Mysterious force harnessed in a silicon chip

By Catherine Zandonella for the Office of the Dean for Research

Getting something from nothing sounds like a good deal, so for years scientists have been trying to exploit the tiny amount of energy found in nearly empty space. It’s a source of energy so obscure it was once derided as a source of “perpetual motion.” Now, a research team including Princeton scientists has found a way to harness this energy using a silicon-chip device, potentially enabling applications.

This energy, predicted seven decades ago by the Dutch scientist Hendrik Casimir, arises from quantum effects and can be seen experimentally by placing two opposing plates very close to each other in a vacuum. At close range, the plates attract each other. Until recently, however, harnessing this “Casimir force” to do anything useful seemed impossible.

A new silicon chip built by researchers at Hong Kong University of Science and Technology and Princeton University is a step toward harnessing the Casimir force. Using a clever assembly of micron-sized shapes etched into the plates, the researchers demonstrated that the plates can instead repel each other as they are brought close together. Constructing this device entirely out of a single silicon chip could open the way to using the Casimir force for practical applications such as keeping tiny machine parts from sticking to each other. The work was published in the February issue of the journal Nature Photonics.

Energy of a vacuum

Image of a Casimir-on-a-chip device
Researchers created a silicon device that enabled them to observe the Casimir force. (Image credit: Nature Photonics)

“This is among the first experimental verifications of the Casimir effect on a silicon chip,” said Alejandro Rodriguez, an assistant professor of electrical engineering at Princeton University, who provided theoretical calculations for the device, which was built by a team led by Ho Bun Chan at Hong Kong University of Science and Technology. “And it also allows you to make measurements of forces in very nontrivial structures like these that cause repulsion. It is a double-whammy.”

The silicon structure looks like two plates lined with teeth that face each other across a tiny gap which is only about 100 nanometers wide. (A human hair is 60,000-80,000 nanometers wide.) As the two plates are pushed closer together, the Casimir force comes into play and pushes them apart.

This repulsive effect happens without any input of energy and to all appearances, in a vacuum. These characteristics led this energy to be called “zero-point energy.” They also fueled earlier claims that the Casimir force could not exist because its existence would imply some sort of perpetual motion, which would be impossible according to the laws of physics.

The force, which has since been experimentally confirmed to exist, arises from the normal quantum fluctuations of the few atoms that persist in the chasm despite the evacuation of all the air.

The team demonstrated that it is possible to build a device in silicon to control the Casimir force.

“Our paper shows that it is possible to control the Casimir force using structures of complex, tailor-made shapes,” said Ho Bun Chan, senior author on the paper and a scientist at the Hong Kong University of Science and Technology. His team drew on earlier work by Rodriguez published in 2008 that proposed shapes that would be expected to yield a Casimir force that could both attract and repel. “This paper is the experimental realization using a structure inspired by Rodriguez’s design,” Chan said.

Rodriguez and his team at Princeton developed techniques that allowed the researchers to compute interactions between two parallel plates as they approach each other. With these tools, they were then able to explore what would happen if more complex geometries were used. This led to some of the first predictions of a repulsive Casimir force in 2008.

The Rodriguez group used nanophotonic techniques, which involved measuring how light would interact with the structures, to get at the complex equations of how the force arises from the interaction of two plates.

The silicon device included a small mechanical spring that the researchers used to measure the force between the two plates, and to verify that the quantum force can be repulsive. The roughly T-shaped silicon teeth are what allow the repulsive force to form. The repulsion comes from how different parts of the surface interact with the opposite surface.

“We tried to think about what kind of shapes Chan’s group would have to fabricate to lead to a significant repulsive force, so we did some background studies and calculations to make sure they would see enough non-monotonicity as to be measurable,” Rodriguez said.

Going forward, the researchers plan to explore other configurations that may give rise to even larger repulsive forces and more well-defined repulsion at larger separations.

Funding for the study came from the Research Grants Council of Hong Kong and the National Science Foundation (grant no. DMR-1454836).

The paper, “Measurement of non-monotonic Casimir forces between silicon nanostructures,” by L. Tang, M. Wang, C. Y. Ng, M. Nikolic, C. T. Chan, A. W. Rodriguez and H. B. Chan was published in the journal Nature Photonics online Jan. 9, 2017 and in the February 2017 issue. Nature Photonics 97–101(2017) doi:10.1038/nphoton.2016.254.

