Big Risks and Bold Science

Livermore Laboratory’s Disruptive Research Program fuels high-risk research seeking high-reward solutions.

Ever evolving as a “big ideas” laboratory, Lawrence Livermore exemplifies bold scientific research through its Laboratory Directed Research and Development (LDRD) Disruptive Research (DR) Program. Launched in 2019, the Laboratory’s DR Program supports research and development activities that are unconventional, innovative, and at the forefront of their fields. Livermore researchers are encouraged to embrace high levels of technical risk to yield transformative scientific advances, even if their projects ultimately fail. Successful DR project outcomes advance research by producing order-of-magnitude improvements, opening new program spaces, or creating positive impacts in Laboratory initiatives.

Now in its second funding cycle, the DR Program has achieved a shift in risk acceptance at Lawrence Livermore. With the first cycle of nine selected projects (see S&TR, August 2021, Embracing Risk for Transformational Results) coming to a close and the second cycle approaching initial go–no-go milestones, the Laboratory will evaluate discoveries made and early successes, review lessons learned and occasional setbacks, and then forge next steps. The following representative projects from both funding cycles demonstrate how Livermore is carving a path in disruptive scientific research.

Laser-Focused on Innovation

In recent years, the Department of Energy (DOE) has designated high- intensity, ultrafast laser research as a key priority for national security and economic growth. Home to the National Ignition Facility (NIF), Jupiter Laser Facility, and some of the world’s most renowned experts in laser research, Lawrence Livermore is well positioned to meet DOE priorities. Under the 2019 DR Program, Thomas Spinka, Program Element Leader for Laser Development in Livermore’s Advanced Photon Technologies program, developed a revolutionary high-energy laser architecture that could open the door for expanding laser research resources and serve as an alternative to current laser facilities.

Spinka and his multidisciplinary team of scientists and engineers combined a previously known but underutilized laser material with laser diodes and new capacitor technology that more effectively stores and releases electrical energy to meet the significant goal of reaching a 100-joule output by the end of the project. The team used yttrium-lithium-fluoride (YLF) crystals doped with thulium ions (Tm:YLF) as the laser’s gain medium. Thulium ions in a YLF crystal lattice can store photons created in the pump laser diodes nearly 50 times longer than more traditional laser gain materials. The next component of this innovation called for pairing Tm:YLF crystals with ultracapacitors, which store electrical energy much more densely and less expensively than traditional electrolytic capacitors. Ultracapacitors cannot be paired with conventional laser materials that lack the ability to release their stored electrical energy rapidly enough.

In the project’s first stage, the team combined ultracapacitors and laser diodes for the first time and demonstrated that the photons produced exhibited the right characteristics for pumping Tm:YLF laser crystals. The second stage proved that the pump photons could be stored efficiently in the laser crystal. Once these concepts had been proven, Spinka and his team needed to effectively extract the energy stored in the Tm:YLF crystal. “We initially targeted 10 joules as a goal, but we were fairly certain we could exceed that target,” says Spinka. “Our stretch goal was 100 joules. When we achieved 120 joules, we were thrilled at the opportunities ahead of us and others in the field.”

Regarding the impact of this project, Spinka explains, “Only a handful of laser materials have demonstrated 100-joule output over the past 60 years. Adding another material to that group at this stage is incredibly exciting.” The outcomes of this project could lead to the development and use of smaller, lighter, more portable, and more cost effective high-energy lasers—a key step for growing a U.S. laser research base. Kilojoule-class lasers are immense and offer limited access to experimental teams. The Tm:YLF high-energy laser Spinka and his team developed under the DR Program shows tremendous promise in storing large amounts of energy in a space small enough to fit on one optical table rather than an entire room. Additionally, the success of Tm:YLF in Spinka’s research may enable new secondary source applications, including nondestructive evaluation with x rays and muons (subatomic particles). The team is also testing ultrashort pulse laser capabilities (pulses mere femtoseconds in duration) and the thermal properties of the Tm:YLF gain material to better understand its operational limits.

Keeping Cool at Mach Speeds

Structural integrity is a critical parameter in flight vehicles, especially those that reach hypersonic speeds (defined as Mach 5 or higher). At these velocities, a vehicle’s leading edge can reach 3,000°C due to shock waves at the tip and friction along the surface. Such intense heat is well above the melting temperatures of most materials, presenting a major technical hurdle for hypersonic flight. Lawrence Livermore’s James Cahill used his 2019 LDRD DR funding to additively manufacture a thermal protection system for hypersonic vehicles that could be a game changer for the field.

