Abstract
Pump-Probe thermoreflectance (Pump-probe TR) is a non-contact detection technique that has been widely used for thermal characterization of materials. In the traditional characterization process, spot detection is usually employed to fit unknown thermal property parameters using non-linear fitting process. However, when processing a large amount of data in a specified area of a sample, the traditional measurement process appears to be time-consuming and labor-intensive. In this work, we propose a high-throughput method for semiconductor thermophysical characterization. The optical path of the measurement system is combined with automatic control components to realize automatic scanning measurements. Deep learning techniques are utilized for the high-throughput data processing. We firstly demonstrated the entire measuring process with Au-sapphire sample, whose interlayers are intentionally controlled by coating different interlayers. The validity of the method can be demonstrated by the measurement results of the thermal boundary conductance of Au-sapphire and the thermal conductivity of sapphire in the scanned area. Then, we demonstrated the application of non-destructive scanning measurement in the industrial production of GaN-on-Si sample, comparing the measurement results at different resolutions. We validate the scanning results demonstrating that this method can measure with high accuracy and speed. Meanwhile the high-resolution scanning measurement can observe the subtle difference of thermal characterization in the area. This method significantly reduces the time and labor required to measure compared to traditional method and it is particularly efficient for thermophysical characterization detection of high-volume wafers.
Funding Information
  • National Natural Science Foundation of China (62104173)

This publication has 53 references indexed in Scilit: