Regiospecific and stereoselective syntheses of (±) morphine, codeine, and thebaine via a highly stereocontrolled intramolecular 4 + 2 cycloaddition leading to a phenanthrofuran system

Article abstract of DOI:10.1021/ja9038505

Total syntheses of the morphine alkaloids are described that use a direct stereoselective formation of the phenanthrofuran system via an intramolecular 4 + 2 cycloaddition of a diene tethered to the 4-position of a 7-methoxybenzofuran-3-carboxylic acid es

Full text of DOI:10.1021/ja9038505

Published on Web 07/22/2009
Regiospecific and Stereoselective Syntheses of (() Morphine,
Codeine, and Thebaine via a Highly Stereocontrolled
Intramolecular 4 + 2 Cycloaddition Leading to a
Phenanthrofuran System
Gilbert Stork,* Ayako Yamashita,† Julian Adams,‡ Gary R. Schulte,§
Richard Chesworth,| Yoji Miyazaki,⊥ and Jay J. Farmer¶
Department of Chemistry, Columbia UniVersity, New York, New York
Received May 12, 2009; E-mail: gjs8@columbia.edu
Abstract: Total syntheses of the morphine alkaloids are described that use a direct stereoselective formation
of the phenanthrofuran system via an intramolecular 4 + 2 cycloaddition of a diene tethered to the 4-position
of a 7-methoxybenzofuran-3-carboxylic acid ester.
substituted benzofuran ring.5 Our work on that eventually
successful approach is sketched in Figure 1.
Introduction
Over half a century ago, the total synthesis of morphine was
accomplished by Marshall Gates.1a A very large number of total
syntheses of natural products have been carried out since that
time, often with the goal of making medically significant natural
substances more available. Efficiency was clearly a major
concern. Frequently, however, the fascination of a synthesis
target comes from the architectural challenges presented by an
intricate structure, and originality of design is the goal.
In the case of the morphine alkaloids,2 the specific structural
challenge of having four rings sharing the same carbon atom is
further complicated by their propensity to undergo sometimes
spectacular rearrangements.3 In any case, it is remarkable that
numerous total syntheses of morphine (1) and/or its relatives,
codeine (2), and thebaine (3), have been reported in the literature
between the 1952 Gates synthesis and the latest publication in
2009.1
Stereochemical Suitability of the Intramolecular Diels-Alder
Route. Two questions arose as we considered such an intramo-
lecular 4 + 2 cycloaddition. First, the nature of R, which
eventually has to be transformed into an ethanamine chain
connected to ring B. A carbomethoxy group appeared to be
(1) (a) Gates, M.; Tschudi, G. J. Am. Chem. Soc. 1952, 74, 1109–1110.
1956, 78, 1380-1393. (b) Elad, D.; Ginsburg, D. J. Am. Chem. Soc.
1954, 76, 312–313. (c) Grewe, R.; Fischer, H. Chem Ber. 1963, 96,
1520–1528. (d) Grewe, R.; Friedrichsen, W. Chem. Ber. 1967, 100,
1550–1558. (e) Barton, D. H. R.; Bhakuni, D. S.; James, R.; Kirby,
G. W. J. Chem. Soc. C 1967, 128–132. (f) Morrison, G. C.; Waite,
R. O.; Shavel, J., Jr. Tetrahedron Lett. 1967, 8, 4055–4056. (g)
Kametani, T.; Ihara, M.; Fukumoto, K.; Yagi, H. J. Chem. Soc. C
1969, 2030–2033. (h) Schwartz, M. A.; Mami, I. S. J. Am. Chem.
Soc. 1975, 97, 1239–1240. (i) Beyerman, H. C.; Lie, T. S.; Maat, M.;
Bosman, H. H.; Buurman, E.; Bijsterveld, E. J. M. Recl. TraV. Chim.
Pays-Bas 1976, 95, 24–25. Lie, T. S.; Maat, L.; Beyerman, H. C. Recl.
TraV. Chim. Pays-Bas 1979, 98, 419–420. (j) Rice, K. C. J. Org. Chem.
1980, 45, 3135–3137. (k) Evans, D. A.; Mitch, C. H. Tetrahedron
Lett. 1982, 285–288. (l) Moos, W. H.; Gless, R. D.; Rapoport, H. J.
