Ohmic contacts to cadmium sulphide films

Ohmic contacts to cadmium sulphide films

NOTES Solid-Stale Electronics, 1973, Vol. 16, pp. 95 l-954. Pergamon Press. Printed in Great Britain Ohmic contacts to cadmium sulphide 6lms (Recei...

247KB Sizes 2 Downloads 56 Views

NOTES Solid-Stale Electronics, 1973, Vol. 16, pp. 95 l-954.

Pergamon Press.

Printed in Great Britain

Ohmic contacts to cadmium sulphide 6lms (Received

9 November

1972)

ACCORDING to the simple energy band model, aluminium electrodes favour an ohmic contact to cadmium sulphide. It is, however, difficult to obtain ohmic aluminium - cadmium sulphide contacts by simple evaporation and this is attributed to the formation of an insulating interfacial layer of aluminium oxide which is brought about by oxygen adsorption at the semiconductor surface or by the effects of residual gases. To overcome this difficulty, trivalent metals such as indium or gallium may be diffused into the surface where they act as donor impurities. Thomson and Cornwall [l] have recently described a diffusion technique for ohmic contact formation. Alternatively, exposure of the crystal to ion bombardment yields surfaces which produce ohmic contacts with most metals[2]. The apparent source of conduction electrons is the highly dis-

ordered region near the crystal surface and, in the model proposed by Kroger er a1.[3], ohmic contact of metal to a highly resistive n-type semiconductor depends on a thin layer of lower resistivity separated from the junction by an extremely narrow barrier through which electrons may tunnel readily. Ohmic contacts have also been produced by sequential multilayer evaporation techniques[4] and by passing large currents through non-ohmic contacts to cause barrier breakdown ]5,61. Allan et al. [7] have recently reported the growth of single crystal epitaxial films of cadmium sulphide on sapphire substrates using a sublimation method. Single crystal layers were grown with exceptionally high resistivities, 107Rm in the dark. Electrical measurements were made on these films using evaporated aluminium electrodes. In spite of its unreliable nature, this method of contact formation is attractive because of its simplicity, especially in the case of thin layers. The results obtained using this method are reported below. In the search for ohmic contacts to cadmium

V

x IO-‘A

Fig. 1. I-V Characteristics

for aluminium electrodes temperature. 951

deposited on CdS films at room

952

NOTES

Type (b) contacts 4

x IO-‘A

Fig. 2. I-L’ Characteristics

Type

ofaluminium

electrodes

deposited

on CdS at 300°C

I

(cl contacts

4-

--II Cl

-1, I-

cz

2

x IO-‘A

Fig. 3. I-V Characteristics

sulphide the aluminium evaporation varied in the following ways:

(4 Evaporation (b) (4

for aluminium

process

electrodes films.

was

of aluminium electrodes on to untreated cadmium sulphide films with substrates at room temperature. Evaporation of aluminium electrodes on to untreated cadmium sulphide films with the substrate at 300°C to reduce gas adsorption. Evaporation of aluminium electrodes on to ion bombarded cadmium sulphide films with the substrate at room temperature.

deposited

on ion bombarded

CdS

Ion bombardment of the cadmium sulphide was performed in an atmosphere of argon at 5 x 1OP’ torr and evaporation of aluminium to form the contact was performed at approximately IOP torr. The aluminium was of SN purity. Only that portion of the cadmium sulphide film on to which the aluminium was evaporated was exposed to the ion bombardment. The nature of the contact formed in each case was examined. To determine the voltage drop across the contact region, samples were probed with tungsten wire etched to a 2 pm dia. point

NOTES

953

50,o=16x10-2~m 40-

30V

20100 k.i2

CdS

IO7

I

I

I,

I 5

0

I

I

I

I

I IO

I

I

,,I,, 15

x lO‘5A

Fig. 4. I-V Characteristics

for a low resistivity CdS film.

IOO-

p=l07x10’Ltm

0

2

4

6

8

-lop, x IO

Fig. 5. I-V

Characteristics for a high resistivity

and indium plated, mounted in an xyz manipulator. In all cases the probe was positioned adjacent to the contact and the contact to probe potential difference was measured by an electrometer of input resistance 1014R over a range of currents passed in both directions. Owing to formation of an insulating inter-facial layer of aluminium oxide it was not expected that contacts produced by the first method would exhibit reproducible electrical characteristics. The results are, however, included as a comparison (Fig. 1). Heating of the substrate to 300°C during the evaporation of aluminium contacts was per-

SSEVol.

