Research of dynamic characteristics of Si-photodiodes in the region of small currents using a computerized measuring complex

Abstract

Studying the current-voltage characteristics (IVC) of photodiodes, as well as any other device with a p–n junction, allows us to determine their quality and compliance with the stated requirements for reverse dark current. However, in the development and production of photodiodes, dark currents are usually investigated with the aim of achieving a certain level. The mechanisms of dark current generation are rarely studied in detail.

Some researchers are faced with the problem of the dynamics of small values of dark current (at the level of 10^–9 A), when the photodiode is a part of a high-resistance electrical circuit.

This study investigated the dynamic IVC of silicon photodiode samples in the zone of small currents using a newly developed, specialised, computerised system.

To carry out the research, we first tested the original computerised setup we had developed and manufactured by measuring the dynamics of dark currents of a UFD 337 type photodiode. The tests allowed establishing that the voltage changes within the range from +1.5 to –9 V (which can be expanded by using other power sources), the number of measurements of dark current values per second is up to 5000 (can be increased when using a higher-frequency controller), and the resolution when measuring dark current is 0.02 nA. During our research, we studied the behaviour of the dark currents of silicon photodiodes over a range their lower values: 10^–10–10^–7 A. The experimental results show that the investigated photodiodes have hysteresis of dynamic I–V characteristics in the range of currents of 1–10 nA at a negative voltage from 10 to 0 V.

If the diode structure is technologically perfect, then for static IVCs such hysteresis should either not exist at all or may be within the dark current measurement error, which is usually about 5%. In our case, the current drop in the hysteresis loop on the dynamic IVCs increases with the voltage change rate. On average, it is about 5 nA, which is much larger than the measurement error of 0.01 nA.

If the photodiode is intended for recording photocurrents below 1 nA (for example, low levels of illumination in photometry: 0.01 – 0.1 lux) or for operation in high-speed devices, then the detected hysteresis phenomenon can significantly distort the results of photocurrent measurement.

Thus, the detection of this hysteresis in the low-current region by the created complex can be used to test manufactured photodiode structures (crystals) prior to assembly into a housing. Such selection will be effective for precision photodiodes designed to measure small fluxes of optical radiation.