Catalytic enantioselective silylation of acyclic and cyclic triols: Application to total syntheses of cleroindicins D, F, and C

Article abstract of DOI:10.1002/anie.200805338

(Chemical Equation Presented) Pick one out of three: Acyclic and cyclic 1,2,3-triols are silylated with exceptional site- and enantioselectivity by a small-molecule catalyst to afford silyl ethers having a neighboring diol moiety. The new process is appli

Full text of DOI:10.1002/anie.200805338

Angewandte
Chemie
DOI: 10.1002/anie.200805338
Enantioselective Silylation
Catalytic Enantioselective Silylation of Acyclic and Cyclic Triols:
Application to Total Syntheses of Cleroindicins D, F, and C**
Zhen You, Amir H. Hoveyda,* and Marc L. Snapper*
Catalysts that promote site-selective modification of poly-
functional molecules can be of significant utility in selective
chemical synthesis.[1] Of particular importance are chiral
catalysts that initiate enantioselective functionalization of
polyoxygenated molecules[2]—entities commonly found
among biologically active agents. Herein, we present methods
for the enantioselective silylations[3] of acyclic and cyclic 1,2,3-
triols [Eq. (1)]; the transformations are promoted by a readily
available small-molecule catalyst and afford silyl ethers that
Scheme 1. A retrosynthesis for cleroindicin D. Si=trialkylsilyl group.
bear a neighboring diol moiety in up to > 99: < less than 1 e.r.
(> 98% ee). Enantiomerically enriched silyl-protected triols
obtained by the protocols described in this report cannot be
selectively prepared by catalytic or stoichiometric dihydroxy-
lations[4] (including the directed variants).[5] The utility of the
new catalytic processes is demonstrated in the context of
enantioselective total syntheses of cleroindicins D, F, and C,
natural products isolated from Clerodendum indicum, a plant
used in China to battle malaria and rheumatism.[6,7]
On the basis of the previously proposed mechanistic
models,[3a] enantioselective conversion of III into II
(Scheme 1) might involve the association of an amino-acid-
based catalyst (e.g., 1) with either a 1,2-diol or 1,3-diol moiety
through H bonding. Whereas silylations of cyclic as well as
acyclic 1,2-diols have largely proven to be highly selective
(>90:10 e.r.), the related transformations of 1,3-diols can
proceed with inferior selectivity; the two examples shown in
Equations (2) and (3) are illustrative. Thus, a complication
regarding reactions of 1,2,3-triols is whether the 1,2-diol (high
selectivity) or the 1,3-diol moiety (potentially inferior selec-
tivity) more predominantly associate with the catalyst
through H bonding.
Our interest in developing methods for the enantioselec-
tive silylation of triols was inspired in part by a retrosynthesis
of cleroindicin D, shown in Scheme 1. We envisioned that
intramolecular conjugative cyclization of enone I (Scheme 1)
would afford the target hydrobenzofuran. Enantiomerically
enriched I would be prepared through a selective b-elimi-
nation by b-hydroxy enone II, which, in turn, could be
accessed through site- and enantioselective silylation of
tetraol III.
[*] Z. You, Prof. A. H. Hoveyda, Prof. M. L. Snapper
Department of Chemistry, Merkert Chemistry Center
Boston College, Chestnut Hill, MA 02467 (USA)
Fax: (+1)617-552-1442
E-mail: amir.hoveyda@bc.edu
marc.snapper@bc.edu
On the basis of the above considerations, we began our
studies by investigating the reactions of meso acyclic 1,2,3-
triols bearing a central tertiary alcohol. As illustrated in
entries 1–3 of Table 1,[8] substrates containing large alkyl
substituents are silylated with high enantioselectivity (96.5:3.5
to > 99: < 1 e.r.) to afford the desired monosilylated diols 6–8
in 78–85% yield after purification. The reactions of the
corresponding aryl-substituted triols (entries 4–6, Table 1) are
[**] Financial support was provided by the NIH (GM-57212) and Z.Y.
acknowledges support as a LaMattina Graduate Fellow. We are
grateful to Dr. Yu Zhao and Jason Rodrigo (Boston College) for
experimental assistance and helpful discussions. Mass spectrom-
etry facilities at Boston College are supported by the NSF (DBI-
0619576).
