US3051876A - Electrolytic transistor - Google Patents

Description

Aug. 28, 1962 J. BARDEEN 3,051,376

ELECTROLYTIC TRANSISTOR Original Filed June 2, 1953 3 Sheets-Sheet 1 ja\ l ,fnunzr'. 170% Bar e en.

Aug. 28, 1962 J. BARDEEN 3,051,876

, ELECTROLYTIC TRANSISTOR Original Filed June 2, 1953 3 Sheets-Sheet 2 Aug. 28, 1962 J. BARDEEN 3,051,876

ELECTROLYTIC TRANSISTOR Original Filed June 2, 1953 3 Sheets-Sheet 3 32 d 34 1 O 32 c 32 Ia l I a 1 rzz/rz Z 01 Zakrz3 r 6 United States Patent Ofiice 3,051,876 Patented Aug. 28, 1962 3,05ll,d76 ELECTROLYTIC TRANSIFsTUR John Bardeen, Champaign, Ill, assignor to The University of Illinois Foundation, a non-profit corporation of Illinois Continuation of application Ser. No. 359,014, June 2, 1953. This appiication May 1.4, 1958, Ser. No. 735,660 2 Claims. (Cl. 317-231) This invention relates to an electroytic device and more particularly to a circuit element that may be termed an electrolytic transistor.

Basic research in the field of semi-conductors has culminated in the development of semi-conductor devices, generally termed transistors, with which currents can be controlled to provide amplification, oscillation and the like. See, tor example, United States Patents 2,524,035 to this applicant, dated October 3, 1950 and 2,569,347 to Shockley, dated September 25, 1951. The device of this invention has some operational characteristics which are similar to those of semi-conductor transistor-s, particularly the so-called junction transistor of the Shockley patent but differs in that the medium in which the controlled current flow takes place is an electrolyte rather than a semi-conductor; this gives rise to the terminology electrolytic transistor. As will appear fully later, the electrolytic transistor has many characteristics which make it suitable for use in applications where semi-conductor transistors are unsatisfactory.

One feature of the invention is that it comprises a first electrode, a second electrode, a material in operative relatiOn with the electrodes which contains particles capable of migrating to one of the electrodes and effecting a change of charge therewith, and a third electrode for alfecting or controlling the migration of the particles. Another "feature is that the third electrode establishes the material or medium used at an operative potential relative to one of the other electrodes. A further feature is that the first and second electrodes are closely spaced or adjacent each other while the third or base electrode is relatively remote therefrom. Another feature is that means are provided, in the material or medium surrounding the electrodes, which accept a charge from one of the first or second electrodes and deliver it to the other. And a further feature is that an ionizable medium is used and a reversible electrochemical reaction occurs at the first electrode which results in a net change of charge with the medium, the ions or neutral atoms formed in this reaction migrating to the other electrode where a reverse reaction occurs resulting in a change of charge with the second electrode; the net result is a transfer of charge from the first electrode to the second electrode with the rate of reaction at the first electrode, and thus the rate of charge transfer to the other electrode, being controlled by the potential of the third electrode.

Another feature is that the current flow in the system, from the first electrode to the second electrode, is independent of the potential of the second electrode and can be controlled by an external resistor so as to be substantially independent of temperature. And a further feature is that the device may he used as a constant current source for a variable load.

Other features and advantages of the invention will be readily apparent from the specification and from the accompanying drawings, in which:

FIGURE 1 is a sectional view of an embodiment of the invention;

FIGURES 2a. and 2b are schematic diagrams of circuits showing the use of two different forms of the invention;

FIGURES 3, 4 and 5 are schematic diagrams of semiconductor transistor amplifiers utilizing the electrolytic transistor of this invention as a bias current source;

FIGURE 6 is a schematic diagram of a circuit using an electrolytic transistor as a current amplifier;

FIGURE 7 is a schematic diagram of a modified form of the invention, having a fourth electrode;

FIGURE 8 is a sectional view of a modified form of the invention having a series of alternate plates for the first and second electrodes;

FIGURE 9 is a sectional view of a modified form of the invention using concentric cylinders for the first and second electrodes;

FIGURE 10 is a diagram showing the rectification characteristic of the device;

FIGURE 11 is a diagram showing the transistor current characteristic of the device; and

FIGURE 12 is a diagram showing the output characteristic of the device.

