Isolable gold(I) complexes having one low-coordinating ligand as catalysts for the selective hydration of substituted alkynes at room temperature without acidic promoters

Article abstract of DOI:10.1021/jo802558e

Hydration of a wide range of alkynes to the corresponding ketones has been afforded in high yields at room temperature by using gold(I)-phosphine complexes as catalyst, with no acidic cocatalysts required. Suitable substrates covering alkyl and aryl terminal alkynes, enynes, internal alkynes, and propargylic alcohols, including enantiopure forms, are cleanly transformed to the corresponding ketones in nearly quantitative yields. Acid-labile groups present in the substrates are preserved. The catalytic activity strongly depends on both the nature of the phosphine coordinated to the gold (I) center and the softness of the counteranion, the complex AuSPhOsNTf₂ showing the better activity. A plausible mechanism for the hydration of alkynes through ketal intermediates is proposed on the basis of kinetic studies. The described catalytic system should provide an efficient alternative to mercury-based methodologies and be useful in synthetic programs.

Full text of DOI:10.1021/jo802558e

Isolable Gold(I) Complexes Having One Low-Coordinating Ligand
as Catalysts for the Selective Hydration of Substituted Alkynes at
Room Temperature without Acidic Promoters
Antonio Leyva* and Avelino Corma*
Instituto de Tecnolog´ıa Qu´ımica UPV-CSIC, AVda. de los Naranjos s/n, 46022, Valencia, Spain
anleyVa@itq.upV.es; acorma@itq.upV.es
ReceiVed NoVember 18, 2008
Hydration of a wide range of alkynes to the corresponding ketones has been afforded in high yields at
room temperature by using gold(I)-phosphine complexes as catalyst, with no acidic cocatalysts required.
Suitable substrates covering alkyl and aryl terminal alkynes, enynes, internal alkynes, and propargylic
alcohols, including enantiopure forms, are cleanly transformed to the corresponding ketones in nearly
quantitative yields. Acid-labile groups present in the substrates are preserved. The catalytic activity strongly
depends on both the nature of the phosphine coordinated to the gold (I) center and the softness of the
counteranion, the complex AuSPhosNTf2 showing the better activity. A plausible mechanism for the
hydration of alkynes through ketal intermediates is proposed on the basis of kinetic studies. The described
catalytic system should provide an efficient alternative to mercury-based methodologies and be useful in
synthetic programs.
SCHEME 1. Catalytic Hydration of Alkynes
Introduction
The hydration of alkynes to give the corresponding carbonyl
compounds (Scheme 1) represents a model of modern, sustain-
able transformation in chemistry: waste-free, water as reagent,
and catalytic.1-3 Curiously, although this reaction has been
known since the 19th century,4 its potential in organic synthesis
and/or in industrial processes has not been conveniently
exploited. The reason for that is the more active catalysts known
to date, HgO-H2SO4 (Kucherov catalyst)4,5 and HgO-BF3
(Hennion-Nieuwland catalyst)6 are highly toxic mercury salts
that, in addition, must be used in combination with either
Bronstead or Lewis acids.7 However, these catalysts were
extensively used in high-scale industrial processes8 until the
discovery of the toxicity of mercury salts.
Alternative catalysts for the hydration of alkynes have been
searched for over the past 30 years, including Bronsted acids,
bases, and metal salts and complexes.2 However, neither of them
has been able to surpass the activity of those systems containing
mercury salts. Among them, it is worth mentioning those
9
systems where simple salts, such as PdCl2 or Zeise’s dimer
[{PtCl2(C2H4)}2],10 were used, along with the recently reported
anti-Markornikov hydration of alkynes catalyzed by ruthenium
complexes.3 However, as far as we know, a common charac-
teristic for most of the catalytic systems described until now,
including the HgO catalysts, indicates that the use of an
additional acid or the heating of the system, or both, are
absolutely necessary in order to obtain good results. Very few
examples can be found where simple alkynes are hydrated by
(1) Alonso, F.; Beletskaya, I. P.; Yus, M. Chem. ReV. 2004, 104 (6), 3079.
(2) Hintermann, L.; Labonne, A. Synthesis 2007, 8, 1121.
(3) Bruneau, C.; Dixneuf, P. H. Chem. Commun. 1997, 507.
