PAT application in the expedited development of a three-step, one-stage synthesis of the dipeptide intermediate of HCV protease inhibitor faldaprevir

Article abstract of DOI:10.1021/op400285y

A concise scalable synthesis of a chiral dipeptide acid, key substructure of the HCV protease inhibitor faldaprevir, has been developed. A green process with an E-factor of 9.2 was achieved utilizing process analytical technology (PAT) to allow effective processing of multiple-steps in a one-stage operation. Mixed anhydride/oxazolone formation, peptide coupling, saponification, and then crystallization of the desired dipeptide acid were completed within 10 h. MultiMaxIR was used to detect the formation and consumption rates of key intermediates and to provide initial safety data which was subsequently confirmed by more comprehensive process safety testing. Further kinetic analysis was performed to determine the range of operability space to ensure conditions for a robust process.

Full text of DOI:10.1021/op400285y

Article
pubs.acs.org/OPRD
PAT Application in the Expedited Development of a Three-Step, One-
Stage Synthesis of the Dipeptide Intermediate of HCV Protease
Inhibitor Faldaprevir
Nizar Haddad,* Bo Qu, Heewon Lee, Jon Lorenz, Rich Varsolona, Suresh Kapadia, Max Sarvestani,
XuWu Feng, Carl A. Busacca, Dominique Hebrault,‡ Simon Rea,‡ Leen Schellekens,‡
and Chris H. Senanayake
Chemical Development, Boehringer Ingelheim Pharmaceuticals, Inc., 900 Ridgebury Rd., Ridgefield, Connecticut 06877, United
States
‡Mettler-Toledo AutoChem, Inc., 7075 Samuel Morse Drive, Columbia, Maryland 21046, United States
S
* Supporting Information
ABSTRACT: A concise scalable synthesis of a chiral dipeptide acid, key substructure of the HCV protease inhibitor faldaprevir,
has been developed. A green process with an E-factor of 9.2 was achieved utilizing process analytical technology (PAT) to allow
effective processing of multiple-steps in a one-stage operation. Mixed anhydride/oxazolone formation, peptide coupling,
saponification, and then crystallization of the desired dipeptide acid were completed within 10 h. MultiMaxIR was used to detect
the formation and consumption rates of key intermediates and to provide initial safety data which was subsequently confirmed by
more comprehensive process safety testing. Further kinetic analysis was performed to determine the range of operability space to
ensure conditions for a robust process.
resulted in a poorly reactive imidazolide intermediate which is
known to require activation by addition of catalysts such as 1-
hydroxybenzotriazole (HOBt) when applied to hindered
carboxylic acids and/or weakly nucleophilic amines.6 Coupling
based on carboxylic-sulfonyl-mixed anhydride intermediate
using TsCl/DMAP7 or TsCl/N-methylimidazole8 protocols
were considered as an alternative to provide a scalable process.
INTRODUCTION
■
Our interest in developing small-molecule inhibitors of HCV
NS3 protease was recently described in the discovery1 and
process development for the preparation of macrocycle
BILN2061.2 Subsequently, faldaprevir, a nonmacrocyclic
inhibitor of HCV protease, has been identified and found to
possess the key features of in vivo potency, safety, and
bioavailability needed for a new HCV therapy.3 A highly
convergent retrosynthetic route was devised as described in
Scheme 1 which was based on the assembly of three advanced
intermediates:4 dipeptide acid 1, thiazole-quinoline 2, and the
aminoester cyclopropane 3. The early stage application of PAT
in expediting the development of a scalable three-step, one-
stage synthesis of dipeptide acid 1 is the highlight of this study.
The Discovery route for the synthesis of dipeptide 1
consisted of a multistep sequence which involved EDC/
HOBT procedure for the peptide coupling of capped-L-tert-
leucine 4 with hydroxyproline methylester (Scheme 2) followed
by saponification. In addition to our primary goal of replacing
HOBT, which is classified as shock sensitive in Europe,5 further
process considerations had to be addressed in developing a
scalable, safe, and economical process. Ultimately, a one-stage
process which consists of using a single organic solvent with
inexpensive reagents was developed. This method allows for a
fast, yet safe, formation of the dipeptide acid without
racemization of the L-tert-leucine, and culminates in a robust
crystallization of 1. Development of this process, which could
serve as the primary means for the preparation and isolation of
key intermediate 1 became our goal.
