The nuclear power industry has a reputation for resisting innovative changes. But Leslie Dewan and a colleague have dared to invent a new type of nuclear reactor. “We were feeling on top of the world. We just passed our qualifying exams for our PhDs,” she says. “We thought, ‘We’re the smartest we’ve been in our lives. We can do anything. Let’s change the world with nuclear.’” Two years later, she’d designed a reactor that solves the main problems facing nuclear power. To commercialize it, she’d cofounded a startup, Transatomic Power.
For decades the nuclear industry has built one type of reactor, called a light-water reactor, almost exclusively. There are significant problems with that technology, which uses ordinary water to cool the fuel rods in which the nuclear reaction takes place. It requires expensive safeguards against a radiation-releasing meltdown if the fuel rods overheat; it produces waste products that are dangerous for 100,000 years. Dewan and a fellow graduate student, Mark Massie, designed an alternative based on molten-salt reactors that were originally proposed in the 1950s as a way to power aircraft. Though nuclear planes never became a reality, the reactor design has several key advantages. For one thing, it can be readily modified so that rather than producing large amounts of waste, it reuses much of the spent nuclear material as fuel.
It is also far safer than the light-water reactors, which require a constant source of electricity to pump in cool water and prevent the runaway nuclear reactions that lead to meltdowns. Molten salt serves as the coolant; it’s mixed with the nuclear materials, so the reactions take place right in the liquid. The heat of those reactions keeps the salt molten. A plug at the bottom of the reaction vessel is made of the same salt, kept solid by cooling it; if the plant’s electricity supply is lost, the plug warms up and liquefies, allowing the contents of the reactor to drain into a large containment tub and spread out so that the nuclear chain reactions come to an almost complete stop. The nuclear material and molten salt then cool down and turn into a contained solid that poses no danger of a meltdown.
The technology had one glaring problem, though: the reactors were large and, thus, expensive for the amount of power they produced. Dewan found a solution. “We realized that with some relatively modest changes to molten-salt reactors we could make them much more power dense and therefore a lot cheaper,” she says. She introduced new materials and a new shape that allowed her to increase power output by 30 times. As a result, the reactor is now so compact that a version large enough for a power plant can be built in a factory and shipped by rail to a plant site, which is potentially cheaper than the current practice of building nuclear reactors on site.
The reactor also makes more efficient use of the energy in nuclear fuel. It can consume about one ton of nuclear waste a year, leaving just four kilograms behind. Dewan’s name for the technology: the Waste-Annihilating Molten-Salt Reactor.
So far, the design exists as a 180-page document, computer simulations, and patent filings. Dewan has designed five experiments, each of which will cost about $1 million, to prove key aspects of the design. If those go well, she’ll still face a decade or more of further tests and U.S. federal certifications that could cost hundreds of millions of dollars. And she suggests that the future for new nuclear-power technology might not be in the United States. She points in particular to China, which is spending far more on new reactor designs and on the construction of nuclear plants.
But though it will be a long and uncertain route to commercialization of the technology, Dewan is driven by what is at stake. She’s part of a new generation of young researchers who see nuclear energy as one of the best hopes for averting disastrous climate change. Dewan originally looked to solar and wind power as ways to reduce carbon dioxide emissions, but “then I looked at the numbers,” she says. “I realized that nuclear power is the best low-carbon energy source that’s available and scalable.”
Caltech professor John Dabiri uses his engineering expertise to try to understand how animals move in their natural environments. While researching the swimming patterns of fish, he recently came to a surprising insight: the way we’re thinking about wind power—specifically, the design of wind farms—is wrong.
Conventional wind farms are designed to minimize the turbulence caused by interactions between turbines. That creates an obvious problem, says Dabiri: “You space them out as far as possible. If you’re talking about a wind turbine that has a 100-meter diameter, then you’re talking about as much as a mile between wind turbines. That’s a lot of space that could be used to generate electricity, but can’t be because of these turbulent interactions.”
Dabiri thought of a solution while researching how fish form schools to minimize drag as they move about. “Fish can reduce the amount of energy that they use if they swim in certain coördinated arrangements as opposed to swimming alone,” he explains. “In fact, fish in large schools form precise, repeating patterns that allow them to move most efficiently. There’s some basic fluid-mechanics theory that you can use to explain why that might be the case. Jotting down the math for urban wind-turbine analysis, there was sort of a eureka moment where I realized that the equations were exactly the same equations that explain fish schooling.
