Enzymatic kinetic resolution of piperidine atropisomers: Synthesis of a key intermediate of the farnesyl protein transferase inhibitor, SCH66336

Article abstract of DOI:10.1021/jo991513v

The resolution of secondary amines via enzyme-catalyzed acylation is a relatively rare process. The kinetic resolution of a series of intermediates of SCH66336 (1), by either enzymatic acylation of the pendant piperidine (4, 5) or hydrolysis of the corresponding carbamate 3, was investigated. In the case of 4, the molecule exists as a pair of enantiomers due to atropisomerism about the exocyclic double bond. The enzymatic acylation of (±)-4 was optimized in terms of acylating agent, solvent, and moisture content. The use of lipase, Toyobo LIP-300, and trifluoroethyl isobutyrate as acylating agent resulted in isobutyrylation of the (+)-enantiomer, which is easily separated from the unwanted (-)-4. Hydrolysis of the isobutyramide 6c yielded the desired (+)-4 in high enantiomeric excess. (-)-4 may be recovered from the resolution step, racemized, and resubjected to enzymatic acylation to increase material throughput.

Full text of DOI:10.1021/jo991513v

VOLUME 65, NUMBER 18
SEPTEMBER 8, 2000
© Copyright 2000 by the American Chemical Society
Articles
En zym a tic Kin etic Resolu tion of P ip er id in e Atr op isom er s:
Syn th esis of a Key In ter m ed ia te of th e F a r n esyl P r otein
Tr a n sfer a se In h ibitor , SCH66336
Brian Morgan,*,† Aleksey Zaks,† David R. Dodds,†,‡ J inchu Liu,† Rama J ain,† Sreeni Megati,†
F. George Njoroge,§ and Viyyoor M. Girijavallabhan§
Biotransformations Group, Schering-Plough Research Institute, U-13-3000, 1011 Morris Avenue,
Union, New J ersey 07083, and Department of Chemical Research and Tumor Biology,
2015 Galloping Hill Road, Kenilworth, New J ersey 07033
E-mail: williambrian.morgan@spcorp.com.
Received September 27, 1999
The resolution of secondary amines via enzyme-catalyzed acylation is a relatively rare process.
The kinetic resolution of a series of intermediates of SCH66336 (1), by either enzymatic acylation
of the pendant piperidine (4, 5) or hydrolysis of the corresponding carbamate 3, was investigated.
In the case of 4, the molecule exists as a pair of enantiomers due to atropisomerism about the
exocyclic double bond. The enzymatic acylation of (()-4 was optimized in terms of acylating agent,
solvent, and moisture content. The use of lipase, Toyobo LIP-300, and trifluoroethyl isobutyrate as
acylating agent resulted in isobutyrylation of the (+)-enantiomer, which is easily separated from
the unwanted (-)-4. Hydrolysis of the isobutyramide 6c yielded the desired (+)-4 in high
enantiomeric excess. (-)-4 may be recovered from the resolution step, racemized, and resubjected
to enzymatic acylation to increase material throughput.
In tr od u ction
To exhibit biological activity, the Ras protein must
undergo binding to the cell membrane, which is depend-
ent on the following sequence of posttranslational
events: (i) farnesylation of the cysteine residue near the
carboxyl terminus, (ii) proteolysis of the three amino
acids at the carboxyl terminus, and (iii) esterification of
the new farnesylated cysteine carboxyl terminus.2 Since
farnesyl protein transferase (FPT) catalyzes the transfer
of a farnesyl group from farnesyl diphosphate to the
cysteine of Ras, it is an attractive target for disrupting
the association of Ras to the cell membrane and thus
Ras proteins play a major role in signal transduction
from the cell surface to the nucleus. In normal signaling,
the signal is terminated by the hydrolysis of bound GTP
to GDP by the intrinsic GTPase activity of the normal
Ras protein. In mutated Ras protein, the GPTase activity
is impaired and the signal remains active, leading to
unregulated cell proliferation. Point mutations in the ras
genes that encode for the Ras proteins have been
observed in ∼30% of all human cancers, including 50%
of colon cancers and 95% of pancreatic cancers.1
* To whom correspondence should be addressed. Telephone: (908)-
820-3545. Fax: (908)-820-6096.
(1) (a) Barbacid, M. Annu. Rev. Biochem. 1987, 56, 779-827. (b) Bos,
J . L. Cancer Res. 1989, 49, 4682-4689. (c) Lowy, D. R.; Willumsen, B.
M. Annu. Rev. Biochem. 1993, 62, 851-891.
(2) Casey, P. J .; Solski, P. A.; Der, C. J .; Buss, J . E. Proc. Natl. Acad.
Sci. U.S.A. 1989, 86, 8323-8327.
† Biotransformations Group.
‡ Present address: Bristol-Myers Squibb, Syracuse, NY 13221.
§ Department of Chemical Research and Tumor Biology.
10.1021/jo991513v CCC: $19.00 © 2000 American Chemical Society
Published on Web 08/10/2000

5452 J . Org. Chem., Vol. 65, No. 18, 2000
Morgan et al.
from one face of the tricyclic system to the other.8
Moreover, in the case of 4, thermal equilibration of the
separated atropisomers had been demonstrated. There-
fore, resolution of either 3 or 4 offers the potential of
recovering, racemizing, and recycling the unwanted
enantiomer, thereby increasing the yield of a valuable
intermediate.
Ch a r t 1
The selective enzymatic acylation and deacylation of
alcohols has become a powerful technique for the prepa-
ration of optically enriched compounds and is practiced
both in laboratory- and large-scale production.9,10 In
contrast, the corresponding resolution of amines is much
less developed. In particular, the selective enzymatic
acylation/deacylation of cyclic secondary amines has only
sporadically been reported,11 even though such com-
pounds (e.g., pyrrolidines, piperidines, piperazines) are
common building blocks in pharmaceuticals.
terminating the signal for cell proliferation.3 However,
since the cell also contains a number of geranylgeranyl-
ated proteins, farnesyl protein transfer inhibitors (FPTI)
must be selective for FPT over geranylgeranyl protein
tranferase-1 (GGPT-1).
SCH66336 (1) (Chart 1) has emerged as a novel,
selective, nonpeptide, nonsulfhydryl farnesyl protein
transferase inhibitor, showing activity in the nanomolar
range.4
The deacylation of carbamate (()-3 by microbial hy-
drolysis was initially evaluated (Scheme 2). Of the 373
cultures that were examined, none showed significant
formation of deacylated product 4, nor of residual non-
racemic starting material 3.
