The increase of device lifetime and reliability of THz photomixers will play an essential role in their possible future application. Therefore, their optimal work conditions/operation range, i.e., the maximal incident optical power should be experimentally estimated. We fabricated and tested THz photomixer devices based on nitrogen-implanted GaAs integrated with a Bragg reflector. Raman spectroscopy was applied to investigate the material properties and to disclose any reversible or irreversible material changes. The results indicate that degradation effects in the photomixer structures/material could be avoided if the total optical power density does not exceed levels of about 0.7 mW/µm2 for 100 min of operation. Furthermore, the investigations performed during 1000 min of optical exposure on the photomixer devices’ central region comprising interdigitated metal-semiconductor-metal (MSM) structures suggest a reversible “curing” mechanism if the power density level of ~0.58 mW/µm2 is not exceeded. Long-term operation (up to 1000 h) reveals that the photomixer structures can withstand an average optical power density of up to ~0.4 mW/µm2 without degradation when biased at 10 V. Besides the decrease of the position of the A1g (LO) Raman mode from ~291 cm−1 down to ~288 cm−1 with increasing optical power density and operation time, broad Raman modes evolve at about 210 cm−1, which can be attributed to degradation effects in the active photomixer/MSM area. In addition, the performed carrier lifetime and photomixer experiments demonstrated that these structures generated continuous wave sub-THz radiation efficiently as long as their optimal work conditions/operation range were within the limits established by our Raman studies.
