Decarboxylative ketone aldol reactions: Development and mechanistic evaluation under metal-free conditions

Article abstract of DOI:10.1021/jo901022j

(Chemical Equation Presented) Malonic acid half thioesters (MAHTs) and malonic acid half oxyesters (MAHOs) are shown to undergo decarboxylative nucleophilic addition reactions with ketone and aldehyde electrophiles in the presence of stoichiometric or catalytic quantities of triethylamine at room temperature. The ability to perform these reactions under metal-free conditions has enabled a detailed mechanistic analysis of the reaction pathway leading to the 1H NMR spectroscopic characterization of a postnucleophilic addition/predecarboxylation intermediate and experimental evidence for a reversible formation of this intermediate followed by an irreversible decarboxylation. Rate constants for each of the bond forming/bond breaking steps in the reaction pathway were also determined, casting light on the differing reactivity betweenMAHOandMAHT nucleophiles in these processes. Finally, the mechanistic insights gained through these studies have been employed in the development of a new decarboxylative coumarin synthesis.

Full text of DOI:10.1021/jo901022j

pubs.acs.org/joc
Decarboxylative Ketone Aldol Reactions: Development and Mechanistic
Evaluation under Metal-Free Conditions
Nicole Blaquiere, Daniel G. Shore, Sophie Rousseaux, and Keith Fagnou*
Centre for Catalysis Research and Innovation, Department of Chemistry, University of Ottawa,
10 Marie Curie, Ottawa, ON, K1N 6N5, Canada
keith.fagnou@uottawa.ca
Received May 25, 2009
Malonic acid half thioesters (MAHTs) and malonic acid half oxyesters (MAHOs) are shown to
undergo decarboxylative nucleophilic addition reactions with ketone and aldehyde electrophiles in
the presence of stoichiometric or catalytic quantities of triethylamine at room temperature. The
ability to perform these reactions under metal-free conditions has enabled a detailed mechanistic
1
analysis of the reaction pathway leading to the H NMR spectroscopic characterization of a
postnucleophilic addition/predecarboxylation intermediate and experimental evidence for a rever-
sible formation of this intermediate followed by an irreversible decarboxylation. Rate constants for
each of the bond forming/bond breaking steps in the reaction pathway were also determined, casting
light on the differing reactivity between MAHO and MAHT nucleophiles in these processes. Finally,
the mechanistic insights gained through these studies have been employed in the development of a
new decarboxylative coumarin synthesis.
Introduction
Claisen condensation of malonyl and acetyl units
(Scheme 1).1 The intimate mechanism of these processes
remains a matter of debate, with three distinct possibilities
being advanced. First, mechanisms involving a concerted
decarboxylation and nucleophilic attack have been proposed
(mechanism 1, Scheme 1). Alternatively, two stepwise addi-
tions have also been proposed, one requiring an initial
decarboxylation step to produce a thioester enolate followed
Chemists have long drawn inspiration from nature’s abil-
ity to form carbon-carbon bonds under mild, aqueous, and
aerobic conditions. Illustrative are the enzymes that exploit
the reactivity of β-ketoacid compounds as a means of facil-
itating enolate formation such as fatty acid decarboxylase
that carries out a chain extension step by the decarboxylative
(1) For reviews on polyketide biosynthesis, see: (a) Smith, S.; Tsai, S.-C.
Nat. Prod. Rep. 2007, 24, 1041. (b) Hill, A. M. Nat. Prod. Rep. 2006, 23, 256.
(c) White, S. W.; Zheng, J.; Zhang, Y.-M.; Rock, C. O. Annu. Rev. Biochem.
2005, 74, 791. (d) Shen, B. Curr. Opin. Chem. Biol. 2003, 7, 285. (e) Staunton,
J.; Weissman, K. J. Nat. Prod. Rep. 2001, 18, 380.
(2) For references containing evidence supporting a concerted mechan-
ism, see: (a) Cane, D. E.; Liang, T.-C.; Taylor, P. B.; Chang, C.; Yang, C.-C.
J. Am. Chem. Soc. 1986, 108, 4957. (b) Sood, G. R.; Robinson, J. A.; Ajaz, A.
A. J. Chem. Soc., Chem. Commun. 1984, 1421. (c) Arnstadt, K. I.; Schindl-
beck, G.; Lynen, F. Eur. J. Biochem. 1975, 55, 561. (d) Vagelos, P. R.;
Alberts, A. W. J. Biol. Chem. 1960, 235, 2786. (e) Brady, R. O. Proc. Natl.
Acad. Sci. U.S.A. 1958, 44, 993. For references that bring additional insight
to the accepted concerted mechanism, see: (f) Kresze, G.-B.; Steber, L.;
Oesterhelt, D.; Lynen, F. Eur. J. Biochem. 1977, 79, 191. (g) Lynen, F.
Biochem. J. 1967, 102, 381.
(3) For references containing evidence supporting a stepwise decarbox-
ylation-addition mechanism, see: (a) Davies, C.; Heath, R. J.; White, S. W.;
Rock, C. O. Structure 2000, 8, 185. (b) Dewar, M. J. S.; Dieter, K. M.
