...... ......... ’ EVADENBE AAA AA A AAAAAACAA. AAAEAAEAAAAE 2;; AA AAA WA: A AAAAAAAAAAAANS {IF PHENYL ' AAAsAA fur AAA Dame AAA PA A3 AAA“ AAAGAN STAAE UNAVERSATY PHALAP ALLEN K359 AA)” * ............. W‘ * LIBRARY Michigan State [IDAVCTSHQV This is to certify that the thesis entitled EVIDENCE FOR A 1,4-BIRADICAL INTERMEDIATE IN THE TYPE II PHOTOREACTIONS 0F PHENYL KETONES presented by Philip Allen Kelso has been accepted towards fulfillment l of the requirements for Ph.D. degree in Chgnifiry ”(or professor Date 8/ 231 71 0-7639 ABSTRACT EVIDENCE FOR A 1,4-BIRADICAL INTERMEDIATE IN THE TYPE II PHOTOREACTIONS 0F PHENYL KETONES By Philip Allen Kelso Several investigations were undertaken in order to establish the intermediacy of a l,4-biradical in both nonradiative decay and type II product fermation from the triplet state of phenyl alkyl ketones. The energetics of phenyl ketone triplet state chemical processes were evaluated. The formation of a l,4-biradical is estimated to be greater than 20 kcal/mole exothermic. Spin-conserving, concerted elimination of a fragment in its triplet state is endothenmic for valerophenone, y-methylvalerophenone and y-phenylbutyrophenone, but slightly exothermic for B,y-diphenylbuterphenone, DPB. Photolysis (366 nm) of DPB produces equal yields of acetophenone and stilbene. Stilbene is formed initially in a trans-c1 ratio of 64:1. The quantum yield of acetophenone formation increases from 0.ll in benzene to 0.19 in the presence of 2 M §:butyl alcohol. A 8 sec was obtained for DPB in benzene. The 8 sec']. triplet lifetime of 5 x lo- rate of l,5-hydrogen transfer is estimated as 4 x l0 These results indicate that triplet stilbene is not fbrmed. Therefbre concerted fragmentation apparently does not contribute to type 11 product formation even when spin conservation is energetically feasible. Irradiation of (+)-y-methylcaprophenone[(4S)-(+)-4~methyl-l-phenyl- l-hexanone], MC, results in racemization and type II fragmentation to acetophenone and cyclization to 2-methyl-2-ethyl-l-phenylcyclobutanols. The sum of the quantum yields fOr product fbrmation (0.26) and racemization (0.77) is unity in benzene. Racemization and product formation are quenched equally by 2,5-dimethyl-2,4-hexadiene. Addition of 0.2 M t:butyl alcohol results in a higher quantum yield for product formation (0.36) and a corresponding lower quantum yield for racemization (0.57). In tybutyl alcohol the quantum yield of product formation is unity and photoracemization is not observed. The cyclobutanols that were isolated from the photolysis of MC in tybutyl alcohol show some optical activity: [aJSS = -0.2 t 0.l° (benzene). This behavior establishes that a 1,4- biradical intermediate forms and returns to ketone after nearly complete equilibration about the B.y~carbon-carbon bond and that there is little, if any, photophysical decay of triplet phenyl ketones possessing tertiary y-hydrogens. In addition, this behavior strongly suggests that the type II products are formed entirely via the l,4-biradical inter- mediate, since alcohol solvent prevents return to ketone and enhances. product formation. The effect of y-deuteration on the photochemical behavior of nonanophenone (l-phenyl-l-nonanone) was examined. In benzene the quantum yield of type II product from nonanophenone (0.30) was slightly lower than that from nonanophenone-y,y-d2 (91.5% y-deuterated) (0.32). The triplet lifetime in benzene of the nonanophenone-y,y-d2 is three times longer than that of nonanophenone indicating a kH/kD value of 4.8. In the presence of t:butyl alcohol the quantum yield of product formation from nonanophenone (0.99) is larger than that of nonanophenone-y,y-d2 (0.83). These data indicate a kH/kD value of 1.5 on reverse hydrogen transfer in the l,4-biradical. The inefficiency of product formation in the presence of alcohol can be attributed to a small amount of l,5-biradical formation via 6-hydrogen abstraction. EVIDENCE FOR A 1,4-BIRADICAL INTERMEDIATE IN THE TYPE II PHOTOREACTIONS OF PHENYL KETONES By Philip Allen Kelso A THESIS Submitted to Michigan State University in partial fulfillment of the requirements fbr the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1971 ACKNOWLEDGEMENTS I would like to thank Professor Peter J. Wagner for his patient direction of an Opinionated and inexperienced graduate student and fbr his assistance in obtaining research assistantships. The assistance of my fellow graduate students in the form of discussion and illustration of chemical techniques is greatly appreciated. Financial support by the Department of Chemistry in the form of teaching or research assistantships made this work possible. TABLE OF CONTENTS Page INTRODUCTION ........................... l l. Photoinduced l,5-Hydrogen Transfer Reactions of Ketones ....................... l 2. Significance ...................... 4 3. Mechanism ........................ 6 a. Excited state responsible for product formation. . . 6 b. l,4-Biradical formation in the primary process . . . 8 c. Reactivity and modes of excited state decay ..... lO 4. General Problem and Specific Approaches ......... l3 a. The pathway of product formation and the mode of excited state deactivation ............. l3 b. Energetic Liminations. ............... l4 c. Biradical fOrmation and triplet state deactivation . l5 d. Intermediacy of the biradical in product formation . 19 e. Concerted pathway .................. 20 RESULTS ............................. 23 l. Thermochemical Estimates ................ 23 2. (S)-(+)-y-Methylcaprophenone .............. 23 a. Products ...................... 25 b. Photoracemization and optical activity in the photoproducts .................... 25 c. Photokinetics .................... 26 TABLE OF CONTENTS (Continued) Page 3. Nonanophenone-y,y—d2 and Nonanophenone .......... 29 a. Products ....................... 29 b. Photokinetics .................... 29 4. B,y-Diphenylbutyrophenone ................ 30 DISCUSSION ............................ 36 l. Energetic Limitations .................. 36 2. (S)-(+)-y-Methylcaprophenone ............... 37 3. y-Deuterium isotope effects ............... 4l 4. 3,y-Diphenylbuterphenone ................ 45 5. Summary ......................... 48 EXPERIMENTAL ........................... 49 A. SPECTRA ........................... 49 B. PREPARATION AND PURIFICATION OF KETONES ........... 49 l. 8,y-Diphenylbutyrophenone (DPB) ............. 50 2. (S)-(+)-y-Methylcaprophenone (MC) ............ 51 3. Nonanophenone (NH) .................... 52 4. Nonanophenone-y,y-d2 (ND) ................ 53 a. Preparation of l-heptanol-2,2-d2 and its tosylate . . 53 b. Preparation of l-phenylethylidenecyclohexylamine. . . 54 c. NonanOphenone-y,y-d2 ................. 55 d. Deuterium content .................. 55 5. Valerophenone ...................... 56 ii TABLE OF CONTENTS (Continued) Page C. OTHER MATERIALS ....................... 56 l. Solvents ........................ 56 2. Internal Standards ................... 56 3. Quencher ........................ 57 4. Miscellaneous ...................... 57 D. IRRADIATION PROCEDURE .................... 57 E. PHOTOLYSATE ANALYSIS .................... 58 F. GENERAL PROCEDURES OF PHOTOKINETIC STUDIES ......... 58 1. Quantum Yields ..................... 58 2. Dependence of Relative Quantum Yield on Quencher Concentrations ..................... 59 3. Dependence of Optical Rotation on Conversion ...... 59 G. ISOLATION OF PHOTOPRODUCTS FROM MC IRRADIATIONS ....... 60 l. Recovered MC ...................... 60 2. Cyclobutanols from MC .................. 60 LITERATURE CITED ......................... 81 APPENDIX I. THE KINETICS OF PHOTORACEMIZATION .......... 88 APPENDIX 11. THERMOCHEMISTRY ................... 93 A. EXCITATION ENERGIES ..................... 93 l. Ketones ......................... 93 2. Olefins and Enols .................... 93 8. REACTION ENTHALPIES ..................... 94 l. Retro-ene Reaction ................... 94 TABLE OF CONTENTS (Continued) Page a. From heats of formation ............... 94 b. From bond energies ................. 97 2. Biradical Formation ................... 104 B. REACTION ENTROPY ...... - ................ 104 iv TABLE OF CONTENTS (Continued) Page a. From heats of formation ............... 94 b. From bond energies ................. 97 2. Biradical Formation ................... 104 B. REACTION ENTROPY ...... - ................ 104 iv TABLE II III IV VI VII VIII IX XI XII XIII XIV XV XVI XVII LIST OF TABLES Survey of Carbonyl Compounds that Yield Type II Fragmentation and/or Isomerization Photoproducts ..... Enthalpies of Some Potential Processes of Triplet State Ketones ...................... Hxnnkinetic Parameters of (S)-(+)-y-Methylcaprophenone. . Photokinetic Parameters of Nonanophenone and Nonanophenone-y,y-d2 ................... Product Distribution: Product Distribution: Nonanophenone and Nonanophenone-y,y-d2 ................... Product Distribution B.y-Diphenylbutyrophenone ...... y-Methylcaprophenone in Benzene ............. y-Methylcaprophenone in Benzene at High Conversion. . . . y-Methylcaprophenone in t;Butyl Alcohol ......... y-Methylcaprophenone y-Methylcaprophenone in Benzene-trButyl Alcohol Mixture . Nonanophenone in Benzene: Dependence of Product Quantum Yields on Initial Ketone Concentration .......... Nonanophenone-y,y-da in Benzene: Dependence of Product i Quantum Yields on I in Benzene: Nonanophenone-y,y-d Dependence of Cyclo- butanol Quantum Yie d on Initial Ketone Concentration . . Nonanophenone, Acetophenone Quantum Yields in the Presence of Varying Amounts of t;8utyl Alcohol ...... Nonanophenone-y,y-d2: Acetophenone Quantum Yield in the Presence of Varying Amounts of trButyl Alcohol ...... y-Methylcaprophenone ....... in Benzene with Triplet Quencher. . tial Ketone Concentration ...... Page 24 28 34 63 64 65 66 67 68 69 7O 71 72 73 74 75 1"“"I i TABLE XVIII XIX XX XXI XXII XXIII XXIV XXV XXVI XXVII XXVIII LIST OF TABLES (Continued) Nonanophenone-y,y-d in Benzene-3:8utyl Alcohol: Dependence of AcetoBhenone Quantum Yield on Initial Ketone Concentration .................. Nonanophenone: Relative Acetophenone Quantum Yield in the Presence of Quencher .............. Nonanophenone-y,y-d : Relative Acetophenone Quantum Yield on the Presenée 0 Quantum Yield of Acetophenone Formation by 0.35 M B.y-Diphenylbutyrophenone ............... B,y-Diphenylbutyrophenone: The Effect of the Addition of t-Butyl Alcohol and Varying Amounts of 2,5-Dimethyl-2,4-hexadiene ............. Reaction Enthalpies of Retro-ene Reaction from Heats of Formation ................... Carbon-Hydrogen and Carbon-Carbon Bond Energies . . . . Carbon-Hydrogen Bond Energies in Radicals . .‘ ..... Pi Bond Energies .................... Reaction Enthalpies of Retro-ene Reaction from Bond Energies ........................ Heats of 1,4-Biradical Formation ............ vi f Quencher ........... Page 76 77 78 79 80 96 99 101 102 103 106 FIGURE 1 LIST OF FIGURES Plots of 1n(ai/a) versus 1n ([KJi/[K]): 0.l M solutions of (+)-y-methylcaprophenone in Q benzene, A benzene plus 0.09 M DMHD, and I benzene plus 0.2 M t—butyl alcohol. . Dependence of the quantum yield of product formation on initial ketone concentration: Acetophenone from nonanophenone—y,y-d2 in benzene (O) and in benzene plus 6.5 M in t-butyl alcohol (0); acetophenone from nonanophenone in benzene (A); cyclobutanols in benzene from nonanophenone-y,y-d2 (<3) and from nonanophenone (£5) ........................... The effect of t:butyl alcohol on the quantum yields of acetophenone formation from nonanophenone-y,y-d2 (O) and nonanophenone (A) ................ The dependence of the relative quantum yields of acetOphenone formation by nonanOphenone-y,y-d2 (O) and nonanophenone (A) on DMHD concentration ....... The dependence of the relative quantum yields of stilbene formation from DPB on DMHD concentration. . . . Pmr spectrum of cyclobutanols collected from irradiation of MC in t:Bu0H ................ vii Page 27 32 33 35 62 INTRODUCTION 1. Photoinduced 1,5-Hydrogen Transfer Reactions of Ketones. In the 1930's, Norrish and co-workers1 examined the photochemistry of a number of acyclic aliphatic ketones in the gas phase and in alkane solutions. They identified two major photoreactions of dialkyl ketones, which they designated the type I and type II reactions. Carbon monoxide and a number of other compounds, mostly hydro— carbons, are fOrmed in the type I reaction. The formation of these products can be accounted fOr by the fragmentation of the ketone to an acyl and an alkyl radical, which form the products in subsequent reactions. In alkane solution, carbon monoxide and compounds arising from the hydrogenation of the free radicals are formed with the corresponding induction of unsaturation in the solvent. The quantum yields of the type I reactions are very small in solution near 25°C, but approach the gas phase values at higher temperatures. All the ketones studied that possess a hydrogen-bearing carbon gammg_to the carbonyl function undergo the type II reaction. A simpler ketone and an alkane, corresponding to the cleavage of an a,B carbon-carbon bond of the parent ketone, are formed in a one to one ratio by the type II reaction in both the gas phase and in alkane solutions. Norrish interpreted the independence of the gross yields R' 0 110 > R/u\ + l 7 CH3 R R = alkyl, R' = hydrogen or methyl and the product ratio from the reaction phase or the temperature as evidence for the formation of the type II products by an intramolecular process rather than by fragmentation to a pair of free radicals. Subsequent observations lend further support for the intra- molecular nature of the type II photoreaction. In the gas phase the quantum yields of the type II photoproducts of 2-hexanone2, methyl neopentyl ketone3, and other dialkyl ketones“ are equal and nearly independent of the temperature. Yang and Yang5 photolyzed 2-octanone and 2-nonanone in cyclohexane solution and obtained equimolar yields of acetone and the corresponding 1-a1kene. The intramolecular nature of the reaction led Rice and Teller6 and Davies and Noyes2 to propose that the type II fragmentation of dialkyl ketones proceeds by transfer of a y-hydrogen to the carbonyl oxygen with cleavage to an olefin and an enol. The enol subsequently tautomerizes to its keto fOrm. o '_,._.