In the embodiment of the invention shown diagrammatically in FIGURE 1, a container 2% has therein a first electrode 21, a second electrode 22 and a third electrode 23. Electrical connections are made to each of these electrodes 21, 22 and 23 by wires 24, 25 and 26, respectively. The container is filled with a suitable transfer material or medium 27. Preferably, in order to control the operation of the device more accurately, only one surface of each of the electrodes is in contact with the medium 27; accordingly, the back face of the electrodes 21 and 22 and the wires 24 and 25 are insulated from the medium 27 by suitable material 28 which should not react with the medium.

The material or medium 27 used in the device contains particles which are capable of migrating through the medium and of effecting a change of charge with the electrodes. A good example of such a medium is an electrolyte; that is a medium which contains ions capable of effecting a change of charge with one or more oi the electrodes under proper conditions.

For example, as shown in FIGURE 2a, the first electrode 21 may be biased positively with respect to the base electrode 23 by a voltage source shown as a battery Evil. The ions in the medium or electrolyte are normally in the reduced state and the desired oxidationreduction equilibrium condition may be maintained by a suitable choice of the material of the base electrode 23. Ignoring for the time the second electrode 22 and its associated circuit, a current will flow from electrode 21 through the material 27 to electrode 23, which current may be controlled by rheostat 31. When a relatively low voltage is impressed between electrodes 21 and 23, the current which flows will be dependent merely on the resistance of the material or medium 27. However, as the voltage increases a series of reactions will occur result ing in a substantially increased how of current through the cell. Depending of course on the medium used, these reactions may be similar to those encountered in the process of electrolysis or electrodeposition. In the circuit of FIGURE 2a for example the electrode 21 is positively biased with respect to electrode 23 and functions as an anode while electrode 23 serves as the cathode. Ancther way of stating this is that an oxidation reaction takes place at electrode 21 while a reduction reaction takes place at electrode 23.

FIGURE 10 shows an operating curve 32 for this circuit, still ignoring the effect of electrode 22 and its as sociated circuit. The portion of the curve 32a illustrates the relatively slow increase of current I as the voltage, V, is increased below the potential sulficient to initiate the electrochemical reactions referred to. The break, 32b, in the curve occurs when the voltage, V, becomes suflicient to sustain the reaction. The current then increases rapidly, "320, with a relatively small incremental increase of the voltage V. In the curve shown in FIG- URE 10, the increase in current falls off in the area 32d when the supply of carriers to the electrode is limited by diffusion. If a more highly concentrated solution were used, this limiting effect would occur at a higher current.

In the circuit shown in FIGURE 2b, the electrolyte is normally in the oxidized rather than the reduced state; the reactions discussed above are reversed as is the polarity of battery 39. Oxidation takes place at electrode 23 and reduction at electrode 21.

When the electrode '22 and its associated circuit 33 are added, it has been found that at least some of the current instead of flowing between electrodes 21 and 23 flows from electrode 21 to 22, While the rate of current flow is still determined by the potential between electrodes 21 and 23. If the electrode 22 is placed sufficiently close to the electrode 21, substantially all the current will flow directly between electrodes 21 and 22 rather than through electrode 23.

As this operation is analogous to that encountered in semi-conductor transistors, analogous terminology and current and voltage convention have been adopted. Electrode Zll will hereinafter be referred to as the emitter, 6, electrode 22 as the collector, c, and electrode 23 as the base, b. Positive currents and potentials will be assumed as shown in FIGURE 2.4 and as specified in the book by Shockley, Electrons and Holes in Semi- Conductors, at page 36.

It is desirable, in order to minimize power loss in the device, that the base-to-material or base-to-electrolyte resistance be as small as practicable. One way of accomplishing this is shown in FIGURE 1 Where the base 23 is relatively large and has a large surface area in contact with the material 27. The entire container 2% may be made the third electrode if desired. The resistivity of the electrolyte should also be low.

