Convenient synthesis and isolation of 1-aminocyclopropane-1-carboxylic acid (ACC) and N-protected ACC derivatives

Article abstract of DOI:10.1055/s-2004-834793

A convenient route to 1-aminocyclopropane-1-carboxylic acid (1, ACC) and N-protected derivatives was developed. This route utilizes a bisalkylation of an O-benzyl glycine derived imine followed by global deprotection via hydrogenation. Direct isolation of ACC from a non-aqueous stream or efficient conversion to N-protected derivatives in a single flask is described.

Full text of DOI:10.1055/s-2004-834793

LETTER
2489
Convenient Synthesis and Isolation of 1-Aminocyclopropane-1-carboxylic
Acid (ACC) and N-Protected ACC Derivatives
Synthesis and
h
Isolation of
1
a
-Aminocyc
w
lopropane-1-carb
n
Michael H. Kress,a Ulf H. Dollingc
a
Merck Research Laboratories, Department of Process Research, 466 Devon Park Drive, Wayne PA 19087, USA
Fax +1(215)9932100; E-mail: shawn_allwein@merck.com
b
Merck Summer Intern
c
Merck Research Laboratories, Department of Process Research, 126 East Lincoln Avenue, Rahway NJ 07065, USA
Received 8 July 2004
O
Abstract: A convenient route to 1-aminocyclopropane-1-carboxyl-
N
Ph
ic acid (1, ACC) and N-protected derivatives was developed. This
route utilizes a bisalkylation of an O-benzyl glycine derived imine
followed by global deprotection via hydrogenation. Direct isolation
of ACC from a non-aqueous stream or efficient conversion to N-
protected derivatives in a single flask is described.
SMe
BnO
Ph
reported
herein
NC
NH2
H2N
CO2H
ref 6e
ref 6a
O
Key words: amino acids, imine alkylation, intramolecular alkyl-
ation, hydrogenation
NH2
HO
ref 6d
1, ACC
ref 6b
NHCOOBz
OMe
TMSO
OEt
ref 6c
HO
1-Aminocyclopropane-1-carboxylic acid (ACC) is a natu-
rally occurring amino acid that is involved in the regula-
tion of ethylene, a plant hormone.1 The initial discovery
that ACC inhibits tumor growth has sparked the evalua-
tion of a number of ACC derivatives in a wide range of
therapeutic areas including antileukemic,2 antitumor,3 an-
tibacterial,4 and neurodegenerative ailments.5
O
O
HO2C
CO2H
Figure 1
ethane electrophile that contained two reactive sites
would allow for bisalkylation resulting in cyclopropana-
tion.9 Choosing the correct halide combination would be
necessary to limit the possible elimination side product.10
The demand for a convenient synthesis of ACC and its
corresponding N-protected derivatives has resulted in a
number of novel approaches being reported (Figure 1).6
While these routes vary widely in transformation and
overall efficiency, a common hurdle has been identified.
ACC is highly water soluble, thus causing its isolation
from aqueous media to be difficult, relying on ion ex-
change or vacuum distillation to remove water. As a tes-
tament to these challenges, ACC continues to be an
expensive material with limited commercial availability.
When considering the N-acylated ACC derivatives, it be-
comes apparent that the most practical synthesis would
avoid the isolation of ACC completely.
The majority of the glycine imine alkylations have uti-
lized benzophenone-derived imines of either glycine ethyl
ester or t-butyl ester. However, the benzyl ester analog
appeared advantageous, as a global hydrogenation would
allow for a more efficient deprotection of both the ester
and imine. Furthermore, the benzyl functionality would
provide a useful label for monitoring the reaction and
competing ester hydrolysis by HPLC. The following syn-
thesis was designed with this in mind (Scheme 1).
From the commercially available glycine benzyl ester p-
toluenesulfonate salt (2), the bisalkylation precursor 3 was
generated via a modified transimination procedure origi-
nally described by O’Donnell.8 This transimination was
easily carried out on 100 g scale and was isolated by direct
We envisioned that various N-acylated ACC derivatives
could ultimately come from the bisalkylation of easily ob-
tainable glycine derivatives. The pivotal transformation in
our sequence was the bisalkylation of a glycine imine,7
which installed the cyclopropane ring in a single step.
This strategy is a logical expansion of the glycine imine
methodology developed by O’Donnell.7,8 Monoalkyla-
tions of a-substituted amino acid imines perform well re-
sulting in a tertiary center. While there are few examples
of intramolecular alkylations of imines, we believed an
O
O
Ph2CNH
N
Ph
NH2-TsOH
BnO
BnO
i-PrOAc
88%
2
3
Ph
O
KOH
NMP
N
Ph
BnO
Br(CH2)2Cl
Ph
SYNLETT 2004, No. 14, pp 2489–2492
Advanced online publication: 20.10.2004
2
2
.1
1
.2
0
0
4
4
80% HPLC Assay
65% Isolated
DOI: 10.1055/s-2004-834793; Art ID: S06504ST
© Georg Thieme Verlag Stuttgart · New York
Scheme 1

