DE1040131B - Method for forming electrical semiconductor devices with a needle electrode provided with foreign matter - Google Patents

Description

GERMAN

Rectifiers in which a semiconductor crystal made of germanium or silicon with a point-shaped Contact is known, as are amplifiers, in which two or more point contacts available. However, difficulties arise in the manufacture of such devices certain properties, and it is common in the manufacture of a series of semiconductor devices, which were made from a given body of semiconductor material and which all of the have undergone the same treatment, such devices with properties that in certain Limits fall, weed out. In addition, many such devices must be discarded.

The invention relates to a method for forming electrical semiconductor arrangements with a needle electrode placed on the semiconductor base body and provided with foreign matter. It is according to the invention characterized in that by the formation in the semiconductor on the needle electrode a zone with a larger radius and acceptor defects is generated that continues to have a zone smaller radius and a conduction type corresponding to the foreign matter of the needle electrode within the first zone is generated and that the difference in zone radii by the foreign matter of the semiconductor base body and the needle electrode, the size, direction and duration of the contact area adapted Forming current is dimensioned so that the desired rectifier properties arise.

It is known that tip contacts on semiconductors made of germanium are caused by brief current surges to form. This formation is reversed at the point of contact and in the immediate vicinity the conductivity type of the semiconductor. Since the conversion area is larger than the contact area, With the known arrangement, several forming points can be arranged next to one another in such a way that that a coherent conversion layer is created on the semiconductor. But it is with this one Method not possible to establish contacts that have certain, predetermined rectifier properties exhibit. In a formation according to the invention, however, by changing the Size, direction and duration of the forming current and through a suitable choice of that contained in the needle electrode Foreign matter influences the rectifier properties in the desired manner.

The invention is to be described below with the aid of examples and with reference to the drawing will.

1 and 2 show, on an enlarged scale and partially schematically, a semiconductor electrode system with a limited contact area before and after the method for forming
of electrical semiconductor devices

with a foreign substance
Needle electrode

ίο applicant:

International

Standard Electric Corporation,
»5 New York, NY (V. St. A.)

Representative: Dipl.-Ing. H. Ciaessen, patent attorney,
Stuttgart-Zuffenhausen, Hellmuth-Hirth-Str. 42

Claimed priority:
Great Britain October 2, 1953

Cyril F. Drake, London,
has been named as the inventor

Action of forming pulses according to the invention;

Fig. 3 shows, on an enlarged scale, the shape of a needle electrode for use therein Devices; in

Fig. 4 shows rectifier characteristics, and in

Fig. 5 is the relationship between the forming current and the properties produced thereby Semiconductor device for an N-type semiconductor and a needle electrode with donor impurities shown;

FIG. 6 shows a diagram similar to FIG. 5 for the case in which the needle electrode contains acceptor impurities contains;

Fig. 7 shows the relationships between the forming current and the continuity and blocking characteristics for certain typical systems with N-germanium; in

Fig. 8 is the relationship between the forming current and the maximum operating reverse voltage for the same systems as shown in Fig. 7;

239 640/360

3 4

Fig. 9 shows characteristics of an electrode system in conversion of the semiconductor to the P-type.

as a collector electrode in a transistor. The temperature required to convert an N-half

It is already known that the contact between conductors in a P-semiconductor is necessary

The higher the initial N-conductivity, the higher the needle electrode and semiconductor crystal for the purpose.

of electrical formation, a current flows through it

send. The present invention assumes the pulse of sufficient amplitude and duration

Realization from the fact that when the current passes through changes - a sufficiently steep drop in the impulse is obtained,

in the needle electrode and the semiconductor is in the case of an N-semiconductor and a needle

going on. electrode with donor impurities in Fig. 2.

If a pointed needle 1 (Fig. 1) is placed under io.

