Effect of high magnetic field on the silver deposition from its aqueous solution

Article abstract of DOI:10.1021/jp971489u

The effects of a high gradient magnetic field on the Ag⁺/Cu liquid-solid redox reaction were studied. A copper wire was placed on a sheet of chromatography paper wet with silver nitrate aqueous solution. At zero field silver dendrites grew uniformly around the wire. In the presence of a gradient magnetic field (ca. 380 T²m^-1), the dendritic pattern changed drastically. The yield of silver metal increased by about 45%. These results are attributable to the convection of aqueous solution on the surface of the paper which is induced by the magnetic force on the paramagnetic copper ions generated in the reaction.

Full text of DOI:10.1021/jp971489u

J. Phys. Chem. A 1997, 101, 7359-7363
7359
Effect of High Magnetic Field on the Silver Deposition from Its Aqueous Solution
Yoshifumi Tanimoto,*,† Akio Katsuki,‡ Hiroshi Yano,† and Shin-ichi Watanabe†
Faculty of Science, Hiroshima UniVersity, Higashi-Hiroshima 739, Japan, and Faculty of Education,
Shinshu UniVersity, Nagano 380, Japan
ReceiVed: May 2, 1997; In Final Form: July 14, 1997X
The effects of a high gradient magnetic field on the Ag+/Cu liquid-solid redox reaction were studied. A
copper wire was placed on a sheet of chromatography paper wet with silver nitrate aqueous solution. At
zero field silver dendrites grew uniformly around the wire. In the presence of a gradient magnetic field (ca.
380 T2m-1), the dendritic pattern changed drastically. The yield of silver metal increased by about 45%.
These results are attributable to the convection of aqueous solution on the surface of the paper which is
induced by the magnetic force on the paramagnetic copper ions generated in the reaction.
1. Introduction
The effect of high magnetic field (>2 T) on chemical
reactions is one of the current topics in basic and applied
chemistry. This is partly because superconducting magnets are
becoming popular today and partly because unexpected magnetic
phenomena which cannot be observed in a low magnetic field
have been discovered in a high magnetic field and these new
phenomena have potential for industrial application.
The effects of high magnetic field (<14 T) on photochemical
reactions of organic compounds have been investigated very
extensively in the last five years.1 The reversal of the magnetic
field effects (MFEs) on the lifetime of short-lived intermediates
has been observed under high magnetic field (>2 T). The
effects of high magnetic field on chemical equilibrium of
hydrogen-ferromagnetic compounds have been reported by
Yamaguchi and his collaborators.2 The Gibbs free energy
change which is induced by a magnetic field results in a change
in the equilibrium constant. The magnetic orientation of fibrin
fiber,3,4 polymers,5 and organic crystals6,7 has been also reported.
Recently, Mogi et al. have reported that the deposition pattern
of silver metal from Ag+/Cu redox reaction is affected by a
high homogeneous magnetic field (8 T).8 This is interpreted
in terms of a magnetohydrodynamics mechanism in which diffu-
sion of ions is affected by a Lorentz force. The electrodeposition
pattern of silver is also affected by a high magnetic field.9
In this paper, we report the effects of a high gradient magnetic
field on the Ag+/Cu liquid-solid redox reaction.10 The dendritic
pattern and yield of silver metal deposited by the reaction are
affected significantly by the high gradient field. The results
are explained in terms of magnetophoresis: A high gradient
magnetic field induces convection in the thin layer of solution
by attracting paramagnetic copper ions generated by the reaction.
Figure 1. Top: The distribution of the magnetic field B(z). The squares
are the observed values, and the solid line is a Gaussian distribution
obtained using a least-squares method, respectively. “z” is the distance
along the magnetic axis from the center of the magnetic field. The
maximum intensity Bmax is 8 T at z ) 0 mm in our experiment.
