Multiple component reactions: An efficient nickel-catalyzed Reformatsky-type reaction and its application in the parallel synthesis of β-amino carbonyl libraries

Article abstract of DOI:10.1021/jo026847q

Multiple-component condensations (MCC) where three or more reactants combine to afford a new core structure possessing the molecular features of its composite building blocks is a powerful method for the preparation of molecular diversity. We have developed an efficient, nickel-catalyzed, Reformatsky-type three-component condensation (3CC) reaction that affords β-amino carbonyl compounds. The scope of the reaction is demonstrated both in the gram and microscale settings; 15 β-amino esters, amides, and a ketone were prepared efficiently at the mmol scale, and a library of 64 β-amino carbonyl compounds was generated at the μmol scale.

Full text of DOI:10.1021/jo026847q

Mu ltip le Com p on en t Rea ction s: An Efficien t Nick el-Ca ta lyzed
Refor m a tsk y-Typ e Rea ction a n d Its Ap p lica tion in th e P a r a llel
Syn th esis of â-Am in o Ca r bon yl Libr a r ies
J ames C. Adrian, J r.*,† and Marc L. Snapper‡
Chemistry Department, Union College, Schenectady, New York 12308, and Department of Chemistry,
Merkert Chemistry Center, Boston College, Chestnut Hill, Massachusetts 02467
adrianj@union.edu
Received December 13, 2002
Multiple-component condensations (MCC) where three or more reactants combine to afford a new
core structure possessing the molecular features of its composite building blocks is a powerful method
for the preparation of molecular diversity. We have developed an efficient, nickel-catalyzed,
Reformatsky-type three-component condensation (3CC) reaction that affords â-amino carbonyl
compounds. The scope of the reaction is demonstrated both in the gram and microscale settings;
15 â-amino esters, amides, and a ketone were prepared efficiently at the mmol scale, and a library
of 64 â-amino carbonyl compounds was generated at the µmol scale.
In tr od u ction
The classic paradigm for drug discovery involves the
iterative screening of potential drug candidates against
biological targets, followed by the sequential optimization
of lead compounds through systematic modifications to
the core structure. The advent of high-throughput bio-
logical assays and the ability to evaluate large numbers
of compounds in parallel, however, has placed a severe
strain on this iterative approach. A complementary
strategy and recognized solution to the synthetic limita-
tions of the classical strategy is the use of multiple-
component condensations (MCC) to provide the requisite
small molecule diversity more efficiently.1 Many unique
structures can be afforded rapidly when three or more
reactants are combined in a single step to afford new
compounds possessing the combined features of the
building blocks. One of many examples of a three-
component condensation (3CC), the Biginelli reaction,
brings together ureas, aldehydes and â-ketoesters, to
afford functionalized pyrimidinones as the core structure
(eq1).2 The combinatorial nature of this reaction allows
for the preparation of numerous compounds from a
relatively few building blocks. In this regard, the MCC
strategy can be a valuable tool for the preparation of large
libraries of compounds based on a common core structure
whose diversity will be proportional to the number and
availability of inputs.
to the high functional group tolerance of organozinc
reagents.5 Reformatsky reagents are prepared easily,
usually in situ, by the reaction of an R-halo ester with
some form of activated zinc metal. The position of the
halogen determines precisely the site of zinc insertion
allowing for regioselective enolate formation in polycar-
bonyl compounds.4 The activated zinc reagents, however,
must be prepared fresh, and often the generation of the
most active metals can be laborious. In light of these
issues, it is surprising that little attention has been
placed on developing a catalytic version of the classic
Reformatsky reaction.
About 10 years ago, Pe´richon and co-workers published
a series of articles where they disclosed a new electro-
chemical Reformatsky reaction in which they reported
that the yields of the â-hydroxy ester products were
markedly increased by the addition of sub-stoichiometric
amounts of Ni(II)Br2(2,2′-bipyridine).6
Since its discovery over 115 years ago, the Reformatsky
reaction has found wide use in synthesis3,4 due, in part,
(4) For recent reviews of the Reformatsky reaction, see: (a) Fu¨rstner,
A. Synthesis 1989, 571-590. (b) Rathke M. W.; Weipert, P. In
Comprehensive Organic Synthesis; Trost, B. M., Fleming, I., Eds.;
Pergamon Press: New York, 1991; Vol. 2, pp 277-299. (c) Fu¨rstner,
A In Organozinc Reagents; Knochel, P., J ones, P., Eds.; Oxford
University Press: New York, 1999; pp 287-305.
(5) (a) Inoue, S.; Imanaka, Y. J . Organomet. Chem. 1972, 35, 1-7.
For leading references, see: (b) Knochel, P.; Singer, R. Chem. Rev.
1993, 93, 3. 2117-2188. (c) Boudier, A.; Bromm, L. O.; Lotz, M.;
Knochel, P. Angew. Chem., Int. Ed. 2000, 39, 4414-4435.
† Union College.
‡ Boston College.
(1) (a) Armstrong, W. R.; Combs, A. P.; Tempest, P. A.; Brown, D.
B.; Keating, T. A. Acc. Chem. Res. 1996, 29, 123-131. For leading
references, see: (b) Kobayashi, S. Curr. Opin. Chem. Biol. 2000, 4,
338-345. (c) Willi, B. Methods Princ. Med. Chem. 2000, 9, 6-21.
(2) Biginelli, P. Ber. Dtsch. Chem. Ges. 1891, 24, 2962.
(3) Reformatsky, S. Ber. Dtsch. Chem. Ges. 1887, 20, 1210.
10.1021/jo026847q CCC: $25.00 © 2003 American Chemical Society
Published on Web 02/06/2003
J . Org. Chem. 2003, 68, 2143-2150
2143

Adrian and Snapper
SCHEME 1
In 1992, Ding and Zhao reported that the reaction of
ethyl bromoacetate (and propanoate) with various alde-
hydes and ketones in the presence of zinc dust and
catalytic amounts of Cp2TiCl2 afforded cleanly the classic
Reformatsky â-hydroxy ester products in good to excel-
lent yields.7 However, the authors propose that the active
nucleophile is actually a titanium enolate. This hypoth-
esis is supported by the observation that only 0.5 equiv
of zinc metal is required to achieve high levels of
conversion.
