A-C Methods in Interfacial Electrical Phenomena

Shizuo VEDA * and Motoko VEDA *

Synopsis

The mechanical vibration of the electrical double layer gives rise to the generation of an alternating voltage. [n the case of glass-liquid interfaces this effect can be used to measure the streaming potential by an a-c technique. For mercury-solution interfaces, an alternating current is generated, which provides new methods for double layer capacity measurements. These phenomena may also be used in various practical applications, such as mechano-electric transducers.

Introduction

Very few d-e methods have been employed in interfacial electrical measurements; most of them, for instance, the streaming potential, electrophoretic mobility measurement, etc., have been carried out by using steady-state methods (l). During the course of experiments on the interfacial electrical phenomena from 1942 to 1943, it was found by S. Ueda that the mechanical vibration of interfaces gives rise to the generation of an a-c voltage. This pheno­menon is named "U-effect" by the author (2).

In this paper stress is placed on the detailed descriptio~ of the methods of these measure­ments rather than on the presentation and discussion of the results obtained from them.

A-( Streaming Potential

The generation of the doc potential difference by the tangential displacement 'at solid -liquid interfaces is widely known as one of the electrokinetic phenomena, the streaming potential (l). The potential difference E between the two electrodes, a and b of Fig. I, is given by

E = P€~ / 41T7]A [I]

where P is the pressure applied to the liquid phase, ~ the so-called electrokinetic zeta potential, and €, 7], and A, are respectively, the dielectric constant, the viscosity, and the specific con­ductivity of the liquid.

It was found in this connection that an a-c potential difference was obtained between these two electrodes when a periodically changing pressure was employed (3). This pheno­

• Lab. Physical Chemistry of Food, Dep. of Food and Nutrition, Kinki Univ., Higashiosaka, Osaka, 577 Japan. IIH~ Toll. IUI,l!J"-ft7-III~'t)

80 Memoirs of the Faculty of Agriculture of Kinki University No .14 (I 981)

a b

+,W'=~ liquidp

·wn; JI' ') ./2...2...d-.

<- - - -" .• I ..... • Fig. 1. Glass capillary for the streaming potential generation.

menon is called "U-effect !" and is the generation of an a-c streaming potential. The poten­tial had the same wave character as the original vibration in all experiments up to a frequency as high as 25,000 cps.

Theoretical. - A simple derivation of the equation of the a-c potential, e = v"2 E exp (jwt), is made by using Helmholtz's model of an electrostatic condenser for the electric double layer at the solid-liquid interface, viz.

a = E~ /41T15 [2)

where a is the interfacial charge density and 5 the gap distance of the equivalent condenser (I). Here, j =~, t is the time, E the complex effective value, and w the circular frequency (i.e.. w = 2 1ft. f being the frequency in cps) of the a·c voltage e. If we take the x-axis in the direction tangent to the interface, the relative displacement at the interface, ax, is given by

ax = .J2 X exp Uwt) [3]

where X is the complex amplitude of vibration. The instantaneous transfer of the electric charge, i.e.. the current i, produced by this mechanical motion is given by

i =2rrra (dlu/dt) =(jWE~'X /v"2 5) exp Uwt) [4]

where r is the radius of the capillary. On the other hand the current i is also given by Ohm's law, viz.

i = e Arr r"2 / I [5]

where I is the distance of the two electrodes. Equating Eq. [4] and [5], we obtain the following expression of the a-c streaming potential

E =jwE~IX / 2rrADr [6]

This means that the a-c voltage is proportional to the zeta potential and the dielectric constant

8 F

Fig. 2. Block diagram of zeta potential measurements by U-effect I.

Shiwo ULnA and Motoko UEJ)A A- Methods in Interfa.:ial Electrical Phenomena XI

and inversely proportional to the specific cond uct.ivity of the Jiq uid. if a constant mechanical vibrat.ion of a given vibrating clement. is employed.

Experi/lu'llIal. -- As an example of streanling potent.ial measurements by U-effect I, results of experiments on a glass capillary in cont.act. wit.h cat.ionic surface active agent solutions will be given.

