A new mechanism of nucleophilic vinylic substitution and unexpected products in the reaction of chlorotrifluoroethylene with [Re(CO)5]Na and [CpFe(CO)2]K

Article abstract of DOI:10.1016/j.jorganchem.2005.12.050

The mechanism of chloride substitution in CF₂{double bond, long}CFCl with [Re(CO)₅]^- and [CpFe(CO)₂]^- anions is investigated experimentally and theoretically. The substitution reaction begins with the nucleophile addition to CF₂{double bond, long}CFCl producing the carbenoid anion [MCF₂CFCl]^- (A) (M = Re(CO)₅, CpFe(CO)₂). This is shown by trapping the intermediate A with electrophiles - proton donor (t-BuOH) to give MCF₂CFClH or with CF₂{double bond, long}CFRe(CO)₅ to give acylmetallate III, and by the formation of the substitution products CF₂{double bond, long}CFM from the anion A, generated by the deprotonation of MCF₂CFClH with t-BuOK. 1,2-Shift of metal carbonyl group concerted with the α-elimination of chloride anion is proposed as the transformation pathway of carbenoid A into CF₂{double bond, long}CFM. A competing process of carbene insertion into Fe-C{triple bond, long}O bond is proposed to explain the formation of {A figure is presented} (XI). The feasibility of these two pathways is confirmed by DFT (B3LYP/SDD and 6-31G^*) calculations of the carbenes [MCF₂CF:] and carbenoid anions [MCF₂CFCl]^-. Transition states (TS) for 1,2-shift (+3.2 kcal/mol) and for nucleophilic addition at C{triple bond, long}O ligand (+5.4 kcal/mol) are located for [(CO)₅ReCF₂CFCl]^-, but only one TS corresponding to carbene insertion into Fe-C{triple bond, long}O bond (+2.1 kcal/mol) is located for [(CO)₂CpFeCF₂CFCl]^-. The formation of other newly observed products, F(CO)CHFRe(CO)₅ (V) and Cp(CO)₂FeC{triple bond, long}CFeCp(CO)₂ (VIII) is also discussed.

Full text of DOI:10.1016/j.jorganchem.2005.12.050

Journal of Organometallic Chemistry 691 (2006) 2346–2357
www.elsevier.com/locate/jorganchem
A new mechanism of nucleophilic vinylic substitution and
unexpected products in the reaction of chlorotrifluoroethylene
with [Re(CO)5]Na and [CpFe(CO)2]K
a,
a
b
a,
*
P.K. Sazonov *, G.A. Artamkina , K.A. Lyssenko , I.P. Beletskaya
a
Moscow State University, Chemistry Department, Leninskie Gory 1, GSP-2, Moscow 119992, Russian Federation
A.N. Nesmeyanov Institute of Organoelement Compounds (INEOS), Russian Academy of Sciences, Vavilov Street 28, Moscow 119991, Russian Federation
b
Received 27 July 2005; received in revised form 13 December 2005; accepted 23 December 2005
Available online 28 February 2006
Abstract
The mechanism of chloride substitution in CF2@CFCl with [Re(CO)5]À and [CpFe(CO)2]À anions is investigated experimentally and
theoretically. The substitution reaction begins with the nucleophile addition to CF2@CFCl producing the carbenoid anion
[MCF2CFCl]À (A) (M = Re(CO)5, CpFe(CO)2). This is shown by trapping the intermediate A with electrophiles – proton donor (t-
BuOH) to give MCF2CFClH or with CF2@CFRe(CO)5 to give acylmetallate III, and by the formation of the substitution products
CF2@CFM from the anion A, generated by the deprotonation of MCF2CFClH with t-BuOK. 1,2-Shift of metal carbonyl group con-
certed with the a-elimination of chloride anion is proposed as the transformation pathway of carbenoid A into CF2@CFM. A competing
process of carbene insertion into Fe–C„O bond is proposed to explain the formation of
(XI). The feasibility
Cp(CO)FeCF=CFC(O)OBu-t
of these two pathways is confirmed by DFT (B3LYP/SDD and 6-31G*) calculations of the carbenes [MCF2CF:] and carbenoid anions
[MCF2CFCl]À. Transition states (TS) for 1,2-shift (+3.2 kcal/mol) and for nucleophilic addition at C„O ligand (+5.4 kcal/mol) are
located for [(CO)5ReCF2CFCl]À, but only one TS corresponding to carbene insertion into Fe–C„O bond (+2.1 kcal/mol) is located
for [(CO)2CpFeCF2CFCl]À. The formation of other newly observed products, F(CO)CHFRe(CO)5 (V) and Cp(CO)2FeC„CFeCp(CO)2
(VIII) is also discussed.
ꢀ 2006 Elsevier B.V. All rights reserved.
Keywords: Transition metal carbonyl anions; SNV reactions; 1,2-Shift (migration); Carbenoids; Iron; Rhenium; DFT calculations
1. Introduction
substitution is the influence of the particular nature of tran-
sition metal centered nucleophiles on the reaction mecha-
nism. Such features of metal carbonyl anions as high
reducing power and halogen affinity, the weakness of car-
bon-metal bond forming in the reaction, and the feasibility
of reactions affecting the ligands (C„O), let alone their prac-
tical importance as catalysts and reagents, make them an
excellent model for our studies.
One of the basic mechanistic criteria in nucleophilic sub-
stitution at sp2-carbon atom is the element effect, the rela-
tive mobility of halogen leaving groups. Higher mobility of
fluorine relative to other halogens is normal for the reac-
tions with ‘‘traditional’’ nucleophiles, which follow the
addition-b-elimination mechanism with the rate limiting
addition step (Scheme 1) [1,2]. Yet, an ‘‘abnormal’’ leaving
Probably no other class of organic reactions is as rich in
mechanistic alternatives and mechanistic criteria to differen-
tiate between them as nucleophilic vinylic substitution. At
present more than 10 main types of reaction mechanisms
are recognised [1], and other pathways are likely to be discov-
ered in the future. However, most mechanistic studies of
vinylic substitution [2] do not go beyond the basic set of ‘‘tra-
ditional’’ nucleophiles, – alkoxides, amines, tiolates and, to a
minor extent, carbanions. The focus of our interest in vinylic
*
Corresponding authors. Tel./fax: +7 095 9393618.
E-mail addresses: beletska@org.chem.msu.ru, petr@elorg.chem.
msu.ru (I.P. Beletskaya).
0022-328X/$ - see front matter ꢀ 2006 Elsevier B.V. All rights reserved.
doi:10.1016/j.jorganchem.2005.12.050

