curve. The quantity of D, CDV, can be obtained from the observed current and the relation i = nZYP,Cn. Finally the quantity of B remaining is obtained from

v(cB + CD f Cl2) = MBo (18)

where MBO is the quantity of B present before decomposition as calculated from the amount of A used in the electrolysis. Equation 18 follows from the stoichiometry of the system, Equations 5 and 6.

The results of calculations of the quantities of B, D, and E as a function of time indicate that CB approaches zero near the current maximum, leaving only C , E, and unreduced D in solution. Thus the assumption that P 2 may be obtained from a plot of In i us. t for the descending portion is valid.

The linearity of the ascending and descending branches of Figure 7 permits calculation of both 8 2 and k and supports the validity of the pinacol decomposition scheme and the derived equations for electrolytic autocatalysis. To test the theory further, additional electrolyses were performed vary- ing the quantity of DBMH, the stirring rate, and the tem- perature.

Table I1 shows the results of 10 experiments with values of the rate constant for the pinacol decomposition determined by the method just discussed. The values of k are relatively independent of stirring rate though they show a small de- pendence on concentration. As expected, k is strongly de- pendent on temperature.

The concentration dependence of k probably reflects an inadequacy in the theory, possibly in the simplifying assump- tion that the decomposition is zero order in pinacol concen- tration. The overall decomposition reaction must involve several steps and a rate equation as simple as that used in the derivation may not be adequate. However, computations of the various concentrations during the current-time maxi-

~~~

Table 11. Kinetics of DBMH Pinacol Decomposition, 25 “C

DBMH, mg k , min-1 5w 0.60 =IC 0.07 80 0.42 0.03

100 0.39 i: 0.04 506 1.38

a Including three different stirring rates. * 40 “C.

~-

mum show that the rate equation used is superior to any other rate equation of comparable simplicity, in particular the case where the rate is first order both in pinacol and catalyst.

Another factor which has not been taken into account in the theory is the fact that the blue solution during the pinacol decomposition slowly becomes green indicating that the benzil radical anion concentration is slowly decreasing. This de- composition of the radical anion has been verified during controlled potential reduction of solutions of benzil under otherwise identical solution conditions. This complication, of course, would affect the derived kinetic equations which assume that benzil radical anion does not decompose.

Nevertheless, the results in Figure 7 and Table I1 lend con- siderable credibility to electrolytic autocatalysis in the decom- position of the pinacol as the source of the unusual current- time maximum in the controlled potential reduction of DBMH and other P-diketones.

RECEIVED for review July 22, 1968. Accepted September 3, 1968. Research supported by the Wisconsin Alumni Re- search Foundation and the National Science Foundation Grant GP-8350. Presented in part, 155th National Meeting, ACS, San Francisco, April 1968.

ffect of Static Magnetic Field on irect Current Polarography Shizuo Fujiwara, Yoshio Umezawa, and Teruhiko Kugo Department of Cliemistry, Faculty of Science, The Unicersity of Tokyo, Bunkyo-ku, Tokyo, Japan

The effect of magnetic field on dc polarography has been investigated with a static magnetic field applied perpendicularly to the electrodes. Maxima of the second kind are decreased by the application of mag- netic field. The factors influencing this effect are examined experimentally. As the mechanism of this phenomenon, the retardation of tangential motion on the electrode surface produced by the application of magnetic field is proposed. The effects of magnetic field on the first kind of maxima are also studied, and are divided into three groups by the magnetic field effect as its criterion. The origin of motion of electro- lytic solutions around the dropping mercury electrode accompanied by the first kind of maxima is not unique. On the other hand, a small but distinct increase in the diffusion or migration current is produced by the application of magnetic field.

