Abstract

We report on the electrical and optical properties of silicon (Si)-doped InP layers grown by solid-source molecular beam epitaxy using a valved phosphorus cracker cell. Within the range of Si effusion cell temperatures investigated (900–1200 °C), the highest electron concentration obtained was 1.1×1020 cm−3. A saturation phenomenon was observed for the electron concentration at higher Si cell temperatures. 300 and 77 K Hall mobility data were used to determine the compensation ratios by comparing them with the theoretical data. Although the Hall data show that the compensation ratio increases with the increase in carrier concentration, the exact values are not certain because the theoretical calculation overestimates the mobility values at higher carrier concentrations. The saturation phenomenon of electron concentration in InP may be considered due to the Si atoms occupying both the In and P lattice sites, or Si donors located at the interstitial sites. The 300 K Hall mobility and the concentration data measured were found to fit the Hilsum expression well. The mobility values obtained in this study are better than or comparable to reported data in the past, indicating good material quality. 5 K photoluminescence (PL) measurements showed two peaks for the undoped and low doped InP layers corresponding to the neutral donor-bound exciton transitions (D0–X) and the acceptor-related transitions (D–A), respectively. When the doping level was increased, the near-band edge (D0–X) recombination peak becomes broadened and asymmetric due to changes in the donor level density of states and relaxation of the wave vector conservation rule. The full-width at half-maximum (FWHM) value of the PL peak position increased when the doping concentration was increased. An empirical equation was developed to fit this variation, which provides a convenient way of determining the dopant concentration from the experimental FWHM value. The near-band edge peak positions shifted to higher energy with the increase of doping level due to the band filling effect. This shift agreed well with the calculations based on the Burstein–Moss shift and the band gap narrowing effect considering a nonparabolic conduction band.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call