Postcondensation modifications of Ugi four-component condensation products: 1-Isocyanocyclohexene as a convertible isocyanide. Mechanism of conversion, synthesis of diverse structures, and demonstration of resin capture

Article abstract of DOI:10.1021/ja953868b

The concept of a 'universal isocyanide' that enables postcondensation modification of Ugi four-component condensation products is introduced. This strategy is suited for the synthesis of libraries. By using 1-isocyanocyclohexene as the isocyanide input in the Ugi reaction, the product cyclohexenamides can be converted to a variety of products. From the original α-(acylamino) amides, new carboxylic acids, esters, and thioesters are produced from acid-activated conversion of the cyclohexenamide moiety. It has been determined that the intermediate in conversion of this type is an oxazolinium-5-one (munchnone) that reacts with many nucleophiles to yield the products above. The munchnone can also undergo cycloaddition with acetylenic dipolarophiles to form pyrroles. Through internal nucleophilic attack, Ugi products are shown to convert to a protected monosaccharide derivative and to 1,4-benzodiazepine-2,5-diones. All of the conversions described consist of a single step. Resin capture of Ugi products is demonstrated, in which a solution condensation reaction is followed by trapping of the products onto solid support resin. Both the trapping step and subsequent cleavage of products from the resin occur in very high yield.

Full text of DOI:10.1021/ja953868b

2574
J. Am. Chem. Soc. 1996, 118, 2574-2583
Postcondensation Modifications of Ugi Four-Component
Condensation Products: 1-Isocyanocyclohexene as a
Convertible Isocyanide. Mechanism of Conversion, Synthesis
of Diverse Structures, and Demonstration of Resin Capture
Thomas A. Keating and Robert W. Armstrong*
Contribution from the Department of Chemistry and Biochemistry, UniVersity of California,
Los Angeles, California 90095
ReceiVed NoVember 17, 1995X
Abstract: The concept of a “universal isocyanide” that enables postcondensation modification of Ugi four-component
condensation products is introduced. This strategy is suited for the synthesis of libraries. By using 1-isocyanocy-
clohexene as the isocyanide input in the Ugi reaction, the product cyclohexenamides can be converted to a variety
of products. From the original R-(acylamino) amides, new carboxylic acids, esters, and thioesters are produced
from acid-activated conversion of the cyclohexenamide moiety. It has been determined that the intermediate in
conversion of this type is an oxazolinium-5-one (mu¨nchnone) that reacts with many nucleophiles to yield the products
above. The mu¨nchnone can also undergo cycloaddition with acetylenic dipolarophiles to form pyrroles. Through
internal nucleophilic attack, Ugi products are shown to convert to a protected monosaccharide derivative and to
1,4-benzodiazepine-2,5-diones. All of the conversions described consist of a single step. Resin capture of Ugi
products is demonstrated, in which a solution condensation reaction is followed by trapping of the products onto
solid support resin. Both the trapping step and subsequent cleavage of products from the resin occur in very high
yield.
Introduction
class of potential drug candidates. Development of libraries of
these compounds is a more challenging endeavor in that (1)
the synthesis should be efficient, (2) the synthesis should offer
ample opportunites for the introduction of input diversity, and
(3) the synthesis should be adaptable to solid support resin or
high throughput solution arrays. Because standard peptides and
oligonucleotides have limitations as bioavailable therapeutics,
the payoff for development of small-molecule libraries is
potentially great.
While very impressive results have been registered in this
area, the strategies to date have been almost exclusively
linear.6,7,9-12 We believe that an important tool of combinatorial
chemistry, the multiple-component condensation (MCC), is
currently being underexploited.13,14,17 A multiple-component
condensation brings three or more reactants together in a single
event to produce a final product that contains aspects of all
inputs. An MCC thus offers greater possibilities for molecular
diversity per step with reduced synthetic time and effort as
compared to linear synthesis. For rapid combinatorial synthesis,
this is a clear advantage.
Combinatorial chemistry has received much attention lately
as a tool for drug discovery.1,2 Unlike traditional approaches
which involve separate synthesis of each individual analog or
desired compound, combinatorial synthesis aims to yield large
numbers of compounds based on a common core structure with
a minimum of time and effort. This may be accomplished
through parallel synthetic methods and readily available, diverse
starting materials. Ideally, complete automation of the synthetic
process would be achieved, thus taking maximum advantage
of the parallel process.
Early work on combinatorial libraries used technologies
readily adaptable to parallel synthesis. Peptides3 and oligo-
nucleotides4,5 were exploited for combinatorial synthesis both
because of the well-developed solid-supported synthetic meth-
odology for each and because each offered a readily available
pool of inputs for each coupling step. For example, for the
creation of a library of all possible tetranucleotides (which
encompasses 44 ) 256 members), all monomers are readily
available, and the linear coupling techniques are quite standard.
More recently, published work has appeared6-12 on libraries
of small-molecule organic compounds, perhaps the most desired
We have recently focused our efforts on the Ugi four-
component condensation (4CC) (Figure 1).15 The Ugi reaction16
brings together a carboxylic acid, an amine, an aldehyde or
X Abstract published in AdVance ACS Abstracts, March 1, 1996.
(1) Gallop, M. A.; Barrett, R. W.; Dower, W. J.; Fodor, S. P. A.; Gordon,
E. M. J. Med. Chem. 1994, 37, 1233-1251.
(9) DeWitt, S. H.; Kiely, J. S.; Stankovic, C. J.; Schroeder, M. C.;
Reynolds Cody, D. M.; Pavia, M. R. Proc. Natl. Acad. Sci. U.S.A. 1993,
90, 6909-6913.
(2) Gordon, E. M.; Barrett, R. W.; Dower, W. J.; Fodor, S. P. A.; Gallop,
M. A. J. Med. Chem. 1994, 37, 1385-1401.
(10) Backes, B. J.; Ellman, J. A. J. Am. Chem. Soc. 1994, 116, 11171-
(3) Jung, G.; Beck-Sickinger, A. G. Angew. Chem., Int. Ed. Engl. 1992,
31, 367-383.
11172.
(11) Baldwin, J. J.; Burbaum, J. J.; Henderson, I.; Ohlmeyer, M. H. J. J.
Am. Chem. Soc. 1995, 117, 5588-5589.
(4) Latham, J. A.; Johnson, R.; Toole, J. J. Nucleic Acids Res. 1994, 22,
2817-2822.
(12) Murphy, M. M.; Schullek, J. R.; Gordon, E. M.; Gallop, M. A. J.
Am. Chem. Soc. 1995, 117, 7029-7030.
(5) Chen, H.; Gold, L. Biochemistry 1994, 33, 8746-8756.
(6) Bunin, B. A.; Ellman, J. A. J. Am. Chem. Soc. 1992, 114, 10997-
10998.
(13) For a recent example, see: Wipf, P.; Cunningham, A. Tetrahedron
Lett. 1995, 36, 7819-7822.
(7) Bunin, B. A.; Plunkett, M. J.; Ellman, J. A. Proc. Natl. Acad. Sci.
U.S.A. 1994, 91, 4708-4712.
(8) Chen, C.; Ahlberg Randall, L. A.; Miller, R. B.; Jones, A. D.; Kurth,
M. J. J. Am. Chem. Soc. 1994, 116, 2661-2662.
(14) For a discussion of MCCs in library synthesis, see: Ugi, I.; Do¨mling,
A.; Ho¨rl, W. EndeaVour 1994, 18, 115-123.
(15) Keating, T. A.; Armstrong, R. W. J. Am. Chem. Soc. 1995, 117,
7842-7843.
0002-7863/96/1518-2574$12.00/0 © 1996 American Chemical Society

Modifications of Ugi Condensation Products
J. Am. Chem. Soc., Vol. 118, No. 11, 1996 2575
As effective as 8 has proven to be in 4CC product modifica-
tions, it should be noted that other universal isocyanides may
turn out to be suited for different and more extensive conver-
sions. For example, the convertible isocyanide strategy de-
scribed here is acid activated. We are investigating other
isocyanide inputs that can be activated for conversion by basic,
neutral, or metal-mediated conditions. Of major importance to
us at this point, however, is the idea of reducing the impact of
a relatively unavailable input in library synthesis by introducing
that input in a convertible form.
We present here evidence for the mechanism of transforma-
tion offered in our initial communication,15 as well as new uses
of Ugi 4CC products for the synthesis of pyrroles, sugars, and
1,4-benzodiazepine-2,5-diones. Finally, we demonstrate the
resin capture of 4CC products of a solution reaction, which
can be a useful tool in solid-supported library synthesis. In
this strategy, a solution phase reaction is followed by trapping
the reaction products onto functionalized resin. This method
can combine the best features of both modes of synthesissease
of monitoring and reduced need for multiple equivalents of
reactants (solution reactions) with ease of isolation and purifica-
tion (solid-supported reactions).
