【摘要】 本文研究了GaP中普遍出现的1.71eV光致发光(PL)宽带。这是一个与多个杂质能级有关的发光带,由吸收谱和光致发光激发谱(PLE)检测到在价带上方0.37—0.089eV范围内与该发光带有关的深能级。时间分辨谱测量表明。随着衰减过程的延续,谱峰移向低能方向,说明该带起源于D—A对复合。讨论了该发光带与背景Cu杂质有关的可能性及其发光机制和激发途径。更多还原

【Abstract】 A broad band peaked at 1.71eV is commonly observed in the PL spectra of n-GaP . Comparing the PL spectra of epitaxial layer with that of the substrate, both have similar behaviours within the interested region. It can be concluded that the band stems from some background impurities or defects in the substrate. The shape of the broad band (Fig. 1) indicates that the band is a superposition of several bands.The absorption spectrum is shown in Fig. 2. The relevant threshold energies ET and corresponding ionization energies of the deep centres D1-D6 are compiled in Tab.1.For the PLE measurement, the apparatus scheme and the filter characteristics used are shown in Fig.3a and 3b, respectively. The PLE spectra obtained are shown in Fig.4, the peak I (at 1.70 or 1.77eV) is distorted by the stray light transmitting filter, but the four thresholds at 1.46, 1.60, 1.68 and 1.91eV are not affected. They are coincident with those obtained from absorption spectrum. The PLE measurement reveals that several excitation processes are related to 1.71eV band.For time-resolved spectrum measurement, Ar+ laser beam (4880A) chopped at 1kHz with a mechanical chopper is used as the excitation source. The result is shown in Fig.5. The peak of the band shifts to lower energy with the evolution of the decay time. It suggests that the band stems from a D-A pair recombination and the 1.71eV component of the band falls faster than the 1.63eV component.X-ray fluorescence measurement reveals that Cu is one of the prominent contaminations in the sample.Based on the results obtained above, we construct a model for the energy levels and electron transition routes related to the 1.71eV luminescence band, which is shown in Fig.6. The D4(Ev + 0.51eV) and D2(Ev +0.71eV) are the well known "A" and "B" levels of Cu respectively . The radiative capture of electron into D4 level gives rise to the 1.65eV band. "B" level is a nonra-diative one, but the two-step excitation through it provides a excitation path for the 1.65eV and other emissions. D3(Ev+0.63eV) is similar to the 0.62eV level related to Cu reported by Wessels. Its electron capture could give rise to the 1.56eV band. D1 is probably the Ec-0.95eV donor often observed in n-GaP. It provides another excitation path similar to D2. Comparing our results with that of Ref.3, and 11, it is reasonable to believe that D6 corresponds to CuGa-2Tep complex as Suggested by Fabre for the 1.70eV band, of course, it should be CuGa-2Sp in our sample. Then, both luminensence band and energy level have little shift. The complex would be the acceptor in the D-A pair, the donor could be the dopant S. Regards to the D5, it can not be rulled out that it is involved in some radiation processes within the energy region near 1.71eV, and it may be the same as the centre reported by Krispin .In brief, our measurements have provided some evidences of Cu contaminant in GaP substrate. The 1.71eV band containes at least three components, which peaked at 1.71, 1.63 and 1.56eV respectively. Among them,both 1.71 and 1.63eV are DAP radiative recombinations. At least six levels D1-D6 were detected from absorption and PLE measurements. Four of them, 0.38, 0.51, 0.62 and 0.72eV above the vallence band edge are related to Cu. It is obvious that Cu is one of contaminants that is harmful for the GaP materials used to made green LED.更多还原