Bu3SnH-Mediated Pinacol Coupling of 1,5- and 1,6-Dicarbonyl Compounds: Synthetic and Mechanistic Studies

Article abstract of DOI:10.1021/jo9809130

A new method is described for the intramolecular pinacol coupling of 1,5- and 1,6-dicarbonyl compounds, employing Bu₃SnH as the stoichiometric reductant. The key steps in this pinacol cyclization are the addition of a tin ketyl to a carbonyl group and a subsequent intramolecular S_H2 reaction. The isolation of 1,3-dioxa-2-stannolanes, along with other product and labeling studies, provides strong support for the proposed homolytic substitution step, which distinguishes the pinacol cyclization from other reductive cyclizations of tin ketyls, all of which proceed through abstraction of hydrogen from Bu₃SnH in the final step. An interesting consequence of the S_H2 pathway is very high cis selectivity in the cyclization of 1,5-dicarbonyl compounds. Mechanistic studies furnish evidence that the steps that precede homolytic substitution, including C-C bond formation, are reversible under the reaction conditions.

Full text of DOI:10.1021/jo9809130

J . Org. Chem. 1998, 63, 6375-6381
6375
Bu
3
Sn H-Med ia ted P in a col Cou p lin g of 1,5- a n d 1,6-Dica r bon yl
Com p ou n d s: Syn th etic a n d Mech a n istic Stu d ies
David S. Hays and Gregory C. Fu*
Department of Chemistry, Massachusetts Institute of Technology, Cambridge, Massachusetts 02139
Received May 12, 1998
A new method is described for the intramolecular pinacol coupling of 1,5- and 1,6-dicarbonyl
compounds, employing Bu
3
SnH as the stoichiometric reductant. The key steps in this pinacol
cyclization are the addition of a tin ketyl to a carbonyl group and a subsequent intramolecular S
H
2
reaction. The isolation of 1,3-dioxa-2-stannolanes, along with other product and labeling studies,
provides strong support for the proposed homolytic substitution step, which distinguishes the pinacol
cyclization from other reductive cyclizations of tin ketyls, all of which proceed through abstraction
3 H
of hydrogen from Bu SnH in the final step. An interesting consequence of the S 2 pathway is
very high cis selectivity in the cyclization of 1,5-dicarbonyl compounds. Mechanistic studies furnish
evidence that the steps that precede homolytic substitution, including C-C bond formation, are
reversible under the reaction conditions.
In tr od u ction
3
The development of radical reactions of Bu SnH has
been the focus of intense interest in recent years.1 In
some instances, the radical-mediated transformations
have no counterpart in heterolytic chemistry, and in other
instances they complement heterolytic chemistry with
respect to functional group tolerance, diastereoselectivity,
and the like.
In 1986, Beckwith established that Bu
3
SnH can effect
the reductive cyclization of a δ,ꢀ-unsaturated aldehyde
F igu r e 1. Bu
enones (Beckwith, Enholm).
3
SnH-mediated reductive cyclization of enals and
to form a cyclopentanol (Figure 1).2
,3
Enholm later
explored this reaction systematically, developing it into
a synthetically useful process for both aldehydes and
ketones.4 We were interested in investigating an analo-
gous reductive cyclization, using a dicarbonyl compound
5
as the substrate (Figure 2). Although a wide array of
metals and inorganic metal complexes effect intramo-
6
lecular pinacol couplings, to the best of our knowledge
there have been no reports that metal hydrides can do
so.
3
In comparing the pathway for Bu SnH-mediated
pinacol coupling that is outlined in Figure 2 with the
pathway followed in the Beckwith-Enholm cyclization
F igu r e 2. Proposed Bu
of dicarbonyl compounds (intramolecular pinacol coupling).
3
SnH-mediated reductive cyclization
(
1) (a) Neumann, W. P. Synthesis 1987, 665-683. (b) RajanBabu,
T. V. In Encyclopedia of Reagents for Organic Synthesis; Paquette, L.
A., Ed.; Wiley: New York, 1995. (c) Davies, A. G. Organotin Chemistry;
VCH: New York, 1997. (d) Pereyre, M.; Quintard, J .-P.; Rahm, A. Tin
in Organic Synthesis; Butterworths: Boston, 1987.
(Figure 1), from the start it is possible to identify
(
2) Beckwith, A. L. J .; Roberts, D. H. J . Am. Chem. Soc. 1986, 108,
893-5901.
3) Sugawara, T.; Otter, B. A.; Ueda, T. Tetrahedron Lett. 1988, 29,
potential pitfalls for the pinacol-coupling reaction. First,
whereas the C-C bond-forming step in the Beckwith-
Enholm cyclization produces a carbon-centered radical,
5
7
(
5-78.
(
4) (a) Enholm, E. J .; Prasad, G. Tetrahedron Lett. 1989, 30, 4939-
a higher-energy oxygen-centered radical would be gener-
4
1
942. (b) Enholm, E. J .; Burroff, J . A. Tetrahedron Lett. 1992, 33,
835-1838.
7-9
ated in a pinacol coupling (Figure 2, B f C);
rather
3
(5) For Bu SnH-mediated reductive cyclizations of carbonyl/oxime
ethers, see: (a) Naito, T.; Tajiri, K.; Harimoto, T.; Ninomiya, I.; Kiguchi,
T. Tetrahedron Lett. 1994, 35, 2205-2206. (b) Kiguchi, T.; Tajiri, K.;
Ninomiya, I.; Naito, T.; Hiramatsu, H. Tetrahedron Lett. 1995, 36,
(7) We are not aware of any precedent for the addition of a tin ketyl
to a carbonyl group.
(8) For discussions concerning the intramolecular addition of a
nonstabilized carbon-centered radical to a carbonyl group, see: (a)
Tsang, R.; Fraser-Reid, B. J . Am. Chem. Soc. 1986, 108, 2116-2117.
