Generation and decay of aryl sulfinyl and sulfenyl radicals: A transient absorption and computational study

Article abstract of DOI:10.1021/jp971532d

Absorption spectra and extinction coefficients of phenylsulfinyl and phenylsulfenyl (thiyl) radicals are determined by nanosecond laser photolysis in various solvents. Direct observation and characterization of arylsulfinyl radicals from the photolysis of

Full text of DOI:10.1021/jp971532d

J. Phys. Chem. A 1997, 101, 6855-6863
6855
Generation and Decay of Aryl Sulfinyl and Sulfenyl Radicals: A Transient Absorption and
Computational Study1
Alexander P. Darmanyan, Daniel D. Gregory, Yushen Guo, and William S. Jenks*
Department of Chemistry, Iowa State UniVersity, Ames, Iowa 50011-3111
X
ReceiVed: May 7, 1997; In Final Form: June 24, 1997
Absorption spectra and extinction coefficients of phenylsulfinyl and phenylsulfenyl (thiyl) radicals are
determined by nanosecond laser photolysis in various solvents. Direct observation and characterization of
arylsulfinyl radicals from the photolysis of several aromatic sulfoxides provides the strongest evidence to
date for R-cleavage as the predominant primary photochemical process for these compounds. The absorption
4
3
-1
-1
spectrum of phenylsulfinyl, with λmax ) 300 and 450 nm and ꢀ ) 1.1 × 10 and 1.3 × 10 M cm , is
practically independent of solvent. Quantum yields of free sulfinyl radicals range from 0.09 to 0.18 in various
solvents. Recombination rate constants very near diffusion control indicate that there is a large spin-orbital
coupling in the radical pair. Rate constants for the reactions of arylsulfinyl radicals with stable nitroxide
2
radicals are among the fastest known, but reactivity with O is very modest. Computations indicate that the
singly occupied molecular orbital is a π* orbital largely localized on the sulfur and oxygen atoms.
Introduction
SCHEME 1
One of the common mechanistic assumptions in sulfoxide
photochemistry is that the primary step after excitation is
homolytic R-cleavage of one of the C-S bonds to form a radical
pair or biradical.2
-24
Recently we have carefully examined a
series of aryl alkyl sulfoxides by steady state photolysis
techniques and found that nearly all of the observed chemistry
could be accounted for by R-cleavage to form a carbon centered
radical and arylsulfinyl radical. Cleavage is followed by
competition between diffusional separation of the radical pair
to form free radicals, recombination to form sulfoxide, and
absorption spectrum attributed to phenylsulfinyl has appeared
4
1
previously. This, however, served as sufficient evidence to
begin a quantitative characterization of the arylsulfinyl spec-
troscopy and reactivity that is reported herein. We have also
carried out a computational study on the structure of phenyl-
sulfinyl, designed to help elucidate the meaning of the two
resonance forms shown for 1.
2
3
recombination to form sulfenic ester (Scheme 1).
Other
secondary reactions, such as hydrogen abstraction by the
radicals, are made available by variation of the sulfoxide
precursor structure.24
Beyond their incidence in sulfoxide photochemistry, smaller
The chemistry of arylsulfenyl radicals 2 has been character-
ized much more thoroughly than that of arylsulfinyl radicals.
Picosecond regime recombination of geminate PhS radical pairs
•
•
sulfinyl radicals, notably HSO and CH3SO , are of interest
2
5,26
•
because of their involvement in atmospheric sulfur cycles.
4
2
They are implicated in the oxidations of H2S, CH3SCH3, CH3-
in hydrocarbons has been reported.
The full absorption
SSCH3, and CH3SH.27 HSO , for instance, has been proposed
•
•
spectrum of PhS was determined in water (λmax ) 295 and
•
4
3
-1 43
to lie along the pathway from HS to atmospheric H2SO4 by
460 nm; ꢀ ) 1.0 × 10 and 2.5 × 10 M cm ), and several
way of reaction with ozone.2
8-32
It has also been unambigu-
•
rate constants for reactions of PhS are known, including an
•
•
ously demonstrated that CH3SO is an intermediate in the gas
extensive set of studies on the reversible reactions of PhS with
phase photolysis of dimethyl sulfoxide.33
44,45
•
olefins.
Summarizing very briefly, PhS reactivity follows
a trend expected for an electrophilic species. No magnetic field
effects were observed on Φesc values for PhS in micelles, which
In this paper we report a solution phase transient absorption
study of aryl sulfoxide photolysis designed to characterize
directly the proposed arylsulfinyl radical intermediate 1. Ex-
tinction coefficients and absolute rate constants for reactions
of 1 have been determined, and structural information can be
retrieved from ab initio and density functional calculations.
Some physical evidence for sulfoxide R-cleavage in solution
•
was attributed to large spin-orbital coupling (SOC) in the
46
radical pairs, consistent with calculations locating the unpaired
4
7
spin localized on the sulfur atom. We report the absorption
spectrum, extinction coefficients, quantum yields, and absolute
•
rate constants for recombination of PhS in organic solvents in
•
addition to the data for ArSO species.
(beyond conclusions drawn from product study) was previously
available. This took the form of weak EPR signals of sulfinyl
radicals at low temperature3
4-37
Experimental Section
and chemically induced dy-
namic nuclear polarization (CIDNP) signals attributed to the
Spectroscopy. All experiments were carried out using a
computer-controlled nanosecond transient absorption spectrom-
eter. The samples were irradiated with the fourth harmonic of
a Continuum Surelite Nd:YAG laser (266 nm, 5 ns, 2-25 mJ/
pulse, 3 mm beam radius). The spectroscopic detection system
includes a pulsed 75 W xenon lamp (τ ∼ 1 ms), an ISA H10
monochromator, an 1P-28 photomultiplier (Rl ) 50 Ω), and a
intermediacy of sulfinyls.3
8-40
However, overinterpretation of
such data without corroborating evidence can be misleading as
they do not necessarily indicate radical pathways for the majority
of material. Only a brief and qualitative report of a transient
X
Abstract published in AdVance ACS Abstracts, August 15, 1997.
S1089-5639(97)01532-6 CCC: $14.00 © 1997 American Chemical Society

6856 J. Phys. Chem. A, Vol. 101, No. 37, 1997
Darmanyan et al.
Textronix TDS-250 200 MHz transient digitizer. Control of
the experiment, data collection, and processing were carried out
using a Macintosh with Labview 2 software. The laser beam
was used at 90° with respect to the probe beam. Investigations
were carried out in a quartz flow cell 1 × 1 × 5 cm or in a
regular quartz cell where solutions were changed after one or
two laser pulses because of extensive photodecomposition.
When it was necessary, the decay kinetics were averaged for
several laser pulses. The optical densities of solutions were
∼
0.3 at 266 nm. The accuracy of quantum yields, ꢀ, and various
rate constants is estimated to be (20%. Except for reactions
with O2, rate constants were determined from lifetimes observed
at no fewer than five quencher concentrations.
Absorption spectra of solutions were recorded on a UV-2101
PC Shimadzu spectrophotometer. The solutions were deoxy-
genated by Ar bubbling or saturated with O2 when required for
•
2
0 min. Experiments were carried out at ambient temperature,
Figure 1. Differential absorption spectrum of PhSO radical after
-
5
excitation of 8 (X ) H, 6.2 × 10 M) in air-saturated cyclohexane
approximately 25 °C.
Solvents. Spectrograde solvents were used as received. The
•
(
solid line). Absorption spectrum of PhSO corrected for ground state
bleaching of sulfoxide (dashed line). Insert: Second order plot of the
decay kinetics at 300 nm in the same solution.