RESULTS AND DISCUSSION
■
The bulkiness of the tert-butyl substituent in capped-L-tert-
leucine 4, accompanied by the carbamate protection of the
aminoacid was expected to limit the possibility of epimerization
during the peptide coupling under mixed anhydride conditions.
Therefore, we first examined the coupling via the mixed-
anhydrides of isobutylchloroformate9 and pivaloyl chloride. In
both cases, incomplete regioselectivity was observed9c with
more than 15% of the undesired products 6 and 7 formed,
which accounted in part for the moderate yield (70%) of the
desired product under these conditions (Scheme 3). ReactIR
monitoring indicated fast formation of the mixed anhydride
with pivaloyl chloride in THF or acetonitrile (1811 and 1745
cm−1).10 However, the slow reaction of the mixed anhydride
with hydroxyproline at 5 °C, required 16 h to reach 95%
conversion.
In an attempt to address the low reactivity and
regioselectivity of the pivaloyl mixed anhydride, sulfonyl
mixed anhydrides of methanesulfonyl chloride (MsCl) and p-
Special Issue: Process Analytical Technologies (PAT) 14
Received: October 11, 2013
While acid chlorides are the most common acylating agents,
the use of a mixed anhydride would be preferred for substrates
that are prone to epimerization. Carbonyl diimidazole (CDI)
© XXXX American Chemical Society
A
dx.doi.org/10.1021/op400285y | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Scheme 1. Retrosynthesis of faldaprevir
Scheme 2. Three-step, one-stage synthesis of dipeptide acid 1
Scheme 3. Unsatisfactory couplings using isobutylchloroformate or pivaloyl chloride conditions
Scheme 4. Preparation of dipeptide 5 under optimized conditions
toluenesulfonyl chloride (TsCl) became an attractive option for
faster reaction11a of a less electrophilic11b leaving group.
Screening of different organic bases and solvents identified N-
methylmorpholine (NMM) as a suitable base in acetonitrile,
resulting in a fast and selective conversion to the desired
dipeptide 5 in >90% isolated yield without detectable
isomerization12 (Scheme 4). Acetonitrile was selected due to
the rapid and clean reaction profile and its compatibility with
the subsequent saponification step.
With these results in hand, TsCl was selected for further
development to eliminate potential complications with MsCl
on scale-up via possible sulfene formation.13 For timely
development of a robust and safe process, the reaction profile
was followed by online IR data recording. The IR data in Figure
1 describe the reaction progress through the different steps. It
first indicates a shift of the carboxylic acid frequency (1744
cm−1) after addition of NMM to capped-L-tert-leucine 4,
followed by fast reaction with TsCl as evidenced by the
disappearance of the carboxylic acid absorption at 1711 cm−1
and increase in the intermediate frequencies at 1825 cm−1 and
It should be noted that these reactions were selective and
rapid in the absence of catalytic DMAP7or N-methylimidazole.8
B
dx.doi.org/10.1021/op400285y | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Figure 1. IR data from in situ monitoring throughout the sequence of preparing dipeptide 1; blue: capped Leu 4, green and purple: oxazolone 9,
black: dipeptide ester 5; red: dipeptide acid 1 + H2O.
Scheme 5. Possible effect of N-methylimidazole in activation of the tosyl-mixed anhydride intermediate; oxazolone 9 formed as
major intermediate in a similar rate, in presence or absence of NMI
1686 cm−1. The IR data is missing the typical carbamate
carbonyl absorption (1745 cm−1)10 that was observed with
pivaloyl chloride reaction and expected in intermediate 8
(Scheme 5). However, the IR data was found consistent with
the structure of cyclic intermediate 2-alkoxy-5(4H)-oxazolone
(9).10b,c A sample of 9 was isolated and fully characterized by
NMR, IR, and MS analyses. Addition of hydroxyproline HCl
salt to the solution of 9 at 0−5 °C resulted once again in a fast
reaction which provided the desired dipeptide ester 5 as
evidenced by the intensity decrease of the oxazolone
frequencies and formation of the amide signals at 1747, 1707,
and 1644 cm−1. The end point of this step was confirmed by
HPLC and NMR analysis which indicated <1% of remaining
starting material. Finally, the saponification step with LiOH was
followed by the decay rate of the methyl ester signal at 1747
cm−1. The desired dipeptide 1 was obtained in >90% yield with
less than 0.05 A% of the product derived from epimerization of
the tert-leucine fragment.