P.K. Sazonov et al. / Journal of Organometallic Chemistry 691 (2006) 2346–2357
2347
Nu
F
F
Nu
F
F
F
F
Nu-
- F-
F
Cl
Cl
Cl
F
Nu- = RO- , RS- , carbanions
Scheme 1.
RHal + [M(CO)nLm] -
RFe(CO)2Cp
O
R - + HalM(CO)nLm
[RCM(CO)4Hal] -
M= Re, Mn
M(CO)nLm = CpFe(CO)2, Re(CO)5, Mn(CO)5
R = CF2=CF, Hal = Br; R = C6F5, Hal = Cl, Br, I
Scheme 2.
group effect (Hal > F, where Hal = Cl, Br, I) is frequently
observed for metal-centered anions, cobaloximes and car-
bonylmetallates, in their reactions with vinylic and aro-
matic halides. As we have previously shown [3,4], such
reactivity pattern of polyfluorinated aryl and vinyl halides
is in fact a manifestation of a different reaction mechanism
– halogen–metal exchange (HME) (Scheme 2). The inter-
mediate carbanion interaction with metal carbonyl halide
gives the nucleophilic substitution (NS) products. In case
of rhenium and manganese carbonyl halides the carbanion
adds to the carbonyl ligand forming halo(acyl)metallate
anions instead of r-alkenyl complexes.
B
A
The present work, however, shows that vinylic substitu-
tion has still a few surprises to offer, and even in such a sim-
ple system as CF2@CFHal there is an alternative pathway
for the substitution of the heavy halogen. It is known that
the reaction of chlorotrifluoroethylene with [CpFe(CO)2]K
and [Re(CO)5]Na leads only to the substitution of chloride
anion [5]. One might suppose that it also follows the halo-
philic pathway, but an entirely different mechanism oper-
ates in this reaction. Our results point to the carbenoid
[L(CO)nMCF2CFCl]À (A) as the key intermediate, which
F
F
F
[Re(CO)5]Na
THF, r.t.
Cl
I
-
F
F
F
O-
C
F
F
Re(CO)5
CO
Na
Re
C
OC
OC
F
C
Re(CO)5
F
CO
F
F
Cl
III
30%
II
60%
Scheme 3.

2348
P.K. Sazonov et al. / Journal of Organometallic Chemistry 691 (2006) 2346–2357
Table 2
is transformed to the substitution products via 1,2-shift of
the metal carbonyl group. The proposed mechanism is val-
idated by computational modeling of the corresponding
intermediates and transition states.
IR spectra of metal carbonyl complexes
Compound IR-spectrum,a
m/cmÀ1
mC@O
mC@C
mC„O
Reactions were usually carried out directly in a NMR
sample tube, so that the product composition, including
the unstable and volatile compounds, could be determined
directly from NMR spectra. Trapping of the intermediate
carbanionic species with proton donors, that has often
helped us to understand the reaction mechanism [3,4], is
another technique used in the current study. We have also
examined the action of t-BuOK(Na)/t-BuOH on the satu-
rated adducts IV and IX in order to trace the genetic rela-
tionship between the reaction products.
II
–
1713 m 2011 vs, 2038 vs, 2083 m, 2152 m
1702 m 1927 vs, 1970 sh, 1984 vs, 2035
vs, 2087 s, 2149 m
–
–
IIIb
1610 m br
IV
V
–
2020 sh, 2037 vs, 2082 m, 2152 m
2013 vs, 2037 vs, 2082 m, 2151 m
1833 s
–
–
–
1605 s
1655 s, br
VII
VIII
IX
XI
XII
a
1717 m 1994 vs, 2044 vs
–
–
–
–
1984 vs, 2027 vs, 2043 vs
1993 vs, 2044 vs
1967 vs
1987 vs, 2039 vs
THF, r.t., lcell = 0.02 cm, range 1600–2300 cmÀ1
In the presence of 18-crown-6.
.
b
2. Results and discussion
2.1. Reaction of CF2@CFCl (I) with [Re(CO)5]Na
F
F
F
[Re(CO)5]Na
THF, r.t.
The 19F NMR spectrum of the reaction solution showed
that besides the CF2@CFRe(CO)5 product (II, 60%), previ-
ously isolated from this reaction [5], another product is also
formed (Scheme 3). In the 19F NMR spectrum of this new
compound (Table 1) one can see both resonances belonging
to trifluorovinyl group and those belonging to XCF2–
CFClY fragment, and in IR (Table 2) and 13C NMR spectra
patterns characteristic for neutral XRe(CO)5 and anionic
cis-X2Re(CO)4 complexes1 [3,4,6,7]. The above data allowed
to identify it as a rhenium acylmetallate salt (III, 30%). Com-
plex of III with 18-crown-6 and dioxane was precipitated
from the reaction mixture as a yellow crystalline solid.
Complex III is an obvious indication of a carbanionic r-
adduct intermediate [(CO)5ReCF2CFCl]À (A) taking part in
the reaction. It may result only from nucleophilic addition of
intermediate A to CO ligand of complex II (Scheme 4).
Nucleophilic addition to CO ligand is highly characteristic
for metal carbonyls, especially electrophilic complexes such
as RRe(CO)5 (R = Hal, Rf or other EWG-group) [6,7].
The role of anionic r-adduct A as an intermediate in the
reaction was further confirmed when the reaction was car-
ried out in the presence of t-BuOH (2 eq), a proton donor,
which can effectively trap carbanions [3,4]. In that ‘‘anion
trapping’’ experiment (Scheme 5) the nucleophilic substitu-
tion products II and III were not observed in the reaction
mixture, instead the saturated adduct (IV, 35%) was formed,
confirming the addition of [Re(CO)5]À anion to I at the CF2-
site. The formation of other products, a-metallated fluoro-
anhydride F(CO)CHFRe(CO)5 (V, 35%), isobutene2 (40%)
Cl
O
C
(CO)5Re
F
Na
III
CF2=CFRe(CO)4
Cl
F
F
II
A
Scheme 4.
O
(CO)5Re
F
H
[Re(CO)5]Na
I
+
(CO)5ReCHFCF
t-BuOH
THF, r.t.
Cl
F
F
IV 35%
35%
V
H3C
H3C
F
+
F
+
F
H
8%
40%
Scheme 5.
and CF2@CFH is not so easy to explain, and will be dis-
cussed later in Section 2.2.
The same reaction when monitored by 19F NMR at low
temperature (À60 to À5 ꢁC) gave only two products. One
was the already mentioned IV, for the other product struc-
ture VI might be surmised3 (Scheme 6). On heating the reac-
tion mixture to r.t. (30 min) the signals ascribed to VI
disappeared, and signals belonging to V and CF2@CFH
became visible in 19F NMR spectrum.
1
Two axial C„O groups in cis-X2Re(CO)4 fragment are diastereotopic
because of the presence of a chiral centre in the structure of III (CFCl
group), and hence X2Re(CO)4 gives four signals of equal intensity in 13C
NMR spectrum instead of 1:2:1 pattern typical of such complexes in
symmetrical ligand environment.
3
It exhibited a signal pattern similar to IV, but with slightly different
2
Isobutene was identified by its characteristic signals in 1H NMR (dH,
chemical shifts (Table 1). Taking into account the tendency of RRe(CO)5
complexes to add nucleophiles at the cis-carbonyl ligand it might be well
to surmise the structure VI for this intermediate.
ppm: 1.676, 4.609, JH–H 1.2 Hz) and 13C NMR spectra (dC, ppm: 24.12,
110.96, 142.55).