2490
S. P. Allwein et al.
LETTER
crystallization of the product from the reaction mixture in
88% yield. Upon formation of imine 3, the bisalkylation
was explored. Initial success using Cs2CO3 as the base in
N-methylpyrrolidine (NMP) gave 33% yield of the cyclo-
propane 4 (Table 1). A series of bases were screened in
different solvents that led to the discovery that pulverized
KOH11 was the base of choice. Further optimization found
the reaction at –5 °C using 1-bromo-2-chloroethane as the
electrophile resulted in an 80% HPLC assay yield of cy-
clopropane 4 (entry 7).12 Surprisingly, no significant elim-
ination product was detected. Isolation by direct
crystallization resulted in 65% yield of cyclopropane 4 as
a white solid. The corresponding alkylation with 1,2-di-
bromoethane resulted in large amounts of monoalkylation
and elimination products prior to stalling (entry 3). Earlier
iterations found when using Bu4NBr as an activator with
1,2-dibromoethane and KOH as the base, up to 30% of the
elimination product formed.13 The 1,2-dichloroethane
electrophile was even less reactive, resulting in no alkyl-
ation products under the reaction conditions, and warmer
temperatures resulted in only decomposition products as
well as unreacted starting material (entry 4).
Figure 2 Water effect on the bisalkylation: formation of benzyl
alcohol
Table 1 Bisalkylation of Imine 3 to Cyclopropane 4
Entry Base
X-(CH2)2-X Temp
(°C)
Time
(h)
Yield
(%)a
1
2
3
4
5
6
7
Cs2CO3
Cl/Br
Cl/Br
Br/Br
Cl/Cl
Cl/Br
Cl/Br
Cl/Br
r.t.
r.t.
0
24
1
33
44
46
0
CsOH
KOH
KOH
KOH
KOH
KOH
Figure 3 Water effect on the bisalkylation via formation of benzyl
alcohol near 9000 ppm water
5
r.t.
r.t.
0
5
the reaction mixture allowed for shorter reaction times (3
h) while limiting the decomposition of the benzyl ester.
These optimized conditions were successfully scaled to
50 g (80% assay yield).
1.5
1.5
6.5
49
75
80
–5
With imine 4 in hand, conversion to ACC was possible via
a global deprotection of the imine and benzyl ester. Hy-
drogenation of isolated imine benzyl ester 4 in methanol
with Pearlman’s catalyst provided a methanolic solution
of ACC with toluene and diphenylethane as the only by-
products.15 These by-products were easily removed in the
workup of the N-protected derivatives 5a–d. To the crude
hydrogenation stream was added a number of electro-
philes to give the N-protected ACC derivatives (5). In situ
alkylation to give 5a,b, and 5d utilized aqueous NaHCO3
as outlined by Paquet.16 The formation of the Fmoc-deriv-
ative 5d was generated in a modest yield when using ACC
(1) as a crude stream (40%). This transformation was
more successful using isolated ACC (1) as its starting ma-
terial (86% unoptimized). TFA–ACC (5c) could be gener-
ated in an 80% overall yield using Et3N and EtOTFA in
MeOH (Table 2).17
a HPLC assay yield.
The major by-product of the bisalkylation reaction was
hydrolysis of the benzyl ester resulting in benzyl alcohol
formation. When trying to suppress this hydrolysis path-
way by removing residual amounts of water, the rate of
the alkylation was slowed drastically (470 ppm water).14
This reduction in rate unfortunately allowed for the for-
mation of significant amounts of benzyl alcohol over
longer reaction times (Figure 2). Similarly, when per-
forming the bisalkylation under very wet conditions
(41000 ppm water), hydrolysis of the benzyl ester was
facile.
Upon recognizing that a relationship existed between the
water content, rate of reaction, and decomposition of the
benzyl ester, a range of optimal water content was estab-
lished. Further experimentation found a region between
8500 to 10000 ppm water to be most effective
(Figure 3).14 These results showed that targeting a Karl
Fischer titration (KF) of approximately 9500 ppm water in
Direct isolation of ACC was also possible by performing
the hydrogenation in wet methanol (5% water). Note,
this hydrogenation was run dilute (<0.1 M) in order to
keep the newly formed ACC in solution. The solubility of
ACC (1) in wet methanol is approximately 10 mg/mL.
Synlett 2004, No. 14, 2489–2492 © Thieme Stuttgart · New York

