Solving Decades-Old Problems in Materials Science

Developing novel approaches for understanding strongly non-equilibrium systems

Computers have revolutionized many branches of science and technology from quantum chemistry to bioimaging, genomics, astrophysics, flight simulations and many others. Despite these impressive advances, there are old-standing problems of Physics, Chemistry, and Biology that do not yield readily to computer modeling. These difficult problems call for radically new and elegant approaches, a motivation that drives Dr. Vassiliy Lubchenko's research at the University of Houston. Being a theoretician gives him the flexibility to think about a wide range of challenges and uncover their mutual connections.

Lubchenko's research efforts focus in three areas:


  • Amorphous Solids and the Theory of the Glass Transition: Dr. Lubchenko is tackling long-standing problems posed by the phase-change materials. The phase change materials are special mixtures of two or more elements from the region in the periodic table centered at antimony and are best known for their use in Blu-ray disks. One can use phase-change materials to record information because of their unique ability to switch electrical and optical properties depending on whether the material is in the crystalline or an amorphous, glassy form. The phase-change materials are potential candidates for the next generation computer memory, ultrafast displays, and smart optics. To realize this potential, we must understand and, ultimately, control the transition between the crystalline and glassy forms of the phase-change material. Achieving this understanding and determining the structure of the phase-change materials are the main goals of this project.


  • Mesoscopic Protein Aggregation: A living cell is a mixture of water and various biomolecules, mostly proteins, enclosed in a flexible, permeable membrane. Except when needed for protein storage, protein aggregation in living cells is harmful, leading to diseases such as sickle cell anemia and Alzheimer's. During sickle cell anemia, deleterious fibers grow uncontrollably inside the red blood cells eventually causing blood clots and other painful symptoms. It turns out these fibers are seeded in very special, tiny droplets of concentrated protein solution. These droplets - which we call mesoscopic clusters - defy traditional views on phase transformations because, naively, they seem to be equivalent to microscopic icebergs in water well above freezing, something we know for sure could not happen! In a collaboration with experimenter, Professor Vekilov, of the University of Houston Chemical Engineering Department, Dr. Lubchenko has proposed a radically new mechanism that could account for the emergence of these clusters. If confirmed, this new knowledge may help prevent harmful types of protein aggregation that leads to diseases. It may also speed up the beneficial kind of protein aggregation, such as protein crystallization, which is the holy grail of structural biology.


  • Predicting the Structure of Complex Inorganic Solids: Predicting the structure of crystalline solids is a very old problem of Chemistry. Solving this problem is key to our progress in designing novel materials with useful, tailored properties, such as high temperature superconductors, which can be used to make loss-free engines and magnets, or thermoelectric materials, which can be used for efficient power generation and refrigeration. Presently, novel materials are designed by trial and error, which is extremely time-consuming. As a result, only a tiny fraction of all possible compounds has been synthesized and characterized. Computers have the potential to greatly reduce the number of possible combinations of distinct elements one needs to sift through in search for a compound with given properties. However, existing computer algorithms have a difficult time bypassing the numerous amorphous configurations of the atoms on their way to the coveted, low energy crystal structure for a given compound. Dr. Lubchenko aims to utilize his expertise in glassy materials to attack this long-standing problem from a new angle in an attempt to minimize the number of amorphous configurations that block the path to the crystal state.