Biochemistry 1988, 27, 3302. For references that bring additional insight to
the accepted stepwise mechanism, see: (c) Olsen, J. G.; Kadziola, A.; von
Wettstein-Knowles, P.; Siggaard-Anderson, M.; Larsen, S. Structure 2001, 9,
233. (d) Price, A. C.; Choi, K.-H.; Heath, R. J.; Li, Z.; White, S. W.; Rock, C.
O. J. Biol. Chem. 2001, 276, 6551. (e) Jez, J. M.; Ferrer, J.-L.; Bowman, M. E.;
Dixon, R. A.; Noel, J. P. Biochemistry 2000, 39, 890. (f) Dreier, J.; Khosla, C.
Biochemistry 2000, 39, 2088. (g) Qiu, X.; Janson, C. A.; Konstantinidis, A.
K.; Nwagwu, S.; Silverman, C.; Smith, W. W.; Khandekar, S.; Lonsdale, J.;
Abdel-Meguid, S. S. J. Biol. Chem. 1999, 274, 36465. (h) Huang, W.; Jia, J.;
Edwards, P.; Dehesh, K.; Schneider, G.; Lindqvist, Y. EMBO J. 1998, 17,
1183. (i) Dewar, M. J. S.; Storch, D. M. Proc. Natl. Acad. Sci. U.S.A. 1984,
82, 2225.
6190 J. Org. Chem. 2009, 74, 6190–6198
Published on Web 07/09/2009
DOI: 10.1021/jo901022j
r
2009 American Chemical Society

Blaquiere et al.
JOCArticle
SCHEME 1. Possible Mechanistic Pathways for Enzymatic Decarboxylative Claisen Condensations
by electrophilic trapping by the acetyl unit (mechanism 2,
Scheme 1) and the other involving enolization and nucleo-
philic addition of the malonyl moiety to the acetyl group
followed by decarboxylation (mechanism 3, Scheme 1).
There is support for both the concerted (mechanism 1)2
and the stepwise pathways (mechanisms 23 and 34), but a
general consensus has not been achieved.
SCHEME 2. Biomimetic Approach for Laboratory Aldol Pro-
cesses
Even in the absence of precise mechanistic knowledge of
the discrete reaction steps in biological decarboxylative
Claisen condensations, this reactivity has been a source of
inspiration for synthetic chemists. For example, in tradi-
tional laboratory aldol and related processes, many methods
for ester enolate formation involve deprotonation with a
strong base followed by trapping as a silyl ketene acetal.5 A
biomimetic approach involving the decarboxylation of half
ester malonates can avoid the need for stoichiometric strong
bases as well as providing a traceless means of activation with
carbon dioxide gas as the only byproduct and providing both
enthalpic and entropic driving forces for the global carbon-
carbon bond-forming process (Scheme 2).
Several transformations have been developed which take
advantage of this decarboxylative reactivity,6,7 the most
renowned being the Galat modification of the Knoevena-
gel-Doebner reaction,7a where the decarboxylative addition
of malonates to aldehydes results in the formation of
(6) For examples of decarboxylative Claisen condensations, see: (a)
Scott, A. I.; Wiesner, C. J.; Yoo, S.; Chung, S.-K. J. Am. Chem. Soc. 1975,
97, 6277. (b) Kobuke, Y.; Yoshida, J.-I. Tetrahedron Lett. 1978, 4, 367. (c)
Brooks, D. W.; Lu, L. D.-L.; Masamune, S. Angew. Chem., Int. Ed. Engl.
ꢀ
1979, 18, 72. (d) Sakai, N.; Sorde, N.; Matile, S. Molecules 2001, 6, 845. (e)
Chen, H.; Harrison, P. H. M. Can. J. Chem. 2002, 80, 601.
(7) For examples of decarboxylative Knoevenagel condensations, see: (a)
Galat, A. J. Am. Chem. Soc. 1946, 68, 376. (b) Ragoussis, N.; Ragoussis, V.
J. Chem. Soc., Perkin Trans. 1 1998, 3529. (c) List, B.; Doehring, A.;
Hechavarria Fonseca, M. T.; Wobser, K.; van Thienen, H.; Rios Torres,
ꢀ
R.; Llamas Galilea, P. Adv. Synth. Catal. 2005, 347, 1558. (d) Berrue, F.;
Antoniotti, S.; Thomas, O. P.; Amade, P. Eur. J. Org. Chem. 2007, 1743.
(4) For references containing evidence supporting a stepwise addition-
decarboxylation mechanism, see: (a) Sedgwick, B.; Cornforth, J. W. Eur. J.
Biochem. 1977, 75, 465. (b) Vagelos, P. R. J. Am. Chem. Soc. 1959, 81, 4119.
(c) Wakil, S. J.; Ganguly, J. J. Am. Chem. Soc. 1959, 81, 2597.
(5) For selected reviews, see: (a) Palomo, C.; Oiarbide, M.; Garcı
´
´
a, J. M.
Chem. Soc. Rev. 2004, 33, 65. (b) Palomo, C.; Oiarbide, M.; Garcıa, J. M.