Walton, R.; Fraser-Reid, B. J . Am. Chem. Soc. 1991, 113, 5791-5799.
(b) Beckwith, A. L. J .; Hay, B. P. J . Am. Chem. Soc. 1989, 111, 230-
234. Beckwith, A. L. J .; Hay, B. P. J . Am. Chem. Soc. 1989, 111, 2674-
2681. (c) Curran, D. P. Synthesis 1988, 417-439.
2
53-256. (c) Naito, T.; Torieda, M.; Tajiri, K.; Ninomiya, I.; Kiguchi,
T. Chem. Pharm. Bull. 1996, 44, 624-626. (d) Tormo, J .; Hays, D. S.;
Fu, G. C. J . Org. Chem. 1998, 63, 201-202.
(6) For a comprehensive review of pinacol couplings, see: Robertson,
G. M. In Comprehensive Organic Synthesis; Trost, B. M., Ed.; Perga-
mon: New York, 1991; Vol. 3, Chapter 2.6.
S0022-3263(98)00913-X CCC: $15.00 © 1998 American Chemical Society
Published on Web 08/12/1998

6
376 J . Org. Chem., Vol. 63, No. 18, 1998
Hays and Fu
Sch em e 1
Ta ble 1. Bu 3Sn H-Med ia ted In tr a m olecu la r P in a col
Cou p lin g
than form C, radical B may simply abstract a hydrogen
from Bu SnH (Scheme 1; B f E). Second, even if C is
accessible, the propensity of alkoxy radicals to fragment
e.g., C f B) could lead to predominant formation of
3
(
acyclic reduction product (e.g., E), rather than desired
pinacol product (D).10
Despite these concerns, we forged ahead with our
attempt to develop a Bu SnH-mediated pinacol coupling
3
reaction. In this paper, we describe an investigation that
has resulted in the discovery of such a process, via an
unanticipated mechanistic pathway.11
Resu lts a n d Discu ssion
a
Average of two runs. b Based on analysis by capillary gas
We were pleased to observe that treatment of a range
c
of 1,5- and 1,6-dicarbonyl substrates with Bu
3
SnH in
chromatography, except as noted. Isolated by crystallization as
d
the 1,3-dioxa-2-stannolane. ∼1:1 mixture of the two diastereo-
refluxing benzene (AIBN as the initiator), followed by a
hydrolytic workup, furnishes pinacol coupling products
in modest to high yield (Table 1). Both dialdehydes
meric cis diols. e Based on analysis by 1H NMR. f Isolated as the
acetonide.
(entries 1-4) and keto aldehydes (entries 5 and 6)
undergo cyclization. In the case of glutaraldehyde, the
reaction must be run dilute (0.025 M) to afford even a
modest yield of cyclopentane-1,2-diol (46%; entry 1); at
higher concentrations, 1,2-reduction of the aldehyde
predominates. Our preliminary attempts to synthesize
larger rings have been unsuccessful.
3
With Bu SnH as the reductant, we have established
that dialdehydes cyclize preferentially in the presence of
the corresponding keto aldehydes and diketones (Figure
3
). The high selectivity is likely a manifestation of
reversible addition of the Bu Sn radical to the carbonyl
group (A a B, Figure 2). Were addition of the Bu Sn
3
3
radical irreversible for both 1 and 2 (Figure 3), then 1
would be predicted to undergo pinacol cyclization perhaps
twice as fast as 2, rather than more than 10 times as
fast.
F igu r e 3. Selective intramolecular pinacol coupling of dial-
dehydes by Bu SnH.
3
3
Inspection of Table 1 reveals that the Bu SnH-medi-
ated intramolecular coupling of 1,5-dicarbonyl compounds
furnishes cis diols with uniformly high diastereoselec-
tivity (>20:1; Table 1, entries 1, 2, and 5), an observation
that stands in sharp contrast to the stereoselection
reported for the intramolecular addition of tin ketyls to
other unsaturated groups (e.g., olefins and oxime
ethers).2-5 We uncovered a clue to the origin of the cis
selectivity when we isolated the prehydrolysis product
of the pinacol cyclization of adipaldehyde, which we
established to be 1,3-dioxa-2-stannolane F (Scheme 2;
12
6
0% yield), not tin alkoxide D.
To determine if tin alkoxide D might be an intermedi-
ate on the pathway to F (Scheme 2), we prepared D
independently and subjected it to the reaction conditions.
(
9) For a recent study of the relative reactivity of oxygen-stabilized
versus nonstabilized carbon radicals, see: J ohnson, C. C.; Horner, J .
H.; Tronche, C.; Newcomb, M. J . Am. Chem. Soc. 1995, 117, 1684-
1
687.
(12) For reviews of the functionalization of 1,3-dioxa-2-stannolane,
see: (a) Pereyre, M.; Quintard, J .-P.; Rahm, A. Tin in Organic
Synthesis; Butterworths: Boston, 1987. (b) Blunden, S. J .; Cusack, P.
A.; Smith, P. J . J . Organomet. Chem. 1987, 325, 141-152. (c) David,
S.; Hanessian, S. Tetrahedron 1985, 41, 643-663.
(
10) Yeung, B.-W. A.; Alonso, R.; Vite, G. D.; Fraser-Reid, B. J .
Carbohydr. Chem. 1989, 8, 413-427.
11) For a preliminary report, see: Hays, D. S.; Fu, G. C. J . Am.
Chem. Soc. 1995, 117, 7283-7284.
(

Bu
3
SnH-Mediated Pinacol Coupling
J . Org. Chem., Vol. 63, No. 18, 1998 6377
3
F igu r e 4. Revised mechanism for the Bu SnH-mediated
intramolecular pinacol coupling of dicarbonyl compounds.
