Machine-learning system finds patterns in materials “recipes,” even when training data is lacking.


MIT Materials Synthesis Web
A new machine-learning system for analyzing materials “recipes” uses a variational autoencoder, which squeezes data (left-hand circles) down into a more compact form (center circles) before attempting to re-expand it into its original form (right-hand circles). If the autoencoder is successfully trained, the compact representation will capture the data’s most salient characteristics. Image, Chelsea Turner, MIT

Last month, three MIT materials scientists and their colleagues published a paper describing a new artificial-intelligence system that can pore through scientific papers and extract “recipes” for producing particular types of materials.
That work was envisioned as the first step toward a system that can originate recipes for materials that have been described only theoretically. Now, in a paper in the journal npj Computational Materials, the same three materials scientists, with a colleague in MIT’s Department of Electrical Engineering and Computer Science (EECS), take a further step in that direction, with a new artificial-intelligence system that can recognize higher-level patterns that are consistent across recipes.

For instance, the new system was able to identify correlations between “precursor” chemicals used in materials recipes and the crystal structures of the resulting products. The same correlations, it turned out, had been documented in the literature.

The system also relies on statistical methods that provide a natural mechanism for generating original recipes. In the paper, the researchers use this mechanism to suggest alternative recipes for known materials, and the suggestions accord well with real recipes.

The first author on the new paper is Edward Kim, a graduate student in materials science and engineering. The senior author is his advisor, Elsa Olivetti, the Atlantic Richfield Assistant Professor of Energy Studies in the Department of Materials Science and Engineering (DMSE). They’re joined by Kevin Huang, a postdoc in DMSE, and by Stefanie Jegelka, the X-Window Consortium Career Development Assistant Professor in EECS.

Sparse and scarce

Like many of the best-performing artificial-intelligence systems of the past 10 years, the MIT researchers’ new system is a so-called neural network, which learns to perform computational tasks by analyzing huge sets of training data. Traditionally, attempts to use neural networks to generate materials recipes have run up against two problems, which the researchers describe as sparsity and scarcity.

Any recipe for a material can be represented as a vector, which is essentially a long string of numbers. Each number represents a feature of the recipe, such as the concentration of a particular chemical, the solvent in which it’s dissolved, or the temperature at which a reaction takes place.

Since any given recipe will use only a few of the many chemicals and solvents described in the literature, most of those numbers will be zero. That’s what the researchers mean by “sparse.”

Similarly, to learn how modifying reaction parameters — such as chemical concentrations and temperatures — can affect final products, a system would ideally be trained on a huge number of examples in which those parameters are varied. But for some materials — particularly newer ones — the literature may contain only a few recipes. That’s scarcity.

“People think that with machine learning, you need a lot of data, and if it’s sparse, you need more data,” Kim says. “When you’re trying to focus on a very specific system, where you’re forced to use high-dimensional data but you don’t have a lot of it, can you still use these neural machine-learning techniques?”

Neural networks are typically arranged into layers, each consisting of thousands of simple processing units, or nodes. Each node is connected to several nodes in the layers above and below. Data is fed into the bottom layer, which manipulates it and passes it to the next layer, which manipulates it and passes it to the next, and so on. During training, the connections between nodes are constantly readjusted until the output of the final layer consistently approximates the result of some computation.

The problem with sparse, high-dimensional data is that for any given training example, most nodes in the bottom layer receive no data. It would take a prohibitively large training set to ensure that the network as a whole sees enough data to learn to make reliable generalizations.

Artificial bottleneck

The purpose of the MIT researchers’ network is to distill input vectors into much smaller vectors, all of whose numbers are meaningful for every input. To that end, the network has a middle layer with just a few nodes in it — only two, in some experiments.

The goal of training is simply to configure the network so that its output is as close as possible to its input. If training is successful, then the handful of nodes in the middle layer must somehow represent most of the information contained in the input vector, but in a much more compressed form. Such systems, in which the output attempts to match the input, are called “autoencoders.”

Autoencoding compensates for sparsity, but to handle scarcity, the researchers trained their network on not only recipes for producing particular materials, but also on recipes for producing very similar materials. They used three measures of similarity, one of which seeks to minimize the number of differences between materials — substituting, say, just one atom for another — while preserving crystal structure.

During training, the weight that the network gives example recipes varies according to their similarity scores.

Playing the odds

In fact, the researchers’ network is not just an autoencoder, but what’s called a variational autoencoder. That means that during training, the network is evaluated not only on how well its outputs match its inputs, but also on how well the values taken on by the middle layer accord with some statistical model — say, the familiar bell curve, or normal distribution. That is, across the whole training set, the values taken on by the middle layer should cluster around a central value and then taper off at a regular rate in all directions.

After training a variational autoencoder with a two-node middle layer on recipes for manganese dioxide and related compounds, the researchers constructed a two-dimensional map depicting the values that the two middle nodes took on for each example in the training set.
Remarkably, training examples that used the same precursor chemicals stuck to the same regions of the map, with sharp boundaries between regions. The same was true of training examples that yielded four of manganese dioxide’s common “polymorphs,” or crystal structures. And combining those two mappings indicated correlations between particular precursors and particular crystal structures.

“We thought it was cool that the regions were continuous,” Olivetti says, “because there’s no reason that that should necessarily be true.”

Variational autoencoding is also what enables the researchers’ system to generate new recipes. Because the values taken on by the middle layer adhere to a probability distribution, picking a value from that distribution at random is likely to yield a plausible recipe.

“This actually touches upon various topics that are currently of great interest in machine learning,” Jegelka says. “Learning with structured objects, allowing interpretability by and interaction with experts, and generating structured complex data — we integrate all of these.”

“‘Synthesizability’ is an example of a concept that is central to materials science yet lacks a good physics-based description,” says Bryce Meredig, founder and chief scientist at Citrine Informatics, a company that brings big-data and artificial-intelligence techniques to bear on materials science research. “As a result, computational screens for new materials have been hamstrung for many years by synthetic inaccessibility of the predicted materials. Olivetti and colleagues have taken a novel, data-driven approach to mapping materials syntheses and made an important contribution toward enabling us to computationally identify materials that not only have exciting properties but also can be made practically in the laboratory.”

The research was supported by the National Science Foundation, the Natural Sciences and Engineering Research Council of Canada, the U.S. Office of Naval Research, the MIT Energy Initiative, and the U.S. Department of Energy’s Basic Energy Science Program.

Larry Hardesty | MIT News Office
December 21, 2017



Tuesday, 19 December 2017 14:36

Engineers create plants that glow

Illumination from nanobionic plants might one day replace some electrical lighting.


MIT Glowing Plants Web
Illumination of a book (“Paradise Lost,” by John Milton) with the nanobionic light-emitting plants (two 3.5-week-old watercress plants). The book and the light-emitting watercress plants were placed in front of a reflective paper to increase the influence from the light emitting plants to the book pages. Image, Seon-Yeong Kwak

Imagine that instead of switching on a lamp when it gets dark, you could read by the light of a glowing plant on your desk.

MIT engineers have taken a critical first step toward making that vision a reality. By embedding specialized nanoparticles into the leaves of a watercress plant, they induced the plants to give off dim light for nearly four hours. They believe that, with further optimization, such plants will one day be bright enough to illuminate a workspace.

