Spy Schools

by Daniel Golden

  Liu stepped into the void. Acting more like a professor than a mere graduate student, he familiarized himself with all of the projects in Smith’s lab, and many in other Duke engineering groups, dispensing ideas and encouragement and initiating collaborations. “He was really more of an organizer and manager than a scientist,” Smith said. “His strongest skills were not in science, but he seemed to have an extraordinary ability to get others to do things.”

  Rarely, though, did Liu knuckle down himself to the rigorous tasks of analyzing and testing his creative notions. “He was a talker,” Schurig said. “He spent all of his time talking to people. He was a good communicator. You never saw him spending long hours focused on his computer. I’m always a little suspicious of people like that. They’re gleaning their ideas from others.”

  * * *

  SOON AFTER JOINING Smith’s group, Liu suggested collaborating with his former team, run by Cui at Southeast University in Nanjing. Liu told Smith that Cui had a large, talented roster of scientists whose projects would mesh with the Duke lab’s research.

  Smith liked the idea. His international association with Pendry had proven fruitful, and Cui was a respected scientist. At San Diego, Smith told me, his adviser “had been extremely paranoid and xenophobic. That really harmed us as a group, and when I began to be more independent, my feeling was to be as open and collaborative as possible.”

  Like most collaborations that bubble up from student or faculty interactions, the one between Smith’s group and Cui’s was unwritten. No formal agreement was drawn up that might have set limits on sharing intellectual property. “The majority of collaborations are just informal, people getting together at meetings and brainstorming,” Smith said. It’s “another nebulous area in the system that is in need of clarity.”

  Liu became the liaison between the Duke and China teams. “He had this kind of advantage of being able to shift between universes, our group and the Chinese group,” Gollub said.

  Then Liu broached another proposal: Smith should enhance the collaboration by inviting Cui’s team to Duke. When Smith objected that he didn’t have the budget, Liu told him to relax: China would pay. So the researchers visited Smith’s lab, where they photographed the equipment: mainly, two big plates of aluminum separated by one centimeter, on mechanical stages. Then they measured the length of the plates, the thickness of the metal, and other dimensions.

  Taking pictures of other teams’ laboratories is controversial in academia. To maintain a competitive advantage, some research groups at U.S. universities ban the practice. Though dismayed, Smith reasoned that his collaborators were entitled to study the equipment. According to the FBI, the lab was later reproduced in China.

  As a member of Smith’s group, and coauthor of the October 2006 Science article, Bryan Justice designed the apparatus. He could tell from later publications by Cui’s group, he told me, that they had “painstakingly duplicated” it and “had everything down to the nuts and bolts … a little more closely than you should be able to by reading our papers.”

  Justice said he “spent probably a semester and a half developing, troubleshooting, and debugging that system before it was ready for prime time.” Cui’s group could “duplicate it in a couple of weeks because we’ve already done the heavy lifting.” However, according to Smith, they failed to reproduce the cloak itself, which limited their ability to collect data with the apparatus.

  In emails to me, Cui didn’t deny that his team photographed and replicated the equipment, but he downplayed the benefit, and said he was surprised that anyone would regard it as a violation of academic ethics. Smith’s group had already published a description and picture of the apparatus, which was simple in concept and design and easy to build, Cui wrote. “By the way, I have seen at least five similar equipment in China, UK, Singapore, and Hong Kong.”

  The collaborations between his group and Smith’s, he added, were “properly scientific.”

  A talk by Cui at Duke also raised suspicion about his—and Liu’s—intentions. At the time, Liu was working with a postdoctoral fellow in Smith’s group who had devised a new application of transformation optics, or using metamaterials to bend light. The postdoc, who asked not to be identified in this book, told me that he supplied files and data from his simulations to Liu, who was “supposed to find an implementable design and to fabricate a structure that would comply with the material parameters I calculated.”

