Angewandte
Communications
Chemie
the ratio of the two alcohol substrates. In the Mn-catalyzed
which enables slow release of 5a and thus ensures high
reaction, a mixture of alkene products was obtained as a result
of the non-selective aldol condensation of the two aldehyde
intermediates generated by alcohol dehydrogenation
(Scheme 1). Nevertheless, the subsequent Ni-catalyzed
hydrogenation of the mixed alkenes gave only the reduced
alkanes 4 (Scheme 3). Awide range of functional groups were
compatible in this reaction sequence, and various unsym-
metrical 1,3-diarylpropanes (4a–i) and functionalized 1-
arylalkanes (4j–s) were obtained in good yields. However,
deoxygenative coupling of the benzyl-substituted arylethanol
gave a low yield of 4t because steric hindrance decelerated
the aldol condensation step. Compared with the reported
stepwise catalytic homocoupling of arylalkanols to diaryl-
alkanes with the use of two different Ir catalysts,[12] this
protocol catalyzed by earth-abundant metals (Mn and Ni) is
more economical and broadly applicable.
efficiency of the following condensation step.
Further mechanistic insights were gained by performing
kinetic studies (Figure 1). The kinetic profile of the standard
To shed light on the reaction mechanism, control experi-
ments and kinetic studies were conducted. It is noteworthy
that Stoermer and co-workers reported the condensation of
phenylacetaldehyde (5a) in basic ethanol solution to produce
1,3-diphenylpropene (2a) in high yield.[17] Nevertheless, this
protocol is highly sensitive to various substituents on the
phenyl ring,[18] which highlights the advantages of our
reported catalytic transformation (see Table S3). On the
basis of this known reaction, we hypothesized that the Mn-
catalyzed dehydrogenation of 1a produced the reaction
intermediate 5a, which underwent a base-mediated conden-
sation to give the final product 2a. To verify this hypothesis,
we first investigated the reactivity of 5a under standard
reaction conditions with or without the Mn catalyst (Scheme
4a,b). Interestingly, 2a was obtained with the same yield
(50%) in these two reactions, and supported aldehyde
intermediate 5a as an intermediate for this Mn-catalyzed
transformation, as well as excluded the assistance of Mn
Figure 1. Kinetic studies.
reaction of 1a to 2a is shown in Figure 1a. No induction
period was observed, and neither intermediate 5a nor 6a was
detected by GC analysis during the entire reaction process. It
illustrated that the dehydrogenation of 1a was much slower
À
catalyst for the condensation of 5a and the following C C
À
bond cleavage process. In addition, another possible reaction
intermediate, 6a, generated by aldol reaction of 5a, was
smoothly transformed into the desired product 2a in 84%
yield without Mn catalyst (Scheme 4c), indicating 6a to be
a key intermediate in this reaction. Based on this observation,
we postulated that this catalytic deoxygenative coupling
reaction is induced by Mn-catalyzed dehydrogenation of 1a,
than the subsequent aldol condensation and C C bond
cleavage steps. This outcome is consistent with our hypothesis
regarding the mechanism for selectivity control as mentioned
above. For comparison, the kinetic behavior of the conden-
sation reaction of 5a to 2a was also studied under catalyst-
free conditions (Figure 1d). Full conversion of 5a and a 38%
yield of 2a were obtained, with a much higher initial rate.
Interestingly, the selectivity and yield of this transformation
could be increased by enhancing the amount of base from 0.5
to 2 equivalents, albeit with similar reaction rate (Figure 1c).
Because of the poor solubility of NaOH in xylene and larger
number of undefined side-reactions of 5a at high temper-
atures, kinetic studies of the known base-mediated conden-
sation reactions in ethanol at lower temperatures were
performed.[17] Specifically, a low reaction rate and yield
were obtained with 0.5 equivalents of NaOH, the same
amount of base as in the catalytic reaction (Figure 1e). In
contrast, when 2 equivalents of base were used, the reaction
rate and yield were significantly improved because of the
better solubility of NaOH in ethanol (Figure 1b). Given these
results, we concluded that the concentration of NaOH must
be much higher than that of 5a to avoid side-reactions during
the condensation process. This condition is achieved by the
Scheme 4. Control experiments to identify reaction intermediates.
Angew. Chem. Int. Ed. 2018, 57, 1 – 6
ꢀ 2018 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
3
These are not the final page numbers!