonic acid; benzenesulfonic acid; and several dimeric and trimeric aromatic compounds. The sulfonamide dye yielded products analogous to all of those listed above, but no nitrosobenzenesulfonamide, and an aminophenol was also found. The authors did not quantify the products. The qualitative product analysis led the authors to propose mechanisms, with dye cleavage either between the azo nitrogens or at the azo bearing carbon of the phenolic moiety (Figure 5). The first steps of the pathways are shared. The peroxidase oxidizes the azo dye to a cation, the cation is nucleophilically attacked by water, and the dye is hydrolyzed. The site of cation stabilization, and therefore the site of azo dye hydrolysis, is the key to whether the dye is cleaved symmetrically or asymmetrically about the azo linkage. Cleavage between the azo nitrogens occurs when the cation resides on a nitrogen atom on the azo linkage, leading to hydrolytic splitting of the azo bond (Figure 5). The initial products of the cleavage are the nitroso compound and an iminoquinone, both of which are relatively unstable. Redox reactions are proposed to convert the nitrosobenzenesulfonic acid to a sulfanilic acid, and the iminoquinone to an aminophenol (Figure 6). Hydrolysis is proposed as a path for the formation of benzoquinone from the iminoquinone. 1 L .------I asymmetric cleavage symmetric cleavage j

hydrolysis HB-W--QN=NH o=Q=o -T HE-wQ-PFO

NH3 oxidation *

k rduction .qYSis oxidation

Ha-weoH *

Figure 4. The products observed (*) during oxidation of a model azo dye by lignin peroxidase and manganese peroxidase from Phanerochaete ch ysosporium. Proposed pathways for the formation of the products is also shown. Substituent assignments in this figure: dye #1: R1 = R2 = CH3, B = 0; Dye #2: R1= H, R2 = OCH3, B = NH. From Goszczynski, et al. (13). [ HB-woN=N+$oH]

Peroxidase I-&ee, H' HB-w*N=Ndo*

Rl JI < R1 .--+N=N+ - HB-w--@-N+o Peroxidase /- eeR2 Peroxidate Rl HB-w+wN+o HB-~+.N-N+~ Fee: Rz HpLH' "'THa

o HB-weziN+o I R2 1 HI^--*^=.] + HN=($=o R2

Figure 5. The proposed mechanism for cleavage of the azo linkage in the peroxidative degradation of two model dyes. Substituent assignments in this figure: dye #1: R1= R2 = CH3, B = NH; Dye #2: Rl = H, R2 = OCH3, B = NH. From Goszczynski, et al. (13). 1. Redox processes

2. Hydrolysis

Figure 6. Proposed redox (1) and hydrolytic (2) processes that account for the formation of several products observed during degradation of two model azo dyes. Substituent assignments in this figure: dye #I: R1= R2 = CH3, B = 0; Dye #2: R1= H, R2 = OCI-13, B = NH. From Goszczynski, et al. (13). Cleavage at the azo bearing carbon of the phenolic moiety occurs when the cation resides on the carbon bearing the azo linkage (Figure 5). Hydrolysis leads to two initial products: a sulfonated phenyldiazene; and a benzoquinone. The authors have invoked redox reactions to rationalize the myriad of products observed (Figures 4 and 6). They propose that phenyldiazene may reduce benzoquinone, iminoquinone, or nitrosobenzene and be converted to benzenediazonium ion. Hydrolysis of benzenediazonium ion is proposed as a route to hydroxybenzenesulfonic acid. Benzenediazoniurn ion is proposed to oxidize phenyldiazene (Figure 6., reaction lc) and lose molecular nitrogen to form a benzenesulfonic acid. The mechanistic steps involved in this last transformation have not been thoroughly explained in the paper (13). As to the formation of a benzenediazonium ion intermediate, no proof, direct or indirect, is offered. More direct degradation of the phenyldiazene to phenyl radical, as suggested in the proposed mechanisms in this thesis, could also lead to the observed benzenesulfonic acid and hydroxybenzenesulfonic acid products (28).

