• Aucun résultat trouvé

OPTICAL STUDY ON BAND EDGE OFFSET IN STRAINED MBE GROWN (InGa)As-GaAs AND (InGa)As-(AlGa)As QUANTUM WELLS

N/A
N/A
Protected

Academic year: 2021

Partager "OPTICAL STUDY ON BAND EDGE OFFSET IN STRAINED MBE GROWN (InGa)As-GaAs AND (InGa)As-(AlGa)As QUANTUM WELLS"

Copied!
6
0
0

Texte intégral

(1)

HAL Id: jpa-00226736

https://hal.archives-ouvertes.fr/jpa-00226736

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.

OPTICAL STUDY ON BAND EDGE OFFSET IN STRAINED MBE GROWN (InGa)As-GaAs AND

(InGa)As-(AlGa)As QUANTUM WELLS

T. Andersson, V. Kulakovski, Z.-G. Chen, A. Uddin, J. Vallin, J. Westin

To cite this version:

T. Andersson, V. Kulakovski, Z.-G. Chen, A. Uddin, J. Vallin, et al.. OPTICAL STUDY ON BAND EDGE OFFSET IN STRAINED MBE GROWN (InGa)As-GaAs AND (InGa)As- (AlGa)As QUANTUM WELLS. Journal de Physique Colloques, 1987, 48 (C5), pp.C5-163-C5-167.

�10.1051/jphyscol:1987532�. �jpa-00226736�

(2)

OPTICAL STUDY ON BAND EDGE OFFSET I N STRAINED MBE GROWN (1nGa)As-GaAs AND (1nGa)As-(A1Ga)As QUANTUM WELLS

T.G. ANDERSSON, V. KULAKOVSKI, Z.-G. CHEN, A. UDDIN, J. VALLIN and J. WESTIN

Department of Physics, Chalmers University of Technology, S-412 96 Gdteborg, Sweden

Abstract

Undoped single and multiple quantum well heterostructures of InxGal-xAs-GaAs and InxGal-,As-A1 Gal-yAs are investigated by

Y

photoluminescence and photoconductivity spectra.The observed conduction-to-valence band offset ratio across the GaAs-strained

(InGajAa intciface is AE,: AEv = 0 . 8 : 0 . 2 , and is iound to be reduced at the AlyGal-yAs-InxGal-xAs interface depending on the Al-and In-concentrations

.

The band edge offset in the (AIGa)As/GaAs system has been intensively studied for a number of years. Today several experiments1 indicate that AEcAE,, = 0.6-0.65. Recently the interest in band edge offsets has been extended to strained quantum wells (OW). The lattice mismatched (InGa)As/GaAs system is such an example . Applications of (1nGa)AslGaAs and

(InGa)Asl(AIGa)As QW and superlattices extended the magnitude of the band offsets. The built-in elastic strain in these systems is due to the 7% lattice mismatch between GaAs and InAs. High quality strained single and multiple QW structures as well as superlattices can be epitaxially grown provided the thickness of the strained layers are kept small enough to avoid generation of misfit dislocation^^.^. The critical thickness LC for dislocation generation is nearly inversely proportional to the In content and being about 200A for an ln0~15Ga0.85A~

QW in GaAs. The substitution of Ga by Al is not expected to influence the critical thickness as the AIAs-GaAs lattice mismatch is only 0.1%.

The conduction-to-valence band discontinuities (AEctlEv ) across the interface determine the depth of the quantum wells. This ratio has been r e p c e d in a few papers4- but the obtained values are not very accurate. We have used photoluminescence (PL) and photoconduclivity (PC) to investigate undoped SOW and MOW heteros:mctures in the (1nGa)As-GaAs and (1nGa)As-(AIGa)As system. The structures were grown by molecular beam epitaxy (MBE).

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:1987532

(3)

C5-164 JOURNAL DE PHYSIQUE

The PL experiments were performed at 2 and 77 K using a He-Ne laser. The luminescence was dispersed by an Jobin-Yvon H-25 grating monochmmator and detected by a photomultiplier (S-1) cooled to liquid nitrogen temperature. The PC measurements were carried out at 77 K using light from a tungsten-halogen projector lamp via the same monochromator. Electrical connection to the layers were made by alloying indium at 400 C.

