Nucleophilicity of metal carbonyl anions in vinylic substitution reactions

Article abstract of DOI:10.1002/poc.1279

Second order rate constants are reported for the reactions of metal carbonyl anions ([M(CO)_nL]^-) with several vinyl halides: PhCCI=C(CN)₂, Z- and E-Ph(CN)C=CHHl (Hal = Cl, Br) which follow the addition-elimination (Ad_NE) substitution mechanism. The obtained data show that the nucleophilic reactivity of [M(CO)_nL]^- anions towards vinyl halides increases in the same order as in aliphatic S _N2 reactions, but much more steeply, by 14 orders of magnitude in the row log{fr_{[M(Co)nL]M′}/K_[CpMo(CO)3]K): [CpFe(CO)₂]^- (～14), [Re(CO)₅]^- (7.8), [Mn(CO)_s]^_ 2.1, [CpW(CO)₃]^- (0.7) > [CpMo(CO)₃]^- (0). A good correlation exists between nucleophilicities of [M(CO)_nL]^- anions towards vinyl (sp²-carbon) and alkyl halides (sp³-carbon) with slope 2.7. The reactivity of [M(CO)_nL]^-in a halogen-metal exchange process (with Z-PhC(CN)=CHI) follows a similar 'large' scale as in the AdNE process. The nucleophilicity of [M(CO)_nL]^- anions correlates better with their one-electron oxidation potentials (E_ox) than with their basicity {pK_a, of [M(CO)_nL]H). Copyright

Full text of DOI:10.1002/poc.1279

Research Article
Received: 25 April 2007,
Revised: 5 September 2007,
Accepted: 14 September 2007,
Published online in Wiley InterScience: 18 January 2008
(www.interscience.wiley.com) DOI 10.1002/poc.1279
Nucleophilicity of metal carbonyl anions
in vinylic substitution reactions
a
a
a
*
*
Petr K. Sazonov , Galina A. Artamkina and Irina P. Beletskaya
Second order rate constants are reported for the reactions of metal carbonyl anions ([M(CO)nL]S) with several vinyl
—
—
halides: PhCCl C(CN) , Z- and E-Ph(CN)C CHHal (Hal ¼ Cl, Br) which follow the addition–elimination (Ad E)
—
—
2
N
substitution mechanism. The obtained data show that the nucleophilic reactivity of [M(CO)nL]S anions towards
vinyl halides increases in the same order as in aliphatic SN2 reactions, but much more steeply, by 14 orders of
magnitude in the row log{k
0 =k½CpMoðCOÞ ŠK}: [CpFe(CO)2]S (ꢀ14), [Re(CO)5]S (7.8), [Mn(CO)5]S 2.1, [CpW(CO)3]S
½MðCOÞ LŠM
n
3
(0.7) > [CpMo(CO)3]S (0). A good correlation exists between nucleophilicities of [M(CO)nL]S anions towards vinyl
(sp2-carbon) and alkyl halides (sp3-carbon) with slope 2.7. The reactivity of [M(CO)nL]S in a halogen–metal exchange
S
—
process (with Z-PhC(CN) CHI) follows a similar ‘large’scale as in the Ad E process. The nucleophilicity of [M(CO) L]
—
N
n
anions correlates better with their one-electron oxidation potentials (Eox) than with their basicity (pKa of [M(CO)nL]H).
Copyright ß 2008 John Wiley & Sons, Ltd.
Supplementary electronic material for this paper is available in Wiley InterScience at http://www.mrw.interscience.wiley.com/
suppmat/0894-3230/suppmat/
Keywords: nucleophilicity; metal carbonyl anions; SNV; nucleophilic vinylic substitution; reaction mechanisms
characterized both by
a remarkably high reactivity of
INTRODUCTION
[CpFe(CO)2]À anion, often called a ‘supernucleophile’, and a very
broad reactivity span between [CpFe(CO)2]À and the least
reactive [Co(CO)4]À which exceeds seven orders of magnitude:
The studies of aliphatic nucleophilic substitution reactions by C.K.
Ingold laid the foundations of physical organic chemistry and the
À
À
À
À
À
À
½CpFeðCOÞ2Š
6 Á 107
>>> ½ReðCOÞ5Š
2 Á 104
>> ½MnðCOÞ5Š
>
½CpWðCOÞ3Š
>
½CpMoðCOÞ3Š
>> ½CoðCOÞ4Š
170
50
35
1
problem of nucleophilic reactivity has been an area of close
interest ever since.[1] An important landmark in this field was the
constant selectivity relationship (log(k/k0) ¼ Nþ) discovered by
Ritchie for the nucleophile addition reactions to stabilized
carbocations.[2] Later it was shown that the Ritchie relationship
also describes the nucleophilicity in the reactions with other
p-acceptors, such as carbonyl compounds, activated alkenes,[3]
transition metal p-complexes[4] and even in nucleophilic
aromatic substitution.[5] Recent progress in the field is
associated with the works of Mayr et al., who demonstrated
that a nucleophile-specific selectivity factor has to be considered
for an accurate description of all p-electrophile–nucleophile
combinations. Using the proposed equation log k ¼ sN Á (N þ E)[6]
they have quantified the reactivity of a great number of
nucleophiles (and p-electrophiles) of various classes and
activity.[7–10] Quite recently it has been shown that the
nucleophilicity in aliphatic SN2 reactions can also be described
by the N parameters provided that the electrophile-specific
selectivity factor (sE ¼ 0.6) is incorporated into the equation: log
k ¼ sE Á sN Á (N þ E).[7]
(Averaged data from Reference [12,13]. Data for [CpFe(CO)2]À taken
from Reference [11]).