Unlike the reported slow esterification of carboxylic acids in
the absence of N-methylimidazole (NMI),8a our studies
indicate fast amide formation with hydroxyproline in the
absence of NMI (Scheme 5). While this observation benefits
the chemoselective preference for amide vs ester formation
upon reaction with hydroxyproline, further evaluation of the
addition of NMI to form the corresponding acyl-imidazolidi-
nium intermediate was tested. IR data indicated fast formation
of oxazolone 9 as the sole intermediate. Complete overlap in
the reaction rates was obtained in the presence and absence of
NMI.8c
While online IR monitoring has effectively provided valuable
kinetic data along with structural confirmation of the key
intermediates, the fast reaction rates emphasized the need for
early calorimetric evaluation of the process to ensure process
safety on multikilogram scale. The Tr and Tr−Tj data obtained
from the MultiMax system provided a preliminary indication on
the heat of reaction as described in Figure 2. In the MultiMax
system, the temperature profile of the reaction in step 1
C
dx.doi.org/10.1021/op400285y | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Figure 2. Heat flow data recorded by the MultiMax system throughout the experiment.
Scheme 6. Scale-up process
indicated a low exotherm (Tr − Tj < 6 °C) after addition of
solid TsCl to the reaction mixture at 0 °C. Similarly, the IR data
along with the observed Tr − Tj < 6 °C in step 2 after addition
of hydroxyproline-HCl solid in one portion indicated a fast, yet
safe, step. The saponification step was also completed in about
30 min following the addition of aqueous LiOH solution. At
this step the value of Tr − Tj =14 °C was attributed to sensible
heat from addition of a higher temperature (RT) solution of
the base. These results were subsequently confirmed by
conducting more comprehensive process safety testing
including differential scanning calorimetry (DSC), adiabatic
calorimetry in the Advanced Reactive Systems Screening Tool
(ARSST) and in reaction calorimetry (RC1e). RC1e experi-
ments indicated an adiabatic temperature rise of 61 °C, 38 °C,
and 17 °C for the three steps in the process, respectively.
With this information in hand, we obtained initial confidence
in the safety and efficiency of this three-step, one-stage process
for the formation of the dipeptide acid 1 and its isolation as
described in Scheme 6.
The following procedure was thus applied on multikilogram
scale: N-methylmorpholine was added to the precooled
solution of capped-tert-leucine 4 in MeCN at 0−5 °C, followed
by addition of TsCl solution while keeping the reaction
temperature below 5 °C (20 min). The mixture was then
stirred for 1 h at 5 °C. A suspension of L-4-hydroxyproline
methylester HCl in MeCN was then added to the reaction
mixture to result in complete conversion to 5 after 1.5 h at 0
°C. Aqueous LiOH solution (3.2 N) was added within 40 min.
Stirring was continued at 25 °C for one hour to reach complete
conversion to 1. MeCN was then distilled under vacuum at 35
°C until GC analysis indicated less than 1% w/w of MeCN in
the reactor. The pH was first adjusted to about pH = 7 by
addition of 6 N HCl followed by heating the mixture to 60 °C.
This operation was then followed by a second addition of 6 N
HCl to reach a pH of 3.5−4. The batch was then seeded, and a
final pH adjustment to about pH = 2 was made with 6 N HCl.
Cooling to 22 °C and filtration then provided the desired
product in 95% yield and purity of >98 wt %.
With the project advancing in development, our efforts
turned to firming up the robustness of the coupling process for
our manufacturing facility. The following topics were
addressed: sensitivity of reaction rate to reactants concentration
(e.g., hydroxyproline concentration) and impact of temperature
on the coupling reaction rate. Blackmond’s approach to kinetic
analysis, Reaction Progress Kinetic Analysis (RPKA),14 and iC
Kinetics, was applied by running a set of experiments geared at
evaluating process robustness. The reaction was investigated
using EasyMax, a synthetic workstation which provides heat
flow information, and ReactIR, an ATR-FTIR instrument that
provides mid-IR spectroscopy information in real time.
