THE PHOTOCHEMICAL REACTIONS OF HEPTACIILOR: KINETICS AND MECHANISMS Thesis for the Degree of Ph. D. MICHIGAN STATE UNIVERSITY RAYMOND R. MCGUIRE 1969 Infiflfi 0-169 This is to certify that the thesis entitled THE PHOTOCHEMICAL REACTIONS OF HEPTACHLOR: KINETICS AND MECHANISMS presented by Raymond R. McGuire has been accepted towards fulfillment of the requirements for Ph .D degree in Chemistry 7 \ZeMIfiLVm/ Dme September 3, 1969 I .1211?“ ' e :- r 4." 1'9 1" 3 7’ V AL- Michlgan 5 Lu. L: University " .l-NDEN‘ .Y ‘\ none a son: I, max ELIE“! mc.‘ ABSTRACT THE PHOTOCHEMICAL REACTIONS OF HEPTACHLOR: KINETICS AND MECHANISMS BY Raymond R. McGuire The purpose of this investigation was to determine the photochemical reaction mechanisms of a class of pesticides whose general structural characteristics can be represented by heptachlor, 1,4,5,6,7,8,8-heptachloro-5a,4,7,7a-tetra- hydro-4,7-methanoindene. The products formed by the photolysis of heptachlor depend upon the conditions under which the reaction is carried out. The irradiation at wavelengths less than 2600A of heptachlor dissolved in non-triplet sensitizing solvents; e.g., hexane, cyclohexane, etc., yields a mixture of two monodechlorination isomers, 1,4,5,7,8,8-hepachloro- 3a,4,7,7a—tetrahydro-4,7-methanoindene and 1,4,6,7,8,8-hexa- chloro-Sa,4,7,7a-tetrahydro-4,7-methanoindene, in equal amounts. These isomers were separated by gas chromatography and characterized by n.m.r. Spectrometry. This reaction proceeds with a quantum yield at 2557A of 0.025 and shows zero order kinetics. A simple, nonchain, free radical mechanism involving only the 5,6 double bond of heptachlor is postulated for this photodechlorination reaction. Raymond R. McGuire When the photolysis of heptachlor was carried out at higher wavelength (5000A) in a triplet sensitizing solvent such as acetone, the product was a cage compound, 2,5,4,4,5,— 6-10-heptachloro-pentacyclo(5.5.04.03’9. 5’8 O )decane. The quantum yield of the cage compound under these conditions was 9.55 x 10"5 based on the absorption of light by acetone. The reaction showed "0" order kinetics. When the reaction was carried out at 5000A in mixtures of cyclohexane and acetone, the major photOproduct was found to be cyclohexyl adduct of heptachlor where a cyclohexyl radical replaced the chlorine on carbon 1. Similar products were found when the photolysis was performed in n-hexane, cyclopentane and ethylacetate rather than cyclohexane. The quantum efficiency of decay of heptachlor was found to in- crease as the amount of acetone decreased. This phenomenon was shown to be a viscosity effect. A nonchain, sensitized triplet mechanism involving only the 2,5 double bond is postulated for the cage and adduct formation. A kinetic mechanism is derived for the sensitized triplet reaction and the specific rate constants are determined for each of the kinetic steps in this mechanism. Finally, the formation of the cage compound was found to be a reversible process. Irradiation of a cyclohexane solution of the cage compound at 2000A gave heptachlor as the photoproduct. This reaction showed zero order kinetics and had a quantum yield of 0.195. THE PHOTOCHEMICAL REACTIONS OF HEPTACHLOR: KINETICS AND MECHANISMS BY \ I Raymond R. McGuire A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1969 ACKNOWLEDGEMENTS The author wishes to express his sincere appreciation to Professor Robert D. Schuetz and Dr. Matthew J. Zabik for their encouragement, council and friendship throughout the course of this investigation. Grateful acknowledgement is also extended to Colonel John D. Hicks, U.S.A.F., without whose help this would not have been possible. ii TABLE OF CONTENTS INTRODUCTION . C O O C O O O O O O O O C O O HISTORICAL BACKGROUND . . . . . . . . . . . Photodechlorination in the Aromatic System Photodeclorination in the Condensed Pon- cyclic System . . . . . . . . . . Cage Compound Formation. . . . . . . . Photodechlorination vs. Cage Formation Toxicity of Photoproducts. . . . . . . BEERIMENTAL O O O O O O O O O O O O O O O 0 Reagents . . . . . . . . . . . . . . . Heptachlor. . . . . . . . . . . . Sensitizers and Solvents. . . . . Instrumentation. . . . . . . . . . . . Gas Chromatograph . . . . . . . . Ultraviolet Spectrophotometer . . Infrared Spectrophotometer. . . . N.M.R. Spectrometer . . . . . . . Mass Spectrometer . . . . . . . . Irradiation Sources. . . . . . . . . . Procedure. . . . . . . . . . . . . . . Determination of Evolved Hydrogen Chloride . . . . . . . . . . Determination of Light Intensity and Quantum Yield. . . . . . . . Kinetic Measurements. . . . . . . Separation and Identification of Products RESULTS AND DISCUSSION. . . . . . . . . . . Photodechlorination. . . . . . . . . . Stoichiometry . . . . . . . . . . Wavelength Dependence . . . . . . Kinetics of the Reaction. . . . . Identification of the Photoisomers. iii Page »> 9- re TABLE OF CONTENTS - Continued Page The Sensitized Triplet Reaction, Cage Formation. . . . . . . . . . . . . . . . . 55 Photolysis of Heptachlor in Pure Acetone . 55 Photolysis of Heptachlor in Mixed Solvents 56 Identification of the Cyclohexyl Adduct. . 48 n-Hexyl Adduct . . . . . . . . . . . . . . 55 Reaction Mechanisms . . . . . . . . . . . . . . 55 Photodechlorination. . . . . . . . . . . . 55 Cage Formation and Adduct Formation. . . . 57 Kinetic Mechanism for the Triplet Reaction 59 Reversibility of Cage Formation . . . . . . . . 64 PRACTICAL IMPLICATIONS . . . . . . . . . . . . . . . 68 LIST OF REFERENCES . . . . . . . . . . . . . . . . . 70 iv LIST OF TABLES TABLE Page I. The NMR Chemical Shifts of Heptachlor and Its Monodechlorination Isomers. . . . . . . . . . . 51 II. Effect of Viscosity on the Rate of Photodecom- position of Heptachlor. . . . . . . . . . . . . 45 III. Mass Spectrum of the Cyclohexyl Adduct of Heptachlor. . . . . . . . . . . . . . . . . . . 49 IV. The IR Spectra of Heptachlor, Cyclohexane and Cyclohexylheptachlor. . . . . . . . . . . . . . 52 V. The NMR Chemical Shifts of Heptachlor and Its Cyclohexyl Adduct . . . . . . . . . . . . . . . 54 VI. Specific Rate Constants for the Triplet Sensi- tized Photodecomposition of Heptachlor. . . . . 65 FIGURE 10. 11. 12. LIST OF FIGURES Rate of photodechlorination of heptachlor at 2557A. . . . . . . . . Rate of formation of cage compound in acetone and 5000A. . . . . . . Reaction of heptachlor cyclohexane. . . . . . Reaction of heptachlor 60% cyclohexane. . . . Reaction of heptachlor 70% cyclohexane. . . . Reaction of heptachlor 80% cyclohexane. . . . Reaction of heptachlor 90% cyclohexane. . . . Effect of viscosity on on cage compound . . . Effect of viscosity on formation. . . . . . . Effect of viscosity on of heptachlor decay. . in 50% acetone, 50% at 5000A in at 5000A in at 5000A in at 5000A in the rate of O O O O C O the rate of the quantum 40% acetone, 50% acetone, 20% acetone, 10% acetone, formation adduct efficiency Comparison of the IR Spectra of cyclohexyl heptachlor, heptachlor and cyclohexane . . . . Rate of cage Opening at 2000A. . . vi Page 50 55 58 59 40 41 42 44 45 46 51 65 INTRODUCTION The use of polychlorinated pesticides has contributed greatly to increases in the efficiency of food production in a major part of the world. These pesticides have the advantages of being broad spectrum; low production costs; storage stability; and low direct toxicity to mammals. However, their field stability has become of increasing con- cern to agriculturists, conservationists, ecologists, legis- lators and the general public. The polychlorinated pesticides do not degrade rapidly under normal environmental conditions. They accumulate in ever increasing amounts as they travel along the various food chains until they reach toxic concen- trations to some birds and fish. Thus they pose a potential danger to man, himself. This persistence has earned for them the label of "hard pesticides." A great deal of time, effort, and funds have been expended in determining the mechanisms of bio-degradation of these highly chlorinated