Chem.;Eur. J. 2002, 8, 36. (c) Machajewski, T. D.; Wong, C.-H. Angew.
Chem., Int. Ed. 2000, 39, 1352. (d) Mahrwald, R. Chem. Rev. 1999, 99, 1095.
(e) Carreira, E. M. In Comprehensive Asymmetric Catalysis; Jacobsen, E. N.,
Pfaltz, A., Yamammoto, H., Eds.; Springer: Heidelberg, Germany; 1999; Vol. 3, p
997. (f) Nelson, S. G. Tetrahedron: Asymmetry 1998, 9, 357.
J. Org. Chem. Vol. 74, No. 16, 2009 6191

Blaquiere et al.
JOCArticle
R,β- and β,γ-unsaturated esters. The concept of decarbox-
ylative enolate addition chemistry has also been investigated
by other research groups, including those of Ricci8 and
Wennemers,9 for additions to imine and nitroolefin electro-
philes. Despite the remarkable success associated with this
approach, the mechanistic ambiguity associated with the
biological processes is mirrored by the uncertainty in the
mechanism of the synthetic applications since multiple pos-
sibilities (and reactive intermediates) have been advanced
across these different reaction classes. In some cases, dec-
arboxylation is proposed to occur prior to nucleophilic
attack6e,8 (analogous to mechanism 2, Scheme 1) and in
others, the stepwise nucleophilic addition/decarboxylation
mechanism is advanced (analogous to mechanism 3,
Scheme 1).6a,6b,7b,7d
SCHEME 3. Copper-Catalyzed Decarboxylative Aldol Reaction
with MAHTs
As direct precedent for the chemistry and the mechanistic
work described herein, Shair and co-workers developed both
racemic and enantioselective decarboxylative aldol reactions
involving the addition of phenyl- and benzylthioesters in the
form of malonic acid half thioesters (MAHTs) to a variety of
aldehydes in the presence of catalytic copper and a bisox-
azolidine ligand (Scheme 3).10 Cozzi and co-workers have
also described an enantioselective decarboxylative aldol
reaction between MAHTs and aldehydes using catalytic
Cu(OTf)2 and enantiopure bisbenzimidazole ligands.11
Mechanistic studies by the Shair group point to the
involvement of a stepwise reversible addition of the MAHT
to the aldehyde electrophile to form intermediate 1 prior to
decarboxylation,12 analogous to that proposed for the enzy-
matic process outlined as mechanism 3 in Scheme 1. While
intermediate 1 could not be detected directly, indirect evi-
dence for its involvement was obtained from kinetic studies
and isotopic labeling experiments. A hypothesis for the role
of copper was advanced, involving the enforcement of a
planar arrangement of the carboxylate and thioester moi-
eties, stereoelectronically favoring deprotonation over dec-
arboxylation and predisposing the system to follow the
stepwise addition/decarboxylation pathway.
The ongoing debate in the literature regarding the me-
chanism of decarboxylative biological processes and the
wide range of proposed pathways for synthetic methods
inspired by this reactivity led us to further investigate the
mechanism of decarboxylative aldol reactions. Herein we
describe (1) the development of amine base-catalyzed dec-
arboxylative aldol reactions between malonic acid half oxye-
sters (MAHOs) and electron-deficient ketones and
aldehydes that occur in the absence of added metal salts
(Scheme 4), (2) proof for the involvement of a stepwise
addition/decarboxylation pathway (analogous to mechan-
ism 3) by direct observation and characterization of the key
postaddition predecarboxylation intermediate in the crude
reaction mixture by DOSY 1H NMR, (3) proof for the
reversible formation of this postaddition predecarboxylation
intermediate based on trapping experiments and kinetic
analysis, (4) evidence that the relative rates of each step in
SCHEME 4. Decarboxylative Aldol Reaction between MAHOs
and Electron-Deficient Ketones and Aldehydes
the reaction pathway must be considered when developing
new reactivity based on this reaction manifold, and (5) use of
these findings and validation of their importance in the
development of a novel decarboxylative coumarin synthesis
from MAHOs and salicylaldehyde substrates. Importantly,
these findings may also shed light on the nature of the
reaction mechanism in biological systems since, contrary to
many previous mechanistic reports,6a,6b,7d,12 these reactions
do not require the presence of a metal catalyst to undergo
decarboxylative nucleophilic addition.
Results and Discussion
To facilitate mechanistic analysis and the detection of
reactive intermediates along the reaction pathway by
NMR spectroscopy, initial studies targeted the development
of metal free decarboxylative aldol condensation reactions.
Gratifyingly, it was determined that activated aldehyde and
ketone electrophiles, such as ethyl pyruvate, undergo effi-
cient decarboxylative coupling with nucleophiles such as
MAHT 2. For example, in under 3 h, the reaction between
MAHT 2 and ethyl pyruvate 4 proceeds smoothly at room
temperature in the presence of 1 equiv of triethylamine to
afford the β-hydroxythioester 5 in 70% yield (Table 1, entry 1).
