Alkoxides of o-silyl benzyl alcohols in cleavage reactions: Approaches to benzyl and silyl anion equivalents

Article abstract of DOI:10.1139/v99-250

The alkoxides of o-silyl benzyl alcohols 4 undergo cleavage reactions under mild conditions allowing transfer of a benzyl or silyl group to an electrophile. Reactions 3a → 8, and 3c or 4c → 11 illustrate their potential as benzyl and silyl anion equivalents.

Full text of DOI:10.1139/v99-250

1421
Alkoxides of o-silyl benzyl alcohols in cleavage
reactions: approaches to benzyl and silyl anion
equivalents
Paul F. Hudrlik, Jose O. Arango, Yousef M. Hijji, Cosmas O. Okoro,
and Anne M. Hudrlik
Abstract: The alkoxides of o-silyl benzyl alcohols 4 undergo cleavage reactions under mild conditions allowing trans-
fer of a benzyl or silyl group to an electrophile. Reactions 3a → 8, and 3c or 4c → 11 illustrate their potential as
benzyl and silyl anion equivalents.
Key words: silyl migration, benzyl anion equivalent, silyl anion equivalent, β-silyl ketone, γ-hydroxysilane.
Résumé : Les alcoolates d’alcools benzyliques o-silylés (4) subissent des réactions de clivage dans des conditions
douces qui permettent d’effectuer le transfert d’un groupe benzyle ou silyle vers un électrophile. Les réactions 3a → 8
et 3c ou 4c → 11 illustrent leur potentiel comme équivalents des anions respectivement benzyle et silyle.
Mots clés : migration d’un silyle, équivalent de l’anion benzyle, équivalent de l’anion silyle, β-silylcétone,
γ-hydroxysilane.
Hudrlik et al. 1427
Introduction
nucleophile could, in principle, be carried through a syn-
thetic sequence, and later activated (deprotected) to form a
more readily cleaved silicon group.
We have previously found that the γ-oxidopropyl group on
silicon is effective as an internal nucleophile. We have re-
ported the use of γ-hydroxysilanes 1 in rearrangement reac-
tions (eq. [1]) (4a), and of γ-hydroxysilanes 2a,b in cleavage
reactions (for generation of benzyl and allyl anion equiva-
lents) (eq. [2]) (4b).
Reactions involving nucleophilic attack at silicon have
been used in organic synthesis for the removal of protecting
groups (1), for the formation of multiple bonds (elimination
reactions) (2), and for the generation of anionic intermedi-
ates (3). We have been interested in the use of
intramolecular nucleophilic attack for some of these reac-
tions. Organosilicon compounds with a protected internal
[1]
[2]
Received July 5, 1999. Published on the NRC Research Press website on October 25, 2000.
Dedicated to Professor Adrian Brook in recognition of his outstanding contributions to organosilicon chemistry.
P.F. Hudrlik,1 J.O. Arango, C.O. Okoro, and A.M. Hudrlik. Department of Chemistry, Howard University, Washington, DC
20059, U.S.A.
Y.M. Hijji. Department of Chemistry, Morgan State University, Baltimore, MD 21251, U.S.A.
1Author to whom correspondence may be addressed. Telephone: (202) 806-4245. Fax: (202) 806-5442.
e-mail: phudrlik@fac.howard.edu
Can. J. Chem. 78: 1421–1427 (2000)
© 2000 NRC Canada

1422
Can. J. Chem. Vol. 78, 2000
Scheme 1.
[3]
A mechanistic rationale appears in Scheme 1. Treatment
of 3 with RLi or of 4 with base should lead to alkoxide 5.
Lithium alkoxide 5a appears to be stable in ether, although
in the presence of THF it yields oxasilacyclopentane 7, most
likely via pentacoordinate intermediate 6. In the case of 3a,
presumably an anionic intermediate such as 6 or Z– was
trapped with water to form toluene, and could in principle be
trapped with a number of electrophiles, allowing 3 and 4 to
be used to generate anion equivalents.
We felt that the use of a more rigid ortho-substituted aro-
matic ring in a γ-hydroxysilane such as 4 should facilitate
the cleavage reactions. We report that this appears to be the
case, and that these reactions can be used for the generation
of benzyl and silyl anion equivalents.
Results and discussion
We were able to demonstrate that the 3a–4a system can
serve as a benzyl anion equivalent by treating silyl ether 3a
with Li in ether (2 h) followed by benzophenone in THF
(12 h, room temperature). Benzyldiphenylcarbinol (8) was
isolated in 57% yield after chromatography (not optimized).
In contrast, the analogous reaction with the acyclic γ-
hydroxysilane 2a required the use of a more polar solvent
(HMPA) (4b).
o-Silyl benzyl alcohol 4a was prepared from o-
bromobenzyl alcohol by treatment with PhCH2SiMe2Cl and
triethylamine (to give silyl ether 3a in 87% yield) followed
by treatment of 3a with lithium wire in ether to give 4a
(92%) (5). Lithium alkoxide 5a (Scheme 1), a presumed in-
termediate in the rearrangement of 3a to 4a, appears to be
relatively stable in ether, but undergoes a facile cleavage re-
action in the presence of THF. Treatment of benzyl silyl
ether 3a with n-BuLi in ether at room temperature resulted
in disappearance of 3a after 1.5 h, with appearance of alco-
hol 4a (analysis of aliquots (with aqueous workup) by GC).
