Aluminum triflate as a highly active and efficient nonprotic cocatalyst in the palladium-catalyzed methoxycarbonylation reaction

Article abstract of DOI:10.1002/anie.200702889

(Chemical Equation Presented) Lewis does it better: Aluminum triflate readily replaces Bronsted acid cocatalysts in the palladium-catalyzed methoxycarbonylation reaction of styrene and 1-pentene, producing catalysts that are stable and more active than those using traditional acids. Catalyst loadings of 0.02% allow conversions of up to 100% to be achieved within three hours with no loss of linear/branched ester selectivity.

Full text of DOI:10.1002/anie.200702889

Communications
DOI: 10.1002/anie.200702889
Methoxycarbonylation
Aluminum Triflate as a Highly Active and Efficient Nonprotic
Cocatalyst in the Palladium-Catalyzed Methoxycarbonylation
Reaction**
D. Bradley G. Williams,* Megan L. Shaw, Michael J. Green, and Cedric W. Holzapfel
Carboxylic acids and their derivatives, such as esters and
anhydrides, are important intermediates and products which
have pharmaceutical and agrochemical applications, amongst
others.[1] There has been substantial interest in the synthesis of
branched esters from vinyl aromatic compounds (for exam-
ple, in the production of anti-inflammatories, such as ibupro-
fen[2]) and the production of commodity chemicals from other
readily available olefins[3] using the catalytic methoxycarbon-
ylation reaction, which is also known as the hydroesterifica-
tion reaction. This reaction of alkenes generally requires a
catalyst, such as palladium, and a Brønsted acid cocatalyst
(usually p-toluenesulfonic acid (p-TsOH) or MeSO3H[4]) in
the presence of carbon monoxide and methanol [Eq. (1)].
We have reported the excellent Lewis acid catalyst
activity of Al(OTf)3 in the alcoholysis[11] and aminolysis[12]
of epoxides. With this in mind, and taking note of the
bimetallic systems mentioned above, the methoxycarbonyla-
tion reaction of styrene and 1-pentene was performed with
Al(OTf)3 as a cocatalyst in place of the more usual Brønsted
acid promoters. To our knowledge, it is the first case of a
strong, stable, recyclable[12] Lewis acid being used as the
cocatalyst in the methoxycarbonylation reaction, and we
present the observed rate enhancements over the benchmark
Brønsted acids and the formation of stable catalyst systems.
Initial experiments in high-pressure Parr reactors
(Table 1) were carried out with styrene (St) as the substrate
and 2% palladium catalyst (PPh3/Pd 4:1)[13] in the presence of
Al(OTf)3 (Pd/Al(OTf)3 1:2), and very high conversions were
observed. A circa threefold rate enhancement was observed
over the p-TsOH benchmark reaction after 15 minutes
(Table 1, entry 1). As anticipated, the reactions slowed as
the substrate was consumed, leading to a convergence of the
rates and a less obvious rate enhancement as the reactions
proceeded (Table 1, entries 1–3). Interestingly, we were able
to reduce the amount of Al(OTf)3 present relative to
palladium to substoichiometric ratios (Pd/Al(OTf)3 as low
as 1:0.25; Table 1, entries 4–7) while maintaining high activ-
ity; no loss of activity was observed at a Pd/Al(OTf)3 ratio of
1:0.75. By comparison, palladium/Brønsted acid ratios of
1:100 are not uncommon for these types of reaction.[8] Under
the reaction conditions used, no reaction was observed at all
in the absence of the promoter. Importantly, palladium black
formation was not observed in any of the runs, indicating
A detailed study on the effect of various acids on the rate
of the methoxycarbonylation reaction of ethene indicated p-
TsOH to be the most active,[5] closely followed by triflic acid
(HOTf, OTf = triflate). When propene was the alkene, the
acids H2SO4, HClO4, and HOTf all showed similar activities.
Metal halides of nickel, copper, iron, or cobalt have been used
as the promoter, but, although these produce reasonable
selectivities, fairly high catalyst loadings (4% Pd) are
required,[6] and overall catalyst activity is not as high as in
the case of other promoters.[7]
Recently, salicylborates have been shown to be effective
promoters in this reaction,[8] and also allow an unusually high
degree of the linear product to be generated.[9] Acidic resins
with SO3H groups have also been used.[10] In all cases
mentioned, the promoter is present in a substantial excess
over the palladium catalyst.