(4) Kutscheroff, M. Chem. Ber. 1884, 17, 13.
(5) Kutscheroff, M. G. Chem. Ber. 1909, 42, 2759.
(6) (a) Hennion, G. F.; Killian, D. B.; Vaughn, T. H.; Nieuwland, J. A. J. Am.
Chem. Soc. 1934, 56, 1130. (b) Nieuwland, J. A.; Vogt, R. R.; Foohey, W. L.
J. Am. Chem. Soc. 1930, 52, 1018.
(7) (a) Killian, D. B.; Hennion, G. F.; Nieuwland, J. A. J. Am. Chem. Soc.
1934, 56, 1786. (b) Killian, D. B.; Hennion, G. F.; Nieuwland, J. A. J. Am.
Chem. Soc. 1936, 58, 80.
(8) Acetaldehyde. In Ullmann’s Encyclopedia of Industrial Chemistry, 7th
ed.; Wiley-VCH: Weinheim, 2006.
(9) Utimoto, K. Pure Appl. Chem. 1983, 55, 1845.
(10) Hiscox, W.; Jennings, P. W. Organometallics 1990, 9, 1997.
10.1021/jo802558e CCC: $40.75
Published on Web 01/26/2009
2009 American Chemical Society
J. Org. Chem. 2009, 74, 2067–2074 2067

Leyva and Corma
TABLE 1. Optimization of the Catalyst, Solvent, and Water
Amount for the Hydration of 1-Octyne to 2-Octanone at Room
Temperature
SCHEME 2. Two Ways To Obtain the Cation Gold(I)
Triphenylphosphine
entry solvent catalysta (mol % of Au) water (equiv) ketoneb (%)
1
2
3
4
5
6
7
8
DCM
AuClPPh3 + AgOTf (5)
AuClPPh3 + AgOTf (5)
AuClPPh3 + AgOTf (5)
AuClPPh3 + AgOTf (5)
[AuPPh3OTf] (5)
24
7
using a transition metal species at room temperature without
help of acidic additives.11
1.5
4
1
1
1
1
1
1
1
1
1
1
1
4
4
4
4
<1
18
76
31
5
Gold chemistry has experienced a great revival in the last
years because of the unexpected properties of the supported
metal in the nanoparticle size12,13 and the reactivity of its
compounds.13 In particular, gold shows a high affinity for
alkynes as a Lewis acid.14 The “alkynephilicity” of gold
resembles that of mercury, and this behavior has been attributed
to the common electronic properties for these two elements in
the valence shell, in part due to relativistic effects.15 Therefore,
the use of gold instead of mercury as catalyst for the hydration
of alkynes is worthy of study. Thus, Utimoto16 and Teles17
applied gold salts and complexes as catalysts for the hydration
of alkynes and, more recently, Hayashi and Tanaka,18 Laguna,19
and others.20 They generally reported good yields, especially
when gold(I) complexes were combined with strong acids as
catalysts under heating. The active catalytic species in these
reactions is the cationic gold complex [(Ph3P)Au]+, generated
in situ from [(Ph3P)AuCH3] after treatment with a strong acid
THF
CH3CN [AuPPh3OTf] (5)
toluene [AuPPh3OTf] (5)
THF
(PPh3Au)3OBF4 (5)
[Au(CF3C6H4)3POTf] (5)
[AuPEt3OTf] (5)
[AuPtBu3OTf] (5)
AuPPh3NTf2 (5)
[AuPPh3OTf] (1)
AuPPh3NTf2 (1)
AuPPh3NTf2 (1)
0
9
30
37
90
87
0
10
11
12
13
14
15
16
17
18
6
19
100
55
22
MeOH AuPPh3NTf2 (1)
EtOHc AuPPh3NTf2 (1)
iPrOH
AuPPh3NTf2 (1)
a AuClPR3 + AgX refers to reactions where the AgCl has not been
removed by filtration; [AuPR3X] refers to the filtrated catalyst, and
AuPR3X refers to the pure, isolated compound. b GC yield.
c Denaturalized EtOH containing ppms of thiols gave results similar to
those for EtOH-grade HPLC.
17
and release of CH4 (Scheme 2). The excess acid can act as
both cocatalyst and stabilizer of the cationic gold (I) species
under heating conditions.19
These complexes will ultimately be isolated and used directly
as stable-bench solids. We report here the results obtained for
the hydration of alkynes using isolable gold(I) cationic com-
plexes at room temperature.
