Supported gold catalyzes the homocoupling of phenylboronic acid with high conversion and selectivity

Article abstract of DOI:10.1002/anie.200462560

(Chemical Equation Presented) Gold on nanocrystalline cerium oxide is a highly active and selective catalyst for the homocoupling of phenylboronic acid derivatives in the presence of K₂CO₃ as a base (see picture). Au³⁺ species are thought to be the active sites for the reaction, and water is required to activate the boronic acid through the participation of surface hydroxy groups.

Full text of DOI:10.1002/anie.200462560

Communications
Supported Catalysts
To investigate our hypothesis, we prepared a supported
gold catalyst containing cationic gold species. The support
consists of nanocrystalline ceria particles, which are able to
stabilize ionic states of gold because of their electron-
withdrawing properties.[8] Indeed, the IR spectrum of CO
adsorbed on an Au/CeO2 catalyst[9] shows a band at 2148 cmꢀ1
that confirms the presence of Au3+ ions. This result was
supported by XPS data, which display a peak at 86.2 eV
corresponding to the binding energy (4f7/2) of Au3+ (Support-
ing Information).
To test the reactivity of gold supported on nanocrystalline
CeO2 for Suzuki cross-coupling, reactions were carried out
between p-iodobenzophenone and phenylboronic acid
(Experimental Section, Reaction a). We observed that all
the boronic acid was transformed into biphenyl, thus indicat-
ing that the homocoupling reaction occurs with practically
100% conversion and selectivity (Supporting Information).
The product of Suzuki cross-coupling condensation was
detected in very small amounts (< 0.5%).
Supported Gold Catalyzes the Homocoupling of
Phenylboronic Acid with High Conversion and
Selectivity**
Silvio Carrettin, Javier Guzman, and Avelino Corma*
Coupling reactions are largely dominated by the use of
palladium catalysts and are a powerful and versatile tool in
synthetic organic chemistry for the formation of carbon–
carbon bonds.[1,2] Among the different coupling reactions, the
Suzuki reaction plays an important role and it is already used
in some industrial processes.[3] The possible products include
biaryls which are interesting because of their presence in
many natural products and biologically active compounds.[4]
This class of compounds can be prepared by two different
types of reactions: homocoupling or cross-coupling.
The general mechanism for the Suzuki cross-coupling
reaction involves the transmetalation of a boronic acid by
Pd2+ ions. Various examples have recently been reported for
the homocoupling of boronic acids by different metals.[5] One
of these[5a] shows that homocoupling of boronic acids occurs
by a double transmetalation at Pd2+ centers formed by
oxidative addition of an a-halocarbonyl compound to an
initial Pd0 complex. The authors found that there was no
reaction in the absence of the a-halocarbonyl compound, thus
confirming that the transmetalation of the boronic acid occurs
at Pd2+ species. Taking this into account, we thought that as
Au3+ is isoelectronic with Pd2+ and able to undergo the redox
cycle Au3+QAu+, it may be possible for a solid catalyst
containing Au3+ species to be active in the cross-coupling as
well as the homocoupling reaction (Scheme 1). To the best of
Interestingly, the reaction also proceeds with the same
conversion and selectivity in the absence of the a-halocar-
bonyl compound and K2CO3. When the reaction is carried out
without K2CO3, the only difference observed is the degrada-
tion of the catalyst. The base is known to activate phenyl-
boronic acid;[1] however, our results indicate that with our
catalytic system base is not needed for the activation of
phenylboronic acid, and its only role is to neutralize the boric
acid. Furthermore, the turnover number (TON) of 20
(calculated as the moles of boronic acid converted divided
by two and by the moles of gold in the catalyst per hour)
indicates that the process is catalytic. It should be noted that
because not all the gold atoms are available for the reaction,
the real TON is likely to be much higher than 20.
Blank experiments without catalyst or using gold-free
nanocrystalline ceria gave no conversion, thus showing that
the catalytically active sites are associated with gold. Possible
Au leaching during the reaction was checked by analyzing
gold in the supernatant liquid by AAS; 0.3 ꢁ 0.1 ppm of gold
was detected (Supporting Information). The activity of the
leached gold was found to be negligible by carrying out an
experiment in which 3 ppm of gold (ten times the amount of
leached gold) in the form of HAuCl4 was introduced as
catalyst. After 6 h at 333 K the conversion was 1.1%. Under
the same conditions using our gold catalyst the conversion
was 100%.
ꢀ
our knowledge the only examples of C C bond formation
catalyzed by gold have been performed using homogeneous
catalytic systems.[6,7]
From a stoichiometric point of view, the formation of
boric acid from boronic acid requires the participation of
either a surface hydroxy group (associated to gold[10] and/or to
the cerium oxide[11]) or molecular water that will dissociate
and subsequently hydrolyze the boronic acid to boric acid. For
either case, it is clear that hydroxy groups participate in the
catalytic reaction, and regeneration of such species requires
dissociation of molecular water on the surface of the support
to restore the surface hydroxy groups and generate molecular
hydrogen.[11] The overall homocoupling reaction is therefore
as described in [Eq. (1)].
Scheme 1. Possible Suzuki and homocoupling reactions catalyzed by
gold; reaction conditions: 2.25 wt% Au/CeO2, toluene, K2CO3,
[*] Dr. S. Carrettin, Dr. J. Guzman, Prof. A. Corma
Instituto de Tecnologꢀa Quꢀmica
UPV-CSIC
Universidad Politꢁcnica de Valencia
Avda. de los Naranjos s/n, 46022 Valencia (Spain)
Fax: (+34)96-387-7809
E-mail: acorma@itq.upv.es
[**] The authors thank MAT 2003-07945-C02-01 and the Auricat EU-
Network (HPRN-CT-2002-00174) for financial support.
Supporting information for this article is available on the WWW
2 ðHOÞ2BPh þ 2 H2O þ Au3þ ! 2 BðOHÞ3 þ PhPh þ Auþ þ H2
ð1Þ
under http://www.angewandte.org or from the author.
T=333 K.
2242
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/anie.200462560
Angew. Chem. Int. Ed. 2005, 44, 2242 –2245