Figure 1.
Results and Discussion
Cyclohexenamide Conversions. Our initial work in this area
demonstrated transformation of Ugi products of type 3 to new
products 4 (eq 2). Ugi has published work18,19 on cleavable
carbonamide moieties R4, but the methods to date require
conversion of the amide group into another, more labile
functionality, and involve at least three steps from the 4CC
product 3. In addition, these methods yield only the carboxylic
acid derivative of 4 (-XR5 ) -OH).
However, Ugi had earlier demonstrated20 the use of 1-iso-
cyanocyclohexene (8) in a 4CC and had converted the enamide
product 5 to the primary amide 6 (eq 3) through acidic
hydrolysis. We decided to reinvestigate this transformation,
Figure 2. Single-step conversions of cyclohexenamide Ugi product
5.
ketone, and an isocyanide in a one-pot condensation reaction
to yield an R-(acylamino) amide (2). While the Ugi reaction is
an excellent tool for a library synthesis of this core structure, it
suffers from a lack of commercially available isocyanides 1.
Unlike the other three readily available components, somewhat
fewer than two dozen isocyanides can be purchased, limiting
potential libraries at the R4 position. One possible solution is
the creation of a sublibrary of isocyanides as inputs into a Ugi
4CC library.17 Another approach will be described herein: we
envisioned the concept of a “universal isocyanide,” an input
for the Ugi 4CC that can be converted, postcondensation, into
different functionalities, thereby circumventing the lack of
commercial isocyanide inputs as well as avoiding the need to
synthesize and store a large number of isocyanides should
commercial sources prove insufficient. The isocyanide that we
present here, 1-isocyanocyclohexene (8), is a remarkably
versatile substrate (see Figure 2) whose 4CC product 5 can be
conVerted in a single step to a Variety of products under acidic
conditions, including many products that are otherwise inac-
cessible from the Ugi reaction.
prompted by a report21 that an enamide could be selectively
cleaved to the corresponding methyl ester under acidic methanol
conditions.
Synthesis of 1-isocyanocyclohexene (8) was straightforward
and proceeded as shown in Scheme 1. Ugi’s original procedure
for synthesis of 8 was modified only in that better yields were
obtained when 7 was purified prior to dehydration and tri-
phosgene was used as the dehydrating agent in place of
phosphorus oxychloride. Other syntheses of 8 have appeared
in the literature.22,23 1-Isocyanocyclohexene can be stored
(18) Ugi, I. Angew. Chem., Int. Ed. Engl. 1982, 21, 810-819.
(19) Geller, J.; Ugi, I. Chem. Scr. 1983, 22, 85-89.
(20) Rosendahl, F. K.; Ugi, I. Ann. Chem. 1963, 666, 65-67.
(21) Fukuyama, T. Presented at the 35th Annual Buffalo Medicinal
Chemistry Symposium, Buffalo, NY, May 1994.
(22) Barton, D. H. R.; Bowles, T.; Huisnec, S.; Forbes, J. E.; Llobera,
A.; Porter, A. E. A.; Zard, S. Z. Tetrahedron Lett. 1988, 29, 3343-3346.
(16) Gokel, G.; Lu¨dke, G.; Ugi, I., In Isonitrile Chemistry; Ugi, I., Ed.;
Academic: New York, 1971; pp 145-199.
(17) Tempest, P. A.; Brown, S. D.; Armstrong, R. W. Angew. Chem.,
Int. Ed. Engl., in press.

2576 J. Am. Chem. Soc., Vol. 118, No. 11, 1996
Keating and Armstrong
Scheme 1a
Table 1. Results of Cleavage of Cyclohexenamide 4CC Products
a Reagents and conditions: (a) NaCN(aq), NH4Cl(aq), Et2O, 61%;
(b) HCOOH, Ac2O, 44%; (c) tBuOK, THF; (d) triphosgene, DABCO,
CH2Cl2, 50% (two steps).
indefinitely at -30 °C under an inert atmosphere, but darkens
in a short time upon exposure to air. We have succeeded in
easily preparing multigram quantities.
Ugi four-component condensations with 8 proceeded smoothly
and in good yield (ca. 50-85%) for a variety of examples
illustrated herein (see the Experimental Section). Conversions
of the cyclohexenamide moiety were then attempted with
various conditions and nucleophiles. When the isolated 4CC
products were treated with the acidic conditions detailed in the
footnotes to Table 1, the products listed were obtained in high
isolated yield. As can be seen, the cyclohexenamide moiety
has been converted under mild conditions to a variety of
carboxylic esters, thioesters, and carboxylic acids. Not only
are products that are not accessible via the Ugi reaction (i.e.,
esters) obtained, but amides can be synthesized from the
carboxylic acids via standard carbodiimide coupling techniques
to afford new Ugi products without requiring the appropriate
isocyanide. This result greatly expands the scope of the 4CC
and reduces the impact of its weakest component, the isocyanide
input.
We next turned to in situ conversions of the cyclohexenamide.
The goal was to perform some of the conversions of Table 1
without isolation of initial cyclohexenamide 4CC product. It
was expected that simple acidification of a crude reaction
mixture that contained the appropriate nucleophile would result
in the cyclohexenamide being converted regardless of other
species present. Results are shown in Table 2. After monitoring
the Ugi reaction in methanol solution and judging it complete,
the introduction of acetyl chloride (to generate HCl) and
subsequent heating resulted in the product methyl esters listed
in Table 2. Hydrolysis products (i.e., carboxylic acids), which
might have arisen from the single equivalent of water produced
in the 4CC, were not observed. By performing the cyclohex-
enamide conversion in a one-pot process, an overall five-
component condensation is achieved, with the isocyanide input
contributing only a single carbonyl carbon atom to the final
structure.
We next focused on the mechanism for this transformation.
It is unusual that a carboxylic acid would result from acidic
hydrolysis of an enamide, as shown in Table 1. One would
expect instead the primary amide, because under aqueous acidic
conditions, protonation of the cyclohexenamide would typically
be followed by hydrolysis of the resulting N-acyliminium species
to the amide and cyclohexanone.24 While cyclohexanone is
indeed observed as a product of these reactions, the primary
amide is not.
a PMB ) p-MeO-benzyl. b Reaction conditions: (A) 1.7% HCl in
THF, 23 °C, overnight; (B) indicated alcohol as solvent, 5 equiv of
AcCl, 55 °C, 3 h; (C) 5 equiv of AcCl, 10 equiv of BnOH in THF, 55
°C, 3 h; (D) EtSH as solvent, 10 equiv of AcCl, 23 °C, overnight; (E)
5 equiv of AcCl, 10 equiv of tBuOH in THF, 55 °C, 48 h. c All yields
are of isolated, purified product.
additional experiments on 26 and 27, it was not only established
that deformylation under these acidic conditions25 was very rapid
(<30 min) but also that the absence of an acylated amine
altogether (28) precludes formation of carboxylic acid as per
Table 1. Thus, it appears that an N-acyl group is essential for
the conversions shown in Tables 1 and 2, and its loss before
the crucial step prevents the formation of the Table 1 products.
This led us to propose15 the mechanism detailed in Figure 3,
which involves an oxazolinium-5-one (mu¨nchnone) intermedi-
ate. Protonation of the enamide gives the activated species 22,
which then cyclizes to the mu¨nchnone26-2930 and eliminates
cyclohexanimine. The reactive mu¨nchnone then is opened by
A clue to the mechanism of transformation is given in Scheme
2. In attempts to convert 24 to the methyl ester under the acidic
methanol conditions of both Tables 1 and 2, only the deformy-
lated primary amide was isolated. This is the expected product
on the basis of a mechanism of hydrolysis of an N-acyliminium
species, but does not fit with the results in Tables 1 and 2. In
(23) Baldwin, J. E.; Yamaguchi, Y. Tetrahedron Lett. 1989, 30, 3335-
3338.
(24) Brossi, A.; Dolan, L. A.; Teitel, S. Organic Syntheses; Wiley: New
York, 1988; Collect. Vol. VI, pp 1-4.
(25) Greene, T. W.; Wuts, P. G. M. ProtectiVe Groups in Organic
Synthesis, 2nd ed.; Wiley-Interscience: New York, 1991; pp 349-350.

Modifications of Ugi Condensation Products
J. Am. Chem. Soc., Vol. 118, No. 11, 1996 2577
Table 2. Results of In situ Acidic Methanolyses of Four
Component Condensation Products
Figure 3. Proposed mechanism of cyclohexenamide conversions.
