A few discovery stories

Discovery of novel and unexpected sodium chlorides: Na3Cl, Na2Cl, Na3Cl2, NaCl3, NaCl7

Zhang W.W., Oganov A.R., Goncharov A.F., et al. (2013). Unexpected stoichiometries of stable sodium chlorides. Science 342, 1502-1505. (link).

This work challenges traditional chemical concepts: while classical rules of chemistry predict that the only possible stable sodium chloride is NaCl, variable-composition USPEX calculations predict that under moderate pressure numerous new compounds become thermodynamically stable. Most of these are metallic. Some of the predicted compouns are two-dimensional metals. These predictions have been verified by experiments, which synthesized and characterized the predicted compounds. Such "forbidden" compounds are expected to become stable in many other systems under pressure. This work has received huge resonance in professional and lay media, below is a small selection of the press.

Prediction of low-energy two-dimensional boron structure with massless Dirac fermions and properties superior to graphene

Zhou X.F., Dong X., Oganov A.R., Zhu Q., Tian Y. and Wang H.T. (2014). Semimetallic two-dimensional boron allotrope with massless Dirac fermions. Phys. Rev. Lett.112, 085502. (link).

It has been assumed the most stable two-dimensional structure formed by boron atoms is the so called planar alpha-sheet, and on its basis structures of boron nanoparticles and nanotubes have been postulated. Using USPEX, we showed that the alpha-sheet is massively unstable and non-planar structures possess the lowest energies. Among these, one stands out: its electronic structure features distorted Dirac cones, with electronic velocities being in one direction substantially higher than in graphene and substantially lower in the perpendicular direction. The instability of the alpha-sheet and stability of completely different structures calls for a reassessment of structural models of boron nanoparticles and nanotubes, while the electronic structure of the newly discovered 2D-boron structure may lead to practical applications.

Prediction of compounds that defy chemical intuition: Mg3O2 and MgO2

Zhu Q., Oganov A.R., Lyakhov A.O. (2013). Novel stable compounds in the Mg-O system under high pressure. Phys. Chem. Chem. Phys. 15, 7696-7700 (link).

MgO, in the form of ferropericlase (Mg,Fe)O, makes up about 10% of the Earth's volume. Until this work, MgO was believed to be the only thermodynamically stable magnesium oxide. Our work showed that at high pressure new oxides MgO2 and Mg3O2 become stable. These new oxides defy standard chemical intuition and may be planet-forming materials in some planets.

Fate of boranes under pressure

Hu C.H., Oganov A.R., Zhu Q., Qian G.R., Frapper G., Lyakhov A.O., Zhou H.Y. (2013). Pressure-Induced Stabilization and Insulator-Superconductor Transition of BH. Phys. Rev. Lett. 110, 165504 (pdf-file).

Using USPEX in the variable-composition mode, we show that above 153 GPa all traditional boranes (including BH3) become thermodynamically unstable and BH (first appearing as a stable compound at 30 GPa) is the only stable compound of boron and hydrogen at pressures above 153 GPa. At 168 GPa BH undergoes a semiconductor-metal transition, the elegant mechanism of which is investigated using the newly developed (by us, and implemented in the USPEX code as well) variable-cell NEB method. The high-pressure metallic phase is predicted to be a superconductor with a Tc of 14-21 K at 175 GPa.

Stability of xenon oxides and the "missing xenon paradox"

Zhu Q., Jung D.Y., Oganov A.R., Gatti C., Glass C.W., Lyakhov A.O. (2013). Stability of xenon oxides at high pressures. Nature Chemistry 5, 61-65 (pdf-file).

Xenon is an inert gas, and its oxides, though known, are not stable againt decomposition into elemental xenon and oxygen (some even decompose explosively). Using USPEX, we show that at pressures above 83 GPa stable oxides XeO, XeO2, and XeO3 are formed. We also show that there are no stable silicates of xenon in the same pressure range. Furthermore, xenon oxides can only be formed in extremely oxidative environments and thus cannot exist in the Earth's lower mantle. However, due to acquired reactivity under pressure and the ability to readily form strong Xe-O bonds, we conclude that xenon can be trapped by defects and grain boundaries of mantle minerals.

