Kinetic studies of the reactions of pentacyanoferrate(III) complexes with L-ascorbic acid

Article abstract of DOI:10.1002/kin.20052

The reduction of Fe(CN)₅L^2- (L = pyridine, isonicotinamide, 4.4′-bipyridine) complexes by ascorbic acid has been subjected to a detailed kinetic study in the range of pH 1-7.5. The rate law of the reaction is interpreted as a rate determining reaction between Fe(III) complexes and the ascorbic acid in the form of H₂A(k₀). HA^-(k₁), and A^2- (k₂), depending on the pH of the solution, followed by a rapid scavenge of the ascorbic acid radicals by Fe(III) complex. With given Ka₁ and Ka₂. the rate constants are k₀ = 1.8, 7.0, and 4.4 M^-1 s ^-1 ; k₁ = 2.4 × 10³. 5.8 × 10 ³, and 5.3 × 10³ M^-1 s^-1 ; k₂ = 6.5 × 10⁸, 8.8 × 10⁸, and 7.9 × 10⁸ M^-1 s^-1 for L = py, isn, and bpy, respectively, at μ = 0.10 M HClO₄/LiClO₄, T=25°C. The kinetic results are compatible with the Marcus prediction.

Full text of DOI:10.1002/kin.20052

Kinetic Studies of the
Reactions of
Pentacyanoferrate(III)
Complexes with
L-Ascorbic Acid
LU-MING LIN, MING-HUEI LIEN, ANDREW YEH
Department of Chemistry, Tunghai Christian University Taichung, Taiwan R.O.C.
Received 12 March 2004; accepted 18 August 2004
DOI 10.1002/kin.20052
Published online in Wiley InterScience (www.interscience.wiley.com).
ABSTRACT: The reduction of Fe(CN)5L2 (L = pyridine, isonicotinamide, 4,4 -bipyridine) com-
plexes by ascorbic acid has been subjected to a detailed kinetic study in the range of pH 1–7.5.
The rate law of the reaction is interpreted as a rate determining reaction between Fe(III) com-
−
ꢀ
−
2−
plexes and the ascorbic acid in the form of H A(k ), HA (k ), and A (k ), depending on the
2
0
1
2
pH of the solution, followed by a rapid scavenge of the ascorbic acid radicals by Fe(III) complex.
−
1
−1
3
With given Ka and Ka , the rate constants are k = 1.8, 7.0, and 4.4 M
s
; k = 2.4 × 10 ,
1
2
0
1
s
3
3
−1 −1
8
8
8
−1 −1
5
.8 × 10 , and 5.3 × 10 M
s
; k = 6.5 × 10 , 8.8 × 10 , and 7.9 × 10 M
for L = py,
2
◦
isn, and bpy, respectively, at µ = 0.10 M HClO /LiClO , T = 25 C. The kinetic results are com-
4
4
patible with the Marcus prediction. ꢁC 2005 Wiley Periodicals, Inc. Int J Chem Kinet 37: 126–133,
005
2
INTRODUCTION
order to prevent the interference of the air even though
the solutions were prepared under nitrogenous atmo-
sphere. However, we have never measured the rates of
reductionoftheascorbicacidonthecomplexeswehave
studied. The reactions of ascorbic acid with several
A number of reactions of L-ascorbic acid with metal
complexes have been under intensive investigations
[1–17]. Most of these reactions proceeded through an
outer-sphere electron transfer process because the oxi-
dants used are substitution inert complexes and do not
possess vacant binding sites. In our previous studies
of the chemistry of pentacyanoferrate(II) complexes,
a small amount of ascorbic acid usually was added in
2−
−
−
Fe(CN)5L (L = CN [13] and NO [14]) complexes
2
have been investigated extensively recently, and the re-
sults were analyzed according to the Marcus theory for
outer-sphere electron transfer. Owing to our longstand-
3−/2−
ing interest in Fe(CN)5L
complexes and by taking
the advantage of the strong absorptions in the visible
region for the N-heterocyclic aromatic ligands of Fe(II)
complexes, we tend to extend our study to the reactions
of ascorbic acid and pentacyanoferrate(III) complexes
with ligands of pyridyl derivatives. In the present work,
Correspondence to: Andrew Yeh; e-mail: adyeh@mail.thu.
edu.tw.
