Materials Science and Engineering, A 168 (1993) 75-79 75

Study of the effects of C O 2 laser irradiation on Cdo.204Hgo.796Te

T h i e r r y V o g t

Centre de Recherches NuclOaires, IN2P3-CNRS, UniversitO Louis Pasteur, Groupe Recherches Physiques et MatOriaux, BP20, F-67037 Strasbourg Cedex 2 (France) Institut Franco-Allemand de Recherches (ISL), BP34, F-68301 Saint-Louis Cedex (France)

R e n d J o e c k l 6

lnstitut Franco-Allemand de Recherches (ISL), BP34, F-68301 Saint-Louis Cedex (France)

C l a u d e S c h w a b

Centre de Recherches Nucldaires, 1N2P3-CNRS, UniversitO Louis Pasteur, Groupe Recherches Physiques et MatOriaux, BP20, F-67037 Strasbourg Cedex 2 (France)

Abstract

The behaviour of p-type Cd0.204Hg0.796Te used in IR detectors was studied under high power CO 2 laser irradiation at room temperature. Its melting threshold was determined for two irradiation conditions. (1) For a pulsed focused laser (pulse duration 3 = 0.3 ms, spot diameter 225/~m), the melting threshold fluence ranged from 13 to 38 J cm-2. (2) For a continuous wave CO 2 focused laser beam (spot diameter 240/~m, I = 360 kW cm -2) swept on the target using a rotating mirror, the threshold fluence was F = 11 J cm -2, corresponding to a scanning speed of 8 m s -~. These experimental results were compared with calculated values assuming either a monodimensional model or a two-dimensional combined absorption and conduction model. The threshold calculated with the monodimensional model was F = 15 J cm -2, whereas the threshold calculated with the two-dimensional model was F = 19 J cm-2 for the pulsed irradiation and F = 17.5 J cm 2 for the continuous irradiation. A decrease in mercury concentration from 39.8 at.%) down to 13.1 at.%, depending on the fluence, was observed using X-ray microanalysis.

1. Introduction

CdxHg~l_x)Te (CMT) is the main detector material used for night vision. Several investigations of the inter- action of laser light with CMT were aimed at the removal of lattice damage associated with doping by ion implantation of this material. However, the first attempts in this direction were not as successful as for silicon. This is related to the different chemical nature of these two materials, i.e. an elemental as opposed to a compound semiconductor. There is indeed the possibility of a compositional change of CMT during heating, owing to volatile element losses even before melting [ 1 ].

The purpose of the present work was to determine the minimum fluence to obtain permanent damage, as revealed by surface melting, under high power laser irradiation conditions, corresponding to typical exposures for a single pixel in an IR camera. It was measured experimentally and compared with theoreti- cal estimations. The rests were done on bulk Cd0204Hg0.796Te suitable for the fabrication of IR devices for the 8-14 #m range [2], to understand the fundamental laser-matter interaction prior to investi- gation of real devices. The illumination conditions with

CO2 lasers (10.6/lm) were selected to cover the two present types of IR camera. The possible modifications of composition were measured by X-ray microanalysis.

2. Experiments

The samples used for the tests were p-type Cd0.2o4Hg0.796Te single crystals supplied by the Colorado Research Laboratory, Walsenburg, CO. All tests were done at room temperature. The minimum energy necessary to obtain a morphological modifica- tion on the surface of the sample was determined by scanning electron microscopy. The laser beams used for the experiments were assumed to be gaussian.

2.1. Pulsed irradiation For a camera with a linear array of detectors and

one-dimensional sweeping, the mean irradiation time of a pixel by a laser beam focused by the camera optics is approximately 0.3 ms. To simulate this irradiation time, a CO2 laser was used for which the high voltage supply of the electrical discharge could be modulated to give an adjustable laser pulse duration. For the tests, this pulse duration was r = 0.3 ms. The laser beam was

0921-5093/93/$6.00 © 1993 - Elsevier Sequoia. All rights reserved

76 T. Vogt et aL / Laser irradiation of Cdo.204Hgo,7~ Te

focused onto the target with a diameter of 225 /~m. Part of the incoming laser beam was reflected by a potassium chloride prism and measured by an energy meter.

Several tests were done with energies ranging from 2.15 to 20 mJ. Figures l(a) and l(b) show two craters resulting from irradiation with 7.6 and 14 mJ energies. The melting threshold corresponded to an energy value between 2.6 and 7.6 mJ. (Unfortunately, it was not possible to obtain intermediate values between these two limts.) The fluence threshold derived from the relation F = 2E/S (where E is the energy and S the laser spot surface) is then within the range 13-38 J c m - 2.

