Photoinduced hydrogen- and electron-transfer processes between chloranil and aryl alkyl sulfides in organic solvents. Steady-state and time-resolved studies

Article abstract of DOI:10.1039/a909372i

The photochemical behavior of three aryl alkyl sulfides, thioanisole (TA), benzyl phenyl sulfide (BPS) and 4-methoxybenzyl phenyl sulfide (MBPS), sensitized by triplet chloranil (CA), was investigated by nanosecond laser flash photolysis and steady-state irradiation in CH₂Cl₂ and MeCN. The nature of the transients detected upon 355-nm laser excitation was independent of the molecular structure of the aryl alkyl sulfides but strongly affected by the solvent polarity. In particular, in CH₂Cl₂ the quenching process of triplet CA by aryl alkyl sulfides was accompanied by H- transfer, with formation of the CAH· and TA(-H)·/BPS(-H)·/MBPS(-H)· radicals. In contrast, a charge transfer process between triplet CA and aryl alkyl sulfides, with formation of the radical anion CA·^- and radical cations of aryl alkyl sulfides, occurred in MeCN. In this solvent, a transient detected at long delay time was tentatively assigned to the anion CAH^- formed by H-transfer between radical ions. In all experiments, transient species were characterized in terms of second-order decay rate constants and quantum yields of formation. Steady-state irradiation of the CA/TA system led to the stable photoadduct C₆H₅SCH₂OC₆Cl₄OH in both CH₂Cl₂ and MeCN with quantum yields of 0.033 and 0.27, respectively. In contrast, aldehydes, thioacetals, and disulfides were the main products obtained upon irradiation of the CA/BPS and CA/MBPS systems. The photoaddition products were not observed, probably owing to their low stability. The nature of the photoproducts formed by irradiation of CA/aryl alkyl sulfides was independent of solvent properties, even though the reactivity was higher in MeCN than in CH₂Cl₂.

Full text of DOI:10.1039/a909372i

View Article Online / Journal Homepage / Table of Contents for this issue
Photoinduced hydrogen- and electron-transfer processes between
chloranil and aryl alkyl sulÐdes in organic solvents. Steady-state and
time-resolved studies
Tiziana Del Giacco,a Fausto Elisei*a and Osvaldo Lanzalungab
a Dipartimento di Chimica, Universita di Perugia, V ia Elce di Sotto 8, 06123 Perugia, Italy.
E-mail: elisei=phch.chm.unipg.it
b Dipartimento di Chimica and Centro CNR di Studio sui Meccanismi di Reazione, Universita L a
Sapienza, Piazzale Aldo Moro 5, 00185, Roma, Italy
Received 29th November 1999, Accepted 14th February 2000
Published on the Web 17th March 2000
The photochemical behavior of three aryl alkyl sulÐdes, thioanisole (TA), benzyl phenyl sulÐde (BPS) and
4-methoxybenzyl phenyl sulÐde (MBPS), sensitized by triplet chloranil (CA), was investigated by nanosecond
laser Ñash photolysis and steady-state irradiation in CH Cl and MeCN. The nature of the transients detected
2
2
upon 355-nm laser excitation was independent of the molecular structure of the aryl alkyl sulÐdes but strongly
a†ected by the solvent polarity. In particular, in CH Cl the quenching process of triplet CA by aryl alkyl
2
2
sulÐdes was accompanied by H-transfer, with formation of the CAH~ and TA(-H)~/BPS(-H)~/MBPS(-H)~
radicals. In contrast, a charge transfer process between triplet CA and aryl alkyl sulÐdes, with formation of the
radical anion CA~~ and radical cations of aryl alkyl sulÐdes, occurred in MeCN. In this solvent, a transient
detected at long delay time was tentatively assigned to the anion CAH~ formed by H-transfer between radical
ions. In all experiments, transient species were characterized in terms of second-order decay rate constants and
quantum yields of formation. Steady-state irradiation of the CA/TA system led to the stable photoadduct
C H SCH OC Cl OH in both CH Cl and MeCN with quantum yields of 0.033 and 0.27, respectively. In
6
5
2
6
4
2 2
contrast, aldehydes, thioacetals, and disulÐdes were the main products obtained upon irradiation of the
CA/BPS and CA/MBPS systems. The photoaddition products were not observed, probably owing to their low
stability. The nature of the photoproducts formed by irradiation of CA/aryl alkyl sulÐdes was independent of
solvent properties, even though the reactivity was higher in MeCN than in CH Cl .
2
2
when the counter ion has a relatively low basicity,9 a homo-
1
Introduction
lytic C ÈH bond cleavage is likely to occur in the interaction
a
Organic sulfur compounds are known to undergo fast one-
electron oxidation reactions, owing to their relatively low ion-
ization potentials. Thus, suitable oxidants can remove an
electron from a lone pair on the sulfur atom, the likely site of
the oxidative attack, to produce molecular radical cations,
important intermediates in a great variety of chemical pro-
cesses, extending from those of industrial importance to the
biological ones.1h3 Most investigations have been carried out
to provide information on the main reaction pathways of
dialkyl sulÐde radical cations, while much less attention has
been given to radical cations from aromatic sulÐdes.1a,4h8
The radical cations of sulÐdes decay through competitive
pathways whose relative rate constants depend strongly on
the structure of the substrate and on experimental conditions
(solvent polarity, additives, etc.). In particular, deprotonation
at a C ÈH bond, CÈS fragmentation, oxidation, aromatic sub-
stitution, and dimerization have been investigated after gener-
ation of the radical cation by photoinduced electron transfer,
radiolysis or chemical and electrochemical oxidation.1h8
The aim of this work was a detailed investigation of the
reactivity and photokinetic behavior of three aryl alkyl
sulÐdes sensitized by triplet chloranil in organic solvents of
di†erent polarity at room temperature. Nanosecond time-
resolved measurements and steady-state irradiations were
carried out to identify the transient species produced upon
light absorption, their decay pathways, and Ðnal photo-
products. As recently shown for alkyl aromatic radical cations
between the chloranil radical anion and aromatic radical
cations of sulÐdes.
