Substituent effects on the photocleavage of benzyl-sulfur bonds. Observation of the "Meta effect"

Article abstract of DOI:10.1021/jo9606923

Benzyl phenyl sulfide has been used to investigate the photocleavage mechanism for benzyl-sulfur bonds. Four experiments have shown that the reaction goes through a radical intermediate. First, the photoproducts observed can all be justified by radical mechanisms. Second, the radical intermediate was trapped with a five hexenyl tether. Third, UV analysis of analogs for the 4-NO₂ derivative indicate no exciplex or electron transfer pathway. Fourth, no strong correlation is observed between a values and the quatum yields for loss of substituted benzyl phenyl sulide. The effect of oxygen on quantum yields is best observed after samples are thoroughly outgassed with consecutive freeze-pump-thaw cycles. It is shown that oxygen diminishes the substituent effect. Upon photolysis of the outgassed samples, the meta-substituted derivatives showed more significant variances than the para derivatives. The meta derivatives are most efficiently cleaved in the following order: 3-CN > 3-NO₂ > 3-CF₃ > 3-CH₃ > 3-OCH₃. These findings are justified by an increase in electron density of the radical ipso to the forming benzyl radical for the 3-OCH₃ derivative and a decrease in the electron density of the radical ipso to the forming benzyl radical for the 3-CN derivative.

Full text of DOI:10.1021/jo9606923

7040
J . Org. Chem. 1996, 61, 7040-7044
Su bstitu en t Effects on th e P h otoclea va ge of Ben zyl-Su lfu r Bon d s.
Obser va tion of th e “Meta Effect”
Steven A. Fleming* and Anton W. J ensen
Department of Chemistry and Biochemistry, Brigham Young University, Provo, Utah 84602
Received April 16, 1996X
Benzyl phenyl sulfide has been used to investigate the photocleavage mechanism for benzyl-sulfur
bonds. Four experiments have shown that the reaction goes through a radical intermediate. First,
the photoproducts observed can all be justified by radical mechanisms. Second, the radical
intermediate was trapped with a five hexenyl tether. Third, UV analysis of analogs for the 4-NO2
derivative indicate no exciplex or electron transfer pathway. Fourth, no strong correlation is
observed between σ values and the quatum yields for loss of substituted benzyl phenyl sulide. The
effect of oxygen on quantum yields is best observed after samples are thoroughly outgassed with
consecutive freeze-pump-thaw cycles. It is shown that oxygen diminishes the substituent effect.
Upon photolysis of the outgassed samples, the meta-substituted derivatives showed more significant
variances than the para derivatives. The meta derivatives are most efficiently cleaved in the
following order: 3-CN > 3-NO2 > 3-CF3 > 3-CH3 > 3-OCH3. These findings are justified by an
increase in electron density of the radical ipso to the forming benzyl radical for the 3-OCH3 derivative
and a decrease in the electron density of the radical ipso to the forming benzyl radical for the 3-CN
derivative.
Ba ck gr ou n d
Excited Sta te Ben zyl P h otoclea va ge. We previ-
ously reported that the photocleavage of benzyl-sulfur
bonds proceeds via a radical mechanism.1 As part of that
study we irradiated benzyl phenyl sulfide (BPS, 1a ) and
several of its derivatives (see Figure 1). Over the last
several years benzyl-heteroatom photocleavage has been
a topic of increasing interest to organic photochemists.
The most recent review related to this topic that we are
aware of was published by Cristol in 1983.2 In general,
two types of photoproducts are formed in these reac-
tions: photosolvolysis products (2) and radical products
such as bibenzyl (3) (see Scheme 1). Typically radical
products are favored, but in certain conditions photosol-
volysis products are formed exclusively. It is not uncom-
mon to have mixtures of radical and photosolvolysis
products. Factors which affect the type of products
formed are solvent,3 photofugacity,4 excited state multi-
plicity,5 counterions (for salts),6 and substituent effects.
The different heteratom leaving group moieties that have
been studied include OAc,7 O2CR,7ij,8 OH,7k,9 O+H2,9b,10
OR,7k,9c,11 O2P(OR)2,7h,12 OSO2R,13 N+(R)3,7h,14 S+(R)2,7h,15
F igu r e 1.
