Mechanistic insight gained into the ligand substitution reactions of bis(o-benzosemiquinonediiminato)(triphenylphosphane)cobalt(III) - Kinetic, solvent, and volume profile studies

Article abstract of DOI:10.1002/ejic.200300107

The ligand substitution reactions of bis(o-benzosemiquinone-diiminato)(triphenylphosphane)cobalt(III), [Co ^III(s-BQDI)₂-(Ph₃P)]⁺, were studied with imidazole (Imid) and 4-dimethyl-aminopyridine (4-Me₂Npy) as entering nucleophiles in MeOH and CH₃CN as solvents (S). The complex [Co^III(s-BQDI)₂(Ph₃P)S]⁺ undergoes dissociation of the solvent to form a five-coordinate intermediate, [Co ^III(s-BQDI)₂(Ph₃P)]⁺, which binds the entering nucleophile in a rate-determining step through a six-coordinate transition state, [Co^III(s-BQDI)₂(Ph₃P)L] ⁺, followed by the release of triphenylphosphane to form [Co ^III(s-BQDI)₂L]⁺. From the temperature and pressure dependence of the substitution of triphenylphosphane by imidazole, the activation parameters for the forward and reverse reactions of [Co ^III(s-BQDI)₂(Ph₃P)]⁺ with imidazole in MeOH were found to be ΔH? = 58±2 and 43.4±0.5 kJ·mol^-1, ΔS? = -116±6 and -73±2 J·K^-1·mo^-1, and ΔV? = -10.6±0.1 and -8.7±0.3 cm³·mo^-1, respectively. In CH₃CN, however, ΔH?, ΔS?, and ΔV? for the forward reaction were found to be 50±3 kJ·mol^-1, -111±9 J·mol^-1· K^-1, and -12.9±0.3 cm³·mol^-1, respectively. The activation parameters for the reaction between [Co ^III(s-BQDI)₂(Ph₃P)]⁺ and 4-dimethylaminopyridine in MeOH for the forward and reverse reactions were found to be ΔH? = 76±2 and 47.9±0.4 kJ·mol^-1, ΔS? = -51±7 and -64±2 J·K^-1·mo^-1, and ΔV? = -10.5±0,3 and -11.1±0.2 cm³·mo^-1, respectively. From these reported rate and activation parameters, the substitution of triphenylphosphane by imidazole and 4-dimethylaminopyridine follows an associative mechanism. The ligand substitution reactions of [Co ^III(s-BQDI)₂(Ph₃P)]⁺ in CH ₃CN were found to be faster than those in MeOH, which is attributed to the potential for hydrogen bond formation with the entering nucleophile in the case of MeOH as solvent. Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2003.

