Remember that diagram of “The Scientific Process” from every high school science textbook? I’ll jog your memory: data is collected that may conflict with a previously held assumption, so a new hypothesis is devised. Experiments are done, the new data is analyzed, the hypothesis appears to hold true, and a conclusion is made. Science! Only later do many of us learn that the course of science doesn’t always run so smoothly.
However, sometimes this model of the scientific process perfectly captures the trajectory of a project. Such was the case for Kathy Hoogeboom-Pot, a recent PhD graduate from the CU Boulder Physics department who published on the subject of nanoscale heat transport.
When Kathy joined the Kapteyn-Murnane Group in 2009, she became interested in the idea that equations describing the movement of heat from its source don’t hold true when the heat source gets very small. The ability to observe this phenomenon is relatively new, due mainly to recently developed techniques using sophisticated lasers.
Using such techniques, the physicists in the Kapteyn-Murnane Group were able to measure nanoscale heat transport from tiny heat sources into underlying substrates. They found exciting results: the trend seen at larger scales was completely reversed when heat sources got much smaller and much closer together.
“Nobody is surprised that nanoscale heat transfer is different from macroscale heat transfer, because everything changes when it gets really small,” says Kathy about this initial data. “But we’re observing it for the first time, and now we can start to understand how and why.”
After multiple confirmations that this process was actually occurring, the scientists began to think about how it could be explained. They considered the process of heat transfer in terms of “phonons”—the physical carriers of heat. Think of the movement of atoms in a given material. Atoms, like springs, like to move in regular, repeating patterns thanks to the laws of thermodynamics.
When heat is applied to a material (like a metal bar), the energy is carried by a wave of excited atoms moving together in repeating patterns. These waves of excited atoms are collectively known as phonons. Thus, phonons carry the excitation energy generated by a heat source through a material—remember that heat is just a form of energy, so we can describe the transfer in terms of movement, another form of energy.
Phonons from two heat sources next to each other probably don’t interact on a macroscale. But as the two sources get smaller and smaller, and the distance between them gets smaller, the phonons do begin to interact. This changes their behavior and leads to the surprising results of heat transport at nanoscale. In her preliminary results, Kathy began to see that mathematically, very small and proximal heat sources start to display features more characteristic of a continuous sheet: that is, one big heat source.
Kathy describes this change in thought as “obvious, once you think about it.” But if true, it leads to some very significant changes in the way physicists and engineers can describe heat sources mathematically on small scales, and the implications could be very important.
Kathy had preliminary data and a new hypothesis, so she was ready to start experimenting. To do this, she worked with a set of nickel nanostructures: lines of metal (the heat source) on sapphire (the substrate). By altering the variables of heat source size and proximity, and then heating the metal with infrared light, she could observe the properties of thermal transfer and measure how they are affected by changing the surrounding variables.
An interesting component of this research is the method of data collection. It’s hard to measure heat on such a miniature scale, so Kathy had to measure something different—diffraction. When a material heats up, it physically changes shape, and this changes the visual results of shining light on it. By shining a laser on the material being heated and measuring its diffraction, or the pattern of scattered light, Kathy could infer the properties of heat transfer.
“The way we capture data on the camera counting photons, and use it to measure completely unrelated phenomenon, is very, very cool,” says Kathy. “And being able to translate what we’re able to measure into what we care about is part of every scientific project.”
Sure enough, Kathy found that her theory held up to testing during the experimental process. Very small heat sources that are close together do in fact behave in new ways, with similarity to one continuous sheet. In fact, they resemble a sheet well enough that they can be modeled as such mathematically. Importantly, this simplifies the equations engineers might use in their calculations at nanoscales, which makes models of heat sources easier to understand and use.
This research could directly affect how our computers are built. “This is exactly what transistors on a microchip look like,” says Kathy. “This is what information storage looks like!” After all, when you’re building a chip to store information on a computer, you’re building tiny arrays of electrical units (the transistors) that heat up as they are used by the machine.
That heat has to go somewhere, and as Kathy explained, “We have to be able to optimize materials and geometry correctly to deal with this heat.” Kathy believes that with this new description of heat at small scales, we will be able to design and build better computer chips.
One of the most significant and counterintuitive implications of this research is the fact that when you put small heat sources close together, they cool faster. This is very important to an engineer in charge of designing microchips, who might be worried about heat control. Of course this work must be tested in more complicated settings—future research will focus on whether or not these principles hold up with greater or fewer heat sources in more complicated designs than a one-dimensional array.
With her conclusion (and her dissertation!) published, Kathy is living a scientist’s dream—her graduate career has almost perfectly followed the arc of the scientific process. She has gone on to work at Intel, where she may get to start applying some of her research to chip design. On her future work, she says: “Instead of asking how do these structures behave, I’ll be asking how do we make those structures.”
For more, you can check out Kathy’s paper here: www.pnas.org/content/112/16/4846.full.pdf
By Alison Gilchrist