JPS5921183B2 - charge coupled device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention is particularly useful in the field of electronic technology, particularly shift registers, delay lines, storage systems, etc., in which the substrate is formed by two semiconductors having mutually different forbidden band widths. Retina (reti)
charge-coupled devices (charg) that can be applied to fields such as
e-coupled devices).

A charge-coupled device (abbreviated as CCD) is a semiconductor system in which charges are accumulated in a potential well formed on the surface of a semiconductor, and these charges are transferred by displacing the potential well. be.

Very generally speaking, these devices include a doped semiconductor substrate, a layer of insulating material, and an array of metal electrodes at an appropriate potential. Therefore, this structure is metal-insulator-
It is of the semiconductor type, and the insulator can be composed of an oxide, in particular. The charges displaced in this way in this type of device are minority carriers in the semiconductor substrate, for example holes if the semiconductor is n-type. For basic information on the general properties of these charge-coupled devices, see the Bell System
Technical Journal” (BellSystemTe)
W.S. Paul (Boyle) and G. Chnical Journal) published in Volume 49, 1970.
, E. Smlth (page 587) and G. F. AmeliO'. M. F. Top set (TO
Olpsett) and J. E. Smith
You can refer to Mr. San's paper titled ``Experimental proof of the idea of a charge-coupled device.''

In these devices, it is necessary for the charge transfer to be unidirectional, thus creating an asymmetrical potential well for the minority carriers that is deeper at the downstream end than at the upstream end. It is necessary.

Many means are known for creating wells of this type. For example, three clocks can be used, each clock connected to one of the three electrodes by three intersecting control lines, and the number of clocks and associated control lines In order to reduce this to two, it is possible to create a surface area in the semiconductor under one electrode that is more heavily doped under the upstream edge of the electrode than under the rest of the electrode. The present invention proposes new means for obtaining unidirectionality of charge transfer within this type of device.

Generally, this consists in placing two different semiconductors with different bandgap widths in the region where the asymmetric well is to be created. As will become clear from the following description, this solution offers many advantages, in particular it provides a very simple manufacturing method and results in an improvement in the properties of the charge-coupled device to which it is applied. In particular, the invention provides a substrate constituted by a first doped semiconductor, an array of metal electrodes arranged sequentially along an axis and separated from the substrate by a thin insulating layer, and a minority charge carrier. means for ensuring unidirectionality of movement of the carrier along the axis by creating an asymmetrical potential well in the surface area of the substrate, the depth of which is greater in the downstream direction than in the upstream direction; means for injecting the at least one electrode below the first electrode;
The present invention relates to a charge-coupled device comprising at least means for detecting the presence of a charge under the electrode and means for placing the electrode at a cyclically varying potential of an appropriate value.

Means for creating an asymmetric potential well in the surface area of the substrate and ensuring unidirectionality of movement of the carrier includes a forbidden band located in the substrate at one end of the area and different from the first semiconductor. a region formed by a second semiconductor having a width, one semiconductor having the smaller bandgap being located downstream of the region; The features and advantages of the invention will become clear from the examples of configurations of the invention which will now be described with reference to the drawings.

This configuration is provided as an example and is not intended to limit the invention in any way. In the following discussion, only the n-type semiconductor Q example will be discussed where the minority carriers (which are the moving charges in the device described here) are holes.

The schematic diagram of FIG. 1 shows the energy distribution of different bands of two semiconductors with two forbidden bands of different widths.

The left part shows the conduction band with energy shown as ECl and the valence band with energy shown as EVl.
nceband). Similarly, the right-hand part shows a semiconductor SC2 containing a valence band with an energy denoted by EC2. At the junction of these two semiconductors, when equilibrium is achieved, the Fermi levels of energies indicated by EF coincide. Eigenenergy, by definition, is the average energy between the conduction band energy and the valence band energy. From conventional considerations in semiconductor theory, the position of the phenolmi level with respect to the valence band is defined by the following relationship. N. , where E8 is equal to the width of the forbidden band, that is, Ec-
Ev, K is the Boltzmann constant, T is the absolute temperature, Nc is a constant determined only by the temperature and the mass of the charge carrier, ND
Negative is the density of the carrier.

If it is desired to make the valence band energy of the semiconductor SC2 larger than the valence band energy of the semiconductor SCl, the following conditions must be satisfied. here,
NDl and ND2 characterize the doping of the semiconductors SCl and SC2.

Inequality (1) is equivalent to the following inequality: If the O dopings NDl and ND2 and the constants NCl and NC2 are closely related, then the inequality (2) becomes the forbidden band Egl and E
It is immediately satisfied that the energies of g2 are slightly different (which is the case).

Considering the case of silicon and germanium as an example, E
g are 1.1ev and 0.7ev, respectively, so
There is a difference of 0.4ev between the forbidden band values. 1 of this model
In the case of paired semiconductors, the magnitude of the exponent in equation (2) is very small and the inequality is satisfied if the doping and the constant Nc are not too different from each other. The state of the valence band is as shown in the diagram of FIG.