Although the results obtained with those acid-generated
gold(I) cationic complexes are excellent, the window of ap-
plication in more elaborated compounds is narrow, since a
moderate acidic media under heating is necessary. A more
favorable media would be obtained if the gold(I) cation is
generated by a well-known procedure21 (Scheme 2) where the
corresponding chloride complexes are treated in situ with a silver
salt, thus precipitating AgCl and forming the new complex. A
soft, noncoordinating anion such a triflate (OTf) or bis(trifluo-
romethanesulfonyl)imidate (NTf2) would make the Au(I) cata-
lytic center sufficiently acidic to perform the reaction without
additives. Moreover, this acidity can be modulated by changing
both the counteranion and the phosphine present in the complex.
Results and Discussion
Study of the Reaction Conditions: Synthesis and Optimi-
zation of the Catalysts. The hydration of 1-octyne to 2-octanone
catalyzed by different gold(I) cationic complexes was selected
as the standard reaction. The results obtained for different
catalysts, water amounts, and solvents are shown in Table 1.
As can be seen, the use of in situ generated AuPPh3OTf in
standard nondry dichloromethane solvent gave a surprising 24%
of ketone (entry 1). However, addition of more water resulted
in a decreased yield, probably due to formation of two phases.
THF, a water-miscible solvent, was found to be a suitable
solvent under similar conditions, and more importantly, an
excellent 76% yield was obtained when the precipated AgCl
was removed from the medium by filtration (compare entries 4
and 5). A systematic study of the catalyst was then accomplished
in this solvent. It was found that a bulky donor phosphine such
as tBu3P was equally suitable when using OTf as counteranion
(entry 11). However, an important increase in the yield of ketone
was observed when the counteranion was varied from OTf to
the softer, less-coordinating NTf2 (see also Figure 1 below).22,23
This reflects the cationic nature of the metallic active site, which
is free to interact with the alkyne when a less-coordinating anion
is present. As expected, the water-miscible solvents are more
suitable for the hydration since they admit more equivalents of
water in order to accelerate the reaction, with MeOH being the
(11) (a) Nishizawa, M.; Skwarczynski, M.; Imagawa, H.; Sugihara, T. Chem.
Lett. 2002, 31 (1), 12. In some cases, thermal heating without acidic promoters
can be enough to achieve the hydration, see: (b) Chang, H.; Datta, S.; Das, A.;
Odedra, A.; Liu, R. Angew. Chem., Int. Ed. 2007, 46 (25), 4744.
(12) (a) Budroni, G.; Corma, A. Angew. Chem., Int. Ed. 2006, 45 (20), 3328.
(b) Corma, A.; Serna, P. Science 2006, 313 (5785), 332. (c) Abad, A.;
Concepcion, P.; Corma, A.; Garcia, H. Angew. Chem., Int. Ed. 2005, 44 (26),
4066.
(13) Hashmi, A. S. K.; Hutchings, G. J. Angew. Chem., Int. Ed. 2006, 45,
7896.
(14) Jimenez-Nun˜ez, E.; Echavarren, A. M. Chem. Commun. 2007, 333.
(15) Gorin, D. J.; Toste, F. D. 2007, 446, 395.
(16) Fukuda, Y.; Utimoto, K. J. Org. Chem. 1991, 56, 3729.
(17) (a) Teles, J. H.; Brode, S.; Chabanas, M. Angew. Chem., Int. Ed. 1998,
37 (10), 1415. (b) Teles, J. H.; Schulz, M. Chem. Abstr. 1997, 127, 121–499;
BASF AG, WO-A1 9721648, 1997.
(18) Mizushima, E.; Sato, K.; Hayashi, T.; Tanaka, M. Angew. Chem., Int.
Ed. 2002, 41 (23), 4563.
(19) Casado, R.; Contel, M.; Laguna, M.; Romero, P.; Sanz, S. J. Am. Chem.
Soc. 2003, 125, 11925.
(20) (a) Roembke, P.; Schmidbaur, H.; Cronje, S.; Raubenheimer, H. J. Mol.
Catal. A 2004, 212, 35. For a recent report about the formation of cyclic acetals
and thioacetals, see: (b) Santos, L. L.; Ruiz, V. R.; Sabater, M. J.; Corma, A.
