Volume 6, number 7
ETCH PIT DENSITY VARIATION AND ELECTRICAL PROPERTIES OF GaSb SINGLE CRYSTALS GROWN BY THE BRIDGMAN METHOD U.N. ROY and S. BASU Semiconductor Materials Laboratory, Materials Science Centre, Indian Institute of Technology, Kharagpur 721302, India
Received 30 September 1987; in final form 15 December 1987
GaSb single crystals were grown by the vertical Bridgman technique. The distributions of the dislocation density along the diameter and the length of the ingot were studied. The average etch pit density was found to be 1.57~ lo5 cm-2, the lowest value reported so far for Bridgman grown crystals. The variation of electrical properties was also studied along the diameter and the length of the crystal. The resistivity was found to vary from 0.05 1 to 0.064 R cm.
Recently GaSb has become an attractive III-V semiconductor for optoelectronic devices in the IR region. High quantum efficiency IR photodetectors in the wavelength range 1.3-1.7 pm have been fabricated using GaSb Schottky and mesa diodes [ 1,2] for fiber optic communications. GaSb is also having good lattice matching with III-V ternary and quaternary alloys like AlGaAs, AlGaAsSb, GaInSb, GaInAsSb, etc. [ 3-71 for fabrication of heterostructure lasers and detectors. For these applications production of good quality GaSb single crystals for substrate is required. Different techniques like Czochralski, travelling heater method (THM), Bridgman, etc., have been employed to grow good quality crystals by several authors [S-l 51. The growth rate is very low in THM technique. Czochralski grown GaSb crystals frequently exhibit heterogeneity of impurity, striations, microfaceted growth and twins. There are few publications on the growth of GaSb using the Bridgman technique [ 12- 151. The Bridgman grown crystals generally show large grain polycrystalline nature. Lendvay et al. [ 141 grew good quality bi-crystals under microgravity conditions. In this communication we report the variation of dislocation density and the electrical properties along the diameter and the length of GaSb 238
single crystals method.
2. Experimental GaSb single crystals were grown by the vertical Bridgman technique from a stoichiometric melt of Ga and Sb ( 5N purity) slightly rich in antimony. The detailed growth procedure has been described elsewhere [ 16 1. Ga and Sb were properly etched, washed and dried. Stoichiometric quantity of Ga and Sb along with a slight excess of antimony was taken in a quartz ampule and heated at 900°C in an argon atmosphere in a rf controlled furnace for 1 t h for synthesis. The material was then etched and placed in a bottom tipped quartz ampule and sealed under a vacuum of 5x 10e6 Torr. The ampule was then heated in a resistance heated vertical Bridgman furnace at 740°C for 2 h and lowered at a rate of 0.94 mm/h up to 5 10°C at the tip end, followed by 9.4 mm/h up to 300°C and finally cooled to room temperature. The grown ingot was of length 2.5 cm and diameter 1.1 cm. The ingot was cut along the length by a diamond saw cutter into two equal halves. The longitudinal face was then lapped, polished in alumina suspension to mirror finish and finally etched in lHN0, + lHF+ 1H20 for 15 s to reveal etch pits. 0167-577x/88/$03.50 0 Elsevier Science Publishers (North-Holland Physics Publishing Division)
Volume 6, number 7
The variations of etch pit density (EPD) along the diameter at different positions and along the length of the crystal were examined by a computerized image analysis method using the Texture Analysis System (Leitz, Germany). Image analysis is an automatic device for determining microstructural characteristics quantitatively. With on-line computer analysis facility a thorough statistical analysis of the measurement data can be accomplished. Over a particular area the same value of etch pit density was obtained and no error of measurement was indicated by the analyses in repeated experiments. The area over which the etch pits are counted and the number of counts are displayed on a TV screen. Thus the density of etch pits was determined in the present investigation. Great care was taken in the sample preparation which permits accurate determination of the etch pit density by the image analysis method. The area as recorded by the computer for each measurement was 4.8283 x 10d4 cm2. To study the variation of electrical properties like resistivity, mobility and carrier concentrations along the length of the crystal, wafers were cut perpendicular to the length of the ingot from different positions as indicated in fig. 2. Samples from the middle and two diametrically opposite portions (indicated in fig. 4) were taken from a wafer cut perpendicularly to the length of the crystal to study the variation of electrical properties along the diameter. The resistivity, mobility and carrier concentrations were measured by the van der Pauw technique and Hall effect measurements using indium as ohmic contact.
3. Results and discussion The single crystallinity was verified by Laue X-ray photography and by high-resolution microscopy where no grain boundary was observed. The material was found to be p-type from hot probe experiments. The resistivity and mobility were found to vary from 0.05 1 to 0.064 n cm and from 4 17 to 445 cm’/V s respectively over the length of the crystal. Fig. 1 shows an optical micrograph of typical etch pits revealed by chemical etching as already described and the average EPD value was found to be 1.57~ lo5 cme2. Fig. 2 shows the variation of EPD along the diameter and along the length of the ingot.
Fig. 1. Optical micrograph of the typical etch pits developed by the etchant lHNO,+ IHF+ lHzO for 15 s.
