SOB as an alternate to BOB: Findings from the preparation of injectable antifungal Sch 59884

Article abstract of DOI:10.1016/S0040-4039(01)01485-X

A facile preparation of 4-silyloxybutyrates (SOB) and their potential use as an alternate to 4-benzyloxybutyrate (BOB) are described.

Full text of DOI:10.1016/S0040-4039(01)01485-X

TETRAHEDRON
LETTERS
Pergamon
Tetrahedron Letters 42 (2001) 7141–7143
SOB as an alternate to BOB: findings from the preparation of
injectable antifungal Sch 59884
P. Renton, D. Gala* and G. M. Lee
Chemical Process Research & Development, Schering-Plough Research Institute, 1011 Morris Avenue, Union, NJ 07083, USA
Received 27 June 2001; accepted 8 August 2001
Abstract—A facile preparation of 4-silyloxybutyrates (SOB) and their potential use as an alternate to 4-benzyloxybutyrate (BOB)
are described. © 2001 Elsevier Science Ltd. All rights reserved.
Recently, Professor Ganem has described the develop-
ment of BOB as a synthetically useful protecting group
towards the synthesis of glycolipids such as keruffarides
and crasserides.1 The use of BOB via Jacobson asym-
metric epoxide opening2 allowed for the preparation of
synthetically useful diol monoesters. After appropriate
manipulation of the free hydroxy group, BOB can be
removed in a stepwise fashion to free the protected
hydroxy group. This report prompts us to describe our
findings towards the preparation of BOB and SOB and
their synthetic utility. The latter can be an attractive
alternate to BOB for larger scale synthesis.
which entrained the intermediates making their isola-
tion difficult with lowered yields upon scale-up.
To overcome the above difficulties, alternate synthetic
strategies were considered. In one of our strategies we
reevaluated the use of BOB, i.e. 4-benzyloxybutyric
acid, 5.4 Compound 5 can be reacted with 2 to generate
ester 6 (Scheme 1). This after debenzylation to alcohol
7 can be directly phosphorylated with a pentavalent
phosphorylating reagent such as dibenzylchlorophos-
phate (DBCP) to the known precursor of 1.3 The
literature conditions,5 described in Scheme 1, allowed
for the preparation of 5 from commercially available
g-lactone 3. The reaction between 3 and excess benzyl-
bromide generated few byproducts and 5 along with
compound 4. This mixture was then hydrolyzed to 5
which was isolated via acid/base work-up. Our previ-
ously described mixed anhydride procedure,2,6 depicted
below Scheme 1, worked well for the coupling of 5 with
2 to form the benzylether 6. The debenzylation of 6
with hydrogen in a variety of solvents, with or without
acid, required long reaction times (1–3 days) even in the
presence of a large (]50 wt%) amount of 5% Pd/C.
Furthermore, debenzylation generated several impuri-
ties which could only be removed via chromatography.
This difficulty in deprotection of 6 does not appear to
be specific to 6 as most of the BOB protected alcohols
reported by Ganem also required large amounts (50–
100 wt%) of Pd/C catalyst. Transfer hydrogenation
with formic acid contaminated 7 with the formate ester
of 7 also requiring chromatographic purification. The
reactions with an excess of ammonium formate were
very slow (incomplete after 7 days, 75 wt% 5% Pd/C).
For a large-scale synthesis of 1, the long reaction times
for the preparation of 5 as well as 7, excesses of BnBr
Our synthesis of the injectable antifungal Sch 59884, 1,
from Posaconazole (Sch 56592), 2, via the p-nitrobenzyl
ester of 4-hydroxybutyric acid is being published.3
Although this synthesis was scaled up to prepare sev-
eral kilos of the antifugal, the p-nitrobenzyl was far
from a desirable protecting group. Not only did it
involve the use of irritant p-nitrobenzyl halide, its
deprotection with either Na2S or Zn/HCl was problem-
atic. The sulfide method required rigorous emission
controls and led to decomposition of intermediates,
whereas the zinc method formed polymeric byproducts
* Corresponding author.
0040-4039/01/$ - see front matter © 2001 Elsevier Science Ltd. All rights reserved.
PII: S0040-4039(01)01485-X

