Enantiopure 1,4-benzoxazines via 1,2-cyclic sulfamidates. Synthesis of levofloxacin

Article abstract of DOI:10.1021/ol0712475

1,2-Cyclic sulfamidates undergo efficient and regiospecific nucleophilic cleavage with 2-bromophenols (and related anilines and thiophenols), followed by Pd(0)mediated amination to provide an entry to substituted and enantiomerically pure 1,4-benzoxazines (and quinoxalines and 1,4-benzothiazines). This chemistry provides a short and efficient entry to (3S)-3-methyl-1,4-benzoxazine 19, a late stage intermediate in the synthesis of levofloxacin.

Full text of DOI:10.1021/ol0712475

ORGANIC
LETTERS
2007
Vol. 9, No. 17
3283-3286
Enantiopure 1,4-Benzoxazines via
1,2-Cyclic Sulfamidates. Synthesis of
Levofloxacin
John F. Bower,† Peter Szeto,‡ and Timothy Gallagher*,†
School of Chemistry, UniVersity of Bristol, Bristol BS8 1TS, U.K., and Synthetic
Chemistry, GlaxoSmithKline, Medicines Research Centre, SteVenage, SG1 2NY U.K.
t.gallagher@bristol.ac.uk
Received May 28, 2007
ABSTRACT
1,2-Cyclic sulfamidates undergo efficient and regiospecific nucleophilic cleavage with 2-bromophenols (and related anilines and thiophenols),
followed by Pd(0)-mediated amination to provide an entry to substituted and enantiomerically pure 1,4-benzoxazines (and quinoxalines and
1,4-benzothiazines). This chemistry provides a short and efficient entry to (3S)-3-methyl-1,4-benzoxazine 19, a late stage intermediate in the
synthesis of levofloxacin.
1,2- and 1,3-cyclic sulfamidates 1, which are readily available
from substituted and often enantiopure 1,2- and 1,3-ami-
noalcohols, are synthetically versatile electrophiles displaying
a reactivity profile analogous to activated aziridines and
azetidines, respectively. Nucleophilic cleavage, which is
highly selective for the C-O bond, occurs in a stereospecific
manner (SN2), and the resulting N-sulfate is readily hydro-
lyzed to the final product under mildly acidic conditions.1
Cyclic sulfamidates are readily exploitable in a number of
ways, and we have developed efficient and flexible entries
to a range of enantiomerically pure N-heterocyclic structures
as illustrated in Scheme 1.2 Of particular value is the
nucleophilic cleavage of 1,2- and 1,3-cyclic sulfamidates with
R-functionalized ester enolates which provides an entry to
substituted pyrrolidinones and piperidinones 2.
Scheme 1. Cyclic Sulfamidates as Precursors to
N-Heterocycles
† University of Bristol.
‡ GlaxoSmithKline.
(1) For a review on the synthesis and reactivity of cyclic sulfamidates,
see: Mele´ndez, R. E.; Lubell, W. D. Tetrahedron 2003, 59, 2581-2616.
Six-ring cyclic sulfamidates, where nitrogen is part of an aziridine ring
system, undergo nucleophilic cleavage preferentially at the C-N bond to
afford substituted seven-ring cyclic sulfamidates: (a) Duran, F.; Leman,
L.; Ghini, A.; Burton, G.; Dauban, P.; Dodd, R. H. Org. Lett. 2002, 4,
2481-2483. (b) Duran, F. J.; Ghini, A. A.; Dauban, P.; Dodd, R. H.; Burton,
G. J. Org. Chem. 2005, 70, 8613-8616.
The scope of this methodology has been defined both in
terms of N-heterocyclic methodologies2b,d and via efficient,
asymmetric routes to biologically active targets, such as the
antidepressant (-)-paroxetine2f and natural products (+)-
10.1021/ol0712475 CCC: $37.00
© 2007 American Chemical Society
Published on Web 07/28/2007

