System Bits: Dec. 17

Realizing the full potential of graphene requires treatments that can be expensive and difficult to apply predictably, but new work by MIT and Berkeley researchers may overcome the hurdles. At the same time, AI could make chemical reactions intelligent, Harvard researchers say.


Simple, Inexpensive Graphene Treatment Could Unleash New Uses
To help realize the promise of graphene in electronics, solar power, and sensors, researchers from MIT and UC Berkeley have created what they said is a simple, inexpensive treatment that they believe may help realize the potential of the material.

While pure graphene lacks some key properties needed for electronic devices, modifying it through the addition of oxygen atoms can provide those properties, the researchers said. However, present methods leave oxygen atoms distributed unpredictably across the graphene’s surface and involve treatment with harsh chemicals, or at temperatures of 700 to 900 degrees Celsius.

The researchers’ approach involves exposing the material to relatively low temperatures, just 50 to 80 C, with no need for additional chemical treatment, which is a mild thermal approach compared to other approaches that have been reported, either thermal or chemical. This is a relatively environmentally-friendly method, with no harsh chemical treatment that generates harmful byproducts. Also, the treatment can easily be applied on a large scale, making commercial applications more feasible.

The low-temperature annealing process modifies the distribution of the oxygen atoms, causing them to form clusters and leaving areas of pure graphene between them, without introducing any disorder to the overall graphene structure — and most importantly, preserving the oxygen content.

The treatment allows the electrical resistance of the material to decrease by four to five orders of magnitude, which could be important for electronics, catalysis, and sensing applications. This is a result of the oxygen clustering, which renders the oxygen-rich regions insulating, but leaves the pure graphene areas in between conducting.

In addition, the pure graphene regions naturally have properties of “quantum dots,” which could find use as highly efficient light emitters, among other applications. The treatment also greatly enhances the material’s ability to absorb visible light, the team said, which could be important for its use in applications such as solar cells.

Making Chemical Reactions Intelligent Through Machine-Learning Algorithms
Computer scientists at the Harvard School of Engineering and Applied Sciences and the Wyss Institute for Biologically Inspired Engineering at Harvard University have joined forces to put powerful probabilistic reasoning algorithms in the hands of bioengineers. Researchers have shown that an important class of artificial intelligence algorithms could be implemented using chemical reactions, which use a technique called “message passing inference on factor graphs,” a mathematical coupling of ideas from graph theory and probability. They represent the state of the art in machine learning and are already critical components of everyday tools ranging from search engines and fraud detection to error correction in mobile phones.

This work demonstrates that some aspects of artificial intelligence (AI) could be implemented at microscopic scales using molecules. In the long term, such theoretical developments could open the door for “smart drugs” that can automatically detect, diagnose, and treat a variety of diseases using a cocktail of chemicals that can perform AI-type reasoning.

Leave a Reply