Practical synthesis of 1-aryl-6-(hydroxymethyl)-2-ketopiperazines via a 6-exo amide-epoxide cyclization

Article abstract of DOI:10.1021/ol048235t

(Chemical Equation Presented) Chiral 1-aryl-6-(hydroxymethyl)-2- ketopiperazines can be prepared via an operationally simple, 6-exo epoxide ring-opening cyclization to form the ketopiperazine C6-N1 bond in high yields and with excellent enantiomeric purity.

Full text of DOI:10.1021/ol048235t

ORGANIC
LETTERS
2004
Vol. 6, No. 22
4069-4072
Practical Synthesis of
1-Aryl-6-(hydroxymethyl)-2-ketopiperazines
via a 6-exo Amide−Epoxide Cyclization
Noel A. Powell,* Fred L. Ciske, Emma C. Clay, Wayne L. Cody,
Dennis M. Downing, Peter G. Blazecka, Daniel D. Holsworth, and
Jeremy J. Edmunds
Pfizer Global Research & DeVelopment, Michigan Laboratories, 2800 Plymouth Road,
Ann Arbor, Michigan 48105
noel.powell@pfizer.com
Received September 1, 2004
ABSTRACT
Chiral 1-aryl-6-(hydroxymethyl)-2-ketopiperazines can be prepared via an operationally simple, 6-exo epoxide ring-opening cyclization to form
the ketopiperazine C6 N1 bond in high yields and with excellent enantiomeric purity.
−
Ketopiperazines are valuable motifs that have found wide-
spread use as conformationally constrained peptidomimetics,
as well as synthetic intermediates for the construction of
biologically active molecules in total synthesis and the
pharmaceutical industry.1 During a recent medicinal chem-
istry program, we became interested in developing a synthesis
towards the chiral 1-aryl-6-hydroxymethyl-2-oxopiperazine
1 for use as a central scaffold for further manipulation.
derivatives as the source of the C6 chiral center. DIBAl-H
reduction of the commercially available ester 2 provided
Garner’s aldehyde 3.3 Reductive amination of the crude
aldehyde 3 with ethyl N-benzylglycine afforded the tertiary
amine 4 in good overall yield. Removal of the N-Boc
protecting group and cyclization to the ketopiperazine were
accomplished in one pot by treatment with aqueous HCl in
refluxing MeOH to provide the chiral 6-hydroxymethyl-2-
ketopiperazine 5 in excellent yield. The N-benzyl protecting
group was converted to a N-Boc carbamate by hydrogenoly-
sis in the presence of Boc2O to afford ketopiperazine 6. The
coupling of 6 and 4-benzyloxy-1-iodobenzene 7 was ac-
complished by treatment with 5 mol % CuI and 10 mol %
N,N′-dimethylethylenediamine,2 giving 8 in good yield. The
phenolic benzyl group was deprotected by hydrogenolysis
to give 1 in excellent yield.
In the early stages of the program, we were able to prepare
gram quantities of 1 via a Buchwald amidation coupling
reaction2 between ketopiperazine 6 and 4-benzyloxy-1-
iodobenzene 7 (Scheme 1). This route utilized serine
Although the above synthesis satisfied our needs during
the early stages of the project, the starting materials proved
(2) Klapars, A.; Huang, X.; Buchwald, S. L. J. Am. Chem. Soc. 2002,
124, 7421-7428.
(3) Garner, P., Park, J. M. Org. Synth., Collect. Vol. IX 1998, 300-305.
(1) Dinsmore, C. J.; Beshore, D. C. Org. Prep. Proc. Int. 2002, 34, 367-
404.
10.1021/ol048235t CCC: $27.50
© 2004 American Chemical Society
Published on Web 10/02/2004