S(O2)R,7h,16 SOR,17 SR,1,9c,18 P+R3,14e,19 As+R3,14e SiR320, F,7k
Cl,7bchk,9c,21 Br,7bchk,9c,21c,22 I,7bc,21c,23 H.9c
A summary of mechanisms which have been considered
in benzyl-heteroatom photocleavages is shown in Scheme
2. For purpose of clarity, multiplicity and the extent of
solvent separation of radical or ionic pairs (6 and 7) are
(7) Penn, J . H.; Zhu, C.; Deng, D.-L.; Petersen, J . L. J . Org. Chem.
1994, 59, 3037.
(8) (a) Zimmerman, H. E.; Sandel, V. R. J . Am. Chem. Soc. 1963,
85, 915. (b) Appleton, D. C.; Brocklehurst, B.; McKenna, J .; McKenna,
J . M.; Smith, M. J .; Taylor, P. S.; Thackeray, S.; Walley, A. R. Chem.
Commun. 1977, 108. (c) Appleton, D. C.; Brocklehurst, B.; McKenna,
J .; McKenna, J . M.; Thackeray, S.; Walley, A. R. J . Chem. Soc., Perkin
Trans. 2 1980, 87. (d) Ivanov, V. L.; Ivanov, V. B.; Kuz'min, M. G. J .
Org. Chem. USSR (Engl.) 1972, 8, 1263. (e) Ivanov, V. B.; Ivanov, V.
L.; Kuz'min, M. G. Mol. Photochem. 1974, 6, 125. (f) Givens, R. S.;
Matuszewski, B.; Neywick, C. V. J . Am. Chem. Soc. 1974, 96, 5547.
(g) J aeger, D. A. J . Am. Chem. Soc. 1975, 97, 902. (h) Arnold, B.;
Donald, L.; J urgens, A.; Pincock, J . A. Can. J . Chem. 1985, 63, 3140.
(i) Hilborn, J . W.; MacKnight, E.; Pincock, J . A.; Wedge, P. J . J . Am.
Chem. Soc. 1994, 116, 3337. (j) Pincock, J . A.; Wedge, P. J . J . Org.
Chem. 1994, 59, 5587. (k) Bartl, J .; Steenken, S.; Mayr, H.; McClelland,
R. A. J . Am. Chem. Soc. 1990, 112, 6918. (l) Kim, J . M.; Pincock, J . A.;
Can. J . Chem. 1995, 73, 885.
(9) (a) DeCosta, D. P.; Pincock, J . A. J . Am. Chem. Soc. 1993, 115,
2180. (b) DeCosta, D. P.; Pincock, J . A. Can. J . Chem. 1992, 70, 1879.
(c) Pincock, J . A.; Wedge, P. J . J . Org. Chem. 1995, 60, 4067. (d)
Yamaguchi, F.; Kuriyama, Y.; Sakuragi, H.; Tokumaru, K. Tetrahedron
Lett. 1992, 33, 5529.
(10) (a) Wan, P.; Chak, B.; Carrier, L. Tetrahedron Lett. 1986, 27,
2937. (b) Wan, P.; Chak, B. J . Chem. Soc., Perkin Trans. 2 1986, 1751.
(c) Manring, L. E.; Peters, K. S. J . Phys. Chem. 1984, 88, 3516.
(11) (a) Wan, P.; Chak, B.; Krogh, E. J . Photochem. Photobiol. A
1989, 46, 49. (b) Diao, L.; Yang, C.; Wan, P. J . Am. Chem. Soc. 1995,
117, 5369.
* Author to whom correspondence should be addressed. Phone )
(801) 378-4054. FAX ) (801) 378-5474. E-mail ) Steve_Fleming@
byu.edu.
X Abstract published in Advance ACS Abstracts, September 1, 1996.
(1) (a) Fleming, S. A.; J ensen, A. W. J . Org. Chem. 1993, 58, 7135.
(b) Fleming, S. A.; Rawlins, D. B.; Samano, V.; Robins, M. J . J . Org.
Chem. 1992, 57, 5968. (c) Fleming, S. A.; Rawlins, D. B.; Robins, M. J .
Tetrahedron Lett. 1990, 31, 4995.
(2) Cristol, S. J .; Bindel, T. H. Photosolvolysis and Attendant
Photoreactions Involving Carbocations. In Organic Photochemistry;
Padwa, A., Ed.; Marcel Dekker; New York, 1983; Vol. 6, Chapter 5.
(3) In general as solvent nucleophilicity increases and radical
abstractability decreases, photosolvolysis products are formed.