Figure 4. Resonance structures of active 1,3-dipoles derived from Ugi
products.
Scheme 3
Scheme 2
dration of N-acyl-N-alkylamino acids using acetic anhydride and
a nitrogenous base at elevated temperatures.31
Therefore, we attempted cycloaddition of cyclohexenamide
Ugi products under various acidic conditions with a selection
of acetylenic dipolarophiles, with the expectation that the
reaction would proceed as depicted in Scheme 3. Protonation
of 5 is followed by cyclization and loss of cyclohexanimine to
30, which loses a proton to form 1,3-dipole 32. This 1,3-dipole
undergoes a [3+2] cycloaddition reaction with the acetylene to
43, which rapidly aromatizes with loss of CO2 to the pyrrole
44. Results of these reactions are shown in Table 3. Several
general trends can be observed. First, unlike cyclodehydration
procedures, which are typically run at 50-60 °C, higher
temperatures and toluene as a solvent were necessary for
improved or even observable yields. Second, the more electron
withdrawing the acetylenic dipolarophile, the better the yields.
Steric hinderance plays a role, as seen in the product ratios of
36 to 37,26 the failure of diphenylacetylene to react to form the
tetraphenylpyrrole,31 and the failure to form the bicyclic pyrrole
expected from the γ-lactam 4CC product. Optimization of
yields by further variation of conditions has not been attempted.
While cycloaddition of mu¨nchnones with electron deficient
nitriles such as ethyl cyanoacetate has also been reported,32 we
were unable to observe any predicted imidazole products with
a variety of conditions and substrates.
a nucleophile to yield the product 23. Lack of an acylated amine
(28), or its rapid loss (24), prevents the formation of 30.
Pyrroles. While these experiments offered some evidence
for the intermediacy of a mu¨nchnone in the cyclohexenamide
conversion, more compelling proof would result from a 1,3-
dipolar cycloaddition of a dipolarophile to the proposed
intermediate. It is well-known that the anhydro-5-hydroxyox-
azolium hydroxide species 31 (Figure 4), which are members
of a larger class of azomethine ylides, undergo cycloadditions
with a wide variety of dipolarophiles.30 Perhaps the most
common method of generating such species is the cyclodehy-
(26) Coppola, B. P.; Noe, M. C.; Schwartz, D. J.; Il Abdon, R. L.; Trost,
B. M. Tetrahedron 1994, 50, 93-116.
(27) Dalla Croce, P.; Rosa, C. L. Heterocycles 1988, 27, 2825-2832.
(28) Padwa, A.; Burgess, E. M.; Gingrich, H. L.; Roush, D. M. J. Org.
Chem. 1982, 47, 786-791.
(29) Potts, K. T. In 1,3-Dipolar Cycloaddition Chemistry; Padwa, A.,
Ed.; Wiley-Interscience: New York, 1984; Vol. 2, pp 1-82.
(30) Lown, J. W. In 1,3-Dipolar Cycloaddition Chemistry; Padwa, A.,
Ed.; Wiley-Interscience: New York, 1984; Vol. 1, pp 653-732.
(31) Huisgen, R.; Gotthardt, H.; Bayer, H. O.; Schaefer, F. C. Angew.
Chem., Int. Ed. Engl. 1964, 3, 136-137.
(32) Brunn, E.; Funke, E.; Gotthardt, H.; Huisgen, R. Chem. Ber. 1971,
104, 1562.

2578 J. Am. Chem. Soc., Vol. 118, No. 11, 1996
Keating and Armstrong
Table 3. 1,3-Dipolar Cycloadditions of Acetylenic Dipolarophiles with Ugi Products
a Reagents and conditions: (A) 5 equiv of dipolarophile, 3 equiv of HCl, toluene, 100 °C; (B) 5 equiv of dipolarophile, 3 equiv of HCl, THF,
55 °C. b Purified, isolated yield.
This 1,3-dipolar cycloaddition reaction represents a novel
synthesis of pyrroles, which are important pharmacophores in
their own right.33 There are several notable advantages to this
method. Instead of reliance upon acylated amino acids as
precursors for mu¨nchnones 30, the larger and more diverse pool
of carboxylic acids, amines, and aldehydes can be called upon
to furnish the 1-, 2-, and 5-substituents of pyrrole 44 resulting
from cycloaddition. Thus, even though the cyclohexenamide
class of Ugi 4CC products can be seen as precursors of acylated
amino acids (as per 9, 10, and 11 in Table 1), it is not necessary
to convert them to the free carboxylic acid to access the standard
acetic anhydride methods of generating active 1,3-dipoles for
cycloaddition; the cyclohexenamide itself is a substrate for
mu¨nchnone formation. Once again, the cyclohexenamide func-
tionality has been erased in the product pyrrole, with even the
single carbon atom present in products 9-21 lost as CO2. Final-
ly, the stereocenter created at the R-carbon in the Ugi 4CC (re-
sulting in an enantiomeric mixture of products) is removed by
aromatization after cycloaddition, resulting in a single product.
Internal Nucleophile. As an extension of the work on
nucleophilic opening of mu¨nchnone intermediates to yield
products 9-21, we investigated tethered nucleophiles as a means
of forming rings (Figure 5). The concept in Figure 5a is
illustrated by using D-arabinose as an aldehyde input in a Ugi
reaction followed by cyclization to form the corresponding
2-acetamido-2-deoxyhexose (Scheme 4).
Protected D-arabinose 46 was prepared using standard
procedures.34-36 This aldehyde reacted smoothly under standard
Ugi conditions to furnish 47 in high yield as a 3.4:1 mixture of
inseparable diastereomers at the R-carbon (as determined by
1H NMR). We expected that acidic treatment of 47 would
cleave the isopropylidene protecting group, protonate the
enamide, promote mu¨nchnone formation, and either open the
mu¨nchnone with methanol or the newly deprotected secondary
hydroxyl, with the exact sequence of events dependent upon
the relative rates. Lactonization should occur under these
(34) Horton, D.; Varela, O. Carbohydr. Res. 1984, 134, 205-214.
(35) Armstrong, R. W.; Teegarden, B. R. J. Org. Chem. 1992, 56, 915-
922.
(33) For a recent representative example, see: Santiago, B.; Dalton, C.
R.; Huber, E. W.; Kane, J. M. J. Org. Chem. 1995, 60, 4947-4950.
(36) Teegarden, B. R. Ph.D. Thesis, University of California, Los
Angeles, 1991.

Modifications of Ugi Condensation Products
J. Am. Chem. Soc., Vol. 118, No. 11, 1996 2579
2,5-diones have appeared very recently in several solid and
solution phase syntheses.42-44
Synthesis of this class of compounds would represent an
intramolecular cyclization of the type shown in Figure 5b. If a
Ugi product of type 51 could be synthesized, cleavage of the
cyclohexenamide to an ester or acid and subsequent lactamiza-
tion with the anthranilic nitrogen would yield the corresponding
1,4-benzodiazepine-2,5-dione 50 (eq 11), in either one or two
Figure 5.
Scheme 4
steps from the Ugi product. The commercially available inputs
for a library of such compounds would then consist of over 40
substituted anthranilic acids, and the legion of amines and
aldehydes. As before, the isocyanide input would be erased
except for a single carbon atom.
In the interests of ease of synthesis, we wished to avoid
employing a protecting group for the anthranilic acid nitrogen
in the Ugi 4CC. This would allow the use of anthranilic acids
“off-the-shelf” and obviate the need for protecting group
removal later. We expected that the precondensation of the
amine and aldehyde components, along with the reduced
nucleophilicity of the anthranilic nitrogen, would avoid this
problem. The results are shown in Scheme 5.
By combining isobutyraldehyde and p-methoxybenzylamine
in methanol with 4 Å molecular sieves and stirring for 1 h,
followed by isocyanide and finally anthranilic acid addition, a
73% isolated yield of 4CC product 52 was obtained, with some
recoverable starting materials, but no observable side products.
Treatment of this product with acetyl chloride in methanol
resulted in 1,4-benzodiazepine-2,5-dione 53 as a single isolable
product in 82% yield. A second example with different inputs
is shown in eq 13. Aside from sluggishness of the 4CC, the
reactions proceeded similarly. This accomplishes the synthesis
of 1,4-benzodiazepine-2,5-diones extremely rapidly, in good
yield, and with remarkable possibilities for diverse functionality.
Resin Capture. As a final exmple of the versatility of the
cyclohexenamide conversion, we demonstrate here the resin
capture of Ugi products onto a solid support. We believe that
this can represent a valuable strategy for library synthesis, in
which a solution reaction is followed by capture of the products
onto a solid support for further transformation, eventually
leading to removal of the final products from the resin.
