Development of an asymmetric hydrogenation route to (S)- N -Boc-2,6-dimethyltyrosine

Article abstract of DOI:10.1021/op200065p

An improved, simpler and potentially more economical route to (S)-N-Boc-2,6-dimethyltyrosine 1, based on a previously published route, is presented. Key modifications were to prepare the dehydroaminoacid hydrogenation substrate 6 in a one-pot process directly from serine methyl ester and 4-iodo-3,5-dimethylphenyl acetate 4 and to identify a significantly more active asymmetric hydrogenation catalyst that allowed a 5-fold reduction in catalyst loading.

Full text of DOI:10.1021/op200065p

ARTICLE
pubs.acs.org/OPRD
Development of an Asymmetric Hydrogenation Route to
(S)-N-Boc-2,6-dimethyltyrosine
Cꢀeline F. B. Praquin,* Pieter D. de Koning, Philip J. Peach, Roger M. Howard, and Sarah L. Spencer
Pfizer Worldwide Research and Development, Ramsgate Road, Sandwich, Kent CT13 9NJ, United Kingdom
S Supporting Information
b
ABSTRACT: An improved, simpler and potentially more economical route to (S)-N-Boc-2,6-dimethyltyrosine 1, based on a
previously published route, is presented. Key modifications were to prepare the dehydroaminoacid hydrogenation substrate 6 in a
one-pot process directly from serine methyl ester and 4-iodo-3,5-dimethylphenyl acetate 4 and to identify a significantly more active
asymmetric hydrogenation catalyst that allowed a 5-fold reduction in catalyst loading.
Scheme 1. Published route to (S)-2,6-dimethyl-tyrosine 81a,b
’ INTRODUCTION
During a recent development program we were required to
prepare a significant quantity of (S)-N-Boc-2,6-dimethyltyrosine
1, and, in order to meet the project timelines, rapid delivery of
material was critical. Several routes to this amino acid have been
published;1 however, most have only been run on moderate scale.
We selected the asymmetric hydrogenation route shown in
Scheme 1 for further investigation as it had been previously
demonstrated on kilogram scale.1a,b In this process, the key step is
an enantioselective hydrogenation of methyl (Z)-2-acetamido-
3-(4-acetoxy-2,6-dimethylphenyl)-prop-2-enoate 6 with the
chiral catalyst, [Rh(1,5-COD)(R,R-DIPAMP)]BF4 under rela-
tively forcing conditions (60 °C, 60 psi) and high catalyst loading
(1 mol %). The hydrogenation substrate was prepared via Heck
coupling between 4-iodo-3,5-dimethylphenyl acetate 4 and
methyl 2-acetamidoacrylate 5. Iodide 4 was synthesized in two
steps from readily available 3,5-dimethylphenol 2,2 and acrylate 5
is commercially available.
’ RESULTS AND DISCUSSION
Whilst this route had been demonstrated on kilogram scale,
we still had some concerns as to whether it would be capable of
delivering sufficient material within the required time frame.
product into ethyl acetate following the acid quench. Further washing
with dilute acid removed any residual pyridine, and the crude product
Specifically, were we to use the current catalyst loading
(1 mol %), the cost of the catalyst would be prohibitive. It was
was concentrated to a thick oil that was then used directly in the Heck
therefore imperative for us to identify conditions that would
coupling.
allow for a significant reduction in catalyst usage. In addition,
Since acrylate 5 was only commercially available in small
while acrylate 5 is commercially available, we could only source
quantities, we made the decision to prepare this, since a review
small quantities (the largest pack size was 5 g) in the time
of the literature indicated that it should be readily prepared from
available and we were then faced with having to make significant
serine methyl ester.3 The key step in many of the published
quantities of this starting material internally.
Synthesis of Dehydroaminoacid 6. The published procedure1a,b
processes was thermal elimination of acetic acid from N,O-
diacetyl-serine methyl ester 9 in the presence of a base. However,
for the preparation of 4 from 3,5-dimethylphenol was repeated.
upon attempting to reproduce these conditions, we found
Unfortunately, however, in our hands the product failed to crystallize
upon quenching the reaction with 0.5 N HCl, as published. Instead,
an oily layer separated from the mixture that, after seeding, converted
to a crystalline mass. However, the solid form was poorly controlled,
Special Issue: Asymmetric Synthesis on Large Scale 2011
and residual solvent and impurities were entrained. In order to
overcome this, we modified the published workup and extracted the
Received: March 15, 2011
Published: May 26, 2011
r
2011 American Chemical Society
1124
dx.doi.org/10.1021/op200065p Org. Process Res. Dev. 2011, 15, 1124–1129
|

Organic Process Research & Development
ARTICLE
Scheme 2. Telescoped synthesis of dehydroamino acid 6 from serine methyl ester
isolation of the desired compound 5 to be rather challenging. In
particular the product was prone to polymerization under the
reaction conditions and was also rather water-soluble, resulting in
significant yield losses during the workup.
