La lecture en ligne est gratuite
Le téléchargement nécessite un accès à la bibliothèque YouScribe
Tout savoir sur nos offres
Télécharger Lire

(Na,K) Aluminosilicate Hollandites [Elektronische Ressource] : structures, crystal chemistry, and high-pressure behaviour / vorgelegt von Jun Liu

De
154 pages
“(Na,K) Aluminosilicate Hollandites: Structures, Crystal Chemistry, and High-pressure Behaviour” Von der Fakultät für Biologie, Chemie und Geowissenschaften der Universität Bayreuth Zur Erlangung des akademischen Grades Doctor der Naturwissenschaften -Dr. rer. nat.- genehmigte Dissertation vorgelegt von Jun Liu Bayreuth, 2007 Table of contents _ Table of contents Summary 1 Zusammenfassung 5 1. Introduction 10 1.1The importance of potassium and sodium in the deep Earth……………..…………10 1.1.1 Potassium-40 and the Earth’s core………………………………………..….11 1.1.2 Importance of potassium and sodium in the lower mantle…….…………….11 1.2 K-Na aluminosilicate hollandite: possible host mineral for K and Na in the Earth’s mantle……………………………………………………………………….………………...13 1.2.1 The hollandite structure…………………………………….………………..13 1.2.
Voir plus Voir moins

“(Na,K) Aluminosilicate Hollandites: Structures, Crystal
Chemistry, and High-pressure Behaviour”






Von der Fakultät für Biologie, Chemie und Geowissenschaften der Universität Bayreuth




Zur Erlangung des akademischen Grades
Doctor der Naturwissenschaften
-Dr. rer. nat.-




genehmigte Dissertation




vorgelegt von

Jun Liu












Bayreuth, 2007




Table of contents _
Table of contents

Summary 1

Zusammenfassung 5

1. Introduction 10
1.1The importance of potassium and sodium in the deep Earth……………..…………10
1.1.1 Potassium-40 and the Earth’s core………………………………………..….11
1.1.2 Importance of potassium and sodium in the lower mantle…….…………….11
1.2 K-Na aluminosilicate hollandite: possible host mineral for K and Na in the Earth’s
mantle……………………………………………………………………….………………...13
1.2.1 The hollandite structure…………………………………….………………..13
1.2.2 The (K,Na)AlSi O hollandite solid solution…………………………….…..15 3 8
1.2.3 The experimental stability field of (K,Na)AlSi O hollandite…………….…16 3 8
1.2.4 The natural occurrence of (K, Na) aluminosilicate hollandite: hollandite in
meteorites……………………………………………………………………………………...18
1.3 Previous study……………………………………………………………….……..19
1.3.1 High-pressure study of the KAlSi O hollandite…………………………….19 3 8
1.3.2 Study of the KAlSi O - NaAlSi O hollandite solid solution……………….21 3 8 3 8
1.4 Aim of this research……………………………………………………….………23

2. Experimental techniques and sample characterization 24
2.1 Samples……………………………………………………………………………24
2.1.1 Starting materials……………………………………………………………24
2.1.2 Multi anvil technique………………………………………………………..26
2.1.3. Characterisation of the run products………………………………………..30
2.1.3.1 Chemical analysis……………………………………….………….30
2.1.3.2 X-ray powder diffraction…………………………………………...32
2.1.3.3 X-ray single-crystal diffraction………………………….………….33
2.2 High-pressure and high-temperature techniques………………………...……….36
Table of contents _
2.2.1 Diamond anvil cells…………………………………………..………….…36
2.2.2 Raman experiments………………………………………………………...39
2.2.3 High pressure X-ray powder diffraction experiments………….…………..40
2.2.3.1 High-pressure X-ray experiments at the BGI……………………...40
2.2.3.2 Synchrotron radiation X-ray powder diffraction…………………..41

3. Results 43
3.1 (K,Na)AlSi O hollandite solid solution………………...………..………..………43 3 8
3.1.1 Chemical composition……………………………………………………...43
3.1.2 X-ray powder diffraction…………………………………………………...46
3.1.3 Raman spectra………………………………………………………………50
3.1.4 Crystal structure…………………………………………………………….53
3.2 High pressure and high temperature behaviour of LiF……………………………57
3.3 High pressure behaviour of Na K AlSi O hollandite in helium as pressure 2 8 3 8
transmitting medium…………………………………………………………………………..58
3.3.1 Unit-cell lattice parameter determination at room T ………………………58
3.3.2 Equation of state…………………………………………………………...63
3.3.2.1 The f-F plot……………………………………………………..65
3.3.3 Spontaneous strain:…………………...…………………………………...67
3.4 High pressure behaviour of K Na AlSi O hollandite in helium as pressure 0.5 0.5 3 8
transmitting medium ………………………………………………………………………….72
3.4.1. The equation of state……………………………………………………...73
3.5 High pressure behaviour of K Na AlSi O hollandite in lithium fluoride as 0.8 0.2 3 8
pressure transmitting medium ………………………..………………………...……………..73
3.5.1 The Equation of state……………………………………………………...74
3.6 High pressure behaviour of K Na AlSi O hollandite in lithium fluoride as 0.6 0.4 3 8
pressure transmitting medium …………………………………………………………….…..74
3.6.1 The Equation of state……………………………………………………..76
3.7 High pressure high-temperature behaviour of K Na AlSi O hollandite in lithium 0.5 0.5 3 8
fluoride as pressure transmitting medium…...………………………………………………...79
3.8 High pressure Raman experiments………………………………………...……..84
Table of contents _
3.8.1 High pressure Raman experiments of KAlSi O hollandite in argon as 3 8
pressure transmitting medium……………………………………………………….………84
3.8.2 High pressure Raman experiment of K Na AlSiO hollandite in 0.6 0.4 8
lithium fluoride as pressure transmitting medium ……………………………….………….89

4. Discussion 93
4.1 The (Na,K)AlSi O hollandite solid solution………………….……………….93 3 8
4.1.1 Solid solution behaviour………………………………….…………..93
4.1.2 The NaAlSi O hollandite………………………………….…………95 3 8
4.1.3 The symmetry in (K,Na)AlSi O hollandite: tetragonal vs. 3 8
monoclini…………………………………………………………………………….………..96
4. 2 High pressure and high temperature behaviour of (Na, K) h….…….………...102
4.2.2 The Spontaneous strain behaviour of the tetragonal to monoclinic high-
pressure phase transition in hollandite……………………………………………………….103
4.2.2.1. Spontaneous strain behaviour under hydrostatic
condition……………………………………………………………………………………..103
4.2.2.2. Spontaneous strain behaviour in lithium fluoride………….104

5. Conclusions 107

References 109

Appendix 127
Appendix 1: Indexed peaks in lattice parameter refinement………………………127
Appendix 2: Rietveld Refinement Patterns from high pressure experiment of
K Na AlSi O hollandite in helium……………………………………………………….136 0.8 0.2 3 8

Erklärung 148





Acknowledgements

There are many people I wish to thank for their help and support. First of all, I want to thank Tiziana
Boffa-Ballaran, the person I owe most to at Bayerisches Geoinstitut. I am so grateful to her, for all her
help, suggestions and support, which made completion of this work possible. Tiziana, thank you so
much! Both of my supervisers Tiziana and Leonid Dubrovinsky have been very supportive and very
patient with me, and have given me the most help. I thank Leonid for all the helps, great suggestions,
and all the work you have put into being my superviser!

I thank Dan Frost for all his help in multi-anvil lab, and for giving me important suggestions and helps
during my study. Thank you Florian Heidelbach for the help and effort you’ve done for my thesis.
Thank you Fritz Seifert for all the kind help and support you’ve given me. Thank you Detlef Krausse
and Anke Potzel for giving me lots of help in the probe lab, and Detlef also fixed all my computer
problems. Thank you Stefan Keyssner for the help and making me feel at home here. Thank you
Innokenty Kantor and Alexander Kurnosov for all your helps in Raman labs. I thank Ahmed El Goresy
for all his help and suggestions. Thank you Nobuyoshi Miyajima for all the help. I thank Gerd Steinle-
Neumann for all the help and knowledge I learned here!

I am also grateful to Andreas Audétat, Xin Nie, and Zhengning Tang for all your support and
encouragement during the writing of this thesis. Thank you Ute Mann, Anastasia Kantor, Innokenty
Kantor, and Ashima Saikia for all your helps during my study and during my writing of the thesis. And
thank Enikö Bali, Gudmundur Gudfinnsson, Lora Armstrong, and Shantanu Keshav for being there
and cheering me up!

Special thanks to our great Secretaries, Petra Buchert and Lydia Kison-Herzingfor all the important
help you’ve given me. Especially Petra, always so nicely and efficiently dealing with all kinds of my
problems and requests, I would have been in much more trouble without your help!

The last and the most important one, I thank Mike Terry for all his help, suggestions, support,
encouragement and everything else… This work was done for both you and me.




Vollständiger Abdruck der von der Fakultät für Biologie, Chemie und Geowissenschaften der
Universität Bayreuth genehmigten Dissertation, zur Erlangung des akademischen Grades
Doktors der Naturwissenschaften (Dr. Rer. Nat.).