The synthesis of 1, which is outlined in Scheme 1, be-
gins with Loratadine (2),5 which is regiospecifically bro-
minated at positions 3 and 10 of the diarylcycloheptane
portion of the molecule.4 Acidic hydrolysis of the carbam-
ate 3 followed by reduction of the exocyclic double bond,6
produces racemic 5, which is resolved by diastereomeric
salt formation with N-Ac-L-phenylalanine followed by
recrystallization from EtOH. After liberation of the free
base, the desired (R)-(+)-5 is obtained in 30-40% molar
yield and 97-98% enantiomer excess. Subsequent cou-
pling with the appropriate 4-substituted piperidine and
several other steps complete the synthesis of 1.
Similarly poor results were obtained for the hydrolysis
of the amide derivatives (()-6a ,b. No product was
detected when the acetamide (()-6a or the phenylaceta-
mide (()-6b were incubated with commercially available
hydrolytic enzymes. In light of these initial negative
results, efforts were concentrated on the kinetic resolu-
tion of 4 and 5 via enzymatic acylation.
A. E n zym a t ic R esolu t ion of (()-4. (a) Enzymatic
Acetylation. The enzymatic resolution of (()-4 was
screened against 233 commercially available enzyme
preparations under acylating conditions, using trifluo-
roethyl acetate (TFEOAc) as acylating agent (Scheme 3).
Five enzymes produced enantiomerically enriched aceta-
mide (6a ) under these conditions, with the best results
being observed with ChiroCLEC PC (Pseudomonas ce-
Resu lts a n d Discu ssion
As a potential alternative to the chemical resolution
of (()-5, the enzymatic resolution of key chiral intermedi-
ates of SCH66336 was explored. The approaches consid-
ered were (a) microbial hydrolysis of the carbamate 3,
(b) enzymatic hydrolysis of the acetamide or phenylac-
etamide of 4, (c) enzymatic acylation of (()-5, and (d)
enzymatic acylation of (()-4. Of these four options, the
most attractive are those that result in the resolution of
derivatives of 3 or 4. Although they do not contain a
chiral center, both 3 and 4 exist as pairs of conforma-
tional enantiomers due to the atropisomerism7 resulting
from the reduced conformational mobility of the diaryl-
[a,d]cycloheptane ring. The bromine substituent at the
10-position prevents flipping of the piperidinylidene ring
(8) Conformational diastereomers due to restricted piperidinylidene
ring flipping have been described in similar tricyclic systems: Piwinski,
J . J .; Wong, J . K.; Chan, T.-M.; Green, M. J .; Ganguly, A. K. J . Org.
Chem. 1990, 55, 3341-3350.
(9) (a) Faber, K., Biotransformations in Organic Chemistry, 3rd ed.;
Springer-Verlag: Berlin, 1997. (b) Drauz, K.; Waldmann, H., Eds.
Enzyme Catalysis in Organic Synthesis: A Comprehensive Handbook;
VCH Publishers: New York, 1995. (c) Wong C.-H.; Whitesides, G.
Enzymes in Synthetic Organic Chemistry; Pergamon: New York, 1994.
(d) Poppe, L.; Novak, L. Selective Biocatalysis; VCH Publishers: New
York, 1992. (e) Kazlauskas, R. J .; Bornscheuer, U. T. Biotechnology,
Vol. 8a: Biotransformations I, 2nd ed.; Kelly, D. R., Ed.; Wiley-VCH:
Weinheim, Germany, 1998; pp 37-191. (f) Preparative Biotransfor-
mations: Whole Cell and Isolated Enzymes in Organic Synthesis;
Roberts S. M.; Wiggins, K.; Casy, G., Phythian, S., Eds.; Wiley:
Chichester, U.K., 1992. (g). Biotransformations CD-ROM; Chapman
and Hall, Electronic Publishing Division: London, U.K. (h) BioCa-
talysis; Synopsis Scientific Systems Ltd: Leeds, U.K.
(3) Levitzki, A. Eur. J . Biochem. 1994, 226, 1-13.
(10) (a) Zaks, A.; Dodds, D. R. Drug Discovery Today 1997, 2, 513-
531. (b) Patel, R. N. Adv. Appl. Microbiol. 1997, 43, 91-140. (c)
Sheldon, R. A. Enzymatic Reactions in Organic Media; Koskinen, A.
M. P., Klibanov, A. M., Eds.; Blackie Academic and Professional:
Glasgow, U.K. 1996; pp 266-307.
(11) (a) Fortier, G.; Mackenzie, S. L. Biotechnol. Lett. 1986, 8, 777-
782. (b) Nakajima, H.; Kitabatake, S.; Tsurutani, R.; Yamamoto, K.;
Tomioka, I.; Imahori, K. Int. J . Pept. Protein Res. 1986, 28, 179. (c)
Asensio, G.; Andreu, C.; Marco, J . A. Tetrahedron Lett. 1991, 32, 4197-
4198. (d) Gutman, A. L.; Meyer, E.; Yue, X.; Abell, C. Tetrahedron Lett.
1992, 33, 3943-3946. (e) Paradkar, V. M.; Dordick, J . S. J . Am. Chem.
Soc. 1994, 116, 5009-5010. (f) Chen, S. Y.; Kao, C. L.; Wang, K. T.
Biorg. Med. Chem. Lett. 1994, 4, 443-448. (g) Herradon, B.; Valverde,
S. Synlett 1995, 599-602. (h) Orsat, B.; Alper, P. B.; Moree, W.; Mak,
C.-P.; Wong, C.-H. J . Am. Chem. Soc. 1996, 118, 712-713. (i)
Takayama, S.; Lee, S. T.; Hung, S.-C.; Wong, C.-H. Chem. Commun.
1999, 127-128.
(4) Njoroge, F. G.; Taveras, A. G.; Kelly, J .; Remiszewski, S.;
Mallams, A. K.; Wolin, R.; Afonso, A.; Cooper, A. B.; Rane, D. F.; Liu,
Y.-T.; Wong, J .; Vibulbhan, B.; Pinto, P.; Deskus, J .; Alvarez, C. S.;
del Rosario, J .; Connolly, M.; Wang, J .; Desai, J .; Rossman, R. R.;
Bishop, W. R.; Patton, R.; Wang, L.; Kirschmeier, P.; Bryant, M. S.;
Nomeir, A. A.; Lin, C.-C.; Liu, M.; McPhail, A. T.; Doll, R. J .;
Girijavallabhan, V. M.; Ganguly, A. K. J . Med. Chem. 1998, 41, 4890-
4902.
(5) Villani, F. J .; Magatti, C. V.; Vashi, D. B.; Wong, J .; Popper, T.
L. Arzneim-Forsch./ Drug Res. 1986, 36, 1311-1314.
(6) Complete details of the reduction of (()-5 will be presented
elsewhere: Njoroge et al., manuscript in preparation.
(7) Since this behavior results from rotation about two bonds instead
of one, it is not atropisomerism as strictly defined. Nevertheless, the
more familiar term will be used: Oki, M. Topics in Stereochemistry;
J ohn Wiley and Sons: New York, 1983; Vol. 14, pp 1-81.