2
-methyl-2-propanol, which though not spectrograde did not
contain any interfering absorbances, was treated with 1% (by
volume) water to prevent freezing. Water was produced with
a Millipore Milli-Q UVPlus deionizer.
CHART 1
Materials. Diphenyl sulfoxide (5) was obtained com-
mercially and recrystallized from ethanol. Diphenyl disulfide,
2,2,6,6-tetramethyl-1-piperidinyloxy (TEMPO), di-tert-butyl ni-
troxide (DTBN), and galvinoxyl were used as received from
commercial sources. The preparation of phenyl tert-butyl
sulfoxide (6), benzyl phenyl sulfoxide (8, X ) H), and benzyl
p-tolyl sulfoxide (8, X ) CH3) have been described previ-
2
3,24
ously.
Phenyl Diphenylmethyl Sulfoxide (7). The sulfoxide was
prepared in 60% isolated yield by oxidation of the sulfide with
the urea-hydrogen peroxide complex followed by column
chromatography on silica.48 H NMR (CDCl3): δ 7.23-7.39
1
13
(
m, 15 H), 4.80 (s, 1 H). C NMR (CDCl3): δ 77.8, 125.0,
1
1
28.2, 128.5, 128.6, 128.8, 129.3, 129.7, 131.1, 134.1, 135.5,
42.9. Analysis of the 13C NMR intensities clearly indicated
that one of the peaks represented two diastereotopic carbons,
but it was not determined which of several it was.
The sulfide was prepared by modification of the method of
Finzi and Bellavita.49 Benzhydrol (11 g, 59 mmol) was
dissolved in 75 mL acetic acid and 25 mL of sulfuric acid at
room temperature. To this mixture, thiophenol (59 mmol) was
added in a dropwise fashion. After 2 h of stirring, the mixture
was filtered, and the precipitate was washed thoroughly with
water and then dried. This provided 14.4 g (59 mmol) of the
with the GAMESS suite of progams50 and using unrestricted
51
Hartree-Fock (UHF) and Becke3LYP methodology on the
5
2
Gaussian 92/DFT suite. Orbitals were visualized with Mac-
MolPlot, which is available as a utility with GAMESS. The
default 6-311G basis sets in Gaussian were modified to conform
with those in GAMESS, as developed by McLean and Chan-
1
sulfide of sufficient purity to carry on the oxidation. H NMR
5
3
dler. The use of this basis set only affects sulfur among the
atoms used here. Rotational barriers include zero point energies,
scaled by a factor of 0.9, and uncorrected for temperature, i.e.,
at 0 K.
(
5
1
CDCl3): δ 7.40 (d, J ) 7.2 Hz, 4 H), 7.12-7.30 (m, 11 H),
.53 (s, 1 H). C NMR (CDCl3): δ 57.3, 126.5, 127.2, 128.4,
28.5, 128.7, 130.4, 136.1, 141.0.
1
3
Benzyl p-chlorophenyl sulfoxide (8, X ) Cl) was prepared
in analogy to phenyl benzyl sulfoxide.23 1H NMR (CDCl3): δ
Results
7
.24-7.42 (m, 7 H), 6.97 (dd, J ) 1.6, 7.6 Hz, 2 H), 4.10 (d,
1
3
J ) 12.6 Hz, 1 H), 3.98 (d, J ) 12.6 Hz, 1 H). C NMR
CDCl3): δ 63.6, 125.9, 128.5, 128.6, 128.7, 129.2, 130.4, 137.4,
The sulfoxides used in the study are shown in the Chart 1.
(
The excitation of all sulfoxides (5-8) led to the formation of a
1
41.3.
transient within the response time of the instrument.
A
Benzyl p-methoxyphenyl sulfoxide (8, X ) CH3O) was
representative example, obtained from compound 8 (X ) H)
in cyclohexane, is shown in Figure 1. The portion of the
transient absorption spectrum with maxima at 300 and 450 nm
is not affected significantly by change of the precursor sulfoxide,
prepared in analogy to phenyl benzyl sulfoxide.23 H NMR
1
(
CDCl3): δ 7.20-7.32 (m, 5 H), 6.89-6.99 (m, 4 H), 4.09 (d,
J ) 12.3 Hz, 1 H), 3.94 (d, J ) 12.3 Hz, 1 H), 3.82 (s, 3 H).
1
3
C NMR (CDCl3): δ 55.6, 63.8, 114.4, 126.4, 128.2, 128.5,
29.4, 130.5, 133.7, 162.1.
solvent, or presence of O (e.g., Figure 2). We did not observe
2
1
any long-lived excited state of the sulfoxide, consistent with
2
3
Computations. Structural optimizations were carried out
using restricted open-shell Hartree-Fock (ROHF) methodology
our previous assignment of singlet cleavage, but a very short-
lived triplet cannot be ruled out.

Sulfinyl and Sulfenyl Radical Generation and Decay
J. Phys. Chem. A, Vol. 101, No. 37, 1997 6857
Figure 3. Absorption spectrum of transients after excitation of 4 (6.5
•
-5
-1 -1
Figure 2. Differential absorption spectrum of PhSO after excitation
× 10
M
s ). The solid line is immediately after the laser pulse in
-4
of 8 (X ) H, 1 × 10 M) in air-saturated 2-methyl-2-propanol (solid
Ar-flushed acetonitrile, and the dashed line is for an O
solution, taken 100 ns after the pulse.
2
-saturated
•
line). Absorption spectrum of PhSO corrected for the ground state
bleaching of sulfoxide (dashed line). Insert: Absorption spectrum of
-
5
transients after excitation of 8 (X ) H, 8.2 × 10 M) in degassed
•
56,60-63
The absorption spectrum of Ph2CH is well-known.
It
acetonitrile immediately after the laser pulse (solid line). Absorption
has maxima at 331 and 318 nm (shoulder) with extinction
•
spectrum of PhSO only in the same air-saturated solution recorded
4
4
-1
-1
coefficients of 4.4 × 10 and 3.1 × 10 M cm , respec-
2
00 ns after laser pulse (dashed line).
6
1
tively. The extinction coefficient at 266 nm is of the order
-
1
-1
of 3000 M cm , and it undergoes only photophysical decay
processes from excited doublet states.
Excitation of 7 (at 266 nm) in degassed acetonitrile led to a
transient with the clear superposition of Ph2CH and PhSO
The decay of the 300 nm transient was very well fit to second
5
6,61,63
order kinetics, and the initial intensity of the signal was directly
proportional to the energy of the incident laser pulse. The decay
kinetics of the transient and its yield were the same in
Ar-flushed, air-saturated, or O2-saturated solutions. For all
sulfoxide precursors in acetonitrile (X ) H), the signal at λmax
•
•
absorptions (Figure 3). The decay kinetics of the peak at 331
nm were well-fit by second order decays with 2kr/ꢀ ) 9.3 ×
4
-1
1
0 cm s . Using Bromberg’s estimate of the extinction
)
300 nm decays with the same rate constant: 2kr/ꢀ ) (5.6 (
5
-1
coefficient, the recombination rate constant of diphenylmethyl
0
.3) × 10 cm s . We assigned the transient to the phenyl-
9
-1 -1
•
was estimated at 2kr ) 4.1 × 10 M s . Ph2CH reacts
sulfinyl radical 1 (X ) H) and the decay kinetics to its self-
5
9,63
effectively with O2. Given the concentration of O2 in air-
recombination.
64
saturated acetonitrile at ambient temperature (1.9 mM) and a
first order lifetime of 350 ns, a rate constant for the consumption
•
In addition to PhSO , the excitation of 5 and 6 should lead
to the simultaneous formation of phenyl and tert-butyl radicals,
9
-1 -1
of benzyl by O2 was roughly estimated at 1.5 × 10 M s ,
5
4
respectively, with absorption maxima at 260 and 230 nm.