Table 1: Palladium-catalyzed methoxycarbonylation ofstyrene with Al-
(OTf)3 as promoter.
Entry
Pd/L/LA/St[a]
% Conv[b]
l/b[c]
t [min]
1
2
3
4
5
6
7
1:4:2:50
1:4:2:50
1:4:2:50
1:4:1:50
1:4:0.75:50
1:4:0.5:50
1:4:0.25:50
26 (9)[d]
81 (52)[d]
93 (81)[d]
95
94
81
1.6:1
1.7:1
1.5:1
2.4:1
2.1:1
2.0:1
2.2:1
15
30
45
60
60
60
60
[*] Prof. Dr. D. B. G. Williams, M. L. Shaw, Prof. Dr. C. W. Holzapfel
Department ofChemistry, University ofJohannesburg
P.O. Box 524, Auckland Park, 2006 (South Africa)
Fax: (+2711)559-2819
77
[a] Reaction conditions: Pd(OAc)2 (0.045 mmol), MeOH/St 36:1 (v/v,
total volume 12.5 mL), 35 atm CO, 80Æ28C internal temperature (118Æ
28C oil bath). (L=PPh3; LA=Al(OTf)3; St=styrene; 1 atmꢀ
101.325 kPa). [b] Conversion into ester products (no detectable by-
product formation) determined by 1H and 13C NMR spectroscopy. [c] l/b
is ratio oflinear to branched isomers determined by 1H and 13C NMR
spectroscopy. [d] Values in parenthesis relate to analogous reactions
using the Brønsted acid p-TsOH.
E-mail: bwilliams@uj.ac.za
Dr. M. J. Green
Sasol Technology Research and Development
1 Klasie Havenga Rd
Sasolburg, 1947 (South Africa)
[**] The authors gratefully acknowledge funding from the University of
Johannesburg, Sasol Ltd, THRIP, and the NRF.
560
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 560 –563

Angewandte
Chemie
catalyst stability even with the very low levels of Al(OTf)3. It
has been postulated that Brønsted acids convert the palla-
dium catalyst into a cationic species, the conjugate base of
which acts as a noncoordinating counterion.[14] It is likely that
the Al(OTf)3, being a hard Lewis acid and consequently
oxophilic, abstracts the acetate ion from the palladium
complex, rendering that complex cationic and providing a
large noncoordinating aluminate counterion [(AcO)Al-
(OTf)3]À. (We do not know at present whether the triflate
ions are inner sphere or outer sphere on the aluminum ion.
For lanthanide triflates, the triflate ions were found to be
outer sphere.[15] However, it is likely that all of the triflate ions
remain in the inner sphere in our system, in light of a crystal
structure of a mixed aluminum–rhenium complex, generated
from Al(OTf)3, in which the three triflate ions remain inner
sphere on the octahedral aluminum ion.)[16] The rates of the
reactions slowly decreased (Table 1, entries 4–7) with
decreasing amounts of cocatalyst, and we established that a
Pd/Al ratio of 1:2gave optimal catalyst stability and activity,
especially at lower catalyst loadings, for a Pd/PPh3 ratio of 1:4
and with styrene as substrate. To our knowledge, similar rates
of reaction (for comparison, see percentage conversion in
parentheses, Table 1, for p-TsOH) have not previously been
observed at such low acid loadings: usually, much more acid
relative to palladium is required to achieve comparable
activities and catalyst stability.
analysis, and only the anticipated linear and branched ester
products were formed. Although there was some sensitivity in
the selectivity of the reaction (ratio of linear/branched
isomers) to the conditions, this proved to be minor. Catalyst
loadings as low as 0.03% (palladium/styrene 1:3000) could be
achieved (Table 2, entries 9–10); such reactions required a Pd/
L/Al ratio of 1:8:4 to afford active catalysts that did not result
in palladium black formation during the 16-hour reaction
periods (palladium black was only observed for the reaction
noted in Table 2, entry 9). The prolonged reaction times
required for essentially quantitative conversion resulted in
the formation of small amounts (< 3%) of an insoluble
polymeric material in addition to the anticipated high yields
of ester. The Lewis acid could be successfully recycled by
simple aqueous extraction of the acid followed by removal of
water and drying of the residue by heating under vacuum, and
using the residue in a new reaction instead of fresh Al(OTf)3.
