Electronic life on the edge: Scientists discover the edge states of graphene nanoribbons

A scanning tunneling microscope determines the topography and orientation of the graphene nanoribbons on the atomic scale. In spectroscopy mode, it determines changes in the density of electronic states, from the nanoribbon's interior to its edge. Credit: Crommie et al, Lawrence Berkeley National Laboratory

As far back as the 1990s, long before anyone had actually isolated graphene ? a honeycomb lattice of carbon just one atom thick ? theorists were predicting extraordinary properties at the edges of graphene nanoribbons. Now physicists at the U.S. Department of Energy's Lawrence Berkeley National Laboratory (Berkeley Lab), and their colleagues at the University of California at Berkeley, Stanford University, and other institutions, have made the first precise measurements of the "edge states" of well-ordered nanoribbons.

A graphene nanoribbon is a strip of graphene that may be only a few nanometers wide (a nanometer is a billionth of a meter). Theorists have envisioned that nanoribbons, depending on their width and the angle at which they are cut, would have unique electronic, magnetic, and optical features, including band gaps like those in semiconductors, which sheet graphene doesn't have.

"Until now no one has been able to test theoretical predictions regarding nanoribbon edge-states, because no one could figure out how to see the atomic-scale structure at the edge of a well-ordered graphene nanoribbon and how, at the same time, to measure its electronic properties within nanometers of the edge," says Michael Crommie of Berkeley Lab's Materials Sciences Division (MSD) and UC Berkeley's Physics Division, who led the research. "We were able to achieve this by studying specially made nanoribbons with a scanning tunneling microscope."

The team's research not only confirms theoretical predictions but opens the prospect of building quick-acting, energy-efficient nanoscale devices from graphene-nanoribbon switches, spin-valves, and detectors, based on either electron charge or electron spin. Farther down the road, graphene nanoribbon edge states open the possibility of devices with tunable giant magnetoresistance and other magnetic and optical effects.

Crommie and his colleagues have published their research in Nature Physics, available May 8, 2011 in advanced online publication.

The well-tempered nanoribbon

"Making flakes and sheets of graphene has become commonplace," Crommie says, "but until now, nanoribbons produced by different techniques have exhibited, at best, a high degree of inhomogeneity" ? typically resulting in disordered ribbon structures with only short stretches of straight edges appearing at random. The essential first step in detecting nanoribbon edge states is access to uniform nanoribbons with straight edges, well-ordered on the atomic scale.

Hongjie Dai of Stanford University's Department of Chemistry and Laboratory for Advanced Materials, a member of the research team, solved this problem with a novel method of "unzipping" carbon nanotubes chemically. Graphene rolled into a cylinder makes a nanotube, and when nanotubes are unzipped in this way the slice runs straight down the length of the tube, leaving well-ordered, straight edges.

Graphene can be wrapped at almost any angle to make a nanotube. The way the nanotube is wrapped determines the pitch, or "chiral vector," of the nanoribbon edge when the tube is unzipped. A cut straight along the outer atoms of a row of hexagons produces a zigzag edge. A cut made at a 30-degree angle from a zigzag edge goes through the middle of the hexagons and yields scalloped edges, known as "armchair" edges. Between these two extremes are a variety of chiral vectors describing edges stepped on the nanoscale, in which, for example, after every few hexagons a zigzag segment is added at an angle.

These subtle differences in edge structure have been predicted to produce measurably different physical properties, which potentially could be exploited in new graphene applications. Steven Louie of UC Berkeley and Berkeley Lab's MSD was the research team's theorist; with the help of postdoc Oleg Yazyev, Louie calculated the expected outcomes, which were then tested against experiment.

Enlarge

By "unzipping" carbon nanotubes, regular edges with differing chiralities can be produced between the extremes of the zigzag configuration and, at a 30-degree angle to it, the armchair configuration. Credit: Hongjie Dai, Stanford University, and Michael Crommie et al, Lawrence Berkeley National Laboratory

Chenggang Tao of MSD and UCB led a team of graduate students in performing scanning tunneling microscopy (STM) of the nanoribbons on a gold substrate, which resolved the positions of individual atoms in the graphene nanoribbons. The team looked at more than 150 high-quality nanoribbons with different chiralities, all of which showed an unexpected feature, a regular raised border near their edges forming a hump or bevel. Once this was established as a real edge feature ? not the artifact of a folded ribbon or a flattened nanotube ? the chirality and electronic properties of well-ordered nanoribbon edges could be measured with confidence, and the edge regions theoretically modeled.

Electronics at the edge

"Two-dimensional graphene sheets are remarkable in how freely electrons move through them, including the fact that there's no band gap," Crommie says. "Nanoribbons are different: electrons can become trapped in narrow channels along the nanoribbon edges. These edge-states are one-dimensional, but the electrons on one edge can still interact with the edge electrons on the other side, which causes an energy gap to open up."

Using an STM in spectroscopy mode (STS), the team measured electronic density changes as an STM tip was moved from a nanoribbon edge inward toward its interior. Nanoribbons of different widths were examined in this way. The researchers discovered that electrons are confined to the edge of the nanoribbons, and that these nanoribbon-edge electrons exhibit a pronounced splitting in their energy levels.

