Effect of Lattice Mismatch on the Decay of Rheed Oscillations During Growth of Strained InGaAs/GaAs Heterostructures
- PDF / 856,237 Bytes
- 6 Pages / 415.8 x 637.2 pts Page_size
- 84 Downloads / 176 Views
225 Mat. Res. Soc. Symp. Proc. Vol. 618 © 2000 Materials Research Society
EXPERIMENTAL TECHNIQUE The GaAs substrates used for growth were nominally (001) oriented and Zn-doped. Prior to growth, an in-situ Ar+ ion bombardment was performed to remove the damaged surface layer using 1-kV acceleration voltage and 30-mA current. Then, an undoped GaAs buffer of about 1 ptm thickness was grown to improve the surface morphology, followed by the undoped InGalixAs layers with different compositions. The growth process was performed under Asrich conditions. The surface showed a (2 x 4) reconstruction throughout the whole growth process. The RHEED pattern remained streaky in the whole composition range we studied, indicating that no islanding occurs. A series of samples have been grown at the deposition temperature of T = 490 'C in the composition range of x ; 0-0.4, and some other samples for comparison at lower temperature, 450 'C. The growth rate, the surface reconstruction and the In composition were studied by a
computer-controlled RHEED system. The intensity oscillations of the RHEED specular beam along the [011] azimuth were recorded by the direct measurement of the electron current using a high angular resolution Faraday cup [8]. This solution has the advantage over the more commonly used fluorescent screen and camera setup of high sensitivity, high linearity, low background and low noise. The electron energy of the RHEED gun was 10 keV, the incidence angle was about 1.50. The variation of the In composition was accomplished by increasing the temperature of the In source while the other parameters remained unchanged. The composition was determined from the growth rate as obtained from the period length of the RHEED intensity oscillations [8], using the GaAs growth rate as a reference, utilizing that the sticking coefficient of both Ga and In is unity under As-rich conditions. The growth rate as a function of composition is thus r(x) = 0.3206(1 + 0.443x) ML/s (monolayer/s) for T-490 °C and r(x) = 0.1659(1 + 0.2535x) ML/s for T = 450 °C. More details about the growth process are described in Ref. [8]. RESULTS AND DISCUSSION The RHEED intensity envelope, taken as a peak-to-peak value between the oscillation period maximum and minimum, 30O 0 T= 4Woc could be fitted by a exponential function, except for the very first 0 T= 4500C S00 period. The time constant of the 2D .. decay determined from the envelopes versus composition is 15 '6ooo plotted in Fig. 1. "10" The decay of the oscillatior ....... of the intensity of the RHEED specular beam is well known in O ........... S0 MBE and occurs even during The growth. .3 .4 homoepitaxial 0.1 6.2 0.0 explanation is the growth front In corrmition roughening: the growth proceeds not in a strictly layer-by-layer In of a function as Fig. 1. Decay time constant manner, the growth of the next composition. The dotted line is a guide to the eye
226
monolayer starts before the previous one is completed. A semi-quantitative model of this process can be formulated as follows. Under usua
Data Loading...