P.K. Sazonov et al. / Journal of Organometallic Chemistry 691 (2006) 2346–2357
2349
composition observed in the reaction of [Re(CO)5]Na with
I in the presence of t-BuOH (Scheme 5).
F
F
F
[Re(CO)5]Na
t-BuOH, THF
Cl
-60 - -5 C (20 min)
˚
˚
2.2. Possible mechanism for the formation of
(CO)5ReCHF(CO)F (V), isobutylene and CF2@CFH
-
OBu-t
O
C
As demonstrated by the low temperature experiment
(Scheme 6) and the reaction of IV with t-BuONa (Scheme
7), fluoroanhydride V, isobutylene and CF2@CFH are
formed together from the saturated adduct IV via the inter-
mediate VI, i.e., in the process beginning with the addition
of t-BuONa to the CO-ligand of IV. The sequence of trans-
formations following may be rationalized as shown in
Scheme 8.
(CO)5Re
F
H
Na+
(CO)4Re
F
H
+
Cl
Cl
F
F
F
F
IV 25%
35%
VI
Scheme 6.
Stage (2) of this mechanism involves c-elimination of
chloride anion from the acylmetallate VI, a reaction reverse
to the nucleophilic attack on a p-coordinated alkene. The
nucleophilic attack of t-BuONa on the p-complex of
CF2@CFH may result either in the substitution of this
weakly bound ligand, producing free CF2@CFH, or it
may result in the nucleophilic addition of t-BuOÀ to the
coordinated fluoroalkene (3). The sequence is completed
by base induced elimination of isobutylene from tert-
butyl-fluoroalkyl ether (5), leading to the fluoroanhydride
It is reasonable to suppose that the anionic r-adduct A
can be generated by the deprotonation of the saturated
adduct IV, and thus the chemistry associated with this
intermediate can be reproduced. Therefore, we studied
the reaction of IV with t-BuONa(K)/t-BuOH mixture
(Scheme 7). The result depended on the counterion. The
reaction of IV with t-BuOK produced r-vinylic complex
II together with V in 2:1 ratio. In case of t-BuONa the
products {V, isobutylene, a minor amount of CF2@CFH
and no r-vinylic complex II (Scheme 7)}, reproduce the
H3C
F
F
F
t-BuONa, t-BuOH
THF, r.t.
+
V
+
H3C
H
(CO)5Re
F
H
Cl
F
F
F
F
t-BuOK, t-BuOH
THF, r.t.
V
+
IV
F
Re(CO)5
II
Scheme 7.
O
F
F
F
t-BuONa
+
[(t-BuO)t-BuOCRe(CO)4]Na
H
(4)
(3)
(1)
(2)
OBu-t
O
OBu-t
O
OBu-t
O
C
C
C
H
t-BuONa
t-BuONa
F
F
Na+
(CO)4Re
H
IV
Na
OBu-t
(CO)4Re
(CO)4Re
- NaCl
F
F
H
F
F
F
Cl
F
F
VI
(5)
(6)
t-BuONa
H2C
H
C
O
O
(CO)5Re
H
t-BuONa
CH3
CH3
(CO)5ReCHFCF
V
Na
-NaF
- (CH3)2C=CH2
- t-BuOH
F
(CO)5ReCFHCF2O
F
F
Scheme 8.

2350
P.K. Sazonov et al. / Journal of Organometallic Chemistry 691 (2006) 2346–2357
F
F
F
F
F
F
Fp
Fp
FpK
THF
+
Fp2
+
Fp
Cl
VII, 5%
VIII, 30%
50%
I
Scheme 9.
F
F
F
F
F
F
FpK
2FpK
FpC CFp
VIII
FC CF
K
Cl
I
2KF
KF
Fp2, KCl
Scheme 10.
Fp
H
t-BuO
+
H
+
F
F
FpK
Fe
F
+
I
Fp2
+
VII
+
C
THF, -50 C
t-BuOH
˚
F
Cl
F
Cl
O
F
H
O
F
F
F
F
F
7%
X, 30%
IX, 35%
6%
20%
t-BuO
XI, 23%
Scheme 11.
V (6), a transformation well known in fluorine chemistry
[8].
(b) The observation of saturated adduct4 (IX, 35%) indi-
cates that [FpCF2CFCl]À intermediate (A-Fe) is
formed in the course of the reaction, i.e., it confirms
the FpK addition to the CF2-group of I.
(c) The formation of the NS product VII (6%) seems to
be unaffected by proton donor.
2.3. Reaction of CF2@CFCl (I) with [CpFe(CO)2]K
(FpK)
Just as described by Bruce et al. [5] the reaction of FpK
with I gives no more than 5% of the corresponding NS
product VII (Scheme 9). No other fluorine-containing com-
pounds were observed by 19F NMR, however there was a
non-fluorinated product not reported previously – a di-
metal substituted acetylene VIII (30%). The IR-spectrum
of acetylene VIII in metal carbonyl region consists not of
two, which is typical for other FpX complexes, but of three
bands of unequal intensity (Table 2). A similar spectral pat-
tern has been observed for the ruthenium analogue of com-
plex VIII, and it was suggested that sin and anti rotamers of
the complex give separate bands in IR-spectrum [9].
Formation of acetylene VIII is a characteristic signature
of the corresponding fluoroalkenyl carbanion, the
[CF2@CF]À anion, which can eliminate b-fluoride to give
the ‘‘superelectrophilic’’ difluoroacetylene (Scheme 10).
The situation changed significantly when the reaction
was carried out in the presence of t-BuOH (Scheme 11):
(d) A new metallocycle product XI (23%) is formed. For-
mation of metallocycle XI largely depends upon the
nature of the cation in the carbonylate salt. The same
reaction carried out with FpLi instead of FpK affor-
ded products IX (45%), VII (5%), CF2@CFH (5%)
and Fp2 dimer (40%), but gave no complex XI.
The treatment of the saturated adduct IX with t-BuOK
also produces metallocycle XI together with r-vinyl com-
plex VII (Scheme 12).
The structure of XI was confirmed by single-crystal X-
ray diffraction analysis and is shown in Fig. 1. Complex
XI has a slightly distorted piano-stool geometry. Its chief
peculiarity is the coordination of acyl oxygen to iron, form-
ing a five-membered conjugated cycle. The metallacycle is
characterized by flattened envelope conformation with
˚
the deviation of the O(1) atom by 0.064(3) A from the
plane of the remaining atoms. Several examples of similar
metallocyclic iron complexes have already been described
(a) The formation of acetylene VIII was suppressed
completely, which, together with the observation
of CF2@CFH (7%), was only to be expected for
a process involving [CF2@CF]À carbanion (Scheme
10).
4
The saturated adduct IX appears to be extremely moisture sensitive
unlike other organofluorine metal carbonyl derivatives. The crystals of IX
become blurred in the open air in a matter of minutes, resulting in a
brownish oil mainly consisting of Fp(CO)CHFCl (XII).

