The o igins o oli ine ab ic ansi ions
and hei e ec s on seismic aniso opy in
he uppe man le
Disse a ion
Zu E langung des G ades eines
Dok o s de Na u wissenscha en
-D . Re . Na .-
Bay eu he G aduie enschule ü Ma hema ik und Na u wissenscha en
de Uni e si ä Bay eu h
„Expe imen al Geosciences“
o geleg on
Sushan Shekha
In .
M.Sc. (Explo a ion Geophysics)
Aus
Kha agpu (
Indien
)
May 2011
P ü ungssausschuß:
P o . F. Langensho s , Uni e si ä Jena
(1. Gu ach e )
P o . Da id Rubie, Uni e si ä Bay eu h
(2. Gu ach e )
P o . L. Dub o insky, Uni e si ä Bay eu h
D . Hen i Samuel, Uni e si ä Bay eu h
P o . Ju gen Senke , Uni e si ä Bay eu h
P o . Ludwig Zölle , Uni e si ä Bay eu h
P o . S. Pei e , Uni e si ä Bay eu h
Table o Con en s
Acknowledgemen s ............................................................................................................................................................................. I
Abs ac .................................................................................................................................................................................................. II
Zusammen assung ....................................................................................................................................................................... V
Lis o Images ........................................................................................................................................................................................X
Lis o Tables .................................................................................................................................................................................... XVI
1 In oduc ion .................................................................................................................................................................... 1
1.1 Seismic aniso opy in he ea h ...................................................................................................................... 3
1.2 CPO in Oli ine ........................................................................................................................................................ 6
1.3 CPO ela ionship wi h mic os uc u e ...................................................................................................... 10
1.4 CPO, man le low and aniso opy ................................................................................................................ 12
1.5 Mine al desc ip ion - Oli ine ......................................................................................................................... 15
1.5.1 C ys al-chemis y ..................................................................................................................................... 15
1.6 Aim o he hesis ................................................................................................................................................. 17
2 Me hodology ................................................................................................................................................................. 19
2.1 De o ma ion expe imen s unde ex eme condi ions ........................................................................ 19
2.1.1 High p essu e de o ma ion appa a us ............................................................................................ 20
2.2 Sample P epa a ion ........................................................................................................................................... 36
2.2.1 Ho p essing San Ca los oli ine .......................................................................................................... 36
2.2.2 Placing pla inum Shea s ain ma ke ............................................................................................. 36
2.3 Analy ical Me hods ............................................................................................................................................ 37
2.3.1 Measu emen o c ys allog aphic p e e ed o ien a ion using Elec on backsca e
di ac ion echnique (EBSD) ................................................................................................................................. 38
2.3.2 S udy o disloca ion s uc u e using T ansmission elec on mic oscope (TEM)........... 44
2.3.3 FTIR ................................................................................................................................................................ 49
2.3.4 Piezoelec ic measu emen s o s ess in he Mul ian il appa a us ..................................... 53
3 Resul s ............................................................................................................................................................................. 58
3.1 De o ma ion expe imen s on San-Ca los oli ine using he D-DIA ................................................. 58
3.1.1 Simple shea de o ma ion expe imen s on d y San Ca los oli ine ...................................... 58
3.2 Cha ac e iza ion o he s a ing ma e ial ................................................................................................. 59
3.3 Measu emen o sample s ain ..................................................................................................................... 61
3.4 SEM and EBSD cha ac e iza ion ................................................................................................................... 67
3.4.1 LPO de e mina ions o d y San Ca los oli ine samples ............................................................ 67
3.4.2 Es ima ion o he mean g ain size ..................................................................................................... 74
3.4.3 TEM cha ac e iza ion .............................................................................................................................. 77
3.4.4 Measu emen o sample s ess ........................................................................................................... 79
3.5 Expe imen s unde we condi ion .............................................................................................................. 83
3.5.1 Measu emen o wa e con en using FTIR ................................................................................... 85
3.5.2 NMR spec oscopy on hyd ous Fo s e i e ..................................................................................... 88
3.5.3 Gene al mic os uc u es ........................................................................................................................ 92
3.5.4 SEM and EBSD cha ac e iza ion ......................................................................................................... 95
3.5.5 TEM cha ac e iza ion ........................................................................................................................... 100
3.6 De o ma ion expe imen on Pe ido i e modal composi ion ......................................................... 103
In-si u measu emen o s ess using piezoelec ic senso ........................................................................... 104
4 Discussion ................................................................................................................................................................... 107
4.1 E ec s o s ess and p essu e on he slip sys ems in oli ine: E idence om de o ma ion
expe imen s on “d y” oli ine .................................................................................................................................... 109
4.2 Fab ic ypes unde wa e ich condi ions ............................................................................................. 116
4.3 Physical basis o slip sys em changes in oli ine ............................................................................... 117
4.3.1 Dominance o (010)[001] slip sys em a highe s esses ..................................................... 118
4.3.2 Highe (100)[001] ac i i y a highe wa e con en ............................................................... 125
4.4 Viscoplas ic sel consis en modelling o ab ic de elopmen in oli ine .................................. 130
4.4.1 Modelling he pole ab ic o d y specimen DD455 ................................................................ 131