"In the quantum world, electrons can be described as waves in addition to being particles," Crommie notes. He says one way to picture how different edge states arise is to imagine an electron wave that fills the length of the ribbon and diffracts off the atoms near its edge. The diffraction patterns resemble water waves coming through slits in a barrier.

Enlarge

Graphene nanoribbons are narrow sheets of carbon atoms only one layer thick. Their width, and the angles at which the edges are cut, produce a variety of electronic states, which have been studied with precision for the first time using scanning tunneling microscopy and scanning tunneling spectroscopy. Credit: Crommie et al, Lawrence Berkeley National Laboratory

For nanoribbons with an armchair edge, the diffraction pattern spans the full width of the nanoribbon; the resulting electron states are quantized in energy and extend spatially throughout the entire nanoribbon. For nanoribbons with a zigzag edge, however, the situation is different. Here diffraction from edge atoms leads to destructive interference, causing the electron states to localize near the nanoribbon edges. Their amplitude is greatly reduced in the interior.

The energy of the electron, the width of the nanoribbon, and the chirality of its edges all naturally affect the nature and strength of these nanoribbon electronic states, an indication of the many ways the electronic properties of nanoribbons can be tuned and modified.

Says Crommie, "The optimist says, 'Wow, look at all the ways we can control these states ? this might allow a whole new technology!' The pessimist says, 'Uh-oh, look at all the things that can disturb a nanoribbon's behavior ? how are we ever going to achieve reproducibility on the atomic scale?'"

Crommie himself declares that "meeting this challenge is a big reason for why we do research. Nanoribbons have the potential to form exciting new electronic, magnetic, and optical devices at the nanoscale. We might imagine photovoltaic applications, where absorbed light leads to useful charge separation at nanoribbon edges. We might also imagine spintronics applications, where using a side-gate geometry would allow control of the spin polarization of electrons at a nanoribbon's edge."

Although getting there won't be simple -- "The edges have to be controlled," Crommie emphasizes -- "what we've shown is that it's possible to make nanoribbons with good edges and that they do, indeed, have characteristic edge states similar to what theorists had expected. This opens a whole new area of future research involving the control and characterization of graphene edges in different nanoscale geometries."

Scientists discover dielectron charging of water nano-droplet

 

This image shows the surface attachment mode of two excess electrons to a water cluster (average diameter of close to 2nm) comprised of 105 molecules. The shown configuration was obtained from first-principles quantum simulations. The two wave functions each occupied by one excess electron, depicted in blue and green, are localized at opposite sides of the cluster and they are shown superimposed on the water molecules. The oxygens and hydrogens of the water molecules are represented by red and gray spheres, respectively. Credit: Uzi Landman/Georgia Tech

Scientists have discovered fundamental steps of charging of nano-sized water droplets and unveiled the long-sought-after mechanism of hydrogen emission from irradiated water. Working together at the Georgia Institute of Technology and Tel Aviv University, scientists have discovered when the number of water molecules in a cluster exceeds 83, two excess electrons may attach to it ? forming dielectrons ? making it a doubly negatively charged nano droplet. Furthermore, the scientists found experimental and theoretical evidence that in droplets comprised of 105 molecules or more, the excess dielectrons participate in a water-splitting process resulting in the liberation of molecular hydrogen and formation of two solvated hydroxide anions. The results appear in the June 30 issue of the Journal of Physical Chemistry A.

It has been known since the early 1980s that while single electrons may attach to small water clusters containing as few as two molecules, only much larger clusters may attach more than single electrons. Size-selected, multiple-electron, negatively-charged water clusters have not been observed ? until now.

Understanding the nature of excess electrons in water has captured the attention of scientists for more than half a century, and the hydrated electrons are known to appear as important reagents in charge-induced aqueous reactions and molecular biological processes. Moreover, since the discovery in the early 1960s that the exposure of water to ionizing radiation causes the emission of gaseous molecular hydrogen, scientists have been puzzled by the mechanism underlying this process. After all, the bonds in the water molecules that hold the hydrogen atoms to the oxygen atoms are very strong. The dielectron hydrogen-evolution (DEHE) reaction, which produces hydrogen gas and hydroxide anions, may play a role in radiation-induced reactions with oxidized DNA that have been shown to underlie mutagenesis, cancer and other diseases.

"The attachment of multiple electrons to water droplets is controlled by a fine balancing act between the forces that bind the electrons to the polar water molecules and the strong repulsion between the negatively charged electrons," said Uzi Landman, Regents' and Institute Professor of Physics, F.E. Callaway Chair and director of the Center for Computational Materials Science (CCMS) at Georgia Tech.

"Additionally, the binding of an electron to the cluster disturbs the equilibrium arrangements between the hydrogen-bonded water molecules and this too has to be counterbalanced by the attractive binding forces. To calculate the pattern and strength of single and two-electron charging of nano-size water droplets, we developed and employed first-principles quantum mechanical molecular dynamics simulations that go well beyond any ones that have been used in this field," he added.

