Project Topics
Get Complete Project Material File(s) Now! »
Xe in clathrates (van der Waals bonds)
A well-known, geologically relevant type of clathrate is ice clathrate, naturally present in the permafrost, in south and north poles, with CO2 and CH4 as guest molecules. There are different clathrate hydrate structures, the most commons (sI, sII (cubic structure I and II) and sH (hexagonal structure)) consist of combination of 12 pentagons of assembled water molecules. They have been extensively studied (Sloan and Koh, 2007) among other due to the risk of potential greenhouse gases release from ice polar clathrates in case of intensive ice melting linked with global warming. Xe is also among the gases that stabilizes clathrate hydrate structures through van Der Waals interactions. Xe ice clathrates were shown to be stable under two phases up to 2.5 GPa before breaking down (Dyadin et al., 1997; Sanloup et al., 2002). Though natural clathrates in polar ice contain very few Xe, stability of Xe-bearing ice clathrate above 50 GPa and 1500 K proves the possibility of their presence at depth in giant planets or in comets (Iro et al., 2003).
Zeolites are microporous, aluminosilicate minerals (more than 200 different types identified so far), characterized by their cage-like structure. They are strictly speaking part of the clathrate group. Their particular structure is the origin of their adsorbent and catalytic properties frequently used in the industry. Xenon is one of the guest molecules which can enter cages of zeolite (Ito 1984; Heo, 1999; Seoung et al, 2014). Seoung et al. (2014) studied incorporation of Xe in Ag natrolite (Ag-NAT) at 1.7 GPa and 250°C. Xe enters the structure of Ag-NAT while nanoparticles of metal Ag are deposited at the crystal surface. In presence of Xe in the structure Ag+ converts to Ag2+ inside the zeolite, linked to an oxidation of water or of lattice oxygen. Ag2+ is very unstable and is normally reduced to Ag +, it is possible that high polarizability of Xe stabilizes Ag2+ (as proposed by Walker et al. (1999)) potentially acting as an adduct, as for Au(I)F (Kurzydłowski and Grochala, 2008). At ambient P Xe is not released and remains inside the Ag-NAT; heating is needed to trigger desorption. Similar experiences with Kr prove that Kr is released, although Seoung et al. (2014) did not report if Ag2+ is also produced with Kr.
Xe in stoichiometric oxides
The Earth is generally-speaking an oxidised environment and bonding with oxygen is predominant in the majority of Earth’s minerals. XeO3 (Smith, 1963; Templeton et al., 1963) and XeO4 with Xe-O bond length of 1.736 Å (Selig et al., 1964; Gundersen, 1970) have been synthesized from Xe-F compounds (Figure B3), although reactions of formation are endothermic and the Xe compounds are highly unstable. A stable xenon dioxide (XeO2) has been synthetized at 273 K by hydrolysis of XeF4 in acid water (Brock and Schrobilgen, 2011).
Though the structure has not been fully resolved, the authors proposed a planar oxygen environment around Xe (Figure B3) based on Raman analysis. Theoretical calculations have been performed to investigate possible structures of xenon oxides at high P. Based on density functional theory (DFT) and within the generalized gradient approximation (GGA), Zhu (2012) predicted stability of XeO, XeO2 and XeO3 respectively above 83 GPa, 102 GPa, and 114 GPa linked with an increasing oxidation state of Xe with P (Xe0, Xe2+, Xe4+, Xe6+) whereas Hermann and Schwerdtfeger (2014) predicted stability of Xe3O2 above 75 GPa.
Recently Dewaele et al. (2016) observed two Xe oxides, at 83 GPa and 97 GPa and 2000 K in diamond anvil cell using in situ X-ray diffraction, Raman and X-ray absorption spectroscopies. Xe2O5, an insulator, is stable under rich O conditions (metastable till 30 GPa) and composed of Xe4+ and Xe6+. Xe3O2, a semimetallic compound, is stable under poor O condition (metastable till 38 GPa) and composed of Xe0 and Xe4+ (Figure B3).
Regarding Xe/H2O phases, above 50 GPa and 1500 K, a stable Xe compound with covalent Xe-O bonds: Xe4O12H12 was found (Sanloup et al., 2013). Xe4O12H12 has a hexagonal structure with 2 Xe2O6H6 units (as shown on figure B3) per unit cell. Existence of stable Xe-H2O phase at these P-T conditions could be relevant for giant planets such as Uranus and Neptune but eventually also for cold slabs subducting into the deep mantle.
High P-T generation
Piston cylinder press was used to dope silicate with noble gases under high P-T conditions. In our work new synthesized samples, as well as samples obtained from the previous study of Alexander Griffith, Master student supervised by Chrystèle Sanloup, are presented (noted as ‘previous’ in experimental tables). To perform in situ studies diamond anvil cells and Paris Edinburgh press were used at synchrotron sources.
