Wilhelm Röntgen’s discovery of x-rays in 1895 is a top moment in scientific serendipity, right up there with the discovery of penicillin. While studying cathode rays — electron streams that light up fluorescent targets — Röntgen happened upon a ray that, oddly enough, passed right through the heavy black cardboard shielding his experiment. Subsequent investigations revealed this unidentified, “x” ray would pass through most substances and cast shadows of solids on photographic plates. When Röntgen emerged from his lab with a ghostly x-ray image depicting the bones of his wife’s left hand, telegraph wires carried news of his discovery around the world.
MU’s Scott Kovaleski, a professor of electrical and computer engineering, can relate. Kovaleski and his research team have gained worldwide attention for an equally serendipitous development related to Röntgen’s rays — the discovery of a new means of x-ray generation that could lead to hand-held, battery-operated x-ray machines.
The possibility of easily transportable x-ray imaging systems has generated excitement not only in the medical community, where doctors foresee health benefits for patients in remote and impoverished regions, but also among those charged with fighting terrorism, improving manufacturing processes and even exploring space.
Not bad for a line of research meant to be merely a stepping-stone along an entirely different path of enquiry. “The x-ray work was on purpose, but the x-ray part was actually a waypoint towards another end,” says Kovaleski.
To better understand those ends, and how the x-ray breakthrough evolved from them, Kovaleski says you need to spend a moment at the beginning — specifically, his beginning working on ion propulsion systems at NASA’s Glenn Research Center in Cleveland, Ohio.
Yes, it’s rocket science and it’s complicated. The first thing to know is that Kovaleski was dealing with plasma, a state of matter that fits alongside the three more familiar states of solid, liquid and gas. Plasma occurs when gas molecules or atoms are ionized, or electrically charged, by adding or removing electrons. To create thrust via ion propulsion, electric and/or magnetic fields are applied to push plasma’s electrically charged ions and electrons.
Kovaleski’s NASA research focused on a class of propulsion called micropropulsion. The “micro” is not a reference to the size of the spacecraft but to the amount of propulsion required. His goal was to create a propulsion system that was extremely small so as to leave maximum room on the spacecraft platform for sensors, cameras, communication devices and other data systems.
Kovaleski continued to experiment with ways to create plasma for ion propulsion after he arrived at MU in 2003. Of particular interest was lithium niobate (LiNb03), a man-made crystal that, when subjected to fast, high-voltage pulses, spits out huge amounts of charge that collects as dense plasma on the crystal’s surface.
Unable to afford a fast, high-voltage pulser, Kovaleski’s team had to settle for a pulser that put out alternating voltage “tunable” to radio frequencies. The upside to this technique was that Kovaleski’s team could use the tunable voltage to see how different voltage frequencies impacted the creation of plasma.
“So my graduate student at the time was tuning the frequency around just to see what would happen,” Kovaleski recalls, “and he comes in one day and says: ‘Oh, hey, when I tune this frequency around, when I get close to this [specific] frequency, I can make these plasmas with much less input voltage than when I’m off that frequency. Isn’t that interesting?’ And I say: ‘Yeah, that is interesting. I wonder what’s going on there.’ ” More digging revealed that the student had, as Kovaleski puts it, “accidentally made a piezoelectric transformer.”
Piezoelectricity results when certain crystalline materials — lithium niobate among them — are subjected to “mechanical stress” and respond with an electrical charge. In Kovaleski’s lab the mechanical stress was not directly applied but rather induced by a low-voltage electric signal from the pulser. At just the right frequency, that signal caused the lithium niobate crystal to vibrate, thus creating a wave in the same way a hammer strike will cause a dinner bell to vibrate and ring. That vibration of the crystal then amplified the low-input voltage to a voltage high enough to form plasma.
Piezoelectric transformers are found in such everyday objects as laptops and TVs, where they can take a small input voltage — say 10 volts —and easily produce 300 volts. Usually, the goal isn’t to create plasma but simply to leverage the crystal’s mechanical wave properties to boost output voltage. But in accidentally using the piezoelectric effect, Kovaleski’s team had found a way to generate significant amounts of plasma from just a small amount of electricity.
Still interested in propulsive thrust, Kovaleski’s team continued experimenting, and eventually, they were able to configure the transformer so that an input of 10 volts generated an output of 10,000 volts. That resulted in enough plasma to create propulsive thrust, the project’s original goal. But Kovaleski wanted to do more.
“We thought, ‘Well, hey, if you can make 10,000 volts to make plasmas and generate space propulsion thrust, could we make much higher voltages and get other interesting radiation phenomena to happen?’ ” Kovaleski says.
At this point, the “interesting radiation phenomena” Kovaleski’s team wanted to explore wasn’t x-rays but energized neutrons, which are often produced by radioactive materials known as radioisotopes. In particular, Kovaleski hoped to discover a new, safer alternative to the radioisotope sources used in oil well logging, a process for discovering underground deposits of oil.