Angewandte
Chemie
To prove experimentally that this reaction is taking place
we monitored the homocoupling reaction by Raman spectros-
copy to detect H2 formation. The results presented in Figure 1
clearly show the presence of molecular hydrogen, as indicated
occur by three different routes: by air, by H+ generated in the
media, or by Ce4+ present in the support. We have seen that
the reaction proceeds in the absence of oxygen with the same
activity, therefore the first possibility can be rejected. To
investigate the second option—oxidation of Au+ by H+ to give
Au3+ and H2—we monitored the homocoupling reaction by
Raman spectroscopy; formation of H2 was observed.[12]
However, it would appear that the redox potential for the
whole reaction (2H+ to H2 and Au+ to Au3+) is not high
enough to cause a spontaneous electron transfer. Neverthe-
less, it should be taken into account that in the case of
nanoparticles, as the number of atoms n per particle
decreases, a nonlinear relationship exists between n, the
redox potential (E0), and the lattice stabilization energy of the
atoms involved. As a consequence, E0 decreases drastically
with decreasing n.[14] Therefore, considering the formation of
H2 and the effect of small particle sizes on the oxidation
potential of gold, it is possible that Au+ present as a gold
nanoparticle could be reoxidized to Au3+ by H+.
by the band at 4104 cmꢀ1 [12]
.
Looking into the third possibility—Ce4+ acting as an
oxidant for Au+—it has been reported[11] that water disso-
ciates on the surface of cerium oxide to form OH groups and
oxidize Ce3+ to Ce4+. This reaction produces the H2 that was
detected by Raman spectroscopy.[12] In this case the redox
potential for the reduction of Ce4+ to Ce3+ is high enough to
oxidize Au+ to Au3+, therefore this is also a promising
candidate for the process that closes the catalytic cycle.
Consistent with this hypothesis, XPS data characterizing the
gold catalyst support the presence of Ce4+ and Ce3+ species
(Supporting information).
We can therefore postulate two possible mechanisms for
the homocoupling of phenylboronic acid on Au/CeO2 cata-
lysts (Scheme 2). For both catalytic mechanisms we infer that
Au3+ is present at the active sites. To confirm this we
investigated a series of Au/CeO2 catalysts with different
Au3+/Au0 and Au+/Au0 ratios, as determined by IR spectros-
copy using CO as a probe molecule. The results presented in
Figure 2 clearly show a direct correlation between the
Figure 1. Raman spectra of a) toluene; b) H2 in toluene; and c) homo-
coupling reaction in toluene.
The presence of H2O is required for this reaction. To
confirm this we performed a set of experiments using
anhydrous materials (Experimental Section, Reaction b)
which lead to only 17% conversion after 7 h. At this point
30 mL of H2O was added, and the conversion increased to
47.5% after another 7 h. Furthermore, when 50 mL of H2O
was added to the dried catalyst and reactants the conversion
increased to 61%. These results clearly support the impor-
tance of H2O in the catalytic reaction, and suggest the
participation of hydroxy groups in hydrolyzing the boronic
acid.
Raman spectroscopy was used to characterize the addi-
tion of water to the gold catalyst and the support alone. The
spectrum of the Au/CeO2 catalyst treated with water shows
the formation of hydroxy groups on the support, as evidenced
by new bands at 3427 and 3594 cmꢀ1 (Supporting Informa-
tion). These hydroxy groups disappeared after addition of
phenylboronic acid, thereby indicating the participation of
OH species in the activation of the phenylboronic acid. In
contrast, when the same experiment was carried out with the
CeO2 support, the Raman bands representing the hydroxy
groups showed no variation after addition of boronic acid.
These results indicate that gold is required to enhance the
interaction between OH species and the boronic acid. It has
been shown previously that the presence of gold weakens the
surface oxygen species of ceria.[13]
When boronic acid was mixed with water or with water
and cerium oxide at the reaction temperature in the absence
of gold, neither product formation nor substrate decomposi-
tion were observed. We can therefore conclude that the OH
groups on the surface of the CeO2 support interact with the
boronic acid to form boric acid, while two phenyl species
interact with Au3+ to form a carbon–metal bond. Thus,
homocoupling occurs and Au3+ is reduced to Au+.
Scheme 2. Catalytic cycle for the homocoupling reaction catalyzed by Au3+
To regenerate the active species and to close the catalytic
cycle, Au+ has to be reoxidized to Au3+. This oxidation can species: a) Au+ reoxidation to Au3+ by H+; b) Au+ reoxidation to Au3+ by Ce4+
.
Angew. Chem. Int. Ed. 2005, 44, 2242 –2245
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2243