2
Experimental
2.1 Starting materials
Chloranil (CA, Fluka) was recrystallized from acetone and
sublimed under vacuum; thioanisole (TA, Aldrich), was dis-
tilled; benzyl phenyl sulÐde (BPS, Aldrich) was recrystallized
from EtOHÈH O; 4-methoxybenzyl phenyl sulÐde (MBPS)
2
was prepared by reaction of 4-methoxybenzyl chloride with
thiophenol as previously described.10 The structures of TA,
BPS and MBPS are shown in Scheme 1. MeCN and CH Cl
2
2
were Aldrich solvents of spectrophotometric grade, used
as received. Tetrachlorohydroquinone (Sigma) and
Me NOH Æ 5H O (Aldrich) were used as received.
a
4
2
2.2 Laser Ñash photolysis
The excitation wavelength of 355 nm from a Nd: YAG laser
(Continuum, third harmonic) was used in nanosecond Ñash
photolysis experiments (pulse width ca. 7 ns and energy \3
mJ per pulse).11 The transient spectra were obtained by a
point-to-point technique, monitoring the absorbance changes
(*A) after the Ñash at intervals of 5È10 nm over the spectral
range 300È850 nm, averaging at least 10 decays at each wave-
DOI: 10.1039/a909372i
Phys. Chem. Chem. Phys., 2000, 2, 1701È1708
This journal is ( The Owner Societies 2000
1701

View Article Online
Scheme 1
length. The t values (the time at which the initial signal is
halved) are reported for transients showing second-order
internal standard. Light intensity was measured by use of pot-
assium ferrioxalate actinometry.14 The photoreaction
quantum yields were averages of at least three independent
experiments (mean deviation ca. 7%).
For BPS and MBPS, quantum yields could not be mea-
sured because the relative amount of photoproducts depends
on irradiation time, even at small conversion percentages.
Ground state absorption spectra were recorded by a
UVKON 923 double-beam spectrophotometer.
1@2
kinetics. The quantum efficiencies of transient formation were
measured by calibrating the experimental set-up with optically
matched solutions of benzophenone (B) in MeCN. The follow-
ing relationship between quantum yields (/), absorption coef-
Ðcients (e), and absorbance changes (*A), measured at the
corresponding absorption maxima of the transient (Tr) and
the triplet benzophenone [/ (B) \ 1 and e (B) \ 6500 M~1
T
T
cm~1 at 520 nm],12 was used to obtain transient quantum
yields:
3
Results and discussion
e (B) *A
T
Tr
/
\ / (B)
(1)
Tr
T
e
*A(B)
Tr
The experimental errors on t were estimated to be ^10%
3.1 Laser Ñash photolysis
1@2
and those on e and / to be ^15%.
Tr
Tr
The lowest triplet state of CA was efficiently quenched by aryl
alkyl sulÐdes in both CH Cl and MeCN; decay rate con-
All solutions were Ñowed through a quartz photolysis cell
while argon was bubbling through them. All measurements
were carried out at 22 ^ 2 ¡C.
2 2
measured at di†erent TA concentrations are
shown in Fig. 1. From the linear plots of k
stants k
obsd
vs. concentra-
obsd
tion of aryl alkyl sulÐdes, quenching rate constants k close to
q
the di†usional limit (k \ 0.49È1.32 ] 1010 M~1 s~1, Table 1)
q
2.3 Photochemical reactions
were obtained. It has to be noted that, within experimental
A solution of CA (2 ] 10~2 M), the TA, BPS and MBPS sub-
strates (2 ] 10~2 M) in CD CN or CH Cl , contained in a
error, the quenching process of triplet CA in CH Cl did not
show any isotope e†ect, at least for TA. In fact, k obtained
with both TA and PhSCD was 1.1 ] 1010 M~1 s~1; hence,
2
2
3
2 2
Pyrex vessel, was capped with a rubber septum, Ñushed with
argon and irradiated without cooling at the instrumental
q
3
operating temperature in a Rayonet photoreactor (j
\
exc
310È390 nm). After irradiation for 60 min (TA) or 30 min
(BPS and MBPS), an internal standard was added and the
raw photolysate was analyzed by GLC, GC-MS and
1H-NMR (directly, for experiments in CD CN, or after evapo-
3
ration of the solvent and addition of CD CN, for experiments
3
in CH Cl ). All products formed were identiÐed by compari-
2
2
son
with
authentic
specimens:
benzaldehyde,
4-
methoxybenzaldehyde and diphenyl disulÐde were commercial
products; benzaldehyde diphenyldithioacetal and 4-
methoxybenzaldehyde diphenyldithioacetal were synthesized
according to literature procedures.13 The adduct
PhSCH OC Cl OH formed by CA and TA was isolated from
2
6 4
the oxidation system on a preparative scale and characterized
on the basis of the following spectral data: MS m/z (rel.
intensity) M` 370,123(100),77,45; 1H-NMR d 7.5È7.3 (m, 5H,
ArH), 5.96 (s, 1H, OH), 5.51 (s, 2H, CH ). Quantitative
2
analysis of reaction products was performed by GLC and
1H-NMR. A good mass balance was obtained in all experi-
ments ([90%). Prolonged irradiation of BPS and MBPS led
to a decrease in the [dithioacetal] : [benzaldehyde] ratio of
the two compounds.
Fig. 1 First-order decay rate constant (k ) of 3CA* in CH Cl (|)
obsd
2 2
and MeCN (…) at di†erent TA concentrations, recorded at 520 nm
(j \ 355 nm).
exc
2.4 Quantum yields
Table 1 Quenching rate constants (k ) of triplet CA by TA, BPS and
A 2.5 ml solution of CA (8.5 ] 10~3 M in MeCN and
q
MBPS in CH Cl and MeCN
1.0 ] 10~2
M in CH Cl ) and TA (2 ] 10~3 and/or
2
2
2
2
1.0 ] 10~2 M), de-aerated by bubbling with argon, was placed
Compound
Solvent
k /1010 M~1 s~1
q
in a quartz cuvette and irradiated with a high pressure Hg
TA
CH Cl
MeCN
1.10
0.49
1.32
0.65
1.07
0.92
lamp (j \ 366 nm, selected by a Balzer interference Ðlter).
