Researchers see potential in wide bandgap, ultra-wide bandgap materials

Caption: BOB KAPLAR (1768) peers at a test circuit built under a Grand Challenge Laboratory Directed Research and Development project to evaluate the switching performance of wide bandgap and ultra-wide bandgap power semiconductor devices. (Photo by Randy Montoya)

by Sue Major Holmes

Sandia researchers are working on wide bandgap materials that someday could replace silicon as the backbone of the power semiconductor industry.

Wide bandgap (WBG) materials such as silicon carbide (SiC) and gallium nitride (GaN) could potentially vastly improve the performance of the electric power grid, solar photovoltaics, and electrical motors, and help meet the aviation and automotive industries’ need to use less energy. GaN has already enabled a revolution in efficient lighting technology and serves as the heart of widely available bright light-emitting diodes. These semiconductor materials have bandgaps significantly wider than that of silicon, the material on which most power systems and modern computers are based.

Bandgap is a fundamental materials property. WBG and ultra-wide bandgap (UWBG) materials are attractive as transistors, or switches, because they can handle higher temperatures and voltages with less degradation.

UWBG materials have potential applications in nuclear weapons and defense systems as well as in future energy systems because of their potentially high radiation resistance and the prospect of enabling smaller and lighter power systems. However, the materials still require a lot of research, says Sandia Fellow Jerry Simmons (1000).

Sandia is researching SiC and GaN, but it’s also working to leapfrog over these next-generation materials to the generation-after-next, UWBG materials such as aluminum nitride (AlN).

AlN and GaN are compatible enough to be mixed. That allows researchers to take small steps toward developing AlN by gradually increasing the amount of AlN versus GaN to study behavior and the effect of lattice mismatch between the semiconductor and the material it’s grown on, says Bob Kaplar (1768), principal investigator for a Grand Challenge Laboratory Directed Research and Development (LDRD) project on UWBG materials.

Wide bandgap materials have performance advantages

Estimates predict SiC could perform 100 times better than silicon for power switching, GaN could be 1,000 times better than silicon, and AlN could be 10,000 times better than silicon. However, their potential can’t be tapped until researchers better understand how the materials work, develop mature techniques to process them, and address reliability concerns, particularly for high-consequence uses.

Sandia researchers have conducted wide bandgap research for nearly two decades. Early work largely focused on solid-state lighting and ultraviolet light emitters, then research expanded into such areas as photodetectors and radio frequency transmitters.

The UWBG grand challenge is the flagship project for Sandia’s Power on Demand Research Challenge aimed at developing electrical power systems with the smallest size and weight, while handling the largest possible amount of energy. The research challenge tackles underlying fundamental science questions, engineering applications, and technical challenges for devices, materials growth, and power systems.

The grand challenge is a major investment and an opportunity to create a team with a single vision, says Manager Rick Schneider (1120), program lead for the grand challenge and part of the research challenge team. Technology from the grand challenge could advance the research challenge’s effort in power electronics. Power on Demand’s other two focus areas are advanced battery research and photovoltaic research.

The grand challenge covers three areas: materials growth; device design, fabrication, and testing; and defects and radiation resistance. It explores ways to grow UWBG materials with fewer defects and different device designs to exploit the properties of materials other than silicon.

Among others leading teams or research focus areas, Andy Allerman (1126) works on growth of wide and ultra-wide bandgap materials; Andy Armstrong (1123) measures densities and energy levels of defects in these materials; Art Fischer (1123) and Albert Baca (1766) research device architectures and fabrication; and Jason Neely (1353) evaluates device performance.

“DOE already supports the wide bandgap manufacturing industry, so there’s a national priority around this technology,” Rick says. Sandia is unique in the DOE complex for its semiconductor capability — the Microsystems and Engineering Sciences Applications (MESA) complex; the Center for Innovative Nanotechnology it operates with Los Alamos National Laboratory; and Sandia’s long history of innovative semiconductor devices and materials. “We come in with a lot of relevant experience and a rich base for power electronics,” Rick says.

The largest support for the Wide Bandgap Grand Challenge within Sandia comes from nuclear weapons programs, and Defense Systems & Assessments is interested as well, he says. Nationally, the research has potential impact on the grid, transportation technology, and energy efficiency.