Establishing the structure of an important high energy-density material: Mg(BH4)2

Zhou X.-F., Oganov A.R., Qian G.R., Zhu Q. (2012). First-principles determination of the structure of magnesium borohydride. Phys. Rev. Lett. 109, 245503 (pdf-file).

Mg(BH4)2 is a prime candidate for hydrogen storage (e.g. for rocket fuel) and was earlier found to undergo a phase transition at 1-2 GPa into the delta-Mg(BH4)2 phase, with a dramatic density increase of >40%. Given its exraordinarily low pressure, this tranition could be used for practical purposes. However, we find that the crystal structure of the delta-phase was incorrectly determined from X-ray powder diffraction by Filinchuk et al. (Angew. Chem. 2011): their structure is energetically unfavorable and dynamically unstable. Using USPEX, we surprisingly found another structure that even better describes experimental X-ray diffraction and is much more stable. We have found not only the structure of the delta-phase, but also that of the elusive delta'-phase. This example shows the danger of relying solely on experimental data when solving crystal structures from powder data, and the importance of coupling such experiments with advanced theory.

Prediction of the densest carbon allotrope

Zhu Q., Oganov A.R., Salvado M., Pertierra P., Lyakhov A.O. (2011). Denser than diamond:ab initio search for superdense carbon allotropes. Phys. Rev. B83, 193410 (pdf-file).

This work illustrates how USPEX method can be used for optimizing target physical properties – here, we searched for the densest possible carbon allotrope. Diamond is the densest known material (in the sense that there are more atoms per unit volume in diamond than in any other known material). Nevertheless, we find that denser (by up to ~3.2%) phases are possible and have extremely interesting optical (very high refractive indices and dispersion of light) and electronic (band gaps from 3.0 to 7.3 eV) properties.

Prediction of the structure of graphane

Wen X.D., Hand L., Labet V., Yang T., Hoffmann R., Ashcroft N.W., Oganov A.R., Lyakhov A.O. (2011). Graphane sheets and crystals under pressure. Proc. Natl. Acad. Sci. 108, 6833-6837 (pdf-file, Supporting Online Materials).

Using USPEX, we showed that for CH the most stable structure is not benzene, but graphane (2D-sheet of the diamond structure, passivated by hydrogen atoms). Graphane was found to have several low-energy isomers. It is surprising that benzene, an archetypal organic molecule known since 1825, is less stable than graphane (discovered only in 2009).

High-pressure behavior of methane and its implications for the interiors of planet Neptune

Gao G., Oganov A.R., Wang H., Li P., Ma Y., Cui T., Zou G. (2010). Dissociation of methane under high pressure. J. Chem. Phys. 133, 144508 (pdf-file).

This work clarifies the puzzle of anomalously large heat flux from planet Neptune. An interesting hypothesis was that at high pressures and temperatures methane (CH4), a major component of Neptune (the composition of which can be schematically represented as H2O:CH4:NH3 = 59:33:8), decomposes with the formation of diamond, sinking of which would produce enormous energy. The main uncertainty was whether the decomposition of methane is thermodynamically favorable. Through structure prediction using USPEX, we confirmed that at high pressures and temperatures methane should first polymerize forming ethane+hydrogen (2CH4=C2H6+H2) and then butane+hydrogen (2C2H6=C4H10+H2), and then diamond+hydrogen (C4H10=4C+5H2). This lends strong support to the idea of massive diamond formation in the interior of Neptune. This could also explain the origin of the recently discovered diamond planets.

Study of exotic structures and superconductivity of calcium under pressure

Oganov A.R., Ma Y.M., Xu Y., Errea I., Bergara A., Lyakhov A.O. (2010). Exotic behavior and crystal structures of calcium under pressure. Proc. Natl. Acad. Sci.107, 7646-7651 (pdf-file, Supplementary Material).

