Development of a practical and convergent process for the preparation of sulopenem

Article abstract of DOI:10.1021/op300131e

Previous synthetic processes for the preparation of sulopenem involved multistep linear sequences in which the chiral sulfoxide side chain was introduced early in the process. This contribution summarizes the development of a practical and convergent process for the large-scale preparation of 1. The key step in the synthesis involves cyclization of an oxalimide intermediate to provide the thiopenem core. This convergent strategy allows for late introduction of the expensive and labile chiral sulfoxide subunit. Additionally, a regioselective sulfur oxidation and an improved deprotection sequence were developed. The latter provides API of high purity without the need for recrystallization.

Full text of DOI:10.1021/op300131e

Article
pubs.acs.org/OPRD
Development of a Practical and Convergent Process for the
Preparation of Sulopenem
Steven J. Brenek,† Stephane Caron,† Esmort Chisowa,†,§ Mark P. Delude,‡ Michele T. Drexler,†
́
Marcus D. Ewing,† Robert E. Handfield,† Nathan D. Ide,† Durgesh V. Nadkarni,*,†,⊥ Jade D. Nelson,*,†
Mark Olivier,‡ Hahdi H. Perfect,† James E. Phillips,† John J. Teixeira,† R. Matt Weekly,†
and John P. Zelina‡
‡
†Chemical Research and Development, Analytical Research and Development, Pfizer Worldwide Research and Development,
Eastern Point Road, Groton, Connecticut 06340, United States
S
* Supporting Information
ABSTRACT: Previous synthetic processes for the preparation of sulopenem involved multistep linear sequences in which the
chiral sulfoxide side chain was introduced early in the process. This contribution summarizes the development of a practical and
convergent process for the large-scale preparation of 1. The key step in the synthesis involves cyclization of an oxalimide
intermediate to provide the thiopenem core. This convergent strategy allows for late introduction of the expensive and labile
chiral sulfoxide subunit. Additionally, a regioselective sulfur oxidation and an improved deprotection sequence were developed.
The latter provides API of high purity without the need for recrystallization.
The second-generation synthesis involved conversion of
trithiocarbonate 2 to thiopenem 8 (analogous to 5) via a base-
induced cyclization, followed by desulfurization with triethyl-
phosphite, in a process reminiscent of an Eschenmoser sulfide
contraction.8,12 Although this approach also comprised a linear
synthetic sequence for the preparation of the penem core,
several of the key issues that prevented larger-scale production
of API were addressed. This route proved satisfactory for the
production of phase 3 supplies; however, it was deemed
unacceptable for commercial manufacture, primarily due to a
poor yield (20−25%) for the key cyclization sequence. As a
result, a process research program was initiated to identify an
improved disconnection strategy. We report herein the
development of an efficient, convergent process for the
synthesis of 1.
INTRODUCTION
■
Penems, a class of β-lactam antibiotics, have been the subject of
intense research over the last three decades.1−6 Sulopenem (1)
is a novel 2-thioalkyl penem that was in development for a
broad range of indications (Figure 1).7,8 In order to support
this research program, a practical process for the large-scale
preparation of this antibacterial was required.
RESULTS AND DISCUSSION
■
Figure 1. Chemical structure of sulopenem.
Process Research. Identification of a Potential Commer-
cial Route. The combination of a high daily intravenous dose, a
challenging β-lactam structure, and an aggressive clinical
development plan for sulopenem (1) required the identification
of scalable technology for the manufacture of rapidly escalating
quantities of active pharmaceutical ingredient (API). To enable
early alignment of suppliers for custom raw materials, 4-
acetoxyazetidinone (9) and the proprietary chiral sulfoxide 10
were selected as building blocks for all potential commercial
processes (Figure 2). Furthermore, we committed to targeting
the late-stage thiopenem intermediate 8, since the end-game
chemistry for conversion of this compound to sulopenem (1)
was well-established by the first- and second-generation
syntheses.9 Conserving the same end-game strategy also
During the early phases of clinical development, sulopenem
(1) was prepared at multikilogram scale via two distinct
chemical processes (Scheme 1).9 Both approaches utilized
trithiocarbonate 2 as a key building block but differed in the
chemistry used for construction of the thiopenem ring. The
first-generation synthesis employed an oxalimide-derived
carbene cyclization for construction of intermediate 5 via
oxalimide 4.10,11 Penem 5 was then converted to sulopenem via
two well-established deprotection steps (desilylation, followed
by Pd-catalyzed deallylation). While this synthesis was fit for
purpose, scale-related challenges prohibited larger-scale appli-
cation. Primary drawbacks included low yield for the cyclization
step (∼40%) and high environmental impact associated with
CHCl3 utilization. It became clear that a more efficient process
was required to support development and commercialization.
Received: May 18, 2012
Published: July 13, 2012
© 2012 American Chemical Society
1348
dx.doi.org/10.1021/op300131e | Org. Process Res. Dev. 2012, 16, 1348−1359

Organic Process Research & Development
Article
Scheme 1. First- and second-generation routes for construction of key intermediates
nucleus with mercaptan side chains.16−18 Unfortunately,
analogous reactions with thiopenems proved largely unsuccess-
ful. Conversely, more reactive enol triflates, or certain alkyl
sulfinate derivatives, have been reported to be effective for
coupling a wide range of thiols to prepare substituted
thiopenems. Preparation and use of an enol triflate (e.g., 14)
would be problematic on large scale for several reasons.19
Therefore, initial efforts were focused on developing alkyl
sulfinate analogues (e.g., 15) for the introduction of the
sulopenem side chain.11,20
Construction of the 2-Thioalkylpenem Core. Numerous
synthetic approaches were evaluated in the laboratory for the
preparation of the challenging penem core (Figure 3). The
most promising hits were further assessed for synthetic viability
on larger scale.
A number of methods for construction of carba- and
thiopenems have been reported in the synthetic chemistry
literature.21−23 Key strategies include C−C,1,3,24−26 C−N,27
and C−S28−30 bond formation for preparation of the five-
membered ring. However, many of the methods that have
proven successful for carbapenems are not applicable to
thiopenems. For instance, incompatibility between sulfur-
containing substrates and a necessary catalyst or promoter
(e.g., Rh) beset the metallo−carbenoid insertion approach.31
Difficulties in the construction and handling of starting
materials for a key transformation complicated the vinyl halide
cross-coupling32 and intramolecular imine formation strategies.
Finally, other routes reached a proof-of-concept phase in the
laboratory, but were either too lengthy (e.g., thiolactonization
and enolate addition to thiocarbonate33 approaches) or
Figure 2. Key building blocks for sulopenem synthesis.
minimized the potential for introducing new process-related
impurities.
After a preliminary cost analysis, it became clear that
commercial processes for the manufacture of 1 would need to
be more convergent than the routes used during development.
A late introduction of the chiral sulfoxide moiety would have a
positive impact, since this subunit was both expensive and
relatively labile. Two complementary approaches for the late-
stage introduction of this group have been reported (Scheme
2).8,13,14 The first method involves S-alkylation of the
thioenolate derived from 11 with a suitable electrophile, such
as 12.15 Unfortunately, this approach resulted in mixtures of
stereoisomers due to either epimerization of the electrophile
12a (X = Br) via halogen−halogen displacement, or via β-
elimination of the leaving group in 12b (X = OTs) to afford a
conjugate acceptor that reacts with poor facial selectivity.
A second, more desirable, approach involves conjugate
addition of thiol 13 (obtained from thioacetate 10) to an
appropriately activated penem derivative (e.g., 14 or 15). For
carbapenems, leaving groups (Y) such as enol phosphates or
enol tosylates have been used for coupling the carbapenem
Scheme 2. Convergent approaches to protected penem 8
1349
dx.doi.org/10.1021/op300131e | Org. Process Res. Dev. 2012, 16, 1348−1359