“The vision is to make a plant that will function as a desk lamp — a lamp that you don’t have to plug in. The light is ultimately powered by the energy metabolism of the plant itself,” says Michael Strano, the Carbon P. Dubbs Professor of Chemical Engineering at MIT and the senior author of the study.

This technology could also be used to provide low-intensity indoor lighting, or to transform trees into self-powered streetlights, the researchers say. MIT postdoc Seon-Yeong Kwak is the lead author of the study, which appears in the journal Nano Letters.

Nanobionic plants

Plant nanobionics, a new research area pioneered by Strano’s lab, aims to give plants novel features by embedding them with different types of nanoparticles. The group’s goal is to engineer plants to take over many of the functions now performed by electrical devices. The researchers have previously designed plants that can detect explosives and communicate that information to a smartphone, as well as plants that can monitor drought conditions.

Lighting, which accounts for about 20 percent of worldwide energy consumption, seemed like a logical next target. “Plants can self-repair, they have their own energy, and they are already adapted to the outdoor environment,” Strano says. “We think this is an idea whose time has come. It’s a perfect problem for plant nanobionics.”

To create their glowing plants, the MIT team turned to luciferase, the enzyme that gives fireflies their glow. Luciferase acts on a molecule called luciferin, causing it to emit light. Another molecule called co-enzyme A helps the process along by removing a reaction byproduct that can inhibit luciferase activity.

The MIT team packaged each of these three components into a different type of nanoparticle carrier. The nanoparticles, which are all made of materials that the U.S. Food and Drug Administration classifies as “generally regarded as safe,” help each component get to the right part of the plant. They also prevent the components from reaching concentrations that could be toxic to the plants.

The researchers used silica nanoparticles about 10 nanometers in diameter to carry luciferase, and they used slightly larger particles of the polymers PLGA and chitosan to carry luciferin and coenzyme A, respectively. To get the particles into plant leaves, the researchers first suspended the particles in a solution. Plants were immersed in the solution and then exposed to high pressure, allowing the particles to enter the leaves through tiny pores called stomata.

Particles releasing luciferin and coenzyme A were designed to accumulate in the extracellular space of the mesophyll, an inner layer of the leaf, while the smaller particles carrying luciferase enter the cells that make up the mesophyll. The PLGA particles gradually release luciferin, which then enters the plant cells, where luciferase performs the chemical reaction that makes luciferin glow.

Video: Melanie Gonick/MIT

The researchers’ early efforts at the start of the project yielded plants that could glow for about 45 minutes, which they have since improved to 3.5 hours. The light generated by one 10-centimeter watercress seedling is currently about one-thousandth of the amount needed to read by, but the researchers believe they can boost the light emitted, as well as the duration of light, by further optimizing the concentration and release rates of the components.

Plant transformation

Previous efforts to create light-emitting plants have relied on genetically engineering plants to express the gene for luciferase, but this is a laborious process that yields extremely dim light. Those studies were performed on tobacco plants and Arabidopsis thaliana, which are commonly used for plant genetic studies. However, the method developed by Strano’s lab could be used on any type of plant. So far, they have demonstrated it with arugula, kale, and spinach, in addition to watercress.

For future versions of this technology, the researchers hope to develop a way to paint or spray the nanoparticles onto plant leaves, which could make it possible to transform trees and other large plants into light sources.

“Our target is to perform one treatment when the plant is a seedling or a mature plant, and have it last for the lifetime of the plant,” Strano says. “Our work very seriously opens up the doorway to streetlamps that are nothing but treated trees, and to indirect lighting around homes.”

The researchers have also demonstrated that they can turn the light off by adding nanoparticles carrying a luciferase inhibitor. This could enable them to eventually create plants that shut off their light emission in response to environmental conditions such as sunlight, the researchers say.

The research was funded by the U.S. Department of Energy.

back to newsletter

Anne Trafton | MIT News Office 
December 12, 2017


More than half of Roxbury, Bunker Hill, students who get summer lab experience at MIT go on to earn a four-year degree.
Susan Rosevear Fall 2017 MRS 0541 DP Web
Community college students who experience a summer of research at MIT develop greater self-confidence and better academic skills, with a majority completing a four-year degree, MIT Materials Research Laboratory Education Officer Susan Rosevear told a symposium at the Materials Research Society Fall meeting in Boston on Monday, Nov. 27, 2017. Photo, Denis Paiste, MIT MRL

A summer of research at MIT gives inner-city Boston community college students a pathway toward greater self-confidence, better academic skills and a four-year college degree, MIT Materials Research Laboratory Education Officer Susan Rosevear said Monday, Nov. 27, 2017, during a symposium at the Materials Research Society [MRS] Fall meeting in Boston.

“Many of them have barely heard about materials science when they come to MIT, and by the end of the summer, they get sort of a full dunk into the world of materials science, so they are better informed to go forward,” Rosevear says. Over the past dozen years, 63 students from Roxbury and Bunker Hill Community Colleges have participated in the program at MIT. Of these, 45 went on to earn a four-year degree, with 34 pursuing degrees in science or engineering. Five continued on to graduate school in science or medicine.

The Research Experience for Undergraduates (REU) program is primarily funded through the MIT Materials Research Laboratory’s National Science Foundation-funded Materials Research Science and Engineering Center [NSF-MRSEC]. Bringing in underrepresented, or non-traditional, students from the community colleges broadens the diversity of students in the REU program.

“We are trying, and I think succeeding, in providing opportunities to community college students that they don’t have at their home institutions,” Rosevear says. Students learn to use electron microscopes, X-ray diffraction spectrometers and other advanced materials science characterization tools. Rosevear addressed a session at MRS highlighting collaborations between community colleges and four-year colleges.

In 2005, the MIT MRSEC, then part of the Center for Materials Science and Engineering, began the partnership with Roxbury Community College with seven students participating during its first year. In recent years, the summer program expanded to include community college professors in materials research on campus led by MIT faculty. So far, nine community college professors have participated. CMSE merged in October 2017 with the Materials Processing Center to form the MIT Materials Research Laboratory.

During the fall 2017 semester, Roxbury Community College Chemistry and Biotechnology Professor Kimberly Stieglitz offered a new course at Roxbury Community College, Research Science, [SCI 281] that brought students to the X-ray diffraction facility at MIT to examine their lab samples. “We keep finding new ways to leverage this partnership,” Rosevear says. Stieglitz and other teachers who have participated in the summer teachers’ program at MIT, also have incorporated material from their summer research into their classroom teaching, Rosevear notes.

Students must complete a basic engineering or science course, such as chemistry or biology, to be accepted into the MIT summer program. Community college teachers select the students based on academic record, statements of interest and faculty letters of recommendation. “They’ve been great partners for us, which is really key to the whole thing,” Rosevear explains. “Kimberly [Stieglitz] has told me, once they are selected, just knowing they are going to MIT changes their performance, they become more serious about themselves, their performance, motivation increases, and they have an increased commitment to STEM,” Rosevear says.

CMSE Scholar Kimberly Stieglitz Jode Lavine 9157 DP Web
Roxbury Community College Chemistry and Biotechnology Professor Kimberly Stieglitz [left] discusses her summer research at MIT with JoDe M. Lavine, Bunker Hill Community College Professor and Chairperson of the Engineering & Physical Sciences Department, during the annual Summer Scholars Poster Session on Aug. 3, 2017. Stieglitz worked in the lab of AMAX Career Development Assistant Professor in Materials Science and Engineering Robert J. Macfarlane. Photo, Denis Paiste, MIT MRL.

