Organometallic gold(III) compounds as catalysts for the addition of water and methanol to terminal alkynes

Article abstract of DOI:10.1021/ja036049x

Different inorganic and organometallic gold(III) and gold(I) complexes have been tested in the addition of water and methanol to terminal alkynes. Anionic and neutral organometallic gold(III) compounds can efficiently mediate these reactions in neutral media in refluxing methanol. The compounds are added in catalytic amounts (1.6-4.5 mol% with respect to the alkyne), Thus, compounds of the general formula Q[AuRCl₃], Q[AuR₂Cl₂], [AuRCl₂]₂, and [AuR₂Cl]₂, (Q = BzPPh₃⁺, PPN: N(PPh₃)₂⁺ or N(Bu)₄⁺; R = C₆F₅ or 2,4,6-(CH₃)₃C₆H₂) seem to behave as Lewis acids in nucleophilic additions to triple bonds. Some intermediates could be detected in the stoichiometric reaction between [Au(C₆F ₅)₂Cl]₂ and phenylacetylene that was followed by variable temperature ¹H, ¹⁹F{¹H}, COSY ¹⁹F{¹H}-¹⁹F{¹H}, and ²H{¹H} NMR experiments. Compound [Au(C₆F ₅)₂Cl]₂ is also able to catalyze the hydration of phenylacetylene at room temperature. A plausible mechanism for the hydration reaction has been proposed.

Full text of DOI:10.1021/ja036049x

Published on Web 09/09/2003
Organometallic Gold(III) Compounds as Catalysts for the
Addition of Water and Methanol to Terminal Alkynes
Raquel Casado, Mar´ıa Contel, Mariano Laguna,* Pilar Romero, and Sergio Sanz
Contribution from the Departamento de Qu´ımica Inorga´nica - Instituto de Ciencia de Materiales
de Arago´n, UniVersidad de Zaragoza-C.S.I.C., Zaragoza, E-50009, Spain
Received May 9, 2003; E-mail: mlaguna@unizar.es
Abstract: Different inorganic and organometallic gold(III) and gold(I) complexes have been tested in the
addition of water and methanol to terminal alkynes. Anionic and neutral organometallic gold(III) compounds
can efficiently mediate these reactions in neutral media in refluxing methanol. The compounds are added
in catalytic amounts (1.6-4.5 mol % with respect to the alkyne). Thus, compounds of the general formula
Q[AuRCl3], Q[AuR2Cl2], [AuRCl2]2, and [AuR2Cl]2 (Q ) BzPPh3+, PPN: N(PPh3)2+ or N(Bu)4+; R ) C6F5 or
2,4,6-(CH3)3C6H2) seem to behave as Lewis acids in nucleophilic additions to triple bonds. Some
intermediates could be detected in the stoichiometric reaction between [Au(C6F5)2Cl]2 and phenylacetylene
that was followed by variable temperature 1H, 19F{1H}, COSY 19F{1H}-19F{1H}, and 2H{1H} NMR
experiments. Compound [Au(C6F5)2Cl]2 is also able to catalyze the hydration of phenylacetylene at room
temperature. A plausible mechanism for the hydration reaction has been proposed.
Introduction
ketones because all known additions of water to alkynes follow
Markovnikov’s rule. More recently, Ru(II) complexes, in the
The hydration of unactivated alkynes, an abundant hydro-
carbon resource, is a safe and environmentally benign route to
form C-O bonds. Although the hydration of alkynes can be
mediated by acids, it generally requires a large excess of acidic
reagent, and only electron-rich acetylene compounds react with
high conversions.1 Transition metals have been used to catalyze
this process as well as the analogous reaction in which alcohols
are added across the triple bond.2 The most extensively
employed catalytic systems consist of toxic mercury salts in
acidic media,3 but other transition-metal-complex catalysts
containing Ru(III),4 Rh,5 Pd(II),6 Pt,7 Cu(II),8 or Ag (I)8 metal
centers as well as some Ir-M (M ) Pd, Pt) clusters9 have been
described. These catalytic systems lead to the obtention of
presence of appropriate auxiliary phosphane ligands, have
efficiently catalyzed the anti-Markovnikov addition of water to
terminal alkynes, yielding mainly aldehydes.10 The use of gold
compounds in homogeneous catalyzed organic reactions has
been undervalued for many years due to the preconceived
opinion that gold is an expensive and extremely chemically inert
metal. However, recent reports have changed this assessment,
and gold salts in small amounts are known to display high
catalytic activity.11 Gold(III) acids or salts such as HAuCl4,
HAuCl4/AgNO3, Na[AuCl4], anhydrous AuCl3, and Au2O3 or
gold(I) complexes such as [AuCl(PPh3)], [Au(PPh3)(NO3)],
[Au(Me)(PPh3)], [AuCl(PPh3)2], [AuCl(dcpe)2] (dcpe ) ethane-
1,2-diylbisdicyclohexylphosphane), [{Au(PPh3)}2S], and gold(I)
compounds with chiral ferrocenylphosphine ligands have be-
haved as good catalysts for different organic reactions. Thus,
asymmetric aldol reactions,12 diboration of vinylarenes,13 de-
hydrogenative dimerizations of trialkylstannanes,14 carbonylation
of olefins15 and amines,16,17 C-C and C-O bond couplings,18-23
(1) (a) Schrohe, A. Chem. Ber. 1875, 8, 367. (b) Baeyer, A. Ber. 1882, 15,
2705. (c) Perkin, W. H., Jr. J. Chem. Soc. 1884, 45, 170. (d) Izumi, Y.
Catal. Today 1997, 33, 371. (e) Kozhevnikov, I. V. Chem. ReV. 1998, 98,
171. (f) Tsuchimoto, T.; Joya, T.; Shirakawa, E.; Kawakami, Y. Synlett
2000, 12, 1777.
(2) Utimoto, K. Pure Appl. Chem. 1983, 55, 1845.
(3) (a) Budde, W. L.; Dessy, R. E. J. Am. Chem. Soc. 1963, 85, 3964. (b)
Storck, G.; Borch, R. J. Am. Chem. Soc. 1964, 86, 935. (c) Uemura, S.;
Miyoshi, H.; Okano, M. J. Chem. Soc., Perkin Trans. I 1980, 1098. (d)
Wasacz, J. P.; Badding, V. G. J. Chem. Educ. 1982, 59, 694. (e) Larock,
R. C.; Leong, W. W. In ComprehensiVe Organic Synthesis; Trost, B. M.,
Flemming, I., Semmelhock, M. F., Eds.; Pergamon Press: Oxford, 1994;
Vol. 4, p 269. (f) March, J. AdVanced Organic Chemistry, 4th ed.; Wiley
Interscience: New York, 1992; pp 762-763.
(9) (a) Masui, D.; Ishii, Y.; Hidai, M. Chem. Lett. 1998, 717. (b) Masui, D.;
Kochi, T.; Tang, Z.; Ishii, Y.; Mizobe, Y.; Hidai, M. J. Organomet. Chem.
2001, 620, 69.
(10) (a) Tokunaga, M.; Wakatsuki, Y. Angew. Chem., Int. Ed. 1998, 37, 2867.
(b) Suzuki, T.; Tokunaga, M.; Wakatsui, Y. Org. Lett. 2001, 3, 735. (c)
AÄ lvarez, P.; Bassetti, M.; Gimeno, J.; Mancini, G. Tetrahedron Lett. 2001,
42, 8467.
(4) (a) Harpern, J.; James, B. R.; Kemp, A. L. W. J. Am. Chem. Soc. 1961,
83, 4097. (b) Harpern, J.; James, B. R.; Kemp, A. L. W. J. Am. Chem.
Soc. 1966, 88, 5142. (c) Khan, M. M. T.; Halligudi, S. B.; Shukla, S. J.
Mol. Catal. 1990, 58, 299.
(5) (a) James, B. R.; Rempel, G. L. J. Am. Chem. Soc. 1969, 91, 863. (b)
Blum, J.; Huminer, H.; Alper, H. J. Mol. Catal. 1992, 75, 153.
(6) (a) Kataoka, Y.; Matsumoto, O.; Ohashi, M.; Yamagata, T.; Tani, K. Chem.
Lett. 1994, 1283. (b) Kadota, I.; Lutete, L. M.; Shibuya, A.; Yamamoto,
Y. Tetrahedron Lett. 2001, 42, 6207.
(7) (a) Hiscox, W.; Jennings, P. W. Organometallics 1990, 9, 1997. (b) Baidosi,
W.; Lahav, M.; Blum, J. J. Org. Chem. 1997, 62, 669. (c) Francisco, L.
W.; Moreno, D. A.; Atwood, J. D. Organometallics 2001, 20, 4237.
(8) Meier, I. K.; Marsella, J. A. J. Mol. Catal. 1993, 78, 31.
(11) Dyker, G. Angew. Chem., Int. Ed. 2000, 39, 4237.
(12) (a) Ito, Y.; Sawamura, M.; Hayashi, T. J. Am. Chem. Soc. 1986, 108, 6405.
