Highly regioselective ring opening of epoxides using NaN3: A short and efficient synthesis of (-)-cytoxazone

Article abstract of DOI:10.1016/j.tetlet.2004.07.157

A convenient and efficient synthesis of 1,2-azido alcohols has been achieved by regioselective ring opening of epoxides using NaN₃ and 4 ? molecular sieves in acetonitrile. The utility of this method has been demonstrated by achieving a short synthesis of (-)-cytoxazone in 48% overall yield.

Full text of DOI:10.1016/j.tetlet.2004.07.157

Tetrahedron
Letters
Tetrahedron Letters 45 (2004) 7355–7358
Highly regioselective ring opening of epoxides using NaN3:
a short and efficient synthesis of (ꢀ)-cytoxazone
Joshodeep Boruwa, Jagat C. Borah, Biswajit Kalita and Nabin C. Barua*
Natural Products Chemistry Division, Regional Research Laboratory (CSIR), Jorhat 785006, Assam, India
Received 14 June 2004; revised 20 July 2004; accepted 30 July 2004
Available online 20 August 2004
Abstract—A convenient and efficient synthesis of 1,2-azido alcohols has been achieved by regioselective ring opening of epoxides
˚
using NaN3 and 4A molecular sieves in acetonitrile. The utility of this method has been demonstrated by achieving a short synthesis
of (ꢀ)-cytoxazone in 48% overall yield.
ꢀ 2004 Elsevier Ltd. All rights reserved.
1,2-Azidoalcohols are versatile intermediates in organic
synthesis, since they are important precursors of b-ami-
noalcohols and vicinal diamines, which are present in
numerous natural products.1 Furthermore, their impor-
tance is significant in the chemistry of carbohydrates
and nucleosides.2 Since the discovery of Sharpless asym-
metric epoxidation3 and asymmetric dihydroxylation4
the ring opening reactions of epoxides have been exten-
sively studied in the context of the chiral synthesis of
biologically active natural products.
as long reaction times, strongly acidic conditions, lack of
regioselectivity, high cost of reagent, difficulty in work-
up and lack of reusability of a catalyst. Therefore a bet-
ter catalyst that would be superior to the existing ones,
especially with regard to recyclability and selectivity, is
still required.
In recent years, organic reactions on solid supports such
as zeolites or mesoporous molecular sieves9 have at-
tracted attention because of the advantages these cata-
lysts have, such as high purity of the product, easy
work-up procedure, reusability and the environment-
friendly nature of the catalyst. In continuation of our
interest in the chemistry of epoxides and their applica-
tion in the synthesis of bioactive natural products,10
we found that a variety of epoxides can be opened in
Generally, azidohydrins are prepared through ring
opening of epoxides using different azides in suitable sol-
vents. The classical method uses5 sodium azide and
ammonium chloride but it requires a long reaction time
(12–48h) and is often accompanied by isomerization,
epimerization and rearrangement of the products.6a
Hence, several modifications have been made, which in-
clude the combined use of NaN3 or TMSN3 and a Lewis
acid or transition metal complex7 or reaction in the pres-
ence of tetrabutylammonium salts.8 Recently, pH-con-
trolled azidolysis of epoxides was also studied in water
using various Lewis acids and NaN3.6
˚
a regioselective manner with NaN3 using 4A molecular
sieves as the catalyst (Table 1).
It is evident from Table 1 that in the case of terminal
epoxides such as 1, 2, 4, 5 and 9, the attack by azide
nucleophile takes place exclusively at the less hindered
position, whereas in the case of a-phenyl epoxides such
as 3, 6, 7 and 8, the attack takes place predominantly at
the benzylic position, although trace amounts of the
other regioisomer were detected in the case of epoxides
6, 7 and 8.
Although these methods have their own advantages,
many of them suffer from one or more drawbacks such
In order to compare the regioselectivity of this ring
opening reaction, a comparative study was carried out
with the classical method. Our findings are shown in
Table 2. The high regioselectivity of our catalyst is evi-
dent from the yields of the individual regioisomers.
Keywords: Molecular sieves; Azidoalcohol; Regioselectivity;
(ꢀ)-Cytoxazone.