Over the course of the summer, community college students attend weekly luncheon meetings covering topics such as crafting a high-quality poster presentation, applying to graduate school, understanding patents and trademarks, and pursuing careers in materials science and other engineering fields.

Interest among MIT faculty in hosting community college students continues to grow. “I have people coming to me and say, how do I get one of these students?
The students have sold themselves, is essentially what’s happened,” Rosevear says.

The community college program is distinct from the Summer Scholars program, which is open to undergraduates in science and engineering from across the U.S. and Puerto Rico who are citizens or legal residents. Applications for summer 2018 must be submitted by Feb. 16, 2018.

 back to newsletter

Denis Paiste, MIT Materials Research Laboratory
December 19, 2017

Thursday, 07 December 2017 17:31

Engineers 3-D print a “living tattoo”

New technique 3-D prints programmed cells into living devices for first time. 
MIT 3D Living Tattoo
MIT engineers have devised a 3-D printing technique that uses a new kind of ink made from genetically programmed living cells. Courtesy of the researchers

MIT engineers have devised a 3-D printing technique that uses a new kind of ink made from genetically programmed living cells.

The cells are engineered to light up in response to a variety of stimuli. When mixed with a slurry of hydrogel and nutrients, the cells can be printed, layer by layer, to form three-dimensional, interactive structures and devices.

The team has then demonstrated its technique by printing a “living tattoo” — a thin, transparent patch patterned with live bacteria cells in the shape of a tree. Each branch of the tree is lined with cells sensitive to a different chemical or molecular compound. When the patch is adhered to skin that has been exposed to the same compounds, corresponding regions of the tree light up in response.

The researchers, led by Xuanhe Zhao, the Noyce Career Development Professor in MIT’s Department of Mechanical Engineering, and Timothy Lu, associate professor of biological engineering and of electrical engineering and computer science, say that their technique can be used to fabricate “active” materials for wearable sensors and interactive displays. Such materials can be patterned with live cells engineered to sense environmental chemicals and pollutants as well as changes in pH and temperature.

What’s more, the team developed a model to predict the interactions between cells within a given 3-D-printed structure, under a variety of conditions. The team says researchers can use the model as a guide in designing responsive living materials.

Zhao, Lu, and their colleagues have published their results Dec. 5, 2017,  in the journal Advanced Materials. The paper’s co-authors are graduate students Xinyue Liu, Hyunwoo Yuk, Shaoting Lin, German Alberto Parada, Tzu-Chieh Tang, Eléonore Tham, and postdoc Cesar de la Fuente-Nunez.

A hardy alternative

In recent years, scientists have explored a variety of responsive materials as the basis for 3D-printed inks. For instance, scientists have used inks made from temperature-sensitive polymers to print heat-responsive shape-shifting objects. Others have printed photoactivated structures from polymers that shrink and stretch in response to light.

Zhao’s team, working with bioengineers in Lu’s lab, realized that live cells might also serve as responsive materials for 3D-printed inks, particularly as they can be genetically engineered to respond to a variety of stimuli. The researchers are not the first to consider 3-D printing genetically engineered cells; others have attempted to do so using live mammalian cells, but with little success.

“It turns out those cells were dying during the printing process, because mammalian cells are basically lipid bilayer balloons,” Yuk says. “They are too weak, and they easily rupture.”
Instead, the team identified a hardier cell type in bacteria. Bacterial cells have tough cell walls that are able to survive relatively harsh conditions, such as the forces applied to ink as it is pushed through a printer’s nozzle. Furthermore, bacteria, unlike mammalian cells, are compatible with most hydrogels — gel-like materials that are made from a mix of mostly water and a bit of polymer. The group found that hydrogels can provide an aqueous environment that can support living bacteria.

The researchers carried out a screening test to identify the type of hydrogel that would best host bacterial cells. After an extensive search, a hydrogel with pluronic acid was found to be the most compatible material. The hydrogel also exhibited an ideal consistency for 3-D printing.

“This hydrogel has ideal flow characteristics for printing through a nozzle,” Zhao says. “It’s like squeezing out toothpaste. You need [the ink] to flow out of a nozzle like toothpaste, and it can maintain its shape after it’s printed

Video Abstract for Advanced Materials, 2017, 29, 1704821. Reproduced with permission. ©2017, Wiley-VCH Verlag GmbH & Co. KGaA.

From tattoos to living computers

Lu provided the team with bacterial cells engineered to light up in response to a variety of chemical stimuli. The researchers then came up with a recipe for their 3-D ink, using a combination of bacteria, hydrogel, and nutrients to sustain the cells and maintain their functionality. “We found this new ink formula works very well and can print at a high resolution of about 30 micrometers per feature,” Zhao says. “That means each line we print contains only a few cells. We can also print relatively large-scale structures, measuring several centimeters.”

They printed the ink using a custom 3-D printer that they built using standard elements combined with fixtures they machined themselves. To demonstrate the technique, the team printed a pattern of hydrogel with cells in the shape of a tree on an elastomer layer. After printing, they solidified, or cured, the patch by exposing it to ultraviolet radiation. They then adhere the transparent elastomer layer with the living patterns on it, to skin.

To test the patch, the researchers smeared several chemical compounds onto the back of a test subject’s hand, then pressed the hydrogel patch over the exposed skin. Over several hours, branches of the patch’s tree lit up when bacteria sensed their corresponding chemical stimuli. The researchers also engineered bacteria to communicate with each other; for instance they programmed some cells to light up only when they receive a certain signal from another cell. To test this type of communication in a 3-D structure, they printed a thin sheet of hydrogel filaments with “input,” or signal-producing bacteria and chemicals, overlaid with another layer of filaments of an “output,” or signal-receiving bacteria. They found the output filaments lit up only when they overlapped and received input signals from corresponding bacteria .

Yuk says in the future, researchers may use the team’s technique to print “living computers” — structures with multiple types of cells that communicate with each other, passing signals back and forth, much like transistors on a microchip. “This is very future work, but we expect to be able to print living computational platforms that could be wearable,” Yuk says.

For more near-term applications, the researchers are aiming to fabricate customized sensors, in the form of flexible patches and stickers that could be engineered to detect a variety of chemical and molecular compounds. They also envision their technique may be used to manufacture drug capsules and surgical implants, containing cells engineered produce compounds such as glucose, to be released therapeutically over time.

“We can use bacterial cells like workers in a 3-D factory,” Liu says. “They can be engineered to produce drugs within a 3-D scaffold, and applications should not be confined to epidermal devices. As long as the fabrication method and approach are viable, applications such as implants and ingestibles should be possible.”
This research was supported, in part, by the Office of Naval Research, National Science Foundation, National Institutes of Health, and MIT Institute for Soldier Nanotechnologies. 

Jennifer Chu | MIT News Office
December 5, 2017   

Monday, 04 December 2017 16:29

Lucia Brunel named 2018 Marshall Scholar

Summer Scholar Lucia Brunel 8899 DP Web
Lucia Brunel uses an active microrheology, optical tweezer setup in the MIT Bioinstrumentation Lab during her 2017 Summer Scholar internship at MIT under Professors Gareth McKinley and Katharina Ribbeck. Photo, Denis Paiste, MIT MRL. 