The experimental device is shown schematically in Fig. 2. The glass capillary element A, filled with the surface act.ive agent solution, is forced to vibrate by an electrodynamic vibrator B. which is driven by an audiofrequency oscillator C. The frequency used was 200 cps in these experiments. The output voltage (! is amplified by an amplifier 0 and read by means of a valve voltmeter E. The valve voltmeter G gives the relative magnitude of the vibrating amplitude X. The measurements of the absolute amplitude of vibration will be shown later. A fine glass rod H serves as the piston of the capillary to prevent the liquid from moving with the glass wall. The cathode ray oscilloscope F is for the purpose of observing the wave form of e.

The solut iuns used were 10-4 N K('I aq ueous solutions. containing dodecyl· or cetylpyri­dinium chloride in various concentrations. The potassium chloride was added as an indifferent electrolyte in order to compress the thickness of the diffuse double layer. since otherwise the zeta potential data would be difficult to interpret (6).

It can be seen from Eq. r6] that the zeta potential is proportional to D. when a constant vibration is employed. since E can be assumed constantin the case of dilute aqueous solu­tions as in the present experiments. The relative ~-value, i.e., ~ / s*, is therefore obtained by the ratio D. /E*A *. where the asterisk denotes the absence of surface active agents.

The S / s* - values thus obtained are plotted against the logarithms of the surface active agent concentration in Fig. 3. On the same figure are plotted. for comparison, the S/ s* ­values obtained by doc measurements. These latler measurements were carried out by the ordinary constant pressure method using a plug of Pyrex glass powder (4). A marked paralle­lism can be found between the two sets of curves obtained by the a-c and doc methods, except

-I ;)

-s/s*-) 0

-05

()1----..4~-----,.a:-+----___i

-) -3 -,)

log c

Fig. 3. Relative zeta potentials as functions of the logarithms of the surf:)'c active agent concentrations ; indifferent ele­ctro�yte, IO-4 NKC I. •. Dodecylpyridiniulll chloride, doc mcthod; ') the same, a-c methud; .... cetylpyridinium chl­oride. d-c mcthod; ,the same.a-c method.

82 Memoirs of the Faculty of Agriculture of Kinki University No.14 (1981)

for the higher absolute values of ~ j ~* at higher concentrations in the case of the a-c data: this wood probably be due to the difference in the qualities of glass used. The curves show the typical behavior of the effect of cationic surface active agents on zeta potentials, causing a change in sign at certain surface active agent concentrations (4,5).

According to the author's theoretical treatment of the zeta potential, using the Stern and the Gouy-Chapman theories of the electrical double layer, the following equation holds (4)

[7]

where

[8]

and

k 2 = exp (L:.GjkT)jSS.6 [9]

Here c and cj are the concentrations, and z and Zj are the valencies (including signs), of the surface active agent and the indifferent electrolyte, respectively. L:.G and N I are the electro­chemical free energy and the nwnber of available sites of adsorption, and k, T, and N are the Boltzmann constant, the absolute temperature, and the Avogadro number. Here, the first order Deby-Hiickel approximation was taken.

It is easy to derive simple equations to be used for estimations of L:.G and N1 in the case of C«c j , viz.

(d~jd log c)o =-2.303~* (1+(~*jkl)] (10]

and

[I I J

Here the subscript zero denotes the isoelectric point, i.e .• ~ = O. The values of L:.G and N\ obtained from the experimental curves in Fig. 3 by using Eq. [10] and (11] are summarized in Table I. It will be seen that N I -values are almost the same for the two cationic surface active agents of different chain lengths. The L:.G-values, on the other hand, show a difference

Table I. Adsorption characteristics of cationic surface active agents on a glass surface at the room temperature

Surface active agent N\. cm- 2 LlC, cal/mole

10 12d·c method 1.06 X 7810

DPe a·c method 1.60 X 10 12 6880

doc method 0.69 X 10 12 9570 epc

a-c method 0.68 X 10 12 9720

DPC. dodecylpyridinium chloride; CPC, cetylpyridinium chloride; indifferent electrolyte, 10-4 N KCI.