__> R1 +(R——>RJ\+K Experimental evidence has been obtained for the formation of an enol. Srinivasan7 determined the deuterium content of the products and recovered starting materials from the gas phase photolysis of acetone and 2-hexanone in a DZO-treated reaction vessel and 2-hexanone- 5,5-d2 in a HZO-treated reaction vessel. The recovered starting materials did not show any exchange but the acetone formed showed significant incorporation of the deuterium or hydrogen, respectively. This indicated the formation of a precursor to acetone which is sufficiently long lived to exchange hydrogens or deuteriums with the walls of the silica reaction vessel. Later, McMillan and co-workers8 detected a transient infrared absorption, whose frequency could be assigned to enolic hydrogen-oxygen stretch, during the gas phase photolysis of 2-pentanone. The decay time of this transient absorption was identical to the rise time of acetone but much less than the rise time for ethylene. Recently, products of a third photoreaction have been detected. Yang and Yang5 obtained l-methylcyclobutanol (12%), gigftrgg§;1-methyl- 2-propylcyclobutanol (17%) and gjsftragsrl-methyl-2—butylcyclobutanol (10%), in addition to the type II fragmentation products, from the photolysis of 2-pentanone, 2-octanone and 2-nonanone, respectively, in cyclohexane. 2-Pentanone also has been reporteda:9 to form l-methyl- cyclobutanol on photolysis in the gas phase. Cyclobutanol formation is H I n #R' hv : “ /’ R = methyl, R' = H, n-propyl or n-butyl most likely an intramolecular reaction which proceeds, as the type II fragmentation, via a 1,5 transfer of a y-hydrogen to the carbonyl oxygen. This photoreaction will be referred to as the type II isomerization reaction or a 1,5-hydrogen-transfer reaction. 2. Significance. Subsequent investigation“:1°:11 has revealed that both type II isomerization and fragmentation are characteristic of a wide structural range of acyclic ketones that possess a saturated, hydrogen-bearing y-carbon. In addition, photoproducts have been obtained from many other types of carbonyl compounds that are formally the result of an analogous type II fragmentation or isomerization reaction. The type II fragmentation products of esterslz, aldehydes13 and presumably most other types of carbonyl compounds are formed intramolecularly. Thus the survey of the various types of carbonyl compounds that yield type II photoproducts (Table I) is a good indication that 1,5-hydrogen transfer is a characteristic excited state process of almost all types of carbonyl chromOphores. 1,5-Hydrogen transfer reactions, formally analogous to the type II photofragmentation, are also induced in ketones, esters and other carbonyl compounds by electron impact30 (the McLafferty rearrangement) and in esters31 and perhaps ketones32'36 by pyrolysis (gjsfelimination or retro-ene reaction). Thus 1,5-hydrogen transfers may be characteristic not only of electronically excited carbonyl compounds but of all high energy forms of carbonyl compounds. The actual mechanism(s) of the photoinduced transfers in the various carbonyl compounds are of obvious interest. TABLE I. Survey of Carbonyl Compounds that Yield Type II Fragmentation and/or Isomerization Photoproducts H 0 CR2 = x_. ._15’ {C\Y/ _ Carbonyl Compound Fragmentation Isomerization Products Products X Y Z Observed Observed H CH2 CR2a yes“:13 yes27 Alkyl CR2 CR2 yes‘HlO’11 yesS CH2 0 yes‘+ yes28 Phenyl CR2 CR2 yes10 yes29 CH2 0 yes28 yes28 CH2 NR yeslk,15 ?lk,15 CH2 S yes16 NIb H 0 CR2 yes12 NI Alkyl 0 CR2 yes“’12’17 NI Phenyl 0 CR2 yes18 NI Alkyl O C=0 yes“ NI Alkoxyl CH2 CR2 yes1+ NI Hydroxy CH2 CR2 yesl9’20 NI Amino CH2 CR2 no19 NI Alkyl C=0 CR2 no21 yes21 Phenyl C=O CR2 no”:23 yes”,23 Hydroxy C=0 CR2 yesz“ n02“ Alkyl C=0 0 yes25 NI Phenyl C=0 0 yes26 NI aR = Hydrogen,alky1 and/or phenyl group. bNI = No indication. In particular, if all photoinduced hydrogen transfers proceed by a common primary photochemical process, a comparison of the selectivity of the various chromophores toward various types of y-hydrogens would be possible. In addition the identity of the primary photochemical process of light induced hydrogen transfers with the elementary step of the electron impact or thermally induced reactions would provide an opportunity to compare selectivities of, and hence the natures of, the different high energy processes. Such comparisons would provide a meaningful experimental background upon which to construct a useful, coherent concept of primary photochemical processes37. Such a beast is painfully absent from photochemistry at present33. However, the evaluation of the Opportunity to make such comparisons still awaits the unequivocal determination of the type II photoreaction mechanisms. 3. Mechanism a. Excited state responsible forgproduct formation. Early attempts to determine the multiplicity of the ketone excited state yielding the hydrogen transfer products were based on the qualitative effects of triplet quenchers on product formation and gave apparently contradictory results. The quantum yields of the type II fragmentation of 2-hexanone39 or methoxyacetone“0 were not quenched by oxygen or by other triplet state quenchers. This observation suggested that the formation of the type II fragmentation products occurs via an excited singlet state. However, biacetyl quenches the phosphorescence, the type II fragmentation and the type II isomerization of 2-pentanone in the vapor phase under conditions such that it does not disturb the fluorescence of 2-pentanone”1. Quantitative studies of the effect of piperylene, a quencher of ketone triplet states“2, on the formation of type II fragmentation products from 2-pentanone and 2-hexanone“3 and from 2-octanone“” indicated that both the singlet and triplet states are involved in the type II fragmentation reactions of these ketones. The ratios of the type II fragmentation quantum yield in the absence of piperylene to those in the presence oflfiperylene (¢E/¢F) increase to a limiting value with increasing piperylene concentration. That is, plots of ¢FI¢F versus piperylene concentration, called Stern-Volmer plots, are nonlinear. Such behavior indicates that part of the product formation is proceeding via an unquenchable excited state in each case. Wagner and Hammond“3 attributed the unquenchable portion of product fOrmation to singlet state reaction. They estimated that 41% and 13% of type II product formation proceeds by the singlet states in 2-hexanone and 2-pentanone, respectively. Subsequent investigations‘*‘5"*6 have confirmed that the type II fragmentation products arise from both the singlet and triplet states of dialkyl ketones. In contrast to the behavior of dialkyl ketones, acetophenone formation from butyrophenone and valerophenone in benzene is quenched with constant efficiency by all concentrations of piperylene; that is, the Stern-Volmer plot is linear33. Wagner and Hammond concluded that type II product formation by these ketones proceeds entirely via triplet states. Subsequent investigations“7"*8 of a number of phenyl ketones have confirmed the generality of this conclusion. Several researchers have examined the effect of triplet quenchers on the quantum yields of cyclobutanol formation. Coulson and Yang“5 found that §j§:dichloroethylene quenches 91% of the cyclobutanol formation but only 62% of the acetone formation from 2-hexanone in hexane solution. Similarly Barltrop and Coyle found that 5 M piperylene quenches 93% and 92% of the cytlobutanol formation from 2-octanone and 5-methyl-2-hexanone in solution while quenching only 53% and 45% of the ketones disappearance, respectively. Thus, in contrast to the type II fragmentation products, more than 90% of cyclobutanol formation by dialkyl ketones occurs via the triplet manifold. Wagner and Hammond"3 observed that the ratio of acetophenone formed to butyrophenone or valerophenone consumed was independent of piperylene concentration. Since cyclobutanols were formed, the ratio of acetophenone to cyclobutanol formation must also be independent of piperylene concentration. This has been confirmed for butyrophenone“9. Thus all the cyclobutanol fbrmation by phenyl alkyl ketones occurs via the same triplet state that yields the type II fragmentation products. b. 1,4-Biradica1 formation in the primary process. The products of the photoinduced hydrogen transfer reactions could a_priori arise directly from the excited state by two distinct primary photochemical processes. ..H” 0” UR' -—> ““f—TR' ////;;7 R \\ 0* li Alternatively, a single primary process, producing a 1,4-biradical, could give rise to both products. OH R -3 * R / 3 OH ‘ R “' \Ri+FR The biradical pathway is attractive, since biradical formation can be regarded as the intramolecular analog of the hydrogen abstraction process which initiates the photoreductionll’12 of acetone, aceto- phenone, and other simple ketones. Several product studies have provided evidence for the fermation of a 1,4-biradical. Yang and co-workers50 found that l-methylcyclohex- 3-enol (1.4%) as well as 2-vinyl-l-methylcyclobutanol (6%) are among the products from the irradiation of 6-hepten-2-one in pentane. They argued that the concurrent formation of a cyclohex-3-enol52 and of a cyclobutanol was indicative of the presence of a common delocalized biradical. However, Schulte-Elte and Ohloff51 found that the cyclohex-3-enol formed in the photolysis of (6S)-2,6—dimethyloct-7-en-3- one was formed with 62% retention of configuration and that the recovered ketone (5%) was 77% racemized. They suggested that the cyclohex—3-enol is fbrmed stereospecificany in a separate concerted process and that the optically inactive portion arose from ketone which had previously under- gone photoracemization via internal disproportionation of the proposed 1,4-biradical. They also found that (-)-2,6-dimethyloctan-3-one 10 underwent significant photoracemization. Previously, Orban and co-workers53 had determined that the §i§_and the trgn§_cyclobutanols obtained from the photolysis of (5R)-5,9-dimethyl-2-decanone were formed with partial retention of configuration (greater than 16% and 12%, respectively). The observed photoracemization of the above ketones suggests that the cyclobutanols may have been formed initially with significantly greater retention of configuration than that observed at high conversion. At the very least the photoracemization demands the formation of a 1,4-biradica1 which can undergo internal disproportionation to starting ketone. The lack of complete retention of configuration in the cyclobutanols suggests that the 1,4-biradical participates in their formation. Since the cyclobutanols originate primarily via the triplet state, the 1,4-biradical is probably also formed by the triplet state. c. Reactivity and modes of excited state decay. The intra- molecular disproportionation of a 1,4-biradical intermediate to the starting ketone has been invoked to explain the inefficiency of product fbrmation by phenyl alkyl ketones, j,g,, product quantum yields appreciably lower than unity“3. . 0 H > M I / R R”' Indeed, the observed photoracemization of dialkyl ketones“,53 confirms that at least part of the inefficiency in product formation from dialkyl ketones arises in this manner. 11 However, most of the inefficiency might be assigned to radiationless (photophysical) decay of the singlet and/or triplet states of dialkyl or phenyl ketones. Indeed this inefficiency of product formation from phenyl ketones must arise from some type of triplet state deactivation since the quantum yields of triplet state formation (intersystem crossing quantum yield a. ) of phenyl ketones are typically unity5“’55. 