laccarin2f and (-)-aphanorphine.2c,e Heteroatom-based nu-
cleophiles (R-amino or thio esters) react with 1,2-cyclic
sulfamidates to give thiomorpholinones and piperazinones
3 in a stereospecific manner.2a
nucleophilic cleavage with the sodium anion of 2-bromophe-
nol (2 equiv) generally occurs between room temperature
and 60 °C (in DMF) to deliver the corresponding adducts
(8a-e) in high yield (88-99%) (Scheme 3 and Table 1).
In this paper, we describe a new application of heteroatom
nucleophiles to provide an efficient and convergent two-step
protocol for the synthesis of substituted and enantiopure 1,4-
benzoxazines and related benzofused heterocycles.3,4
This chemistry capitalizes on the highly efficient cleavage
of 1,2-cyclic sulfamidates with readily available 2-bro-
mophenolate nucleophiles (4; X ) O),5 allowing direct access
to adducts 5 (after N-sulfate hydrolysis) which afford the
target heterocycles 6 under Pd(0)-mediated Buchwald-
Hartwig amination conditions (Scheme 2).6
Scheme 3. Synthesis of (3S)-3-Benzyl-1,4-benzoxazine 9aa
a For other examples based on these procedures, see Table 1.
The only exception was cyclic sulfamidate 7d, a substrate
which is sensitive to â-elimination,2d which afforded 22%
of N-benzyl cinnamylamine in addition to the desired adduct
8d (in 57% yield). The efficient SN2 nature of the initial
nucleophilic displacement step is clearly demonstrated by
comparing entries 2 and 3 (leading to 8b and 8c) in which
none of the alternative diastereomer was detectable.
Scheme 2. Tandem Ring-Opening/Pd(0)-Mediated Amination
Pd(0)-catalyzed cyclization of phenylalanine-derived ad-
duct 8a was investigated using a range of ligands under
The scope of this methodology is reported as is its
application to a concise, high-yielding, and asymmetric entry
to the potent antibiotic drug levofloxacin. In addition, the
use of nucleophiles based on 2-bromoaniline (4; X ) NH)
and 2-bromothiophenol (4; X ) S) provides direct access to
substituted and enantiomerically pure quinoxaline and 1,4-
benzothiazine variants, respectively.7
Table 1. Synthesis of 2- and 3-Substituted and 2,3-Disubstituted
1,4-Benzoxazines
Using a structurally representative range of substituted 1,2-
and 1,3-cyclic sulfamidates 7a-e,2a,b,d we have found that
(2) (a) Williams, A. J.; Chakthong, S.; Gray, D.; Lawrence, R. M.;
Gallagher, T. Org. Lett. 2003, 5, 811-814. (b) Bower, J. F.; Sˇvenda, J.;
Williams, A. J.; Charmant, J. P. H.; Lawrence, R. M.; Szeto, P.; Gallagher,
T. Org. Lett. 2004, 6, 4727-4730. (c) Bower, J. F.; Szeto, P.; Gallagher,
T. Chem. Commun. 2005, 5793-5795. (d) Bower, J. F.; Chakthong, S.;
Sˇvenda, J.; Williams, A. J.; Lawrence, R. M.; Szeto, P.; Gallagher, T. Org.
Biomol. Chem. 2006, 4, 1868-1877. (e) Bower, J. F.; Szeto, P.; Gallagher,
T. Org. Biomol. Chem. 2007, 5, 143-150. (f) Bower, J. F.; Riis-Johannessen,
T.; Szeto, P.; Whitehead, A. J.; Gallagher, T. Chem. Commun. 2007, 728-
730.
(3) For a recent review covering the synthesis and biological importance
of 1,4-benzoxazines, see: Ilasˇ, J.; Anderluh, P. Sˇ.; Dolenc, M. S.; Kikelj,
D. Tetrahedron 2005, 61, 7325-7348.
(4) For recent approaches to 1,4-benzoxazines and related heterocyclic
scaffolds, see: (a) Wolfer, J.; Bekele, T.; Abraham, C. J.; Dogo-Isonagie,
C.; Lectka, T. Angew. Chem., Int. Ed. 2006, 45, 7398-7400. (b) Xu, D.;
Chiaroni, A.; Fleury, M.-B.; Largeron, M. J. Org. Chem. 2006, 71, 6374-
6381. (c) Feng, G.; Wu, J.; Dai, W.-M. Tetrahedron 2006, 62, 4635-4642.
(d) Shinkevich, E. Y.; Novikov, M. S.; Khlebnikov, A. F. Synthesis 2007,
225-230.
(5) Openings of 1,2-cyclic sulfamidates with 2-methoxyphenolate have
previously been reported in 47-82% isolated yield: Okuda, M.; Tomioka,
K. Tetrahedron Lett. 1994, 35, 4585-4586.
(6) For a review on Pd-catalyzed amination, see: (a) Yang, B. H.;
Buchwald, S. L. J. Organomet. Chem. 1999, 576, 125-146. For a previous
approach to benzoxazines using Ni(0)-mediated intramolecular amination
of aryl chlorides, see: (b) Omar-Amrani, R.; Thomas, A.; Brenner, E.;
Schneider, R.; Fort, Y. Org. Lett. 2003, 5, 2311-2314. For an example
using Pd(0)-mediated intramolecular amination of an aryl chloride, see: (c)
Omar-Amrani, R.; Schneider, R.; Fort, Y. Synthesis 2004, 2527-2534.
(7) For a previous synthesis of a quinoxalinone via intramolecular Pd-
(0)-mediated amination of an aryl iodide, see: Kitagawa, O.; Takahashi,
M.; Yoshikawa, M.; Taguchi, T. J. Am. Chem. Soc. 2005, 127, 3676-
3677.
a
b
Isolated yield. C-Br reduction and polymerization occurred.
3284
Org. Lett., Vol. 9, No. 17, 2007

