Heterogeneous transfer hydrogenation involves pairwise hydrogen transfer from the same position of two molecules of formic acid

Article abstract of DOI:10.1039/a803809k

Using the reduction of an alkyne to cis-alkene as hydrogen trap, differentially deuterium labelled formic acid is shown to deliver a pair of hydrogen atoms either from the formyl or the carboxy position, which suggests that palladium diformate is a key intermediate in heterogeneous transfer hydrogenation.

Full text of DOI:10.1039/a803809k

View Article Online / Journal Homepage / Table of Contents for this issue
Heterogeneous transfer hydrogenation involves pairwise hydrogen transfer
from the same position of two molecules of formic acid
Jinquan Yu and Jonathan B. Spencer*†
University Chemical Laboratory, Lensfield Road, Cambridge, UK CB2 1EW
Using the reduction of an alkyne to cis-alkene as a hydrogen
trap, differentially deuterium labelled formic acid is shown
to deliver a pair of hydrogen atoms either from the formyl or
the carboxy position, which suggests that palladium di-
formate is a key intermediate in heterogeneous transfer
hydrogenation.
provide both hydrogen atoms for the reduction of one molecule
of carbon dioxide to give the same intermediate 1.
To investigate whether 1 is the active intermediate that
provides the hydrogen on the catalytic cycle, the reduction of
the triple bond of phenylpropiolate to the cis-double bond was
employed to trap the hydrogen liberated by the collapse of 1
[Scheme 1(b)]. If differentially deuterium labelled formic acid
(HCO2D or DCO2H) is used then the distribution of deuterium
across the cis-double bond can determine the origin of the
hydrogen transferred to the triple bond. The reduction of a
double bond was not used as a reporter molecule for the
hydrogen transfer because of the possible additional incorpora-
tion of deuterium through isomerization of the double bond
before saturation occurs.15,16
When the reduction was carried out with HCO2D the results
show that the major product is the cis-double bond containing
two hydrogen atoms (Table 1), rather than mono-deuterated cis-
double bond which would be predicted to dominate if both
hydrogen atoms came from the same molecule of formic acid as
proposed previously for the homogeneous system.9 It is
possible that the major product containing two hydrogen atoms
on the double bond could be formed by the preferential donation
of hydrogen from the metal surface (H–Pd–D, intermediate 2,
Scheme 1) owing to a favourable kinetic isotope effect.
However, when 3 was reduced with an equal mixture of
hydrogen and deuterium the results show that there is no
The oxidation of formic acid to carbon dioxide has been widely
used as a source of hydrogen for the reduction of organic
compounds.1–3 This reaction has been shown to be one of the
most effective methods for transfer hydrogenation that could
potentially replace traditional hydrogenation conditions, such as
asymmetric reduction where high pressure is often required.4
The reverse reaction involving conversion of carbon dioxide to
formic acid using catalytic hydrogenation has attracted much
interest in recent years because it offers an environmentally
friendly approach to the use of carbon resources as a raw
material for the chemical industry.5–8 The advantage of using
heterogeneous catalysts in industrial processes has prompted us
to investigate the mechanism of this reversible reaction
catalysed by palladium on carbon.
A major effort has already been directed to the understanding
of the mechanism of this reaction in homogeneous systems with
the formato-(hydrido) metal complex proposed as a key
intermediate 1 [Scheme 1(a)], based on deuterium labelling
studies and ab initio calculations.9–11 Recent studies using
NMR spectroscopy have provided evidence for the existence of
this intermediate,12,13 however, as discussed by Halpern,14 such
thermodynamically stable compounds may not be the active
species on hydrogenation pathways.
Table 1 Reduction of methyl phenylpropiolate
Product distribution (%)a
It has been proposed that the formato-(hydrido) metal
complex 1 collapses to give a metal dihydrido species 2
[Scheme 1(a)] which can then carry out the reduction.9 If this
mechanism is operating in heterogeneous systems then the two
hydrogen atoms from the same formic acid molecule will
always be transferred as a pair in one catalytic cycle.
Correspondingly, when the catalytic cycle proceeds from
carbon dioxide to formic acid, one hydrogen molecule will
Hydrogen source
H + H
H(D) + D(H)
D + D
HCO2Db
DCO2Hb
XCO2Dc
72
25
28
25
10
18
37
51
18
57
35
24
d
H2 + D2
a
b
Errors for the data are ±5%. Labelled formic acids were prepared by
HCl or DCl,
treating sodium formate (either DCO2Na or HCO2Na) with 1
M
extracted with Et2O and dried over anhydrous Na2SO4. Distillation of the
extracts gave DCO2H or HCO2D (99% atom excess as determined by H
(a)
1
O
NMR analysis and mass spectrometry). Pd/C (10%, 38 mg) was added to a
mixture of HCO2D or DCO2H (235 mg, 5.0 mmol), Et3N (2.8 g) and
phenylpropiolate 3 (80 mg, 0.5 mmol) and vigorously stirred under a
blanket of argon. The progress of the reaction was monitored by TLC and
stopped at approximately 40% conversion to avoid formation of the fully
saturated product. This precaution was taken because the protons a to the
carbonyl group of the saturated product can exchange under these
conditions and could potentially lead to scrambling of the label. The
reaction mixture was filtered to remove the catalyst and evaporated to give
residue (81 mg) which was purified by column chromatography (eluted with
hexane–Et2O, 50:1) to afford 25 mg of cis-alkene. The distribution of
deuterium across the double bond of the cis-alkene was determined by 1H
NMR analysis and mass spectrometry. c X = a mixture of D and H (2:1).
M
H
O
+ CO2
M
HCO2H*
M
H
H*
2
H*
1
O
H
(b)
O
D
Ph
CO2Me
Pd
Ph
CO2Me
H
D
Pd/C
3
HCO2D
d
Pd/C (10%, 19 mg) was added to a mixture of Et3N (2.8 g) and
Ph
D
CO2Me
H
H
D
phenylpropiolate 3 (80 mg, 0.5 mmol) and purged with hydrogen and
deuterium (1:1). The reaction mixture was stirred at room temperature and
stopped at approximately 40% conversion. The isolation and analysis of the
cis-alkene was carried out as detailed above.
Pd
+ CO2
Scheme 1
Chem. Commun., 1998
1935