Although library generation has traditionally followed an either/
or approach to solid and solution phase synthesis, a mixed
strategy in which difficult to monitor or low-yielding reactions
are performed in solution, and then the following reactions on
a solid support, could prove superior.
conditions whether or not the methyl ester is initially formed,
and cyclization should occur via the secondary hydroxyl to form
a six-membered ring rather than via the primary hydroxyl
(seven-membered ring). Treatment of 47 with acidic methanol
conditions (0.1 M HCl, 23 °C) yielded the protected 2-aceta-
mido-2-deoxy-D-mannono-δ-lactone 48 in 29% yield. Attempts
to improve the yield through either elevated temperatures or
increased acidicity resulted in â-elimination and formation of
the R,â-unsaturated 2-acetamido lactone. The p-methoxybenzyl
amide protecting group could be oxidatively removed using ceric
ammonium nitrate, and the benzyl ethers by catalytic hydro-
genation over Pearlman’s catalyst.
1,4-Benzodiazepine-2,5-diones. 1,4-Benzodiazepines37 49
have been the target of several solid phase combinatorial
strategies.6,7,9,38 The structurally similar 1,4-benzodiazepine-
2,5-diones, however, have been reported as important pharma-
We employed the Wang45 p-(benzyloxy)benzyl alcohol resin
as the nucleophile in cyclohexenamide conversion (Scheme 6),
which was expected to perform similarly to benzyl alcohol in
product 15. An excess of Ugi 4CC products 5 was incubated
cophores in their own right, as anticonvulsants,39 as antitumor
agents,40 and as glycoprotein antagonists.41 1,4-Benzodiazepine-
(41) McDowell, R. S.; Blackburn, B. K.; Gadek, T. R.; McGee, L. R.;
Rawson, T.; Reynolds, M. E.; Robarge, K. D.; Somers, T. C.; Thorsett, E.
D.; Tischler, M.; Webb, R. R., II; Venuti, M. C. J. Am. Chem. Soc. 1994,
116, 5077-5083.
(42) Boojamra, C. G.; Burow, K. M.; Ellman, J. A. J. Org. Chem. 1995,
60, 5742-5743.
(43) Goff, D. A.; Zuckermann, R. N. J. Org. Chem. 1995, 60, 5744-
5745.
(37) Sternbach, L. H. J. Med. Chem. 1979, 22, 1.
(38) Plunkett, M. J.; Ellman, J. A. J. Am. Chem. Soc. 1995, 117, 3306-
3307.
(39) Cho, N. S.; Song, K. Y.; Pa´rka´nyi, C. J. Heterocycl. Chem. 1989,
26, 1807-1810.
(40) Jones, G. B.; Davey, C. L.; Jenkins, T. C.; Kamal, A.; Kneale, G.
G.; Neidle, S.; Webster, G. D.; Thurston, D. E. Anti-Cancer Drug Des.
1990, 5, 249.
(44) Karp, G. M. J. Org. Chem. 1995, 60, 5814-5819.
(45) Wang, S.-S. J. Am. Chem. Soc. 1973, 95, 1328.

2580 J. Am. Chem. Soc., Vol. 118, No. 11, 1996
Keating and Armstrong
Scheme 5
Scheme 6
of carboxylic acid and methyl ester derivatives arising from
hydrolysis/methanolysis of the excess 4CC starting material
during these washes.
Interestingly, when we attempted to capture 1,4-benzodiaz-
epine-2,5-dione precursors 52 and 54 on Wang resin, subsequent
cleavage yielded no product. Characterization of the wash
solutions revealed high yields of the 1,4-benzodiazepine-2,5-
diones. Apparently, the cyclization to produce the 1,4-benzo-
diazepine-2,5-diones from their 4CC precursors is facile enough
that the initially formed Wang benzyl ester is very rapidly
displaced by the anthranilic nitrogen; thus, the product cleaves
itself off the resin.9 This was observed even under milder
temperatures and acidity.
Conclusions
Table 4. Results of Resin Capture of 4CC Products and
Subsequent Cleavage
The Ugi four-component condensation and multiple-compo-
nent condensations in general are powerful tools for the
generation of libraries of diverse compounds based on a common
core. There are several advantages to MCCs over linear
synthesis: fewer steps, greater diversity at each step from higher
numbers of inputs,46 and (in some cases) more readily available
starting materials. The drawbacks to MCCs are not trivial: the
libraries are wedded to a single core functionality, and in the
case of the Ugi 4CC, some inputs have only limited availability.
We have sought to address these shortcomings specifically. Use
of 1-isocyanocyclohexene as the isocyanide input yields products
that can enjoy great variability at a formerly restricted site; not
only are different amide functionalities possible, but also
heretofore inaccessible moieties. The pyrrole synthesis pre-
sented above represents transformation into an entirely new class
of compounds. But the scope of the cyclohexenamide conver-
sion is not limited to external nucleophiles. We have demon-
strated nucleophiles attached to both the aldehyde and acid
inputs, which resulted in a monosaccharide and benzodiaz-
epinediones, respectively. More generally, we believe the
concept of a universal isocyanide can extend beyond the
1-isocyanocyclohexene presented here. Although 1-isocyano-
cyclohexene is remarkably versatile, different isocyanides can
potentially lead to further modifications that open access to
with Wang resin under anhydrous acidic THF conditions, and
then the resin was washed repeatedly with methanol and CH2-
Cl2. Cleavage from the resin was with a 20% (v/v) trifluoro-
acetic acid in methylene chloride solution. The products were
characterized without purification and were >95% pure by TLC
and 1H NMR. Perhaps most remarkable, as displayed in Table
4, are the yields of products recovered from the resin. The
yields were calculated on the basis of the manufacturer’s stated
loading level of the Wang resin. We initially employed a 4-fold
excess of 4CC products, but discovered that high yields of
recovered carboxylic acid were maintained even when the excess
of Ugi products was reduced to 1.5 equiv! Inspection of the
solution washes of the resin prior to cleavage revealed a mixture
(46) In a linear strategy, bifunctional or bireactiVe inputs are required at
each step in order to provide reactive sites for succeeding steps. On the
other hand, an MCC can proceed successfully with monofunctional and
monoreactiVe inputs. This concept is illustrated by comparison of peptide
synthesis with the Ugi 4CC. Each amino acid input for linear peptide
synthesis incorporates two orthogonally reactive sitessone to couple to the
existing peptide and the other to provide a site for further attachment.
However, the four inputs for the Ugi reaction all require only one reactive
site. Thus, the 4CC reaction can employ a greater number and variety of
inputs than can a linear strategy.

Modifications of Ugi Condensation Products
J. Am. Chem. Soc., Vol. 118, No. 11, 1996 2581
and the residue dissolved in pH 10 water and washed with methylene
chloride. The aqueous layer is then acidified and reextracted with
methylene chloride. The organics are combined and dried over sodium
sulfate, and the solvent is evaporated.
General Procedure for Pyrrole Synthesis (Method A). The
cyclohexenamide (0.05 mmol) is azeotropically dried with toluene, and
then dissolved in 1 mL of toluene. The acetylene (0.25 mmol, 5 equiv)
is then added, followed by HCl (3 equiv as a 1.0 M solution in
anhydrous ether). The flask is then capped and heated to 100 °C for
4 h. The solvent is then evaporated after cooling, the residue taken up
in methylene chloride and filtered, and the soluble portion purified by
preparative TLC (0.5 mm thickness, 20 cm × 20 cm, 1-5% MeOH in
methylene chloride as eluent).
classes of compounds beyond those shown here. For instance,
we have investigated metal-mediated transformations of Ugi
products incorporating a 2-isocyanopyridine input.
Finally, polymer capture of activated cyclohexenamides, while
not fully exploited for library synthesis here, represents a new
strategy for solid phase synthesis: beginning not with attachment
of the first input to the solid support, but rather with a solution
reaction whose products are amenable to being subsequently
attached to the resin. Careful inspection of the system presented
here reveals that the only species trappable on the resin is the
fully-formed Ugi 4CC product; unreacted starting materials and
potential side products are inert to the capture because they lack
the required cyclohexenamide functionality which is only present
in the 4CC product. Furthermore, while it has not been
attempted here, “captured” 4CC products are available to
undergo further synthetic transformations on a solid support prior
to being cleaved from the resin. We believe this method can
lead to hybrid strategies for library generation that combine the
ease of solution synthesis with the ease of solid-supported
isolation and purification.
General Procedure for Pyrrole Synthesis (Method B). This
procedure is identical to that above except that THF is substitued for
toluene as solvent and the reaction is heated to 55 °C.