Pressure is applied to the surface of a semiconductor body 2. A change in the lattice structure occurs within

stirs, the tip is more or less off a certain isothermal shell 5, which the

flat, as shown at 3. If now an area surrounds 4. Up to an isothermal shell 6

Current pulse from the needle 1 via the contact 3, a conversion from N-type to P-type occurs.

flows between needle and semiconductor, the effect is 15 The isotherm 6 then forms a P-N transition.

If the amplitude and duration of the current are sufficient, the germanium remains outside the isotherm 6

impulses a threefold: of the N-type. Nevertheless, outside of this transfer

1. The material at the end of the needle or the half-turn, as shown in FIG. 2, changes conductor in the immediate vicinity or both are due to the fact that the shell 5 is melted outside the. When the material of the semiconductor 20 is shell 6. The creation of holes alone melts, the material of the needle dissolves and the recombination ratio is so increased in it. In any case, an alloy is formed only within that of the transition 6, ie up to between the material of the needle and that of the isotherm at approximately 300 ⁰ C germanium entHalbleiters. speaks, no noticeable effects occur. Of the

2. Lattice disturbances are formed in the semiconductor around a radius of 25 of the hemisphere 5 is the alloy area with r _{2 in the following.} designated.

3. Contaminants get into the semiconductor in a. It should be mentioned that some appearances immediately around the alloy zone, if there are indications that there are no other, not the needle electrode a certain amount of doping impurities, such as. B. copper, pre-contaminants contains. 30 handesn are responsible for the thermal conversion of the

In Fig. 2, the state after the N-type semiconductor is formed into a P-type semiconductor

depicted how he could have meaning near the end of the needle. With germanium or silicon

trains. however, not a single sample was found in which

The first of the above effects produces a thermal conversion that has not occurred

Area 4 of approximately hemispherical shape, which had 35.

rend of the current passage has melted. However, there were differences in the appearance area 4 contains material of the needle 1 of the halftone with forward and reverse currents in the formation of conductor 2, the ratio of which was determined by the metallurgy of germanium and silicon. This (phase diagram) of the two-component system can relate to such impurities as z. B. Kuphangs. The area 4 becomes conductive and listens to a 40 fer, which behave as exchange ions and to be in a semiconductor. Be transported outside of the area 4 in half-direction, which rises from the directional conductor, the temperature does not depend on the melting sense of the forming pulse, returned point, so that isothermally different shells of.

hemispherical training arise. It was made by if the material of the needle 1 a doping

Experiments found that the radius contains a certain impurity, so this impurity is also in the

th isotherm is directly proportional to the forming current area 4, which consists of an alloy of the

is, if one assumes that the thermal identical needle material exists with the semiconductor material,

weight is reached. This doping impurity migrates into a hemispherical

When a semiconductor is heated, the shaped region 7 surrounding the region 4 becomes. Of the

Lattice structure is more and more destroyed, the higher the transport of contaminants through diffusion is through

Temperature increases. In further training of the invention electrolytic line, which is from the sense of direction of the

The forming current is dependent on the forming pulse in the form of impulses, either promoted or

applied with a short fall time. The so generated inhibited. The distance at which the concentration

Lattice disturbances are caused by the rapid cooling down to a certain value, depends on the

Frozen after the impulse has subsided and two 55 starting concentration of the substance in needle 1 decreases.

Effects generated: In the case of N-type contaminants, an N-P over-

a) The impurities can act as centers of strong BiI- in area 7 at a distance at which the generation of negative charge carriers and strong N-type impurities from needle 1 in the same way Recombination work. In this way the concentration is present as the acceptors are in Service life of the carrier reduced. 60 Area 5. The radius of area 4 is given in the following

b) Suitable impurities act as acceptors and denoted by V _1.

form holes. When there is an N-type semiconductor and the If the semiconducting material 2 is of the N-type, wire 1 contains impurities of the P-type (acceptors), for example N-germanium, the material generally contains those shown in FIG originally an excess of donor impurity conditions, but none is produced in area 7 (which generate electrons). So can the P-N junction. This area consists of one thermally generated acceptors partially or fully- semiconductors of the P-type with very high conductivity, constantly making the donors ineffective. If the The effect of the forming current is then that Concentration of the generated acceptors that changes the conductivity of the semiconductor in aboriginal exceeds existing donors, there is a dependence on the current in the forward direction

