Influence of the β-alkoxy group on the diastereoselectivity of aldol reactions of tetrahydro-4H-thiopyran-4-one with 4-alkoxytetrahydro-2H-thiopyran-3-carboxaldehydes

Article abstract of DOI:10.1021/jo016291r

The diastereoselectivity of the aldol reaction of tetrahydro-4H-thiopyran-4-one (3) with 1,4-dioxa-8-thiaspiro[4.5]decane-6-carboxaldehyde (9a) under a variety of conditions is examined. Under optimized conditions, three of the four possible diastereomers from this aldol reaction can be obtained selectively (3-16:1). Reactions of 9a with the Li, B, Mg(II), and Ti(IV) enolates of 3 and with the corresponding trimethylsilyl enol ether 4b in the presence of BF₃·OEt₂, SnCl₄, or TiCl₄ as promoters gave the Felkin adducts exclusively (>95%) as mixtures of syn (11a) and anti (12a) diastereomers. Use of the "amine-free" Li enolate of 3 gave 12a with a much higher diastereoselectivity (9:1) and yield (70%) than that obtained using the lithium diisopropylamide-generated Li enolate of 3 (2-3:1; 15-40%). The TiC1₄-promoted reaction of 4b with 9a gave 11a with excellent selectivity (16: 1). In contrast, the MgBr₂·OEt₂-promoted reaction of 4b with 9a gave the anti-Felkin adducts exclusively as a 3:1 mixture of syn (13a)/anti (14a) diastereomers. Similar aldol reactions of 3 with the cis and trans isomers of 4-(methoxy)methoxytetrahydro-2H-thiopyran-3-carboxaldehyde (9b and 9c) were examined to probe the influence of the ketal protecting group in 9a on the observed aldol diastereoselectivity. The results are rationalized by applying Evans' stereochemical model for merged 1,2- and 1,3-asymmetric induction (nonchelation), with the exception of the MgBr₂· OEt₂-promoted reactions of 4b with 9a, 9b, and 9c, which are accommodated by assuming chelation control. Comparison of the reactions of 9a, 9b, and 9c suggests that the ketal group in 9a uniquely allows high levels of either Felkin or anti-Felkin selectivity to be achieved.

Full text of DOI:10.1021/jo016291r

1618
J . Org. Chem. 2002, 67, 1618-1629
In flu en ce of th e â-Alk oxy Gr ou p on th e Dia ster eoselectivity of
Ald ol Rea ction s of Tetr a h yd r o-4H-th iop yr a n -4-on e w ith
4-Alk oxytetr a h yd r o-2H-th iop yr a n -3-ca r boxa ld eh yd es
Dale E. Ward,* Marcelo Sales, Chuk C. Man, J ianheng Shen, Pradip K. Sasmal, and
Cheng Guo
Department of Chemistry, University of Saskatchewan, 110 Science Place,
Saskatoon SK S7N 5C9, Canada
dale.ward@usask.ca
Received November 18, 2001
The diastereoselectivity of the aldol reaction of tetrahydro-4H-thiopyran-4-one (3) with 1,4-dioxa-
8-thiaspiro[4.5]decane-6-carboxaldehyde (9a ) under a variety of conditions is examined. Under
optimized conditions, three of the four possible diastereomers from this aldol reaction can be obtained
selectively (3-16:1). Reactions of 9a with the Li, B, Mg(II), and Ti(IV) enolates of 3 and with the
corresponding trimethylsilyl enol ether 4b in the presence of BF3‚OEt2, SnCl4, or TiCl4 as promoters
gave the Felkin adducts exclusively (>95%) as mixtures of syn (11a ) and anti (12a ) diastereomers.
Use of the “amine-free” Li enolate of 3 gave 12a with a much higher diastereoselectivity (9:1) and
yield (70%) than that obtained using the lithium diisopropylamide-generated Li enolate of 3 (2-
3:1; 15-40%). The TiCl4-promoted reaction of 4b with 9a gave 11a with excellent selectivity (16:
1). In contrast, the MgBr2‚OEt2-promoted reaction of 4b with 9a gave the anti-Felkin adducts
exclusively as a 3:1 mixture of syn (13a )/anti (14a ) diastereomers. Similar aldol reactions of 3 with
the cis and trans isomers of 4-(methoxy)methoxytetrahydro-2H-thiopyran-3-carboxaldehyde (9b
and 9c) were examined to probe the influence of the ketal protecting group in 9a on the observed
aldol diastereoselectivity. The results are rationalized by applying Evans’ stereochemical model
for merged 1,2- and 1,3-asymmetric induction (nonchelation), with the exception of the MgBr2‚
OEt2-promoted reactions of 4b with 9a , 9b, and 9c, which are accommodated by assuming chelation
control. Comparison of the reactions of 9a , 9b, and 9c suggests that the ketal group in 9a uniquely
allows high levels of either Felkin or anti-Felkin selectivity to be achieved.
In tr od u ction
pionate synthesis involves aldol reactions of tetrahydro-
4H-thiopyran-4-one derivatives, followed by desulfuriza-
tion.5,6 We have previously reported on the linear and
two-directional iterative aldol homologations of tetrahy-
dro-4H-thiopyran-4-one (3)with 1,4-dioxa-8-thiaspiro[4.5]-
decane-6-carboxaldehyde (9a ) as a means to prepare
stereochemically complex polypropionate synthons (six
to eight stereogenic centers) in only a few steps.6b The
ability to control the stereoselectivity of the aldol cou-
plings is crucial to the success of this approach. Although
the development of methods for stereoselective aldol
reactions has been intensively investigated for more that
20 years,7 studies on reactions of cyclic ketones or cyclic
aldehydes are rare. In this contribution, we report
optimized conditions that allow three of the four possible
diastereomers from the aldol reaction of 3 with 9a to be
The use of cyclic templates to facilitate and control
various chemical transformations is a well-established
synthetic strategy. Cyclic sulfides have been widely used
for this purpose and offer several advantages, including
ready preparation, versatile chemistry, and relative ease
of removal of the sulfur from the final product.1 In
particular, thiopyran-derived scaffolds have been ex-
ploited in the construction of a variety of synthetic
targets.2,3 For example, Woodward et al. took advantage
of the concave topology of a cis-fused dithiadecalin
template to install 8 of the 10 stereogenic centers present
in the macrocyclic lactone of erythromycin A.4 An alter-
native and potentially more general strategy for polypro-
* Corresponding author.
(1) (a) Belen’kii, L. I. In Chemistry of Organosulfur Compounds;
Belen’kii, L. I., Ed.; Ellis Horwood: Chichester, 1990; p 193. (b) Ingall,
A. H. In Comprehensive Heterocyclic Chemistry; Katritzky, A. R., Rees,
C. W., Eds.; Pergamon: Oxford, 1984; Vol. 3, p 885. (c) Ingall, A. H. In
Comprehensive Heterocyclic Chemistry II; Katritzky, A. R., Rees, C.
W., Scriven, E. F. W., Eds.; Pergamon: Oxford, 1984; Vol. 5, p 501. (d)
Kuthan, J .; Sebek, P.; Bo¨hm, S. In Advances in Hetereocyclic Chemistry;
Katritzky, A. R., Ed.; Academic Press: San Diego, 1994; Vol. 59, p 179.
(e) Thiophene and its Derivatives. In Chemistry of Heterocyclic
Compounds; Gronowitz, S., Ed.; Wiley & Sons: New York, 1991; Vol.
44(4).
(2) Reviews: (a) Vedejs, E.; Krafft, G. A. Tetrahedron 1982, 38,
2857-2881. (b) Vedejs, E. Stud. Nat. Prod. Chem. 1991, 8, 205-218.
(3) (a) Samuel, R.; Nair, S. K.; Asokan, C. V. Synlett 2000, 1804-
1806, and references therein. (b) Ward, D. E.; Gai, Y.; Lai, Y. Synlett
1995, 261-262.
(4) Woodward, R. B.; Logusch, E.; Nambiar, K. P.; Sakan, K.; Ward,
D. E.; Au-Yeung, B.-W.; Balaram, P.; Browne, L. J .; Card, P. J .; Chen,
C. H.; Cheˆnevert, R. B.; Fliri, A.; Frobel, K.; Gais, H.-J .; Garratt, D.
G.; Hayakawa, K.; Heggie, W.; Hesson, D. P.; Hoppe, D.; Hoppe, I.;
Hyatt, J . A.; Ikeda, D.; J acobi, P. A.; Kim, K. S.; Kobuke, Y.; Kojima,
K.; Krowicki, K.; Lee, V. J .; Leutert, T.; Malchenko, S.; Martens, J .;
Matthews, S. R.; Ong, B. S.; Press, J . B.; RajanBabu, T. V.; Rousseau,
G.; Sauter, H. M.; Suzuki, M.; Tatsuta, K.; Tolbert, L. M.; Truesdale,
E. A.; Uchida, I.; Ueda, Y.; Uyehara, T.; Vasella, A. T.; Vladuchick, W.
C.; Wade, P. A.; Williams, R. M.; Wong, H. N.-C. J . Am. Chem. Soc.
1981, 103, 3210-3213.
(5) Hayashi, T. Tetrahedron Lett. 1991, 32, 5369-5372.
(6) (a) Ward, D. E.; Man, C. C.; Guo, C. Tetrahedron Lett. 1997, 38,
2201-2202. (b) Ward, D. E.; Guo, C.; Sasmal, P. K.; Man, C. C.; Sales,
M. Org. Lett. 2000, 2, 1325-1328. (c) Ward, D. E.; Sales, M.; Sasmal,
P. K. Org. Lett. 2001, 3, 3671-3673.
10.1021/jo016291r CCC: $22.00 © 2002 American Chemical Society
Published on Web 02/01/2002

Diastereoselectivity of Aldol Reactions
J . Org. Chem., Vol. 67, No. 5, 2002 1619
Sch em e 1. P r ep a r a tion of Sta r tin g Ma ter ia lsa
Sch em e 2. Ald ol Rea ction s of 4 w ith 9
examine aldol reactions of 4 with 9a . 9a was readily
prepared from 2, as outlined in Scheme 1. Despite
considerable experimentation,10 we were unable to re-
produce the reported5 aldol reaction of 4a with 9a ; in our
hands, aldol products were obtained as separable 2-3:1
mixtures of 12a /11a in 15-40% combined yield. Initially,
we suspected that the “profound tendency toward ret-
roaldolization” reported5 for aldol adducts of 3 was
responsible for our difficulties; however, both 11a and
12a were recovered unchanged (>85% yield) after treat-
ment with LDA under the above conditions.11
a
Reagents: (a) NaOMe (90%). (b) 10% H2SO4 (80%). (c) LDA,
THF, -78 °C (4a ; >95% assumed). (d) TMSCl, Et3N, rt (4b; >95%).
(e) R2B-X, i-Pr2EtN, -78 °C, 3 h (>95% assumed). (f) (i) TiCl4;
(ii) i-Pr2EtN, -78 °C, 2 h (4f). (g) (HOCH2)2, p-TsOH, C6H6, reflux
(90%). (h) NaBH3CN, citric acid, EtOH (73%). (i) LiAlH4 (86-94%).
(j) MOM-Cl, i-PrEt2N (>95%). (k) From 8: (i) DMSO, (COCl)2; (ii)
i-Pr2EtN (55-89%). (m) From 5c: DIBAl-H (69%). (n) MeLi, THF/
ether, rt (>95%).
We decided to examine the reaction of 4c with 9a on
the basis of a note12 in Woodward’s erythromycin A
synthesis indicating that a key aldol-coupling of a dithia-
decalone proceeded in substantially higher yield when
using the “amine-free” Li enolate than when using the
LDA-generated Li enolate. It is well established that the
i-Pr2NH product formed from the reaction of LDA with
ketones is associated with the enolate (e.g., 4a ) and that
the presence of an amine can influence the ensuing
enolate reactions;13 however, any such effects on aldol
diastereoselectivity have not been well-documented. We
were delighted to find that the reaction of 4c (prepared
from 4b and MeLi)14 with 9a produced a separable 9:1
mixture of 12a /11a in 70% isolated yield (Table 1, entry
2). This increase in aldol diastereoselectivity using an
obtained selectively. We also report on the unique influ-
ence of the ketal protecting group in 9a on the stereose-
lectivity of the aldol reactions.
Resu lts
Our ongoing interest in desymmetrization8,9 and the
ready availability of 2 and 3 (Scheme 1) prompted
consideration of various coupling reactions as a means
to prepare symmetric bifunctional substrates that could
be used as tri- or tetrapropionate synthons. Initially, we
attempted aldol reactions of the lithium diisopropyla-
mide-generated (lithium diisopropylamide ) LDA) eno-
late 4a with 10 but were unable to obtain adducts.
Subsequently, Hayashi reported that aldol reactions of
4a with a variety of aldehydes gave adducts with high
anti diastereoselectivity.5 Of particular interest to us, the
reaction of 4a with aldehyde 9a was reported to give 12a
(Scheme 2) in 85% yield with excellent diastereoselec-
tivity (50:1).5 In light of this information, we decided to
(10) The reported5 procedure is typical: addition of 3 to LDA (1
equiv) in THF at -78 °C; after 30 min, addition of 9a (1.2 equiv); after
0.5 min, quenching by addition of NH4Cl. We calibrated these condi-
tions using cyclohexanone and benzaldehyde and obtained the aldol
product (75% yield; 4:1 anti/syn) as reported: Hirama, M.; Noda, T.;
Takeshi, S.; Ito, S. Bull. Chem. Soc. J pn. 1988, 61, 2645-2646;
Majewski, M.; Gleave, D. M. Tetrahedron Lett. 1989, 30, 5681-5684.
Addition of TMSCl instead of 9a gave 4b (88% yield). Addition of
benzaldehyde instead of 9a gave a mixture of aldol diastereomers (30-
50%; 4-6:1 anti/syn), a result inferior to that reported5 (98%; 9:1 anti/
syn). We attempted numerous modifications of the procedure (i.e.,
reaction time with 9a (0.5-60 min), concentration (0.05-0.5 M),
stoichiometry (9a /3 ) 0.5-3:1), quench (HOAc, TESOTf, DIBAl-H),
and work-up procedure).
(7) (a) Evans, D. A.; Nelson, J . V.; Taber, T. R. Top. Stereochem.
1982, 13, 1-115. (b) Heathcock, C. H. In Asymmetric Synthesis;
Morrison, J . D., Ed.; Academic Press: New York, 1984; Vol. 3, pp 111-
212. (c) Heathcock, C. H. Aldrichimica Acta 1990, 23, 99-111. (d)
Nelson, S. G. Tetrahedron: Asymmetry 1998, 357-389. (e) Arya, P.;
Qin, H. Tetrahedron 2000, 56, 917-947. (f) Machajewski, T. D.; Wong,
C.-H. Angew. Chem., Int. Ed. 2000, 39, 1352-1374.
(11) We have found various aldol adducts of 3 to be stable to
retroaldol reaction; see reference 6c.
(12) See footnote 16 in reference 4.
(8) Review: Willis, M. C. J . Chem. Soc., Perkin Trans 1 1999, 1765-
1784.
(9) (a) Ward, D. E.; Liu, Y.; Rhee, C. K. Synlett 1993, 561-563. (b)
Ward, D. E.; How, D. B.; Liu, Y. J . Am. Chem. Soc. 1997, 119, 1884-
1894. (c) Ward, D. E., Sales, M.; Hrapchak, M. J . Can J . Chem. 2001,
79, 1715-1785.
(13) Review: Seebach, D. Angew. Chem., Int. Ed. 1988, 27, 1624-
1654.
(14) Stork, G.; Hudrlik, P. F. J . Am. Chem. Soc. 1968, 90, 4462-
4464; J . Am. Chem. Soc. 1968, 90, 4464-4465.