83 Shizuo UEDA and Motoko UEDA A-C Methods in Interfacial Electrical Phenomena

To amplifier

.; To ground

(A)

a - .... b-'+-... ba -1'-:~~_£-... ~

Hg IN KClaq

A (8)

Supersonic F-"'-!!!!!Li~ a ........I-b a-"""{ ) :( ) r }-bwave 'liquid ..,..." Hg IN KCI aq

-'"

Fig. 5. Capillary elements of U~ffect n.Fig. 4. Unit cell of the hydrophone of

U~ffect I.

of approximately 2~3 kcal/mole for an increase in chain length from 12 to 16. These values of tiC and the increase with chain length are sufficiently reasonable if we consider the rela­tively poor accuracy of interfacial electrical measurements in general (4,5). The values of

2N1 (2!'IO I2 cm- ) are considerably smaller than those obtained on negatively charged silver iodide sols (5), indicating that only a very small portion of the glass surface provides adsorp­tion sites for these cationic surface active agents.

It has been difficult to do electrokinetic measurements by doc methods, especially in the case of solid-organic liquid interfaces, due to the high resistance of the plug, and hence of the circuit. However, this kind of difficulty can be reduced to a considerable amount by using the a-c method as given above. Figure 4 shows another type of such measurements. This is the unit cell of a hydrophone of a glass filter vibrating element A filled with various liqUids, i.e., distilled water, methanol, ethanol, acetone, pyridine, acetic acid anhydride, carbon tetrachloride, or benzene. The element has a pair of electrodes, a and b, immersed on both sides of the capillary system C, and attached vertically to the diaphragm D of the unit cell.

A supersonic wave of a constant power and frequency, here 13,500 cps, forces the dia­phragm and hence the element to vibrate, giving rise to the generation of an a<· voltage by U-effect I at the two electrodes. This voltage is amplified and read by a valve voltmeter.

It was found that no a< voltage generation was observed when benzene or carbon tetrachloride was used as the liquid phase. Moreover the relative value of zeta potential, E"A./e (see Eq. [6)) was an increasing function of the dipole moment of the molecule of the liquid phase (6). This suggests that the zeta potential is intimately related to the dipole orientation of the organic molecules at the interface. This orientation takes place in order to minimize the free energy of the system, which gives rise to the "dipole double layer ," including the diffuse part due to thermal agitation (6).

Memoirs of the Faculty of Agriculture of Kinki University NO.14 (1981 )

A-C Capacity Current

U-effects are not resticted to the case of solid-liquid interfaces as shown above. If a glass capillary containing a mercury-salt solution interface, as shown in Fig. SA, is forced to vibrate mechanically: an a-c voltage is generated bel ween the two electrodes. a and b. The voltage has the same wave character as the mechanical vibration. This we called "U-effect 0" (7). The principle is mOre or less the same as in the case of the condenser microphone, since the electrical double layer at the mercury-solution interface behaves as a perfect condenser (R). However, the total double layer capacity changes periodically by virtue of lhe interfacial area change in the present case. while it is due to the change in the condenser gap distance in the case of the condenser microphone.

The inner resistance of the element can be reduced by using a solution of high sail content. e.g., I N KCI or H2 S04 , In addition it was found that the output voltage increased linearly with the number of interfaces, and hence a very efficient transd ucer could be made by using an element as is shown in Fig. 5B.

Theuretical. - The mercury-solution interface is equivalent to a perfect condenser, having a potential difference E, I'S. lhe electrocapillary maximum. Its instantaneous integral capacity c is given by

c=Ca=CA(I+p) [I ~ I

where C is the integral double layer capacity per unit area. a and A are the instantaneous and the average interfacial areas, and p is the time dependent term of Vibration. The instantane­ous current of the circuit i is given hy

i=(d/dt) [c(E+e -iRo)J

= (d/dt) [CAE(J +p){1 +(e -iRo)/E}J

= CA [E(dp/dt) + dee - iRo)/dt] r13J

where Ro is the solution resistance and e the instantaneous potential drop at the load Z (see Fig.6). Here the terms of higher orders in p have been neglected. In the .:ase of the

~f-."~f: c Ro

-----e--­Z

Fig. 6. Equivalent circuit of U-cffect fl.

stationary state, complex representations of the instantaneous quantities can be employed, as

i = ..J21 exp (jwt)

e =y'"2 E exp (jwl)