15C Such values suggest a high rate of intersystem crossing for phenyl ketones. Since the singlet lifetime of acetophenone has been shown to be less than 0.1 nsec56 and ¢. = 1.0, the rate of intersystem 15C 10 se -1 crossing is greater than 10 c . The fluorescence lifetime of benzophenone (apparently in methylcyclohexane at room temperature) has ‘2 sec57. Thus the rate of intersystem crossing in benzophenone is m2 x 10]] sec“. been measured as m5 x 10' If these rates of intersystem crossing are typical, phenyl ketones can not be expected to exhibit significant nonradiative decay from their excited n,n-singlet states. It can be shown58 that the slopes of the Stern-Volmer plots obtained by Wagner and Hammond“3 in their study of the reaction multiplicity, are the products of the triplet state lifetime in the absence of quencher ¢O F _ ¢F - l + quT [Quencher] (1) (IT) and a quenching rate constant (kq). Since k had been shown to q be equal to or close to the rate of diffusion53, values of 1.3 x 10'7 and 7 x 10'9 sec.1 can be estimated for the triplet state lifetimes of butyrophenone and valerophenone, respectively. In order for a photophysical decay process to account for the inefficiency of product 12- formation by valerophenone, it would have to have a rate of 7 x 107 sec-1. Since this rate is much faster than known photophysical decay processes, Wagner and Hammond invoked the reversal of hydrogen transfer in the proposed 1,4-biradica1 to explain the inefficiency. This conclusion requires that the triplet lifetime equal the reciprocal of the rate for the hydrogen transfer process. The data then yield relative rates of primary to secondary (butyrophenone to valerophenone) hydrogen transfer of approximately 1:18. This is the same relative selectivity shown by benzophenone and by trbutoxy radicals towards hydrogen donor559. Wagner“:61 found that the quantum yields for the decomposition (¢_K) of butyrophenone and valerophenone aneunity in t:butyl alcohol compared to m0.5 in benzene or hexane. Since, the lifetime of valerophenone, estimated from the efficiency of the type II product quenching of 2,5-dimethyl-2,4-hexadiene, was lower in t;butyl alcohol than hydrocarbon solvents by only a factor of two, Wagner concluded that the polar solvent quenches the decay process rather than increasing the rate of reaction. The addition of small amounts of t;butyl alcohol produce a significant increase in the quantum yieldkcf'acetophenone formation from valerophenone. Wagner concluded that t;butyl alcohol must be solvating either thelexcited state or the postulated 1,4- biradical intermediate. Wagner did not find any precedent for the former possibility but pointed out that involvement of the biradical hydroxyl proton in hydrogen bonding with t7butyl alcohol could reasonably be expected to retard reverse hydrogen transfer and increase the proportion of the 1,4-biradicals proceeding to products. These l3 solvent effects strongly suggest that most of the inefficiency of product formation by phenyl ketones in hydrocarbon solvents arises by the internal disproportionation of the supposed 1,4-biradica1 intermediate. Recently it has been shown that the inefficiency of product formation from dialkyl ketones arises in both the singlet and triplet states52 and that the photoracemization of dialkyl ketones with an asymmetric center at the y-carbon arises predominately via the triplet state50. Apparently most of the inefficiency of triplet state product formation by dialkyl ketones is quenched by tfbutyl alcohol“6’5°’51’52. Thus a 1,4-biradical is responsible for part and perhaps all the inefficiency of product formation from the triplet states of dialkyl ketones. y-Deuterium substitution in 2-hexanone results in both an increased quantum yield of acetone formation from the triplet state and an increased triplet 1ifetime“5. This can be interpreted as evidence for the formation of the fragmentation products via the 1,4-biradical. A normal primary isotOpe effect on the hydrogen transfer process and on the reverse hydrogen transfer process of the biradical would increase the triplet lifetime and decrease the inefficiency due to reverse hydrogen transfer, respectively. However, part or all of the increase in product formation can be accounted for in terms of an increase in the efficiency of intersystem crossing from the singlet state“5. 4. General Problem and Specific Approaches. a. The pathwayiof product formation and the mode of excited state deactivation. In spite of the evidence fer the fermation of a biradical 14 intermediate it has not been established that the products are formed entirely through such an intermediate. Nor has it been established that the inefficiency of product formation is due totally to the internal disproportionation of this intermediate. If the reactivities of the excited states toward 1,4-biradical formation are to be evaluated from excited state lifetimes and from product quantum yields, it will be necessary to partition (1) the rate of product formation between the concerted and biradical pathways and (2) the inefficiency of product fermation between excited state deactivation and internal dispropor- tionation of the 1,4-biradica1. Since product formation occurs entirely via the triplet state of phenyl ketones, an examination of the mechanism of their type II reactions should be the most straightforward undertaking. Knowledge of the triplet state behavior of phenyl ketones would then provide a basis for the analysis of dialkyl ketone triplet state behavior. Therefore an investigation of the pathway(s) of type II product formation and the mode(s) of triplet state deactivation of phenyl ketones was undertaken as the subject matter for this thesis. b. Energetic Liminations. The thermochemistry of various 3 prior1_processes of the triplet ketone and the biradical were estimated for a number of phenyl alkyl ketones in order to determine if any of these processes are sufficiently endothermic that their intervention could be dismissed. In particular it was suspected that internal disproportionation of the triplet biradical to triplet state ketone would be appreciably endothermic. Therefore the reaction enthalpy fer 3 isomerization of triplet ketone K to triplet biradical 3BR (process BR) was evaluated for several phenyl ketones. 15 3 K :> BR (Process BR) It was also suspected that the concerted fragmentation of triplet ketone occurring with conservation of spin, i.e., fragmentation to triplet enol or triplet olefin, would be endothermic for most phenyl ketones but might be exothermic for B,y-diphenylbutyrophenone (l,3,4-triphenyl-l-butanone), DPB. Accordingly, the reaction enthalpies DPB of the concerted fragmentation of triplet ketone (3K*) to triplet enol (3E*) and ground state olefin (0) (Process TE), to ground state enol (E) and triplet olefin (30*) (Process TO), and to ground state enol and olefin (Process SP) were evaluated for several phenyl ketones. 3K* -——————{;> 3E* + 0 (Process TE) 3K* -——————{§> E + 30* (Process TO) K* -——————{}> E + 0 (Process SP) 3 c. Biradicgj formation and triplet state deactivation. The observation of photoracemization of a phenyl ketone with a center of optical activity at the y-carbon would affirm the formation of a 1,4-biradica1 and its internal disproportionation in these ketones. Also the quantum yield of photoracemization would provide a lower l6 limit for the amount of internal disproportionation occurring and, in conjunction with the quantum yield of product formation, an upper limit on the efficiency of triplet state photophysical decay. Therefore, (S)-(+)-y-methylcaprophenone [(4S)-(+)-4-methy1-1-phenyl-1-hexanone], MC, was prepared and its photochemical and photokinetic behavior assessed. §2H5 ('1 H “‘ CH l 3 c H C6H5/ \mZ/C 2 MC The method fer obtaining the quantum yield of racemization requires further conment. If internal disproportionation of the proposed 1,4- biradical is in fact responsible for all the inefficiency of product formation by phenyl ketones in benzene solution, photoracemization of a ketone such as MC will be efficient and will fall off rapidly with increasing conversion. Consequently, a reliable extrapolation to zero conversion is required. The quantum yield of racemization can be obtained from the dependence of the relative optical purity or perhaps the optical rotation of the photolysate on conversion and on the quantum yield of product formation. The following general scheme for the photoracemization of a phenyl ketone can be written on the basis of previous mechanistic studies: l7 SCHEME I .EEQESEE. _J§flg;_ Ks + 110 ——> K; ([KSJ/[KDI * 3 9: KS —9 P kP[ KS] 9: 3 * Ks '—‘9 Ks "d[ Ks] * 3 * Ks ”"9 KR kIN[ Ks] + no ——> K; ([KR1/[K])I :3‘ K; “"9 KR kd[3KR] K; ‘9 l(s kINEBKR] KS and KR represent the S and R enantiomers of the optically active ketone, respectively. An asterisk indicates a triplet excited state or states. Since the extinction coefficients of KS and KR for nonpolarized light are equal, the rate of light absorption by an enantiomer is equal to the absorbed intensity times the ratio of the enantiomer concentration to the total ketone concentration [K]. No distinction is made between the rates of enantiomeric processes. P represents the type II photoproducts. l8 Steady state analysis of Scheme I (see Appendix I) yields the following expression for the dependence of optical purity of the ketone, (OP), on the ketone concentration: L . a [K]. 1n OP 1 = ( (tiff: )IH—WL- (2) The subscript i, indicates an initial value. ¢RAC is the quantum yield of racemization in the limit of zero conversion. 2k ¢RAc ‘ kIN + kP +kd ¢+P is the quantum yield of product formation. k = P (4) kIN + kP + kd ¢+P If the optical rotations of the products are minor then the rotation of the photolysate (a) will be due to the ketone, OP will be directly proportional to a/[K] and the dependence of a on [K] will be given by Equation 5. Thus if the optical rotations of the products are L “' ¢RAC [K31 1n ; = (l + ¢+P )1n—IR]——- (5) small the ratio ¢RAC/¢+P and hence ¢RAC can be obtained directly from the photolysate's optical rotation. 19 d. Intermediacy of the biradical in product formation. Neither the observation of photoracemization or the kinetics of photoracemization would require the intermediacy of the 1,4-biradical in product formation. Since the quantum yield of intersystem crossing is unity for phenyl ketones and internal disproportionation of the proposed 1,4-biradical to triplet ketone is probably sufficiently endothermic that it does not occur, an inverse y-deuterium isotope effect on the quantum yield of product formation could arise only from a primary isotope effect on the rate of internal disproportionation by the 1,4-biradical. The observation of such an isotope effect on product formation would thus provide good evidence for the intermediacy of the supposed 1,4-biradical in product formation. In addition, if the radiative decay of the triplet is minor as suspected, the isotope effect on the rate of hydrogen transfer can be evaluated directly from the effect on the triplet state's lifetimes. The relative lifetimes can be assessed from the quenching efficiency of an appropriate triplet quencher. A phenyl alkyl ketone possessing secondary y-hydrogens was chosen for a y-deuterium isotope study fer several reasons. The previous isotope effect studies were performed with dialkyl ketones that possessed secondary y-hydrogens“5’53. Also phenyl alkyl ketones with secondary y-hydrogens have sufficiently short lifetimes to reduce the problem of impurity quenching to reasonable proportions but sufficiently long to undergo measurable quenching before the onset of static quenching. Previous studies“7"‘8 had shown that nonanophenone exhibited quantum yields and a triplet lifetime typical of long-chain phenyl alkyl 2O ketones. Thus nonanophenone-y,y-d [l-phenyl-l-nonanone-4,4-d2], ND’ 2 was prepared and its photokinetic behavior was compared with that of nonanophenone. C5Hll e. Concerted pathway. It is possible that a concerted pathway contributes to product fermation. Concerted product fermation could g_prjgrj_proceed by either a primary process in which the system's spin is conserved or by a primary process in which there is a change in the spin of the system. Comparison of ketone excitation energies with the excitation energies of the olefinic fragments suggests that a concerted spin- conserving pathway for product formation is endothermic for most ketones, one likely exception being B,y-diphenylbutyrophenone (DPB). Therefore, if a concerted pathway were contributing to product formation it would be a pathway involving a spin change for most phenyl ketones but could be a spin-conserving pathway for DPB. Since it is expected65 that a spin-conserving concerted process will be more rapid than a similar process proceeding with a spin change, triplet biradical formation should be intrinstically faster than the supposed concerted process which must proceed with a spin change. However, there is little substantial information regarding the magnitude of the retarding effects of incorporating a change in spin on the rate of an elementary chemical process. 21 Gill and Laidler66 have compared the best experimental value to several theoretical values of the pre-exponential factor for the thermally induced decomposition of singlet N20 to singlet N2 and a triplet oxygen atom. The experimental value was less than the theoretical values by only a factor of 10 to 100. The thermal isomerization of olefins, frequently discussed in the earlier literature65 as proceeding through the triplet state, is no longer believed to do 5057. Alternatively, if one adopts the view that primary photochemical processes are essentially radiationless decay processe568a59, then the retarding effects of a multiplicity change on the rate of a primary photochemical process should be analogous to those on the rates of internal conversions between electronic excited states. The rates of internal conversion between excited states of similar energies but different multiplicities are generally slower than the rates of the analogous process between excited states of the same multiplicity by a factor70 of 102 to 106. This suggests that the incorporation of a spin change into an elementary chemical process will probably retard that process by at least a factor of 100. Unfortunately, this factor is too small to rule out altogether the possible competition of the concerted pathway (spin not conserved) with the formation of triplet biradical. However, it is large enough to suggest that if, in general, concerted product formation does complete significantly with biradical formation it should be a major pathway for product formation by DPB (fer which spin conservation is energetically feasible). 