Physical organic chemistry of transition metal carbene complexes. 19. Kinetics of reversible alkoxide ion addition to substituted (methoxyphenylcarbene)pentacarbonylchromium(0) and (methoxyphenylcarbene)pentacarbonyltungsten(0) in methanol and aqueous ace

Article abstract of DOI:10.1021/ja994174w

Rate and equilibrium constants for the nucleophilic addition of MeO^- in methanol and in 90% MeCN-10% MeOH, of HC≡CCH₂O^- and CF₃CH₂O^- in 50% MeCN-50% water, and of OH^- in 50% MeCN

Full text of DOI:10.1021/ja994174w

J. Am. Chem. Soc. 2000, 122, 3821-3829
3821
Physical Organic Chemistry of Transition Metal Carbene Complexes.
19.1 Kinetics of Reversible Alkoxide Ion Addition to Substituted
(Methoxyphenylcarbene)pentacarbonylchromium(0) and
(Methoxyphenylcarbene)pentacarbonyltungsten(0) in Methanol and
Aqueous Acetonitrile
Claude F. Bernasconi* and Luis Garc´ıa-R´ıo
Contribution from the Department of Chemistry and Biochemistry, UniVersity of California,
Santa Cruz, California 95064
ReceiVed NoVember 29, 1999
Abstract: Rate and equilibrium constants for the nucleophilic addition of MeO- in methanol and in 90%
MeCN-10% MeOH, of HCtCCH2O- and CF3CH2O- in 50% MeCN-50% water, and of OH- in 50%
MeCN-50% water (rate constants only) to Fischer carbene complexes of the type (CO)5MdC(OMe)C6H4-
4-Z (M ) Cr and W) are reported. The reactions lead to the formation of tetrahedral adducts of the type
(CO)5Mh -C(OMe)(OR)C6H4-4-Z. The kinetics of general acid catalyzed loss of MeO- from the MeO- adducts
by phenol, 4-bromophenol, and 3,5-dichlorophenol, to regenerate the carbene complex, were also investigated.
Hammett F and Fn values as well as approximate ânuc (ânnuc) values were determined. They suggest that the
transition state for alkoxide ion addition to the carbene complexes is imbalanced in the sense that charge
delocalization into the CO ligands lags somewhat behind C-O bond formation. This finding is consistent
with the fact that the intrinsic rate constants for these reactions are substantially lower than those for alkoxide
ion addition to carboxylic esters. Another type of imbalance is seen for the transition state of the general acid
catalyzed MeO- departure from the MeO- adducts. Here proton transfer to the leaving group is ahead of
C-O bond cleavage.
Introduction
The substitution of an alkoxy group in Fischer type carbene
complexes such as 1-M2 by nucleophiles3-8 has generally been
assumed to occur by the stepwise mechanism of eq 1. This
mechanism is analogous to the one established for the reactions
of carboxylic esters with nucleophiles. However, there is a major
(1) For part 18, see: Bernasconi, C. F.; Ali, M.; Lu, F. J. Am. Chem.
Soc. 2000, 122, 1352.
(2) When using the symbols 1-M, 2-M, etc., both the Cr and W derivative
will be meant. If only one derivative is referred to, the symbols 1-Cr, 1-W,
2-Cr-, 2-W-, etc. will be used.
(3) Do¨tz, K. H.; Fischer, H.; Hofmann, P.; Kreissl, F. R.; Schubert, U.;
Weiss, K. Transition Metal Complexes; Verlag Chemie: Deerfield Beach,
FL, 1983.
(4) Thiolate ions: (a) Fischer, E. O.; Leupold, M.; Kreiter, C. G.; Mu¨ller,
J. Chem. Ber. 1972, 105, 150. (b) Lam, C. T.; Senoff, C. V.; Ward, J. E.
H. J. Organomet. Chem. 1974, 70, 273. (c) Aumann, R.; Schro¨der, J. Chem.
Ber. 1990, 123, 2053. (d) Bernasconi, C. F.; Flores, F. X.; Kittredge, K.
W. J. Am. Chem. Soc. 1998, 120, 7983. (e) Bernasconi, C. F.; Kittredge,
K. W.; Flores, F. X. J. Am. Chem. Soc. 1999, 121, 6630.
difference between the two classes of electrophiles in that 2-Cr-
or 2-W- are much more stable than the respective tetrahedral
intermediates in the ester reactions. As a result, 2-Cr- or 2-W-
are directly detectable in the reactions of 1-Cr and 1-W with
thiolate ions in 50% aqueous acetonitrile.4e In another investiga-
tion, the symmetrical dimethoxy adducts 4-M- were generated
in measurable concentrations in the reactions of 1-Cr and 1-W
with MeO- in methanol.9
(5) OH- and/or water: Aumann, R.; Hinterding, P.; Kru¨ger, C.; Goddard,
R. J. Organomet. Chem. 1993, 459, 145. (b) Bernasconi, C. F.; Flores, F.
X.; Kittredge, K. W. J. Am. Chem. Soc. 1997, 119, 2103.
(6) Amines: (a) Klabunde, U.; Fischer, E. O. J. Am. Chem. Soc. 1967,
89, 7141. (b) Connor, J. A.; Fischer, E. O. J. Chem. Soc. A 1969, 578. (c)
Fischer, E. O.; Kollmeier, H.-J. Chem. Ber. 1971, 104, 1339. (d) Fischer,
E. O.; Leupold, M. Chem. Ber. 1972, 102, 599. (e) Fischer, E. O.; Heckl,
B.; Werner, H. J. Organomet. Chem. 1971, 28, 359. (f) Werner, H.; Fischer,
E. O.; Heckl, B.; Kreiter, C. G. J. Organomet. Chem. 1971, 28, 367. (g)
Bernasconi, C. F.; Stronach, M. W. J. Am. Chem. Soc. 1993, 115, 1341.
(7) Carbanions: (a) Fischer, E. O.; Riedmu¨ller, S. Chem. Ber. 1976, 109,
3358. (b) Fischer, E. O.; Held, W.; Kreissl, F. R. Chem. Ber. 1977, 110,
3842. (c) Burkhardt, T. J.; Casey, C. P. J. Am. Chem. Soc. 1973, 95, 5833.
(d) Fischer, E. O.; Held, F. R.; Kreissl, F. R.; Frank, A.; Hattner, G. Chem.
Ber. 1977, 110, 656. (e) Casey, C. P.; Burkhardt, T. J.; Bunnell, C. A.;
Calabrese, J. C. J. Am. Chem. Soc. 1977, 99, 2127.
Besides providing the most compelling evidence for the
stepwise mechanism, these studies allowed the determination
of the rate constants of the various steps (k1, k-1, k2) and their
(8) Other nucleophiles: (a) Fischer, E. O.; Kreis, G.; Kreissl, F. R.;
Kreiter, C. G.; Mu¨ller, J. J. Chem. Ber. 1973, 106, 3910. (b) Casey, C. P.;
Brunsvold, W. P. Inorg. Chem. 1977, 16, 391. (c) Bell, R. A.; Chisholm,
M. H.; Couch, D. A.; Rankel, L. A. Inorg. Chem. 1977, 16, 677.
(9) Bernasconi, C. F.; Flores, F. X.; Gandler, J. R.; Leyes, A. E.
Organometallics 1994, 13, 2186.
10.1021/ja994174w CCC: $19.00 © 2000 American Chemical Society
Published on Web 04/07/2000

3822 J. Am. Chem. Soc., Vol. 122, No. 16, 2000
Bernasconi and Garc´ıa-R´ıo
dependence on a variety of structural parameters. Two important
conclusions that have emerged from this work can be sum-
marized as follows.