2.2. Continuous irradiation The optical source used in these experiments was a

76 W continuous wave CO2 laser. In a single detector camera, the irradiation time of a pixel is several micro- seconds. To obtain such short times, the focused laser beam was swept onto the sample by a rotating mirror. The adjustabl e revolution speed of the mirror allowed the sweeping speed of the laser beam over the sample to be varied. Figure 2 shows the sweeping system. The laser spot diameter was 240 ktm, the maximum power density w a s / m a x = 360 kW cm-: . The spot diameter is divided by the scanning speed to obtain the irradiation time, the fluence is given by F = It.

Several sweeps were done on the sample with scan- ning speeds ranging from 4.0 to 8.5 m s-l , the corre- sponding fluences varying from 10 to 22 J cm -2. The resulting damage was a groove whose width increases with fluence. Figure 3(a) shows grooves obtained under these conditions; Fig. 3(b) shows the groove resulting from a scanning speed of 5.5 m s-~ with a maximum fluence Fma x = 16 J cm- 2

The melting threshold was obtained for a scanning speed of 8 m s-1. The corresponding irradiation time and fluence were respectively r = 30 / i s and F = 11 J c m - 2.

COz laser beam

rllllTor

ax~ ofrota~on "~- of the mirror

Fig. 1. Craters obtained with pulsed irradiation: fluence at the centre of the spot (a) Fma x = 38 J cm-2, (b) Fma x = 70 J cm-2 (spot diameter 225/~m, z = 0.3 ms).

s of sweeping of the laser beam

Fig. 2. Experimental set-up for the sweeping tests.

T. Vogt et al. / Laser irradiation of Cdo 2o4 Jig 0 7,)6 Te 77

3. Theoretical melting threshold

T h e melting thresholds were estimated for both conditions of irradiation, using a conventional model and a more sophisticated model developed by Gautier at ISL [3].

3.1. M o n o d i m e n s i o n a l absorpt ion m o d e l For a first approximation, monodimensional behav-

iour is assumed, i.e. we assume no radial heat conduc- tion. Since at room tempera ture the absorpt ion coefficient of C M T for a wavelength of 2 = 10.6/~m is ct = 100 c m - 1 [4], the optical absorpt ion depth is about 100 ,urn. For an irradiation time of r = 0 . 3 ms, the thermal conduct ion depth given by the relation D = 2(gr) °'5 [5] is D = 35 /~m; for r = 30 /~s, we have D = 10 ktm (7¢ is the thermal diffusivity, i.e. the thermal conductivity divided by the specific weight and the specific heat). Since for both durations the absorpt ion depth is larger than the thermal depth, we infer that an absorpt ion model can be used and the conduct ion can be neglected. Further, since the thickness of the sample is ten times larger than the absorpt ion depth, it can be assumed as an infinite half-space. In this case, the increase in tempera ture at the surface of the sample is given by the relation [5]:

6 T( t )= (1 - R ) t a r / p C p

where I is the laser power density, R is the reflection factor, a is the absorpt ion coefficient, r is the irradia- tion time, and Cp is the specific heat of the material. Our aim was to calculate the laser power density needed to reach the melting tempera ture at the surface of the material. T h e thermal and optical parameters are collected in Table 1.

For an irradiation t ime of 0.3 ms, the calculated power density is I = 49 kW c m - 2 with a corresponding fluence F = 15 J cm 2. For a scanning speed of 8 m s - ( r = 30/zs) , the power density is I = 490 kw cm -2 and the fluence is then F = 15 J cm 2

Fig. 3. (a) Grooves obtained under scanned irradiation; (b) groove obtained with scanned irradiation, fiuence at the centre of the spot Fmax = 16 J cm -2 (spot diameter 240 /zm, r= 44/~s).

TABLE 1. Thermal and optical parameters of CMT

(I

K

Tmelting p k Cp R

100 cm-1 at room temperature for A = 10.6/~m [4] 0.01 cm z s-I at room temperature [6] 707 °C for x = 0.204 [6] 7.95 g cm- 3 at room temperature" 0.015 W cm- ~ K- 1 at room temperature ~' 0.18 J g -l K -1 [7] 0.30 at room temperature and 10.6 #mb

aColorado Research Laboratory, PO Box 692, Walsenburgh, CO 81089, USA. ~'Measured in our laboratory.