Tetrahedron 2008, 64 (34), 7902. (c) Liu, B.; De Brabander, J. K. Org. Lett.
2006, 8, 4907–4910.
(22) Mezailles, N.; Ricard, L.; Gagosz, F. Org. Lett. 2005, 7 (19), 4133. In
the present work, the commercially available, dimeric toluene adduct AuPPh3NTf2
complex (Sigma-Aldrich) is used.
(21) Kennedy-Smith, J. J.; Staben, S. T.; Toste, F. D. J. Am. Chem. Soc.
2004, 126, 4526.
(23) Mathieua, B.; Ghosez, L. Tetrahedron 2002, 58, 8219.
2068 J. Org. Chem. Vol. 74, No. 5, 2009

Isolable Gold(I) Complexes as Catalysts
FIGURE 1. Kinetic study for different gold catalysts: AuSPhosNTf2 (*), AuSPhosCl + AgNTf2 (×), AuSPhosCl + AgOTf (2), AuPPh3Cl +
AgNTf2 (9), AuPPh3NTf2 (().
TABLE 2. Hydration of 2-Methylbuten-3-yne
TABLE 3. Hydration of Oct-1-ynylbenzene
entry
catalyst (mol % of Au)
ketonesa (%)
1
2
3
AuPPh3NTf2 (1)
AuSPhosOTf (10)
AuSPhosNTf2 (1)
30 (50:50)
100 (60:40)
100 (60:40)
entry
catalyst (mol % of Au)
ketonea (%)
1
2
3
AuPPh3NTf2 (2)
AuPPh3NTf2 (5)
AuSPhosNTf2 (5)
32
73
90
a GC yield. Molar ratios are shown in parentheses.
a GC yield.
Moreover, we managed to obtain the less stable triflate derivative
AuSPhosOTf in pure form as well. Thus, a second study
including these isolated catalysts was undertaken under the
optimal conditions found in Table 1 (MeOH as solvent, 4 equiv
of water). The results obtained are shown in Table S1 (Sup-
porting Information). A similar behavior in the catalytic activity
for different gold(I) complexes, depending on the nature of the
phosphines and the counteranion, was obtained by using MeOH
as solvent instead of THF (compare, for instance, entries 5 and
8 in Table S1, Supporting Information with entries 11 and 9 in
Table 1, respectively). Incidentally, the presence of AgCl does
not have as a dramatic effect in MeOH as in THF (compare
entries 3 and 4, 6 and 7). More importantly, the use of bulky
solvent of choice (entry 16). However, the use of others
alcoholic solvents clearly decreased the reaction rate, which is
related to the formation of ketals as intermediates (entries
16-18, see the mechanism study below).
The bis(trifluoromethanesulfonyl)imidate triphenylphosphine
gold(I) complex is an easy-to-make air-stable solid recently
described by Gagosz and co-workers.22 By analogy to its
procedure, and in order to improve the activity of the catalyst,
we synthesized complexes containing this weakly coordinated
counteranion and bulkier, better donor phosphines such as tBu3P
or 2-dicyclohexylphosphino-2′,6′-dimethoxybiphenyl (SPhos).
J. Org. Chem. Vol. 74, No. 5, 2009 2069

Leyva and Corma
TABLE 4. Hydration of Different Propargylic Alcohols
SCHEME 3. Hydration of the Two Enantiomers of a Chiral
Propargylic Alcohol
such as the camphor derivative in Table 4 (entries 10 and 11)
did not react even in the presence of 5 mol % of AuSPhosNTf2.
The results obtained for propargylic alcohols are particularly
relevant, since these compounds are prone to give side reactions
under acidic or heating conditions, namely the Meyer-Schuster
or Rupe rearrangements,24 or simply polymerize. As mentioned
above, the majority of reports for the hydration of alkynes need
such conditions, and scarce examples about efficient hydration
of propargylic alcohols can be found in the literature except
when Kucherov conditions are used.25 In contrast, the results
in Table 4 show that the use of the neutral, mild conditions
reported herein avoid side reactions, giving cleanly the corre-
sponding R-hydroxy ketones in high yields. These products are
of paramount importance as building blocks for more elaborated
molecules, especially those having the hydroxylic carbon as a
chiral center. Therefore, the hydration of the chiral propargylic
alcohol 3-butyn-2-ol, in the two enantiomerically pure forms S
and R, was tested under the reaction conditions depicted in
Scheme 3.
a GC yield. Yields of byproducts are shown in parentheses. b After
48 h. c When AuPPh3NTf2 (5 mol %) was used, no solvent was required
to obtain a 72% yield. d System diluted four times, 16 equiv of water.
e 53% product recovered after workup, together with a 15% of the
Schuster-Meyer rearrangement product formed during workup.
f Byproduct in some cases.