Addition of THF to the reaction mixture resulted in almost
complete disappearance of 4a within 1 h with the appear-
ance (after aqueous workup) of toluene and a product as-
signed as oxasilacyclopentane 7 by the mass spectrum [m/z
164 (18%, M+), 149 (100%, M+ – Me)]. By contrast, treat-
ment of the acyclic analog 2a with n-BuLi–THF resulted in
no reaction in 1 day (although cleavage of the benzyl group
was observed under more polar conditions, e.g., KH–ether or
NaH–THF) (4b).
Evidence for the intermediacy of oxasilacyclopentane 7 in
the cleavage reactions was obtained by trapping with MeLi.
Treatment of γ-hydroxysilane 4c (5) with KH–THF–0°C fol-
© 2000 NRC Canada

Hudrlik et al.
1423
lowed by excess MeLi–ether (15 h, room temperature) gave
γ-hydroxysilane 4d (5, 7) in 82% yield. Initial attempts to
isolate and more fully characterize oxasilacyclopentane 7 led
to samples which appeared pure by GC and GC/MS, but had
sizeable OH signals in the IR, suggesting the oxasilacyclo-
pentane might be forming a dihydroxy compound during
workup or partially forming oligomers.
good (72%) yields of 4c, but the Grignard reaction was hard
to initiate and keep going. Use of activated (Rieke) magne-
sium (11) in THF (room temperature) resulted in a facile re-
action to give 89% crude yield of a product which appeared
to be pure 4c. Use of lithium wire (containing 1% sodium)
in ether (0°C → room temperature) resulted in 78–92%
yields of 4c. In preliminary experiments, lithium powder
(~0.5% Na) was also found to give 4c in a similar manner.
The lithium wire procedure was convenient, and was the
procedure we used to prepare alcohol 4c. In all cases, the
progress of the reaction was followed by GC or GC/MS of
aliquots before aqueous workup of the entire reaction. The
pentamethyldisilyl ether of benzyl alcohol (hydrolysis prod-
uct of the intermediate from the metal-exchange reaction)
could be seen in small amounts in some of the early aliquots
from the reactions with lithium wire, and in larger amounts
in early aliquots from the reactions with magnesium and ac-
tivated magnesium.
Relative reactivities of the γ-hydroxysilanes were studied
in reactions which were followed by GC and GC/MS of
aliquots using an internal hydrocarbon standard (with aque-
ous workup). γ-Hydroxysilane 4c was considerably more re-
active to cleavage than the acyclic analog 2c (and 2c was
less reactive than the corresponding benzyl compound 2a).
For example, with NaH in THF at room temperature, reac-
tion of 4c was complete by 30 min, while reaction of 2c was
complete by 4 h (and reaction of 2a by 30 min (4b)). With
KH in Et2O at room temperature, cleavage reactions of 2a
and 4c were complete within 3–5 min, and that of 2c was
mostly complete by 90 min.
As mentioned above, the lithium alkoxide of 4a (gener-
ated from 3a) appeared to be stable in ether, but disappeared
within an hour upon the addition of THF. The lithium
alkoxide of 4c was more stable; 70% of alcohol 4c remained
one hour after treatment of 4c with MeLi in THF. Treatment
of alcohol 2c with MeLi–THF for 22 h resulted in no reac-
tion (after workup).
To study the possible utility of these substrates as silyl an-
ion equivalents, alcohols 2c and 4c were treated with base
and bromo silyl ether 3c was treated with lithium wire to
generate the corresponding alkoxides (5 in the case of 3c
and 4c), and then they were treated with electrophiles, some-
times in a more polar solvent.
To isolate oxasilacyclopentane 7, we decided to use γ-
hydroxy(allyl)silane 4b (5) as a substrate because the ex-
pected by-product, propene, would be volatile. Compound
4b was prepared by treatment of silyl ether 3b with Li–Et2O
(5). Treatment of 4b with KH in ether (1 h, room tempera-
ture), followed by removal of solvent under vacuum (no
aqueous workup) and bulb-to-bulb distillation (room temper-
1
ature, oil pump vacuum) gave a product with an H NMR
spectrum expected for 7 (singlets at δ 0.40 and 5.15, and a
multiplet at 7.2–7.6 in a ratio of 6:2:4), but with an IR spec-
trum having a rather large OH (3438 cm–1). Because we felt
the allylmetallic species generated in this reaction might be
causing oligomerization of oxasilacyclopentane 7, we added
1,2-dibromoethane as a trapping agent. Treatment of 4b with
t-BuLi (1.1 equiv) in THF (0°C, 1.5 h) followed by 1,2-
dibromoethane (removal of cold bath, 30 min) followed by
removal of solvent (no aqueous workup) and distillation
gave a sample of 7 having an IR spectrum which did not
1
have the OH peak. The H NMR spectrum was essentially
identical to that above, except for considerable amounts of
dibromoethane (as well as small amounts of ether and THF);
the GC showed two major peaks corresponding to
dibromoethane (26%) and oxasilacyclopentane 7 (70%).