22 If concerted product formation is indeed the major pathway in DPB the quantum yield of the type II fragmentation products (aF) from DPB should be larger and its triplet lifetime (1T) shorter then the values of ¢F and TT exhibited by other comparable phenyl ketones. In addition, the triplet olefinic fragment generated (presumedly stilbene) will undergo its typical triplet state processes: decay to gi§_and trag§_ isomers. Since acyclic olefins typically undergo photosensitized gig/traps isomerization to gis/trgp§_mixtures distinctly different than those induced by thermolysis or free radical catalysis, the departure of the initial gig/trap§_ratio of stilbene formed from the ratio of the thermal equilibrium mixture would indicate the presence of triplet stilbene and concerted product formation. Therefore, B,y-diphenylbutyrophenone was prepared and its photochemical and photokinetic behavior investigated. RESULTS 1. Thermochemical Estimates. The estimates obtained for the reaction enthalpies AHBR; AHTO; AHTE and AHSP of processes BR, T0, TE and SP, respectively, of several ketones are presented in Table II. The reaction enthalpy of the corresponding retro-ene reaction (AHRE) was employed together with the appropriate singlet-triplet vertical excitation energies (AEST) of the excited species to evaluate AHTO’ AHTE and AHSP. The values of AHRE and AEST that were employed are also presented in Table II. The methods and sources used in arriving at values of AHRE and AHBR are given in Appendix II. The sources for the values of AEST employed are also given in Appendix II. Estimates of the potential energy difference between the ground state and the twisted triplet of the olefin71, are given in parentheses below the corresponding vertical excitation energies. Values of‘AHTO obtained from these values are also presented in parentheses below the values obtained from the vertical excitation energies. These values of AEST are probably too low and hence the values of AHTO obtained from them, represent an upper limit for the exothermicity of that process. 2. (S):(+)-y-Methylcaprophenone. 25 The preparation of (S)-(+)-y-methylcaprophenone72, MC, [aJD = +9.8° (benzene), is described in the experimental section. 23 24 .mmm + «in .mm 2.8 o + «mm + «.xm .m._. 9.0 .pqwgumaam us» an zq new xmgmcm cowpmuwuxm umpawgunpmpmcpm use up m + N + «gm.oe no + N +.sgm.¢m mo + N +.g mmuauPucw um Saveumnamu wmumuvucw ammuocn mg» mo aapmgucm covuummg mg» mp mmc "mumpm vmuwuxm ca xmwemuma cm .mamam uopawgp a mmumowucw m unwgumgmaam a "umsgom Poem on» m .umsgoe cwwmpo one o .mcoamx mucmmoeamg x ”mpoasamn .mpos\pmux cw mmzpm> pp.u mzmu o Am +V ANSV mm- e_+ N_+ we- mm Fm m.NN ¢.NN+ mINUNIUNzoNzuomxou acacagaoaxuzn % -pxcmgqu> Amp+v Ammv m N N m 8 MN. ep+ Nm+ «e- mm Fm m.NN ¢.NN+ zuzu =0 :8 z u acccaga028Fa> m2“ -nguue-» ANF+V Ammv m N N N m m mN- o_+ mm+ Ne- mm _m m.NN ¢.om+ :0 =8 :8 18¢ I u acocmgaoaapa> e_- NN+ om+ Pm- NN _N 8.0» N.mN+ :8 :8 =8 20w :8 acocaxa=-N «axe NPIQ oeza amxa tum tam sxm Nazq mgauuagum van acouux .o.a.m mmcoumx mumpm umpavgp No mammmuosm P~_u=ouoa meow $o mmwapngucm .HH wgm<~ 25 a. Products. Irradiation (313 nm) of MC in degassed benzene or t;butyl alcohol results in the formation of a mixture of 2-methyl-2-ethy1-1- maylcyclobutanols and acetophenone in ratios of 0.12 and 0.06, respectively. These were the only significant products detectable by gas-liquid phase chromatography. The 2-methyl-1-butene, presumedly formed in equilimolar yield with acetOphenone, could not be observed under the conditions of the analysis. b. Photoracemization and optical activity in the photoproducts. The irradiation (313 nm) of 0.1 M benzene solutions of MC resulted in a rapid decrease in the optical rotation displayed by the solution. The optical rotation (10 cm path, A = 578 nm) of a 0.10 solution drops from +0.198° before irradiation to +0.120° at 9% conversion and reaches a residual value of +0.006° at greater than 99% conversion. The specific rotation of a sample of unreacted ketone, isolated from several 0.1 M benzene solutions that had been irradiated to ~16% conversion, was only 69% of the original value. A mixture of the cyclobutanols was isolated from a number of 0.1 M benzene solutions of MC that had been photolyzed to greater than 98% conversion. The isolated cyclobutanol mixture displayed a very small positive rotation ([6];5 = +0.9 1 0.9°, benzene) which apparently contributes to part or most of the residual rotation at high conversions. In contrast, the decrease in the optical rotation of a 0.1 M t:butyl alcohol solution of MC parallels the decrease in the ketone concentration. A mixture of cyclobutanols was isolated from a 0.18 M t;butyl alcohol solution of MC that had been irradiated to greater than 90% conversion. 25 The mixture displayed a negative rotation ([a]D = -0.2 t O.T§ benzene) 26 that was too small to contribute significantly to the rotation of the solution. The opposite rotations of the isolated mixtures probably reflects different relative proportions of the two diasteriomers produced in benzene and in trbutyl alcohol. For example, the ratio of gig; to trap§:l-methyl-l-phenylcyclobutanol produced in the photolysis of valerophenone in benzene and tfbutyl alcohol are 4:1 and 2:1, respectively51. Since both the rotation of the mixture of cyclobutanols and the residual rotation of the photolyzed solutions of MC are small, the optical rotation of the solution (a) should be a good measure of the optical rotation of the remaining ketone in either benzene or t:butyl alcohol. Thus the ratio ¢RACI¢+P can be obtained directly from the slope of a plot of 1n(ai/a) versus ln([K]i/[K]) as indicated by Equation 4. c. Photokinetics. Accordingly, degassed samples of a 0.1 M solution of MC were irradiated for various periods of time and the relative ketone concentrations and the optical rotation, a, of each sample were measured. The resulting plots of 1n(ai/a) versus ln([K]i/[K]), are presented in Figure l and the slopes of the plots in Table III. In the course of determining the dependence of the optical rotation on ketone concentration, the quantum yields of acetophenone formation and molar ratio of cyclobutanol to acetophenone were determined. The quantum yield of acetophenone formation from MC (0.1 M in benzene) and the 27 1.0— 1 A 0.54 + '1 0.0 i l 1 l T T l fiF # I 0.0 0.1 0.2 1n([K],./[K] FIGURE 1. Plots of ln(a /a) versus ln([K]./[K]): 0.1M solution; of (+)-y-methylc3prophenone in O benzene, A benzene plus 0.09 M DMHD, and I benzene plus 0.2 M _t_-butyl alcohol. 28 TABLE III. Photokinetic Parameters of (S)-(+)-y-Methy1caprophenonea. L f b c ¢RAc d e L Solvent ¢F 01 (l + ¢+P ) ¢+P ¢RAC benzene 0.228 0.027 4.0 0.255 0.77 benzene + 0.09 M 0 0.122 0.014 4.0 0.136 0.39 benzene + 4 2% t-BuOH 0.32 0.03 2.6 0.35 0.57 - 4 2 6 t:BuOH 0.944 0.055 1.0h 1.00 0h aDegassed, 0.10 M solutions irradiated at 313 nm and 25°C. bQuantum yield of acetophenone formation. These values have a precision of 14%. cQuantum yield of cyclobutanol formation. These values have a precision of 17%. dThe slope of the 1n(oi/a) versus ln([K]i/[K]) plot. eThe sum of 0F and 01. fThe quantum yield of racemization of zero conversion. The precision of these values is estimated as t7%. 90 is 2,5-dimethyl-2,4-hexadiene. hAt 17% conversion. 29 ratio (yf acetophenone formation to ketone disappearance were found to be constant throughout the range of conversions employed (3 to 20%). The quantum yields of acetophenone formation (0F) obtained at moderate conversions in each solution are presented in Table IV with the quantum yields of cyclobutanol formation (¢I) obtained from them. 3. Nonanophenone-y,y—d2 and Nonanophenone. The preparation of nonanophenone-y,y-d2 (ND) is described in the experimental section. Mass spectral analysis of the parent ion peaks indicated that the ND contained 91.5% y-deuteriums. a. Products. Acetophenone and compounds that are presumed to be cyclobutanols are the only significant products observed on the photolysis (313 nm) of ND or nonanophenone (NH) in degassed benzene or t:butyl alcohol solutions. The 1-heptene expected could not be observed under the conditions of analysis. Only one additional photoproduct (~0.2%) was detected at low or moderate conversions by gas-liquid phase chromatography. The molar ratio of cyclobutanols to acetophenone was 0.28 and 0.30 for 0.07 M NH and 0.07 M ND in benzene, respectively, and 0.16 for both 0.07 M NH and ND in tfbutyl alcohol. These ratios are similar to those reported for other phenyl ketones with secondary y-hydrogens72. b. Photokinetics. The quantum yields of acetophenone formation (0F) and cyclobutanol formation (01) were determined at several ketone concentrations and extrapolated to zero ketone concentration in order to correct for solvation of the biradical by the parent. ketone7“ and for potential impurity quenching. Plots of ¢F and 0C versus ketone 30 concentration are presented in Figure 2 and the extrapolated values in Table IV. The effect of added tfbutyl alcohol on ¢F was examined fer both NH and ND (Figure 4). ¢F reached 0.85 for 0.07 M NH but only 0.69 f0r 0.07 M ND in the presence of 6.8 M trbutyl alcohol. A study of the effect of starting ND concentration on 0F in the presence of 6.3 M tybutyl alcohol, was undertaken. 0F showed only a slight dependence (Figure 2) with an extrapolated value of 0.71. The linear Stern-Volmer plots shown in Figure 4 were obtained from the photolysis of 0.03 M NH and ND in degassed benzene solutions in the presence of varying amounts of 2,5-dimethy1-2,4-hexadiene. The relative slopes of the Stern-Volmer plots indicate that 91% y-deuteration enhanced the triplet lifetime by a factor of 3.05. 4. B,y-Diphenylbutyrophenone. Acet0phenone and stilbene were formed in equal yields on photolysis of benzene solutions of B,y-diphenylbutyrophenone (DPB) at 366 nm. A traps/21§_ratio of 64.0 was obtained from the 366 nm irradiation of 0.051 M DPB in benzene to 2.4% conversion. This ratio was lower on irradiation of 0.35 M DPB in benzene for longer periods of time. Quantum yieldscfi'0.ll and ~0.09 were obtained fer the formation of acet0phenone on photolysis (366 nm) of 0.35 M and 0.05 M DPB in degassed benzene. The quantum yield of acetophenone formation from 0.05 M DPB in benzene increased 1.8 times on the addition of 2.0 M trbutyl alcohol. The formation of stilbene by 0.05 M DPB in benzene was quenched by the presence of 2,5-dimethyl-2,4-hexadiene (DMHD). A Stern-Volmer plot of the relative quantum yield of stilbene formation 1 versus the quencher concentration was linear with a slope of 2.4 M- (Figure 5). Quantum Yield 31 . 70‘ 0 R .30] .28‘ .261 .24“ .22-i .08“ .07‘w’ll— .06 I r l f I i I T T I i 0.0 0.10 0.20 x (\ [Nonanophenone or Nonanophenone-y,y-d2] FIGURE 2. Dependence of the quantum yield of product formation on initial ketone concentration: Acetophenone from nonanophenone-y,y-d2 in benzene (O) and in benzene plus 6.5 M in t-butyl alcohol ((9); acetophenone from nonanophenone in benzene (A); cyclobutanols in benzene from nonanophenone-y,y-d2 (C)) and from n0nanophenone (£5). quantum yield of acetophenone formation 32 ——4l 0.8“ “ a 0.5~ 0.4-4 0.2 F I l 1 T | l 1 0.0 2.0 4.0 6.0 [t;butyl alcohol] FIGURE 3. The effect of t-butyl alcohol on the quantum yields-of acetophenone formation from nonanophenone-y,y-d2 (O) and nonanophenone (A). 33 .cowumgucoucou oxzo co A4v 32.2858: EB ADV «viifmcocmfiocmcoc .3 5.5228 mcocmsaopmum we mupmwx.szucm:a m>wumpmg ms» mo mocmucmamu one mozzou no.0 No.0 Po.o _ . r _ _ .e mmeHu o.~ (9/02) 34 TABLE IV. Photokinetic Parameters of Nonanophenone and Nonanophenone-y,y-déq Ketone Benzene solution tyButyl alcohol added ¢F I +P qTT f f f g 9 NH 0 232 0.069 0 30 28.8 0.85 0.99 f f f f,h f,h ND 0 245 0.077 0 32 87.9 0.714 0.83 aDegassed solutions irradiated at 313 nm and 25°C. bQuantum yields of acet0phenone formation. Individual values have a precision of 14%. cQuantum yields of cyclobutanol formation. Individual values have a precision of 17%. dTotal quantum yield of product formation. eStern-Volmer quenching slope for 0.03 M NH and ND in benzene. fExtrapolated to zero ketone concentration. 96.8 M trButyl alcohol. h6.3 M tyButyl alcohol. ¢°/¢ 35 2.0- 1.8.4 1.6-i 7. 1.4-1 1.2 ~ 1.0 . 