Contract grant sponsor: National Science Council of the Republic
of China.
c 2005 Wiley Periodicals, Inc.
ꢁ

KINETIC STUDIES OF THE REACTIONS OF PENTACYANOFERRATE(III) COMPLEXES
127
we report the kinetic results of our study of the reduc-
HP 8453 spectrophotometer or on the Hi-Tech CU-61
stopped-flow apparatus, depending on the rate of the re-
action. Unless otherwise specified, measurements were
performed at an ionic strength of 0.10 M under pseudo-
first-order conditions with ascorbic acid in excess. Re-
actions were monitored by following the formation of
tion of Fe(CN)5L2 (L = pyridine (py), 4,4 -bipyridine
−
ꢀ
(bpy), and isonicotinamide (isn)) over the pH range
from 1 to 7.5. Theoretical calculations based on the
Marcus theory will also be discussed.
3
−
Fe(CN)5L complexes at their band maxima [23].
EXPERIMENTAL
Materials
There is a bathochromatic shift to ꢁmax = 513 nm
3
−1
−1
ꢀ
(
εmax = 4.70 × 10 M cm ) for the 4,4 -bpy com-
plex of Fe(II) in the acidic medium (0.01 to 0.10 M
+
ꢀ
ꢀ
[H ]) due to the protonation of the remote nitrogen of
4
,4 -Bipyridinium(4,4 -bipyridine)pentacyanoferrate
the ligand. The ln(A∞ − At ) vs time plots were linear
for more than three half lives, and the observed rate
constants were obtained from the slopes of the linear
least square fits of these plots.
(III) trihydrate was prepared according to the literature
method[18]. (BpyH2)[Fe(CN)5L] (L = py, isn) were
prepared similarly with some modifications. 0.13 g
(
0.4 m mole) of Na3[Fe(CN)5NH3] · 3H2O [19,20] was
added to 20 mL predeaerated solution containing 3
mmole of ligands. The reaction was allowed to proceed
for 20 min under argon atmosphere. At this point 0.95 g
RESULTS
(
4 m mole) of Na2S2O8 was added to the solution,
Stoichiometry
which was keptbubbling withargon foranother15 min.
to allow the oxidation to be complete. The solution was
acidified by adding dropwise ∼0.5 mL concentrated
2−
The stoichiometry for the reduction of Fe(CN)5L
complexes by the ascorbic acid was determined by a
spectrophotometric titration of the fixed concentrations
of the complex with the ascorbic acid. Typical plot for
the reduction of Fe(CN)5isn complex was shown in
Fig. 1.
ꢀ
HCl. 4,4 -Bpy was then added gradually with stirring
until a permanent turbidity was observed. After cooling
in the ice bath for 1 h, the desired solid precipitated out
and was washed with ethanol and ether. The products
were recrystallized from minimum amount of 0.10 M
2−
The break in the plot at [H2A]tot/[Fe(III)] ratio of
◦
HCl at 40–50 C. Yields: 100 mg (60%) for L = py and
∼
0.5 indicates that the ascorbic acid is a two elec-
8
0 mg (43%), for L = isn. Anal. for L = py, Calcd for
tron reductant as is expected, and the stoichiometry of
the reaction under consideration follows Eq. (1), where
H2A is the ascorbic acid and A is the dehydroascorbic
acid.
FeC20H15N8: C, 56.8; N, 26.5; H, 3.60. Found: C, 56.6;
N, 26.9; H, 3.70. For L = isn, Calcd for FeC21H16N9O:
C, 54.1; N, 27.0; H, 3.50. Found: C, 53.8; N, 26.9; H,
− −1
1
3
.80. The ꢀCN at 2110 cm (py), 2110 cm (isn),
−1
ꢀ
and 2118 cm (4,4 -bpy) of the IR spectra further
confirmed the formation of Fe(III) complexes [21]. All
other chemicals were of reagent grade and were used
withoutfurtherpurifications. Thedoublydeionizedwa-
ter was obtained by passing house-line distilled water
through an Osmonics P-12 water purification system.
Fe(CN)5L + H2A → 2Fe(CN)5L3− + A + 2H+
2−
2
(
1)
Instruments
UV-visible spectra were measured on a Hewlett
Packard HP 8453 spectrophotometer. Infrared spectra
were recorded on a Perkin-Elmer 1725 FT-IR spec-
trophotometer in KBr pellets. The pH measurements
were performed on an Orion 420A pH meter. Electro-
chemistry was carried out on a PAR model 273A po-
tentiostat/galvanostat system as described before [22].
