Towards the end of July, Berlin was the meeting place for the by far largest International Conference on the Physics of Semiconductors to discuss the most recent progress in the field. Seven years after breaking down "the wall" scientists from east and west met for five days at the Technical University of Berlin. A record number of 1873 abstracts was received, 33% more than in Vancouver (1994). They split into "1. Bulk and Dynamics" 20% (13% in 1994), "2. Growth, Surface and Interface" 13% (12%), "3. Heterostructures, QW and SL" 31% (35%), "4. Small Quantum Systems" 15% (15%), "5. Defects and Impurities", 14% (10%), and "6. New Materials and Devices" 8% (15%). The accepted 43% of the papers were presented in 196 oral presentations and 608 posters. 1291 Participants were registered, 505 from Germany, 119 from Japan, 102 from USA, 94 from UK, 72 from Russia, and 49 from France. The total number of participants from the former Soviet Union countries was 104 (!). A total of about 47 nitride related papers was presented.
The titles and authors of the eight plenary talks given were K. v. Klitzing "Latest developments in quantum Hall effect", Sh. Nakamura "III-V nitride based blue/green LEDs and LDs", N.N. Ledentsov "Ordered arrays of quantum dots", M.A. Kastner "mesoscopic physics with artificial atoms", A. Heeger "Plastic optoelectronic devices", L. Eaves "Magnetotunnelling into quantum wells, wires and chaotic stadia", Phaedon Avouris "Atomic and nanometer scale modification of semiconductor surfaces", and T. Ando "Mesoscopic transportin low-dimensional systems".
In his Monday morning talk Nakamura presented a historical overview of the development of the nitrides and how the LEDs and laser diodes developed. The data cited for LEDs are: blue 450nm, 4mWopt @ 20mA, external efficiency 7.3%; green 525nm 3mWopt @ 20mA, external efficiency 6.3%, 12cd, and yellow emission at 600nm was noted. Laser diode structures were presented using a InxGa1-xN/InyGa1-yN superlattice with 30Å well width for the active layer. The In concentration for green emission was x=0.45, for blue emission x=0.20. In both cases y=0.05. Cladding layers were provided by AlzGa1-zN, z=0.15. Nakamura demonstrated a pulsed laser diode (T=300K) in the near UV and produced the colors red, green, and blue on fluorescent cardboards. Typical threshold data were given by Ith=4-10kA/cm2, Uth=20-30V. Nakamura then focused on the optical transition involved in the emission of InGaN Laser diodes. For various In concentrations a Stokes shift between optical absorption and emission of 100 up to 600meV was observed. At a threshold carrier concentration of 1.3x1019cm-3 a carrier lifetime of 3ns and a luminescence lifetime of the 2-4ns for the transition in the InGaN was reported. From this long lifetime, in contrast to the excitons in GaN, and the strong Stokes shift Nakamura concluded that the transition involved in the stimulated emission were within localized states of the InGaN, possibly caused by the fluctuations of the alloy. During the discussion, the origin of the large threshold voltage drop was attributed to the high resistivity of the p-contact, made out of AuNi.
In a symposium on wide-gap materials and devices Tuesday afternoon chaired by A.V. Nurmikko two nitride related talks by M. Asif Khan and J.M. Baranovski followed two talks of Akira Ishibashi and A. Waag reporting on the progress of II-VI based systems.
Asif Khan reported on AlGaN/GaN heterostructure transistors taking advantage of modulation doping in the AlGaN achieving high electron concentrations in the GaN layer. Solar blind detectors were presented on AlGaN with various Al concentration allowing a tuning of the cut-off energy at present up to 250nm, 4.9eV.
J.M Baranovski presented the homoepitaxy of MOCVD GaN on small bulk GaN crystals from high pressure synthesis. Growth was better on the smooth side of the bulk crystals which was assumed to be nitrogen terminated. X-ray data indicated a remaining mismatch of the c-lattice constant of only 5x10-4. Low temperature photoluminescence showed donor and acceptor bound excitons and the A- and B- free excitons. Fine structure around 3.40eV was interpreted due to discretely separated donor acceptor pairs.
On Friday morning a session was devoted to Defects in GaN and chaired by J. Schneider. J. Neugebauer presented first principle total energy calculations of electronic levels and formation energies of all native defects and donor and acceptor impurities such as Si and Mg. Due to the large bandgap, native defects and clusters thereof are found to be compensating defects only. A passivation by H is found to be supportive for the incorporation of Mg.
W.E. Carlos reported on magnetic resonance studies of recombination processes in LEDs. Two resonances at g=2.00 and 2.02 are found. The signal at 2.02 was seen in both electrical transport and in optical recombination and therefore was attributed to a deep donor in the active region. The 2.00 signal occured in the electrical measurment only and was assigned to a defect in the depletion region of the contacts.
A. Wysmolek reported on Mg-doped p-type homoepitaxial GaN layers. From the temperature dependence of the Mg-related donor-acceptor transition in photoluminescence, an optically determined thermal ionization energy of the Mg acceptor was derived to be ~254meV. The participating residual donor was found to be as deep as 58meV.
Y.G. Shreter discussed dislocation related luminescence in covalent crystals and compared with ionic GaN. Spatially resolved cathodoluminescence contributions in the range of 3.3 to 3.4eV were associated with dislocation related transitions. His model description made use of a free but immobile hole and electrons trapped in the strain field near dislocations.
C. Wetzel presented a study of the dominant donor by Raman scattering. An electron freeze-out at large hydrostatic pressure was observed in O doped samples, but not in Si doped ones. Therefore Si should be a good hydrogenic donor but O induces a deep gap state at large pressure or higher Al concentration in AlGaN similar to a DX behavior.
M. Ramsteiner presented internal shallow donor transitions in electronic Raman scattering. Scattering efficiencies are especially strong when exciting in resonance with the yellow luminescence which indicates the involvement of this shallow donor in the luminescence band. An extrinsic dominant donor such as oxygen was proposed.
In the very lively three poster sessions, nitride topics such as exciton and phonon dynamics, strain effects, and defect related absorption, luminescence, magnetical resonance, and level calculations were presented. Crystal growth and morphology were both addressed experimentally and in theory.
The table of contents of the proceedings is available at http://www.rz-berlin.mpg.de/th/icps23/CONTENT.html
Christian Wetzel C_Wetzel@LBL.GOV