Proton Transfer to Nickel-Thiolate Complexes. 1. Protonation of [Ni(SC 6H4R-4)2(Ph2PCH2CH 2PPh2)] (R = Me, MeO, H, Cl, or NO2)

Article abstract of DOI:10.1021/ic030322e

The kinetics of the equilibrium reaction between [Ni(SC₆H ₄R-4)₂(dppe)] (R= MeO, Me, H, Cl, or NO₂; dppe = Ph₂PCH₂CH₂PPh₂) and mixtures of [lutH]⁺ and lut (lut = 2,6-dimethylpyridine) in MeCN to form [Ni(SHC₆H₄R-4)-(SC₆H ₄R-4)(dppe)]⁺ have been studied using stopped-flow spectrophotometry. The kinetics for the reactions with R = MeO, Me, H, or Cl are consistent with a single-step equilibrium reaction. Investigation of the temperature dependence of the reactions shows that AG? = 13.6 ± 0.3 kcal mol^-1 for all the derivatives but the values of ΔH? and ΔS? vary with R (R = MeO, ΔH? = 8.5 kcal mol^-1, ΔS? = -16 cal K^-1 mol ^-1; R = Me, ΔH? = 10.8 kcal mol^-1, ΔS? = -9.5 cal K^-1 mol^-1; R = Cl, ΔH? = 23.7 kcal mol^-1, ΔS? = +33 cal K ^-1 mol^-1). With [Ni(SC₆H₄-NO ₂-4)₂(dppe)] a more complicated rate law is observed consistent with a mechanism in which initial hydrogen-bonding of [lutH] ⁺ to the complex precedes intramolecular proton transfer. It seems likely that all the derivatives operate by this mechanism, but only with R = NO₂ (the most electron-withdrawing substituent) does the intramolecular proton transfer step become sufficiently slow to result in the change in kinetics. Studies with [lutD]⁺ show that the rates of proton transfer to [Ni(SC₆H₄R-4)₂(dppe)] (R = Me or Cl) are associated with negligible kinetic isotope effect. The possible reasons for this are discussed. The rates of proton transfer to [Ni(SC ₆H₄R-4)₂(dppe)] vary with the 4-R-substituent, and the Hammett plot is markedly nonlinear. This unusual behavior is attributable to the electronic influence of R which affects the electron density at the sulfur.

Full text of DOI:10.1021/ic030322e

Inorg. Chem. 2004, 43, 3098−3105
Proton Transfer to Nickel−Thiolate Complexes. 1. Protonation of
[Ni(SC H R-4)2(Ph2PCH CH PPh2)] (R ) Me, MeO, H, Cl, or NO )
6 4
2
2
2
Valerie Autissier, William Clegg, Ross W. Harrington, and Richard A. Henderson*
Chemistry, School of Natural Sciences, UniVersity of Newcastle,
Newcastle upon Tyne, NE1 7RU U.K.
Received November 14, 2003
The kinetics of the equilibrium reaction between [Ni(SC H R-4)2(dppe)] (R) MeO, Me, H, Cl, or NO ; dppe )
6
4
2
Ph2PCH CH PPh2) and mixtures of [lutH]+ and lut (lut ) 2,6-dimethylpyridine) in MeCN to form [Ni(SHC H R-4)-
2
2
6 4
(SC H R-4)(dppe)]+ have been studied using stopped-flow spectrophotometry. The kinetics for the reactions with
6
4
R ) MeO, Me, H, or Cl are consistent with a single-step equilibrium reaction. Investigation of the temperature
q
q
dependence of the reactions shows that ∆G ) 13.6 ± 0.3 kcal mol-1 for all the derivatives but the values of ∆H
and ∆S vary with R (R ) MeO, ∆H ) 8.5 kcal mol-1, ∆S ) −16 cal K-1 mol-1; R ) Me, ∆H ) 10.8 kcal
q
q
q
q
mol-1, ∆S ) −9.5 cal K-1 mol-1; R ) Cl, ∆H ) 23.7 kcal mol-1, ∆S ) +33 cal K-1 mol-1). With [Ni(SC H -
q
q
q
6
4
NO -4)2(dppe)] a more complicated rate law is observed consistent with a mechanism in which initial hydrogen-
2
bonding of [lutH]+ to the complex precedes intramolecular proton transfer. It seems likely that all the derivatives
operate by this mechanism, but only with R ) NO (the most electron-withdrawing substituent) does the intramolecular
2
proton transfer step become sufficiently slow to result in the change in kinetics. Studies with [lutD]+ show that the
rates of proton transfer to [Ni(SC H R-4)2(dppe)] (R ) Me or Cl) are associated with negligible kinetic isotope
6
4
effect. The possible reasons for this are discussed. The rates of proton transfer to [Ni(SC H R-4)2(dppe)] vary with
6
4
the 4-R-substituent, and the Hammett plot is markedly nonlinear. This unusual behavior is attributable to the electronic
influence of R which affects the electron density at the sulfur.
Introduction
ally complicated, because protonation can occur at more
than one site in these complexes: either metal or ligand.5,6
In many complexes rapid protonation at one site (kinetically
favored protonation site) is followed by intra- or intermo-
lecular rearrangements in which the proton “moves” to
another site (thermodynamically favored protonation site) to
give the product. Much of this chemistry has been defined
with complexes containing metal-carbon bonds, where the
rates of protonation of both the metal and carbon sites are
appreciably slower than the diffusion-controlled limit (kdiff
) 1 × 1010 dm3 mol-1 s-1).7 It is generally accepted that
proton transfer at carbon sites is slow because of the change
in hybridization of the carbon, and proton transfer at metal
sites is slow because of the changes in the coordination
geometry and the diffuse nature of the electron density at a
metal (i.e., no stereochemical lone pair of electrons).8
Understanding the factors which control the rates and
mechanisms of protonation of metal complexes is funda-
mental in defining the elementary reactions both of certain
metalloenzymes, including nitrogenases and hydrogenases,1,2
and of certain industrially and economically important
catalysts, such as those involved in the isomerization or
hydrocyanation of alkenes and alkynes.3,4 Previous kinetic
studies have revealed that apparently simple protonation
reactions of transition metal complexes can be mechanistic-
* To whom correspondence should be addressed. E-mail: r.a.henderson@
ncl.ac.uk.
(1) (a) Henderson, R. A. Nitrogen Fixation at the Millenium; Leigh, G.