P.K. Sazonov et al. / Journal of Organometallic Chemistry 691 (2006) 2346–2357
2351
ular insertion of a coordinated aryne (alkyne) into Fe–C
bond of R(CO)Fe group [10]. However, this mechanism
is inconsistent with our observation, that the action of t-
BuOK on iron r-vinylic complex VII did not give any trac-
table products. The mechanism that would be acceptable
has to take into account the transformation of the satu-
rated adduct IX into complex XI under the action of t-
BuOK (Scheme 12), and this suggests [FpCF2CFCl]À anion
(A-Fe) as an intermediate.
Fp
F
H
t-BuOK, t-BuOH
THF, r.t.
Cl
F
F
IX
F
F
F
Fe
F
F
C
+
O
O
Fp
t-BuO
2.4. The mechanism of chloride substitution in CF2@CFCl
(I) by metal carbonyl anions
XI
Scheme 12.
VII
We can exclude the HME pathway (Scheme 2) from the
start, as it would have produced haloacylrhenate
[CF2@CF(CO)Re(CO)4Cl]Na instead of r-vinylic complex
II actually formed in the reaction with [Re(CO)5]Na. The
results of anion trapping experiments both for iron
(Scheme 11) and rhenium (Schemes 5 and 6) are not consis-
tent with the HME mechanism either. On the other hand:
(a) formation of complex III (Scheme 3); (b) formation
of saturated adducts IV (Scheme 5) and IX (Scheme 11)
in the presence of t-BuOH, while the formation of NS
products II and III is completely suppressed; (c) formation
of NS products II and VII directly from the corresponding
adducts IV and IX under the action of t-BuOK (Schemes 7
and 12) – all this points to carbanionic r-adduct
[MCF2CFCl]À (A) as the intermediate on the NS pathway.
The problem, however, is that anion A cannot transform
into NS products II and VII by b-elimination of a leaving
group, which would be b-elimination of fluoride as is the
case with ‘‘ordinary’’ nucleophiles (ROÀ, RSÀ, R2NÀ, car-
banions, Scheme 1). There must be some other pathway for
the transformation of the intermediate A, much more facile
than b-elimination of fluoride.
Fig. 1. A general view of the molecular structure of metallocycle XI (50%
probability ellipsoids). Selected bond distances (A): Fe–C(1) 1.903(2), Fe–
O(1) 2.012(2), Fe–C(8) 1.755(2), O(1)–C(3) 1.257(3), C(1)–C(2) 1.347(3),
C(2)–C(3) 1.432(3), O(2)–C(3) 1.312(3); and bond angles (ꢁ): C(8)–Fe–C(1)
91.8(1), C(8)–Fe–O(1) 99.92(9), C(1)–Fe–O(1) 81.14(8).
˚
A question arises: could the exclusive substitution of
chloride in I be explained by a mechanism with parallel
and reversible formation of two isomeric carbanions A
and B (Scheme 13)? The NS products can be formed from
a less stable carbanion B, while the anion trap protonates
the more stable carbanion A present in much higher
concentration.
[10]. A mechanism proposed in the literature for their for-
mation usually consists in (a) attack of the nucleophile at
the CO ligand of r-aryl complex (r-vinyl in our case);
(b) c-elimination of leaving group resulting in a p-coordi-
nated aryne (in our case it would be alkyne); (c) intramolec-
(kCl · kel)
kobs
=
Without t-BuOH:
F
F
L(CO)nM
Cl
M(CO)nL
F
(k-1 + kel)
kel
- Cl -
F
F
kCl
F
kobs ≤ kCl
k-1
B
[M(CO)nL]–
I
k
obs = kF
With t-BuOH:
kF
k-1
L(CO)nM
L(CO)nM
H
kF >> kCl
t-BuOH
t-BuO-
F
F
Cl
F
F
Cl
F
F
A
Scheme 13.

2352
P.K. Sazonov et al. / Journal of Organometallic Chemistry 691 (2006) 2346–2357
L(CO)nM
F
H
Cl
F
F
t-BuO-
t-BuOH
F
F
L(CO)nM
F
F
F
Cl
F
M(CO)nL
F
kel
- Cl -
[M(CO)nL]–
I
F
F Cl
A
Scheme 14.
-
O
FpK
C
OBu-t
O
F
Fe
K+
I
K+
Fe
Fe
C
t-BuOH
- KCl
O
C
O
C
O
O
F
Cl
F
F
F
F
F
F
F
A-Fe
C
t-BuOK
t-BuOH t-BuOK
Fp
F
H
- KF
Fe
F
C
O
Cl
O
F
F
F
IX
t-BuO
XI
Scheme 15.
Our answer is no. Firstly, if this scheme is true, it is dif-
ficult to conceive how NS products could be formed
directly from the saturated adducts IV and IX under the
action of t-BuOK. Secondly, we have shown that the rate
of the reaction of I with [Re(CO)5]Na does not change in
the presence of t-BuOH (k2 = 0.7 0.15 l molÀ1 sÀ1), while
a significant rate increase (proportional to kF/kCl) has to be
expected according to Scheme 13. And finally, this mecha-
nism does not explain the formation of metallocycle prod-
uct XI from carbanion A.
Carbanion intermediate A, like all a-halogen-substituted
carbanions [11], is a carbenoid, and this is the key point in
understanding the reaction mechanism. Its transformation
into the NS products II or VII can proceed via 1,2-shift of
metal carbonyl group concerted with a-elimination of chlo-
ride anion (Scheme 14).
bonds to carbon, such as silicon [14,15]. The same must
be true for transition metal groups, Re(CO)5 and
Fe(CO)2Cp. Concerted 1,2-shift of metal carbonyl group
appears to provide additional driving force for the a-elim-
ination of chloride anion, since the analogous carbanion
intermediates in the reactions of I with sulphur or oxygen
nucleophiles do not eliminate chloride (Scheme 1).
Moreover, if we consider that carbenes (and conse-
quently carbenoids) are easily inserted into metal-C„O
bond [16], formation of metalacycle XI from the carbenoid
anion A-Fe can be also explained. Insertion of the carbene
equivalent into Fe–C„O bond will produce the iron ketene
complex C (Scheme 15). Nucleophilic addition of alkoxide
While 1,2-shift in carbanions is considered not feasible
due to the antiaromaticity of the corresponding 4-electron
transition state, no molecular orbital symmetry restrictions
apply to the 1,2-shift in a carbenoid if it is concerted with
the elimination of chloride anion [12]. It is the key step in
Fritsch–Buttenberg–Wiechell rearrangement, and is analo-
gous to Beckmann rearrangement of oximes [12]. In case of
free carbenes the 1,2-shift of the b-substituent to the car-
bene centre is usually a rapid process [13]. The ease of this
rearrangement increases for b-substituents forming weak
Fe1
O
O10
C9
Fe
1.195
F7
O
F
F
1.435
F
C5
F6
C
F8
Fig. 2. The calculated structure of keten intermediate C (bond lengths in
˚
A).

