Photochemical Oxidation of Thioketones by Singlet Molecular Oxygen Revisited: Insights into Photoproducts, Kinetics, and Reaction Mechanism

Article abstract of DOI:10.1021/acs.joc.5b01710

Photosensitized oxidation of trimethyl[2.2.1]bicycloheptane thioketones by ¹O₂ can yield more photoproducts than exclusively ketones and sulfines. Moreover, the ketone/sulfine ratio can be reversed when protic conditions and high thioketone concentrations are used, conversely to earlier results reporting ketones as the main photoproducts. A new mechanistic proposal for sulfine formation is suggested following intermolecular oxygen transfer from a peroxythiocarbonyl intermediate to a second thioketone molecule. Reaction quantum yields (10^-5-10^-2) depend on the reaction conditions and time. Sulfine production reaches a maximum at short irradiation times, whereas decomposition to the corresponding ketone is observed at long reaction times. When the thioketone substrate has a hydrogen atom at the α position a peroxyvinylsulfenic acid intermediate can be formed by proton transfer. Reaction of this intermediate with another thioketone molecule can yield more sulfine and its tautomeric vinylsulfenic acid, which dimerizes in situ to the thiosulfinate. The hydroperoxyl group of the peroxyvinylsulfenic acid can also rearrange to the α position, and by reaction with the starting thioketone, α-hydroxy thioketone and additional sulfine can be formed, while dehydration yields the α-oxo thioketone. In situ [2 + 2] and [4 + 2] self-cycloaddition of the α-oxo thioketone yields significant amounts of the corresponding adducts at prolonged irradiation times.

Full text of DOI:10.1021/acs.joc.5b01710

Article
pubs.acs.org/joc
Photochemical Oxidation of Thioketones by Singlet Molecular
Oxygen Revisited: Insights into Photoproducts, Kinetics, and
Reaction Mechanism
Antonio J. Sanchez-Arroyo, Zulay D. Pardo, Florencio Moreno-Jimenez, Antonio Herrera,
́
́
Nazario Martín,* and David García-Fresnadillo*
Complutense University of Madrid, Department of Organic Chemistry, Faculty of Chemical Sciences, E-28040, Madrid, Spain
S
* Supporting Information
ABSTRACT: Photosensitized oxidation of trimethyl[2.2.1]-
1
bicycloheptane thioketones by O2 can yield more photo-
products than exclusively ketones and sulfines. Moreover, the
ketone/sulfine ratio can be reversed when protic conditions
and high thioketone concentrations are used, conversely to
earlier results reporting ketones as the main photoproducts. A
new mechanistic proposal for sulfine formation is suggested
following intermolecular oxygen transfer from a peroxythio-
carbonyl intermediate to a second thioketone molecule.
Reaction quantum yields (10−5−10−2) depend on the reaction
conditions and time. Sulfine production reaches a maximum at
short irradiation times, whereas decomposition to the
corresponding ketone is observed at long reaction times.
When the thioketone substrate has a hydrogen atom at the α position a peroxyvinylsulfenic acid intermediate can be formed by
proton transfer. Reaction of this intermediate with another thioketone molecule can yield more sulfine and its tautomeric
vinylsulfenic acid, which dimerizes in situ to the thiosulfinate. The hydroperoxyl group of the peroxyvinylsulfenic acid can also
rearrange to the α position, and by reaction with the starting thioketone, α-hydroxy thioketone and additional sulfine can be
formed, while dehydration yields the α-oxo thioketone. In situ [2 + 2] and [4 + 2] self-cycloaddition of the α-oxo thioketone
yields significant amounts of the corresponding adducts at prolonged irradiation times.
Scheme 1. Reaction Mechanism Proposed by Ramamurthy
and Coworkers for the Photosensitized Oxidation of
INTRODUCTION
■
Photosensitized and self-photosensitized oxidation of thioke-
tones in solution and in the solid phase was mainly studied by
Ramamurthy and co-workers in the 1980s, in order to develop
clean oxidation methods and to reveal the mechanism of
thiocarbonyls transformation into other functional groups of
interest in organic synthesis.1−7 Photosensitized oxidation of
thioketones 1 yielded ketones 2 and sulfines (thiocarbonyl S-
oxides) 3 as the isolated photoproducts, and the proposed
reaction mechanism (Scheme 1) involved intramolecular SO
and atomic oxygen release for ketone and sulfine formation,
respectively. The ketone/sulfine ratio was governed by
stereoelectronic factors operating on a zwitterionic or diradical
intermediate, and in general, carbonyl compounds correspond-
ing to desulfurization were reported as the main photoproducts.
However, it was recognized that several aspects of the original
mechanistic proposal were yet to be substantiated and new
experimental evidence was necessary to confirm the reaction
mechanism, its intermediates, and, in particular, the fate of
atomic oxygen apparently produced during the reaction course
as a result of sulfine formation.5 Since then, the mechanism of
photosensitized oxidation of organic compounds containing
sulfur is still open to debate.8 However, the mechanism
a
[2.2.1]Bicycloheptane Thioketones 15
a
(1R)-Thiocamphor 1a, R1 = Me; R2 = H. (1R)-Thiofenchone 1b, R1
= H; R2 = Me.
proposed by Ramamurthy has been accepted for decades and
described in several reviews and photochemistry books.9−11
Photosensitized oxidation of alicyclic thioketones 1 ((1R)-
thiocamphor 1a and (1R)-thiofenchone 1b) was reported to
Received: July 23, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acs.joc.5b01710
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
a
Scheme 2. New Proposal of Reaction Scheme for the Photosensitized Oxidation of [2.2.1]Bicycloheptane Thioketones 1
a
New proposal of reaction pathways: (i) Low substrate concentration; ketones 2 are the major photoproducts. (ii) High substrate concentration;
sulfines 3 are the major photoproducts in protic solvent or acidic conditions, while the corresponding ketones 2 are the major photoproducts in
aprotic solvent. (iii) At high substrate 1a concentration in aprotic solvent an intramolecular proton transfer from α-position to the thiocarbonyl
group occurs producing the vinylpersulfenic acid 4a. (iv) Enethiolization promoted by reaction of 1a and 4a. (v) Dehydration of the vinylsulfenic
acid 5a dimer to thiosulfinate 6a. (vi) Sulfine 3a and α-hydroxythioketone 7a formation via intermolecular reaction of 1a and the vinylpersulfenic
acid 4a. (vii) Dehydration of 4a to α-oxothioketone 8a. (viii) [2 + 2] self-cycloaddition of 8a producing 9a. (ix) [4 + 2] self-cycloaddition of 8a
yielding 10a.
give, intramolecularly, the corresponding ketones 2a and 2b
and sulfines 3a and 3b as the only major and minor products,
respectively. However, the yield of isolated compounds for
100% conversion after >48 h irradiation was low (<50%) and
attributed to difficulties in the isolation procedures and not due
to any side products.5 New oxidation experiments performed
with thioketones 1a and 1b in the presence of photosensitized
singlet oxygen allow us to give an insight into a more plausible
and complete reaction mechanism (Scheme 2 and Schemes
S1−S5, SI) which clarifies Ramamurthy’s proposal and
accounts for the formation of novel reaction products, not
previously described in the literature. Moreover, we also report
on the experimental conditions that can be used to photo-
chemically promote formation of thiocarbonyl S-oxides as the
major reaction products, instead of the corresponding ketones.
It is worth mentioning that sulfines are a unique class of
sulfur-centered heterocumulenes with Z/E isomerism constitut-
ing valuable synthetic intermediates in organic chemistry, due
to the variety of reactions in which they can take part (e.g.,
different types of cycloadditions, as reaction substrates of
thiophilic and carbophilic reagents, in α-carbanion formation
via tautomerization to vinylsulfenic acids, etc.).12−14 For
instance, the stereospecific [4 + 2] cycloaddition of
thiocarbonyl S-oxides with 1,3-butadienes, where the geometry
of the sulfine is retained in the cycloadduct, might be a shorter
alternative route for the preparation of new dihydrothiopyran
S-oxides and related heterocyclic compounds such as Aprikalim,
a potassium channel activator with antihypertensive and
antianginal activity, or derivatives thereof.15−19
RESULTS AND DISCUSSION
■
Evidences supporting the new proposal of reaction mechanism
were collected from photosensitized oxidation experiments of
thioketones 1 in (deuterated) acetonitrile or methanol in the
presence of [Ru(dip)3]Cl2 dye (ΦΔ = 0.97
0.08 in
methanol,20 0.75 0.06 in acetonitrile) irradiated with a blue
LED lamp (Figure S1, SI). The reaction crudes were analyzed
B
DOI: 10.1021/acs.joc.5b01710
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
by a UV−vis technique at low substrate 1 concentrations (ca.
