Studies on the chemistry and reactivity of α-substituted ketones in isonitrile-based multicomponent reactions

Article abstract of DOI:10.1021/jo8019708

(Chemical Equation Presented) Using the Passerini and Ugi reactions as representative tests, the utility of several α-substituted ketones R-CO-CH₂-X (X = sulfonyloxy, acyloxy, azido, halo, hydroxy, and sulfonyl) in isonitrile-based multicompone

Full text of DOI:10.1021/jo8019708

Studies on the Chemistry and Reactivity of r-Substituted Ketones in
Isonitrile-Based Multicomponent Reactions†
Lijun Fan, Ashley M. Adams, Jason G. Polisar, and Bruce Ganem*
Department of Chemistry and Chemical Biology, Baker Laboratory, Cornell UniVersity,
Ithaca, New York 14853-1301
bg18@cornell.edu
ReceiVed September 8, 2008
Using the Passerini and Ugi reactions as representative tests, the utility of several R-substituted ketones
R-CO-CH2-X (X ) sulfonyloxy, acyloxy, azido, halo, hydroxy, and sulfonyl) in isonitrile-based
multicomponent reactions was explored. In a relative rate study (R ) PhCH2CH2), each of the R-substituted
ketones underwent Passerini condensation more rapidly than the parent ketone. Short, highly convergent
routes to oxazoline, ꢀ-lactam, di-O-acylglyceramides, and other molecular frameworks were developed.
Introduction
The Passerini and Ugi reactions, which represent the two most
widely employed isonitrile-based MCRs, also need simple
carbonyl compounds as integral reactants. As part of our efforts
to synthesize new molecular frameworks by using the “single
reactant replacement” (SRR) strategy to evolve existing MCRs,
we have investigated aldehydes and ketone surrogates that
combine carbonyl electrophilic properties with additional em-
bedded reactivity elements. For example, we recently discovered
that replacing simple carbonyl compounds in the Passerini
reaction with acyl cyanides made it possible to redirect the
outcome to afford ꢀ-aminoacid diamides instead of the con-
ventional R-acyloxyamides,5 and further empowered the as-
sembly of densely functionalized oxazoles and tetrazoles.6
We also investigated the use of R-diazoketones in isonitrile-
based MCRs, finding them resistant to Passerini and Ugi
reactions. However, by exploiting the reactivity of diazocarbo-
nyls to Bronsted acids or transition metal catalysis, we recently
achieved successful one-pot, four-component condensations that
furnished substituted oxazolines via R-sulfonyloxyketones7 as
well as substituted di-O-acylglyceramides from R-acyloxyke-
tones.8 Following those communications, we now report a full
account of our studies using R-substituted ketones as surrogates
for simple carbonyl compounds in the Ugi and Passerini
reactions.
The discovery in 1850 that R-amino acids could be prepared
in one step from simple reactants using the Strecker reaction,
the first reported example of a multicomponent reaction (MCR),
launched a period of intensive synthetic chemical development
that remains active today. Generally defined as one-pot processes
that produce a desired target from the combination of three or
more reactants,1 MCRs can create simple and convergent paths
to structurally complex products in atom-economical2 fashion.
With the advent of combinatorial chemistry and its widespread
adoption by the pharmaceutical industry, MCR chemistry
experienced a renaissance of interest in the past two decades,
triggered initially by the demand for screening libraries in drug
discovery programs, and stimulated more recently by efforts to
assemble natural product-like scaffolds exhibiting specific
pharmacological profiles.
To date, MCR reactions based on the chemistry of isonitriles
have been particularly widely used in these endeavors, despite
the well-known drawbacks to using this family of compounds.
Besides their unpleasant olefactory properties, isonitriles have
generally been limited in their commercial availability and have
limited shelf stability. In recent years, however, versatile new
syntheses of isonitriles have been developed,3 including routes
to convertible isonitriles possessing pleasant aromas,4 that
promise to expand the repertoire of known MCRs.
The rationale for this study was to investigate SRR outcomes
when reactive functionality was embedded at the R-position of
† Dedicated to the memory of the late Professor A. I. Meyers.
(1) Armstrong, R. M.; Combs, A. P.; Tempest, P. A.; Brown, S. D.; Keating,
T. A. Acc. Chem. Res. 1996, 29, 123–131.
(5) Oaksmith, J. M.; Peters, U.; Ganem, B. J. Am. Chem. Soc. 2004, 125,
13606–13607.
(6) Cle´menc¸on, I. F.; Ganem, B. Tetrahedron 2007, 63, 8665–8669.
(7) Fan, L.; Lobkovsky, E.; Ganem, B. Org. Lett. 2007, 9, 2015–2017.
(8) Fan, L.; Adams, A. M.; Ganem, B. Tetrahedron Lett. 2008, 49, 5983–
5985.