Kinetic Measurements
The rates of reductions of Fe(CN)5L2 complexes
Figure 1 Stoichiometric plot for the reaction of L-ascorbic
−
2−
−4
acid with Fe(CN) isn complex. [Fe(III)] ꢂ 2 × 10 M,
5
at various pH values were carried out either on the
pH 5.0, µ = 0.10 LiClO4.

128
LIN, LIEN, AND YEH
Although reduction of the ascorbic acid is a two
electron process, it is well recognized that the rate-
determining step is the one electron reduction of the
ascorbic acid to ascorbic acid radical [5,7], and the
rate of the reaction is expected to be first order both
in the Fe(III) complex and in the ascorbic acid. This is
further evidenced by our kinetic observations. A nice
pseudo-first-order kinetics was observed in our sys-
tem with either Fe(III) complex or ascorbic acid as the
limiting reagent, as shown in Fig. 2, where a linear re-
lationship was obtained for the ln(A∞ − At) vs time
plots.
Figure 3 kobs vs [H2A] plots in acidic medium for
2
−
the reduction of Fe(CN)5 isn
HClO4/LiClO4. ꢀ 0.01 M, ꢁ 0.02 M, ꢂ 0.03 M,
0.05 M, 0.08 M, ∗ 0.09 M, × 0.10 M.
0.06 M, ꢄ 0.07 M,
complex. µ = 0.10 M
Kinetics for the Reduction
•
0.04 M, ꢃ
2
−
Complexes
of Fe(CN)5L
ꢀ
◦
The pseudo-first-order rate constants, kobs, increased
linearly with increasing ascorbic acid concentrations
+
The second-order rate constants at various acid concen-
trations, k, as obtained from the one-parameter linear
at constant [H ], as illustrated in Fig. 3 for the reduc-
2−
tion of Fe(CN)5isn complex. Therefore, at constant
+
least square fits of the kobs vs [H A] plots, are listed in
2
[
H ], the rate law according to Eq. (1) can be expressed
Tables I and II. The values of k decrease with increas-
by
+
ing acid concentrations, but not linearly. At [H ] =
+
ꢀ
−ꢁ
0
.01–0.10 M, k vs 1/[H ] plots are linear, as shown
−
d Fe(CN)5L2
ꢀ
−ꢁ
2k Fe(CN)5L2 [H2A] (2)
in Fig. 4. However, for pH 4.0–5.25, unlike that of the
acid solutions, linear relationships were observed be-
=
dt
+
tween 1/k and [H ], as shown in Fig. 5. Unfortunately,
we were unable to measure the rate of the reactions as
a function of pH at high pH values due to the reactivity
and the sensitivity of the ascorbic acid toward air in
basic solution. The highest pH where the rate can be
measured in this system is at pH 7.5, and the results are
also shown in Table II.
Table I Specific Rate Constants of the Reductions of
2
−
Fe(CN) L Complexes by Ascorbic Acid in Acid
Medium
5
a
−
1
−1
s )
k (M
L
+
[
H ] (M)
Py
Isn
Bpy
0
0
0
0
0
0
0
0
0
0
.01
.02
.03
.04
.05
.06
.07
.08
.09
.10
8.81 ± 0.09
5.30 ± 0.08
3.92 ± 0.09
3.53 ± 0.08
3.14 ± 0.05
3.07 ± 0.09
2.92 ± 0.01
2.71 ± 0.08
2.62 ± 0.09
2.33 ± 0.05
32.6 ± 0.4
20.3 ± 0.4
16.4 ± 0.2
13.9 ± 0.6
11.6 ± 0.2
11.3 ± 0.7
10.2 ± 0.2
10.2 ± 0.7
9.6 ± 0.3
63.3 ± 0.2
33.8 ± 0.9
24.5 ± 0.5
18.7 ± 0.6
15.1 ± 0.4
13.4 ± 0.4
12.9 ± 0.6
12.0 ± 0.3
11.1 ± 0.4
11.2 ± 0.2
Figure 2 ln(A∞ − At) vs time plots for the reduction
2
−
of Fe(CN)5isn
complex at pH 5.0, µ = 1.0 M LiClO4.