85 Shizuo UEDA and Motoko UEDA A-C Methods in Interfacial Electrical Phenomena

and

p =-./2 P exp (jwt) [14]

where the phase differences are included in the corresponding complex amplitude terms, I, E, and P. Substituting Eq. [14] into Eq. [13], followed by rearrangements, we obtain

1= jwCA(V + E -IRo) [I S]

where

V=EP [16]

V is the "electromotive force of U-effect D." On the other hand, Ohm's law states that

E = -I Z [17]

Equating [1 S] with [17], we obtain the final equation, as

I = VI [(R+Ro) +i{ X - 1I (wCA)}] [18]

where

Z=R+jX [19]

R and X being the resistive and reactive parts of the load. The a-c current I is, therefore, proportional to the electromotive force V, if a constant load is used. Since both V and C are functions of the polarization E, 1 is a complicated function of E.

Experimental. - Figure 7 is the block diagram of the circuit for the measurement of the I-E relation. The mercury-solution interface in a glass-capillary vibrating element A is pola­rized by a potentiometer B. The a-c current i generated by the mechanical vibration of this interface is measured by the a-c potential drop across a small load resistance R, amplified by the amplifier C, and read by the valve voltmeter D. The vibrating mechanism of the element is the same as in the case of Fig. 2, i.e., E and Fare, respectively, the vibrator and the oscilla-

Fig. 8. A-c currents of U-effect IT as functions of polarizations of the mercury surface. I is in an arbitrary scale.A, IN Klaq, 300 cps; A'IN

Fig. 7. Block diagram of the measurement of the KCI aq, 500 cps; B, IN KI aq, 300 cps; B'IN

currentpo!arization relationship of U-effect n. KI aq, 500 cps.

86 Memoirs of the Faculty of Agriculture of Kinki University NO.14 (1981)

tor. A constant amplitude of vibration was always used for each frequency. Figure 8 shows the experimental relative I-values as functions of E. The curves A and A'

are for IN KCI aqueous and Band B' are for IN KI aqueous. As expected from the theory, Eq. [16] and [18], the current I is zero at the electrocapillary maximum, where E = 0, and increases with increasing lEI. The difference in the curves for KI and for KCI is explained by the higher integral capacity of the mercury-Kl solution interface, as compared with that for the mercury-KCI solution interface (8, 10).

Double Layer Capacity Measurements by U~ffect n

Impedance matching method. - If a purely resistive load is used, i.e., X = 0 and hence Z =R in Eq. [18] , the power W supplied to the load is given by

[20]

where the quantities shown by italic types are the moduli of the corresponding vectors. The condition of the maximum power supply, i.e., aWl aR = 0, is then given by

[21 ]

This is generally known as impedance matching since, when the modulus of the inner imped­ance and the load resistance are the same, the maximum power is supplied to the load.

This principle can be applied to double layer capacity measurements at mercury-solution

interfaces (II). If we find the matched load resistance at two frequencies, it is an easy matter to solve for C and Ro by using Eq. [21] .

Resonance method. - For an inductive load, i.e., Z = R + .iwL, the current is given by

1= V/((R + Ro) +i {wL- I/(wCA)}] [22]

The condition of the maximum current is widely known as the resonance phenomenon and

is given ~y

wL = I/(wCA) [23]

This principle is also applicable to double layer capacity measurements (I2). Experimental. - Figure 9 is the block diagram of the apparatus for double layer capacity

measurements using U-effect n. The capillary element A containing a mercury-solution interface is inserted through a small hole into a glass vessel B. This vessel contains the salt

Fig. 9. Block diagram of the impedance matching and the resonance methods for the double layer capacity measurements at mercury-solution interfaces.

87 Shizuo UEDA and Motoko UEDA A-C Methods in Interfacial Electrical Phenomena

solution to be investigated. e.g., IN KCl aqueous. The interface in the capillary is polarized by a potentiometer D, the large pool of mercury C being the reference electrode.

The a-<: current produced by the mechanical vibration of the element is supplied to the load, which consists of a variable resistance R I (40 k ohm var.) and a small resistance R 2

(J 00 ohm). The a-c potential drop e at the load (R 1 + R 2 ) is amplified (by E) and read by means of a valve voltmeter F,while the current i is read by another amplifier (G)-valve volt­meter (H) system from the potential drop at the resistance R 2 .By varying R.continuously, the condition of the maximum power supply is determined, i.e., the value of (R.+ R2 ) which gives the maximum power. The vibration of the element is produced by means of a vibrator I driven by an oscillator J.

