Mitsunobu inversion of a secondary alcohol with diphenylphosphoryl azide. application to the enantioselective multikilogram synthesis of a HCV polymerase inhibitor

Article abstract of DOI:10.1021/op200002u

The development of a practical synthesis of the hepatitis C virus polymerase inhibitor 1 was necessary to support preclinical safety and human clinical studies. Significant challenges face the process chemist in developing a route to 1 that is amenable to

Full text of DOI:10.1021/op200002u

ARTICLE
pubs.acs.org/OPRD
Mitsunobu Inversion of a Secondary Alcohol with
Diphenylphosphoryl azide. Application to the Enantioselective
Multikilogram Synthesis of a HCV Polymerase Inhibitor
Jeremy P. Scott,*,† Mahbub Alam,† Nadine Bremeyer,† Adrian Goodyear,† Thientu Lam,‡ Robert D. Wilson,†
and George Zhou‡
†Department of Process Research, Merck Sharp & Dohme Research Laboratories, Hertford Road, Hoddesdon, Hertfordshire EN11
9BU, U.K.
‡Department of Chemical Process Development and Commercialisation, Merck and Co. Inc., 126 East Lincoln Avenue, Rahway, New
Jersey 07065, United States
ABSTRACT: The development of a practical synthesis of the hepatitis C virus polymerase inhibitor 1 was necessary to support
preclinical safety and human clinical studies. Significant challenges face the process chemist in developing a route to 1 that is
amenable to multikilogram operation. In particular, an efficient construction of the eight-membered dihydroindolobenzox-
azocine ring and enantioselective synthesis of the secondary amine stereocenter are required. This article describes our process
development of a Mitsunobu protocol to achieve the latter goal which uses diphenylphosphoryl azide at ambient temperature to
invert a scalemic secondary alcohol. The hazard evaluation performed to establish the safety of this protocol and allow pilot-plant
introduction at >8.0 kg scale is discussed. Overall, an enantioselective synthesis of 1 by way of seven isolated intermediates in 32%
overall yield was developed from commercially available materials. This allowed us to prepare over 3 kg of the targeted drug
candidate.
’ INTRODUCTION
available indole fragment 4 to form the eight-membered dihy-
droindolobenzoxazocine ring in moderate to good yield.3
Both of the Medicinal Chemistry approaches allowed for the
multigram synthesis of 1, facilitating early biopharmaceutical,
biological, and toxicological evaluation. However, there were issues
with these synthetic routes that would prohibit further scale-up.
In particular, the thermal instability of aziridine 5 was a concern
as was the number of steps required to prepare this intermediate.4
The stereoselective conversion of secondary alcohol 3 to amine 2
required activation as either mesylate or tosylate derivative and
displacement with sodium azide at high temperature (95-100 °C,
DMF) to introduce the nitrogen functionality. Concerns with
regard to the potential formation of hydrazoic acid in the reactor
headspace together with the thermal stability of the product azide
formed would require addressing this to scale without risk to
equipment and personnel. Both Medicinal Chemistry routes also
entailed a relatively long six-step sequence to elaborate the
N,N-ethylenediamine side chain from amine 2 to deliver the
candidate 1, contributing to an overall longest linear sequence of
16 steps in the best case.
The treatment of hepatitis C virus (HCV) infections repre-
sents a major unmet medical need that impacts an estimated 170
million people throughout the world.1a It is the leading cause of
liver transplantation in the developed world and results in over
10000 deaths annually in the United States. At present there is no
marketed vaccine to prevent the disease nor is there a specific
antiviral agent effective against HCV infection. The standard of
care for chronic hepatitis C patients is based on the combination
of subcutaneous pegylated interferon (Peg-IFN) along with the
oral nucleoside drug, ribavirin. However, side effects and poor
patient response rates, particularly amongst those with genotype
1, render the development of novel anti-HCV therapies as urgent.1b
As part of a programme within Merck Research Laboratories
directed toward the identification of hepatitis C virus NS5B-
polymerase allosteric inhibitors,2 candidate 1 was identified for
development, and a scaleable synthetic route was required to
support early preclinical and clinical evaluation.
The polymerase inhibitor 1 poses a number of challenges with
regard to development of a concise asymmetric synthesis suitable
for multikilogram scale up. Most notably, the construction of the
eight-membered heterocyclic biaryl-containing ring, as well as
the single amine stereocenter were expected to be challenging.
Previous investigations from our Medicinal Chemistry laboratories
had identified two asymmetric routes to 1 proceeding by way of
primary amine 2 (Scheme 1).2 The first-generation route involved
the five-step preparation and use of the bifunctional aziridine reagent
5. A second-generation approach replaced the aziridine with
the commercially available enantiopure epoxide 6. Both of these
bifunctional electrophiles underwent reaction with the commercially
A more expedient route to 1 which addressed a number of the
issues was rapidly developed within the Process Research group
and enabled us to obtain 1 in a total of nine steps on multigram
scale from the biaryl core 4 (Scheme 2). Medicinal chemistry
intermediate alcohol 3 was oxidized to the ketone 7 under modified
Special Issue: Asymmetric Synthesis on Large Scale 2011
Received: January 4, 2011
Published: March 02, 2011
r
2011 American Chemical Society
1116
dx.doi.org/10.1021/op200002u Org. Process Res. Dev. 2011, 15, 1116–1123
|

Organic Process Research & Development
ARTICLE
Scheme 1. Medicinal Chemistry retrosyntheses
Scheme 2. Preparative separation approach via rac-8
Scheme 3. Construction of the eight-membered ring
Pfitzner-Moffatt conditions and then underwent facile reductive
amination with commercially available N,N-dimethylethylenedia-
mine using sodium triacetoxyborohydride (STAB) to afford rac-8.
Preparative chiral phase separation of 85 followed by reductive
methylation with formaldehyde and hydrolysis afforded the desired
1. Unfortunately, the preparative separation of the enantiomers of 8
or intermediates downstream on multikilogram scale proved in-
feasible due to low volume productivity, and an alternative route was
required to support early-phase kilogram-scale preparation of 1.
The Mitsunobu reaction is a well-established transformation
to stereochemically invert secondary alcohols, facilitating a
displacement by activation of the alcohol through the adduct
of an azodicarboxylate and a tertiary phosphine.6 A range of
amine nucleophiles can be used in this reaction, and the forma-
tion of azides is possible at low temperatures. With this in mind,
we envisaged the alcohol 3 as attractive for multikilogram pre-
paration of 1, and our efforts focused on exploiting this inter-
mediate and developing a means to introduce the required nitrogen
side chain to prepare 1 enantioselectively.
