Reversibility of titanocene-catalyzed hydromagnesation of silylacetylenes

Article abstract of DOI:10.1016/0022-328X(91)86189-W

The regioisomeric silylvinylmagnesium compounds A and B, products of titanocene-catalyzed hydromagnesation of trimethylsilylacetylenes, are in a rapid dynamic equilibrium with starting acetylene and, therefore, with each other due to the reversibility of the reaction.The existence of this equilibrium is demonstrated by transfer, under the conditions of hydromagnesation, of 'HMgBr' from the vinylmagnesium compounds A/B to other acetylenes with regeneration of the starting acetylenes.Me3Si(BrMg)C=C(H)R (A)<*>Me3SiC<*>Cr<*>Me3Si(H)C=C(MgBr)R (B).The position of the equilibrium between A and B was determined by deuterolysis and found to depend on the substituent, being 77 : 23 for R = n-C5H11 and 95 : 5 for R = cyclo-CH2C6H11 in diethyl ether.Reactions of the magnesium reagents A/B with electrophiles produce mixtures of linear (type A) and branched (type B) products.The ratios of regioisomeric products differ depending on the electrophile and the conditions (from 98 : 2 to 20 : 80).This is due to the preferential reaction of one of the regioisomers and to rapid equilibration.

Full text of DOI:10.1016/0022-328X(91)86189-W

Journal of Organometallic Chemistry, 401 (1991) 1-11
Elsevier Sequoia S.A., Lausanne
JOM 21064
Reversibility of titanocene-catalyzed hydromagnesation
of silylacetylenes
M a r i n a A. Djadchenko, Kasimir K. Pivnitsky *
Institute of Experimental Endocrinology of The All-Union Endocrinology Scientific Centre, USSR
Academy of Medical Sciences, ul. Moskvorechje 1, Moscow 115522 (USSR)
Jorg Spanig and Hans Schick
Central Institute of Organic Chemistry, Academy of Sciences of the G.D.R., Rudower Chaussee 5,
Berlin-O-1199 (Germany)
(Received May 14th, 1990)
Abstract
The regioisomeric silylvinylmagnesium compounds A and B, products of titanocene-catalyzed hydro-
magnesation of trimethylsilylacetylenes, are in a rapid dynamic equilibrium with starting acetylene and,
therefore, with each other due to the reversibility of the reaction. The existence of this equilibrium is
demonstrated by transfer, under the conditions of hydromagnesation, of 'HMgBr' from the vinylmag-
nesium compounds A/B to other acetylenes with regeneration of the starting acetylenes.
Me3Si(BrMg)C--C(H)R (A)~-Me3Si~--CR~ Me3Si(H)C=C(MgBr)R (B). The position of the equi-
librium between A and B was determined by deuterolysis and found to depend on the substituent, being
77 : 23 for R = n-CsH11 and 95 :5 for R = cyclo-CH2C6Hll in diethyl ether, Reactions of the magnesium
reagents A/B with electrophiles produce mixtures of linear (type A) and branched (type B) products. The
ratios of regioisomeric products differ depending on the electrophile and the conditions (from 98:2 to
20 : 80). This is due to the preferential reaction of one of the regioisomers and to rapid equilibration.
Introduction
In recent years hydromagnesation of silylacetylenes catalyzed by titanocene
dichloride has been increasingly used in organic synthesis. This is due to its high
regio- and stereoselectivity and to the versatility of the resulting 1-(1-silylvinyl)mag-
n e s i u m reagents (see ref.
1 and references therein). In our investigations of the
application of these reagents to the synthesis of eicosanoids, we have been faced
with some results w h i c h we report below, w h i c h can be most easily interpreted by
assuming the reversibility of the hydromagnesation reaction. As such a reversibility
c o u l d have significant consequences for the applicability of the acetylene hydromag-
nesation reaction, we have studied the question in more direct experiments. Here,
we report on the reversibility of silylacetylene hydromagnesation when catalyzed by
0022-328X/91/$03.50
© 1991 - Elsevier Sequoia S.A.

titanocene dichloride, and present examples of its influence on the reactivity of the
products, the silylvinylmagnesium bromides.