Organic Process Research & Development
Article
Figure 3. Survey of methods for thiopenem synthesis.
Scheme 3. Preparation of oxalimide derivatives from 9
technically challenging/costly for commercial-scale application
(azomethine ylide34 and Eschenmoser sulfide contraction9
routes).
These salts were subsequently reacted with azetidinone 9 to
provide 16−18 in good yields. Compounds 16−18 were each
crystalline solids that exhibited improved stability relative to
trithiocarbonate 2, which was utilized in the first- and second-
generation syntheses.
The trithiocarbonates 16−18 were then treated with allyl
oxalyl chloride 19 in the presence of triethylamine to form the
desired oxalimides 20−22. These oxalimides were each carried
into the carbene cyclization step, in order to compare their
performances in this key reaction.
After extensive evaluation of potential methods, a variation of
the oxalimide carbene cyclization approach was selected.10
Despite poor performance of this strategy in the first-generation
process, the use of more stable oxalimide cyclization precursors
led to significantly increased viability as a commercial approach.
Key to the success of this modification was introduction of the
chiral sulfoxide side chain after thiopenem construction, which
offered several advantages.
Preparation of Oxalimides from Azetidinone 9. As in
the first-generation synthesis, the commercially available
azetidinone 9 was utilized in the preparation of several
oxalimides for evaluation in the key carbene cyclization step
(Scheme 3).8 Azetidinone 9 was converted to a range of alkyl
trithiocarbonates 16−18 utilizing simple, inexpensive thiols and
carbon disulfide. Aliphatic thiols were chosen due to their
performance in the cyclization sequence, coupled with the
knowledge that the resulting 2-thioalkylpenems could be
further converted to sulopenem (1) via established methods.20
Methyl, n-propyl, and tert-butyl thiols were treated with
sodium hydride in MtBE, followed by addition of carbon
disulfide to afford the corresponding trithiocarbonate salts.
Cyclization of Oxalimides to Thiopenems. Cyclizations of
oxalimides of this type have traditionally been performed using
trialkyl phosphites under high dilution in chlorinated
solvents.10,11 During the first-generation process (Scheme 1)
the optimal conditions for conversion of 4 to 5 involved the use
of triethylphosphite in CHCl3.8 Toluene was also identified as
being compatible with the carbene cyclization, and was chosen
as a more environmentally friendly alternative for the current
study. Oxalimides 20−22 were treated with two molar
equivalents of triethylphosphite in toluene at elevated temper-
ature to provide the corresponding thiopenems 23−25
(Scheme 4) in similar in situ yields (∼50%).
Upon completion of this initial screen, the advantages of
employing 1-propanethiol in this sequence became apparent.
1350
dx.doi.org/10.1021/op300131e | Org. Process Res. Dev. 2012, 16, 1348−1359

Organic Process Research & Development
Article
(e.g., triethylamine) promoted the reaction to provide
thiopenem 8 (Scheme 8), an advanced intermediate in the
second-generation process to sulopenem (1).9
Scheme 4. Cyclization of oxalimide to penem
Process Optimization for the Preparation of Penem 8.
Telescoped Synthesis of Oxalimide 21. While trithiocarbonate
17 could be isolated via crystallization from heptanes, several
crops were required to maximize recovery. Therefore, a
telescoped process for the preparation of oxalimide 21 was
developed that took advantage of the high in situ yield of 17. 1-
Propanethiol was treated with sodium hydride to afford the
corresponding thiolate. After addition of carbon disulfide, the
resulting intermediate was combined with 4-acetoxyazetidinone
(9) to provide trithiocarbonate 17 in nearly quantitative yield
(Scheme 9).
This thiol is inexpensive and nonvolatile (vs MeSH) and led to
highly crystalline, isolable compounds at both the oxalimide
(21) and thiopenem (24) stages. Therefore, intermediate 21
was selected to evaluate downstream chemistry. As described in
Scheme 5, conversion of thiopenem 24 to the known late-stage
intermediate 8 was envisioned to include selective sulfur
oxidation, followed by displacement of the sulfinate of
intermediate 26 with the thiol derived from deacetylation of 10.
Formation of Penem Sulfoxide 26 by Oxidation of Penem
24. A survey of the literature indicated that the most common
methods for oxidation of compounds of type 24 employed
either m-chloroperoxybenzoic acid (m-CPBA)20 or bleach, and
proceeded in modest yields.24,25,35 Due to its simplicity and
reported regioselectivity favoring the exocyclic sulfur, m-CPBA
was initially selected for conversion of 24 to 26. This oxidant
provided the desired sulfoxide 26 as the major product, albeit as
a mixture of diastereoisomers contaminated by significant levels
of sulfone 27 and regioisomer 28 (Scheme 6).
Nevertheless, the ability to produce sulfoxide 26 in
multigram quantities enabled us to focus on preparation of a
suitable nucleophile for introducing the chiral side-chain
moiety. Initial trials using historical methods8 involved
treatment of thioacetate 10 with aqueous sodium hydroxide
to liberate the corresponding sodium thiolate 29 (Scheme 7).
The reaction mixture was then acidified to afford thiol 13,
which unfortunately proved to be highly water-soluble.
Although 13 could be partially extracted into CH2Cl2,
concentration of solutions resulted in the formation of disulfide
30 (Figure 4). Disulfide formation occurred primarily during
concentration of the solution of 13 even when care was taken
to exclude oxygen. In order to avoid handling thiol 13, direct
addition of thiolate 29 to sulfoxide 26 was investigated.
Unfortunately, this approach led to poor yields of the desired
coupling product.
To avoid the issues experienced with extraction of thiol 13
from an aqueous solution, thioacetate 10 was converted to
thiolate 29 with sodium ethoxide in ethanol. This was followed
by acidification with acetic acid to give crude thiol 13
containing minimal disulfide 30. Although thiol 13 proved
incapable of addition to 26 directly, addition of nitrogen bases
The crude reaction mixture in MtBE was filtered through a
pad of silica gel and then treated with allyl oxalyl chloride 19
and triethylamine at −10 °C to provide oxalimide 21 in 89−
92% yield from 9.
Carbene Cyclization to Prepare Thiopenem 24. Although
the conversion of oxalimide 21 to thiopenem 24 in toluene had
been demonstrated, the reaction was far from optimal in terms
of yield, throughput and scalability. Initial cyclization conditions
utilizing triethylphosphite in toluene at 75 °C resulted in 35−
40% in situ yield of penem 24. Cyclization proceeded in a range
of solvents, but CHCl3, toluene and trifluorotoluene gave the
highest yields. Evaluation of reaction parameters including
concentration, temperature and phosphite addition rate did not
lead to substantial improvement. Alternative phosphite reagents
(e.g., P(OiPr)3, P(OPh)3) gave inferior results, but a break-
through was realized when dialkylphosphonite reagents (P-
(OR)2R′) were used (Table 1).36,37 For example, the use of
dimethyl phenylphosphonite or diethyl methylphosphonite, in
toluene, provided thiopenem 24 in an improved 69% and 68%
in situ yield, respectively (entries 4 and 5). Further evaluation
of this class of reagents showed generally enhanced perform-
ance compared to phosphites. Of the phosphonites evaluated,
diethyl methylphosphonite provided the best combination of
performance and bulk reagent cost, so it was selected for
further development.
With the substantial increase in yield attained by using a
more effective phosphorous reagent, focus shifted to evaluation
of the mass balance for this cyclization. As suspected from
previous experience on the first-generation synthesis, the allyl
protecting group on oxalimide 21 led to 5−10% of the
cyclopropanation byproduct 31 (Scheme 10).10 Oxalimide 32,
which contains a chloroallyl protecting group, was prepared in
order to compare performance in the cyclization.11 Chloroallyl
protection was found to minimize intramolecular cyclo-
propanation byproducts in the first-generation carbene syn-
thesis through added steric hindrance, as well as by rendering
the alkene less electron-rich. Under equivalent conditions,
Scheme 5. Proposed introduction of the chiral side chain
1351
dx.doi.org/10.1021/op300131e | Org. Process Res. Dev. 2012, 16, 1348−1359