LETTER
Synthesis and Isolation of 1-Aminocyclopropane-1-carboxylic Acid
2491
(4) (a) Walsh, C. T.; Pascal, R. A.; Johnston, M.; Raines, R.;
Dikshit, D.; Krantz, A.; Honma, M. Biochemistry 1981, 20,
7509. (b) Chu, D. T. W.; Claiborne, A. K.; Clement, J. J.;
Plattner, J. J. Can. J. Chem. 1992, 70, 1328.
(5) (a) Spadoni, G.; Balsamini, C.; Bedini, A. Il Farmaco 1993,
48, 1663. (b) Berkowitz, D. B.; Pedersen, M. L. J. Org.
Chem. 1994, 59, 5476.
(6) (a) Rich, D. H.; Tam, J. P. Synthesis 1978, 46. (b) Fadel, A.
Tetrahedron 1991, 47, 6265. (c) Zhu, X.; Gan, P. Synth.
Commun. 1998, 28, 3159. (d) Strazewski, P.; Tamm, C.
Synthesis 1987, 298. (e) O’Donnell, M. J.; Bruder, W. A.;
Eckrich, T. M.; Shullenberger, D. F.; Staten, G. S. Synthesis
1984, 127.
(7) O’Donnell, M. J. Aldrichimica Acta 2001, 34, 3.
(8) O’Donnell, M. J.; Polt, R. L. J. Org. Chem. 1982, 47, 2663.
(9) Jabin, I.; Monnier-Benoit, N.; Gac, S. L.; Netchitailo, P.
Tetrahedron Lett. 2003, 44, 611.
Table 2 Formation and in situ Trap of ACC (1)
O
O
Pd(OH)2/C
MeOH
N
Ph
NH2
BnO
HO
Ph
4
1
O
H
N
Electrophile
Base
HO
R
5
Product R
Electrophile
Base
Yield (%)a
1
Direct isolation of 1
CBz-O-succinimide
Boc2O
96
99
75
80
40
5a18
5b19
5c
CBz
Boc
NaHCO3
NaHCO3
Et3N
(10) The elimination by-product is shown in Scheme 3.
COCF3 EtOTFA
Fmoc Fmoc-O-succinimide
O
O
5d
NaHCO3
N
Ph
N
Ph
base
BnO
BnO
a Isolated yield from 4.
Ph
X = halide
Ph
X
Filtration of the hydrogenation catalyst and concentration
of the filtrate gave ACC in high overall yield (96%).
Scheme 3
In addition, a through process for the synthesis of the N-
protected ACC derivatives was developed due to the high
mother liquor losses in the isolation of imine 4. To this
end, the bisalkylation reaction was extracted into toluene,
hydrogenated and treated with ethyltrifluoroacetate to
allow for direct isolation of TFA–ACC (5c) in a 52%
overall yield from 3 (Scheme 2).
(11) (a) in the heterogeneous bisalkylation, surface area and
particle size of the KOH appeared to be a key variable.
(b) Pulverized KOH was obtained by running KOH flakes in
a food processor for 2 min in a N2 filled glove bag to achieve
a powder.
(12) A representative procedure for the bisalkylation to 4 is as
follows: To a 1 L round bottom flask containing NMP (300
mL) was charged(diphenylmethylene)-glycine benzyl ester
(25.0 g, 76.0 mmol) and cooled to –5 °C. 1-Bromo-2-
chloroethane (6.90 mL, 83.6 mmol) was added followed by
portionwise addition of pulverized KOH (5 × 4.27 g, 380
mmol). The resulting orange slurry was aged for 9 h at –5 °C
at which time <0.5% starting material remained. The
reaction was then diluted with cold (2 °C) toluene (270 mL)
and quenched with H2O (250 mL). The organic layer was
separated and washed with H2O (2 × 250 mL).
1. Br(CH2)2Cl
O
O
H
KOH, NMP
N
Ph
N
CF3
BnO
HO
2. H2, Pd(OH)2/C
Toluene;
EtOTFA, Et3N
52% overall
Ph
O
3
5c
Scheme 2
Concentration and recrystallization from a heptane–toluene
(9:1) solution gave 49.3 g (65%) of a white crystalline solid.
Mother liquor losses accounted for 12% of 4, mp 50 °C. 1H
NMR (300 MHz, CDCl3): d = 7.66–7.64 (m, 2 H), 7.44–7.27
(m, 11 H), 7.22–7.18 (m, 2 H), 4.97 (s, 2 H), 1.56 (dd,
In conclusion, an efficient synthesis to ACC and its corre-
sponding N-protected derivatives has been developed.
This strategy utilizes a bisalkylation with 1-bromo-2-
chloroethane to insert the necessary cyclopropane ring.
Hydrogenation and direct isolation of ACC from a non-
aqueous solution has been demonstrated. Alternatively, a
single flask hydrogenation and alkylation of cyclopropane
imine 4 resulted in the formation of N-protected ACC de-
rivatives.
J = 7.5, 4.2 Hz, 2 H), 1.21 (dd, J = 7.5, 4.2 Hz, 2 H). 13
C
NMR (75 MHz, CDCl3): d = 174.7, 172.3, 140.1, 137.8,
135.9, 130.6, 129.0, 128.8, 128.6, 128.3, 128.2, 128.1, 66.5,
45.4, 20.5.
(13) Detected by crude NMR.
(14) Water content determined by coulometric Karl Fischer
titration with Metrohm 756 titrator.
(15) Attempts to bisalkylate the corresponding benzaldehyde
imine failed. This route would have been advantageous in
the hydrogenation step by generating a single inert by-
product, toluene.
(16) (a) The N-acylation to give 5a,b, and 5d from the crude
ACC stream followed: Paquet, A. Can. J. Chem. 1982, 60,
976. (b) Isolated materials matched the literature by 1H
NMR and 13C NMR (see ref.18,19).
(17) A representative procedure for the hydrogenation and in situ
N-protection to 5c is as follows: A 500 mL Parr hydro-
genation vessel was charged with [(diphenylmethyl-
ene)amino]cyclopropyl benzyl ester (4, 50.0 g, 141 mmol)
References
(1) (a) Yang, S. F.; Hoffman, N. E. Annu. Rev. Plant Physiol.
1984, 35, 155. (b) Pirrung, M. C.; Cao, J.; Chen, J. J. Org.
Chem. 1995, 60, 5790.
(2) Evano, G.; Schaus, J. V.; Panek, J. S. Org. Lett. 2004, 6, 525.
(3) (a) Ichihara, A.; Shiraishi, K.; Sakamura, S. Tetrahedron
Lett. 1977, 269. (b) Nara, S.; Toshima, H.; Ichihara, A.
Tetrahedron 1997, 53, 9509. (c) Braslau, R.; Anderson, M.
O. Tetrahedron Lett. 1998, 39, 4227.
Synlett 2004, No. 14, 2489–2492 © Thieme Stuttgart · New York