General Procedure for Polymer Capture and Cleavage Using
Wang Resin. The cyclohexenamide (0.11 mmol) is azeotropically
dried with toluene, dissolved in THF, and then added to a flask
containing Wang resin (0.027 mmol, 0.25 equiv, or 0.073 mmol, 0.67
equiv). HCl (0.55 mmol as a 1 M solution in anhydrous ether) is added
and the flask capped and then heated to 55 °C for 5 h. After cooling,
the resin is filtered off and washed three times with methylene chloride,
three times with methanol, and three times with methylene chloride
again. The carboxylic acid product is cleaved from the resin by
incubating with a 20% trifluoroacetic acid in methylene chloride
solution for 20 min, at which point the pink to dark purple resin is
washed three times with methylene chloride. The solvent is evaporated,
and the product is characterized.
Experimental Section
General Procedures. All reactions unless otherwise indicated were
performed in oven- or flame-dried glassware under an inert atmosphere.
Solvents were distilled immediately prior to use: THF from sodium/
benzophenone ketyl, methanol from magnesium turnings, CH2Cl2 from
P2O5, and toluene from calcium hydride. Anhydrous DMF was
purchased from Aldrich and used directly. Wang p-(benzyloxy)benzyl
alcohol resin (100-200 mesh) was purchased from Advanced ChemTech.
The loading level was 0.9 mmol/g. The resin was dried prior to use
by azeotropically removing water with toluene. Thin layer chroma-
tography was performed on silica gel with precoated glass plates (E.
Merck Brinkman, Kieselgel 60 F254, 0.25 mm) and visualized with UV
light, p-anisaldehyde, and/or ninhydrin staining. NMR spectra were
obtained with a Bruker ARX-500, ARX-400, or AM-360 spectrometer
in a CDCl3 solvent and referenced to residual CHCl3. Coupling
constants are listed in hertz. IR spectra were obtained with a Nicolet
510P FT-IR spectrometer. Absorbances are stated in inverse centi-
meters. Optical rotations were obtained with a Perkin-Elmer 241MC
polarimeter. Concentrations are reported in grams per milliliter.
General Procedure for Ugi Four-Component Condensation. The
carboxylic acid (1.25 equiv), amine (1.25 equiv), and aldehyde (1.0
equiv) are dissolved in methanol to an approximate concentration of 1
M in each. This solution is allowed to stand for 10 min, and then is
added in one portion to a flask containing the isocyanide (1.0 equiv).
The resulting solution is allowed to stir at room temperature for 12 h.
When reaction is complete by TLC (1-5% methanol in CH2Cl2), the
solvent is removed in vacuo, and the residue purified by flash column
chromatography on silica gel, eluting with a 0-5% methanol in CH2-
Cl2 gradient.
General Procedure for Ugi Reaction Followed by in Situ
Methanolysis. The carboxylic acid (1.25 equiv), amine (1.25 equiv),
and aldehyde (1.0 equiv) are dissolved in methanol to an approximate
concentration of 1 M in each. This solution is allowed to stand for 10
min, and then is added in one portion to a flask containing the
isocyanide (1.0 equiv). The resulting solution is allowed to stir at room
temperature for 12 h. When reaction is complete by TLC (1-5%
methanol in CH2Cl2), the methanol volume is doubled and acetyl
chloride (10 equiv) is added in one portion. The flask is equipped
with a reflux condenser and heated to 55 °C for 3 h. When TLC shows
complete conversion to methyl ester, the reaction is cooled to room
temperature, the solvent removed in vacuo, and the residue taken up
in CH2Cl2 and filtered. Purification is by flash column chromatography
on silica gel, eluting with a 0-5% methanol in CH2Cl2 gradient.
General Procedure for Acidic Hydrolysis of Cyclohexenamide
Products. The starting cyclohexenamide (0.1 mmol) is dissolved in 1
mL of a stock solution of 0.5 mL concentrated HCl in 9.5 mL of THF.
This reaction mixture is stirred for 12-18 h, neutralized with solid
sodium bicarbonate, and then filtered. The solvent is removed in vacuo
Characterization and experimental data for compounds 9-10, 13-
21, and 25 have been previously reported.15 Data for these compounds
include full experimental details, complete listings of proton and carbon-
13 magnetic resonances, IR absorbances, results of high-resolution mass
1
spectrometric analysis, and copies of H NMR and 13C NMR spectra.
1-Amino-1-cyanocyclohexane.20 Cyclohexanone (103 mL, 1.0 mol)
was dissolved in 33 mL of diethyl ether in a 1 L flask. Ammonium
chloride (60.4 g, 1.13 mol) was then added in 183 mL of water. The
solution was cooled in an ice bath, followed by sodium cyanide (50.5
g, 1.03 mol) addition in 133 mL of water dropwise via addition funnel.
The biphasic solution was stirred overnight at 23 °C, and then the
aqueous layer was acidified to pH 1 with concentrated HCl. (CAU-
TION: Hydrogen cyanide is liberated.) The hydrochloride salt of the
product was filtered and washed with ether. The remaining aqueous
solution was extracted with ether three times to remove unreacted
cyclohexanone, and then the pH was raised to 12 with 1 M NaOH and
the solution extracted with ether three times to collect the remaining
product. Organics were combined and dried over Na2SO4 and the
solvent removed in vacuo to leave a clear oil. The hydrochloride salt
was dissolved in 1 M NaOH and extracted with ether; after repeating
the above procedures, the oils were combined to yield 76 g (61%) of
product.
1-Cyano-1-formamidocyclohexane (7).20 To 1-amino-1-cyanocy-
clohexane (76 g, 613 mmol) in an ice bath cooled flask was added 107
mL of 98% formic acid. To this stirred, cooled solution was added a
mixture of 107 mL of 98% formic acid and 63.5 mL of acetic anhydride
dropwise via addition funnel. (CAUTION: Formation of the mixed
anhydride occurs with considerable heat evolution and should be
performed slowly and in a cooled flask.) After stirring for 12 h at 23
°C, the reaction was quenched with 10 mL of water, and as much
solvent as possible was removed with a rotary evaporator. The resulting
yellow oil was dissolved in methylene chloride and washed with pH
10 water. This aqueous layer was extracted with fresh methylene
chloride twice, and the combined organics were washed with pH 1
water to remove starting material. Neutralization of the acidic water
and CH2Cl2 extraction yielded, after standard drying and solvent
removal, 5.0 g (6.5%) of starting material. The former organic portions
yielded 55 g of crude product, which was recrystallized from benzene
to provide 40 g (53%) of shiny, white leaves. The title compound
exists as two rotamers in a 7.7:1 ratio at room temperature in
chloroform. 1H NMR (400 MHz): (major rotamer) δ 8.13 (d, 1 J )
1.1), 6.88 (br, 1), 2.32 (m, 2), 1.73-1.58 (m, 8); (minor rotamer) δ
8.49 (d, 1, J ) 11), 7.02 (d, 1, J ) 11), 2.15 (m, 2), 1.31-1.27 (m, 8).

2582 J. Am. Chem. Soc., Vol. 118, No. 11, 1996
Keating and Armstrong
1-Formamidocyclohexene.22 1-Cyano-1-formamidocyclohexane (7)
(2.0 g, 13.2 mmol) and anhydrous potassium tert-butoxide (5.9 g, 52.6
mmol) were dissolved in 50 mL of THF. After heating to reflux for
6 h, the reaction mixture was poured into 200 mL of 0.5 M Na2CO3.
The layers were separated and extracted 3× with ethyl acetate. The
organics were combined and dried over Na2SO4. Silica gel chroma-
tography (1:1 to 1:2 hexanes/ethyl acetate) yielded 1.22 g (74%) of a
white crystalline solid along with 177 mg (9%) of starting material.
The title compound exists as two rotamers in a 3:1 ratio at room
temperature in chloroform. 1H NMR (400 MHz): (major rotamer) δ
8.35 (d, 1, J ) 11.5), 7.55 (br, 1), 5.30 (t, 1, J ) 3.9), 2.15-2.07 (m,
4), 1.73-1.58 (m, 4); (minor rotamer) δ 8.21 (d, 1, J ) 13.5), 8.15 (d,
1, J ) 1.6), 6.12 (t, 1, J ) 3.9), 2.33 (m, 4), 1.85 (m, 4).
6.87 (d, 2, J ) 8.6), 4.37 (s, 1), 4.12 (dd, 1, J ) 18.3, 6.0), 4.00 (dd,
1, J ) 18.3, 5.3), 3.82 (d, 1, J ) 13.0), 3.80 (s, 3), 3.78 (d, 1, J )
13.0), 3.77 (s, 3), 2.45 (br, 1). HRMS (EI): m/z (M+) calcd 343.1658,
found 343.1659.