EFFECTS of external electric or magnetic fields on electrolytic reactions at the electrodes have been reported (1-5). Kryu- kova and Popovam ( 1 , 2 ) reported the observation of polaro- graphic maxima of the first kind produced by a static electric field applied perpendicularly to the dropping mercury elec- trode (DME). Imai, Inouye, and Chaki (3) studied the effect of an ac electric field on electrolytic reactions at the DME. Furthermore, Pekhteleva and Smirnov ( 4 ) reported the

influence of transverse and longitudinal static magnetic fields on electrolytic reactions at the plane electrode, and showed that the electrolytic solution is hydrodynamically stirred by the magnetic field. Antweiler (5) tried, but failed, to observe the effect of magnetic field on maximum wave currents of the first kind in polarography. In a preceding paper (6), it was reported that the magnetic field influences dc polarography and the maximum wave currents of the first and second kinds decrease with the application of magnetic field. Experimental details and the interpretation of this effect will be given in this report. Furthermore, it will be shown that the diffusion or migration current is increased slightly by the application of magnetic field.

(1) T. I. Popova and T. A. Kryukova, Zh. Fiz. Khim., 5,283 (1951). (2) V. Levich, “Physicochemical Hydrodynamics,” Prentice-Hall,

(3) H. Imai, S. Inouye, and S . Chaki, Bull. Chem. SOC. Japan 32,

(4) N. I. Pekhteleva and A. G. Smirnov, Magnitn. Gidrodinam.,

( 5 ) H. Antweiler, 2. Elektrochem., 44, 836 (1938). (6) S. Fujiwara, H. Haraguchi, and Y. Umezawa, ANAL. CHEM., 40,

Englewood Cliffs, N. J., 1962.

994(1959); 33,296(1960).

Akad. Nauk. Laic., SSR, 1965(2), 89; C.A., 64,279h (1966).

249 (1968).

21 86 ANALYTICAL CHEMISTRY

\ -

, . . . . . . . . . . . . . . . . . . . . . . . . I . . , . ( '

, . .! + J ..... +;..+.. L------ 1 . \ ;,. & J ..,.,

. . . . . . . . . 8 . i i.:. 2;

i ..... m j _j I + ! : : L ; , ' + -1 / -w - t5v

p- XmGpr. . r K a c a n . u y 1 0 3 O n . , 1 2 u D ~ , l Y M G = s . l8mlGau . . . . . . . . . . . . . . . I .

,l.ll.-ll-li. , . ., .."..l,i, . I j , . ~ . . . . . . " _ , L , . r

Figure 1. Dependence of polarographic current (the maximum of the second kind) on magnetic field strength

0.5mM Mn(I1) and 1M KSCN

EXPERIMENTAL

Apparatus. Direct current polarograms were taken with the use of a Yanagimoto Polarograph Model PA-102. The static magnetic field varied from 0 to 18,000 Oe and was applied perpendicularly to the electrodes. The electromagnet used in this experiment was manufactured by the Japan Electron Optics Lab. Co., and has 30-cm diameter pole- pieces with a 60-mm gap. The homogeneity of the magnetic field is about 10-5 over the whole volume of the glass cell. All measurements were performed at 23.0 =k 0.5 'C.

Capillary. Capillary characteristics were as follows. The drop time of mercury ( t ) was 3.52 sec for a system of 0.2mM Mn(IP), 1M KC1 solution at a potential of -1.58 V, which was not altered by the applied magnetic field. The flow rates (m) were 0.736 and 0.740 mgisec with and without magnetic field, respectively, for the above solution with an open circuit. The drop time and flow rate mentioned above were obtained with a 70-cm mercury height. Two other capil- laries with similar characteristics were used occasionally.

Sample solutions used in this experiment were prepared from chemicals of analytical reagent grade. The water used for all solutions was deionized and distilled.

Chemicals.