Full text of DOI:10.1002/ejic.200300107

FULL PAPER
Mechanistic Insight Gained into the Ligand Substitution Reactions of
Bis(o-benzosemiquinonediiminato)(triphenylphosphane)cobalt(III) ؊
Kinetic, Solvent, and Volume Profile Studies
Basam M. Alzoubi,[a] Mohamed S. A. Hamza,[a][‡] Carlos Dücker-Benfer,[a] and
Rudi van Eldik*[a]
Keywords: Cobalt / Kinetics / Ligand substitution / Mechanism / Pressure dependence / Solvent effects
The ligand substitution reactions of bis(o-benzosemiquinone- CH3CN, however, ∆H‡, ∆S‡, and ∆V‡ for the forward reaction
diiminato)(triphenylphosphane)cobalt(III),
[CoIII(s-BQDI)2-
were found to be 50 3 kJ·mol−1, −111 9 J·mol−1·K−1, and
−12.9 0.3 cm3·mol−1, respectively. The activation parameters
for the reaction between [CoIII(s-BQDI)2(Ph3P)]+ and 4-di-
methylaminopyridine in MeOH for the forward and reverse
(Ph3P)]+, were studied with imidazole (Imid) and 4-dimethyl-
aminopyridine (4-Me2Npy) as entering nucleophiles in
MeOH and CH3CN as solvents (S). The complex [CoIII(s-
BQDI)2(Ph3P)S]+ undergoes dissociation of the solvent to
form a five-coordinate intermediate, [CoIII(s-BQDI)2(Ph3P)]+,
which binds the entering nucleophile in a rate-determining
reactions were found to be ∆H‡ = 76 2 and 47.9 0.4 kJ·mol−1
,
∆S‡ = −51 7 and −64 2 J·K−1·mol−1, and ∆V‡ = −10.5 0.3 and
−11.1 0.2 cm3·mol−1, respectively. From these reported rate
step through
a
six-coordinate transition state, [CoIII(s- and activation parameters, the substitution of triphenylphos-
BQDI)2(Ph3P)L]+, followed by the release of triphenylphos-
phane to form [CoIII(s-BQDI)2L]+. From the temperature and
pressure dependence of the substitution of triphenylphos-
phane by imidazole, the activation parameters for the for-
phane by imidazole and 4-dimethylaminopyridine follows an
associative mechanism. The ligand substitution reactions of
[CoIII(s-BQDI)2(Ph3P)]+ in CH3CN were found to be faster
than those in MeOH, which is attributed to the potential for
ward and reverse reactions of [CoIII(s-BQDI)2(Ph3P)]+ with hydrogen bond formation with the entering nucleophile in
imidazole in MeOH were found to be ∆H‡ = 58 2 and
43.4 0.5 kJ·mol−1, ∆S‡ = −116 6 and −73 2 J·K−1·mol−1, and
the case of MeOH as solvent.
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
∆V‡ = −10.6 0.1 and −8.7 0.3 cm3·mol−1, respectively. In Germany, 2003)
Introduction
The chemistry of transition metal complexes of benzose-
miquinonediimine (s-BQDI), derived from o-phenylenedia-
mine (OPD), are rather important because of their interest-
ing redox properties[1,2] and their relation to biologically rel-
evant catechol and benzoquinone derivatives.[3]
(1)
The bidentate ligand o-benzosemiquinonediiminato(Ϫ1),
position becomes occupied by a weakly bound solvent mol-
ecule, which cannot be observed in the solid-state struc-
ture.[3,6] Complexes such as [CoIII(s-BQDI)2(Ph3P)]PF6,
derived from the one-electron oxidation of o-phenylenedia-
mine, can be stabilised by complex formation.[4] By removal
of a second electron it is converted into the unstable o-
benzoquinonediimine (BQDI) [Equation (1)], which can
also be stabilised through complexation.[2]
[CoIII(s-BQDI)2I],
[CoIII(s-BQDI)2SCN],
or
[Fe(s-
BQDI)2(Ph3P)] are potentially five-coordinate in CH3CN
as solvent.[2]
X-ray diffraction shows that the structure of [CoIII(s-
BQDI)2(Ph3P)]ClO4 is square-pyramidal in the solid state,
with PPh3 occupying one of the axial positions.[5] Upon dis-
solution in a coordinating solvent such as MeOH, the sixth
We are interested generally in equilibria between five- and
six-coordinate species in such complexes, since this should
provide a better understanding of their mechanistic behav-
iour in ligand displacement processes. Alkyl cobalamines
and cobinamides (vitamin B12 derivatives) with stronger do-
nor ligands (e.g., vinyl, methyl, sulfite, dialkyl phosphite)
exist as temperature-dependent mixtures of yellow five-co-
ordinate and red six-coordinate aqua (solvent) complexes.
Increasing the temperature favours the five-coordinate form
with diethyl phosphite as ligand, whereas increasing the
[a]
Institute for Inorganic Chemistry, University of Erlangen-
Nürnberg,
Egerlandstr. 1, 91058 Erlangen, Germany
[‡]
On leave from: Department of Chemistry, Ain Shams
University, Cairo, Egypt
Supporting information for this article is available on the
WWW under http://www.eurjic.org or from the author.
2972
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
DOI: 10.1002/ejic.200300107
Eur. J. Inorg. Chem. 2003, 2972Ϫ2978