  Instead, the fellow was stunned to hear Cui deliver the simulation results, as if the Chinese team had discovered them. “For me, it was obvious that Ruopeng reported my results to T. J. Cui and most likely sent him the files,” he continued. “I was really upset and I think one day after the talk I reported the case to David Smith.”

  Again, Smith shrugged off the concerns. “With me believing that we were in a collaboration, I felt that we just needed to tune things a bit,” he told me. He told the postdoc that “he was not obliged to share details of his project if he felt uncomfortable with the situation.” The postdoc ended the collaboration with Liu and virtually stopped speaking to him. In retrospect, Smith said, “he made a very good choice to be cautious with his results around the lab!”

  Surfing the Internet, the postdoctoral student noticed that Liu was coauthoring scientific articles with Cui’s team. The fellow notified Smith, who didn’t know about the publications. “Being new to all of this, I assumed the best, and thought they just didn’t understand that you shouldn’t do that,” Smith said. He talked with Liu, who said it was a misunderstanding, and wouldn’t happen again. But it would.

  After the Chinese team’s visit to Duke, Liu began inviting his colleagues in Smith’s group to reciprocate. Still angry that Cui had appropriated his data, the postdoctoral student declined. But Smith and several other members of his lab agreed to go to China. The offers were hard to resist, because the trips were free: China’s government footed the bill.

  Early in 2008, Jonah Gollub accompanied Liu to China and gave talks on his research at five universities. “I was confused about why he was bringing me over there,” Gollub told me. “We saw many, many different groups. It was kind of weird. Even at the time, I was saying, ‘Why are we doing this?’” Gollub added that “it was clear they had money for this sort of thing.”

  In November 2008, Smith, Liu, Schurig, and other Duke researchers participated in an international metamaterials workshop that Cui organized at the Jinling Riverside Conference Hotel in Nanjing. Along with Cui, Smith and Pendry were listed as cochairs. “I’m not positive that I agreed,” Smith told me. “In retrospect, I think they were trying to sell us as active collaborators in China.”

  “A whole bunch of us visited China,” recalled Schurig, who is now a professor at the University of Utah. “At the time, we were pretty excited about it. We’d never been to China. They were great hosts. It was a really fun trip. A lot of research was presented by the Chinese and the visitors. China paid for everything.”

  Smith almost backed out of the conference, he told me. He was hoping to save one day for sightseeing, but Liu was demanding that he cancel his touring. Since Cui and China were picking up the Duke group’s expenses, Liu told him, he should devote all of his time to giving technical advice and talks to Cui’s group. The two of them argued in Chicago’s O’Hare International Airport as their flight to China was about to depart. “He began trying to add constraints and things to all of the invited speakers,” Smith recalled. “I told him that my understanding was that this was a conference and that we were invited speakers, and we weren’t there to work for Professor Cui. Ruopeng got upset and said that it was important for the program that we provide some technical input to Cui’s program.”

  Smith finally threatened to turn around and go back to North Carolina. Liu “clearly was under pressure from China to make our trip worthwhile,” Smith said.

  It’s likely that Chinese officials anticipated Smith’s full cooperation because Liu had persuaded him to join a program called Projec
t 111. Cui was associated with Project 111, and set up the metamaterials workshop under its aegis.

  Liu had explained to Smith that Project 111 would strengthen the collaboration with Cui and fund research. But he didn’t divulge its true purpose, or the commitment that Smith was making. China’s Ministry of Education and its State Administration of Foreign Experts Affairs established Project 111 in 2006 to spur “scientific renewal” of Chinese universities by recruiting renowned scientists as “overseas academic masters.” In return for travel fees, allowance, housing, and medical services, each master was expected to work at least one month at “innovation centers” established on Chinese campuses.

  Liu sought to enlist other Western metamaterials experts besides Smith in Project 111. One of them, a China-born professor at the University of California, became suspicious. He translated the contract into English and warned Smith that it required working in China. Smith went back to Liu, who assured him that everything was fine, and not to worry about it.