We were especially interested to study the possibility of
using these reactions to generate silyl anion equivalents.
Silylmetallic reagents (8) such as trimethylsilyllithium and
phenyldimethylsilyllithium have found considerable use in
synthesis, but existing methods for the preparation of
silylmetallic reagents have disadvantages. Trimethylsilyllith-
ium is most often prepared from the reaction of hexamethyl-
disilane with MeLi in HMPA (9), which is highly polar,
nonvolatile, and a suspected carcinogen. Trimethylsilyllith-
ium can also be made from silylmercury compounds which
are undoubtedly highly toxic. Phenyldimethylsilyllithium is
easily prepared by the reaction of the disilane (or the silyl
chloride) with lithium, but this type of reaction cannot be
used for trialkylsilylmetallic reagents (8), and has been
found to fail with a number of substituted aryldimethylsilyl
groups (10). We are interested in developing general meth-
ods to generate trimethylsilyl- and other silylmetallic re-
agents in solvents which are not highly polar, nonvolatile, or
toxic.
Compounds 2c, 3c, and 4c were prepared as potential sub-
strates for generation of silyl anion equivalents. Alcohol 2c
was prepared from Me3SiSi(Me2)F by treatment with
allylmagnesium bromide (54% yield) followed by
hydroboration-oxidation of the product (64% yield). Silyl
ether 3c was prepared from o-bromobenzyl alcohol by treat-
ment with Me3SiSi(Me)2Cl and triethylamine in 92% yield.
Alcohol 4c was prepared via rearrangement of 3c (eq. [3])
with lithium or magnesium followed by aqueous
workup. This transformation was studied with several differ-
ent conditions, and the yields have been improved over those
reported (5). Use of magnesium turnings in ether–THF gave
Compounds 2c and 4c were each treated with base fol-
lowed by Ph2C=O or PhCHO. The products were the Me3Si
ethers of benzhydrol (9) and of benzyl alcohol (10), respec-
tively, presumably from trapping of the Me3Si group fol-
lowed by Brook (12) rearrangement. Benzhydrol and benzyl
alcohol were also present in the product mixtures, presum-
ably from hydrolysis of the silyl ethers.
© 2000 NRC Canada

1424
Can. J. Chem. Vol. 78, 2000
For example, treatment of alcohol 4c with KH–THF and
followed by analysis of aliquots. Two major products were
seen in the GC and GC/MS, oxasilacyclopentane 7 and silyl
ketone 11. Silyl ketone 11 was generally not observed at low
temperatures, but began to appear above 0°C. Two reactions
were run in the presence of tridecane as an internal standard,
and in two other reactions tridecane was added to a portion
of the crude product. The yields of silyl ketone 11 at com-
pletion were determined by GC to be 61–65%.
Following the progress of the reaction by aliquots ap-
peared to be important. In one case the reaction time (3–4
days) was longer than the others (~1 day), and a reaction in
which 1.2 equiv of CuCN was used was faster (~7 h), but
the GC yields at completion were all about the same.
The use of alcohol 4c as a precursor to silyl ketone 11
was also investigated. Reactions of 4c with n-BuLi followed
by cyclohexenone were carried out using CuCN and Et2Zn.
The use of CuCN gave good yields of silyl ketone 11. (An
attempt to improve the reaction using Me3SiCl as an additive
(19) led to the trimethylsilyl ether of alcohol 4c.) The use of
Et2Zn instead of CuCN resulted in mostly recovered 4c; in
one reaction silyl ketone 11 was formed in 14% yield by
GC.
benzaldehyde (mixed at 0°C, then room temperature, 1 h)
gave a mixture of silyl ether 10, benzyl alcohol, and
oxasilacyclopentane 7 in GC area ratios of 13%, 33%, and
47%, respectively. Silyl ether 10 (13) was isolated by
chromatography in 39% yield. Treatment of 4c with KH–
THF followed by benzophenone similarly gave a mixture
containing silyl ether 9. Use of MeLi–THF followed by ben-
zophenone in THF–HMPA (mixed at 0°C, then room tem-
perature, 15 h) gave a similar mixture from which silyl ether
9 (13) was isolated in 35% yield.
When silyl ether 3c was treated with lithium wire in ether
followed by benzyl bromide in THF, and when alcohol 4c
was treated with n-BuLi followed by benzyl bromide or
chloride in THF, little or no benzyltrimethylsilane was
formed. However, when 4c in THF was treated with MeLi
followed by benzyl bromide in THF–HMPA (0°C, 1 h; room
temperature, 14 h), benzyltrimethylsilane was formed, along
with comparable amounts of oxasilacyclopentane 7 and
benzyl bromide, and small amounts of bibenzyl.