Such radioisotopes work similarly to x-ray imaging systems in that both use radiation to detect and analyze materials enclosed within solids. But unlike x-ray sources, radioisotope sources do not turn off with a switch: Even when pulverized or immersed in water, radioisotopes continue to emit environmentally hazardous radiation. This makes radioisotope sources dangerous in the wrong hands.
And there’s another problem: It’s not uncommon for radioisotope sources to get stuck in the deep boreholes necessary for oil well logging. “You can’t just take a drill and break up that radioisotope source because it could get in the water supply,” Kovaleski says. All that’s left to do is to cap off a very expensive hole and start over.
Kovaleski’s team aimed to solve these problems by coming up with a compact neutron source that would emit radiation only when turned on and intact, which would drastically limit its potential destructiveness. Such a source, they figured, would be an invention not only with environmental appeal, but one with a real market value, as those in the oil well logging business would love to be free of the administrative costs associated with keeping track of traditional radioisotopes sources.
Building such a device depended chiefly on overcoming one daunting hurdle — developing a piezoelectric transformer that could generate energies in the 120,000 volt range. That is the voltage benchmark for generating energized neutrons, the penetrating particle used in applications like well logging.
Kovaleski hoped to discover a new, safer alternative to the radioisotope sources used in oil well logging, a process for discovering underground deposits of oil.
Kovaleski had a plan to get and use those energized neutrons: First, he would generate deuterium ions using the plasma generation of the earlier propulsion technology. Then — once he got his piezoelectric transformer to generate 120,000 volts — he would accelerate those deuterium ions and direct them to a target also containing deuterium. This would lead to nuclear fusion: When the deuterium ions, moving at the energy of 120,000 volts, smacked into the deuterium-containing target, some of the accelerated deuterium ions would fuse with the target deuterium. Following that fusion, the ions would immediately break back apart, releasing the desired energetic neutron.
And the energetic neutron is desired because with high energy and no charge, it penetrates extremely well, “so it has no problem going through concrete walls or steel containers,” Kovaleski explains. Further, if an energetic neutron does interact with something, what it interacts with will kick out other ionizing radiation, in particular gamma rays and neutrons, which can then be analyzed to reveal the composition of the radiation’s source. This is how radioisotope sources provide information about hidden materials, such as rock in an oil well or a uranium brick sitting inside a truck.
Kovaleski looked at what other people had done in piezoelectric transformers and decided that his best bet for getting to 120,000 volts was to switch his design to a configuration known as a Rosen-type transformer, which is about the size and shape of a stick of gum. Kovaleski differed from his predecessors, however, in using lithium niobate in his configuration rather than ceramics, the most commonly used material. He went with lithium niobate because it’s a tougher material and able to withstand plasma generation. Kovaleski’s team also placed the transformer into a vacuum. With these changes, he was able to generate 30,000 volts.
That is how far the project had advanced when graduate student Brady Gall joined the lab in 2009. He was charged with continuing to optimize the piezoelectric transformer to generate upwards of 100,000 volts. Gall researched and tinkered for two years, but nothing brought the voltage above the 30,000 volts his predecessor, doctoral student Andrew Benwell, had achieved.
Gall began to think it was impossible. “I was like: ‘That’s it. There’s no way these things are going to go more than 30,000 volts,’ ” he recalls. “But we kept pushing at it and kept making improvements.”
The breakthrough came after a redesign that moved elements around on the transformer, Gall says. “It was not a steadily marching up to 120,000 volts; it was nothing, nothing, nothing, nothing. Then, finally, boom, there it was. And I was like: ‘Yes! I can write my thesis! I can finally graduate!’ ”
Still, neither Gall nor Kovaleski realized their success would lead to a paper on a compact x-ray source. And it wouldn’t have, had there not been a problem with measuring the piezoelectric transformer’s output voltage.
We thought, ‘Hey, we’re actually making x-rays! What if we also used this as an x-ray source, not just as a convenient little trick for us, but an x-ray source in itself?’
X-ray compactors Scott Kovaleski and Brady Gall Usually, to measure the electrical characteristics of a device one simply hooks up a “parasitic” element, such as a multimeter or voltage tester. But a piezoelectric transformer is too finicky for that. Traditional diagnostics sap so much energy that the transformer won’t operate at its intended output.
“So it won’t be at 120,000 volts; it will pull it down even to 1,000 volts or a few hundred volts,” Kovaleski says. “So we had to come up with a new way of seeing the voltage of this piezoelectric crystal.” And that’s when x-rays entered the picture.