J., Ed.; Elsevier: Amsterdam, 2002; Chapter 9. (b) Burgess, B. K.;
Lowe, D. J. Chem. ReV. 1996, 96, 2983.
(2) (a) Henderson, R. A. Recent AdVances in Hydride Chemistry;
Peruzzini, M., Poli, R., Eds.; Elsevier: Amsterdam, 2001; Chapter
16. (b) Cammack, R.; Vliet, P. Bioinorganic Catalysis, 2nd ed.;
Reedijk, J., Bouwmann, E., Eds.; Marcel Dekker: New York, 1999;
Chapter 9. (c) Henderson, R. A. J. Chem. Res., Synop. 2002, 407.
(3) Masters, C. Homogeneous Transition Metal Chemistry; Chapman and
Hall: London, 1981; pp 135-158.
(5) (a) Kramarz, K. W.; Norton, J. R. Prog. Inorg. Chem. 1994, 42, 1.
(b) Kristjansdottir, S. S.; Norton, J. R. Transition Metal Hydrides:
Recent AdVances in Theory and Experiment; Dedieu, A., Ed.; VCH:
New York, 1992; Chapter 9.
(6) Henderson, R. A. Angew. Chem., Int. Ed. Engl. 1996, 35, 946.
(7) Eigen, M. Angew. Chem., Int. Ed. Engl. 1964, 3, 1.
(4) Jolly, P. W.; Wilke, G. Organic Chemistry of Nickel; Academic
Press: New York, 1975; Vol. 2, Chapter 5.
3098 Inorganic Chemistry, Vol. 43, No. 10, 2004
10.1021/ic030322e CCC: $27.50 © 2004 American Chemical Society
Published on Web 04/20/2004

Proton Transfer to Nickel-Thiolate Complexes
Table 1. Analysis and Spectroscopic Characteristics of
[Ni(SC6H4R-4)2(dppe)]+ (R ) NO2, Cl, H, Me, or MeO; dppe )
Ph2PCH2CH2PPh2)
It might be anticipated that proton transfer to transition
metal complexes containing ligands with group 16 or 17
donor atoms would be less complicated than protonation of
organometallic complexes. In general, protonation of stereo-
chemical lone pairs of electrons on group 16 or 17 atoms in
ligands is diffusion-controlled, and is much faster than the
rate of protonation of the metal. In addition, group 16 or 17
donor atoms are often markedly more basic than the metal.
Thus, protonation of such ligands is favored (both kinetically
and thermodynamically) over protonation of the metal.
Using the few general principles outlined above, it is
possible to predict the rates of proton transfer in a variety
of different metal complexes. However, some exceptions to
the general patterns are emerging. We have been particularly
interested in the rates of proton transfer to sulfur sites in
compounds which mimic active sites of metalloenzymes.
Thus, we have established that protonation of synthetic
Fe-S-based clusters which contain dimeric {MFe(µ-S)2}2+
or cuboidal {MFe3(µ3-S)4}n+ (M ) Fe, Mo or W) cores is
an entirely general phenomenon.9 Protonation has been
attributed to reactions of the µn-S sites, and the pKa’s of the
protonated clusters are essentially independent of the topol-
ogy, charge, nuclearity, or terminal ligands on the clusters
(pKa ) 18.4 ( 0.5).10 More recently, the rates of protonation
of these clusters have been shown to be slower than the
diffusion-controlled limit and are markedly dependent on the
metal composition of the cluster.11
Thiolate sulfur (cysteinate) is a ligation common in
bioinorganic chemistry, and protonation of such ligands is a
reaction which is invoked in the reactions of many metal-
loenzymes, most notably nitrogenases,12 hydrogenases,13 and
Fe-S-based hydrolases such as aconitase.14 Little is known
about the protonation of simple coordinated thiolates. It is
generally assumed that the reactions are diffusion-controlled.
However, some recent studies by us have indicated that
proton transfer reactions involving simple transition metal
thiolates are, at least in certain cases, slow.15
In this paper, we report studies on the protonation reactions
of [Ni(SC6H4R-4)2(dppe)] (R ) MeO, Me, H, Cl, or NO2;
dppe ) Ph2PCH2CH2PPh2) with [lutH]+ (lut ) 2,6-dimeth-
ylpyridine) in MeCN and show that protonation of the
thiolate sulfur is typified by the following observations: (i)
proton transfer is slow; (ii) the metal effectively “levels” the
acidity of the coordinated thiol; and (iii) with most derivatives
elemental analysis/%a
H
NMR spectroscopy
R
C
N
1Hb,c
31Pd
NO2
59.80
(59.96)
61.14
(61.29)
66.63
(67.55)
68.15
(68.27)
65.35
4.19
(4.18)
4.18
(4.30)
4.83
(5.04)
5.23
(5.40)
5.16
3.68
(3.67)
56.95
Cl
57.38
56.50
56.40
55.70
H
Me
MeO
2.1 (Me)
5.7 (MeO)
(65.31)
(5.17)
a Calculated values shown in parentheses. b Chemical shifts relative to
TMS. c Peaks due to dppe ligands are present in all spectra at δ 7.0-8.0
(multiplets, Ph groups) and δ 2.2-3.0 (doublet, CH2); peaks due to
4-RC6H4S are present in all spectra at δ 6.0-7.0 (multiplets). d Chemical
shifts relative to H3PO4, singlets.
the kinetics are simple, but when R ) NO2 a more com-
plicated rate law is observed, consistent with a mechanism
involving rapid formation of a hydrogen-bonded precursor
prior to rate-limiting intramolecular proton transfer. In the
following paper16 we elaborate further on the mechanistic
aspects of proton transfer to nickel thiolate complexes by
studies on [Ni(SC6H4R-4)(triphos)]+ {triphos ) PhP(CH2-
CH2PPh2)2}. In this class of complexes, the kinetics of the
protonation reactions of all derivatives are consistent with a
mechanism in which a hydrogen-bonded adduct between the
acid and complex can be detected prior to the rate-limiting
intramolecular proton transfer to form [Ni(SHC6H4R-4)-
(triphos)]2+. Studies on [Ni(SC6H4R-4)(triphos)]+ allow the
kinetic characterization of intramolecular proton transfer
reactions.
Experimental Section
All preparations and manipulations were routinely performed
under an atmosphere of dinitrogen using Schlenk or syringe tech-
niques as appropriate. All solvents were dried and freshly distilled
from the appropriate drying agent immediately prior to use.
