Formation and Stability of N-Heterocyclic Carbenes in Water: The Carbon Acid pKa of Imidazolium Cations in Aqueous Solution

Article abstract of DOI:10.1021/ja039890j

We report second-order rate constants k_DO (M^-1 s ^-1) for exchange for deuterium of the C(2)-proton of a series of simple imidazolium cations to give the corresponding singlet imidazol-2-yl carbenes in D₂O at 25 °C and l = 1.0 (KCl). Evidence is presented that the reverse protonation of imidazol-2-yl carbenes by solvent water is limited by solvent reorganization and occurs with a rate constant of k_HOH = k_reorg = 10¹¹ s^-1. The data were used to calculate reliable carbon acid pK_as for ionization of imidazolium cations at C(2) to give the corresponding singlet imidazol-2-yl carbenes in water: pK_a = 23.8 for the imidazolium cation, pK _a = 23.0 for the 1,3-dimethylimidazolium cation, pK_a = 21.6 for the 1,3-dimethylbenzimidazolium cation, and pK_a = 21.2 for the 1,3-bis-((S)-1-phenylethyl)benzimidazolium cation. The data also provide the thermodynamic driving force for a 1,2-hydrogen shift at a singlet carbene: K₁₂ = 5 × 10¹⁶ for rearrangement of the parent imidazol-2-yl carbene to give neutral imidazole in water at 298 K, which corresponds to a favorable Gibbs free energy change of 23 kcal/mol. We present a simple rationale for the observed substituent effects on the thermodynamic stability of N-heterocyclic carbenes relative to a variety of neutral and cationic derivatives that emphasizes the importance of the choice of reference reaction when assessing the stability of N-heterocyclic carbenes.

Full text of DOI:10.1021/ja039890j

Published on Web 03/12/2004
Formation and Stability of N-Heterocyclic Carbenes in Water:
The Carbon Acid pKa of Imidazolium Cations in Aqueous
Solution
Tina L. Amyes,* Steven T. Diver,* John P. Richard, Felix M. Rivas, and
Krisztina Toth
Contribution from the Department of Chemistry, UniVersity at Buffalo,
SUNY, Buffalo, New York 14260
Received December 1, 2003; E-mail: tamyes@chem.buffalo.edu
Abstract: We report second-order rate constants kDO (M-1 s-1) for exchange for deuterium of the C(2)-
proton of a series of simple imidazolium cations to give the corresponding singlet imidazol-2-yl carbenes
in D2O at 25 °C and I ) 1.0 (KCl). Evidence is presented that the reverse protonation of imidazol-2-yl
carbenes by solvent water is limited by solvent reorganization and occurs with a rate constant of kHOH
)
kreorg ) 1011 s-1. The data were used to calculate reliable carbon acid pKas for ionization of imidazolium
cations at C(2) to give the corresponding singlet imidazol-2-yl carbenes in water: pKa ) 23.8 for the
imidazolium cation, pKa ) 23.0 for the 1,3-dimethylimidazolium cation, pKa ) 21.6 for the 1,3-
dimethylbenzimidazolium cation, and pKa ) 21.2 for the 1,3-bis-((S)-1-phenylethyl)benzimidazolium cation.
The data also provide the thermodynamic driving force for a 1,2-hydrogen shift at a singlet carbene: K12
) 5 × 1016 for rearrangement of the parent imidazol-2-yl carbene to give neutral imidazole in water at 298
K, which corresponds to a favorable Gibbs free energy change of 23 kcal/mol. We present a simple rationale
for the observed substituent effects on the thermodynamic stability of N-heterocyclic carbenes relative to
a variety of neutral and cationic derivatives that emphasizes the importance of the choice of reference
reaction when assessing the stability of N-heterocyclic carbenes.
Scheme 1
It is 40 years since the striking report of Olofson and co-
workers of facile deuterium exchange of the C(2)-proton of the
1,3-dimethylimidazolium cation (DMI) in D2O (Scheme 1).1
The deprotonation of imidazolium cations at C(2) results in the
formation of formally neutral carbon bases which are examples
of nucleophilic singlet carbenes that are strongly stabilized by
the presence of two heteroatoms at the carbenic center (Scheme
1).2-6 The electron-rich nature of N-heterocyclic carbenes has
led to their wide-ranging application in organometallic cataly-
sis,3,6,7 and they also serve as nucleophilic catalysts in several
important reactions such as benzoin condensation8-12 and acyl
Despite the isolation and characterization of a vast array of
transfer.13,14
stable N-heterocyclic and other diamino carbenes,2-4 there are
no systematic data for the formation and stability of imidazol-
2-yl carbenes in aqueous solution at room temperature. More-
over, the kinetic and thermodynamic acidity of the C(2)-proton
of simple imidazolium cations has not been examined in the
light of modern theories of proton transfer at carbon.
(1) Olofson, R. A.; Thompson, W. R.; Michelman, J. S. J. Am. Chem. Soc.
1964, 86, 1865-1866.
(2) Alder, R. W. In Carbene Chemistry; Bertrand, G., Ed.; FontisMedia S. A.;
Marcel Dekker Inc.: New York, 2002; pp 153-176.
(3) Bourissou, D.; Guerret, O.; Gabbaie, F. P.; Bertrand, G. Chem. ReV. 2000,
100, 39-91.
(4) Arduengo, A. J., III Acc. Chem. Res. 1999, 32, 913-921.
(5) Warkentin, J. In AdVances in Carbene Chemistry; Brinker, U. H., Ed.; JAI
Press Inc.: Stamford, CT, 1998; Vol. 2, pp 245-295.
(6) Herrmann, W. A.; Ko¨cher, C. Angew. Chem., Int. Ed. Engl. 1997, 36, 2162-
2187.
(7) Herrmann, W. A. Angew. Chem., Int. Ed. Engl. 2002, 41, 1290-1309.
(8) Breslow, R. J. Am. Chem. Soc. 1958, 80, 3719-3726.
(9) Wanzlick, H. W.; Kleiner, H. J. Chem. Ber. 1963, 96, 3024-3027.
(10) Lappert, M. F.; Maskell, R. K. J. Chem. Soc., Chem. Commun. 1982, 580-
581.
We report here a study of the deuterium exchange reactions
of the C(2)-proton for a series of simple imidazolium cations
(Chart 1) in D2O at 25 °C and I ) 1.0 (KCl). The data are used
to obtain reliable carbon acid pKas for ionization of these
imidazolium cations at C(2) to give the corresponding singlet
imidazol-2-yl carbenes in water. We also report the first deter-
mination of the thermodynamic driving force for a 1,2-hydrogen
shift at a singlet carbene: K12 ) 5 × 1016 for rearrangement of
(11) Teles, J. H.; Melder, J.-P.; Ebel, K.; Schneider, R.; Gehrer, E.; Harder,
W.; Brode, S.; Enders, D.; Breuer, K.; Raabe, G. HelV. Chim. Acta 1996,
79, 61-83.
(12) Miyashita, A.; Suzuki, Y.; Kobayashi, M.; Kuriyama, N.; Higashino, T.
Heterocycles 1996, 43, 509-512.
(13) Grasa, G. A.; Kissling, R. M.; Nolan, S. P. Org. Lett. 2002, 4, 3583-
3586.