have also reported a 5-exo example to form a fused bi-
cyclic pyrrolidinone ring, although formation of the 6-endo
product competed with the desired 5-exo pathway.5 Although
all of the above examples utilized conformationally con-
strained bicyclic cyclization precursors, we were encouraged
to investigate this route toward the ketopiperazine ring
structure.
Scheme 1
Our revised synthesis began with the acylation of 4-ben-
zyloxyaniline 12 with chloroacetyl chloride to provide the
chloroacetamide 13 in excellent yield (Scheme 2). The use
of an inorganic base allowed for the simple filtration and
precipitation of 13. Alkylation of BnNH2 with 13 provided
the secondary amine 10 in 93% yield. Condensation of amine
10 with (S)-epichlorohydrin 11 in the presence of MgSO4
afforded the desired chlorohydrin 14 in quantitative crude
yield.6 Attempts to purify the crude chlorohydrin by chro-
matography on silica gel gave variable yields of pure 14.
The crude product was therefore used without purification
and briefly treated with cold 5% aqueous NaOH for 20 min
to initiate formation of the terminal epoxide 9. Treatment of
9 with NaH in anhydrous DMF initiated the desired 6-exo
cyclization; however, the desired ketopiperazine product 15
was isolated in only a disappointing 26% yield. The
remaining mass balance consisted of 4-benzyloxyaniline and
numerous other unidentified polar byproducts. The hydrolytic
mechanism responsible for formation of the 4-benzyl-
oxyaniline byproduct under the anhydrous conditions
remains unclear. Further, the low yield of 15 proved to be
highly variable under these conditions and was success-
ful only in DMF. Other solvents (THF, THF/DMF mixtures,
and toluene) led exclusively to the formation of polar
byproducts.
to be cost prohibitive and of low atom efficiency for synthesis
of multigram quantities of 1. To achieve a more practical
synthesis, we chose to investigate an alternative retrosynthesis
that formed the ketopiperazine ring at the N1-C6 bond
through an intramolecular 6-exo cyclization of amide-
epoxide 9 (Figure 1). Further retrosynthetic cleavage of the
However, we were greatly pleased to discover that
prolonged exposure of chlorohydrin 14 to aqueous 5%
NaOH at room temperature led to cyclization and the
formation of ketopiperazine 15 in a greatly improved 77%
yield. TLC monitoring of the reaction indicated that the
addition of NaOH to chlorohydrin 15 led to the rapid
formation of epoxide 9, followed by a slow cyclization and
formation of 15. Hydrogenation of 15 in the presence of
Boc2O led to facile formation of the desired product 1 in an
excellent yield.
Figure 1. Revised retrosynthesis.
This synthetic route to ketopiperazine 1 satisfied our
requirements for a concise and robust synthesis that was
amenable to the preparation of >200 g quantities of 1. No
chromatography was required, as all intermediates could be
isolated by precipitation or used without purification. To
determine the enantiomeric purity of this route, the enan-
tiomeric ketopiperazine 16 was prepared in an analogous
fashion, using (R)-epichlorohydrin. Chiral HPLC analysis of
the enantiomers 1 and 16 indicated that the epichlorohydrin
N4-C5 bond revealed the R-aminoacetamide 10 and (S)-
epichlorohydrin 11 as the source of the C6 chiral center. The
intermolecular nucleophilic opening of epoxides by amides
is well precedented in the literature; however, fewer ex-
amples of intramolecular ring-opening reactions of un-
activated epoxides with amides exist. A synthesis of hetero-
norbornanes via a 5-exo amide-epoxide cyclization was
reported by Spurlock and co-workers, as well as an ex-
ample of a conformationally constrained 6-exo cycliza-
tion to form an azaadamantanol.4 Kibayashi and co-workers
(5) Abe, H.; Aoyagi, S.; Kibayashi, C. J. Am. Chem. Soc. 2000, 122,
4583-4592.
(6) (a) Bergeron, R. J.; Yao, G. W.; Yao, H.; Weimar, W. R.; Sninsky,
C. A.; Raisler, B.; Feng, Y.; Wu, Q.; Gao, F. J. Med. Chem. 1996, 39,
2461-2471. (b) Bergeron, R. J.; Ludin, C.; Mu¨ller, R.; Smith, R. E.;
Phanstiel, O. J. Org. Chem. 1997, 62, 3285-3290.
(4) Schultz, R. J.; Staas, W. H.; Spurlock, L. A. J. Org. Chem. 1973,
38, 3091-3093.
4070
Org. Lett., Vol. 6, No. 22, 2004

