1152

J . Phys. Chem. 1984,88, 1152-1156

A Theory for the Apparent Negative Second Wien Effect Observed in

Electric-Field-Jump Studies of Suspensions
Z. A. Schelly* and R. D. Astumian
Department of Chemistry, University of Texas at Arlington, Arlington, Texas 76019-0065
(Received: June 27, 1983; In Final Form: August 15, 1983)

Transient conductance decrease has recently been observed during E-jump relaxation kinetic studies of adsorption-desorption
processes in several colloidal suspensions. A theoretical analysis of these observations is given and an interpretation is described,
where electric-field-induced increased adsorption of ions is shown to be responsible for the apparent negative second Wien
effect. A dipole model system is discussed, which exhibits a true negative second Wien effect. It is suggested that such
a dipole mechanism may provide an interpretation for the hitherto unexplained negative second Wien effects in homogeneous
solutions, which had been observed previously.

Introduction
In relaxation kinetic studies2 of fast rate processes, an equilibrium system is perturbed by a small, rapid, stepwise (transient
or jump methods) or periodic (stationary methods) variation of
a thermodynamic variable, followed by the measurement of the
rate and amplitude of the subsequent reequilibration, by monitoring a physical parameter that is proportional to the concentrations. During the perturbation, energy is introduced into the
system. For a reversible exchange of energy between a system
and its surrounding, the total differential of the Gibbs free enthalpy
G, in general, is given by3
dG = COj dFj - A dC;
J

(1)

where Oj and Fj are conjugate extensive (generalized displacements)

and intensive (generalized forces) variables, respectively. A is
the affinity, and C; is the extent of the reaction. Clearly, depending
on the nature of the system, the change of any of the conjugate
variables in (l), i.e.
SdT

fidg

ni dpi

? dl?

can affect the shift of the equilibrium. In addition ot the familiar

thermodynamic symkols (S, T, V, P , ni) in <2), pi stands for the
chemical potential, is the electric field, P is the polarization
of dielectric media, D* is related to_the electric displacement,2
M denotes the magnetization, and H is the magnetic field. It was
found theoreticall? that the rate 7 of reequilibration is faster (and
therefore possibly less easily measurable) if the perturbation is
accomplished by the change of an intensive variable Fj while the
extensive variables 0, are kept constant. Nevertheless, experimentally it is much easier to suddenly change an Fj, which led
to the development2 of the T-jump, P-jump, E-jump, and solvent(pi)-jump5 methods, as well as the sound absorption and
dispersion and different dielectric6 relaxation methods.
In the present paper we shall focus on some aspects of the
electric field (E)-jump method. This method is based on the
sensitivity of chemical equilibria involving ions, dipolar, or po( I ) Abstracted in part from R.D.A.s Dissertation, University of Texas at
Arlington, 1983.
(2) Eigen, M.; DeMaeyer, L. In Technique of Organic Chemistry, Part
2; Friess, S. L.; Lewis, E. S.; Weisberger, A,, Eds.; Interscience: New York,
1963; Vol. 8.
(3) Schelly, Z. A. J . Chem. Educ. 1980, 57, 247.
(4) Schlogel, R. Z . Phys. Chem. (Wiesbnden) 1956, 9 , 259.
( 5 ) Chao, D. Y.; Schelly, Z. A. J . Phys. Chem. 1975, 79, 2734.
(6) DeMaeyer, L.; Persoons, A. In Techniques in Chemistry, Part 2;
Hammes, G. G., Ed.; Wiley: New York, 1973; Vol. 6.

0022-3654/84/2088-1152$01.50/0

larizable species to the electric field strength, according to the

relation

(a In Ka/ait)p,T= A ~ * / R T

(3)

where K, is t_he thermodynamic equilibrium constant of the reaction and AD* is the change in the macroscopic electric moment
of the system per mole equivalent transformation of reaction
partners

