Pd-catalyzed aldehyde to ester conversion: A hydrogen transfer approach

Article abstract of DOI:10.1021/ol303298g

Aliphatic and aromatic aldehydes are successfully converted into their corresponding esters using Pd(OAc)₂ and XPhos. This approach utilizes a hydrogen transfer protocol: concomitant reduction of acetone to isopropanol provides an inexpensive and sustainable approach that mitigates the need for other oxidants.

Full text of DOI:10.1021/ol303298g

ORGANIC
LETTERS
2
013
Vol. 15, No. 3
00–503
Pd-Catalyzed Aldehyde to Ester
Conversion: A Hydrogen Transfer Approach
5
†
Brittany A. Tschaen, Jason R. Schmink, and Gary A. Molander*
Roy and Diana A. Vagelos Laboratories and Penn/Merck Laboratory for High
Throughput Experimentation, Department of Chemistry, University of Pennsylvania,
Philadelphia, Pennsylvania 19104-6323, United States
gmolandr@sas.upenn.edu
Received November 30, 2012
ABSTRACT
2
Aliphatic and aromatic aldehydes are successfully converted into their corresponding esters using Pd(OAc) and XPhos. This approach utilizes a
hydrogen transfer protocol: concomitant reduction of acetone to isopropanol provides an inexpensive and sustainable approach that mitigates
the need for other oxidants.
6
7
Historically, the most direct synthetic routes to esters
either couple an activated carboxylic acid derivative with
the appropriate alcohol or employ an equilibrium me-
diated esterification/transesterification protocol. These
methods require stoichiometric use of toxic coupling
2
reagents (e.g., DCC, HOBt), with concomitant formation
metric oxidant such as MnO , azobenzene, a substituted
2
8
9
quinone, or an Fe/O system, with concomitant reduc-
2
1
0
tion elsewhere in the molecule; an electrochemical
oxidation has also been used in conjuction with NHC
1
activation, (3) use of transition metals such as Ir, Rh,
and Ru in the presence of an external or internal oxidant.
Our focus led to exploring the use of Pd catalysis to convert
aldehydes directly to the corresponding esters in the pres-
ence of the appropriate alcohol, simply using acetone as
the terminal oxidant.
1
1
12
13
14
of byproducts that can be difficult to remove during
isolation. To address these limitations, there has been a
recent emergence of investigations into direct oxidative
routes to esters, allowing practitioners to select synthetic
precursors in alternative oxidation states. The most widely
investigated is the direct conversion of aldehydes to esters
in the presence of an alcohol, and three representative
approaches include (1) oxidation of the aldehyde in the
presence of an alcohol employing stoichiometric oxidants
Recent investigations have shown that Pd is a capable
catalyst to effect the oxidative esterification of aldehydes,
although these studies have demonstrated the requirement
for stoichiometric benzyl chloride to close the Pd(II)ꢀPd(0)
1
5
catalytic cycle. Under similar conditions using Pd/NHC
complexes and air as the terminal oxidant, Cheng et al.
3
4
suchasV O •H O , oxone, orpyridinium hydrobromide
2
5
2
2
5
perbromide and H O; (2) use of NHC catalysts to activate
2
aldehydes in situ, which in turn undergo esterification in
the presence of the appropriate alcohol and stoichio-
(
8) Sarkar, S.; Grimme, S.; Studer, A. J. Am. Chem. Soc. 2010, 132,
190.
9) Reddy, R. S.; Rosa, J. N.; Veiros, L. F.; Caddick, S.; Gois,
1
(
P. M. P. Org. Biomol. Chem. 2011, 9, 3126.
†
Present address: Department of Chemistry, Bryn Mawr College, Bryn
(10) Sohn, S. S.; Bode, J. W. Org. Lett. 2005, 7, 3873.
(11) Finney, E. E.; Ogawa, K. A.; Boydston, A. J. J. Am. Chem. Soc.
2012, 134, 12374.
(12) (a) Kiyooka, S.; Ueno, M.; Ishii, E. Tetrahedron Lett. 2005, 46,
4639. (b) Kiyooka, S.; Wada, Y.; Ueno, M.; Yokoyama, T.; Yokoyama,
R. Tetrahedron 2007, 63, 12695.
Mawr, PA 19010.
(
1) Otera, J.; Nishikido, J. Esterification. Methods, Reactions, and
Applications, 2nd ed.; Wiley-VCH: Weinheim, 2010.
2) Neises, B.; Steglich, W. Angew. Chem., Int. Ed. Engl. 1978, 17,
22.
(
5
(
(
3) Gopinath, R.; Patel, B. K. Org. Lett. 2000, 2, 577.
4) Travis, B. R.; Sivakumar, M.; Hollist, G. O.; Borhan, B. Org.
(13) Grigg, R.; Mitchell, T.; Sutthivaiyakit, S. Tetrahedron 1981, 37,
4313.
Lett. 2003, 5, 1031.
(14) Murahashi, S.; Naota, T.; Ito, K.; Maeda, Y.; Taki, H. J. Org.
Chem. 1987, 52, 4319.
(15) (a) Heropoulos, G. A.; Villalonga-Barber, C. Tetrahedron Lett.
2011, 52, 5319. (b) Liu, C.; Tang, S.; Zheng, L.; Liu, D.; Lei, A. Angew.
Chem., Int. Ed. 2012, 51, 5662.
(
(
(
5) Sayama, S.; Onami, T. Synlett 2004, 15, 2739.
6) Maki, B. E.; Scheidt, K. A. Org. Lett. 2008, 10, 4331.
7) Noonan, C.; Baragwanath, L.; Connon, S. J. Tetrahedron Lett.
2008, 49, 4003.
1
0.1021/ol303298g r 2013 American Chemical Society
Published on Web 01/15/2013