Recently, Honda and colleagues reported a rhodium-
catalyzed Reformatsky-type reaction in which diethylzinc
acted as the zinc source.8 While the overall yields of the
â-hydroxy esters for this transformation were modest, the
development of this type of catalytic system represents
a significant advancement in the Reformatsky reaction.
Similarly, Dolbier and co-workers reported that the
reaction of ethyl bromofluoroacetate with several alde-
hydes and ketones in the presence of zinc dust and a
catalytic amount of CeCl3‚7H2O afforded a limited num-
ber of R-fluoro-â-hydroxy ester products in good to
excellent yields.9
That imines can be substituted for the classic Refor-
matsky partners, aldehydes or ketones, was first dis-
closed by Gilman and Speeter nearly 60 year ago.10
However, this transformation can be problematic, often
affording a mixture of â-amino esters and â-lactams.11
Recently, we have shown that imines prepared from
2-methoxyaniline, such as 1, afford the â-amino esters
as the sole products.12
Despite the high functional group tolerance of the
Reformatsky reagent and convenience of in situ genera-
tion of the nucleophile and imine, there remains a
surprising dearth of MCC examples; we are aware of only
three.13 In all cases, the yields have been modest and
required the use of a potentially explosive reagent13b or
a fairly expensive catalyst.13c To date, none of these
methods demonstrate sufficient synthetic efficiency to be
applied effectively to the preparation of a combinatorial
library of Reformatsky products.
We wish to disclose our results in this area: the
development of an efficient, nickel-catalyzed Reformatsky-
type three-component condensation that combines an
aldehyde, aniline, and an R-bromocarbonyl to afford
compounds with a â-amino carbonyl core structure.14
TABLE 1. Com p a r ison of Ca ta lytic Activity
esterb ethyl adductb convnb timec
entry
catalysta
3a (%)
4 (%)
(%)
(h)
1
2
RhCl(PPh3)3
Pd(dba)2
98
71
2
29
100
100
100
100
100
100
80
100
60
95
100
100
≈20
≈20
≈20
≈20
2.0
0.5
0.5
0.5
0.5
0.5
2.0
2.0
24
24
2.0
2.0
24
3
4
5
PdCl2(PPh3)2
61
>95
99
39
<5
1
d
Ni(acac)2
e
Ni(acac)2
6
7
8
NiCl2(PPh3)2
Cr(acac)3
Cr(acac)3
99
50
50
1
50
50
e
9
Fe(acac)2
Fe(acac)2
Co(acac)3
Co(acac)3
Cr(CH3O2)2
Fe(CO)5
Cu(CF3SO3)2
ZrCp2Cl2
60
55
55
45
<1
<1
<1
<1
40
45
45
55
>99
>99
>99
>99
e
10
11
12
13
14
15
16
e
24
24
24
a
b
Catalyst loading: 5 mol %. Based on 1H NMR analysis of
reaction mixtures directly after workup. c Time of reaction quench,
d
based on TLC determination of conversion or after 24 h. After 5
min, a black precipitate, presumably nickel metal, was observed.
e 10 mol % of PPh3 was added to the reaction as a co-ligand.
Herein, we demonstrate the scope of this new nickel-
catalyzed process at both macro- and microscales. We
prepared 15 compounds at the mmol scale with typical
yields of g90%, and we conducted a parallel synthesis of
a library consisting of 64 â-amino carbonyl compounds
at a µmole scale.
Resu lts a n d Discu ssion
Ca ta lyst Activity. Our primary objective was to
improve the efficiency and expand the scope and utility
of the Reformatsky-type addition of ester, amide, and
ketone enolates to imines.4 Our previous experience also
indicated that Reformatsky additions to imines were
significantly more facile when conducted in CH2Cl2.12
Thus, we began our efforts to improve further the
transformation by establishing a metal-catalyzed reaction
of imine 1 and methyl bromoacetate (2) (Scheme 1). In
this regard, a catalytic Reformatsky reaction using
Wilkinson’s catalyst and diethylzinc in CH2Cl2 afforded
the â-amino ester 3a (Table 1, entry 1). The initial success
was encouraging, but three observations gave us pause.
First, the color of the product was nearly black and
required several purification cycles (in this case, recrys-
tallization), which afforded, at best, a gray crystalline
material and resulted in a significant reduction in the
isolated yield (61% from crude yield of 96%). Second,
while Honda and co-workers did not report the formation
(6) (a) Sibille, S.; d’Incam, E.; Leport, L.; Massebiau, M.-C.; Pe´richon
J . Tetrahedron Lett. 1987, 28, 55-58. (b) Mcharek, S.; Sibille, S.;
Ne´de´lec, J .-Y.; Pe´richon J . J . Organomet. Chem. 1991, 401, 211-215.
(c) Conan, A.; Sibille, S.; Pe´richon J . J . Org, Chem. 1991, 56, 2018-
2024.
(7) Ding, Y.; Zhao, G. J . Chem. Soc., Chem. Commun. 1992, 941-
942.
(8) Kanai, K.; Wakabayashi, H.; Honda, T Org. Lett. 2000, 2, 2549-
2551.
(9) Ocampo, R.; Dolbier, W. R.; Abboud, K. A.; Zuluaga, F. J . Org.
Chem. 2002, 67, 72-78.
(10) Gilman, H.; Speeter, M. J . Am. Chem. Soc. 1943, 65, 2255-
2256.
(11) Dardoize and co-workers reported that â-amino esters and
lactams were formed in ratios that were temperature dependent. For
leading references, see: Dardoize, F.; Moreau, J .-L.; Gaudemar, M Bull.
Soc. Chim. Fr. 1972, 3841-3846.
(12) Adrian, J . C., J r.; Barkin, J . L.; Hassib, L. Tetrahedron Lett.
1999, 40, 2457-2460.
(13) (a) Nishiyama, T.; Kishi, H.; Kitano, K.; Yamada, F. Bull. Chem.
Soc. J pn. 1994, 67, 1765-1768. (b) Saidi, M. R.; Azizi, N.; Zali-Boinee,
H. Tetrahedron 2001, 57, 6829-6832. (c) Honda, T.; Wakabayashi, H.;
Kanai, K. Chem. Pharm. Bull. 2002, 50, 307-308.