(R,S)-1-(1-Piperidino)phenylacetamide (29). IR (neat): 3403,
3185, 2934, 1661. 1H NMR (500 MHz): δ 7.35-7.27 (m, 5), 7.16
(br, 1), 5.51 (br, 1), 3.85 (s, 1), 2.39 (br, 4), 1.63 (br, 4), 1.42 (br, 2).
HRMS (EI): m/z [(M + H)+] calcd 219.1497, found 219.1499.
Dimethyl 1-(4-Methoxybenzyl)-2-methyl-5-phenylpyrrole-3,4-di-
carboxylate (34). Yield: 63%. IR (neat): 1707, 1514, 1248, 1175.
1H NMR (400 MHz): δ 7.33-7.24 (m, 5), 6.80 (d, 2, J ) 8.7), 6.77
(d, 2, J ) 8.9), 4.91 (s, 2), 3.83 (s, 3), 3.79 (s, 3), 3.65 (s, 3), 2.37 (s,
3). 13C NMR (101 MHz): δ 166.1, 165.4, 158.9, 135.6, 135.3, 130.4,
130.3, 128.6, 128.2, 126.8, 115.0, 114.2, 112.1, 55.2, 51.6, 51.4, 47.2,
11.4. HRMS (EI): m/z [(M + H)+] calcd 394.1654, found 394.1652.
Dimethyl 1-butyl-2,5-diphenylpyrrole-3,4-dicarboxylate (35). Yield
) 35%. IR (neat): 2953, 1717, 1198, 1167. 1H NMR (400 MHz): δ
7.45-7.40 (m, 10), 3.68 (t, 2, J ) 7.7), 3.65 (s, 6), 1.14 (tt, 2, J ) 7.5,
7.5), 0.83 (tq, 2, J ) 7.5, 7.5), 0.49 (t, 3, J ) 7.4). 13C NMR (101
MHz): δ 165.4, 136.5, 131.0, 130.5, 128.7, 128.2, 114.3, 51.6, 44.6,
32.3, 19.3, 13.1. HRMS (EI): (M+H)+ calcd 392.1862, found
392.1866.
1-Isocyanocyclohexene (8).20,22,23 1-Formamidocyclohexene (6.0 g,
48.0 mmol) and diazabicyclo[2.2.2]octane (DABCO, triethylenedi-
amine) (16.2 g, 144.0 mmol) were dissolved in 120 mL of CH2Cl2. To
the cooled solution (0 °C), 10% w/v triphosgene in CH2Cl2 (95 mL,
32.0 mmol) was added dropwise. After stirring for 0.5 h, the reaction
mixture was poured into 500 mL of a 0.5 M Na2CO3 solution. This
was extracted 3× with CH2Cl2, dried over Na2SO4, and rapidly
chromatographed on triethylamine-deactivated silica gel (9:1 hexanes/
ethyl acetate) to yield 3.1 g (60%) of a clear, colorless oil. The product
is conveniently stored under argon at -30 °C as a 1 M solution in
hexanes. 1H NMR: δ 6.05 (m, 1), 2.23 (m, 2), 2.11 (m, 2), 1.69 (m,
2) 1.57 (m, 2).
(R,S)-2-(N-Butylbenzamido)phenylacetic Acid (11). NMR spectra
of the title compound suffer extreme peak broadening due to rotational
isomerism. This compound was characterized by 1H NMR and HRMS,
and by conversion into 12. 1H NMR (400 MHz): δ 7.47-7.38 (m,
5), 5.67 (s, 1), 3.38 (br, 1), 3.24 (br, 1), 1.41 (br, 1), 1.24 (br, 1), 0.96-
0.86 (br, 2), 0.67 (br, 3). HRMS (EI): m/z (M+) calcd 311.1521, found
311.1521.
Methyl 1-(4-Methoxybenzyl)-2-methyl-5-phenyl-3-pyrrolecar-
boxylate (36). Yield: 24% (isolated as a 3.3:1 mixture with 37).
1
Assignment of major and minor regioisomers was on the basis of H
NMR chemical shifts, coupling constants, observed NOE of the pyrrole
proton from irradiation of the pyrrole methyl in the minor regioisomer,
and published data on analogous compounds.27,28 IR (neat): 2928, 1701,
1514, 1248. 1H NMR (400 MHz): δ 7.33-7.26 (m, 5), 6.84 (s, 4),
6.65 (s, 1), 5.07 (s, 2), 3.82 (s, 3), 3.79 (s, 3), 2.46 (d, 3, J ) 3.3).
HRMS (EI): m/z (M+) calcd 335.1521, found 335.1526.
Methyl 1-(4-Methoxybenzyl)-5-methyl-2-phenyl-3-pyrrolecar-
boxylate (37). Yield: 24% (isolated as a 1:3.3 mixture with 36). IR
(neat): 2928, 1701, 1514, 1248. 1H NMR (400 MHz): δ 7.33-7.26
(m, 5), 6.80 (d, 2, J ) 8.9), 6.75 (d, 2, J ) 8.9), 6.47 (d, 1, J ) 0.7),
4.85 (s, 2) 3.78 (s, 3), 3.65 (s, 3), 2.12 (d, 2, J ) 0.6). HRMS (EI):
m/z (M+) calcd 335.1521, found 335.1526.
(R,S)-2-(N-Butylbenzamido)phenylacetic Acid Methyl Ester (12).
Yield: 98%. IR (neat): 2957, 1748, 1638. 1H NMR (500 MHz): δ
7.44-7.38 (m, 10), 5.97 (br, 1), 3.80 (s, 3), 3.17 (br, 2), 1.63 (br, 2),
0.81 (br, 2), 0.52 (br, 3). 13C NMR (126 MHz): δ 170.9, 136.5, 134.4,
129.5, 128.8, 128.7, 128.4, 126.5, 62.0, 52.4, 47.5, 31.5, 19.6, 13.2.
HRMS (EI): m/z (M+) calcd 325.1678, found 325.1681.
Methyl 1-Butyl-2,5-diphenyl-3-pyrrolecarboxylate (38). Yield:
24%. IR (neat): 2957, 2928, 1717, 1475, 1196. 1H NMR (400
MHz): δ 7.46-7.36 (m, 10), 6.67 (s, 1), 3.83 (t, 3, J ) 7.5), 3.65 (s,
3), 1.18 (tt, 2, J ) 7.3, 7.3), 0.85 (tq, 2, J ) 7.4, 7.4), 0.53 (t, 3, J )
7.4). 13C NMR (101 MHz): δ 165.2, 139.7, 134.6, 133.1, 132.3, 130.7,
129.2, 128.5, 128.3, 128.0, 127.6, 113.0, 110.5, 50.8, 44.7, 32.5, 19.3,
13.2. HRMS (EI): m/z (M+) calcd 333.1719, found 333.1722.
(R,S)-N-(1-Cyclohexenyl)-2-(N′-(4-methoxybenzyl)formamido)-
phenylacetamide (24). Yield: 72%. IR (neat): 3308, 2932, 1655,
1514, 1248. The title compound exists at room temperature as a 1.3:1
mixture of rotamers. When peaks corresponding to the same proton-
(s) from each rotamer can be identified, they are listed separately as
major and minor. 1H NMR (400 MHz): (major rotamer) δ 8.23 (s,
1), 7.28-7.27 (m, 5), 7.13 (d, 2, J ) 8.4), 6.71 (d, 2, J ) 8.4), 6.03
(br, 1), 5.77 (s, 1), 4.51 (d, 1 J ) 15.4), 4.29 (d, 1, J ) 15.3), 3.73 (s,
3), 2.30 (t, 1, J ) 6.4), 2.06-1.52 (m, 7); (minor rotamer) δ 8.15 (s,
1), 7.28-7.27 (m, 5), 6.89 (d, 2 J ) 7.5), 6.81 (d, 2, J ) 7.5), 6.03
(br, 1), 4.97 (s, 1), 4.71 (d, 1, J ) 14.7), 4.27 (d, 1 J ) 14.9), 3.75 (s,
3), 2.06-1.52 (m, 8). 13C NMR (101 MHz): δ 167.0, 166.6, 163.8,
159.0 (2), 134.6, 134.1, 132.3, 132.2, 129.9 (2), 129.5, 129.3, 129.0
(2), 128.7, 128.6, 128.5 (2), 128.3, 128.2, 114.2, 114.0, 113.8, 113.6,
64.8, 61.0, 55.1, 49.7, 46.6, 41.8, 27.6, 27.5, 26.9, 24.9, 23.9, 23.8,
22.3 (2), 21.8, 21.7. HRMS (EI): m/z (M+) calcd 378.1943, found
378.1939.
