2D materials and their heterostructures are prominent for fabricating next-generation optical and photonic devices. The optical, electrical, and mechanical properties of 2D materials largely depend on atomic layer numbers. Although machine learning techniques are implemented to identify large-area thickness distribution using microscopic images, the existing work mainly focuses on rough identification of thicknesses with in-house datasets which limits fair and comprehensive comparisons of new machine learning approaches. Here, first a microscopic dataset is collected and released for three fundamental image processing tasks including multilabel classification, segmentation, and detection. Then three deep-learning architectures DenseNet, U-Net, and Mask-region convolutional neural network (RCNN) are benchmarked on three tasks and their robustness is evaluated on the augmented 2D microscopic images with different optical contrast variations. Deep learning models are trained and evaluated to identify mono-, bi-, tri-, multilayer and bulk flakes using microscopic images of MoS 2 fabricated on the SiO 2 /Si substrate by chemical vapor deposition. The relation between model performances and statistics of datasets is studied based on the international commission on illumination (CIE) 1931 color space and red, green, blue (RGB) histograms of optical contrast differences. Finally, the robust pretrained models are integrated into a graphic user interface for the on-site use of full field-of-view images captured by bright-field microscopes.

Wafer-scale two-dimensional (2D) semiconductors with atomically thin layers are promising materials for fabricating optic and photonic devices. Bright-field microscopy is a widely utilized method for large-area characterization, layer number identification, and quality assessment of 2D semiconductors based on optical contrast. Out-of-focus microscopic images caused by instrumental focus drifts contained blurred and degraded structural and color information, hindering the reliability of automated layer number identification of 2D nanosheets. To achieve automated restoration and accurate characterization, deep-learning-based microscopic imagery deblurring (MID) was developed. Specifically, a generative adversarial network with an improved loss function was employed to recover both the structural and color information of out-of-focus low-quality images. 2D MoS2 grown by the chemical vapor deposition on a SiO2/Si substrate was characterized. Quantitative indexes including structural similarity (SSIM), peak signal-to-noise ratio, and CIE 1931 color space were studied to evaluate the performance of MID for deblurring of out-of-focus images, with a minimum value of SSIM over 90% of deblurred images. Further, a pre-trained U-Net model with an average accuracy over 80% was implemented to segment and predict the layer number distribution of 2D nanosheet categories (monolayer, bilayer, trilayer, multi-layer, and bulk). The developed automated microscopic image deblurring using MID and the layer number identification by the U-Net model allow for on-site, accurate, and large-area characterization of 2D semiconductors for analyzing local optical properties. This method may be implemented in wafer-scale industrial manufacturing and quality monitoring of 2D photonic devices.

Furfurylation with a low concentration of furfuryl alcohol (FA) promotes the improvement of the properties and the effectiveness of FA on cell–wall action without darkening the furfurylated wood to the point that it affects its applications. In this paper, the effects of furfurylation on the hygroscopicity and water uptake dimensional stability of poplar (Populus sp.) and Chinese fir (Cunninghamia lanceolata) were analyzed. Meanwhile, the distribution of FA resin, the relationship between wood and water, the change in pore size distribution, and the weight percentage gain and cell wall bulking coefficient of wood were also investigated. The results were as follows: (1) A low concentration of FA could better enter the cell walls of the Chinese fir than the poplar, as FA resin was almost cured in the secondary walls, cell corners, and compound middle lamellae when a 10% concentration of FA was applied to the Chinese fir and poplar. When the FA concentration was increased to 30%, there were no significant increases in the amount of FA entering the cell walls and the amounts of FA cured in the cell lumen of the poplar were greater than those of the Chinese fir. Meanwhile, the modification of cell walls was more suitable in poplar than in Chinese fir. (2) The pointed ends of the pit chambers and the pit apertures (800–1000 nm) in the poplar and the small pores of the pit membranes and the pit apertures (1–6 μm) in the Chinese fir were partially deposited by the FA resin, which formed new pores in the size ranges of 80–600 nm and 15–100 nm, respectively. The porosity of the poplar was greater than that of the Chinese fir, and the bulk density of the poplar was less than that of the Chinese fir before and after modification. (3) Furfurylation with a low concentration of FA was able to better reduce the equilibrium moisture content, improve the anti-swelling efficiency, and enhance the dimensional stability of the poplar wood compared to the Chinese fir. Furfurylation effectively reduced water uptake due to the hydrophobic property of the FA resin. The water uptake of the Chinese fir increased by 17%–19% in second cyclic water soaking when treated with FA with various concentrations, which indicated the loss and leaching of FA resin during the test. Low-field NMR was used to demonstrate that the furfurylation not only reduced the amount of water but also affected the combination state of bound and free water with wood. Thus, furfurylation at a low concentration is a feasible method by which to extend applications of furfurylated wood.