“Why not use how fish form schools as a starting point for understanding how to design wind farms?” asks Dabiri. “We began to use the same tools that were used to determine the optimal configuration for fish schools to optimize the wind farm. We looked at an arrangement that’s been identified as optimal for fish, and we found that if we, in our computer models, arranged our wind turbines exactly in the same kind of diamond pattern that fish form, you get significant benefits in the performance of a wind farm.”
To maximize that performance, Dabiri would use vertical wind turbines, which have been around for years but are much less common than the familiar horizontal–axis turbines. Vertical turbines can perform better when they are packed together—at least if they are arranged in the optimal pattern Dabiri discovered. That raises the possibility of redesigning wind farms to increase the amount of power they produce and lower the cost. Dabiri says the turbines could be squeezed into existing wind farms so that they produce more power without taking up any more land. It’s a solution that could greatly reduce the drag on an industry that often seems to be swimming upstream.
When Roozbeh Ghaffari was five years old, his only sibling—a brother named Soran—was born three months prematurely. A few things would eventually emerge about Soran: he was blind and mildly intellectually impaired, he had remarkably acute hearing (and perfect pitch), and he was his older brother’s best friend, superfan, and inspiration. As the elder Ghaffari became an expert in the science and technology of body-machine interfaces and devices that can be integrated into the body (he now leads advanced technology development at MC10, a startup in Cambridge, Massachusetts), Soran remained physically at home in Los Angeles but frequently at his brother’s virtual side, standing by for nightly telephone updates and reading up on his work using text-to-speech software.
Indeed, while Roozbeh Ghaffari’s lifelong interest in the merger of biology and engineering was shaped partly by his parents—his mother is a microbiologist and his father an architect—it was inspired mainly by his brother, whose blindness was caused by retinal damage from excessive oxygen exposure in the neonatal intensive-care ward. “What I found was that my goals were all driven by him,” he says. “I wanted to work on the retinal implant project at MIT, right from year one as an undergraduate.”
He didn’t succeed at finding funding for that daunting challenge. But as a graduate student in the Harvard-MIT Division of Health Sciences and Technology, he shifted to another critical problem, this time inspired by his brother’s sharp hearing rather than his deficient eyesight. Ghaffari’s goal was to unravel the mysteries of the cochlea—that “biological black box with thousands of moving parts in a fluid” that transforms vibrations in the inner ear into nerve signals. “How do we hear with such remarkable sensitivity?” he asks. “How can we process both the roaring sound of jet engines and the sound of pins dropping?”
Focusing on a cochlear structure called the tectorial membrane, he managed to build a system that could measure what the structure was actually doing. “It supports a traveling wave of energy that can propagate along the cochlea. That hadn’t been known before,” he says. And this could help explain how the human ear can detect both very loud and very soft sounds, as well as a wide range of pitches.
After taking a Harvard Business School class focusing on commercializing science, and meeting venture capitalist Carmichael Roberts and postdocs in the lab of Harvard chemist George Whitesides, Ghaffari helped develop a business plan for a company called Diagnostics for All, which is commercializing paper-based diagnostics invented in Whitesides’s lab.
Next, with Roberts and Whitesides serving as the matchmakers, Ghaffari met John Rogers, a materials scientist at the University of Illinois. Rogers was fabricating stretchable electronic devices using polymers and ultrathin semiconductors, such as silicon. But the technology was looking for an important problem to solve. So in 2008 Ghaffari was brought on as cofounder of MC10. The founders considered what kinds of flexible or stretchable products they might enhance with electronics (they even considered contact lenses), but within a couple of months they had settled on balloon catheters and health-monitoring skin patches.
Today, the devices under development add electronics and sensors to balloon catheters. Existing versions of these devices are snaked into coronary arteries and inflated to compress accumulated plaques that can block blood flow. The new versions can, among other things, sense misfiring cardiac tissue that causes irregular heartbeats called arrhythmias. They can even ablate tiny patches of such tissue without harming the healthy tissue nearby. As always, Soran is eager to hear all about it. “He’ll go look up ‘ventricular tachycardia’ and grill me night and day: ‘What is this disorder? What are you guys doing?’” Ghaffari says.
Someday Ghaffari may yet build what his brother needs: a bionic replacement for his damaged retina. In the meantime, he is finding new ways to create other devices that promise to help others. While its catheters still need regulatory approval, MC10 has launched a thin $150 cap that athletes, such as football or hockey players, can wear inside their helmets to indicate the severity of blows to the head. By lighting up red, yellow, or green LEDs on the cap, the technology could indicate whether the wearer might have suffered a worrisome head impact.
