Protonation of CpW(CO)2(PMe3)H: Is the metal or the hydride the kinetic site?

Article abstract of DOI:10.1021/ja002395s

We have compared the rate at which a proton is transferred to the hydride ligand of CpW(CO)₂(PMe₃)H (3) with the rate at which one is transferred to the metal. Qualitative evidence that protonation of the hydride ligand is faster is offered by observation of fast exchange between 3 and CF₃SO₃D and selective broadening of the trans hydride resonance of 3 during exchange with anilinium. Rate constants were obtained by ¹H NMR for H/D exchange between 3 and 4-tert-butyl-N,N-dimethylanilinium-N-d₁ (5-d₁); an unusual EIE of 0.19 was observed and explained by location of the vibrational normal modes of 3, 5, and their deuterated analogues via IR and Raman spectroscopy. A pK_a of 5.6(1) in CH₃CN was determined for CpW(CO)₂(PMe₃)-H₂⁺ (4a) by IR; the importance of homoconjugate pair interactions in protonation equilibria is illustrated and discussed. The exchange rate data and the rate constant for deprotonation of 4a by 4-tert-butyl-N,N-dimethylaniline (6), combined with the pK_a data, provide quantitative evidence that the kinetic site of protonation of 3 is the hydride ligand.

Full text of DOI:10.1021/ja002395s

J. Am. Chem. Soc. 2000, 122, 12235-12242
12235
Protonation of CpW(CO)2(PMe3)H: Is the Metal or the Hydride the
Kinetic Site?
Elizabeth T. Papish,† Francis C. Rix,‡ Nikos Spetseris,‡ Jack R. Norton,*,† and
Robert D. Williams§
Contribution from the Departments of Chemistry, Columbia UniVersity, 3000 Broadway,
New York, New York 10027, Colorado State UniVersity, Fort Collins, Colorado 80523, and
Princeton UniVersity, Princeton New Jersey 08544
ReceiVed July 3, 2000
Abstract: We have compared the rate at which a proton is transferred to the hydride ligand of CpW(CO)2-
(PMe3)H (3) with the rate at which one is transferred to the metal. Qualitative evidence that protonation of the
hydride ligand is faster is offered by observation of fast exchange between 3 and CF3SO3D and selective
broadening of the trans hydride resonance of 3 during exchange with anilinium. Rate constants were obtained
by 1H NMR for H/D exchange between 3 and 4-tert-butyl-N,N-dimethylanilinium-N-d1 (5-d1); an unusual EIE
of 0.19 was observed and explained by location of the vibrational normal modes of 3, 5, and their deuterated
analogues via IR and Raman spectroscopy. A pKa of 5.6(1) in CH3CN was determined for CpW(CO)2(PMe3)-
H2+ (4a) by IR; the importance of homoconjugate pair interactions in protonation equilibria is illustrated and
discussed. The exchange rate data and the rate constant for deprotonation of 4a by 4-tert-butyl-N,N-
dimethylaniline (6), combined with the pKa data, provide quantitative evidence that the kinetic site of protonation
of 3 is the hydride ligand.
Introduction
the former site will involve less geometric and electronic
rearrangement.1,5,6 The result of hydride ligand protonation will
be a dihydrogen complex.6,7
The reaction of transition-metal hydride complexes with
protons1 is important in the catalysis of the reduction of H+ to
By observing that NEt3 broadens the hydride resonances of
[Cp(dmpe)Ru(H2)]+ but not [Cp(dmpe)RuH2]+,8 Heinekey has
shown6b that the kinetic acidity of a dihydrogen complex will
be greater than that of the corresponding dihydride. Several
experiments in the literature suggest that the microscopic reverse
is true, i.e., that a dihydrogen intermediate will generally be
the kinetic product of the protonation of a hydride complex.
Heinekey has observed that Cp(*)Ru(PPh3)2H is protonated at
the hydride ligand at low temperature but isomerizes to a metal
dihydride upon warming.8 In a similar study by Morris et al.,
when Cp*Ru(dppp)H is protonated at -60 °C, a dihydrogen
complex is initially formed, and, at room temperature, the
dihydrogen complex slowly isomerizes to the trans dihydride.9
Upon protonation of Re(NO)(CO)L2H2 with CF3COOH at
low temperature, an intermediate dihydrogen complex is
observed by NMR prior to the loss of H2.10 A dihydrogen
intermediate is observed at low temperature in the protonation
H2 and in ionic hydrogenation reactions.4 The latter involve
2,3
the sequential transfer of H+ and H- to olefins, ketones, or
imines (eq 1).
H+ + R2CdE h R2C+sEH
R2C+sEH + HsM f R2CHsEH + M+
(E ) CR2, NR, or O)
(1)
One of us,1,5c and others,6 have suggested that there will be
a kinetic preference for protonating a hydride complex at the
hydride ligand rather than the metal center, because reaction at
* To whom correspondence should be addressed.
† Columbia University.
‡ Colorado State University.
§ Princeton University.
(1) For a review of the protonation reactions of transition-metal
complexes, see: Kristja´nsdo´ttir, S. S.; Norton, J. R. Acidity of Hydrido
Transition Metal Complexes in Solution. In Transition Metal Hydrides;
Dedieu, A., Ed.; VCH: New York, 1991.
(2) Michaile, S.; Hillman, M.; Eisch, J. J. Organometallics 1988, 7,
1059-1065 and references therein.
(3) (a) Collman, J. P.; Wagenknecht, P. S.; Hembre, R. T., Lewis, N. S.
J. Am. Chem. Soc. 1990, 112, 1294-1295. (b) Collman, J. P.; Wagenknecht,
P. S.; Lewis, N. S. J. Am. Chem. Soc. 1992, 114, 5665-5673. (c) Collman,
J. P.; Wagenknecht, P. S.; Hutchison, J. E. Angew. Chem., Int. Ed. Engl.
1994, 33, 1537-1554.
(4) (a) Smith, K.; Norton, J. R.; Tilset, M. Organometallics 1996, 15,
4515-4520. (b) Bullock, R. M.; Song, J. S. J. Am. Chem. Soc. 1994, 116,
8602-8612. (c) Song, J. S.; Szalda, D. J.; Bullock, R. M.; Lawrie, C. J.
C.; Rodkin, M. A.; Norton, J. R. Angew. Chem., Int. Ed. Engl. 1992, 31,
1233-1235.
(5) (a) Jordan, R. F.; Norton, J. R. J. Am. Chem. Soc. 1982, 104, 1255-
1263. (b) Jordan, R. F.; Norton, J. R. ACS Symp. Ser. 1982, 198, 403. (c)
Inorganic Reactions and Methods; Zuckerman, J. J., Ed.; VCH: New York,
1987; p 207.
(6) (a) Jessop, P. G.; Morris, R. H. Coord. Chem. ReV. 1992, 121, 155.