Prüfungssausschuß:

Prof. S. Peiffer, Universität Bayreuth (Vorsitzender)
PD. Dr. L. Dubrovinsky, Universität Bayreuth (1. Gutachter)
Prof. F. Langenhorst, Universität Jena (2. Gutachter)
Prof. D. Rubie, Universität Bayreuth
Prof. J. Breu, Universität Bayreuth


Datum der Einreichung der Dissertation: 23. Mai 2007 des Wissenschaftlichen Kolloquiums: 23. Juli 2007












Summary _ 1
Summary

Aluminosilicates with the composition (Na,K)AlSi O and the dense hollandite-type structure, 3 8
in which all Si and Al are in six-fold coordination, are considered as a possible repository of
potassium and sodium in the Earth’s mantle. Behaviour of Na and K in the deep mantle are of
40considerable interest from geophysical and geochemical points of view, because K is one of
the important heat sources during the evolution of the Earth, and alkali elements play an
important role in stabilizing a variety of Al-rich phases in the lower mantle. In previous study
of the phase relations in the system KAlSi O - NaAlSi O , the maximum solubility of 3 8 3 8
NaAlSi O component into KAlSi O hollandite-type structure was found to be 40 – 51mol% 3 8 3 8
(Yagi et al., 1994; Liu, 2006). For higher Na content the high-pressure phase appears to be that
of the calcium-ferrite type structure. However, natural occurrences of NaAlSi O hollandite 3 8
have been reported in shock-induced melt veins of chondrite. In recent high-pressure studies
of KAlSi O hollandite, a phase transformation from tetragonal to monoclinic structure at 3 8
about 20 GPa has been reported (Ferroir et al., 2006). Until now there has been no report on
high-pressure behavior of mixed compositions in the system KAlSi O - NaAlSi O . The aim 3 8 3 8
of this research is to explore the phase relation of the K-Na system at different temperatures
and pressures, and to determine the physical-chemical properties and high-pressure behaviour
of silicate hollandite-type structures containing K and Na in different concentrations.

The (Na,K)AlSi O hollandite solid solution has been synthesised using multi-anvil apparatus 3 8
in the pressure range between 13 and 26 GPa and temperatures between 1500 and 2200 °C,
using (Na , K )AlSi O glasses, NaAlSi O glass, and Na K Ca AlSi O glass as 0-0.6 1-0.4 3 8 3 8 0.75 0.05 0.1 3 8
starting materials. With increasing pressure, the solubility of Na component into the KAlSi O 3 8
hollandite end-member increases. Also by increasing temperature, the stability field of
(Na,K)AlSi O hollandite becomes larger. Homogeneous assemblages with a pure hollandite 3 8
phase (and maximum 1-2% of stishovite) were synthesized at temperature of 1700 °C and
different pressures with 0%, 10%, 30%, 40%, 50% of NaAlSi O component. Further pressure 3 8
and temperature increases had no effect on increasing the Na solubility. No pure NaAlSi O 3 8
hollandite end-member was succsessfully synthesized. In experiments with
Na K Ca AlSi O glass as starting material used to simulate the composition of natural 0.75 0.05 0.1 3 8
Summary _ 2
occurring NaAlSi O hollandite in shock induced melt veins in meteorites, a complex mixture 3 8
of phases, among which a hollandite with composition K Na Ca Al Si O was 0.26 0.53 0.16 1.15 2.86 8
synthesized at 22 GPa and 1900 °C. Although the Na component in this hollandite is still not
higher than 53 mol%, the ratio Na/(Na+K+Ca) = 0.56 is greater than the Na/(Na+K) ratio
obtained from any other experiments performed in this study. Considering the difference in
heat dissipation between the shock events in meteorites and the multi-anvil presses, it appears
likely that NaAlSi O hollandite forms as a result of local high pressure and high temperature 3 8
conditions and really fast quenching under non-equilibrium conditions.

All synthesized hollandite samples have tetragonal I4/m symmetry at ambient conditions. The
unit-cell volume and lattice parameters of the (Na,K)AlSi O hollandite decreases linearly 3 8
with increasing Na content. The a cell parameter decreases more rapidly than the c cell
parameter, suggesting that changing the cation size in the tunnels of the hollandite structure
affects more the a axis than the c axis, probably because the c axis length depends mostly on
the cation to cation repulsion across the shared octahedral edges of the octahedral double
chains, and only in a minor way on the size of the tunnel cations. Structural refinements of
single-crystal data collected for KAlSi O and K Na AlSi O hollandites are consistent with 3 8 0.8 0.2 3 8
Si and Al disorder among the octahedral sites. The major difference between the KAlSi O 3 8
hollandite end-member and the K Na AlSi O sample is the presence in the latter of a split 0.8 0.2 3 8
site away from the 4th-fold axis. This position, occupied by ~ 75% of the total Na content, is
closer to the framework walls and has a very distorted coordination polyhedron with only 5
Na1-O bond distances between 2.4 and 2.6 Å whereas all other Na1-O bond distances are
larger than 3 Å. The Si(Al)O octahedra in KAlSi O and K Na AlSi O hollandite are more 6 3 8 0.8 0.2 3 8
distorted with respect to those in stishovite, and the O1b-Si(Al)-O2b angle subtended by the
shared horizontal edge is larger than the corresponding angle in stishovite, probably due to the
3+ 4+substitution of Al for Si .

