2,6-Disubstituted and 2,2,6-trisubstituted piperidines from serine: Asymmetric synthesis and further elaboration

Article abstract of DOI:10.1021/jo101010c

(Figure presented) 4-Hydroxy-3,4-dihydro-2H-pyridine-1-carboxylic acid benzyl esters, which are readily prepared from serine and terminal acetylenes, undergo Claisen rearrangement to piperidine derivatives when heated with butyl vinyl ether in the presence of Hg(OAc)₂ and Et₃N. This route to optically pure piperidines having substituents α to nitrogen is general, and the rearrangement products are versatile intermediates for making a broad range of amines containing a substituted piperidine subunit.

Full text of DOI:10.1021/jo101010c

pubs.acs.org/joc
2,6-Disubstituted and 2,2,6-Trisubstituted Piperidines from Serine:
Asymmetric Synthesis and Further Elaboration
Hukum P. Acharya and Derrick L. J. Clive*
Department of Chemistry, University of Alberta, Edmonton, Alberta T6G 2G2, Canada
derrick.clive@ualberta.ca
Received May 23, 2010
4-Hydroxy-3,4-dihydro-2H-pyridine-1-carboxylic acid benzyl esters, which are readily prepared
from serine and terminal acetylenes, undergo Claisen rearrangement to piperidine derivatives when
heated with butyl vinyl ether in the presence of Hg(OAc)2 and Et3N. This route to optically pure
piperidines having substituents R to nitrogen is general, and the rearrangement products are versatile
intermediates for making a broad range of amines containing a substituted piperidine subunit.
Introduction
tuted piperidines2 are important in medicinal chemical
research3 and also feature in many natural products;4 there-
fore, we have made a more extensive study of this piperidine
synthesis, and the results are reported here, together with
a number of further transformations that lead to azaspiro
compounds and the indolizidine skeleton.
During studies on the total synthesis of the marine alka-
loid halichlorine,1 a method was developed in this laboratory
for constructing optically pure piperidines with a quaternary
center adjacent to nitrogen. The method was used to make
aldehyde 1, which was elaborated further for the halichlorine
project, but the related compounds 2 and 3 were also
prepared as evidence that the approach is general. Substi-
(1) Liu, D.; Acharya, H. P.; Yu, M.; Wang, J.; Yeh, V. S. C.; Kang, S.;
Chiruta, C.; Jachak, S. M.; Clive, D. L. J. J. Org. Chem. 2009, 74, 7417–7428.
(2) Synthesis of piperidines. Reviews: (a) Buffat, M. G. P. Tetrahedron
2004, 60, 1701–1729. (b) Felpin, F.-X.; Lebreton, J. Eur. J. Org. Chem. 2003,
3693–3712. (c) Weintraub, P. M.; Sabol, J. S.; Kane, J. M.; Borcherding, D. R.
Tetrahedron 2003, 59, 2953–2989. (d) Bates, R. W.; Sa-Ei, K. Tetrahedron 2002,
58, 5957–5978. (e) Enders, D.; Thiebes, C. Pure Appl. Chem. 2001, 73, 573–578.
(f) Husson, H.-P.; Royer, J. Chem. Soc. Rev. 1999, 28, 383–394. (g) Laschat, S.;
Dickner, T. Synthesis 2000, 1781–1813. (h) Deiters, A.; Martin, S. F. Chem.
Rev. 2004, 104, 2199–2238. (i) Rubiralta, M.; Giralt, E.; Diez, A. Piperidines,
Structure, Preparation, Reactivity, and Synthetic Applications of Piperidine
and its Derivatives; Elsevier: Amsterdam, 1991. Routes to non-racemic piper-
idines: (j) Coombs, T. C.; Zhang, Y.; Garnier-Amblard, E. C.; Liebeskind, L. S.
J. Am. Chem. Soc. 2009, 131, 876–877. (k) Lebold, T. P.; Leduc, A. B.; Kerr,
M. A. Org. Lett. 2009, 11, 3770–3772. (l) Abrunhosa-Thomas, I.; Roy, O.; Barra,
M.; Besset, T.; Chalard, P.; Troin, Y. Synlett 2007, 1613–1615. (m) Adriaenssens,
L. V.; Hartley, R. C. J. Org. Chem. 2007, 72, 10287–10290. (n) Gotchev, D. B.;
Comins, D. L. J. Org. Chem. 2006, 71, 9393–9402. (o) Back, T. G.; Nakajima, K.
Results and Discussion
Our route to piperidines starts with the construction of
dihydropyridinones of type 4. Stereoselective reduction of
the carbonyl group gives the tetrahydropyridinol 5, which
forms a vinyl ether that undergoes Claisen rearrangement5
(5 f 6 f 7) when heated in butyl vinyl ether in the presence of
Hg(OAc)2 and Et3N (Scheme 1).
A number of routes to the starting dihydropyridinones 4
are known,6 but the method of Turunen and Georg7 seemed
most appropriate and, with some modification, proved to be
€
J. Org. Chem. 1998, 63, 6566–6571. (p) Waldmann, H.; Braun, M.; Drager, M.
Angew. Chem., Int. Ed. 1990, 29, 1468–1471. (q) Kunz, H.; Pfrengle, W. Angew.
Chem., Int. Ed. Engl. 1989, 28, 1067–1068.
(3) For comments on the pharmaceutical significance of piperidines, see:
(4) See, for example: Daly, J. W.; Spande, T. F.; Garraffo, H. M. J. Nat.
Prod. 2005, 68, 1556–1575.
(5) Review of the Claisen rearrangement: Castro, A. M. M. Chem. Rev.
2004, 104, 2939–3002.
Watson, P. S.; Jiang, B.; Scott, B. Org. Lett. 2000, 2, 3679–3681.
DOI: 10.1021/jo101010c
r
Published on Web 06/30/2010
J. Org. Chem. 2010, 75, 5223–5233 5223
2010 American Chemical Society

Acharya and Clive
JOCArticle
SCHEME 1. General Plan
SCHEME 2. Route to Acetylenic Ketones
ideal for our purpose. A key example of their work is shown
in eq 1, the product 9 being formed with an ee of >90%.
structure seemed appropriate for the halichlorine synthesis.
When it was subjected to the first step of the reported
conditions for cyclization to a dihydropyridinone (4 N
HCl in dioxane, or Me3SiI in CH2Cl2)7 only a complex
mixture was obtained. Removal of the Boc group could be
achieved, however, with ZnBr2 to release the N-benzyl
amine 26, which was unstable and did not appear to cyclize
under the conditions we tried (K2CO3, MeOH; Na2CO3,
DMF), our impression being that it underwent loss of
benzylamine.
The initial requirements in the context of our route to
halichlorine were suggestive of serine as a starting material
for making the dihydropyridinones of type 4, and such a
route was readily developed,1 as summarized in Scheme 2. In
principle, amino acids with other subunits besides the CH2OH
of serine should also be appropriate starting materials, but
we have not investigated this possibility.
Following a literature procedure,8 serine methyl ester
hydrochloride was converted into the oxazolidinone 11,
which was treated successively with NaBH4 and TsCl
(11 f 12 f 13). The tosyloxy group was then displaced by a
Finkelstein reaction using NaI, and the resulting iodide 14
was treated with vinylmagnesium bromide in the presence
of CuBr SMe2. These operations gave the allyl-substituted
3
oxazolidinone 15,9 and ozonolysis then provided the key
aldehyde 16 that was used for much of our work. All of
the yields in the steps leading to 16 were satisfactory
(>79%).
The aldehyde reacted efficiently with a range of lithium
salts derived from terminal acetylenes to afford the expected
propargylic alcohols (16 f 17), which were oxidized with
MnO2 to generate ketones of type 18. These ketones were
then converted into the corresponding dihydropyridinones
4, as described later. Most of the acetylenic alcohols (of
type 17) and ketones (of type 18) that we made are listed in
Table 1, which also includes the two examples1 (entries i, ii)
investigated during the synthesis of halichlorine.
We next evaluated a different set of protecting groups by
making the acetylenic ketone 27. This compound did cyclize
to 28 (80%) when treated with Me3SiI, followed by workup
with aqueous NaHCO3 and then treatment with K2CO3 in
MeOH for 1 h at room temp, but the two attempts we made
to remove the isopropylidene group (FeCl3 SiO2;10 CuCl2
3
3
2H2O11) were unsuccessful.
