Nitrilase-catalysed desymmetrisation of 3-hydroxyglutaronitrile: Preparation of a statin side-chain intermediate

Article abstract of DOI:10.1021/op050257n

An efficient, scaleable synthesis of ethyl (R)-4-cyano-3-hydroxybutyrate, a potential intermediate in the synthesis of Atorvastatin (Lipitor), has been developed. The three-stage process starts with reaction of low-cost epichlorohydrin with cyanide to give 3-hydroxyglutaronitrile (3-HGN). The second stage utilises a nitrilase-catalysed desymmetrisation of 3-HGN. The nitrilase reaction has been optimized to work at 3 M (330 g/L) substrate concentration, pH 7.5,27 °C. Under these conditions, with an enzyme loading of 6 wt %, 100% conversion and 99% ee product is obtained in 16 h. This material is then esterified to give the target compound, ethyl (R)-4-cyano-3-hydroxybutyrate. The cost-effectiveness of the process is determined by three factors: use of a low-cost starting material, the introduction of the chiral centre by desymmetrisation as opposed to kinetic resolution, and the use of Pfenex Expression Technology to allow a lower-cost supply of biocatalyst.

Full text of DOI:10.1021/op050257n

Organic Process Research & Development 2006, 10, 661−665
Nitrilase-Catalysed Desymmetrisation of 3-Hydroxyglutaronitrile: Preparation of
a Statin Side-Chain Intermediate
Sophie Bergeron, David A. Chaplin, John H. Edwards,† Brian S. W. Ellis,† Catherine L. Hill, Karen Holt-Tiffin,*
Jonathan R. Knight, Thomas Mahoney, Andrew P. Osborne, and Graham Ruecroft‡
Dowpharma, Chirotech Technology Ltd., a Subsidiary of The Dow Chemical Company, Cambridge Science Park,
Milton Road, Cambridge CB4 0WG, UK
Scheme 1
Abstract:
An efficient, scaleable synthesis of ethyl (R)-4-cyano-3-hydroxy-
butyrate, a potential intermediate in the synthesis of Atorvas-
tatin (Lipitor), has been developed. The three-stage process
starts with reaction of low-cost epichlorohydrin with cyanide
to give 3-hydroxyglutaronitrile (3-HGN). The second stage
utilises a nitrilase-catalysed desymmetrisation of 3-HGN. The
nitrilase reaction has been optimized to work at 3 M (330 g/L)
substrate concentration, pH 7.5, 27 °C. Under these conditions,
with an enzyme loading of 6 wt %, 100% conversion and 99%
ee product is obtained in 16 h. This material is then esterified
to give the target compound, ethyl (R)-4-cyano-3-hydroxy-
butyrate. The cost-effectiveness of the process is determined
by three factors: use of a low-cost starting material, the
introduction of the chiral centre by desymmetrisation as
opposed to kinetic resolution, and the use of Pfjenex Expression
Technology to allow a lower-cost supply of biocatalyst.
Scheme 2
Scheme 3
Introduction
Statins are HMG-CoA reductase inhibitors that are used
for the treatment of hypocholesterolemia and atherosclerosis.
Atorvastatin (Lipitor),1 a statin launched in 1997, is now the
world’s largest-grossing drug with 2004 sales of $12 billion.
A number of different routes to statin side chains have been
reported using a biocatalytic step to introduce one of the
two chiral centres.2 Recently, scientists at Codexis described
another process whereby ethyl-4-chloroacetoacetate is re-
duced to ethyl (S)-4-chloro-3-hydroxybutyrate by a ketore-
ductase, followed by an enzymatic cyanation to give ethyl
(R)-4-cyano-3-hydroxybutyrate (1), a useful intermediate in
statin side-chain synthesis (Scheme 1).3 In this paper we will
describe development of an alternative efficient route to 1.
The key step employs a highly volume-efficient nitrilase4-
catalysed desymmetrisation of the meso-dinitrile species
3-HGN (2), allowing the product 1, to be obtained in three
steps from epichlorohydrin, an inexpensive starting material
(Scheme 2).
* To whom correspondence should be addressed. Telephone: +44 (0)1223
728010. Fax: +44 (0)1223 728070. E-mail: keholt-tiffin@dow.com.
† Hampshire Chemical Ltd, The Dow Chemical Company, Seal Sands,
Middlesbrough TS2 1UD, U.K.