Results and discussion
The accepted mechanism of titanocene-catalyzed hydromagnesation of acetylenes
(see Scheme 1) [2] postulates that the latter are hydrotitanated (step 1) by the
hydridotitanium(III) species B generated in situ by the Grignard reagent from
titanocene dichloride via the intermediate titanium(Ill) species A. The product C
exchanges titanium for magnesium (step 2) producing the final vinylmagnesium
compound D and titanium species A, thus closing the catalytic cycle. After comple-
tion of the process, perhaps after all Grignard reagent has been consumed, the
hydridotitanium species B ceases to be regenerated. However, if steps 1 and 2 are
reversible, B continues to exist in its equilibrium concentration (which is most
probably very low) and could be trapped continuously by adding another acetylene,
producing new hydromagnesated species similar to D and regenerating the starting
acetylene (see Scheme 2). Of course, this process should also be reversible. Therefore
the demonstration of (reversible) hydridomagnesium transfer from species D to an
other acetylene, with the regeneration of starting acetylene, could be regarded as a
proof of the reversibility of acetylene hydromagnesation reactions.
The results of the experiments observing the hydridomagnesium transfer are
shown in Table 1. The first, starting acetylene 1 taken in small excess (1.06 molar
equivalent), was hydromagnesated by isobutylmagnesium bromide in the presence
of 8 tool% of titanocene(IV) dichloride, under the exact conditions described by
Sato [2], except that excess catalyst was present, to accelerate the process. Comple-
tion of the reaction takes 20 rain to 18 hours, depending on the acetylene, and
results in formation of an equivalent quantity of the corresponding vinylmagnesium
compound II. Then an equivalent of the other acetylene I* was added and the
course of the process was followed by GLC analysis of protonated aliquots. To
make the diagram more clear the second, recipient acetylenes and their products
bear asterisks and the corresponding formulas in Scheme 2 are in italics. In most of
the I*+ !! pairs the gradual appearance of the starting acetylene 1 was observed
with concomitant formation of the new vinylmagnesium compound II*. The forma-
i-BuMgBr
Cp2TiCI2
'
Cp2TiBu'i
(A)
i-BuMgBr
l
R
R
R
R
¢pffi. (,),
o)
~
,
RC~-~CR
Scheme 1
(2)
H
TiCp2
H
MgBr
(c)
(D)

3
Table 1
Reactions of trans-hydromagnesation of vinyimagnesium bromides 11 and acetylenes !* (see Scheme 2)
Entry
Starting
materials
Reaction
time (h)
Composition of
reaction mixture
Yield of !i*,
%'s on reacted II
I
I*
87
II
82
II*
1
2
3
lla+l*b
lla + l*c
lib + l*a
7.5
94.5
18
52
13
25
72
49
75
48
24
95
63
91
44
27
37
9
58
73
92
80
4.25
94.3
29
75
83
59
71
25
17
41
58
55
4
5
6
lib + l*c
ile+l*a
llc + l*b
28.5
25
96
0
10
100
100
4
100
100
90
0
94
0
25
0
0
0
tion of the latter was demonstrated by reactions with water as well as with
trimethylsilyltriflate, both of which gave the same products (vide infra) as obtained
from the vinylmagnesium reagent synthesized by the direct hydromagnesation. The
process approached equilibrium in 28-95 hours, but true equilibrium was not
reached owing to deactivation of the catalyst.
The rate of the trans-hydromagnesation and its extent should both depend on the
relative susceptibilities to hydromagnesation of the recipient acetylenes. The de-
termination of relative rates of hydromagnesation of the acetylenes la-c in competi-
tive experiments was confounded by the exchange process (cf. Scheme 2). A rough
estimation was obtained by comparing the reaction periods needed for a 95%
hydromagnesation under standard conditions (see Experimental). These relative
rates were found to be 3 : 1 : 60 for l a - c , respectively, and correlate well with the
extents of trans-hydromagnesation. Transfers to the most effective recipient Ic
(entries 2 and 4 in Table 1) were almost complete within a short time, while no
RC~--- CR' (la-c)
(i)
l
R
H
R'
R
R'
===~
+
RC-~ CR'
•
R C ~ C R ' +
+I~IgBr
H + MgBr
(n,,-e)
(ra-¢)
0,,-e)
( l l * a - c )
a: R = Me3Si, R' - n-CsH11
b: R = Me3Si, R' = cyclo-CH2C6H11
c: R = n-Pr, R' = n-CsHla
Scheme 2. (i) i-BuMgBr, Cp2TiCI2, Et20.

transfer was detected from the corresponding magnesium compound lle (entries 5
and 6). The transfer of magnesium from the magnesium compounds lla,b to the
acetylenes la,b was of medium extent, and, compared with the reverse process
IIa---, lb (entry 1), the transfer to the better recipient IIb ~ la (entry 3) was
significantly more complete. We consider these observations to be additional
evidence of the reversible trans-hydromagnesation approaching an equilibrium
under the chosen experimental conditions, at least in the case of trimethylsilyl-sub-
stituted acetylenes.
