382
J. Eur. Opt. Society-Rapid Publ. 22, 39( 2026)
Table 1. Physical models used in SILVACO ATLAS simulation for nBn InGaAs photodetector design.
Model |
Syntax |
Description |
Mobility |
CVT |
To enable transverse field, doping, and temperature-dependent parts of mobility |
Optical |
OPTR |
To enable band-to-band recombination for direct band semiconductors |
Shockley – Read – Hall |
SRH |
To enable recombination. It uses fixed minority carrier lifetimes |
AUGER recombination Auger |
To enable direct transition of three carriers |
Fermi distribution |
Fermi |
To enable carrier statistics. Suitable for highly doped regions |
Quantum tunneling |
QTUNN |
To enable quantum tunneling through the conduction and valence band barrier due to a semiconductor |
Quantum tunneling |
QTUN. BBT |
To enables the band-to-band mode of the direct quantum tunneling model |
Band to Band |
|
|
Non-local trap assisted |
TAT. |
To enables the non-local tunneling model in the calculation of the field effect |
tunneling |
NONLOCAL enhancement factors. |
Impact ionization |
IMPACT SELB |
To enables the impact ionization. Recommended for most case, includes temperature dependent parameters. |
Al 0. 57 Ga 0. 43 As barrier layer, which is hypothesized to provide near-flat valence-band alignment while maintaining sufficient conduction-band offset to block electron transport. Numerical simulations are performed using SILVACO ATLAS TCAD at 300 K, enabling accurate modeling of band profiles, carrier distributions, and current transport mechanisms. The optimized device exhibits a significant reduction in dark current compared to conventional pn InGaAs structures while preserving responsivity, indicating that proper band alignment through barrier engineering can lead to improved detectivity and thermal stability in SWIR photodetectors.
2 Simulation and methodology
A detailed computational framework was developed to investigate the electrical transport and band-alignment characteristics of the proposed nBn-InGaAs photodetector. Simulations were performed using the SILVACO ATLAS, employing the physical models listed in Table 1 and the material parameters given in Table 2 [ 14 – 17 ].
In the simulation, the CVT mobility model was employed to account for doping, temperature, and transverse field effects on carrier mobility. Optical recombination was included via the OPTR model, while SRH and Auger models were used to describe trap-assisted and high-injection recombination processes, respectively. Fermi statistics were considered to ensure accurate carrier distribution in highly doped regions. Additionally, QTUNN and QTUN. BBT models were activated to capture quantum tunneling effects, including band-to-band tunneling across the barrier layer. To further improve the accuracy of dark current analysis, the TAT. NONLOCAL model was incorporated to account for non-local trap-assisted tunneling mechanisms, and the IMPACT SELB model was included to consider impact ionization effects with temperature dependence.
Vertical device stack( from substrate up) was modeled as InP( n + contact) / InGaAs( absorber) / Al x Ga 1-x As
( barrier) / InGaAs( n � contact) shown in Figure 1. Owing to its low lattice mismatch with the InGaAs and the advancements in epitaxial growth techniques, the AlGaAs barrier has emerged as the most suitable barrier material for InGaAs-based nBn photodetectors [ 18 ]. The reduced lattice strain at the heterointerface ensures high crystalline quality and stable electronic properties, facilitating efficient carrier transport while effectively suppressing dark current.
During the design of the nBn structure, systematic simulations were conducted to assess how different Al – Ga molar fractions( Al x Ga 1-x As, with x ranging from 0.50 to 0.60) influence the electronic band alignment within the barrier layer. This compositional tuning was particularly crucial, as the barrier must balance two competing requirements: minimizing the valence band offset( VBO) to ensure unhindered hole transport, while simultaneously maintaining a sufficiently large conduction band offset( CBO) to suppress the flow of photogenerated electrons. Achieving this balance is essential for reducing dark current without compromising photocurrent extraction. The simulation results indicate that an Al composition of x = 0.57 provides the most favorable trade-off among the examined cases. At this composition, the VBO reaches its minimum value( DE v 35 meV), effectively eliminating the barrier for holes, whereas the CBO( DE c 1.26 eV) remains sufficiently high to block majority-carrier leakage. Figure 2a illustrates the equilibrium band diagrams for the tested molar fractions, clearly demonstrating how incremental changes in Al content reshape both conduction and valence band edges. This optimized band configuration is particularly advantageous for nBn photodetectors, as it ensures that photogenerated minority carriers contribute efficiently to the photocurrent while majority carriers are strongly suppressed, ultimately leading to enhanced detectivity and improved bias stability.
In Figure 2b, the 150 mV reverse bias confines Fermi level band bending within the large bandgap AlGaAs barrier, leaving the InGaAs absorption layer largely undepleted and thus effectively mitigating G-R dark current, in line