2
2
exc
To minimize the inner Ðlter, the conversion to photoproduct
BPS
MBPS
CH Cl
MeCN
was kept below 10%. To obtain the percentage of conversion,
the solvent of the irradiated solution was evaporated and the
2
2
2
CH Cl
2
residue was dissolved in CDCl before analysing it by
MeCN
3
1H-NMR spectrometry and/or GC. Bibenzyl was used as an
1702
Phys. Chem. Chem. Phys., 2000, 2, 1701È1708

View Article Online
for the radical CAH~ in CH Cl was
the main bimolecular deactivation process of triplet CA does
not involve a H-transfer process, even if the isotope e†ect
could be hidden by high quenching rate constants.
The product e
/
Rad Rad
2 2
obtained by a comparison between the absorbance changes
(*A) recorded at 435 nm (maximum absorption for CAH~)
upon laser excitation of CA in the presence of aryl alkyl sul-
Ðdes, and that recorded at 520 nm (maximum tripletÈtriplet
absorption) for an optically matched solution of benzophe-
none in MeCN (see experimental section). A value of 0.06 was
obtained for quantum yields of formation of the radical CAH~,
Laser photolysis of oxygen-free CH Cl solutions of CA
2
2
(4.0 ] 10~3 M) was carried out by exciting at 355 nm in the
presence of aryl alkyl sulÐdes (1.0 ] 10~2 M). Under these
experimental conditions, the lifetime of triplet CA (j \ 520
max
nm)15 was too short to be detected (q O 10 ns). Indeed, small
signals were recorded just after the laser pulse with absorption
maxima at 340È360, 420(sh) and 435 nm (see Fig. 2 for the
CA/TA system), very similar to those previously reported for
CAH~,16,17 TA(-H)~,7 and the radicals PhCH~SPhCH SO ~
by use of the absorption coefficient e \ 7300 M~1 cm~1
Rad
(ref. 16 and 18) for CAH~ at 435 nm. The quantum yields of
formation of the radicals TA(-H)~, BPS(-H)~ and MBPS(-H)~
(/ ) were obviously taken to be equal to those of CAH~. It
2
3
Rad
and 4-MeOPhCH~SPhCH SO ~.7 Taking also into account
has to be noted that /
is rather low and remains constant
2
3
Rad
the small e†ect of the ÈCH SO ~ group on the absorption
for the three systems investigated. In fact, also the rate con-
stant of radical formation (k D 7 ] 108 M~1 s~1) does not
change with the structure of aryl alkyl sulÐdes (see below).
These results are likely related to a transition state of the CÈH
homolytic cleavage which is close to the structure of the
reagents. The possibility that, in the presence of high concen-
trations of aryl alkyl sulÐdes, the radical CAH~ is formed by
H-abstraction from the solvent is ruled out by the very short
lifetime of triplet CA (q \ 10 ns to be compared with q B 3
2
3
spectra of similar radicals,7 the transient species were assigned
to radicals produced by a fast H-transfer process from aryl
alkyl sulÐdes to CA within a short-lived triplet complex.
The properties of the radicals CAH~, TA(-H)~, BPS(-H)~ and
MBPS(-H)~ are summarized in Table 2. The absorption spec-
trum of TA(-H)~ is centered at ca. 340 nm (Fig. 2), to be com-
pared with the j
of 330 nm previously reported for the
same transient in water,7 while the main absorption bands of
max
T
T
BPS(-H)~ and MBPS(-H)~ are overlapped with the Ðrst
ls in the absence of aryl alkyl sulÐdes) and by the presence of
the partner radicals [TA(-H)~, BPS(-H)~ and MBPS(-H)~]
which absorb at 340È360 nm.
absorption of CAH~ (j \ 360 nm).
max
The *A values measured at j
for TA(-H)~, BPS(-H)~ and
MBPS(-H)~ (340 or 360 nm, depending on the sulÐde) were
max
used to obtain the absorption coefficient e
by use of eqn (2):
of each radical
Rad
*A
e /
j
T
T
[ e (CAH~) ] e (CA)
e
\
(2)
Rad
j
G
*A (B) /
520
Rad
where *A (B) is the absorbance change recorded at 520 nm
520
for the optically matched solution of benzophenone (B) in ace-
tonitrile, the product e / \ 6500 M~1 cm~1 refers to the
T
T
tripletÈtriplet absorption of B at 520 nm,12 e (CAH~) is the
j
absorption coefficient of CAH~ at the wavelength of analysis
(equal to 2400 and 7800 M~1 cm~1 at 340 and 360 nm,
respectively)19 and e (CA) is the absorption coefficient of
G
chloranil in the ground state. It has to be noted that at these
wavelengths e (CA) B 200 M~1 cm~1 and, therefore, the
G
e†ect of ground state absorption on e
mental error (^15%).
is within the experi-
Fig. 2 Time-resolved absorption spectra obtained by laser photoly-
sis of 4.0 ] 10~3 M CA in the presence of 1.0 ] 10~2 M TA in
CH Cl , recorded 0.22 (L), 1.9 (|) and 8.0 ()) ls after the laser pulse
Rad
The e
values measured in this work (in the range 4400È
Rad
9000 M~1 cm~1, see Table 2) are similar to those previously
2
2
(j \ 355 nm). Inset: changes in absorbance recorded at 435 nm.