Catalytic base will also induce decarboxylative addition,
with 10 mol % triethylamine resulting in an 86% yield of
product overnight (entry 2). Encouraged by this promising
reactivity, MAHO 3, which has not been previously shown to
undergo decarboxylative aldol reactions, was also evaluated
as a potential coupling partner. We were pleased to see that over
a slightly longer reaction time (4 h) and otherwise identical
conditions, MAHO 3 also reacts smoothly with ethyl pyruvate
(entries 3 and 4).
(8) Ricci, A.; Petterson, D.; Bernardi, L.; Fini, F.; Fochi, M.; Perez
Herrera, R.; Sgarzani, V. Adv. Synth. Catal. 2007, 349, 1037.
(9) Lubkoll, J.; Wennemers, H. Angew. Chem., Int. Ed. 2007, 46, 6841.
(10) (a) Lalic, G.; Aloise, A. D.; Shair, M. D. J. Am. Chem. Soc. 2003, 125,
2852. (b) Magdziak, D.; Lalic, G.; Lee, H. M.; Fortner, K. C.; Aloise, A. D.;
Shair, M. D. J. Am. Chem. Soc. 2005, 127, 7284.
(11) Orlandi, S.; Benaglia, M.; Cozzi, F. Tetrahedron Lett. 2004, 45, 1747.
(12) Fortner, K. C.; Shair, M. D. J. Am. Chem. Soc. 2007, 129, 1032.
6192 J. Org. Chem. Vol. 74, No. 16, 2009

Blaquiere et al.
JOCArticle
TABLE 1. Optimization of the Decarboxylative Aldol Reaction of
Malonic Acid Half Thioesters and Esters with Ethyl Pyruvatea
in the absence of paramagnetic copper(II) salts that can
significantly limit the use of NMR spectroscopic analysis
of the crude reaction mixture. Employing the newly estab-
1
lished reaction conditions and substrate combinations, H
NMR analysis was performed with a range of substrates. We
were pleased to find that 1H NMR analysis of reaction
progress between MAHO 3 and ethyl pyruvate 4 with
triethylamine-d15 in benzene-d6 revealed the appearance of
peaks that were tentatively assigned as the two diastereomers
of the postaddition/predecarboxylation intermediate 8
(Figure 1). As might be anticipated, all attempts to isolate
this postulated intermediate failed. Consequently, to identify
and characterize this intermediate species, DOSY
(Diffusion-Ordered SpectroscopY) NMR analysis of the
reaction mixture was employed.14 As shown in Figure 1,
unreacted ethyl pyruvate 4 and product 6 can be easily
identified.15 Furthermore, by extracting the peaks that cor-
respond to the third compound in the reaction mixture, this
species was identified as the postaddition/predecarboxyla-
tion intermediate 8, providing compelling evidence that the
reaction follows a pathway involving a stepwise addition-
decarboxylation process (mechanism 3, Scheme 1). While
reactions with MAHT substrates proceed too rapidly at
room temperature to permit an analogous analysis, by
performing the reaction at 0 °C, a similar postaddition/
entry
X
Et3N (equiv)
temp (°C)
yield (%)b
1
2
3
4
S
S
O
O
1.0
0.1
1.0
1.0
23
23
50
23
70
86c
70
82
aConditions: malonic acid half thioester or oxyester (1 equiv), ethyl
pyruvate (1 equiv), triethylamine, THF (0.5 M), 2-4 h. bIsolated yields.
cOvernight reaction.
TABLE 2. Scope of the Decarboxylative Aldol Reaction with MAHO
Nucleophilesa
entry
Ar
R
R0
yield (%)b
1
2
3
4
5
6
7
8
9
10
Ph
Ph
Ph
Ph
Ph
Ph
Me
CF3
Ph
H
H
Me
Me
Me
Me
Me
Et
Et
Et
Et
iPr
Me
Et
Et
Et
Et
82
84
46
65
46
47
42
70
50
70
1
predecarboxylation intermediate could be detected by H
NMR, supporting a common mechanism for both thioester
and ester nucleophiles (Figure 2).
p-CN-C6H4
p-F-C6H4
3,5-di(OMe)-C6H4
p-OMe-C6H4
To probe the potential reversibility of the initial addition
in this stepwise mechanism, a competition experiment was
devised in which a second, and more reactive, electro-
phile was added to a reaction between MAHO 3 and ethyl
pyruvate following maximal conversion to the postaddition/
predecarboxylation intermediate 8 (Figure 3). Ethyl trifluor-
opyruvate was chosen as the second electrophile because of
its clean and fast reaction with MAHO 316 along with its
structural similarity to ethyl pyruvate. The reaction progress
as well as the relative ratios of the unreacted starting
materials, the post addition/predecarboxylation intermedi-
ate, and the products were monitored by 1H NMR
(Figure 3).
To establish baseline reactivity and the optimal moment
for the addition of the second electrophile, a kinetic profile
for the reaction of MAHO 3 with ethyl pyruvate was
obtained. Under standard conditions, maximal conversion
(approximately 75%) to the postaddition/predecarboxyla-
tion intermediate 8 occurs 130 min into the reaction
(Figure 3, solid lines). After 130 min, aldol product 6 is
present in 10% yield and 15% of unreacted starting material
3 remains. This point in the reaction profile was chosen for
the time of addition of the second electrophile, ethyl trifluoro-
pyruvate, in the competition experiment.
aConditions: malonic acid half oxyester (1 equiv), ethyl pyruvate (1
equiv), triethylamine (1 equiv), THF (0.5), 2-4 h. bIsolated yields.