Sch em e 2
F igu r e 5. Rationalization of the high cis stereoselectivity that
3
is observed in the Bu SnH-mediated intramolecular pinacol
coupling of 1,5-dicarbonyl compounds.
homolytic substitution (Figure 4, C f F ). For this study,
we employed Oct SnD, rather than Bu SnH, as the
3 3
stoichiometric reductant; this obviates complications
associated with the volatility of butane and allows us to
unambiguously establish the fate of the hydrogen (deu-
3
terium) initially bound to tin. As illustrated in eq 1, Oct -
SnD-mediated pinacol cyclization of adipaldehyde affords
an equal mixture (within experimental error) of pinacol
product and Oct-D, a result consistent with a reaction
pathway proceeding essentially exclusively according to
Figure 4.
Based on our observation that F is not formed, we have
concluded that the Bu SnH-mediated intramolecular
3
pinacol coupling of dicarbonyls likely proceeds through
the pathway outlined in Figure 4.
The first two steps of the revised mechanism are
identical to those in the original pathway (Figure 2).
However, when alkoxy radical C is produced, rather than
abstracting hydrogen from Bu
reaction, we now believe that the alkoxy radical effects
an intramolecular homolytic substitution (S 2) at tin,
thereby furnishing 1,3-dioxa-2-stannolane F (Figure 4;
3
SnH in an intermolecular
The S
H
2 reaction is a favorable process from the
standpoints of both entropy and enthalpy (Sn-O bond:
1
c
∼90 kcal/mol; Sn-C bond: ∼65 kcal/mol ). In contrast
to the original mechanism, the revised pathway accounts
for the observation that reductive cyclizations of 1,5-
dicarbonyl substrates afford the pinacol products with
high cis diasteroselectivity (Figure 5). Thus, the ketyl
radical of B can add to the pendant carbonyl group to
generate either cis-C or trans-C. cis-C can readily
undergo an intramolecular S
3.3.0] ring system (cis-F ), whereas the analogous reac-
tion of trans-C is much less facile, since it would furnish
a strained trans [3.3.0] ring system (trans-F ). Conse-
quently, trans-C instead fragments to regenerate B,
which can in turn provide cis-C and then cis-F .
H
C f F ).1
propagates the chain through abstraction of hydrogen
from Bu SnH.
To provide further support for the S
3-15
The butyl radical that is liberated then
3
H
2 mechanism, as
well as to determine if the original hydrogen-abstraction
pathway (Figure 2, C f D) is intervening to any
appreciable extent, we quantified the formation of the
chain-carrying alkyl radical that is generated upon
H
2 reaction to produce a cis
[
(13) A related process involving a carbon-centered radical has been
observed as a minor (e10%) side reaction in the stannylformylation
of 1,6-dienes: Ryu, I.; Kurihara, A.; Muraoka, H.; Tsunoi, S.; Kambe,
N.; Sonoda, N. J . Org. Chem. 1994, 59, 7570-7571 and references
therein.
The competition experiment that was presented earlier
(Figure 3) furnishes evidence that the first step of the
(14) (a) Davies has shown that the presence of electronegative
substituents on tin facilitates homolytic displacement of an alkyl group
by an alkoxy radical; for an overview, see: Chemistry of Tin; Harrison,
P. G., Ed.; Chapman and Hall: New York, 1989. (b) For a recent review
of free-radical homolytic substitution, see: Schiesser, C. H.; Wild, L.
M. Tetrahedron 1996, 52, 13265-13314.
Bu
3
SnH-mediated intramolecular pinacol coupling pro-
Sn radical to a carbonyl group, is
cess, addition of a Bu
3
reversible under the reaction conditions (Figure 4, A a
B). To investigate the reversibility of the next step of
the reaction sequence, C-C bond formation (Figure 4, B
a C), we synthesized both diastereomers of 1,2-cyclo-
hexanediol derivative 4 (Figure 6). In the presence of
(
15) For examples of intramolecular reactions of oxygen-centered
radicals with tributyltin alkoxides that lead to Bu Sn transfer (rather
than cyclization), see the following. (a) 1,5-Bu Sn transfer: Davies, A.
G.; Tse, M.-W. J . Organomet. Chem. 1978, 155, 25-30. Kim, S.; Lee,
S.; Koh, J . S. J . Am. Chem. Soc. 1991, 113, 5106-5107. (b) 1,6-Bu Sn
transfer: Kim, S.; Do, J . Y.; Lim, K. M. Chem. Lett. 1996, 669-670.
3
3
3
Bu Sn radicals, nitrate esters undergo addition-frag-
3

6
378 J . Org. Chem., Vol. 63, No. 18, 1998
Hays and Fu
F igu r e 7. Common mechanisms for pinacol coupling.
compounds, employing Bu SnH as the stoichiometric
3
reductant. To the best of our knowledge, this is the first
example of the use of a metal hydride to effect an
intramolecular pinacol reaction. Whereas most methods
for pinacol coupling employ 2 equiv of a one-electron
reducing agent (Figure 7), this mechanistically distinct
3
Bu SnH-mediated process requires just one equivalent
of reducing agent (Figure 4).
The key steps in this pinacol cyclization are an
unprecedented addition of a tin ketyl to a carbonyl group
(Figure 4, B f C) and a subsequent intramolecular S 2
H
F igu r e 6. Stereochemical evidence for reversible C-C bond
formation under the intramolecular pinacol coupling condi-
tions.
mentation to produce alkoxy radicals (e.g., C).16 Thus,
cis-4 and trans-4 provide access to cis-C and trans-C,
respectively. If the reactions of cis-4 and trans-4 with
reaction (C f F ). The isolation of 1,3-dioxa-2-stanno-
lanes (F ), along with other product and labeling studies,
provides strong support for the proposed homolytic
substitution step, which distinguishes the pinacol cy-
clization from other reductive cyclizations of tin ketyls,
all of which proceed through abstraction of hydrogen from
3
Bu Sn radical each affords diastereomerically pure 1,3-
dioxa-2-stannolane F , then C-C bond formation in the
pinacol coupling proceeds irreversibly; on the other hand,
if isomeric mixtures of F are produced, then C-C bond
formation is reversible. As illustrated in Figure 6, the
data strongly support reversible C-C bond formation
under the pinacol-coupling conditions.