Piston cylinder press and gas loading devices
The piston cylinder press is a solid medium apparatus used to compress large volume samples (> 1 mm3), it can generally reach P up to 5 GPa and T close to 2000°C (Figures C1, C2). A comprehensive historical evolution of piston cylinder devices can be found in Hall (1980). Sir Charles Parsons developed the first piston cylinder type apparatus motivated by the disappointed aim of diamond synthesis (1880-1928). Further development by Coes (1962) was marked by the synthesis of coesite (the high P polymorph of silica). Piston cylinder presses are now routinely used in Earth’s sciences and are nearly fully described by Boyd and England (1960).
The piston cylinder press is based on a Bridgman-type system: a piston slides into an end-loaded cylinder (both generally made of Tungsten Carbide (WC)). The upper plate remains fixed whereas a hydraulic ram applies strength on the bottom plate due to injection of pressurized oil into a chamber: P is generated through compression of the sample. Resistive-heating is achieved through a graphite furnace which enables to get high and rather homogeneous T through the whole capsule. T is read with a thermocouple placed near the top of the capsule.
In situ synchrotron-based probes at high P-T conditions
In situ experiments were performed at synchrotron sources during five beamtimes: indeed a highly focused and highly brilliant beam was needed to obtain high-quality X-ray diffraction, X-ray absorption, and Infrared data, on small samples under extreme conditions.
For X-ray absorption spectroscopy at the Kr and Xe K-edges (14.3 keV and 34.6 keV respectively) a high energy, high brilliance X-ray beam was needed in combination with a scanning in energy and a high beam stability due to long data-collection. Experiments were performed at beamline BM23 (a bending magnet beamline) at the ESRF, Grenoble, France.
For X-ray diffraction, experiments were performed at beamline P02.2 (an undulator beamline) at PETRA III, Hamburg, Germany and at beamline I15 (a wiggler beamline) at Diamond Light Source, Oxfordshire, United Kingdom.
For Infrared spectroscopy, experiments were performed on SMIS beamline (edge and bending magnet beamline) at SOLEIL Synchrotron, Gif-sur-Yvette, France. Synchrotron beam is obtained through injection in a storage ring of electrons at near light speed. These electrons are accelerated in a magnetic field in curved sections by bending magnets or in straight sections via arrays of magnets called insertion devices. A broad X-ray spectrum is created. X-rays are emitted in the direction of the electron beam and are polarised in the plane of the ring. Quality of synchrotron beam is defined by the brilliance (in photon/s/mrad²/mm²/0.1%BW), which is the intensity (photon per second) divided by the angular divergence (mrad²), the beam size (mm²) and the spectra distribution (the bandwidth expressed in 0.1%). Insertion devices, wigglers and undulators, impose beam brilliance. Use of wigglers, magnets of opposite polarity in a row making electrons wiggling back and forth, multiply brilliance by two, whereas undulators, magnets of opposite polarity create small deflection inducing constructive interference between electrons, multiply brilliance by four11. In addition to brilliance, stability of the beam is also a crucial parameter.
Three generations of synchrotron source are generally defined. The first generation made use of a storage ring initially built for particle physics purposes. The second generation was initially built to produce X-rays. The third generation implemented insertion devices.
In a synchrotron source there are various beamlines (49 at the ESRF for instance), depending on their location, they use X-rays created by bending magnets or insertion devices. Each beamline is dedicated to specific techniques and experiments.
X-ray diffraction on crystalline material
Here the principles of X-ray diffraction on crystalline material are briefly presented before focusing on powder X-ray diffraction and its treatment (Rietveld refinement). This overview of a more than 100 year-old technique is based on review by Lavina et al. (2014) and the work of Hunter and Howard (1998) for Rietveld refinement.
X-ray diffraction is an elastic (Thomson) scattering method i.e. incident and diffracted X-rays have the same wavelength comprised between 0.1 nm and 10 nm for X-rays. X-rays are scattered by electrons implying that scattering power of an atom is function of its atomic number (atomic scattering factors can be found in Brown et al., 2006b). Interference between scattered waves coming from the electronic cloud explains why scattered intensity decreases at large angle. Distribution of the electrons is supposed to be spherical around the nucleus, although thermal vibrations as well as static disorder (substitution on an atomic site for instance) can affect this distribution.
X-rays are diffracted if there is a periodicity in the electron distribution, i.e. the material is ordered enough. Furthermore, the periodic structure needs to have a length comparable to X-ray wavelength, which is normally the case for lattice parameters of a crystal. Diffraction phenomenon can be described within the Bragg’s representation, which defines diffraction as the reflection of X-rays by a family of crystallographic planes noted (hkl), separated from each other by a distance ℎ . ℎ can be calculated for each family .