78 T. Vogt et aL / Laser irradiation of Cdo zt~4Hgo 7~ Te

TABLE 2. Comparison of experimental and theoretical thresholds according to the models

Threshold power density (kW cm- 2) and fluence (J cm- 2) Irradiation time and spot Experimental v a l u e Monodimensional model Bidimensional model diameter

0.3 ms, 225/zm 43-127 49 62 13-38 15 19

30/zs, 240/~m 360 490 584 11 15 17.5

3.2. Two-dimensional combined absorption and conduction model

For an irradiation time r = 0.3 ms, the thermal depth is only three times smaller than the absorption depth. Therefore, it is likely that the thermal conduction has some effect on the results. Furthermore, the radius of the spot is comparable with the thermal depth, hence the thermal radial conduction may also interfere. Therefore, a two-dimensional model developed at ISL [3] was used to calculate the thresholds for the two con- ditions.

The laser spot diameters are 225/zm for r = 0.3 ms, and 240/zm for r = 30/zs. Further we assume that the optical and thermal parameters remain constant during the process. The model then calculates the power density for which the surface temperature reaches the melting temperature for a predetermined duration.

For r = 0.3 ms, the calculated power density is I-- 62 kW cm-2 corresponding to a fluence F = 19 J cm-2, whereas for r = 30/~s the power density is I = 584 kW cm - 2 and the fluence F = 17.5 J cm- 2.

The experimental and theoretical thresholds are listed in Table 2 and compared in Section 5.

4. X-ray microanalysis

It has been reported previously that some mercury losses can occur under laser irradiation of CMT [1, 8]. This is the reason why the damage resulting from irradiation was also investigated by X-ray micro- analysis. Our samples had an initial value x = 0.204 + 0.01. From this value, we derive the follow- ing atomic proportions of the constituent elements: Hg 39.8%, Cd 10.2%, Te 50.0%. Our preliminary analysis indeed shows that significant mercury losses occur under laser irradiation.

4.1. Pulsed irradiation We analysed a crater obtained with a maximum

fluence at the centre of the spot Fmax=38 J cm -2, which is slightly above the threshold value (crater shown in Fig. l(a)). At the centre of the crater, the

40

35

30

25

%Hg 20

15

10

5

0

10 20 30 40 50 60 70 80

Ruence (J/¢rn2l

Fig. 4. Variation of mercury content (at.%) as a function of fluence.

mercury content decreased to 14.1% while at the edge of the crater, it was only reduced to 29.3%. On a crater obtained with a maximum fluence of 70 J cm-2, (Fig. l(b)) the mercury content was 13.1% at the centre and 20.5% at the edge.

4.2. Continuous irradiation The best shaped groove obtained with a maximum

fluence of 16 J cm -2 (slightly above the threshold value) was analysed (Fig. 3(b)). The mercury content decreased to 16.4% at the centre of the groove, and to 20.7% at its edge.

This first series of experiments shows that a very large amount of mercury is already outgased for a fluence of 16 J cm-2. Moreover, the mercury losses do not seem to increase further for higher fluences (Fig. 4). In future, it would be interesting to determine the mercury losses at fluences smaller than the melting threshold values.

5. Discussion

The experiments reveal that the melting threshold fluence is smaller than both theoretical estimations. Perhaps other parameters should be taken into account in the assumptions, such as possible variation of the absorption coefficient of the material during heating or

T. Vogt et al. / Laser irradiation of Cdo_,~4Hgo 7,~ Te 79

possible modification of its thermal and optical charac- teristics due to the mercury losses during irradiation.

6. Conclusion

The minimum fluence required to obtain surface melting of C MT under high power laser irradiation was determined for two experimental conditions. The experimental values are lower than the theoretical estimations. The reasons for these differences could be possible variations of the properties of the material during laser illumination. X-ray microanalysis revealed the expected mercury losses induced by laser irradia- tion. They are very important and further work will analyse the composit ion of the material submitted to fluences lower than the melting threshold value.

References

1 G. Bahir and R. Kalish, Appl. Phys. Lett., 39 ( 1981 ) 730-732. 2 J. L. Schmitt and F. L. Stelzer, J. Appl. Phys., 40 (1969)

4865-4869. 3 B. Gautier, Rapport ISL S-R 908/84 (1984) (ISL, BP34

F-68301 Saint-Louis, France). 4 J.A. Mroczkowski and D. A. Nelson, J. Appl. Phys., 54 (1983)

2041-2051. 5 H. S. Carlslaw and J. C. Jaeger, Conduction of Heat in Solids',

Oxford University Press, London, 1959. 6 L R. Holland, in J. Brice and E Capper (eds.), EMIS Data-

review RN 15633, EMIS Datareviews series 3, October 1986, pp. 18-19.

7 J. C. Brice, in J. Brice and P. Capper (eds.), EMIS Datareview RN 15620, EMIS Datareviews series 3, January 1986, pp. 13-15.

8 A. A. Zaginei, B. K. Kotlyarchuk, G. V. Plyatsko and V. G. Savitskii, Inorganic Mater., 25 (1989) 932-935.