2017 Summer Scholar Lucia Brunel is one of 43 outstanding American undergraduate students selected as Marshall Scholars for up to three years of study at leading British universities in a wide variety of disciplines beginning in September 2018.

During summer 2017, Brunel interned at MIT, where she worked on a joint project under Professors Gareth McKinley and Katharina Ribbeck. “My research project as a Summer Scholar was an investigation of the self-healing behavior of certain biological gels. This project was a great opportunity for me to participate in interdisciplinary materials science research with both mechanical and biological engineering components,” Brunel says.

"The whole lab was so excited to hear the great news,” says MIT graduate student Caroline Wagner, who supervised Brunel’s internship. “I had such a good time working with Lucia through the Summer Scholars program, and I wish her all the best in her future studies."

Brunel plans to study materials science at Cambridge University after she graduates from Northwestern University in June 2018 with both B.S. and M.S. degrees in Chemical and Biological Engineering. Brunel plans to work in the laboratory of Cambridge Centre for Medical Materials Co-Directors and Professors Ruth Cameron and Serena Best, whose research aims to improve performance of biopolymer scaffolds for tissue engineering. “My work in particular will focus on chemically modifying collagen-based scaffolds to achieve more selective and controllable cell affinity that can be tailored for the type of tissue that must be regenerated,” Brunel explains. Brunel previously received a Goldwater Scholarship, which supports students in science, mathematics, and engineering.

“My materials science research experience at MIT taught me many of the material characterization skills that I will utilize at the University of Cambridge,” says Brunel, who intends to pursue a PhD in the U.S. in the area of biomaterials after her Marshall Scholar program. Three MIT seniors also were named 2018 Marshall Scholars. 

The Materials Research Laboratory sponsors the Summer Research Internship Program through the NSF REU program. Brunel participated through the former Materials Processing Center and Center for Materials Science and Engineering, which merged in October 2017 to form the Materials Research Laboratory at MIT.

The Summer Scholars program started in 1983, and has brought hundreds of the best science and engineering undergraduates in the country to MIT for graduate-level materials research. A wide range of project areas is available. Applications for summer 2018 must be submitted by Feb. 16, 2018.

back to newsletter – Materials Research Laboratory
December 5, 2017



Forbes calls this year's 30 Under 30 lists an "encyclopedia of creative disruption."
mit forbes 30 under 30 2018 Web
At least 30 MIT faculty, research staff, and alumni are listed throughout Forbes’ seventh annual 30 Under 30 edition, featuring some of the world’s best young innovators.

Forbes calls its 2018 30 Under 30 lists an “encyclopedia of creative disruption featuring 600 young stars in 20 different industries.” So it should come as no surprise that these lists are heavily populated by recent MIT graduates and other members of the Institute community.

Similar to past years, at least 30 MIT faculty, research staff, and alumni are listed throughout Forbes’ seventh annual edition of the world’s best young innovators. Read about the MIT community members who made this year’s list below:

Omar Abudayyeh ’12 and Jonathan Gootenberg ’13 (health care), doctoral candidates at the Broad Institute of Harvard and MIT. “Abudayyeh and Gootenberg pioneered two advances: a new enzyme for editing genes and a new technique for editing RNA.”

David Bierman SM ’14, PhD ’17 (energy), founder of Marigold Power, Inc. “At MIT he helped to develop a thermophotovoltaic converter that absorbs sunlight and converts it to a form of light.”

Greg Brockman ’13 (enterprise technology), chief technology officer of OpenAI. “The boundary-breaking nonprofit is dedicated to building safe AI and ensuring AI’s benefits are widely and evenly distributed.”

Ritchie Chen SM ’13, PhD ’16 (science), postdoc at the MIT Institute for Medical Engineering and Science. “Chen’s research found that dysfunctional brain regions could be stimulated by metal nanoparticles powered by magnetic fields.”

Tiffany Chu ’10 (enterprise technology), cofounder of Remix. “Chu is cofounder of Remix, a public transit platform used by more than 200 agencies worldwide…that evaluates transit data and suggests improvements.”

Lisa Conn MBA ’17 (law and policy), strategic partner and manager at Facebook. “Conn joined Facebook to lead the civic leadership team in its community partnerships program. Previously at the MIT Media Lab, Conn was program manager of the Electome Project.”

Cody Daniel ’11 (science), director of research at 3Scan. “Instead of fewer than 10 slices, 3Scan’s … robotic microscope can turn a small tissue sample into up to 60,000 slices.”

Maher Damak SM ’15 and Karim Khalil SM ’14 (energy), cofounders of Infinite Cooling. “Infinite Cooling … recaptures 80 percent of the water vapor that normally escapes from cooling towers attached to big power plants.”

Karen Dubbin ’12 (manufacturing and industry), science director at Aether. “Dubbin is the science director at Aether, which builds 3-D-printers capable of creating living tissue. She’s responsible for creating the 'bio-inks' that Aether uses to build tissues.”

Gregory Falco (enterprise technology), graduate student in the MIT Department of Urban Studies ans Planning and cofounder of NeuroMesh. “NeuroMesh provides endpoint security for smart devices and re-engineers malware to become a vaccine for the Internet of Things.”

Alistair Johnson (health care), postdoc in the MIT Laboratory for Computational Physiology. “Johnson created a database of ICU records used by 4,000 researchers from 30 countries to conduct clinical research.”

Brent Keller PhD ’16 (manufacturing), cofounder of Via Separations. “Via Separations develops membrane materials for separation processes. Keller [is] part of MIT’s The Engine accelerator program.”

Weihua Li ’15, MEng ’16 and Arun Saigal ’13, MEng ’13 (consumer technology), cofounders of Thunkable. “Saigal and Li decided spin-out MIT’s App Inventor tool, the drag-and-drop service for building your own app.”

Karthish Manthiram (science), assistant professor in the MIT Department of Chemical Engineering. “Manthiram’s research is focused on providing farmers with fertilizer by manufacturing it out of thin air, literally, by using air, water, and solar power.”

Jess Newman MBA ’17 (energy), director of U.S agronomy at Anheuser Busch InBev. "Her team of 15 agronomists advise barley, rice, and hop farmers on how to become more efficient."

Christina Qi ’13 and Jonathan Wang ’13, MEng ’15 (finance), partners at Domeyard LP. “[Domeyard] is a small hedge fund that is using high-frequency strategies to trade U.S. equity futures and currencies.”

Ritu Raman (science), postdoc at MIT's Koch Institute for Integrative Cancer Research. “Raman’s research focuses on understanding the dynamic interactions between biological and synthetic materials and developing bio-hybrid systems to tackle different applications. ”

Yichen Shen PhD ’16 (energy), postdoc in the MIT Research Laboratory of Electronics. “Has contributed to nanophotonic breakthroughs that could shape the future of energy. Light-AI designs computer chips powered by light rather than electricity.”

Hao Sun (science), research affiliate in the MIT Department of Civil and Environmental Engineering and assistant professor at the University of Pittsburgh. “Sun’s research uses analytics and machine learning combined with internet-of-things enabled sensors to track the health of buildings.”

