The ring expansion of a dimethyl-substituted vinylcyclobutene derivative by metal complexes that can undergo an inner-sphere electron transfer

Article abstract of DOI:10.1016/S0022-328X(00)00054-1

The ring expansion of a dimethyl-substituted vinylcyclobutene derivative to the corresponding dimethyl-substituted cyclohexadiene or aromatic compound was studied. It was found that metal complexes of Ni(I), Ti(III), Sm(II), and Fe(II), which can undergo

Full text of DOI:10.1016/S0022-328X(00)00054-1

www.elsevier.nl/locate/jorganchem
Journal of Organometallic Chemistry 601 (2000) 147–152
The ring expansion of a dimethyl-substituted vinylcyclobutene
derivative by metal complexes that can undergo an inner-sphere
electron transfer
Herman L. Holt Jr., Tom Russo, Allan R. Pinhas *
Department of Chemistry, Uni6ersity of Cincinnati, PO Box 210172, Cincinnati, OH 45221-0172, USA
Received 13 October 1999; accepted 17 January 2000
Abstract
The ring expansion of a dimethyl-substituted vinylcyclobutene derivative to the corresponding dimethyl-substituted cyclohexa-
diene or aromatic compound was studied. It was found that metal complexes of Ni(I), Ti(III), Sm(II), and Fe(II), which can
undergo an inner-sphere electron-transfer reaction, allow the rearrangement to occur at room temperature. Other oxidation states
of these metals and complexes that can only undergo an outer-sphere electron-transfer process do not promote this ring expansion
reaction. © 2000 Elsevier Science S.A. All rights reserved.
Keywords: Rearrangement reaction; Inner-sphere electron transfer; Ring expansion
1
. Introduction
One possible explanation of why the Ni(I) reaction is
catalytic and the Ni(0) reaction is not catalytic is that
the Ni(0) complex decomposes under the higher-tem-
perature conditions required for this reaction. Our evi-
dence for this decomposition includes the fact that
Over the past several years, we have studied the
nickel-promoted ring expansion of vinylcyclobutene
derivatives 1 to cyclohexadiene derivatives 2 [1–3].
Scheme 1 summarizes our results. (1) When compound
1
is treated with a Ni(II) complex, no reaction occurs
and starting compound 1 can be re-isolated. (2) When
compound 1 is treated with a stoichiometric amount of
a Ni(0) complex, compound 2 is generated, but only at
elevated temperatures (e.g. refluxing THF or benzene).
(
3) In contrast, when compound 1 is treated with a
Ni(I) complex, the rearrangement reaction is catalytic
in nickel and occurs at room temperature.
In the case in which R=H (compound 1a) and the
Ni(0) complex is (Ph P) Ni(C H ), complex 3 can be
3
2
2
4
isolated when the reaction is performed at room tem-
perature under an inert atmosphere. Complex 3 can be
converted to compound 2a either at elevated tempera-
ture by refluxing in THF, or at room temperature by
reacting complex 3 with a mild oxidant such as air,
+
3 2 2 2 2
(
Ph P) NiBr , or Cp Fe . A strong oxidant such as I
converts complex 3 back to compound 1a [2].
*
Corresponding author. Fax: +1-513-556 9239.
E-mail address: allan.pinhas@uc.edu (A.R. Pinhas)
Scheme 1.
0022-328X/00/$ - see front matter © 2000 Elsevier Science S.A. All rights reserved.
PII: S 0 0 2 2 - 3 2 8 X ( 0 0 ) 0 0 0 5 4 - 1

148
H.L. Holt, Jr. et al. / Journal of Organometallic Chemistry 601 (2000) 147–152
reduction of the readily available titanium(IV) complex
Cp TiCl (which is red in solution) [5]. The titanium(II)
2
2
complex Cp Ti was prepared by a butyllithium reduc-
Scheme 2.