IV. AZO DYE DEGRADATION BY CHEMICAL OXIDATION

Advanced oxidation processes are being considered for treating azo dye wastes. Bench scale reactors have already been designed (15). Several researchers have recently explored azo dye oxidation by Fenton (Fe(II)/H202) and Fenton-like reagents (Fe(m)/H202). Gregor reports that several dyes are degraded by (Fe@)/H202) to an extent of 85-loo%, but that disperse dyes are particularly resistant to oxidation, being degraded only O- 0.5% in at least one case (14). Flaherty and Huang found that room temperature oxidations by both Fe(lI)/H202 and Fe(III)/H202 yielded excellent decolorization and reduction of chemical oxygen demand (COD) in four textile industry wastewaters, each of a different color (15). For one of the waste streams, the ratio of M2:Fe2+had to be nearly ten times greater than the ratio of H202:Fe3+in order to achieve a similar loss in chemical oxygen demand, but for the oxidation of the other three waste streams the Fez+ and ~e3+systems were quite comparable. Extensive loss of color occured in all waste streams even at the lowest ratios of H202:Fe2+or H202:Fe3+(15). Powell, et al., demonstrated that defined dye streams containing Reactive Black 5 are quite effectively decolorized by (Fe(II)/H202) (10:l or 20:l ~202:Fe~+)(16). Adsorption of the organic products to the iron hydroxide sludges occurs. An interesting result is that addition of ferric iron (Few))to the completed Fe(II)/H202 reaction for degradation of a Navy Jet dye significantly increased dye decolorization during further reaction. Excepting the work presented in this thesis (chapter 5), neither reaction products nor mechanistic pathways have been suggested for Fenton degradation of azo dyes. In several studies on photocatalytic degradation of azo dyes the authors have suggested that a benzenediazonium ion is formed during azo cleavage (Figure 7) (3031). However, no evidence for the formation of benzenediazonium ion has been offered. However, quinones and benzene compounds are found as degradation products of N,N-dialkyl-4- phenylazoanilines and phenylazonaphthols (30,31). This may suggest that mechanisms similar to those described in chapters three, four, and five are operative in the photocatalytic oxidation systems. Figure 7. Azo dyes subjected to photocatalysis, and their degradation products. From Griffiths and Hawkins (30) and Albini, et al. (31). V. FUTURE WORK

Several questions pertaining to the degradation of azo dyes by chemical and biological systems remain. Areas of concern include: improvements in the rate and extent of degradation; in testing the actual loss of toxicity or mutagenicity of the dye wastes after treatment; and in defining the proper conditions for treating actual waste streams. With these concerns in mind, the following 8 projects should be pursued.

1. The radical mechanism for azo dye degradation should be tested in at least three ways:

a. Foremost, electron paramagnetic resonance (EPR)should be used to detect the formation of various radicals, but especially of the phenyl radical.

b. The proposed release of the azo linkage as nitrogen during enzymatic and chemical oxidation could remove potential azo dye toxicity. Using azo dyes isotopically labeled in the azo linkage with 15N, degradation reactions should be performed in such a way as to capture evolved nitrogen and then to subject the nitrogen to mass spectral analysis.

c. The source of oxygen that is incorporated into the benzoquinone products of azo dye degradation should be studied. 1802, HZ18@, and ~2180could be used to do this.

2. Thus far, most degradation studies have been performed under defined laboratory conditions. Obviously, future degradation studies need to be made on dyes mixed into real effluents or waste streams. Such studies will help in judging the true efficacy of the biological and chemical oxidation treatments.

3. The reduction in azo dye toxicity by these biological and chemical oxidation processes must be studied. As well, the toxicity of dye oxidation products must be assessed. 4. Azobenzene formation from anilines (derived from acylanilide herbicides and pesticides) in the environment is of concern (27). Studies should be made of the ability of soil-borne fungi, bacteria, and peroxidases to oxidatively degrade these azobenzenes.

5. The fungal degradation of certain azo dyes under non-ligninolytic conditions is intriguing (1). Aust and coworkers have shown that reduction of aromatic nitro groups to amines occurs under non-ligninolytic conditions (23). Enzymes such as azo reductases may also exist. We have also observed trace amounts of azo dye hydroxylation occuring under non-ligninolytic conditions. Extensive product analyses should be made, and new azo dye degrading enzymes purified and characterized.