Ligand Substitution Reactions of Bis(o-benzosemiquinonediiminato)(triphenylphosphane)cobalt(III)
FULL PAPER
pressure favours the six-coordinate form.[7] The reaction (ε ϭ 2.2·104, 1.1·104, and 7.9·103 Ϫ1·cmϪ1, respectively).
volume for the formation of the six-coordinate species in Upon mixing of the complex with imidazole or 4-(Me)2Npy
water as solvent was found to be Ϫ12.5 cm3 molϪ1, suggest- in the stopped-flow instrument, there is a small and very
ing that the entering water molecule is fully bound in the rapid initial increase in absorbance prior to the reaction
six-coordinate complex.[7]
with the entering nucleophile. This rapid increase can be
The ligand substitution reactions of [CoIII(s-BQDI)2L]ϩ, ascribed to a fast equilibration between the complex and
where
Ph3Sb and Ph3As, and of [CoIII(s- the solvent (methanol) as a result of the equilibrium be-
L
ϭ
CBQDI)2(Ph3As)]ϩ, where s-CBQDI ϭ 4,5-dichloro-1,2- tween the five- and six-coordinate complexes in solution as
benzosemiquinonediiminato, were studied in methanol with discussed above. This rapid absorbance change is also ob-
different nucleophiles such as pyridine, pyrrolidine, and imi- served in the absence of imidazole or 4-(Me)2Npy, just by
dazole as entering ligands.[3,6] It was found that these nucle- dilution of a solution of the complex with solvent in a
ophiles easily displace the coordinated solvent molecule to stopped-flow instrument, and occurs within the mixing time
form the six-coordinate [CoIII(s-BQDI)2LLЈ]ϩ complexes, of the instrument (2 ms) at room temperature. With the use
where LЈ represents the entering nucleophile.[3] In a sub- of a Biologic low temperature stopped-flow unit, however,
sequent step, substitution of L by the entering nucleophile the spectral changes associated with this reaction could be
takes place, and an associative interchange mechanism recorded at Ϫ40 °C, and a typical kinetic trace (see Figure
was suggested.
S-2, Supporting Information) indicates that this reaction is
In this investigation we performed a detailed analysis of over within 2 s at Ϫ40 °C. Subsequently, there is a much
the axial ligand substitution reactions of [CoIII(s- larger absorbance decrease at 710 and 450 nm, and an in-
BQDI)2(Ph3P)S]ϩ by imidazole and 4-(Me)2Npy, where S is crease at 540 nm, associated with the coordination of imi-
MeOH or CH3CN, in order to study the intimate nature of dazole. The [CoIII(s-BQDI)2(Imid)]ϩ complex shows two
the ligand displacement mechanism and the effect of the bands at 716 and 496 nm, with a shoulder at 420 nm (ε ϭ
selected solvent on these reactions.
1.1·105, 5.7·104, and 3.9·104 Ϫ1·cmϪ1
 , respectively),
whereas [CoIII(s-BQDI)2(4-Me2Npy)]ϩ has characteristic
UV/Vis bands at 714 and 520 nm, with a shoulder at
422 nm (ε ϭ 6.7·104, 5.2·104, and 3.1·104 Ϫ1·cmϪ1, respec-
tively).
Results and Discussion
We started our studies by investigating the effect of tem-
perature and pressure on the UV/Vis spectra of [CoIII(s-
BQDI)2(Ph3P)]ϩ in methanol as solvent. The UV/Vis spec-
trum of 5·10Ϫ5  [CoIII(s-BQDI)2(Ph3P)]ϩ was recorded as
a function of temperature, as shown in Figure S-1 (Support-
ing Information). With decreasing temperature, from 60 to
5 °C, it exhibits isosbestic points at 508, 600, and 760 nm,
a significant absorbance decrease at 712 and 450 nm, and
an increase in absorbance at 550 nm. Similar spectra were
obtained either on decreasing the temperature or on in-
creasing the pressure.
The effect of temperature and pressure on the UV/Vis
spectra suggests an equilibrium between five- and six-coor-
dinate forms of the complex [CoIII(s-BQDI)2(Ph3P)S]ϩ,
where the sixth coordination site is occupied by a solvent
molecule. The six-coordinate form is favoured by increasing
pressure or decreasing temperature, and vice versa for the
five-coordinate complex. Similar trends could be observed
in both solvents (MeOH and CH3CN). In the case of
CH3CN as solvent, however, the change in the UV/Vis spec-
trum was not as significant as in that of MeOH as solvent,
which is attributed to the fact that CH3CN is a much
stronger nucleophile than MeOH and will form a less labile
six-coordinate species. For this reason a smaller fraction of
the five-coordinate species will exist in equilibrium with the
six-coordinate species in the case of CH3CN as solvent. In
the solid state the complex is square-pyramidal with a Ph3P
ligand occupying the axial position.[5]
The
reaction
between
2.5·10Ϫ5

[CoIII(s-
BQDI)2(Ph3P)]ϩ and an excess of imidazole (0.05 to 1.0 )
was studied in methanol as a function of temperature
(15Ϫ45 °C). Kinetic traces for this reaction revealed perfect
pseudo-first order behaviour, and the kinetic results as a
function of the imidazole concentration are shown in Fig-
ure 1. It follows that good linear plots with significant inter-
cepts are obtained within the experimental error limits. Fur-
thermore, the plots do not indicate any saturation at high
[Imid]. This kinetic behaviour can be expressed by the rate
law given in Equation (2), where ka and kb represent the
rate constants for the forward (slope) and reverse (intercept)
reactions, respectively. Values of ka and kb as a function of
temperature, along with the corresponding activation par-
ameters, are summarised in Table 1. Values of ∆H‡ and ∆S‡
for the forward reaction (ka) were found to be
58Ϯ2 kJ·molϪ1 and Ϫ116Ϯ6 J·KϪ1·molϪ1, whereas those
for the reverse reaction (kb) were found to be
68Ϯ1 kJ·molϪ1 and Ϫ82Ϯ3 J·KϪ1·molϪ1, respectively.
kobsd. ϭ ka [L] ϩ kb
(2)
The results in Figure 1 clearly indicate that we are dealing
with an unfavourable displacement of triphenylphosphane
by imidazole, as can be seen from the magnitude of the
intercepts of the plots that represent the contribution of the
reverse reaction. Under the selected experimental con-
ditions (no free phosphane added), the reverse displacement
Reaction between [CoIII(s-BQDI)2(Ph3P)]؉ and Imidazole
in Methanol
The UV/Vis spectrum of [CoIII(s-BQDI)2(Ph3P)]ϩ exhib- of imidazole by phosphane will follow second-order kine-
its two bands at 710 and 478 nm, with a shoulder at 423 nm tics, with the result that the kinetic data measured under
Eur. J. Inorg. Chem. 2003, 2972Ϫ2978
www.eurjic.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2973