The reversibility of the acetylene hydromagnesation that we have discovered
would be expected to have several consequences on the chemistry of the resulting
vinylmagnesium compounds II. Firstly, a dynamic equilibrium may exist between
the regioisomeric vinylmagnesium compounds llAa,b and llBa,b formed from
silylacetylenes la,b (Scheme 3). Indeed, due to their equilibrium with the parent
acetylenes la,b [reaction (1)], the regioisomers should be in equilibrium with each
other, and the equilibrium position be determined solely by thermodynamics.
Although inter-molecular trans-hydromagnesation (Scheme 2) takes several hours to
achieve equilibrium, the analogous process in Scheme 3 could proceed intramolecu-
larly, without dissociation of delivered hydridotitanium species from an acetylene
molecule (so-called "internal return"), much more rapidly. Therefore, an equi-
librium in Scheme 3 might be reached already while hydromagnesation is proceed-
ing. Analyzing the ratios of regioisomeric IIAa and IIBa by deuterolysis of the
aliquots of the reaction mixtures during the acetylene la hydromagnesation showed
the constant ratio of 77 : 23 after 2, 5 or 22 hours (entry 1 in Table 2). In a similar
experiment with the acetylene Ib the ratio of the vinylmagnesium compounds lIAb
to llBb was found to be 95:5 after 22 hours (entry 2). The preponderance of the
1-silyl-l-vinylmagnesium regioisomers IIAa,b is to be expected from the well-known
stabilisation of carbanions by silyl substituents, and the lower proportion of the
minor regioisomer IIBb in comparison with IIBa is consistent with the destabilisa-
tion of the former by steric repulsion of the bromomagnesium group with the
bulkier cyclohexylmethyl substituent in IIBb. All these data agree with the notion of
Me3Si
BrMg
R
Me3Si
R
(1_____~) Me3SiC~CR
(1)
~
X - - C H O
H
H
MgBr
(llAa,b)
(la,b)
(llBa,b)
(XIII)
[
a: R = n . C , Hu
[
0
X = O
(2) (i)
b: R = cyclo-CH2CrHll
(3) (i)
i
:
El
H
H
E1
"
(HI-VU)
(vm-Ydl)
Scheme 3. (i) Electrophile.

I
v ~
I
I
~o
+
-t-
+
+
+
+
+
.-
~
~ °
°
?
~J
o~
~m

a rapidly established equilibrium between the regioisomeric silylvinylmagnesium
compounds IIA and liB, the position of which is governed by the thermodynamic
stability of the components. In his pioneering paper [2] Sato described a regioi-
someric ratio of 95 : 5 for silylvinylmagnesium compounds, which was found to be
slightly dependent on the procedure of hydromagnesation. Sato's ratio is close to
ours for the acetylene Ib, but differs from that for the acetylene la, for unknown
reasons. We can only suggest minor differences in the procedures influence the
quantity that is generated of the active catalyst (species B in Scheme 1) and thus the
rate of establishing the equilibrium.
We should note that in the other regioselective acetylene hydromagnesations
(arylacetylenes, silylarylacetylenes, propargyl alcohols) that have been described, the
main regioisomer formed is the most stable thermodynamically [2-4], in other
words the results could also be explained by the position of the equilibrium between
the regioisomers of vinylmagnesium compounds.
The second possible consequence of the reversibility of acetylene hydromagnesa-
tion lies in the ratios of regioisomers formed in the reactions of vinylmagnesium
compounds IIA/B with various electrophiles (reactions 2 and 3 in Scheme 3). If the
rates of these reactions are both much greater than the rate of reaction 1, the ratio
of linear products III-VII to corresponding branched ones VIII-Xll would be equal
to the ratio of magnesium compounds IIA and lIB in equilibrium. Deuterium oxide
is believed to be such an electrophile. However, if the rates of reactions 2 and/or 3
are smaller than that of reaction 1, the ratio of linear and branched products should
depend on the ratio of these reaction rates being different from the equilibrium ratio
and influenced by the reaction conditions. We investigated the reactions of the
vinylmagnesium compounds IIA/B with several electrophiles (see Table 2). Reac-
tions with trimethylsilyltriflate (trimethylsilylchloride does not react) with a 77:23
mixture of the magnesium compounds IIA/Ba provided mixtures of the correspond-
ing silylated compunds IVa and IXa, with compositions close to those of the starting
mixture, with dependence on the reaction temperature (entry 3). Another extreme
was found in the reaction with carbon tetrabromide which resulted in almost pure
bromide Va with traces only of regioisomeric Xa (entry 5). Intermediate cases were
found in reactions with aldehyde XIII (a prostanoid synthon [5]) and cyclo-
heptanone (taken as an example of a ketone with hardly any tendency to enolisa-
tion) where the proportions of minor regioisomers were in the range of 6-19%,
being higher for ketone (entries 6, 7). An analogous silylation of a 95 : 5 mixture of
the magnesium compounds IIA/Bb resulted in the mixture of IVb and IXb with the
same ratio of regioisomers (entry 4). The most striking result was encountered when
the bromomagnesium residue in the compounds IIA/Ba was initially substituted for
tri(isopropyloxy)titanium(IV) and the resulting titanium reagent reacted with the
aldehyde XlII (entry 8). The carbinol Xlla, which was the minor isomer in a
corresponding reaction of the magnesium reagent IIA/Ba, now became by far the
major isomer. One or more of the following reasons could explain this result.