1620 J . Org. Chem., Vol. 67, No. 5, 2002
Ta ble 1. Dia ster eoselectivity of Ald ol Rea ction s of 4 w ith 9
Ward et al.
enolate
derivative
(equiv)
aldol
c
ratio of aldols
aldehyde
(equiv)
promoter
(equiv)
adducts
yield
(%) b
entry
(series)a
11
12
13
14
d
d
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
4a (1)
4c (1.5)
4c (2)
4d (1)
4e (1)
4f (1)
4b (2)
4b (2)
4b (1.5)
4b (2)
4b (2)
4c (2)
4c (2)
4b (1.5)
4b (2)
4b (2)
4b (2)
9a (1.2)
9a (1)
9a (1)
9a (4)
9a (4)
9a (1.2)
9a (1)
9a (1)
9a (1)
9a (1)
9a (1)
9b (1)
9c (1)
9b (1)
9c (1)
9b (1)
9c (1)
a
a
a
a
a
a
a
a
a
a
a
b
c
b
c
b
c
15-40
70
75
1
1
1
1
2.5
9
2.5
10
15
1
1
3
1
MgBr2‚OEt2 (2)e
f
82
f
84
1
g
46h
70
3.6h
2
BF3‚OEt2 (1)i
SnCl4 (2)j
72
87
23
89
61
64
27
81
1
16
TiCl4 (1)i
MgBr2‚OEt2 (1)j
1.7
3
1
1
MgBr2‚OEt2 (3)k
d
1
4
1
1
1.5
d
TiCl4 (1)i
1.3
2
TiCl4 (1)i
20
4
1
MgBr2‚OEt2 (3)k
MgBr2‚OEt2 (3)k
84
74
3.5
7
1
1
a
b
The series of adducts 11-14 as shown in Scheme 2. The total isolated yield of all aldol adducts. c Determined by isolation or by 1H
NMR of purified products (relative error estimated at (10%); see Experimental Section. These ratios were consistent with those suggested
from 1H NMR of the crude reaction mixtures. The detection limit for minor diastereomers is estimated to be 3%. Reaction at -78 °C for
d
1-5 min. e Reaction of 4c with MgBr2 for 5 min at -78 °C and then with 9a for 5 min at -78 °C. f Reaction at -78 °C for 3 h. Reaction
g
h
i
j
k
at -78 °C for 5 h. See text. Reaction at -78 °C for 1 h. Reaction at 0 °C for 4 h. Reaction at 0 °C for 1 h.
amine-free Li enolate may have some generality and
should be studied further.15
lectivity (Table 1, entry 9). All of the above aldol reactions
of 9a with 4a , 4b, or 4c gave only the 1′,6′′-syn diaster-
eomers 11a and 12a . By contrast, reaction of 9a with 4b
in the presence of MgBr2‚OEt2 exclusively gave the 1′,6′′-
anti diastereomers 13a and 14a . The diastereoselectivity
in favor of 13a was improved with excess MgBr2‚OEt2
(cf. entries 10 and 11); this effect has been noted in other
processes.18
Reactions of Ti, Mg, and B enolates of 3 with 9a were
also investigated (Table 1). Compared to the Li enolate
4c, inferior diastereoselectivity in favor of 12a was
obtained with the putative Mg(II) enolate (Table 1, entry
3). Although the boron enolates 4d and 4e performed
slightly better than did 4c in terms of yield and stereo-
selectivity of the aldol reaction (compare entry 2 with
entries 4 and 5), the simplicity of the experimental
procedure using 4c was an advantage in reactions carried
out on a preparative scale. The reaction of 9a with the
Ti(IV) enolate 4f (prepared by addition of i-Pr2EtN to a
mixture of 3 and TiCl4) gave 11a (26%) in addition to a
1:1 mixture of two bisaldol adducts (20%).6b On the basis
of the structures of the bisaldols adducts, we estimate
that the initial aldol reaction between 4f and 9a produced
a 3.6:1 mixture of 11a /12a .16
To the best of our knowledge, stereoselective aldol
reactions of ketals of â-ketoaldehydes have not been
previously reported. Because of the unusual and useful
diastereoselectivity observed in the reactions of 9a with
4, we sought to ascertain the role of the ketal group in
this selectivity. Toward this end, we prepared the cis and
trans (methoxy)methyl (MOM) ether-protected â-hy-
droxyaldehydes 9b and 9c (Scheme 1) and examined
their aldol reactions with 4b and 4c under the conditions
that gave the most diastereoselective reactions with 9a
(Table 1, entries 12-17). Interestingly, 9a and 9c had
very similar stereoselectivities in their reactions both
with 4c (i.e., high selectivity for the 1′,3-anti-1′,3′′(6′′)-
syn products 12a and 12c) and with 4b in the presence
of TiCl4 (i.e., high selectivity for the 1′,3-syn-1′,3′′(6′′)-
syn products 11a and 11c) but very dissimilar stereose-
lectivities in their reactions with 4b in the presence of
MgBr2‚OEt2. By contrast, 9a and 9b had very similar
stereoselectivities in their reactions with 4b in the
presence of MgBr2‚OEt2 (i.e., high selectivity for 1′,3′′-
(6′′)-anti products 13a /14a and 13b/14b) but very dis-
similar selectivities in the other reactions.19 In no case
was a 1′,3-anti-1′,3′′(6′′)-anti aldol (i.e., 14) the major
product.
Several Lewis acid-mediated reactions of 4b with 9a
were examined (i.e., Mukaiyama reactions).17 Whereas
the reactions mediated by BF3‚OEt2 and SnCl4 gave
mixtures of 11a and 12a with modest stereoselectivity
(Table 1, entries 7 and 8), the TiCl4-promoted reaction
gave 11a in good yield and with excellent diastereose-
(15) Improved diastereoselectivity was noted in other reactions.
Reaction of the “amine-free” Li enolate of cyclohexanone (prepared from
the trimethylsilyl enol ether and MeLi as described for 4c) with
benzaldehyde under the same conditions (-78 °C, 5 min) gave a 12:1
mixture of anti and syn aldol adducts, respectively (75%), and a similar
reaction of 4c with benzaldehyde gave a >10:1 mixture of anti and
syn aldol adducts, respectively (60%). See reference 10 for the results
of the same reactions using LDA-generated enolates.
(16) We have previously established (ref 6b) that the same bisaldol
adducts are produced from reactions of 9a with the Ti(IV) enolates of
11a and 12a . One of the bisaldols was identical to the major product
from 11a , and the other was the major product from 12a . Thus, it is
plausible that the 1:1 mixture of bisaldols (20%) comes from an equal
mixture of 11a and 12a and, consequently, an estimated 3.6:1 ratio of
11a to 12a in the initial aldol reaction. Presumably, the bisaldols result
from incomplete enolate formation from 3, thereby leaving excess base
to generate the Ti(IV) enolate from the initially formed Ti(IV) aldolates
corresponding to 11a and 12a .
Det er m in a t ion of St er eoch em ica l Con figu r a -
tion s. In previous work, Hayashi5 proposed the 1′,3-anti-
(18) Ward, D. E.; Hrapchak, M. J .; Sales, M. Org. Lett. 2000, 2, 57-
60.
(17) (a) Mukaiyama, T.; Banno, K.; Narasaka, K. J . Am. Chem. Soc.
1974, 96, 7503-7509. (b) Genari, C. In Comprehensive Organic
Synthesis; Trost, B. M., Ed.; Pergamon: Oxford, 1993; Vol. 2, Chapter
2.4, p 629.
(19) Aldehyde 9b was not stable to TiCl4 (mainly elimination); the
low yields of 14b and 11b obtained from the TiCl4-promoted reaction
of 4b with 9b prevent us from drawing firm conclusions about the
diastereoselectivity in this reaction.

Diastereoselectivity of Aldol Reactions
J . Org. Chem., Vol. 67, No. 5, 2002 1621
Sch em e 3. Deter m in a tion of Ald ol Ster eoch em istr ya
a
Reagents: (a) DIBAL-H. (b) NaBH(OAc)3. (c) NaBH4. (d) NaBH3CN. (e) Im2CO. (f) H+. (g) NaOH, MeOH. (h) MOM-Cl, i-PrEt2N.
3
1′,6′′-syn relative configurations20 (i.e., 12a ) for the aldol
was tenuous because the apparent J HC-1′/HC-3 coupling
constants for the major (12a ) and minor (11a ) isomers
were 6.5 and 8.5 Hz, respectively, and the validity of
those apparent J values was questionable because of
possible second-order effects due to the similar chemical
shifts for HC-3 and H2C-2.21 The assumption of Cram
diastereoface selectivity also seemed uncertain in light
of numerous examples of “anti-Cram” diastereoselectivity
in aldol reactions of aldehydes with â-alkoxy substitu-
ents.22 Thus, to unambiguously establish the relative
stereochemical configurations for 11a and 12a , we pre-
pared a series of chemical derivatives, as summarized
in Scheme 3.6a A similar approach allowed determination
of the relative stereochemical configurations for all of the
aldol adducts 11-14.
adduct obtained from the reaction of 4a with 9a on the
basis of a “large” (the value was not reported) 3J HC-1′/HC-3
coupling constant (indicating a 1′,3-anti relative config-
uration)21 and assuming Cram diastereoface selectivity
for the addition to the aldehyde group in 9a (leading to
a 1′,6′′-syn relative configuration). In our hands (vide
supra), the same reaction gave a 2-3:1 mixture of two
aldol adducts (i.e., 12a /11a ); assignment of the C-1′/C-3
relative stereochemical configurations in our products
(20) For simplicity and to allow easy comparison to similar aldol
reactions, the syn (same side) and anti (opposite sides) stereochemical
descriptors for aldols 11-14 refer to the relative disposition of the
designated pairs of non-hydrogen substituents with respect to the plane
defined by the carbon chain C-4-C-3-C-1′-C-3′′(or C-6′′)-C-4′′(or
C-5′′) in an extended (zigzag) conformation.
The C-1′/C-3 relative configuration for each aldol
adduct was determined by preparing a cyclic carbonate
derivative (23-28). Reduction of the aldols 11-14 to give
(21) According to the Stiles-House rule for intramolecularly H-
3
bonded aldols, the J H-C(OH)-C(CO)-H coupling constant for the threo
(anti) isomer (6-9 Hz) is larger than that for the erythro (syn) isomer
(2-4 Hz): (a) Stiles, M.; Winkler, R. R.; Chang, Y.-L.; Traynor, L. J .
Am. Chem. Soc. 1964, 86, 3337-3342. (b) House, H. O.; Crumrine, D.
S.; Teranishi, A. Y.; Olmstead, H. D. J . Am. Chem. Soc. 1973, 95, 3310-
3324.
(22) Reviews: (a) Mahrwald, R. Chem. Rev. 1999, 99, 1095-1120.
(b) Mengel, A.; Reiser, O. Chem. Rev. 1999, 99, 1191-1223.