(2) Bienayme´, H.; Hulme, C.; Oddon, G.; Schmitt, P. Chem.-Eur. J. 2000,
6, 3321–3329.
(3) Domling, A.; Ugi, I. Angew. Chem., Int. Ed. 2000, 39, 3168–3210.
(4) Pirrung, M. C.; Ghorai, S. J. Am. Chem. Soc. 2006, 128, 11772–11773.
9720 J. Org. Chem. 2008, 73, 9720–9726
10.1021/jo8019708 CCC: $40.75 2008 American Chemical Society
Published on Web 10/22/2008

Substituted Ketones in Multicomponent Reactions
SCHEME 1
FIGURE 1. Relative rates of a prepresentative Passerini condensation
of R-substituted ketones.
prepared from 1a by Cu(acac)2-catalyzed reaction with acetic
acid.12 Chloroketone 5a was obtained from 1a by reaction with
HCl in ether.13
Azidoketones 6 and arylsulfonylketones 7 were both prepared
according to literature methods from the corresponding chlo-
roketones 5. Thus, azidoketone 6a arose from the reaction of
5a with NaN3 in DMF.14 The synthesis of R-sulfonylketone 7a
(Ar ) Ph) was achieved by reaction of 5a with thiophenol and
NaOH in the presence of oxone.15 A sample of the parent
unsubstituted ketone, benzylacetone 8a, was obtained com-
mercially and used as a control compound for the kinetic study.
Relative Rate Study of r-Substituted Ketones in the
Passerini Reaction. Compounds 2a and 4a-8a were subjected
to a representative Passerini condensation with acetic acid and
t-butyl isonitrile (1.0 equiv each) under neat conditions at rt.
Aliquots of each reaction mixture were removed for NMR
analysis, and the progress of the reaction was monitored by the
relative amounts of starting ketone (as judged by integration of
the singlet resonance for the R-methylene carbon) and the
Passerini product (as measured by the AB-quartet for the same
methylene group). The results are tabulated in Figure 1.
As the data indicate, each of the R-substituted ketones
underwent Passerini condensation more rapidly than the parent
carbonyl compound 8a. Mesyloxyketone 2a was the most
reactive carbonyl compound tested. Although tosylate 3a was
not included in the rate study, independent side-by-side com-
parisons of 2a and 3a in other Passerini and Ugi condensations
(Vide infra) established that 3a displayed comparable reactivity
in isonitrile-based MCRs. Chloroketone 5a and azidoketone 6a
showed similar rate profiles, and were somewhat more reactive
than acyloxyketone 4a. Sulfonylketone 7a was the least reactive
of the substituted ketones tested, undergoing Passerini conden-
sation only slightly faster than 8a. The kinetic data in Figure 1
were consistent with the expected enhancement of carbonyl
electrophilicity caused by electronegative substituents, as judged
by known Pauling atom or group electronegativity values for
each of the substituents tested.16
the ketone. While several reactive R-substituents were of
potential interest (sulfonyloxy, acyloxy, azido, halo, hydroxy
and sulfonyl), almost nothing was known about their MCR
chemistry, let alone relative reactivities. Prior to our own
oxazoline-forming synthesis, there were no reports of Passerini
or Ugi condensations on R-mesyloxy- or tosyloxyketones.
Likewise, no examples of MCRs of R-azido-, R-hydroxy-, or
R-sulfonylketones have been described in the literature. Enan-
tioselective, copper-catalyzed Passerini reactions of alkoxy-
substituted aldehydes have recently been described,9 but no
examples of the corresponding alkoxyketones were included.
Alpha-chloroketones have been reported to undergo the Passerini
condensation, forming chloro-substituted acyloxyamides, which
after subsequent base treatment were transformed into the
corresponding acyloxy-ꢀ-lactams.10
It occurred to us that any pronounced rate differences between
these various substituted ketones might make it possible to
achieve selectivity when using mixtures of two or more carbonyl
compounds, thus broadening the opportunities for incorporating
novel functionality in MCRs. Anchimeric effects involving the
neighboring R- substituent might also be exploited to produce
new molecular frameworks. Here we report the relative rates
of condensation of a representative series of R-substituted
ketones in the Passerini condensation. That kinetic study
provides useful new insights into the relative reactivity of each
family and suggests additional directions for new work.