Bottom: The outline of experimental handling: (a) a copper wire (6 φ
× 250 mm), (b) a piece (37 mm × 300 mm) of chromatography paper
soaked with silver nitrate solution (0.5 mol dm-3), and (c) a plastic
vessel (40 mm × 380 mm × 10 mm).
m-1 at z ) 50 mm. In the following, for simplicity, Bmax is
used as the representative of the magnetic field strength.
Silver nitrate (Nacalai, GR grade) and chromatography paper
(Whatman, 1Chr) were used as supplied. Deionized water was
used. Copper wire (99.9%, 6 φ) and zinc wire (99.99%, 6 φ)
were polished mechanically and chemically before use.
Liquid-solid redox reaction was carried out in a plastic vessel
(37 mm × 380 mm × 10 mm) at room temperature (Figure 1).
A copper wire (6 φ × 250 mm) was placed on a piece of
chromatography paper (40 mm × 300 mm) which was wet with
a silver nitrate solution (0.5 mol dm-3, 0.4 mL), as depicted in
Figure 1. The vessel was then placed in a bore tube of the
magnet so that the center of the vessel coincided with that of
the magnetic field. A silver nitrate solution was soaked into a
piece of chromatography paper, since a fluid solution is attracted
to or repelled from a high gradient magnetic field due to its
magnetic susceptibility.11
2. Experimental Section
The magnetic fields were applied using a superconducting
magnet (Oxford Instrument, Spectromag-1000) whose bore tube
(50 φ × 370 mm) was installed horizontally. Distribution of
the magnetic field B(z) was approximated with a Gaussian
distribution (Figure 1), the direction (z-axis) of the magnetic
field being taken to be along the bore tube. The maximum
strength of the magnetic field (Bmax) is 8 T at z ) 0 mm and
that of the gradient magnetic field, B(z) dB(z)/dz, is ca. 380 T2
The dendrite of silver metal deposited on the paper was
washed twice with dilute hydrochloric acid, 0.1 mol dm-3
sodium thiosulfate solution, and water and was dried.
† Hiroshima University.
‡ Shinshu University.
X Abstract published in AdVance ACS Abstracts, September 15, 1997.
S1089-5639(97)01489-8 CCC: $14.00 © 1997 American Chemical Society

7360 J. Phys. Chem. A, Vol. 101, No. 40, 1997
Tanimoto et al.
For the purpose of comparison, the effects of the high gradient
magnetic field on the Ag+/Zn redox reaction were examined
analogously.
3. Results
3.1. MFE on the Deposition Pattern of Silver Metal and
Copper Ion. The liquid-solid redox reaction studied here is
given by the following equation:
2Ag+ + Cu f 2AgV + Cu2+
(2)
Because of the order of ionization tendency, the dendrite of
silver metal deposits from a silver nitrate solution when a copper
wire is placed on the paper. Figure 4 shows the effects of a
magnetic field (Bmax ) 8 T) on the deposition pattern of silver
metal as a function of time. At zero field, silver metal deposits
uniformly around the copper wire with time. When the reaction
is carried out in the magnetic field, the dendritic pattern of silver
metal is affected drastically: The silver metal does not deposit
in the neighborhood of the center of the wire after the 1 h
reaction. The dendritic pattern of the metal at z ) ca. 75 mm
changes remarkably, as shown in Figure 4, although, in the
absence of the magnetic field, the dendrite grows almost
perpendicular to the long axis (z-axis) of the wire. In the
presence of the magnetic field the dendrite tilts to the high
magnetic field. The MFE was observed only in the high
magnetic field (Bmax g ca. 5 T).
Figure 2. Experimental setup for size effects of chromatography paper
and copper wire on the yield of copper ions. (a) Small pieces of copper
wire (6 φ × 15 mm) were placed on small pieces of the paper (37 mm
× 20 mm) at z ) 0, 50, and 115 mm. (b) A copper wire (6 φ × 250
mm) was placed on pieces of the paper (37 mm × 20 mm) at the same
positions. (c) A copper wire (6 φ × 250 mm) was placed on a long
piece of the paper (37 mm × 300 mm). In cases a and b the relative
concentration of copper ions was determined at each piece of paper in
the absence and presence of a magnetic field (Bmax ) 8 T). In the case
of c, after the reaction the paper was cut as indicated by broken lines
in the figure, and then the copper concentrations were determined
analogously.