Enzymatic Kinetic Resolution of Atropisomers
J . Org. Chem., Vol. 65, No. 18, 2000 5453
Sch em e 1
Sch em e 2
carbonate was sluggish, and trifluoroethyl benzyloxycar-
bonate (TFEOBOC) showed moderate selectivity (E )
30).13
Moderate to good enantioselectivity was observed for
the trifluoroethyl esters of straight-chain acids (acetic,
butyric, hexanoic, and lauric) (E ) 10-80), and for
triacetin and tributyrin (E ) 20-80).
(b) Enzymatic Isobutyrylation. (i) Acylating Agent. To
enhance the enantioselectivity of the reaction, bulkier
acylating agents were examined. It was previously
observed that increasing the steric bulk of the acyl group
can result in a significant increase in enzyme selectivity.14
Repeating the Toyobo LIP-300 catalyzed acylation with
trifluoroethyl isobutyrate (TFEOiBu) in either TBME or
MeCN gave dramatically different results. The reaction
in MeCN was extremely slow compared to a similar
reaction using TFEOAc as acylating agent. In contrast,
the reaction with TFEOiBu in TBME showed excellent
enantioselectivity, with the isobutyramide 6c being
formed with 98% ee at 37% conversion (E > 100). The
reaction with trifluoroethyl benzoate also showed high
selectivity (E > 100) but was slower. Trifluoroethyl
dimethylbutyrate showed no sign of reaction under these
conditions, and trifluoroethyl adamantanecarboxylate
resulted in a low conversion to racemic product.
While more selective, the enzymatic isobutyrylation
reaction was also considerably slower, requiring large
loadings of catalyst to effect a useful conversion within
24 h.15 However, more reactive isobutyrylating reagents
showed poor results.16 A comparison of three acylating
agents under similar conditions confirmed that the
pacia) (enantiomer ratio,12 E ) 8) and Toyobo LIP-300
(Pseudomonas aeruginosa) (E ) 18). All enzymes showed
the same selectivity; the desired atropisomer (+)-4, which
eventually produces (R)-(+)-5 upon reduction, was acety-
lated. This resulted in a less convenient direct resolution,
since an extra step was required to remove the acetamide
moiety from 6a . Since the enantiomeric excess (ee) of the
product decreases with increasing conversion during a
direct resolution, a very selective process (high E value)
was required for such a resolution to be practical.
On the basis of catalyst cost, the acylation was further
investigated using Toyobo LIP-300, and those factors
known to affect enzyme selectivity and reactivity (e.g.,
solvent, temperature, acylating agent, moisture content,
enzyme source) were examined.
From an examination of 12 common solvents, the best
combination of reactivity, selectivity, and solubility was
observed in TBME. Since the reactivity and selectivity
of enzymatic acylations are in large part dictated by the
stereo and electronic properties of the acylating agent,
26 acyl transfer agents were examined in TBME.
(13) The moderate selectivity of TFEOBOC suggested the enzymatic
introduction of the Cbz group followed by catalytic hydrogenation to
remove the Cbz group and reduce the olefin to yield (R)-(+)-5 directly.
However transfer hydrogenation of (()-4 with cyclohexane/10% Pd/C
in refluxing EtOH resulted in complete removal of the 3-Br substituent,
with the olefin remaining intact.
(14) Increasing the size of the acylating agent from vinyl acetate to
isobutyric anhydride resulted in a dramatic increase in the selectivity
of a Novozyme 435 catalyzed desymmetrization of a prochiral 1,3-diol.
See: (a) Nielsen, C. M.; Sudhakar, A. U.S. Patent 5,756,830, 1998. (b)
Morgan, B.; Dodds, D. R.; Zaks, A.; Andrews, D. R.; Klesse, R. J . Org.
Chem. 1997, 62, 7736-7743.
(15) Toyobo LIP-300 is a lipoprotein lipase (P. aeruginosa) im-
mobilized on Hyflo Supercel. The commercial catalyst contains ∼1%
enzyme. See: Nakatani, T.; Hiratake, J .; Yoshikawa, K.; Nishioka, T.;
Oda, J . Biosci. Biotech. Biochem. 1992, 56, 1118-1123.
(16) Vinyl isobutyrate resulted in decomposition of starting material,
whereas acetone oxime isobutyrate and isobutyric anhydride showed
little or no selectivity at low temperature.
As previously observed for amine substrates, the use
of vinyl acetate resulted in extensive decomposition of
the starting material, whereas isopropenyl acetate, both
neat and in solution, resulted in complete acetylation.
Poor selectivity was also observed for ethyl trifluoroac-
etate, methyl methoxyacetate, neat methyl acetate and
the trifluoroethyl esters of chloroacetic, dichloroacetic,
and trifluoroacetic acids. The reaction with dibenzyl
(12) Chen, C.-S.; Fujimoto, Y.; Girdaukas, G.; Sih, C. J . J . Am. Chem.
Soc. 1982, 104, 7294-7299.

5454 J . Org. Chem., Vol. 65, No. 18, 2000
Morgan et al.
Sch em e 3
Ta ble 1. En zym a tic Acyla tion of (()-4a
immobilized on Hyflo Supercel, i.e., LIP-300/301, en-
hanced reactivity and selectivity were observed, and the
reactions proceeded to ∼40% conversion with E > 100
(runs 1, 5). Similarly good results were obtained with
LIP-321 (run 3), a recombinant form of the lipase, which
is also purified and immobilized. Whether these results
were due to differences in protein stability/conformation
under immobilized versus nonimmobilized conditions,
differences in enzyme preparation (e.g., “pH memory”),18
or whether they were due to differences in moisture
content of the various preparations (vide infra) was not
established.
run
R
time, h
ees, %
eep, %
conv., %
E
1
2
3
Me
nPr
iPr
2.75
5
26
81.1
71.1
96.3
89.3
95.6
97.0
47.6
42.7
49.8
44
95
>200
a
Conditions: (()-4, 2.3 g; LIP-300, 5 g; acylating agent, 3 equiv;
TBME, 46 mL; 200 rpm; room temperature.
reaction with TFEOiBu showed highest selectivity (Table
1). Reaction with TFEOAc was rapid but showed poor
selectivity, whereas the reaction with TFEOBu was
slower but more selective (E ) 95). Although much
slower, the reaction with TFEOiBu was clearly the most
selective: when run under these conditions, the product
was typically formed in g97% ee (E > 100) after 24 h.
Following removal of the enzyme, the mixture of (+)-6c
and (-)-4 could conveniently be separated by exploiting
the difference in pKa between the pyridine and the
piperidine. Extraction of the reaction mixture with 0.5
M H2SO4 selectively removed the more basic unreacted
(-)-4. Subsequent extraction with 6 M H2SO4 isolated
the product (+)-6c which was refluxed in the same acidic
solution to remove the isobutyramide group and yield (+)-
4. The recovered (-)-4 could be racemized and subjected
to further resolution (vide infra), and the acylation
reaction solution after both acidic extractions could be
washed, dried, and replenished with additional TFEOiBu
for further reaction.