0.0 011 0.2 0.3 074 [DMHD] FIGURE 5. The dependence of the relative quantum yields of stilbene fonmation from DPB on DMHD concentration. DISCUSSION 1. Energetic Limitations. The thermochemical evaluation suggests that triplet biradical f0rmation by triplet phenyl ketones is on the order of 20 kcal/mole or more exothermic. Since this value is well outside the uncertainty in the thermochemical estimates, there is little possibility that any of the biradical returns to triplet ketone. As expected the results of the thermochemical evaluations suggest that the concerted fragmentation of phenyl ketones such as y-methyl- valerophenone are endothermic if spin is conserved whether spectro- scopic (planar) or nonspectroscopic (twisted) triplets of a fragment are generated. A better lower limit for the endothermicity awaits a better evaluation of the potential energy difference between the twisted triplet state and the ground state [AEST (twisted)] of the carbon-carbon double bond species generated. However, the endothermicity of the spin-conserving process for these ketones is apparently real since values used for AEST (twisted) are probably too low by at least 5 kcal/mole or more71. In contrast to the other phenyl ketones, the fragmentation of triplet B,y-diphenylbutyrophenone to either spectro- scopic or twisted triplet stilbene and ground state enol is apparently exothermic . 36 37 2. (S)-(+)-y-Methylcaprophenone. Since 0 is unity, the observed photoracemization of (S)-(+)-y- lSC methylcaprophenone (MC) requires the formationiof an intermediate by a triplet state of MC which can proceed to racemized ketone. The identical values of 3.0 obtained for ¢RACl¢+P in benzene solution and in benzene solution with sufficient 2,S-dimethyl-2,4-hexadiene added to reduce ¢+P by half, indicates that product formation and racemization proceed via the same triplet state and perhaps via the same intermediate. Since the proposed 1,4-biradical is the only plausible intermediate which could result in photoracemization, the photoracemization of MC is best discussed in terms of a 1,4-biradical which can undergo internal disproportionation to ketone and possibly product formation. As discussed previously, the thermochemical evaluation strongly indicates that none of the 1,4-biradical returns to triplet ketone. Thus the total quantum yields of product formation and of racemization cannot exceed unity and will provide a lower limit for the quantmm yield of all chemical processes. Accordingly, racemization was not observed in t:butyl alcohol in which ¢+P was unity. The values of ¢+P = 0.26 and ¢RAC = 0.77 obtained in benzene indicate that the total of all chemical processes from the triplet state is unity and that the rate of photophysical decay from the triplet state must be much less than the rate of hydrogen transfer in the triplet state of MC in benzene solution. 0n the addition of a small amount of tfbutyl alcohol (2%), ¢+P increases to 0.36, ¢RAC correspondingly decreases to 0.57 and ¢RAC + ¢+P is again near unity. Therefore, the rate of photophysical decay from the 38 triplet state must also be much less than the rate of hydrogen transfer in benzene-trbutyl alcohol mixtures. Thus the inefficiency of product formation by MC in benzene or benzene-tfbutyl alcohol mixtures arises from the internal dispropor- tionation of a 1,4-biradical intermediate. The effect of the presence of trbutyl alcohol is entirely compatible with Wagner's conclusion61 that EbetYI alcohol quenches the internal disproportionation of l- hydroxy-l,4-biradicals rather than a photophysical decay process of the ketone triplet state. The effect of t:butyl alcohol also lends further support to the fermation of the products primarily through the 1,4-biradical inter- mediate. The only plausible alternative processes for product formation are the set of concerted 1,5-hydrogen transfer processes that have been discussed previously. Since Wagner61 observed that valerophenone has only a slightly longer triplet lifetime in tybutyl alcohol than in benzene (this is also true of y-methylvalerophenonee”’“7, which should be a good model for MC) solvation of the excited triplet state by t:butyl alcohol would have to retard the formation of the 1,4-biradical while enhancing the rate of concerted 1,5-hydrogen transfer processes by a similar factor. Such equal and opposite solvent effects on such similar and rapid processes are unlikely and with the magnitude required are without precedent in transition state chemistry or the photophysical decay of excited states. These results also substantiate the conclusion of Wagner and co-workers that the triplet lifetime of those phenyl ketones for which the quantum yield of product formation is unity in the presence of 39 t:butyl alcohol equals the reciprocal of the rate of 1,5-hydrogen transfer“7:51. Thus the relative rates of 1,5-hydrogen transfer by the triplet states of these ketones can be assessed from the relative slopes of their Stern-Volmer quenching plots. Since Wagner found that the lifetime of valer0phenone in tybutyl alcohol is similar to that in benzene, the maximum quantum yields of product formation by ketones in the presence of tfbutyl alcohol should provide a reasonable measure of the probability of biradical formation for those ketones whose quantum yields of product formation do not reach unity. The photolysis of MC can be represented by the set of processes presented in Scheme II. SCHEME II 40 In Scheme II, KS and KR represent (S)- and (R)-MC, respectively. BRS and BRR represent those biradicals formed by (S)-MC and by (R)-MC, respectively, that have not yet lost their initial conformations about the B,y-carbon—carbon bond. BRe represents the biradicals that have "forgotten" their initial conformation about the B,y-carbon-carbon bond and P represents the photoproducts. A superscript 3 indicates a triplet state and an asterisk indicates an electronic excited state. No distinction is made between the rate constants of enantiomeric processes. Also no distinction is made between singlet and triplet biradicals since (1) all spin-conserving pathways from triplet biradicals are appreciably endothermic and (2) the interaction between the radical centers will be small and interconversion between singlet and triplet states should accordingly be facile. The following expressions for the quantum yield of product formation (0+P), the quantum yield of racemization at the limit of zero conversion (¢RAC) and the dependence of the optical rotation of the photolysate (a) on the ketone concentration ([K]), can be obtained by a steady state analysis of Scheme II entirely analogous to that made for Scheme 1: ¢+P = ¢ischRPP (51 ¢L = 0 P P P (7) RAC isc BR eK RAC L 0. ¢ [K]. 1 RAC 1 = + ¢isc is the quantum yield of intersystem crossing to the triplet state PBR = kH/(kH + kd) is the probability that the triplet state proceeds 41 to the biradical. PP is the overall probability that the biradicals will proceed to products. Pek is the probability that the equilibrated biradical (i.e., BRe) returns to ketone. PRAC is the probability that an initially formed biradical (BRS or BRR) undergoes loss of its initial conformation before reacting. PRAC = kRAC/(kRAC + kiP T kiK) (9) L An estimate of PRAC can be obtained from ¢RAC and ¢+P if it 15 assumed that PeK = 1 - PP. Substituting 1 - ¢+P for PeK into Equation 7 and rearranging yields: L PRAC % ¢RAc/(‘ ' 45+13) (‘0) The average value of PRAC’ from the three studies in which racemization occurred, is m0.96. This value is reasonable in view of the optical activity observed in the isolated cyclobutanols and values of approximately unity obtained for ¢RAC + ¢+P‘ It is unlikely that an accurate assessment of PRAC can be made from photokinetic studies. The cyclobutanols will have to be separated and their optical purities established in order to obtain an accurate estimate of PRAC' 3. [IfDeuterium Isotope Effects. Little or no inefficiency of product f0rmation from nonanophenone (NH) was observed in the presence of 6.8 M tfbutyl alcohol. In contrast product formation from nonanophenone-y,y-d2 (ND) was inefficient in the presence of 6.3 M t;butyl alcohol. The quantum yield of product formation (¢+P) by ND extrapolated to only 0.83 at zero ketone concentration. 42 This requires that either y-deuteration induces some f0rm of triplet state deactivation or that internal disproportionation in the deuterated 1,4-biradical is inefficiently quenched by t:butyl alcohol. The latter seems unlikely because of the rapid initial rise in 0F at low concentration of added alcohol (Figure 3). An estimate of the ratio of the hydrogen abstraction rate constant (kH) to the deuterium abstraction rate constant (kD) can be made from to ND (TH/TD). As usual the H quantum yield of product formation by NH or ND (aflp and ¢2P’ respectively) the ratio of the triplet lifetime of N can be expressed as in Equation 11 as the product of the quantum yield of intersystem crossing (a. ), the probability of 1,4-biradical lSC formation (PBR), and the probability of product f0rmation from the biradical (PP). The expression for the probability of 1,4-biradica1 formation from triplet ND (PER) is given by Equation 12, where FD and FH ¢+P = ¢iscPBRPP (1]) D PBR = (kDFD + kHF”)TD (12) are the fractions of y-deuteriums and y-hydrogens, respectively, in ND. The probability of biradical formation from triplet NH (P3P) can be expressed as the product kHTH. Combining the expressions for PER and PER and rearranging yields Equation 13. F0 k /k = (13) H D («cH/r“) (PER/P3,.) - F“ 43 In one extreme the inefficiency of product formation in tgbutyl alcohol, could be attributed totally to some kind of triplet state deactivation. If this is so, PER/PgR would be equal to the value of .fiP/.2p obtained in tfbutyl alcohol and kH/kD = 4.8. In the other extreme, the inefficiency could be attributed totally to residual formation of ketone by the biradical. If this were so, PgR/PgR would be equal to unity and kH/kD = 3.8. Thus one can conclude that kH/kD lies between 3.8 and 4.8 without assigning the source of the inefficiency. Deuterium isotope effects of m7, 6 and 1.7 have been reported for the triplet type II fragmentation reactions of 2-hexanone45, 5-decanone63 and y-hydroxy-y-phenylbutyrophenone75, respectively. Thus the upper limit of 4.8 appears to be in better agreement with previous results. The rate of y-deuterium transfer (kD) may be sufficiently slow that o-hydrogen abstraction (1,6-hydrogen transfer) can compete. Products arising from o-hydrogen abstraction (tetrahydrofuranols) have been observed in B-alkoxy phenyl ketones that also have y-hydrogens75. However, the reasonable material balance obtained in t:butyl alcohol (coupled with the expectation that the products of o-hydrogen abstraction, cyclopentanols, and ND should have the same detector response per mole of material) and the lack of other significant products requires that the intermediate 1,5-biradical proceed very inefficiently to products in t;butyl alcohol. Recently Zepp77 has observed that the quantum yields of cyclopentanols from two different ketones are low and insensitive to the presence of polar solvents, including tfbutyl alcohol. 44 If the inefficiency is due totally to o-hydrogen abstraction and subsequent reversion to ketone an estimate of the ratio of k to H the rate constant for 6-hydrogen abstraction (k6) can be made. The expressions for the triplet lifetimes are given by Equations 14 and 15. 1h” = k + k (14) D - 1/T - kDFD + kHFH + k6 (15) Combining Equations 14 and 15 and solving for kH/ko yields Equation 16. 1 - (TH/TD) ( ) k/k = 16 H 5 (TH/TD) - F” - FD(kK/kH) Taking kH/kD as 4.8 yields kH/k6 ~14. This corresponds to a quantum yield of product formation of ~0.94 for NH in t;butyl alcohol compared to the value of 0.99 obtained. Walling and Padwa78 found that 1,6-shifts of secondary hydrogens occur ~1/15 as readily as 1,5-shifts of secondary hydrogens in tfalkoxy radicals produced in the photolysis of the corresponding tfalkylhypo- chorites. Zepp77 has estimated that kH/k6 ~20 for secondary hydrogens glphg to a methoxy group in phenyl ketones. Thus it appears reasonable to assign most of the inefficiency of product formation from HD in t;butyl alcohol to 6-hydrogen abstraction and subsequent efficient reversion of the resulting 1,5-biradical to ketone. However, some of the inefficiency could arise from the inefficient quenching of internal disproportionation of the 1,4-biradical by trbutyl alcohol. Recent work by Bonner79 indicates that hydrogen- 45 bonding is more extensive than deuterium-bonding in CH3OD-CH3OH mixtures. Thus trbutyl alcohol may solvate the deuteriobiradical less efficiently than the protiobiradical. This could be tested by examining the effect of pyridine on the quantum yield of product formation by ND. If the inefficiency of product formation is assigned to 6-hydrogen _ Alc BR ' ¢+P proceeds to products (Pp) in benzene is m0.39 for ND and ~0.30 for NH. abstraction, then P and the probability that the 1,4-biradica1 The probability that the 1,4-biradical reverses biketone (PK) is 60.70 for NH and m0.6l for ND (note that PP + PK = 1). This requires at most a positive isotope effect on the rate of internal disproportionation (reverse hydrogen transfer) of only 1.5. This low value probably reflects little bond breaking in the transition state for and hence the high rate of reverse hydrogen transfer in the 1,4-biradical. 