compounds. Their metabolisms have been studied in biological systems ranging from bacteria to mammals. However, another, and perhaps more important method of environmental degradation, namely photolysis, was largely unexplored until about 1960. Since then, work in this area has remained fragmentary and has been undertaken largely from the point of view of toxicology rather than that of the fundamental photochemistry involved. DeSpite this some important information has been accumulated on the photolysis of these materials. It has been determined that the "hard pesticides" do, indeed, undergo environmental photolysis. A number of the photolytic products have been isolated, identified and reproduced in the laboratory. The toxicity of these photo- products has been examined and found to be vastly different .from those of the parent compounds; some being as much as five times as toxic. These environmental photo reactions fall generally into two categories: cage compound formation where the structure of the parent pesticide permits such structure formation to occur; e.g., heptachlor, dieldrin, aldrin, isodrin etc.; and dechlorination where cage formation is not possible, e.g., DDT. The present investigation was undertaken to determine the mechanisms by which these types of compounds undergo photolysis, to add to the understanding of the primary photo- chemical processes; and to advance the time when it will be possible to design and synthesize a pesticide which is highly toxic to insects and will environmentally degrade to non- toxic materials in a functional period of time. Heptachlor was chosen as the model compound because of its close structural relationship to most other "hard pesticides," DDT and its derivatives excepted, and because the structures of two of its major phot0products had already been determined by R. D. Flotard, working in these laboratories. HISTORICAL BACKGROUND The so-called "hard pesticides" are polychlorinated hydrocarbons which can be divided into two general systems: aromatic and condensed polycyclic systems. The aromatic system is exemplified by DDT, 1,1,1-trichloro-2,2-bis- (p-chlorophenyl)ethane, and its derivatives. The second system, of which heptachlor is a member, is composed of a series of structurally related compounds produced by the Diels-Alder reaction of cyclopentadiene with hexachloro— cyc10pentadiene or of cyclopentadiene, ethylene and hexa- chlorocyclopentadiene. The photochemical reactions of these systems can also be separated into two general categories: Photodechlorina- tion and photocyclization (cage formation). Examples of photodechlorination have been found in both systems while only the condensed polycyclics have been shown to undergo cage formation. Photodechlorination in the Aromatic System The prolonged residual action of DDT (I) is due to its low vapor pressure, stability to oxidation and to biodegrada- tion. It was shown by Fleck in 1949 (11) that 4.4'- 4 01 c1 /H \ /c1 c / \c1 c1 c1 dichlorobenzophenone (II) results from the eXposure of DDT to ultraviolet radiation in ethanol solution. This reaction was first thought to proceed by the dehydrochlorination of DDT to yield III, followed by its oxidation to give the chlorinated benzophenone. However, the intermediate Ar\\./’H hv Ar\\ |O|a /C\cc13 -HCl Ar/C CCIZ Ar I III Ar\\ Ar = p-chlorophenyl dehydrochlorination product (III) could not be isolated even when air was rigorously excluded from the system. The product actually isolated in the absence of air was 2,5- dichloro-1,1,4,4-tetrakis(p-chlorophenyl)butene-2 (IV). It was suggested that (IV) could then lose two molecules of Ar Ar >33: E1- §< IV Ar Ar hydrogen chloride to form the butatriene (V) which has been shown to yield (II) on oxidation. Mosier et a1. (21) have recently reported on further investigations of the photodegradation of DDT at 2557A both as a solid and in a n-hexane solution and have proposed the following mechanisms for the photochemical reaction. I A ”\E $1 + Cl. //"“‘ -C1 Ar - [Ia] Ar:>C/,H Ar\ //Cl Ia + I -*>' 1 + - ' ;__ [VI] Ib DDT (I) is photodechlorinated by ultraviolet light to give (Ia) and a chlorine radical. (Ia) then reacts with I to give a second radical (Ib) and 1,1,dichloro-2,2—bis- (p-chlorophenyl)ethane (VI) (DDD). (Ia) can also react with a chlorine radical A Ia + C1' ——+ JG>c = cc12 + HCl Ar III to give 1,1-dichloro-2,2(p-chlor0phenyl)ethylene (III) or (DDE) and hydrogen chloride. DDT can react with a chlorine radical to yield Ib and HCl. The radicals Ia and I + Cl: —* Ib-I-HCl C1' + solvent -—-*' HCl + solvent radical Ia+Ib—-> I+III Ib can react to give DDT (I) and DDE (III). A final source of HCl would be the reaction of a chlorine radical with the solvent. III and VI were isolated from their reaction mixtures. These reactions were also run in the presence of the free-radical scavangers iodine and n-butylmercaptan. It was found that the presence of iodine decreased both the rate of disappearance of DDT (I) and the rate of forma- tion of DDD (VI). However, the presence of n-butylmercaptan had no effect on the rate of decomposition of DDT (I) and actually increased the rate of formation of DDD (VI). The authors attribute this to the abstraction of a proton from the mercaptan by the radical (Ia) to form DDD (VI). Although the conclusion of a nonchain, free radical mechanism is probably correct, the proposed steps cannot be substantiated from the data presented by the authors. In the first place, the proposed mechanism does not account for the formation of (IV) found by Fleck (11), secondly, most of the photolysis products were not identified, e.g., in one reaction DDT was irradiated in n-hexane for four hours yielding 5% of unreacted DDT, 4%IDDD and three uni- dentified products: thirdly, the fate of the scavangers was not traced and finally, the reported quantum yield (¢) was improperly calculated. Although photochemical dehalogenation involving the halogen attached directly to an aromatic ring has been shown to occur quite easily (25,55,56) it has not been noted in the case of DDT. The normal course of these reactions is the abstraction of a proton from the solvent by the radical. However, the aromatic radical can add a solvent radical. i.e., solvent minus proton. wolf and Kharash (40) have shown that the irradiation of 4-iodophenylbenzene VII in benzene yields 91% p—terphenyl (VIII). In addition, I —h_y___)- @ VII VIII Crosby and Tutass (6), have shown that the irradiation of a water solution of the herbicide 2,4-dichlorophenoxy acetic acid (IX) yields 1,2,4-benztriol (X) and, eventually, polymeric humic acids (XI). I OCHgCOOH F—- 0H 0H h 14a ; H20 I .110 _I. OH X IX (x) (XI) Photodechlorination in the Condensed Polycyclic System Photodechlorination in the condensed polycyclic sys- tem has only recently been discovered. Henderson and Crosby (15) reported in 1967 that aldrin (XII) and its epoxide dieldrin (XIII) undergo photolysis in hexane solu- tion at 2557A to give the monodechlorination products XIV and XV respectively. These dechlorination products were C1 C1 C1 ———-> Cl C Cl XII XIV Cl C1 C1 Cl C1 C]. I d ———>' C1 C1 C1 XIII XV 10 not produced at wavelengths above 2600A. Since solar radia- tion cuts off at about 2865A (19), it is not surprising that they have not been found under field conditions. Rosen (26) has also reported the photodechlorination of aldrin (XII). Flotard (12) has shown a similar reaction for heptachlor (XVI) while Anderson et a1. (1) have investigated the photo- Cl c1 an c1 Cl C. c1 XVI dechlorination of other alicyclic systems. Cage Compound Formation The more commonly found products for the photolysis of the condensed polyclic system are cage structures. While cage structures are often formed upon irradiation of the polycyclic systems, even under field conditions, they have not been found as photo-products of the aromatic systems under any conditions. Mitchell (20) in 1961 reported that dieldrin (XIII) and aldrin (XII) were decomposed by 2557A energy radiation and Roburn (25), in 