Oxathiirans as intermediates in the photolysis of sulphines

摘要

1404 J.C.S. Perkin IOxathiirans as Intermediates in the Photolysis of SulphinesBy Lars Carlsen, Niels Harrit, and Arne Holm,’ Department of General and Organic Chemistry, University ofIrradiation of thiobenzophenone S-oxide (8) in the region corresponding to its long-wavelength absorption (Amax.329 nm), in various solvents a t room temperature, gave benzophenone in quantitative yield. At 85 K the photolytictransformation was monitored by electronic absorption spectroscopy. In EPA glass or PVC film a monomericcompound (390) with a long-wavelength absorption at 390 nm was formed. Heating gave rise to benzo-phenone in quantitative yield. On irradiation of (390) in its long-wavelength absorption region, a monomericblue intermediate (550). with a long-wavelength absorption a t 550 nrn, is formed.Heating the glass containing(550) led to benzophenone and S- and O-phenyl thiobenzoate in 85, 14, and 1% yields, respectively (g.1.c.).Compound (390) i s identified as 3.3-diphenyloxathiiran. Structures for the intermediate (550) are suggested.Copenhagen, The H. C. 0rsted Institute, Universitetsparken 5, DK-2100 Copenhagen 0, DenmarkOXATHIIRANS are a hitherto unknown class of com-pounds, whose intervention has been proposed in thethermal conversion of thiophosgene S-oxide into chloro-thioperoxoformyl chloride and of thiopropenal S-oxideinto propionaldehyde.2 A theoretical study of theelectrocyclic formation of the oxathiiran ring has beenundertaken by S n ~ d e r . ~ Recently oxathiiran S-oxidehas been suggested as the primary product from thereaction of methylene and sulphur dioxide.*Several authors have suggested oxathiirans as inter-mediates in the photochemical conversion of sulphinesJ.SilhAnek and M. Zbirovsky, Chem. Comm., 1969, 878.W. F. Wilkens, Ph.D. Thesis, Cornell University, Ithaca,New York, 1961; W. F. Wilkens, Cornell Agricultural Experi-ment Station, Memoir 385, Ithaca, New York, January 1964.J. P. Snyder, J . Amer. Chem. Soc., 1974, 96, 6005.H. Hiraoka, J.C.S. Chem. Comm., 1974, 1014.(thione S-oxides) into the corresponding carbonyl com-pounds [equation (i)].5 Schlessinger and Schultz 5g(1) (21 (3)(a) J. F. King and T. Durst, J . Amer. Chem. SOC., 1963, 85,2676; (b) J. Strating, L. Thijs, and B. Zwanenburg, Rec. Trav.chinz., 1964,83,631; ( c ) B.Zwanenburg, L. Thijs, and J. Strating,Tetrahedron Letters, 1967, 3453; ( d ) A. M. Hamid and S. Trippett,J . Chem. SOC. ( C ) , 1968, 1612; (e) A. G. Schultz and R. H.Schlessinger, Chem. Comm., 1969, 1483; (f) A. G. Schultz, C . D.DeBoer, and R. H. Schlessinger, J . Amer. Chem. SOL, 1968, 90,5314; (g) R. H. Schlessinger and A. G. Schultz, TetrahedronLetters, 1969, 4513; ( h ) A. G. Schultz and R. H. Schlessinger,Chem. Comm., 1970, 1051; (2) B. Zwanenburg and J. Strating,Quart. Reports Sulphur Chem., 1970, 5A, 791976 1405determined the quantum yields for photolysis of thesulphine (4) in benzene and in chloroform as a functionof concentration (10-1-10-3~). The products are thesyn- and anti-isomers (4) and (5), the thione (6), and theketone (7).At low concentrations, where the thione isnot formed, the results are in agreement with an uni-molecular process supposedly involving an oxathiiran.At high concentrations a bimolecular process seems tocompete, giving rise to an intermediate which affords thethione.C/ \\s+I Ph0 -( 4 ) pc 0-/ \yPh(5) pPh As Ph A.