Choice of Nitrogen Protecting Groups. Table 1 shows our
optimized approach, which is discussed in detail below;
however, we first examined the acetylenic ketone 25, whose
At the same time as these experiments were underway, we
also prepared the acetylene 29 and found that it could be
cyclized to 30 in good yield (84%) by a seemingly small but
crucial modification of the conditions reported by Turunen
(6) For two powerful modern methods, see: (a) Back, T. G. Can. J. Chem.
2009, 87, 1657–1674. (b) Back, T. G.; Hamilton, M. D.; Lim, V. J. J.; Parvez,
M. J. Org. Chem. 2005, 70, 967–972. (c) Comins, D. L. J. Heterocycl. Chem.
1999, 36, 1491–1500. (d) Comins, D. L.; Joseph, S. P.; Goehring, R. R. J. Am.
Chem. Soc. 1994, 116, 4719–4728. (e) Comins, D. L; Hong, H.; Salvador,
J. M. J. Org. Chem. 1991, 56, 7197–7199.
(7) Turunen, B. J.; Georg, G. I. J. Am. Chem. Soc. 2006, 128, 8702–8703.
(8) Sibi, M. P.; Rutherford, D.; Renhowe, P. A.; Li, B. J. Am. Chem. Soc.
1999, 121, 7509–7516.
(10) Kim, K. S.; Song, Y. H.; Lee, B. H.; Hahn, C. S. J. Org. Chem. 1996,
51, 404–407.
(9) We found this method more convenient than that reported in the
literature( Kim, S.-G.; Ahn, K. H. Synth. Commun. 1998, 28, 1387–1397).
(11) Saravanan., P.; Chandrasekhar, M.; Anand, R. V.; Singh, V. K.
Tetrahedron Lett. 1998, 39, 3091–3092.
5224 J. Org. Chem. Vol. 75, No. 15, 2010

Acharya and Clive
JOCArticle
TABLE 1. Cyclization to Dihydropyridinonesa
aReagents and conditions: (a) Tris(1-methylethyl)(4-pentynyloxy)silane, BuLi, THF, -78 °C. (b) MnO2, CH2Cl2. (c) Cs2CO3, MeOH, 65 °C, 12 h.
(d) t-BuMe2SiCl, Et3N, DMAP, CH2Cl2. (e) 1-Methoxy-4-[(2-propyn-1-yl-oxy)methyl]benzene, BuLi, THF, -78 °C. (f) Cs2CO3, MeOH, 80 °C, 6 h.
(g) (1,1-Dimethylethyl)(4-pentynyloxy)diphenylsilane, BuLi, THF, -78 °C. (h) 1-Pentyne, BuLi, THF, -78 °C. (i) Cs2CO3, MeOH, 80 °C, 5 h.
(j) Trimethylsilylacetylene, BuLi, THF, -78 °C. (k) Ethynylbenzene, BuLi, THF, -78 °C.
and Georg. This modification involved operating at a lower
temperature (-15 to -40 °C) and introducing i-PrNEt2
after the initial treatment (30 min at -15 to -40 °C) with
Me3SiI.
factory conditions for cyclization of 29 had been found,
we reexamined the acetylene 25. In the event, this com-
pound underwent cyclization satisfactorily (76%) using the
modified procedure but, unfortunately, attempts to reduce
the carbonyl group (DIBALH; NaBH4-CeCl3 7H2O;
3
(13) The same modified procedure also worked for conversion of i into
20d and ii into 23d, but we measured the ee (>99%) only for 20d.
€
In the absence of Hunig’s base only a trace of 30 was
detected after warming the mixture to room temperature,
removing volatile material, and treating the residue with
€
K2CO3 in MeOH. The inclusion of Hunig’s base is part of a
standard method for making vinyl iodides.12,13 Once satis-
(12) (a) Jasperse, C. P.; Curran, D. P. J. Am. Chem. Soc. 1990, 112, 5601–
5609. (b) Cheon, S. H.; Christ, W. J.; Hawkins, L. D.; Jin, H.; Kishi, Y.;
Taniguchi, M. Tetrahedron Lett. 1986, 27, 4759–4762.
J. Org. Chem. Vol. 75, No. 15, 2010 5225

Acharya and Clive
JOCArticle
DIBALH-[4-methyl-2,6-di(tert-butyl)phenol]; LiAlH4; Red-Al)
of the product (31) were not successful. Related dihydropyri-
the pendant originating from the terminal acetylene. In
contrast, other compounds in which an i-Pr3SiO or t-BuPh2-
SiO group was present on a three-carbon chain (as in 19a and
21a) were not desilylated.
dinones in which the nitrogen is protected as a carbamate
have been reduced,14 and so our next approach was to
prepare the cyclic carbamate 19b (Scheme 3), expecting that
it might afford 32. This expectation was not realized, as we
obtained 19c from the cyclization, which was done using
Cs2CO3-MeOH at reflux. The initial choice of Cs2CO3 was
arbitrary, but a subsequent brief optimization survey, using
20b, showed that the Cs2CO3-MeOH combination is one of
the best conditions we examined (Table 2). The liberation of
the hydroxyl and amino groups (as in 19c) was of no
consequence, as each could be reprotected.
The hydroxyl group of the cyclized products from
19c-24c was protected, usually by silylation (Table 1), and
the nitrogen was converted into a carbamate (Table 3) by
reaction with CbzCl or Boc2O, except for one case (Table 3,
entry iv, 19c f 19n), where triphosgene was used to mu-
tually protect the hydroxyl and amino functions. Introduc-
tion of the Cbz group was done by deprotonation with
BuLi, followed by reaction with CbzCl. The same proce-
dure with Boc2O appeared to be unsuccessful (at least with
19d), but the Boc group was easily introduced in all cases by
a more conventional procedure (Boc2O, DMAP, MeCN).
Attempts to protect the hydroxyl of 19c as a MOM ether were
unsuccessful and the MOM group was not examined with any
of our other substrates.
SCHEME 3. Cyclization and Hydrolysis of Oxazolidinone
Reduction of the Carbonyl Group. The fully protected
dihydropyridinones shown in Table 3 could be reduced
stereoselectively to the indicated R-alcohols, using NaBH4-
CeCl3 7H2O at a low temperature (-40 °C), the yield gene-
3
rally being >80%. In the case of 24e, reduction gave 8.5%
of the undesired β epimer and 71% of the desired R isomer.
The stereochemical outcome for the reductions 19e f 19f
(Table 3, entry i) and 19k f 19l (Table 3, entry iii) was est-
ablished by comparison of the 1H NMR spectra with those of
the C-4 epimeric alcohols 35 and 36, respectively, obtained
by use of DIBALH-[4-methyl-2,6-di(tert-butyl)phenol],14b,15
it being known14b that reduction of related dihyropyridi-
TABLE 2. Optimization Studies
nones with NaBH4-CeCl3 7H2O and di-isopropylalumi-
3
num hydride-[4-methyl-2,6-di(tert-butyl)phenol] (we used
DIABLH-[4-methyl-2,6-di(tert-butyl)phenol]15) gives oppo-
site stereochemical results. As expected for the R alcohol 19f,
bath
temp time
solvent/
additive
yield
(%)
base
(°C)
(h)
i
ii
iii
iv
v
Cs2CO3 MeOH
K2CO3 MeOH
Cs2CO3 CH2Cl2-H2O
K2CO3 CH2Cl2-H2O
K2CO3 PhMe
80
80
80
80
80
80
rt
5
6
6
5
5
5
12
5
79
80
no reaction
no reaction
no reaction
20b þ trace 20c
decomp of 20b, no 20c
decomp of 20b, no 20c
TROESY measurements showed an NOE enhancement
between the carbinyl hydrogen (CHOH) and the CH2 of
the CH2OCbz side chain. The stereochemical assignment
could not be made on the basis of coupling constants because
the presence of rotamers broadened the key signals. The
chemical shifts of the carbinyl hydrogen CHOH of the R and
β isomers are distinctly different. Access to the β alcohol 35
via Mitsunobu inversion (we did not try to make 36 in this
vi
Cs2CO3 MeCN
vii Cs2CO3 PhMe, Aliquat
viii Cs2CO3 PhMe, Aliquat
0
The conditions used (see Scheme 3) to make 19c were
applied to all the other ketones listed in Table 1 to afford the
unprotected hydroxy amines 20c-24c, respectively. We did
find one restriction on the structure of the starting acetylenic
ketones: attempts to apply our standard cyclization condi-
tions to 33 and 34 showed that neither silicon protecting
group survived, and so we used other forms of protection for
(14) (a) Comins, D. L.; Foley, M. A. Tetrahedron Lett. 1988, 29, 6711–
ꢀ
6714. (b) Comins, D. L.; Kuethe, J. T.; Miller, T. M.; Fevrier, F. C.; Brooks,
C. A. J. Org. Chem. 2005, 70, 5221–5234.
(15) Iguchi, S.; Nakai, H.; Hayashi, M.; Yamamoto, H.; Maruoka, K.
