Concerning the deprotonation of the trimethylsulfonium ion by the dimethylsulfinyl anion

Article abstract of DOI:10.1039/c1ob05889d

As shown by deuterium labelling experiments, the deprotonation of the trimethylsulfonium ion (1) by the dimsyl anion (8) is accompanied by extensive hydrogen exchange. This cannot be explained by an acid-base equilibrium between the trimethylsulfonium ion (1) and the dimsyl anion (8) on one side and dimethylsulfonium methylide (2) and DMSO on the other side, because for thermodynamic reasons this process is irreversible due to the limited life-time of 2. Therefore, the isotopic exchange that accompanies the deprotonation is an indicator of a more complex deprotonation process. It is suggested that in a kinetically controlled reaction, a proton of 1 is transferred to the O-atom of 8 rather than to the carbanionic centre. This means that instead of DMSO, its tautomer, hydroxy-methylsulfonium methylide (10), is obtained in the deprotonation process. Similarly, in the acid-base interaction between DMSO and its conjugate base 8, the formation of the DMSO tautomer 10 is kinetically favoured. The intermediate 10 produced in this way transfers a DMSO-derived proton to 1 when it intervenes in the back reaction 10 + 2 → 8 + 1. An alternative mechanism based on methyl group exchange between 1 and 8 could be excluded by a ¹³C-labelling experiment. The hydrogen exchange according to the suggested scenario is taking place in competition with the reaction of dimethylsulfonium methylide (2) with electrophilic substrates. This explains the different degrees of isotopic exchange when compounds of different electrophilicities are used to scavenge 2 from the deprotonation-hydrogen distribution equilibria.

Full text of DOI:10.1039/c1ob05889d

View Article Online / Journal Homepage / Table of Contents for this issue
Organic &
Biomolecular
Chemistry
Dynamic Article Links
Cite this: Org. Biomol. Chem., 2011, 9, 7748
www.rsc.org/obc
PAPER
Concerning the deprotonation of the trimethylsulfonium ion by the
dimethylsulfinyl anion
Peter Haiss and Klaus-Peter Zeller*
Received 3rd June 2011, Accepted 8th August 2011
DOI: 10.1039/c1ob05889d
As shown by deuterium labelling experiments, the deprotonation of the trimethylsulfonium ion (1) by
the dimsyl anion (8) is accompanied by extensive hydrogen exchange. This cannot be explained by an
acid–base equilibrium between the trimethylsulfonium ion (1) and the dimsyl anion (8) on one side and
dimethylsulfonium methylide (2) and DMSO on the other side, because for thermodynamic reasons this
process is irreversible due to the limited life-time of 2. Therefore, the isotopic exchange that
accompanies the deprotonation is an indicator of a more complex deprotonation process. It is
suggested that in a kinetically controlled reaction, a proton of 1 is transferred to the O-atom of 8
rather than to the carbanionic centre. This means that instead of DMSO, its tautomer,
hydroxy-methylsulfonium methylide (10), is obtained in the deprotonation process. Similarly, in the
acid–base interaction between DMSO and its conjugate base 8, the formation of the DMSO tautomer
10 is kinetically favoured. The intermediate 10 produced in this way transfers a DMSO-derived proton
to 1 when it intervenes in the back reaction 10 + 2 → 8 + 1. An alternative mechanism based on methyl
group exchange between 1 and 8 could be excluded by a 13C-labelling experiment. The hydrogen
exchange according to the suggested scenario is taking place in competition with the reaction of
dimethylsulfonium methylide (2) with electrophilic substrates. This explains the different degrees of
isotopic exchange when compounds of different electrophilicities are used to scavenge 2 from the
deprotonation–hydrogen distribution equilibria.
Intriguingly, we observed an almost complete deuterium deple-
Introduction
tion of the transferred methylene group of octadeuteriodimethyl-
sulfonium methylide ([D8]2) when the perdeuteriated sulfur ylide
was prepared from [D9]1 with the dimethysulfinyl (dimsyl) carban-
ion (8) as base in DMSO.4 By way of contrast, the dideuteriated
methylene group was introduced intact into 7 when [D9]1 was
deprotonated by sodium hydride in DMF.