We first evaluated the consistency between the information
provided by EasyMax heat flow and changes in mid-IR signal.
As shown in Figure 3, good correlation was obtained between
the two plots following the formation of amide 5.
D
dx.doi.org/10.1021/op400285y | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Figure 5. Empirical model for oxazolone 9 reaction with hydroxypro-
line, built using iC Kinetics
Figure 3. Correlation between conversion of oxazolone 9 to amide 5
vs time, data obtained by heat flow and IR.
further by using the large number of data points made available
by real time reaction monitoring while applying a more
complex fit equation software.
Using the Different Excess strategy (DE), as described by
Blackmond,14 we investigated the coupling power law rate
equation (Figure 4).
We tested the developed model (rate constant and reaction
orders) using the Different Excess strategy on a central point
and obtained good levels of prediction for time-course
concentration change. Furthermore, the calculated rate
equation that best fits the model remains the same across the
conversion range, confirming a stable reaction and a consistent
mechanism throughout the step. This observation increased the
level of confidence in the robustness of the process. The
reaction was then tested following the consumption of
oxazolone 9 at different temperatures across the −10 °C to
+30 °C range. Faster conversion was obtained as the
temperature increased with good fit (R2 = 0.998) of the data
points into an Arrhenius model.16 Interestingly, when the
reaction sequence was tested at 30 °C, 1.3% epimerization on
the tert-butyl substituent was detected by HPLC analysis.
Considering that the reaction investigated is unlikely an
elementary step, the extension of the model outside the −10
to 30 °C temperature range might lead to erroneous
conclusions.
In summary, a concise, safe, and scalable multistep synthesis
of chiral dipeptide acid 1 was effectively achieved by early-stage
utilization of PAT to allow processing of three steps in one
operation which is completed within 10 h. MultiMaxIR was
used to identify the formation and consumption rates of
oxazolone 9 as a key intermediate and temperature profile of
the reaction which indicated safe scale-up as subsequently
confirmed by reaction safety calorimetry. Oxazolone 9 was
effectively formed in the absence of N-methylimidazole, and
provided dipeptide 1 without detectable epimerization when
the reaction temperature was kept below 10 °C.
Figure 4. Oxazolone 9 concentration vs time in Different Excess
experiments.
The data shown in Figure 4 was then used to build an
empirical model using the iC Kinetics algorithm,15 and
determine reaction time to 90% conversion for 400 simulated
experimental conditions (Figure 5).
A good linear regression correlation value (0.99) was
obtained, which indicates that the amide formation reaction
lends itself well to preliminary kinetic investigation and
modeling. The model that best fits the experimental data
gave partial orders in oxazolone 9 and hydroxyproline (0.78
and 0.69 respectively, rate constant 0.0115 M−1 s−1) which
likely indicates several elementary steps, or the presence of
chemical equilibrium, involved in the formation of the
oxazolone intermediate 9. It should be noted that the initial
rate of a fast reaction like this may not provide a full picture for
reaction rate compared to a full data set of up to 90%
conversion as calculated by the iC Kinetics. Reaction Progress
Kinetic Analysis (RPKA) and iC Kinetics allow for going a step
EXPERIMENTAL SECTION
■
General. Capped-tert-leucine 4 was prepared according to
our reported procedure.4 Data collection of adiabatic
calorimetry was performed using ARSST manufactured by
Fauske and Associates. Kinetic experiments were performed
using METTLER TOLEDO EasyMax synthesis workstation
equipped with 100 mL vessels in conjunction with iControl
software. The accuracy of temperature control of this system is
0.01 °C. A METTLER TOLEDO ReactIR 15, based on
attenuated total reflectance-Fourier transform IR spectroscopy
E
dx.doi.org/10.1021/op400285y | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
(ATR-FTIR) and equipped with diamond-composite insertion
fiber probe was used for real-time in situ process monitoring.