0.07 mM) or NMR and GC-MS spectrometry when
irradiations were carried out at high thioketone 1 concen-
trations (1−80 mM).
acetonitrile proceeded faster. A new absorption band was
observed only in methanol which matched the UV absorption
peak of the corresponding thioketone S-oxides obtained by
thermal synthesis (Table S1, Figure S3, SI). Isosbestic points
(272 and 268 nm in acetonitrile and methanol, respectively)
were observed except for the photooxidation of 1b in
acetonitrile, due to the expected low yield of sulfine 3b,
according to previous reports.5 Kinetic analysis of the
photosensitized oxidation in methanol at low substrate
concentration (ca. 0.07 mM), where sulfine formation is
small though relevant, was possible using the UV−vis spectral
features of compounds 1, 2, and 3 (Table S1, SI). It has to be
noted that, under continuous irradiation and assuming
negligible quenching of the excited sensitizer by thioketones
1, the quantum yield and the rate of 1 disappearance should be
first order with respect to [1].23 In agreement with the previous
statement, and assuming that formation of the zwitterionic
peroxythiocarbonyl intermediate A is the rate-determining step
of the reaction, the experimental data points for the
photooxidation of 1 in methanol could be fitted to first order
kinetics with respect to [1] (Table 1, Figure 1 inset and Figure
S4, SI). The asymptotic behavior at long irradiation times
reveals that quasi-quantitative phototransformation of the
thioketones (>96%) had been achieved, and yields of sulfines
3a and 3b were (24.0 0.9)% and (11.0 0.3)%, respectively.
The quality of these results is also supported by the fraction of
conversion (f) of 1 into 3, which can be determined from the
ratio of absorption coefficients of the substrate and the product
at the isosbestic point wavelength (f = ε1λIP/ε3λIP, Figure S5,
SI).24 The f values in methanol for the photooxidation of 1a
into 3a and 1b into 3b allowed an estimation of the yields of
The zwitterionic peroxythiocarbonyl intermediate A in
Scheme 2 has been proposed to be the first formed reaction
intermediate,5 similarly to persulfoxide in reactions of singlet
oxygen with organic sulfides. Persulfoxide can be described as a
resonance hybrid of zwitterionic and diradical structures,
although previous experimental and theoretical results are
consistent with dominance of the zwitterionic form.21 This key
intermediate in the photooxidation of sulfides is a weakly
bound species with a sufficient lifetime to undergo different
inter- and intramolecular reactions (e.g., nucleophile/base at
oxygen but also showing a trend to interconvert to
intermediates often behaving as electrophilic oxidizing
agents).22 Therefore, by analogy with persulfoxide, the
peroxythiocarbonyl intermediate A can reasonably be expected
to play a significant role in the photooxidation of thioketones.
Experimental evidence in favor of the parallel pathways (i)
intramolecular and (ii) intermolecular in Scheme 2, arising
from A, and their different prevalence depending on the
reaction conditions, includes the following:
1. Evidence from UV−vis Spectrophotometry Experi-
ments. UV spectral changes showed a decrease of the
absorption band of thioketones 1 in acetonitrile or methanol
with the irradiation time (Figures 1 and S2, SI). Reactions in
sulfine (28
4)% and (14
2)%, respectively, in good
agreement with the yields of 3a and 3b determined from the
kinetic analysis. A similar calculation for the photooxidation of
1a into 3a in acetonitrile gave ca. 15% of 3a for 93% conversion
of 1a (Table 1 and Figure S2a).
On the other hand, the occurrence of an isosbestic point is
compatible with conversion of the reaction substrate in
different products via parallel reactions.25 Therefore, the
presence of the isosbestic point and the first order kinetics
observed would support a reaction mechanism where parallel
pathways allow competitive formation of 2 and 3 from the same
parent precursor A, ruling out, under these conditions, a fast
consecutive reaction causing decomposition of sulfines 3 into
their corresponding ketones 2 and elemental sulfur, via the
oxathiirane intermediate D.14,26
Figure 1. UV−vis monitorization of the photosensitized oxidation of
1b in methanol at different reaction times (1:0 min, 2:30 min, 3:1 h,
4:2 h, 5:3 h, 6:4 h, 7:5 h, 8:6 h, 9:7 h). Inset: Kinetics of thioketone 1b
disappearance (blue) and sulfine 3b formation (red).
2. Evidence from 1H and 13C NMR Spectrometry
Experiments. Results of the photosensitized oxidation of 1 in
CD3CN and CD3OD (1 < [1] < 80 mM), monitored by NMR
(accounting for 97−99% of the reaction mass) are collected in
ab c
, ,
Table 1. Experimental Results of the Photooxidation Process at Low Substrate Concentration in Methanol
compound
conversion
yield
k/h−1
R2
Φphotooxidation
Φphotoproduction
1a
3a
1b
3b
96.2%
−
96.1%
−
24.0%
0.499 0.004
0.47 0.02
0.999
0.996
0.999
0.992
1.24 × 10−4
−
−
0.30 × 10−4
−
−
11.0%
0.431 0.003
0.61 0.05
1.28 × 10−4
−
−
0.15 × 10−4
a
b
[1a]0 = (7.07 0.02) × 10−5 M; [3a]t→∞ = (1.72 0.04) × 10−5 M; [1b]0 = (7.44 0.02) × 10−5 M; [3b]t→∞ = (8.2 0.2) × 10−6 M. (1.6
c
0.2) × 10−3 mol of photons absorbed during an irradiation time of 7.05 h. In acetonitrile, with experimental data collected from Figure S2(a), SI;
[1a]0 = (6.1 0.1) × 10−5 M, k = (0.65 0.01) h−1, R2 = 0.999. Φphotooxidation = 2.7 × 10−5 after 1.5 h of substrate 1a irradiation. A 15.1% yield of 3a
was estimated for 93% conversion of 1a.
C
DOI: 10.1021/acs.joc.5b01710
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
Table 2. Experimental Results of the Photooxidation of Thioketones 1 at High Substrate Concentration in CD3CN, CD3CN +
a
H+, and CD3OD
compound
solvent
conversion
yield
Φphotooxidation
Φphotoproduction
b
e
e
1a
CD3CN
34%
−
8.5 × 10−3
−
−
−
c
c
c
c
c
i
CD3CN + H+
35%
36%
−
−
−
−
−
−
−
−
−
−
−
−
−
8.8 × 10−3
d
CD3OD
9.9 × 10−3
b
e
2a
3a
CD3CN
14.6%
9.0%
4.5%
−
−
−
−
−
−
−
−
−
−
−
−
3.6 × 10−3
2.2 × 10−3
1.3 × 10−3
1.2 × 10−3
3.4 × 10−3
7.1 × 10−3
1.7 × 10−3
1.5 × 10−3
0.6 × 10−3
1.5 × 10−3
1.5 × 10−3
0.6 × 10−3
−
CD3CN + H+
d
CD3OD
b
e
CD3CN
4.6%
CD3CN + H+
13.6%
25.6%
d
CD3OD
f
b
e
alkene-a1
CD3CN
6.6%
CD3CN + H+
6.2%
2.0%
d
CD3OD
g
b
e
alkene-a2
CD3CN
6.0%
CD3CN + H+
6.0%
2.0%
−
−
−
8.0%
3.6%
3.3%
0.0%
7.2%
25.7%
d
CD3OD
−
h
1b
2b
3b
CD3CN
8%
11%
29%
−
−
−
−
−
−
2.7 × 10−3
CD3CN + H+
2.8 × 10−3
−
−
j
CD3OD
10.1 × 10−3
h
CD3CN
−
−
−
−
−
−
2.7 × 10−3
0.9 × 10−3
1.1 × 10−3
−
i
CD3CN + H+
j
CD3OD
h
CD3CN
i
CD3CN + H+
1.8 × 10−3
j
CD3OD
8.9 × 10−3
a
b
1.2 × 10−3 mol of photons absorbed in 3 h of irradiation. 0.8 mM concentration of [Ru(dip)3]Cl2 photosensitizer in all experiments. [1a]0 = 50 ×
c
d
e
10−3 M. [1a]0 = 50 × 10−3 M, [CF3CO2H] = 50 × 10−3 M, in CD3CN. [1a]0 = 55 × 10−3 M. After 5 h of 1a irradiation in CD3CN, 59%
f
g
conversion of 1a was achieved, yielding 32% of 2a, 7.0% of 3a, 5.2% of alkene-a1, and 15% of alkene-a2. Alkene-a1, 6.30 ppm. Alkene-a2, 6.05
h
i
j
ppm. [1b]0 = 68 × 10−3 M, 3 h of irradiation. [1b]0 = 50 × 10−3 M, [CF3CO2H] = 50 × 10−3 M, 3 h of irradiation. [1b]0 = 70 × 10−3 M, 4 h of
irradiation.