Le. the difference of partial molar_polarization of products and

reactants. The quantity D = tO5Eis usually called electric displacement, which differs from D* by the volume V of the field
space between the electrodes. The symbols to and t in (4) are the
permittivity of vacuum (rational system of units) and the macroscopic dielectric constant of the homogeneous medium, respectively. The implication of eq 3 is that, if a reaction in question
involves a net change in the dipole moments or the number of ions,
an electric field can be used to perturb the system. Generally,
field strengths of the order of 5C-100 kV/cm must be used to cause
observable effects in homogeneous solutions.
In the first E-jump experiment, perturbation was caused by
a single, strongly damped sinusoidal pulse of the field that also
drove the Wheatstone bridge for conductivity detection. The
application of a square pulse8 instead of the sinusoidal pulse has
the advantage that the relaxation time can be determined in a
single experiment. Subsequent instrumental developments included
the replacement of the several hundred meters of coaxial delay
cables by a low-inductance plastic capacitorg and the introduction
of photometric detection,I0 where the use of a pulsed or synchronized flashI2 light source significantly improved the signalto-noise ratio. In spite of the advanced instrumentation, however,
only a relatively small number of chemical systems have been
studied by the E-jump method.13 The majority of these systems
involved weak electrolytes (in water as well as o r g a n i ~ ~solvents)
J~
Eigen, M.; Schoen, J. 2.Elektrochem. 1955, 59, 483.
Eigen, M.; DeMaeyer, L. Z . Elektrochem. 1955, 59, 986.
. . Olsen, S. L.; Silver, R.L.; Holmes, L. P.; Auborn, J. J.; Warrick, P.,
Jr .; Eyring, E. M. Rev. Sci. Znstrum. 1971, 42, 1247.
(10) Ilgenfritz, G. Dissertation, George August University, Goettingen,
West Germany, 1966.
(11) Olsen, S . L.; Holmes, L. P.; Eyring, E. M. Rev. Sri. Instrum. 1974,
45, 859.
(12) Savage, C. R.; Herrmann, U.; Sumdani, G.; Schelly, Z. A. Rev. Sci.
Instrum. 1980, 51, 1190.
(13) For recent reviews, see: Eyring, E. M. In Chemical and Biological
Applications of Relaxation Spectrometry; Wyn-Jones, E., Ed.; Reidel:
Dordrecht, Holland, 1975. Purdie, N.; Eyring, E. M.; Rodriquez, L. In
Techniques of Chemistry; Weissberger, A.; Rossiter, B., Eds.; Wiley: New
York, 1980; Vol. 9.
(14) Massey, M. W.; Schelly, Z. A. J . Phys. Chem. 1974, 78, 2450.
0 1984 American Chemical Societv

Apparent Negative Second Wien Effect

The Journal of Physical Chemistry, Vol. 88, No. 6,1984 1153

for which the field-induced equilibrium shift is typically much

larger than for dipolar equilibria, and the effect is more amenable
to detection.
Recently, the E-jump method was also used to study heterogeneous systems, namely the adsorption-desorption equilibria of
ions at the solid-liquid interface in aqueous suspensions of colloidal
metal oxide particles.16i8 Interestingly, the expected exponential
increase of the conductance during the application of the high
electric field was preceded by a transient decrease of conductance,
with relaxation times ranging from 5 to 50 ps. In the present paper
we propose a field-induced increased adsorption of ions on the
particles as an explanation for the apparent negative second Wien
effect.

Discussion
Wiens experimental observations of conductivity increases upon
the application of high electric fields to both strongig and weak20
electrolytes are now referred to as the first and second Wien
effects, respectively. Onsager, Kim, and Fuoss have provided
theoretical interpretations for both types of deviation from Ohms
law. It was found that the major effect of the field on strong
electrolytes21is the acceleration of ions to such an extent that they
are able to break away from their surrounding ion atmosphere,
which leads to increased mobilities. For weak
where
the dissociation is described by eq 5, the major effect of the electric
k

AB
k-l

(A+-B;)
ion pair

=A + + Bk2A

Figure 1. Schematic representation of the conductometric detection

(Wheatstone) bridge in E-jump experiments. For the balanced bridge,

before the application of the perturbation square pulse, the approximate
values of the circuit components are as follows for the sample (s) and
reference (r) cells: C, = C, 2 25 p F R,2 R,= 104-10s a. R,= R2 =
50 0. OSC = oscilloscope.