The foundation of Cahill’s concept is metal transpiration cooling, in which heat is carried away from a material via liquid evaporation. He and his research team created a vascular structure that houses solid metal intended to melt and vaporize through the structure’s branching tubes. “The process is similar to the way organisms cool their bodies through sweating,” says Cahill. “In our concept, we found substantial cooling potential from both the melting and vaporization of the solid metal,” says Cahill. The structure’s dendritic channels extend outward, then narrow to increase capillary pressure and wick the melted metal toward the surface.

To proceed, the team needed to identify a robust material for the vasculature shell as well as an appropriate metal to melt within the structure. High-temperature ceramics, which have ideal properties including high melting points and strength at elevated temperatures, showed the most promise for the structure’s exterior. The metal interior required a melting point of approximately 1,000°C to wet the ceramic shell, yet not oxidize, which would introduce additional phases and thermodynamics and complicate the cooling process. “We narrowed the choice down to silver and gold,” says Cahill. “As noble metals, they are difficult to oxidize, and their melting temperatures are both in the right range. Gold happens to wet the ceramic much better than silver, so it was the ideal candidate to test our concept without adding too many other factors.”

Cahill’s team then manufactured the vasculature structure to test the concept. They fabricated an intricate ceramic structure with some sections no more than 100 to 200 micrometers wide. They began by printing a sacrificial polymer mold of the vasculature. Then, they used a refined casting process to create a ceramic slurry, similar in consistency to modeling clay, to infiltrate into the polymer mold. Once the ceramic was dried and set, they carefully burned out the polymer mold.

Following the components’ manufacturing, Cahill’s team faced a decisive milestone in the DR funding scheme to determine whether the metal transpiration cooling concept was effective. “If the structure didn’t work, that was pretty much it for the project,” says Cahill. Working with components only 1 to 2 centimeters in size, Cahill interrogated samples using a torch to simulate the necessary thermal gradients and high temperatures experienced in hypersonic flight. The data showed both melting and vaporization of the gold along with a significant 200°C drop in sample temperature. With this favorable result, Cahill now anticipates working to achieve scalability of manufacturing, technology transfer, and eventually real-world applications.

As Fast as a CHEETAh

Optoelectronic devices are found in many everyday items, including fiber optic cables, satellite communications, LED lights, solar cells, and photodiodes. Designing and implementing advanced optoelectronic technology that is orders of magnitude faster and operates at much higher voltages is a national security focus area that comes with many challenges. Lars Voss, an engineering group leader, led a DR project to develop a laser-driven semiconductor that could theoretically operate at higher speeds and higher voltages than existing photoconductive devices, which could enable next-generation satellite communication systems capable of transferring more data at a faster rate and over longer distances.

The Compact High Efficiency Electrically Tunable Amplifier (CHEETAh) —in essence, a radio frequency amplifier—began as an innovative optoelectronic semiconductor device designed to achieve advanced performance metrics such as a 100 gigahertz (GHz) frequency range far surpassing existing photoconductor and semiconductor switches. Voss and his team sought to demonstrate a semiconductor switch that uses a high- powered laser doped with the base material gallium nitride (GaN) to generate an electron charge cloud while under extreme electric fields. Although the team was ultimately unable to demonstrate results toward this goal by the end of the project, they gained critical knowledge for future optical electronic research and applications.

Semiconductor devices commonly use gallium arsenide (GaAs) as a doping solution to improve laser emissions due to its high refractive index. One unexpected result from the team’s earlier experiments using GaAs was a phenomenon called negative differential mobility (NDM). Normally, as increased voltage or field is applied to a semiconductor device, the electrons move faster until they reach velocity saturation, meaning they cannot move any faster because they collide with atoms in the material. However, in this project—still using GaAs—instead of the velocity saturating as voltage is increased, the velocity decreased. The electrons moved slower, and Voss found that he was generating a confined region of electrons that would drift under bias. In normal electron charge clouds, the carriers (electrons) at the front of the cloud shield carriers at the back of the cloud from view and move faster. However, under NDM conditions—when operating in the right regime of laser energy, number of carriers generated, and the applied voltages—the carriers at the back move faster and generate an output voltage pulse that is faster than the input optical pulse. Although the team was able to generate these results using GaAs, they were unable to do so with GaN.

