J. Eur. Opt. Society-Rapid Publ. 22, 39( 2026) 383
Table 2. Material parameters used in nBn InGaAs photodetector design.
Parameter |
InP |
InGaAs |
AlGaAs |
Bandgap |
1.35 eV |
0.734 eV |
1.99 eV |
Electron affinity |
4.40 eV |
4.67 eV |
3.41 eV |
Permittivity |
12.5 |
13.9 |
11.5 |
Electron SRH lifetime |
|
1 ns |
|
Hole SRH lifetime |
|
20 ns |
|
Trap density |
|
1 10 14 cm �3 |
|
Electron Auger coefficient |
3.7 10 �31 cm 6 / s |
1.8 10 �28 cm 6 / s |
5 10 �30 cm 6 / s |
Hole Auger coefficient |
8.7 10 �30 cm 6 / s |
1.8 10 �28 cm 6 / s |
1 10 �31 cm 6 / s |
Electron effective mass |
0.0795 m 0 |
0.0332 m 0 |
0.77 m 0 |
Hole effective mass |
0.6 m 0 |
0.471 m 0 |
0.734 m 0 |
Figure 1.( a) Planar design, and( b) band diagram of nBn structure.
with the core design principles of nBn photodetectors [ 19 ]. Therefore, an Al composition of 0.57 was adopted for subsequent electrical and optical simulations to ensure optimal device performance.
Following the determination of the optimal Al composition ratio for the nBn InGaAs structure, the influence of varying N-type doping concentrations within the AlGaAs barrier layer was examined Figure 3 presents the corresponding energy band diagrams under a 150 mV reverse bias for different doping levels. As simulated, increasing the N-type doping concentration leads to a higher VBO, whereas the undoped AlGaAs barrier exhibits the most favorable band alignment. Nevertheless, to ensure that the simulation results accurately reflect realistic device behavior and considering that background doping cannot be completely avoided during epitaxial growth the electrical simulations were performed assuming a maximum donor concentration of 1 10 16 cm �3 in the AlGaAs barrier. This consideration ensures a closer correspondence between the simulated and experimentally achievable device performance [ 20 ].
Basedontheoptimizedbandalignmentparameters, the device structure was defined as follows: 0.2 lm n �-In 0. 53- Ga 0. 47 As top n-contact layer( 5 10 17 cm �3), 0.1 lm n �- Al 0. 57 Ga 0. 43 As barrier layer( 1 10 16 cm �3), 5 lmn �-In 0. 53- Ga 0. 47 As absorber layer( 5 10 15 cm �3), 0.5 lm n +-InP bottom buffer n-contact layer( 5 10 18 cm �3) and 500 lm n +-InP substrate( 5 10 18 cm �3). The barrier thickness was determined according to the photocurrent versus potential curves presented in Figure 4 of Section 3. In addition, the absorber layer was designed to be relatively thick to achieve low capacitance, and lightly n-doped to further reduce the dark current [ 21 ].
3 Results and discussion
In this section, all simulations were conducted at 300 K, with the lateral device active area set to 300 lm toreflect the geometry of commercial detectors, and all current values were normalized accordingly. Figure 4a shows the simulated current as a function of applied bias for different barrier thicknesses. Increasing the barrier thickness clearly reduces the dark current, but at the same time causes a relative decrease in photocurrent. Detailed analysis revealed that a barrier thickness of 100 nm provides the optimal balance: it maintains photocurrent levels comparable to those of thinner barriers while yielding nearly an order-ofmagnitude reduction in dark current. In contrast, barrier layers thicker than 100 nm introduce a more pronounced photocurrent degradation, indicating that additional thickness imposes unnecessary transport resistance without a