Compressed calcium is the highest-temperature elemental superconductor (Tc = 25 K) and its rich polymorphism is still mysterious both from the experimental and theoretical viewpoints. We have predicted, through global optimization using USPEX, the stability of the beta-Sn-like phase of Ca – and this prediction was recently confirmed. Higher-pressure phases have also been predicted, giving the ab initio (DFT) optimal solution for comparison with experiments.

Prediction of stable compounds LiH8, LiH6 and LiH2

Zurek E., Hoffmann R., Ashcroft N.W., Oganov A.R., Lyakhov A.O. (2009). A little bit of lithium does a lot for hydrogen. Proc. Natl. Acad. Sci. 106, 17640-17643. (pdf-file, Supplementary Material).

High pressure fundamentally changes chemical bonding and can lead to violations of basic rules of chemistry. We have predicted, using USPEX, that at pressures around 100 GPa and above LiH will not be the only stable lithium hydride – LiH2, LiH6 and LiH8 will also be thermodynamically stable.

Structure of "superhard graphite"

a - Li Q., Ma Y., Oganov A.R., Wang H., Wang H., Xu Y., Cui T., Mao H.-K., Zou G. (2009). Superhard monoclinic polymorph of carbon. Phys. Rev. Lett. 102, 175506. (pdf-file); b - Oganov A.R., Glass C.W. (2006). Crystal structure prediction using ab initio evolutionary techniques: principles and applications. J. Chem. Phys. 124, art. 244704 (pdf-file).

In 1963, Aust and Drickamer have found a new superhard carbon allotrope. It was obtained by room-temperature compression of graphite and confirmed by numerous subsequent experiments, but its structure could not be solved for nearly 50 years. In 2006, Oganov and Glass reported a low-enthalpy structure of carbon found with USPEX as a metastable state, which closely corresponded to the 2x1 reconstruction of diamond (111) surface, but surprisingly was never proposed before as a 3D-structure. In 2009, we (Li et al.) found that this structure, named M-carbon, perfectly matches all observations for this new carbon allotrope. Several other structures were shown to match experiments as well, however. The dispute was resolved by our calculations (Boulfelfel et al., 2012), which found that among all possible structures M-carbon has the lowest barrier of formation from graphite upon room-temperature compression, and is thus the likeliest phase to form. Shortly afterwards, experimentalists from Yale (Wang et al., 2012) found that only M-carbon matches their high-resolution experimental data.

Theoretical/experimental discovery of a transparent high-pressure phase of sodium

Ma Y., Eremets M.I., Oganov A.R., Xie Y., Trojan I., Medvedev S., Lyakhov A.O., Valle M., Prakapenka V. (2009). Transparent dense sodium. Nature 458, 182-185. (pdf-file, Supporting Online Material).

In 2008, using USPEX, we made a startling prediction that sodium (a nearly free-electron metal) loses its metallic character under pressure, forming an electride insulator at pressures ~2 Mbar, with a wide band gap that rapidly increases with pressure. This prediction was confirmed by experiment. This prediction was marked as one of the major discoveries ever made by Stony Brook University researchers.

Theoretical/experimental discovery of a new superhard and partially ionic phase of boron

a - Oganov A.R., Chen J., Gatti C., Ma Y.-Z., Ma Y.-M., Glass C.W., Liu Z., Yu T., Kurakevych O.O., Solozhenko V.L. (2009). Ionic high-pressure form of elemental boron. Nature 457, 863-867. (pdf-file, Supporting Online Material); b - Solozhenko V.L., Kurakevych O.O., Oganov A.R. (2008). On the hardness of a new boron phase, orthorhombic gamma-B28. J. Superhard Mater. 30, 428-429. (pdf-file).

In 2006, using USPEX, we found the structure of a new stable and superhard boron allotrope, gamma-boron. The paper was published in January 2009 (Oganov et al., Nature 2009) and attracted immediate attention. This discovery was listed as one of the major chemical discoveries of 2009 by Chemistry World (Royal Society of Chemistry). An unusual result was that there is significant charge transfer between different positions of boron atoms (i.e. there is partial ionic character). This result was confirmed by later experimental study (Mondal et al., 2011). Gamma-boron is the hardest known boron allotrope and one of 5-6 hardest solids known to date, as follows from our measurements of its hardness (Solozhenko et al., 2008).