(5)

k-2

field is the asymmetric perturbation of the ionic Brownian motion,

leading only to an increase of the rate constant kz. This results
in an overall shift to the right of eq 5, which produces ions. The
relative increase of the equilibrium constant K2 = k2/k-2due to
this dissociation field effect is given byZZ

l o

where Kz(E) and Kz(E=O) are the dissociation constants at a field

E and E = 0, respectively. The quantity q = -z,zzeo2/8mq,kT
is known as the Bjerrum radius. The energy of Coulombic
interaction between two ions bearing charges of zleOand zzeOin
a medium with dielectric constant e becomes equal to kT when
they approach each other to a distance of 29. The quantity 2@
represents a reciprocal distance and is defined as

20 = IEeo(ziu1 - z2uz)/(ui + u2)k71

t=O

t,

-t
Figure 2. Schematic representation of the several different transient

(7)

where uiare the mechanical mobilities (reciprocal of the friction

constants) of the ions, in contrast to the ionic electrical mobilities
which are given by zieoui. Two opposite charges of equal magnitude zeo must be separated to a distance of 1/2p to form a dipole
that acquiresJhe energy -kT if aligned in the direction of the
applied field E . Thus, in a first approximation, the ratio 209 of
the two characteristic distances q and 1/2P describes the probability of two ions escaping from an ion pair under the influence
of an external field.23
~

~~~

(15) Marshall, D. B.; Strohbusch, F.; Eyring, E. M. J. Phys. Chem. 1981,

85, 2270.
(16) Sasaki, M.; Moriya, M.; Yasunaga, T.; Astumian, R. D. J. Phys.
Chem. 1983,87, 1449.
(17) Astumian, R. D.; Moriya, M.; Sasaki, M.; Yasunaga, T., submitted

for publication.
(18) Hachiya, K.; Ashida, M.; Sasaki, M.; Kan, H.; Inoue, T.; Yasunaga,
T. J . Phys. Chem. 1979,83, 1866.
(19) Wien, M. Phys. Z . 1928, 29, 751.
(20) Wien, M. Phys. Z . 1931, 32, 545.
(21) Onsager, L.; Kim, S. K. J . Phys. Chem. 1957,61, 198. Fuoss, R. M.;
Onsager, L. Ibid. 1957, 61, 668.
(22) Onsager, L. J . Chem. Phys. 1934, 2, 599.
(23) DeMaeyer, L. In Methods of Enzymology; Kustin, K., Ed.; Academic Press: New York, 1969; Vol. 14.

signals observed or expected for the models discussed in the text and their
temporal relationship to the perturbation square pulse (a), that is triggered at time t = 0 and terminated at t l . (b) Typical conductance
increase caused by the second Wien effect for a weak electrolyte. (c)
Solid line: Typical signal observed in the suspensions discussed (ref
16-18). The initial conductance decrease is due to the sorption of ions
on the colloidal particle dipoles induced by the external field 8. The
relaxation time and the amplitude of the decrease are dependent on E .
The subsequent conductance increase is caused by the simultaneous
second Wien effect on the surface of the particles. Dashed line: Expected
signal for a system where only the sorption of ions on the induced dipoles
occurs, in the absence of the second Wien effect. (d) Expected signal
for a system where polarization is the only effect. In contrast to (a), (b),
and (c), the time scale for the period between t = 0 and t l has been
expanded by about a factor of 10 in this figure.

In an actual E-jump experiment, of course, one usually wishes

to establish the time dependence of the major (second Wien) effect.
However, upon the application of the external field, several secondary effects such as the first Wien effect, Joule heating, electrode
polarization, electrolysis, and space charge effects are also present.
The last is due to the buildup of double layers at the electrodes
in less than 1 ps at the high fields
Another effect, which,
(24) Schelly, Z. A.; Lundy-Douglas, K.; Eyring, H. Proc. Narl. Acad. Sci.

U.S.A. 1978, 75, 2549.

Schelly and Astumian

1154 The Journal of Physical Chemistry, Vol. 88, No. 6,1984

Figure 3. Schematic representation of the models discussed. The dielectric constant e and conductance u of the medium (1) and the particles
(2) are valid for all pictures. (a) A positively charged particle at equilibrium with the electrolyte solution in the absence of an external field.
(b) Maxwell-Wagner polarization of an uncharged particle for the case
u1 > u2. (c) Illustration of the quantities relevant for surface. conductance
us. (d) Bound and diffuse layer polarization of a positively charged
particle at steady state in an external field. Charges in the double layer
represent excess local charges.

as we shall show, can become a major factor, is the polarization

of the particles of the system under investigation.
All the secondary influences can be reduced to a negligible
minumum by suitable difference measurements, the proper design
of the instrument, and the choice of the experimental conditions.