In all QW:s we observed the ground state transition ( & ) in the PL measurements and three transitions in the PC measurements. As shown in Fig. 1 these are the excitonic transitions between the ground states ( & ) and the first excited states of electrons and (i, =

+

312) holes (

d

) and also a transition between the electronic ground state and the strained split of

j, = 11: 112 hole subband (

1

). The same spectral position of& lines in the PC and PL spectra

suggests the interpretation of these lines as free excitonic transitions indicating excitonic localisation effects are small enough to be neglected. Two additional lines

13

and d a r e at energies smaller than that of the exciton in GaAs as shown in fig.2. With decreasing QW thickness the

d

peak moves to the higher energies faster then the d peak and it disappears after. reaching the position of the GaAs free exciton. But 13 peak moves to higher energy slightly slower than the d peak with OW decreasiilg for ~ ~ t ~ s l d n t In concentration.

Therefore,the

P

peak can not be associated with the excited sublevels of the electron and hole states. The energy of

fi

peak Suggests that it may be connected with the exdtonic transition

involving an electron from the QW ground state and a hole from the valence subband split off by strain as proposed by ~arzin'. The In,Gal -#.s valence band splitting due to lattice mismatch indicate that no strain relaxation is present in the SQW and MQW structures for thicknesses smaller than the critical one. Using an extensive fitting procedure between calculated levels in QWs and experiments the AEcAEv in the (InGafAslGitAs system was found to be 0.83:0.17.

1.4

ENERGY (eV) G a A s (InGa)As G a A s

Fig 1. The band edge configuration in a strained F1g.2 The photoconductivity spectra of (1nGa)As-G~As

(InGa)As OW in an unstrained GaAs matrix. In

the (1nGa)As layer the valence band is split SQW samples with L=100 A and L=180 A at 77 K

into JZ= + 312 and tll2-hole bands. The

A)'

observed excitonic transitions are shown by

P

(4)

To study the (InGa)Asl(AIGa)As system a specially designed MQW structure

.

shown in Fig.3, was made. This contains one unstrained GaAs and two strained (1nGa)As layers. In addition there is one strained (1nGa)As QW (close to the critical thickness) in a GaAs layer. The thin GaAs QW was made for determination of the GaAs growth rate whereas the (1nGa)As QW in the GaAs cladding layer is for measuring the energy gap Eg of the strained (1nGa)As using the above value for AEc:AEv.

Figure 4 represents typical PL spectra of the MQW structure shown in Fig. 3. In the spectra (at 2 K) from strained layers the very strong narrow line, do dominates. Its halfwidth for the wells with width = 30- 40 A does not exceed 3 meV in the case of a GaAs matrix and 5 meV in the case of (AIGa)As (x-0.2-0.4). These are peaks above 1.7 eV (not shown) which come from the thick (AIGa)As and the GaAs buffer layer. The lines at 1.492 eV and 1.51 2 eV o b s e ~ e d at 2K in all samples, correspond to emission of neutral donor-acceptor pairs

(Do,Ao) and bound excitons (BE) in the GaAs. When the temperature increases to 77 K these two lines disappear due to thermal dissociation of the excitons.

G a A s ( I n G a I A s G a A s ( A I G a I A s O W O W ' S l a v e r m a t r i x

Figd.The confined energy levels in the, MQW samples and the observed transitions.

Fig4. Photoluminescence spectra of two MOW samples with different composition ol AlyGal -yAs and lnxGal*As recorded at 2K.

(5)

C5-166 JOURNAL DE PHYSIQUE

The lines denoted b y d a r e due to emission from the four different QW:s. Evidently the intensity of the lines decreases with energy. The strongest line.& 0, originates from the (InGa)As QW in the GaAs layer. The l i n e d % due to emission from the thin GaAs QW. The thickness of this was intentionally grown close to the width of the thinnest (1nGa)As layer in order to avoid difficulties with identification of the emission. The lines

4

and d,correspond to emission from the two (1nGa)As QW:s in the (A1Ga)As matrix.

