MULTIPLE CHARACTERIZATION OF STRUCTURAL PERFECTION IN GaAs/AlAs SUPERLATTICE

HAL Id: jpa-00226677

https://hal.archives-ouvertes.fr/jpa-00226677

Submitted on 1 Jan 1987

HAL is a multi-disciplinary open access archive for the deposit and dissemination of sci- entific research documents, whether they are pub- lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

MULTIPLE CHARACTERIZATION OF STRUCTURAL PERFECTION IN GaAs/AlAs

SUPERLATTICE

I. Umebu, S. Komiya, T. Nakamura, S.-I. Mutoh, A. Iida

To cite this version:

I. Umebu, S. Komiya, T. Nakamura, S.-I. Mutoh, A. Iida. MULTIPLE CHARACTERIZATION OF

STRUCTURAL PERFECTION IN GaAs/AlAs SUPERLATTICE. Journal de Physique Colloques,

1987, 48 (C5), pp.C5-41-C5-44. �10.1051/jphyscol:1987506�. �jpa-00226677�

JOURNAL DE PHYSIQUE

Colloque

suppl6men-t au nOll, Tome 48, novembre 1987

MULTIPLE CHARACTERIZATION OF STRUCTURAL PERFECTION IN GaAs/AlAs SUPERLATTICE

I. UMEBU, S. KOMIYA, T. NAKAMURA, S.-I. MUTOH and A. IIDA*

Fujitsu Limited, 10-1 Morinosato-Wakamiya Atsugi Kanagawa-ken 243-01, Japan

* ~ a t i o n a l Laboratory for High Energy Physics, Oho-machi, Tsukuba-gun Ibaraki-ken 305, Japan

Abstract

A small angle X-ray diffraction using synchrotron radiation, microscopic Raman spectroscopy and

high-resolution

were used to characterize superlattices (SLs) grown at 500, 600, and 700°C. The periodicity of the SLs was not degraded, but transition-layer thickness increased with the growth temperature. The transition layer was revealed to be atomic-layer steps with a monolayer per length ,100

for the 500°C-, and with one or two monolayers per length less than 30 A for the 700°C-grown SLs.

1. Introduction

Recently much attention has been paid to extremely thin multilayers for developing quantum effect devices (e.g. Re2onant Hot Electron Transistors

new materials (e.g. ordered alloys

and so forth. In these materials, atomic-scale structural perfection, especially at heterointerfaces, is crucial because each layer consists of only a few monolayers.

The heteroinferfaces have been chargcterized in many ways,5e.g. X- ray diffraction , photoluminescenge , Raman scattering , and transmission electron microscopy (TEM) , but only independently.

In this paper we study growth-temperature dependence of structure of ~aAs/AlAs superlattices (SLs), periodicity and transition layer at the heterointerfaces, by small angle X-ray diffraction, Raman scattering, and high-resolution TEM. The first two methods give the quality of the wafer as a whole nondestructively and quickly, but it is impossible to gain an insight on an atomic scale on what is happening within the wafer. TEM is a powerful tool to get information on atomic scale, but its field of view is very narrow. So there is always a fear in applying the observed results to the whole area of the wafer. We threfore consider characterization by multiple means is the only way to get a true picture of structural perfection in superlattlces.

2. Experiments and Results samples

~ a A s / ~ l A s SLs with 9-by-9 atomic layers were grown for 60 periods on (001 )-oriented GaAs substrates by conventional molecular beam epitaxy (MBE) at substrate temperatures of 500, 600, and 700°C.

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:1987506

C5-42 JOURNAL DE PHYSIQUE

X-ray small angle diffraction

A 2.5 GeV synchrotron radiation is used as an X-ray source. The beam divergence is about 0.03 mradian and the shape of the beam is 0.5 mm

5 mm on the sample. The wavelength is 0.961 A so the Bragg angle is around 0.55 degree for those SLs.

Figure 1 shows the X-ray reflectivity vs. the incident angle. The reflectivity of the 500°C-grown SL is 11% and the full width of half maximum (FWHM) is 37 msec. For the 600°C and 700°C-grown SLs, the reflectivity was only slightly above the background level. For the 500°C-grown SL, the Ga fluorescent X-ray excited by the standing wave in the SL was detected around the Bragg reflection angle, which suggests the high periodicity of this SL.

Raman Scattering

Measurements were made at room temperature by using Ar ion laser with a wavelength of 5145 A and a spot size of 100 pm on the samples.

Figure 2 shows the growth temperature dependence of longitudinal acoustic (LA)-yhonon lines. The first order phonon frequencies are 29.3 and 29.4 cm for 500°C- and 600°C-grownlSLs. They agree very well with the calculated frequencies 29.44 cm- for an SL with an intendedl thickness of 51 A. For the 700°C-grown SL, line frequency shifted 2 cm higher, meaning that the period

shorter than the intended one.

In all the samples, the phonon lines were observed up to the sixth order.

Figure 3 shows the longitudinal optical (L01 phonon spectra. They are the localized L0 phonons in the GaAs layers, found by the polarization measurement. The L0 phonon frequencies were read from this graph and the thicknesses of the transition layer were obtained by using a relation given in Ref. 7. They were 0 (abrupt) for 500°C-, 2.2 A for 600°C-, and 5.3 A for 700°C-grown SLs.

High resolution transmission electron microscopy

The Transmission electron microscope used has an acceleration voltage of 200 kV and a point resolution of 1.8 & (AKASHI 002B).