Communications
Table 1: Homocoupling of phenylboronic acids.[a]
Reaction
Yield [%]
100
100
97[b]
[a] Reaction conditions: 5 mL toluene, 0.3 mmol of phenylboronic acid,
5 mol% of Au, T=333 K, 15 h; yield determined by GC-MS with
n-decane as internal standard. [b] 3% Acetophenone was formed as
byproduct.
reaction of phenylboronic acid with very high activity and
selectivity in the absence of an a-halocarbonyl compound and
presence of a base. Au3+ species are the active sites for the
reaction and water is required to hydrolize the boronic acid
through surface hydroxy groups.
Experimental Section
Au/CeO2 catalysts were prepared by precipitation of HAuCl4 with
NaOH following a procedure described elsewhere.[9] The total Au
content of the final catalyst was determined by chemical analysis to be
2.25 wt%.
The solvent (5 mL) and the catalyst (the amount depends on the
mol percent of gold required) were added to a 10-mL round-
bottomed flask containing phenylboronic acid (0.3 mmol) and K2CO3
(0.4 mmol). The reactions were carried out at 333 K (unless otherwise
specified) for 15 h with vigorous stirring. The yields and conversions
were determined by GC-MS analysis using decane as internal
standard.
Figure 2. a) Correlation between Au3+ and Au0 species and TON for
the homocoupling reaction catalyzed by gold; b) correlation between
Au+ and Au0 species and TON for the homocoupling reaction cata-
lyzed by gold. The frequency and intensity of the IR band of CO adsor-
bed on gold catalysts were used to identify Au3+ (band at 2148 cmꢀ1
representing Au3+-bound CO), Au+ (2130 cmꢀ1, Au+-bound CO) and
Au0 (2104 cmꢀ1, Au0-bound CO).
Reaction
a (Suzuki cross-coupling): The catalyst (100 mg;
5 mol% of Au per mol of PhB(OH)2) was added to a round-
bottomed flask together with DMF (5 mL), p-iodobenzophenone
(0.2 mmol), and phenylboronic acid (0.3 mmol). The reaction was
carried out at 433 K.
Reaction b: The catalyst and reactants were dried at 333 K and
10ꢀ3 Torr and anhydrous solvent was used; the reaction was carried
out at 333 K under nitrgoen.
concentration of Au3+ species and catalytic activity; no
correlation was found between catalytic activity and the
concentration of Au+ or Au0.
Gold on nanocrystalline CeO2 is a much more active
catalyst than gold on precipitated CeO2 for the homocoupling
reaction. Consistent with these results, our group has recently
demonstrated that the characteristics of the cerium oxide
surface are extremely important for CO oxidation on an Au/
CeO2 catalyst.[9] We found that nanocrystalline CeO2 (nano-
particles of about 4 nm in size with fluorite structure)
increases the activity of gold for CO oxidation by two
orders of magnitude with respect to a precipitated CeO2
support. The results obtained for gold on different metal-
oxide catalysts tested for CO oxidation and for homocoupling
show a direct correlation between the activities (Supporting
Information).
We also studied the homocoupling of other aryl boronic
acids to determine the possible influence of substituents on
the aromatic ring. The Au/CeO2 catalyst gave practically
100% conversion and selectivity in all cases (Table 1).
In conclusion, we have found that gold supported on
nanocrystalline cerium oxide catalyzes the homocoupling
Received: November 9, 2004
Revised: January 10, 2005
Published online: March 10, 2005
Keywords: C–C coupling · gold · heterogeneous catalysis ·
.
nanostructures · supported catalysts
[1] a) A. de Meijere, F. Diederich, Metal-Catalyzed Cross-Coupling
Reactions, Vol. 1–2, Wiley-VCH, Weinheim, 2004; b) G. C. Fu,
A. F. Littke, Angew. Chem. 2002, 114, 4350 – 4386; Angew.
Chem. Int. Ed. 2002, 41, 4176 – 4211.
[2] K. C. Nicolaou, E. I. Sorensen, Classics in Total Synthesis, VCH,
New York, 1996.
[3] G. B. Smith, G. C. Dezeny, D. L. Hughes, A. O. King, T. R.
Verhoeven, J. Org. Chem. 1994, 59, 8151 – 8156.
[4] J. Hassan, V. Penalva, L. Lavenot, C. Gozzi, M. Lamaire,
Tetrahedron 1998, 54, 13793 – 13804.
2244
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
Angew. Chem. Int. Ed. 2005, 44, 2242 –2245

