DOI: 10.1002/cctc.201500493
Communications
Reductive Transformations of Carbonyl Compounds
Catalyzed by Rhodium Supported on a Carbon Matrix by
using Carbon Monoxide as a Deoxygenative Agent
Niyaz Z. Yagafarov,[a] Dmitry L. Usanov,[b] Alexey P. Moskovets,[a] Nikolai D. Kagramanov,[a]
Victor I. Maleev,[a] and Denis Chusov*[a]
An efficient method for the rhodium on carbon matrix cata-
lyzed preparation of secondary and tertiary amines, cyanoest-
ers, and nitriles through the reductive amination/alkylation of
carbonyl compounds was developed, including a convenient
procedure for the tandem formal reductive addition of acetoni-
trile to aldehydes. The catalyst could be reused, and at least
three consecutive reaction cycles were performed with compa-
rable efficiency. The method was shown to be compatible with
functional groups prone to reduction by hydrogen and com-
plex hydrides.
while working with carbon monoxide because of its toxicity, its
higher flammability limit relative to that of hydrogen repre-
sents an important advantage especially on a larger reaction
scale.[5] Whereas the use of CO has been systematically investi-
gated in the reduction of, for example, aromatic nitro
groups,[6,7] its potential as a reductive agent still remains con-
siderably underexplored.
Heterogeneous catalysis is oftentimes much more conven-
ient with respect to homogeneous alternatives, for example, in
terms of catalyst recovery and purification of the products;[8]
and the intrinsic merits of heterogeneous catalytic systems are,
therefore, responsible for wide applications thereof in nearly
all areas of chemical industry owing to the improved economic
and environmental profiles. Given that our growing toolbox of
methods employing CO as a deoxygenative agent have so far
involved homogeneous protocols,[2] we were particularly inter-
ested in the development of a heterogeneous system of com-
parable potential that would render our method more compat-
ible with the general requirements of industrial-scale produc-
tion. Thus, herein we report an atom-economical heterogene-
ous catalytic system equally applicable to the reductive forma-
tion of both CÀN and CÀC bonds that employs rhodium
supported on carbon matrix as a catalyst and carbon monox-
ide as a deoxygenative agent.
Reductive CÀC and CÀN bond-forming reactions are widely
used as powerful and convenient tools to build up molecular
complexity rapidly.[1] Whereas in the paradigm of fine laborato-
ry synthesis classical synthetic approaches are being constantly
improved and refined to meet ever-increasing stringency of re-
quirements for reactivity, selectivity and cost, we strongly be-
lieve that the academic community should pay greater atten-
tion to optimization of synthetic tools in terms of the integral
economic and environmental effects they might have if con-
ducted on semi-industrial or industrial scale. In this context,
the use of side products of the chemical industry as starting
materials for novel synthetic protocols represents a very valua-
ble concept. Our group recently discovered catalytic reductive
methodology[2] that takes advantage of the deoxygenative po-
tential of carbon monoxide and does not require an external
hydrogen source, unlike conventional approaches such as re-
ductive amination and reductive Knoevenagel condensation.
As carbon monoxide is produced in multiton quantities as
a side product of steelmaking,[3] it represents an abundant
source for chemical synthesis and is a more atom-economical[4]
alternative to many of the established synthetic agents. More-
over, despite the fact that certain precautions have to be taken
For the initial studies on reductive amination, we decided to
work on the formation of a reduced adduct of p-fluorobenzal-
dehyde with p-anisidine as a model reaction; the choice of
a fluorinated aldehyde as a substrate was made primarily be-
cause of the convenience of NMR spectroscopy monitoring. Ex-
pectedly, no reductive amination reaction proceeded in the
presence of unmodified activated carbon, whereas the use of
activated carbon-supported ruthenium and rhodium (5% w/w,
0.25 mol% metal) led to the detection of a trace amount of
the desired product (Table 1, entries 2 and 3).
Rhodium on the carbon support of less developed surface
area demonstrated marginally greater yield, which could prob-
ably be explained by a lower degree of product absorption
(Table 1, entry 4). The use of alumina-supported metals did not
lead to an improvement in reaction efficiency (Table 1, en-
tries 5 and 6). We then decided to switch to matrix supports
(Table 1, entries 7 and 8); to our delight, rhodium supported
on carbon matrix (Degussa Co.) demonstrated substantially im-
proved performance: the desired product was formed in 37%
yield, which increased up to 82% by changing the catalyst
loading to 1 mol% (Table 1, entry 10). Solvent screening for the
process catalyzed by Rh/carbon matrix showed superiority of
[a] N. Z. Yagafarov, A. P. Moskovets, N. D. Kagramanov, Prof. Dr. V. I. Maleev,
Dr. D. Chusov
A. N. Nesmeyanov Institute of Organoelement Compounds of the
Russian Academy of Sciences
Vavilova Street 28, Moscow 119991 (Russia)
[b] Dr. D. L. Usanov
Department of Chemistry and Chemical Biology
Harvard University
12 Oxford Street, Cambridge, MA 02138 (USA)
Supporting Information for this article is available on the WWW under
ChemCatChem 0000, 7, 2590 – 2593
2590
ꢀ 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim