Synthesis of a peroxime proliferator activated receptor (PPAR) α/γ agonist via stereocontrolled Williamson ether synthesis and stereospecific SN2 reaction of S-2-chloro propionic acid with phenoxides

Article abstract of DOI:10.1021/jo050268e

The stereospecific synthesis of the PPAR α/γ agonist 1 was accomplished via ethylation of the optically pure trihydroxy derivative 6, itself derived via an enzymatic resolution. The ethylation can be accomplished without epimerization only under strict control of the reaction conditions and the choice of base (sodium tert-amylate), temperature (-30°C), order of addition, and solvent (DMF). The key diastereospecific S_N2 reaction of the phenol 4 with S-2-chloropropionic acid is best achieved via the sodium phenoxide of 4 derived from Na⁰ as the reagent of choice. The structure elucidation and key purification protocols to achieve pharmaceutical purity will also be described

Full text of DOI:10.1021/jo050268e

Synthesis of a Peroxime Proliferator Activated Receptor (PPAR)
r/γ Agonist via Stereocontrolled Williamson Ether Synthesis and
Stereospecific SN2 Reaction of S-2-Chloro Propionic Acid with
Phenoxides
James A. Aikins,* Michael Haurez, John R. Rizzo, Jean-Pierre Van Hoeck, Willy Brione,
Jean-Paul Kestemont, Christophe Stevens, Xavier Lemair, Gregory A. Stephenson,
Eric Marlot, Mindy Forst, and Ioannis N. Houpis*
Chemical Product Research and Development, Lilly Development Centre, rue Granbonpre 11, B-1348
Mont-St-Guibert, Belgium, and Lilly Technology Center, B-110, Indianapolis, Indiana 46285
houpis_ioannis@yahoo.com
Received February 10, 2005
The stereospecific synthesis of the PPAR R/γ agonist 1 was accomplished via ethylation of the
optically pure trihydroxy derivative 6, itself derived via an enzymatic resolution. The ethylation
can be accomplished without epimerization only under strict control of the reaction conditions and
the choice of base (sodium tert-amylate), temperature (-30 °C), order of addition, and solvent (DMF).
The key diastereospecific SN2 reaction of the phenol 4 with S-2-chloropropionic acid is best achieved
via the sodium phenoxide of 4 derived from Na0 as the reagent of choice. The structure elucidation
and key purification protocols to achieve pharmaceutical purity will also be described
Introduction
attempt to balance clinical benefits by the simultaneous
activation of two of the above receptors.
Non-insulin-dependent diabetes mellitus (type 2 dia-
betes) is a serious disease that affects tens of millions of
people in the developed world. It is estimated that 120
million people worldwide suffer from this condition that
can affect the eyes, kidneys, and periferal nervous
system. In addition to the hyperglycemia that character-
izes the disease, there is frequently dyslipidemia associ-
ated with this condition, which accelerates coronary
arteriosclerosis and is thus linked to increased mortality
of these patients.1 Recent reports have identified that
selective agonists of a nuclear receptor family, the
peroxime proliferator activated receptors (PPAR) R, γ,
and δ, can concurrently normalize glucose levels and
lower lipid levels, thus making such agonists attractive
targets for the treatment of type 2 diabetes.2 Significant
attention has also been paid to dual agonists in an
In this paper, we would like to describe the total
synthesis of 1, the pharmacophore of our PPAR R/γ dual
agonist that is potentially useful for the treatment of type
2 diabetes and dyslipidemia.3
The initial goal of our synthetic strategy aimed to
install, stereospecifically, the two stereogenic centers, at
C-2 and C-10, in a way that avoids resolution or chiral
chromatography. Indeed initial syntheses of the molecule
(2) (a) Lehman, J. M.; Moore, L. B.; Smith-Oliver, T. A.; Wilkinson,
W. O.; Willson, T. M.; Kliewer, S. A. J. Biol. Chem. 1995, 270, 12953.
(b) Staels, B.; Dallongeville, J.; Auwerx, J.; Schoonjans, K.; Leitersdorf,
E.; Fruchart, J.-C. Circulation 1998, 41, 5020. (c) Oliver, M. F.; Heady,
J. A.; Morris, J. N.; Cooper, J. Lancet 1984, 2, 600.
(1) (a) Porte, D., Jr.; Schwartz, M. W. Science 1996, 27, 699. (b) Cobb,
J.; Dukes, I. Annu. Rep. Med. Chem. 1998, 33, 213. (c) Goldberg I. J.
J. Clin. Endocrinol. Metab. 2001, 86, 965 (d) Grag, A. Am. J. Cardiol.
1998, 81, 47 B.
10.1021/jo050268e CCC: $30.25 © 2005 American Chemical Society
Published on Web 05/05/2005
J. Org. Chem. 2005, 70, 4695-4705
4695

Aikins et al.
SCHEME 1. Retrosynthetic Options
relied on such techniques for the isolation of the desired
2-R,10-S diastereomer.4
consistent with scale-up principles. For example, inter-
mediate 2 could provide an excellent purification point
especially if perfect stereocontrol could not be achieved
in the earlier transformations. Consequently this step-
wise approach was selected for the synthesis of 1 and is
detailed below.
The key transform in this strategy (Scheme 1, option
1) involves a stereocontrolled displacement transforma-
tion to yield the activated lactic acid derivative 5 and the
phenol 4. The challenge to effect this transformation, in
the synthetic direction, would lie in effecting complete
control of the inversion or retention of the stereogenic
center in 5 (eventually C-10) under conditions that do
not induce epimerization of the chiral center in 4 (even-
tually C-2).
The coupling partner, 4, could be derived from 7, which
was readily available in our laboratories via an enzymatic
kinetic resolution of the racemic methyl ether, methyl
ester 8.6
An alternative approach to 6 could be also envisioned
(option 1, path B) via an asymmetric reduction of the
commercially available 4-hydroxyphenyl pyruvic acid
derivative 8a.
In addition, it was necessary to develop a process that
preserved the enantiomeric purity of the readily epimer-
izable stereogenic centers in subsequent reactions. Fur-
thermore, convenient purification points had to be es-
tablished to avoid chromatographic purifications that are
inconsistent with large-scale synthesis. This latter task
was not trivial because of the lipophilic nature of this
molecule and the absence of heteroaromatic rings.5
Herein we report the details of our synthetic and devel-
opment work that succeeded in meeting all of the above
goals.
Retrosynthetic Analysis. Two retrosynthetic strate-
gies were considered for the synthesis of 1 (Scheme 1
options 1 and 2).
Disconnection of bond A (Scheme 1, option 2) would
lead to a convergent approach to 1; however, that would
place the key transformation, setting the C-10 stereo-
center, at the last step of the synthesis. Consequently,
although on first principles convergent syntheses are
sought, in our case the strategy shown in option 1 is more
(3) A number of excellent reports for the action of PPAR agonists
have recently appeared in the literature. Further details on the
biological significance of these receptors can be found therein: (a)
Haigh, D.; Birrell, H. C.; Cantello, B. C. C.; Eggleston, D. S.;
Haltiwanger, R. C.; Hindley, R. M.; Ramaswamy, A.; Stevens, N. C.