The generation of anionic intermediates for a conjugate
addition reaction was studied using both silyl ether 3c and
alcohol 4c. As mentioned above, silyl ether 3c could be con-
verted to alcohol 4c (or alkoxide 5c) with either lithium or
magnesium. Since both copper and zinc reagents are known
to promote conjugate addition reactions (14), we have stud-
ied several combinations of these reagents with 3c and 2-
cyclohexenone to try to effect conjugate addition to form 3-
trimethylsilylcyclohexanone (11). β-Silyl ketones such as 11
are of interest (for a recent method for preparing β-silyl ke-
tones via silylcopper reagents see ref. 15) because of the di-
recting effect of the silyl group on reactions such as the
Baeyer–Villiger reaction (16), the Beckman fragmentation
reaction (9b, 17), and alkylation reactions (18).
The role of copper in these reactions was demonstrated by
the reaction of 4c with n-BuLi in ether followed by
cyclohexenone in THF. The major product was
oxasilacyclopentane 7; there was no evidence for silyl
ketone 11, the 1,2-addition product (1-trimethylsilyl-2-
cyclohexenol (20)), or its Brook rearrangement product, the
trimethylsilyl ether (21) of 2-cyclohexenol.
In the best procedure, the reactions were run in a very
similar manner to those with silyl ether 3c and lithium wire
followed by CuCN: An ice-cooled ether solution of alcohol
4c was treated with n-BuLi, 0.5–0.6 equiv of CuCN, cooled
to about –78°C, then cyclohexenone in THF was added. The
mixture was gradually warmed to room temperature follow-
ing the reaction by GC and (or) GC/MS of aliquots. GC and
GC/MS showed only oxasilacyclopentane 7 and silyl ketone
11 as major products. In one reaction, the crude product was
purified by chromatography, yielding silyl ketone 11 in 50%
yield. In another reaction, tridecane was added as an internal
standard, and GC yields were determined. As in the case
with the reactions of silyl ether 3c with lithium wire and
CuCN, silyl ketone 11 did not begin to appear in the aliquots
until the reaction temperature rose above 0°C (e.g., 11%
yield of 11 at 14°C, 8.75 h after addition of cyclohexenone).
At reaction completion, the GC yield of silyl ketone 11 was
73%.
Reactions of silyl ether 3c with lithium, CuCN, and
cyclohexenone gave moderate yields of silyl ketone 11. The
corresponding reactions with activated magnesium instead of
lithium gave no evidence for the formation of 11. The reac-
tions of 3c with lithium, Et2Zn, and cyclohexenone gave
small amounts of silyl ketone 11 in two reactions. In most
cases (Mg–CuCN and Li–Et2Zn reactions) alcohol 4c and
oxasilacyclopentane 7 were major products.
In summary, we have found that easily available o-silyl
benzyl alcohols 4 and their precursor silyl ethers 3 can be
used to generate alkoxides 5 which undergo facile cleavage
reactions. Alcohols 4 and silyl ethers 3 are therefore of po-
tential use for the generation of anionic intermediates.
The best procedure of those we tried used lithium wire
and CuCN. An ice-cooled solution of 3c in ether was treated
with lithium wire until conversion to alkoxide 5c was com-
plete (from GC or GC/MS analysis of aliquots); the resulting
solution was cannulated away from the unreacted lithium
into a flask containing about 0.5 equiv of CuCN and ether or
THF. Cyclohexenone in THF was added at low temperature
(–90 to –75°C), and the solution was allowed to gradually
warm to room temperature. The progress of the reaction was
Experimental section
General procedures
All reactions were carried out under a nitrogen or argon
atmosphere, and liquids were transferred with nitrogen- or
argon-flushed syringes. The verb “concentrated” refers to re-
moval of the solvent by a rotary evaporator. Column chro-
matography was carried out as flash chromatography (22)
© 2000 NRC Canada

Hudrlik et al.
1425
Reaction of silyl ether 3a with lithium and benzophenone:
Formation of benzyldiphenylcarbinol (8)
using 200–425 mesh silica gel. Nuclear magnetic resonance
(NMR) spectra were obtained in CDCl3 (1H at 300 MHz and
13C at 75 MHz) and chemical shifts are reported in δ using
To a solution of 0.20 g (0.60 mmol) of silyl ether 3a (5) in
5 mL of anhydrous ether was added 0.1 g (14 mmol) of lith-
ium wire (flattened and washed with petroleum ether). The
reaction was stirred at room temperature for 2 h (following
the progress by GC analysis of an aliquot). Then 0.12 g
(0.66 mmol) of benzophenone was added, and the reaction
mixture was stirred at room temperature for 30 min (aliquots
showed little or no reaction). Then 5 mL of THF was added,
and the progress of the reaction was followed by aliquots
taken after 5 min, 30 min, and 12 h. The reaction mixture
was then added to saturated NaHCO3 and extracted twice
with ether. The organic layer was dried (MgSO4), concen-
trated, and chromatographed on 20 g of silica gel using 8:2
pet ether:ether giving 0.094 g (57%) of benzyldiphenylcar-
binol (8) as a white solid: mp 88–89°C (lit. (27) mp 89–
90°C).