B. M. Aizoubi, M. S. A. Hamza, C. Dücker-Benfer, R. van Eldik
FULL PAPER
Figure 2. Plot of kobsd. versus [Ph3P] for the reaction between
1.25·10Ϫ4  [CoIII(s-BQDI)2(Imid)]ϩ and Ph3P in methanol at 25
°C; the best fit (solid line) gives kb ϭ 23.3 Ϫ1·sϪ1
Figure 1. Plots of kobsd. versus [Imid] for the reaction between [CoIII
-
(s-BQDI)2(Ph3P)]ϩ and imidazole in methanol as a function of
temperature; experimental conditions: [CoIII] ϭ 2.5·10Ϫ5 , tem-
perature: 15.0 (A), 25.0 (B), 35.0 (C), 45.0 °C (D)
Table 2. Kinetic data for the reaction between [CoIII(s-
BQDI)2(Imid)]ϩ and Ph3P in methanol as a function of tempera-
ture and pressure
Table 1. Kinetic data for the reaction between [CoIII(s-
BQDI)2(Ph3P)]ϩ and imidazole in the absence of added phosphane
in methanol as a function of temperature and pressure {[CoIII(s-
BQDI)2(Ph3P)]ϩ ϭ 2.5·10Ϫ5 }
Pressure, MPa
0.1
Temp., °C
kb/[Ph3P], Ϫ1 sϪ1 [a]
5.0
10.0
15.0
20.0
25.0
25.0
6.2Ϯ0.3[b]
8.8Ϯ0.2[b]
12.5Ϯ0.1[b]
17.2Ϯ0.3[b]
23.2Ϯ0.1[b]
23.6Ϯ0.6[c]
26.9Ϯ0.3[c]
30.7Ϯ2.0[c]
36.0Ϯ1.5[c]
43.4Ϯ0.5
Pressure, MPa
0.1
Temp., °C ka ϫ 104, Ϫ1 sϪ1 kb ϫ 104, sϪ1 [a]
15.0
25.0
35.0
45.0
15.0
1.79Ϯ0.08
4.5Ϯ0.3
9.5Ϯ0.5
19.3Ϯ1.4
2.39Ϯ0.06
2.85Ϯ0.01
3.41Ϯ0.08
4.06Ϯ0.10
58Ϯ2
1.21Ϯ0.05
3.2Ϯ0.2
8.6Ϯ0.3
19.4Ϯ0.8
1.30Ϯ0.04
1.47Ϯ0.01
1.63Ϯ0.05
1.87Ϯ0.07
68Ϯ1
10
50
90
130
10
50
90
130
∆H‡, kJ·molϪ1
∆S‡, J·KϪ1·molϪ1
∆V‡, cm3·molϪ1
Ϫ73Ϯ2
Ϫ8.7Ϯ0.3
∆H‡, kJ·molϪ1
∆S‡, J·KϪ1·molϪ1
∆V‡, cm3·molϪ1
Ϫ116Ϯ6
Ϫ10.6Ϯ0.1
Ϫ82Ϯ3
Ϫ7.2Ϯ0.3
[a]
Second-order rate constant calculated from the slope of plots of
[b]
kb versus [Ph3P].
[CoIII(s-BQDI)2(Imid)]ϩ ϭ 1.25·10Ϫ4 ,
[a]
First-order rate constant obtained from the intercept of plots of
[Ph3P] ϭ 2.5·10Ϫ2 .
[Ph3P] ϭ 1.25·10Ϫ2 .
[CoIII(s-BQDI)2(Imid)]ϩ ϭ 1.25·10Ϫ4 ,
[c]
kobsd. versus [Imid].
pseudo-first order conditions at low imidazole concen-
trations could become unreliable due to a contribution of in Table 2. The ∆H‡ and ∆S‡ values for kb were found to be
the second-order reverse process. We therefore studied the 43.4Ϯ0.5 kJ·molϪ1 and Ϫ73Ϯ2 J·KϪ1·molϪ1, respectively.
rate and activation parameters for the reverse substitution
The
reaction
between
2.5·10Ϫ5

[CoIII(s-
of imidazole by Ph3P separately, by first treating 2.5·10Ϫ4  BQDI)2(Ph3P)]ϩ and imidazole ([Imid] ϭ 0.05Ϫ1 ) was
of [CoIII(s-BQDI)2(Ph3P)]ϩ with 1  imidazole to form the studied at different pressures (10Ϫ130 MPa). The results
imidazole complex, followed by the addition of an excess of are shown in Figure 3, from which it follows that good lin-
Ph3P to form the [CoIII(s-BQDI)2(Ph3P)]ϩ complex again. ear plots with significant intercepts are obtained within the
Figure 2 shows a linear plot of kobsd. versus [Ph3P] for the experimental error limits. The values of ka and kb as a func-
reaction between [CoIII(s-BQDI)2(Imid)]ϩ and an excess of tion of pressure, along with the corresponding activation
Ph3P (0.005Ϫ0.025 ). The linear plot exhibits a negligible volumes, are summarised in Table 1. Plots of ln ka and ln
intercept, indicating that the displacement of imidazole by kb versus pressure were linear within the error limits (Fig-
phosphane goes to completion under the selected con- ures S-3 and S-4, Supporting Information), and values of
ditions. The slope of the line in Figure 2 represents kb, for ∆V‡ for the forward and the reverse reactions were found
which the results are reported as a function of temperature to be Ϫ10.6Ϯ0.1 and Ϫ7.2Ϯ0.3 cm3·molϪ1, respectively.
2974
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2003, 2972Ϫ2978