If we use instead of R. a variable inductance L, and measure the current by the potential drop at R 2 , we can easily determine the resonance condition by varying L, and hence the C-value by using Eq. [23]. It is also possible to work out the solution resistance Ro from the analysis of the shape of the resonance curve (9, 13).

Practical Applications of U-effects

Several characteristics of V-effect D. - In addition to the contributions to interfacial electrical measurements in supplying various new methods as given above, U-effects I and n can also be applied to various mechano-electric transducers (7). It is then important to know several characteristics of the capillary vibrating element for these purposes. From the view point of the practical application, however, U-effect n is far superior to the other due to the lower internal impedance of the element, the higher a-<: voltage gain, etc. We shall, there­fore, examine the characteristics of the elements of U-effect n in this section, although it has been proved by experiments that U-effect I is also applicable to various kinds of trans­ducer devices (14).

Figure 10 shows the block dIagram of the measurement circuit. The a-c voltage obtained by the mechanical vibration of the capillary element A is measured by an amplifier (B)

.attenuator(C)-amplifier(D)-valve voltmeter(E) system. The current is measured by the poten-

Fig. 10. Block diagram of characteristic measurements of U-effect D.

88 Memoirs of the Faculty of Agriculture of Kinki University No.l4 (1981)

O.5mm ~~

To J

Fig. 11. Vibrating condenser for vibrational amplitude measurements.

tial drop across a small constant resistance R2 , which is read by another amplifier.{F)-attenu­ator (G)-amplifier(H)-valve voltmeter(I) system.

The amplitude of vibration was measured by a frequency modulation circuit J. The princi­ple is as follows. The frequency of a self-oscillator is a function of the capacity of its tank circuit. Hence, the mechanical vibration of a condenser can be converted into a frequency change, which is further converted into a voltage change by a detector. This voltage change

is supplied, after amplification, to the vertical axis of the cathode ray oscilloscope K, and read by the height of the trace on the screen. Land Mare, respectively, the oscillator and vibrator for the mechanical vibration. Figure II shows the vibrating condenser ex attached to the end of the vibrating rod. This condenser constitutes a part of the tank circuit of the self -oscillator set in the FM circuit. The symbols A, L, M, and J have the same meanings as in Fig.IO.

The inner impedance of the element was measured by the impedance matching method. The voltage at the load (R I +R2 ) in Fig. 10 is measured as before, and the current by the potential drop at R 2 • Since that decibel readings, Db I and Db 2 , of the attenuators C and G are the measures of logarithms of the voltage and current, the sum (DOl +Db 2 ) is the measure of the power supplied to the load (R I +R2 ). By changing the load resistance by varying R I ,

and maintaining the readings of the voltmeters E and 1at constant values, a condition can be found which gives the maximum (Db I +Db 2 ) value. The value of (R I +R 2 ) under this condi­tion is equal to the inner impedance of the element at the frequency used.

Table II shows the values of the inner impedance of capillary elements of large, medium, and small cross sections, i.e., (I) 0.76 (D) 0.49, and (ill) 0.37 mm in diameter, respectively. Each element contains forty mercury-1 N HCI aqueous interfaces. It is seen from this table that the internal impedance is of the order of several thousand to several tens of thousand of ohms at 1000 cps and is inversely proportional to the cross-sectional area of the element. It was also found that the internal impedance was proportional to the number of interfaces.

Tibia II. Inner impedance of the capillary elements of U-effect IT at 1000 cps

Elment No. Diameter, mm Impedance, k ohm

1 0.76 5 IT 0.49 20 ill 0.37 30

Number of interfaces, 40; solution phase, 1N Hel aqueous.