Artificial topological matter opens new research directions

By Catherine Zandonella, Office of the Dean for Research

An international team of researchers have created a new structure that allows the tuning of topological properties in such a way as to turn on or off these unique behaviors. The structure could open up possibilities for new explorations into the properties of topological states of matter.

“This is an exciting new direction in topological matter research,” said M. Zahid Hasan, professor of physics at Princeton University and an investigator at Lawrence Berkeley National Laboratory in California who led the study, which was published March 24th in the journal Science Advances. “We are engineering new topological states that do not occur naturally, opening up numerous exotic possibilities for controlling the behaviors of these materials.”

The new structure consists of alternating layers of topological and normal, or trivial, insulators, an architecture that allows the researchers to turn on or off the flow of current through the structure. The ability to control the current suggests possibilities for circuits based on topological behaviors, but perhaps more importantly presents a new artificial crystal lattice structure for studying quantum behaviors.

Theories behind the topological properties of matter were the subject of the 2016 Nobel Prize in physics awarded to Princeton University’s F. Duncan Haldane and two other scientists. One class of matter is topological insulators, which are insulators on the inside but allow current to flow without resistance on the surfaces.

In the new structure, interfaces between the layers create a one-dimensional lattice in which topological states can exist. The one-dimensional nature of the lattice can be thought of as if one were to cut into the material and remove a very thin slice, and then look at the thin edge of the slice. This one-dimensional lattice resembles a chain of artificial atoms. This behavior is emergent because it arises only when many layers are stacked together.

Artificial topological matter
The researchers made different samples where they could control how the electrons tunnel from interface to interface through alternating layers of trivial and topological insulators, forming an emergent, tunable one-dimensional quantum lattice. The top panel (A, B, C, and D) shows a structure where the trivial layer is relatively thin, enabling electron-like particles to tunnel through the layers (topological phase). The bottom panel (G, H, I, and J) shows a structure where the trivial insulator is relatively thick and blocks tunneling (trivial phase). (Image courtesy of Science/AAAS)

By changing the composition of the layers, the researchers can control the hopping of electron-like particles, called Dirac fermions, through the material. For example, by making the trivial-insulator layer relatively thick – still only about four nanometers – the Dirac fermions cannot travel through it, making the entire structure effectively a trivial insulator. However, if the trivial-insulator layer is thin – about one nanometer – the Dirac fermions can tunnel from one topological layer to the next.

To fashion the two materials, the Princeton team worked with researchers at Rutgers University led by Seongshik Oh, associate professor of physics, who in collaboration with Hasan and others showed in 2012 that adding indium to a topological insulator, bismuth selenide, caused it to become a trivial insulator. Prior to that, bismuth selenide (Bi2Se3) was theoretically and experimentally identified as a topological insulator by Hasan’s team, a finding which was published in the journal Nature in 2009.

“We had shown that, depending on how much indium you add, the resulting material had this nice tunable property from trivial to topological insulator,” Oh said, referring to the work published in Physical Review Letters in 2012.

Graduate students Ilya Belopolski of Princeton and Nikesh Koirala of Rutgers combined two state-of-the-art techniques with new instrumentation development and worked together on layering these two materials, bismuth selenide and indium bismuth selenide, to design the optimal structure. One of the challenges was getting the lattice structures of the two materials to match up so that the Dirac fermions can hop from one layer to the next. Belopolski and Suyang Xu worked with colleagues at Princeton University, Lawrence Berkeley National Laboratory and multiple institutions to use high resolution angle-resolved photoemission spectroscopy  to optimize the behavior of the Dirac fermions based on a growth to measurement feedback loop.

Photo of research team
Princeton research team from L to R: Guang Bian, M. Zahid Hasan, Nasser Alidoust, Hao Zheng, Daniel Sanchez, Suyang Xu and Ilya Belopolski (Image credit: Princeton University)

Although no topologically similar states exist naturally, the researchers note that analogous behavior can be found in a chain of polyacetylene, which is a known model of one-dimensional topological behavior as described by the 1979 Su-Schrieffer-Heeger’s theoretical model of an organic polymer.