View Article Online
(a)
O
trapped by the triple bond, to give deuterium which can then
mix with the hydrogen from the carboxy end to form HD. The
formation of HD from hydrogen and deuterium on the surface of
the metal has been shown to readily occur in a control
experiment (Table 1).
O
O
CH3
CH3
Pd + 2 CH3CO2H
Pd
+
H2
O
The combined results with HCO2D and DCO2H strongly
suggests that the hydrogen comes directly from the palladium
diformate intermediate 5 rather than intermediate 1, since this
latter species would give the same distribution of deuterium on
the double bond in both cases. The high level of pairwise
addition of hydrogen from HCO2D or deuterium from DCO2H
demonstrates the greater reactivity of the formyl position. It is
noteworthy that there is a higher incorporation of hydrogen
from the formyl position of HCO2D than deuterium from the
same position of DCO2H, suggesting an isotope effect is
involved in the cleavage of the carbon–hydrogen bond of the
palladium diformate (Table 1).
4
O
(b)
O
O
D
Pd + 2 DCO2H
+
H2
Pd
D
O
5
D
The theory was further tested by using a mixture of DCO2D
and HCO2D in a ratio of 2:1 to reduce the triple bond. The
results show that there is a large increase in product containing
one hydrogen and one deuterium compared to the product
obtained using soley HCO2D (Table 1). The increase in
monodeuterated cis-alkene can only be accounted for by
pairwise transfer from the formyl position rather than from the
formato-hydride intermediate 1 which would be predicted to
produce less of the monodeuterated product with the addition of
DCO2D.
The results clearly show that heterogeneous transfer hydro-
genation involves the transfer of a pair of hydrogen atoms either
from the formyl or the carboxy position of two molecules of
formic acid. This provides evidence that palladium diformate is
a key intermediate in this reaction and suggests that the
reduction of carbon dioxide must also proceed through the
diformate intermediate.
Pd
+
2 CO2
D
Scheme 2
substantial isotope effect (Table 1), which was also found in an
earlier study using a homogeneous catalyst.17 Furthermore, this
mechanism cannot account for the fact that more double
deuterated than mono-deuterated cis-double bond is formed.
These results would be consistent with a direct pairwise
hydrogen transfer from either the formyl or the carboxy position
of two different formic acid molecules. It is feasible that two
hydro-formato species 1 on the surface of the metal could line
up in such a way that the two adjacent formyl groups can donate
hydrogen in a pairwise manner. However, palladium is known
to react with acetic acid to form palladium diacetate 4 with the
liberation of hydrogen from the carboxy position [Scheme
2(a)]18 which has prompted us to propose that palladium
diformate 5 is formed in a similar manner and is reponsible for
pairwise transfer of two formyl hydrogens [Scheme 2(b)].
Although palladium diformate has not been isolated, maybe
owing to its instability, other metal diformate species have been