1
CHCl3 (δ 7.26) for H NMR, and CDCl3 (δ 77.00) for 13C
NMR as internal standard. Methylene chloride (δ 5.32) was
1
used as an internal standard for H NMR for compounds
containing aromatic hydrogens. Gas chromatography (GC)
analyses were carried out with a flame ionization detector
and using helium as the carrier gas using a Shimadzu GC-14A
instrument with a 15 m × 0.32 mm × 25 µm cross-linked
methyl silicone gum capillary column, and the following
temperature program: 50°C (5 min), 10°C/min to 230°C
(15 min) unless otherwise noted. Most reactions were fol-
lowed by GC or GC/MS of aliquots (using the aqueous
workup of the reaction). GC yields were determined using
tridecane calibrated with an authentic sample (9) of 3-
trimethylsilylcyclohexanone (11). In a few cases where
noted, GC analyses were carried out with a Hewlett-Packard
HP 5880 instrument with 12.5 and 25 m cross-linked poly-
methyldisiloxane capillary columns and the following tem-
perature program: 50°C (1 min), 10°C/min to 250° (5 min).
o-Bromobenzyloxypentamethyldisilane (3c)
A solution of 6.30 g (33.7 mmol) of o-bromobenzyl alco-
hol, 14 mL of triethylamine, and 110 mL of anhydrous ether
was cooled in an ice bath, and 5.80 g (34.8 mmol) of
pentamethylchlorodisilane in 15 mL of anhydrous ether was
added dropwise. The mixture was stirred at room tempera-
ture for about 2 h, and then added to 170 mL of aqueous
NaHCO3 overlaid with 50 mL of ether. The organic layer
was washed with water (4 × 30 mL), dried (MgSO4), con-
centrated, and placed under oil pump vacuum (3 mm, 4 h).
Chromatography on silica gel (26.7 × 4.4 cm) using petro-
leum ether gave 10.2 g (92%) of silyl ether 3c (5): IR (neat)
3066, 2950, 2891, 1443, 1375, 1247, 1090, 1026, 837, 806,
Materials
Chloropentamethyldisilane was prepared by a procedure
of Ishikawa et al. (23) and was purchased from Gelest. An-
hydrous ether and tetrahydrofuran (THF) were distilled from
sodium and benzophenone. Lithium wire contained 1% so-
dium. 2-Cyclohexenone was dried over magnesium sulfate
and fractionally distilled under reduced pressure. Triethyl-
amine was distilled from calcium hydride.
1
749 cm–1; H NMR δ 0.17 (s, 9 H), 0.34 (s, 6 H), 4.76 (s, 2
3-Hydroxypropylpentamethyldisilane (2c)
Allylpentamethyldisilane (24) was prepared from fluoro-
pentamethyldisilane (25) and allylmagnesium bromide, and
H), 7.15 (crude t, J = 7 Hz, 1 H), 7.37 (crude t, J = 7 Hz, 1
H), 7.5–7.6 (m, 2 H); 13C NMR (assignments by DEPT) δ
–2.01 (CH3), –0.75 (CH3), 64.99 (CH2), 121.08 (C), 127.23
(CH), 127.62 (CH), 128.15 (CH), 131.98 (CH), 140.12 (C).
GC analysis showed one major peak at 15.3 min (100%).
The mass spectrum was equivalent to that reported (5).
1
had IR, H NMR, and mass spectra corresponding to those
reported (24).
To an ice-cooled solution of 2.43 g (14 mmol) of
allylpentamethyldisilane in 10 mL of THF was added 42 mL
(42 mmol) of BH3⋅THF (1.0 M in THF) dropwise. The ice
bath was replaced with a water bath, and the mixture was
stirred for 24 h. The mixture was recooled in ice, and the
following additions were made: (1) 10 mL of H2O,
(2) 10 mL of 3 M NaOH, (3) 10 mL of 30% H2O2 (dropwise).
