Optical determination of layered-materials InSe thickness via RGB contrast method and regression analysis

Indium selenide (InSe) features intriguing thickness-dependent optoelectronic properties, and a simple, and precise way to identify the thickness is essential for the rapid development of InSe research. Here, a red, green, and blue (RGB) color contrast method with regression analysis for quantitative correlation of three optical contrasts from RGB channels with the InSe thickness (1-35 nm), is demonstrated. The lower accuracy of the thickness identification obtained from the individual channels was discussed. Moreover, the effective refractive indices in the three RGB regions can be extracted from the Fresnel equation and numerical analysis by finding the best fit to the experimental optical contrast. After further consideration of the wavelength-dependent refractive indices, the slope of the regression line between the estimated thickness and that obtained from the atomic force microscope was improved from 1.59 +/- 0.05 to 0.97 +/- 0.02. The complex refractive index spectra of InSe (1-10 layers) generated from ab initio numerical calculation results were also adopted to identify the InSe thickness. Compared to dispersion, the evolution of the band structure had less effect on thickness identification. This work could be extended to other layered materials, facilitate the thickness-dependent study of layered materials, and expedite the realization of their practical applications.

Automated summary

This study demonstrates a red, green, and blue (RGB) color contrast method for identifying the thickness of Indium selenide (InSe). By analyzing the optical contrasts from the RGB channels, the thickness of InSe can be quantitatively correlated. The effective refractive indices are extracted and the wavelength-dependent refractive indices are considered to improve the accuracy of thickness estimation. In addition, the complex refractive index spectra generated from ab initio numerical calculation results are used for thickness identification. This work has significant implications for the study and practical applications of layered materials.