2144 J . Org. Chem., Vol. 68, No. 6, 2003

Nickel-Catalyzed Multicomponent Reformatsky Reaction
SCHEME 2
of an ethyl adduct, we observed the ethyl adduct 4 as a
minor component (2%).15 This difference in reactivity may
be due to differences in reaction times. Honda and co-
workers quench their reactions with carbonyl electro-
philes after about 5 min8 and with in situ generated
imines after about 10 min.13c,16 In contrast, our imine
required 2 h to achieve 100% conversion as determined
by TLC analysis.17 Third, and perhaps most significantly,
was the relatively high cost of Wilkinson’s catalyst.
Given these above-mentioned concerns, we decided to
evaluate other metal salts for potential catalytic activity
in this reaction system. The results of this study are
presented in Table 1; it is clear from these data that
several metals catalyze the transformation. We focused
our attention on those salts that afforded 100% conver-
sion in the shortest time (i.e., Pd and Ni salts). The two
palladium catalysts, Pd(dba)2 and PdCl2(PPh3)3 (entries
2 and 3), both lead to complete conversion within 30 min;
however, despite attempts at further optimization, these
catalysts also generated a significant fraction of the
undesired ethyl adduct 4.
problem. A pleasant surprise was that the inclusion of
the co-ligand also reduced significantly the formation of
4. For practical considerations, we decided to try NiIICl2-
(PPh3)2 (entry 6). We were delighted to find that it was
extremely active, there was no appreciable black con-
taminate, and most important, the ethyl adduct 4 was
1
barely detectable by H NMR.
To shift further the unavoidable competition between
the two in situ nucleophiles, the zinc enolate and the
alkyl zinc, in favor of the addition of the zinc enolate, we
switched from diethylzinc to the considerably less nu-
cleophilic dimethylzinc.5 While for the most part this was
a cosmetic change, it turned out to be not without
consequence. Since this reaction worked well with the
methyl bromoacetate 2, we were curious to see if it was
equally reactive with the methyl chloroacetate 6. For
example, Pe´richon and co-workers conducted their nickel-
catalyzed electrochemical Reformatsky reactions with
R-chloro esters or nitriles.6 The results of this experiment
are presented in Scheme 2. The methyl chloroacetate
proved to be active only when diethylzinc was used as
the zinc source and was completely inert when dimeth-
ylzinc was used. It is important to note that while no
Reformatsky adduct was formed when dimethylzinc was
used, within the time frame of the experiment (vida infra)
no methyl adduct 5 was detected either.
With these results in hand, we next turned our
attention to the effect of catalyst loading of the reaction
of imine 1 with methyl bromoacetate 2. The results of
these experiments are presented in Table 2. These data
indicate that optimum catalyst loading for this process
is 4-5 mol %. At this level of catalyst loading, there is
no detectable methyl adduct 5 formed. It is not surprising
that as the catalyst loading decreases there is a concomi-
tant increase in the ratio of methyl adduct and the time
required to achieve 100% conversion. Note that at zero
catalyst loading, the reaction (by this we mean the
formation of methyl adduct) attains only 80% conversion
within 3 days, thus confirming the importance of the
nickel catalyst.
Consistent with the observations of Pe´richon and co-
workers, the nickel catalysts (entries 4-6) were ex-
tremely active,6c affording the â-amino ester 3a as the
major product in all cases. Initially, we found that while
NiII(acac)2 alone (entry 4) was an effective catalyst, the
reaction was beset by two problems. First, as the reaction
proceeded, a black precipitate was formed that appeared
to carry through to the product infecting it with a dark
impurity, similar to that observed with Wilkinson’s
catalyst. The second problem was the formation of the
ethyl adduct 4, while less than that observed with the
palladium catalysts by a factor of 6, it still represented
a significant fraction of the product mixture. We reasoned
that the black precipitate observed during the reaction
was metallic nickel that had been reduced as a result of
the catalytic process and that this might be countered
by the addition of a stabilizing ligand to the reaction
mixture. To our satisfaction, the addition of 2 equiv
(relative to catalyst loading, entry 5) of PPh3 to the
reaction mixture nearly eliminated the precipitation
On the basis of a detailed electrochemical study by
Pe´richon et al.6c and a complimentary mechanistic study
by Heathcock and co-workers on ethylzinc enolate forma-
tion from an R-bromo ketone and diethylzinc, a reason-
able mechanistic picture emerges (Scheme 3). The cata-
lytic cycle is initiated by the in situ generation of the
active Ni(0), which then adds oxidatively into the R-halo
carbonyl. After formation of the initial Ni(II) complex,
ligand exchange with the dimethylzinc affords a halozinc
enolate (or methylzinc enolate) and a second Ni(II)
complex. The zinc enolate goes on to react with the imine,
generated in situ (vida infra) or not, which upon hydroly-
sis gives the â-amino carbonyl. The second Ni(II) complex
undergoes a reductive elimination to return the active
Ni(0) catalyst.
(14) The â-amino carbonyl products of the multicomponent Refor-
matsky method reported herein are complimentary to those obtained
from the Mannich reaction. For recent examples of multicomponent
Mannich reactions, see: (a) List, B.; Pojarliev, P.; Biller, W. T.; Martin,
H. J . J . Am. Chem. Soc. 2002, 124, 827-833. (b) Loh, T.-P.; Liung, S.
B. K. W.; Tan, K.-L.; Wei, L.-L. Tetrahedron 2000, 56, 3227-3237. (c)
Takaya, J .; Kagoshima, H.; Akiyama, T. Org. Lett. 2000, 2, 1577-1579.
(d) Akiyama, T.; Takaya, J .; Kagoshima, H. Synlett 1999, 1426-1428.
(15) Based on the relative integration in the 1H NMR of the reaction
mixture after workup.
(16) In their initial paper (ref 8), Honda and co-workers specifically
comment that no ethyl adducts were observed. In their subsequent
paper (ref 13c), they make no comment on the formation of ethyl
adducts.
(17) In this initial case and in all subsequent cases, the reactions
were judged to be complete based on the disappearance of the limiting
aldehyde as determined by TLC. We made no attempt to determine
the relative rate of formation of the ester 3a beyond following the
course of the reaction by TLC.
J . Org. Chem, Vol. 68, No. 6, 2003 2145

Adrian and Snapper
TABLE 2. Effect of Ca ta lyst Loa d in ga
After an additional 15 min, methyl 2-bromoacetate (1.05
equiv) was added followed immediately by a CH2Cl2
solution of the catalyst. The reaction was followed by TLC
and reached 100% conversion within 1.5 h. This order of
addition afforded the desired â-amino ester 3a in a 96%
yield (entry 1). To our delight, the purity of the crude
1
product, as determined by H NMR, was also excellent.