(1) The equilibrium constants, K1 ) k1/k-1, for nucleophilic
addition to 1-M are many orders of magnitude higher than those
for addition to esters or carboxylic acids. For example, K1 for
MeO- addition to 1-Cr (70.1 M-1)9 is 2 × 108 to 109 fold higher
than that for MeO- addition to methyl benzoate10 and K1 for
ture-reactivity coefficients such as Brønsted R and â values.15
So far, no such evidence has been presented for reactions of
Fischer carbene complexes. A major objective of the present
work is to look for such evidence by studying the reaction of
phenyl-substituted Fischer carbene complexes, 6-M, with alkox-
ide ions.
n-PrS- addition to 1-Cr (1.06 × 104 M-1 4e
)
is ≈1017 fold higher
than that for EtS- addition to acetic acid (≈7.9 × 10-14 M-1).11
These comparisons show that the (CO)5M moieties are much
more electron withdrawing than the carbonyl oxygen of an ester
or carboxylic acid. This is, in large measure, due to the
π-acceptor properties of the (CO)5M moieties which, in the
adducts, lead to delocalization of the negative charge into the
CO ligands.12
(2) As one would expect, the rate constants, k1, for nucleo-
philic attack on 1-Cr and 1-W are also higher than those for
nucleophilic attack on the corresponding acyl compounds.
However, the extent to which the k1 values for 1-Cr and 1-W
exceed those for the acyl compounds is less dramatic than what
one might expect on the basis of the high ratios of the K1 values.
This indicates that the intrinsic rate constants14 for nucleophilic
addition to the carbene complexes are substantially lower than
those for addition to the acyl compounds. The lowering of the
intrinsic rate constant is a manifestation of the well-documented
inverse relationship between intrinsic rate constants of anion-
forming reactions and the degree of resonance stabilization of
the anion.15 The underlying reason for this inverse relationship
is that the development of the resonance effect lags behind bond
formation or charge transfer at the transition state (“transition
state imbalance”). As a result, the stabilization of the transition
state by the resonance effect is disproportionately small
compared to the stabilization of the anion and hence the intrinsic
rate constant is depressed.16
For the addition of a nucleophile to a Fischer carbene
complex, the situation may be described by eq 3, which shows
the charge at the transition state mostly localized on the metal
while in the adduct the charge is mostly delocalized into the
CO ligands, a point highlighted by the brace symbol on 2-M-.
In reactions leading to carbanions stabilized by more con-
ventional π-acceptors such as nitro groups, carbonyl groups,
and others, there is independent evidence for the imbalanced
nature of the transition state. This evidence is based on struc-
Results
Reactions in Methanol. A. Chromium Carbene Com-
plexes. The reaction of the substituted chromium carbene
complexes 6-Cr-Z (Z ) 4-OMe, 4-Me, H, 4-F, 4-Cl, 3-Cl, and
4-CF3) with NaOMe in methanol follows the general pattern
described for the reaction with the parent compound (6-Cr-H
≡ 1-Cr9). There is a rapid process that leads to the reversible
formation of the corresponding adducts (7-Cr-Z-) followed by
a slower reaction that leads to unidentified decomposition
products. Reversibility was demonstrated by recovering the
carbene complex upon acidification of the adduct immediately
after its formation; when acidification was delayed there was
only partial or no recovery as judged from the absorption spectra.
Two types of kinetic experiments were performed. In the first,
reaction was initiated by mixing a methanolic solution of the
carbene complex with a methanolic solution of NaOMe in a
stopped-flow spectrophotometer. NaOMe was always in large
excess over the substrate, ensuring pseudo-first-order conditions.
The pseudo-first-order rate constants showed a linear depen-
dence on [MeO-] according to eq 5. Representative plots of
kobsd ) k1MeO[MeO-] + kMeO
(5)
-1
kobsd vs [MeO-] are shown in Figure 1. The k1MeO and kMeO
values are reported in Table 1; for the most reactive carbene
complexes (Z ) 4-Cl and 4-CF3), the intercepts were too small
to yield a reliable k
-1
(10) K1 for MeO- addition to methylbenzoate has been estimated to be
MeO
value.
between >1.1 × 10-7 and 5.0 × 10-7 M-1 9
.
-1
In the second type of experiment, equilibrium was approached
from the adduct side by reacting 7-Cr-Z- with phenol,
4-bromophenol, and 3,5-dichlorophenol buffers. Because the
adducts are quite unstable, the reaction was initiated within
0.5-2 s after the adduct had been generated with 0.1 M NaOMe
in a sequential stopped-flow apparatus. In these experiments,
(11) Guthrie, J. P. J. Am. Chem. Soc. 1978, 100, 5892.
(12) The strong electron-withdrawing effect of M(CO)n moieties has also
been expressed in terms of a high electronegativity.13
(13) Labinger, J. A.; Bercaw, J. E. Organometallics 1988, 7, 926.
(14) For a reaction with the forward rate constants k1 and the reverse
rate constant k-1, the intrinsic rate constant is defined as ko ) k1 ) k-1
when the equilibrium constant, K1 ) 1 (∆G° ) 0).
(15) (a) Bernasconi, C. F. Acc. Chem. Res. 1987, 20, 301. (b) Bernasconi,
C. F. Acc. Chem. Res. 1992, 25, 9. (c) Bernasconi, C. F. AdV. Phys. Org.
Chem. 1992, 27, 119.
k
obsd is given by eq 6, with B- and BH representing the buffer
base and buffer acid, respectively.17 Plots of kobsd vs [BH] at
(16) This relationship between intrinsic rate constants and transition state
imbalances is the manifestation of a more general principle which is not
restricted to resonance effects and has been called the principle of nonperfect
synchronization (PNS).15
(17) In principle, eq 6 should include terms for spontaneous MeOH
H
addition (kM1 eOH) and its reverse, H+-catalyzed MeO- departure (k-1aH ),
+
but under the reaction conditions these terms are negligible.

Kinetics of ReVersible Alkoxide Ion Addition
J. Am. Chem. Soc., Vol. 122, No. 16, 2000 3823
Table 1. Rate and Equilibrium Constants for Reversible MeO- Addition to 6-Cr-Z and 6-W-Z in Methanol at 25 °C
MeO
-1
MeO
-1
MeO
-1
Z
kM1 eO,a M-1 s-1
k
,a s-1
k
,b s-1
k
,c s-1
KM1 eO ) kM1 eO/kMeO, M-1
kM1 eO-(90% MeCN), M-1 s-1
-1
6-Cr-Z
4-Me2N
4-MeO
4-Me
H
4-F
4-Cl
45.7 ( 0.6
14.4 ( 1.3
31.2 ( 1.3
77.1 ( 3.4
188 ( 5
3.80 ( 0.20
3.03 ( 0.06
0.98 ( 0.28
0.74 ( 0.02
0.27 ( 0.02
3.80 ( 0.21
2.07 ( 0.06
0.67 ( 0.08
0.61 ( 0.10
0.36 ( 0.03
3.80 ( 0.20
2.55 ( 0.48
0.82 ( 0.16
0.68 ( 0.07
0.31 ( 0.04
3.68 ( 0.45
12.2 ( 2.6
(1.06 ( 0.02) × 103
(2.65 ( 0.03) × 103
(4.04 ( 0.05) × 103
(7.47 ( 0.25) × 103
(1.44 ( 0.08) × 104
(2.51 ( 0.03) × 104
(3.36 ( 0.12) × 104
94 ( 22
276 ( 36
330 ( 6
(1.06 ( 0.15) × 103
3-Cl
4-CF3
554 ( 4
817 ( 14
0.090 ( 0.012d
(9.07 ( 1.4) × 103 e
6-W-Z
4-MeO
4-Me
H
4-F
4-Cl
46.3 ( 0.6
105 ( 1
4.42 ( 0.65
3.95 ( 0.14
1.60 ( 0.21
1.80 ( 0.08
0.72 ( 0.09
10.5 ( 1.7
26.6 ( 4.7
(2.96 ( 0.06) × 103
(4.86 ( 0.06) × 103
(9.03 ( 0.09) × 103
(1.63 ( 0.03) × 104
(3.32 ( 0.06) × 104
186 ( 3
116 ( 17
683 ( 40
806 ( 15
379 ( 41
(1.12 ( 0.19) × 103
a From experiments with NaOMe (eq 5). b From experiments in phenol or 4-bromophenol buffers (eq 7), see text. c Adopted value for the reactions
of 6-Cr-Z, see text. k-1 calculated as k1MeO/K1MeO, with K1MeO obtained from eq 7, see text. e K1 obtained from eq 7, see text.