exc
Table 2 Spectral and kinetic properties and quantum yields of the transients sensitized by triplet CAa in the presence of aryl alkyl sulÐdesb
Compound
TA
Solvent
CH Cl
j
/nm
t
c/ls
1@2
14
13
2k /1010 M~1 s~1
ed/M~1 cm~1
/
Transient
max
340
2
Tr
1.1
1.7
3.1
3.8
3.5
1.3
1.7
3.4
3.8
3.5
0.34
0.23
3.4
4.5
3.8
4400
7300e
9700f
6100g
13500
8200
0.06
0.06
1.0
PhSCH~
2
2
2
360, 420(sh), 435
325, 430(sh), 450
530
330, 360, 450
360
360, 420(sh), 435
325, 430(sh), 450
320, 530
330, 360, 450
360
CAH~
MeCN
3.5
CA~~
2.9
13.5
13
1.0
PhSCH ~`
3
0.24
0.06
0.06
0.83
0.83
0.22
0.06
0.06
0.75
0.75
0.23
CAH~
BPS
CH Cl
PhSCH~Ph
CAH~
2
2
10
MeCN
3.5
3.0
14
CA~~
4800
9000
PhSCH Ph~`
2
CAH~
MBPS
CH Cl
116
123
MeOPhCH~SPh
CAH~
2
2
360, 425(sh), 435
325, 430(sh), 450
330, 520
MeCH
3.9
3.0
12
CA~~
4500
MeOPhCH SPh~`
2
330, 360, 450
CAH~
a [CA] \ 4.0 ] 10~3 and 3.6 ] 10~3 M in CH Cl and MeCN, respectively. b 1.0 ] 10~2 M aryl alkyl sulÐdes in CH Cl and MBPS in MeCN
2
2
2 2
and 2.0 ] 10~3 M TA and BPS in MeCN. c t is the time at which the initial signal was halved. d Absorption coefficient at the underlined
1@2
maximum. e From ref. 16 and 18. f From ref. 16 and 17. g From ref. 23.
Phys. Chem. Chem. Phys., 2000, 2, 1701È1708
1703

View Article Online
obtained by pulse radiolysis of TA and sulfonate derivatives of
BPS and MBPS in aqueous solutions,7 thus indicating that
the medium does not a†ect signiÐcantly the transition prob-
abilities of these transients.
The transient absorptions recorded at any wavelength (see,
e.g., inset of Fig. 2) decayed by second-order laws, with half-
lives (t ) on the microsecond timescale and second-order rate
1@2
constants (2k ) in the range 109È1010 M~1 s~1. The decay
2
rate constant of the radicals is very close for TA(-H)~ and
PS(-H)~ (t \ 10È14 ls and 2k \ 1.1È1.7 ] 1010 M~1 s~1)
1@2
2
and decreases by about one order of magnitude for MBPS(-H)~
(t B 120 ls and 2k B 0.3 ] 1010 M~1 s~1), thus indicating
1@2
2
that the methoxy group stabilizes signiÐcantly the radical
species. The second-order decay kinetics recorded for the rad-
icals CAH~, TA(-H)~, BPS(-H)~ and MBPS(-H)~, together with
the lack of further transients in the absorption spectra, sug-
gests that CAH~ recombines with the counterpart radical to
produce stable products (see below).
Fig. 4 Absorption spectrum of CAH~ obtained by photolysis of
CAH (0.025 M) and di-tert-butyl peroxide (1 M) in MeCN (j
\
2
exc
355 nm).
The photochemical behavior of CA in the presence of aryl
alkyl sulÐdes investigated in CH Cl is in general agreement
2
2
with that reported for chloranil-sensitized oxidation of aro-
matic hydrocarbons in benzene and 1,2-dichloroeth-
ane,16,17,20 where CAH~ was the main transient produced
after the decay of charge-transfer complexes in the triplet
manifold.
Photolysis of CA (3.6 ] 10~3 M) in Ar-saturated MeCN
was carried out in the presence of aryl alkyl sulÐdes
(2.0 ] 10~3 M), conditions where triplet CA was efficiently
quenched (q O 150 ns). The time-resolved absorption spectra
recorded for the CA/TA system are reported in Fig. 3. Two
transients appeared on the absorption spectra just after the
resembles the longer-lived transient that was, therefore, tenta-
tively assigned to the anion CAH~. The absorption spectra of
CA~~ and CAH~ are similar to each other. This was con-
Ðrmed by quantum mechanical calculations carried out with
the INDO/1-CI model (ZINDO module of the Cerius2
package, rel. 3.8) on molecular structures optimized by use of
the set of bases 3È21G (Gaussian 98).23
ConÐguration interaction (CI) calculations including singly
excited conÐgurations built from 20 highest occupied and 20
lowest virtual molecular orbitals have shown that the energies
of the allowed S ] S transitions do not change signiÐcantly
decay of triplet CA; absorption bands with j \ 325, 430(sh)
0
n
max
for CA~~ and CAH~ and that, except for the forbidden n,p*
and 450 nm were conÐdently attributed to the anion radical
states, the two species have the same state order.
CA~~ (ref. 16 and 18) while the broad band centered at 520È
530 nm was assigned to the radical cations of aryl alkyl sul-
Ðdes.7 Remarkably, the presence of the radicals CAH~ and
PhSCH ~, the transients recorded in CH Cl , can be clearly
Because of the strong overlap between transient absorptions
in the 300È550 nm range, the kinetic analysis of signals
recorded at any wavelength could not give direct information
on the formation process of CAH~ but, by analogy with CA/
bis(4-methoxyphenyl)methane in MeCN,9 the ions CAH~ and
TA(-H)`/BPS(-H)`/MBPS(-H)` were probably formed by H-
transfer from the radical cations of the aryl alkyl sulÐdes to
CA~~. It has to be noted that the cations were not detected in
the photolysis experiments, because they absorb only at short
wavelengths (see below).
The formation quantum yields of CA~~ in MeCN were
determined by calibrating the experimental set-up with an
optically matched solution of benzophenone in the same
solvent and by use of the absorption coefficient e \ 9700 M~1
cm~1 at 435 nm for CA~~ (Table 2). For CA/TA, also the
formation quantum yield of the radical cation TA~` was
2
2 2
excluded in MeCN, if one compares the time-resolved absorp-
tion spectra of CA/TA (Fig. 3) with that of CAH~ in MeCN
(Fig. 4).21
At long delay time, the absorptions of the ion radicals were
replaced by that of a further transient with maxima at 330,
360(sh) and 450 nm (Fig. 3, spectrum recorded 40 ls after the
pulse). This spectrum resembles that of CA~~, but the much
longer decay time shows that it must be assigned to another
transient. The absorption spectrum of CAH~, produced both
by (a) photolysis of CA/bis(4-methoxyphenyl)methane in
MeCN9 and (b) deprotonation of tetrachlorohydroquinone
(CAH , 1 ] 10~4 M) in the presence of the base
2
Me NOH É 5H O (2.4 ] 10~4 M) in MeCN (see Fig. 5), also
4
2
Fig. 3 Time-resolved absorption spectra obtained by laser photoly-
sis of 3.6 ] 10~3 M CA in the presence of 2.0 ] 10~3 M TA in
MeCN, recorded 0.5 (L), 3.5 (|) and 40 ()) ls after the laser pulse
Fig. 5 Absorption
spectrum
of
1 ] 10~4
M
tetra-
(j \ 355 nm).