To confirm the improvement in enzyme selectivity, 17
enzyme preparations were reexamined using TFEOiBu
as acylating agent. These included other preparations of
P. aeruginosa, selected Pseudomonas sp. lipases, and
those enzymes that had shown some selectivity in the
initial screen using TFEOAc as acylating agent. The data
in Table 2 confirm that the best results were achieved
with Toyobo LIP-300. Essentially no reaction was ob-
served for the enzymes from C. antarctica, C. rugosa and
Alcaligenes sp. (entries 8-13). Boehringer-Mannheim
Chirazyme L4, Lipase (Pseudomonas sp.), and Altus
ChiroCLECTM PC all showed low conversions and poor
selectivities (entries 5-7). High selectivity was observed
with Boehringer-Mannheim Chirazyme L6 (E ) 110), but
the reaction was slower. The most surprising result was
the variation in the performance, both reactivity and
selectivity, of the various Toyobo preparations (Table 2
and Figure 1).The crude lipase preparation, LPL-701,
showed good selectivity (E ∼ 100) but only about 20%
conversion after 24 h (run 4, Figure 1).17 In contrast, the
purified enzyme, LPL-311, showed poor selectivity and
poor reactivity (run 2). However, when LPL-311 is
(ii) Moisture Content. It was quickly established that
the degree of conversion of the enzymatic isobutyrylation
was sensitive to the moisture content of the reaction:
reactions carried out in the presence of 4 Å molecular
sieves typically proceeded to ∼10% higher conversion
after 24 h. No reaction was observed when the reaction
was carried out in water-saturated TBME or in the
presence of Na2SO4/Na2SO4‚10H2O, a salt hydrate buffer
reported to maintain the water activity (aw) of solutions
at 0.76.19
Toyobo LIP-300 gave better conversion if dried im-
mediately before use, although the effect was somewhat
batch dependent. Substrate solutions in TBME could be
dried by azeotropic distillation, moisture contents <200
ppm being conveniently achieved in this fashion.20 Figure
2 illustrates that addition of increasing amounts of water
to such an azeotropically dried solution of (()-4 resulted
in a corresponding decrease in conversion. Nonetheless,
under these conditions, using 2:1 (w/w) enzyme/substrate,
37-48% conversion could be obtained in 24 h if the
moisture content of the system was <800 ppm.
The inhibitory effect of high moisture levels could be
attributed to hydrolysis of the trifluoroethyl isobutyrate;
under standard conditions, the presence of 1% water in
the bulk solvent resulted in 90% hydrolysis of the
TFEOiBu within 4 h. The higher the moisture content
of the system, the less acylating agent remains available
and the reaction slows down. Hydrolysis of TFEOiBu
releases isobutyric acid which can also inhibit the
enzyme21 or protonate the piperidine ring of (()-4 so that
(18) Zaks, A.; Klibanov, A. M. Proc. Natl. Acad. Sci. U.S.A. 1985,
82, 3192-3196.
(19) Halling, P. J . Biotech. Tech. 1992, 6, 271-276. Hogberg, H.-E.;
Edlund, H.; Berglund, P.; Hedenstrom, E. Tetrahedron: Asymm. 1993,
4, 2123-2126.
(20) Azeotropic distillation of the substrate solution in the presence
of the enzyme resulted in ∼10% lowered conversion, but without
diminution of enzyme selectivity.
(21) For the effect of acids on the enantioselectivity of C. cylindracea
catalyzed acylations with anhydride. See: Berger, B.; Rabiller, C. G.;
Konigsberger, K.; Faber, K.; Griengl, H. Tetrahedron: Asymm. 1990,
1, 541-546.
(17) Sawa is a supplier of the Toyobo enzymes. LIP-300 and LIP-
301 are the same enzyme; the different designations are for different
pricing schedules.

Enzymatic Kinetic Resolution of Atropisomers
J . Org. Chem., Vol. 65, No. 18, 2000 5455
Ta ble 2. Isobu tyr yla tion of (()-4 w ith Va r iou s En zym es in TBMEa
run
vendor
enzyme
source
ees, %
eep, %
conv. , %
E
1
2
3
4
5
6
7
8
9
10
11
12
13
Toyobo
Toyobo
Toyobo
Boehringer-Mannheim
Boehringer-Mannheim
Boehringer-Mannheim
Altus
Boehringer-Mannheim
Amano
Meito
Meito
Meito
Novo
LIP-300
P. aeruginosa
P. aeruginosa
P. aeruginosa
72.4
26.7
5.9
26.0
51.7
9.2
5.3
2.1
1.7
2.0
98.7
97.9
92.0
97.7
79.1
91.0
90.8
n/d
n/d
n/d
n/d
n/d
42.3
21.4
6.0
21.0
39.6
9.2
5.5
n/d
n/d
n/d
>200
123
25
110
14
23
22
n/d
n/d
n/d
n/d
n/d
n/d
lipoprotein lipase (LPL-701)
lipoprotein Lipase (LPL-311)
Chirazyme L6
Lipase
Chirazyme L4
ChiroClec PC
Chirazyme L3
Lipase AY-30
Lipase MY
Pseudomonas sp.
Pseudomonas sp.
Pseudomonas sp.
Pseudomonas cepacia
C. rugosa
C. cylindracea
C. cylindracea
Alcaligenes sp
Alcaligenes sp
C. antarctica, type B
Lipase QLC
Lipase QLG
Novozyme 435
1.4
1.3
1.1
n/d
n/d
n/d
n/d
a
Conditions: (()-4, 25 mg, 50 mM; TFEOiBu, 5 equiv; enzyme, 6-27 mg; 4 Å sieves, 25-40 mg; TBME, 1.0 mL; 250 rpm; room
temperature; 23.5 h.
F igu r e 1. Calculated conversion at 4.75 and 24 h for the
isobutyrylation of (()-4 using various P. aeruginosa prepara-
tions: 1, LIP-300; 2, LPL-311; 3, LIP-321; 4, LPL-701 (Sawa);
5, LIP-301 (Sawa). Conditions: (()-4 (53 mg, 60 mM), enzyme
(50 mg), TFEOiBu (5 equiv), TBME (2.0 mL), 4 Å sieves (47-
68 mg), 250 rpm, room temperature.
F igu r e 2. Effect of increasing moisture content on the
enzymatic isobutyrylation of (()-4.
it is unavailable for reaction.22 In addition the hydrolysis
of TFEOiBu releases trifluoroethanol, which is also
capable of inhibiting the reaction. The effects of the
concentration of TFEOiBu, isobutyric acid, and trifluo-
roethanol are shown in Figures 3-5.