4.4.2 Modelling he pole ab ic o d y specimen DD456 ................................................................ 134
Seismic aniso opy in he uppe man le – Implica ions om his s udy ............................................... 136
Seismic aniso opy in he uppe man le ......................................................................................................... 137
Oli ine LPO ansi ions and changes in seismic aniso opy wi h dep h ........................................... 138
5 Conclusion ................................................................................................................................................................... 143
Re e ences .............................................................................................................................................................................. 145
Appendix ................................................................................................................................................................................. 152
E klä ung ............................................................................................................................................................................... 165
[X]
LIST OF IMAGES
FIGURE 1-1: PYROLITIC MANTLE MINERALOGY AS A FUNCTION OF MINERAL VOLUME FRACTION AND DEPTH VARIATION. (FIGURE COURTESY: DAN FROST)
.................................................................................................................................................................................................................................................................... 2
FIGURE 1-2: A WAVE TRAVELLING THROUGH A ELASTICALLY ANISOTROPIC MEDIA SPLITS INTO TWO ORTHOGONALLY POLARIZED WAVE. MAGNITUDE OF
THE SHEAR WAVE SPLITTING IS GIVEN BY THE TIME DELAY (ΔT) BETWEEN THE FAST WAVE AND THE SLOW WAVE. FIGURE SOURCE: - ED GARNERO
- HTTP://GARNERO.ASU.EDU/RESEARCH_IMAGES ........................................................................................................................................................................... 4
FIGURE 1-3: PHYSICAL AND CHEMICAL STRUCTURE AND RADIAL SEISMIC ANISOTROPY OBSERVED IN THE EARTH. SOURCE OF SIGNIFICANT ANISOTROPY
IN THE UPPER MANTLE IS BELIEVED TO BE THE CRYSTALLOGRAPHIC PREFERRED ORIENTATION OF MANTLE MINERAL, MAINLY OLIVINE
(COURTESY: D. MAINPRICE). ................................................................................................................................................................................................................ 5
FIGURE 1-4: CRYSTALLOGRAPHIC PREFERRED ORIENTATION DEVELOPMENT IN OLIVINE DUE TO SHEARING NATURE OF THE MANTLE FLOW. CPO OF
ELASTICALLY ANISOTROPIC MINERALS IS THE PRINCIPAL CAUSE FOR SEISMIC ANISOTROPY OBSERVED IN THE UPPER MANTLE. ................................... 7
FIGURE 1-5: DOMINANT SLIP SYSTEMS IN OLIVINE AS A FUNCTION OF STRAIN RATE AND TEMPERATURE (AT P = 1.5 GPA) (FROM CARTER & AV´E
LALLEMANT 1970). RESULTS SHOWN HERE SUGGEST STRESS-INDUCED TRANSITIONS IN THE DOMINANT SLIP SYSTEMS. (B) A COMPARISON OF
CREEP STRENGTH FOR DIFFERENT ORIENTATIONS OF SINGLE CRYSTAL AND POLYCRYSTAL AT ˙Ε ≈ 10−5 S−1 (FROM GOETZE 1978). THE [110]C
ACTIVATES THE [100] (010) SLIP SYSTEM, THE [101]C ORIENTATION, THE [100] (001) AND [001] (100) SLIP SYSTEMS, AND THE [011]C AND
[001] (010) SLIP SYSTEMS. FIGURE SOURCE: KARATO (2008) .................................................................................................................................................. 8
FIGURE 1-6: DOMINANT SLIP SYSTEM IN OLIVINE AS A FUNCTION OF STRESS AND WATER CONTENT [T = 1400 TO 1570K]. AS EVIDENT FROM THE PLOT,
HIGHER CONTENT OF WATER PROMOTES (100)[001] SLIP WHEREAS AT HIGHER STRESS PROMOTES (010)[001] SLIP SYSTEM. FIGURE SOURCE:
JUNG & KARATO (2001) ....................................................................................................................................................................................................................... 9
FIGURE 1-7: LIKELY DISTRIBUTION OF OLIVINE FABRICS IN THE UPPER MANTLE AS A RESPONSE TO CHANGING STRESS, TEMPERATURE AND WATER
CONTENT OF THE PARTS OF UPPER MANTLE (FIGURE SOURCE: KARATO 2008). .................................................................................................................... 12
FIGURE 1-8: IDEALIZED FORSTERITE STRUCTURE PROJECTED ON (100) PLANE (REDRAWN FROM DEER ET AL., 1997). SI ATOMS ARE AT THE CENTRE OF
THE TETRAHEDRONS. SMALL BLACK CIRCLE, SI; LARGER GRAY CIRCLE, OXYGEN; BLACK CIRCLE, M1; DIAGONALLY HATCHED CIRCLE, M2 ................... 16
FIGURE 1-9: FORSTERITE STRUCTURE PERPENDICULAR TO (100) SHOWING THE APPROXIMATELY HEXAGONAL CLOSE-PACKING STRUCTURE (REDRAWN
FROM DEER ET AL., 1997) .................................................................................................................................................................................................................. 16
FIGURE 2-1: SCHEMATIC DIAGRAMS OF ORIGINAL DIA AND DEFORMATION-DIA. A) ORIGINAL DIA CONSISTS OF UPPER AND LOWER GUIDE BLOCKS,
FOUR WEDGE SHAPE SIDE WEDGES AND SIX TUNGSTEN CARBIDE ANVILS. B) A DEFORMATION-DIA HAS TWO ADDITIONAL HYDRAULIC ACTUATORS
CALLED DEFORMATION RAMS WHICH PROVIDES A MEAN TO ACHIEVE CONTROLLED DEFORMATION. (SOURCE: Y. WANG) ........................................... 22
FIGURE 2-2: A VERTICAL CROSS-SECTION OF D-DIA SHOWING THE TWO SIDE WEDGES AND THE DIFFERENTIAL RAM. PRESENCE OF DIFFERENTIAL RAMS
PROVIDES CONTROLLED DEFORMATION OF THE CUBIC SAMPLE AT A CONSTANT PRESSURE. ................................................................................................ 22
FIGURE 2-3: PRESSURE-TEMPERATURE-STRAIN PROFILE OF TYPICAL EXPERIMENTAL RUN IN D-DIA PRESS. AFTER COMPRESSING THE PRESSURE CELL
TO THE REQUISITE PRESSURE, SAMPLE IS HEATED UP TO THE DESIRED TEMPERATURE AND IT IS ALLOWED TO HEAT FOR AT LEAST 30 MIN TO
RELEASE THE INITIAL STRESS BUILD-UP, IF ANY PRESENT IN THE SAMPLE. THEN, THE SAMPLE IS DEFORMED AT A CONSTANT STRAIN RATE. ONCE
THE TARGET AMOUNT OF STRAIN IS ACHIEVED, DEFORMATION IS STOPPED AND THE SAMPLE IS QUENCHED RIGHT AFTER THAT. THEREAFTER THE
PRESSURE IS RELEASED SLOWLY. ....................................................................................................................................................................................................... 23
FIGURE 2-4: CARTOON COMPARING DEFORMATION BY PURE SHEAR AND SIMPLE SHEAR .................................................................................................................. 26
FIGURE 2-5: ORIENTATION CONTRAST IMAGE OF AN EXCESSIVELY DEFORMED SAMPLE AT 8GPA (SAMPLE NO. DD407). ALUMINA PISTONS HAVE FAILED,
OWING TO THE LARGE SHEARING STRESS ACTIVE ON THE WEDGE SHAPED ALUMINA PISTONS. THE APPLIED SHEAR STRAIN WAS MORE THAN
200%. ..................................................................................................................................................................................................................................................... 28
FIGURE 2-6: SCHEMATIC DIAGRAM OF AN 8/6 MM D-DIA ASSEMBLY. ................................................................................................................................................... 29
FIGURE 2-7 : A SCHEMATIC DIAGRAM OF A 4/6 MM D-DIA ASSEMBLY SHOWING ITS MAJOR COMPONENTS ................................................................................. 30
FIGURE 2-8: SCHEMATIC OF ASSEMBLY USED FOR PRESSURE CALIBRATION USING BISMUTH AND MANGANIN ............................................................................. 31
FIGURE 2-9: CALIBRATED CELL PRESSURE HAS BEEN PLOTTED AS A FUNCTION OF OIL PRESSURE. ROOM TEMPERATURE CALIBRATION HAS BEEN DONE BY
USING PHASE TRANSITIONS IN BISMUTH AND MANGANIN RESISTIVITY METHOD. HIGH TEMPERATURE (1000°C) PRESSURE CALIBRATION WAS
DONE USING PHASE TRANSITION IN QUARTZ (QUARTZCOESITE AND COESITESTISHOVITE). 700 BAR OIL PRESSURE IS EQUIVALENT TO 500
TONNE LOAD FOR D-DIA PRESS AT BGI. .......................................................................................................................................................................................... 33
FIGURE 2-10: MEASURED TEMPERATURE ALONG THE SAMPLE LENGTH. CENTER OF THE SAMPLE RECORDED THE HIGHEST TEMPERATURE WITH APPROX.