Once Sandia establishes a core ultra-wide bandgap capability, “We have the potential to create very unique radiation-hardened components. It may be a very good opportunity to think about a trusted manufacturing enterprise within Sandia in partnership with MESA. That’s a 5- to 10-year vision for Sandia with the potential for external collaboration and industry outreach,” Rick says.

Performance issues include defects, integration into larger systems

Although some devices using SiC and GaN are on the market, thorny problems remain, Bob says. Common performance issues include defects, incompatibility with the microelectronics substrates on which the materials are grown, and the impact of integrating a device into a larger system. Sandia researchers can evaluate those problems impartially, building on knowledge and capabilities gained in decades of nuclear weapons work.

Improvements in WBG or UWBG materials could cascade into improvements in an entire system.

For example, WBG materials could potentially reduce the estimated 10 percent energy loss that occurs between generating electricity and transmitting it to a home or business. A wide bandgap allows faster switching, which could reduce the size of bulky passive elements, Bob says. Thus, WBG devices could mean higher reliability, smaller size, and less expensive systems.

Likewise, if electric vehicles could tap the potential for WBG and UWBG power electronics to withstand higher temperatures, they might not need liquid cooling systems, resulting in smaller and lighter electronic systems. That could reduce the system’s complexity and improve vehicle range because the car would weigh less.

But materials compatibility also can be a problem in part of the transistor called the gate, which turns the switch on and off. A critical part of that is the gate oxide, or insulating material. Researchers are exploring various substances as possible gate oxides.

For silicon and SiC, the gate oxide is silicon dioxide. While it’s a match for silicon, silicon dioxide is a more complex interface for SiC because the presence of carbon results in more material defects that can affect the transistor’s electrical behavior, Bob says. Oxide physics on GaN pose an extremely complicated materials science problem that will require atomic-level understanding before such devices will be practical, he says.

Researchers studying how to grow wide bandgap materials better

There also is no widely available, low-cost, large-area GaN substrate, the crystal on which semiconductor material is grown. That lack affects the type of transistor structure that can be made. Bob led an LDRD project to explore how to design, grow, and fabricate GaN devices on GaN substrates.

Silicon and SiC are grown on very pure single-crystal silicon or SiC wafers purchased commercially, meaning the starting material and grown layers are basically the same. However, without low-cost, large-area GaN wafers, GaN traditionally is grown on slices of sapphire, silicon, or SiC. Since the atomic spacing between the wafer and GaN isn’t the same, GaN ends up with tiny defects, Jerry says. Defects limit the type of transistor structure that can be made.

That’s why device performance doesn’t live up to predictions, Jerry says. Defects also impair reliability, a critical issue for uses such as the electric grid and transportation.

Until researchers can figure out how to grow large-area bulk GaN substrates inexpensively, they use such tricks as etching minuscule fingers into the substrate so GaN grows on top of the fingers and merges, reducing the strain inside, Jerry says.

WBG and UWBG devices must be integrated into larger systems. A switch, for example, fits with other parts into a system that includes high-voltage switches and low-voltage sections such as microprocessors, with a buffer or driver in between. Everything in a highly integrated system must withstand high temperatures. While WBG materials are robust at such temperatures, the packaging or semiconductor chip casing is not as resilient, Bob says. Sandia is studying robust packaging approaches.

“With expertise spanning fundamental materials growth all the way through systems integration, Sandia is uniquely qualified to tackle the entire range of challenges in this emerging field and is well-positioned to be the leading laboratory in an area of increasing national importance,” Bob says.


-- Sue Major Holmes

Back to top of page

‘Laboratory Biorisk Management’ details safety, security methods for biosciences sites

Jennifer Gaudioso and Ren Salerno are editors of a new book, Laboratory Biorisk Management, that aims to aid hospitals and bioscience labs assess, mitigate, and manage biological risks.  (Photo by Randy Montoya)

by Heather Clark

Recent mishaps at laboratories that mishandled potentially dangerous biological substances and the transmission of the Ebola virus in a US hospital are symptoms at bioscience facilities that two Sandia researchers seek to prevent in a new book on biorisk management.

The 228-page “Laboratory Biorisk Management” published by CRC Press was edited by two senior managers, Ren Salerno of the Biological Sciences and Technologies Program (8630) and Jennifer Gaudioso of the International Biological and Chemical Threat Reduction Program (6820).