P.K. Sazonov et al. / Journal of Organometallic Chemistry 691 (2006) 2346–2357
2353
C3
Re1
Re1
O6
2.036
C4
O7
C12
F14
C12
F16
F14
1.427
F16
Re1
F15
C13
F15
C13
Na18
Re1
C4
Na18
Cl17
F14
F16
C13
Re1
C12
Cl17
C12
F15
F16
C12
F16
C4
O7
F15
C13 1.536
F15
O7
F14
F14
Na18
Na18
C13
Cl17
Cl17
Na18
Cl17
II·NaCl
Re-acyl
A-Re
TS-Re
TS-acyl
Energy, kcal/mol:
-12.5
-38.5
3.2
0.0
4.9
Fig. 3. The calculated structures and energies of rhenium carbenoid [(CO)5ReCF2CFCl]Na (A-Re), its transformation products (IIÆNaCl and Re-acyl) and
the corresponding transition states (TS-Re and TS-acyl).
to iron ketene complexes has been described [17]. It is
directed at the central carbon atom of ketene ligand and
gives alcoxycarbonyl iron complexes. Thus, the reaction
of C with t-BuOK will give the metallocycle XI, and the
formation of XI supports the proposed carbenoid mecha-
nism (Scheme 14).
carbenes rearrange in the process of geometry optimisa-
tion, but rearrangement pathways are different for iron
and rhenium. 1,2-Migration of Re(CO)5 group in (CO)5R-
eCF2CF: results in r-vinyl complex (II) (DE = 46.1 kcal/
mol),5 but in its iron analogue carbene inserts into Fe–
CO bond (DE = 22.5 kcal/mol). The resulting structure
(Fig. 2), may be regarded as iron ketene complex, which
was proposed as an intermediate (C) in the formation of
metallacycle XI (Scheme 15).
The calculation of rhenium analogue of ketene complex
C showed that its formation from (CO)5ReCF2CF: is less
exothermic than C from FpCF2CF: by 5.4 kcal/mol. This,
however, does not seem to be the only reason for the pre-
ferred insertion in FpCF2CF:. Because of the shorter Fe–
C bonds and greater steric crowding imposed by Cp ring,
the 1,2-migration path intersects with the slopes of the
‘‘hollow’’ corresponding to ketene complex C.
2.5. DFT study of carbenes [MCF2CF:] and carbenoids
[MCF2CFCl]Na (M = Re(CO)5Na and CpFe(CO)2):
1,2-shift via carbene insertion into metal–CO bond
Carbenes and more recently carbenoids [11] have been
the subject of extensive quantum chemical studies, yet very
little data are available on 1,2-shift of groups other than
hydrogen or simple organic radicals. As for the b-element
substituted carbenes we can find some data only for silicon.
It was shown (at MP2 or B3LYP levels of theory) that b-
silicon substituted carbenes represent no minima on poten-
tial energy surface rearranging to the corresponding
alkenes without an activation barrier [14,15].
The calculation method used in this study (B3LYP/
SDD(Fe, Re) 6-31G(d) for all other elements) has shown
very good performance in various organometallic systems
[18]. We have checked its adequacy for the systems under
consideration by computing the structures of iron metallo-
cycle XI and halo(acyl)rhenate anion [CF2@CF(CO)-
Re(CO)4Br]À [4]. In both cases the computed structures
were in good agreement with the X-ray data.
In contrast to free carbenes energy minima correspond
to the carbenoids [(CO)5ReCF2CFCl]Na (A-Re) and
[FpCF2CFCl]Na (A-Fe). A number of isomeric structures
for these carbenoids were located, differing in the OC–
Fe–CF2–CFCl dihedral angle and the coordination mode
of Na (with CF or CO bonds). The most stable are shown
in Figs. 3 and 4. The structures show that they are typical
˚
˚
carbenoids, with elongated C–Cl bond (2.18 A and 2.09 A
for A-Fe and A-Re, respectively) and a three-centre coordi-
nation of Na cation with Cl, carbenoid-C and b-F atoms.
At first we chose the carbenes (CO)5ReCF2CF: and
FpCF2CF: as more simple models of the key intermediate
in the proposed mechanism – carbenoid A. However, for
the free carbenes no energy minima could be located. Both
5
Relative to the energy of the starting carbene optimised with Re(Fe)–
CF2–CF: angle constrained to 110ꢁ (obtained in molecular mechanics
geometry optimisation with mm+ force field).

2354
P.K. Sazonov et al. / Journal of Organometallic Chemistry 691 (2006) 2346–2357
F7
Fe1
F12
C4
C5
C8
Cl10
1.152
F6
O9
Cl10
Na11
F12
F7
Fe1
17
2.8
C5
C4
Cl10
Fe1
Na11
C4
F6
TS-Fe
C8
F7
O9
Na11
F6
C5
A-Fe
F12
C·NaCl
Energy,
kcal/mol
0.0
2.1
-21.8
Fig. 4. The calculated structures and energies of iron carbenoid [(CO)2CpFeCF2CFCl]Na (A-Fe), keten complex (CÆNaCl) and the connecting transition
state (TS-Fe).
A transition state (TS) for 1,2-migration of Re(CO)5
group (TS-Re in Fig. 3) was easily located. The activation
barrier for 1,2-migration is only 3.2 kcal/mol, and TS-Re
can be characterized as early and asymmetrical. The devel-
3. Conclusions
The title reaction became a classical example of the
so-called ‘‘abnormal halogen leaving group effect’’ – the
preferred substitution of a heavy halogen in a vinyl halide
usually associated with the metal-centered nucleophiles
[19]. The present study shows that this reaction is a clear
case of the ‘‘carbenoid’’ mechanism of nucleophilic vinylic
substitution, a process involving 1,2-shift of the entering
nucleophile (Scheme 14). This mechanism proposed at first
from purely experimental evidence was then validated by
quantum chemical DFT calculations. Besides, our experi-
ments have also revealed an unexpectedly rich and varied
chemistry with such compounds as III, IV, V, VIII, IX,
XI, XII not reported previously.
It appears that, the halophilic mechanism (Scheme 2) is
not the only pathway that can lead to the substitution of a
heavier halogen, and the lower reactivity of I in the halo-
philic process compared to bromotrifluoroethylene [4]
opens the way for the alternative ‘‘carbenoid’’ reaction
route.
1,2-Shift of the entering nucleophile is not exactly a new
idea for the NS in unsaturated systems. In fact b-addition –
a-elimination – 1,2-shift sequence was among the first
mechanisms to be considered for the NS in acetylenes as
early as 1970s [20]. Later, it was proposed for vinylic sys-
tems to account for the substitution of a leaving group
attached to the same carbon atom as the EWG-group
[21], and was proved kinetically for the substitution with
thiolates anions in a-bromovinylsulphones [22]. However,
it remained rather a mechanistic hypothesis, and its
relation to the carbenoid rearrangements has not been
˚
oping Re–CF bond is much longer (2.77 A), than the
˚
breaking Re–CF2 bond (2.38 A). Another TS (TS-acyl in
Fig. 3) higher in energy (4.9 kcal/mol) than TS-Re corre-
sponds to the attack of carbenoid anion at the CO ligand,
however, it is not an insertion of carbene as it does not lead
to the rhenium ketene complex analogous to C. It corre-
sponds instead to the intramolecular nucleophilic addition
of the carbanion to CO group, resulting in the acylrhenate
salt (Re-acyl in Fig. 3). We see here obvious parallel with
the intermolecular nucleophilic addition of [(CO)5Re-
CF2CFCl]Na to the CO ligand of II observed in the exper-
iment (Scheme 4).
Only one TS (TS-Fe in Fig. 4) was found for iron carb-
enoid A-Fe corresponding to the insertion of carbene into
the Fe–CO bond and leading to the cyclic ketene product
C, or rather its complex with NaCl (CÆNaCl in Fig. 4). It
is an extremely low lying (2.1 kcal/mol above A-Fe) and
early TS, with a very little change in Fe–C„O angle and
bond lengths. The distance between CO group and the
˚
carbenoid carbon in TS-Fe is as long as 2.56 A.
The most important conclusion that follows from the
above data are that different transformation pathways are
predicted for rhenium and iron carbenoids. It qualitatively
reproduces the different chemistries characteristic of
[Re(CO)5]Na and [CpFe(CO)2]K in the reaction with I.
Thus, the computational results support the proposed
mechanisms (Schemes 14 and 15) with the carbenoid
anions as the key intermediates.