RESULTS AND DISCUSSION

Decrease in Maximum Wave Current. The effect of mag- netic field is observed as a decrease in the maximum wave current. A typical example of the magnetic field effect on the maximum wave of the second kind is shown in Figure 1. For the second kind of maxima, this decrease in current is always observable. On the other hand, the appearance of the magnetic field effect for the first kind of maxima depends on conditions as described later. Two possible mechanisms for this effect are (i) the direct interaction of the applied magnetic field with induced or permanent magnetic dipoles arising from mercury drops and electrolytes, and (ii) the Lorentz force due to the interaction between magnetic field and the charge of electrons in mercury and/or that of electro- lytes in motion. Mechanism (i) is eliminated, because the effect of magnetic field was similar in both cases for diamagnetic and paramagnetic samples, and the drop time of mercury does not change with magnetic field. Mechanism (ii) wi!l be observed in two different ways-Le., one as a Hall effect, and the other a magnetohydrodynamic force. Of these two, the Hall effect can be eliminated, for ordinary diffusion current has to be reduced by the magnetic field for the Hall effect. I t has been shown that the polarographic maxima are always associated with tangential motion at the surface of the DME (2, 7). Because the effect of magnetic

(7) J. Heyrovsk5 and J. K h a , "Principles of Polarography," Academic Press, New York and London, 1966.

Figure 2. Tangential motion at interface of mercury and solu- tion, and vector diagram of magnetohydrodynamics

V. Velocity of conducting fluid H. Applied magnetic field Find. Induced electromotive force

field is observed as a reduction in maximum wave current, the most plausible mechanism for this is the suppression of the tangential motion due to the magnetohydrodynamic force.

Maxima of the Second Kind. Maxima of the second kind are of purely hydrodynamic origin and have been inter- preted as due to the tangential motion of the surface of the mercury drop under the action of the high flow rate of mercury (2, 7). The polarographic limiting current for this maximum wave can be expressed as the sum of the Ilkovi; diffusion and maximum wave currents, the latter being given by Equation 1 (2, a>,

i,,, = KnFD1'2Cu3/2V1'2 (1 )

where K is a proportionality constant, yt the number of elec- trons involved in electrochemical reaction at the surface, F the Faraday constant, D the diffusion coefficient, C the con- centration of electroactive species, r the mean radius of a mercury drop, and V the velocity of the tangential motion at the mercury/solution interface as drawn in Figure 2.

It is assumed that the present system of mercury and electrolytic solutions at the interface can be treated as a magneto-fluid in a magnetic field. According to the theory of magnetohydrodynamics (9), a force Find acting on a con- ducting fluid in motion with a velocity V induced by a magnetic field H is given by

Find = ~ ( & l X H) X H (2) where K is the electroconductivity of the conducting fluid.

As shown in Figure 2, the direction of Find is opposite to that of V,-i.e,, V is reduced by the magnetic field, resulting in a decrease in i,,, of Equation 1. Since K for mercury is very much greater than that for electrolytic solu- tions, this reduction is important only in the surface layer and inside the mercury drop.

For the validity of the assumption mentioned above, several experimental results were examined, which will be summarized in the following.

DEPENDENCE ON MAGNETIC FIELD STRENGTH. As shown in Figure 1, i,,, decreases gradually as field strength increases, as suggested by Equation 2. By a rough calculation using Equations 1 and 2, it can be shown that the current observed is inversely proportional to the magnetic field strength. The details of this calculation will be published elsewhere.

(8) Y. Okinaka and 1. M. Kolthoff, J . Amer. Clzem. Soc., 79, 3326

(9) See, for example, A. B. Carnbel, "Plasma Physics and Magneto- ( 1957).

fluidrnechanics," McGraw-Hill, New York, 1963.

VOL. 40, NO. 14, DECEMBER 1968 0 2187

1 1 M KCI

0

3 i

2.1

1 4 l e

0:5 LbM [KCII

- Figure 3. Dependence of mag- netic field effect on concentra- tion of supporting electrolyte

0.5 mA4 Mn(I1)

DEPENDENCE ON CONCENTRATION OF SUPPORTING ELECTRO- LYTE. The magnetic field effect AH is defined as follows,

(3)

where iE and io,$ are the current intensities of the maximum waves with and without magnetic field, respectively. As the concentration of a supporting electrolyte is decreased, A H defined in Equation 3 decreases. This is shown in Figure 3 taking Mn(1I) as an example. As discussed above, the reduction in A H implies that tangential motion becomes slow, Kryukova (2) reported that, at low concentrations of sup- porting electrolyte, tangential motion can only be observed at the potential of electrocapillary zero as shown in Figure 4. Her result is consistent with the present one in that tan- gential motion is negligible for low concentrations of sup- porting electrolyte at the decomposition potential of Mn(I1) (about - 1.6 V) as shown in Figure 3.