2
-5
Bu
3
SnH in the final step (cf. Figure 2, C f D).
An
interesting consequence of the S
H
2 pathway is very high
cis selectivity in the cyclization of 1,5-dicarbonyl com-
pounds. Mechanistic studies furnish persuasive evidence
that the steps that precede homolytic substitution,
including C-C bond formation, are reversible under the
reaction conditions.
Other group 14 hydrides also effect intramolecular
pinacol coupling of dicarbonyl compounds, although they
3
are less efficient than Bu SnH (eq 2; yields according to
GC).1
7,18
Simple 1,2-reduction is the predominant side
reaction.
Exp er im en ta l Section
Gen er a l Meth od s. AIBN (Eastman) was used without
purification. Bu SnH (Aldrich) was distilled prior to use.
3
Solvents were distilled from the indicated drying agents:
benzene (sodium/benzophenone); toluene (molten sodium);
CH
2
19
Cl
2
(calcium hydride).
Sn chemical shifts are reported in ppm downfield from
SnMe (neat, external reference, δ scale) and were determined
1
4
1
with pulse intervals of 0.3 s; broad-band H NMR decoupling
was only applied during acquisition. Coupling constants (J )
are given in Hz.
Su m m a r y a n d Con clu sion s
All reactions were carried out under an atmosphere of
nitrogen or argon in oven-dried glassware with magnetic
stirring, unless otherwise indicated.
We have discovered an interesting new method for the
intramolecular pinacol coupling of 1,5- and 1,6-dicarbonyl
Note: The data reported below for intramolecular pinacol-
coupling reactions may differ slightly from the data reported
in the text, since the latter are the average of two runs.
Ta ble 1, En tr y 1. Commercially available glutaric dial-
dehyde (50 wt % solution; Aldrich; 5-6 mL) was dissolved in
(
16) Vite, G. D.; Fraser-Reid, B. Synth. Commun. 1988, 18, 1339-
342.
17) (a) For a report of a triplet ketone reacting with (Me
an S 2 pathway, see: Alberti, A.; Dellonte, S.; Paradisi, C.; Roffia, S.;
1
(
3 4
Si) Si via
H
Pedulli, G. F. J . Am. Chem. Soc. 1990, 112, 1123-1129. (b) For related
work, see: Kulicke, K. J .; Chatgilialoglu, C.; Kopping, B.; Giese, B.
Helv. Chim. Acta 1992, 75, 935-939. Miura, K.; Oshima, K.; Utimoto,
K. Chem. Lett. 1992, 2477-2478.
4
benzene (75 mL) and dried with MgSO . After filtration and
concentration, the resulting viscous, colorless oil was distilled
bulb-to-bulb to afford the anhydrous dialdehyde as a clear,
colorless oil that remained predominantly monomeric for at
least 48 h.
(
18) In the reaction of (Me
one molecule of the pinacol coupling product is generated for each
molecule of (Me Si) SiH that is consumed (see Experimental Section).
We believe that this is attributable to the preference of the Me Si
radical, generated upon S 2 cyclization, for effecting pinacol coupling
of adipaldehyde rather than hydrogen atom abstraction from (Me
Si) SiH.
3 3
Si) SiH with adipaldehyde, more than
3
3
3
A solution of the dialdehyde (200 mg, 2.00 mmol), Bu
699 mg, 2.40 mmol), and AIBN (33 mg, 0.20 mmol) in benzene
(80 mL) was heated to reflux. Additional AIBN (33 mg, 0.20
3
SnH
H
(
3
-
3

Bu
3
SnH-Mediated Pinacol Coupling
J . Org. Chem., Vol. 63, No. 18, 1998 6379
mmol) was added every 3 h. After 12 h, the reaction mixture
was cooled to room temperature, and an aliquot was removed
for GC analysis (acetylated and analyzed as the bisacetate
derivative; 98:1 cis/trans; comparison with authentic cis and
trans diols from Aldrich). The reaction mixture was concen-
trated, and the resulting white solid was washed with ice-cold
toluene (3 × 5 mL) and dried, affording 313 mg (47%) of the
comparison with authentic cis product prepared by dihydroxy-
lation of 4,4-bis[[(tert-butyldimethylsilyl)oxy]methyl]-1-meth-
ylcyclopentene). The reaction mixture was concentrated to a
pale yellow oil and purified by flash chromatography, which
1
afforded 64 mg (63%) of the cis diol as a colorless oil: H NMR
(300 MHz, CDCl
3.28 (s, 2H), 2.68 (d, 1H, J ) 10.8), 1.88 (dd, 1H, J
) 7.4), 1.72 (d, 1H, J ) 14.4), 1.65 (d, 1H, J ) 14.1), 1.59 (dd,
1H, J ) 13.4, J ) 8.9), 1.22 (s, 3H), 0.92 (s, 9H), 0.87 (s, 9H),
0.10 (s, 6H), 0.02 (s, 6H); C NMR (75 MHz, CDCl
77.7, 69.0, 68.5, 45.2, 43.4, 37.2, 25.9, 25.7, 23.4, 18.4, 18.1,
3
) δ 4.05 (s, 1H), 3.64 (m, 1H), 3.42 (s, 2H),
1
) 13.7, J
2
1
,3-dioxa-2-stannolane as a white solid. The 1H NMR spec-
1
9
trum was identical to the published spectrum.