In our studies of nucleophilic vinylic substitution reac-
tions[15–18] we became increasingly aware that the differences
in reactivity between metal carbonyl anions are much greater
than found in aliphatic SN2 reactions. The reactivity differences
(vide infra) were so large that we could not compare all the metal
carbonyl anions in the reactions with a single vinyl halide. Data for
iron and rhenium carbonylates were obtained with a,b,b-
trifluorostyrene, weakly nucleophilic carbonylates of manganese,
tungsten and molybdenum had to be compared in the reactions
with a highly electrophilic a-chlorobenzylidenemalononitrile.
Bridging these two reactivity series turned out to be a most
difficult task, and was finally achieved by comparing [Re(CO)5]À
and [Mn(CO)5]À in the reactions with b-halo-a-phenylacry-
lonitriles (Hal ¼ Cl, Br, I).
*
Chemistry Department, M.V. Lomonosov Moscow State University, Leninskie
Gory 1, Moscow 119992, Russian Federation.
E-mail: petr@elorg.chem.msu.ru
Metal carbonyl anions represent a family of useful and easily
accessible model metal-centre nucleophiles.[11] Their nucleophi-
licity was studied in aliphatic SN2 reactions by Dessy et al.,[12]
Pearson and Figdore[13] and later Atwood and co-workers.[14] It is
a
P. K. Sazonov, G. A. Artamkina, I. P. Beletskaya
Chemistry Department, M.V. Lomonosov Moscow State University, Leninskie
Gory 1, Moscow 119992, Russian Federation
J. Phys. Org. Chem. 2008, 21 198–206
Copyright ß 2008 John Wiley & Sons, Ltd.

NUCLEOPHILICITY OF METAL CARBONYL ANIONS
Scheme 1. The addition–elimination (AdNE) mechanism of nucleophilic vinylic substitution
Scheme 2. The halogen–metal exchange (HME) mechanism of nucleophilic vinylic substitution
A comparison of reactivity makes sense only if the reactions
under consideration follow the same mechanism. This is
particularly true of the nucleophilic vinylic substitution reactions
in view of a large diversity of possible mechanisms.[19] However,
in the reactions with metal carbonyl anions the choice is usually
between two alternative pathways, addition–elimination (AdNE
Scheme 1) and halogen–metal exchange (HME). The first one
(AdNE) is the most common pathway of substitution in vinyl
halides activated with b-EWG.[20–22]
– formation of s-vinylic complexes instead of halo(acyl)rhenate
complexes in the reactions with [Re(CO)5]Na, which is charac-
teristic of HME mechanism (Scheme 2);
In our opinion, they can be interpreted in the framework
of AdNE mechanism (Scheme 1). The observed retention of Z/
E-configuration of the starting vinyl halide is consistent with this
mechanism and indicates that the s-adduct intermediate is very
short lived, or even does not exist.[20,23,24] (In this latter case, a
The HME pathway (Scheme 2) comes into play for the
substrates with ‘heavier’ halogen leaving groups (Br, I and
sometimes even Cl), with substituents stabilizing the vinyl
carbanion (e.g. polyfluorinated) and steric hindrance to the attack
of nucleophile at the double bond.[18]
one-step nucleophilic substitution with
a transition state
resembling a s-adduct would occur.) The vinyl halides we have
used are activated with b-EWG and have been shown previously
to react via AdNE mechanism with other nucleophiles such as
amines or thiolates.[25–27]
Reactions we used for the comparison of nucleophilicity do not
follow the HME mechanism, which is evident from
The possibility of an alternative single-electron transfer (SET)
pathway should not be overlooked (Scheme 3), especially since
metal carbonyl anions are potent reducing agents.[28] Similarly a
SET-version may be drawn for HME pathway.
There is, however, no experimental evidence of a pathway
involving free radicals. Reactions follow second order kinetics
without inductive period, the yield of nucleophilic substitution
products is high and no homocoupling or protodehalogenation
– high yield of s-vinylic complexes as products, which does
not become lower in the presence of ‘anion traps’. Anion
traps protonate the vinyl carbanions, the formation of nucleo-
philic substitution products in case of HME mechanism
(Scheme 2);
Scheme 3. The single-electron transfer (SET) mechanism of nucleophilic vinylic substitution
www.interscience.wiley.com/journal/poc
Copyright ß 2008 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2008, 21 198–206

P. K. SAZONOV, G. A. ARTAMKINA AND I. P. BELETSKAYA
Table 1. The observed rate constants (kobs) for the reaction of [CpFe(CO)2]K and Re(CO)5Na with a,b,b-trifluorostyrene (1),
b-chloro-a,b-difluorostyrene (1-Cl)[15,16] and perfluoro-3,3-dimethyl-1-butene (2)[17]
Vinyl halide
Entry
R
Hal
[M(CO)nL]M0
T (8C)
kobs (L Á molÀ1 Á sÀ)1
1
2
3
4
5
6
1
1
1-Cl
1
2
2
Ph
Ph
Ph
Ph
F
F
Cl
F
F
F
[CpFe(CO)2]K
[CpFe(CO)2]K
[CpFe(CO)2]K
[Re(CO)5]Na
[Re(CO)5]Na
[Mn(CO)5]K
À75
À23
À23
þ21
þ21
þ21
1.6
30
0.37
0.00020
0.0012
ꢁ10À7
(CF3)3C–
(CF3)3C–
products are formed (Supplementary material). When the reactions
are monitored by 1H NMR, no CIDNP effects are observed. Retention
of vinyl halide Z/E-configuration is not consistent with a radical
anion intermediate either. When the reactions of 1, 1-Cl or 2 with
[CpFe(CO)2]K were performed with the electrochemical one-
electron reduction of the vinyl halide, mixtures of Z- and E-isomers
of the substitution products were formed.[29]
[CpMo(CO)3]À (1)[12–14]), although the order of reactivity is
retained. The four-fold difference between [CpW(CO)3]À and
[CpMo(CO)3]À is particularly notable, since these anions have very
similar steric requirements and almost the same reactivity in
aliphatic SN2 reactions.