Bull. Chem. Soc. Jpn. 1981, 64, 3033–3041.
5226 J. Org. Chem. Vol. 75, No. 15, 2010

Acharya and Clive
JOCArticle
TABLE 3. Claisen Rearrangement
aSi* = t-BuMe2Si. Reagents and conditions: (a) BuLi, CbzCl, THF, -78 °C. (b) NaBH4, CeCl3 7H2O, MeOH, -40 C, 30 min. (c) Hg(OAc)2,
3
Et3N, n-butyl vinyl ether, 80 °C, 36 h. (d) (Boc)2O, DMAP, MeCN, 6 h. (e) Hg(OAc)2, Et3N, n-butyl vinyl ether, 110 °C, 38 h. (f) NaBH4, CeCl3 7H2O,
3
MeOH, -40 C, 45 min. (g) Hg(OAc)2, Et3N, n-butyl vinyl ether, 110 °C, 36 h. (h) BuLi, triphosgene, THF, -78 °C. (i) NaBH4, CeCl3 7H2O, MeOH, -40
3
C, 1 h. (j) Hg(OAc)2, Et3N, n-butyl vinyl ether, 110 °C, 48 h. (k) (Boc)2O, DMAP, MeCN, 5 h. b8.5% of β-alcohol also isolated.
way) was not successful because of the instability of the
intermediate ester, but reduction of 19k with DIBALH-[4-
methyl-2,6-di(tert-butyl)phenol] was satisfactory, although
the yield was lower than that obtained in the reduction with
(syn to the silyloxy substituent.
NaBH4-CeCl3 7H2O leading to the isomeric R alcohol
3
19l. We assume that all of the Luche reductions involve the
conformation 37, which is adopted to alleviate A1,3 strain.
If complexation of the carbonyl occurs anti to the silyloxy
substituent, then hydride delivery would occur axially
Optical Purity of 20d and 19d. As described in our pub-
lication on the synthesis of halichlorine,1 the optical purity of
J. Org. Chem. Vol. 75, No. 15, 2010 5227

Acharya and Clive
SCHEME 4. Formation of an Azaspiro[4,5]decadiene
JOCArticle
20d was measured by N-acylation with (S)-(þ)-R-methoxy-
R-trifluoromethylphenylacetyl chloride to give a product
with a vinyl signal in the H NMR spectrum (400 MHz) at
1
5.93 ppm. Acylation of 20d with the enantiomeric reagent gave
a signal at 5.88 ppm; this signal is absent in the spectrum of the
derivative made with the S-reagent and so 20d must be a single
enantiomer. The same procedure was used to establish1 that
19d had an ee of >99%, and we assume that the other dihydro-
pyridinones are likewise formed without compromising the
stereochemical purity of the starting serine. We also hydrolyzed
the cyclic carbamate 19n (derived from serine with ee of 97%)
back to the parent dihydropyridinone 19c using the conditions
we employed for the original cyclization 19b f 19c and showed
(using the above Mosher amide procedure, after O-silylation)
that there was no loss of optical purity.
Claisen Rearrangement. The tetrahydropyridinols listed in
Table 3 (third column of structures) reacted with butyl vinyl
ether on heating for ca. 36 h in the presence of Hg(OAc)2 and
Et3N to afford the product of Claisen rearrangement. Ob-
viously, an intermediate vinyl ether is generated and rear-
ranges in situ (cf. 5 f 6 f 7). Our attempts to isolate such an
intermediate, using 19l, were unsuccessful, as the compound
appeared (1H NMR) to suffer elimination of the vinyloxy
group during chromatography. The formation and rearran-
gement of the vinyl ethers were all conducted by heating the
substrate alcohol in butyl vinyl ether (bp 94 °C) as solvent,
using a Teflon-sealed thick-walled tube partially immersed
in an oil bath set at 100-115 °C. When the reaction was
attempted with 19f in refluxing butyl vinyl ether under Ar, but
at 1 atm, some of the derived vinyl ether remained (1H NMR)
after the normal reaction time of 40 h (but could not be
isolated). We did try a higher-boiling vinyl ether (cyclohexyl
vinyl ether, bp 147 °C) at a bath temperature of 140 °C to avoid
the use of a sealed vessel, but the reaction, at least with 19f and
20f, was slower and this protocol did not appear to offer any
advantage. The use of Et3N in the Claisen rearrangement was
simply a precaution suggested by prior observations with
Pd(OCOCF3)2-mediated Claisen rearrangements;16 in a single
experiment done at a somewhat lower temperature without the
amine, the yield was significantly lower.
The alcohol stereochemistry is important and the R-alcohol
must be used; tests with the β-alcohols 35 and 36 were
unsuccessful, presumably because Claisen rearrangement re-
quires two substituents to adopt an axial conformation,
whereas only one substituent need be axial for rearrangement
of the vinyl ether derived from the R-alcohol. This situation is
illustratedin diagrams 38 and 39. The restriction toR-alcohols
means that a careful choice must be made for the substituent
on the acetylene used in constructing the dihydropyridinones
of type 4 (group R) so that, after the Claisen rearrangement,
the appendages on the fully substituted carbon R to the
nitrogen can be modified in the desired way.
Claisen rearrangement. An acyclic version of a related
Ireland ester enolate rearrangement has been reported,17
and it was noted that the intermediate esters are unstable
to chromatography. In the acyclic series the diastereoselec-
tivity was usually in the range 2:1 to 3:2 (with one exception,
95:5).
Alcohol 19l was converted into its acetate 19q by DCC-
mediated coupling; acylation using Ac2O/Et3N was unsuc-
cessful because the derived acetate decomposed in situ. The
material could not be purified because of its instability to
silica chromatography, but it did rearrange on treatment with
(Me3Si)2NLi in the presence of Me3SiCl and HMPA in THF
(-78 to þ70 °C). The yield of ester obtained by methylation
of the resulting acid was only 46%. Use of (Me3Si)2NNa
under the same conditions gave a comparable yield of ester
(43%). Corresponding experiments were done with the meth-
oxy acetate 19s and the (phenylthio)acetate 19u, each made by
DCC coupling, but yields of the derived esters 19t and 19v
were always below 60%; consequently, this route was not
investigated further, even though the products do have func-
tionality that is potentially useful for further elaboration.
Elaboration of the Claisen Rearrangement Products. Most
of the Claisen rearrangements shown in Table 3 afford
compounds with two pendant functionalized chains R to
nitrogen. These chains should be modifiable in a number of
ways, and we have illustrated a few of the possibilities. Wittig
homologation of 20g with methylenetriphenylphosphorane
gave 40 (Scheme 4), and the PMB group was then removed.
Oxidation of the resulting alcohol to the unstable aldehyde
42 and reaction with vinylmagnesium bromide generated
dienyl alcohol 43, which was obtained as a single isomer
whose stereochemistry at the hydroxyl-bearing carbon was
not established. Ring-closing metathesis (Grubbs II) gave
the azaspirocycle 44 in just over 50% yield overall from 41.
(16) Wei, X.; Lorenz, J. C.; Kapadia, S.; Saha, A.; Haddad, N.; Busacca,
C. A.; Senanayake, C. H. J. Org. Chem. 2007, 72, 4250–4253.
(17) Ylioja, P. M.; Mosley, A. D.; Charlot, C. E.; Carbery, D. R.
Tetrahedron Lett. 2008, 49, 1111–1114.