Org. Chem. 1983, 48, 227–238. (m) White, J. D.; Caravatti, G.; Kline,
T. B.; Edstrom, E.; Rice, K. C.; Brossi, A. Tetrahedron 1983, 39,
2393–2397. (n) Toth, J. E.; Fuchs, P. L. J. Org. Chem. 1987, 52, 473–
475. J. Org. Chem. 1988, 53, 4694. (o) Tius, M. A.; Kerr, M. A. J. Am.
Chem. Soc. 1992, 114, 5959–5966. (p) Parker, K. A.; Fokas, D. J. Am.
Chem. Soc. 1992, 114, 9688–9689; J. Org. Chem. 1994, 59, 3927-
3933; J. Org. Chem. 2006, 71, 449-455. (q) Hong, C. Y.; Kado, N.;
Overman, L. E. J. Am. Chem. Soc. 1993, 115, 11028–11029. Hong,
C. Y.; Overman, L. E. Tetrahedron Lett. 1994, 35, 3453–3456. (r)
Mulzer, J.; Duerner, G.; Trauner, D. Angew. Chem., Int. Ed. Engl.
1996, 35, 2830–2832. Mulzer, J.; Bats, J. W.; List, B.; Opatz, T.;
Trauner, D. Synlett 1997, 441–444. Trauner, D.; Bats, J. W.; Werner,
A.; Mulzer, J. J. Org. Chem. 1998, 63, 5908–5918. Mulzer, J.; Trauner,
D. Chirality 1999, 11, 475–482. (s) White, J. D.; Hrnciar, P.;
Stappenbeck, F. J. Org. Chem. 1997, 62, 5250–5251. J. Org. Chem.
1999, 64, 7871-7884. (t) Nagata, H.; Miyazawa, N.; Ogasawara, K.
Chem. Commun. 2001, 1094–1095. (u) Taber, D. F.; Neubert, T. D.;
Rheingold, A. L. J. Am. Chem. Soc. 2002, 124, 12416–12417. (v)
Trost, B. M.; Tang, W. J. Am. Chem. Soc. 2002, 124, 14542–14543.
Trost, B. M.; Tang, W.; Toste, F. D. J. Am. Chem. Soc. 2005, 127,
14785–14803. (w) Uchida, K.; Yokoshima, S.; Kan, T.; Fukuyama,
T. Org. Lett. 2006, 8, 5311–5313. Heterocycles 2009, 77, 1219-1234.
(x) Omori, A. T.; Finn, K. J.; Leisch, H.; Carroll, R. J; Hudlicky, T.
Synlett 2007, 2859–2862. (y) Verin, M.; Barre´, E.; Iorga, B.; Guillou,
C. Chem.sEur. J. 2008, 14, 6606–6608. (z) Tanimoto, H.; Saito, R.;
Chida, H. Tetrahedron Lett. 2008, 49, 358–362.
Our own interest in the synthesis of the morphine alkaloids4
arose because the numerous approaches followed to their
syntheses did not include the direct formation of the phenan-
throfuran ring system of morphine by the intramolecular 4 + 2
cycloaddition of a diene tethered to the C-4 of a suitably
† Columbia University, New York, NY.
‡ Infinity Pharmaceuticals, Cambridge, MA.
§ Kahuna Scientific Consulting, LLC., Stonington, CT.
| Envivo Pharmaceuticals, Watertown, MA.
⊥ University of Tokyo, Tokyo, Japan.
¶ Present address: Novomer, Ithaca, NY.
9
11402 J. AM. CHEM. SOC. 2009, 131, 11402–11406
10.1021/ja9038505 CCC: $40.75  2009 American Chemical Society

Syntheses of (() Morphine, Codeine, and Thebaine
A R T I C L E S
find that, here again, 2 + 4 cycloaddition had followed the path
that led to b and producing structure 6 (established by X-ray of
its acetate; cf Supporting Information) with the resulting correct
relative arrangement of the three contiguous centers in ring C.
Figure 1
suitable for that eventual transformation, and should facilitate
the planned cycloaddition. Most important was whether the
addition would take place with the diene endo to the ester,
leading to adduct a, with the unwanted relative stereochemistry
at the starred center, or exo to it (an addition probably more
relevantly described as endo to the benzofuran), thus leading
to the desired outcome shown in b.