(b) Hayashi, T.; Sawamura, M.; Ito, Y. Tetrahedron 1992, 48, 1999. (c)
Togni, A.; Pastor, S. D. J. Org. Chem. 1990, 55, 1649. (d) Togni, A.; Pastor,
S. D. Chirality 1991, 3, 331. (e) Hughes, P. F.; Smith, S. H.; Olson, J. T.
J. Org. Chem. 1994, 59, 5799.
(13) Baker, R. T.; Nguyen, P.; Marder, T. B.; Westcott, S. A. Angew. Chem.,
Int. Ed. Engl. 1995, 34, 1336.
(14) Ito, H.; Yajima, T.; Tateiwa, J.; Hosomi, A. Tetrahedron Lett. 1999, 40,
7807.
(15) Xu, Q.; Imamura, Y.; Fujiwara, M.; Souma, Y. J. Org. Chem. 1997, 62,
1594
9
10.1021/ja036049x CCC: $25.00 © 2003 American Chemical Society
J. AM. CHEM. SOC. 2003, 125, 11925-11935
11925

A R T I C L E S
Casado et al.
oxidation of sulfides,24 or, more recently, the selective oxidation
of thioethers25 can be catalyzed or mediated by gold(III) or
gold(I) compounds. Gold has been an effective catalyst in the
addition of nucleophiles to triple bonds.26-30 In 1991, Utimoto
et al. described that Na[AuCl4] (2 mol %) could catalyze the
addition of water and methanol to terminal and internal alkynes
in neutral media.26 However, the process was not regioselective
for nonsymmetric internal alkynes. In 1998, Teles and co-
workers reported the addition of methanol to alkynes catalyzed
by Au(I) species in conjunction with acidic cocatalysts.29 More
recently, Tanaka et al. have described that Au(I)-acid systems
in aqueous methanol serve as useful catalysts which promote
the hydration of alkynes and have high turnover frequencies.30
Nevertheless, the authors did not isolate or identify possible
metallic intermediates in these reactions.
Scheme 1
(2) in 91% yield. More recently, Tanaka and co-workers have
achieved the hydration of 1 to give 2 in 98% yield by using the
gold(I) catalytic system [AuMe(PPh3)] (0.2 mol %):H2SO4 (50
mol %) in aqueous methanol at 70 °C (1 h).30 We wanted to
test the catalytic activity of inorganic and organometallic
gold(III) compounds in neutral media due to the fact that it has
been reported that acids, such as triflic acid, are able to catalyze
the hydration of terminal alkynes.1f
We decided to test the catalytic activity of gold(III) organo-
metallic complexes in the addition of water and alcohols to
alkynes in neutral media as the incorporation of organic
fragments to the gold(III) centers may help to stabilize and to
identify possible intermediates by spectroscopic methods. We
initiated our studies with gold(III) anionic compounds bearing
one or two pentafluorophenyl (C6F5) or mesityl (2,4,6-
(CH3)3C6H2) radicals that had been previously described by
us.31,32 Some of the complexes displayed catalytic activity, and
we tested neutral gold(III) inorganic and organometallic com-
pounds as well as gold(I) and gold(II) compounds in the absence
of acidic cocatalysts. Some gold(III) and gold(I) organometallic
complexes were tested in conjunction with acids. Here, we report
the results of these studies as well the identification of some
gold(III) intermediates in the stoichiometric additions of phenyl-
acetylene to neutral gold(III) organometallic complexes.
In our studies, the amount of gold compound used was
generally 2 mol % (with respect to the alkyne), and the reaction
was carried out in refluxing methanol over 1 h 30 min. Utimoto
and co-workers reported that the pH of Na[AuCl4] in aqueous
methanol was slighty acidic (around 5), implying somehow that
it was an acid-based catalysis.26a However, there was preliminary
evidence that one of the plausible reaction pathways may be
the cleavage of a Au-Cl bond and subsequent coordination of
the alkyne as the first step.26b
All of the catalytic results are shown in Tables 1 and 2. We
studied the hydration of 1 in neutral media (Table 1) and in
conjuction with acids (Table 2). The gold(III) compounds chosen
were inorganic neutral compounds and anionic and neutral
organometallic complexes in which one or more chloride ligands
have been replaced by organic groups. Thus, inorganic neutral
gold(III) derivatives such as [AuCl3(tht)]33 and [AuCl3(PPh3)]
do not give conversion of 1 to 2 or the conversion is very poor
(11%).34 However, anionic organometallic gold(III) complexes
of the type Q[AuRCl3] (Q ) BzPPh3 (benzyl triphenyl phos-
phonium), R ) mes31a or C6F532a) as well as neutral [AuRCl2]2
compounds in which a chloride has been replaced by an organic
mononegative radical behave as good catalysts (Table 1). We
chose the mesityl (2,4,6-(CH3)3C6H2) and pentafluorophenyl
(C6F5) radicals for two main reasons: (a) their electronic and
steric properties are quite different (resembling the electronic
properties of the C6F5 group as compared to those of a chloride
ligand); and (b) our research group has prepared a large variety
of gold(I), gold(III), and even gold(II) complexes with these
radicals.
Results and Discussion
1. Hydration of Phenylacetylene. 1.1. Hydration with
Gold(III) Complexes. The hydration of phenylacetylene (1) in
refluxing aqueous methanol has already been described by
Utimoto et. al with the gold(III) salt Na[AuCl4] (2 mol %, 1 h).
The resulting product (Scheme 1) is the phenyl methyl ketone
(16) Shi, F.; Deng, Y.; Yang, H.; Sima, T. Chem. Commun. 2001, 345.
(17) Shi, F.; Deng, Y. Chem. Commun. 2001, 443.
(18) Hashmi, A. S. K.; Schwarz, L.; Choi, J.-H.; Frost, T. M. Angew. Chem.,
Int. Ed. 2000, 39, 2285.
(19) Hashmi, A. S. K.; Frost, T. M.; Bats, J. W. J. Am. Chem. Soc. 2000, 122,
1554.
(20) Hoffmann-Ro¨der, A.; Krause, N. Org. Lett. 2001, 3, 2537.
(21) Hashmi, A. S. K.; Frost, T. M.; Bats, J. W. Org. Lett. 2001, 3, 3769.
(22) Hashmi, A. S. K.; Frost, T. M.; Bats, J. W. Catal. Today 2002, 72, 19.
(23) Asao, N.; Takahashi, K.; Lee, S.; Kasahara, T.; Yamamoto, Y. J. Am. Chem.
Soc. 2002, 124, 12650.
(24) Gasparrini, F.; Giovannoli, M.; Misiti, D.; Natile, G.; Palmieri, G.
Tetrahedron 1983, 39, 3184.
(25) Boring, E.; Geletii, Y. V.; Hill, C. L. J. Am. Chem. Soc. 2001, 123,
1625.
(26) (a) Fukuda, Y.; Utimoto, K. J. Org. Chem. 1991, 56, 3729. (b) Norman,
R. O. C.; Parr, W. J. E.; Thomas, C. B. J. Chem. Soc., Perkin Trans. 1976,
1983.
Gold(III) organometallic compounds that behave as good
catalysts or that we use in subsequent stoichiometric reactions
have been numbered in the tables and throughout the text.
Neutral complexes were prepared in situ by reaction with silver
salts (Scheme 2 and Experimental Section) and loaded in a 4.5
mol % ratio.
(27) Fukuda, Y.; Utimoto, K. Bull. Chem. Soc. Jpn. 1991, 64, 2013.
(28) Fukuda, Y.; Utimoto, K. Synthesis 1991, 975.
(29) (a) Teles, J. H.; Schulz, M. (BASF AG), WO-A1 9721648, 1997; Chem.
Abstr. 1997, 127, 121499. (b) Teles, J. H.; Brode, S.; Chabanas, M. Angew.
Chem., Int. Ed. 1998, 37, 1415.
Whereas the compound with one mesityl group (4) is not
catalytically active (entry 2), the compound with a pentafluoro-
phenyl radical (8) gives 100% yield in the above-mentioned
reaction conditions (entry 4). The reaction is sensitive to the
amount of catalyst used (entries 1 and 2).
(30) Mizushima, E.; Sato, K.; Hayashi, T.; Tanaka, M. Angew. Chem., Int. Ed.
2002, 41, 4563.
(31) (a) Contel, M.; Jime´nez, J.; Jones, P. G.; Laguna, A.; Laguna, M. J. Chem.
Soc., Dalton Trans. 1994, 2515. (b) Contel, M.; Edwards, A. J.; Garrido,
J.; Hursthouse, M. B.; Laguna, M.; Terroba, R. J. Organomet. Chem. 2000,
697, 129.
(32) The synthesis of BzPPh3[Au(C6F5)Cl3] is carried out by oxidative addition
of PhCl2I to BzPPh3[Au(C6F5)Cl] in a way similar to that for the obtention
of BzPPh3[Au(mes)Cl3].31a
(33) (a) [AuCl3(tht)] obtained by reaction of [AuCl(tht)]33b and PhCl2I (1:1 molar
ratio). (b) Uso´n, R.; Laguna, A.; Vicente, J. J. Organomet. Chem. 1977,
131, 471.