Figure 5 shows the effect of the magnetic field (Bmax ) 8 T)
on the Cu2+ ion distribution pattern as a function of time. At
zero field the Cu2+ ions distribute uniformly around the copper
wire. When the magnetic field is applied, the distribution pattern
of Cu2+ changes. Even after 10 min of reaction, the Cu2+ ions
shift to the neighborhood of the center part of the paper, which
coincides with the center of the magnetic field.
For the purpose of comparison, similar solid-liquid redox
reactions shown below were examined:
Cu2+ + Zn f CuV + Zn2+
2Ag+ + Zn f 2AgV + Zn2+
(3)
(4)
Figure 3. Experimental setup for the electromotive force. Two sets
of copper wire (6 φ × 50 mm × 4) were placed on a piece of
chromatography paper at z ) 0 and z ) 125 mm.
The distribution of the copper ion generated was detected
using the following reaction,
In the case of reaction 3, the deposition of copper metal shifts
significantly to the neighborhood of the center of the paper.12
Figure 6 shows the MFE on the deposition pattern of silver
metal generated from reaction 4. In marked contrast to the
results of reactions 2 and 3, a high magnetic field (Bmax ) 8 T)
does not affect reaction 4.
3.2. MFE on the Yields of Silver Metal and Copper Ion.
The MFE on the yield of silver metal was determined by
gravimetry. The yield after the 1 h reaction is 53 ( 7 mg at
zero field, whereas it is 77 ( 10 mg in the presence of a
magnetic field (Bmax) 8 T). The yield increases by ca. 45%.
Figure 7 shows the effect of the magnetic field (Bmax ) 8 T) on
the yield of the copper ion. In accordance with the change in
the silver metal yield, the yield of the copper ion after the 1 h
reaction increases by about 36% in the magnetic field. The
Ag+/Cu redox reaction is enhanced by application of the
magnetic field. It is interesting to note that, by application of
the magnetic field, the net yields of both silver metal and copper
ion increase, even though the deposition of silver is localized
at the two ends of the wire.
Cu2+ + 4NH3 f [Cu(NH3)4]2+
(1)
Since the tetraammine copper complex ([Cu(NH3)4]2+) had an
absorption band at ca. 630 nm, the area of the paper where Cu2+
ion existed became bright blue. Actually, after the redox
reaction, the chromatography paper was brought into contact
with another chromatography paper wet with aqueous ammonia
so that the copper ion on the paper reacted with ammonia before
diffusion.
The yield of silver metal was determined by gravimetry. The
relative yield of Cu2+ was determined using reaction 1. After
the redox reaction, the solution adsorbed by the chromatography
paper was extracted with water. Ammonia water was added to
the extract, and the relative yield of copper ion was calculated
from the absorbance of the copper-ammine complex at 630
nm.
Size effects of the chromatography paper and copper wire
were examined under the experimental condition schematically
shown in Figure 2.
The electromotive force was measured using a high-imped-
ance electric recorder. As shown in Figure 3, the two pieces
of copper wire (6 φ × 50 mm × 4) were placed on a piece of
chromatography paper at z ) 0 and z ) 125 mm. Magnetic
field strengths were 8 (z ) 0 mm) and 2 T (z ) 125 mm).
For the purpose of comparison, the yield of silver metal
deposited from the Ag+/Zn redox reaction (eq 4) was also
examined. Since the dendrite deposited from this reaction sticks
to both the zinc wire and chromatography paper, the yields could
not be determined accurately. The yields were obtained to be
64 ( 12 mg at zero field and 74 ( 21 mg at 8 T after 1 h of
reaction. Taking these large standard deviations into account,

Ag Deposition from Its Aqueous Solution
J. Phys. Chem. A, Vol. 101, No. 40, 1997 7361
Figure 4. Deposition of silver metal from the Ag+/Cu reaction in the absence (left) and presence (right) of a magnetic field (Bmax ) 8 T). Reaction
time is indicated on the left side.