9.8 ± 0.6
−4
−5
(
[
a) [Fe(III)] = 5.17 × 10 M, [H2A] = 5.10 × 10 M; (b)
Fe(III)] = 4.88 × 10− M, [H2A] = 5.56 × 10 M.
4
−3
a
µ = 0.10 M HClO
/LiClO
, T = 25 C.
◦
4
4

KINETIC STUDIES OF THE REACTIONS OF PENTACYANOFERRATE(III) COMPLEXES
129
Table II Specific Rate Constants of the Reductions of Fe(CN) L Complexes by Ascorbic Acid at pH 4 to 7.5a
2
−
5
−
1
−1
s )
k (M
Ligand
Isn
Ph
Py
Bpy
2
3
3
3
3
4
3
3
3
3
3
5
3
4
4
4
5
5
7
.00
(6.68 ± 0.08) × 10
(1.04 ± 0.02) × 10
(1.69 ± 0.03) × 10
(2.31 ± 0.03) × 10
(2.68 ± 0.03) × 10
(9.90 ± 0.05) × 10
(2.19 ± 0.03) × 10
(3.01 ± 0.01) × 10
(4.45 ± 0.08) × 10
(5.35 ± 0.07) × 10
(5.49 ± 0.07) × 10
(1.33 ± 0.08) × 10
(2.41 ± 0.03) × 10
(3.47 ± 0.05) × 10
(3.99 ± 0.04) × 10
(4.60 ± 0.05) × 10
(4.90 ± 0.06) × 10
(1.19 ± 0.05) × 10
3
3
3
3
5
.25
.50
.00
.25
.52
a
◦
µ = 0.10 M LiClO
4
, T = 25 C.
k1
Cation Ion Effect
Fe(CN) L + HA −→ Fe(CN) L + HA·
2−
−
3−
(6)
(7)
5
5
The reduction of Fe(CN)5isn2 complex has been car-
ried out using various electrolytes at pH 5.00 and
µ = 0.10 M. The values k, which are obtained from
the slopes of kobs vs [H2A] plots, are listed in Table III.
−
k
2
Fe(CN) L + A2− −→ Fe(CN) L + A·−
2
−
3−
5
5
fast
Fe(CN) L + H2A·+ −→ Fe(CN) L + 2H + A
2−
3−
+
5
5
(
8)
fast
2
−
·
3−
+
Fe(CN) L + HA −→ Fe(CN) L + H + A
5
5
DISCUSSION
(
9)
fast
The mechanism comprising Eqs. (3)–(10), together
with the elaborations which will be introduced, ac-
counts for our results.
Fe(CN) L + A·− −→ Fe(CN) L + A
2−
3−
(10)
5
5
According to this mechanism, the expression of k in
Eq. (2) will be
Ka1
ꢁ
H2A ꢀ H + HA−
+
(3)
(4)
+
2
+
k0[H ] + k1 Ka1[H ] + k2 Ka1 Ka2
Ka2
−
+
2−
HA
ꢀ H + A
k =
(11)
ꢁ
+
2
+
[
H ] + Ka1[H ] + Ka1 Ka2
k0
Fe(CN) L + H2A −→ Fe(CN) L + H2A·+ (5)
2
−
3−
5
5
+
+
Figure 4
Fe(CN)5L
k
vs 1/[H] ] plot for the reduction of
Figure 5 1/k vs [H ] plots for the reduction of
2
−
+
2−
complexes at [H ] = 0.01–0.10 M HClO4
Fe(CN)5L
isn).
complexes at pH 4.0–5.25 (•
py, ꢁ bpy, ꢅ
(
•
py, ꢁ bpy, ꢅ isn).

130
LIN, LIEN, AND YEH
Table III Kinetic Effect of the Alkali Cation for the
Table IV. Although in reasonable agreement with each
2−
a
Reduction of Fe(CN) isn Complex
other, the results of k obtained in this pH range are
5
1
expected to be more reliable and more representative
−
3
−1 −1
M
10 k (M s )
for the values of k1 as compared to those obtained in
the acidic medium due to the sensitivity of the Fe(II)
complexes to air at pH < 4 [24]. The slightly slower
rates obtained in the acid solution may arise from the
interference of the oxygen.
Li
Na
K
6.9 ± 0.1
5.2 ± 0.1
6.1 ± 0.1
a
pH 5.00, µ = 0.10 M Cl.
