Dynamic photoconductive gain effect in shallow-etched AlGaAs/GaAs quantum wires

We report on a dynamic photoconductive gain effect in quantum wires which are lithographically fabricated in an AlGaAs/GaAs quantum well via a shallow-etch technique. The effect allows resolving the one-dimensional subbands of the quantum wires as maxima in the photoresponse across the quantum wires. We interpret the results by optically induced holes in the valence band of the quantum well which shift the chemical potential of the quantum wire. The nonlinear current-voltage characteristics of the quantum wires also allow detecting the photoresponse effect of excess charge carriers in the conduction band of the quantum well. The dynamics of the photoconductive gain are limited by the recombination time of both electrons and holes.

Figure 1

(a) Experimental circuit. A lateral constriction of a 2DEG forms a quantum wire between source and drain contacts. The central area of the device is covered with an opaque gate (bright rectangle). An aperture in this top gate close to the constriction defines the position where the underlying 2DEG of the source region is optically excited. (b) and (c) The quantum wire is defined by two adjacent trenches A and B in the 2DEG via a shallow-etch technique. (d) Conductance and photoresponse data of sample A at $T_{\text{bath}}=2.3\text{K}$, $f_{\text{chop}}=117\text{Hz}$, and $V_{\text{SD}}=1.5\text{mV}$. See text for details.

Figure 2

(a) Conductance data of sample A as a function of $V_{\text{SD}}$ for gate voltages in the range of $V_G=545$ and 623 mV $\left(T_{\text{bath}}=2.1\text{K}\right)$. (b) Corresponding linear grayscale plot of the absolute transconductance data ($\text{black}=\text{zero}$ transconductance, $\text{white}=\text{high}$ absolute value). (c) Linear grayscale plot of the photoresponse of sample A as a function of $V_{\text{SD}}$ and $V_G$ at $T_{\text{bath}}=2.3\text{K}$ $\left(\text{black}=0\text{nA},\text{white}=25\text{nA}\right)$. (d) Linear grayscale plot of the photoresponse of sample B as a function of $V_{\text{SD}}$ and $V_G$ at $T_{\text{bath}}=2.2\text{K}$ $\left(\text{black}=0\text{nA},\text{white}=55\text{nA}\right)$. (e) Single photoresponse curve which corresponds to the dotted line in Fig. 2c. (f) Photoresponse curves which correspond to the white-dashed, black-dashed, and dashed-dotted lines in Fig. 2d. See text for details.

Figure 3

(a) Squares: photoresponse $\left|I_{\text{PR}}\right|$ of sample A as a function of excitation photon energy $E_{\text{photon}}$ with a maximum at $E_{\text{PR}}\sim 1.55\text{eV}$. Line graph: spectrally resolved photoluminescence (PL) of the quantum well with a maximum $E_{\text{PL}}=1.545\text{eV}$. Data taken in a microPL setup at $T_{2\text{DEG}}\sim 10\text{K}$. (b) Photoresponse $\left|I_{\text{PR}}\right|$ of sample A at the first photoresponse maximum in Fig. 1d as a function of the chopping frequency $f_{\text{chop}}$ in a logarithmic representation. The two lines are fits to exponential decay functions. See text for details. $T_{\text{bath}}=2.7\text{K}$, $P=80\text{mW}/{\text{cm}}^2$.

Figure 4

(a) Simple model to describe the photoresponse of the quantum wire: the 2DEGs in the source and drain contacts are described by Fermi distributions, while the transfer function of the quantum wire is described by a Heaviside step function. (b), (d), and (f) Calculated results of Eqs. (1, 2, 3). (c) and (e) Experimental data at $T_{\text{bath}}=2.0\text{K}$. (g) Experimental data at $T_{\text{bath}}=2.7\text{K}$. See text for details.