However, the low extinction coefficients for these radicals and
6
3
which compares well with reported values.
Further evidence for the assignment of the transient absorption
spectrum in Figure 3 is the observation of fluorescence of Ph2-
CH . The radical is created and then excited within the same
laser pulse, and an emission spectrum was observed with
maxima at 520 and 540 nm, in agreement with observations of
3
-1
-1
their peroxy derivatives (e10 M cm ) imply that those
species will not significantly distort the sulfinyl radical signals
in deaerated or air-saturated solutions. Because of the strong
overlap of these absorption spectra with those of the sulfinyl
radical and the starting materials, they could not be recorded.
We did observe the absorption of unidentified products at
•
6
1-63
other workers.
Molecular oxygen quenches excited Ph2-
CH even faster than it does the ground state. The observed
•
∼
320-600 and ∼320-350 nm in photolysis of 5 and 6,
6
-1
decay rate of the fluorescence was 3.3 × 10 s in Ar-flushed
respectively. The contributions of these products to the spectra
were small, and the lifetimes were g200 µs in deoxygenated
or air-saturated solutions.
7
-1
acetonitrile and 2.9 × 10 s in air-saturated solution, giving
1
0
-1 -1
a two-point rate constant of ca. 1.5 × 10
with the diffusion-controlled quenching rate constant (3.7 ×
M
s , consistent
Photolysis of benzyl phenyl sulfoxide 8 (X ) H) at 266 nm
10
-1 -1 65
1
0
M
s ) mitigated by a spin-statistical factor of 1/3 for
•
leads to the formation of PhSO and benzyl radicals in
63
the reaction
acetonitrile (Figure 2). The absorption spectrum of benzyl
5
5-58
radical has two strong maxima, at 258 and 316 nm.
It
•
3
•
1
1
Ph CH * + O f Ph CH + O ( ∆ )
reacts very quickly with O2 and the pseudo-first order lifetime
2
2
2
2
g
5
9
of benzyl in air-saturated acetonitrile should be about 170 ns.
•
•
Thus, in principle, one could determine the extinction coef-
Because of the rapid quenching of Ph2CH * and Ph2CH by
molecular oxygen and the relative unreactivity of PhSO with
•
•
ficients for absorption by PhSO from the data in the insert of
Figure 2. In practice this was not possible due to the strong
overlap of the two absorption spectra. Additionally, the
absorption spectrum of the sulfenic ester formed by geminate
oxygen, the transient absorption spectrum observed 100 ns after
the laser pulse in O2-saturated solution was strongly dominated
•
by PhSO with its characteristic 300 nm λmax. The decay of
•
•
5
-1
recombination of PhCH2 and PhSO has a maximum at 310
this transient was second order: 2kr/ꢀ ) 5.4 × 10 cm . Since
nm, causing further interference.2
3,24
Therefore, precursor 7 was
the concentration of Ph2CH and PhSO radicals escaped from
geminate recombination in the solvent cage should be equal,
• •
chosen for quantitative evaluation of the sulfinyl radical
absorption, as the absorption of the diphenylmethyl radical (Ph2-
•
the extinction coefficient of PhSO could be estimated from the
•
CH ) was shifted far enough away to not interfere.
spectra at early times and the equation

6858 J. Phys. Chem. A, Vol. 101, No. 37, 1997
Darmanyan et al.
TABLE 1: Photophysical Properties of the Phenylsulfinyl
Radical
TABLE 2: Quantum Yields of Sulfinyl Radicals and
Consumption of Sulfoxides
η(25 °C)a 2kr/ꢀ (10 2k
5
b
(10
s
9
k
dif(25 °C)a
sulfoxide
solvent
Φ
•
Φ
a
r
PhSO
chem
solvent
(cP)
cm s-
1
)
M
-1 -1
)
(10 M
9
-1 -1
s
)
Φ
•
PhSO
PhSOPh (5)
cyclohexane
2-methyl-2-propanol
cyclohexane
acetonitrile
cyclohexane
0.06
0.05
0.12
0.18
0.11
0.09
ethyl ether
hexane
acetonitrile
cyclohexane
water
0.24c
0.29
0.34
0.90
0.89
1.08
1.94
2.04
4.31
12.0
9.0
5.8
5.65
3.6
3.2
2.3
2.8
1.55
13.0
10.0
6.4
6.2
4.0
3.5
2.5
3.1
1.7
27
22
19
0.18
0.16
0.12
0.12
0.10
0.14
0.12
0.12
0.09
0.034
PhSOC(CH
PhSOCHPh
3
)
3
(6)
2
(7)
7.4
PhSOCH
2
Ph (8)
7.4
6.1
3.4
3.2
1.5
2-methyl-2-propanol
0.20
ethanol
a
The quantum yield for chemical consumption of the sulfoxide.23,24
1
2
2
-propanol
-propanol
-methyl-2-
propanol
d
•
•
TABLE 3: Rate Constants for Reaction of PhS and PhSO
with Stable Radicals
a
At 25 °C.64,97 These values are estimated from 2k
b
r
/ꢀ using
. At 20 °C. The true value of
rxn (10 M s-1
8
-1
k
)
4
-1
-1
c
d
ꢀ
k
PhSO•(300 nm) ) 1.1 × 10 M cm
radical
σ+
0.11 310, ∼460
300, 450
λ
max (nm) TEMPO DTBN
solvent
dif should be slightly higher due to the addition of 1% water.
1
1
1
1
(X ) Cl)
(X ) H)
(X ) CH
15.6
9.4
8.4
14.4
10.0
9.7
8.1
32
29
23
acetonitrile
acetonitrile
acetonitrile
acetonitrile
hexane
0
b
3
)
-0.31 310
(X ) OCH
3
)
)
-0.78 320, 515c
0.11 310, ∼460
7.8
where ꢀ0 ) 5 × 10 M cm-1 and is the extinction coefficient
2
-1
1 (X ) Cl)
1
1
1
(X ) H)
(X ) CH
0
300, 450
hexane
hexane
hexane
of the ground state of sulfoxide at 300 nm. The dependences
of the yield of both radicals on the energy of the laser pulse
were strictly linear. From data presented in Figure 3, we
-0.31 310b
3
)
(X ) OCH
3
-0.78 320, 515
-0.31 310b
0.11
23
1 (X ) CH
2 (X ) Cl)
3
)
13
16
cyclohexane
cyclohexane
acetonitrile
cyclohexane
4
-1
-1
a
obtained ꢀPhSO•(300 nm) ) (1.1 ( 0.1) × 10 M cm .
2
2
2
2
(X ) H)
(X ) H)
(X ) CH
0
0
295
295
14
17
13
As measured by this technique, the extinction coefficient of
a
•
PhSO is practically independent of solvent. Thus, the quantum
yield of PhSO avoiding geminate recombination and escaping
into the bulk (ΦPhSO•) in different solvents can be easily
estimated. The yield of PhSO was measured in comparison
3
)
-0.31
-0.78
7.5a cyclohexane
•
a
(X ) OCH
3
)
3.3
cyclohexane
a
Reference 45. b Very small absorbance about 450 nm. c Ratio of
•
maxima in differential absorption spectrum is 2.8:1.
with the triplet-triplet absorption of anthracene in degassed
cyclohexane with these parameters: ΦT ) 0.71 and ꢀT(422.5
4
-1
-1 66-68
nm) ) 6.47 × 10 M cm .
The optical density at the
excitation wavelength and the energy of laser pulse (7 mJ) were
equal for solution of sulfoxide and anthracene. These data,
together with the decay rate constants, are presented in Table
1
. The ΦPhSO• values for the different sulfinyl precursors are
shown in Table 2.