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/anie.200805338.
Angew. Chem. Int. Ed. 2009, 48, 547 –550
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
547

Communications
Table 1: Enantioselective silylation reactions of acyclic triols.[a]
Table 2: Enantioselective silylation reactions of cyclic triols.[a]
Entry Product
T [8C] t [h] Yield e.r.[c]
ee
[%][b]
[%][c]
Entry
1
Product
Yield [%][b]
85
e.r.[c]
ee [%][c]
>98
1
2
3
4
ꢀ30
ꢀ30
ꢀ30
ꢀ50
96
96
96
78
85
81
96.5:3.5
97:3
93
94
>99:<1
2
3
60
84
>99:<1
>99:<1
>98
>98
>99:<1 >98
>99:<1 >98
120 70
120 68
5
6
ꢀ50
ꢀ50
97.5:2.5
98:2
95
96
[a] Conditions: 0.5m in diol, 1.5 equiv DIPEA, 1.25 equiv TESCl;>98%
secondary silyl ether product obtained in all cases. [b–c] See Table 1.
TES=triethylsilyl.
120 75
120 62
120 57
7
8
ꢀ50
ꢀ50
94.5:5.5
75:25
89
50
formed with TESCl (versus TBSCl); reactions of this set of
relatively hindered carbinols must be performed with a less
sterically demanding silylating agent to achieve maximum
efficiency.
[a] Conditions: 1.0m in diol, 1.5 equiv DIPEA, 1.5 equiv TBSCl; >98%
primary silyl ether observed in all cases. [b] Yield of isolated product after
purification. [c] Enantiomeric ratios and ee values determined by GLC or
HPLC analysis (see the Supporting Information for details). TBS=tert-
butylsilyl; DIPEA=N,N-diisopropylethylamine; Cy=cyclohexyl.
Two trends regarding enantioselective silylations of acy-
clic and cyclic triols shown in Table 1 and Table 2 are
noteworthy: 1) Reactions of acyclic and cyclic substrates
proceed with the opposite sense of asymmetric induction
(e.g., 12 in Table 1 versus 14 in Table 2). Such findings can be
rationalized by the modes of reaction involving complexes IV
and V as illustrated in Scheme 2. The sense and level of
equally efficient and enantioselective. The silylation process
shown in entry 7 of Table 1, involving a triol that bears a less
sterically demanding allyl substituent, proceeds with equal
efficiency but with a lower enantioselectivity (12 formed in
94.5:5.5 e.r.). Additional diminution of the selectivity in the
silylation depicted in entry 8 of Table 1 (13 obtained in
75:25 e.r.) underlines the influence of the size of the central
carbinol substituent on the degree of the enantiodifferentia-
tion.[9] The reversal of the selectivity observed for diol product
13 (entry 8, Table 1) is likely because of the smaller size of the
methyl substituent versus the hydroxy methylene unit (versus
entries 1–7, Table 1 where alkyl and aryl substituents are
larger than CH2OH; see Scheme 2 for mechanistic models).
The findings shown in Table 1, in view of previous observa-
tions regarding reactions of 1,2-[3] and 1,3-diols [see Eqs. (1)
and (2)], suggest that the silylation of triols proceed
predominantly via complexes established through H bonding
between the chiral catalyst and the two adjacent alcohols of
the substrate.