electron-poor (including 10a and 3-hydroxypyridine 10b) and
electron-rich (10c and 10d10) nucleophiles are highly efficient
in this process. The subsequent Pd-catalyzed arylation step
occurs in moderate to excellent yield leading to 12a-d; the
most difficult substrates were those which possess electron-
donating substituents on the aromatic (e.g., 11d f 12d), and
in such cases, cyclization was both slower and less efficient.
Extension of this two-step protocol also allows access to
thio and aza variants 12e and 12f. Cyclic sulfamidate 7a
reacted efficiently with 2-bromothiophenol 10e to deliver 11e
in quantitative yield. Cyclization of this species was facile
and afforded benzothiazine 12e in 89% yield. Opening of
7a with the anion of aniline 10f was less efficient, possibly
due to competing elimination, and adduct 11f was isolated
in 56% yield. Cyclization of 11f was also slow, and
quinoxaline 12f was isolated in only 22% yield (30% based
on recovered 11f). Obviously this process requires further
optimization, but the results shown in Table 2 clearly
demonstrate the breadth and potential applicability of this
chemistry.
Table 2. Synthetic Scope: O-, N-, and S-Based Nucleophiles
Scheme 4. Synthesis of Benzoxazine 19 and Formal Synthesis
of Levofloxacin
a
Isolated yield.
standard conditions (t-BuONa, PhMe, 5 mol % of Pd(OAc)2,
100 °C), and of those screened, xantphos (7.5 mol %) was
most efficient affording benzoxazine 9a in 88% yield.8 These
conditions have successfully been applied to adducts 8b-d
to deliver the target heterocycles 9b-d in good to excellent
yield. A representative example based on the conversion of
7a to 9a is shown in Scheme 3, and other substrates are
summarized in Table 1 (full details are given in the
Supporting Information). Unfortunately, the annulation ad-
duct 8e derived from 1,3-cyclic sulfamidate 7e did not cyclize
(to form 9e), and only reduction and polymerization were
observed.9
Using the phenylalanine-derived cyclic sulfamidate 7a, a
more comprehensive assessment of the scope of the nucleo-
philic component and the subsequent Pd-catalyzed cyclization
involved in this process has been conducted (Table 2).
Phenolates are generally well tolerated, and a range of both
The blockbuster antibiotic levofloxacin 20 (2006 sales of
US $1.41 billion) provides an appropriate focus for this
methodology. The major challenge associated with develop-
ing an asymmetric entry to 20 resides in identifying efficient
routes to the key chiral benzoxazine core 19, and in this
regard, several approaches have been reported.11 We have
achieved this goal through the union of alanine-derived cyclic
sulfamidate 15 and 6-bromo-2,3-difluorophenol 16 (Scheme
4).
(8) Other ligands screened included BINAP which gave 9a in 76% yield
and Verkade’s TTPU ligand (Urgaonkar, S.; Xu, J. H.; Verkade, J. G. J.
Org. Chem. 2003, 68, 8416-8423) which gave 9a in 86% yield.
(9) An example of a successful Pd-catalyzed cyclization to form a related
seven-ring system has been reported.6c
(10) For the preparation of 10d, see: Novak, Z.; Timar, G.; Kotschy,
A. Tetrahedron 2003, 59, 7509-7513.
Org. Lett., Vol. 9, No. 17, 2007
3285

