Silver-catalysed protodecarboxylation of carboxylic acids

Article abstract of DOI:10.1039/b912509d

A silver-based catalyst system has been discovered that effectively promotes the protodecarboxylation of various carboxylic acids at temperatures of 80-120 °C - more than 50 °C below those of the best known copper catalysts. The Royal Society of Chemistry 2009.

Full text of DOI:10.1039/b912509d

View Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Silver-catalysed protodecarboxylation of carboxylic acidsw
Lukas J. Gooßen,* Christophe Linder, Nuria Rodrıguez, Paul P. Lange and Andreas Fromm
Received (in Cambridge, UK) 25th June 2009, Accepted 7th October 2009
First published as an Advance Article on the web 28th October 2009
DOI: 10.1039/b912509d
A silver-based catalyst system has been discovered that effectively
promotes the protodecarboxylation of various carboxylic acids at
temperatures of 80–120 1C—more than 50 1C below those of the
best known copper catalysts.
acids in which only catalytic amounts of a copper mediator
are necessary.9 The new protocol involves a copper/1,10-
phenanthroline-complex as the catalyst for an effective
decarboxylation of a wide spectrum of aromatic carboxylic
acids within 16 h at 170 1C. Independently, Kozlowski
et al.10 disclosed a catalytic protodecarboxylation protocol
for particularly electron-rich bis-ortho-substituted aromatic
carboxylic acids.11 It proceeds readily at 70 1C, but involves
the use of as much as 20 mol% of expensive Pd(O2CCF3)2 as
the catalyst.
In recent years, numerous catalytic transformations have been
developed that have substantially extended the use of
carboxylic acids as building blocks in organic synthesis.1
Decarboxylative coupling reactions in particular are of high
preparative utility, as they allow using carboxylate salts as
substitutes for costly organometallic reagents in C–C bond
forming reactions. In these transformations, the carbon
nucleophiles are generated via extrusion of CO2 from carboxylate
salts and can subsequently be coupled with a range of carbon
electrophiles.2 The decarboxylation of carboxylic acids is
also the initiating step in preparatively valuable oxidative
couplings, e.g. decarboxylative Heck reactions,3 C–S couplings,4
and tandem decarboxylation/C–H activation reactions.5 To
date, the practical applicability of all these transformations is
limited by their requirement of rather high temperatures, as
well as their tolerance of only a narrow spectrum of functional
groups. Considering that the extrusion of CO2 is believed to be
the rate-determining step in all these processes, the development
of more effective decarboxylation catalysts is the key towards
lowering the reaction temperatures and thereby extending the
scope of these transformations to more sensitive derivatives.
The protodecarboxylation of aromatic carboxylates is an ideal
test reaction for the development of more effective co-catalysts
for decarboxylative coupling reactions (Scheme 1), as it is
easily performed and mechanistically relatively well-understood.
Moreover, it is itself an important synthetic tool, allowing
removal of surplus carboxylate groups left behind as a result
of the chosen synthetic route. The decarboxylation of simple
arenecarboxylates usually involves extreme reaction temperatures.
Moreover, in the early protocols, stoichiometric amounts of
copper6 or silver7 complexes are required to mediate this
process. Mercury-mediated protodecarboxylations readily
proceed at temperatures of 100 1C, but their synthetic utility
is limited by the toxicity of the organomercury(II) compounds.8
Recently, we disclosed a protodecarboxylation of arenecarboxylic
We herein present an effective and economical silver-based
low-temperature protodecarboxylation catalyst.12 In our
search for a new catalyst generation, we performed numerous
DFT calculations modelling the extrusion of CO2 from
various copper and silver carboxylate complexes. In this
context, we calculated the structures and energies for the
starting materials, transition states and products for the
decarboxylation of both a 1,10-phenanthroline/copper(I)-2-
fluorobenzoate complex and the corresponding silver(I) system
(Table 1).13 Under standard conditions at 298 K, the extrusion
of CO2 from the copper carboxylate 3a is endergonic and has
an activation barrier of DGa
= 31.1 kcal molꢀ1. For the
silver(I) carboxylate 30a, this transformation is exergonic
298
and has a significantly lower activation barrier of only of
DGa = 29.6 kcal molꢀ1
.