2
tion of Cp TiCl [6]. The complex TiCl can be pre-
2
2
3
pared by a zinc reduction of the titanium(IV) complex
unreacted nickel complex cannot be isolated from the
reaction mixture, the formation of solids as the reaction
progresses, and in some cases, the formation of a nickel
mirror on the glassware. The decomposition of the
metal complex leads to the question: is it really Ni(0)
that is causing the rearrangement to occur at elevated
temperatures, or is the Ni(0) generating small amounts
of Ni(I) and it is Ni(I) that is catalyzing the reaction?
In this article, we shall show that Ni(0) simply is a
source of Ni(I) and that Ni(I) causes the rearrangement
reaction to occur. In addition, we shall show that other
metal complexes that can be oxidized easily by an inner
sphere process cause similar rearrangement reactions to
occur.
TiCl [7], or is commercially available. The complexes
4
5
SmI , SmCl , FeCl , FeCl , and Cp*Fe [Cp*=h -
2
3
2
3
2
C (CH ) ] are all commercially available.
5
3 5
Organic compound 1b (RꢀCH ) was synthesized in a
3
manner similar to that previously reported for com-
pound 1a [2], except 2,3-dimethylbutadiene was used in
place of butadiene in the photoreaction (Scheme 3).
Significant improvements in two of the steps have made
the synthesis easier. Firstly, for the esterification step,
we have found that simply adding a weak base and
methyl iodide generates the ester in good yield [8].
More importantly, we have found that the dechlorina-
tion reaction can be accomplished by using zinc, rather
than
with
the
expensive
nickel
complex
(
Ph P) Ni(C H ). However, this process works only if
3
2
2
4
the zinc is activated by removing the oxide coating with
a sulfuric acid wash. After the acid wash, the zinc then
must be rinsed with water to ensure that all the acid is
removed, and then rinsed with dry THF to insure that
all the moisture is removed.
Also, as shown in Scheme 3, an authentic sample of
compound 2b was easily synthesized from a Diels–
Alder reaction of 2,3-dimethylbutadiene and dimethyl
acetylenedicarboxylate.
2
. Synthesis of starting materials
In our previous work, as mentioned in Scheme 1, we
used the complex (Ph P) Ni(C H ). Unfortunately, the
3
2
2
4
synthesis of this nickel complex is very time consuming.
In addition, the complex is very air sensitive and cannot
be stored for extended periods. For the research dis-
cussed in this manuscript, we have employed the
analogous butene complex, (Ph P) Ni(C H ), which, as
3
2
4
8
shown in Scheme 2, is much easier to synthesize [4] and
duplicates all of the chemistry of the ethylene complex.
This chemistry is not restricted to only triphenyl phos-
phine as a ligand, the butyl phosphine analog,
3
. Ni(I) or Ni(0) as the active metal complex
To determine if Ni(0) is simply generating Ni(I),
which in turn is causing the rearrangement to occur, we
needed a method to quench immediately any Ni(I) that
could be forming under the reaction conditions by
reducing it back to Ni(0). If all the Ni(I) is reduced and
(
Bu P) Ni(C H ), can be synthesized in a similar man-
3 2 4 8
ner and works well in all of our nickel chemistry.
5
The titanium(III) complex Cp TiCl [Cp=h -C H ]
2
5
5
(
which is green in solution) was prepared by a zinc
Scheme 3.

H.L. Holt, Jr. et al. / Journal of Organometallic Chemistry 601 (2000) 147–152
149
Scheme 4.
the rearrangement reaction still occurs, then both Ni(I)
and Ni(0) are active metal complexes for this reaction.
In contrast, if reducing all the Ni(I) to Ni(0) stops the
rearrangement reaction, then the most straightforward
explanation is that Ni(I) is the only oxidation state of
the nickel which is active for this reaction.
either Cp TiCl or TiCl3 as the Ti(III) complex, as
2
shown in Scheme 4, aromatic compound 4 was ob-
tained in about a 75% yield. (The remainder of the
material is mainly unreacted starting material 1b.) Con-
trol experiments show:
1
. As with Ni(II), the Ti(IV) complex Cp TiCl does
2 2
Previously [2], we have shown that decamethyl fer-
not catalyze nor promote any reaction.