1622 J . Org. Chem., Vol. 67, No. 5, 2002
Ta ble 2. Der iva tives Used to Esta blish th e Rela tive Ster eoch em ica l Con figu r a tion s of 11-14
Ward et al.
relative
configurationsa
(C-1′/C-3,
HC-4/HC-4a
relative
orientation
carbonate
derivative
3J HC-4/HC-4a
triol
derivative
symmetry
c
entry
aldol
C-1′/C-3′′(6′′))
(Hz)b
(no. of δC’s)d
1
2
3
4
5
6
7
8
9
11a
12a
13a
14a
11b
12b
13b
14b
11c
12c
13c
14c
syn, syn
anti, syn
syn, anti
anti, anti
syn, syn
anti, syn
syn, anti
anti, anti
syn, syn
anti, syn
syn, anti
anti, anti
23a
25a
27a
28a
24b
25b
27b
28b
24c
25c
27c
28c
2
10
4.5
10
4.5
10
cis
trans
cis
trans
cis
trans
cis
trans
cis
trans
cis
16e
18f
19e
22f
29b
17e
19e
21e
18f
30c
20f
22f
symmetric (6)
unsymmetric (11)
unsymmetric (11)
symmetric (6)
symmetric (8)
unsymmetric (11)
unsymmetric (11)
symmetric (6)
3.5
8.5
7e
symmetric (6)
10
11
12
11e
6e
unsymmetric (17)
unsymmetric (11)
unsymmetric (11)
10e
trans
a
b
See ref 20. The coupling constant was determined assuming first-order behavior. c With respect to the carbonate derivative; see ref
25. Symmetry of the triol derivative; see ref 26. e The spin systems for the two thiopyran fragments in the c series of aldols (11-14) are
very similar; the assignment was confirmed by NOE (correlation between HC-4′ and the MOM-CH2).
d
1,3-diols 15-22 was accomplished with a variety of
reagents: DIBAL-H (1,3-syn selective),23 NaBH(OAc)3
(1,3-anti selective),24 NaBH4, and NaBH3CN (unselec-
tive). The C-3/C-4 relative configuration (i.e., cis or trans)
for each of the products (15-22) was readily determined
trans hydroxycarbonates 25f and 28f gave the unsym-
metric triol 18f (identical to 20f) and the symmetric triol
22f, respectively, thereby fully defining the stereochem-
ical configurations for 12a and 14a . The stereostructures
of aldols 11a and 13a were established using the same
procedure but starting from the derived cis carbonates
23a and 26a ; after hydrolysis of the ketals and reduction
of the resulting ketones, base hydrolysis of the corre-
sponding cis hydroxycarbonates 23e and 26e gave triols
15e (symmetric) and 19e (unsymmetric), respectively.
For aldols 11b-14b derived from 9b, the corresponding
cis diols 17b, 19b, and 21b were used to prepare triols
17e (identical to 19e; unsymmetric) and 21e (symmetric)
by hydrolyses of the MOM ethers; the bis-MOM ether
derivative 29b was prepared from cis diol 15b by reaction
with MOM-Cl. For the aldols 11c-14c derived from 9c,
the corresponding trans diols 16c, 20c, and 22c were
used to prepare triols 16f (symmetric), 20f (identical to
18f; unsymmetric), and 22f (symmetric), respectively; the
same unsymmetrical tris-MOM derivative 30c was ob-
tained from the trans diols 18c and 20c. Table 2
summarizes the carbonate and triol derivatives used to
assign the relative stereochemical configurations in aldols
11a -14a , 11b-14b, and 11c-14c.
3
by analysis of the J H-H coupling constants observed for
HC-4, assuming a chair conformation for the 3-substi-
tuted tetrahydro-2H-thiopyran-4-ol fragment. For each
aldol (except 11a ), a trans alcohol derivative (i.e., 16, 18,
20, or 22) was obtained selectively or was isolated from
a mixture of cis and trans products. Reactions of the 11
trans alcohols 16b,c, 18a -c, 20a -c, and 22a -c with
1,1′-carbonyldiimidazole (Im2CO) gave the corresponding
trans fused cyclic carbonates 24b,c, 25a -c, 27a -c, and
28a -c, respectively. The relative orientations of HC-4/
HC-4a (i.e., cis or trans) in these carbonates were
3
established from their respective J HC-4/HC-4a coupling
constants,25 thereby establishing the C-1′/C-3 relative
configurations in the precursor aldols (Table 2). For 11a ,
the carbonate 23a (prepared from the cis alcohol 15a )
3
had a J HC-4/HC4a coupling constant of 2 Hz, indicating a
cis relationship between HC-4 and HC-4a (Table 2, entry
1); this assignment was confirmed by the observation of
a positive NOE for HC-4 upon irradiation of HC-8a.
The relative configurations at C-3/C-6′′ in aldols 11a -
14a and at the equivalent C-3/C-3′′ positions in aldols
11b-14b and 11c-14c (i.e., syn or anti)20 were estab-
lished from the symmetry of triol derivatives prepared
with identical relative configurations at C-3/C-4 and at
C-3′′/C-4′′ (i.e., both cis or both trans).26 Thus, for aldols
12a and 14a , acid hydrolysis of the ketals in the derived
trans carbonates (25a and 28a ) gave the corresponding
ketones 25d and 28d , respectively.27 Nonselective reduc-
tions of 25d and 28d (NaBH4) gave cis/trans mixtures of
alcohols 25e,f and 28e,f, respectively.28 Hydrolyses of the
3
Table 3 shows the apparent J HC-1′/HC-3 and δH HC-1′
3
values for aldols 11-14. Clearly, the use of the J HC-1′/HC-3
coupling constant to assign simple aldol diastereoselec-
tivity (i.e., syn or anti for C-1′/C-3)21 is completely
unreliable for these systems.29 Half of the 1′,3-anti aldols
(entries 4, 8, and 14) have “small” coupling constants (e5
Hz), and two of the syn aldols (entries 1 and 9) have
“large” coupling constants (>8 Hz). Comparing pairs of
syn/anti diastereomers (i.e., entries 1/2, 3/4, 5/6, 7/8, 9/10,
3
and 11/12), four of the six have similar J HC-1′/HC-3
coupling constants (difference is e1 Hz). Only two
examples (entries 1/2 and 11/12) have as much as a 2
Hz difference between their J HC-1′/HC-3 coupling con-
stants, and the syn isomer has the greater J value in
one of these (cf. entries 1 and 2). On the other hand, the
(23) Kiyooka, S.; Kuroda, H.; Shimasaki, Y. Tetrahedron Lett. 1986,
27, 3009-3012
3
(24) Evans, D. A.; Chapman, K. T.; Carreira, E. M. J . Am. Chem.
Soc. 1988, 110, 3560-3578.
(25) The trans ring fusion in carbonates 24, 25, 27, and 28 ensures
that HC-4a has an axial orientation; thus, for HC-4/HC-4a, trans H’s
will have an axial-axial disposition (J ) 8.5-11 Hz) and cis H’s will
have an axial-equatorial disposition (J ) 3.5-7 Hz).
(26) Aldols with a syn relative configuration at C-3/C-6′′ or C-3/C-
3′′ give symmetrical (Cs) triol derivatives. Aldols with an anti relative
configuration at C-3/C-6′′ or C-3/C-3′′ give unsymmetrical (pseudo-C2-
symmetric) triol derivatives.
(28) The C-3′′/C-4′′ relative configurations (i.e., cis or trans) of the
3
product alcohols were readily determined by analysis of the J H-H
coupling constants observed for HC-4′′, assuming a chair conformation
for the 3-substituted tetrahydro-2H-thiopyran-4-ol fragment.
(29) For a detailed discussion of the pitfalls of using 1H NMR to
assign stereochemical configurations of aldols of cyclic ketones, see:
Kitamura, M.; Nakano, K.; Okada, M.; Noyori, R. J . Am. Chem. Soc.
2001, 123, 8939-8950.
(27) Ketal hydrolyses were shown to occur without epimerization
by isolating products without deuterium incorporation from reactions
using DClO4/D2O.

Diastereoselectivity of Aldol Reactions
J . Org. Chem., Vol. 67, No. 5, 2002 1623
3
Ta ble 3. Ap p a r en t J HC-1′/HC-3 a n d δH HC-1′ for 11-14
C-1′/C-3
relative
3J HC-1′/HC-3
entry
aldol
configurationa
δ
H HC-1′
(Hz)b
1
2
3
4
5
6
7
8
9
11a
12a
13a
14a
11b
12b
13b
14b
11c
12c
13c
14c
syn
anti
syn
anti
syn
anti
syn
anti
syn
anti
syn
4.79
4.50
4.57
4.24
4.23
3.96
4.20
3.77
4.84
4.47
4.52
4.07
8.5
6.5
3.5
3.5
5.5
6
3.5
3.5
8.5
9.5
3
10
11
12
anti
5
a
b
See reference 20. The coupling constant was determined
assuming first-order behavior.
δH HC-1′ values for the syn aldols are consistently greater
(i.e., at lower field) than the values for the corresponding
anti aldols and within each series; the order (low field to
high field) of the δH HC-1′ values is 1′,3-syn-1′,3′′(or 6′′)-
syn > 1′,3-syn-1′,3′′(or 6′′)-anti > 1′,3-anti-1′,3′′(or 6′′)-
syn > 1′,3-anti-1′,3′′(or 6′′)-anti.20
Discu ssion
The directed aldol reaction of 3 (via 4) with 9 can
produce four diastereoisomeric products, 11-14. The
control elements influencing the stereochemical outcome
of this reaction involve the diastereoface selectivity for
addition to the aldehyde carbonyl group in 9 (i.e., to give
1′,3′′(6′′)-syn or 1′,3′′(6′′)-anti products) and the simple
diastereoselectivity (i.e., relative topicity) for the aldol
reaction (i.e., to give 1′,3-syn or 1′,3-anti products). The
stereoselectivity of aldol reactions of enolate derivatives
from cyclic ketones is often low and unpredictable.30
Nonetheless, in the only previously described examples
of aldol reactions of 3,5 a high propensity for anti relative
topicity (10-99:1) was reported for reactions of the
corresponding Li,10 B, Sn(IV), and Ti(IV) enolates with
simple aldehydes. Similarly, both the B31 and Li (espe-
cially amine-free)10,15 enolates of cyclohexanone react with
aldehydes to give the anti adducts with high selectivity.
On the other hand, Mukaiyama-type reactions17 of cy-
clohexanone enol trimethylsilyl ether with aldehydes
using a variety of Lewis acid promoters typically give
adducts with modest stereoselectivity (syn/anti from 3:1
to 1:3).32 These results can be rationalized by the prefer-
ence for “closed” chairlike Zimmerman-Traxler transi-
tion states33 in aldol reactions of Li and B enolates,
whereby the minimization of nonbonded interactions
favors the formation of anti aldol adducts from the
necessarily (E)-enolates derived from cyclic ketones.7a-c,34
By contrast, Mukaiyama aldol reactions are thought to
F igu r e 1. Proposed nonchelation transition-state models for
reactions of 4 with (a) 9a , (b) 9b, and (c) 9c.
proceed via open transition states having either anti-
periplanar or synclinal geometries, and both the direction
and degree of simple diastereoselectivity are dependent
on a number of factors.17,22a,35
Models to rationalize and predict the diastereoface
selectivity in additions of enolates to chiral aldehydes
have been discussed extensively in the literature.22 The
Felkin-Ahn model36 is widely used to account for the
effect of a stereogenic center R to the aldehyde carbonyl
(1,2 induction), and the influence of a heteroatom sub-
stituent â to the carbonyl group (1,3 induction) has been
reconciled using chelation37,38 or nonchelation39-41 models.
The stereochemical model recently proposed by Evans40
merges the effects of 1,2 and 1,3 asymmetric induction
and thus is well suited for application to aldehydes such
as 9. Figure 1 shows the proposed nonchelation transi-
tion-state (TS) models for aldol reactions of 4 with 9,
according to the Evans model.40 Thus, 9c is analogous to
an acyclic “anti” R-methyl-â-alkoxyaldehyde (cf. 32),
(30) Baldwin, S. W.; Chen, P.; Nikolic, N.; Weinseimer, D. C. Org.
Lett. 2000, 2, 1193-1196 and references therein.
(35) Denmark, S. E.; Lee, W. J . Org. Chem. 1994, 59, 707-709.
(36) (a) Cherest, M.; Felkin, H.; Prudent, N. Tetrahedron Lett. 1968,
2199-2204. (b) Ahn, N. T.; Eisenstein, O. Nouv. J . Chim. 1977, 1, 61-
70. (c) Ahn, N. T. Top. Curr. Chem. 1980, 88, 145-162.
(37) Reetz, M. T. Acc. Chem. Res. 1993, 26, 462-468.
(38) Evans, D. A.; Allison, B. D.; Yang, M. G. Tetrahedron Lett. 1999,
40, 4457-4460.
(39) (a) Leitereg, T. J .; Cram, D. J . J . Am. Chem. Soc. 1968, 90,
4011-4018; J . Am. Chem. Soc. 1968, 90, 4019-4026. (b) Reetz, M. T.;
Kesseler, K.; J ung, A. Tetrahedron Lett. 1984, 25, 729-732.
(40) Evans, D. A.; Dart, M. J .; Duffy, J . L.; Yang, M. G. J . Am. Chem.
Soc. 1996, 118, 4322-4343.
(41) Bonini, C.; Esposito, V.; D’Auria, M.; Righi, G. Tetrahedron
1997, 53, 13419-13426.
(31) (a) Evans, D. A.; Vogel, E.; Nelson, J . V. J . Am. Chem. Soc.
1979, 101, 6120-6123. (b) Evans, D. A.; Nelson, J . V.; Vogel, E.; Taber,
T. R. J . Am. Chem. Soc. 1981, 103, 3099-3111.
(32) (a) Nakamura, H.; Matsuhashi, H.; Arata, K. Synlett 2000, 668-
670. (b) Raju, S. V. N.; Ponrathnam, S.; Rajan, C. R.; Srinivasan, K.
V. Synlett 1996, 239-240 and references therein.
(33) (a) Zimmerman, H. E.; Traxler, M. D. J . Am. Chem. Soc. 1957,
79, 1920-1923. (b) Denmark, S. E.; Henke, B. R. J . Am. Chem. Soc.
1991, 113, 2177-2194.
(34) Braun, M. In Stereoselective Synthesis: Methods in Organic
Chemistry, ed. E21; Helmchen, G., Hoffman, R., Multzer, J ., Schrau-
mann, E., Eds.; Thieme: Stuttgart, 1996; Vol. 3, pp 1603-1612.

