No abstract

Porous anodic alumina has been of an increasing interest to applications of on-chip biosensors and bioseparations. However, the characteristics of molecular diffusion in alumina nanopores have not been fully explored. Here, we have investigated an ultra-thin freestanding alumina membrane fabricated on a silicon substrate as a new on-chip diffusion system. Sub-1,000 Da molecules such as caffeine were diffused through alumina pores with a pore diameter of 40 nm and a pore length of 1.2 lm. The diffusion dynamics was characterized by modeling the molecular transport as one-dimensional convective Fickian flow. The diffusion coefficients were calculated to be on the order of 10 -8 cm 2 s -1 . The transport rate of coions was enhanced by increasing the ionic strength of diffusion solution. Relative to thick alumina membranes, the thin-film alumina was able to achieve a significantly higher flux rate, making it more favorable for rapid molecular transport. The characterizations we present here have contributed important experimental data to alumina nanofluidics, and are believed to be valuable for applications such as drug deliveries, molecular separations, and membrane biosensors.

We have created hybrid devices consisting of nanoscale fabricated inorganic components integrated with and powered by a genetically engineered motor protein. We wish to increase the assembly yield and lifetime of these devices through identification, measurement, and improvement of weak internal bonds. Using dynamic force spectroscopy, we have measured the bond rupture force of (histidine)6 on a number of different surfaces as a function of loading rate. The bond sizes, lifetimes, and energy barrier heights were derived from these measurements. We compare the (His)6−nickel bonds to other bonds composing the hybrid device and describe preliminary measurements of the force tolerances of the protein itself. Pathways for improvement of device longevity and robustness are discussed.

A novel method has been developed for fabricating full or partial freestanding anodic alumina. In our method a sacrificial metal layer is introduced between an Al film and a Si 3 N 4 substrate. A freestanding alumina film at wafer scale is successfully achieved by anodizing the double metal layer, during which the alumina is spontaneously stripped off the Si 3 N 4 substrate due to the anodic oxidation of the sacrificial layer. The barrier oxide of the alumina film is effectively removed either by H 3 PO 4 dissolution or by CF 4 reactive ion etching. The freestanding alumina film is utilized as a contact mask to transfer its nanoporous pattern to a Si substrate. By patterning the sacrificial metal layer with contact lithography, a partial freestanding alumina film is successfully achieved on the silicon chip, producing a unique micro/nanofluidic channel. Compared with previous techniques, the method reported here is advantageous for its simplicity and flexibility.Index Terms-Anodic alumina, contact mask, double-layer anodization, freestanding, nanofluidic channel, sacrificial metal layer, thin film.

No abstract

No abstract

Classic image scaling (e.g. bicubic) can be seen as one convolutional layer and a single upscaling filter. Its implementation is ubiquitous in all display devices and image processing software. In the last decade deep learning systems have been introduced for the task of image super-resolution (SR), using several convolutional layers and numerous filters. These methods have taken over the benchmarks of image quality for upscaling tasks. Would it be possible to replace classic upscalers with deep learning architectures on edge devices such as display panels, tablets, laptop computers, etc.? On one hand, the current trend in Edge-AI chips shows a promising future in this direction, with rapid development of hardware that can run deep-learning tasks efficiently. On the other hand, in image SR only few architectures have pushed the limit to extreme small sizes that can actually run on edge devices at realtime. We explore possible solutions to this problem with the aim to fill the gap between classic upscalers and small deep learning configurations. As a transition from classic to deep-learning upscaling we propose edge-SR (eSR), a set of one-layer architectures that use interpretable mechanisms to upscale images. Certainly, a one-layer architecture cannot reach the quality of deep learning systems. Nevertheless, we find that for high speed requirements, eSR becomes better at trading-off image quality and runtime performance. Filling the gap between classic and deeplearning architectures for image upscaling is critical for massive adoption of this technology. It is equally important to have an interpretable system that can reveal the inner strategies to solve this problem and guide us to future improvements and better understanding of larger networks.

No abstract