Flushed with success, we explored the scope of this
novel MCC protocol, and the results are presented in
Table 3. In all but one case (entry 13), the isolated yields
are greater than 80%, and in virtually all cases the
product purity out of the reaction workup was sufficient
for preparative purposes without further purification.
We looked initially at the R-bromocarbonyl component
(entries 1-5) and found that esters (entry 1), tert-butyl
ketone (entry 2), and various tertiary amides18 (entries
3-5) were all active participants in the reaction. The
1-bromo-3,3-dimethyl-2-butanone (entry 2) was found to
be the only active ketone. Attempts with the com-
mercially available 2-bromoacetophenone resulted in low
yields of the â-amino ketone. Presumably, this is due to
the undesired reaction of the acetophenone carbonyl with
the enolate.6c,8 This disadvantage is mitigated, however,
by the fact that the Weinreb amide (entry 3) was found
to be an active reaction partner, thus affording access to
a wide array of â-amino ketones as well as aldehydes.19,20
All 2-methoxyaniline derivatives tested (all entries
except 7 and 8) were sufficiently activating and afforded
either the â-amino ester or amide. We also found that
the 2-(N,N-dimethylamino)aniline (entry 7) and 2-amino-
6-methylpyridine (entry 8) were equally reactive, giving
the â-amino esters as the only observed products in both
cases.
catalyst
loading (%)
ester 3a b
methyl adduct
time
(h)
entry
(%)
5b (%)
1
2
3
4
5
6
7
8
0.0
0.1
0.5
1.0
2.0
3.0
4.0
5.0
0
48
72
88
94
99
g99
g99
80c
52
28
12
6
72
40
16
16
4
2
2
1.5
1
e1d
e1d
a
Unless otherwise noted, all reactions reached 100% conversion
as determined by TLC. Determined by 1H NMR of the reaction
b
mixture directly after workup. c Only 80% conversion at quench.
Undetectable in the 1H NMR of the reaction mixture directly
d
after workup.
SCHEME 3. P r op osed Ca ta lytic Cycle
The need for an electronegative group in the ortho
position was clearly demonstrated in the reaction with
imine 7 derived from benzaldehyde and aniline (eq 2).
Imine 7 lacks an electronegative group in the ortho
position and reacted sluggishly, achieving only an 80%
conversion after 10 h as well as affording a mixture of
the ester 8 and lactam 9.21,22 This observation is consis-
tent with our previous observation that imines derived
from 2-methoxyaniline afford the â-amino esters as the
sole products under more traditional Reformatsky condi-
tions.12
Nick el-Ca ta lyzed Refor m a tsk y 3CC. At this point,
we initiated our effort to improve the efficiency further
by converting this reaction to a multicomponent process
(Table 3). In our preliminary experiments, we combined
all reagents used in entry 1 of Table 3, less the catalyst.
After allowing time for imine formation, approximately
30 min, a solution of the Ni(II) catalyst was added. While
these experiments afforded the desired â-amino ester 3a ,
yields never exceeded 60% and the crude product mix-
tures were troubled with impurities. These results sug-
gested that the order of addition might be important.
We reasoned that the imine should be prepared first,
followed by addition of the bromo compound and catalyst.
To test this hypothesis, we stirred the 4-chlorobenzalde-
hyde (1 equiv) together with the 2-methoxyaniline (1.02
equiv) in CH2Cl2. After 30 min, a toluene solution of the
dimethylzinc (3.5 equiv) was added, which served the
dual role as both the zinc source and dehydrating agent.
All aldehydes tested, both enolizable (entries 11, 14,
and 15) and nonenolizable (all other entries of Table 3)
were found to be active. The reaction appears to be
sensitive to steric demand at the R-position, as evidenced
by the significantly lower yield of 2,2-dimethylpropanal
(entry 13). This is particularly apparent when the yields
of the two TBDMS-protected hydroxyaldehydes, 2-hy-
2146 J . Org. Chem., Vol. 68, No. 6, 2003

Nickel-Catalyzed Multicomponent Reformatsky Reaction
TABLE 3. 3CC Nick el-Ca ta lyzed Refor m a tsk y Rea ction Affor d in g Com p ou n d s w ith th e â-Am in o Ca r bon yl Cor e
Str u ctu r ea
a
All reactions were run on a 1 mmol scale in which the aldehyde was the limiting reagent. Typical molar ratios of starting materials:
aldehyde/aniline/bromocarbonyl 1:1.02:1.05. Isolated. c Determined from 1H NMR of the reaction mixture immediately after workup.
b
droxyethanal (entry 11, 81%) and 3-hydroxypropanal
(entry 14, 96%), are compared with that of 2,2-dimeth-
ylpropanal (58%).
Ch em ica l Libr a r y. In light of the high efficiency
achieved, the Ni-catalyzed 3CC Reformatsky reaction is
ideally suited for the parallel synthesis of a combinatorial
library of compounds based on the â-amino carbonyl core
structure. Using a 96-well microtiter plate, a small
library consisting of 64 members (4 aldehydes × 4
R-bromocarbonyls × 4 anilines) was prepared and dis-
tributed such that one product was formed per well
(Figure 1). Inputs for this library were chosen from the
building blocks used in Table 3, representing 54 novel
combinations. Of the possible 64 compounds, 10 had been
prepared previously on scale-up.
(18) We have since determined that secondary amides are also
effective enolate sources. The following example has not been opti-
mized:
The array was conducted on a 25 µmol scale in which
the aldehyde was the limiting reagent. An LCMS-UV
analysis of the reaction wells indicated that in every case
the major, if not only, product was the predicted â-ami-
nocarbonyl compound. In all cases, the M + H signal
observed was that of the expected molecular weight for
(21) The “activating” effect of an electron-withdrawing group at the
ortho position of the N-phenyl group has also been observed for addition
of lithium enolates to imines; see: Saito, S.; Hatanaka, K.; Yamaoto,
H. Org. Lett. 2000, 2, 1891-1894.
(22) Ester and lactam ratio was based on the relative integration
of the â-proton signal (lactam 5.01 ppm, ester 4.84 ppm) in the 1H
NMR of the reaction mixture after workup.
(19) Nahm, S.; Weinreb, S. M.; Tetrahedron Lett. 1981, 22, 3815-
3818.