The results from a reaction using recycled Al(OTf)3 (Table 2,
entry 5b) was almost indistinguishable from the first cycle of
the Lewis acid (Table 2, entry 5a), indicating that the
promoter was recovered intact.
Ligand consumption by alkylation to form the methyl-
triphenylphosphonium cation is problematic with protic
acids,[17] necessitating the use of a large excess of ligand to
render the catalyst stable and an excess of the acid as it is
consumed in the process. One exception is the salicylborates,
which catalyze this process only slowly but which themselves
are deactivated by esterification of the salicylic moiety.[8]
Even in this instance, the acid is used in large excess. The
extent of quaternization of PPh3 by Al(OTf)3 and some
Brønsted acids was investigated in situ by 31P NMR spectros-
copy. Table 3 shows that PPh3 is consumed very slowly under
conditions that resemble those of the reaction, and only 7%
alkylation is observed after 20 hours in the presence of one
equivalent of the Lewis acid. Even with five equivalents of
Al(OTf)3, only 27% alkylation was detected (under our
reaction conditions the ligand is in excess of the Lewis acid).
This result is another demonstrable benefit over the tradi-
tional Brønsted acids, which catalyzed this reaction to a
significantly greater extent under identical conditions. Impor-
tantly, ligand recovery from a typical reaction (Table 2,
entry 8) employing Al(OTf)3 demonstrated ligand alkylation
of less than 1%.
To test the limits of the system, the catalyst and cocatalyst
loadings were systematically reduced, as was the MeOH/
styrene ratio (Table 2). The latter move was especially
important at low catalyst loadings (0.08% Pd: palladium/
styrene 1:1200 and lower) to prevent palladium black
formation. We established that the Pd/Al(OTf)3 combination
provided a stable active catalyst for this reaction, affording
high yields of ester product (> 93% in all cases; the remaining
7% was unreacted starting material). In all instances, no by-
products were observed by NMR spectroscopy or GC-FID
Table 2: Palladium-catalyzed methoxycarbonylation ofstyrene at low
catalyst loadings with Al(OTf)3 as promoter.
Entry Pd/L/LA/St[a] MeOH/Styrene %Conv[b]
l/b[c]
t [h]
1
2
3
4
5a
5b[d]
6
7
8
1:4:2:100
1:4:2:150
1:4:2:200
1:4:2:200
1:4:2:400
1:4:2:400
1:4:2:600
1:4:2:800
1:4:2:1200
1:8:2:3000
1:8:4:3000
18
12
12
95
93
93
95
96
94
>98
93
1.9:1
2.1:1
1.9:1
2.0:1
2.1:1
2.1:1
2.2:1
2.8:1
2.8:1
1
1
1
1
1
1
2
2
4
4.75
Following these successful outcomes, we directly com-
pared the Al(OTf)3 cocatalyst, using a standard set of reaction
conditions, with the more commonly used high-performance
Brønsted acids[5] such as p-TsOH and HOTf, and were
gratified to note that Al(OTf)3 substantially outperformed
the Brønsted acids in all cases in time-limited (2h) reactions
4.75
4.75
4.75
4.75
2.4
95
9
10
1.5
1.5
97+Polymer[e]
3.0:1 16
>98+Polymer 2.7:1 16
Table 3: Acid-catalyzed quaternization ofPPh 3 in MeOH.[a]
[a] Reaction conditions: Pd(OAc)2 (0.045 mmol), 35 atm CO, 80Æ28C
internal temperature (118Æ28C oil bath), total volume 12.5 mL. (L=
PPh3; LA=Al(OTf)3; St=styrene). [b] Conversion into ester products
(no detectable by-product formation except as cited for entries 9 and 10)
Acid
L/A 1:1 [%]
L/A 1:2 [%]
L/A 1:5 [%]
Al(OTf)3
p-TsOH
TfOH
7
22
51
61
75
27
85
91
95
28
46
58
1
as determined by H and 13C NMR analysis. [c] l/b is ratio oflinear to
branched isomers determined by 1H and 13C NMR spectroscopy.
