Kinetics and mechanism of cyclopropanation of cyclooctene by [Fe(η5-C5H5)(CO)2CH 2SPh2]BF4

Article abstract of DOI:10.1021/om980472a

The kinetics of reaction of [Fe(η⁵-C₅H₅)(CO)₂CH ₂SPh₂]BF₄ with cyclooctene in CH₂Cl₂ solution have been studied under pseudo-first-order conditions at room temperature. The variation of k_obs is nonlinear in both cyclooctene and added Ph₂S. The results are interpreted in terms of a two-step reaction in which the cation [Fe(η⁵-C₅H₅)(CO)₂CH ₂SPh₂]⁺ undergoes reversible, dissociative loss of Ph₂S followed by competitive capture of the methyleneiron complex by cyclooctene to produce bicyclo[6.1.0]nonane. The methyleneiron intermediate is 4-5 times more reactive toward diphenylsulfide than cyclooctene. In the absence of added Ph₂S, millimolar solutions of the iron complex react with cyclooctene at room temperature with a half-life of about 36 min to give >85% yields of bicyclo[6.1.0]nonane after 3 h, and it is thus more reactive than the dimethyl- and methylphenylsulfonium salt analogues.

Full text of DOI:10.1021/om980472a

Organometallics 1998, 17, 4645-4648
4645
Kin etics a n d Mech a n ism of Cyclop r op a n a tion of
Cycloocten e by [F e(η5-C5H5)(CO)2CH2SP h 2]BF 4
Paul McCarten and E. Kent Barefield*,†
School of Chemistry and Biochemistry Georgia Institute of Technology,
Atlanta, Georgia 30332-0400
Received J une 8, 1998
The kinetics of reaction of [Fe(η5-C5H5)(CO)2CH2SPh2]BF4 with cyclooctene in CH2Cl2
solution have been studied under pseudo-first-order conditions at room temperature. The
variation of kobs is nonlinear in both cyclooctene and added Ph2S. The results are interpreted
in terms of a two-step reaction in which the cation [Fe(η5-C5H5)(CO)2CH2SPh2]+ undergoes
reversible, dissociative loss of Ph2S followed by competitive capture of the methyleneiron
complex by cyclooctene to produce bicyclo[6.1.0]nonane. The methyleneiron intermediate
is 4-5 times more reactive toward diphenylsulfide than cyclooctene. In the absence of added
Ph2S, millimolar solutions of the iron complex react with cyclooctene at room temperature
with a half-life of about 36 min to give >85% yields of bicyclo[6.1.0]nonane after 3 h, and it
is thus more reactive than the dimethyl- and methylphenylsulfonium salt analogues.
In tr od u ction
salts.6 Two limiting pathways can be envisioned: an
SN1 pathway in which dissociation of sulfide produces
a carbene, or an SN2 pathway in which the sulfide is
displaced by the olefin. In principle, the latter pathway
might allow the use of a chiral sulfonium salt as a
means of directing the stereochemistry of cyclopropa-
nation of a prochiral substrate7,8 whereas the former
would not.
In connection with our investigation of thallium(I)-
promoted substitution of chloride in chloromethyl alkyl
complexes by neutral nucleophiles9 we generated sev-
eral new [CpFe(CO)2CH2SR2]+ salts including [CpFe-
(CO)2CH2SPh2]BF4. Not unexpectedly, the diphenyl-
sulfonium salt proved to be considerably more reactive
for cyclopropanation of olefins than either the dimethyl-
or methylphenylsulfonium salts studied by Helquist.1
The complex reacted with 1 equiv of C8H14 in methylene
chloride (0.1 M) at room temperature (t1/2 ∼36 min) to
give >85% of bicyclo[6.1.0]nonane after 3 h. The
convenient time scale of this reaction led us to perform
a detailed kinetics study, which is the primary subject
A variety of molecular complexes have been utilized
as reagents for conversion of alkenes to cyclopropanes.