1
2
1
3
Ta ble 1, En tr y 2. A solution of the dialdehyde (218 mg,
.890 mmol), Bu SnH (311 mg, 1.07 mmol), and AIBN (15 mg,
.089 mmol) in benzene (36 mL) was heated to reflux.
3
) δ 78.1,
Si [M
0
0
3
-5.61, -5.64; IR (neat) 3412; HRMS calcd for C20
H
45
O
4
2
+
Additional AIBN (15 mg, 0.089 mmol) was added at 3 h
intervals. After 12 h, the reaction mixture was cooled to room
temperature, concentrated, and purified by flash chromatog-
raphy, which afforded 147 mg (67%) of a mixture of diaster-
+ H] 405.2856, found 405.2845.
Ta ble 1, En tr y 6. A solution of the keto aldehyde (176 mg,
1.00 mmol), Bu SnH (437 mg, 1.50 mmol), and AIBN (33 mg,
3
0.20 mmol) in benzene (100 mL) was heated to reflux in an oil
bath. Additional AIBN (33 mg, 0.20 mmol) was added at 3 h
intervals. After 9 h, the reaction mixture was cooled to room
temperature, concentrated to a yellow oil, and passed through
a pad of silica gel. Treatment of the crude diol with 2,2-
dimethoxypropane (1.0 g, 10 mmol) and catalytic PPTS,
eomeric cis diols as a clear, colorless oil. Less polar diastere-
1
omer: H NMR (300 MHz, CDCl
3
) δ 3.88 (br t, 2H), 3.54 (d,
2
2
6
4
H, J ) 3.0), 3.25 or 2.80 (br s, 2H), 2.24 (m, 1H), 2.03 (m,
H), 1.49 (dt, 2H, J ) 14.3, J ) 5.5), 0.93 (s, 9H), 0.09 (s,
H). More polar diastereomer: H NMR (300 MHz, CDCl
.07 (m, 2H), 3.43 (d, 2H, J ) 5.1) 3.25 or 2.80 (br s, 2H), 2.42
1
2
1
3
) δ
followed by aqueous workup and flash chromatography, af-
1
(
m, 1H), 1.77 (m, 2H), 1.64 (m, 2H), 0.09 (s, 9H), 0.01 (s, 6H);
forded 116 mg (53%) of the acetonide as a colorless oil:
H
1
3
C NMR (75 MHz, CDCl
3
) δ 74.2, 74.0, 66.6, 66.4, 36.8, 35.8,
NMR (300 MHz, CDCl ) δ 7.50 (dd, 1H, J 1 ) 7.4, J ) 1.4),
3
2
3
4.1, 33.4, 25.8, 18.3, 18.2, -5.4, -5.5; IR (neat) 3383; HRMS
7.24-7.13 (m, 2H), 7.06 (d, 1H, J ) 7.1), 4.17 (dd, 1H, J )
1
+
calcd for C12
H
27
O
3
Si [M + H] 247.1729, found: 247.1727.
4.2, J 2 ) 2.1), 3.06 (ddd, 1H, J ) 17.1, J 2 ) 12.7, J ) 5.4),
1
3
GC analysis of an acetylated aliquot of the unpurified
reaction mixture revealed a ratio of 1.0:1 for the diastereomeric
cis diols and a ratio of >99:1 for cis/trans diols (comparison
with authentic products prepared by dihydroxylation of 4-[(tert-
butyldimethylsilyl)oxymethyl]cyclopentene).
2.62 (ddd, 1H, J 1 ) 16.5, J ) 5.4, J 3 ) 2.4), 2.31-2.22 (m,
2
1H), 2.01-1.89 (m, 1H), 1.58 (s, 3H), 1.44 (s, 3H), 0.98 (s, 3H);
1
3
C NMR (75 MHz, CDCl ) δ 140.2, 134.9, 127.8, 127.7, 126.8,
3
126.4, 107.9, 79.04, 79.00, 27.5, 27.2, 27.1, 24.1, 23.6; IR (neat)
2982, 2931, 2870; HRMS calcd for C H O [M + H] 219.1385,
found 219.1378.
+
1
4
19
2
Ta ble 1, En tr y 3. A solution of the dialdehyde (114 mg,
1
0
3
.00 mmol), Bu SnH (349 mg, 1.20 mmol), and AIBN (16 mg,
.10 mmol) in benzene (10 mL) was heated to reflux for 1 h.
An aliquot of the unpurified reaction mixture was passed
through a plug of silica gel and analyzed by GC, which revealed
a 20:1 (cis/trans) ratio of diastereomers (comparison with
authentic cis product prepared by dihydroxylation of 4-methyl-
Upon cooling to room temperature, the 1,3-dioxa-2-stannolane
precipitated as a white solid. CH Cl was added until the
2
2
reaction mixture was homogeneous, and an aliquot was then
removed for GC analysis. The reaction mixture was concen-
trated to a white solid/colorless oil and purified by flash
chromatography, which provided 97 mg (84%) of a mixture of
cis and trans diols as a white solid. GC analysis of the
acetylated aliquot revealed a 1:2.4 (cis/trans) mixture of diols
1
,2-dihydronaphthalene).
Com p etition Exp er im en t (F igu r e 3). A solution con-
taining the dialdehyde (114 mg, 1.00 mmol), keto aldehyde
(128 mg, 1.00 mmol), diketone (142 mg, 1.00 mmol), Bu
3
SnH
(349 mg, 1.20 mmol), and tetradecane (65 µL; internal stan-
dard) in benzene (10 mL) was prepared and analyzed by GC.
AIBN (16 mg, 0.10 mmol) was added, and then the solution
was heated to reflux for 90 min. Upon cooling to room
temperature, a white precipitate formed. The reaction mixture
(comparison with authentic cis and trans diols from Aldrich).