The resulting mixture was heated at 50°C for 3 h, cooled to
room temperature, and then 9.0 g of K2CO3 was added. The
mixture was stirred an additional 5 min, then the layers were
allowed to separate, and the aqueous layer was extracted
with ether (2 × 10 mL), and the combined organic layers
were washed with saturated NaHCO3, dried (MgSO4), con-
centrated, chromatographed (50 g of Florisil, ether:petroleum
ether (3:7)), and Kugelrohr distilled (oven temperature 70°C,
oil pump vacuum) to give 1.72 g (64%) of 3-hydroxy-
propylpentamethyldisilane (2c) (26): IR (film) 3340 (b),
2956, 1461, 1246 (m), 1069, 834 cm–1; mass spectrum m/z
(relative intensity, tentative assignment) 190 (M+, not visi-
ble), 147 (5, (Me3SiOSiMe2)+), 133 (20), 131 (20,
Me3SiMe2Si+), 117 (60, Me2Si+CH2CH2CH2OH), 75 (100,
(Me2SiOH)+). GC analysis (HP 5880, dodecane = 6.42 min)
o-Pentamethyldisilylbenzyl alcohol (4c)
A solution of 10.77 g (33.9 mmol) of o-bromo-
benzyloxypentamethyldisilane (3c) in 130 mL of anhydrous
ether was cooled in an ice bath. Lithium wire was wiped
with paper towels to remove oil, weighed, placed between
weighing papers, flattened with a hammer, and cut into
small pieces. To the above solution 2.18 g (314 mmol) of the
lithium was added. Typically, the reaction was stirred at ice
temperature until the slurry turned from slightly turbid to
light brown-yellow or until GC or GC/MS indicated the re-
action had begun. Then the ice bath was removed, and the
reaction was stirred at room temperature until GC and (or)
GC/MS indicated the reaction was complete (and the reac-
tion mixture had turned cloudy gray). In this case, the reac-
tion was stirred at ice temperature 1.5 h after the addition of
lithium, and 1 h after removal of the ice bath. The reaction
mixture was transferred to a beaker, and the remaining lith-
ium pieces were quickly removed. The mixture was then im-
mediately poured into 120 mL of aqueous NaHCO3 overlaid
with 50 mL of ether. The organic layer was dried (MgSO4),
concentrated, placed under oil pump vacuum (0.75 mm,
1
showed one peak at 6.13 min (100%). The H NMR spec-
trum is the same as that reported (26).
© 2000 NRC Canada

1426
Can. J. Chem. Vol. 78, 2000
Table 1. GC yields of aliquots.
2.5 h), and chromatographed on silica gel (35 × 4.4 cm) us-
ing petroleum ether:ether (1:0, 7:3, and then 1:1) giving
Time (h)
Temp (°C)
% GC yield of 11
1
6.82 g (84%) of alcohol 4c. The IR, H NMR, 13C NMR,
1.1
4.1
7.3
9.8
–83
5
16
15
0
and mass spectra were equivalent to those reported (5).
33.0
61.0
60.6
Reaction of silyl ether 3c with lithium, CuCN, and 2-
cyclohexenone: Formation of 3-trimethylsilylcyclohexanone
(11)
A solution of 1.10 g (3.47 mmol) of silyl ether 3c and
40 mL of anhydrous ether was cooled in an ice bath, and
0.22 g (31.7 mmol) of lithium wire (for handling, see the
synthesis of 4c) was added. The progress of the reaction was
followed visually and by GC/MS as described for the syn-
thesis of 4c. In this case, the reaction was stirred in an ice
bath for 35 min, and then the ice bath was removed and stir-
ring was continued (total 2 h). The reaction mixture was
transferred via cannula into an ice-cooled mixture of 0.20 g
(2.23 mmol) of CuCN in 40 mL of THF. The resulting dark
brown slurry was stirred for 35 min at ice temperature, and
then cooled to –75°C. After 15 min, 0.6 mL (6.2 mmol) of
cyclohexenone in 20 mL of THF was added dropwise. The
slurry was kept at –80°C for 1 h, then allowed to warm up
gradually to room temperature, and the progress of the reac-
tion was monitored by GC/MS and GC analysis of aliquots.
After 21 h, the mixture was added to 160 mL of aqueous
NH4Cl overlaid with 50 mL of ether. The organic layer was
dried (MgSO4) and concentrated. The GC analysis showed
peaks at 7.91 min (59%, oxasilacyclopentane 7), 9.35 min
(29%, silyl ketone 11), and 16.32 min (12%), and the
GC/MS showed two peaks in the GC chromatogram having
mass spectra corresponding to 7 (major) and 11. The product
was placed under oil pump vacuum (2.5 mm, 0.5 h) giving
1.65 g of a light brown crude product. The IR, 1H NMR, and
13C NMR were consistent with the presence of predomi-
nantly 7 and 11. A portion of the product was mixed with
tridecane in ether, and the GC yield of 11 was calculated to
be 65%.
An analogous reaction was done in which the reaction
mixture was cannulated into CuCN in ether rather than in
THF. GC analysis of the crude product showed two major
peaks at 7.99 min (57%, 7) and 9.45 min (43%, 11). The
spectra were all consistent with predominantly 7 and 11. A
portion of the crude product was mixed with tridecane in
ether, and the GC yield of silyl ketone 11 was calculated to
be 62%.
A similar reaction was done in which the standard was
present when the aliquots were taken (and 1.2 equiv of
CuCN were used), and GC yields were calculated at various
reaction times: 0.76 g (2.39 mmol) of silyl ether 3c in 45 mL
of anhydrous ether, 0.25 g (36 mmol) of lithium wire, 0°C
for 2.5 h; mixture transferred via cannula into an ice-cooled
mixture of 0.25 g (2.79 mmol) of CuCN, 0.074 g of
tridecane, and 15 mL of anhydrous ether (0°C for 35 min),
then cooled to –90°C for 20 min; cyclohexenone (0.55 mL,
5.68 mmol) in 35 mL of THF added, –80 to –85°C for about
an hour, and then mixture warmed gradually to room tem-
perature, monitoring progress of the reaction by GC and
GC/MS of aliquots. GC yields are given in Table 1.