b-Halo-a-phenylacrylonitriles (4, Hal ¼ Cl, Br, I) proved to be
especially useful models. Both Z- and E-isomers of 4-Cl and 4-Br
(Hal ¼ Cl, Br) gave the s-vinylic complexes with quantitative
yields. Reaction with Z or E isomers of 4 gave in each case only
one isomer of substitution product. We assume that this result
indicates retention of Z/E configuration of the starting vinyl
halide. The assignment of Z/E configuration of the products is
supported by the lower nCN in the IR spectra of E-isomers of the
products in accordance with the literature data on b-substituted
a-phenylacrylonitrile derivatives.[25] It is also supported by the
results of DFT calculations of the vibrational frequencies in the
corresponding products. The geometry assignment is also
confirmed by chemical shift of vinyl protons in the 1H NMR
spectra, which are observed in higher field for the E-isomers of
the products, the same as in the starting vinyl halides. Anyway, it
RESULTS AND DISCUSSION
The first set of reactivity data given in Table 1 were obtained for
the reactions with trifluorovinyl derivatives (1) and (2). Reactions
with [CpFe(CO)2]K were very fast even at À758C, and the reaction
rate could be measured only for the less reactive trifluorostyrene
(entries 1,2). Substitution of one b-fluorine in trifluorostyrene for
chloride (in 1-Cl) lowers the reactivity by two orders of magnitude
(entry 3). This ‘element effect’ (kF/kCl >> 1) provides additional
support for the AdNE mechanism.[20]
In contrast, reactions of 1 and 2 with [Re(CO)5]Na were slow at
room temperature (entries 4, 5), and even the more reactive
substrate 2 did not react with [Mn(CO)5]K (entry 6). Only an
estimate of [CpFe(CO)2]K: [Re(CO)5]Na reactivity ratio could be
obtained by the extrapolation to 218C of the [CpFe(CO)2]K data
(kobs at À758C and À238C) according to Arrhenius equation. This
Table 2. The observed rate constants (kobs) for the reaction of
a-chlorobenzylidenemalononitrile (3) with metal carbonyl
anions ([M(CO)nL]M0), THF, 228C
gives the
k
½CpFeðCOÞ ŠK ꢀ 2  102 L Á molÀ1 Á sÀ1 at 218C and
2
k
½CpFeðCOÞ ŠK=k½ReðCOÞ ŠNa ꢀ 106 which is nearly three magnitudes
2
5
higher than observed in aliphatic SN2 reactions.[12]
.
a-Chlorobenzylidenemalononitrile (3) belongs to the highly
electrophilic vinyl halides that can react with moderately weak
nucleophiles, such as anilines.[26,27] In the reactions of 3 with
carbonylmetallates the nucleophilic substitution products are
formed quantitatively. Their formation is not suppressed in the
presence of t-BuOH – a negative test result for the HME
mechanism (Scheme 2).
The reactivity difference between the metal carbonyl anions in
the reaction with 3 (Table 2) is significantly larger than in
aliphatic SN2 reactions ([Mn(CO)5]À (5) > [CpW(CO)3]À (1.4) >
[M(CO)nL]M0
kobs (L Á molÀ1 Á sÀ)1
Reactivity ratio
[Re(CO)5]Na
[Mn(CO)5]K
[CpW(CO)3]K
[CpMo(CO)3]K
Too fast
30
0.7
—
170
4
0.17
1
J. Phys. Org. Chem. 2008, 21 198–206
Copyright ß 2008 John Wiley & Sons, Ltd.
www.interscience.wiley.com/journal/poc

NUCLEOPHILICITY OF METAL CARBONYL ANIONS
Table 3. The observed rate constants (kobs) for the reaction of b-halo-a-phenylacrylonitriles (4, Hal ¼ Cl, Br) with metal carbonyl
anions ([M(CO)nL]M0), THF, 228C
Vinyl halide
Hal
[M(CO)nL]M0
kobs (L Á molÀ1 Á sÀ1
)
Reactivity ratios
4-Z-Cl
4-Z-Cl
4-E-Cl
4-E-Cl
4-Z-Br
4-Z-Br
4-Z-Br
4-E-Br
4-E-Br
4-E-Br
Cl
Cl
Cl
Cl
Br
Br
Br
Br
Br
Br
Re(CO)5Na
Mn(CO)5K
Mn(CO)5K
[CpW(CO)3]K
Mn(CO)5K
[CpW(CO)3]K
[CpMo(CO)3]K
Mn(CO)5K
[CpW(CO)3]K
[CpMo(CO)3]K
350
5 Â 105:1
0.0007
0.0028
0.00010
0.011
0.0013
0.00016
0.04
28:1
70:8:1
170:5:1
0.0011
0.00024
is highly unlikely that reaction with activated vinyl halide such as
4 will proceed with complete inversion of configuration.