The favorable result just described suggested that one could
easily complete a formal synthesis of the morphine alkaloids
from adduct 6. This would require straightforward transforma-
tion of the carbomethoxy group to an N-tosylethanamine chain,
followed by its attachment to ring B, after dehydration to a
styrene system, by application of the Parker piperidine ring
closure.1p We successfully carried out these and other necessary
transformations to (() codeine, but we will describe fully here
a less predictable route in which the starting diene system is
introduced into benzofuranacetadehyde 11. This variation was
made in the hope that eventual closure of the piperidine ring
could be simply achieved by dispacement of a derivative of
the secondary carbinol in ring B of the adduct (cf. Scheme 1).
The synthesis of 11 is outlined in Scheme 1. It starts with
the acetal 8 of the readily available iodoisovanillin 7.8 Conjugate
Scheme 1a
Early experiments in our laboratory were actually performed
with the benzofuran diester 4, and showed (X-ray)6 that the
adduct isolated from 4 was the desirable isomer b above. This
favorable result led us, eventually, to examine whether the
methoxycarbonyl at C-2 (which would eventually have to be
removed) was required for the subsequent 4 + 2 cycloaddition
and, if not, whether one would still get an endo to benzofuran
cycloaddition. We selected the 3-carbomethoxybenzofuran
derivative 57 to explore these questions, and were pleased to
(2) For several excellent recent reviews related to morphine (biological
action, biogenesis, total syntheses, and various approaches, see, inter
alia: Blakemore, P. R.; White, J. D. Chem. Commun. 2002, 1159–
1168. Taber, D. F.; Neubert, T. D.; Schlecht, M. F. in Strategies and
Tactics in Organic Synthesis; Harnata, M., Ed.; Elsevier: London,
2004, Vol. 5, pp 353-389. Zezula, J.; Hudlicky, T. Synlett 2005, 388–
405. Hudlicky, T.; Reed, J. W. The Way of Synthesis; Wiley-VCH:
Weinheim, 2007, pp 726-757.
a Conditions: (a) HO(CH2)3OH, p-TSOH, toluene; (b) methyl propiolate;
Et3N, THF; (c) Pd(OAc)2, Ph3P, NaOAc, n-Bu4NCl, DMF, 125°; (d) HCl,
THF; (e) (1) Ph3PCH2OCH3Cl, KHMDS, THF, (2) HCl, THF.
(3) “The star performers in the team of molecular acrobats are undoubtedly
the alkaloids of the morphine group.” Robinson, R. Proc. R. Soc.
London 1947, B135, xiv.
addition of 8 to either methyl propiolate or methyl 3-chloro-
acrylate gave what proved to be a suitable precursor of the desired
(4) Starting with early unsuccessful synthesis attempts: Stork, G.; Conroy,
H. J. Am. Chem. Soc. 1951, 73, 4748–4751. See also: Stork, G. in
The Alkaloids; Manske, R. H. F.; Holmes, H. L., Eds.; Academic Press:
New York, 1952; Vol. II, Chapter 8, Part II, pp 171-203; Stork, G.
The Alkaloids; Vol. VI, chapter 7, pp 219-245.
(5) For intermolecular additions to benzofurans, see: Wenkert, E.; Piettre,
S. R. J. Org. Chem. 1988, 53, 5850–5853, For intramolecular, B-seco
morphine related, 4 + 2 additions to benzofurans, see. (a) Ciganek,
E. J. Am. Chem. Soc. 1981, 103, 6261–6262. (b) France, S.;
Boonsombat, J.; Leverett, C. A.; Padwa, A. J. Org. Chem. 2008, 73,
8120–8123. The possibility of a direct formation of a phenanthrofuran
system via an intramolecular Diels Alder was considered by Parsons,
P. J. Chem. ReV. 1996, 96, 197. For different approaches to phenan-
throfuran precursors of morphine alkaloids, see inter alia: Fuchs, P. M.
(ref 1n). Parker, K. A. (ref 1p).
(7) Experimental details of the synthesis of 1,3-pentadienecarbinol 5 from
the benzofurancarboxaldehyde 10 are described in the Supporting
Information, as is the X-ray of the acetate of Diels-Alder adduct 6.
That X-ray showed that selective formation of a specific diastereomer
of the secondary hydroxyl in ring B of 6 had taken place. This is
potentially useful because, even though that hydroxyl would have to
be removed at some future point, it means that one could select the
correct enantiomer of starting carbinol 5 to produce either enantiomer
of the synthetic alkaloid. We did not take advantage of that possibility
because enantioinduction was not a goal of our work. The synthetic
substances synthesized in this work are racemic: the structural drawings
in this work indicate their relative stereochemistry.