Table 3. Influence of the Reaction Conditions on the Ketone/Sulfine (2/3) Ratio at Low Substrate Conversions; Effect of the
a
Initial Concentration of Thioketone in CD3OD and of the Presence of CF3CO2H 50 mM in CD3CN
b
Substrate
CD3OD
CD3CN
CD3CN + H+
c
1a
[1a]0/mM
1
10
25
50
80
50
50
d
2a/3a
3.20
1
0.45
10
0.26
40
0.28
70
0.24
80
3.20
68
0.71
50
e
1b
[1b]0/mM
d
f
2b/3b
0.50
0.24
0.14
0.11
0.12
−
0.50
a
b
c
Data from 1H NMR experiments. 0.8 mM concentration of [Ru(dip)3]Cl2 photosensitizer in all experiments. [CF3CO2H] = 50 mM in CD3CN. 3
h irradiation. Uncertainty of 2/3: 10%. 4 h irradiation. Only ketone 2b was detected.
d
e
f
Tables 2, 3 and Figures S6−S26, SI. Ketones 2 and sulfines 3
were the main reaction products but not the only ones, since
new alkene-type, carbonyl and thiocarbonyl derivatives were
also detected in the case of substrate 1a (Tables 2, 3 as well as
Figures S11−S14, S24−S26 and Table S3, SI). Moreover,
depending on the reaction conditions (i.e., reaction time,
solvent used, concentration of substrate or presence of H+) the
ketone to sulfine ratio 2/3 could be reversed (Figure 2 and
Table 3). These experimental findings are in contrast to
previous works claiming only one type of major product
(ketone) regardless of the changes in the reaction conditions
(initial concentration of substrate or aprotic/protic nature of
the solvent), since only a slight dependency of the product
distribution on the aprotic/protic character of the solvent was
observed in previous reports.4,5
the NMR experiments slowed down the kinetics of the
photooxidation process considerably (e.g., at [1a]0 50 × 10−3
M, k (0.26 0.01) h−1 in CD3CN; while at [1a]0 6.1 × 10−5 M,
k (0.65
0.01) h−1 in CH3CN, Table 1). Therefore, NMR
monitorization was typically maintained at 3−5 h of irradiation
or performed when the thioketone substrate was quantitatively
depleted, as confirmed by GC-MS.
Evolution of the diagnostic signals of each compound, i.e.
methyls and methylene protons in α-position to the functional
group (Figures S6−11a, SI), allowed the determination of the
1
2/3 ratio by H NMR under different reaction conditions
(Table 3). Conversely to previous results, the product
distribution was found to be strongly dependent on the solvent
used, since the thioketone S-oxide can be the major product at
low substrate conversions and the amount of sulfine is always
higher in CD3OD than in CD3CN (Figure 2, Tables 2, 3, and
Figures S6−S8 and S11a, SI). For example, photooxidation of
thiofenchone 1b in CD3CN gave fenchone 2b as the only
product, with no evidence of sulfine 3b formation, in agreement
It has to be mentioned that, despite the longer lifetimes
1
expected for *[Ru(dip)3]2+ and, especially, O2 in deuterated
solvents, increased quenching (mainly physical, Table S2, SI) of
the excited states at the high substrate concentrations used in
D
DOI: 10.1021/acs.joc.5b01710
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
with UV−vis results (Figure 2 and Figures S2c and S6, SI).
However, when a similar experiment was carried out in CD3OD
sulfine 3b was the major product with an overall ratio 3b/2b =
7.1 (Figure 2 and Figure S7, SI), in contrast with UV−vis
results at low substrate concentration where 3b was a minor
product (Table 1 and Figure S2d, SI).
These observations would agree with the existence of an
intermolecular reaction pathway for sulfine formation and also
with the stabilization of the zwitterionic peroxythiocarbonyl
intermediate A by hydrogen (deuterium) bonding with
methanol molecules (Schemes S1a and S1b, SI), as proposed
by Ramamurthy.5 Therefore, in protic solvents such as
methanol, intramolecular ring closure to 2 by pathway (i) via
the highly unstable 1,2,3-dioxathietane intermediate B would be
retarded, while production of 3 at sufficient substrate
concentration would be favored by the intermolecular pathway
(ii), through the intermediate C. As a result, sulfines could be
major reaction products in protic solvents at short reaction
times, in contrast to general assumptions (Figure 2). Similarly
to the role played by m-chloroperbenzoic acid in thermal
oxidations of thioketones, or by persulfoxide species in the
photooxidation of organic sulfides, the zwitterionic peroxythio-
carbonyl intermediate A would also behave as an oxygen atom
donor to another thiocarbonyl molecule in its surroundings.
1
Concerning thiocamphor 1a photooxidation, the H NMR
spectra in CD3CN (Figure S11, SI) evidenced formation of
ketone 2a as the major product and, surprisingly, sulfine 3a was
also detected, in contrast with 1b photooxidation in the same
solvent, where sulfine 3b product was not present. On the other
hand, besides 2a and 3a formation, other unknown compounds
(Figure 3 and Figures S8, S11−S13, SI) showing doublet
signals evolving in the alkene region between 6.00 and 7.00
ppm (Tables 2 and S3, SI) were detected in CD3CN and, to a
minor extent, in CD3OD as well. It has to be noted that no
signals in the alkene region were detected when thiofenchone
1b was photooxidized (Figure S6, SI). The observed coupling
Figure 2. Distribution of the photooxidation products of thioketones 1
under different reaction conditions. Data from H NMR experiments.
(a) 1a photooxidation; (b) 1b photooxidation. Conversions and
product yields (%) are collected in Table 2.
1
Figure 3. Evolution of the 1H NMR signals in the alkene region by irradiation of 1a in CD3CN for 22 h. 1H NMR spectra (500 MHz) collected at 0
h, red (spectrum 1); 1 h, olive (spectrum 2); 2 h, green (spectrum 3); 3 h, cyan (spectrum 4); 5 h, blue (spectrum 5); 22 h, purple (spectrum 6, 300
MHz). [1a]0 = 50 × 10−3 M.
E
DOI: 10.1021/acs.joc.5b01710
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
constant for the alkene signals detected in 1a photooxidation
was always 3.3 Hz, corresponding to coupling of alkenic and
bridgehead protons, as demonstrated by COSY correlations
(Figure S12, SI) and also by experimental data from other
structurally related compounds (Experimental section). A signal
at 6.30 ppm appeared at shorter irradiation times, while a
doublet at 6.05 ppm showed up at longer times (>1.5 h).
Surprisingly, the signal at 6.05 ppm increased at the expense of
that at 6.30 ppm and a new doublet at 6.45 ppm was detected
after 22 h of irradiation, when 1a was quantitatively depleted.
This new signal grew at the expense of that at 6.05 ppm (Figure
3), and other additional signals were also detected after 22 h of
irradiation (Figure S1, SI).
methanol and in acidic conditions as well (Table 3 and Figure
2).
Considering the previous results, a study on the influence of
the initial concentration of substrates 1 on the products
distribution was undertaken in CD3OD. Experimental results
are collected in Table 3. The intense singlet signals of products
2 and 3 near 1.0 ppm, corresponding to the methyl groups of
the ketones and thiocarbonyl S-oxides, were used to determine
the relative ratios of the photoproducts (Figures S19 and S20).