MeO
MeO
d
Figure 1. Representative plots of kobsd vs [MeO-] (eq 5) for the
reactions of MeO- with 6-Cr-F (9), 6-W-Me (O), 6-W-MeO (b),
and 6-Cr-MeO (0).
Figure 2. Representative plots of kobsd vs buffer acid concentration
for the reactions of phenol buffers with 6-Cr-MeO (O), 6-Cr-Me (9),
and 6-Cr-H (b) and of 4-bromophenol buffers with 6-Cr-Me (0).
kobsd ) k1MeO[MeO-] + k-M1eO + k1B[B-] + kBH[BH] (6)
BH
-1
MeO
[BH] yield k and k
directly from slope and intercept,
-1
-1
MeO
respectively. The k
values (3.80 ( 0.21 s-1 for the phenol
-1
constant buffer ratio provide slopes given by eq 7, with KaBH
and 3.74 ( 0.36 s-1 for the 4-bromophenol buffers) are in
MeO
-1
perfect agreement with k
the experiments in NaOMe solutions (Table 1). The intercept
from the plot of kobsd vs [3,5-dichlorophenol] are too small to
) 3.80 ( 0.20 s-1 obtained from
kB1 KBa H
KM1 eO
aH
K
slope )
+ kB-H1 ) kBH
S + 1
(7)
-1
(
)
aH
+
+
MeO
give a reliable k
value. The slopes of the buffer plots yield
-1
BH
being the acidity constant of the buffer, KS the ionic product of
k
from which kB is calculated as k-1KMeOKS/KBH
.
-1
1 1 a
the solvent,18 and K1MeO ) kM1 eO/kMeO the equilibrium constant
Z ) 4-Me. The k1MeO[MeO-] and kB1 [B-] terms are again
-1
for MeO- addition. The intercepts of these plots are given by
kM1 eO[MeO-] + k-M1eO. Some representative plots are shown in
Figure 2.
MeO
-1
negligible. The k
value (2.07 ( 0.06 s-1) obtained in
phenol buffers is in reasonably good agreement with that
obtained in NaOMe solutions (3.03 ( 0.6 s-1); the average value
(2.55 s-1) will be adopted. No reliable k-M1eO values could be
obtained in the other buffers because the intercepts of the buffer
The amount of information obtained from these plots varied
depending on Z because of the large variation in the K1MeO
values (Table 1). The salient features can be summarized as
follows.
plots were too small. The kBH and kB1 values were obtained as
-1
for Z ) MeO.
Z ) 4-MeO. Due to the small KM1 eO values, the k1MeO[MeO-]
and kB1 [B-] terms in eq 6 are negligible and the plots of kobsd vs
Z ) H. The kM1 eO[MeO-] and kB1 [B-] terms are negligible
in the 4-bromophenol and 3,5-dichlorophenol buffers but
not in the phenol buffers. From the pH dependence of the inter-
(18) pKS ) 16.92, Rochester, C. H.; Rossall, B. Trans. Faraday Soc.
1969, 65, 1004.
cepts of the plots of kobsd vs [BH] in the phenol buffers, kMeO
)
1

3824 J. Am. Chem. Soc., Vol. 122, No. 16, 2000
Bernasconi and Garc´ıa-R´ıo
Table 2. Rate and Equilibrium Constants for Buffer-Catalyzed Addition of Methanol to 6-Cr-Z at 25 °C
BH
-1
pKBa H
kB1 , M-1 s-1
k
, M-1 s-1
K1B
a
Z
BH
PhOH
4-BrC6H4OH
3,5-Cl2C6H3OH
PhOH
4-BrC6H4OH
3,5-Cl2C6H3OH
PhOH
4-BrC6H4OH
3,5-Cl2C6H3OH
PhOH
4-BrC6H4OH
3,5-Cl2C6H3OH
PhOH
4-BrC6H4OH
3,5-Cl2C6H3OH
4-BrC6H4OH
3,5-Cl2C6H3OH
4-MeO
14.46
13.5
12.1
14.46
13.5
12.1
14.46
13.5
12.1
14.46
13.5
12.1
14.46
13.5
12.1
13.5
12.1
(1.94 ( 0.30) × 10-1
(1.19 ( 0.20) × 10-1
(7.56 ( 0.85) × 10-2
(5.74 ( 1.32) × 10-1
(3.69 ( 0.88) × 10-1
(2.35 ( 0.61) × 10-1
2.70 ( 0.82
15.3 ( 0.6
(1.27 ( 0.15) × 10-2
(1.40 ( 0.17) × 10-3
(5.56 ( 0.68) × 10-5
(4.22 ( 0.90) × 10-2
(4.64 ( 0.99) × 10-3
(1.84 ( 0.39) × 10-4
(3.25 ( 0.76) × 10-1
(3.57 ( 0.82) × 10-2
(1.42 ( 0.33) × 10-3
1.03 ( 0.12
84.9 ( 4.4
(1.36 ( 0.10) × 103
13.6 ( 0.2
4-Me
H
79.6 ( 2.5
(1.28 ( 0.07) × 103
8.32 ( 0.61
1.51 ( 0.41
42.3 ( 1.7
1.04 ( 0.27
(7.34 ( 0.22) × 102
4.27 ( 1.14
4-F
4.40 ( 1.67
3.53 ( 0.56
31.2 ( 0.6
(1.13 ( 0.13) × 10-1
(4.48 ( 0.53) × 10-3
3.67 ( 0.52
2.32 ( 0.38
(5.18 ( 0.24) × 102
1.61 ( 0.50
4-Cl
4-CF3
5.91 ( 2.65
5.88 ( 1.11
14.6 ( 0.7
(4.03 ( 0.57) × 10-1
(1.60 ( 0.22) × 10-2
3.45 ( 0.53
3.50 ( 0.66
(2.19 ( 0.12) × 102
9.66 ( 2.90
2.80 ( 0.41
7.63 ( 1.26
55.7 ( 1.1
(1.37 ( 0.20) × 10-1
a KB1 ) KM1 eOKS/KBH with KS ) 1.20 × 10-17 (pKS ) 16.92). b kB1 ) kBHK1B.
a
-1
84.2 ( 12.6 M-1 s-1 and k-M1eO ) 0.67 ( 0.08 s-1 are obtained,
in fair to good agreement with kM1 eO ) 77.1 ( 3.4 M-1 s-1 and
kM-1eO ) 0.98 ( 0.28 s-1 determined in NaOMe solutions.9 The
average values with will be adopted. The intercepts of the buffer
plots with 4-bromophenol and 3,5-dichlorophenol were too small
MeO
-1
BH
to yield k
values. The k value for phenol was obtained
-1
according to eq 7 while kB1 was calculated in the usual way.