(8) Markovich, K. M.; Tantishaiyakul, V.; Hamada, A.; Miller, D. D.;
Romstedt, K. Y.; Shams, G.; Shin, Y.; Fraundorfer, P. F.; Doyle, K.;
Feller, D. R. J. Med. Chem. 1992, 35, 466–479.
(6) Carried out in 1982 by Ms. Gayle Schulte in Professor Jon Clardy’s
Laboratory at Cornell University.
9
J. AM. CHEM. SOC. VOL. 131, NO. 32, 2009 11403

A R T I C L E S
Stork et al.
Scheme 2a
Scheme 3a
a Conditions: (a) (1) Super-hydride, THF, (2) Dess-Martin, CH2Cl2. (b)
Ph3PCH2OCH3Cl, KHMDS, CH2Cl2-THF.
Scheme 4a
a Conditions: (a) (1) 3-Buten-1-yne, ZrCp2(H)Cl, AgOTf, CH2Cl2, (2)
TESCl, imidazole. (b) Decalin, Et3N, 230 °C. (c) (1) TBAF, THF, (2) super
hydride, CH2Cl2.
a Conditions: (a) (1) TBAF, THF, (2) Dess-Martin, CH2Cl2, 0°. (b) (1)
L-Selectride, THF, 0°, (2) MsCl, Et3N, CH2Cl2, 0°. (c) (1) HCl, THF, (2)
MeNH2 ·HCl, Et3N, Ti(OiPr)4, MeOH, (3) NaBH4.
benzofuran-3-carboxylic acid derivative 9, via a Heck reaction
catalyzed by palladium acetate.9 Dilute acid hydrolysis of the acetal,
followed by homologation by the usual Wittig-Levine reaction,10
now gave the desired benzofuranacetaldehyde 11.
to be inverted to that of minor isomer 13a to take part in the
closure planned for the piperidine ring. Fortunately, however,
the mixture of epimers 13 and 13a could be used without
separation because both should converge to the same ketone
(()16.14 As the sequel shows, this proved to be the case.
Elaboration of the Ethanamine Chain. Construction of the
piperidine ring of the opium alkaloids has been carried out
following a variety of pathways, but, surprisingly, not by a
simple intramolecular displacement involving an ethanamine
chain. This presents an interesting challenge here, because the
basicity of the ethanamine chain suggests competition between
the desired displacement process and a base-catalyzed elimination.
Conversion of the ester group in 13,13a to an ethanamine
was carried out by standard operations (cf Scheme 3), to produce
enol ethers 15 of the precursor acetaldehyde chain.
The assumption that both 13 and 13a would lead to the same
ketone (16) was verified,14 and as anticipated from its shape,
reduction with L-Selectride (Scheme 4) took place from the more
accessible face and gave the R-hydroxyl stereochemistry
required for the formation of mesylate 17. The ethanamine chain
needed for the planned cyclization was now established by
hydrolysis of the methyl enol ether in 17, followed by reaction
of the resulting aldehyde with methylamine in the presence of
titanium tetraisopropoxide, and in situ reduction of the resulting
imine with sodium borohydride.15 Methylaminomesylate 18,
thus obtained in 87% yield from 17, was ready for the crucial
issue of displacement versus elimination. This was anticipated
to be a serious problem because, in 18, the relevant hydrogens
adjacent to the mesylate can attain antiperiplanar arrangement
with it, and they are readily reached by the methylamine group,
thus leading to unwanted arylethylene or conjugated diene.
(() Deoxycodeine. There have been a large number of
experiments and discussions in the literature about the competi-
Intramolecular 4 + 2 Cycloaddition. Our initial target was
the TES-protected dienyl carbinol 12. Participation of the
butadiene system of 12 in a Diels-Alder addition requires that
the disubstituted allylic double bond has the (E) stereochemistry
shown: the Suzuki process11 was selected as a particularly
effective method of achieving that result. Hydrozirconation of
vinylacetylene with the Schwartz reagent,12 followed by in situ
reaction of the (E) vinylzirconium intermediate with benzo-
furanacetaldehyde 11, and trapping with chlorotriethylsilane
gave the desired 12 in excellent yield (Scheme 2).