Organic Process Research & Development
Article
Scheme 6. Selective oxidation of thiopenem 24
Scheme 7. Preparation of thiolate 29 and thiol 13
Table 1. Phosphorous reagent screen
a,b
entry
phosphorous reagent
triethyl phosphite, P(OEt)3
yield
1
2
3
4
5
6
7
35−40
15−20
trace
69
Figure 4. Disulfide impurity 30.
triisopropyl phosphite, P(Oi-Pr)3
triphenyl phosphite, P(OPh)3
Scheme 8. Preparation of thiopenem 8
dimethyl phenylphosphonite, PhP(OMe)2
diethyl methylphosphonite, MeP(OEt)2
diisopropyl phenylphosphonite, PhP(Oi-Pr)2
68
53
diethyl phenylphosphonite, PhP(OEt)2
39
a
b
In situ yields determined by HPLC analysis. Conditions: 21 (2
mmol), phosphorous reagent (2.25 equiv) added over 4 h, toluene (30
mL/g), 90 °C.
products. Although the intramolecular carbene cyclization is
optimally carried out under high dilution, reasonable yields
could be achieved with as little as 20 volumes (L/kg substrate)
of toluene.10 Utilizing the findings from the experimental
design, optimal conditions were determined. This process
involved addition of 2.5 equiv of methyl diethylphosphonite
over 4−5 h to a solution of 21 in toluene (30 volumes) at 90
°C. In situ yields of 63−69% were reproducibly achieved with
isolated yields typically 54−60%. After completion of the
reaction, toluene was displaced with i-PrOH under reduced
pressure. The product was subsequently crystallized from an i-
PrOH concentrate as large, high density laths, which effectively
excluded process-related impurities (see Figure 6). The high
bulk density of this material also enabled high throughput
manufacture.
Identification of Optimal Sulfur Oxidation Conditions.
Initial oxidation of penem 24 with m-CPBA was suitable for
proof of concept; however, only low yields of desired sulfoxide
26 were obtained. Despite significant optimization, the m-
CPBA oxidation unfortunately resulted in maximum yields of
only 55−59% of sulfoxide 26. In addition, safety concerns
associated with the use of m-CPBA on large scale,39 led to an
oxalimide 32 was converted to thiopenem 33 in a slightly
improved yield (∼10% greater than 21→24), as the
corresponding chlorocyclopropane 34 was not observed.
Since 2-chloroallyl oxalyl chloride is significantly more
expensive than the allyl derivative, the minimal yield improve-
ment did not justify a change in strategy. Therefore, allyl
oxalimide 21 was utilized for further development.
In addition to the cyclopropanation byproduct, the thiol
derivative 11, and S-ethyl penem 35 were each identified
(Figure 5). These two compounds contributed <10% to the
overall mass balance. These impurities formed in detectable
levels under all reaction conditions explored, but neither
compound was problematic in downstream chemistry.38
A design of experiment (DoE) approach was utilized to
further optimize reaction conditions for the cyclization of
oxalimide 21 in toluene with diethyl methylphosphonite. These
investigations revealed that concentration and reaction temper-
ature were important factors. The optimal reaction temperature
appeared to be in the range of 80−90 °C. Below 65 °C, the
reaction failed to reach completion after extended periods,
while elevated temperatures led to thermal decomposition
Scheme 9. Telescoped synthesis of oxalimide 21
1352
dx.doi.org/10.1021/op300131e | Org. Process Res. Dev. 2012, 16, 1348−1359

Organic Process Research & Development
Article
Scheme 10. Formation of cyclopropanation byproduct 31
required.46 Unfortunately, optimal yield and selectivity for the
formation of sulfoxide 26 required at least 4−5 volumes of
HFIP. The use of polar cosolvents (e.g., THF, ethyl acetate,
acetonitrile, and ethanol) inhibited the reaction, while nonpolar
cosolvents (e.g., toluene and heptanes) could be used
successfully. The use of toluene significantly slowed the
reaction rate, but reactions eventually reached completion
with similar yield. Heptanes and HFIP are immiscible, but use
of the biphasic mixture produced desired sulfoxide 26 in similar
yield and selectivity to that using neat HFIP, without a
significant rate reduction. Presumably, the desired oxidation
occurs in the HFIP layer. Although the use of heptanes
cosolvent did not decrease the required HFIP volume, it
enabled the development of a suitable recycling process for this
expensive solvent.
Figure 5. Cyclization byproducts 35 and 11.
The reaction could be run in the presence of heptanes.
However, the optimal process was carried out in neat HFIP
(4−5 L/kg), and upon reaction completion, heptanes were
added. The biphasic solution could then be distilled under
reduced pressure.47 The distillate was collected as a biphasic
mixture, and the lower HFIP layer could be easily separated and
reused without further purification. The concentrated reaction
mixture was carried on to the next step without isolation.
Coupling thiol 13 with sulfoxide 26. Thioacetate 10 is
efficiently converted to thiol 13 via hydrolysis to thiolate 29
with sodium ethoxide followed by acidification with acetic acid
(Scheme 7). Direct coupling of thiol 13 with sulfoxide 26, in
the presence of a trialkylamine, provided thiopenem 8 in
scouting experiments. In order to maximize efficiency at large
scale, a telescoped process was desired. Therefore, an extensive
screen of base and solvent combinations was carried out.
Isopropyl acetate emerged as the optimal solvent for hydrolysis
of 10 to 13 and also proved suitable for the coupling step. For
the coupling, no amine base proved to be superior to
triethylamine in terms of cost and effectiveness. A two-step
telescoped process was developed that provided thiopenem 8 in
80−83% in situ yield. This process tolerated crude solutions of
sulfoxide 26, allowing a multistep telescoped sequence to be
utilized (Scheme 12), which further improved the efficiency of
the overall synthesis. After aqueous workup and isolation from
heptanes, 8 was obtained in 55% overall yield from thiopenem
24.
Figure 6. Crystals of thiopenem 24.
evaluation of alternatives. Extensive screening of catalytic (e.g.,
V, Mo, W, Au) and stoichiometric oxidants was performed.
These efforts resulted in identification of two highly
regioselective oxidation systems for the formation of sulfoxide
26. A catalytic system utilizing methyltrioxorhenium
(MTO)40,41 and urea−hydrogen peroxide (UHP) in trifluor-
oethanol (TFE), as well as a noncatalytic system using urea−
hydrogen peroxide (UHP) in hexafluoroisopropanol
(HFIP),42−44 each produced sulfoxide 26 in 70−75% yield.
Due to the limited availability and high cost of MTO, the latter
method was selected for further development (Scheme 11).
HFIP is a highly polar specialty solvent, commercially
available on multiton scale from several suppliers.45 However,
due to the relatively high cost of HFIP compared to that of
more traditional solvents, it was evident that maximizing
concentration and/or recycling this material would be
End-Game Development. Developing an Ideal Depro-
tection Strategy. The synthesis described herein proved
efficient and robust at moderate-to-large laboratory scale,
providing thiopenem 8 in an overall yield of 30% (five steps)
from 4-acetoxyazetidinone (9). Therefore, focus shifted to
improvements in the end game. The end-game strategy utilized
in both first- and second-generation syntheses of sulopenem
(1) involved desilylation of the secondary hydroxyl group,
followed by palladium-catalyzed deallylation to provide the free
carboxylic acid (Scheme 13).9,11,48
Scheme 11. Selective oxidation of 2-thioalkyl penems with
UHP in HFIP
1353
dx.doi.org/10.1021/op300131e | Org. Process Res. Dev. 2012, 16, 1348−1359