(14) Nyce, G. W.; Lamboy, J. A.; Connor, E. F.; Waymouth, R. M.; Hedrick,
J. L. Org. Lett. 2002, 4, 3587-3590.
9
4366
J. AM. CHEM. SOC. 2004, 126, 4366-4374
10.1021/ja039890j CCC: $27.50 © 2004 American Chemical Society

N-Heterocyclic Carbenes in Water
A R T I C L E S
Chart 1
benzimidazolium cation (DMBI) and the 1,3-bis-((S)-1-phen-
ylethyl)benzimidazolium cation (DPEBI) in buffered D2O (pD
4.3-8.9) at 25 °C and I ) 1.0 (KCl) was followed by 1H NMR
spectroscopy at 500 MHz. There was no detectable hydrolysis
or decomposition of these substrates during deuterium exchange,
monitored for up to 40 days for imidazole, 60 days for DMI,
44 days for DMBI and 28 days for DPEBI.
the parent imidazol-2-yl carbene to give neutral imidazole in
water at 298 K, which corresponds to a favorable Gibbs free
energy change of 23 kcal/mol. Finally, we present a simple
rationale for the observed substituent effects on the thermody-
namic stability of N-heterocyclic carbenes relative to a variety
of neutral and cationic derivatives.
Experimental Section
The syntheses of DMI, DMBI, and DPEBI, the preparation of
buffers, and the NMR methods used to monitor deuterium exchange
are described in the Supporting Information.
Determination of pD and pKBD. Solution pD and the concentration
of deuterioxide ion at any pD were determined as described in our
previous work.15 The apparent value of pKBD ) 7.72 for ionization of
the imidazolium cation at nitrogen in D2O at 25 °C and I ) 1.0 (KCl)
was determined as the average value of pKBD ) pD - log([B]/[BD+])
for six different gravimetrically prepared solutions of 30 mM imidazole
buffer in D2O (I ) 1.0, KCl) at [B]/[BD+] ) 0.11-19.
Deuterium Exchange. All reactions were carried out in D2O at 25
°C and I ) 1.0 (KCl). The reactions of DMI, DMBI, and DPEBI were
initiated by dissolving the crystalline substrate in 2 mL of the
appropriate buffer in D2O to give a final substrate concentration of 10
mM for DMI and DMBI, and 5 mM for DPEBI. The reactions of
imidazole were initiated by bringing a previously frozen solution of
30 or 60 mM imidazole buffer in D2O to room temperature. Tetra-
methylammonium hydrogensulfate, which served as an internal stan-
dard, was added by making a g200-fold dilution of a stock solution in
D2O into the reaction mixture to give a final concentration of 1-2
mM. 700 µL of the reaction mixture was immediately transferred to
an NMR tube and incubated at 25 °C. The remainder was incubated at
25 °C and was used to monitor the pD of the reaction mixture during
deuterium exchange. There was no change ((0.03 units) in the pD of
the reaction mixtures during the first two-halftimes for deuterium
exchange.
Deuterium exchange was followed by 1H NMR spectroscopy during
the disappearance of 60-85% of the C(2)-proton of the substrate.
Values of the reaction progress, R, were calculated from the integrated
areas of the signals due to the C(2)-proton of the substrate (AC2H) and
the tetramethylammonium ion internal standard (AIS) at zero time and
time t, according to eq 1. The observed first-order rate constants for
exchange, kex (s-1), were determined from the slopes of linear
semilogarithmic plots of reaction progress against time according to
eq 2, which generally covered at least 2 halftimes. The values of kex
(s-1) were reproducible to (10%.
Figure 1. Representative 1H NMR spectra at 500 MHz of imidazole buffer
(30 mM, 80% free base, pD 8.30) obtained during deuterium exchange of
the C(2)-proton in D2O at 25 °C and I ) 1.0 (KCl). (A) Deuterium exchange
results in the disappearance of the broad singlet at 8.002 ppm due to the
C(2)-proton. The extent of deuterium exchange is indicated above each
spectrum. (B) Deuterium exchange results in the disappearance of the
doublet (J ) 1.1 Hz) at 7.215 ppm due to the C(4,5)-protons and the
appearance of an upfield-shifted singlet at 7.210 ppm (∆δ ) 0.005 ppm)
due to these protons at imidazole containing deuterium at C(2).
Figure 1 shows representative 1H NMR spectra at 500 MHz
of imidazole (30 mM, self-buffered at pD 8.30) obtained during
exchange for deuterium of the C(2)-proton in D2O at 25 °C.
The signal for the C(4,5)-protons at 7.215 ppm appears as a
doublet (J ) 1.1 Hz) because there is a small long-range coup-
ling to the C(2)-proton (Figure 1B, top spectrum). The signal
for the C(2)-proton appears as a broad singlet at 8.002 ppm
(Figure 1A, top spectrum). Deuterium exchange at C(2) of imi-
dazole results in the disappearance of the signal for the C(2)-
proton and of the doublet at 7.215 ppm due to the C(4,5)-
protons, and the appearance of a broad upfield-shifted singlet
at 7.210 ppm (∆δ ) 0.005 ppm) due to the same protons at
imidazole containing deuterium at C(2).
(AC2H/AIS)t
R )
(1)
(2)
(AC2H/AIS)o
lnR ) -kext
Results
Deuterium Exchange Followed by 1H NMR Spectroscopy.
The exchange for deuterium of the C(2)-proton of imidazole,
the 1,3-dimethylimidazolium cation (DMI), the 1,3-dimethyl-
(15) Richard, J. P.; Williams, G.; O’Donoghue, A. C.; Amyes, T. L. J. Am.
Chem. Soc. 2002, 124, 2957-2968.
9
J. AM. CHEM. SOC. VOL. 126, NO. 13, 2004 4367

A R T I C L E S
Amyes et al.
Figures S1 and S2 of the Supporting Information show the
1
changes in the H NMR spectra of DMI and DMBI during
deuterium exchange at C(2) in buffered D2O at 25 °C.
Rate Constants for Deuterium Exchange. Deuterium
exchange at C(2) of imidazole, DMI, DMBI and DPEBI in
the presence of various concentrations of buffer at pD 4.33-
8.88 in D2O at 25 °C and I ) 1.0 (KCl) was followed by
monitoring the disappearance of the broad singlet due to the
1
C(2)-proton of the substrate by H NMR spectroscopy. There
was no change in the total integrated area for all of the signals
due to the other protons of these substrates during deuterium
exchange at C(2). The first-order rate constants kex (s-1) for
deuterium exchange of the C(2)-proton of these substrates are
reported in Tables S1-S4 of the Supporting Information.
The essentially identical values of kex ((10%) for the
deuterium exchange reactions of 30 mM imidazole at pD 7.64
and 60 mM imidazole at pD 7.69 shows that, under these
conditions, the values of kex (s-1) for imidazole are equal to ko
(s-1) for solvent-catalyzed deuterium exchange. Figure 2 (0)
shows the pD-rate profile of the values of ko for the deuterium
exchange reaction of imidazole. There is a downward break at
the pKa of N-protonated imidazole which shows that the reactive
form of the substrate for the deuterioxide-ion-catalyzed reaction
is the cationic N-protonated imidazole.16,17 The plot of ko/fN+
against [DO-] according to eq 3 is linear, where fN+ is the
fraction of the substrate present in the reactive N-protonated
form that was calculated from the pD and pKBD ) 7.72 for the
apparent pKa of imidazole in D2O at 25 °C and I ) 1.0 (KCl).