DEPENDENCE ON MERCURY RESERVOIR HEIGHT. As shown in Figure 5, A H decreases with decrease in mercury reservoir height. There is a linear relation between the flow rate of mercury and the speed of motion at the rnercury/solution interface. Hence, the speed of motion becomes negligible when the height of the mercury reservoir is low (2 ,7) , resulting a decrease in AH as suggested from Equation 2.

When a transverse ac elec- tric field (about 30 kHz) with a voltage of 0.5 - 1 V/cm is applied to the DME, a rise in limiting current like the maxi- mum of the second kind appears and is reduced by the applica- tion of magnetic field. This experiment shows that an ac electric field may cause tangential motion.

AE = ( L t - iH)/iout

EFFECT OF AC ELECTRIC FIELD.

l e

5.01 I

a e

tdn(II) Q5mM KCI 1 M

e

60 70 cm Figure 5. Dependence of magnetic field effect on mercury reservoir height

0.5mM Mn(I1) and 1M XCI

0.1 M KCI I

App(ied Potential. Volt vs. N.C.E.

Figure 4. Relative velocity of motion of potassium chloride solution around a mercury drop

Arrows indicate experimental values, and curves are constructed from calculated values, both obtained by Kryukova et al. (2). L ! ~ is velocity at potential of electrocapillary zero, L! is velocity at any other potentials

EFFECT OF CAPILLARY ACTIVE SUBSTANCES. When surface active substances are added to solutions, the effect of magnetic field becomes small as shown in Figure 6. This is attributed to the retardation of the tangential motion by the presence of surface active substances. The magnitude of the effect of magnetic field is dependent on the ionic species. This result is interpreted in terms of the control of movement of the interface by capillary active ions.

Maxima of the First Kind. Sharp-shaped maxima ob- served in the rising portion of polarograms are generally designated as maxima of the first kind. Maxima of the first kind have been investigated by many workers (7). It is now well established that maxima of the first kind accompany the motion of electrolytic solutions around the DME. How- ever, the origin of this motion still remains to be clarified.

10. %I iouf-i 4 + Q

l o

2 5' 0

0.0025 Mi5% [~e~atin]

Figure 6. Dependence of mag- netic field effect on concentra- tion of surface-active substance

0.5mM Luteo salt [Co(II) ---f

Co(O)] and 1MKC1

2 1 88 e ANALYTICAL CHEMISTRY

Figure 7. Relation between magnetic field effect on first kind of maxima and mercury reservoir height

1 mM CuC12 and 1M KC1

Frumkin and von Stackelberg (2 , 7, IO) proposed the “surface tension theory,” where they suggested that the uneven current density over the surface of a mercury drop produces a surface tension gradient and results in the motion of the mercury/ solution interface. Barker and Faircloth (11) proposed that specific adsorption is the origin of the surface tension gradient. According to Heyrovskg (7, 12, 13), however, it is outside the DME that the motion arises and not at the interface. From this observation, he concluded that the motion is due to the nonhomogeneity of the electric field in the neighborhood of the DME.

In the present work, the effects of magnetic field on the first kind of maxima are examined. The results may be divided into three groups-Le., the magnetic field effect is (i) observable only when the height of mercury reservoir is high, (ii) not observable at all, and (iii) observable, but almost independent of the height of mercury reservoir. These results suggest that the origin of the first kind of maxima is not unique.