The thiols 4-RC6H4SH (R ) NO2, Cl, Me, or MeO), lut (lut )
2,6-dimethylpyridine), and Ph2PCH2CH2PPh2 (dppe) were pur-
chased from Aldrich and used as received. NaSC6H4R-4,17 [lutH]-
BPh4,18 and [lutD]BPh4 were prepared by literature methods.
[NiCl2(dppe)] was prepared by the method reported earlier.19
Preparation of [Ni(SC6H4R-4)2(dppe)] (R ) NO2, Cl, H, Me,
or MeO). The complexes in the series [Ni(SC6H4R-4)2(dppe)] were
all prepared by the same method.20 The complexes were character-
ized by elemental and spectroscopic analysis as shown in Table 1,
and (for R ) H, Cl, or Me) by X-ray crystallography. A typical
preparation is described below for [Ni(SC6H4Me-4)2(dppe)].
To a suspension of [NiCl2(dppe)] (1 g, 1.95 mmol) in ethanol
(ca. 30 mL) was added an excess of NaSC6H4R-4 (6.0 mmol). The
(8) Bell, R. P. The Proton in Chemistry, 2nd ed.; Chapman and Hall:
London, 1973; Chapter 10 and references therein.
(9) (a) Henderson, R. A.; Oglieve, K. E. Chem. Commun. 1994, 377. (b)
Almeida, V. R.; Gormal, C. A.; Gro¨nberg, K. L. C.; Henderson, R.
A.; Oglieve, K. E.; Smith, B. E. Inorg. Chim. Acta 1999, 291, 212.
(c) Henderson, R. A.; Oglieve, K. E. J. Chem. Soc., Dalton Trans.
1998, 1731.
(10) Gro¨nberg, K. L. C.; Henderson, R. A. J. Chem. Soc., Dalton Trans.
1996, 3667.
(11) (a) Henderson, R. A.; Oglieve, K. E. J. Chem. Soc., Dalton Trans.
1999, 3927. (b) Henderson, R. A.; Dunford, A. J. Chem. Commun.
2002, 360. (c) Bell, J.; Dunford, A. J.; Hollis, E.; Henderson, R. A.
Angew. Chem. 2003, 42, 1149.
(12) Burgess, B. K.; Lowe, D. J. Chem. ReV. 1996, 96, 2983.
(13) Peters, J. W.; Lanzilotta, W. N.; Lemon, B. J.; Seefeldt, L. C. Science
1998, 282, 1853
(16) Accompanying paper: Autissier, V.; Zarza, P. M.; Petrou, A.;
Henderson, R. A.; Harrington, R. W.; Clegg, W. Inorg. Chem. 2004,
43, 3106-3115.
(17) Hagen, K. S.; Holm, R. H. Inorg. Chem. 1984, 23, 418.
(18) Gro¨nberg, K. L. C.; Henderson, R. A.; Oglieve, K. E. J. Chem. Soc.,
Dalton Trans. 1998, 3093.
(19) Busby, R.; Hursthouse, M. B.; Jarrett, P. S.; Lehmann, C. W.; Abdul
Malik, K. M.; Phillips, C. J. Chem. Soc., Dalton Trans. 1993, 3767.
(20) Hayter, R. G.; Humiec, F. S. J. Inorg. Nucl. Chem. 1964, 26, 807.
(14) Flint, D. H.; Allen, R. M. Chem. ReV. 1996, 96, 2315.
(15) Clegg, W.; Henderson, R. A. Inorg. Chem. 2002, 41, 1128.
Inorganic Chemistry, Vol. 43, No. 10, 2004 3099

Autissier et al.
Table 2. Crystal Data and Structure Refinement Parameters for
[Ni(SC6H4R-4)2(dppe)] (R ) H, Me, or Cl; dppe ) Ph2PCH2CH2PPh2)
Table 3. Summary of Bond Lengths and Bond Angles in
[Ni(SC6H4R-4)2(dppe)]+ (R ) Me, H, or Cl; dppe ) Ph2PCH2CH2PPh2)
bond length
bond angle
[Ni(SC6H5)2-
(dppe)]
[Ni(SC6H4-
Me-4)2(dppe)] (dppe)]‚0.25CH2Cl2
[Ni(SC6H4Cl-4)2-
R
Ni-P
Ni-S
P-Ni-P
P-Ni-S S-Ni-S Ni-S-C
formula
C38H34NiP2S2 C40H38NiP2S2
C38H32Cl2NiP2S2‚
0.25CH2Cl2
765.53
triclinic
P1h
150
13.6352(7)
16.5127(8)
18.9762(10)
65.255(1)
75.748(1)
71.427(1)
3646.4(3)
4
26364
12693, 0.0360
838
0.0482
0.1322
Ha
2.1624(7) 2.2086(7) 87.05(4)
86.24(2)
169.99(3)
84.70(3)
2.1933(9) 90.77(4) 113.16(10)
169.82(4)
85.61(4)
89.49(4)
168.36(5)
169.72(4)
87.29(4)
101.33(4) 111.81(9)
fw
675.42
703.47
Me
Clb
2.1583(10) 2.2486(9) 86.72(4)
98.84(3) 99.47(11)
cryst syst
space group
T, K
a, Å
b, Å
monoclinic
C2/c
monoclinic
P21/n
169.02(4)
2.1891(9)
150
150
2.1556(11) 2.1892(11) 87.26(4)
2.1710(11) 2.2389(11)
98.91(4) 100.50(13)
111.24(13)
14.3764(16)
15.9241(12)
14.1357(13)
11.4276(6)
17.6539(9)
17.2914(8)
c, Å
2.1537(11) 2.1946(11) 87.17(4)
2.1601(11) 2.2244(11)
101.10(4) 109.19(13)
111.48(14)
R, deg
â, deg
γ, deg
V, Å3
87.87(4)
164.35(5)
164.27(5)
92.756(2)
100.337(1)
3232.4(5)
4
3431.8(3)
4
Z
a The molecule has crystallographic C2 symmetry. b Two independent
reflns measured
unique data, Rint 4737
4737
24602
molecules in the asymmetric unit.