While the reactions can be carried out in the presence of
catalytic triethylamine (10 mol %), stoichiometric triethyla-
mine was used in most cases to allow for shorter reaction
times. Conveniently, no precautions for the use of inert
atmosphere are required since the reactions can be per-
formed at room temperature in an open flask. The reactions
proceed well with electron-withdrawing groups on the R-
ketoesters (Table 2, entry 2), and also occur with the corre-
sponding phenyl ketone, albeit in lower yields plausibly due
to the diminished electrophilicity of this substrate (entry 3).
Aldehydes are also compatible, although yields are more
variable ranging from 46% to 65% (entries 4 and 5).
Distillation of these aldehydes just prior to use was found
to be crucial for success since they are known to undergo
polymerization at room temperature. Substitution on the
aromatic ring of the MAHO with both electron-rich and
electron-poor groups also results in moderate to good yields,
with electron-rich functionalities generally providing better
conversions (entries 7-10). The lower yield in entry 7 may be
attributed to the instability of the product due to the
heightened electrophilic nature of the ester.13
(14) Morris, K. F.; Johnson, C. S. Jr. J. Am. Chem. Soc. 1993, 115, 4291.
This technique is an effective two-dimensional spectroscopic method that
permits spectroscopic separation of crude reaction components in a complex
mixture according to diffusion constants that depend on variations in size
and charge. By extracting horizontal traces from points along the diffusion
axis corresponding to the reaction components for a reaction between
MAHO 3 (or MAHT 2) and ethyl pyruvate, structural elucidation of the
different species present in the reaction mixture may be achieved.
(15) The diffusion constant for unreacted MAHO 3 is very close to that of
intermediate 8. This made the extraction of a clean 1H NMR spectrum of
MAHO 3 from the DOSY spectrum difficult.
Mechanistic Studies. Crucial to the ability to follow the
reaction progress and evaluate the involvement of unstable
reactive intermediates is the ability to perform these reactions
(13) Nonaromatic esters were generally incompatible with the reaction,
with only trace amounts of product being observed with malonic acid half
ethyl and half tert-butyl esters by NMR.
(16) See the Supporting Information.
J. Org. Chem. Vol. 74, No. 16, 2009 6193

Blaquiere et al.
JOCArticle
FIGURE 1. DOSY and extracted 1H NMR spectra (500 MHz) of the reaction of MAHO 3 and ethyl pyruvate. For peak assignment of 8, see
the Supporting Information.
FIGURE 2. DOSY and extracted 1H NMR spectra (500 MHz) of the reaction of MAHT 2 and ethyl pyruvate. For peak assignment of 9, see
the Supporting Information.
A kinetic profile was then obtained for the reaction of
MAHO 3 and ethyl pyruvate following the addition of ethyl
trifluoropyruvate after 130 min, and a notable diversion
from the behavior of the baseline reaction was observed
(Figure 3). A significant and rapid drop in MAHO 3 con-
centration occurs immediately after ethyl trifluoropyruvate
addition compared to the baseline reaction due to a fast
reaction with this more reactive electrophile. More impor-
tantly, a significant decrease in the concentration of inter-
mediate 8 also occurs without concomitant appearance of
product 6. This is a strong indication that, in addition to
decarboxylative product formation, there must be another
decomposition pathway for intermediate 8, most likely via a
retro-aldol reaction. This would regenerate starting material
3, which will then undergo fast reaction with the more
reactive ethyl trifluoropyruvate to generate a new post
addition/predecarboxylation intermediate 10, and result in
a net decrease in the concentration of 8.
The final product distribution also provides strong evidence
for a reversible formation of the postaddition/predecarboxyla-
tion intermediate 8. After 650 min, product 6, arising from
reaction with ethylpyruvate, is formed in 25% yield and 11,
arising from reaction with ethyl trifluoropyruvate, is formed in
41% yield. If the formation of intermediate 8 were irreversible,
up to 85% yield of product 6 would be anticipated (res-
ulting from the 75% conversion to intermediate 8 and 10%
6194 J. Org. Chem. Vol. 74, No. 16, 2009

Blaquiere et al.