(20) For reviews on the growing synthetic utility of the Weinreb
amide, see: (a) Sibi, M. P. Org. Prep. Proced. Int. 1993, 25, 15-40. (b)
Singh, J .; Satyamurthi, N.; Aidhen, I. S. J . Prakt. Chem. 2000, 342,
340-347.
J . Org. Chem, Vol. 68, No. 6, 2003 2147

Adrian and Snapper
F IGURE 1. Chemical inputs for a three-dimensional three-component array and example positional decoding for the product
structure Ad 1.
15 min, the methyl bromoacetate (100 µL, 1.05 mmol) was
added, followed immediately by a freshly prepared solution of
bistriphenylphosphine nickel(II) dichloride (2.5 mL, 0.02M,
0.05 mmol) in CH2Cl2. The reaction progress was followed by
TLC. However, unless otherwise noted, all reactions were
quenched after 1-3 h by the addition of aqueous HCl (2.0 mL,
2 M). The organic phase was separated, washed sequentially
with saturated NaHCO3 (10 mL) and brine (10 mL), dried with
MgSO4, filtered, and concentrated in vacuo to afford 302 mg
(96%) of 3a as a yellow-orange oil that crystallizes: mp ) 84-
85 °C; Rf ) 0.11 (SiO2, 40% hexane/CH2Cl2); IR (thin film)
the given components, and in a few cases, an M + Na
signal was also observed.23
Su m m a r y
We have developed a highly efficient, nickel-catalyzed,
Reformatsky-type three-component condensation reaction
that affords â-amino carbonyl compounds as its core
structure. This reaction uses an inexpensive nickel
catalyst and is competent at both gram and microscale
settings. In addition, we have demonstrated its utility
in the parallel generation of molecular diversity by
preparing a 64 â-amino carbonyl compound library. Given
the large number of commercially available aldehydes,
there is significant potential for preparing extremely
large libraries. Adaptation of this reaction to the solid
phase, as well as the development of other multicompo-
nent condensation reactions, are underway and will be
reported in due course.
1
3410.6, 1734.2 cm-1; 300 MHz H NMR δ 7.32 (d, 2H, J ) 9.0
Hz), 7.28 (d, 2H, J ) 9.0 Hz), 6.77 (dd, 1H, J ) 9.0, 3.0 Hz),
6.71 (td, 1H, J ) 9.0, 2.0 Hz), 6.45 (td, 1H, J ) 9.0 Hz, 3.0
Hz), 6.36 (dd, 1H, J ) 9.0, 2.0 Hz), 5.03 (d, 1H, J ) 6.0 Hz),
4.82 (m, 1H), 3.87 (s, 3H), 3.65 (s, 3H), 2.86 (dd, 1H, J ) 15.0,
5.0 Hz), 2.79 (dd, 1H, J ) 15.0, 4.0 Hz); 75 MHz 13C NMR δ
171.1, 146.8, 140.8, 136.2, 132.9, 128.8, 127.6, 121.0, 117.1,
111.1, 109.4, 55.4, 54.0, 51.8, 42.7. Anal. Calcd for C17H18
-
ClNO3: C, 63.85; H, 5.67; N, 4.38. Found: C, 64.12; H, 5.54;
N, 4.17.
1-(2-Meth oxyp h en yl)a m in o-1-(4-ch lor op h en yl)-3,3-d i-
m eth yl-3-p en ta n on e (3b). Compound 3b was isolated as a
brown oil (96%): Rf ) 0.42 (SiO2, 20% hexane/CH2Cl2); IR (thin
Exp er im en ta l Section
1
film) 3415.9, 1702.3 cm-1; 300 MHz H NMR δ 7.31 (d, 2H, J
Gen er a l Con sid er a tion s. Melting points were measured
in open capillary tubes and are uncorrected. Robertson Microlit
Laboratories, Inc. of Madison, NJ performed the elemental
analyses. The Mass Spectrometry Service of the University of
Illinois performed the low- and high-resolution mass spectra.
LCMS analysis was preformed by the Mass Spectrometry
) 8.0 Hz), 7.25 (d, 2H, J ) 8.0 Hz), 6.77-6.60 (m, 4H), 6.37
(dd, 1H, J ) 8.0, 2.0 Hz), 4.97 (s, 1H), 4.88 (m, 1H), 3.87 (s,
3H), 3.04 (dd, 1H, J ) 17.0, 6.0 Hz), 2.95 (dd, 1H, J ) 17.0,
6.0 Hz), 1.02 (s, 9H); 75 MHz 13C NMR δ 212.9, 146.8, 141.6,
136.4, 132.5, 128.6, 127.8, 121.0, 116.9, 111.1, 109.3, 55.3, 53.4,
44.4, 44.2, 25.6; HRMS calcd for C20H24ClNO2 345.1496, found
345.1491.
1
Service of Boston College. H NMR spectra were determined
at 300 MHz, and 13C NMR spectra were determined at 75
MHz. Chemical shifts are expressed in parts per million (δ
units) downfield from tetramethylsilane used as an internal
reference. Dichloromethane (CH2Cl2) was distilled from cal-
cium hydride under N2 immediately prior to use. All other
commercially available reagents and solvents were used
without further purification unless otherwise noted.
3-(2-Meth oxyph en yl)am in o-3-(4-ch lor oph en yl)-N-m eth -
yl-N-m eth oxyp r op a n a m id e (3c). Compound 3c was isolated
as a brown oil (86%): Rf ) 0.11 (SiO2, 20% hexane/CH2Cl2);
1
IR (thin film) 3394.7, 1654.6 cm-1; 300 MHz H NMR δ 7.37
(d, 2H, J ) 9.0 Hz), 7.28 (d, 2H, J ) 9.0 Hz), 6.78 (dd, 1H, J
) 8.0, 2.0 Hz), 6.67 (td, 1H, J ) 8.0, 2.0 Hz), 6.63 (td, 1H, J )
8.0, 2.0 Hz), 6.36 (dd, 1H, J ) 8.0, 2.0 Hz), 5.31 (s, 1H), 4.87
(m, 1H), 3.88 (s, 3H), 3.58 (s, 3H), 3.15 (s, 3H), 3.01 (dd, 1H,
J ) 15.0, 8.0 Hz), 2.92 (dd, 1H, J ) 15.0, 6.0 Hz); 75 MHz 13C
NMR δ 171.2, 146.9, 141.6, 136.5, 132.7, 128.7, 127.8, 120.9,
116.8, 111.1, 109.3, 61.2, 55.4, 54.1, 40.0, 31.9; HRMS calcd
for C18H21ClN2O3 338.1241, found 348.1238.