Figure 1. Nucleophiles evaluated for Mitsunobu inversion of alcohol 3.
’ RESULTS AND DISCUSSION
O-Alkylation and Cyclisation to Prepare Alcohol 3. (S)-
Glycidyl tosylate 9 was used as the alkylation agent in preference
to the cheaper (S)-epichlorohydrin, as the latter resulted in
epimerization of the stereocenter under all reaction conditions
evaluated (Scheme 3). A slow, reverse addition procedure was
developed to minimise the formation of dialkylated and dimeric
impurities. Polar, aprotic solvents were optimal and the reaction
was carried out in DMAc with potassium tert-butoxide as the base
of choice. Thus, the preformed phenoxide anion was added as a
solution in DMAc over 2 h to the epoxide in DMAc at 65 °C. The
reaction was complete in 30 min and with typical assay yields of
1117
dx.doi.org/10.1021/op200002u |Org. Process Res. Dev. 2011, 15, 1116–1123

Organic Process Research & Development
ARTICLE
Scheme 4. One-pot Mitsunobu inversion and Staudinger reduction
85%. The reaction mixture was then cooled to 45 °C, and ethyl
acetate (1 mL/g of product) was added to avoid gumming during
the subsequent direct crystallisation carried out by the addition of
water (10 mL/g) at 45 °C. Cooling to ambient temperature and
filtration were followed by a purity upgrade of the crude product
in MTBE (4 mL/g).7 Some ring closure to 3 was observed (up to
10 area % (A%)) in this process, and since this was the desired
product of the next step, this upgrade was optimised to retain as
much of this as possible. The bulk scale reaction afforded epoxide
10 in 76% yield and with 95% ee.
Telescoping of the alkylation and cyclisation steps into a one-
pot process was evaluated but resulted in unsatisfactory impurity
levels and low overall assay yields. Consequently, the cyclisation
to form 3 was developed as a discrete step. A reverse addition
protocol was also applied in this step, with slow addition of the
epoxide in DMAc to a slurry of cesium carbonate (0.5 equiv) in
DMAc at 65 °C. This promoted intramolecular over intermole-
cular cyclisation by addition of the substrate into the basic solu-
tion. The reaction was complete in 30 min and was cooled to
45 °C before the addition of isopropyl acetate to aid the crystal-
lisation. This was preferred to ethyl acetate to minimise forma-
tion of the acetate derivative of alcohol 3 which formed by way of
transesterification on aging of the reaction mixture at this elevated
temperature. Addition of water (10 mL/g) over 1 h with seeding
at 40-50% addition ensured smooth crystallisation. Cooling to
ambient temperature and aging of the slurry overnight gave
mother liquor losses of less than 1%, and alcohol 3 was isolated in
88% yield with a typical 3 A% of the seven-membered ring bypro-
duct present. The optical purity was typically 93% ee.
Figure 2. Equipment setup for hydrazoic acid headspace evalution
using FT-IR.
Mitsunobu Inversion. The secondary alcohol 3, after activa-
tion as a tosylate or mesylate, was inert to displacement to a range
of primary and secondary nitrogen nucleophiles evaluated
(MeNH2, MeONH2, or NH3) or underwent elimination to an
olefinic byproduct when employing sodium phthalimide as nu-
cleophile. Displacement of the secondary alcohol using diphe-
nylphosphorylazide (DPPA) and DBU was investigated in toluene
at room temperature; however, no formation of the desired azide
was observed.8 The nucleophiles 11-15 were tested under
Mitsunobu reaction conditions to generate the chiral amine
intermediate (Figure 1). Unfortunately, direct introduction of
any imide or amide functionality as nucleophile was unsuccessful,
and the generation of an azide intermediate proved unavoidable.
Under Mitsunobu conditions, the alcohol 3 was activated with
1.2 equiv of diisopropylazodicarboxylate (DIAD) and 1.2 equiv
of triphenylphosphine in THF (Scheme 4). Following addition
of 1.2 equiv of diphenylphosphorylazide (DPPA),6e smooth con-
version to the azide 16 took place at ambient temperature
(20-25 °C) and with typical assay yields of 95%. The DPPA
was added after the triphenylphosphine had fully reacted with
DIAD; this was critical since triphenylphosphine reacts with
DPPA to form the corresponding iminophosphorane with release
Figure 3. Hydrazoic acid measured in the headspace before and after
DPPA addition.
of nitrogen.10 Diisopropylethylamine (1.0 equiv) was also added
at the reaction outset to ensure the pH of the system remained
basic and thus avoid the release of hydrazoic acid in the reactor
headspace. Staudinger reduction of the intermediate azide to the
iminophosphorane was developed as a through-process to avoid
any significant handling of the azide intermediate 16. Thus the
addition of further PPh3 led to complete consumption of the
azide 16 within 24 h at 30 °C (nitrogen off-gassing observed).
Hydrolysis with water at 50 °C then afforded the primary amine 2
which was isolated as the hydrochloride salt. The 2 molar equiv of
triphenylphosphine oxide formed during the reaction was com-
pletely rejected to the methanol/isopropanol mother liquors.
This process gave the hydrochloride salt of primary amine 2 in
1118
dx.doi.org/10.1021/op200002u |Org. Process Res. Dev. 2011, 15, 1116–1123

Organic Process Research & Development
ARTICLE
Table 1. RC-1 Calorimetry Data
heat of reaction (kJ mol-1 of
adiabatic temperature rise
batch operation
addition of DIAD
alcohol 3
(°C)
comments
-173.1
23.3
addition rate controlled; accumulation <5% (addition over 20 min at
10 °C)
addition of DPPA
addition of THF solution of
Ph3P
-149.3
-277.7
18.5
30.5
70% accumulation (addition over 6 min at 15 °C)
∼50% accumulation (addition over 12 min at 25 °C)
addition of water
-11.8
1.3
-
Scheme 5. Reductive amination and hydrazinolysis
84% isolated yield in excellent chemical purity and with minimal
erosion of enantiopurity (98 A%, 91% ee).10
timecycles employed. A further safety precaution introduced for
the reaction workup was to assay the residual azide content of the
THF layer of the amine 2 prior to acidification. The level was
typically less than 10 ppm based on ion chromatographic analysis
indicating acidification could proceed without incident.