As an alternative to drying the reaction mixture by
azeotropic distillation or addition of molecular sieves,
running the reaction in 10-20% Et3N/TBME also re-
sulted in higher conversions. Presumably the added base
neutralized the isobutyric acid formed by adventitious
hydrolysis and reversed any enzyme inhibition or sub-
strate protonation. The beneficial effect of organic base
on both the reactivity and selectivity of enzymatic acyl-
ations has been reported,23 but in the present case, there
was no further improvement in the selectivity of the
reaction.
(22) At the concentrations shown in Figure 5, to produce 0.5 equiv
of isobutyric acid by hydrolysis of TFEOiBu would require 486 ppm
water.
F igu r e 3. Effect of TFEOiBu concentration on the enzymatic
isobutyrylation of (()-4.
(23) (a) Rakels, J . L. L.; Straathof, A. J . J .; Heijnen, J . J . Tetrahe-
dron: Asymm. 1994, 5, 93-100. (b) Boaz, N. W.; Zimmerman, R. L.
Tetrahedron: Asymm. 1994, 5, 153-156. (c) Stead, P.; Marley, H.;
Mahmoudian, M.; Webb, G.; Noble, D.; Ip, Y. T.; Piga, E.; Rossi, T.;
Roberts, S.; Dawson, M. J . Tetrahedron: Asymm. 1996, 7, 2247-2250.
(d) Parker, M.; Brown, S. A.; Robertson, L.; Turner, N. J . Chem.
Commun. 1998, 2247-2248.
(iii) Enzyme Recovery and Reuse. The use of a bulky
acylating agent to enhance the enantioselectivity of the
enzyme resulted in a sluggish reaction which required
2:1 (w/w) catalyst/substrate to achieve ∼40% reaction in
24 h. At such high catalyst loadings, recovery and reuse

5456 J . Org. Chem., Vol. 65, No. 18, 2000
Morgan et al.
F igu r e 4. Effect of added isobutyric acid on the enzymatic
isobutyrylation of (()-4.
F igu r e 6. Reuse of Toyobo LIP-301 in a packed column (5.5
× 1.0 cm enzyme bed); flow rate 1.0 mL/min ((()-4 (0.87 g) in
TBME (19 mL, 0.1 M)).
F igu r e 7. Performance of recovered enzyme after various
treatments. (See text for details).
tions, poor reactivity was observed. Further drying of this
enzyme did not restore activity, but complete activity
could be restored by suspending the enzyme in damp
TBME, followed by evaporating the solvent, and drying
the enzyme under vacuum.
The behavior of recovered and reactivated enzyme is
shown in Figure 7. Using the recovered enzyme without
rehydration (run 1), or with only further drying (run 2),
resulted in poor activity. Suspending the enzyme in
TBME alone, followed by decanting the supernatant
solvent, and evaporating and drying the residue, did not
fully restore performance (run 4). Only when suspended
in damp TBME, filtered, and subsequently dried (run 3)
did the catalyst perform comparably to unused enzyme
samples.
F igu r e 5. Effect of added trifluoroethanol on the enzymatic
isobutyrylation of (()-4.
of the enzyme becomes essential; column containment is
a convenient way to reuse enzyme over a large number
of cycles.
As shown in Figure 6, the enzyme performed consis-
tently over 15 cycles, producing isobutyramide 6c in 99%
ee and 44-49% conversion after 24 h. The results
indicated that the enzyme was stable when submerged
in TBME for 3-4 weeks.
Despite the encouraging results of the packed enzyme
column, due to accelerated project time-lines, the enzy-
matic step was scaled up to run in batch mode. A single
50 kg batch of the enzyme behaved as expected when
used in the acylation of two 25 kg batches of (()-4 in
quick succession. The enzyme was then recovered, stored
as a TBME damp filter cake for ∼1 month, then finally
dried, and stored at room temperature. When samples
of this recovered enzyme were used in subsequent reac-
Our experience using Toyobo LIP-300 lipase, both fresh
and recovered enzyme, emphasizes the importance of
moisture content as a parameter in enzyme-catalyzed
transesterifications. While multiple reuse was demon-
strated for enzyme samples contained in a column,
enzyme recovered from batch reactions showed a large

Enzymatic Kinetic Resolution of Atropisomers
J . Org. Chem., Vol. 65, No. 18, 2000 5457
Sch em e 4
Ta ble 3. En zym a tic Resolu tion a n d Recyclin g of (()-4
decrease in activity but no major decrease in enantiose-
lectivity. Since the reactivity could be restored by sus-
pending the enzyme in damp TBME (0.5-7.5% water/
enzyme v/w) and then redrying, it was concluded that
the enzyme had become dehydrated. However, the mech-
anism of such dehydration remains obscure.
B. Ra cem iza tion a n d Recyclin g. The perceived
advantage for carrying out the enzymatic resolution of
(()-4 rather than that of (()-5 (whether by enzymatic
resolution or by diastereomeric salt formation), was the
potential for racemizing the unwanted (-)-4 and recy-
cling the racemate through further resolutions, increas-
ing the yield of SCH66336.
(()-4
(+)-4
yield, %
42.2
racemization
yield, %
resolution
1st res.
ee, %
95.6
1st racem.
2nd racem.
72.0
81.5
2nd res.
41.3
98.0
98.7
(-)-4 could be rapidly racemized in refluxing phenyl
ether (260 °C), and the (+)-4 isolated from a subsequent
enzymatic resolution displayed an impurity profile which
was essentially indistinguishable from that of the initial
racemic material. However, the longer heating and
cooling cycle required when carrying out the racemization
on large scale resulted in unacceptable levels of several
impurities that could not easily be removed subsequently.
The racemization was therefore studied in a set of
seven high-boiling solvents,24 the best results being
obtained in refluxing di(ethyleneglycol) dibutyl ether
(DEGDBE). Moreover, impurity levels could be controlled
by carrying out the reaction at lower temperature; the
racemization was complete after heating in DEGDBE (5:1
v/w) at 210 °C for 12-18 h. Although the levels of some
impurities increased under these conditions, several were
completely removed during the subsequent enzymatic
resolution and the others were reduced to acceptable
levels by existing downstream processing. Multiple ra-
cemizations and resolutions of the same sample of (-)-4
have been demonstrated. A sample of (()-4 was subjected
to three rounds of enzymatic resolution, using the same
sample of enzyme, with the recovered (-)-4 subjected to
two rounds of racemizations, resulting in a 65% overall
yield of (+)-4, with an acceptable impurity profile (Table
3).