1624 J . Org. Chem., Vol. 67, No. 5, 2002
Ward et al.
where the R and â substituents reinforce addition to the
same aldehyde diastereoface (Figure 1, TS D) and high
1′,3′′-syn selectivity (i.e., Felkin-type) is predicted and
observed (Table 1, entries 13 and 15). Reaction of 9c with
the Li enolate 4c via the expected closed TS D2 gives the
1′,3-anti aldol 12c, whereas the TiCl4-promoted reaction
of 9c with 4b via the open antiperiplanar TS D1 gives
the 1′,3-syn aldol 11c. In contrast, 9b is analogous to an
acyclic “syn” R-methyl-â-alkoxyaldehyde (cf. 31), where
the substituents direct addition to opposite faces of the
aldehyde (Figure 1, Felkin TS B has an unfavorable
dipole interaction and anti-Felkin TS C presents an
unfavorable steric interaction with the nucleophile),
thereby offering a rationale for the low 1′,3′′-syn/anti
selectivity observed (Table 1, entries 12 and 14). For
aldehyde 9a , TS A predicts 1′,3′′-syn selectivity (i.e.,
Felkin-type), as is observed (Table 1, entries 1-9).
Although TS A has an unfavorable interaction between
the carbonyl and cis â-oxygen of the ketal group, all other
possible conformations with staggered relationships be-
tween the forming bond and the R-substituents have
additional unfavorable steric or dipole interactions.42
Reactions of 9a with the Li (4c), B (4d and 4e), and Mg
enolates via the expected closed TS A2 would preferen-
tially give the 1′,3-anti aldol 12c, as observed (Table 1,
entries 1-5). The reactions of 4b with 9a promoted by
BF3‚OEt2, SnCl4, or TiCl4 give only the Felkin adducts
11a and 12a (Table 1, entries 7-9), but the wide
variation in the observed 11a /12a ratio suggests that
reactions can occur via both antiperiplanar (i.e., via A1
to give 11a ) and synclinal TSs (i.e., via A2 with M ) TMS
and without coordination to give 12a ). Perhaps the
relative propensity for formation of 12a in these reactions
is dependent upon the nature of the interaction between
the coordinated Lewis acid and the TMSO group in the
synclinal TS. In any event, the use of TiCl4 gave 11a with
high selectivity (Table 1, entry 9).
F igu r e 2. Proposed chelated transition-state models for
reactions of 4b with (a) 9a , (b) 9b, and (c) 9c.
established by NMR;45 however, the use of these Lewis
acids does not necessarily result in a reaction via a
chelated intermediate.38 For example, chelate organiza-
tion is unlikely in both the TiCl4- and SnCl4-mediated
reactions of 4b with 9a because the exclusive Felkin
diastereoselectivity in these cases is also obtained in the
presence of BF3‚OEt2, where chelation is prevented (Table
1, entries 7-9). On the other hand, the dramatic reversal
of the diastereoface selectivity for addition of 4b to 9 in
the presence of MgBr2‚OEt2 strongly supports the hy-
pothesis of a chelated intermediate in these reactions.46
Figure 2 shows proposed chelated transition-state (TS)
models that can account for these results. The exclusive
anti-Felkin selectivity observed in the reactions of 4b
with 9a and 9b (Table 1, entries 11 and 16) can be
rationalized by examination of the chelated models E47
and F , which suggest high selectivity because the least-
hindered approach and the “axial” face of the carbonyl
carbon are coincident. Similarly, the modest selectivity
observed in the reaction of 9c (Table 1, entry 17) is
accounted for in the chelated model G, which indicates
that the faces of the carbonyl are not differentiated
decisively.
In contrast to the above reactions that give the Felkin
aldol adducts 11 and 12 with high selectvity,19 the MgBr2‚
OEt2-promoted reactions of 4b with 9a , 9b, and 9c
predominantly give the anti-Felkin adducts 13 and 14
(Table 1, entries 10, 11, 16, and 17). These results can
be rationalized by assuming that the Mg(II)-mediated
reactions proceed through a chelated intermediate (i.e.,
chelation-controlled selectivity).38 Chelated intermediates
have often been invoked to explain the formation of anti-
Felkin products in aldol reactions of â-alkoxy alde-
hydes;22,43 however, Evans et al. have established that
the same products can also result from nonchelated
intermediates.40,44 The ability of MgBr2, TiCl4, and SnCl4
to form chelates with â-alkoxyaldehydes has been firmly
It is instructive to compare the diastereoselectivities
observed in aldol reactions of 4 with 9b and 9c to those
reported48 for similar aldol reactions of the structurally
related acyclic analogues 31-33 (Scheme 4). The trans
(45) (a) Keck, G. E.; Castellino, S. E. J . Am. Chem. Soc. 1986, 108,
3847-3849. (b) Keck, G. E.; Castellino, S. Tetrahedron Lett. 1987, 28,
281-284.
(42) The sulfur in 9 can also be considered a â-heteroatom substitu-
ent in this model. If the tetrahydrothiopyran ring adopts a chair
conformation and the formyl group has an equatorial orientation as
expected, then the C-S bond is necessarily anti to the C-CHO bond
and neither steric nor dipole interactions are possible. If the formyl
group has an axial orientation, then the staggered conformation of 9a
without unfavorable dipole interactions between the carbonyl and ketal
oxygens has an unfavorable dipole interaction with the C-S bond; this
conformation would also suggest Felkin selectivity for addition to the
carbonyl. For examples of aldol reactions with acyclic â-thio-substituted
aldehydes, see: Annunziata, R.; Cinquini, M.; Cozzi, F.; Cozzi, F. G.;
Consolandi, E. J . Org. Chem. 1992, 57, 456-461.
(43) Braun, M. In Stereoselective Synthesis, Methods in Organic
Chemistry, ed. E21; Helmchen, G., Hoffman, R., Multzer, J ., Schrau-
mann, E., Eds.; Thieme: Stuttgart, 1996; Vol. 3, pp 1713-1722.
(44) Evans, D. A.; Halstead, D. P.; Allison, B. D. Tetrahedron Lett.
1999, 40, 4461-4462.
(46) The superior ability of MgBr2 (compared to SnCl4 and TiCl4) to
chelate â-alkoxy aldehydes has been noted previously; see reference
45.
(47) TS E (Figure 2) indicates chelation of the carbonyl oxygen and
the cis oxygen of the ketal group. In principle, chelation involving the
trans oxygen of the ketal is also possible. Such a structure (cf. G with
the H geminal to the O-MOM group replaced with an OCH2) would
also suggest predominant formation of anti-Felkin products with
increased selectivity compared to G because of the increased steric
hindrance to Felkin approach caused by the ketal. The very close
similarity in the diastereoselectivities of the reactions of 9a and 9b
under these conditions favors cis chelation as indicated in TS E.
(48) (a) Evans, D. A.; Yang, M. G.; Dart, M. J .; Duffy, J . L.; Kim, A.
S. J . Am. Chem. Soc. 1995, 117, 9598-9599. (b) Evans, D. A.; Yang,
M. G.; Dart, M. J .; Duffy, J . L. Tetrahedron Lett. 1996, 37, 1957-1960.

Diastereoselectivity of Aldol Reactions
J . Org. Chem., Vol. 67, No. 5, 2002 1625
Sch em e 4. Ald ol Dia ster eoselectivity for Acyclic
Su bstr a tes48
the diastereoselectivities for aldol reactions of 9b and 9c
are strikingly similar to those obtained with the related
acyclic analogues 31 and 32, and the major products can
be rationalized using established TS models.
In conclusion, comparison of the aldol reactions of 9a
with those of 9b and 9c suggests that the ketal group in
9a is a useful stereocontrol element for aldehyde dias-
tereoface selectivity. Under nonchelating conditions, 9a
exhibits very high Felkin selectivity (similar to 9c but
much higher than 9b); under chelating conditions, very
high anti-Felkin selectivity is obtained (similar to 9b but
much higher than 9c). Under optimized conditions, three
of the four possible diastereomers from the reaction of
4b with 9a can be produced selectively by varying the
mediator: with MeLi (via 4c), 12a (9:1); with TiCl4, 11a
(16:1); with MgBr2‚OEt2, 13a (3:1). The fourth diastere-
omer 14a can be obtained in good yield by isomerization
of 13a .6c In the context of polypropionate synthesis,
additional aldol reactions of 11a -14a with 9a (two-
directional or linear) can be envisaged.6b Such reactions
raise several interesting stereoselectivity issues (e.g.,
double diastereoselection, mutual kinetic enantioselec-
tion, etc.). Our efforts in this area will be reported in due
course.
Exp er im en ta l Section 50
Ma ter ia ls. Numerous methods for the preparation of 2 (and
related esters), usually by Dieckmann cyclization of 1, have
been reported.51,52 Most of these methods involve the use of
NaOMe or NaH in various modifications of Fehnel and
Carmack’s improvement51a of the original procedure by Ben-
nett and Scorah.51b Compared to the Fehnel and Carmack
protocol (2 equiv of alcohol-free NaOMe in ether; 64% yield
on a 1.4 mol scale), the modified procedures are typically
reported on a much smaller scale (<0.1 mol), and the yields
obtained vary widely (13-81%; most around 75%). Our experi-
ments suggest that 2 decomposes readily under basic condi-
tions in protic solution, perhaps explaining the variable yields
noted above. The procedure detailed below is simple, reliable,
and easily scalable. The important modifications over the
previous method involve the in situ generation of NaOMe, the
use of sufficient THF to have a (nearly) homogeneous reaction
mixture, and careful quenching of the reaction. Ketone 3 was
prepared from 2 by the known procedure.51a,52e All other
reagents were commercially available and, unless otherwise
noted, were used as received.
Meth yl Tetr a h yd r o-4-oxo-2H-th iop yr a n -3-ca r boxyla te
(2). Anhydrous methanol (freshly distilled from Mg(OMe)2;
43.0 mL, 34.0 g, 1.06 mol) was added dropwise to a stirred
suspension of finely cut sodium metal (23.0 g, 1.00 mol) in
ether (100 mL) and THF (100 mL) at 0 °C under argon
(caution: H2 evolution). The mixture was stirred for 12 h at
rt (some Na metal remained), and then a solution of the diester
1 (100 g, 0.485 mmol) in ether (200 mL) was added dropwise
at 0 °C. The reaction mixture was stirred at rt until the Na
aldehyde 9c and the related anti aldehyde 32 give very
similar diastereoselectivities under similar conditions.
Thus, reactions of the Li (E)-enolates 4c (with 9c) and
33 (with 32) are highly selective for Felkin addition and
anti relative topicity to give the adducts 12c (exclusive)
and 35a s (86% ds),49 respectively; the related Mukaiyama
reactions (i.e., 4b and 9c, TiCl4; 33 (M ) TMS) and 32,
BF3‚OEt2) are also highly selective for Felkin addition,
but with syn relative topicity to give adducts 11c (87%
ds)49 and 35ss (95% ds).49 Interestingly, the minor adduct
from the Mukaiyama reaction of 32 also results from a
Felkin addition, whereas the minor adducts from the
reaction of 9c are anti-Felkin. The diastereoselectivities
of the reactions of the cis aldehyde 9b and the related
syn aldehyde 31 are also similar to each other but are
much lower than those discussed above.19 For example,
reactions of the Li (E)-enolates 4c (with 9b) and 33 (with
31) gave only modest diastereoselectivity. In both cases,
the major product arose from Felkin addition with anti
relative topicity to give adducts 12b (60% ds)49 and 34a s
(51% ds).49 The anti/syn diastereoselectivities among the
Felkin adducts are also similar in the two reactions (i.e.,
12b/11b, 4:1; 34a s/34ss, 4.5:1); however, the anti-Felkin
syn adduct 34sa is the second most-abundant product
(30%) from 31, but the analogous diastereomer (i.e., 13b)
is not detected in the reaction of 9b. Thus, considering
the presence of the sulfur atom and the associated ring
that restricts the number of possible transition states,
(50) See Supporting Information for general methods and proce-
dures.
(51) (a) Fehnel, E. A.; Carmack, M. J . Am. Chem. Soc. 1948, 70,
1813-1817. (b) Bennett, G. M.; Scorah, L. V. D. J . Chem. Soc. 1927,
194-200.
(52) (a) Conroy, J . L.; Sanders, T. C.; Seto, C. T. J . Am. Chem. Soc.
1997, 119, 4285-4291. (b) Li, C.-J .; Chen, D.-L. Synlett 1999, 735-
736. (c) Ghosh, A. K.; Thompson, W. J .; Munson, P. M.; Liu, W.; Huff,
J . R. Bioorg. Med. Chem. Lett. 1995, 5, 83-88. (d) Tamai, S.;
Ushirogochi, H.; Sano, S.; Nagao, Y. Chem. Lett. 1995, 295-296. (e)
Rule, N. G.; Detty, M. R.; Kaeding, J . E.; Sinicropi, J . A. J . Org. Chem.
1995, 60, 1665-1673. (f) Ghosh, A. K.; Liu, W. J . Org. Chem. 1995,
60, 6198-6201. (g) Dowd, P.; Choi, S. C. Tetrahedron 1991, 47, 4847-
4860. (h) Casy, G.; Sutherland, A. G.; Taylor, R. J . K.; Urben, P. G.
Synthesis 1989, 767-769. (i) Liu, H. J .; Ngooi, T. K. Can. J . Chem.
1982, 60, 437-439. (j) Duus, F. Tetrahedron 1981, 37, 2633-2640.
(49) Percent diastereoselectivity (ds) is defined here as the mole
fraction of the major (or designated) diastereomer.