HPLC analysis was carried out on an Agilent 1100 system
equipped with a Waters YMC ODS-AQ, S-5, 120A, 250 mm ×
4.6 mm column and detected at 210 nm. Mobile phase: A:
mixture of 0.1% H3PO4 and 0.05% HClO4 in HPLC grade H2O
with B: HPLC grade MeCN, was run in a gradient of 90%A/5%
A, flow rate = 1.0 mL/min. HPLC retention times: 4 (15.73
min); 5 (15.6 min); (1S,3R,5S)-1 (11.72 min), (1S,3R,5R)-
diastereomer 1 (11.26 min); (1R,3R,5S)-diastereomer 1 (11.45
min); (1S,3S,5S)-diastereomer 1 (12.28 min); All NMR spectra
were collected on Bruker spectrometers equipped with a 5 mm
BBI probe (1H,13C). 1H and 13C chemical shifts were calibrated
vs the deuterated solvent used.
added in over 20−30 s; then the resulted mixture was stirred
for 1 h at 10 °C. L-4-Hydroxyproline methylester HCl (2.04 g,
5.71 mmol) was added in one portion. The reaction was
followed by recording IR (256 scans/min, 1 spectrum/min
acquisition) for 1.5 h.
The experiment was repeated while varying the L-4-
hydroxyproline methylester-HCl amounts as follows: Experi-
ment 1: 2.04 g, 1.1 equiv; Experiment 2: 2.78 g, 1.5 equiv;
Experiment 3: 3.71 g, 2.0 equiv.
(S)-4-(tert-Butyl)-2-cylopentyloxy-5(4H)-oxazolone (9).
The above tosyl chloride procedure was repeated on 1.2 g of
capped-tert-leucine 4. A 2 mL sample was removed from the
reaction mixture after 1 h from the addition of tosyl chloride.
The sample was passed through a short column containing 10 g
silica gel and eluted with 30 mL MeCN. The volatiles were
removed under reduced pressure (30 mbar) to give 120 mg of
Methyl (S)-2-(Cyclopentyloxycarbonylamino)-3,3-di-
methylbutanoate (5). Pivaloyl Chloride Procedure. To a
50 mL vessel, equipped with nitrogen inlet, mechanical stirrer,
IR probe recording every 30 s, and thermocouple, was added
NMM (5.5 mL, 49.0 mmol) to a cold (5 °C) solution of
capped-tert-leucine 4 (2.38 g, 9.8 mmol) in THF or MeCN (10
mL) while stirring at 300 rpm. A solution of pivaloyl chloride
(1.3 mL, 10.76 mmol) in MeCN (1.5 mL) was added in over 2
min. The reaction mixture stirred for an additional 1 h at 5 °C.
L-4-Hydroxyproline methylester HCl (1.95 g, 1.1 equiv) was
added in one portion. The reaction stirring was continued at 5
°C for 16 h, at which point IR indicated the end of the reaction.
Water (10 mL) was added, the organic solvent was removed
under reduced pressure, and then the product was extracted
with EtOAc (30 mL), washed with water (5 mL), dried over
MgSO4, and then concentrated under reduced pressure (30
mmHg). Purification by silica gel column, using 25% EtOAc in
hexanes (Rf = 0.3), provided 2.5 g of the product in 70% yield.
Tosyl Chloride Procedure. To a 50 mL reaction vessel,
equipped with a nitrogen inlet, mechanical stirrer, IR probe
recording every 30 s, and thermocouple, was added NMM (5.5
mL, 49.0 mmol) to a cold (5 °C) solution of capped-tert-
leucine 4 (2.38 g, 9.8 mmol) in THF or MeCN (10 mL) while
stirring at 300 rpm. Tosyl chloride (2.05 g, 10.76 mmol) was
added in one portion. The resultant mixture was stirred for 1 h
at 5 °C until IR data indicated complete consumption of 4. L-4-
Hydroxyproline methylester HCl (1.95 g, 1.1 equiv) was then
added in one portion. The reaction stirring was continued at 5
°C and was followed by recording IR for 1.5 h. Water (10 mL)
was then added, the organic solvent was removed under
reduced pressure, and the product was then extracted with
EtOAc, washed with water (5 mL), dried, and then
concentrated under reduced pressure. Purification by SGC
(25% EtOAc in hexanes, Rf = 0.3) provided 3.2 g (89%) of the
1
oxazolone 9 in good purity and ∼80% yield: H NMR (400
MHz, CDCl3, 5.3 (m, 1H), 4.0 (s, 1H), 2−1.5 (m, 8H), 1.05 (s,
9H). 13C NMR (100 MHz, CDCl3): 174.8, 156.9, 83.4, 73.7,
31.9, 24.8, 22.9. Exact Mass: Calculated, [C12H19NO3 + H+]
226.1438; Experimental, 226.1428.