85°C/MM TEMPERATURE GRADIENT AS WE MOVE TOWARDS THE EXTREMITIES. .................................................................................................................. 34
FIGURE 2-11: EMPLACEMENT OF PLATINUM STRAIN MARKER FOR SHEAR STRAIN MEASUREMENT. APPROXIMATELY 100 NM THICK PLATINUM-LAYER IS
SPUTTER COATED ON THE SIDES OF THE TWO CUT HALVES OF THE HOTPRESSED SAMPLE. ROTATION OF THE STRAIN MARKER IS DIRECTLY
RELATED TO THE SHEAR STRAIN ........................................................................................................................................................................................................ 37
FIGURE 2-12: FORMATION OF BACKSCATTERED KIKUCHI PATTERNS BY EBSD IN THE SEM. (A) ORIGIN OF KIKUCHI LINES FROM THE EBSD (I.E., TILTED
SPECIMEN) PERSPECTIVE. (B) EBSD PATTERN FROM OLIVINE (ACCELERATING VOLTAGE 20 KV)..................................................................................... 39
FIGURE 2-13: SCHEMATIC SETUP OF AN EBSD SYSTEM SHOWING ITS PRINCIPAL COMPONENTS ..................................................................................................... 40
FIGURE 2-14: DIAGRAM ILLUSTRATING THE EVALUATION OF AN EBSD PATTERN. SAMPLE COORDINATE HAS BEEN REPRESENTED BY SUPERSCRIPT “S”
WHERE AS SCREEN COORDINATE HAS BEEN REPRESENTED BY SUPERSCRIPT “SCREEN”. SPECIMEN-TO-SCREEN DISTANCE IS . (X*I,Y*I)
ARE COORDINATES OF THE CENTER OF THE PATTERN AND (XPC,YPC ) ARE THE COORDINATES OF THE CENTER OF THE SCREEN. ................................ 42
FIGURE 2-15 : SCHEMATIC ILLUSTRATION OF DIFFRACTION AROUND A DISLOCATION CORE. IN THIS CASE THE ELECTRON BEAM IS DIFFRACTED MORE
STRONGLY TILED LATTICE PLANES TO ONE SIDE OF THE DISLOCATION CORE THAN IN THE UNDISTORTED PARTS OF THE CRYSTAL. THE
TRANSMITTED BEAM IS DEPLETED AROUND THE DISLOCATION LINE, AND IN A BRIGHT-FIELD IMAGE THE DISLOCATION LINE WILL APPEAR DARKER

[XI]
THAN THE REST OF THE CRYSTAL. ON THE OTHER HAND, THE DIFFRACTED INTENSITY IS GREATER AROUND THE DISLOCATION LINE AND IN A DARK
FIELD IMAGE USING THE DIFFRACTED BEAM; THE DISLOCATION LINE WILL APPEAR LIGHTER THAN THE REST OF THE CRYSTAL. UNDER WEAK
BEAM (WBDF) CONDITION, THE OVERALL INTENSITY OF THE IMAGE IS REDUCED IN COMPARISON TO A DARK FIELD IMAGE. ..................................... 45
FIGURE 2-16: ILLUSTRATION OF EDGE AND SCREW DISLOCATIONS IN A HYPOTHETICAL CRYSTAL. BURGERS VECTOR “B”, THE LATTICE VECTOR THAT
CLOSES THE CIRCUIT AROUND THE DISLOCATION CORE AND DISLOCATION LINE HAS BEEN REPRESENTED BY “DL”, A). IN CASE OF EDGE
DISLOCATION, BURGERS VECTOR IS NORMAL TO THE DISLOCATION LINE. IN THIS CASE, SLIP PLANE IS DEFINED AS THE PLANE CONTAINING THE
DISLOCATION LINE AND THE BURGERS VECTOR, B). IN CASE OF SCREW DISLOCATION, THE DISLOCATION LINE AND BURGERS VECTOR ARE
PARALLEL. ............................................................................................................................................................................................................................................... 49
FIGURE 2-17: DETAILS OF THE FTIR MICROSCOPE (REDRAWN FROM BOLFAN-CASANOVA, 2000) .............................................................................................. 50
FIGURE 2-18: PIEZOELECTRIC CRYSTAL CONFIGURATIONS SHOWING DIFFERENT ORIENTATIONS OF THE APPLIED FORCE WITH RESPECT TO THE CHARGE
POLARIZATION. ...................................................................................................................................................................................................................................... 54
FIGURE 2-19: A SIMPLIFIED CIRCUIT DIAGRAM OF THE CHARGE AMPLIFIER PRODUCED BY COMBINING AN OPERATIONAL AMPLIFIER WITH AN RC
NETWORK. ............................................................................................................................................................................................................................................... 55
FIGURE 2-20: FINAL ASSEMBLY DESIGN FOR PIEZOELECTRIC EFFECT MEASUREMENTS AT HIGH PRESSURE IN THE D-DIA AND 6-AXIS MULTIANVIL
PRESSES. CUBE IS 8 MM IN EDGE LENGTH AND IS COMPRESSED USING 6 MM EDGE LENGTH TRUNCATIONS. THE CRYSTAL IS 1.2 MM IN DIAMETER
AND 0.4 MM THICK AND IS COATED WITH AU USING VAPOUR DEPOSITION. .............................................................................................................................. 57
FIGURE 3-1: TOP: SPECIMEN ID B206 – POLE FIGURE FOR POLYCRYSTALLINE OLIVINE SAMPLE HOT-PRESSED AT 1 𝑮𝑷𝒂 AND 1200°C USING PISTON
CYLINDER PRESS (TALC-PYREX ASSEMBLY). WE FOUND NO LPO IN THIS SPECIMEN. BOTTOM: SPECIMEN ID H3115 – POLE FIGURE FOR OLIVINE
SAMPLE HOT PRESSED AT 8.5 𝑮𝑷𝒂 AND 1200°C USING AN 8-6 MULTI-ANVIL APPARATUS. THIS HOT PRESSED SPECIMEN EXHIBITS A WEAK LPO
RESULTING FROM THE ACTIVITY OF THE (𝟎𝟏𝟎)[𝟏𝟎𝟎] SLIP SYSTEM. ........................................................................................................................................ 60
FIGURE 3-2: PLATINUM SHEAR MARKER IN THE SAMPLE DD402 IS SHOWN. A). SIDEWISE DISPLACEMENT (140 µM) OF THE ALUMINA PISTONS CAN BE
SEEN. B) FAINTLY VISIBLE PLATINUM MARKER IS SHOWN FOR THE SAME ASSEMBLY. C). A CLOSE-UP LOOK AT THE PLATINUM AND ITS AVERAGE
ROTATION DUE TO SAMPLE SHEAR (16.4°); NOTE THAT THE ROTATION OF THE MARKER IS MORE PRONOUNCED NEAR THE PISTON. ...................... 62
FIGURE 3-3: ROTATION OF PLATINUM STRAIN MARKER Θ AND AMOUNT OF SHEAR ∆L FOR A STRAIN MARKER INITIALLY ORIENTED AT 45° TO THE BASE
OF THE SPECIMEN. DOTTED PARALLELOGRAM DEPICTS THE INITIAL ORIENTATION OF A HYPOTHETICAL PLANAR ELEMENT OF THICKNESS “T” THAT
UNDERGOES SHEARING DUE TO THE SIDEWISE MOVEMENT OF THE ALUMINA PISTONS. SOLID LINES INDICATE THE NEW ROTATED POSITION OF THE
