Mechanistic diversity covering 15 orders of magnitude in rates: Cyanide exchange on [M(CN)4]2- (M = Ni, Pd, and Pt)

Article abstract of DOI:10.1021/ic010917e

Kinetic studies of cyanide exchange on [M(CN)₄]^2- square-planar complexes (M = Pt, Pd, and Ni) were performed as a function of pH by ¹³C NMR. The [Pt(CN)₄]^2- complex has a purely second-order rate law, with CN^- as acting as the nucleophile, with the following kinetic parameters: (k₂^Pt,CN)²⁹⁸ = 11 ± 1 s^-1 mol^-1 kg, ΔH₂? ^Pt,CN = 25.1 ± 1 kJ mol^-1, ΔS₂? ^Pt,CN = -142 ± 4 J mol^-1 K^-1, and ΔV₂? ^Pt,CN = -27 ± 2 cm³ mol^-1. The Pd(II) metal center has the same behavior down to pH 6. The kinetic parameters are as follows: (k₂^Pd,CN)²⁹⁸ = 82 ± 2 s^-1 mol^-1 kg, ΔH₂? ^Pd,CN = 23.5 ± 1 kJ mol^-1, ΔS₂? ^Pd,CN = -129 ± 5 J mol^-1 K^-1, and ΔV₂? ^Pd,CN = -22 ± 2 cm³ mol^-1. At low pH, the tetracyanopalladate is protonated (pK_a ^Pd(4,H) = 3.0 ± 0.3) to form [Pd(CN)₃HCN]^-. The rate law of the cyanide exchange on the protonated complex is also purely second order, with (k₂^PdH,CN)²⁹⁸ = (4.5 ± 1.3) × 10³ s^-1 mol^-1 kg. [Ni(CN)₄]^2- is involved in various equilibrium reactions, such as the formation of [Ni(CN)₅]^3-, [Ni(CN)₃HCN]^-, and [Ni(CN)₂(HCN)₂] complexes. Our ¹³C NMR measurements have allowed us to determine that the rate constant leading to the formation of [Ni(CN)₅]^3- is k₂^Ni(4),CN = (2.3 ± 0.1) × 10⁶ s^-1 mol^-1 kg when the following activation parameters are used: ΔH₂? ^Ni,CN = 21.6 ± 1 kJ mol^-1, ΔS₂? ^Ni,CN = -51 ± 7 J mol^-1 K^-1, and ΔV₂? ^Ni,CN = -19 ± 2 cm³ mol^-1. The rate constant of the back reaction is k_-2^Ni(4),CN = 14 × 10⁶ s^-1. The rate law pertaining to [Ni(CN)₂(HCN)₂] was found to be second order at pH 3.8, and the value of the rate constant is (k₂ ^Ni(4,2H),CN,)²⁹⁸ = (63 ± 15) × 10⁶ s^-1 mol^-1 kg when ΔH₂? ^Ni(4,2H),CN = 47.3 ± 1 kJ mol^-1, ΔS₂? ^Ni(4,2H),CN = 63 ± 3 J mol^-1 K^-1, and ΔV₂? ^Ni(4,2H),CN = -6 ± 1 cm³ mol^-1. The cyanide-exchange rate constant on [M(CN)₄]^2- for Pt, Pd, and Ni increases in a 1:7:200 000 ratio. This trend is modified at low pH, and the palladium becomes 400 times more reactive than the platinum because of the formation of [Pd(CN)₃HCN]^-. For all cyanide exchanges on tetracyano complexes (A mechanism) and on their protonated forms (III_a mechanisms), we have always observed a pure second-order rate law: first order for the complex and first order for CN^-. The nucleophilic attack by HCN or solvation by H₂O is at least nine or six orders of magnitude slower, respectively than is nucleophilic attack by CN^- for Pt(II), Pd(II), and Ni(II), respectively.

Full text of DOI:10.1021/ic010917e

Inorg. Chem. 2002, 41, 1717−1727
Mechanistic Diversity Covering 15 Orders of Magnitude in Rates:
Cyanide Exchange on [M(CN)4]2- (M ) Ni, Pd, and Pt)1
Florence J. Monlien, Lothar Helm, Amira Abou-Hamdan, and Andre´ E. Merbach*
Contribution from the Institut de Chimie Mole´culaire et Biologique, Ecole Polytechnique Fe´de´rale
de Lausanne, EPFL-BCH, CH-1015 Lausanne, Switzerland
Received August 20, 2001
Kinetic studies of cyanide exchange on [M(CN)4]2- square-planar complexes (M ) Pt, Pd, and Ni) were performed
as a function of pH by 13C NMR. The [Pt(CN)4]2- complex has a purely second-order rate law, with CN- as acting
as the nucleophile, with the following kinetic parameters: (k2Pt,CN 298
)
) 11 ± 1 s-1 mol-1 kg, ∆H q Pt,CN ) 25.1 ±
2
1 kJ mol-1, ∆S q Pt,CN ) −142 ± 4 J mol-1 K-1, and ∆V q Pt,CN ) −27 ± 2 cm mol-1. The Pd(II) metal center
3
2
2
has the same behavior down to pH 6. The kinetic parameters are as follows: (k2Pd,CN 298
)
) 82 ± 2 s-1 mol-1 kg,
∆H q Pd,CN ) 23.5 ± 1 kJ mol-1, ∆S q Pd,CN ) −129 ± 5 J mol-1 K-1, and ∆V q Pd,CN ) −22 ± 2 cm mol-1. At
3
2
2
2
low pH, the tetracyanopalladate is protonated (pK Pd(4,H) ) 3.0 ± 0.3) to form [Pd(CN)3HCN]-. The rate law of the
a
cyanide exchange on the protonated complex is also purely second order, with (k2PdH,CN 298 ) (4.5 ± 1.3) × 103 s-1
)
mol-1 kg. [Ni(CN)4]2- is involved in various equilibrium reactions, such as the formation of [Ni(CN)5]3-, [Ni(CN)3HCN]-,
and [Ni(CN)2(HCN)2] complexes. Our 13C NMR measurements have allowed us to determine that the rate constant
leading to the formation of [Ni(CN)5]3- is k2Ni(4),CN ) (2.3 ± 0.1) × 106 s-1 mol-1 kg when the following activation
parameters are used: ∆H q Ni,CN ) 21.6 ± 1 kJ mol-1, ∆S q Ni,CN ) −51 ± 7 J mol-1 K-1, and ∆V q Ni,CN ) −19
2
2
2
± 2 cm mol-1. The rate constant of the back reaction is k-2Ni(4),CN ) 14 × 106 s-1. The rate law pertaining to
[Ni(CN)2(HCN)2] was found to be second order at pH 3.8, and the value of the rate constant is (k2Ni(4,2H),CN 298
(63 ± 15) ×106 s-1 mol-1 kg when ∆H q Ni(4,2H),CN ) 47.3 ± 1 kJ mol-1, ∆S q Ni(4,2H),CN ) 63 ± 3 J mol-1 K-1, and
3
)
)
2
2
Ni(4,2H),CN
3
∆V q
) − 6 ± 1 cm mol-1. The cyanide-exchange rate constant on [M(CN)4]2- for Pt, Pd, and Ni
2
increases in a 1:7:200 000 ratio. This trend is modified at low pH, and the palladium becomes 400 times more
reactive than the platinum because of the formation of [Pd(CN)3HCN]-. For all cyanide exchanges on tetracyano
complexes (A mechanism) and on their protonated forms (I/Ia mechanisms), we have always observed a pure
second-order rate law: first order for the complex and first order for CN-. The nucleophilic attack by HCN or
solvation by H O is at least nine or six orders of magnitude slower, respectively than is nucleophilic attack by CN-
2
for Pt(II), Pd(II), and Ni(II), respectively.
Introduction
a rare example of a mechanistic study (eq 1) that can be
performed on the same Pt(II), Pd(II), and Ni(II) complexes.
We have investigated the kinetics of oxygen and cyanide
exchange on the dioxotetracyanometalate complexes of
Os(VI), Re(V), Tc(V), W(IV), and Mo(IV).3 We now extend
this study to cyanide exchange on the tetracyanometalate
complexes [M(CN)4]2- (M ) Ni, Pd and Pt) that are
thermodynamically stable yet kinetically labile in water.4
These square-planar tetracyano complexes provide us with
* Corresponding author. E-mail: andre.merbach@epfl.ch.