Results and Discussion
Preparation of Substrates. The starting material for all
R-substituted ketones was the corresponding R-diazoketone 1
(Scheme 1), which could be conveniently prepared by reaction
of the cognate acid chloride with diazomethane. Initially, the
readily available 1-diazo-4-phenyl-2-butanone 1a (prepared from
hydrocinnamoyl chloride) was used to synthesize a representa-
tive example of each family of substituted ketones. Reaction
of 1a with methanesulfonic acid or toluenesulfonic acid
furnished the mesyloxy and tosyloxyketones 2a and 3a, respec-
tively.11 The corresponding acyloxyketone 4a (R2 ) CH3) was
Results with compounds 2a, 5a, and 6a suggested that
selective Passerini condensations on these substances might be
possible in the presence of a sulfonyl-substituted or unsubstituted
ketone. As a preliminary test of this hypothesis, a 1:1 mixture
of chloroacetone 5b (R1 ) CH3) and phenylsulfonylacetone 7b
(12) Shinada, T.; Kawakami, T.; Sakai, H.; Takada, I.; Ohfune, Y. Tetra-
hedron Lett. 1998, 39, 3757–3760.
(9) Andreana, P. R.; Liu, C. C.; Schreiber, S. L. Org. Lett. 2004, 6, 4231–
4233.
(10) Sebti, S.; Foucaud, A. Synthesis 1983, 546–549.
(11) Mesylates: (a) Nogrady, T.; Doyle, T. W.; Morris, L. J. Med. Chem.
1965, 8, 656–659Tosylates: (b) Ogawa, K.; Terada, T.; Muranaka, Y.; Hamakawa,
T.; Ohta, S.; Okamoto, M.; Fujii, S. Chem. Pharm. Bull 1987, 35, 3276–3281.
(13) Ames, A. F.; Ames, D. E.; Coyne, C. R.; Grey, T. F.; Lockhart, I. M.;
Ralph, R. S. J. Chem. Soc. 1959, 3388–3400.
(14) Mckervey, M. A.; O’Sullivan, M. B.; Myers, P. L.; Green, R. H. J. Chem.
Soc., Chem. Commun. 1993, 94–96.
(15) Davis, R. Syn. Commun. 1987, 17, 823–827.
(16) Huheey, J. E. J. Phys. Chem. 1966, 70, 2086–2092.
J. Org. Chem. Vol. 73, No. 24, 2008 9721

Fan et al.
TABLE 2. Synthesis of Acyloxy-ꢀ-lactams 11 from Sulfonates 9
and 10
SCHEME 2
sulfonate
product (yield)
9c
11a (62%)
11b (46%)
11c (59%)
11d (64%)
11e (56%)
10a
10b
10c
10d
SCHEME 4
TABLE 1. Passerini Reactions of Mesyloxy- and Tosyloxyketones
mesylate/tosylate
R2CO2H
R3NC
product
(yield)
R1 )
R2 )
R3 )
2a Ph(CH2)2
2a
2a
2b n-C7H15
2c cyclo-C6H11
2d Ph
3a Ph(CH2)2
3a
3b n-C5H11
3b
3c cyclo-C6H11
3d Ph
CH3
Ph
CH3
CH3
Ph
CH3
Ph
Ph
CH3
Ph
CH3
Ph
t-Bu
9a (88%)
9b (82%)
9c (80%)
9d (91%)
9e (80%)
9f (18%)
10a (99%)
10b (97%)
10c (97%)
10d (99%)
10e (80%)
10f (28%)
cyclo-C6H11
cyclo-C6H11
cyclo-C6H11
n-Bu
cyclo-C6H11
t-Bu
cyclo-C6H11
cyclo-C6H11
cyclo-C6H11
t-Bu
comparable whether from the mesylate or tosylate precursor,
the cyclization was noticeably more rapid in the case of mesylate
9c.
cyclo-C6H11
We also investigated Ugi reactions of sulfonyloxyketones 2
and 3. When using ammonia as the amine, the MCR led to a
novel 4-component synthesis of substituted 2-oxazolines 13
(Scheme 4) by way of the intermediate diamides 12.7 Apparently
the formation of imines derived from 2 or 3 was faster than
ammonolysis of the mesylate or tosylate group, thus triggering
the desired Ugi reaction to form 12. The fact that 12 cyclized
spontaneously to 13 during the reaction finds precedent in the
known biosynthesis of peptide-based oxazolines17 by cyclization
of activated N-acylserine derivatives.18
Initial studies of this interesting transformation with 2a as a
representative mesyloxyketone utilized methanol as solvent, a
common medium for Ugi condensations, and also afforded
significant quantities of the corresponding Passerini product 9.
This byproduct could be effectively suppressed by replacing
methanol with 2,2,2-trifluoroethanol, a solvent that was recently
reported to improve Ugi reactions involving ammonia as the
amine component.19 Overall, the new oxazoline synthesis, for
which full experimental details on 14 examples have been
reported,7 accommodated a broad range of sulfonyloxyketones,
isonitriles and carboxylic acids.
Our initial communication noted that when NH3 was replaced
with benzylamine, the MCR of 2a with acetic acid and
cyclohexyl isonitrile afforded aminoester 15a (Scheme 5) as
the exclusive product. Particularly characteristic was the strong
ester carbonyl band (1740 cm-1) in the IR spectrum of 15a.
