BAND STRUCTURE ENGINEERING OF SEMICONDUCTOR LASERS

HAL Id: jpa-00226763

https://hal.archives-ouvertes.fr/jpa-00226763

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.

BAND STRUCTURE ENGINEERING OF SEMICONDUCTOR LASERS

E. Yablonovitch, E. Kane

To cite this version:

E. Yablonovitch, E. Kane. BAND STRUCTURE ENGINEERING OF SEMICONDUCTOR LASERS.

Journal de Physique Colloques, 1987, 48 (C5), pp.C5-275-C5-276. �10.1051/jphyscol:1987559�. �jpa-

00226763�

JOURNAL DE PHYSIQUE

Colloque C5, suppl6ment au noll, Tome 48, novembre 1987

BAND STRUCTURE ENGINEERING OF SEMICONDUCTOR LASERS

E. YABLONOVITCH and E.O. KANE

Bell Communications Research, Navesink Research Center, Red Bank, NJ 07701-7020, U.S.A.

There is tremendous room for improvement in semiconductor lasers. In the spirit of band structure engineering, we will discuss those ideas which could lower the threshold current density to

5 10 ~ m p s / c m ~ and significantly improve the efficiency.

T h e band structure of today's semiconductor lascrs is far from ideal. There is a serious asymmetry between the very light conduction band m a s and the very heavy valence band mass. (Ideally both masses should be as light as possible.) Under laser threshold conditions, the hole occupation remains classical even while the electrons are degenerate. This results in a significant penalty in terms of threshold current density, carrier injection level, and excess Auger and other nonradiative recombination This is only one of a series of problems which can be addressed by band structure engineering.

Unwanted electron-hole recombination is the main enemy of all minority carrier devices including semiconductor lasers. Such recombination manifests itself as:

a Defect recombination in the bulk or at heterojunction interfaces.

b. Spontaneous electron-hole radiative recombination c. Auger recombination due to the high carrier density.

We have found in a survey of all the major growth methods, that the defect recombination rate in modern 111-V epitaxy is negligible, with lifetimes > 1 psec being commonplace. Therefore band structure engineering may safely concentrate on such intrinsic properties as spontaneous emission and Auger recombination, while trusting that the material quality will be adequate.

Auger recombination is proportional to carrier density cubed Therefore one of our main goals is to reduce the threshold carrier density. Due to the heavy valence band mass, the carrier density required to satisfy the Bernard-Duraffourg condition is higher than would otherwise be the case.

Proposals have been made12 to employ a combination of strain and quantum confinement to reduce the valence band effective mass and to lessen the laser threshold requirements. If the sign of the strain is such as to put compression on the quantum well active layer, then the effcct on the topmost valence sub-band will be to make the effective mass heavy in the z-direction but light in the x-y plane. In this way the Bernard-Duraffourg condition may be satisfied at the lowest possible carrier density.

Spontaneous radiative recombination presents a more tricky issue. According to Einstein, any reduction in spontaneous emission must of necessity be accompanied by a corresponding reduction in stimulated gain Therefore we must bc carcful to recognizc that emission into the laser waveguide is unavoidable, but that there are approaches3" to inhibit spontaneous emission into modcs cxtcrnal to the waveguide.

T h e basic idea behind our approach to band structure engineering is illustrated in Figures 1 and 2 at the bottom of this page. In Fig. l(a) we compare the actual situation existing in bulk 111-V semiconductors with the situation I(b) we desire to achieve by a combination of strain and quantum confinement to artificially modify the band structure. The result is that the iso-energy contour at the top of the valence band will be shaped as in Figure 2

Article published online by EDP Sciences and available at http://dx.doi.org/10.1051/jphyscol:1987559

C5-276 JOURNAL

DE

T h e specific numerical calculations focused on the InxGaL-,As material system since its emission wavelength is approximately optimal for optical communications, and the moie fraction x could be adjusted to compensate for the strain and quantum confinement induced shifts in the band gap. The calculations were made in the effective mass approximation and incorporated the actual valence band off- sets for both InP and InAlAs barrier layers. We found that it was very beneficial to maintain the highest passible potential barrier in the valence band in order to achieve the lowest effective mass and this was most readily achieved with InP barrier layers.

We will present work on both theoretical modelling and preliminary experimental results on optically pumped semiconductor quantum well lasers incorporating these concepts.

1. "Reduction of Lasing Threshold Current Density by the Lowering of Valence Band Effective Mass", E. Yablonovitch and E. 0. Kane, Journal of Lightwave Technology, vol. LT-4, 504 (1986).

Erratum LT-4, 961 (1986).

2. "Band Structure Engineering for Low Threshold High Efficiency Semiconductor Lasers", A. R Adains, Electronics Lett. vol. 22, 249 (1986).

3. "Inhibited Spontaneous Emission", D. Kleppner, Phys. Rev. Lett. 47, 233 (1981).

4. "Inhibited Spontaneous Emission in Three Dimensionally Modulated, Periodic Dielectric Structures", E. Yablonovitch abstract submitted to the MSS-111 (1987 Modulated Semiconductor Structures Conference).

la1 l b l

Fig I la) A n ard~nary Ill-V cemiconductorar galn Ihreshold In >hlch !he

~ . o n d u c l ~ o n band crrrlcr Jenblty I\ de$enerare and the valence band car- r w r dcn\lty I\ nondegcner~te due to the elYkcttvc ma\\ a*>mn,ttp Ib, A n tdcallred bcmtconducror with equal eflective ma\hex that .irrtr.c\ at galn threshold wtlh a lower carrier lnjecrlon dens~ty than care l a ) .

Fly 2 The I\o-enurgy Lontc,ur 01 thc hrghc,~ lygng \ ~ l e n c e band 01 ,I \ ~ . t r l l -

conductor whjccr to compre*,ctc ,lrain in the r and t dlrect:on\ I n [he : dtrectaon. the ma\\ I \ the heavy hole mas,. I n the r and ! dlrect~i,n\.

!he maas I S much l e \ *