Figures 4 and 5 show lattice images of (110) cross section for 500°C- and 700°C-grown SLs. The AlAs layers look brigheter than the GaAs layers. One white dot corresponds to one molecule of GaAs or AlAs due to a shortage of resolution(the projected distance between Ga and As or A1 and As is 1.4 A while the resolution of the TEM is 1.8 A). A series of brighter spots at the interfaces belongs not to the GaAs layer but to the AlAs layer, and appears only on one side of the two AlAsIGaAs and ~ a ~ s / ~ l A s interfaces. This asymmetry comes from the asymmetric alignment of A1 and Ga atoms at two interfaces, t'hat is shown by our computer simulation.

The period is found to be

monolayers for both SLs by counting the number of bright dots though Raman scattering measurement gives 17 monolayers (48 A) for the 700°C-grown SL .

3. Discussion

Even the highest reflectivity of X-ray is only a half of that of an ideal SL as shown in Fig.

To investigate the causes, we calculated reflectivity for t$ree cases assuming SL to be a multilayer consisting of continuum media with 1) periods distributed around the average following the binomial distribution b y f l , or f 2 and

monolayers, 2) roughgess at the heterointerfaces which follows the Gaussian distribution , and 3) GaAlAs mixed-crystal layers at the heterointerfaces (where the period is kept constant).

Figures 6 and 7 show the calculated reflectivity for case 1 and 2.

In case

the reflectivity has dropped sharply and the FWHM increased

from that of the ideal SL. In case 2, the reflectivity has dropped to a

half of the ideal case for a standard deviation ~ = 0 . 5 monolayer and to

less than a few percents f o r r > l monolayer where the FWHMs have barely

changed. In case 3, unrealistically thick mixed-crystal layer was

necessary to reduce the reflectivity to less than a half of that of the ideal SL.

As a consequence, roughness at the interfaces is considered to be the most probable cause of the reduced reflectivity: transition layer thickness ( = 2 4 ) is about one monolayers (=2.8 A) for the 500°C-grown SL and over two monolayers (>5.6 A) for the 600°C and 700°C-grown SLs.

They are bigger than those obtained by Raman scattering, especially for the 500°C-grown SL.

We made a computer simulation of high-resoluti?~ transmission electron microscopy by Cowley-Moodie multi-slice method using Fujitsu super computer FACOM VP-200 and VP-400. A monolayer step made a clear contrast at the heterointerface of ~aAs/AlAs SL. We estimate there are atomic steps on the interfaces where the bright lines show irregularity. The estimated height and density of atomic-layer steps are a monolayer (2.8 A) per length longer than 100 A for the 50O0C- grown SL, and one or two monolayers(5.6 A) per length less than 30 A for the 700°C-grown SL.

Among three characterization methods, the X-ray reflectivity measurement gives the severest mark. One reason for this may be that the X-ray sees a large area at a time and that the uniformity of SLs is not enough if viewd on a millimeter range. A basis for that estimation is that the reflection electron microscopy for MBE-grown GaAs shows high-density small steps on undulations on the surface.

Combining X-ray diffraction and TEM observation results, sparcely- distributed one-monolayer transition layer exists at the heterointerfaces of 500°C-grown SL, though Raman scattering has not shown any sensitivity to such irregularities under our measurement conditions. At

700°C growth temperature, the transition layer thickness is about two monolayers as the three characterization results coincide.

4. Conclusion

The GaAs/~lAs SLs grown at 500, 600 and 700°C were characterized by X-ray reflection, Raman scattering, and TEM. The periodicity of the SLs was not degraded, but transition layer thickness increased with the growth temperature. X-ray reflection and Raman scattering with a combination of theoretical calculations gave quantitative indexes for the structural perfection, -where the transition layer was revealed to be the atomic layer steps by TEM. Importance of characterization by multiple methods has been shown as each method has a different characteristic in sensitivity to the atomic steps, observation area, and resolution.

This work was performed under the management of the R

D Association for Future Electron Devices as a part of the R

D Project of Basic Technology for Future Industries sponsored by Agency of Industrial Science and Technology, MITI.

References

1. N. Yokoyama, K. Imamura, S. Muto, S. Hiyamizu, and H. Nishi, Jpn. J. Appl. Phys. 24, L853(1985).

2.

Y. Yao, Jpn.

Appl. Phys. 22, L680(1983).

3. A. Segmiiller, P. Krishna, and^. Esaki,

Appl. Cryst. 10, l(1977).

4. L. Goldstein, Y. Horikoshi, S. Tarucha, and H. OI~:amoto, Jpn. J. Appl Phys. 22, 1489(1983).

5. C. Colvard, T. A. Grant, M. V; Klein, R. Merlin, R. Fisher, H. Morkoc, and A. C. Gossard, Phys. Rev. g , 2080(1985).

6. Y. Suzuki and H. Okamoto, Jpn. J. Appl. Phys. 24, L696(1985).

7. B. Jusserand, F. Alexandre, D. Paquet, and G. Roux, Appl. Phys.

Lett. 47, 301 (1 985).

8. L. G. Parratt, Phys. Rev. 2, 359(1954).

9. H. Davies, Proc. Inst. Elec. Engrs. 101, 209(1954)

10.J. M. Cowley and A. F. Moodie, Acta. Cryst. 10, 609(1957).

CS-44 J O U R N A L D E PHYSIQUE

0.3

growth temp.[%]

- 500

U

Incident angle [degl

I ^{' '} A ^{' ' ~;~wthi} Temperature

-

Fig.4. Lattice image(500°C-grown S L )

I i I l I i I

4- ^Growth

Temperature

0.3

fluctuation of period [MLI

C C

7O0I0C I

.-

U) C

6OOI'Cl

C 0

E

L0

30 20

Romon shift [cm-'1

Fig.2. L A phonon spectra

,,----.

0.54 0.56 0.58 0.60 Incident angle [deg.]

Calculated reflectivity for

with

0.3

Fig.3. L 0 phonon spctra S L with roughness