The results of 1H NMR experiments after 3−4 h of
photosensitized irradiation in CD3OD showed that the 2/3
ratio decreases as the initial concentration of substrate 1
increases, again suggesting that an intermolecular reaction
pathway is involved in sulfine formation.
Formation of alkene derivatives from substrate 1a could be
explained by the tendency of the terminal oxygen in the
zwitterionic peroxythiocarbonyl intermediate A to stabilize its
negative charge via intramolecular proton transfer when there
are hydrogen atoms available in the α-position to the
thiocarbonyl group attacked by singlet oxygen (Scheme S2,
SI). By analogy with the Schenck-ene reaction (Scheme S6, SI)
and also in agreement with the typical reactivity of persulfoxide
intermediates (removal of a α-hydrogen leading to hydro-
peroxysulfonium ylide formation),8,27,28 proton transfer in the
zwitterionic peroxythiocarbonyl intermediate A and double
bond migration would lead to production of a vinylpersulfenic
acid derivative 4a. Formation of 4a would be promoted in
acetonitrile but minimized in methanol due to stabilization of A
by the protic solvent (Figure 3 and Scheme S1, SI). Chemical
evolution of 4a via intermolecular reactions could explain the
development of additional 1H NMR signals in the alkene region
(Schemes 2 and S2, S4, SI).
It is noteworthy that the broad singlet signal corresponding
to residual water was slightly deshielded from 2.2 to 2.3 ppm
during the reaction course (Figure S14, SI). It must also be
noted that when photooxidation experiments of thioketones 1
were carried out in the presence of DABCO (1,4-diazabicyclo-
[2.2.2]octane), a well-known singlet oxygen quencher,29 the
formation of products was slowed down and, in the case of 1a
in CD3CN, alkene signals were completely suppressed (Figures
S15 and S16, SI), demonstrating that the evolution of alkene
byproducts is linked to singlet oxygen production in the
reaction system.
It has to be noted that as the reaction time of 1a
photooxidation in CD3CN increases, the 2a/3a ratio also
tends to increase (e.g., 2a/3a = 3.0 after 3 h, while 2a/3a = 4.6
after 5 h of irradiation, and whereas 2a was the major reaction
product after 22 h of irradiation, 3a signals were undetectable
by 1H NMR; Table 2 and S3, SI). Significant decomposition of
sulfine 3a at room temperature and in the dark was confirmed
after 2 days in CD3CN solution (Figure S21, SI). Therefore,
slow decomposition of 3a could occur in the reaction crude at
long reaction times or, alternatively, by its participation in
subsequent secondary reactions.
13C NMR spectra of the reaction crudes at long irradiation
times evidenced the presence of new signals of little intensity in
the alkene region (Figure S22, SI) in agreement with results
1
from H NMR, but also in the carbonyl region at 211.1 and
212.6 ppm (Figure S23, SI), and thiocarbonyl region 277.0
ppm (Figure S24, SI), that could be attributed to new keto and
thioketo species formed during the thermal evolution of the
reaction crude.
3. Evidence from GC-MS Spectrometry Experiments.
GC-MS spectrometry was used to analyze the reaction crude of
photooxidized 1a in CD3CN, in the search for the new
byproducts (Figures S25−S41, SI) formed by dark reactions
occurring simultaneously with the photooxidation process but
whose structure could not be elucidated by NMR. After
continuous irradiation for 22 h, thiocamphor 1a was
quantitatively depleted (undetectable by NMR) but gave a
residual peak (m/z: exp. 168.1, Figure S27, SI) by GC-MS.
Camphor 2a (m/z: exp. 152.1, Figure S26, SI) was found by
both 1H NMR and GC-MS (major peak of the chromatogram,
Figure S25a, SI). On the other hand, no signals of thiocamphor
It has been previously reported that protic conditions in
photooxidation reactions of sulfur compounds can stabilize
reaction intermediates bearing electrostatic charge,30 and this
effect could be of major importance in those cases where
different products can be formed from the same reaction
intermediate, since its stabilization can modify the ratio
between products. In this sense, and taking into account that
the 2/3 ratio is higher in CD3OD than in CD3CN, the addition
of protons to an aprotic solvent such as CD3CN could help to
reverse the sulfine/ketone ratio in favor of the thioketone S-
oxide product. Addition of an equimolar amount of trifluoro-
acetic acid to a 50 mM solution of thioketones 1 in CD3CN,
and subsequent irradiation, evidenced improved production of
the corresponding sulfines 3 (Table 3, Figure 2, and Figures
S17 and S18, SI). A control experiment carried out for 3 h in
1
S-oxide 3a were found by H NMR, but a GC-MS peak (m/z:
exp. 184.1, Figure S30, SI) that could be attributed to sulfine 3a
or its isomeric vinylsulfenic acid 5a was detected. However, due
to the expected thermal instability of sulfines and sulfenic acids
at the high temperatures used in gas chromatography, this peak
could most likely correspond to the α-hydroxy thioketone 7a,
isobaric with 3a and 5a, in agreement with the occurrence of a
new thiocarbonyl peak observed by 13C NMR (277.0 ppm,
Figure S24, SI) corresponding, most likely, to the endo
diastereoisomer. Formation of α-hydroxythiocamphor 7a
could be explained via the vinylpersulfenic acid 4a generated
by proton transfer from intermediate A (Scheme S2, SI), and a
subsequent reaction through intermediates I−L and M
according to the reaction pathway (vi) depicted in Scheme
S3, SI. GC-MS clues supporting the formation of the
vinylpersulfenic acid 4a or its isobaric isomer L have been
found (m/z: exp. 200.1, Figure S31; and m/z: exp. 199.9,
Figure S32, SI), although these compounds are also isobaric
1
the dark showed no evolution of the reaction substrates by H
NMR. Moreover, formation of alkenes was also significant
when trifluoracetic acid was added to 1a in CD3CN. Therefore,
the intermolecular pathway leading to sulfine 3a is prevalent at
high substrate 1a concentrations in protic solvents such as
F
DOI: 10.1021/acs.joc.5b01710
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
with the highly thermally unstable and less likely detectable
1,2,3-dioxathietane intermediate B (Scheme S1, SI). Evidence
of a cyclic derivative of 4a generated during the analysis by loss
of molecular hydrogen at the high temperatures in the
chromatograph, most likely from intermediate I (Scheme S3,
SI), has also been detected (m/z: exp. 198.1, Figure S29, SI).
Moreover, α-oxo thioketone 8a (m/z: exp. 182.1, Figure S28,
SI) has been detected as well, and its formation could be
justified by the intramolecular dehydration process shown in
the reaction pathway (vii) (Scheme S3, SI).
Four intense peaks related to intermolecular reactions of 8a,
which can be attributed to the different diastereoisomers of
adducts 9a and 10a formed via the [2 + 2] and [4 + 2] self-
cycloadditions of 8a, have been detected as well (m/z: exp.
364.1, Figures S37 and S39−S41, SI). Three also intense
chromatographic peaks corresponding to decomposition by-
products of these adducts by loss of one (m/z: exp. 332.2,
Figures S35 and S36, SI) or two sulfur atoms (m/z: exp. 300.3,
Figure S33, SI) were also detected, as could be expected for the
two possible byproducts related to the loss of one sulfur atom,
and only one byproduct when two sulfur atoms are extruded
from 10a. The two carbonyl peaks observed by 13C NMR at
211.1 and 212.6 ppm (Figure S23, SI) could therefore be
assigned to the most abundant diastereoisomers of these
adducts. Although several carbonyl signals could be expected
from 8a, 9a (C2 symmetry group) and 10a products, since self-
cycloaddition reactions of 8a are expected to occur rapidly after
8a formation, it is likely that the observed carbonyl peaks
correspond to the most stable diastereoisomers among the four
possible structures of adducts 9a and 10a. Fast in situ [2 + 2]
and hetero-Diels−Alder [4 + 2] self-cycloaddition reactions of
α-oxo thioketones have previously been described.31−34
chromatograph by decomposition of 10a adducts. In addition,
the fragmentation pattern of the peak of m/z 262.0 includes a
major peak at m/z 198.1, in good agreement with the possible
formation of the proposed cyclic derivative of 4a.