(4) Photofugacity refers to the rate of leaving group dissociation from
the excited state. For a discussion, see ref 7h.
(5) Theoretically photosolvolysis should only occur from the singlet
since heterolytic cleavage of the benzyl-heteroatom bond would involve
concerted movement of unpaired electrons into the same bond.
However, some evidence indicates that although spin forbidden this
process may occur.7c,k,21b,c See: Cristol, S. J .; Schloemer, G. C. J . Am.
Chem. Soc. 1972, 94, 5916.
(6) Some counterions have been shown to reduce intermediate
radical or ionic pairs. For an example, see ref 19b.
S0022-3263(96)00692-5 CCC: $12.00 © 1996 American Chemical Society

Photocleavage of Benzyl-Sulfur Bonds
J . Org. Chem., Vol. 61, No. 20, 1996 7041
Sch em e 1
Sch em e 3
Sch em e 2
Sch em e 4
not indicated in the scheme. Intermediates 4 and 5,
involving intramolecular single electron transfer (SET),
have received limited attention.24
Although other pathways are possible, the most logical
explanation for radical pair (6) formation is homolytic
cleavage directly from the excited state of 1. Formation
of the ionic pair (7) has been rationalized by two main
theories. The original proposal, suggested by Zimmer-
man and Sandel,7a invoked formation of 7 via heterolytic
cleavage from an excited state “resonance structure” of
1 (Scheme 3). This theory was based on substituent
effect studies of benzyl acetate photocleavage which
showed increased formation of solvolysis products (2)
upon photolysis of m-methoxy derivatives relative to the
p-methoxy derivative. Electron density calculations in-
dicate a higher electron density at the meta position than
there is at the para position in the excited state for
benzene substituted with electron-donating groups (EDG).
More recent analysis further supports this excited state
phenonemon.25 The impact of substituents on excited
state reactions has frequently been called the “meta
effect” to explain particularly the formation of photosol-
volysis products (2) arising from the heterolytic cleavage
pathway.
The second theory for formation of the ionic pair (7) is
that homolytic cleavage is the predominate pathway and
the radical intermediate 6 can then form 3 and other
radical products or undergo SET to form 7. This theory
was originally suggested by McKenna et al.7c and recently
furthered by Pincock et al.7i,j The Pincock group uses
Marcus electron transfer theory to explain increased
photosolvolysis. They recently examined substituted
benzyl acetates and pivalates (see Scheme 4) and found
that when oxidation potentials of substituted benzyl
radicals are plotted versus the rate of electron transfer,
a classical Marcus curve with an inverted region is
obtained. Their studies indicate that the 3-OMe benzyl
(12) (a) Givens, R. S.; Matuszewski, Athey, P. S.; Stoner, R. M. J .
Am. Chem. Soc. 1990, 112, 6016. (b) Givens, R. S.; Singh, R.
Tetrahedron Lett. 1991, 32, 7013. (c) Givens, R. S.; Kueper, L. W. Chem.
Rev. 1993, 93, 55. (d) Furuta, T.; Torigai, H.; Osawa, T.; Iwamura, M.
Chem. Lett. 1993, 1179. (e) Furuta, T.; Torigai, H.; Sugimoto, M.;
Iwamura, M. J . Org. Chem. 1995, 60, 3953.
(13) J aeger, D. A. J . Am. Chem. Soc. 1976, 98, 6401.
(14) (a) Ratcliff, M. A.; Kochi, J . K. J . Org. Chem. 1971, 36, 3112.
(b) Hyomaki, J .; Koskikallio, J . Acta Chem. Scand. Ser. A 1977, 31,
321. (c) Appleton, D. C.; Bull, D. C.; Givens, R. S.; Lillis, V.; McKenna,
J .; McKenna, J . M.; Thackeray, J .; Walley, A. R. J . Chem. Soc., Perkin
Trans. 2 1980, 77. (d) Lillis, V.; McKenna, J .; McKenna, J . M.; Smith,
M. J .; Taylor, P. S.; Williams, J . H. J . Chem. Soc., Perkin Trans. 2
1980, 83. (e) Breslin, D. T.; Saeva, F. D. J . Org. Chem. 1988, 53, 713.
(f) Foster, B.; Gaillard, B.; Mathur, N.; Pincock, A. L.; Pincock, J . A.;
Sehmbey, C. Can. J . Chem. 1987, 65, 1599. (g) Michelsen, R. M.S.