6. W (300nm) irradiation (2 h/day) of P. chrysosporium cultures increases the mineralization rate of 2,3,7,8-tetrachloroyV-14Cldibenzodioxin (19). Degradation rates for several other recalcitrant pollutants (ie. DDT, tetrachlorobiphenyl) in fungal culture also increase when the cultures are irradiated at either 2snrn or 300 nm. The effects are synergistic in that degradation is faster than by either fungal cultures or UV irradiation alone. In addition, some P. chrysosporium strains appear to be particularly UV and fungicide tolerant when compared to several other fungal and bacterial species (19). The use of W irradiation to enhance azo dye degradation by fungal cultures should be examined. W irradiation may also increase the rate and extent of azo dye degradation by purified peroxidases, and this should be examined, as well.

7. Degradation of azo dyes by peroxidases appears to produce phenyl radical intermediates, as shown in chapters 3 and 4. Phenyl radicals are known to cause inactivation of peroxidases by forming adducts with the heme (17,18). The possible inactivation of peroxidases during azo dye degradation reactions should be studied. Radiolabeled azo dyes should be used to investigate the formation of w-14Clphenyl-heme adducts.

8. With respect to the Fe(III)/H202 system, several experiments also remain: a. It has been shown that irradiation with UV light enhances 2,4D degradation (24). Enhancement of azo dye degradation should be attempted in a similar manner.

b. Better identification and quantitation of degradation intermediates needs to be made. The reactions should be done in a slow, controlled manner so that particularly oxidizable intermediates, such as quinones, can be isolated.

c. The oxidation mechanism proposed in chapter 5, Figure 5, which appears to be comparable to that found for the peroxidases, should be tested via deuterium labeling studies and experiments ~singl80~,H21Q2, and H2180. VI. REFERENCES

1. Spadaro, J.T., Gold, M.H., and Renganathan, V. (1992) Appl. Environ. Microbiol. 58(8), 2397-2401.

2. Cripps, C., Bumpus, J.A., and Aust, S.D. (1990) Appl. Environ. Microbiol, 56(4), 1114-1118.

3. Kulla, H.G., Krieg, R., Zimmermam, T., and Leisinger, T. (1983) In Current Perspectives in Microbial Ecology, Proc. Int. Symp., (Klug, M.J., and Reddy, C.A., Eds.), pp. 663-667, Am. Soc. Microbiol. Washington, D.C.

4. Chung, K.-T., and Stevens, S.E., Jr. (1993) Environ. Toxicol. Chem. 12, 2121-2132.

5. Paszczynski, A., Pasti-Grigsby, M.B., Goszczynski, S., Crawford, D.L., and Crawford, R.L. (1991) Enzyme Miuobiol. Technol. 13,378-384.

6. Pasti-Grigsby, M.B., Paszczynski, A., Goszczynski, S., Crawford, D.L., and Crawford, R.L. (1992) Appl. Environ. Microbiol. 58(11), 3605-3613.

7. Paszczynski, A., Pasti-Grigsby, M.B., Goszczynski, S., Crawford, RL., and Crawford, D.L. (1992) Appl. Environ. Microbiol. 58(11), 3598-3604.

8. Leisinger, T., and Brumer, W. (1981) In Biotechnology, (Rehm, H.-J., and Reed, G., Eds.), Vol. 8, pp. 475-513, VCH, New York.

9. Stiborova, M., Afsaw, B., and Anzenbacher, P. (1988) FEBS Lett. 232(2), 387-390.

10. Paszczynski, A., and Crawford, R.L. (1991) Biochem. Biophys. Res. Comrnun. 178,1056-1063.

11. Cao, W., Mahadevan, B., Crawford, D.L., and Crawford, R.L. (1993) Enzyme Microbiol. Technol. 15, 819-817.

12. Ollika, P., Alhonmaki, K., Leppanen, V.-M.,Glumoff, T., Raijola, T., and Suominen, I. (1993) Appl. Environ. Miuobiol. 59(12), 4010-4016.

13. Goszczynski, S., Paszczynski, A., Pasti-Grigsby, M.B., Crawford, RL., and Crawford, D.L. (1994) J. Bacteriol. 176(5), 1339-1347.