When Soran comes to Boston, Ghaffari brings him to the MC10 lab and lets him hold and feel the electronic skullcaps and the instrumented catheters, with their intricate patterns of ultrathin, stretchable sensors. But even when Soran is home in Los Angeles, Ghaffari says, the following question he inspires is in the back of his mind every day: “How can we turn technology into something useful that integrates with the human body?” And every evening, Soran is on the other end of the line, helping him answer that question.
Innovation: Kuniharu Takei, a professor at Japan’s Osaka Prefecture University, has led the development of cheap and robust methods for “printing” uniform, ultrathin patterns of different types of nanoelectronics on a wide range of surfaces.
Why it matters: Nanoscale components made of materials other than silicon could lead to more versatile, less expensive electronic devices. Transistors made from so-called compound semiconductors, for instance, could be up to twice as fast and 10 times as energy efficient as silicon transistors.
Takei’s goal is to build circuits and sensor networks that simultaneously exploit the properties of several materials, each chosen because it offers a specific advantage. Nanomaterials made of compound semiconductors could be used to add high-speed radio-frequency components and efficient light emitters to silicon chips. But there is not yet a way to cheaply and reliably add such nanoscale components. Existing strategies involve highly specialized procedures for growing these materials on silicon or attaching them to silicon wafers; such methods are expensive and may not be practical for manufacturing. Printing processes like Takei’s could be an attractive alternative.
Methods: In the process he uses to print compound-semiconductor nanomaterials, Takei grows thin films of the chosen material on a suitable substrate, uses a lithography technique to create strips in the material, and releases the patterns from the substrate with a chemical etchant. He can then transfer the nanomaterial to a range of new surfaces, including silicon wafers and bendable plastics, by using a silicone rubber stamp that picks up the material and prints it.
Next steps: Takei’s printing methods could be used to produce electronic devices that exploit the properties of multiple materials. For example, he says, organic light-emitting diodes could be combined with transistors made of inorganic nanomaterials to make low-power, bendable displays. He’s now working on a smart bandage that would be able to sense and respond to things like glucose level and skin temperature.
Problem: Scientists have worked for years to increase the longevity of targeted drugs, which promise to deliver treatment to a specific tissue within the body. These targeted treatments require new drug carriers such as polymers that are designed to evade the immune system. But too often, these carriers are destroyed before the drug can effectively target tumors and other localized sites of disease. Though the body’s own cells are protected from the immune system by their protein-studded outer membrane, it’s not possible to re-create this complex matrix for synthetic particles used in drug delivery.
Solution: Why not cloak therapeutics in natural membranes? That’s the idea of Liangfang Zhang, a nanoengineering professor at UC San Diego.
Zhang derives red-blood-cell membranes from blood samples and uses them to coat polymer nanoparticles. Because these particles look like red blood cells on the surface, they can fool the immune system; loaded with drugs, they serve as robust and long-lived drug carriers. An unexpected bonus: they can also act like nanoscale sponges to suck up toxic proteins produced by infectious bacteria or introduced by snake or insect venom. If the particles flood the bloodstream, they will divert most of the toxin away from actual cells.
Born in Wuwei County, 45 minutes by plane from Shanghai, Zhang left home for the prestigious Tsinghua University in Beijing when he was just 15. By the time he was 20 years old, he could have opened a factory to produce exceptionally tough rubber materials he’d helped invent as a student. But Zhang says he “didn’t want to run a rubber factory all my life.” And he knew if he started a factory, some other young upstart would come up with a better technology and he might not be able to compete. So he decided to pursue an advanced graduate degree in engineering in the United States. Despite his accomplishments as a scientist, however, he has never lost his desire to turn laboratory advances into practical breakthroughs.
“There have been a lot of taboos about psychiatric diseases,” says Feng Zhang. “People would think depressed people are not mentally strong enough. But that’s not true. In this and the next decade, we will learn much more about the mechanisms that lead to these neurological problems. And that will change our way of interacting with these people, and it will also change how we can treat them.”
Zhang is an assistant professor at MIT and one of only 11 core faculty members at the Broad Institute, a leading center of genomic research. He’s spent much of his brief but impressive career developing tools to understand how the brain functions, including what goes wrong in people with mental illnesses. As a graduate student at Stanford, he played a key role in developing optogenetics, which uses light to affect the behavior of living animals by controlling specific neurons; he then used the technique in mice to pinpoint brain cells that are associated with depression.
