Hi Tom Becker,

It is difficult to say something without seeing the spectra. I think you should check transients, if their are exponential, signal/noise ratio, etc

Dear Tom Becker,

The negative DLTS peaks are often related to a non-ideal back ohmic contact of a sample. I'd recommend reading this article.

Article Signature of a back contact barrier in DLTS spectra

Since the temperature is quite high (600 K), I would suspect some degradation of the ohmic contact. Some remaining oxide layer between the substrate and the back contact may also be a reason for this behavior.

Hi Dovile Meskauskaite ,

thank you for your answer!

Attached is a picture of the spectra and the transient at the tip of the negative peak. If you have any suggestion, i would be very greatful! :)

Kind regards,

Tom

Edit: pictures removed

Dear I. Kolevatov ,

thank you for the paper! I will definetly look into it.

I also did multiple measurements with the same and different chips of the same typ and always have this effect. Shouldn´t a degradation affect further measurements with the same sample (at lower temperatures) as well?

Kind regards,

Tom

Tom Becker, as you can see there is almost no changes of the amplitude when applying different reverse bias or filling pulses. Also, the peak shape is non-gaussian. Therefore, this peak might be not related to carrier emission from deep levels. The capacitance dependence on temperature for these spectra and C-V/I-V characteristics at that temperature, as well as the paper suggested above by Ilia, might helps you to find the reason for such behavior.

Dovile Meskauskaite, thank you very much! I will try to identify the origin of this peak with your tips.