in-house was quickly conducted,5 and from this we identified
[Rh-(1,5-COD)(S,S-Et-FerroTANE)]BF4 as a potential alter-
6
native. Under screening conditions the reaction employing this
catalyst proceeded to completion (unlike many of the other
catalysts screened) and gave the product in >90% ee. In addition,
the S/C could be lowered to 500:1 with reaction completion still
observed within an acceptable time and without any reduction of
the ee. Longer term, the cost of using of this proprietary catalyst
(at lower catalyst levels) instead of the more readily available Rh-
DIPAMP would need to be carefully evaluated; however, at this
stage it was suitable for our purposes.
Given the difficulties encountered with isolation of acrylate 5
and our desire to reduce the number of processing steps, we
considered the possibility of telescoping the acrylate formation
and Heck coupling steps. Since the Heck coupling is done under
basic conditions using triethylamine in hot acetonitrile, it seemed
plausible that elimination of acetic acid from N,O-diacetyl serine
methyl ester 9 would occur under these conditions to generate
the desired acrylate 5 in situ, thus avoiding the need to prepare
and isolate this material in a separate step. In addition, since
the isolation of N,O-diacetyl serine methyl ester 9 is also
challenging, with significant degradation occurring during water
washes to remove triethylammonium salts (as well as product
losses to the aqueous phase), we decided to extend the reaction
telescope to incorporate this step, as shown in Scheme 2.
To our delight, this telescoped one-pot, three-stage process
worked extremely well. Readily available serine methyl ester
hydrochloride was treated with acetic anhydride and triethyla-
mine in acetonitrile at ambient temperature, generating a solution
of diacetate 9. Next, iodide 4, palladium(II) acetate, tri-o-tolylpho-
sphine, and further triethylamine were added; the mixture was
heated to reflux, whereupon the desired tandem eliminationÀ
Heck coupling sequence occurred smoothly. The process af-
forded the desired dehydroaminoacid 6 in 69% yield after an
extended workup required to remove potential catalyst poisons
(see Experimental Section for more details). At this stage, given
the time constraints and the need to deliver material quickly, we
simply used the previously published reaction conditions for the
Heck coupling (2 mol % Pd, 5 mol % phosphine) without
optimization. We hope that further development might increase
our yield to match that of the published process with acrylate 5
(69% vs 85%), as well as potentially delivering a reduction in
catalyst loading. While we have not been able to test the generality
of this telescoped process, it only requires cheap and readily
available starting materials, and we would expect it to provide a
practical route to a wide range of aromatic dehydroamino acids.
Enantioselective Hydrogenation of 6. The published pro-
cedure for the asymmetric hydrogenation of 6 using [Rh-(1,5-
COD)(R,R-DIPAMP)]BF4 was run at a substrate to catalyst
ratio (S/C) of 100:1 (1 mol %).1a,b Given the catalyst cost,4
conducting the hydrogenation at this catalyst loading was
deemed prohibitively expensive. Furthermore, our initial studies
with this catalyst indicated that we would not be able to
significantly reduce the loading, as the reaction rate and ee
dropped precipitously as the catalyst level was reduced. A small
screen of catalysts that were readily available in sufficient quantity
During our initial development work, we were plagued with
inconsistent conversion under nominally identical conditions.
After some investigation we discovered that this process is
particularly oxygen-sensitive (in comparison to the hydrogena-
tion of less hindered dehydroamino acids), and careful applica-
tion of the thorough oxygen-removal protocol described in the
published process1a was essential to ensure reproducibility.
Our initial studies were performed in a hydrogenation vessel,
employing disposable glass inserts. However, as we were required
to use a multipurpose stainless steel reactor for our large-scale
processing, we conducted a series of experiments in an inter-
mediate-scale reactor to test the robustness of the process and the
effectiveness of our cleaning protocol. On the basis of previous
experience, the stainless steel reactor was fully dismantled and
cleaned with acetic acid, followed by multiple water and ethyl
acetate boil-outs prior to use. A series of experiments were carried
out which demonstrated that robust and reproducible conversion
and ee could be achieved at S/C 500:1. When lower catalyst
levels were examined, a slight erosion of ee was noted, 88À90%
at S/C 750:1. At S/C 1000:1 this became significant (68À76%),
and quite variable. Therefore, we chose to conduct our scale-up
experiments at S/C 500:1 to ensure acceptable quality was
achieved. Further optimization of this step should hopefully
allow for reduction in the catalyst loading.