Scott Sundvor ’12 (consumer technology), cofounder of Nima. “Nima is a portable bluetooth-enabled device that tests foods for allergens before you eat. The company has raised more than $20 million between venture funding and government grants.”

Michael Tomovich SM ’14 (manufacturing), cofounder of Kuvee. “Kuvee has engineered a patented, smart wine bottle that prevents oxygen from reaching the wine inside, and has raised $10 million in venture funding to roll it out.”

Sin Wang SM ’15 (science), cofounder of TetraScience. “Wang is cofounder and CTO of TetraScience, which is bringing the internet of things to daily lab work. The company has raised over $10 million in venture capital.”

Andrew Warren PhD ’16 (health care), founding scientist at Glympse Bio. “Based on Andrew Warren’s PhD at MIT, Glympse Bio uses modular nanoparticle sensors to create diagnostics for diseases potentially including cancer. ”

You Wu SM ’14 (manufacturing), cofounder of Pipeguard Robotics. “Wu’s company, Pipeguard Robotics, manufactures a shuttlecock-shaped robot that travels through water pipes to detect leaks.”

Jenny Xu ’19 (games), founder of JCSoft Inc. “Xu has released nine mobile games that have been downloaded over 3.5 million times, including Can You Escape Fate.”

Julia Yu ’10 (finance), investment analyst at Millennium Management. “Emerging markets trader with big role on a large macro team at billionaire Israel Englander’s $34 billion hedge fund.”

A version of this article originally appeared on the Slice of MIT blog.

back to newsletter

Jay London | MIT Alumni Association 
November 21, 2017
MIT News 

Tuesday, 28 November 2017 16:50

Turning emissions into fuel

MIT-developed method converts carbon dioxide into useful compounds.
MIT Fuel From CO2 Wu Web
XiaoYu Wu pictured with the reactor his team used for the research. MIT researchers have developed a new system that could potentially be used for converting power plant emissions of carbon dioxide into useful fuels. The method may not only cut greenhouse emissions; it could also produce another potential revenue stream to help defray its costs. Image, Tony Pulsone

MIT researchers have developed a new system that could potentially be used for converting power plant emissions of carbon dioxide into useful fuels for cars, trucks, and planes, as well as into chemical feedstocks for a wide variety of products.

The new membrane-based system was developed by MIT postdoc Xiao-Yu Wu and Ahmed Ghoniem, the Ronald C. Crane Professor of Mechanical Engineering, and is described in a paper in the journal ChemSusChem. The membrane, made of a compound of lanthanum, calcium, and iron oxide, allows oxygen from a stream of carbon dioxide to migrate through to the other side, leaving carbon monoxide behind. Other compounds, known as mixed ionic electronic conductors, are also under consideration in their lab for use in multiple applications including oxygen and hydrogen production.

Carbon monoxide produced during this process can be used as a fuel by itself or combined with hydrogen and/or water to make many other liquid hydrocarbon fuels as well as chemicals including methanol (used as an automotive fuel), syngas, and so on. Ghoniem’s lab is exploring some of these options. This process could become part of the suite of technologies known as carbon capture, utilization, and storage, or CCUS, which if applied to electicity production could reduce the impact of fossil fuel use on global warming.

The membrane, with a structure known as perovskite, is “100 percent selective for oxygen,” allowing only those atoms to pass, Wu explains. The separation is driven by temperatures of up to 990 degrees Celsius, and the key to making the process work is to keep the oxygen that separates from carbon dioxide flowing through the membrane until it reaches the other side. This could be done by creating a vacuum on side of the membrane opposite the carbon dioxide stream, but that would require a lot of energy to maintain.

In place of a vacuum, the researchers use a stream of fuel such as hydrogen or methane. These materials are so readily oxidized that they will actually draw the oxygen atoms through the membrane without requiring a pressure difference. The membrane also prevents the oxygen from migrating back and recombining with the carbon monoxide, to form carbon dioxide all over again. Ultimately, and depending on the application, a combination of some vaccum and some fuel can be used to reduce the energy required to drive the process and produce a useful product.

The energy input needed to keep the process going, Wu says, is heat, which could be provided by solar energy or by waste heat, some of which could come from the power plant itself and some from other sources. Essentially, the process makes it possible to store that heat in chemical form, for use whenever it’s needed. Chemical energy storage has very high energy density — the amount of energy stored for a given weight of material — as compared to many other storage forms.

At this point, Wu says, he and Ghoniem have demonstrated that the process works. Ongoing research is examining how to increase the oxygen flow rates across the membrane, perhaps by changing the material used to build the membrane, changing the geometry of the surfaces, or adding catalyst materials on the surfaces. The researchers are also working on integrating the membrane into working reactors and coupling the reactor with the fuel production system. They are examining how this method could be scaled up and how it compares to other approaches to capturing and converting carbon dioxide emissions, in terms of both costs and effects on overall power plant operations.

In a natural gas power plant that Ghoniem’s group and others have worked on previously, Wu says the incoming natural gas could be split into two streams, one that would be burned to generate electricity while producing a pure stream of carbon dioxide, while the other stream would go to the fuel side of the new membrane system, providing the oxygen-reacting fuel source. That stream would produce a second output from the plant, a mixture of hydrogen and carbon monoxide known as syngas, which is a widely used industrial fuel and feedstock. The syngas can also be added to the existing natural gas distribution network.

The method may thus not only cut greenhouse emissions; it could also produce another potential revenue stream to help defray its costs.

The process can work with any level of carbon dioxide concentration, Wu says — they have tested it all the way from 2 percent to 99 percent — but the higher the concentration, the more efficient the process is. So, it is well-suited to the concentrated output stream from conventional fossil-fuel-burning power plants or those designed for carbon capture such as oxy-combustion plants.

“It is important to use carbon dioxide to produce carbon monoxide for the conversion of sustainable thermal energies to chemical energy,” says Xuefeng Zhu, a professor of chemical physics at the Chinese Academy of Sciences, in Dalian, China, who was not involved in this work. “Using an oxygen-permeable membrane can significantly reduce the reaction temperature, from 1,500 C to less than 1,000 C, indicating a great energy saving compared to the traditional carbon dioxide decomposition process,” he says. “I think their work is important to the field of sustainable energy and membrane processes.”

The research was funded by Shell Oil and the King Abdullah University of Science and Technology.

back to newsletter

David L. Chandler | MIT News Office
November 28, 2017

Tuesday, 28 November 2017 12:33

How to get sprayed metal coatings to stick

When spraying metal coatings, melting hurts rather than helps, MIT research reveals.
MIT Melting Bonding 01
Micrographs of a metal surface after impact by metal particles. Craters are formed due to melting of the surface from the impact. Courtesy of the researchers

When bonding two pieces of metal, either the metals must melt a bit where they meet or some molten metal must be introduced between the pieces. A solid bond then forms when the metal solidifies again. But researchers at MIT have found that in some situations, melting can actually inhibit metal bonding rather than promote it. 

The surprising and counterintuitive finding could have serious implications for the design of certain coating processes or for 3-D printing, which both require getting materials to stick together and stay that way. The research, carried out by postdocs Mostafa Hassani-Gangaraj and David Veysset and professors Keith Nelson and Christopher Schuh, was reported in two papers, in the journals Physical Review Letters and Scripta Materialia.

