We develop a closed-loop solution to design, to optimize and to expansively fabricate the desirable quasi-random nanostructures(QRNs). In contrast to the current non-deterministic manufacturing process that cannot be deployed in large dimensionality, we innovatively import binary quasi-random sequences to generate QRNs deterministically without the restriction of the pattern size. Note that all 2D quasi-random patterns such as particle and channel types can also be converted into binary sequences by digitizing their 2D pattern images. Moreover, to bridge the gap between the nanostructure spatial arrangement and its optical performance, the star discrepancy calculation is employed as a guidance to evaluate and to optimize these binary QNRs given that the nanostructures’ uniformity is a key factor for light trapping. Finally, these binary QRNs are generated in a “pit-and-land” morphology so that they can be facilely and directly fabricated via optical disk recording technology.
|