Selective generation of the titanium reagent of type B, its preponderance in
equilibrium and enhanced reactivity, but the primary reason is again the dynamic
equilibrium between the regioisomers of silylvinylmagnesium(titanium) compounds.
It is rapid in this example at - 7 8 ° C due to additional catalysis by tri(isopropyl-
oxy)titanium(IV) compounds [6] present in stoichiometric quantity.
We have demonstrated the reversibility of titanocene-catalyzed hydromagnesa-

tion only in the case of silylalkylacetylenes, but see no reason why the same
phenomenon could not exist with other acetylenes.
Experimental
Spectra conditions: IR, Specord IR-75 (G.D.R.), CCI4 solutions; PMR, Tesla
BS-487C (80 MHz) (C.S.S.R.), and Bruker WH-360 (360 MHz) (F.R.G.), CCI4
solutions, internal standard: tetramethylsilane; mass-spectra (MS), LKB 2091
(Sweden), E1 at 22.5 eV, in the process of chromatography on open tubular glass
capillary column, 25 m × 0.36 mm, SE-30 liquid phase 0.4 micron, carrier gas
(helium) velocity of 60 cm/s, column, separator and ion source temperatures
110-203°C (depending on the substance), 280°C and 250°C, respectively. GLC
analyses: Chrom-5 apparatus (C.S.S.R.); condition No. 1: open tubular fused silica
column, 50 m x 0.25 mm, OV-1 0.2 micron, helium 35 cm/s, l l 0 ° C for 2 rain, then
heating to 270°C at 20°/min; condition No. 2: same column, helium 30 cm/s,
100°C for 4 rain, then heating to 270°C at 20°/min; condition No. 3: open
tubular glass capillary, 25 m × 0.35 mm, SP-1000 0.3 micron, helium 50 cm/s,
isothermally at 234 ° C. For TLC on silica gel Silufol UV-254 (C.S.S.R.) plates were
used.
All experiments with organometallic compounds were performed in the static
atmosphere of purified argon in diethyl ether purified by distillation from benzo-
phenone-sodium. Usual treatment of reaction mixtures consisted in dilution by
hexane or diethyl ether, separation of the aqueous phase, additional extraction by
the same organic solvent, washing of combined extracts with brine, drying with
anhydrous Na2SO4, and rotoevaporation at aspirator vacuum. Isobutylmagnesium
bromide in ether was prepared by a standard method [7].
Synthesis of starting materials
l-Trimethylsilylhept-l-yne (la). Prepared from commercial 1-heptyne according
to ref. 8, yield 89%, b.p. 75-76°C/20 mmHg, n~ 1.4342 (lit. [8]: b.p. 176°C, n~
1.4344).
1-Trimethylsilyl-3-cyciohexylprop-l-yne (Ib). 8.6 ml of 1.4 M solution (12.0 mmol)
of n-butyllithium were added to a stirred solution of 1.7 ml (12.1 mmol) of
diisopropylamine in 20 ml abs. ether at -20 °C and after 15 min a solution of 1.22
g (10.0 retool) of cyclohexylallene (prepared according to ref. 9, b.p. 50-51°C/17
mmHg, n~ 1.4789) was added. The reaction mixture was maintained at 30-36°C.