When all these experimental facts are taken into consid-
eration, a new mechanistic proposal for the photooxidation of
thioketones 1 with sensitized singlet oxygen is summarized in
Scheme 2 and Schemes S1−S5. Evidence collected from UV−
vis, NMR, and GC-MS experiments allows us to propose the
formation of ketones 2 and sulfines 3 by intramolecular and
intermolecular reaction pathways, respectively, in contrast to
previous reports proposing monomolecular reactions only.
Therefore, sulfine formation can now be explained without the
need for intramolecular release of atomic oxygen but
intermolecular oxygen transfer to another thioketone molecule
from the first formed peroxythiocarbonyl zwitterionic inter-
mediate A. Moreover, and also conversely to previous
statements, the ketone/sulfine ratio can be reversed by
controlling the reaction conditions, thus enabling the formation
of thioketone S-oxides as the major reaction products at
sufficiently high substrate 1 concentrations and relatively low
substrate 1 conversions. The aprotic/protic nature of the
solvent or the presence of a protic organic acid influences the
stability of the first formed peroxythiocarbonyl zwitterionic
intermediate A and, therefore, the ketone/sulfine ratio. Under
aprotic conditions, intramolecular closure of the zwitterion A to
a 1,2,3-dioxathietane intermediate B and a subsequent retro-[2
+ 2] process yields ketone 2 plus SO fragments (intramolecular
pathway (i) prevailing), as previously described by Ramamur-
thy. However, at sufficiently high substrate 1 concentrations
and under protic or acidic conditions, the peroxythiocarbonyl
intermediate A is stabilized or protonated to a hydroperoxy-
thiocarbonyl intermediate AH+, and the intermolecular pathway
(ii) prevails. As a result, the probability of ring closure to 2 is
decreased, while the chances of a bimolecular reaction with
another molecule of thioketone 1 are increased, leading to
production of thioketone S-oxide 3 as the major photoproduct
(Scheme S1, SI) through solvated or protonated intermediates
C or CH+, respectively. However, and in agreement with
previous reports, the studied sulfines 3 tend to slowly
decompose in the dark, due to their thermal instability, giving
their corresponding ketones via intermediate D. Therefore,
efficient production of thioketone S-oxides 3 is limited to
relatively low conversions of the reaction substrates 1.
Of particular interest is the case where the thione bears a
hydrogen atom at the α-position (e.g., thiocamphor 1a) and its
concentration is high enough, since experimental evidence of
proton transfer to the peroxythiocarbonyl group have been
found in an aprotic solvent such as acetonitrile (pathway (iii),
Scheme S2, SI). Proton transfer in 1a produces a vinyl-
persulfenic acid intermediate 4a which, upon reaction with
another thiocamphor molecule, promotes formation of 5a, the
usually not observed vinylsulfenic acid tautomer of 3a (pathway
(iv)). Evidence proving the photoinduced formation of
unsaturated intermediates and products related to 4a and 5a
and of vinylsulfenic acid 5a dimerization and dehydration to its
thiosulfinate derivative 6a (pathway (v), Scheme S4, SI) has
been found by NMR and GC-MS spectrometry. In parallel, 1,3-
migration of the hydroperoxyl group in 4a to the α-position,
followed by intermolecular reaction with another 1a molecule
would yield 3a and the α-hydroxy thioketone 7a (pathway (vi),
Scheme S3, SI). Reaction pathways (ii), (iv), and (vi) can
explain the remarkable production of 3a from 1a in an aprotic
In order to explain the different signals observed by 1H NMR
in the alkene region between 6 and 7 ppm, and their evolution
with time, the GC-MS peaks of low intensity (m/z: exp. 200.1,
Figure S31; and m/z: exp. 199.9, Figure S32, SI) would support
the formation of vinylpersulfenic acid 4a (Scheme S2, SI), since
at least one if not both peaks could be tentatively assigned to 4a
and/or its L isomer (Scheme S3, SI). On the other hand, the
enethiolization of sulfines to vinylsulfenic acids has previously
been described using catalytic amounts of pyridinium p-
toluenesulfonate,35,36 and it is likely that reaction of 4a with
1a could play a similar role yielding 3a in equilibrium with 5a
(Scheme S2, SI). In this sense, it is quite remarkable the
presence of a peak corresponding to one of the two possible
diastereoisomers of thiosulfinate 6a (m/z: exp. 350.2, Figure
S38, SI) formed by dimerization of 5a and subsequent
dehydration via an intermediate such as N, in agreement with
the typical chemistry of sulfenic acids (Scheme S4, SI).37−40
Taking into account the presence of alkene-type protons in the
different reaction intermediates shown in Schemes S2 and S4
(SI), 4a, E−H, 5a, and 6a, several signals of alkenic protons
1
with coupling constants of 3.3 Hz could be expected in the H
NMR spectrum acquired at prolonged reaction times, in
agreement with the experimental results (Figure 3 and Figure
S13, SI). Further support about the occurrence of the
vinylpersulfenic acid intermediate 4a could be found in a
peak (m/z: exp. 262.0, Figure S34, SI) that could be attributed
to addition of inorganic sulfur to the alkene group in the cyclic
derivative of 4a generated by loss of molecular hydrogen (m/z:
exp. 198.1, Figure S29, SI).41 It has to be noted that inorganic
sulfur can be produced by decomposition of sulfine 3a;
moreover, elemental sulfur can also be produced in the
G
DOI: 10.1021/acs.joc.5b01710
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
periodically removing the caps of the quartz cell and NMR tube.
Reproducibility of the illumination geometry was possible with a piece
of graph paper on top of the magnetic stirrer and an appropriate
holder for the NMR tube. The whole illumination system was
mounted inside a rigid box equipped with a sliding cover. The interior
of the illumination system was lined with pieces of black card.
Temperature during the irradiation experiments was 21 2 °C.
Chemical Actinometry. In order to determine the amount of
photons absorbed by the specific geometries of the irradiated systems,
i.e., quartz cells of 10 mm optical path length (3 mL) and standard
NMR tubes (0.6 mL), the following procedure assuring total
absorption of incident light was used:42,43
Six samples of 3 mL (V1) each were prepared, under red light, from
a 0.15 M stock solution of potassium ferrioxalate in H2SO4 0.05 M.
Each sample was added, under red light, to a 10 mm quartz cell
containing a 3 mm × 1 mm magnetic bar. The quartz cell was placed,
under red light, inside the box used for the irradiation experiments and
care was taken to ensure reproducibility of the relative geometry with
respect to the blue LED lamp. The samples were exposed to blue light
for 5, 10, 30, 60, 90, and 120 s, respectively. Conversions of Fe(III) to
Fe(II) were always below 5% for all the irradiation times used. A blank
sample was also prepared under red light with 0.1 mL of the
aforementioned stock solution in a glass vial, which was kept in the
dark. A buffered solution of 1,10-phenanthroline 0.1% (w/v) was
prepared in a 25 mL volumetric flask with 0.0250 g of 1,10-
phenanthroline, 3.3944 g of sodium acetate, and H2SO4 0.5 M. 0.1 mL
(V2) of each irradiated solution were added to a vial. One mL of
distilled and deionized water and 0.5 mL of 1,10-phenanthroline
buffered solution were added to the vials containing each irradiated
sample and also to the vial with the blank sample, for a total volume of
1.6 mL (V3) in each vial. These solutions were stirred and kept in the
dark for 1 h to allow quantitative formation of the tris(1,10-
phenanthroline)iron(II) complex. Absorbance of each solution at 510
nm was measured (triplicates) in quartz cells of 1 mm optical path
length in order to determine the amount of Fe(II) formed as a result
of the radiation absorbed by the illuminated system. The same
procedure was used for the standard NMR tube geometry, except V1 =
0.6 mL. Calibration graphs corresponding to each geometry are shown
in Figure S42, SI.
solvent such as acetonitrile, while production of 3b from 1b was
not observed in the same solvent. 1,3-Migration of the
hydroperoxyl group in 4a and subsequent intramolecular
dehydration account for relevant α-oxo thioketone 8a
formation (pathway (vii), Scheme S3, SI). Self-cycloaddition
reaction of 8a yields cycloadducts 9a (pathway (viii)) and 10a
(pathway (ix)) (Scheme S5, SI), which have been detected by
their moderately intense peaks in GC-MS experiments (Figures
S25, S37, and S39−41, SI) and the intense peaks of the
decomposition products of 10a (Figures S33, S35, and S36, SI).