(1) Part 98 of the series on high-pressure NMR kinetics.
(2) Monlien, F. J.; Helm, L.; Abou-Hamdan, A.; Merbach, A. E. Inorg.
Chim. Acta. In press (part 97 of the series on high-pressure NMR
kinetics).
The reactions involving the displacement of ligands in
square-planar d8 transition-metal complexes can be described
10.1021/ic010917e CCC: $22.00 © 2002 American Chemical Society
Published on Web 02/08/2002
Inorganic Chemistry, Vol. 41, No. 7, 2002 1717

Monlien et al.
at T ) 298.1 K and I ) 0.60 mol kg-1. If the pH differed by more
than 0.1 unit between the beginning and the end of an NMR
measurement, the data were not accepted.
by the two-term equation5 kobs ) k1 + k2 [ligand]. The
second-order rate constant k2 refers to a direct attack of the
ligand on the substrate, and the pseudo-first-order rate
constant k1, to the attack of the solvent. The value of the k1
term depends only on the nature the solvent. The square-
planar complexes have characteristic kinetic behavior involv-
ing a five-coordinate intermediate or transition state. Early
exchange studies6,7 with heavy-metal cyanide complexes
were monitored with counting techniques using 14C-labeled
radiocyanide. Today, the rate of cyanide exchange is acces-
sible by kinetic NMR methods and by preliminary results
of the exchange on Pt(II) and Pd(II) for pH ) 10.3-10.8.8
The rate law was found to be second order, and no cyanide-
independent pathway was observed. Under these conditions,
the cyanide exchange for [Ni(CN)4]2- was found to be too
fast on the NMR time scale to be determined.
NMR Measurements. NMR spectra were recorded with Bruker
DPX-400 and ARX-400 spectrometers. All solutions used for 13C
measurements contained 2% D2O as the internal lock substance
and 13CH3OH as an internal chemical shift reference (49.3 ppm
with respect to TMS). Methanol was also used as reference for B0
homogeneity. The temperature was controlled by a Bruker B-VT
3000 unit and measured by substituting the sample tube for one
containing a platinum Pt-100 resistor (accuracy ) 0.5 K).9 Slow
isotopic-exchange kinetics were followed after mixing two ther-
mostated solutions at the same pH using a fast injection unit10 or
manual injection. The mixing time was less than 0.5 s. The number
of spectra (15-50), the number of scans for each spectrum, and
the supplementary delay between two successive spectra were
adjusted according to the rate of the exchange reaction.
Magnetization transfer measurements were performed using the
“inversion recovery technique” as described in the literature.11 This
technique can be performed when the longitudinal relaxation rate
1/T1 is less than or equal to the exchange rate. The exchange rate
between two sites can be deduced by selectively inverting the signal
of one and by monitoring the intensity of the signals at both sites
as a function of the delay τ between the perturbation and the
acquisition pulses. Selective inversion of the bound or free signal
was achieved using the so-called “1-3-3-1” pulse train.12 The
return of the magnetization to equilibrium is then governed by both
the 1/T1 value of the exchanging species and the exchange rate
between the sites.13
We have extended the 13C NMR study to the cyanide
exchange from pH ) 1-12.5 to determine if HCN or H2O
can act as nucleophiles. Because of the large numerical range
of the rate constants, kinetic studies were monitored by line-
broadening using the Kubo-Sack formalism as well as
magnetization-transfer and isotopic-labeling techniques. Vari-
able pressure experiments were performed to assign the
cyanide exchange mechanism to these metal centers. The
variation of the pH leads to mechanistic diversity involving
pentacoordinated species and protonation of the tetracyano-
complexes.
High-pressure high-resolution NMR spectra were monitored with
a home-built narrow bore probe (5 mm NMR tube).14 Tetra-
chloroethylene (above 298 K) and Fluobrene (below 298 K) were
used as pressurization liquids. By pumping a thermostated liquid
(synthetic oil above 298 K and ethanol below 298 K) through the
high-pressure vessel, the temperature was stabilized to ( 0.2 K.
The experiments were performed between 273.1 and 353.1 K.
Programs. Line widths of NMR signals were obtained by fitting
Lorentzian functions to the experimental spectra using the NM-
RICMA 2.7 program15for Matlab.16 The adjustable parameters are
the resonance frequency, intensity, line width, baseline, and phasing.
Complete line-shape analysis based on the Kubo-Sack formalism
using modified Bloch equations was also performed with NM-
RICMA 2.7 to extract rate constants from experimental spectra. Data
analysis was carried out with the nonlinear least-squares fitting
program Scientist.17 The reported errors correspond to 1σ.
Experimental Section
Materials and Solutions. K2[Pt(CN)4]‚3H2O, K2[Pd(CN)4]‚
3H2O, and K2[Ni(CN)4]‚H2O (Aldrich) and K13CN (Cambridge
isotope laboratory, 99% enriched) were of the highest quality
available (p.a.) and were used without further purification. Aqueous
solutions of about 0.1 mol kg-1 of each complex with variable
concentrations of KCN were freshly prepared before each experi-
ment. The ionic strength of each solution was fixed to 0.65 ( 0.10
mol kg-1 by adding KNO3 or KCF3SO3. The pH was adjusted by
adding concentrated HNO3 or CF3SO3H and KOH.
pH Measurements. pH values were determined by the poten-
tiometric technique using a titroprocessor (Titrino 716 from
Metrohm) for the addition of titrant and for mV readings. A
combined glass/reference electrode with a symmetrical electrode
chain (Radiometer Analytical S. A., pHC2406L) and a Metrohm
713 pH meter were calibrated from titrations of HCl with NaOH
Results
Cyanide Exchange on [M(CN)4]2- as a Function of pH.
Cyanide exchange on Pt(II), Pd(II), and Ni(II) was studied
(3) (a) Roodt, A.; Leipoldt, J. G.; Helm, L.; Merbach, A. E. Inorg. Chem.
1992, 31, 2864. (b) Roodt, A.; Leipoldt, J. G.; Helm, L.; Merbach, A.
E. Inorg. Chem. 1994, 33, 140. (c) Roodt, A.; Leipoldt, J. G.; Helm,
L.; Abou-Hamdan, A.; Merbach, A. E. Inorg. Chem. 1995, 34, 560.
(d) Abou-Hamdan, A.; Roodt, A.; Merbach, A. E. Inorg. Chem. 1998,
37, 1278. (e) Roodt, A.; Abou-Hamdan, A.; Engelbrecht, H. P.;
Merbach, A. E. AdV. Inorg. Chem. 2000, 49, 59.
(9) Ammann, C.; Meier, P.; Merbach, A. E. J. Magn. Reson. 1982, 46,
319.
(10) Bernhard, P.; Helm, L.; Ludi, A.; Merbach, A. E. J. Am. Chem. Soc.
1985, 107, 312.
(11) (a) Dahlquist, F. W.; Longmuir, K. J.; Du Vernet, R. B. J. Magn.
Reson. 1975, 17, 406. (b) Alger, J. R.; Prestegord, J. H. J. Magn.
Reson. 1977, 27, 137. (c) Led, J. J.; Gesmar, H. J. Magn. Reson. 1982,
49, 444.
(4) Sharpe, A. G. The Chemistry of Cyano Complexes of the Transition
Metals; Academic Press: New York, 1976.
(5) (a) Lincoln, S. F.; Merbach, A. E. AdV. Inorg. Chem.; Sykes, A. G.;
Academic Press: 1995, 1; (b) Tobe, M. L.; Burgess, J. Inorganic
Reaction Mechanisms; Longman Group: Harlow, U.K., 1999. (c)
Basolo, F.; Pearson, R. G. Mechanism of Inorganic Reactions, 2nd
ed; Wiley & Sons: New York, 1965. (d) Cross, R. J. AdV. Inorg.
Chem. 1989, 34, 219.
(6) Adamson, A. W.; Welker, J. P.; Volpe, M. J. Am. Chem. Soc. 1950,
73, 4030.
(7) Long, F. A. J. Am. Chem. Soc. 1951, 78, 537.
(12) Hore, P. J. J. Magn. Reson. 1983, 55, 283.
(13) (a) Led, J. J.; Gesmar, H. J. J. Magn. Reson. 1982, 49, 444. (b) Mann,
B. E Annu. Rep. NMR Spectrosc. 1982, 12, 263.
(14) (a) Cusaneli, A.; Nicula-dadci, L.; Frey, U.; Merbach, A. E. Inorg.
Chem. 1997, 36, 2211. (b) Frey, U.; Helm, L.; Merbach, A. E.; Roulet,
R. AdVanced Applications of NMR to Organometallic Chemistry;
Wiley & Sons: New York, 1996.
(15) Helm, L.; Borel, A. NMRICMA 2.7; University of Lausanne: 2000.
(16) Matlab, version 5.3.1; Mathworks Inc.: 1999.
(17) Scientist, version 2.0; Micromath, Inc.: Salt Lake City, UT, 1995.
(8) Pesek, J. J.; Mason, W. R. Inorg. Chem. 1983, 22, 2958.
1718 Inorganic Chemistry, Vol. 41, No. 7, 2002

Cyanide Exchange on [M(CN)4]2-
The 13C NMR spectrum of the solution containing K2[Ni-
(CN)4] and KCN is shown in Figure 1c. At low pH, the HCN
(112.4 ppm) and [Ni(CN)4]2- (135.5 ppm) signals are broad
because of cyanide exchange. These lines broaden when the
pH increases, and the signals coalesce at pH 8. The single
signal at pH 8 sharpens, and its resonance frequency moves
downfield because of the deprotonation of the ligand. Below
pH 4, it is not possible to monitor the spectrum because of
decomposition and precipitation of the substrate.
Proton Exchange on HCN. The kinetics of the proton
exchange on HCN have been reported.2,19 For clarity, the
three proposed reaction pathways are shown (eqs 2-4).
HCN,H
0
HCN + H2O { k
} CN- + H3O+
(2)
(3)
HCN,H
k-0
HCN,H
HCN + OH- { k1
} CN- + H2O
} HCN + CN-
HCN,H
k-1
HCN,H
HCN + CN- {k2
(4)
HCN,H
By combining the rate constants k0HCN,H, k1HCN,H, and k2
of the three pathways, the overall rate law for the proton
exchange on HCN, kobsHCN,H, was determined as a function
of pH (eq 5). This determination led to the following
q
parameters at 298.1 K: (k0
)
HCN,H 298 ) 113 ( 17 s-1, ∆H0
)
38 ( 2 kJ mol-1 kg, ∆S0 ) -80 ( 6 J mol-1 kg K-1,
q
HCN,H 298
(k1
(k2
)
) (2.85 ( 0.7) × 109 s-1 mol-1 kg, and
HCN,H 298
)
) (0.6 ( 0.2) × 106 s-1mol-1kg.
kobsHCN,H ) - d[HCN]/[HCN] dt )
k0HCN,H + k1HCN,H [OH-] + k2HCN,H [CN-] (5)
Figure 1. 13C NMR spectra measured as a function of pH at 298.1 K
(D2O/H2O ) 2%) for solutions containing 0.1 mol kg-1 of K2[M(CN)4]
and 0.4 mol kg-1 K13CN. (a) M ) Pt(II). (b) M ) Pd(II). (c) M ) Ni(II).
Determination of pKaHCN. Because of the fast proton
exchange on the NMR time scale, the variation of the
resonance frequency of the coalesced signal of HCN/CN-
as a function of pH can be used to determine the dissociation
at T ) 298.1 K as a function of pH using 13C NMR for
solutions containing 0.1 mol kg-1 of complexes and 0.4 mol
kg-1 of free cyanide (KCN). The pH profiles are shown in
Figure 1. The signal of the Pt(II) complex is composed of a
central line (125.0 ppm) with two satellite lines that are due
constant of HCN, KaHCN
.
KaHCN
KaHCN + [H+]
[H+]
[H+] + KaHCN
1
to cyanide that is bound to the 195Pt isotope with spin /2
δobs
)
δCN
+
δHCN (6)
(1JCPt ) 1030 Hz). The satellite integral represents 33.8%
of the total integrated signal, and the central line, 66.2%.
The resonance frequency and the line shape of the complexed
ligand are not affected by exchange with free cyanide over
the entire pH domain. At low pH, the free ligand is
protonated (pKaHCN ) 9.5). The spectrum of HCN shows a
doublet (112.4 ppm) with a coupling constant of 1JCH ) 272
Hz. With increasing pH, the two doublet signals broaden
and coalesce (at pH 7) because of the decreasing lifetime of
H+ on HCN. From pH 8 to 9.5, the coalesced signal broadens
and moves toward the CN- frequency (165.1 ppm), and then
the line narrows at basic pH when CN- predominates.
In Figure 1b, the 13C NMR spectra show a pH-independent
resonance frequency of the palladium complex [Pd(CN)4]2-
(131.2 ppm). Its line width is affected by cyanide exchange
only for pH > 9, with a small amount of line broadening.
The free ligand shows the same behavior with pH as does
the platinum complex; however, the free-ligand line width
is slightly affected by the cyanide exchange at basic pH.
This NMR titration was performed for solutions containing
CN- and HCN, with and without2 [Pt(CN)4]2-and [Pd(CN)4]2-.
The fitted values for δCN and δHCN, the resonance frequencies
of the CN- and HCN forms, are 165.1 ( 0.2 and 112.4 (
0.2 ppm, respectively, with pKaHCN ) 9.5 ( 0.2. The results
were the same for all three solutions, within experimental
error.
Cyanide Exchange on [Pt(CN)4]2-. The tetracyanoplati-
nate(II) complex is very stable in aqueous solution (and even
at low pH with excess ligand), with an overall stability
constant of log â4 ) 40.20 The cyanide-exchange rate law
was determined at 298.1 K using pH adjustments to monitor
CN- variation. The exchange rate constants were determined
(18) Bednar, R. A.; Jenks, W. P. J. Am. Chem. Soc. 1985, 107, 7117.
(19) Banyai, I.; Blixt, J.; Glaser, J.; Toth, I. Acta Chem. Scand. 1992, 46,
142.
(20) Grinberg, A. A.; Gelman, M. I. Dokl. Akad. Nauk 1960, 133, 1081.
Inorganic Chemistry, Vol. 41, No. 7, 2002 1719

Monlien et al.
Figure 3. (a) Evolution of 13C NMR spectrum after a rapid mixing of
H13CN with [Pt(CN)4]2- at pH ) 6.0 and 298.1 K. The signal intensity of
the complex increases while that of HCN decreases as a function of time.
CPt ) 0.100 mol kg-1, CCN ) 0.262 mol kg-1, I ) 0.56 mol kg-1. (b)
Figure 2. (a) Evolution of high-pressure 13C NMR spectra after selective
inversion of the complex signal (125.0 ppm) as a function of the delay
between the perturbation and the acquisition pulses at P ) 160 MPa and
Time evolution of the 13C populations of HCN and [Pt(CN)4]2-
.
change reactions were followed by rapidly mixing a H13CN
solution with a complex solution at the same temperature
(298.1 K) and pH using a fast injection unit. We paid special
attention to the long relaxation time (T1) of bound CN-.
Experimental T1 values as a function of pH for [Pt(CN)4]2-,
HCN/CN-, and mixtures of [Pt(CN)4]2- and HCN/CN- are
from 15 to 22 s. In the presence of cyanide and water
containing dissolved O2, [Pt(CN)4]2- had a shorter 13C
relaxation time T1 than that measured by Pesek et al.22 By
monitoring the increase in the intensity of the complex signal
and the decrease in the intensity of the free-ligand signal,
the kinetics constants were determined with the modified
McKay equation (eq 7).23,24 A typical example of the 13CN-
signal increase observed with time is given in Figure 3.
pH ) 11.5. (b) Least-squares fit of the integrals. CPt ) 0.101 mol kg-1
,
CCN ) 0.413 mol kg-1, I ) 0.69 mol kg-1, D2O/H2O ) 2%, and T )
298.1 K.
using the magnetization transfer technique from pH 11.7 to
9.7. The system includes six sites (one for CN-, two
for HCN because of spin-spin coupling, and three for
[Pt(CN)4]2-) and two exchange processes (for proton ex-
change on HCN and for cyanide exchange on [Pt(CN)4]2-).
Because of the fast exchange between HCN and CN-, the
signal of the free ligand can be attributed to one unique site.