Commercially available alaninol 13 was N-Boc protected
under standard conditions to give 14, which was then
converted to cyclic sulfamidate 15 in 90% yield over three
steps.12 Reaction of 15 (in DMF) with the sodium anion of
phenol 16 (1.2 equiv) (prepared by bromination of com-
mercially available 2,3-difluorophenol in 91% yield13) led
to smooth nucleophilic cleavage to afford the intermediate
N-sulfate 17. In this case, direct and concomitant removal
of both the N-sulfate and N-Boc moieties was achieved by
employing 10% H2SO4 in p-dioxane for the hydrolysis step
which afforded cyclization precursor 18 in 99% overall yield
from 15. Cyclization of 15 (5 mol % of Pd(OAc)2, 7.5 mol
% of xantphos, t-BuONa, PhMe, 100 °C) then cleanly
afforded the key benzoxazine intermediate 19 in 84% yield
20
23
{[R]D -9.1 (c 1.3, CHCl3); lit. [R]D -7.8 (c 0.7,
CHCl3)11b}. This sequence is highly efficient affording
benzoxazine 19 in 74% overall yield over six steps. Conver-
sion of this intermediate in three steps to levofloxacin 20
has previously been reported.11a,d
In summary, we have demonstrated an efficient and
modular two-step protocol for the coupling of 1,2-cyclic
sulfamidates with readily available 2-bromophenols as a
means of accessing substituted and enantioenriched 1,4-
benzoxazines and have shown that analogous chemistry is
equally applicable to the synthesis of quinoxaline and 1,4-
benzothiazine scaffolds. This methodology provides a direct
and high-yielding entry to the drug levofloxacin, and its
fundamental simplicitysthe use of readily available nucleo-
philic and electrophilic componentssmakes it well suited
to further applications within both academic and industrial
settings.
(11) For selected previous synthetic studies on 20, see: (a) Hayakawa,
I.; Atarashi, S.; Imamura, M.; Yokohama, S.; Higashihashi, N.; Sakano K.;
Ohshima, M. US patent, 1986, 5,053,407. (b) Atarashi, S.; Yokohama, S.;
Yamazaki, K.; Sakano, K.; Imamura, M.; Hayakawa, I. Chem. Pharm. Bull.
1987, 35, 1896-1902. (c) Sakano, K.; Yokohoma, S.; Hayakawa, I.;
Atarashi, S.; Kadoya, S. Agric. Biol. Chem. 1987, 51, 1265-1270. (d)
Mitscher, L. A.; Sharma, P. N.; Chu, D. T. W.; Shen, L. L.; Pernet, A. G.
J. Med. Chem. 1987, 30, 2283-2286. (e) Atarashi, S.; Tsurumi, H.;
Fujiwara, T.; Hayakawa, I. J. Heterocycl. Chem. 1991, 28, 329-331. (f)
Kang, S. B.; Ahn, E. J.; Kim, Y.; Kim, Y. H. Tetrahedron Lett. 1996, 37,
9317-9320. (g) Satoh, K.; Inenaga, M.; Kanai, K. Tetrahedron: Asymmetry
1998, 9, 2657-2662. (h) Adrio, J.; Carretero, J. C.; Ruano, J. L. G.; Pallare´s,
A.; Vicioso, M. Heterocycles 1999, 51, 1563-1572.
Acknowledgment. We thank the EPSRC and GSK for
financial support.
Supporting Information Available: Full experimental
1
details, compound characterization data, and copies of H
(12) Tewson has previously reported the synthesis of cyclic sulfamidate
15 in 78% yield from 14: Posakony, J. J.; Grierson, J. R.; Tewson, T. J. J.
Org. Chem. 2002, 67, 5164-5169.
(13) Pearson, D. E.; Wysong, R. D.; Breder, C. V. J. Org. Chem. 1967,
32, 2358-2360.
and 13C NMR are available. This material is available free
of charge via the Internet at http://pubs.acs.org.
OL0712475
3286
Org. Lett., Vol. 9, No. 17, 2007