One class of reagents developed by Helquist for this
purpose are salts of the cations {Fe(η5-C5H5)(CO)2[CH-
(R1)SR2R3]}+.1 The salt [FeCp(CO)2CH2S(CH3)2]BF4
has been most extensively studied.2 It is readily
prepared in excellent yield from inexpensive starting
materials, has an extensive shelf life, and under ap-
propriate conditions, converts olefins to cyclopropanes
in moderate to good yields with high stereoselectivity.
Unlike related cyclopropanating reagents, such as those
generated by induced R-ionization of {Fe(η5-C5H5)(CO)-
(L)[C(X)RR′]}3 or protonation of [Fe(η5-C5H5)(CO)(L)C-
(R)dCR2],4 for which there is good evidence that a
metal-carbene complex is generated in at least some
instances,5 there is no direct evidence currently avail-
able concerning the pathway followed by the sulfonium
† E-mail address: kent.barefield@chemistry.gatech.edu. Fax: (404)
894-7452.
(1) An exhaustive review of Helquist’s work on these reagents and
of other methods of cyclopropanation of alkenes up to 1987 is given
in: O’Connor, E. J .; Brandt, S.; Helquist, P. J . Am. Chem. Soc. 1987,
109, 3739.
(2) This compound is currently available from Aldrich Chemical Co.,
and its preparation and procedures for cyclopropanation have appeared
in: Organic Syntheses; Mattson, M. N.; O’Connor, E. J .; Helquist, P.
Org. Synth. 1992, 70, 177.
(5) Evidence for the formation of “free” metal-carbenes and a
discussion of their reactions is given in: (a) Brookhart, M.; Studabaker,
W. B. Chem. Rev. 1987, 87, 411, and (b) Petz, W. Iron-Carbene
Complexes; Springer-Verlag: Berlin, 1993. [CpFe(CO)2CH2]+ has never
been spectroscopically detected. When generated in the absence of a
trapping agent, the ion undergoes disproportionation to give [CpFe-
(CO)2(C2H4)]+ and “[CpFe(CO)2]+” (ref 3b and references cited).
(6) A bimolecular process was initially suggested for the reaction of
olefins with {Fe(Cp)(CO)2[CH(CH3)S(CH3)Ph]}+ (Kremer, K. A. M.;
Helquist, P. J . Organomet. Chem. 1985, 285, 231), but see ref 1 for
alternative mechanistic considerations.
(7) The efficiency of such a process would likely be limited, although
the extent of bond breaking in the transition state would be very
important. It would also be inefficient in that the chiral center would
be lost upon displacement of the sulfide.
(8) A highly enantioselective cyclopropane synthesis has been ac-
complished using chiral forms of [CpFe(CO)(PR3)dCHCH3]+: Brookhart,
M.; Liu, Y.; Goldman, E. W.; Timmers, D. A.; Williams, G. D. J . Am.
Chem. Soc. 1991, 113, 927. References to earlier less successful
attempts at enantioselective cyclopropane synthesis using chiral [CpFe-
(CO)(PR3)dCH2]+ are cited.
(3) Such reactions include: (1) Ag+ ion induced ionization of
CpFe(CO)2CH2Cl (a) J olly, P. W.; Pettit, R. J . Am. Chem. Soc. 1966,
88, 5044, (b) Bodnar, T. W.; Cutler, A. R. Organometallics 1985, 4,
1558. (2) H+ reaction with CpFe(CO)(L)C(OR)R′R′′; ref 3a; (c) Brookhart,
M.; Nelson, G. O. J . Am. Chem. Soc. 1977, 99, 6099, (d) Green, M. L.
H.; Ishag, M.; Whiteley, R. N. J . Chem. Soc., A 1967, 1508. (3) Ph3C+
abstraction of RO- from CpFe(CO)(L)C(OR)R′R′′, ref 3c (this is not a
general reaction). (4) TMSOTf silylation of CpFe(CO)(L)C(OR)R′R′′, R
) alkyl: (e) Brookhart, M.; Humphrey, M. B.; Kratzer, H. J .; Nelson,
G. O. J . Am. Chem. Soc. 1980, 102, 7802, R ) Me3Si, (f) R. M.; Theys,
R. D.; Hossain, M. M. J . Am. Chem. Soc. 1992, 114, 777.
(4) (a) Casey, C. P.; Miles, W. H.; Tuikada, H.; O’Connor, J . M. J .