Using these conditions, multiple scale-up runs of the hydro-
genation step (from 130 g to a maximum of 800 g) were carried
out without incident. In addition, we were also able to reduce the
hydrogen pressure from 10 to 5 bar.
Preparation of (S)-N-Boc-2,6-dimethyltyrosine 1. The pub-
lished procedure for deprotection of N-acetyl amino acid 7 using
refluxing concentrated (12 M) HCl at relatively high concentra-
tion (4.3 mL/g) was deemed unsuitable for scale-up in our
facility. At this high HCl concentration a significant amount of
hydrogen chloride gas is evolved at reflux, requiring scrubbing. In
addition, upon cooling the product 8 (hydrochloride salt)
crystallizes from the reaction mixture as an unstirrable solid
mass. To minimize HCl off-gassing, less concentrated (6 M) HCl
was used, and the reaction temperature was held at just below the
1125
dx.doi.org/10.1021/op200065p |Org. Process Res. Dev. 2011, 15, 1124–1129

Organic Process Research & Development
ARTICLE
Scheme 3. Optimized route to (S)-N-Boc-2,6-dimethyltyrosine 1 (Method B in the Experimental Section)
boiling point (90À95 °C). Neither of these changes had any
impact on the reaction time. Furthermore, in order to overcome
the problem of uncontrolled precipitation of the hydrochloride
salt upon cooling, the reaction mixture was cooled to 40 °C and
then sodium hydroxide was added to adjust the pH to 4. At this
pH, 2,6-dimethyltyrosine 8 precipitates, which is much easier to
handle than its hydrochloride salt. For process simplicity, the pH
was then adjusted to pH 9À10, and Boc anhydride was added. As
the Boc protection proceeded, additional sodium hydroxide was
added to keep the pH above 8 until the reaction was complete.
Upon completion the pH of the reaction mixture was lowered
to 4, and the desired product (S)-N-Boc-2,6-dimethyltyrosine 1
was extracted into ethyl acetate. This solution was then concen-
trated to low volume, and the residue was slurried in a mixture of
ethyl acetate and heptane to give pure product 1 (Method A in
the Experimental Section).
To further improve the synthesis and to avoid handling
refluxing hydrochloric acid and the laborious pH adjustments
of the reaction mixture in the workup, we decided to investigate
the method described by Burk7 for the direct conversion of N-
acetyl amino ester 7 to the N-Boc amino acid 1 (Scheme 3).
Treatment of N-acetyl amino ester 7 with Boc anhydride in
presence of a catalytic amount of DMAP gave the expected
N(Ac)Boc amino ester, and, upon treatment with 5 M NaOH,
chemoselective hydrolysis of the acetamide occurred in addition
to hydrolysis of the methyl ester and phenolic acetate. The
sequence afforded N-Boc amino acid 1 in an excellent 91% yield
after isolation and purification (Method B in Experiment
Section).
Since our asymmetric hydrogenation process provided ma-
terial of acceptable optical purity (the internal target was
>90% ee), no work was done to investigate the possibility of
upgrading the chiral purity (for example, via crystallization of a
salt). It was noted, however, that no enhancement of chiral purity
was observed during the crystallization of either N-acetyl amino
ester 7 or N-Boc amino acid 1.
In parallel to the asymmetric hydrogenation route described in
Scheme 3, we also investigated whether the (S)-amino acid 8
could be accessed via an enzyme catalyzed hydrolytic resolution
of racemic N-acetyl-2,6-dimethyltyrosine (see Supporting Infor-
mation for full details). In summary, after an extensive screen,
only one commercially available enzyme (porcine kidney acylase)
was identified as being able to hydrolyze the acetamide with high
enantioselectivity; however, this required extended reaction times
(3 days), the presence of cobalt chloride, and a high enzyme
loading (40 wt %). Fortunately, since the asymmetric hydro-
genation route afforded suitable quality 1, this process was not
required. A single gram-scale experiment was conducted to
provide an initial proof-of-concept for an enzymatic hydrolysis
process using optically enriched material (Method C in the
Experimental Section).
’ CONCLUSION
In summary, we have modified the published route towards
(S)-N-Boc-2,6-dimethyltyrosine 1 to make it more practical to
operate. Key changes to the published process were (1) the
discovery and implementation of a novel one-pot process to the
dehydroamino acid 6 from readily available serine methyl ester
and (2) the identification of an alternative, more active catalyst
for the asymmetric hydrogenation step.5
Although the overall yield of the revised process is slightly
lower than that of the method previously published,1a,b we feel
that it offers significant potential for further improvement. The
modified Heck protocol removes the need to prepare methyl
2-acetamidoacrylate 5, and the reduced catalyst loading in the
asymmetric hydrogenation step has the potential to minimize
catalyst costs. Future development of the enzymatic deacylation
offers scope for upgrading the chiral purity as well as providing
milder deprotection conditions.