Shizuo Ell and M 10k ~.I)A A- Method in Interfacial 1 ctrical Phenomena

.-.t 1

III > > III/ EE I{IIc.i ::­01, ;{II:="" '" "- II '0 :";11 "0

>> :':11:; :;

0.0. 111

0'" I '" 111

II II I ll. I:.: I I II

Amplitude of vibration

10' Imml I umber of interfa~e

Fig. 12. Output voltages of ·cffc~1 n a fun ·tion of Fig. 13. Output vohage of -ef'fecI 11 a a function of

amplitude fibration at 100 cps. the number of interfaces at 00 cp .l:.lcment ! . J I.

Fic . 12 sh w the output v ltage a fun tions I' the amplitude I' vibration for the three elements. This is nol a linear relation. since the amplitude I' the periodic interfacial area change is of an order higher than that of the one-dimen ional amplitude mea ured y the FM ircuit.

Figure 13 shows the linear relation between the output voltage and the number of inter­f~ e contained in the capillar element. II. Here the element n wa u ed. and the amplitude and fr quency of vibration were kept onstant at 2.6 X 10-3 mm and 500 cps, respectively. The load wa alway matched to the internal impedan e I' the clemen!. since otherwise this impedance would bec me larger than th I ad r r higher 11- alue. and a saturation effect would be observed.

The curves is Fi . 14 ar I g rithmic plots of the voltage-frequency relation at a conslant amplit ude of vibrat ion, i.., 1.2 X 10-3 mm. 1n each case, Ihe load re i Ian e was matched t the internal impedan e of the lemenl at 1000 cps. It i clear fr ITl the e urve that the respoll e al lower frequell ie is higher for hlrger elements, while that at higher frequencies is hi her for smaller elements. Thi kind I' frequen 'y chara ler is alw3. s shown by the free t pe of vibralion employed here. which is mainly due 10 the natural frequencies of the el ments. It will be shown later that nat frequen y character curve are obtained by empl ying a pi tnt pe of vibrat in.

:1lI ,---------,

:!o > E ](I

ab Uc.i /Y­O/) E~, B:3 0

:) ..­:; r-+ _~A = _ Hg 0'" Paraffin =.-. - I~~~ __ . 2NH,SO....~~8:

(G las cap liar ies)C -. HI(

100 51Wl I.lKH) :.!.IKlO

Freq uenc '. cps. Fig. 15. ooting experiments f -effed IJ.

Fig. 14. I:requenc char3cter curve of -effect n for

the free type of vibration of the clements I. U. and ill. A mplitude of vibration: 1.2 x 10-3

mOl.

~ It, Memoirs of the Faculty of Agriculture of Kinki University No.14 (1981)

6 .-..-....... .. .. OJ

'"0

oc-~ 0­; 2 0

20 10 o -10 -20 -30 -4() Temperature,OC

Fig. 16. Output voltage of U~ffect n as function of temperature.

Effect of cooling (IS). - Figure 15 shows the experimental device. An element of U -effect n, A, made of glass filter No.2, was dipped in an alcoholic bath B, cooled by the dry ice C, and the receiving sensitivity measured as a function of the bath temperature. A con­stant mechanical pulse is delivered to the element by knocking the upper end of the glass pole D, and the pulse voltage generated at the two electrodes, a and b, was measured by the height of the trace at the screen of a cathode ray oscilloscope. E is the thermometer.

Figure 16 shows the results obtained. The output voltage is practically constant from 20 to ca.-40°C.The melting points of the mercury and the 2N H2 S04 aqueous, which was used as the solution phase of the element, were -38.89°C and -19°C, respectively. The results indicate, therefore,that even if the solution is in the solid state ,the output voltage of U-effect Ddoes not diminish.

Hydrophones. - An example of the practical application of U-effect 1 to a hydrophone device has already been given in Fig. 4. A hydrophone using U-effect D is shown in Fig. 17.