“This is the first full-length manuscript on the detailed implementation of biorisk management,” Ren says. “Laboratory biorisk management is fundamentally a culture of rigorously assessing risks, deciding how to mitigate those risks deemed unacceptable, and establishing mechanisms to constantly evaluate the effectiveness of the control measures.”

Salerno, Gaudioso, and the other authors advocate a cultural shift in how laboratories, hospitals, and other bioscience facilities approach safety and security. They say biorisk management should:

  • prioritize an intellectually sound, evidence-based decision-making process using substantive risk assessments to evaluate a facility’s risk based on the unique operating environment.
  • require implementation of mitigation measures according to the risks of specific activities, experiments, or projects.
  • constantly assess performance.
  • emphasize more meaningful roles and responsibilities for all personnel within a facility.
  • assign ultimate responsibility for safety or security performance to top management.
  • be scalable from the smallest hospital or clinical lab to the largest research institution.

About a dozen other Sandia experts in the field paired up with their international counterparts to develop and advance a practical set of concepts relevant and able to be implemented by labs worldwide, Jen says.

Other Sandia authors include William Arndt, Susan Boggs, Ben Brodsky, contractor Mark Fitzgerald, Laura Jones, William Pinard, and Laurie Wallis (all 6824); Lisa Astuto Gribble, Susan Caskey, and Monear Makvandi (all 6825); and LouAnn Burnett, Lora Grainger, and Cecelia V. Williams (all 6826).

In addition to explaining biorisk management and providing a model, the book includes chapters on risk assessments, facility design and controls, training, operations and maintenance, how to evaluate biorisk management performance, communications issues, case studies, and future directions and challenges for biorisk management.

Authors say growth of biosciences necessitates need for change

The time to rethink the safety and security of biosciences facilities is now because of the expansion in scope, scale, and sophistication of the biosciences field over the last 15 years, Ren says. Examples of this expansion include the rapid advance of synthetic biology and, following the 2001 anthrax attacks on the White House and Congress, the deep integration of biosciences within national security research.

“The times are changing and what we have never done in the biosafety community is take a good hard look at why we do what we do and ask ourselves if the system needs to be radically reshaped in light of all the changes in biology,” he says. “From our perspective, this is way overdue.”

Today’s biosafety guidelines were created in the early 1980s. The Centers for Disease Control and Prevention partners with the National Institutes of Health to publish biosafety guidelines to protect workers and prevent exposures in biological laboratories, Ren says.

The current guidelines tier biological agents into four risk groups and designate work with those agents into one of four biosafety levels. Ren says use of the guidelines has become perfunctory and their nuances are not widely understood by many personnel at bioscience facilities. For example, he says, it has become common practice in the field to share risk assessments or material safety data sheets between facilities, so they no longer take into account the unique circumstances of each facility, including location, the type of work done there, and the expertise and training of its personnel.

“I believe the events of the last year in this field demonstrate exactly what we’ve argued: that the current system is broken. It’s a systemic problem,” Ren says. “We’ve created an administrative-based safety culture in biology that is way too simplistic for the level of complexity of today’s science.”

Global assistance in lab security brings issues to light

Sandia scientists became more aware of the issues through their work over the past 15 years with laboratories around the globe.

In 2008, the European Committee for Standardization hosted an international workshop that published an agreement among 24 countries, introducing an overview of biorisk management. The World Health Organization, which quickly adopted the biorisk management framework, asked Sandia and other technical advisers to create a two-week Biorisk Management Advanced Trainer Programme, which Sandia experts helped teach in 2010-2011.

“We were barely scratching the surface and everybody wanted more information, more detail, and wanted to understand how to implement the concept,” Ren says. “That’s when we began talking about the need for a manuscript.”

In addition to the book, the Labs also curates the Global Biorisk Management Curriculum, which contains 47 separate courses developed by Sandia and others. It is being taught by 500 trainers worldwide, Jen says.

Book says focus on performance can prevent problems before they happen

Ren says the book promotes the idea that a good biorisk management system determines ahead of time the metrics that will show a project, experiment, or activity is being done safely and securely.

The risk assessment completed before an activity has begun sets leading safety and security performance indicators. Then, regular monitoring and documentation will show whether the activity is achieving the safety and security goals, enabling scientists to identify things that are working fairly well, but perhaps not perfectly, while the activity is in progress, he says.