Am. Chem. Soc. 1982, 104, 3761. (b) Kremer, K. A. M.; Kuo, G. H.;
O’Connor, E. J .; Helquist, P.; Kerber, R. C. J . Am. Chem. Soc. 1982,
104, 9. (c) Kuo, G. H.; Helquist, P.; Kerber, R. C. Organometallics 1984,
3, 806. (d) Casey, C. P.; Miles, W. H.; Tukada, H. J . Am. Chem. Soc.
1985, 107, 2924.
(9) Barefield, E. K.; McCarten, P.; Hillhouse, M. C. Organometallics
1985, 4, 1682.
S0276-7333(98)00472-5 CCC: $15.00 © 1998 American Chemical Society
Publication on Web 09/12/1998

4646 Organometallics, Vol. 17, No. 21, 1998
McCarten and Barefield
Ta ble 1. Yield s a n d Ha lf-Lives for
Cyclop r op a n a tion of Va r iou s Olefin s by
[Cp F e(CO)2CH2SP h 2]BF 4 a t Room Tem p er a tu r e
of this report. Also reported are some qualitative
investigations of the reactivity of the diphenylsulfonium
complex toward a number of olefins with substituents
that were either more or less electron donating than
alkyl.
a
olefin
cyclopropane/yield, %
t1/2, h
cyclooctene
styrene
bicyclo[6.1.0]nonane/92
0.6
1
4
cyclopropylbenzene18/82
trans-stilbene
trans-1,2-diphenylcyclopropane19/98
Exp er im en ta l Section
2,3-dihydropyran 1-oxabicyclo[4.1.0]heptane20/75
0.25
methyl acrylate
trichlorethene
no reaction
no reaction
All synthetic operations and manipulations of solutions of
organometallics were conducted under a nitrogen atmosphere.
Methylene chloride and hexane were purified by standard
methods.10 Diphenylsulfide (Aldrich Chemical Co.) was dis-
tilled from sodium under vacuum. Cyclooctene (Aldrich
Chemical Co.) was passed through a column of alumina and
distilled from calcium hydride. All other olefinic substrates
(all from Aldrich Chemical Co.) were used as received. Infra-
red spectra were recorded using 0.1 mm NaCl cells and a
Beckman 4240 spectrophotometer. Gas chromatographic analy-
ses were performed at 140 °C using a 10% OV-101 on
Chromosorb G (1 m × 1/8 in.) column and a thermal conduc-
tivity detector.
P r ep a r a tion of [Cp F e(CO)2CH2SP h 2]BF 4. A solution of
4.0 mL of Ph2S (4.5 g, 24 mmol) and 1.20 g (5.3 mmol) of
CpFe(CO)2CH2Cl in 40 mL of CH2Cl2, which was prepared and
maintained under a nitrogen atmosphere, was combined with
2.70 g (10.0 mmol) of TlBF4, and the resulting slurry was
stirred for 20 h. The insoluble thallium salts were removed
by filtration and washed with CH2Cl2. The combined meth-
ylene chloride solutions were concentrated to a small volume
(ca. 15 mL), and hexane was added to complete precipitation
of the yellow-orange product. The solid was redissolved in a
minimum amount of methylene chloride and crystallized by
slow addition of hexane. The product was collected by filtra-
tion and dried under vacuum to give a nearly quantitative
yield of product (2.4 g). Anal. Calcd for C20H17BF4FeO2: C,
51.76; H, 3.69. Found: C, 51.92; H, 3.76. IR (CH2Cl2): νCO
2034, 1990 cm-1. NMR (acetone-d 6): δ 5.31 (s, 5H, C5H5), 3.53
(s, 2H, CH2), 7.52-7.75 (m, 6H, m,p-C6H5), 7.93-8.17 (m, 4H,
o-C6H5).