Use of the Ireland Ester Enolate Rearrangement. Using 19l
as a test substrate, we examined the possibility of applying
the Ireland ester enolate rearrangement instead of the
5228 J. Org. Chem. Vol. 75, No. 15, 2010

Acharya and Clive
JOCArticle
TABLE 4. Ireland Ester Enolate Rearrangement
SCHEME 5. Preparation of an Azaspiro[5,5]undecadiene
SCHEME 6. Formation of an Indolizidine
the resulting imine would then undergo hydrogenation (47 f
50). To our surprise, an experiment along these lines (eq 5)
failed to afford the indolizidine and gave only a complex
mixture, but the indolizidine structure could be reached by
the steps summarized in Scheme 6. Pinnick oxidation of the
aldehyde group (47 f 51) and methylation of the resulting
ester gave 52, and at that point the Cbz groups were removed
by hydrogenolysis. Finally, the required cyclization to gen-
erate the indolizidine18 skeleton was induced by heating
amine 53 in PhMe.
As summarized in Scheme 5 the related azaspirocycle 49
was accessible from 19m, again using conventional proce-
dures. Wittig homologation of 19m (19m f 45), followed by
deprotection, oxidation, and Wittig homologation of the
siloxypropyl side arm (45 f 46 f 47 f 48) set the stage for
ring-closing metathesis, which proceeded in high yield (89%)
to afford the spirocycle 49. We found that both Wittig re-
actions must be done at a low temperature and for 30 min
only, in order to avoid loss of the Cbz groups.
Finally, we converted the olefinic aldehyde 47 into the
simple indolizidine 54 (Scheme 6). Our expectation (eq 5) was
that simple hydrogenation of 47 would serve to saturate the
carbon-carbon double bonds, remove the Cbz groups, and
allow the nitrogen to cyclize onto the pendant aldehyde group;
(18) For recent approaches to indolizidines, see, for example: (a) Reference
6a. (b) Friedman, R. K.; Rovis, T. J. Am. Chem. Soc. 2009, 131, 10775–10782.
(c) Barbe, G.; Pelletier, G.; Charette, A. B. Org. Lett. 2009, 11, 3398–3401.
(d) Reviews: Perreault, S.; Rovis, T. Chem. Soc. Rev. 2009, 38, 3149–3159.
(e) Michael, J. P. Nat. Prod. Rep. 2008, 25, 139–165.
Mechanistic Studies. As described in the full report on the
synthesis of halichlorine,1 a brief mechanistic study indi-
cated, at least for the test example (19b), that the pathway
J. Org. Chem. Vol. 75, No. 15, 2010 5229

Acharya and Clive
JOCArticle
SCHEME 7. Attempted Isolation of Intermediates
5 H), 2.33-2.41 (m, 2 H), 3.73 (t, J = 6.0 Hz, 2 H), 3.97-4.23 (m,
2 H), 4.43-4.60 (m, 2 H), 5.47-5.57 (m, 1 H), 7.34-7.48 (m, 6 H),
7.62-7.70 (m, 4 H); exact mass (electrospray) m/z calcd for
C26H33NNaO4Si (M þ Na) 474.2071, found 474.2071.
(4R)-4-[7-[[(1,1-Dimethylethyl)diphenylsilyl]oxy]-2-oxo-3-heptyn-
1-yl]-2-oxazolidinone (21b). Activated MnO2 (0.578 g, 6.67 mmol)
was added to a stirred solution of 21a (0.31 g, 0.69 mmol) in
CH2Cl2 (8 mL). Stirring was continued for 1.5 h, and the slurry
was filtered through a short pad of Celite (3 ꢀ 3 cm), using EtOAc
(3 ꢀ 10 mL) as a rinse. Evaporation of the solvent and flash
chromatography of the residue over silica gel (2 ꢀ 12 cm), using
1:5 EtOAc-hexane, gave 21b (0.253 g, 81% yield) as a colorless
oil: [R]25D þ9.1 (c 0.75, CHCl3); FTIR (CHCl3 cast) 3294, 2214,
1758, 1671 cm-1;1H NMR (CDCl3, 400 MHz) δ1.06 (s, 9 H), 1.83
(quintet, J = 7.0 Hz, 2 H), 2.56 (t, J = 7.0Hz, 2 H), 2.75-2.90 (m,
2 H), 3.75 (t, J = 6.0 Hz, 2 H), 3.98 (dd, J = 9.0, 6.0 Hz, 1 H),
4.18-4.26 (m, 1 H), 4.54 (t, J = 9.0 Hz, 1 H), 5.42 (br s, 1 H),
7.35-7.48 (m, 6 H), 7.63-7.70 (m, 4 H); 13C NMR (CDCl3, 100
MHz) δ15.7(t), 19.2(s), 26.8 (q), 30.4(t), 48.0(d), 50.3 (t), 62.0 (t),
69.2 (t), 80.4 (s), 96.6 (s), 127.7 (d), 129.7 (d), 133.5 (s), 135.5 (d),
158.6 (s), 184.2 (s); exact mass (electrospray) m/z calcd for
C26H31NNaO4Si (M þ Na) 472.1915, found 472.1913.
probably was via 55 (arbitrary double bond geometry
shown), and although we could not specify the timing of
the oxazolidinone hydrolysis, we suspect it happens after
cyclization because 19n gave 19c on exposure to the cycliza-
tion conditions, but the reference model 15 (see Scheme 2)
was largely unchanged under these conditions; 19n is an
imide, whereas 15 is a carbamate, and we would expect the
hydrolytic stability of 15 to accurately reflect the hydrolytic
stability of 19b, while an imide would be more sensitive to
hydrolysis. When 55 was subjected to the cyclization condi-
tions, but only for 1.5 h, so that some 55 remained, only 19c
and 55 were detectable in the mixture (1H NMR), apart from
obviously minor impurities.
(2R)-6-[3-[[(1,1-Dimethylethyl)diphenylsilyl]oxy]propyl]-2,3-
dihydro-2-(hydroxymethyl)-4(1H)-pyridinone (21c). Cs2CO3 (0.432 g,
1.33 mmol) was added in two portions (at start and after 20 min)
to a stirred solution of 21b (0.205 g, 0.456 mmol) in MeOH
(9 mL). Stirring was continued for 1 h after the second addition
and then at 80 °C for 3 h. The solution was cooled and eva-
porated. The residue was diluted with water (10 mL) and EtOAc
(10 mL), and the aqueous phase was extracted with EtOAc
(2 ꢀ 10 mL). The combined organic extracts were dried (MgSO4)
and evaporated. Flash chromatography of the yellow residue
over silica gel (2 ꢀ10 cm), using 1:10 MeOH-EtOAc, gave 21c
(0.135 g, 70% yield) as a greasy solid: [R]25D -81.3 (c 0.57, CHCl3);
FTIR (CHCl3 cast) 3263, 1602, 1570, 1524 cm-1; 1H NMR (CDCl3,
500 MHz) δ 1.07 (s, 9 H), 1.80 (quintet, J = 6.0 Hz, 2 H), 2.20-2.42
(m, 5H), 3.61(t, J= 11.0 Hz, 1 H), 3.66-3.78(m, 4H), 4.99(s, 1H),
5.57 (br s, 1 H), 7.35-7.48 (m, 6 H), 7.63-7.70 (m, 4 H); 13C NMR
(CDCl3, 125 MHz) δ 19.2 (s), 26.9 (q), 30.4 (t), 31.5 (t), 37.2 (t),
54.4 (d), 62.6 (t), 64.2 (t), 98.4 (d), 127.7 (d), 129.8 (d), 133.5 (s),
135.5 (d), 165.9 (s), 191.6 (s); exact mass (electrospray) m/z calcd for
C25H33NNaO3Si (M þ Na) 446.2122, found 446.2125.
Conclusion
Heating readily available and optically pure 4-hydroxy-
3,4-dihydro-2H-pyridine-1-carboxylic acid benzyl esters in
butyl vinyl ether is a general method for preparing piperidine
derivatives having a quaternary center adjacent to nitrogen.
The method, which was developed specifically for the synthe-
sis of the complex marine alkaloid halichlorine, involves
Claisen rearrangement of a vinyl ether that is generated in
situ. Corresponding Ireland ester enolate rearrangements are
much less efficient. The substituted piperidines can be ela-
borated in a number of ways, and several possibilities were
illustrated in the present work by making two spiro com-
pounds and an indolizidine.