(b) Heinekey, D. M.; Oldham, W. J., Jr. Chem. ReV. 1993, 93, 913.
(7) (a) Kubas, G. J. J. Chem. Soc., Chem. Commun. 1980, 61-62. (b)
Kubas, G. J. Comments Inorg. Chem. 1988, 7, 17-40. (c) Kubas G. J. Acc.
Chem. Res. 1988, 21, 120-128. (d) Crabtree, R. H. Acc. Chem. Res. 1990,
23, 95-101. (e) Koelle, U. New J. Chem. 1992, 16, 157-169.
(8) Chinn, M. S.; Heinekey, D. M. J. Am. Chem. Soc. 1990, 112, 5166-
5175.
(9) Jia, G.; Lough, A. J.; Morris, R. H. Organometallics 1992, 11, 161-
171.
(10) Feracin, S.; Bu¨rgi, T.; Bakhmutov, V.; Eremenko, I.; Vorontsov,
E. V.; Vimenitis, A. B.; Berke, H. Organometallics 1994, 13, 4194-4213.
10.1021/ja002395s CCC: $19.00 © 2000 American Chemical Society
Published on Web 11/21/2000

12236 J. Am. Chem. Soc., Vol. 122, No. 49, 2000
Papish et al.
of the iron hydride in eq 2, although a dihydride complex is
the sole thermodynamic product.11
Scheme 1
Scheme 2
When (CMe5)2WH2 is protonated by DCl/D2O, the major
product ([Cp*2WH2D]+) has a central deuteride ligand.12 This
product cannot result from direct metal protonation because there
is no occupied tungsten orbital between the hydride ligands, so
the charge-controlled formation of an intermediate η3-H2D or
an (η2-HD)(H) dihydrogen complex (eq 3) has been suggested.12
A kinetic study has confirmed the presence of an intermediate.13
in an iron dihydride which isomerized to a dihydrogen complex
upon warming. Loss of H2 above 10 °C precluded isolation of
the iron dihydrogen complex. This is the only example in the
literature of the kinetic site of protonation being the metal despite
the thermodynamic site being the hydride ligand. Therefore, the
kinetic preference for hydride ligand protonation is very strong,
and a unique combination of steric and electronic factors are
needed to reverse this preference.
We decided to search for a dihydrogen intermediate in the
protonation of CpW(CO)2(PMe3)H (3), known eventually to
form the cationic dihydride [CpW(CO)2(PMe3)H2]+ (4). This
system had the advantage over those previously studied in that
both 3 and 4 had been fully characterized. The hydride complex
Chaudret has recently shown that the hydride ligand of 1
undergoes intramolecular proton exchange with the ammonium
acid (eq 4b).14 Although the facile loss of H2 in eq 4a (which
1
3 was first synthesized in 1970 and identified by H and 31P
NMR and IR spectroscopy, and exists in solution in both cis
(3c) and trans (3t) forms which interconvert rapidly on the NMR
time scale (Scheme 1).18 The structure of the cationic dihydride
4 has been determined by X-ray diffraction.19 The hydride
ligands of the cationic dihydride 4 interconvert rapidly at room
temperature, and two inequivalent hydride resonances are not
observed until the temperature is lowered to -112 °C.20
If the kinetic site of protonation of 3 is the hydride ligand,
H/D exchange between the hydride complex 3 and an acid will
occur faster than protonation of the W to form the dihydride
cation 4 (Scheme 2). We have therefore compared the rate of
H/D exchange between 3 and various acids with the rate at
which these acids protonate the metal.
presumably goes through 2) shows that the cationic H2 complex
2 cannot be an intermediate in eq 4b, an intermediate like that
drawn in eq 4bswith amine coordination stabilizing the
coordinated H2sremains a possibility.15 A dihydrogen complex
has been observed during the protonation of ruthenium com-
plexes similar to 1.16
Experimental Section
General. All manipulations were carried out using Schlenk, high-
vacuum, or inert-atmosphere-box techniques, unless otherwise indicated.
CpW(CO)2(PMe3)H (3) was prepared by a literature procedure.21 CpW-
(CO)2(PMe3)D (3-d1) (90% D) was formed via reaction of 3 with KH,
followed by quenching with CF3COOD. [CpW(CO)2(PMe3)H2][BF4]
(4a) was formed by treating 3 with HBF4‚OMe2.20
In an attempt to make a silane σ complex, Brookhart et al.
inadvertently generated the iron hydride in eq 5 in situ from
[PhNH3‚(OEt2)1-2][B(Arf)4] was prepared by treating [H‚(OEt2)2]-
[B(Arf)4] [Arf ) 3,5-C6H3(CF3)2]22 with PhNH2 in ether. [CpWH2(CO)2-
PMe3][B(Arf)4] (4b) was similarly prepared by treating [H‚(OEt2)2]-
[B(Arf)4] [Arf ) 3,5-C6H3-(CF3)2]22 with 3 in CH2Cl2. 4-tert-Butyl-
N,N-dimethylaniline (6) was distilled from sodium and treated with
HBF4‚OMe2 in ether to form the anilinium tetrafluoroborate (5), which
was recrystallized from CH2Cl2 and hexane. The deuterated acid
(5-d1) was prepared by repeated exchange of 5 with EtOD, yielding
97% D incorporation. The chloride salt of 5 was prepared by treating
6 with HCl/ether, followed by ether removal under vacuum. 4-tert-
CpFe(PEt3)(SiEt3)CO, H(OEt2)2 BAr′4-, and adventitious H2O.17
+
Protonation, most likely by Et3SiOH2+ produced in situ, resulted
(11) Hamon, P.; Toupet, L.; Hamon, J.; Lapinte, C. Organometallics
1992, 11, 1429-1431.
(12) Parkin, G.; Bercaw, J. E. J. Chem. Soc., Chem. Commun. 1989,
255-257.
(13) Henderson, R. A.; Oglieve, K. E. J. Chem. Soc., Dalton Trans. 1993,
3431-3439.
(14) Ayllon, J. A.; Sayers, S. F.; Sabo-Etienne, S.; Donnadieu, B.;
Chaudret, B. Organometallics 1999, 18, 3981-3990.
(15) This possibility is supported by the effect on the proton exchange
rate of the base remaining in solution after 1 is formed (Bu3P makes
exchange faster than does Et3N).13 A similar effect (of counterion on H/D
exchange rate) was reported by Berke and co-workers with Re(NO)(CO)-
L2H2 in ref 10.
(18) Kalck, P.; Pince, R.; Poiblanc, R.; Roussel, J. J. Organomet. Chem.
1970, 24, 445-452.
(19) Bullock, R. M.; Song, J. S.; Szalda, D. J. Organometallics 1996,
15, 2504-2516.