From this figure, it follows that the holes in the semiconductor SCl and therefore the charges transferred in the charge-coupled device are constrained to move from the valence band of the semiconductor SCl towards the valence band of the semiconductor SC2. Thus, an electric field is created which acts in the minority carrier in a unidirectional manner, which is the main object of the invention. Although Figure 1 concerns the general case, when doping and constant NC are closely related, in other words NDl+ND
・If NCl≠NO2 in 2, the conduction band energy EC
l and EOi will have very close values.

Similarly, by calculating the difference between the Fermi energy and the characteristic energy, it can be seen that the two characteristic energies of the semiconductors SCl and SC2 coincide in the specific case below. This can be written as follows.

nl is the density of charges in the conduction band (or equivalently the number of holes in the valence band) for the corresponding native semiconductor.

Therefore, it is clear from the above explanation that the conditions controlling the doping of the two semiconductors are not critical, as long as there is a sufficient energy difference between the two forbidden bands.

In practice, it is advantageous to adopt a situation in which the two dopings are almost the same, thereby simplifying the fabrication.
In FIG. 2a, an asymmetric potential well is created by adding to the first semiconductor SCl a region constituted by a second semiconductor SC2 having a forbidden band width different from that of the first semiconductor. A cross-section of the area of the charge-coupled device being constructed is shown.

In the diagram of FIG. 2, semiconductor S. 2 is located downstream with respect to the direction D of movement of the minority carrier.

From the diagram in FIG. 1, it can be seen that the semiconductor SC2 has the smallest forbidden band width. If the semiconductor SCl is silicon, the semiconductor SC2 is, for example, germanium. Besides the two semiconductors SCl and SC2, the device comprises an insulating layer 10, which has any desired shape and can support other devices not shown, and which is characteristic of a charge device. be. 2b shows the shape of the surface potential V8 along the area shown in FIG. 2a.

This potential is the potential of the interface between the semiconductors 10. Since minority carriers are attracted to semiconductor SC2, the potential well has a greater depth at the level of semiconductor SC2 than in the rest of this area. This therefore results in a stepped configuration of the form shown in Figure 2a. semiconductor SC
a second semiconductor SC2 having a smaller forbidden band in l;
By adding conduct so as to obtain. 3 and 4 show the application of the above means to a charge-coupled device.

FIG. 3 shows a charge coupled device with a single control line.

FIG. 3a is a cross-sectional view of a device of this type, in which the first semiconductor S
Cl, on which are placed an insulating layer 12 and an electrode 14, the electrodes 14 all being configured to deliver a voltage VG varying cyclically between two limit values V1 and V2. It is connected to the lead wire 16 from the terminal.

Furthermore, the device of FIG. 3a, on the one hand, has a second
on the other hand, and on the other hand, it has an identical region 22 formed of the same semiconductor SC2 in the gap between the electrodes.

The surface potential of the interface between the semiconductor SCl and the insulator 12 is the third
b varies along the device as shown in figure b.

For a given voltage VG, an asymmetrical potential well is observed under each electrode, which is deeper in the downstream direction than in the upstream direction, depending on the nature of the two semiconductor regions located under each electrode. . The depth of these wells is determined by the voltage VG applied to electrode 14. The two values of these potentials for the two limit values V1 and V2 of the voltage VG are shown in FIG. 3b. For the same reason, an asymmetrical potential well with a deeper depth in the downstream direction than in the upstream direction is observed within the interelectrode gap, but the surface potential within the interelectrode gap is almost independent of the potential applied to the electrodes. The well is permanent since it is determined to be the average value of V, and 2 at the end of a sufficient time interval, regardless of the surface resistance being properly chosen. It should be noted here that the use of a second set of electrodes interposed between the first electrodes is connected by a line similar to lead 16 to a constant potential source equal to said average value. It is also possible. The device is then ready for operation, but with added complexity. The operation of the charge-coupled device according to the invention is as follows. Applied voltage VG
is equal to V1 (in principle, this voltage is negative for an n-type substrate), the depth of the potential well created under the electrode is at most such that positive charges are trapped in it in the downstream area. Ru.
When the voltage G takes a value of 2, the depth of the potential well below the electrode 14 is reduced and the charge that can be trapped therein is reduced due to the permanent asymmetry of the immediately adjacent downstream electrode gap. potential is moved into the well. These charges are transferred into the potential well created under the subsequent electrode when the voltage G returns to the value 1. The voltage G can have any waveform, for example a rectangular wave. The only condition to be satisfied is that the values of V1 and V2 be such that the minimum depth of the potential well downstream of the electrode is less than the depth of the potential well upstream of the interelectrode gap; and the maximum depth of the potential well upstream of the electrodes must be greater than the depth of the potential well downstream of the interelectrode gap. Under such conditions,
The transfer of charge can take place during the voltage cycle, first from the area located under the electrode towards the interelectrode gap, and then from this interelectrode gap towards the area located under the subsequent electrode. The first means according to the invention for producing unidirectional charge transfer applies not only to charge-coupled devices with a single control line, but also to two charge-coupled devices of the type shown in FIG. It also applies to charge-coupled devices with control lines.