The platinum system is therefore reduced in a two-site
complex. Inversion was performed on the three lines of the
complex (Figure 2). For each of the spectra collected during
the magnetization transfer experiments, Lorentzian functions
were fitted to the signals of the free ligand and to the central
line as a function of τ. The contribution of the satellite lines
to the integral of the central line and the ratio of all the signal
line widths were taken into account. The equations given
by Led and Gesmar13 were fitted to the experimental
integrals21 using an iterative nonlinear least-squares routine.
Nine parameters were adjusted: the final and the initial
integrals of the free (f) (CN-/HCN) and bound (b) ligands,
their spin-lattice relaxation rates 1/T1f and 1/T1b, the
populations of the two sites, and the exchange rate constant.
Below pH 8, magnetization transfer cannot be used; however,
at pH e 6, the reaction exchange rate is slow enough to be
determined using isotopic labeling experiments. The ex-
x(t) ) x∞ + (x0 - x∞) exp(-kobsPd,CNt/(1 - x∞)) (7)
x(t) represent the mole fraction of coordinated 13CN as a
function of time, and x0 and x∞ are the limiting conditions
at t ) 0 and for the exchange equilibrium, respectively.
Lorentzian functions were fitted to both NMR signals with
NMRICMA to determine the values of the integrals as a
function of time. kobsPt,CN, x∞ , and x0 were determined by
fitting the 13C NMR data to eq 7. Isotopic exchange
measurements were performed at pH 6.0, 3.6, 2.6, and 1.1
(Figure 4).
(22) Pesek, J. J.; Mason R. Inorg. Chem. 1979, 18, 924.
(23) McKay, H. J. Am. Chem. Soc. 1943, 65, 702.
(24) Lincoln, S. F.; Merbach, A. E. AdV. Inorg. Chem.; Sykes A. G., Ed.;
Acad. Press: San Diego. 1995.
(21) Frey, U.; Elmroth, S.; Moullet, B.; Elding, L.; Merbach, A. E. Inorg.
Chem. 1991, 30, 5033.
1720 Inorganic Chemistry, Vol. 41, No. 7, 2002

Cyanide Exchange on [M(CN)4]2-
ibility coefficient of activation ∆âq ) -(10 ( 2) × 10-2
cm3 mol-1 MPa-1.
q
Ɖq
∆V2
k2Pt,CN ) (k2Pt,CN) exp -
P +
P2
(10)
(
)
0
RT
2RT
Cyanide Exchange on [Pd(CN)4]2-. The tetracyano-
palladate(II) is also very stable in aqueous solution. Its overall
formation constant as determined by Fasman et al. is log â4
) 51.7.25 The rate law of the cyanide exchange on [Pd(CN)4]2-
was determined at pH ) 11.4. 13C NMR spectra of 0.05 mol
kg-1 [Pd(CN)4]2- solutions containing variable KCN con-
centrations (0.053-0.427 mol kg-1) at fixed ionic strength
(I ) 0.6 ( 0.1 mol kg-1) were measured at 298.1 K. The
resonance frequencies of the complexed (131.2 ppm) and
the free (165.1 ppm) cyanide did not change, whereas their
line widths broaden with increasing free-cyanide concentra-
tion. The inverse of the mean lifetime of cyanide on
[Pd(CN)4]2-, 1/τPd ) krPd, were extracted from the NMR
spectra by line-shape analysis using the Kubo-Sack formal-
ism with the following 4 × 4 exchange matrix D2.
Figure 4. (a) Variation of the 13C NMR line width of the [Pt(CN)4]2-
signal with pH at 298.1 K. (b) Variation of the concentration-normalized
(CCN ) 0.4 mol kg-1) exchange rate constant k′obsPtCN with pH for cyanide
exchange on [Pt(CN)4]2-. CPt ) 0.100 mol kg-1, I ) 0.69 mol kg-1, D2O/
H2O ) 2%, and T ) 298.1 K.
Analysis of the pH dependence (eq 8) indicates a purely
Pt,CN 298
second-order rate law (Figure 4) with (k2
s-1 mol-1 kg.
)
) 11 ( 1
2-
-d[Pt(CN)4
]
) k2Pt,CN [CN-] )
Pt,CN
kobs
)
2-
dt [Pt(CN)4
]
KaHCN
CN KaHCN + [H+]
k2Pt,CNC
(8)
D2 is composed of the three sites for the ligand (HRCN,
HâCN, and CN-) and one site for the complex [Pd(CN)4]2-,
with the corresponding populations PHCN/2, PHCN/2, PCN, and
PCNb. The resonance frequencies of HRCN, HâCN, CN-, and
[Pd(CN)4]2- were fixed to 11 441, 11 169, 16 638, and
13 200 Hz, respectively, on a 400 MHz NMR spectrometer.
The inverse of the mean lifetime of cyanide at the HCN site,
krHCN, was fixed at values calculated from activated param-
eters given previously. The inverse of the mean lifetime of
cyanide at the [Pd(CN)4]2- site, krPd, was determined as a
function of the free-cyanide concentration by fitting the
experimental spectra. krPd is defined for the exchange of a
particular cyanide on the metal center Pd(II). The observed
where CCN is the analytical concentration of the free cyanide.
A variable-temperature study of the exchange rate constant,
kobsPt,CN, was performed by 13C NMR magnetization transfer
from 271.1 to 317.1 K at pH ) 11.5 on a solution containing
0.096 mol kg-1 of the Pt complex and 0.392 mol kg-1 of
the free cyanide. Higher temperatures cannot be used because
of cyanide decomposition. Fitting eq 9 to the data led to the
Pt,CN 298
following activation parameters: (k2
)
) 11 ( 1 s-1
q Pt,CN
q Pt,CN
mol-1 kg, ∆H2
) 25.1 ( 0.4 kJ mol-1, and ∆S2
) -142 ( 2 J mol-1 K-1.
q
q
kBT
h
∆S2
∆H2
RT
T
Pt,CN
Pd,CN
k2
)
exp
-
)
kinetic rate constant of cyanide exchange on Pd(II), kobs
[
]
R
Pd,CN
(eq 12), is given by krPd (kobs
) krPd). The analysis of
q
∆H2
[CN-] dependence shows a negligible first-order k1Pd,CN term,
1
1
T
Pt,CN 298
(k2
)
exp
-
(9)
[
]
)
(
298.1
R
298.1
and a purely second-order rate law was fitted to the
Pd,CN
experimental data, leading to k2
) 78 ( 2 s-1 mol-1
Variable-pressure experiments on 13C NMR magnetization
transfer were performed as a function of pressure at pH 11.5
on a solution containing 0.101 mol kg-1 of the Pt complex
and 0.413 mol kg-1 of the free cyanide. Equation 10 was
fitted (Figure 2) to the exchange rate constant obtained as a
function of pressure, leading to the rate constant at zero
pressure (k2Pt,CN)0 ) 10.2 ( 0.3 s-1 mol-1 kg, the activation
volume ∆V2q Pt,CN ) -27 ( 2 cm3 mol-1, and the compress-
kg.
2-
-d[Pd(CN)4
]
Pd,CN
kobs
)
) k2Pd,CN[CN-]
(12)
2-
dt [Pd(CN)4
]
(25) Fasman, A. B.; Kutyukov, G. C.; Sokolskii, D. V. Zh. Neorg. Khim.
1965, 10, 1338.
Inorganic Chemistry, Vol. 41, No. 7, 2002 1721

Monlien et al.
Scheme 1. Cyanide Exchange on Pd(II) Center
8.1 to 6.5. Selective inversions were performed on the signal
of the free or the bound ligand. Below pH 6, the exchange
rate is slow compared to the longitudinal relaxation rate, and
Pd,CN
kobs
could not be determined using magnetization
transfer. At pH e 6.7, the exchange rate was determined
using isotopic labeling experiments. The isotopic labeling
experiment was performed at pH ) 6.7 (the highest value)
to overlap with the domain where magnetization transfer
techniques were applied. The value of rate constant does not
vary with the modification of the technique. Twelve experi-
ments were performed from pH 6.7 to 0.8 (Figure 5).
All observed rate constants are represented as a function
Pd,CN
of pH in Figure 5. In basic solution (pH > 10), kobs
pH-independent. For pH < 10, kobs
is
Pd,CN
decreases with pH
because of the decrease in the concentration of CN- upon
protonation. The rates reach a second plateau between pH 6
and 3 and decrease again for pH < 3. Pd(II) behavior is
different than that of Pt(II). This difference can be explained
by the formation of a protonated complex whose influence
on the observed rate constant can be deduced from Scheme
1.
