Resistance measurement artifact during sudden change in resistance

[bua123]

Hi,
I am having trouble with the measurement of resistivity. I am doing a temperature-dependent resistance measurement. Basically I am acquiring voltage and current data with the temperature sweep of 100 K to 360 K range. I am using keithley 2450 SMU. I configured it for 4 wire sensing mode, where the source is constant current and measuring the voltage across the sample. My material undergoes metal to insulator transition depending on temperature. So, when it transforming the resistance (metal to insulator or vice versa) suddenly with temperature, there is an artifact of the measured voltage close to the transition temperature for both during cooling and heating. I looked clearly, the current source value is proper but the measured voltage value is not. I am attaching a few figures to make it clear. Please if anybody has any suggestions on it help me. Thanking you.

[bua123]

Hi all,
I am could not fix the problems yet. I am waiting for a suggestion. If anyone has an idea for fixing this above problem, please inform.
Thanks

[Andrea C]

To be honest, it’s not clear to me what the graph should look like. The material has hysteresis evidently. Can you give a mock up of what you are expecting?
From the third graph, you gave us the current and the voltage vs the temperature.
At low temp, there is almost 1 volt; at high temp about 1mV.

The 10uA of sourced current does not seem as flat as I’d expect. Was the instrument warmed up before the experiment was performed?
There is one current value at about 250K that dramatically changes. The graph scale does not tell us how much any of this change really is.
Are you using 10uA range?

For the source current, you specify a voltage compliance limit. How big is that? Needs to be larger, by at least 10%, than the max voltage you expect.

What about ranging? Auto ranging the voltage measure? Try fixed range (2v) to see if that gives you a more expected result.