2. Zn, which is used to reduce Ti(IV) to Ti(III), and
5
rocene [Cp*Fe, Cp*=h -C (CH ) ] will reduce Ni(I) to
2
5
3 5
+
Ni(0) but that decamethyl ferrocinium [Cp*Fe ] will
2
ZnCl , the by-product of the reduction reaction,
2
not oxidize Ni(0) to Ni(I). A major concern was the
relative rate of the reduction of Ni(I) in comparison to
the rate of the rearrangement reaction. For this reac-
were found to have no effect on compound 1b.
3
. The yield of compound 4 for the Cp TiCl reaction is
2
approximately the same whether or not the Zn and
ZnCl2 are filtered off prior to the reaction of
tion to be successful, Cp*Fe must reduce Ni(I) to Ni(0)
2
faster than the Ni(I) catalyzes the rearrangement reac-
tion. Given the approximate relative rates of these two
reactions, which are based on the length of time each
reaction must stir to obtain a good yield (less than an 1
h for the reduction reaction and overnight for the
rearrangement reaction), we felt confident that the reac-
tion rates were in our favor.
Cp TiCl with compound 1b.
2
4
. The Ti(II) complex Cp Ti was found to be too
2
strong a reagent. When 1b was subjected to Cp Ti,
only intractable materials were observed.
To test whether compound 2b could be an intermedi-
ate in this conversion, 2b was put under the same
Ti(III) reaction conditions and was found to aromatize
faster than compound 1b converts to compound 4.
Thus, Ti(III) causes chemistry similar to that observed
with Ni(I) to occur; however, Ti(III) is not catalytic due
to the net oxidation of the cyclohexadiene to the aro-
matic product.
2
When compound 1b and either the butyl or the
phenyl
bis(phosphine)nickel
butene
complex
[
(R P) Ni(C H )] were mixed in THF and heated to
3
2
4
8
reflux overnight, compound 2b was obtained. In con-
trast, when the same reaction was run with the addition
of Cp*Fe, for the same or even for a longer period of
2
time, not even a trace of compound 2b or any other
product could be detected by gas chromatography. As a
control experiment, the reaction of compound 1b and
5
. Samarium-promoted rearrangement
Cp*Fe was run under similar conditions and again not
even a trace of compound 2b or any other product
could be detected.
Therefore, under reaction conditions such that all the
Ni(I) is reduced to Ni(0), no rearrangement reaction
occurs. Therefore, it was concluded that Ni(I) is causing
the rearrangement of 1 to 2. Simply put, Ni(0) is
nothing more than a source of Ni(I). (For a subsequent
section of this manuscript, it is important to note that
2
Next, we decided to try SmI and SmCl . In contrast
2
3
to the nickel and titanium cases in which a complex
with an odd number of electrons on the metal causes
the rearrangement to occur, when compound 1b was
treated with the odd electron Sm(III) complex, no
reaction was observed. However, when SmI and 1b
2
reacted at room temperature, aromatic compound 4
was obtained in an even higher yield than with tita-
nium. Thus, it seems that it is not whether the active
metal has an even or an odd number of electrons that
determines if it will cause the rearrangement reaction to
occur, but if the metal has a readily available next
higher oxidation state.
Cp*Fe does not cause a rearrangement reaction to
2
occur.)
4. Titanium-promoted rearrangement
We next decided to try another odd-electron species,
Ti(III), to see if it would cause a similar rearrangement
of compound 1b. Ti(III) was chosen because it is
known to initiate other radical-type reactions [9]. When
compound 1b was treated at room temperature with
6. Iron-promoted rearrangement
To further investigate the ‘next higher oxidation state
theory’, we needed another metal with a readily avail-

150
H.L. Holt, Jr. et al. / Journal of Organometallic Chemistry 601 (2000) 147–152
able pair of oxidation states in which the lower oxida-
tion state has an even number of electrons on the metal.
