Jeonbuk National University (JBNU) research team led by Professor Kwang-wook Park of the Electronic Materials Engineering Program presented an image-based screening method that uses photographs taken with a smartphone to quickly check the degradation state of two-dimensional layered semiconductor hafnium disulfide (HfS2) thin films prior to detailed analysis.

HfS2 is a layered semiconductor in the transition metal dichalcogenide (TMD) family with strong potential for electronic and optoelectronic device applications. However, it is sensitive to oxygen and moisture in air and suffers from rapid oxidation. Under high humidity, its characteristic color fades within days and the Raman signal decreases significantly, indicating large changes in its intrinsic physical and chemical properties.

This study is a follow-up to the team’s previously published work on ambient degradation of HfS2 thin films (DOI: 10.1016/j.mssp.2025.109471). In the prior study, JBNU PhD candidate Joo-chan Hwang analyzed the growth of HfS2 thin films and their humidity-dependent degradation, showing that the material degrades rapidly under high humidity and that protective coatings can suppress this degradation.

In the follow-up study, the team reanalyzed the previously accumulated photographs of HfS2 samples and the corresponding Raman measurements from a new perspective to assess whether the degradation state could be predicted in advance using only color and image information extracted from the photographs.

The research idea originated from the observation that HfS2 fades in color as it degrades. While discussing this phenomenon with his longtime friend and hobbyist coder, patent attorney Yong-cheol Jeon, Professor Park proposed that the correlation between color change and actual material degradation could be used to quickly check sample condition before Raman or other detailed measurements. Yong-cheol Jeon, now CEO of MyStar24 Patent Law Office, participated in the quantitative image analysis and data processing, helping to concretize the study.

The team quantified color, distribution, and texture information from photographs of HfS2 samples taken with a Samsung Galaxy Note10 5G (SM-N971N) smartphone. In particular, they used the b* value in the CIE Lab color space and the overall color difference ΔE to analyze changes as a function of air exposure time.

The photographs were taken without professional equipment, using a smartphone camera, indoor LED lighting, and a white A4 sheet. The analysis showed that the color changes observed in degraded HfS2 samples correlated with degradation trends confirmed by Raman measurements. For samples with clear optical changes, a simple analysis using only the b* value could estimate the degradation onset with an error of about 6.78 days.

The study demonstrates that if image databases of well-grown and degraded HfS2 are established, newly grown samples can be photographed under the same conditions with a smartphone and rapidly compared to determine their current condition. This can serve as an initial screening tool to check sample status and set analysis priorities before conducting detailed analyses such as Raman or XPS.

Professor Park explained, “HfS2 can oxidize rapidly in air even after successful growth, so its state at the time of measurement may differ from the initial state. This method is not a substitute for detailed analysis. It is an initial check-up tool to quickly determine the degree of degradation and decide which samples should be prioritized for analysis.”

The results were published under the title "Raman-calibrated image-based screening of ambient aging in CVD-grown HfS2 thin films" in the Q1 SCIE journal Materials Science in Semiconductor Processing (DOI: 10.1016/j.mssp.2026.111122).