Details are given in ref 2 and 23. For conductivity detection, the
essential part of the experimental setup is the Wheatstone bridge
(Figure 1) that was used also in the studies of the colloidal syswhich we wish to analyze. In a typical E-jump experiment, the sample cell contains the weak electrolyte studied, and
the reference cell is fitted with a solution of strong electrolyte that
has the same specific conductivity as the weak electrolyte. Prior
to the application of the high-voltage perturbation square pulse,
the bridge is balanced to UDB= 0 with a small alternating potential
applied across AC (see Figure 1). Under this condition the
impedances of the arms of the bridge fulfill the relationship
WZr =

(8)
During the perturbation, the dissociation field effect increases the
conductivity of the sample cell, resulting in a typical relaxation
signal given in Figure 2b.
Physical Model. To interpret the transient conductivity decrease (Figure 2c) observed in colloidal systems, let us consider
one of the spherical particles of colloidal size (radius r > 100 A)
carrying a free charge Q, suspended in a medium of conductance
ul and dielectric constant t l between the plates of a capacitor
(Figure 3a). The particle itself has a conductance of u2 and a
dielectric constant c2. The medium is a dilute electrolyte, and the
overall system is electrically neutral. In the absence of an electric
field, the Coulombic interaction between the particle and the ions
will be balanced by their thermal motions after the formation of
an electrical double layer around the particle. At equilibrium,
the net fluxes of both positive and negative ions toward the sphere
are zero, J* = 0. If the charge Q of the particle is due to
chemisorbed ions, the double layer may be broken down into two
portions.25 The dense part at the sphere surface is the layer of
tightly bound (chemisorbed) ions, which essentially determines
Zl/ZZ

(25) Dukhin, S. S. Surf. Colloid Sci. 1973, 3, 83.

the surface potential of the particle. The outside diffuse part is

made up of the supporting counterions, which are held by Coulombic attraction.
Upon the application of a homogeneous, constant electric field
?!, between the plates of the capacitor, the particle undergoes rapid
polarization, during which process work w

is done on the system (cf. (2)). Besides the negligibly small, very
rapid electronic and atomic polarization of the particle, the relevant
polarization mechanisms are all migratory in nature;26 Le., they
involve the movement of ions over one or more molecular distances
or lattice sites. The most important polarization mechanisms are
the Maxwell-Wagner (interfacial) p o l a r i z a t i ~ npolarization
,~~~~~
by surface cond~ctivity,2~
and bound layer30 and diffuse layerz5
polarization.
Maxwell-Wagner polarization occurs if the conductivity of the
medium is different from that of the particle (cf. Figure 3b). If
ul < a2,charges arrive from the particle interior faster than they
can be carried away toward the electrodes through the solution.
If u1> u2,charges arrive from the solution at the interface faster
than they can be transported through the solid. In both cases a
charged interface results, with charge accumulation at the two
polar regions of the sphere facing the electrodes. For the systems
under investigation a1 > az is the case, resulting in a situation
depicted in Figure 3b. The relaxation time of the polarization
process is given by3
2q
7Mw

= 2a1

+ 2 - v(t2 + a2 - v(a2 - a1)

1)

(10)

(where u denotes the volume fraction of the spheres), and the final
effective induced dipole moment pMwbyz5

Surface conductivity is a frequency-independent mechanism

by which charge is transferred from one side of the sphere to the
other, along the surface,32within a thin shell of thickness d and
conductivity a, (Figure 3c). Polarization results only if a, > ul.
Surface conductivity can be treatedz9as a part of the total conductance u2/ of the particle
ai = a2

+ 2u,d/r

(12)

If it is assumed that the outer surface of the conducting shell

provides still another surface upon which charge accumulation
can occur in the Maxwell-Wagner sense, for small volume fraction
v, the relaxation time for the surface-conductance modified interfacial polarization is given by33
2

rMws = a2 + 2A,/r

+ 2tl - U ( 2 -

1)

+ 2a1 - u(u2 + 2AJr

- al)

(13)

where A, = a,d is the effective surface conductance of the shell.

The anticipated effective induced dipole moment p of a charged
sphere with a surface conductance is found34to be

( ;+ ; :7)

p = po + p , = qr3E - -

-1

(14)

where po and ps are the induced dipole moment components of

(26) Pohl, H. A. Dielectrophoresis;Cambridge University Press: Cam-