*
Corresponding author. Tel.: +91 376 2370121; fax: +91 376 237
0011; e-mail: ncbarua2000@yahoo.co.in
0040-4039/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2004.07.157

7356
J. Boruwa et al. / Tetrahedron Letters 45 (2004) 7355–7358
˚
Table 1. Regioselective ring opening of epoxides using NaN3/CH3CN/4A MS
Entry
1
Epoxide
Product
Time (h)
1.097
Yielda (%)
O
OH
Cl
N3
Cl
1
1a
OH
O
OBn
N3
OBn
2
3
4
5
1.2
96
2
2a
OH
O
1.096
1.3
N3
3
3a
4a
OH
OH
OH
O
92
90
N3
4
OH
OH
OH
O
1.5
N3
Cl
Cl
5a
5
N3
O
COOEt
O
COOEt
6
7
1.5
1.6
94 (3)
95 (2)
OH
6
N3
6a
COOEt
COOEt
OH
MeO
MeO
7
7a
N3
O
OH
OH
8
9
1.2
1.2
95 (2)
OH
8
8a
O
O
92
OH
O
9
9a N3
a Yields of pure isolated product. Yields in the parentheses correspond to the yield of the other regioisomer.
˚
The reason for high regioselectivity exhibited by 4A
molecular sieve in the epoxide opening may be attribu-
ted to potassium ion, which is the cation constituent
reported that (ꢀ)-cytoxazone exhibits cytokine modulat-
ing activity by inhibiting the signalling pathway of Th2
cells13 and because of interesting biological activity, this
molecule has been the target of various recent synthetic
efforts.14 Sunjic and co-workers synthesized this mole-
cule in the optically pure form using a strategy involving
epoxide ring opening with NaN3. However in their syn-
thesis the yield of the desired regioisomer of the azido
alcohol was only 56% and to obtain the final product
they employed a kinetic resolution step using Candida
antarctica lipase (CAL).14b
11
˚
of 4A molecular sieves. Finally we turned our atten-
tion towards the possible recycling of the catalyst. The
catalyst was filtered off, washed with acetone, dried
and reused. In these experiments products were obtained
in almost the same yields and regioselectivity.
Seeing the near exclusive formation of the regioisomer
7a (Table 2, entry 4), we decided to employ this method
for a synthesis of (ꢀ)-cytoxazone 15, a natural product
possessing cytokine modulating activity isolated from
Streptomyces sp. by Osada and co-workers.12 It is
We started the synthesis of (4R,5R)-cytoxazone from
ethyl p-methoxycinnamate 10, which was prepared on

J. Boruwa et al. / Tetrahedron Letters 45 (2004) 7355–7358
7357
Table 2. Comparative results of regioselectivity
as a gum in 96% yield and 99% ee without epimerization
at the C-2 centre, which is a serious problem with several
other base/solvent combinations.18 Azidolysis of this
Entry
1
Epoxide
Product
Conditionsa
Yieldb (%)
2
2a
A
B
A
B
A
B
A
B
A
B
96
˚
epoxide was performed with NaN3 and 4A molecu-
88 (2)
95 (2)
76 (12)
94 (3)
78 (13)
95 (2)
79 (10)
95 (2)
83 (11)
lar sieves in acetonitrile to obtain the azido alcohol
(2S,3R)-14 in 95% yield and 98% ee. The carbocationic
character in the transition state needed for regioselective
attack by the nucleophile, at the benzylic position is
encouraged by the electron donating effect of the p-
methoxy group.19 Conversion of (2S,3R)-14 to (ꢀ)-cyt-
oxazone was achieved following the procedure described
by Nakata and co-workers14a giving (ꢀ)-cytoxazone in
48% in overall yield, the spectral data of which were
identical to those reported.12
2
3
4
5
3
6
7
8
3a
6a
7a
8a
a
˚
A: 1.5equiv NaN3, 4A molecular sieves, CH2Cl2. B: 2.2equiv NaN3,
5equiv NH4Cl, MeOH/H2O (8:1) reflux.
b Yields in parentheses are of the other regioisomer.