The diagnostic tool Kovaleski’s team chose was the bremsstrahlung measurement. Bremsstrahlung refers to the process behind common x-ray generation: When a charged particle undergoes an acceleration change in an electric field, electromagnetic radiation in the form of an x-ray is emitted. Because the voltages of escaping x-rays correspond to the voltages powering their release, Kovaleski says, the bremsstrahlung process can be used as a diagnostic method. For Kovaleski’s team, the plan was to use their piezoelectric transformer to generate x-rays — even weak x-rays, too weak for imaging — then measure the x-rays’ energies to determine the transformers’ output voltage.
Turning the transformer into an x-ray source was easy enough because generating x-rays with even 20,000 volts is a well-known process with simple steps. First, Kovaleski’s team took little sharp points of platinum-iridium wire — “They are so sharp, they are atomically sharp,” Kovaleski says — and affixed them to the crystal. Next, they placed a dense metallic target a centimeter from the wire points and grounded it. Then, it was time to fire up the transformer, which, in the beginning, sent 30,000 volts through the wires and into the centimeter of space; later that 30,000 volts became 120,000 volts.
“That’s an electric field, and it’s a fairly high electric field, and because these wire tips are really small, they’ll actually increase the electric field locally to the tip,” Kovaleski says.
The idea is to generate an electric field strong enough to pull the electrons out of the platinum-iridium wire. Those released electrons then shoot across a centimeter of space and strike the grounded metallic target — or, to put it another way, the charged particles (electrons) undergo a change in acceleration (by striking the dense target), and voila, the target kicks out an x-ray. During lab tests, 120,000 volts created x-rays that were strong enough for imaging. “And that’s how our x-ray source worked in the paper that we published in January [2013 in IEEE Transaction on Plasma Science],” Kovaleski says.
Publishing was really just an afterthought, Gall adds. “The x-ray diagnostic was intended just to be a diagnostic. It was just a neat little way to measure the output voltage without impacting the performance of the device. And then we thought: ‘Hey, we’re actually making x-rays! What if we also used this as an x-ray source, not just as a convenient little trick for us, but an x-ray source in itself?’ ”
The answer to that “what if” came after the University touted the findings in a press release. The story showed up everywhere from Discovery.com to Reddit, where the focus tended to be on the device’s similarities to the tricorders of the Star Trek universe. But it wasn’t just Trekkies who found the story fascinating.
James Lucas, president and CEO of Epic Medical Concepts and Innovations, a company that helps researchers and inventors bring medical devices to market, made one of the first industry calls. Specific applications that Lucas wants to explore are small dental x-ray machines — so small, they could even fit inside patients’ mouths — and portable x-ray machines that could go out with ambulances or onto battlefields. Already, Lucas has met with Kovaleski, and their meeting went really well.
“We always look at three things,” Lucas says. No. 1 is the technology: Is it a breakthrough? Could it have an important impact on human health? No. 2 is profitability: Will sales cover the cost of development? And No. 3 is the partner: Will the researcher stay involved to see the technology developed?
Lucas says it’s too early to answer the financial question, “but we definitely like what we see of the technology, and we definitely like what we see of Scott.”
Another call Kovaleski has received was from a totally different field, manufacturing.
Manufacturers use x-rays to diagnose such problems as weak materials and bad welds, but currently, getting those x-rayed requires taking samples to a central radiography facility. If instead, manufacturers could bring portable x-ray machines to production lines or, perhaps even better, install x-ray devices on the lines themselves, that would save man hours and allow manufacturers to test 100 percent of their products, which would result in minimal scrap.
“I can tell you manufacturers will go crazy for this,” says Tony Dellacecca, director of quality at Soldy, a manufacturer of aluminum and zinc die castings. He has already had one conversation with Kovaleski about Soldy’s interest.
“I think we both were excited,” Dellacecca says. “I think from his standpoint, he hadn’t considered this particular application, and from my standpoint, it saves the company money. … Pennies add up, and when you are looking at the ability to detect these things quickly and get that information quickly, I think there’s some huge commercial appeal to it. I can imagine most of the automotive manufacturers, the aerospace companies and ammunition companies, all of these companies that use aggressive inspection techniques, will probably adopt these [portable x-ray machines]. If you can get 100 percent of the product x-rayed, that changes the game a lot.”
Other applications Kovaleski sees are in law enforcement agencies and security screeners, or on space vehicles, such as the Mars Rover, that could never haul around a traditional x-ray machine.
Kovaleski has said it would be possible to develop a prototype of a portable x-ray imaging device in three to five years. So far, however, the neutron work still commands center stage.
“Our primary focus and the funding we’ve received so far has been for the [neutron] work,” Kovaleski explains. “Because of the interest that has arisen this year, we have started looking more closely at the x-ray potential on our own; we’re not funded to do that work, but we’re interested in seeing where we could go with that.”
“You know, I’ve written lots of papers,” he says, “and they sort of go out there, and some people read them, usually people in my field, but this particular one struck a chord with people.”