(6) (7)RESULTS AND DISCUSSIONWe report here the photochemical formation of twothermally unstable monomeric intermediates from thio-benzophenone S-oxide (8) at 85 K. In room temperaturephotolysis benzophenone (9) is formed quantitatively inall solvents investigated ( 10-3-104~ ; aerated oroxygen-free solutions) [equation (ii)] . Thiobenzophenone*- " A'3!233 K)Ph,C = 5' rh 330 nml+ Ph,C=O t s (ii)(loo'/: }was not detected.The conversion was investigated bymeans of low temperature photolysis in EPA glass * at85 K combined with electronic absorption spectroscopy.The sulphine (8) exhibits a long wavelength absorptionwith maximum at 329 nm (E ca. 1.2 x lo4 1 mol-l cm-1),6the shape of which does not change on cooling. Irradia-tion at wavelengths within this band produced a newabsorption with maximum at 390 nm (Figure 1) [thecorresponding intermediate is referred to as (390) in thefollowing]. Owing to the low intensity of this band arather high concentration of sulphine was required forits observation. This resulted in total extinction in theshort wavelength region, rendering it impossible tomonitor the absorptions of the starting material in theinitial stages.On the assumption of quantitativetransformation of the sulphine (€9, the extinction co-efficient of (390) is estimated as 200 1 mol-l cm-1.* Diethyl ether-isopentane-alcohol (5 : 5 : 2 ) .B. Zwanenburg, L. Thijs, and J. Strating, Rec. Trav. chim.,1967, 86, 577.Compound (390) is stable at 85 K but decomposeseither on heating to the m.p. of the glass (around 140 K)or on irradiation at 390 nm. In the first case benzo-phenone is formed in lOOyo yield [equation (iii)]. In thelatter, benzophenone is formed together with a blue-1 I 1 L 2 -,-I300 400 500 600 A /nmFIGURE 1 U.v.-visible spectra recorded during stepwise photo-lysis (A ca. 330 nm) of the sulphine (8) in EPA glass a t 85 K:(a) initial curve; (b) partial transformation; (c) total trans-formation into (390)intermediate exhibiting a broad absorption with maxi-mum at 550 nm [referred to as (550) in the following].(8 1Intermediate (550) likewise showed photoreactivity,and decomposed on irradiation at 550 nm.This de-composition was not accompanied by appearance of newabsorptions in the region open to observation (A 350nm). On heating, (550) decomposed at ca. 100-110 Kbut no new absorptions were observed. In preparativeexperiments ( l O A 3 ~ ) in EPA glass at 77 K, (550) washeated to room temperature. G.1.c. showed that 85%benzophenone had been formed together with S-phenylthiobenzoate (14%) and 0-phenyl thiobenzoate (1 yo).It has not yet been possible by means of absorption oremission spectroscopy to determine whether the estersare produced in a single photolytic step from (390) or viathermal degradation of (550).However, we find thesame amount of esters whether or not (550) is irradiatedat 550 nm before heating. It seems unlikely that (550)should give rise to the same two esters in the sameyields and relative amounts in both thermal and photo-chemical reactions. This suggests that the esters areformed by a light-induced reaction of (390) [equation(iv)l*( C d . Iho/o) +Ph- C -0PhSII( c a . 