The solid-state electrochemistry of metal octacyanomolybdates, octacyanotungstates, and hexacyanoferrates explained on the basis of dissolution and reprecipitation reactions, lattice structures, and crystallinities

Article abstract of DOI:10.1021/ic9909330

The electrochemical behavior of solid microparticles of metal (Ag⁺, Cd²⁺, Co²⁺, Cr²⁺, Cu²⁺, Fe²⁺, Mn²⁺, Ni²⁺, Pb²⁺, and Zn²⁺) octacyanomolybdates, octacyanotungstates, and hexacyanoferrates has been studied by voltammetry, electrochemical quartz crystal microbalance, and microscopic diffuse reflectance spectroelectrochemical measurements. The solid microparticles have been immobilized on the surface of graphite electrodes prior to the electrochemical measurements. A comparative study of the cyclic oxidation and reduction of these compounds in the presence of potassium ions revealed that any interpretation of the electrochemistry requires the solubility equilibria of the reduced compounds to be taken into account, such as in the case of the silver salts {Ag₃K[X]} and {Ag₄[X]} (with X = Fe(II)(CN)₆/^4-, M(IV)(CN)₈/^4- (M = Mo, W)). Because {Ag₄[X]} has a lower solubility than {Ag₃K[X]}, the electrochemistry is accompanied by a conversion of solid {Ag₃K[X]} into solid {Ag₄[X]}. Two distinct voltammetric signal systems are generated by these two compounds according to {Ag₃K[X]} ? {Ag₃-[X]} + K⁺ + e^- and {Ag₄[X]} ? {Ag₃[X]} + Ag⁺ + e^-. When silver ions are present in the solution adjacent to the microparticles, the silver octacyanometalates and silver hexacyanoferrate show a chemically reversible and very stable voltammetric behavior. Despite the fact that the electrochemistry is based upon a single-electron/single-ion transfer reaction ({Ag₄[X]} ? {Ag₃[X]} + Ag⁺ + e^-), more than one electrochemical signal is observed because of the simultaneous presence of amorphous and crystalline particles. This study shows that the interplay of solubility equilibria and electrochemical equilibria is generally observed for the other metal octacyanomolybdates, octacyanotungstates, and hexacyanoferrates as well.

Full text of DOI:10.1021/ic9909330

1
006
Inorg. Chem. 2000, 39, 1006-1015
The Solid-State Electrochemistry of Metal Octacyanomolybdates, Octacyanotungstates, and
Hexacyanoferrates Explained on the Basis of Dissolution and Reprecipitation Reactions,
Lattice Structures, and Crystallinities
†
Uwe Schr o1 der and Fritz Scholz*
Institut f u¨ r Chemie und Biochemie, Ernst-Moritz-Arndt-Universit a¨ t, Soldmannstrasse 23,
17489 Greifswald, Germany
ReceiVed August 4, 1999
+
2+
2+
2+
2+
2+
2+
2+
The electrochemical behavior of solid microparticles of metal (Ag , Cd , Co , Cr , Cu , Fe , Mn , Ni ,
2+
2+
Pb , and Zn ) octacyanomolybdates, octacyanotungstates, and hexacyanoferrates has been studied by voltammetry,
electrochemical quartz crystal microbalance, and microscopic diffuse reflectance spectroelectrochemical measure-
ments. The solid microparticles have been immobilized on the surface of graphite electrodes prior to the
electrochemical measurements. A comparative study of the cyclic oxidation and reduction of these compounds in
the presence of potassium ions revealed that any interpretation of the electrochemistry requires the solubility
equilibria of the reduced compounds to be taken into account , such as in the case of the silver salts {Ag3K[X]}
II
4-
IV
4-
and {Ag4[X]} (with X ) Fe (CN) , M (CN) (M ) Mo, W)). Because {Ag4[X]} has a lower solubility than
Ag3K[X]}, the electrochemistry is accompanied by a conversion of solid {Ag3K[X]} into solid {Ag4[X]}. Two
6
8
{
distinct voltammetric signal systems are generated by these two compounds according to {Ag3K[X]} a {Ag3-
+
-
+
-
[
X]} + K + e and {Ag4[X]} a {Ag3[X]} + Ag + e . When silver ions are present in the solution adjacent
to the microparticles, the silver octacyanometalates and silver hexacyanoferrate show a chemically reversible and
very stable voltammetric behavior. Despite the fact that the electrochemistry is based upon a single-electron/
+
-
single-ion transfer reaction ({Ag4[X]} a {Ag3[X]} + Ag + e ), more than one electrochemical signal is observed
because of the simultaneous presence of amorphous and crystalline particles. This study shows that the interplay
of solubility equilibria and electrochemical equilibria is generally observed for the other metal octacyanomolybdates,
octacyanotungstates, and hexacyanoferrates as well.
Introduction
could not be explained thus far, such as, for example, the origin
of some voltammteric signals, their shape and stability. How-
Solid-state electrochemical reactions, which are not confined
to only the surface, are accompanied by ion-transfer processes
between the redox-active solid phase and the adjacent electrolyte
solution.1 The reason for this ion transfer is to maintain the
electroneutrality found inside the solid phase. Such ion insertion/
expulsion processes require a scaffold, which forms a tunnel or
layer structure to enable the ion transport inside the solid phase.
All very sparingly soluble metal octacyanomolybdates, octa-
cyanotungstates, and hexacyanoferrates, considered in this work,
possess a so-called zeolitic structure with channels and acces-
sible cavities for ion transport and housing. This is a prerequisite
for their insertion electrochemistry with a coupled electron and
ion transfer.
ever, a clear understanding of all electrochemical reactions is
necessary as several metal cyanometalates find applications in
sensors, electrochromic displays, electrocatalysis, and so forth.
It is very interesting to compare the electrochemistry of solid
metal octacyanomolybdates (ocm), octacyanotungstates (oct),
and hexacyanoferrates (hcf), as such study reveals common
features that have yet to be described. The electrochemistry of
the dissolVed anions comprises a single electron-transfer step:
,2
II
4-
III
3-
-
IV
4-
[
Fe (CN)6] a [Fe (CN)6] + e and [M (CN)8]
a
V
3-
-
[
M (CN)8] + e . Although the solution chemistry and elec-
trochemistry of the hexacyanoferrates and octacyanometalates
have been the subjects of thorough investigations,1
0,11
little is
known about the physical, chemical, and electrochemical prop-
erties of solid octacyanometalates. Cox et al. described a
chemical irreversibility of the cyclic oxidation and reduction
Although the electrochemistry of metal hexacyanoferrates has
3-9,15,30
been extensively studied,
several experimental results
*
To whom correspondence should be addressed. Tel: +49(03834)-
8
64450. Fax: +49(03834)864451. E-mail: fscholz@mail.uni-greifswald.de.
(7) Feldmann, B. J.; Murray, R. W. Inorg. Chem. 1987, 26, 1702.