GROUP (i). A typical example of group (i) is shown in Figure 7. In order that the magnetic field effect can be observed in group (i), the additional condition is required that the concentration of supporting electrolytes is high. The conditions for the magnetic field effect observed in group (i) is nothing but the conditions where maxima of the second kind are obtained. It has been shown that, under these conditions, maxima of the first kind often accompany those of the second kind (7). Therefore, the peak part observed with a 70-cm mercury height shown in Figure 7 should be taken as a superposition of the real maximum of the first kind, and a maximum of the second kind, the latter being reduced by magnetic field. With a 60-cm mercury height, only the real maximum of the first kind can be observed as shown in Figure 7. Because this maximum is little affected by magnetic field, the origin of this maximum is similar to that of group (ii), as described successively.

GROUP (ii). The magnetic field effect is always unobserv- able in this group. Examples for this group are systems of

(10) M. von Stackelberg, Fortschr. Chem. Forsch., 2, 229 (1951), and M. von Stackelberg and R. Doppelfeld, “Advances in Polarography,” Pergamon Press, London, 1960, p 68.

(11) G. C. Barker and R. L. Faircloth, “Advances in Polarog- raphy,” Pergamon Press, London, 1960, p 313.

(12) J. Heyrovsk?: 2. Phys. Chem., Sonderheft Juli, 1958, 7. (13) J. Heyrovsk?, Z . Phys. Chem., 229, 49 (1965).

Figure 8. Limiting current increase upon application of a magnetic field of 7000 Oe

0.5mM NiCls and no supporting electrolyte at -1.65 V constant + Application of magnetic field - Switch off magnetic field

0.5 - 2.5mM 02, Cu(II), Mn(II), Ni(II), Zn(II), Co(II), and Hg(I1) at low concentrations (0 - 0.005N) of KC1 or ”Os, all of which have remarkable maxima. However, according to Frumkin and von Stackelberg (2 , 7, IO), there is motion of the mercury/solution interface. If the inter- pretation proposed above to explain the magnetic field effect on maxima of the second kind is correct, the motion of the mercury/solution interface mentioned above should also be affected by magnetic field. This implies that Frumkin and von Stackelberg’s explanation about the origin of maxima of the first kind is incorrect. One possible explanation is that motion does not exist on the surface of the mercury drop, and motion of the electrolytic solution is little affected by the magnetic field as discussed above. The absence of motion on the surface of the mercury drop in this case is consistent with the result obtained by Heyrovskg (12).

Examples of this group are found in systems of 2mM Cu(II)-0.05-0.1N KSCN, 0.5-lmM T1 (I)-1M KC1 and 1 mM Cu(I1)-O.01N KC1. A few other systems belong to this group, The effect of magnetic field can be observed, even when the mercury reservoir is lowered. In addition to this, the ratio for the concentration of sup- porting electrolytes to that of depolarizers has to be properly chosen to observe the magnetic field effect. Kolthoff and Okinaka (14) reported that in dilute (<O.lM) KSCN solu- tion, copper(I1) yields cuprous thiocyanate which is adsorbed on the DME forming a monomolecular layer. At the re- duction potential of this adsorbent, a remarkable maximum of the first kind appears. Barker and Faircloth (11) showed that there also is appreciable specific adsorption of the ion of thallium(1) from chloride solution giving maxima of the first kind, In the present work, both of these maxima are affected by the magnetic field as mentioned above. From the above considerations, it can be concluded that motion of the interface arising from the surface tension gradient of adsorptive-origin (11, 15) is an important origin of maxima of the first kind of group (iii).

GROUP (iii).

(14) I. M. Kolthoff and Y . Okinaka, J. Amer. Chem. Soc., 82,

(15) S . Sathyanarayana, J . Elecrroanal. Chem., 10, 56 (1965). 3528 (1960).

VOL. 40, NO. 14, DECEMBER 1968 0 2189

An Increase in Diffusion or Migration Current. In all investigations described above, the magnetic field effect on polarography has been observed as a decrease in the maximum wave current, and has been explained using the theory of magnetohydrodynamics. However, it has recently been found that the limiting current of migration or diffusion is also affected by the magnetic field. Contrary to the previous case, the present effect is observed as an increase in the migra- tion or diffusion current as shown in Figure 8, and is small in magnitude compared with the previous one. Favorable con- ditions for observing this effect are a higher mercury reservoir height, the absence of surface active substances, and a lower concentration of supporting electrolyte. A possible mecha- nism can be obtained through the analogy of a magnetic mercury cathode (16)-namely, the magnetic field causes the electrolyte or the mercury drop to rotate or agitate like a rotor in a motor when the polarographic current flows. The third condition mentioned above may concern the removal of the tangential motion. However, explanations for the other two conditions remain to be worked out.