After 10.3 h, the reaction was quenched by adding it to
aqueous NH4Cl overlaid with ether, the organic layer was
dried (MgSO4), and the GC yield of 11 was determined to
Table 2. GC yields of aliquots.
Time (h)
Temp (°C)
% GC yield of 11
0.5
3.5
6.75
8.75
11.25
21.75
–80
–7
9
14
16
22
0
0
3.2
11.3
39.8
73.1
1
be 62%. After removal of solvent, the spectra (IR, H NMR,
13C NMR, GC/MS) were all consistent with the presence of
oxasilacyclopentane 7, silyl ketone 11, and tridecane, and
the GC analysis showed peaks at 7.93 min (31%, 7),
9.38 min (16%, 11), and 11.22 min (10%, tridecane).
Reaction of alcohol 4c with n-BuLi, CuCN, and 2-
cyclohexenone: Formation of 3-trimethylsilylcyclohexanone
(11)
A solution of 1.05 g (4.40 mmol) of alcohol 4c in 100 mL
of anhydrous ether was cooled in an ice bath and 3.9 mL
(6.24 mmol) of n-butyllithium (1.6 M in hexane) was added.
The reaction mixture was stirred and cooled in an ice bath
for about 0.5 h at which time the slurry turned light yellow.
Then 0.23 g (2.6 mmol) of CuCN was added, within 5 min
40 mL of THF was added, and the mixture was stirred at
0°C for 20 min. Then the slurry was cooled to –80°C, and
after 35 min, a solution of 0.74 mL (7.64 mmol) of
cyclohexenone in 80 mL of THF was added. The reaction
mixture was allowed to warm up gradually, and the progress
of the reaction was monitored by GC analysis and GC/MS
of aliquots. The reaction mixture was quenched after 13 h by
adding it to a mixture of 200 mL of aqueous NH4Cl overlaid
with 85 mL of ether, and the organic layer was dried
(MgSO4) and concentrated. GC analysis showed two major
peaks at 6.17 min (55%, oxasilacyclopentane 7) and
7.82 min (19%, silyl ketone 11), and a number of small
peaks from 10–25 min. The product was placed under oil
pump vacuum (6–7 mm, 2 h). The spectra were consistent
with the presence of 7 and 11. The crude product was
chromatographed on silica gel (24 × 2.9 cm). Elution with
petroleum ether:ether (4:1) gave 0.372 g (50% yield) of silyl
ketone 11 as a colorless liquid: 13C NMR (assignments by
APT) δ –3.90 (CH3), 25.85 (CH2), 27.82 (CH), 29.75 (CH2),
41.77 (CH2), 42.19 (CH2), 212.41 (C=O); GC/MS m/z (rela-
tive intensity, tentative assignment) 170 (16, M+), 169 (15),
155 (24, M+ – Me), 127 (20), 75 (65), 73 (100, Me3Si+). The
spectra were essentially identical to those of an authentic
1
sample of silyl ketone 11 (9), and the IR and H NMR were
identical to those reported (15).
An analogous reaction was done in which the standard
was present when the aliquots were taken, and GC yields
© 2000 NRC Canada

Hudrlik et al.
1427
11. R.D. Rieke, S.E. Bales, P.M. Hudnall, T.P. Burns, and G.S.
Poindexter. Org. Syn. Coll. Vol. VI, 845 (1988).
12. (a) A.G. Brook. Acc. Chem. Res. 7, 77 (1974); (b) A.G. Brook
and A.R. Bassindale. In Rearrangements in ground and excited
states. Vol. 2. Edited by P. de Mayo. Academic Press, New
York, 1980. pp. 149–227.
13. W. Gerrard and K.D. Kilburn. J. Chem. Soc. 1536 (1956).
14. (a) W. Tückmantel, K. Oshima, and H. Nozaki. Chem. Ber.
119, 1581 (1986); (b) R.A.N.C. Crump, I. Fleming, and C. J.
Urch. J. Chem. Soc. Perkin Trans. 1, 701 (1994).
15. H. Ito, T. Ishizuka, J. Tateiwa, M. Sonoda, and A. Hosomi. J.
Am. Chem. Soc. 120, 11 196 (1998).