Considerable experimentation was involved in determining
the conditions that were eventually selected for the cycloaddi-
tion: degassed decalin solution heated for 5 h, under argon, at
∼230 °C (bath temperature) in a sealed pressure flask, in the
presence of some triethylamine. Under those conditions, phenan-
throfurans 13 and 13a (89% yield in a 4/1 ratio) were obtained.
The structure of the desilylated minor isomer 13a, was
established by X-ray crystallography (cf Supporting Informa-
tion). That of the major isomer 13 was tentatively written as
shown because it readily formed a lactone on desilylation, and
because we assumed (correctly, as it turned out) that it, like
13a, again came from the now familiar endo benzofuran
transition state. The formation of 13 as the major adduct further
implies that, in the 4 + 2 cycloaddition leading to it, the
benzofuran and the OTES group are positioned so they end
mostly cis to each other.13
If the relationship of the asymmetric centers in 13 is as that
in 13a, then they both would have the required all cis
arrangement of the three crucial contiguous centers around ring
C. Obviously, however, the protected hydroxyl of 13 would have
(9) For a related case, cf: Henke, B. R. et al J. Med. Chem. 1997, 40,
2706-2725.
(10) Levine, S. G. J. Am. Chem. Soc. 1958, 80, 6150–6151.
(11) Suzuki, K. Pure Appl. Chem. 1994, 66, 1557–1564. Maeta, H.;
Hashimoto, T.; Hasegawa, T.; Suzuki, K. Tetrahedron Lett. 1992, 33,
5965–5968. Maeta, H.; Suzuki, K. Tetrahedron Lett. 1993, 34, 341–
344.
(14) Actually, because the enol ether system precursor of the acetaldehyde
chain is itself an (operationally irrelevant) E/Z mixture, convergence
to a single ketone was cleanly observable only after hydrolysis of enol
ether 17 to the related aldehyde.
(15) (a) Mattson, R.; Pham, K. M.; Leuck, D. J.; Cowen, K. A. J. Org.
Chem. 1990, 55, 2552–2554. (b) Neidigh, K. A.; Avery, M. A.;
Williams, J. S.; Bhattacharyya, S. J. J. Chem. Soc. Perkin Trans. I
1998, 2527–2532.
(12) Schwartz, J.; Labinger, J. A. Angew. Chem., Int. Ed. Engl. 1976, 15,
333–340.
(13) Possibly as a result of attraction between the carbonyl of the ester
and the OTES group.
9
11404 J. AM. CHEM. SOC. VOL. 131, NO. 32, 2009

Syntheses of (() Morphine, Codeine, and Thebaine
A R T I C L E S
Scheme 5
Scheme 6a
tion between SN2 displacements and E2 eliminations, particu-
larly in simple ꢀ-phenethyl systems. The aminomesylate 18
involves a similar competition, although, in our case, both
pathways are intramolecular.
While the existing evidence may not be applicable to our
case, it strongly suggests that a departing sulfonate favors
substitution compared with a bromide or iodide.16 We were,
nevertheless, unprepared to find (Scheme 5) that, in refluxing
acetonitrile, essentially complete elimination of mesylate 18 (cf,
19) was observed. Fortunately, in benzene solution at 75 °C
for 24 h, displacement took place,17 with the formation of (()
6-deoxycodeine (20), in 53% yield, accompanied by ∼30%
elimination. A possible interpretation of these observations is
that there may be, at least in this particular case, more charge
separation in the transition state for the elimination reaction than
in that for the desired substitution. Additional factors are likely
to be involved (conformational?) because in the presence of a
methoxyl at C-6 (in our precursor of codeine methyl ether, Vide
infra), the similar reaction in benzene appears to lead to the
desired substitution with significantly less elimination.
The above construction of (() 6-deoxycodeine may be
claimed to constitute a formal synthesis of (() codeine because
we found it possible to introduce the missing oxygen at C-6
via selenium dioxide oxidation of the N-methylcarbamate
derived from our (() 6-deoxycodeine to the corresponding
N-derivative of isocodeine in 61% yield.18 The transformation
of isocodeine into codeine19 (and the latter into morphine20) is
well-known. It was, however, thought interesting to examine
whether the eventually required oxygen at C-6 might be
introduced at a much earlier stage, as a terminal substituent on
the conjugated diene of 12. This approach, starting with a
methoxydiene addition to benzofuran 11, was successful. It
allowed a synthesis of both thebaine (3) and codeine (2) from
an intermediate codeine methyl ether. This is described below.