After 4.5 h GLC analysis (condition No. 1) of the water-quenched reaction mixture
demonstrated the conversion into the 8:2 mixture of 3-cyclohexylprop-l-yne and
presumed 3-cyclohexylprop-2-yne, retention times (RT) being 3.93 and 4.29 rain,
respectively, for starting allene 4.18 rain. The bright-yellow reaction mixture was
cooled to - 7 8 ° C , 1.7 ml (13.4 mmol) of trimethylsilylchloride were added, the
mixture was heated for 15 rain to 23°C and maintained at this temperature for 1.5
h, which resulted in the formation of a white precipitate. Reaction was quenched at
0 o C by adding 25 ml of saturated aqueous ammonium chloride solution and treated
in the usual manner (ether). Fractional distillation isolated 1.43 g (74%) of silyla-
cetylene Ib, b.p. 93-98.5°C/9 mmHg, n2D3"5 1.4619. IR (cm-1): 2127, 1450, 1251,
1029, 843. PMR (CDCI3, 80 MHz), 8, ppm: 0.137 s (9H, SiMe3), 2.06 d (2H, J 6.2
Hz, CH2C~-C). MS, m/z (rel. intensity, %): 194 (5, [M]+), 179 (100, [ M - Me]+),

120 (5, [M-HSiMe3I+), 112 (11, [Me3SiC:::--CMe]+), 83 (5, [CrH~I]+), 73 (16,
[Me3Si]+).
Titanocene-catalyzed hydromagnesation of acetylenes la-c.
A. For use in transfer hydromagnesation reactions. To the suspension of 10 mg
(0.04 mmol) of finely ground titanocene dichloride in solution of 0.50 mmol of an
alkyne and 25-27 mg of n-decane (internal standard for GLC analysis) in 0.5 ml of
abs. ether was added an ether solution of isobutylmagnesium bromide (2.16 M, 0.22
ml, 0.47 mmol) under stirring at 23°C. The reaction mixture turned out to be
brown-blackish with a complete dissolution of titanocene. The course of the
reaction was followed by GLC analysis of aliquots quenched by pouring into a
hexane-phosphate buffer solution (pH 6.8), using the percentage of the correspond-
ing (Z)-olefine formed after protonation as a measure of hydromagnesation com-
pleteness. Data for the acetylenes Ia-e: reaction times 5-6 h, 18 h, 15-20 min;
conversions 95%, 90-95%, 94-95%; RT (condition No. 2) of starting acetylenes
5.92, 9.00, 5.57 rain, of resulting (Z)-olefins 6.09, 9.29, 5.49 rain, resp. Spectral data
for protonation products. 1-Trimethylsilylhept-l(Z)-ene. IR (cm-1): 1611, 1250, 853,
836. PMR (80 MHz), 8, ppm: 0.10 s (gH, SiMe3), 0.89 t (3H, J 4.5 Hz, Me), 2.04 q
(2H, J 6.5 Hz, C3H2), 5.39 dt [1H, J 14.5 + 1 Hz (compare with [10]), CIH], 6.24 dt
(1H, J 14.5 + 7.2 Hz, C2H). MS, m/z (%): 170 (3, [M]+), 155 (100, [ M - Me]+),
114 (37, [Me3SiCH--CHMe]+), 99 (32, [114-Me]+), 95 (20), 85 (15), 73 (64,
[Me3Si]+), 59 (52, [Me2SiH]+). 1-Trimethylsilyl-3-cyclohexylprop-l(Z)-ene. IR
(cm-l): 1606, 1448, 1258, 1248, 854, 833, 686. PMR (80 MHz), 8, ppm: 0.10 s (3H,
SiMe3), 1.93 t (2H, J 6.5 Hz, C3H2), 5.42 dt (1H, J 14 + 1 Hz, CIH), 6.25 dt (1H,
J 14 + 7 Hz, C2H). MS, m/z (%): 196 (0.5, [M]+), 181 (100, [M - Me]+), 153 (19),
114 (89, [Me3SiCH=CHMe]+), 99 (20, [Me3SiCH=CH]+), 83 (20, [C6HH]+), 73(89),
59(19). 4(Z)-Decene. IR (cm-a): 1656, 715. PMR (80 MHz), 8, ppm: 0.89 t (6H, J
6.5 Hz, 2 × Me), 1.95 q (4H, J 5 Hz, 2 × CH2C=), 5.24 m (2H, CH=CH).
B. For use in reactions with electrophiles. Procedure as above but using 0.02
mmol of titanocene dichloride and 0.85 mmol of isobutylmagnesium bromide and
reaction times (for Ia,h) of 18-20 h.
Reactions of transfer hydromagnesation
To the solution of alkenylmagnesium bromide prepared as described above a
solution of the equivalent quantity (0.50 mmol) of an acetylene la-e in 0.3 ml abs.
ether was added and the course of trans-hydromagnesation was followed by GLC
analysis (condition No. 2) of aliquots quenched by pouring into the mixture of
phosphate buffer (pH 6.8) and hexane. For the results see Table 1.