The research presents a foray into making artificial topological materials, Hasan said. “In nature, whatever a material is, topological insulator or not, you are stuck with that,” Hasan said. “Here we are tuning the system in a way that we can decide in which phase it should exist; we can design the topological behavior.”

The ability to control the travel of light-like Dirac fermions could eventually lead future researchers to harness the resistance-less flow of current seen in topological materials. “These types of topologically tunable heterostructures are a step toward applications, making devices where topological effects can be utilized,” Hasan said.

The Hasan group plans to further explore ways to tune the thickness and explore the topological states in connection to the quantum Hall effect, superconductivity, magnetism, and Majorana and Weyl fermion states of matter.

In addition to work done at Princeton and Rutgers, the research featured contributions from the following institutions: South University of Science and Technology of China; Swiss Light Source, Paul Scherrer Institute; National University of Singapore; University of Central Florida; Universität Würzburg; Diamond Light Source, Didcot, U.K.; and Synchrotron SOLEIL, Saint-Aubin, France.

Work at Princeton University and synchrotron-based ARPES measurements led by Princeton researchers were supported by the U.S. Department of Energy under Basic Energy AQ29 Sciences grant no. DE-FG-02-05ER46200 (to M.Z.H.). I.B. was supported by an NSF Graduate Research Fellowship. N.K., M.B., and S.O. were supported by the Emergent Phenomena in Quantum Systems Initiative of the Gordon and Betty Moore Foundation under grant no. GBMF4418 and by the NSF under grant no. NSF-EFMA-1542798. H.L. acknowledges support from the Singapore National Research Foundation under award no. NRF-NRFF2013-03. M.N. was supported by start-up funds from the University of Central Florida. The work acknowledges support of Diamond Light Source, Didcot, U.K., for time on beamline I05 under proposal SI11742-1. Some measurements were carried out at the ADRESS beamline (24) of the Swiss Light Source, Paul Scherrer Institute, Switzerland. This study was in part supported by grant AQ30 no. 11504159 of the National Natural Science Foundation of China (NSFC), grant no. 2016A030313650 of NSFC Guangdong, and project no. JCY20150630145302240 of the Shenzhen Science and Technology Innovations Committee.

The paper, “A novel artificial condensed matter lattice and a new platform for one-dimensional topological phases,” by Ilya Belopolski, Su-Yang Xu, Nikesh Koirala, Chang Liu, Guang Bian, Vladimir Strocov, Guoqing Chang, Madhab Neupane, Nasser Alidoust, Daniel Sanchez, Hao Zheng, Matthew Brahlek, Victor Rogalev, Timur Kim, Nicholas C. Plumb, Chaoyu Chen, François Bertran, Patrick Le Fèvre, Amina Taleb-Ibrahimi, Maria-Carmen Asensio, Ming Shi, Hsin Lin, Moritz Hoesch, Seongshik Oh and M. Zahid Hasan, was published in the journal Science Advances on March 24, 2017. (Belopolski et al., Sci. Adv. 2017;3: e1501692 24 March 2017)


Study reveals the multitasking secrets of an RNA-binding protein

RNA-binding domains
Two views of one of Glo’s RNA-binding domains highlight the amino acids required for binding G-tract RNA (left) and U-A stem structures (right). Courtesy of Cell Reports.

By Staff, Department of Molecular Biology

Researchers from Princeton University and the National Institute of Environmental Health Sciences have discovered how a fruit fly protein binds and regulates two different types of RNA target sequence. The study, published April 4 in the journal Cell Reports, may help explain how various RNA-binding proteins, many of which are implicated in cancer and neurodegenerative disease, perform so many different functions in the cell.

There are hundreds of RNA-binding proteins in the human genome that together regulate the processing, turnover and localization of the many thousands of RNA molecules expressed in cells. These proteins also control the translation of RNA into proteins. RNA-binding proteins are crucial for maintaining normal cellular function, and defects in this family of proteins can lead to disease. For example, RNA-binding proteins are overexpressed in many human cancers, and mutations in some of these proteins have been linked to neurological and neurodegenerative disorders such as amyotrophic lateral sclerosis. “Understanding the fundamental properties of this class of proteins is very relevant,” said Elizabeth Gavis, the Damon B. Pfeiffer Professor in the Life Sciences and a professor of molecular biology.

