We recently used the 3 omega method to detect the thermal conductivity of a film with about 100nm thickness on a STO substrate with thickness about 500um. In the detection, we strangely found the calculated substrate temperature enhancement is higher than that of the heater line, leading to a negative film temperature enhancement so a negative thermal conductivity. It is unreasonable in our opinion. We have checked the calcualtion process carefully several times. Can anyone give some suggestions?