3rd res.
overall
39.8
65
Ta ble 4. Acyla tion of (()-5 Usin g LIP -300 in Va r iou s
Solven tsa
run
solvent
TBME
(()-5, mg/mL ees, % eep, % conv., %
E
1
2
3
4
5
6
7
8
25
5
81.8
83.5
47.4
4.4
20.2
0.4
23.6
12.3
97.1
96.7
88.9
98.0
96.2
97.9
93.2
97.2
45.7
46.3
34.8
4.3
17.3
0.4
20.2
11.3
>100
>100
27
THF
50
5
n/d
63
toluene
MeCN
50
5
n/d
36
25
5
81
a
Conditions: (()-5, 50 mg; TFEOiBu, 5 equiv; LIP-300, 50 mg;
room temperature; 250 rpm; 42 h.
tion; the third enzyme, Novozyme 435 (Candida antarc-
tica) acylated the undesired (S)-(-) enantiomer, resulting
in a potentially more useful subtractive resolution.
However, Novozyme 435 showed poor selectivity (E < 25)
when tested with a series of short-chain acylating agents
(alkyl or trifluoroethyl esters including TFEOiBu, or
anhydrides) either neat or in TBME, or when examined
in 10 common solvents with TFEOAc as acylating agent.
Similarly, poor enantioselectivity (E < 20) was displayed
by ChiroCLEC PC using TFEOAc as acylating agent in
eight common solvents.
C. En zym a tic Acyla tion of (()-5. Our first ap-
proach to the enzymatic kinetic resolution of racemic
SCH66336 intermediates had been the acylation of (()-5
(Scheme 4). From an initial screen of 236 enzyme
preparations, three enzymes showed enhanced reactiv-
ity.25 Two of these, Toyobo LIP-300 (P. aeruginosa) and
ChiroCLEC PC (P. cepacia), acylated the desired (R)-(+)
enantiomer, resulting in a less convenient direct resolu-
Mindful of the enhancement of reactivity and selectiv-
ity in the resolution of (()-4 when using dried enzyme
and a bulky acylating agent, the acylation of (()-5 was
reexamined using trifluoroethyl isobutyrate (TFEOiBu)
and freshly dried Toyobo LIP-300. The results of reac-
tions in four solvents at two concentrations are shown
in Table 4. In THF and toluene under dilute conditions,
the extent of reaction was negligible (runs 4, 6), whereas
it was slower but more selective in MeCN at higher
dilution (runs 7, 8). Excellent selectivity (E > 100) and
reactivity were observed in TBME under both conditions
(runs 1, 2).
(24) A 5:1 solvent/(-)-4 mixture was maintained at reflux, or else
heated to 240 °C over a period of 2.5 h, maintained at 240 °C for 2 h,
then cooled, and subjected to a standard workup. Solvents (bp, °C) were
Decalin (189-191), diethylbenzene (180-182), di(ethyleneglycol) dibu-
tyl ether (256), heptamethylnonane (240), hexadecane (287), N-
methylacetamide (204-206), phenyl ether (259).
The enzymatic resolution of (()-5 on a 1 g scale was
demonstrated using the same conditions (dried enzyme,
azeotropically dried solvent) as used successfully for (()-
(25) Conditions: (()-5, 5 mg; isopropenyl acetate, 10 equiv; EtOAc
or TBME, 2 mL; room temperature; 250 rpm.

5458 J . Org. Chem., Vol. 65, No. 18, 2000
Morgan et al.
doubled due to rotamers): δ 19.27, 19.38, 29.80, 30.00, 30.64,
30.79, 31.13, 31.68, 32.16, 41.90, 42.35, 45.16, 45.58, 118.88,
123.24, 126.85, 129.96, 130.10, 130.14, 130.60, 133.56, 134.30,
134.43, 139.32, 139.49, 140.81, 141.70, 141.83, 147.54, 147.65,
151.96, 152.10, 175.25, 175.30. MS (CI+CH4) (m/z): 539 (M+1,
100).HRMS (FAB)(m/z)Calcd for C23H23N2OBrCl81Br: 539.9924.
Found: 539.9917. Anal. Calcd for C23H23N2Br2ClO‚0.5 H2O:
C, 50.44; H, 4.42; N, 5.11; Br, 29.18; Cl, 6.47. Found: C, 50.46;
H, 4.58; N, 5.10; Br, 27.17; Cl, 6.70.
The combined 0.5 M H2SO4 extracts were added slowly to a
mixture of 50% NaOH (2 mL) and water (18 mL), and the
precipitated solid, (-)-4, was filtered, washed with water, and
dried under vacuum (1.00 g, 43.3%; 96.3% ee). [R]2D5 ) -181.84
(c 1.02, MeOH).
4, and proceeded with high enantioselectivity (ees, 98.5%;
eep, 94.3%; c, 51.1%; E >100). Again, by choosing the
appropriate acylating agent, a highly selective acylation
could be effected. However, since there was no possibility
for recovery and racemization of the unwanted enanti-
omer, this procedure was not pursued.
Con clu sion
A highly selective enzymatic kinetic resolution of
4-substituted piperidines was developed by a careful
choice of reaction conditions. In particular, the choice of
the bulky trifluoroethyl isobutyrate as acylating agent
resulted in formation of the isobutyramide product in
high enantiomeric excess and allowed its facile separation
from the unreacted starting material by a simple acid
extraction. In the case of the compound (()-4 which exists
as a pair of conformational enantiomers, the unreacted
(-)-enantiomer could be recovered, racemized thermally,
and resubmitted to enzymatic acylation. The enzyme
could be recovered and reused but was sensitive to the
water content of the system. While fresh enzyme had to
be dried before initial use, the recovered enzyme required
rehydration and redrying for optimum activity. This
enzymatic step was successfully integrated into the
synthesis of a novel farnesyl protein transferase inhibitor,
SCH66336.
En zym a tic Bu tyr yla tion of (()-4. Toyobo LIP-300 (5 g)
and TFEOBu (2.4 mL, 15.8 mmol) were added to an azeotro-
pically dried stock solution of (()-4 (2.5 g, 5.3 mmol) in TBME
(50 mL), and the mixture was shaken at 200 rpm at room
temperature for 5 h. The reaction was submitted to the same
workup as described above to yield (+)-6d (1.18 g, 41.3%;
95.6% ee). [R]2D5 ) +173.13 (c 1.06, MeOH). IR (KBr pellet):
1640 (CdO), cm-1 1H NMR (400 MHz, CDCl3): δ 0.95, 0.96
.
(2t, 3H, J ) 7.4 Hz), 1.65 (m, 2H), 2.07 (m, 2H), 2.24-2.42 (m,
1H), 2.59, 2.69 (2m, 1H), 2.75 (dt, 1H, J ) 13.6, 4.0, 4.0 Hz),
2.87 (bt, 2H), 3.20-3.36 (m, 3H), 3.41 (td, 1H, J ) 13.6, 13.6,
4.4 Hz), 3.72 (m, 1H), 4.03 (bsex, 1H), 7.20 (d, 1H, J ) 1.6
Hz), 7.40 (d, 1H, J ) 2.0 Hz), 7.54 (bd, 1H), 8.54 (bs, 1H). 13C
NMR (75 MHz, CDCl3) (signals doubled due to rotamers): δ
13.97, 18.76, 29.88, 30.72, 31.16, 31.60, 32.24, 35.28, 41.72,
42.16, 45.47, 45.88, 119.07, 123.37, 123.43, 126.95, 130.26,
130.35, 133.73, 133.80, 134.55, 134.77, 139.24, 139.40, 139.99,
140.16, 141.33, 141.54, 141.66, 141.84, 147.34, 147.56, 151.77,
151.81, 171.46, 171.53. MS (CI+CH4) (m/z): 539 (M+1) (100).
