News & Events

The Material Times: What’s New in the World of Materials Science?

Simple smart glass recognizes digits: future of artificial vision

Here’s a five minute read to bring you up to date with the latest developments in the field of materials, spanning research, business, economics and culture. Here we go ☺️

Simple smart glass recognizes digits: future of artificial vision

Scientists at the University of Wisconsin Madison and MIT have proved through simulations that carefully designed glass with strategically placed air bubbles can work as a neural network to identify digits.

Fig. 1. © Erfan Khoram, Ang Chen, Dianjing Liu, Lei Ying, Qiqi Wang, Ming Yuan, and Zongfu Yu – (a) NNM trained to recognize handwritten digits. The input wave encodes the image as the intensity distribution. On the right side of the NNM, the optical energy concentrates to different locations depending on the image’s classification labels. (b) Two samples of the digit 2 and their optical fields inside the NNM. As can be seen, although the field distributions differ for the images of the same digit, they are classified as the same digit. (c) The same as (b) but for two samples of the digit 8. Also, in both (b) and (c), the boundaries of the trained medium have been shown with black borderlines.

You’re in need of a material that doesn’t exist yet. What if glass is the key element to your development? Challenge glass, challenge SCHOTT – get in touch with glass experts.

Once realized in practice, this can be used to create glasses that can recognize faces and function as biometric locks, which would not require any external energy while working at the speed of light. The simulation model currently has an accuracy of 79%.

Fig. 2. © Erfan Khoram, Ang Chen, Dianjing Liu, Lei Ying, Qiqi Wang, Ming Yuan, and Zongfu Yu Author Information – (a) Training starts by encoding an image as a vector of current source densities in the FDFD simulation. This step is followed by an iterative process to solve for the electric field in a nonlinear medium. Next, we use the ASM to calculate the gradient, which is then used to update the level-set function and consequently, the medium itself. Here we use mini-batch SGD (explained in the supplementary materials section of Ref. [17]). In training with mini-batches, we sum the cost functions calculated for different images in the same batch and compute the gradients. (b)–(d) show an NNM in training after 1, 33, and 66 training iterations, respectively. (After iteration 66, the medium has already seen each of the training samples at least once, since we are using batches of 100 images.) At each step, the boundary between the host material and the inclusions is shown, along with the field distribution for the same randomly selected digit 8. Also, the accuracy of the medium on the test set can be seen for that particular stage in training.

30 shades of steel: scientists develop ‘cheat sheet’ for the creation of new steels

Researchers at the National University of Science and Technology have created the most detailed thermodynamic database so far of the Lanthanum – Iron – Carbon and Lanthanum-Iron phase diagram. This is a collection of necessary thermodynamic parameters that are the starting point for certain specialized software that model new materials. 

This is a “cheat sheet” for material scientists, with which they can design new materials with required properties. The data helps optimize the development of new steels, by minimizing the time to search for new compositions and conduct the necessary experiments by a considerable margin.

With the database, the period of development of new steel grades can be reduced from 1 year to 1-2 months.

30 shades of steel: scientists develop ‘cheat sheet’ for the creation of new steels

Sunlight harvested by nanotubes

Nanotubes of tungsten disulfide have been demonstrated to be photovoltaic – they generate current from incident radiation. This is significantly different from the current semiconductor-based photovoltaic technology both in principle and in practice. 

These nanotube-based devices have a much higher current density than typical solar cells; however, their conversion efficiency leaves much to be desired. 

Solar cells using this technology could be made from an array of semiconductor nanotubes that could be scaled to increase output.

Two types of solar cell

Fig. 3. Two types of the solar cell. (a) A conventional solar cell is made of a semiconductor such as silicon. Electrical transport occurs through electron vacancies called holes in one region (bottom), and through electrons in another region (top). An electric field is generated across the junction between these two regions. When this junction is illuminated by sunlight, electron-hole pairs are produced. The electrons and holes are then separated by the electric field, giving rise to an electric current. (b) Zhang et al.1 report a junction-free solar cell that is made of a non-centrosymmetric semiconductor — one whose structure lacks symmetry under a transformation known as spatial inversion. Under illumination, electron-hole pairs are produced and separated because of a phenomenon called the bulk photovoltaic effect, generating an electric current.

What are the principles of a photovoltaic solar energy and what are the key materials used in solar cells? Learn more here.

New self-cleaning nanostructured glass inspired by butterfly wings

Butterfly wings have nanostructured designs on either side that repels water, oils and dirt while lending colour. New glass with applications in technologies like displays, tablets, laptops, smartphones, and solar cells has generously borrowed this design.  

Such glass is self-cleaning, antifogging, self-healing, super clear, highly transparent and stain-resistant – all because of random nanostructures on its surface that are smaller than the wavelengths of visible light.

New self-cleaning nanostructured glass inspired by butterfly wings

On the left, various liquids pool when poured on the regular glass. On the right, liquids bead, due to the superomniphobic properties created by Haghanifar and Leu. (Pittwire, University of Pittsburgh)

Researchers from the University of Pittsburgh’s Swanson School of Engineering took inspiration from natural surfaces like lotus leaves, moth eyes and butterfly wings that are self-cleaning, bacterial-resistant and water-repellant—adaptations for survival that evolved over millions of years. Combining that with machine learning, they were able to reproduce such properties in synthetic material.

Using machine learning to expedite design testing, associate professor Paul Leu (left) and doctoral candidate Sajad Haghanifar, both of the Swanson School of Engineering, have developed a glass that’s durable, clear, anti-fogging and liquid resistant. (Maggie Pavlick/University of Pittsburgh)

Time is running out for sand

Underlining what we already know and fear, a new report in Nature has once again highlighted the growing global scarcity of sand. Rapidly developing economies in China, India and Africa have an unsatiated appetite for sand to build new cities and to house their growing population. 

For manufacturing materials such as concrete, glass, and electronics, about 32 billion to 50 billion tonnes of sand are used annually. At this rate, demand might outstrip supply by the middle of this century. Such an unsustainable utilization of sand resources is ascribed to a shortage of knowledge and oversight. 

The authors have called on the United Nations Environment Programme (UNEP) and the World Trade Organization (WTO) to set up and oversee a global monitoring programme for sand resources.

Hot off the Press: unsupervised word embeddings capture latent knowledge from materials science literature

Natural language processing-based deep learning techniques were used to analyze relationships among words in 3.3 million materials science abstracts spanning decades and predict discoveries of new thermoelectric materials years in advance, recommend materials for functional applications before discovery, and suggest yet unknown materials. 

The authors at the Lawrence Berkeley National Lab have demonstrated that word embeddings capture not only syntactic, but also semantic relationships between materials, properties, processes and structures

Among other examples, the word embeddings can answer questions such as ‘Zirconium is to Hexagonal, as Chromium is to __? 

Hot off the Press: unsupervised word embeddings capture latent knowledge from materials science literature

So, these were 6 of the most recent studies in the world of material science. Which one do you think is the most promising of all? Which one do you relate to? Or is there any other recent innovation in material science that you believe is more worthy of mentioning?

🔥 Either way, let us know by writing your thoughts in the comment section below.

Matmatch Blog

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.