+
2
+
At pH 7.52, k0[H ] ꢄ k1 Ka1[H ] + k2 Ka1 Ka2,
+
2
+
+
+
and [H ] + Ka1 Ka2 ꢄ Ka1[H ], Eq. (11) therefore
At [H ] = 0.01–0.10 M, since [H ] ꢃ Ka1, Ka2,
can be reduced to
Eq. (11) can be simplified as
k2 Ka2
+
k = k +
(15)
k0[H ] + k1 Ka1
1
+
[
H ]
k =
(12)
+
H ]
[
−
12
Taking the value of Ka2 = 4.57 × 10
(pKa2 =
According to Eq. (12), we expect linear plots between
k and 1/[H ], as is observed in Fig. 4. Taking the value
1
1.34 [4]), and k1 values measured, k2 can thus be cal-
+
culated and the results also are shown in Table IV.
−5
2
−
of Ka1 = 8.32 × 10 [7], the values of k0 and k1,
Since Fe(CN)5L complexes are substitution inert
and therefore the reductions of these complexes can be
viewed as outer-sphere electron transfer reactions that
obtained from the nonlinear least square fits of k vs
+
[
H ] according to Eq. (12), are listed in Table IV.
Under the conditions pH 4.0–5.25, Ka1 Ka2 ꢄ
obey Eqs. (16)–(20) [5,7]
+
2
+
[
H ] + Ka1[H ], and Eq. (11) will be reduced to
ꢂ
K12 = k11k22 K12 f12W12
(16)
+
2
+
k0[H ] + k1 Ka1[H ] + k2 Ka1 Ka2
2
k =
(13)
[ln K12 + (w12 − w21)/RT ]
+
2
+
[
H ] + Ka1[H ]
ln f12 =
22
4[ln(k11k22)/10 + (w11 + w22)/RT ]
However, since there is a linear relationship between
(17)
+
1
/k vs [H ] plots as shown in Fig. 5, we would ex-
+
2
W12 = exp[−(w12 + w21 − w11 − w )/2RT ]
pect that in this pH range k0[H ] + k2 Ka1 Ka2 ꢄ
22
+
k1 Ka1[H ]. Therefore Eq. (13) can further be reduced
(18)
to
zi z j e2
k1 Ka1
wi j =
β =
ꢃ
ꢄ
(19)
1
2
k = [
(14)
DSai j 1 + ꢂai j µ
+
H ] + Ka1
ꢅ
ꢆ
2
1
/2
+
Nonlinear least square fits of k against [H ], accord-
ing to Eq. (14), yielded the values of k1 as shown in
8Nπe
(20)
1000D kT
s
Table IV Rate Constants for the Reductions of Fe(CN)5L2 Complexes by Ascorbic Acida
−
k0 (M 1
−
s
−1
)
k1 (M
−1 −1
s
)
k2 (M
Meas
−1 −1
s
)
Fe(CN)5L2
L
−
/
E0
(V vs NHE)
Meas
Calcd
Meas
Calcd
Calcd
3 b
3 b
2 d
3
8 f
8
8
8
py
0.44
0.50
1.79 ± 0.05 4.78 × 10−
(8.4 ± 0.1) × 10
1.88 × 10 (6.54 ± 0.09) × 10
e
3.06 × 10
5.95 × 10
5.42 × 10
3
(
2.4 ± 0.2) × 10
−
c
3 d
3 e
3
8 f
isn
7.0 ± 0.26 2.65 × 10
(3.1 ± 0.1) × 10
7.57 × 10 (8.76 ± 0.09) × 10
(
8.53)
(5.7 ± 0.2) × 10
3
e
3
8 f
bpy
bpyH
0.49
0.54
(5.3 ± 0.1) × 10
(7.1 ± 0.1) × 10
6.17 × 10 (7.85 ± 0.09) × 10
+
−2 b
3 d
4
4.4 ± 0.3
3.34 × 10
1.14 × 10
c
(
8.42)
a
µ = 0.10 M HClO
4 4
/LiClO .
b
c
d
e
f
0
·+
·+
Using E (H
2
A
A
/H
2
A) = 1.17 V.
0
Using E (H
2
/H
2
A) = 0.955 V, see text.
Calculated according to Eq. (12).
Calculated according to Eq. (14).
Calculated according to Eq. (15).