(2S,4R)-1-((S)-2-(Cyclopentyloxycarbonylamino)-3,3-
dimethylbutanoyl)-4-hydroxy-pyrrolidine-2-carboxylic
Acid (1). Pilot Plant Procedure. A 100-gal glass reactor was
charged with capped-tert-leucine 4 (15.0 kg, 61.65 mol, 1
equiv) and MeCN (30.0 L). The mixture was cooled to 0 °C
followed by addition of NMM (31.4 kg, 308.3 mol, 5.0 equiv)
to give a colorless solution. The internal temperature was then
reduced to −20 °C, and a solution of tosyl chloride (12.15 kg,
63.75 mol, 1.03 equiv) in MeCN (30.0 L) was added over 80−
100 min. The internal temperature was maintained below 5 °C.
The resulting mixture was stirred for an additional 1 h from the
end of the addition at 0 °C, then cooled to below −2 °C. A
slurry of L-4-hydroxyproline methylester HCl (12.25 kg, 67.50
mol, 1.1 equiv) in MeCN (30.0 L) was then added over 25−30
min, maintaining the temperature below 10 °C during the
addition. The mixture was held for 1.5 h at 0−15 °C, and
HPLC analysis showed that less than 0.22% of 9 remained
(checked as the benzylamide derivative). LiOH (3.2 M, 90.0 L,
12.10 kg in 85.5 L H2O; 288.0 mol,4.67 equiv) was then added
to this mixture of ester 5 over 40 min, while maintaining the
internal temperature below 15 °C. The mixture was then
warmed to 23 °C and stirred for an additional 1 h; HPLC
analysis showed complete consumption of the intermediate
ester. The MeCN was removed by distillation under reduced
pressure (30−45 °C/30 mbar), then HCl (6 N, 25.5 L, 28.4 kg,
153.0 mol) was added over 25 min while the internal
temperature was maintained below 35 °C. The final pH value
was 7−8. Residual MeCN was then removed by distillation (45
°C/30 mbar) to below 1% w/w. The mixture was then heated
to 60 °C, and HCl (6 N, 28.75 kg, 154 mol) was slowly added
to reach pH 3.6; then the batch was seeded with a slurry of 75 g
dipeptide acid 1 seeds in 1 L H2O. The agitation was continued
at 60 °C for 30 min to establish a seed bed. An additional
charge of HCl (6 N, 11.05 kg) was then added over 1 h to
reach pH 0.99 at 60 °C (pH 1.3 at 23 °C). After stirring for an
additional 1 h at 60 °C, the mixture was cooled to 23 °C over 1
h and then aged at that temperature for an additional 1 h. The
slurry was then filtered (about 50 min), and the filter cake was
washed with HCl (0.1 N, 2 × 25.0 L), then with H2O (25.0 L).
The solid was dried at 50 °C/30 mbar with an N2 bleed to give
20.65 kg of dipeptide acid 1 as a white crystalline solid in 91%
overall yield. HPLC purity: 99.91%; wt % purity: 101.14%;
1
desired product. H NMR (400 MHz, CDCl3, major rotamer
reported): 5.44 (d, J = 9.4 Hz, 1H), 5.02 (br s, 1H), 4.65 (t, J =
8.5 Hz, 1H), 4.50 (s, 1H), 4.25 (d, J = 9.5 Hz, 1H), 3.97 (d, J =
11.1 Hz, 1H), 3.73 (m, 1H), 3.71 (s, 3H), 2.45 (m, 1H), 2.0
(m, 1H), 1.9−1.5 (m, 8H), 1.02 (s, 9H). 13C NMR (100 MHz,
CDCl3, both rotamers): 172.6, 170.9, 156.8, 78.1, 70.1, 59.0,
57.8, 56.5, 52.2, 37.5, 35.7, 32.9, 32.6, 26.2, 23.7. Exact Mass:
Calculated, [C18H30N2O6 + H+] 371.2165; Experimental,
371.2177.