Thesis, Brigham Young University, 1991. (h) Hassoon, S.; Sarker, A.;
Rodgers, M. A. J .; Neckers, D. J . Am. Chem. Soc. 1995, 117, 11369.
(15) (a) Maycock, A. L.; Berchtold, G. A. J . Org. Chem. 1970, 35,
2532. (b) Saeva, F. J . Am. Chem. Soc. 1991, 113, 5333.
(16) (a) Amiri, A. S.; Mellor, J . M. J . Photochem. 1978, 9, 902. (b)
Givens, R. S.; Matuszewski, B. Tetrahedron Lett. 1978, 861. (c) Givens,
R. S.; Hrinczenko, B.; Liu, J . H.-S.; Matuszewski, B.; Tholen-Collison,
J . J . Am. Chem. Soc. 1984, 106, 1779. (d) Gould, I. R.; Tung, C.; Turro,
N. J .; Givens, R. S.; Matuszewski, B. J . Am. Chem. Soc. 1984, 106,
1789. (e) Pitchumani, K.; Velusamy, P.; Banu, H. S.; Srinivasan, C.
Tetrahedron Lett. 1995, 36, 1149.
(17) Guo, Y.; J enks, W. S. J . Org. Chem. 1995, 60, 5480.
(18) (a) Thyrion, F. C. J . Phys. Chem. 1973, 77, 1478. (b) Sucholeiki,
I. Tetrahedron Lett. 1994, 35, 7307.
(19) (a) Griffin, C. E.; Kaufman, M. L. Tetrahedron Lett. 1965, 773.
(b) Imrie, C.; Modro, T. A.; Rohwer, E. R.; Wagener, C. C. P. J . Org.
Chem. 1993, 58, 5643.
(20) (a) Steinmetz, M. G. Chem. Rev. 1995, 1527. (b) Fasani, E.;
d’Alessandro, N.; Albini, A.; Mariano, P. S.; J . Org. Chem. 1994, 59,
829. (c) Kako, M.; Kakuma, S.; Hatakenaka, K.; Nakadaira, Y.; Yasui,
M.; Iwasaki, F. Tetrahedron Lett. 1995, 36, 6293. (d) Baciocchi, E.;
DelGiacco, T.; Elisei, F.; Ioele, M. J . Org. Chem. 1995, 60, 7974.
(21) (a) Cristol, S. J .; Bindel, T. H. J . Org. Chem. 1980, 45, 951. (b)
Cristol, S. J .; Bindel, T. H. J . Am. Chem. Soc. 1981, 103, 7287. (c)
Slocum, G. H.; Schuster, G. B. J . Org. Chem. 1984, 49, 2177.
(22) Rezenda, D. de B.; Campos, I. P. de A.; Toscano, V. G.; Catalani,
L. H. J . Chem. Soc., Perkin Trans. 2 1995, 1857.
(23) Slocum, G. H.; Kaufmann, K.; Schuster, G. B. J . Am. Chem.
Soc. 1981, 103, 4625.
(24) For a recent example of time-esolved studies on benzyl halide
photocleavage, see: McGowan, W. M.; Hilinski, E. F. J . Am. Chem.
Soc. 1995, 117, 9019.
(25) Zimmerman, H. E.; J . Am. Chem. Soc. 1995, 117, 8988.

7042 J . Org. Chem., Vol. 61, No. 20, 1996
Fleming and J ensen
Ta ble 1. Φ Va lu es for 1a -k
radical is at the maximum of the curve. Hence, this
substitution could lead to more efficient SET from the
initial benzyl radical compared to other substituted
derivatives. Unfortunately, the study was able to com-
pare only a few substituents due to undesired side
reactions of other substituted benzyl acetates.26
Gen er a l Su bstitu en t Effects. In ground state ionic
reactions, resonance effects on benzyl ions are greater
in the para position than the meta, and in general,
inductive effects are greater for meta than para substitu-
tion. In ground state radical studies, para substitution
of any type tends to stabilize a benzyl radical and meta
substituents have a minimal effect or destabilize the
benzyl radical.27 Ground state radical stabilization scales
(σ• scales) have been proposed for a variety of reaction
types.28
compd
N2Φ
F-P-T Φ
1a
1b
1c
1d
1e
1f
1g
1h
1i
0.12 ( 0.045
0.31 ( 0.021
0.054 ( 0028
0.18 ( 0.073
0.11 ( 0.009
0.10 ( 0.059
0.096 ( 0.022
0.068 ( 0.015
0.139 ( 0.078
0.063
0.033 ( 0.015
0.071 ( 0.026
<0.020
0.18 ( 0.031
0.047 ( 0.003
0.078 ( 0.037
0.053 ( 0.002
0.044 ( 0.011
0.39 ( 0.01
0.023
1j
1k
0.081 ( 0.017
0.071 ( 0.016
outgassing was used for deoxygenation, were only de-
tected when the sample was irradiated to higher conver-
sions.