(8) Kulesza, P. J.; Malik, M. A.; Berrettoni, M.; Giorgetti, M.; Zamponi,
S.; Schmidt, R.; Marassi, R. J. Phys. Chem. B 1998, 102, 1870-1876.
(9) Kulesza, P. J.; Faszynska, M. J. Electroanal. Chem. 1988, 252, 461-
466.
†
Current address: Theoretical and Physical Chemistry Department,
Oxford University, South Parks Road, OX1 3QZ, UK.
1) Sch o¨ llhorn, R. Angew. Chem. 1980, 92, 1015; Sch o¨ llhorn, R. Angew.
(
(
(
(
(
(
Chem., Int. Ed. Engl. 1980, 19, 983.
2) Bruce, P. G., Ed. Solid State Electrochemistry; Cambridge University
Press: Cambridge, England, 1995.
(10) Samotus, A.; Szklarzewicz, J. Coord. Chem. ReV. 1993, 125, 63-74.
(11) Oyama, N.; Ohsaka, T.; Yamamoto, N.; Matsui, J.; Hatozaki O. J.
Electroanal. Chem. 1989, 265, 297-304.
3) Dostal, A.; Meyer, B.; Scholz, F.; Schr o¨ der, U.; Bond, A. M.; Marken,
F.; Shaw, Sh. J. J. Phys. Chem. 1995, 99, 9, 2096-2103.
4) Kahlert, H.; Retter, U.; Lohse, H.; Siegler, K.; Scholz, F. J. Phys.
Chem. B 1998, 44, 8757-8765.
(12) Cox, J. Anal. Chem. 1985, 57, 2739-2740.
(13) (a) Scholz, F.; Meyer, B. Voltammetry of Solid Microparticles
Immobilized on Electrode Surfaces. In Electroanalytical Chemistry;
Bard, A., Rubinstein, I., Eds.; Marcel Dekker: New York, 1998; Vol.
20. (b) Voltammetry of Microparticles. www.iic.cas.cz/∼grygar/
AbrSV.html.
5) Dostal, A.; Schr o¨ der, U.; Scholz, F. Inorg. Chem. 1995, 34, 1711-
1717.
6) Xidis, A.; Neff, V. D. J. Electrochem. Soc. 1991, 138, 3637.
1
0.1021/ic9909330 CCC: $19.00 © 2000 American Chemical Society
Published on Web 02/18/2000

Electrochemistry of Cyanometalates
Inorganic Chemistry, Vol. 39, No. 5, 2000 1007
of films of silver octacyanomolybdate, which appeared to be a
peculiarity of silver octacyanomolybdate and tungstate.
Mikroskopie und Systeme GmbH, Germany) with a 100 W halogen
incandescent lamp. A 20-fold magnification was used for the spectro-
electrochemical measurements. Two crossed linear polarizing filters
minimized the specular reflectance. The spectrometer, which was
coupled via fiber optics, was a transputer-integrated diode array system
1
2
Here, it will be shown that the electrochemistry of these
complex cyanides is strongly affected by the solubility equilibria
of the involved compounds. It will be demonstrated that the
complex and chemically irreversible electrochemistry of silver
octacyanomolybdate in the presence of potassium ions, as
observed by Cox et al.,12 is caused by the competition of the
(TIDAS) (J&M Analytische Mess- und Regeltechnik GmbH, Ahlen,
Germany) with a spectral range from 320 to 950 nm. It was interfaced
to a Pentium personal computer.
The detailed description of the experimental setup is given in a
IV
IV
solubility equilibria of Ag4[Mo (CN)8] and Ag3K[Mo (CN)8].
This gives rise to an internal reprecipitation process accompanied
by a partial dissolution. Under certain conditions, analogous
reactions can also be observed in case of silver hexacyanoferrate
and several other solid-metal hexacyanoferrates and octacya-
nometalates. However, these reactions differ to some extent
because of the different structures of the solid compounds.
For this study, the Voltammetry of microcrystals proved to
be the most useful electroanalytical technique because it enabled
us to investigate a wide number of synthesized compounds that
could easily be characterized using the necessary analytical
15
previous work.
Preparation. All chemical operations were carried out using
IV
bidistilled water and analytical or reagent grade chemicals. K
(CN) ] was synthesized by the procedure described by Furman et al.
Anal. Calcd: C, 19.33; N, 22.55; H, 0.80. Found: C, 18.81; N, 21.76;
4
[Mo -
1
6
8
V
H, 0.85. K
[
3
[Mo (CN)
8
8 4
] was obtained by oxidizing a solution of K -
IV
Mo (CN)
was produced by the procedure described by Brauer Anal. Calcd: C,
6.96; N, 19.78; H, 0.35. Found: C, 16.91; N, 19.33; H, 0.30. This is
] with manganate(VII). Potassium octacyanotungstate(IV)
17
1
IV
in agreement with the monohydrate K
4
[W (CN)
8 2
]‚H O.
Potassium hexacyanoferrate(II) and -(III) (Laborchemie Apolda,
Germany) were used to precipitate the silver hexacyanoferrates.
A certain percentage of silver-ion sites in the lattice of silver
octacyanomolybdate(IV) (Agocm(IV)) and silver hexacyanoferrate(II)
1
3
methods. This is a great advantage over electrodes, on the
surface of which films of compounds have been deposited by
chemical or electrochemical methods.
(Aghcf(II)) can be occupied by potassium ions. Depending on the
IV
x 8 x
desired content of alkali ions (Ag4-xK [Mo (CN) ] and Ag4-xK -
II
Experimental Section
[Fe (CN) ] with 0 < x < 1), the following procedures have been
6
applied:
Electrochemical Measurements. The voltammetric experiments
were performed with Autolab PGSTAT 20 and PSTAT 10 instruments
+
Maximized K Content (Procedure I). To obtain the composition
4-
4-
6
Ag
3
K[X] (X ) [Mo(CN)
8
]
and [Fe(CN) ] ) stoichiometric amounts
(Eco-Chemie, Utrecht, The Netherlands) in conjunction with Metrohm
of 0.01 M silver nitrate solution were added dropwise to a stirred
solution containing 0.01 M potassium octacyanomolybdate or hexacy-
anoferrate, respectively, dissolved in a saturated potassium nitrate
solution. The precipitate was washed with a 1 M potassium nitrate
VA 663 electrode stands (Metrohm, Switzerland) and a 486 personal
computer. The electrochemical cell consisted of a three-electrode setup.