6044, 0.0635
408
params
196
R (F, F2 > 2σ)
Rw (F2, all data)
GOF on F2
max, min
0.0456
0.1287
1.059
0.0388
0.0890
1.091
1.063
1.23, -0.94
0.76, -0.48
0.33, -0.35
electron
density (e/Å3)
solution changed rapidly from orange to dark red and was stirred
overnight to become homogeneous. The next day, half of the solvent
was removed in vacuo, and the dark red microcrystalline solid was
removed by filtration, washed with a large amount of ethanol, and
dried in vacuo. Crystals of the R ) Me, H, and Cl derivatives
suitable for X-ray diffraction were obtained by recrystallization from
CH2Cl2/EtOH at room temperature.
X-ray Crystallography. All data were collected on a Bruker
SMART CCD area diffractometer, using Mo KR radiation (λ )
0.71073 Å), by the ω-scan method.21 Crystal data and other
experimental information are given in Table 2, with further details
in the Supporting Information. Semiempirical absorption corrections
were applied in all cases, on the basis of repeated and symmetry-
equivalent reflections.21 The structures were solved by direct
methods and refined by full-matrix least-squares on all unique F2
values.22 Anisotropic displacement parameters were assigned to all
the non-hydrogen atoms. Hydrogen atoms were placed in idealized
positions and allowed to ride on their respective parent atoms. The
molecule of the unsubstituted compound (R ) H) lies on a
crystallographic 2-fold rotation axis, passing through the Ni atom,
so the asymmetric unit is half of a molecule. This structure was
found to be twinned; intensities were integrated for both twin
components and were retained as independent data without merging
of symmetry-equivalent reflections. The ratio of the two components
was refined to 0.539:0.461(2). The chloro derivative has two
independent molecules in the asymmetric unit, together with a
partially occupied site for a solvent dichloromethane molecule,
which has been assigned half occupancy on the basis of refinement
of the displacement parameters. The largest difference map features
lie close to this solvent molecule, which is probably somewhat
disordered. The two molecules of the complex show only minor
differences in geometry.
Figure 1. Absorbance-time curve for the reaction of [Ni(SC6H4Me-4)2-
(dppe)] (0.2 mmol dm-3) with [lutH]+ (12.5 mmol dm-3) and lut (1.0 mmol
dm-3) in MeCN at 25.0 °C, measured at λ ) 400 nm.
Kinetic Studies. All kinetic studies were performed using an
Applied Photophysics SX.18MV stopped-flow spectrophotometer,
modified to handle air-sensitive solutions. The temperature was
maintained at 25.0 ( 0.1 °C using a Grant LT D6G thermostated
recirculating pump.
All solutions were prepared under an atmosphere of dinitrogen
and transferred by gastight, all-glass syringes into the stopped-flow
spectrophotometer. Solutions containing mixtures of lut and [lutH]-
BPh4 were prepared from freshly prepared stock solutions of the
components, and used within 1 h.
Kinetics were studied under-pseudo first-order conditions with
all reagents in a large excess (>10-fold) over the concentration of
[Ni(SC6H4R-4)2(dppe)].23 Under all conditions, the absorbance-
time curve for the reactions is an excellent fit to a single exponential
for at least 4 half-lives, indicating a first-order dependence on the
concentration of complex24 (see Figure 1). The entire curve was
fitted using the Applied Photophysics computer program to obtain
the observed rate constants (kobs). The rate laws were determined
by graphical analysis as described in the Results and Discussion
section.
Selected bond lengths and angles for the three complexes are
reported in Table 3.
(21) SMART (control), SAINT (integration), GEMINI (twinning), and
SADABS (absorption correction) software; Bruker AXS Inc.: Madison,
WI, 2001.
(22) Sheldrick, G. M. SHELXTL version 6; Bruker AXS Inc.: Madison,
WI, 2001.
(23) Espenson, J. H. Chemical Kinetics and Reaction Mechanisms;
McGraw-Hill: New York, 1981; p 12.
(24) Wilkins, R. G. Kinetics and Mechanisms of Reactions of Transition
Metal Complexes; VCH: Weinheim, 1991; Chapter 1.
3100 Inorganic Chemistry, Vol. 43, No. 10, 2004

Proton Transfer to Nickel-Thiolate Complexes
Protonation of [Ni(SC6H4R-4)2(dppe)] (R ) MeO, Me,
H, or Cl). The reactions between all [Ni(SC6H4R-4)2(dppe)]
(R ) MeO, Me, H, or Cl) and [lutH]+ in the presence of
lut, in MeCN as solvent, show all the characteristics of an
equilibrium reaction. This is particularly obvious when the
reaction is monitored at a single wavelength in the stopped-
flow spectrophotometric studies. Thus, when the concentra-
tion of [lut] is kept constant, the absorbance change increases
with increasing concentration of acid. Similarly, when the
concentration of [lutH]+ is kept constant, increasing the
concentration of [lut] results in a decrease in the absorbance
change. The kinetics of the reactions are also consistent with
an equilibrium reaction.
When studied at a single wavelength using a stopped-flow
spectrophotometer, the reaction between [Ni(SC6H4R-4)2-
(dppe)] (R ) MeO, Me, H, or Cl) and a large excess of
[lutH]+ and lut (in MeCN as solvent) is typified by an
absorbance-time curve which is a single exponential. The
dependence on the concentrations of both [lutH]+ and lut
was determined graphically. A typical example is shown in
Figure 3.
A plot of kobs/[lut] against [lutH+]/[lut] is a straight line
with a small but finite intercept, with an associated depen-
dence on the concentrations of [lutH]+ and lut as shown in
eq 1, from which the rate law shown in eq 2 is derived.
Figure 2. The molecular structure of [Ni(SC6H4Me-4)2(dppe)], with 50%
probability displacement ellipsoids. H atoms are omitted for clarity.
In the kinetic studies with [lutD]+, the degree of deuterium-
labeling of the acid was established by 1H NMR spectroscopy. We
have found that even small amounts of protic impurity lead to rapid
exchange even in the solid state, presumably because of the strength
of this acid. Certainly, proton exchange in [lutH]+ (pKa ) 15.4 in
MeCN)25 is more marked than for [NHEt3]+ (pKa ) 18.46 in
MeCN).26 Although it is impossible to entirely eliminate all protic
1
kobs k1 [lutH+]
R
impurities in CD3CN, H NMR spectroscopic studies show that
R
the [lutD]BPh4 used in these studies is at least 90% deuterium-
labeled. In particular, the broad peak centered at δ 12.2 ppm in the
1H NMR spectrum of [lutH]+ is not present in the spectrum of
[lutD]+.