To determine the band edge offsets we compared the measured transition energies with those calculated from a square well potential using the equations

where mZ and Mz are the z-components of electron (hole) effective mass in the QW and barrier, respectively and V is the potential depth. The details of the calculation is found in ref.

7.

The energy gap and thus,y, of the thick AlyGal-yAs cladding layer for every sample was obtained from its PC and PL sprctra recorded at 77K. The energy gap of the InGaAs was determined from the spectral position of the excitonic emission of the InxGal -,As Qw in a GaAs environment. The sum of electron and hole confinement energies for this LO =I50 A, is small and depends weakly on the ratio AEJAEg. This sum was calculated from Eq.(l) with AE,JAEg=0.83 obtained for strained (1nGa)As-unstrained GaAs interface.

The GaAs growth rate, )'(~aAs),was determined by the excitonic emission from the thin GaAs QW in an (AIGa)As matrix. The width of this QW, L' ,was obtained by fitting the calculated excitonic transition energy to the observed using Eq.(l) and the latest literature value AE,JAEg=0.64. The width,Li,of the InxGal-xAs QW:s were then obtained from the relation

j(lnx~al-,~s)= '1](~a~s)l(l-x). Therefore, we could determine the necessary

heterostructure characteristics for all QW:s and used Eq.(l) to estimate the conduction band offset at the (InGa)As/(AIGa)As interface. For a set of InxGal-xAs QW:s in an AlyGal-yAs matrix with different L (25-95A), X (0.12-0.35) and Y (0.2-0.35) appreciable disagreement was found for AEJAEg larger than 0.75 or smaller than 0.6. Therefore, the ratio AEJAEg across the unstrained AlyGal .yAs (y=0.2-0.35)-strained InxGal-,As (x=0.12-0.35) interface is limited to the region 0.6 to 0.75.

ACKNOWLECGEMUICT

The Swedish National Board for Technical Development (STU), The Swedish Natural Science Research Council (NFR) and The Royal Swedish Academy of Sciences (KVA) are acknowledged for their financial support.

(6)

1. H Heinrich. J M Langer, FrstkcJrperprobleme XXVI, 251 (1986).

2. J W Matthews, A E Bladeslee, J. Cryst. Growlh,'27,118 ( 1974),29,273(1975), 32, 265(1976).

3. G C Osbourn, R M Briefiekl, P L Gourley, Appl. Phys. Lett. 41,178(1982).

4. I J Fritz, S T Picraux, L R Dawson. T J Drummond, W 0 Laiding. N G Anderson, Appl.

Phys. Lett. 46,967 (1985 ).

5. J Y Marzin, E V K Rao, Appl. Phys. Lett. 43,560 (1983).

6. J Y Marzin, M N Charasse, 8 Sermage, Phys.Rev. B 31,8298 (1 985).

7. T G Andersson, Z G Chen, V D Kulakovskii, A Uddin, J T Vallin. " to be published in Solid State Comminications".

Références

Documents relatifs

It has to be pointed out that a relaxed, dislocation free InGaAs/GaAs mu1 tilayer taken as a whole, would have a lattice parameter in the direction parallel to the surface

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

The agreement between theory and experiment provides support for the envelope-function approximation as a suitable method to treat the electronic structure of

INVESTIGATIONS OF THE NEG- ATIVE DIFFERENTIAL CONDUCTIVITY AND CURRENT BISTABILITY IN DOUBLE BAR- RIER n+ GaAs/(AlGa)As/GaAs/(AlGa)As/n+ GaAs RESONANT TUNNELLING DEVICES US- ING

The light and heavy hole masses, the relative band alignments of GaAs and AlGaAs and the quantum well width were treated as fitting parameters.. Figure 2 shows

The dashed line The first possibility to investigate the shift of the higher displays the shift of the fundamental band subbands is the comparison of the emission

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des