Results and Discussion
‡ Present address: C3 Technology, Accentus plc, Harwell Business Centre,
B551, Didcot, Oxfordshire OX11 OQJ, U.K.
Preparation of 3-Hydroxyglutaronitrile (2). The syn-
thesis of 2 was investigated by looking at two main routes
(Scheme 3). For the direct, one-pot route, it was discovered
that precise pH control was necessary for the reaction to work
efficiently. Since the pH control required would be difficult
(1) Davignon, J. Atherosclerosis: ID Research Alert 1997, 2(6), 243.
(2) (a) Mu¨ller, M. Angew. Chem., Int. Ed. 2005, 44, 362. (b) Moen, A. R.;
Hoff, B. H.; Hansen, L. K.; Anthonsen, T.; Jacobsen, E. E. Tetrahedron:
Asymmetry. 2004, 15, 1551.
(3) (a) Davis, S. C.; Grate, J. H.; Gray, D. R.; Gruber, J. M.; Huisman, G. W.;
Ma, S. K.; Newman, L. M. WO 04015132, 2004. (b) Davis, S. C.; Jenne,
S. J.; Krebber, A.; Huisman, G. W.; Newman, L. M. WO 05017135, 2005.
(c) Davis, S. C.; Fox, R. J.; Gavrilovic, V.; Huisman, G. W.; Newman, L.
M. WO 05017141, 2005. (d) Davis, S. C.; Grate, J. H.; Gray, D. R.; Gruber,
J. M.; Huisman, G. W.; Ma, S. K.; Newman, L. M.; Sheldon, R.; Wang, L.
A. WO 05018579, 2005.
(4) (a) DeSantis, G.; Zhu, Z.; Greenberg, W.; Wong, K.; Chaplin, J.; Hanson,
S.; Farwell, B.; Nicholson, L.; Rand, C.; Weiner, D.; Robertson, D.; Burk,
M. J. Am. Chem. Soc. 2002, 124, 9024. (b) DeSantis, G.; Wong, K.; Farwell,
B.; Chatman, K.; Zhu, Z.; Tomlinson, G.; Huang, H.; Tan, X.; Bibbs, L.;
Chen, P.; Kretz, K.; Burk, M. J. Am. Chem. Soc. 2003, 125, 11476.
10.1021/op050257n CCC: $33.50 © 2006 American Chemical Society
Published on Web 04/01/2006
Vol. 10, No. 3, 2006 / Organic Process Research & Development
•
661

Chart 1. Effect of the method of nitrilase addition on
reaction progress
at plant scale, this route was quickly eliminated. Therefore,
a two-step process was developed.
Step 1. The synthesis of 4-chloro-3-hydroxybutyronitrile
(CHBN) was optimised to give yields of 93%. The major
impurity found in the CHBN was the target molecule, 2.
The main consideration in this step is the highly reactive
nature of the reaction substrates, epichlorohydrin and hy-
drogen cyanide. It was considered unsafe to charge the
reactor with epichlorohydrin.5 Accordingly, the reactants
were co-fed into an aqueous solution containing base catalyst.
The reaction is initially two phase since epichlorohydrin is
not water soluble. As the reaction proceeds and the water-
soluble CHBN concentration increases, the system becomes
homogeneous which can lead to a significantly increased and
uncontrollable rate of reaction. It was found that the addition
of phase transfer catalyst increases the rate of reaction in
the early stages and avoids any sudden uncontrollable
exothermic activity.
Step 2. The highest yield achieved for the displacement
of the chloride of CHBN by cyanide was 68%. DOE was
used to broadly identify important parameters in this stage
of the process. Stoichiometry, temperature, reaction volume,
and pH were investigated in a series of screening reactions.
The temperature range 40-55 °C was found to be optimum;
below 40 °C the reaction becomes prohibitively slow, and
above 55 °C side reactions become significant. Dilute
conditions were found to reduce impurities arising from
intermolecular interactions, e.g., dimers and polyols. At pH
10 the reaction yield is highest, although polymerisation of
cyanide may be increased. In all cases a yield maximum
(typically 68%) was achieved with further reaction time only
resulting in more byproduct formation. Thus, a yield of 68%
is accompanied with a significant residual concentration of
CHBN; typically 15%. Isolation of 2 proved problematic.