Although an obvious choice is Fe(II), as discussed
All IR spectra were recorded on a Perkin–Elmer
1600 Series FTIR spectrophotometer using KBr or
CaF2 cells. All NMR spectra were recorded on a
Bruker 250-MHz spectrometer and referenced to TMS
at 0.00 ppm, deuterated chloroform was the solvent
unless otherwise noted. Mass spectra were obtained on
a Hewlett–Packard 6890 GC/MS instrument using a
Supelco SPB-1 column and temperature programming.
above, Cp*Fe, which is an Fe(II) complex, does not
2
react with compound 1b. However, due to the reactivity
of the titanium and samarium halides, we decided to try
the iron halides anyway.
As expected, when compound 1b was treated with
FeCl3 at room temperature, no reaction occurred.
When the FeCl was treated with the activated zinc, to
3
8.2. Spectral data for compound 1b
generate Fe(II), prior to the addition of compound 1b,
compound 2b was formed in addition to small amounts
of two cyclohexadiene double bond isomers. (The reac-
1
H-NMR: l 1.50 (s, 3 H), 1.73 (s, 3 H), 2.55 (dd,
J=30, J=14.9 Hz, 2 H), 3.79 (s, 6 H), 4.78 (s, 1 H),
tion of 1,4-cyclohexadiene 2b with FeCl and Zn gener-
13
3
4.80 (s, 1 H); C-NMR: l 19.5, 21.9, 41.6, 50.1, 51.9,
10.8, 127.1, 127.2, 128.7, 188.3; IR (neat): 3021 (w),
2951 (s), 2870 (m), 1735 (vs), 1653 (m), 1431 (s), 1261
ates the exact same mixture of these three
cyclohexadiene isomers.) Thus, FeCl2 will cause the
1
rearrangement reaction to occur but not Cp*Fe. This
−1
2
(s), 1214 (s) 1068 (s) cm ; MS (m/z): 224 (0.8%), 192
32%), 177 (37%), 165 (20%), 133 (49%), 105 (100%),
suggests that the electron transfer must be an inner-
sphere process because the ferrocene derivatives are
known to undergo only outer sphere processes [10].
(
9
1, (73%), 77 (44%), 65 (34%), 59 (68%).
8.3. Spectral data for compound 2b
7
. Other evidence for an inner-sphere process and a
1
H-NMR: l 1.66 (s, 6 H), 2.88 (s, 4 H), 3.78 (s, 6 H);
proposed mechanism
13
C-NMR: l 28.1, 52.6, 66.1, 121.2, 132.6, 171.0; IR
(
THF): 2895 (s), 1740 (vs), 1437 (s), 1260 (s), 1043 (s)
There are a number of reported cases [11] in which
electrochemical reduction of a metal complex causes an
isomerization reaction to occur more easily. However,
this is not one of them. Previously [2], we have shown
that compound 1a is very difficult to reduce and that
when it is reduced electrochemically, only high molecu-
lar weight compounds are observed. Similarly, a potas-
sium naphthalide reduction of compound 1b only
generates intractable materials. Combining these results
with the decamethylferrocene result, we believe that an
inner-sphere electron transfer process is necessary for
the isomerization of 1 to 2. Other metal complexes with
readily available next higher oxidation states and which
can undergo an inner-sphere process, e.g. Cu(I) and
Cr(II), are under investigation.
−
1
cm ; MS (EI): 224 (1%), 192 (50%), 177 (80%), 106
33%), 91 (30%), 77 (100%), 71 (44%), 59 (66%).
(
8
.4. Spectral data for compound 4
1
H-NMR: l 1.74 (s, 6 H), 3.71 (s, 6 H), 7.77 (s, 2 H);
1
3
C-NMR: l 15.3, 50.1, 128.6, 130.2, 143.0, 167.1; IR
(
1
(
(
THF): 2993 (m), 1750 (vs), 1645 (s), 1455 (s), 1267 (s),
−
1
070 (s) cm ; MS (EI): 222 (3%), 207 (10%), 190
30%), 163 (90%), 105 (45%), 96 (29%), 75 (100%), 59
60%).