(16) E. J. Center, R. C. Overbeck, and D. L. Chase, ANAL. CHEM., 23, 1134 (1951).

CONCLUSION

By the use of a theory of magnetohydrodynamics, the mechanism of the effect of magnetic field on the second kind of maxima is explained in terms of the retardation of tan- gential motion of the mercury surface.

Using the magnetic field effect as a criterion, it is concluded that motion of electrolytic solutions accompanying maxima of the first kind is classified into two types of different origin. The first type of motion, originating at the surface of a mercury drop, is affected appreciably by a magnetic field. On the contrary, the second type of motion which arises in electrolytic solutions near the mercury drop, is little influenced by a magnetic field.

ACKNOWLEDGMENT

The authors thank Hiroki Haraguchi for his valuable dis- cussion and help in the experimental work and Yoji Arata for his helpful comments.

RECEIVED for review April 2, 1968. 1968.

Accepted August 12,

Fluorescence of Prostaglandin C. L. Gantt, L. R. Kizlaitis, D. R. Thomas, and J. G. Greslin Clinical Research Center, Department of Medicine, Unieersity of Illinois College o j Medicine, Chicago, 111. 60612

PROSTAGLANDINS, originally isolated and identified from sheep seminal fluid, comprise a large group of naturally- occurring compounds found in most body organs and fluids. These are of great potential biologic interest because most have profound effects at very small dosages. Several recent reviews on the nomenclature, chemistry, physiology, and pharmacology of prostaglandins have appeared (1-5). The structure of prostaglandin El (9-Keto-1 la,l5a-dihydroxy- prost-13-enoic acid) is shown in Figure 1. Because the structure of the prostaglandins suggested the possibility for several resonance forms of the molecule, this study was undertaken to determine their possible fluorescence.

EXPERIMENTAL

Prostaglandin El (PGEJ was weighed and dissolved in methanol at a concentration of 1 mg/ml and stored at 4 "C. Freshly-prepared dilutions of the stock solution were made in methanol at concentrations of 5 pg/ml and less. Aliquots of these were evaporated to dryness in 2-rnl test tubes, under nitrogen, with a sand bath at 40 "C. One-half milliliter of

(1) E. W. Horton and C. J. Thompson, Brit. J . Pharmacol Chemo-

(2) S. Bergstrom and B. Samuelsson, J. Bid. Chem., 237, PC3005

( 3 ) M. Hamberg and B. Samuelsson, ibid., 241, 257 (1966). (4) K. Green and B. Samuelsson, J. Lipid Res., 5, 117 (1964). (5) S. Bergstrom, Science, 157, 382 (1967).

-

fher., 22, 183 (1964).

(1962).

20

OH

Figure 1. Structure of PGEl

sulfuric acid-water reagent (70%, 30% v/v) was pipetted into these tubes which were then mechanically swirled. The samples and appropriate blanks of sulfuric acid were closed with ground glass stoppers and incubated for one-half hour at 65 "C. They were immediately immersed in 15 "C water for one-half hour. Fluorescence was determined on an Aminco-Bowman spectrophotofluorometer with quartz micro-cuvettes, previously acid cleaned and stored in con- centrated nitric acid. The samples were activated at 366 mp with a Xenon lamp source and read at 420 mp. The stability of the instrument is checked with a standard solution of hydrocortisone.

All glassware was cleaned in concentrated sulfuric acid- dichromate solution and rinsed thoroughly with distilled water. Meticulous care and cleanliness were necessary to obtain reproducible and accurate results when measuring small quantities of the prostaglandins. In these studies, the turret of the spectrophotofluorometer was set at 3 and the sensitivity at 30.

2 198 e ANALYTICAL CHEMISTRY