Downstream processing required that the HGN stream should
have low ionic strength and be as clean as possible. Attempts
to distill 2 from aqueous salt solution resulted in significant
tar formation and low recovery. Azeotropic removal of water,
as a strategy to remove inorganic salts, was attempted. This
approach resulted in a tarry phase that was impossible to
filter. Solvent extraction was found to be the only practical
solution. Following a brief series of trials, 2-methyl-propan-
1-ol was selected as the extraction solvent since it was
efficient, inert to the process, easy to recover, and unlikely
to have a detrimental effect in the downstream biotransfor-
mation. Due to the hydrophilic polar nature of 2, it was found
that multiple extractions were necessary. Further purification
of 2 was necessary to remove volatile impurities which
inhibit the nitrilase enzyme in the next stage. Small-scale
batch distillations gave poor yields of good-quality material,
but which performed well in the biotransformation. However,
thermal screening analysis (using the Thermal Screening
Unit6 instrument supplied by Hazard Evaluation Laboratories)
showed the crude 2 to decompose exothermically at tem-
peratures approaching 150 °C; as a result, batch distillation
was excluded as an option for large-scale purification on
safety grounds. A scaleable solution was to perform a single
pass through a wiped film evaporator to remove the trouble-
some volatile impurities. This was achieved at an acceptable
temperature and pressure with reasonable recovery (96 °C,
1.4 mbar, 67% recovery). Material treated in this way was
shown to perform well in the nitrilase desymmetrisation.
GC-MS identified the principal impurities removed in the
distillate as 4-hydroxy-but-3-enenitrile and CHBN. The
economics for the overall process are important, and the
modest yield for the preparation of 2 will be further addressed
in future work. However, on the positive side, the raw
materials are inexpensive when compared to the product and
processing costs.
Preparation of (R)-4-Cyano-3-hydroxybutyric acid (3).
A nitrilase enzyme (BD9570, Diversa) was used in the
desymmetrisation of the prochiral substrate 2 to afford 3. A
series of small-scale experiments were performed to optimize
the nitrilase-catalysed step. First, the method of nitrilase
addition to the reaction mixture was investigated. The
nitrilase (Diversa BD9570, 0.1 unit mg-1 of lyophilised
powder) was either added directly to the reaction mixture
as a lyophilised powder or first rehydrated with buffer
(sodium phosphate, 100 mM, pH 7.2) and then added to the
reaction mixture containing 2. Addition of lyophilized
nitrilase directly to a solution consisting of 2 and buffer
resulted in the enzymatic reaction stalling at 13% conversion
after 10 h (Chart 1). However, addition of a solution
consisting of nitrilase and buffer to 2 resulted in complete
conversion within 14 h (Chart 1). Therefore, it is critical to
rehydrate the lyophilised nitrilase prior to the addition of
substrate to preserve the nitrilase activity.
(5) Epichlorohydrin is unstable in the presence of acid or base. In this reaction,
base catalyst and water are charged to the reaction vessel. Hydrolysis of
the epichlorohydrin is possible, but in the worst case, autopolymerisation
may occur. For these reasons the inventory of epichlorohydrin present is
minimized by batch feeding. HCN is batch fed in all processes. When acid
stabilised, HCN may be considered inert. However, in the presence of
catalytic quantities of any base, cyanide ion is generated which initiates
polymerisation of HCN. In the worst case, explosive polymerisation may
occur. For this reason the inventory of HCN is minimized by batch feeding.
The enthalpy of reaction (for the required reaction) was estimated at ∼100
kJ/mol. The reaction diluent, water, provides a good heat sink, but the
reaction power is controlled by feed rate and efficient cooling.
Next, optimisation of the loading of the nitrilase enzyme
(20, 30, 40, and 60 mg of lyophilized nitrilase, 0.1 unit mg-1
of lyophilised powder) with 1 g of 3-HGN (3 M) was
(6) Singh, J.; Simms, C. Institution of Chemical Engineers Symposium Series
148, Hazards XVI; Institution of Chemical Engineers, 2001; pp 67-79.
662
•
Vol. 10, No. 3, 2006 / Organic Process Research & Development

Chart 2. Effect of nitrilase loading on reaction progress
As a result, under the optimized conditions of the
biotransformation using rehydrated nitrilase BD9570 (6 wt
%), the 3-HGN (3 M) substrate was shown to be converted
to the hydroxy acid (3) in ∼99% conversion within 16 h at
pH 7.5. These conditions were used for scale-up.