Tetrahedron: Asymmetry 1999, 10, 1353. (b) Liu, K. G.; Smith, J. S.;
Ayscue, A. H.; Henke, B. R.; Lambert, M. H.; Leesnitzer, L. M.; Plunket,
K. D.; Willson, T. M.; Sternbach, D. D. Bioorg. Med. Chem. J. 2001,
11, 2385. (c) Sauerberg, P.; Pettersson, I.; Jeppeses, L.; Bury, P. S.;
Mogensen, J. P.; Wassermann, K.; Brand, C. L.; Sturis, J.; Woeldike,
H. F.; Fleckner, J.; Andersen, A.-S. T.; Mortesen, S. B.; Svensson, L.
A.; Rasmussen, H. B.; Lehmann, S. V.; Polivka, Z.; Sindelar, K.;
Panajotova, V.; Ynddal, L.; Wulff. E. M. J. Med. Chem. 2002, 45, 789.
(4) Potlapally, R. K.; Siripragada, M. R.; Kotra, N. M.; Sirisilla, R.;
Mamillapalli, R. S.; Gaddam, O. R. PTC WO 02/24625 A2, 2002.
(5) Initial attempts to identify suitable pharmaceutical and/or
“technical” salts for 1, 2c, and 4a met with limited success. Significant
experimentation was required to crystallize the salts mentioned herein.
Despite the advantages derived from the availability
of 7 in multikilo quantities, triggering the S-goal ret-
rosynthetic approach described above,7 considerable effort
was expended in the implementation of this strategy. The
problem lay in the fact that any base-mediated ethylation
reaction, on the way to 4, could potentially involve
significant risk of epimerization at C-2.
In the next section we describe a variety of solutions
that permit the use of 6. Recently, and in order to
(6) Rizzo, J.; Zhang, T. Personal communication. Deusen, H.-J.;
Zundel, M.; Valdois, M.; Lehmann, S. V.; Weil, V.; Hjort, C. M.;
Ostergaard, P. R.; Marcussen, E.; Ebdrup, S. Org. Process Res. Dev.
2003, 7, 82.
4696 J. Org. Chem., Vol. 70, No. 12, 2005

SCHEME 2. Synthesis of Intermediate 10a-ca
a (a) (1) NaI (1.2 equiv), 12 M HCl, reflux; (2) filtration at ambient temperature. (b) (1) K2CO3, EtOH, reflux; (2) removal of EtOH then
NaOH, H2O; (3) acidification and filteration. (c) K2CO3, EtOH, H2O for in situ conversion to 10a. (d) EtOH, H2S04 (cat.) for 10b. (e)
iPrOH, H2S04 (cat.) for 10c.
overcome the shortcomings of this synthetic strategy, a
totally different approach to 4 was discovered and will
be detailed elsewhere.8
added slowly to a solution of concentrated HCl in water,
containing NaI, and the mixture was heated to 100 °C
for 16 h. The product was easily obtained in 85-95% yield
by filtration at ambient temperature. On a small scale,
where heating times are short, significant frothing was
observed, presumably due to rapid evolution of CH3Cl/
CH3I. On a large scale this problem was easily avoided
by controlling the rate of heating and ensuring rigorous
stirring of the reaction mixture. On a 6 kg scale the
results were reproducible; however, care was required in
the isolation of 6 as the volume of solvent used for the
wash of the crude crystals and the temperature during
the wash needed to be controlled to minimize losses in
the mother liquors.
It is noteworthy that even under these relatively harsh
reaction conditions no epimerization was observed at C-2,
as shown by chiral HPLC analysis of the product. It is
indeed true, in all the cases we have investigated, that
the C-2 center is not prone to epimerization under acidic
conditions.
Selective protection of the phenol could be accom-
plished smoothly, in ca. 82% yield, by reacting 6 with
benzyl chloride in ethanol in the presence of K2CO3 at
reflux for ca. 24 h. The resulting bis-benzyl derivative 9
was not isolated; rather, the ethanol was evaporated in
vacuo and the mixture was treated with aqueous NaOH
to effect hydrolysis of the benzyl ester. Acidification of
the reaction mixture induced precipitation of the product,
which was isolated by filtration. Interestingly the trans-
formation of 6 to 10 can be accomplished directly by
performing the benzylation in aqueous EtOH. Under
these conditions 9, which is still formed as indicated by
HPLC, hydrolyzed in situ to afford 10a. The enantiomeric
purity of the product was established to be identical to
the starting material by chiral HPLC analysis.
Alternatively, following the literature procedure12 (eq
1) to produce enantiomerically enriched 10a, p-hydroxy-
phenyl pyruvic acid was reduced with (+)-DIP-Cl to
afford 6 in 90% ee and 92% chemical yield. Upgrade of
the enantiomeric purity can be performed after monoben-
zylation (as described above) by a single recrystallization
Results and Discussion
Synthesis of Ethoxy Cinnamate Derivative 4. The
intense interest in PPAR agonists and the presence of
the R-alkoxy dihydrocinnamate pharmacophore in a
number of these drug candidates have spurred significant
research activity to discover an efficient and scaleable
synthesis of these molecules. The need for convenient
syntheses of these molecules is augmented by recent
reports that these pharmacophores could be useful for
the treatment of inflammatory disease.9
The initial focus on the synthesis of these R-alkoxy
esters was centered on the preparation of R-hydroxy
esters followed by alkylation, and so both enzymatic and
chemical methods have been designed to give the phenol,
monoprotected, C-2 hydroxy derivatives, e.g., 10b or 10c,
with high enantiomeric excess.10 These compounds can
be prepared via chem-enzymatic methodology developed
by Deng and co-workers, and a number of these deriva-
tives have become commercially available.11
In our laboratories, the synthesis of 4 commenced from
7 (Scheme 2), which was in turn synthesized from 8 via
a sequence developed in the Lilly Research Laboratories
involving (a) an enzymatic hydrolysis of rac-8, (b) de-
benzylation via catalytic hydrogenation, and (c) hydroly-
sis and crystallization of the sodium salt to afford
enantiomerically pure 7. This key intermediate was
(7) Corey, E. J.; Cheng X.-M. The Logic of Chemical Synthesis; John
Wiley & Sons: New York, 1989.
(8) Alt, C.; Haurez, M.; Forst, M.; Houpis, I. N.; Patterson, L.; Rizzo,
J.; Zhang, T. Org. Lett. 2005, in press.
(9) See ref 3 and Storz, T.; Dittmar, P. Org. Process Res. Dev. 2003,
7, 559
(10) (a) Wong, H. N. C.; Xu, Z. L.; Chang, H. M.; Lee, C. M. Synthesis
1992, 793 (b) Camps, P.; Perez, F.; Soldevilla, N. Tetrahedron:
Asymmetry 1997, 8, 1877 (c) Jeppesen, L.; Bury, P. s.; Sauerberg, P.
PCT WO 00/23415 A1, 2000. (d) Ebdrup, S.; Deusen, H.-J. W.; Zundel,
M. PCT WO 01/11072 A1, 2001. (e) Boije, M. Horvath, K. Inghardt, T.
PCT WO 01/40171 A1, 2001. (f) Nemoto, T.; Kakei, H.; Gnanadesikan,
V.; Tosaki, S.; Ohshima, T.; Shibasaki, M. J. Am. Chem. Soc. 2002,
124, 14544. (g) Linderberg, M. T.; Moge, M.; Sivadasan, S. Org. Process
Res. Dev. 2004, 8, 838.
