RESEARCH AREAS

Multiscale methods

Biopolymer translocation through a nanopore, simulated through coupled Molecular Dynamics for the polymer and the Lattice Boltzmann method for the fluid, the latter not shown explicitly (image created by Simone Melchionna).

Many phenomena in nature are intrinsically multiscale, that is, they involve processes on several scales. A typical example is the propagation of a crack in brittle fracture, in which bonds are breaking at the atomic scale in the crack tip region, defects like dislocations and stacking-fault planes propagate into the plastic zone around the tip, and far away from the crack the material remains in the elastic state. Another example is the motion of polymers in fluids, in which the molecular scale motion of the polymer may exhibit interesting behavior, but the motion of the surrounding fluid molecules is not particularly interesting. For these complex systems, we must devise a new way of simulating them, keeping only the interesting degrees of freedom and capturing the effect of the remaining ones in an effective way; this is referred to as 'coarse graining' of the unisteresting degrees of freedom. We have been developing multiscale methodologies to treat these types of phenomena, both for solids and for biomolecules (such as DNA) in a fluid.
The figure shows the translocation of a biopolymer (DNA) through a nanopore, investigated through a multiscale method in which the motion of the moelcule is followed at the molecular level but the motion of the fluid is treated mesoscopically, through the Lattice Botlzmann method which simulates the equations of hydrodynamics.

Biomolecules

Electronic states in stretched DNA: the wavefunctions of the highest occupied state are shown as color coded contours for the value of the wavefunction magnitude, with purlpe the highest value, on a cylindrical surface co-axial with the DNA double helix; on the left is a stretched form and on the right the unstretched form of DNA (image created by Paul Maragakis).

Biomolecules have many interesting properties that derive from their structure; small changes in the structure can have a very siginficant effect on the properties. Through these properties, biomolecules are able to perform the biological function for which they have been selected. We investigate the electronic and optical properties of several types of biomolecules, including DNA bases, melanin and flavonoids. DNA bases have very specific optical and electronic properties that may help identify them through electrical or optical measurements. This might be useful for devices designed to perform ultrafast DNA sequencing. Melanin, on the other hand, has a very broad and featureless optical absorption spectrum, which is very unusual for a biomolecule, but may be essential for its photo-protective role. Flavonoids are anti-oxidants whose electronic properties and interaction with metal ions play a crucial role in the ability of cells to defend their DNA against oxidative stress. For all these systems, detailed understanding of their microscopic structure and electronic properties can only be attained through accurate quantum mechanical calculations. We have been developing methods to perform such calculations, including time-dependent density functional theory, and applying them to the types of problems mentioned.
The figure shows an example of electronic states related to the DNA double helix and their dependence on the degree of DNA stretching. Using multiscale quantum mechanical simulations, we were able to show that stretching induces localization of electrons in DNA, which accurately explains exprimental observations.

Nanostructures

Electronic states related to a ring of palladium atoms around a carbon nanotube. These states fall in the band gap of the originally semi-conducting nanotube (image created by Wenguang Zhu).

There are several systems that have nano-meter size and can be produced controllably in the laboratory. These systems, referred to as `nanostructures', exhibit interesting behavior and their size makes them ideal components for several types of applications, for instance, in future electronic devices. Carbon nanotubes are one example of such systems that have attracted much interest because of their robustness and their extraordinary properties. Other types of nanotubes may be equally robust and possibly even more useful in the context of electronic devices. It is also possible to form nanostructures on carefully prepared surfaces, as, for example, at the edges of a stepped surface. We have been investigating the properties of these systems, and their interaction with other components such as metallic leads and DNA bases.
The figure shows electronic states associated with a ring of palladium atoms deposited on a carbon nanotube: palladium seems to be one of the best materials to make a metal contact with a semiconducting carbon nanotube. This may be an ideal combination for the basic component of future electronic devices.

Mechanochemistry

A small crack in an aluminum crystal, treated quantum mechanically: the contours show the density of electrons, with red the highest, and the blue spheres the positions of atoms (image created by Gang Lu).