)
+ k-1
(1)
[lut]
[lut]
-d[Ni(SC6H4R-4)2(dppe)]
)
dt
R
{k1 [lutH+] + k-1R[lut]}[Ni(SC6H4R-4)2(dppe)] (2)
Results and Discussion
Structures of [Ni(SC6H4R-4)2(dppe)] (R ) H, Me, or
Cl). The complexes, [Ni(SC6H4R-4)2(dppe)] (R ) MeO, Me,
H, Cl, or NO2), were prepared by the literature methods19
and characterized by elemental analysis and 1H and 31P NMR
spectroscopies, together with the X-ray crystal structures for
the R ) H, Cl, and Me derivatives. The structures of the
three derivatives are essentially identical, and Figure 2 shows
that of [Ni(SC6H4Me)2(dppe)].
Equation 2 is consistent with the simple equilibrium
reaction28 shown in eq 3 involving proton transfer to and
R
from the nickel complex, where k1 is the rate constant for
R
proton transfer to sulfur and k-1 is the rate constant for
R
deprotonation of the coordinated thiol. The values of k1
R
and k-1 for R ) MeO, Me, H, or Cl are summarized in
Table 4.
The structures show that the geometry about the nickel in
[Ni(SC6H4R-4)2(dppe)] is best described as distorted square-
planar. The major bond lengths and bond angles for [Ni-
(SC6H4R-4)2(dppe)] (R ) H, Cl, or Me) are compared in
Table 3. The Ni-P bond lengths fall in the range 2.1537-
(11)-2.1891(9) Å, and the Ni-S bond lengths are 2.1892-
(11)-2.2486(9) Å. It is worth noting that the Ni-S-C angles
are somewhat variable: 111.81(9)° (R ) H); 100.50(13)-
111.48(14)° (R ) Cl), and 99.47(11)-113.16(10)° (R ) Me).
These angles, as has been noted before, are normal for
transition metal thiolates27 and indicate little strain in the
bound thiolates. The P-Ni-S, S-Ni-S, and P-Ni-P
angles are unexceptional.
When considered in more detail both the forward (proto-
nation) and back (deprotonation) reactions shown in eq 3
must occur by intimate mechanisms involving two distinct
steps. For the protonation reaction the first step must involve
approach of the acid toward the sulfur and the formation of
a precursor hydrogen-bonded adduct. The protonation is
completed in the second step in which intramolecular proton
transfer occurs from the acid to the sulfur, within the
(25) Cauquis, G.; Deronzier, A.; Srve, D.; Vieil, E. J. Electroanal. Chem.
Interfacial Electrochem. 1975, 60, 205.
(27) Orpen, A. G.; Brammer, L.; Allen, F. H.; Kennard, O.; Watson, D.
C.; Taylor, R. J. Chem. Soc., Dalton Trans. 1989, 51.
(28) Reference 23, p 45
(26) Izutsu, K. Acid-Base Dissociation Constants in Dipolar Aprotic
SolVents; Blackwell Scientific: Oxford, 1990.
Inorganic Chemistry, Vol. 43, No. 10, 2004 3101

Autissier et al.
Figure 3. Graph of kobs/[lut] against [lutH+]/[lut] for the reaction of [Ni(SC6H4OMe-4)2(dppe)] with [lutH]+ and lut in MeCN at 25.0 °C, measured at λ
) 400 nm. The data points correspond to [lut] ) 1-40 mmol dm-3 [lutH+] ) 12.5 mmol dm-3 (b); [lutH+] ) 25 mmol dm-3 (9) and [lutH+] ) 50 mmol
dm-3 (2). The straight line is defined by eq 2. The plot is typical of those observed for the R ) Me, H, and Cl derivatives as well.
Table 4. Summary of Elementary Rate Constants and Activation
Parameters for the Reactions of [Ni(SC6H4R-4)2(Ph2PCH2CH2PPh2)]+
with [lutH]+ and lut in MeCN at 25.0 °C
k1R/dm3
k-1R/dm3
∆Hq/kcal
∆Sq/cal
R
R
R
mol-1 s-1 mol-1 s-1 K1
pKa
mol-1
K-1 mol-1
NO2
Cl
H
Me
MeO
1980a
360
338
474
924
4000b
300
220
250
400
0.5
1.2
1.5
1.9
2.3
15.1
15.5
15.6
15.7
15.8
23.7
+33
10.8
8.5
-9.5
-16
a k1 ) K2Rk3 (R ) NO2). k-1 ) k-3 (R ) NO2).
R
R
b
R
R
hydrogen-bonded adduct. The deprotonation reaction must
involve analogous steps.
The simplicity of the kinetics observed with [Ni(SC6H4R-
4)2(dppe)] (R ) MeO, Me, H, or Cl) means that we have no
kinetic evidence for the intimate mechanism. However, for
the R ) NO2 derivative direct kinetic evidence for the
formation of a precursor adduct is obtained.
Figure 4. Graph of kobs/[lut] against [lutH+]/[lut] for the reaction of [Ni-
(SC6H4NO2-4)2(dppe)] with [lutH]+ and lut in MeCN at 25.0 °C, measured
at λ ) 400 nm. The data points correspond to [lut] ) 0.1-1.0 mmol dm-3
[lutH+] ) 12.5 mmol dm-3 ([); [lutH+] ) 25 mmol dm-3 (9) and [lutH+]
) 50 mmol dm-3 (2). The straight lines are defined by eq 4.
Protonation of [Ni(SC6H4NO2-4)2(dppe)]. The reaction
between [Ni(SC6H4NO2-4)2(dppe)] and an excess of [lutH]+
and lut also shows all the characteristics of an equilib-
rium reaction. However, the kinetics of this reaction are
more complicted than those observed for the other deriva-
tives. The kinetics are illustrated in Figure 4. It is important
to note that the studies with [Ni(SC6H4NO2-4)2(dppe)] were
performed using the same range of [luH]+ and lut concentra-
tions as used in the studies with all the other derivatives.
Consequently, the change in kinetics observed with [Ni-
(SC6H4NO2-4)2(dppe)] is not because of a change of reaction
conditions, but rather the electronic influence of the nitro-
substituent.
For the reaction of [Ni(SC6H4NO2-4)2(dppe)], as with all
the derivatives, a plot of kobs/[lut] against [lutH+]/[lut] is a
straight line. The difference is that the graph is only a straight
3102 Inorganic Chemistry, Vol. 43, No. 10, 2004

Proton Transfer to Nickel-Thiolate Complexes
line if the concentration of [lutH]+ is kept constant. At
different concentrations of acid, the gradient of the line is
different, such that the higher the concentraion of [lutH]+,
the smaller the gradient. Analysis of the data yields the rate
law shown in eq 4.
