Scalable nanomaterial-based circuits and optoelectronics

将新材料转化为对社会有益的实际手段通常需要大量的时间和资本投资。基于孤立的低维材料纳米结构的器件，如0D光学活性缺陷、1D单独的纳米线或2D异质结，具有独特的光电子特性，可以实现传统块体材料无法实现的电子和光子应用。在为集成系统放大这些设备时，需要知道每个组成纳米结构的确切位置。因此，制造和测量纳米级设备，特别是早期新兴材料的设备，是出了名的劳动密集型设备。该项目将建立基于计算机视觉的方法，如目标检测和跟踪，以及多模式数据关联，用于高通量材料测试和评估、设备布线和集成系统原型。主要目标是：i)展示多技术、多尺度的表征，在加工步骤之间对相同的纳米结构自动执行，以关联材料属性，跟踪加工引起的变化，定位高质量的结构，并解决成品率和可扩展性方面的挑战。Ii)使用计算机可调电极设计的纳米材料集成系统原型，其中机器学习算法自动在芯片上对单个设备进行布线。

Translating new materials into practical devices of benefit to society typically requires substantial time and capital investment. Devices based on isolated low-dimensional material nanostructures, such as 0D optically active defects, 1D individual nanowires, or 2D heterojunctions, have unique optoelectronic characteristics which can enable electronic and photonic applications not possible with conventional bulk materials. When scaling up these devices for integrated systems, the exact position of each constituent nanostructure needs to be known. As such, fabricating and measuring nanoscale devices, particularly for emerging materials in the early stages, is notoriously labour-intensive. This project will build upon computer-vision-based approaches such as object detection & tracking, and multi-modal data correlation, for high-throughput materials test and evaluation, device routing, and integrated system prototyping. The main objectives are: i) Demonstrate multi-technique, multi-scale characterisation, automatically performed on the same nanostructures between processing steps, to correlate material properties, track processing-induced changes, locate high-quality structures, and address the challenges of yield and scalability. ii) Prototype nanomaterial integrated systems using computer-adjustable electrode designs where a machine-learning algorithm automatically routes individual devices across a chip.