chlorohydroquinone and 2.4 ] 10~4 M Me NOH É 5H O in MeCN.
exc
4
2
1704
Phys. Chem. Chem. Phys., 2000, 2, 1701È1708

View Article Online
obtained by use of its absorption coefficient at 530 nm (6100
M~1 cm~1).24 However, since the e values of BPS~` and
MBPS~` in MeCN are unknown, their formation quantum
yields were taken to be equal to the corresponding values of
CA~~; this assumption allowed the absorption coefficients of
these radical cations to be obtained (Table 2) by relationship
(3):
*A
*A
RC
RA
e
\ e
RA
(3)
RC
where *A and *A represent the absorbance changes mea-
RC RA
sured at the absorption maxima of the radical cation of an
aryl alkyl sulÐde and CA~~, respectively. The e values
RC
obtained in this way (4800 and 4500 M~1 cm~1 for BPS~`
and MBPS~`, respectively) are of the same order of magnitude
as those previously reported for similar radical cations in
water.7 In all cases, the radical ions were produced with high
Fig. 6 Changes in absorbance with time following photolysis of
3.6 ] 10~3 M CA and 2.0 ] 10~3 M TA in MeCN solution at (a) 450
and (b) 530 nm.
efficiency (/ P 0.75, see Table 2).
RI
The high values of the formation quantum yield of CAH~
(/ \ 0.24È0.30, see Table 2) show that, surprisingly, the
An
hydrogen transfer process competes efficiently with other pos-
while the radicals TA(-H)~, BPS(-H)~, and MBPS(-H)~ formed
sible decay channels. It is known that radical ions of aryl alkyl
sulÐdes are able to undergo deprotonation and/or CÈS cleav-
age.3,5,6,8 However, such processes were not operative under
our experimental conditions. In fact, the lack of the typical
absorption band at ca. 360 nm (characteristic of CAH~) in the
time-resolved spectra of CA/aryl alkyl sulÐdes recorded in
MeCN (see Fig. 3, as an example) excludes the occurrence of
an efficient proton transfer process between CA~~ and radical
cations of aryl alkyl sulÐdes to give the corresponding radicals
CAH~ and TA(-H)~/BPS(-H)~/MBPS(-H)~.21 As pointed out in
a previous work,9 this is likely due to the low basicity of CA~~
in CH Cl were quenched by molecular oxygen (k D 1 ] 107
2
2
q
M~1 s~1).
3.2 Absorption spectra of cations
In the time-resolved absorption spectra of CA/aryl alkyl sul-
Ðdes there is no absorption that can conÐdently be ascribed to
the counterpart cation TA(-H)`/BPS(-H)`/MBPS(-H)`. Since
to our knowledge, the absorption spectra of these cations have
not been reported in the literature, quantum mechanical cal-
culations were performed to predict transition energies and
probabilities. The geometry of TA(-H)`, BPS(-H)`, and
MBPS(-H)` was optimized at a semi-empirical level by use of
the AM1 method (Gaussian 98) followed by geometry opti-
mization with an ab initio calculation by use of the set of bases
3È21G (Gaussian 98).23 The transition energies and oscillator
strengths ( f ) were calculated by the INDO/1-CI model
(ZINDO module of the Cerius2 package, rel. 3.8) including
soluteÈsolvent interactions in the Hamiltonian by use of a self-
consistent reaction Ðeld (SCRF) model;25 conÐguration inter-
action (CI) calculations included singly excited conÐgurations
built from 20 highest occupied and 20 lowest virtual molecu-
lar orbitals. Calculations predicted main absorption maxima
at 204 nm ( f \ 0.21), 312 nm ( f \ 0.93) and 328 nm ( f \ 1.0)
for TA(-H)`, BPS(-H)`, and MBPS(-H)`, respectively. It has
to be noted that the errors on these absorption maxima are
likely to be in the range of some tens of nanometers. On the
basis of these predictions, cations are therefore expected to
absorb essentially below 350 nm, in regions where our experi-
mental set-up is not sensitive or where strong absorptions of
other transients and of the ground state molecule can make it
difficult to distinguish their presence.
in MeCN (pK \ 6.8) which is not able to deprotonate radical
a
cations of aryl alkyl sulÐdes. Furthermore, the inefficiency of
CÈS cleavage for BPS~` and MBPS~` was supported by both
time-resolved and steady-state investigations. In fact, neither
absorption bands characteristic of PhS~ (j \ 300 and 460
max
nm)7 nor photoproducts attributable to the cations PhCH `
2
and 4-MeOPhCH ` (alcohols and acetamides) were obtained
2
by time-resolved and steady-state investigations, respectively,
of CA/BPS and CA/MBPS in MeCN.
The concentrations of TA and BPS in MeCN were kept at
2 ] 10~3 M to avoid formation of dimer radical cations
which would make the reaction mechanism more compli-
cated; strong absorptions in the 600È850 nm range, due to
dimers, were observed at higher concentrations.24 Instead,
analogously with 4-methoxythioanisole,24 dimer radical
cations of MBPS were not detected even at a concentration of
1 ] 10~2 M. It has to be noted that the presence of the dimer
does not a†ect the reactivity of the system, since the quantum
yield of CAH~ (/ ) and of the Ðnal photoadduct (/ ) are
An
P
virtually independent of TA concentration: in fact, almost
constant / and / values were obtained at 2 ] 10~3 and
An
P
1 ] 10~2 M TA.