F -P -T vs N2 Ou tga ssin g. F-P-T deoxygenation
decreases the Φ of nine of the compounds (BPS, 4-OMe,
3-OMe, 3-Me, 4-CF3, 3-CF3, 4-CN, 4-NO2, 3-NO2) with
respect to those determined after N2 outgassing. In one
case (3-CN) the Φ is significantly higher for F-P-T. The
Φ remains unchanged for the p-methyl studies.
In excited state chemistry, little work has been done
on substituent effect scales. Yates and McEwen have
shown that data from photohydration of styrene (pre-
sumed to involve an excited state benzyl cation interme-
diate) can be used to generate a σ scale for excited state
ionic reactions (σhν).29 It is noteworthy that meta sub-
stituents demonstrated the greatest impact in the pho-
tohydration reaction. We are unaware of any attempt
to develop a σ scale based on excited state radical
reactions.
In this paper we report the substituent effects on an
excited state radical reaction. This reaction is the
photocleavage of benzyl sulfides. On the basis of our
previous work,1 and transient absorption studies,9c,18a it
is apparent that benzyl sulfides undergo almost exclusive
radical photocleavage. No products resulting from benzyl
cations are detected.
From the data it is apparent that the degree of
deoxygenation has a significant impact on the reaction
efficiency for several of the BPS derivatives. Oxygen
facilitates intersystem crossing (ISC) which is normally
demonstrated by its triplet quenching ability; although
it is highly unlikely, oxygen may also help the excited
state singlet-to-triplet conversion (ISC), and perhaps it
is the triplet state pathway that leads to cleavage so that
the decrease in oxygen concentration (F-P-T) decreases
the Φ. Alternatively, a decrease in F-P-T Φs relative
to N2 Φs may be indicative of a long caged-radical
lifetime, which in the absence of oxygen returns to
starting material (i.e., recombination) since there is no
oxygen for conversion to the corresponding benzaldehyde.
Another explanation for diminished reactivity in the
absence of oxygen is that there may be formation of an
exciplex or a ground state complex between the sulfide
and oxygen which enhances photocleavage of the benzyl
bond.
We also reasoned that it might be possible to see
photooxidation of the sulfur resulting in formation of
sulfone or sulfoxide and then benzyl photocleavage as a
secondary photochemical step. However, no oxidized
sulfur products have been detected in this study or in
that performed previously in our lab.1 Furthermore,
recent publications concerning the quantum yield for
photocleavage of this type indicates that such reactions
are relatively inefficient.30
Resu lts a n d Discu ssion
P r oced u r e. We irradiated 1a -k in cyclohexane with
a low-pressure Hg arc lamp (254 nm) and measured the
quantum yields (Φs) for loss of starting material at low
conversions (typically 5-15%). We determined Φ values
1
after outgassing with N2 for /2 h. Under these conditions
we observed that benzaldehydes are formed as minor
photoproducts at longer conversions (up to 95% conver-
sion). Therefore, we reasoned that more accurate values
would be obtained if oxygen were more thoroughly
removed from the samples prior to photolysis. We
subsequently irradiated 1a -k after freeze-pump-thaw
(F-P-T) degassing of the samples and redetermined the
Φs on the basis of loss of starting benzyl sulfide. Both
sets of Φs are shown in Table 1.
It is clear from comparing Φs after N2 bubbling and
F-P-T cycles that it is important to remove oxygen in
order to more accurately evaluate the substituent effects.
This ought to be a priority in all attempts to compare
the pathways for the general reaction of benzyl photo-
cleavage. The remainder of our analysis for this discus-
sion will consider the F-P-T quantum yields only.