Safety Evaluation of Mitsunobu Process. The Mitsunobu
process engendered the potential formation of the toxic and
highly explosive hydrazoic acid11 as well as the stoichiometric
formation of an organic azide intermediate. An explosive gas
mixture can be formed with air or nitrogen when hydrazoic acid
concentrations exceed 10%.11b Extensive safety studies were
carried out to mitigate the risks involved with operating this
chemical process. Online monitoring of the gas phase by FT-IR
was carried out (Figure 2) and indicated that, prior to the addi-
tion of water to hydrolyse the iminophosphorane, there is less
than 1 ppm HN3 in the headspace at any point in time. Indeed
HN3 was only observed during the DPPA addition and this is
likely due to the above-surface addition mode (Figure 3) and rep-
resented the level of HN3 present in the commercially available
DPPA used. After the addition of water and subsequent aging at
50 °C over 24 h, no significant amount of HN3 is released to the
headspace. Apparently, the addition of 1.2 equiv of diisopropy-
lamine ensures any azide remains in the solution phase as the ion.
Any risk of potential accumulation of hyrazoic acid in the
headspace was further reduced by applying a constant sweep of
nitrogen over the head space during plant operation. In addition,
to avoid any opportunity for explosive liquid hydrazoic acid
condensates to form upstream of the reactor, the reactor con-
denser system was not used during the process. Any presence of
heavy metals which could catalyse azide decomposition was also
avoided through ICP analysis for metal residues in the input
materials and through diligent cleaning of the reactor.
N,N,N0-Trimethylethylenediamine side-chain installation.
The elaboration of amine 2 to the desired candidate 1 required
six steps using the original medicinal chemistry synthetic routes;
consequently, it was desirable to shorten this. A four-step route to
achieve this goal was developed.
Reductive Amination and Hydrazinolysis. Phthalimidohy-
drate 17 was envisaged to undergo a reductive amination with
primary amine 2 as the first step to installing the ethylenediamine
side chain (Scheme 5). Commercially available phthalimidoace-
taldehyde diethylacetal was hydrolysed to the hydrate 17 in a
procedure developed to avoid any risk of uncontrolled exother-
micity. Thus, a reverse addition protocol, whereby the diethyla-
cetal in THF (2 mL/g) was added to 6 M HCl heated to 40 °C
and led to complete acetal cleavage after a 1-h age, and 17 was
crystallized directly in 71% isolated yield from the reaction mix-
ture by addition of further water.
The reductive amination of the hydrate 17 with the free base of
amine 2 was initially optimized with STAB in isopropylacetate as
solvent, but retention of boron residues in the product were
undesirable for the subsequent chemical step (Scheme 5). Switching
to dichloromethane combined with a pH-controlled workup
procedure removed the boron residues successfully. The STAB
was added portionwise as a solid to a solution of the hydrate 17,
amine hydrochloride 2, and i-Pr2EtN in DCM. This allowed for
control of the reaction exothermicity given STAB is relatively
insoluble in dichloromethane and could not therefore be added
as a solution. Crystallisation of 18 from MTBE (9 mL/g) at 40 °C
led to rapid crystallisation with initially small particle size, and a heat
cycle (aging of batch at 50 °C for 4 h, then cooling to ambient) was
used to increase the particle size to give a more rapid filtration. The
isolated yield was 85%, and fortuitously, the optical purity of the
isolated solids were upgraded to 98% ee at this point.
Density scanning calorimetry (DSC), advanced reactive sys-
tem screening tool (ARSST) experiments as well as RC-1 cal-
orimetry were performed to evaluate potential pressurisation and
thermal properties of this chemical process (Table 1). Small
exotherms are observed upon the additions of DIAD, DPPA,
PPh3 and water. When DIAD is added, the exotherm is addition-
rate controlled, whilst nonproblematic accumulation of reagent
occurs during the DPPA and PPh3 additions based on the charging
1119
dx.doi.org/10.1021/op200002u |Org. Process Res. Dev. 2011, 15, 1116–1123

Organic Process Research & Development
ARTICLE
Scheme 6. Triple methylation and ester hydrolysis
procedure to substitute for the hazardous utilization of sodium
azide at high temperature. In addition, a more expedient elabora-
tion of the ethylenediamine side chain was developed. Imple-
mentation on scale allowed for the preparation of >3 kg of 1 at
>99 A% chemical and >99% optical purity.
’ EXPERIMENTAL SECTION
General. Melting points were determined by closed-cell DSC.
HPLC assays were carried out using a C-18 reversed-phase
column eluted with 0.1% H3PO4 (aq) and acetonitrile. Assay yields
were obtained by HPLC using pure compounds as standards.
Isolated yields refer to yields corrected for purity on the basis of
HPLC assays using purified standards. All reagents and solvents
were used as received without further purification. The 1H and
13C NMR spectra of compounds containing the dihydroindolo-
benzoxazocine ring are complicated by the presence of atropi-
somerism, and only the signals observed at 25 °C are reported.
3-Cyclohexyl-2-(2-oxiranylmethoxy-phenyl)-1H-indole-
6-carboxylic Acid Methyl Ester (10). Methyl ester 4 (9.808 kg)
was dissolved in DMAc (74.7 kg) at room temperature and
potassium tert-butoxide (3.315 kg) added portionwise, maintain-
ing a temperature of 20-25 °C. Separately (2S)-glycidyl tosylate
(9.60 kg) was dissolved in DMAc (18.6 kg) and the solution
heated to 65 °C. Slow addition of the phenol anion solution was
then made to the (2S)-glycidyl tosylate solution, held at 65 °C,
over 2 h. The reaction was aged 30 min, ethyl acetate (9 kg) was
added, and the temperature was adjusted to 45 °C. Water
(105 kg) was added slowly over 30 min, maintaining the batch
temperature at 45 °C. The batch was allowed to cool to ambient
temperature, filtered, and washed with DMAc/water (1:3; 7:22.5
kg) and water (30 kg) and dried in vacuo at 60 °C. The solid
(11.717 kg) was then swished in MTBE (34.7 kg) overnight at
ambient temperature, filtered, and washed with MTBE (6.7 kg)
and dried overnight at 40 °C to give 8.647 kg (76% yield) of 10 as
a white solid. 1H NMR (400 MHz, d6-DMSO) δ 11.33 (1H, br s),
7.98 (1H, d, J = 1.3 Hz), 7.80 (1H, d, J = 8.5 Hz), 7.59 (1H, dd,
J = 8.5 Hz, J = 1.5 Hz), 7.47-7.42 (1H, m), 7.32 (1H, dd, J = 7.5
Hz, J = 1.6 Hz), 7.20 (d, 1H, J = 8.2 Hz), 7.11 (1H, t, J = 7.4 Hz),
4.34 (1H, dd, J = 11.5 Hz, J = 2.7 Hz), 3.94-3.89 (1H, m), 3.86
(3H, s), 3.25-3.21 (1H, m), 2.79-2.76 (1H, m), 2.63-2.57 (2H,
m), 1.94-1.84 (2H, m), 1.75-1.66 (5H, m), 1.38-1.16 (3H, m);
13C NMR (100 MHz, d6-DMSO) δ 167.8, 156.6, 135.9, 134.9,
132.3, 130.5, 130.4, 122.2, 121.8, 121.3, 120.1, 119.1, 199.0, 115.5,
113.4, 69.9, 52.1, 50.1, 44.2, 36.6, 33.1, 27.2, 26.2; mp 184-
186 °C. Chiracel AS-H 250 mm ꢀ 4.6 mm, 5 μm; isocratic 97:3
hexane/ethanol, flow 1.5 mL/min, 254 nm, Rt (min) 24.1
(undesired), 27.6 (desired). Enantiomeric excess 95%. HRMS
(ES) Calcd for C25H28NO4 (MHþ) 406.2018. Found 406.2019.