Anion trap (t-BuOH) had no effect on the course (and rate) of
the reactions with 4-Br (both Z- and E-isomers).
The observed rate constants summarized in Table 3 again show
a very large difference in reactivity between manganese,
tungsten and molybdenum carbonylmetallates, similar to those
observed in the reactions with 3. More importantly, the rate
constant for the reaction of 4-Z-Cl with [Re(CO)5]Na allows to
compare its reactivity with that of less reactive carbonylmetal-
lates. The [Mn(CO)5]K: [Re(CO)5]Na reactivity ratio is 1:500 000,
which may be compared to 1:120 ratio typical for aliphatic SN2
reactions.[13,14]
Both Z and E-isomers of 4-Br turned out to be more reactive
than the corresponding chlorides 4-Cl. The intermolecular
element effect kBr/kCl reaches 15, the highest value observed
so far in AdNE nucleophilic vinylic substitution reactions. Typically
kBr/kCl lie in the range 0.6–7,[20] the maximum value of 11 having
being reported by Rappoport, Shainyan and co-workers.[24]
The great difference between the reactivity span of carbo-
nylmetallates in aliphatic SN2 and vinylic AdNE substitution
reactions is shown visually in Fig. 1, where the observed
log(kVinHal) values are plotted against log(kCH I). It can be seen that
3
the lines for all the vinyl halides except the trifluorostyrene have
almost the same and rather steep slope varying from 2.5 to 3.2.
The slope for triflurostyrene (reactions with [CpFe(CO)2]K and
[Re(CO)5]Na) is lower (1.7) but still much higher than unity.
A similar selectivity observed for the reactions of 3, 4-Cl and
4-Br with manganese, tungsten and molybdenum carbonylme-
tallates allows one to average the corresponding reactivity ratios.
Taking into account the [Re(CO)5]Na:[Mn(CO)5]K ratio (5 Â 105)
from the reaction with 4-Z-Cl and [CpFe(CO)2]K:[Re(CO)5]Na
(ꢀ106) ratio from the reaction with 1, the following ranking of
Figure 1. The observed rate constants {log(kVinHal)} for the reaction of
metal carbonyl anions with vinyl halides plotted against the rate con-
stants with methyl iodide {log(kCH )
[13,14]}.
I
3
www.interscience.wiley.com/journal/poc
Copyright ß 2008 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2008, 21 198–206

P. K. SAZONOV, G. A. ARTAMKINA AND I. P. BELETSKAYA
metal carbonyl anions reactivity in AdNE nucleophilic vinylic
0
reactivity of ordinary nucleophiles in AdNE vinylic substitution
reactions usually increases in the same (or slightly larger)
substitution reactions is obtained (k½MðCOÞ LŠM =k½CpMoðCOÞ ŠK):
n
3
À
À
À
À
À
½CpFeðCOÞ2Š
ꢀ 1014
>> ½ReðCOÞ5Š
7 Á 107
>> ½MnðCOÞ5Š
130
>
½CpWðCOÞ3Š
>
½CpMoðCOÞ3Š
5
1
It is shown graphically in Fig. 2. A good linear correlation with
slope 2.16 is observed between log {k
proportion as in aliphatic SN2 reactions, that is, the slopes are near
or slightly above unity. The different slopes make the comparison
of [M(CO)nL]À anion nucleophilicity to that of ordinary
nucleophiles ambiguous, the reactivity order being dependent
on the reference substrate. However, the [CpFe(CO)2]À anion still
holds the lead, confirming its ‘supernucleophile’ status, as can be
seen from the following rankings:
0 =k½CpMoðCOÞ ŠK} and
½MðCOÞ LŠM
n
3
log (kCH I) for all the data (r ¼ 0.992), which improves (r ¼ 0.9996,
3
slope 2.75) if the data point for the [CpFe(CO)2]K (reaction with
trifluorostyrene) is excluded. Such steep slopes are uncommon
for the reactions of ordinary carbon, nitrogen or sulphur
nucleophiles with the same (or similar) vinyl halides. The
À
À
À
À
À
in reactions with 1 :15;30 ½CpFeðCOÞ2Š
>
½9 À MeFlŠ
12
>
½PhSŠ
>
½EtOŠ
>
½ReðCOÞ5Š
2 Á 10À4
kobsðL Á molÀ1 Á sÀ1Þ :
ꢀ 2 Á 102
ꢀ 10À3ðRef:31Þ
½9 À MeFlŠÀis the 9 À methylfluorenide anion;
À
À
À
À
in reactions with 4 À Z À Cl :25 ½ReðCOÞ5Š
>
½p À TolSŠ
34
>
½EtOŠ
ꢂ
ꢀ 0:10
piperidine
>
½MnðCOÞ5Š
0:0007;
kobsðL Á molÀ1 Á sÀ1Þ :
350
À
in reactions with aÀhaloarylidenemalononitriles :26;27 ½MnðCOÞ5Š
30
ꢃ
piperidine
>
ArNH2
kobsðL Á molÀ1ÁsÀ1Þ :
10À2 À 10À3
:
The relative magnitude of nucleophilicity changes in different
reactions can be compared through the corresponding Brønsted
bNuc values. In the reaction of 9-substituted fluorenyl carbanions
with trifluorostyrene (1) the slope of the Brønsted plot
(bNuc ¼ 0.37)[32] is almost the same as in SN2 reactions of these
carbanions (bNuc ¼ 0.35).[32,33] The ratio of reactivity of morpho-
line and piperidine in reactions with 4-Cl[25] and the p-nitro
analogue of 3[26] allows one to estimate bNuc ¼ 0.3–0.35. These
values fall in the range quite normal for aliphatic SN2 reactions.