(20) Ryan, O. B.; Tilset, M.; Parker, V. D. J. Am. Chem. Soc. 1990,
112, 2618-2626.
(21) (a) Tate, D. P.; Knipple, W. R.; Augl, J. M. Inorg. Chem. 1962, 1,
433-434. (b) Keppie, S. A.; Lappert, M. F. J. Chem. Soc. (A) 1971, 3216-
3220. (c) Kalck, P.; Poiblanc, R. J. Organomet. Chem. 1969, 19, 115-
121.
(16) (a) Chu, H. S.; Lau, C. P.; Wong, K. Y.; Wong, W. T. Organo-
metallics 1998, 17, 2768. (b) Caballero, A.; Jalon, F. A.; Manzano, B. R.
J. Chem. Soc., Chem. Commun. 1998, 1879.
(17) Scharrer, E.; Chang, S.; Brookhart, M. Organometallics 1995, 14,
5686-5694.
(22) Brookhart, M. S.; Grant, B.; Volpe, A. F. Organometallics 1992,
11, 3920-3922.

Protonation of CpW(CO)2(PMe3)H
J. Am. Chem. Soc., Vol. 122, No. 49, 2000 12237
Table 1. H/D Exchange between 3 and 5
c
c
d
[5-d1] (M)a
[3] (M)a
[5] (M)a
[3-d1] (M)a
T (K)b
kf (M-1 s-1
)
kr (M-1 s-1
)
Rc
Keq
0.061
0.116
0.117
0
0
0
0
0.051
0.058
0.053
0
0
0
0
0
0
0
0
0
0
283
293
293
273
263
283
283
8.1 × 10-4
1.5 × 10-3
1.8 × 10-3
3.6 × 10-4
1.9 × 10-4
7.5 × 10-4
8.4 × 10-4
4.2 × 10-3
8.8 × 10-3
9.4 × 10-3
1.7 × 10-3
1.0 × 10-3
4.3 × 10-3
4.6 × 10-3
0.006
0.01
0.01
0.04
0.02
0.01
0.01
0.19
0.17
0.2
0.21
0.19
0.18
0.18
0.105
0.104
0.127
0.127
0.051
0.048
0.058
0.055
a Initial concentrations. b Temperature at which kinetics were monitored. c Determined by MacKinetics. Keq ) kf/kr.
Table 2. Stopped-Flow Kinetics of Deprotonation of 4a
d
kr mp
kr mp
kr mp
kr mp
kr mp
[6] (M)
(223 Kc, s-1
)
(233 Kc, s-1
)
(243 Kc, s-1
)
(253 Kc, s-1
)
(263 Kc, s-1
)
0.00511a
0.00652b
0.00835b
0.0102b
9.2(4)
11.0(9)
12(1)
16.3(5)
21(3)
25(2)
29(3)
34(3)
42(5)
53(5)
50(3)
66(6)
77(8)
4.5(3)
6.1(3)
a Four kinetic runs were done at each temperature. b Between 9 and 11 kinetic runs were done at each temperature. c Represents the average
temperature; the true temperature used in extracting kinetic data was within 1 K of this value.
Butyl-N,N,N-trimethylanilinium tetrafluoroborate was prepared by a
variation of a literature method:23 6 was treated with [Me3O][BF4] in
CH2Cl2, followed by precipitation with hexanes. The tetrafluoroborate
salts of 2,4-dichloroanilinium, 4-cyanoanilinium, and 4-carboethoxy-
anilinum were prepared by treating HBF4‚OMe2 with the corresponding
aniline in ether, followed by recrystallization from ethanol and drying
under vacuum.
CH3CN and CD3CN were purified by a series of steps that our
previous work had shown to be effective:24,25 they were first treated
with anhydrous cupric sulfate to remove amines; then, they were
fractionally distilled from P4O10, discarding the first 5-10% of the
distillate to remove acrylonitrile; last, they were distilled from CaH2,
again discarding the first 5% of the distillate. Et2O, THF, and hexanes
were distilled under N2 from sodium/benzophenone. CH2Cl2 and CD2Cl2
were distilled under N2 from P4O10.
a CD2Cl2 solution of 3-d1 (0.104 M, 300 µL). The 3-d1 solution was
transferred by cannula onto the anilinium solution, which was at -78
°C. An 1H NMR spectrum was taken at 223 K to ensure that no
exchange had taken place, and then the probe was warmed to an
appropriate temperature (see Table 1). The progress of the reaction
was observed by monitoring the hydride resonances (δ -7.9, 3t; δ
-8.0, 3c). The forward and backward rate constants were obtained using
MacKinetics27 simulation software, with initial concentrations for all
species and the time dependence of the tungsten hydride as input.
Stopped-Flow System. The variable-temperature stopped-flow
system has been previously described.28 One spectrum was taken each
millisecond with an RSM 1000 monochromator, calibrated with
holmium oxide. The photomultiplier tube output from the SF-41 was
converted to digital form with an Analog Devices A/D converter and
analyzed (on a Gateway 2000 Pentium PC) with OLIS-RSM software.
Kinetics of Deprotonation of 4a by 6. A CH2Cl2 solution of 4a
(1.18 × 10-3 M) was placed in one reservoir by syringe. The other
reservoir was filled with a CH2Cl2 solution that contained an excess of
6 (1.71 × 10-2 M). The reaction was monitored by observing the
hydride absorbance at 346 nm. The aniline concentration and the
temperature were varied in several experiments (see Table 2).
Raman Spectra. The spectra were obtained in vacuum-sealed NMR
tubes containing crystals of 3c and 3c-d1, using a 135° backscattering
geometry and a triple-stage spectrograph (Spex Triplemate) equipped
with a Princeton Instruments diode array detector. Excitation source
was a Spectra Physics model 2020 Ar+ laser. The spectra were
referenced with DMF and processed with Grams/32 software (Galactic
Corp.).
NMR. 1H, 19F, and 13C NMR spectra and kinetic data were recorded
on a Bruker 300-MHz instrument. Temperatures (223-283 K) were
checked with a methanol thermometer;26 the uncertainty in temperature
was estimated to be (1 K between 223 and 283 K. NMR line shape
simulations were done with Cherwell Scientific’s gNMR 4.1 software.
Kinetics of Cis/Trans Isomerization. 1H NMR spectra of a CD2Cl2
solution of 0.065 M 3 were taken in 5 K increments from 268 to 298
K, and the rates were determined by line shape simulation.
H/D Exchange between 3 and Triflic Acid. A CD2Cl2 solution
(600 µL) of 3 (0.0742 M) was placed in a J. Young NMR tube and
cooled at -80 °C. A separate vacuum bulb was charged with 2.7 µL
of CF3SO3D, and the latter was vacuum transferred onto the hydride
1
solution. H NMR spectra were taken at 223 K; the progress of the
reaction was observed by monitoring the hydride resonances (δ -7.9,
3t; δ -8.0, 3c; δ -2.5, 4a).