8.5. Preparation of acti6ated zinc
To zinc dust (4.0 g, 62 mmol) in a 50 ml Erlenmeyer
flask with a stir bar was added 10% sulfuric acid (20
ml), and this mixture was stirred. When the zinc formed
a sponge-like solid (after about 10 min), the sulfuric
acid was decanted off. The remaining zinc was washed
with water until the rinse water was no longer acidic.
The moist zinc was washed several times with distilled
THF to dry the zinc. The remaining THF was removed
by a stream of Ar gas or in vacuo. The zinc is now
activated, and should be stored as a solid in a glove box
or made into a slurry using freshly distilled solvent.
8
. Experimental
8
.1. General procedure
All reactions were carried out with oven-dried
(
120°C) glassware, under an argon atmosphere. Pho-
toreactions were carried out using a Rayonet photore-
actor employing 350 nm bulbs and Pyrex reaction
tubes. Solvents were purchased from Fisher Scientific
and were freshly distilled from potassium metal. The
complexes (Ph P) Ni(C H ), Cp TiCl, Cp Ti, and TiCl
8.6. Synthesis of 1b
3
2
4
8
2
2
3
were prepared by literature methods [4–7]. All other
reagents were purchased from Aldrich, Fluka, or Strem
and used without further purification.
To a thick-walled phototube was added dichloroma-
leic anhydride (1.87 g, 11.1 mmol) and benzophenone

H.L. Holt, Jr. et al. / Journal of Organometallic Chemistry 601 (2000) 147–152
151
(
0.15 g, 0.82 mmol) under an inert atmosphere. To the
Then, 2,3-dimethylbutadiene (1.64 g, 19.5 mmol) was
added to the tube. The tube was then sealed and the
solution was allowed to react for 6 days at r.t. The
contents of the tube were poured into a 100 ml flask,
and the solvent and unreacted diene were removed in
vacuo. Compound 2b was isolated (4.02 g, 17.8 mmol,
88% yield).
phototube was added THF or acetonitrile (21 ml). The
mixture was purged with argon for 10 min. To the
purged solution was added 2,3-dimethyl-1,3-butadiene
(
2.0 ml). The phototube was then sealed, placed in a
Rayonet reactor, and irradiated for 88 h. The volume
of the reaction solution was then reduced in vacuo.
Then 60°C water was added and the solution stirred for
10 min. The resulting mixture was made basic with 2N
8.8. Nickel-promoted rearrangement of 1b to 2b
sodium hydroxide, then extracted with ether (4×25
ml). The aqueous portion was made acidic with 10%
sulfuric acid, and extracted with ether (4×20 ml). The
combined ether layers were dried with magnesium sul-
fate, filtered, and concentrated to yield the diacid (2.47
g, 9.25 mmol) in 83% yield from the anhydride.
In a round-bottomed flask, equipped with a stir bar,
was added the newly formed diacid (2.47 g, 9.25 mmol).
The diacid was dissolved in DMF (21 ml), and then
reacted with potassium carbonate (5.31 g, 38.4 mmol).
After stirring for 5 min, methyl iodide (10.0 ml, 177
mmol) was added. The combined mixture was stirred at
room temperature for 30 h. The resulting mixture was
added to water (50 ml). The bottom viscous layer was
removed and saved as an organic layer. The remaining
layer was extracted with ether (2×20 ml). The com-
bined ether extracts were combined with the organic
layer which was removed initially, and washed with
brine (2×20 ml). The organic layer was dried with
anhydrous potassium carbonate, filtered, and concen-
trated, yielding a diester (1.56 g, 5.29 mmol) in 57%
yield.
In a three-neck round-bottomed flask containing a
stir bar was added bis(triphenylphosphine)nickel
dichloride (0.66 g, 1.0 mmol) and THF (30 ml). The
resulting slurry was cooled to ice-bath temperatures,
and butyllithium (1.6 M, 1.25 ml, 2.00 mmol) was
slowly added over the period of 1 min via a syringe.