Organic Process Research & Development
Article
Scheme 12. Preparation of thiopenem 8
Scheme 13. Historical deprotection sequence
Scheme 14. Reordered deprotection sequence
Utilizing this strategy, the final API consistently required
remediation due to high levels of residual palladium. Aqueous
reprocessing of crude 1 was effective in purging Pd to an
acceptable level; however, this resulted in unfortunate product
losses of 15−20%. Late in the development of the second-
generation route, it was discovered that the addition of a slight
excess of triphenylphosphine to the deallylation led to isolated
API that met tight Pd specifications.49 This addition did not
hinder the deprotection step, but it precluded the need for
reprocessing by maintaining a soluble form of Pd throughout
isolation. High Pd and/or triphenylphosphine oxide levels
nevertheless remained a potential risk and were closely
monitored. For commercial manufacturing, a reordered end-
game sequence was desired (Scheme 14).
Deprotection of 8 Leading to Formation of Sulopenem
(1). Performing the deallylation before the desilylation provided
a two-fold advantage. The step involving Pd is further removed
from the end of the synthesis, and the resulting TBS-protected
acid is a fine crystalline solid that is relatively insoluble in water
and most common organic solvents. The new strategy would
provide another opportunity for purification via crystallization
prior to isolation of sulopenem (1).
The Pd-catalyzed deallylation of 8 was studied closely to
optimize catalyst, solvent, and allyl scavenger.50 Initial screening
provided conditions that included Pd(Ph3P)4 in dichloro-
methane/water with the use of a phase transfer catalyst
(Scheme 14). Sodium benzene sulfinate was an effective allyl
scavenger, affording free acid 38 as a crystalline solid after
acidification of the aqueous phase. While these conditions were
adequate and gave reproducibly high yields, several factors
urged us to reevaluate, including high cost and poor stability for
the Pd catalyst and the use of a chlorinated solvent.
Furthermore, the sulfinate was a relatively expensive stoichio-
metric reagent that required a phase transfer catalyst for
solubility in the organic phase. Optimization of these
parameters would not only affect the chemistry output, but
there were also questions around scalability, commercial
availability of reagents, and overall cost.
The most significant parameter for both scalability and cost
was the choice of Pd catalyst precursor. Typically, Pd(II)
catalyst precursors are preferred, owing to their increased
stability under standard handling conditions. Therefore, several
Pd(II) compounds, ligands, and additives were evaluated.51,52
After extensive screening, it was determined that a mixture of
Pd(OAc)2 and P(OEt)3 could serve as a viable replacement for
Pd(Ph3P)4 in terms of performance, economics, and stability.
The catalyst prepared in situ from these two components
proved effective for conversion of 8 to 38, leading to a less
expensive, more robust process.
An extensive screen of alternatives to the sodium benzene
sulfinate allyl scavenger was largely unsuccessful; most led to
either a significant decline in rate or a poor purity profile for
acid 38. Triethylammonium acetate proved suitable for the
deallylation but caused complications during workup. With
sodium benzene sulfinate as the best choice for allyl scavenger,
a survey of alternative solvents was carried out. It was
determined that THF/water was an excellent replacement for
the CH2Cl2/water combination. This miscible pair allowed for
deallylation of 8 without the need for a PTC. With this solvent
system change, the reaction conditions were further simplified
and made more environmentally friendly. Final conditions for
deallylation involved reaction of ester 8 with P(OEt)3,
Pd(OAc)2 (5 mol % catalyst), and sodium benzene sulfinate
in a mixture of THF and water (2:1) at 20−25 °C, which
provided acid 38 in 80−85% yield (Scheme 15).
Preparation of Sulopenem Acid 1 from 38. For the final
step of the synthesis of sulopenem acid 1, a desilylation is
performed on TBS ether 38. Adopting conditions gleaned from
1354
dx.doi.org/10.1021/op300131e | Org. Process Res. Dev. 2012, 16, 1348−1359

Organic Process Research & Development
Article
its salt form. However, simple acidification of the aqueous
phase failed to result in precipitation of acid 1, despite the
success of a related protocol in the first- and second-generation
routes. Presumably, tetrabutylammonium salts were inhibiting
the crystallization of sulopenem (1) from the aqueous solution.
After extensive screening, it was discovered that addition of
sodium benzenesulfinate to the aqueous solution, followed by
CH2Cl2 extractions, allowed for smooth crystallization of 1,
upon acidification of the aqueous mixture (e.g., H2SO4 or HCl).
This final process for workup and isolation of sulopenem (1)
provided API of very high quality in up to 82% yield. While the
role of sodium benzenesulfinate has not been definitively
established, we suspect that dynamic salt metathesis allows for
extraction of the more organic soluble tetrabutylammonium
benzenesulfinate into the CH2Cl2 layer.
Scheme 15. Improved deallylation conditions
the first-generation synthesis, tetrabutylammonium fluoride
(TBAF) in acetic acid was investigated for the reaction.9,11,53
Under these conditions, significant decomposition of both
starting material and product was observed. Therefore, other
fluoride reagents (e.g., aqueous HF, CsF, KF, Et3N·3HF) were
evaluated.33,54−59 These reagents also proved ineffective for the
reaction, showing poor product conversion or significant
substrate decomposition. After carrying out pH-dependent
stability studies on 38 and 1, it became apparent that the acetic
acid cosolvent was primarily responsible for the degradation
observed. Therefore, TBAF in THF was reevaluated in the
absence of AcOH and was found to be effective. Silyl ether 38
was cleanly converted to alcohol 1 at 20−25 °C, resulting in
excellent in situ yield (Scheme 16). Elevated temperatures
resulted in thermal degradation of the β-lactam. Therefore,
longer reaction times (∼18 h) near 25 °C proved optimal.
CONCLUSION
■
Following extensive route-scouting studies, a practical, cost-
effective, and convergent third-generation process to sulope-
nem (1) was identified and developed for potential commercial
application (Scheme 17). Conditions were discovered for
telescoping the first two steps of the synthesis, affording
oxalimide 21 in high yield from commercially available 4-
acetoxyazetidinone (9). The cyclization of oxalimide 21 to
penem 8 was accomplished in good yield by replacing a
phosphite reagent with a phosphonite, which enabled the
carbene cyclization to proceed in a more environmentally
friendly solvent. A highly regioselective oxidation procedure
was developed for the preparation of sulfoxide 26 from penem
17, utilizing UHP/HFIP. The practical utility of this oxidation
was enhanced by development of a recovery and recycle
process for the HFIP solvent. Due to the relative instability of
thiol 13, mild conditions were developed that enabled efficient
coupling of thiol 13 with sulfoxide 26 to give thiopenem 8 via a
telescoped process from penem 24. Additionally, an alternative
end-game process was identified for palladium-catalyzed
deallylation of ester 8, affording the penultimate penem acid
38. The reordered end game alleviated concerns over residual
palladium and triphenylphosphine oxide in the API, in addition
Scheme 16. Desilylation of TBS ether 38
The isolation of 1 was surprisingly complicated. Partition of
the crude reaction mixture between organic and aqueous phases
allowed the product to be separated from organic impurities as
Scheme 17. Final process scheme
1355
dx.doi.org/10.1021/op300131e | Org. Process Res. Dev. 2012, 16, 1348−1359