(11) Liang, T.; Deng, L. J. Am. Chem. Soc. 2002, 124, 2870. This
technology has been licensed to the Diacel corporation, which has made
these derivatives commercially available.
(12) Wang, Z.; La, B.; Fortunak, J. M.; Meng, X.-J.; Kabalka, W.
Tetrahedron Lett. 1998, 39, 5501.
J. Org. Chem, Vol. 70, No. 12, 2005 4697

Aikins et al.
TABLE 1. Attempts at Alkylation of 10a-c
starting
material
solvent
mixture
alkylation
agenta
temp
(°C)
conversion
(yield, %)
entry
product
base
S:R ratiob
1
2
3
4
6
5
7
8
10b
10b
10b
10b
10b
10b
10b
10b
11b
11b
11b
11b
11b
11b
11b
11b
DMF
NaH, 1.05 equiv
NaH, 2 equiv
NaHMDS
EtI
-30
0
0
DMF
EtI
70-80
55
82:18
DMF
EtOTs
EtI
0
THF/DMF
THF/DMF
PhCH3/DMF
PhCH3
NaHMDS
0
-30
0
95
-30
80
80:20c
97:3d
70:30
NaHMDS
EtI
80-90
85
KHMDS
EtI
K2CO3
EtI
0
PhCH3/DMF
NaOtAmyl, 1.4 equiv
EtI
99 (70)e
96:4
a Routinely 5 equiv of alkylating agent was used. b Ratio by HPLC area % at 220 nm. c EtI added last to the mixture of the base and
hydroxy ester at -30 °C. d The base was added slowly to a mixture of the hydroxy ester and EtI. e Material balance is due to hydrolysis
of the product to give 11a.
from iPrOH to afford the desired product (10a) in 98%
gated that afforded mixtures of 10a, 11a, and 11b.
Furthermore the chiral assays of the 11b thus formed
indicated that up to 10% epimerization of the ethoxy
center had occurred.15 Although our downstream purifi-
cation protocols could cope with the undesired enanti-
omer (up to 11%), a greater issue was that the results
seemed to depend on the quality of NaH used. Both the
enantiomeric excess and conversion varied with different
lots of NaH, whereas in some cases, when incomplete
reaction was observed, further addition of NaH resulted
in extensive epimerization (g20%). To avoid the complex-
ity of effecting both transformations concurrently, the
ethyl and isopropyl esters were formed (see Scheme 2)
by acid-catalyzed esterification. In this manner, 10b and
10c can be conveniently prepared without racemization.
Alkylation attempts on those substrates are shown in
Table 1.
ee and 76% overall yield.13
Our screen initially revealed that the best conditions
involed NaHMDS in THF and EtI as the alkylating agent
at -30 °C, affording 11b in ca. 60% isolated yield and
>94% ee (Table 1, entry 6). However closer examination
of the reaction mixture indicated that competing silyla-
tion of the alkoxide occurred even at -30 °C to afford
the silyl ether 12 in 20-30% yield.16
Formation of Ethyl Ether 11. Contrary to the
plethora of synthetic approaches to the R-hydroxy ester
function, the ether formation at C-2 is considerably more
challenging. Despite intense research, significant epimer-
ization has been observed with a number of the alkylation
procedures described, and naturally, heavy reliance on
resolution or chiral chromatography protocols was neces-
sary for several of the syntheses. We have also been
successful in developing an efficient resolution by forma-
tion of the D-alaninol salt of 11a (11d); however, we
endeavored to effect this reaction with minimal epimer-
ization, to be able to productively use our optically pure
starting material 10.
Initial attempts to effect simultaneously alkylation of
both the alcohol and carboxylate functions were met with
modest success. Specifically, in our hands procedures
from the patent literature14 describing high yields and
enantiomeric excess for the transformation 10a f 11b
(PhCH3, K2CO3 EtI reflux) failed to produce any of the
desired product. Several other procedures were investi-
Fortunately, further screening of bases identified that
sodium tert-amylate (1 M PhCH3 solution) effected com-
(15) Alkylation screens for the conversion of 10a to 11b were
performed using the following combinations of components. Bases: Na,
K and Cs2CO3, Li, Na, K-HMDS, KOH, NaOH with or without phase
transfer catalysts, and NaH. Solvents: THF, PhCH3, DMF, NMP, CH2-
Cl2. Alkylating agents: EtI, Et2SO4.
(13) We thank Dr. Vincent Mancuso and his team for the execution
of the transformation and discovery of the crystallization protocol.
(14) Potlapally, R. K.; Sipirpagada, M. R.; Kotra, N. M.; Sirisilla,
R.; Mamillapalli, R. S.; Gaddam, O. R. PTC WO 02/24625 A2, 2002.
4698 J. Org. Chem., Vol. 70, No. 12, 2005

TABLE 2. Optimization Screen for the Synthesis of 11c
SCHEME 3. Impurities Produced in the Ethylation
Step
rate of
equiv equiv temp addition % H2O yield S:R
entry of base of EtI (°C)
(min)
in DMF (%)
ratio
1
2
3
4
5
1.4
1.5
1.2
1.2
1.2
5
5
5
2
5
-30
-30
-10
-30
-30
60
60
0.14
0.52
0.05
0.05
0.05
91
96:4
87
98:2
90:10
85:15
93:7
45
60
170
95
transesterification product 13. The latter two compounds
are anticipated to afford the desired final product.
Remarkably the orthometalation/alkylation product 14
was also detected, albeit in small amounts (<0.5%).
On a larger scale, the reaction performed with some
variability. Indeed some initial reactions, on the 1 kg
scale, were performed by addition of solid NaOtAmyl to
the reaction mixture instead of a toluene solution. Three
different batches were run to give 11c with enantiomeric
excess between 99.5% and 89%. It was speculated that
the variability was due to uncontrolled local warming as
the base dissolved in DMF, and consequently a solution
of NaOtAmyl in toluene was used, to afford a more
consistent 95-97% ee (5 batches, 700 g scale). Although
under strict control of the reaction conditions we obtained
useful ee’s of 11c, we developed a crystallization proce-
dure that could accommodate the ca. 89% ee values
observed under less rigorous reaction control. Specifically,
11c was hydrolyzed in EtOH and NaOH solution at
ambient temperature back to 11a followed by salt forma-
tion with D-alaninol in 2-propanol. In this manner,
material of 88% ee can be upgraded to 98% ee. Resub-
jecting the salt to H2SO4 in 2-propanol reforms 11c
without loss of enantiomeric purity. The D-alaninol salt,
11d, gave crystals suitable for X-ray crystallographic
analysis, and thus structural proof and the absolute
configuration of these intermediates was established via
X-ray crystallography.18
This variability along with the difficulties in using
reactive alkylating agents (EtI) prompted the search for
alternative approaches to the synthesis of 11b or 11c.