[d] Lewis acid recycle reaction: aqueous extraction ofAl(OTf) 3, removal
ofwater (150 8C, 0.1 mm Hg), residue added to fresh palladium catalyst,
ligand, solvent, and substrate. [e] palladium black formation observed.
MsOH
[a] Values in the table are percentage alkylation ofPPh to [MePPh4]+.
3
Reaction conditions: PPh3 (0.0475 mm), MeOH (1.0 mL), 708C, 20 h
(L=PPh3; A=Lewis or Brønsted acid; MsOH=CH3SO3H).
Angew. Chem. Int. Ed. 2008, 47, 560 –563
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
561

Communications
Table 5: Palladium-catalyzed methoxycarbonylation of1-pentene using
various acid promoters.
(Table 4). At palladium catalyst loadings of 0.08% (palla-
dium/styrene 1:1200) and 0.0625% (palladium/styrene
1:1600), the reactions in which Al(OTf)3 was used as
promoter were a factor of 1.6–1.9 faster than the correspond-
ing Brønsted acid-promoted reactions.
Entry Pd/L/HA[a] Promoter T [8C] TOF[b] (”102) % Conv[c] l/b[d]
[e]
[e]
1
2
3
4
5
6
7
8
1:9:2
1:9:4
1:9:6
1:9:2
HCl
HCl
HCl
115
115
115
3.2
6.0
3.0
3.6
5.7
8.5
7.7
9.5
–
–
42
–
–
–
62
–
2.8:1
–
–
–
3.0:1
–
3.2:1
CH3SO3H 105
CH3SO3H 105
CH3SO3H 105
CH3SO3H 105
CH3SO3H 105
HOTf100
HOTf100
HOTf100
HBF4
HBF4
HBF4
Al(OTf)3
Al(OTf)3
Al(OTf)3
Al(OTf)3
Table 4: Palladium-catalyzed methoxycarbonylation ofstyrene with var-
ious acid promoters.
1:9:4
1:9:14
1:9:18
1:20:30
1:9:2
1:9:14
1:9:18
1:9:2
1:9:6
1:9:8
1:9:2
1:9:4
Entry Pd/L/LA/St[a] Promoter % Conv[b] l/b[c]
TOF[d] (”102)
70
1
2
3
4
5
6
1:4:2:1200
1:4:2:1200
1:4:2:1200
1:4:2:1600
1:4:2:1600
1:4:2:1600
Al(OTf)3
p-TsOH
HOTf42
Al(OTf)3
p-TsOH
HOTf22
62
39
2.8:1 3.7
9
9.3
10.4
8.8
–
74
–
–
2.6:1 2.3[e]
2.1:1 2.5
10
11
12
13
14
15
16
17
18
3.1:1
–
31
18
2.7:1 2.5
2.5:1 1.4
2.3:1 1.3
100
4.5
11.2
9.9
6.2
9.4
–
83
–
–
100
100
100
100
100
100
3.4:1
-
[a] Reaction conditions: Pd(OAc)2 (0.045 mmol), 4.75:1 MeOH/St (v/v,
total volume 12.5 mL), 35 atm CO, 80Æ28C internal temperature (118Æ
28C oil bath), 2 h (L=PPh3; LA=Al(OTf)3; St=styrene). [b] Conversion
into ester products (no detectable by-product formation) as determined
by 1H and 13C NMR analysis. [c] l/b is ratio oflinear to branched isomers
determined by 1H and 13C NMR spectroscopy. [d] TOF=mol ester
produced/mol Pd.h, based on total run, average ofthree runs. [e] TOF =
2.0”102 for 0.4% Pd catalyst, Pd/PPh3 1:4, Pd/p-TsOH 1:10, 35 atm
CO.[13]
–
-
68
100
100
3.4:1
3.5:1
3.5:1
1:9:8
1:18:18
14.4
15.8
[a] Reaction conditions: 0.1 mmol Pd(OAc)2, Pd/1-pentene 1:4750,
MeOH/1-pentene 1:1 (v/v, total volume 100 mL), 50 atm CO, 3 h, (L=
PPh3). [b] TOF: mol CO absorbed/mol Pd.h, based on the linear part of
the CO uptake curve (to ca. 80% conversion), average ofthree runs.