Figure 5. Distribution of copper ions for the Ag+/Cu reaction in the absence (left) and presence (right) of a magnetic field (Bmax ) 8 T). Reaction
time is indicated on the left side.
Figure 6. Deposition of silver metal from the Ag+/Zn reaction in the
absence (top) and presence (bottom) of a magnetic field (Bmax ) 8 T).
Reaction time is 1 h.
the MFE on the deposition yield of silver metal for the Ag+/Zn
Figure 7. Relative yields of copper ions vs time for the Ag+/Cu
reaction in the absence and presence of a magnetic field (Bmax ) 8 T).
redox reaction seems insignificant, if any.
3.3. Influence of the Sizes of Chromatography Paper and
Copper Wire on the Yield of Copper Ion. To examine the
influence of the sizes of the paper and the wire, the yield of
Cu2+ was observed under three different experimental conditions
(Figure 2). The results are summarized in Table 1. (a) When
small pieces of the wire (6 φ × 15 mm) are placed separately
on the paper (37 mm × 20 mm) (Figure 2a), the magnetic field
does not affect the relative yield of Cu2+. (b) When a copper
wire (6 φ × 250 mm) is placed on the three pieces of paper,
placed separately as depicted in Figure 2b, the magnetic field
does not affect the yield of Cu2+. (c) Only when the long piece
(37 mm × 300 mm) of paper is used (Figure 2c) does the yield
increase at the high magnetic field region (z ) 0 mm), whereas
it decreases at the low magnetic field region (z ) 115 mm).
3.4. Electromotive Force. The present redox reaction in
the gradient magnetic field generates an electromotive force. It
was measured using the experimental setup depicted in Figure
3. Two sets of copper wire (6 φ × 50 mm × 4) are placed at
the positions of z ) 0 and z ) 125 mm on a piece of
chromatography paper wet with silver nitrate aqueous solution.
The results are shown in Figure 8. In the absence of a magnetic
field no electromotive force is generated. When a magnetic

7362 J. Phys. Chem. A, Vol. 101, No. 40, 1997
Tanimoto et al.
Figure 8. Electromotive force for the Ag+/Cu redox reaction induced by a magnetic field (Bmax ) 8 T). A closed box around “8 T” indicates the
time period when the reaction system is exposed to the magnetic field. Unless otherwise noted, the reaction is carried out at zero field. Experimental
setup is given in Figure 3.
TABLE 1: Influence of the Size of the Paper and Copper
Wire on the Relative Yields Φ(8 T)/Φ(0 T) of Copper Ion.
Φ(8 T) and Φ(0 T) Are the Yields in the Presence and
Absence of a Magnetic Field (Bmax ) 8 T). For
Experimental Setup, See Figure 2
reaction is carried out on a small piece of separate paper.
Furthermore, the electromotive force between the two sets of
copper wire shown in Figure 3 is estimated to be 2.0 × 10-6
V
from the above-mentioned free energy difference. This value
is too small to explain the observed electromotive force (ca.
200 mV) shown in Figure 8.
z/mm
B(z)/T
0
8
50
6.6
115
2.6
The MHD force FMHD is given by the following equation:
case a
case b
case c
0.98
1.02
1.63
0.95
1.07
1.43
0.92
0.92
0.46
FMHD ) eW × B
(6)
where e is the electric charge, W is its velocity, and B is the
homogeneous magnetic field. The MHD force FMHD gives
charged species torque perpendicular to both W- and B-axes.
When W and B are perpendicular to each other, FMHD is
maximum, whereas FMHD is zero when W and B are parallel.