In summary we have described a novel procedure for the
regioselective ring opening of epoxides, which does not
require tedious preparation of the catalyst or harsh con-
ditions like other solid-supported catalysts employed for
this purpose. The reusability of this catalyst, high degree
of regioselectivity, mild reaction conditions, low cost of
the reagent and easy work-up procedure make it an
attractive methodology. Further we have described an
efficient synthesis of (ꢀ)-cytoxazone using this method-
ology as one of the steps.
a large scale in 82% yield by Wittig olefination of p-anis-
aldehyde with triphenyl carbethoxymethyl phospho-
nium chloride supported on basic alumina under
microwave irradiation (Scheme 1).
Instead of resorting to a kinetic resolution step, we
decided to construct the epoxide with the desired chiral-
ity at this stage. a,b-Unsaturated esters are very good
substrates for asymmetric dihydroxylation resulting in
excellent enantiomeric excess. Thus when compound
10 when subjected to ADH4 using (DHQD)2PHAL, diol
(2S,3R)-11 was isolated in 94% yield and 99% ee15 as a
solid, which was treated with p-toluenesulfonyl chloride
and triethylamine in dichloromethane16 to give the
monotosylated product (2S,3R)-12 as a crystalline solid
in 95% yield and 98% ee. The next step in the synthesis
was the formation of the epoxide (2R,3S)-13, which was
effected by treating (2S,3R)-12 with K2CO3 in metha-
nol17 whereupon the epoxide (2R,3S)-13 was obtained
Spectral data of some selected compounds: 2a. IR (chlo-
roform, cmꢀ1): 3420, 2917, 2866, 2101, 1496, 1454,
1279, 1099, 738, 698. 1H NMR (300MHz, CDCl3):
d = 7.21 (5H), 4.43 (s, 2H), 3.81 (m, 1H), 3.45 (d,
J = 4Hz, 2H), 2.83 (br s). MS (ESI) m/z = 230.1
(M++Na). Compound 6a. IR (chloroform, cmꢀ1):
3462, 2983, 2107, 1736, 1736, 1637, 1311, 1256, 1069,
1
769, 702. H NMR (300MHz, CDCl3) d = 7.29 (5H),
4.85 (d, J = 2.7Hz, 1H), 4.21 (m, 1H), 4.11 (q,
J = 4.4Hz, 2H), 1.37 (t, J = 7.1Hz, 3H), 2.11 (br s).
OH
OH
3
COOEt
COOEt
b
COOEt
a
2
OTs
94%
95%
OH
MeO
MeO
MeO
(2S,3R) 12
(2S,3R) 11
10
N3
O
COOEt
e
COOEt
d
c
67%
95%
OH
95%
MeO
MeO
(2S,3R) 14
(2R,3S) 13
O
HN
O
4
5
OH
MeO
(4R,5R) 15
Scheme 1. Reagents and conditions: (a) (DHQD)2PHAL, OsO4, t-BuOH/H2O (1:1), 0ꢁC 16h; (b) TsCl, Et3N, CH2Cl2, 5ꢁC, 72h; (c) K2CO3, MeOH,
˚
rt, 12h; (d) NaN3, 4A molecular sieves, CH3CN, rt, 1.6h; (e) (i) PhCOCl, pyridine, CH2Cl2, (ii) PPh3, THF/H2O, 50ꢁC, (iii) NaBH4, MeOH, 0ꢁC
(67% in three steps).

7358
J. Boruwa et al. / Tetrahedron Letters 45 (2004) 7355–7358
MS (ESI) m/z = 258.0(M ++Na). Compound 8a. IR
(chloroform, cmꢀ1): 3408, 2931, 2103, 1443, 1279,
1104, 944, 873, 771. 1H NMR (300MHz, CDCl3)
d = 7.21 (5H), 4.22 (d, J = 6Hz, 1H), 3.59 (m, 1H),
3.28 (d, J = 2.6Hz, 1H), 2.08 (br s, 1H). MS (ESI)
chromatography over silica gel using ethylacetate/petro-
leum ether as eluent.
Acknowledgements
+
m/z = 216.0(M +Na).
The authors are thankful to Director R. R. L. Jorhat for
providing facilities.