SAME ELEMENT AFTER THE SHEAR STRAIN OF Γ. ............................................................................................................................................................................ 63
FIGURE 3-4: VARIATION IN STRAIN EXPERIENCED BY THE SAMPLE DD402 ALONG ITS THICKNESS. TOP-LEFT: THE PARTS CLOSER TO THE ALUMINA
PISTON ARE STRAINED MORE THAN THOSE ARE CLOSE TO THE NEUTRAL LINE N´N. LOCAL ORIENTATION OF THE PT STRAIN MARKER IS SHOWN
USING A SOLID WHITE LINE WHEREAS ORIGINAL ORIENTATION OF PT-MARKER IS SHOWN USING A DOTTED RED LINE. SENSE OF SHEAR IS AS
INDICATED BY THE TWO RED ARROWS ON THE TOP AND BOTTOM. TOP-RIGHT: LOCAL INCREASE IN THE SHEAR STRAIN IN THE SAMPLE NEAR
ALUMINA PISTON HAS BEEN MARKED BY A CURLY BRACKET. BOTTOM (LEFT AND RIGHT): THESE IMAGES SHOW THE DIFFERENCE IN THE ROTATION
ANGLE AS WE MOVE AWAY FROM THE NEUTRAL LINE TOWARDS THE ALUMINA PISTON. ....................................................................................................... 64
FIGURE 3-5: REACTION OF OLIVINE WITH ALUMINA FORMS A LAYER OF SPINEL AND GARNET AT THEIR INTERFACE. THIS MAY ENHANCE THE COUPLING
BETWEEN THE PISTON AND THE SPECIMEN MATERIAL (OLIVINE) .............................................................................................................................................. 65
FIGURE 3-6: DRY SAMPLES DEFORMED AT 3 GPA AND 1300°C. SAMPLE DEFORMED AT LOWER STRAIN RATE (TOP) SHOWS DOMINANT SLIP SYSTEM TO
BE (𝟎𝟏𝟎)[𝟏𝟎𝟎]. OLIVINE A-AXES ARE PREFERENTIALLY ALIGNED SUB-PARALLEL TO THE SHEAR DIRECTION WHEREAS B-AXES ARE ALIGNED
SUBNORMAL TO THE SLIP PLANE. (BOTTOM) SAMPLE DEFORMED UNDER HIGHER STRAIN RATE ALSO SHOW THE PRESENCE OF (𝟎𝟏𝟎)[𝟏𝟎𝟎] SLIP
SYSTEM ALONG WITH (𝟎𝟏𝟎)[𝟎𝟎𝟏] SLIP SYSTEM. ......................................................................................................................................................................... 67
FIGURE 3-7 : DRY SAMPLES DEFORMED AT 5 GPA AND 1300°C. SAMPLE DEFORMED AT LOWER STRAIN RATE (TOP) SHOWS DOMINANT SLIP SYSTEM TO
BE (𝟎𝟏𝟎)[𝟏𝟎𝟎]. OLIVINE A-AXES ARE PREFERENTIALLY ALIGNED SUB-PARALLEL TO THE SHEAR DIRECTION WHEREAS B-AXES ARE ALIGNED
SUBNORMAL TO THE SLIP PLANE. (BOTTOM) SAMPLE DEFORMED UNDER HIGHER STRAIN RATE ALSO HAS BOTH (𝟎𝟏𝟎)[𝟏𝟎𝟎] SLIP AND
(𝟎𝟏𝟎)[𝟎𝟎𝟏] SLIP SYSTEM ACTIVE. 20° GAUSSIAN SMOOTHING WAS APPLIED TO THE POLE FIGURE OF SPECIMEN DD350. ...................................... 68
FIGURE 3-8: DRY SAMPLES DEFORMED AT 5 GPA AND 1400°C. SAMPLE DEFORMED AT LOWER STRAIN RATE (TOP) SHOWS HAS AN LPO RESULTANT OF
SIGNIFICANT STRAIN CONTRIBUTION FROM BOTH (𝟎𝟏𝟎)[𝟏𝟎𝟎] AND 𝟎𝟏𝟎𝟎𝟎𝟏 SLIP SYSTEM. (BOTTOM) SAMPLE DEFORMED UNDER HIGHER
STRAIN RATE HAS (𝟎𝟏𝟎)[𝟎𝟎𝟏] SLIP SYSTEM DOMINANT. .......................................................................................................................................................... 69
FIGURE 3-9: DRY SAMPLES DEFORMED AT 8.5 GPA AND 1300°C. SAMPLE DEFORMED AT SLOWER STRAIN RATE (BOTTOM) SHOWS DOMINANT SLIP
SYSTEM TO BE (𝟎𝟏𝟎)[𝟏𝟎𝟎] AND (𝟎𝟏𝟎)[𝟎𝟎𝟏]. OLIVINE A-AXES AND C-AXES ARE PREFERENTIALLY ALIGNED SUB-PARALLEL TO THE SHEAR
DIRECTION WHEREAS B-AXES ARE ALIGNED SUBNORMAL TO THE SLIP PLANE. (BOTTOM) SAMPLE DEFORMED UNDER HIGHER STRAIN RATE SHOW
THE PRESENCE OF (𝟎𝟏𝟎)[𝟎𝟎𝟏] SLIP SYSTEM. 20° GAUSSIAN SMOOTHING WAS APPLIED TO THE POLE FIGURE OF DD335. ..................................... 70
FIGURE 3-10: SUBSETS OF POLE FIGURES INDICATED A PARTICULAR CRYSTALLOGRAPHIC AXIS PARALLEL TO A SELECTED SPECIMEN AXIS. [100] || X0
IMPLIES THAT THE SUBSET CONTAINS ONLY THE DATA POINTS SUCH THAT OLIVINE [100] AXES ARE ALIGNED PARALLEL (OR SUB-PARALLEL) TO X-
AXIS OF THE SPECIMEN. ........................................................................................................................................................................................................................ 71
FIGURE 3-11: DRY SAMPLES DEFORMED AT 8GPA AND 1500°C. NO RECOGNISABLE LPO IS PRESENT IN THESE SAMPLES. .................................................... 73
FIGURE 3-12 : TOP: EBSD MAP FOR SPECIMEN DD350 WITH NON-INDEXED DATA POINTS; GRAINS HAVE BEEN ASSIGNED COLOUR ACCORDING TO THEIR
EULER ANGLES 1 TO 3; BOTTOM: EBSD MAP AFTER GRAIN RECONSTRUCTION USING THE NEAREST NEIGHBOUR NOISE-REDUCTION METHOD...... 74
FIGURE 3-13: TOP -TEM MICROGRAPH FOR SPECIMEN D384 THAT WAS DEFORMED AT 8.5 GPA AND 1300°C. C-DISLOCATIONS (ONLY EDGE SEGMENTS
ARE VISIBLE) ARE VISIBLE IN THE TOP-LEFT IMAGE (𝒈= [𝟎𝟎𝟒]). NO A-DISLOCATION COULD BE SEEN FROM 𝒈= [𝟏𝟏𝟎] IMAGING DIRECTION,
WHICH IMPLIES THAT C-SLIP WAS THE DOMINANT SLIP SYSTEM. BOTTOM – TEM MICROGRAPHS FOR SPECIMEN DD391 DEFORMED AT 8.5 GPA
AND 1500°C. BOTH A- AND C-DISLOCATIONS CAN BE SEEN IN THE LEFT IMAGE. RIGHT IMAGE SHOWS ONLY THE C-DISLOCATION FOR THE SAME
SPECIMEN. WHITE DOUBLE-ARROWS IN THE PICTURE INDICATE THAT SENSE OF SHEAR FOR THE BULK SAMPLE............................................................ 78
FIGURE 3-14: TEM MICROGRAPHS FOR THE DRY SAMPLE DD455 DEFORMED SLOWLY AT 1300°C. ACTIVE SLIP SYSTEMS ARE (010)[100], (100)[001]
AND (010)[001]. FIGURE ON THE LEFT SIDE SHOWS LARGE NUMBER OF B = [100] DISLOCATIONS PRESENT IN ONE GRAIN. WHITE DOUBLE-
ARROWS IN THE PICTURE INDICATE THAT SENSE OF SHEAR FOR THE BULK SAMPLE. ............................................................................................................. 79
[XII]
FIGURE 3-15: DISLOCATION DENSITY VERSUS STRESS RELATIONSHIP [JUNG AND KARATO, 2001A]. THE SOLID LINE IS THE STRESS VERSUS DISLOCATION
DENSITY RELATIONSHIP FOR A SINGLE CRYSTAL WITH THE SCHMIDT FACTOR = 0.5 [KOHLSTEDT ET AL., 1976B]. ....................................................... 80
