I am measuring the resistance of a CNT thin film deposited on Si substrate with a four probe attachment setup. The setup consist of a nanovoltmeter and current source with copper probes. The I-V curve in a direct DC sweep measurement is in the figure. I was hoping if anyone may have an explanation for such a behavior or guide me into the potential measurement errors from the described setup (Such as effect of substrate, probes, etc) that may have caused this.