5 6

strengthen as the current flows in the forward direction (while trode is immersed in an electrolytic bath, its

the subsequent normal operation) of the defect negative electrode consisted of a platinum coil,

electrons that pass junction 6- The electrodes were connected to a direct current source

ren, and thus connected by the presence of these, their current constant and adjustable

Holes was the conductivity of the bulk of the 5 (i.e., independent of stress). The wire

N-conducting semiconductor increased. Under these circum- stances up to a certain depth in the electrical

levels, the effect of the forming current is lytes, namely perpendicular to its surface,

an improvement in the conductivity of the electrodes. After switching on the current, the

systems in the forward direction without deterioration process of tip formation that the wire in the area

the blocking property. worn away from the contact between the wire and the electrolyte

When a P-type semiconductor is used,. As a result, the resistance increases between

intensifies the thermal transformation in the area 5 needle and electrolyte as well as the voltage. Of the

the P-type properties of the semiconductor, and it's voltage rise is used to limit

there is no transition corresponding to limit 6. triggering relay when a certain voltage

In this case, the needle 1 must be contaminated by the 15 voltage. The relay gets the current

N-type included, and there is only one rectifying switched off.

Transition according to 7 available. Such a surface of the semiconductor, on which the electrode system is then suitable for diodes, the identification needle 1 is placed, is subjected to a chemical or lines according to the curve A in Fig. 4 have electrolytic etching to surface and act as emitter electrodes in a transistor ao borrowed impurities and disturbed surfaces are suitable, layers, caused by the mechanical treatment

In all of the cases considered, one may have to remove it.

given initial system (unformed) the radius r _x In order to obtain a certain characteristic curve enlarged by increasing the forming current in the electrode system, the base electrode must be coffin, which at the same time is an enlargement of the 35 fold attached and should in many cases correspond to a possible radius r ₂ . The difference T ₂ -T ₁ can be inert through borrowed. For this purpose, therefore, the forming current must be changed to any desired value that is suitable when the base electrode is manufactured, and the necessary precautions must be taken. This can, for example, depend on the electrode system by washing and degreasing (after grinding) the difference. 30 of the surface to which the base electrode is attached

The difference T ₂ -T ₁ is determined by: will, be achieved. After that the area will be with

a) the initial content of the semiconductor to sturgeon a copper plate immersed in a bath, the one materials, suitable solution for the production of a copper

b) the composition of the needle 1, contains coating. The current will be approx

c) the initial contact area 3 (Fig. 1) switched on for between 35 15 seconds, using the crystal needle 1 and semiconductor 2, plate forms the anode. That way becomes a

d) the size of the forming current, small amount of the semiconductor material in solution

e) the direction of the forming current. brings. The direction of the current is then reversed, When using an N-type semiconductor and around copper on the immersed face of the crystal

a needle 1 with n-type contaminants hangs the 40 to knock down. The current must be about 5 minutes difference r _% -T ₁ from the concentration of contaminants in the flow in order to produce a semi-glossy copper coating on needle 1. produce. The crystal is then taken out of the bath. The required properties can be that of a rectifier and the coating washed with water or that of a rectifying Kon- and dried. After removing the protective clock, for example a collector electrode in a coating, the copper-plated surface will be a crystal amplifier or transistor. Alloy of 60% lead and 40% tin with the use of a resin flux when the individual factors of formation are fixed. The base electrodes created in this way have the successively produced systems and the base electrode can then be soldered to a suitable self-supporting base that is completely identical after the formation. If a half shaft. 50 conductors of a certain resistance with an inert.The initial contact area depends on the shape of the base electrode is provided and the needle 1 is designed and attached as described above to the original tip of the needle 1, described, can change the pressure between the tip and the semiconductor surface different properties by changing the and to a certain extent on the base material of the amplitude duration and the direction of the forming needle 1. In order to obtain reproducible values, 55 pulses are generated. In all cases, the required characteristic curve was reproducible with needle electrodes from the in.
Fig. 3 used form shown. The values given below, within certain limits, can also all relate to characteristic curves for different initial con-needle electrodes of this type. clock range through the use of forming pulses The needle electrode consists of a wire with 60 with a correspondingly adapted amplitude and the diameter d, which will be two equal bends 8. In other words: within certain and 9 on opposite sides of the axis of the limit the Am-wire must have for a certain characteristic. The height α and the width b of the amplitude of the forming pulse are increased if the bend is precisely defined, as is the distance c from the contact area, and vice versa, the bend 9 from the tip of the wire. Since the 65 If a rectifier with a high directivity shape of the tip of the wire is also desired the contact, the forming current must be influenced in the area, a uniform process was used for the forward direction or in the direction of low production of the tips. A wire, the resistance of the contact flow, that is the direction about 0.127 mm in diameter and made of a device in which the needle electrode is made when using suitable material, was made positive as a positive electrode from N-semiconductor material and

7 8

the needle electrode is used more than 0.01% of doping crystal triode, the emitter electrode of which

P-type or acceptor-type impurities emit a constant emitter current,

contains. These contaminants can be found in the needle material. There is no scale given on the ordinate

or as a coating on the needle. only the characteristic shape of the curves is shown

The question of the contact area between the 5 should be. A forming current, which the tip of the needle and the semiconductor must still corresponds size T, produces a characteristic entfolgendes be said: When low passband speaking curve B resistance at higher current desired in a rectifier (size (Fig. 4) U) curve C appears, while with still the initial contact area must be as large as higher currents (size W) of the rectifier. However, it cannot be destroyed indefinitely. Characteristic curves of type B can become one size. Area between the sizes T and U can be generated,

The electrical properties of a crystal and in the adjoining part of the area between rectifiers are so complicated that it is necessary to select the sizes U and W , the characteristic curve can have some characteristic features for defining times both types B and C together to select the properties. Three such features, ie one area of negative resistance each, both of particular interest for rectifiers in the forward direction and in the reverse direction.
are as follows: The current gain, which is represented by the curve H

1. The forward current is set to be relatively smaller, has a defined maximum. The voltage V _x , for example 1 volt, is generally curve G has two minima, one of which in the area is marked with V (1). 20, in which type B characteristics are generated.

2. The return current at a relatively high voltage V ₂ , these two minima correspond to a maximum, for example 30 volts, generally referred to as V at collector impedance, if the electrode system is (-30). used as a collector in a transistor.

3. The maximum of the working voltage in reverse You can see from the curves G and H , direction, that generally not less than 25 which forming current is required to be a leg voltage V ₃ , which is greater than V ₂ . correct desired characteristic with regard to the current

These three properties are used below to obtain gain and / or impedance,

Called "specific standard identifier". The curves of Fig. 5 are following the ordinate

The maximum working voltage in the reverse direction pushed together when the resistance of the

is the voltage at which a relative 30 Germaniunis or the content of interfering substances in the

sharp bend in the resistance curve of the barrier needle electrode increases or when the forming current

characteristic occurs. "acts in the reverse direction on the rectifier.