Ta ble 1, En tr y 4. A solution of the dialdehyde (172 mg,
0
3
.486 mmol), Bu SnH (170 mg, 0.583 mmol), and AIBN (16
mg, 0.097 mmol) in benzene (4.9 mL) was heated to reflux.
After 3 h, AIBN (8 mg, 0.05 mmol) was added, and the reaction
was refluxed for an additional 1 h. The reaction mixture was
concentrated to a cloudy, colorless oil and purified by flash
chromatography, which provided 152 mg (88%) of diols as a
2 2
was made homogeneous by the addition of CH Cl , and it was
then analyzed by GC. The amount of each dicarbonyl com-
pound that remained, as a percentage of its original amount,
follows: dialdehyde (9%); keto aldehyde (87%); diketone
(
100%).
Isola tion of a 1,3-Dioxa -2-sta n n ola n e (Sch em e 2). A
solution of adipaldehyde (114 mg, 1.00 mmol), Bu SnH (349
1
colorless oil that solidified on standing: H NMR (300 MHz,
CDCl
3
) δ 7.36-7.27 (m, 10H), 4.45-4.39 (m, 4H), 3.90-3.31
3
(
m, 6H), 3.15 (br s, OH), 2.85 (br s, OH), 2.60 (br s, OH), 2.34-
mg, 1.20 mmol), and AIBN (16 mg, 0.10 mmol) in benzene (10
mL) was heated to reflux for 1 h. Stirring was then stopped,
and the reaction mixture was permitted to slowly cool to room
temperature, thereby allowing the 1,3-dioxa-2-stannolane to
crystallize as white needles. After 34 h, the solvent was
1
3
1
1
1
6
3
.23 (m, 6H); C NMR (75 MHz, CDCl ) δ 138.5, 138.3, 138.2,
37.4, 128.5, 128.4, 128.3, 128.0, 127.9, 127.7, 127.65, 127.60,
27.5, 75.1, 73.4, 73.1, 73.08, 73.02, 72.2, 71.9, 71.8, 70.7, 69.6,
9.2, 68.5, 37.9, 35.1, 34.4, 33.4, 31.9, 30.6; IR (neat) 3385;
+
HRMS calcd for C22
H
29
O
4
[M + H] 357.2066, found 357.2068.
removed, and the crystals were washed with cold benzene
GC analysis of an acetylated aliquot of the unpurified
reaction mixture indicated that a 20:17:63 (cis/cis/trans)
mixture of diols was obtained (comparison with authentic
products prepared by dihydroxylation of meso-4,5-bis(benzyl-
oxymethyl)cyclohexene and acid-induced ring opening of meso-
1
(three times) and dried under vacuum (199 mg, 60%; H NMR
2
0
spectrum identical to the published spectrum ).
Exa m in a tion of th e Ch em ica l Com p eten ce of D f F
(
3
Sch em e 2). Bu SnOEt (3.35 g, 10.0 mmol) was added to a
suspension of trans-1,2-cyclohexanediol (1.16 g, 10.0 mmol) in
benzene (10 mL). The reaction was stirred under vacuum until
no ethoxy peaks remained in the 1H NMR spectrum. The
spectrum revealed that a mixture of tin alkoxides was present.
4
,5-bis(benzyloxymethyl)cyclohexene oxide).
Ta ble 1, En tr y 5. A solution of Bu SnH (218 mg, 0.750
3
mmol) and AIBN (16 mg, 0.10 mmol) in benzene (1 mL) was
added via syringe pump over 36 h to a refluxing solution of
the keto aldehyde (101 mg, 0.250 mmol) in benzene (0.50 mL).
The reaction mixture was cooled to room temperature, and an
aliquot was passed through a plug of silica gel and analyzed
AIBN (1.0 mg, 0.0080 mmol) and Bu
mmol) were added to a solution of the tin alkoxide mixture
6 6
32.4 mg, 0.0800 mmol) in C D (800 µL) in a J . Young tube.
3
SnH (27.9 mg, 0.096
(
1
After 2.5 h in an oil bath maintained at 80 °C, the reaction
mixture was examined by H NMR, which indicated that no
by H NMR (no trans diol was detectable w >20:1 cis/trans;
1
(
19) Gmelin Handbook of Inorganic Chemistry, Organotin Com-
pounds, 8th ed.; Kruerke, U., Ed.; Springer-Verlag: Berlin, 1988; Vol.
5, p 52.
(20) (a) Reference 19. (b) See also: Grindley, T. B.; Thangarasa, R.
J . Am. Chem. Soc. 1990, 112, 1364-1373.
1

6
380 J . Org. Chem., Vol. 63, No. 18, 1998
,3-dioxa-2-stannolane had formed. PhSiH (9 mg, 0.08 mmol)
Hays and Fu
1
3
tions.16 Unfortunately, these reactions are slow, and they do
2
3
was then added to the reaction mixture; after 105 min at room
temperature, no Bu SnH was observed (PhSiH reacts rapidly
at room temperature with R Sn(OR) to form R SnH ), verify-
ing that no 1,3-dioxa-2-stannolane was present.
Syn th esis of Oct Sn D. A solution of Oct SnCl (948 mg,
.92 mmol) in Et O (4 mL) at 0 °C was added to a solution of
LiAlD (80.6 mg, 1.92 mmol; Aldrich, 98% D) in Et O (2 mL).
After the mixture was stirred at room temperature for 90 min,
the Oct SnD was isolated under an atmosphere of nitrogen
not always proceed cleanly. In the following stereochemical
tests for reversible C-C bond formation, we regard the
observation that both cis and trans diol acetates are formed
(following acetylation), as well as cis- and trans-1,3-dioxa-2-
stannolanes, to be strong evidence for reversibility. However,
due to side reactions of the starting material under the
reaction conditions, we believe that the ratio of diol acetates
that we measure may be biased toward the starting isomer.