Gen er a l P r oced u r e. Met h yl 3-(2-Met h oxyp h en yl)-
a m in o-3-(4-ch lor op h en yl)p r op a n oa te (3a ). To a stirred
solution of 4-chlorobenzaldehyde (139 mg, 0.98 mmol) in CH2-
Cl2 (7.5 mL) at room temperature was added 2-methoxyaniline
(115 µL, 1.02 mmol). After 30 min, a solution of dimethylzinc
(1.75 mL, 2M, 3.5 mmol) in toluene was added all at once. After
3-(2-Met h oxyp h en yl)a m in o-3-(4-ch lor op h en yl)-N,N-
d im eth ylp r op a n a m id e (3d ). Compound 3d was isolated as
(23) See the Supporting Information.
2148 J . Org. Chem., Vol. 68, No. 6, 2003

Nickel-Catalyzed Multicomponent Reformatsky Reaction
a brown oil (96%): Rf ) 0.27 (SiO2, 10% diethyl ether/CH2-
103.5, 99.2, 55.6, 53.3, 51.9, 42.4. Anal. Calcd for C17H20N2O4:
C, 64.54; H, 6.37; N, 8.86. Found: C, 64.76; H, 6.21; N, 8.64.
Met h yl 3-(2-Met h oxyp h en yl)a m in o-5-p h en yl-4-p en -
ten oa te (3j). Compound 3j was isolated as a yellow oil (96%):
Rf ) 0.29 (SiO2, 20% hexane/CH2Cl2); IR (thin film) 3410.6,
1734.2 cm-1; 300 MHz 1H NMR δ 7.36-7.20 (m, 5H), 6.87-
6.77 (m, 2H), 6.72-6.68 (m, 2H), 6.57 (d, 1H, J ) 16.0 Hz),
6.21 (dd, 1H, J ) 16.0, 6.0 Hz), 4.67 (s, 1H), 4.51 (m, 1H), 3.87
(s, 3H), 3.69 (s, 3H), 2.81 (dd, 1H, J ) 15.0, 7.0 Hz), 2.72 (dd,
1H, J ) 15.0, 7.0 Hz), 2.13 (s, 3H); 75 MHz 13C NMR δ 171.4,
146.8, 136.4, 130.6, 129.7, 128.3, 127.4, 126.4, 121.0, 116.8,
1
Cl2); IR (thin film) 3384.1, 1644.0 cm-1; 300 MHz H NMR δ
7.36 (d, 2H, J ) 9.0 Hz), 7.27 (d, 2H, J ) 9.0 Hz), 6.75 (dd,
1H, J ) 9.0, 3.0 Hz), 6.69 (td, 1H, J ) 9.0, 3.0 Hz), 6.62 (td,
1H, J ) 9.0, 3.0 Hz), 6.32 (dd, 1H, J ) 9.0, 3.0 Hz), 5.37 (s,
1H), 4.83 (m, 1H), 3.88 (s, 3H), 2.93-2.78 (m, 8H); 75 MHz
13C NMR δ 169.8, 146.9, 141.5, 136.3, 132.5, 128.5, 127.7,
120.7, 116.7, 111.0, 109.2, 55.3, 54.2, 40.6, 37.1, 35.2; HRMS
calcd for C18H21ClN2O2 332.1292, found 332.1283.
N-[3-(2-Met h oxyp h en yl)a m in o-3-(4-ch lor op h en yl)-1-
oxop r op a n e]m or p h olin e (3e). Compound 3e was isolated
as a brown oil which crystallized on standing (94%): mp )
145-146 °C; Rf ) 0.30 (SiO2, 20% diethyl ether/CH2Cl2); IR
(thin film) 3384.1, 1638.7, 1224.9, 1113.4 cm-1; 300 MHz 1H
NMR δ 7.34 (d, 2H, J ) 9.0 Hz), 7.28 (d, 2H, J ) 9.0 Hz), 6.76
(dd, 1H, J ) 8.0, 2.0 Hz), 6.69 (td, 1H, J ) 8.0, 2.0 Hz), 6.63
(td, 1H, J ) 8.0, 2.0 Hz), 6.32 (dd, 1H, J ) 8.0, 2.0 Hz), 5.30
(s, 1H), 4.85 (m, 1H), 3.88 (s, 3H), 3.62-3.52 (m, 4H), 3.44-
3.41 (m, 2H), 3.29-3.26 (m, 2H), 2.92 (dd, 1H, J ) 14.0, 7.0
Hz), 2.80 (dd, 1H, J ) 14.0, 6.0 Hz); 75 MHz 13C NMR δ 168.7,
146.9, 141.4, 136.4, 133.0, 128.9, 127.8, 121.0, 117.0, 111.1,
109.3, 66.7, 66.3, 55.5, 54.5, 46.3, 41.9, 40.2. Anal. Calcd for
110.1, 109.5, 55.2, 552.0, 51.5, 40.2; HRMS calcd for C19H21
NO3 311.1521, found 311.1521.
-
Met h yl
3-(2-Met h oxyp h en yl)a m in o-4-ter t-b u t yld i-
m eth ylsiloxybu ta n oa te (3k ). The reaction was quenched
with saturated ammonium chloride, and 3k was isolated as a
yellow oil (81%): Rf ) 0.37 (SiO2, 20% hexane/CH2Cl2); IR (thin
1
film) 3409.9, 1736.3 cm-1; 300 MHz H NMR δ 6.85 (td, 1H, J
) 8.0, 2.0 Hz), 6.75 (dd, 1H, J ) 8.0, 2.0 Hz), 6.69-6.33 (m,
2H), 4.68 (s, 1H), 3.94 (m, 1H), 3.82 (s, 3H), 3.75 (dd, 1H, J )
10.0, 3.0 Hz), 3.68 (dd, 1H, J ) 10.0, 5.0 Hz), 3.67 (s, 3H),
2.64 (d, 2H, J ) 6.0 Hz), 0.90 (s, 9H), 0.04 (s, 3H), 0.03 (s,
3H); 75 MHz 13C NMR δ 172.6, 147.2, 136.5, 121.2, 116.7,
110.7, 109.7, 63.8, 55.3, 51.5, 51.0, 35.7, 25.7, 18.2, -5.6, -5.5;
HRMS calcd for C18H31NO4Si 353.2022, found 353.2018.