Figure 5. (a) Variation of the populations PPd of [Pd(CN)4]2- and PPdH of
Pd
[Pd(CN)3HCN]- with pH. (b) Evolution of the complex line width, 1/T2
,
as a function of pH. CPd ) 0.095 mol kg-1 and CCN ) 0.374 mol kg-1. (c)
Variation of the concentration-normalized (CCN ) 0.4 mol kg-1) rate
constant k′obsPd,CN with pH for cyanide exchange on [Pd(CN)4]2-, with the
contribution of both pathways. (- - -) PPdk2Pd,CN[CN-] and (- ‚‚ -)
PPdHk2PdH,CN[CN-]. CPd ) 0.095 mol kg-1, I ) 0.66 mol kg-1, and D2O/
H2O ) 2%.
The rate law (eq 13) is thus composed by two terms:
k2Pd,CN, the second-order rate constant of cyanide exchange
on [Pd(CN)4]2-, and k2PdH,CN, the second-order rate constant
of cyanide exchange on the protonated complex [Pd(CN)3-
(HCN)]-. KaPd(4,H) and KaHCN are the proton dissociation
constants of [Pd(CN)3(HCN)]- and HCN, respectively.
A variable-temperature study (287.3-328.9 K) of the
exchange reaction was performed on a solution containing
0.1 mol kg-1 of the Pd complex and 0.472 mol kg-1 of the
free cyanide at pH ) 11.4 by extracting the rate constants
from the line-shape analysis of the 13C NMR spectra. Higher
temperatures could not be used without decomposition of
the cyanide. The exchange rate constant at 298.1 K,
kobsPd,CN ) PPdk2Pd,CN [CN-] + PPdHk2PdH,CN [CN-] )
Pd,CN 298
(k2
)
) 88 ( 3 s-1 mol-1 kg, and the activation
KaHCN
q Pd,CN
parameters ∆H2q Pd,CN) 23.5 ( 1.1 kJ mol-1 and ∆S2
) -141.9 ( 4.0 J mol-1 K-1 were obtained with an Eyring
fit to the data.
(PPdk2Pd,CN + PPdHk2
)
CCN (13)
PdH,CN
KaHCN + [H+]
A variable-pressure study at pH ) 11.4 and 323.1 K was
PPd ) KaPd(4,H)/(KaPd(4,H) + [H+]) and PPdH ) [H+]/(KaPd(4,H)
+ [H+]) are the populations of [Pd(CN)4]2- and [Pd(CN)3-
(HCN)]-, respectively (Figure 5a). Equation 13 was fitted
to the observed rate constants kobsPd,CN between pH 11.7 and
performed on a 0.1 mol kg-1 [Pd(CN)4]2- solution containing
Pd,CN
0.371 mol kg-1 of the free ligand. The variation of k2
with pressure was calculated from the observed rate constant
kobsPd,CN, which was determined using the Kubo-Sack
0.8, leading to the proton dissociation constant KaPd(4,H)
)
)
q
formalism (D2 matrix). The activation volume ∆V2 ) -22
Pd,CN
3.0 ( 0.3 and to the second-order rate constants k2
Pd,CN
( 2 cm3 mol-1 and the zero-pressure rate constant (k2
)
0
67 ( 13 s-1 mol-1 kg and k2PdH,CN ) (4.5 ( 1.3) × 103 s-1
) 197 ( 15 s-1 mol-1 kg were determined by fitting eq 11
to the variable-pressure data. The value of the compressibility
coefficient of activation, ∆âq ) - (8 ( 1) × 10-2 cm3 mol-1
MPa-1, is noteworthy.
mol-1 kg (Figure 5). The contributions of both pathways are
values were normalized
with respect to concentration CCN ) 0.4 mol kg-1, leading
Pd,CN
shown in Figure 5c. The kobs
Pd,CN
to k′obs
values as a function of pH.
A variable-pH study gave us another opportunity to study
Cyanide Exchange on [Ni(CN)4]2-. The ionic strength
was fixed with potassium triflate to avoid complex formation
with the nitrate.26 Two 13C NMR signals are obtained if
KNO3 is added to a solution of [Ni(CN)4]2-. The overall
formation constant of [Ni(CN)4]2- (yellow) in aqueous
solution has been determined by UV-visible and IR
how the rate law changes with CN- concentration. The 13
C
NMR spectra in Figure 1b were used to extract the observed
rate constants, kobsPd,CN, using the Kubo-Sack formalism with
the D2 matrix in the basic domain (pH 11.7-9.1). Below
pH 9, the rate constant was too small to induce detectable
line-broadening in the signal of the coordinated CN-. Thus,
magnetization transfer experiments were performed from pH
(26) Beck, M.; Bjerrum, J. Acta Chem. Scand. 1962, 16, 2050.
1722 Inorganic Chemistry, Vol. 41, No. 7, 2002

Cyanide Exchange on [M(CN)4]2-
spectroscopy and leads to a log â4 ) 30.3 ( 0.1.27 At high
CN- concentration, a pentacoordinated [Ni(CN)5]3- (red)
species is formed, and the formation constant K5 is small
(K5 ≈ 0.15 mol l-1) and depends slightly on the ionic
strength.28
Scheme 2. Cyanide Exchange on Ni(II) Center
Ni(4),CN
[Ni(CN)4]2- + CN- { 4k2
} [Ni(CN)5]3-
Ni(4),CN
4k-2
K5 ) [Ni(CN)53-]/[Ni(CN)42-][CN-] (14)
The [Ni(CN)4]2- ion is very stable in water (nickel bis-
(dimethyl)glyoximate, for instance, is soluble in aqueous
cyanide solution). However, it is rapidly dissociated at pH
< 4, with the formation of Ni(CN)2 and finally Ni2+.29-31
This process is reversible by adding KOH and raising the
pH to 4. Before and during the dissociation process,
[Ni(CN)4]2- can be protonated. Margerum and Kolski29 have
measured the following three successive proton dissociation
constants: pKaNi(4,H) ) 5.4 for [Ni(CN)3(HCN)]-, pKaNi(4,2H)
) 4.5 for [Ni(CN)2(HCN)2], and pKaNi(4,3H) ) 2.6 for
[Ni(CN)(HCN)3]+.
The cyanide exchange on [Ni(CN)4]2- at very high pH is
known to be fast on the NMR time scale, and previously
only the limit8 (>5 × 105 s-1mol kg-1) of the exchange rate
constant was given. The cyanide exchange induces line
broadening over the entire pH domain (Figure 1c); therefore,
the intermolecular-exchange rate constant can be extracted
from line-shape analysis using the Kubo-Sack formalism
from pH 10.3 to 3.8. At pH > 7, the exchange matrix must
be modified relative to the palladium system to take into
account the pentacoordinated species. The protonation of
[Ni(CN)4]2- can be neglected under these basic condition.
The new exchange matrix D3 is given in eq 15. It is
composed of five sites: two for HCN, one for CN-, one for
the tetracyano complex, Ni(4), and one for the pentacyano-
nickelate(II), Ni(5), site.
of the mean lifetime of cyanide at the [Ni(CN)4]2- site.
According to this matrix, cyanide exchange takes place only
via the formation of the pentacoordinated species. The
resonance frequencies of HRCN, HâCN, CN-, and [Ni(CN)4]2-
were fixed at 11 441, 11 169, 16 638, and 13 645 Hz,
respectively, on a 400 MHz spectrometer. The populations
were calculated as functions of the nickel concentration, CNi,
the total cyanide concentration, Ctot (Ctot ) CCN + 4CNi),
the formation constant of [Ni(CN)5]3-, K5 (eq 15), the
dissociation constant of HCN, KaHCN, and the pH.
To be able to use matrix D3, the resonance frequency of
[Ni(CN)5]3- was first determined. Six solutions were pre-
pared at pH 11.4 with 0.05 mol kg-1 of [Ni(CN)4]2- and
variable amounts of KCN (0.06-0.52 mol kg-1). The 13C
NMR spectra of these solutions showed a coalesced singlet
that shifted downfield as the cyanide concentration increased.