FIGURE 3-16: LEFT-DURING ARGON MILLING PROCESS, ARGON STREAM BOMBARDS THE SAMPLE FROM TOP AND BOTTOM (ONLY TOP STREAM IS SHOWN
IN THE FIGURE). THIS GIVES THE MILLED GRAIN SHAPE OF A WEDGE (MARKED BY THE PRESENCE OF THICKNESS FRINGES) WITH HALF-ANGLE
BEING EQUAL TO THE ANGLE OF INCIDENCE OF ARGON STREAM (~5°). APPROXIMATE THICKNESS OF THE PLATEAU OF THE GRAIN CAN BE
CALCULATED FROM THIS SIMPLE MODEL. RIGHT-WEDGE SHAPED PART AND PLATEAU TOP (REGION ENCLOSED BY WHITE RECTANGLE) OF ARGON-
MILLED OLIVINE GRAIN FOR SPECIMEN DD384 IS SHOWN HERE. THIS SAMPLE WAS DEFORMED AT 8.5 GPA AND 1300°CNOTE THAN BASE (B) OF
THE WEDGE PART IS APPROXIMATELY 2 µM. .................................................................................................................................................................................... 81
FIGURE 3-15 SHOWS THIS RELATIONSHIP FOR DRY AND WET SPECIMENS [JUNG AND KARATO, 2001B]. FIGURE 3-17: STRESS VERSUS RECRYSTALLIZED
GRAIN-SIZE RELATIONSHIP FROM JUNG AND KARATO 2001. STRESS MAGNITUDES IN THE SAMPLES FROM THIS STUDY WERE ESTIMATED FROM
DISLOCATION DENSITIES. THE SOLID LINES INDICATE THE RESULTS OF THE LEAST SQUARE FIT FOR THE ‘DRY’ AND ‘WET’ CONDITION. THE SIZE OF
RECRYSTALLIZED OLIVINE DEFORMED UNDER ‘WET’ CONDITIONS IS SIGNIFICANTLY LARGER THAN THAT UNDER ‘DRY’ CONDITIONS AT THE SAME
STRESS. .................................................................................................................................................................................................................................................... 82
FIGURE 3-18: BACKGROUND CORRECTION OF THE RAW FTIR DATA. A BASELINE WAS CREATED USING PIECEWISE CUBIC INTERPOLATION METHOD.
WATER SOLUBILITY VALUES ARE SENSITIVE TO THE CHOICE OF THE BASELINE AND RANGE OF WAVENUMBER USED FOR INTEGRATION (2950 TO
3780 CM-1 IN OUR CASE). .................................................................................................................................................................................................................... 85
FIGURE 3-19: FTIR SPECTRA OF HYDROUS OLIVINE SPECIMEN AFTER THE EXPERIMENTS. ABSORBANCE OF THE SPECTRA WAS NORMALIZED FOR 1 CM
THICK SPECIMEN. ................................................................................................................................................................................................................................... 86
FIGURE 3-20: WATER SOLUBILITY IN SAN CARLOS OLIVINE (MODIFIED AFTER KEPPLER AND BOLFAN-CASANOVA [2006]). OUR RESULTS ARE SHOWN
ALONG WITH THE EXPERIMENTAL DATA FROM MOSENFELDER ET AL. (2006; BLUE DIAMOND) AND KOHLSTEDT ET AL. (1996; RED SQUARES).
THE H2O CONTENTS FROM THIS STUDY EMPLOY THE PATERSON CALIBRATION SO AS TO COMPARE THEM DIRECTLY WITH THE WORK OF
KOHLSTEDT ET AL., 1996 WHERE OLIVINE WAS SATURATED WITH EXCESS H2O.THIS COMPARISON INDICATES THAT THE OLIVINE FROM THIS
STUDY HAD H2O CONTENTS LESS THAN THE SATURATION LEVEL (25-35%). THE STUDY OF MOSENFELDER ET AL (2006) REPORTED HIGHER
H2O CONTENTS MAINLY BECAUSE OF USING THE NEWER BELL ET AL CALIBRATION. ............................................................................................................ 87
FIGURE 3-21: NMR SPECTRA FOR GYPSUM AND SYNTHETIC FORSTERITE SAMPLE. GYPSUM WITH ITS KNOWN WATER CONTENT HAS BEEN USED AS THE
STANDARD. SYNTHETIC FORSTERITE WAS SYNTHESIZED BY ADDING SMALL AMOUNT OF EQUIMOLAR MIXTURE OF BRUCITE AND SILICA WITH
FORSTERITE (SEE TABLE 3-2) AT 11 GPA AND 1150°C USING 8-6 TYPE MA APPARATUS. SAMPLE Z769 WAS ADDED WITH 4 TIMES MORE
BRUCITE-SILICA MIXTURE THAN Z771. CHEMICAL SIFT FOR PEAKS ARE MARKED BY THE ARROW. THE VALUE IN THE PARENTHESIS FOR Δ 6.73 IS
THE CORRESPONDING VALUE OF THE ORDINATE. ........................................................................................................................................................................... 89
FIGURE 3-22: LEFT: RELATION BETWEEN O-H STRETCHING FREQUENCY AND D (O...O) [LIBOWITZKY, 1999]. OPEN SYMBOLS REPRESENT STRAIGHT H
BONDS, SHADED SYMBOLS MARK BENT H BONDS, AND FILLED ONES DENOTE COPPER COMPOUNDS; CIRCLES - SILICATES, SQUARES -
(OXY)HYDROXIDES, HEXAGONS - CARBONATES, DIAMONDS - SULFATES, TRIANGLES - PHOSPHATES AND ARSENATES. RIGHT: ISOTROPIC CHEMICAL
SHIFTS VERSUS O – H...O DISTANCE FOR VARIOUS CRYSTALLINE COMPOUNDS (ECKERT ET AL. 1988)............................................................................. 90
FIGURE 3-23: FTIR SPECTRA OF THE SPECIMEN Z769 CONTAINING FE-FREE SYNTHETIC FORSTERITE WITH TWO WT. PERCENTAGE BRUCITE-SIO2
EQUIMOLAR MIXTURE. WATER CONTENT IN THIS SAMPLE WAS MEASURED USING FTIR IS 559 WT. PPM USING CALIBRATION BY PATERSON
(1982) .................................................................................................................................................................................................................................................... 91
FIGURE 3-24: SEM ORIENTATION CONTRAST IMAGES OF THE SPECIMENS DEFORMED UNDER WET CONDITION. IN GENERAL, THE AVERAGE GRAIN SIZES IN
THE HYDROUS SPECIMENS ARE SMALLER IN COMPARISON TO THEIR DRY COUNTERPARTS IN TERMS OF P-T CONDITIONS. GRAINS IN WET
SPECIMENS HAVE SERRATED BOUNDARIES. ...................................................................................................................................................................................... 93
FIGURE 3-25: WET SAMPLES DEFORMED AT 3 GPA AND 1300°C. SAMPLE DEFORMED AT LOWER STRAIN RATE (TOP) SHOWS TWO ACTIVE SLIP SYSTEMS
– (010)[100] AND (100)[001]. (BOTTOM) SAMPLE DEFORMED UNDER HIGHER STRAIN SHOWS ONLY (100)[001] SLIP SYSTEM TO BE ACTIVE.