Figure 2. Reaction monitoring: effect of various amounts of
base (0.2 equiv of K CO proves ideal). Yields determined by
2
3
GC using biphenyl as an internal standard. Average of two runs
agreeing to within 5%.
Figure 1. Catalyst loading and substrate concentration screen.
Results interpreted using HPLC analysis with biphenyl as the
internal standard.
and 5% XPhos in a 50:50 ethanol/acetone mixture at 50 °C
overnight (Figure 1, entry b). This yielded a crude reaction
mixture that was 73% ester 2, 14% alcohol 3, and 13%
unreacted starting material 1. Increasing the Pd loading led
to increased reactivity (5% unreacted starting material)
but provided lower selectivity (18% conversion to the
alcohol; Figure 1, entry a), while lowering the Pd loading
saw sluggish conversion (Figure 1, entries c and d). Thus, we
selected a 2.5% Pd loading and sought further optimization.
Our next step was to undertake an “additive screen”,
recently demonstrated that aromatic aldehydes could be
1
6
oxidatively esterified with phenols. Other Pd catalyzed
aldehyde to ester conversions include 1,2 additions of
1
7
18
nucleophilic siloxanes or arylboronic acids followed
by oxidative esterification.
In various cross-coupling projects, our group has ob-
served that aldehydes can be problematic substrates, pro-
viding lower-than-anticipated yields in otherwise high
yielding protocols, especially when run in methanol or
ethanol while using XPhos (2-dicyclohexylphosphino-
2
0
using High-Throughput Experimentation techniques, in
which a screen comprised of 95 various additives at 20 mol %
loading was compared to our control. Screens such as this
are often quite insightful, as they illustrate conditions that
both improve and hinder the desired outcome. To accen-
tuate any impacts, this screen was performed at rt, and the
reaction was quenched after only 2 h. To our delight,
catalytic amounts of inorganic bases (K CO , KHCO ,
0
0 0
2
,4 ,6 -triisopropylbiphenyl) as a catalyst. Detailed analy-
sis of a model reaction using benzaldehyde provided evi-
dence of disproportionation of the aldehyde, and we
observed equimolar amounts of the benzyl alcohol and
ethyl/methyl benzoate derivatives. Although generally pre-
sent in <5% of the overall crude reaction mixture, we
hypothesized that the Pd/XPhos catalyst system might play
some role in this byproduct formation.
2
3
3
K PO , etc.) showed a significant positive impact on the
3
4
outcome of the reaction. Under these conditions, potas-
sium carbonate provided complete conversion of the alde-
hyde and formation of the desired ester in 89% assay yield,
with the remainder as the reduced alcohol. This additive
screen also indicated that oxidants such as oxone greatly
Seeking to maximize the conversion to the ester, we
reasoned that the reduction of the aldehyde might
be limited by adding a competitive “hydrogen acceptor.”
Simply running the reaction with acetone as a cosolvent
did just this. By preparing a reaction under strictly anaer-
interfered with the reactivity, while V(O)(acac) and iodo-
2
sobenzene diacetate had little impact either way. This screen
gave us the first indications that the aldehyde/hemiacetal
equilibrium might be important in this reaction mechanism.
The general procedure involved preparing a solution of
1
9
obic conditions (glovebox, N2), a 0.2 M solution of
-phenylpropanal was allowed to react with 2.5% Pd(OAc)2
3
the Pd(OAc) and XPhos in acetone before then adding it
2
(
16) Zhang, M.; Zhang, S.; Zhang, G.; Chen, F.; Cheng, J. Tetra-
to the aldehyde solution in ethanol. As it aged, a color
change of the palladium/ligand solution was observed.
We thought this color change might be an indication that
the catalyst needed to preform prior to its exposure to
hedron Lett. 2011, 52, 2480.
(
17) Wolf, C.; Lerebours, R. J. Am. Chem. Soc. 2006, 128, 13052.
(18) Qin, C.; Wu, H.; Chen, J.; Liu, M.; Cheng, J.; Su, W.; Ding, J.
Org. Lett. 2008, 10, 1537.
19) The reactions were initially prepared in a glovebox, and the
(
solvents were degassed to eliminate the possibility that oxygen played
any role in the catalytic cycle. Reactions can be performed in air, but
product yields are dramatically lowered.
(20) Dreher, S. D.; Dormer, P. G.; Sandrock, D. L.; Molander, G. A.
J. Am. Chem. Soc. 2008, 130, 9257.
Org. Lett., Vol. 15, No. 3, 2013
501