Ligand Substitution Reactions of Bis(o-benzosemiquinonediiminato)(triphenylphosphane)cobalt(III)
FULL PAPER
Because of the possible unreliability of kb under these con-
ditions, as indicated above, its pressure dependence was also
measured directly by treatment of the imidazole complex
with an excess of phosphane, the results of which are in-
cluded in Table 2. A plot of ln kb versus pressure gave a
good linear relationship, from which ∆V‡ was calculated to
be Ϫ8.7Ϯ0.3 cm3·molϪ1
.
Figure 4. Volume profile for the reaction (the given structures are
schematic presentations)
within the experimental error limits. Furthermore, the plots
do not indicate any saturation at high 4-Me2Npy concen-
trations. In this case the intercepts are smaller and the
slopes of the plots larger than in the case of the reaction
with imidazole, indicating that a more favourable equilib-
rium situation is reached in the reaction with 4-Me2Npy.
The same rate law and data treatment as used for the reac-
tion with imidazole can be adopted, and the results are
summarized as a function of temperature and pressure in
Table 3. The values of ∆H‡ and ∆S‡ for the forward reaction
(ka) were found to be 76Ϯ2 kJ·molϪ1 and Ϫ51Ϯ7
J·KϪ1·molϪ1, respectively, and those for the reverse reaction
to be 61Ϯ4 kJ·molϪ1 and Ϫ115Ϯ12 J·KϪ1·molϪ1, respec-
tively. Because of the possible contribution of a second-or-
der reverse reaction at low 4-Me2Npy concentration, we
also determined the rate and activation parameters for the
substitution of 4-Me2Npy by Ph3P to form [CoIII(s-
BQDI)2(Ph3P)]ϩ directly, by treatment of 2.5·10Ϫ4  [CoIII-
(s-BQDI)2(Ph3P)]ϩ with 1  4-Me2Npy, followed by the
reaction with phosphane. A plot of kobsd. versus [Ph3P] for
the reaction between [CoIII(s-BQDI)2(4-Me2Npy)]ϩ and ex-
cess Ph3P (0.005Ϫ0.025 ) is shown in Figure S-6 (Support-
ing Information), from which it follows that the back reac-
tion does not contribute significantly under these con-
ditions. The data for kb are summarised in Table 4. The val-
ues of ∆H‡ and ∆S‡ were found to be 47.9Ϯ0.4 kJ·molϪ1
and Ϫ 64Ϯ2 J·KϪ1·molϪ1, respectively.
Figure 3. Plots of kobsd. versus [Imid] for the reaction between [CoIII
-
(s-BQDI)2(Ph3P)]ϩ and phosphane in methanol as a function of
pressure; experimental conditions: [CoIII] ϭ 2.5·10Ϫ5 , pressure:
10 (A), 50 (B), 90 (C), 130 MPa (D)
From the available kinetic and spectroscopic data, we
suggest that the reaction between [CoIII(s-BQDI)2(Ph3P)S]ϩ
and imidazole (or 4-Me2Npy, see later) takes place in two
consecutive steps. There is rapid dissociation of coordinated
solvent in Equation (3) to form a five-coordinate intermedi-
ate, followed by the rate-determining associative ligand ex-
change process Equation (4) to form the product complex.
[CoIII(s-BQDI)2(Ph3P)S]ϩ ǟǞ [CoIII(s-BQDI)2(Ph3P)]ϩ ϩ S
(3)
(4)
From the activation parameters and the constructed vol-
ume profile shown in Figure 4, it is reasonable to conclude
that the reaction between [CoIII(s-BQDI)2(Ph3P)]ϩ and imi-
dazole in methanol follows an associative mechanism
through the formation of a six-coordinate intermediate/
transition state.
The
reaction
between
2.5·10Ϫ5

[CoIII(s-
BQDI)2(Ph3P)]ϩ and 4-Me2Npy (0.05 to 1 ) as a function
of pressure (10Ϫ130 MPa) was also studied, and the results
are reported in Figure S-7 (Supporting Information). The
data show good linear plots with significant intercepts
Reaction between [CoIII(s-BQDI)2(Ph3P)]؉ and 4-(Me)2Npy
in Methanol
The
reaction
between
2.5·10Ϫ5

[CoIII(s- within the experimental error limits. The values of ka and
BQDI)2(Ph3P)]ϩ and an excess of 4-Me2Npy (0.05 to 1 ) kb as a function of pressure, along with the corresponding
in methanol as a function of temperature (15Ϫ45 °C) was activation parameters, are summarised in Table 3. Plots of
studied in a similar way. The results are presented in Figure lnka and lnkb versus pressure gave good linear fits, from
S-5 (Supporting Information), from which it follows that which the values of ∆V‡ for the forward and reverse reac-
good linear plots with significant intercepts are obtained tions were calculated to be Ϫ10.5Ϯ0.3 and Ϫ8.7Ϯ0.6
Eur. J. Inorg. Chem. 2003, 2972Ϫ2978
www.eurjic.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2975