The corresponding S and R acetoins were cleanly obtained
at room temperature; no racemization was found by chiral GC
and optical rotation analysis. The same results were obtained
with half the catalyst (1 mol %) after 48 h. This simple and
practical way to obtain chiral R-hydroxy ketones represents an
alternative to current processes such as enzymatic resolution
or the enantiomeric hydrogenation of diketones, since the
starting materials can be easily prepared by controlled addition
of the corresponding alkyne to enantiomerically pure epoxides,
which are obtained by Jacobsen’s kinetic resolution.
A last comparison study of the two catalysts AuPPh3NTf2
and AuSPhosNTf2 was performed with molecules containing
acid-labile protecting groups. It would be extremely useful in
synthesis to have a general protocol for the hydration of triple
bonds in the presence of different protecting groups, something
lacking in the literature. Thus, alkynes containing different
silicon-protected groups as well as a trityl group were prepared
and tested under optimized conditions (Table 5). Unfortunately,
the TMS group was found to be completely deprotected as either
a hydroxyl or an alkynyl protecting group. However, both the
more robust TBDPS and the trityl protecting groups showed a
good stability under the reaction conditions.
Finally, oncewe determined that AuSPhosNTf2 is the catalyst
of choice, different terminal and internal alkynes containing
functionalities such as hydroxyl, aryl and alkyl halides, nitro,
methyl ester, and ether were hydrated at room temperature. The
results obtained are shown in Table 6.
The hydration of the different alkynes is performed in high
yields, preserving the functionalities present in the molecules.
This means an important improvement compared to previous
reports. When the substrates were not soluble solids in the
reaction medium at room temperature, such as the trityl-
donor phosphines such tBu3P or SPhos instead of PPh3 improved
the results again (entries 5, 6, 13, and 14). Similarly, the use of
the less-coordinating NTf2 led to optimal results (entry 14).
Unfortunately, the use of a even less-coordinating anion such
as B(C6F5)4 just inhibited the reaction (entry 15). Since the yield
obtained after prolonged reaction times depends not only on
the intrinsic activity but also on the stability of the catalyst, a
kinetic study was accomplished in order to assess the initial
rate and, therefore, the intrinsic activity of the catalysts
(Figure 1).
As expected, AuSPhosNTf2 gave the best result in terms of
both activity and stability over time, improving in 1 order of
magnitude the initial rate obtained with AuPPh3NTf2 (entry 1
vs 4). The result shows clearly the importance of the phosphine
for the catalytic activity of the complex, being even more
relevant than the coordination ability of the counteranion
(compare NTf2 vs OTf, entries 2 and 3).
Scope of the Reaction. AuSPhosNTf2 was found to be the
best catalyst for all cases studied. This is particularly interesting
since this complex is prepared in nearly quantitative yield from
the corresponding starting materials (see the Experimental
Section) and can be weighed in the open air. Under the optimal
experimental conditions found for the hydration of 1-octyne,
AuSPhosNTf2 was further compared to AuPPh3NTf2 for the
hydration of a wide range of alkynes, including an enyne (Table
2), an internal alkyne (Table 3), and different propargylic
alcohols (Table 4).
The presence of 1 mol % or less of AuSPhosNTf2 in the
reaction is sufficient to produce a complete hydration of the al-
kynes to the corresponding ketones in most of the cases. The
same catalytic amounts of AuPPh3NTf2 gave systematically
poorer results. Higher amounts of AuSPhosNTf2 are necessary
when sterically demanding alkynes are used (Table 2, entry 3;
Table 4, entries 8 and 9). A highly bulky propargylic alcohol
(24) Swaminathan, S.; Narayanan, K. V. Chem. ReV. 1971, 71, 429.