Figure 2 (9) shows the pD-rate profile of the values of ko/fN+
for the DO--catalyzed deuterium exchange reaction of N-
protonated imidazole.
Figure 2. pD-rate profiles for the deuterium exchange reactions of the
C(2)-proton of imidazolium cations in D2O at 25 °C and I ) 1.0 (KCl).
Key: (0) Values of ko for imidazole; (9) Values of ko/fN+ for imidazole
where fN+ is the fraction of the substrate present in the reactive N-protonated
form that was calculated from the pD and pKBD ) 7.72 (see text); (b)
Values of ko for DMI (fN+ ) 1); (2) Values of ko for DMBI (fN+ ) 1);
([) Values of ko for DPEBI (fN+ ) 1). The solid lines of unit slope show
the fits of the data to eq 4 which gave the second-order rate constants kDO
(M-1 s-1) for DO--catalyzed exchange of the C(2)-proton listed in Table
1.
DMBI, and DPEBI. The second-order rate constants kDO (M-1
s-1) for DO--catalyzed exchange of the C(2)-protons of the
imidazolium cations obtained from these fits are given in Table
1.
Discussion
ko
) kDO[DO-]
(3)
(4)
Isotope Exchange at C(2) of Imidazolium and Benzimi-
dazolium Cations. The observed changes in the NMR spectra
for the imidazolium cation, DMI and DMBI upon incubation
in buffered D2O (pD 4.3-8.9) at 25 °C provide direct evidence
that disappearance of the C(2)-proton is accompanied by
deuterium incorporation at C(2) (Figures 1, S1, and S2). There
is no evidence for any hydrolysis of these imidazolium cations
or formation of the corresponding carbene dimers19 during
deuterium exchange at C(2) under these conditions.
fN+
ko
kDOKw
log
) log
+ pD
(
)
(
)
fN +
γOL
A 2-fold increase in the concentration of 20% free base
phosphate buffer from 50 mM (pD 6.34) to 100 mM (pD 6.35)
results in no significant change ((10%) in kex (s-1) for DMI.
Therefore, at these low buffer concentrations, the values of kex
(s-1) for DMI are equal to ko (s-1) for solvent-catalyzed
deuterium exchange. Figure 2 (b) shows the pD-rate profile of
the values of ko for the DO--catalyzed deuterium exchange
reaction of DMI determined in the presence of 25 mM acetate
and 50-100 mM phosphate buffer.
Deuterium exchange at C(2) of DMI in D2O under a variety
of conditions has been reported previously. A value of kDO
)
480 M-1 s-1 at 25 °C can be calculated from the data of Wong
and Keck,17,20 while the observed rate constant for deuterium
exchange at pD 8.9 and ca. 31 °C reported by Olofson et al.1
can be combined with Kw ) 10-14.70 for the ion product of D2O
at 30 °C18 to give kDO ) 1600 M-1 s-1 at 31 °C. These data
are in qualitative agreement with kDO ) 247 M-1 s-1 at 25 °C
and I ) 1.0 (KCl) determined in this work (Table 1). Deuterium
exchange at C(2) of the imidazolium cation in D2O has been
studied at elevated temperatures and values of kDO ) 107 M-1
s-1 at 37 °C,21 360 ( 120 M-1 s-1 at 33 °C,16 and 1300 M-1
Figure 2 also shows the pD-rate profiles of the values of ko
) kex (s-1) for the DO--catalyzed deuterium exchange reactions
of DMBI in the presence of 50 mM acetate and phosphate buffer
(2) and of DPEBI in the presence of 50 mM acetate buffer
([).
The solid lines of unit slope in Figure 2 show the fits of the
data to eq 4, where Kw ) 10-14.87 is the ion product of D2O at
25 °C,18 γOL ) 0.75 is the activity correction for lyoxide ion
under our experimental conditions,15 and fN+ ) 1.0 for DMI,
(19) Bo¨hm, V. P. W.; Herrmann, W. A. Angew. Chem., Int. Ed. Engl. 2000, 39,
4036-4038.
(20) Table 1 of ref 17 reports kDO ) 270 M-1 s-1 for deuterium exchange at
C(2) of the 1,3-dimethylimidazolium cation in D2O at 25 °C. However, an
analysis of the experimental data reveals considerable scatter and the fit to
(16) Vaughan, J. D.; Mughrabi, Z.; Wu, E. C. J. Org. Chem. 1970, 35, 1141-
1145.
(17) Wong, J. L.; Keck, J. H., Jr. J. Org. Chem. 1974, 39, 2398-2403.
(18) Covington, A. K.; Robinson, R. A.; Bates, R. G. J. Phys. Chem. 1966, 70,
3820-3824.
eq 4 of this work (fN+ ) 1, γOL ) 1) gives kDO ) 476 M-1 s-1
.
(21) Bradbury, J. H.; Chapman, B. E.; Pellegrino, F. A. J. Am. Chem. Soc. 1973,
95, 6139-6140.
9
4368 J. AM. CHEM. SOC. VOL. 126, NO. 13, 2004

N-Heterocyclic Carbenes in Water
A R T I C L E S
Table 1. Kinetic and Thermodynamic Acidity of the C(2)-Proton
of Substituted Imidazolium Cations in Water at 25 °C and I ) 1.0
(KCl)
dielectric relaxation of the solvent, which occurs with a rate
constant of kreorg ≈ 1 × 1011 s-1 15,25
.
This solvent reorganization
which places a molecule of DOD in a position to deliver a
deuteron to C(2) of the carbene is irreversible because it dilutes
the molecule of DOL, and so it necessarily results in formation
of the isotope exchange product.
Figure 3. Reaction coordinate profile for hydron transfer from C(2) of
imidazolium and thiazolium cations C(2)-L+ (L ) H, D) to deuterioxide
ion in D2O to give the free carbene that is limited by solvent reorganization
(kp > kreorg).
The following provide good evidence that the reverse proton
transfer from solvent water to C(2) of imidazol-2-yl carbenes
to give imidazolium cations is limited by the reorganization of
the solvent (kp > kreorg, Figure 3) and occurs with a limiting
rate constant of kHOH ) kreorg ) 1011 s-1
.
a Second-order rate constant for deprotonation of the imidazolium cation
at C(2) by deuterioxide ion in D2O determined by monitoring the
disappearance of the C(2)-proton by 1H NMR spectroscopy. b Second-order
rate constant for deprotonation of the imidazolium cation at C(2) by
hydroxide ion in H2O, calculated from the value of kDO using a secondary
solvent isotope effect of kDO/kHO ) 2.4 (see text). c Carbon acid pKa for
ionization of the imidazolium cation at C(2) in H2O, calculated from kHO
(M-1 s-1) and kHOH ) 1011 s-1 for the reverse protonation of the conjugate
base (carbene) by solvent water according to eq 5. The estimated uncertainty
in these pKa values is (0.5 units (see text).