The reference electrode was Ag/AgCl (saturated KCl with E ) 0.204
V vs SHE), and the auxiliary electrode was a platinum wire. Electrodes
used in this study were paraffin-impregnated graphite electrodes (PIGE)
prepared from graphite rods (electrodes for spectrographic analysis,
VEB Elektrokohle, Lichtenberg, Germany) impregnated in melted
solution and was then vacuum-dried.
+
Minimized K Content (Procedure II). To precipitate the Ag
4
[X]
compounds, an excess of silver nitrate solution was added to a solution
of K [Mo(CN) ] or K [Fe(CN) ] dissolved in water. The precipitate
1
3,14
4
8
4
6
paraffin.
The sample preparation was carried out as follows: 1-3
was then allowed to age for 2 h in the presence of silver ions before
mg of the sample powder was placed on a glass plate. The electrode
was gently rubbed over the material in order to immobilize some
compound at the electrode surface. After measurement, the electrode
surface was cleaned with a razor blade. All electrolyte solutions were
degassed with high-purity nitrogen for 10 min prior to the electrochemi-
cal measurements.
being washed with distilled water.
The octacyanomolybdates of Cd +, Co2+, Cr2+, Cu2+, Fe2+, Mn2+
2
,
Ni2 , Pb , and Zn were precipitated by dropwise adding a 0.01 M
metal nitrate solution to the 0.01 M solution of K [Mo(CN) ]. Cr [Mo-
CN) ] was precipitated under a nitrogen atmosphere to prevent the
oxidation of chromium(II).
+
2+
2+
4
8
2
(
8
For the ring-disk experiment, a four-electrode potentiostat PG 287
The elemental composition of the synthesized compounds was
determined by CHN analysis and atomic absorption spectrometry. The
following results were obtained for the silver compounds:
(HEKA Elektronik, Lambrecht, Germany) with the following experi-
mental setup was used: A conventional rotating disk electrode with
the glassy carbon electrode replaced by a paraffin-impregnated graphite
rod was used in conjunction with a platinum wire (L 0.5 mm) winded
around the graphite electrode over a distance of approximately 1 mm.
The ring electrode was in-plane with the disk electrode. The sample
was immobilized at the graphite electrode as usual. This electrode could
then be polarized to electrolyze the immobilized compound. At the
platinum wire, a constant potential of -100 mV was applied in order
to accumulate silver ions that were expelled from the solid. In the
subsequent step, the accumulated silver was stripped off anodically.
Electrochemical Quartz Crystal Microbalance (ECQM) Mea-
surements. ECQM measurements were performed using a µ-Autolab
Silver octacyanomolybdate: precipitated following procedure I,
Ag3.36
K
0.64[Mo(CN)
8
]; precipitated following procedure II, Ag
]. Silver hexacyanoferrate: precipitated following procedure I,
]; precipitated following procedure II, Ag [Fe(CN) ].
4
[Mo-
(CN)
8
Ag3.05K
0.95[Fe(CN)
6
4
6
Structural Chemistry of the Silver Octacyanometalates and
Hexacyanoferrate. As is known for AgCN, silver(I) preferentially
forms linear bonds to its neighbor atoms. Thus, the nearly perfect
octahedral orientation of the cyanide ligands in the hexacyanoferrate
ion is an ideal basis for a three-dimensional structure of the silver
hexacyanoferrate lattice. The scaffold of silver hexacyanoferrate(II/
-
/0
(Eco-Chemie, Utrecht, The Netherlands) combined with a Fluke
III) is based on an Ag
3
[Fe(CN)
6
]
unit (Figure 1) with empty
+
+
PM6680B frequency counter (Fluke, The Netherlands) and a 10 MHz
oscillator Model 230 with AT-cut gold coated plano-convex quartz
crystals (Technical Department, Institut Chemii Fizycznej Polskiej
Akademii Nauk, Poland).
Microscopic in Situ Diffuse Reflectance Spectroelectrochemical
Measurements. A special electrochemical cell was used to perform
electrochemical measurements with solid particles immobilized on the
surface of graphite electrodes with in situ recording of diffuse
reflectance spectra under an incident-light microscope. The microscope
was a Leitz Laborlux 12 POL S incident-light microscope (Leica
interstitials, in the case of the hexacyanoferrate(III), and K or Ag
1
8
occupied interstitials, in the case of the hexacyanoferrate(II). It
possesses a hexagonal elementary cell with a ) 7.02 Å and c ) 7.25
Å. The C-N bond length is 1.12 Å, and the Fe-C bond length is 1.85
Å. In contrast to the hexacyanoferrate ion, the dodecahedral, or square
(15) Schr o¨ der, U.; Scholz, F. J. Solid State Electrochem. 1997, 1, 62-67.
(16) Furman, N. H.; Miller, C. O. In Inorganic Synthesis; Audrieth, L.F.,
Ed.; McGraw-Hill: New York, 1950; Vol. III, p 160.
(
17) Brauer, G. Handbuch der Pr a¨ paratiVen Anorganischen Chemie;
Enke: Stuttgart, 1954; p 1068.
(
18) Kahlert, H. Ph.D. Thesis, Humboldt-Universit a¨ t, Berlin, Germany,
1998.
(14) Scholz, F.; Lange, B. Trends Anal. Chem. 1992, 11, 359.

1008 Inorganic Chemistry, Vol. 39, No. 5, 2000
Schr o¨ der and Scholz
Figure 1. Structure of silver(I) hexacyanoferrate(III) along the [001]
1
8
axis (reproduced with permission ).
antiprismatic, configurations of the octacyanometalate ions allow a
crystalline lattice to be formed only with cations, which do not require
a linear preferential orientation. Hence, only crystal data of the soluble
1
9
20
4
alkali salts (e.g., K [Mo(CN)
8
]
and H
4
[Mo(CN)
8
]
and the corre-
sponding tungsten compounds) and the relatively sparingly soluble Tl
4
-
2
1
8
[Mo(CN) ] can be found in the literature. In the latter compound,
which has been reported to consist of square antiprismatic anion
polyhedra, the C-N-Tl bond angles vary between 83 and 164°.21 This
flexibility of the bond angles is a prerequisite for the good crystallinity
of the compound.