“In other words, by evaluating performance you can adjust your safety measures before something happens,” Ren says. “You don’t want a bad thing to happen to determine whether or not your system is working.”

Some might view this as added paperwork, but Ren and Jen point out that experience in other high-consequence industries shows that when processes are more effective and efficient, a more effective safety system is the result, which in turn leads to decreased costs and improved productivity.

Jen explains that a lot of the risk assessment and mitigation work in the book should help institutions solidify good practices and fill in gaps in their procedures.

“The burden should be proportionate to the risk, so that you’re not asking too much from people who are carrying out activities that don’t present a lot of risk to themselves or the community,” she says. “But for people whose activities carry more significant risk, then yes, they have to do a little bit more to make sure they are managing those risks appropriately. I don’t think that’s an unreasonable thing to ask.”

Culture change in biosciences required

Ren recognizes the system outlined in the book won’t work unless stakeholders in the biosciences community buy into the concepts.

“If someone takes this book, agrees that the performance chapter makes some good points, but then adds a large number of additional, perhaps arbitrary, requirements, the system will look like yet another administrative checklist. That would be counterproductive,” he says.

In the final chapter, author Ben Brodsky (6824) and a co-author wrote that biorisk management is a relatively young approach that faces challenges to being implemented broadly.

More evidence is needed to show whether biorisk management works, so they call on more organizations to develop ways to measure the performance of biorisk management and to show how it benefits an organization.

“This will enable the biorisk management community to continue creating tangible benefits for the bioscience community, including keeping society and the environment safe while more efficiently facilitating the delivery of science,” they wrote.


-- Heather Clark

Back to top of page

Workhorse gamma ray generator HERMES III reaches significant milestone

FORMER CONTRACTOR JJ Montoya (foreground) and Chris Kirtley (1342) work atop the HERMES III Accelerator, making final adjustments on a newly rebuilt cavity. (Photo by Randy Montoya)

by Neal Singer

The High-Energy Radiation Megavolt Electron Source, better known as HERMES III, fired its 10,000th shot on July 14. HERMES III is the world’s most powerful gamma ray generator. It produces a highly energetic beam that tests the capability of electronics to survive a burst of radiation that approximates the output of a nuclear weapon. The machine can accommodate targets that range in size from a single transistor to large vehicles.

The machine does its work by generating an intense electron beam at energies approaching 20 mega-electron volts. The electron beam is then guided into a high atomic-number target, where it is slowed down and produces copious amounts of gamma rays. The thinness of the target permits the majority of the beam’s energy to pass through it; thus, the passage causes minimum damage. This enables HERMES to fire multiple shots at a time without breaking vacuum. “HERMES III has gone hundreds of shots without any damage to the tantalum Bremsstrahlung diode,” says manager Ray Thomas (1342).

To achieve its high voltage, HERMES III uses 20 inductively isolated modules arranged in series. The machine resembles a short subway train in size and shape — 17 feet wide, 50 feet long, and 16.5 feet high. Each “car,” or unit, adds 1 million volts in series, reaching a total of 20 million volts. Its linear, voltageadding geometry is distinct from the wagon-wheel shaped architecture favored by Saturn and Z, which is useful for adding current.

Also, HERMES III places its targets at an end of the machine rather than its center.

“Our customers bring their own targets, place them at the front of the machine as we request, and then remove them after the shot,” says technician Gary Tilley (1342), who’s worked on HERMES III for 20 years. The other two facilities have to clean up the remnants of exploded targets placed at the center of their energy flows. Juan Diego Salazar (1342) is part of the team that watches to make sure each module receives the proper dose of power, at the right moment in time, to accelerate the beam.

“Every firing is different,” he says. “The targets always change.” Continual reevaluation of the electrical power feeding the beam as it flows through its modules, and continual recalibration of the beam’s line of sight to the target, are necessary because an unobserved power or alignment failure somewhere within the system could mistakenly show a target more radiation- resistant than it actually is.

Real-time adjustments would be too late: The achieved beam flashes for 20 billionths of a second, about the time it takes light to travel 20 feet. “Accurate results are important,” says Ray. “That’s what we’re about.”

-- Neal Singer

Back to top of page

Download Lab News September 4, 2015 (PDF, 4MB)