Organic Process Research & Development
Article
to providing a crystalline intermediate for purification. Finally,
desilylation conditions were developed, for the conversion of
38 to API, that addressed poor chemical stability and unique
solubility properties of 1, while obviating the need for
recrystallization. The process to sulopenem described herein
was intended for commercial application. Unfortunately, the
program was discontinued before the chemistry was performed
beyond laboratory scale.
3H). 13C NMR (100 MHz, CDCl3): δ 218.3, 163.4, 159.1,
154.4, 130.4, 120.2, 67.4, 66.1, 64.5, 58.9, 39.0, 25.6, 22.0, 21.3,
17.8, 13.4, −5.3, −4.3. HRMS (ESI) exact mass calcd for (M +
H) 492.1363; found: 492.1365.
(5R,6S)-Allyl 6-((R)-1-(tert-Butyldimethylsilyloxy)-
ethyl)-7-oxo-3-(propylthio)-4-thia-1-azabicyclo[3.2.0]-
hept-2-ene-2-carboxylate (24). A solution of toluene (3.0
L) and 21 (100 g, 0.20 mol) was heated to 90 °C under
nitrogen. Diethyl methyl phosphonite (30% w/w solution in
toluene; 230 mL, 0.46 mol) was added over 2 h, and the
solution was stirred for an additional 2 h (IPC target
specification <5% 21 according to HPLC). The reaction
mixture was cooled to 50 °C and concentrated to 1.5 L under
reduced pressure. The solution was further cooled to 20−25 °C
and sequentially washed with aqueous 0.1 N HCl (500 mL),
5% aqueous sodium bicarbonate (500 mL), and 23% aqueous
sodium chloride (500 mL). The organic layer was concentrated
under reduced pressure at 50 °C to 200 mL. Isopropyl alcohol
(500 mL) was added, and the mixture was concentrated under
reduced pressure at 50 °C to 150 mL. The product slurry was
cooled to 30 °C and stirred for 30 min. The solids were then
filtered, washed with cold isopropyl alcohol (150 mL), and
dried under vacuum at 20−25 °C to afford of penem 24 (54.1
g, 60%) as a yellow solid.
EXPERIMENTAL SECTION
■
General. (R)-4-Acetoxyazetidinone 9 and allyl oxalyl
chloride 19 were obtained from Kaneka Corporation and
Interchem Corporation, respectively. All reactions were
monitored by reverse phase liquid chromatography using an
Agilent 1100 series HPLC equipped with a Halo C18 (50 mm
× 4.6 mm, 2.7 μm) column utilizing an isocratic elution of 25%
acetonitrile and 75% aqueous methanesulfonic acid with a flow
rate of 1.0 mL/min and a column temperature of 30 °C.
Approximate retention times (min): 1 (0.45), 38 (3.4), 8 (6.2),
26 (8.4, 8.6), 24 (11.6) and 21 (12.0).
Allyl 2-((3S,4R)-3-((R)-1-(tert-Butyldimethylsilyloxy)-
ethyl)-2-oxo-4-(propylthiocarbonothioylthio)azetidin-1-
yl)-2-oxoacetate (21). To sodium hydride (60% w/w in
mineral oil, 28 g, 690 mol) was added MtBE (1.0 L), and the
resulting slurry was cooled to 20 °C. To the heterogeneous
reaction mixture, was added slowly at ≤25 °C a solution of 1-
propanethiol (34 g, 450 mmol) in MtBE (200 mL).
(CAUTION: hydrogen gas evolution!) The resulting slurry
was stirred at 20 °C for 1 h, and then carbon disulfide (53 g,
690 mmol) was added at ≤35 °C. The resulting yellow slurry
was stirred at 20 °C for 1 h, and then anhydrous magnesium
sulfate (6.0 g) was added; the reaction mixture was filtered
through a pad of dry diatomaceous earth under a nitrogen
atmosphere. The filter cake was rinsed with MtBE (100 mL).
The combined filtrates were transferred to a clean and dry
reactor to which 9 (100 g, 370 mmol) was added as a solid. The
resulting slurry was stirred at 20 °C for 1 h and then filtered
through a pad of silica gel under nitrogen atmosphere. The
filter cake was rinsed with MtBE (100 mL), and then the
combined filtrates containing 17 were cooled to −10 °C. Allyl
oxalyl chloride (19) was then added at a rate to ensure the
internal temperature was maintained below 0 °C. To the
resulting solution was added a solution of triethylamine in
MtBE at a rate to maintain temperature below 0 °C. After the
addition of triethylamine was complete, the resulting mixture
was warmed to 20 °C and stirred for 1 h. The reaction was
quenched with water (200 mL), and the layers were separated.
The organic layer was then washed with aqueous 10% sodium
bicarbonate solution (2× 200 mL), dried over magnesium
sulfate (6.0 g), and filtered. The filtrate was diluted with
heptanes (500 mL). The solution was then distilled at
atmospheric pressure to a final temperature of 93 °C to ensure
removal of MtBE. The resulting solution was cooled to 0 °C
and the product precipitated. The resulting slurry was stirred
for a minimum of 12 h, then was filtered and washed with
heptanes (200 mL), and dried under reduced pressure for a
minimum of 8 h to give 21 as an off-white solid (152 g, 89%).
1H NMR (400 MHz, CDCl3): δ 6.77 (d, 1H, J = 3.4 Hz),
6.00−5.92 (m, 1H), 5.41 (dd, J = 17.1, 1.2 Hz, 1H), 5.33 (dd, J
= 10.4, 1.1 Hz, 1H), 4.80−4.79 (m, 2H), 4.40 (dddd, J = 6.3,
6.3, 6.3, 2.4 Hz, 1H), 3.58 (t, J = 3.0 Hz, 1H), 3.45−3.36 (m,
2H), 1.78 (sxt, J = 7.4 (×5) Hz, 2H), 1.26 (d, J = 6.1 Hz, 3H,),
1.04 (t, J = 7.4, 7.4 Hz, 3H), 0.88 (s, 9H), 0.11 (s, 3H), 0.05 (s,
1H NMR (500 MHz, CDCl3): δ 5.99−5.91 (m, 1H), 5.61 (d,
J = 1.2 Hz, 1H), 5.42 (ddt, J = 17.2, 1.6, 1.6 Hz, 1H), 5.23 (ddt,
J = 10.5, 1.2, 1.2) Hz, 1H), 4.77−4.66 (m, 2H), 4.25 (dq, J =
6.3 (x4), 1H), 3.68 (dd, J = 5.1, 1.5 Hz, 1H), 3.00−2.87 (m,
2H), 1.75 (sxt, J = 7.2 Hz, 2H), 1.27 (d, J = 6.2 Hz, 3H), 1.05
(t, J = 7.2 Hz, 3H), 0.90 (s, 9H), 0.09 (s, 3H), 0.09 (s, 3H). 13
C
NMR (125 MHz, CDCl3): δ 172.1, 159.8, 155.5, 132.0, 118.1,
117.0, 71.5, 65.3, 63.7, 37.9, 25.7, 23.4, 22.5, 17.9, 13.2, −5.1,
−4.3. HRMS (ESI) exact mass calcd for (M + H)
C20H34NO4S2Si+ 444.1693; found: 444.1701.
(5R,6S)-Allyl 6-((R)-1-(tert-Butyldimethylsilyloxy)-
ethyl)-7-oxo-3-(propylsulfinyl)-4-thia-1-azabicyclo-
[3.2.0]hept-2-ene-2-carboxylate (26). Urea-hydrogen per-
oxide (5.75 g, 60 mmol) was added to penem 24 (25 g, 56
mmol) in hexafluoroisopropanol (HFIP) (125 mL) at 20−25
°C. The mixture was stirred for 12−14 h (IPC target
specification <7% 24 according to HPLC), and then heptanes
(38 mL) were added. The reaction mixture was distilled under
reduced pressure at 35 °C to remove HFIP. Solvent
displacement was continued by adding heptanes and
concentrating until the HFIP was removed. The distillate was
collected as two layers; the lower HFIP layer was separated for
reuse in subsequent batches. The reaction mixture was
concentrated to a slurry under reduced pressure to a final
volume of 100−125 mL. To this was added isopropyl acetate
(50 mL), and the heterogeneous mixture containing sulfoxide
26 (70−75% in situ yield according to HPLC analysis) was
carried into the next step without purification.
A portion of the crude product was purified by silica gel
column chromatography using ethyl acetate/heptane (30:70) as
eluent to afford an analytically pure sample of a mixture of
sulfoxide diastereomers 26.
1H NMR (700 MHz, CDCl3): δ 5.93−5.86 (m, 1H), 5.80 (d,
J = 1.4 Hz, 0.6H), 5.67 (d, J = 1.4 Hz, 0.4H), 5.43−5.38 (m,
1H), 5.27 (br d, J = 10.8 Hz, 1H), 4.73−4.64 (m, 2H), 4.26−
4.22 (m, 1H), 3.86 (dd, J = 4.0, 1.7 Hz, 0.4H), 3.84 (dd, J = 4.2,
1.7 Hz, 0.6H), 3.14 (ddd, J = 12.8, 9.2, 5.1 Hz, 0.6H), 3.09
(ddd, J = 12.6, 9.2, 5.1 Hz, 0.6H), 3.03−2.96 (m, 1H), 1.98−
1.84 (m, 2H), 1.24 (d, J = 6.3 Hz, 1.8H), 1.24 (d, J = 6.3 Hz,
1356
dx.doi.org/10.1021/op300131e | Org. Process Res. Dev. 2012, 16, 1348−1359