Methyl(7)-14-cyclohexyl-7-hydroxy-7,8-dihydro-6H-indolo-
[1,2-e][1,5]benzoxazocine-11-carboxylate (3). A solution of
methyl ester 10 (8.6 kg) in DMAc (24.4 kg) was added to a slurry
of cesium carbonate (3.48 kg) in DMAc (56.7 kg) at 65 °C over
2 h and the resulting slurry heated at 65 °C for 30 min. The batch
was then cooled to 45 °C and isopropyl acetate (7.5 kg) added.
Slow addition of water (86.5 L) was then made to the batch over
30 min, maintaining the temperature at 45 °C. The batch was
allowed to cool to ambient temperature (21 °C), aged for 30 min,
and then filtered. The cake was washed with DMAc/water (1:3;
12 kg), water (12 kg) and dried in vacuo at 60 °C to give a pale-
yellow solid (8.3 kg, 92.7% ee, 88% yield). 1H NMR (400 MHz,
d6-DMSO) δ 8.22-8.20 (1H, m) 7.90-7.83 (1H, m), 7.69-7.60
The phthalimide group was removed by hydrazinolysis to
form diamine 19 (Scheme 5). Optimisation of this procedure led
to the use of 3 equiv of hydrazine to minimize formation of
dimeric impurities. Tetrahydrofuran was the solvent of choice,
and the reaction proceeded to completion in a typical 12-14 h at
30 °C with assay yields of 81-91%. The reaction workup was
developed to remove the stoichiometric phthalohydrazide by-
product as the sodium salt through a wash with NaOH. Water
and brine washes of the organic layer of 19 reduced the residual
hydrazine levels to 2-3 ppm, ensuring distillation of the batch
could then take place safely.13 Concentration to ∼4 mL/g total
volume was followed by a controlled crystallisation by solvent
switching to heptane via feed-and-distill, giving the product in
92% isolated yield.
Triple Reductive Methylation and Final Hydrolysis. For-
mation of the penultimate intermediate 20 was completed
efficiently from diamine 19 in a single step by triple reductive
methylation using aqueous acetic acid, formaldehyde, and Pd-
catalysed hydrogenation (Scheme 6). A significant increase of
conversion was observed moving from 30 to 80 psi of hydrogen,
and 4 equiv of formaldehyde were required to drive the reaction
to complete formation of the trimethylated amine. Heating the
reaction mixture above ambient temperature led to inferior
purity profiles. The catalyst of choice was Pd/C JM paste 91,
and a typical 18 mol % was required to obtain complete con-
version. MeOH proved to be the best solvent for the transforma-
tion with ethanol resulting in slower conversion, and the reaction
profile was poor in THF or ethyl acetate. A simple direct crys-
tallization of 20 in 86% yield was achieved following filtration of
the catalyst and addition of 10 M NaOH (0.9 mL/g) to adjust the
pH to 14.
Hydrolysis of penultimate 20 was achieved using 2 M NaOH
in MeOH. Other solvents such as THF afforded slower conver-
sion, although nonalcoholic solvents were preferred to mitigate
any potential risk of formation of the known genetoxic toluene-
sulfonic acid methyl ester.14 The hydrolysis was complete in 3 h
at 70 °C, and salt formation was telescoped into the process. DCM
was used to extract the zwitterionic form of 1 after neutralization
with HCl. After washing with water, addition of toluenesulfonic
acid hydrate as a solution in acetonitrile to the dichloromethane
layer formed the tosylate salt. Impurity rejection on batch con-
centration and isolation was excellent, routinely giving >99 A%
tosylate salt of 1 which was isolated in 86% yield and with >99% ee.
’ CONCLUSION
A concise enantioselective synthesis of the hepatitis C poly-
merase inhibitor 1 was achieved in seven isolated intermediates
with 32% overall yield from commercially available indole 4. The
key issues that prohibited scale up of the medicinal chemistry
route were addressed, most notably by way of the development
of a one-pot Mitsunobu inversion and Staudinger reduction
1120
dx.doi.org/10.1021/op200002u |Org. Process Res. Dev. 2011, 15, 1116–1123

Organic Process Research & Development
ARTICLE
(1H, m), 7.54-7.44 (1H, m) 7.32-7.17 (3H, m), 5.67 (0.7H, s
br), 5.26 (0.3H, s br), 4.60-4.56 (0.3H, m), 4.42-4.38 (0.7H,
m), 4.35-4.31 (0.7H, m), 4.02 (1H, s br), 3.89-3.84 (4H, m),
3.79-3.61 (2H, m), 2.75-2.64 (1H, m), 2.05-1.10 (10 H, m);
13C NMR (100 MHz, d6-DMSO) δ 167.8, 167.7, 159.8,
158.5, 138.0, 137.8, 137.5, 135.9, 132.9, 132.1, 131.7, 131.4,
129.9, 129.8, 124.0, 123.1, 122.8, 122.7, 122.6, 122.1, 122.07,
121.0, 120.6, 120.1, 119.9, 119.3, 118.8, 118.3, 114.3, 111.8,
76.7, 75.3, 67.6, 67.2, 52.3, 52.2, 48.5, 46.7, 36.8, 36.6, 33.3,
33.2, 33.0, 27.1, 26.0, 22.1; mp 225-226 °C. Chiracel OJ-H
5 μm 250 mm ꢀ 4.6 mm, 5 μm; isocratic 95:5:0.1 hexane/
ethanol/diethylamine, flow 1.5 mL/min, 55 °C, 254 nm; Rt (min)
11.5 (undesired), 15.6 (desired). Enantiomeric excess 93%.