(1) There are only small primary isotope effects of kH/kT )
2.94 and kD/kT ) 1.58 for hydron transfer from C(2) of the 3,4-
dimethylthiazolium cation 1a (L ) H, D, T) to lyoxide ion at
30 °C and these quantities exhibit a significant deviation from
the Swain-Schaad relationship.24 This shows that there is
significant internal return of the transferred hydron to the
thiazolium ylide/carbene 2 (kp/kreorg ≈ 3 for L ) H, Figure 3),24
so that the reverse protonation of 2 by solvent water is limited
largely by the physical “encounter” of a molecule of HOH with
the carbene. The values of kDO (M-1 s-1, Table 1) for C(2)
deprotonation of the imidazolium cations studied here, which
are structurally similar to thiazolium cations, are 20-8000-fold
smaller than that for deprotonation of 1a (L ) H).26,27 Therefore,
relative to the cationic azolium ion ground state, simple
imidazol-2-yl carbenes should be less stable than the thiazol-
2-yl carbene 2, so that their protonation by solvent water should
be even more limited by the solvent reorganization step with
s-1 at 50 °C22 have been reported. A temperature correction of
these rate constants using an activation energy of 21.4 kcal/
mol23 results in estimated values of kDO for the imidazolium
cation that are in qualitative agreement with kDO ) 36.9 M-1
s-1 at 25 °C and I ) 1.0 (KCl) determined in this work (Table
1). The differences between the earlier literature data and those
reported here are likely due to the fact that many of the earlier
experiments were conducted with only approximate control of
solution pD, ionic strength and temperature, and to the improved
sensitivity and sophistication of modern NMR spectroscopy
which allows for the use of ca. 100-fold lower substrate
concentrations.
Lifetime of Imidazol-2-yl Carbenes in Solvent Water.
Figure 3 shows the reaction coordinate profile for hydron
transfer from C(2) of imidazolium cations C(2)-L+ (L ) H,
D) to deuterioxide ion. Hydron transfer results initially in
formation of the corresponding imidazol-2-yl carbene C: in
close contact with a molecule of DOL. This complex may either
collapse with hydron transfer to regenerate the original substrate
(“internal return”, kp),24 or it may undergo reorganization by
kp > kreorg and kHOH ) kreorg
.
(2) There is a systematic increase in the secondary solvent
isotope effect for hydron transfer to lyoxide ion from C(2) of
N-substituted 4-methylthiazolium cations with their decreasing
(25) Kaatze, U. J. Chem. Eng. Data 1989, 34, 371-374.
(26) A value of kDO ) 3.1 × 105 M-1 s-1 for deprotonation of 1a (L ) H) by
deuterioxide ion in D2O at 25 °C can be estimated from the literature24
values of kDO ) 4.27 × 105 M-1 s-1 and ∆Hq ) 11.7 kcal/mol (Ea ) 12.3
kcal/mol) at 30 °C (I ) 2.0).
(27) Haake, P.; Bausher, L. P.; Miller, W. B. J. Am. Chem. Soc. 1969, 91, 1113-
1119.
(22) Takeuchi, Y.; Kirk, K. L.; Cohen, L. A. J. Org. Chem. 1978, 43, 3570-
3578.
(23) Noszal, B.; Rabenstein, D. L. J. Phys. Chem. 1991, 95, 4761-4765.
(24) Washabaugh, M. W.; Jencks, W. P. J. Am. Chem. Soc. 1989, 111, 683-
692.
9
J. AM. CHEM. SOC. VOL. 126, NO. 13, 2004 4369

A R T I C L E S
Amyes et al.
Scheme 2
reactivity from kDO/kHO ) 1.30 for 1b (L ) T) to 2.35 for 1a
(L ) T).24 Therefore, the secondary solvent isotope effect for
proton transfer from C(2) of the structurally similar but less
reactive (see above) imidazolium cations to lyoxide ion should
be essentially equal to the limiting maximum value of kDO/kHO
) 2.4.28 This is consistent with proton transfer that is limited
kHOH
kHO
pKa ) pKw + log
(5)
by the physical step of solvent reorganization with kp > kreorg
(
)
15,29
and kHOH ) kreorg
.
(3) The values of kex (s-1) for deuterium exchange at C(2) of
DMI and DMBI in the presence of 1.0 M acetate buffer at pD
4.40 or 5.78 are no more than 10% larger than the calculated
values of ko ) kDO[DO-] for the DO--catalyzed reactions of
these substrates at these pDs (Tables S2 and S3). This is on the
order of the experimental uncertainty in the values of kex ((10%)
so that our data provide no evidence for significant Brønsted
base catalysis by acetate anion of deuterium exchange at C(2)
of simple imidazolium and benzimidazolium cations in D2O.
The absence of detectable buffer catalysis of exchange strongly
supports the conclusion that kHOH ) kreorg for the protonation
of imidazol-2-yl carbenes by solvent water.15,29,30
The values of pKa ) 21.2-23.8 for the C(2)-proton of simple
imidazolium cations show that these cations are relatively weak
carbon acids whose carbon acidities in water are intermediate
between those of the prototypical neutral carbonyl carbon acids
acetone (pKa ) 19.3)34 and ethyl acetate (pKa ) 25.6).35
The value of pKa ) 23.0 for the C(2)-proton of DMI in water
(Table 1) is substantially higher than the previous estimate of
pKa ≈ 17 for the very similar 1,3,4-trimethylimdazolium cation27
that has been has been propagated in the contemporary
literature.36 However, it is very similar to the values of pKa )
22.7 for 337 and pKa ) 24.0 for 438 determined in DMSO. This
is consistent with the small solvent effect on the acidities of a
variety of thiazolium and alkylammonium cations, for which
the pKas in DMSO and water lie within 1 pK unit of each other.39
Intrinsic Reactivity of Imidazol-2-yl Carbenes in Water.
Figure 4 shows Brønsted-type rate-equilibrium correlations of
rate constants for deprotonation of several classes of carbon
acids by hydroxide ion in water, log kHO, with the pKa of the
carbon acid, with statistical corrections for the number of protons
p at the carbon acid. The extended linear correlation with a slope
of -0.40 for proton transfer from neutral monocarbonyl carbon
acids over a range of 20 pKa units (Figure 4, b) was discussed
in our earlier work.15,35,40 This correlation was expected to
exhibit a downward break to a slope of -1.0 at pKa ) 31 where
the reverse protonation of the carbanion by solvent water is
limited equally by kreorg for solvent reorganization and kp for
the proton-transfer step (Figure 3), so that kHOH ) kreorg/2 ) 5
× 1010 s-1. The downward break was verified for acetate anion
(pKa ) 33.5), whose dianionic enolate undergoes protonation
(4) The rate constants for proton transfer from several alcohol
solVents to singlet diphenylcarbene Ph2C: generated by fem-
tosecond laser flash photolysis of diphenyldiazomethane show
a good correspondence with the solvation time (dielectric
relaxation time) of the solvent.31 For example, kMeOH ) 1.1 ×
1011 s-1 (τ ) 9 ps) for proton transfer to singlet Ph2C: from
solvent methanol is very similar to the solvation time of
methanol, τMeOH ) 6.8 ps.31 At the MP2/DZ level the gas-phase
proton affinity of singlet diphenylcarbene (275 kcal/mol)32 is
computed to be 18 kcal/mol greater than that of unsubstituted
imidazol-2-yl carbene (257 kcal/mol).33 Therefore, even singlet
carbenes with a very highly basic carbon lone pair undergo
proton transfer from hydroxylic solvents that is limited by
solvent reorganization with a rate constant on the order of 1011
s-1
.