298
As phenanthroline is not a common ligand for silver, we
suspected at first that this result might be an artefact caused
by an unrealistic ligand environment. We thus performed
additional calculations on the decarboxylation of a silver(I)-
2-fluorobenzoate complex which is ligated solely by two NMP
molecules (5). The structures and energies obtained are depicted
Table 1 Relative energies for the decarboxylation of 2-fluorobenzoic
acid with silver and copper catalystsa
M
DER
DGR
DEa
DGa
298
tot
298
tot
Scheme 1 Metal-mediated decarboxylation of benzoic acids.
3a - 4a
Cu
Ag
13.91
11.04
2.10
ꢀ0.45
32.36
31.36
31.11
29.56
30a - 40a
a
Computational
conditions:
Gaussian03;14
B3LYP15
/
FB Chemie – Organische Chemie, TU Kaiserslautern,
Erwin-Schroedinger-Strasse, Geb. 54, 67663, Kaiserslautern, Germany.
E-mail: goossen@chemie.uni-kl.de; Fax: +49 (631) 205-3921
w Electronic supplementary information (ESI) available: Experimental
data, procedures for all compounds and computational details. See
DOI: 10.1039/b912509d
6–311+G(2d,p)16//B3LYP/6–31G(d)17 for H, C, N, O, F; Stuttgart
RSC 1997 ECP18 for Cu, Ag, scaling factor for thermal corrections:
f = 0.9804.19 DERtot, DGR298, DEa and DGa
are expressed in
298
tot
kcal molꢀ1
.
ꢁc
This journal is The Royal Society of Chemistry 2009
Chem. Commun., 2009, 7173–7175 | 7173

View Online
Table 2 Optimization of the reaction conditionsa
Entry Catalyst Ligand
Additive Solvent
T/1C 2a(%)
1
2
3
4
5
6
7
8
Cu2O
Ag2O
phen
phen
—
—
—
—
—
—
—
—
—
—
—
—
—
—
—
NMP/quin 170 24
00
00
00
22
85
00
00
NMP
00
00
00
00
00
00
00
120 60
00
00
00
00
00
00
00
00
00
00
00
00
00
AgOTf
AgO
Ag2CO3
AgOAc
00
0
20
00
00
39
70
51
47
46
21
11
0
24
98
57
77
79
Fig.
1 Calculated reaction path for metal-catalyzed proto-
00
decarboxylation.
00
9
5-NO2-phen
4,7-di-Ph-phen
2,20-bipyridine
quin
AsPh3
PPh3
—
—
—
—
—
—
—
00
00
00
00
00
00
00
00
00
00
00
00
00
10
11
12
13
14
15
16
17
18
19b
20
21c
in Fig. 1. The reaction to the NMP-stabilized 2-fluorophenyl
silver complex 6 was found to also be exergonic (ꢀ9.3 kcal molꢀ1),
and the activation barrier was as low as 28.8 kcal molꢀ1. We
were positively surprised by these results, as silver complexes
had so far been inferior to copper complexes in all our
experimental decarboxylation studies. Moreover, silver is
an untypical mediator for protodecarboxylations,20 and is
required in overstoichiometric amounts as a co-mediator in
decarboxylative couplings.2,3,5
00
00
00
00
00
00
00
00
00
K2CO3
00
00
00
00
00
DMSO
DMF
NMP
00
100 17
80 60
00
Intrigued by the predictions of our DFT studies, we
performed a series of test reactions using the decarboxylation
of 2-methoxybenzoic acid as a model system. This substrate
was chosen due to its similar electronic properties to the
2-fluoro derivative used in the calculations, but leading to a
less volatile product that is more easily analyzed and isolated.
In order to get a first idea of the relative reaction rates
for copper- and silver-based catalysts, we compared the
respective conversions for the protodecarboxylation of
2-methoxybenzoic acid obtained at 10 mol% catalyst loading
after 15 min at 140 1C. A copper(I) oxide/1,10-phenanthroline
complex gave only trace amounts of the product anisole (2a),
whereas the corresponding silver(I) oxide/1,10-phenanthroline
system afforded 20% conversion. This result confirms the
improved catalytic activity of silver(I) over copper(I).
a
Reaction conditions: 1 mmol 2-methoxybenzoic acid, 10 mol% catalyst
(5 mol% Cu2O, Ag2O and Ag2CO3), 10 mol% ligand, 15 mol% additive,
2 mL solvent, 16 h. Conversions were determined by GC analysis using
n-tetradecane as internal standard. phen
b
quin = quinoline. After 8 h. Using 2-nitrobenzoic acid.