Results and Discussion
III
The powder diffractograms of Ag3[Fe (CN)6] and Ag3K-
II
[
Fe (CN)6] are given in Figure 2, parts a and b. They reflect
that both compounds possess a very similar structure, despite
the presence of potassium ions in the reduced compound. This
indicates that the potassium ions are only weakly interacting
with the cyanide ligands of the Aghcf(II) (hcf ≡ hexacyano-
ferrate), and thus, they only slightly perturb the host scaffold,
if at all.
II
Figure 2. Powder diffraction patterns of (a) Ag [Fe(CN) ], (b) Ag K-
When some micrograms of microcrystalline Ag3K[Fe (CN)6]
3
6
3
[
Fe(CN)
8
], and (c) Ag
4
[Fe(CN)
6
3 8
] obtained by storing Ag K[Fe(CN) ]
were mechanically immobilized on a gold-plated quartz crystal
of an ECQM and immersed into a 0.1 M KNO3 solution, an
addition of silver nitrate (the resulting silver concentration was
in a 0.1 M AgNO
3
for 10 h.
-
3
Equations 1 and 2 can be understood on the basis of the
solubility equilibria of Ag3K[Fe(CN)6] and Ag4[Fe(CN)6]:
1
0
M) decreased the frequency of the oscillator crystal by
about 2000 Hz. Obviously, this is due to the increased mass of
immobilized compound on the surface of the crystal. This mass
increase is caused by the exchange of the incorporated potassium
ions by silver ions according to the following reaction:
II
II
4-
+
{Ag K[Fe (CN) ]} a [Fe (CN) ] + K +
3
6
6
Ag+
3
(K ) (3)
L
II
+
II
+
II
II
4-
+
{
Ag K[Fe (CN) ]} + Ag f {Ag [Fe (CN) ]} + K (1)
{Ag [Fe (CN) ]} a [Fe (CN) ] + 4 Ag
(KL′) (4)
3
6
4
6
4
6
6
(
Braces denote solid phases.)
This is a simple ion-exchange reaction known from litera-
with KL ) 2.1‚10-9mol/L and KL′ ) 3.3‚10-5mol/L.23,24 In
the case of Ag4[Fe(CN)6], these solubility products are the re-
sult of a strong, partially covalent interaction of the extra sil-
ture.22 The ion exchange is accompanied by very pronounced
lattice alterations, that is, an amorphization, which can be seen
from the powder diffraction pattern of the reaction product (cf.
Figure 2c).
Electrochemical measurements (see below) show that the
storage of Ag3K[Fe(CN)6] in water causes the following
conversion of one solid compound into another to occur:
2
5,26
ver ions with the cyanide ligands.
In the case of Ag3K-
[Fe(CN)6], they are the result of comparably weak, mainly
electrostatic bonding of the potassium ions . Later, it will be
shown that similar reprecipitation reactions also occur with silver
and all metal(II) octacyanomolybdates, octacyanotungstates, and
(
(
21) Meske, W.; Babel, D. J. Alloys Compd. 1992, 183, 158-167.
22) (a) Tananaev, I. V.; Kachukhashvili, S. I. Zh. Neorg. Khim. 1962, 7,
1516-1520; Tananaev, I. V.; Kachukhashvili, S. I. Russ. J. Inorg.
Chem. 1962, 7, 782-785. b) Tananaev, I. V.; Levina, I. M. Zh. Neorg.
Khim. 1957, 2, 576-585.
II
II
4
{Ag K[Fe (CN) ]} f 3{Ag [Fe (CN) ]} +
3 6 4 6
II
4-
+
[
Fe (CN) ] + 4K (2)
6
(
(
23) Zasepa, W. Rocz. Chem. 1968, 42, 1129-1135.
24) Bellomo, A. Talanta 1970, 17, 1109-1114.
(
(
19) K o¨ nig, E. Z. Naturforsch. 1968, 23a, 853-860.
20) Basson, S. S.; Bok, L. D. C.; Leipoldt, J. G. Acta Crystallogr. 1970,
B26, 1209.
(25) Kerler, W.; Neuwirth, W., Fluck E. Z. Phys. 1963, 175, 200-220.
(26) Nekrasov, B. V.; Seifer, G. B.; Kharitonov, Yu. Ya. Bull. Akad. Sci.
USSR DiV. Chem. Sci. 1970, 219-224.

Electrochemistry of Cyanometalates
Inorganic Chemistry, Vol. 39, No. 5, 2000 1009
IV
Figure 4. Powder diffraction pattern of the conversion of (a) Tl
4
[Mo -
IV
(CN)
8 4 8
] to (b) Ag [Mo (CN) ]. The spectra have been recorded before
IV
(
a) and after an overnight storage of Tl
4
8 3
[Mo (CN) ] in 0.1 M AgNO
solution.
In Figure 4, the powder diffraction patterns of thallium
octacyanomolybdate(IV) before (Figure 4a) and after (Figure
4
b) an overnight storage in silver nitrate solution are shown.
The product of the overnight storage can be identified as Ag4-
Mo(CN)8]. The conversion of the thallium to the silver
compound, according to the following equation,
[
IV
+
IV
+
{
Tl [Mo (CN) ]} + 4Ag f {Ag [Mo (CN) ]} + 4Tl
4 8 4 8
V
3 8 4
Figure 3. Powder diffraction pattern of (a) Ag [Mo (CN) ], (b) Ag -
[
(6)
IV
V
8 3 8
Mo (CN) ], and of (c) Ag [Mo (CN) ] treated in a ball mill.
is brought about by the much lower solubility of Ag4[Mo(CN)8].
This reaction provokes a dramatic decrease of crystallinity of
the solid material. This deterioration of the long-range order
can be understood to be a result of the rigid bond angle of the
silver ions, in contrast to the flexible bonding between the
thallium ions and the cyanide polyhedra. It was found that the
powder diffraction patterns of silver octacyanomolybdate and
silver octacyanotungstate (Agoct) are nearly identical for the
oxidized and the reduced compounds, respectively. This explains
the strong similarities of the chemical and electrochemical
properties.
copper hexacyanoferrate. It is highly probable that these reac-
tions also occur with all other sparingly soluble metal hexacy-
anoferrates, although this might not be easily detectable due to
the special relation of the solubility products to each other.
Thus, eq 2 can be generalized for the silver salts as follows:
As in the case of the Aghcf(II) lattice, potassium ions can be
incorporated into the lattice of Agoct(IV) and Agocm(IV) (see
the results of analysis). Unlike silver hexacyanoferrate(II),
Agocm(IV) can only be produced with a maximum content of
II
IV
IV
where X represents Fe (CN)6, Mo (CN)8, or W (CN)8.
+
+
Depending on the activity of K and Ag in the solution,
the equilibrium will be shifted to either the left or the right side.