The less selective reactions of triols which contain a linear
alkyl substituent at the central carbinol site (e.g., 12 in entry 7,
Table 1), did not bode well for the projected plans for total
syntheses of the cleroindicins (Scheme 1). Catalytic enantio-
selective silylations of related cyclic triols, summarized in
Table 2, however, demonstrate that the desired monosilylated
cyclic diols 14–16 are isolated in 60–85% yield and with
> 99: < 1 e.r. (> 98% ee) regardless of the size of the central
alkyl group.[10] Notably, the silylations in Table 2 are per-
Scheme 2. Proposed models for enantioselective silylations in Tables 1
and 2.
enantiodifferentiation in the transformations of acyclic triols
are likely controlled by the complexation of the catalyst and
the substrate in a manner which leads to minimal unfavorable
steric repulsion (large substituent, RL, positioned away from
the amino-acid-based structure). The suggested scenario is
consistent with the trend that higher selectivities require the
presence of a more sizeable alkyl or aryl substituent (e.g.,
compare entries 1–3 to 7–8 in Table 1). In contrast, enantio-
selective silylations of cyclic triols in Table 2 might be largely
governed by the exo mode of substrate–catalyst association,
as depicted in V (Scheme 2). The relative insensitivity of the
enantioselectivity to the size of the central carbinol substitu-
ent in the silylations shown in Table 2, versus the reactions in
548
www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 547 –550

Angewandte
Chemie
Table 1, supports the above proposal. 2) The exceptional
enantiopurity with which the six-membered ring 16 (entry 3,
Table 2) is obtained (> 99: < 1 e.r., > 98% ee) is in contrast to
the inferior level of enantioslectivity with which the corre-
sponding silylation of cyclohexane-1,3-diol proceeds (38% ee
under with TBSCl).[11] Such findings underscore the higher
efficiency with which 1,2-diols associate with the silylation
catalyst as opposed to 1,3-diols. That is, although both modes
of H bonding are illustrated in the proposed model (V in
Scheme 2), it is likely that catalyst–substrate association
involving H bonding with the central carbinol unit is more
critical to the high selectivity.
We next turned our attention to catalytic desymmetriza-
tions of all-secondary 1,2,3-triols.[12] Such transformations,
which deliver products used in the enantioselective synthesis
of natural products,[13] present an additional challenge, since
the three carbinols units now reside in a less differentiable
environment. Enantioselective silylations of all-secondary
triols, summarized in Table 3, uniformly proceed with excep-
Table 3: Enantioselective silylation reactions of all-secondary triols.[a]
Scheme 3. Enantioselective total syntheses of cleroindicins D, F, and
C. a) 1.1 equiv of TBSCl, 1 equiv of imidazole, THF, 08C, 1 h.
b) 1.2 equiv of PhI(OAc)2, CH3CN/H2O (1:1), 08C, 20 min; 69% overall
yield for 2 steps. c) 10.0 equiv of H2O2, 8.0 equiv of K2CO3, 08C, 6 h;
92% yield. d) 1 atm. H2, 4 wt% PtO2, 228C, 12 h. e) 10 mol% ppTs,
MeOH, THF, ꢀ788C, 48 h; 83% yield. f) 20 mol% 1, 2.25 equiv of
TESCl, 2.5 equiv of DIPEA, THF, ꢀ788C, 48 h; 83% yield. g) 2.5 equiv
of MsCl, 3.5 equiv DIPEA, CH2Cl2, 08C!228C; 92% yield. h) 5.0 equiv
HCl, THF/H2O (1:1), 08C!228C, 4 h; 45% yield. i) 1.2 equivalents of
MsCl, 2.2 equiv of DIPEA, CH2Cl2, 08C, 16 h; 92% yield. j) 1 atmos-
phere H2, 50% wt Pd/C, MeOH, 12 h; >98% yield. THF=tetrahydro-
furan; ppTs=pyridinium p-toluenesulfonate; Ms=methanesulfonyl.