Table 1. Alcohol Scope
Table 2. Aromatic Aldehyde Scope with Methanol, Ethanol,
and Isopropanol
a
Isolated yields after column chromatography.
22
1
:1 ester/alcohol ratio (disproportionation) was observed.
2
3
24
Use of either Cu(OAc)2 or iron/benzoquinone as a
co-oxidant resulted in disproportionation of the starting
aldehyde. These results suggest that acetone is likely acting
as a formal molecular H-acceptor in the catalytic cycle.
Alternative H-acceptors were screened, including 2,3-
butanedione and hexafluoroacetone. Both proved to be less
effective, and the low price and high reactivity combination
of acetone could not be matched.
With our optimal reaction conditions in hand, a range of
aliphatic alcohols was screened. When coupled with the
aliphatic aldehyde, primary alcohols provided the desired
product in moderate to good yield in 2 h at rt (Table 1, entries
a
b
Isolated yields after column chromatography. 0.5% Pd(OAc)
2
,
c
1
% XPhos, 24 h, 7.0 mmol scale. 50 °C.
the aldehyde, base, and alcohol to maximize the desired
reactivity. Indeed, without this preformation process, re-
action yields were ∼30% lower.
Multitime point sampling of the reaction indicated that
both the initial rate and overall conversion to the ester were
dependent on the loading of the K CO base (Figure 2).
2
3
With a full equivalent, the reaction shows no induction
period, but the desired reactivity quickly stalls, equilibrat-
ing at 72% conversion. However, employing 0.2 equiv of
K CO leads to maximum conversion to the ester after a
short induction period, and the reaction is complete after
0 min. When base is excluded, the reaction is very slow,
though eventually 34% conversion to the ester is observed
after 25 h. This may provide further evidence that the
aldehydeꢀhemiacetal equilibrium is important to the cata-
lytic cycle, or may result from the necessity to deprotonate
the hemiacetal, which is required to generate the alkoxy-
palladium intermediate required for β-hydride elimina-
1a, 1b). The reaction scales well, affording a higher isolated
yield while simultaneously allowing a 5-fold reduction of the
Pd/XPhos loading at the 7 mmol scale (entry 1b). Sterically
more demanding secondary alcohols (Table 1, entries 1c, 1e)
provided the desired ester, although requiring extended
reaction times at 50 °C. tert-Butanol did not engage in the
oxidative esterification (entry 1d). Similarly, electronic im-
pacts could be seen: trifluoroethanol (Table 1, 1f) was less
reactive than ethanol, requiring a higher temperature and
longer reaction time, yet still provided the desired ester albeit
in a very modest 39% isolated yield.
2
3
3
21
tion. Concerning the abrupt deactivation of the catalytic
system when using a full equivalent of the base, we ruled out
ester to carboxylic acid degradation of the product: we used
rigorously dried K CO , and no loss of ester or appearance
of the carboxylic acid was detected by GC analysis.
Next, a series of control reactions was executed to
investigate mechanistic possibilities. Under an atmosphere
ofoxygen(balloon) and withacetone replacedwithTHF, a
The inability to incorporate CꢀC double bonds within
substrates proved to be a significant limitation of this
method. Alkenes may serve as effective ligands for the Pd
catalyst, disrupting the catalytic cycle. Thus, no oxidative
esterification was observed on substrates containing CꢀC
double bonds that were not part of the aromatic systems.
Further, when cinnamaldehyde was substituted for hydro-
cinnamaldehyde, the deep garnet color indicative of the
active catalyst immediately was quenched to a pale yellow,
and no formation of the desired ester was observed.
2
3
(
21) Almeida, M. L. S.; Beller, M.; Wamg, G.-Z.; B €a ckvall, J.-E.
Chem.;Eur. J. 1996, 2, 1533.
22) (a) Shue, R. S. Chem. Commun. 1971, 1510. (b) Dams, M.; De Vos,
D. E.; Celen, S.; Jacobs, P. A. Agnew. Chem., Int. Ed. 2003, 42, 3512.
23) Wang, J.-R.; Yang, C.-T.; Liu, L.; Guo, Q.-X. Tetrahedron Lett.
007, 48, 5449.
(
(24) (a) B €a ckvall, J. E.; Hopkins, R. B. Tetrahedron Lett. 1988, 29,
2885. (b) B €a ckvall, J. E.; Hopkins, R. B.; Grennberg, M. M.; Awasthi,
A. K. J. Am. Chem. Soc. 1990, 112, 5160. (c) Piera, J.; B €a ckvall, J. E.
Angew. Chem., Int. Ed. 2008, 47, 3506.
(
2
5
02
Org. Lett., Vol. 15, No. 3, 2013