In aprotic, dipolar solvents the trimethylsulfonium ion (1) is easily
deprotonated to yield dimethylsulfonium methylide (2), which is a
versatile reagent in organic synthesis.1,2
(1)
The loss of deuterium in the attempted generation of [D8]2
from [D9]1 and the dimsyl anion (8) cannot be explained with
the establishment of the acid–base equilibrium (eqn (2)). With
the pKa values for DMSO (35)6 and the trimethylsulfonium ion
(24.5),7,8 an equilibrium constant of K = 1010.5 follows for this acid–
base reaction. Therefore, the rate constant for the back reaction
k2 = K-1k1 ª 10-10.5k1 is too small to account for a rapid hydrogen
exchange within the mean life-time of a few minutes reported for
the dimethylsulfonium methylide (2).2
In a more recent application, Kitano and Ohashi3 reported on
the nucleophilic ortho-methylation of ortho-substituted nitroben-
zenes by action of 2. Based on deuterium labelling experiments,4
we could show that the reaction proceeds essentially according to
the concept of vicarious nucleophilic substitution.5 The reaction
sequence is outlined for 2-nitroanisole (3) as substrate in Scheme 1.
H
The formation of the Meisenheimer complex (s -adduct) 4 is
followed by the E1-like b-elimination via 5 to give the elimination
product 6 with the sulfur ylide 2 acting as base. The concomitantly
formed trimethylsulfonium ion (1) protonates 6 to give the
methylation product 7.
(2)
The hydrogen exchange accompanying the formation of 2
clearly indicates that the deprotonation of 1 with the dimsyl
anion (8) is more complex than expressed by eqn (2). Because
the deprotonation of sulfonium ions by the dimsyl anion in
Institut fu¨r Organische Chemie, Auf der Morgenstelle 18, 72076 Tu¨bingen,
Germany. E-mail: kpz@uni-tuebingen.de; Fax: +49 (0)7071 29 5076;
Tel: +49 (0)7071 29 72444
7748 | Org. Biomol. Chem., 2011, 9, 7748–7754
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Scheme 1 The nucleophilic ortho-methylation of 2-nitroanisole (3) by dimethylsulfonium methylide (2).4
DMSO is the classical protocol for the generation of sulfur
ylides,2 it is of some interest to obtain deeper insights into this
process. In the present article, the results of 13C- and D-labelling
experiments are presented and a mechanism is suggested to
account for the unexpected hydrogen exchange that accompanies
the deprotonation of the trimethylsulfonium ion (1) by the dimsyl
anion (8).
(0.75), which corresponds to 87.8% CD3, 11.5% CD2H and 0.7%
CDH2. In the 1H NMR spectrum, a singlet signal for the 3-methyl
group is absent and replaced by a complex multiplet of weak
intensity composed of a quintet at d = 2.255 ppm (CD2H) and
a triplet at 2.264 ppm (CH2D). The above deuterium labelling
experiments are summarized in Scheme 2.
The substitution of the hydrogen atoms in dimethylsulfonium
methylide (2) by hydrogen atoms originating from DMSO was
further demonstrated by addition of an equimolar amount
of trimethylsulfonium iodide (1·I-) to a solution of [D5]8 in
[D6]DMSO. After standing for 1 h, to allow the generated
dimethylsulfonium methylide to decompose, 1H NMR spec-
troscopy indicated an increase of the signal for the residual protons
in [D6]DMSO corresponding, within experimental error, to 9
1H-atoms per added trimethylsulfonium iodide (1·I-) (referenced
against internal dioxane) (Fig. 1). The hydrogen transfer from 1
to DMSO is coupled to the presence of dimsyl anion, because a
solution of 1·I- in pure [D6]DMSO is stable over several weeks.
The experimental results leave it beyond doubt that the
trimethylsulfonium-derived hydrogen atoms of the sulfur ylide 2
become superseded by hydrogen atoms descending from DMSO,
before the reagent 2 can introduce the ortho-methyl group into
2-nitroanisole according to Scheme 1.
Results and discussion
As explained in the introduction, only a small portion of the
deuterium labelling in [D9]1 survived in the introduced methyl
group of compound 7 when the transferring reagent 2 was
prepared by deprotonation with the dimsyl anion 8 in DMSO.