Schuh, who is the Danae and Vasilis Salapatas Professor of Metallurgy and head of the Department of Materials Science and Engineering, explains that one of the papers outlines “a revolutionary advance in the technology” for observing extremely high-speed interactions, while the other makes use of that high-speed imaging to reveal that melting induced by impacting particles of metal can impede bonding.

The optical setup, with a high-speed camera that uses 16 separate charged-coupled device (CCD) imaging chips and can record images in just 3 nanoseconds, was primarily developed by Veysset. The camera is so fast that it can track individual particles being sprayed onto a surface at supersonic velocities, a feat that was previously not possible. The team used this camera, which can shoot up to 300 million frames per second, to observe a spray-painting-like process similar to ones used to apply a metallic coating to surfaces in many industries.

While such processes are widely used, until now their characteristics have been determined empirically, since the process itself is so fast “you can’t see it, you can’t tell what’s happening, and no one has ever been able to watch the moment when a particle impacts and sticks,” Schuh says. As a result, there has been ongoing controversy about whether the metal particles actually melt as they strike the surface to be coated. The new technology means that now the researchers “can watch what’s happening, can study it, and can do science,” he says.

The new images make it clear that under some conditions, the particles of metal being sprayed at a surface really do melt the surface — and that, unexpectedly, prevents them from sticking. The researchers found that the particles bounce away in much less time than it takes for the surface to resolidify, so they leave the surface that is still molten.

If engineers find that a coating material isn’t bonding well, they may be inclined to increase the spray velocity or temperature in order to increase the chances of melting. However, the new results show the opposite: Melting should be avoided.

MIT Melting Bonding 02
The top row of photos shows a particle that melts the surface on impact and bounces away without sticking. The bottom row shows a similar particle that does not melt and does stick to the surface. Arrows show impact sprays that look like liquid, but are actually solid particles. Courtesy of the researchers

It turns out the best bonding happens when the impacting particles and impacted surfaces remain in a solid state but “splash” outward in a way that looks like liquid. It was “an eye-opening observation,” according to Schuh. That phenomenon “is found in a variety of these metal-processing methods,” he says. Now, it is clear that “to stick metal to metal, we need to make a splash without liquid. A solid splash sticks, and a liquid one doesn’t.” With the new ability to observe the process, Hassani-Gangaraj says, “by precise measurements, we could find the conditions needed to induce that bond.”

The findings could be relevant for processes used to coat engine components in order to reuse worn parts rather than relegating them to the scrap-metal bin. “With an old engine from a large earth-moving machine, it costs a fortune to throw it away, and it costs a fortune to melt and recast it,” Schuh says. “Instead, you can clean it off and use a spray process to renew the surface.” But that requires that the sprayed coating will remain securely bonded.

In addition to coatings, the new information could also help in the design of some metal-based additive manufacturing systems, known as 3-D printing. There, as with coatings, it is critical to make sure that one layer of the printing material adheres solidly to the previous layer.

“What this work promises is an accurate and mathematical approach” to determining the optimal conditions to ensure a solid bond, Schuh says. “It’s mathematical rather than empirical.”

The work was supported by the U.S. Army through MIT’s Institute for Soldier Nanotechnologies, the U.S. Army Research Office, and the U.S. Office of Naval Research.

back to newsletter

David L. Chandler | MIT News Office
November 22, 2017


Wednesday, 01 November 2017 17:55

Poster Highlights

Materials Day poster presenters give two-minute introductions to their research during annual symposium.

Materials Day Poster Session presenters capped off the annual Materials Day Symposium with brief highlights of research ranging from solar energy and alternative fuels to spinal cord injury and neural networks for artificial intelligence.

Postdoc Grace Han, in Prof. Jeffrey Grossman’s group, Department of Materials Science and Engineering, described progress in creating materials which absorb photons from sunlight and convert them into heat energy through the charging and discharging cycle of organic photo switching molecules. “This is quite different from just heating water or concrete block by solar radiation in that we can actually store the energy and release energy by triggering,” Han said. These organic coatings can be integrated onto car windshields for deicing, fabrics for personal heating, or building materials for temperature control. Han’s poster also described a new process to harness waste heat from industrial furnaces, and store it for later release.

Janille Maragh, a graduate student in Professor Admir Masic’s lab, Department of Civil and Environmental Engineering, presented her work on sustainable construction materials. To study ancient Roman concrete from an archaeological site in Italy, she used Energy Dispersion Spectroscopy and Raman spectroscopy to map centimeter scale samples at microscopic resolution. “What we are trying to do is understand exactly what our sample is made of so can we understand this phenomenal material. … So we understand not only the bulk composition of our material but also their fracture surface.”

“Carbon monoxide is responsible for more than half of all fatal poisonings worldwide,” Vera Schroeder, a graduate student in Professor Timothy Swager’s lab, said. “Exposure to this odorless, colorless and tasteless gas is very difficult to detect for humans, which is compounded by the fact that the initial symptoms of poisoning – headache, dizziness, and confusion are non-specific.” Schroeder is developing bio-inspired carbon monoxide sensors that use a transistor-based design to activate a chemical change in iron atoms to detect carbon monoxide, even in air. “This new mode of sensor allows us to have a voltage activated, enhanced and highly specific response and we can detect carbon monoxide in air with much higher sensitivity than we detect CO2, oxygen or water,” she said.


Alfonso Juan Carrillo, a postdoc in the lab of Jennifer L. M. Rupp, the Thomas Lord Assistant Professor of Materials Science and Engineering, presented results of work on perovskite materials for solar-driven transformation of CO2 and water into fuels. Carillo selected the best candidate perovskite materials, synthesized these perovskites, analyzed their microstructure, and tested them in a fixed bed high-temperature reactor. “We use what are called thermochemical cycles,” Carillo said. As the perovskite absorbs oxygen, it can transform water and carbon dioxide into hydrogen and carbon monoxide.

Minghui Wang, a postdoc in Professor Karen Gleason’s lab, is creating thin-film microporous polymers for gas separation using chemical vapor deposition. Gas separation is important for industrial gas needs and carbon capture but heat-based methods are energy intensive,” he said. “One challenge is that you need to achieve both high flux and high gas selectivity for membrane materials. To do so, usually you need a rigid and microporous structure and also you need to fabricate very thin films, but to do both of them is kind of difficult. In our lab, we use chemical vapor deposition to deposit pinhole-free thin films by using this technique and using porphyrin as a monomer.” He achieved high selectivity for carbon dioxide and nitrogen separation using polymerized porphyrin on a flexible substrate.

Andrew Dane, a graduate student in Professor Karl Berggren’s group, discussed progress on improving speed and efficiency in superconducting nanowire single photon detectors. Two competing available materials tilt toward either speed or efficiency. “We changed the material deposition and made some devices and showed that we kind of combined the best of both worlds,” Dane said. “There is a quantum phase transition in the material that we’re working with and a lot of other interesting things.”

About a million Americans undergo hernia repair surgery each year and for one in four or them, hernia will re-occur. About half will experience some degree of chronic pain, said Sebastian Pattinson, a postdoc in the lab of Associate Professor of Mechanical Engineering A. John Hart. The surgical mesh used to mechanically reinforce the tissue as it heals causes many of these complications. Pattinson described a new 3D printing process that allows local customization of mechanical response in a surgical mesh and in particular allows for non-linear mechanical response in a way that mimics tissue. “We hope that these meshes will help alleviate the complications suffered by many patients all around the world,” Pattinson said.