1626 J . Org. Chem., Vol. 67, No. 5, 2002
Ward et al.
metal was completely consumed and the diester was not
detected by TLC (30% ethyl acetate in hexane) (ca. 24-60 h).
The reaction mixture was transferred via cannula to a vigor-
ously stirred solution of acetic acid (85 mL, 89 g, 1.5 mol) in
water (200 mL) at 0 °C. The organic layer was separated, and
the aqueous layer was extracted with ether (3×). The combined
organic layers were washed sequentially with saturated aque-
ous NaHCO3 (2×), water, and brine, dried over Na2SO4, and
concentrated to give 2 as a pale-yellow viscous oil (74-83 g,
88-98%). Spectral data for the product was in accord with that
previously reported.52a
and TLC and was used in the next step without further
purification. IR ν˜max 1738 cm-1; 1H NMR (300 MHz, CDCl3) δ
4.65 (1H, d, J ) 7 Hz), 4.60 (1H, d, J ) 7 Hz), 4.37-4.33 (1H,
m), 3.71 (3H, s), 3.33 (3H, s), 3.17 (1H, dd, J ) 12, 13.5 Hz),
3.03-2.92 (1H, m), 2.77 (1H, ddd, J ) 3, 3.5, 12 Hz), 2.61 (1H,
ddd, J ) 2, 3, 13.5 Hz), 2.38-2.25 (2H, m), 1.87-1.75 (1H,
m); 13C NMR (75 MHz, CDCl3) δ 172.7, 95.9, 72.2, 55.8, 51.9,
48.1, 31.9, 23.7, 22.2; HRMS m/z calcd for C9H16O4S 220.0769,
found 220.0768 (EI).
Meth yl (3R*,4R*)-tetr ah ydr o-4-(m eth oxym eth oxy)th io-
p yr a n -3-ca r boxyla te (5c). Using the above procedure, 7 (5.32
g, 30.2 mmol) was converted into 5c (7.0 g, quantitative), which
3,6-Dih yd r o-4-tr im eth ylsilyloxy-2H-th iop ya n (4b).53
A
1
solution of 3 (5.34 g, 46.0 mmol), Et3N (64 mL, 46 g, 0.46 mol),
and TMSCl (29 mL, 25 g, 0.23 mol) in CH2Cl2 (50 mL) was
stirred at rt under argon in a capped vessel for 10 days. The
reaction mixture was concentrated, diluted with ether, and
filtered through Celite. The combined filtrate and washings
were concentrated, and the residue was placed under high
vacuum (0.5 Torr) for several hours to give 4b as a yellow oil
(8.0-8.5 g, 92-98%), which was homogeneous by 1H NMR and
was used without further purification. The material slowly
decomposed (mainly by hydrolysis) upon storage under argon
at -15 °C. Thus, if the material was not used promptly, a
convenient method of storage involved making a solution of
known concentration in benzene (ca. 1 M) containing 2 equiv
of Et3N. This solution could be stored for months at -15 °C
with negligible decomposition. The product was recovered as
required by the concentration of aliquots. 1H NMR (300 MHz,
CDCl3) δ 5.06-5.04 (1H, m), 3.15-3.14 (2H, m), 2.76-2.72 (2H,
m), 2.27-2.23 (2H, m), 0.17 (9H, s); 13C NMR (75 MHz, CDCl3)
δ 151.3 (s), 102.2 (d), 31.2 (t), 25.7 (t), 25.1 (t), 0.3 (q × 3);
HRMS m/z calcd for C8H16OSSi 188.0708, found 188.0705 (EI).
Gen er a l P r oced u r e for P r ep a r a tion of th e Am in e-F r ee
Lith iu m En ola te 4c.53a Methyllithium (1-1.5 M in diethyl
ether, 1 equiv) was added dropwise via syringe to a stirred
solution of the trimethylsilyl enol ether of the tetrahydro-4H-
thiopyran-4-one in ether (2 mL/mmol of enol ether) at 0 °C
under argon. The reaction mixture was warmed to rt (note:
the Li enolate precipitated from the solution), and after 1 h,
THF (2 mL/mmol of enol ether) was added via syringe. After
stirring for 5 min (note: enolate dissolved), the reaction
mixture was cooled to -78 °C.
Meth yl 1,4-Dioxa -8-th ia sp ir o[4.5]d eca n e-6-ca r boxyla te
(5a ).3b A solution of â-ketoester 2 (30 g, 0.17 mol), ethylene
glycol (43 g, 0.69 mol) and p-TsOH‚H2O (1.7 g, 8.9 mmol) in
benzene (800 mL) was heated under reflux for 16 h with
removal of water via a Dean-Stark trap. The cooled (rt)
reaction mixture was concentrated under reduced pressure,
diluted with ether, washed sequentially with saturated aque-
ous NaHCO3 (2×), water, and brine, dried over Na2SO4, and
concentrated to give 5a as a colorless oil (32-37 g, 85-98%),
which was homogeneous by TLC and NMR and was used
without further purification. IR ν˜max 1733 cm-1; 1H NMR (300
MHz, CDCl3) δ 4.00-3.86 (4H, m), 3.71 (3H, s), 3.11 (1H, ap
dd, J ) 8.5, 13.5 Hz), 2.96-2.78 (3H, m), 2.67 (1H, dddd, J )
1.5, 3.5, 7, 13.5 Hz), 2.25 (1H, ddd, J ) 3.5, 7, 13.5 Hz), 1.82
(1H, ddd, J ) 3.5, 9.5, 13.5 Hz); 13C NMR (75 MHz, CDCl3) δ
171.0 (s), 107.2 (s), 65.1 (t), 64.7 (t), 51.8 (d), 51.2 (q), 36.0 (t),
29.6 (t), 26.7 (t); HRMS m/z calcd for C9H14O4S 218.0613, found
218.0604 (EI). Anal. Calcd for C9H14O4S: C, 49.53; H, 6.46.
Found: C, 49.58; H, 6.39.
was homogeneous by H NMR and TLC and was used in the
1
next step without further purification. IR ν˜max 1736 cm-1; H
NMR (300 MHz, CDCl3) δ 4.69 (1H, d, J ) 7 Hz), 4.63 (1H, d,
J ) 7 Hz), 3.80 (1H, ddd, J ) 4, 9, 10 Hz), 3.73 (3H, s), 3.34
(3H, s), 2.88-2.58 (5H, m), 2.42-2.32 (1H, m), 1.76 (1H, dddd,
J ) 4, 10, 10, 14 Hz); 13C NMR (75 MHz, CDCl3) δ 173.4, 95.7,
75.7, 55.7, 52.0, 50.2, 32.8, 29.7, 26.9; HRMS m/z calcd for
C9H16O4S 220.0769, found 220.0771 (EI).
Meth yl (3S*,4R*)-tetr a h yd r o-4-h yd r oxy-2H-th iop yr a n -
3-ca r boxyla te (6) a n d Meth yl (3R*,4R*)-tetr a h yd r o-4-
h yd r oxy-2H-th iop yr a n -3-ca r boxyla te (7).52f,54 NaCNBH3
(14.0 g, 0.223 mol) was added in four equal portions at 5 min
intervals to a stirred solution of â-ketoester 2 (40.0 g, 0.225
mol) and citric acid (48 g, 0.228 mol) in ethanol (200 mL) at 0
°C. (Note: addition of NaCNBH3 is exothermic.) The ice bath
was removed, and the progress of the reaction was monitored
by TLC (50% ethyl acetate in hexane); after ca. 40 min, 2 had
been consumed. The reaction mixture was concentrated and
then taken up in ethyl acetate and washed with H2O and brine,
dried over Na2SO4, and concentrated to give the crude hydrox-
1
yesters (a 2:1 mixture of 6/7 by H NMR) as a light-yellow oil
(32.6 g). (Note: If the reaction was left for longer than 1 h, a
gel-like suspension that impeded the removal of ethanol by
rotary evaporation formed. In these cases, the mixture was
diluted with ethyl acetate and washed with brine (3×). The
aqueous phases were extracted with ethyl acetate, and the
combined organic layers were dried over Na2SO4 and concen-
trated to give the crude hydroxyesters 6 and 7.) The crude
product was fractionated by DFC (5-50% ethyl acetate in
hexane) to give the cis hydroxyester 6 as a colorless oil (19.89
g, 49%) and the trans hydroxyester 7 as a colorless solid (9.85
g, 24%). 6: IR ν˜max 3503, 1719 cm-1 1H NMR (300 MHz,
;
CDCl3) δ 4.17 (1H, ddd, J ) 3, 3, 5.5 Hz), 3.72 (3H, s), 3.16
(1H, dd, J ) 10.5, 13 Hz), 3.10-2.90 (1H, m), 2.99 (1H, ddd, J
) 3, 11.5, 14 Hz), 2.85 (1H, ddd, J ) 3, 3, 10.5 Hz), 2.57 (1H,
dd, J ) 3, 13.5 Hz), 2.32 (1H, dddd, J ) 1.5, 4, 5, 13.5 Hz),
2.16 (1H, dddd, J ) 3, 5.5, 5.5, 14 Hz), 1.89 (1H, dddd, J ) 3,
3, 11, 14 Hz); 13C NMR (75 MHz, CDCl3) δ 174.5 (s), 66.0 (d),
52.4 (q), 47.6 (d), 33.5 (t), 25.3 (t), 23.1 (t); HRMS m/z calcd
for C7H12O3S 176.0507, found 176.0508 (EI). 7: IR ν˜max 3452,
1728 cm-1; 1H NMR (300 MHz, CDCl3) δ 3.72 (1H, ddd, J ) 4,
11, 11 Hz), 3.70 (3H, s), 3.00 (1H, br s), 2.83 (1H, ddd, J ) 2,
3, 12 Hz), 2.71-2.61 (3H, m), 2.62 (1H, ddd, J ) 3, 11.5, 11.5
Hz), 2.28 (1H, dddd, J ) 3.5, 3.5, 3.5, 13.5 Hz), 1.68 (1H, dddd,
J ) 5, 11, 11, 13 Hz); 13C NMR (75 MHz, CDCl3) δ 174.0 (s),
70.3 (d), 52.3 (q), 52.0 (d), 35.3 (t), 29.8 (t), 27.7 (t); HRMS m/z
calcd for C7H12O3S 176.0507, found 176.0508 (EI).
1,4-Dioxa -8-th ia sp ir o[4.5]d eca n e-6-m eth a n ol (8a ).3b
A
solution of 5a (22.5 g, 103 mmol) in anhydrous ether (100 mL)
was added dropwise via syringe to a stirred suspension of
LiAlH4 (4.5 g, 113 mmol) in ether (200 mL) at 0 °C under
argon. After 1 h, the reaction was quenched55 by sequential
dropwise addition of water (4.5 mL), 15% (w/v) sodium
hydroxide (4.5 mL), and water (13.5 mL). The resultant gray
suspension was stirred until granular white flakes formed (ca.
1-1.5 h) and then filtered through Celite using 1:1 ether/
hexane. The combined filtrate and washings were washed with
brine, dried over Na2SO4, and concentrated to give 8a as
Meth yl (3S*,4R*)-tetr a h yd r o-4-(m eth oxym eth oxy)-2H-
th iop yr a n -3-ca r boxyla te (5b). MOM-Cl (3.2 mL, 3.4 g, 42
mmol) was added dropwise over 2 min to a stirred solution of
6 (3.76 g, 21.3 mmol) and DIEA (7.4 mL, 5.5 g, 43 mmol) in
CH2Cl2 (20 mL) at 0 °C. The reaction was allowed to stand at
rt with periodic monitoring by TLC (50% ethyl acetate in
hexane). After 3 days (reaction complete by TLC), the mixture
was diluted with CH2Cl2 (200 mL), washed with 1 M HCl (2
×) and H2O, dried over Na2SO4, and concentrated to give 5b
(4.75 g, quantitative), which was homogeneous by 1H NMR
(54) Moser, U.; Lambrecht, G.; Mutschler, E. Arch. Pharm. (Wein-
heim, Ger.) 1983, 316, 670-677.
(55) Micovic, V. M.; Mihailovic, M. L. J . J . Org. Chem. 1953, 18,
1190-1200.
(53) (a) Aoki, S.; Fujimura, T.; Nakamura, E. J . Am. Chem. Soc.
1992, 114, 2985-2990. (b) Biondi, S.; Piga, E.; Rossi, T.; Vigelli, G.
Bioorg. Med. Chem. Lett. 1997, 7, 2061-2067.