We chose a reductive alkylation method in which a silyl
ether is reacted with an aldehyde in the presence of BiBr3
and Et3SiH to afford the corresponding alkylation product
(eq 2). The mechanistic considerations and interesting
applications of this reaction have appeared recently.19
In our hands this reaction proceeded to afford useful
yields of the product 11b only when the hindered tBuMe2-
Si ether was used. With the less hindered Me3Si and Et3-
Si derivatives, the yield was ca. 15% and 30%, respec-
tively, with 10b comprising the remainder of the material
plete conversion to the desired product, although initially
with modest yield (Table 1, entry 8). Not surprisingly,
control experiments indicated that 11b epimerizes under
the reaction conditions, albeit at a slower rate than the
alkylation reaction. A short optimization study was
undertaken (Table 2) that allowed us to arrive at the
following conclusions: (a) The ethyl ester (10b) was
sensitive to both hydrolysis (to give 10a) and epimeriza-
tion, and thus the isopropyl ester 10c was chosen for this
reaction. (b) Five equivalents of EtI are required for
optimal reaction rate and enantiomeric excess. (c) It is
crucial to ensure that the base is consumed as rapidly
as possible to avoid epimerization of the product. Con-
sequently the EtI and 10c or 10b are premixed at -30
°C, followed by slow addition of the base while maintain-
ing the temperature of the reaction at e-30 °C to avoid
epimerization. This slow addition of the solution of tert-
amylate was particularly important toward the end of
the reaction where solid (NaI) precipitation caused a
sudden increase in the internal reaction temperature.
The amount of water in the reaction medium was also
monitored, as extensive hydrolysis was observed with
increasing amounts of water. Gratifyingly, this effect was
less severe when the isopropyl ester 10c was utilized.
Ironically, the selectivity of the reaction increased with
increasing amounts of H2O as less “live” sodium tert-
amylate was present in the medium. It is noteworthy that
the anhydrous NaOH generated during the reaction by
the presence of adventitious water does not seem to cause
epimerization of the product 11c (or 11b). At this point,
we cannot exclude the possibility that the H2O or result-
ing NaOH played a different beneficial role in this
transformation.
Thus the optimal conditions for the reaction involve
addition of sodium tert-amylate in toluene to a mixture
of 10c and EtI at -30 °C over 60 min, followed by quench
with AcOH and extractive workup. The product is
isolated in 86-88% yield with 96-98% ee.17 Several side
products were identified by mass spectrometry and
correlation to known compounds and are shown in
Scheme 3. In addition to the starting material 10c and
the hydrolysis byproduct 11a, 11b was also detected,
presumably from ethylation of 11a, as well as the
(18) Chiral analysis of 11d was established with chiral HPLC. X-ray
crystallographic data can be found in Supporting Information.
(19) We thank Prof. P. A. Evans for pointing out to us the reactivity
trends of the silyl ether derivatives in this reaction. (a) Bajwa, J. S.;
Jiang, X.; Slade, J.; Prasad, K.; Repic, O.; Blacklock, T. J. Tetrahedron
Lett. 2002, 43, 6709 (b) Evans, P. A.; Cui, J.; Gharpure, J. Org. Lett.
2003, 5, 3883. (c) Evans, P. A.; Cui, J.; Gharpure, J.; Hinkle, R. J. J.
Am. Chem. Soc. 2003, 125, 11456. (d) Evans, P. A.; Cui, J.; Gharpure,
J.; Polosukhin, A.; Zhang, H.-R. J. Am. Chem. Soc. 2003, 125, 14702
and references therein.
(16) In addition to the significant yield loss, all downstream
purification protocols developed could not remove 10b or 10c derived
byproducts from the final drug substance.
(17) The enantiomeric excess was determined by chiral electro-
phoresis analysis.
J. Org. Chem, Vol. 70, No. 12, 2005 4699

Aikins et al.
reaction of 16a or 16b with phenol 4b in the presence of
Cs2CO3 in DMF at ambient temperature resulted in
partial epimerization at C-10 as determined by chiral
HPLC analysis (eq 4).
To determine what species was responsible for this
epimerization several control experiments were per-
formed that showed that 16a or 17 do not readily
epimerize under the reaction conditions, so it is assumed
that the cesium phenoxide acted as the base responsible
for the observed loss of optical purity. This reaction was
not investigated further, as a more effective solution was
discovered.
In a related project in our laboratories, aimed at the
synthesis of arylated lactic acid derivatives, it was
discovered that coupling of phenol 18a (eq 5) with
R-chloropropionic acid (19a) in the presence of 2 equiv
of NaH afforded 20 in good yield. Furthermore the
product was shown to be a single enantiomer based on
chiral HPLC analysis.
balance. Nonetheless, in all three cases, the optical purity
at the C-2 center in 10b is perfectly preserved in 11b.
Although the chemical yield is relatively low, with some
significant hydrolysis of the silyl ether during the reac-
tion, this method provided an attractive alternative to
the base-mediated alkylation reaction and is being
explored further.
Debenzylation of 11c. Facile debenzylation of 11b
or 11c can be effected by heterogeneous catalytic hydro-
genation (eq 3) with 20 wt % of 10% Pd/C at 50 psi at
ambient temperature to afford the product in nearly
quantitative yield at the multikilo scale (eq 3).
Unfortunately we were, initially, unable to determine
the absolute stereochemistry of the product (20), as most
of the derivatives studied at the time failed to afford
crystalline chiral salts appropriate for single-crystal
X-ray crystallographic analysis. Nonetheless, we were
Coupling of 4c with Activated Lactic Acid De-
rivatives. Our initial attempts for the synthesis of 2 in
a diastereoselective manner involved the coupling of the
phenoxide of 11c with readily available lactic acid
derivatives, which would be appropriately activated for
the substitution reaction. Although initially this appeared
to be a fairly straightforward approach with literature
precedence,20 several technical problems needed consid-
eration. For example, one would need to avoid epimer-
ization or elimination reaction of the activated lactic acid
derivative 16 (eq 4) or epimerization of 4 under the basic
coupling conditions. The product, of course, is also
susceptible to epimerization. Indeed in preliminary work,
(20) Significant research has been devoted to the stereospecific
substitution of activated lactic acid derivatives with carbon, nitrogen,
oxygen. or sulfur nucleophiles. Some select examples are: (a) Effen-
berger, F.; Burkard, U.; Willfahrt, J. Angew. Chem. 1983, 95, 50. (b)
Burkard, U.; Effenberger, F. Chem. Ber. 1986, 119, 1594. (c) Sato, T.;
Otera, J. J. Org. Chem. 1995, 60, 2627 and references therein. (d)
Otera, J.; Nakazawa, K.; Sekoguchi, K.; Orita, A. Tetrahedron 1997,
53, 13633. (e) Larcheveque, M.; Petit, Y. Synthesis 1991, 162. (f)
Nestler, H. J.; Hoerlein, G.; Handte, R.; Bieringer, H.; Schwerdtle, F.;
Langelueddeke, P.; Frisch, P. U.S. Patent US005712226A, 1998 and
references therein
4700 J. Org. Chem., Vol. 70, No. 12, 2005

TABLE 3. Reaction of Phenol 4b-c with
gratified and intrigued by such complete control of
stereochemistry and excellent yield at refluxing THF (no
elimination of epimerization of 19a or 20) and postulated,
as have others, that formation of 21 followed by opening
of the R-lactone by the phenolate was responsible for the
control of the stereochemistry and stabilization of the
reactive intermediate (eq 6). Naturally, if such a mech-
anism were in effect, overall retention of the stereochem-
istry of the chloropropionic acid stereogenic center would
have been observed. An SN2 mechanism would, of course,
give inversion at that center.21
S-2-Chloropropionic Acid under Basic Conditionsa
starting
entry material
temp
(°C)
2a
(%)
4
(%)
24
base
solvent
(%)
1
2
3
4
5
6
7
8
9
4b
4b
4c
4b
4b
4b
4c
4c
4c
tBuONa 65
THF/PhCH3 38
THF/PhCH3 34
14
23
38
16
tBuOK
tBuOK
K2CO3
LDA
65
65
65
65
THF
18
b
THF
THF/PhCH3
DMF
THF
THF
b
Cs2CO3 65
tr >90
NaH
Na0
Na0
65
50
68
10
18
94
89
1.7 2.5
4.5 2.5
rt f 50 THF
a All results in this screen are given in HPLC area % at 210
nm. b No reaction was observed under these reaction conditions,
although some hydrolysis of the starting material due to adventi-
tious water was observed. c The chloropropionic acid was added
to a solution of the preformed phenoxide 23.
further established by React IR experiments, in which
the expected R-lactone resonance at 1800 cm-1 was not
observed.