It should be mentioned that the actual extent of these reactions
will be strongly influenced by their kinetics. Although potassium
ions are immediately exchanged by silver ions (as found from
the quartz crystal microbalance experiment), the exchange of
+
IV
0.64 K , that is, Ag3.36K0.64[Mo (CN)8] when precipitated from
a saturated potassium nitrate solution. As shown later, this
formula describes the net composition of a mixture of Ag4[Mo-
(CN)8] and Ag3K[Mo(CN)8]. Because of the ratio of the
solubility products of these two compounds, it is probably
impossible to achieve a higher potassium net content. Although
only three of the four silver ions of Aghcf(II) are part of the
lattice framework, and therefore, the fourth (interstitial) ion can
be replaced by a potassium ion without a dramatic loss of
+
+
Ag by K is overtly slow.
The powder diffraction pattern of Agocm(V) (ocm ≡ octa-
cyanomolybdate) and Agocm(IV) are shown in Figure 3, parts
IV
V
a and b. They reveal that both Ag4[Mo (CN)8] and Ag3[Mo -
CN)8] are characterized by a considerably high content of
amorphous material, and a poor crystallinity in the case of
Agocm(IV). The latter is caused by the extremely low solubility
and the relative rigidity of the bond angle of the silver(I) ions.
(
II
III
stability (the solubility of Ag4[Fe (CN)6] and Ag3[Fe (CN)6]
are nearly equal), it takes four cations to connect the cyanide
ligands of the octacyanomolybdate polyhedra. Replacing a silver

1010 Inorganic Chemistry, Vol. 39, No. 5, 2000
Schr o¨ der and Scholz
III
V
electrochemical behaviors of Ag3[Fe (CN)6] and Ag3[Mo -
(CN)8] under these conditions. The voltammetry of Aghcf(III)
(Figure 5a) is chemically reversible. It is characterized by the
signal pair of A and B. The dependence of the formal potential,
EfA/B, on the activity of the potassium ions in the electrolyte
solution, that is, ∆EfA/B/∆ log aK+ ) 60.5 mV, indicates the
following reaction:
A
III
+
-
II
{
Ag [Fe (CN) ]} + K + e {
\
} {Ag K[Fe (CN) ]} (7)
3
6
3
6
B
An additional oxidation peak occurs at about 1100 mV in the
cyclic voltammogram of the analogue molybdenum compound
(D′, Figure 5b), which is in contrast to the results found for
hexacyanoferrate. With the help of a modified ring-disk experi-
ment that was carried out in a 1 M KNO3 solution, and which
is described in detail in the Experimental Section, the origin of
this peak could be revealed. While the rotating graphite
electrode, which was previously modified with a sample having
IV
a net composition of Ag3.6K0.4[Mo (CN)8], was polarized
toward positive potential, the surrounding platinum wire was
held at a constant potential of -100 mV in order to trap silver
expelled from the solid during the oxidation. This experiment
showed that only the oxidation process D′ is accompanied by
the expulsion of silver ions into the electrolyte solution. The
presence of the peak D′ is also accompanied by a fast decrease
of the voltammetric signals. Restricting the oxidation sweep to
a limit of 900 mV results in a stable and chemically reversible
electrochemistry (Figure 5c) with the dependence ∆Ef/∆ log
+
aK ) 59.5 mV. From this, it follows that the redox process,
analogous to reaction 7, is based upon the following reaction:
A′
V
+
-
IV
{
Ag [Mo (CN) ]} + K + e {
\
} {Ag K[Mo (CN) ]} (8)
3
8
3
8
B′
II
Figure 6 shows the cyclic voltammograms of Ag4[Fe (CN)6]
IV
(
Figure 6a) and Ag4[Mo (CN)8] (Figure 6b), which were
recorded in the presence of a 0.1 M KNO3 electrolyte. In the
first cycle, both figures exhibit the same characteristics. During
the first oxidation half-scan, the compounds are oxidized,
accompanied by a loss of silver ions:
+
-
{
Ag [X]} f {Ag [X]} + Ag + e
(9)
4
3
The expelled silver ions escape by diffusion into the bulk
solution, and hence, they cannot be completely reinserted into
+
+
the compound during the next reduction. Instead of Ag , K is
inserted during the subsequent reduction:
Figure 5. Cyclic voltammograms of (a) Ag
3
[Fe(CN)
6 3
] and (b, c) Ag -
+
-
V
{Ag [X]} + K + e f {Ag K[X]}
(10)
[Mo (CN)
8
] mechanically attached on the surface of a paraffin-
3
3
impregnated graphite electrode and immersed in a 0.1 M KNO
rate 5 mV/s.
3
, scan
At this point, the differences between Agocm(IV) and
Aghcf(II) become very obvious: Although Aghcf(II) is able to
form Ag3K[Fe(CN)6], which is stable when in contact with a
0.1 M KNO3 solution, the position of the precipitation equilib-
rium (5) of Agocm(II) leads to an internal reconstruction under
partial re-formation of Ag4[Mo(CN)8]:
ion with a potassium ion thereby weakens the lattice structure
and raises the solubility of the compound.
The Electrochemistry of Aghcf and Agocm in Potassium
Nitrate Electrolyte. Along with the structures, the electro-
chemical properties of octacyanotungstates and octacyanomo-
lybdates are very similar. Therefore, it suffices to explain the
electrochemical behavior using the example of silver octacya-
nomolybdate. Cases in which differences are observed will be
noted.
The voltammetry of solid Aghcf(II/III) or Agocm(IV/V),
immobilized on the electrode surface and adjacent to a potassium
nitrate solution, is strongly affected by the precipitation
equilibria presented in eq 5. Figure 5 shows the very different
The result of this reaction is a mixture of the two compounds
Ag4[Mo(CN)8] and Ag3K[Mo(CN)8]. Instead of giving the ratio
of these two compounds, one can write the following formula:
Ag3.36K0.64[Mo (CN)8]. The oxidation of these two compounds
occurs at well-separated potentials, as it comprises the oxidation
IV

Electrochemistry of Cyanometalates
Inorganic Chemistry, Vol. 39, No. 5, 2000 1011
Table 1. Formal Potentials E
by Cyclic Voltammetry as E
f
[mV vs Sat. Ag/AgCl] Determined
) (Epc + EPa)/2
f
0
.1 M KNO 0.1 M TlNO
3
3
0.1 M AgNO
3
Ag
Ag
Ag
3
3
3
[Mo(CN)
[W(CN)
[Fe(CN)
8
]
831.2 ( 0.8
574.4 ( 3.4
671.0 ( 4.0
948.3 ( 5.0
663.5 ( 4.5
788.6 (3.2
1145.4 ( 4.8
871.6 ( 3.0
1007.4 ( 4.8
8
]
]
6
+
Ag plays a crucial role. The potassium ions are only electro-
statically bonded, whereas the silver ions are partially covalently
bonded. The silver ions reduce the electron density of the redox
active central atom (e.g., of the 3d orbital of the iron(II) in the
2
5
hexacyanoferrate ) leading to a shift of the redox potential
toward more positive values, as shown in Table 1. Table 1 also
+
includes data for the Tl insertion electrochemistry. The
potential shift turns out to be nearly constant (about 300 mV)
for all investigated compounds. This influence of the metal ions
coordinated to the complex ion (hcf, ocm, oct) has already been
studied in detail in case of metal hexacyanoferrates.30
The process of continuous re-formation and dissolution finally
leads to a total degradation of the solid material at the electrode
surface and a disappearance of the voltammetric response.