Entry Product
Mol% T [8C] t [h] Yield e.r.[c]
of 1
ee
[%][b]
[%][c]
1
2
3
30
ꢀ50 120 72
>99:<1 >98
30
ꢀ30 96
70
51
98:2
96
commercially available para-substituted phenol 21; protec-
tion of the primary alcohol and subsequent conversion into
cyclic dienone 22 through oxidative dearomatization[15] pro-
ceeds in 69% overall yield. Directed epoxidation of the two
electrophilic alkenes in 22 proceeded with exceptional
diastereoselectivity,[16] affording bisepoxide 23 in 92% yield
after silica gel chromatography. Site-selective reduction of the
100
0
24
>99:<1 >98
>99:<1 >98
4
30
ꢀ30 120 65
[17]
ꢀ
two electronically activated C O bonds in 23
and con-
version into the derived dimethylacetal, which proceeds with
concomitant removal of the primary silyl ether, delivered
tetraol 24. Enantioselective silylation of 24 in the presence of
20 mol% 1 and 2.25 equivalents of TESCl led to the
formation of 25 in > 99: < 1 e.r. and 83% yield. Subsequent
conversion into mesylate 26 was performed under standard
conditions, affording the desired product in 92% yield.
Treatment of silyl ether 26 with five equivalents of HCl in
aqueous THF (08C!228C) for four hours led to the
formation of enantiomerically pure cleroindicin D, which
was isolated in 45% overall yield and > 99: < 1 d.r. Con-
version of 26 into the target molecule under the aforemen-
tioned acidic conditions constitutes a five-step sequence
involving removal of the two silyl groups, conversion of the
acetal into the ketone, and b-elimination of the mesylate
group to furnish the requisite enone, which undergoes
intramolecular conjugate addition to afford the furan ring of
the natural product. As also illustrated in Scheme 3, cler-
[a–c] See Table 1; <2% silylation of the central hydroxy group observed
in all cases.
tional selectivity (from 98:2 to > 99: < 1 e.r.), regardless of the
substrate ring size; the outcomes of these transformations are
therefore consistent with the mechanistic proposals outlined
above (Scheme 2). The transformation illustrated in entry 3 of
Table 3 requires 100 mol% of the chiral catalyst and a
relatively elevated temperature (08C) because of the low
solubility of the cyclohexyl triol.[14] Notably, processes shown
in Table 3 proceed with exceptional site-selectivity: less than
2% of the product is derived from the silylation of the central
secondary alcohol is observed.
With the protocols for the enantioselective silylation of
triols in hand, we turned our attention to the total syntheses
of enantiomerically enriched cleroindicins D, F, and C
(Scheme 3). We began by using a two-step sequence involving
Angew. Chem. Int. Ed. 2009, 48, 547 –550
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
549

Communications
hoe, K. Blades, P. R. Moore, M. J. Waring, J. J. G. Winter, M.
Helliwell, N. J. Newcombe, G. Stemp, J. Org. Chem. 2002, 67,
7946 – 7956.
oindicin D can be easily and efficiently converted into
cleroindicins F and C.
We thus demonstrate that the range of substrates that
efficiently undergo catalytic enantioselective silylation
extends beyond the previously reported 1,2-diols.[3] The
processes detailed above deliver otherwise difficult-to-
access polyoxygenated small molecules of exceptional enan-
tiomeric purity, significantly expanding the utility of this
practical class[3] of catalytic reactions. Development of new
and more effective chiral silylation catalysts and additional
methods for enantioselective silylations are subjects of
ongoing studies.
[6] J. Tian, Q.-S. Zhao, H.-J. Zhang, Z.-W. Lin, H.-D. Sun, J. Nat.
Prod. 1997, 60, 766 – 769.
[7] For a total synthesis of rac-cleroindicin D, culminating in a
structural revision of the natural product, see: S. Barradas, M. C.
Carreꢀo, M. Gonzꢁlez-Lꢂpez, A. Latorre, A. Urbano, Org. Lett.
2007, 9, 5019 – 5022.
[8] For details regarding preparation of triol substrates, see the
Supporting Information.
[9] Varying amounts of bis-silylation products are formed in
reactions involving substrates that bear a smaller substituent at
the central carbinol unit. For example, 8%, 32%, and 36% bis-
silyl product is isolated from the transformations shown in
entries 3, 7, and 8 of Table 1, respectively. Such bis-silylations
increase the enantiomeric purity of the chiral monosilyl prod-
ucts.