For kB-H1 and kB1 for 4-bromophenol and 3,5-dichlorophenol, the
procedures described for Z ) MeO were used.
Z ) 4-F. The situation is quite similar to that for Z ) H. In
phenol buffers kM1 eO ) 138 ( 18 M-1 s-1 and k-M1eO ) 0.61 (
0.10 s-1 are obtained, in reasonably good agreement with
the values determined in NaOMe solutions (kM1 eO ) 199 ( 3
M-1 s-1 and k-M1eO ) 0.74 ( 0.02 s-1). The latter rate constants
BH
-1
appear to be more accurate and will be adopted. The k
and kB1 values were obtained as described for Z ) H.
Z ) 4-Cl. The kM1 eO[MeO-] and kB1 [B-] terms are important
in phenol as well as in 4-bromophenol buffers. For the phenol
buffers, one obtains kM1 eO ) 239 ( 24 M-1 s-1 and kMeO
)
-1
0.20 ( 0.13 s-1, and for the 4-bromphenol buffers, kM1 eO ) 419
( 32 M-1 s-1 and kM-1eO ) 0.36 ( 0.03 s-1, respectively. This
Figure 3. Plot of slope vs a-H1 according to eq 7 for the reaction of
compares with kM1 eO ) 330 ( 6 M-1 s-1 and kMeO ) 0.27 (
+
-1
6-Cr-CF3 with 4-bromophenol buffers.
0.02 s-1 obtained in NaOMe solutions; for k-M1eO the average of
Reactions with MeO- in 90% MeCN-10% MeOH. Rate
0.27 and 0.36 will be adopted. The kBH and k1B values for phe-
-1
constants for MeO- addition to 6-Cr-Z and 6-W-Z (kM1 eO
)
nol and 4-bromophenol were evaluated as described for Z ) F
with the phenol buffers; kBH and kB1 for 3,5-dichlorophenol
were determined in 90% MeCN-10% MeOH with NaOMe as
the base. In this solvent the equilibrium for nucleophilic addition
is more favorable than in pure methanol. This allowed inclusion
of the rather unreactive 6-Cr-NMe2 in our series. The plots
of kobsd vs [MeO-] were all linear with intercepts ≈0 or too
small to yield a reliable k-M1eO value. The results are included in
Table 1.
-1
were obtained as described for Z ) MeO.
Z ) CF3. In phenol buffers, the kMeO[MeO-] terms is much
1
larger than kM-1eO and only kM1 eO ) (1.01 ( 0.09) × 103 M-1 s-1
could be obtained which is in good agreement with k1MeO
)
820 ( 14 M-1 s-1 determined in NaOMe solutions. In 4-bromo-
phenol buffers, k1MeO ) 810 ( 110 M-1 s-1 is obtained but no
Reactions with OH-, HCtCCH2O-, and CF3CH2O- in
50% MeCN-50% Water (v/v). A. OH-. Rates were measured
in the range of 0.01 to 0.1 M KOH. The slopes of the linear
plots of kobsd vs [KOH] yielded the kO1 H values reported in
Table 3. Because 6-Cr-CF3 is quite unstable in 50% MeCN-
50% water, the experiments with this compound had to be
performed using stock solutions in pure MeCN which may have
introduced some experimental errors as described in the
Experimental Section.
reliable value for kM-1eO could be determined. A different source
MeO
-1
of k
comes from treating KM1 eO in eq 7 as an unknown.
From the plot of slope vs a-H1 (4-bromophenol) shown in
+
Figure 3, one obtains KM1 eO ) (9.07 ( 0.14) × 103 M-1 from
MeO
-1
which k
) 0.090 ( 0.012 s-1 is calculated. All kB-H1 and kB1
values are summarized in Table 2.
B. Tungsten Carbene Complexes. Rate constants (k1MeO
MeO
and k ) for the reaction of MeO- with the tungsten carbene
-1
B. HCtCCH2O- and CF3CH2O-. Alkoxide ion solutions
were made by adding the respective alcohol to a KOH solution.
[RO-] and [OH-] were calculated on the basis of the pKRa OH of
complexes 6-W-Z (Z ) MeO, Me, H, 4-F, and 4-Cl) were
determined as described for the Cr analogues. The results are
included in Table 1.

Kinetics of ReVersible Alkoxide Ion Addition
J. Am. Chem. Soc., Vol. 122, No. 16, 2000 3825
Table 3. Rate and Equilibrium Constants for the Reactions of 6-Cr-Z and 6-W-Z with OH-, HCtCCH2O-, and CF3CH2O- in 50%
MeCN-50% Water (v/v) at 25 °C
OPGYL
-1
OTFE
-1
Z
kO1 H, M-1 s-1
kO1 PGYL
,
a M-1 s-1
k
,a s-1
KO1 PGYL,a M-1
kO1 TFE,b M-1 s-1
k
,b s-1
KO1 TFE,b M-1
6-Cr-Z
4-Me2N
4-MeO
4-Me
H
4-F
4-Cl
0.131 ( 0.004
3.35 ( 0.18
13.2 ( 0.2
26.6 ( 0.3c
37.1 ( 0.9
76.8 ( 1.9
80.1 ( 1.3
105 ( 1
25.3 ( 0.6
47 ( 1
0.496 ( 0.003
0.281 ( 0.001
0.250 ( 0.002
0.150 ( 0.008
51.0 ( 1.5
167 (4
87 ( 1
348 ( 7
807 ( 69
121 ( 4
162 ( 4
240 ( 3
43.6 ( 0.2
61 ( 2
0.493 ( 0.003
0.199 ( 0.005
88.4 ( 1.3
307 ( 18
3-Cl
4-CF3
0.087 ( 0.003
2759 ( 136
6-W-Z
4-MeO
4-Me
H
4-F
4-Cl
6.85 ( 0.20
17.2 ( 0.9
26.3 ( 0.2c
48.0 ( 1.0
55.7 ( 1.4
50.0 ( 3.9
84.0 ( 1.8
158 ( 3
0.988 ( 0.006
0.504 ( 0.007
0.465 ( 0.003
0.272 ( 0.006
50.6 ( 4.2
167 ( 6
340 ( 9
853 ( 41
232 ( 6
62.9 ( 1.9
1.08 ( 0.01
58.2 ( 2.3
a PGYL ) HCtCCH2; pKaPGYLOH ) 14.73. b TFE ) CF3CH2; pKTa FEOH ) 13.68. c Reference 4e.
erating the adducts in pure acetonitrile in the presence of 5 ×
10-4 M NaOR and then adding these adducts to a triethylamine
buffer at pH 10.5. At this pH the concentration of the alkoxide
ion is quite small. Hence the equilibrium for adduct formation
becomes very unfavorable and the adduct decomposes by
RO
competing loss of RO- to revert back to reactants (k ) and
-1
loss of MeO- to form the RO-substituted carbene complex
8-M-Z (kM2 eO, special case of k2 in eq 1). Thus, kobsd is given by
eq 9. In principle, determination of the ratio of recovered
substrate to the substitution product by HPLC could provide
the kRO/k2MeO ratio19 which, in conjunction with eq 9, would
-1
allow dissection of kRO and kM2 eO. However, both 6-M-Z and
-1
Figure 4. Representative plots of kobsd - kOH[OH-] vs [RO-] accord-
kobsd ) k-RO1 + k2MeO
(9)
1
ing to eq 8 for the reactions of 6-W-F with HCtCCH2O- (b), of
6-W-H with HCtCCH2O- (O), and of 6-Cr-Cl with CF2CH2O- (9).