Carbon Acid pKa of Imidazolium Cations in Water. Table
1 gives the second-order rate constants kHO (M-1 s-1) for proton
transfer from C(2) of the imidazolium cation, DMI, DMBI, and
DPEBI to hydroxide ion in water that were calculated from
the values of kDO (M-1 s-1) using a secondary solvent isotope
effect of kDO/kHO ) 2.4 for proton transfer that is limited by
the solvent reorganization step.15,29 These values of kHO were
used to calculate carbon acid pKas for ionization of the
imidazolium cations at C(2) to give imidazol-2-yl carbenes in
water (Table 1), using eq 5 derived for Scheme 2, with Kw )
10-14 for the ion product of water at 25 °C and kHOH ) 1011
s-1 for the reverse protonation of the carbene by solvent water
(see above). The range of error in these pKa values is estimated
to be (0.5 units and stems largely from the uncertainty in the
value of kHOH for the reverse protonation of the carbenes by
solvent water.
by solvent water with kHOH ) kreorg ) 1011 s-1 15
The more
.
limited data for proton transfer from a series of cationic
monocarbonyl carbon acids (Figure 4, 9) define a linear
correlation of slope ) -0.44 that lies ca. 3.5 log units above
that for neutral monocarbonyl carbon acids, and which is
expected to exhibit a downward break to a slope of -1.0 at
around pKa ) 27.41-43
(34) Keeffe, J. R.; Kresge, A. J. In The Chemistry of Enols; Rappoport, Z., Ed.;
John Wiley and Sons: Chichester, 1990; pp 399-480.
(35) Amyes, T. L.; Richard, J. P. J. Am. Chem. Soc. 1996, 118, 3129-3141.
(36) Sauers, R. R. Tetrahedron Lett. 1996, 37, 149-152.
(37) Kim, Y.-J.; Streitwieser, A. J. Am. Chem. Soc. 2002, 124, 5757-5761.
(38) Alder, R. W.; Allen, P. R.; Williams, S. J. J. Chem. Soc., Chem. Commun.
1995, 1267-1268.
(39) Bordwell, F. G.; Satish, A. V. J. Am. Chem. Soc. 1991, 113, 985-990.
(40) Richard, J. P.; Amyes, T. L.; Toteva, M. M. Acc. Chem. Res. 2001, 34,
981-988.
(41) Rios, A.; Amyes, T. L.; Richard, J. P. J. Am. Chem. Soc. 2000, 122, 9373-
9385.
(28) Gold, V.; Grist, S. J. Chem. Soc., Perkin Trans. 2 1972, 89-95.
(29) Richard, J. P.; Williams, G.; Gao, J. J. Am. Chem. Soc. 1999, 121, 715-
726.
(30) Fishbein, J. C.; Jencks, W. P. J. Am. Chem. Soc. 1988, 110, 5087-5095.
(31) Peon, J.; Polshakov, D.; Kohler, B. J. Am. Chem. Soc. 2002, 124, 6428-
6438.
(32) Pliego, J. R., Jr.; De Almeida, W. B. J. Chem. Soc., Faraday Trans. 1997,
93, 1881-1883.
(33) Dixon, D. A.; Arduengo, A. J., III J. Phys. Chem. 1991, 95, 4180-4182.
9
4370 J. AM. CHEM. SOC. VOL. 126, NO. 13, 2004

N-Heterocyclic Carbenes in Water
A R T I C L E S
The systematic shift in the position of the downward break
to a slope of -1.0 for the correlations shown in Figure 4 to
lower carbon acid pKa on moving from neutral to cationic
monocarbonyl compounds and to azolium cations is a direct
consequence of the increasing intrinsic reactivity and decreasing
Marcus intrinsic barrier to proton transfer along this series of
carbon acids.15 It shows a strong correspondence with the
expected increasing extents of localization of negative charge
at the central carbon of the conjugate base.48 The tendency
toward a greater localization of negative charge at the R-carbon
of the formally neutral enolates of cationic monocarbonyl
compared with those of neutral monocarbonyl carbon acids has
been discussed in our earlier work.40,41 The small Marcus
intrinsic barrier to proton transfer from C(2) of imidazolium
cations is consistent with evidence for a high degree of
localization of the lone pair at the in-plane sp2-orbital of the
carbene conjugate base.33,49,50
Figure 4. Rate-equilibrium correlations of kHO (M-1 s-1) for deprotonation
of carbon acids by hydroxide ion with the pKa of the carbon acid in water
at 25 °C. The values of kHO and pKa were statistically corrected for the
number of acidic protons p at the carbon acid. (b) Correlation for neutral
monocarbonyl carbon acids.15,35,41,43 (9) Correlation for cationic monocar-
bonyl carbon acids.41-43 (2) Data for cyanoalkanes which define a slope
of -1.0.29 ([) Data for simple imidazolium cations from this work which
define a slope of -1.0 (Table 1). (O) Data for the 3-cyanomethyl-4-
methylthiazolium cation 1b (L ) H), pKa ) 16.9.24,77 The Eigen/Marcus
curve through the data for the imidazolium ([) and 3-cyanomethyl-4-
methylthiazolium (O) cations was constructed using an estimated Marcus
intrinsic barrier of 5.0 kcal/mol, as described in ref 47.
Figure 4 illustrates the kinetic and thermodynamic acidity of
carbon acids of varying structure over a wide range of carbon
acid pKa and substrate reactivity. It serves to emphasize the
extremely low intrinsic reactivity of neutral carbonyl compounds
as carbon acids. For example, a neutral monocarbonyl carbon
acid of pKa ) 15.74 for water undergoes proton transfer to
hydroxide ion 108-fold more slowly does than an imidazolium
cation of the same pKa.51 Therefore, imidazol-2-yl carbenes are
expected to be up to 108-fold more reactive toward water than
a typical enolate carbanion of the same thermodynamic basicity,
which implies a difference in the Marcus intrinsic barriers to
proton transfer of around 11 kcal/mol. The large intrinsic
reactivity of imidazol-2-yl carbenes toward proton donors such
as water suggests that their addition and expulsion from carbonyl
electrophiles is also intrinsically fast, which is consistent with
Figure 4 ([) shows the linear correlation with a slope of -1.0
for proton transfer to hydroxide ion from C(2) of the simple
imidazolium cations studied here (Table 1), whose conjugate
bases undergo protonation by solvent water with kHOH ) kreorg
) 1011 s-1. Therefore, the downward break for proton transfer
from C(2) of simple imidazolium cations is anticipated to occur
at pKa < 21.2 determined for DPEBI. Our earlier data29 for
proton transfer from simple cyanoalkanes (Figure 4, 2), whose
conjugate bases also undergo protonation by solvent water with
kHOH ) kreorg ) 1011 s-1 29
, necessarily lie on the same “limiting”
correlation of slope -1.0 defined by the data for imidazolium
cations.