20
(2S,3R)-11: Solid; mp: 72–73ꢁC; ½aŠ ꢀ6:1 (c 1 metha-
D
nol); ee: 99%. IR (chloroform, cmꢀ1): 3449, 2982,
2937, 2838, 1735, 1613, 1514, 1465, 1389, 1370, 1302,
1249, 1178, 1031, 834, 771. 1H NMR (300MHz, CDCl3)
d = 7.38 (d, J = 8.5Hz, 2H), 6.84 (d, J = 8.5Hz, 2H),
4.92 (d, J = 5.5Hz, 1H), 4.12 (q, J = 4.4Hz, 2H), 3.85
(s, 3H), 3.25 (d, J = 5.5Hz, 1H), 2.85 (br, 1H), 1.35 (t,
J = 7.17Hz, 3H). MS (ESI) m/z = 263.0(M ++Na). 13C
NMR (75MHz, CDCl3) d = 13.09, 54.29, 61.11, 73.23,
73.8, 112.8, 124.89, 126.62, 131.07, 158.35, 171.80. Anal.
Calcd for C12H16O5: C, 59.99; H, 6.71. Found: C, 59.80;
H, 6.88.
References and notes
1. Schuberet, J.; Schwesinger, R.; Prizbach, H. Angew.
Chem., Int. Ed. Engl. 1994, 23, 167–169.
2. Scriven, E. F. V.; Turnbull, K. Chem. Rev. 1988, 88,
297–368.
3. Katsuki, T.; Sharpless, K. B. J. Am. Chem. Soc. 1980, 102,
5976–5978.
4. Jacobsen, E. N.; Marko, I.; Mungall, S.; Schroder, G.;
Sharpless, K. B. J. Am. Chem. Soc. 1988, 110, 1968–
1970.
5. (a) The Chemistry of Azido Group; Patal, S., Ed.; Wiley:
New York, 1971; (b) Nugent, T. C.; Hudlicky, T. J. Org.
Chem. 1998, 63, 510–520.
6. (a) Fringuelli, F.; Piermatti, O.; Pizzo, F.; Vaccaro, L. J.
Org. Chem. 2001, 64, 6094–6096; (b) Fringuelli, F.;
Piermatti, O.; Pizzo, F.; Vaccaro, L. J. Org. Chem. 2001,
66, 4719–4722.
7. (a) Nugent, W. A. J. Am. Chem. Soc. 1992, 114,
2768–2769; For review: (b) Paterson, L.; Barrisford, D.;
Organic, J. In Synthesis Highlights III; Mulzer, C.,
Waldmann, H., Eds.; Wiley: VCH, 1998.
8. Schnieder, C. Synlett, 2000, 1840–1842.
9. Sen, S. E.; Smith, S. M.; Sullivan, K. A. Tetrahedron 1999,
55, 12657–12698.
20
(2S,3R)-12: Solid; mp 97–98ꢁC; ½aŠ ꢀ46 (c 1 metha-
D
nol); ee: 98%. IR (chloroform, cmꢀ1): 3426, 3025,
2930, 1742, 1614,1514, 1370, 1250, 1176, 1030, 882,
767. 1H NMR (300MHz, CDCl3) d = 7.57 (d,
J = 8.2Hz, 2H), 7.19 (d, J = 8.1Hz, 2H), 7.10(d,
J = 8.6Hz, 2H), 6.71 (d, J = 8.6Hz, 2H), 4.98
(d, J = 4.9Hz, 1H), 4.76 (d, J = 5Hz, 1H), 4.03 (q,
J = 7Hz, 2H), 3.75 (s, 3H), 2.82 (br, 1H), 2.41 (s, 3H),
0.98 (t, J = 7Hz, 3H). MS (ESI) m/z = 417.1 (M++Na).
13C NMR (75MHz, CDCl3) d = 12.81, 20.63, 54.05,
60.81, 72.26, 75.59, 76.01, 76.44, 80.57, 95.13, 112.64,
126.65, 127.00, 128.48, 128.55, 131.79, 143.67, 158.57,
165.88.
10. Kalita, B.; Bezbarua, M. S.; Barua, N. C.; Bez, G. Synlett,
2001, 1411–1414.
11. Paquette, L. A., Ed. Encyclopaedia of Reagents for
Organic Synthesis; John Wiley & Sons, 1995; Vol. 5, pp
3646.