Mass spectrometric measurements reveal that only 5% [D1]7 is
present. The 1H NMR spectrum confirms that the 3-methyl group
of 1-methoxy-3-methyl-2-nitrobenzene (7) is almost completely
free of deuterium. The strong singlet for CH3 at 2.29 ppm is
accompanied by a partly overlapping triplet of very low intensity at
~2.27 ppm for CH2D. In a complementary experiment, unlabelled
1 was deprotonatedby [D5]8 in [D6]DMSOand then reactedwith 3.
The molecular ion peak region of the methylation product exhibits
the following isotopic distribution: m/z 170 (100), 169 (13), 168
Scheme 2 Methylation of 2-nitroanisole (3) with dimethylsulfonium methylide prepared from (CD3)3S+I- in DMSO (a) and (CH3)3S+I- in [D6]DMSO
(b).
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 7748–7754 | 7749
©

View Article Online
(3)
The potential nucleophilicity of the central S-atom in the dimsyl
anion (8) could be the base for an exchange of methyl groups
between the sulfonium ion 1 and the anion 8. Initiated by an
SN2 attack on 1, an equilibrium with the dimethyl-oxosulfonium
methylide (9) with two equivalent methyl substituents could be
established (Scheme 3). In the back reaction, a DMSO-derived
methyl group can be transferred with the same probability as the
return of the original methyl group. Because the dimsyl anion
equilibrates with its parent DMSO, which is present in excess, this
scenario could result in an almost complete exchange of the methyl
groups in 1 for methyl groups originating from DMSO. Provided
this equilibrium operates fast enough parallel to the practically
irreversible formation of dimethylsulfonium methylide (2), the
steps summarized in Scheme 3 could give a possible rationalization
of the observed isotopic scrambling results.
Fig. 1 (a) 1H NMR spectrum of a mixture of [D6]DMSO (550 mL,
7.775 mmol, d = 2.50 ppm) and dioxane (5 mL, 0.059 mmol, d = 3.55 ppm)
showing the residual protons. From the intensity ratio it follows that the
[D6]DMSO contains 0.92 atom% 1H atoms. The deuteriation degree of this
[D6]DMSO used was given as 99.5% by the supplier. (b) 1H NMR spectrum
obtained after addition of NaH (12.5 mg, 0.52 mmol) to a mixture of
[D6]DMSO (1.500 mL, 21.205 mmol) and dioxane (43 mL, 0.505 mmol)
followed after 15 min by trimethylsulfonium iodide (1·I-) (107 mg,
0.52 mmol corresponding to 4.68 mmol 1H-atoms). All manipulations
were performed under a stream of nitrogen and the spectrum was measured
after 1 h to allow for decomposition of the deprotonation product 2 into
ethene and dimethyl sulfide. From the intensity ratio, it follows that the
signal at 2.50 ppm represents, after correction for the residual protons in
[D6]DMSO, 4.70 mmol 1H-atoms.
Scheme 3 Methyl group exchange vs. deprotonation as competing
parallel reactions.
To put this hypothesis to the test, [13C]methyl-
dimethylsulfonium iodide ([13C]1·I-) was prepared from dimethyl
sulfide and [13C]methyl iodide (99% 13C) and applied to the
1
methylation of 3. In the H NMR spectrum of the methylation
product, the signal for the 3-methyl substituent is split into a
singlet (d = 2.28 ppm) and a doublet (1J(13C,H) = 129 Hz) in the
ratio 2 : 1. The NMR data and a 13C-labelling degree of 32.6%
derived from the mass spectrum indicate that one labelled and
two unlabelled carbon atoms are utilized in the methylation
of 3 (Scheme 4). Thus, the 13C-labelling experiment excludes a
methyl group exchange mechanism (Scheme 3) to rationalize
the hydrogen exchange between dimsyl/DMSO on one side and
Formally, the isotopic exchange in the trimethyl sulfo-
nium/dimsyl/DMSO system can occur by two modes. (1) The
appearance of DMSO-derived hydrogen atoms in the sulfur ylide
(2) and vice versa could be the result of a methyl group exchange. (2)
The carbon–sulfur framework of the interacting species remains
intact, but the hydrogen atoms migrate between the two species.
