CHAPTER II

CONSIDERATION OF CELL TYPES AND THEIR CHARACTERISTICS

=A selenium cell= consists essentially of two electrodes of brass or copper bridged by a thin layer of metallic selenium. When connected into a circuit with batteries and other apparatus the current flows from one electrode to the other thru this selenium bridge. Since the resistance of the selenium to an electric current depends upon the amount of light falling upon it the flow of current thru the cell will be controlled by the brilliancy of the illumination.

=Metallic selenium= being opaque, the light penetrating but ¹/₅₀,₀₀₀th of an inch as calculated by Marx, it is necessary that the selenium layer be extremely thin in order that the light may affect an appreciable proportion of the total conducting area. This condition is never reached when the electrodes lie parallel to each other with the selenium between them. However by arranging the electrodes so that the current flows at right angles to the plane of the selenium surface we can cause all the current to flow thru the light affected area. This can only be accomplished by making use of a transparent conductor for one electrode.

To realize the importance of the above factors a description of the various types of cells developed by the many investigators in this field will be of great assistance. The different workers made use of various arrangements of the electrodes but the cells fall into certain classes. These types have been named after the inventor or the one most prominent in the work on them.

=The Bildwell cell= is possibly the best known type. It is made by winding two bare wires of copper, brass, german silver or platinum on a sheet of mica or slate. The wires are spaced about ¹/₃₂nd of an inch apart. The size of the wire is of little importance, the usual practice being to use #28 wire on a form measuring two by one inches. In Fig. 1 is shown this type of construction using a mica form, the wires being fastened by passing them thru holes at the ends of the sheet.

The selenium is applied to the cell by melting it over the wires. The cell is laid on a mica covered copper plate supported over a bunsen burner. The temperature of the cell is raised to the point where a stick of selenium when touched to the cell melts. The entire surface of the cell is coated with the selenium in a very thin layer, smoothing out the lumps with a sheet of mica or a steel knife. To get a satisfactory coating the temperature must be regulated closely, if too low the selenium turns grey and the temperature must be increased to melt it, if too high the selenium collects in drops due to surface tension and is as difficult to spread as mercury. The proper state is a semi-fluid condition which it attains at 220° C when it can be easily manipulated.

When a satisfactory surface has been obtained the cell is transferred to a copper plate to cool while the bunsen burner is turned down to give a temperature of 120° C. When cool the cell is replaced on the hot copper plate and allowed to heat up again. Shortly the whole surface will turn grey in color due to the selenium crystallizing. The temperature is now slowly increased till the selenium shows signs of melting, this will be indicated by the edges turning black. The bunsen burner is immediately withdrawn and the edges allowed to recrystallize. The burner is turned down a trifle and replaced under the hot plate. The cell is watched carefully for signs of melting and if none appear it is left so for three or four hours. If it melts again the burner should be further lowered, just sufficient to keep the cell a trifle below the melting point of the selenium. The cell is then allowed to cool by lowering the burner by small amounts extending over a period of an hour. This prolonged heating and slow cooling is known as annealing.

After the above treatment the cell is complete save for mounting. The usual method is to mount the cell in a small wooden box fitted with a glass window to admit the light, leads being brought from the electrodes to two binding posts mounted on the box. This protects the cell from moisture and dust.

It will be apparent that with the above method of construction it is impractical to get the extremely thin layer of selenium necessary if the light is to affect a relatively large proportion of the total area. This will be even more clear from an examination of the cross sectional view of this type of cell as shown in Fig. 1. Here we have a comparatively thick layer of selenium bridging the space between the wires. Of this layer only the thin surface film facing the light drops in resistance while the interior part is unaffected. This means that should the surface layer drop to even ¹/₅₀₀th of its dark resistance the total drop of the cell would be much less.

=The Ruhmer cell= is similar in construction to the Bildwell, differing only in the form of support. A porcelain or glass tube is used to support the parallel wires as shown in Fig. 2. When porcelain is used the constructor can fasten the wires at the ends by slipping them into slots cut with a hack saw. With glass some other means are necessary to hold the wire while winding.

The selenium is applied to the cell in the same manner as the Bildwell and then annealed. The method of mounting the cells as devised by Ruhmer is worthy of mention. The cell unit is enclosed in a glass tube and then the air exhausted. By attaching the leads to an incandescent lamp base a very convenient arrangement results. The cell is well protected from all external influences and is therefore more stable and reliable. This is perhaps the most important improvement in this type of cell.

We have in the Ruhmer cell conditions almost identical to that in the Bildwell, namely a large area of conducting selenium that is beyond the range of the light and hence not affected thereby. This is offset to a certain extent by the large area exposed to the light as these cells can be employed with a parabolic reflector to cause the light to fall on all sides. This type of cell was employed by Ruhmer in his experiments with the Photophone. He succeeded in transmitting speech for a distance of four miles using a speaking arc at the transmitting station.

