Influence of temperature on photoconductivity of single Se nanowires has been studied. Time response of photocurrent at both room temperature and low temperature suggests that the trap states play an important role in the photoelectrical process. Further investigations about light intensity dependence on photocurrent at different temperatures reveal that the trap states significantly affect the carrier generation and recombination. This work may be valuable for improving the device optoelectronic performances by understanding the photoelectrical properties.
Se nanowiresTrap statesPhotoconductivityTemperature effects
Semiconductor nanowires have great potential applications as the building blocks for nanoscale electronic and optoelectronic devices. Selenium (Se) is an important semiconductor, and Se nanowires have attracted enormous attentions due to their unique physical properties, such as high photoconductivity, large piezoelectric, and thermoelectric effects [1–14]. In particular, photodetectors and photoelectrical switchers based on individual Se nanowires have been fabricated [10–14]. In order to further improve the performance of the Se nanowires optoelectronic devices, well understanding of the photoelectrical properties is desirable. Measurement of the temperature dependence of photoconductivity is an efficient method to study the photoelectrical properties because it can yield more information about the carrier generation and recombination. However, the detailed temperature effects on photoconductivity of single Se nanowires are not yet reported. In this work, we study the photoconductivity of single Se nanowires by measuring the time response on photocurrent and the light intensity dependence on photocurrent at different temperatures. It is found that the trap states are sensitive to the temperature.
The Se nanowires were grown by solution method as described previously . Briefly, solid Se spheres were first fabricated by the dismutation of Na2SeSO3 solution. And then, the Se spheres were dispersed in ethanol to finally form the solid Se nanowires. Field-emission scanning electron microscopy (SEM, FEI DB 235) was used to image the Se nanowires. The SEM image shown in Fig. 1a reveals the nanowires having lengths of about several micrometers with dendritic structures. The Se nanowires were characterized through Raman spectroscopy and photoluminescence (PL) spectroscopy using a Renishaw inVia Raman-PL microscope with a 514 nm laser excitation. Figure 1b shows the Raman spectra taken on the Se nanowires. The sole peak is centered at 238 cm−1 suggesting that the selenium nanowires are with the trigonal phase . A typical PL spectrum of the Se nanowires is shown in Fig. 1c. The PL spectrum is dominated by a peak centered at 706 nm corresponding to 1.76 eV, which is well consistent with the band-gap of trigonal-Se .
In order to perform the measurement of photoelectrical properties of individual Se nanowires, the Se nanowires were mechanically transferred onto the SiO2(500 nm)/Si substrate. Ti/Au (10/70 nm) electrodes contacting onto individual Se nanowires were fabricated by electron beam lithography, metal sputtering deposition, and lift-off. SEM image of the fabricated two-terminal nanowire device is shown in Fig. 1d. Conductance measurements were carried out using a Keithley 4200 Semiconductor Characterization System and the samples were placed on a Janis micro-cryostat under vacuum (~10−6 Torr). For photoconductivity measurements, illumination was provided by an Ar ion laser with 514 nm wavelength guided by a Renishaw Raman microscope.
Figure 2a shows the time course of the photocurrent measured at 300 K and with constant 0.1 V bias voltage as the laser illumination was turned on and off. The photocurrent rapidly increased upon light exposure and then reduced to a constant, as seen the sharp peaks in Fig. 2a. When illumination was removed, the current quickly jumped down to the level of initial dark current value. The photocurrent–time characterization suggest the existence of trap states in the Se nanowires. Our previous work also indicated that the surface-absorbed oxygen molecules on the Se nanowire can capture electrons and induce the p-type conductivity.  At room temperatures and dark conditions, the trap states are mostly not occupied due to the thermal activation. At the beginning of the illumination, a mass of electron–hole pairs are generated resulting in the sudden jump of current in Fig. 2a. The photo-generated electron–hole pairs upset the original carrier balance and will fill in the trap states. The filling of the trap states leads to the decay of the photocurrent in Fig. 2a. The rebalance of the carrier concentration makes the photocurrent toward saturation. Figure 2a also shows that the current is continued to increase slightly after the illumination was turned off, which is attributed to the fact that the carriers are returned slowly from the trap states with the assistant of thermal activation.