HRMS (ES) Calcd for C25H28NO4 (MHþ) 406.2018. Found
406.2033.
vacuo at 40 °C gave 4.037 kg (98.5 GC A%, 75% corrected yield)
of a white solid. 1H NMR (400 MHz, d6-DMSO) δ 7.91-7.75
(4H, m), 6.32 (1H, d, J = 8.0 Hz), 5.12 (1H, dd, J = 5.6, 13.2 Hz),
3.58 (1H, dd, J = 5.6, 13.6 Hz), 3.50 (1H, dd, J = 6.0, 14.0 Hz);
13C NMR (100 MHz, d6-DMSO) δ 168.1, 134.8, 132.1, 123.5,
89.2, 43.0; GC method: RTX5 amine 30 mm ꢀ 320 mm, 1 μm;
helium at 2 mL/min, constant flow, gradient 100 to 250 °C @
20 deg min-1, hold 5 min, FID, injector temperature: 50 °C, split
10:1 detector temperature 250 °C. Diluent as an internal standard
was prepared by making a 0.5 mg/mL solution of benzophenone
in acetonitrile. Rt (min) 17.2.
Methyl-14-cyclohexyl-7-{[2-(1,3-dioxo-1,3-dihydro-2H-
isoindol-2-yl)ethyl]amino}-7,8-dihydro-6H-indolo[1,2-e]-
[1,5]benzoxazocine-11-carboxylate (18). Hydrochloride
salt of amine 2 (6.6 kg) and 2-(2,2-dihydroxyethyl)isoin-
dole-1,3-dione 17 (3.22 kg) were slurried in dichloromethane
(87.5 kg), and the mixture was cooled to 15 °C. Disopropy-
lethylamine (1.93 kg) was then added in one portion over
5 min and the batch aged a further 25 min to give a thin, yellow
slurry. Sodium triacetoxyborohydride (4.22 kg) was then
added in four equal portions (1.3 kg total) at 15-min intervals
at 15 °C and the batch aged a further 80 min. The batch was
cooled to 15 °C and quenched by addition of 1 M HCl
(1.83 kg 37% HCl in 16.4 kg water) over 30 min. NaOH (6
M [10.2 kg, 10 M NaOH in 4.8 kg water]) was then added
while maintaining T < þ20 °C, the phases were separated, and
the lower organic layer was washed with water (9.3 kg), and
the phases were cut. The organic layer was then solvent
switched to MTBE to a final volume of 53 L (9 mL/g based
on starting material input) at final temperature of 40 °C to
give a slurry which was cooled to 20 °C and aged overnight.
Filtration, washing with MTBE (7.3 kg), and drying in vacuo
at 55 °C for 16 h overnight under a nitrogen sweep gave 7.453
kg (85% yield, 97.6% ee) of a white solid. 1H NMR (400 MHz,
d6-DMSO, 4:1 mixture of atropisomers) δ 8.22-8.18 (1H,
m), 7.95-7.65 (6H, m), 7.43-7.31 (1H, m), 7.28-7.05
(3H, m), 4.68 (0.2H, m), 4.50-4.43 (0.8H, m), 4.23-4.18
(0.2H, m), 3.98-3.57 (8H, m), 3.18-2.95 (3H, m), 2.80-
2.76 (1H, m), 2.10-1.85 (4H, m), 1.80-1.50 (5H, m) 1.45-
1.30 (3H. m); 13C NMR (100 MHz, d6-DMSO) δ 137.9,
137.6, 137.4, 135.6, 134.0, 133.8, 133.7, 133.3, 132.2, 132.0,
130.7, 130.5, 130.4, 130.0, 123.4, 123.2, 123.0, 122.7, 122.5,
121.6, 121.5, 120.4, 120.4, 120.2, 120.0, 119.7, 117.9, 112.7,
111.672.5, 68.3, 56.3, 56.1, 52.0, 45.2, 45.1, 44.8, 38.9, 36.8,
36.8, 36.7, 33.3, 33.2, 33.1, 27.0, 27.0; mp 143-145 °C.
Chiracel OD-H 250 mm ꢀ 4.6 mm, 5 μm, isocratic 95:5
hexane/propan-2-ol, flow 1.0 mL/min, 254 nm, Rt (min) 31.9
(undesired), 34.8 (desired). Enantiomeric excess 98%. HRMS (ES)
Calcd for C35H36N3O5 (MHþ) 578.2655. Found 578.2648.
Methyl-7-[(2-aminoethyl)amino]-14-cyclohexyl-7,8-dihy-
dro-6H-indolo[1,2-e][1,5]benzoxazocine-11-carboxylate
(19). The phthalimide derivative 18 (7.35 kg) was dissolved in
tetrahydrofuran (65.3 kg) and hydrazine hydrate (3.45 kg)
added. The mixture was heated to 30 °C and aged overnight.
The reaction mixture was cooled to ambient temperature and
diluted with isopropyl acetate (64 kg), followed by addition of
2 M sodium hydroxide (88 L) and water (44 L), and this mixture
was thoroughly stirred. The phases were cut, and then the
organic layer was washed with distilled water (60 L), and phases
were cut again and washed with 50% saturated brine solution (60 L).
The batch was concentrated to low bulk (26.5 L) by distillation
and flushed with isopropyl acetate (44 kg). The batch was seeded
Methyl (7)-7-Amino-14-cyclohexyl-7,8-dihydro-6H-
indolo[1,2-][1,5]benzoxazocine-11-carboxylate hydrochlor-
ide (2 HCl). Carboxylate 3 (8.44 kg) and PPh3 (6.55 kg,
3
25.0 mol) were slurried in THF (74.1 kg), diisopropylethylamine
(2.69 kg, 20.8 mol) was added and the mixture cooled to 10 °C.