Angewandte
Chemie
[5] a) A. Lei, X. Zhang, Tetrahedron Lett. 2002, 43, 2525 – 2528, and
references cited therein; b) J. R. Falck, S. Mohapatra, M.
Bondlela, S. K. Venkataraman, Tetrahedron Lett. 2002, 43,
8149 – 8151; c) A. S. Demir, ꢀ. Reis, M. Emrullahoglu, J. Org.
Chem. 2003, 68, 578 – 580; d) J. P. Parrish, Y. C. Jung, R. J. Floyd,
K. W. Jung, Tetrahedron Lett. 2002, 43, 7899 – 7902; e) A. Demir,
ꢀ. Reis, M. Emrullahoglu, J. Org. Chem. 2003, 68, 10130 –
10134; f) C. J. Li, Angew. Chem. 2003, 115, 5004 – 5006;
Angew. Chem. Int. Ed. 2003, 42, 4856 – 4858.
[6] a) G. Dyker, Angew. Chem. 2000, 112, 4407 – 4409; Angew.
Chem. Int. Ed. 2000, 39, 4237 – 4239, and references cited
therein; b) A. S. K. Hashmi, L. Ding, J. W. Bats, P. Fisher, W.
Frey, Chem. Eur. J. 2003, 9, 4339 – 4345.
[7] One of the referees pointed out an ASAP publication where
colloidal gold is used as a catalyst for the homocoupling of
phenylboronic acids. However, the activity is low and, more
importantly, the selectivity is always below 80%: H. Tsunoyama,
H. Sakurai, N. Ichikuni, Y. Negishi, T. Tsukuda, Langmuir 2004,
20, 11293 – 11296.
[8] a) Q. Fu, H. Saltsburg, M. Flytzani-Stephanopoulos, Science
2003, 301, 935 – 938; b) F. Zhang, S. W. Chan, J. E. Spanier, E.
Apak, Q. Jin, R. D. Robinson, I. P. Herman, Appl. Phys. Lett.
2002, 80, 127 – 129; c) A. N. Pestryakov, V. V. Lunin, A. N.
Kharlanov, N. E. Bogdanchikova, I. V. Tuzovskaya, Eur. Phys.
J. D 2003, 24, 307 – 309; d) A. N. Pestryakov, V. V. Lunin, J. Mol.
Catal. A 2000, 158, 325 – 329.
[9] S. Carrettin, P. Concepciꢁn, A. Corma, J. M. Lꢁpez-Nieto, V. F.
Puntes, Angew. Chem. 2004, 116, 2592 – 2594; Angew. Chem. Int.
Ed. 2004, 43, 2538 – 2540.
[10] E. D. Park, J. S. Lee, J. Catal. 1999, 186, 1 – 11.
[11] a) T. Tabakova, F. Boccuzzi, M. Manzoli, D. Andreeva, Appl.
Catal. A 2003, 252, 385 – 397; b) Lj. Kundakovic, D. R. Mullins,
S. H. Overbury, Surf. Sci. 2000, 457, 51 – 62.
[12] a) D. Gracin, I. Bogdanovic, V. Bojanovic, M. Jaksic, Z.
Pastuovic, J. M. Dutta, V. Vlahovic, R. J. Nemanich, Vacuum
2001, 61, 303 – 308; b) A. W. R. Leitch, J. Weber, V. Alex, Mater.
Sci. Eng. B 1999, 58, 6 – 12.
[13] D. Andreeva, V. Idakiev, T. Tabakova, L. Ilieva, P. Falaras, A.
Bourlinos, A. Travlos, Catal. Today 2002, 72, 51.
[14] L. D. Burke, A. J. Ahern, A. P. OꢂMullane, Gold Bull. 2002, 35,
3 – 10.
Angew. Chem. Int. Ed. 2005, 44, 2242 –2245
www.angewandte.org
ꢀ 2005 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2245