Cyclop r op a n a t ion of Cyclooct en e b y [Cp F e(CO)2-
CH2SP h 2]BF 4. To a solution of 0.250 g of [CpFe(CO)2CH2-
SPh2]BF4 (0.539 mmol) in 5 mL of methylene chloride was
added 0.0594 g (0.540 mmol) of C8H14 and 0.0791 g (0.468
mmol) of diphenylmethane as an internal standard for gas
chromatography. The mixture was stirred at room tempera-
ture for 3 h, at which time an infrared spectrum of the CO
stretching region indicated that the peaks due to starting
material (2034, 1990 cm-1) had been replaced by new absorp-
tions at 2069 and 2027 cm-1 (due to [CpFe(CO)2(SPh2)]+, vide
infra). The inorganic material was precipitated by the addition
of hexane, and the supernatant was concentrated and analyzed
by gas-liquid chromatography. Bicyclo[6.1.0]nonane (80%
yield based on iron) and 1-methylcyclooctene (15%) were the
only products detected.11 Washing of the supernatant with
aqueous sodium bicarbonate and drying over sodium sulfate
prior to concentration as suggested by Brookhart12 gave
increased yields of bicyclononane (up to 92%) and smaller
amounts of 1-methylcyclooctene (as low as 3%). Survey
experiments indicated that the rate of disappearance of the
iron complex increased at higher concentrations of C8H14, but
decreased sharply when Ph2S was added to the reaction
mixture.
a
Approximate half-life for disappearance of iron complex; based
upon infrared monitoring of the reaction mixture.
Cyclop r op a n a tion of Oth er Olefin s w ith [Cp F e(CO)2-
CH2SP h 2]BF 4. Reactions were done by combining 1.0 mmol
of the olefin with 1.0 mmol of the iron complex in 5 mL of
methylene chloride at room temperature. After the iron
starting material had disappeared (determined by infrared
spectroscopy), 2 g of alumina (Brockman Activity 1) was added
to the reaction mixture, the solvent was evaporated, and the
alumina was transferred to the top of a 20 cm × 2 cm alumina
column. The organic products were isolated by elution with
100 mL of 2-methylbutane. The elutant was evaporated, the
residue was weighed, and the products were identified by
comparison of their NMR spectra with data recorded in the
literature. The olefins tested, approximate half-lives for
reaction, and yield of cyclopropanation product are given in
Table 1.
Kin etics of th e Rea ction of [Cp F e(CO)2CH2SP h 2]BF 4
w ith Cyclocten e. The infrared spectrum of a solution of
0.0374 g (0.081 mmol) of [CpFe(CO)2CH2SPh2]BF4 in 1.00 mL
of methylene chloride gave an absorbance of 0.73 for the CO
stretch at 1990 cm-1, which decreased linearly with concentra-
tion as the sample was diluted. The CO absorptions slowly
decreased in intensity as solutions aged, but this was com-
pletely inhibited by the addition of small amounts of Ph2S.
For kinetic runs samples of [CpFe(CO)2CH2SPh2]BF4 were
weighed into Schlenk vessels, which were then flushed with
nitrogen and further charged with measured amounts of
methylene chloride and a standard solution of diphenylsulfide.
The vessel was capped with a septum and placed in a constant-
temperature bath (22.4 ( 0.2 °C). A measured amount of a
standard solution of C8H14 in methylene chloride thermostated
to the same temperature was added, and the disappearance
of the 1990 cm-1 carbonyl absorption monitored on samples
taken at intervals from the reaction vessel. The final concen-
tration of the iron complex was 0.0434 M after addition of the
cyclooctene solution. Three trials were run at five concentra-
tions of C8H14 (10, 15, 20, 30, 40 equiv relative to complex),
each in the presence of 10 equiv of added Ph2S.13 Two trials
were run at four additional concentrations of Ph2S (7.5, 15,
20, 30 equiv relative to complex), each in the presence of 20
equiv of C8H14
.