Organic Process Research & Development
Article
1.2H), 1.11 (t, J = 7.4 Hz 1.2H), 1.11 (t, J = 7.4 Hz 1.8H), 0.87
(s, 3.6H), 0.86 (s, 5.4H), 0.06 (s, 3.6H), 0.06 (s, 2.4H). 13C
NMR (176 MHz, CDCl3): δ 173.0, 172.0, 164.6, 164.2, 158.5,
158.4, 131.0, 130.0, 121.7, 119.3, 119.2, 73.9, 73.6, 66.6, 65.2,
65.0, 62.7, 57.6, 57.2, 27.1, 25.8, 22.5, 22.4, 18.1, 16.6, 16.3,
13.2, −4.1, −5.1. HRMS (ESI) exact mass calcd for (M + H)
C20H34NO5S2Si 460.1642 found: 460.1645.
1H), 3.10−3.07 (m, 1H), 3.01−2.97 (m, 1H), 2.84 (ddd, J =
12.7, 12.7, 6.5 Hz, 1H), 2.70 (ddd, J = 14.4, 5.7, 2.1 Hz, 1H),
2.66−2.62 (m, 1H), 2.41−2.36 (m, 1H), 1.18 (d, J = 6.5 Hz,
3H), 0.85 (s, 9H), 0.05 (s, 3H), 0.04 (s, 3H). 13C NMR (176
MHz, DMSO-d6): δ 172.8, 160.9, 151.3, 117.8, 70.7, 64.9, 63.7,
60.6, 52.2, 46.3, 33.4, 25.6, 21.8, 17.7, −4.5, 5.2. HRMS (ESI)
exact mass calcd for (M + Na); C18H29O5NNaS3Si 486.0869
found: 486.0875.
(3S)-3-({(5R,6S)-6-[(1R)-1-{[tert-Butyl(dimethyl)silyl]-
oxy}ethyl]-7-oxo-2-[(prop-2-en-1-yloxy)carbonyl]-4-
thia-1-azabicyclo[3.2.0]hept-2-en-3-yl}sulfanyl)-
tetrahydrothiophenium-1-olate (8). A solution of sodium
ethoxide in ethanol (35 mL, 90 mmol) was charged to a
solution of 10 (16.05 g, 90 mmol) in degassed isopropyl acetate
(110 mL) at 0−5 °C. The reaction was stirred for 30 min, and
then the pH was adjusted to 6−7 with acetic acid (5.4 mL) to
provide a solution of thiol 13. Triethylamine (13.8 mL, 10
mmol) was then charged, and the resulting suspension was
added to a crude solution of sulfoxide 26 (∼45 mmol) in
heptanes/isopropyl acetate (∼150 mL) at −20 to −30 °C over
1 h. The reaction was warmed to 0−5 °C and stirred for 1 h.
HPLC analysis of the reaction mixture indicated ∼1% sulfoxide
26 remained. The reaction mixture was then quenched by
addition of 1 M HCl (25 mL), and the layers were separated.
The organic layer was washed with saturated aqueous sodium
bicarbonate solution (100 mL) followed by saturated aqueous
sodium chloride (100 mL). The organic layer was then
concentrated under reduced pressure to a final volume of 25−
50 mL. Heptanes (100 mL) were added, and the slurry was
stirred at −20 °C for 30 min, filtered, washed with additional
cold heptanes (100 mL), and dried under vacuum at 25 °C for
16 h to provide 8 (15.6 g, 55%) as an off-white solid.
(3S)-3-({(5R, 6S)-2-Carboxy-6-[(1R)-1-hydroxyethyl]-7-
oxo-4-thia-1-azabicyclo[3.2.0]hept-2-en-3-yl}sulfanyl)-
tetrahydrothiophenium-1-olate (1). Tetrabutylammonium
fluoride (1 M solution in THF) (113 mL, 113 mmol) was
added to TBS ether 38 (35 g, 75.5 mmol) and THF (105 mL)
at 25 °C, and the mixture was stirred for 18 h under nitrogen.
After reaction completion was confirmed by HPLC analysis
(IPC target specification <2% 38), dichloromethane (245 mL)
and a solution of sodium benzene sulfinate (13.3 g, 79.3 mmol)
dissolved in water (245 mL) were added. The resulting biphasic
mixture was separated, and the aqueous layer was washed with
dichloromethane (2 × 87.5 mL). The aqueous layer was treated
with activated carbon (3.5 g) for 20 min at 20−25 °C and then
was filtered and washed with water (35 mL). The filtrates were
combined and cooled to 5 °C. A 1.0 M HCl solution was added
to obtain pH 2.5. The resulting slurry was stirred at 5 °C for 5
min, filtered, washed with water (35 mL), and dried under
reduced pressure for 24 h to afford sulopenem (1) as a white
solid (21.1 g, 72%).
1H NMR (700 MHz, DMSO-d6) 5.71 (br s, 1H), 5.21 (br s,
1H), 4.00−3.96 (m, 1H), 3.90−3.85 (m, 1H), 3.80 (dd, J = 6.0,
1.2 Hz, 1H), 3.76 (dd, J = 14.4, 8.9 Hz, 1H), 3.01−2.98 (m,
1H), 2.84 (ddd, J = 12.7, 12.7, 6.6 Hz, 1H), 2.72 (ddd, J = 14.4,
5.6, 1.9 Hz, 1H), 2.65 (ddt, J = 13.1, 6.5, 2.3 Hz, 1H), 2.39
(ddt, J = 12.6, 9.2, 6.4 Hz, 1H), 1.16 (d, J = 6.4 Hz, 3H). 13C
NMR (176 MHz, CDCl3) ppm 173.34, 160.86, 151.56, 117.75,
71.09, 64.31, 63.94, 60.53, 52.17, 46.26, 33.46, 21.51. HRMS
(ESI) exact mass calcd for C12H16O5NS3 (M + H) 350.0185,
found: 350.0188.
1H NMR (700 MHz, CDCl3) 5.92 (ddt, J = 17.3, 10.6, 5.4
Hz, 1H), 5.65 (d, J = 1.4 Hz, 1H), 5.39 (ddt, J = 17.2, 1.5 (x3)
Hz, 1H), 5.23 (ddt, J = 10.5, 1.4 (×3) Hz, 1H), 4.74−4.65 (m,
2H), 4.24 (dq, J = 6.3, 6.3, 6.3, 4.5, 1H), 3.96 (dd, J = 14.3, 8.4
Hz, 1H), 3.70 (dd, J = 4.7, 1.5 Hz, 1H), 3.64 (dt, J = 8.3, 8.3
Hz, 1H), 3.13−3.10 (m, 1H), 2.79 (ddd, J = 14.3, 8.7, 1.8 Hz,
1H), 2.74−2.64 (m, 3H), 1.24 (d, J = 6.3 Hz, 3H), 0.88 (s,
9H), 0.08 (s, 3H), 0.07 (s, 3H). 13C NMR (176 MHz, CDCl3)
ppm 172.02, 159.60, 150.96, 131.81, 118.63, 71.92, 65.87,
65.34, 64.21, 61.58, 52.87, 46.86, 33.32, 25.85, 22.67, 18.11,
4.13, 4.97. HRMS (ESI) exact mass calcd for C21H34NO5S3Si
(M + H) 504.1368, found: 504.1380.
ASSOCIATED CONTENT
■
S
* Supporting Information
1
Copies of H and 13C NMR spectra, X-ray crystal structure for
compound 24, and design of experiment parameter and result
information for the preparation of compound 24. This material
is available free of charge via the Internet at http://pubs.acs.org.
(3S)-3-({(5R,6S)-6-[(1R)-1-{[tert-Butyl(dimethyl)silyl]-
oxy}ethyl]-2-carboxy-7-oxo-4-thia-1-azabicyclo[3.2.0]-
hept-2-en-3 yl}sulfanyl)tetrahydrothiophenium-1-olate
(38). Pd(OAc)2 (4.46 g, 20 mmol), triethylphosphite (20.2 g,
120 mmol), THF (800 mL), and water (400 mL) were
combined and stirred at 20−25 °C for 10 min under nitrogen.
Solid penem 8 (200 g, 400 mmol) was added, followed by
sodium benzenesulfinate (73.2 g, 440 mmol). The reaction was
stirred for 4 h at 25 °C. After reaction completion was
confirmed (IPC target specification <1% 38 according to
HPLC), the mixture was cooled to 5 °C and 1 M aqueous HCl
(495 mL) was added over 60 min (pH 2.5). The slurry was
stirred at 0−5 °C for 30 min, and then filtered. The solids were
reslurried in THF (800 mL), filtered, washed with THF (400
mL) and dried under reduced pressure for 16 h to afford 38
(176.02 g, 95.6%, potency =85.8%, assay corrected yield
=82%).
AUTHOR INFORMATION
Corresponding Author
*durgesh.nadkarni@pfizer.com (D.V.N.); jade.nelson@pfizer.
com (J.D.N.).
■
Present Addresses
§Pfizer Global Supply, Kalamazoo, MI 49001.
⊥Pfizer Bioprocess Research & Development, Pearl River, NY
10965.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
We gratefully acknowledge Brian Jones for HRMS analysis, Bao
Nguyen and Silke Wunderwald for NMR analysis, the Pfizer
Chemical Research & Development and Analytical Research &
Development Management Teams and the R&D Team at IDT
Australia Limited (Melbourne, Australia) for valuable con-
1H NMR (700 MHz, DMSO-d6): δ 12.82 (br s, 1H), 5.70 (d,
J = 1.1 Hz, 1H), 4.20 (dq, J = 6.2, 6.0 Hz, 1H), 3.93 (dd, J =
4.2, 1.3 Hz, 1H), 3.89−3.84 (m, 1H), 3.77 (dd, J = 14.4, 8.9 Hz,
1357
dx.doi.org/10.1021/op300131e | Org. Process Res. Dev. 2012, 16, 1348−1359