JOCArticle
FIGURE 3. Competition experiment between ethyl pyruvate and ethyl trifluoropyruvate electrophiles in the decarboxylative aldol reaction to
determine the reversibility of intermediate formation. In the graph, the relative concentrations, monitored by 1H NMR, of MAHO 3,
intermediate 8, and product 6 under standard conditions (solid lines) are compared to the relative concentrations of the same compounds (blue,
dark green, and red data points, respectively) when ethyl trifluoropyruvate is added at 130 min into the reaction. The yellow and light green data
points represent the relative concentrations of the newly formed intermediate 10 and product 11 upon addition of this more reactive
electrophile. The dashed line represents the time of addition of ethyl trifluoropyruvate.
conversion to product 6 that had occurred prior to addition of
the second electrophile), but no more than 15% of 11, arising
from reaction with the second electrophile and unreacted
MAHO 3, could form. Since the experimentally observed yield
of 11 surpasses the amount of remaining unreacted 3 at the
point of ethyl trifluoropyruvate addition, the difference must be
accounted for by a reversion to starting material 3 from the key
postaddition/predecarboxylation intermediate 8, followed by
electrophilic trapping by ethyl trifluoropyruvate.
experimental validation of a reversible nucleophilic addition
step, but also permits a more detailed kinetic analysis of
this reaction pathway casting light on the relative importance
of individual steps in the overall process. The integrated rate
law of a model system for a reversible addition to the
electrophile was evaluated using COPASI (Complex Path-
way Simulator) kinetics analysis software.17 The second step
in the reaction pathway was assumed to be irreversible since
(17) Hoops, S.; Sahle, S.; Gauges, R.; Lee, C.; Pahle, J.; Simus, N.;
Singhal, M.; Xu, L.; Mendes, P.; Kummer, U. Bioinformatics 2006, 22, 3067.
For further information or to download COPASI, see http://www.copasi.
org/tiki-index.php.
Kinetic Studies. The ability to follow the consumption of
starting materials and appearance of the postaddition/pre-
decarboxylation intermediate over time not only allowed for
J. Org. Chem. Vol. 74, No. 16, 2009 6195

Blaquiere et al.
JOCArticle
SCHEME 5. Calculated Rate Constants for the Reversible Reaction of MAHO 3 or MAHT 2 with Ethyl Pyruvate in Benzene-d6
carbon dioxide can bubble out of the reaction and the
concentration of CO2 remaining in the reaction mixture
was assumed to be too low to significantly influence the
kinetics of the reaction. The integrated rate equation for the
reversible model was fit to the experimental data obtained by
1H NMR analysis of the reaction profile in benzene-d6
(Figure 4a) and THF-d8 (Figure 4b) and an excellent accord
was observed. The reaction in benzene-d6 provides a closer fit
to the reversible model than the reaction in THF-d8, which
can be explained by the increased amount of side reactions
(decarboxylation, etc.) that occur in more polar solvents,
affecting starting material concentrations.
The theoretical rate expressions for the reversible model
were also solved allowing a determination of the correspond-
ing rate constants (Scheme 5). These values not only point to
the important reaction parameters that must be considered
during the development of analogous processes, but may
also begin to explain the superior reactivity of MAHT
compared to MAHO nucleophiles. Reaction with the
MAHT nucleophile is associated with faster values for each
elemental reaction step. Furthermore, the relative rates for
the decarboxylation of intermediates 9/8 compared to their
formation is larger for MAHT compared to MAHO nucleo-
philes as well. This agrees with the shape of the plots
previously obtained (Figures 3 and 4) where a significant
buildup of the postaddition/predecarboxylation intermedi-
ate is observed (dark green data points). When envisioning
the development of new decarboxylative processes, these
results indicate that in addition to the rates of nucleophilic
addition, the relative rates of decarboxylation and reversion
to starting materials of the first step must be considered. This
concept has been illustrated below in the development of a
novel decarboxylative coumarin synthesis.
Development of a Decarboxylative Coumarin Synthesis.
While evaluating the scope of the amine-catalyzed decarbox-
ylative aldol reactions described herein, it was determined
that less reactive carbonyl electrophiles such as benzalde-
hyde failed to efficiently participate. For example, under
standard conditions the addition of MAHO 3 to benzalde-
hyde results in only 13% yield of the desired aldol product
SCHEME 6. Decarboxylative Aldol Reaction with Benzaldehyde
FIGURE 4. Relative concentrations of MAHO 3 (blue), intermedi-
ate 8 (green), and β-hydroxyester 6 (red) over 11 h analyzed by 1H
NMR in (a) benzene-d6 and (b) THF-d8 overlapped with the
theoretical curves (solid lines) obtained from the integrated rate
equation predicted by COPASI for a reversible mechanism. Data
points represent the mean of three trials, and error bars represent the
standard deviation. Integrations were measured relative to an
internal standard (pentachlorobenzene).
6196 J. Org. Chem. Vol. 74, No. 16, 2009

Blaquiere et al.
JOCArticle
SCHEME 7. Reaction of Salicylaldehyde with MAHO 3
(Scheme 6). Following our mechanistic studies, this was
hypothesized as not only being a consequence of a slow initial
nucleophilic attack (k1) resulting in low concentrations of
intermediate 12, but perhaps also as a consequence of
unfavorable kinetics at the reversible addition (k-1) and
decarboxylation steps (k2). We reasoned that if this inter-
mediate could follow a third reaction pathway that is fast
(compared to k-1) and irreversible, then the overall process
may become more favorable (Scheme 7).