bridge, UK, 1978; and references therein.

(27) Maxwell, J. C. A Treatise on Electricity and Magnetism, 2nd ed.;

Clarendon Press: Oxford, UK, 1881; Vol. 1 .
(28) y g n e r , K. W. Arch. Electrotech. (Berlin) 1914, 2, 271.
(29) 0 Konski, C. T. J . Phys. Chem. 1960,64, 605.
(30) Schwarz, G. J . Phys. Chem. 1962, 66, 2636.
(31) van Beek, L. Progr. Dielectr. 1970, 7, 69.
(32) Miles, J. B.; Robertson, H. P. Phys. Rev. 1932, 40, 583.
(33) Schwan, H. P.; Schwarz, G.; Maczuk, J.; Pauly, H. J . Phys. Chem.
1962, 66, 2626.
(34) Dukhin, S.S.; Shilov, V. N. Ado. Colloid Znterface Sci. 1980, 13, 153.

The Journal of Physical Chemistry, Vol. 88, No. 6, 1984 1155

Apparent Negative Second Wien Effect

the sphere uncharged and that due to charge flow on the surface,
respectively, and y = u,/u,r. When y is very small and u1 > u2,
the surface conductance component of the dipole is small and the
particle polarizes by the Maxwell-Wagner mechanism, opposite
to the field. When y >> 1, the charge component dominates and
the particle polarizes in the direction of the field. At some value
of y the induced dipole moment is zero, which is called the isopolarization state.
Bound and diffuse layer polarization both represent mechanisms where ionic motion tangent to the surface is relatively
unhindered, but diffusion away from the surface requires overcoming a considerable activation barrier. Thus, these mechanisms
predict a sizeable polarization with little or no direct contribution
to the dc conductivity, but they increase the surface capacitance.
Several authors considered the possibility of each of the bound
and diffuse layer polarizations occurring alone; however, Dukhin
et alesanalysis34suggests that both must be present and that the
latter is more important. They found the expression for the
relaxation time of polarization as
TBDL

= r2/2DJ4

(15)

where D, is the coefficient of bound ion diffusion and M denotes

hf = 1 + Z + z - ( Z +

+ Z-)Kpb/co

(16)

where z+ and z- are the ionic charges, K is the Debye-Hiickel

reciprocal distance, & represents the surface charge equilibrium
density of the bound ions, and co is the ion concentration in the
bulk solution. For the induced dipole moment due to doublelayer
polarization, the following expression was derived:34

where

3 = [e~p(e,+~/ZkT)+ 3m exp(eo{/2kT)]/~r

(18)

with the dimensionless parameter m being

m = el(kT)2/6~qDeo2

(19)

In eq 18 and 19, qd represents the potential at the outer Stern

plane, {the zeta potential, 7 the viscosity of the medium, and D
the mean diffusion coefficient of the counterions.
Comparison of the several theories shows that simple bulk
factors such as considered in the Maxwell-Wagner approach, even
in the surface-conductance modified version, are inadequate to
account for the experimental facts of the anomalous giant polarization of suspensions in polar solvents.26 The more successful
treatment^^^,^^ involve the consideration of the electrically flexible
parts of the ionic double layer, with the
that the major
effect is the polarization of the double layer, where the contribution
of the bound layer is muted by that of the diffuse portion (Figure
3d).
The net result of the sudden (in less than 0.1 ps) application
of the external dc field, at this point, is the production of giant
dipoles aligned in the direction of the field. For particles of r <
80 nm, with the systems and experimental conditions of ref 18-18,
all the mechanisms discussed above predict a relaxation time of
polarization T~~ < 2 ps, and for r = 500 nm, T~~ < 50 fis. If the
polarization of our model system was the only effect, to a good
approximation, it would only increase the capacitance of the
sample cell (Figure l ) , which would result in a pulse-shaped
response (Figure 2d) of the Wheatstone bridge, with the rise time
of the first pulse approximately equal to the relaxation time of
polarization T ~ ~ .
During the formation of the giant particle dipoles, the free ions
in the solution are exposed not only to the external field but also
to the local fields of the dipoles. The concerted effect of the fields
is the preferred motion of the ions toward the oppositely charged
ends of the dipoles where they adhere to the particles (Figurg 4).
The details of the sorption process with respect to the depth of
penetration of the double layer by the ions is less important than
the fact that the mobility of a significant number of originally

Figure 4. Schematic representation of the sorption of ions on an induced-dipole particle (positively charged). Charges in the double layer
represent excess local charges.

z
I

Figure 5. Illustration of a permanent-dipole particle oriented by an

external field. The charges within the equipotential surface re represent