All these reaction pathways, not previously reported in the
literature about photosensitized oxidation of thioketones 1, are
supported by experimental facts and are coherent with our new
mechanistic proposal based on the competition between intra-
and intermolecular reaction pathways starting from the first
formed peroxythiocarbonyl reaction intermediate A.
CONCLUSIONS
■
A new reaction scheme for the photosensitized oxidation of
thioketones 1 with singlet molecular oxygen is proposed, which
explains the product distribution and its strong dependence on
the reaction conditions (substrate concentration, solvent
nature, and acidic medium) and time of irradiation. In addition,
our mechanistic proposal can explain, in the case of thioketones
with hydrogens in α position to the thiocarbonyl group, such as
thiocamphor 1a, the formation of reaction products not
previously described in the literature for the photosensitized
oxidation of thiocarbonyls (vinylsulfenic acid derivatives, α-
hydroxy thioketones and α-oxo thioketones, and their 1,3-
dithietane and 1,3,4-oxadithiin cycloadducts).
EXPERIMENTAL SECTION
■
Reagents and Solvents. D-(+)-Camphor (+97%) and L-
(−)-fenchone (+98%) were the substrates used in the preparation of
the corresponding thioketones. Tris(4,7-diphenyl-1,10-
phenanthroline)ruthenium(II) dichloride ([Ru(dip)3]Cl2)20 was used
as the singlet oxygen photosensitizer. DABCO (1,4-diazabicyclo-
[2.2.2]octane) (97%) was used as a singlet oxygen quencher.
Acetonitrile (99.85%, H2O < 0.02%), deuterated acetonitrile (H2O <
0.05%, 99.8% D), methanol (99.9%, H2O < 0.02%), and deuterated
methanol (d4) (H2O < 0.03%, 99.8% D) were used as the solvents in
photooxidation experiments, and trifluoroacetic acid (≥99.5%) was
used for those photooxidation experiments in acidic conditions.
Lawesson’s reagent (99%) and m-chloroperbenzoic acid (75%) were
used in the preparation of D-(+)-camphor and L-(−)-fenchone
thioketones and sulfines, respectively. Sulfuric acid (95−97%), 1,10-
phenanthroline (≥99%), sodium acetate (99%), and potassium
ferrioxalate were used in actinometry experiments to determine the
photon flux absorbed by the irradiated systems where the photo-
oxidation reactions were carried out. Potassium ferrioxalate was
synthesized according to literature procedures.42,43 Briefly, 1.5 M
solutions of iron(III) chloride (97%), FeCl3 (15 mL), and potassium
oxalate monohydrate (98.8−101.0%), K2(C2O4)·H2O (45 mL), in
distilled and deionized water were mixed in a beaker under red light
Photooxidation Experiments and Analysis of Reaction
Crudes. Photosensitized oxidation of thioketones was carried out in
quartz cells (10 mm optical path length) and standard NMR tubes (5
mm outer diameter) in the presence of [Ru(dip)3]Cl2 photosensitizer
(Figure S1, SI). The photooxidation process was monitored at
1
different time intervals by UV−vis spectrophotometry, H and 13C
NMR spectrometry (300, 500, and 700 MHz for 1H, and 75, 125, and
176 MHz for 13C, respectively; spectra were referenced to residual
solvent signals at δH/C 1.94/1.39 and 118.3 ppm (acetonitrile-d3) or
3.31/49.15 ppm (methanol-d4) relative to tetramethylsilane as the
internal standard) or GC-MS (injection of 1 μL of the irradiated
reaction crude on a GC system coupled to a mass selective detector
under the following conditions: column (5% phenylmethylsiloxane) 30
m × 0.25 mm × 0.25 μm; carrier gas, helium; injector temperature,
220 °C; oven temperature, 70 °C, 2.5 °C/min, 300 °C; interface
temperature, 150 °C; ion source, 230 °C; electron energy, 70 eV).
The volumes of the irradiated solutions (containing thioketone
substrate, photosensitizer, and the corresponding solvent) were 3 and
0.6 mL for the quartz cells and NMR tubes, respectively. In order to
avoid sample dilution, UV−vis monitorization was carried out with
max
(GU10 red LED lamp 230 V; λem 638.0 nm, 22 nm fwhm). A
solutions of ca. 7 × 10−4 M thioketone and ca. 5 × 10−6
M
precipitate was formed after 10 min of magnetic stirring. The
amorphous solid was filtered and recrystallized three times in water.
Light-green crystals were obtained (33%) and stored in an amber vial.
Experimental Setup Used in the Irradiation Experiments. A
[Ru(dip)3]Cl2 photosensitizer (using quartz cells of 10.00 mm or 1.00
mm optical length). The fraction (F) of incident photons absorbed by
the singlet oxygen photosensitizing dye was calculated using the
equation F = 1−10−A, where A is the absorbance of the photosensitizer
in the 420−530 nm range (average value A = 0.276, i.e. ∼47% of the
incident photons are absorbed by the dye in the quartz cell). On the
other hand, NMR monitorization was performed with 1−80 mM
thioketone 1 solutions containing [Ru(dip)3]Cl2 0.80 mM. This
concentration of photosensitizer assured total absorption of the
incident photons entering the NMR tube (determined by chemical
max
LED lamp (GU10 blue LED 240 V, 1.8 W T/C; λem 463.5 nm, 22
nm fwhm, Figure S1, SI) was used to illuminate either a quartz cell
(10.00 mm optical length) or a standard NMR tube placed 5 cm away
from the LED lamp. Homogenization of the illuminated solutions and
equilibration with air was achieved with a 3 mm × 1 mm magnetic bar
and a magnetic stirrer placed below the irradiated vessels (Figure S1,
SI). Air equilibration was assured during the experiments by
H
DOI: 10.1021/acs.joc.5b01710
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
(s, 3H); 0.95 (s, 3H); 0.83 (s, 3H). 13C-RMN (CD3CN, 75
MHz) δ/ppm: 210.2; 56.1; 50.1; 44.3; 38.1; 35.5; 26.9; 19.2;
18.2; 12.0. UV−vis (CH3OH) λabsmax/nm (ε/M−1 cm−1
( 7%)): 269 (7750). Spectral data of 3b: 1H-RMN
(CD3CN, 300 MHz) δ/ppm: Z diastereoisomer methyl groups:
1.54 (s, 3H); 1.51 (s, 3H); 1.37 (s, 3H) and E diastereoisomer
methyl groups: 1.80 (s, 3H); 1.36 (s, 3H); 1.35 (s, 3H). 13C-
RMN and DEPT-135 (CD3CN, 75 MHz) δ/ppm: Z
diastereoisomer: 218.0 (C, CSO); 55.4 (C); 52.8 (C);
50.1 (CH); 45.3 (CH2); 38.0 (CH2); 26.0 (CH2); 24.2 (CH3);
23.0 (CH3); 18.1 (CH3) and E diastereoisomer: 215.1 (C, C
SO); 60.9 (C); 50.3 (CH); 49.7 (C); 46.5 (CH2); 37.0
(CH2); 31.2 (CH3); 27.8 (CH3); 25.5 (CH2); 19.2 (CH3).
HMQC and HMBC spectra confirmed the signals of the Z and
E diastereoisomers (Figure S46, SI). 4:1 mixture of Z/E
diastereoisomers, in agreement with literature data.45 UV−vis
(CH3OH) λabsmax/nm (ε/M−1 cm−1 ( 7%)): 270 (7800).
actinometry). The 220−300 nm spectral region was chosen for UV−
vis monitorization since compounds 1 and 3 show strong absorption
bands with distinct peaks in the selected wavelength interval, while
ketones 2 display very weak absorption (ε ≤ 30 M−1 cm−1) and,
therefore, have a negligible spectral contribution compared to that of
compounds 1 and 3.