It has been found that the potential of the collector, V has, at least within certain ranges, no effect on the collector cur-rent, I This is shown in FIGURE 12 Where the collector current, l is plotted as a function of the collector voltage V for a fixed emitter voltage. The straight line portion of the curve, 34a indicates that the collector current I is independent of the collector voltage V through a fairly wide range. The portions of the curve 34b and 340, where the current changes markedly, occur when the voltage V exceeds the potential necessary to carry on the electrochemical reactions. Within the range shown however the collector current is dependent only on the emitter voltage, V. Thus, the voltage applied to the collector and accordingly the load connected in the collector circuit may vary considerably without affecting the current flowing therethrough.

In one particular embodiment of the invention which has been operated, the casing 20 was made of Lucite, while the emitter 21 and collector 2.2 were circular platinum discs. The base electrode 23 was formed by a mercury pool in the bottom of the container. All me tallic parts which would come into contact with the solution, except for the faces of the emitter and collector, were coated with Lucite-acetone cement to insulate them from the solution and to prevent undesired reactions. Polystyrene-CCL, cement might also "be used.

The solution used as a medium contained 0.25 M FeSO and 1.5 M HCl. It was found necessary to increase the emitter voltage, V,, to about 0.4 volt to initiate the desired reaction. In this case, the ferrous ion, Fe++ is oxidized at the emitter, losing an electron, to form a ferric ion, Fe+++. The oxidized ferric ions then migrate or travel by a process of diffusion to the collector 22 where the reverse reaction occurs, the ferric ion gaining an electron, being reduced, to form a ferrous ion. Thus, a flow of current is established between the emitter and the collector.

The base electrode 23 establishes the potential of the medium or electrolyte 27 and effectively controls the flow of current. Technically, the potential difierencebetween the emitter 21 and the base 23, V determines the rate of the electrochemical reaction at the emitter, and the flow of current, as shown in FIGURE 10. How ever, as substantially all of the ions which are oxidized at the emitter migrate directly to the collector, the potential of the base 23, by controlling the rate of creation of oxidized ions, will hereinafter be said to affect or control the migration of the ions or particles, in the sense of the quantity migrating to the other electrode, since this is for all practical purposes a direct function of their rate of creation.

The hydrochloric acid used in this solution performs Several functions. First, a number of the Clanions react with the mercury pool forming the base 23 to provide a thin layer of mercurous chloride, Hg Cl This calomel electrode provides a low electrode-to-medium resistance and helps to maintain the electrolyte in the reduced state. The acidic solution also results in an improved conductivity and prevents iron hydroxides from precipitating out of the solution.

Another solution which has been used consists of l M NaCl and 0.2 M HCl. With this solution, the reactions are based on the chloride-chlorine couple. The chloride ion, Cl, is oxidized at the emitter to form chlorine, at least part of which is reduced at the collector. When large currents are passed through the electrolyte, some of the chlorine molecules form a gas given off as bubbles at the emitter so that the reaction is not completely reversible.

The bromide-bromine couple has also been studied and has been found to operate successfully. No gas evolution was observed.

It is desirable, of course, to use an electrolyte in which the reactions are reversible rather than one in which a gas is given off or a solid deposited to prevent deterioration and possible exhaustion of the cell. Thus, it is more desirable to use the ferrous-ferric or bromide-bromine systems rather than the chloride-chlorine system as the first mentioned reactions are reversible.

The operation of both mediums which have been described involves an oxidation reaction at the emitter, 21, and accordingly the emitter must be biased positively with respect to the base 23 or medium 27, FIGURE 2a. If the analogy to semi-conductor transistors is extended, this device corresponds to a pup junction transistor. An electrolytic transistor corresponding to an npn junction transistor may be provided by using a medium containing a salt which has a positive ion that is predominately in the oxidized state, for example, soluble salts of V+++. With a medium of this type, the emitter 21 would be biased negatively with respect to the base 23 as shown in FIGURE 2b; the carrier ions would be reduced at the emitter, each receiving an extra electron, and then migrate by diffusion to the collector where they would be oxidized to their former state, giving up the electron. Again, the potential of the base controls the rate of creation of reduced ions at the emitter and hereinafter will be said to affect or control the migration thereof.