[c] Conversion into ester products (excluding pentene isomers) as
1
determined by H NMR and GC-FID analysis (no detectable by-product
formation). [d] l/b is ratio of linear to branched isomers determined by
1H and 13C NMR spectroscopy. [e] Analyses performed only for the best
reaction in a given set ofreactions using a particular Brønsted/Lewis
acid.
Encouraged, we compared various Brønsted acids with
Al(OTf)3 using 1-pentene as a representative nonaromatic
substrate (Table 5). Somewhat modified conditions were
applied at lower catalyst loadings than were used with
styrene, in line with a previous observation on hydroformy-
lation reactions that 1-alkenes are approximately three times
more reactive than styrene,[18] as was also found in this case. In
these reactions, it was found that a Pd/L ratio of greater than
1:5 was imperative for catalyst stability (upon which the
observed rate of reaction relies); below this level of ligand,
significant formation of palladium black was observed. As
demonstrated above, this is unlikely to be related to ligand
alkylation. Furthermore, a minimum palladium/acid ratio of
1:2was required, even with Al(OTf) 3, for acceptable catalyst
activity to be obtained, which is possibly a consequence of the
higher ligand concentrations present.
It can be seen from Table 5 that the optimum palladium/
acid ratio, as determined from the rates of the reactions, is
(sometimes critically) acid-dependent. These optimum ratios
range from 1:4 in the case of HCl to 1:30 or more in the case of
CH3SO3H. With Al(OTf)3, the optimum Pd/Al conversion
ratio is 1:8, although some further rate enhancement was
observed with ratios up to 1:18. Again, no reaction was
observed in the absence of cocatalyst, as was the case with
styrene. Of the Brønsted acids tested, anhydrous HBF4
proved to be superior; however, Al(OTf)3 outperformed
that cocatalyst, again from a rate perspective, by a factor of
1.4, with no loss of the selectivity, and by a factor of 1.9 over
the more commonly used CH3SO3H. The reactions using
Al(OTf)3 showed a linear uptake of CO to a conversion of
approximately 80%; thereafter the rates of the reactions
slowed. This reaction is apparently unaffected by the rapid
double-bond isomerization of 1-pentene. This isomerization
reached equilibrium within one hour under the conditions
employed: the CO uptake curves showed no deviation from
linearity at all during the course of the reaction for the cited
portion of the curve (pertaining to approximately 80%
conversion), which would otherwise be expected to be the
case if the reaction was in fact sensitive to this process.
Further rate enhancements (not presented in Table 5)
were obtained by changing the MeOH/1-pentene ratio from
1:1 to 2:1 v/v, which resulted in a rate enhancement of
approximately 1.25 and was accompanied by a slight increase
in selectivity for the linear product (l/b 4:1).
It is unlikely that triflic acid, generated by methanolysis of
Al(OTf)3, is the active species. Recent work on the Al(OTf)3-
catalyzed cyclization of alkenols demonstrated activity of this
catalyst in the presence of water and 2,6-lutidine as base in
CH3CN solution, conditions under which HOTf was shown to
be inactive.[19] Furthermore, our previous work using this
Lewis acid in alcohol solvents has also shown that triflic acid is
not the active catalyst,[11] and the ligand alkylation study
reported herein also contradicts this notion. Nonetheless,
[19]
À
Al(OTf)3 does polarize the O H bond of alcohols, which
+
À
possibly provides the source of the {Pd H} species generally
accepted to be required to initiate and sustain the catalytic
cycle involved with faster catalysts[20] (PdII has also been
À
shown to react directly with MeOH to form {Pd H}
species[21]), and at the same time this Lewis acid presumably
also functions as a noncoordinating aluminate counterion, as
mentioned above.