B. M. Aizoubi, M. S. A. Hamza, C. Dücker-Benfer, R. van Eldik
FULL PAPER
6.7·10Ϫ4 and 4.5·10Ϫ4

Ϫ1·sϪ1 at 25 °C, respectively,
Table 3. Kinetic data for the reaction between [Co(s-
BQDI)2(Ph3P)]ϩ and 4-Me2Npy in the absence of added phos-
phane in methanol as a function of temperature and pressure
{[CoIII(s-BQDI)2(Ph3P)]ϩ ϭ 2.5·10Ϫ5 }
whereas the rate constants for the reverse reactions are 12.1
and 23.2 Ϫ1·sϪ1 at 25 °C, respectively. This indicates that
the forward rate constant (ka) increases with increasing ba-
sicity of the entering ligand, whereas the reversible reaction
rate constant (kb) decreases with increasing basicity due to
the stronger cobaltϪligand bond.
Pressure, MPa
0.1
Temp., °C ka ϫ 104, Ϫ1 sϪ1 kb ϫ 104, sϪ1 [a]
Reaction between [CoIII(s-BQDI)2(Ph3P)]؉ and Imidazole
in Acetonitrile
15.0
25.0
35.0
45.0
15.0
2.5Ϯ0.2
6.7Ϯ0.5
19Ϯ1
0.54Ϯ0.02
1.6Ϯ0.2
3.1Ϯ0.2
The
reaction
between
2.5·10Ϫ5

[CoIII(s-
BQDI)2(Ph3P)]ϩ and excess imidazole ([Imid] ϭ 0.025Ϫ1
) in acetonitrile was studied at different temperatures
(11Ϫ32 °C). The results are presented in Figure 5 and show
a significant deviation from linearity at high imidazole con-
centrations, as well as negligible intercepts. This behaviour
can be accounted for in terms of the interchange mecha-
nism presented in reactions according to Equations (5) and
(6), for which the rate law is given in Equation (7), where K
represents the equilibrium constant for precursor forma-
tion, and k the interchange rate constant. Since acetonitrile
is a much more strongly coordinating solvent than meth-
anol, this presumably prevents formation of a five-coordi-
nate intermediate and forces the reaction to proceed
through an interchange instead of an associative mecha-
nism involving a five-coordinate intermediate.
54Ϯ2
6.9Ϯ0.4
10
50
90
130
2.5Ϯ0.1
3.10Ϯ0.08
3.65Ϯ0.09
4.3Ϯ0.1
76Ϯ2
0.54Ϯ0.04
0.64Ϯ0.05
0.74Ϯ0.06
0.83Ϯ0.06
61Ϯ4
∆H‡, kJ·molϪ1
∆S‡, J·KϪ1·molϪ1
∆V‡, cm3·molϪ1
Ϫ51Ϯ7
Ϫ10.5Ϯ0.3
Ϫ115Ϯ12
Ϫ8.7Ϯ0.6
[a]
First-order rate constant obtained from the intercept of plots of
kobsd. versus [4-Me2Npy].
Table 4. Kinetic data for the reaction between [CoIII(s-BQDI)2(4-
Me2Npy)]ϩ and Ph3P in methanol as a function of temperature
and pressure
[CoIII(s-BQDI)2(Ph3P)S]ϩ ϩ Imid ǟǞ
(5)
Pressure, MPa
0.1
Temp., °C
kb/[Ph3P], Ϫ1 sϪ1 [a]
{[CoIII(s-BQDI)2(Ph3P)S]ϩ·Imid}; K
5.0
2.8Ϯ0.1[b]
4.2Ϯ0.1[b]
6.0Ϯ0.1[b]
8.6Ϯ0.1[b]
12.1Ϯ0.1[b]
15.4Ϯ0.8[c]
18.6Ϯ1.6[c]
22.0Ϯ1.3[c]
26.5Ϯ1.1[c]
47.9Ϯ0.4
{[CoIII(s-BQDI)2(Ph3P)S]ϩ·Imid} Ǟ
(6)
10.0
15.0
20.0
25.0
25.0
[CoIII(s-BQDI)2(Imid)S]ϩ ϩ Ph3P; k
10
50
90
130
(7)
∆H‡, kJ·molϪ1
∆S‡, J·KϪ1·molϪ1
∆V‡, cm3·molϪ1
Ϫ64Ϯ2
Ϫ11.1Ϯ0.2
[a]
Second-order rate constant calculated from the slope of plots of
[b]
kb versus [Ph3P].
[CoIII(s-BQDI)2(4-Me2Npy)]ϩ ϭ 1.25·10Ϫ4 ,
[Ph3P] ϭ 2.5·10Ϫ2 . [CoIII(s-BQDI)2(4-Me2Npy)]ϩ ϭ 1.25·10Ϫ4
, [Ph3P] ϭ 1.25·10Ϫ2 .
[c]
cm3·molϪ1, respectively. The values of kb were also meas-
ured directly as a function of pressure, with use of the dis-
placement of 4-Me2Npy by phosphane, and a plot of lnkb
versus pressure resulting in a ∆V‡ value of Ϫ11.1Ϯ0.2
cm3·molϪ1 (see Table 4). From the reported activation par-
ameters, we can conclude that the reverse reaction between
[CoIII(s-BQDI)2(Ph3P)]ϩ and 4-(Me)2Npy follows the same
associative reaction mechanism, in which the entering nu-
cleophile is coordinated to the CoIII centre in the six-coordi-
nate intermediate/transition state.
Figure 5. Plots of kobsd. versus [Imid] for the reaction between [CoIII
-
(s-BQDI)2(Ph3P)]ϩ and imidazole in acetonitrile as a function of
The forward rate constants for the reaction between [CoIII-
temperature; experimental conditions: [CoIII] ϭ 2.5·10Ϫ5 , tem-
(s-BQDI)2(Ph3P)] and 4-(Me)2Npy or imidazole are perature: 11.0 (A), 18.0 (B), 25.0 (C), 32.0 °C (D)
2976
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2003, 2972Ϫ2978