To amplifier

~~~ ,. V"-­

Supersonic wave B

Fig. 17. Unit ceU of the hydrophone of U-effect n.

The unit cell has twenty capillary elements, AI ,A 2 , •.. and A2o , which are fixed to a resonat­ing plate B and connected in series. Each element contains twenty mercury-solution phases. The output voltage is amplified and read by a valve voltmeter.

The comparison of the receiving sensitivities of the unit cells of U-effects I and II and also of the Rochelle salt was made by 10-meter water tank tests. The unit cell and the supersonic source, using the magnetostriction of a nickel pole with the anode loss of about 1w and the frequency of 13,500 cps, confronted each other in the tank, and the readings of the output voltage were observed. Outdoor tests were also made at the Yodo river, in which the maxi­mum receivable distance was tested by using the same supersonic source, with the anode loss of about lOw.

Shizuo VEDA and Motoko VEDA A-C Methods in Interfacial Electrical Phenomena

Table III shows the average results thus obtained. It is clear that the unit cell of U-effect II has a receiving sensitivity of twice to 1.5 times as high as that of the unit cell of the Rochelle salt .

Pickups of the electrophonograph. - In applying the U-effects to pickups of the electro­phonograph. a uniform frequency response is demanded, avoiding the resonance of the vibrating system. A pickup of the free vibration type of the capillary element of U-effect II is shown in Fig. 18A. The curve I of Fig. 19 is its frequency character curve. indicating a marked peak at about 400 cps. It was also proved experimentally that this peak shifted toward the high·frequency region for a fine capillary (curve IT) and toward the low frequency region for a large one (curve ill). The internal diameters of these capillaries are 0.135.0.148. and 0.182 mm. respectively.

The influence of the electrolyte of the solution phase of the element is also examined. A solution of high electrolytic conductivity and low viscosity is desirable. For instance the response at lower frequencies is lower in the case of sulfuric acid for low temperatures than in the case of hydrochloric acid. This is due to the larger viscosity increase at low tempera-

Table III. Receiving sensitivities of the hydrophones of U-effects I and 0 and the Rochelle salt

IQ·meter water tank Maximum receivable

Element test. (Readings of distance, km

the valve-voltmeter)

U-effect I 42 1.9 U-effect 0 143 3.8 Rochelle salt 68 2.5

.0 30Cl ill ,_~ , ....,.oj

I '-.,',bO .-..~- .... II ~ / --. ~

",'.'. 20 rv-· ~:--_> / ,.

¥"0

I

I

..."5 I ­"".~ __ .... I

fr I /. (B) ./;:I~ ('.A) 0 10 - ..

~~ o ~~

~ 1.500 3.000 6. Frequency. cps.

Fig. 18. Pickups of the electrophonograph using Fig. 19. Frequency character curves of the pickups U~ffect 0: A, free type; B, piston type. of U-effect D. The output voltage is in an

arbitrary scale.

tures for the former solution. A pickup of very high quality is made by employing the piston type of vibration to one

of the electrodes as shown in Fig. 18B. The frequency character curve of this type of pickup is given by the curve IV in Fig. 19, showing a satisfactory linearity over the frequency range examined.

Memoir f th Fa ullyof gri ullure f Kinki niversity No.14 (19 I)

Summary