The decay kinetics of radical ions (see Fig. 6 for decays of
CA/TA recorded at 450 and 530 nm) were well Ðtted by
second-order laws with half-lives on the microsecond time-
3.3 Steady-state irradiation
scale (t \ 2.9È3.9 ls) and decay rate constants close to the
Photolysis of CA/aryl alkyl sulÐdes in Ar-saturated solutions
of CH Cl and MeCN was carried out by steady-state irradia-
1@2
di†usional limit (2k \ 3.1È4.5 ] 1010 M~1 s~1) (Table 2).
2
2 2
The anion CAH~ also decayed by second-order kinetics with
tion at j \ 310È390 nm, where only CA absorbs. The
exc
a half-life of 12È14 ls, longer than those observed for radical
results of these experiments are compiled in Table 3. In partic-
ions, and a di†usion controlled decay rate constant (2k D 3.5
ular, for the CA/TA system the photoreaction was clean and
led exclusively to the photoadduct PhSCH OC Cl OH, while
2
] 1010 M~1 s~1, see Table 2). These results show that there
is no signiÐcant e†ect of the sulÐde structure on decay rate
constants of charged transients, thus suggesting that their
behavior is Ñattened by chloranil properties.
2
6 4
for BPS and MBPS di†erent photoproducts were obtained,
namely benzaldehyde (or 4-methoxybenzaldehyde), benz-
aldehyde diphenyldithioacetal (or 4-methoxybenzaldehyde
diphenyldithioacetal) and diphenyl disulÐde (Table 3). These
products might derive from hydrolysis of unstable photoad-
ducts [ArCH(SPh)OC Cl OH] that produce CAH and
The formation quantum yields and decay rate constants of
charged species produced in MeCN and of CAH~ in CH Cl
2
2
remained virtually unchanged in air-equilibrated solutions,
6
4
2
Phys. Chem. Chem. Phys., 2000, 2, 1701È1708
1705

View Article Online
Table 3 Products formed in the CA-sensitized photolysis of TA, BPS
the presence of CA (2 ] 10~2M) in MeCN, a partial oxidation
and MBPS in CH Cl and MeCN together with the chemical yields
of this substrate to give benzaldehyde and diphenyl disulÐde
was observed.
2
2
Chemical
Compound
TAb,c
Solvent
Product
yielda (%)
3.4 Photoreaction mechanisms
CH Cl
2
PhSCH OC Cl OH
7.3
38.5
3.0
1.1
0.9
15
4.7
4.5
5.3
3.8
0.8
33.1
11.6
9.0
2
2
2
6 4
On the basis of the above assignments, the results of the time-
resolved and steady-state measurements on the photochemical
behavior of CA/TA, the only system where a stable photo-
product was formed, can be rationalized in terms of the reac-
tion kinetics shown in Scheme 2. Steps 1È4 refer to the less
polar CH Cl and steps 5È8 to the more polar MeCN. The
MeCN
PhSCH OC Cl OH
2
PhCHO
6 4
BPSb,d
CH Cl
2
PhCH(SPh)
PhSSPh
PhCHO
PhCH(SPh)
PhSSPh
2
MeCN
2
2
2
efficiency r(i) and the second-order rate constant 2k (i) (with
MBPSb,d
CH Cl
4-CH OPhCHO
2
2
2
3
i \ 1È8) of each step are compiled in Table 4.
4-CH OPhCH(SPh)
3
2
2
PhSSPh
In CH Cl , the quenching process of triplet CA is mainly
2
2
MeCN
4-CH OPhCHO
followed by return to the ground state (step 1), with very high
3
4-CH OPhCH(SPh)
quantum efficiency [r(1) \ 1 [ r(2) \ 1 [ / \ 0.94] and
3
PhSSPh
Rad
2k (1), corresponding to the product k r(1), close to the di†u-
2
q
a Yields referred to the substrate. Average of at least two determi-
nations. The error is ^10%. b [TA] \ [BPS] \ [MBPS] \ 0.01 M.
c Irradiation time of 60 min. d Irradiation time of 30 min.
sional limit. In contrast, the parallel formation of radical
species (step 2) is rather inefficient [(r(2) \ / \ 0.06)] and
Rad
the corresponding rate constant [2k (2) \ k r(2)] is about one
2
2
q
order of magnitude smaller than 2k (1) (Table 4). This is con-
Ðrmed by the lack of a deuterium isotope e†ect on the quen-
ching rate constant of triplet CA by TA. The radical
recombination can give both the Ðnal photoadduct [step 3,
with r(3) \ / // ] and the return H-transfer process [step
hemithioacetals [ArCH(OH)SPh] which in turn lead to benz-
aldehydes and thiophenol. A reaction between the latter pro-
ducts might lead to dithioacetal, while thiophenol can be
oxidized to the disulÐde.
P
Rad
4, with r(4) \ 1 [ r(3)] with rate constants [2k (3) \
2
2k (CAH~)r(3) and 2k (4) \ 2k (CAH~)r(4)] slightly smaller
2
2
2
than the di†usional limit.
For the CA/TA system, reaction quantum yields (/ ) of
In MeCN, the bimolecular interaction between TA and
triplet CA produces radical ions (step 5) with unitary efficiency
and rate constant 2k (5) \ k r(5) \ 0.49 ] 1010 M~1 s~1.
P
0.033 and 0.27 were obtained in CH Cl and MeCN, respec-
2
2
tively, by exciting at 366 nm. In MeCN, / were measured at
P
2
q
two TA concentrations (2 ] 10~3 and 1 ] 10~2 M) without
observing any appreciable change. It was not possible to
The main decay pathway of the radical ions CA~~ and TA~`
is return electron transfer [step 6, r(6) \ 1 [ r(7) \ 1
measure / of the other two sulÐdes because the relative
[ / \ 0.76] which occurs with a very high rate constant
P
An
amount of photoproducts was strongly dependent on the irra-
[2k (6) \ 2k (RI)r(6) \ 3.45 ] 1010 M~1 s~1, where 2k (RI)
2
2
2
diation time, thus indicating that the primary products were
not photostable. Indeed, when benzaldehyde diphenyl-
dithioacetal, PhSCH(SPh)Ph (2 ] 10~2M), was irradiated in
is the average of decay rate constants of radical ions]. At this
point one should expect the occurrence of proton transfer
from TA~` to CA~~, however, the absence of the signal at ca.