Plots of ln(At - A∞) vs t were linear for at least 3 half-lives
for each reaction. A∞ was determined after 16 h. Values of
kobs were obtained from the slopes of these plots; kobs values
from duplicate runs were averaged14 to prepare Figures 1 and
2.
Id en tifica tion of Ir on -Con ta in in g P r od u cts of Cyclo-
p r op a n a tion Rea ction s. The observation of a high-fre-
quency pair of carbonyl absorptions (ca. 2060 and 2030 cm-1
)
for some spent reaction mixtures suggested that a cationic
[CpFe(CO)2L]+ species was formed during the cyclopropanation
reaction. A yellow crystalline solid precipitated from the
reaction of 40 equiv of C8H14 with [CpFe(CO)2CH2SPh2]BF4
(10) Perrin, D. D.; Armarego, W. L. F.; Perrin, D. R. Purification of
Laboratory Chemicals; Pergamon Press Ltd: Oxford, 1966.
(11) Identified by comparison of their 1H NMR spectra with those
of authentic materials. Bicyclo[6.1.0]nonane, prepared according to the
procedure given by Simmons, H. E.; Cairns, T. L.; Vladuchick, S. A.;
Hoiness, C. M. Org. React. 1973, 20, 1, was provided by J ames G.
Davidson. 1-Methylcyclooctene was obtained from Aldrich Chemical
Co.
(13) The actual concentrations of diphenylsulfide are less than the
number of equivalents indicated by a factor of 0.424/0.434. The correct
molar concentrations were used in all calculations of rate constants.
(14) The rate constants (units of s-1 × 104) used in constructing
Figures 1 and 2 are 0.89 ( 0.09, 1.21 ( 0.09, 1.25 ( 0.07, 1.58 ( 0.02,
1.66 ( 0.04, 2.03 ( 0.12, 2.50 ( 0.01, 2.93 ( 0.17, and 3.35 ( 0.09.
(12) Brookhart, M.; Tucker, J . R.; Husk, G. R. J . Am. Chem. Soc.
1983, 105, 258.

Cyclopropanation of Cyclooctene
Organometallics, Vol. 17, No. 21, 1998 4647
Resu lts
P r elim in a r y Cyclop r op a n a tion Stu d ies. [CpFe-
(CO)2(SPh2)]BF4 reacted with 1 equiv of C8H14 (ca. 0.1
M in methylene chloride) with a half-life of about 36
min to give up to 92% bicyclo[6.1.0]nonane and a small
amount of 1-methylcyclooctene after suitable workup.
[CpFe(CO)2(SPh2)]+ was the major iron-containing prod-
uct. Survey experiments indicated that the rate of
disappearance of the iron complex increased at higher
cyclooctene/complex ratios, but decreased markedly
when Ph2S was added to the reaction mixture.
Kin etics of Cyclop r op a n a tion of Cycloocten e by
[CpFe(CO)2CH2SP h 2]BF4. The observations described
above suggest the following reaction sequence:
[CpFe(CO)2CH2SPh2]1+ yk1z
k-1
[CpFe(CO)2CH2]1+ + SPh2 (1)
F igu r e 1. Plots of 1/kobs vs [diphenylsulfide]/[cyclooctene]
where 1/kobs) k-1[Ph2S]/k1k2[C8H14] + 1/k1. The intercept
is 1/k1, and the slope is proportional to k-1/k2.
k2
[CpFe(CO)2CH2]1+ + C8H14
98
[CpFe(CO)2]1+ + C9H16 (2)
[CpFe(CO)2]1+ + L fast8 [CpFe(CO)2L]1+
(3)
where L in the third step is either diphenylsulfide or
cyclooctene. If it is assumed that reaction 2 is irrevers-
ible, that reaction 3 is fast compared to 1 and 2, and
that the steady-state approximation holds for the con-
centration of [CpFe(CO)2CH2]1+, then the rate law for
the reaction is:
k1k2[complex][C8H14]
k-1[Ph2S] + k2[C8H14]
-d[complex]
dt
)
In this expression complex refers to [CpFe(CO)2CH2-
SPh2]1+. Defining kobs ) k1k2[C8H14]/(k-1[Ph2S] + k2-
[C8H14]) allows for the extraction of values for both k1
and k-1/k2 by plotting 1/kobs vs the ratio [Ph2S]/[C8H14
]
when the reaction is examined under pseudo-first-order
conditions varying either olefin or sulfide while holding
the other constant (both C8H14 and Ph2S in excess).
