6484
K. F. Albizati et al. / Tetrahedron Letters 42 (2001) 6481–6485
as the exclusive product as a mixture of diastereomers.
One of these diastereomers has been completely charac-
5. For a recent review on cyclic sulfate chemistry, see: Byun,
H.; He, L.; Bittman, R. Tetrahedron 2000, 56, 7051–7091.
6. Mash, E. A.; Nelson, K. A.; Van Deusen, S.; Hemperly,
S. B. Org. Synth. 1990, 68, 92–103.
1
terized by H NMR.16 When the TFA salt of 24 was
subjected to the thiophenol reaction conditions, a 27:73
ratio of 22 to 23 was obtained, showing that quater-
nization is a pathway that leads to the undesired iso-
mer. Again, no products arising from nucleophilic
attack on the six-membered ring were observed.
7. Takano, S.; Kurotaki, A.; Sekiguchi, Y.; Satoh, S.;
Hirama, M.; Ogasawara, K. Synthesis 1986, 811–817.
8. All new compounds exhibited satisfactory spectral (300
1
MHz H NMR, 75 MHz 13C NMR, IR), and high-reso-
lution mass spectral characteristics. Spectroscopic data
The reaction mixture containing 22 and 23 was treated
with ethanolamine at reflux to cleave the phthalimide.
After an aqueous workup, an MTBE solution of the
free amine was treated with benzoic acid and the ben-
zoate salt 25 crystallized from the mixture upon dilu-
tion with hexanes in 71% yield. This salt formation
gave us a very good purification opportunity for the
nelfinavir synthesis, being the penultimate intermediate.
The product formed from ethanolamine cleavage of the
undesired isomer 23 was completely removed by this
crystallization. Conversion of 25 to the freebase,
acylation with the acid chloride 17, and hydrolysis of
the phenolic acetate afforded nefinavir 1 in 79% yield
after two acetone/water reslurries. The nelfinavir passed
all purity specifications. The above sequence has been
demonstrated on a multi-kilogram scale.
for 5: mp 103°C; [h]D 54.7 (c=1.0, MeOH); IR (KBr)
1
wmax 1397, 1211, 1179, 814 cm−1; H NMR (DMSO-d6) l
7.80 (4H, d, J=8.4 Hz), 7.51 (4H, d, J=8.4 Hz), 5.24
(2H, m), 4.48 (4H, m), 2.43 (6H, s); 13C NMR (CDCl3) l
146.5, 131.8, 130.7, 128.54, 78.8, 65.7, 22.2; HRMS calcd
for C18H21O10S3 493.0297, found 493.0294.
9. Kim and co-workers have reported the desymmetrization
of a 1,4-dichloro-cyclic sulfate with azide ion, see: Kim,
B. M.; Bae, S. J.; Seomoon, G. Tetrahedron Lett. 1998,
39, 6921–6922. For other work concerning the desym-
metrization of cyclic sulfates, see: (a) Yun, G.; Sharpless,
K. B. J. Am. Chem. Soc. 1988, 110, 7538–7539; (b) Kim,
B. M.; Sharpless, K. B. Tetrahedron Lett. 1989, 30,
655–658; (c) Lohray, B. B.; Gao, Y.; Sharpless, K. B.
Tetrahedron Lett. 1989, 30, 2623–2626.
10. Kim and Sharpless have reported that only one equiva-
lent of water is sufficient for these kinds of hydrolyses:
Kim, B. M.; Sharpless, K. B. Tetrahedron Lett. 1989, 30,
655–658.
In conclusion, we have developed an effective synthesis
of nelfinavir utilizing a key nucleophilic opening of a
cyclic sulfate. The cyclic sulfate is readily derived from
11. Spectroscopic data for 13: mp 82°C; [h]D 24.2 (c=1.0,
1
D
-tartaric acid through a series of simple transforma-
MeOH); IR (KBr) wmax 3580, 2092, 1360 cm−1; H NMR
tions in high yield. We have also replaced the use of
hazardous sodium azide through the use of potassium
phthalimide. Additionally, the potassium phthalimide
nucleophile served a dual role as protecting group and
oxazoline precursor.