X-ray Structure Determination of 3. Crystals of 3 were grown
from hexane at -30 °C. Single-crystal data collection and refinement
parameters are summarized in Table 3. Data were collected on a Bruker
P4 diffractometer equipped with a SMART CCD detector. The structure
was solved using direct methods and standard difference map techniques
and refined by full-matrix least-squares procedures using SHELXTL.29
Hydrogen atoms on carbon were included in calculated positions.
Crystals of 3 were ground to a powder and sprinkled onto a Vaseline-
coated glass lens, which was rotating during powder X-ray data
Proton Exchange between 3 and [PhNH3‚(OEt2)1-2][B(Arf)4]. In
a typical experiment, the 1H NMR spectra of 3 (0.019 M) were
examined in the presence of [PhNH3‚(OEt2)1-2][B(Arf)4] [Arf ) 3,5-
C6H3(CF3)2] (0.0084 M) in CD2Cl2 at 206-273 K. The exchange was
monitored by observing the line broadening of the hydride resonances
(δ -7.9, 3t; δ -8.0, 3c).
1
Deprotonation of [CpWH2(CO)2PMe3][B(Arf)4] (4b). H NMR
Line Widths of the Resulting 3. In a typical experiment, a solution
of 4b (0.042M) in CD2Cl2 and increasing amounts of PhNH2 (0.014-
0.087 M) were combined in a NMR tube. The trans hydride resonance
was broadened into the baseline until the concentration of aniline was
0.087 M, at which point the rates of exchange and metal protonation
of the resulting 3 were determined by observing the broadening of the
hydride resonance (δ -7.9, 3t; δ -8.0, 3c) and the PMe3 resonance (δ
1.63, 3t; δ 1.59, 3c) at 273 K. The relative rate constants for metal
and hydride protonation were determined as described in the text.
H/D Exchange between 3 and 5. A CD2Cl2 solution (300 µL) of 5
(0.232 M) and C6H18O3Si3 (as an internal standard) (0.06 M) was placed
in a septum-sealed NMR tube. A separate NMR tube was charged with
(23) Petti, M. A.; Shepodd, T. J.; Barrans, R. E., Jr.; Dougherty, D. A.
J. Am. Chem. Soc. 1988, 110, 6825-6840.
(24) Moore, E. J.; Sullivan, J. M.; Norton, J. R. J. Am. Chem. Soc 1986,
108, 2257-2263.
(25) (a) Burfield, D. R.; Lee, K. H.; Smithers, R. H. J. Org. Chem. 1977,
42, 3060-3065. (b) The authors thank Professor V. L. Pecoraro for
suggesting distillation from CaH2 to neutralize any acids present after the
P4O10 treatment.
(26) Van Geet, A. L. Anal. Chem. 1970, 42, 679.
(27) McKinney, R. J.; Weiher, J. F.; Leipold, W. S. MacKinetics; E. I.
du Pont de Nemours, Inc.: Wilmington, DE, 1992-1995.
(28) Eisenberg, D. C.; Lawrie, C. J. C.; Moody, A. E.; Norton, J. R. J.
Am. Chem. Soc 1991, 113, 4888-4895.

12238 J. Am. Chem. Soc., Vol. 122, No. 49, 2000
Table 3. Summary of Crystallographic Data of 3c
Papish et al.
empirical formula
formula weight
temperature, K
crystal system
space group
a, Å
C10H15O2PW
382.04
238(2)
monoclinic
P21/c
8.3686(10)
12.4131(14)
12.1808(14)
90
105.602(2)
90
1218.7(2)
4
2.082
0.71073
9.584
6920
2684
2684/0/153
1.047
0.0386, 0.1053
0.1143, 0.1462
b, Å
c, Å
R, deg
â, deg
γ, deg
V, Å3
Z
D
calc, g/cm3
λ(Mo KR), Å
µ, mm-1
no. of data collected
no. of unique data
data/restraints/param
GOF on F2
Figure 1. Thermal ellipsoid plot of CpW(CO)2PMe3H (3).
R1, wR2 (I > 2σ(I))
R1, wR2 (all data)
Table 4. Determination of the CH3CN pKa(4a) from the
Equilibrium between 3 and BH+
acid (BH+)
pKa(BH+)
Keq
pKa(4a)
4-carboethoxyanilinum
8.2(1)
7.6(1)
7.0(2)a
2.5 × 10-3
5.6(1)
5.6(1)
5.5(1)
4-cyanoanilinium (8-H+)
2,4-dichloroanilinium (7-H+)
0.011
0.029
a This pKa had been previously reported as 8.0,31 but the observation
of such a high Keq between 3 and 7-H+ led us to reinvestigate this pKa.
collection. The radiation source was Cu KR, λ1 ) 1.54060 Å, λ2 )
1.54439 Å, and data were collected from 2θ ) 1.025 to 49.975 (Figure
2a). A calculated X-ray diffraction powder pattern of 3c was obtained
using X-Pro software.29
Figure 2. XRD powder pattern of 3 (Cu radiation): (a) experimental
Calculation of the Normal Modes of 5 and 5-d1. These calculations
were done for the gas-phase cations using Jaguar version 3.5, release
42, basis set 6-31G**.30 These results were helpful in assigning the
bending modes in the experimental IR spectra; see Supporting Informa-
tion for details.
Determination of the CH3CN pKa of 4a. All pKa determinations
were done in CH3CN using a CaF2 cell. The equilibria between 3 and
the three anilinium acids listed in Table 4 were observed by IR. For
each acid used, the Keq was determined from the absorbance of the
νCO bands (3 absorbs at 1919 and 1831 cm-1, with known ꢀ;24 4a
absorbs at 2073 cm-1 with ꢀ ) 1600 M-1 cm-1 and at 2015 cm-1 with
ꢀ ) 1700 M-1 cm-1). Two experiments at different starting acid
concentrations were done with each acid.
and (b) calculated.
Table 5. Kinetics of Cis-Trans Isomerization of 3
kc f t
kt f c
∆Hq (kcal/mol)
∆Sq (eu)
15.5(2)
-0.2(9)
15.51(2)a
0.72(3)a
15.4(3)
0.2(1)
∆Gq263K (kcal/mol)
15.39(2)a
0.91(4)a
k263K (s-1
)
a Extrapolated from rate measurements at 268-298 K.
obtained (Figure 2b) that agreed with the observed powder
pattern of crystals of 3 (Figure 2a). Thus, in the solid phase, 3
is exclusively 3c. Hence, the cis isomer of 3 preferentially
crystallizes from the mixture of cis and trans isomers found in
solution.
Determination of the CH3CN pKa of 5. The equilibrium between
1
5 and pyridinium triflate was observed from the averaged H NMR
chemical shift of the aromatic protons of 5 and 6 in fast exchange during
six experiments at different starting concentrations; see Supporting
Information for details.