KINETIC STUDIES OF THE REACTIONS OF PENTACYANOFERRATE(III) COMPLEXES
131
where k11 and k22 are self-exchange rate constants for
the ascorbic acid and the transition metal complexes,
respectively; k12 and K12 are the rate constant and
the equilibrium constant for the cross reaction, zi and
z j are charges of the reacting species, the values of
−
8
−8
radii are 3.5 × 10 cm and 5.0 × 10 cm for ascor-
2−
bic acid [5] and Fe(CN)5L complexes (assuming
them to be equal to that of the Fe(CN)5py2 complex
18]); D is the static dielectric constant of water, k
is the Boltzmann constant. The reduction potentials of
−
[
2
−/3−
Fe(CN)5L
complexes are listed in Table IV. Since
the self-exchange rate constants of the substituted pyri-
dine complexes of pentacyanoferrate(II,III) are rather
unaffected by the nature of the ligands [25], we may
assume that the self-exchange rate constant for the
Figure 6 log k1 vs log K plots for the reduction of
Fe(CN)5L complexes by ascorbic acid.
2
−
2
−/3−
5
−1 −1
Fe(CN)5L
couple to be 7.0 × 10 M
s , taken
to be the same as pyridine complex [23]. Taking the re-
duction potentials and self-exchange rate constants of
2
−
3
−1 −1
·+
like the reductions of Fe(CN)5L complexes, none
ascorbicacidas1.17V, 2 × 10 M
s
(H2A /H2A),
3
+
5
−1 −1
·
−
of the Ru(NH3)5L complexes exhibited a detectable
contribution of an acid independent path (i.e. k0) be-
cause of the high-reactivity ratio of k1/k0. Taking
0
2
.71 V, 1.6 × 10 M
s
(HA /HA ), and 0.015 V,
5
−1 −1 ·− 2−
× 10 M
s
(A /A ) [5,14], the rate constants
k0, k1, and k2 can be calculated with the aid of
Eqs. (16)–(20). The results are also listed in Ta-
ble IV. As shown in Table IV, values of k1 and
k2 agree reasonably well with our measured values.
However, the calculated k0 are smaller than the ex-
perimental values by nearly two orders of magni-
tude. Since various values of reduction potential for
3
+
the reduction of Ru(NH3)5py complex, for exam-
ple, the calculated values of k0 and k1 according to the
0
·+
Marcus theory by using E (H2A /H2A) = 0.96 V are
−
2
2
−1 −1
3
.0 × 10 and 2.7 × 10 M
s , respectively. This
3
1
gives k1/k0 = 9.0 × 10 . However, for the reaction of
2
−
Fe(CN)5py complex with ascorbic acid the reactivity
3
·
+
ratio is 1.1 × 10 according to the Marcus theory calcu-
the H2A /H2A couple have been previously reported
3,4,13,26], we have used our measured k0 in the re-
lation. This is smaller than that of the Ru(NH3)5py3
+
[
2−
complex by a factor 8. The difference in the rate ra-
tio may arise predominantly from the charge effect of
the reactants. For an outer-sphere electron transfer re-
action, the reactants are brought together in close ap-
proach to form an outer-sphere precursor before the
electron transfer takes place. The ion pair formation
process is diffusion controlled and the ion pair forma-
tion constant can be estimated from Eq. (21) [14].
duction of Fe(CN)5py complex to predict the re-
·
+
duction potential for the H2A /H2A couple, and the
value is 0.955 V, which is in agreement with the value
of 0.95 V, estimated by Funahashi et al. [13]. Using
this reduction potential, the calculated k0 for isonicoti-
ꢀ
namide and 4,4 -bipyridine complexes agree fairly well
with our experimental results, as also are indicated in
Table IV.
The values of k1 measured in the present system
are slightly greater than that of the reduction of the
ꢃ
ꢄ
4
πNa3
−wi j
QIP =
exp
(21)
3−
2
related system of Fe(CN)5NO complex (6.45 × 10
3
RT
2
−1
−1
M
s
) [14]. Since the rates of outer-sphere electron
When the reaction was carried out in the acidic
medium where the neutral species H2A is the reduc-
tant, the ion pair formation constant is rather unaf-
fected by the charge of the complex. However, when
the ascorbic acid is in the HA form, the reaction with
Ru(NH3)5py complex corresponds to bringing to-
gether species of opposite charges, which will be more
transfer reaction are rather insensitive to the electro-
static effect for reactants of like charges, the difference
may arise from, at least for the large part, the differ-
ence in the equilibrium barrier of the reaction. If we
plot log k1 vs log K for the reactions after the correc-
tion for the charge effect, they all fall on a straight line
with a slope of 0.48 ± 0.04, as shown in Fig. 6, which
is in agreement with the value of 0.50, predicted by the
Marcus–Hush theory for outer-sphere electron transfer
reaction [27].