16. (a) P.F. Hudrlik, A.M. Hudrlik, G. Nagendrappa, T. Yimenu,
E.T. Zellers, and E. Chin. J. Am. Chem. Soc. 102, 6894
(1980). Correction: J. Am. Chem. Soc. 104, 1157 (1982);
(b) P.F. Hudrlik, A.M. Hudrlik, T. Yimenu, M.A. Waugh, and
G. Nagendrappa. Tetrahedron, 44, 3791 (1988); (c) T. Aida,
M. Asaoka, S. Sonoda, and H. Takei. Heterocycles, 36, 427
(1993).
were calculated at various reaction times: 0.63
g
(2.64 mmol) of alcohol 4c in 45 mL of anhydrous ether and
0.135 g of tridecane, 1.7 mL (3.4 mmol) of n-butyllithium
(2.0 M in cyclohexane) (0.5 h at 0°C), 0.135 g (1.5 mmol)
of CuCN (30 min), cooled to –75°C for 15 min, dropwise
addition of 0.45 mL (4.65 mmol) of cyclohexenone in
40 mL of THF, gradual warming to room temperature, mon-
itoring the reaction by GC analysis of aliquots. GC yields
are given in Table 2.
After 22.75 h the reaction was quenched by adding it to
aqueous NH4Cl overlaid with ether, the organic layer was
dried (MgSO4), and the GC yield of 11 was determined to
1
be 73%. After removal of solvent, the spectra (IR, H NMR,
13C NMR, GC/MS) were all consistent with the presence of
oxasilacyclopentane 7, silyl ketone 11, and tridecane.
Acknowledgements
Financial support from the National Science Foundation
17. H. Nishiyama, K. Sakuta, N. Osaka, H. Arai, M. Matsumoto,
(CHE-9505465
and
CHE-9007879)
and
NIGMS
and K. Itoh. Tetrahedron, 44, 2413 (1988).
(SO6GM08016) is gratefully acknowledged.
18. R.A.N.C. Crump, I. Fleming, J.H.M. Hill, D. Parker, N.L.
Reddy, and D. Waterson. J. Chem. Soc. Perkin Trans. 1, 3277
(1992).
19. (a) B.H. Lipshutz, S.H. Dimock, and B. James. J. Am. Chem.
Soc. 115, 9283 (1993), and refs. therein; (b) S.H. Bertz, G.
Miao, B.E. Rossiter, and J.P. Snyder. J. Am. Chem. Soc. 117,
11 023 (1995).
20. (a) G. Altnau, L. Rösch, and G. Jas. Tetrahedron Lett. 24, 45
(1983); (b) P.F. Hudrlik, L. Tafesse, and A.M. Hudrlik. J. Am.
Chem. Soc. 119, 11 689 (1997).
References
1. Review: T.D. Nelson and R.D. Crouch. Synthesis, 1031
(1996).
2. Review: D.J. Ager. Org. React. (N.Y.) 38, 1 (1990).
3. For example, see A.S. Pilcher and P. DeShong. J. Org. Chem.
61, 6901 (1996).
4. (a) P.F. Hudrlik, Y.M. Abdallah, and A.M. Hudrlik. Tetrahe-
dron Lett. 33, 6743 (1992); see also ref. 6; (b) P.F. Hudrlik,
Y.M. Abdallah, and A.M. Hudrlik. Tetrahedron Lett. 33, 6747
(1992).
21. P.F. Hudrlik, C.-N. Wan, and G.P. Withers. Tetrahedron Lett.
1449 (1976).
22. W.C. Still, M. Kahn, and A. Mitra. J. Org. Chem. 43, 2923
(1978).
23. M. Ishikawa, M. Kumada, and H. Sakurai. J. Organomet.
Chem. 23, 63 (1970).
24. (a) Yu.P. Egorov, G.G. Kirei, S.A. Samoilenko, E.A.
Chernyshev, and N.G. Tolstikova. Bull. Acad. Sci. USSR Div.
Chem. Sci. (Eng. Trans.) 515 (1963); (b) T.A. Dixon, K.P.
Steele, and W.P. Weber. J. Organomet. Chem. 231, 299 (1982).
25. M. Kumada, M. Yamaguchi, Y. Yamamoto, J. Nakajima, and
K. Shiina. J. Org. Chem. 21, 1264 (1956).
5. Y.M. Hijji, P.F. Hudrlik, C.O. Okoro, and A.M. Hudrlik.
Synth. Commun. 27, 4297 (1997).
6. Y.M. Hijji, P.F. Hudrlik, and A.M. Hudrlik. Chem. Commun.
(Cambridge), 1213 (1998).
7. N. Meyer and D. Seebach. Chem. Ber. 113, 1304 (1980).
8. For a review, see J. Belzner and U. Dehnert. In The chemistry
of organic silicon compounds. Vol. 2. Part 1. Edited by Z.
Rappoport and Y. Apeloig. Wiley, New York, N.Y. 1998.
pp. 779–825.
9. (a) W.C. Still. J. Org. Chem. 41, 3063 (1976); (b) P.F. Hudrlik,
M.A. Waugh, and A.M. Hudrlik. J. Organomet. Chem. 271, 69
(1984).
26. Y. Nakadaira, A. Sekiguchi, Y. Funada, and H. Sakurai. Chem.
Lett. 327 (1991).
27. R.O.C. Norman and R.A. Watson. J. Chem. Soc. B, 692
(1968). See also ref. 4b.
10. N. Abd. Rahman, I. Fleming, and A.B. Zwicky. J. Chem. Res.
Synop. 292 (1992).
© 2000 NRC Canada