12. Kakeya, H.; Morishita, M.; Kobinata, K.; Osono, M.;
Ishizuka, M.; Osada, H. J. Antibiot. 1998, 51, 1126–1128.
13. Kakeya, H.; Morishita, M.; Koshino, H.; Morita, T.;
Kobayashi, K.; Osada, H. J. Org. Chem. 1999, 64,
1052–1053.
14. (a) Sakamoto, Y.; Shiraishi, A.; Soenhee, J.; Nakata, T.
Tetrahedron Lett. 1999, 40, 4203–4206; (b) Hamersak, Z.;
Ljubovic, E.; Mercep, M.; Mesic, M.; Sunjic, V. Synthesis,
2001, 1989–1992; (c) Vanketeswara Rao, B.; Madhan, A.;
Kumar, R. Tetrahedron: Asymmetry 2001, 12, 2009–2011;
(d) Miyata, O.; Asai, H.; Naito, T. Synlett 1999,
1915–1916.
20
(2R,3S)-13: Gum; ½aŠ þ9:1 (c 0.8, chloroform); ee:
D
98%. IR (chloroform, cmꢀ1): 2926, 2936, 2838, 1733,
1612, 1513, 1463, 1301, 1249, 1111, 1030, 833. 1H
NMR (300MHz, CDCl3) d = 7.31 (d, J = 8.5Hz, 2H),
6.88 (d, J = 8.5Hz, 2H), 4.24 (d, J = 7Hz, 1H), 4.02
(q, J = 7Hz, 1H), 3.82 (d, J = 4.7Hz, 1H), 3.70(s,
3H), 1.30(t, J = 7Hz, 3H). MS (ESI) m/z = 222.0
(M+). Anal. Calcd for C12H14O4: C, 64.85; H, 6.35.
Found: C, 64.63; H, 6.12.
20
(2S,3R)-14: Liquid; ½aŠ ꢀ23 (c 1.0chloroform); ee:
D
98%. IR (chloroform, cmꢀ1): 3490, 2965, 2107, 1744,
1612, 1514, 1428, 1252, 1213, 1131, 1028, 827. 1H
NMR (300MHz, CDCl3) d = 7.30(d, J = 8.5Hz, 2H),
6.88 (d, J = 8.5Hz, 2H), 4.85 (d, J = 2.7Hz, 1H), 4.23
(q, J = 7Hz, 2H), 3.75 (s, 3H), 3.08 (d, J = 6.7Hz,
1H), 1.30(t, J = 7Hz, 3H). MS (ESI) m/z = 265.1
(M+). 13C NMR (75MHz, CDCl3) d = 12.82, 54.60,
56.73, 67.45, 77.51, 77.45, 77.85, 112.81, 125.49,
127.13, 156.43, 168.29. Anal. Calcd for C12H15O4N3:
C, 54.33; H, 5.70; N, 15.84. Found: C, 54.49; H, 5.83;
N, 15.97.
15. The enantiomeric excess (ee) was measured by HPLC
analysis carried out on a Waters 510HPLC system.
Chiracel OD packed in
a SS column of 4.6mm
i.d. · 250mm dimension was used. Isocratic elution was
applied with a mobile phase n-hexane 90% and isopropa-
nol 10% at a flow rate 0.8mL/min, and pressure 125psi.
UV was monitored at 243nm.
16. Denis, J.-N.; Correa, A.; Green, A. E. J. Org. Chem. 1990,
55, 1957–1959.
17. Flemming, P. R.; Sharpless, K. B. J. Org. Chem. 1991, 56,
2869–2875.
18. Hoffman, R. V.; Kim, H. J. Org. Chem. 1991, 56,
6759–6764.
19. (a) Caron, M.; Sharpless, K. B. J. Org. Chem. 1985, 50,
1557–1560; (b) Caron, M.; Carlier, R. R.; Sharpless, K. B.
J. Org. Chem. 1988, 53, 5185–5187.
General method for epoxide ring opening: NaN3 in 2mL
˚
CH3CN were added to powdered 4A molecular sieves
(100mg) and the mixture stirred for specific time. The
reaction mixture was then filtered, the filter pad was
washed with CH3CN and the combined filtrates was
evaporated. The crude product was then purified by