Anal. Calcd for C23H25N2Br2ClO‚1 H2O: C, 49.62; H, 4.53; N,
5.03; Br, 28.70; Cl, 6.37. Found: C, 49.72; H, 4.17; N, 4.79;
Br, 28.34; Cl, 6.24. (-)-4 was obtained as described above (1.35
g, 54.0%; 71.1% ee). [R]2D5 ) -128.18 (c 1.04, MeOH)
Exp er im en ta l Section
Gen er a l. HPLC was carried out on a Waters 715 Ultra
Wisp. Chiral HPLC was carried out on a Chiralpak AD column
(0.46 × 25 cm) (Chiral Technologies, Inc.) (20% iPrOH/heptane/
0.2% DEA; 1.0 mL/min; 240 nm; 30 °C for 4 and 6c and 15%
EtOH/heptane/0.5% DEA/1.0 mL/min; 240 nm; 30 °C for 6a ).
Retention times were as follows: for 4 (5.80, 17.53 min), 6c
(4.95, 10.08 min), 6a (9.26, 10.92 min). Nonchiral reverse-
phase HPLC was carried out on a Symmetry C-18 column (0.46
× 25 cm) (Waters) (33% MeCN/50 mM phosphate buffer (pH
3); 1.5 mL/min; 240 nm; room temperature). Optical rotations
were determined on a Perkin-Elmer 243 B polarimeter. Flash
chromatography was carried out with Sorbisil C60 (40/60A)
or Selecto 32-63. All chemicals were used as received from
commercial sources; trifluoroethyl esters were prepared from
trifluoroethanol and the corresponding acid chloride or anhy-
dride under standard conditions. Trifluoroethyl isobutyrate
was obtained from ChemoDynamics Inc., Sayreville, NJ . With
the exception of Toyobo LIP-300/301, all enzymes were used
as received from suppliers; Toyobo LIP-300/301 was typically
dried under vacuum or under a N2 stream before use.
En zym a tic Isobu tyr yla tion of (()-4. (()-4 (2.34 g, 5.0
mmol) was dissolved in TBME (75 mL) and filtered, and the
filtrate was dried by azeotropic distillation, removing 30 mL
of solvent (KF 169 ppm). Toyobo LIP-300 (5 g) was added, and
the mixture was stirred for a further 0.5 h (KF 283 ppm).
TFEOiBu (2.4 mL, 14.9 mmol) was then added, and the
mixture shaken at room temperature at 200 rpm. After 26 h,
the reaction was filtered, and the enzyme cake washed with
TBME (20 mL). The combined filtrate was extracted three
times with 0.5 M H2SO4. The organic layer was then extracted
twice with 6 M H2SO4. The combined 6 M H2SO4 extracts were
added slowly to a mixture of 50% NaOH (9 mL) and ice (25 g).
The solid which precipitated, (+)-6c, was filtered, washed with
water, and dried under vacuum (1.15 g, 42.6%; 97.0% ee).
[R]2D5 ) +176.53 (c 1.03, MeOH). IR (KBr pellet): 1625 (CdO)
cm-1. 1H NMR (400 MHz, CDCl3): δ 1.09-1.15 (m, 6H), 2.00-
2.12 (m, 1H), 2.24-2.44 (m, 2H), 2.59,2.64 (2m, 1H), 2.75 (dt,
1H, J ) 14.0, 4.4, 4.4 Hz), 2.74-2.84 (m, 1H), 2.87 (bt, 1H),
3.20-3.36 (m, 3H), 3.42 (td, 1H, J ) 13.6, 13.6, 4.4 Hz), 3.76
(m, 1H), 4.05 (m, 1H), 7.20 (d, 1H), 7.47 (d, 1H, J ) 2.0 Hz),
7.52 (bs, 1H), 8.51 (bs, 1H). 13C NMR (75 MHz, CDCl3) (signals
En zym a tic Acetyla tion of (()-4. Toyobo LIP-300 (4.6 g)
and TFEOAc (1.7 mL, 14.5 mmol) were added to an azeotro-
pically dried stock solution of (()-4 (2.3 g, 5.3 mmol) in TBME
(46 mL), and the mixture was shaken at 200 rpm at room
temperature for 2.75 h. The reaction was submitted to the
same workup as described above to yield (+)-6a (1.0 g, 40.3%;
89.3% ee). [R]2D5 ) +140.82 (c 1.02, MeOH). IR (KBr pellet):
1639 (CdO), cm-1. 1H NMR (400 MHz, CDCl3): δ 2.0-2.1 (m,
1H), 2.09, 2.11 (2s, 3H), 2.29 (m, 1H), 2.37 (m, 1H), 2.59, 2.66
(2m, 1H), 2.75 (dt, 1H, J ) 12.0, 4.4, 4.4 Hz), 2.87 (bt, 1H, J
) 16 Hz), 3.22-3.36 (m, 4H), 3.42 (td, 1H, J ) 13.8, 13.8, 4.4
Hz), 3.37 (bsept, 1H), 3.40 (bsept, 1H), 7.20 (t 1H J ) 1.6 Hz),
7.48 (d, 1H J ) 2.0 Hz), 7.52 (bs, 1H), 8.5 (t, 1H, J ) 2.4 Hz).
13C NMR (75 MHz, CDCl3) (signals doubled due to rotamers):
δ 21.46, 29.72, 30.49 30.58, 31.22, 31.44, 32.23, 32.27, 41.63,
42.05, 46.24, 46.66, 118.97, 119.03, 123.28, 123.32, 126.94,
130.22, 130.30, 130.89, 133.66, 133.70, 134.34, 134.48, 139.26,
139.30, 139.37, 139.49, 140.76, 140.89, 141.72, 141.88, 147.67,
147.80, 152.01, 152.18, 168.84, 168.91. MS (CI+CH4)(m/z): 511
(M+1) (100). Anal. Calcd for C21H19N2Br2ClO‚1 H2O: C, 47.71;
H, 4.00; N, 5.30; Br, 30.23; Cl, 6.71. Found: C, 47.96; H, 3.57;
N, 5.09; Br, 29.89; Cl, 6.56. (-)-4 was obtained as described
above (1.32 g, 57.9%. 81.1% ee). [R]2D5 ) -124.50 (c 1.00,
MeOH).