Before citing quantitative results, let Fig. 6 show qualitative results that were obtained

4, 5 and 6 will initially be qualitative when the needle electrode is used in place of the

An idea of the effects will be given to the 35 donor impurities those of the acceptor type

by varying the electrical formation and contains. Again, there is no scale on the ordinate

other conditions can be achieved. In Fig. 4 indicated as the curves only the main features

is a number of voltage-current characteristics intended to represent.

a single rectifying contact with N-Ger- the forward and reverse currents represented by the manium. The positive and negative curves / and A 'shown have maxima, values do not have the same scale as a result of the different magnitudes of the current values in the current for slightly different values of the forming. The reverse current corresponds to the forward and reverse directions. Curve A represents the curve K has a minimum with a relatively large forming characteristic of a normal rectifier of relatively currents. From these curves the formation high resistance in blocking direction can be taken. The maximum conditions can be taken that are necessary to mum the working voltage in the blocking direction, for example a rectifier with a mini-voltage at the bend Q of the curve. In the case of curve B mum of forward resistance, a maximum of, the reverse voltage R is significantly reduced and reverse resistance or a maximum of directional resistance increases the size of the negative resistance, to be able to be established. It was found that at rend the positive resistance, which corresponds to the following formation with a needle electrode, which corresponds to the acceptor region, is very small. This curve contains decontamination, the directional ratio corresponds to a rectifier as described in the German patent specification 919 303, which has been enlarged beyond the extent possible up to now. The curve C can.

has an area of negative resistance in the positive.

Ouadrants and corresponds to a rectifier, such as 55, are shown in FIGS. 7 and 8. The conditions

he has already been proposed. The curves D and E under which these curves were obtained

show other characteristics, which after the procedure were the following:

can be obtained according to the invention. Average resistance of a

Quantitative results are shown in Fig. 5, N-type germanium crystal 2.6 ohm · cm

as they do with N-germanium using a 60 "

Needle electrode with donor impurities the same size as the needle electrode

more than 0.01% was obtained. The curves represent ^{vgL * ^lg ' ^ά) ^ ^l ' j? ^mm

the relationship between the amperage (amplitude) ^ ' ^mm

of the forming impulses in the forward direction and the ) '^ ^mm

Forward current of the formed rectifier at 65> · · '"'' ^{u> mm}

constant small positive voltage (curve F), advance of the needle after the

the reverse current with a constant, relatively large " ^first " cycle .... U 1/5 mm

negative voltage (curve G) and the associated ^{duration of the} shape pulse 25 milliseconds

Current gain (curve H) represents when the needle formation was carried out with two formation pulses

electrode designed as a collector electrode of a semiconductor 70.

The curves in FIGS. 7 and 8 relate to needle electrodes made of three different materials were made, namely:

1. Platinum-ruthenium alloy with 10% ruthenium,

2. as under 1, the tip coated with indium, which acts as an acceptor impurity,

3. Phosphor bronze with 0.2% phosphorus as a donor impurity.

The initial contact region at 0.175 mm feed of the needle after the first contact was about 6.5 x 10- ⁷ cm ² for the needle (1), 19.5 x 10 ^-7 cm ² for the needle (2) and 3.9 x 10 - ⁷ cm ² for the needle (3).

Fig. 7 shows the relationships between the forming current (abscissa) and the forward current at 1 volt V (1) and the reverse current at 30 volts V (-30) (ordinate) under the above conditions. The curve marked L applies to an electrode made of platinum-ruthenium, the curve marked M for platinum-ruthenium coated with indium, and the curve marked N for phosphor bronze. The standards for forward and reverse directions are different.

In Fig. 8 are the relationships between forming current (abscissa) and the maximum working voltage in locking (ordinate) for the same needle materials as shown in FIG.

For the method according to the invention, the needle material should be somewhat softer than the semiconductor. Therefore, platinum hardened with ruthenium is useful.

The material to be selected for the needle electrode and the shape of the spring system are determined by a Counting factors determined. When the needle is placed on the surface of the semiconductor and around a If a certain amount is pressed further against the semiconductor, the size of the flattened area depends from the last pressure applied. Also the elastic constants of the spring system must be taken into account. As mentioned above, when the Formierstromes a melting of the needle tip and the semiconductor in the contact zone, and that How far the needle penetrates the molten material is determined by the inertia of the Needle and the time that the semiconductor remains molten. It is therefore important in particular in cases where there must be a small contact area after formation, ζ. B. in devices for high frequency that the movement or penetration of the needle tip to a minimum is reduced. The extent of the molten region is due to the nature of the alloy system formed determined in the transition area between needle and semiconductor.