In the context of possibility (1), a study by Buncel and co-
workers12 on the denticity of the dimsyl anion toward 1,3,5-
H
trinitrobenzene is of interest. Three dimsyl s -adducts could be
identified by NMR spectroscopy, which demonstrates the unique
tridentate character of the dimsyl anion as an O-, S-, and C-
nucleophile. The O- and S-adducts are the kinetically favoured
complexes, whereas the C-adduct represents the thermodynami-
cally preferred product. It has been concluded that the structure
8 fully represents the tridentate nature of the dimsyl anion.
In terms of natural resonance theory, its weight amounts to
77.3% and some delocalization by admixture of double bond–no
bond structures, amongst them 8a contributes 8%, stabilizing the
anion.13
Scheme 4 Methylation of 2-nitroanisole with DMSO/NaH/13CH3-
(CH3)2S+I-.
7750 | Org. Biomol. Chem., 2011, 9, 7748–7754
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
trimethylsulfonium/dimethylsulfonium methylide on the other
side.
However, provided that the kinetically favoured proton transfer
to the O-atom of 8 is a general principle for the acid–base
interaction of CH-acids and the ambident dimsyl anion (8), a
rationalization of the H-exchange is possible which requires no
additional assumption such as intermediate complex formation.
Analogously to the kinetically preferred formation of 10 in the
deprotonation of 1 by the dimsyl anion (8), the interaction between
DMSO and its conjugate base 8 should also generate a stationary
concentration of the DMSO tautomer 10 (equilibrium (c) in
Scheme 5).
The DMSO tautomer 10 generated according to equilibrium (c)
contains a hydroxy proton originating from DMSO, thus, when it
intervenes in equilibrium (a), a DMSO-derived H-atom becomes
incorporated into the trimethylsulfonium ion (1). Provided that
the kinetically favoured proton transfers from 1 and DMSO,
respectively, to the O-atom of 8 work fast enough against the
thermodynamic sink (formation of 2 and DMSO, (b) in Scheme 5)
the substitution of the hydrogen atoms in 2 by hydrogen atoms
introduced with DMSO would find a feasible explanation.
At this point it may be necessary to address the possible
involvement of water traces in the hydrogen exchange process.17
The residual water in the solvents DMSO ([D6]DMSO) and THF
is ꢀ40 and ꢀ20 ppm, respectively. With regard to some additional
water introduced by the glassware etc., the H2O concentration in
the reaction mixtures should not exceed 50 ppm.
The remaining possibility (2) requires a hydrogen jumping pro-
cess between the two interacting species coupled with a reversible
deprotonation to guarantee the return to trimethylsulfonium ions
with partly exchanged hydrogen atoms. When this is repeated many
times the hydrogen atoms introduced by the sulfonium ion 1 and
DMSO are finally statistically distributed over the deprotonation
product dimethylsulfonium methylide (2) and the solvent DMSO.
Reversibility of the deprotonation of the trimethyl sulfonium ion
(1) should become possible when the negatively charged oxygen
atom of the dimsyl anion (8) instead of the carbanionic center
abstracts a proton from 1 (equilibrium (a) in Scheme 5).14 In this
case, in equation (2), dimethyl sulfoxide is replaced by its formal
“enol” hydroxy-methylsulfonium methylide (10), which according
to ab initio and DFT calculations is ca. 100 kJ mol-1 less stable
than DMSO.15 However, reversibility is not sufficient to explain
the observed hydrogen exchange because in the back reaction the
same proton would return to its origin. Therefore, the reversibility
of the deprotonation has to be supplemented by a process allowing
for hydrogen interchange.
The dimsyl reagent used to deprotonate 1 is prepared by
reaction of DMSO with sodium hydride. Adventitious water traces
dragged in by solvents will be transformed into NaOH under these
conditions. The amount of free water formed by the equilibrium
(4) is ꢀ 1% of the moisture introduced by the solvents. This
follows from the equilibrium constant of (4) (~10-3.7) and the
large excess of 8 produced by the deprotonation of DMSO with
NaH. The deuterium loss in the case of [D6]DMSO caused by this
equilibrium can be neglected because [D6]DMSO is several orders
of magnitude more abundant than the proton source, water.
