Autonomous driving technology put forward an urgent requirement for eye-safe light detection and ranging (LiDAR). The In_0.53Ga_0.47As/InP APD avalanche photodiodes (APDs) distinguished themselves with superior performances such as high internal gain, high reliability, and high sensitivity working in the band covering the eye-safe wavelength, making it one of the core devices for LiDAR system. 

The extensively used separate absorption, grading, charge and multiplication(SAGCM) structure In_0.53Ga_0.47As/InP APD presented distinct spectral response characteristics under front- and back-illumination before its multiplication layer thoroughly punched through. The insight into the interior principle is significant for the application of the In_0.53Ga_0.47As/InP APD.

In a study published in Materials Letters, a research group led by Prof. SONG Hang from the Changchun Institute of Optics, Fine Mechanics and Physics (CIOMP) of the Chinese Academy of Sciences (CAS) proposed a planar SAGCM structure bi-incident In_0.53Ga_0.47As/InP APD for the derivation of the difference in photoelectric characteristics.

The planar SAGCM structure bi-incident In_0.53Ga_0.47As/InP APD was obtained by selective-area zinc diffusion method using a Zn₃P₂/Zn/SiO₂ multilayer as the dopant. Based on the bi-incident APD, the performance comparison was conducted between front-illumination and back-illumination, which revealed the typical bandpass spectral response characteristic in back-illumination while the transition from two single band to a bandpass spectral response characteristic in front-illumination.

Based on the findings, it is concluded that the different performances in the electrical and spectral aspects between the two approaches originated from the PIN junction in the InP cap layer with high electric field, which acted as short-wavelength photoelectric conversion before the InP multiplication layer punching though.

The research on the performance difference of the planar SAGCM structure In_0.53Ga_0.47As/InP APD operating in different approaches is helpful to guide its practical applications, such as in LiDAR system.