8-M-Z are too unstable under the reaction conditions to allow
such analysis. But HCtCCH2O- and CF3CH2O- are better
the alcohol, the total concentration of ROH, and the measured
pH. In the experiments that were initiated by mixing the carbene
complex with the RO-/OH- solutions, kobsd can be approximated
by eq 8; it is an approximation because only the OH- reaction
leaving groups than MeO- and the push provided by the
RO
-1
methoxy group left behind for the k process is stronger than
the push by the alkoxy groups left behind in the kM2 eO process.
Both factors contribute to render kR-O1 >(.) k2MeO, especially for
kobsd ) k1OH[OH-] + kR1 O[RO-]
(8)
RO- ) CF3CH2O-.21 It is therefore reasonable to approximate
RO
-1
kobsd by k in eq 9 (more on this in the Discussion section).
is irreversible (see Discussion) while the RO- reaction leads to
reversible adduct formation. kR1 O (Table 3) was obtained as the
Discussion
slope of plots of {kobsd - kOH[OH-]} vs [RO-]. Representative
Mechanism. The reactions of 6-M-Z with MeO-, HCtCC-
H2O-, and CF3CH2O- all lead to nucleophilic addition to form
the corresponding 7-M-Z- type adducts. Upon standing for
extended periods of time (several minutes or longer), the adducts
decompose into unidentified byproducts as discussed in more
1
examples are shown in Figure 4. For the reactions with
CF3CH2O-, the equilibrium for adduct formation was only
favorable enough to allow measurements with the most strongly
activated carbene complexes (6-Cr-Cl, 6-Cr-CF3, 6-W-Cl),
while with HCtCCH2O- these measurements were feasible
with all but the least activated carbenes (6-M-OMe). For the
reactions with 6-Cr-CF3, the experiments again had to be
performed using stock solutions in pure MeCN.
(19) This type of analysis was successfully applied to the competing
loss of MeO- and CF3CH2O- from PhC(OMe)(OCH2CF3)C(Ph)dNO-.20
(20) Bernasconi, C. F.; Schuck, D. F.; Ketner, R. J.; Eventova,2 I.;
Rappoport, Z. J. Am. Chem. Soc. 1995, 117, 2719.
Rate constants for the loss of RO- from the respective
(21) For an example where CF3CH2O- and HCtCCH2O- are shown to
be much better leaving groups than MeO-, see ref 22.
RO
-1
alkoxide ion adducts, k (Table 3), were obtained after gen-

3826 J. Am. Chem. Soc., Vol. 122, No. 16, 2000
Bernasconi and Garc´ıa-R´ıo
Table 4. Hammett F and Normalized F (Fn) Valuesa for Alkoxide and Hydroxide Ion Addition to 6-M-Z
OR
-1
F(kO1 R
)
F(k
)
F(KO1 R
)
Fn(kOR
)
Fn(kOR
)
-1
b
b
reaction
6-Cr-Z + MeO- in MeOH
1
2.20 ( 0.23
(2.67 ( 0.10)
2.15 ( 0.14
(2.36 ( 0.13)
1.34 ( 0.16
(1.39 ( 0.14)
2.20 ( 0.20
(2.23 ( 0.29)
2.58 ( 0.48
(2.37 ( 0.22)
2.11 ( 0.14
(2.07 ( 0.13)
1.73 ( 0.31
(1.68 ( 0.17)
1.78 ( 0.34
(1.68 ( 0.32)
-2.04 ( 0.10
(-2.24 ( 0.15)
4.22 ( 0.35
(4.92 ( 0.12)
0.52 ( 0.02
(0.54 ( 0.02)
-0.48 ( 0.02
(-0.46 ( 0.02)
6-Cr-Z + MeO- in 90% MeCN-10% MeOH
6-Cr-Z + HCtCCH2O- in 50% MeCN-50% H2O
6-Cr-Z + HO- in 50% MeCN-50% H2O
6-W-Z + MeO- in MeOH
-0.90 ( 0.20
(-1.06 ( 0.11)
2.25 ( 0.29
(2.45 ( 0.24)
0.59 ( 0.05
(0.57 ( 0.01)
-0.41 ( 0.05
(-0.43 ( 0.01)
-1.60 ( 0.20
(-1.67 ( 0.17)
4.18 ( 0.31
(4.04 ( 0.08)
0.63 ( 0.07
(0.59 ( 0.05)
-0.37 ( 0.07
(-0.41 ( 0.05)
6-W-Z + MeO- in 90% MeCN-10% MeOH
6-W-Z + HCtCCH2O- in 50% MeCN-50% H2O
6-W-Z + HO- in 50% MeCN-50% H2O
-1.38 ( 0.11
(-1.39 ( 0.16)
3.11 ( 0.26
(3.07 ( 0.01)
0.56 ( 0.06
(0.55 ( 0.05)
-0.44 ( 0.06
(-0.45 ( 0.05)
a First number, F based on all substrates; second number (in parentheses), F based on excluding 4-F and 4-CF3, see text. b Normalized values
obtained from plots of log kO1 R vs log K1OR and log kOR vs log KO1 R, respectively.
-1
Figure 5. Representative Hammett plots for the reaction of 6-Cr-Z
Figure 6. Representative Brønsted plots for the reactions of 6-Cr-Z
with MeO- in MeOH (O, b) and of 6-Cr-Z with HCtCCH2O- in
50% MeCN-50% water (0, 9).
with MeO- in methanol (O, kM1 eO; b, kMeO; 0, K1MeO) and in 90%
-1
MeCN-10% MeOH (2, kM1 eO).
detail elsewhere.9 However, on a much shorter time scale, the
adducts revert back to reactants upon removal of the alkoxide
ion from equilibrium by rapid acidification of the solution. This
is the case for the symmetrical MeO- adducts and at least to a
good approximation for the HCtCCH2O- and CF3CH2O-
adducts as well; as mentioned in the Results section, the
competing loss of MeO- from these latter adducts is expected
to be substantially slower than the loss of HCtCCH2O- and
CF3CH2O-, respectively.
log K1). Representative Hammett and Brønsted plots are shown
in Figures 5 and 6, respectively. On most plots, there is a
tendency for the 4-F derivatives to deviate positively and for
the 4-CF3 derivative to deviate negatively from the lines defined
by the other substituents. Whether this represents a systematic
error or indicates true curvature is difficult to decide; ex-
perimental problems in obtaining highly accurate results with
6-Cr-CF3 (see Results section) suggest that, at least for this
compound, there may be some systematic error involved. For
the 4-F derivatives, the positive deviations are more pronounced
in the Hammett plots than in the Brønsted plots, especially in
methanol, suggesting that the standard σ value for the 4-F
substituent may not be appropriate in methanol.
Table 4 summarizes all F (Hammett plots) and normalized F
values (Fn, Brønsted plots) for the reactions with OH- and
alkoxide ions, while Table 5 gives a summary of the normalized
F values for the reversible buffer-catalyzed MeOH addition to
6-Cr-Z in methanol.23 Because of the problems with the 4-F
and 4-CF3 derivatives mentioned above, the F and Fn values
were calculated with and without including these derivatives;
inclusion or exclusion of these points has only a small influence
The reactions of 6-M-Z with OH- presumably also lead to
adducts of the type 7-M-Z- with R ) H. However, loss of
MeO- to form 8-M-Z (R ) H) with subsequent further
decomposition is much faster than loss of OH- to regenerate
6-M-Z5b and hence no adducts can be detected.