−
3
+
2
−
−
favorable than that of the Fe(CN)5py –HA system,
where both reactants carry the charges of the same
The reductions of Ru(NH3)5L3 complexes by
+
1
◦
·+
7
If E (H
2
A
/H
2
A) = 1.17 V is used, the ratio will be 1.3 × 10 ,
ascorbic acid have been studied extensively [7]. Un-
as is reported in [7].

132
LIN, LIEN, AND YEH
sign. The values of QIP as calculated from Eq. (21) are
potassium is used as the cation because of the size of
the ion. This will stabilize the ion triplet thus formed
due to the large separation of the anions. The second
stage of reduction further supports this argument. The
−1
2−
0
.633 and 5.20 M for the reductions of Fe(CN)5py
3+
and Ru(NH3)5py systems, respectively. The differ-
ence in QIP between two complexes by a factor of 8
agrees with that of the k1/k0 reactivity ratio. If we cor-
rect k1 for the contribution of QIP, the reactivity ra-
5
−
rate of reduction of [Fe2(CN)10] in the potassium ion
medium is ∼7-fold greater than that measured in the
lithium ion medium, while the reduction in the first
stage, where the charge of the complex is −4, the dif-
ference in rates between two ions is less than a factor
of 3. In the present system, the charge of the complex
in −2, and the charge effect will not affect the stability
of the ion triplet when it is formed. The rate of the re-
duction therefore is rather independent of the cations
used.
3
3
2−
tio will be 2.0 × 10 and 2.5 × 10 for Fe(CN)5py
and Ru(NH3)5py3 complexes, which agree with each
+
other fairly well.
Elding et al. [16,17] in their study of the reduc-
tion reaction of the haloam(m)ine platinum(IV) com-
plexes by ascorbate has found that the reactivity of the
2−
ascorbate ion, A , is greater than that of the hydrogen
−
ascorbate, HA , by about seven orders of magnitude.
5
However, in our system k2/k1 ratio is only 10 . The dif-
ference may arise from the difference in reaction mech-
anisms. In the reduction of Fe(CN)5L2− complex, the
SUPPLEMENTARY MATERIAL
electron transfer process is outer sphere, and the higher
2
−
−
Tables S1–S3, listing the observed pseudo-first-order
rate constants for the ascorbic acid reduction of
rates for A than for HA are attributable, at least in
large part, to the much larger driving force for the re-
2
−
2
−
Fe(CN) L complexes at various pH, and Table S4,
action with A species. Since the intrinsic reactivity
for the two couples is much the same, as is evidenced
from the values of kex, the ratio therefore is expected to
vary approximately with the square root of the ratio of
5
listing the observed pseudo-first-order rate constants
2
−
for the reduction of Fe(CN)5isn complex with vari-
ous alkali metal ions at µ = 0.10 M and pH 5.00, are
available at http://www.interscience.wiley.com/jpages/
0538-8066/suppmat.
equilibrium quotients according to the Marcus theory
√
5
−
(
Eq. (16)). The ratio of K of 7 × 10 between HA –
2
−
Fe(III)andA –Fe(III)systemsagreesreasonablywell
with our observation. In the reduction of Pt(IV) com-
plexes, the electron is transferred intramolecularly due
to the formation of a bridge-activated binuclear com-
plex between the reductant and a coordinated chloride
of the complex. In this case, the strong electron with-
drawing power of the proton in hydrogen ascorbate will
make the transfer of the electron through the bridge to
the metal center much more difficult than the ascorbate
ion in addition to the thermal equilibrium barrier factor
which has been favored the ascorbate ion already.
Beckford et al. in their study of the reduction of
The support of this work by the National Science Council of
the Republic of China is gratefully acknowledged.
BIBLIOGRAPHY
1
2
3
4
. Pelizzetti, E.; Mentastic, E.; Pramauro, E. Inorg Chem
976, 15, 2898.
