T. Vielhaber and C. Topf
Applied Catalysis A, General 623 (2021) 118280
combined Mn-mediated dehydrogenation-hydrogenation sequence is
also conceivable [64]. Indeed, when a pristine portion of commercial
d36 was subjected to the reaction conditions from Fig. 5, but in the
absence of H2 gas, the given allyl alcohol was completely converted to
the ketone c36 [65]Fig. 5.
Table 2
Optimization of the reaction conditions for the reduction of acetophenone
catalyzed by the in situ formed manganese catalysta.
2.4. Hydrogenation of cinnamic acid ester derivatives
Entry
[Mn(CO)5Br]
(mol%)
t-BuOK
t
p
Conversion
(%)
Yield
(%)
The pleasing results obtained in the hydrogenation of trans-chalcone
a36 encouraged us to test the in situ-generated [Mn(NN)(CO)3Br] for its
ability to selectively reduce cinnamate derivatives (Table 4). Applying
the optimized reaction conditions established for the acetophenone
hydrogenation resulted in virtually full conversion of the methyl cin-
namate c1 (95 %, Table S6 in the supporting information). On
exchanging THF for 1,4-dioxane as solvent we actually achieved com-
plete substrate conversion while the corresponding saturated ester
product was isolated in very good yield (86 %). The alkyl cinnamates c2-
c4 gave rise to the same excellent yields, in fact irrespective of the de-
gree of branching in the esterifying alcohol. Rather bulky substrates
such as c5 and c6 did not hamper the course of the catalytic trans-
formation either, allowing the synthesis of d5 and d6 in almost quan-
titative yields without debenzylation taking place in the ester part. Quite
remarkably, the in situ-prepared Mn-catalyst smoothly converted allyl
cinnamate c7 into the targeted hydrogenation product without
(mol%)
(h)
(bar)
1
3
9
12
12
12
12
12
5
50
50
40
40
30
20
10
40
40
50
50
67
>99
89
98
>99
42
26
18
79
27
4
67
99
89
97
99
42
25
17
79
27
4
2
3
4
3
3
2
4b
5
3
3
3
3
6
3
3
7
3
3
16
12
12
18
18
8
2.5
2.5
2
5
9
2.5
2
10
11
1
2
a
0.5 mmol of acetophenone were used. Conversions and yields were deter-
mined by GC analysis using n-dodecane as internal standard.
b
The reaction temperature was 100 ◦C. M:L denotes the molar ratio of the
manganese versus the picolylamine-ligand.
–
–
C bond of the allyl motif. However, a higher
compromising the C
–
electronegativity of the attached halide species which is, however,
intrinsically related to the size of the respective atom.
catalyst loading (5 mol% versus 3 mol%) and slightly reinforced reaction
conditions had to be applied in order to achieve this important result
(50 bar H2 and 20 h reaction time versus the standard values of 30 bar H2
and 12 h).
Our Mn-based catalytic system also mediated the reduction of
pharmaceutically relevant CF3-tagged acetophenones, viz. a22 and its
disubstituted congener a24, to generate the corresponding
trifluoromethyl-alcohols in excellent yields. Interestingly, we observed a
subtle but decisive effect of the alkyl chain length that is in direct
proximity of the ketone group. If we subjected the immediate homolo-
gous propiophenones a23 and a25 to the same hydrogenation process,
the catalytic transformation was sluggish and the alcohols b23 and b25
were produced in very low yields (17 % and 23 %, respectively).
Along with aromatic ketones, (benzannulated) cyclic substrates that
incorporate the carbonyl group directly in the ring (b26-b30) were also
completely converted to afford the corresponding secondary alcohol in
very good yields. Notably, the ring size affected the catalyst activity in
that six-membered rings required higher H2 pressure and prolonged
reaction time (b26 and b27 versus b28).
Placing electron-releasing groups on the phenyl ring of the cinnamic
acid ester did not markedly affect the catalyst performance. Yet, in case
of the methyl cinnamates c8 and c9 the conditions had to be modified
(50 bar H2, 100 ◦C and 20 h) so as to maintain full substrate conversion
and decent yield.
Regarding the halide-functionalized esters c11-c14 we did not
observe any unwanted hydro-dehalogenation processes for the fluoro-
and chloro-compounds whereby the corresponding products were iso-
lated in good yields ranging from 80 % to 89 %. Most notably, the
respective bromo-derivatives were also amenable to the given hydro-
genation which is in stark contrast to the results obtained for the
brominated acetophenones a18-a20 (vide supra). However, the reaction
conditions had to be adjusted (6 mol% catalyst loading, 50 bar, 24 h,
100 ◦C) which again demonstrated a somewhat detrimental decreased-
electronegativity-effect of the appendant bromide substituent. With
respect to d13 and d14 we detected 3-phenyl methyl cinnamate in the
GC–MS spectra which is indicative of concomitant hydro-debromination
that occurred during the course of the Mn-catalyzed hydrogenation.
The introduction of a cyano group on the arene ring was not as well
accommodated as was the case with the related CN-labelled acetophe-
none a33 (vide infra) and hence cinnamate d15 was only isolated in a
mediocre yield of 75 %. Diesters c16 and c17 displayed considerable
reactivity towards the conjugate hydrogenation of their C = C-bonds
though the yield of the former was surprisingly low considering its
rather simple molecular architecture.
To our regret, methyl ketones with an attached pyridine core were
either not at all (a31) or poorly (a32, a33) converted into the desired
products b31-b33 presumably due to inhibition of the active Mn-center
through the strongly ligating sp2 nitrogen atom of the pertinent
heterocycle.
Finally, the chemoselectivity of the manganese in situ-system was
examined by testing substrates that incorporate one additional reducible
motif, viz. CN-functionalized acetophenone a34, isophorone a35, and
the aldehyde-tagged ketone derivative a37 (see Section 4 in the sup-
porting information). With the exception of the latter, the ketone group
was selectively reduced to produce the target alcohols b34 and b35 in
very good yields. On the contrary, the aldehyde-bearing acetophenone
a37 was exhaustively hydrogenated to the bis-alcohol b37. Quite
remarkably, on applying benzaldehyde derivatives devoid of an acetyl
group on the arene moiety as substrates, we observed either only traces
of the respective primary alcohol or no product at all.
The most surprising feature of the in situ-prepared [Mn(NN)(CO)3Br]
complex is its ability to even reduce tetra-substituted C = C-bonds at
100 ◦C in crowded esters such as c18 and c19. Upon applying a H2-
pressure of 50 bar and a reaction time of 20 h the isolated yields of both
desired products were well above 70 %. Accordingly, the conjugated
In stark contrast to the selectivity observed in the case of the
reduction of isophorone 34a, the hydrogenation of trans-chalcone a36
produced a mixture of saturated alcohol b36 and ketone c36 (Fig. 5).
Since no evidence for the formation of allyl alcohol d36 was found, it is
tempting to assume that the given Mn-catalyzed transformation is
considerably biased towards the reduction of the C = C-bond. However,
the intermediate formation of d36 through hydrogenation of the ketone
group in enone a36 and its direct isomerization to the ketone c36 via a
C = C-bond of the less sterically congested α-phenyl cinnamyl nitrile c20
was fully hydrogenated upon applying lower H2-pressure (30 bar) and a
shorter reaction time (12 h).
To further explore the limitations of the Mn-based in situ-system
presented herein, we expanded the substrate panoply by ethyl-
β-methyl cinnamate c21 (Table 5). Regrettably, we soon realized that
this compound cannot compete with those listed in Table 4 with respect
5