Codeine Methyl Ether and Thebaine. Realization of the
scheme just mentioned required the synthesis of 4-methoxy-3-
butene-1-yne, more specifically of its 3,4 (E) isomer 21 (Scheme
6). As it happens, the 3,4 (Z) isomer is the one commercially
available, but the required (E) isomer is known to be the major
product of the reaction of the dimethyl ether of 2-butyne-1,4-
a Conditions: (a) (1) 11, ZrCp2(H)Cl, AgOTf, CH2Cl2, (2) TESCl,
imidazole. (b) Decalin, Et3N, 240°, 24 h. (c) TBAF, THF.
diol with sodamide in liquid ammonia.21 The small amount of
(Z) isomer formed simultaneously can be readily separated, but
this is unnecessary because it is expected to be much less
reactive than the (E) isomer in the subsequent 4 + 2 cycload-
dition. The Suzuki reaction sequence,11 involving hydrozir-
conation and in situ addition to the benzofuranacetaldehyde 11,
proceeded, again in excellent yield, to the anticipated meth-
oxydienyl carbinol, isolated as its TES derivative 22, precursor
of the planned 4 + 2 cycloaddition (Scheme 7). The 4 + 2
cycloaddition once more took place in the required “endo-
benzofuran” mode, thus leading to the correct relatiVe stereo-
chemistry of the fiVe contiguous asymmetric centers shown in
23. Possibly as a result of some steric interference by the
methoxy substituent on the “endo benzofuran” transition state,
cycloaddition appeared to be slower than in the previous
desmethoxy case: 24 h in benzene at 235-240 °C (bath
temperature) gave 21% recovered starting material, in addition
to 69% of a mixture (again 4:1) of a major and a minor adduct:
X-ray structure determinations (see Supporting Information) of
the crystalline alcohol obtained by removal of the TES protecting
group of the minor adduct 23a and of the crystalline lactone
from 23 established that cycloaddition of 22 had proceeded,
similarly to the previously described desmethoxy case, to give
a 4:1 ratio of 23 to 23a in favor of 23.
(16) Veeravagu, P.; Arnold, R. T.; Eigenmann, E. W. J. Am. Chem. Soc.
1964, 86, 3072–3075.
(17) For an early example of an uneventful intermolecular displacement
of the primary halide in 2-bromoethylbenzene by a secondary amine,
in benzene, see: Foreman, E. L.; McElvain, S. M. J. Am. Chem. Soc.
1940, 62, 1435–1438.
As in the desmethoxy series, the ∼4:1 mixture of isomeric
TES-protected alcohols 23, 23a did not require separation
because both proceed to the same (()ketone 24 (Scheme 7).
Following the procedures described earlier in the deoxycodeine
synthesis, DIBAL reduction and mesylation to 25, followed by
(18) As described in the Supporting Information, the 61% yield in that
oxidation with tert-butyl hydroperoxide/selenium dioxide in methylene
dichloride was obtained only in the presence of a small amount of
water. The observed regiochemistry may reflect the smaller structural
distortion required for that transition state over the alternative. See
also ref 1v.
(19) (a) Ijima, I.; Rice, K. C.; Silverton, J. V. 1977, 6, 1157–1165 (same
process, but starting with the corresponding ethyl carbamate). (b)
Schwartz, M. A.; Pham, P. T. K. J. Org. Chem. 1988, 53, 2318–2322.
(20) Rice, K. C. J. Med. Chem. 1977, 20, 164–165.
(21) Brandsma, L. PreparatiVe Acetylene Chemistry; Elsevier: Amsterdam,
1971; p 181; Adams, H.; Anderson, J. C.; Bell, R.; Jones, D. N.; Peel,
M. R.; Tomkinson, N. C. O. J. Chem. Soc., Perkin Trans. I 1998,
3967–3974.
9
J. AM. CHEM. SOC. VOL. 131, NO. 32, 2009 11405

A R T I C L E S
Stork et al.