2
2
3
2
2
2
2
3
3
1
2
4
2
3
Rea ction of cis-4 (F igu r e 6). Bu SnH (29.1 mg, 0.100
3
mmol) and methyl butyl ether (internal 1H NMR standard)
were added to a solution of cis-4 (37.4 mg, 0.0830 mmol) and
AIBN (1.3 mg, 0.0080 mmol) in C D (800 µL) in a J . Young
by filtering, concentrating, passing through Celite (benzene
washes), concentrating, filtering through an Acrodisc, and then
concentrating. This procedure furnished Oct
cloudy, colorless oil. No Oct SnH was detectable by H NMR:
) δ 1.68 (m, 6H), 1.43-1.52 (m, 30H),
3
SnD as a slightly
6
6
1
3
tube. The reaction was heated for 4 h in an oil bath
maintained at 80 °C, at which time 73% of cis-4 remained, as
determined by integration versus the internal standard. GC
analysis of an acetylated aliquot revealed a 1.3:1 (trans:cis)
mixture of bisacetates of trans- and cis-1,2-cyclohexanediol.
The NMR tube was immersed in an ice bath, resulting in the
precipitation of a white solid, which was washed with benzene
(1 × 1 mL) and dried. The white solid was then dissolved in
1
6 6
H NMR (300 MHz, C D
1
3
1
3
.04 (m, 6H), 0.94-0.90 (m, 9H); C NMR (75 MHz, C
6
D
6
) δ
Sn NMR (111.9
) δ -90.5 (t, J ) 243); H NMR (46 MHz, C ) δ
.14 (s, J ( Sn/ H) ) 243, J ( Sn/ H) ) 232); IR (neat) 2918,
850, 1465, 1378, 1340, 1300.
P in a col Cycliza tion of Ad ip a ld eh yd e w ith Oct
Eq 1). Oct SnD (613 mg, 1.33 mmol), nonane (142 mg, 1.11
1
19
5.0, 32.8, 30.2, 30.1, 28.6, 23.5, 14.7, 9.1.
2
MHz, C
5
2
6
1
D
6
119
6 6
H
2
1
117
2
3
Sn D
1
119
(
3
CDCl and examined by H and
Sn NMR, which revealed
3
mmol; internal GC standard), and tetradecane (220 mg, 1.11
mmol; second internal GC standard) were added to a solution
of adipaldehyde (127 mg, 1.11 mmol) and AIBN (18 mg, 0.11
mmol) in benzene (11 mL). The resulting reaction mixture
was heated to reflux for 75 min. An aliquot was then analyzed
by GC, which revealed an 82% yield of octane. Analysis by
GCMS indicated that the octane was 97% deuterated. GC
analysis of an acetylated aliquot revealed an 80% yield of
acetylated pinacol products (2.3:1 trans/cis).
that both cis- and trans-1,3-dioxa-2-stannolanes were present.
Rea ction of tr a n s-4 (F igu r e 6). Bu SnH (349 mg, 1.20
3
mmol) was added to a solution of trans-4 (450 mg, 1.00 mmol)
and AIBN (16 mg, 0.10 mmol) in benzene (10 mL). The
reaction mixture was heated to reflux for 5.5 h. Then, more
AIBN (16 mg, 0.10 mmol) was added, and the reaction was
refluxed for 7 h. GC analysis of an acetylated aliquot revealed
a 3.9:1 trans/cis mixture of bisacetates of 1,2-cyclohexanediol.
The bulk reaction mixture was concentrated to a white solid/
colorless oil. The solid was washed with pentane (3 × 1 mL),
2
1
tr a n s-1,2-Cycloh exa n ed iol Mon on itr a te. A solution of
fuming HNO (2.52 g, 40.0 mmol) in Ac O (5 mL) was added
to a suspension of trans-1,2-cyclohexanediol (4.64 g, 40.0 mmol)
in Ac O (25 mL) at 0 °C. The reaction was allowed to warm
to room temperature over 2.5 h, and then the Ac O was
destroyed by slow addition of saturated NaHCO (700 mL) at
°C. After the mixture was stirred at room temperature for
4 h, the product was extracted from the aqueous layer with
O, dried (MgSO ), filtered, and concentrated. The residue
1
119
3
2
dried, and analyzed by H and Sn NMR, which revealed that
both cis- and trans-1,3-dioxa-2-stannolanes were present.
2
P in a col Cycliza tion of Ad ip a ld eh yd e w ith Bu
). A solution of adipaldehyde (228 mg, 2.00 mmol), Bu
699 mg, 2.40 mmol), AIBN (33 mg, 0.20 mmol), and tetrade-
cane (198 mg, 1.00 mmol; internal GC standard) in benzene
20 mL) was heated to reflux for 1 h, at which time GC analysis
3
Sn H (Eq
2
2
3
SnH
3
(
0
2
(
Et
2
4
indicated that 8% of the dialdehyde remained. The reaction
was refluxed for an additional 1.75 h, cooled to room temper-
ature, and made homogeneous by the addition of CH Cl (20
2 2
was purified by flash chromatography, followed by bulb-to-
bulb distillation, which afforded trans-1,2-cyclohexanediol
1
13
mononitrate as a white solid. The H and C NMR spectra
were identical to published spectra.
mL). GC analysis of an acetylated aliquot revealed that the
bisacetate of 1,2-cyclohexanediol was present in 95% yield
2
2
cis-1,2-Cycloh exa n ed iol Mon on itr a te. This material
was prepared as above from cis-1,2-cyclohexanediol (667 mg,
2.1:1 trans/cis by GC and by 1H NMR) and that 6-acetoxy-
(
hexanal was present in 3% yield.