Diastereoselectivity of Aldol Reactions
J . Org. Chem., Vol. 67, No. 5, 2002 1627
a pale-yellow viscous oil (17.0-18.5 g, 86-94%), which was
homogeneous by TLC and NMR and was used without further
purification. IR ν˜max 3442 cm-1; 1H NMR (300 MHz, CDCl3) δ
4.03-3.97 (4H, m), 3.91 (1H, dd, J ) 6.5, 11 Hz), 3.66 (1H,
dd, J ) 4.5, 11 Hz), 2.80-2.71 (3H, m), 2.65 (1H, ddd, J ) 4,
7, 13.5 Hz), 2.25 (1H, br s), 2.19-2.13 (1H, m), 2.07 (1H, ddd,
J ) 3.5, 7, 13.5 Hz), 1.77 (1H, ddd, J ) 4, 9.5, 13.5 Hz); 13C
NMR (75 MHz, CDCl3) δ 109.5 (s), 64.7 (t), 64.4 (t), 62.3 (t),
47.0 (d), 35.1 (t), 29.3 (t), 26.6 (t); HRMS m/z calcd for C8H14O3S
190.0664, found 190.0672 (EI).
(1H, d, J ) 7 Hz), 4.40 (1H, ddd, J ) 2.5, 2.5, 5 Hz), 3.31 (3H,
s), 3.04 (1H, ap dd, J ) 12.5, 12.5 Hz), 2.93 (1H, ap dd, J )
12, 12.5 Hz), 2.70-2.60 (2H, m), 2.41-2.26 (2H, m), 1.90-1.78
(1H, m); 13C NMR (75 MHz, CDCl3) δ 202.1, 95.3, 70.6, 53.9,
46.9, 31.3, 23.2, 22.9; HRMS m/z calcd for C8H14O3S 190.0664,
found 190.0666 (EI).
(3S*,4S*)-Tetr a h yd r o-4-(m eth oxym eth oxy)-2H-th iop y-
r a n -3-ca r boxa ld eh yd e (9c). Sw er n Oxid a tion of 8c. Using
the above procedure, 8c (761 mg) was converted into the trans
aldehyde 9c (170 mg, 23%).
DIBAH Red u ction of 5c.56 Cold DIBAL-H (0.5 M in
toluene; 26.2 mL, 13 mmol) was added dropwise via syringe
pump over 3 h to a stirred solution of the ester 5c (1.581 g,
7.18 mmol) in toluene (10 mL) at -78 °C under argon. This
addition was achieved by having the output from the syringe
pump run down the side of a coldfinger condenser (dry ice/
acetone) mounted above the reaction mixture. After 3 h, the
reaction was quenched by the addition of cold MeOH (9 mL)
via syringe pump over 2 h as above. The reaction mixture was
allowed to warm to rt over several hours and then ice-cold 1
M HCl (50 mL) was added. After 10 min of vigorous stirring,
the mixture was diluted with brine and extracted with CH2-
Cl2 (3×). The combined organic layers were dried over Na2-
SO4, concentrated, and fractionated by MPC (30% ethyl acetate
in hexane) to give 8c (193 mg, 14%) and 9c (939 mg, 69%). IR
ν˜max 2824, 2725, 1721 cm-1; 1H NMR (300 MHz, CDCl3) δ 9.80
(1H, d, J ) 1.5 Hz), 4.73 (1H, d, J ) 7 Hz), 4.62 (1H, d, J ) 7
Hz), 3.94 (1H, ddd, J ) 3.5, 8, 8 Hz), 3.35 (3H, s), 2.97 (1H, br
d, J ) 12.5 Hz), 2.86-2.66 (3H, m), 2.52 (1H, ddd, J ) 3, 9,
13.5 Hz), 2.23 (1H, dddd, J ) 3, 3.5, 7.5, 13 Hz), 1.85 (1H,
dddd, J ) 3, 8, 9, 13.5 Hz); 13C NMR (300 MHz, CDCl3) δ 202.6
(d), 95.4 (t), 73.5 (d), 55.9 (q), 54.1 (d), 31.7 (t), 26.1 (t), 25.7
(t); HRMS m/z calcd for C8H14O3S 190.0664, found 190.0664
(EI).
Ald ol R ea ct ion of 4b a n d 9a (TiCl4). Freshly distilled
TiCl4 (0.30 mL, 0.51 g, 2.7 mmol) was added dropwise over 1
min to a stirred solution of 9a (500 mg, 2.66 mmol) in CH2Cl2
(35 mL) at -78 °C under argon. The resulting fine, yellow
suspension was stirred for 10 min, and then a solution of 4b
(736 mg, 3.91 mmol) in CH2Cl2 (7.5 mL) was added dropwise
via syringe over 2 min, whereupon the yellow suspension
turned to a clear dark-orange and then to a red solution. After
1 h at -78 °C, the reaction was allowed to warm to rt over 30
min before being quenched by the sequential addition of a
solution of Et3N (80 mg, 0.79 mmol) and MeOH (100 mg, 3.1
mmol) in CH2Cl2 (1 mL) and then saturated NH4Cl (5 mL).
The mixture was diluted with water and extracted with CH2-
Cl2 (3×), and the combined organic layers were dried over Na2-
SO4, concentrated, and fractionated by FCC (20-50% ethyl
acetate in hexane) to give 12a (18 mg, 5%) and 11a (663 mg,
82%) as white solids.
(3′R*,4′R*)-[Tetr a h yd r o-4-(m eth oxym eth oxy)-2H-th io-
p yr a n -3-yl]m eth a n ol (8b). Using the above procedure, re-
duction of the cis ester 5b (2.423 g, 11.0 mmol) with LiAlH4
(543 mg, 14.3 mmol) in ether (50 mL) and THF (7 mL; used to
dissolve 5b) gave 8b as a clear oil (1.899 g, 90%), which was
1
homogeneous by H NMR and TLC and was used in the next
step without further purification. IR ν˜max 3426 cm-1; 1H NMR
(300 MHz, CDCl3) δ 4.67 (1H, d, J ) 6.5 Hz), 4.60 (1H, d, J )
6.5 Hz), 3.95 (1H, br t, J ) 2.5, 2.5 Hz), 3.63 (1H, dd, J ) 7.5,
11 Hz), 3.61 (1H, dd, J ) 7.5, 11 Hz), 3.38 (3H, s), 2.96-2.71
(2H, m), 2.40-2.25 (3H, m), 2.18 (1H, dddd, J ) 2, 4, 4, 13
Hz), 2.03 (1H, m), 1.77 (1H, dddd, J ) 2.5, 4, 11.5, 14 Hz); 13
C
NMR (75 MHz, CDCl3) δ 95.8, 73.2, 64.2, 56.1, 43.8, 31.3, 25.9,
23.6; HRMS m/z calcd for C8H16O3S 192.0820, found 192.0820
(EI).
(3′R*,4′S*)-[Tetr a h yd r o-4-(m eth oxym eth oxy)-2H-th io-
p yr a n -3-yl]m eth a n ol (8c). Using the above procedure, 5c
(1.016 g, 4.62 mmol) was converted to 8c (832 mg, 94%), which
1
was homogeneous by H NMR and TLC and was used in the
1
next step without further purification. IR ν˜max 3451 cm-1; H
NMR (300 MHz, CDCl3) δ 4.73 (1H, d, J ) 7 Hz), 4.60 (1H, d,
J ) 7 Hz), 3.82 (1H, dd, J ) 4.5, 11 Hz), 3.66 (1H, dd, J ) 4.5,
11 Hz), 3.44 (1H, ddd, J ) 3.5, 10, 10 Hz), 3.40 (3H, s), 2.75-
2.53 (4H, m), 2.45 (1H, br s), 2.31 (1H, dddd, J ) 3.5, 4, 4, 13
Hz), 1.96-1.87 (1H, m), 1.79-1.66 (1H, m); 13C NMR (75 MHz,
CDCl3) δ 95.5 (t), 77.8 (d), 64.7 (t), 56.0 (q), 46.3 (d), 33.4 (t),
30.1 (t), 27.4 (t); HRMS m/z calcd for C8H16O3S 192.0820, found
192.0820 (EI).
1,4-Dioxa-8-th iaspir o[4.5]decan e-6-car boxaldeh yde (9a).
Sw er n Oxid a tion . A solution of DMSO (2.3 mL, 2.5 g, 32
mmol) in CH2Cl2 (5.0 mL) was added dropwise via syringe to
a stirred solution of oxalyl chloride (1.4 mL, 16 mmol) in CH2-
Cl2 (30 mL) at -78 °C under argon. After 30 min at -78 °C, a
solution of 8a (2.00 g, 10.5 mmol) in CH2Cl2 (10 mL) was added
dropwise via syringe to the reaction mixture. After another
30 min, i-Pr2EtN (9.2 mL, 53 mmol) was added, and the
reaction mixture was allowed to warm to rt. The mixture was
diluted with CH2Cl2, washed sequentially with aqueous NH4-
Cl and NaHCO3, dried over Na2SO4, and concentrated. Frac-
tionation by FCC (35% ethyl acetate in hexane) gave 9a as a
pale-yellow oil (1.30-1.70 g, 66-86%).
P CC Oxid a tion . PCC (4.90 g, 22.7 mmol) was added in
three equal portions at 15 min intervals to a vigorously stirred
suspension of 8a (2.00 g, 10.5 mmol), NaOAc (2.58 g, 31.5
mmol), and Celite (5 g) in CH2Cl2 (50 mL) at 0 °C under argon.
The reaction mixture was allowed to warm to rt, and after 3
h, it was filtered through a short plug of SiO2 using ether. The
combined filtrate and washings were concentrated and frac-
tionated by FCC (25% ethyl acetate in hexane) to give 9a as a
colorless oil (1.1-1.4 g, 55-70%). IR ν˜max 2840, 2737, 1721
Ald ol Rea ction of 4b a n d 9a (BF 3‚OEt2). Using the same
procedure as above but replacing TiCl4 with BF3‚OEt2, 9a (73
mg, 0.039 mmol) gave a 2:1 mixture of 11a /12a (70%).
Ald ol Rea ction of 4b a n d 9b (TiCl4). The cis aldehyde
9b was unstable in the presence of TiCl4 (elimination). Using
the above procedure, 9b (73 mg, 0.38 mmol) gave 12b (11 mg,
9%) and 14b (15 mg, 12%) and recovered 9b (16 mg, 22%) after
fractionation by MPC (30% ethyl acetate in hexane).
Ald ol Rea ction of 4b a n d 9c (TiCl4). Using the above
procedure with TiCl4 (0.018 mL, 31 mg, 0.16 mmol), 9c (30
mg, 0.16 mmol) and 4b (60 mg, 0.32 mmol) in CH2Cl2 (5 mL)
gave, after workup, a light-orange oil (78 mg), which contained
a 10:1 mixture of 11c/14c and ca. 8% of the remaining 9c (1H
NMR). The crude product was fractionated by MPC (35% ethyl
acetate in hexane) to give a 20:2:1 mixture of 11c/14c/13c (40
mg, 81%). A pure sample of 11c was obtained by fractionation
of a portion of the above mixture by PTLC (50% ethyl acetate
in hexane).
cm-1 1H NMR (300 MHz, CDCl3) δ 9.85 (1H, s), 4.07-3.94
;
(4H, m), 2.96 (1H, dd, J ) 9.5, 13.5 Hz), 2.86 (1H, br d, J )
13.5 Hz), 2.81-2.72 (2H, m), 2.64 (1H, m), 2.08 (1H, ddd, J )
3, 6, 13.5 Hz), 1.89 (1H, ddd, J ) 3.5, 10, 13.5 Hz); 13C NMR
(75 MHz, CDCl3) δ 201.3 (d), 107.8 (s), 64.9 (t), 64.7 (t), 56.6
(d), 36.2 (t), 26.7 (t), 26.4 (t); HRMS m/z calcd for C8H12O3S
188.0507, found 188.0512 (EI). Anal. Calcd for C8H12O3S: C,
51.04; H, 6.43. Found: C, 51.20; H, 6.58.
(3S*,4R*)-Tetr a h yd r o-4-(m eth oxym eth oxy)-2H-th iop y-
r a n -3-ca r boxa ld eh yd e (9b). Swern oxidation of 8b (701 mg,
3.65 mmol), according to the above procedure, gave a light-
yellow oil (672 mg) after workup, which was fractionated by
FCC (30% ethyl acetate in hexane) to give the cis aldehyde
Ald ol Rea ction of 4b a n d 9a (MgBr 2). MgBr2‚OEt2 (4.45
g, 17.4 mmol) was added to a stirred solution of 9a (1.08 g,
5.74 mmol) in CH2Cl2 (26 mL) at rt under argon. After 2 min,
9b (617 mg, 89%). IR ν˜max 2823, 2722, 1726 cm-1 1H NMR
;
(300 MHz, CDCl3) δ 9.63 (1H, s), 4.69 (1H, d, J ) 7 Hz), 4.57
(56) Garner, P.; Park, J . M. Org. Synth. 1992, 70, 18-26.