4. B,y-Diphenylbutyrophenone. Since triplet stilbene decays80 to a traps/gi§_ratio of 40/60, the trgps/gi§_ratio of 64/1 obtained on the photolysis of DPB to low conversions is incompatible with the formation of more than 2% of the stilbene in its triplet state. The ratio obtained probably reflects a conformational preference for a "trapsfl configuration of the phenyl groups about the B,y-carbon-carbon bond in the 1,4-biradical. Such a ratio could also arise in the concerted fragmentation of the triplet to ground state olefin and enol. In any case, the decrease in the tggps/gj§_ratio observed at higher conversions is apparently due to photosensitization by the ketone. Photosensitization was not a problem at short irradiation times due to the short lifetime (i.e., low Stern- Volmer quenching slope) of DPB. Although the thermochemical estimates 46 indicate that triplet enol might be formed in a concerted fragmentation of triplet DPB, its formation is 7 kcal/mole less exothermic than the fermation of triplet stilbene. Since there is no reason to expect a concerted electronic rearrangement to proceed via the less exothermic of two otherwise similar pathways the concerted spin-conserving product fermation by DPB is unlikely. Since the substitution of a B-phenyl group should produce only a weak inductive effect on the rate of formation of a 1,4-biradical, a triplet lifetime and an acetophenone quantum yield similar to those of y-phenylbutyrophenone (y-PB) were expected. However, the quantum yield (0.11) and the lifetime (5 x 10'8 sec) obtained in benzene were only ml/S as large as those for y-PB“7. Similarly the maximum quantum yield of acetophenone formation from y-PB81 in the presence of trbutyl alcohol was m5 times larger than that for DPB. The magnitude of this effect is much too large to be due to an impurity in thesample of DPB used. The lower acetophenone quantum yield is not due to a change in the proportion of isomerization to fragmentation products since Caldwell and Fink82 found that acetophenone formation accounted for 80% of DPB disappearance on photolysis of 0.27 M benzene solutions. A similar value is obtained for y-PBZ9. Alternatively, such an effect could be due to a triplet state deactivation process induced by the presence of a B-phenyl group. Indeed the quantum yield of acetophenone formation and the triplet lifetime of B-phenylbutyrophenone (B-PB) are only ml/13 of the values for butyrophenoneel. If it is assumed that the quantum yield of product fermation in the presence of t:butyl alcohol (¢§;c) equals the 47 probability of biradical formation by the triplet state in benzene, the rate constant for biradical formation (kH) is given by oig7tT and the rate constant of triplet state deactivation (kd) by (1 - ¢£;c)/TT- In 9 1 this way Wagner and co-workers81 obtained kd k l x 10 sec- for both 8 1 B-PB and DPB and k = 4 x 10 sec' for DPB. The similarity of the kd H values obtained for B-PB and DPB coupled with other observations of carbonyl triplet deactivation by B-phenyl substitution“,81+ lend support to this interpretation. 8 The kH value of 4 x 10 sec-1 thus obtained for DPB is identical to the value obtained for y-PB. The fact that the k exhibited by H DPB, for which the concerted spin-conserving formation of products is energetically feasible, is the same as the kH exhibited by a ketone (y-PB) for which such a process is energetically unfeasible lends further support to the conclusion that a concerted pathway does not contribute to the formation of type II fragmentation products in phenyl ketones. This interpretation of the results f0r DPB is valid, of course, only if the incorporation of a spin-change into a process actually retards that process, that is, if singlet products are not f0rmed directly by the triplet state. Recently Stephenson and co-workers85 have shown that the triplet state type II fragmentation products of dialkyl ketones are formed with only a small degree of stereospecificity. By analogy, little concerted product formation by a pathway involving a spin-change is to be expected for phenyl ketones. Thus it appears that even when a spin-conserving, concerted pathway is energetically feasible little, if any, concerted product formation takes place. 48 5. Summary. The photoracemization of MC establishes that an intermediate is formed from triplet MC that reverts to racemized ketone. A 1,4-biradical is the most likely such intermediate. Since photoracemization and photoproduct from MC are quenched equally by 2,5-dimethyl-2,4-hexadiene, 1,4-biradical formation and product formation must occur via the same triplet state. The value of unity obtained for the sum of the quantum yields of racemization and product formation coupled with the endo- thermicity of the return of the 1,4-biradical to triplet ketone indicates that photophysical decay of triplet MC does not occur. The lack of photophysical decay and the quenching of racemization by tfbutyl alcohol suggests that the photoproducts are formed entirely via the 1,4-biradical intermediate. The optical activity displayed by the cyclobutanols suggests that the 1,4-biradical does not achieve complete equilibration about the B,y-carbon-carbon bond. The inverse y-deuterium isotope effect on the inefficiency of product formation from nonano- phenone further substantiates the intermediacy of the 1,4-biradical in product formation. The lack of triplet stilbene formation by DPB and the lack of an enhancement on the rate of 1,5-hydrogen transfer from DPB relative to y-PB, indicate that concerted fragmentation does not compete with 1,4-biradical formation even when concerted fragmentation with conservation of spin is energetically feasible. These results are in accord with Wagner and co-workers' suggestions”,61 that a 1,4-biradical is formed with 100% efficiency from the triplet state of those phenyl ketones for which the quantum yield of product formation is unity in t:butyl alcohol. EXPERIMENTAL A. SPECTRA. Infrared (ir) spectra were recorded on a Perkin-Elmer 237B grating spectrophotomer as liquid films between NaCl plates or as solutions in chlorofbrm or carbon tetrachloride. All ir spectra were calibrated at 1601.4 cm'1 with polystyrene film. Proton magnetic resonance (pmr) spectra were recorded on a Varian A-60, a Varian A-56/60, a Jeolco C-60—H or a Varian T-60 spectrometer at ambient probe temperature as solutions in carbon tetrachloride using tetramethylsilane as an internal standard. The mass spectra were determined by Mrs. Lorraine Guile of this department on a Hitachi-Perkin-Elmer RMU-6 mass spectro- meter. Ultraviolet spectra and solution absorbances were measured with a Gilford 200 series spectrophotometer. Optical rotation was measured in a l-dm x 1-cm cylindrical cell with a Perkin-Elmer Model 141 automatic polarimeter at 589 nm. 8. PREPARATION AND PURIFICATION OF KETONES. All extracts in the product isolation procedures were dried over anhydrous magnesium sulfate and condensed at ~25torr and 95% ND. Repeated recrystallization at low temperature, from purified pentane followed by chromatography on neutral alumina gave nonanophenone of >99.8% purity by glp chromatography. d. Deuterium content. The mass spectra (15 ev ionization potential) of NH and ND were each obtained repeatedly under as similar conditions as possible in the region of the parent peaks. NH showed peaks of mass units 217, 218, 219 and 222 with averaged relative intensities of 4.6, 100.0, 21.8 and 4.7, respectively. ND showed peaks of mass units 218, 219, 220, 221, and 22 with relative intensities of 4.0, 17.5, 100.0, 20.1, and 3.2, respectively. If it is assumed that 56 relative intensities of the P-l, P and P+l peaks are the same for each deuterio compound, then the deuterated compound contains 3% nondeuterated, 11% monodeuterated and 86% dideuterated nonanophenone. That is, the nonanophenone contains 91.5% y-deuteriums. 5. Valerophenone. Valerophenone (Aldrich Chemical Company) was recrystallized from pentane, passed through neutral activity I alumina and distilled at reduced pressure. C. OTHER MATERIALS. 1. Solvents. Gallon quantities of comercial thiophene free benzene were washed with approximately 200 ml portions of concentrated sulfuric acid until the acid layer no longer showed color. The benzene was washed several times with distilled water, with 10% aqueous NaOH, and with saturated NaCl solution, then dried and fractionally distilled from phosphorous pentoxide at atmospheric pressure. Gallon quantities of commercial (Fischer Scientific Company) p-butyl alcohol were fractionally distilled from sodium at atmospheric pressure. Gallon quantities of commercial pentane (MCB) were treated in the same manner as benzene. 2. Internal Standards. peTetradecane (Columbia Organic Chemicals) was treated in a manner analogous to benzene. pyEicosane (Matheson, Coleman and Bell) was 57 recrystallized from absolute ethanol and dried ip_vacuo. p:0ctadecane was washed with concentrated sulfuric acid and recrystallized from absolute ethanol. 3. Quencher Commercial 2,5-dimethyl-2,4-hexadiene (DMHD) was sublimed once at 5°C and 1 atm pressure. 4. Miscellaneous An exceptionally pure commercial sample (MCB) of acetophenone was used for standardization without purification. Commercial samples of trag§;stilbene (Eastman Organic Chemicals) and gisfstilbene (Aldrich Chemical Company) were used for comparison and for standardization without purification. D. IRRADIATION PROCEDURE. Identical portions (2.8 or 3.0 ml) of solutions containing the ketone and the appropriate additives were placed into constricted, 100 x l3-mm pyrex culture tubes with a 5-ml syringe. The samples were degassed by three freeze-pump-thaw cycles and sealed jp_vggug_at 5 0.003 torr. The set of samples composing a run were irradiated in parallel in a "merry-go-round" apparatus89 with a Hanovia 450-w medium-pressure mercury lamp contained in a water-cooled, quartz immersion well. Corning N-7-83 filter combinations were used to isolate the 366 nm line. A 1 cm path of 0.002 M potassium chromate in 5% aqueous potassium carbonate was used to isolate the 302.5-313 nm lines. The merry-go-round ensures that the same intensity of radiation impinges on each sample during a run. 58 E. PHOTOLYSATE ANALYSIS. All quantitative anflyzes of the photolysates for product formation or ketone disappearance were performed with an Aerograph Model 1200 Hy-Fi III glp chromatograph employing a flame ionization detector, on-column injection, and a 9-ft x 1/8-in column packed with 4% QF-l and 1% carbowax 20 M on 60-80 mesh Chromosorb P. The area ratios of the material being determined to an internal standard were measured by a DISC integrator-Leeds and Northrup Model H recorder combination. The area ratios of acetophenone-tetradecane and trap§;stilbene-eicosane were converted to mole ratios when necessary by calibration with known mixtures. The molar responses of the cyclobutanols were assumed to equal the molar response of the isomeric ketone. The fraction of unreacted ketone was determined by direct comparison of ketone-internal standard area ratios of the irradiated sample and an unirradiated sample. F. GENERAL PROCEDURES OF PHOTOKINETIC STUDIES. 1. Quantum Yields. The quantum yields (6) of acetophenone formation from the ketones were all determined by valerophenone actinometry. A solution of valerophenone in benzene (usually 0.10 M) containing pftetradecane and a ketone solution containing petetradecane and other materials were prepared, using the same fiftetradecane stock solution whenever possible. Samples of identical size were prepared from both solutions. At low conversions a set of ketone samples and a set of valerophenone samples were irradiated in parallel for the same length of time. At high conversions, a succession of several sets of valerophenone samples were 59 irradiated simultaneously with the set of ketone samples, in order to limit the conversion in the valerophenone samples to less than 10%. The quantum yield of cyclobutanol formation from MC was determined from the ratio of cyclobutanols f0rmed to acetophenone formed and the quantum yield of acetophenone formation under those conditions. The quantum yield of cyclobutanol formation from the nonanophenones in the presence of high concentrations of tebutyl alcohol was determined from the ratio of the cyclobutanols formed to the acetophenone formed in trbutyl alcohol and the quantum yield of acetophenone formation. The quantum yield of cyclobutanol formation from the nonanophenones in benzene was determined by direct comparison with a valerophenone actinometer solution. 