(34) [AuCl3(PPh3)] obtained by reaction of [AuCl(PPh3)]45 and PhCl2I (1:1 molar
ratio).
9
11926 J. AM. CHEM. SOC. VOL. 125, NO. 39, 2003

Organometallic Gold(III) Complexes as Catalysts
A R T I C L E S
Table 1. Hydration of Phenylacetylene (1) with Organometallic Gold(III) Compounds
entry
catalyst
cat. loading (mol %)
time
conversiona (%)
isolated yieldb (%)
1
2
3
4
5
6
7
8
9
BzPPh3[Au(mes)Cl3] (3)
BzPPh3[Au(mes)Cl3] (3)
[“Au(mes)Cl2”] (4)
BzPPh3[Au(C6F5)Cl3] (7)
[“Au(C6F5)Cl2”] (8)
[Au(C6F5)Cl2(tht)]
cis-PPN[Au(mes)2Cl2] (5)
cis-[Au(mes)2Cl]2 (6)
trans-NBu4[Au(C6F5)2Cl2] (9a)
[“Au(C6F5)2Cl”] (10)
trans-NBu4[Au(C6F5)2Br2] (9b)
trans-NBu4[Au(C6F5)2Br2] (9b)
1
2
4.5
2
4.5
2
1 h 30 min
1 h 30 min
1 h 30 min
1 h 30 min
1 h 30 min
1 h 30 min
1 h 30 min
1 h 30 min
1 h 30 min
1 h 30 min
3 h 30 min
4 h
3c
100
2c
81
72c
100
0c
70
98
2
11c
33c
100
100
94c
96
4.5
1.6
4.5
1.6
2.5
98
98
83
90
10
11
12
1
a Conversion to phenyl methyl ketone (2) determined by GCMS. b Isolated compound (see Experimental Section) characterized by H NMR. c Starting
material was recovered.
Table 2. Hydration of Phenylacetylene (1) with Au(III) or Au(I) Compounds in the Presence of Acid (1:10 Molar Ratio) or Silver Salts (1:1
Molar Ratio)
entry
catalyst
acid or silver salt
cat. loading (mol %)
time
conversiona (%)
1
2
3
4
5
6
7
8
9
BzPPh3[Au(mes)Cl3] (3)
BzPPh3[Au(C6F5)Cl3] (7)
BzPPh3[Au(C6F5)Cl3] (7)
t-NBu4[Au(C6F5)2Cl2] (9a)
t-NBu4[Au(C6F5)2Cl2] (9a)
BzPPh3[Au(mes)Cl3] (3)
[Au(Me)PPh3] (11)
[Au(Me)PPh3] (11)
[Au(Me)PPh3] (11)
[Au(Me)PPh3] (11)
[Au(Me)PPh3] (11)
HSO3CF3
H2SO4
HSO3CF3
H2SO4
HSO3CF3
H2SO4
H2SO4
H2SO4
H2SO4
HSO3CF3
HSO3CF3
AgOSO2CF3
0.5
0.5
0.5
0.5
0.5
0.5
1
0.5
0.1
0.5
0.1
1
1 h 30 min
1 h 30 min
1 h 30 min
1 h 30 min
1 h 30 min
1 h 30 min
1 h 30 min
1 h 30 min
1 h 30 min
1 h 30 min
1 h 30 min
1 h 30 min
23b
17b
45b
100
75b
0b
100
100
0b
10
11
12
100
15b
16b
[AuClPPh3]
a Conversion to phenyl methyl ketone (2) determined by GCMS. b Starting material was recovered.
Scheme 2
The compound [Au(C6F5)Cl2(tht)],35 that can be obtained as
a solid, does not catalyze the hydration of 1 (Table 1, entry 6).
This is similar to what we observed (no conversion at all) with
the inorganic compound [AuCl3(tht)].
Anionic and neutral organometallic compounds with two
mesityl or pentafluorophenyl groups such as Q[AuR2Cl2] or
[AuR2Cl]2 were also studied. We have recently described the
compound cis-PPN[Au(mes)2Cl2] (5),31b and it is obtained via
organomercurial transmetalation. Here, we present an alternative
Scheme 3
synthesis of 5 via a less toxic way. The reaction involves a
transmetalation from a mesityl gold(I) compound to PPN[AuCl4]
in a way similar to the transmetalation of [Au(η1-CNN)(PPh3)]
(NCN ) [2,6-(CH2-NMe2)2-C6H3]) to Au(III), Ni(II), Pd(II),
Pt(II), Ti(IV), and Fe(III) metallic centers that has been recently
reported by one of us.37 We obtain pure 5 in 92% yield (in
refluxing acetone for 2 h) and [AuCl(PPh3)], that can be
reutilized, as a subproduct from the reaction (Scheme 3,
Experimental Section).
However, 5 and its neutral derivative [Au(mes)2Cl]2 6 (whose
dimeric structure has been confirmed by X-ray analysis)31b do
not catalyze the hydration of 1 or achieve poor conversions,
even in quantities of 3 mol % or after 4 h (Table 1, entries 7
and 8). Similarly, cationic or neutral derivatives of 5 (previously
described)31b such as cis-[Au(mes)2{P(tol)3}Cl], cis-[Au(mes)2-
{AsPh3}Cl], or cis-[Au(mes)2(bipy)]ClO4 do not catalyze the
reaction of hydration at all. Nevertheless, anionic com-
pounds with two pentafluorophenyl groups such as trans-
NBu4[Au(C6F5)2X2] (X ) Cl (9a) or Br (9b))38 are able to give
the ketone product 2 in high yields. In the case of 9a, the total
conversion is achieved with 1.6 mol % of catalyst. The influence
of the different catalytic activity observed for compounds trans-
NBu4[Au(C6F5)2X2] (X ) Cl (9a), Br (9b), I (9c))38 can be
related to the electronegativity for the different halides as well
as the strength of Au-X bonds. The hydration reaction of 1
does not take place when X ) I (9c) under standard reaction
conditions.
Neutral [Au(C6F5)2Cl]2 (10)39 is prepared from 9a, and that
can be isolated as a solid (Scheme 4). In situ generated 10 can
be used successfully for catalytic purposes (catalyst loading 4.5
mol %, entry 10).
(35) Uso´n, R.; Laguna, A.; Bergareche, B. J. Organomet. Chem. 1980, 184,
411.
(36) (a) PPN[AuCl4] is obtained by oxidative addition of PhCl2I to PPN-
[AuCl2]36b (1:1 molar ratio). (b) Braunstein, P.; Clark, R. J. H. J. Chem.
Soc., Dalton Trans. 1973, 1845.
(37) Contel, M.; Stol, M.; Casado, M. A.; van Klink G. P. M.; Ellis, D. D.;
Spek, A. L.; van Koten, G. Organometallics 2002, 21, 4556.
(38) Uso´n, R.; Laguna, A.; Garc´ıa, J.; Laguna, M. Inorg. Chim. Acta 1979, 37,
201.
(39) [Au(C6F5)2Cl] is obtained by reaction of trans-NBu4[Au(C6F5)2Cl2)] and 1
equiv of AgClO4 or AgTfO.
9
J. AM. CHEM. SOC. VOL. 125, NO. 39, 2003 11927

A R T I C L E S
Casado et al.
Scheme 4
We performed the hydration reaction with selected organo-
metallic gold(III) compounds 3, 7, and 9a in conjunction with
acids. The addition of H2SO4 (in a molar ratio of 1:10, gold:
H2SO4) does not affect the hydration reaction in the case of
BzPPh3[Au(mes)Cl3] (3) or not much in the case of BzPPh3-
[Au(C6F5)Cl3] (7) (Table 2, entries 1 and 2), but it helps the
amount of catalyst to be decreased to 0.5 mol % to give total
conversion in the case of trans-NBu4[Au(C6F5)2Cl2] (9a) (entry
3). The addition of HSO3CF3 seems to improve a little the
catalytic activity of 3 and 7 (entries 4 and 5). Again, the behavior
of organometallic complexes containing mesityl or pentafluoro-
phenyl ligands is different, pointing out perhaps the different
reaction pathways.
1.2. Hydration with Gold(I) Complexes. The high catalytic
activity displayed by the system [Au(Me)(L)]:acid (in a 1:10
molar ratio, L ) phosphine, phosphite, or arsine ligand) or
[AuCl(PPh3)]:silver salt, when water and MeOH were added
to alkynes, has been described by Teles29 and Tanaka.30
We studied the catalytic activity of inorganic and organo-
metallic gold(I) compounds in neutral media (and under standard
reaction conditions) for the hydration of phenylacetylene 1. The
complexes tested were anionic organometallic compounds with
one or two organic radicals (such as PPN[Au(mes)Cl]31a or
NBu4[Au(C6F5)2]43) or neutral inorganic and organometallic
complexes with the triphenylphosphine ligand (such as [AuCl-
(PPh3)]44 or [AuR(PPh3)], R ) C6F5,45 mes,31a Me47). Under
the standard reaction conditions (2 mol % of catalyst, 1.5 h in
refluxing MeOH), these compounds either do not catalyze the
hydration at all or do so very poorly (20% of conversion for
[AuCl(PPh3)]). This also gives evidence for the hypothesis that
gold(III) compounds in neutral media do not catalyze the
reactions because they decompose to gold(I) complexes. Com-
pounds that could form supposedly catalytically active [Au-
(PPh3)]+ species easily under the reaction conditions such as
[O{Au(PPh3)}3](BF4)47 or [Au(PPh3)2]ClO4 were also studied.48
However, only with [O{Au(PPh3)}3](BF4) is a conversion of
55% obtained. That perhaps may imply that the acidic cocatalyst
is not only necessary to form [Au(PPh3)]+ fragments but it also
could help to prevent the reduction from Au(I) to Au(0) and
the decomposition of the catalyst.