Mogi et al. have reported that a high magnetic field affects the
deposition pattern of this Ag+/Cu redox reaction.8 In their
experiment, the direction of the magnetic field is perpendicular
to the motion of ions. In the present experiment, the direction
of the flow of ions is parallel to that of the magnetic field as a
whole. It might also be possible to consider that the flow in
the vicinity of the wire might take an important role in the
MFEs. In this case, the flow is considered to be perpendicular
to the magnetic field from the dendritic pattern at zero field
shown in Figure 4. Under this circumstance, the force is
perpendicular to the surface of the paper, and therefore, the
MHD force does not induce a convective flow in the present
redox reactions. Furthermore, the Ag+/Zn reaction should be
affected by the external magnetic field, if the MHD mechanism
is operative. Contrary to this prediction, the magnetic field does
not affect the dendritic pattern deposited from the Ag+/Zn redox
reaction (Figure 6). Therefore, the MHD mechanism does not
explain the experimental results shown in Figures 4 and 5.
Magnetic properties of reactions 2 and 3 are different from
that of reaction 4. In reaction 2, diamagnetic reactants (Ag+
and Cu) generate a paramagnetic product (Cu2+), and in reaction
3, a paramagnetic reactant (Cu2+) changes into a diamagnetic
product (Cu). In reaction 4, which does not exhibit MFEs, both
reactants and products are diamagnetic. Thus it is reasonable
to consider that the MFEs observed in reaction 2 are associated
with the paramagnetic property of the product (Cu2+). Fur-
thermore, as shown in Table 1, the redox reaction carried out
on a small piece of paper exhibits no MFE (cases a and b). In
these cases, the difference in the magnetic field strengths ∆B
at the two ends of the wire is small and the reaction is considered
to be carried out in a locally homogeneous magnetic field. In
case c, which exhibits significant MFEs, ∆B between the center
and edge of the wire is about 6 T, and therefore, the reaction is
carried out under a high gradient magnetic field as a whole.
Since not a homogeneous field but a gradient one affects reaction
2, the magnetic force is considered to be operative in the MFEs
shown in Figures 4 and 5.
field (Bmax ) 8 T) is applied, an electromotive force of ca. 200
mV is generated between the two sets of identical wire.
The electromotive force for the Ag+/Zn redox reaction was
also examined. The force for this reaction is less than 20 mV
under the analogous condition. The electromotive force gener-
ated from the Ag+/Zn reaction is insignificant, as compared with
that generated from the Ag+/Cu reaction.
4. Discussion
Now let us consider the mechanism of MFEs mentioned
above. In general there are three mechanisms to explain the
observed effects. (a) The first is the difference in the Gibbs
free energies in the magnetic fields,2 (b) the second is the
magnetohydrodynamics (MHD) mechanism,8 and (c) the third
is the magnetic force caused by a gradient magnetic field
(magnetophoresis).13
The Gibbs free energy change ∆G for reaction 2, which is
induced by a magnetic field, is
∆G ) (-1/2µ0)∆øVB(z)2
(5)
where ∆ø is the difference between magnetic susceptibilities ø
of ions and metals of eq 2, and V and µ0 are the volume of the
sample and magnetic permeability, respectively. B(z) is the
magnetic field strength at the position z. The difference in the
Gibbs free energy ∆G at the center and edge of the wire may
induce a shift of chemical equilibrium of the redox reaction.
By assuming that molar magnetic susceptibilities of copper,
copper ion, silver, and silver ion are - 4π × 5.46 × 10-12
,
+4π × 1.28 × 10-9, - 4π × 2.05 × 10-11, and - 4π × 2.4
× 10-11 m3 mol-1 (all values are described using SI units),
respectively, and magnetic fields at the center and the edge of
the wire are 8 and 2 T, respectively, ∆G’s are estimated to be
-1.04 × 10-1 at 8 T and -6.47 × 10-3 J at 2 T, respectively,
when all silver ions (0.5 mol dm-3) react with copper metal
(0.25 mol dm-3). Thus, the difference in free energy change
between the center and the edge of the wire is -9.8 × 10-2 J.