  “I was incredibly naïve about all this,” Smith said.

  * * *

  IN A NUMBER of high-profile cases, the U.S. government has wrongly accused Chinese-American scientists of economic espionage. Most notably, Professor Xi Xiaoxing, chairman of Temple University’s physics department, was cleared in 2015 of sending a secret design to China of a superconductor device known as a pocket heater, when it turned out that the blueprints were for entirely different equipment. A study by Thomas Nolan, a California defense attorney specializing in economic espionage cases, shows that people with Chinese surnames were sentenced on average to 32 months in prison for stealing trade secrets, as against 15 months for everyone else.

  False accusations and unfair sentencing are inexcusable. Nevertheless, they stem from a disturbing reality: foreign theft of American science and technology is running rampant, and China is the leading culprit. The 2013 report of the Commission on the Theft of American Intellectual Property, cochaired by former U.S. ambassador to China Jon M. Huntsman Jr., concluded that China accounts for between 50 percent and 80 percent of the more than $300 billion in intellectual property that the United States loses annually. Almost two-thirds of all economic espionage cases alleging a foreign destination for stolen trade secrets involve China. Of China-related cases between 1997 and 2016, twenty-four led to convictions or guilty pleas, three were reduced to lesser charges, eight were dropped or dismissed, and thirteen were pending.

  Project 111 is one of a vast array of Chinese “brain gain” programs that, intentionally or not, encourage theft of intellectual property from U.S. universities. These initiatives to attract overseas scientists, especially those born in China, offer such generous salaries, laboratory facilities, and other incentives that a borderline candidate may be tempted to improve his chances by bringing back somebody else’s data or ideas.

  Chinese recruitment programs “pose a serious threat to US businesses and universities through economic espionage and theft of IP,” according to a September 2015 FBI report. Koerner, the FBI’s former head of counterintelligence in Tampa, sums up the implicit message to Chinese researchers in the United States: “Don’t come home empty-handed.”

  In March 2013, Huajun Zhao, a research assistant at Medical College of Wisconsin, was arrested and charged with stealing three vials of a cancer-fighting compound from his professor, Marshall Anderson, who had patented it. Zhao, who claimed that he invented the compound and wanted to bring it to China for further study, had applied for funding from Chinese agencies that support research and overseas recruitment. One application was an “exact translation” of an old grant proposal by Anderson, according to the 2015 FBI report, which was distributed only to law enforcement and corporate security members of its Domestic Security Alliance Council. Zhao later pleaded guilty to a reduced charge of illegally downloading research data and was sentenced to the four-and-a-half months he’d already served plus two years’ probation.

  Since its inception in 1949, the People’s Republic of China has recognized the importance of foreign-trained scientists in accelerating technological progress. Established that same year, the Chinese Academy of Sciences “soon welcomed over 200 returning scientists who contributed to CAS the high-level expertise they had acquired abroad,” according to its website. Driven out of the United States as a suspected communist on flimsy evidence, California Institute of Technology rocket scientist Qian Xuesen returned to China in 1955 and built its space and missile programs.

  After the Cultural Revolution, when Chinese leader Deng Xiaoping decided to send throngs of students to the United States, he hoped that 90 percent would return and foster China’s technological prowess. Instead he spurred a brain drain that peaked after the 1989 Tiananmen Square massacre. Students who had opposed the crackdown feared prosecution in China, and the U.S. government allowed them to stay.

  In response, China’s national, provincial, and municipal governments embarked on aggressive efforts to lure back the most successful expatriates. Of the slew of initiatives, the best known are the Hundred Talents Program and the Thousand Talents Program. Hundred Talents seeks up-and-coming scholars under age forty. Thousand Talents, established in 2008 by the Communist Party’s powerful Organization Department, woos prominent professors of Chinese ethnicity under age fifty-five. Aside from salaries, laboratories, and research funds, the perks include housing, medical care, jobs for spouses, and top schools for children. The government also rewards Chinese universities for landing foreign experts.