.................................................................................................................................................................................................................................................................. 96
FIGURE 3-26: WET SAMPLES DEFORMED AT 5 GPA AND 1300°C. BOTH THE HIGH STRAIN RATE AND LOW STRAIN RATE SAMPLE EXHIBIT ONLY ONE
ACTIVE SLIP SYSTEM – (100)[001]. ................................................................................................................................................................................................. 96
FIGURE 3-27: WET SAMPLES DEFORMED AT 5 GPA AND 1400°C. SAMPLE DEFORMED AT LOWER STRAIN RATE (TOP) SHOWS MAINLY ONE ACTIVE SLIP
SYSTEMS – (100)[001]. WHEREAS, (BOTTOM) SAMPLE DEFORMED UNDER HIGHER STRAIN HAS TWO (010)[001] AND (100)[001] SLIP
SYSTEMS ACTIVE. ................................................................................................................................................................................................................................... 97
FIGURE 3-28: WET SAMPLES DEFORMED AT 8.5 GPA AND 1300°C. IRRESPECTIVE OF THE STRAIN RATE, BOTH THE SPECIMENS DEFORMED AT 8.5 GPA
AND 1300°C SHOW TWO ACTIVE SLIP SYSTEMS –(010)[100] AND (100)[001]. THIS OBSERVATION IS CONSISTENT WITH ACTIVITY OF
(010)[001] SLIP SYSTEM AT RELATIVELY HIGHER STRESSES AND (100)[001] SLIP SYSTEM UNDER HYDROUS CONDITION. ..................................... 98
FIGURE 3-29: WET SAMPLES DEFORMED AT 5 GPA AND 1500°C. SAMPLE DEFORMED AT LOWER STRAIN RATE (TOP) SHOWS TWO ACTIVE SLIP SYSTEMS
– (010)[100] AND (100)[001]. (BOTTOM) SAMPLE DEFORMED UNDER HIGHER STRAIN SHOWS ONLY (100)[001] SLIP SYSTEM TO BE ACTIVE.
.................................................................................................................................................................................................................................................................. 99
FIGURE 3-30: TEM MICROGRAPHS FOR THE WET SPECIMEN DD456. DEFORMATION EXPERIMENT WAS CARRIED OUT AT 8.5 GPA AND 1300°C WITH A
STRAIN RATE OF 5X10-4. TOP FIGURE SHOWS THE PRESENCE OF C-DISLOCATIONS. (100)[001] DISLOCATIONS ARE MOSTLY OF EDGE NATURE
WHERE AS THE [001] SCREW DISLOCATIONS ARE MOST LIKELY FROM (010)[001] DISLOCATION. EVIDENCE OF CROSS-SLIP CAN BE ALSO SEEN AS
INDICATED BY MARKER 1 IN TOP IMAGE AND WHITE ARROW IN THE BOTTOM-LEFT IMAGE. BOTTOM-RIGHT FIGURE ALSO SHOWS STRAIGHT C-
SCREW DISLOCATION FROM (010)[001] SLIP SYSTEM. ............................................................................................................................................................. 101
FIGURE 3-31: A TYPICAL HRTEM IMAGE (UPPER AND LOWER RIGHT) AND THE FAST FOURIER TRANSFORMED IMAGE (LOWER RIGHT) OF THE
DISSOCIATED C-EDGE DISLOCATION VIEWING ALONG THE {110} ZONE AXIS OF A DEFORMED HYDROUS OLIVINE. THE IMAGE CONTRAST IN THE
DISLOCATION CORE REGIONS IS DIFFERENT FROM THAT IN THE SURROUNDING BULK, WHICH INDICATES THAT THE CORE IS EXPANDED. .............. 102
FIGURE 3-32: PERIDOTITE SAMPLES DEFORMED AT 8.5GPA AND 1300°C. OLIVINE IN THE SLOWLY DEFORMED AGGREGATE LIKELY HAS BOTH
(010)[100] AND (010)[001] SLIP SYSTEMS ACTIVE WHEREAS IN THE EXPERIMENT CONDUCTED AT HIGHER STRAIN RATE THE SLIP SYSTEM IS
(010)[001]. PYROXENE IN BOTH THE CASES SHOW (100)[001] SLIP SYSTEM .................................................................................................................. 103
[XIII]
FIGURE 3-33: OUTPUT VOLTAGE FROM THE CHARGE AMPLIFIER AS A FUNCTION OF TIME FOR AN EXPERIMENT WHERE A GAPO4 CRYSTAL WAS
COMPRESSED TO 2 GPA AND THEN HELD AT CONSTANT STATIC PRESSURE FOR 80 MIN. ................................................................................................... 104
FIGURE 3-34: OUTPUT VOLTAGE AS A FUNCTION OF TIME FOR AN EXPERIMENT HELD STATICALLY AT 2 GPA AND THEN DEFORMED BY DRIVING OUT THE
ANVILS IN THE HORIZONTAL DIRECTION SIMULTANEOUSLY AFTER APPROXIMATELY 60 S BY 20 MICRONS. THE DRIFT BEFORE 60 S IS LINEAR AND
IS REMOVED BY SUBTRACTING A LINEAR BACKGROUND AS SHOWN IN B. ............................................................................................................................... 105
FIGURE 3-3-35: STRESS AND ANVIL DISPLACEMENT VERSUS TIME FOR 4 DEFORMATION EVENTS PERFORMED AT 2 GPA. .................................................... 106
FIGURE 4-1: SUMMARY OF FABRICS OBSERVED IN SAN-CARLOS OLIVINE DEFORMED UNDER DRY AND WET CONDITION AT DIFFERENT STRAIN RATES.