Table of contents :
AKNOWLEDGEMENTS
SUMMARY
RESUME EN FRANCAIS
ABSTRACT
ORGANIZATION OF THE PRESENT WORK
Chapter A: INTRODUCTION (GEOCHEMISTRY)
I/ Noble gases in geosciences
1/ Helium
2/ Neon
3/ Argon
4/ Krypton
5/ Xenon
6/ Xe and Kr as nuclear fission products
II/ Evolution of the Earth-atmosphere system and noble gases
1/ Origin of noble gases
2/ Atmospheric loss
3/ Degassing processes
4/ Recycling at depth
III/ The ‘Missing Xenon’ and the Xenon Paradox
1/ The ‘Missing Xenon’
2/ Xe paradox
3/ Hypotheses to explain the ‘Missing Xenon’ and the Xenon paradox
Chapter B: STATE OF THE ART: XE REACTIVITY AND INCORPORATION IN EARTH’S RELEVANT MATERIALS
I/ Xe physical properties
1/ Noble gases atomic radii
2/ Diffusion and adsorption properties
3/ Xe phase diagram
4/ Xe in metallic phases at high P
II/ Xe in oxides and silicates
1/ Xe in clathrates (van der Waals bonds)
2/ Xe in stoichiometric oxides
3/ Xe as a minor or trace element in silicates
4/ Xe in silicate melts
III/ Prospects
Chapter C: EXPERIMENTAL METHODS
I/ High P-T generation
1/ Piston cylinder press and gas loading devices
2/ Diamond anvil-cell (DAC)
3/ Paris Edinburgh Press (PEP)
4/ P-T calibration
II/ In situ synchrotron-based probes at high P-T conditions…
1/ Synchrotron source
2/ X-ray absorption spectroscopy (XAS)
3/ X-ray diffraction on crystalline material
4/ Infrared (IR) spectroscopy
III/ Analytical methods
1/ Scanning Electron Microscopy (SEM)
2/ Transmission Electron Microscopy (TEM)
3/ Electron microprobe analyses (EMPA)
4/ Raman spectroscopy
IV/ Theoretical modelling of Xe incorporation in minerals
1/ Density functional theory (DFT) in condensed matter
2/ Infrared and Raman spectra modeling
3/ Computational details
4/ Previous modeling of Xe within the DFT framework
Chaper D: XE INCORPORATION IN OLIVINE
I/ Ex situ analyses
1/ Sample synthesis
2/ SEM / EMP analyses
II/ New constraints on Xe incorporation in olivine from First-principles calculations (Crépisson et al., 2018)
1/ Introduction
2/ Methods
3/ Results
a/ Reference compounds and pure forsterite
b/ Xe incorporation models
c/ Cell parameters vs. Xe concentration
d/ Raman spectra
4/ Conclusion
III/ In situ IR spectroscopy on Xe -bearing olivine
1/ Methods
2/ Results and discussion
IV/ New in situ X-ray diffraction on synthetic olivine
1/ Methods and samples
2/ Results and discussion
V/ Conclusion
Chapter E: XE INCORPORATION IN QUARTZ
I/ Sample synthesis
II/ The Xe-SiO2 system at moderate P and high T (submitted to G3)
1/ Methods
a/ X-ray diffraction
b/ Infrared spectroscopy
c/ Ab initio calculations
2/ Results and discussion
a/ Increase of unit-cell volume of quartz in presence of Xe
b/ A new (Xe,Si)O2 phase at high temperature
III/ Conclusion
Chapter F: XE INCORPORATION IN HIGH-TEMPERATURE FELDSPAR (SANIDINE)
I/ Ex situ analyses
1/ Sample synthesis
2/ SEM / TEM / EMP / Raman analyses
II/ In situ X-ray diffraction
1/ Methods and samples
2/ Results for dry sanidine
3/ Results for K-cymrite and sanidine at high water fugacity
III/ Conclusion
Chapter G: XE AND KR INCORPORATIONS IN FELDSPATHIC GLASS AND MELT
I/ Sample synthesis
II/ Kr environment in a feldspathic glass and melt: a high P-T XAS study (published in Chem. Geol.)
1/ Introduction
2/ Methods
a/ Sample: synthesis and characterization
b/ Experimental set-up
c/ X-ray diffraction and XAS measurements
d/ Processing of XAS data
3/ Results and Discussion
a/ XAS data at the Kr K-edge
b/ Resolving Kr environment in sanidine glass and melt
4/ Conclusion …
III/ In situ XAS study of Xe incorporation in a feldspathic glassy and crystalline sample
1/ Methods
2/ XAS at the Xe K-edge in sanidine (glass and crystal)
3/ Conclusion
Chapter H: GENERAL CONCLUSIONS AND PERSPECTIVES
REFERENCES