(45) Bednar, R. A.; Jencks, W. P. J. Am. Chem. Soc. 1985, 107, 7117-7126.
(46) Marcus, R. A. J. Phys. Chem. 1968, 72, 891-899.
Scheme 3 shows the Eigen mechanism44 for proton transfer
from a carbon acid to hydroxide ion, where kreorg (s-1) is the
first-order rate constant for “encounter” of the conjugate base
of the carbon acid with water by solvent reorganization that
places a molecule of HOH in a “reactive” position (see Figure
3). The solid curve through the data for simple imidazolium
cations in Figure 4 ([) shows the calculated fit to this mech-
anism obtained using kd ) kHO ) 109.9 M-1 s-1 for the ther-
modynamically favorable diffusion-limited proton transfer be-
(47) Proton transfer from C(2) of the 3-cyanomethyl-4-methylthiazolium cation
1b (L ) H, pKa ) 16.9) to lyoxide ion is limited in part by proton transfer
with kp/kreorg ) 0.29 (Figure 3).24 The value of kreorg ) 1011 s-1 then gives
kp ) 2.9 × 1010 s-1. This value of kp was substituted into the Marcus ex-
pression of eq 7 derived at 298 K (Kw ) 10-14 M2, Ka )10-16.9 M) to give
an estimated Marcus intrinsic barrier46 for proton transfer between C(2)
2
1.36log(Kw/Ka)
4Λ
1
1.36
log kp )
17.44 - Λ 1 -
(7)
(8)
{
(
) }
kdk-pkreorg
kHO
)
k-pkreorg + k-dkreorg + k-dkp
tween HCN (pKa ) 9.0) and hydroxide ion,45 kreorg ) 1011 s-1
,
and an estimated Marcus intrinsic barrier46 for the actual proton-
transfer step (k-p or kp) of 5.0 kcal/mol.47 This Eigen/Marcus
curve shows that the downward break in the correlation for pro-
ton transfer between C(2) of imidazolium and thiazolium cations
and hydroxide ion is expected to occur at around pKa ) 18.4.
k-d Ka
k-p
)
kp
(9)
( k )(K )
d
w
of azolium cations and hydroxide ion of Λ ) 5.0 kcal/mol. The solid curve
through the data for imidazolium cations in Figure 4 ([) was calculated
according to eq 8, which is the steady-state equation for the Eigen
mechanism shown in Scheme 3, with values of kp and k-p calculated using
eq 7 (Λ ) 5.0 kcal/mol) and eq 9, respectively, kreorg ) 1011 s-1, kd ) 109.9
M-1 s-1 and k-d ) 109.9 s-1 (we assume no work term for the encounter
of hydroxide ion and the carbon acid so that kd ) k-d).
Scheme 3
(48) Bernasconi, C. F. AdV. Phys. Org. Chem. 1992, 27, 119-238.
(49) Heinemann, C.; Mueller, T.; Apeloig, Y.; Schwarz, H. J. Am. Chem. Soc.
1996, 118, 2023-2038.
(50) Arduengo, A. J., III; Dias, H. V. R.; Dixon, D. A.; Harlow, R. L.; Klooster,
W. T.; Koetzle, T. F. J. Am. Chem. Soc. 1994, 116, 6812-6822.
(51) The linear correlation for proton transfer from neutral monocarbonyl carbon
acids (Figure 4, b) is given by log(kHO/p) ) 6.496-0.401(pKa + log p),15,43
which gives kHO ) 1.53 M-1 s-1 for proton transfer from a carbon acid of
pKa ) 15.74. Interpolation of the calculated Eigen/Marcus curve47 for proton
transfer from C(2) of imidazolium cations (Figure 4, [) gives kHO ) 1.51
× 108 M-1 s-1 for proton transfer from C(2) of a hypothetical imidazolium
cation of pKa ) 15.74.
(42) Rios, A.; Crugeiras, J.; Amyes, T. L.; Richard, J. P. J. Am. Chem. Soc.
2001, 123, 7949-7950.
(43) Rios, A.; Richard, J. P.; Amyes, T. L. J. Am. Chem. Soc. 2002, 124, 8251-
8259.
(44) Eigen, M. Angew. Chem., Int. Ed. Engl. 1964, 3, 1-72.
9
J. AM. CHEM. SOC. VOL. 126, NO. 13, 2004 4371

A R T I C L E S
Amyes et al.
their efficacy as nucleophilic catalysts of benzoin condensation8-12
and acyl transfer.13,14
rationalize the apparent greater stability of thiazolium than of
imidazolium ylides/carbenes.36,60-63
1,2-Hydrogen Shift at Imidazol-2-yl Carbene. A signature
reaction channel of singlet carbenes is the 1,2-hydrogen shift
in which the carbenic carbon is converted to an olefinic carbon
at the product.52,53 The thermodynamic cycle in Scheme 4 (X
) NH) shows that the equilibrium constant for the 1,2-hydrogen
shift at singlet imidazol-2-yl carbene to give neutral imidazole
can be obtained directly from the acidity constants for ionization
of the imidazolium cation at nitrogen (KNH) and at C(2) (KCH),
according to eq 6.
More recently, the isolation of a wide variety of imidazol-
2-yl and other diamino carbenes as stable species2-4 has fuelled
computational investigations of their stability.33,36,49,56-59,64 One
conclusion from this work is that N-substituted imidazol-2-yl
carbenes 5a are more stable than their S-substituted thiazol-2-
yl counterparts 5b.36,58
Scheme 4
The data obtained in this work allow for the first comparison
of the thermodynamic acidities of imidazolium and thiazolium
cations in water. Table 2 gives the kinetic and thermodynamic
acidities of the C(2)-proton of several simple azolium cations
in water at 25 °C that were determined in this work or calculated
from literature data. The thermodynamic cycle in Scheme 4
shows that the substituent effect on the carbon basicity of
N-heterocyclic carbenes is the sum of the substituent effects
on the 1,2-H shift at the carbene to give the corresponding
neutral azole and on N-protonation of the neutral azole to give
the azolium cation. The difference in the pKas for ionization of
the unsubstituted imidazolium cation (Scheme 4, X ) NH) and
log K12 ) pKCH - pKNH
(6)
The values of pKCH ) 23.8 (Table 1) and pKNH ) 7.154 for
ionization of the imidazolium cation at carbon and nitrogen,
respectively, were substituted into eq 6 to obtain K12 ) 5.0 ×
1016 as the equilibrium constant for the 1,2-H shift at the parent
imidazol-2-yl carbene to give imidazole in water at 25 °C. This
corresponds to a favorable Gibbs free energy change of ∆Go )
-22.7 kcal/mol and it represents the first determination of the
thermodynamic driving force for a 1,2-H shift at a singlet
carbene.55 The concerted 1,2-H shift at singlet imidazol-2-yl
carbene is symmetry-forbidden3 and the kinetic barrier to this
reaction in the gas phase has been calculated to be in the range
∆Eq ) 40 to 47 kcal/mol56-59 so that this reaction cannot occur
directly. Rather, the 1,2-H shift at imidazol-2-yl carbene to give
imidazole in water occurs by proton transfer to the carbenic
carbon to give the imidazolium cation followed by loss of a
proton from nitrogen to give the neutral imidazole product
(Scheme 4).
the thiazolium cation (Scheme 4, X ) S) at C(2), ∆log KCH
)
3.9 (Table 2), and at nitrogen, ∆log KNH ) 4.7,65 then gives
∆log K12 ) -0.8 as the difference in the equilibrium constants
for the 1,2-H shift at the parent imidazol-2-yl and thiazol-2-yl
carbenes to give the corresponding azoles in water. This is
illustrated in Figure 5 which shows that the 3.9 pK unit (5.3
kcal/mol) smaller C(2) carbon acidity of the imidazolium cation
than of the thiazolium cation in water reflects the 6.4 kcal/mol
greater stabilization of the imidazolium than of the thiazolium
cation relative to the neutral azole, and the offsetting 1.1 kcal/
mol greater stability of the imidazol-2-yl carbene than of the
thiazol-2-yl carbene relative to the neutral azole.