The resonance frequency of [Ni(CN)5]3- can be deduced
from the variation in the shift of the coalesced signal by
fitting eq 16 as a function of Ctot to the 13C data. δCN (165.1
ppm), δNi(4) (135.6 ppm), and δNi(5) are the resonance
frequencies of cyanide in CN-, [Ni(CN)4]2-, and [Ni(CN)5]3-,
respectively.
CN
Ni(4)
Ni(5)
δobs ) P
δ
+ 4P
δ
+ 5P
δ
Ni(5)
(16)
CN
Ni(4)
The populations were calculated as functions of Ctot, CNi,
K5, and pH. K5 was fixed to the value found by Hugel, 0.168
( 0.003 mol-1 kg.28c A Lorentzian function was fitted to
each spectrum, and the analysis of the dependence of the
resonance frequency on the concentratin of CN- led to the
adjusted resonance frequency of δNi(5) ) 230 ( 26 for
[Ni(CH)5]3-.
pH Variation on [Ni(CN)4]2-. The rate of the cyanide
exchange on [Ni(CN)4]2- was determined as a function of
pH. At basic pH, D3 is used because of the formation of the
pentacoordinated species. However, at low pH, the penta-
coordinated complex can be neglected, but the protonated
complexes must be taken into account. The cyanide exchange
can take place via the three pathways given in Scheme 2:
Surprisingly, the protonation does not affect the chemical
shift of the 13C NMR complexed-cyanide signal. Therefore,
the three complexes have the same [Ni(CN)4]2- signal in
13C NMR. For this reason, D2 was used for pH < 8. The
four sites correspond to HRCN, HâCN, CN-, and Ni(4)tot,
the tetracoordinated complexes of Ni(II) with or without
protonation. The observed rate constant of the cyanide
exchange at low pH is equal to krNi(4) in D2. This observed
rate constant can be expressed as a function of the free-
The value of the rate constant krHCN is the same as that in
the palladium D2 matrix. The rate constant krNi(4) is the inverse
(27) Freund, H.; Schneider, C. R. J. Am. Chem. Soc. 1959, 81, 4780.
(28) (a) Penneman, R. A.; Bain, R.; Gilbert, G.; Jones, L. H.; Nyholm, R.
S.; Reddy, K. N. J. Am. Chem. Soc. 1963, 85, 6. (b) Van Geet, A. L.;
Hume, D. N. Inorg. Chem. 1964, 4, 523. (c) Pierrard, J. C.; Hugel, R.
ReV. Chim. Miner. 1971, 8, 831. (d) McCullough, R. L.; Jones, H. L.;
Penneman, R. A. J. Inorg. Nucl. Chem. 1960, 13, 286.
(29) Kolski, G. B.; Margerum, D. W. Inorg. Chem. 1968, 7, 2239.
(30) Persson, H. Acta Chem. Scand. A 1974, 28, 885.
(31) Persson, H.; Ekstro¨m, C. G. Acta Chem. Scand., Ser. A 1976, 30, 31.
Inorganic Chemistry, Vol. 41, No. 7, 2002 1723

Monlien et al.
pathway can be suppressed without significantly affecting
the values of the other two rate constants (Figure 6c and d).
Concentration Dependence. At pH ) 7.5, cyanide
exchange takes place predominately (more than 98%) via
the [Ni(CN)4]2- complex. The cyanide exchange kinetics
were determined as a function of the concentration of free
cyanide to verify the rate law. At this pH, the presence of
the protonated complexes could be neglected, but that of the
protonated ligand could not (Figure 6a). Five solutions of
K2[Ni(CN)4] (0.1 mol kg-1) with variable KCN concentra-
tions (0.059-0.952 mol kg-1) were analyzed by 13C NMR
at 298.1 K. Using the Kubo-Sack formalism and D2, the
Ni(4),CN
kobs
values were extracted. They are linearly propor-
tional to the concentration of the free-CN- ligand (without
the intercept) and lead to the second-order rate constant
k2Ni(4)CN ) (2.5 ( 0.5) × 106 s-1 mol-1 kg. At pH ) 3.8, the
cyanide exchange (>85%) mainly occurs via the [Ni(CN)2-
(HCN)2] species. Thus, the purely second-order rate law was
also verified by measuring the rate constant as a function of
the free-cyanide concentration. 13C NMR spectra of five
solutions containing 0.05 mol kg-1 of K2[Ni(CN)4] and
0.066-0.528 mol kg-1 of KCN were obtained at 298.1 K,
and rate constants were extracted from line-shape analysis
Ni(4,2H),CN
using D2. The iteration volume led to the value k2
) (67 ( 4) × 106 s-1 mol-1 kg with a negligible first-order
Ni(4,2H),CN
term. This value confirms that the value of k2
)
(50 ( 18) × 106 s-1 mol-1 kg that was obtained with the
pH-variation method is correct.
Temperature Dependence. A variable-temperature study
Ni,CN
of kobs
was performed at pH ) 7.5 (and 272.1-323.1
K, mainly for [Ni(CN)4]2-) and pH ) 3.8 (and 274.6-329.8
K, mainly for [Ni(CN)2(HCN)2]) on solutions containing
0.100 mol kg-1 of K2[Ni(CN)4] and 0.385 ( 0.005 mol kg-1
of KCN. Equation 9 was fitted to the values of k2Ni(4),CN that
Figure 6. (a) Variation of the populations PNi(5) of [Ni(CN)5]3-, PNi(4) of
[Ni(CN)4]2-, PNi(4,H) of [Ni(CN)3HCN]-, and PNi(4,2H) of [Ni(CN)2(HCN)2]
with pH. (b) Evolution of the complex line width, 1/T2Ni, as a function of
pH. CNi ) 0.092 mol kg-1 and CCN ) 0.384 mol kg-1. (c) Variation of the
concentration-normalized (CCN ) 0.4 mol kg-1) rate constant k′obsNi,CN with
pH for cyanide exchange on [Ni(CN)4]2-, with the contribution of three
pathways: (i) PNi(4)k2Ni(4),CN[CN-] ) PNi(5)k-2Ni(4),CN, (ii) PNi(4,H)k2Ni(4,H),CN[CN-]
for k2Ni(4,H),CN ) 10 × 106 s-1 mol-1 kg, and (iii) PNi(4,2H)k2Ni(4,2H),CN[CN-].
CNi ) 0.092 mol kg-1, I ) 0.60 mol kg-1, and D2O/H2O ) 2%.
were obtained at pH 7.5, leading to the following param-
q
eters: (k2
)
Ni(4),CN 298 ) (2.3 ( 0.1) × 106 s-1 mol-1 kg, ∆H2
q Ni(4),CN
) 21.6 ( 2 kJ mol-1 kg, and ∆S2
) -51 (
Ni(4),CN
7 J mol-1 kg K-1. For the experiment at pH 3.8, we obtained
the following parameters for the cyanide exchange on [Ni-
(CN)2(HCN)2]: (k2
)
Ni(4,2H),CN 298 ) (63 ( 15) × 106 s-1 mol-1
cyanide concentration (eq 17).
q
Ni(4,2H),CN
q
kg, ∆H2
) 47.3 ( 1.0 kJ mol-1 kg, and ∆S2
) 63 ( 3 J mol-1 kg K-1.
Ni(4,2H),CN
Ni(4,H),CN
kobsNi,CN ) (PNi(4,2H)k2Ni(4,2H),CN + PNi(4,H)k2
+
Pressure Dependence. Variable-pressure 13C NMR spec-
tra were acquired at pH ) 8.0 and T ) 297.5 K for a solution
[CN-]
Ni(4,2H) + PNi(4,H) + PNi(4)
Ni(4),CN
P
Ni(4)k2
)
(17)
containing 0.102 mol kg-1 of [Ni(CN)4]2- and 0.382 mol
P
Ni(4),CN
kg-1 of KCN. The rate constants kobs
were obtained
q
13C NMR spectra were measured as a function of pH at 298.1
K for a solution containing 0.093 mol kg-1 of [Ni(CN)4]2-
and 0.384 mol kg-1. Equation 17 was fitted to the values of
kobsNi,CN obtained with D3 for pH ) 10.3-7 and with D2 for
using D2. These values give an the activation volume of ∆V2
) -19 ( 2 cm3 mol-1, with a negligible compress-
Ni(4),CN
ibility coefficient of activation. Variable-pressure 13C NMR
experiments were also performed at pH ) 3.5 and T ) 298.1
K on a solution of 0.1 mol kg-1 of K2[Ni(CN)4] and 0.393
pH ) 8-3 (Figure 6). The iteration routine led to the
q Ni(4,2H),CN
mol kg-1 of KCN. An activation volume of ∆V2
following values: k2
kg, k2
) (2.7 ( 0.4) × 106 s-1 mol-1
Ni(4),CN
) -6 ( 1 cm3 mol-1 kg was calculated for the cyanide
exchange on [Ni(CN)2(HCN)2] at acidic pH.