A mechanism of formation of 15a paralleling that of 13 could
be envisioned in which hydrolysis of the oxazolinium intermedi-
ate 14a led to the observed product. Interestingly, 15a did not
undergo 1,2-acyl shift, leading to the corresponding amidoal-
cohol 16a. Further confirmation of the structure of 15a was
obtained by single crystal X-ray diffraction analysis.
SCHEME 3
(R1 ) CH3) were reacted with 1 equiv each of acetic acid and
t-butyl isonitrile (24 h, rt) to afford an 8:1 ratio of products
from 5b and 7b, respectively.
Isonitrile-Based MCRs of Substituted Ketones. Mesyloxy
and tosyloxyketones 2 and 3 proved to be very good carbonyl
components in the Passerini reaction, typically forming the
expected products 9 and 10, respectively, in high yields
(Scheme 2).
The results of a representative sampling of condensations
using various carboxylic acids and isonitriles (1.1 equiv each)
are summarized in Table 1.
Condensations of 2 and 3 were generally completed in 12-18
h and successfully furnished the expected products in very good
to excellent yields. Products 9 and 10 were readily purified by
flash column chromatography. Furthermore, the possibility of
directly converting the parent R-diazoketone to mesylate 9a was
realized in a simple one-pot process by reacting a CH2Cl2
solution of diazoketone 1a with methanesulfonic acid (1.1 equiv,
10 min, 0 °C). After addition of NaOAc (0.5 equiv) to neutralize
residual acid, the solution was treated with acetic acid and
cyclohexyl isonitrile (rt 24 h) to afford 9a in 86% yield.
With their reactive sulfonate leaving groups, sulfonates 9 and
10 were also well-suited precursors for the synthesis of
substituted ꢀ-lactams 11 (Scheme 3) following a procedure
developed for cyclizations of the corresponding chloro com-
pounds.10
Table 2 summarizes results obtained in our laboratory using
this approach to substituted ꢀ-lactams from representative
sulfonates 9 and 10. While the yields of 11a-e were generally
Structures such as 16a embody the hydroxyethylene or
hydroxyethyl structural unit found in transition state isosteres
(17) Review: Wipf, P. Alkaloids: Chemical and Biological PerspectiVes;
Pelletier, S. W., Ed.; Pergamon: New York, 1998; pp 187-228.
(18) For a recent summary of synthetic methods, see: Benito, J. M.;
Christensen, C. A.; Meldal, M. Org. Lett. 2005, 7, 581–584.
(19) Kazmaier, U.; Hebach, C. Synlett. 2003, 1591–1594.
9722 J. Org. Chem. Vol. 73, No. 24, 2008

Substituted Ketones in Multicomponent Reactions
SCHEME 5
SCHEME 7
SCHEME 8
TABLE 3. Ugi Reactions of Mesyloxyketones 2 with Primary
Amines
mesylate 2;
R1 )
R2CO2H
R3NH2
R4NC
product
(yield)
R2 )
R3 )
R4 )
2a Ph(CH2)2
2a
2b n-C5H11
2b
CH3
i-Pr
Ph
i-Pr
CH3
i-Pr
CH3
Bn
allyl
Bn
allyl
i-Pr
n-Bu
Bn
cyclo-C6H11
cyclo-C6H11
t-Bu
cyclo-C6H11
cyclo-C6H11
t-Bu
15a (71%)
15b (66%)
15c (67%)
15d (61%)
9e (70%)
15e (39%)
15f (74%)
TABLE 4. Four-Component Condensations of Diazoketones 1
Leading to Di-O-acylglyceramides 20
R2CO2H
R3NC
R4CO2H
product
(yield)
1 R1 )
R2 )
R4 )
R4 )
2b
2c cyclo-C6H11
2d CH3
1a Ph(CH2)2
1a
1a
1b n-C5H11
1b
1c cyclo-C6H13 C7H15
1c
CH3
CH3
Ph
(CH3)2CH
Ph
t-Bu
t-Bu
EtO2CCH2 C7H15
t-Bu
cyclo-C6H11 CH3
n-Bu CH3
EtO2CCH2 Ph
n-Bu
CH3
(CH3)2CH
20a (94%)
20b (90%)
20c (72%)
t-Bu
CbzNHCH2 20d (70%)
SCHEME 6
20e (80%)
20f (70%)
20g (76%)
NCCH2
(CH3)2CH
H(CH2O-CH2)3 t-Bu
1d CH3
1d
CbzNHCH2 20h (71%)
C7H15 20i (58%)
In an alternative approach (Scheme 7), the known21 hydrox-
yketone 19 was subjected to the Ugi reaction with benzylamine,
acetic acid and cyclohexyl isonitrile. Instead of the expected
amidoalcohol 16a, the sole product of that reaction (56%) was
aminoester 15a. Taken together, the results depicted in Schemes
6 and 7, while not conclusive, were consistent with the rapid
conversion of 16a to 15a and its hydrolysis product 18a.