Rates of Cis/Trans Isomerization. In solution, 3c and 3t
interconvert rapidly on the NMR time scale (Scheme 1). The
published line shape analysis showed that the first-order rate
of isomerization kcis f trans of 3 is 4.62 s-1 at 280 K in toluene-
d8.18 The kinetics of interconversion between 3c and 3t (Scheme
2) were redetermined by variable-temperature NMR in CD2Cl2
(a more practical solvent for proton-transfer reactions). Line
shape analysis of the PMe3 resonance (δ ) 1.58 ppm) gave the
rate of cis-trans isomerization and the kinetic parameters found
in Table 5. The activation parameters were very close to those
found in toluene by Poiblanc et al. (kcis f trans is 4.56 s-1 at 280
K in CD2Cl2);18 a solvent effect is not expected for an
intramolecular process. Above 263 K, the rate of cis/trans
isomerization is sufficiently fast to maintain equilibrium between
3c and 3t throughout proton exchange with deuterated acid.
H/D Exchange between 3 and Triflic Acid. Since triflic
acid is known to protonate 3,19 we began our study of the
Redetermination of the CH3CN pKa of 2,4-Dichloroanilinium
(7-H+). The equilibrium between 7-H+ and 4-cyanoaniline (8) was
1
observed from the averaged H NMR chemical shift of the aromatic
protons of 8 and 8-H+ in fast exchange during two experiments at
different starting concentrations; see Supporting Information for details.
Results and Discussion
Solid State Structure of CpWH(CO)2PMe3 (3). An X-ray
structure of a single crystal of 3 revealed that the solid structure
is cis, as shown in Figure 1. The hydrogen atoms were located
and refined; the apparent W-H distance was 1.30(8) Å. From
the single-crystal structure, a predicted powder pattern was
(29) Sheldrick, G. M. SHELXTL, An Integrated System for Solving,
Refining and Displaying Crystal Structures from Diffraction Data; University
of Go¨ttingen, Go¨ttingen, Federal Republic of Germany, 1981.
(30) Jaguar v. 4.0; Schro¨dinger, Inc.: Portland, OR, 1999.

Protonation of CpW(CO)2(PMe3)H
J. Am. Chem. Soc., Vol. 122, No. 49, 2000 12239
Figure 4. Plot of the average chemical shift of PhNH2 and
[PhNH3‚(OEt2)1-2][B(Arf)4] (δ) versus mole fraction (X) of PhNH2. A
linear plot would indicate that homoconjugation is negligible.
Effect of Added Aniline on Equilibrium between 3 and
Anilinium. When [PhNH3‚(OEt2)1-2][B(Arf)4] was added to 3
in the presence of 1 equiv of PhNH2, the formation of
[CpWH2(CO)2PMe3][B(Arf)4] (Arf ) 3,5-C6H3(CF3)2) (4b) was
suppressed. From this point we will distinguish between 4a,
Figure 3. 1H NMR of hydride region of 3 (0.019 M) broadened
by exchange with [PhNH3‚(OEt2)1-2][B(Arf)4] (0.0084 M) in CD2Cl2
at 263 K. Signals marked with * and [ are due to 3t and 3c,
respectively.
the BF4 salt of the cation 4, and 4b, the B(Arf)4 salt of 4.
The trans hydride signal still showed significant broadening
relative to the signal of the cis isomer. However, because our
PhNH3+/Et2O experiments had shown that ether was not an
innocent spectator, it was impossible to deduce meaningful
quantitative data from these exchange studies.
-
-
protonation of 3 by examining the rate of isotopic exchange
between 3 and TfOD. Preliminary NMR experiments in CD2Cl2
at 223 K showed that proton exchange was faster than addition
of D+ (or H+) to the metal. However, both reactions proved to
be too fast to quantify even at that temperature.
Proton Exchange between 3 and [PhNH3‚(OEt2)1-2]-
[B(Arf)4]. To slow both protonation reactions, and ideally to
quantify their rates, the 1H NMR spectrum of 3 was examined
in the presence of [PhNH3‚(OEt2)1-2][B(Arf)4] [Arf ) 3,5-C6H3-
(CF3)2] (pKa ) 10.5 in acetonitrile31) in CD2Cl2. (The anion
was chosen because it enhances solubility in CD2Cl2,22 and the
ethers cocrystallized with the PhNH3+ salt.) This acid protonated
3 to a small extent (no signal attributable to a dihydrogen
complex was seen), and it gave selectiVe broadening of the
hydride resonance of the trans hydride 3t by ∼7.5 Hz at 263 K
(Figure 3).
Deprotonation of [CpWH2(CO)2PMe3][B(Arf)4] (4b) by
Aniline. To eliminate ether, 3 was generated by incomplete
deprotonation of [CpWH2(CO)2PMe3][B(Arf)4] (4b) by aniline
(eq 7). Separate signals were observed for 3, 4b, and the acidic
hydrogens of the solution. The hydride resonance of 3t was
broadened more (about 60 Hz) than its PMe3 resonance (2 Hz)
at 0 °C. The resulting first-order rate constants (6.5(7) s-1 for
metal protonation, 180(30) s-1 for exchange) show that the
hydride ligand is protonated more rapidly than the tungsten
(conversion of 3 to 4b will broaden both signals, but exchange
will only broaden the hydride resonance).
This broadening is surely the result of selectiVe protonation
of the hydride ligand of 3t. There is no plausible reason for
protonation at the metal to be faster for 3t than for 3c, while
the hydride ligand of 3t is known to be more reactive toward
electrophiles than the hydride ligand of 3c.4a,32
PhNH3+ is capable of protonating 3 only because of its strong
interaction with its conjugate base, PhNH2. The overall reaction
is thus the combination of eqs 6 and 7 or eq 8, with equilibrium
constant K2 (eq 9).
Evidence of Aniline/Anilinium Homoconjugate Pair For-
+
mation. To check for interactions between PhNH2 and PhNH3
in the preceding experiment, we plotted the average chemical
shift of aniline and [PhNH3‚(OEt2)1-2][B(Arf)4] versus aniline
mole fraction in CD2Cl2 (Figure 4). A linear relationship would
+
have indicated that no species other than PhNH2 and PhNH3
were present. The nonlinear relationship observed in Figure 4
suggests extensive homoconjugate pair formation (eq 6). The
[3][PhNH2‚‚‚H+‚‚‚H2NPh]
K2 ) K1KH )
[4b][PhNH2]2
[3]2
)
(9)
[4b]{[PhNH2]0 - 2[3]}2
addition of ether varied the chemical shift of [PhNH3‚(OEt2)1-2]-
[B(Arf)4], implying that there are additional interactions between
PhNH3 and Et2O.