The high pressure behaviour of hollandite samples with compositions of KAlSi O , 3 8
K Na AlSi O , K Na AlSi O and K Na AlSi O have been studied using diamond 0.8 0.2 3 8 0.6 0.4 3 8, 0.5 0.5 3 8
anvil cells and different pressure transmitting media, by means of X-ray powder diffraction
and Raman spectroscopy. High temperature behaviour of K Na AlSi O hollandite at high 0.5 0.5 3 8
Summary _ 3
pressures has also been explored by means of X-ray powder diffraction. At high pressures, all
tetragonal hollandite samples transform to a monoclinic (hollandite II) structure with space
group I2/m. The transition pressure decreases with increasing Na component. Na substitution,
thus, stabilizes the monoclinic phase, likely because the framework walls are more distorted
than in the tetragonal phase and therefore more apt to accommodate the smaller Na atom.
Second order Birch- Murnaghan equations of state were calculated for the tetragonal and
monoclinic phases. The bulk moduli obtained for the tetragonal phases from fitting of the data
collected using LiF as pressure transmitting medium are very large (up to values obtained for
stishovite (Ross et al. 1990)). However, if only experiments using He as pressure transmitting
medium are compared, it appears that Na has little effect on the bulk modulus value of the
tetragonal aluminosilicate hollandite. Monoclinic hollandites are more compressible, and are
stable up to the highest pressures reached during the experiments, suggesting that they may be
possible host minerals for Na and K in transition zone and even down to the Earth’s lower
mantle, if continental crust is subducted to those depths. The high temperature and high
pressure behaviour of K Na AlSi O hollandite indicates qualitatively a positive Clapeyron 0.5 0.5 3 8
slope for the tetragonal to monoclinic high-pressure transition. A simple extrapolation of a
linear fit through the transition pressures of KAlSi O and K Na AlSi O hollandite 3 8 0.8 0.2 3 8
indicates that the phase transition pressure of NaAlSi O hollandite may be at ~7.5 GPa at 3 8
room temperature.

The lattice strains associated with the tetragonal I4/m to monoclinic I2/m transition have been
determined. The phase transition is proper ferroelastic with negligible volume strain. The
a − b a
symmetry breaking strains e − e = and e = cosγ are proportional to the order 1 2 6
a a0 0
parameter Q associated with the transition and their squared values vary linearly with pressure
indicating that the transition is second-order in character. The variation with pressure of the
symmetry breaking strains is similar in K Na AlSi O and KAlSi O hollandites, suggesting 0.8 0.2 3 8 3 8
that Na substitution mainly affects the transition pressure but not the transition mechanism.

Results from the high pressure experiments show that the tetragonal to monoclinic phase
transition is very sensitive to deviatoric stresses present during the experiments due to the
Summary _ 4
pressure transmitting medium. Three different compounds: helium, argon and lithium fluoride
were used as pressure transmitting media in this study. Helium seemed to provide ideal
hydrostaticity and the tetragonal to monoclinic transition was quantitatively described
following a second order Landau excess free-energy expansion. Argon has been reported to
maintain quasi-hydrostaticity up to 9 GPa, and in high-pressure Raman experiments conducted
on KAlSi O hollandite end-member a change in slope of the vibrational modes was observed 3 8
at about 15 GPa (~5 GPa lower than expected). Experiments using LiF have more stressed
environments. The observed transition pressures (i.e. observation of splitting of diffraction
lines indicating the transition to the monoclinic structure) in LiF are much lower than those
observed in hydrostatic condition and are not consistent among different experiments. A stress
term, hQ, added to the Landau excess free-energy expansion shows that, even for “mild”
stresses created in a soft medium such as LiF, the monoclinic structure is stable at any
pressure, preventing a quantitative characterization of the transition. These results might also
give an indication of the possible effects arising from stresses on the mineral transitions in the
Earth’s mantle.













Un pour Un
Permettre à tous d'accéder à la lecture
Pour chaque accès à la bibliothèque, YouScribe donne un accès à une personne dans le besoin