It appears that the predominant state of, or equilibrium condition for, the ions in solution (i.e., reduced or oxidized) is primarily a function of two properties of the cell; the oxidation-reduction potential of the ion couple used, and the oxidation-reduction potential of the base or reference electrode used. It is my belief that if the oxidation-reduction potential of the base is more positive than that of the ion couple in the medium, the ions will tend to be predominately in the reduced state when the system is in equilibrium. For example, if a calomel (Hg CI base electrode is used with a ferrous-ferric ion couple, the ions will tend to be in the reduced (Fe++) state as the oxidation-reduction potential of the calomel base is 0.270 volt and that of the ferrous-ferric couple is 0.77 volt. On the other hand, if the oxidation-reduction potential of the base is more negative than the oxidation-reduction potential of the ion couple, the ions in solution will tend to the oxidized state. For example, if a solution containing vanadium ions (the V++ -V+ oxidation-reduction potential is 0.20 volt) is used with a calornel base electrode, the ions in solution will be predominately oxidized.

The mobility of the ions in the solution is relatively poor as compared with the mobility of electrons or holes in semi-conductors. The device will not respond to frequencies much above 1 c.p.s., the cut-off depending on the mobilities and on electrode spacing. This characteristic gives rise to several possible uses for the device.

Broadly, the electrolytic transistor can be used as a constant current source for a variable load. In the circuits shown in FIGURES 2a and 2b, the collector current, I flows through a load illustrated as a two-terminal network 33. The magnitude of this current is determined by the voltage between the emitter 21 and the base 23 and is, within limits, completely independent of the magnitude of the load 33.

One specific example of such a load is a semi-conductor transistor. In transistor circuits, it is necessary to bias at least one of the transistor elements by a current of the proper polarity in order that the transistor will operate in the proper range. Particularly with the junction transistor, the DC. resistance between elements varies rather markedly with temperature changes; sometimes as much as per degree centigrade. It is diflicult with ordinary power supplies to provide a constant current through a resistance which varies so greatly. The electrolytic transistor is particularly well adapted for use as a transistor bias source as it has a constant output with a varying load, it can be designed to have extremely low impedance to alternating currents, and to have the bias current relatively unaffected by temperature changes.

FIGURE 3 shows an electrolytic transistor 35, of the same type shown in FIGURE 2a, used as a bias source for an npn semi-conductor transistor amplifier 36 connected for grounded base operation. The bias current is controlled by the battery 37 and rheostat 38 in the circuit of the emitter of the electrolytic transistor 35. The base 40 of the electrolytic transistor is connected to the emitter 41 of the semi-conductor transistor, while the collector 42 of the electrolytic transistor is connected to the base 43 of the semi-conductor transistor. The signal to be amplified may be applied to the transformer 44 in the circuit of the emitter of the semi-conductor transistor while the output of the stage may be developed across resistor 45 in the circuit of the collector 46 of the semiconductor transistor.

FIGURE 4 shows a similar arrangement with a pup semi-conductor transistor, 47. Here the collector 42 of the electrolytic transistor is connected to the emitter 48 of the semi-conductor transistor while the base 40 of the electrolytic transistor is connected to the base 49 of the semi-conductor transistor. Again the bias current is controlled by potentiometer 38. An input signal may be coupled to the amplifier through transformer 44 and the output signal is developed across load resistor 45.

FIGURE 5 shows another modification in which an npn semi-conductor transistor 50 is connected for grounded emitter operation. Here, the base 40 of the electrolytic transistor 35 is connected to the emitter 51 of the semiconductor transistor, while the collector 42 of the electrolytic transistor is connected to ground 52. A signal may be fed to the amplifier between terminals 53, connected to base 54 of the semi-conductor transistor and terminal 55, connected to ground. The output of the stage is developed across load resistor 45 in the circuit of the collector 56 of the semi-conductor transistor. Again, the bias current is controlled by rheostat 38 in the emitter circuit of the electrolytic transistor.