This free energy difference is too small to explain the MFE
shown in Figures 4 and 5. As shown in Table 1, the yield of
copper ion is independent from the magnetic field when the

Ag Deposition from Its Aqueous Solution
J. Phys. Chem. A, Vol. 101, No. 40, 1997 7363
The magnetic force FGMF induced by the gradient magnetic
field is given by
Faraday constant, and m1 and m2 are the concentrations of the
solution at the electrode 1 and 2, respectively. As shown in
Figure 8, E is ca. 200 mV in the present system. From eq 8,
the concentration ratio m1/m2 is estimated to be ca. 2400. This
value seems too large to explain the MFEs shown in Figures 4,
5, and 7. If this is the case, the observed MFEs should be very
significant. Therefore, it is unlikely that magnetic-field-induced
convection results in a concentration cell. In case 2, the
observed emf may be explained as follows: In the presence of
a high gradient magnetic field, the solution in the vicinity of
the two sets of wire flows from the low field (2 T) to the high
field (8 T), since it receives attractive force from the gradient
field. This convection of a paramagnetic solution generates the
emf in the gradient magnetic field. In Figure 8, upon application
of the magnetic field, the voltage increases up to its maximum
value with a time constant of ca. 10 s. This time constant seems
to reflect a time by which the convective flow reaches a steady
state. The induced electric current is estimated to be about 4
× 10-6 A, since the emf is about 200 mV and the impedance
between the two sets of wire is about 50 kΩ. This corresponds
to a flow of 1.5 × 10-7 mol of positive ions per 1 h. If the
electric charge of Ag+ is the origin of the electric current, this
value corresponds to a flow of 1.6 × 10-5 g of Ag+ per 1 h.
Since, in the experiment shown in Figure 4, the amount of the
deposited Ag metal is (5-7) × 10-2 g after 1 h of reaction, a
flow rate of 1.6 × 10-6 g h-1 of Ag+ between the two sets of
copper wire seems attainable under the experimental condition
shown in Figure 3. Thus, it is conceivable that the convective
flow generates directly the emf shown in Figure 8.
FGMF ) (ø/µ0)B(z) dB(z)/dz
(7)
Silver ion and water are diamagnetic, whereas copper ion is
paramagnetic. Silver ion and water are repelled by the gradient
field, whereas copper ion is attracted by the field. Suppose the
maximum value of magnetic field B(z)(dB(z)/dz) is estimated
to be ca. 380 T2 m-1 (z ) 50 mm, see Figure 1), the magnetic
force for a 0.25 mol dm-3 aqueous solution of Cu2+ is estimated
to be 1.2 N dm-3, whereas that for a 0.5 mol dm-3 solution of
Ag+ is estimated to be -4.6 × 10-2 N dm-3. The force for
the Cu2+ solution is about 1/8 of gravity. Thus, the paramag-
netic copper ion is the key compound for the MFEs shown in
Figures 4 and 5. As shown in Table 1, significant MFEs are
observed only when the copper wire is placed on a long sheet
of paper. This is a strong piece of evidence that a convective
flow of the solution is induced along the z-axis by the gradient
magnetic field. Neither conduction of electric current in the
wire nor the Gibbs free energy difference takes an important
role in the MFEs.
It is considered that the surface of chromatography paper wet
with aqueous solution is covered with a thin layer of the solution.