In situ generated cationic complexes of the type [Au-
(C6F5)2S*2]X (X ) ClO4, TfO; S* ) solvent molecule)40 with-
out chloride ligands showed a lower catalytic activity (32%
and 25%, respectively) than that found for 10. Neutral com-
pounds with two C6F5 groups and without halide ligands such
as cis- and trans-[Au(C6F5)2(CH2PPh3)2]41 or the complex
[Au(C6F5)3(tht)]42 do not catalyze the hydration of phenyl-
acetylene.
In the cases where there is partial or total conversion to 2,
metallic gold is formed after the hydration reaction has taken
place. After these preliminary tests, we can conclude the
following: (a) Anionic organometallic compounds with one
organic radical of the type BzPPh3[AuRCl3] (R ) mes (3), C6F5
(7)) are efficient catalysts in the hydration of phenylacetylene
1 in neutral media under standard reaction conditions (with a
catalytic activity similar to that observed for Na[AuCl4]26a). The
neutral derivative [Au(C6F5)Cl2]2 (8) also behaves as an efficient
catalyst, whereas the analogous complex with one mesityl group
(4) does not. (b) Only compounds with two pentafluorophenyl
radicals (such as trans-NBu4[Au(C6F5)2X2] (X ) Cl (9a), Br
(9b)) or [Au(C6F5)2Cl]2 (10)) are efficient catalysts. Complexes
with two mesityl groups do not catalyze the hydration under
the same reaction conditions. (c) The presence of at least one
chloride ligand coordinated to the gold(III) center seems
necessary for the reaction to proceed. Complexes with three
C6F5 groups such as [Au(C6F5)3(tht)] or neutral or cationic
compounds with two C6F5 radicals (like cis- or trans-
[Au(C6F5)2(CH2PPh3)2] or [Au(C6F5)2S*2]X) do not catalyze the
reactions, or they catalyze them poorly. (d) The presence of
coordinating ligands (even weakly coordinating ligands) inhibits
the catalytic process with the gold(III) compounds (see the
results with [AuCl3(tht)], [Au(C6F5)Cl2(tht)], cis-[Au(mes)2ClL]
(L ) phosphine or arsine), or cis-[Au(mes)2(bipy)]ClO4) unless
stable [AuL]+ fragments are formed.
The influence of acids as cocatalysts in the hydration of 1
was studied (Table 2). We found that, for [AuMe(PPh3)] (11)
and acid (H2SO4 or HSO3CF3) in a 1:10 molar ratio, at least
0.5 mol % of the gold compound (5 mol % of acid with respect
to the alkyne) was necessary for the hydration of phenylacet-
ylene (1) to take place (Table 2, entries 7-12).
It also seems evident that the organometallic gold(III)
complexes that behave as good hydration catalysts do not simply
decompose to gold(I) active catalytic species. We believe that
“AuCl3”, “AuRCl2”, or “AuR2Cl” (R ) C6F5) may be the active
catalytic species. As we will describe later on, stoichiometric
reactions between [Au(C6F5)Cl2]2 or [Au(C6F5)2Cl]2 and phenyl-
acetylene 1 may support this assumption. From these preliminary
results, we observe a different behavior between complexes with
the pentafluorophenyl and those containing the mesityl group.
The behavior of organometallic complexes with one or two C6F5
groups resembles closely the behavior of the Na[AuCl4] salt.
Moreover, it seems that a Au(III)-C6F5 bond can behave
similarly to a Au(III)-Cl bond.
To test the catalytic activity of [Au(PPh3)]+ species in neutral
media (that were formed according to Teles et al. by addition
of acids such as HSO3CH3, H2SO4, or HBF4 to [AuMe(PPh3)]
(11)),29 we prepared them in situ by reaction of 1 mol % of
[AuCl(PPh3)] and AgOSO2CF3 (in a 1:1 molar ratio). However,
the conversion of 1 to 2 was only 16% (entry 12). This could
also support the hypothesis that the acid helps the hydration
reaction not only due to the protonation of the [AuMeL] species.
(43) Coates, G.; Parkin, C. J. Chem. Soc. 1962, 3220.
(44) Uso´n, R.; Laguna, A.; Vicente, J. J. Organomet. Chem. 1977, 131, 471.
(45) Uso´n, R.; Laguna, A. Organomet. Synth. 1985, 3, 325.
(46) Uso´n, R.; Laguna, A.; Laguna, M.; Colera, I.; De Jesu´s, E. J. Organomet.
Chem. 1984, 263, 121.
(47) Bruce, M. I.; Low, P. J.; Skelton, B. W.; White, A. H. J. Chem. Soc., Dalton
Trans. 1993, 3145.
(48) Laguna, A.; Laguna, M.; Jime´nez, J.; Lahoz, F. J.; Olmos, E. J. Organomet.
Chem. 1992, 435, 235.
(40) [Au(C6F5)2S2]X is obtained by reaction of trans-NBu4[Au(C6F5)2Cl2)] and
2 equiv of AgClO4 or AgTfO in diethyl ether.
(41) Uso´n, R.; Laguna, A.; Laguna, M.; Uso´n, A.; Gimeno, M. C. J. Chem.
Soc., Dalton Trans. 1988, 701.
(42) Uso´n, R.; Laguna, A.; Laguna, M.; Ferna´ndez, E. J.; Jones, P. G.; Sheldrick,
G. M. J. Chem. Soc., Dalton Trans. 1983, 1971.
9
11928 J. AM. CHEM. SOC. VOL. 125, NO. 39, 2003

Organometallic Gold(III) Complexes as Catalysts
A R T I C L E S
Scheme 5
Table 4. Hydration of n-Heptyne (15) with Organometallic Gold(III)
Compounds
cat. loading
(mol %)
conversiona isolated
entry
catalyst
time
(%)
yieldb (%)
1
2
3
4
5
6
7
BzPPh3[Au(mes)Cl3] (3)
[Au(mes)Cl2]2 (4)
BzPPh3[Au(C6F5)Cl3] (7)
[Au(C6F5)Cl2]2 (8)
t-NBu4[Au(C6F5)2Cl2] (9a)
[Au(C6F5)2Cl]2 (10)
BzPPh3[Au(mes)Cl3] (3)/
10 equiv of H2SO4
BzPPh3[Au(mes)Cl3] (3)/
10 equiv of HSO3CF3
BzPPh3[Au(C6F5)Cl3] (7)/
10 equiv of H2SO4
2
4.5
2
4.5
2
4.5
1
1 h 30 min
1 h 30 min
1 h 30 min
1 h 30 min
1 h 30 min
1 h 30 min
1 h 30 min
8
0c
100
100
100
100
40
87
85
90
90
Table 3. Addition of Anhydrous Metanol to Phenylacetylene (1)
with Organometallic Gold(III) Compounds
cat. loading
(mol %)
conversiona isolated
entry
catalyst
time
(%)
yieldb (%)
1
BzPPh3[Au(mes)Cl3] (3)
3
1 h 30 min
26 (12)
64 (13)
45 (12)
55 (13)
8
9
1
1
1
1
1
1 h 30 min
1 h 30 min
1 h 30 min
1 h 30 min
1 h 30 min
43
100
73
2
3
BzPPh3[Au(C6F5)Cl3] (7)
t-NBu4[Au(C6F5)2Cl2] (9a)
3
1 h 30 min
93
69
95
90
2.5
1 h 30 min 100 (12)
98
10 BzPPh3[Au(C6F5)Cl3] (7)/
10 equiv of HSO3CF3
11 t-NBu4[Au(C6F5)2Cl2] (9a)/
10 equiv of H2SO4
12 t-NBu4[Au(C6F5)2Cl2] (9a)/
10 equiv of HSO3CF3
a Conversion to products determined by GCMS. b Isolated 12 (see
1
100
100
Experimental Section) characterized by H NMR.
Scheme 6
a Conversion to 2-heptanone (16) determined by GCMS. b Isolated
compound (see Experimental Section) characterized by 1H NMR. c Starting
material was recovered.