1961, reported finding phot0products on grass that had been treated with dieldrin and aldrin. Robinson et al. (24) and Rosen et a1. (50) isolated these phot0products 11 and hypothesized cage structures on the bases of I.R. spectra. HarriSon et al. (14) proposed (XVII) as the structure for the cage phot0product of dieldrin (XIII). This structure is :6? XVI I inconsistent with the nmr spectrum and Parsons and Moore (22) have proposed what is now the accepted structure (XVIII). Cl XVIII Cookson et al. (5,6) had earlier found that isodrin underwent cage formation, as shown by the infrared and ultraviolet Spectrum, when irradiated in ethylacetate solution while its isomer aldrin did not. This was taken as proof of the endo- endo structure (XIX) for isodrin and the endo-exo structure 12 C1 C1 C1 C1 XIX (XII) for aldrin. Zabik et al. (41) have also shown that the isodrin epoxide endrin (XX) forms a cage compound upon photolysis. C1 C1 . c1 1 ' C1 A Cl"c> l —+ c , C I c1 C1 C1 XX XXI Rosen et al. (27,29) have made the cage phot0products of aldrin, dieldrin, isodrin and heptachlor and reported them to be sensitized by benzoPhenone. This is not an estab- lished fact, however, since the reactions were carried out in benzene solution and sensitization by benzene cannot be ruled out by the eXperiments they performed. While it is likely that a reaction sensitized by benzophenone, which has a lowest triplet energy of 69 kcal/mole, would also be sensitized by benzene with a triplet energy of 85 kcal/mole 15 (57) the converse is not necessarily true. Cage formation, or 4-cycloaddition, has been shown to Stedman and Miller (55) 5605,10. occur in nonpesticide systems. have formed the cage ketone, hexacyclo[5.4.1.0 08’9- 0.8’11]dodecane-4-one (XXII) by irradiating the diene ketal hv g_ acetoner XXIII XXII (XXIII) in acetone. It will also occur in the absence of chlorinated double bonds. Barborak and Pettit (2) have obtained homocubanol (XXIV) by irradiating an acetone solu- tion of the diene XXV. The reaction seems to be H OH H OH hv k_ a acetoner XXV XXIV regiosPecific. Dilling et al. (10) have obtained the diene (XXVIIby treating the symmetrical chlorinated pentacyclo- decane XXVII with lithium metal in tertiary butanol: however, 14 Li \ ——_C112 t-BuOH ’ I. XXVII XXVI irradiation of the diene XXVI in acetone yields only the unsymmetrical pentacyclodecane XXVIII (51,52). hv . r acetone XXVI XXVIII Photodechlorination vs. Cage Formation In their study of the photodecomposition of dieldrin and aldrin, Henderson and Crosby (15) found that they did not get photodechlorination on irradiating at wavelengths above 2600A. They also found that when they did get photo- dechlorination they did not get cage formation. Anderson et a1. (1) have recently studied the reaction path selectiv- ity in some related alicyclic systems. They have found that the tetrachloro ketal XXIX gave only the trichloro ketal XXX on irradiation at wavelengths above 2100A in ether solu- tion and did not react at all when irradiated above 2900A in 15 Me OMe MeO OMe Cl hV C C]. I C]. XXIX XXX the presence of benzophenone. However the ketal XXXI gave the cage compound (XXXII) only when irradiated either in acetone or in carbon tetrachloride with benzophenone added. Me 0 oMe M630 1 OM62 ' C l c 1 C 1 h v C1 C1 C1 XXXI XXXII They have also shown that the urazole XXXIII is stable under reaction conditions which normally leads to photodechlorination MeO OMe hv . -——€b no reaction Cl XXXIII 16 but that the urazole XXXIV yields the cage compound XXXV when /N hv ,_ / k) aceton? V ‘3 H3 XXXIV . XXXV irradiated in acetone solution. They concluded that the photodechlorination takes place by way of a singlet excited state which is quenched by the urazole ring and that cage formation, unquenched by a urazole ring but sensitized by both benzoPhenone (triplet energy = 69 kcal/mole) and acetone (triplet energy = 76 kcal/mole)(54), goes through an excited triplet. This conclusion (singlet vs. triplet) has been confirmed by Flotard (12) in his study of the heptachlor system. He has shown that heptachlor XVI gives the cage compound XXXVI hv xi acetone 17 when irradiated in acetone solution but that a mixture of photodechlorination isomers XXXVIIa and XXXVIIb are pro— duced when the reaction is carried out in hexane or cyclo- hexane. C1 C1 C1 c1/ C1Cl XXXVIIa XXXVIIb Toxicity of Photoproducts The increased toxicity of the cage phot0products is a cause for great concern. Brown et al. (4) and Rosen and Sutherland (28) have reported that the cage photodieldrin (XVIII) is from 2 to 10 times as toxic to several vertebrates as is dieldrin (XIII) itself. The cage compound (XVIII) has also been shown to be more toxic to insects (28,50). The cage photoaldrin (XXXVIII) has been shown to be eleven times more toxic to mosquito larvae than aldrin (XII) itself (54). 18 Kahn et al. (16) have attributed this increased toxicity to the formation of XXXIX in the insect metabolism of both cage photoaldrin (XXXVIII) and cage photodieldrin (XVIII). This metabolite is not found in the insect metabolism of either aldrin (XII) or dieldrin (XIII). This metabolite (XXXIX) was first discovered by Klein et al. (18) in the metabolism of dieldrin by male mice and was shown to be significantly more toxic than dieldrin itself. As yet the toxicities of the photodechlorination products have not been reported. XXXIX EXPERIMENTAL Reagents Heptachlor The heptachlor used in this study was obtained from R. D. Flotard, of these laboratories, who had prepared it as follows. A commercial sample of heptachlor, 25% by weight, was dissolved in acetone, filtered and the solvent evaporated. The resulting solid was redissolved in g-hexane and chromatographed on an activated alumina column. Fifty milliliter fractions were collected and tested by vapor phase chromatography. Fractions three and four, which contained the heptachlor, were combined and evaporated to dryness. Heptachlor purified in this manner was shown to be identical, when gas chromatographed using an electron capture detector, with a sample purchased from City Chemical Corp., of 99+% purity ("ESA" Pesticide Reference Standard). Sensitizers and Solvents The gfhexane, cyclohexane and acetone used in this study were "Distilled in Glass" solvents purchased from Burdick and Jackson Laboratories Inc., and were used as received. 19 20 The benZOphenone was purified by vacuum sublimation just prior to use. Instrumentation Gas Chromatograph All gas chromatograms were obtained using a Beckman Model GC-4 Gas Chromatograph equipped with a fraction collector. Two types of detectors were used: a hydrogen flame ionization detector for concentrated samples such as those used for fraction collecting: and an arc discharge electron capture detector for more dilute solutions such as those used for kinetic studies. The column packing was prepared by hand shaking approxi- mately 9 grams of 60/80 mesh Gas Chrom Q with approximately 1 gram of DC-11 silicone grease dissolved in 500 ml of ethylacetate. The slurry was then vacuum filtered, air dried and heated in a vacuum oven at 100°C for 24 hrs. The material was then packed into two stainless steel columns. one 1/8" by 6', the other 1/4" by 5' using a vibrator. The packed columns were conditioned on a GC-4 for 5 days at 220°C with periodic injections of cyclohexane and a 1% solu- tion of heptachlor in cyclohexane. Ultraviolet Spectrophotometer Ultraviolet spectra were determined in a Beckman Model DB-G Ultraviolet Grating Spectrophotometer. 