Reactions with Nitroxides. Rate constants for reactions of
sulfinyl radicals with stable nitroxide radicals were determined
•
and compared with the corresponding rate constants for PhS
from the literature. Both TEMPO and DTBN were used. Rate
constants were found by the usual first order expression
•
kobs ) k + k [NO ]
0
rxn
•
and are shown in Table 3. The values obtained for PhS (Vide
4
5
infra) are in good agreement with previous accounts.
•
Figure 4. Differential absorption spectrum of PhS after excitation of
It is known that the photolysis of DTBN goes by way of
C-N homolytic cleavage with formation of 2-methyl-2-nitroso-
propane and the tert-butyl radical (which is quickly trapped by
-
5
PhSSPh (5.5 × 10 M) in air-saturated acetonitrile (solid line).
•
Absorption spectrum of PhS corrected for ground state bleaching of
sulfide. Insert: Second order plot of the decay kinetics at 295 nm.
69
another mole of nitroxide). By contrast, TEMPO is known
to be an efficient hydrogen abstractor.7 In principle, such
radical intermediates themselves or the decomposition products
therefrom could interact with the sulfur-centered radicals under
investigation, especially at higher nitroxide concentrations.
However, the dependence of the sulfinyl radical decay rates on
the nitroxide concentration was linear and independent of laser
pulse energy, and such potential problems were neglected. Direct
excitation of the nitroxides in the absence of sulfoxides did not
result in observation of any transients from 260 to 600 nm.
A similar experiment was carried out with galvinoxyl in
acetonitrile. Only an upper limit for the rate constant of
0
Phenylsulfenyl Radicals. Phenylsulfenyl radicals (PhS•)
were produced by the direct photolysis of diphenyl disulfide.
In relatively inert solvents, the only significant decay pathway
42,45,47
is recombination, both geminate and random.
The absorp-
•
tion spectrum of PhS varied with solvent (Figures 4 and 5) but
was characterized by maxima at about 295 and 460 nm. The
microsecond-domain kinetics were strictly second order, and
neither the initial intensity nor the decay kinetics depended on
the presence of O2.
2
Assuming that the oscillator strength f is proportional to n(n
-
2
+ 2) ∫ꢀ dν (where n is the refractive index of the solvent)
•
•
71
•
interaction between PhS and PhSO radicals with galvinoxyl
and is a constant in various solvents, ꢀ values for PhS were
9
-1 -1
(
k e 1 × 10 M s ) could be established because the
obtained from the current data and previous reports from
4
3
galvinoxyl was destroyed by the laser pulse.
aqueous solution.
They are reported in Table 4. Also

Sulfinyl and Sulfenyl Radical Generation and Decay
J. Phys. Chem. A, Vol. 101, No. 37, 1997 6859
•
Figure 5. Differential absorption spectrum of PhS after excitation of
PhSSPh in air-saturated 2-propanol (solid line). The dashed-line
spectrum is corrected for bleaching of ground state PhSSPh.
•
Figure 6. Computed geometry for PhSO . Values in parentheses are
TABLE 4: Photophysical Properties of the Phenylsulfenyl
•
from Becke3LYP/6-31G+(d,p) calculations, and others are from ROHF/
Radical (PhS )
6
-31G+(d).
5
9
64
2
kr/ꢀ (10
ꢀ(295nm)
2kr (10
kdif(25 °C)
-
1
(10 M cm-1) M
4
-1
-1 -1
9
-1
s
-1) ΦPhS•
solvent
cm s
)
s
) (10 M
of the C-S bond. The planar form was a true minimum at all
acetonitrile
cyclohexane
11.0
6.5
4.8
2.7
0.70
0.90
0.67
0.67
7.7
5.9
3.2
1.8
19
7.4
3.2
1.5
1.1
1.3
1.2
0.85
computational levels.
•
Preliminary calculations on PhSO were carried out at the
2
2
-propanol
-methyl-2-
propanol
a
UHF/3-21G(d) level. These served as useful starting geometries
for other calculations. However, it was evident that the UHF
model would not be appropriate given the significant spin
a
The actual value may be slightly higher since 1% water was added
2
contamination obvious from an S value of 1.31. (When the
to the solvent.
phenyl group was substituted with a non-conjugating group such
as CH3, the spin contamination was much less.) Subsequent
ab initio calculations were done with the ROHF method, and
energies were reevaluated with single point calculations with
second order Møller-Plesset perturbation theory, truncated at
estimated were recombination rate constants and quantum yields
•
of PhS radicals in the bulk solvent (ΦPhS•). It is worthwhile
to note that the quantum yield of disulfide decomposition and
•
escape leading to formation of two fully separated PhS radicals
7
7
the second order (RMP2). The Becke3LYP hybrid density
functional was also used to optimize structures. Single point
energies were obtained with a larger basis set, as indicated
below.
in solvent equals ΦPhS•/2. Also, while the self-termination rate
•
•
constants 2kr are essentially identical for PhS and PhSO , the
ΦPhS• values are much larger than those of ΦPhSO•.
Computations. In order to further characterize the sulfinyl
radicals, computations were carried out to determine the
character of the singly-occupied molecular orbital (SOMO) of
•
The optimized geometry for PhSO at ROHF/6-31+G(d) is
shown in Figure 6, and several computed quantities are given
in Table 5. Including zero point energies, the barrier to rotation
•
PhSO . An extensive computational project regarding smaller
(i.e., the difference in energy between this conformation and
sulfinyl radicals and related species will be published else-
the transition state with a 90° dihedral angle) is 2 kcal/mol.
With the RMP2 electron correlation correction, this barrier rises
to 13 kcal/mol. The computed singly occupied molecular orbital
is essentially a π* S-O orbital, with minor delocalization (ca.
7
2
•
•
where.
Both HSO and CH3SO are of Cs symmetry. In
2
2
principle, the ground state could be either A′ or A′′, corre-
sponding to a σ-type or π-type singly occupied orbital. Our
CASSCF calculations on these species showed both ground
states to be of A′′ symmetry, consistent with previous theoreti-
1
%) on the ortho and para carbon atoms. Mulliken analysis of
the SOMO showed a 48% contribution from S p-orbitals and a
nearly identical 48% contribution from the O p-orbitals.
Similarly, Mulliken analysis places a +0.55 charge on the sulfur
and -0.52 on the oxygen. Interestingly, although the delocal-
ization of the SOMO into the phenyl ring is minimal, conjuga-
tion apparently draws spin density onto the sulfur atom, at least
at the ROHF level; the 90° transition state showed a 76%
contribution from the O atom and 24% from S. Mulliken
charges on S and O and the bond order are all slightly lower at
the transition state. A surprisingly large S-O bond lengthening
(0.06 Å) for the transition state is also observed at the ROHF
level.
2
8,73,74
26,75,76
2
•
cal
and experimental
determinations on HSO .
ROHF calculations also give a A′′ ground state in all cases,
and the singly occupied orbitals obtained from the CASSCF
calculations on the smaller systems were very similar to those
from ROHF. Thus, MCSCF calculations were not carried out
•
on PhSO when less expensive methods were expected to give
qualitatively similar results.
•
Three general conformations of PhSO were considered: a
C1 conformation, where the CSO plane is rotated to an arbitrary
degree relative to the phenyl plane, and two limiting Cs
conformations with the CSO plane parallel and perpendicular
to the phenyl plane. No stationary points were found that were
not essentially Cs in symmetry. (Shallow potentials allowed
the structure to “optimize” without coming to the exact 0 or
Using the Becke3LYP hybrid density functional method, the
structures were optimized with the 6-31+G(d,p) basis set, and
single point energies were done with the 6-311+G(3df,2p) basis
set. This basis set was chosen because of results with other,
smaller sulfinyl radicals that indicated it was necessary for good
9
0° dihedral angle.) The geometry with the CSO plane at a 90°
dihedral angle was found to be a transition state with a single
imaginary vibrational frequency whose vectors involved rotation
7
2
energies. The calculated rotational barriers are effectively

6
860 J. Phys. Chem. A, Vol. 101, No. 37, 1997
TABLE 5: Computed Parameters for PhSO•
ROHF/
Darmanyan et al.