Table 5. Addition of Anhydrous Methanol to n-Heptyne (15) with
Organometallic Gold(III) Compounds
2. Addition of MeOH to Phenylacetylene. The addition of
MeOH to phenylacetylene (1) gave two compounds with the
catalytic system [AuMe(PPh3)]:HSO3CH3 (molar ratio 1:10).29
The enol ether (13) and the acetal (12) that are shown in Scheme
5 were obtained in a ratio of 1:2, respectively.
cat. loading
(mol %)
conversiona isolated
entry
catalyst
time
(%)
yieldb (%)
1
BzPPh3[Au(mes)Cl3] (3)
3
1 h 30 min
24 (17)
72 (18)
27 (17)
73 (18)
50 (17)
50 (18)
2
3
4
BzPPh3[Au(C6F5)Cl3] (7)
BzPPh3[Au(C6F5)Cl3] (7)
t-NBu4[Au(C6F5)2Cl2] (9a)
3
1 h 30 min
3 h
We tested the most representative organometallic gold(III)
catalysts that had been efficient catalysts for the hydration of 1
(3, 7, and 9a). The results are collected in Table 3. To achieve
a total conversion of 1 to the addition products, the catalyst
loading had to be of 3 mol % in the case of anionic complexes
with one organic radical (BzPPh3[AuRCl3], R ) mes (3), C6F5
(7)) and 2.5 mol % in the case of trans-NBu4[Au(C6F5)2Cl2]
(9a). The enol ether 13 was obtained as the major product for
3, whereas it was almost obtained in a 1:1 ratio for 7 (Table 3,
entries 1 and 2). However, the selectivity improved when we
used 9a as catalyst, and the acetal 12 was obtained as the only
product of the addition reaction (entry 3). In the case of
Na[AuCl4], only 12 was obtained.26a We know that under
workup conditions, 13 is converted into the ketone 2, and thus
5
2.5
1 h 30 min 100 (17)
95
a Conversion to products determined by GCMS. b Isolated 17 (see
1
Experimental Section) characterized by H NMR.
Scheme 7
reaction, and the amounts of catalysts can be decreased to 1
mol % in the case of compounds 7 and 9a. The addition of dry
MeOH to n-heptyne (15) to give the diacetal (17) or the enol
ether product (18), Scheme 7, was tested. The results (Table 5)
are quite similar to those obtained in the addition of MeOH to
phenylacetylene (1, Table 3). Complexes BzPPh3[AuRCl3]
afforded a mixture of acetal and enol ether (18), the latter being
the major product of the reaction (entries 1 and 2). However, if
we increase the amount of catalyst to 5 mol % and the reaction
time from 1.5 to 3 h, the amount of the diacetal (17) can be
increased (entry 3). As in the case of phenylacetylene (1), the
compound with two pentafluorophenyl groups t-NBu4[Au-
(C6F5)2Cl2] (9a) gives total conversion to the diacetal 17 in 1.5
h (2.5 mol %).
We can conclude that the behavior of organometallic gold(III)
compounds is very similar for terminal alkynes (with arylic or
aliphatic groups) in neutral media. The addition of acidic
cocatalysts improves slightly their catalytic activity.
4. Stoichiometric Reactions between Neutral Organo-
metallic Complexes and Phenylacetylene. One of the goals
1
a mixture of 12 and 2 is identified by H NMR.
3. Hydration and Addition of MeOH to n-Heptyne. The
catalytic activity of these organometallic gold(III) compounds
with a terminal aliphatic-chain alkyne was tested. The alkyne
of choice was n-heptyne (15), and its hydration product (Scheme
6) was 2-heptanone (16). In Table 4, we collect the hydration
results for anionic organometallic gold(III) compounds such as
3, 7, and 9a or neutral compounds such as 4, 8, and 10. The
results are almost identical to those obtained for the hydration
of phenylacetylene (1) with the exception of compound
BzPPh3[Au(mes)Cl3] (3) that catalyzes the hydration of n-
heptyne (15), entry 1, but does so more poorly than it catalyzes
the hydration of alkyne 1 (2 mol %), Table 1 (entry 2).
Complexes with either one or two pentafluorophenyl groups
(anionic or neutral) are able to give the hydration of 15 to 16
under the standard reaction conditions (Table 4, entries 3-6).
Again, the addition of acids such as H2SO4 or HSO3CF3 (10
mol % with respect to the alkyne) improves the hydration
9
J. AM. CHEM. SOC. VOL. 125, NO. 39, 2003 11929

A R T I C L E S
Casado et al.
of the present study was to identify possible metallic intermedi-
ates in the addition of water and methanol to alkynes. As was
stated in the Introduction, the incorporation of organic ligands
into the metallic gold(III) centers may help to stabilize and to
identify (by spectroscopic methods such as NMR) possible
intermediates. We already knew that the hydration and MeOH
addition reaction conditions were difficult to reproduce with
NMR techniques as it involved the use of refluxing methanol.
We were also aware that, most probably, the first step in the
reaction may be the cleavage of a Au(III)-Cl bond (and
elimination of QCl) and that this step required an activation
energy; the conversion of 1 with t-NBu4[Au(C6F5)2Cl2] at room
temperature is very low (30% after 11 days, and it remains the
same after 24 days), whereas 1 can be hydrated totally with the
compound [Au(C6F5)2Cl]2 at room temperature after 11 days.
That is why neutral compounds such as [Au(C6F5)Cl2]2 (8) and
[Au(C6F5)2Cl]2 (10) reacted with phenylacetylene (1) in a
stoichiometric way. Complexes 8 and 10 can be obtained from
the anionic 7 or 9a by reaction with silver salts (Schemes 2
and 4). We assumed that the reaction of these neutral compounds
(that have been able to give the addition of water and MeOH
to terminal alkynes) with 1 could enable us to identify some
gold(III)-alkyne intermediates.
Figure 1. 19F{1H} NMR spectrum of the reaction of 1 and 10 in toluene-
d8 at -60 °C.
CHCl3), and it decomposes to gold(I) compounds and metallic
gold after a few minutes at room temperature. Compound 10 is
insoluble in less polar solvents such as toluene or benzene and
soluble in Et2O, THF, and MeOH. We carried out the reaction
between 10 and 1 at low temperature. The solvent of choice
was toluene for 10, which was insoluble. However, the addition
of 1 at -80 °C gives a partial solution of 10 and a change of
color of the resulting solution (colorless to orange). When trying
to remove solvents (above -50 °C) to isolate gold compounds,
we observed a very fast darkening of the solution and
decomposition to metallic gold. We followed the reaction by
NMR techniques (1H and 19F{1H}), and thus we carried out
the stoichiometric reaction of 10 and 1 in dry toluene-d8 at -80
To the best of our knowledge, although some gold(I)-alkyne
complexes have been detected and isolated at low temperature,
the reactions of alkenes and alkynes with gold(III) halides are
often complex and lead to at least partial reduction of gold(III)
to gold(I) in most cases (even at low temperatures).49 There
are not any gold(III)-alkyne products known to date, although
some gold(III)-vinyl intermediates have been isolated in reac-
tions of alkyl-gold(I) compounds with alkynes50 or AuCl3 with
dimethylacetylene.51
1
°C. At this temperature, we mostly observed in the H NMR
spectrum the signals of the solvent and unreacted phenylacet-
ylene (1). When the temperature is increased to -60 °C, some
new signals (appart from those corresponding to unreacted
phenylacetylene at 2.42 (s, PhCCH), 6.90 (m), and 7.25 ppm
The addition of PhCCH (in a molar ratio of 1:1) to [Au(C6F5)-
Cl2]2 (8), prepared in situ in Et2O as detailed in Scheme 2 and
the Experimental Section, at room temperature results in a rapid
change of color of the solution from colorless to bright yellow,
orange, and then subsequent decomposition to metallic gold in
a few seconds. However, the reaction can be carried out at -80
°C, and in a few minutes the solution of 8 becomes yellow by
addition of 1. When we tried to isolate the possible metallic
products (by removal of solvent), abundant metallic gold is
produced within a few seconds. The obtained residue is
1
(d) (PhCCH)) can be observed in the H NMR spectrum. The
signals appear at 3.23 (s), 5.80 (s), and 7.9 (s, br) ppm and
have approximately the same intensity. A broad signal (with
less intensity) appears at 3.76 (s) ppm. The 19F{1H} NMR at
-60 °C spectrum (Figure 1) is complicated and shows different
signals that can be assigned to F (in ortho, para, and meta
positions) from the C6F5 radicals. In general, for C6F5 groups,
the signals in the ortho and meta positions can appear to be
very complicated (second-order spectra) due to the couplings
between F in ortho and meta positions. From the chemical shifts
of the signals assigned to F in the ortho position, we can propose
that they belong to the gold(III) species (as they appear between
-120 and -124 ppm). The number of signals in the F para
zone is more indicative of how many different C6F5 groups we
have (or else of how many different C6F5-containing compounds
we may have).
In this case, we observe the existence of five signals with
appreciable intensity. The signals could be assigned to two
different gold(III) compounds with the two C6F5 groups in a
cis configuration (and thus with different substituents). The fifth
signal could be assigned to a gold(III) species with the two C6F5
groups in a trans configuration. The signals appear as broad
triplets at -154.4 and -156.2 ppm (for a gold(III) compound
with two cis-C6F5 groups, species A, Figure 2), at -156.8 and
-157.1 ppm (for the other gold(III) complex with two C6F5
1
dissolved in C6D6 and analyzed by H and 19F{1H} NMR at
room temperature. The obtained spectra are very complicated,
and we can only deduce that phenylacetylene (1) has reacted
completely and that there are mixtures of gold(III) complexes
containing C6F5 groups, C6F5-containing organic products, and
phenyl methyl ketone (2).