1628 J . Org. Chem., Vol. 67, No. 5, 2002
Ward et al.
the resulting creamy, off-white suspension was placed in an
ice bath, and after 15 min, a solution of 4b (2.16 g, 11.5 mmol)
in CH2Cl2 (1 mL) was added dropwise via syringe over 3 min.
After stirring for 1 h at 0 °C, the reaction mixture was poured
into ice-cold phosphate buffer (pH 7, 50 mL) with vigorous
stirring. The mixture was diluted with water and extracted
with CH2Cl2 (3×). The combined organic layers were dried over
Na2SO4 and concentrated to give an orange oil that contained
a 3:1 mixture of 13a /14a , in addition to 3 and 9a (ca. 3%).
The relatively volatile 3 (621 mg, 47% based on 4b) was easily
recovered from the crude mixture by sublimation at high
vacuum (rt). The remaining residue was crystallized from
methanol to give 13a (mp 152-153 °C; 1.07 g, 61%). The
mother liquor (ca. 4:1 mixture of 14a /13a ) was concentrated
and fractionated by MPC (25-50% ethyl acetate in hexane)
to give 13a (101 mg, 6%) and 14a (381 mg, 22%).
diastereomers was not detected by 1H NMR analysis of the
crude product.
Ald ol Rea ction of 4d (4e) w ith 9a . A solution of 3 (20
mg, 0.17 mmol) in CH2Cl2 (0.5 mL) was added to a stirred
solution of Bu2BOTf (1 M in CH2Cl2; 0.26 mL) and i-Pr2EtN
(0.090 mL, 67 mg, 0.52 mmol) in CH2Cl2 (5 mL) at -78 °C
under argon. After 3 h, a solution of 9a (130 mg, 0.69 mmol)
in CH2Cl2 (0.5 mL) was added to the reaction mixture. After
3 h, the reaction was quenched by the sequential addition of
phosphate buffer (pH 7, 0.5 mL), methanol (3.5 mL), and 30%
H2O2 (0.2 mL). The mixture was stirred at 0 °C for 15 min,
and then aqueous Na2SO3 was added to reduce the H2O2. The
mixture was diluted with water and extracted with CH2Cl2
(3×), and the combined organic layers were dried over Na2-
SO4, concentrated, and fractionated by FCC (20-40% ethyl
acetate in hexane) to give recovered 9a (58 mg, 45%), 11a (4
mg, 8%), and 12a (39 mg, 74%). A similar experiment using
(c-C6H11)2BCl in place of Bu2BOTf gave a 15:1 mixture of 11a /
12a (44 mg, 84%).
Ald ol Rea ction of 4f w ith 9a . TiCl4 (0.030 mL, 52 mg,
0.28 mmol) was added to a stirred solution of 3 (29 mg, 0.25
mmol) in CH2Cl2 (4 mL) at -78 °C under argon. After 5 min,
i-Pr2EtN (0.052 mL, 39 mg, 0.30 mmol) was added, and after
2 h at -78 °C, a solution of 9a (56 mg, 0.30 mmol) in CH2Cl2
(0.5 mL) was added. After 5 h, the reaction was quenched by
the addition of aqueous NH4Cl. The mixture was diluted with
water and extracted with CH2Cl2 (3×), and the combined
organic layers were dried over Na2SO4, concentrated, and
fractionated by FCC (20-50% ethyl acetate in hexane) to give
11a (20 mg, 26%) and a 1:1 mixture of Cs/C1 bisaldols (26 mg,
20%).6b
Ald ol Rea ction of 4b a n d 9a (Sn Cl4). Using the same
procedure as above but using SnCl4 (1 M in CH2Cl2, 2 equiv)
in place of MgBr2‚OEt2 (3 equiv), 9a (25 mg, 0.13 mmol) gave
a 3:1 mixture of 12a /11a (72%).
Ald ol Rea ction of 4b a n d 9b (MgBr 2). Using the above
procedure, the cis aldehyde 9b (573 mg, 3.01 mmol) gave an
orange-yellow oil on workup that contained a 4:1 mixture of
13b/14b in addition to 3 and 9b (ca. 2%). The relatively volatile
3 (321 mg, 46% on the basis of 4b) was easily recovered from
the crude mixture at high vacuum (rt). The remaining residue
was crystallized from methanol to give 13b (525 mg, 57%),
and the mother liquor was fractionated by MPC (30% ethyl
acetate in hexane) to give a 2.5:1 mixture of 14b/13b (257 mg,
27%). A pure sample of 14b was obtained by fractional
crystallization of the above mixture from methanol, resulting
in a mother liquor enriched in 14b (ca. 5:1) that was further
fractionated by PTLC (2% methanol in CH2Cl2, multiple
development).
(1′S*,3S*,6′′R*)-3-[(1,4-Dioxa -8-th ia sp ir o[4.5]d ec-6-yl)-
h yd r oxym eth yl]tetr a h yd r o-4H-th iop yr a n -4-on e (11a ). IR
1
ν˜max 3507, 1703 cm-1; H NMR (300 MHz, CDCl3) δ 4.79 (1H,
Ald ol Rea ction of 4b a n d 9c (MgBr 2). Using the above
procedure, the trans aldehyde 9c (206 mg, 1.08 mmol) gave a
7:4:1 mixture of 13c/11c/14c(245 mg, 74%) after fractionation
by MPC (30% ethyl acetate in hexane). A pure sample of 13c
was obtained by further fractionation of a portion of the
mixture by PTLC (2% methanol in CH2Cl2). A pure sample of
14c could not be obtained. Extensive fractionation of the
mixture by PTLC (2% methanol in CH2Cl2) gave a 1.5:1
mixture of 14c/13c.
dd, J ) 3, 8.5 Hz, HC-1′ (3J HC-1′/HC-3 ) 8.5 Hz)), 4.10-3.96
(4H, m), 3.19-2.88 (7H, m), 2.81-2.67 (4H, m), 2.58-2.54 (1H,
m), 2.11 (1H, ddd, J ) 3, 5.5, 14 Hz), 1.93 (1H, ddd, J ) 3, 3,
10.5 Hz), 1.73 (1H, ddd, J ) 3.5, 11.5, 14 Hz); 13C NMR (75
MHz, CDCl3) δ 210.1 (s), 109.8 (s), 66.1 (d), 64.6 (t), 64.2 (t),
56.0 (d), 46.2 (d), 44.2 (t), 35.6 (t), 32.6 (t), 31.2 (t), 26.4 (t),
26.3 (t); HRMS m/z calcd for C13H20O4S2 304.0803, found
304.0799. Anal. Calcd for C13H20O4S2: C, 51.29; H, 6.62.
Found: C, 51.43; H, 6.44.
Ald ol Rea ction of 4c a n d 9a . A solution of 9a (2.32 g, 12.3
mmol) in THF (5 mL) was added rapidly via syringe to a
stirred solution of the amine-free 4c (prepared as above from
4b: 3.48 g, 18.5 mmol) at -78 °C. After 5 min, it was quenched
by the rapid addition of a solution of glacial acetic acid (1.5
mL) in THF (5 mL). The reaction mixture was removed from
the cooling bath, and CH2Cl2 (100 mL) and water (50 mL) were
added. The aqueous layer was extracted with CH2Cl2 (2×), and
the combined organic layers were washed sequentially with
aqueous NaHCO3 and water, dried over Na2SO4, and concen-
trated to give a pale-yellow semisolid. Trituration of the crude
product with ether gave a yellowish powder that was recrys-
tallized from benzene to give the pure 12a (2.06 g, 55%; mp
148-149 °C). The ether from trituration and the mother liquor
from recrystallization were combined and concentrated, and
3 (910 mg, 42%) was recovered by sublimation (rt, 0.5 Torr).
The residue was fractionated by FCC (10-50% EtOAc in
hexane) to recover 9a (541 mg, 23%) and to give 11a (270 mg,
7%) and 12a (300 mg, 8%).
Ald ol Rea ction of 4c a n d 9b. Using the same procedure
as above, the reaction of 9b (580 mg, 3.05 mmol) with the
amine-free 4c (prepared as above from 4b: 1.15 g, 6.11 mmol)
gave, after workup, a yellow oil (1.413 g) that contained a 4.1:
1.7:1 mixture of 12b/14b/11b (1H NMR). Fractionation of the
crude product by MPC (25% ethyl acetate in hexane) gave 11b
(88 mg, 9%), 12b (220 mg, 24%), and a 1.1:1 mixture of 14b/
12b (258 mg, 28%).
(1′S*,3R*,6′′R*)-3-[(1,4-Dioxa -8-th ia sp ir o[4.5]d ec-6-yl)-
h yd r oxym eth yl]tetr a h yd r o-4H-th iop yr a n -4-on e (12a ). IR
ν˜max 3488, 3409, 1711 cm-1; 1H NMR (300 MHz, CDCl3): δ 4.50
(1H, dd, J ) 4.5, 6.5 Hz, HC-1′ (3J HC-1′/HC-3 ) 6.5 Hz)), 4.05-
3.92 (4H, m), 3.08-2.58 (12H, m), 2.12 (1H, ddd, J ) 4.5, 4.5,
9 Hz), 2.03 (1H, ddd, J ) 3.5, 6.5, 13.5 Hz), 1.74 (1H, ddd, J
) 4, 9.5, 13.5 Hz); 13C NMR (75 MHz, CDCl3): δ 211.5 (s),
109.3 (s), 69.3 (d), 64.4 (t), 64.3 (t), 55.5 (d), 47.0 (d), 44.4 (t),
35.5 (t), 34.3 (t), 31.4 (t), 27.4 (t), 26.5 (t); HRMS m/z calcd for
C
C
13H20O4S2 304.0803, found 304.0801. Anal. Calcd for
13H20O4S2: C, 51.29; H, 6.62. Found: C, 51.59; H, 6.55.
(1′R*,3R*,6′′R*)-3-[(1,4-Dioxa -8-th ia sp ir o[4.5]d ec-6-yl)-
h yd r oxym eth yl]tetr a h yd r o-4H-th iop yr a n -4-on e (13a ). IR
1
ν˜max 3497, 1706 cm-1; H NMR (300 MHz, CDCl3) δ 4.57 (1H,
ddd, J ) 2.5, 3.5, 7.5 Hz, HC-1′ (3J HC-1′/HC-3 ) 3.5 Hz)), 4.10-
3.89 (4H, m), 3.85 (1H, d, J ) 2.5 Hz, OH), 3.13 (1H, dd, J )
11, 13.5 Hz), 3.0 (1H, ddd, J ) 4, 10.5, 13.5 Hz), 2.91-2.57
(9H, m), 2.22-2.13 (1H, m), 2.07 (1H, ddd, J ) 3.5, 7.5, 8 Hz),
1.82-1.71 (1H, m); 13C NMR (75 MHz, CDCl3) δ 209.1 (s), 110.8
(s), 69.7 (d), 64.8 (t), 64.1 (t), 55.3 (d), 46.2 (d), 44.1 (t), 34.9
(t), 29.7 (t), 29.4 (t), 29.3 (t), 26.7 (t); HRMS m/z calcd for
C
C
13H20O4S2 304.0803, found 304.0807. Anal. Calcd for
13H20O4S2: C, 51.29; H, 6.62. Found: C, 51.34; H, 6.66.
(1′R*,3S*,6′′R*)-3-[(1,4-Dioxa -8-th ia sp ir o[4.5]d ec-6-yl)-
h yd r oxym eth yl]tetr a h yd r o-4H-th iop yr a n -4-on e (14a ). IR
1
ν˜max 3496, 1702 cm-1; H NMR (300 MHz, CDCl3) δ 4.24 (1H,
ddd, J ) 3.5, 3.5, 7.5 Hz, HC-1′ (3J HC-1′/HC-3 ) 3.5 Hz)), 4.12-
3.95 (4H, m), 3.93 (1H, d, J ) 3.5 Hz, OH), 3.25 (1H, dd, J )
11.5, 14.5 Hz), 3.02 (1H, ddd, J ) 4, 10, 13.5 Hz), 2.95-2.56
(9H, m), 2.41 (1H, ddd, J ) 3, 7.5, 7.5 Hz), 2.15 (1H, ddd, J )
4, 9, 13.5 Hz), 1.79 (1H, ddd, J ) 3.5, 7, 13.5 Hz); 13C NMR
(75 MHz, CDCl3) δ 210.1 (s), 110.5 (s), 72.9 (d), 64.7 (t), 64.2
Ald ol Rea ction of 4c a n d 9c. Using the same procedure
as above, the reaction of the trans aldehyde 9c (525 mg, 2.76
mmol) with the amine-free 4c (prepared from 4b: 1.08 g, 5.73
mmol) gave 12b (542 mg, 64%) after fractionation by MPC
(35% ethyl acetate in hexane). The presence of other aldol