It became apparent from efforts to prepare 16a that Ugi
condensations of acyloxyketones 4, while feasible, suffered
certain limitations. Therefore, we decided to study the reactivity
of acyloxyketones in the corresponding Passerini reactions. A
recently published preliminary report summarized our findings
in this area.8 Since acyloxyketones of general structure 4 were
easily prepared by copper-catalyzed insertion of diazoketone 1
into a carboxylic acid,22 we were able to develop a versatile
one-pot 4-component condensation in two stages leading to a
new family of structurally diverse di-O-acylglyceric acid amides
20 (Scheme 8).
such as statine or norstatine, which have been developed as
inhibitors of retroviral proteases.20 Given their potential bio-
medical significance, we have investigated this MCR further.
As indicated in Table 3, several other primary amines and
mesyloxyketones formed the corresponding amidoesters 15. In
each case, only the amidoester isomer 15 (and none of the
diamide 16) was detected.
To determine whether the products observed in each case
were formed under kinetic control, aminoester 15a was subjected
to basic (DBU-CH3CN, Et3N, NaH-DMF) equilibration condi-
tions. However, no evidence for the formation of diamide 16a
was obtained from these experiments.
The complementary equilibration experiment required an
authentic sample of 16a, which we attempted to synthesize from
acetoxyketone 4a in two steps using the Ugi reaction as shown
below (Scheme 6). Little is known about Ugi condensations of
such acyloxyketones. In fact, reaction of 4a with ammonia,
acetic acid, and cyclohexyl isonitrile returned mostly starting
ketoacetate 4a, along with the corresponding Passerini product
(not shown) in 14% yield. However, replacing ammonia with
benzylamine did afford the desired MCR product 17a in 67%
yield.
The initial insertion stage required heating of 1 with R2CO2H
(30-60 min, toluene, 60-90 °C), usually in the presence of
1.0 mol-% Cu(acac)2. Once gas evolution ceased, the toluene
was removed in vacuo and the crude acyloxyketone 4 was
subjected to a Passerini condensation with R3NC and R4CO2H.
Table 4 summarizes the range of new glyceramide diesters that
could be synthesized using this method.
One advantage of the method over stepwise acylation of the
Vic-diol moiety in glyceramides is that it avoids the potential
Treatment of 17a with NH3-CH3OH was expected to furnish
16a by methanolysis of the more reactive acetate ester. Instead,
two products were formed: aminoester 15a (20%) and aminoal-
cohol 18a (60% yield). Unfortunately, none of the hoped-for
amidoalcohol 16a could be detected by IR.
(20) Sawant, R. L.; Bhatia1, M. S.; Sawant, M. R.; Wadekar, J. B. Curr.
Trends Biotechnol. Pharm. 2008, 2, 133–141.
(21) Muthusamy, S.; Babu, S. A.; Gunanathan, C.; Jasra, R. V. Tetrahedron
Lett. 2001, 42, 5113–5116.
(22) Shinada, T.; Kawakami, T.; Sakai, H.; Takada, I.; Ohfune, Y. Tetra-
hedron Lett. 1998, 39, 3757–3760.
J. Org. Chem. Vol. 73, No. 24, 2008 9723

Fan et al.
SCHEME 9
SCHEME 11
SCHEME 12
SCHEME 10
SCHEME 13
for ester interchange by competing O,O-acyl transfer side
reactions. Moreover, by using one hydrophobic and one
hydrophilic carboxylic acid (e.g., 20i), the approach depicted
in Scheme 8 can be used to synthesize facially amphiphilic
glyceramides for potential biomedical use, whether as cell-wall
disrupting agents displaying antifungal activity or as encapsulat-
ing agents in micellar drug delivery. The method can be adapted
to introduce different chain lengths in each ester group to fine-
tune desired hydrophobic/hydrophilic properties.
The use of R-chloroketones 5 in isonitrile-based multicom-
ponent reactions was also of interest in this investigation. Since
Passerini reactions of 5 have already been described,10 we turned
our attention to Ugi 4-component condensations of chloroac-
etone 5b (R ) CH3) as a representative chloroketone using
conditions optimized for the corresponding mesyloxyketones
2. Thus, reaction of 5b with ammonium acetate (4 equiv) and
t-butyl isonitrile (1.1 equiv, rt, 24 h) in trifluoroethanol solvent
afforded the normal Ugi product 21a in very good yield.
Likewise, condensation of 2b with NH4OAc and cyclohexyl
isonitrile also afforded the expected Ugi product 21b, along with
small quantities of the corresponding Passerini product 22.