= 1,10-phenanthroline,
c
carbonate bases as additives. Unligated silver in the presence
of 15 mol% potassium carbonate, in particular, indeed led to
substantially improved yields over all silver/ligand combinations
(entry 16). Comparative experiments confirmed that NMP was
the best choice of solvent (entries 17–19).
Under thus optimized conditions, 79% of the product was
formed already within only 8 h reaction time (entry 19).
Any further reduction in temperature considerably affected
the yields for this particular substrate (entry 20). However,
activated benzoic acids such as 2-nitrobenzoic acid gave
satisfactory yields even at a reaction temperature of only
80 1C (entry 21).
We next investigated the generality of the optimized protocol
using various aromatic and heteroaromatic carboxylic acids at
a temperature of 120 1C. Selected results are summarized in
Table 3. The yields are shown in comparison to those obtained
with Cu2O/phenanthroline at 170 1C.9 We were pleased to find
that the protodecarboxylation proceeded in high selectivities
with, at most, traces of side products arising from homocoupling
or transesterification reactions. The reaction time of 16 h was
sufficient to achieve good conversions for most substrates. The
examples demonstrate that the scope of the new Ag-catalyst is
broader than that of the published Pd-system.10 Moreover, its
performance nicely complements that of the Cu-catalyst, the
latter being particularly suited for meta- and para-substituted
benzoic acids, while the Ag-system is highly effective for
many ortho-substituted or heterocyclic derivatives that gave
low or no yields with the Cu-catalyst. Examples include
We next optimized the catalyst and conditions in a series of
screening experiments. Selected results are summarized in
Table 2. As previously reported, the protodecarboxylation of
the deactivated 2-methoxybenzoic acid with Cu(I) oxide/1,10-
phenanthroline in a mixture of NMP and quinoline at 170 1C
was low. Under identical conditions, silver(I) oxide initially
gave disappointing results, but the reaction became very
effective in the absence of the basic co-solvent quinoline
(entries 2, 3). The rate acceleration achieved by using silver
allowed us to reduce the reaction temperature to 120 1C
without obtaining substantially decreased yields (entry 4).
The counter-ion on the silver had a marked effect on the
reaction outcome, the basic acetate giving best results
(entries 5–8). Some of the mono- and bidentate nitrogen
ligands, triarylphosphines and -arsines tested including 1,10-
phenanthroline were found to have a beneficial effect on the
reaction yields (entries 9–14) in comparison to silver acetate
alone (entry 15). Suspecting that the principal role of such
ligands might be to act as a mild base, we also investigated
ꢁc
This journal is The Royal Society of Chemistry 2009
7174 | Chem. Commun., 2009, 7173–7175

View Online
Table 3 Scope of the transformationa
may be combined with other types of more temperature-
sensitive functionalizations.
We thank W. R. Thiel and C. van Wullen for help with the
DFT calculations, the Deutsche Forschungsgemeinschaft,
Saltigo GmbH, and NanoKat for funding, and the A. v.
Humboldt Foundation for a scholarship to N. R.
AgOAc: Cu2O:
Ar–H Yield/% Yield/%
Entry
Ar–COOH
1
2
3
4
5
6
7
8
1a 2-MeO–C6H4–COOH
1b 2-NO2–C6H4–COOH
1c 2-NO2-5-MeO–C6H3–COOH 2c
1d 2,4,5-MeO–C6H2–COOH
1e 2,4-MeO–C6H3–COOH
1f 2,6-MeO–C6H3–COOH
1g 2-Br–C6H4–COOH
2a
2b
83
92
88
43
89
87
76
95
76
74
85
89
(91)
71
58
(11)
17
45
(80)
(36)
(76)
(41)
(38)
(14)
249
879
909
—
—
—
0
—
0
—
Notes and references
1 L. J. Gooßen, N. Rodrıguez and K. Gooßen, Angew. Chem., Int.
Ed., 2008, 47, 3100; L. J. Gooßen, K. Gooßen, N. Rodrıguez,
M. Blanchot, C. Linder and B. Zimmermann, Pure Appl. Chem.,
2008, 80, 1725.