(4)
After the addition of trimethylsulfonium iodide (1·I-), the de-
protonation of 1 by OH- (reaction (5)) contributes an insignificant
amount to the formation of dimethylsulfonium methylide (2),
when compared with the deprotonation of 1 by the dimsyl anion
(8). However, reaction (5) becomes important with respect to the
formation of water. The trimethylsulfonium ion (1) is >10 orders
of magnitude more acidic than DMSO, therefore, this process
overrides reaction (4) as water source.
Scheme 5 Kinetic (a) vs. thermodynamic controlled (b) deprotonation of
the trimethylsulfonium ion (1) and kinetically favoured O-protonation of
8 by its conjugated acid DMSO (c). The hydrogen atoms of 1 are marked
as H* to indicate the difference of the hydroxy proton of 10 in equilibria
(a) and (c).
In a first attempt to explain the hydrogen scrambling, we envis-
aged the possibility of hydrogen transfer in some sort of interme-
diate complex between DMSO- and trimethylsulfonium-derived
species. The reversible formation of the sulfurane complex16 11
by nucleophilic attack of the carbanionic centre in 10 at the
sulfur atom in 2 may illustrate this hypothesis. In the proper
conformation, the sulfurane 11 would accomplish the structural
conditions for hydrogen exchange between the two combined
entities via a six-membered transition state.
(5)
The water concentration now present in the system corresponds
to the water content of solvents used, however, each H2O molecule
now contains a proton originating from the trimethylsulfonium
ion (1). In combination with the protonation of 8 by water, this
could give rise to a catalytic cycle for the transfer of one hydrogen
atom from 1 into DMSO (Scheme 6), provided that this process is
competitive with the deprotonation of 1 by 8.
However, the almost complete exchange of the hydrogen atoms
of 1 by hydrogen atoms of DMSO cannot be rationalized with
such a catalytic cycle, because this scenario does not include a way
back from 2 to 1, which is a condicio sine qua none for extensive
hydrogen scrambling.
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 7748–7754 | 7751
©

View Article Online
2
are seen for CH2 and CHD (the J(H,D) coupling is too small
to be resolved). This indicates the formation of ~15% monodeu-
teriated oxirane 14. The mass spectrometric determination of the
deuterium content is rendered difficult due to the presence of a
strong [M - H]+ peak in the EI spectrum of 14, which could not
be brought to disappearance by lowering the electron voltage. The
partly deuteriated oxirane 14 was therefore reduced by LiAlH4
into 1,1-diphenylethanol.18 Mass spectrometric analysis of the
reduction product indicates the presence of 16.0% [D1] and 0.3%
1
[D2]. In the H NMR spectrum of the reduction product, the
methyl group is represented by a singlet at 1.98 ppm (CH3) and
a triplet (CH2D), high field shifted by 0.017 ppm, in the ratio of
~3 : 0.5, which corresponds to ~20% monodeuteration. Because
no base line separation could be achieved, this is in acceptable
agreement with the mass spectrometric determination.
The comparison of the deuterium incorporation into the methyl
group of 7 and the methylene group of 14 shows a marked
difference in the progress of deuterium incorporation in the
methylene transferring reagent dimethylsulfonium methylide (2)
generated from 1 in [D6]DMSO, when 2 is scavenged with 3
(formation of 7) and 12 (formation of 14), respectively. In the first
case, more than 99% of the introduced methyl groups of 7 consist
of CD2X (X = D or H), whereas interception by benzophenone
(12) results in 2,2-diphenyloxirane (14) with less than 20% CHD.
This means that the isotopic exchange between 1 and [D6]DMSO
is more or less complete before dimethylsulfonium methylide (2)
reacts with 2-nitroanisole (3) (Scheme 1). By way of contrast, the
first step of the Corey–Chaykowski reaction (formation of 13,
Scheme 8) is fast enough to scavenge 2 at an early state of the
H/D exchange.
Scheme 6 Possible catalytic cycle for the transfer of a hydrogen atom
from 1 to DMSO.
As discussed above, the ambident nature of the dimsyl anion
(8) may afford the DMSO tautomer 10 as kinetic protonation
product. For H2O as Brønsted acid, this is illustrated in Scheme 7.