Using benzylamine instead of ammonia, the Ugi condensation
of 5b took a different path, forming aminoester 15f (Scheme
10) in 74% yield, along with a small amount of the correspond-
ing Passerini product (14%, structure not shown). As depicted
in Scheme 10, the formation of aminoester 15f likely proceeded
by an initial Ugi reaction leading to 23, followed by a
rearrangement via 14f similar to that invoked for related
structures 15a-e (Scheme 5).
embody the key structural element of penicillins, cephalosporins,
and other ꢀ-lactam antibiotics at the forefront of modern
antimicrobial chemotherapy.23 Not surprisingly, however, ex-
posure of 21a to mild base instead formed the known17
oxazoline 13a.
We reasoned that replacing the acetamide group in 21a with
a phthalimide would block the undesired formation of an
oxazoline. Moreover, if the Ugi condensation of 5b using
phthalic acid in combination with ammonia and t-butylisonitrile
were successful, the initally formed phthalimidic acid 25
(Scheme 12) might spontaneously dehydrate to the phthalimide
26 either during workup or upon standing, thus paving an
approach to the desired ꢀ-lactam 27.
After testing the solubility of ammonium phthalate in various
solvents, we chose methanol as the solvent for the Ugi reaction
depicted in Scheme 11. Chloroketone 5b was added to a
suspension of ammonium phthalate (4 equiv, 0 °C, 10 min),
followed by t-butyl isonitrile (1.1 equiv, rt. 16 h). Extractive
workup afforded acylaziridine 28 (Scheme 13) in 33% yield.
However, when the Ugi reaction was repeated using the less
nucleophilic trifluoroethanol as solvent, only traces of phthal-
imide 26 were detected by mass spectrometry. Instead, the major
product was hydroxyamide 29, apparently the result of a
Chloroketone-derived Ugi products like 21a and 21b would
be particularly useful if they underwent cyclization to the
corresponding acylamino-ꢀ-lactams 24 (Scheme 11), which
(23) Buynak, J. D. Biochem. Pharmacol. 2006, 71, 930–940.
9724 J. Org. Chem. Vol. 73, No. 24, 2008

Substituted Ketones in Multicomponent Reactions
SCHEME 15
TABLE 5. Passerini Reactions of r-Sulfonylketones 7
R2CO2H
R3NC
product
(yield)
7
R2 )
R3 )
7a R1 ) Ph(CH2)2 Ar ) Ph
CH3
CH3
Ph
t-Bu
n-Bu
t-Bu
32a (46%)
32b (40%)
32c (50%)
7b R1 ) CH3 Ar ) Ph
FIGURE 2. ORTEP diagram of the X-ray crystal structure of 29.
7b
7b
(CH3)2CH cyclo-C6H11 32d (69%)
cyclo-C6H11 32e (40%)
Ph
SCHEME 14
7c R1 ) CH3 Ar ) p-FC6H4 Ph
7c
t-Bu
32f (36%)
SCHEME 16
TABLE 6. Ugi Reactions of r-Sulfonylketones 7
R2CO2H R3NH2
R4NC
product
(yield)
2-component Passerini condensation. The structure of 29 was
confirmed by single crystal X-ray analysis (Figure 2).
The relatively rapid Passerini reaction of 6a (Figure 1)
suggested that such azidoketones, although heretofore unstudied
in isonitrile-based MCRs, would be good substrates for such
complexity-generating processes. Furthermore, alkyl azides are
particularly useful components in the synthesis of triazoles and
tetrazoles by copper-catalyzed cycloadditions with alkynes using
“click” chemistry.24 As shown in Scheme 14, representative Ugi
condensations of 6a and 6d (R ) CH3) with ammonium acetate
and t-butyl isonitrile afforded the expected diamides 30 and 31
(Scheme 14) in reasonable yield.
We also investigated isonitrile-based MCRs of R-arylsulfo-
nylketones having general structure 7. Unlike other representa-
tive R-substituted ketones in this study, sulfonylketones (also
known as ꢀ-ketosulfones) are distinguished by their relatively
acidic methylene hydrogens and have widespread applications,
in the synthesis of disubstituted alkynes,25 alkenes,26 allenes,27
and heterocycles.28 Some ꢀ-ketosulfones function as aromatase
inhibitors29 and also exhibit fungicidal activity.30
7
R2 )
R3 )
R4 )
7b R1 ) CH3 Ar ) Ph
CH3
Ph
CH3
CH3
H
H
H
H
t-Bu
t-Bu
33a (54%)
33b (<5%)
7b
7b
CH2CO2Et 33c (64%)
n-Bu 33d (51%)
7c R1 ) CH3 Ar ) p-FC6H4
1 d) to afford the expected product 32a (46% yield) as an
inseparable 1:1 mixture with 7a, based on NMR analysis.
The process seemed to be general, judging from the repre-
sentative cases shown in Table 5, in which variations to the
arylsulfonyl group and to the carbon framework of the ketone
were introduced. Unlike 32a, products 32b-f were obtained
pure after flash chromatography. Interestingly, using 2 equiv
of acid and isonitrile shortened the overall reaction time, but
failed to improve the yield of product.