(DMSO-d6) 7.79 (2H, d, J=8.1 Hz), 7.76 (2H, d, J=8.1
Hz), 7.50 (2H, d, J=8.1 Hz), 7.48 (2H, d, J=8.1 Hz),
5.75 (1H, br s), 4.28 (1H, dd, J=2.8, 10.7 Hz), 4.08 (1H,
dd, J=7.2, 10.5 Hz), 4.02 (1H, dd, J=3.3, 10.5 Hz), 3.94
(1H, dd, J=5.4, 10.5 Hz), 3.76 (1H, ddd, J=3.0, 7.2, 7.5
Hz), 3.63 (1H, ddd, J=3.0, 5.1, 7.8 Hz); 13C NMR
(DMSO-d6) l 145.7, 145.4, 132.3, 132.1, 130.6, 130.5,
128.0, 71.7, 69.8, 67.4, 60.8, 21.5; HRMS calcd for
C18H22N3O7S2 456.0899, found 456.0893.
Acknowledgements
12. Available from Austin Chemical Company, Inc. 1565
Barclay Blvd. Buffalo Grove, IL 60089, USA.
The authors wish to thank David Stirling and Chris
Albizati for analytical support and Mark Wilson for
help in sourcing materials. The authors also wish to
thank Dave Kucera and Trevor Laird for helpful
discussions.
13. Available from the Aldrich Chemical Company, Inc.
14. Weigel and Astleford have used the pyrrolidine adduct to
attenuate the reactivity of the phthalimide group. Astle-
ford, B.; Weigel, L. O. Tetrahedron Lett. 1991, 32, 3301–
3304.
15. Inaba and co-workers (Ref. 4) have also proposed a
spirocyclic ammonium intermediate as the source of a
related regioisomer.
References
16. NMR experiments (1H NMR, DQF-COSY, HMQC and
HMBC) were acquired at 500 MHz on the mixture of
diastereomers in DMSO-d6 at 30°C. These data clearly
established the presence of the quaternary spirocycle for a
single diastereomer in the mixture. The DQF COSY
1. Kaldor, S. W.; Kalish, V. J.; Davies, II, J. F.; Shetty, B.
V.; Fritz, J. E.; Appelt, K.; Burgess, J. A.; Campanale, K.
M.; Chirgadze, N. Y.; Clawson, D. K.; Dressman, B. A.;
Hatch, S. D.; Khalil, D. A.; Kosa, M. B.; Lubbehusen, P.
P.; Muesing, M. A.; Patick, A. K.; Reich, S. H.; Su, K.
S.; Tatlock, J. H. J. Med. Chem. 1997, 40, 3979–3985.
2. Gawronski, J.; Gawronski, G. Tartaric and Malic Acids
in Synthesis; Wiley-Interscience: New York, 1999.
3. Toray Industries, Inc. 2-1, Nihonbashi-Muromachi 2-
chome, Chuo-ku, Tokyo 103, Japan.
1
defined the H spin system involving the benzamide pro-
ton NHa and protons attached to carbons C1–C4, along
with two additional spin systems assigned to protons at
C5–C6 and C12–C13. The HMQC and HMBC experi-
ments provided unambiguous proton and carbon reso-
nance assignments for NHa, C1–C7, C12–C13, C19, and
the t-butylamide nuclei. Over 20 long range 1H–13C
correlations involving these nuclei were observed, includ-
ing key correlations from the protons at C1 to carbons
4. Inaba, T.; Birchler, A. G.; Yamada, Y.; Sagawa, S.;
Yokota, K.; Ando, K.; Uchida, I. J. Org. Chem. 1998, 63,
7582–7583.