Mechanochemistry is a term that denotes the effects of chemical impurities on the mechanical behavior of solids. In most real materials there are impurities either from the way the material was manufactured or from exposure to the environment. These impurities can have a dramatic impact on mechanical behavior, sometimes turning a normally ductile and tough solid into a brittle and fragile one. The presence of impurities at the tip of a crack alters the bonding between atoms and helps propagate the crack resulting in brittle fracture. Corrosion and external stress combine to have this effect on several materials, hence the so called 'stress corrosion cracking' phenomenon.
The figure shows an example of the structure of a small crack in an aluminum crystal, the prototypical ductile solid. The presence of a small amount of impurities at the crack tip is sufficient, under the right stress conditions, to turn aluminum into a brittle solid. This is only a part of a more elaborate multiscale simulation, going from the atomistic to the continuum scale (see also Multiscale methods).

SAMPLING OF RESEARCH PROJECTS

- Superhydrophilic Diamond Surface

- Carbon Nanotube Contact With Metal Leads

- Nanowires on Stepped Metal Surfaces

- DNA Interaction with Carbon Nanotubes

- Effect of Chemical Impurities on Mechanical Behavior

Superhydrophilic Diamond Surface

Top and side views of the chemically modified diamond (111) surface with several layers of water molecules on it. The metal atoms in this case are sodium atoms (M = Na, blue spheres) and the non-metal atoms X are fluorine atoms (X = F, yellow spheres).

Surfaces exhibit either hydrophobic (water repelling) or hydrophilic (water attracting) behavior depending on their structure and chemical composition.
The (111) surface of diamond, terminated by H atoms, has extreme hydrophobic behavior (super-hydrophobic). Interestingly, the atomic structure of this surface is compatible with the crystal structure of ice.
We proposed, based on extensive first-principles calculations, that by chemically modifying this surface with the proper type of elements, which partially substitute the H atoms, it would be possible to create a super-hydrophilic surface.

Our simulations show that the ice layers on this surface should be stable at temperatures in the range of 310 K (body temperature). If this system can be formed experimentally, it would be an exciting possibility for biomedical applications.

This work was carried out by Postdoctoral Fellow Sheng Meng in collaboration with Zhenyu Zhang, and the water simulations were further extended by graduate student Alex Wissner-Gross.

Carbon nanotube contact with metal leads

Two views of a carbon nanotube embedded in a palladium crystal.

Nanostructures provide unique opportunities for vastly improving the performance of electronic devices through the realization of molecular computing, based on components with ultra-high density of elements and speed of operation.

Among the various possible nanoscale elements for molecular devices, carbon nanotubes (CNTs) have emerged as one of the most promising building block because of their size, structural strength and extraordinary electronic properties. Since they were first synthesized, these intriguing carbon structures have received extensive attention.

Recently, it has been demonstrated that an individual semiconducting CNT can operate either as a conventional MOSFET (metal oxide semiconductor field effect transistor) or an unconventional Schottky barrier transistor when it forms a contact with a metal electrode, with the type of behavior depending on the properties of the metal-CNT contact. The interaction between the carbon nanotube and the metal leads, and the resulting electronic structure effects, are crucial for the performance of the carbon nanotube field-effect transistor (CNFET).

Contours of the calculated electrostatic potential (red corresponds to values higher than the Fermi level, blue to lower values).

In this project, we used first-principles calculations to study the electronic structure of the contact between semiconducting single-wall carbon nanotubes (SWCNTs) and Pd metal leads, in a fully covered geometry that resembles closely the experimental setups. Our analysis showed that the fully-covered CNT exhibits metallic character in its contact to the Pd lead. Further, we showed that when rings of Pd atoms cover the CNT, a Schottky barrier naturally arises between the covered and uncovered portions of the CNT.

This work was carried out by Postdoctoral Fellow Wenguang Zhu. We have benefitted greatly from interaction with the experimental group of Dr. Phaedon Avouris (IBM T.J. Watson Research Center).

Nanowires on Stepped Metal Surfaces

Deposition of iron (FE) atoms on a stepped copper (Cu) surface: Fe atoms are not simply attached to the edge of steps, they are actually embedded in the step behind a row of Cu atoms. This unique arrangement (Fe atoms shown in red, on the stepped Cu surface) makes the Fe line of atoms very stable.

Nanowires are one of the "holy grail" items of nanotechnology: the ultimate nanowire is a system of atoms essentially infinite in length but only one atom wide. It is exceedingly difficult to produce such systems in the laboratory, because it is not possible to confine, and keep in a straight line, a single row of atoms. Theoretical studies of our group proposed a possible realization of such a single-atom-wide wire of iron (Fe) atoms on a surface of copper (Cu). It is quite common for atoms deposited on a surface to seek out and attach themselves to the edge of surface steps; Cu surfaces can be prepared with lots of atom-high steps. However, something unique occurs, our theoretical calculations predicted, when Fe atoms are deposited on a stepped Cu surface. The Fe atoms are not simply attached to the edge of steps, they are actually embedded in the step behind a row of Cu atoms. This unique arrangement, illustrated in the first figure on the left makes the line of Fe atoms very stable.