(25) (a) Jirkovski, I.; Cayen, M. N. J. Med. Chem. 1982, 25, 1154. (b) Somers,
P. K.; Wandless, T. J.; Schreiber, S. L. J. Am. Chem. Soc. 1991, 113, 8045.
2070 J. Org. Chem. Vol. 74, No. 5, 2009

Isolable Gold(I) Complexes as Catalysts
TABLE 5. Hydration of Alkynes Containing Protecting Groups
a GC yield. Deprotection is shown in parentheses.
TABLE 6. Scope for the Hydration of Alkynes Catalyzed by AuSPhosNTf2
a GC yield. Isolated yield is shown in parentheses. b AuPPh3NTf2 used as catalyst. When NaHCO3 (aq, 0.14 M, 72 µL, 1 mol %) was used instead of
neat water, the hydration did not proceed at all. c AuPPh3NTf2 (1 mol %) gave no complete conversion. d Reaction time: 48 h. e Isolation gave the
corresponding cyclized methyl ketal as major product together with a mixture that could contain the ketone. f AuPPh3OTf as catalyst in wet
dichloromethane. g H2O (16 equiv) and THF (3 mL) added to the mixture, reaction time: 48 h. Pure ketone obtained after treatment of the reaction
mixture with 0.83 mL of concd aqueous HCl (5 equiv) for 5 h at rt. h Mixture of isomers E/Z (2:1 molar ratio). HCl treatment, as for previous entry, did
not return the ketone, the mixture seeming to be stable.
J. Org. Chem. Vol. 74, No. 5, 2009 2071

Leyva and Corma
FIGURE 2. Kinetic study for different nucleophiles: MeOH (1 equiv, 9); H2O (1 equiv, (); MeOH (0.2 equiv) + H2O (0.8 equiv, 2); iPrOH (0.2
equiv) + H2O (0.8 equiv, ×).
SCHEME 4. Intermediates Claimed for the Hydration of
Triple Bonds Catalyzed by Gold(I) or Gold(III) Complexes
to display the same selectivity. Ab initio calculations from the
same authors17 showed that the relative affinities of the different
reagents for a theoretical cationic gold(I) center [Me3PAu]+
follow the order alkynes > methanol > water. This order
correlates with the signals observed in low temperature 31P NMR
measurements when methanol and 3-hexyne are sequentially
added to in situ formed [Ph3PAu]+[MeSO3]-.17
On the other hand, Hayashi, Tanaka, and co-workers18
claimed that, although they cannot completely exclude dimethyl
ketals as intermediates under their reaction conditions (strong
acid as cocatalyst, 70 °C), they bank on a direct attack of water
to the alkyne since the reaction proceeds in nonalcoholic solvents
as well. Besides that, Pt-catalyzed hydration has also been
reported to work without any formation of methanol adducts.10
Finally, Laguna and co-workers19 propose that the hydration
of alkynes when gold(III) complexes are used as catalysts goes
through direct attack of water.
protected alkyne (entry 7) or diphenylacetylene (entry 13), the
addition of DCM or THF as cosolvents solves the problem and
does not hamper the final yield. For highly hydrophobic liquids,
the dispersion in the aqueous methanol was good enough to
produce the reaction smoothly (entries 1 and 10). The amounts
of product recovered after isolation correlate with the yields
obtained by GC analyses.
Mechanistic Studies: Ketals as Intermediates. Previous
works on the hydration of alkynes catalyzed by gold(I)
complexes agree to report methanol as the solvent of choice,
but different hypotheses for the reaction mechanism are
proposed. On one hand, Teles and co-workers17 have described
enol ethers and ketals as plausible intermediates in the hydration
of alkynes, based on the isolation of these compounds when
water is not present in the reaction medium (Scheme 4).
According to that, coordination of methanol to the previously
formed alkyne-cationic gold(I) phosphine would lead to an
unstable tricoordinated transition species, which would evolve
into the addition of the hydroxyl group to the alkyne to form
the enol ether. This crowded transition state would explain the
low reactivity of secondary alcohols when compared to primary
alcohols, although an outer-sphere pathway would be expected
To shed light on all these, apparently, contradictory conclu-
sions, we performed a series of kinetic experiments using
MeOH, H2O, or both as nucleophiles. Diphenylacetylene was
used as substrate since the corresponding methyl enol ethers
are stable enough to be observed at room temperature, using
AuSPhosNTf2 as catalyst, in dry THF. The advantage of our
methodology over previously reported reaction conditions lies
in the absence of any acid promoter, the hydroxyl group
interacting freely over the catalyst. The results obtained are
shown in Figures 2 and 3.