Ligand Substitution Reactions of Bis(o-benzosemiquinonediiminato)(triphenylphosphane)cobalt(III)
FULL PAPER
Table 5. Kinetic data for the reaction between [CoIII(s-BQDI)2(Ph3P)]ϩ and imidazole in acetonitrile as a function of temperature and
pressure
Pressure, MPa
0.1
Temp., °C
k·K ϫ 103, Ϫ1 sϪ1 [a]
K, Ϫ1 [b]
k ϫ 103, sϪ1 [b]
11.0
18.0
25.0
32.0
5.0
5.8Ϯ0.1[c]
10.4Ϯ0.5[c]
18Ϯ4[c]
0.9Ϯ0.1
1.0Ϯ0.2
1.1Ϯ0.1
1.3Ϯ0.1
5.8Ϯ0.5
10Ϯ2
19Ϯ1
26Ϯ2[c]
29Ϯ3
10
50
90
130
4.9Ϯ0.2[d]
6.1Ϯ0.3[d]
7.5Ϯ0.1[d]
9.6Ϯ0.1[d]
50Ϯ3[c]
∆H‡, kJ·molϪ1
∆S‡, J·molϪ1·KϪ1
∆V‡, cm3·molϪ1
Ϫ111Ϯ9[c]
Ϫ12.9Ϯ0.3[d]
[a]
[b]
k·K was calculated from kobsd./[Imid] at [Imid] ϭ 0.05 .
K and k were calculated from the nonlinear least-squaresfit by use of
Equation (7). [CoIII(s-BQDI)2(Ph3P)]ϩ ϭ 2.5·10Ϫ5 . [CoIII(s-BQDI)2(Ph3P)]ϩ ϭ 2.5·10Ϫ5 , [Imid] ϭ 0.025 .
[c]
[d]
The values of k and K as a function of temperature were cleophile, directly or through hydrogen bonding, can cause
calculated from a nonlinear fit of the measured rate con- a significant decrease in the nucleophilicity of the entering
stants and are summarised, along with the corresponding nucleophile. Specific solvation of the nucleophile by a protic
activation parameters, in Table 5. The values of ∆H‡ and solvent strongly stabilizes the nucleophile and lowers its re-
∆S‡ for the overall reaction between [CoIII(s- activity toward an electrophilic metal centre in an associat-
BQDI)2(Ph3P)]ϩ and imidazole in CH3CN were calculated ive reaction mode. In contrast, polar non-protic solvents
for k·K ϭ kobsd./[Imid] at [Imid] ϭ 0.05  (i.e., where the such as acetonitrile favour associative reaction modes.[8]
imidazole concentration dependence is still linear), and
In the case of the benzosemiquinonediimine chelate, de-
were found to be 50Ϯ3 kJ·molϪ1 and Ϫ111Ϯ9 localization of electron density from the unsaturated equa-
J·KϪ1·molϪ1, respectively. The pressure dependence of the torial ligand (s-BQDI) onto the CoIII centre could partially
reaction was studied at 0.025  imidazole and 5 °C, and the induce CoII character, which could in turn account for the
plot of lnkobsd. vs. pressure gave a good linear fit from which higher liability and the observed associative mechanism.
the value of ∆V‡ was calculated to be Ϫ12.9Ϯ0.3 The strongly negative activation entropies and volumes sup-
cm3·molϪ1. The results are summarized in Table 5. The re- port the associative nature of the substitution mechanism
ported activation parameters favour an associative inter- for these complexes. By way of comparison, in complexes
change (Ia) mechanism.
of ethylenediamine, the chelate is fully saturated and an Id
mechanism is more likely to account for the observed kine-
tic behaviour. It seems clear that the phenylenediamine li-
gand in the equatorial plane controls the ligand displace-
ment mechanism. The induced electron density at the metal
centre favours the unsaturated five-coordinate state and
lowers the energy barrier to bind the entering nucleophile
in the absence of a strongly coordinating solvent to favour
the associative reaction pathway. In the presence of a more
strongly coordinating solvent such as acetonitrile, the ab-
sence of a vacant coordination site will force the system to
follow an interchange substitution mechanism.
Mechanistic Interpretation
Most ligand substitution reactions of CoIII complexes,
such as those in vitamin B12, its derivatives and model com-
plexes, follow a D or Id mechanism.[9Ϫ17] Surprisingly, the
substitution reactions investigated in the current study be-
tween [CoIII(s-BQDI)2(Ph3P)]ϩ and imidazole or 4-Me2Npy
clearly follow an associative mechanism. Simandi et al.
studied the ligand substitution reactions of [CoIII(s-
CBQDI)2(Ph3As)]ϩ with pyridine, pyrrolidine, and imidaz-
ole in methanol and suggested an Ia mechanism, but no
activation parameters were reported.[3]
The second-order rate constant for the reaction between
[CoIII(s-BQDI)2(Ph3P)]ϩ and imidazole in methanol at 25
Experimental Section
°C (4.5·10Ϫ4