2d
2e
2e
2f
2 C. Peschko, C. Winklhofer and W. Steglich, Chem.–Eur. J., 2000,
6, 1147; L. J. Gooßen, G. Deng and L. M. Levy, Science, 2006, 313,
662; P. Forgione, M. C. Brochu, M. St-Onge, K. H. Thesen,
M. D. Bailey and F. Bilodeau, J. Am. Chem. Soc., 2006, 128,
11350; L. J. Gooßen, N. Rodrıguez, B. Melzer, C. Linder, G. Deng
and L. M. Levy, J. Am. Chem. Soc., 2007, 129, 4824; J.-M. Becht,
C. Catala, C. Le Drian and A. Wagner, Org. Lett., 2007, 9, 1781;
L. J. Gooßen, F. Rudolphi, C. Oppel and N. Rodrıguez,
Angew. Chem., Int. Ed., 2008, 47, 3043; L. J. Gooßen,
B. Zimmermann and T. Knauber, Angew. Chem., Int. Ed., 2008,
47, 7103; L. J. Gooßen, N. Rodrıguez and C. Linder, J. Am. Chem.
Soc., 2008, 130, 15248; J. Moon, M. Jeong, H. Nam, J. Ju,
J. H. Moon, H. M. Jung and S. Lee, Org. Lett., 2008, 10, 945.
3 A. G. Myers, D. Tanaka and M. R. Mannion, J. Am. Chem. Soc.,
2002, 124, 11250; A. Maehara, H. Tsurugi, T. Satoh and
M. Miura, Org. Lett., 2008, 10, 1159; P. Hu, J. Kan, W. Su and
M. Hong, Org. Lett., 2009, 11, 2341.
4 Z. Duan, S. Ranjit, P. Zhang and X. Liu, Chem.–Eur. J., 2009, 15,
3666.
5 A. Voutchkova, A. Coplin, N. E. Leadbeater and R. H. Crabtree,
Chem. Commun., 2008, 6312; C. Wang, I. Piel and F. Glorius,
J. Am. Chem. Soc., 2009, 131, 4194.
6 A. F. Shepard, N. R. Winslow and J. R. Johnson, J. Am. Chem.
Soc., 1930, 52, 2083; M. Nilsson, Acta Chem. Scand., 1966, 20, 423;
T. Cohen and R. A. Schambach, J. Am. Chem. Soc., 1970, 92,
3189; A. Cairncross, J. R. Roland, R. M. Henderson and
W. F. Shepard, J. Am. Chem. Soc., 1970, 92, 3187.
1h 2-Br-4,5-MeO–C6H2–COOH 2g
9
1i 2,6-Cl–C6H3–COOH
1j 2,4-Cl–C6H3–COOH
1k 2-Cl–5-NO2–C6H3–COOH
1l 2-MeS(O)2–C6H4–COOH
1m 2-CF3–C6H4–COOH
1n 2-iPrOC(O)–C6H4–COOH
1o 2-Ac–C6H4–COOH
1p 2-CN–C6H4–COOH
1q 4-HO–C6H4–COOH
1r 1-naphthyl–COOH
1s 2-thienyl–COOH
2h
2h
2i
2j
2k
2l
2m
2n
2o
2p
2q
2q
2r
2s
2a
2a
10
11
12
13
14
15
16
17
18
19
20
—
609
(22)9
829
879
—
—
839
(58)9
—
1t 3-thienyl–COOH
21b,c 1u cinnamic acid
—
—
—
809
22c
23d
24d
1v 2-thiopheneglyoxylic acid
1w 3-MeO–C6H4–COOH
1x 4-MeO–C6H4–COOH
a
Reaction conditions: 2 mmol carboxylic acid, 10 mol% AgOAc,
15 mol% K2CO3, 4 mL NMP, 120 1C, 16 h, isolated yields. GC yields
are given in parentheses and were determined using n-tetradecane as
b
c
d
the internal standard. Using 4 mL of DMAC. 140 1C. 160 1C.