16No.S-H

CdS film.

formed with a view to reducing gas adsorption on the semiconductor and, although this procedure yielded more reproducible results (Fig. 2), some traces of non-ohmic behaviour persisted. The effect of ion bombardment on the electrical characteristics of the aluminium contacts is illustrated in Fig. 3, which, in contrast to Fig. 1, indicates reproducible contacts of an ohmic nature. Resistivity measurements were performed on the films using the circuit indicated in Fig. 4, which employs a digital voltmeter of input resistance 1Oron or an electrometer of input resistance 1Ol4n for voltage measurements. The samples

NOTES

954

were enclosed in a light tight compartment during investigation. The I-V characteristics for low and high resistivity films, which were grown at different substrate temperatures. are presented in Figs. 4 and 5. Ion bombardment of the cadmium sulphide film is considered to increase the surface conductivity by increasing the contribution of surface states through an increase in their density; in addition it aids the removal of any adsorbed impurity atoms. It is not possible to identify the dominant mechanism, but the ohmic nature of the contacts fabricated using the ion bombardment technique is clearly illustrated. The method is shown to be useful over a wide range of resistivities. i.e. 1.6 X lo-'- 1O7Rm. are grateful to the Science Research Council for providing support for this work.

Acknowledgement-We

Department ofElectrical Engineering University of Edinburgh Scotland

DOUGLAS ALASTAIR MALCOLM

D. M. ALLAN J. HAY A. REID

REFERENCES 1. M. J. Thomson and M. G. Cornwall, Solid-St. Elect tron. 15,861 (1972). 2. W. M. Butler and W. Muscheid, Ann. Phys. 14, 21.5 (1954); IS,82 (1954). 3. F. A. Kroger, G. Deimer and H. A. Klasens, Phys. Rev. 103,279 (1956). 4. K. W. Boer and R. B. Hall, J. appl. Phys. 37, 4739 (1966). 5. Y. T. Sihvonen and D. R. Boyd, J. crppl. Pi1y.s. 29,

1143 (1958). 6. M. Itakura and H. Tayoda, Jap. J. crppl. t’hys. 3. 197 (1964). 7. D. D. M. Allan, M. A. Reid and W. E. J. Farvis. Nnture, t’hys. Sri. 236, 64 (1972).

S,,l,d-srarcE/P<,10,1,< .1.1973,Vol

16.~~

954-955

Pergamon PI-C+\\.

Printed m Great Britain

The uniformity of vapor deposited S&N, film* (Received

1November

1972: in revisedform 1973)

4 January

use is made of the masking and passivating properties of S&N4 in the production of semiconductor devices. In recent years, a considerable amount of work has been carried out in determining INCREASING

*This work was performed at Raytheon Co., 2 Wayside Rd., Burlington, Mass.

the properties of this film. A minimum of information is available on the factors that govern the uniformity of film thickness within a certain length of the furnace zone. In the present work this problem is investigated using SiCl, and NH3 as the sources. Recently Kohler[l] observed large differences in the growth rate of the film along a deposition boat using SiCl, and NH,,. To eliminate this difference, a sloping temperature profile of the boat was required. Wohlheiter and Whitner[2] used exceedingly high N, carrier gas flow rate (> 100 l/min) to obtain uniform deposition. This required the use of additional heating chamber for N, carrier gas prior to entering the furnace tube. In the present work the deposition boat was maintained at a constant temperature while relatively low gas flow rates were used, thus eliminating additional heating chamber. The deposition system is indicated in Fig. 1 [3]. It consists of a source bottle designed NH3+N2

NZ

Fig. I Geometry of the SiCl, source bath and deposition furnace tube. such that N, gas flowing parallel to the surface of SiClj carries its vapor into the furnace tube. The concentration of SiCl, in the furnace tube is determined by the temperature of the source bottle. The reaction between SiCI, and NH, occurs at point A which is at approximately similar temperature as the constant zone temperature. The advantages of this type of deposition geometry are as follows: (1) The bubbling of N, gas through SiCI, is eliminated. This avoids droplets of SiCI, entering the furnace tube and hence prevents non-uniform deposition. Wohlheiter and Whitner[2] observed that the deposition rate was a function of NT gas bubbling through SiCI,. In the present system, N, gas flows parallel to the surface of SiCI,. A fluctuation in the rate of N, gas flow does not effect the uniformity or the rate of deposition.