Ϫ1·sϪ1) is significantly smaller than that ob-
tained in the case of CH3CN as solvent (1.78·10Ϫ2 Ϫ1·sϪ1).
This suggests either that hydrogen bonding plays an import-
ant role in methanol as solvent by deactivating the attack-
ing nucleophile, or that the stronger donor properties of
acetonitrile labilizes the trans position and induces a rapid
displacement of phosphane.
The rates of associative substitution reactions can in
principle be affected by the nature of the solvent. Especially
in the case of protic solvents, partial protonation of the nu- in the literature.[5]
Materials: All chemicals were of p.a. grade and were used as re-
ceived. Imidazole was purchased from Aldrich and 4-dimethylami-
nopyridine
was
purchased
from
Across.
[Co(s-
BQDI)2(Ph3P)]ClO4·0.5C6H6 was prepared as described pre-
viously[5] by treatment of Co(ClO4)2·6H2O with o-phenylenediam-
ine and triphenylphosphane in the presence of oxygen. The com-
plex was characterised by elemental analysis and by IR and NMR
spectroscopy, and the results were in agreement with those reported
Eur. J. Inorg. Chem. 2003, 2972Ϫ2978
www.eurjic.org
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
2977

B. M. Aizoubi, M. S. A. Hamza, C. Dücker-Benfer, R. van Eldik
FULL PAPER
[4]
A. L. Balch, R. H. Holm, J. Am. Chem. Soc. 1966, 88,
Instrumentation: UV/Vis spectra and kinetic measurements were re-
corded in thermostatted cell compartments of Shimadzu UV-2101
and Cary 5 spectrophotometers. Measurements at high pressures
were carried out by use of the Shimadzu spectrophotometer
equipped with a home-made high-pressure unit.[18] For faster reac-
tions a high pressure, stopped-flow instrumentation was used.[19]
Kinetic data were analysed with the OLIS KINFIT programme.
All instruments used were thermostatted to the desired temperature
(Ϯ0.1 °C). Kinetic measurements were carried out under pseudo-
first order conditions (i.e., the ligand concentration was in at least
a tenfold excess).
5201Ϫ5209.
[5]
S. Nemeth, L. I. Simandi, G. Argay, A. Kalman, Inorg. Chim.
Acta 1989, 166, 31Ϫ33.
[6]
S. Nemeth, L. I. Simandi, Inorg. Chem. 1994, 33, 5964Ϫ5966.
[7]
M. S. A. Hamza, R. van Eldik, P. L. S. Harper, J. M. Pratt, E.
A. Betterton, Eur. J. Inorg. Chem. 2002, 580Ϫ583.
J. McMurry, Organic Chemistry, 2nd ed., Brooks/Cole, Cali-
[8]
fornia, USA, 1988.
B. M. Alzoubi, M. S. A. Hamza, C. Dücker-Benfer, R. van
[9]
Eldik, Eur. J. Inorg. Chem. 2002, 968Ϫ974.
[10]
G. Stochel, R. van Eldik, Inorg. Chem. 1990, 29, 2075Ϫ2077.
[11]
F. F. Prinsloo, M. Meier, R. van Eldik, Inorg. Chem. 1994,
33, 900Ϫ904.
M. Meier, R. van Eldik, Inorg. Chem. 1993, 32, 2635Ϫ2639.
F. F. Prinsloo, E. L. J. Breet, R. van Eldik, J. Chem. Soc., Dal-
[12]
[13]
Acknowledgments
ton Trans. 1995, 685Ϫ688.
[14]
The authors gratefully acknowledge financial support from the
Deutsche Forschungsgemeinschaft,
H. M. Marques, J. Chem. Soc., Dalton Trans. 1991, 339Ϫ341.
[15]
a
fellowship from the
H. M. Marques, J. C. Bradley, L. A. Campbell, J. Chem. Soc.,
Alexander von Humboldt foundation (M. S. A. H.), and sabbatical
leave from the Ain Shams University, Cairo (M. S. A. H.).
Dalton Trans. 1992, 13, 2019Ϫ2027.
[16]
H. M. Marques, J. C. Bradley, K. L. Brown, H. Brooks, J.
Chem. Soc., Dalton Trans. 1993, 23, 3475Ϫ3478.
M. S. A. Hamza, C. Dücker-Benfer, R. van Eldik, Inorg. Chem.
[17]
2000, 39, 3777Ϫ3783.
[1]
[18]
S. M. Peng, C. T. Chen, D. S. Liaw, C. I. Chen, Y. Wang, Inorg.
Chim. Acta 1985, 101, L31ϪL33.
L. F. Warren, Inorg. Chem. 1977, 16, 2814Ϫ2819.
L. I. Simandi, S. Nemeth, Inorg. Chim. Acta 1998, 270,
326Ϫ329.
M. Spitzer, F. Gartig, R. van Eldik, Rev. Sci. Instrum. 1988,
59, 2092Ϫ2093.
R. van Eldik, W. Gaede, S. Wieland, J. Kraft, M. Spitzer, D.
A. Palmer, Rev. Sci. Instrum. 1993, 64, 1355Ϫ1357.
[2]
[19]
[3]
Received February 23, 2003
2978
 2003 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjic.org
Eur. J. Inorg. Chem. 2003, 2972Ϫ2978