A representative Ugi condensation using phenylsulfonyl
ketone 7b with ammonium acetate and t-butyl isonitrile was
performed as an initial reactivity test, and afforded the expected
product 33a (Scheme 16, 54%). However, additional examples
summarized in Table 6 defined a narrow scope of reactivity.
In entry 2, the low yield obtained of product 33b could not
be improved by switching the solvent to the more commonly
used methanol, and was apparently due to the poor solubility
of benzoic acid in either solvent. Attempts to use benzylamine
in place of ammonia produced complex mixtures in which
neither the Ugi (expected) nor Passerini product could be
identified. Generally, the reaction was successful with a number
of isonitriles including t-butyl or n-butyl isonitrile as well as
ethyl isocyanoacetate.
The Passerini condensation of phenylsulfonyl ketone 7a
(Scheme 15) with acetic acid and t-butyl isonitrile (1 equiv each)
reported in the kinetic study (Figure 1) proceeded smoothly (rt,
(24) Bock, V. D.; Hiemstra, H.; van Maarseveen, J. H. Eur. J. Org. Chem.
2006, 51–68.
(25) (a) Lythgoe, B.; Waterhouse, I. Tetrahedron Lett. 1978, 19, 2625–2628.
(b) Bartlett, P. A.; Green, F. R.; Rose, E. H. J. Am. Chem. Soc. 1978, 100,
4852–4858. (c) Mandai, T.; Yanagi, T.; Araki, K.; Morisaki, Y.; Kawada, M.;
Otera, J. J. Am. Chem. Soc. 1984, 106, 3670–3672.
(26) Ihara, M.; Suzuki, S.; Taniguchi, T.; Tokunaga, Y.; Fukumoto, K.
Tetrahedron 1995, 51, 9873–9890.
(27) Baldwin, J. E.; Adlington, R. M.; Crouch, N. P.; Hill, R. L.; Laffey,
T. G. Tetrahedron Lett. 1995, 36, 7925–7928.
Conclusion
(28) (a) Marco, J.-L.; Fernandez, I.; Khiar, N.; Fernandez, P.; Romero, A. J.
Org. Chem. 1995, 60, 6678–6679. (b) Marco, J. L. J. Org. Chem. 1997, 62,
6575–6581.
(29) Xiang, J.; Ipek, M.; Suri, V.; Tam, M.; Xing, Y.; Huang, N.; Zhang,
Y.; Tobin, J.; Mansour, T. S.; McKew, J. Bioorg. Med. Chem. 2007, 15, 4396–
4405.
In conclusion, we have shown that numerous R-functionalized
ketones having general structures 2-7 can serve as useful
carbonyl components in isonitrile-based MCRs, occasionally
with unexpected and surprising outcomes. The successful
application of Passerini and Ugi condensations to families of
(30) Wolf, W. M. J. Mol. Struct. 1999, 474, 113–124.
J. Org. Chem. Vol. 73, No. 24, 2008 9725

Fan et al.
The oily residue was blanketed in nitrogen, then treated with iso-
butyric acid (140 µL, 1.5 mmol, 1.5 equiv) and t-BuNC (170 µL,
1.5 mmol, 1.5 equiv). The resulting reaction mixture was stirred at
rt under N2 for 20 h. The product was purified by silica gel flash
column chromatography (1:1 EtOAc:hexanes, Rf ) 0.3) to afford
5b (338 mg, 90%) as a pale-yellow oil.
ketones 2-7 enhances the utility of these two powerful name
reactions, and adds another dimension to the structural complex-
ity that can be achieved using multicomponent reactions.
Experimental Section
Representative Procedure for the Passerini Reaction of
r-Sulfonylketones 7. Compound 32d. To a stirred solution under
N2 of freshly dried phenylsulfonylacetone 7b (67 mg, 0.34 mmol)
in CH2Cl2 (0.02 mL, 19 M) was added isobutyric acid (37 µL, 0.37
mmol) followed by cyclohexylisonitrile (45.5 µL, 0.37 mmol). The
reaction mixture was stirred at rt for four days, then volatile residues
were removed under reduced pressure. After TLC analysis,31 the
product was purified by silica gel flash chromatography (2:1 ethyl
acetate:hexanes, Rf ) 0.5) to furnish pure 32d (92 mg, 69%) as a
colorless oil.