1 %1Formation of benzophenone at 85 K during the variousstages of photolysis could not be monitored by meansof electronic absorption spectroscopy owing to tota1406 J.C.S. Perkin Iextinction in the pertinent region.This was possibleby means of phosphorescence spectroscopy a t 77 K inmore dilute solutions ( 10-4~). A linear relation betweenbenzophenone concentration and phosphorescence inten-sity was observed in the range studied. Neither S-nor O-phenyl thiobenzoate showed interfering phos-phorescence. Photolysis of the sulphine (8) in EPA glassat 85 K with 330 nm light was accompanied by forma-tion of ca. 10% of benzophenone, according to thephosphorescence intensity. Subsequent irradiation inthe region of maximum absorption of (390) resulted inthe formation of ca. 50% of benzophenone. A furtherincrease of ca. 10% of benzophenone was observed aftersubsequent irradiation at 550 nm, where only (550)absorbs. These figures correspond to an overall yieldof ca.Toy0, which is to be compared with the 85% inthe preparative experiment. The discrepancy is be-lieved to be due to incomplete transformation in theindividual steps of photolysis in the phosphorescenceexperiment. From the above figures it follows that (550)is converted into benzophenone in high yield on irradi-ation.In order to establish firmly that (390) and (550) at85 K are products of unimolecular reactions and not theresult of molecular migrations, the matrix material waschanged from EPA to the extremely rigid polyvinylchloride (PVC). This did not alter the course of photo-lysis at 85 K (Figures 1 and 2), although the thermal/\\t l . . i _ _ _ i _ L _ _ I I 1 I300 400 500 600 A InrnFIGURE 2 U.v.-visible spectra recorded during stepwise photo-lysis (A ca.390 nm) of (390) in EPA-glass a t 85 K: (a) initialcurve ; (b) partial transformation ; (c) optimal transformationinto (550)stability of (550) seems to be greater in PVC than inEPA. As already mentioned, the electronic absorptionspectrum of the sulphine (8) in EPA was unaltered duringcooling to 85 K, indicating lack of aggregation. Theseobservations in conjunction with the simplicity of theproducts obtained on heating the intermediates, restrictsthe reactions to unimolecular processes.Of the two separate photochemical processes observedat 85 K only the first can play a role in the room tempera-ture photolysis of the sulphine (8). Thus when the* Since this work was completed a cyclic five-membered mixedaliphatic-aromatic sulphenate has been reported.8 The electronicabsorption spectrum features a long wavelength transition a t 383nm (E 621, but conjugation with the aromatic ring is most likelyreflected in this value.Direct comparison with the present ali-phatic system is therefore hardly possible.sulphine is photolysed at room temperature or photolysedat 85 K and then heated, benzophenone is formed in100~o yield, whereas photolysis of (390) at 85 K andthen heating gives benzophenone as well as thiobenzoatesas already mentioned. We suggest on the basis of thethermal properties of the intermediate (390) and on thebasis of its absorption spectrum, discussed below, that(390) is 3,3-diphenyloxathiiran.No other cyclic sulphenates, to which class of com-pounds the oxathiirans belong, are reported.* How-ever, relevant comparison with open-chain sulphenatesand with cyclic and non-cyclic disulphides can be made.The longest wavelength transitions of compounds ofthese types are shown in the Scheme.A significant redBut S*OE t Bu'O * S . CC13 EtS-SEt277 nm ( E 390)' 266 nm ( c 71)' 254 nm ( e 398)'n S-S S-S i) s- s330 nmic 147)' 286 nm ( t 295)' 259 n r n ( ~ c a . 