Dimethyl 2-Isopropyl-1-(4-methoxybenzyl)-5-methylpyrrole-3,4-
dicarboxylate (39). Yield: 13% (method A), 5% (method B). IR
(neat): 2951, 1705, 1514, 1215. 1H NMR (400 MHz): δ 6.84 (d, 2,
J ) 8.9), 6.80 (d, 2, J ) 8.9), 5.03 (s, 2), 3.84 (s, 3), 3.78 (s, 6), 3.01
(qq, 1, J ) 7.1, 7.1), 2.35 (s, 3), 1.19 (d, 6, J ) 7.1). 13C NMR (101
MHz): δ 168.1, 165.3, 159.0, 138.8, 134.3, 128.4, 126.6, 114.3, 111.1,
55.3, 51.9, 51.2, 46.4, 29.7, 26.1, 21.7, 11.1. HRMS (EI): m/z (M+)
calcd 359.1733, found 359.1735.
2,3-Di-O-benzyl-4,5-O-isopropylidene-D-arabinose Diethyl Dithio-
acetal. 4,5-O-Isopropylidene-D-arabinose diethyl dithioacetal34 (2.0 g,
6.75 mmol) was dissolved in anhydrous DMF (50 mL). Deoiled sodium
hydride (324 mg, 13.5 mmol) was added in one portion to the stirred
solution. After 1 h, the solution was cooled in an ice bath, and benzyl
bromide (3.2 mL, 27.0 mmol) was added slowly. Another 324 mg of
sodium hydride was added after 1 h, and the solution was stirred for
an additional 12 h. The reaction was quenched with methanol and
then with saturated aqueous ammonium chloride. As much solvent as
possible was removed with a rotary evaporator, and then the residue
was partitioned between saturated aqueous ammonium chloride and
methylene chloride. The aqueous layer was extracted three times with
methylene chloride, and then the organic extracts were combined and
dried over sodium sulfate. After removal of solvent in vacuo,
purification was via flash column chromatography (silica, hexanes to
3:1 hexanes/ethyl acetate gradient), to obtain 2.48 g (77%) of the title
compound. 1H NMR (360 MHz): δ 7.41-7.27 (m, 10), 4.93 (d, 1, J
) 10.9), 4.80 (d, 1, J ) 11.4), 4.76 (d, 1, J ) 11.5), 4.75 (d, 1, J )
11.0), 4.25 (dd, 1, J ) 12.7, 6.4), 4.18-4.09 (m, 2), 4.04 (dd, 1, J )
(R,S)-Methyl
2-(2-(N-(4-Methoxybenzyl)formamido)phen-
ylacetamido)acetate (26). Yield: 43%. IR (neat): 3304, 2953, 1754,
1663, 1514, 1250, 1211, 1179. The title compound exists at room
temperature as a 1:1 mixture of rotamers. When peaks corresponding
to the same proton(s) from each rotamer can be identified, they are
listed together. 1H NMR (400MHz): δ 8.19 and 8.11 (s, 1), 7.28-
7.25 (m, 5), 7.15 and 7.04 (br t, 1), 7.12 and 6.89 (d, 2, J ) 8.7 and
8.6), 6.80 and 6.70 (d, 2, J ) 8.5 and 8.6), 5.75 and 5.04 (s, 1), 4.77
and 4.48 (d, 1, J ) 14.7 and 15.4), 4.24 and 4.21 (d, 1, J ) 15.3 and
14.7), 4.02-3.86 (m, 2), 3.74 and 3.72 (s, 3), 3.68 and 3.67 (s, 3). 13
C
NMR (101 MHz): δ 169.8, 169.7, 169.2, 169.1, 163.8, 163.7, 159.0,
158.9, 134.1, 133.8, 129.9, 129.4, 129.0, 128.9, 128.7, 128.6 (2), 128.6,
128.2, 128.1, 113.9, 113.8, 64.0, 60.2, 55.1, 52.2, 52.1, 49.6, 46.5, 41.2,
41.1. HRMS (EI): m/z (M+) calcd 370.1529, found 370.1532.
(R,S)-Methyl 2-(2-((4-Methoxybenzyl)amino)phenylacetamido)-
acetate (27). IR (neat): 3322, 2926, 1752, 1671, 1514, 1248. 1H NMR
(400MHz): δ 7.86 (br t, 1), 7.42-7.30 (m, 5), 7.27 (d, 2, J ) 8.5),

Modifications of Ugi Condensation Products
J. Am. Chem. Soc., Vol. 118, No. 11, 1996 2583
8.3, 6.3), 3.90 (dd, 1, J ) 8.3, 6.7), 3.81 (dd, 1, J ) 6.5, 4.4), 2.75-
2.62 (m, 4), 1.43 (s, 3), 1.34 (s, 3), 1.26 (t, 3, J ) 7.4), 1.24, t, 3, J )
7.5).
mmol) and p-methoxybenzylamine (16 µL, 0.125 mmol) were combined
in 500 µL of methanol with 4 Å molecular sieves. The solution was
stirred at 23 °C for 1 h, and then 8 (100 µL of 1 M solution in hexanes,
0.1 mmol) was added, followed by anthranilic acid (14 mg, 0.1 mmol).
After stirring at 23 °C for 18 h, the solution was filtered through Celite,
the solvent removed in vacuo, and the residue purified via flash column
chromatography (silica, hexanes to 1:1 hexanes/ethyl acetate gradient)
to yield 32 mg (73%) of a clear glass. Characterization by NMR was
made difficult by amide rotational isomerism that broadened all peaks.
IR (neat): 2934, 1684, 1615, 1514, 1246. 1H NMR (400 MHz): δ
8.25 (br, 1), 7.20-6.72 (br, 8), 5.94 (br, 1), 4.7 (br d, 1), 4.44 (d, 1, J
) 14.9), 4.15 (br d, 1), 3.74 (s, 3), 2.65 (br, 1), 2.05 (br, 4), 1.63-1.54
(m, 4), 0.97-0.86 (m, 6). HRMS (EI): m/z (M+) calcd 435.2522, found
435.2528.
2,3-Di-O-benzyl-4,5-O-isopropylidene-D-arabinose (46).36 The
diethyl dithioacetal of the title compound (766 mg, 1.61 mmol) was
dissolved in 41 mL of acetonitrile and 11 mL of water. 2,6-Lutidine
(1.31 mL, 11.3 mmol) that had been passed through basic alumina was
added. To the stirred solution was added a slurry of N-bromosuccin-
imide (1.72 g, 9.65 mmol) in acetonitrile (7.5 mL). The solution turned
orange-brown. After 10 min, the reaction was quenched by dropwise
addition of aqueous NaHSO3 until the color disappeared. The reaction
mixture was diluted with saturated aqueous sodium bicarbonate and
extracted with methylene chloride three times. The organic layers were
combined and dried over sodium sulfate, and the solvent was removed
in vacuo. Purification was via flash column chromatography (silica,
hexanes to 3:1 hexanes/ethyl acetate gradient). Yield: 296 mg (50%).
1H NMR (400 MHz): δ 9.72 (d, 1, J ) 1.3), 7.37-7.27 (m, 10), 4.77
(d, 1, J ) 11.8), 4.65 (d, 1, J ) 11.8), 4.57 (s, 2), 4.25 (dd, 1, J )
12.6, 6.3), 4.07 (dd, 1, J ) 8.5, 6.3), 4.03 (dd, 1, J ) 3.1, 1.3), 3.98
(dd, 1, J ) 6.4, 3.1), 3.94 (dd, 1, J ) 8.5, 6.2), 1.42 (s, 3), 1.35 (s, 3).
13C NMR (101 MHz): δ 203.0, 137.3, 136.8, 128.4, 128.3, 128.2, 128.1,
128.0, 127.9, 108.9, 83.9, 79.6, 75.1, 74.3, 73.6, 66.5, 26.5, 25.1.
D-Arabinose 4CC Product 47. Compound 47 was prepared using
the general procedures described above for 4CC, except that the reaction
was run at 0 °C for 18 h in order to improve the diastereoselectivity.