The isolation of 3 from the biotransformation was
challenging due to its water solubility and the presence of
cell debris and salts in the reaction mixture. Filtration of the
acidified mixture, or centrifugation, followed by continuous
liquid-liquid extraction with MTBE was the best way to
achieve high yield in the laboratory (ca. 89%). However,
long extraction times (>20 h) meant that it was necessary
to investigate alternative isolation procedures. In some cases
the calcium salt could be formed by treatment of the reaction
mixture directly with calcium hydroxide. This resulted in a
poorly filterable solid, albeit with a slight upgrade in
enantiomeric excess (98.6% to 99.4%). The calcium salt
could be used directly in the next esterification step, if
sufficient acid was added, but was also accompanied by a
slow filtration during product isolation. In other cases the
biotransformation reaction mixture was telescoped into the
esterification step by azeotropic removal of water by co-
distillation with ethanol or methyl ethyl ketone followed by
treatment with sulphuric acid-ethanol at reflux. For reasons
of volume efficiency and throughput, the workup of choice
involved acidification and filtration with Celite filter aid,
followed by conventional batch extraction with methyl ethyl
ketone. This gave good yields (81%), albeit with the principal
byproduct of the nitrilase hydrolysis, 3-hydroxypentanedioc
acid, still present.
Chart 3. Effect of pH on the nitrilase-catalysed reaction
Chart 4. Effect of alkaline pH on the nitrilase-catalysed
reaction
Preparation of Ethyl (R)-4-Cyano-3-hydroxybutyrate
(1). The esterification of 3 was accomplished by treatment
with ethanol and sulphuric acid catalyst at reflux, and was
generally complete within 2 h. However, it was noted that
larger-scale laboratory experiments occasionally stalled at
ca. 90% conversion. In these cases complete conversion was
achieved by evaporation of ethanol and resubmitting the
partially converted substrate to the reaction conditions. On
larger-scale this would be a somewhat cumbersome and
inefficient approach. Thus, care is needed in the biotrans-
formation workup to remove as much water as possible
from the intermediate acid. On plant scale this esterifica-
tion would be performed by concomitant distillation from,
and resparging of, ethanol into the reaction vessel. Final
purification of 1 was achieved by two passes through a
wiped film evaporator. This was the purification method
of choice since small-scale fractional vacuum distillation
had resulted in significant yield loss due to product decom-
position.
investigated at pH 7.2. In each case the nitrilase enzyme was
rehydrated prior to addition to the reaction mixture. At a
nitrilase loading of 20 mg the reaction stalled at 29%
conversion (Chart 2). Whereas, with a nitrilase loading of
30 mg the reaction proceeded to 71% conversion within 31
h (Chart 2). At nitrilase loadings of 40 and 60 mg the
reactions proceeded to completion within 34 and 13 h,
respectively (Chart 2). The nitrilase loading of 60 mg was
chosen for all subsequent optimisation experiments with
3-HGN (1 g, 3 M).
The effect of pH on nitrilase activity was also investigated
(Charts 3 and 4). Under acidic conditions of pH 6.0 and pH
6.5, the reaction stalled at <50% conversion, whereas, at
pH 7.5 and pH 8.0 the reaction proceeded to completion
within 12.5 h. Under more alkaline conditions a slowing in
the rate of conversion was observed. Thus, an optimum pH
range for this reaction is pH 7.5-8.
Biocatalyst Production. A key aspect of process eco-
nomics is the cost of the catalyst. With this in mind, the
nitrilase was expressed in a strain of Pseudomonas fluore-
scens (Pfejnex Expression Technology), a robust gene expres-
sion host developed by The Dow Chemical Company.7 The
organism has proved to have broad application. It grows to
very high cell densities, and target proteins are commonly
(7) Squires, C. H.; Retallak D. M.; Chew, L. C.; Ramseier, T. M.; Schneider J.
C.; Talbot, H. W. BioProcess Int. 2004, December, 54-59.
Vol. 10, No. 3, 2006 / Organic Process Research & Development
•
663

obtained in very high yield and soluble, active form. When
applied to the nitrilase used in this process, excellent re-
sults were obtained. The enzyme was obtained solely as a
soluble, active multimer in excess of 25 g L-1 of fermenta-
tion, a quantity that represented >50% of the total cell
protein. An added advantage of such a high level of
expression of protein is the greatly simplified downstream
processing of the enzyme, a further contributing factor to
the enzyme cost.
and 2.4 w/w% HGN (3.5% yield). No attempt is made to
remove residual cyanide since this can be part of the charge
in step 2.