According to the initial rates obtained, MeOH adds to
diphenylacetylene 2.5 times faster than H2O. A slight increase
in the initial rate occurs when 20 mol % of MeOH or iPrOH in
2072 J. Org. Chem. Vol. 74, No. 5, 2009

Isolable Gold(I) Complexes as Catalysts
FIGURE 3. Evolution of the Z and E isomers of the corresponding methyl enol ether when MeOH or H2O are the major nucleophile.
SCHEME 5. Plausible Mechanism for the Hydration of Alkynes Using AuPR3NTf2 Compounds as Catalysts
water is used as reagent mixture instead of pure water. These
results fit those obtained for the hydration of 1-octyne to
2-octanone in MeOH, EtOH, iPrOH, and THF as solvents (100,
55, 22, and 19%, respectively, see Table 1, entries 15-18),
which suggests a possible precoordination of the hydroxyl to
the metal center in the transition state.17 Concerning the
formation of Z or E enol ethers, it is observed that the Z/E ratio
increases rapidly over the time when MeOH is used as only
reagent. In contrast, the E isomer is predominant over the Z
isomer when pure H2O is used, although the tendency over the
time is similarly to increase the Z/E ratio. It has to be noted
that the Z isomer is the more stable thermodynamically in this
particular case, and this correlates with the formation of the
more stable thermodynamically E isomer of dimethylacetylene
dicarboxylate26 as major product (see Table 6, entry 14). With
this data in hand, a plausible mechanism for the hydration of
alkynes using AuPR3NTf2 compounds as catalysts at room
temperature is depicted in Scheme 5.
The formation of the enolic species (IV) occurs more rapidly
when MeOH is used instead of water. Considering than the
coordination of MeOH to the alkyne-metal complex (I) is
favored over H2O, and that this could be the controlling step,
we cannot conclude which of them, MeOH or H2O, adds
preferentially to the alkyne on the tricoordinated transition state
(II), as we suppose that protodeauration in (III) to regenerate
the catalyst is much faster. Concerning the regioselectivity, the
presence of a constant amount of the less thermodynamically
stable E isomer over the time, regardless if water is present or
not in the medium, seems to point out a cis-addition of the
alcohol across the triple bond to form E-(IV). Clearly, isomer-
ization of this E to the Z isomer occurs more rapidly over time
in the absence of water (see Figure 3, MeOH 1 equiv), in
contrast with the direct formation of the ketone in the presence
of excess H2O. Although more data should be compiled to check
the following hypothesis, we propose that double addition of
MeOH to form the corresponding dimethyl ketal (V) and
subsequent demethoxylation of this intermediate would explain
the formation of the Z from the E isomer. This path would be
overridden when H2O is present in the medium, since the
corresponding hemiketal of (V) leads to the ketone directly. The
gold catalyst could play a role in these last steps as well.
Comparison to Previous Methods. A brief study comparing
our methodology to those previously reported17,18 was carried
out. The results obtained are shown in Table 7.
Both methods require similar amounts of gold at room
temperature to complete the reaction (compare entries 1-4).
However, our method is more selective and preserves acid-labile
functional groups in a higher extent.
Conclusions
The high-efficient hydration of a wide range of alkynes to
the corresponding ketones by using exclusively gold(I) phos-
phine complexes has been afforded at room temperature.
AuSPhosNTf2, an air-weighable solid, has been synthesized in
high yields and proved to be the catalysis of choice in MeOH
as solvent. Ketals are intermediates in this transformation since
the attack of MeOH to the alkyne is faster compared to H2O.
The absence of acidic promoters allows the use of alkynes
containing acid-labile groups. This methodology is milder, more
(26) Baag, M. M.; Kar, A.; Argade, N. P. Tetrahedron 2003, 59, 6489.
J. Org. Chem. Vol. 74, No. 5, 2009 2073

Leyva and Corma
TABLE 7. Comparison between Gold-Catalyzed Hydration Methodologies at Room Temperature
a GC yield. b Complete conversion; several byproducts were found. c 0.5 M solution, DCM used as cosolvent (1:1). d The rest of material was
deprotected.