Selected as one of major discoveries in "Cutting edge chemistry of 2009" (Chemistry World, published by the Royal Society of Chemistry, 18 December 2009)

Prediction of a superconducting state (with Tc=64 K) in germane, GeH4

Gao G., Oganov A.R., Bergara A., Martinez-Canalez M., Cui T., Iitaka T., Ma Y., Zou G. (2008). Superconducting high pressure phase of germane. Phys. Rev. Lett. 101, 107002 (pdf-file).

Using USPEX, we have predicted a high-pressure superconducting phase of germane with an unusually high superconducting Tc of ~64 K.

Evolutionary methodology for crystal structure prediction

a - Oganov A.R., Glass C.W. (2006). Crystal structure prediction using ab initio evolutionary algorithms: principles and applications. J. Chem. Phys. 124, art. 244704 (pdf-file); b - Lyakhov A.O., Oganov A.R., Valle M. (2010). How to predict very large and complex crystal structures. Comp. Phys. Comm. 181, 1623-1632 (pdf-file); c - Oganov A.R., Lyakhov A.O., Valle M. (2011). How evolutionary crystal structure prediction works - and why. Acc. Chem. Res. 44, 227-237 (pdf-file).

Until 2006, crystal structure prediction was widely thought to be an insoluble problem. This all changed when the USPEX method (Oganov & Glass, 2006; Glass et al., 2006) was developed. Our USPEX code has since been the widest used and the most powerful code in this new field, and most existing codes (XtalOpt, Calypso, Muse) have started from older versions of USPEX. The development of USPEX has been called “revolutionary” (Chaplot & Rao, 2006), as it has opened the new field of computational materials discovery and has led to many fundamental and applied discoveries.

Polytypes of MgSiO3 and implications for Earth's mantle

Oganov A.R., Martonák R., Laio A., Raiteri P., Parrinello M. (2005). Anisotropy of Earth’s D” layer and stacking faults in the MgSiO3 post-perovskite phase. Nature 438, 1142-1144 (pdf-file).

Using metadynamics simulations, we show that perovskite and post-perovskite phases of MgSiO3 are related and can be considered as polytypes.

Discovery of a new high-pressure phase of alumina, Al2O3

(Oganov A.R., Ono S. (2005). The high pressure phase of alumina and implications for Earth’s D” layer. Proc. Natl. Acad. Sci. 102, 10828-10831 (pdf-file)).

We have predicted that Al2O3 has a high-pressure phase isostructural with MgSiO3 post-perovskite. Increased electrical conductivity observed for Al2O3 at conditions where we predict this phase to be stable suggests that MgSiO3 post-perovskite is likely to be an ionic conductor. High electrical conductivity of post-perovskite D” layer would, by electromagnetic coupling with the Earth’s magnetic field, explain the observed decadal variations of the length of day. These conclusions were confirmed by experimental measurements of Ohta et al. (2008).

Discovery of the post-perovskite phase of MgSiO3

Oganov A.R. & Ono S. (2004).Theoretical and experimental evidence for a post-perovskite phase of MgSiO3 in Earth's D" layer. Nature 430, 445-448. (pdf-file).

Until 2004, anomalies of the Earth’s D” layer, known since 1950, had no convincing explanation. Guided by experimental insight of Shigeaki Ono, I found that the D” boundary corresponds to a new phase transition of MgSiO3 – from well-known perovskite to the new post-perovskite phase with a CaIrO3-type structure. This prediction was verified by experiments of S. Ono (Oganov & Ono, Nature 2004) and by independent experiments (Murakami et al., Science 2004). Properties of MgSiO3 post-perovskite have explained the anomalous properties – as we showed (Oganov & Ono, 2004), the seismic velocity jumps at the D” boundary, unusually large topography of the D” boundary, seismic anisotropy of this layer and anticorrelations between seismic wave velocities are all quantitatively explained by the post-perovskite transition. This transition also has implications for the Earth’s evolution and probably also for the origin of life.