Unfortunately, the stereochemistry of the C-2 and C-10
centers could not be established directly as none of the
salts of the final product 2b or 2c produced X-ray quality
crystals. The stereochemistry was thus established at the
final product in an indirect way as discussed later.
Alternative Coupling Agents. Despite the success
in the stereochemical outcome of the coupling reaction,
there was a significant yield loss due to hydrolysis of the
ester function during the course of the reaction. This
significant side reaction was not due to adventitious
water in the medium but, once more, due to the presence
of NaOH on the surface of the NaH.
To avoid these yield losses, we investigated alternative
bases for this coupling. Once more the choices of reagents
were limited by the presence of the epimerizable chiral
centers; however, our previous experience in the ethyla-
tion reaction guided our experimental design (Table 3).
As seen in the table, Na0 offered significant advantages
over NaH and was chosen as an interim solution for our
development work (Scheme 4).
Despite the fact that the stereochemistry was un-
known, since a single product was obtained from the
reaction we felt that we could successfully incorporate
this chemistry into our synthetic strategy. If the reaction
proceeded with retention of configuration, the R-chloro-
propionic acid would be used, whereas inversion would
require the S-isomer. Both isomers of chloropropionic acid
are commercially available.22
Consequently, according to our original procedure,
phenol 4b was dissolved in THF and treated with 2.1
equiv of NaH, and the resulting phenoxide was in turn
treated with 1.2 equiv of S-chloropropionic acid (22a). A
thick suspension (presumably 22b) was formed, which
was heated to 75 °C to form the product 2b (R ) Et) in
ca. 70% yield (eq 7).
Phenol 4c in THF was added to Na0 cubes in THF and
stirred for 1 h until 23 was produced as judged by the
cessation of the H2 evolution. The solution was heated
to 50 °C, and 22a was added slowly to form a thin, easily
stirred suspension. After overnight reaction (98-99%
conversion) the mixture was quenched with acetic acid
to ensure destruction of Na0, followed by normal aqueous
workup.
The resulting product had sufficient chemical and
diastereomeric purity (g97%) to be carried forward crude;
however, crystallization as the R-naphthylethylamine
salt considerably enhanced its diastereomeric purity. For
instance, material produced in some of our earlier batches
with significant amounts of the wrong C-2 isomer (7.4%,
from the NaH ethylation reaction) could be successfully
upgraded to pharmaceutically acceptable levels via the
R-naphthylethylamine salt 2d as shown in Scheme 5.
Next, we examined the parameters that influenced the
robustness of this key transformation. Not surprisingly,
the water content of the reaction solvent and starting
material had to be controlled in order to avoid the
formation of diacid 24. Although this impurity did not
impact the purity of the final product, its formation
diminished the overall yield.
To our delight, once more a single diastereomer (2b)
was formed. Thus, it has been established that the
reaction proceeded with complete inversion of stereo-
chemistry at the chloropropionic acid stereogenic center.
This clearly established that formation of R-lactone 21
does not occur to any significant degree. This fact was
(21) Formation of transient R-lactones has been accomplished via
photochemical means: Chapman, O. L.; Wojtkowski, P. W.; Adam, W.;
Rodriquez, O.; Rucktaeschel, R. J. Am. Chem. Soc. 1972, 94, 1365. They
have also been postulated in solvolysis reactions of R-bromo pro-
pionates: Grunwald, E.; Winstein, S. J. Am. Chem. Soc. 1948, 70, 841.
(22) Both R-(+)- and S-(-)-2-chloropropionic acid are available from
Sigma-Aldrich Co. On a production scale the S-derivative is available
at ca. $100/kg and 96% ee.
J. Org. Chem, Vol. 70, No. 12, 2005 4701

Aikins et al.
SCHEME 4. Diastereoselective Coupling of 4c with S-2-chloropropionic Acid
SCHEME 5. Diastereomeric Excess Upgrade of 2c
It is noteworthy that the temperature of the reaction
and the temperature of addition of 22a have no impact
to the chemical or diastereomeric purity, although the
rate and operational convenience are indeed affected. For
example, addition of the chloropropionic acid to the
solution of 23 at ambient temperature resulted in a thick
suspension, which covered the surface of the Na0, causing
some variable reaction rates depending on the rate of
agitation and vessel configuration. These problems were
avoided by slow addition of the acid at the reaction
temperature (50 °C), affording reproducible results (rate
and yield) in the laboratory as well as kilogram scale (3
kg scale).
SCHEME 6. Impurities in the Sodium-Induced
Coupling of 4c with S-2-Chloropropionic Acid
Remarkably the concentration of the reaction had a
significant effect on product purity. The optimal reaction
conditions, described in Scheme 4, required fairly dilute
conditions at ca. 30 mL of THF per gram of 4c. Under
these conditions the des-ethoxy impurity 25 (Scheme 6)
was formed at a manageable level of 2-2.5%, presumably
via a single election transfer (SET) mechanism. However,
when the reaction was attempted in a concentration of
20 mL per gram, not only 25 but also the over-reduced
alcohol 26 was produced in substantial amounts (g10%).
The dimeric products 27 and 28 were also detected at
ca. 0.3% and 0.4%, although they were not observed
under our standard protocol (30 mL per gram).
higher amounts of these impurities. Preliminary results
in our laboratories showed that the use of (CH3)3SiONa
in THF was capable of effecting the transformation with
complete inversion of configuration. These results will
be reported elsewhere upon further development.
Despite the fact that the level of these impurities was
not significant at this stage, their existence raised
concerns about the manufacturability of this reaction. In
particular the quality of Na0 (and its K0 content) could
enhance the degree of the SET component, leading to
The Final Sequence. Formation of the amide bond
proceeded in good yield, with no epimerization at C-10.
Although the reaction proceeded smoothly, care had to
be taken to add the EDC slowly or in portions in a
4702 J. Org. Chem., Vol. 70, No. 12, 2005

SCHEME 7. Final Reaction Sequence
SCHEME 8. Impurities Identified in the Final
Product
cally established. Thus by exclusion our compound is the
required 2-S,10-R, biologically active compound.23
Alternative Coupling Results. Finally, some pre-
liminary investigation was conducted to evaluate the
convergent approach outlined in Scheme 1. We were
particularly interested to examine whether in this reac-
tion the intermediacy of an R-lactam24 could be observed,
in contrast to the absence of the R-lactone with chloro-
propionic acid.