Reactions of the alkenylmagnesium bromides l l a , b with electrophiles (conditions and
results see Table 2)
A. lla with deuterium oxide. The aliquots (0.5 mmol) of the reaction mixture in
the synthesis of magnesium reagent lla were withdrawn and quenched by pouring
them into the mixture of deuterium oxide (1 ml) and hexane. Usual work-up
followed by filtration in hexane through 0.2 g Kieselgel 60 (ASTM) provided the
mixtures of the deuterated olefines Ilia + Villa for analysis by PMR.
B. lib with deuterium oxide. A portion of the magnesium reagent lib (0.5 mmol)
was quenched by deuterium oxide (1 ml) and the product isolated as above.

C. lla with trimethylsilyl trifluoromethanesulphonate. To the solution of 0.50
mmol of lla (see above) 1.20 ml of a 30% solution of silyl reagent in benzene were
added and the reaction was followed by GLC. After quenching with phosphate
buffer (pH 6.8), usual treatment and filtration in hexane through a small Kieselgel
60 column the mixtures of 1,1- and 1,2-bis(trimethylsilyl)hept-l(E)-enes (IVa + IXa)
were isolated. IR of the mixture (cm-l): 1573, 1246, 859, 836, 686. Compound IVa.
GLC (condition No. 3), RT: 8.09 rain. PMR (80 MHz), 6, ppm: 0.06 s (9H,
cis-SiMe3), 0.14 s (9H, trans-SiMe3), 0.87 t (3H, J 5 Hz, Me), 2.10 m (2H, C3H2),
6.59 t (1H, J 7 Hz, C2H). Compound IXa. GLC: 7.57 min. PMR: 0.06 s (9H,
CISiMe3), 0.11 s (9H, C2SiMe3), 5.94 b.s (1H, CIH) (compare with [11]). MS, m/z
(% for IVa and IXa, respectively): 242 (5, 0.3, [M]÷), 227 (41, 6, [ M - Me]+), 186
(25, 41, [(Me3Si)2C----CHMe]+), 153 (100, 29, [227 - HSiMe3]+), 98 (54, 48), 73 (70,
100, [SiMe3]+); metastable ions 143.0 (242 ~ 186), 103.1 (227 ~ 153), 62.8 (153
98).
D. lib with trimethylsilyl trifluoromethanesulphonate. Magnesium reagent lib,
treated analogously to the above procedure, produced 1,1-bis(trimethylsilyl)-3-
cyclohexylprop-l-ene (IVb) with a small (see Table 2) admixture of 1,2-bis(trimeth-
ylsilyl)-3-cyclohexylprop-l(E)-ene (IXb). GLC (condition No. 2), RT: 8.86 and 8.42
min, resp. Spectra of the mixture. IR (cm-~): 1566, 1450, 1262, 1250, 893, 856, 833,
683. PMR (80 MHz), 6, ppm: 0.06 s (9H, cis-SiMe3), 0.15 s (9H, trans-SiMe3), 2.06
t (2H, J 6.5 Hz, C3H2), 6.62 t (1H, J 7.2 Hz, C2H) (signals of the minor isomer
could not be identified as they coincided with signals of impurities, mainly protona-
tion product). MS, m/z (% for major and minor isomer, resp.): 268 (6, 0, [M}+),
253 (39, 6, [ M - Me]+), 186 (45, 65, [(MeaSi)2C----CHMe or isomer]+), 179 (100, 9,
[253 - HSiMe3]+), 98 (91, 64), 73 (64, 100).
E. lla with carbon tetrabromide. To a solution of 0.49 mmol of lla a solution of
185 mg (0.56 mmol) CBr4 in 0.3 ml abs. ether was added at ambient temperature.
Exothermic reaction resulted in red-brown discoloration. The quenching by phos-
phate buffer and usual treatment (ether) afforded after filtration in hexane through
a small Kieselgel 60 column a mixture of 1- and 2-bromo-l-trimethylsilylhept-
l(E)-enes (Va + Xa), a yellow oil, with 15-20% of la, protonated and unidentified
compounds as impurities. IR of the mixture (cm-I): 1604, 1470, 1460, 1250, 1106,
853, 839, 696, 673, 626. Compound Va. GLC (condition No. 2), RT: 8.29 min. PMR
(80 MHz), 6, ppm: 0.26 s (9H, SiMe3), 0.90 t (3H, J 5 Hz, Me), 2.04 q (2H, J 6.5
Hz, C3H~), 6.71 t (1H, J 8.2 Hz, C2H). Compound Xa. GLC: 7.95 rain. PMR: 0.17
s (9H, SiMe3), 5.88 b.s (1H, C1H). MS, m/z (% for Va and Xa, resp., ions with
89Br): 248 (3, 15, [M]+), 153 (5, 7, [M - Me - HBr]÷), 137 (99, 88, [BrSiMe2]+), 95
(45, 39, [ M - HBr - SiMe3]+), 73 (84, 81, [SiMe3]+), 67 (11, 7).