En zym a tic Isobu tyr yla tion of (()-5. A mixture of (()-5
(1.01 g, 2.2 mmol) in TBME (100 mL) was dried by azeotropic
distillation, reducing the volume to 50 mL. This solution was
transferred to a flask containing Toyobo LIP-300 (2 g) which
had been held under a stream of N2 for 2 h. After being shaken
for 0.5 h, TFEOiBu (1.0 mL, 6.2 mmol) was added, and the
reaction was shaken at 200 rpm at room temperature for 20
h. The reaction mixture was filtered, the enzyme cake was
washed with CH2Cl2, and the filtrate was evaporated to dry-
ness. The products were separated by silica gel chromatogra-
phy, eluting with 5-20% MeOH/CH2Cl2/0.2% Et3N and frac-
tions of 20-50 mL were collected. Fractions 3-5 were com-
bined and evaporated to yield the isobutyramide (+)-7b as a

Enzymatic Kinetic Resolution of Atropisomers
J . Org. Chem., Vol. 65, No. 18, 2000 5459
white foam (0.49 g, 42.2%; 94.3% ee). [R]2D5 ) +70.20 (c 1.04,
MeOH). IR (KBr pellet): 1636 (CdO) cm-1. 1H NMR (400 MHz,
CDCl3): δ 1.06-1.14 (m, 6H), 1.28-1.60 (m, 4H), 2.38 (m, 1H),
2.70-3.20 (m, 4H), 3.26 (dt, 1H, J ) 17.6, 4.4 Hz), 3.63 (bt,
1H, J ) 13.8 Hz), 3.90 (bt, 1H, J ) 13.0 Hz), 4.61 (bd, 1H, J
) 12.8 Hz), 4.89 (dd, 1H, J ) 10.4 Hz), 7.14 (bs, 1H,), 7.49 (d,
1H, J ) 2.0 Hz), 7.54 (bs, 1H), 8.44 (d, 1H, J ) 2.4 Hz). 13C
NMR (75 MHz, CDCl3) (signals doubled due to rotamers): δ,
19.18, 19.53, 29.93, 30.05, 31.04, 31.30, 31.70, 32.02, 32.25,
41.73, 41.89, 41.98, 42.23, 45.31, 45.46, 57.96, 58.08, 118.77,
127.04, 127.15, 129.03, 130.99, 133.02, 135.01, 135.17, 136.98,
137.34, 141.21, 141.36, 142.44, 142.61, 147.21, 147.32, 154.69,
154.85, 175.06. MS (FAB-MS) (m/z): 541.0 (M+1). Anal. Calcd
for C23H25 N2Br2ClO: C, 51.09; H, 4.66; N, 5.18; Br, 29.55; Cl,
6.56. Found: C, 50.96; H, 4.71; N, 5.17; Br, 29.15; Cl, 6.92.
Fractions 18-23 were combined and evaporated to yield
unreacted starting material (-)-5 as a pale yellow powder (0.36
g, 36.0%; 98.5% ee). [R]2D5 ) -48.8 (c 1.01, MeOH).
7.52 (d, 1H, J ) 2.4 Hz), 8.44 (d, 1H, J ) 2.4 Hz). 13C NMR
(75 MHz, CDCl3): δ 32.97, 33.28, 34.29, 43.37, 47.89, 48.11,
60.01, 119.72, 128.31, 130.00, 131.98, 133.81, 136.05, 138.65,
142.28, 143.55, 148.23, 156.40. MS (ESI) (m/z): 471.1 (M+1).
Anal. Calcd for C19H19N2Br2Cl: C, 48.49; H, 4.07; N, 5.95; Br,
33.96; Cl, 7.53. Found: C, 48.54; H, 4.24; N, 5.78; Br, 33.79;
Cl, 7.40.
Ra cem iza tion of (-)-4. A mixture of (-)-4 (100 g, 0.21
mole) (83.0% ee. [R]2D5 ) -159.4 (c 1.01, MeOH)) and diethyl-
eneglycol dibutyl ether (500 mL) was heated at 210 °C under
N2 for 18 h, by which time racemization was complete. After
being cooled to room temperature, the dark solution was dil-
uted with EtOAc (2 L). The reaction flask was rinsed with 0.5
M H2SO4 and EtOAc which were combined with the diluted
reaction mixture. The mixture was stirred for 0.5 h, and the
aqueous acidic layer was separated. The organic layer was fur-
ther extracted with two portions of 0.5 M H2SO4. Decolorizing
charcoal (50 g) was added to the combined acidic extracts
which were then refluxed for 1 h. After being cooled, the solu-
tion was filtered through Celite which was rinsed with 0.5 M
H2SO4. The filtrate was added slowly to an ice cold mixture of
50% NaOH (100 mL) and water (750 mL), and the precipitated
solid was collected, washed until the effluent was neutral, and
dried under vacuum at 60 °C to yield a light yellow solid (75.8
g, 75.8%). Anal. Calcd for C19H17N2Br2Cl‚0.5 H2O: C, 47.78;
H, 3.80; N, 5.87. Found: C, 47.87; H, 3.63; N, 5.73.
A sample of (+)-7b (0.37 g, 0.68 mmol) was refluxed in 6 M
H2SO4 (5 mL) for 3.5 h, by which time HPLC indicated com-
plete hydrolysis. The cooled solution was diluted with water
(25 mL) and extracted with EtOAc (2 × 25 mL). The aqueous
layer was made basic with 50% NaOH and extracted with
EtOAc (25 mL), and the EtOAc layer was washed (2 × 25 mL
of water), dried (Na2SO4), filtered, and evaporated (0.25 g,
79.4%; 93.9% ee). [R]2D5 ) +49.41 (c 1.02, MeOH). IR (KBr pel-
let): 3500 (b), 3333, 2946, 2910, 2815, 2737, 1436, cm-1 1H
.
NMR (400 MHz, CDCl3): δ 1.30 (td, 2H, J ) 9, 3.6 Hz), 1.38-
1.54 (m, 2H), 1.66 (bs, 1H), 2.27 (m, 1H), 2.45 (m, 2H), 2.79
(dt, 1H, J ) 14.8, 4.4, 4.4 Hz), 2.97 (ddd, 1H, J ) 17.4, 13.2,
4.4 Hz), 3.05 (m, 2H, J ) 12.8 Hz), 3.27 (dt, 1H, J ) 13.6, 4.4,
4.4 Hz), 3.64 (td, 1H, J ) 17.2, 13.9, 4.4 Hz), 4.90 (d, 1H, J )
10.4 Hz), 7.12 (d, 1H, J ) 2.0 Hz), 7.48 (d, 1H, J ) 2.0 Hz),
Su p p or t in g In for m a t ion Ava ila b le: 1H and 13C NMR
spectra of compounds (+)-4, (+)-6a , (+)-6b, (+)-6c, and (+)-
7b. This material is available free of charge via the Internet
at http://pub.acs.org.
J O991513V