Formation of the amide took place smoothly under our
standard coupling conditions to produce the desired
S-chloropropionic acetamide in high yield and with
complete preservation of optical purity. Unfortunately
coupling attempts under our established conditions did
not give the superior results realized with the propionic
acid itself. Both Na0 and NaH afforded a mixture of the
coupling product 29 and its hydrolysis product 1 in
modest yields. More disappointing was the fact that
significant erosion of the enantiomeric excess was ob-
served. Alternative bases and solvents such as NaOH in
DMSO (Scheme 9) improved the overall yield of the
coupling reaction, unfortunately at the expense of selec-
tivity. React IR experiments failed to reveal the inter-
mediacy of the R-lactam, and so it is assumed that the
reaction did not proceed exclusively via SN2 displacement
or that the chloropropionamide racemizes under the
reaction conditions. This approach was not considered
further for both technical and strategic reasons discussed
before, as with this approach the purification opportunity
of the intermediate 2c is no longer possible and thus a
valuable control point has been eliminated.
mixture of 2c, 3, and HOBT in order to avoid the heavy
precipitate that resulted when the addition was done all
at once.
An extractive workup afforded the product (29), which
was taken directly to the next step. The isopropyl ester
moiety was readily hydrolyzed with 1 NaOH in EtOH to
give 1 in quantitative yield. No epimerization was
observed at C-2 or C-10, and indeed performing the
hydrolysis at 70 °C also gave no epimerization of our
stereogenic centers.
It is noteworthy that impurities resulting from 24 and
26 can readily be removed during workup via base and
acid extractions (Scheme 8)
Finally, purification of 1 to afford purity consistent
with API requirements can be effected by formation of
the dicyclohexylammonium salt (30).
Structure Determination. Unfortunately, we were
unable to establish directly the absolute stereochemistry
of the final product, so an indirect method was used. First
the center at C-2 was firmly established by X-ray
crystallography as discussed above. Next both possible
diastereomers (10-R, desired, and 10-S) of 1 were pre-
pared by separate reaction of 4c with S- and R-chloro-
propionic acid, respectively (Scheme 5), and the resulting
2c was carried forward to the final product (Scheme 8).
The undesired 10-S,2-S diastereomer did give X-ray
quality crystals, and its stereochemistry was unequivo-
In conclusion, we have devised a number of synthetic
strategies for the synthesis of 1 with control of the
absolute and relative stereochemistry without the need
for resolution or chiral chromatography. Purification of
all intermediates was accomplished without the need for
column chromatography. Both stepwise and convergent
(23) The X-ray structure determination and chiral separation of all
four possible diastereomers of 1 are shown in Supporting Information.
(24) A thorough discussion with extensive references on the forma-
tion and reactivity of R-lactams is described: Tantilla, D. J.; Houk, N.
K.; Hoffman, R. V.; Tao, J. J. Org. Chem. 1999, 64, 3830.
J. Org. Chem, Vol. 70, No. 12, 2005 4703

Aikins et al.
(t, J ) 7.33 Hz, 2H), 7.42-7.46 (m, 2H), 12.42 (bs, 1H). 13C
NMR (100 MHz, DMSO-d6) δ ppm: 69.1, 71.2, 114.3, 127.5,
128.4, 130.2, 130.3, 137.3, 156.8, 175.1. HRMS: calcd 273.1127,
found 273.1118
SCHEME 9. Attempts at a Convergent Synthesis of 1
Preparation of 2-Hydroxy-3-(4-benzyloxyphenyl)-pro-
pionic Acid Isopropyl Ester (10c). The solid 10a (20 g, 73.5
mmol) was suspended in 400 mL of 2-propanol, and 2.2 mL of
H2SO4 (41.3 mmol) was added. After 1 h of stirring at ambient
temperature, the mixture was heated at 45 °C overnight. The
HPLC showed <4% (HPLC area) of remaining starting mate-
rial. The product can be isolated in two different ways. The
first is an extractive workup with addition of toluene (300 mL)
and NaHCO3 (200 mL, saturated solution) to remove the traces
of unreacted starting material. The layers are separated, and
the organic layer is dried by azeotropic distillation followed
by a clarifying filtration. The dried solution can be used as-is
in the next step after a switch of solvent to DMF. Quantitative
assay by HPLC indicated 21.5 g of product 10c (93%).
Alternatively, the material can be isolated as a low melting
solid by precipitation from the reaction mixture by the addition
of water at pH 8.5 (0.35 g of K2CO3, 2.1 g of NaHCO3), again
to remove traces of starting material. Ratio of 2-propanol to
aqueous ) 1:3. Average yield at the kilo scale: 95% (5 batches).
1H NMR (400 MHz, CDCl3) δ ppm: 1.15-1.31 (m, 6H), 2.54
(s, 1H), 2.90 (dd, J ) 14.02, 6.44 Hz, 1H), 2.98-3.13 (m, 1H)
4.35, t, J ) 5.56 Hz, 1H), 4.99-5.10 (m, 3H), 6.90 (d, J ) 8.34
Hz, 2H), 7.15 (d, J ) 8.34 Hz, 2H), 7.28-7.35 (m, 1H), 7.37 (t,
J ) 7.33 Hz, 2H), 7.40-7.44 (m, 2H). 13C NMR (100 MHz,
CDCl3) δ ppm: 21.8, 39.6, 69.6, 70.0, 71.3, 114.7, 127.41, 127.9,
128.5, 130.6, 137.1, 157.8, 173.7. HRMS: calcd 315.1596, found
315.1603.
Preparation of 2-Ethoxy-3-(4-benzyloxyphenyl)-pro-
pionic Acid Isopropyl Ester (11c). The isolated 10c (105
g, 0.334 mol) was placed in an inerted 6 L reactor and dissolved
in anhydrous DMF (1 L, <200 ppm H2O). The resulting
solution was cooled to -30 °C and treated with ethyl iodide
(261 g, 1.674 mol). This homogeneous mixture was subse-
quently treated with a solution of sodium tert-amylate in
toluene (420 mL, 0.42 mol, 1 M solution) at such a rate that
the temperature was maintained between -27 to -30 °C. After
addition of approximately three-fourths of the total charge of
base, a precipitate formed that caused an increase in the
internal temperature, requiring a decrease in the addition rate
to maintain the prescribed temperature. After 1.5 h of stirring
at -30 °C, an in-process assay indicated 6% of 10c still
remained. An additional portion of 1 M sodium tert-amylate
(30 mL, 0.03 mol) was added, and the reaction mixture was
stirred for an additional l h. Acetic acid (32.5 g, 1.54 mol) was
added dropwise, and the mixture was warmed to ca. 10 °C.