Analysis of the Distribution of Products and the Reaction
Kinetics. Using the absorption coefficients (ε/M−1 cm−1) of sulfines 3
in the wavelength interval from 285 to 295 nm, where the thioketone
substrates 1 and the ketone photoproducts 2 present negligible
absorption, the sulfine concentrations at different times during the
irradiation process could be determined by UV−vis spectrophotom-
etry for the photooxidation experiments in methanol (Figures S2−S4,
SI). Similarly, taking into account the concentration of photogenerated
sulfine, the concentration of the remaining thioketone could be
estimated.
1
In H NMR experiments, the characteristic signals of the reaction
substrates 1 and products 2 and 3 in the regions of the α-methylene
and methyl protons were used to determine the product distributions
and their concentrations at different time intervals (Figures S6−S11
and S17−S20, SI). Examples of 1,7,7-trimethylbicyclo[2.2.1]hept-2-
ene derivatives with sulfur substituents at position 2, which are
structural analogues of 4a, 5a, or 6a having alkenic protons showing
coupling constants of ca. 3.3 Hz with their corresponding bridgehead
protons, are given in Figure S43, SI, together with data from their
NMR spectra.44
Photophysical Characterization of the System under Study.
The experimental setup used for the photophysical characterization
has been described elsewhere.46 Emission decay traces (10 shots/
sample, λexc = 532 nm, λem = 620 nm for the photosensitizer and λem
=
1270 nm for singlet oxygen) were analyzed using the Marquardt
algorithm for decay analysis included in the software package of the
instrument. The singlet oxygen production quantum yield of the
[Ru(dip)3]Cl2 photosensitizer in acetonitrile was determined relative
to Rose Bengal in acetonitrile, which was used as the reference
photosensitizer (ΦΔ = 0.71).47 The singlet oxygen luminescence decay
curves at 1270 nm following pulse laser excitation were recorded at
different laser powers in the 1−10 mW range. Experimental runs at
different incident powers were carried out by measuring the singlet
oxygen emission decay from the solution of reference photosensitizer
(Rose Bengal) initially and, consecutively, from the [Ru(dip)3]Cl2
photosensitizer. The relative ratio of the slopes of the linear plots of
singlet oxygen luminescence vs the excitation laser radiant power was
used to determine the singlet oxygen production quantum yield
(Figure S47, SI).23 The luminescence intensities at 1270 nm were
measured using quartz cells under identical experimental conditions
from identical laser pulses, and from solutions with matched
A multipeak fitting procedure using a Voight function was employed
to calculate the area under the respective methyl peaks of the 1a
reaction crudes in CD3OD (Figure S20, SI).
Synthesis of thioketones and their sulfines:
- Synthesis of (1R)-thiocamphor 1a and (1R)-thiofenchone 1b.
Thioketones 1a and 1b were prepared by refluxing 13.5 mmol
of the corresponding ketones D-(+)-camphor and L-(−)-fen-
chone, respectively, with Lawesson’s reagent (6.75 mmol) in
dry toluene (50 mL) under an argon atmosphere for 15 days.
Monitorization of the reaction crude by gas chromatography
(carrier gas: nitrogen; injector and detector temperature: 220
°C; column temperature: 100 °C; injected volume: 0.60 μL)
allowed detection of the remaining ketone. Solvent removal
under vacuum and subsequent column chromatography (silica
gel, hexane) yielded 80% of an orange solid 1a or liquid 1b,
respectively. Characterization of the pure products agreed with
absorbance at 532 nm (absorbance = 0.100
0.009 at 532 nm).
Stern−Volmer analysis of the emission lifetimes allowed the
determination of the bimolecular deactivation rate constants collected
in Table S2, SI.
5
1
literature data. Spectral data of 1a: H-RMN (CD3CN, 300
MHz, Figure S44, SI) δ/ppm: 2.81 (dm, J = 21.0 Hz, 1H,
CHexo−CO); 2.44 (d, J = 21.0 Hz, 1H, CHendo−CO); 2.22 (t, J
= 4.2 Hz, 1H); 2.10−2.00 (m, 1H); 1.89−1.79 (m, 1H); 1.47−
1.38 (m, 1H); 1.30−1.21 (m, 1H); 1.10 (s, 3H); 1.08 (s, 3H);
0.72 (s, 3H). 13C-RMN (CD3CN, 75 MHz) δ/ppm: 274.1;
69.6; 55.7; 49.1; 45.4; 34.2; 27.1; 19.4; 19.3; 13.0. UV−vis
(CH3CN) λabsmax/nm (ε/M−1 cm−1 ( 7%)): 479 (12), 247
(8750), 214 (3030). UV−vis (CH3OH) λabsmax/nm (ε/M−1
cm−1 ( 7%)): 480 (11), 247 (9320), 214 (3960). Spectral data
ASSOCIATED CONTENT
■
S
* Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.joc.5b01710.
Schemes of reaction depicting the proposed reaction
mechanism. Experimental setup used in the photo-
oxidation experiments. UV−vis spectra and analysis of
1
of 1b: H-RMN (CD3CN, 300 MHz, Figure S45, SI) δ/ppm
1
the information retrieved from the spectral data. H and
(J/Hz): 2.33 (s, 1H); 1.87 (d, J = 9.9 Hz); 1.80−1.63 (m, 4H);
1.28 (s, 3H); 1.24−1.18 (m, 1H); 1.16 (s, 3H); 1.12 (s, 3H).
13C-RMN (CD3CN, 75 MHz) δ/ppm: 280.3; 67.3; 58.7; 47.8;
44.2; 36.3; 29.0; 26.8; 25.6; 19.5. UV−vis (CH3CN) λabsmax/nm
(ε/M−1 cm−1 ( 7%)): 477 (12), 244 (8915), 218 (sh) (4160).
UV−vis (CH3OH) λabsmax/nm (ε/M−1 cm−1 ( 7%)): 477
(11), 244 (8640), 218 (sh) (3380).
13C 1D and 2D NMR spectra and analysis of the
information retrieved from the spectral data. GC-MS
data. Photophysical characterization of the sensitizer
excited state and singlet oxygen in the system under
study (PDF)
- Thermal synthesis of the (1R)-thiocamphor sulfine 3a and
(1R)-thiofenchone sulfine 3b.45 The experimental procedure
was carried out according to literature data. 3a 72% yield (E
diastereoisomer only), 3b 46% yield (80:20, Z/E mixture of
diastereoisomers). Structural characterization of 3a and 3b
AUTHOR INFORMATION
■
Corresponding Authors
*E-mail: nazmar@ucm.es (N.M.).
*E-mail: dgfresna@ucm.es (D.G.-F.).
agreed with literature data.5,45 Spectral data of 3a: H-RMN
1
(CD3CN, 300 MHz) δ/ppm: 2.95 (dm, J = 21.0 Hz, 1H,
CHexo−CO); 2.46 (d, J = 21.0 Hz, 1H, CHendo−CO); 2.04 (t, J
= 4.2 Hz, 1H); 1.92−1.82 (m, 2H); 1.54−1.39 (m, 2H); 1.10
Notes
The authors declare no competing financial interest.
I
DOI: 10.1021/acs.joc.5b01710
J. Org. Chem. XXXX, XXX, XXX−XXX

The Journal of Organic Chemistry
Article
(26) Carlsen, L.; Holm, A.; Koch, E.; Stilkerieg, B. Acta Chem. Scand.
1977, 31b, 679−682.
ACKNOWLEDGMENTS
■
The work was supported by the Ministry of Science and
Innovation (Project Reference CTQ2011-24652). A.J.S.-A.
thanks the Complutense University for a predoctoral grant
(ref. CT4/14).
(27) Prein, M.; Adam, W. Angew. Chem., Int. Ed. Engl. 1996, 35, 477−
494.
(28) Clennan, E. L. Tetrahedron 2000, 56, 9151−9179.
(29) Wilkinson, F.; Helman, W. P.; Ross, A. B. J. Phys. Chem. Ref.
Data 1995, 24, 663−1021.
(30) Clennan, E. L.; Greer, A. J. Org. Chem. 1996, 61, 4793−4797.
(31) Back, T. J.; Dyck, B. P.; Parvez, M. J. Org. Chem. 1995, 60, 703−
710.
(32) Crossland, I. Acta Chem. Scand. 1977, 31b, 890−894.