’ EXPERIMENTAL SECTION
Most of the intermediates have been previously described,1a
and in-house characterization data matched the published data.
All reagents were obtained commercially and used as received
unless otherwise noted.
Most of the reactions were monitored by HPLC analysis using
an Agilent 1100 series using an Agilent Zorbax SB C18 column,
3.0 mm  50 mm  1.8 μm, at 50 °C. UV detection was set at
210 nm with a rate of <0.01 min. The mobile phases used were
water with 0.05% trifluoroacetic acid and acetonitrile at a flow
rate of 1.2 mL/min. Two methods were used. A short method
(starting level of acetonitrile at 5% increased to 100% in 3.5 min,
maintained at 100% for 1 min then decreased to 5% in 0.1 min
and finally maintained at 5% for 0.4 min) and a long method
(starting level of acetonitrile maintained at 5% for 1 min,
increased to 100% in 8 min, maintained at 100% for 2.5 min,
1126
dx.doi.org/10.1021/op200065p |Org. Process Res. Dev. 2011, 15, 1124–1129

Organic Process Research & Development
ARTICLE
then decreased to 5% in 0.1 min, and finally maintained at 5% for
0.4 min).
The resulting solution was stirred at 5 °C (external) for 1 h.
Separately, a slurry of potassium iodide (925.6 g, 5.58 mol) and
potassium iodate (562.4 g, 2.63 mol) in water (7.5 L) was
prepared. The phenol solution was further cooled to 0 °C
(external) before the oxidant mixture was slowly added at such
a rate that the internal temperature was kept below 15 °C. Upon
contacting the acidic methanol, the iodate solution turned dark
red, and over the course of the addition some precipitation
occurred. The vessel used for preparing the oxidant solution was
washed twice with water (500 mL), and the washes were
transferred into the main vessel. After the addition of the oxidant,
the reaction mixture was stirred at 0 °C (external) for 30 min, at
10 °C (external) for 2 h, and then at 20 °C overnight. The
reaction mixture consisted of an off-white precipitate in a dark-
orange solution. Sodium metabisulfite (74 g, 0.39 mol) was
added, resulting in the solution turning pale yellow (no oxidant
remained as determined by testing with an indicator strip). The
reaction mixture was stirred at room temperature for 30 min
before being filtered. The solid was washed with a solution of
sodium metabisulfite (54 g, 0.28 mol) in water (500 mL), then
with water (500 mL) followed by further washing with a mixture
of water (1.5 L) and methanol (1.5 L) and finally with water
(1 L). The resulting powder was dried in a vacuum oven at 50 °C
overnight to give the title compound 3 (1190.2 g, 4.8 mol, 59%
yield).
Retention times for the short method: 3,5-dimethylphenol 2, 2.4
min; 4-iodo-3,5-dimethylphenol 3, 2.8 min; 4-iodo-3,5-dimethyl-
phenyl acetate 4, 3.4 min; methyl-2-acetamido acrylate 5, 1.15 min;
methyl (Z)-2-acetamido-3-(4-acetoxy-2,6-dimethylphenyl)-2-pro-
penoate 6, 2.4 min; 4-acetoxy-N-acetyl-2,6-dimethyltyrosine meth-
yl ester 7, 2.3 min; N-acetyl-2,6-dimethyltyrosine 10, 1.6 min;
2,6-dimethyltyrosine methyl ester, 0.2À0.6 min; N-Boc-2,6-di-
methyltyrosine 1, 2.4 min.
Retention times for long method (used only for hydrogenation
step): Methyl (Z)-2-acetamido-3-(4-acetoxy-2,6-dimethylphenyl)-
2-propenoate 6: 4.5 min, 4-acetoxy-N-acetyl-2,6-dimethyltyrosine
methyl ester 7: 4.4 min.
The conversions for the enzymatic deacetylation were fol-
lowed by HPLC analysis using an Agilent 1200 series using
agilent Zorbax SB C18 column, 3.0 mm  50 mm  1.8 μm, at
70 °C. UV detection was set at 204 nm with a rate of <0.0025 min
(or 0.02 s). The mobile phases used were (A) 95% water/5%
acetonitrile/0.1% trifluoroacetic acid and (B) 95% acetonitrile/
5% water/0.1% trifluoroacetic acid at a flow rate of 1.8 mL/min.
The method consisted of a gradient from 5% to 100% B over 3
min and 100% B maintained for 0.5 min. The mobile phase was
then switched back to the 5% B over 0.1 min and held to reach a
total run time of 4 min.
Retention times: 4-acetoxy-N-acetyl-2,6-dimethyltyrosine
methyl ester 7, 1.31 min; intermediates in the in situ chemical
hydrolysis, 0.89 and 1.07 min; N-acetyl-2,6-dimethyltyrosine 10,
0.6 min; 2,6-dimethyltyrosine 8, 0.27 min.
