The study of material properties at extreme conditions (megabar pressures and thousands of degrees Kelvin) has gained considerably attention recently, in part due to the emergence of new techniques and the development of large lasers at facilities such as the National Ignition Facility (NIF) at Livermore, USA [1], and the 100 TW facilities at LULI, France [2]. These emerging facilities will allow the study of material properties at conditions until now inaccessible to experiments. Brighter x-ray sources and advances in static laser-heated diamond-anvil-cell (LHDAC) technology, pyrometry, and spectroradiometry measurement techniques have extended the pressure-temperature regime accessible via LHDAC experiments. These advances have dramatically increased the study of pressure-induced phase transitions within the last decade from both theoretical and experimental camps. In turn, advances in large-scale computing and computational speeds have led to an increase in the accuracy and predictive power of simulations [3] of phase transformations at high pressures, both statically and under dynamic shock-loading conditions.

Ab initio simulations are now commonly used in the investigation of material deformation at extreme conditions, where the predictions of new crystalline phases have increased significantly in the last decade. They have been used with varied degrees of success in the prediction of the equation of state of elemental solids (notably H, C, Fe and Al), semiconductors, oxides, and many other systems. This type of calculations can be particularly useful in aiding experimental efforts in the search and identification of new high-pressure phases. Simple strategies, such as random sampling [4] and simulated annealing [5] are commonly employed for searching/finding the stable structure of systems with a relatively small number of atoms in the unit cell. More sophisticated algorithms [6] are required to deal with much larger systems and to allow for a more reliable and efficient structure prediction at a fully

ab initio level.

The study of dynamics and kinetics of high-pressure phase transitions is more challenging. The development of first-principles, constant-pressure Quantum Molecular Dynamics (QMD) simulation methods [7] have allowed the investigation of transformation paths and relative stability of high pressure phases in many systems. It has also been used in predicting the pressure dependence of the melt temperature of elemental solids, including Fe [8] and some transition metals [9,10] to name but a few applications. The time and length scale limitations of these methods can, in principle, be overcome by creating classical, semi-empirical potentials, which reproduce data from first-principles calculations. This has been tried with great success for iron, whose mechanical properties were calculated in the pressure regime of a few GPa [8]. Classical, large-scale MD simulations, which make use of semi-empirical inter-atomic potentials, have been used with great success in tackling problems in shock-induced plasticity, phase transformation, and melting. The simulated time and length scales are approaching those of current laser-driven shock experiments and have yielded excellent agreement with experiments in some cases [11]. These simulations can complement experiments on high-pressure phase transitions in two completely different regimes: (1) equilibrium, such as probed by static diamond-anvil cells; and (2) non-equilibrium, such as probed by dynamical shockwave techniques. An important difference between these two regimes involves the kinetics of phase transitions in the shock process, where kinetic temperature overshoots the equilibrium value in the shock front. Thus, the possibility exists in shock experiments for inhomogeneous, transient regions where phases could exist that would never be possible under equilibrium conditions. Each of these MD approaches -- quantum and classical large-scale -- has its place in studying high-pressure, high temperature phase transformations: QMD can give relatively accurate equilibrium equation-of-state information, but not kinetic effects; large-scale NEMD can provide information on intermediate time scales, particularly if local shear stresses lower the thermal barriers to transitions and produce sufficient heterogeneities, so that one has hope of seeing believable kinetics. In summary, while much progress has been achieved in this field in the last decade, many theoretical and experimental challenges remain. It is clear that in many instances experimental efforts can benefit greatly from the details of atomistic computer simulations. One goal of this workshop is to facilitate the synergy between experiment and theory, review current developments and challenges and envision ways to deal with new ones.

Scientific Objectives

The aim of the workshop is to bring together experimental, theoretical, and computational experts in high-pressure phase transitions, in order to tackle current and new challenges in the study of solids under extreme hydrostatic and non-hydrostatic pressures and high temperatures. Novel mechanisms for melting at high-pressures under hydrostatic [12] and shock-loading conditions have been proposed but remain difficult to confirm experimentally. Some groups have proposed the theoretical existence of intermediate phases in some transitions metals prior to melting [13,14]; however, no experimental observation of these phases has so far been observed. The accuracy of current computational techniques for evaluating the solid-liquid and solid-solid phase boundaries should be revisited.

An area of growing interest is the use of efficient evolutionary algorithms for high-pressure structure predictions [6]. However, all the existing

ab initio structure predictions are usually only applicable near zero temperature. Finite-temperature structure predictions are possible in principle, but are impractical because of the extreme computational costs of

ab initio free energy calculations. Methods of incorporating temperature into current search algorithms (genetic and random) is expected to play a major role in identifying, at a fully

ab initio level, structural phases under extreme conditions and hence help in extending the regime of validity and improving the accuracy of current and new generation of semi-empirical potentials used in large-scale simulations.

Topics to be considered will include:

- Solid-solid and solid-liquid (melting) phase transitions.

- Comparison among theory and experiments; drawbacks and advantages of different techniques.

- Kinetics and transitions paths in phase transformations under dynamic uniaxial and hydrostatic loading.

- Mechanical properties and instabilities in crystalline solids(metals, semiconductors, oxides) under extreme conditions.

- Methods for incorporating temperature in search algorithms for identifying structural phases and locating phase boundaries.

References

[1]

https://lasers.llnl.gov/[2]

http://www.luli.polytechnique.fr/index_en.htm[3] A. Mujica, A. Rubio, A. Muñoz and R.J. Needs, Rev. of Moderm Phys. 75, 863 (2003).

[4] C.J. Pickard and R.J. Needs, Phys. Rev. Lett. 97, 045504 (2006).

[5] J.C. Shön and M. Jansen, Angew. Chem. – Int. Ed. 35, 1287–1304(1996).

[6] A.R. Oganov and C.W. Glass, J. Chem. Phys. 124, 244704 (2006).

[7] R. Carr and M. Parrinello, Phys. Rev. Lett., 55, 2471 (1985).

[8] A. Laio et al, Science, 287, 1027 (2000).

[9] S. Taioli, C. Cazorla, M. J. Gillan, and D. Alfe`, Phys. Rev B 75, 214103 (2007).

[10] Z.-L. Liu, L.-C. Cai, X.-R. Chen, and F.-Q. Jing, Phys. Rev. B 77, 024103 (2008).

[11] K.Kadau et al, Science, 296, 1681 (2002).

[12] C. J. Wu, P. Söderlind, J. N. Glosli, and J. E. Klepeis, Nature Mater. 8, 223 (2009).

[13] L. Burakovsky, et al, Phys. Rev. Lett. 104, 255702 (2010).

[14] A. B. Belonoshko et al, Phys. Rev. Lett. 100, 135701 (2008).