MP2/6-31+G(d)//
ROHF/6-31+G(d)
Becke3LYP/
-31G+(d,p)
Becke3LYP/6-311G+(3df,2p)//
Becke3LYP/6-31G+(d,p)
quantity
6-31+G(d)
S-O bond length (Å)
1.49
0.48
0.48
2.0
1.53
0.49
0.41
4.4
a
spin density, S
0.52
0.39
5.0
a
spin density, O
rotational barrier (kcal/mol)
13.0
S-O bond length (transition state) (Å)
1.55
0.23
0.76
1.54
0.49
0.48
a
spin density, S (transtion state)
0.54
0.46
a
spin density, O (transition state)
a
Mulliken approximation.
identical for the two basis sets with Becke3LYPsabout 5 kcal/
mol. The Becke3LYP calculations show a S-O bond length
which varies much less with C-S bond rotation than in the
ROHF calculations but is much larger in the ground state
structure than for the ROHF calculation.
Like the ROHF calculations, Becke3LYP places the unpaired
electron in an orbital that is essentially S-O π*. The planar
structure has some delocalization (e10%) onto the ortho and
para positions on the ring, but this delocalization is severely
limited in the transition state, where the S-O π* orbital has no
overlap with the phenyl π system.
Among the compounds tested, the maximum quantum yield
of free PhSO radical production is observed for phenyl
•
diphenylmethyl sulfoxide, 7 (Table 2). Because of the time
scale of the experiment, the value of ΦPhSO• is actually Φcleave
times the fraction of radical pairs which escape the geminate
cage without some kind of reaction. The structural features of
4 are consistent with both favorable cleavage and steric
hindrance to recombination, and we are unable to separate these
aspects from the data quantitatively. However, the trend of
values of ΦPhSO• as a function of precursor is clearly in order
with expectations.
Because of the disagreement between S-O bond lengths
obtained at ROHF/6-31+G(d) and those at Becke3LYP/6-
There is also the matter of the value of ΦPhSO• as a function
of solvent (Table 1). Over a little more than 1 order of
magnitude variation in viscosity, the quantum yield drops by
about a factor of 2. The simplest notion correlating free radical
quantum yields and viscosity from a solvent continuum model
3
1+G(d,p) (1.49 Å Vs 1.53 Å), the planar structure was also
optimized at other levels. Several basis sets were examined
with the ROHF model. Moving from double-ú to triple-ú had
no significant effect; neither did adding p functions to the
hydrogens or removing the diffuse functions. However, there
was some variation with the number of d polarization functions.
With one, two, or three d functions on the heavy atoms, the
S-O bond length was found to be 1.49, 1.47, and 1.49 Å,
respectively. Using the BLYP functional (Vide infra) and the
predicts that Φ • should monotonically approach zero with
R
increasing η. However, this model neglects important specific
8
5
solvent-solute interactions. Given the computational results
that put a significant negative charge on the oxygen of the
sulfinyl radicals, it is likely, for instance, that they are solvated
much better by alcohols than by alkanes. It is thus interesting
to note that ΦPhSO• values are very similar in all of the solvents,
save the least viscous (η < 0.3 cP, higher ΦPhSO•) and most
viscous (η > 4 cP, lower ΦPhSO•), especially noting that the
more viscous solvents within the middle group are alcohols.
Specific solvation would be expected to increase the ΦPhSO•.
6
-31+G(d,p) basis set, an S-O bond length of 1.55 Å was
78,79
obtained.
Discussion
Role of Sulfinyl Radicals in Sulfoxide Photochemistry.
This work represents the strongest and most direct evidence
for the intermediacy of sulfinyl radicals in sulfoxide photo-
chemistry, as outlined in Scheme 1. The observation of the
same transient from precursors 5-8 and in the variety of
solvents leaves little doubt of the assignment of the transient
illustrated in Figure 1.
•
Computational Characterization of PhSO . Previous work-
•
ers have described PhS as having its spin essentially localized
on the S atom on a p-type orbital, with a single bond between
4
7
the carbon and sulfur. In a sense, the computational results
•
for PhSO are similar. Fundamentally, we find that the unpaired
spin lies almost entirely outside the ring in a π* orbital
constructed almost entirely of S and O p-orbitals (Table 5). Only
a trivial delocalization of the spin is observed into the phenyl
ring.
Typically, sulfenic esters that are formed by sulfoxide
photolysis undergo secondary photolysis to give sulfenyl/alkoxyl
radical pairs.1
3,23,80-84
In principle, then, a single intense laser
pulse could carry out two sequential photochemical reactions,
generating a sulfenyl radical for spectroscopic observations.
While we cannot rule out this occurrence to a minor extent, the
It has long been known that inclusion of polarization functions
in the basis set is necessary for accurate calculations of the
properties of hypervalent sulfur compounds such as sulfoxides.86
Given the results at the various basis sets, we believe 1.49 Å is
the best estimate of the ROHF value of the S-O bond length.
Experimental and computed bond lengths of typical sulfoxides
•
•
distinct spectral characteristics of the PhSO and PhS radicals
and the linear dependence of the signal strengths on laser pulse
energy seem to rule it out as a major problem with these
•
87
experiments. Only in the fluorescence of Ph2CH do we see
are 1.48-1.49 Å.
evidence of a two photon process of any sort.
Significant variation is observed between the ROHF,
Becke3LYP, and BLYP S-O bond lengths for PhSO (Figure 6
•
The current data are consistent with our previous steady state
results. For instance, from ref 23, it can be derived that the
quantum yield of PhSSO2Ph (the ultimate dimerization product
and BLYP/6-31+G(d,p) ) 1.55 Å). Comparison with the work
of Cramer and co-workers may shed light on this issue. They
carried out a series of computational studies on phosphorus-
containing radicals, some of which have considerable structural
•
of PhSO ) from photolysis of 8 (X ) H) is about 0.055. In
•
Table 1, the quantum yield of free PhSO is shown as 0.12.
•
• 88,89
Given a pure second order decay of the free PhSO , a
analogy to the present sulfinyl radicals (e.g., Me2PO ).
thiosulfonate quantum yield is predicted as 0.06. Since these
numbers are arrived at quite independently, their consistency
serves as a good check on the extinction coefficient value
Using the 6-31G(d,p) basis set, they found that UHF geometries
are more reliable than BLYP and other density functionals for
predicting hyperfine coupling constants, with the latter producing
P-X bonds longer than other methods by 0.02-0.06 Å. Even
•
determined for PhSO .

Sulfinyl and Sulfenyl Radical Generation and Decay
J. Phys. Chem. A, Vol. 101, No. 37, 1997 6861
•
the hyperfine coupling constants calculated with Becke3LYP
were more accurate when done at the UHF geometry, though
the difference is not as large as with other density functional
methods. Using this experience as a guide, one may suggest
that the 1.53 Å value is an upper limit, and the true length may
be closer to 1.49 Å, that is, about the same as, or a little longer
than, that of the sulfoxide.
previous results) and similarity to the PhSO values lend
confidence to the assignment.