7142
P. Renton et al. / Tetrahedron Letters 42 (2001) 7141–7143
Scheme 1. Preparation of 1 via the use of BOB (5): (a) 4 equiv. BnBr, KOH, tol. reflux, 3 days; (b) tol., aq. NaOH, reflux, 1 day;
85% from 3; (c) TsCl, DMAP, THF, 0°C as in Refs. 3 and 6; 90%; (d) 100 wt% 5% Pd/C, H2 100 psi, THF, 24 h 100 psi, 75%
soln; 60% isolated after chromatography;7 (e) Refs. 3 and 10.
and Pd/C, and chromatographic purifications for the
removal of impurities formed in theses reactions were
undesirable.
This forced evaluation of several alternate conditions
for the conversion of 10 to 7, of which the preferred
one entailing the use of dil. aq. HCl leading to good
isolated yield (85%) is depicted in Scheme 2.
The above difficulties were overcome by the use of SOB
9, in place of BOB, 5, as described in Scheme 2. It was
observed in our laboratory that under very mild condi-
tions the reaction of commercially available 4-hydroxy
sodium butyrate 8 with a variety of silylchlorides led
directly to the formation of 4-silyloxybutyric acids
(SOB, 9) in good to excellent isolated yield (70–90%).
To the best of our knowledge the formation of 9
proceeds via an intramolecular carboxy to a hydroxy
silyl transfer reaction which is novel. This new finding
was then applied to the synthesis of 1. For the purpose
of synthesizing 1, we chose TBDMS as the protecting
group of choice.8 Once again, the mixed anhydride
coupling procedure3,6 was well suited for the conversion
of 9 to 10, and in fact it led to an excellent isolated
yield of 10, as shown in Scheme 2. The conversion of 7
from 10 with fluorides generated a minimum amount of
the desired product and its implications are listed later.
The removal of the silyl group from 10 with TBAF
(tetrabutylammoniumfluoride) was investigated first for
the conversion of 10 to 7. Even under mild conditions
(−10 to 0°C) and with limited amount of this reagent,
the formation 7 was immediately followed by the for-
mation of 2 and g-lactone. Interestingly, this facile
desilylation and elimination overcomes difficulties asso-
ciated with the removal of BOB. Thus, a facile prepara-
tion, subsequent easy deprotection, and apparent
compatibility of SOB to mild Jacobsen asymmetric
epoxide opening conditions should allow for the use of
SOB in epoxide opening leading to an alternate to BOB
for synthetic purposes.
A detailed description of the intramolecular carboxyl to
hydroxyl group transfer in 8, identification of the inter-
Scheme 2. Preparation of 1 via use of SOB (9): (c) 2, TsCl, DMAP, THF, 0°C to rt, 24 h; 98%; (e) Refs. 3 and 10; (f) DMF, rt,
8–20 h, 70–90% yield; (g) 5% aq. HCl, THF, 0–22°C, 85%; (h) 1.1 equiv. TBAF, THF, rt, 75% unoptimized;9 (e) as in Scheme
1.

P. Renton et al. / Tetrahedron Letters 42 (2001) 7141–7143
7143
mediates involved in this transfer via modern spectral
techniques, preparation of 9, as well as development of
an efficient alternative (to DBCP) phosphorylation
method for the conversion of 7 to 1 is the subject of a
separate publication.10
5. (a) Szczepankiewicz, B. G.; Heathcock, C. H. Tetra-
hedron 1997, 53, 8853; (b) Hoye, T. R.; Murth, M. J.; Lo,
V. Tetrahedron Lett. 1981, 22, 815.
6. Andrews, D. R.; Gala, D.; Gosteli, J.; Gunter, F.;
Mergelsberg, I.; Sudhakar, A. US Patent 5,625,064; Pro-
cess for the Preparation of Triazolones, April 29, 1997.
7. A number of byproducts were seen via HPLC. The large
excess of catalyst adsorbed product as well as starting
materials. The use of methylenechloride was necessary for
sampling as well as obtaining good mass balance from
the catalyst.
In summary, the preparation of SOB via a novel silyl
transfer reaction is described. The preparation and
removal of SOB is accomplished under mild conditions.
SOB may provide an attractive alternate to the use of
BOB for synthetic purposes.
8. Of the three silyl groups evaluated, TBDMS was com-
mercially less expensive than the others, and had all the
desirable properties for the synthesis of 1, hence it was
the group of choice. Although TMS-Cl is readily avail-
able and less expensive than TBDMS-Cl, the former was
expected to be unstable as a carboxylic acid protecting
group. Hence, TMS was not considered as a suitable silyl
protecting reagent for a large scale preparation of 1.
9. Compound 7 accounted for the rest of the mass balance.
In theory, extended reaction time or slight excess of
TBAF could lead to complete conversion to 2, which was
not our interest.
10. Renton, P.; Shen, L.; Eckert, J.; Lee, G. M.; Gala, D.;
Chen, G.; Pramanik, B.; Schumacher, D. Na-GHBA: An
Intramolecular Carboxy to Hydroxy Silyl Transfer and
its Application to the Synthesis of Injectable Antifungal
Posaconazole Derivative Sch 59884.
References
1. Clark, M. A.; Ganem, B. Tetrahedron Lett. 1999, 41,
9523.
2. Jacobsen, E. N.; Kakiuchi, F.; Konsler, R. G.; Larow, J.
F.; Tokunaga, M. Tetrahedron Lett. 1997, 38, 773.
3. Lee, G. M.; Gala, D.; Eckert, J.; Schwartz, M.; Renton,
P.; Pergamen, E.; Whittington, M.; Schumacher, D.; Hei-
mark, L.; Shipkova, P. Synthesis of Injectable Antifungal
Sch 59884, OPRD, submitted.
4. Bannett, F.; Girijavallabhan, V.; Patel, N. M.; Saksena,
A.; Ganguly, A. US Patent 6, 043, 245; Tetrahydrofuran
Antifungal Phosphate, March 28, 2000. Here phsophoryl-
ation was carried out with a trivalent phosphorous
reagent. The intermediate thus formed was then oxidized
with a peroxide to the phosphate.