Ni(4,H),CN
Ni(4,2H),CN
< 10 × 106 s-1 mol-1 kg, and k2
)
(50 ( 18) × 106 s-1 mol-1 kg. The populations were
expressed as functions of Ctot, CNi, K5, pKaHCN, pH, pKaNi(4,H)
,
Discussion
and pKaNi(4,2H). Literature values of pKaNi(4,H) and pKaNi(4,2H)
were used. We gave an upper limiting value to k2
because in the fitting procedure the [Ni(CN)3(HCN)]-
-
[M(CN)4]2- Complexes. The chemical shift data for the
square-planar [M(CN)4]2- (Table 1) show a definitive trend
Ni(4,H),CN
1724 Inorganic Chemistry, Vol. 41, No. 7, 2002

Cyanide Exchange on [M(CN)4]2-
Table 1. 13C NMR Chemical Shifts, log â4 Stability Constants, and
FCN and FMC Force Constants of the [M(CN)4]n- Square-Planar
Complexes
δ (ppm) log â4 FCN (mdynes/Å) FMC (mdynes/Å)
CN-
165.1
135.5
131.2
125.0
112.4
104.2
16.90e
17.20f
17.44f
17.41f
18.70e
17.44g
[Ni(CN)4]2-
[Pd(CN)4]2-
[Pt(CN)4]2-
HCN
31a
2.25f
2.32f
2.75f
51.7b
40c
[Au(CN)4]-
85d
2.99g
a Reference 27. b Reference 25. c Reference 20. d Reference 36. e Ref-
erence 33. f Reference 34. g Reference 35.
to higher field as the progression from Ni(II) to Pd(II) to
Pt(II) is made. The resonance frequency of the Au(III)
complex lies at 104.2 ppm, which is below that of HCN and
can be attributed to the higher charge of the metal center. A
substantial shift to higher field was already observed for the
13C resonance frequencies of isoelectronic cyano complexes
as the metal oxidation state is increased.32 Another important
factor affecting the chemical shift is the geometry of the
cyano complex, which explains the high chemical shift value
(230 ( 26 ppm) of the pentacyanonickelate(II) complex. In
these complexes, the σ donation of the anionic CN- likely
dominates the π-backbonding; therefore, the variation of the
chemical shift (from the shielding of the 13C nucleus) also
reflects the variation of the σ-bond strength: Au-C > Pt-C
> Pd-C > Ni-C. The trend in the chemical shifts of the
square-planar complexes is clearly observed for the M-C
force constants33-35 but is hardly observed for the C-N force
constants. These conclusions are consistent with the variation
in log â4.36
It is well-known that [Ni(CN)4]2- (pKaNi(4,H) ) 5.4) can
be protonated in acidic medium to form [Ni(CN)3HCN]- and
further-protonated species.29 Our kinetic results confirm this
behavior and allow us, for the first time, to determine the
pKaPd(4,H) () 3.0 ( 0.2) of the palladium analogues kineti-
cally. Surprisingly, for the palladium system, which under-
goes slow intermolecular cyanide exchange at low pH, no
variation of the chemical shift was observed with changes
in pH, within experimental error. This result implies that
resonance frequency variations due to protonation are small.
Interestingly, the 13C NMR line width of the bound cyanide
at pH 3.5 increases slightly (from 1 to 10 Hz), in agreement
with the pKaPd(4,H) value determined from this work. A similar
broadening is observed at pH 1.5 for the 13C (from 2 to 7
Hz) and 195Pt NMR signals (from 4 to 7 Hz) of the platinum
complex. These observations are in line with the expected
trend in pKaM(4,H) values: Ni > Pd > Pt.
Figure 7. Variation of kobsM,CN on [M(CN)4]n- in water at 298.1 K where
M ) Au(III), Pt(II), Pd(II), and Ni(II). CCN ) 0.4 mol kg-1 and CM ) 0.1
mol kg-1
.
decrease of ˜4 orders of magnitude as the pH decreases to
6. At low pH, the tetracyanopalladate (pKaPd(4,H) ) 3.0 (
0.3) is protonated to [Pd(CN)3HCN]-. Because of a much
faster cyanide exchange on [Pd(CN)3HCN]- (k2
≈ 55
PdH, CN
k2Pd,CN), the contribution of [Pd(CN)3HCN]- to the overall
exchange becomes detectable around pH 6.5 and is predomi-
nant at pH < 5.5. The rate law of the cyanide exchange on
the protonated complex is also purely second order. The
[Ni(CN)4]2- is involved in various equilibrium reactions
(formation of a pentacoordinated and of protonated com-
plexes). At high pH, [Ni(CN)4]2- and [Ni(CN)5]3- are in
equilibrium, as previously shown by UV-visible spectros-
copy.27,28 Our 13C NMR measurements have allowed to verify
Ni(4),CN
this behavior and to determine the rate constant (k2
)
(2.3 ( 0.1) × 106 s-1 mol-1 kg) leading to the formation of
Ni(4),CN
[Ni(CN)5]3- as well as the backward rate constant (k-2
) 14 × 106 s-1). The overall CN- exchange rate decreases
down to pH 6, where the protonated complexes are formed.
The reactivity of these complexes is higher because of
protonation (Table 2). The rate law for [Ni(CN)2(HCN)2]
was found to be second order at pH 3.8.
For all cyanide exchanges on tetracyano complexes and
on their mono- and diprotonated forms, we have always
observed a pure second-order rate law: first order for the
complex and first order for CN-. However, this rate law does
not apply to square-planar complexes. For example, the
chloride exchange on [PtCl4]2- takes place only via a
reversible aqueous solvation process37 (first-order rate law),
but two pathways (first and second order) were observed
for the chloride exchange on [AuCl4]-.38 We set the
following upper limits to the first-order terms of the cyanide
exchange (from Figure 7): >10-9 s-1 for Pt(II), 10-5 s-1
for Pd(II), and 10 s-1 for Ni(II). These limits imply that
nucleophilic attack by HCN or solvation by H2O for Pt(II),
Pd(II), and Ni(II) is at least 9, 6, and 6 orders of magnitude
Cyanide Exchange as a Function of pH. The [Pt(CN)4]2-
complex has a purely second-order rate law, with CN- acting
as a nucleophile. Therefore, the observed exchange rate
decreases ˜9 orders of magnitude the pH decreases (Figure
7). The Pd(II) metal center exhibits similar behavior, with a
(32) Pesek, J. J.; Mason, W. R. Inorg. Chem. 1979, 18, 924 and references
therein.
(33) Jones, L. H. Inorg. Chem. 1963, 2, 777.
(34) Kubas, G. J.; Jones, L. Inorg. Chem. 1974, 13, 2816.
(35) Jones, L. H.; Smith, J. M. J. Chem. Phys. 1964, 41, 2507.
(36) Remick, A. E. J. Am. Chem. Soc. 1947, 69, 94.
(37) Grantham, L. F.; Elleman, T. S.; Martin, D, S. J. Am. Chem. Soc.
1955, 77, 2965.
(38) Rich, R.; Taube, H. J. Phys. Chem. 1954, 58, 1.
Inorganic Chemistry, Vol. 41, No. 7, 2002 1725

Monlien et al.