In those cases where a small contact area is needed, in practice it still is it is desirable that the needle material be as hard as possible, with the caveat that it still ate it is slightly softer than the semiconductor. The hardening of the original material can, for example, by Addition of an inert hardener or, in some cases, the addition of the required active interfering substance self-generated, such as a needle made of silver-phosphorus.

It is desirable in all cases that the thermal and electrical conductivity of the needle be high is to allow the passage of strong forming currents without melting the needle. This is from of particular importance when semiconductors (e.g. silicon) are used, in which a high temperature is necessary for the conversion to the other conductivity type.

It is advantageous to use a needle electrode made of silver in which donor impurities such as phosphorus or arsenic, for example, are homogeneously distributed, which simultaneously serve as hardening agents. Silver also has the advantage of great electrical and thermal conductivity. Alloys with, for example, 0.025 ^fl / o phosphorus are particularly suitable for producing devices with characteristics of the type B and C of FIG.

The formation of an electrode system with precisely specified properties is illustrated by the following values, which relate to electrode systems with a needle made of silver with 0.1% phosphorus on the one hand and silver with 1.0% by volume of arsenic on the other hand, when these are combined as a chamber electrode The characteristics of such an electrode system are illustrated by the curves shown in Fig. 9, which show the current-voltage characteristics of such an electrode system at different emitter currents, the currents on the abscissa and the voltages on the ordinate. Curve X is the characteristic curve for an emitter current of zero, curve Y for an emitter current of 0.5 mA and curve Z for an emitter current of 1.5 mA. These curves apply to germanium with a resistance of 8 ohm · cm and a distance between emitters - and collector needle of 0.127 mm.

The horizontal distance for each ordinate, for example between curves Y and Z, is denoted by α and represents the current gain of the device at this voltage.

The values of the point at which the curve Z has a pronounced knee are denoted by I _C s and V _cs.

In the following table et is given for a voltage of 10 volts and the values / _cs for two different materials of the needle electrode, for different contact areas of the initial contact between the needle and the semiconductor and for two different values of the forming current.

TC O TTf "3 Κ Γ— Silver with 0.1 per cent phosphorus 3.3 Forming current
+ 100 mA ¹ CS Silver with 1 _ , 0 ° / o arsenic 4th J.XL / JLl LctJuhl '
diameter Forming current
+ 400 mA 3.9 α at 10 V 2.4 Forming current
+ 40OmA .— - Forming current
+ 100 mA - mm α at 10 V 4.7 0.9 α at 10 V I _es 3.6 α at 10 V! J
CS 5.7 0.004 1.6 3.8 2.6 -. 2.4 - · 0.005 1.9 2.8 1.0 - - - - 0.0065 2.0 1.6 2.2 2.3 0.0075 1.5 0.7 - - 0.010 . 1.5 - 0.8

209 640/360

11 12

Column 1 contains the initial contact area. If the duration of the impulse is not sufficient,

between the end of the needle and the germanium. A second pulse can be applied.

Columns 2, 3, 4 and 5 contain the values for a. It has already been mentioned that curve B of

Silver wire with 0.1% phosphorus, the columns 6, 7, 8 Fig. 4 shows a rectifier characteristic that a

and 9 for a silver wire with 1% arsenic. 5 rectifiers according to German patent specification 919 303