Meth yl 3-(2-Meth oxyp h en yl)a m in o-3-fu r a n ylp r op a n -
oa te (3l). Compound 3l was isolated as a yellow oil which
crystallized on standing (97%): mp ) 40-42 °C; Rf ) 0.22
C
20H23ClN2O3: C, 64.08; H, 6.18; N, 7.47. Found: C, 64.22; H,
6.13; N, 7.46.
Meth yl 3-(2-Meth oxy-5-m eth ylp h en yl)a m in o-3-(4-ch lo-
r op h en yl)p r op a n oa te (3f). Compound 3f was isolated as a
yellow oil (93%): Rf ) 0.20 (SiO2, 40% hexane/CH2Cl2); IR (thin
1
film) 3415.9, 1734.2 cm-1; 300 MHz H NMR δ 7.32 (d, 2H, J
(SiO2, 20% hexane/CH2Cl2); IR (thin film) 3394.7, 1734.2 cm-1
;
) 9.0 Hz), 7.28 (d, 2H, J ) 9.0 Hz), 6.66 (d, 1H, J ) 9.0 Hz),
6.44 (dd, 1H, J ) 9.0, 3.0 Hz), 6.25 (d, 1H, J ) 3.0 Hz), 4.95
(s, 1H), 4.82 (m, 1H), 3.84 (s, 3H), 3.65 (s, 3H), 2.86 (dd, 1H,
J ) 15.0, 8.0 Hz), 2.79 (dd, 1H, J ) 15.0, 6.0 Hz), 2.13 (s, 3H);
75 MHz 13C NMR δ 171.0, 144.8, 140.8, 135.9, 132.8, 130.2,
128.7, 127.6, 117.2, 112.1, 109.4, 55.5, 53.9, 51.7, 42.6, 20.9;
HRMS calcd for C18H20ClNO3 333.1132, found 333.1135.
1
300 MHz H NMR δ 7.33 (d, 1H, J ) 2.0 Hz), 6.85-6.76 (m,
2H), 6.72-6.66 (m, 2H), 6.27 (dd, 1H, J ) 3.0, 2.0 Hz), 6.20
(d, 1H, J ) 3.0 Hz), 5.03 (m, 1H), 4.80 (d, 1H, J ) 9.0 Hz)),
3.84 (s, 3H), 3.67 (s, 3H), 3.00-2.83 (m, 2H); 75 MHz 13C NMR
δ 171.3, 154.5, 147.1, 141.8, 136.2, 121.1, 117.4, 111.1, 110.2,
109.7, 106.2, 55.4, 51.8, 48.6, 20.3. Anal. Calcd for C15H17
-
NO4: C, 64.55; H, 6.22; N, 5.09. Found: C, 65.67; H, 6.35; N,
5.04.
Met h yl 3-(2-N,N-Dim et h yla m in op h en yl)a m in o-3-(4-
ch lor op h en yl)p r op a n oa te (3g). Compound 3g was isolated
as a brown oil (97%): Rf ) 0.25 (SiO2, 20% hexane/CH2Cl2);
Meth yl 3-(2-Meth oxyp h en yl)a m in o-4,4-d im eth ylp en -
ta n oa te (3m ). Compound 3m was isolated as a yellow oil
(58%): Rf ) 0.33 (SiO2, 20% hexane/CH2Cl2); IR (thin film)
1
IR (thin film) 3352.3, 1734.2 cm-1; 300 MHz H NMR δ 7.31
(d, 2H, J ) 9.0 Hz), 7.28 (d, 2H, J ) 9.0 Hz), 7.03 (dd, 1H, J
) 9.0, 3.0 Hz), 6.84 (td, 1H, J ) 9.0, 3.0 Hz), 6.46 (td, 1H, J )
9.0, 3.0 Hz),6.35 (dd, 1H, J ) 9.0, 3.0 Hz), 5.48 (d, 1H, J ) 7.0
Hz), 4.81 (m, 1H), 3.65 (s, 3H), 2.89-2.75 (m, 2H), 2.67 (s, 6H);
75 MHz 13C NMR δ 171.0, 141.1, 141.0, 140.4, 132.8, 128.7,
127.5, 124.4, 119.1, 117.2, 111.1, 54.2, 51.7, 43.9, 43.1; HRMS
calcd for C18H21ClN2O2 332.1291, found 332.1288.
1
3415.5, 1738.4 cm-1; 300 MHz H NMR δ 6.86-6.73 (m, 3H),
6.62-6.57 (m, 1H), 4.26 (d, 1H, J ) 9.0 Hz), 3.83 (s, 3H), 3.76
(m, 1H), 3.53 (s, 3H), 2.64 (dd, 1H, J ) 15.0, 6.0 Hz), 2.36 (dd,
1H, J ) 15.0, 9.0 Hz), 0.96 (s, 9H); 75 MHz 13C NMR δ 173.1,
146.3, 138.3, 121.2, 115.8, 110.5, 109.6, 59.1, 55.5, 51.1, 37.0,
35.7, 26.4; HRMS calcd for C15H23NO3 265.1678, found 265.1684.
Met h yl
3-(2-Met h oxyp h en yl)a m in o-5-ter t-b u t yld i-
Meth yl 3-(3-Meth yl-2-pyr idyl)am in o-3-(4-ch lor oph en yl)-
p r op a n oa te (3h ). The reaction was quenched with saturated
ammonium chloride, and 3h was isolated as a brown oil that
crystallized on standing to afford tan prisms (95%): mp )
101-102 °C; Rf ) 0.33 (SiO2, 20% diethyl ether/CH2Cl2); IR
(thin film) 3394.7, 1734.2 cm-1; 300 MHz 1H NMR δ 7.34-
7.22 (m, 5H), 6.63 (d, 1H, J ) 7.0 Hz), 6.32 (d, 1H, J ) 8.0
Hz), 5.21 (d, 1H, J ) 7.0 Hz), 5.10 (m, 1H), 3.64 (s, 3H), 2.89
(dd, 1H, J ) 15.0, 7.0 Hz), 2.80 (dd, 1H, J ) 15.0, 6.0 Hz),
2.34 (s, 3H); 75 MHz 13C NMR δ 171.1, 156.9, 156.8, 140.4,
137.9, 133.1, 128.8, 127.7, 113.0, 103.6, 52.4, 51.9, 41.9, 24.2.