Scheme 2
1706
Phys. Chem. Chem. Phys., 2000, 2, 1701È1708

View Article Online
Table 4 Quantum efficiencies (r) and second-order rate constants (2k ) of reactions 1È4 (in CH Cl ) and 5È8 (in MeCN), see Scheme 2
2
2
2
TA
BPS
MBPS
Step
r
2k /1010 M~1 s~1
r
2k /1010 M~1 s~1
r
2k /1010 M~1 s~1
2
2
2
1
2
3
4
5
6
7
8
0.94
0.06
0.55
0.45
1.0
0.76
0.24
1.0
1.03
0.07
0.94
0.77
0.49
3.45
1.2
0.94
0.06
1.24
0.08
0.94
0.06
1.01
0.06
0.83
0.73
0.27
0.54
2.63
0.97
0.75
0.70
0.30
0.69
2.76
1.2
1.25
360 nm which is characteristic of CAH~ clearly excludes the
deprotonation process. Thus, a H-transfer process from TA~`
to CA~~ probably occurs, by analogy with CA/bis(4-methoxy-
phenyl)methane in MeCN, and, accordingly, evidence of the
absorption spectrum of CAH~ has been found at long delay
References
1
For reviews on radical cations of organosulfur compounds see:
(a) R. S. Glass, Xenobiotica, 1994, 25, 637; (b) Sulfur Centered
Reactive Intermediates in Chemistry and Biology, ed. C. Chatgilia-
loglu and K. D. Asmus, Nato ASI Series, Plenum Press, New
York, 1990.
times (t D 13 ls). On the basis of Scheme 2, one expects
2
(a) Y. Watanabe, T. Numata, T. Iyanagi and S. Oae, Bull. Chem.
Soc. Jpn., 1981, 54, 1163; (b) S. Oae, Y. Watanabe and K. Fuji-
mori, T etrahedron L ett., 1982, 23, 1189; (c) Y. Watanabe, T.
Iyanagi and S. Oae, Bull. Chem. Soc. Jpn., 1982, 55, 188; (d) T.
Takata, R. Tajima and W. Ando, Phosphorus Sulfur, 1983, 16, 67;
(e) S. Oae, A. Mikami, T. Matsuura, K. Ogawa-Asada, Y. Wata-
nabe, K. Fujimori and T. Iyanagi, Biochem. Biophys. Res.
Commun., 1985, 131, 567; ( f ) S. Kobayashi, M. Nakano, T.
Kimura and A. P. Schaap, Biochemistry, 1987, 26, 5019; (g) J. R.
Cashman, J. Proudfoot, Y.-K. Ho, M. S. Chin and L. D. Olsen, J.
Am. Chem. Soc., 1989, 111, 4844; (h) J. R. Cashman and L. D.
Olsen, Mol. Pharmacol., 1990, 38, 573; (i) E. Baciocchi, O. Lanza-
lunga and F. Marconi, T etrahedron L ett., 1994, 32, 9771; ( j) E.
Baciocchi, M. Ioele, O. Lanzalunga, M. Malandrucco and S.
Steenken, J. Am. Chem. Soc., 1996, 118, 8973.
1@2
that the half-life of the cation species must be equal to that of
CAH~. This value, which might appear particularly high,26 is
a†ected by the remarkable stabilization of the methyl cation
induced by the phenylthio substituent as shown by the kinetic
data for the hydrolysis of aryl chloromethyl sulÐdes.28
The H-transfer process [step 7, r(7) \ / \ 0.24] is a reac-
An
tion slightly slower than return electron transfer and competes
efficiently in the decay of radical ions to produce the ions
CAH~ and TA`. The Ðnal product is probably formed by
recombination of ions [step 8, r(8) \ / // and 2k (8) \
R
An
2
2k (CAH~)r(8)].
2
From the kinetic point of view (nature of transients,
quantum efficiencies and rate constants), no signiÐcant
changes were obtained by use of BPS and MBPS, instead of
TA, as quencher of triplet CA. Only slightly smaller values of
r(5) were obtained by use of BPS and MBPS (0.83 and 0.75,
respectively). In contrast, a remarkable change in photo-
chemical products was observed by replacing TA with BPS
and MBPS. In our opinion, this is probably due to the insta-
bility of the photoadduct, the precursor of the stable photo-
products obtained by irradiation of CA/BPS and CA/MBPS
(Table 3).
3
4
E. Baciocchi, O. Lanzalunga and B. Pirozzi, T etrahedron, 1997,
36, 12287.
(a) M. Platen and E. Steckhan, Justus L iebigs Ann. Chem., 1984,
1563; (b) H. Garcia, S. Iborra, M. A. Miranda and J. Primo, New
J. Chem., 1989, 13, 805; (c) M. A. Fox and A. A. Abdel-Wahab,
T etrahedron L ett., 1990, 31, 4533; (d) G. A. Empling and Q.
Wang, T etrahedron L ett., 1992, 33, 5909; (e) E. Baciocchi, E.
Fasella, O. Lanzalunga and M. Mattioli, Angew. Chem. Int. Ed.
Engl., 1993, 32, 107; ( f ) M. Kamata, Y. Murakami, Y. Tama-
gawa, M. Kato and E. Hasegawa, T etrahedron, 1994, 50, 1282; (g)
K. Chiba, Y. Yamaguchi and M. Tada, T etrahedron L ett., 1999,
39, 9035.
In conclusion, it has been clearly shown that, even though
solvent polarity does not modify the nature of the products
obtained by chloranil-photosensitization of aryl alkyl sulÐdes,
the reaction mechanism operative in the two solvents is com-
pletely di†erent. In fact, only radical transients were detected
in the less polar CH Cl , while charged species were formed
5
6
7
8
9
E. Baciocchi, C. Rol, E. Scamosci and G. V. Sebastiani, J. Org.
Chem., 1991, 56, 5498.
E. Baciocchi, C. Crescenzi and O. Lanzalunga, T etrahedron,
1997, 53, 4469.
M. Ioele, S. Steenken and E. Baciocchi, J. Phys. Chem., 1997, 101,
2979.
2
2
W. Adam, J. E. Arguello and A. B. Penenory, J. Org. Chem., 1998,
63, 3905.