Step 2: 3-Hydroxyglutaronitrile, 2. Water (855 g, 19.0
mol) and 30% aq sodium cyanide (408 g, 1.0 mol) are
charged to a reaction vessel having heating and cooling
control as previously described. Hydrochloric acid (35%) is
added (typically 15.6 g, 0.06 mol) to adjust to pH 10. This
solution is heated to and maintained at 50-55 °C whilst
charging the prepared CHBN solution (431 g, 1 mol
epichlorohydrin basis) over 4 h. The extent of reaction is
monitored by GC.9 The reaction is stirred out at 50-55 °C,
typically 6 h, until the yield maximum is achieved. Hydro-
chloric acid, (35% , 6.5 g, ∼0.025 mol) is added to adjust
to pH neutral. The dark reaction mixture, (1716 g, 1687 mL)
so obtained, contains ∼11.1% w/w 2; 68% yield based on
epichlorohydrin input.
Purification of 2. Aqueous reaction liquor (4.39 kg) was
extracted with 2-methylpropan-1-ol (4 × 1.6 L). The
combined organic layers were evaporated in vacuo (45 °C,
10 mmHg) to yield crude 2 as a brown liquid (390 g).
Purification was achieved by passing the liquid through a
wiped film evaporator and retaining the residue (jacket
temperature ) 96 °C; vacuum ) 1.4 mbar; throughput )
46 g h-1; jacket surface area ) 0.02 m2). Yield of residue
was 263 g (67% recovery) containing 2 at 91.2% purity (GC);
1H NMR (400 MHz, D2O) δ 4.25-4.33 (m, 1H), 2.65-2.80
(m, 4H).
Conclusion
An efficient three-stage synthesis of ethyl (R)-4-cyano-
3-hydroxybutyrate (1) from low-cost epichlorohydrin in
23% overall yield, 98.8% ee, and 97% purity has been
developed. Further scale-up will allow additional process
improvements to produce the optimum process. The key step
is the introduction of the chiral centre using a nitrilase-
catalysed desymmetrisation of 3-HGN. The nitrilase reac-
tion has been optimized to work at 3 M (330 g/L) sub-
strate concentration, pH 7.5, 27 °C. Under these conditions,
with an enzyme loading of 6 wt %, 100% conversion and
99% ee product is obtained in 16 h. The use of Pfejnex
Expression Technology to produce the enzyme for this
transformation allows a lower-cost supply of biocatalyst. All
of these factors allow a cost-efficient synthesis of a key chiral
building block.
Experimental Section
Preparation of (R)-4-Cyano-3-hydroxybutyric Acid
(3). To an aqueous solution of 100 mM NaH2PO4 at pH
7.5 (510 mL; adjusted to pH 7.5 by addition of 1 N so-
dium hydroxide solution) was added lyophilized nitrilase
enzyme powder (15.15 g). The mixture was stirred at
27 °C (to rehydrate the lyophilised enzyme powder) for
40 min. 3-Hydroxyglutaronitrile (252.5 g, 2.29 mol) was
then charged over ∼10 min. The mixture was stirred at
27 °C for 16 h and then cooled to 2 °C prior to acidifica-
tion by controlled addition of concentrated sulphuric acid
(∼56 mL) to give pH 2 (exotherm is observed). Celite
was charged (25 g) and the slurry filtered. The filtrate
was extracted with methyl ethyl ketone (4 × 400 mL).
The combined methyl ethyl ketone extracts were evap-
orated in vacuo (15 mbar, 40 °C) to yield the product 3 as
1
General Experimental Procedures. H NMR and 13C
NMR were recorded on a Bruker Avance 400 spectrometer
operating at 400 MHz for proton and 100 MHz for carbon.
Chemical shifts are reported in ppm using Me4Si or residual
nondeuterated solvent as reference. GC-MS data were
obtained using a HP 5890 series 2 GC fitted with a HP 5972
series Mass Selective Detector using electron impact ioniza-
tion.
Preparation of 3-Hydroxyglutaronitrile, (2). Caution!
Free HCN may be present at all stages of the following
process description. All work must be carried out in an
efficient fume cupboard. Waste streams should be rendered
cyanide free by treatment with base and formaldehyde or
hypochlorite. During step 1, exothermic activity should be
detected soon after the co-feeds have been initiated. Absence
of any exotherm indicates an accumulation of unreacted
materials which may lead to a subsequent uncontrollable
exotherm.