2. Dependence of Relative Quantum Yield on Quencher Concentrations. A stock solution that contained the ketone and the internal standard(s) and a stock solution that contained the quencher (DMHD) were prepared. Then a series of solutions having identical ketone concentrations but varying DMHD concentrations were prepared by pipetting the appropriate amounts of these stock solutions into 5 or 10 m1 volumetric flasks and diluting to the mark with solvent. Samples of the solutions were prepared and irradiated in parallel for the same length of time. 3. Dependence of Optical Rotation on Conversion. A solution containing MC, the internal standards and additives was prepared and its optical rotation determined. Then 3 m1 samples of the solution were prepared. The samples were irradiated in parallel. Sets of two tubes were removed at intervals. The contents of each set 60 of tubes were combined and the rotation of the resulting solution determined. These solutions were then analyzed for acetophenone and MC. G. ISOLATION OF PHOTOPRODUCTS FROM MC IRRADIATIONS. l. Recovered MC. Preparative glp chromatography of the condensate from 12 ml of 0.1 M MC irradiated to 16% conversion with a Hewlett-Packard Model 776 chromatograph fitted with a 20 ft x 3/8 in column packed with 4% QF-l and 1% carbowax 20 M on 30-60 mesh AW/DMCS Chromosorb P gave 0.050 g of recovered MC. The recovered MC was quantitatively transferred to a 10 ml volumetric flask and filled to the mark with benzene. This solution exhibited a rotation of +0.034° at 589 nm; [GJES = 6.8°(benzene) for recovered MC. 2. Cyclobutanols from MC. The condensate of a high conversion irradiation of 25 ml of ~O.18 M MC in tebutyl alcohol was chromatographed on activity I, neutral alumina using pentane as the eluant. Fractions were examined by analytical glp chromatography, as described above. Acetophenone and residual MC were eluted in the early fractions. The later fractions contained mostly f cyclobutanols and their condensates exhibited typical infrared O-H absorptions. Preparative glp chromatography (see above) of the cyclo- butanol rich fractions gave 0.093 g of the cyclobutanols (99% by glp chromatography). These were transferred quantitatively to a 6 m1 volumetric flask (non-class A) and diluted to mark with benzene. The 61 resulting solution (0.082 M) exhibited a rotation of -0.003 1 0.001n at 589 nm. Therefore, [a];5 (MD cyclobutanols in benzene) = 0.2 t 0.l°. The pmr spectrum (Figure 6) of the isolated cyclobutanols was consistent with a 50/50 mixture of 2-methyl-2-ethy1-l-phenylcyclobutanols, I and II: pmr (CC14) broad singlet (5H) at T 2.75 (C6H5), multiplet (1H) centered at T 7.4 (Hb,Hb.); sharp singlet (1H) at 7.82 (0H), overlapping absorptions (4H) at T 7.8-8.8 [Ha’ Ha', Hc’ Hc', Hd, Hd', EHgCH3(II)] and overlapping absorptions (7H) at T 8.8-9.6 [CH3(I), CHZCH3(I), -CH2CH3(II), CHZCH3(I), CH3(II)] with sing1ets at-t 8.87—[CH3II)] and T 942—1011301117— OH CH3 C6H5“ '"CHZCH3 Ha n Ill Hd Hb Hc I Cyclobutanols from the irradiation (313 nm) of 9 m1 of 0.1 M MC in benzene to >75% conversion were also collected by preparative glp chromatography (see above). The cyclobutanols (0.0112 g) were washed into a 10 ml volumetric and diluted to the mark with benzene. This solution exhibited a rotation of +0.001 t 0.001° at 589 nm: [0135 = +0.9 1 O.9° (benzene). 62 .Io:mum cw oz mo cowpmwumeew sob» umpomF—oo mFocmpsno_uxu mo Ezeuumam can .0 mmond m 14 a _ 3 e .mm.o u ovum; mFoE-Lma-mmcoamme-mmem oxU\u<;a amazmm mo mucmmmea we» see um:?5exm mew; Qamw use a: mpm pm meowu=Fom commmmmc mo cowamwumeew any seem mopemxpouoza been o m m A :95.» eoe.o - mpo.o oemo.o oo.N N_.o - Ne.F eoo.p \aeaNeaeV - - - emao.o - - - PN.P an onzo mm eNF.o - ep_o.o Nwo.o mo.N 1N.o - _e._ NF.F \aeaNeaav - - - Foe.o - - - Nm._ e0 amo.o Naooo.o eee_o.o omeo.o eo.N - No_o.o mm._ mNm.o Axosmnmv - - - mN25 - - - mNp.P om Ne.o e_m_o.o Nemeo.o NF.N - eNo.o Noe.P mo.~ AaeaNeaav - - - eemo.o - - - aN._ be eF-U\osu dab oeea oz obwmno owpmm mmg< . eme m¢m.>-m:ocmgaocmcoc was mampiuv mcmumumuuo ”oceuceumu Auemmno mowumm mme< . N m v->.>-mcocm;aocmcoz use ococmsgocmcoz -mmcoammaieoe< ”companmepmmo poauoee .H> m4m.m z Nmo.o "m cowuapomu .aeaNeam em oN-u z NONoo.o ace ep-u z omNoo.o .oeoeoeaoamb=m_meoeamo-s.m z mm.o we combspomm .. .AON-U "eeepmebmu Am-ov aeoa_mbmumflm nee HAON-UV aeemaomo .: ”useucepmu Amuuv mcmnpmumimceeu .AeF1uv memomueeumpuc “neeucmumg Augmmao mompmm eme< . .aoeoeaeaosmbsmpmeaeamm-».m eoebseesbmmm boseoaa .Hm> momem .N_.N n 0000; mpoe-ema-wmcoammeumme0 u<;0\epuu emeammmzv .mcm~=m0 cm epic z momoo.o 0:0 wcocmgaoempm> z oo_.o "emumeocmuucou mzomem> o» Agomm .E: mFmV cmpmeumeem meowparom oz 00 HA0F-QV mcmumumeumunm ”venucmumu umseoe Au<00v mcocmzaoumum mo 0.0000 000 new ”A0N10V ocemouwm-c “uemvcmpmu Auzv mcocmgaoeamopxgpmei> we cowbmepcmucou mgu .Aav 0000000; _000000 0:00 66‘ mNN.0 000.0 .0m_0.0 0mN0.0 mNN.0 mm0.N 000.0 N00.0 00.0 .1N0.0 . 0 . 0 . 0 . 0 . N . 0 . 0 . 00.0 0000.0 0N_ 0 00 0 _N_0 0 0N0 0 NN 0 00 _ m0 0 0_ 0 00.1 0N00.0 000.0 _0m.0 n00000.0 0N00.0 mNN 0 000.0 N00.0 N00.0 my.” mefl.m 000.0 N00.0 0WN00.0 Nm00.0 00N.0 _N0.0 000.0 0N0.0 0PH_ .mpe 0 me 0 o0N0 0 000.0 _0N.0 m0300.0 0000.0 NNN.0 F00.0 F00.0 F00.0 0N._ 0000.0 00.0 00.0 - 0000.0 - - - - 0N.P .mNP.0 0_-0\o<0a .Ihmpe_ .Hwe_ 0w«mmliiinipzu o<0a eoeoeoewbo< 01-0\o<;0 00-0\o<;0 00n0\oz wnxu 00 2 0.0000 upmw> Sauceaa 0000; 00cm oemumsocmuo< acompabom oz 0 cessam um>bmmno mowumm mme< . 000N000 :0 00000000L0000000021» .000> m0m z NoF.o "emposocwuuPumpme as» 000 mfiomiuv mnemoummnm uuemucepmu 0.0000 Auzv mcocmgqoeaequgumsu> we» .Aav 0000000; 0000000 0:00 000.0 m0_.0 0mN0.0 0m0.0 mm.. mm.0 0N0.0 .000.0 . . 0 . - - - 0 . . 00 0 00 0 N0, 0 N0 _ o0NN 0 .Hwnr_0 .mwep 02 o 2 «00.0 "000000000000 .mcwncmn 00 000000x00 -0.N-_0000000-m.m z m_00.0 000 00-0 2 0000.0 .0010 2 00000.0 .0: 2 000.0 "00000000 020 .m0opmem>000 000000> 0» 00000000 02 00 AUomN .00 0000 00000000000 00 000000 HA0? 00 0000000000» 0 muemucmpmg ~0<000 000000000000 00 0000» w>000000 .mfiom 00 000m0000 0 .0000000mu 0. 0000 Auzv 000000000000000amsn> 00» A00 00000000 0000000 0000 0m0.0 00m.0 F000.0 - - .0._ 00.0 0000.0 PNF.0 000.0 0000.0 - - 0F._ 0_._ .m__.0 - - 0000.0 - - 000.0 NN.0 - 000.0 000.0 _N00.0 NNP 0 00m N 0000.0 NmN._ 002.0 00.0 00.0 0000.0 - - - 0N0._ o000.0 .100000 .0w0_ 00 a.0o000 0P-0\o<00 -0\o 0 0 - 00 a 0000 00 100 0.0000 0000» 0000000 0000000000o0 0000000m 00 00>00000 00000 000< .00000000 0000000 0003 m0m~0mm 00 000000000000 00000:-» .Hx ugm<0 7O .000~000 00 00-0 2 «0000.0 000 000000000000> 2 000.0 "0000500000<0 .000N000 00 0000000 quzncp z 00~.o 000 omuo 2 0000.0 .0puu z N00oo.o .0: z mowmo.o 000000000 0 .00000000 0003 0000000>000 000000> 00 Auomm .50 mpmv 0000000000 000000000 00 0000-00 00000000000 -0 "000000um0 000000 00<00v 000000000000 00 0000» 0>000—00 000 .mflomuuv 00000000nm 0000000000 02 00 00000000000000 000 .000000000 0020 000000000000000000-» 000 00 Adv 00000000 0000000 0000 0 . . 000.0 000.0 00000.0 - - - 0mm.w mefl.m m . . 000.0 000.0 00000.0 - - - 0ww.” Mwmfl.m 0 . . 000.0 000.0 00000.0 - - - 0mm.” ”mm”.m 0 . . 000.0 0000.0 00000.0 000.0 000.0 000.0 0mm.” ”wwfl.w 00.0 00.0 00000.0 - - - 000.0 0000.0 .lnpzp00 .0000 00 0000 00-000000 000-000000 00-000: 00020 00 020 0.0000 00000 0000000 000000000000 000000000000, 0 00>00000 000000 000< .00000x0z 0000000 quzmnm-000~000 00 00000000000000000z-> .HHx 000<0 71 .o .mmm.— u ovum; mpoe-gma-mmcoammgummgm u<;a\vp-u umgammmz .m=m~=mn c_ w_-u z mmmmoo.o nae «F-u z .wcmNcmn cw vpuu z NmmSod new mcocmgaosmpm> z «op.o ”memsocwpu< u mm—oo.o -.P u ompmg mpos-gmq-mmcoammg-mmgm :2\mpuu umgammmzm u .:z ”meowp=_omu a .ummmmmmm mgmz floomm new E: mpmV cowum?umggw co memsocwuum mcocmnqogwpm> a xa umsgoy u_pm_mg mchm mko.o «mm.o Nepoo.o mmmoo.o mNo.P mw~.o 0:2 2 mo~.o mmo.o Aemm.oV P¢_oo.o - - ump.o 0:2 2 _m_.o mko.o mom.o mmpoo.o mmcoo.o “mm._ omfi.o uzz z N__.o Nfio.o _mm.o mwpoo.o F$85 mm¢.F P~_.o m6:: 2 memo.o - flmm.oH - mmmoo.o «mw.F - agmumeo=wpu< may; Esmfim a? 2.2WWE Ewan? mosmmfiaa 5.528 .cowumgpcmucou ocouox mepwcH co «upmw> saucmac puzuoga mo mucmncmamo "ocmNcmm cw mcocmsaocmcoz .HHHx m4m<~ 72 .No.m u ovum; m—osugma-mmcoammg-mmgm u<;a\¢~-um .m=m~cmn cw mpno z Pp_oo.o new ep-u z upmoo.o .22 go Qz "mcowpzpomu .mco~cmn cw ¢F-u z upmoo.o ucm mcocmsqogmpm> z Pop.o "m gmumsocwuu z oop.o u< memsocwuu x3 mm5gom u.>-m:ocm;mocmcoc an umELow mfimpuuv mcmuoumuuouc “ugmucmumu onuv mpocmuznopuau vcm mA¢F-uv mcmumumgpmuuc "ugmccmumu Auwumpmg mchm .. omm.o ¢¢ooo.o N8; - - uxz z mmmo.o - NNN.o mmmoo.o moo._ - - uaz z oom.o - mmm.o N88.0 mqo._ - - no: 2 omp.o - ooN.o mmooo.o _No._ - - uaz z mop.o “No.0 mm~.o Neooo.o ¢oo._ o¢.F “.mm coz z mmho.o N88.0 NmN._ - - um gmgmsocwpu< - mmm.oH mmmoo.o Nom.F - - n< gmpmsocwpu< uxu oqcm u Eapcmac wqumwmmmm mowpmm mwg< cowuzFom .mcowpmgpcmocou mcopmx meuwcH co nape?» saucmao uozuogm yo mucmvcmamo ”acmncom c? ~u->.>-oco:m;aocmcoz .>Hx u4mu\mp-om .mmm.p u ovum; mFoe-gma-mmcoammg-mmgm u<;a\e_-u m .o u .m—nu z Nmmoo.o tam vpuo z mmpoo.o z "mcowuzpomu .ep-u mmpoo.o new mcocmgnogmpm> z cop.o "Lmumsocwpu z: HA¢F-UV mcwomumgpmpuc "ugmuxmpmu cmsgo$ Au.>-mco:m;qocmcoc >a fiAmP-Uv mcmumumuuo-= "ugmucmpmu wmego$ Auxuv wrocmpznoruzu mo uFme m>wumpmg mghm E; - @285 - - m8; 002 2 2:5 E; - . $80.0 .- - $25 uaz 2 $3 Rod - $8.0 - - #25 go: 2 $2.0 - 5.8 - 88: gm; - £32853 98 we; 03 2% 3-22.: 3.20% 55.28 upmw> Eaacmzo m:.o:ou vw>gmmao mowpma mmg< .mcowumgpcmucou ocoumx mepmcH co saucmso Focmpsnopuao mo mocmccmamo ”memNcmm cw Nu->.>-mcoco;aocmcoz .>x m4m

>zzmm C. . G. D. . S. J . R. A Lin and K. J. Laid1er, Can. J. Chem., 38, 973 (1968). . Hammond, Adv. Photochem., Z, 373 (1969). . Rice, Adv. Chem. Phys., 88, 153 (1971). . Turro, “MoTecuTar Photochemistry“, Benjamin, 1967, p. 65. S A J . C. A E J Baird, M01. Photochem., 8, 53 (1970). . Shaikh and K. A. Thaker, J. Indian Chem. Soc.,,gg, 340 (1966). . Kemppainen, Ph.D. Thesis, Michigan State University, 1971. . Wagner and G. Capen, M01. Photochem., 8, 173 (1969). D. Lewis, J. Amer. Chem. Soc., 88, 5602 (1970). . Yates and J. M. Pa1, Chem. Commun., 553 (1970). Zepp, Michigan State University, unpub1ished resu1ts. Wa11ing and A. Padwa, J. Amer. Chem. Soc.,,88, 1597 (1963). Bonner, J. Chem. Thermodynamics, 8, 577 (1970). Hammond, et.a1., J. Amer. Chem. Soc.,,88, 3197 (1964). . Wagner, P. A. Ke1so, A. E. Kemppainen, A. Hang and Graber, M01. Photochem., 8, 81 (1970). . Ca1dwe11 and P. M. Fink, Tetrahedron Lett., 2987 (1969). 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97. 98. 99. 100. 86 D. G. Whitten and W. E. Punch, M01. Photochem., 8, 77 (1970). F. R. Stermitz, D. E. Nicodem, V. P. Mura1idharan and C. M. O'Donne11, M01. Photochem., 8, 87 (1970). L. M. Stephenson, P. R. Cavig1i and J. LiPar1ett, J. Amer. Chem. Soc., 93, 1984 (1971). ——'v1. P. J. Hammrick, Jr. and C. R. Hauser, J. Amer. Chem. Soc., 81, 493 (1959). G. Stork and S. R. Dowd, J. Amer. Chem. Soc., 88, 2178 (1963). J. March, "Advanced Organic Chemistry: Reactions, Mechanisms and Structure", McGraw-Hi11, New York, 1968, p. 655. F. G. Moses, R. S. H. Liu and B. M. Monroe, M01. Photochem., l, 245 (1969). W. D. CTark, A. D. Litt and C. Stee1, J. Amer. Chem. Soc., 81, 5413 (1969). R. F. Borkman and D. R. Kearns, J. Chem. Phys., 33, 945 (1966). J. Sa1tie1, et.a1., J. Amer. Chem. Soc., 88, 410 (1970). P. J. Wagner, M. J. May, A. Haug and D. R. Graber, J. Amer. Chem. §9§,, 88, 5269 (1970). D. F. Evans, J. Chem. Soc., 1351 (1957). R. H. Dyck and D. S. McC1ure, J. Chem. Phys., 88, 2326 (1962). . By1inda and Z. R. Grabowski, Trans. Faraday Soc., 88, 458 (1969). . A. LamoTa and G. S. Hammond, J. Chem. Phys., 88, 2129 (1965). . Itoh and R. S. Mu11iken, J. Phys. Chem.,,a8, 4332 (1969). A A D. F. Evans, J. Chem. Soc., 1735 (1960). M R . E. Ke11ogg and W. T. Simpson, J. Amer. Chem. Soc.,,8z, 4230 (1965). 101. 102. 103. 104. 105. 106. 107. 108. 109. 110. 111. 112. 113. 87 . W. Benson, et.a1., Chem. Rev., 88, 279 (1969). . Sunner, Acta Chem. Scand., ll, 1757 (1960). S S A. Gero, J. Org, Chem., 18, 1960 (1954). S. Marantz and G. T. Armstrong, J. Chem. Eng. Data, 88, 118 (1968); 599 (1968). J. A. Kerr, Chem. Rev., 88, 465 (1966). D. M. Golden and S. W. Benson, Chem. Rev., 88, 125 (1969). S. Furuyama, D. M. Golden and S. W. Benson, Int. J. Chem. 51223:: 8) 83 (1970). W. Tsang, Int. J. Chem. Kinet., 8, 311 (1970). K. D. King, D. M. Golden and S. W. Benson, J. Amer. Chem. Soc., 88, 5541 (1970). S. W. Benson and R. Shaw, 939, Peroxides, l, 105 (1970). A. B. Trenwith, Trans. Faraday Soc., 88, 2805 (1970). F. R. Cruickshank and S. W. Benson, J. Amer. Chem. Soc., 88, 1289 (1969). F. R. Cruickshank and S. W. Benson, J. Amer. Chem. Soc., 88, 2487 (1969). APPENDICES APPENDIX I THE KINETICS 0F PHOTORACEMIZATION The following expressions can be written for the rate of change in the concentrations of the transient species in Scheme I (p. 16) at some time t: 3 * -5!En§5L- = -%:§3—1 ' (kP * kd * kIN)[3K5] d__[____:K;1 _ [KR] -IET——I - (kp + kd + kIN)[3KR] Setting d[3K*]/dt = d[3K*]/dt = 0, 1/T = kp + kd + k1", and solving for the steady state concentrations of the transient species yields 3 [K5] [ K‘k] : Tj—K IT (17) 3 W I: K*] = wl‘r (18) The rate of ketone disappearance and of product formation are stoichio- metrically equal and according to Scheme I can be expressed as +1—di = -—£—]—‘“,‘ = kpt3n<§1 + 8.136;] (19) 88 89' Substituting in the expressions for the concentrations of the transient species (Equations 17 and 18) into Equation 19 and rearranging (note, [K] = [KS] + [KR]) yields d P __E_J_dt = -_ng§ = kPrI (20) It is assumed that the initial concentration of ketone ([K]) is sufficient1y large that >99,9% of the light is absorbed over the duration of the photolysis. Note that no correction will be necessary for the absorption of radiation by acet0phenone (type II fragmentation product) at lower conversions. The similar excitation energies of phenyl ketones and the long triplet lifetime of aceto- 6 secgo) relative to the triplet lifetime of phenyl 9 s phenone (3.5 x 10' ketones with tertiary y-hydrogens (2.2 x 10' ec47) resu1ts in the effective flow (by energy transfer) of a11 excitation to the parent ketone at lower conversions. Thus the rate of absorption of light by the starting ketone (I) will be constant and the quantum yield of product formation (¢+P) is _ d P 1 ¢+p ' ‘851‘7— (2” = k r (22) and from Equations 20 and 22 -d[K] = Idt (23) 90 The quantum yield of racemization (¢RAC) is defined as the ratio of the rate of formation of racemate to the rate of light absorption. The rate of formation of racemate is twice the rate of formation of the minor enantiomer. Taking KS as the enantiomer initially present in excess, it follows that d[K 1 _ 2 R ¢RAC " 1 dt (24) According to Scheme I the rate of formation of the minor enantiomer can be expressed as: d[KR] [KR] 4 4 "EE"" = - "Til—"J + kd[KR] + kINEKS] (25) Substituting the expressions for [K3] and [Kg] (Equations 17 and 18) into Equation 25 yields d[ 1 [ ] [K 1 d§&__ - (de - 1) .1%§}__1 + kINT ‘TE1"{ (26) Substituting the expression for d[KR]/dt (Equation 26) into that for ¢RAC (Equation 24) yields 2 (de - 1) [KR] + 2 kINT [KS] ¢RAC = [K] (27) In the limit of zero conversion [KR] + 0 and [KS] + [K] therefore: lim _ L _ t + 0(¢RAC) ’ ¢RAO ' 2 kINT (28) 91 The enantiometeric or optical purity (0P) of the ketone in the photolysate can be expressed in terms of the enantiomer's concentration as: (OP) = ([KS] - [KR1)/[K1 = ([K] - 2[KR11/[K] (29) Thus the rate of decrease in the optical purity of the ketone in the photolysate is given by d‘OP) = d(1'2[KR]/[K]) = 2 [K] d[KR] - 2 [KR] d[K] (30) dt dt [K]2 dt [K]2 dt Substituting the expressions for d[KR]/dt (Equation 25) and d[K]/dt (Equation 20) into Equation 30 and rearranging yields d 0P g 21 [ (de I kPT ' 1) [KR] + k1MT [Ks] ] (3]) t ‘[R]“' [K] Substituting (de + kpr - 1) = 'kINT (this follows from the definition of 1) into Equation 31 yields d(OP) _ 21 [Ks]‘[KR] dt ' -[K]—kINT [K] 132’ Substituting the equalities expressed by Equations 29, 23, and 28 into Equation 32 and rearranging yields ¢L (0P) ¢+P K 92 Integrating Equation 33 from t = O to t = t yields (0P), 41W; 00, ”To—P1— : "EIT‘"‘[R]— (34’ where i indicates an initial value. If the optical rotation of the photolysate (a) is due to entirely to the ketone, then 0P can be expressed as: 0P = _§[K1"' (35) S is the ratio of the rotation of an optica11y pure sample of the ketone enantiomer initially in excess to its concentration (under the same conditions of measurement as a). Substitution of this expression f0r 0P into Equation 34 yields on rearrangement L . “i ) = (1 + ¢RAc [K31 ln( 0 ¢+P )1n(-rfi]-9 (36) 93 APPENDIX 11 THERMOCHEMISTRY A. EXCITATION ENERGIES. l. Ketones. Borkman and Kearns91 have estimated that the origin of acetone's singlet-triplet transition is at 80.6 t 1 kcal/mole from phosphorescence emission studies in rigid solutions at 77°K. This value was used for the singlet-triplet excitation energy of 2-hexanone. Recently values of 72.0 and 72.9 kcal/mole92990 have been obtained for the singlet-triplet transition of acetophenone from the phosphorescence emission studies in iso-octane and carbon tetrachloride at 20°C. B,y-Diphenylbuterphenone31, y-phenylbutyrophenone81 and other pheny1 alkyl ketones93 all have 0-0 phOSphorescence emission bands near 74.5 kcal/mole in rigid hydrocarbon solutions at low temperature. Therefore a value of 72.5 kcal/mole was adopted for the sing1et-triplet excitation energies of all the pheny1 ketones considered. 2. Olefins and Enols. It is well established that the vertical singlet-triplet excitation energy of trans-stilbene is ~50 kcal/mole9”’95:95. Evans91+ has obtained a value of 61.8 kcal/mol for the vertical singlet-triplet excitation energy of styrene. Lamola and Hammond97 have obtained a value of 94 59C8 kcal/mole for the vertical singlet-triplet excitation energy of .§I§g§;B-methylstyrene. Thus a value of 61 kcal/mol was adopted for the vertical singlet-triplet excitation of styrene. Evans98 obtained a vertical singlet-triplet excitation energy of S 82 kcal/mole for ethylene. The onset of the sing1et-triplet absorption spectra of propene and 2-methylpropene99 can be estimated to be at 80.5 kcal/mole. Thus a value of 81 kcal/mole was adapted for the vertical singlet— triplet excitation energy of alkenes. There are no values for the singlet-triplet excitation energies of acetophenone or acetone enol or their ethers available in the literature. However, phenol's phosphorescence O-O band lies ~2.5 kcal below the that of benzene. The 0-0 band of 1-methoxybutadiene is singlet-triplet absorption spectrum lies m3.2 kcal below that of butadieneloo. Examination of singlet-singlet absorption spectra of enol ethers, reveals that substitution of a methoxy group at the end or middle of an unsaturated system results in a similar shift of the maximum of the singlet-singlet absorption spectra. On this basis, vertical singlet-triplet excitation energies were assigned to the enols that are 3 kcal/mole lower than the value for the corresponding olefin. B. REACTION ENTHALPIES. l. Retro-ene Reaction. a. From heats of formation. The reaction enthalpies of the retro-ene reaction (AHRE) of the selected ketones were obtained from the difference between the product's heats of formation and the ketone's heat of formation. 95 AHRE = z AHf(Products) - 0Hf(Ketone) (37) The heats of formation employed and the resulting estimates of AHRE are given in Table XXIII. The heats of f0rmation which had not been experimentally determined were calculated, with the exception of the heat of f0rmation of acetone enol, by the group additivity method using the group values of Benson gt,§1,1°1. An estimate of the that of formation of acetone enol can be obtained from the sum of the heat of formation of acetone and its heat of enolization [AHE (acetone)]. AHE (acetone) has been estimated as 10.0 kcal/mole from a comparison of the heats of hydrolysis of isopropenyl acetate and m-methylphenyl acetateloz. Taking AHE(acetone) as 10 kcal/mole yields -41.7 kcal/mole f0r AH;(acetone enol). Using this value the enthalpy of formation contribution for the group [O-(C)(H)] was estimated to be -48.5 kcal/mole. This group contribution, in conjunction with Benson's 23:21: group valueleI, was used to estimate AH;(acetophenone enol) as -11 kcal/mole This value may be too positive. It requires that AHE(C6H5COCH3) = +11 kcal compared AHE(CH3COCH3) = +10 kcal in contrast to the larger enol content of neat acetophenone relative to neat acetone103. 96 .40_ mucmgmmmmm .AFOF mocmgmmmmv umumFauerI .Fop mucmcmmwmm .uxmu mmmu .mpmu mcu sage umcwmuao o + m + x com omcmsu xangpcmo .spo I use comN um acaoaeou mzommmm mo cowpmsgom we pom: a .mFOE\PMU¥ Cw mmzpm> FF I.II+ ce.m¢- 6o.4 - Io._F- III III I IIIQ meechIeIm_8> PIIpes-» 4.8+ Imam- 694+ I9:. III I I IIII 8655222, I.IN+ 88.88- mm.e + Io.Ie- mIo I I III ecocmxnI-I mm Apoe\_muxv APoE\~muIV AFoE\quxv I c I d < I I IAIV I Ifiov I8 IAIV I4 mI NI _I I Icaegeeo o I I I _I ,7, (x P HU\\I + / m I I I I N m I mI cOmeEc—om Lb mumm: =51; cowuommm $5.10.»qu $0 mmII_mIo=I IeIoeIeI .HHHxx IIIII 97 b. From bond energies. Estimates of the reaction enthalpy for the retro-ene reactions (AHRE) were also obtained from the difference between the bond energies (D) of the bonds broken and the bond energies of the bonds f0rmed. AHRE = 2[D(bond broken)] - z[D(bond formed)] (38) It was assumed that the bond energy of a bond depends only on the bonds immediate molecular environment. Thus the bond energies used in these calculations are bond energies of a representative bond which possesses the same immediate molecular environment. The bond energies used were computed from the latest data in the literature. The single bond dissociation energy, D; (A-B), is defined as the reaction enthalpy for the formation of A-(g) and B-(g) from the molecule AB(g) and is related to their standard heats formation(AHf) of these species by Equation 39. 02(4-31 H; [A (9)] + H;[B (9)] - H; [43(9)] (39) Bond energies that have not been measured can be obtained from heats of formation of the parent compounds and the measured bond energies by the application of Hess' law. Table XXIV contains the carbon-hydrogen and oxygen-hydrogen bond energies that were employed as representative bond energies or in the calculation of representative bond energies. A good experimental value was not available for D:[C6H5CH(CH3)——H]. This value was estimated as follows: Comparison of the well established valueslos»106 for DO[06HSCH2-—H] and DO[CH3CH2-—H] indicates that 98 substitution of a phenyl group in place of a methyl group at the developing radical center lowers the C-H bond energy by 13 kcal, i.e. by the benzylic resonance energy. Although there is considerable disagreement over the absolute values of the analogous allylic resonance energies (ARE), the various values all indicate that the ARE value increases 1 to 2 kcal for each substitution of a methyl group for a hydrogen at the initial radical centerlll. Accordingly, the benzylic resonance energy was taken to be 14 kcal and D:[C6H5CH(CH3)—H] = 81 kcal/mole. Similarly, D;[C6H50(CH3)2—H] was estimated to be 78 kcal/mole. There was no experimental value for D:[HOCH(C6H5)——H]. A survey of the literature (see Table XXIV) indicates that substitution of an hydroxy or alkoxy group for an alkyl group at the radical center formed lowers the C-H bond strength by 3 kcal/mole. Accordingly, a value of 78 kcal/mole was adopted for D;[HOCH(C6H5)-H] taking D;[C6H5CH(CH3)——H] = 81 kcal/mole. There is, of course, no direct experimental value for D;[CH2=C(CH3)0——H]. The strength of this bond will depend on the resonance energy of the radical formed (note that it is the "electronc tautomer" of the acetonyl radical). However there does not appear to be any resonance energy in this system109. Therefore, DO[CH2=C(CH3)0I—H] = 104 kcal/mole. Note, that the errors arising in DO[CH2=C(CH3)0-H] and in D;[CH3COCH2-—H] from the neglect of any resonance energy will be equal and cancel in the computation of AHRE. 99 TABLE XXIV. Carbon-Hydrogen and Carbon-Carbon Bond Energiesa. A - 8 0° (A-B) Source H; (A8)b Hf(A-)c ll-H 104.2 105 0 +52.1 CH3CH2-H 98 106 -20 .2 +25.7 (CH3)20H-H 95 106 -24.8 +18.1 gg-CSHg-H 95 107 -18.5 +24 4 (CH3)3C-H 93 106,108 -32.2 + 8.7 CH3COCH2-H 98 109 -51.7 - 5.8 CGHSCHz-H 85 106 +12.0 +44.9 C6H5CH(CH3)-H 81 text + 7.1 +36.0 C6H5C(CH3)2-H 78 text + 0.9 +26.8 HOCHz-H 96 106 -48 0 - 4.1 CH3OCH2-H 93 106 -44 0 - 3.1 [:75_+' 92 106 -44.0 - 4.1 HOC(CH3)2-H 90 106 -65.2 -27.3 H0CH(C6H5)-H 78 text -22.5 + 3.4 (0H3)20H0-H 104 110 -65.2 -13.3 CGHSCHZO-H 104 110 -22.5 +29.4 CH2=C(CH3)0-H 104 text -- -- CH3COCH2-CHZCH3 82 above -6l.9 -- CH3COCH2-CH(CH3)06H5 66 above -35.6 -- 06H5CH2CH(C6H5)-H 81 above +32.4 +6l.3 aA11 values in kcal/mole. bReference lOl. cFrom the data according to Equation 39. 100 The pj_bond energy (Du) of an unsaturated compound (RZC=X, where X = CR2, 0 and R = H, alkyl, phenyl, hydroxy, or alkoxy) is defined in terms of single bond energies according to Equation 40. D"(R C=X) DO(R CHX——H) - DO(R2CX——H) (40) DO(R2CX——H) is defined as the reaction enthalpy for the fOrmation of cnz=x and H- by RZCXH and is related to the standard heats of formation of these species by Equation 41. The heats of formation of the DO(R2CX——H) = AHf(R2C=X) + AHf(H-) - AHf(CR2XH) (41) requisite free radicals have been computed from the corresponding D;(A-B) and are listed in Table XXIV. The D;(RZCX——H) are listed in Table XXV. The D:(RZC=X) of representative compounds are presented in Table XXVI. The pi bond energy of an enol, H0CR=CH2, was taken to equal D:[CH3CR=CH2] where R = CH3 or C6H5' This is reasonable since Cruickshank and Benson112 have obtained D:( C?) = 59 kcal/mole and D:(CH300H=CH) = 59 kcal/mole. It can be seen from Table xxvr that these values are equal to those for D:( Q) and D:(CH3CH=CH2). There is insufficient data available to calculate D:[C6H5C(CH3)=O). However, D:[06HSCH=O) could be estimated to be 61 kcal/mole and this value was used. The values of AHRE obtained are listed in Table XXVII. These are close to the values obtained from heats of formation (Table XXIII). The values obtained from heats of formation are the ones listed in Table II. 101 TABLE XXV. Carbon-Hydrogen Bond Energies in Radicals. ' b o _ c o . d o ' e CRZX-H AHf(CR2-X) Hf[ CRZX-H] DO(CR2X-H) CH3CHCH2-H + 4.9 +18.1 39 (CH3)2CCH2-H - 4.0 + 8.7 39 O [:>>——41 + 8.2 +24.4 36 C6HSCHCH2-H +35.2 +36.0 51 C6H50(CH3)CH2-H +27.0 +26.8 52 C6HSCHCH(C6H5)-H +53.4 +6l.3 44 (CH3)ZCO-H -51.7 -27.3 28 C6H5CH0-H - 6.0 + 3.4 43 aAll values are in kcal/mole. bx = CR2, 0 and R = H, alkyl, pheny1, hydroxy and/or alkoxy. cReference lOl. dFrom Table XXIV. eComputed according to Equation 41 taking AHf[H-(g)] = 52.1 kcal/mole. 102 TABLE xxv1. Pi Bond Energiesa. _ b o c o ' d o _ e CRz-X DO(HCR2X——H) DO(CR2X-—H) D"(CR2-X) CH20H=CH2 98 39 59 (CH3)2C=CH2 98 39 59 (:1? 95 36 59 06H50H=0H2 98 51 47 C6H50(CHé)=CH2 98 52 46 06H5CH=CH06H5 81 44 37 (CH3)ZC=O 104 28 76 C6H5CH=O 104 43 61 aAll values in kcal/mole. bX = CR2, 0 and R = H, alkyl, phenyl, hydroxy and/or alkoxy. CFrom Table xx1v. dFrom Table xxv, eComputed from data according to Equation 40. .mm cowpmzam o“ mcwueooom name on» seem umuaasouu .>Hxx m_nmh soguu . Z? 3%» soc“: .m_os\_ooI cw mos—m> P—xx m4mHxx opnMF Eogm a .mFoE\Fmox cw mozpm> ~P.m mm- m.~n me Pm mIou I I mImu mcocmcaoeausn -chocai> MI- m.II me mm mIo mIo I mIIo oeeeoIIoeo_8> -panoe-» mm- m.NI me mm MID I I mImo oeoeoIIogopms e_- I.om II mm III I I mIo oeeeaon-~ II em a e < I IAIIMV we oAIIIOUIvao IAIIIIMINIvaa mI NI _I I I o\~_ _. m NI . I I II I .aeoaaneoI _aoaoaewm-e._ co moIoI .HHH>xx m4m