Similar bNuc values are estimated for the p-MeC6H4SÀ and
p-ClC6H4SÀ pair in reactions with 4-Cl.[25] While the nucleophi-
licity of metal carbonyl anions poorly correlates with their basicity
(Fig. 3), the reactivity differences found in nucleophilic vinylic
substitution with metal carbonyl anions (Fig. 2) correspond to
bNuc values which are 2.16–2.75 times higher than in aliphatic SN2
reactions, that is, bNuc about unity.
However, unusually high Brønsted bNuc values were observed
in nuclephilic vinylic and aromatic substitution with weak soft
nucleophiles in highly activated systems. Values of bNuc higher
than unity (up to 1.4) have been recently reported for the
reactions of 4-nitrobenzofurazan derivatives with 4-substituted
anilines.[34] Analysis of earlier published literature data[27] on the
reactions of anilines with a-halo(p-dimethylaminobenzylidene)
malononitriles also gives high bNuc values that are about unity.
One should bear in mind the importance of rate-limiting
stage of the AdNE substitution mechanism for the interpretation
of the nucleophilicity data. The observed rate constants reflect
the nucleophilicity of metal carbonyl anions (kobs ¼ k1) as long as
the first (addition) stage is rate limiting. But if it is not, the
observed rate constant becomes composite (Scheme 1) and in
Figure 2. Correlation between the nucleophilicity of metal carbonyl
k
1
the extreme case
k
>> kel ) kobs
¼
Á kel ¼ K Á kel will
À1
0
anions in AdNE vinylic substitution reactions {log(k
K)} and SN2 aliphatic nucleophilic substitution {log(kCH
=k
k
½MðCOÞ LŠM
½CpMoðCOÞ
Š
reflect the thermodynamic affinity (equiÀli1brium, K) of the
nucleophiles. A priori metal carbonyl anions should be good
n
3
)
[13,14]}. This figure
is available in colour online at www.interscience.wiley.com/journal/poc
I
3
J. Phys. Org. Chem. 2008, 21 198–206
Copyright ß 2008 John Wiley & Sons, Ltd.
www.interscience.wiley.com/journal/poc

NUCLEOPHILICITY OF METAL CARBONYL ANIONS
leaving groups (though their nucleofugality has never been
quantified), however, it seems highly unlikely that they would be
better leaving groups than chloride or bromide ions, and the
condition of the rate-limiting addition kÀ1 ꢄ kel probably holds
true for the studied systems. Yet, the alternative possibility,
cannot be completely ruled out.
Vinyl halides with heavier halogens (Hal ¼ Br and I, but
sometimes even Cl) can react with metal carbonyl anions via
the HME mechanism (Scheme 2).[18] While no indications of HME
were found in the reactions with Z- and E-isomers of 4-Br, the
corresponding iodide 4-Z-I reacted with metal carbonyl anions by
two competing pathways AdNE and HME. In the presence of anion
trap (t-BuOH) 4-Z-I gave mixtures of substitution (AdNE) and
protodehalogenation products (Table 4), the latter resulting from
the protonation of the vinyl carbanion generated by HME. The
carbanionic adduct intermediate in the AdNE process is not
protonated by t-BuOH most probably because of its low basicity.
In the reaction with [Re(CO)5]Na the HME is evident even without
the anion trap by the formation of halo(acyl)rhenate complex
instead of the ‘normal’ substitution product (Scheme 4). The vinyl
carbanions generated by HME attack Re(CO)5Hal at the
coordinated CO instead of the metal centre:[18]
—
Nucleophilic substitution products, Z-Ph(CN)C CHM(CO) L,
—
n
are formed primarily by the AdNE pathway, since their yields do
not change much in the presence of t-BuOH.