E. Baciocchi, T. Del Giacco, F. Elisei and O. Lanzalunga, J. Am.
Chem. Soc., 1998, 120, 11800.
in the polar MeCN. Furthermore, radical cations of aryl alkyl
sulÐdes do not give deprotonation or fragmentation when
they react with the chloranil radical anion, but a H-transfer
process is likely to be operative.
10 G. A. Russell and J. M. Pecoraro, J. Am. Chem. Soc., 1979, 101,
3331.
The above results have shown the lack of structural e†ects
on the primary steps of the chloranil sensitized reactivity of
TA, BPS and MBPS, in line with the high rate constants of
the decay channels of radicals (in CH Cl ) and radical ions (in
11 (a) A. Romani, F. Elisei, F. Masetti and G. Favaro, J. Chem. Soc.,
Faraday T rans., 1992, 88, 2147; (b) H. Gorner, F. Elisei and G. G.
Aloisi, J. Chem. Soc., Faraday T rans., 1992, 88, 29.
12 I. Carmichael and G. L. Hug, J. Phys. Chem. Ref. Data, 1986, 15,
1.
13 W. E. Truce and F. E. Roberts, J. Org. Chem., 1963, 28, 961.
14 N. J. Bunce, in Handbook of Organic Photochemistry, ed. J. C.
Scaiano, Academic Press, New York, 1989, vol. 1, p. 243.
15 E. F. Hilinski, S. V. Milton and P. M. Rentzepis, J. Am. Chem.
Soc., 1983, 105, 5193.
16 H. Kobashi, M. Funabashi, T. Kondo, T. Morita, T. Okada and
N. Mataga, Bull. Chem. Soc. Jpn., 1984, 57, 3557.
17 H. Kobashi, T. Okada and N. Mataga, Bull. Chem. Soc. Jpn.,
1986, 59, 1975.
2
2
MeCN). This suggests a transition state very close to the
reagent structure.
Acknowledgement
Thanks are due to Professor E. Baciocchi for his advice and to
the Italian Consiglio Nazionale delle Ricerche and Ministero
della Universita e della Ricerca ScientiÐca e Tecnologica
(Programma di Ricerca di Interesse Nazionale) for Ðnancial
support.
18 J. J. Andre and G. Weill, Mol. Phys., 1968, 15, 97.
19 These values were measured in this work by laser Ñash photolysis
of tetrachlorohydroquinone (D2 ] 10~3 M) in the presence of
di-tert-butyl peroxide (1.0 M) in CH Cl , k \ 355 nm, by use
2
2
exc
of the known e \ 7300 M~1 cm~1 for CAH~ at 435 nm.
Rad
Phys. Chem. Chem. Phys., 2000, 2, 1701È1708
1707

View Article Online
20 (a) G. Jones, W. A. Haney and X. T. Phan, J. Am. Chem. Soc.,
1988, 110, 1922; (b) G. Jones and N. Mouli, J. Phys. Chem., 1988,
92, 7174; (c) H. Kobashi, S. Okabe, Y. Ohsugi and H. Shizuca,
Bull. Chem. Soc. Jpn., 1990, 63, 2173; (d) G. Jones, N. Mouli,
W. A. Haney and W. R. Bergmark, J. Am. Chem. Soc., 1997, 119,
8788.
21 As far as we know, no clear evidence for the formation of CAH~
in MeCN by deprotonation of aromatic radical cations has been
shown in the literature by transient absorption techniques.
Therefore, to be sure that the absorption spectrum of CAH~ is
not a†ected signiÐcantly by solvent properties, Ñash photolysis of
Mennucci, C. Pomelli, C. Adamo, S. Cli†ord, J. Ochterski, G. A.
Petersson, P. Y. Ayala, Q. Cui, K. Morokuma, D. K. Malick,
A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. Cioslowski,
J. V. Ortiz, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I.
Komaromi, R. Gomperts, R. L. Martin, D. J. Fox, T. Keith,
M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, C. Gonzalez, M.
Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong,
J. L. Andres, C. Gonzalez, M. Head-Gordon, E. S. Replogle and
J. A. Pople, Gaussian, Pittsburgh, PA, 1998.
24 H. Yokoi, A. Hatta, K. Ishiguro and Y. Sawaki, J. Am. Chem.
Soc., 1998, 120, 12728.
CAH (0.025 M) and di-tert-butyl peroxide (1 M) was carried out
25 (a) K. K. Stavrev, T. Tamm and M. C. Zerner, Int. J. Quant.
Chem., 1996, 60, 372; (b) M. M. Karelson and M. C. Zerner, J.
Phys. Chem., 1992, 96, 6949.
26 The half-life value of ca. 13 ls is comparable to that reported for
4-CH -C H CH ] C H generated from the corresponding
2
in MeCN. In fact, upon irradiation, the latter compound is able
to produce tert-BuO~ that extracts
quinones to generate corresponding semiquinone radicals.22 The
spectrum recorded under these experimental conditions (k
a H-atom from hydro-
\
max
3
6
4
6 5
365, 420 and 435 nm, with relative intensities of 1:0.64:1.11,
respectively) was virtually identical with that previously reported
for CAH~ in CH Cl .
chloride in MeCN (k
^ 5.9 ] 105 s~1).27
decay
27 J. Bartl, S. Steenken, H. Mayr and R. A. McClelland, J. Am.
Chem. Soc., 1990, 112, 6918.
28 F. G. Bordwell, G. D. Cooper and H. Morita, J. Am. Chem. Soc.,
1957, 79, 376.
2
2
22 P. K. Das, M. V. Encinas, S. Steenken and J. C. Scaiano, J. Am.
Chem. Soc., 1981, 103, 4142.
23 GAUSSIAN 98, Revision A.6, M. J. Frisch, G. W. Trucks, H. B.
Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, V. G.
Zakrzewski, J. A. Montgomery, Jr., R. E. Stratmann, J. C. Burant,
S. Dapprich, J. M. Millam, A. D. Daniels, K. N. Kudin, M. C.
Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi, B.
Paper a909372i
1708
Phys. Chem. Chem. Phys., 2000, 2, 1701È1708