Resolving the trap levels of $\mathrm{Se}$ and ${\mathrm{Se}}_{1-x}{\mathrm{Te}}_x$ via deep-level transient spectroscopy

Benefiting from the facile fabrication and tunable optoelectronic properties, Se and ${\mathrm{Se}}_{1-x}{\mathrm{Te}}_x$ are promising candidates for optoelectronic applications, e.g., thin film solar cells and short-wavelength infrared detectors. However, the trap features of selenium based semiconductors are complicated mainly due to the low crystallinity induced dispersive nature. In particular, the underlying charge transport properties of ${\mathrm{Se}}_{1-x}{\mathrm{Te}}_x$ have not been systematically investigated, which is crucial for device optimization in real applications. In this work, we first introduce deep-level transient spectroscopy to characterize Se and Se-based semiconductors, and we also compare it with reverse-bias deep-level transient spectroscopy. It was found that the latter technique can result in much higher transient capacitance signal and better energy resolution. Based on the full analysis of the transient capacitance, we found the introduction of Te into Se can easily form two additional deep trap states around 0.55 eV and 0.75 eV, respectively, and it also explains why the device performance of ${\mathrm{Se}}_{1-x}{\mathrm{Te}}_x$ are still suffering from the large dark current and recombination losses.

Figure 1

(a) $J\text{−}V$ curves of Se-based photodiode measured in the dark and under 100 mW/${\mathrm{cm}}^2$; (b) $C\text{−}V$ curves of Se-based photodiode, recorded transient capacitance decays of Se-based photodiode at various temperature based on (c) DLTS and (d) RDLTS measurements.

Figure 2

(a) Comparison of DLTS and RDLTS spectra of Se-based photodiodes at various pulse filling widths, and the label “H” represents the hole traps. Arrhenius plots of the emission rates as a function of $1\setminus \mathrm{T}$ of (b) DLTS and (c) RDLTS measurements, and (d) the comparison of retrieved activation energy. (e) Pulse width dependent amplitude of DLTS measurement.

Figure 3

DLTS spectra of (a) ${\mathrm{Se}}_{0.9}{\mathrm{Te}}_{0.1}$ and (b) ${\mathrm{Se}}_{0.8}{\mathrm{Te}}_{0.2}$ based photodiodes. RDLTS spectra of (c) ${\mathrm{Se}}_{0.9}{\mathrm{Te}}_{0.1}$ and (d) ${\mathrm{Se}}_{0.8}{\mathrm{Te}}_{0.2}$ based photodiodes. The labels “E1,” “E2,” and “E3” represent the electron traps.

Figure 4

(a) Arrhenius plot of the emission rates of E1 extracted from DLTS as a function of $1\setminus \mathrm{T}$, (b) Laplace-RDLTS spectra of ${\mathrm{Se}}_{0.9}{\mathrm{Te}}_{0.1}$-based photodiodes, (c) Arrhenius plot of the emission rates of E2 and E3 extracted from RDLTS as a function of $1\setminus \mathrm{T}$, and (d) obtained activation energy of all traps extracted both from DLTS and RDLTS.