1980[lutH+]
-d[Ni(SC6H4NO2)2(dppe)]
dt
)
+
Figure 5. The intimate mechanism of proton transfer for the reactions of
[Ni(SC6H4R-4)2(dppe)] (R ) MeO, Me, H, Cl, or NO2), as indicated by
the kinetics for the R ) NO2 derivative.
+
{
1 + 200[lutH ]
4000[lut] [Ni(SC6H4NO2-4)2(dppe)] (4)
}
1, eq 7 simplifies to eq 8. Equation 8 is identical to that
R
R
R
R
R
shown in eq 2, with k1 ) K2 k3 and k-1 ) k-3
.
Equation 4 is consistent with a mechanism involving two
coupled equilibria29 as illustrated in Figure 5. In this mech-
anism, the initial step in the reaction involves binding of
[lutH]+ to [Ni(SC6H4NO2-4)2(dppe)] but without proton
transfer. It seems likely that this initial binding involves
hydrogen-bonding, in which the protic end of [lutH]+
hydrogen-bonds to the thiolate sulfur. It is possible that, in
addition to the hydrogen-bond, the binding of [lutH]+ to [Ni-
(SC6H4NO2-4)2(dppe)] could also involve aromatic π-π-
stacking of the phenyl groups of the phosphine with the
[lutH]+.
The mechanism shown in Figure 5 is simply an elaboration
of the single-step reaction shown in eq 3. In contrast to the
other derivatives, with [Ni(SC6H4NO2-4)2(dppe)] the forma-
tion of the initial hydrogen-bonded adduct is kinetically
distinct from the subequent intramolecular transfer of the
proton. It seems likely that the different kinetics observed
with [Ni(SC6H4NO2-4)2(dppe)] are because of the strongly
electron-withdrawing effect of the nitro-group. The nitro-
group is the most strongly electron-withdrawing 4-R-
substituent employed in this series of complexes.
R
R
kobs ) K2 k3 [lutH+] + k-3R[lut]
(8)
Basicity of Coordinated Thiolates. From the rate con-
R
R
stants k1 and k-1 shown in Table 4, we can calculate the
R
R
protonation equilibrium constants, K1 ) k1 /k-1R. Further-
more, since we know the pKa of [lutH]+ in MeCN is 15.4,25
the pKaR of each [Ni(SHC6H4R-4)(SC6H4R-4)(dppe)]+ can
be calculated. The values are presented in Table 4. It is
evident that although the pKaR values of the complexes vary
in the expected manner (i.e., lowest for R ) NO2 and highest
for R ) MeO), the value is remarkably insensitive to the
4-R-substituent. Thus, the pKaR varies only from 15.1 (R )
NO2) to 15.8 (R ) MeO). This variation is much smaller
than anticipated with the free thiols. Athough there is no
quantitative data on the strengths of 4-substituted thiols, there
is information about the corresponding phenols. Thus, for
4-MeC6H4OH (pKa ) 10.2), C6H5OH (pKa ) 9.9), and 4-O2-
NC6H4OH (pKa ) 7.1),30 the acid strengths follow the
expected order and cover more than 3 pKa units. It is
reasonable that the corresponding thiols would show an
analogous trend. It appears that, when coordinated to the
nickel site, the acidities of the thiols are effectively “leveled”.
It is worth noting that Enemark and co-workers have
observed analogous behavior in complexes of the type
[Mo(Tp)(E)(tdt)]n- {Tp ) hydrotris(3,5-dimethyl-1-pyra-
zolyl)borate; E ) O, S, or NO; tdt ) 3,4-toluenedithiolate,
n ) 0-3}.31 In all these complexes, the first ionization
potentials are very similar, indicating that the changes in the
formal oxidation state (+2 to +5) and π-effects of the
E-group are neutralized by changes to the bonding of the
ene-dithiolate ligand. The values of the contributing rate
constants shown in Table 4 indicate that changes in the value
Consideration of the mechanism shown in Figure 5 gives
the rate law shown in eq 5. The corresponding dependence
of kobs on the concentrations of [lutH]+ and lut is shown in
eq 6, from which eq 7 is readily derived. Equation 7 is the
basis of the graph shown in Figure 4. Comparison of eqs 4
and 5 gives the values of the elementary rate constants for
the reaction between [Ni(SC6H4NO2-4)2(dppe)] and [lutH]+
or lut, as shown in Table 4.
R
R
-d[Ni(SC6H4NO2)2(dppe)]
dt
K2 k3 [lutH+]
)
+
R
1 + K2 [lutH+]
k-3R[lut][Ni(SC6H4NO2)2(dppe)] (5)
R
R
of k1 are mirrored by a similar change in k-1
.
K2 k3 [lutH+]
R
R
The electronic origin of the “leveling” effect is consistent
with a compensatory effect on the basicity of the sulfur atom
in the complexes. In order to understand such a compensatory
effect we need to consider the bonding of the thiolate and
the corresponding thiol to the nickel which has been
described by others.32-34 The lone pair of electrons on the
thiolate is of the correct symmetry to interact with the t2g
d-orbitals of the metal, as shown in Figure 6. However the
t2g-orbitals in nickel are full, and a four-electron repulsion
kobs
)
+ k-3R[lut]
(6)
(7)
1 + K2 [lutH+]
R
kobs K2 k3 [lutH+]/[lut]
R
R
R
)
+ k-3
1 + K2 [lutH+]
R
[lut]
Equation 5 describes the full rate law for the mechanism
shown in Figure 5. However, this rate law describes the
condition when the hydrogen-bonded adduct accumulates.
The adduct will only attain appreciable concentrations if both
(30) Sykes, P. A Guidebook to Mechanism in Organic Chemistry, 3rd ed.;
Longmans: London, 1970; p 60. pKa values are in water.
(31) Westcott, B. L.; Gruhn, N. E.; Enemark, J. H. J. Am. Chem. Soc. 1998,
120, 3382.
R
R
k3R and k-2R are small compared to k2 . When K2 [lutH+] <
(29) Reference 24, pp 33-37.
Inorganic Chemistry, Vol. 43, No. 10, 2004 3103

Autissier et al.