7 J. Chodowska-Palicka and M. Nilsson, Acta Chem. Scand., 1970,
24, 3353.
8 H. Gilman and G. F. Wright, J. Am. Chem. Soc., 1933, 55, 3302.
9 L. J. Gooßen, W. R. Thiel, N. Rodrıguez, C. Linder and B. Melzer,
Adv. Synth. Catal., 2007, 349, 2241; L. J. Gooßen, F. Manjolinho,
B. A. Khan and N. Rodrıguez, J. Org. Chem., 2009, 74, 2620.
10 J. S. Dickstein, C. A. Mulrooney, E. M. O’Brien, B. J. Morgan and
M. C. Kozlowski, Org. Lett., 2007, 9, 2441.
Scheme 2 Ag/Pd-catalysed decarboxylative biaryl synthesis.
11 For acid mediated decarboxylations of such compounds, see:
R. W. Hay and M. J. Taylor, Chem. Commun., 1996, 525b.
12 For an independently developed silver-based protodecarboxylation
see: J. Cornella, C. Sanchez, D. Banawa and I. Larrosa,
Chem. Commun., 2009, DOI: 10.1039/b916646g.
13 See in comparison: N. R. Gunawardena and T. B. Brill, J. Phys.
Chem. A, 2001, 105, 1876; J. Li and T. B. Brill, J. Phys. Chem. A,
2003, 107, 2667; M. Staikova and D. J. Donaldson, J. Phys. Chem. A,
2005, 109, 597.
2-methoxybenzoic, 2-halobenzoic, and heteroarenecarboxylic
acids. At a more elevated temperature of 140 1C, an even wider
range of carboxylates can be converted, e.g. a-oxocarboxylic
acids and cinnamic acid, and above 160 1C, turnover is
achieved even for some meta- and para-substituted
benzoic acids.
The new decarboxylation catalyst is likely to open up
entirely new opportunities for low-temperature decarboxylative
cross-couplings. In a first test reaction at 120 1C, we observed
a 71% yield in the reaction of 2-nitrobenzoic acid with 4-tolyl
triflate using a non-optimized catalyst system consisting of
5 mol% of silver carbonate, 3 mol% of PdCl2 and 9 mol% of
PPh3 (Scheme 2). Having thus achieved a reduction in the
temperature by 50 1C also for decarboxylative cross-couplings,
we are optimistic that after careful adjustments in the Pd
co-catalyst, an effective, generally applicable, low-temperature
decarboxylative cross-coupling is in close reach.
14 GAUSSIAN 03, Revision E.01, Gaussian, Inc., Wallingford CT,
2004; for full citation see the ESIw.
15 C. Lee, W. Yang and R. G. Parr, Phys. Rev. B: Condens. Matter
Mater. Phys., 1988, 37, 785; A. D. Becke, J. Chem. Phys., 1993, 98,
5648; P. J. Stephens, J. F. Devlin, C. F. Chabalowski and
M. J. Frisch, J. Phys. Chem., 1994, 98, 11623.
16 R. Krishnan, J. S. Binkley, R. Seeger and J. A. Pople, J. Chem.
Phys., 1980, 72, 650.
17 P. C. Hariharan and J. A. Pople, Theor. Chim. Acta, 1973, 28, 213.
18 M. Dolg, U. Wedig, H. Stoll and H. Preuss, J. Chem. Phys., 1987,
86, 866; D. Andrae, U. Haußermann, M. Dolg, H. Stoll and
H. Preuss, Theor. Chim. Acta, 1990, 77, 123.
19 M. W. Wong, Chem. Phys. Lett., 1996, 256, 391.
20 For related silver-mediated protodecarboxylation see: J. M. Anderson
and J. K. Kochi, J. Am. Chem. Soc., 1970, 92, 1651; J. M. Anderson
and J. K. Kochi, J. Org. Chem., 1970, 35, 986.
Moreover, now that aryl nucleophiles are accessible from
carboxylic acids at such low temperatures, their generation
ꢁc
This journal is The Royal Society of Chemistry 2009
Chem. Commun., 2009, 7173–7175 | 7175