If it is presumed that the O-protonation competes efficiently with
the thermodynamically controlled formation of DMSO and that
the DMSO tautomer 10 is consumed in the back reaction of
equilibrium (a) of Scheme 5, the water traces present in the reaction
mixture would initiate a sequence by which hydrogen atoms of the
trimethylsulfonium ion (1) are transferred via H2O and 10 back to
1. Therefore, it is concluded that the involvement of water traces
in the acid–base chemistry of the reaction mixture has no marked
influence on the deuterium labelling results.
Conclusions
Deprotonation of the trimethylsulfonium ion (1) to dimethylsul-
fonium methylide (2) by action of dimsyl anion (8) in DMSO
is accompanied by hydrogen exchange. This follows from the
paradoxical result that the nucleophilic ortho-methylation of 2-
nitroanisole (3) yields essentially unlabelled 7, when reagent 2
is generated from perdeuteriated [D9]1 with the dimsyl anion
in DMSO, whereas the reagent 2 obtained from unlabelled 1 in
[D6]DMSO transfers more than 99% CD2X (X = H, D) into the
introduced methyl group of 7. The acid–base reaction between
the trimethylsulfonium ion (1) and dimsyl anion (8) on one
side and dimethylsulfonium methylide (2) and DMSO on the
other side cannot account for the isotopic interchange, because
for thermodynamic reasons this acid–base reaction is practically
irreversible in terms of the life-time of 2.
To explain this apparent contradiction it is suggested that
the potentially ambident conjugate base of DMSO abstracts
a proton from 1 by its negatively charged oxygen rather than
by its carbanionic center. This replaces DMSO in the acid–
base reaction by its tautomer hydroxy-methylsulfonium methylide
(10), which according to theory is ~100 kJ mol-1 less stable.
Therefore, a reversible deprotonation–reprotonation equilibrium
is established, which in combination with a hydrogen exchange
process could be responsible for the observed labelling results. We
suggest that the DMSO tautomer 10 is also formed by O-mediated
proton transfer from DMSO to its conjugate base 8. The tautomer
10 generated in this way transfers a DMSO-derived H-atom to the
Scheme 7 O-Protonation vs. C-protonation of 8 by H2O.
The mechanistic rationalizations given above describe the
hydrogen exchange as a competitive process to the reaction of
the sulfur ylide 2 with electrophilic substrates like 3. This means
that compounds of different electrophilicity should react with the
sulfur ylide 2 to give products of varying extents of isotopic ex-
change. This was verified by the generation of dimethylsulfonium
methylide (2) in [D6]DMSO followed by addition of benzophenone
(12) according to the Corey–Chaykowski reaction2 (Scheme 8).
Scheme 8 Corey–Chaykowski reaction.
The deuterium incorporation into the methylene group of 2,2-
diphenyloxirane (14) is relatively small. In the 1H NMR spectrum
two singlets at 3.30 ppm and 3.29 ppm in the ratio of ~1 : 0.09
7752 | Org. Biomol. Chem., 2011, 9, 7748–7754
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
trimethylsulfonium ion 1 when it participates in the reprotonation
of dimethylsulfonium methylide (2). The hydrogen exchange
operates in competition with the product-forming reaction of
dimethylsulfonium methylide (2) with electrophilic substrates.
Therefore, the extent of hydrogen exchange in the transferred
methylene group depends on the electrophilicity of the substrates.
From the results presented in this article, it can be concluded
that the hydrogen/deuterium atoms of DMSO/[D6]DMSO can be
exchangeable even in contact with much weaker Brønsted bases
than the conjugate dimsyl anion 8.
Reaction of [13C]methyl-dimethylsulfonium iodide
[13C]1·I-/DMSO/NaH with 3-nitroanisole (3)
Application of [13C]1·I- and DMSO in the above procedure gave
the product [methyl-13C]7 containing ca. 33% 13C in the 3-CH3
1
group; H NMR (400 MHz, CDCl3) d = 2.29 ppm consisting
1
of a singlet and a doublet J(13C,1H) = 129.1 Hz, s : d ratio =
2 : 1); 13C NMR (100 MHz, CDCl3): d = 16.9 ppm (1J(13C,3-13C) =
43.9 Hz).
Partly deuteriated (at C3) 2,2-diphenyloxirane (14)
Experimental
To a mixture of [D6]DMSO (1.5 mL) and THF (1 mL) sodium
hydride (27 mg, 1.12 mMol) is added under nitrogen. After 20
min the vial was immersed in an ice–water bath (5–10 ◦C) and
trimethylsulfonium iodide (1·I-, 225 mg, 1.10 mmol) was added
under stirring. This was followed by addition of benzophenone
(12, 100 mg, 0.55 mmol) in THF (0.5 mL). Usual work-up
furnished 85 mg (79%) 14; m.p. 54–55 C, ref. 2. m.p. 54–56 C;
1H NMR (400 MHz, CDCl3) d = 3.29, 3.30 (two s for CHD and
CH2, ratio ~0.18 : 2), 7.29–7.31 (m, 10H).