4.1. Reaction of [Au(C6F5)2Cl]2 with Phenylacetylene.
Compound [Au(C6F5)2Cl]2 (10) can be obtained as a solid
(Scheme 4, Experimental Section), although with a small
impurity of NBu4ClO4. Compound 10 is very unstable in
solution (especially in chlorinated solvents such as CH2Cl2 or
(49) Puddephatt, R. J. In ComprehensiVe Organometallic Chemistry; Wilkinson,
G., Stone, F. G. A., Abel, E. W., Eds.; Pergamon Press: Elmsford, NY,
1982; Vol. 7, pp 809-814 and references therein.
(50) (a) Mitchell, C. M.; Stone, F. G. A. J. Chem. Soc., Dalton Trans. 1972,
102. (b) Jarvis, J. A.; Johnson, A.; Puddephatt, R. J. J. Chem. Soc., Dalton
Trans. 1973, 373. (c) Johnson, A.; Puddephatt, R. J. J. Chem. Soc., Dalton
Trans. 1977, 1384. (d) Johnson, A.; Pudephatt, R. J. J. Chem. Soc., Dalton
Trans. 1978, 980.
(51) Hu¨ttel, R.; Forkl, H. Chem. Ber. 1972, 105, 2913.
9
11930 J. AM. CHEM. SOC. VOL. 125, NO. 39, 2003

Organometallic Gold(III) Complexes as Catalysts
A R T I C L E S
Figure 4. Plausible gold(III)-alkyne intermediate (1H NMR signal of
terminal H at 3.76 ppm).
Figure 5. Plausible gold(III)-vinyl intermediate.
Figure 2. Different gold(III) species identified as major products in the
19F{1H} NMR spectrum of the reaction of 1 and 10 in toluene-d8 at -60
°C (square, triangle, and circle represent different substituents).
Figure 6. Proposed gold(III)-vinyl intermediate in the reaction of Au2Cl6
with DMA and pyridine.51
temperature, the 19F{1H} NMR is quite similar to that recorded
at -20 °C, although new signals (whose F in ortho positions
that appears at around -140 ppm can be assigned to organic
C6F5-containing products) are observed.
The 1H NMR spectra recorded between -80 and 0 °C allowed
us to propose that the signals observed at 3.66 and 5.99 ppm
(and that remain from -80 to 0 °C) may correspond to the
proton from phenylacetylene (1) that has reacted with [Au-
(C6F5)2Cl]2 (10). The signal that appears at 3.76 (at -80 °C)
should also correspond to a proton from 1 that has reacted with
10. We think that this signal may possibly correspond to a signal
of a terminal proton of a phenylacetylene moiety (1) coordinated
to a metallic center (in this case, gold(III)) as a two-electron
ligand (Figure 4).52
The two signals that appear at 3.23 and 5.80 ppm (at -80
°C) or at 3.66 and 5.99 ppm at (0 °C) with identical intensity
could correspond to those of a terminal proton and a hydroxyl
group from a coordinated phenylacetylene to a gold(III) center
such as a vinyl ligand (Figure 5), most plausibly generated by
nucleophilic attack of an OH- to the gold(III)-alkyne intermedi-
ate shown in Figure 4.
The signal of the OH proton should be that at 5.80 ppm,
whereas the signal of the terminal proton (coming from the
phenylacetylene moiety and in a cis disposition with respect to
the OH group) may be that at 3.23 ppm. Similar chemical shifts
have been found in π-vinyl alcohol complexes of Os(II) formed
by quantitative reaction of the alkyne complex [Os(NH3)5(η2-
CH3CCCH3)]2+ with water.53 Moreover, this complex seems
to be a plausible intermediate as it has been reported that the
reaction of Au2Cl6 with dimethylacetylene (DMA) and pyridine
affords a gold(III)-vinyl complex that can be spectroscopically
characterized by 1H NMR at -40 °C (CH2Cl2).51 The proposed
structure of this vinyl intermediate is shown in Figure 6. Besides,
the gold(III) compounds cis-[trans-AuMe2(PMe3){(F3C)Cd
CH(CF3)] and cis-[cis-AuMe2(PMe3){(F3C)CdCH(CF3)] could
Figure 3. 1H NMR spectrum of the reaction of 1 and 10 in toluene-d8 at
0 °C.
groups in a cis disposition, species B, Figure 2) and at -157.9
ppm (for the compound with two C6F5 groups in a trans
disposition, species C, Figure 2). The chemical shifts of the F
in ortho positions are consistent with gold(III) compounds and
not with gold(I) species.
1
The H and 19F{1H} NMR spectra at -50 °C are almost
identical. At this temperature, a 19F{1H}-19F{1H} COSY NMR
experiment was carried out. From this spectrum, we can make
the assignment of related F ortho and F meta belonging to each
F para (see Supporting Information).
When we increased the temperature from -50 to -20 °C,
1
the changes observed in the H NMR were minimal. In the
19F{1H} NMR spectrum at -20 °C, slight changes in the region
of F ortho can be observed (some signals become bigger),
whereas in the region that corresponds to F in para positions
only one signal modifies its intensity, becoming smaller. This
signal is the one that we assigned to the trans gold(III) species
C.
1
However, at 0 °C the H NMR spectrum changes, and we
observed how the phenylacetylene reacts totally (the signal of
the terminal proton at 2.42 ppm disappears, and the signals
corresponding to the phenylic protons appear as broad multiplets
between 6.8 and 7.6 ppm). The signals at 3.66 and 5.99 (s)
ppm (see arrows in Figure 3) appear with almost the same
intensity, whereas the signal at 3.85 (s) ppm (3.76 ppm at 0
°C) has almost disappeared. Signals corresponding to the NBu4
cation also become more apparent when we increased the
temperature. Other new signals (less intense) appear at 5.28,
5.37, 6.09, 6.27, 6.03, 6.47, and 6.57 ppm (as singlets). At this
(52) Carbo´, J. J.; Crochet, P.; Esteruelas, M. A.; Jean, Y.; Lledo´s, A.; Lo´pez,
A. M.; On˜ate, E. Organometallics 2002, 21, 305.
(53) Harman, W. D.; Dobson, J. C.; Taube, H. J. Am. Chem. Soc. 1989, 111,
3061.
9
J. AM. CHEM. SOC. VOL. 125, NO. 39, 2003 11931

A R T I C L E S
Casado et al.
Figure 8. Proposed structures for 19 and 20.
C6F5-C6F5 (19).56 We also observe signals at -139.3 (d, Fo),
-154.3 (t, Fp), and -162.6 (q, Fm). We propose that this
1
compound is that showed in Figure 8 (20). In the H NMR
spectrum at 80 °C, the signals of the vinyl protons (the phenyl
protons cannot be observed as they are overlapped by other
signals) appear as a complicated signal at 5.91 that can be
described as a triplet of triplets of doublets. The coupling
between the two protons must be very small, and thus the signal
appears as a ttd (4JFo-H ) 10 Hz, JFm-H ) 7 Hz, JH-H ) 3
Hz). Signals corresponding to cis- and trans-stylbene can also
be observed (although phenyl protons are overlapped by other
major products).
At 90 °C, the 1H NMR and 19F{1H} NMR spectra are almost
identical. The nature of the final products can be confirmed by
a GC-MS analysis of the sample (after removal of the toluene-
d8, addition of MeOH, and filtration of abundant metallic gold
through silica). We can identify complexes 2 (retention time
6.69 min), C6F5-C6F5 (19, 7.29 min), cis- and trans-stylbene
(10.10 and 11.50 min), and cis-pentafluorophenylphenylethylene
(20, 13.01 min). Although the molecular peak of 20 does not
appear in its mass spectrum, peaks that can be assigned to
species such as [C7H7] (m/z ) 91), [C6F5 + C] (m/z ) 179),
and [M - Ph] (m/z ) 194) can be identified.
To gain more information about the nature of the plausible
gold(III) intermediates, we carried out the stoichiometric reaction
of compound 10 and deuterated phenylacetylene (21) at the
NMR scale (molar ratio 1:1). The reaction was followed by
2H{1H} NMR in nondeuterated toluene at different temperatures.
However, in our hands, we did not get satisfactory data. To
obtain a proper deuterium spectrum, long acquisition times
(minimum of 0.5 h) are necessary, and, even at low tempera-
tures, the reaction evolves quite quickly.
4.2. Proposed Mechanism for the Hydration of Phenyl-
acetylene with [Au(C6F5)2Cl]2. Although we have not been
able to isolate gold(III) intermediates for the addition of water
and MeOH to terminal alkynes, we believe that we have
identified spectroscopically at low temperature the gold(III)-
alkyne and the gold(III)-vinyl intermediates for these reactions.
We have confirmed that organometallic gold(III) compounds
with pentafluorophenyl as ligand do not decompose to gold(I)
intermediates in their reaction with phenylacetylene. Gold(III)
C6F5-containing species (cis and trans) could be identified at
low temperatures, whereas at higher temperatures decomposition
to metallic gold and organic C6F5-containing products did occur.