Structure-Reactivity Relationships. The major objective
of this study was to determine substituent effects on the reactions
of 6-Cr-Z and 6-W-Z with MeO-, HCtCCH2O-, CF3CH2O-,
and OH- in various solvents. The results were treated in terms
of Hammett correlations and, for the reactions that allowed
determination of both rate and equilibrium constants, also in
terms of Brønsted correlations (log k1 vs log K1 and log k-1 vs

Kinetics of ReVersible Alkoxide Ion Addition
J. Am. Chem. Soc., Vol. 122, No. 16, 2000 3827
Table 5. Normalized F Values (Fn) for Buffer-Catalyzed Methanol
Addition to 6-Cr-Za
errors or a real difference between the two kinds of complexes
cannot be decided. Because eq 8 is only an approximation, the
kR1 O values, and with them the KR1 O values, may indeed be
subject to significant systematic errors that are not reflected in
their standard deviations.
B-
Fn(kB)
Fn(kBH
)
1
-1
PhO-
0.62 ( 0.07
(0.60 ( 0.09)
0.58 ( 0.07
(0.46 ( 0.09)
0.59 ( 0.06
(0.67 ( 0.06)
-0.38 ( 0.07
(-0.40 ( 0.09)
-0.42 ( 0.07
(-0.54 ( 0.09)
-0.41 ( 0.06
(-0.33 ( 0.06)
4-BrC6H4O-
A major factor that is likely to contribute to the smaller
F(K1OR) and F(k1OR) values for the HCtCCH2O- reactions is
the higher polarity of 50% MeCN-50% water (ꢀ ) 58.8)25
compared to that of methanol (ꢀ ) 32.6). The change in
nucleophile may constitute an additional factor. Since the HCt
CCH2O group in 7-M-Z- is more electron withdrawing than
the MeO group, it contributes more to the stabilization of the
negative charge in 7-M-Z- and the transition state leading to it
than the methoxy group. This, just as is the case with a more
polar solvent, reduces the requirement for stabilization of the
charge by the Z-substituent.
3,5-Cl2C6H3O-
a First number, Fn based on all substituents; second number (in
parentheses), Fn based on excluding 4-F and 4-CF3, see text.
on the actual F and Fn values except for the buffer-catalyzed
reactions with 4-BrC6H4O- and 3,5-Cl2C6H3O- (Table 5), but
in most cases the standard deviations for the Fn values are
smaller than for the F values.
Hammett G Values. A. Reactions with MeO- and OH-.
On the basis of the Hammett plots including all substituents,
Gn and ânnuc. Is There a Transition State Imbalance?
The normalized F values for all reactions are all very close,
especially the ones obtained by excluding the 4-F and 4-CF3
substrates which are virtually indistinguishable, with an average
the F values for k1MeO, kMeO, and KM1 eO are, within the rather
-1
large experimental uncertainties, about the same for the
chromium and tungsten carbene complexes, i.e., in MeOH on
value of 0.56 for Fn(kOR) and -0.44 for Fn(k-RO1). Interestingly,
the order of 2.2-2.6 for kM1 eO, -1.6 to -1.0 for kMeO, 4.2 for
1
this is also very close to Fn(kRS) ) 0.54 for the reaction of
-1
KM1 eO, and 2.1 for k1MeO in 90% MeCN-10% MeOH. On the
other hand, based on the plots from which the 4-F and 4-CF3
derivatives are excluded, the standard deviations of most F
values are significantly smaller which allows some distinctions
1
HOCH2CH2S- with 9-W-Z26 even though the Hammett F values
between the two types of carbene complexes to be made. In
MeO
-1
particular, F(k ) for 6-Cr-Z (-2.24) appears to be signifi-
cantly more negative than that for 6-W-Z (-1.67) while
F(KM1 eO) is significantly larger for 6-Cr-Z (4.92) than for
6-W-Z (4.04); F(kM1 eO) is probably also larger for the chro-
mium carbene complexes, as suggested by the results in
methanol as well as in 90% MeCN-10% MeOH, although the
standard deviations are still too large for a definite conclusion.
The stronger substituent effect on the rate constant for oxyanion
addition to the chromium carbene complexes is also seen in
the larger F(kO1 H) values for 6-Cr-Z (2.23) compared to 6-W-Z
(1.68).
for this latter reaction are quite different. It appears that the
fraction of negative charge seen by the Z-substituent at the
transition state relative to the charge seen on the adduct is about
the same for all reactions, i.e., independent of the metal, the
leaving group or the nucleophile. What is less clear is whether
Fn(kRO) can be considered a measure of C-O bond formation
The smaller F(K1MeO) and F(kM1 eO) values for 6-W-Z com-
pared to 6-Cr-Z are consistent with the somewhat higher K1MeO
and kM1 eO values for the tungsten derivatives, especially those
with electron-donating substituents (Table 1). The enhanced
reactivity of the tungsten carbene complexes which has been
noted before4e probably reflects a somewhat stronger stabiliza-
tion of the negative charge by the (CO)5W moiety in 7-W-Z-
than by the (CO)5Cr moiety in 7-Cr-Z-. This reduces the
demand for further stabilization by the Z-substituent and lowers
the F values.
1
at the transition state. If the transitions state is imbalanced (see
eq 3), Fn(kRO) would not be such a measure. This is because
1
the closer proximity of the negative charge to the phenyl group
at the transition state compared to the situation in the adduct
tends to enhance Fn(kRO), i.e., bond formation will be less than
1
suggested by Fn(kRO).
1
The normalized ânuc values (ânnuc) summarized in Table 6
can shed some light on this question. Following traditional
views,27 ânnuc may be considered an approximate measure of
bond formation or charge transfer at the transition state.28
For the reactions of 6-Cr-Cl and 6-W-Cl with HCtCCH2O-
and CF3CH2O-, the ânnuc values are 0.46 (6-Cr-Cl) and 0.49
B. Reactions with HCtCCH2O- in 50% MeCN-50%
Water. The most notable feature of the reactions with
HCtCCH2O- is that the F(k1RO) and F(KR1 O) values are signi-
ficantly smaller than those for the reactions with MeO-,
especially for the chromium carbene complexes (Table 4). For
example, F(K1OR) ) 2.25 (2.45)24 for HCtCCH2O- vs 4.22
(4.92)24 for MeO-, or F(k1OR) ) 1.73 (1.68)24 for HCtCCH2O-
vs 2.20 (2.67)24 for MeO-. For the tungsten carbene com-
plexes, the differences in the F values for the two reactions
are smaller; whether this is an artifact caused by systematic
(6-W-Cl), respectively, i.e., somewhat lower than Fn(kRO) for
1
the reactions of HCtCCH2O- with 6-Cr-Z (0.59 (0.57)) and
6-W-Z (0.56 (0.55)). ânnuc might actually be still lower than the
values reported in Table 6, i.e., the latter should be considered
upper limits because the K1RO values (Table 3) were obtained
under the assumption that in eq 9 kM2 eO is negligible compared
RO
-1
to k
(see Results section). If kM2 eO were not negligible
(22) Bernasconi, C. F.; Ketner, R. J.; Chen, X.; Rappoport, Z. J. Am.
Chem. Soc. 1998, 120, 7461.
(25) Moreau, C.; Douhe´ret, G. J. Chem. Thermodyn. 1976, 8, 403.
(26) Bernasconi, C. F.; Ali, M. J. Am. Chem. Soc. 1999, 121, 11384.
(27) (a) Leffler, J. E.; Grunwald, E. Rates and Equilibria of Organic
Reactions; Wiley: New York, 1963; p 128. (b) Kresge, A. J. In Proton-
Transfer Reactions; Caldin, E. F., Gold, V., Eds.; Wiley: New York, 1975;
p 179. (c) Jencks, W. P. Chem. ReV. 1985, 85, 511.
(23) Fn values can also be obtained as F(kRO)/F(KR1 O) and F(kRO)/F(KR1 O
)
1
ratios, etc., but the direct determination from Brønsted plots r-ed1uces the
experimental uncertainty substantially.