P in a col Cycliza tion of Ad ip a ld eh yd e w ith TTMS (Eq
5
.74 mmol) and fuming HNO
3 2
(362 mg, 5.74 mmol) in Ac O
(
20 mL). Flash chromatography afforded cis-1,2-cyclohex-
2
(
). A solution of adipaldehyde (228 mg, 2.00 mmol), TTMS
597 mg, 2.40 mmol), AIBN (33 mg, 0.20 mmol), and tetrade-
cane (198 mg, 1.00 mmol; internal GC standard) in benzene
20 mL) was heated to reflux for 3 h. GC analysis of the
reaction mixture indicated that 5% of adipaldehyde remained
1
anediol mononitrate as a clear, colorless oil: H NMR (500
MHz, CDCl
1
3
) δ 5.09 (m, 1H), 4.03 (m, 1H), 2.05-1.98 (m, 1H),
.83-1.65 (m, 6H), 1.46-1.38 (m, 2H); 13C NMR (125 MHz,
(
CDCl
3
) δ 83.4, 68.0, 30.7, 25.5, 22.0, 20.6.
SnOEt (1.41 g, 4.19 mmol) was
added to a solution of trans-1,2-cyclohexanediol mononitrate
676 mg, 4.19 mmol) in pentane (4 mL). The reaction mixture
was stirred under vacuum until no ethoxy peaks remained in
Syn th esis of tr a n s-4. Bu
3
1
and that 1.11 mmol of the TTMS had been consumed.
H
NMR analysis revealed polymeric material in the carbinol,
aliphatic, and TMS regions. An aliquot was treated with
excess TBAF, acetylated, and analyzed by GC. The bisacetate
of 1,2-cyclohexanediol was found to be present in 71% yield
(
1
1
the H NMR spectrum: H NMR (300 MHz, C
6
D
6
) δ 4.87 (m,
H), 3.58 (m, 1H), 1.90-1.80 (m, 2H), 1.64-1.56 (m, 6H), 1.40-
.28 (m, 9H), 1.13-0.90 (m, 18H); 13C NMR (75 MHz, C
) δ
1
1
8
(1.4:1 trans/cis), as determined by integration versus the
6
D
6
internal standard, along with 4% of acetylated acyclic 1,2-
reduction products (corrected for 5% adipaldehyde).
1
19
9.1, 74.1, 37.0, 29.0, 28.6, 27.8, 24.4, 24.1, 15.5, 14.2; Sn
NMR (112 MHz) δ 100.2.
Syn th esis of cis-4. This material was prepared as above
P in a col Cycliza tion of Ad ip a ld eh yd e w ith Bu
). A solution of adipaldehyde (9.1 mg, 0.080 mmol), Bu
23.5 mg, 0.0960 mmol), AIBN (1.5 mg, 0.010 mmol), n-butyl
3
GeH (Eq
2
(
3
GeH
from cis-1,2-cyclohexanediol mononitrate (161 mg, 1.00 mmol)
1
and Bu
3
SnOEt (335 mg, 1.00 mmol): H NMR (300 MHz, C
6
D
6
)
methyl ether (1.20 mg, 0.0133 mmol; from stock solution;
δ 4.76 (m, 1H), 3.98 (m, 1H), 2.01-1.95 (m, 1H), 1.78-1.53
1
internal H NMR standard), and tetradecane (8.0 mg, 0.040
1
3
(
m, 8H), 1.45-1.26 (m, 8H), 1.13-0.89 (m, 18H); C NMR (125
MHz, C ) δ 86.5, 70.5, 34.8, 28.6, 27.8, 25.4, 23.9, 20.7, 15.5,
4.3.
Note: Nitrate esters are known to afford oxygen-centered
radicals when treated with Bu SnH under free-radical condi-
6 6
mmol; internal GC standard) in C D (0.800 mL) was heated
6
D
6
to 80 °C in a J . Young tube. The reaction was monitored by
1
1
H NMR. Additional AIBN (2 mg) was added after 9.3 and
2
4.3 h (total time). After 35 h, 10% of the dialdehyde
3
remained. The reaction mixture was acetylated and analyzed
by GC, which revealed a 27% yield (2.5:1 trans/cis) of the
(
21) Basavaiah, D.; Pandiaraju, S.; Muthukumaran, K. Tetrahedron:
Asymmetry 1996, 7, 13-16.
22) Cristol, S. J .; Franzus, B. J . Am. Chem. Soc. 1957, 79, 2488-
489.
(
(23) Beckwith, A. L. J .; Hay, B. P. J . Am. Chem. Soc. 1988, 110,
4415-4416.
2

Bu
3
SnH-Mediated Pinacol Coupling
J . Org. Chem., Vol. 63, No. 18, 1998 6381
bisacetate of 1,2-cyclohexanediol, along with 16% and 21% of
acetylated 1,6-hexanediol and 6-hydroxyhexanal, respectively.
Corporation. Acknowledgment is made to the donors of
the Petroleum Research Fund, administered by the
ACS, for partial support of this research.
Ack n ow led gm en t. Support has been provided by
the American Cancer Society, the Camille and Henry
Dreyfus Foundation, Eli Lilly, Firmenich, Glaxo
Wellcome, the National Science Foundation (predoctoral
fellowship to D.S.H.; Young Investigator Award, with
funding from Merck, Pharmacia & Upjohn, Bristol-
Myers Squibb, DuPont, Bayer, Rohm & Haas, and
Novartis), Pfizer, Procter & Gamble, and the Research
Su p p or tin g In for m a tion Ava ila ble: 1H NMR spectra of
the reaction products illustrated in Table 1 (6 pages). This
material is contained in libraries on microfiche, immediately
follows this article in the microfilm version of the journal, and
can be ordered from the ACS; see any current masthead page
for ordering information.
J O9809130