The conversion of serine methyl ester to methyl-2-acetamido
acrylate was followed by GC/MS using an Agilent GC system
6890N with MS model 5975. An HP-5MSi column, 30 m Â
0.250 mm  0.25 μm, was used at a constant flow of 0.9 mL/min.
The inlet temperature was set at 250 °C, and the oven was
warmed from 50 to 280 °C at a rate of 20 °C/min and maintained
at 280 °C for 1 min. Helium was used as carrier gas with hydrogen
used for FID detection at 300 °C.
1H NMR (400 MHz, CDCl3): δ 2.42 (s, 6H), 4.57 (s, 1H),
6.60 (s, 2H).
4-Iodo-3,5-dimethylphenyl Acetate 41a. 4-Iodo-3,5-di-
methylphenol 3 (2310.4 g, 9.3 mol) was charged to a jacketed
30-L reactor containing pyridine (2.3 L, 28.44 mol) to give a pale-
orange solution that was cooled to 10 °C (external). Acetic
anhydride (1.2 L, 12.7 mol) was added portionwise at such a rate
that the internal temperature was kept below 39 °C. The reaction
mixture was warmed to 55 °C and stirred at this temperature,
with further charges of acetic anhydride (120 mL, 1.7 mol) and
acetic anhydride (120 mL, 1.7 mol) and pyridine (200 mL,
2.4 mol) being added after 100 and 170 min, respectively. After a
further 75 min at 55 °C the reaction was cooled to 25 °C and
stirred for 72 h. Ethyl acetate (17.5 L) was added to the reaction
mixture, and the organic layer was washed with 0.5 N HCl (10 L)
until the pH of the aqueous layer was 1 (four washes were
needed), followed by a water wash (10 L). The organic layer was
dried with magnesium sulfate and then concentrated to dryness
on a rotary evaporator to give the title compound 4 as a yellow oil
that solidified on standing (2644 g, 9.1 mol, 98% yield).
1H NMR (400 MHz, DMSO-d6): δ 2.23 (s, 3H), 2.36 (s, 6H),
6.90 (s, 2H).
Methyl (Z)-2-Acetamido-3-(4-acetoxy-2,6-dimethylphenyl)-
2-propenoate 61a. DL-Serine methyl ester hydrochloride (1477 g,
9.49 mol) was added to a 30-L jacketed reactor containing
acetonitrile (12.5 L) to give a white slurry. Triethylamine (5.3 L,
38.0 mol) was added to the reaction mixture and the slurry cooled
to 0 °C. Acetic anhydride (2.26 L, 23.9 mol) was then added slowly,
at such a rate that the internal temperature stayed below 15 °C. At
the end of the addition, the reaction mixture was warmed to 20 °C
and stirred for 30 min. Iodide 4 (2.5 kg, 8.62 mol), triethylamine
(2.65 L, 19.0 mol), tri-o-tolylphosphine (131.8 g, 0.43 mol), and
palladium acetate (38.9 g, 0.17 mol) were then added. The mixture
was heated to 73 °C and stirred for 13 h. The reaction mixture was
cooled to room temperature and filtered through a bed of Arbocel,
washing with acetonitrile (7.5 L). The filtrate was concentrated on a
rotary evaporator to remove approximately 20 L of solvent. The
Retention times: serine methyl ester, 5.8 min; N,O-diacetyl-
serine methyl ester 9, 8.8 min; methyl-2-acetamido acrylate 5,
6.8 min.
Chiral analysis by HPLC using an Agilent 1100 system. Two
methods were used.
Chiralpak AD-H column, 4.6 mm  100 mm  5 μm, was
used at 25 °C with UV detection set at 210 nm. The mobile
phases used were heptane and a 50:50 mixture of methanol/
ethanol at a flow rate of 2.5 mL/min. Starting level of heptane of
98% decreased to 30% in 5 min, increased to 98% in 0.1 min, and
was maintained at that level for 2 min.
4-Acetoxy-N-acetyl-2,6-dimethyltyrosine methyl ester 7: (R),
1.8 min; (S), 1.9 min
N-Boc-2,6-dimethyltyrosine 1: (R), 1.8 min; (S), 2.2 min
Chiralcel QN-AX column, 4.6 mm  100 mm  5 μm, was
used at 25 °C with UV detection at 210 or 254 nm . The mobile
phase used was a mixture of methanol/acetic acid/ammonium
acetate 98/2/0.5 (v/v/w) at a flow rate of 1.5 mL/min.
