State-of-the-art experiments reveal “hidden” details when transforming materials that affect faster microelectronics

Phase change is at the heart of the world around us. Perhaps the most familiar example is when ice melts in water or boils into steam in water, but the phase change is also underlying heating systems and digital memory, such as those used in smartphones.

Some materials triggered by light or electrical pulses can switch between two different phases representing binary codes 0 and 1S to store information. Understanding how a material is converted from one state or phase to another is key to adjusting a material using specific properties that can increase switching speeds or operate at low energy costs, for example.

However, researchers were unable to directly visualize how these transformations unfold in real time. We often assume that the materials are perfect and look the same everywhere, but “part of the task is that different parts of the material are often heterogeneous, with different parts of the material changing in different ways, including scales and timescales of different lengths,” said co-author Aaron Lindenberg, professor at SLAC and Stanford University.

In the new study, researchers and colleagues at the Department of Energy’s SLAC National Accelerator Laboratory used a new technique called X-ray photon correlation spectroscopy (XPCS) on a Linac coherent light source (LCL) to probe the phase transformation with the complex structure of lead-tau and strontium layers.

“The material is equipped between multiple states and forms a kind of frustrating system that is highly reconfigurable at the atomic and nanoscale,” Lindenberg said. “A single pulse of light switches it to another crystal structure known as a “supercrystalline.” ”

What they found challenged traditional wisdom: the time that the conversion process took place lasted up to 100,000 times longer than previously thought. They also found that transformations progress non-uniformly and can link the movement of boundaries between phases to much longer transformation timescales.

Research teams from SLAC, Stanford University, Penn State University, Argonne National Institute, Lawrence Berkeley National Institute, Indian Institute of Science, University of California, Berkeley and Rice University have published their findings on the minutes of the National Academy of Sciences.

The team initially set out to explore the possibility of using XPC to study the possibility of studying in greater detail how stages move from one to the other. This technique shoots a single pulse of visible light, called a pump, to trigger a transition and track the conversion with subsequent X-ray pulses, called a probe, at various delay time intervals.

A pattern from X-ray light scattered by a sample (speckle pattern) is a condyst in a seemingly random region of bright paxels that encode information about how the sample is converted. The researchers then repeated this “pump probe” method in other areas around the sample. Compared to previous studies, XPCS draws a more detailed picture (time and space) of how phase transitions unfold.

Using XPCS, the team was able to investigate the phase transitions of the material on a length scale from the level of individual atoms to the approximate diameter of human hair.

I expect the conversion to complete in hundreds of nanoseconds. Nanoseconds are one billionth of a second. Instead, the team discovers that at that point only half of it has been done, and the complete process takes dozens of times and nearly 100,000 times more.

“This is similar to Google Maps, which says you’ve reached your destination. In fact, you’re halfway through there and the other half takes a lot longer than you’d previously thought,” said Venkatraman Gopalan, co-author and professor at Pennsylvania State University.

It also reveals new details about how phases and their boundaries form, grow and interact with each other, and are associated with the heterogeneity of transformations. According to Gopalan, these uneven processes are like unexpected traffic jams, and it takes time to slow you down and clear.

“Overall, our findings show how these materials systems are complex and involve movements that correlate at different phase boundaries at multiple time and length scales,” Lindenberg said. “Through this approach, we can visualize and reveal details that were blinded when using traditional approaches.”

Successful use of XPC opens the door to investigate similar processes in other material systems and devices. “This advanced technique provides a way to investigate and visualize heterogeneous and dynamic processes,” Lindenberg said. “We can apply knowledge gained to technically relevant goals, such as designing materials for information storage technologies that operate with faster switching devices and lower energy costs.”