MIT researchers have created a new combination of techniques that can offer thorough information regarding the microstructure of polycrystalline metals.
Such materials contain a random matrix of numerous minute crystals rather than a sole big crystal. These materials are generally used for applications in aircraft, nuclear reactors, and civil infrastructure. Nevertheless comprehending their crystal structure details and the boundaries among the crystal areas is difficult.
Their discoveries are reported in the Nature Computational Materials journal, as a research paper by Matteo Seita, an MIT postdoc; Michael Demkowicz, Professor of materials science and engineering; Christopher Schuh, the Danae and Vasilis Salapatas Professor of Metallurgy, and five other researchers.
“This is a unique combination of different technologies,” Seita elucidates. The novel approach he and his colleagues developed deals with “one of the most common problems in materials science: How do we quantify the characteristics of materials in a high-throughput fashion?”
Certain techniques offer extensive information about structures, but they require time to execute and cannot disclose fast changes within the material. Others work quickly but offer very little structural detail and a few other methods offer both temporal and spatial detail but are extremely costly or available in certain places only. This new grouping of techniques could help in resolving these limitations by offering rapid, high-resolution, and cheap imaging of the materials, says Seita.
In polycrystalline metals, which contain numerous small crystal grains, it is essential to know the dimensions, location, angles of contact, and other features of the various grains cmaking up the material. Especially, the interfaces among the crystal grains, known as grain boundaries, “happen to be critical,” Seita states, “to many individual properties of the material — its strength, radiation tolerance, hardness, electrical resistance, and so on — but they are very difficult to characterize experimentally, because they are very complex.”
Scientists wish to quantify five fundamental characteristics about these grain boundaries, but the majority of the tools for studying the materials can only give up some subset of two or three of those. A way to get all five characteristics immediately is high-energy synchrotron radiation, which is available in some facilities that are costly and are likely to be oversubscribed.
Our solution was to try to create a very simple technology that can be used by anyone, in their own lab, using software and hardware that are easily available.
And that’s what they achieved, by using a blend of two existing methods, the optical microscopy and the electron microscopy.
“We take two different datasets and combine them using our numerical image analysis,” Seita explains. The researchers used sheets of polycrystalline metal foil, which were adequately thin that each grain could be seen from both sides. The researchers then took optical microscopic images from one side of the foil, turned the sheet over, and imaged the next side, and with the help of software connected the grain boundaries from one side to the other side. From that, Seita says, “We can reconstruct the 3-D orientation of these grain boundaries.”
The data is then combined with electron microscopic images that explain the real pattern of atoms inside the grains, depicting the orientation of the individual crystal lattices inside every grain and how they correlate to those of neighboring grains. This collective data provides all five characteristics of the grain boundaries in the metal foils.
“The beauty of this is that it’s high-throughput technology,” Seita explains. “On one sample, we can measure up to 500 grain boundaries or so, and can build up datasets rapidly. And it’s nondestructive,” unlike current methods that use the sample in the process. Thus the sample can then undergo other tests, like tests of electrical or mechanical properties, whose outcomes can be correlated with the information of grain boundaries.
According to Seita, the new method “is very versatile, so many groups out there can use it.” Though the early tests were conducted using polycrystalline metals, the technique “is materials agnostic,” and is applicable to semiconductors or insulators and metals. “We can test for different kinds of properties and build up large datasets,” he states, and eventually utilize that information to predict the features of new polycrystalline materials.
“We can figure out what kind of grain boundaries we want to have” for a material designed for a specific application, “and figure out how to make a material with those grain boundaries.” Altering the features of these grain boundaries, by changing the material to enhance their abundance or relative orientations, could lead to noteworthy changes in the property of the material. This technique may be used to decipher how to decrease the rate of corrosion of metals which are exposed to harsh environments, like gas or oil drilling tools, Seita says.
This study is “an inspiring step forward in rapid, data-rich characterization of the structure of crystalline materials,” comments Brad Boyce, a distinguished associate of the technical staff at Sandia National Laboratories in New Mexico, who was not part of this research. “Grain boundaries, which are interfacial disruptions in the crystalline lattice of polycrystals, influence a wide range of material phenomena ranging from how the material deforms to the electrical resistivity … yet materials scientists possess a limited range of techniques to explore the grain boundary character.”
Using this novel method, Boyce states, “I am excited to see how this work will inspire further developments that provide rapid, high-throughput characterization, especially techniques that can be used to decipher local grain boundary character below the spatial resolution limits of optical microscopy.”
The other members of the research group are Marco Volpi and Maria Vittora Diamanti of the Polytechnic University of Milan, Italy; Srikanth Patala at North Carolina State University; and Ian McCue and Jonah Erlebacher at Johns Hopkins University, Baltimore. The study was supported by the MISTI Seed Fund, the National Science Foundation, and the U.S. Department of Energy.