N-acetyl-2,6-dimethyltyrosine 11: (R), 6.5 min; (S), 8.7 min
4-Iodo-3,5-dimethylphenol 32. 3,5-Dimethylphenol (1000.9 g,
8.19 mol) was added to a 30-L jacketed reactor containing
methanol (10 L) to give a clear orange solution that was cooled
down to 10 °C (external temperature). Hydrochloric acid
solution (36.5% w/w, 5 L, 50.5 mol) was added portionwise at
such a rate that the internal temperature was kept below 20 °C.
1127
dx.doi.org/10.1021/op200065p |Org. Process Res. Dev. 2011, 15, 1124–1129

Organic Process Research & Development
ARTICLE
residue was carefully diluted with a mixture of ethyl acetate (25 L)
and saturated sodium bicarbonate (10 L). The layers were sepa-
rated, and the organic solution was diluted with further ethyl acetate
(2.5 L) before being washed with saturated sodium bicarbonate
(10 L) followed by water (10 L). The organic solution was treated
with magnesium sulfate (2 kg), filtered (washing with 5 L of ethyl
acetate), and held for 72 h. During this time a small amount of
precipitate formed, which was isolated by filtration and washed with
ethyl acetate (2.5 L) to give dehydroamino acid 6 (112.2 g, 0.4 mol,
4% yield) of excellent purity. The filtrate was stirred with Carbon
Darko KB wet powder 100 mesh (1 kg) for 2.5 h at room
temperature and then filtered through a filter pad consisting of
Arbocel and silica, washing with ethyl acetate (5 L). The filtrate was
separated into two roughly equal portions that were concentrated
to dryness on a rotary evaporator. The resulting yellow solid was
slurried in a mixture of ethyl acetate (6.5 L) and heptane (13 L) for
2 h at reflux and then for 16 h at 20 °C. The precipitate was isolated
by filtration, washed with a mixture of ethyl acetate (800 mL) and
heptane (1.6 L) and then with heptane (1.4 L) to give the title
product 6 as a white solid in two crops (848.7 g, 2.7 mol, 32% yield
and 873.1 g, 2.8 mol, 33% yield).
was added slowly until the mixture reached pH 4, maintaining the
internal temperature between 40 and 47 °C. The resulting slurry
was stirred at 20 °C overnight. NaOH (40%, 200 mL) was added
to adjust the pH to 10, followed by di-tert-butyldicarbonate
(752 g, 3.45 mol). The reaction mixture was stirred at 20 °C for
72 h, maintaining the pH between 8 and 10 by regular addition of
40% NaOH (700 mL used in total). Hydrochloric acid (36.5%
w/w; 1.15 L) was added (to adjust to pH 4), followed by ethyl
acetate (10 L). After stirring for 10 min, the phases were separated.
The organic layer was dried with magnesium sulfate and then
concentrated to dryness to give a beige solid. This residue was
stirred in a mixture of ethyl acetate (4 L) and heptane (8 L) for
16 h, isolated by filtration, and washed with a mixture of ethyl acetate
(500 mL) and heptane (1 L), followed by heptane(600 mL), to give
the title product 1 (689 g, 2.2 mol 70% yield, 95% ee) as an off-white
powder after drying in a vacuum oven at 50 °C.
1H NMR (400 MHz, DMSO-d6): δ 1.14 (s, 2H), 1.29 (s, 7H),
2.14 (s, 6H), 2.73 (dd, J = 12 and 8 Hz, 1H), 2.91 (dd, J = 12 and
8 Hz, 1H), 3.95 (m, 1H), 6.33 (s, 2H), 7.00 (d, J = 8 Hz, 1H),
8.85 (s,1H).
Method B. To a slurry of (S)-4-acetoxy-N-acetyl-2,6-dimethyl-
tyrosine methyl ester (120 g, 390 mmol, 93% ee) in THF
(400 mL) was added di-tert-butyldicarbonate (170.8 g, 782
mmol) and THF (200 mL) followed by DMAP (9.5 g, 77 mmol)
and THF (200 mL). The clear solution was stirred at 20 °C for 16
h, and then sodium hydroxide solution (5 M; 460 mL, 2.30 mol)
was added. After stirring at 20 °C for 5 h, the phases were
separated. The aqueous phase was diluted with water (150 mL)
The pH was then adjusted to 4 with HCl (36.5%; 190 mL, 2.21
mol), keeping the reaction temperature below 30 °C. Ethyl
acetate (500 mL) was added, and the phases were separated.
The organic layer was dried with magnesium sulfate and then
concentrated to dryness to give an orange oil that was refluxed in
ethyl acetate (180 mL) for 1 h to give a white slurry. Heptane
(480 mL) was added and the reaction mixture left under reflux
for a further 2.5 h. The mixture was then cooled to 20 °C and held
for 16 h. The solid was isolated by filtration, washed with heptane
(120 mL) and dried to give the title product 1 (110.2 g,
356 mmol 91% yield, 93.4% ee).