The resulting solution was stirred at 0°C for 15 min.
Vinylcyclobutene (1b) (0.24 g, 1.1 mmol) was added to
the newly formed nickel(0) complex. The resulting solu-
tion was allowed to reflux 19 h. GC–MS analysis of
this solution showed the formation of compound 2b.
Alternatively, in a three-neck round-bottomed flask
containing a stir bar was added nickel dichloride
(NiCl ) (0.18 g, 1.4 mmol), THF (15 ml), and tri-n-
2
butylphosphine (0.7 ml, 2.5 mmol). This mixture was
allowed to stir at r.t. for 1.5 h (a slight modification of
the literature procedure [4]). The solution then was
cooled 0°C and butyllithium (1.6 M, 1.25 ml, 2.00
mmol) was added slowly over a period of 1 min via a
syringe. The resulting solution was stirred at 0°C for 10
min. The nickel–butene complex was then added via
cannula to compound 1b (0.305 g, 1.36 mmol) and the
mixture was refluxed. The reaction was monitored by
GC.
For the reactions with added decamethylferrocene,
the ferrocene (0.23 g, 7.0 mmol) and compound 1b were
mixed together prior to the addition of the nickel–
butene complex. For these reactions, GC–MS analysis
showed no formation of compound 2b or any other
new compound even when the solution was allowed to
reflux 39 h.
Five ml of 1,4-dioxane was added to a 10 ml pear-
shaped flask containing a magnetic spin bar. The
dichloro diester (0.25 g, 0.85 mmol) was added by
Pasteur pipet. The oil was stirred until it dissolved.
Argon gas was bubbled through the solution for 2 min.
The flask was then sealed with a rubber septum. Then,
0.50 ml (0.25 g, 3.8 mmol) of a zinc–hexane slurry was
added to the flask through the rubber septum. The
mixture was stirred for 5–6 h at r.t. The mixture was
allowed to settle without stirring for at least 30 min.
The liquid was removed from the flask via a syringe
and placed in a argon filled 2 dram vial. The vial was
spun down in a centrifuge for 2 min at high speed. The
liquid was transferred to a 5 ml flask and the solvent
was removed. The yield of compound 1b was nearly
quantitative (0.188 g, 0.84 mmol).
8.9. Reaction of titanium, samarium, iron, or potassium
naphthalide with compounds 1b and 2b
Ten millilters of 1,4-dioxane was added to a 25 ml
flask containing a magnetic stir bar. Either compound
1
b (0.25 g, 1.1 mmol) or compound 2b (0.25 g, 1.1
8
.7. Synthesis of authentic 2b
mmol) was added and stirred under an argon atmo-
sphere for 10 min to complete solution. The flask was
then sealed with a rubber septum. Two milliliters of
Forty milliliters of freshly distilled THF was placed
−
1
in an 100 ml Erlenmeyer flask with a magnetic stir bar.
Dimethyl acetylenedicarboxylate (2.88 g, 20.3 mmol)
was added and the mixture stirred until the acetylene
dicarboxylate was dissolved. The mixture was added to
a thick-walled phototube under an argon atmosphere.
zinc suspension (0.2 g ml , 0.4 g, 6 mmol) in hexane,
as prepared above, was added by syringe to the flask.
Biscyclopentadienyl titanium dichloride (Cp TiCl )
2
2
(0.25 g, 1.0 mmol) dissolved in 2 ml of 1,4-dioxane was
added by syringe in the same way as the zinc suspen-

152
H.L. Holt, Jr. et al. / Journal of Organometallic Chemistry 601 (2000) 147–152
sion. When commercially available TiCl , SmI , or
FeCl (0.20 g, 1.2 mmol), yield=no reaction; 1b with
K C H (0.30 g, 1.8 mmol), yield=intractable solids.