The origin of the heightened activity of the catalyst system
most probably lies in its acceleration of the rate-limiting step
562
www.angewandte.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2008, 47, 560 –563

Angewandte
Chemie
of the reaction, which has been shown to be the alcoholysis of
the acetylpalladium species.[20] In the present case, the
Al(OTf)3 may either activate this species towards migratory
nucleophilic attack by palladium-bound MeOÀ/MeOH
(Lewis acid coordination to the carbonyl oxygen) and/or
possibly provide a source of MeOÀ in the form of the ion
[MeOAl(OTf)3]À for direct nucleophilic attack. The reaction
had a fractional order dependence on the Lewis acid
concentration: a rate enhancement of 1.27 was found over
the range Pd/LA 1:1 to 1:8 (relative rates for 1:1 = 1, 1:2 =
1.14, 1:4 = 1.25, and 1:8 = 1.27). The most notable feature of
this part of the study was a decrease in the induction period
from approximately 30 minutes for the lower Al(OTf)3 ratios
to only about 10 minutes for the higher ratios.
other Lewis acids to this reaction, and the investigation will
also include the industrially important[3] production of methyl
methacrylate (MMA) from ethene by the methoxycarbonyl-
ation product methyl propanoate, producing MMA upon
condensation with formaldehyde or its dimethyl acetal.[22]
Received: June 29, 2007
Revised: October 4, 2007
Published online: November 28, 2007
Keywords: aluminum · carbonylation · catalysis · Lewis acid ·
.
palladium
[1] K. Nozaki, I. Ojima in Catalytic Asymmetric Synthesis, 2nd ed.,
Wiley, New York, 2000, p. 429.
The unusual observation of reaction rate linearity (ap-
parent zero-order kinetics in alkene for a significant portion
of the reaction) for both styrene and 1-pentene led us to
investigate this phenomenon more fully. The observation of
rate linearity, together with the fact that, in general, no
palladium black formation was observed under the reaction
conditions given which led to successful reactions, even at the
end of the reaction after depressurization of the reactor, also
decidedly points to high catalyst stability and a stable,
constant-composition, catalytically active palladium species.
We established that the rates of the reactions were indeed
apparently independent of the alkene concentration to
approximately 80% conversion (for the higher initial con-
centrations, tapering off to about 70% for the lower initial
concentrations) after which the CO uptake curve gradually
flattened off until complete conversion was obtained (regres-
sion analysis for the linear section of the CO uptake curve
provided a fit of better than R2 = 0.996 in all cases, after an
induction period of approximately 20–30 minutes) for initial
styrene concentrations in the window 0.77–2.88m. Further
studies on CO dependence showed a fractional order depend-
ence. Relative rates were determined for Pco (defined as 1 at
25 atm): Pco 35 atm = 1.07, Pco 45 atm = 1.20, Pco 55 atm =
1.25. Additionally, as shown in Table 1 (entries 4–7) and
Table 5 (entries 15–18), rate enhancements were also
observed with increasing levels of the Al(OTf)3 (as was also
observed for the Brønsted acids, Table 5), the results of which
point to a fractional order dependence on Lewis acid
concentration.
[2] G. Papadogianakis, G. Verspui, L. Maat, L. R. A. Sheldon, Catal.
Lett. 1997, 47, 43.
[3] a) W. Clegg, G. R. Eastham, M. J. Elsegood, R. P. Tooze, W. L.
Wang, K. Whiston, Chem. Commun. 1999, 1877; b) W. Clegg,
G. R. Eastham, M. J. Elsegood, B. T. Heaton, J. A. Iggo, R. P.
Tooze, R. Whyman, S. Zacchini, Organometallics 2002, 21, 1832.
[4] G. Kiss, Chem. Rev. 2001, 101, 3435.
[5] E. Drent (Shell), EP 0,106,379, 1988.
[6] B. S. Wan, S. J. Liao, Y. Xu, D. R. Yu, J. Mol. Catal. A 1998, 136,
263.
[7] a) T. Fuchikami, K. Ohishi, I. Ojima, J. Org. Chem. 1983, 48,
3803; b) D. H. Nguyen, Y. Coppel, M. Urrutigoꢀty, P. J. Kalck, J.
Organomet. Chem. 2005, 690, 2947.