excess local charges. The two surfaces re are not spherical, in contrast
to the simplified drawing given.

free ions is drastically reduced. Either because they actually

adsorb at the bound layer or because they become incorporated
in the diffuse layer, there is a decrease of free charge carriers.
We propose this field-induced increased adsorption of ions to be
the source of the transient conductivity decrease observed (Figure
2c). The amount of ions adsorbed should be proportional to the
induced dipole moment p of the particles and the rate 7-l of their
adsorption to the average bulk ion current I in the solution, if the
polarization is very fast. Since both p and I , to a good approximation, linearly increase with the field E, one expects both the
amplitude and the rate of the conductance decrease to do so too,
as was borne out by experiments (see Figures 2 and 3 of ref 16).
During the binding of ions, the strength of the dipoles is
decreasing because of partial charge neutralization at the poles.
If no other processes (second Wien effect) occurred simultaneously,
the system would reach a steady state (in contrast to an equilibrium state in the absence of an external field) with respect to
the thermal motions and fluxes, characterized by div J* = 0 for
both positive and negative ions. The expected signal for this
situation is depicted by the dashed line in Figure 2c.
The Binding Constant of Free Ions to a Dipole. It is instructive to investigate how the number of ions associated with
a dielectric spherical particle changes upon its polarization.
Following Onsager, under associated we mean that the potential
energy of interaction between the two moieties is greater than 2kT.
For illustration, we shall investigate only the equilibrium case,
Le. how a permanent dipole (without an external field) binds ions.
Thus, in essence, we are comparing two systems: a spherical,
uncharged particle with a spherical permanent-dipole particle,
each suspended in a dilute electrolyte solution. In the first system,
clearly, no ions are associated with the particle.
In the second system, with reference to Figure 5 , if the dipole
particle consists of two point charges *Q that are located at the
surface, its dipole moment is p = 21Q. Since the ions cannot
penetrate the sphere, we are interested only in the outside region,
where the potential of the dipole is

+ = ( Q / 4 i ~ e ~ q ) ( r-~r2-l)
-~

(20)

which according to the law of cosines can be written as

+ +

+(r,O) = (Q/4iTol)[(r2 Iz - 21r cos 0)-1/2 - (r2 I2

21r cos ~ ) ] - l /(21)
~
Although, in the bulk solution, where r

>> I, eq 21 reduces to

+ = 2Q1 cos 0 / h r 2 e o q

(22)

in our case, close to the surface, this approximation cannot be used.

1156 The Journal of Physical Chemistry, Vol. 88, No. 6,1984

Since the potential energy of an i-type ion with charge zpo is just
E, =

(23)

Z$O*

by setting E, = 2kT we can obtain an equipotential surface re(O,

$ = constant), within which ions are considered bound. To find
re, eq 21 must be expanded as a polynomial in r a n d solved, after
setting J/ = 2kT/z,eo. Then, if the bulk number density of the
i-type ions in the solution is oni = ,JVi/V,the number iN,l of the
bound i-type ions per dipole can be obtained from
iN,l =
8 s " s r e ~ r ' 2 0 nexp(-z,eoJ/(r,0)/kT)r2
i
sin 0 d r d0 d 4 (24)
8=0

r=l

Schelly and Astumian

signal (solid line in Figure 2c). It is important to note that this
relaxation takes place between two steady states. In the colloidal
systems studied,l6-l8 reaction 5 occurs on the surface of the
suspended particles, involving ionic groups permanently attached
to the surface. Thus, the particles S themselves are part of the
weak-electrolyte equilibrium shifted by the second Wien effect.
Hence, if all the possibly different types of cations and anions
present are denoted by C+ and A-, respectively, eq 5 can be
rewritten for the colloidal systems as
S-C-A

k
2
(S-C+-A-)
k-l

ion pair

S-C+

+ A-

(26)

k-2

$=O

Since the problem has cylindrical symmetry with respect to the

y axis and planar symmetry with respect to the xz plane, the
integration is carried out only for the first octant. The angle
runs around the axis of the dipole. The integration limit Be
corresponds to the point where the equipotential surface intersects
the particle.
If the total number of dipoles in the solution is Nd, the number
of bound i-type ions is ii?b= NdiN<and the total number of bound
ions is N d c i i N { . Thus, the "binding constant* can be written
as the fraction
Kb

[bound ions]
[free ions]

in which the reaction is, at least physically (in case sometof the
free ions in the solution are different from those in (26)), but