The presence of final product 20 (although in small amounts)
confirms that the alkyne has been coordinated to the organo-
metallic gold(III) center at some point. The obtention of phenyl
methyl ketone (2) as the hydration product also leads to the
hypothesis of a gold(III)-alkyne intermediate that may suffer a
Figure 7. 19F{1H} NMR spectrum of the reaction of 1 and 10 in toluene-
d8 at 60 °C.
be isolated (although they were not properly characterized) by
reaction of cis-[(Me3P)Au{(F3C)CdC(CF3)}trans-AuMe2(PMe3)]
or cis-[(Me3P)Au{(F3C)CdC(CF3)}cis-AuMe2(PMe3)] with HCl,
respectively.50d The latter (that could be characterized by X-ray
crystallography) were obtained by reaction of hexafluorobut-
2-yne and [AuMe(PMe3)] at low temperature (-80 °C).50
6
3
Although it is quite difficult to connect the H signals
(observed in the proton NMR spectra) to the signals assigned
to pentafluorophenyl groups (observed in the 19F{1H} NMR
spectra), we may make the following assumption. At low
temperature (from -80 to 0 °C), we mainly observed signals
that can be assigned to pentafluorophenyl groups of two different
gold(III) species that are in a cis disposition. We believe that
these compounds (shown in Figure 2, species A and B) are also
those represented in Figures 4 and 5 (whose terminal protons
are detected by 1H NMR between -80 and 0 °C). The alkyne-
gold(III) intermediate seems to be more unstable, and it almost
disappears at 0 °C (converting itself into the gold(III)-vinyl
species). The gold(III) species that contains the pentafluoro-
phenyl groups in a trans disposition (Figure 2, species C) is in
minor proportion, and we have not found a connection between
this species and signals observed in the 1H NMR spectra.
However, C can be in equilibrium with A or B as shown by
the 19F{1H} NMR at low temperature (see description for spectra
at -50 and -20 °C). When the temperature is increased, the
signals in the 1H NMR spectra (assigned to the proposed
gold(III)-alkyne and gold(III)-vinyl species) become smaller at
the same time that the signals assigned to gold(III) species A
and B (19F{1H}) spectra do.
At 20 °C, we observe in the 19F{1H} NMR spectrum how
the signals corresponding to organic products containing C6F5
groups increase their intensity with respect to Au(III)-C6F5
1
species. The change is more evident at 40 °C. In the H NMR
spectrum (40 °C), the signals at 3.67 and 5.99 ppm become
very small, and the signals corresponding to the hydration
product PhCOCH3 (2) (mainly the signal at 2.12 (s) PhCOCH3)
begin to appear.
At 60 °C, the signals at 3.67 and 5.99 disappear completely,
and the 19F{1H} NMR spectrum shows mainly signals that can
be assigned to the above-mentioned organic C6F5-containing
products (Figure 7).
1
At 80 °C, we mostly observe in the H NMR spectrum the
signals corresponding to PhCOCH3 (2), whereas in the 19F{1H}
NMR spectrum the intensities of the signals at -138.1 (m, 2F,
Fo), -150.1 (t, 1F, Fp), and -160.9 (q, 2F, Fm) are very high
and correspond to a major product, the previously described
(54) von Firsk, G.; Hausmann, H.; Francke, V.; Gu¨nther, H. J. Org. Chem. 1997,
62, 5074.
(55) An example: Bennett, M. A.; Bhargava, S. K.; Hockless, D. C. R.; Welling,
L. L.; Willis, A. C. J. Am. Chem. Soc. 1996, 118, 10469.
(56) Long, D. A.; Seele, D. Spectrochim. Acta 1968, 24A, 1125.
9
11932 J. AM. CHEM. SOC. VOL. 125, NO. 39, 2003

Organometallic Gold(III) Complexes as Catalysts
A R T I C L E S
Scheme 8
nucleophilic attack of a OH- molecule. With all of the collected
spectroscopic and chromatographic data, we would like to
propose the following mechanism for the hydration of phenyl-
acetylene (1) with [Au(C6F5)2Cl]2 (10) as catalyst (Scheme 8).
First, the alkyne coordinates to the organometallic gold(III)
complex as a two-electron ligand to give 22. The water molecule
(dissociated in H+ and OH-) in polar solvents such as MeOH
reacts with 22 to give 23 by nucleophilic attack of the OH-
group. The attack is produced on the carbon that contains the
aryl substituent and that would generate, therefore, the more
stable carbocation. The H+ can generate a stable ionic complex
23 as shown in Scheme 8, or it can generate HCl, giving a
neutral complex with a coordinated solvent molecule. A
subsequent keto-enol equilibrium would afford intermediate
24 that would give phenyl methyl ketone (2) and regenerate 10
as the catalyst by reaction with H+ that is present in the medium
(or HCl). The presence of at least one chloride ligand coordi-
nated to the gold(III) center seems essential for the reaction to
proceed.
In the case of ionic gold(III) compounds, the first step may
imply the dissociation of one chloride ligand and elimination
of QCl from the gold(III) precatalyst to generate the plausible
gold(III) active [“AuCl3”], [“AuCl2R”], or [“AuClR2”] catalytic
species. The reactions need refluxing MeOH temperature to
proceed, although in the case of neutral [Au(C6F5)2Cl]2 (10)
total conversion of 1 to 2 can be achieved at room temperature
after 11 days. These gold(III) catalytic species decompose to
gold(0), and, in the case of organometallic compounds, they
give organic R-R compounds by reductive carbon-carbon
coupling under reaction conditions. Examples of such gold-
mediated couplings are well documented in the literature.55
We believe that the addition of some acidic cocatalysts to
these gold(III) complexes improve their catalytic activity a little
because they prevent the reduction to metallic gold.
coordinate an alkyne molecule (as claimed by Teles and co-
workers29 and Tanaka et al.30). However, the addition of acidic
cocatalysts in this case not only helps to form stable [AuL]+
fragments (generated from [AuMeL]/acid) but may also prevent
the decomposition of these catalytic gold(I) systems to gold(0)
species.
Conclusion
In summary, we have proven that anionic and neutral
organometallic gold(III) complexes (with one or two organic
radicals) are able to catalyze the addition of water and methanol
to terminal alkynes with a catalytic activity similar to that
reported for Na[AuCl4]. The addition reactions follow the
Markovnikov rule, giving the corresponding ketones or diacetal
products. The reactions take place in neutral medium, although
the addition of acidic cocatalysts improves their catalytic activity
(plausibly by avoiding the decomposition of these catalysts to
metallic gold(0)). The presence of at least one chloride ligand
coordinated to the gold(III) center seems essential for the
reaction to proceed, whereas the presence of even weakly
coordinating neutral ligands seems to inhibit the catalytic
process. Anionic complexes need refluxing methanol temper-
ature, whereas neutral complexes are able to catalyze the
hydration of alkynes at room temperature in neutral medium
(although very slowly). The pentafluorophenyl ligand coordi-
nated to gold(III) centers behaves in a way similar to chloride
ligands. It seems that gold(III) compounds are the active catalytic
species for these processes as gold(I) intermediates were not
identified by spectroscopic methods. Moreover, the spectro-
scopic study of stoichiometric reactions between the complex
[Au(C6F5)2Cl]2 and phenylacetylene at variable temperature has
allowed us to identify two plausible gold(III) intermediates for
the hydration reactions. Thus, we propose that the terminal
alkyne coordinates to a gold(III) center as the first step of the
reaction and ulterior nucleophilic attack of an hydroxyl group
to the coordinated alkyne gives a gold(III)-vinyl intermediate
The mechanism of gold(I) complexes is different. We believe
that the catalytic species are [AuL]+ stable fragments that can
9
J. AM. CHEM. SOC. VOL. 125, NO. 39, 2003 11933

A R T I C L E S
Casado et al.
mmol) in 20 mL of Et2O. After 20 min of stirring at room temperature,
the AgCl formed was removed by filtration through Celite. The resulting
solutions were concentrated to ca. 5 mL, and when 20 mL of cold di-
ethyl ether was added, BzPPh3ClO4 precipitated and was removed by
filtration. Solutions of [Au(mes)Cl2]2 (4) and [Au(C6F5)Cl2]2 (8) in cold
Et2O could be used subsequently. The complex cis-PPN[Au(mes)2Cl2]
(5)31b was prepared via an alternative gold(I)-transmetalation method
(which prevents the use of toxic [Hg(mes)2]).
Preparation of cis-PPN[Au(mes)2Cl2] (5). To a solution of
PPN[AuCl4] (0.438 g, 0.5 mmol) in 30 mL of acetone was added
[Au(mes)(PPh3)] (0.723 g, 1.25 mmol). The reaction mixture was
refluxed for 2 h. The resulting colorless solution was then cooled at
room temperature and concentrated to ca. 5 mL. When 40 mL of toluene
was added, the compound cis-PPN[Au(mes)2Cl2] (5) was obtained as
a white solid. The crude product was filtered and washed with 3 × 20
mL of toluene. After being dried under vacuum, cis-PPN[Au(mes)2Cl2]
(5) was obtained as a pure compound and could be used without further
purification (0.485, 92% yield). The toluene filtrate was concentrated
to ca. 5 mL, and when 20 mL of cold n-hexane was added, [AuCl-
(PPh3)] precipitated as a white solid. The solid was obtained in 85%
yield (0.841) as an analytically pure compound.
that by subsequent keto-enol equilibrium is able to form the
ketone and regenerate the initial gold(III) catalytic species. We
were able to prove that the decomposition of the catalyst is due
to reductive elimination at the gold(III) center, giving the
coupling organic product bis-pentafluorophenyl and metallic
gold.