As expected, the same situation arises for Aghcf(II/III) for
cases in which the potassium ion concentration in the electrolyte
-
2
-1
solution is lower than 10 mol‚L (Figure 6c):
Figure 7, parts a and b, show that copper hexacyanoferrate(II)
exhibits a similar behavior. Two distinct oxidation signals can
be observed which are based upon the following redox processes
(13, 14).
II
III
2+
-
2
{Cu [Fe (CN) ]} a 2{Cu [Fe (CN) ]} + Cu + 2e
2 6 1.5 6
(13)
II
III
+
-
{
Cu K[Fe (CN) ]} a {Cu [Fe (CN) ]} + K + e
1
.5
6
1.5
6
(14)
Both reactions are chemically reversible in 0.1 M solutions of
the respective cation.
Electrochemical Determination of the Potassium Content
of Solid Cyanides. The phenomenon that the release of two
different interstitial cations during oxidation prompts the oc-
currence of two different voltammetric signals can be used for
the quantitative evaluation of the composition of silver octa-
cyanomolybdates. As outlined in Figure 8, parts a and b, it is
based upon a two-step chronocoulometric oxidation of samples
Figure 6. Cyclic voltammograms of (a) Ag
4
[Fe(CN)
6
] and (b) Ag
4
-
IV
with a net composition Ag4-xKx[Mo (CN)8].
[
Mo(CN)
and of (c) Ag
mV/s.
8
] recorded in the presence of 0.1 M KNO electrolyte solution
3
-
3
The sample was immobilized on the surface of a paraffin-
impregnated graphite electrode, which was then immersed into
a potassium nitrate solution. To minimize the reprecipitation
3
6 3
K[Fe(CN) ] in the presence of 10 M KNO , scan rate
5
of two differently coordinated octacyanomolybdate ions. Fur-
thermore, the two distinct oxidation signals unambiguously
-1
reaction (5), a relatively concentrated (1 mol‚L ) potassium
nitrate solution was used. In the first step, the electrode was
IV
prove that no mixed crystals have formed between Ag3K[Mo -
polarized toward a potential at which the oxidation of Ag3K-
IV
(
CN)8] and Ag4[Mo (CN)8]. In that case, according to mixed
IV
[
Mo (CN)8] proceeds accompanied by the release of potassium
phase thermodynamics, and in accordance with experimental
ions (see Figure 8a). The potential was held at this value for
50 s in order to allow a complete conversion of the sample.
2
7-29
results,
one would observe a single signal shifting from
2
the position of one compound to the position of the other. Here,
the difference between the bonding of the inserted K and the
IV
The second potential step was aimed to oxidize Ag4[Mo (CN)8]
accompanied by the release of silver ions.
+
The results of the chronocoulometric determination of the
composition of differently precipitated silver octacyanomolyb-
(
27) Bond, A. M.; Scholz, F. Langmuir 1991, 7, 3197-3204.
(28) Reddy, S. J.; Dostal, A.; Scholz, F. J. Electroanal. Chem. 1996, 403,
2
09-212.
29) Meyer, B.; Zhang, S.; Scholz, F. Fresenius’ J. Anal. Chem. 1996, 68,
, 267-270.
(
(30) Scholz, F.; Dostal, A. Angew. Chem. 1995, 107, 2876-2878 ( Scholz,
F.; Dostal, A. Angew. Chem., Int. Ed. Engl. 1995, 34, 2685-2687).
6

1012 Inorganic Chemistry, Vol. 39, No. 5, 2000
Schr o¨ der and Scholz
Figure 8. (a) Linear sweep voltammogram (scan rate 10 mVs-1) of
the oxidation of Ag3.36
solution. (b) chronocoulometric plot of the oxidation of a Ag4-x
Mo(CN)
K
0.64[Mo(CN)
8
] in the presence of 1 M KNO
3
K
x
-
[
8
]; Initial potential: 0.6 V, first potential step: 1.05 V, second
potential step: 1.25 V.
the single-electron/single-ion transfer reaction that is expected
for the reaction of silver octacyanomolybdate in the presence
of a silver nitrate solution.
Figure 7. Succeeding differential pulse voltammograms of the
oxidation of a sample of Cu1.5K[Fe(CN) ] immobilized on the surface
of a paraffin-impregnated graphite electrode and immersed in a 0.2 M
Cu(NO solution (A). The compound was rereduced at 200 mV after
each scan. After twenty redox cycles, the electrode was immersed in a
.1 M KNO solution and differential pulse voltammograms of the
6
IV
V
+
-
{
Ag [Mo (CN) ]} a {Ag [Mo (CN) ]} + Ag + e (15)
4 8 3 8
3 2
)
0
3
For this simple reaction, a single voltammetric system can be
expected. Figure 9 shows that the reality is more complex. It
can be seen that, depending on the scan rate, two main signal
systems appear (Aox/Ared and Box/Bred). Additionally, the peaks
of the system Box/Bred are split. A similar peak splitting can be
observed in the voltammogram of silver hexacyanoferrate
oxidation (B) were recorded as described for A. (Step potential ) 5
mV, modulation amplitude ) 25 mV.)
IV
dates, with a net composition Ag4-xKx[Mo (CN)8], are in good
agreement with the data obtained by atomic absorption spec-
trometry (AAS). For the calculation, it was assumed that the
(Figure 10).
-
unit {Ag3[Mo(CN)8] } remains unchanged. On the basis of the
The peaks of the systems Aox/Ared and Box/Bred (Figure 9)
+
+
formula Ag1-xKx{Ag3[Mo(CN)8]}, the ratio Ag /K was cal-
culated following QAg+ + QK+ ) 1, where Q is the charge of
the single-oxidation processes normalized to the total oxidation
charge.
differ in their half width, peak separation, and scan rate behavior.