14. Gregor, et a1 (1994) In Chemical Oxidation: Technologies for the Nineties, Vol. 2, (Eckenfelder, W.W., Bowers, A.R., and Roth, J.A., Eds.), pp. 161-193, Technornic, Lancaster, PA. Flaherty, K.A., and Huang, C.P. (1994) In Chemical Oxidation: Technologies for the Nineties, Vol. 3, (Eckenfelder, W.W., Bowers, A.R., and Roth, J.A., Eds.), pp. 112-133, Technomic, Lancaster, PA.

Powell, et a1 (1994) In Chemical Oxidation: Technologies for the Nineties, Vol. 2, (Eckenfelder, W.W., Bowers, A.R., and Roth, J.A., Eds.), pp. 278-300, Technomic, Lancaster, PA.

Ortiz de Montellano, P.R. (1987) Acc. Chem. Res. 20, 289-294.

Harris, R.Z., Wariishi, H., Gold M.H., and Ortiz de Montellano, P.R. (1991) J. Biol Chem. 266(14), 8751-8758.

Katayama, A., and Matsumara, F. (1991) Environ. Sci. Technol. 25, 1329- 1333.

Eriksson, K.-E.L., Blanchette, R.A., and Ander, P. (1990) In Microbial and Enzymatic Degradation of Wood and Wood Components, pp. 80- 84, Springer-Verlag, Berlin.

Valli, K., Wariishi, H., and Gold, M.H. (1990) Biochemistry 29,8535- 8539.

Harvey, P.J., Gilardi, G.F., and Palmer, J.M. (1989) In Enzyme Systems for Lignocellulose Degradation, (Coughlan, M.P., Ed.), pp. 111-120, Elsevier, London.

Stahl, J.D., and Aust, S.D. (1993) Biochem. Biophys. Res. Commun. 192(2), 477-482.

Sun, Y., and Pignatello, J.J. (1993) Environ. Sci. Technol. 27,304-310.

Meyer, U. (1981) In Microbial Degradation of Xenobiotics and Recalcitrant Compounds, (Leisinger, T., Cook, A.M., Nuesch, J. and Hutter, R., Eds.), pp. 371-385, Academic Press, London.

Shaul, G.M., Holdsworth, T.J., Dempsey, C.R., and Dostal, K.A. (1991) Chemosphere 22(1,2), 107-119.

Bartha, R, and Pramer, D. (1970) Advan. Appl. Microbiol. 13,317-341.

Russell, G.A., and Bridger, R.F. (1963) J. Am. Chem. Soc. 85,3765-3766.

Spadaro, J.T., and Renganathan, V. (1994) Arch. Biochem. Biophys. (in press). 30. Griffiths, J., and Hawkins, C. (1977) J. Chem. Soc. Perkin Trans. 11 747- 752.

31. Albini, A., Fasani, E., Pietra, S., and Sulpizio, A. (1984) J. Chem. Soc. Perkin Trans. I1 1689-1692.

32. Kirk, T.K., and FarreU, R.L. (1987) Ann. Rev. Microbial. 41,465-505.

33. Brown, M.A., and DeVito, S.C. (1993) Crit. Rev. Environ. Sci. Technol. 23(3), 249-324. BIOGRAPHY

The author was born in West Islip, New York on January 7,1965. He made his home for 18 years in Bay Shore, New York, where he graduated from high school in 1983. During this time he also attained the rank of Eagle Scout in the Boy Scouts of America. He acquired the degree Bachelor of Science in Chemistry at the Worcester Polytechnic Institute (WPI) in Worcester, Massachusetts, which he attended from 1983 to 1987. The year from 1987 to 1988 was spent as a laboratory technician at Eco-Test Laboratories, an environmental testing firm in North Babylon, New York. The move to graduate study at the Oregon Graduate Institute (OGI) came in the fall of 1988. The author now resides in the southeast section of Portland, Oregon. He enjoys organic gardening, hiking, backpacking, skiing, jazz, quaffing fine microbrewed beer, and many other aspects of living in the great Northwest.

The work in chapters two, three, and five has been published:

Chapter Two - Spadaro, J.T., Gold, M.H., and Renganathan, V. (1992) Appl. Environ. Microbial. 58(8), 2397-2401.

Chapter Three - Spadaro, J.T., and Renganathan, V. (1994) Arch. Biochem. Biophys., in press.

Chapter FIve - Spadaro, J.T., Isabelle, L. ,and Renganathan, V. (1994) Environ. Sci. Technol., in press.