Chemistry postdoc Zhou Lin, in Professor Troy A. Voorhis’ group, presented research on a process to double electric current in organic solar cells by splitting single excitons into pairs, a process that is called singlet fission. “We can generate two electric currents out of one high-energy photon so we can promote the efficiency of organic photovoltaics, that’s what we want,” Lin said. “Based on our electronic structure theory calculations, we are able to reproduce the experimental trend for the fission rate using three different isomers that can undergo this intramolecular singlet fission,” she said.

Yukio Cho, a graduate student in Prof. Harry L. Tuller’s lab, is working on mixed ionic and electronic conductor [MIEC] cathode materials for solid oxide fuel cells. Using electrochemical methods, Cho and colleagues synthesized n-type cathode material to improve the surface exchange. “We control the defects, control the electronic defects, and for the current result, we successfully synthesized n-type materials,” Cho said. “The expected good surface exchange capability is also confirmed through transfer diffusion measurements.”

Frank McGrogan, a graduate student in Professor Krystyn J. Van Vliet’s lab, presented his work with the Chemomechanics of Far-From-Equilibrium Interfaces [COFFEI] group on all-solid electrolytes in lithium ion battery systems. “One of the main sticking points is we have this problem of lithium metal unevenly plating the electrodes, crossing the electrolyte and shorting the cell. Our group has been treating this as a fracture issue. … We’ve demonstrated experimentally that fracture is indeed a mechanism for this lithium plating and shorting problem.”

“We’ve gone ahead and measured some mechanical properties including fracture properties of several important solid electrolytes and used these inputs in simulations to predict damage evolution,” McGrogan said. “I think that the way that our group has approached this problem and how we’re getting to the mechanism is going to change the way our field thinks about failure in all solid-state lithium ion batteries.”

Postdoc Dena Shahriari, who works with Professors Yoel Fink and Polina Anikeeva, shared an update on efforts to repair spinal cord damage by optically stimulating and guiding the growth of injured neurons. “We’re using a thermal drawing process, which is a high throughput technique which will allow us to create kilometer-long fibers in just one experiment,” Shahriari said. These highly flexible probes deliver light to the lesion of the spinal cord, and record at multiple sites of these neurons.

“For the tissue engineering part we needed to bridge the nerve gap, we needed to create porosity into these scaffolds, and for that we need to add a twist to this thermal drawing process that will allow us to not only create, but also control, porosity in that,” Shahriari said.

Gerald Wang, a graduate student in mechanical engineering under Prof. Nicolas Hadjiconstantinou, invited attendees to learn more about his poster by arranging it so that the first letters of each line spelled out “C-O-M-E.” He is exploring the fluid-solid interface atop a sheet of graphene. “It turns out when you put fluid in this environment under the right conditions, it will spontaneously arrange into a layered structure that mimics the solid below it. This layered fluid structure, practically indistinguishable from tiramisu or the layered cake of your preference, imparts upon the fluid remarkable fluid properties including enhanced heat transfer, remarkably long slip lengths and highly modified surface diffusivities very different from the bulk fluid.”

“It’s a very exciting story with some of the great actors and actresses of today including Van der Waals, high through-put simulation and molecular self-assembly. So there’s something for everybody whether you’re an experimentalist, a theorist, a computationalist, or you just like a good scaling relation, you should make like the letters and come on by,” Wang said.

Mary Elizabeth Wagner, a graduate student in the group of Associate Professor of Metallurgy Antoine Allanore, is working on a sustainable way to refine precious metals from nature and from recycled materials. “The problem is these expensive elements, silver, gold, platinum, are found in very, very tiny amounts, comparatively to copper, but they make up so much of the cost,” Wagner said.

“My idea in my research focuses on one system that can host electrochemistry for gold, for silver, and for platinum group metals,” Wagner said. Molten sulfide electrolytes are one promising system. “We should be able to treat all of these metals in one go, which should be able to provide an environmentally sustainable as well as a cost-effective way to treat these metals,” she said.

Vrindaa Somjit, a graduate student under Prof. Bilge Yildiz, is examining the effect of dopants on hydrogen solubility in alumina using a computational, first principles approach. Hydrogen may become a fuel of the future, but one of the main problems in making this a reality is the storage and transport of hydrogen. Hydrogen can penetrate steel and cause it to fail.

“One way to mitigate this problem of hydrogen embrittlement is by the use of permeation barrier coatings, and alpha-alumina is a promising candidate,” Somjit said. She set out to determine if dopants, extra chemical elements added to a compound, could improve the performance of alpha-alumina in resisting hydrogen penetration. “What we found is that actually dopants do not help in decreasing the hydrogen solubility because alpha-alumina itself lies at the bottom of the hydrogen solubility valley,” Somjit said.

Graduate student Chang Sub Kim, in Professor Harry Tuller’s group, conducts research to electrochemically pump oxygen in and out of a thin film of layered cuprate, which has potential as a cathode material. “An interesting fact is that it can accommodate both oxygen vacancies as well as interstitials. So in this study, I show you that I can control the region where I can access oxygen-access and also oxygen-deficient regions, and then show that I can simultaneously measure different materials properties such as oxygen surface reaction kinetics as well as in-plane conductivity, which agrees very well with the expected defect chemistry.”

Postdoc Yuming Chen in Professor Ju Li’s group, spoke about a project to develop a sodium-ion battery anode using nitrogen-doped carbon. Chen introduces nitrogen atoms into the structure of hollow carbon tubes to create larger spacing that allows sodium to penetrate the carbon tube and yield higher performance. These carbon tubes can be used as freestanding electrodes with long cycling life.

Ananya Balakrishna, a postdoc in Professor W. Craig Carter’s group, developed theoretical and computational models to investigate the link between material properties and microstructure. “In my research, I probe questions like what determines microstructural patterns, can we engineer microstructures to control macroscopic material properties,” she said. Her poster featured two projects describing microstructure in ferroelectric materials and in lithium battery electrodes.

“In lithium batteries, microstructures form during a typical charge/discharge cycle. In these microstructures, the underlying lattice symmetry has an effect on material properties, for example, certain lattice arrangements facilitate the faster propagation of diffusion of lithium ions and certain lattice arrangements cause non-uniform expansion of electrodes,” Balakrishna said. She is working on a phase field crystal model that couples lattice symmetry with the concentration field to describe electrode microstructure.

Menghsuan Sam Pan, a graduate student in Professor Yet-Ming Chiang’s group, focuses on using water-based sulfur batteries for low-cost energy storage. “It’s one of the lowest cost per stored charge in any electrochemically active materials, only behind water and oxygen,” Pan said. “When we work in soluble electrodes, we found that the sulfur can only be reversibly cycled between a di-sulfide and a tetra-sulfide regime, and with this we did some technical economic modeling to see the installed costs of the electrode. What surprised us is that the component that’s used to hold the electrode is more costly than the active material itself.”

Experiments showed these sulfide species cycle reversibly, precipitating into the electrode and then dissolving very well when they are cycled back, Pan said. “We cycled for more than 1,600 hours, more than two months,” he said, noting a 30 percent cost reduction in terms of cost per stored capacity.