P.K. Sazonov et al. / Journal of Organometallic Chemistry 691 (2006) 2346–2357
2355
explicitly considered. We have demonstrated that the real-
ization of this mechanism depends on the nature of nucle-
ophile involved, the ease of its 1,2-migration. This puts it
among the mechanisms of nucleophilic vinylic substitution
more specific for metal centered nucleophiles.
begins immediately (a complex with 18-crown-6 and diox-
ane, 100 mg). The compound can be purified by dissolving
in diethyl ether (3 ml) and precipitating with dioxane
(0.4 ml). 13C NMR (THF, r.t., dC, ppm):6 252.25 dm
(C@O, Jd ꢁ 45 Hz), 193.94 t (C„O, JC–F ꢁ 6 Hz, 1C),
192.45 t (C„O, JC–F ꢁ 3 Hz, 1C), 191.68 d (C„O,
JC–F ꢁ 8 Hz, 1C), 191.11 m (C„O, 1C), 183.16 t (C„O,
JC–F ꢁ 6 Hz, 5C). MALDI MS (À), m/z: 822.7 [MÀCO,
ÀNa]À (5%), 800.8 (5%), 766.8 [MÀ3CO, ÀNa]À (8%),
738.8 [MÀ4CO, ÀNa]À (12%), 646.7 [Re2(CO)6C4F3]À
(15%), 386.8 [CF2@CFRe(CO)3Cl]À (100%), 379.8 [CF2@
CFRe(CO)4]À (25%).
Complex II was isolated from mother liquor by column
chromatography on silica gel (L 60/200). The first fraction
eluted with petroleum ether containing Re2(CO)10 was
discarded, the second fraction eluted with petroleum
ether–CH2Cl2 (4:1) gave 110 mg of colourless crystals.
The product is extremely volatile and part of it may be lost
while removing the solvent. 13C NMR (THF, r.t., dC,
4. Experimental
General experimental techniques and procedures were
exactly the same as described in the preceding communica-
1
tion [4]. H (400 MHz), 19F (376.3 MHz) and 13C (100.57
MHz) NMR spectra were obtained on a Varian VXR-400
spectrometer at r.t. MALDI mass spectrum of complex III
was obtained using a Bruker Daltonics Autoflex II
spectrometer equipped with N2 laser (337 nm), accelerat-
ing voltage 19 kV, in negative ion mode. Dithranol was
used as a matix, however similar results were obtained
with anthracene. Mass spectra of other compounds were
obtained under EI ionisation on a Cratos MS-890 spec-
trometer, operating in direct inlet method, with a source
temperature of 150 ꢁC and 70 eV ionising energy. The
mass numbers of cluster peaks are given for 35Cl and
187Re isotopes.
ppm): 181.08 s br (C„O, 4C), 182.66 d (C„O, JC–F
=
6 Hz, 1C), 164.18 ddd (@CF2, JC–F = 257, 306, 42 Hz),
126.73 ddd (@CFRe, JC–F = 273, 83, 6 Hz).
Chlorotrifluoroethylene (b.p. À28 ꢁC) was measured out
by vapor pressure as described earlier [4]. The yields of the
products discussed above are purely spectroscopic, based
upon the relative integral intensity of the signals of the
products and the signal of internal standard (PhCF3,
4.2. Reaction of I with [Re(CO)5]Na in the presence of
t-BuOH: the isolation of CHFClCF2Re(CO)5 (IV) and
F(CO)CHFRe(CO)5 (V)
Alkene I (ꢁ1 mmol) was vacuum transferred to a NMR
sample tube containing the frozen THF solution (1 ml) of
0.46 mmol [Re(CO)5]Na and 58 mg t-BuOH. The tube
was heated to r.t. and 19F NMR spectra were recorded,
which showed that in 8 min the reaction was completed.
The products were separated by preparative TLC on Silu-
fol-UV-254 plates eluting with CH2Cl2-petroleum ether
(1:1). The uppermost band contained trace amounts of
Re2(CO)10, product IV (25 mg) was isolated (extraction
with Et2O) from the second band (Rf ꢀ 0.5) and product
V (51 mg) – from the bottom band (Rf ꢀ 0.25). The prepa-
ration was also repeated on a larger scale the products
being separated by column chromatography on silica gel
(gradient elution with petroleum ether–CH2Cl2). Yet in
that case the isolated yield of the products was lower, espe-
cially V (ꢁ10%), and the colourless bands of IV and V were
closely followed on the column by yellow bands of the
products of their decomposition.
Crude IV was recrystallised from hexane (+20 to
À70 ꢁC) to give off-white solid. EI MS, m/z: 425 [MÀF]+
(3%), 408 [MÀHCl]+ (1,5%), 397 [MÀF, ÀCO]+ (9%),
381 [MÀCl, ÀCO]+ (16%), 362 [Re(CO)5Cl]+ (60%), 353
[MÀCl, À2CO]+ (13%), 349 [MÀCClFH, ÀCO]+ (34%),
346 [Re(CO)5F]+ (40%), 334 [Re(CO)4Cl]+ (55%), 327
[Re(CO)5]+ (70%), 306 [Re(CO)3Cl]+ (70%), 304
[ReCF2CFClH]+ (80%), 299 (60%) [Re(CO)4]+, 290
[Re(CO)3F]+ (45%), 278 [Re(CO)2Cl]+ (50%), 271
1
C6H5F) in 19F and H NMR spectra of reaction solutions.
The signals were referenced to the individual products by
comparison to the spectra of the isolated compounds or
(in case of VII) with the previously obtained data [4].
1
The parameters of NMR (19F and H) and IR spectra of
the products are given in Tables 1 and 2, respectively
(except for V, VII and XII). [Re(CO)5]Na was purified by
the low temperature crystallisation from THF, unless spe-
cially noted.
4.1. Reaction of I with [Re(CO)5]Na: the isolation of
CF2@CFRe(CO)5 (II) and [CF2@CFRe(CO)4C(O)-
CFClCF2Re(CO)5]Na (III)
In a typical procedure a solution of 0.44 mmol of I in
0.26 ml THF was added to 1.33 ml of THF solution con-
taining 0.42 mmol [Re(CO)5]Na at r.t. The reaction solu-
tion was transferred into NMR sample tube, which was
then frozen in liquid nitrogen and sealed off with flame.
The NMR spectra showed the reaction to be completed
in less then 15 min after heating to r.t. Products were iso-
lated from a larger scale experiment, carried out with
non-crystallised [Re(CO)5]Na, obtained from 327 mg
Re2CO10 and 0.65 ml of 0.5% NaHg in ꢁ8 ml THF, and
1.0 mmol I. After 30 min at r.t., THF was replaced with
diethyl ether (2 ml), and the obtained orange solution
was filtered through Celite pad. 18-Crown-6 ether
(150 mg) and dioxane (0.4 ml) were consecutively added
to the filtrate. The precipitation of the yellow crystals of IIIÆ
6
The weaker signals of fluorinated backbone were not observed.