The project’s first two go–no-go milestones involved demonstrating the physical principles of the semiconductor and demonstrating the operation at 50 GHz. Voss and his team achieved these two milestones, but the final proof of concept—demonstrating an integrated module before the end of the three-year funding period—was too ambitious. Voss says, “We encountered new regimes of physics that we hadn’t anticipated around the negative differential mobility, and we needed to spend more time investigating what that behavior meant. We took an optical pulse and output a shorter electrical pulse, which hadn’t been done before and made our aim to reach the high-power levels we initially identified more difficult.” Semiconductor device development is an iterative, time- consuming process, a reality that slowed the project’s progress. “In hindsight, given the constraints of developing entirely new technology for semiconductors, combined with the constraints to onsite work associated with the COVID-19 pandemic, we were overly optimistic in what we could achieve,” says Voss.

The team sees that its work has significant value for future research, which is one of the key benefits of the DR portfolio. “We’ve learned about the underlying physics of the devices that we’re interested in pursuing in the future,” he says. “In particular, we learned a lot about operating regimes outside the bounds of where we normally operate. This project will inform how we think about accessing higher frequency modes of operation when building semiconductor devices.”

Powering up Access

Livermore staff scientist Buddhinie Jayathilake doesn’t think of an age range when considering her retirement. Instead, she says, “As soon as the entire world has access to low-cost energy, then I’ll be ready to retire.” Jayathilake’s previous work in carbon dioxide (CO2) conversion and reducing carbon impact has positioned her well to tackle the challenges of her project aimed at finding a cheap and efficient grid-level energy storage solution. A successful outcome could address U.S. national security concerns and drastically increase access to affordable renewable energy around the world.

Realizing cost-effective and efficient renewable energy grid storage has long been a challenge for scientists and engineers. Renewable power sources such as wind or solar are not available all day. According to DOE, viable grid solutions must be able to store excess energy for at least 10 to 12 hours, requiring large-scale use of batteries. Vanadium redox flow batteries are a well-known and reliable battery technology for this purpose. However, at $150 per kilowatt- hour (kWh), vanadium electrolytes are too costly for widespread adoption. Jayathilake is focusing on enhancing iron flow redox battery technology with the use of a device to redistribute lost energy to the system and ensure durability over time. She calls the device an “artificial kidney” because the method in which the device redistributes energy is similar to the way kidney dialysis machines clean and recirculate blood.

Like other flow batteries, iron redox flow batteries use electrochemical reactions of electrolytes (in this case, dissolved iron salts) to store and release energy. Two tanks of electrolyte solutions—one positively and one negatively charged—hold different chemical configurations of iron ions (Fe, Fe²+, and Fe³+). The electrolyte solutions are pumped into a cell stack, wherein a thin membrane prevents the different solutions from mixing. When the battery discharges, iron metal dissolves to the negative solution, releasing electrons (oxidation) and Fe²+, while on the positive electrode the iron ion changes from Fe²+ to Fe³+ (reduction). To recharge the battery, the process is reversed.

Iron redox flow batteries have significant advantages over other battery technologies: notably, iron is the most abundant transition metal and significantly cheaper compared to electrolytes in vanadium redox flow batteries. However, traditional iron flow batteries are not durable enough for widespread grid storage use. They tend to lose charging capacity over time as the negatively charged side’s reaction differs from that of the positively charged side, resulting in energy loss. To combat this capacity imbalance, Jayathilake’s artificial kidney device integrates with the iron redox flow battery system. The device, if successful, will capture the lost energy and store it before electrochemically converting it back to the discharge state and adding it back to the system. Jayathilake says, “This is more sustainable than continuously adding more acids to the systems, which has been the main durability workaround for iron flow batteries to date.”

Since the project initiated in April 2022, Jayathilake has successfully tested the artificial kidney device and battery system for quick and intermediate storage tests (1 hour and 5 hours discharging, respectively). Next, her team will test for improved performance over 100 charging cycles, and eventually work toward DOE’s storage requirements. If the project is successful and can achieve scale to grid- level solutions, Jayathilake may be able to retire sooner than expected.

Climate Friendly Fertilizer

Food insecurity is a daunting challenge made increasingly severe by climate change. Ironically, the process for manufacturing agricultural fertilizers is a major contributor to CO₂ emissions. Industrial fertilizers are manufactured by converting atmospheric nitrogen (N₂) and hydrogen (H₂) into ammonia (NH₃). This process has been instrumental in boosting agricultural production to support the growing global population, but it also emits 1 gigaton of CO₂ annually, all the while consuming 2 percent of the world’s energy and 5 percent of the world’s natural gas and producing large amounts of greenhouse gas emissions and pollutants.