F. 11a with cycloheptanone. 199 mg (1.63 mmol) of cycloheptanone in 1.2 rnl of
abs. ether were used for 2 mmol of magnesium compound lla. Quenching with
saturated aqueous NH4CI solution, usual treatment (ether) and chromatography
using 20 g of silica gel (Woelm, TSC) and hexane as eluent resulted in the isolation
of 201 mg (44%) of the mixture of 1-[l'-(trimethylsilyl)hept-l'(Z)-en-l'-yl]cyclo-
heptan-l-ol (Via) and 1-[l'-(trimethylsilyl)hept-l'(E)-en-2'-yl]cycloheptan-l-ol
(Xla), a light-yellow oil. IR of the mixture (cm-I): 3607, 3485, 1603, 1460, 1448,
1250, 1023, 853, 839. Compound Via. TLC (hexane-ethylacetate, 9:1), R r 0.42.
GLC (condition No. 2), RT: 14.09 min. PMR (80 MHz), 6, ppm: 0.20 s (9H,
Cl'SiMe3), 0.89 t (3H, J 5.5 Hz, Me), 2.06 q (2H, J 7 Hz, C3'H2), 6.01 t (1H, J 7.4

10
Hz, C2'H). Compound Xla. TLC: 0.34. GLC: 13.73 min. PMR: 0.10 s (9H, SiMe3),
5.78 s (1H, C2'H). MS, m//,~ (~ for Vla and Xia, resp.): 282 (24, 2, [M]+), 267 (14,
10, [ M - M e ] + ) , 266 (14, 1), 264 (40, 5, [ M - H 2 0 ] + ) , 249 (21, 16, [ M - M e -
H20]÷), 223 (40, 1), 209 (42, 9, [ M - Bu - H20]+), 208 (29, 8), 197 (71, 4), 195 (42,
3), 191 (28, 2, [M - H20 - SiMe3]÷), 190 (29, 2), 155 (55, 60, [Me2SiCH=CHAm]+),
153 (31, 10), 136 (16, 33), 121 (16, 39), 114 (13, 14, [Me3SiCH=CHMe]+), 113 (55,
45, [Me3SiCH=CHCH2]+), 95 (17, 19, [cycloheptenyl]+), 75 (68, 23), 73 (100, 100).
G. IIa with 2-oxa-cis-bicyclo[3.3.0]octane-3-one-6-exo-carbaidehyde (XIII). 227
mg (1.47 mmol) of the aldehyde XIll in 1.5 ml of abs. THF were added at -78 °C
to the solution of 2.14 mmol of the magnesium compound lla, the mixture was
stirred for 150 min at - 2 0 ° C and quenched with saturated aqueous NH4C1
solution. Usual treatment (ethyl acetate) and high-performance flash chromatogra-
phy (HPFC) [12] yielded a mixture of epimeric/isomeric alcohols a + fl-VIIa and
a + fl-Xila, separable by GLC (condition No. 3), RT: 13.2, 13.8, 12.0 and 12.6 min,
ratio 15.0 : 78.5 : 1.4 : 5.1, resp. (stereochemistry determination will be published
elsewhere in due course). The mixture of epimeric alcohols a + fl-VIIa was isolated
by repeated HPFC, Rf 0.57 and 0.50, respectively (methylene chloride-acetone-
hexane, 8:1:1). IR (cm-l): 3602, 3440, 1773, 1661, 1613, 1251, 1172, 1047, 852,
837. PMR (360 MHz), 8, ppm: 0.18 s (9H, SiMe3), 0.90 t (3H, J 6.8 Hz, Me), 4.00 d
and 4.10 d (total 1H, J 7.4 and 6.1 Hz, CI'H), 4.94 t and 4.96 t (total 1H, J 6.3 and
6.5 Hz, CIH), 6.18 td and 6.21 td (total 1H, J 7.6 + 0.8 and 7.6 + 1.0 Hz, C3'H).
MS, m/z (% for a- and B-epimer, resp.) 324 (1.5, 1.6, [M]+), 309 (0.4, 0.3,
[ M - M e ] + ) , 306 (0.5, 0.5, [ M - H 2 0 ] ÷ ) , 291 (2, 2, [ 3 0 9 - n 2 0 ] + ) , 253 (3, 3,
[ M - Am]+), 225 (2, 2), 199 (20, 21, [AmCH--CSiMe3CHOH]+), 183 (100, 100), 129
(8, 2), 113 (10, 9), 75 (18, 16), 73 (14, 16); metastable ion: 168.2 (199 -, 183).