DIAD (5.05 kg, 25.0 mol) was then added slowly over 15 min at
T e þ15 °C and the reaction mixture stirred for 10 min. DPPA
(6.87 kg, 25.0 mol) was then added dropwise over 10 min at
15 °C, and the reaction mixture was warmed to 25 °C over a
period of 30 min and the reaction mixture aged for at least 2 h. A
second batch of PPh3 (7.10 kg, 27.0 mol) dissolved in THF
(6.3 kg) was added over 30 min at T e þ30 °C. [CAUTION:
evolution of nitrogen gas!] The reaction mixture was stirred for at
least 8 h at 25 °C. Water (4.22 kg) was then added, and the batch
was heated to 50 °C and stirred overnight for 18 h. NaOH (2 M,
200 mL) was added, the mixture stirred for 5 min, and the phases
were settled and cut. Ion chromatographic analysis of the organic
layer at this point indicates <4-5 ppm of residual azide (N3-).
MeOH (59.3 kg) was added to the organic layer and a solution of
concentrated HCl (4.70 kg) in IPA (96.5 kg) added dropwise,
maintaining the temperature 20-25 °C; the batch aged at 20 °C
for 19 h. The solid was filtered, washed with IPA twice (21.3 and
25 kg), and dried at 50 °C in vacuo for 18 h to give 6.64 kg
1
(91% ee, 84% yield) of a white solid. H NMR (400 MHz, d6-
DMSO) δ 8.90 (2H, br s), 8.44 (1H, s), 7.91 (1H, d, J = 8.4 Hz),
7.71 (1H, dd, J = 8.4 Hz, J = 1.2 Hz), 7.58-7.54 (1H, m), 7.35-
7.28 (3H, m), 4.75-4.68 (1H, m), 4.52-4.48 (1H, m), 4.15-
4.10 (1H, m), 4.02-3.93 (1H, m), 3.88 (3 H, s), 3.61-3.56 (1H,
m), 2.71-2.65 (1H, m), 2.01-1.93 (3H, m), 1.84-1.82 (1H,
m), 1.76-1.67 (2H, m), 1.54-1.52 (1H, m), 1.37-1.25 (2H,
m), 1.18-1.08 (1H, m); 13C NMR (100 MHz, d6-DMSO) δ
167.7, 159.1, 137.1, 135.9, 132.7, 132.0, 130.1, 124.8, 124.0,
123.2, 122.3, 122.1, 120.7, 120.0, 119.3, 112.4, 52.4, 48.7, 43.8,
36.9, 36.3, 33.2, 33.0, 27.1, 26.0, 25.6; mp >250 °C. Chiracel OD-
H 250 mm ꢀ 4.6 mm, 5 μm; isocratic 98:2 hexane/ethanol
þ0.1% isobutylamine, flow 1.2 mL/min, 254 nm, Rt (min) 24.9
(undesired), 29.3 (desired). Enantiomeric excess 91%. HRMS
(ES) Calcd for C25H29N2O3 (MHþ) 405.2178. Found 405.2196.
2-(2,2-Dihydroxyethyl)-isoindole-1,3-dione (17). Concen-
trated HCl (14.96 L, 12.68 kg) was added slowly to water
(13.84 L) to form 6 M HCl that was aged for 15 min and then
heated to 40 °C. A solution of phthalimidoacetaldehyde diethy-
lacetal dissolved in THF (9.60 kg) was added slowly to the 6 M
HCl over 0.5 h, maintaining the temperature at eþ41 °C. Water
(36 L) was added to the reaction mixture slowly at 40 °C, over
1 h. After cooling and aging overnight at 17 °C, the batch was
filtered and washed with water (28.8 kg). Drying the batch in
1121
dx.doi.org/10.1021/op200002u |Org. Process Res. Dev. 2011, 15, 1116–1123

Organic Process Research & Development
ARTICLE
(5 g), and a slurry formed. Heptane (38.3 kg) was added slowly
over 30 min and aged overnight. The batch was concentrated to
∼26 L and flushed twice with heptane (2 ꢀ 56 L) and then
diluted with heptane (56 L). The solid was filtered, washed with
heptane (12.3 kg), and dried in vacuo at 45 °C to give 5.36 kg
(92% yield and 97.9% ee). 1H NMR (400 MHz, d6-DMSO, 4:1
mixture of atropisomers) δ 8.28 (0.2H, s), 8.13 (0.8H, s), 7.88-
7.82 (1H, m), 7.70-7.55 (1H, m), 7.55-7.40 (1H, m), 7.33 -
7.15 (3H, m), 4.70-4.55 (0.2H, m), 4.55-4.45 (0.8H, m),
4.32-4.25 (0.8H, m), 4.10-4.02 (0.2H, m), 3.98-3.65 (4H,
m), 3.55-3.45 (1H, m), 3.05-2.55 (6H, m), 2.05-1.05 (13 H);
13C NMR (100 MHz, d6-DMSO) δ 167.7, 167.6, 159.8, 158.6,
138.0, 137.8, 135.7, 133.3, 132.1, 131.6, 131.4, 129.9, 129.7,
123.7, 122.8, 122.6, 122.4, 122.0, 121.9, 120.5, 120.2, 120.1,
119.7, 119.4, 118.6, 118.5, 114.2, 111.6, 75.6, 72.4, 57.0, 56.9,
52.3, 52.2, 50.6, 50.5, 46.4, 45.4, 42.4, 42.1, 36.8, 36.6, 33.3, 33.1,
27.1, 20.1, 26.0, 22.1; mp 136-138 °C. Chiralpak AD-H 4.6
mm ꢀ 250 mm, 5 μm, isocratic 75:25 hexane/ethanol þ0.1%
isobutylamine, flow: 1 mL/min, 40 °C, 254 nm, Rt (min) 12
(undesired), 15 (desired). HRMS (ES) Calcd for C27H34N3O3
(MHþ) 448.2618. Found 448.2600.
(7R)-Methyl 14-cyclohexyl-7-[[2-(dimethylamino)ethyl]-
(methyl)amino]-7,8-dihydro-6H-indolo[1,2-e][1,5]benzoxa-
zocine-11-carboxylate (20). Formaldehyde (3.79 kg) and acetic
acid (2.81 kg) were added to a slurry of 19 (5.23 kg) in MeOH
(105 L). A slurry of Pd/C (2.53 kg) in MeOH (15 L) in an inert
container was then added. and the batch was flushed with nitrogen
(twice) and set under a hydrogen atmosphere (flushed three times,
80 psi, ∼5.5 barg). The reaction was aged under a constant
hydrogen atmosphere of 80 psi for 15 h, maintaining T ≈ 28 °C.