But truly understanding the genetics of mental illnesses will mean identifying the mutations causing the abnormal behavior. After getting his PhD, Zhang invented two new ways to “edit” animal genomes that were far cheaper and more effective than the existing technology. One method in particular, called CRISPR, promises to change how genomic engineering is done. It allows researchers to precisely snip out a short sequence of DNA so that they can substitute other genetic material or simply delete the sequence.
By inserting genetic mutations that others have linked to autism and schizophrenia into human stem cells that mature into neurons, Zhang is able to create brain cells with the specific genetic errors linked to those conditions. This makes it possible to study the abnormal cells directly: Do the neurons look different? Are there biochemical clues to what is going wrong? He has also engineered mice with the mutations to study how the changes affect behavior. Such research could not only help identify the causes of the disorders but suggest ways to identify and test drugs to treat them—and his genome-editing tools may even one day provide a way to “fix” the mutations.
Zhang has been interested in ways to “repair” diseases since he was in high school in Iowa, when he spent every afternoon working with a medical researcher at the Human Gene Therapy Research Institute, a part of Methodist Hospital in Des Moines. Though the gene therapies available at the time turned out to be too risky for widespread use in humans, Zhang never gave up hope of finding ways to directly fix the genetic mutations behind many diseases, using the increasing capabilities of genomic engineering. These days, bolstered by the success of his editing tools and other genomic advances, he is working to translate the technology into actual human therapies and exploring opportunities to start a company.
Bowen Zhao dropped out of Beijing’s top high school to take a job at BGI-Shenzhen, the world’s largest DNA-sequencing organization. Soon after joining the company, he became involved in a new research effort: investigating the genetic basis of human cognitive abilities, including intelligence. “We want to know the genetic basis of IQ,” he says. Zhao thinks human intelligence is from 40 to 80 percent inheritable, and he wants to know which genes may influence the trait he calls “high cognitive ability.”
Zhao’s team is sequencing the DNA of more than 2,000 people with high IQs. Zhao is not looking for an IQ gene; rather, he expects to pinpoint multiple small variations in thousands of genes that shape the inheritable aspect of intelligence. Perhaps uniquely in the world, BGI has both the massive computing power and the manpower to handle a data–intensive approach to combing through the genetic clues. “We’re data driven, not hypothesis driven,” says Zhao.
The project involves sequencing more than six trillion DNA bases. This is not the first attempt to map the biological roots of human intelligence. But now, Zhao points out, DNA sequencing technology is so advanced that it’s possible to sequence and compare thousands of minute variations in extremely large samples.
Zhao is keenly aware that research into the heritability of intelligence is controversial and fraught with ethical dangers. But he says that it is far too early to make any decisions or judgments based on his genomic studies. For the foreseeable future, he adds, if you want to identify high-IQ individuals, it will be far easier and more accurate to conduct a standard IQ test than to sequence the person’s DNA.
Stanford professor Xiaolin Zheng often works in the esoteric fringes of nanoscience, but she also likes to find simple ways to fabricate complex materials that can be put to use in practical applications like solar-fuel systems, solar cells, and batteries. Last year she created solar cells in the form of flexible stickers—only a 10th as thick as plastic wrap—that can be applied to a window, a piece of paper, the back of a mobile phone, or anything else you want. These solar cells produce just as much electricity as rigid ones made of the same materials.
Zheng got the inspiration for this invention from her father. One day when they were talking on the phone—he in China, she in California—he said that it should be possible to put solar cells on the walls of buildings, not just the roof. And Zheng’s daughter, like many kids, loves stickers.
All this was in the back of Zheng’s mind when she read a research paper about graphene, a novel type of nanomaterial. The researchers grew the material on a layer of nickel on top of a silicon wafer. When they put the whole thing in water, the nickel separated from the surface, taking the graphene with it. “I couldn’t believe that soaking in water would do this,” she says.
Zheng has demonstrated this water–soaking approach as a way to peel off thin-film silicon solar cells grown on a rigid substrate. It turns out the phenomenon—called water-assisted subcritical debonding—had been known since the 1960s, but no one before had tried using it to make flexible electronics. She hopes the technology will be scaled up beyond the one-square-centimeter devices she’s made so far, so that the sides of buildings can one day be papered with solar cells as her father suggested.