EXPERIMENTS WERE PERFORMED BETWEEN 3 TO 8.5 GPA AND 1300°C TO 1500°C. WIDTH OF EACH COLOUR BAR IS PROPORTIONAL TO THE
APPROXIMATE NUMBER OF GRAINS THAT WERE PRESENT IN THE SUBSET CONTAINING DATA POINTS FOR THAT SLIP SYSTEM. REFER TO SECTION
3.4.14 FOR MORE DETAILS. AS SHOWN IN THE TABLE AT TOP-RIGHT CORNER OF THE PAGE, THE LOWER ROW IN THE 2X2 MATRIX CONTAINS
RESULTS FROM DRY EXPERIMENTS WHILE UPPER ROW CONTAINS RESULTS FROM WET EXPERIMENTS. THE LEFT COLUMN IN 2X2 MATRICES HAS
RESULTS FROM SLOWLY DEFORMED SAMPLES WHEREAS SAMPLES DEFORMED AT RELATIVELY HIGHER STRAIN RATE HAVE THEIR FABRICS SHOWN
IN THE RIGHT COLUMN. ..................................................................................................................................................................................................................... 111
FIGURE 4-2: POLE FIGURES FOR TWO POLYCRYSTALLINE OLIVINE SPECIMEN HOTPRESSED AT 8.5 GPA (H3115) AND 11 GPA (H3354). SPECIMENS
WERE ANNEALED AT 1400°C. BOTH POLE FIGURES RESEMBLE A-TYPE FABRIC WHICH IS OFTEN OBSERVED UNDER LOW STRESS AND DRY
DEFORMATION ENVIRONMENT. PRESENCE OF A-TYPE FABRIC IN THESE HOTPRESSED SPECIMEN IS INDICATIVE OF (010)[100] SLIP SYSTEM
ACTIVITY. .............................................................................................................................................................................................................................................. 113
FIGURE 4-3: DEFORMATION DATA FROM THIS STUDY AND OTHER STUDIES ARE SHOWN AS A FUNCTION OF STRESS AND WATER CONTENTS (T ∼ 1470–
1670 K). LARGER SYMBOLS WITH BLACK BOUNDARIES REPRESENT DATA FROM THIS STUDY WHEREAS REST OF DATA ARE FROM KATAYAMA ET
AL. 2004. EXCEPT, ONE OF THE DATA FOR D-TYPE FABRIC IS FROM BYSTRICKY ET AL. (2001). WATER CONTENT WAS ESTIMATED USING THE
PATERSON (1982) CALIBRATION. BROKEN GRAY LINES INDICATE THE LIKELY TRANSITION LINE BETWEEN TWO DIFFERENT FABRIC TYPES
(MODIFIED AFTER KARATO ET AL., 2008) ................................................................................................................................................................................... 115
FIGURE 4-4: CRITICAL RESOLVED SHEAR STRESSES (CRSS) OF THE (010)[100] AND (010)[001] SLIP SYSTEMS AS A FUNCTION OF TEMPERATURE.
DATA (CORRESPONDING TO A STRAIN RATE OF 10-5 S-1) FROM EXPERIMENTS PERFORMED ON SINGLE CRYSTALS ORIENTED ALONG [011]C
(BLACK-FILLED SYMBOLS) TO PROMOTE [001](010) GLIDE AND ALONG [110]C (OPEN SYMBOLS) TO PROMOTE [100](010) GLIDE. (SOURCE-
PHD THESIS – HELEN COUVY, 2005) ............................................................................................................................................................................................ 118
FIGURE 4-5: TEMPERATURE DEPENDENCE OF THE CRITICAL SHEAR STRESS ΤC(T) OF COVALENT CRYSTALS MEASURED UNDER HIGH OR ATMOSPHERIC
PRESSURE. THE DATA ARE TAKEN FROM THE REFERENCES: LAGERLOF ET AL. (1994) FOR Α-AL2O3, CASTAING ET AL. (1981B) FOR SI AND
BOIVIN ET AL. (1990) FOR GAAS OF INTRINSIC AND P-TYPE. (FIGURE SOURCE: KOIZUMI ET AL., 1994) ..................................................................... 119
FIGURE 4-6: LEFT IMAGE SHOWS A KINK (DARK LINE) LYING ACROSS A POTENTIAL VALLEY. BROKEN LINES INDICATE THE POTENTIAL MAXIMA WITH
MINIMA REPRESENTED BY THE SOLID LINES. RIGHT: A KINK IN THE PRESENCE OF EXTERNAL STRESS HAS ITS EQUILIBRIUM POSITION DISPLACED
AWAY FROM THE UNSTRESSED POSITION. SIZE OF THE KINK IS REPRESENTED BY KINK HEIGHT H AND WIDTH 2K+L IN CASE OF A TRAPEZOIDAL
KINK MODEL. (SOURCE: SUZUKI ET AL. 1995) ............................................................................................................................................................................. 120
FIGURE 4-7: ENTHALPY CHANGE ASSOCIATE WITH THE CONTRIBUTION FROM THERMAL PERTURBATION AT TEMPERATURE T AND MECHANICAL WORK
DONE BY STRESS Σ. ............................................................................................................................................................................................................................. 121
FIGURE 4-8: THE FUNCTION G(X) GIVING THE SHAPE OF THE PEIERLS POTENTIAL. IT IS SINUSOIDAL FOR A = 0, DAM-LIKE WITH A ROOF TOP FOR A=0.5,
AND CAMEL-HUMP SHAPED, WITH AN INTERMEDIATE MINIMUM FOR A = 0.8. (SOURCE: KOIZUMI 1994) ................................................................... 122
FIGURE 4-9: PREDICTED CRSS VALUES FOR A-SLIP AND C-SLIP USING DOUBLE KINK NUCLEATION THEORY. AT AROUND 1300°C, 300 MPA STRESS
WOULD BARELY ACTIVATE C-SLIP WHEREAS AT AROUND 600 MPA, BOTH C-SLIP AND A-SLIP ARE ACTIVE. IN THIS CASE, ACTIVITY OF (010)[001]
SLIP SYSTEM WOULD BE HIGHER BECAUSE THIS SLIP SYSTEM HAS EXTRA THERMAL ENERGY AVAILABLE AT ITS DISPOSAL. ....................................... 124
FIGURE 4-10: (001) PROJECTION OF THE OLIVINE STRUCTURE. ONLY THE OXYGEN IONS ARE SHOWN, BUT THE POSITIONS OF THE SILICON IONS ARE
INDICATED BY THE SI 04 TETRAHEDRA. PERIODIC JOGS IN A (100) PLANE ARE INDICATED BY THE BROKEN LINE. THE ATOM POSITIONS ARE THOSE
OF THE PAPER BY HANKE (1965). (FIGURE SOURCE: OLSEN AND BIRKELAND, 1973) ..................................................................................................... 127
FIGURE 4-11: FTIR SPECTRA FOR HYDROUS SAMPLES SHOW PEAKS AT 3477, 3448, 3629 AND 3676 CM-1. THESE PEAKS COULD BE ARISING FROM
HYDROGEN ASSOCIATED WITH VACANT SILICON SITES. .............................................................................................................................................................. 128
FIGURE 4-12: DRY SAMPLES DEFORMED AT 8.5 GPA AND 1300°C. SAMPLE DEFORMED AT SLOWER STRAIN RATE SHOWS DOMINANT SLIP SYSTEM TO
BE (𝟎𝟏𝟎)[𝟏𝟎𝟎] AND (𝟎𝟏𝟎)[𝟎𝟎𝟏]. ............................................................................................................................................................................................. 131