121.5, 117.5, 104.5 (2C), 55.1 (2C), 37.0, 36.6, 30.6, 30.5, 29.5
(2C), 26.7, 26.5 (2C), 26.3, 25.7 (2C). 31P NMR (δ, ppm): 38.6.
Anal. Calcd for C28H35AuF6NO6PS2: C, 37.89; H, 3.97; N, 1.58;
S, 7.22. Found: C, 37.61; H, 3.99; N, 1.57; S, 7.05. FAB+: [M+;
calcd for C28H35AuF6NO6PS2: 887]: found m/z 887 (M+), 607 (M
- N(SO2CF3)2). HRMS (ESI): [(M - N(SO2CF3)2)+ calcd for
C26H35AuO2P: 607.2040], found m/z 607.2033
Typical Hydration Procedure. AuSPhosNTf2 (17.8 mg, 1 mol
%) was placed in a round-bottomed flask under nitrogen. MeOH
(2 mL), 1-ethynyl-4-phenoxybenzene (361 µL, 2 mmol), and H2O
(144 µL, 4 equiv) were sequentially added, and the mixture was
magnetically stirred at rt for 24 h. Then, a sample was taken for
GC analysis, and the mixture was diluted in dichloromethane (20
mL) and extracted with water (10 mL). The water was extracted
with dichloromethane (10 mL), and the combined organic phases
were washed with brine (10 mL), dried over MgSO4, filtered, and
concentrated under reduced pressure to give 1-(4-phenoxyphenyl)e-
thanone (390 mg, 92%).
selective and operationally easier than those reported before and
should be of interest in the synthesis of elaborated molecules.
Experimental Section
Synthesis of AuSPhosNTf2: Following a reported but modified
procedure,27 tetrachloroauric acid trihydrate (394 mg, 1 mmol) was
dissolved in distilled water (1 mL) under nitrogen and the solution
cooled in ice. Then, 2,2′-thiodiethanol was slowly added over 45
min until disappearance of the color (ca. 300 µL). After that,
2-dicyclohexylphosphino-2′,6′-dimethoxybiphenyl (SPhos, 410 mg,
1 mmol) was added followed by 3 mL of ethanol. The mixture
was stirred at rt for 3 h. The solid was filtered off and washed with
methanol, redissolved in dry dichloromethane, and filtered again.
The solution was concentrated to dryness to obtain AuSPhosCl as
a white crystalline powder (570 mg, 89%). Following a related
procedure,22 AuSPhosCl (1.33 g, 2.07 mmol) was dissolved in dry
dichloromethane (50 mL) and AgNTf2 (820 mg, 2.07 mmol) was
added. The mixture was stirred at rt under nitrogen for 30 min and
filtered over Celite. The solution was concentrated to dryness to
obtain AuSPhosNTf2 as a white solid (1.83 g, quantitative). IR
(cm-1): 3444, 2947, 2933, 2862, 2848, 2837, 1599, 1489, 1475,
1390, 1213, 1192, 1130, 1113, 962. 1H NMR (δ, ppm; J, Hz): 7.62
(1H, t, J ) 7.5), 7.58 (1H, t, J ) 5.9), 7.51 (1H, dt, J ) 6.0, 1.5),
7.43 (1H, t, J ) 8.3), 7.21 (1H, ddd, J ) 7.3, 4.3, 1.1), 6.71 (2H,
d, J ) 8.5), 3.69 (6H, s), 2.16 (2H, mult), 2.00 (2H, mult),
1.86-1.60 (10H, mult), 1.47-1.18 (12H, mult). 13C NMR (δ, ppm):
157.0 (2C), 142.6, 133.3, 131.7, 131.4, 129.8, 127.2, 125.8, 125.0,
Acknowledgment. A.L. thanks CSIC for financial support
under the JAE-doctors program.
Supporting Information Available: General methods, ex-
perimental procedures including syntheses and characterization
of catalysts and substrates, kinetics and reaction procedures, and
1H, 13C, DEPT, and 31P NMR spectra of compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
(27) Nieto-Oberhuber, C.; Lopez, S.; Echavarren, A. M. J. Am. Chem. Soc.
2005, 127, 6178.
JO802558E
2074 J. Org. Chem. Vol. 74, No. 5, 2009