10 8
3
2
3
+
−
SmCl , were used or when potassium naphthalide was
3
used, the solution was removed from the stock bottle
using cannula techniques to exclude air. The zinc addi-
tion was eliminated. When FeCl was used, the zinc
Acknowledgements
3
step was not included for the iron(III) reaction and was
included for the iron(II) reaction. The mixture was
allowed to react for 240 h at r.t. with constant stirring.
Either samples were taken at 48 h intervals for GC–MS
monitoring, or at the end of the reaction, the solution
was worked-up by adding 20 ml of water and then
extracting the water with ether. The ether extracts were
combined, dried with Na SO , and the solvent was
The authors are deeply indebted to Professor James
Hershberger of Miami University for many valuable
discussions.
References
2
4
[
[
1] D. DiFrancesco, A.R. Pinhas, J. Org. Chem. 51 (1986) 2098.
2] H. Choi, J.W. Hershberger, A.R. Pinhas, D.M. Ho, Organo-
metallics 10 (1991) 2930.
removed.
8
.10. Amounts and yields for the abo6e reactions
[3] H. Choi, A.R. Pinhas, Organometallics 11 (1992) 442.
[
[
4] (a) S. Saito, M. Sakai, N. Miyaura, Tetrahedron Lett. 37 (1996)
993. (b) B.H. Lipshutz, S. Kim, P. Mollard, K.L. Stevens,
Tetrahedron 54 (1998) 1241.
5] T.V. RajanBabu, W.A. Nugent, J. Am. Chem. Soc. 116 (1994)
2
1
b with TiCl (1.0 M, 2.0 ml, 2.0 mmol), yield=
3
0.190 g, 0.86 mmol, 78%; 2b with TiCl (1.0 M, 2.0 ml,
3
2
.0 mmol), yield=0.179 g, 0.81 mmol, 74%; 1b with
9
86.
Cp TiCl (0.20 g, 0.80 mmol), yield=no reaction; 2b
[6] N. Mezailles, N. Avarvari, D. Bourissou, F. Mathey, P. Le
Floch, Organometallics 17 (1998) 2677.
2
2
with Cp TiCl (0.20 g, 0.80 mmol), yield=no reaction;
2
2
[
[
[
7] W. Chen, J. Zhang, M. Hu, X. Wang, Synthesis (1990) 701.
8] J.E. Shaw, D.C. Kunerth, J. Org. Chem. 39 (1974) 1968.
9] See for example: (a) S. Ogoshi, J.M. Stryker, J. Am. Chem. Soc.
120 (1998) 3514. (b) R.P. Spencer, J. Schwartz, J. Org. Chem. 62
(1997) 4204. (c) A. Gansauer, D. Bauer, J. Org. Chem. 63 (1998)
1
b with SmI (1.0 M, 1.0 ml, 1.0 mmol), yield=0.193 g,
2
0
.87 mmol, 79%; 2b with SmI (1.0 M, 1.0 ml, 1.0
2
mmol), yield=0.189 g, 0.85 mmol, 77%; 1b with SmCl3
(
0.2 g, 0.78), yield=no reaction; 2b with SmCl (0.2 g,
3
2
070.
0
.78), yield=no reaction; 1b with FeCl (0.20 g, 1.2
3
[
[
10] A.R. Pinhas, J.W. Hershberger, Organometallics 9 (1990) 2840.
11] See for example: (a) J.M. Mevs, T. Gennett, W.E. Geiger,
Organometallics 10 (1991) 1229. (b) W.E. Geiger, M.J. Shaw, M.
Wunsch, C.E. Barnes, F.H. Foersterling, J. Am. Chem. Soc. 119
(1997) 2804.
mmol), Zn (0.20 g, 3.1 mmol), yield=0.133 g, 0.60
mmol, 55%; 2b with FeCl (0.20 g, 1.2 mmol), Zn (0.20
3
g, 3.1 mmol), yield=0.25 g, 1.1 mmol, 100%; 1b with
FeCl (0.20 g, 1.2 mmol), yield=no reaction; 2b with
3
.