The nature of the initial structure formed by sulfinyl dimer-
ization is not known for certain, but it has been assumed that it
is the mixed sulfenic-sulfinic anhydride 9, as opposed to the
In a previous computational work on sulfoxides in which
S-O bond dissociation energies were determined by isodesmic
exchange of the oxygen between a test sulfoxide and dimethyl
sulfide, it was found that certain structural features led to
significant differences in computed bond energies between
O-O or S-S bonded dimers.34,54 Whatever the initial structure,
the isolated structure for sulfinyl dimerization is a thiosulfonate,
in this case 10. There clearly is a second rearrangement which
occurs to generate these compounds. As alluded to previously,
steady state photolysis of 7 leads to product mixtures which
contain the appropriate thiosulfonate, particularly in less viscous
87
Hartree-Fock and Møller-Plesset methods. Moreover, Har-
tree-Fock methods are not generally reliable for predicting
transition state energies. Therefore, it is not surprising to see
•
a significant difference in the rotational barriers for PhSO
2
3,24
solvents.
calculated by ROHF and RMP2 methods, though the magnitude
of difference (2 kcal/mol Vs 13 kcal/mol) is large in this case.
The Becke3LYP data (ca. 5 kcal/mol) are intermediate. One
may speculate that very likely the ROHF number is too low,
and the true value may lie between the density functional and
ROMP2 results.
•
Finally, we return to the reactions of ArSO with other
molecules. Using steady state experiments and E/Z isomeriza-
tion of alkenes as a probe, previous workers have suggested
that sulfinyl radicals add reversibly to alkenes, but we are aware
of precious little other data regarding sulfinyl reactivity in
3
4,93
solution.
Intermolecular Reactivity of Sulfinyl Radicals. There are
three types of reactions reported here for sulfinyl radicals: that
with O2, that with themselves, and that with nitroxides. We
begin with the reaction with O2. An upper limit of ap-
We have obtained the first absolute rate constants for reactions
of sulfinyl radicals with other molecules, as reported in Table
3
. Unlike the self-termination reactions, these reactions, while
9
-1 -1
fast (ca. 10 M s ), are dinstinctly below the diffusion
controlled rate constant, and the rate constants do not vary much
with the range of substituents shown in the table. The nitroxides
are expected to behave as electron rich radicals, as illustrated
by the following resonance structures and by the well-known
increase of the nitrogen hyperfine coupling constant with solvent
polarity.
7
-1 -1
proximately 10 M
s
can be estimated for the apparent rate
•
constant for the reaction of O2 and PhSO . We make no claim
that PhSO is “inert” to O2, merely that it is effectively so under
•
these conditions because of the limited lifetime due to self-
termination. Furthermore, a rapidly reversible reaction between
•
PhSO and O2 is also consistent with our data.
•
The self-termination reaction of PhSO , however, is interesting
for its speed, rather than the lack thereof. Previous to this work,
two reports of sulfinyl radical self-termination reactions were
available; both used kinetic EPR to determine rate constants at
low temperature. The tert-butylsulfinyl radical dimerizes with
7
-1 -1
90
rate constant 2kr ) 6 × 10 M
s
at -100 °C. More
•
44
PhS , in its reactions both with nitroxides and with olefins,
directly comparable, 1 (X ) Cl) and its 2,5-dichloro analog
were studied over a temperature range of about 185-250 K in
toluene.91 The activation energy of the dimerization reaction
exhibits a polarity effect consistent with an electron poor
reactivity. This trend is evident in Table 3 for the different
substituents of 2. Whereas there is a range of about a factor of
(
1.7 and 2.4 kcal/mol) is consistent with a diffusion controlled
4
over the sulfenyl series from X ) Cl to OCH3, the trend,
9
process. The rate constants 2kr extrapolate up to 2 × 10 and
while in the same direction, is less than a factor of 2 in either
polar or nonpolar solvents for the sulfinyl series. Presumably
this is due to the presence of the negatively charged oxygen
atom; it can be speculated that the charge on S and O is
dominated by their mutual interaction and that the perturbation
by the remote substitutions is smaller.
9
-1 -1
1
× 10 M
s
at 298 K, respectively, in quite reasonable
agreement with the current results, particularly given the
extrapolation.
The results presented in Tables 1 and 4 show that an increase
in solvent viscosity leads to a decrease of the recombination
rate constant (2kr) of the sulfur-centered radicals. In 2-propanol
and 2-methyl-2-propanol 2kr ) kdif, where kdif is the diffusion
controlled rate constant. None of the 2kr values dip below 0.33
kdif. For radicals without steric hindrance and a low value of
spin-orbital coupling, 2kr should be 0.25kdif because of the spin
statistical effect of doublet recombination.92 Our results thus
point out that there is a significant spin-orbit coupling in radical
pairs of both types (sulfenyl and sulfinyl) of sulfur-centered
radicals investigated. Previous workers, who observed a lack
It is interesting to note that the rate constants for reactions
of nitroxides with ArS and ArSO radicals, though below
diffusion control, are higher than any of the published values
for similar sized carbon centered radicals.
•
•
94
Benzyl, for
instance, is trapped by TEMPO with a rate constant of 9.5 ×
7
-1 -1
1
0 M
s
in acetonitrile (at 18 °C), about an order of
• 95
magnitude more slowly than in PhSO . This difference is
contributed to by both electronic structure and steric hindrance.
We presume that the effect is not entirely steric, since the rate
constant for radical trapping by TEMPO decreases only by a
factor of 4 from the benzyl to the cumyl radical.
Although we did not attempt to isolate any products, we
assume the reaction of sulfinyls and nitroxides certainly proceeds
by coupling between the O atom of the nitroxide to the S and/
or O of the sulfinyl radical. This leads to an analogy between
the nitroxide and sulfinyl structures that may be useful in
supporting the presumed mode of sulfinyl dimerization.
•
of magnetic field effects for the ArS radical pairs in micelles,
came to similar conclusions for that radical.46
•
Our inclusion of data for PhS is mainly for comparison to
PhSO . Our estimates of the extinction coefficients for the
absorption spectra in various solvents depend on the previous
assignment of the spectrum in water.43 However, since the rate
constants depend on the chosen value of ꢀ, their proper
magnitude (given the expectations for spin-orbit coupling from
•

6862 J. Phys. Chem. A, Vol. 101, No. 37, 1997
Darmanyan et al.
The nitroxides are at least qualitatively isolobal with the
(17) Kowalewski, R.; Margaretha, P. HelV. Chim. Acta 1993, 76, 1251-
1
257.
sulfinyl radicals (see resonance structures above), and the
(18) Kato, H.; Arikawa, Y.; Hashimoto, M.; Masuzawa, M. J. Chem.
9
6
unpaired spin is expected to reside in an N-O π* orbital.
Soc., Chem. Commun. 1983, 938-938.
Preliminary calculations confirm this. However, nitroxides can
only react at the partially negative oxygen center for energetic
and steric reasons. Therefore, the coupling between the
nitroxide and the sulfinyl can be seen as a model for the sulfinyl
self-termination reaction in which one of the sulfinyls is
constrained to react at its O atom. Given that the rate constants
for sulfinyl-nitroxide reaction are within less than an order of
magnitude of the dimerization rate constants (and considering
the steric bulk of the nitroxide) and the comparable dipole
relationship of the N-O and S-O bonds, we suggest that these
data support the mode of reaction shown above.
(19) Kellogg, R. M.; Prins, W. L. J. Org. Chem. 1974, 39, 2366-2374.
(20) Thiemann, C.; Thiemann, T.; Li, Y.; Sawada, T.; Nagano, Y.;
Tashiro, M. Bull. Chem. Soc. Jpn. 1994, 67, 1886-1893.
(21) Larson, B. S.; Kolc, J.; Lawesson, S.-O. Tetrahedron 1971, 27,
5163-5176.