FIGURE 6 shows the electrolytic transistor used as a current amplifier. Here again, the electrolytic transistor is of the pnp type shown in FIGURE 20: with the emitter 61 positively biased. A current to be amplified may be connected between terminals 62 and 63 in the circuit of the base 64. The amplified output will appear in the load 65 connected in the circuit of the collector 66.

FIGURE 7 again shows an electrolytic transistor of the type shown in FIGURE 2a, with the addition of a fourth electrode, a second base, 70. This electrode 70 is established at a different potential from the base 23 by a battery 71, in this case making electrode 70 more positive than base 23. An electrostatic field will be set up between electrode '70 and electrode 23, with electrode 70 being more positive than electrode 23. As the charge carrying particles in this embodiment of the invention are positively charged, the electric field between electrodes 70 and 23 will cause the particles to concentrate in that portion of the medium relatively adjacent electrode 23, resulting in a decreased resistance between the base 23 and the medium or electrolyte 27.

FIGURE 8 shows a modified arrangement of the emitter and collector electrodes. As shown here the emitter comprises three physically spaced but electrically connected plates 21 which are sandwiched between four larger, electrically connected plates 22 forming the collector. The collector plates 22' are substantially larger than the emitter plates 21 and this together with the sandwiched arrangement insures that substantially all the charged particles leaving the emitter will migrate or diffuse to the collector.

Another arrangement of the emitter and collector electrodes is shown in FIGURE 9 where the emitter is a single cylinder El and the collector comprises a pair of cylinders 22" one inside and the other outside the emitter. Again, practically all charged particles leaving the emitter will migrate to the collector.

This application constitutes a continuation of my application Serial No. 359,014 filed June 2, 1953, and now abandoned.

While I have shown and described certain embodiments of my invention, it is to be understood that it is capable of many modifications. Changes therefore, in the con- Stl'llCllOIl and arrangement may be made without departing from the spirit and scope of the invention as disclosed in the appended claims.

Having now described the invention, What is claimed is:

1. An electrical system of the character described comprising: a device having a first plate-like electrode, a second plate-like electrode adjacent and in close proximity thereto While substantially uniformly spaced therefrom, a medium containing ions in operative relation with said electrodes, and a base for establishing said medium at an operative potential; means for biasing said first electrode with respect to said base for causing a change in charge of ions at that electrode, said ions being capable of migrating to said other electrode and effecting a change in charge therewith, the electrode base potential controlling the migration of said ions; and a transistor having a plurality of elements, the circuit of at least one of said elements being connected in series with said second electrode and base for biasing said transistor.

2. An electrical system of the character described, comprising: a device having a substantially plate-like first electrode, a substantially plate-like second electrode of substantially like size positioned adjacent and parallel thereto, a medium containing ions in operative relation with said electrodes, and a base for establishing said medium at an operative potential; a source of direct current potential connected in series with a variable resistance between said first electrode and base for bias ing said first electrode with respect to the base to cause a change in charge of ions at the first electrode, said ions being capable of migrating to said second electrode and affecting a change in charge therewith, the first electrodebase potential controlling the migration of said ions; and a transistor having a plurality of elements, the circuit of at least one of said elements being connected in series With said second electrode and base for biasing said transistor.

References Cited in the file of this patent UNITED STATES PATENTS 1,409,383 Lane Mar. 14, 1922 1,439,526 Mershon Dec. 19, 1922 1,497,430 Chubb June 10, 1924 1,877,140 Lilienfeld Sept, 13, 1932 8 Lilienfeld Mar. 7, 1933 Hammond Aug. 29, 1933 Mershon Feb. 27, 1934 Van Geel June 18, 1935 Lilienfeld Feb. 19, 1952 Trent Ian. 12, 1954 Root July 27, 1954 FOREIGN PATENTS Germany Oct. 29, 1919 France Feb. 12, 1925

Images