Furthermore, the present study demonstrates that organo-
metallic gold(III) complexes behave similarly to Lewis acids.
More importantly, the incorporation of organic fragments to the
gold(III) centers not only seems to stabilize plausible metallic
intermediates in nucleophilic additions to triple bonds but it also
has helped to identify these intermediates by spectroscopic
methods. A rational design and synthesis of organic and
inorganic ligands able to coordinate to gold(III) and gold(I)
centers that can modify electronic and steric properties and, thus,
improve their catalytic activity in different chemical processes
is, therefore, anticipated.
Additions of nucleophiles to triple and double bonds catalyzed
by gold(III) and gold(I) organometallic and inorganic com-
pounds are currently under investigation by our research group.
General Procedure for the Hydration of Terminal Alkynes. 1.
Neutral Media. To a stirring solution of 1 mmol of phenylacetylene
(1) (0.110 mL) or n-heptyne (15) (0.130 mL) and water (0.5 mL) in 5
mL of MeOH was added the amount of gold compound as specified in
Tables 1, 2, and 4, and the mixture was heated at reflux for 90 min for
4 h. Prior to workup, the metallic gold that formed was removed by
filtration through a short silica column, and the resulting filtrate was
analyzed by GCMS. If conversions were higher than 90%, the reaction
mixture was concentrated under reduced pressure and the residue was
diluted with Et2O and washed with a 1:1 mixture of brine and aqueous
ammonia. The ethereal solution was dried over anhydrous MgSO4 and
concentrated to give the products (phenyl methyl ketone (2) or
Experimental Section
General Methods. All reactions with air- or water-sensitive
compounds were carried out under standard Schlenk line techniques.
Reactions of gold compounds with silver salts must be carried out
avoiding light exposure until the silver chloride is removed. Caution:
perchlorate salts of metal complexes with organic ligands are potentially
explosive. Only small amounts of material should be prepared, and all
1
samples should be handled with great caution. H and 19F{1H} NMR
spectra were obtained on a 300 MHz Varian GEMINI 2000, Varian
UNITY-300, or Bruker ARX-300 apparatus in CDCl3 or toluene-d8
solutions; chemical shifts were quoted relative to SiMe4 (external, 1H)
1
2-heptanone (16)) as pure colorless oils as characterized by H NMR
2
and CFCl3 (external, 19F{1H}). H{1H} NMR spectra were obtained
(see Supporting Information).
2. Acidic Media. To a stirring solution of 1 mmol of phenylacetylene
(1) (0.110 mL) or n-heptyne (15) (0.130 mL) and water (0.5 mL) in 5
mL of MeOH was added the amount of gold compound and acid (HtfO
or H2SO4) in a molar ratio 1:10 as specified in Tables 2 and 4, and the
mixture was heated at reflux for 90 min. Prior to workup, the metallic
gold formed was removed by filtration through a short silica column,
and the resulting filtrate was analyzed by GCMS. If conversions are
90% or higher, we followed the previously described procedure to
obtain 2 or 16 as pure colorless oils.
General Procedure for the Addition of Methanol to Terminal
Alkynes. To a solution of 1 mmol of phenylacetylene (1) (0.110 mL)
or n-heptyne (15) (0.130 mL) in 5 mL of anhydrous MeOH was added
the amount of gold compound as specified in Tables 3 and 5, and the
mixture was heated at reflux under Ar for 90 min. The reaction mixture
was analyzed by GCMS (after addition of 0.5 mL of triethylamine and
elimination of metallic gold). If the conversion was higher than 90%,
we follow the previously described workup to obtain compounds 12
and 17 as pure colorless oils or mixtures of 2/12 and 16/17 (see Tables
3 and 5 and the Results and Discussion section) as characterized by
1H NMR (see Supporting Information).
Stoichiometric Reaction between [Au(C6F5)2Cl]2 (10) and Phenyl-
acetylene (1). To a solution of 0.023 g (0.04 mmol) of [Au(C6F5)2Cl]2
(10) in 0.5 mL of dry toluene-d8 at -80 °C (NMR tube) was added
phenylacetylene (1) (5 µL, 0.04 mmol) at this temperature, and the
tube was placed immediately in the NMR probe that had been cooled
to -80 °C. 1H and 19F{1H} NMR spectra were recorded at -80, -60,
-50, -20, 0, 20, 40, 60, 80, and 90 °C. A 19F{1H}-19F{1H} COSY
NMR experiment was carried out at -50 °C (before warming the
reaction mixture to 0 °C, see tabulated data in Supporting Information).
The final mixture was concentrated to dryness and dissolved with 2
mL of MeOH. Metallic gold was removed by filtration through a silica
on a 300 MHz Varian UNITY-300 apparatus in dry toluene solutions;
chemical shifts were quoted relative to Si(CD3)4. 19F{1H}-19F{1H}
COSY NMR experiments were recorded on a Bruker ARX-300
apparatus. GCMS spectra were obtained on a Hewlet-Packard 6890
(with an HP 5973 selective mass detector) apparatus using an HP-
1900915-433 (5% phenyl methyl siloxane) column. The method
employed was as follows: 2 min at 50 °C, 4 min at 15 °C/min, 2 min
at 110 °C, 8 min at 20 °C/min, and 3 min at 270 °C.
Materials. Commercially available toluene-d8 (Aldrich) was dried
with a sodium/potassium amagalm under argon and degassed by three
freeze-pump-thaw cycles. Toluene-d8 was stored over activated 4 Å
molecular sieves. Toluene, diethyl ether, and tetrahydrofuran were
distilled from sodium benzophenone ketyl under argon. Methanol was
distilled from magnesium under argon. Phenylacetylene (1), phenyl-
acetylene-D (21), and n-heptyne (14) were purchased from Aldrich and
further purified by distillation under argon prior to use. They were stored
under argon over activated 4 Å molecular sieves. H2SO4 and HSO3CF3
were purchased from Aldrich. [AuCl3(tht)],33 [AuCl3(PPh3)],34 BzPPh3-
[Au(mes)Cl3] (3),31a BzPPh3[Au(C6F5)Cl3] (7),32a [Au(C6F5)Cl2(tht)],35
PPN[AuCl4],36 [Au(mes)(PPh3)],31a cis-[Au(mes)2Cl]2 (6),31b cis-
[Au(mes)2{P(tol)3}Cl],31b cis-[Au(mes)2{AsPh3}Cl],31b cis-[Au(mes)2-
(bipy)]ClO4,31b trans-NBu4[Au(C6F5)2X2] (X ) Cl 9a, Br 9b, I 9c),39
[Au(C6F5)2Cl]2 (10),39 [Au(C6F5)2S*2]X (X ) ClO4-, SO3CF3-),40 cis-
[Au(C6F5)2(CH2PPh3)2],41 trans-[Au(C6F5)2(CH2PPh3)2],41 [Au(C6F5)3-
(tht)],42 PPN[Au(mes)Cl],31a NBu4[Au(C6F5)2],43 [AuCl(PPh3)],44
[Au(Me)(PPh3)] (11),45 [Au(C6F5)(PPh3)],46 [O{Au(PPh3)}3]BF4,47
[Au(PPh3)2]ClO4,48 AgClO4, and AgOSO2CF3 were prepared as
31a
described previously. [Au(mes)Cl2]2 (4) and [Au(C6F5)Cl2]2 (8) were
prepared in situ in the following way: to a solution of BzPPh3[AuRCl3]
(0.311 g, 0.4 mmol, R ) mes (3); 0.329 g, 0.4 mmol, R ) C6F5 (7))
in 20 mL of CH2Cl2 was added a solution of AgClO4 (0.091 g, 0.44
9
11934 J. AM. CHEM. SOC. VOL. 125, NO. 39, 2003

Organometallic Gold(III) Complexes as Catalysts
A R T I C L E S
column, and the resulting MeOH solution was subsequently analyzed
by GC-MS chromatography. The collected NMR and chromatographic
data are presented in the Discussion section, and detailed tabulated data
have been collected in the Supporting Information section.
Supporting Information Available: Tables of spectroscopic
and analytical data from varying gold catalysts; spectral and
chromatographical data for all organic reaction products (PDF).
This material is available free of charge via the Internet at
http://pubs.acs.org.
Acknowledgment. We thank the MCYT for financial support
(BQU2002-0409-CO2-01). M.C. thanks the MCYT-Universidad
de Zaragoza for her Research Contract “Ramo´n y Cajal”.
JA036049X
9
J. AM. CHEM. SOC. VOL. 125, NO. 39, 2003 11935