[8] A. Ferreira, R. Crous, L. Bennie, A. M. M. Meij, K. Blann,
B. C. B. Bezuidenhout, D. A. Young, M. J. Green, A. Roodt,
Angew. Chem. 2007, 119, 2 32 3A; ngew. Chem. Int. Ed. 2007, 46,
2273.
[9] T. O. Vieira, M. J. Green, H. Alper, Org. Lett. 2006, 8, 6143.
[10] H. Ooka, T. Inoue, S. Itsuno, M. Tanaka, Chem. Commun. 2005,
1173.
[11] D. B. G. Williams, M. Lawton, Org. Biomol. Chem. 2005, 3, 3269.
[12] D. B. G. Williams, M. Lawton, Tetrahedron Lett. 2006, 47, 6557.
[13] A. Seayad, A. A. Kelkar, L. Toniolo, R. V. Chaudhari, J. Mol.
Catal. A 2000, 151, 47.
[14] a) G. P. C. Dekker, C. J. Elsevier, K. Vrieze, P. W. N. M. van
Leeuwen, C. F. Roobeek, J. Organomet. Chem. 1992, 430, 357;
b) B. A. Markies, D. Kruis, M. H. P. Rietveld, K. A. N. Verkerk,
J. Boersma, H. Kooijman, M. Lakin, A. L. Spek, G. van Koten, J.
Am. Chem. Soc. 1995, 117, 5263.
[15] a) J. M. Harrowfield, D. L. Kepert, J. M. Patrick, A. H. White,
Aust. J. Chem. 1983, 36, 483; b) A. Chatterjee, E. N. Maslen, K. J.
Watson, Acta Crystallogr. Sect. B 1988, 44, 381.
[16] A. A. Danopoulos, G. Wilkinson, T. K. N. Sweet, M. B. Hurst-
house, J. Chem. Soc. Dalton Trans. 1996, 2995.
[17] E. Drent, M. Hasselaar (Shell), WO 97/03943, 1997.
[18] A. Van Rooy, E. N. Orij, P. J. C. Kamer, P. W. N. M. van Leeu-
wen, Organometallics 1995, 14, 34.
[19] L. Coulombel, M, Rajzmann, J. M. Pons, S. Olivero, E. Duꢁach,
Chem. Eur. J. 2006, 12, 6356.
[20] P. W. N. M. van Leeuwen, M. A. Zuideveld, B. H. G. Swannen-
huis, Z. Freixa, P. C. J. Kamer, K. Goublitz, F. Fraanje, M. Lutz,
A. L. Spek, J. Am. Chem. Soc. 2003, 125, 5523, and references
therein.
[21] a) P. W. N. M. van Leewen Homogeneous Catalysis, Kluwer,
Dordrecht, 2004, p. 259; b) R. A. M. Roberston, D. J. Cole-
Hamilton, Coord. Chem. Rev. 2002, 225, 67; c) C. Bianchini, A.
Meli, W. Oberhauser, S. Parisel, A. V. Gusev, A. M. Kalꢂsin, N. V.
Vologdin, F. M. Dolgushin, J. Mol. Catal. A 2004, 224, 35.
[22] J. P. Spivey, M. R. Gogate, J. R. Zoeller, R. D. Colberg, Ind. Eng.
Chem. Res. 1997, 36, 4600.
We have presented the unprecedented application of a
strong, hydrolytically stable[12] Lewis acid such as Al(OTf)3 in
the methoxycarbonylation reaction of styrene and 1-pentene,
and that this material may readily take the place of the
commonly used Brønsted acids to form the anticipated ester
products in high yields. This aluminum-based promoter
significantly improves the activity of the catalyst, leading to
rates 1.4–1.9 times higher than those observed for even the
best benchmark Brønsted acids under comparable conditions.
Importantly, Al(OTf)3 is a stable Lewis acid, which we have
previously shown to be easily recoverable and reusable
without loss of activity,[12] and which we have also demon-
strated herein. Stable palladium catalysts are produced in situ
and the reaction conditions may be readily manipulated so
that no palladium black is formed even at low catalyst
loadings. We are currently investigating the application of
Angew. Chem. Int. Ed. 2008, 47, 560 –563
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
563