Method C. (S)-4-Acetoxy-N-acetyl-2,6-dimethyltyrosine meth-
yl ester 6 (1.0 g, 3.25 mmol, 94.6% ee) was added to phosphate
buffer (0.1 M, pH 8.0) containing 1 mM cobalt chloride (40 mL)
and stirred at 35 °C. Sodium hydroxide (5 M, 1.478 mL) was
autotitrated to the slurry to maintain the reaction at pH 10. The
addition stopped once the starting material’s ester groups had
hydrolyzed. Hydrochloric acid (6 N, 0.235 mL) was then charged
to adjust the reaction mixture to pH 8 and porcine kidney acylase
(Sigma A8376; 408.8 mg) added. The reaction was stirred at
35 °C for three days. Sodium hydroxide solution (10 M, 0.5 mL)
was added to adjust to pH 10, followed by the addition of di-tert-
butyldicarbonate (0.79 g, 3.62 mmol). The reaction mixture was
stirred at 20 °C for 21 h, holding the pH between 8 and 10 by
regular addition of 10 M NaOH (1 mL charged in total). HCl
(36.5%, 1.1 mL) was added to adjust to pH 4 (some off-gassing
noted), followed by addition of ethyl acetate (10 mL). The
resultant biphasic mixture was filtered to remove enzyme residues
which were further washed with ethyl acetate (10 mL). The
organic phase was separated and dried over anhydrous magne-
sium sulfate before being concentrated under reduced pressure to
give a dark foam. This residue was dissolved in ethyl acetate
(5 mL) before heptane (10 mL) was slowly added to effect
precipitation. The slurry was stirred at 20 °C for 16 h; then the
1H NMR (400 MHz, DMSO-d6): δ 1.84 (s, 3H), 2.12 (s, 6H),
2.25 (s, 3H), 3.70 (s, 3H), 6.85(s, 2H), 6.91 (s, 1H), 9.30 (s,1H).
(S)-4-Acetoxy-N-acetyl-2,6-dimethyltyrosine Methyl Ester
71a. Dehydroamino acid 6 (130 g, 425 mmol) and [Rh-(1,5-
COD)(S,S-Et-FerroTANE)]BF4 (0.63 g, 0.85 mmol, 0.002
equiv) were added to a 5-L hydrogenation vessel containing
ethyl acetate (1.95 L) to give a slurry. The vessel was purged
three times with nitrogen (5 bar) without stirring. The vessel was
then pressurized with nitrogen (5 bar) and stirred for 10 min
before the pressure was released; this was repeated twice. The
vessel was pressurized with nitrogen (5 bar) and warmed to
60 °C with stirring. The pressure was released. The vessel was
pressurized again with nitrogen (5 bar), stirred at 60 °C for 10
min, and then vented. This process was repeated twice. The
vessel was then purged with hydrogen (5 bar, three times)
without stirring, followed by pressurization with hydrogen
(5 bar) and stirring at 60 °C for 16 h. The vessel was cooled to
20 °C and was purged three times with nitrogen (5 bar) with
stirring. SiliaBond thiol (Silicycle Inc., 39 g) was added to the
reaction mixture and the slurry stirred at 20 °C for 2 h. The
mixture was filtered and washed with ethyl acetate (2 Â 200 mL),
and the filtrate was concentrated under reduced pressure to give
an off-white solid. Ethyl acetate (130 mL) and heptane (650 mL)
were added to the residue, and the slurry was warmed to reflux.
Upon cooling, a precipitate formed that was stirred at room
temperature overnight. The precipitate was isolated by filtration
and washed with heptane (2 Â 130 mL) to give the title
compound 7 as a white solid (125.2 g, 407 mmol, 95% yield,
93% ee) after drying.
1H NMR (400 MHz, DMSO-d6): δ 1.81 (s, 3H), 2.22 (s, 3H),
2.26 (s, 6H), 2.93 (dd, J = 16 and 8 Hz), 3.05 (dd, J = 16 and
8 Hz), 3.51 (s, 3H), 4.44 (q, J = 8 Hz), 6.75 (s, 2H), 8.84 (d, J =
4 Hz, 1H).
(S)-N-Boc-2,6-dimethyltyrosine 1. Method A. (S)-4-Acet-
oxy-N-acetyl-2,6-dimethyltyrosine methyl ester 7 (975.8 g, 3.1
mol, 95% ee) was added to a 30-L jacketed reactor containing
water (4.4 L) to give a slurry. Hydrochloric acid (36.5% w/w;
4.3 L, 50.0 mol) was slowly added at such a rate that the internal
temperature was kept below 30 °C. The reaction mixture was
heated to 90 °C and held for 6.5 h before cooling to 40 °C. The
jacket temperature was then set at 10 °C, and 40% NaOH (3.6 L)
1128
dx.doi.org/10.1021/op200065p |Org. Process Res. Dev. 2011, 15, 1124–1129

Organic Process Research & Development
ARTICLE
solid was isolated by filtration and washed with a mixture of ethyl
acetate/heptane (1:1, 5 mL) to give the title product 1 (0.55 g,
54% yield, >99% ee, (R)-enantiomer not detected) as an off-white
powder after drying in a vacuum oven at 50 °C.