(2R)-2-[[[(1,1-Dimethylethyl)dimethylsilyl]oxy]methyl]-6-[3-
[[(1,1-dimethylethyl)diphenylsilyl]oxy]propyl]-2,3-dihydro-4(1H)-
pyridinone (21d). DMAP (8.3 mg, 0.067 mmol), Et3N (0.15 mL,
1.1 mmol), and t-BuMe2SiCl (38.3 mg, 0.255 mmol) were added
to a stirred solution of 21c (90.1 mg, 0.212 mmol) in CH2Cl2
(4 mL), and stirring was continued overnight. The mixture was
quenched with saturated aqueous NaHCO3, and the aqueous
phase was extracted with Et2O (2 ꢀ 10 mL). The combined
organic extracts were washed with brine, dried (MgSO4), and
evaporated. Flash chromatography of the residue over silica gel
(1 ꢀ 10 cm), using 1:4 EtOAc-hexane, gave 21d (86.5 mg, 78%
Experimental Section
(4R)-4-[7-[[(1,1-Dimethylethyl)diphenylsilyl]oxy]-2-hydroxy-
3-heptyn-1-yl]-2-oxazolidinone (21a). n-BuLi (2.5 M in THF,
1.02 mL, 2.56 mmol) was added dropwise to a stirred and cooled
(-78 °C) solution of (1,1-dimethylethyl)(4-pentynyloxy)-
diphenylsilane19 (0.753 g, 2.33 mmol) in THF (8 mL). Stirring at
-78 °C was continued for 30 min, and a solution of 16 (0.148 g,
1.15 mmol) in THF (2 mL plus 1 mL as a rinse) was added
dropwise. Stirring at -78 °C was continued for 2 h, and then
saturated aqueous NH4Cl (10 mL) was added. The mixture was
extracted with EtOAc (3 ꢀ 10 mL), and the combined organic
extracts were dried (MgSO4) and evaporated. Flash chromato-
graphy of the residue over silica gel (2 ꢀ 10 cm), using 1:5
EtOAc-hexane, gave 21a (0.427 g, 81%) as an oil which was a
mixture of two isomers: FTIR (CHCl3 cast) 3448, 3317, 2210, 1748
cm-1; 1H NMR (CDCl3, 400 MHz) δ 1.06 (s, 9 H), 1.71-2.02 (m,
yield) as a colorless oil: [R]25 -90.5 (c 1.40, CHCl3); FTIR
D
1
(CHCl3 cast) 3244, 1612, 1578, 1531 cm-1; H NMR (CDCl3,
500 MHz) δ 0.08 (s, 3 H), 0.09 (s, 3 H), 0.91 (s, 9 H), 1.06 (s, 9 H),
1.75-1.82 (m, 2 H), 2.19 (dd, J = 16.5, 12.5 Hz, 1 H), 2.27-2.39
(m, 3 H), 3.60 (t, J = 10.5 Hz, 1 H), 3.64-3.75 (m, 4 H), 4.98 (s, 1
H), 5.18 (br s, 1 H), 7.36-7.47 (m, 6 H), 7.62-7.69 (m, 4 H); 13
C
NMR (CDCl3, 125 MHz) δ -5.4 (q), -5.3 (q), 18.3 (s), 19.3 (s),
25.92 (q), 26.9 (q), 30.5 (t), 31.7 (t), 37.6 (t), 54.5 (d), 62.7 (t), 64.9
(t), 98.7 (d), 127.8 (d), 129.8 (d), 133.6 (s), 135.57 (d), 135.58 (d),
165.2 (s), 191.7 (s); exact mass (electrospray) m/z calcd for
C31H47NNaO3Si2 (M þ Na) 560.2987, found 560.2985.
(2R)-2-[[[(1,1-Dimethylethyl)dimethylsilyl]oxy]methyl]-6-[3-
[[(1,1-dimethylethyl)diphenylsilyl]oxy]propyl]-3,4-dihydro-4-oxo-
1(2H)-pyridinecarboxylic Acid Phenylmethyl Ester (21e). n-BuLi
(2.5 M in hexane, 60 μL, 0.15 mmol) was added dropwise to
(19) Baldwin, J. E.; Romeril, S. P.; Lee, V.; Claridge, T. D. W. Org. Lett.
2001, 3, 1145–1148.
5230 J. Org. Chem. Vol. 75, No. 15, 2010

Acharya and Clive
JOCArticle
a stirred and cooled (-78 °C) solution of 21d (75.3 mg, 0.139
mmol) in THF (2 mL). Stirring at -78 °C was continued for
5 min, and CbzCl (25 μL, 0.177 mmol) was added in one lot.
Stirring at -78 °C was continued for 15 min, and then MeOH
(1 mL) and saturated aqueous NH4Cl (5 mL) were added. The
cooling bath was removed, and the mixture was allowed to
warm to room temperature (ca. 15 min). The aqueous phase was
extracted with EtOAc (2 ꢀ 10 mL), and the combined organic
extracts were washed with brine (5 mL), dried (MgSO4), and
evaporated. Flash chromatography of the residue over silica gel
(1 ꢀ 10 cm), using 1:8 EtOAc-hexane, gave 21e (79.2 mg, 84%)
as an oil: [R]25D þ92.0 (c 1.08, CHCl3); FTIR (CHCl3 cast) 1725,
1669, 1596 cm-1; 1H NMR (CDCl3, 500 MHz) δ -0.02 (s, 3 H),
-0.01 (s, 3 H), 0.84 (s, 9 H), 1.06 (s, 9 H), 1.71 (quintet, J = 7.0
Hz, 2 H), 2.50-2.62 (m, 2 H), 1.74 (dd, J = 17.5, 6.5 Hz, 1 H),
3.01-3.10 (m, 1 H), 3.54-3.65 (m, 3 H), 3.72 (dd, J = 10.0, 7.0
Hz, 1 H), 4.80-4.85 (m, 1 H), 5.20 (AB q, J = 12.0, ΔνAB = 27.5
Hz, 2 H), 5.42 (s, 1 H), 7.32-7.47 (m, 11 H), 7.62-7.69 (m, 4 H);
13C NMR (CDCl3, 125 MHz) δ -5.6 (q), 18.1 (s), 19.2 (s), 25.7
(q), 26.9 (q), 31.3 (t), 32.4 (t), 37.3 (t), 56.8 (d), 60.1 (t), 62.9 (t),
68.7 (t), 112.5 (d), 127.7 (d), 128.5 (d), 128.65 (d), 128.70 (d),
129.6 (d), 133.7 (s), 135.1 (s), 135.5 (d), 153.1 (s), 158.7 (s), 192.9
(s); exact mass (electrospray) m/z calcd for C39H53NNaO5Si2
(M þ Na) 694.3355, found 694.3353.
(0.150 g, 1.16 mmol) in THF (2 mL plus 1 mL as a rinse) was
added dropwise. Stirring at -78 °C was continued for 1.5 h, and
then saturated aqueous NH4Cl (8 mL) and EtOAc (10 mL) were
added. The mixture was extracted with EtOAc (3 ꢀ 10 mL),
and the combined organic extracts were dried (MgSO4) and
evaporated. Flash chromatography of the residue over silica gel
(2 ꢀ 10 cm), using 1:2 EtOAc-hexane, gave 23a (0.217 g,
82%) as a colorless oil which was a mixture of two isomers:
FTIR (CHCl3 cast) 3381, 3270, 2180, 1728 cm-1 1H NMR
;
(CDCl3, 300 MHz) δ 0.18 (s, 9 H), 1.82-2.30 (m, 2 H), 2.49 (br s,
0.5 H), 2.67 (br s, 0.5 H), 4.00-4.35 (m, 2 H), 4.50-4.65 (m,
2 H), 5.90 (brs, 0.5H), 6.08 (brs, 0.5H);exactmass(electrospray)
m/z calcd for C10H17NNaO3Si (M þ Na) 250.0870, found
250.0869.
(4R)-4-[2-Oxo-4-(trimethylsilyl)-3-butyn-1-yl]-2-oxazolidinone
(23b). Activated MnO2 (0.803 g, 9.24 mmol) was added to a
stirred solution of 23a (0.211 g, 0.928 mmol) in CH2Cl2 (20 mL).