A very stable, one-atom-wide iron nanowire should be formed on the Cu surface, involving this double-line of Fe atoms: one line buried in the surface behind a step, the second line on top of the first.

Moreover, the next set of Fe atoms deposited on the surface, the calculations again predicted, should be strongly attracted to the buried Fe atoms, because Fe-Fe bonds are stronger than Fe-Cu bonds. As a result, a very stable, one-atom-wide iron nanowire should be formed on the Cu surface, involving this double-line of Fe atoms, one line buried in the surface behind a step, the second line on top of the first, as shown in the second figure on the right.
Recently, experiments confirmed this theoretical prediction. Additional theoretical work proved that the experimental images from Scanning Tunneling Microscopy are in perfect agreement with the calculations for the electronic profile of this system. These findings open the possibility for manufacturing Fe nanowires of width equal to the size of one atom, which will be stable and reproducible on a carefully prepared Cu surface. The implications for nanoscale device components employing these iron wires, possibly to transmit spin-polarized electronic signals since Fe atoms are spin polarized, are truly exciting.
This work was carried out by graduate student Yina Mo, in collaboration with Zhenyu Zhang and experimental colleagues J. Guo and H. Weitering.

DNA Interaction with Carbon Nanotubes

In this project we study the interaction of Carbon nanotubes (CNT's) with DNA. Our first attempt to elucidate the physics of this system considered a periodic array of CNT's of the (10,0) type on which an infinite double-strand of DNA is placed. The CNT's have just the right size to allow them to fit snuggly in the major groove of the DNA.

The first picture on the left shows a realization of this system in top view and side view (the latter along the axis of the nanotubes). Note that the angle between the nanotube and DNA axes needs to be chosen carefully to make it possible for the parallel array of nanotubes to be inserted in the DNA groove.
Our electronic structure calculations revealed that this combined system has interesting properties which may render it useful as an electronic device. Moreover, the interaction between the two components may allow the sequencing of DNA through the effect that each pair of bases has on the electronic properties of the adjacent CNT.

The second figure below shows electronic states on the CNT and DNA system (described by blue spheres which represent the concentration of electronic charge). Notice that the electrons in both subsystems are delocalized, that is, they extend through the entire system.

Since a small number of base pairs is in intimate contact with the adjacent CNT, it might be possible to differentiate between pairs of bases according to their electronic properties. Further calculations are under way to quantify and prove these claims.
This work was carried out by Postdoctoral Fellow Gang Lu (now a faculty member at California State University at Northridge) and Paul Maragakis (now at Shaw Research, New York). The work is being continued in collaboration with Prostdoctoral Fellow Sheng Meng and Costas Papaloukas, a visitor from the University of Ioannina.

Effect of Chemical Impurities on Mechanical Behavior

Many metals and metal alloys can change behavior from tough, ductile solids to brittle ones due to the presence of chemical impurities. In some cases it takes only minute amounts of chemical impurities to induce dramatic changes in the mechanical behavior. Since the effect of chemical impurities is local (as all chemistry is) but its manifestations are macroscopic, several length scales must be simultaneously described in this system. Moreover, since chemical effects must be described by quantum mechanics while mechanical behavior at the macroscopic scale need only involve classical description, a successful methodology must couple classical and quantum mechanics.
Our recent work on this topic has focused on Aluminum, the prototypical ductile metal and a material used widely in technological applications. Aluminum can become brittle due to hydrogen impurities. We have successfuly produced a multiscale methodology, coupling the Quasicontinuum Method (for the classical mechanics) with Density Functional Theory (for the quantum mechanics) to describe the effect of hydrogen impurities on dislocations in Aluminum. We find that the presence of H impurities in the dislocation core splits the dislocation into partials with considerably wider separation than in the pure metal (see figure on the left, with atomic planes identified on either side of the dislocation core, in the pure metal, top, and in the metal with H impurities, bottom). This is a first step in a realistic description of hydrogen embrittlement effects in ductile metals.
This work was carried out with former members of the group, Gang Lu (now a faculty member at California State University at Northridge), Nick Choly (now at Bank of America, Boston) and Ellad Tadmor (now a faculty member at the University of Minnesotta), in an on-going collaboration.