Surprisingly, the fraction of HME products grows up with the
increase of metal carbonyl anion nucleophilicity from [CpMo
(CO)3]K to [Re(CO)5]Na (Table 4). It means that the reactivity of
[M(CO)nL]À in the HME process increases even more steeply than
in AdNE reactions (Fig. 4). However, the differences in HME/
AdNE ratio are much smaller than the overall increase in reactivity
from [CpMo(CO)3]K to [Re(CO)5]Na. So, the overall result is that
reactivity of metal carbonyl anions in HME and AdNE reactions
Figure 3. Correlation between the nucleophilicity of metal carbonyl
anions in AdNE vinylic substitution reactions {log(k
0 =k
ŠK)}
½CpMoðCOÞ
½MðCOÞ LŠM
n
3
and their basicity {pKa of [M(CO)nL]H}.[43] This figure is available in colour
online at www.interscience.wiley.com/journal/poc
Table 4. The observed rate constants (kobs) and product ratios in the reactions of Z-b-iodo-a-phenylacrylonitrile (4-Z-I) with metal
carbonyl anions ([M(CO)nL]M0), THF, 228C
Yields of (%)
kobs
(L Á molÀ1 Á sÀ)1
Reactivityb of
[M(CO)nL]M0 in HME
[M(CO)nL]M0
% of HMEa
[Re(CO)5]Na
[Mn(CO)5]K
[CpW(CO)3]K
[CpMo(CO)3]K
60c
30
8
ꢁ2
50
75
65
97
40
10
15
Too fast
0.02
0.003
ꢀ1 Â 1010d
300
10
1
12
0.0002
a
—
The percent of HME is calculated from the ratio of the yield of HME product (Ph(CN)C CH ) to the total product yield.
—
2
0
b Calculated as kH½MMðECOÞ LŠM =kH½CMpEMoðCOÞ ŠK, where kHME ¼ kobs Á ð% of HMEÞ=100:
n
3
c
—
Without t-BuOH 65% of halo(acyl)rhenate [Z-PhC(CN) CH(CO)Re(CO) I]Na is formed.
—
4
½MnðCOÞ ŠK
kHME calculated from the reactivity ratio in the AdNE reactionðk½ReðCOÞ ŠNa=kAd
Þ with 4-Z-Cl:
d
5
5
Ad
N
E
E
N
½ReðCOÞ ŠNa
5
k
½MnðCOÞ ŠK
½ReðCOÞ ŠNa
5
½ReðCOÞ ŠNa
½MnðCOÞ ŠK
Ad
E
ð1À%HMEÞ
5
5
%HME
5
N
kHME
¼
Á ð1À%HMEÞ
Á
Á kobs
.
½MnðCOÞ ŠK
½ReðCOÞ ŠNa
100
5
5
k
Ad
E
N
www.interscience.wiley.com/journal/poc
Copyright ß 2008 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2008, 21 198–206

P. K. SAZONOV, G. A. ARTAMKINA AND I. P. BELETSKAYA
Scheme 4. The HME mechanism of the halo(acyl)rhenate formation.
follows a similar scale in spite of the entirely different mecha-
nisms and reaction centres involved (sp2-carbon vis halogen).
As can be seen from Fig. 5 nucleophilicity of metal carbonyl
anions better correlates with their one-electron oxidation
potentials (Eox)[28] than with their basicity (Fig. 3). The slope of
this correlation is again unusually steep. If both the rate data and
the reduction potentials are converted into free energy terms, the
slope of the correlation (for the AdNE substitution) will be near unity
(0.97, Eqn (1)). It indicates that the changes in the free energies of
It is tempting to interpret these data in terms of a SET mecha-
nism for both the AdNE (Scheme 3) and HME processes. Further
still, steep nucleophilicity slopes (Figs 1 and 2) are equivalent to
high Brønsted bNuc values (ꢃ1), which previously have been
interpreted as an indication of a complete SET prior to the
coupling of nucleophilic and electrophilic partners.[34–38] Similar
nucleophilicity slopes for HME and AdNE (Fig. 4) also suggest a
common or similar rate-limiting stage for these processes.
However, the outer sphere SET pathway is not supported by any
‘chemical’ evidence and does not seem feasible, the difference
between the Eox of the donor ([M(CO)nL]À)[14,28] and the Ered of
the acceptor (4-Hal) being too large. Eox of [M(CO)nL]À decreases
from À0.52 V for [Re(CO)5]À to À0.33 V for [CpW(CO)3]À,[28] Ered
are À1.70 V for 4-Z-Cl, À1.65 V for 4-Z-Br and À1.60 V for 4-Z-I
(O.M. Nikitin and T.V. Magdesieva, personal communication, All E
values are converted to SHE scale). The inner-sphere SET pathway,
which in this context implies an electron-transfer more or less
concerted with the bond formation between the electrophile and
the nucleophile, is a viable alternative. In the framework of single
¼
activation DG of the AdNE nucleophilic vinylic substitution
reaction are almost the same as the changes in the free energies
of one-electron oxidation of metal carbonyl anions DG8ET.
!
½MðCOÞnLŠM0
k
RT
AdN
E
log
Á
½CpMoðCOÞ3ŠK
logð2:718Þ
k
AdN
E
½MðCOÞnLŠM0
¼ 0:97 Á F Á EOX
À 38500
(1)
(F is the Faraday constant).
Figure 5. Correlation between the nucleophilicity of metal carbonyl
Figure 4. Nucleophilicity of metal carbonyl anions in AdNE vinylic sub-
—
—
anions {log(k
0 =k
ŠK)} and their one-electron oxidation
½CpMoðCOÞ
stitution and halogen–metal exchange (HME) reactions with Z-PhCCN
½MðCOÞ LŠM
n
3
potentials Eox (from Reference [28], versus ferrocene/ferrocenium).
d
CHI. The point for [Re(CO)5]À is an estimate (footnote to Table 4).
