LLNL physicists Adam Bernstein (right) and Steve Dazeley
Detecting Nuclear Blasts
June / July 2008
Volume 6 Number 3
For several years, two Lawrence Livermore National Laboratory scientists have helped conduct research that could yield major dividends for the nation and the world in security and faster computers.
Adam Bernstein, a 45-year-old experimental particle physicist, has led a team of researchers who have developed an antineutrino detector that could determine whether plutonium has been produced in a clandestine manner inside a nuclear reactor.
Paul Mirkarimi, a 42-year-old materials scientist, has been a member of a large team of scientists and engineers who have made important strides toward manufacturing computer chips in a new way, using extreme ultraviolet lithography (EUVL). Their work might open the door within several years for producing computer chips that are dozens of times faster and have hundreds of times more memory than today's chips.
The research by Bernstein, Mirkarimi and their colleagues represent two projects among a number at LLNL and other national laboratories that could result in transformational breakthroughs for the United States and the international community.
After earning his Ph.D. in experimental high energy physics from Columbia University, Bernstein came to Sandia National Laboratories/California in 2000, where he worked as a postdoctoral fellow for two years. Bernstein started studying the use of antineutrinos to detect nuclear explosions. He discovered the concept wouldn't work well, as it would be technically difficult, highly expensive and require huge detectors.
"In the course of my research, I came to the conclusion that nuclear reactor monitoring would be a much more practical use of an antineutrino detector," Bernstein said.
In 2002, Bernstein moved across the street to join LLNL. By that time, Sandia and LLNL had established a joint program that continues to this day to study monitoring nuclear reactors using antineutrinos.
Recently, the LLNL-Sandia team demonstrated that the operational status and thermal power of reactors can be quickly and precisely monitored over hour-to-month-time scales, using a cubic-meter-scale antineutrino detector.
"I'm particularly pleased by our team's efforts. This work has run the entire circuit—€”from being an afterthought for another program to becoming something that the International Atomic Energy Agency (IAEA) is actively considering as a new safeguards tool. This is one of the most satisfying things in which I've been involved in my career," Bernstein said.
Bernstein envisions at least two applications for antineutrino detectors. First, the detectors could provide a ubiquitous presence at nuclear reactors for ensuring compliance with existing IAEA safeguards. Second, in the longer term, it might be possible to make detectors that could help uncover reactors, including even hidden ones, at long distances. "The long range detection would require much more research and be more expensive, but the payoff may be greater," he says.
"Through our work, and that of a growing community of researchers around the world, it is appropriate to speak of a new kind of applied physics—€”applied antineutrino physics. Assuming this technology is broadly adopted as a nonproliferation tool, it's not much of a stretch to imagine a small industry springing up in a few years, churning out antineutrino detectors for nuclear safeguards."
The LLNL-Sandia team, in Bernstein's view, understands how to build antineutrino detectors and the needs of the international safeguards agencies. "We believe it will be possible to patent and/or license several detector designs that will be appropriate for various safeguards applications within the next couple of years."
The leader of the Advanced Detectors Group at the lab, Bernstein is the principal investigator for the antineutrino detector program. He works on the project with colleagues Nathaniel Bowden, Steven Dazeley and Robert Svoboda, along with David Reyna, Jim Lund and Lorraine Sadler from Sandia and Professor Todd Palmer and graduate student Alex Misner from Oregon State University.
Because reactors consume uranium and are the source of all of the world's plutonium, they are a critical nuclear fuel cycle element within the jurisdiction of the IAEA safeguard regime. The regime was put in place by international treaty (the Nuclear Nonproliferation Treaty) to detect the diversion of fissile materials from civil nuclear fuel cycle facilities into weapons programs. Part of the nonproliferation program involves comparing the actual operations of a reactor—€”specifically its changing plutonium and uranium inventories —€”with operator declarations of what the reactor is expected to produce during its normal operations.
That's where the new detector comes in. It provides a direct measurement of the operational status (on/off) of the reactor, measures the reactor thermal power and places a direct constraint on the fissile inventory of the reactor throughout its lifecycle. All three parameters are derived directly from the antineutrino rate, measured nonintrusively and continuously by the detector. The data can be acquired directly by the safeguards agency (for example, the IAEA) without any intervention or support from the reactor operator. The detector is located on the reactor site in an out-of-the-way location tens of meters from the core, and outside the containment dome.