H. Ila with tri(isopropyloxy)titanium(IV) chloride followed by aldehyde XIII. An
ether solution of 2.76 mmol of the magnesium compound IIa was cooled at - 78o C
under stirring and 1.30 ml of a 2.12 M hexane solution (2.76 mmol) of tri(isopro-
pyloxy)titanium(IV) chloride were added, followed by the solution of 100 mg (0.65
mmol) of aldehyde XIII in 1.4 ml of abs. THF after 10 min. The mixture was stirred
at the same temperature for 2 h, heated to 23°C and, after 30 rain, diluted by
saturated aqueous NaHCO3 solution and ether, and filtered. After usual treatment
(ether) and flash chromatography the mixture of the alcohols a + fl-Vlla and
a + fl-Xlla (ratio 5.4 : 14.6 : 33.6 : 46.4, respectively) was isolated, yield 98 mg (47%).
IR of the mixture (cm-l): 3590, 3470, 1788, 1613, 1253, 1162, 1045, 855, 835. PMR
(360 MHz), 8, ppm (signals of epimeric a- and fl-XIIa): 0.12 s (9H, SiMe3), 0.90 t
(3H, J 6.7 Hz, Me), 3.97 d (1H, J 5.8 Hz, C1'H), 4.93 m (1H, C~H), 5.51 s (1H,
SiCI"H). MS, rn/z (% for a- and fl-epimer, resp.): 324 (12, 12), 309 (3, 3), 291 (27,
18), 267 (7, 8, [ M - Bu]+), 253 (19, 14), 251 (8, 7, [ M - SiMe3]÷), 199 (28, 24), 183
(22, 22), 153 (13, 11), 143 (30, 23), 133 (36, 26), 127 (42, 35), 126 (30, 24), 119 (24,
17), 118 (21, 16), 75 (97, 81), 73 (100, 100) (interpretation of the ions as above).
Acknowledgements
The authors are greatly indebted to Drs. M.A. Lapitskaja, E.A. Erperina and
D.A. Lemenovskii for generous samples of cyclohexylallene, 4-decyne and titano-
cene dichloride, respectively, 360 MHz NMR spectra were obtained in The All-Un-
ion Oncology Scientific Centre AMS USSR by courtesy of Dr. I.V. Yartzeva.

11
References
1 F. Sato, J. Organomet. Chem., 285 (1985) 53.
2 F. Sato, H. lshikawa and M. Sato, Tetrahedron Lett., 22 (1981) 85.
3 F. Sato, H. Watanabe, Y. Tanaka and M. Sato, J. Chem. Soc,, Chem. Commun., (1982) 1126.
4 J. Kang, W. Cbo and W.K. Lee, J. Org. Chem., 49 (1984) 1838.
5 P. Crabbe and A. Guzman, Tetrahedron Lett., (1972) 115.
6
H.L. Finkbeiner and G.D. Cooper, J. Org. Chem., 27 (1962) 3395.
7 B.A. Kazanskij (Ed.), Syntesi Organicheskich Preparatov (in Russian), Vol. 4, Inostrannaja Literatura,
Moscow, 1953, p. 314.
8 R.A. Benkeser and R.A. Hickner, J. Am. Chem. Soc., 80 (1958) 5298.
9 L. Brandsma and H.D. Verkruijsse, Synthesis of Acetylenes, Allenes and Cumulenes, Vol. 8, Elsevier,
Amsterdam-Oxford-New York, 1981, p. 183.
10 H. Shiragami, T. Kawamoto, K. Imi, S. Matsubara, K. Utimoto and H. Nozaki, Tetrahedron, 44
(1988) 4009; J.A. Soderquist and S.-J.H. Lee, ibid., 44 (1988) 4033.
11 K. Fugami, J. Hibino, S. Nakatsukasa, S. Matsubara, K. Oshima, K. Utimoto and H. Nozaki,
Tetrahedron, 44 (1988) 4277.
12 L.L. Vasiljeva and K.K. Pivnitsky, manuscript deposited in VINITI (Moscow), 1986, No. 6390-B86.