(22) Kropp, P. J.; Fryxell, G. E.; Tubergen, M. W.; Hager, M. W.; Harris,
G. D., Jr.; McDermott, T. P., Jr.; Tornero-Velez, R. J. Am. Chem. Soc.
1
991, 113, 7300-7310.
23) Guo, Y.; Jenks, W. S. J. Org. Chem. 1995, 60, 5480-5486.
(24) Guo, Y.; Jenks, W. S. J. Org. Chem. 1997, 62, 857-864.
25) Tyndall, G. S.; Ravishankara, A. R. Int. J. Chem. Kin. 1991, 23,
83-527.
26) Schurath, U.; Weber, M.; Becker, K. H. J. Chem. Phys. 1977, 67,
(
(
4
(
110-119.
(27) Trunipseed, A. A.; Barone, S. B.; Ravishankara, A. R. J. Phys.
Chem. 1993, 97, 5926-5934.
Summary
(
28) Xantheas, S. S.; Dunning, T. H., Jr. J. Phys. Chem. 1993, 97, 18-
19.
The transient absorption spectrum of arylsulfinyl radicals has
been observed, confirming with direct evidence the photochemi-
cal pathway of R-cleavage for aromatic sulfoxides. A deter-
mination of extinction coefficients for the parent phenylsulfinyl
radical was made. It was found that, though kinetically
unreactive to molecular oxygen, the sulfinyl radical self-
quenches with rate constants quite closely approaching the
diffusion controlled limit, implying that large spin-orbit
couplings overcome typical spin-statistical concerns. Reactivity
with nitroxide radicals is very high but is not as variable as
that of corresponding arenesulfenyl radicals when substituted
in the para position. Computational studies indicate that the
singly occupied orbital is largely localized on S and O in a π*
configuration lying most heavily on O.
(
(
29) Wang, N. S.; Howard, C. J. J. Phys. Chem. 1990, 94, 8787-8794.
30) Friedl, R. R.; Brune, W. H.; Anderson, J. G. J. Phys. Chem. 1985,
89, 5505-5510.
31) Wang, N. S.; Lovejoy, E. R.; Howard, C. J. J. Phys. Chem. 1987,
1, 5743-5749.
32) Lovejoy, E. T.; Wang, N. S.; Howard, C. J. J. Phys. Chem. 1987,
(
9
(
5749-5755.
(
Jenks, W. S. J. Chem. Phys. 1997, 106, 86-93.
(
Patai, S., Rappoport, Z., Stirling, C. J. M., Eds.; John Wiley & Sons Ltd.:
New York, 1988; pp 1081-1087.
(
33) Zhao, H.-Q.; Cheung, Y.-S.; Heck, D. P.; Ng, C. Y.; Tetzlaff, T.;
34) Chatgilialoglu, C. In The Chemistry of Sulfones and Sulfoxides;
35) Chatgilialoglu, C.; Gilbert, B. C.; Gill, B.; Sexton, M. D. J. Chem.
Soc., Perkin Trans. 2 1980, 1141-1150.
36) Gilbert, B. C.; Kirk, C. M.; Norman, O. C.; Laue, H. A. H. J. Chem.
Soc., Perkins Trans. 2 1977, 497-501.
37) Gilbert, B. C.; Gill, B.; Sexton, M. D. J. Chem. Soc., Chem.
Commun. 1978, 78-79.
38) Khait, I.; L u¨ dersdorf, R.; Muszkat, K. A.; Praefcke, K. J. Chem.
(
(
(
Acknowledgment. We gratefully acknowledge financial
support from the National Science Foundation (Grant CHE 94-
Soc., Perkins Trans. 2 1981, 1417-1429.
(39) Muszkat, K. A.; Praefcke, K.; Khait, I.; L u¨ dersdorf, R. J. Chem.
Soc., Chem. Commun. 1979, 898-899.
1
2964) and the Research Corporation. Y.G. is the recipient of
(40) L u¨ dersdorf, R.; Khait, I.; Muszkat, K. A.; Praefcke, K.; Margaretha,
a Dow Chemichal fellowship at Iowa State University and
expresses his gratitude for their support. D.D.G. gratefully
acknowledges a GAANN fellowship, administered through Iowa
State University. We are also most grateful to Prof. Mark
Gordon and Dr. Mike Schmidt for continuing assistance with
the computational chemistry.
P. Phosphorus Sulfur 1981, 12, 37-54.
(41) Thyrion, F. C. J. Phys. Chem. 1973, 77, 1478-1482.
(
42) Scott, T. W.; Liu, S. N. J. Phys. Chem. 1989, 93, 1393-1396.
(43) Bonifacic, M.; Weiss, J.; Chandhri, S. A.; Asmus, K.-D. J. Phys.
Chem. 1985, 89, 3910-3914.
(44) Ito, O. Res. Chem. Intermed. 1995, 21, 69-93.
(
45) Nakamura, M.; Ito, O.; Matsuda, M. J. Am. Chem. Soc. 1980, 102,
6
98-701.
(46) Jeschke, G.; Wakasa, M.; Sakaguchi, Y.; Hayashi, H. J. Phys. Chem.
References and Notes
1
994, 98, 4069-4075.
(
47) Tripathi, G. N. R.; Sun, Q.; Armstrong, D. A.; Chipman, D. M.;
(
(
(
1) Photochemistry and Photophysics of Aromatic Sulfoxides. 8.
2) Block, E. Q. Rep. Sulfur Chem. 1969, 4, 315-319.
3) Still, I. W. J. In The Chemistry of Sulfones and Sulfoxides; Patai,
Schuler, R. H. J. Phys. Chem. 1992, 96, 5344-5350.
(48) Balicki, R.; Kaczmarek, L.; Nantka-Namirski, P. Liebigs Ann. Chem.
1992, 883-884.
S., Rappaport, Z., Stirling, C. J. M., Eds.; John Wiley & Sons Ltd.: New
(49) Finzi, C.; Bellavita, V. Gazz. Chim. Ital. 1933, 62, 699-708.
(50) Schmidt, M. W.; Baldridge, K. K.; Boatz, J. A.; Elbert, S. T.;
Gordon, M. S.; Jensen, J. H.; Koseki, S.; Matsunaga, N.; Nguyen, N.; Su,
S. J.; Windus, T. L.; Dupuis, M.; Montgomery, J. A. J. Comput. Chem.
1993, 14, 1347-1363.
York, 1988; pp 873-887.
(4) Jenks, W. S.; Gregory, D. D.; Guo, Y.; Lee, W.; Tetzlaff, T. In
Molecular and Supramolecular Photochemistry, Vol. 1, Organic Photo-
chemistry; Ramamurthy, V., Schanze, K. S., Eds.; Marcel Dekker Inc.: New
York, 1997; pp 1-56.
(51) Becke, A. D. J. Chem. Phys. 1993, 98, 5648-5652.
(52) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.;
Johnson, B. G.; Wong, M. W.; Foresman, J. B.; Robb, M. A.; Head-Gordon,
M.; Replogle, E. S.; Gomperts, R.; Andres, J. L.; Raghavachari, K.; Binkley,
J. S.; Gonzalez, C.; Martin, R. L.; Fox, D. J.; Defrees, D. J.; Baker, J.;
Stewart, J. J. P.; Pople, J. A. Gaussian 92; Gaussian, Inc.: Pittsburgh, PA,
(
(
5) Still, I. W. J.; Thomas, M. T. Tetrahedron Lett. 1970, 4225-4228.
6) Still, I. W. J.; Cauhan, M. S.; Thomas, M. T. Tetrahedron Lett.
1
973, 1311-1314.