  “In fact, the Chinese government has been the most assertive government in the world in introducing policies targeted at triggering a reverse brain drain,” David Zweig, a professor at Hong Kong University of Science and Technology, and Huiyao Wang, director general of the Center for China and Globalization in Beijing, wrote in 2012.

  Such initiatives have attracted numerous foreign scientists. In the 296 national research programs approved by China’s Ministry of Science and Technology from 2006 to 2011, 47 percent of the chief scientists earned their doctorates abroad, and 32 percent came to China via the Thousand Talents Program. Of Chinese students who received science and engineering doctorates at American universities in 1996, 98 percent remained in the United States in 2001. Among the corresponding cohort a decade later, 15 percent left the United States, likely due to China’s enticements.

  Hong Ding was one of the Thousand Talent Program’s biggest prizes. The tenured full professor of physics at Boston College accepted what Ding called a “very attractive” package, including a $147,000 relocation allowance, to lead two projects for the Chinese Academy of Sciences.

  Boston College faculty “were shocked,” Ding told China Daily. “People thought staying in the US was good for my career. But I wanted to contribute to basic research going on in China.” China made Ding “unbelievable offers of salary and equipment,” a Boston College physics professor told me. “He was an up and coming star.”

  Still, China has snared relatively few scientists of Ding’s status from U.S. universities. Most are reluctant to uproot families and leave tenured sinecures in the creative disorder of American academia to work in an authoritarian society. As a result, Thousand Talents and similar programs have modified their rules to allow recruits to keep their jobs abroad and come to China two or three months a year, much like Smith’s contract with Project 111. Yitang Zhang, a University of California, Santa Barbara, math professor who has won a MacArthur “genius” award, rejected a full-time offer from the Chinese Academy of Sciences in 2013. Instead, Zhang, a former Chinese dissident, teaches graduate students at CAS in Beijing for two months in the summer. The Chinese have a phrase for such summer returnees: “migratory birds.”

  Even a part-time commitment was too much for Xiaochun He, a nuclear physicist at Georgia State University and Brookhaven National Laboratory. For both family and career reasons, he rejected a Thousand Talents offer to spend three months a year at a Chinese university with a higher salary and bigger laboratory.
<
br />   “My friends in China, they have a whole floor of lab space,” he told me. “My biggest concern is the research environment, and whether it’s dominated by government policy instead of my free will. I’m a scientist. I do what I’m interested in.”

  Disdained by elite scientists, the talent programs appeal primarily to those on the fringes of American academia: untenured or adjunct professors, postdoctoral students with uncertain prospects. They realize what China wants—rapid progress in fields where it lags behind the West—and may be tempted to enhance their credentials by plundering American research.

  “The structure, particularly of the Hundred Talents, encouraged people to bring back technology that was not available in China if they wanted to get a position,” Zweig told me. “People do plan to go out looking for something they know is missing. They know that’s their comparative advantage.”

  * * *

  LATE IN 2008, Liu and a Duke graduate student in statistics, Chunlin Ji, developed a new invisibility cloak. Ji wrote the computer code for the design, and Liu translated the algorithms into the layout of wires needed to make the cloak. It could hide a bump on a flat surface from a broad spectrum of wave frequencies, bringing tantalizingly closer to fulfillment the dream of making objects invisible to the human eye.

  Smith was pleased. Still, after the paper was submitted to Science, he asked Liu, the lead author, for a clearer explanation of the breakthrough. “I really had wanted him to be able to present—for his own education—all of the details in his own words,” Smith said. Liu stalled him, and Smith wondered why. “This is the point where I started becoming very suspicious, because Ruopeng would never produce the technique by which he achieved the design of the cloak. I probably asked him fifty times, and he kept saying he’d get to it. It was okay for a while, as we were finishing up the paper and it clearly worked.… But he kept stonewalling.”