Stirring was continued for 2 h, and the slurry was filtered
through a short pad of Celite (1.5 ꢀ 2 cm), using EtOAc (3 ꢀ
10 mL) as a rinse. Evaporation of the solvent and flash chro-
matography of the residue over silica gel (2 ꢀ 15 cm), using 1:4
EtOAc/hexane, gave 23b (0.185 g, 89% yield) as an oil: [R]25
D
þ20.0 (c 0.63, CHCl3); FTIR (CHCl3 cast) 3260, 3161, 2146,
1739, 1677 cm-1; 1H NMR (CDCl3, 500 MHz) δ 0.26 (s, 9 H),
2.93 (dd, J = 15.0, 6.0 Hz, 1 H), 2.99 (dd, J = 18.5, 8.0 Hz, 1 H),
4.02 (dd, J = 9.0, 6.0 Hz, 1 H), 4.26 (quintet, J = 6.0 Hz, 1 H),
4.57 (t, J = 9.0 Hz, 1 H), 5.67 (br s, 1 H); 13C NMR (CDCl3, 125
MHz) δ -0.9 (q), 48.0 (d), 50.3 (t), 69.3 (t), 100.8 (s), 101.1 (s),
158.8 (s), 183.9 (s); exact mass (electrospray) m/z calcd for
C10H15NNaO3Si (M þ Na) 248.0713, found 248.0712.
(2R,4S)-2-[[[(1,1-Dimethylethyl)dimethylsilyl]oxy]methyl]-6-
[3-[[(1,1-dimethylethyl)diphenylsilyl]oxy]propyl]-3,4-dihydro-4-
hydroxy-1(2H)-pyridinecarboxylic Acid Phenylmethyl Ester
(21f). NaBH4 (8.3 mg, 0.216 mmol) was added to a stirred and
cooled (-40 °C) slurry of 21e (65.0 mg, 0.097 mmol) and
CeCl3 7H2O (79.4 mg, 0.213 mmol) in MeOH (5 mL). Stirring
3
(2R)-2-[[[(1,1-Dimethylethyl)dimethylsilyl]oxy]methyl]-2,3-di-
hydro-4(1H)-pyridinone (23d). Cs2CO3 (0.266 g, 0.817 mmol)
was added in two portions (at start and after ca. 10 min) to a
stirred solution of 23b (46.2 mg, 0.205 mmol) in MeOH (8 mL).
Stirring was continued for 1 h after the second addition and then
at 80 °C. When the reaction was complete (6 h, TLC control),
the solution was cooled and evaporated, and the residue (23c)
was kept under oilpump vacuum for 2 h to remove MeOH
completely.
at -40 °C was continued for 30 min, and the mixture was
quenched with acetone (0.1 mL) and saturated aqueous NH4Cl.
The cold bath was removed and replaced by an ice bath, and
stirring was continued for 30 min. The aqueous phase was
extracted with Et2O (2 ꢀ 10 mL), and the combined organic
extracts were dried (MgSO4) and evaporated. Flash chroma-
tography of the residue over silica gel (1 ꢀ 10 cm), using 4:1
hexane-EtOAc, gave 21f (44.0 mg, 68% yield) as an oil, which
was used directly in the next step.
DMAP (7.7 mg, 0.063 mmol), Et3N (0.14 mL, 1.0 mmol) and
t-BuMe2SiCl (62.2 mg, 0.408 mmol) were added to a stirred
solution of the above crude 23c in CH2Cl2 (5 mL), and stirring
was continued overnight. The mixture was quenched with satu-
rated aqueous NaHCO3 (5 mL), and the aqueous phase was
extracted with EtOAc (2 ꢀ 10 mL). The combined organic
extracts were washed with brine, dried (MgSO4), and evapo-
rated. Flash chromatography of the residue over silica gel
(1 ꢀ 10 cm), using 1:1 EtOAc-hexane, gave 23d (32.3 mg, 65%
yield over the two steps) as an oil: [R]25D -203.5 (c 0.39, CHCl3);
FTIR (CHCl3 cast) 3254, 1627, 1578 cm-1; 1H NMR (CDCl3,
500 MHz) δ 0.08 (s, 3 H), 0.09 (s, 3 H), 0.91 (s, 9 H), 2.29 (dd, J =
16.0, 13.0 Hz, 1 H), 2.34 (dd, J = 16.0, 5.5 Hz, 1 H), 3.62 (t, J =
10 Hz, 1 H), 3.71 (dd, J = 10.0, 4.0 Hz, 1 H), 3.71-3.80 (m, 1 H),
5.02 (d, J = 7.5 Hz, 1 H), 5.32 (br s, 1 H), 7.20 (t, J = 7.5 Hz, 1
H); 13C NMR (CDCl3, 125 MHz) δ -5.5 (q), 18.3 (s), 25.8 (q),
38.3 (t), 54.6 (d), 64.9 (t), 99.6 (d), 150.6 (d), 191.9 (s); exact mass
(electrospray) m/z calcd for C12H24NO2Si (M þ H) 242.1571,
found 242.1573.
(2R)-2-[[[(1,1-Dimethylethyl)dimethylsilyl]oxy]methyl]-3,4-di-
hydro-4-oxo-1(2H)-pyridinecarboxylic Acid Phenylmethyl Ester
(23e). n-BuLi (2.5 M in hexane, 0.112 mL, 0.282 mmol) was
added dropwise to a stirred and cooled (-78 °C) solution of 23d
(61.2 mg, 0.253 mmol) in THF (3 mL). Stirring -78 °C was
continued for 5 min, and CbzCl (47 μL, 0.33 mmol) was added in
one lot. Stirring at -78 °C was continued for 15 min, and then
hexane (5 mL) and saturated aqueous NH4Cl (5 mL) were
added. The cooling bath was removed, and the mixture was
allowed to warm to room temperature (ca. 15 min). The aqueous
(2S,6R)-6-[[[(1,1-Dimethylethyl)dimethylsilyl]oxy]methyl]-2-[3-
[[(1,1-dimethylethyl)diphenylsilyl]oxy]propyl]-5,6-dihydro-2-(2-
oxoethyl)-1(2H)-pyridinecarboxylic Acid Phenylmethyl Ester
(21g). A mixture of 21f (39.0 mg, 0.058 mmol), Hg(OAc)2
(1.5 mg, 0.005 mmol), and Et3N (1 drop) in n-butyl vinyl ether
(2 mL) was heated at 110 °C (oil bath temperature) in a sealed
glass tube (Teflon seal) for 36 h. Volatile material was evapo-
rated, and flash chromatography of the residue over silica gel
(1 ꢀ 10 cm), using 1:5 EtOAc-hexane, gave 21g (28.6 mg, 69%)
as a pale yellow oil: [R]25D -11.3 (c 0.35, CHCl3); FTIR (CHCl3
cast) 1722, 1700, cm-1; 1H NMR (CDCl3, 500 MHz) δ -0.04 (s,
3 H), -0.02 (s, 3 H), 0.85 (s, 9 H), 1.05 (s, 9 H), 1.35-1.46 (m,
2 H), 1.71-1.86 (m, 1 H), 2.10-2.24 (m, 2 H), 2.40-2.51 (m,
2 H), 3.41 (t, J = 10 Hz, 1 H), 3.42-3.51 (m, 1 H), 3.51-3.70 (m,
3 H), 4.30-4.40 (m, 1 H), 5.13 (AB q, J = 12.5, ΔνAB = 19.0 Hz,
2 H), 5.59 (dd, J = 10.0, 2.5 Hz, 1 H), 5.88 (t, J = 10.5 Hz, 1 H),
7.26-7.46 (m, 11 H), 7.62-7.69 (m, 4 H), 9.59 (br s, 1 H); 13
C
NMR (CDCl3, 125 MHz) δ -5.5 (q), -5.4 (q), 18.2 (s), 19.2 (s),
23.8 (t), 25.8 (q), 26.9 (q), 27.6 (t), 35.1 (s), 51.3 (t), 52.3 (t), 58.6
(d), 63.4 (t), 63.5 (t), 67.1 (t), 123.8 (d), 127.7 (d), 127.9 (d), 128.1
(d), 128.5 (d), 128.6 (d), 129.6 (d), 131.8 (d), 133.9 (d), 135.5 (d),
136.4 (s), 149.4 (s), 154.9 (s), 201.4 (d); exact mass (electrospray)
m/z calcd for C41H57NNaO5Si2 (M þ Na) 722.3668, found
722.3669.