7) Still, I. W. J.; Arora, P. C.; Chauhan, M. S.; Kwan, M.-H.; Thomas,
M. T. Can. J. Chem. 1976, 54, 455-470.
(
1
993; pp.
(
8) Kobayashi, K.; Mutai, K. Tetrahedron Lett. 1981, 22, 5201-5204.
9) Kobayashi, K.; Mutai, K. Phosphorus Sulfur 1985, 25, 43-51.
(53) McLean, A. D.; Chandler, G. S. J. Chem. Phys. 1980, 72, 5639-
(
5
648.
(
10) Kobayashi, K.; Mutai, K. Chem. Lett. 1983, 1461-1462.
(54) Chatgilialoglu, C. In Handbook of Organic Photochemistry; Sca-
(11) Kimura, T.; Ishikawa, Y.; Ueki, K.; Horie, Y.; Furukawa, N. J.
iano, J. C., Ed.; CRC Press: Boca Raton, FL, 1989; Vol. 2, pp 3-11.
Org. Chem. 1994, 59, 7117-7124.
(
55) Huggenberger, C.; Fischer, H. HelV. Chim. Acta 1981, 338-353.
(
12) Gajurel, C. L. Indian J. Chem. B 1986, 25, 319-320.
(56) Meisel, D.; Das, P. K.; Hug, G. L.; Bhattacharyya, K.; Fessenden,
(13) Schultz, A. G.; Schlessinger, R. H. J. Chem. Soc., Chem. Commun.
R. W. J. Am. Chem. Soc. 1986, 108, 4706-4710.
1
970, 1294-1295.
14) Schultz, A. G.; DeBoer, C. D.; Schlessinger, R. H. J. Am. Chem.
Soc. 1968, 90, 5314-5315.
15) Schultz, A. G.; Schlessinger, R. H. J. Chem. Soc., Chem. Commun.
969, 1483-1484.
16) Capps, N. K.; Davies, G. M.; Hitchcock, P. B.; McCabe, R. W.;
Young, D. W. J. Chem. Soc., Chem. Commun. 1983, 199-200.
(57) Christensen, H. C.; Sehested, K.; Hart, E. J. J. Phys. Chem. 1973,
(
77, 983-987.
(58) Hiratsuka, H.; Okamura, T.; Tanaka, I.; Tanizaki, Y. J. Phys. Chem.
1980, 84, 285-289.
(
1
(59) Maillard, B.; Ingold, K. U.; Scaiano, J. C. J. Am. Chem. Soc. 1983,
105, 5095-5099.
(
(60) Bromberg, A.; Meisel, D. J. Phys. Chem. 1985, 89, 2507-2513.

Sulfinyl and Sulfenyl Radical Generation and Decay
J. Phys. Chem. A, Vol. 101, No. 37, 1997 6863
(
61) Bromberg, A.; Schmidt, K. H.; Meisel, D. J. Am. Chem. Soc. 1985,
07, 83-91.
62) Bromberg, A.; Schmidt, K. H.; Meisel, D. J. Am. Chem. Soc. 1984,
06, 3056-3057.
63) Scaiano, J. C.; Tanner, M.; Weir, D. J. Am. Chem. Soc. 1985, 107,
396-4403.
64) Murov, S. L.; Carmichael, I.; Hug, G. L. Handbook of Photochem-
istry, 2nd ed.; Marcel Dekker, Inc.: New York, 1993.
(80) Pasto, D. J.; Hermine, G. L. J. Org. Chem. 1990, 55, 5815-5816.
(81) Pasto, D. J.; L’Hermine, G. Tetrahedron 1993, 49, 3259-3272.
(82) Pasto, D. J.; Cottard, F. Tetrahedron Lett. 1994, 35, 4303-4306.
(83) Pasto, D. J.; Cottard, F. J. Org. Chem. 1994, 59, 4642-4646.
1
1
4
(
(
(84) Pasto, D. J.; Cottard, F.; Picconatto, C. J. Org. Chem. 1994, 59,
7
172-7177.
(
(
85) Booth, D.; Noyes, R. M. J. Am. Chem. Soc. 1960, 82, 1868-1872.
(86) Amato, J. S.; Karady, S.; Reamer, R. A.; Schlegel, H. B.; Springer,
(
65) Ware, W. R. J. Phys. Chem. 1962, 66, 455-458.
J. P.; Weinstock, L. M. J. Am. Chem. Soc. 1982, 104, 1375-1380.
87) Jenks, W. S.; Matsunaga, N.; Gordon, M. J. Org. Chem. 1996, 61,
275-1283.
88) Lim, M. H.; Worthington, S. E.; Dulles, F. J.; Cramer, C. J. In
(66) Birks, J. B. Photophysics of Aromatic Molecules; Wiley-Inter-
(
science: New York, 1970.
67) Bensasson, R.; Land, E. J. Trans. Faraday Soc. 1971, 67, 1904-
915.
1
(
(
1
ACS Symposium Series 629; Laird, B. B., Ross, R. B., Eds.; American
Chemical Society: Washington, D.C., 1996; pp 402-422.
(
(
(
68) Amand, B.; Bensasson, R. Chem. Phys. Lett. 1975, 34, 44-48.
69) Anderson, D. R.; Koch, T. H. Tetrahedron Lett. 1977, 3015-3018.
(89) Cramer, C. J.; Lim, M. H. J. Phys. Chem. 1994, 98, 5024-5033.
(90) Howard, J. A.; Furimsky, E. Can. J. Chem. 1974, 52, 555-556.
(91) Bennett, J. E.; Brunton, G.; Gilbert, B. C.; Whittall, P. E. J. Chem.
70) Johnston, L. J.; Tencer, M.; Scaiano, J. C. J. Org. Chem. 1986, 51,
2
806-2808.
71) Calvert, J. C.; Pitts, J. N. Photochemistry; John Wiley & Sons,
Ltd.: New York, 1966.
(
Soc., Perkin Trans. 2 1988, 1359-1364.
(
92) Khudyakov, I. V.; Serebrennikov, Y. A.; Turro, N. J. Chem. ReV.
(
(
(
(
72) Gregory, D. D.; Jenks, W. S. Manuscript in preparation.
73) Plummer, P. L. M. J. Chem. Phys. 1990, 92, 6627-6634.
74) Luke, B. T.; McLean, A. D. J. Phys. Chem. 1985, 89, 4592-4596.
75) Endo, Y.; Saito, S.; Hirota, E. J. Chem. Phys. 1981, 75, 4379-
1
993, 93, 537-570.
(
(
93) Iino, M.; Matsuda, M. J. Org. Chem. 1983, 48, 3108-3109.
94) Bowry, V. W.; Ingold, K. U. J. Am. Chem. Soc. 1992, 114, 4992-
4
996.
4
4
384.
(
72.
76) White, J. N.; Gardiner, W. C., Jr. Chem. Phys. Lett. 1978, 58, 470-
(95) Beckwith, A. L. J.; Bowry, V. W.; Ingold, K. U. J. Am. Chem.
Soc. 1992, 114, 4983-4992.
(
77) Knowles, P. J.; Andrews, J. S.; Amos, R. D.; Handy, N. C.; Pople,
(96) Aurich, H. G. In Nitrones, Nitronates, and Nitroxides; Patai, S.,
Rappoport, Z., Eds.; Wiley & Sons, Ltd.: New York, 1989.
(97) Weast, R. C.; Astle, M. J. CRC Handbook of Chemistry and Physics;
CRC Press, Inc.: Boca Raton, FL, 1996.
J. A. Chem. Phys. Lett. 1991, 186, 130-136.
(
78) Becke, A. D. Phys. ReV. A 1988, 38, 3098-3100.
(79) Lee, C.; Yang, W.; Parr, T. G. Phys. ReV. B 1988, 37, 785-789.