21 Infrared Spectrophotometer All infrared spectra were determined either as potassium bromide pellets (solids), or as smears on potassium bromide pellets (viscous liquids). Spectra were determined on a Perkin—Elmer model 557 Grating Spectrophotometer. N.M.R. Spectrometer Nuclear magnetic resonance spectra were determined on a Varian Model 56/60 Spectrometer in deuterated chloroform with tetramethylsilane as an internal standard. Mass Spectrometer Mass spectra were determined in a L.K.B. Model 9000 Mass Spectrometer equipped with mass marker, peak matcher and gas chromatographic inlet. Samples were injected as acetone or methylene chloride solutions through the gas Chromatograph. The column was 6' by 1/8" packed 2%IOV 225 on 120 mesh gas chrom Q. Irradiation Sources Four sources of ultraviolet radiation were used in the course of this study. Large scale preparations were run using a 200 watt, medium pressure, wide band mercury discharge immersion lamp manufactured by the Hanovia Lamp Division of Engelhart-Hanovia Inc. For exploratory irradiations and for kinetic determination at 2000A and 5660A a high energy 22 deuterium source for the Beckman DB-G spectrophotometer (with an effective band width of 55A) was used. This lamp yields about three times the energy of the normal hydrogen source used in this instrument and was found to be adequate for these irradiations. The irradiations at 5000A were carried out in a Rayonet Photochemical Reactor manufactured by the Southern N. E. Ultraviolet Co. This reactor is equipped with filtered medium pressure lamps having a peak output at 5000A. The band width is unknown. The final source was a NFU-500 low pressure mercury discharge lamp manufactured by the Nester-Faust Co. This lamp yields 96% of its total energy as a single line at 2557A as measured with the DB-G. This lamp was further filtered by a K2Cr04, K2C03 solution as described by wagner (58) to give approxi- mately 99% pure 2557A radiation. Procedure Determination of Evolved Hydrogen Chloride A 150 ml volume of a 1.0% solution of heptachlor in cyclohexane was placed in a reaction vessel having a fritted glass bottom and two sidearms for introducing gas through the solution. The solution was irradiated with the 200 watt immersion lamps for periods of 90 to 150 min. Dry nitrogen was continuously passed through the reaction mixture, then through a dry ice cold trap and into 150 ml. of 0.2805 N sodium hydroxide. After completion of the reaction purging 25 with dry nitrogen was continued for 15 to 20 min. at which time the bottom stopcock of the reaction vessel was opened and the reaction mixture was drained into 50 ml of standard base. The reaction flask was washed out with distilled water and the washings were drained into the base. The aqueous layer was separated, combined with the base removed from the trap and back titrated with 0.2184 N hydrochloric acid to a phenolphthalein endpoint. . Qgtermination of Light Intensity and Quantumyxield The intensities of each of the sources, with the exception of the wide band prep source, were measured actinometrically using a potassium trisoxalatoferrate III actinometer according to the method described by Calvert and Pitts (5). The quam- tum yields (¢) were calculated from the formula, ¢ = n/Ig Z. where n is the number of molecules of product produced per minute or the number of molecules of reactant disappearing per minute, I: is the intensity of light at the inside face of the reaction vessel as measured by the actinometer and expressed in quanta/minute, and Z is the fraction of this light which is absorbed by the reactant. This equation can be expressed as I = koA V/Ig X where k0 is the zero order rate constant fin moles/liter minute), A is Avagadro's number and V is the volume of the reaction vessel in liters. Kinetic Measurements Kinetic measurements were determined using the gas 24 Chromatograph. The 6' by 1/8" column described above was used in conjunction with the electron capture detector. The flow of prepurified helium and the temperature were adjusted to give adequate peak separation for accurate measurements without excessive broadening or tailing. Under the conditions normally employed, 1500C and 40 ml/min. helium flow, the photodechlorination isomers eluted in two minutes: heptachlor, 5 minutes; cage compound, 4.2 minutes and the cyclohexyl adduct in 5.5 minutes from the solvent peak. The general procedure used in these rate measurements is as follows. A sample of heptachlor, 1.0 x 10"4 molar, dissolved in the solvent to be studied was pipetted into the reaction vessel. For irradiations at 5000A, spectronic '20' sample tubes containing 6.0 ml each were used as reaction vessels. All other irradiations were carried out in silica DB cells containing 5.7 ml of sample. The vessels were stoppered. specially designed septum stoppers were used for the DB cells, and placed in the reactors. A carrosel was used with the 2557A source to insure even irradiation of all samples. (The irradiation sources were all Operated for an hour prior to exposure of the reaction mixture to allow them to stabilize.) A 0.5 microliter sample of the starting material was injected into the gas Chromatograph and the supression voltage adjusted so as to give a peak height of 60 to 80% of the full scale. At various times, depending on 25 the rate at which the reaction was progressing, 0.5 micro- liter samples were removed from the reaction flasks and injected into the gas Chromatograph. The areas of the peaks were measured with a planimeter,.normalized and ex— pressed as a percentage of the total peak area. This pro- cedure was deemed to be valid since the reactions were allowed to proceed only to the point of initiating competing reactions. Separation andAIdentification of Products The photodechlorination isomers and the cage compound had been identified and characterized by Flotard (12). Other products were obtained by collecting fractions from the gas Chromatograph using the 5' by 1/4" column described above. These samples were identified either by spectroscopic methods or by comparison of their chromatographic behavior with authentic samples. RESULTS AND DISCUSSION Flotard (12) reported that the products of the photolysis of heptachlor (1,4,5,6,7,8,8-heptachloro—5a,4,7,7a-tetrahydro- 4,7-methanoindene)(I) were dependent upon the reaction con- ditions. When the photolysis was carried out with a high pressure, broad spectrum lamp in either hexane or cyclohexane the predominant products were a pair of monodechlorination isomers, 1,4,5,7,8,8-hexachloro-5a,4,7,7a-tetrahydro-4,7- methanoindene (II) and 1,4,6,7,8,8-hexachloro-5a,4,7,7a- tetrahydro-4,7-methanoindene (III). Hydrogen chloride was also evolved in the course of this reaction. Cl 1 Cl 1 C c1 ‘ M > hexane or Cl ““VV cyclohexane Cl + HCl 26 27 When the reaction was carried out in acetone, a triplet sensitizing solvent, the photolysis product was the "cage" compound 2,5,4,4,5,6,10-heptachloro-pentacyclo(5,5.02'6.03’9- 5,8 .0 )decane, IV. C1 C1 4’ I hv q» \ 1 acetone C ClCl IV Rosen et al. (29) have obtained the same product (IV) from the photolysis of I in the presence of benzophenone in benzene as a reaction solvent. Flotard (12) suggested that, by analogy to the findings of Anderson et al. (1), that the photodechlorination of heptachlor (I) proceeds through a singlet transition state, while the 4-cyclo addition ("cage" formation) proceeds by way of a triplet state. This study was undertaken to gain a more complete understanding of the mechanism or mechanisms of these reactions. Photodechlorination Stoichiometry An initial experiment was performed in an attempt to establish a 1:1 relationship between the number of moles of hydrogen chloride produced and the number of moles of hepta- chlor reacting. It quickly became apparent that the results had little significance since the hydrogen chloride to hepta- chlor ratios obtained were usually high and depended upon the 28 length and intensity of irradiation. On only one occasion was the expected 1:1 relationship found. This indicates that the secondary reactions, under these conditions, evolve hydrogen chloride much faster than does the primary, mono- dechlorination reaction. Further experimentation along these lines was abandoned at this point. Iggyglengthygependence Henderson and Crosby (15) in their investigation of the photodechlorination of dieldrin and aldrin (see Historical Section) found that the reaction did not occur at wavelengths above 2600A. The analogous