SCHEME 8. Experimental Support for the Proposed Mechanism
Outlined in Scheme 6 Where Dehydration Occurs Prior to Decar-
boxylation
To test this hypothesis, salicylaldehyde was selected as a
suitable electrophile since it was envisioned that upon for-
mation of intermediate 13 (Scheme 7), the o-hydroxyl moiety
might undergo transesterification with the phenyl ester to
give intermediate 14. Once this has occurred, the retro-aldol
step from intermediate 14 would become strongly disfavored
thus introducing an alternative “irreversible” step in the
overall reaction scheme. A subsequent decarboxylation
could then generate coumarin 16. This notion is supported
by the fact that when salicylaldehyde was reacted with 2
equiv of MAHO 3 under the standard reaction conditions,18
compound 19 was isolated in 64% yield. Subsequent experi-
ments support the involvement of a pathway similar to that
depicted in Scheme 7 where dehydration precedes decarbox-
ylation (vide infra). Importantly, the success of this experi-
ment compared to the use of benzaldehyde validates the need
to consider all of the reaction steps when proposing the
application of this approach in a new process.
On the other hand, if dehydration were to occur prior to
decarboxylation, a new intermediate coumarin-3-carboxylic
acid 17 could be formed that might undergo a second
reaction with a second MAHO molecule. Consequently,
intermediate 17 was synthesized independently, and in the
presence of MAHO 3 or 20, rapidly reacts to give compounds
19 and 23 in good yields (Scheme 8). Finally, reaction of
methyl-substituted MAHO 20 with salicylaldehyde provides
methyl-substituted coumarin 22 exclusively in 40% yield
(Scheme 8). In this instance, the presence of the methyl
substituent prevents dehydration prior to decarboxylation
at intermediate 21, leading to the formation of 22. This novel
reaction pathway, and the enhanced reactivity of carboxyl-
substituted coumarin 17 compared to 16, offers a second
To explain the formation of 19 instead of the predicted
compound 16, two additional reactions were carried out.
A control reaction between MAHO 3 and coumarin 16 was
performed and was found to not result in the formation of
19, ruling out the possibility of the product arising from
1,4-addition of the malonate to a coumarin intermediate.
(18) Two equivalents of MAHO 3 were used to maximize product yield.
J. Org. Chem. Vol. 74, No. 16, 2009 6197

Blaquiere et al.
JOCArticle
1
useful and conceptually interesting role for the carboxylic
acid moiety: in addition to acting as a traceless activating
group for the generation of nucleophilic enolates, it may also
be used to activate electrophiles in a similar traceless fashion.
NMR tube. A baseline H NMR spectrum was obtained. All
NMR spectra were taken with sample spinning enabled to
ensure that the reaction mixture was sufficiently mixed. Et3N
(23 μL, 1.0 equiv) was then added to the NMR tube via syringe,
the cap was replaced, and the tube was inverted several times to
ensure adequate mixing. A 1H NMR spectrum was then taken to
ensure the starting material had been completely deprotonated.
Lastly, the electrophile of interest (1.0 equiv) was added via
syringe, the time was noted, the tube was inverted to mix, and a
starting point 1H NMR spectrum was obtained. The NMR tube
cap was replaced with one in which a hole had been pierced to
Conclusion
In conclusion, an amine base-catalyzed decarboxylative
aldol reaction has been developed. Mechanistic studies
established the reversible formation of a postaddition/pre-
decarboxylation intermediate in the decarboxylative aldol
reaction of both MAHTs and MAHOs. The information
garnered from these studies shed light on the importance of
considering reversibility and relative rates of elementary
reaction steps in reaction development. As well, the applica-
tion of this decarboxylative reactivity to the synthesis of
coumarin derivatives benefited from this mechanistic insight
and in turn resulted in the use of the carboxylate moiety as a
traceless activating group for the 1,4-addition of malonic
acid derivatives to R,β-unsaturated esters. Finally, these
mechanistic findings, performed in the absence of metal
catalysts and under conditions more closely resembling the
biological systems, could influence current opinions as to the
pathway of decarboxylative reactions in nature.
1
allow the escape of CO2 gas. Ten H NMR spectra were then
taken at 1.5 min intervals. Once this program was complete, 20
1H NMR spectra were taken at 2 min intervals. Upon comple-
tion of these spectral scans, spectra were taken at 20-30 min
intervals as the sample reacted overnight (8-18 h). Concentra-
tions of reaction components at each time interval were then
determined relative to internal standard.