Reduction with LiAlH4 according to ref. 18 furnished 1,1-
diphenylethanol, m.p. = 79–80 ◦C, ref. 18. m.p. = 80–81 ◦C.
The 1H NMR spectrum shows a singlet at 1.98 ppm (CH3) and
a triplet at 1.96 ppm (CH2D) for the partly deuteriated methyl
group (ratio ~3 : 0.5). From the molecular ion peak region of the
EI mass spectrum a deuteriation degree of 16% D1 and 0.3% D2
was calculated.
1H and 13C NMR spectra were recorded with a Bruker Avance
spectrometer. EI mass spectra were obtained with a TSQ-70
triple stage mass spectrometer. Commercial anhydrous DMSO
and [D6]DMSO was further dried by sonification over CaH2 for
several hours followed by vacuum distillation. The residual H2O
traces in [D6]DMSO were determined by spiking with toluene
or dioxane and the integration of the 1H signals in the 1H
NMR spectrum. The H2O concentrations obtained range from
20 to 40 ppm.19 The glassware used in the experiments was
thoroughly dried in a stream of nitrogen before use. [13C]Methyl-
dimethylsulfonium iodide ([13C]1·I-) was obtained from dimethyl
sulfide and [13C]methyl iodide (99 atom% 13C) according to the
procedure for the unlabelled sulfonium iodide.2 1H NMR (400
MHz, [D6]DMSO) d = 2.88 ppm (d, 6H (3J(13C,1H) = 3.48 Hz and
d, 3H, 1J(13C,1H) = 144.66 Hz).
◦
◦
Reaction of [D9]1·I-/DMSO/NaH with 2-nitroanisole (3)
Notes and references
To a mixture of anhydrous DMSO (3.25 mL) and anhydrous THF
(H2O < 20 ppm) (1.25 mL)—to avoid freezing—sodium hydride
(30 mg, 1.25 mmol) was added under a nitrogen atmosphere.
After 15 min the mixture was cooled in an ice–water bath
and perdeuteriated trimethylsulfonium iodide ([D9]1·I-, 255 mg,
1.25 mmol) was added in one portion under stirring. At a
temperature < 10 ◦C, 2-nitroanisole (3, 100 mg, 0.65 mmol) in
1 mL DMSO was added dropwise. The reaction mixture was
allowed to warm slowly to room temperature and stirring was
continuedfor 16 h. After mixingwithice water the reaction mixture
was extracted with petroleum ether (40–60 ◦C) (5 ¥ 7.5 mL). The
extracts were washed with brine (5 ¥ 7.5 mL) and dried over
sodium sulfate. Flash chromatography with mixtures of PE (40–
1 V. Aggarwal and J. Richardson, in Science of Synthesis (ed., A. Padwa
and D. Bellus), Thieme Verlag, Stuttgart, New York, 2004, vol. 27, pp.
21–104.
2 (a) V. Franzen and H.-E. Driesen, Chem. Ber., 1963, 96, 1881–1890;
(b) E. J. Corey and M. Chaykowski, J. Am. Chem. Soc., 1965, 87, 1353–
1364.
3 M. Kitano and N. Ohashi, Synth. Commun., 2000, 30, 4247–4254.
4 P. Haiss and K.-P. Zeller, Eur. J. Org. Chem., 2011, 295–301.
5 M. Makosza, Chem. Soc. Rev., 2010, 39, 2855–2868, and references
cited therein.
6 F. G. Bordwell, Acc. Chem. Res., 1988, 21, 456–463.
7 Y. Fu, H.-J. Wang, S.-S. Chong, Q.-X. Guo and L. Liu, J. Org. Chem.,
2009, 74, 810–819.