First, the measured rate constants (k1 and k-1R) are com-
posite values comprising the elementary reactions of (i)
binding of the acid to form the hydrogen-bonded species
followed by (ii) intramolecular proton transfer. Second, the
Hammett σ+ parameter reflects only electronic resonance
effects but neglects electronic inductive effects. Recent
theoretical calculations36 have used the Natural Bond Order
package to calculate the electron density on the sulfur atoms
R
Figure 6. Illustration of the electronic influence of the 4-R-substituent
on the interaction between the full p-orbitals on the sulfur and the full t2g
d-orbitals on the nickel in complexes of the type [Ni(SC6H4R-4)2(dppe)].
R
of thiols and disulfides. However, the values of k1R and k-1
do not correlate well with these calculated values of the
electron density on the sulfur.
It is evident that the effect the various 4-R-substituents
have on the rates of proton transfer is not so simplistic as
the modulation of the electron density (basicity) of the sulfur
site. As we have seen above, the basicity of the site is
relatively insensitive to the substituent. As the kinetics of
the reaction between [Ni(SC6H4NO2-4)2(dppe)] and [lutH]+
reveal, the mechanism for the protonation of all [Ni(SC6H4R-
4)2(dppe)] consists of two steps: the formation of a hydro-
gen-bonded adduct followed by intramolecular proton trans-
fer. Consequently, the values of k1R and k-1R reflect the net
effect of the 4-R-substituent on both the adduct formation
and intramolecular proton transfer. It seems reasonable that
the binding constant for the formation of the hydrogen-
bonded adduct would depend on the electron density at the
sulfur and hence correlate with the pKaR. As discussed above,
the pKaR of [Ni(SHC6H4R-4)(SC6H4R-4)(dppe)]+ is rather
insensitive to the nature of the 4-R-substituent, and hence,
it seems likely that Figure 7 reflects the effect that the 4-R-
substituent has on the rate of intramolecular proton transfer.
This proposal is justified in the succeeding paper, where
studies on [Ni(SC6H4R-4)(triphos)]+ show that the rate of
intramolecular proton transfer with these complexes cor-
relates with the electron density on the sulfur.16
Figure 7. Hammett plot for the rate constants for the protonation of [Ni-
(SC6H4R-4)2(dppe)] (R ) MeO, Me, H, Cl or NO2) by [lutH]+ (k1R, 9)
and deprotonation of [Ni(SHC6H4R-4)(SC6H4R-4)(dppe)]+ by lut (k-1R, b).
The curves are a guide to the trend of the data.
results. When the lone-pair of electrons is protonated to form
the corresponding thiol, there is a reduction in the repulsion
(leading to a shortening of the Ni-S bond length), and a
weakening of the σ-donation since thiol is a neutral ligand
but thiolate is anionic (leading to a lengthening of the Ni-S
bond length). Thus, electron-donating 4-R-substituents would
be expected to increase the electron density of the sulfur in
thiolates, but this would be tempered by the added repulsion
with the t2g orbitals on nickel. In contrast, electron-withdraw-
ing 4-R-substituents diminish the electron density at the
sulfur but also permit closer interaction between the nickel
and sulfur. These electronic effects result in an effective
leveling of the pKaR of the coordinated thiol through a
modulation of the degree of covalency of the Ni-S bond.
Rates of Proton Transfer. The rates of proton transfer
are affected by the 4-R-substituent (Table 4). The influence
It is evident from Figure 7 that the same factors which
R
facilitate proton transfer to the sulfur (k1 ) also facilitate
proton transfer from the coordinated thiol (k-1R). How both
electron-releasing and electron-withdrawing 4-R-substituents
facilitate proton transfer to and from sulfur is rationalized
in terms of the bonding description shown in Figure 6.
Proton transfer to thiolate sulfur is facilitated by electron-
donating substituents because of the increased electron-
density on the sulfur (reflected in pKaR). Effectively, this is
a ground state destabilizing effect. As discussed above,
electron-withdrawing 4-R-substituents will diminish the
unfavorable interaction between the full p-orbitals on the
sulfur and the full t2g-orbitals on the nickel.37 Protonation of
a thiolate ligand has a similar effect on the Ni-S orbital
interaction. Consequently, electron-withdrawing substituents
facilitate the protonation reaction by effectively making the
nickel-thiolate orbital interaction more like that of the
nickel-thiol. Effectively, this is a transition state stabilization
effect.
R
R
of the 4-R-substituent on k1 and k-1 is illustrated in the
Hammett plot shown in Figure 7. We have used the Hammett
σ+ parameter for the correlation, since this parameter allows
for conjugation between the 4-R-substituent and the reaction
site (sulfur). Previous studies35 have indicated that this
parameter is more appropriate in the reactions of ligands on
transition metal complexes than the Hammett σ parameter.
Interpretation of the effect that the 4-R-substituent has on
the rate of proton transfer is complicated by two features.
(32) Grapperhaus, C. A.; Pothrovic, S.; Mashata, M. S. Inorg. Chem. 2002,
41, 4309.
We consider now the rate of deprotonation of the thiol
complex. Proton transfer from thiol sulfur is facilitated by
(33) Ashby, M. T.; Enemark, J. H.; Lichtenberger, D. L. Inorg. Chem. 1988,
27, 191.
(34) Grapperhaus, C. A.; Patra, A. K.; Mashata, M. S. Inorg. Chem. 2002,
41, 1039.
(35) Hussain, W.; Leigh, G. J.; Mohd Ali, H.; Pickett, C. J.; Rankin, D. A.
J. Chem. Soc., Dalton Trans. 1984, 1703.
(36) Sengar, R. S.; Nemykin, V. N.; Basu, P. New. J. Chem. 2003, 7, 1115.
(37) Sellmann, D.; Sutter, J. Acc. Chem. Res. 1997, 30, 460 and references
therein.
3104 Inorganic Chemistry, Vol. 43, No. 10, 2004

Proton Transfer to Nickel-Thiolate Complexes
electron-withdrawing substituents because of the increased
acidity of the coordinated thiol (i.e., ground state destabiliza-
tion). In complexes where the coordinated thiol contains an
electron-donating 4-R-substituent, the unfavorable orbital
interactions in the nickel-sulfur bond are enhanced as they
are in the thiolate product (i.e., transition state stabilization).