1
Procedure for the H NMR-Monitored Competition Experi-
ment. MAHO 3 (30 mg, 1.0 equiv) and internal standard
pentachlorobenzene (41.7 mg, 1.0 equiv) were weighed into a
disposable glass vial. The vial was then charged with 0.75 mL
(0.22 M) of C6D6. The reaction mixture was then stirred until
homogeneous and the entire solution was transferred by pipet
into a clean, oven-dried NMR tube. A baseline 1H NMR
spectrum was obtained. All NMR spectra were taken with
sample spinning enabled to ensure that the reaction mixture
was sufficiently mixed. Et3N (23 μL, 1.0 equiv) was then added
to the NMR tube via syringe, the cap was replaced, and the tube
was inverted several times to ensure adequate mixing. A 1H
NMR spectrum was then taken to ensure the starting material
had been completely deprotonated. Ethyl pyruvate 4 (1.0 equiv)
was added via syringe, the time was noted, the tube was inverted
to mix, and a 1H NMR spectrum was obtained. Maximum
intermediate concentration is observed at approximately t =
130 min. As such, the NMR tube cap was replaced with one in
which a hole had been pierced to allow the escape of CO2 gas and
10 1H NMR spectra were taken at 1.5 min intervals. Once this
Experimental Section
Representative Procedure for the Synthesis of Malonic Acid
Half Esters: Monophenyl Malonate (3). A neat solution of 2,2-
dimethyl-1,3-dioxane-4,6-dione (5.0 g, 34.7 mmol, 1.0 equiv)
and phenol (3.26 g, 34.7 mmol, 1.0 equiv) was heated to 120 °C
for 2 h, then cooled to room temperature. The reaction mixture
was then concentrated in vacuo and loaded directly on a silica
gel column. Chromatography on silica gel (20-40% EtOAc in
hexanes) gave 4.7 g of product as a white solid in 75% yield. 1H
NMR (300 MHz, CDCl3, 293 K, TMS) δ 3.66 (s, 2H), 7.11-7.14
(m, 2H), 7.22-7.27 (m, 1H), 7.35-7.41 (m, 2H), 8.85 (br s, 1H).
General Procedure for the Decarboxylative Aldol Addition of
Malonic Acid Half Esters to Electrophiles: 2-Hydroxy-2-methyl-
succinic Acid 1-Ethyl Ester 4-Phenyl Ester (6). Monophenyl
malonate (0.100 g, 0.56 mmol, 1.0 equiv) was weighed into a
dry screw-capped vial equipped with a magnetic stir bar. THF
(1.1 mL, 0.5 M), triethylamine (77 μL, 0.56 mmol, 1.0 equiv),
and ethyl pyruvate (62 μL, 0.56 mmol, 1.0 equiv) were added via
syringe. The resulting mixture was stirred at room temperature
for 4 h, with a hole pierced in the septum to allow release of built
up pressure. Upon completion, the solvent was evaporated in
vacuo and the reaction mixture was purified by silica gel flash
column chromatography (15% EtOAc in hexanes), which af-
forded 114 mg of product as a clear oil in 81% yield. 1H NMR
(400 MHz, CDCl3, 293 K, TMS) δ 1.30 (t, J = 7.5 Hz, 3H), 1.52
(s, 3H), 2.93 (d, J = 16.0 Hz, 1H), 3.22 (d, J = 16.6 Hz, 1H), 3.77
(br s, 1H), 4.22-4.33 (m, 2H), 7.05-7.09 (m, 2H), 7.20-7.26 (m,
1H), 7.35-7.39 (m, 2H). 13C NMR (100 MHz, CDCl3, 293 K,
TMS) δ 14.1 (CH3), 26.5 (CH3), 44.4 (CH2), 62.3 (CH2), 72.5
(C), 121.5 (CH), 126.1 (CH), 129.5 (CH), 150.3 (C), 169.4 (C),
175.4 (C). IR (νmax, cm-1) 3192, 3075, 2936, 1755, 1742, 1592,
1459, 1195, 1115, 759. HRMS calculated for C13H16O5 (Mþ)
252.0998, found 252.0984.
1
program was complete, H NMR spectra were taken at 2 min
intervals until t = 130 min. At this point, the second electro-
phile, ethyl trifluoropyruvate (1.0 equiv), was added via syringe,
the tube was inverted, and a starting point 1H NMR spectrum
was obtained. Ten 1H NMR spectra were then taken at 1.5 min
intervals. Once this program was complete, 20 1H NMR spectra
were taken at 2 min intervals. Upon completion of these spectral
scans, spectra were then taken at 20-30 min intervals as the
sample reacted overnight (8-18 h). Concentrations of reaction
components at each time interval were then determined relative
to internal standard.
Acknowledgment. We thank NSERC and the University
of Ottawa for support of this work. The Research Corpora-
tion, Boehringer Ingelheim (Laval), Merck Frosst Canada,
Merck Inc., Eli Lilly, Amgen, and Astra Zeneca Montreal
are thanked for additional unrestricted financial support. N.
B. thanks NSERC for a postgraduate scholarship (CGS-M),
and S.R. thanks NSERC for an undergraduate summer
research award. Dr. Glenn Facey is thanked for NMR
ꢀ
spectroscopy assistance. Dr. Benoıt Liegault is thanked for
assistance in preparing several figures.
General Procedure for the Determination of Reaction Profile
by 1H NMR. Monophenyl malonate (30 mg, 1.0 equiv) and
internal standard pentachlorobenzene (41.7 mg, 1.0 equiv) were
weighed into a disposable glass vial. The vial was then charged
with 0.75 mL (0.22 M) of deuterated solvent. The reaction
mixture was then stirred until homogeneous and the entire
solution was transferred by pipet into a clean, oven-dried
Supporting Information Available: Detailed experimental
procedures for the synthesis of all compounds, characterization
data and 1H and 13C NMR spectra for all new compounds, and
derivation of the integrated rate laws. This material is available
free of charge via the Internet at http://pubs.acs.org.
6198 J. Org. Chem. Vol. 74, No. 16, 2009