From columns 2 and 3, in which α corresponds to 10 volts. According to the method according to the invention and Ies for an emitter current of 1.5 mA, such electrode systems can be shown in large numbers, it can be seen that constant values are always produced.
Dimensions such as the contact area at a given The characterizing of such devices, which is enlarged for forming current ^{, α and Ics DIS} are suitable for io breakover circuits, the exit of a maximum rise and then fall again. Defects from the base electrode (which in Das results from the fact that, in this case, an active electrode and not an inert electrode as the initial contact area is enlarged), which for the distance between area 5 and 7 (Fig. 2) tip contact flow which reduces a bias in itself. The factor α is inversely proportional. The result is similar to this distance (r ₂ - r _t ), since α depends on the number of results as when using an inert one of the electrons that come from the area T ₁ via the base electrode, if the necessary negative carriers reach the area r ₂ - r _{x can get} into the bulk of the semiconductor through the impact of light or some other beam material without them being captured by the development on a point near the tip cone impurities in the area r ₂ - r _x . 20 cycles can be generated. The light generated pairs of Ics depends on α and also on the collector current for electrons and holes, and the hole in the hole depends on the emitter voltage zero. trons are conveyed to the needle electrode and

Above a certain contact area, an effect similar to that of defect electrons is sufficient,

the forming current does not generate a molten zoner that is generated by the active base electrode at one of the

of sufficient size, and the value α 25 of the rectifier is assumed,

drops sharply. The pairs of electrons and holes

From column 4 and 5 it can be seen that, apart from in the base electrode or in others, with a smaller forming current the enlargement electrodes are also generated in the body of the semiconductor material of the contact area an initial increase of α, whereby the amount of generated and Ics causes and that Then these values again 30 pairs, among other things a function of the current that falls, but the forming current is insufficient to flow through, the applied voltage and a sufficiently large molten zone r ₁ to produce the material with suitable properties. and all values of α are small. A tilting effect occurs in such devices,

If now the Störstoffgehailt the needle 1 (Fig. 2) without electrons being injected or light is increased tenfold as in columns 6, 7, 8 35 rays must be noticed.

and 9, the value of α increases to the extent that it has been found by experiments that the device distance of the transition 7 (Fig. 2) from the origin in which the semiconducting material increases through the lines, and it becomes a smaller one Forming current forming pulses was suitably converted to required for a maximum of α , because as a result of producing suitable layer systems, and also increasing contaminant content and because of a molten zone that produces only a small amount of electrons and holes as a result of the pair in such devices With a low forming current, the higher impurity content can be obtained, which was able to have a good PN transition and therefore offer TU of FIG. 5. Lattice perturbations that produce a good value of α in front of. the formation are present, are known to work

The two different contaminants that have been used as centers for the generation of electron defect electrodes, (Phosphorus and arsenic) are of the same type. tron pairs, however, this effect is small in the ver-id it is not possible to use a silver needle with 1% phosphorus - the same as the one deliberately used by the beschriephor therefore arsenic was produced for the bit-forming process when the half-games used in columns 6 to 9. head is selected appropriately.

The results, which are given in the above table 5 ° The effect of the forming current, an electrode are, as a means of several ver system for a device of the type described search in every case, but the results of the ver generating can be better understood from Fig. 2 Different attempts do not differ very much from becoming mediocre. If the semiconductor is N-type and the worth from. The reproducibility of the needle electrode obtained contains N-type contaminants, then Properties can be derived from the following listing 55 must be the distance between the boundary of the N-leivon six different examples of the third series of tendon area 7 made by injection of N-sturgeon from Column 2 can be removed. substances is generated, and the active base electrode

(Hole source) comparable or smaller

Average _se j _na j _{s is} the diffusion length of the holes im

α 1.9 2.3 2.0 1.8 2.3 1.8 2.0 6o semiconductor material. When using another

/ _cs 2.8 5.6 4.8 4.8 5.7 4.6 4.7 Hole sources are the same conditions

authoritative.