Φ Or d er . The overall Φ order for loss of starting
material is 3-CN > 4-Me > 4-CF3 > 4-OMe ) 3-NO2 >
3-CF3 > 3-Me > 4-CN > H > 4-NO2 > 3-OMe (see Table
1). Interestingly, when the meta substituents are sepa-
rated from the para substituents a trend becomes evi-
dent. The differences from one meta to the other are all
significant (greater than 1 standard deviation difference).
In general there are only two photoproducts detected
at low conversions, substituted bibenzyls and diphenyl
disulfide. Other products, including benzaldehydes if N2
(26) Personal communication with P. J . Wedge.
(27) (a) Creary, X.; Mehrsheikh-Mohammadi, M. E.; McDonald, S.
J . Org. Chem. 1987, 52, 3254. (b) Wagner, D. D. M.; Arnold, D. R. Can.
J . Chem. 1985, 63, 2378. (c) Nicholas, A. M. de P.; Arnold, D. R. Can.
J . Chem. 1986, 64, 270. (d) Arnold, D. R. The Effect of Substituents
on Benzylic Radical ESR Hyperfine Coupling Constants. The σ°a Scale
Based Upon Spin Delocalization. In Substituent Effects in Radical
Chemistry; Viehe, H. G., J anousek, Z., Merenyi, R., Eds.; D. Reidel:
Dordrecht, 1986; pp 171-188.
(28) (a) J ackson, R. A. σ• Revisited. In Substituent Effects in Radical
Chemistry; Viehe, H. G., J anousek, Z., Merenyi, R., Eds.; D. Reidel:
Dordrecht, 1986; pp 325-328. (b) Walling, C.; Briggs, E. R.; Wolfstirn,
K. B.; Mayo, F. R. J . Am. Chem. Soc. 1948, 70, 1537. (c) Agirbas, H.;
J ackson, R. A. J . Chem. Soc., Perkin Trans. 2 1983, 739. (d) Dinc¸tu¨rk,
S.; J ackson, R. A. J . Chem. Soc., Perkin Trans 2 1981, 1127.
(29) McEwen, J .; Yates, K. J . Phys. Org. Chem. 1991, 4, 193.
(30) J enks, W. S.; Taylor, L. M.; Guo, Y.; Wan, Z. Tetrahedron Lett.
1994, 35, 7155.

Photocleavage of Benzyl-Sulfur Bonds
J . Org. Chem., Vol. 61, No. 20, 1996 7043
Sch em e 5
F igu r e 2.
Their relative quantum yields are 3-CN . 3-NO2 > 3-CF3
> 3-Me > 3-OMe. The cyano and nitro compounds may
involve the triplet excited state, while the others are
presumably singlet pathways.31 This order clearly obeys
the following sequence: EWG with delocalizable electrons
> EWG without delocalizable electrons > EDG without
delocalizable electrons > EDG with delocalizable elec-
trons.
Among the para-substituted compounds, with the
exception of 4-CF3, the EWGs appear to result in a less
efficent reaction while the EDGs appear to help reaction
efficiency. If only the mean values (not standard devia-
tions) are considered, then the order is 4-Me . 4-CF3 >
4-OMe > 4-CN > 4-NO2. However, the difference from
one para derivative to the next is only statistically
significant for the 4-Me and 4-CF3. This is similar to a
trend seen in ground state σ• scales in which para EWGs
stabilize and meta EWGs destabilize the radical to the
greatest extent.27
Our Φ order for the meta-substituted compounds can
be explained by Zimmerman’s meta effect. In addition
to previous excited state calculations that describe the
excited state for substituted aromatic rings, we have
performed semiempirical calculations using MOPAC32
which support the proposed deformation of the aromatic
ring in the excited state. Our results indicate that there
is an increase for C-2-C-6 bonding in anisole and
significant positive charge buildup on the ipso carbon in
addition to an increase in electron density at C-3 and C-5
(see Figure 2).33 We interpret our results and these
calculations as clear indication that there is a “meta
effect”. A substitutent in the meta position will stabilize
the excited state and/or provide a pathway of reactivity
that is not available for the para-substituted analogue.
Since an ipso carbon could be stabilized in the excited
state by an electron-withdrawing as well as an electron-
donating group, we propose that a reverse polarized
excited state is possible.
compounds. It should be emphasized that the meta effect
does not predict formation of a full charge at the meta
position; our calculations imply that the degree of charge
is a function of the substituent on the ipso carbon. It is
more accurate to view the anions and cations in the
resonance structures as merely increased or decreased
p-orbital electron density relative to one electron.