This device has a substrate made of a first semiconductor SCl and an insulating layer 30, and an electrode 3 is placed on the insulating layer.
2, 34 are arranged, which are connected to lead wires 32', 34', and these lead wires are connected to two clocks 321, 34'.
341. The means for producing unidirectionality of charge transfer is the second one having a smaller bandgap than the semiconductor SCl.
by adding a region 40 made up of semiconductor SC2 and placed under the downstream end of each electrode. In this situation, an asymmetrical potential well with a greater depth in the downstream direction than in the upstream direction is created under each electrode, and the device thus formed operates in the same way as a charge-coupled device with two control lines. do. As an example 1
Patent Application No. 47-6408 filed on January 14, 1972, or “1EEE Journal of Solid State.
circuit” (1EEEJ0urna10fS011d
``Charge Coupled Digital Circuits,'' published in October 1971, Vol. 6, No. 5. F. Kosonotsuki (KOsOnOck)
y) and G. E. Reference may be made to the paper by Carmes. In the above two charge-coupled devices containing one or two control lines, a correction is made across the semiconductor substrate, thus adjusting the voltage Vs so that the correct values of VO and V2 are symmetrical with respect to ground. is always possible.

If the semiconductor is n-type as assumed earlier for explanation, then
For this purpose, it would be possible to apply a uniform p-type deposit to the semiconductor soil, for example by ion implantation of boron. In addition to the advantage of having symmetrical voltages with respect to ground, this structural variant makes it possible to "bury" the moving channels of minority carriers, thereby eliminating recombination at the surface of the semiconductor. Improve equipment performance. A first advantage of the measure according to the invention is that it provides a planar charge-coupled device, in other words, the form of the potential is not stepped and is extremely simplified compared to comparable devices of the prior art. The goal is to provide a device that is A second advantage lies in the fact that charge-coupled devices with a single control line are extremely easy to manufacture. In any case, the solution proposed by the invention does not involve strict conditions regarding doping. In particular, the doping need not be light, so this limits the influence of carrier recombination. In order to form the semiconductor substrate necessary for the application of the invention, small-width crystals of two different semiconductors may be juxtaposed, or a large-sized crystal of a first type of semiconductor may be placed inside a large-sized crystal of a semiconductor of another type. It is clear that it is easy to introduce arrays of small width crystals of semiconductors.

Such a form of construction appears to be more theoretical than practical, and the most convenient way to obtain the desired result is to
Starting from a large-sized crystal of a semiconductor of the type mentioned above, by introducing atoms of the other type of semiconductor, a row of areas of small width can be formed locally within the bandgap of the first type of semiconductor mentioned above. The goal is to bring about significant change. The width of the forbidden zone of the area thus formed can be varied sequentially up to a maximum width corresponding to the second type of semiconductor. The production of semiconductors according to the method described above can be carried out by diffusion techniques, but is more easily achieved by ion implantation.

The examples of silicon and germanium mentioned above are not intended to limit the invention, and in particular, silicon substrates such as GaAS, . Gap. GaSpe1nAs
、． 1np,. l-type such as 1nSb or 0dSe
10dTr,. ZnSe,. It is possible to create areas constituted by -type compound semiconductors such as ZnTe.

In the above case, it is only necessary to carry out two successive implantations of ions of the two elements of the composite semiconductor.

Finally, the best feature of the device according to the invention is the silicon-on-insulator technology (SiliOn-0n-1ns).
ulat0rtechn010gy).

[Brief explanation of the drawing]

FIG. 1 is a schematic diagram showing the energy distribution in different bands of two semiconductors used together according to the invention to obtain unidirectionality of charge transfer.

Claims (1)

[Claims] 1 a substrate constituted by a first doped semiconductor, an array of metal electrodes arranged sequentially along one axis and separated from the substrate by a thin insulating layer, and an upstream substrate for minority carriers; means for creating an asymmetrical potential well in the surface area of the substrate with a greater depth at the downstream end than at the end, thus obtaining unidirectionality of movement of the carrier along said axis; and at least the first electrode. means for injecting minority carriers under the at least one last electrode; means for detecting the presence of a charge under at least the last electrode; a voltage source; means for creating an asymmetrical potential well in the surface area of the substrate and obtaining unidirectionality of the carrier movement, comprising:
formed by a second semiconductor having a different bandgap than the first semiconductor, and constituted by a region located in the substrate at one end of the area; A charge-coupled device characterized in that the semiconductor having the forbidden band is located downstream of said zone.