Scheme 2
condensation and cyclization steps occurred without any
racemization of the enantiomeric center.7
Table 1. Scope of the Cyclization of N-Aryl Acetamides 17
Having developed a novel intramolecular 6-exo amide-
epoxide cyclization route to 1-aryl-6-(hydroxymethyl)-2-
ketopiperazine 1, we were interested in exploring the scope
and limitations of this methodology. We prepared a number
of 2-benzylamino-N-aryl-acetamides 17a-f and subjected
them to the standard epichlorohydrin condensation/cycliza-
tion conditions (Table 1). N-Aryl acetamides 17a-c, con-
taining halides and alkyl groups, underwent smooth cycliza-
tion under the standard reaction conditions to afford the
desired ketopiperazines 18a-c in excellent yields. We were
particularly pleased to find that the sterically hindered ortho-
substituted substrates 17d-f underwent smooth cyclization
to yield the corresponding ketopiperazines 18d-f in good
yield.
A significant limitation of this methodology was discov-
ered when chlorohydrin 19, containing an electron-withdraw-
ing 4-diethylsulfonamide substituent, was subjected to the
standard NaOH cyclization conditions (Scheme 3). In this
case, diol 20 was the major product formed, with only trace
amounts of the desired ketopiperazine 21 observed. Other
substrates containing para electron-withdrawing groups (4-
(7) Absolute configurations of 1 and 16 were determined unambiguously
following conversion into the corresponding final analogues and X-ray
crystallography of the enzyme-analogue complexes. See: Powell N. A.;
Clay, E. H.; Holsworth, D. H.; Edmunds, J. J.; Jalaie, M.; Bryant, J. W.;
Zhang, E., Bioorg. Med. Chem. Lett. 2004, submitted.
Org. Lett., Vol. 6, No. 22, 2004
4071

formation of the desired ketopiperazine 21 in 54% yield.
Formation of diol 20 was not observed under these condi-
tions. This result contrasts with the poor yields encountered
in the attempted deprotonation and cyclization of epoxide 9
with NaH (Scheme 2). While NaH is compatible with
electron-poor aryl acetamides, we hypothesize that the
presence of electron-donating aryl substituents inductively
destabilizes the amide anion and promotes undesirable side
reactions. The equilibrium conditions established with the
weaker NaOH base in a protic medium circumvent these side
reactions by avoiding the formation of acetamide anion.
Scheme 3
In conclusion, we have developed a concise and practical
synthesis of chiral 1-aryl-6-(hydroxymethyl)-2-ketopiper-
azines via an intramolecular 6-exo amide-epoxide cycliza-
tion that is amenable to large-scale synthesis. This methodol-
ogy is applicable to a variety of electron-rich N-aryl
acetamides. Conditions have also been developed to allow
for the successful cyclization of substrates with electron-
withdrawing substituents. Further elaboration of the 1-aryl-
6-(hydroxymethyl)-2-ketopiperazines in the context of our
medicinal chemistry program will be reported in due course.
CN and 4-NO2) behaved similarly (data not shown). The
competitive formation of the corresponding diols was not
observed in the previous relatively electron-rich substrates
17a-f. It was expected that the cyclization of N-aryl
acetamides with aryl electron-withdrawing substituents would
be more facile, due to the inductive decrease of the acetamide
NH pKa by the para electron-withdrawing aryl substituents.
The preferential formation of diol 20 indicates that nucleo-
philicity of the amide anion is reduced by the inductive
stabilization by the electron-withdrawing aryl substituent,
resulting in a slower rate of cyclization and competitive
epoxide ring opening by hydroxide anion.
Acknowledgment. We thank Mark Lovdahl and Norm
Colbry for technical assistance in the high-pressure reactions,
Partha Mukherjee and Frank Riley for chiral HPLC analysis,
and Dana DeJohn for obtaining the mass spectral data.
Supporting Information Available: Experimental pro-
cedures and characterization information for compounds 1,
8, 15, 16, 18a-f, and 21. This material is available free of
charge via the Internet at http://pubs.acs.org.
We theorized that the use of a nonnucleophilic base in an
aprotic solvent would circumvent the competitive epoxide
ring-opening side reaction. Accordingly, treatment of chloro-
hydrin 19 with NaH in DMF led to smooth cyclization and
OL048235T
4072
Org. Lett., Vol. 6, No. 22, 2004