Table 2. Activation Parameters of the Cyanide Exchange on the
[M(CN)4]n-/HCN/CN- System in an Aqueous Medium at 298.1 K
M,X
k2
(s-1 mol-1
kg)a
∆Hq(kJ ∆Sq(J mol-1 ∆Vq(cm3
mol-1 K-1 mol-1
metal
)
)
)
Ni(II)b
k2
k2
k2
k2
k2
k2
(2.3 ( 0.1) × 106 21.6 ( 2
-51 ( 7
-19 ( 2
Ni(4),CN
Ni(4,H),CN
Ni(4,2H),CN
Pd,CN
<10 × 106
(63 ( 15) × 106 47.3 ( 1
+63 ( 3
-6 ( 1
Pd(II)c
Pt(II)e
82 ( 2d
23.5 ( 1d -129 ( 5d -22 ( 2
25.1 ( 1d -142 ( 4d -27 ( 2
(4.5 ( 1.3) × 103
PdH,CN
Pt,CN
11 ( 1d
Au(III)e,f k2
k2
6240 ( 85
0.56 ( 0.03
40.0 ( 1
65.1 ( 1
-38 ( 3
-31 ( 3
+2 ( 1
-14 ( 2
Au,CN
Au,Cl
a The rate constants reported are for the exchange of a particular
coordinated X-. The rate constant for an unspecified X- is therefore four
times greater than the rate constant for the coordinated X-, and the
corresponding ∆Sq value is 11.5 () R ln 4) times greater than ∆Sq for the
coordinated X-. b The rate law is given in equation 17. c The rate law is
given in equation 13. d The kinetic parameters obtained previously for pH
Pd,CN
) 10.3 and 10.8 are for Pd (Pt): k2
) 124(25) s-1 mol-1 kg; ∆Hq )
17(24) kJ mol-1, and ∆Sq ) -147(137) J mol-1K-1 from references 8 and
e
M,L
M,L
42. kobs
) k2
[L]. f Reference 2.
slower, respectively, than nucleophilic attack by CN- for
Pt(II), Pd(II) and Ni(II), respectively.
Figure 8. Effect of pressure on RT ln(k/k20) for the cyanide exchange (a)
on [M(CN)4]n- where M ) Au(III), Pt(II), Pd(II), and Ni(II), CM ) 0.10 (
We have recently reported the rate law and the pH
variation of the cyanide exchange on Au(III) square-planar
complex in water.2 The observed cyanide-exchange rate
constant on [Au(CN)4]-, kobsAu,CN, is also presented in Figure
7. The Au(III) and Pt(II) complexes are not kinetically
affected by the protonation of the complex, whereas Pd(II)
and Ni(II) show an important change at low pH because of
the faster cyanide exchange on the protonated species. The
increase of the rate at low pH can be attributed to the more
positive charge of the protonated complex and to a weaker
M-CN bond. No kinetic effect of the protonation of
[Pt(CN)4]2- is observed, probably because of a lower pKaPt(4,H)
value, as discussed above.
0.05 mol kg-1, CCN ) 0.40 ( 0.5 mol kg-1, and I ) 0.6 ( 0.1 mol kg-1
.
For Au(III) and Pt(II), pH ) 11.5, T ) 298.1 K. For Pd(II), pH ) 11.4 and
T ) 323.1 K. For Ni(II), pH ) 10.4 and T ) 279.9 K and (b) on [Ni-
(CN)2(HCN)2]. CM ) 0.10 ( 0.05 mol kg-1, CCN ) 0.40 ( 0.5 mol kg-1
I ) 0.6 ( 0.1 mol kg-1, pH ) 3.5, and T ) 298.1 K.
,
same order of magnitude as the value obtained for the
isonitrile ligand and confirms the much higher selectivity of
the Pt(II) center with respect to the Pd(II) center. The divalent
complexes [Ni(CN)4]2-, [Pd(CN)4]2-, and [Pt(CN)4]2- have
approximately the same activation enthalpy; however, the
difference in reactivity among these three complexes is
reflected in the entropy term. The negative values of ∆Sq
are indicative of an associative activation mode. The most
negative values are observed for platinum and palladium,
implying the formation of a fully pentacoordinated species
at the transition state. The much less negative value for
Ni(II) indicates that the coordination of the transition-state
species is intermediate between tetra- and pentacoordination,
as confirmed by the true thermodynamic equilibrium between
[Ni(CN)4]2- and [Ni(CN)5]3-.
Variable Pressure. The activation volumes are clearly
negative and show the same trend as do the entropies (Table
2, Figure 8). Using the same argument as above and taking
into account the fact that the pentacoordinated species is
observed for nickel, we conclude that this species participates
in a limiting A exchange pathway. We observed that the
magnitude of the activation volume for cyanide exchange
on Pd(II) is always smaller than that for the corresponding
exchange on Pt(II).42 Furthermore, the ∆Vq values for CN-
exchange on [Pt(CN)4]2- (-27 cm3 mol-1) and [Pd(CN)4]2-
(-22 cm3 mol-1) are the most negative ever observed for
homoleptic square-planar ligand-exchange reactions. For
example, the following ∆Vq values were reported for Pt(II)
and Pd(II), respectively, in Me2S and in H20: -22 and -9.4
Kinetic Parameters. All the activation parameters, rate
constants, and rate laws are listed in Table 2. The cyanide-
exchange rate constant on [M(CN)4]2- increases in a 1:7:
200 000 ratio for Pt/Pd/Ni (Figure 7). This trend is modified
at low pH; the palladium becomes 400 times more reactive
than the platinum because of the formation of [Pd(CN)3HCN]-.
The high reactivity of [Ni(CN)4]2- is explained by the
stability of the pentacoordinate intermediate that becomes
stable for nickel. [Ni(CN)5]3- has been isolated with both
regular square-pyramidal and distorted trigonal-bipyramidal
geometries present in the same crystal.39 The high reactivity
of Au(III) with respect to the isoelectronic Pt(II) is explained
by the higher oxidation state. The relative reactivity of
palladium with respect to platinum depends strongly on the
nature of the ligand involved in ligand exchange on their
Pt,L
homoleptic complexes. The values of k2Pd,L/k2 for H2O,
Me2S, and MeNC are 1.4 × 106, 1.39 × 103, and 1.7,
respectively.40-44 k2Pd,L/k2Pt,L ) 7 for CN- exchange is of the
(39) (a) Raymond, K. N.; Corfield, P. W. R.; Ibers, J. A. Inorg. Chem.
1968, 7, 1362. (b) Jurnak, F. A.; Raymond, K. N. Inorg. Chem. 1974,
13, 2387.
(40) Helm, L.; Elding, L. I.; Merbach, A. E. Inorg. Chem. 1985, 24, 1719.
(41) Helm, L.; Elding, L. I.; Merbach, A. E. HelV. Chim. Acta 1984, 67,
1453.
(43) Olsson, A.; Kofod, P. Inorg. Chem. 1992, 31, 183.
(44) Hallinan, N.; Besanc¸on, V.; Forster, M.; Elbaze, G.; Ducommun, Y.;
Merbach, A. E. Inorg. Chem. 1991, 30, 1112.
(42) Frey, U.; Elmroth, S.; Moullet, B.; Elding, L.; Merbach, A. E. Inorg.
Chem. 1991, 30, 5033.
1726 Inorganic Chemistry, Vol. 41, No. 7, 2002

Cyanide Exchange on [M(CN)4]2-
cm3 mol-1 (Me2S) and -4.6 and -2.2 cm3 mol-1 (H2O).
Therefore, we attribute our results to a limiting A mechanism,
even though an intermediate could not be formally detected.
The ∆Vq value for cyanide exchange on [Ni(CN)2(HCN)2]
is much less negative than that for the deprotonated species,
but the value for the entropy of cyanide exchange on
[Ni(CN)2(HCN)2] is less positive than that for the deproto-
nated species. These results are indicative of a type I/Ia
interchange process with a small amount of associative
character.
or transition states as well as protonated complexes. The
reactivities of these species are strongly pH-dependent,
covering 15 orders of magnitude in reaction rates. The
nucleophilic attack by HCN is at least 6-9 orders of
magnitude slower than is the nucleophilic attack by CN- on
[M(CN)4]2-.
Acknowledgment. We thank the Swiss National Science
Foundation for financial support.
Supporting Information Available: All exchange rate con-
stants as a function of pH, concentration, temperature, and pressure;
tables of the resonance frequencies and line widths; fitted parameters
for magnetization transfer and isotopic labeling experiments. This
material is available free of charge via the Internet at http://
pubs.acs.org.
Conclusions
The cyanide exchange on [M(CN)4]2- with M ) Pt, Pd,
and Ni is a rare case in which mechanistic comparisons
between 3d, 4d, and 5d transition-metal complexes. Surpris-
ingly, the behavior of these metal square-planar centers leads
to mechanistic diversity involving pentacoordinated species
IC010917E
Inorganic Chemistry, Vol. 41, No. 7, 2002 1727