Figure 1 shows the dependence of 1/kobs on the ratio
[Ph2S]/[C8H14]. The value of k1 obtained from the
intercept is 7.0 ( 0.6 × 10-4 s-1, and the value of k-1/k2
obtained from the slope is 4.8 ( 0.5.16 When k-1[Ph2S]
≈ k2[C8H14], the dependence of the rate on [C8H14] is
nonlinear and added diphenylsulfide will retard the
reaction. Figure 2 shows the effect on kobs of increasing
the concentration of either cyclooctene (in the presence
of 10 equiv of diphenylsulfide) or diphenylsulfide (in the
presence of 20 equiv of cyclooctene).
F igu r e 2. Plot of kobs for cyclopropanation of cyclooctene
by [Fe(η5-C5H5)(CO)2CH2SPh2]+ in the presence of added
diphenylsulfide: (b) as a function of [C8H14] in the presence
of 10 equiv of added Ph2S; (O) as a function of [Ph2S] in
the presence of 20 equiv of C8H14. The value of kobs of 2.03
is common to both plots. The lines are curves calculated
using the expression for kobs and the rate constants
determined in Figure 1.
in the presence of 10 equiv of Ph2S after allowing the reaction
mixture to stand for 24 h. This compound was filtered, washed
with 2-methylbutane, and dried in vacuo. It was identified
as [CpFe(CO)2(cyclooctene)]BF4 by comparison of its infrared
spectrum (CO ) 2060, 2035 cm-1) and 1H NMR spectra with
the literature values.15 A yellow crystalline solid precipitated
from the reaction of 20 equiv of C8H14 with [CpFe(CO)2CH2-
SPh2]BF4, in the presence of 30 equiv of Ph2S after 48 h. This
solid was filtered, washed with 2-methylbutane, and dried in
vacuo. The product was formulated as [CpFe(CO)2(SPh2)]BF4
In the iron product [CpFe(CO)2L]+ L is either cy-
15
clooctene or Ph2S.17 [CpFe(CO)2(C8H14)]BF4 was iso-
lated from the reaction with 40 equiv of C8H14 in the
presence of 10 equiv of Ph2S, whereas [CpFe(CO)2-
(SPh2)]BF4 was isolated from the reaction with 20 equiv
1
on the basis of its H NMR spectrum: 1H NMR (acetone-d6) δ
7.48-7.65 (m, 10H, Ph2S), 5.80 (s, 5H, η5-C5H5). The CO
(16) Errors in k1 and k-1/k2 are derived from the estimated standard
errors of the intercept and slope from regression analysis (SigmaPlot).
The value for k-1/k2 is dimensionless, as it is the ratio of second-order
rate constants.
stretching absorptions (CH2Cl2 solution) were at 2069 and 2027
cm-1
.
(15) Cutler, A.; Ehntholt, D.; Giering, W. P.; Lennon, P.; Raghu, S.;
Rosan, A.; Rosenblum, M.; Tancrede, J .; Wells, D. J . Am. Chem. Soc.
1976, 98, 3495.
(17) Helquist reports isolation of [CpFe(CO)2SMe2]+ from spent
cyclopropanation reactions involving [CpFe(CO)2CH2SMe2]+; see refer-
ence 1.

4648 Organometallics, Vol. 17, No. 21, 1998
McCarten and Barefield
of C8H14 and 30 equiv of Ph2S. Infrared spectra of
reaction mixtures suggest that both of these complexes
are formed in some cases, but no effort was made to
establish the relative stabilities of these two complexes
or to determine their relative solubilities.
as long as [CpFe(CO)2CH2Cl] is required as an inter-
mediate in the preparation of the diphenyl derivative,
it is not likely to see much application except possibly
in cyclopropanation of thermally sensitive substrates.