Organic Process Research & Development
Article
(32) McCombie, S. W.; Ganguly, A. K.; Girijavallabhan, V. M.;
Jeffrey, P. D.; Lin, S.; Pinto, P.; McPhail, A. T. Tetrahedron Lett. 1981,
22, 3489.
tributions to the program and assistance in the preparation of
this manuscript.
(33) Hanessian, S.; Rozema, M. J. Bioorg. Med. Chem. Lett. 1994, 4,
2279.
(34) Planchenault, D.; Wisedale, R.; Gallagher, T.; Hales, N. J. J. Org.
Chem. 1997, 62, 3438.
(35) Leanza, W. J.; Christensen, B. G.; Dininno, F. P.; Ratcliffe, R. W.
6-Substituted-2-carbamimidoylpen-2-em-3-carboxylic acids, and their
antibiotic compositions. EP 83-101923 87792, 1983.
(36) Budt, K. H.; Duerckheimer, W.; Fischer, G.; Hoerlein, R.;
Kirrstetter, R.; Lattrell, R. Preparation of penems by cyclization of N-
alkoxalyl 4-acylthioazetidin-2-ines using dialkyl alkylphosphonites. DE
89-3914389 3914389, 1990.
REFERENCES
■
(1) Franceschi, G.; Perrone, E.; Alpegiani, M.; Bedeschi, A.; Della
Bruna, C.; Zarini, F. Spec. Publ. - R. Soc. Chem. 1989, 70, 222.
(2) Hamilton-Miller, J. M. T. Pharmacotherapy 2003, 23, 1497.
(3) McCombie, S. W.; Ganguly, A. K. Med. Res. Rev. 1988, 8, 393.
(4) Perrone, E.; Franceschi, G. Recent Prog. Chem. Synth. Antibiot.
1990, 613.
(5) Singh, G. S. Mini-Rev. Med. Chem. 2004, 4, 93.
(6) Singh, G. S. Mini-Rev. Med. Chem. 2004, 4, 69.
(7) Kosowska-Shick, K.; Ednie, L. M.; McGhee, P.; Appelbaum, P. C.
Antimicrob. Agents Chemother. 2009, 53, 2239.
(37) Budt, K. H.; Fischer, G.; Hoerlein, R.; Kirrstetter, R.; Lattrell, R.
Tetrahedron Lett. 1992, 33, 5331.
(8) Volkmann, R. A.; O’Neill, B. T. Strategies Tact. Org. Synth. 1991,
3, 495.
(38) The mechanism for formation of thiol 11 has not been
thoroughly investigated; however, the authors propose that non-
selective desulfurization of the epi-sulfide intermediate 39 may be
involved. S-ethyl penem 35 may result from alkylation of 11 by various
potential ethylating agents present in the reaction mixture
(9) See the preceding article for an extensive discussion on enabling
chemistry for this candidate. Brenek, S. J.; Caron, S.; Chisowa, E.;
ColonCruz, R.; Delude, M. P.; Drexler, M. T.; Handfield, R. E.; Jones,
B.; Nadkarni, D. V.; Nelson, J. D.; Olivier, M.; Weekly, R. M.;
Bellinger, G.; Brkic, Z.; Choi, N.; Desneves, J.; Lee, M.; Pearce, W.;
Watson, J. Org. Process Res. Dev. 2012, DOI: 10.1021/op300130p.
(10) Afonso, A.; Hon, F.; Weinstein, J.; Ganguly, A. K.; McPhail, A.
T. J. Am. Chem. Soc. 1982, 104, 6138.
(11) Volkmann, R. A.; Kelbaugh, P. R.; Nason, D. M.; Jasys, V. J. J.
Org. Chem. 1992, 57, 4352.
(12) Eschenmoser, A. Quart. Rev., Chem. Soc. 1970, 24, 366.
(13) Brighty, K. E., Preparation of 1-aza-4,5-dithiabicyclo[4.2.0]oct-2-
ene-8-oxo-2-carboxylate esters as intermediates for penems. US 86−
891493 4695626, 1987.
(14) Brighty, K. E.; Lindner, D. L., Preparation of allyl 2-
thioxopenam-3-carboxylates as intermediates for penem antibiotics.
EP 87-304669 249358, 1987.
(15) Leanza, W. J.; DiNinno, F.; Muthard, D. A.; Wilkening, R. R.;
Wildonger, K. J.; Ratcliffe, R. W.; Christensen, B. G. Tetrahedron 1983,
39, 2505.
(16) Sletzinger, M.; Liu, T.; Reamer, R. A.; Shinkai, I. Tetrahedron
Lett. 1980, 21, 4221.
(17) Sakurai, O.; Ogiku, T.; Takahashi, M.; Hayashi, M.; Yamanaka,
T.; Horikawa, H.; Iwasaki, T. J. Org. Chem. 1996, 61, 7889.
(18) Sakurai, O.; Ogiku, T.; Takahashi, M.; Horikawa, H.; Iwasaki, T.
Tetrahedron Lett. 1994, 35, 2187.
(19) High cost, toxicity, moisture sensitivity, and a need for cryogenic
reaction temperatures all add to the complexity of handling triflic
anhydride on large scale. Alternatives to triflic anhydride were not
evaluated due to high cost and/or bulk availibility.
(20) DiNinno, F.; Muthard, D. A.; Ratcliffe, R. W.; Christensen, B. G.
Tetrahedron Lett. 1982, 23, 3535.
(21) Kametani, T.; Fukumoto, K.; Ihara, M. Heterocycles 1982, 17,
463.
(22) Palomo, C. Recent Prog. Chem. Synth. Antibiot. 1990, 565.
(23) Sasaki, A.; Sunagawa, M. Chem. Heterocycl. Compd. 1999, 34,
1249.
(24) Girijavallabhan, V. M.; Ganguly, A. K.; Versace, R. W. 2-
(Heterocyclylalkylthio)penems. EP 84-114070 144041, 1985.
(25) Hamanaka, E. S. 2-Alkylthiopenem derivatives. EP 84-304083
130025, 1985.
(26) Phillips, D.; O’Neill, B. T. Tetrahedron Lett. 1990, 31, 3291.
(27) Wasserman, H. H.; Han, W. T. J. Am. Chem. Soc. 1985, 107,
1444.
(39) m-CPBA is an organic peroxide that is known to undergo
hazardous thermal decomposition near 60 °C.
(40) Adam, W.; Mitchell, C. M.; Saha-Moeller, C. R. Tetrahedron
1994, 50, 13121.
(41) Herrmann, W. A.; Rost, A. M. J.; Mitterpleininger, J. K. M.;
Szesni, N.; Sturm, S.; Fischer, R. W.; Kuehn, F. E. Angew. Chem., Int.
Ed. 2007, 46, 7301.
(42) Begue, J.-P.; Bonnet-Delpon, D.; Crousse, B. Synlett 2004, 18.
(43) Ravikumar, K. S.; Begue, J.-P.; Bonnet-Delpon, D. Tetrahedron
Lett. 1998, 39, 3141.
(44) Ravikumar, K. S.; Zhang, Y. M.; Begue, J.-P.; Bonnet-Delpon, D.
Eur. J. Org. Chem. 1998, 2937.
(45) According to HFIP manufacturer, DuPont Chemicals, “HFIP is
corrosive to eyes and may cause permanent eye injury on contact. It is also
corrosive to skin, and vapors can cause respiratory tract irritation. Care
should be taken to avoid contact by wearing appropriate gloves, safety
glasses and other protective equipment as needed. Refer to the MSDS
for further handling information. This product may also contain a very
low level of hexafluoroacetone, which is a potential reproductive
hazard”.
(46) Heydari, A.; Khaksar, S.; Tajbakhsh, M. Synthesis 2008, 3126.
(47) The reduced-pressure distillation was maintained below 35 °C
to avoid thermal degradation of sulfoxide 26.
(48) Honda, M.; Morita, H.; Nagakura, I. J. Org. Chem. 1997, 62,
8932.
(49) For current guidance on acceptable levels of residual metals in
active pharmaceutical ingredients, refer to the European Medicines
Agency Web site: http://www.ema.europa.eu/ema/.
(50) Trost, B. M. Acc. Chem. Res. 1980, 13, 385.
(28) Ishiguro, M.; Nakatsuka, T.; Tanaka, R.; Namikawa, K.; Matsuki,
S. Preparation of penem derivatives and antimicrobial agents
containing them. WO 96-JP366 9625417, 1996.
(29) Ishiguro, M.; Tanaka, R.; Namikawa, K.; Nasu, T.; Inoue, H.;
Nakatsuka, T.; Oyama, Y.; Imajo, S. J. Med. Chem. 1997, 40, 2126.
(30) Ghosez, L.; Marchand-Brynaert, J.; Vekemans, J.; Bogdan, S.;
Cossement, E. Tetrahedron 1983, 39, 2493.
(31) Ratcliffe, R. W.; Salzmann, T. N.; Christensen, B. G. Tetrahedron
Lett. 1980, 21, 31.
1358
dx.doi.org/10.1021/op300131e | Org. Process Res. Dev. 2012, 16, 1348−1359