5 0 0 ) ~ @SCHEMEa J. A. Barltrop, P. M. Hayes, and M. Calvin, J . Amer. Chem.Soc., 1954, 76,4348. R. S. Irwin and N. Kharasch, J . Amer.Chem. Soc., 1960,82, 2502. C. W. N. Cumper, 3. F. Read,and A. I. Vogel, J . Chem. Soc. ( A ) , 1966,239. V. Ramakrish-nan, S . D. Thompson, and S. P.McGlynn, Photochem. Photobiol.,1965, 4, 907.shift with decreasing ring size is evident. This effectcan be correlated with the twisting of the CSSC torsionangle.s If a similar picture prevails for cyclic sulphen-ates, the absorption of (390) [A,,, 390 nm (E ca. 2001 mol-l cm-l)] is well within the region expected for theoxathiiran ring. CNDO-S calculations performed inthis laboratory lo predict that the electronic absorptionspectrum of unsubstituted oxathiiran should feature aweak maximum at 396 nm. Furthermore, the bisectedand eclipsed conformations of 3-phenyloxathiiran arepredicted to show absorption maxima at 399 and 394 nm,respectively.For the intermediate (550) some structures may beconsidered. Diphenylmethylene, arising by loss of SOfrom (390), could have been formed but is reported toabsorb at 301 and 465 nm; l1 and SO is not coloured.12Thiobenzophenone absorbs at 595 nm, thus ruling outthis substance.13 Sulphur is produced in the photolysisof the sulphine (S), and we considered whether one of itsallotropes might be responsible for the absorption a tC.A. Parker and C. G. Hatchard, Analyst, 1962, 87, 664.G. W. Astrologes and J. C. Martin, J . Amer. Chem. Sac.,D. B. Boyd, J . Anzev. Chew. Sac., 1972, 94, 8799.1975, 97, 6909.lo J. P. Snyder, in preparation.l1 A. M. Trozzolo and W. A. Gibbons, J . Anzer. Chem. Soc.,l2 Gmehn, ' Handbuch der anorg. Chemie,' 9, Teil B, Lfg. 1,l3 G. Oster, L. Citarel, and M. Goodman, J . Amer. Chem. Soc.,1967, 89, !39.167, 1953.1962, 84, 7031976 1407550 nm.S, Is not stable in an MPH glass * a t 77 K,14but forms S, which exhibits a broad band at 530 nm.15The effect of solvent shift on this transition is notknown, and we therefore irradiated S,C1, in EPA a t85 K. This produced an absorption a t 525 nm assignedto S,. Further irradiation in this wavelength regionproduced broad bands around 625 nm, due to higherallotropes of ~u1phur.l~ In contrast, irradiation of, (550)eradicated the blue colour completely.It is noteworthy that the three-membered oxiransand thiirans on irradiation form coloured intermediates.In the case of 1,2-diphenyloxiran a band is producedwith maximum at 503 nm; for tetraphenyloxiran thereare two bands at 605 nm, and for tetraphenylthiiranthere is a broad structureless band with maximum a t650 nm.Accumulated evidence indicates that the inter-mediates are diradicals resulting from C-0 and C-Sc1eavage.l6 Only weak e.s.r. signals are 0btained.l’ Inthe present case we have not been able to detect freeradicals in an irradiated sample at 77 K when either(390) or (550) was present. This result cannot, however,be used to exclude structures with diradical character,since anisotropic spin-spin coupling forces may producestrong relaxation effects causing the e.s.r. signals fromasymmetric 1,3-diradicals to broaden and thus escapedetection.18 Although the diradical hypothesis reg.(lo)] appears best to account for the properties of (550)we cannot unambiguously rule out a ketone sulphideintermediate (1 1).Ph S-P h X O .S-4 / Ph2C = 0(10) 111)EXPERIMENTALLOZJ Temperature Spectroscofiy.-The low temperature U.V.cell used has been described elsewhere.*O U.V.spectra wererecorded with a Cary 14 instrument for samples in EPAglass at 85 K, and irradiations were performed outside thespectrograph with a Bausch and Lomb SP-200 mercurypoint source equipped with monochromator (typical band-width 20 nni). Emission spectra were recorded with aPerkin-Elmer RIPF-3 fluorescence spectrophotometer,equipped with phosphorescence accessory, for samples inEPA glass at 77 K. Irradiation was performed in thespectrograph by using the xenon arc of the instrument(bandwidths used for photolysis 40 nm : for recording 3 nm) .The position of the probe was not changed during theexperiment.