possibly also chemically, coupled with the reaction
CA&C+
k-3

+ A-

If K3 is large, as for a strong electrolyte, the applied field has

negligible effect on equilibrium 27 and thus on its contribution
to the conductance change. The field, however, as we have shown,
rapidly transforms the particles to large dipoles S+ (Figure 3d),
and so, in an electric field, eq 26 should be written as

Ndcrly,'
i

CiNo- NdciNb)
i

K3

(25)

Thus, increasing the dipole moment from zero t o p increases the

binding constant from zero to Kb.
If the permanent dipole in equilibrium with its two different
ionic atmospheres at the poles is exposed to an external dc field,
the dipole will orient in the direction of the field. During the
rotation, the ionic atmospheres are somewhat distorted, and also
additional induced polarization may occur. Furthermore, note
that in the final aligned position, the two ionic atmospheres face
the likecharged electrodes, which can be contrasted to the situation
depicted in Figure 3d. With respect to the ion binding, in this
hypothetical model, the electric field will shift the equilibrium
toward increased association, which would correspond to a true
negative second Wien effect.
In case of a field-induced dipole of the same magnitude p
(instead of starting with a permanent dipole), the final steady state
reached will be the same, however, with an even larger expected
true negative second Wien effect. A detailed analysis of dipolar
systems will be published in a forthcoming paper. At this point
we only would like to mention that the dipolar mechanisms we
have outlined may provide an interpretation for the hitherto
unexplained negative second Wien effects in homogeneous systems,
which had been observed in ~ t a t i c as
~ ~well
9 ~as kinetic context^.^'
Chemical-Physical Model. Let us now turn back to the relaxation signal in Figure 2c. During the application of the external
field, simultaneously with the polarization and ion-sorption processes also, the, in this case slower, chemical reactions (eq 5) take
place because of the second Wien effect. As mentioned before,
ions are produced during the chemical relaxation, and the resulting
increase of conductance ultimately may become dominant, which
is responsible for the conductance-increase part of the observed
(35) Bailey, F. E.; Spinnler, J. F.; Patterson, A., Jr. J . Am. Chem. SOC.
1961,83, 1761.

(36) Spinnler, J. F.; Patterson, A., Jr. J . Phys. Chem. 1965, 69, 500.
(37) Cole, D. L.; Eyring, E. M.; Rampton, D. T.; Silzars, A,; Jensen, R.
P.J . Phys. Chern. 1967, 71, 2771.

ki

S?-C-A + (Sf-C+.**A-)
k-1
ion pair

kz

S2-C'

+ A-

(28)

k-2

It is important to remember that the and - symbols on S?

represent a large number of elementary charges and that the two
polar regions of the particle are not equivalent. They differ in
polarity, surface potential, local charge density, ionic concentrations, etc., and the two polar regions are separated by a large
distance on the molecular scale. While oppositely charged ions
are deposited at the poles by the field (Figure 4), at rates of the
ion currents in the solution, simultaneously the second Wien effect
(- in (28)) produces ions at the slower rates of the chemical
relaxation(s) of (28), governed by the rate constants. Naturally,
the deposited ions can ultimately reach the opposite sides of the
particle through surface conductance, where a fraction of them
may desorb because of the favorable direction of the field, until
the overall process reaches a steady state. In this final steady state,
the conductance does not change any more (Figure 2c); however,
the fluxes are maintained by the continuous input of energy, until
the termination of the perturbing square pulse.
Generalization to Homogeneous Systems. It seems that the
role of the particle S can also be assumed by a molecular moiety
R of large polarizability a. In the case of a homogeneous system,
the corresponding stoichiometric equations can be obtained by
simply replacing S by R in (26) and (28). Systems involving
biological macromolecules with many adsorption sites, polyelectrolytes, micelles, etc. seem to be possible prime candidates for
the observation of apparent negative second Wien effects in homogeneous solutions.
Acknowledgment. This work was partially supported by the
R. A. Welch Foundation and the Organized Research Fund of
UTA. Acknowledgment is made to the donors of the Petroleum
Research Fund, administered by the American Chemical Society,
for partial support of this work and to the Japanese Ministry of
Education for a visiting research scholarship to R.D.A. We are
grateful for stimulating discussions with Professors H. Berg and
T. Yasunaga, as well as Dr. K. Hachiya, M. Sasaki, and C. R.
Savage.