Electroluminescence in Unipolar-Doped ${\mathrm{In}}_{0.53}{\mathrm{Ga}}_{0.47}\mathrm{As}/\mathrm{Al}\mathrm{As}$ Resonant-Tunneling Diodes: A Competition between Interband Tunneling and Impact Ionization

We measure and analyze the light emission from a room-temperature n-type unipolar-doped ${\mathrm{In}}_{0.53}{\mathrm{Ga}}_{0.47}\mathrm{As}/\mathrm{Al}\mathrm{As}$ double-barrier resonant-tunneling diode (RTD) that occurs just above the ${\mathrm{In}}_{0.53}{\mathrm{Ga}}_{0.47}\mathrm{As}$ band edge and peaks around 1631 nm. The emission is attributed to electron-hole recombination emission made possible by holes generated in the high-field region on the collector side of the device by interband tunneling and impact ionization, which contribute comparable hole densities, according to our analysis. Although the external quantum efficiency (EQE) in our experimental configuration is rather low (≈2 × ${10}^{-5}$ at 3.0-V bias), limited by suboptimal output coupling, the internal quantum efficiency (IQE) is much higher (≈6% at 3.0-V bias), as derived from the experimental EQE and a radiometric analysis. To check this value and better understand the transport physics, we also carry out an independent estimate of the IQE using a combined interband-tunneling impact-ionization transport model, which yields an IQE of 10% at 3.0-V bias. The satisfactory agreement of theory with experimental data suggests that a RTD designed for better hole transport and superior optical coupling could become a useful light-emitting device, while retaining the intrinsic functionality of high-speed negative differential resistance, and all without the need for resistive p-type doping.

Figure 1

(a) MBE-growth stack. (b) Microphotograph of fabricated device with vertical emission aperture. (c) Experimental setup for measurement of electrical and optical characteristics.

Figure 2

Experimental I-V curve (left vertical axis) and L-V curve (right vertical axis) at 300 K.

Figure 3

(a) Experimental light-emission spectra at 300 K plotted versus wavelength for two bias voltages (2.8 and 3.0 V). Peak of both spectra occurs at about 1631 nm, and the full width at half maximum is about 148 nm. (b) Same light spectra as in (a) but plotted versus photon energy, including the ideal spectrum (offset for clarity and plotted in green) for ${\mathrm{In}}_{0.53}{\mathrm{Ga}}_{0.47}\mathrm{As}$ as a direct-band-gap semiconductor.

Figure 4

Methodology for estimating IQE by two separate approaches.

Figure 5

EQE is obtained directly from experiments (left vertical axis). Internal quantum efficiency is calculated using the radiometric approach of Fig. 4 (right vertical axis).

Figure 6

Energy diagram showing a qualitative model for electron-hole generation by interband tunneling, G_IT, or by impact ionization, G_II, on the collector side of the ${\mathrm{In}}_{0.53}{\mathrm{Ga}}_{0.47}\mathrm{As}/\mathrm{Al}\mathrm{As}$ RTD and radiative recombination at either the emitter side or in the quasi-neutral or n⁺ regions on the collector side.

Figure 7

Computed band bending (left vertical axis) and internal electric field (right vertical axis) for the qualitative models shown in Fig. 6 and a bias voltage of 2.9 V. Also shown is the energy range, ΔE = eΔV (“tunneling window”), over which interband tunneling can occur in the high-field region on the collector side.

Figure 8

Computed available current density for interband tunneling and impact ionization (left vertical axis) in comparison to photocurrent (right vertical axis). Interband-tunneling curve assumes U_G = 0.747 eV and m_r = 0.023 m_e. Impact-ionization curve assumes electron-ionization coefficients from Fig. 9.

Figure 9

(a) Table of expressions for electron impact-ionization coefficient over four contiguous regions of internal electric field. (b) Graphical representation of (a).

Figure 10

(a) Internal quantum efficiency computed by radiometric and charge-transport approaches, showing the individual contributions by interband tunneling and impact ionization to the charge-transport approach. (b) Magnification of the high-bias region of (a) and plotted on a linear scale.

Figure 11

Terminal electric current (from Fig. 2) and available injection efficiency (right vertical axis).

Figure 12

Transmission probability for electrons (EL), light holes (LH), and heavy holes (HH) through the ${\mathrm{In}}_{0.53}{\mathrm{Ga}}_{0.47}\mathrm{As}$ /$\mathrm{Al}\mathrm{As}$ double-barrier structure at zero bias, by a computation described in Appendix pp1.