The reaction mixture was then filtered through Celite and washed
with MeOH (2 ꢀ 20 L). NaOH (10 M ,4.70 L) was added dropwise
at ambient temperature (23 °C) to the reaction filtrate over 20 min
and then aged for 15 h. Water (2 ꢀ 21 kg) was then added. The
solid was then filtered, washed with water (20 L), and dried at 50°C
in vacuo for 16 h to give 4.96 kg (86% yield) as a white solid. 1H
NMR (400 MHz, d6-DMSO, 4:1 mixture of atropisomers) δ 8.27
(0.2H, s), 8.14 (0.8H, s), 7.93-7.83 (1H, m), 7.72-7.58 (1H, m),
7.55-7.42 (1H, m), 7.33-7.14 (3H, m), 4.71-4.63 (0.2H, m),
4.53-4.45 (0.8H, m), 4.31-4.22 (0.8H, m), 4.10-4.02 (0.2H, m),
3.97-3.55 (3.2H, m), 3.53 (3.45 (1H, m), 3.06-3.00 (0.2H, m),
2.98-2.88 (0.8H, m), 2.86-2.75 (1H, m), 2.75-2.55 (4H, m),
2.48 (3H, br s), 2.01-1.05 (14 H); 13C NMR (100 MHz,
d6-DMSO) δ 167.7, 167.6, 159.8, 158.6, 138.0, 137.8, 137.7,
135.7, 133.3, 132.1, 131.6, 131.4, 129.9, 129.7, 123.7, 122.8, 122.6,
122.6, 122.4, 122.0, 121.9, 120.6, 120.2, 120.1, 119.7, 119.4, 118.6,
118.5, 114.2, 111.6, 75.6, 72.3, 57.0, 56.8, 52.3, 52.2, 50.6, 50.5, 46.4,
45.4, 42.4, 42.1, 36.8, 36.6, 33.3, 33.1, 33.1, 27.1, 27.1, 26.0, 22.1;
mp 139-141 °C. HRMS (ES) Calcd for C30H39N3O3 (MHþ)
490.3070. Found 490.3077.
∼46 L maintaining T < 45 °C. After cooling to ambient temp-
erature, and aging for 1 h, the slurry was filtered and the solid
washed with acetonitrile (15.4 kg). Drying in vacuo at 65 °C
for 16 h gave 5.346 kg (98.0 A%, >99% ee) of 1 as a white solid
1
in 86% yield. H NMR (400 MHz, d6-DMSO) δ 12.62 (1H,
br s), 8.95 (1H, br s), 8.17 (1H, s), 7.88 (1H, d, J = 8.4 Hz),
7.69 (1H, d, J = 8.4 Hz), 7.59-7.7.47 (3H, m), 7.35-7.26
(3H, m), 7.09 (1H, d, 7.6 Hz), 4.66-4.53 (1H, m), 4.36-4.30
(1H, m), 4.10-4.00 (1H, m), 3.85-3.75 (1H, m), 3.33-3.25
(1H, m), 3.20-3.03 (3H, m), 2.93-2.75 (7H, m), 2.70-2.65
(1H, m), 2.33 (3H, s), 2.26 (3H, s), 2.05-1.75 (5H, m),
1.75-1.63 (2H, m), 1.55-1.45 (1H, m), 1.38-1.25 (2H, m),
1.18-1.08 (1H, m); 13C NMR (100 MHz, d6-DMSO) δ
138.2, 137.1, 135.6, 132.0, 131.8, 129.8, 128.5, 125.9, 124.2,
123.7, 123.4, 122.8, 120.5, 120.1, 118.8, 112.0, 61.9, 54.0,
49.2, 44.5, 43.0, 33.7, 33.1, 27.1, 26.0, 21.2; mp 193.6 °C.
Chiracel AD-H 250 mm ꢀ 4.6 mm, 5 μm, isocratic 95:5
hexane/ethanol þ0.1% isobutylamine, flow 1.2 mL/min,
254 nm, Rt (min) 19 (undesired), 29 (desired). Enantiomeric
excess >99%. HRMS (ES) Calcd for C29H37N3O3 (MHþ)
476.2913. Found 476.2911.
FT-IR Hydrazoic acid Headspace Measurements. An FT-IR
analyzer (React-IR 4000) equipped with an online gas cell
(30 mL) was connected to a 250-mL jacketed resin kettle
(Figure 2). The gas phase during the reaction was examined
continuously by purging dry nitrogen gas through the headspace
at a rate of 16 mL/min. The tubing connecting the headspace of
the reactor to the IR gas-cell was heat-traced with a set tem-
perature at 56 or 68 °C, thus preventing the condensation of
HN3 (bp 37 °C) in the tubing. In addition, another FT-IR
analyzer equipped with an in situ Dicomp probe was inserted in
the reaction mixture to monitor the reaction progress in the
liquid phase. The typical absorption bands of HN3 in the vapour
phase are at 2126 and 2154 cm-1 and are the linear stretches of
HN3. In a calibration experiment, 0.05 mol NaN3 was driven out
of the aqueous phase by HCl over a period of time and purged
through the IR gas-cell by nitrogen gas at 12 mL/min.
’ AUTHOR INFORMATION
Corresponding Author
jeremy_scott@merck.com.
’ ACKNOWLEDGMENT
We thank Jos Brands, Simon Hamilton, Sophie Strickfuss,
John Edwards, Steven Oliver, Gavin Stewart, Debra Wallace,
Brian Bishop, Sarah Brewer, Andrew Gibson, Peter Sajonz,
Mirlina Biba, William Leonard, and Michael Moore for addi-
tional support.
(7R)-14-Cyclohexyl-7-[(2-dimethylamino)ethyl](methyla-
mino)-7,8-dihydro-6H-indolo[1,2-e][1,5]benzoxacine-11-
carboxylic Acid 4-Methylbenzenesulfonate (1). To a slurry
of methyl ester 20 (4.94 kg) in MeOH (24.7 L, 19.5 kg) was
added 2 M NaOH (aq) (10.1 L) and the reaction heated to
∼72 °C for 3 h. The reaction mixture was then cooled to 20 °C
and CH2Cl2 (52.3 kg) added, followed by 2 M HCl (aq) (10.1 L),
maintaining T < 25 °C. The phases were then separated, and
the organic layer was washed with water (8.9 kg). A slurry of
TsOH monohydrate (1.864 kg) in acetonitrile (10.3 kg) was
then added to the organic layer. The batch was then flushed
twice with acetonitrile (2 ꢀ 36.2kg) andthebatchconcentrated to
’ REFERENCES
(1) (a) Sarbah, S. A.; Younossi, Z. M. J. Clin. Gastroenterol. 2000, 30,
125–143. (b) Hayashi, N.; Takehara, T. J. Gastroenterol. 2006, 41, 17–27.