Note that [PhNH2‚‚‚H+‚‚‚H2NPh] ) [3] and [PhNH2] )
+
(31) (a) Edidin, R. T.; Sullivan, J. M.; Norton, J. R. J. Am. Chem. Soc
1987, 109, 3945-3953. (b) Coetzee, J. F. Prog. Phys. Org. Chem. 1967, 4,
45.
(32) For a discussion of the increased kinetic hydricity of trans and cis
hydrides, see: Cheng, T. Y.; Brunschwig, B. S.; Bullock, R. M. J. Am.
Chem. Soc. 1998, 120, 13121-13137.

12240 J. Am. Chem. Soc., Vol. 122, No. 49, 2000
Table 6. Kinetics of Proton Exchange between 3 and 5-d1
Papish et al.
Table 7. Experimental Vibrational Modes of 3c, 3c-d1, 5, and 5-d1
kf ex
10.7(5)
kr ex
11.2(6)
ν
WH(3)
ν
WD(3-d1)
(cm-1
1322a
472a
381a
ν
NH(5)
ν
ND(5-d1)
(cm-1
2320b
1016c
1062c
(cm-1
)
)
(cm-1
)
)
∆Hq (kcal/mol)
∆Sq (eu)
stretching
bending
bending
1850a
539a
456a
3117b
1482c
1183c
-35(2)
-30(2)
∆Gq263K (kcal/mol)
19.86(4)
19.01(4)
k
263K (M-1 s-1
)
1.8(1) × 10-4
8.9(8) × 10-4
a From the Raman spectra of crystalline 3c and 3c-d1. b From the
IR spectra of 5 and 5-d1 in CH2Cl2 in a KBr cell. c From the IR spectra
of 4-tert-butyl-N,N-dimethylanilinium chloride and 4-tert-butyl-N,N-
dimethylanilinium-N-d1 chloride in CH3CN in a KBr cell (the chloride
salts were not soluble in CH2Cl2). These bending modes could not be
located in 5 and 5-d1 due to strong B-F stretches from 1150 to 950
Scheme 3
cm-1
.
3c. While IR spectra of 3 in solution would thus give the
vibrational frequencies of 3t along with those of 3c, the Raman
spectrum of crystalline 3 (shown above to be exclusively 3c)
gives the vibrational frequencies of 3c only. We have thus given
only the stretching and bending modes of the H and D ligands
of 3c and 3c-d1 in Table 7sclearly identifiable from the Raman
spectra. The frequencies of the normal modes of 3t are probably
similar to those of 3c; their carbonyl stretches differ by less
[PhNH2]0 - 2[3]. A plot of ([PhNH2]0 - 2[3])2 versus ([3]2/
[4b]) is linear (Figure S-1, Supporting Information) with slope
K2 ) K1KH ) 88.3(4) M-1. Unfortunately, KH is unknown in
CD2Cl2, so it is not possible to quantify K1 from this experiment
or to estimate the value of pKa(4b) in any solvent.
Homoconjugate pair formation also makes it impossible to
obtain from the line-broadening experiments a meaningful
second-order rate constant for either reaction (exchange or
protonation at the metal) of 3t. The nature and concentration
of the effective acids are uncertain.
than 10 cm-1 18
.
A theoretical calculation of the normal modes for 5 and 5-d1
gas-phase cations showed that we should expect one stretch and
two bends. The condensed-phase spectra showed additional
peaks, so the Teller-Redlich product rule and the frequencies
of the calculated bending modes (see Supporting Information)
were used to select the experimental frequencies given in Table
7. These were used along with the vibrational frequencies of
3c and 3c-d1 to estimate an EIE value.
1H NMR Spectrum of 3 in the Presence of 4-tert-Butyl-
N,N-dimethylanilinium Tetrafluoroborate (5). To eliminate
homoconjugate pair formation, we then tried a dimethyl-
substituted anilinium salt. A tert-butyl substituent enhanced
solubility in CD2Cl2 so that tetrafluoroborate could be used as
a counteranion and a smaller mass of acid was required for the
1
There is precedent35 for using only the W-H(D) and N-H(D)
stretching frequencies to estimate Keq (eq 10). The N-H, N-D,
W-H, and W-D frequencies, as shown in Table 3, were 3117,
2320, 1850, and 1322 cm-1, respectively, suggesting a Keq of
0.50 at 283 K.
concentrations desired. The H NMR spectrum of 3 was little
affected by the presence of varying amounts of 5 at temperatures
from 300 to 223 K, although slight broadening of the hydride
peaks (<1 Hz) was observed. The line width and chemical shift
changes were much smaller than those observed when there is
hydrogen bonding between hydride ligands and acids.33 Fur-
thermore, similar line broadening and chemical shift changes
were observed in the presence of 4-tert-butyl-N,N,N-trimethy-
lanilinium tetrafluoroborate, which has no acidic protons; the
slight effect of 5 on the 1H NMR spectrum of 3 thus arose from
a nonspecific ionic interaction. No differential broadening by 5
of the hydride resonance of 3t relative to that of 3c was
observed. As subsequent experiments showed, H/D exchange
between 3 and 5 is too slow to cause line broadening.
Keq ≈ ZPE(stretching modes)
) exp{-(hc/2kBT)(νWH - νWD - νNH + νND)}
(10)
Of course, eq 10 does not consider all normal modes of the
reactants and products. A more complete description of equi-
librium isotope effects (EIEs) can be found in eqs 11-13,34,36
which consider all isotopically sensitive normal modes.
H/D Exchange Between 3 and 5. Table 6 summarizes the
results of H NMR studies in CD2Cl2 of the isotope exchange
EIE ) kf ex/kr ex ) Keq ) MMI × EXC × ZPE
1
≈ EXC × ZPE
(11)
(12)
in Scheme 3; as expected, no metal protonation was observed.
Seven experiments were done (Table 1), starting from both sides
of the equilibrium. The activation parameters are summarized
in Table 6; ∆Hq is ∼11 kcal/mol; ∆Sq is ∼-30 eu, as expected
for a bimolecular process; ∆Gq is ∼19 kcal/mol at 263 K. Values
of Keq were calculated from the ratio of kf ex/kr ex as well as from
the final integrated intensities. Keq was consistently found to
be 0.19(2), a surprisingly large equilibrium isotope effect (EIE).