* Isopentane-methylcyclohexane (1 : 2).l4 A. Morelle, P1i.D.Thesis, University of Washington,Seattle, 1971; B. Meyer, M. Gouterman, D. Jensen, T. V.Oommen, K. Spitzer, and T. Stroyer-Hansen, Amer. Chem. SOC.Advances in Chemistry Series, No. 110, ‘ Sulfur Research Trends,’1972.l5 €3. Meyer, T. Stroyer-Hansen, and T. V. Oommen, J . MoZ.Spectroscopy, 1972, 42, 335.Electron Spin Resonance.-Spectra were obtained with aJEOL JES-ME-1X instrument, by use of a liquid nitrogenDewar flask. g Values were standardized against Mn2+ inMnO .Diaryl Thioketone S-Oxides.-These were prepared fromthe corresponding thioketones by oxidation with m-chloro-perbenzoic acid.6Iwadz’ation of Thiobenzophenone S-Oxide (8) in VariousSoZvents at Room Temperature.-(a) The sulphine (8) inabsolute ethanol or methylene chloride ( M) was irradi-ated (A ca.330 nm) until conversion was complete. Quanti-tative formation of benzophenone was observed in bothcases (u.v. spectroscopy).(b) The sulphine (8) (5 mg) and tetraphenylethylene(TPE) (20 mg) were dissolved in methylene chloride (4 ml)and irradiated until transformation was complete (ca. 3 h).Only benzophenone and TPE were then present (t.1.c.).Irradiation of the Sulplziize (8) at 77 K.--No precautionswere taken t o exclude oxygen from the solutions used inlow temperature experiments.(a) A solution of the sulphine (8) in EPX glass (lop3 M)at 77 K was irradiated ( A 330 nm) until transformationinto (390) was complete (ca.1 h). After heating t o roomtemperature, u .v. analysis revealed the formation of benzo-phenone in 99% yield.(b) A solution of the sulphine (8) in EPA4 glass (10-3 w)[containing octadecanol ( M) as internal standard forg.l.c.1 was irradiated ( A 330 nm) until partial transformationof (8) (ca. 90”/0) into (390) had occurred (ca. 30 min). Oncontinued irradiation a t 390 nni and a t 77 K, (390) wasconverted into (550) (ca. 60 min). After heating to roomtemperature the mixture was submitted to g.1.c. analysis[Pye-Unicam 104 chromatograph, with dual flame ionizationdetector, connected to a Varian Aerograph 477 electronicintegrator: 2 m x $ in column of 3% OV 1 on GaschromQ(100-120 mesh); nitrogen as a carrier gas]. From thegas chromatogram the unconverted amount of sulphinecould be calculated (calibration curves) and on this basisthe amounts of benzophenone and S- and O-phenyl thio-benzoate (calibration curves) formed were calculated to be85, 14, and l%, respectively.(c) A solution of the sulphine (8) in EPA glass ( l o - 3 ~ )(containing octadecanol as internal standard) was irradiatedat 330 nm (30 min), then at 390 nm (60 min), and finally at550 nin (5 min) until the blue colour was removed. Afterheating t o room temperature the mixture was analysed byg.1.c. The same products were found in the same amountsas described in (b).We thank the Danish Satural Science Research Councilfor financial support.[5/1706 Received, 4th September, 19761R. S. Becker, J. Kolc, R. 0. Bost, H. Kietrich, P. Petrellis,and G. Griffin, J . Amev. Chem. SOC., 1968,90, 3292; R. S. Becker,R. 0. Bost, J. Kolc, N. R. Bertoniere, R. L. Smith, and G. W.Griffin, ibid., 1970, 92, 1302.l7 A.M. Trozzolo, W. A. Yager, G. W. Griffin, H. Kristinnsson,and I. Sarkar, .J. Amer. Chem. Soc., 1967, 89, 3357.l8 A. Carrington and A. D. McLachlan, ‘ Introduction toMagnetic Resonance,’ Harper, New York, 1967.l9 A. Holm, N. Harrit, and N. Toubro, J . Amev. Chern. Soc.,1975,97, 6179