The kinetics of the reaction of [CpFe(CO)2CH2SPh2]+
with cyclooctene are entirely consistent with a two-step
mechanism, with reversible dissociative loss of Ph2S
occurring in the first step. Most likely the other
sulfonium salts react in a similar fashion. Helquist’s
observation that addition of [Cu(NCCH3)4]+ increased
the rate of reaction of [CpFe(CO)2CH2SMe2]+ with
Su r vey of Rea ctivity of [Cp F e(CO)2CH2SP h 2]BF 4
in Cyclop r op a n a tion of Olefin s. The reactivity of
the iron complex was surveyed in a qualitative fashion
for cyclopropanation of some other olefins besides cy-
clooctene. Half-lives and yields of cyclopropanation
products for the olefins surveyed are given in Table 1.
The reactivity pattern is similar to that observed by
Helquist and co-workers for their sulfonium salts.1
None of the olefins chosen for our studies provide
information concerning the stereoselectivity of addition
of the methylene to the double bond. One item of
interest is the reasonable yield of the cyclopropanation
product derived from 2,3-dihydropyran. Although it is
not clear if dihydropyran was examined in earlier
studies with other sulfonium salts, it has been reported
that whereas their reactivity is high toward electron-
rich olefins, the anticipated cyclopropanes were gener-
ally not observed.1
1
Ph2CdCH2 is most likely a result of trapping of
dissociated Me2S and thus eliminating its competition
for the carbene.
Assuming that it is the free iron methylene carbene
that reacts with the olefin for all of the sulfonium salts,
their similar reactivity pattern toward substituted
olefins is predictable. The fact that these sulfonium salt
reagents can be used to prepare cyclopropanes in high
yields without competitive formation of the dispropor-
tionation products [CpFe(CO)2C2H4]+ and [CpFe(CO)2L]+,
which occurs when the methylene complex is generated
rapidly by other methods, is surely because the dispro-
portionation reaction is second order.22 Thus when the
rate of formation of the methylene species is low and a
suitable trapping agent is present, its concentration
never builds up sufficiently for disproportionation to
become a competitive process.
Discu ssion
As expected, based upon Helquist’s results with the
[CpFe(CO)2CH2S(Me)R]+ salts (R ) Me, Ph),1 the diphe-
nylsulfonium analogue is more reactive toward nucleo-
philes than either of these salts. Although equally good
yields of cyclopropanation products can be obtained with
any one of the three reagents, the dimethyl derivative
requires reflux temperatures,21 whereas the diphenyl
derivative reacts at room temperature. Unfortunately,
Ack n ow led gm en t. We would like to thank H. M.
Neumann for helpful discussions on the analysis of
kinetics data.
OM980472A
(21) The best solvent for synthetic applications of [CpFe(CO)2CH2-
SMe2]BF4 appears to be nitromethane; see ref 1 and also Ba¨ckvall,
J .-E.; Lo¨fstro¨m, C.; J untunen, S. K.; Mattson, M. Tetrahedron Lett.
1993, 34, 2007.
(18) Adams, W.; Birke, A.; Cadiz, C.; Diaz, S.; Rodriquez, A. J . Org.
Chem. 1978, 43, 1154.
(19) Casey, C. P.; Polichnowski, S. W.; Schusterman, A. J .; J ones,
C. R. J . Am. Chem. Soc. 1979, 101, 7282.
(20) Friedrich, E. C.; Domek, J . M.; Pong, R. Y. J . Org. Chem. 1985,
50, 4640.
(22) The kinetics of disproportionation of [CpFe(CO)2CH2]+ have not
been studied. However, the disproportionation of [CpRe(PPh3)(NO)-
CH2]+ is second order: Merrifield, J . H.; Lin, G.-Y.; Kiel, W. A.;
Gladysz, J . A. J . Am. Chem. Soc. 1983, 105, 5811.