Working under Professor Jeehwan Kim, graduate student Scott Tan is developing hardware for neural networks for artificial intelligence. He makes silicon-germanium cross-bar arrays with a reversible silver conductance channel to toggle the conductance state of these synaptic devices. “We’ve also used these devices in a simulation and showed that they can perform handwriting recognition with accuracy up to 95 percent,” Tan said.

Mechanical engineering graduate student Nicholas T. Dee presented work in Professor A. John Hart’s group on scalable roll-to-roll graphene production for membrane applications. “We’ve developed a roll-to-roll CVD reactor for this process that is unique in that it has two different zones, one specifically for annealing the substrate and catalyst and one zone for growth of the graphene,” Dee said. The researchers tuned the gas composition to achieve full coverage of monolayer graphene and explored the tradeoff between production rate and quality of the graphene. “We have demonstrated using our graphene produced in this high-throughput manner to produce nano-porous, atomically thin membranes for potential desalination applications,” he said.

Brad R. Nakanishi, a graduate student in Professor Antoine Allanore’s group, introduced his research on high-temperature materials chemistry in refractory metals. “What we’ve done, where experiment by conventional methods or prediction by first principles prove very complex and challenging, we’ve basically modified a floating zone furnace which has provided us with enhanced experimental throughput and also very unique ability to see and probe the properties of these refractory liquids,” Nakanishi said. His poster showed an image of the first direct electrolytic decomposition of aluminum oxide to oxygen gas and aluminum metal. “We’ve been using this approach to make fundamental thermodynamic property measurements like chemical potential,” he said. This work has implications for discovery of new materials for applications from aerospace to nuclear as well as discovery of new processes for materials extraction.

Chosen by guests who attended the Materials Day Poster Session, this year's Poster Session prize winners were Postdoc Dena Shahriari, electrical engineering and computer science; graduate student Vera Schroeder, chemistry; and Postdoc Sebastian Pattinson, mechanical engineering.

The annual MIT Materials Research Laboratory [MRL] Materials Day Symposium and Poster Session were held on Wednesday, Oct. 11, 2017.

back to newsletter

Related: A magical dimension

Interdisciplinary materials science model offers key to progress



Wednesday, 01 November 2017 17:43

A magical dimension

Engineering at the nanoscale opens new doors to control optical, electronic and magnetic behaviors of materials and enable new multi-functional devices

Materials Day Panel 9584 DP Web
MIT MRL External Advisory Board Chair Julia Phillips [far left] moderated the Materials Day Symposium panel on “Frontiers in Materials Research.” She was joined by [from second left] Professors Karen Gleason, Caroline Ross, Timothy Swager, and Vladimir Bulović. The session was held Wednesday, Oct. 11, 2017.

Newly discovered optical, electronic and magnetic behaviors at the nanoscale, multifunctional devices that integrate with living systems, and the predictive power of machine learning are driving innovations in materials science, a panel of MIT professors told the MIT Materials Research Laboratory [MRL] Materials Day Symposium.

“The development of new material sets is a key to the launch of new physical technologies,” Professor Vladimir Bulović, founding director of MIT.nano, said. “Once we get down to the nanoscale, we can start inducing quantum phenomena that were never quite accessible. So that scale between 1 nanometer, the typical size of a molecule, and on the order of, let’s say, 20 nanometers, that’s a magical dimension, where you can fine tune your optical, electronic and magnetic properties.”

Professor Caroline Ross, Associate Head of the Department of Materials Science and Engineering, cited a trend of harnessing nature to self assemble complex structures. “As we want to make things smaller and smaller, we need to have nature helping out,” she said. Ross noted progress on a range of new multi-functional materials that use, for example, extremely low voltage levels  to control magnetism or that use strain to control electronic properties. “All of these can enable new kinds of devices from those materials, so you can imagine devices which are smart that can have memory or logic functions, that can have analog instead of just digital type of behavior, that can work together to make smart circuits. … The difficulties of integrating those types of materials will be well paid for by the new sorts of functionality we can get from the devices we make.”

MIT MRL External Advisory Board Chair Julia Phillips moderated the Materials Day Symposium panel on Wednesday, Oct. 11, 2017. Phillips is a former Sandia National Laboratories executive.

Professor Timothy Swager, Director of the Deshpande Center, said the expectation that new medical devices, for example, are compatible with our bodies demands different requirements than previous generations of electronics. “Thinking about how we interface complex dynamic chemically reactive systems with a material is really a very important area that, I think, will continue to be of importance and many good discoveries are going to come about as result of the interest in that area,” he said.

Associate Provost and Professor Karen Gleason spoke of the growing influence of machine learning on materials advances and the potential for one-dimensional and two-dimensional materials to provide better computers and memory storage. “It’s going be incredible for materials discovery as we learn how to use machine learning to predict what materials are optimal, but there’s also a credible place for materials in making this technology grow. Now computational power and memory and databases have gotten large enough that the predictive power is actually great.”

“The biggest component is you need the data so you need all of these sensors for accurate positioning, for detection of gases, for health. People want wearables,” Gleason said. “So I think this is an enormous field with tremendous impact in many different ways that materials can play.”

Bulović said while it takes a lot of perseverance to implement a new idea on the nanoscale, “It’s important to highlight that the invention of an idea happens in a moment, that eureka moment, but to actually scale that idea up so a million people can hold it in their hands, that takes a decade sometimes, especially if it’s in the materials space. Recognition of that is important in order to support the evolution of the new ideas.”

The annual Materials Day Symposium was hosted for the first time by the MIT Materials Research Laboratory, which formed from the merger of the Materials Processing Center and the Center for Materials Science and Engineering, effective Oct. 1, 2017. The MIT MRL will work hand-in-hand with MIT.nano, the central research facility being built in the heart of the MIT campus due to open in June 2018. MIT will receive a $2.5 million gift from the Arnold and Mabel Beckman Foundation to help develop a state-of-the-art cryo-electron microscopy (cryo-EM) center to be housed at the MIT.nano facility.

“I don’t think we can underestimate the value of the tool sets in providing us the direction to what we need to do to advance life as we know it,” Bulović said. “I get struck by the example of DNA … It took 80-plus years to obtain the first inkling that there was something twisted inside our cells. Then we debated for another decade, is this thing really a twisted molecule inside our cells. If you add it all up, 80, 90 years of debate. Today that’s reduced to a couple of hours of work by one graduate student who can take a cell, pull out a nucleus, put it under a scanning tunneling microscope or cryo electron microscope and see a twisted molecule we call DNA now.”

Swager noted that biologists also will use MIT.nano. “They are going to be using the cryo-EM in the basement, so nano is not only for engineers and molecule builders. … I think that’s going to be really exciting and where that fusion leads us, who knows.”

Moderator Phillips asked the panelists what tool sets that would like to see in MIT.nano. Gleason said she would like to see chemical vapor deposition for thin polymer films. Ross said that MIT needs to be at the forefront for materials characterization tools. “We need to have the best tools to do the best work,” Ross said. She would like to see MIT.nano get the best possible electron microscope and advanced deposition tools for oxide molecular beam epitaxy and building up complex materials layer by layer. Swager said it is important for the shared facility to house tools for rapid prototyping and fabrication of devices.back to newsletter

Denis Paiste, Materials Research Laboratory
November 27, 2017

Related: Poster Highlights

Interdisciplinary materials science model offers key to progress

Page 1 of 2