Calculation from their Vibrational Frequencies of the EIE
To Be Expected for H/D Exchange between 3 and 5. While
equilibrium isotope effects can be calculated from a knowledge
of the vibrational frequencies of all species involved,34 such a
calculation is complicated in the case of the equilibrium in
Scheme 3 by the fact that 3 in solution is a mixture of 3t and
WH/kBT
NH/kBT
ND/kBT
3Nr - 6
3Np - 6
1 - e-hcν
1 - e-hcν
1 - e-hcν
i
j
EXC )
∏
∏
WD/kBT
1 - e-hcν
i
j
/
i
j
WH/kBT
WD/kBT
NH/kBT
3Nr - 6
3Np - 6
ehcν
ehcν
i
j
ZPE )
(13)
∏
∏
ND/kBT
ehcν
ehcν
i
j
/
i
j
The EIE is the product of three factors: the rotational and
translational factor, MMI, is close to unity and can be ignored;
the EXC factor contains contributions from vibrational excited
states and is also close to unity; the ZPE factor takes into account
(34) (a) Bender, B. R. J. Am. Chem. Soc. 1995, 117, 11239-11246. (b)
Bender, B. R.; Kubas, G. J.; Jones, L. H.; Swanson, B. I.; Eckert, J.; Capps,
K. B.; Hoff, C. D. J. Am. Chem. Soc. 1997, 119, 9179-9190. (c) Abu-
Hasanayn, F.; Krogh-Jespersen, K.; Goldman, A. S. J. Am. Chem. Soc. 1993,
115, 8019-8023. (d) Rabinovich, D.; Parkin, G. J. Am. Chem. Soc. 1993,
115, 353-354.
(33) (a) Crabtree, R. H. J. Organomet. Chem. 1998, 577, 111-115. (b)
Ayllon, J. A.; Gervaux, C.; Sabo-Etienne, S.; Chaudret, B. Organometallics
1997, 16, 2000-2002. (c) Shubina, E. S.; Belkova, N. V.; Krylov, A. N.;
Vorontsov, E. V.; Epstein, L. M.; Gusev, D. M.; Niedermann, M.; Berke,
H. J. Am. Chem. Soc. 1996, 118, 1105-1112.

Protonation of CpW(CO)2(PMe3)H
J. Am. Chem. Soc., Vol. 122, No. 49, 2000 12241
Table 8. Kinetics of Deprotonation of 4a by 6
Scheme 4
a
kr mp
kf mp
∆Hq (kcal/mol)
∆Sq (eu)
7.0(2)
-13.3(8)
10.55(2)
16.1(2)a
-13.3(8)a
19.57(2)a
∆Gq263K (kcal/mol)
k263K (M-1 s-1
)
9.5(3) × 103
2.8(1) × 10-4 a
the differences in ground-state vibrational energy between the
reactants and the products and accounts for the bulk of the EIE.34
Inclusion of the one stretching frequency and two bending
frequencies for compounds 3c, 3c-d1, 5, and 5-d1 listed in Table
7 leads to a calculated EIE of 0.18 at 283 K, in excellent
agreement with the experimentally observed EIE of 0.19. Since
the bending modes of the H(D) ligand of 3 are much lower in
frequency than those of the H(D) on the nitrogen in 5, the
stretching frequencies alone do not account for the EIE. (The
relative energy difference in bending frequencies between 3 and
5 may be atypical of other EIEs, because the bending modes of
3 appear to be lower compared to those of other hydride
complexes.)
a Extrapolated from kf mp ) Keq mp 263 Kkr mp
.
Scheme 5
Determination of the Rate Constant for 3 f 4a from the
Equilibrium Constant and the Rate Constant for 4a f 3.
As only triflic acid had proven strong enough to protonate the
W of 3 (the rate of that protonation had proven too fast to
measure) and protonation by anilinium introduced the complica-
tion of homoconjugation, we set out to determine the rate
constant for protonation by 5 of the W of 3 (kf mp) from the
be made from the aqueous pKa (5.1541) of N,N-dimethyl-
anilinium (the difference between the CH3CN pKa of anilinium
ions and their aqueous pKa is approximately 7.5).1
equilibrium constant (Keq mp) and the reverse rate constant (kr mp
)
(Scheme 4 and eq 14).
Kinetics of Deprotonation of 4a by 4-tert-Butyl-N,N-
dimethylaniline (6). The formation of 3 and 5 from 4a and 6
(rate constant kr mp) was monitored under pseudo-first-order
conditions (excess of 6) by stopped flow at low temperatures
in CH2Cl2. The reaction proved quite rapid and first order in
[6].42
The Ka values determined at room temperature (as described
in the preceding section) give (eq 15) a Keq mp in CH3CN at
300 K of 3.98 × 106. If we assume that an equilibrium constant
for proton transfer between two large bases should not vary
appreciably from CH3CN to CH2Cl2,43 and correct Keq mp 300 K
to Keq mp 263 K by assuming that the variation of ∆Gmp with
temperature is negligible, then kf mp in CH2Cl2 can be calculated
from eq 15 (see Scheme 4 and Table 8).
Keq mp ) kf mp/kr mp ) Ka(5)/Ka(4a)
(14)
Determination of the CH3CN pKa of 4a. An early estimate
placed pKa(4a) > 8.9, from the observation that 5 equiv of dry
HCl (pKa ) 8.9 in CH3CN31b) was able to protonate 3.20
However, HCl is a more effective acid than its pKa implies,
because of homoconjugate pair formation between HCl and
-
Cl-;24 Co(CO)4 is protonated by HCl although the pKa of
HCo(CO)4 is 8.3.24
In the absence of homoconjugate pair formation, the hydride
3 proved to be surprisingly difficult to protonate: it took 5 equiv
of 4-cyanoanilinium to protonate 10% of 3. The pKa of 4a was
determined by observing the protonation equilibria between 3
and three anilinium acids by IR, as shown in Table 4.
(Homoconjugate pair formation is negligible in CH3CN for such
weakly basic anilines.)37 Measurements with three different
anilinium acids placed the CH3CN pKa of 4a at 5.6(1).
The pKa of 5 in CH3CN. The equilibrium between 6 and
pyridinium triflate (CH3CN pKa ) 12.338) was observed by
NMR in CH3CN, and pKa(5) was found to be 12.2(1). The
known39 Kf[py2H+] ) 7.0 was used to correct for homoconjugate
pair formation between pyridine (py) and pyridinium (pyH+).40
The observed pKa(5) agrees with the estimate of 12.65 that can
Keq mp 263 K ) kf mp/kr mp
) exp{(300 K/263 K) ln(Keq mp 300 K)}
(15)
Comparison of the Rate of Metal Protonation with That
of Hydride Protonation. To compare the second-order rate
constant for metal protonation, kf mp, with that for hydride
protonation, khp (as defined in Scheme 5), we must calculate
the latter from kf ex. The rate of hydride protonation, khp, is
related to kf ex by a primary isotope effect on H(D)+ transfer to
W-H from the anilinium ion and by the fractional probability
F that formation of W(H-D)+ will result in detectable rear-
rangement (to W-D and N-H+).