The line of scanning at different positions is shown in the insert of fig. 2a. The open circle indicates the EPD distribution near the top of the ingot, whereas the closed circle indicates the distribution near the middle portion of the ingot. The distribution of EPD near the bottom end is shown by the triangle. It is observed from fig. 2a that the EPD increases from the outer region to the core of the crystal. Harsy et al. [ 15 ] also observed higher defect density near the core and interpreted radial heat loss to be the source of thermal stresses responsible for the formation of dislocations. However, the EPD values were reported to be higher towards the outer surface of the ingot for Czochralski grown crystals [ 91. Quenching effects on the growing crystal due to the large temperature difference with the ambient was shown to be the reason for the generation of dislocations. Fig. 3 shows the variation of resistivity. mobility and carrier concentrations along the length of the crystal. The resistivity was found to decrease from the tip end to the upper portion of the ingot, while the carrier concentration was found to increase in the same direction indicating thus the segregation of impurities towards the upper end. The mobility was found to be slightly lower at the tip end. However there was little variation along the length of the crystal. Fig. 3 also shows the variation of EPD along the length of the ingot, where each point gives the average EPD values along the line perpendicular to the length of the crystal. It is seen from the figure that the EPD values increase slowly from the tip end up to the middle of the ingot, and then increase rapidly towards the upper end. The radial variation of re239
Volume 6, number 7
ol ALONG THE DIAMETER
I E "
0 4. w lo5
1.2 I cm
Fig. 2. Etch pit density distribution along the length and diameter of GaSb single crystal grown by Bridgman technique. The inserts show the longitudinal cross section of the ingot and the line of scanning at different positions, along which the etch pits were counted.
wafer cut perpendicular to the length of the crystal is shown in fig. 4. The carrier concentration was found to be minimum at the central portion of the wafer and was found to increase towards the outer portions, showing the non-uniformity of impurity distribution along the diameter. Though the EPD values were higher near the core, the mobility was found
to be maximum,
result due to the
low carrier concentrations near the core. The resistivity however was found to increase from one end to the other.
4. Conclusion The EPD value averaged along the diameter was found to increase towards the upper end of the crys-
Volume 6, number 7
tal. The variation of EPD along the diameter showed a maximum near the core. The average EPD was found to be less than the lowest value reported so far for Bridgman grown GaSb crystals.
Acknowledgement The authors are thankful to Dr. SK Steel Authority of India (SAIL) measurements.
Sen of the for EPD
Fig. 3. Variation of resistivity, mobility, carrier concentration and average EPD along the length of the crystal.
2.si 0.058 E ”
r5 t m
[ 1 ] Y. Nagao, T. Harm and Y. Shibata, IEEE Trans. Electron Devices ED 28 ( 198 I 1407.  F. Capasso, M.B. Panish, S. Sumski and P.W. Foy, Appl. Phys. Letters 36 ( 1980) 165. [ 31 H. Kane and K. Su~yama, Electron. Letters 16( 1980) 146. (41 L.M. Dolginov, L.V. Druzhinina, P.G. Elisev, A.N. Lapshin, M.G. Milvidskij and B.N. Sverdlov. Soviet J. Quantum Electron. 8 (1978) 416. [ 51 T. Sukegawa, T. Hiraguchi, A. Janaka and M. Hagino, Appl. Phys. Letters 32 ( 1978) 3 16.  H. Temkin and W.T. Tsang, J. Appl. Phys. 55 ( 1983) 1413. [ 7 ] C. Caneau, A.K. Srivastava, A.G. Dentai, J.L. Zyskind, C.A. Burrus and M.A. Pollack, Electron. Letters 22 ( 1986) 992. [S] M. Kumagawa, Y. Asabe and S. Yamada, J. Crystal Growth 41 (1977) 245.  S. Kondo and S. Miyazawa, J. Crystal Growth 56 ( 1982) 39. [ lo] W.A. Sunder, R.L. Barns, T.Y. Komedani, J.M. Parsey Jr. and R.A. Laudise, J. Crystal Growth 78 ( 1986) 9. [ 111 K.W. Benz and G. Muller, J. Crystal Growth 46 ( 1979) 35. [ 121 M. H&my, T. GGrog, E. Lendvay and F. Koltai, J. Crystal Growth 53 ( 198 1) 234. [ 131 M. Harsy, F. Koltai, I. Gyuro, T. Go&g and E. Lendvay, Acta Phys. Acad. Sci. Hung. 53 ( 1982) 133. [ 141E. Lendvay, M. Harsy, T. Gijriig, I. Gyuro, I. Pozsgai. F. Koltai, J. Gyulai, T. Lohner, G. Mezey, E. Kotai. F. Paszti, V.T. Hrjapov, N.A. Kultchisky and L.L. Regel. J. Crystal Growth 71 (1985) 538. [IS] M. Harsy, F. Koltai and T. Goriig, Acta Phys. Hung. 57 (1985) 245. [ 161 U.N. Roy and S. Basu, presented at the E-MRS Conference on Growth, Characterization, Processing of III-V Materials with Correlation to Device Performances, 2-5 June 1987, Strasbourg, France.
z 0.052 I. 2 DISTANCE
L cm 1
4 Fig. 4. Variation of resistivity, mobility and carrier concentration along the diameter of the crystal.