Temperature Dependence of the Reactions. The effect
of temperature on the reactions of [Ni(SC6H4R-4)2(dppe)]
(R ) MeO or NO2) with [lutH]+ have been studied.38 The
activation parameters determined from analysis of the data
are characterized by an effectively constant ∆Gq ) 13.6 (
0.3 kcal mol-1. However, the contributing values of ∆Hq
and ∆Sq vary significantly, as shown in Table 4. Thus, as
the 4-R-substituent becomes more electron-withdrawing, ∆Hq
becomes larger and ∆Sq becomes more positive (R ) MeO,
∆Hq ) 8.5 kcal mol-1, ∆Sq ) -16 cal K-1 mol-1; R ) Me,
∆Hq ) 10.8 kcal mol-1, ∆Sq ) -9.5 cal K-1 mol-1; R )
Cl, ∆Hq ) 23.7 kcal mol-1, ∆Sq ) +33 cal K-1 mol-1).
The values of ∆Hq and ∆Sq give further insight into the
intimate mechanism of proton transfer in these complexes.
The trend observed in the values of ∆Hq and ∆Sq is consistent
with a description of the transition state in which the position
of the proton between [lutH]+ and the thiolate sulfur during
the proton transfer depends on the nature of the 4-R-
substituent.
When the 4-R-substituent is strongly electron-releasing
(e.g., R ) MeO) the sulfur is sufficiently basic that, upon
approach of the [lutH]+, the proton is attracted (and hence
is close) to the sulfur. The subsequent transfer of the proton
to the sulfur thus involves only proton transfer over a short
distance. We propose that ∆Hq reflects the distance the proton
has to transfer. The negative value of ∆Sq reflects a tight,
productlike transition state. For electron-withdrawing 4-R-
substituents (e.g., R ) Cl), approach of [lutH]+ toward the
sulfur results in a hydrogen-bonded adduct in which the
proton is still predominantly associated with the nitrogen of
lut. In order to complete the transfer, the proton has to move
an appreciable distance (∆Hq is large), and the transition state
is more reactant-like (∆Sq is positive).
Negligible kinetic isotope effects have been observed in
other systems where, similar to that presented herein, the
reaction only involves the transfer of a hydrogen, or proton.
Most notably the addition of dihydrogen40 to trans-[IrCl-
(CO)(PPh3)2] to form [Ir(H)2Cl(CO)(PPh3)2] has an isotope
effect kH/kD ) 1.09 at 25.0 °C. This small value has been
attributed to the compensatory effect of an unusually large
mass and moment of inertia term (arising from translational
and rotational partition functions), a moderate inverse excita-
tion term (vibrational partition functions), and a substantial
inverse zero point energy factor. In the proton transfer
reactions of 2,4-dinitrophenol with NEt3 or piperidine, there
is no kinetic isotope effect; this has been attributed to the
lack of hydrogen bonding between the phenol and these
strong bases. In contrast, the reaction with the weaker base,
pyridine, where hydrogen bonding does occur, is associated
with a significant isotope effect.41
Summary
In this paper we have described the synthesis, character-
ization, and protonation reactions of [Ni(SC6H4R-4)2(dppe)]
(R ) MeO, Me, H, Cl, or NO2). The kinetics of the reaction
between [Ni(SC6H4NO2-4)2(dppe)] and [lutH]+ indicate a
mechanism in which initial rapid formation of a hydrogen-
bonded adduct occurs prior to the intramolecular transfer of
the proton within the hydrogen-bonded adduct. Analysis of
the kinetics allows determination of the rate constants and
equilibrium contants associated with each of the steps. To
date, intramolecular proton transfer reactions have been little
studied, despite the fact that such reactions play crucial roles
in enzymes where the access of protons to the active site
needs to be limited.42 Unfortunately, only with the nitro-
derivative is the intramolecular proton transfer step kineti-
cally distinguishable from the formation of the precursor
hydrogen-bonded adduct. In the succeeding paper, we explore
intramolecular proton transfer reactions in nickel thiolate
complexes in more detail. In particular, we will show that
the reactions of all the derivatives of [Ni(SC6H4R-4)-
(triphos)]+ (R ) MeO, Me, H, Cl, or NO2) with [lutH]+
exhibit kinetics which allow determination of the rate
constants for intramolecular proton transfer within the
hydrogen-bonded precursor, providing a complete kinetic
description of the factors affecting the rates of the largely
unexplored intramolecular proton transfer reactions.
Isotope Effects. We have studied the effect on the rate of
using [lutD]+ for selected complexes. The kinetics for the
reactions of [Ni(SC6H4Me-4)2(dppe)] or [Ni(SC6H4Cl-4)2-
(dppe)] with [lutD]+ in the presence of lut show that the
reaction is associated with a negligible kinetic isotope effect.
The use of the aprotic solvent MeCN means that the isotopic
label is not lost in rapid exchange with the solvent, and hence,
Supporting Information Available: X-ray crystallographic files
in CIF format for the three crystal structures and kinetic data for
the reactions of [Ni(SC6H4R-4)2(dppe)] (R ) MeO, Me, H, Cl, or
NO2). This material is available free of charge via the Internet at
http://pubs.acs.org.
H
D
the small kinetic isotope effect (k1 /k1 ) 1.0 ( 0.1) is
unambiguously associated with the transfer of a proton.
Previous studies39 have shown that weak kinetic isotope
effects are observed because of (i) limiting asymmetry in
the transition state; (ii) heavy atom participation in the reac-
tion coordinate resulting in an increase in isotope-sensitive
zero point energy of the transition state; or (iii) bending
modes orthogonal to the reaction coordinate which add addi-
tional isotope zero point energy to the transition state. We
are not able to distinguish which of these aspects is dominant
in the reactions of [Ni(SC6H4R-4)2(dppe)] with [lutH]+.
IC030322E
(38) Reference 23, pp 116-122.
(39) Ozaki, A. Isotopic Studies in Heterogeneous Catalysis; Academic
Press: New York, 1976; pp 172-178.
(40) Zhou, P.; Vitale, A. A.; San Filippo, J., Jr.; Saunders, W. H., Jr. J.
Am. Chem. Soc. 1985, 107, 8049.
(41) Bell, R. P.; Crooks, J. E. J. Chem. Soc. 1962, 3513.
(42) Stubbe, J.; Nocera, D. G.; Yee, C. S.; Chang, M. C. Y. Chem. ReV.
2003, 103, 2167 and refs therein.
Inorganic Chemistry, Vol. 43, No. 10, 2004 3105