8 The theoretically derived pKa = 24.5 of ref. 7 is used to estimate the
equilibrium constant of the acid–base reaction (2), although in ref. 6,
an experimental value of pKa = 18.2 is tabulated for the equilibrium
acidity of 1 in DMSO. This value appears problematic, because the
introduction of the (CH3)2S+ group in hydrocarbons should cause an
acidifying effect of ca. 25 pKa units.7,9 Thus, going from methane to 1,
a pKa ≥ 25 should be expected. Suspiciously, a pKa = 18.2 is associated
with the trimethylsulfoxonium ion in the literature10 and in a database.11
Prof. Reich informed us (12 May 2011), that he obtained the value
of 18.2 in a private communication from Bordwell as pKa for the
trimethylsulfoxonium ion and that he considers the trimethylsulfonium
formula in Bordwell’s account6 as a typographical error, because most
probably the pKa of 1 cannot be measured by Bordwell’s method (rapid
decomposition of the conjugate base 2).2.
◦
60 C) and CH2Cl2 furnished unreacted starting material and 1-
methoxy-3-methyl-2-nitrobenzene (7) (yield: 40–45%, calculated
for consumed starting material). For spectral data see ref. 4. The
presence of 5% D in the 3-methyl group follows from a weak triplet
signal at 2.27 ppm overlapping with the strong singlet signal for
CH3 at 2.29 ppm in the 1H NMR spectrum and a corresponding
increase of the M + 1 satellite peak in the EI spectrum.
Reaction of 1·I-/[D6]DMSO/NaH with 2-nitroanisole (3)
9 J.-P. Cheng, B. Liu and X.-M. Zhang, J. Org. Chem., 1998, 63, 7574–
7575.
Replacement of [D9]1·I- by unlabelled sulfonium salt 1·I- and
DMSO by [D6]DMSO resulted in a product consisting of 87.8%
[D3]7, 11.5%[D2]7, and 0.7% [D]7 as derived from the molecular
ion peak region in the mass spectrum. In the 1H NMR the singlet
peak for the 3-CH3 group is absent and replaced by a quintet for
CD2H at 2.255 ppm superimposed by a weak triplet for CH2D.
10 R. Appel, N. Hartmann and H. Mayr, J. Am. Chem. Soc., 2010, 125,
17894–17900.
11 H. J. Reich, Bordwell pKa Table ((Acidity in DMSO),
http://www.chem.wisc.edu/areas/reich/pkatable/).
12 E. Buncel, K.-T. Park, J. M. Dust and R. A. Manderville, J. Am. Chem.
Soc., 2003, 125, 5388–5292.
This journal is
The Royal Society of Chemistry 2011
Org. Biomol. Chem., 2011, 9, 7748–7754 | 7753
©

View Article Online
13 E. D. Glendening and A. L. Shrout, J. Phys. Chem. A, 2005, 109, 4699–
4972.
Functional Groups (ed.: S. Patai and Z. Rappoport), J. Wiley and Sons,
Chichester, 1993, pp. 799–956.
17 We thank the referees for directing our attention to this point.
18 E. L. Eliel and D. W. Delmonte, J. Am. Chem. Soc., 1958, 80, 1744–
1752.
19 The water traces in DMSO cannot be determined by Karl-
Fischer titration, because the relatively large quantities of DMSO
act as oxidizing agent on the iodide formed: W. Fischer,
S. Beil and K.-D. Krenn, J. Prakt. Chem., 1995, 337, 266–
268.
14 In the case of the formally analogous enolates it has been shown that O-
protonation is kinetically favoured over C-protonation; see M. Eigen,
Angew. Chem., 1963, 75, 489–508, (Angew. Chem., Int. Ed. Engl., 1964,
3, 1).
15 F. Turecˇek, J. Phys. Chem. A, 1998, 4703–4713.
16 For the role of the sulfurane intermediates in organic sulfur chemistry,
see: J. Drabowicz, P. Lyzwa and M. Mikolajczyk, in The Chemistry
of Functional Groups Suppl. S, The Chemistry of the Sulfur-Containing
7754 | Org. Biomol. Chem., 2011, 9, 7748–7754
This journal is
The Royal Society of Chemistry 2011
©