The viability of epitaxial regrowth of non-intentionally doped InP to passivate lateral mesa surfaces of InGaAs photodiodes lattice-matched to InP is investigated, evaluating whether the residual doping of the regrown layer can be responsible for an unexpected increase of the surface current. The effect of residual doping is evaluated via numerical calculations of dark current, considering the range of doping concentrations expected for non-intentionally doped InP. The calculations show that the increase in dark current due to the residual doping of the regrown InP layer is not enough to justify the observed increase in surface current. On the other hand, the technique is still valid as a passivation method if the photodetector pixel is isolated by etching only the top contact layer.

One of the key parameters determining detection properties of silicon PIN detector structures (p⁺-ν-n⁺ or n⁺-ν-p⁺) is minority carrier diffusion length in p-n junction regions p-n (p⁺-ν or n⁺-ν). The parameter concerned strongly depends on quality of the starting material and technological processes conducted and has a significant impact on detector parameters, in particular dark current intensity. Thus, the parameter must be determined in order to optimise the design and technology of detectors.

The paper presents a method for measuring the spatial distribution of effective carrier diffusion length in silicon detector structures, based on the measurement of photoelectric current of a non-polarised structure illuminated (spot diameter of 250 μm) with monochromatic radiation of two wavelengths λ₁ = 500 nm (silicon penetration depth of around 0.9 μm) and λ₂ = 900 nm (silicon penetration depth of around 33 μm). The value of diffusion length was determined by analysing the spatial distribution of optical carrier generation and values of photoelectric currents.