Scheme 7a
Scheme 8
nature of the methoxy group led to only a minor amount of
codeine. The major cleavage, when the reaction was run for
a very short time, was the 6-ꢀ-bromo compound. Steric
hindrance on the underside (R) side of the molecule precluded
simple inversion displacement, as by acetate. A number of
exotic possibilities, such as two sequential SN2′ reactions,
were explored with only limited success. We eventually
recognized that the conformation factors involved (vide supra)
in the formation of thebaine from codeine methyl ether with
manganese dioxide,22 and with the selenium dioxide oxidation
of deoxycodeine (vide supra) might well allow direct
selenium dioxide oxidation of codeine methyl ether to
codeinone. This was successful. After quantitative transfor-
mation of 27 to its methyl carbamate by reaction with methyl
chloroformate, selenium dioxide accomplished, in 72% yield,
the hoped-for conversion into the carbamate of codeinone
(28) (Scheme 8). The latter, as previously described,19
underwent simultaneous reduction of the ketone carbonyl to
the correct 6-hydroxyl stereochemistry, as well as of the
urethane to regenerate the N-methyl group, thus producing
the desired (() codeine (2). Its spectral properties (NMR,
IR, ESI-LCMS) matched those of natural (-) codeine.
Conversion of codeine to morphine (1) by cleavage of the
aromatic methoxy group to the free phenol is known, both in
the natural20 and the (()1w series: The 4 + 2 route to codeine,
morphine, and thebaine was now complete.
a Conditions: (a) (1) Super-hydride, THF, (2) Dess-Martin, CH2Cl2, (3)
Ph3PCH2OCH3Cl,KHMDS,CH2Cl2-THF,(4)TBAF,THF,(5)Dess-Martin,
CH2Cl2, 0°. (b) (1) L-Selectride, THF, 0°, (2) MsCl, Et3N, CH2Cl2, 0°. (c)
(1) HCl, THF, (2) MeNH2 ·HCl, Et3N, Ti(OiPr)4, MeOH, (3) NaBH4. (d)
K2CO3, benzene, 75°, 24 h. (e) Reference 22.
conversion of the enol ether side chain of 25 into the ethanamine
chain, led to 26, the precursor of the piperidine ring.
Intramolecular closure of the basic methylaminoethyl chain
of 26 by displacement of the mesylate now involved the possible
effect of the 6-methoxyl on the cyclization/elimination ratio from
26. We were pleasantly surprised that in refluxing benzene,
under the conditions used in the desmethoxy series (24 h in
benzene, at 75 °C bath temperature) 26 gave (() codeine methyl
ether 27 in 73% yield.
(() Thebaine. The specific formation of (() codeine methyl
ether just described permits that of (() thebaine in just one
additional step because the C-6 configuration of the (() methyl
ether of codeine (and not that of (() isocodeine) can take
advantage of the easy oxidation of the former to thebaine by
manganese dioxide described many years ago by Barber and
Rapoport22 who also reported that, by contrast, the epimeric
methyl ether of isocodeine was unchanged under their oxidation
conditions.
Codeine from Codeine Methyl Ether. The completion of our
synthetic project now required cleavage of the allylic methyl
ether of 27 either directly to codeine, or to codeinone, since
the reduction of codeinone, or its desN-methyl N-carbamate
(cf 28) to codeine is well-known,19 involving simple hydride
addition to the accessible face of the codeinone mole-
cule simultaneously with the reduction of the carbamate.
This proved to be more of a problem than we had
anticipated. Cleavage of the methyl ether proceeded very
readily, and selectively, with boron tribromide, but the allylic
Acknowledgment. The early part of this work was supported
by grants from the National Institutes of Health and the National
Science Foundation. Support has more recently come from private
grants, including the Arun Guthikonda Endowment, as well as from
Amgen, Inc. We also express our thanks to Drs. Somarajan Nair,
Kyoji Furuta, and Chengde Wu, as well as to Jonathan R. Scheerer
for valuable contributions to some aspects of this project. Thanks
are also due to Professor Gerard Parkin of this department, and to
several members of his group, who generously performed most of
the X-ray structure determinations mentioned in this paper.
Supporting Information Available: Complete ref 9; experi-
mental procedures and spectral data for all compounds from 5
onward. This material is available free of charge via the Internet
at http://pubs.acs.org.
(22) Barber, R. B.; Rapoport, H. J. J. Med. Chem. 1975, 18, 1074–1077.
JA9038505
9
11406 J. AM. CHEM. SOC. VOL. 131, NO. 32, 2009