(24) The numbers in parentheses are based on excluding the 4-F and
4-CF3 derivatives from the Hammett plots.
(28) The traditional view is not universally accepted.29

3828 J. Am. Chem. Soc., Vol. 122, No. 16, 2000
Bernasconi and Garc´ıa-R´ıo
Table 6. Brønsted Coefficients and Intrinsic Rate Constants for the
Reactions of HCtCCH2O- and CF3CH2O- with 6-Cr-Cl, 6-Cr-CF3
and 6-W-Cl in 50% MeCN-50% Water
Table 7. Brønsted Coefficients for Buffer Catalysis of MeOH
Addition to 6-Cr-Z
BH
Z
â (kB1 )
R (k )
-1
6-Cr-Cl
6-W-Cl
6-Cr-CF3
4-MeO
4-Me
H
4-F
4-Cl
0.17 ( 0.02
0.16 ( 0.02
0.17 ( 0.04
0.12 ( 0.02
0.10 ( 0.04
0.83 ( 0.02
0.84 ( 0.02
0.83 ( 0.04
0.88 ( 0.02
0.90 ( 0.04
d log kO1 R
d pKRa OH
ânuc
)
0.42
0.54
0.57
RO
-1
d log k
e-0.49
g0.91
e-0.57
g1.11
e-0.34
g0.91
âlg )
d pKRa OH
Table 8. 1H NMR Data for the Carbene Complexes in CDCl3
d log KR1 O
d pKRa OH
4-Me2N 4-MeO 4-Me 4-F 4-Cl
3-Cl
4-CF3
âeq
)
6-Cr-Z
CH3O
Ar
4.81
6.63
7.99
3.11
4.87
6.85
7.77
3.89
4.84 4.80 4.77 4.75
4.75
d log kR1 O
d log KR1 O
7.24 7.08 7.35 7.18, 7.25 7.30
7.36 7.48
2.40
ânnuc
)
e0.46
e0.42
e0.62
7.36
7.65
Z
6-W-Z
RO
-1
d log k
CH3O
Ar
4.76
6.92
7.89
3.79
4.77 4.78 4.77
7.23 7.10 7.43
7.63 7.71 7.54
2.40
ânlg )
log ko
e-0.54
g0.74
e-0.51
g0.94
e-0.38
g0.23
d log KR1 O
Z
RO
-1
RO
-1
compared to k , the true k values would be somewhat
As elaborated upon previously,9 the log ko values are about
2 log units lower than those for MeO- addition to carboxylate
esters. This is consistent with the notion that the transition state
in the carbene complex reactions is imbalanced while no such
imbalance occurs in the ester reactions because there is no
charge delocalization in the tetrahedral adducts.
lower and the true KR1 O somewhat higher than the reported
ones, especially for RO ) HCtCCH2O. This would mean that
the actual ânnuc values would be somewhat lower than those
listed in Table 6.
The fact that ânnuc is lower than Fn(kRO) is suggestive of a
1
small imbalance, but more definite conclusions will have to
await further study for two reasons. (1) The ânnuc value for
6-Cr-CF3 does not fit the pattern seen for 6-Cr-Cl and 6-W-
Cl. This may or may not be related to the problems with
systematic errors in the kinetic data for 6-Cr-CF3 mentioned
earlier. (2) The fact that ânnuc is based on only two nucleophiles
leaves too much room for uncertainty in these values. Neverthe-
less, the notion that the transition state of these reactions is
imbalanced in the sense shown in eq 3 is consistent with
observations for the reaction of thiolate ions with 9-W-H and
9-Cr-H26 where the ânuc values are very much lower than
Fn(kR1 S).30 It is also consistent with the intrinsic rate constants
discussed in the next section.31
Intrinsic Rate Constants. The logarithmic values of the
intrinsic rate constants for the addition of MeO- to 6-Cr-H and
6-W-H in methanol were estimated previously to be around
0.96 and 1.25, respectively.9 From the two-point Brønsted plots
for HCtCCH2O- and CF3CH2O- addition to 6-Cr-Cl and 6-W-
Cl in 50% MeCN-50% water, one obtains log ko ) 0.74 and
0.94, respectively (Table 6); these values are virtually the same
as those for the above reactions of MeO-. For the reaction of
6-Cr-CF3, log ko ) 0.23 is somewhat lower, part of which may
again be an artifact.
On the other hand, the intrinsic rate constants for alkoxide
ion addition to the carbene complexes are not as low as for
MeO- addition to some other electrophiles such as 2,6-dinitro-
4-methylsulfonylanisole (log ko ≈ -0.76)34 or â-methoxy-R-
nitrostilbene (log ko ≈ -3.9).35 A similar finding was reported
for thiolate ion addition to 6-Cr-H and 6-W-H where the
intrinsic rate constants were estimated to be about 2 orders of
magnitude higher than for thiolate ion addition to â-methoxy-
R-nitrostilbene.4e These results imply that resonance stabilization
of 7-M-Z- resulting from charge delocalization is not as strong
as in the adducts derived from these other electrophiles. This,
in turn, suggests that the transition state imbalance for the
carbene complex reactions should not be very large, consistent
with the relatively small differences between Fn(kRO) and ânnuc
1
discussed in the previous section.
Buffer-Catalyzed MeOH Addition to 6-Cr-Z. Fn(kB) val-
1
ues for reversible methanol addition to 6-Cr-Z catalyzed by
PhO-, 4-BrC6H4O-, and 3,4-Cl2C6H3O- buffers are summarized
in Table 5. Although the experimental uncertainties are larger
than for the Fn(kRO) values for MeO- addition (Table 4), thus
1
precluding firm conclusions, it nevertheless appears that the Fn-
(kB) values are somewhat larger than the Fn(kOR) values and
1
1
the Fn(kBH) values somewhat smaller than the Fn(k-RO1) values.
-1
(29) (a) Johnson, C. D. Tetrahedron 1980, 36, 3461. (b) Pross, A. J.
Org. Chem. 1984, 49, 1811. (c) Bordwell, F. G.; Hughes, D. L. J. Am.
Chem. Soc. 1985, 107, 4737.
(30) In these reactions, desolvation of the nucleophile complicates the
picture but still allows the conclusion that bond formation has made very
little progress at the transition state.26
(31) A reviewer has raised the question as to whether the observed
transition state imbalance might “be an effect imposed by the specific solvent
chosen rather than an inherent feature of the reaction system.” On the basis
of numerous studies that show the characteristics of imbalanced transition
states for a variety of reactions in a variety of solvents15,32 as well as in the
absence of any solvent,33 we conclude that imbalances are inherent features
of the reaction. On the other hand, the magnitude of the imbalance often
depends on the solvent.15,32,33
This suggests transition states with more advanced charge
development on the (CO)5Cr moiety and more bond formation
than in the MeO- reactions. Since methanol is a weaker
nucleophile than MeO-, even when assisted by a general base,
these results are consistent with the Hammond-Leffler
principle.27a,36
Additional information about the transition state is obtained
from the dependence of kB1 and kBH on the pKBa H of the buffer,
-1
which yields the Brønsted â and R values summarized in Table
7. The â values are quite small (0.10-0.17) and the R values
(32) Bernasconi, C. F.; Wenzel, P. J. J. Am. Chem. Soc. 1996, 118, 11446.
(33) (a) Bernasconi, C. F.; Wenzel, P. J. Am. Chem. Soc. 1994, 116,
5405. (b) Bernasconi, C. F.; Wenzel, P. J.; Keeffe, J. R.; Gronert, S. J. Am.
Chem. Soc. 1997, 119, 4008.
(34) Terrier, F.; Millot, F.; Morel, J. J. Org. Chem. 1976, 41, 3892.