Toluene (2 L) was added followed by water (1 L), added at 10
°C (exothermic dissolution of the solids caused a temperature
increase up to 22 °C). The organic layer was separated, and
the aqueous layer was diluted with an additional 0.5 L of water
and re-extracted by 0.5 L toluene. The combined organic layers
are washed twice with 0.5 L of water followed by 15% NaCl
solution (0.5 L). The toluene was evaporated under reduced
pressure to afford crude 11c (ca 130 g). The crude material
was then purified on silica gel (330 g) using cyclohexane/ethyl
acetate (90/10 v/v) as the eluent. Purified 11c was isolated as
a pale yellow oil in 90% yield (122.8 g, 0.3 mol). Enantiomeric
purity data for two different runs: run 1 ) 1.2% of wrong
isomer; run 2 ) 2.2% of the wrong isomer. 1H NMR (400 MHz,
CDCl3) δ ppm: 1.11-1.2 (m, 6H), 1.23 (d, J ) 6.32 Hz, 3H),
2.94 (d, J ) 6.57 Hz, 2H), 3.30-3.42 (m, 1H), 3.53-3.65 (m,
1H), 3.94 (t J ) 6.69 Hz, 1H), 4.96-5.09 (m, 2H), 6.89 (d, J )
8.59 Hz, 2H), 7.16 (d, J ) 8.59, 2H), 7.31 (t, J ) 7.07 Hz, 1H),
7.37 (t, J ) 7.33 Hz, 2H), 7.40-7.45 (m, 2H). 13C NMR (100
MHz, CDCl3) δ ppm: 15.1, 21.7, 21.8, 38.4, 66.0, 68.2, 70.0,
80.4, 114.6, 127.41, 127.9, 128.5, 129.6, 130.4, 137.2, 157.6,
172.0. HRMS: calcd 343.1909, found 343.1942
approaches were evaluated, the former proving superior
in terms of overall selectivity and purity control.
The selective synthesis of 4 without the need for the
sensitive alkylation step is under investigation in our
laboratories and will be reported shortly.
Experimental Section
Preparation of 2-Hydroxy-3-(4-hydroxyphenyl)-pro-
pionic Acid (6). Solid sodium salt 7 (204 g, 0.94 mol) and
sodium iodide (375 g, 2.5 mol) were dissolved in 1 L of
concentrated hydrochloric acid, at ambient temperature, in an
inerted reaction vessel equipped with a caustic scrubber. The
solution was heated gently to 110 °C for 24 h. Upon completion
of the reaction, as judged by HPLC analysis, the mixture was
cooled in an ice-water bath for 2 h. The slurry was filtered,
and the filter cake was washed with 300 mL of cold water.
The white solid was collected and dried in a vacuum oven at
60 °C with nitrogen purge to provide 178.7 g of crude solid.
The crude solid, which contains some NaI, can be used as-is
in the next step. Further purification could be effected by
washing the filter cake with further amounts of cold water,
with slight yield losses. Average yield at 5 kg scale (2
batches): 91%. 1H NMR (400 MHz, DMSO-d6) δ ppm: 2.66
(dd, J ) 13.77, 7.96 Hz, 1H), 2.83 (dd, J ) 13.77, 4.42 Hz,
1H), 4.05 (dd, J ) 7.71, 4.67 Hz, 1H), 5.20 (s, 1H), 6.64 (d, J
) 8.08 Hz, 2H), 7.00 (d, J ) 8.34 Hz, 2H), 9.12, s, 1H), 12.35
(bs, 1H). 13C NMR (100 MHz, DMSO-d6) δ ppm: 71.4, 114.8,
128.1, 130.2, 155.7, 175.2.
Preparation of 3-(4-Benzyloxyphenyl)-2-hydroxy-pro-
pionic Acid (10a). A solution of 6 (43.32 g, 0.238 mol) in
absolute ethanol (950 mL) was treated with K2CO3 (65.73 g,
0.476 mol) and benzyl chloride (55 mL, 0.476 mol), and the
suspension was refluxed for 12-24 h. Upon completion of the
reaction (note: the phenol benzyl ether, benzyl ester compound
was formed at this stage), the ethanol was removed by vacuum
distillation (65% of the total volume). Water (1 L) and sodium
hydroxide (44 mL) were added (pH ) 13-14) in order to
hydrolyze the benzyl ester formed during the reaction. The
hydrolysis was complete after 2 h at ambient temperature. The
remainder of the ethanol was removed followed by the addition
of water (500 mL) and MTBE (1000 mL) before acidification
by dropwise addition of HCl 37% (88 mL). Carbon dioxide
evolution was accompanied by the precipitation of the product
as an off-yellow solid. The solid was filtered, rinsed with water
(500 mL), and dried overnight under reduced pressure and
nitrogen purge at 50 °C to give 10a as an off-white solid (60.81
g, 94%). At the kilo scale, 5 batches, lower yields (ca 82%) were
obtained as a result of higher losses in the mother liquors. 1H
NMR (400 MHz, DMSO-d6) δ ppm: 2.73 (dd, J ) 13.77, 8.21
Hz, 1H), 2.90 (dd, J ) 13.77, 4.42 Hz, 1H), 4.10 (dd, J ) 7.83,
4.55 Hz, 1H) 5.06 (s, 2H), 5.30 (s, 1H) 6.91 (d, J ) 8.34 Hz,
2H), 7.14 (d, J ) 8.34 Hz, 2H) 7.33 (d, J ) 7.07 Hz, 1H), 7.38
Preparation of 2-Ethoxy-3-(4-hydroxyphenyl)-pro-
pionic Acid Isopropyl Ester (4c). A dry and inerted 2 L
4704 J. Org. Chem., Vol. 70, No. 12, 2005

Parr hydrogenator was charged with Pd/C 10% (anhydrous, 8
g) and 11c (80 g, 0.232 mol) dissolved in 800 mL of absolute
ethanol. The vessel was purged five times with 1.5 bar nitrogen
and set under 40 psi H2 pressure keeping the temperature
under 27 °C. When no more hydrogen consumption was
observed and HPLC analysis confirmed consumption of start-
ing material, the mixture was filtered through 20 g of Celite.
The filter cake was rinsed twice with 50 mL of ethanol, and
the combined solutions were concentrated under reduced
pressure. Toluene (200 mL) was added to the residue, and
distillation was continued at ca. 80 mbar at about 40 °C in
order to remove any residual ethanol. The volatiles were
concentrated at 50 °C at 10 mbar pressure. The product 4c
was isolated as a yellow oil (59.7 g, 0.236 mol, 100% yield)
and stored under nitrogen, protected from light. 1H NMR (400
MHz, CDCl3) δ ppm: 1.08,-1.21 (m, 6H), 1.22 (d, J ) 6,0.32
Hz, 3H), 2.92 (d, J ) 6.57 Hz, 2H), 3.28-3.43 (m, 1H), 3.50-
3.65 (m, 1H), 3.95 (t, J ) 6.57 Hz, 1H), 4.93-5.11 (m, 1H),
6.71 (d, J ) 8.34 Hz, 2H), 7.07 (d, J ) 8.08 Hz, 2H). 13C NMR
(100 MHz, CDCl3) δ ppm: 15.0, 21.6, 21.8, 38.4, 66.1, 68.6,
80.5, 115.2, 128.7, 130.5, 154.6, 172.4. HRMS: calcd 253.1440,
found 253.1444
Preparation of S-2-Ethoxy-3-[4-(R-1-carboxy-ethoxy)-
phenyl]-propionic Acid Isopropyl Ester (2c). A 500 mL
dry and inerted vessel was charged with freshly cut pieces of
Na0 (24.1 g, 1.04 mol, 2.65 equiv) in THF (1 L). The starting
material 4c (99.84 g, 395.7 mmol, 1.0 equiv) was azeotropically
dried from toluene and added as a solution in anhydrous THF
(2 L). Controlled gas evolution was observed on the surface of
the metal. The brownish red mixture was stirred at ambient
temperature for 1 h. The surface of the sodium turned black.