(4R)-4-[2-Hydroxy-4-(trimethylsilyl)-3-butyn-1-yl]-2-oxazoli-
dinone (23a). n-BuLi (2.5 M in THF, 0.700 μL, 1.75 mmol) was
added dropwise to a stirred and cooled (-78 °C) solution of
trimethylsilylacetylene (0.245 mL, 1.74 mmol) in THF (8 mL).
Stirring at -78 °C was continued for 30 min, and a solution of 16
J. Org. Chem. Vol. 75, No. 15, 2010 5231

Acharya and Clive
JOCArticle
phase was extracted with EtOAc (3 ꢀ 5 mL), and the combined
organic extracts were washed with brine, dried (MgSO4), and
evaporated. Flash chromatography of the residue over silica gel
(1 ꢀ 8 cm), using 1:4 EtOAc-hexane, gave 23e (73.9 mg, 78%)
as an oil: [R]25D þ32.4 (c 0.51, CHCl3); FTIR (CHCl3 cast) 1728,
1676, 1607 cm-1; 1H NMR (CDCl3, 500 MHz) δ -0.01 (s, 6 H),
0.85 (s, 9 H), 2.03 (d, J = 17.0, Hz, 1 H), 2.80 (dd, J = 17.0, 7.5
Hz, 1 H), 3.62-3.74 (m, 2 H), 4.60-4.70 (m, 1 H), 5.18-5.41 (m,
3 H), 7.34-7.44 (m, 5 H), 7.72-7.88 (m, 1 H); 13C NMR
(CDCl3, 125 MHz) (one signal is missing) δ -5.6 (q), 18.2 (s),
25.7 (q), 36.9 (t), 54.2 (d), 61.7 (t), 69.0 (t), 107.3 (d), 128.5 (d),
128.6 (d), 128.7 (d), 128.8 (d), 135.0 (s), 141.7 (d), 192.4 (s); exact
mass (electrospray) m/z calcd for C20H29NNaO4Si (M þ Na)
398.1758, found 398.1758.
extracted with Et2O (2 ꢀ 5 mL), and the combined organic extr-
acts were dried (MgSO4) and evaporated. Flash chromatogra-
phy of the residue over silica gel (2 ꢀ 10 cm), using 1:5 EtOAc-
hexane, gave 40 (50.1 mg, 85%) as a colorless oil: [R]25D -18.3
(c 0.68, CHCl3); FTIR (CHCl3 cast) 1700, 1670, 1613 cm-1; 1H
NMR (CDCl3, 400 MHz) δ -0.08 (s, 6 H), 0.84 (s, 9 H), 2.11 (dd,
J = 14.0, 6.0 Hz, 2 H), 2.32 (dd, J = 17.0, 7.0 Hz, 1 H), 3.10-
3.48 (m, 3 H), 3.54 (t, J = 10.0 Hz, 1 H), 3.80 (s, 3 H), 3.82-4.10
(m, 1 H), 4.26-4.42 (m, 3 H), 4.96 (d, J = 12.0 Hz, 2 H), 5.13 (s,
2 H), 5.60 (dd, J = 10.0, 3.0 Hz, 1 H), 5.58-5.71 (m, 1 H), 5.79-
5.87 (m, 1 H), 6.84 (d, J = 8.5 Hz, 2 H), 7.17 (d, J = 8.5 Hz, 2 H),
7.27-7.38 (m, 5 H); 13C NMR (CDCl3, 125 MHz) δ -5.5 (q),
18.2 (s), 23.7 (t), 25.9 (q), 38.7 (t), 52.0 (d), 55.2 (q), 61.2 (t), 63.4
(t), 66.4 (t), 73.0 (t), 73.4 (s), 113.7 (d), 117.8 (t), 122.9 (d), 127.8
(d), 128.4 (d), 129.2 (d), 130.6 (s), 131.7 (d), 133.2 (d), 137.1 (s),
154.5 (s), 159.0 (s); exact mass (electrospray) m/z calcd for
C32H45NNaO5Si (M þ Na) 574.2959, found 574.2958.
(2R,4S)-2-[[[(1,1-Dimethylethyl)dimethylsilyl]oxy]methyl]-3,4-
dihydro-4-hydroxy-1(2H)-pyridinecarboxylic Acid Phenylmethyl
Ester (23f). NaBH4 (12.5 mg, 0.330 mmol) was added to a stirred
and cooled (-40 °C) slurry of 23e (62.3 mg, 0.166 mmol) and
(2S,6R)-6-[[[(1,1-Dimethylethyl)dimethylsilyl]oxy]methyl]-5,6-
dihydro-2-(hydroxy-methyl)-2-(2-propenyl)-1(2H)-pyridinecar-
boxylic Acid Phenylmethyl Ester (41). DDQ (33.8 mg, 0.148
mmol) was added to a stirred and cooled (0 °C) mixture of 40
(40.9 mg, 0.074 mmol), CH2Cl2 (6.5 mL), and water (0.5 mL),
and vigorous stirring was continued for 1.5 h. The mixture was
diluted with saturated aqueous NaHCO3 and Et2O (5 mL). The
aqueous phase was extracted with Et2O (2 ꢀ 5 mL), and the
combined organic extracts were dried (MgSO4) and evaporated.
Flash chromatography of the residue over silica gel (1 ꢀ 10 cm),
using 1:4 EtOAc-hexane, gave 41 (28 mg, 88% yield) as an oil:
CeCl3 7H2O (123 mg, 0.330 mmol) in MeOH (4 mL). Stirring
3
at -40 °C was continued for 1 h, and the mixture was quenched
with acetone (0.1 mL) and saturated aqueous NH4Cl. The cold
bath was removed and replaced by an ice bath, and stirring was
continued for 30 min. The aqueous phase was extracted with
Et2O (3 ꢀ 5 mL), and the combined organic extracts were dried
(MgSO4) and evaporated. Flash chromatography of the residue
over silica gel (1 ꢀ 10 cm), using 3:1 hexane-EtOAc, gave 23f
(50.3 mg, 81%) as an oil: [R]25 þ2.9 (c 1.58, CHCl3); FTIR
D
1
(CHCl3 cast) 3414, 1712, 1655 cm-1; H NMR (CDCl3, 500
[R]25 -7.9 (c 0.18, CHCl3); FTIR (CHCl3 cast) 3448, 1700,
MHz) (signals broad due to the presence of rotamers) δ 0.003 (s,
6 H), 0.87 (s, 9 H), 1.45 (br s, 1 H), 1.55-1.75 (m, 1 H), 2.50 (dd,
J = 12.0, 6.5 Hz, 1 H), 3.45-3.79 (m, 2 H), 4.20-4.40 (m, 1 H),
4.41-4.55 (m, 1 H), 4.80-5.00 (m, 1 H), 5.19 (AB q, J = 12.0,
ΔνAB = 20.5 Hz, 2 H), 6.73-6.95 (m, 1 H), 7.29-7.48 (m, 5 H);
13C NMR (CDCl3, 125 MHz) δ -5.6 (q), 18.2 (s), 25.8 (q), 31.7
(t), 52.7 (d), 61.5 (t), 61.7 (d), 67.8 (t), 109.9 (d), 125.0 (d), 128.2
(d), 128.3 (d), 128.6 (d), 135.9 (s), 153.2 (s); exact mass (electro-
spray) m/z calcd for C20H31NNaO4Si (M þ Na) 400.1915, found
400.1916.
D
1666 cm-1; 1H NMR (CDCl3, 400 MHz) δ 0.03 (s, 6 H), 0.88 (s, 9
H), 1.96-2.10 (m, 2 H), 2.18-2.32 (m, 1 H), 3.10-3.30 (m, 1 H),
3.38-3.50 (m, 1 H), 3.52-3.64 (m, 3 H), 4.20 (d, J = 10.8 Hz, 1
H), 4.56-4.68 (m, 1 H), 4.88-4.99 (m, 2 H), 5.02-5.10 (m, 1 H),
5.24 (d, J = 12.5 Hz, 1 H), 5.50 (dd, J = 10.5, 3.2 Hz, 1 H),
5.54-5.67 (m, 1 H), 5.90 (t, J = 8.0 Hz, 1 H), 7.27-7.40 (m,
5 H); exact mass (electrospray) m/z calcd for C24H37NNaO4Si
(M þ Na) 454.2384, found 454.2387.
