A. Corma, M. J. Sabater et al.
aldehyde takes place through ionic intermediate species.[15a]
If this is so, a further decrease in the formation of 2a could
be achieved by adding a radical scavenger to the reaction
system. Then, phenyl N-tert-butylnitrone (PBN) was used as
radical scavenger and, surprisingly, selectivity to 2a did not
decrease but increased, whereas the rate of formation of
for 1 h to ensure the formation of metal hydrides from the
very beginning. The total amount of palladium hydrides
(0.6 mmol) was estimated on the basis of the decrease in hy-
drogen pressure. Then, stoichiometric amounts of benzalde-
hyde and benzenethiol were reacted to give the expected
thioether 1a as major product, the hydrogenolysis product
toluene as secondary product, and traces of the hydrogena-
tion product trifluoromethylcyclohexane. Disulfide 2a was
also detected at the trace level. These results confirm that
formation of thioether 1a will effectively occur through the
direct reductive thiolation of in situ generated aldehydes
with thiols by palladium hydride species according to
Scheme 3. Then, to discern the nature of the key intermedi-
ate that will be reduced during the reaction, compound II
was synthesized[17] and reacted with benzyl alcohol (1 mmol)
in the presence of Pd/MgO at 1808C. Under these experi-
mental conditions thioacetal II remained unreacted after
8 h, albeit 1a could be obtained with low yields (ꢀ5%)
when the thioacetal was treated with H2.
thioACHTUNGTRENNUNGether 1a was strongly inhibited (Table 1, entry 5) up to
the point that in the presence of PBN, 2a became the major
reaction product. It can be assumed that the strong inhibit-
ing effect of PBN for the formation of the thioether is
a clear indication that radical intermediates are formed in
some elementary step of the reaction. Following the above,
and because the nucleophilic attack of the thiol to the alde-
hyde has been shown to proceed through ionic intermedi-
ACHTUNGTRENNUNG
ates,[15b] the obvious conclusion is that the dehydrogenation
of benzyl alcohol to benzaldehyde must involve radical in-
termediates. This conclusion is in agreement with mechanis-
tic studies on the alcohol dehydrogenation reaction based
on DFT calculations.[13c] Moreover, the fact that the forma-
tion of the disulfide 2a was not affected by the presence of
the radical scavenger indicates that this reaction involves
partially ionic intermediates.
Mechanistically the formation of the thioether 1a can be
explained by nucleophilic attack of benzenethiol to the
benzACHTUNGTRENNUNGaldehyde formed after dehydrogenation of the alcohol,
This experimental fact suggests that the major pathway is
most likely formation of the thionium ion III from the hemi-
thioacetalic compound I, followed by a very rapid hydrogen-
ation of III by the palladium hydrides (see Scheme 3).[10] In
ACHTUNGTRENNUNG
fact, these results are also consistent with the observed ki-
netics: first-order dependence on benzaldehyde and first-
order dependence on benzenethiol (see below).
which gives a hemithioacetal intermediate (RCH(OR)SR;
I) in equilibrium with thioacetal (II; not detected by gas
chromatography (GC)) if water is present (see Scheme 3).
Taking into account that thioacetals can be desulfurized to
Rate-controlling step: The study of the reaction scheme and
mechanism of the synthesis of thioethers through the direct
reductive thiolation of in situ
generated aldehydes as well as
the study of the scope of the re-
action were carried out with
a
PdACTHNGUETRNNU(G 0.8 wt%)/MgO catalyst
(Pd average particle size, d=
2 nm; see Figure 3S in the Sup-
porting Information). In a first
approach, the elucidation of the
rate-determining step in the
global process could be helpful
to indicate what reaction step
should be accelerated and con-
sequently what modifications
should be made on the catalyst
to improve activity and selectiv-
ity.
Scheme 3. Plausible mechanistic pathways for the palladium-catalyzed synthesis of 1a starting from benzyl al-
cohol and benzenethiol.
To do that, the two elementa-
the corresponding sulfides or hydrocarbons by using diverse
reducing agents (versus H2 and Raney nickel, pyridine-
borane in trifluoroacetic acid, and so forth),[16] it may very
well occur that the final product 1a could be formed by re-
ductive hydrogenolysis of the thioacetal II by the metal hy-
drides formed on the metal surface, or even by reduction of
the hemithioacetal I and/or of a thionium ion (III) inter-
mediate formed during the reaction.
ry steps for the S-monoalkylation of benzenethiol with
benzyl alcohol, that is, alcohol dehydrogenation to afford
benzaldehyde and reductive thiolation of the latter to form
the thioether 1a, have been formulated by assuming that
the hemithioacetal (or the thionium ion), which was not de-
tected by GC, is an intermediate that reacts very fast (see
Scheme 4).
Two kinetic rate expressions were derived by assuming
that the dehydrogenation of benzyl alcohol to give benzal-
dehyde and a metal hydride [Eq. (1)], or the reductive
To test this hypothesis, a fixed amount of Pd/MgO
(0.8 wt% Pd) was treated initially with H2 (5 bar) at 1808C
17466
ꢃ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2013, 19, 17464 – 17471