Investigations on controlled size-selected clusters allow explorations of intrinsic properties of finite-sized material aggregates, as well as probing of the size-dependent evolution of materials properties from the molecular nano-scale to the condensed phase regime.

In the 1980s Landman, together with senior research scientists in the CCMS Robert Barnett, the late Charles Cleveland and Joshua Jortner, professor of chemistry at Tel Aviv University, discovered that there are two ways that single excess electrons can attach to water clusters ? one in which they bind to the surface of the water droplet, and the other where they localize in a cavity in the interior of the droplet, as in the case of bulk water. Subsequently, Landman, Barnett and graduate student Harri-Pekka Kaukonen reported in 1992 on theoretical investigations concerning the attachment of two excess electrons to water clusters. They predicted that such double charging would occur only for sufficiently large nano-droplets. They also commented on the possible hydrogen evolution reaction. No other work on dielectron charging of water droplets has followed since.

That is until recently, when Landman, now one of the world leaders in the area of cluster and nano science, and Barnett teamed up with Ori Chesnovsky, professor of chemistry, and research associate Rina Giniger at Tel Aviv University, in a joint project aimed at understanding the process of dielectron charging of water clusters and the mechanism of the ensuing reaction - which has not been observed previously in experiments on water droplets. Using large-scale, state-of-the-art first-principles dynamic simulations, developed at the CCMS, with all valence and excess electrons treated quantum mechanically and equipped with a newly constructed high-resolution time-of-flight mass spectrometer, the researchers unveiled the intricate physical processes that govern the fundamental dielectron charging processes of microscopic and the detailed mechanism of the water-splitting reaction induced by double charging.

Enlarge

This image shows the internal attachment mode of two electrons to a water cluster (average diameter of close to 2nm) comprised of 105 molecules. The shown configuration was obtained from first-principles quantum simulations. The wave function of the two excess electrons is depicted in pink and it is shown superimposed on the water molecules of the cluster. The compact dielectron distribution is localized in a hydration cavity in the interior of the cluster. The configuration shown here corresponds to the start of the dielectron hydrogen-evolution reaction. The protons of the two reacting neighboring water molecules, located approximately in the middle of the figure, are depicted by smaller blue spheres. Together with the dielectron these protons would form, in the course of the reaction, a hydrogen molecule. The oxygens and hydrogens of the water molecules are represented by red and gray spheres, respectively. Credit: Uzi Landman/Georgia Tech

The mass spectrometric measurements, performed at Tel Aviv, revealed that singly charged clusters were formed in the size range of six to more than a couple of hundred water molecules. However, for clusters containing more than a critical size of 83 molecules, doubly charged clusters with two attached excess electrons were detected for the first time. Most significantly, for clusters with 105 or more water molecules, the mass spectra provided direct evidence for the loss of a single hydrogen molecule from the doubly charged clusters.

The theoretical analysis demonstrated two dominant attachment modes of dielectrons to water clusters. The first is a surface mode (SS'), where the two repelling electrons reside in antipodal sites on the surface of the cluster (see the two wave functions, depicted in green and blue, in Figure 1). The second is another attachment mode with both electrons occupying a wave function localized in a hydration cavity in the interior of the cluster ? the so-called II binding mode (see wave function depicted in pink in Figure 2). While both dielectron attachment modes may be found for clusters with 105 molecules and larger ones, only the SS' mode is stable for doubly charged smaller clusters.

"Moreover, starting from the II, internal cavity attachment mode in a cluster comprised of 105 water molecules, our quantum dynamical simulations showed that the concerted approach of two protons from two neighboring water molecules located on the first shell of the internal hydration cavity, leads, in association with the cavity-localized excess dielectron (see Figure 2), to the formation of a hydrogen molecule. The two remnant hydroxide anions diffuse away via a sequence of proton shuttle processes, ultimately solvating near the surface region of the cluster, while the hydrogen molecule evaporates," said Landman.

"What's more, in addition to uncovering the microscopic reaction pathway, the mechanism which we discovered requires initial proximity of the two reacting water molecules and the excess dielectron. This can happen only for the II internal cavity attachment mode. Consequently, the theory predicts, in agreement with the experiments, that the reaction would be impeded in clusters with less than 105 molecules where the II mode is energetically highly improbable. Now, that's a nice consistency check on the theory," he added.

As for future plans, Landman remarked, "While I believe that our work sets methodological and conceptual benchmarks for studies in this area, there is a lot left to be done. For example, while our calculated values for the excess single electron detachment energies are found to be in quantitative agreement with photoelectron measurements in a broad range of water cluster sizes ? containing from 15 to 105 molecules ? providing a consistent interpretation of these measurements, we would like to obtain experimental data on excess dielectron detachment energies to compare with our predicted values," he said.

"Additionally, we would like to know more about the effects of preparation conditions on the properties of multiply charged water clusters. We also need to understand the temperature dependence of the dielectron attachment modes, the influence of metal impurities, and possibly get data from time-resolved measurements. The understanding that we gained in this experiment about charge-induced water splitting may guide our research into artificial photosynthetic systems, as well as the mechanisms of certain bio-molecular processes and perhaps some atmospheric phenomena."