Representative Procedure for the Ugi Reaction of r-Sulfo-
nylketones 7. Compound 33a. A solution of NH4OAc (101 mg,
1.31 mmol) and freshly dried phenylsulfonylacetone 7b (65 mg,
0.33 mmol) in anhydrous 2,2,2-trifluoroethanol (add vol, 0.4 M)
under N2 was stirred at 0 °C for 1 h. The, t-butyl isonitrile (41 µL,
0.36 mmol) was added and stirring at rt was continued for two
days. Volatiles were removed under reduced pressure, and after
TLC analysis,31 the crude oily product was purified by silica gel
flash chromatography (10:1 ethyl acetate: hexanes, Rf 0.3) to furnish
pure 33a (60 mg, 54% yield) as a white solid (mp 175-176 °C).
Representative Procedure for the Passerini Reaction of
r-Mesyloxyketones 2. Compound 9a. To a sample of R-mesy-
loxyketone 2a (48 mg, 0.2 mmol) under N2 in a 5 mL round-bottom
flask was added acetic acid (13 µL, 0.22 mmol, 1.1 equiv) and
t-butyl isonitrile (25 µL, 0.22 mmol, 1.1 equiv). The resulting
homogeneous solution was stirred at rt under N2 for 20 h. The
product was purified by silica gel flash column chromatography
(3:7 EtOAc:hexanes, Rf ) 0.2) to afford 9a (68 mg, 88%) as a
pale-yellow oil.
Representative Procedure for the Synthesis of Acyloxy-ꢀ-
lactams. Compound 11a. To a suspension of freshly washed
(pentane) sodium hydride (9 mg of 80% dispersion, 0.3 mmol, 1.5
equiv) in 4:1 CH2Cl2:DMF (3 mL) in a dry 25 mL round-bottom
flask under N2 at 0 °C was added a solution of 9c (82 mg, 0.2
mmol) in 4:1 CH2Cl2:DMF (3 mL) dropwise. The reaction mixture
was stirred at rt for 2 h. Saturated NH4Cl (6 mL) was then added
and the mixture stirred for 30 min at rt. After separating the layers,
the aqueous phase was extracted with CH2Cl2 (5 mL) and the
combined organic layers were washed with H2O (5 mL) and
saturated NaCl (5 mL), dried over MgSO4, filtered, and concentrated
in Vacuo. The crude product was purified by silica gel flash column
chromatography (3:7 ethyl acetate:hexanes, Rf ) 0.3) to afford the
desired product 11a as a pale-yellow oil.
Representative Procedure for the Ugi Reaction of r-Mesyl-
oxyketones 2 with Primary Amines. Compound 15a. To a
solution of R-mesyloxyketone 2a (48 mg, 0.2 mmol, 1 equiv) in
anhydrous 2,2,2-trifluoroethanol (0.5 mL) at 0 °C was added
benzylamine (178 µL, 0.8 mmol, 4.0 equiv) dropwise. The reaction
mixture was stirred for 10 min at 0 °C, then acetic acid (23 µL,
0.4 mmol, 2.0 equiv) and cyclohexyl isonitrile (27 µL, 0.22 mmol,
1.1 equiv) were added via syringe with stirring. The reaction mixture
was then warmed to rt and stirred for 18 h. After removing the
solvent in Vacuo, the residual oil was purified by silica gel flash
column chromatography (3:7 ethyl acetate:hexanes, Rf ) 0.3) to
afford 15a as a white solid (57 mg, 71%).
Representative Procedure for the Synthesis of Di-O-acyl-
glyceramides. Compound 20b. A mixture of acetic acid (57 µL,
1.0 mmol) and Cu(acac)2 (2.6 mg, 0.01 mmol, 1 mol %) toluene
(2 mL) in a 10 mL round-bottom flask was heated to 60 °C for 10
min under nitrogen. To it was added dropwise a solution of
diazoketone 1a (226 mg, 1.3 mmol, 1.3 equiv) in toluene (2 mL).
Once gas evolution was judged complete, the reaction mixture was
stirred an additional 5 min, then cooled and concentrated in Vacuo.
Acknowledgment. Generous support from the Johnson &
Johnson Focused Giving Program is gratefully acknowledged.
One of us (LF) wishes to thank the Sien Moo Tsang endowment
for a graduate fellowship. Support of the Cornell NMR Facility
has been provided by NSF (CHE 7904825; PGM 8018643) and
NIH (RR02002).
Supporting Information Available: Supporting spectro-
1
scopic data, including copies of H and 13C NMR spectra, for
all new compounds. Also included is complete data (collection,
structure solution, refinement, unit cell metrics, atom coordinates
and thermal parameters, bond lengths and angles) for the
structure of 29 as a file in Crystallographic Information File
format. This material is available free of charge via the Internet
at http://pubs.acs.org.
JO8019708
(31) In cases where the product had the same Rf value as the starting ketone,
the crude product was dissolved in CH2Cl2 and washed with 0.2 M NaOH
saturated with NaCl. The aqueous layer was back-extracted twice with CH2Cl2
and the combined organic layers were dried (MgSO4) and concentrated prior to
chromatography.
9726 J. Org. Chem. Vol. 73, No. 24, 2008