(35) (a) Bullock, R. M.; Headford, C. E. L.; Hennessy, K. M.; Kegley,
S. E.; Norton, J. R. J. Am. Chem. Soc. 1989, 111, 3897-3908. (b)
Wolfsberg, M. Acc. Chem. Res. 1972, 5, 225. (c) Ritchie, C. D. Physical
Organic Chemistry: The Fundamental Concepts; Marcel Dekker: New
York, 1975; Chapter 8. (d) Lowry, T. H.; Richardson, K. S. Mechanism
and Theory in Organic Chemistry, 3rd ed.; Harper: New York, 1987;
p 255.
(40) The averaged aromatic chemical shift of 5 and 6 in CD3CN, observed
1
by H NMR, gave a linear relationship between δ and mole fraction of 6
(36) (a) McLennan, D. J. In Isotopes in Organic Chemistry; Buncel, E.,
Lee, E., Eds.; Elsevier: New York, 1987; Vol. 7, Chapter 6, p 395. (b)
Melander, L.; Saunders, W. H. Reaction Rates of Isotopic Molecules;
Wiley: New York, 1980; p 20.
(37) Homoconjugate pair formation is negligible in CH3CN for other
anilinium acids with pKa < 8 in CH3CN; see: Izutsu, K. Acid-Base
Dissociation Constants in Dipolar Aprotic SolVents; Blackwell Scientific
Publications: Oxford, 1990.
(X6) (see Supporting Information for more details), so homoconjugation is
negligiblesas expected in view of the methyl substituents.
(41) Serjeant, E. P.; Dempsey, B. Ionization Constants of Organic Acids
in Aqueous Solution; Pergamon: Oxford, 1979.
(42) We have assumed that the stopped-flow experimentsswhich involve
substantial concentrations of the base, [6]smeasure the rate at which a
proton is removed directly from 4a. We have not observed the saturation
kinetics (in [6]) that would have indicated deprotonation via an intermediate
dihydrogen complex.
(38) Coetzee, J. R.; Padmanabhan, G. J. Am. Chem. Soc. 1965, 87, 5005.
(39) Chantooni, M. K.; Kolthoff, I. M. J. Am. Chem. Soc. 1968, 90, 3005.
(43) Jia, G.; Morris, R. H. Inorg. Chem. 1990, 29, 581-582.

12242 J. Am. Chem. Soc., Vol. 122, No. 49, 2000
Papish et al.
If secondary isotope effects are neglected, the ratio k-2/k1
can be estimated from the zero-point energies of the N-H and
N-D bonds by assuming a linear transition state for proton
transfer (eq 16).31a,44 Keq and k-2/k1 can be used to estimate F
in eqs 17-19. At 263 K, a temperature at which both kf ex and
kr mp were measured, this gives khp ) 2.7 × 10-3 M-1 s-1 and
kf mp ) 2.8 × 10-4 M-1 s-1 from eqs 20 and 21, respectively.
On this basis, hydride protonation is 9.6 times faster than metal
protonation.
Homoconjugate interaction can be prevented by using 4-tert-
butyl-N,N-dimethylaniline (6) and 4-tert-butyl-N,N-dimethyl-
anilinium tetrafluoroborate (5) instead of PhNH2 and [PhNH3‚
(OEt2)1-2][B(Arf)4]. Isotopic exchange between CpW(CO)2-
(PMe3)H (3) and 4-tert-butyl-N,N-dimethylanilinium (5) shows
an equilibrium isotope effect (EIE) larger than expected on the
basis of the stretching frequencies. Inclusion of the bending
modes in the EIE calculation rationalizes the experimental value.
From the forward and backward rate constants for exchange,
kf ex and kr ex, the rate of hydride protonation, khp, can be
estimated.
The CH3CN pKa of 5 and of 4a, and the measured rate
constant kr mp for the deprotonation of 4a by 6 in CH2Cl2, make
it possible to estimate the rate constant kf mp for protonation at
the metal by 5. While some uncertainty is introduced by using
the CH3CN value of the proton-transfer equilibrium constant
Keq mp in CH2Cl2,45 kf mp appears to be an order of magnitude
smaller than khp, the rate constant for hydride protonation.
Confirmation that the kinetic site of protonation for CpW-
(CO)2(PMe3)H is the hydride ligand is offered by the line widths
in an equilibrium mixture of 3, 4a, and [PhNH2‚‚‚H+‚‚‚H2NPh].
The line width of 3t, and thus the first-order rate constant for
exchange, is 27.5 times larger than that for metal protonation.
k-2/k1 ) exp{(hc/2kBT)(νNH - νND)} ) 8.9
Keq ) k1k2/(k-1k-2) ) 0.19
(16)
(17)
F ) k2/(k-1 + k2)
(18)
1/F ) 1 + k-1/k2 ) 1 + (1/(Keqk-2/k1)) ) 1.6 (19)
khp ) kf ex(1/F)(kH/kD) ) (1.8 × 10-4 M-1 s-1)(1.6)(8.9)
) 2.7 × 10-3 M-1 s-1
(20)
kf mp ) kr mpKeq mp ) (9500 M-1 s-1)(3.0 × 10-8)
Acknowledgment. The authors thank the NSF for financial
support (Grant CHE-99-7446) and Dr. B. R. Bender, Dr. G.
Kaminsky, Dr. B. Bridgewater, Dr. C. J. Harlan, G. P. Abramo,
J. P. Roth, A. Hanley, and Profs. L. Brammer, R. Friesner, and
A. K. Rappe´ for helpful discussions and assistance with
experiments.
) 2.8 × 10-4 M-1 s-1
(21)
Conclusion
In CH3CN, the protonation of CpW(CO)2(PMe3)H (3) is
difficult (pKa ) 5.6), with 5 equiv of 4-cyanoanilinium capable
of generating only 10% of the dihydride cation (4a). In CD2Cl2
the situation is quite different, with [PhNH3‚(OEt2)1-2][B(Arf)4]
capable of protonating some 3 because of the strong homocon-
jugate interaction between the resulting PhNH2 and additional
PhNH3+. The change in the extent of protonation as the solvent
is varied emphasizes the extent to which protonation equilibria
in nonionizing solvents are affected by secondary interactions
among the acids and bases present.
Supporting Information Available: Experimental details
from the determination of the CH3CN pKa of 5, redetermination
of the CH3CN pKa of 2,4-dichloroanilinium (7-H+), assignment
of the isotopically dependent normal modes of 5 and 5-d1, linear
plot to determine K2 for deprotonation of [CpWH2(CO)2PMe3]-
[B(Arf)4] (4b) by aniline, and structural data for CpWH-
(CO)2PMe3 (3c) (PDF). This material is available free of charge
via the Internet at http://pubs.acs.org.
JA002395S
(44) Cheng, T. Y.; Bullock, R. M. J. Am. Chem. Soc. 1999, 121, 3150-
3155.
(45) If Keq mp is in error by 150%, hydride protonation is still faster by
at least a factor of 3.