Dr. Lubchenko grew up in the Soviet Union. His early interest in science and mathematics was spurred by 1970's advances in Cosmology and theories of unification of fundamental interactions. Dr. Lubchenko's research interests include supercooled liquids and glasses, in which molecular motions span an extremely broad dynamic range that is rivaled only by astrophysical phenomena. Of particular interest are the glassy alloys that are currently used in optical drives but have also shown a potential to revolutionize computer, video, and medical technology. Molecular interactions in these alloys exhibit a diversity much like that seen in particle physics. Controlling the competition between these interactions is key to our ability to create new materials. Dr. Lubchenko also uses his expertise in strongly non-equilibrium phenomena and chemical interactions to tackle the problem of protein aggregation, which underlies many diseases awaiting cure. Website: Dr. Lubchenko's work will also provide answers to these difficult challenges: One of the oldest materials problems is the physics of amorphous materials such as the common window glass or the Gorilla glass used in a smartphone or the alloys used to make Blu-ray disks. When forming, these glassy materials flow on the timescale of hours, while the quickest atomic motions, the vibrations, occur within femtoseconds, i.e., 18 orders of magnitude faster! Not only such a tremendous dynamical range is nine orders of magnitude too broad for a computer to handle, the computer will not be able to explain in human terms the mechanism of those slow motions, much the same way Chemistry explains in relatively simple terms -by using the concept of the chemical bond - the complicated quantum mechanics of the electrons and nuclei. It is the mechanism of those slow motions that we need to understand in order to tailor the properties of amorphous materials to make them suit our needs. At stake is our ability to create faster, more powerful computers and materials with better mechanical, electric, and thermal properties. Our strategy is multi-prong: First, we eliminate the need to simulate the vibrations of the atoms so that we only have to take care of their average positions. This is being achieved by mapping the dynamics in glassy liquids onto a class of physical models that are used to describe magnets, among other things. In addition to reducing the computational complexity of the problem, we can also take advantage of the insight gained into those magnet models in the 1970s. Additionally, we are designing novel, efficient algorithms to generate amorphous structures that take advantage of the progress made lately in generation of densely-packed amorphous arrays of perfect spheres. Adding realistic chemical interactions is tricky but doable. Very different, but equally challenging questions arise with regard to protein aggregation. The computers are just beginning to be able to handle the folding of an individual protein molecule; folding is requisite for the biomolecule's functionality. Now, how about a protein aggregate that contains hundreds of thousands of partially unfolded proteins? This is too much for a computer to attack head on. Yet this challenge needs to be faced if we were to control the deleterious kinds of protein aggregation that lead to many diseases, such as sickle cell anemia, Alzheimer's and many others. We break this difficult problem into three components. On the one hand, we have made a strong case that a novel mechanism of aggregation, by way of forming transient protein complexes, could be at play in protein solutions. This fundamental result clears the way for more detailed studies, in which we try to detect the protein complexes directly (as part of collaboration with experimenter Peter G. Vekilov) and, at the same time use the computer to detect such complexes in a simulation. Recent advances in protein structure prediction are providing us with novel tools to tackle this computationally intensive problem. The chemical formula for a compound is much like the list of ingredients in a recipe. For instance, the best known phase-change material, GST, consists of two parts germanium per two parts antimony per five parts tellurium: Ge2Sb2Te5. As in cooking, it often takes skill to put together the ingredients. In addition, an experienced chef will often know without cooking what a specific combination of ingredients will taste like, thus eliminating the necessity to try all possible combinations of food stuffs and allowing one to focus on a small, most promising subset of combinations. Synthesizing new materials in the lab is usually difficult and time-consuming, while the number of possible recipes is just enormous even for only two-three ingredients. But things often become interesting, application-wise, only starting at four ingredients, as is the case for many high temperature super-conductors. This suggests we may not ever be able to discover some of the most useful materials simply for the lack of time! Could we, instead, use the computer to help us out and predict whether a given combination of elements would be stable and have specific properties of interest? It turns to put together a solid using a computer generally takes as much time as would be needed for somebody to do it physically in the lab. What is the computational bottleneck? The bottleneck seems to be very same states we wanted to generate in the project on the amorphous materials. It is those amorphous states where the computer tends to linger instead of cruising toward the low energy, crystalline state. Our strategy is to modify the interactions between the atoms in a way that the lowest energy state is still recognized as such but, at the same time, to diminish the occurrence of the glassy states in the simulation and thus remove or, at least, mitigate the computational bottleneck toward the crystal structure.


Microscopically Based Calculations of the Free Energy Barrier and Dynamic Length Scale in Supercooled Liquids: The Comparative R


Molecular Binoculars: How to Spatially Resolve Environmental Fluctuations by Following Two or More Single-Molecule Spectral Trai


Microscopic calculation of the free energy cost for activated transport in glass-forming liquids.


Ostwald-like ripening of the anomalous mesoscopic clusters in protein solutions.


Liquid State Elasticity and the Onset of Activated Transport in Glass Formers.



Alfred P. Sloan Research Fellow, 2011-2013

NSF CAREER Award, 2010

Beckman Young Investigator Award, 2008

Postdoctoral Fellowship, Massachusetts Institute of Technology, 2003-2005

Hovorka Fellowship, University Fellowship, University of Illinois at Urbana-Champaign, 1996-1998