Received: October 31, 2008
Published online: December 15, 2008
[10] For enantioselective syntheses of a related cyclohexane-based
triols by enzymatic resolution of meso-2-(2-propynyl)cyclohex-
ane-1,2,3-triol, see: T. Matsumoto, T. Konegawa, H. Yamaguchi,
T. Nakamura, T. Sugai, K. Suzuki, Synlett 2001, 1650 – 1652.
[11] It should be noted that 15% of bis-silylated product is also
obtained in the enantioselective silylation of 1,2,3-triol 16; this
byproduct is easily removed by using silica gel chromatography.
Such a process serves as a “correcting” mechanism in the
catalytic process to remove the minor product enantiomer.
[12] For enantioselective syntheses of cyclohexane-1,2,3-triols
through enzymatic kinetic resolution, see: L. Dumortier, J.
Van der Eycken, M. Vandewalle, Tetrahedron Lett. 1989, 30,
3201 – 3204.
[13] For example, see: a) X. Deruytterre, L. Dumortier, J. Van der
Eycken, M. Vandewalle, Synlett 1992, 51 – 52; b) S. K. Vadivel, S.
Vardarajan, R. I. Duclos, Jr., J. T. Wood, J. Guo, A. Makriyannis,
Bioorg. Med. Chem. Lett. 2007, 17, 5959 – 5963.
[14] In the presence of 30 mol% 1 (08C) and after 72 h of reaction
time, silyl ether 19 is isolated in 32% yield and 98.5:1.5 e.r.
(97% ee).
Keywords: enantioselectivity · homogeneous catalysis ·
natural products · silanes · total synthesis
.
[1] For an application of enantioselective catalysis to site- and
enantioselective synthesis, see: C. A. Lewis, S. J. Miller, Angew.
Chem. 2006, 118, 5744 – 5747; Angew. Chem. Int. Ed. 2006, 45,
5616 – 5619.
[2] For copper-catalyzed enantioselective benzoylation reactions of
1,2,3-triols, see: a) B. Jung, M. S. Hong, S. H. Kang, Angew.
Chem. 2007, 119, 2670 – 2672; Angew. Chem. Int. Ed. 2007, 46,
2616 – 2618; b) B. Jung, S. H. Kang, Proc. Natl. Acad. Sci. USA
2007, 104, 1471 – 1475.
[3] For previous reports regarding catalytic enantioselective silyla-
tion reactions, see: a) Y. Zhao, J. Rodrigo, A. H. Hoveyda, M. L.
Snapper, Nature 2006, 443, 67 – 70; b) Y. Zhao, A. W. Mitra,
A. H. Hoveyda, M. L. Snapper, Angew. Chem. 2007, 119, 8623 –
8626; Angew. Chem. Int. Ed. 2007, 46, 8471 – 8474. For an
overview of enantioselective silylation processes, see: S. Ren-
dler, M. Oestreich, Angew. Chem. 2008, 120, 254 – 257; Angew.
Chem. Int. Ed. 2008, 47, 248 – 250.
[4] Catalytic enantioselective dihydroxylations of cyclic allylic
alcohols afford the derived anti diastereomers. See: H. C. Kolb,
M. S. VanNieuwenhze, K. B. Sharpless, Chem. Rev. 1994, 94,
2483 – 2547.
[15] F. Felpin, Tetrahedron Lett. 2007, 48, 409 – 412.
[16] a) A. McKillop, R. J. K. Taylor, R. J. Watson, N. Lewis, Synlett
1992, 1005 – 1006. For a review of hydroxyl-directed epoxida-
tions, see:b) A. H. Hoveyda, D. A. Evans, G. C. Fu, Chem. Rev.
1993, 93, 1307 – 1370.
[17] X. Lei, N. Zaarur, M. Y. Sherman, J. A. Porco, Jr., J. Org. Chem.
2005, 70, 6474—6483.
[5] Hydroxy-directed dihydroxylation reactions of cyclic allylic
alcohols afford meso triols diastereoselectively. See: T. J. Dono-
550
www.angewandte.org
ꢀ 2009 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2009, 48, 547 –550