Kinetics Experiments. To a 200 mL reaction vessel,
equipped with nitrogen inlet, mechanical stirrer, IR probe
and thermocouple, was added NMM (2.8 mL, 25.5 mmol) to a
cold (0 °C) solution of capped-tert-leucine 4 (1.2 g, 5.19
mmol) in MeCN (20 mL) while stirring at 250 rpm. A solution
of tosyl chloride (1.0 g, 5.2 mmol) in MeCN (2.5 mL) was
F
dx.doi.org/10.1021/op400285y | Org. Process Res. Dev. XXXX, XXX, XXX−XXX

Organic Process Research & Development
Article
Grand-Maitre, C.; Grenier, L.; Guindon, Y.; Guse, I.; Plamondon, L.;
Soucy, F.; Valois, S.; Wernic, D.; Yoakim, C. J. Org. Chem. 1997, 62,
3440−3448. (b) Herrmann, M.; Ehrler, J.; Kayser, H.; Rindlisbacher,
A.; Hofle, G. Eur. J. Org. Chem. 1999, 3381−3392. (c) Valeur, E.;
Bradley, M. Chem. Soc. Rev. 2009, 38, 606−631.
(10) (a) Wilder, R.; Mobashery, S. J. Org. Chem. 1992, 57, 2755−
2756. (b) Coste, J.; Frerot, E.; Jouin, P. J. Org. Chem. 1994, 59, 2437−
2446. (c) Benoiton, N. L.; Chen, F. M. F. Can. J. Chem. 1981, 59,
384−389.
(11) (a) Relative rates are consistent with the pKa values in
acetonitrile for sulfonic acids (TsOH = 8.5) and carboxylic acids
(acetic acid = 23.5); pKa values from: Eckert, F.; Leito, I.; Kaljurand, I.;
Kutt, A.; Klamt, A.; Diedenhofeni, M. J. Comput. Chem. 2008, 30,
799−810. (b) Um, I.; Chun, S. M.; Bae, S. K. Bull. Korean Chem. Soc.
2005, 26, 457−460.
(12) HPLC method that separates dipeptide acid 1 from its
corresponding diasteromers (1S,3R,5R), (1R,3R,5S), and (1S,3S,5S)
was developed. Please refer to Supporting Information for details.
(13) (a) Opitz, G.; Ehlis, T.; Rieth, K. Chem. Ber. 1990, 123, 1989−
1998. (b) King, J. F.; Lam, J. Y. L.; Skonieczny, S. J. Am. Chem. Soc.
1992, 114, 1743−1749.
(14) Blackmond, D. G. Angew. Chem., Int. Ed. 2005, 44, 4302−4320.
(15) iC Kinetics is a graphical method for analyzing how different
experimental conditions influence the speed of a chemical reaction. It
imports concentration data from in situ spectroscopic measurements
or conversion values from reaction calorimetric for kinetic analysis.
The kinetic model created by the software can be used to simulate the
effect of concentration and temperature parameters on the perform-
ance of the reaction. An extensive overview of the Reaction Progress
Kinetic Analysis (RPKA) approach, adopted in iC Kinetics, is
described in reference 14.
H2O = 0.093% (KF); thermogravimetric analysis loss = 0.854%.
Mp 117−124 °C; H NMR (400 MHz,CDCl3, major rotamer
1
reported): 5.51 (d, J = 9.2 Hz, 1H), 5.02 (br s, 1H), 4.70 (t, J =
8.8 Hz,1H), 4.52 (s, 1H), 4.31 (d, J = 9.2 Hz, 1H), 4.01 (d, J =
11.2 Hz, 1H), 3.68 (dd, J = 3.6, 11.6 Hz, 1H), 3.06 (s, 1H, 0.33
MTBE), 2.08 (m,1H), 1.87−1.48 (m, 9H), 1.09 (s, 3H, 0.33
MTBE), 0.92 (s, 9H). 13C NMR (100 MHz, CDCl3, both
rotamers): 173.5 (s), 172.5 (s), 157.1 (s), 78.5 (d), 69.8 (d),
59.1 (d), 58.2 (d), 56.8 (t), 49.6 (MTBE), 36.7 (t), 35.5 (s),
32.8 (t), 32.7 (t), 26.3 (d), 23.7 ppm (t). Elemental Analysis:
Calculated, C: 57.29; H: 7.92; N: 7.86; Experimental, C: 57.16;
H: 7.88; N: 7.86.
