Xenon oxides under pressure

Transcription

1 Xenon oxides under pressure Richard Needs and Nicholas Worth (Uni of Cambridge, UK) Agnès Dewaele (CEA, DAM, DIF, France) Chris Pickard (UCL, UK) Sakura Pascarelli (ESRF, France) Olivier Mathon (ESRF, France) Mohamed Mezouar (ESRF, France) Tetsuo Irifune (Ehime University, Japan)

2 Xenon oxides Xenon is not very reactive at low pressures Xenon can be made to form metastable oxides at low pressures which readily decompose DFT study by Zhu, Jung, Oganov, Glass, Gatti and Lyakhov [Nature Chem 5, 61 (2013)] considered XeO 2, XeO 3 and XeO 4 Zhu et al. predicted stable xenon oxides above 83 GPa with oxidation states of +2, +4, +6 No experimental evidence for the formation of xenon oxides at high pressures has been reported

3 Questions about xenon oxides What are the stable stoichiometries of Xe/O mixtures at high pressures? What are the stable structures of xenon oxides? What are the oxidation states of xenon atoms in the oxides? At what pressure do xenon oxides become thermodynamically stable? Metals or insulators? Role of Xe 4d electrons?

4 Methodology Diamond anvil cell experiments at ESRF Use laser heating to try and synthesise equilibrium phases Characterise xenon oxides using X-ray diffraction (XRD), X-ray absorption, and Raman spectroscopy Use Ab initio random structure searching (AIRSS) within DFT to identify stable stoichiometries and structures

5 The missing xenon paradox Chondrite: a stony meteorite unmodified by melting or differentiation of the parent body Xenon depletion of >90% in the Earth s atmosphere, compared with its abundance in chondritic meteorites Possible resolutions of the paradox include: Xe could have escaped into space Xe could be stored within Earth s mantle Xe could be stored within Earth s core Solving the paradox will require improved understanding of the high-pressure chemistry of xenon We have studied the stability of compounds of xenon with the most abundant element in the Earth s mantle, oxygen

6 XRD data 33% Xe+67% O 2 89% Xe+11% O 2 before heating after heating Xe Xe(O 2 ) 2 Re Xe 2 O 5 before heating after heating Xe Xe(O 2 ) 2 Re Xe 2 O 5 Xe 3 O 2 x10 6 P~84 GPa, after heating P~97 GPa, after heating P~84 GPa, before heating P~95 GPa, before heating q (deg.) q (deg.) 15

7 Convex hull at 83 GPa Formation Enthalpy (ev/atom) x in Xe x O 1-x Structures on the convex hull are stable against decomposition Structures above the convex hull are unstable to decomposition Blue: Structures of Zhu et al. recalculated at high accuracy

8 Combining theory and experiment The Xe 2 O 5 phase was discovered in XRD experiments and identified using AIRSS The Xe 3 O 2 phase was predicted using AIRSS and subsequently verified by XRD experiments Fruitful dialogue between theory and experiment

9 Structure of Xe 2 O 5 with +4 and +6 xenon atoms

10 Structure of Xe 3 O 2 with 0 and +4 xenon atoms

11 Cell parameters PathH:Agnes:XeO2:NewPhaseXeO2:XeO2_8_EoS; file: cc1 lun. 24 févr a K0=176, K'0=4 c/2 K0=120, K'0=4 180 c/a b/a d (Å) 5.2 V/3 f.u. (Å 3 ) P (GPa) P (GPa) P (GPa) Xe 2 O 5 Xe 3 O 2

12 Raman Data PathH:Agnes:XeO2:NewPhaseXeO2:Raman_78_reaction_EHPRG; file: lun. 3 févr PathH:Agnes:XeO2:NewPhaseXeO2:Raman_78_reaction_EHPRG; file: LambdamaxSauv jeu. 6 févr Xe 2 O 5 after decompression XeO 4 x10 3 XeO 3 XeO Intensity Raman shift (cm -1 ) Xe 3 O 2, P=97 GPa Raman frequency (cm -1 ) 400 XeO2 XeO3 XeO4 Xe 2 O 5, P=88 GPa Raman shift (cm -1 ) V/V Xe 2 O 5 and Xe 3 O 2 Xe 2 O 5

13 Bandstructures of Xe 2 O 5 and Xe 3 O 2 at 83 GPa Xe2O5-P4nnc-83GPa.BU Xe3O2-Immm-83GPa ε (ev) 0 ε (ev) ,0,0 0,0,1/2 1/2,1/2,1/2 1/2,1/2,0 0,1/2,0 0,1/2,1/2 0,0,0-40 0,1/2,0 0,0,0 1/4,1/4,0 1/4,1/4,1/4 1/4,0,1/4 0,1/4,1/4 0,0,0 Xe 2 O 5 Xe 3 O 2

14 Influence of the Xe 4d electrons Xe 4d orbitals lie 45 ev below the HOMO and 30 ev below the bottom of the sp bands Surely they contribute very little to the bonding and they can be replaced by the pseudopotential?

15 Influence of the Xe 4d electrons The Xe 4d orbitals have a strong influence on the energies of structures Surely they contribute very little to the bonding and they can be replaced by the pseudopotential?

16 Acknowledgements: ESRF, EPSRC, ARCHER, UKCP, Cambridge High Performance Computing Service (Darwin)