(c) Huang, Z.; Murray, M. G.; Secrist, J. A. Antiviral Res. 2006, 71,
351–362. (d) Gordon, C. P.; Keller, P. A. J. Med. Chem. 2005, 48,
1–20.
(2) Narjes, F.; Crescenzi, B.; Ferrara, M.; Habermann, J.; Colarrusso,
S.; Del Rosario, M. C.; Stansfield, I.; Mackay, A. C.; Conte, I.; Ercolani,
C.; Zaramella, S.; Palumbi, M. C.; Meuleman, P.; Leroux-Roels, G.;
Giuliano, C.; Fiore, F.; Di Marco, S.; Baiocco, P.; Koch, U.; Migliaccio,
G.; Altamura, S.; Laufer, R.; De Francesco, R.; Rowley, M. J. Med. Chem.
2011, 54, 289–301.
1122
dx.doi.org/10.1021/op200002u |Org. Process Res. Dev. 2011, 15, 1116–1123

Organic Process Research & Development
ARTICLE
(3) (a) Oka, T.; Ikegashira, K.; Hirashima, S.; Yamanaka, H.; Noji, S.;
Niwa, Y.; Matsumoto, Y.; Sato, T.; Ando, I.; Nomura, Y. PCT Int. Appl.
WO/2005/080399, 2005. (b) Hudyma, T. W.; Zheng, X.; He, F.; Ding,
M.; Bergstrom, C. P.; Hewawasam, P.; Martin, S. W.; Gentles, R. G. PCT
Int. Appl. WO/2006/020082, 2006. (c) Conte, I.; Ercolani, C.; Narjes,
F.; Pompei, M.; Rowley, M.; Stansfield, I. PCT Int. Appl. WO/2006/
046030, 2006. (d) Ikegashira, K.; Oka, T.; Hirashima, S.; Noji, Satoru, Y.;
Hiroshi, H.; Adachi, T.i; Tsuruha, J.; Doi, S.; Hase, Y.; Noguchi, T.;
Ando, I.; Ogura, N.; Ikeda, S.; Hashimoto, H. J. Med. Chem. 2006,
24, 6950–6953.(e) Hudyma, T. W.; Zheng, X.; He, F.; Ding, M.;
Bergstrom, C. P.; Hewawasam, P.; Martin, S. W.; Gentles, R. G. PCT
Int. Appl. WO/2007/092000, 2007. (f) Apito, E.; Habermann, J.; Narjes, F.;
Rico Ferreira, M. R.; Stansfield, I. PCT Int. Appl. WO/2007/129119, 2007.
(g) Stansfield, I.; Koch, U.; Habermann, J.; Narjes, F.. PCT Int. Appl. WO/
2008/075103, 2008. (h) Conte, I.; Haberman, J.; Mackay, A.; Narjes, F.;
Ricco Ferreira, M. R.; Stansfield, I. Br. UK Pat. Appl. GB 2451184.
(i) Conte, I.; Habermann, J.; Mackay, A.; Narjes, F.; Rico Ferreira, M. R.;
Stansfield, I. U.S. Pat. Appl. Publ. U.S. 2009 048239, 2009.
(4) Storage of aziridine 5 at -20 °C was necessary to avoid
degradation.
(5) Conditions for sCO2 preparative separation: Chiralpak OD-H,
150 mm ꢀ 4.6 mm, 15% methanol þ 25 mM isobutyl amine/CO2, 200
bar, 35 °C, 2.5 mL/min, 6 min run time. Rt (min) 3.9 (undesired), 4.3
(desired).
(6) (a) Mitsunobu, O. Synthesis 1981, 1–28. (b) Hughes, D. L. Org.
React. 1992, 42, 335–656. (c) Hughes, D. L. Org. Prep. 1996, 28, 127–164.
(d) Swamy, K. C. K.; Kumar, N. N. B.; Balaraman, E.; Kumar, K. V. P. P.
Chem. Rev. 2009, 109, 2551–2651. (e) Lal, B.; Pramanik, B. N.; Manhas,
M. S.; Bose, A. K. Tet. Lett. 1977, 23, 1977–1980.
(7) This purity upgrade is primarily to remove unreacted (S)-glycidol
tosylate 9 which is detrimental to the impurity profile of the subsequent
ring closure.
(8) Thompson, A. S.; Humphrey, G. R.; DeMarco, A. M.; Mathre,
D. J.; Grabowski, E. J. J. J. Org. Chem. 1993, 58, 5886–5888.
(9) Larrꢀe, C.; Donnadieu, B.; Caminade, A.-M.; Majoral, J.-P. Eur.
J. Inorg. Chem. 1999, 601–611.
(10) A one-pot Mitsunobu, Staudinger, and aza-Wittig procedure
was demonstrated to be viable but was not implemented due to time
constraints.
(11) (a) Wiss, J.; Fleury, C.; Onken, U. Org. Process Res. Dev. 2006,
10, 349–353. (b) Wiss, J.; Fleury, C.; Heuberger, C.; Onken, U. Org.
Process Res. Dev. 2007, 11, 1096–1103.(c) Gosselin, R. E.; Smith, R. P.
Hodge, H. C.; Braddock, J. E. Clinical Toxicology of Commercial Products;
.
Williams and Wilkings: Baltimore, 1984; pp II-114-II-115.
(12) Kaplan, M. A.; Granatek, A. P. U.S. Pat. Appl. Publ. U.S.
3840535, 1974.
(13) Giguere, P. A.; Rundle, R. E. J. Am. Chem. Soc. 1941, 63, 1135–
1137.
(14) (a) Robinson, D. Org. Process Res. Dev. 2009, 13, 391–396.
(b) Pierson, D. A.; Olsen, B. A.; Robbins, D. K.; DeVries, K. M.; Varie,
D. L. Org. Process Res. Dev. 2009, 13, 285–291. (c) Elder, D. P.; Teasdale,
A.; Lipczynski, A. M. J. Pharm. Biomed. Anal. 2008, 46, 1–8.
1123
dx.doi.org/10.1021/op200002u |Org. Process Res. Dev. 2011, 15, 1116–1123