situation was found in the photode- chlorination of heptachlor. Irradiations of heptachlor solu- tions were carried out at wavelengths ranging from 5000A to 2000A with the principal work carried out at 2557A. Even though the extinction coefficient of heptachlor is essentially zero at wavelengths above 2800A and is low at 2557A (eoitfi75), the reaction proceeds smoothly at the latter wavelength. Solutions irradiated at 5000A remained unchanged after three hours. Thus, the results for the heptachlor system are in agreement with those found by Henderson and Crosby for related systems. .Kinetics of the Reaction The photolysis of a 10—4 M solution of heptachlor in cyclohexane proceeds smoothly at 2557A. For irradiation times of up to three hours (9510% reaction) the only products detected were the pair of monodechlorination isomers reported by Flotard (12). No attempt was made to separate these 29 isomers and they were treated, kinetically, as one product. Hydrogen chloride and bicyclohexyl were also identified in the reaction mixture. The reaction was carried out in both Open and closed vessels. As the rates were essentially the same, showing no effect by the evolved hydrogen chloride, the data was combined. A plot Of concentration versus time gives a straight line showing the reaction to be "0" order as expected (Figure 1). The rate, determined from the slope of the least squares line, is 6.25 x 10’8 moles/liter-minute. 015 quanta/minute Based on an effective intensity of 5.66 x 1 (only 2% Of the incident radiation is absorbed), this gives a "0" order rate constant k0 ===1.11 x 10'23 moles/liter-quanta or a quantum yield O = 0.025. Flotard's findings (12) showing that this reaction is not sensitized by acetone were corrobo- rated in this study. Identification Of the Photoisomers Flotard (12), in his study of the photodechlorination of heptachlor, referred to the monodechlorination isomers by their order of elution from the gas Chromatograph (peak 1 and peak 2) without attempting to assign definite structures to them. How- ever, a close inspection of the nmr spectra Of the isomers (Table I) makes it possible to assign structures to these isomers. Examination of the molecular model of heptachlor (I) shows that the allyl hydrogen (HC in Table I) should be shielded by the electronic clouds of both the chlorinated double bond and the chlorine attached to carbon 6. Replacement Of the Concentration (moles/liter x 105) 50 11.0 100 10.45—- ’ : -— 95 8 0 0 9.90-— Heptachlor 90 1.10r- 10 ‘0 Photodechlorination 0 Products I I I I I 0 50 60 90 120 150 180 Time (minutes) Figure 1. Rate Of photodechlorination of heptachlor at 2557A. Percent of Reaction Mixture 51 TABLE I The NMR Chemical Shifts of Heptachlor and ' Its Monodecthrination Isomers (I) Proton Chemical Shifts (tau) Heptachlor Peak 1 Peak 2 Ha , 4.1 (s) 4.25 (s) 4.50 (s) “b - . Hc 5.2 (m) 5.42 (m) 5.5 (m) Hd 6.5 (m) 6.5 (m) 6.5 (m) He 5.9 (m) 6.07 (m) 6.09 (m) Hf —-- 4.20 (s) 4.20 (s) (s) = singlet: (m) = multiplet 52 electronegative chlorine atom from either carbon 5 or carbon 6 by a hydrogen (Hf) should cause an increase in size of the electronic cloud associated with the 5,6 double bond and, consequently, increase the shielding Of HC causing an upfield shift of its nmr absorption (an increase in the tau value Of the chemical shift). Table I shows that this upfield shift Of HC does, indeed, occur in the nmr spectra of both peak 1 and peak 2. However, the replacement Of the chlorine on carbenmswuwhile‘it increases the shielding due to the double bond, eliminates the shielding due to the replaced chlorine. Consequently, the replacement of the chlorine on carbon 6 would have less of an overall shielding effect on HC than replacement of the chlorine on carbon 5. On this basis structure II is assigned to peak 1 and structure III to peak 2. Rosen (26) has recently noted the same phenomenon (change in the chemical shift of a nearby proton upon dechlorination) in his study of the photolysis of aldrin (V). Replacement Of one of the vinyl chlorines (they are equivalent due to the symmetry Of aldrin) caused a shift in the tau value for Hb. 55 The shift was, however, in the Opposite direction from that found on dechlorination Of heptachlor (THb = 8.40 in aldrin, THb = 7.54 in the phot0product). Examination of the molecular model of aldrin shows that Hb lies in the deshielding field of the double bond rather than being shielded by it. Further deshielding by replacement Of one of the vinyl chlor- ines by hydrogen is, therefore, to be expected. The Sensitized Triplet Reaction, Cage Formation Anderson et al. (1) have shown that cage formation occurs through an excited triplet state and is susceptible tO sensi- tization. Flotard (12) and Rosen et al. (29) have confirmed this finding for the heptachlor system. Rosen claims that the cage formation Of heptachlor is sensitized by benZOphenone (ET = 69 kcal/mole) (57). Our investigations have failed to substantiate this claim. A 10-4 M cyclohexane solution of heptachlor, 10‘2 M in benzophenone, remained unchanged after a two hour irradiation at 5660A (n-w* transition for benzo- phenone). However it should be noted that Rosen Carried out his irradiations in benzene solutions. Benzene is, itself, a triplet sensitizer (ET = 85 kcal/mole) (57) although a poor one. It is possible, then, that what Rosen observed was, actually, sensitization by benzene. Photolysis of Heptachlor in Pure Acetone The irradiation Of 5000A of a 10"4 M solution of hepta- chlor in acetone proceeds smoothly yielding the cage compound 54 IV as the sole product during irradiation periods of up to 60 minutes (:710% raction). A plot Of cage compound concen— tration versus time gives a straight line, showing "0" order kinetics (Figure 2). The reaction rate was determined to be 1.21 x 10'7 moles/liter-minute giving a "0" order rate constant (kc) and quantum yield (p) Of 2.57 x 10'26 moles/ liter-quanta and 9.55 x 10-5 respectively. These values are based on the absorption Of the light, measured at 4.72 x 1018 quanta/minute, by the acetone. (The dependency Of the rate on the concentration Of sensitizer will be discussed in the section on mixed solvents.) “It should be noted here that the low value Of the quantum efficiency is not a true measure Of the facility Of this reaction. This term, in a sensitized reaction, must include a term for the quantum efficiency of intersystem crossing of the sensitizer. This is a very inefficient process in acetone C3310'3). Thus, the quantum yield Of a sensitized reaction is, primarily, a measure of the efficiency Of the sensitizer in converting the incident energy to a usable form. Henderson and Crosby (15) and Anderson et al. (1), in their investigations Of the potolytic reactions of systems similar to heptachlor, have found that the two processes, cage formation and photodechlorination, do not occur under the same reaction conditions. Under reaction conditions where photodechlorination occurs, lower wavelengths and no sensi- tizer, no cage compound formation is detected and when cage formation does occur, higher wavelengths in the presence of Concentration (moles/liter x 105) 55 11.00 100 10.89—— —- 99 10.78—— _. 