Scheme 5
Substituent Effects on the Stability of N-Heterocyclic
Carbenes. The literature on the stability of N-heterocyclic
carbenes 5 presents conflicting conclusions about the effect of
a change from a nitrogen to a sulfur substituent X on the stability
of these species. Along with the classic report of Haake et al.
that DO--catalyzed deuterium exchange at C(2) of the thiazo-
lium cation 6b in D2O is ca. 3000-fold faster than for the
imidazolium cation 6a,27 there have been several attempts to
(52) Bonneau, R.; Liu, M. T. H. In AdVances in Carbene Chemistry; Brinker,
U. H., Ed.; JAI Press Inc.: Stamford, CT, 1998; Vol. 2, pp 1-28.
(53) Moss, R. A. In AdVances in Carbene Chemistry; Brinker, U. H., Ed.; JAI
Press Inc.: Stamford, CT, 1994; Vol. 1, pp 59-88.
(54) Catalan, J.; Claramunt, R. M.; Elguero, J.; Laynez, J.; Menendez, M.; Anvia,
F.; Quian, J. H.; Taagepera, M.; Taft, R. W. J. Am. Chem. Soc. 1988, 110,
4105-4111.
(55) To the best of our knowledge, the only other thermodynamic data for such
a process is ∆H ) -42.1 kcal/mol for the 1,2-H shift at methylchloro-
carbene to give vinyl chloride in heptane at 295 K, determined using
photoacoustic calorimetry.75
(56) Heinemann, C.; Thiel, W. Chem. Phys. Lett. 1994, 217, 11-16.
(57) Boehme, C.; Frenking, G. J. Am. Chem. Soc. 1996, 118, 2039-2046.
(58) Maier, G.; Endres, J. Eur. J. Org. Chem. 1998, 1517-1520.
(59) McGibbon, G. A.; Heinemann, C.; Lavorato, D. J.; Schwarz, H. Angew.
Chem., Int. Ed. Engl. 1997, 36, 1478-1481.
Scheme 5 shows a series of isodesmic reactions that result
in the interconversion of imidazol-2-yl and thiazol-2-yl carbenes
(60) Olofson, R. A.; Landesberg, J. M. J. Am. Chem. Soc. 1966, 88, 4263-
4265.
9
4372 J. AM. CHEM. SOC. VOL. 126, NO. 13, 2004

N-Heterocyclic Carbenes in Water
A R T I C L E S
Table 2. Effect of Nitrogen, Sulfur and Oxygen Heteroatom
Substituents on the Kinetic and Thermodynamic Acidity of the
C(2)-Proton of Simple Azolium Cations in Water at 25 °C
Figure 5. Effect of a change from a nitrogen to a sulfur heteroatom
substituent on the Gibbs free energy change for the 1,2-hydrogen shift
reaction of N-heterocyclic carbenes to give the neutral azole (|∆∆G| ) 1.1
kcal/mol) and for proton transfer to C(2) to give the azolium cation (|∆∆G|
) 6.4-1.1 ) 5.3 kcal/mol) in water at 298 K.
hydrogenation of the two carbenes computed at the B3LYP/6-
311G(d,p) level.58 It shows that, relative to the corresponding
hydrogen adduct, an imidazol-2-yl carbene is substantially more
stable than a thiazol-2-yl carbene. This reflects the greater
π-donating ability to carbon of nitrogen than of sulfur66,67 and
the greater stabilization of the carbene by electron donation of
the lone pair at nitrogen than of that at sulfur to the p-orbital at
C(2). By contrast, no such π-stabilization is possible at the
hydrogen adducts where C(2) is sp3-hybridized, so that the large
positive value of ∆Eo reflects the complete loss of π-interactions
of the lone pair(s) at the N or S heteroatom substituent with
C(2) on moving from the carbene to the hydrogen adduct.
The value of ∆Go ) +1.1 kcal/mol in water for Scheme 5B
(∆logK12 ) -0.8, this work) is close to the value of ∆Eo )
+2.5 kcal/mol in the gas phase calculated as the difference in
the energy changes for the 1,2-H shift at the two carbenes com-
puted at the B3LYP/6-31G** level.59,68 It shows that, relative
to the corresponding neutral azole, an imidazol-2-yl carbene is
of similar stability to a thiazol-2-yl carbene (Figure 5). This is
consistent with a similar stabilization of the carbene and the
neutral azole, both of which are formally aromatic, by π-interac-
tions of the lone pair(s) at the N or S heteroatom substituent.56
Finally, the value of ∆Go ) -5.3 kcal/mol in water for
Scheme 5C (∆log KCH ) 3.9, Table 2) shows that, relative to
the corresponding azolium cation, an imidazol-2-yl carbene is
substantially less stable than a thiazol-2-yl carbene. This is
a Second-order rate constant for deprotonation of the azolium cation at
C(2) by hydroxide ion in H2O. b Carbon acid pKa for ionization of the
azolium cation at C(2) in H2O to give the corresponding N-heterocyclic
carbene. c Data from Table 1. d Calculated from kDO ) 3.1 × 105 M-1 s-1
for the 3,4-dimethylthiazolium cation at 25 °C [ref 26] using a secondary
solvent isotope effect of kDO/kHO ) 2.35 [ref 24] with the assumption that
the combined effects of N-methylation (2.6-fold increase in kDO per
N-methyl group observed for the imidazolium cation, Table 1) and 4-
methylation (2-4-fold decrease in kDO at 50 °C, ref 22) result in essen-
tially no change in kHO for azolium cations. e Calculated from kHO (M-1
s-1) and kHOH ) 1011 s-1 for the reverse protonation of the conjugate base
(carbene) by solvent water according to eq 5. f Calculated from kHO ) 1.3
× 105 M-1 s-1 for the thiazolium cation (Table 2) and the 2.6-fold increase
in kDO per N-methyl group observed for the imidazolium cation (Table 1).
g Calculated from kHO ) 103 M-1 s-1 for the 1,3-dimethylimidazolium
cation (Table 1) and the 3 × 105-fold greater value of kDO for the 3,4-
dimethyloxazolium cation than for the 1,3,4-trimethylimidazolium cation
at 34 °C [ref 27]. h Estimated from kHO ) 2.73 × 107 M-1 s-1 for the
3-cyanomethyl-4-methylthiazolium cation 1b (L ) H) at 25 °C [ref 77] for
which pKa ) 16.9 [ref 24], with the assumption that proton transfer from
oxazolium cations to hydroxide ion follows the same Eigen correlation as
that shown in Figure 4 for the closely related imidazolium and thiazolium
cations.
and the products of their hydrogenation (Scheme 5A), 1,2-
hydrogen shift (Scheme 5B) and proton transfer (Scheme 5C)
reactions.