Step 1: 4-Chloro-3-hydroxybutyronitrile. Water (540
g, 2.5 mol), triethanolamine (44.7 g, 0.025 mol), and
tetrabutylammonium bromide (19.3 g, 0.005 mol) are pre-
mixed and charged to a reaction vessel fitted with a reflux
condenser and efficient stirring and temperature controls.
Hydrogen cyanide (356.4 g, 1.10 mol) and epichlorohydrin
(1110 g, 1 mol) are co-fed from separate feeds to the reaction
vessel over 4 h, maintaining the temperature at 45-50 °C.
The reaction is then allowed to stir out at this tempera-
ture for up to 8 h. The progress of the reaction is moni-
tored by GC.8 The resultant product (2070 g, 1680 mL) is a
yellow solution of typically 63.7 w/w% CHBN (92% yield)
1
a brown liquid (240.8 g, 81% yield; ee 98.8%); H NMR
(400 MHz, CDCl3) δ 4.25-4.34 (m, 1H), 2.50-2.77 (m,
4H).
Preparation of Ethyl (R)-4-Cyano-3-hydroxybutyrate
(1). 3 (239.5 g, 1.855 mol) was dissolved in absolute ethanol
(480 mL). Concentrated sulphuric acid (3.3 mL) was dosed
in (exotherm observed). The mixture was heated at reflux
(80 °C) for 1 h, after which time the reaction was found to
have stalled at ca. 90% conversion. The ethanol was
(8) GC assay: J&W-DB17 column; injector temperature, 250 °C; detector
temperature, 275 °C; carrier gas, He @ 0.95 kg/cm2; oven program: 75 °C
(hold 5 min), heat to 275 °C @ 10 °C/min, hold for 5 min; CHBN elutes
at 10.7 min.
(9) GC assay: J&W-DB17 column; injector temperature, 250 °C; detector
temperature, 275 °C; carrier gas, He @ 0.95 kg/cm2; oven program: 75 °C
(hold 5 min), heat to 275 °C @ 10 °C/min, hold for 5 min; CHBN elutes
at 10.7 min, HGN elutes at 14.5 min.
664
•
Vol. 10, No. 3, 2006 / Organic Process Research & Development

Table 1
The final product 1 was obtained as a colourless liquid
(181.5 g, 62% yield); purity (GC peak area10) 97%; ee
98.8%;11 1H NMR (400 MHz, CDCl3) δ 4.31-4.37 (m, 1H),
4.23 (q, 2H), 2.60-2.67 (m, 4H), 1.27 (t, 3H); 13C NMR
(100 MHz, CDCl3) δ 171.9, 117.4, 64.5, 61.6, 40.4, 25.4,
1st pass
2nd pass
input ester/g
jacket temp/°C
vacuum/mbar
condensor temp/°C
throughput/g h-1
distillate yield/%
residue yield/%
290
51
1.4
6
62
5
275
113
1.4
6
57
66
14.5; [R]25 -26.2° (c ) 1, CHCl3).
D
Acknowledgment
We thank Diversa for provision of the nitrilase BD9570
enzyme.
95
34
subsequently evaporated in vacuo. Ethanol (400 mL) and
concentrated sulphuric acid (3.3 mL) were recharged, and
heating resumed for 1.25 h, after which time conversion was
complete. Ethanol was evaporated in vacuo to yield the crude
product 1 (290 g, yield 99%; ee 98.7%). The crude product
was purified by two passes through a wiped film evaporator.
The residue from the first pass acted as input for the second
as shown in Table 1.
Received for review December 22, 2005.
OP050257N
(10) GC assay: DB-1701 column; injector temperature, 250 °C; detector
temperature, 300 °C; carrier gas, He @ 15 psi; oven program: 60 °C (hold
5 min), heat to 220 °C @ 5 °C/min; 1 elutes at 19.6 min.
(11) GC assay: Chiraldex G-TA column; injector temperature, 180 °C; detector
temperature, 180 °C; carrier gas, He @ 14 psi; oven program: 150 °C (hold
for 10 min), heat to 180 °C @ 10 °C/min, hold for 10 min; (R)-enantiomer
elutes at 9.66 min; (S)-enantiomer elutes at 9.44 min.
Vol. 10, No. 3, 2006 / Organic Process Research & Development
•
665