The resulting phenoxide was heated to 50 °C and was treated
with chloropropionic acid (36 mL, 414.7 mmol, 1.05 equiv
added neat) added dropwise over 10 min. The sodium surface
cleaned again and precipitation of an easily stirred solid was
observed. It is worth noting that addition of the chloropropionic
acid at ambient temperature resulted in the formation of a
gel that may present stirring problems at larger scales.
After 6 h at ca. 50 °C, HPLC analysis indicated that the
reaction was 95% complete, while <2% of the diacid 24 could
be observed. The mixture was cooled to ambient temperature
and transferred into 2.5 L of a 5% aqueous solution of NaH2-
PO4 (pH ) 4.2). The pH after the quench is 5.8 (which avoids
ester hydrolysis). Toluene was added (1.5 L), and the pH of
the aqueous layer was adjusted to 2.4 with 6 N HCl (ca. 110
mL). The layers are separated, and the organic layer was
extracted twice with NaHCO3 (2.2 L, saturated aqueous
solution) to remove unreacted starting material. The aqueous
layers (containing the desired product) were combined, fresh
toluene (4.4 L) was added, and the mixture was acidified with
HCl (6 N) to pH ) 2.4. The desired compound is quantified in
the different layers by HPLC against a standard. The yield in
the organic layer was thus estimated at 109 g, 86%, with the
following impurity profile: 1.2% (area% at 220 nm) of starting
material, 1.0% of the diacid impurity 24, and 92.9% of the
desired compound 2c. This solution was used directly in the
next reaction. Alternatively, the toluene was removed by
vacuum distillation and replaced with isopropyl acetate (1.1
L) followed by R-naphthylethylamine (58 g, 340 mmol), and
the mixture was stirred overnight. Filtration and washing of
the salt with isopropyl acetate afforded 2d. 1H NMR (400 MHz,
CDCl3) δ ppm: 1.02-1.15 (m, 9H), 1.19 (d, J ) 6.32 Hz, 3H),
1.46 (d, J ) 6.06 Hz, 3H), 2.63-2.83 (m, 2H), 3.15-3.29 (m,
1H), 3.45-3.51 (m, 1H), 3.68-3.77 (m, 1H), 4.14-4.23 (m, 1H),
4.94-4.99 (m, 2H), 6.47 (d, J ) 8.08 Hz, 2H), 6.82 (d, J ) 8.08
Hz, 2H), 7.34 (t, J ) 7.58 Hz, 1H), 7.42-7.52 (m, 2H), 7.69-
7.81 (m, 3H), 7.85 (m, 1H), 7.90 (br s, 3H). 13C NMR (100 MHz,
CDCl3) δ ppm: 15.0, 18.5, 21.5, 21.8, 38.2, 46.6, 65.9, 68.2,
74.5, 80.3, 115.0, 122.2, 122.6, 125.80, 126.6, 128.5, 129.0,
129.2, 130.1, 135.9, 156.8, 172.0, 178.7. HRMS: calcd 325.1651,
found 325.1644.
Preparation of S-2-Ethoxy-3-(4-{R-1-[2-(4-ethyl-phen-
yl)-ethylcarbamoyl]-ethoxy}-phenyl)-propionic Acid Iso-
propyl Ester (29). Anhydrous THF (2.75 L), ethylpheneth-
ylamine (118.6 g), HOBT (127 g), and EDCI (90.2 g) were
added to the toluene solution of 2c (above) between 20 and 30
°C. After 15 min, a second portion of EDCI (90.2 g) was added,
and the reaction mixture was stirred at room temperature for
2 h. H2O (1.15 L) and HCl 37% (8.2 g) were added to a pH of
ca. 2. The layers were separated, and the organic layer was
washed with an aqueous solution of NaCl (10% solution, 250
mL), an aqueous solution of Na2CO3 (20% solution, 780 mL),
and finally with a second aqueous solution of NaCl (10%
solution, 250 mL). The organic layer was partially distilled (2.2
L) under reduced pressure. Ethanol (820 mL) was added, and
a second distillation was performed (430 g of distillate) under
reduced pressure Finally the ethanol volume was adjusted to
25 mL per gram of 29 (determined by quantitative HPLC
analysis), and the solution was used directly in the next step.
Preparation of S-2-Ethoxy-3-(4-{R-1-[2-(4-ethyl-phen-
yl)-ethylcarbamoyl]-ethoxy}-phenyl)- propionic Acid (1).
The solution of 29 above was treated with aqueous NaOH
solution (84 g of NaOH/2.1 L of H2O), and the reaction mixture
was strirred at 30 °C for 5 h. The weight of the reaction
mixture was 8.5 kg (3.29% w/w ) 279.5 g of 1). The ethanol
was partially distilled (direct aqueous workup did not succeed
because of the large volume of ethanol present), and H2O (3.80
L) and MTBE (2.70 L) were added to the residue. The aqueous
phase was acidified with HCl 37% (ca. 200 mL) to pH ) 2 and
extracted with MTBE (2 × 2.8 L). The MTBE was partially
evaporated under reduced pressure to give 5336 g of a solution
of 1 (4.636% w/w ) 247.3 g of 1). The MTBE was removed in
vacuo and replaced with isopropyl acetate (2.5 L, ca. 10 mL/
g). The resulting homogeneous solution was treated with neat
dicycloxexylamine (109 g, 1.02 equiv) at ambient temperature
overnight. The resulting slurry was filtered, and the filter cake
was washed with isopropyl acetate (300 mL) and dried in vacuo
1
at 40 °C to afford 1 as the DCHA salt (326 g, 92%). H NMR
(400 MHz, CDCl3) δ ppm: 1.09-1.35 (m, 12 H), 1.44-1.61 (m,
7H), 1.69 (d, J ) 10.86 Hz, 2H), 1.83 (d, J ) 12.13 Hz, 4H),
2.07 (d, J ) 11.12 Hz, 4H), 2.60-2.76 (m, 3H), 2.77-2.86 (m,
1H), 2.91 (dd, J ) 13.26, 9.73 Hz, 1H), 2.96-3.12 (m, 3H),
3.26-3.41 (m, 1H), 3.47-3.54 (m, 1H), 3.57-3.63 (m, 1H),
3.66-3.72 (m, 1H), 3.86 (d, J ) 7.33 Hz, 1H), 4.57-4.7 (m,
1H), 6.54 (t, J ) 5.18 Hz, 1H), 6.78 (d, J ) 7.33 Hz, 2H), 7.02
(d, J ) 6.82 Hz, 2H), 7.12 (d, J ) 7.07 Hz, 2H), 7.25-7.32 (m,
2H), 8.50 (s, 2H). 13C NMR (100 MHz, CDCl3) δ ppm: 15.2,
15.5, 18.9, 24.8, 25.3, 28.4, 29.3, 29.4, 35.3, 39.0, 40.1, 52.6,
65.2, 75.1, 82.5, 114.9, 128.0, 128.7, 130.5, 133.4, 135.7, 142.4,
155.3, 172.2, 177.6. HRMS: calcd 414.2280, found 414.2254
Supporting Information Available: NMR and high
resolution mass spectral data for all new isolated compounds;
HPLC and chiral HPLC data for the key intermediates; X-ray
data in CIF format for the determination of the absolute
configuration of 4 and support of the absolute stereochemistry
of the final product. This material is available free of charge
via the Internet at http://pubs.acs.org.
JO050268E
J. Org. Chem, Vol. 70, No. 12, 2005 4705