J. Phys. Org. Chem. 2008, 21 198–206
Copyright ß 2008 John Wiley & Sons, Ltd.
www.interscience.wiley.com/journal/poc

NUCLEOPHILICITY OF METAL CARBONYL ANIONS
electron shift concept[39] it may be regarded as the initial stage in
the transition form polar to SET pathway. However, the distinction
between the inner-sphere SET and the polar AdNE pathway
remains unclear at present.
were calculated by linear regression according to Eqn (2) giving
good fit (r > 0.99) with experimental data.
1
½MÀŠ0 Á ½VinHalŠ
½VinHalŠ0 Á ½MÀŠ
Á ln
¼ kobs Á t
(2)
½VinHalŠ0 À ½MÀŠ
0
EXPERIMENTAL
The progress of reactions too fast for NMR method
(kobs > 5 Â 10À2 L Á molÀ1 Á sÀ1) was followed by UV spectroscopy
by measuring the absorbance (A) changes in 280–360 nm range.
The quartz cell was washed with a large excess of metal carbonyl
salt solution that was then diluted to working concentration
(about 5 Â 10À4 mol Á LÀ1) and the cell was sealed under vacuum.
The experiments were usually run with a slight excess of vinyl
General
1H NMR (400.13 MHz) and 13C NMR (100.61 MHz) spectra were
obtained on a Bruker Avance spectrometer at 228C and
referenced to the signals of the solvent. UV-spectra and kinetic
measurements were run on Hewlett-Packard-8452A diode array
spectrophotometer. Preparation of metal carbonyl anion salts
and their reactions with alkenyl halides were carried out in
‘all-fused’ glassware using vacuum-line techniques. Break-seal
ampoules were used for the transfer of substances. Experiments
with polyfluorinated vinyl halides 1, 1-Cl and 2 have been
described previously.[15–17]
halide. Reaction coordinate
x was calculated for several
wavelengths as x ¼ ðA À A0Þ=ðA1 À A0Þ from which the con-
centrations of reagents were determined ([MÀ] ¼ [MÀ]0 Á (1 À x);
[VinHal] ¼ [VinHal]0 À [MÀ] Á x) and fitted to the Eqn (2) to get
second order rate constants.
Acknowledgements
Materials
The work was supported by the Grant of the President of the
Russian Federation for the State Support of the Leading Research
Schools (Grant No. HIII-6059.2006.3).
THF was stored over sodium benzophenone ketyl and
vacuum-transferred to the reaction vessels. Potassium salts of
metal carbonyl anions were obtained quantitatively (95–98%) by
reductive cleavage of the corresponding dimers [M(CO)nL]2 with
excess of NaK2.8 alloy (0.10–0.15 mL per 0.5 mmol of dimer) in
THF.[40] [Re(CO)5]Na was prepared by the reduction of Re2(CO)10
with 0.5% NaHg (30–50% excess) and purified by the low
temperature crystallization from THF.[16] Vinyl chlorides 3,[41]
4-Z-Cl and 4-E-Cl[42] were obtained according to published
procedures. Compounds 4-Z-Br and 4-E-Br were prepared
similarly to 4-Cl and vinyl iodide 4-Z-I was obtained from
4-Z-Br and NaI as described in the supporting material. Small
portions of reagents required for a single experiment were sealed
under vacuum into thin-walled glass vials.
REFERENCES
[1] M. Harris, S. P. McManus, (eds). In Nucleophilicity, Adv. Chem. Ser, Vol.
215, American Chemical Society: Washington, DC, 1987.
[2] C. D. Ritchie, Acc. Chem. Res. 1972, 5, 348–354.
[3] S. Hoz, D. Speizman, J. Org. Chem. 1983, 48, 2904–2910.
[4] L. A. P. Kane-Maguire, E. D. Honig, D. A. Sweigart, Chem. Rev. 1984, 84,
525–543.
[5] C. D. Ritchie, M. Sawada, J. Am. Chem. Soc. 1977, 99, 3754–3761.
[6] H. Mayr, M. Patz, Angew. Chem. Int. Ed. 1994, 33, 938–957.
[7] T. B. Phan, M. Breugst, H. Mayr, Angew. Chem. Int. Ed. 2006, 45,
3869–3874.
[8] H. Mayr, B. Kempf, A. R. Ofial, Acc. Chem. Res. 2003, 36, 66–77.
[9] S. Minegishi, H. Mayr, J. Am. Chem. Soc. 2003, 125, 286–295.
[10] A. R. Ofial, K. Ohkubo, S. Fukuzumi, R. Lucius, H. Mayr, J. Am. Chem.
Soc. 2003, 125, 10906–10912.
[11] R. B. King, Acc. Chem. Res. 1970, 3, 417–427.
[12] R. E. Dessy, R. L. Pohl, R. B. King, J. Am. Chem. Soc. 1966, 88,
5121–5124.
Reaction products
The nucleophilic substitution products (metal carbonyl s-vinyl
complexes) were isolated by column chromatography on SiO2
and characterized by NMR, IR, and MS or analytical data as
described in the supporting material. The Product yields were
determined by the integration of the 1H NMR spectra of the crude
reaction solutions using an internal standard (durene). The
signals were referenced to the individual products by comparison
to the spectra of the isolated compounds or with the literature
data for the compounds previously described (such as metal
carbonyl dimers).
[13] R. G. Pearson, P. E. Figdore, J. Am. Chem. Soc. 1980, 102, 1541–1547.