(2S,6R)-6-[[[(1,1-Dimethylethyl)dimethylsilyl]oxy]methyl]-2-
formyl-5,6-dihydro-2-(2-propenyl)-1(2H)-pyridinecarboxylic Acid
Phenylmethyl Ester (42). Dess-Martin periodinane (70.8 mg,
0.167 mmol) was added to a stirred and cooled (0 °C) slurry of 41
(24.0 mg, 0.056 mmol) and NaHCO3 (14.1 mg, 0.167 mmol) in
CH2Cl2 (3 mL). Stirring at 0 °C was continued for 1 h, the cold
bath was removed, and stirring was continued for 1 h. Brine
(5 mL) and Et2O (5 mL) were added, and the aqueous phase was
extracted with Et2O (2 ꢀ 5 mL). The combined organic extracts
were dried (MgSO4) and evaporated, and the residue was passed
through a short column of flash chromatography silica gel (2 ꢀ
5 cm), using 1:1 EtOAc-hexane, to afford crude 42, which was
used without characterization.
(2S,6R)-6-[[[(1,1-Dimethylethyl)dimethylsilyl]oxy]methyl]-5,6-
dihydro-2-(2-oxoethyl)-1(2H)-pyridinecarboxylic Acid Phenyl-
methyl Ester (23g). A mixture of 23f (30.2 mg, 0.080 mmol),
Hg(OAc)2 (2.5 mg, 0.008 mmol), and Et3N (1 drop) in n-butyl
vinyl ether (1.5 mL) was heated at 110 °C (oil bath temperature)
in a sealed glass tube (Teflon seal) for 36 h. Volatile material was
evaporated, and flash chromatography of the residue over silica
gel (1 ꢀ 10 cm), using 1:4 EtOAc-hexane, gave 23g (21.3 mg,
66%) as a colorless oil: [R]25 þ70.3 (c 053, CHCl3); FTIR
D
(CHCl3 cast) 1701, 1404 cm-1; 1H NMR (CDCl3, 400 MHz) δ
-0.02 (s, 6 H), 0.85 (s, 9 H), 2.20-2.30 (m, 1 H), 2.35-2.52 (m, 1
H), 2.71 (dd, J = 17.0, 7.0 Hz, 1 H), 2.88-3.02 (m, 1 H), 3.44 (t,
J = 9.0 Hz, 1 H), 3.45-3.64 (m, 1 H), 4.14-4.23 (m, 1 H),
4.52-4.61 (m, 1 H), 5.15 (q, J = 12.5 Hz, 2 H), 5.81-5.94 (m, 2
H), 7.29-7.42 (m, 5 H), 9.74 (br s, 1 H); 13C NMR (CDCl3, 125
MHz) δ -5.5 (q), -5.4 (q), 18.2 (s), 23.9 (t), 25.8 (q), 48.1 (t), 49.4
(d), 52.5 (d), 62.2 (t), 67.3 (t), 124.0 (d), 128.09 (d), 128.14 (d),
128.3 (d), 128.5 (d), 136.3 (s), 156.0 (d), 200.3 (d); exact mass
(electrospray) m/z calcd for C22H33NNaO4Si (M þ Na) 426.2071,
found 426.2074.
(2S,6R)-6-[[[(1,1-Dimethylethyl)dimethylsilyl]oxy]methyl]-5,6-
dihydro-2-[1-hydroxy-2-propenyl]-2-(2-propenyl)-1(2H)-pyridi-
necarboxylic Acid Phenylmethyl Ester (43). Vinylmagnesium
bromide (1.0 M in THF, 75 μL, 0.075 mmol) was added drop-
wise to a stirred and cooled (0 °C) solution of 42 (21.8 mg, 0.048
mmol) in THF (5 mL). Stirring at 0 °C was continued for 1 h,
and then saturated aqueous NH4Cl (5 mL) and Et2O (5 mL)
were added. The aqueous phase was extracted with Et2O (2 ꢀ
5 mL), and the combined organic extracts were dried (MgSO4)
and evaporated. The crude product (43) was used in the next step
without characterization.
(2S,6R)-6-[[[(1,1-Dimethylethyl)dimethylsilyl]oxy]methyl]-5,6-
dihydro-2-[[(4-methoxyphenyl)methoxy]methyl]-2-(2-propenyl)-
1(2H)-pyridinecarboxylic Acid Phenylmethyl Ester (40).(Me3Si)2NLi
(1.0 M in PhMe, 0.30 mL, 0.30 mmol) was added dropwise
to a stirred and cooled (0 °C) solution of Ph3PCH3Br (0.113 g,
0.317 mmol) in THF (4 mL). Stirring at 0 °C was continued for
15 min, and a solution of 20g (60.4 mg, 0.106 mmol) in THF
(1 mL plus 2 mL as a rinse) was added dropwise. Stirring at 0 °C
was continued for 45 min, and then saturated aqueous NH4Cl
(5 mL) and Et2O (5 mL) were added. The aqueous phase was
(5S,7R)-7-[[[(1,1-Dimethylethyl)dimethylsilyl]oxy]methyl]-1-
hydroxy-6-azaspiro-[4,5]deca-2,9-diene Carboxylic Acid Phenyl-
methyl Ester (44). A solution of the above crude 43 and Grubbs
second generation catalyst20 (4.2 mg, 0.005 mmol) in CH2Cl2
(20) [1,3-Bis(2,4,6-trimethylphenyl)-2-imidazolidinylidene]dichloro(phenyl-
methylene)(tricyclohexylphosphine)ruthenium.
5232 J. Org. Chem. Vol. 75, No. 15, 2010

Acharya and Clive
JOCArticle
(4 mL) was refluxed for 8 h, and then evaporated. Flash
chromatography of the residue over silica gel (1 ꢀ 8 cm), using
1:4 EtOAc-hexane, gave 44 (12.0 mg, 51% over three steps) as
an oil which was a single isomer: [R]25D þ22.4 (c 0.36, CHCl3);
FTIR (CHCl3 cast) 3505, 1683, 1665 cm-1; 1H NMR (CDCl3,
400 MHz) δ -0.010 (s, 3 H), -0.002 (s, 3 H), 0.86 (s, 9 H), 2.13
(dt, J = 17.6, 2.5 Hz, 1 H), 2.22-2.42 (m, 2 H), 3.23 (d, J = 8.0
Hz, 1 H), 3.58 (t, J = 10.0 Hz, 1 H), 3.65 (dd, J = 10.0, 4.0 Hz,
1 H), 3.67-3.77 (m, 1 H), 4.10 (d, J = 11.0 Hz, 1 H), 4.50-4.61
(m, 1 H), 5.14 (AB q, J = 12.4, ΔνAB = 26.6 Hz, 2 H), 5.41 (dd,
J = 10.0, 3.0 Hz, 1 H), 5.54-5.62 (m, 1 H), 5.81-5.86 (m, 1 H),
5.92-5.98 (m, 1 H), 7.27-7.40 (m, 5 H); 13C NMR (CDCl3, 125
MHz) δ -5.4 (q), 18.2 (s), 23.6 (t), 25.9 (q), 44.2 (t), 51.7 (d), 62.0
(d), 62.7 (t), 67.3 (t), 84.3 (s), 117.0 (d), 128.0 (d), 128.1 (d), 128.5
(d), 131.6 (s), 132.8 (d), 132.9 (d), 136.4 (d), 157.8 (s); exact
mass (electrospray) m/z calcd for C24H35NNaO4Si (M þ Na)
452.2228, found 452.2230.
Acknowledgment. We thank Dr. D. Liu for exploratory
experiments and NSERC for financial support.
Supporting Information Available: Experimental proce-
dures and copies of NMR spectra for new compounds. This
material is available free of charge via the Internet at http://
pubs.acs.org.
J. Org. Chem. Vol. 75, No. 15, 2010 5233