In Figure 11 a, the result of an ECQM measurement is presented.
V
It shows that the reduction of Ag3[Mo (CN)8] is accompanied
by a decrease in the oscillation frequency, indicating an increase
in the mass of the immobilized compound. The subsequent
oxidation proceeds with a decrease in mass, indicating a release
of cations. The shape of the derivative of the frequency change
and the corresponding cyclic voltammogram (cf. Figure 11b)
are quite similar, proving that the mass-to-charge ratio remains
constant during the potential sweep. This supports the assump-
tion that only one cation is involved in the different peaks. These
results, and the dependence of the formal potentials of both
The net composition of two different samples of Ag4-xKx-
IV
[
Mo (CN)8] is determined by:
AAS
chronocoulometry
Ag3.36
K
0.64[Mo(CN)
0.22[Mo(CN)
8
8
]
]
Ag3.41
K
0.59[Mo(CN)
0.23[Mo(CN)
8
]
]
Ag3.78K
Ag3.77K
8
The outlined approach can probably be applied to other complex
cyanides as well.
The Electrochemistry of Silver Octacyanomolybdate in
the Presence of Silver Nitrate Electrolyte. Equation 15 depicts
redox systems on the silver ion activity, that is, ∆Ef
/
/∆ lg
Aox Ared
+
/∆ lg aAg+ ) 53.3 mV, indicate
aAg ) 55 mV and ∆Ef
/
Box Bred

Electrochemistry of Cyanometalates
Inorganic Chemistry, Vol. 39, No. 5, 2000 1013
Figure 10. Cyclic voltammogram of a sample of Ag
immobilized on the surface of a paraffin-impregnated graphite electrode
and immersed in a 0.1 M AgNO solution, scan rate 0.1 mV/s.
3 6
[Fe(CN) ]
3
3 8
Figure 9. Cyclic voltammograms of a sample of Ag [Mo(CN) ]
immobilized on the surface of a paraffin-impregnated graphite electrode
and immersed in a 0.1 M AgNO solution.
3
Figure 11. (a) Frequency response for the reduction and reoxidation
of Ag [Mo(CN) ] deposited onto a 5 mm diameter gold-coated quartz
crystal oscillator; 0.1 M AgNO , 5 mV/s. (b) Derivative of the frequency
over the potential (dashed line) in comparison to a corresponding cyclic
voltammogram.
3
8
that both of the signal systems are based on reaction 15. To
elucidate the differences between the systems Aox/Ared and Box/
Bred, a variation of the scan rate was helpful. Figure 12a depicts
these dependencies for the oxidation peaks Aox and Box. With
3
1
i ∝ v , peak Aox obeys a typical thin-layer behavior, whereas
of this experiment in the form of a plot of the logarithm of the
molar ratio C /C versus potential. The C /C ratio was
1/2
Box exhibits a diffusion control, that is, i ∝ v .
ox red
ox red
Another important feature of the two systems is their
reversibility. Aox/Ared was found to be electrochemically revers-
ible, having a zero-peak separation at scan rates lower than 0.5
mV/s. The considerable peak separation of Box/Bred (45 mV at
a scan rate of 0.1 mV/s) indicates a kinetic control. The different
reversibility can also be deduced from multistep chronocoulo-
metric experiments. To study the redox systems under equilib-
rium conditions, Agocm(IV) was oxidized stepwise (the potential
step was 10 mV, and the step time was 300 s), and the
chronocoulograms were recorded. Figure 12b shows the results
calculated from the integrated charge related to the total
oxidation charge. According to this plot, the system Aox/Ared
follows the Nernst equation for a one-electron process (with a
slope of 1/59 mV). However, the system Box/Bred has a slope
of 1/23 mV. The latter figure points to a very sudden conversion
of the reduced form to the oxidized form and vice versa. This
is also obvious from the cyclic voltammogram (Figure 9a),
which shows very steep peaks of the system Box/Bred. This
sudden conversion, exhibiting a super-Nerstian slope, may be
caused by a nucleation and growth rate control. Nucleation

1014 Inorganic Chemistry, Vol. 39, No. 5, 2000
Schr o¨ der and Scholz
Figure 13. Kubelka-Munk function of Agocm(IV) immobilized on
the surface of a paraffin-impregnated graphite electrode and immersed
in a 0.1 M AgNO solution. The Kubelka-Munk function was recorded
3
in situ during the cyclic oxidation and rereduction of the compound.
Figure 12. (a) Scan rate dependence of the oxidation peaks (see Figure
0a) of Ag [Mo(CN) ]/electrolyte solution: 0.1 M AgNO . (b) Plot of
the ratio of the molar ratios of the oxidized compound (Ag [Mo(CN) ])
to the reduced compound (Ag [Mo(CN) ]) derived from a multistep
chronocoulometric experiment of the oxidation of Ag [Mo(CN) ]/
electrolyte solution: 0.1 M AgNO
1
3
8
3
3
8
4
8
4
8
3
.
overpotential may exist, and surpassing it will lead to a very
rapid conversion of the compound, limited only by the diffusion
of cations in the solid phase (cf. last paragraph).
Summarizing these results and taking the poor crystallinity
and high content of amorphous compound into consideration
(cf. Figure 3), the following conclusions can be made. At the
redox system Aox/Ared, the conversion of amorphous particles
takes place. The conversion of amorphous material proceeds
very easily as the redox centers are better accessible; the cation
transport is less hindered and less energy is necessary to recon-
struct the lattice of these extremely small particles. This explains
the high electrochemical reversibility of this redox process. On
the other hand, the peak system Box/Bred, is caused by the elec-
trolysis of larger particles of much better crystallinity. To prove
this hypothesis, the raw material was milled in a ball mill until
it became almost X-ray amorphous. The effect of this treatment
on the crystallinity can be seen in Figure 3c. All of the sharp
reflections of the powder diffraction pattern are eliminated, and
only weak and broad signals remain, thus indicating a pro-
nounced defect in structure and large amounts of amorphous
Figure 14. Linear sweep voltammogram (full line) of Agocm(IV/V)
in the presence of 0.1 M AgNO
3
, scan rate 1 mV/s. First derivative of
the reflectance over the dR/dE potential (dotted line) versus potential.