The value of ∆Eo ) +14.6 kcal/mol in the gas phase for
Scheme 5A was calculated as the difference in the energies of
(65) Calculated using the statistically corrected value of pKa ) 7.41 for the
imidazolium cation in water at I ) 0.1 and 25 °C54 and pKa ) 2.68 for the
thiazolium cation in water at 25 °C.76
(61) Olofson, R. A.; Landesberg, J. M.; Houk, K. N.; Michelman, J. S. J. Am.
Chem. Soc. 1966, 88, 4265-4266.
(62) Aldrich, H. S.; Alworth, W. L.; Clement, N. R. J. Am. Chem. Soc. 1978,
100, 2362-2366.
(66) Kapp, J.; Schade, C.; El-Nahasa, A. M.; Schleyer, P. v. R. Angew. Chem.,
Int. Ed. Engl. 1996, 35, 2236-2238.
(63) Scheffers-Sap, M. M. E.; Buck, H. M. J. Am. Chem. Soc. 1979, 101, 4807-
(67) Bernardi, F.; Bottoni, A.; Venturini, A. J. Am. Chem. Soc. 1986, 108, 5395-
4811.
5400.
(64) Alder, R. W.; Blake, M. E.; Oliva, J. M. J. Phys. Chem. A 1999, 103,
11 200-11 211.
(68) McGibbon, G. A.; Hrusak, J.; Lavorato, D. J.; Schwarz, H.; Terlouw, J. K.
Chem. Eur. J. 1997, 3, 232-236.
9
J. AM. CHEM. SOC. VOL. 126, NO. 13, 2004 4373

A R T I C L E S
Amyes et al.
Scheme 6
Summary. (1) The determination of reliable values of pKa
) 21.2-23.8 for ionization of the C(2)-proton of simple
imidazolium and benzimidazolium cations in water reveal these
cations to be relatively weak carbon acids whose acidities are
intermediate between those of the prototypical neutral carbonyl
carbon acids acetone and ethyl acetate.
(2) Imidazol-2-yl carbenes exhibit a high intrinsic reactivity
toward protonation by solvent water and the Marcus intrinsic
barrier to this reaction is estimated to be ca. 11 kcal/mol smaller
than that for protonation of the enolates of simple monocarbonyl
carbon acids.
(3) A thermodynamic cycle gives the thermodynamic driving
force for the 1,2-hydrogen shift at singlet imidazol-2-yl carbene
to give imidazole in water as ∆Go ) -22.7 kcal/mol (298 K).
(4) An analysis of substituent effects on the hydrogenation,
1,2-hydrogen shift and proton transfer reactions of simple N-
heterocyclic carbenes emphasizes the importance of the choice
of reference reaction when assessing the stability of N-hetero-
cyclic carbenes relative to a variety of neutral and cationic
derivatives.
Acknowledgment. We acknowledge the National Science
Foundation (CHE-0092434 to S. T. Diver) and the National
Institutes of Health (GM 39754 to J. P. Richard) for generous
support of this work.
consistent with the conclusion that imidazol-2-yl carbenes are
strongly stabilized by π-interactions, but that this stabilization
is even more important at the imidazolium cation.49 These data
serve to emphasize that substituent effects on organic reactions
depend not only on the stability of reactive intermediates (e.g.,
carbenes), but that they can also exhibit a very strong depen-
dence on the stability of the reacting ground state.69,70
The above discussion emphasizes the importance of the choice
of reference reaction when assessing the “stability” of N-het-
erocyclic carbenes. For example, the value of ∆Eo ) +18.9
kcal/mol for the reaction shown in Scheme 6A,71 which has
been taken as a measure of the relative stability of the formally
aromatic imidazol-2-yl carbenes and their saturated imidazolin-
2-yl counterparts,57,72 can be attributed to the greater stabilization
of the former carbenes by cyclic electron delocalization.49,57
However, this difference in thermodynamic stability is largely
eradicated when the reference reaction is the 1,2-H shift to give
the neutral azole (Scheme 6B, ∆Eo ) -4.0 kcal/mol73), or
proton transfer to give the azolium cation (Scheme 6C, ∆Eo ≈
0 kcal/mol74), both of which are sp2-hybridized at C(2) and can
therefore derive stabilization from cyclic electron delocaliza-
tion.49,56
Supporting Information Available: Details of the syntheses
of DMI, DMBI, and DPEBI, the preparation of buffers and
the NMR methods. Tables S1-S4: Rate Constants kex (s-1
)
for deuterium exchange at C(2) of imidazole, DMI, DMBI, and
DPEBI in buffered D2O at 25 °C. Figures S1-S2: 1H NMR
spectra of DMI and DMBI obtained during deuterium exchange
in buffered D2O at 25 °C. Figure S3: Representative semiloga-
rithmic plots of reaction progress against time. (PDF). This
material is available free of charge via the Internet at
http://pubs.acs.org.
JA039890J
(73) Calculated as the difference in the energy changes for the 1,2-H shift at
the two carbenes computed at the MP4/6-311G(d,p)//MP2/6-31G(d) level.57
(74) At the B3LYP/6-31G* level the 1,3-dimethylimidazol-2-yl carbene and the
corresponding saturated 1,3-dimethylimidazolin-2-yl carbene are computed
to have identical proton affinities of 266.0 kcal/mol.2
(69) Richard, J. P.; Amyes, T. L.; Rice, D. J. J. Am. Chem. Soc. 1993, 115,
2523-2524.
(70) Apeloig, Y.; Biton, R.; Abu-Freih, A. J. Am. Chem. Soc. 1993, 115, 2522-
(75) LaVilla, J. A.; Goodman, J. L. J. Am. Chem. Soc. 1989, 111, 6877-6878.
(76) Barszcz, B.; Gabryszewski, M.; Kulig, J.; Lenarcik, B. J. Chem. Soc., Dalton
Trans. 1986, 2025-2028.
2523.
(71) Calculated as the difference in the energies of hydrogenation of the two
carbenes computed at the MP4/6-311G(d,p)//MP2/6-31G(d) level.57 A value
of ∆Eo ) +14.4 kcal/mol has been computed at the B3LYP/6-31G(d)//
B3LYP/6-31G(d) level.72
(77) A value of kHO ) 2.73 × 107 M-1 s-1 for deprotonation of 1b (L ) H) by
hydroxide ion in H2O at 25 °C was calculated from the following literature
data24 at 30 °C (I ) 2.0): kDO ) 4.62 × 107 M-1 s-1, a secondary solvent
isotope effect of kDO/kHO ) 1.3, and ∆Hq ) 7.9 kcal/mol (Ea ) 8.5 kcal/
mol).
(72) Denk, M. K.; Rodezno, J. M.; Gupta, S.; Lough, A. J. J. Organomet. Chem.
2001, 617-618, 242-253.
9
4374 J. AM. CHEM. SOC. VOL. 126, NO. 13, 2004