The duration of the forming current is determined by other examples for the formation of diodes

the time that is required to finally equalize with special characteristic curves are given below.

to generate weight in the area r ₂ , but there will be 65.

no upper limit for this. On the other hand, germanium is left with a resistance of 2 ohm · cm

the forming current do not flow longer than necessary. The and a needle electrode made of silver with 0.1% phosphor

maximum time is, for example, with the Ger- phor and 0.125 cm in diameter was elec-

manium only a few milliseconds. The Formierstrom trischem way provided with a tip that 35 to must have the shape of a pulse with a steep slope of 7 ° 55 ° and at the end a through-

had a knife of 0.0025 mm. The needle electrode was shaped as shown in Figure 3 and was as previously described Dimensions. One or two forming pulses of 900 mA were from a power source high impedance applied for 25 msec. The electrode system obtained with an inert base electrode has an area of negative resistance in the pass band of the characteristic.

For a diode with a low forward resistance, germanium with 0.4 ohm · cm resistance and uses an electrolytically sharpened aluminum wire with an angle of 35 to 55 ° at the tip, in which the diameter of the tip was less than 0.0025 mm. One or two pulses from 500 mA and a duration of 25 msec was applied in such a way that the needle electrode was positive. In this way, diodes having a low on-resistance were obtained.

Even if devices were described in the examples in which germanium as a semiconductor is used, the invention is not limited to this material. It can after Process also electrode systems are produced in which one needle electrode with another Semiconductors, such as silicon, is in contact. With other semiconductor material, the The amplitude of the forming current differs from that used for germanium due to the different thermal conductivity in the semiconductor material and the other position of the isotherms through which the lattice disturbances are generated from. For silicon z. B. are the temperatures at which lattice disturbances are generated higher than with germanium.

The invention is not intended to relate solely to the exemplary embodiments described and illustrated the substances mentioned must be restricted.

Claims (7)

Patent claims: 1. A method for forming electrical semiconductor arrangements with a needle electrode placed on the semiconductor body and provided with foreign matter, characterized in that a zone with a larger radius (r ₂ ) and acceptor defects is generated by the formation in the semiconductor on the needle electrode, that furthermore a zone with a smaller radius (r _t ) and a conductivity type corresponding to the foreign matter of the needle electrode is generated within the first zone and that the difference in the zone radii (r ₂ - r _t ) by the foreign matter of the semiconductor body and the needle electrode, the size, direction and duration of the The forming current adapted to the contact surface is dimensioned in such a way that the desired rectifier properties arise. 2. The method according to claim 1, characterized in that that to produce a high directivity and a low forward resistance a needle electrode with more than 0.01% of an impurity of the P-type (acceptor) and a N-type semiconductors are used and a forming current of 50 mA to 2.5 A so through that Electron system is sent that the needle electrode is positive. 3. The method according to claim 1 or 2, characterized in that for generating a collector electrode for a transistor, an N-type semiconductor material and a needle electrode with more than 0.01% donor impurities are used and that the amplitude of the forming current is chosen so that a predetermined ratio between current gain and collector impedance is achieved. 4. The method according to claim 1 or 2, characterized in that to achieve rectifiers with the same negative slope in the pass band of the characteristic a semiconductor from N-type and a needle electrode with more than 0.01% N-type impurities are used and a forming current of 0.75 mA to 1.25 A is applied in the forward direction. 5. The method according to claim 1 to 4, characterized in that germanium as the semiconductor is used. 6. The method according to claim 3, characterized in that the needle electrode consists of silver. 7. The method according to claim 1, characterized in that to achieve a large directional ratio an N-type semiconductor and a needle electrode made of platinum-ruthenium alloy containing 10% ruthenium are used, and that the amplitude and direction of the forming current according to the absence or the Presence of acceptor impurities in the needle electrode is selected. Considered publications: German Patent Nos. 861282, 879 270; German patent application R 5804 VIIIc / 21g, 11/02 (published December 6, 1951; British Patent Nos. 595 601, 686 958; French Patent No. 930,783; Journal of Appl. Physics, Vol. 24, February 1953, pp. 162 to 166; RCA-Reviev, March 1953, pp. 17 to 21; Journal of Applied Physics, Vol. 5, 1953, pp. 163, 164. For this purpose 2 sheets of drawings 809 640/360 9.58