(33) El-Essawy, F. A. G.; Yassin, S. M.; El-Sakka, I. A.; Khattab, A. F.;
Sotofte, I.; Madsen, J. O.; Senning, A. J. Org. Chem. 1998, 63, 9840−
9845.
(34) Hegab, M. I.; Abd El-Galil, A. A.; Abdel-Megeid, F. M. E. Z.
Naturforsch. (B) 2002, 57, 922−927.
(35) Mazzanti, G.; van Helvoirt, E.; van Vliet, L. A.; Ruinaard, R.;
Masiero, S.; Bonini, B. F.; Zwanenburg, B. J. Chem. Soc., Perkin Trans. 1
1994, 3299−3304.
(36) Mazzanti, G.; Ruinaard, R.; van Vliet, L. A.; Zani, P.; Bonini, B.
F.; Zwanenburg, B. Tetrahedron Lett. 1992, 33, 6383−6386.
(37) Block, E.; O'Connor, J. J. Am. Chem. Soc. 1974, 96, 3929−3944.
(38) Davis, F. A.; Jenkins, R. H., Jr. J. Am. Chem. Soc. 1980, 102,
7967−7969.
REFERENCES
■
(1) Rajee, R.; Ramamurthy, V. Tetrahedron Lett. 1978, 19, 5127−
5130.
(2) Jayathirtha Rao, V.; Ramamurthy, V. Indian J. Chem., Sect. B 1980,
19B, 143−145.
(3) Ramnath, N.; Ramesh, V.; Ramamurthy, V. J. Chem. Soc., Chem.
Commun. 1981, 112−114.
(4) Rao, V. J.; Muthuramu, K.; Ramamurthy, V. J. Org. Chem. 1982,
47, 127−131.
(5) Ramnath, N.; Ramesh, V.; Ramamurthy, V. J. Org. Chem. 1983,
48, 214−222.
(6) Arjunan, P.; Ramamurthy, V.; Venkatesan, K. J. Org. Chem. 1984,
49, 1765−1769.
(7) Rao, V. P.; Ramamurthy, V. Tetrahedron 1985, 41, 2169−2176.
(8) Clennan, E. L.; Liao, C. Tetrahedron 2006, 62, 10724−10728.
(9) Maciejewski, A.; Steer, R. P. Chem. Rev. 1993, 93, 67−98.
́
(10) Corsaro, A.; Pistara, V. Tetrahedron 1998, 54, 15027−15062.
(39) Gupta, V.; Carroll, K. S. Biochim. Biophys. Acta, Gen. Subj. 2014,
1840, 847−875.
(40) Li, X. B.; Xu, Z. F.; Liu, L. J.; Liu, J. T. Eur. J. Org. Chem. 2014,
2014, 1182−1188.
(11) Gilbert, A.; Baggott, J. Essentials of Molecular Photochemistry;
Blackwell Scientific Publications: Oxford, 1991; pp 518−520.
(12) Zwanenburg, B. J. Sulfur Chem. 2013, 34, 142−157.
(13) Ulrich, H. Cumulenes in Click Reactions; Wiley: Chichester, 2009;
pp 13−23.
(41) Bulman Page, P. C.; Wilkes, R. D.; Reynolds, D. Alkyl
Chalcogenides: Sulfur-based Functional Groups. 2.03.2.1.7 Dialkyl
Sulfides. Formation from alkenes. In Synthesis: Carbon with One
Heteroatom Attached by a Single Bond; Ley, S. V., Vol. Ed. In
Comprehensive Organic Functional Group Transformations; Katritzky, A.
R., Meth-Cohn, O., Rees, C. W., Eds-in-chief; Pergamon: Oxford,
1995; Vol. 2, pp 137−138.
(14) Zwanenburg, B. Recl. Trav. Chim. Pays-Bas-J. R. Neth. Chem. Soc.
1982, 101, 1−27.
(15) Lucassen, A. C. B. Sulfine-based Synthesis of Four-, Five- and
Six-Membered Heterocycles. Ph.D. Thesis, Radboud University
Nijmegen, Netherlands, November 2003. http://repository.ubn.ru.
nl/handle/2066/19352 (accessed July 1, 2015).
(16) Philipse, H. J. F. Silyl-Mediated and Oxidative Synthesis of
Sulfines. Ph.D. Thesis, Radboud University Nijmegen, Neherlands,
October 2003. http://repository.ubn.ru.nl/bitstream/handle/2066/
19347/19347_silyanoxs.pdf?sequence=1 (accessed July 1, 2015).
(17) Damen, T. J. G. Synthesis and Reactions of α-Oxo Sulfines and
3,6-Dihydro-2H-thiopyran S-oxides. Ph.D. Thesis, Radboud University
Nijmegen, Netherlands, March 2002. http://repository.ubn.ru.nl/
dspace31xmlui/bitstream/handle/2066/19104/19104_syntanreo.
pdf?sequence=1 (accessed July 1, 2015).
(42) Kuhn, H. J.; Braslavsky, S. E.; Schmidt, R. Pure Appl. Chem.
2004, 76, 2105−2146.
(43) Montalti, M.; Credi, A.; Prodi, L.; Gandolfi, M. T. Handbook of
Photochemistry, 3rd ed.; CRC Press: Boca Raton, FL, 2006; Chapter
12, pp 601−616.
́
́
(44) Alvarez García, A. M. Sıntesis estereoselectiva y aplicaciones de
́
nuevos tioderivados con esqueleto norbornanico. Ph.D. Thesis,
Complutense University of Madrid, Spain, September 2008. http://
eprints.ucm.es/8558/ (accessed July 1, 2015).
(45) Maccagnani, G.; Innocenti, A.; Zani, P. J. Chem. Soc., Perkin
Trans. 2 1987, 1113−1116.
(18) Pinto, I. L.; Buckle, D. R.; Rami, H. K.; Smith, D. G. Tetrahedron
Lett. 1992, 33, 7597−7600.
(19) Bastin, R.; Albadri, H.; Gaumont, A. C.; Gulea, M. Org. Lett.
2006, 8, 1033−1036.
(20) García-Fresnadillo, D.; Georgiadou, Y.; Orellana, G.; Braun, A.
M.; Oliveros, E. Helv. Chim. Acta 1996, 79, 1996−1222.
(21) Jensen, F.; Greer, A.; Clennan, E. L. J. Am. Chem. Soc. 1998, 120,
4439−4449.
́
(46) Díez-Mato, E.; Cortezon-Tamarit, F. C.; Bogialli, S.; García-
Fresnadillo, D.; Marazuela, M. D. Appl. Catal., B 2014, 160−161, 445−
455.
(47) Ortiz, M. J.; Agarrabeitia, A. R.; Duran-Sampedro, G.; Prieto, J.
B.; Lopez, T. A.; Massad, W. A.; Montejano, H. A.; Garcia, N. A.;
Arbeloa, I. L. Tetrahedron 2012, 68, 1153−1162.
(22) Clennan, E. L. Acc. Chem. Res. 2001, 34, 875−884.
(23) Wilkinson, F.; Helman, W. P.; Ross, A. B. J. Phys. Chem. Ref.
Data 1993, 22, 113−263. It has to be noted that <3% of *[Ru(dip)3]2+
dye is quenched when [1] = 1 mM compared to the very efficient
quenching (> 95%) by dissolved molecular oxygen in both acetonitrile
and methanol (Table S2, SI). Therefore, the steady-state approx-
imation can be applied to all the excited states involved in the
photoprocess (*[Ru(dip)3]2+ and 1O2). The rate of 1 disappearance is
first-order with respect to [1] when the deactivation rate constant, kd,
1
of O2 by the solvent (kd = 1/τΔ, where τΔ is the emission lifetime of
singlet oxygen in the solvent) is much higher than the product kr·
1
[substrate] (where kr is the overall rate constant of O2 depletion in
the presence of the substrate, either by physical quenching or chemical
reaction).
(24) Berlett, B. S.; Levine, R. L.; Stadtman, E. R. Anal. Biochem. 2000,
287, 329−333.
(25) Croce, A. E. Can. J. Chem. 2008, 86, 918−924.
J
DOI: 10.1021/acs.joc.5b01710
J. Org. Chem. XXXX, XXX, XXX−XXX