2356
P.K. Sazonov et al. / Journal of Organometallic Chemistry 691 (2006) 2346–2357
[Re(CO)3]+ (85%), 250 [ReC2F2H]+ (45%), 243 [Re(CO)2]+
(60%), 237 [ReCF2]+ (40%), 218 [ReCF]+ (45%), 215
[Re(CO)]+ (45%), 206 [ReF]+ (35%), 199 [ReC]+ (50%),
187 [Re]+ (100%). A peak pattern of a compound with
m/z = 714, presumably C2F4Re2(CO)10, is also observed
in the spectrum except for several first scans.
[FeCp]+ (10%), 108 (29%), 89 (18%), 57 (17%), 56
(13%), 41 (17%), 39 (15%). Anal. Calc. for C13H14F2FeO3:
C, 50.03; H 4.52. Found: C, 50.19; H, 4.60%. The satu-
rated adduct IX was not isolated in this experiment, it
decomposes during column or plate chromatography on
silica gel.
Crude V was sublimed at 60 ꢁC/5 · 10À2 Torr to give a
1
white solid. H NMR (THF, r.t., d, ppm): 6.184 dd (JH–F
4.4.1. Crystal data for XI
At 298 K crystals of XI (C13H14F2FeO3) are triclinic,
51, 1.5 Hz). 19F NMR (THF, r.t., dF, ppm): 9.85 d (JF–F
61 Hz), À184.19 dd (JH–F 51 Hz) À140.86 d (JH–F 55 Hz).
EI MS, m/z: 406 M+ (2), 359 [MÀCOF]+ (2%), 327
[Re(CO)5]+ (40%), 299 [Re(CO)4]+ (50), 271 [Re(CO)3]+
(90%), 243 [Re(CO)2]+ (60%), 238 [ReCHF2]+ (60%), 199
[ReC]+ (50%), 187 [Re]+ (100%), 60 (50%), 51 (50%)
[CHF2]+. Anal. Calc. for C7HF2O6Re: C, 20.75; H 0.25.
Found: C, 20.75; H, 0.26%.
ꢀ
˚
˚
space group P1, a = 8.186(3) A, b = 9.192(3) A, c =
˚
9.560(4) A, a = 76.82(3)ꢁ, b = 89.43(3)ꢁ, c = 70.76(3),
3
˚
V = 659.6(5) A , Z = 2 (1), M = 312.09, dcalc = 1.571
g cmÀ3, l(Mo Ka) = 11.67 cmÀ1, F(000) = 320. Intensities
of 3990 reflections were measured with a Siemens P3/Pc
using graphite monochromated Mo Ka radiation (k =
˚
0.71073 A, h/2h-scans) and 3736 independent ones were
used in the further refinement. The structure was solved
by direct method and refined by the full-matrix least-
squares technique against F2 in the anisotropic-isotropic
approximation. Hydrogen atoms were located from the
Fourier synthesis and refined using riding model. The
refinement converged to wR2 = 0.1267 and GOF = 0.991
for all independent reflections (R1 = 0.0489 was calculated
against F for 3445 observed reflections with I > 2r(I)). All
calculations were performed using SHELXTL PLUS 5.0 on
IBM PC AT.
4.3. Reaction of I with FpK: the isolation of FpCCFp (VIII)
FpK, obtained from 0.166 mmol Fp2 and 30 ll NaK2.8
in 4 ml THF, was mixed with I (0.33 mmol) in 0.15 ml
THF at r.t. The solution immediately turned from
orange-red to dark cherry. Part of the reaction mixture
was subjected to column chromatography on silica gel,
but the dark-yellow band eluted from the column with
ethyl acetate–CH2Cl2 mixture (1:1) contained only decom-
position products instead of acetylene VIII. The rest of the
reaction mixture was placed on a column with neutral alu-
mina (Brockman grade II), the column was first eluted with
petroleum ether–CH2Cl2 (1:1) to remove VII and Fp2, then
a yellow band was eluted with CH2Cl2–ethyl acetate (10:1)
mixture which after the evaporation of the solvent afforded
4.5. Reaction of I with FpLi in the presence of t-BuOH: the
isolation of CHFClCF2Fp (IX)
The solution of FpK, obtained from 176 mg Fp2 and
120 ll NaK2.8 in 10 ml THF, was stirred for 20 min with
51 mg LiCl, which gradually dissolved and KCl precipi-
tated forming a light slurry. To the obtained orange-brown
solution t-BuOH (280 mg) was added, the mixture was
cooled to À50 ꢁC and 1.0 mmol of I (in 0.6 ml THF) was
introduced. After 15 min at À50 ꢁC the reaction mixture
was allowed to reach the r.t. Product IX (90 mg) was iso-
lated from the reaction mixture by vacuum sublimation
at 30–50 ꢁC/5 · 10À2 Torr/2 h and purified by low temper-
ature crystallization (+20 to À80 ꢁC) from 2 ml petroleum
ether–Et2O (8:1). EI MS, m/z: 294 M+ (9), 266 [MÀCO]+
(7%), 258 [MÀHCl]+ (5%), 240 [MÀF, ÀCl]+ (5), 230
[MÀHCl, ÀCO]+ (5%), 228 [MÀCFCl]+ (5%), 212
[MÀCO, ÀF, ÀCl]+ (8%), 205 (17%), 187 (20%)
[FeCp2H]+, 186 (95%) [FeCp2]+, 177 [Fe(CO)2Cp]+
(50%), 156 [FeCpCl]+ (100%), 140 [FeCpF]+ (95%), 121
[FeCp]+ (100%), 91 [FeCl]+ (55%), 82 (60%), 75 [FeF]+
(85%), 65 (60%), 63 (90%), 56 [Fe]+ (95%).
1
a few mg of a microcrystalline solid identified as VIII. H
NMR (THF, r.t., d, ppm): 4.913. EI MS, m/z: 378 M+
(8%), 350 [MÀCO]+ (10%), 322 [MÀ2CO]+ (10%), 294
[MÀ3CO]+ (6%), 266 [MÀ4CO]+ (60%), 210 [MÀ4CO,
ÀFe]+ (40%), 186 (80%) [FeCp2]+, 121 [FeCp]+ (100%).
4.4. Reaction of I with FpK in the presence of t-BuOH: the
isolation of
(XI)
Cp(CO)FeCF=CFC(O)OBu-t
Alkene I (0.75 g) was vacuum transferred to the frozen
solution of FpK, obtained from 310 mg Fp2 and 185 ll
NaK2.8 in 14 ml THF. The reaction mixture was allowed
to thaw in cold bath (À50 ꢁC), and after stirring for 5 min
heated to r.t. Metallocycle XI was isolated by column
chromatography on silica gel. The first yellow band eluted
with petroleum ether–CH2Cl2 mixture (8:1) contained fer-
rocene, it was followed by a brown band which after the
removal of the solvent gave 95 mg of XI. m.p. 78 ꢁC. 13C
NMR (acetone-d6, r.t., dC, ppm): 228.03 dd (JC–F = 369,
15 Hz, @CFFe), 218.6 d (JC–F = 2 Hz, C„O), 175.80 dd
(JC–F = 24, 14 Hz, C@O), 127.64 dd (JC–F = 277, 9 Hz,
@CF), 82.00 s (Cp), 28.27 s (t-BuO). EI MS, m/z: 312
M+ (5%), 284 [MÀCO]+ (5%), 256 [MÀ2CO]+ (2%),
228 [MÀCO, ÀC4H8]+ (100%), 164 [FeC3F2H2O2]+
(5%), 163 [FeC3F2HO2]+ (4%), 140 [FeCpF]+ (9%), 121
In the open air the solid IX is rapidly hydrolysed to give
Fp(CO)CHFCl (XII), which was isolated as a yellow oil by
column chromatography on silica gel, eluting with petro-
1
leum ether–CH2Cl2–ethyl acetate (4:4:1). H NMR (THF,
r.t., d, ppm): 5.085 s (Cp, 5H), 5.726 d (CHF, JH–F
= 55 Hz, 1H). 19F NMR (THF, r.t., dF, ppm): À140.86 d
(JH–F = 55 Hz). EI MS, m/z: 205 [Fe(CO)3Cp]+ (30%),
177 [Fe(CO)2Cp]+ (25%), 156 [FeCpCl]+ (5%), 149