characterised, such as vanadyl diformate,19 triarylbismuth
diformate20 and ruthenium diformate.21 This last example is
particularly important because it was identified during the
hydrogenation of carbon dioxide using a homogeneous ruthe-
nium catalyst.
The formation of the palladium diformate would give one
molecule of hydrogen solely from the carboxy end of the two
formic acid molecules that can be used for the reduction of the
triple bond (Scheme 3). The palladium diformate can then
transfer another pair of hydrogen atoms from the formyl
positions to another triple bond. The results with DCO2H also
show that the major cis-alkene produced from the alkyne
contains two deuterium atoms on the double bond (Table 1).
The minor monodeuterated product is probably formed by
scrambling of the deuterium label. This could occur either via
reduction of the carbon dioxide to give formic acid9 or by
collapse of intermediate 5 [Scheme 2(b)], if not immediately
We thank the Royal Society for the award of a Royal Society
University Research Fellowship (J. B. S.) and the British
Council and the Chinese Government for the award of Sino-
British Friendship Scholarship (J. Y.).
Notes and References
† E-mail: jbs20@cam.ac.uk
1 A. Fujii, S. Hashiguchi, N. Uematsu, T. Ikariya and R. Noyori, J. Am.
Chem. Soc., 1996, 118, 2521.
2 N. Uematsu, A. Fujii, S. Hashiguchi, T. Ikariya and R. Noyori, J. Am.
Chem. Soc., 1996, 118, 4916.
3 K. Matsumura, S. Hashiguchi, T. Ikariya and R. Noyori, J. Am. Chem.
Soc., 1997, 119, 8738.
4 R. Noyori and S. Hashiguchi, Acc. Chem. Res., 1997, 30, 97.
5 E. T. Sundquist, Science, 1993, 259, 934.
6 P. G. Jessop, I. Takao and R. Noyori, Nature, 1994, 368, 231.
7 P. G. Jessop, I. Takao and R. Noyori, Chem. Rev., 1995, 95, 259.
8 W. Leitner, Angew. Chem., Int. Ed. Engl., 1995, 34, 2207.
9 W. Leitner, J. M. Brown and H. Brunner, J. Am. Chem. Soc., 1993, 115,
152.
10 P. G. Jessop, Y. Hsiao, T. Ikariya and R. Noyori, J. Am. Chem. Soc.,
1996, 118, 344.
11 F. Hutschka, A. Dedieu, M. Eichberger, R. Fornika and W. Leitner,
J. Am. Chem. Soc., 1997, 119, 4432.
12 J. Tsai and K. M. Nicholas, J. Am. Chem. Soc.,1992, 114, 5117.
13 E. Lindner and B. Keppeler, P. Wegner, Inorg. Chim. Acta, 1997, 258,
97.
O
O
O
Ph
CO2Me
14 J. Halpern, Science, 1982, 217, 401.
Ph
H
CO2Me
H
3
H
H
15 J. Yu and J. B. Spencer, J. Am. Chem. Soc., 1997, 119, 5257.
16 J. Yu and J. B. Spencer, J. Org. Chem., 1997, 62, 8618.
17 J. A. Osborn, F. H. Jardine, J. F. Young and G. Wilkinson, J. Chem. Soc.
(A), 1996, 1711.
18 S. M. Morehouse, A. R. Powell, J. P. Heffer, T. A. Stephenson and
G. Wilkinson, Chem. Ind., 1964, 544.
Pd
O
5
Pd/C
2 HCO2D
19 D. Mootz and R. Seidel, Acta Crystallogr., Sect. C, 1987, 43, 1218.
20 H. Suzuki, T. Ikegami, Y. Matano and N. Azuma, J. Chem. Soc., Perkin
Trans. 1, 1993, 2411.
21 M. K. Whittlesey, R. N. Perutz and M. H. Moore, Organometallics,
1996, 15, 5166.
D
Ph
CO2Me
Ph
D
CO2Me
D
3
Pd
D
Scheme 3
Received in Liverpool, UK, 20th May 1998; 8/03809K
1936
Chem. Commun., 1998