Gavis and colleagues are particularly interested in a protein called Glorund (Glo), a type of RNA-binding protein that performs several functions in fruit fly development. This protein was originally identified due to its ability to repress the translation of an RNA molecule called nanos to protein in fly eggs. By binding to a stem structure formed by uracil and adenine nucleotides in the nanos RNA, Glo prevents the production of Nanos protein at the front of the embryo, a step that enables the fly’s head to form properly.

Like many other RNA-binding proteins, however, Glo is multifunctional. It regulates several other steps in fly development, apparently by binding to RNAs other than nanos. The mammalian counterparts of Glo, known as heterogeneous nuclear ribonucleoprotein (hnRNP) F/H proteins, bind to RNAs containing stretches of guanine nucleotides known as G-tracts, and, rather than repressing translation, mammalian hnRNP F/H proteins regulate processes such as RNA splicing, in which RNAs are rearranged to produce alternative versions of the proteins they encode.

To understand how Glo might bind to diverse RNAs and regulate them in different ways, Gavis and graduate student Joel Tamayo collaborated with Traci Tanaka Hall and Takamasa Teramoto from the National Institute of Environmental Health Sciences to generate X-ray crystallographic structures of Glo’s three RNA-binding domains. As expected, the three domains were almost identical to the corresponding domains of mammalian hnRNP F/H proteins. They retained, for example, the amino acid residues that bind to G-tract RNA, and the researchers confirmed that, like their mammalian counterparts, each RNA-binding domain of Glo can bind to this type of RNA sequence.

However, the researchers also saw something new. “When we looked at the structures, we realized that there were also some basic amino acids that projected from a different part of the RNA-binding domains that could be involved in contacting RNA,” Gavis explained.

The researchers found that these basic amino acids mediate binding to uracil-adenine (U-A) stem structures like the one found in nanos RNA. Each of Glo’s RNA-binding domains therefore contains two distinct binding surfaces that interact with different types of RNA target sequence. “While there have been examples previously of RNA-binding proteins that carry more than one binding domain, each with a different specificity, this represents the first example of a single domain harboring two different specificities,” said Howard Lipshitz, a professor of molecular genetics at the University of Toronto who was not involved in the study.

To investigate which of Glo’s two RNA-binding modes was required for its different functions in flies, Gavis and colleagues generated insects carrying mutant versions of the RNA-binding protein. Glo’s ability to repress nanos translation during egg development required both of the protein’s RNA-binding modes. The researchers discovered that, as well as binding the U-A stem in the nanos RNA, Glo also recognized a nearby G-tract sequence. But Glo’s ability to regulate other RNAs at different developmental stages only depended on the protein’s capacity to bind G-tracts.

“We think that the binding mode may correlate with Glo’s activity towards a particular RNA,” said Gavis. “If it binds to a G-tract, Glo might promote RNA splicing. If it simultaneously binds to both a G-tract and a U-A stem, Glo acts as a translational repressor.”

The RNA-binding domains of mammalian hnRNP F/H proteins probably have a similar ability to bind two different types of RNA, allowing them to regulate diverse target RNAs within the cell. “This paper represents an exciting advance in a field that has become increasingly important with the discovery that defects in RNA-binding proteins contribute to human diseases such as metabolic disorders, cancer and neurodegeneration,” Lipshitz said. “Since these proteins are evolutionarily conserved from fruit flies to humans, experiments of this type tell us a lot about how their human versions normally work or can go wrong.”

The research was supported in part by a National Science Foundation Graduate Research Fellowship (DGE 1148900), a Japan Society for the Promotion of Science fellowship, the National Institutes of Health (R01 GM061107) and the Intramural Research Program of the National Institute of Environmental Health Sciences. The Advanced Photon Source used for this study is supported by the US Department of Energy, Office of Science, Office of Basic Energy Sciences, under contract W-31-109-Eng-38.

The study, “The Drosophila hnRNP F/H Homolog Glorund Uses Two Distinct RNA-binding Modes to Diversify Target Recognition,” by Joel Tamayo, Takamasa Teramoto, Seema Chatterjee, Traci Tanaka Hall, and Elizabeth Gavis, was published in the journal Cell Reports on April 4, 2017.