NONUNIFORM ELECTRIC FIELDS - by Herbert A Pohl
When a piece of iron jumps toward a magnet, it is responding to a nonuniform field. In approaching the pole faces it travels in a direction of increasing magnetic strength. Placed in a uniform field, no matter how strong, it would not move at all. Much less familiar is an analogous effect of nonuniform electric fields. They too can set matter in motion.
The behaviour of a nonuniform field can best be understood by considering first the simpler case of a uniform field, such as the one between a pair of flat, parallel metal plates that are oppositely charged. A charged body freely suspended between the plates - for example, in a nonconducting liquid - will move parallel to the field, toward the plate bearing the opposite charge. A neutral body, on the other hand, is not impelled in either direction; it stays put.
Even though it appears to ignore the field, however, the neutral body is not completely unaffected. It acquires, in effect, a negative charge on the side facing the positive electrode and a positive charge on the side facing the negative electrode. The reason for this polarization, as it is called, is that the atoms composing the neutral body are made up of separate electric charges - positive nuclei and negative electrons. Under the influence of the outside field the electrons and nuclei are pulled in opposite directions, so that the center of negative charge no longer coincides with the center of positive charge. The amount of separation produced by a given electric force (the 'polarizability') varies widely for different materials, but all are influenced to some degree.
The net effect is an excess of positive charge on one half of the body and an equal excess of negative charge on the other. Therefore the two sides of the gross body are also pulled in opposite directions by the field. Since the charges are equal and the field is the same on both sides, the opposing forces exactly cancel.
If, however, the field is made stronger on one side than the other, the forces are no longer in balance, and the body is pulled in the direction of the stronger field. The effect can be demonstrated with electrodes in the form of a pair of concentric cylinders [see fig.1]. In running from the larger to the smaller electrode the lines of electric force converge. This means that the field grows stronger from the outside in. An uncharged body suspended in the space between is seen to move toward the inner electrode.
Note that it travels the same way no matter which electrode is positive and which is negative ! The polarity of the field makes no difference; the only thing that matters is how its strength varies. Thus an alternating voltage applied to the electrodes produces the same result as a direct voltage. This is because the polarization induced in the body switches with the field. Each half is always charged oppositely to the electrode it faces, and the pull of the inner electrode is always greater than that of the outer one.
The motion of electrically polarized matter in nonuniform fields is called DIELECTROPHORESIS [attraction to the strongest part of electric flux]. Compared to the movement of charged particles (ELECTROPHORESIS) [attraction to opposite polarity electrode], it is a mild effect, which is why it has been so long neglected.
When the central electrode is extremely thin, the field immediately around it reaches very high values. The molecules of the suspending liquid can no longer stand the electrical stress; they break down, becoming charged. In effect the concentrated charge on the wire is diffused over a much larger region in a sort of incipient corona discharge. Particles entering the region acquire a charge themselves and are then pushed away from the electrode rather than attracted toward it. A similar breakdown occurs if the applied voltage is too high. Thus there is a critical minimum for the diameter of the electrode and a critical maximum for the voltage.
In fig. 3 is shown a sketch of the action of a properly placed, sharp-pointed electrode with its strongly divergent field. The field apparently repels the liquid below. In reality, this is but a secondary manifestation of the field effect upon the air in the neighbourhood of the pointed electrode. The air molecules are attracted to the point by dielectrophoretic action, then charged at the electrode and subsequently repelled strongly. The pressure of the 'electric wind' pushes down the liquid below. By slightly changing conditions, as by adjusting the height of the wire, or using a wire loop instead of the point, the liquid may be made to rise to the wire instead of appearing to be repelled.
In fig. 4 is shown a sketch of properly arranged electrodes attracting the liquid [of CCL4, benzene, etc.]. This particular effect is capable of widely varied character. Differing arrangements of the electrodes will cause the liquid to move quietly or vigorously at the same applied voltage. By using a sharply pointed central electrode passing up through the liquid, the motion may be made to pump the liquid.
In figs. 5 and 6 are shown sketches of several arrangements in which liquid is made to leave the main body of the liquid at rather high velocities. Fig. 5 shows the drops leaving the dish and 'hanging' in the air around the electrode. Occasionally the individual drops will remain suspended or circling around the lead-in wire for as long as 15 seconds. It would appear that most of these drops have become charged, hence the effect is a combination of dielectro- and electrophoresis. The voltage applied in this experiment was 11,000 volts DC.In another experiment, shown diagrammatically in fig. 6, where CCL4 liquid was thrown over four feet up into the air at about 50 cc/sec, the voltage was applied from below by a small Van de Graaf generator. The voltage was approximately 200,000 volts (negative) at the fine wire electrode. The electrical power input, at 20 ľA was therefore about 4W. The power expended in the liquid rising at the indicated rate and height is about one joule/sec, indicating the electromechanical 'pump' to be about 20 to 25% efficient.
From: Journal of Applied Physics 22, 869-871(1951); 29, 1182-1188 (1958): Scientific American (Dec 1960) 108-116: Journal Electrochemical Soc. 107,390 (1960).