2492
S. P. Allwein et al.
LETTER
while adding heptane to 144 mg/mL. Concentration and
crystallization from heptane–i-PrOAc (8:1) gave 22.2 g
(80%, 113 mmol) of TFA–ACC(5c) as a white crystalline
solid, mp 174 °C. 1H NMR (300 MHz, CDCl3): d = 12.74 (s,
1 H), 9.91 (s, 1 H), 1.37 (dd, J = 7.8, 4.5 Hz, 2 H), 1.09 (dd,
J = 7.8, 4.5 Hz, 2 H). 13C NMR (75 MHz, CDCl3): d = 172.8,
157.8 (t, J = 36 Hz, 1C), 116.3 (t, J = 293 Hz, 1 C), 118.2,
114.3, 33.2, 16.5. 19F NMR: d = –75.0.
followed by MeOH (300 mL) and wet Pd(OH)2/C (12.5 g, 20
wt.% Pd). The mixture was shaken under 40 psi of H2 for 1.5
h then assayed to show quantitative formation of toluene and
diphenylmethane. The crude ACC stream was charged with
MeOH (36 mL), ethyl trifluoroacetate (50.4 mL, 422 mmol)
and Et3N (58.8 mL, 422 mmol) and heated to 50 °C for 3 h.
The slurry was cooled to r.t., filtered through a pad of solka
floc, and concentrated. The resulting residue was treated
with HCl (3 M, 166 mL), stirred for 30 min and washed with
heptane (3 × 140 mL). This removes the diphenylmethane
by-product. The aqueous solution was saturated with NaCl
and the TFA–ACC was extracted with i-PrOAc (3 × 140
mL). The combined TFA–ACC extracts were concentrated
(18) Chu, D. T. W.; Claiborne, A. K.; Clement, J. J.; Plattner, J.
J. Can. J. Chem. 1992, 70, 1328.
(19) Pirrung, M. C.; Cao, J.; Chen, J. J. Org. Chem. 1995, 60,
5790.
Synlett 2004, No. 14, 2489–2492 © Thieme Stuttgart · New York