ASSOCIATED CONTENT
■
S
* Supporting Information
Experimental procedures, IR-spectra, selected 1H NMR spectra
1
for reactions study, H and 13C NMR spectra for all new
compounds, full structure determination of oxazolone 9 and
calorimetric data. This material is available free of charge via the
Internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*E-mail: nizar.haddad@boehringer-ingelheim.com
■
Notes
The authors declare no competing financial interest.
REFERENCES
■
(16) Please refer to Supporting Information for additional details.
(1) (a) Llinas-Brunet, M.; Bailey, M. D.; Bolger, G.; Brochu, C.;
Faucher, A. M.; Ferland, J. M. J. Med. Chem. 2004, 47, 1605−1608.
(b) Lamarre, D.; Anderson, P. C.; Bailey, M.; Beaulieu, P.; Bolger, G.;
Bonneau, P. Nature 2003, 426, 186−189. (c) LaPlante, S. R.; Llinas-
Brunet, M. Curr. Med. Chem.: Anti-Infective Agents 2005, 4, 111−132.
(d) Tsantrizos, Y. S. Bioploymers 2004, 76, 309−323.
(2) Yee, N. K.; Farina, V.; Houpis, I. N.; Haddad, N.; Frutos, R. P.;
Gallou, F.; Wang, X.-J.; Wei, X.; Simpson, R. D.; Feng, X.; Fuchs, V.;
Xu, Y.; Tan, J.; Zhang, L.; Xu, J.; Smith-Keenan, L. L.; Vitous, J.;
Ridges, M. D.; Spinelli, E. M.; Johnson, M.; Donsbach, K.; Nicola, T.;
Brenner, M.; Winter, E.; Kreye, P.; Samstag, W. J. Org. Chem. 2006, 71
(19), 7133−7145.
(3) Highleyman, L. Experimental HCV Protease Inhibitor BI 201335
Shows Robust Antiviral Activity in Genotype I Non-Responders. In
Proceedings of the 45th Annual Meeting of the European Association
for Study of the Liver (EASL), Vienna, Austria, April 14−18, 2010.
(4) Busacca, C. A.; Wei, X.; Haddad, N.; Kapadia, S.; Lorenz, J. C.;
Saha, A. K.; Varsolona, R. J.; Berkenbusch, T.; Campbell, S. C.; Farina,
V. Asian J. Org. Chem. 2012, 1, 80−89.
(5) Weisenburger, G. A.; Vogt, P. F. Org. Process Res. Dev. 2006, 10
(6), 1246−1250.
(6) (a) Dunn, P. J.; Hoffmann, W.; Kang, Y.; Mitchell, J. C.;
Snowden, M. J. Org. Process Res. Dev. 2005, 9, 956. (b) Carey, J. S.;
Laffan, D.; Thomson, C.; Williams, M. T. Org. Biomol. Chem. 2006, 4,
2337−2347.
(7) (a) Wakasugi, K.; Nakamura, A.; Tanabe, Y. Tetrahedron Lett.
2001, 42, 7427−7430. (b) Wakasugi, K.; Nakamura, A.; Iida, A.;
Nishii, Y.; Nakatani, N.; Fukushima, S.; Tanabe, Y. Tetrahedron 2003,
59, 5337−5345. (c) Hamada, T.; Horikoshi, H.; Ishibashi, K.; Kojima,
K.; Kurata; H. U.S. Patent 5,536,714, 1996.
(8) (a) Wakasugi, K.; Iida, A.; Misaki, T.; Nishii, Y.; Tanabe, Y. Adv.
Synth. Catal. 2003, 345, 1209−1214. (b) Angelaud, R.; Davies, I.;
Maguire, C.; Lau, S.; O’Shea, P.; Shultz, S.; Patent WO/2010/11584
A2, 2010. (c) Please refer to Supporting Information for a comparison
graph of reaction rates in the presence and absence of N-
methylimidazole.
(9) (a) Beaulieu, P. L.; Lavallee, P.; Abraham, A.; Anderson, P. C.;
Boucher, C.; Bousquet, Y.; Duceppe, J. S.; Gillard, J.; Gorys, V.;
G
dx.doi.org/10.1021/op400285y | Org. Process Res. Dev. XXXX, XXX, XXX−XXX