Description of research
Our research in computational materials science focuses on atomistic computation of defects, phase transitions, electronic properties and mechanical behavior of materials. We balance two complementary goals, how do complex structures form in nature and how do these structures give rise to a range of properties including transitions from one structure to another. To be useful in materials and device design, materials modeling must accurately predict crucial microscopic behavior with scales of Angstrom and nanoseconds, and feed mesoscale models to predict macroscopic response of devices on time scales up to decades.
We aim to develop computational techniques that both accurately predict materials properties and provide an estimate of their accuracy and to apply these methods to accelerated materials development and enhanced understanding of the effect of atomic-scale processes on meso and macroscale behavior. The flexible approaches we employ - ranging from efficient molecular dynamics simulations to more accurate density-functional techniques to highly accurate quantum Monte-Carlo methods - enable simulations at relevant length and time scales and accuracy estimates of the underlying methods.
Current projects range from modeling martensitic transformations in transition metals, to mantle minerals under pressure, to defects in organic and compound semiconductors.
There is one position open in my research group for a graduate student. If you are a interested in a career in computational materials science, please apply for the graduate program in Materials Science at Cornell University and contact me.
If you are an undergraduate or graduate student at Cornell University and interested in a career in computational materials science, please contact me.
Computational methods
We systematically combine methods to both increase the accuracy and efficiently scan phase space to discover the evolution of defects and phases in materials. Our approach relies on transition-state techniques such as accelerated molecular dynamics, nudged-elastic band and the dimer method to efficiently scan the phase space and discover new defect structures and transition pathways. In addition we optimize fast empirical and tight binding potentials to accurately reproduce electronic structure results. Electronic structure methods estimate formation energies, diffusion and transition rates and describe the electronic properties of materials. Highly accurate quantum Monte-Carlo calculations benchmark the accuracy of approximations in electronic structure techniques.
Publications:
- Alleviation of the Fermion-Sign Problem by Optimization of
Many-Body Wave Functions.
C. J. Umrigar, J. Toulouse, C. Filippi, S. Sorella and R. G. Hennig.
Phys. Rev. Lett. 98, 110201 (2007). [PDF] - Comparison of screened hybrid density functional theory to
diffusion Monte Carlo in calculations of total energies of
silicon phases and defects.
E. R. Batista, J. Heyd, R. G. Hennig, B. P. Uberuaga, R. L. Martin, G. E. Scuseria, C. J. Umrigar, and J. W. Wilkins.
Phys. Rev. B 74, 121102(R) (2006). [PDF]
Martensitic phase transformations
Martensitic phase transitions are abundant in nature and commonly used in engineering technologies. They are diffusionless structural transformations proceeding near the speed of sound, governed by strongly correlated displacive motion of atoms. Onset and kinetics of martensitic transformations are controlled by impurities trapped during the transformation. Specifically, the pressure driven hcp to omega martensitic transformation in titanium is important in the aerospace industry because formation of omega-Ti lowers toughness and ductility. Oxygen impurities with concentrations as low as 1% shut down this transformation.
We use ab initio, tight-binding, and classical potentials to determine the pathway and study the effect of impurities. A systematic algorithm enumerates all possible Ti hcp to omega pathways and sorts them according to their energy barrier. A new, homogeneous pathway emerges with a barrier at least 4 times lower than other pathways. The pathway is favorable in any nucleation model and molecular dynamics simulates the motion of the phase boundary at a fraction of the speed of sound. Using the new pathway we study interstitial O, N, C; substitutional Al and V; and Ti vacancies and interstitials. Ab initio calculations yield the changes in both the relative stability of and energy barrier between the phases. The resulting microscopic picture explains the observations, specifically the suppression of the transformation in A-70 and Ti-6Al-4V titanium alloys by O and Al impurities.
Publications:
- Impurities block the alpha to omega martensitic transformation in titanium.
R. G. Hennig, D. R. Trinkle, J. Bouchet, S. G. Srinivasan, R. C. Albers, and J. W. Wilkins.
Nature Materials 4, 129 (2005). [PDF] - A new mechanism for the alpha to omega martensitic transformation in pure titanium.
D. R. Trinkle, R. G. Hennig, S. G. Srinivasan, D. M. Hatch, M. D. Jones, H. T. Stokes, R. C. Albers, and J. W. Wilkins.
Physical Review Letters 91, 025701 (2003). [PDF] - Systematic pathway generation and sorting in martensitic
transformations: Titanium alpha to omega.
D. R. Trinkle, D. M. Hatch, H. T. Stokes, R. G. Hennig and R. C. Albers.
Physical Review B 72, 014105 (2005). [PDF] - An empirical tight-binding model for titanium phase transformations.
D. R. Trinkle, M. D. Jones, R. G. Hennig, S. P. Rudin, R. C. Albers and J. W. Wilkins.
Phys. Rev. B, 73 094123 (2006). [PDF]
Defect properties for realistic simulations of semiconductor devices
Realistic simulations of silicon devices require multiscale methods whose success hinges on accurate atomistic input parameters. Interstitials, vacancies and extended defects limit device properties. The large number of degrees of freedom in realistic materials and the unknown error bars for atomistic simulations reduce the reliability of calculated materials parameters. The inability of electronic structure methods alone to determine defect structures, energies and diffusion rates requires an atomistic multiscale approach that both increases the accuracy and efficiently scans the phase space.
The discovery of previously unknown ground state structures and low-energy diffusion pathways reveals the complexity of interstitial defect evolution in silicon and demonstrates the power of our computational approach. The new interstitial defects serve as steps along the nucleation path for ion-implantation-induced extended defect structures. This general approach can be applied to defects in other materials and even diffusive phase transformations. Our study extends the accuracy of electronic structure to truly macroscopic length scales by providing reliable materials parameters for device simulations.
Publications:
- Diffusion mechanisms for silicon di-interstitials.
Y. A. Du, R. G. Hennig, and J. W. Wilkins.
Phys. Rev. B 73, 245203 (2006). [PDF] - Fast diffusion mechanism of silicon tri-interstitial defects.
Y. A. Du, S. A. Barr, K. R. A. Hazzard, T. J. Lenosky, R. G. Hennig, and J. W. Wilkins.
Physical Review B Rapid Communications 72, 241306(R) (2005). [PDF] - Complexity of small silicon self-interstitial defects.
D. A. Richie, J. Kim, S. A. Barr, K. R. A. Hazzard, R. G. Hennig, and J. W. Wilkins.
Physical Review Letters 92, 45501 (2004). [PDF]