The cell developed by Bell and Taintor in their experiments is rather novel in the arrangement of the electrodes. As shown in Fig. 3 the electrodes take the form of brass disks separated by thin mica disks supported by two brass rods, the whole being clamped together by nuts on the ends of the rods. The disks are one inch in diameter, eighteen or twenty being sufficient for a small cell. By drilling the holes in the disks of different sizes and assembling them as shown it is possible to have alternate disks connected to the same rod. After assembling and clamping the cell skeleton is chucked in a lathe and the surface turned smooth and polished.

The selenium is applied by heating the form and melting it on, by rolling the cell back and forth over the hot plate it is possible to get an extremely thin film of selenium on the smooth surface offered by the cell. The coating is then annealed in the regular manner. This cell can well be mounted in a glass tube and the air exhausted.

The advantage gained by this form of construction is the thin film of selenium obtainable due to there being no spaces between the electrodes into which the selenium can flow. A cross section of this type is given in the illustration. It will be seen that although the main defect of the cells mentioned previously has been reduced to some extent still it has not been removed entirely.

=The Mercadier cell= is similar in many respects to the Bell but is easier of construction. Two strips of thin copper one half inch wide are wound into a spiral being separated from each other by strips of mica. One face of the flat spiral is filed flat and then polished. The selenium is melted onto the cell and then smoothed off with a strip of mica. The cell then being annealed as described previously.

We have here a condition analogous to that in the Bell cell, the only advantage being ruggedness and simplicity against a loss in active surface area. For experimental purposes this cell is entirely satisfactory for if it does not prove sensitive the selenium coating can be filed off and another applied.

We come now to the consideration of cells wherein the current flows at right angles to the surface of the selenium. This implies the use of at least one electrode on the surface of the selenium. The Gripenberg cell has both electrodes on the surface. As shown in Fig. 5 the electrodes are made by depositing a thin film of gold on a glass plate and with a sharp tool removing narrow strips of it to form a grid, alternate bars of which are connected to the same terminal. The grid arrangement is shown in the detail illustration.

The selenium is not applied to the grid in the molten state. A thin plate of metallic selenium is obtained by melting the metal on a glass plate and then applying pressure with a cold glass plate. After annealing the selenium will be found to adhere closely to the hot plate in a thin film. The cell is then assembled by placing the plate with the gold grid in a frame having a portion of one side cut out to form a window. The plate with the selenium adhering is placed over the grid and forced into contact with same by means of a small screw.

A study of the cross section of this cell shows that the current inflowing from one electrode to the other must pass the surface of the selenium. Since the gold film is semi-transparent the light passes thru it to affect the selenium and will have a maximum effect upon the resistance of the cell. This object is attained at the expense of loss of illumination since the gold film cuts off all but the green rays of light. The advantage just mentioned more than outweighs this loss. Could a transparent conductor be found this defect would be removed entirely.

=The Fritts Cell= is little known except by name but is superior to the others both in simplicity of construction and correctness of design. In this case the selenium is melted directly on a copper plate that serves as one electrode. While soft, pressure is applied by a non-adherent plate to obtain the thin film necessary. The selenium enters into chemical combination with the copper and adheres firmly to it. After cooling the selenium film is covered with gold leaf to form the other electrode. The cell unit can be mounted between two strips of fibre as shown in Fig. 6, a sheet of very thin mica serving as a protection to the gold foil surface.

We have here the ideal condition. All the current must flow thru the light affected area in passing from one electrode to the other. The light passes thru the semi-transparent gold foil to effect the change in resistance of the selenium film. The only disadvantage lies in the diminishing of the light strength by the gold film. Despite this however the Fritts cell has proved the most sensitive cell made having in one case a ratio of 337 to 1, that is, the resistance in the light is but ³/₁₀ths of one per cent of that in the dark.

The six types of cells described cover all the types worthy of special mention and will enable one to select a cell for experimental purposes. In view of the fact that the Gripenberg and Fritts cells are superior in point of design the selection of the cell becomes a question of mechanical difficulties to be overcome. The construction of the grid in the Gripenberg type is rather difficult unless an engraving machine is obtainable whereas the Fritts cell requires very simple apparatus for its construction.

A factor often overlooked in considering the design of selenium cells is the relation that light and electricity bear to each other. These are manifestations of the same force and their interaction can be taken advantage of in the Fritts cell due to it being possible to cause the current to flow in the same or opposite direction to that of the light vibrations. The importance of this will be covered in detail later.

No mention has been made of just how the light affects the selenium to reduce its resistance. This is still a moot point, one theory being that the light being electromagnetic in nature causes the molecules of the selenium to cohere in a manner similar to that of the radio coherer used in the early days of radio telegraphy. However the conduction thru a selenium cell is similar to that in an electrolyte and differs from metallic conduction. Considering this the light may act to ionize the selenium in some manner and make possible the more rapid interchange of the ions from the opposite terminals. Although a thorough understanding of the action taking place may lead to the improvement of selenium cells it is not within the scope of the present work to consider the various phases of this part of the problem, confining itself as it does more to the practical production of the cells.