(
A) The reflectance was measured for a sample layer of ∼1 µm
thickness. (B) The reflectance was measured for a sample layer of ∼10
µm thickness.
compound. In the cyclic voltammogram of this sample, of Ag3-
V
[
Mo (CN)8], the sharp voltammetric signals (Box/Bred) are
changes of the Kubelka-Munk functions during a cyclic
oxidation and rereduction are presented in Figure 13. For the
two differently thick layers, a comparison was made between
the voltammetric response and the first derivative of the
reflectance over potential, that is, dR/dE. In the case of a thin
absent, and only a rather broad system is present. The same
IV
results were obtained from a milled sample of Ag4[Mo (CN)8].
This designation of the systems Aox/Ared and Box/Bred is
supported by spectroelectrochemical measurements. The diffuse
reflectance spectra of Agocm(IV) (which is a yellow compound)
1
5,31
and Agocm(V) (which is a dark brown compound)
were
(31) Schr o¨ der, U.; Meyer, B., Scholz F. Fresenius’ J. Anal. Chem. 1996,
simultaneously measured during the cyclic voltammetry. The
68, 295-298.

Electrochemistry of Cyanometalates
Inorganic Chemistry, Vol. 39, No. 5, 2000 1015
does not require any significant change of the lattice. However,
+
the insertion of Ag tremendously changes the lattice structure
and crystallinity (cf. Figure 2c). It seems reasonable to assume
that a nucleation growth mechanism, similar to the one found
for octacyanomolybdate, is responsible.
The Electrochemistry of Solid Octacyanomolybdates of
Bivalent Cations. In the presence of potassium nitrate solution,
the octacyanomolybdates of the bivalent cations show an
electrochemistry (Figure 15) very similar to that of the silver
octacyanometalates, which is based upon the following reac-
tions:
Redox system Rred/Rox
II
IV
II
V
+
-
{
M K[Mo (CN) ]} a {M [Mo (CN) ]} + K + e
1.5 8 1.5 8
IV
(16)
2 8
Figure 15. Cyclic voltammograms of Co [Mo (CN) ] mechanically
attached on the surface of a paraffin-impregnated graphite electrode
and immersed in a 0.1 M KNO , scan rate 50 mV/s.
Redox system Sred/Sox
3
Table 2. Formal Potentials E
Systems of Octacyanomolybdates of Selected Bivalent Cations in
the Presence of 0.1 M KNO Determined by Cyclic Voltammetry as
) (Epc + EPa)/2 (Recorded with a Scan Rate of 50 mV/s)
f
[mV vs Sat. Ag/AgCl] of the Redox
II
IV
II
V
2+
-
2
{M [Mo (CN) ]} a 2{M [Mo (CN) ]} + M + 2e
2 8 1.5 8
(
17)
3
E
f
The rapidly decreasing response, which is visible in Figure 15
for the case of Co2[Mo (CN)8], can be explained by the high
solubility of the M [Mo (CN)8] compounds and the disinte-
gration of M K[Mo (CN)8] under a partial dissolution analo-
redox system redox system
IV
R
red/Rox (mV)
Sred/Sox (mV)
II
V
1.5
Cd
Co
2
[Mo(CN)
[Mo(CN)
8
]
]
]
590.5
600.0
571.5
578.5
560.0
588.0
612.0
581.5
594.0
not reversible
838
917
880
not reversible
981
873
not reversible
not reversible
II
1.5
IV
2
8
Cr
2
[Mo(CN)
[Mo(CN)
[Mo(CN)
[Mo(CN)
8
gous to eq 11.
Cu
2
8
]
The formal potential of the peak system Rred/Rox, that is, Ef
(Epa+Epc)/2, of all studied compounds (see Table 2) was close
to the formal potential of the dissolved octacyanomolybdate in
a 0.1 M KNO3 solution.
Fe
2
8
]
8
)
Mn
2
]
Ni
Pb
Zn
2
[Mo(CN)
[Mo(CN)
[Mo(CN)
8
]
]
]
2
8
2
8
Conclusions
layer (∼1 µm), the dR/dE curve tracks the voltammetric curve
only in case of peak Box (Figure 14a). There is no significant
change in the reflectance at peak Aox. This indicates that the
peak Aox is caused by the oxidation of such small, and therefore
probably amorphous, particles, that the light penetrates them
without being detectably absorbed, even by the oxidized form.
The changes of the optical properties of these particles can only
be observed in the case of a thicker substance layer (about 10
µm) (cf. Figure 14b). In that case, the small particles contribute
to the diffuse reflection as well. In contrast to the amorphous
particles, the conversion of crystalline particles always leads to
changed reflectance spectra.
This study of the electrochemical behavior of various solid
metal hexacyanoferrates, octacyanomolybdates, and octacyano-
tungstates revealed that their behavior is governed by some
general rules and reaction schemes, which have not been
described thus far:
The electrochemical behavior of the studied solid metal
cyanometalates in the presence of potassium nitrate solution can
be explained on the basis of precipitation equilibria of the
involved compounds, as exemplified in the reaction scheme (5).
There is no mixed crystal formation between the two forms
of the reduced metal cyanometalates Me′3K[X] and Me′4[X]
+
2+
(Me′ ) Ag ), and Me′′1.5K[X] and Me′′2[X] (Me′′ ) Me ,
2+
2+
2+
The Electrochemistry of Aghcf in Silver Nitrate Electro-
lyte. The cyclic voltammograms of silver hexacyanoferrate(II/
III) in the presence of silver nitrate solution are very similar
for the compounds Ag3[Fe(CN)6] and Ag4[Fe(CN)6]. Figure 10
shows that the cyclic voltammogram, very similar to the
octacyanomolybdate, consists of a split and steep peak system
e.g., Cu , Co , Cu ), respectively. Each of these compounds
gives its own voltammetric signal, allowing the determination
of their ratio in mixtures.
It could be shown that amorphous and crystalline particles
of silver octacyanomolybdate and octacyanotungstate samples
produce separate oxidation and reduction signals. Although the
conversion of the amorphous compounds is thermodynamically
controlled, the electrolysis of crystalline constituents is under
kinetic control and associated with a nucleation overpotential.
′
B′ox/B red, the underneath of which is obviously a very broad
peak that can be compared to the Aox/Ared system of the
octacyanomolybdate. A comparison of the cyclic voltammo-
grams recorded in the presence of silver nitrate solution (Figure
Acknowledgment. Uwe Schr o¨ der was supported with a
scholarship by Hans-B o¨ ckler-Stiftung. The Fonds der Chemis-
chen Industrie is acknowledged for their constant support of
the research.
1
0) with the voltammogram recorded in the presence of
potassium nitrate solution (Figure 5a) shows that the peak
splitting only occurs with the transfer of silver ions. As the
powder diffraction patterns (Figure 2, parts a and b) indicate,
the reduction of Ag3[Fe(CN)6] in the presence of potassium ions
IC9909330