Yield: 85% of a 3.4:1 mixture of inseparable diastereomers at the C-2
position. IR (neat): 2934, 1684, 1638, 1514, 1248. NMR data are
listed for each diastereomer, when resonances from each can be
identified. 1H NMR (500 MHz): (major diastereomer) δ 7.41-7.27
(m, 10), 7.10 (d, 2, J ) 8.8), 6.80 (d, 2, J ) 8.6), 5.96 (br, 1), 4.95
(dd, 1, J ) 15.5, 11.6), 4.79-4.26 (m, 6), 4.25-4.00 (m, 4), 4.08 (td,
1, J ) 7.7, 2.1), 3.80 (s, 3), 2.11 (s, 3), 2.12-1.97 (m, 4), 1.66-1.55
(m, 4), 1.49 (s, 3), 1.38 (s, 3); (minor diastereomer) δ 7.56 (br, 1),
7.41-7.27 (m, 10), 6.89 (d, 2, J ) 7.2), 6.87 (d, 2, J ) 7.2), 6.05 (br,
1), 5.22 (d, 1, J ) 11), 4.79-4.26 (m, 6), 4.25-4.00 (m, 3), 3.82 (s,
3), 3.69 (d, 1, J ) 6.3), 3.60 (dd, 1, J ) 7.0, 2.0), 2.07 (s, 3), 2.12-
1.97 (m, 4), 1.66-1.55 (m, 4), 1.48 (s, 3), 1.35 (s, 3). 13C NMR (126
MHz): (both diastereomers) δ 173.1, 172.9, 167.2, 167.1, 166.9, 166.6,
159.0, 158.8, 138.6, 138.3, 137.8, 137.7, 132.3, 132.1, 128.3 (2), 128.2
(2), 128.1 (4), 127.9, 127.8, 127.7, 127.5 (2), 127.4, 127.3, 114.2, 114.1
(2), 113.4, 113.1, 112.8, 112.6, 109.1, 108.4, 108.2, 80.1, 78.9, 77.4,
76.3, 76.0, 75.1, 74.7, 74.6, 74.5, 74.1, 73.1, 72.4, 66.6, 66.5, 66.2,
60.0, 55.1 (2), 50.7, 27.7 (2), 26.6, 26.4, 25.2, 25.1, 24.8, 23.9, 23.8,
22.6, 22.4 (3), 22.3, 22.9, 21.8. HRMS (EI): m/z [(M + H)+] calcd
657.3495, found 657.3516.
(R,S)-3-Isopropyl-4-(4-methoxybenzyl)-1,4-benzodiazepine-2,5-di-
one (53). Compound 52 (33 mg, 0.076 mmol) was azeotropically dried
with benzene and then dissolved in 1 mL of methanol. After adding
distilled acetyl chloride (27 µL, 0.38 mmol), the flask was capped and
then heated to 55 °C for 6 h. After removal of solvent in vacuo, the
residue was purified via flash column chromatography (silica, 2:1
hexanes/ethyl acetate to 1:2 gradient). Yield: 21 mg (82%) of a clear
glass. IR (neat): 3220, 2967, 1686, 1632, 1514, 1248. 1H NMR (500
MHz): δ 8.77 (s, 1), 8.00 (dd, 1, J ) 7.9, 1.5), 7.43 (td, 1, J ) 7.9,
1.5), 7.32 (d, 2, J ) 8.6), 7.23 (td, 1, J ) 7.8, 0.9), 6.88 (d, 1, J )
8.0), 6.83 (d, 2, J ) 8.6), 5.15 (d, 1, J ) 14.4), 4.45 (d, 1, J ) 14.4),
3.74 (s, 3), 3.62 (d, 1, J ) 11.5), 1.71-1.63 (m, 1), 0.82 (d, 3, J )
6.6), 0.65 (d, 3, J ) 6.6). 13C NMR (126 MHz): δ 171.9, 166.1, 159.3,
134.6, 132.5, 131.5, 130.2, 128.6, 126.8, 124.8, 119.7, 114.0, 71.0,
55.2, 54.7, 27.5, 19.6, 19.4. HRMS (EI): m/z (M+) calcd 338.1630,
found 338.1633.
(R,S)-N-(1-Cyclohexenyl)-2-(N′-butyl-2-amino-5-iodobenzamido)-
phenylacetamide (54). See the procedures for 52. Yield: 28%. IR
(neat): 3349, 2930, 1680, 1613, 1483, 1424. 1H NMR (400 MHz): δ
7.44-7.37 (m, 8), 6.63 (br, 1), 6.46 (d, 1, J ) 10.3), 6.02 (br, 1), 5.77
(br, 1), 4.57 (br, 1), 3.45-3.33 (m, 1), 3.24-3.21 (m, 1), 2.09-2.04
(m, 4), 1.66-1.63 (m, 2), 1.57-1.53 (m, 2), 0.90-0.86 (m, 4), 0.53
(m, 3). 13C NMR (101 MHz): δ 170.2, 168.0, 143.6, 138.7, 135.4,
134.8, 132.3, 129.9, 129.1, 129.0, 123.5, 117.9, 114.4, 64.8, 48, 31.6,
27.8, 23.9, 22.4, 21.8, 19.6, 14.2. HRMS (EI): m/z (M+) calcd
531.1383, found 531.1374.
(R,S)-4-Butyl-7-iodo-3-phenyl-1,4-benzodiazepine-2,5-dione (55).
See the procedures for 53. Yield: 66%. IR (neat): 3214, 2957, 1688,
1620, 1480. 1H NMR (500MHz): δ 8.49 (s, 1), 8.04 (s, 1), 7.44 (d, 1,
J ) 8.0), 7.18-7.07 (m, 5), 6.46 (d, 1, J ) 8.4), 5.36 (s, 1), 4.08 (br,
1), 3.66 (ddd, 1, J ) 13.7, 9.0, 5.5), 1.81 (br, 1), 1.75-1.70 (m, 1),
1.43 (tq, 2, J ) 7.4, 7.4), 0.98 (t, 3, J ) 7.3). 13C NMR (126 MHz):
δ 171.2, 165.4, 140.6, 139.6, 133.6, 132.9, 129.6, 129.0, 128.7, 128.2,
124.4, 121.4, 88.1, 66.8, 30.1, 20.0, 13.8. HRMS (EI): m/z (M+) calcd
434.0491, found 434.0480.
2-(N-(4-Methoxybenzyl)acetamido)-2-deoxy-3,4-di-O-benzyl-D-
mannono-δ-lactone (48). Compound 47 (3.4:1 mixture of diastereo-
mers, 63 mg, 0.10 mmol) was azeotropically dried with toluene and
then dissolved in 4.4 mL of methanol. Distilled acetyl chloride (34
µL, 0.48 mmol) was then added and the solution stirred at 23 °C for 2
h. The reaction mixture was poured into saturated aqueous sodium
bicarbonate and extracted three times with ethyl acetate. After drying
with sodium sulfate, the solvent was removed in vacuo, and the residue
was purified by preparative TLC (silica, 0.5 mm thickness, 20 cm ×
20 cm) with 1:2 hexanes/ethyl acetate as eluent. Yield: 14 mg (29%)
Acknowledgment. This work has been supported in part
by an Office of Naval Research Graduate Fellowship (T.A.K.)
and by funding from the National Institutes of Health (Grant
GM51095).
of the major diastereomer. [R]22 ) +44.2° (c ) 0.013, CHCl3). IR
D
(neat): 3418, 2930, 1748, 1642, 1514, 1248. 1H NMR (400 MHz): δ
7.39-7.27 (m, 12), 6.84 (d, 2, J ) 8.7), 4.89 (d, 1, J ) 11.7), 4.86 (d,
1, J ) 11.0), 4.79 (d, 1, J ) 11.0), 4.64 (d, 1, J ) 11.7), 4.40 (d, 1, J
) 15.6), 4.37 (dt, 1, J ) 9.5, 2.3), 4.29 (d, 1, J ) 16.1), 4.28 (t, 1, J
) 9.3), 3.96 (dd, 1, J ) 12.8, 2.3), 3.84 (dd, 1, J ) 12.8, 2.4), 3.82-
3.77 (m, 1), 3.77 (s, 3), 3.55 (d, 1, J ) 8.8), 1.95 (s, 3). 13C NMR
(101 MHz): δ 170.9, 167.1, 159.2, 137.9, 137.5, 129.0, 128.6, 128.5,
128.4, 128.1, 128.0, 127.9, 127.4, 114.0, 80.1, 78.3, 75.9, 75.0, 74.9,
62.1, 60.2, 55.2, 54.5, 21.3. HRMS (FAB): m/z [(M + H)+] calcd
520.2335, found 520.2328.
Supporting Information Available: Text giving charac-
terization data for compounds 40 and 41 and figures showing
1H NMR spectra of compounds 11, 12, 24, 26, 27, 29, 34-41,
46-48, and 52-55 and several unnumbered intermediates (24
pages). This material is contained in many libraries on
microfiche, immediately follows this article in the microfilm
version of the journal, can be ordered from the ACS, and can
be downloaded from the Internet; see any current masthead page
for ordering information and Internet access instructions.
(R,S)-N-(1-Cyclohexenyl)-2-(N′-(4-methoxybenzyl)-2-aminoben-
zamido)-3-methylbutanamide (52). Isobutyraldehyde (9 µL, 0.1
JA953868B