P.K. Sazonov et al. / Journal of Organometallic Chemistry 691 (2006) 2346–2357
2357
[Fe(CO)Cp]+ (30%), 140 [FeCpF]+ (5%), 121 [FeCp]+
(100%), 95 (20%), 78 (80%), 56 [Fe]+ (80%).
[7] (a) C.P. Casey, D.M. Scheck, J. Am. Chem. Soc. 102 (1980) 2723;
(b) C.P. Casey, C.J. Czerwinski, R.K. Hayashi, Organometallics 15
(1996) 4362.
[8] V.F. Snegirev, K.N. Makarov, Bull. Acad. Sci. USSR Div. Chem.
Sci. (Engl. Transl.) 35 (1986) 91.
5. Computational details
[9] G.A. Koutsantonis, J.P. Selegue, J. Am. Chem. Soc. 113 (1991) 2316.
[10] (a) M. Akita, M. Tekada, S. Oyama, S. Sugimoto, Y. Moro-oka,
Organometallics 10 (1991) 1561;
Geometries were optimised using the B3LYP hybrid
DFT method, using triple-f basis set with the Stuttgart/
Dresden effective core potentials for iron and rhenium
[23] and the standard 6-31G(d) basis set for other elements,
as implemented in GAUSSIAN-98 program package [24]. Har-
monic frequencies were calculated at the same level of the-
ory to characterise the stationary points and transition
states (at 298.15 K, 1 atm). The connectivity of the TS
was confirmed by IRC calculations near the TS, followed
by geometry optimization of both reactants and products.
All energies are zero point corrected.
(b) A.N. Chernega, A.J. Graham, M.L.K. Green, J. Haggitt, J.
Lloyd, C.P. Mehnert, N. Metzler, J. Souter, J. Chem. Soc., Dalton
Trans. (1997) 2293.
[11] G. Boche, J.C.W. Lohrenz, Chem. Rev. 101 (2001) 697.
[12] (a) S.P. Samuel, T.-Q. Niu, K.L. Erickson, J. Am. Chem. Soc. 111
(1989) 1429;
(b) K.L. Erickson, J. Org. Chem. 38 (1973) 1463.
[13] (a) D.L.S. Brahms, W.P. Dailey, Chem. Rev. 96 (1996) 1585;
(b) W. Sander, G. Bucher, S. Wierlacher, Chem. Rev. 93 (1993)
1583.
[14] X. Creary, M.A. Butchko, J. Org. Chem. 67 (2002) 112.
[15] T.J. Barton, J. Lin, S. Ijadi-Maghsoodi, M.D. Power, X. Zhang, J.
Am. Chem. Soc. 117 (1995) 11695.
6. Supplementary material
[16] (a) W.A. Herrmann, J. Plank, M.L. Ziegler, K. Weidenhammer, J.
Am. Chem. Soc. 101 (1979) 3133;
(b) A. Miyashita, T. Kihara, K. Nomura, N. Hiroyuki, Chem. Lett.
(1986) 1607;
(c) L.S. Hegedus, G. de Weck, S. D’Andrea, J. Am. Chem. Soc. 110
(1988) 2122.
Crystallographic data (excluding structure factors) for
the structures reported in this paper have been deposited
with the Cambridge Crystallographic Data Center as sup-
plementary No. CCDC 278043. Copies of the data can
be obtained free of charge on application to CCDC, 12
Union Road, Cambridge CB2 1EZ, UK (fax: (internat.)
+44 1223 336 033; e-mail: deposit@ccdc.cam.ac.uk or
www: http://www.ccdc.cam.ac.uk).
[17] (a) E.J. Crawford, C. Lambert, K.P. Menard, A.R. Cutler, J. Am.
Chem. Soc. 107 (1985) 3130;
(b) T.W. Bodnar, A.R. Cutler, J. Am. Chem. Soc. 105 (1983)
5926.
[18] (a) M.N. Glukhovtsev, R.D. Bach, C.J. Nagel, J. Phys. Chem. A 101
(1997) 31;
(b) S. Zalis, H. Stoll, E.J. Baerends, W. Kaim, Inorg. Chem. 38
(1999) 6101;
(c) V.P. Ananikov, D.G. Musaev, K. Morokuma, Organometallics
20 (2001) 1652.
Acknowledgements
This work was financially supported by the Grant of the
President of the Russian Federation for the State Support
of the Leading Research Schools (Grant No. HI-
1611.2003.3). The authors gratefully acknowledge the help
of Dr. Navruz G. Akhmetov in obtaining the 19F and 13C
NMR spectra. The authors thank the P&M Company
(Moscow) for a gift of chlorotrifluoroethylene.
[19] P.K. Sazonov, V.A. Ivushkin, G.A. Artamkina, I.P. Beletskaya,
Arkivoc (2003) Part (X). Available from: <http://www.arkat-usa.org/
ark/journal/2003/I10_Rossi-Ruveda/RR-839C/839C.pdf> and refer-
ences cited therein.
[20] H.G. Viche, S.Y. Delavarenne, Chem. Ber. 103 (1970) 1216.
[21] (a) G.A. Russel, D.F. Dedolph, J. Org. Chem. 50 (1985) 3878;
(b) B.A. Shainyan, A.N. Mirskova, V.K. Belskiy, Zh. Org. Khim. 22
(1986) 1923.
[22] B.A. Shainyan, J. Phys. Org. Chem. 6 (1993) 59.
[23] (a) M. Dolg, U. Wedig, H. Stoll, H. Preuss, J. Chem. Phys. 86 (1987)
866;
References
(b) D. Andrae, U. Haeussermann, M. Dolg, H. Stoll, H. Preuss,
Theor. Chim. Acta 77 (1990) 123.
[1] Z. Rappoport, Rec. Trav. Chim. 104 (1985) 309.
[2] (a) For reviews see Ref. [1] and: Z. Rappoport, Acc. Chem. Res. 25
(1992) 474;
(b) Z. Rappoport, Acc. Chem. Res. 14 (1981) 7;
(c) Z. Rappoport, Adv. Phys. Org. Chem. 7 (1969) 1.
[3] (a) G.A. Artamkina, P.K. Sazonov, V.A. Ivushkin, I.P. Beletskaya,
Chem. Eur. J. 4 (1998) 1169;
[24] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb,
J.R. Cheeseman, V.G. Zakrzewski, J.A. Montgomery Jr., R.E.
Stratmann, J.C. Burant, S. Dapprich, J.M. Millam, A.D. Daniels,
K.N. Kudin, M.C. Strain, O. Farkas, J. Tomasi, V. Barone, M.
Cossi, R. Cammi, B. Mennucci, C. Pomelli, C. Adamo, S. Clifford,
J. Ochterski, G.A. Petersson, P.Y. Ayala, Q. Cui, K. Morokuma,
D.K. Malick, A.D. Rabuck, K. Raghavachari, J.B. Foresman, J.
Cioslowski, J.V. Ortiz, A.G. Baboul, B.B. Stefanov, G. Liu, A.
Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R.L. Martin,
D.J. Fox, T. Keith, M.A. Al-Laham, C.Y. Peng, A. Nanayakkara,
M. Challacombe, P.M.W. Gill, B. Johnson, W. Chen, M.W. Wong,
J.L. Andres, C. Gonzalez, M. Head-Gordon, E.S. Replogle, J.A.
Pople, GAUSSIAN-98, Revision A.9, Gaussian, Inc., Pittsburgh, PA,
1998.
(b) V.A. Ivushkin, P.K. Sazonov, G.A. Artamkina, I.P. Beletskaya,
J. Organomet. Chem. 597 (2000) 77.
[4] P.K. Sazonov, G.A. Artamkina, V.N. Khrustalev, M.Yu Antipin,
I.P. Beletskaya, J. Organomet. Chem. 681 (2003) 59.
[5] M.I. Bruce, P.W. Jolly, F.G.A. Stone, J. Chem. Soc. A (1966) 1602.
[6] (a) K.P. Darst, C.M. Lukehart, J. Organomet. Chem. 171 (1979) 65;
(b) D.W. Parker, M. Marsi, J.A. Gladysz, J. Organomet. Chem. 194
(1980) C1.