Anal. Calcd for C16H17ClN2O2: C, 63.05; H, 5.62; N, 9.19.
Found: C, 63.17; H, 5.69; N, 9.25.
m eth ylsiloxyp en ta n oa te (3n ). The reaction was quenched
saturated ammonium chloride, and 3n was isolated as a yellow
oil (96%): Rf ) 0.11 (SiO2, 20% hexane/CH2Cl2); IR (thin film)
3409.4, 1737.9 cm-1; 300 MHz 1H NMR δ 6.85 (td, 1H, J )
8.0, 2.0 Hz), 6.75 (dd, 1H, J ) 8.0, 1.0 Hz), 6.70-6.61 (m, 2H),
4.55 (s, 1H), 4.01 (m, 1H), 3.82 (s, 3H), 3.75-3.68 (m, 2H), 3.65
(s, 3H), 2.68 (dd, 1H, J ) 15.0, 6.0 Hz), 2.54 (dd, 1H, J ) 15.0,
7.0 Hz), 1.92-1.75 (m, 2H), 0.90 (s, 9H), 0.03 (s, 3H), 0.02 (s,
3H); 75 MHz 13C NMR δ 172.3, 146.9, 136.9, 121.2, 116.3,
110.3, 109.5, 60.1, 55.3, 51.5, 47.5, 39.0, 37.3, 25.9, 18.2, 12.5,
-5.5; HRMS calcd for C19H33NO4Si 367.2179, found 367.2169.
Meth yl 3-(2-Meth oxyph en yl)am in o-5-ph en ylpen tan oate
(3o). Compound 3o was isolated as a yellow oil (87%): Rf )
0.11 (SiO2, 1/40/59% diethyl ether/hexane/CH2Cl2); IR (thin
Met h yl 3-(2,4-Dim et h oxyp h en yl)a m in o-3-(3-p yr id yl)-
p r op a n oa te (3i). The reaction was quenched with saturated
ammonium chloride, and 3i was isolated as a yellow oil which
crystallized on standing to afford small tan needles (95%): mp
) 93-94 °C; Rf ) 0.14 (SiO2, 20% diethyl ether/CH2Cl2); IR
(thin film) 3384.1, 1734.2 cm-1; 300 MHz 1H NMR δ 8.63 (s
1H), 8.49 (d, 1H, J ) 6.0 Hz), 7.68 (d, 1H, J ) 9.0 Hz), 7.23
(dd, 1H, J ) 9.0, 6.0 Hz), 6.43 (d, 1H, J ) 3.0 Hz), 6.33 (d, 1H,
J ) 9.0 Hz), 6.25 (dd, 1H, J ) 9.0, 3.0 Hz), 4.84 (s, 1H), 4.69
(m, 1H), 3.85 (s, 3H), 3.70 (s, 3H), 3.66 (s, 3H), 2.90 (dd, 1H,
J ) 15.0, 8.0 Hz), 2.83 (dd, 1H, J ) 15.0, 6.0 Hz); 75 MHz 13C
NMR δ 171.0, 152.5, 148.2, 138.1, 134.3, 130.0, 123.7, 112.0,
1
film) 3405.3, 1728.9 cm-1; 300 MHz H NMR δ 7.30-7.25 (m,
2H), 7.21-7.16 (m, 3H), 6.86 (td, 1H, J ) 8.0, 1.0 Hz), 6.78
(dd, 1H, J ) 8.0, 1.0 Hz), 6.66 (td, 1H, J ) 8.0, 1.0 Hz), 6.60
(dd, 1H, J ) 8.0, 1.0 Hz), 4.30 (s, 1H), 3.89-3.81 (m, 4H), 3.64
(s, 3H), 2.81-2.62 (m, 3H), 2.49 (dd, 1H, J ) 15.0, 7.0 Hz),
2.05-1.85 (m, 2H); 75 MHz 13C NMR δ 172.2, 146.9, 141.5,
136.8, 128.3, 128.2, 125.8, 121.2, 116.5, 110.3, 109.6, 55.3, 51.5,
49.4, 39.3, 36.7, 32.3; HRMS calcd for C19H23NO3 313.1678,
found 313.1679.
Mu lticom p on en t Ar r a y. The array was run using a
standard 96-well plate in a glovebox under an inert atmo-
J . Org. Chem, Vol. 68, No. 6, 2003 2149

Adrian and Snapper
sphere of N2. Vials containing solutions of the aldehydes (0.5
M), anilines (0.52 M), and the R-bromo compounds (0.52 M)
in 1,2-dichloroethane were transferred into the glovebox. Using
a stepping pipet, a 50 µL aliquot of the aldehyde (25 µmol)
followed by a 50 µL aliquot of the aniline (26 µmol) were
combined in the appropriate wells of the plate. The wells were
then covered to minimize evaporation. After 30 min, a 45 µL
aliquot of a solution of Zn(CH3)2 (2 M in toluene) was added
to the aldehyde and aniline mixtures with an eight-tip pipet,
and again the wells were covered. After 15 min, a 50 µL aliquot
of the R-bromo compound solutions (26 µmol) was added to
the wells using the stepping pipet. The Ni catalyst was then
dissolved in CH2Cl2 (0.025 M), and 50 µL of the solution (1.25
µmol) was added to each well using an eight-tip pipet. The
wells were then covered. After 1.5 h, 10 µL aliquots were taken
from six random wells for TLC analysis, all indicated complete
reaction. The 96-well plate was then taken from the glovebox,
and using an eight-tip pipet, the reactions were quenched by
passing them across neutral alumina that had been covered
with a thin layer of activated charcoal. Vacuum was applied
to pull the reaction mixtures through the alumina, which was
subsequently rinsed with a total of 600 µL of acetonitrile (ACN)
per well. Aliquots of the resulting product solutions were
diluted further in ACN for LCMS analysis.
Ack n ow led gm en t. We thank the NSF for an ROA
(CHE-0132221) and the NIH (GM-57212) for support
of this research. Many thoughtful conversations with
Professor Amir H. Hoveyda are gratefully acknowl-
edged.
1
Su p p or tin g In for m a tion Ava ila ble: Representative H
NMR and 13C NMR spectra; UV and total ion chromatograms
and low-resolution mass spectra for all array compounds. This
material is available free of charge via the Internet at
http://pubs.acs.org.
J O026847Q
2150 J . Org. Chem., Vol. 68, No. 6, 2003