The mecl1anic:d vibration of the ele trical double layer in a olass capillary give rise to the generation of an a·c voltage between the two ends. This phenomenon is called ingeneral "U-effect". In the ase of glass·liquid interfaces. we call it .. -effect J", while in the case of mercury· alt solution interfa es" -effect 0". The former is an effect of an a·' streaming potential generation, and the latter is an a-c capacity urrent of [he double byer at mercury -solution interfaces.

U-effect I an be used to measure streaming potentials by an a-c technique. providing an alternative methode to mpare with the rdinary d-<: method. U-effect II has its applica­tion in providing new methods for the double layer capa it measurement at mercury-solu· tion interfaces.

U-effects, both I and II. are applicable to all sorts of mechano-e1ectric transducers and it was shown that the hydrophone and pick up using -effe t IJ were equal to or even superior to those using Rochelle salt.

( 1947). References 9.� A. Watanabe, F. Tsuji. and S. Ueda. J. Elec­

I.� H. A. Abram n, "Electrokinetic Pheno· trochem. Soc.,Japan, 22.179 (19 4).

mena." hem. atalog, ew York (1934); 10. A. Watanabe. F. Tsuji. and S. Ueda,J. Elec­

J. A. V. Butler. "Electrocapillarity,"� trochem. S c., Japan, 22, 521 (1954);

Methuen. London (1940). Proc. 2nd Intern. Congr. Surface ctivity, London. 3,94 (1957); Bull. in I Chem.2. S. Ueda. F. Tsuji and A. Watanabe: Bull. Res., Kyoto Univ .. 34, i. 65 (I956).Inst. Chern. Res., Kyoto Univ., 19, 40

II. S. eda. A. Watanabe. and F. Tsuji, 1. Elec­(1948); Proc. :!nd Intern. ongr. Surface trochem. Soc.. Japan, 21, 267 (1953);Activity, London, 3, 3 (1957 ; Mem. Call. Bull. Inst. ('hem. Res., Kyoto Univ .. 25,Agr., Kinki Univ .. 9, 84(1976 . 30 (19 -I); Mem. \011. Agr., Kyoto Univ., 3. S. Ueda,� A. Watanabe, and F. Tsuji, Bull. 60,13 (1951); 67,53 (1954).Inst. Chern. Res., Kyoto Univ .. 20, :!8

(1950); Mem. II. Agr.. Kyoto Univ .• 12. A. Watanabe, F. Tsuji. K. ishizawa, and

57,22 (J 950); J. Electrochem. Soc.. Japan,� S. Ueda, Bull. Inst. Chem. Res., Kyoto Univ., 32. -4 (1954); Mem. Coil. Agr ..19.142(195~).

4.� A. Watanabe, F. Tsuji, and S. Ueda, 1. Ekc­ Kyoto Univ., 67,79 (1954).

trochem. Soc., Japan. 29,777 (J96I). 13. A. Watanabe, F. Tsuji. K. ishizawa. and

5.� R. H. Ottewill and M. . Rastogi. Trans. S. Ueda, Bull. Inst. Chem. Res .. Kyoto

Faraday Soc., 56,866 (1960). niv. 32,6_ (I9S4).

6. S. Ueda, F. Tsuji. and A. Watanabe, 1. Elec­ 14. S. eda, A. Watanabe. F. Tsuji, and K.

trochem. Soc., Japan 24,74 (1956). Nishizawa, J. Electrochem. Soc., Japan, 20,

7. S.� Ueda F. Tsuji and A. Watanahe, Bull. 605 (1952); Bull. Ins!. Chem. Res.. K oto

Inst. Chern. Soc., Kyoto Univ., 20, 28 Univ .. 28. 47 (195 ); 31, 103 (1953);

(19-0); Mem. 011. Agr .. Kyoto Univ.. 57, Mem. Call. Agr.. Kyoto Univ .. 67, 61

22 (1950); Chem. & Chern. Ind. 4,316 (1954).

(1951); Mem. Coil. Agr., Kinki Univ., 12, IS. S. Ueda. F. Tsuji. and A. Watanabe. J. Elec­

71 (I979). troche1l1. Soc.. Japan, 24. 116 (1956).

8.� D. C. Grahame, Chem. Revs., 41. 441

dMo oUE

I(一効果

界面電',-現象における交流法 ことができ,測定法の簡易さ測定値の正確度に於て

静男

u

ktooUEAanDizuhS AD A-C Me nI err El i ltecrcailacatnhd iost menahenoP 93

Iは交流法によって弧軌屯位 を測定する

･臥.1 は.従来の在洗法におけるよりも勝れている.

効果 H を利用 して水銀一溶液界面の屯(7ミニ兎~

,'J( l_ 唇谷最測定の新 Lい方法が考え出された.に ニt(桐を線繊的に振動すると

u硝 /･七線 中のL'･17'

Iih端に交旋'tr.J ';Fが発セする.=の鋭敏を-船に" 一u 一効果 lJE.に Ilu は色々の純期の横根一花気勢力

効火｡と買う.この内かラスー液体界面の梯今はや

列火 1

-u 変換恭敬t二用いらiLる.例えば.我々は水中聴曹機 ,

"と呼 ul.水銀一塩郡満液界Lrの時は"uil 一助 劾氷

'L正位を莞'kL, 詐すべきことは. これらu効果利用の機器は従来使

-ビックア･･/ブ等に u 11を採用 した.そして特

_ :. f;m薪の坊介は交統統動Il.･n と叶

後者の IサーTに淡を発生する. 良い特作 をもっている.

･AHま水銀一溶液HHの二相屑の交洗コンデン 用 さJLl=ロ･/シュル塩利用織器 と同等かそtL以上のf