Diastereoselectivity of Aldol Reactions
J . Org. Chem., Vol. 67, No. 5, 2002 1629
(t), 54.7 (d), 46.4 (d), 44.2 (t), 34.5 (t), 33.2 (t), 30.0 (t), 29.4 (t),
26.8 (t); HRMS m/z calcd for C13H20O4S2 304.0803, found
304.0801.
(1H, ap ddd, J ) 5, 5, 8.5 Hz), 2.79 (1H, ddd, J ) 2, 3, 13.5
Hz), 2.72-2.65 (3H, m), 2.64 (1H, dd, J ) 9.5, 13.5 Hz), 2.57
(1H, ddd, J ) 3, 11, 13.5 Hz), 2.20 (1H, dddd, J ) 3, 3.5, 6, 13
Hz), 1.79 (1H, dddd, J ) 3.5, 9, 11, 13 Hz), 1.72 (1H, dddd, J
) 3, 4, 9, 9.5 Hz); 13C NMR (100 MHz, CDCl3) δ 210.29, 96.33,
76.79, 66.30, 56.55, 55.26, 45.61, 44.19, 33.32, 33.11, 31.50,
26.85, 26.24; HRMS m/z calcd for C13H22O4S2 306.0960, found
306.0962 (EI).
(1′R*,3S*,3′′R*,4′′R*)-3-[Hyd r oxy(4-(m eth oxym eth oxy)-
tetr a h yd r o-2H-th iop yr a n -3-yl)m eth yl]tetr a h yd r oth iop y-
r a n -4-on e (11b). IR ν˜max 3496, 1701 cm-1; 1H NMR (500 MHz,
CDCl3) δ 4.74 (1H, d, J ) 7 Hz), 4.60 (1H, d, J ) 7 Hz), 4.23
(1H, br t, J ) 5.5, 5.5 Hz, HC-1′ (3J HC-1′/HC-3 ) 5.5 Hz)), 3.96
(1H, ddd, J ) 2, 2.5, 5 Hz), 3.36 (3H, s), 3.07-2.85 (8H, m),
2.75-2.71 (2H, m), 2.57 (1H, br dd, J ) 3, 13 Hz), 2.34-2.28
(2H, m), 1.90 (1H, dddd, J ) 2.5, 3, 5.5, 11.5 Hz), 1.77 (1H,
m); 13C NMR (125 MHz, CDCl3) δ 211.37, 94.86, 73.92, 71.17,
56.36, 54.77, 45.27, 43.74, 31.57, 31.29, 31.23, 23.86, 22.49;
HRMS m/z calcd for C13H22O4S2 306.0960, found 306.0956 (EI).
(1′R*,3R*,3′′R*,4′′R*)-3-[Hyd r oxy(4-(m eth oxym eth oxy)-
tetr a h yd r o-2H-th iop yr a n -3-yl)m eth yl]tetr a h yd r oth iop y-
r a n -4-on e (12b). IR ν˜max 3508, 1700 cm-1; 1H NMR (500 MHz,
CDCl3) δ 4.71 (1H, d, J ) 7 Hz), 4.59 (1H, d, J ) 7 Hz), 3.96
(1H, ddd, J ) 3, 6, 7.5 Hz, HC-1′ (3J HC-1′/HC-3 ) 6 Hz)), 3.89
(1H, ddd, J ) 2.5, 3, 5 Hz), 3.41 (3H, s), 3.15-2.94 (7H, m),
2.90 (1H, br dd, J ) 12, 13 Hz), 2.81-2.71 (2H, m), 2.64 (1H,
br d, J ) 13.5 Hz), 2.34 (1H, br ddd, J ) 3.5, 4, 13 Hz), 2.26
(1H, dddd, J ) 2.5, 4, 5, 14.5 Hz), 2.14 (1H, dddd, J ) 3, 3.5,
6, 11 Hz), 1.77 (1H, dddd, J ) 2.5, 3.5, 12, 14.5 Hz); 13C NMR
(125 MHz, CDCl3) δ 211.84 (s), 94.89 (t), 74.97 (d), 74.37 (d),
56.39 (q), 54.06 (d), 45.12 (d), 45.05 (t), 35.20 (t), 31.69 (t), 31.15
(1′R*,3R*,3′′R*,4′′S*)-3-[Hyd r oxy(4-(m eth oxym eth oxy)-
tetr a h yd r o-2H-th iop yr a n -3-yl)m eth yl]tetr a h yd r oth iop y-
r a n -4-on e (12c). IR ν˜max 3516, 1701 cm-1; 1H NMR (500 MHz,
CDCl3) δ 4.72 (1H, d, J ) 6.5 Hz), 4.70 (1H, d, J ) 6.5 Hz),
4.47 (1H, ddd, J ) 2.5, 4, 9.5 Hz, HC-1′ (3J HC-1′/HC-3 ) 9.5 Hz)),
3.61 (1H, ddd, J ) 4, 10, 10.5 Hz), 3.43 (3H, s), 3.22 (1H, d, J
) 4 Hz, OH), 3.03-2.93 (2H, m), 2.92-2.83 (2H, m), 2.82-
2.65 (5H, m), 2.58 (1H, m, J ) 2.5, 3.5, 4, 13.5 Hz), 2.51 (1H,
ddd, J ) 2.5, 3, 13.5 Hz), 2.42 (1H, dddd, J ) 3, 4, 4, 12.5 Hz),
1.77 (1H, dddd, J ) 3, 3, 9.5, 10 Hz), 1.74 (1H, dddd, J ) 3,
10.5, 12, 12.5 Hz); 13C NMR (125 MHz, CDCl3) δ 212.35 (s,
CO), 96.52 (t), 76.18 (d), 67.72 (d), 56.03 (q), 55.45 (d), 45.73
(d), 44.99 (t), 34.90 (t), 32.42 (t), 31.07 (t), 27.75 (t), 26.31 (t);
HRMS m/z calcd for C13H22O4S2 306.0960, found 306.0954 (EI).
(1′S*,3R*,3′′R*,4′′S*)-3-[Hyd r oxy(4-(m eth oxym eth oxy)-
tetr a h yd r o-2H-th iop yr a n -3-yl)m eth yl]tetr a h yd r oth iop y-
r a n -4-on e (13c). IR ν˜max 3483, 1701 cm-1; 1H NMR (500 MHz,
CDCl3) δ 4.70 (1H, d, J ) 6.5 Hz), 4.67 (1H, d, J ) 6.5 Hz),
4.52 (1H, ddd, J ) 3, 5, 8 Hz, HC-1′ (3J HC-1′/HC-3 ) 3 Hz)), 3.95
(1H, ddd, J ) 3, 6, 7 Hz), 3.40 (3H, s), 3.16 (1H, d, J ) 5 Hz,
OH), 3.09-2.87 (7H, m), 2.80-2.75 (2H, m), 2.43 (1H, dddd, J
) 1, 3.5, 7, 12.5 Hz), 2.32 (1H, ddd, J ) 1, 6.5, 13.5 Hz), 2.20
(1H, dddd, J ) 3, 3.5, 10, 13.5 Hz), 1.94-1.85 (2H, m); 13C
NMR (100 MHz, CDCl3) δ 211.45, 95.65, 74.40, 69.99, 55.95,
55.66, 45.10, 42.02, 30.86, 30.46, 29.86, 27.44, 24.85; HRMS
m/z calcd for C13H22O4S2 306.0960, found 306.0962 (EI).
(1′S*,3S*,3′′R*,4′′S*)-3-[Hyd r oxy(4-(m eth oxym eth oxy)-
tetr a h yd r o-2H-th iop yr a n -3-yl)m eth yl]tetr a h yd r oth iop y-
r a n -4-on e (14c). A 1.5:1 mixture of 14c/13c was obtained.
Spectral data for 14c was deduced from that of the mixture
by comparison with the data for pure 13c. 1H NMR (500 MHz,
CDCl3) δ 4.70 (2H, ap s), 4.07 (1H, ddd, J ) 5, 5, 6.5 Hz, HC-
1′ (3J HC-1′/HC-3 ) 5 Hz)), 3.87 (1H, ddd, J ) 3, 8, 8 Hz), 3.40
(3H, s), 3.33 (1H, d, J ) 6.5 Hz, OH), 3.15-2.88 (5H, m), 2.85-
2.68 (4H, m), 2.55 (1H, dd, J ) 8, 13.5 Hz), 2.51 (1H, ddd, J )
2.5, 9, 12.5 Hz), 2.29 (1H, dddd, J ) 3, 3.5, 8, 13.5 Hz), 2.06
(1H, dddd, J ) 3, 5, 8, 8 Hz), 1.83 (1H, dddd, J ) 3, 8, 9, 13.5
Hz); 13C NMR (125 MHz, CDCl3) δ 212.62, 95.86, 74.92, 74.41,
56.24, 56.06, 45.08, 43.80, 34.10, 31.78, 31.06, 29.98, 25.86.
(t), 24.98 (t), 23.09 (t); HRMS m/z calcd for
C13H22O4S2
306.0960, found 306.0961 (EI).
(1′S*,3R*,3′′R*,4′′R*)-3-[Hyd r oxy(4-(m eth oxym eth oxy)-
tetr a h yd r o-2H-th iop yr a n -3-yl)m eth yl]tetr a h yd r oth iop y-
r a n -4-on e (13b). IR ν˜max 3464, 1697 cm-1; 1H NMR (300 MHz,
CDCl3) δ 4.71 (1H, d, J ) 6.5 Hz), 4.67 (1H, d, J ) 6.5 Hz),
4.23 (1H, m), 4.20 (1H, ddd, J ) 3.5, 5, 9 Hz, HC-1′ (3J HC-1′/HC-3
) 3.5 Hz)), 3.43 (3H, s), 3.21 (1H, d, J ) 5 Hz, OH), 3.11 (1H,
dd, J ) 10.5, 14 Hz), 3.06-2.82 (5H, m), 2.80-2.66 (3H, m),
2.35-2.25 (2H, m), 2.10 (1H, dd, J ) 2, 13.5 Hz), 1.83 (1H,
dddd, J ) 2, 3, 9, 11 Hz), 1.73 (1H, m); 13C NMR (75 MHz,
CDCl3) δ 211.0, 96.3, 71.4, 69.2, 56.3, 54.9, 44.7, 43.6, 31.9,
30.4, 29.9, 24.7, 22.4; HRMS m/z calcd for C13H22O4S2 306.0960,
found 306.0962 (EI).
(1′S*,3S*,3′′R*,4′′R*)-3-[Hyd r oxy(4-(m eth oxym eth oxy)-
tetr a h yd r o-2H-th iop yr a n -3-yl)m eth yl]tetr a h yd r oth iop y-
r a n -4-on e (14b). IR ν˜max 3501, 1701 cm-1; 1H NMR (500 MHz,
CDCl3) δ 4.72-4.69 (2H, m), 4.24 (1H, br s), 3.77 (1H, ddd, J
) 3.5, 8.0, 10 Hz, HC-1′ (3J HC-1′/HC-3 ) 3.5 Hz)), 3.42 (3H, s),
3.29 (1H, d, J ) 10 Hz), 3.16 (1H, dd, J ) 9.5, 13 Hz), 3.04-
2.88 (6H, m), 2.83-2.67 (2H, m), 2.39-2.28 (2H, m), 2.22-
2.10 (2H, m), 1.82-1.73 (1H, m); 13C NMR (125 MHz, CDCl3)
δ 211.41, 96.25, 74.14, 71.60, 56.28, 54.29, 45.12, 45.01, 35.10,
31.95, 31.14, 25.73, 22.66; HRMS m/z calcd for C13H22O4S2
306.0960, found 306.0956 (EI).
Ackn owledgm en t. Financial support from the Natu-
ral Sciences and Engineering Research Council (Canada)
is gratefully acknowledged.
(1′R*,3S*,3′′R*,4′′S*)-3-[Hyd r oxy(4-(m eth oxym eth oxy)-
tetr a h yd r o-2H-th iop yr a n -3-yl)m eth yl]tetr a h yd r oth iop y-
r a n -4-on e (11c). IR ν˜max 3472, 1702 cm-1; 1H NMR (400 MHz,
Su p p or tin g In for m a tion Ava ila ble: Experimental pro-
1
cedures and spectroscopic data for 15-30 and H NMR spectra
CDCl3) δ 4.84 (1H, ddd, J ) 4, 5, 8.5 Hz, HC-1′ (3J HC-1′/HC-3
)
for all reported compounds. This material is available free of
charge via the Internet at http://pubs.acs.org.
8.5 Hz)), 4.73 (1H, d, J ) 6.5 Hz), 4.63 (1H, d, J ) 6.5 Hz),
3.47 (1H, ddd, J ) 3.5, 9, 9, Hz), 3.47 (3H, s), 3.17 (1H, d, J )
5 Hz, OH), 3.12 (2H, ap d, J ) 5 Hz), 2.99-2.94 (2H, m), 2.86
J O016291R