Lawrence Livermore staff scientist Jeremy Feaster seeks to make the conversion of nitrogen to ammonia greener, more sustainable, and more effective. His novel solution is to convert the air we breathe into ammonia using renewable energy sources and state-of-the-art, additively manufactured electrochemical reactors. If successful, Feaster’s technology could initiate a paradigm shift in industrial fertilizer manufacturing and utilization.

Converting nitrogen to ammonia currently involves inducing a thermochemical reaction using a metal catalyst under high temperatures and pressures (between 400 and 500°C and above 10 megapascals, respectively). Nitrogen and hydrogen gases repeatedly pass over four beds of the catalyst and achieve up to 97 percent efficient conversion to ammonia. Although traditional machining methods to fabricate the reactors and catalysts have been effective, efficiency improvements using these methods are limited. Feaster’s research leading up to and including the DR project is aimed at accelerating true innovations in electrochemistry.

Feaster’s project uses computer assisted design stereolithography to design and fabricate entirely new electrochemical reactors and catalysts in a fraction of the time and for a fraction of the cost of traditional reactors. “With advanced manufacturing, we can design and print systems that only a few years ago were impossible using conventional methods,” says Feaster. Traditional reactors require thousands of dollars and weeks to manufacture, and manufacturing new iterations can be a lengthy process. Feaster’s research alleviates these limitations, enabling a team to design, print, and test a reactor in a matter of days. If any element does not work, the same process can be used to rapidly develop and manufacture a new iteration for $10 worth of material.

Using the improved process, Feaster and his team are now designing and fabricating electrodes and catalysts in complex structures that offer a much wider range of capabilities in testing new chemical reactions and subsequent outputs. So far, the team’s experiments have shown that breaking the extremely strong nitrogen– nitrogen bond is less challenging than activating the nitrogen, which conflicts with past research in this field. Feaster and his team will review these unexpected results and focus their efforts on approaches to activate the nitrogen. The first go–no-go milestone will be tested in April 2023, ideally proving that Feaster’s team can achieve a 10-millimolar concentration of activated nitrogen, the same concentration level found in commercial fertilizers. If the trial is successful, the team will advance their focus to converting the activated nitrogen into ammonia.

Looking ahead, Feaster is optimistic about the potential real-world impact of his research. “We have to think about how we’re going to scale this demonstration,” he says. “The demonstration is cool, but how do we go from cool to commercial?” Scaling the results will require Feaster and his team to consider size, production, and time requirements to ensure the conversion process can run for weeks and months, not just hours. Feaster has already begun discussions with local farmers about their needs and how the research effort might be tailored to conditions in the field and not just in the laboratory. If the team achieves its project goals, the opportunities for sustainable fertilizer production to support a growing population will expand enormously.

Innovation for the Future

Looking ahead, the DR Program will continue to explore what disruptive research truly represents. The first cycle’s projects were almost all successful, owing to the caliber of the research teams’ expertise and skills. But, LDRD Program leaders will consider if future projects should push further and embrace even more risk. Livermore’s research efforts, scientists, and facilities haveall contributed to these successes and demonstrate that the Laboratory can be even more disruptive than imagined. The cultural shift in how scientists picture success is still in its early stages, but Livermore’s scientists are increasingly able to be bold and creative. Says Feaster, “Being offered the opportunity to take risks and the grace to fail allows us to shift from a cautious to a more daring approach in which we achieve significant science. We can’t solve tomorrow’s problems with today’s technology. For the technology and solutions needed in the future, we must take correspondingly huge risks.”

—Sheridan Hyland

Key Words: Compact High Efficiency Electrically Tunable Amplifier (CHEETAh), computer assisted design, disruptive research (DR) Program, gallium arsenide (GaAs), gallium nitride (GaN), iron redox flow battery, Laboratory Directed Research and Development (LDRD) Program, laser gain medium, metal transpiration cooling, negative differential mobility (NDM), stereolithography, thulium:yttrium-lithium-fluoride (Tm:YLF), vanadium redox flow battery.

For further information contact Chris Spadaccini (925) 423-3185 (spadaccini1@llnl.gov) or Doug Rotman (925) 422-7746 (rotman1@llnl.gov).