98 O 10.67b-t O -‘ 97 10.56-— —- 96 10,45— 0 ‘ 95 10.54— -—< 94 10.25m- " 95 10.12— —- 92 Heptachlor 0.88I _a 8.0 0.77 d 7.0 . —* 6.0 0 66 0 0055 G — 500 0.44 -‘ 4.0 0.55 0 -— 5.0 0 0.22 —— 2.0 0.11 Cage Compound __ 1.0 I I I I I 10 20 50 40 50 Time (minutes) Figure 2. Rate of formation of cage compound in acetone and 5000A. Percent of Reaction Mixture 56 a triplet sensitizer, there is no photodechlorination. This is in contrast to Flotard's finding that when hepta- chlor was photolyzed in hexane solution (photodechlorination) cage compound was also Obtained in::75% yield. Our investi- gation has shown that the irradiation of undegassed solutions of heptachlor in cyclohexane at 2200A yielded a small amount of cage formation. This is probably due to sensitization by oxygen. wSince Flotard's reactions were carried out with undegassed solutions using a high pressure lamp (continuum of wavelengths to about 1850A) it is understandable that a small amount of cage formation was obtained in nonsensitizing solvents. Photolysis Of Heptachlor in Mixed Solvents Based on the above discussion it is possible to conclude that either the photodechlorination and cage formation pro- ceed through different and distinct mechanisms or that the efficiency Of intersystem crossing (S' -—+ T') is very low. Based on the study of the photolysis Of heptachlor in mixed solvents it is possible to determine which of these possibili- ties is correct. The photodecomposition Of heptachlor proceeds rapidly in mixed cyclohexane/acetane solutions (10% to 50% acetone by volume) at 5000A. Unexpectedly, the rate of decay of heptachlor is much greater in the mixed solvents (2.49 x 10-6 moles/liter-minute for 10% acetone) than in pure acetone (1.21 x 10‘7 moles/liter-minute), and the major reaction 57 product is not the cage compound (IV) but a solvent adduct (VI) where a cyclohexyl group replaces the allyl chlorine attached to carbon 1. (The identification and structure determination of this compound is discussed in a separate section below.) ~ The rates of formation Of cage compound (IV) and cyclohexyl adduct (VI) and the rate of decay of hepta- chlor (I) are shown in Figures 5 through 7 and summarized in Table II. Inspection of Table II shows that the values Of the "0" order rate constants for the reaction in pure acetone are out Of line with the other values. It is usually assumed in photochemical reactions that quenching or sensitization reactions are diffusion controlled and that the rate constant, kq, varies as a function Of 1/n. Wagner and Kochever (59) have shown that for solutions Of low viscosity Kq 7.7 _- 70 6.6 —' — 60 0’ .E 5.5 -— 5 * 50 C] <> 4.4 -- 5 6 <‘ 40 5.5—— - 50 2.2 _ El —4 20 I! 1.1 —. n 10 ' 61 l I I 0 10 20 50 40 50 60 Figure 5. Time (minutes) Reaction Of heptachlor in 50% acetone. 50% cyclohexane. Percent of Reaction Mixture Concentration (moles/liter x 105) 59 E] Adduct 0 Heptachlor 11.0) —-100 O Cage compound 9.9.. __ 90 8 .8 — — 80 7.7~‘ <> —- 70 6.6w— - 50 C] E 5.5‘_ <> J 50 4 .4 —- E] ‘ 40 9 7 5 .5 — -I 50 2.2m— _. 20 1 .1 ._ ' _ 10 o o’J/q l I l I 0 10 20 50 40 50 60 Time (minutes) Figure 4. Reaction of heptachlor at 5000A in 40% acetone, 60% cyc thexane . Percent of Reaction Mixture Concentration (moles/liter x 105) 40 I3 Adduct Q) Heptachlor 11 '0 G Cage compound 9.9.— ~O 8 .8 '— 7.7”— 0 I3 6 06 —' B 5 95‘— 5‘ o 4.4" I: 0 0 5.5?— 2.2—~ II 100 90 80 70 -E 60 —- 50 -— 4O —- 5O _. 20 10 o 10 26 so 40 50 Time (minutes) 60 Figure 5. Reaction Of heptachlor at 5000A in 50% acetone, 70% cyclohexane. Percent of Reaction Mixture Concentration (moles/liter x 105) 41 E] Adduct C) Heptachlor 11.0 _. C) Cage compound __ 9.9 h- .1 8.8 h 9 —— 7.7 p. __ B d 6.6 t' C] __ Q 'd 5.5 - __ <> 4.4 _. n _ 5.5 - o _N *0 2.2 - , __ J 1.1 -~ _ 0 . 0 10 20 50 40 50 60 Time (minutes) 100 90 80 7O 60 50 40 50 20 10 Figure 6. Reaction of heptachlor at 5000A in 20% acetone, 80% cyclohexane. Percent Of Reaction Mixture Concentration (moles/liter x 105) 42 E) Adduct 11 .0 __ O Heptachlor "100 O Cage compound 9,9‘... _, 90 8.8— —- 80 0 u 7.7-— —~ 70 6.6— —g 60 \ 5.5— I - 50 U. 4.4 _— -- 40 O 5.5 - — 50 n ‘9 2.2 —— ¢ —-‘ 20 1.1 —- o __ 10 44 M I Y i I 0 10 20 50 40 50 60 Time (minutes) Figure 7. Reaction Of heptachlor at 5000A in 10% acetone, 90% cyc lohexane . Percent Of Reaction Mixture 45 .wuscwfi\m»cmsv mao d x N>.¢ mo wuwmcmHCw m>wuommmm am so Ummmm* momdn N>.N IIII NNJN ooo.d o.mm m.mN hm.¢ mm¢.o ¢mm.o m.wm d.mm mm.m mmm.o mm¢.o N.H¢ >.>m >>.N mmm.0 hmm.o n.md ¢.mw mm.m m>m.0 mmm.o m.mm a.om >m.m mm~.o dwa.o c x >mumn Acanomumwm cowumEHom uos©©< coauMEHom ommo wuwmoomw> mcoumo< wucmsw.uwuwa\mmHOE owed x ox ucmumcoo mumm HmUHO =O:* cowuomum 0H0: HoHSUMummm mo coflpwmomeoowoouonm mo mumm mnu co wuwmoomfl> mo uommmm HH mqmfla 44 .chomEoo mmmo mo coflumEHom mo mumu may no mpfimoomfl> mo uommwm .0 musmwm x mo QNOfi x ox H N.H a.H 0.6 0.0 0.0 >.0 0.0 0.0 ¢.0 mno N.0 «.0 0 _ d _ 4 fl _ fl _ _ _ _ 4 l.0d.0 |.0N.0 J 0m.0 I:0¢.0 _ T.- o ,b l 8.0 u an mcoo w on u no . = u u u .u. $6 .0 CM .nl Owoo occumom mo cowuomnm maoe u x coauSHOm mo muflmoomw> n c .L 0>.0 .I 00.0 45 .GOwUMEHom uoswom mo mumu mnu no huwmoomfl> mo uomwmm .m musmwm x 00H mNOH X OM o.m md 04 m..o o d _ _ q ucmumcoo mum“ “mono :0: n x wGOanUM MO COMUUMHM. NHOE I II C’ N newusaom mo muflmoomw> I «.0 N.0 m.0 ¢.0 0.0 u 501— 46 .mmomo HOHSUwummn mo mocmwowmwm Enucmsv on“ so mufimoomw> mo uommmm .OH musmwm N 00H m0d x a 0.¢ 0.m 0.N 0.H 0 _ _ n A I!H.0 I.N.0 Ijm.0 I.¢.0 l:m.0 havoc HoH£U Imummn mo mocmwuwmmm Esucmso u e wcouoom mo :ofluomnm mace u x I20 0 GOHuSHom mo muwmoumw> u c 5.0 Izm.0 u 601- 47 1* rate of cage formation depends upon X n while the rate of formation of the cyclohexyl adduct and the rate of decay of heptachlor depend on, approximately, X fi§ (n = 0.182 for heptachlor and 0.225 for the adduct). Values for the viscosities for the binary mixtures were calculated by the method of Kendall and Monroe (17). That the adduct formation is a general reaction is shown by the formation of similar products from the photolysis of heptachlor in mixtures of acetone and hexane, cyclopentane, and ethylacetate. R = cyclohexyl, cyclo- pentyl, hexyl, etc The formation of the solvent adduct at carbon 1 shows that the sensitized triplet reaction or cage formation in— volves excitation of the 2,3 double bond. Since the photo- dechlorination involves excitation of the 5,6 double bond, the two reactions are not closely related but proceed through two, separate and discrete transition states. It is under- standable, therefore, that the two reactions have not been found to occur simultaneously. 48 Identification of the chlohexyl Adduct The identification of the major product formed by the irradiation of heptachlor at 5000A in mixtures of acetone and cyclohexane was made on the basis of the mass spectrum, nuclear magnetic resonance spectrum and infrared spectrum of the pure compound. (The sample was collected as a viscous liquid from the gas Chromatograph. Reinjection of the sample under a different set of conditions into a different column showed only one peak.) Some of the key features of the mass spectrum are sum— marized in Table III. The expected values for the isotOpe effects should be considered as only qualitative due to the complexity of the molecule. Additional large peaks, for which isotope effects were not calculated due to the com- plexity of the spectra, appear at m/e values of 500 (F8 = C10H5C15+), 299 (F9 = C10H4C15+), 265 (F10 = C10H5Cl4+). 264 (F11 = C10H4C14+) and 65 (F12 = C5H5). The presence of these fragments can be explained by the following scheme. F; ——>' F2 + Cl‘ F2 ——> F3 +Hc1 F1 ——-> F4 + ceHn- (cyclohexyl) F7 F1‘--€>' F5 + F5 (reverse Diels-Alder) F4——> F3 + -C1 F4—-—>- F9 + HCl F8 —> F10 + Cl' 49 TABLE III Mass Spectrum of the Cyclohexyl Adduct of Heptachlor M/e Percent of P Percent of P P ion (found) (expected) 9+6 424 81 70 9+4 422 176 161 + 9+2 420 204 196 F1 CIBHIBCIB 9 418 100 100 9+6 369 37 35 9+4 367 106 106 + 9+2 365 161 163 F2 C16H18C15 9 363 100 100 9+6 353 23 14 9+4 351 69 64 + 9+2 349 116 131 F3 C16H15C14 9 347 100 100 9+6 341 69 70 9+4 339 157 161 + 9+2 337 173 196 F4 C1°H5C16 9 335 100 100 9+6 241 41 35 9+4 239 91 106 9+2 237 149 161 F5 C5C15 9 235 100 100 9+2 150 0.95 0.67 9+1 149 11.8 12.1 96 011913 9 146 100 100 9+2 65 0.13 0.19 P+1 84 8.2 6.7 F7 C3H11 9 63 100 100 50 F8 __"* F11 + HC]. Or F9 —_")' F11 + Cl' and F6 -———*'F12 + F7 The mass spectrum of heptachlor (8) shows a very similar fragmentation pattern including a large contribution of the retro Diels-Alder type of cleavage giving peaks at m/e = 270 (C5C15+) analagous to F5 and at m/e = 100 (C5H5Cl+) analogous to F5. Further support for the assignment of carbon 1 as the site of reaction is given by the complete absence in the mass spectrum of the adduct of any fragments at m/e's of 518 (C11H11C15++) and 285 (C11H11Cl4+). These fragments would certainly be formed by the retro Diels-Alder cleavage of the adduct if the cyclohexyl group replaced any chlorine other than the one attached to carbon 1. Examination of the infrared spectrum of the adduct (above 1200 cm‘l) shows it to be almost a superimposition of the spectra of cyclohexane on that of heptachlor (Figure 11 and Table IV). Close analysis of Table IV and Figure 11 shows two significant features. First of all, the 1610 cm'1 absorption of heptachlor (v C=C). is also present in the spectrum of the adduct indicating that the double bonds are unaffected by the reaction. Secondly, 3070 cm“1 absorption (v C-H) of heptachlor is shifted to lower frequency (5050 cm'l) in the adduct. This is accompanied by a shift of the 2890 cm'1 and 2800 cm"1 (v-CH) bands by cyclohexane to higher .mGMXosoHowo 0cm HoHSUmumms .HoHnomummc awxmsoaomo mo muuommm mH 0:» Ho comflnmmeoo .aa musmwm AHIEUV mocmsvmnm . Coma 000m 00mm 000m 00mm fl 1 ,. _ _ ; IIIIIIIII\>( , mcmxmsoHowofi \\\I\\\)(\\/\\l\mmflmmmmmmm 51 r\\)/ri . Ho no ummm HuxmaoHomo o.m o.m o.m 0.9 m mCOHUHS 52 TABLE IV The IR Spectra of Heptachlor, Cyclohexane and Cyclohexylheptachlor Cm’l Cm'l ' Cm"1 Adduct Heptachlor Cyclohexane 5050 (m) 5070 (W) 2960 (sh) 2960 (w) 2925 (vs) 2890 (vs) 2900 (sh) 2860 (Sh) 2850 (vs) 2800 (vs) 3725 (W; 2223 2:; 2333 2500 (w) 1720 (W) 1725 (W) 1610 (vs) 1610 (vs) 1455 (vs) 1445 (vs) 1550 (m) 1540 (S) 1500 (w) 1295 (m) 1275 (sh) 1280 (m) 1260 (vs) 1240 (vs) (w) = weak; (m) = medium; (5) = strong: (vs) = very strong; (sh) = shoulder. 55 frequencies (2925 cm-1 and 2850 cm‘l, respectively). This is consistent with the replacement of the chlorine attached to carbon 1 in heptachlor (I) by a cyclohexyl radical to form the adduct (VI). Final and conclusive proof that carbon 1 is the re- action site is provided by a comparison of the nuclear magnetic resonance spectrum of heptachlor with that of the adduct (Table V). The addition of a broad multiplet at 8.55 T'having an area indicative of eleven hydrogens together with the splitting of H and the drastic upfield shift b (toward higher tau values) of HC show quite unequivocally, when taken along with the mass and infrared spectral evidence, that the chlorine attached to carbon 1 in heptachlor (I) is replaced by an unrearranged cyclohexyl group to form VI. n-Hexyl Adduct The acetone sensitized photolysis of heptachlor in n-hexane is not so simple as that in cyclohexane due to the ease of rearrangement of the n-hexyl radical. Indeed the reaction yields at least three major adducts in proportions which vary depending on the exact reaction conditions, e.g., length of irradiation and amount of air in contact with the reaction mixture. The first product shows a parent peak in the mass spectrum at M/e = 420 and appears to be a rather straightforward substitution product similar to that formed in the cyclohexyl case. (Although this product has not been completely characterized, it is probable that n-hexyl 54 TABLE V The NMR Chemical Shifts of Heptachlor and Its Cyclohexyl Adduct C1 C1 I R= Cl VI R= Cyclohexyl Proton Chemical Shift (tau) Heptachlor Adduct Ha 4.75 (S) (1H) 4.1 (s) (2H) Hb 4.28 (m) (1H) HC 5.20 (m) (1H) 7.55 (m) (1H) Hd 6.50 (m) (1H) 6.90 (m) (1H) He 5.90 (m) (15) 6.20 (m) (1H) HR 8.55 (m) (11H) (m) = multiplet (s) = singlet 55 radical has rearranged.) The second product is what ap- pears to be, from its infrared and mass spectra, an alcohol formed by a partial oxidation following the substitution reaction. The third product is quite interesting in that it shows a drastic rearrangement of the n-hexyl radical along with unsaturation. Its mass spectrum shows a parent peak at M/e = 418 instead of M/e = 420; its infrared spectrum shows the presence of a terminal methylene group; and its nmr spectrum shows the presence of two equivalent vinyl protons and three methyl groups, two of which are equivalent. This evidence points to a structure such as VII for the hexyl radical. Reaction Mechanisms On the basis of the evidence presented above it is possible to postulate mechanisms for both the photodechlor- ination and cage formation photoreactions of heptachlor. Photodechlorination The monodechlorination of heptachlor under the influ- ence of ultraviolet light can be viewed as a simple 56 non-chain free radical process. Heptachlor (I) is activated 1) or + Cl’ by high energy ultraviolet light (<(2800A) to give the acti- vated complex 1*. The exact nature of 1* is only specula- tion at this time. Although it is probably a singlet state, a high energy triplet (ET > 86 Kcal/mole) cannot be elimi- nated. This excited state can then decompose to yield either free radical Ia or Ib and a chlorine radical. Radicals Ia and lb can then abstract a proton from the solvent (cyclohexane) 2)Ia+®—+II+®. 5)Ib+@—+III+® to give the monodechlorination isomers II and III respective- ly plus solvent radicals. The chlorine radicals can likewise 57 4) Cl' -+ [::j -———>' HCl + [::j abstract a proton from the solvent to give hydrogen chloride and a solvent radical. Finally, two cyclohexyl radicals can combine to form bicyclohexyl. Processes such as those 5)2®6—->- shown in equations 6 through 9 are unlikely in that they a) (s; _,. (j... ——+ 7) C1' + I Ia”yor Ib + 012 o l 8) [::J +- I '—-%> Ia or Ib + 9) H- + I —-> Ia or Ib + HCl would lead to a chain process and result in a quantum yield much higher than that observed (0 = 0.025). Cage Formation and Adduct Formation The sensitized triplet reaction of heptachlor to form cage compound or add solvent radical can be viewed as pro- ceeding by a mechanism such as the following. Heptachlor(I) C1 C1 C1 1 C1 10) -———+>- 01 C' sens. 44—44———, Cl 01“ J I I 58 is activated by ultraviolet light through the sensitizer to form the triplet biradical IC. IC can do either of two things then; it.can close to form the cage compound IV C1 C1 11) E C1 C1 S .1 ( C1 .1 I IV Cl Cl (equation 11) or it can eliminate a chlorine radical to form the stable allyl radical Id (equation 12) . The allyl radical .firgcq IC Id (Id) can then react with a solvent radical, formed by the abstraction of a proton by the chlorine radical, to form 15) Cl' + R-H ——-> HCl + R' c C1 C1 ‘ . 1 14) Id + R -——> 01/ 0 VI 59 the solvent adduct VI (equations 15 and 14). The radical Id is, evidently, stable enough that it allows for rearrangement of an n-hexyl radical (R = n-hexyl). Consequently proton abstraction by Id from the solvent is not exPected and, indeed, products such as VIII have not been found. This mechanism has also been written as a non-chain process 15) Id + R.H ——> VIII because of the low quantum efficiency with which it proceeds. Kinetic Mechanism for the Triplet Reaction The above mechanism for the triplet reaction can be rewritten in a little different manner as follows: cage adduct (This diagram is not meant to indicate relative energy levels except in a qualitative and intuitive way.) The equations for each of the separate steps can be written: 60 k 16) Ho -‘—'5—> [H3] rate = kg = RS 3 kd 17) [H ]-‘-€?' HO rate = kd [H3] = Rd kc 18) [H3] -—-)- Cage rate = kc [H3] = RC k 19) [H3] —9—>- Adduct rate = ka [H3] = Ra where Ho refers to the ground state heptachlor and H3 is the excited triplet biradical. (The rate constant for sensitization is dependent, to some extent, on the amount of sensitizer and the viscosity of the solution as has been discussed previously.) If a steady state approximation is now applied to the triplet state, 3 20) $1 = 0 = ks - kdtH3] - kCIH") - kalH3] . the concentration of the triplet state, [H3], can be expressed as 21) [H3] = ks4ka + kC + kd} The rate of formation of cage compound (equation 18) can now be written as 22) RC = kskcflta + kc + kd). In like manner the rate of adduct formation and the rate of decay of heptachlor can be written as 25) Ra = kskaAka + kc + kd) and 24) RH_ = ks - kdks4> ks! kal kc then R f: I c 29) 7c - ‘EE and k5 50) ¢H- ‘: 'Eg Now, dividing equation 29 by equation 50 and substituting the values obtained for the 10% acetone solution k - 31) 6c/¢H_ '§'-43 2: 8'25 X410 5 :29 .0427 ks 1.93 x 10"3 or 62 32) kc 2’, .0427 ks Now, since, kc :5: 0.045 ka' from equation 25, 33) ka = «égigl ks = 0.946 ks and from 30) kd 9.9 kS/¢H_ = 516 ks all of the rate constants can be written in terms of ks: 30) kd 3’ 516 ks 52) kc x 0.0427 kS 33) ka = 0.946 ks If these values are substituted into equation 24 and the rate determined for the 10% acetone solution is used for RH-' one obtains a value for kS of 34) ks a: 1.243 x 10-3 moles/liter-minute. The values of these constants for all the reactions carried out in mixed solvents (cyclohexane-acetone) are summarized in Table VI. That the sensitization step is not diffusion controlled in this case can be seen by the low value for ks even though the rate of diffusion does have some effect on this step as seen above. For a truly diffusion controlled reaction, the value of k8 should be on the order of 1010 as predicted by the Debye equation (9). 