We have established that the benzyl cleavage process
for the sulfides is a radical pathway. It is necessary,
therefore, to compare the ease of radical formation next
to a cationic center with radical formation next to an
anionic center. Ground state σ• values indicate that a
radical which is able to delocalize with an electron
deficient group is more stable than a radical which is able
to delocalize with an electron rich group, as long as the
substituent does not contain d orbitals adjacent to the
radical.27a,d These established trends are consistent with
our results. A meta EWG will favor radical cleavage of
the sulfur bond compared to the meta EDG (see Scheme
5). One can rationalize this selectivity with orbital-
following similar to the recent work by Wagner on
photocleavage of aryl halide bonds.34
This analysis is consistent with significant differences
observed in the Φs of the meta-substituted benzyl phenyl
sulfides and relatively small differences between the para
(31) In order to determine the reaction multiplicity, we irradiated
1a with acetone as the solvent. The result was that we saw no
significant change in the quantum yield. However, similar attempts
to determine reaction multiplicity21a,b have been justifiably criticized
due to competitive absorption of the starting material and quencher.7f,h
It has been suggested that a better method for determining reaction
multiplicity in benzyl-heteroatom photocleavages is to use the naph-
thyl system instead of benzyl. Among the substituents we studied, it
is likely that the nitro (1j and 1k ) and perhaps the cyano (1h and 1i)
involve a triplet excited state. See: Blakemore, D. C.; Gilbert, A. J .
Chem. Soc., Perkin Trans. 1 1992, 16, 2265. For a recent example of
excited state multiplicity playing a role in homolytic vs heterolytic
cleavages, see: J ime´nez, M. C.; Miranda, M. A.; Tormos, R. Heterocycles
1996, 43, 339.
Recent work by Pincock also supports this explanation
for the “meta effect”. He reports, “substituents also exert
a strong electronic effect on the rate of the homolytic
cleavage step...”.8c
P r op osa l of a New σ‚h ν Sca le. Since we believe that
this reaction is an excited state radical reaction, we are
in a position to propose, using these quantum yields as
the basis, a new excited state radical scale. We can
(32) CACheMOPAC version 94 with multiplicity ) singlet, param-
eter ) AM1, and CI level ) 2; CAChe Scientific, Inc.
(33) For a similar proposed polarity in the excited state of aromatic
compounds, see: Weber, G.; Runsink, J .; Mattay, J . J . Chem. Soc.,
Perkin Trans. 1 1987, 11, 2333.
(34) Wagner, P. J .; Sedon, J .; Waite, C.; Gudmundsdottir, A. J . Am.
Chem. Soc. 1994, 116, 10284.

7044 J . Org. Chem., Vol. 61, No. 20, 1996
Fleming and J ensen
Ta ble 2. P r op osed σ•h ν Va lu es for Excited Sta te Ra d ica l
In an effectively oxygen-free environment, meta substit-
uents are shown to have a significant effect on overall
cleavage efficiency. The meta EWGs appear to facilitate
formation of 6 while meta EDGs appear to supress it.
This finding is different from ground state reactions and
supports some type of excited state meta stabilization.
In fact, it may be easily rationalized by the “meta effect”
originally proposed by Zimmerman. Semiempirical cal-
culations support the partial charge density at the ipso
carbon and C-3/C-5.
Rea ction s
substituent
σ•hν
substituent
σ•hν
H
0.00
4-CF3
3-CF3
4-CN
3-CN
0.37
0.21
0.12
1.07
4-OMe
3-OMe
4-Me
3-Me
0.33
<-0.22
0.74
0.15
generate a new scale based on quantum yields by dividing
the quantum yield of the substituent by the quantum
yield of BPS and taking the log of the resulting fraction,
in the same manner that the original σ values were
obtained by Hammett using equilibrium constants.35 The
values generated by performing this transformation are
shown in Table 2.
We do not claim that these σ values will have universal
correlation with all excited state homolytic cleavage
reactions. For example, it is clear that the benzyl
cleavage reactions from the excited state vary with
leaving group, solvent, and concentration of dissolved
oxygen. Presumably this is due to the extent of polariza-
tion of the aromatic ring in the excited state. In order
to test the validity of our scale, it would be desirable to
compare these values against other reactions that are
thought to go through excited state radical pathways.