(35) Bernasconi, C. F.; Fassberg, J.; Killion, R. B.; Schuck, D. F.;
Rappoport, Z. J. Am. Chem. Soc. 1991, 113, 4937.
(36) Hammond, G. S. J. Am. Chem. Soc. 1955, 77, 334.

Kinetics of ReVersible Alkoxide Ion Addition
J. Am. Chem. Soc., Vol. 122, No. 16, 2000 3829
Table 9. 13C NMR Data for the Carbene Complexes in CDCl3
complexes are more stable than in methanol or aqueous acetonitrile.
Reaction solutions in methanol or aqueous acetonitrile were only
prepared immediately prior to an experiment; the stability of these
solutions could be enhanced by adding 2.5 × 10-3 M CH3SO3H in
methanol or 2.5 × 10-3 M HCl in 50% MeCN-50% water. For 6-Cr-
CF3, the decomposition in aqueous acetonitrile even in the presence
of HCl was too rapid to allow the preparation of reaction solutions
outside the stopped-flow apparatus. Hence the experiments were
conducted by mixing a solution of 6-Cr-CF3 in pure MeCN with that
of the appropriate nucleophile in water, yielding a 50% MeCN-50%
water solution in the course of the stopped-flow experiment. Because
the mixing of water with MeCN leads to a decrease in temperature,
the initial temperature of the separate solutions was set at 29.5 °C in
order to achieve a temperature of 25 °C after mixing. The assumed
4.5° temperature drop was based on an Arrhenius plot of the reaction
of 6-Cr-Cl with 0.1 M KOH. This attempt at correcting for the cooling
of the reaction solution is somewhat crude and may have contributed
to the rate constants for the reactions of 6-Cr-CF3 to be deviating
negatively from the Hammett plots.
4-Me2N 4-MeO 4-Me 4-F 4-Cl 3-Cl
4-CF3
6-Cr-Z
C )
326.0 351.2 348.3 346.8 348.0 348.8 350.0
(CO)5 trans 223.6 224.8 224.0 223.7 223.8 223.8 223.7
cis 217.7 215.7 216.5 216.2 216.1 215.8 215.6
CH3O
Ar
66.1
65.7 67.2 67.4 67.4 67.4 67.4
153.8 154.0 151.0 162.3 151.7 154.8 156.2
139.9 132.1 141.9 149.9 137.0 134.0 130 or 128
132.3 129.6 128.8 127.3 128.4 129.9 125.4
123.8 124.8 115.0 125.4 122.0 122.4
Z
39.9
55.7 21.4
130 or 128
6-W-Z
C )
313.4 319.4 318.1 319.1
203.0 203.3 203.1 203.2
197.8 197.6 197.2 197.1
69.8 70.0 70.1 70.1
164.0 152.2 167.1 153.1
146.9 143.6 163.1 131.3
132.1 128.9 130.1 128.8
124.3 128.0 130.0 128.1
55.6 21.6
(CO)5 trans
cis
CH3O
Ar
For the experiments in the sequential mixing mode, the methoxide
ion adducts were generated in the presence of 0.1 M NaOMe. After a
delay time of 0.5-5 s, the solution containing the adduct was mixed
with the buffer solution; the buffer solution was such as to both fully
neutralize the NaOMe and leave enough free buffer acid to give a
convenient [B]/[BH] ratio.
Z
quite large (0.83-0.90), suggesting that the proton is close to
the methanol oxygen at the transition state (10). This, then, is
The reactions of the HCtCCH2O- and CF3CH2O- adducts with
triethylamine buffers were conducted in a different manner. Because
of competition with the OH- reaction, the adducts could not easily be
generated at sufficient concentrations in 50% MeCN-50% water. Hence
they were generated in pure acetonitrile instead by adding small amounts
of HCtCCH2O-/HCtCCH2OH and CF3CH2O-/CF3CH2OH (5 × 10-4
M) to the carbene complex. The adducts generated in this manner were
stable for approximately 20 min, thus obviating the need for sequential
mixing stopped-flow experiments. The kinetics of their reversal to
reactants were conducted by mixing the acetonitrile solution of the
adduct with an aqueous solution of the triethylamine buffer in the
stopped-flow apparatus. As described above for the reactions of 6-Cr-
CF3, the initial temperature was set at 29.5 °C in order to compensate
for the cooling during the mixing process.
another type of transition state imbalance where proton transfer
to the buffer base apparently lags behind C-O bond formation.
pKa of HCtCCH2OH and CF3CH2OH in 50% MeCN-50%
Experimental Section
Water. These pKa values were obtained as pKa ) pKw - log KOH
,
Materials. The various carbene complexes, 6-Cr-Z (Z ) 4-Me2N,
4-MeO, 4-Me, 4-F, 4-Cl, 3-Cl, and 4-CF3) and 6-W-Z (Z ) 4-MeO,
4-Me, and 4-Cl), were synthesized as described by Fischer et al.37 Tables
with pKw being the ionic product of the solvent38 and KOH being defined
by eqs 10 and 11. KOH was determined by measuring the pH of solutions
KOH
1
8 and 9 summarize their H and 13C NMR data which were recorded
ROH + OH- y z RO- + H2O
KOH ) [RO-]/[ROH][OH-]
(10)
(11)
in CDCl3 on a Bruker 250 MHz spectrometer. 6-Cr-H and 6-W-H
were available from a previous study.9 Phenol was recrystallized from
toluene; 4-bromophenol and 3,5-dichlorophenol (Aldrich) were used
without further purification. Triethylamine was refluxed over Na/CaH2
and freshly distilled prior to use. Methanol and reagent grade acetonitrile
were purified by fractional distillation and stored over 4-Å molecular
sieves. Water was taken from a Milli Q purification system.
Solutions of NaOMe in methanol were prepared by adding freshly
cut Na metal to methanol at 0 °C and allowing the reaction to go to
completion under nitrogen or argon. The concentrations of NaOMe
stock solutions were determined by titration against 0.08 M HCl using
phenolphthalein as indicator.
Stopped-Flow Kinetics. In the conventional mode, rates were
measured by monitoring the disappearance of the carbene complexes
around 400 nm (450 nm for 6-M-NMe2) in an Applied Photophysics
DX.17MV stopped-flow apparatus. For UV spectra of 1-M and adducts
of the general structure 2-M-, see refs 4d, 4e, and 9. Stock solutions
of the substrate were initially prepared in acetonitrile where the carbene
prepared by adding 1.60 × 10-2 M KOH and various concentrations
of the respective alcohol to the solvent, and making use of the
relationships [RO-] + [OH-] ) [KOH]0 and [RO-] + [ROH] )
[ROH]0. In evaluating the actual pH, 0.18 unit was added to the
measured pH according to Allen and Tidwell.40 These measurements
were made with an Orion 611 pH meter equipped with a glass electrode
and a Sure Flow (Corning) reference electrode.
Acknowledgment. This research was supported by Grant
CHE-9734822 from the National Science Foundation. L.G.R.41
thanks the Xunta of Galicia for a postdoctoral fellowship.
JA994174W
(38) pKw ) 15.19.39
(39) Bernasconi, C. F.; Sun, W. J. Am. Chem. Soc. 1993, 115, 12526.
(40) Allen, A. D.; Tidwell, T. T. J. Am. Chem. Soc. 1987, 109, 2774.
(41) Permanent address: Departamento de Quimica F´ısica, Universidad
de Santiago, Santiago de Compostela, Spain.
(37) Fischer, E. O.; Kreiter, C. G.; Kollmeier, H.; Mu¨ller, J.; Fischer, R.
C. J. Organomet. Chem. 1971, 28, 237.