65 4.09 x 6+.m x 6.09 x m.s ox . M +uoa x o m x A m +709 9 mm.m +709 2 06.5 +uoa x 04.5 6-09 x mo.a 6.09 x +N.H . x cm . 09 on om oa mc0u00¢ unmUHQM, H0H300ummm mo cowuwmomeoomcouonm cmuwuwmcmm umamwua mnu How mucmumcoo mumm vanaummm H> fimdn. 64 Reversibility of Cage Formation Irradiation at 2000A of a mixture of cage compound and heptachlor in cyclohexane, without the addition of sensitizer, yielded a decrease in the concentration of the cage compound and a corresponding increase in the concentration of the heptachlor. This reaction proceeds at a rate of 5.22 x 10‘7 moles/liter-minute (Figure 12), giving a quantum yield of 0.195 based on absorption of 2.5% of the 1.6 x 1017 quanta/ minute available energy. Although this is the first demonstration of the reversi- bility of cage formation in pesticide systems it has been shown in somewhat related systems. Hammond et al. (15) have shown the photoisomerization of [2.2.1]-bicycloheptane (IX) to [2.2.1.02'°.03'5] tetracycloheptane (X) to be a reversible reaction with a quantum efficiency of 0.08 for the cage . 1‘) hv \ fl gensitizer IX X opening. The authors have proposed two possible mechanisms for this process; §ensT a +—— IX 65 exnixrw uoIqoeeu go iuaoxea omd ONH .dooom um maficmmo 0000 no mumm Om Amwuscwev mafia 00 .NH musmfim 0H7 omfiasomeoo wmmu 0 HOHSUMumwm Av _ ¢.m N.>N (501 X JGQTI/SBTOW) uotqelqueouoa 66 or w 7 X Xa IX Although Hammond shows both forward and reverse processes to be sensitized reactions, it must be noted that he gives no data on reactions carried out in the absence of sensi- tizer. Indeed our investigation has shown that the reverse process, cage opening, proceeds in the absence of sensitizer in the heptachlor system. Our investigation has also shown that in the heptachlor system the classical, nondelocalized biradical IC, analogous to Xa, is more probable than a delocalized transition state (Ie) analogous to IXa. If an intermediate such as Ie were important in the reaction mechanism, one would be hard pressed to account for the photodechlorination and cage 67 formation being independent from one another and for the loss of chlorine from carbon 1 in forming the solvent adduct. Since a stable, delocalized system would be already present there would be no tendency to form an allyl free radical by the loss of chlorine from carbon 1. PRACTICAL IMPLICATIONS The results of these investigations have contributed a great deal to the understanding of the phenomena observed in the photodegradation of chlorinated policyclic pesticides. The differentiation between mechanism of photodechlorination and that of cage formation has explained the failure to ob- serve these reactions concurrently. Indeed, the failure to observe photodechlorination under environmental condi- tions is understandable in view of the energy requirements for this process. The inability to react by this mode coupled with the requirement for a sensitizer with a rela- tively high triplet energy explains, to a large degree, the persistency of these pesticides under environmental condi- tions. This investigation has also Opened up new paths of applied research in the pesticide area. It would, for example, be of great potential value to investigate the ef- fect of sensitizers added to the spray formulations on the persistency of the pesticides. Another area of interest would be the ability of these pesticides to form products, similar to the solvent adducts, with compounds commonly found in nature, especially those found in plant surfaces. 68 69 This study has, however, been only a step, although an important one, toward the ultimate goal of pesticide research; the ability to produce a pesticide which is toxic to insects but harmless to animal life. The attainment of this goal must await, for the most part, the better under- standing of the relationship between the structures of these compounds and their toxicity. LIST OF REFERENCES 1) 2) 5) 4) 5) 6) 7) 8) 9) 10) 11) 12) 15) 14) LIST OF REFERENCES Anderson, C. M., J. B. Bremner, I. W. McCay and R. N. Warrener, Tetrahedron Letters, 1255 (1968). Barborak, J. C. and R. Pettit, J. Am. Chem. Soc., 89, 5080 (1967). Bird, C. W., R. C. Cookson, and E. Crundwell, J. Chem. Soc., 4809 (1961). Brown, V. K. H., J. Robinson and A. Richardson, Food Cosmet. Toxicol., 1, 155 (1966). Calvert, J. G. and J. N. Pitts Jr., "Photochemistry" Wiley, New York (1966), 785. Cookson, R. C. and E. Crundwell, Chem. and Ind., 1004 (1958). Crosby, D. G. and H. O. Tutass, J. Agr. Food Chem., 22, 596 (1966). Damico, J. N., R. P. Brown and J. M. Ruth, Organic Mass Spectrometry, 1, 551 (1968). Debye, P. Trans. Electrochem. Soc., 82, 265 (1942). Dilling, W. L., H. P. Braendlin and E. T. McBee, Tetrahedron, 25, 1211 (1967). Fleck, E. 2., J. Am. Chem. Soc., 11, 1034 (1949). Flotard, R. D., Masters Thesis, Michigan State Univer— sity. Hammond, G. S., P. wyatt, C. D. DeBoer and N. J. Turro, J. Am. Chem. Soc., 86, 2552 (1964). Harrison, R. B., D. C. Holmes, J. Roburn and J. Tatton, J. Sci. Food Agr., 18, 10 (1967). 70 15) 16) 17) 18) 19) 20) 21) 22) 25) 24) 25) 26) 27) 28) 29) 30) 51) 52) 71 Henderson, G. L. and D. G. Crosby, J. Agr. Food Chem., ‘15. 888 (1967). Kahn, M. A. 0., J. D. Rosen and D. J. Sutherland, Science, 164, 518 (1969). Kendall, J. and K. P. Monroe, J. Am. Chem. Soc., 52, 1787 (1917). Klein, A. K., J. D. Link and N. F. Ives, J. Assoc. Offic. Agr. Chem., 51, 805 (1968). Koller, L. R. "Ultraviolet Radiation" 2nd Ed. Wiley, New York, N. Y. (1965). Mitchell, L. C., J. Assoc. Offic. Agr. Chem., 11, 645 (1961). Mosier, A. R., W. D. Guenzi and L. L. Miller, Science, 164, 1085 (1969). Parsons, A. M. and J. D. Moore, J. Chem. Soc., 2026 (1966). Plimmer, J. R. and B. E. Hummer, J. Agr. Food Chem., 11, 85 (1969). Robinson, J., A. Richardson, B. Bush and K. Elgar, Bull. Environ. Contam. Toxicol., 1, 127 (1966). Roburn, J., Chem. and Ind., 1555 (1965). Rosen, J. D., Chem. Comm., 189 (1967). Rosen, J. D. and W. F. Carey, J. Agr. Food Chem., 16, 556 (1968). Rosen, J. D. and D. J. Sutherland, Bull. Environ. Contam. Toxicol. 2, 1 (1967). Rosen, J. D., D. J. Sutherland and M. A. Q. Kahn, J. Agr. Food Chem., 11, 404 (1969). Rosen, J. D., D. J. Sutherland and G. R. Lipton, Bull. Environ. Contam. Toxicol., 1, 127 (1966). Schenck, G. 0., and R. Steinmetz, Bull. Soc. Chem. Belg., 11. 781 (1962). Schenck, G. 0., and R. Steinmetz, Chem. Ber., 96, 520 (1965). 55) 54) 55) 56) 57) 58) 59) 40) 41) 72 Stedman, R. J. and L. S. Miller, J. Org. Chem., 32, 55 (1967). Sutherland, D. J. and J. D. Rosen, Mosquito News, 28, 155 (1968). Szychlinski, J., Roczniki Chem.,'§§, 1709 (1961). Szychlinski, J. and L. Litwin, Roczniki Chem., 51, 671 (1965). Turro, N. J., "Molecular Photochemistry," W. A. Benjamin, New York (1965). Wagner, P. J., Tetrahedron Letters, 1755 (1967). Wagner, P. J. and I. Kochever, J.-Am. Chem. Soc..‘29, 2252 (1968). WOlf, W. and N. Kharasch, J. Org. Chem., 59, 2495 (1965). Zabik, M. J., R. D. Schuetz, and W. Burton to be pub- lished. l,‘ll‘,“ N“ Uflo WW3 MMQ "2 Slll|4l N” Am3 I." w"