Table 3. Additional Aldehyde Scope with Ethanol
a
Isolated yields after column chromatography.
Finally, the inhibition by excess base (Figure 2) may also
trace its roots to the same phenomenon. Thus, we hypothe-
size that aldol condensation products form between acet-
one and the aldehyde substrates when a full equivalent of
the carbonate base is used for the reaction, eventually
deactivating the catalytic system in a similar manner.
Next, we screeneda range of aromatic aldehydes (Table 2),
and the importance of the aldehyde/hemiacetal equilibrium
again was apparent, as aromatic aldehydes required ex-
tended reaction times and elevated temperatures. Reacting
the aldehydes overnight at 50 °C resulted in moderate to
good yields when using primary alcohols such as methanol
and ethanol, but yields suffered greatly with the bulkier
isopropanol. The lone exception to this was when using
a
Figure 3. Comparison of alternative catalysts.
25
L16ꢀL19). Somewhat unexpectedly, the aminobiphenyl-
derived Pd/XPhos precatalyst (Figure 3) is significantly less
reactive, showing poor conversion at 50 °C and providing
none of the desired product at rt.
In conclusion, a new method of direct ester synthesis
from aldehydes has been developed. This approach utilizes
inexpensive acetone as a H-acceptor and provides a sus-
tainable and cost-effective option to synthesize esters
from the corresponding aldehydes. Initial investigations into
this catalytic cycle indicate the importance of the aldehyde/
hemiacetal equilibrium on the observed reactivity. The com-
4-nitrobenzaldehyde, presumably because the electron-with-
drawing nature of the ring shifted the aldehyde/hemiacetal
equilibrium enough to compensate somewhat for the low
reactivity of the sterically more demanding isopropanol.
Additional aldehydes including heterocyclic motifs were
converted to the corresponding ethyl esters in good to modest
yields (Table 3, 8bꢀ11b). Unsurprisingly, the attempted
conversion of 4-chlorobenzaldehyde to the corresponding
ester resulted in a messy reaction from which only 5% of the
desired ester could be isolated (Table 3, 15b). Competitive
oxidative addition of the Pd catalyst to the aryl chloride is the
likely source of the product mixture.
bination of XPhos with Pd(OAc) proved to be the most
2
effective catalyst for this transformation. The results de-
scribed herein also have implications for cross-coupling
reactions of aldehyde-containing substrates in alcohol sol-
vents, where the competing redox esterification may compete
with the desired cross-coupling event.
Acknowledgment. Support by the National Science
Foundation (NSF-GOALI 0848460) and NIGMS
(
R01 GM035249-27, R01 GM081376) is gratefully
acknowledged. Dr. Rakesh Kohli (University of
Pennsylvania) is acknowledged for obtaining HRMS data.
Finally, to investigate the importance of XPhos/
Pd(OAc) in this transformation, a range of other ligands
2
were examined in the conversion of benzaldehyde to ethyl
benzoate using our optimized conditions. Not only did
XPhos prove superior to the other phosphine ligands
screened, but it seems that the XPhos/Pd(OAc) combina-
2
tion is crucial to effect the desired reactivity (Figure 3,
Supporting Information Available. Additional screen-
ing data, experimental procedures and characterization
data for all compounds. H and C NMR spectra. This
material is available free of charge via the Internet at
http://pubs.acs.org.
1
13
(25) Kinzel, T.; Zhang, Y.; Buchwald, S. L. J. Am. Chem. Soc. 2010,
32, 14073.
1
The authors declare no competing financial interest.
Org. Lett., Vol. 15, No. 3, 2013
503