[14] C. K. Lai, W. G. Feighery, Y. Q. Zhen, J. D. Atwood, Inorg. Chem. 1989,
28, 3929–3930.
[15] G. A. Artamkina, M. M. Shtern, P. K. Sazonov, I. P. Beletskaya, Zhurn.
Org. Khimii. 1996, 32, 1319–1328.
[16] P. K. Sazonov, M. M. Shtern, G. A. Artamkina, I. P. Beletskaya, Zhurn.
Org. Khimii. 1998, 34, 1807–1821.
[17] P. K. Sazonov, G. A. Artamkina, I. P. Beletskaya, Russ. J. Org. Chem.
2000, 36, 943–950.
Kinetic measurements
[18] P. K. Sazonov, V. A. Ivushkin, G. A. Artamkina, I. P. Beletskaya, Arkivoc
2003, Part (X), 323–334.
Relatively slow reactions (kobs < 5 Â 10À2 L Á molÀ1 Á sÀ1) were
monitored by 1H NMR spectroscopy, with initial concentration of
metal carbonyl anion 0.05–0.20 mol Á LÀ1 and a slight excess of
vinyl halide. The reactions were carried out in undeuterated THF
in vacuum-sealed glass tubes (ꢀ3.5 mm diameter) which were
placed into standard NMR tubes containing acetone-d6 for the
lock signal. The concentrations of the starting [VinHal] and
resulting [VinM] vinyl compounds were determined by the
integration of the spectra using internal standard-durene. The
concentration of metal carbonyl salt [MÀ] was determined
similarly except for [M(CO)5]K that was calculated as difference:
[19] Z. Rappoport, Rec. Trav. Chim. 1985, 104, 309–349.
[20] Z. Rappoport, Acc. Chem. Res. 1981, 14, 7–15.
[21] Z. Rappoport, Acc. Chem. Res. 1992, 25, 474–479.
[22] B. A. Shainyan, Uspekhi. Khimii. 1986, 55, 942–973.
[23] Y. Apeloig, Z. Rappoport, J. Am. Chem. Soc. 1979, 101, 5095–5098.
[24] M. Beit-Yannai, Z. Rappoport, B. A. Shainyan, Y. S. Danilevich, J. Org.
Chem. 1997, 62, 8049–8057.
[25] Z. Rappoport, A. Topol, J. Chem. Soc. Perkin Trans. 2, 1975, 982–988.
[26] Z. Rappoport, A. Topol, J. Chem. Soc. Perkin Trans. 2, 1975, 863–874.
[27] Z. Rappoport, R. Ta-Shma, J. Chem. Soc. B. 1971, 871–881.
[28] M. Tilset, V. D. Parker, J. Am. Chem. Soc. 1989, 111, 6711–6717.
[29] I. I. Kukhareva, T. V. Magdesieva, G. A. Artamkina, I. P. Beletskaya, K. P.
Butin, Russ. Chem. Bull. 1996, 45, 1452–1457.
[MÀ] ¼ [MÀ]0À[VinM]. The second order rate constants (kobs
)
www.interscience.wiley.com/journal/poc
Copyright ß 2008 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2008, 21 198–206

P. K. SAZONOV, G. A. ARTAMKINA AND I. P. BELETSKAYA
[30] G. A. Artamkina, M. M. Shtern, P. K. Sazonov, I. P. Beletskaya, Russ. J.
Org. Chem. 1996, 32, 1281–1285.
[36] F. G. Bordwell, A. H. Clemens, D. E. Smith, J. Begemann, J. Org. Chem.
1985, 50, 1151–1156.
[31] H. F. Koch, A. J. Kielbania, J. Am. Chem. Soc. 1970, 92, 729–730.
[32] A. A. Solov’yanov, M. M. Shtern, I. P. Beletskaya, O. A. Reutov, Zhurn.
Org. Khimii. 1983, 19, 2233–2240.
[33] A. A. Solov’yanov, P. I. Dem’yanov, I. P. Beletskaya, O. A. Reutov, Zhurn.
Org. Khimii. 1976, 12, 2489–2496.
[34] F. Terrier, M. Mokhtari, R. Goumont, J.-C. Halle, E. Buncel, Org. Biomol.
Chem. 2003, 1, 1757–1763.
[35] F. G. Bordwell, A. H. Clemens, J. Org. Chem. 1981, 46, 1035–
1037.
[37] V. M. Vlasov, I. A. Os’kina, D. Landini, A. Maia, Russ. Chem. Bull. 1995,
44, 2211–2215.
[38] V. M. Vlasov, I. A. Os’kina, Russ. J. Gen. Chem. 2004, 74, 600–605.
[39] A. Pross, Acc. Chem. Res. 1983, 18, 212–219.
[40] J. E. Ellis, E. A. Flom, J. Organomet. Chem. 1975, 99, 263–268.
[41] K. Friedrich, H. K. Thieme, Synthesis 1973, 111–112.
[42] M. Cariou, Bull. Soc. Chim. Fr. 1969, 217–222.
[43] E. J. Moore, J. M. Sullivan, J. R. Norton, J. Am. Chem. Soc. 1986, 108,
2257–2263.
J. Phys. Org. Chem. 2008, 21 198–206
Copyright ß 2008 John Wiley & Sons, Ltd.
www.interscience.wiley.com/journal/poc