’ ASSOCIATED CONTENT
S
Supporting Information. This material is available free
b
of charge via the Internet at http://pubs.acs.org.
’ AUTHOR INFORMATION
Corresponding Author
*celine.praquin@pfizer.com.
’ ACKNOWLEDGMENT
We thank our colleagues John Deering, Asayuki Kamatani, and
Matthew Morland without whom no large-scale work would
have been possible, and John Wong for helpful discussions
regarding the enzymatic deacetylation. The members of the
analytical department are thanked for the support they gave in
providing instruments and methods. We also thank the reviewers
for their thoughtful comments.
’ REFERENCES
(1) (a) Dygos, J. H.; Yonan, E. E.; Scaros, M. G.; Goodmonson, O. J.;
Getman, D. P.; Periana, R. A.; Beck, G. R. Synthesis 1992, 741.(b)
Getman, D. P.; Periana, R. A.; Riley, D. P. US Patent 4879398, 1987. (c)
Tang, X.; Soloshonok, V. A.; Hruby, V. J. Tetrahedron 2001, 57, 6375.
(d) Tang, X.; Soloshonok, V. A.; Hruby, V. J. Tetrahedron: Asymmetry
2000, 11, 2917. (e) Balducci, D.; Contaldi, S.; Lazzari, I.; Porzi, G.
Tetrahedron: Asymmetry 2009, 20, 1398.(f) Hansen, D. W.; Jones, D. A.;
Mazur, R. H.; Schlatter, J. M. US Patent 4722922, 1988.
(2) Hunig, S.; Schwarz, H. Liebigs Ann. Chem. 1956, 599, 131.
(3) (a) Gallou-Dagommer, I.; Gastaud, P.; Rajanbabu, T. V. Org. Let.
2001, 3, 2053. (b) Adams, L. A.; Aggarwal, V. K.; Bonnert, R. V.; Bressel,
B.; Cox, R. J.; Shepherd, J.; de Vincente, J.; Walter, M.; Whittingham,
W. G.; Winn, C. L. J. Org. Chem. 2003, 68, 9433. (c) Gangadasu, B.;
Narender, P.; Kumar, S. B.; Ravinder, M.; Rao, B. A.; Ramesh, Ch.; Raju,
B. C.; Rao, V. J. Tetrahedron 2009, 62, 8398. (d) Wang, H.; Zhang, J.;
Xian, M. J. Am. Chem. Soc. 2009, 131, 13238. (e) Crestey, F.; Collot, V.;
Stiebing, S.; Rault, S. Synthesis 2006, 3506.
(4) From a range of catalogue companies, the price quoted varied
from £600/g to £900/g.
(5) A selection of hydrogenation catalysts were screened under 20
bar hydrogen at 60 °C in ethyl acetate for 20 h. Either 10 wt % of a
preformed metalÀligand complex (Method A), or 4 wt % Rh(COD)2-
(acac)2 and 6 wt % ligand (Method B) were used. No conversion was
observed with the following: [(R)-TCFP-Rh-COD]BF4 (Method A) or
Josiphos SL-J001-1; Josiphos SL-J002-1; Mandyphos SL-M001-1; Mandyphos
SL-M002-1; (R,R)-Norphos; Solphos SL-A00101; (R,R)-MeDuPHOS;
(S,S,R,R)-Duanphos and (R)-Xylylphanephos (Method B). Partial con-
version was observed with the following: Walphos SL-W001-1 (58%
conversion); Walphos SL-W002-1 (25% conversion); Walphos SL-
W003-1 (26% conversion); Mandyphos SL-M002-1 (14% conversion)
(all Method B) and complete conversion was obtained with [(R,R)-
MeBPE-Rh-COD]BF4 (9% ee, (R)) and [(S,S)-Et-Ferrotane-Rh-
COD]BF4 (93% ee, (S)) (all Method A).
(6) This catalyst currently costs £394/500 mg from Strem
(catalogue number 45-0155). As we had sufficient quantities of this
catalyst in-house and a quick turnaround was required, no consideration
was given to the long-term supply of this catalyst.
(7) Burk, M. J.; Allen, J. G. J. Org. Chem. 1997, 62, 7054.
1129
dx.doi.org/10.1021/op200065p |Org. Process Res. Dev. 2011, 15, 1124–1129