. Rickman, R. A.; Sorensen, R. L.; Watkins, K. O.; Davies,
G. Inorg Chem 1977, 16, 1570.
1
. Kimura, M.; Yamamoto, M. J Chem Soc, Dalton Trans
4
−
binuclear complex by ascorbic acid
[Fe2(CN)10]
1
982, 423.
. Williams, N. H.; Yandell, J. K. Aust J Chem 1982, 35,
133.
has found a significant cation effect with a reactiv-
+
+
+
ity sequence K > Na > Li [28]. They attributed
1
this cation effect to the formation of an ion triplet,
5. Macartney, H.; Sutin, N. Inorg Chim Acta 1983, 74,
221.
6. Martinez, P.; Zuluaga, J.; Uribe, D.; van Eldik, R. Inorg
Chim Acta 1987, 136, 11.
4
−
+
−
[
Fe2(CN)10] · M · A , with cation facilitating the
association of the anions. However, we do not ob-
serve obvious difference in the rates of reduction in the
present system, at least under our experimental con-
ditions, when various alkali salts are used, as shown
in Table III. The contradiction between the two sys-
tems may arise predominantly from the difference in
7
8
. Akhtar, M. J.; Haim, A. Inorg Chem 1988, 27, 1608.
. Hoddenbagh, J. M. A.; Macartney, D. H. J Chem Soc,
Dalton Trans 1990, 615.
9. Xu, J.; Jordan, R. B. Inorg Chem 1990, 29, 4180.
1
0. B a¨ nsch, B.; Martinez, P.; Uribe, D.; Zuluaga, J.;
van Eldik, R. Inorg Chem 1991, 30, 4555.
the charge of the complex. For Fe2(CN)4 , the complex
−
10
is highly negatively charged even after the formation of
an ion pair with the metal ion, and a strong electrostatic
repulsion may still be expected in the ion triplet. How-
ever, this repulsion might be greatly reduced when the
1
1. Saha, S. K.; Ghosh, M. C.; Gould, E. S. Inorg Chem
1992, 31, 5439.
12. B a¨ nsch, B.; van Eldik, R.; Martinez, P. Inorg Chim Acta
1992, 201, 75.

KINETIC STUDIES OF THE REACTIONS OF PENTACYANOFERRATE(III) COMPLEXES
133
1
1
1
1
1
3. Kagayama, N.; Sekiguchi, M.; Inada, Y.; Takagi, H. D.;
Funahashi, S. Inorg Chem 1994, 33, 1881.
4. Wanat, A.; van Eldik, R.; Stochel, G. J Chem Soc, Dalton
Trans 1998, 2497.
5. Stolarczyk, K.; Bilewicz, R.; Siegfried, L.; Kaden, T.
Inorg Chim Acta 2003, 348, 129.
6. Lemma, K.; Sargeson, A. M.; Elding, L. I. J Chem Soc,
Dalton Trans 2000, 1167.
7. Lemma, K.; House, D. A.; Retta, N.; Elding, L. I. Inorg
Chim Acta 2002, 331, 98.
20. Jwo, J. J.; Haim, A. J Am Chem Soc 1976, 98, 1172.
21. Dows, D. A.; Wilmarth, W. K.; Haim, A. J Inorg Nucl
Chem 1961, 21, 33.
22. Chen, M. H.; Lee, S.; Liu, S.; Yeh, A. Inorg Chem
1996, 35, 2627.
23. Toma, H. E.; Malin, J. M. Inorg Chem 1973, 12, 1039.
24. Yeh, A.; Haim, A. J Am Chem Soc 1985, 107, 369.
25. Hrepic, N. V.; Malin, J. M. Inorg Chem 1979, 18,
409.
26. Mark, J.; Hoddenbagh, A.; Macartney, D. H. J Chem
Soc, Dalton Trans 1990, 615.
1
8. Phillips, J.; Haim, A. Inorg Chem 1980, 19, 1616.
9. Brauer, G. Handbook of Preparative Inorganic Chem-
istry, 2nd ed.; Academic Press: New York, 1965; Vol. II,
p. 1511.
1
27. Sutin, N. Prog Inorg Chem 1983, 30, 441.
28. Beckford, F. A.; Dasgupta, T. P.; Stedman, G. J Chem
Soc, Dalton Trans 1995, 2561.