Figure 2b shows the time course of photocurrent measured at 100 K and 0.2 V bias. The photocurrent has a quick response to the laser illumination switching. No transient decay of photocurrent was observed after the device was exposed under illumination. At low temperatures and at dark, the carriers captured in the trap states are almost frozen. Under illumination, the carriers generated from the trap states and the bandgap contribute to the photoconductivity. The dynamic balance between the carrier generation and recombination results in the relatively steady photocurrent but with some noise. After turning off the illumination, the current declines quickly, and then the dark current slowly decreases with a long relaxation time, as shown in Fig. 2b. The relaxation of the dark current is due to the recapture of the carrier by the trap states.
Figure 3a shows the time response of a single Se nanowire device at different excitation intensities measured at 10 K and 0.5 V bias voltage. As the illumination intensity was varied from 0 to 2.3 × 104 mW/mm2, the photocurrent was increased from 0.03 to 22 nA. The light intensity dependences of the photocurrent at different temperatures are shown in Fig. 3b with log–log scales. It is interesting to notice the photocurrent toward the similar values at high light intensities for the different temperatures. At high illumination intensities, the number of photogenerated carrier is overwhelmingly larger than the thermal activated carriers. Therefore, the current is governed by the light intensity but not the temperature. The dependence of photocurrent (Iph) on laser intensity (P) can be well fitted by a power law, Iph ∝ Pα, where exponent α can help to reveal the dynamics of carrier generation and recombination. Fitting the power law dependence to the experimental data gives α = 0.64, 0.49, and 0.07 at temperatures of 10, 200, and 300 K, respectively. The power law dependence can be further analyzed by inspecting the variation of the density of free carriers (N) in the nanowire 
where, F is the photon absorption rate and is proportional to the illumination intensity PC is the probability of a charge to be captured, and Ntrap is the density of trapped carriers. Under steady-state conditions, (dN/ dt) = 0 and we can obtain that
Assuming the photocurrent Iph proportional to the free carrier density N and considering two extreme conditions, if Ntrap >>N, then N ∝ P and thus Iph ∝ P; but if Ntrap <<N, then N ∝ P0.5 and thus Iph ∝ P0.5. The exponent α = 1 and α = 0.5 are corresponding to the monomolecular recombination and bimolecular recombination, respectively . For our experimental results, at 10 K, α = 0.64, a signature of existence of both monomolecular recombination and bimolecular recombination. At low temperatures, the free carriers are not thermally activated, and therefore, the response is dominated by the trap states. At elevated temperature of 200 K, α = 0.49, which is characteristic of bimolecular recombination. The free carriers increase as a result of increased thermal activation, which induces the transition from monomolecular to bimolecular recombination. However, the exponent α of the photocurrent dependence on the excitation intensity changes to 0.07 at 300 K, which may be due to the fact that the trap states become recombination centers. At 300 K, the trap states are thermally activated and act as recombination centers under illumination, leading to the weak light intensity dependence of photocurrent.
In summary, the temperature effects on photoelectrical properties of single Se nanowires have been investigated. The light intensity dependence of photocurrent suggests that the Se nanowire photoconductivity is dominated by trap states at low temperatures, while a weak dependence of photocurrent on incident light intensity was observed at room temperature.
We thank Prof. Yadong Li of Tsinghua University for supplying the Se nanowires. This work was supported by NSFC (No. 10804002), MOST (Nos. 2007CB936202, 2009CB623703), and the Research Fund for the Doctoral Program of Higher Education (RFDP).
This article is distributed under the terms of the Creative Commons Attribution Noncommercial License which permits any noncommercial use, distribution, and reproduction in any medium, provided the original author(s) and source are credited.