Organic Process Research & Development
Article
(51) Huang, J.-P.; Chen, X.-X.; Gu, S.-X.; Zhao, L.; Chen, W.-X.;
Chen, F.-E. Org. Process Res. Dev. 2010, 14, 939.
(52) Gronquist, M. R.; Meinwald, J. J. Org. Chem. 2001, 66, 1075.
(53) Alpegiani, M.; Bedeschi, A.; Perrone, E.; Franceschi, G.
Tetrahedron Lett. 1984, 25, 4171.
(54) Biondi, S.; Pecunioso, A.; Busi, F.; Contini, S. A.; Donati, D.;
Maffeis, M.; Pizzi, D. A.; Rossi, L.; Rossi, T.; Sabbatini, F. M.
Tetrahedron 2000, 56, 5649.
(55) Cabri, W.; Zarini, F.; D’Anello, M.; Marchi, M.; Bedeschi, A.;
Franceschi, G. Tetrahedron Lett. 1993, 34, 3491.
(56) Hanessian, S.; Desilets, D.; Bennani, Y. L. J. Org. Chem. 1990,
55, 3098.
(57) Hu, X. E.; Kim, N. K.; Grinius, L.; Morris, C. M.; Wallace, C. D.;
Mieling, G. E.; Demuth, T. P., Jr. Synthesis 2003, 1732.
(58) King, S. A.; Pipik, B.; Thompson, A. S.; DeCamp, A.;
Verhoeven, T. R. Tetrahedron Lett. 1995, 36, 4563.
(59) Lee, K. S.; Lim, J. H.; Kang, Y. K.; Yoo, K. H.; Kim, D. C.; Shin,
K. J.; Kim, D. J. Eur. J. Med. Chem. 2006, 41, 1347.
1359
dx.doi.org/10.1021/op300131e | Org. Process Res. Dev. 2012, 16, 1348−1359