A Microscope as a Shovel? UConn Researchers Dig It

excerpted from UConn Today by
Kim Krieger – UConn Communications

Bryan Huey’s lab used the tip of an Atomic Force Microscope (AFM) as a chisel to scrape away the surface of bismuth ferrite and map the electric landscape of the interior. (Image courtesy of the Huey Lab)

Using a familiar tool in a way it was never intended to be used can open up a whole new method to explore materials, report UConn researchers in the Proceedings of the National Academies of Science. Their specific findings could someday create more energy efficient computer chips. But more broadly, their approach should spur scientists worldwide into trying to use this new approach for a wide range of other materials and eventual applications.

Image courtesy of Jack Dumala, Daniel Schwartz, UConn SquaredLabs
Image courtesy of Jack Dumala, UConn SquaredLabs

The research is based on Atomic force microscopes (AFM), which materials scientists and other researchers use to carefully trace an ultra sharp tip across the surface of all kinds of materials. The tip can ‘feel’ where the surface is, and sometimes can also sense properties like electric and magnetic forces emanating from the material. Then, in the same way a farmer methodically drives a plow back and forth or up and down a rolling field, an AFM can scan the hills and valleys at the surface of a material, developing maps of its holes and protrusions, and even its properties, all at length scales a thousand times smaller than a grain of salt.

Unlike the farmer’s plow, AFMs are generally designed to barely touch the surface in order to prevent damage to the sample (churning up the field). But sometimes it happens anyway.

A few years ago, Yasemin Kutes and Justin Luria, graduate student and postdoc members of UConn materials scientist Bryan Huey’s lab, dug into solar cells they were studying. At first thinking this was an irritating mistake, they noticed that the properties of the material looked different from pictures of the original surface alone. That wasn’t too surprising—for materials used in real-world applications, often the surface is actually engineered to have different properties. Yet before, there had simply been no way to measure such underlying properties with the resolution offered by AFM.

In fact, in the 30 years since AFMs were invented, only a handful of groups worldwide have reported such measurements. This was usually either to finely shape a surface, or to map where electricity flows in a part of a computer chip or in a solar cell like at UConn. But another graduate student in Huey’s group, James Steffes, was inspired to take advantage of this discovery for an entirely different class of materials and materials properties. Could he intentionally use the tip of an AFM like the farmer’s plow, progressively digging deeper into the material, and at the same time map the electrical or magnetic properties for deeper and deeper layers of a ‘functional ceramic?’

The answers, as Steffes, Huey, and their colleagues report in the highly competitive journal PNAS, are yes and yes. To demonstrate the approach, they dug into a sample of bismuth ferrite (BiFeO3), which is a room temperature multiferroic provided by project collaborator Ramamoorthy Ramesh of UC Berkeley. Multiferroics are materials that support both electric and magnetic properties at the same time. For example, “BFO” is antiferromagnetic—it responds to magnetic fields, but overall does not exhibit a North or South magnetic pole—and ferroelectric, meaning it has switchable electric polarization. Such ferroelectrics usually comprise tiny ‘domains’ that all have similarly oriented electric fields. Think of a whole bunch of tiny batteries, clusters of which are aligned with their positive terminals pointing in one direction, alongside other clusters pointing another direction. These are very valuable for computer memory, because the computer can flip the domains, ‘writing’ data into the surrounding material. These domains can be fine enough to be serious contenders for replacing the enormous market of thumb drives and other solid state memory that is now in every smartphone, tablet, camera, and most computers.

But when a material scientist “reads” or “writes” such data in BFO, they can normally only see what happens on the surface. Yet they really need to know what lies beneath as well—if that is understood, it might be possible to engineer more efficient computer chips that run faster and use less energy than those available today. That’s a very important goal for society—already ~5% of all energy consumed in the US goes just to running computers.

domains in the bismuth ferrite changed in shape with depth
A new domain. The researchers were able to show how the domains (tiny sections of material with the same electric polarization) in the bismuth ferrite changed in shape with depth. (Image courtesy J. Steffes, B. Huey)

So Steffes, MSE Department Head Huey, and the rest of the team used an AFM tip to meticulously dig through a film of BFO and measure the interior piece by piece. They found they could map the individual domains all the way down, exposing patterns and properties which weren’t always apparent at the surface. Sometimes a domain narrowed with depth until it vanished, or split into a y-shape, or merged with another domain. No one had ever been able to see inside the material in this way before. It was revelatory, like looking at a 3-Dimensional CT scan of a bone for the first time, when you’d only been able to read 2-D x-ray films before.

“The systems we have in the IMS are special in many ways, including one we are now developing to advance Tomographic AFM even further thanks to a $1M grant from the National Science Foundation alongside support from UConn, the School of Engineering, and UConn. But worldwide there are something like 30,000 AFMs already installed. A big fraction of those are going to try Tomographic AFM in 2019 as our community realizes that we have literally just been scratching the surface all this time” predicts Huey. He also thinks more labs will buy AFMs if 3D mapping works for their materials, and some microscope manufacturers in this substantial high-tech industry will shift their focus to volumetric instead of surface scanning.

Steffes, who drove the project for his PhD research, has subsequently graduated from UConn with his PhD and is applying his skills and knowledge at computer chip maker GlobalFoundries. Researchers at Intel, muRata, and others are also intrigued with what the group discovered, as they seek new materials to extend computing and mobile devices beyond the current state of the art. Meanwhile, Huey’s current team of postdoc, graduate, and undergraduate researchers are continuing to use AFMs to dig into all kinds of materials, from concrete to bone to a host of other computer components. Huey says, “Working with academic and corporate partners, we can use our new insight to understand how to better engineer these materials to use less energy, optimize their performance, and improve their reliability and lifetime—those are examples of what Materials Scientists strive to do every day.”