In the absence of the magnetic field the liquid-solid redox
reaction is diffusion-controlled. When it is carried out in the
high gradient field, it is allowed to suffer from the magnetic-
field-induced convection as the reaction time proceeds. This
is because the concentration of paramagnetic copper ion in
solution increases upon the reaction. The MFEs shown in
Figures 4 and 5 are explained as follows: (1) Before the
reaction, the aqueous solution is diamagnetic. At the initial stage
of the reaction, the copper ion is generated uniformly in the
vicinity of the copper wire. (2) When the concentration of
copper ion becomes higher, the solution in the vicinity of the
wire becomes paramagnetic as a whole and then it starts to move
to the higher gradient field. This induces convection of solution
on the surface of the chromatography paper. Because of this
convection, mass transfer to the front of the reaction area (copper
metal surface and silver dendrite) is enhanced. This results in
the change in deposition pattern of the dendrite as well as the
increase in the reaction yields in the presence of the gradient
magnetic field. From the magnetic susceptibilities of water and
ions, the solution in the vicinity of the copper wire is considered
to become paramagnetic when the local concentration of copper
ion becomes higher than ca. 0.6 M. It must be pointed out that
a homogeneous solution composed of paramagnetic and dia-
magnetic compounds receives a magnetic force proportional to
the sum of magnetic susceptibilities of components. In a
homogeneous solution each component does not move inde-
pendently in the magnetic field.
In conclusion, the MFEs on silver metal deposition from the
Ag+/Cu redox reaction are attributable chiefly to the convection
of solution which is induced by the magnetic force upon
paramagnetic copper ions generated in the reaction.
Acknowledgment. This work was supported in part by
Grants-in-Aid for Scientific Research (07NP0101, 08218243,
09874157) from the Ministry of Education, Science, Sports, and
Culture of Japan.
References and Notes
(1) For a review paper, see: Tanimoto, Y.; Fujiwara, Y. J. Synth. Org.
Chem. Jpn. 1995, 53, 413.
(2) Yamaguchi, M.; Yamamoto, I.; Miura, S. Phys. Lett. A 1989, 134,
504.
(3) Torbet, J.; Freyssinet, J.-M.; Hudry-Clergeon, G. Nature 1981, 289,
91.
(4) Yamagishi, A.; Takeuchi, T.; Higashi, T.; Date, M. J. Phys. Soc.
Jpn. 1989, 58, 2280.
(5) Ito, E.; Sata, H.; Yamato, M. Mem. Fac. Technol., Tokyo Metrop.
UniV. 1993, 43, 4677.
The electromotive force (emf) generated from the Ag+/Cu
redox reaction shown in Figure 8, then, may be explained by
the convection of a solution on the surface of the chromatog-
raphy paper, since no remarkable emf is generated in the Ag+/
Zn reaction where both reactants and products are diamagnetic.
There are two possibilities of the effect of convection on the
generation of the emf. (1) The convection induces a concentra-
tion differencce between the two sets of copper wire, resulting
in the formation of a concentration cell. (2) The convective
flow between the two sets of wire induces the emf.
(6) Katsuki, A.; Tokunaga, R.; Watanabe, S.; Tanimoto, Y. Chem. Lett.
1996, 607.
(7) Sazaki, G.; Yoshida, E.; Komatsu, H.; Nakada, T.; Miyashita, S.;
Watanabe, K. J. Cryst. Growth, in press.
(8) Mogi, I.; Okubo, S.; Nakagawa, Y. J. Phys. Soc. Jpn. 1991, 60,
3200.
(9) Mogi, I.; Kamiko, M.; Okubo, S. Physica B 1995, 211, 319.
(10) A preliminary result has appeared: Katsuki, A.; Watanabe, S.;
Tokunaga, R.; Tanimoto, Y. Chem. Lett. 1996, 219.
(11) Ueno, S.; Iwasaka, M. J. Appl. Phys. 1994, 75, 7177.
(12) Tanimoto, Y.; Katsuki, A.; Watanabe, S.; Tokunaga, R. Abstracts
of IV International Symposium on Magnetic Field and Spin Effects in
Chemistry and Related Phenomena, 1996, Novosibirsk, p 39.
In case 1, the emf of the concentration cell, E, is given by
the equation14
(13) Weiss, A.; Witte, H. Magnetochemie; Verlag Chemie: Weinheim,
1973; Chapter 3.
E ) -(RT/F) ln(m1/m2)
(8)
(14) Atkins, P. W. Physical Chemistry; Oxford University Press: Oxford,
1990; Chapter 10.
where R is the gas constant, T is the temperature, F is the