Unfortunately, most other benzyl-heteroatom photo-
cleavages involve some type of partitioning between
radical and ionic intermediates. Thus, the σ•hν values
may suffer because they have been defined by too unique
of a reaction path.
Exp er im en ta l Section
Gen er a l P r oced u r es. Benzyl phenyl sulfide (1a ) was
purchased from Lancaster Synthesis and used as received. All
other derivatives of 1a (1b-k ) were synthesized as previously
reported.1 Cyclohexane was washed with 1:1 concentrated
nitric and sulfuric acid, washed with water, passed through a
basic alumina column, dried with CaCl2, and distilled.
Gas chromatography (GC) was performed on a Hewlett
Packard 5890 instrument equipped with a flame ionization
detector (FID) and fused methyl silicone stationary columns
(50 m and 18 m × 0.25 mm × 0.25 µm). Hydrogen was used
as the carrier gas with compressed air mixing at the detector.
Qu a n tu m Yield s. (a ) Qu a n tu m Yield for Va ler op h e-
n on e. The quantum yield of dissappearance of valerophenone
at 313 nm is known (0.46).37 Potassium ferrioxylate actinom-
etry38 was used on an optical bench equipped with an ISA
Instruments monochromater, Oriel Model 66056 lamp, and
Oriel photomultiplier tubes. Valerophenone (41.5 mg) was
added to purified cyclohexane and outgassed with nitrogen for
1
/
h. The photolysis was allowed to run for 66 h, and GC
2
analysis indicated that 5.6 mg of valerophenone reacted.
A
quantum yield of disappearance of 0.4 was calculated for
valerophenone at 254 nm.
Ba ck Rea ction . One aspect of the benzyl-sulfur
photocleavage reaction that is undetermined is the extent
of back reaction or recombination of the radical pair. It
might be possible to measure the rate of recombination
using a chiral benzyl sulfide for each derivative and
measuring the percent racemization. A few benzyl
cleavages have been examined using chiral reagents;
however, when chiral benzyl acetates with isotopically
enriched carbonyl oxygens were photolyzed in a solvent
cage, scrambling of the 18O was observed in the absence
of racemization.36 This indicates that recombination
within the solvent cage can be fast enough to retain
benzyl stereochemistry. Such a chiral sulfide might then
be useful only to estimate out-of-cage recombination. We
are interested in continuing our study of these photo-
cleavage pathways and expect that time-resolved studies
will provide the best analysis of back reactions.
(b) Gen er a l Qu a n tu m Yield P r oced u r e. Samples were
prepared by dissolving 1-30 mg of material in about 6 mL of
purified cyclohexane. The samples were deoxygenated and
irradiated for 5-10 min with valerophenone actinometry.
Light was generated by a Rayonet RPR, 253.7 nm lamp, in a
merry-go-round apparatus (Rayonet miniphotochemical reac-
tor). The lamp was allowed to warm up for about 1/2 h prior
to photolysis. The samples were cooled by a fan in the bottom
of the merry-go-round.
(c) Deoxygen a tion of Qu a n tu m Yield Sa m p les. The
first method used was deoxygenation with purified nitrogen.39
The second method involved freeze-pump-thaw degassing (1
× 10-5 Torr for at least three consecutive cycles).
Ca libr a tion Cu r ves for An a lytica l Wor k . To determine
the amount of starting material lost for the quantum yield
determinations, calibration curves were generated by prepar-
ing four-five samples of material. These samples contained
between 0 and 30 mg of starting material spiked with a
consistent amount of benzophenone internal standard. The
correlation constants for the calibration curves of the deriva-
tives ranged from 0.995 to 1.000.
Su m m a r y
J O9606923
Quantum yield determination has shown that the
extent of deoxygenation plays a significant role in the
efficiency of homolytic cleavage from the excited state.
(37) Wagner, P. J . J . Am. Chem. Soc. 1967, 89, 5898.
(38) Hatchard, C. G.; Parker, C. A. Proc. R. Soc. London A 1956,
235, 518.
(39) Nitrogen was purified by being passed through an Ace-Burlitch
inert atmosphere system containing a column packed with a BASF
R3-11 catalyst followed by another column packed with Aquasorb
drying agent.
(35) Hammett, L. P. J . Am. Chem. Soc. 1937, 59, 96.
(36) Givens, R. S.; Matuszewski, B.; Levi, N.; Leung, D. J . Am.
Chem. Soc. 1977, 99, 1896.