This research has been conducted through the LLNL Global Security's Nonproliferation Program, with sponsorship from the National Nuclear Security Administration's Office of Nonproliferation Research and Development.
Faster computer chips
The other potential transformational breakthrough, aimed at yielding computer chips with dozens of times more speed and hundreds of times more memory, also was undertaken as a collaborative venture between multiple national laboratories.
In 1997, a virtual national laboratory—€”with LLNL, Sandia California and Lawrence Berkeley national laboratories—€”was established to develop the extreme ultraviolet lithography (EUVL) technology for the commercial manufacturing of computer chips.
The research, set up under a $250 million cooperative research and development agreement (CRADA), one of the largest in Department of Energy history, was funded by a consortium of semiconductor firms. Eventually, the consortium included Intel Corp., Motorola Inc., Advanced Micro Devices Inc., Micron Technology Inc., Infineon Technologies and IBM.
At the height of the research venture, about 175 scientists, engineers and technicians at the three national laboratories worked on overcoming the hurdles to allow EUVL to be utilized for producing faster computer chips with more memory.
During the six principal years of the CRADA, between 1997 and 2003, LLNL scientists and engineers involved in the EUVL effort garnered or played roles in winning seven prestigious R&D 100 awards. In addition, Livermore researchers received or assisted in the development of at least 50 patents.
"The EUVL project was extremely interesting and enabled us to perform some particularly innovative research," said materials scientist Paul Mirkarimi. "In the early years, we had major resources and that was a huge help."
Since the EUVL research effort has concluded, Mirkarimi sees some uncertainty about whether the process will be deployed to produce computer chips—€”but he is hopeful and wants to see it succeed. "I really want to see this technology make it. I am now consulting for a company that is bringing an important part of the process to the market and I want to see the EUVL technology move forward."
In April, the EUVL technology won an important endorsement at a technology symposium in San Jose, as the Taiwan Semiconductor Manufacturing Co. (TSMC) outlined its lithography roadmap, according to EE Times.
A silicon foundry giant, TSMC has apparently taken a new position on the EUVL technology and provided its backing. It has placed both EUVL and maskless lithography on its roadmap for 15-nanometer and perhaps 22-nanometer feature-size computer chips. (Currently, semiconductor manufacturers are producing computer chips with 45-nanometer feature sizes).
In the five years since the main EUVL research ended at the national labs, the semiconductor consortium has continued to fund some projects, including work at LLNL to fabricate zero defect EUVL masks.
In late 2007, the laboratory received a letter from three Intel Components Research officials praising the "outstanding contributions" of Mirkarimi, lab mechanical technician Jeff C. Robinson, engineering associate Sherry Baker and participating guest Eberhard Spiller.
"Specifically, we appreciate their efforts in enabling one significant approach toward the realization of defect free EUV mask blanks, and their effective project management during 2007," the Intel officials wrote.
In the end, the LLNL researchers and engineers developed a process to planarize substrate particles and pits on the masks simultaneously to, or below, one nanometer in height.
"When we started," Mirkarimi recalls, "almost no one thought we could achieve this level."
A Woodbury, N.Y.-based firm, Veeco Instruments, has designed and built a tool to perform the LLNL smoothing process on masks used in EUVL work. The machine achieves the same results as LLNL has produced. In addition, the instrument has been installed at SEMATECH's Nanotech Center in Albany, N.Y.
One important challenge remains for the technology's mask efforts. While particles and pits can be smoothed and it can be done quickly, this work still needs to be done more cleanly, according to Mirkarimi.
Other key advances also are needed, Mirkarimi believes, for the source, which will produce the EUVL light that travels through the lithography system and hits the wafer; and for the resist, the polymer-based compound that enables small feature sizes to be created on the wafer.
The Livermore materials scientist sees the chances of the EUVL process some day producing computer chips at 50-50—€”not bad odds for what some would consider a transformational breakthrough.
Stephen Wampler is an LLNL public information officer.
Copyright © 2012 | Innovation America