FIGURE 4-13: POLE FIGURES FOR MODELS DESCRIBED IN THE TABLE 5-3. MODELS WHICH ASSUME VERY SIMILAR CRSS VALUE FOR (010)[100] AND
(010)[001] AND AT LEAST THREE TIMES HIGHER CRSS VALUE FOR OTHER TWO SLIP SYSTEMS CAN MIMIC THE EXPERIMENTAL POLE FIGURE. 133
FIGURE 4-14: NORMALIZED ACTIVITY VERSUS EQUIVALENT STRAIN PLOT FOR VARIOUS MODEL. MODEL 1 TO 6 IS SHOWN HERE. ACTIVITY OF SLIP
SYSTEMS CAN CHANGE WITH INCREASING STRAIN BECAUSE OF GEOMETRICAL CONSTRAINTS. IN THIS SENSE, MODEL 4 AND 5 APPEAR VERY STABLE
............................................................................................................................................................................................................................................................... 134
FIGURE 4-15: WET SAMPLES DEFORMED AT 8.5 GPA AND 1300°C. THE SPECIMENS SHOWS TWO LIKELY ACTIVE SLIP SYSTEMS – (010)[100] AND
(100)[001] WHICH HAS ALSO BEEN CONFIRMED BY TEM STUDY ON THIS SAMPLE........................................................................................................... 135
FIGURE 4-16: POLE FIGURES FOR MODELS DESCRIBED IN THE TABLE 5-4. MODELS WHICH ASSUME (100)[001] TO BE THE EASIEST AND (010)[100] AS
SLIGHTLY HIGHER THAN THE FORMER ALONG WITH VERY HIGH VALUE OF CRSS FOR (010)[100] AND (001)[100] I.E. FOR A-SLIP CAN
REPRODUCE WELL THE POLE FIGURE FOR THE SPECIMEN DD456. ......................................................................................................................................... 135
FIGURE 4-17: LEFT: SHEAR WAVE ANISOTROPY IN THE UPPER MANTLE AS A FUNCTION OF DEPTH. RIGHT: P-WAVE ANISOTROPY AS A FUNCTION OF
DEPTH (SOURCE: PHD THESIS – HELEN COUVY, 2005). .......................................................................................................................................................... 138
FIGURE 4-18: VARIATION OF WATER CONTENT OF MAJOR MINERAL PHASES IN THE UPPER MANTLE. CHANGES IN THE WATER CONTENT ARE RESULT OF
VARIATION IN THE PORTIONING COEFFICIENT OF WATER FOR VARIOUS PHASES WITH DEPTH. ......................................................................................... 140
FIGURE 4-19: VARIATION IN OLIVINE FABRIC WITH CHANGES IN WATER CONTENT AS A FUNCTION OF DEPTH. PRESENCE OF C-TYPE FABRIC CAN EXPLAIN
THE NATURE OF THE SEISMIC ANISOTROPY IN THE LOWER PARTS OF THE UPPER MANTLE. NUMBERS IN THE PARENTHESIS ARE THE VSH/VSV
RATIOS (FROM KARATO ET AL., 2008) THAT ARE OBSERVED IN NATURAL OLIVINE SPECIMENS EXHIBITING CORRESPONDING FABRIC TYPES. ..... 141
LIST OF TABLES
TABLE 1-1: FABRIC TYPE AND NATURE OF SLIP SYSTEM (JUNG AND KARATO, 2001) .............................................................................................. 9
TABLE 1-2: SUMMARY OF SLIP SYSTEM (DURHAM AND GOETZE, 1977) ................................................................................................................. 11
TABLE 1-3: SHEAR WAVE SPLITTING (DIRECTION OF THE POLARIZATION OF THE FASTER, VERTICALLY TRAVELING SHEAR WAVES) (FROM
KARATO, 2008) .................................................................................................................................................................................................. 13
TABLE 1-4: VSH/VSV ANISOTROPY (FROM KARATO, 2008) .................................................................................................................................... 13
TABLE 1-5: LATTICE CONSTANTS AND DENSITIES OF OLIVINES (DEER ET AL., 1997) ........................................................................................... 16
TABLE 2-1: LIST OF DEFORMATION DEVICES AND PROPERTIES (MODIFIED AFTER KARATO 2008) .................................................................... 21
TABLE 2-2: LIST OF EXPERIMENTS AND THE END PRODUCTS - CALIBRATION OF CELL PRESSURE AT 1000°C USING PHASE TRANSITION IN
QUARTZ ................................................................................................................................................................................................................ 32
TABLE 2-3 : OPTICS SETTINGS FOR DIFFERENT FREQUENCY RANGES USED TO ANALYZE WATER SPECIES ............................................................. 50
TABLE 3-1: EXPERIMENTAL CONDITIONS AND RESULTS OF DRY SAN CARLOS OLIVINE EXPERIMENTS .................................................................. 59
TABLE 3-2: MEASUREMENT OF STRESS USING RECRYSTALLIZED GRAIN SIZE ........................................................................................................... 76
TABLE 3-3: LIST OF EXPERIMENTS AND EXPERIMENTAL CONDITIONS UNDER WET CONDITION ............................................................................. 83
TABLE 3-4: STARTING MATERIAL FOR DEFORMATION EXPERIMENTS ON HYDROUS OLIVINE .................................................................................. 84
TABLE 3-5: DESCRIPTION OF THE STARTING MATERIAL AND WATER CONTENT FROM 1H MAS NMR AND FTIR MEASUREMENTS ................ 89
TABLE 3-6: O – H...O DISTANCE FOR DIFFERENT STRETCHING FREQUENCIES PRESENT IN THE FTIR SPECTRA OF THE HYDROUS FORSTERITE
(Z769) AND OLIVINE SAMPLE USING RELATION CORRELATION PROPOSED BY LIBOWITZKY (1999). CHEMICAL SHIFT VALUES
OBSERVED USING 1H MAS NMR AND CORRESPONDING O – H...O DISTANCE IN THE HYDROUS FORSTERITE SAMPLE (ECKERT,1988)
HAS BEEN SHOWN IN THE BOTTOM TWO ROWS. ............................................................................................................................................... 90
TABLE 3-7: DEGREE OF RECRYSTALLIZATION AND RECRYSTALLIZED GRAIN SIZE FOR WET SPECIMENS ................................................................ 93
TABLE 3-8: EXPERIMENTAL CONDITIONS FOR PERIDOTITE DEFORMATION EXPERIMENTS AND LIKELY ACTIVE SLIP SYSTEMS ........................ 103
TABLE 4-1: FABRIC TYPE AND NATURE OF SLIP SYSTEMS (JUNG AND KARATO, 2001) ........................................................................................ 110
TABLE 4-2: VALUE OF CONSTANTS THAT DESCRIBE WELL THE CRSS-TEMPERATURE RELATION FOR THE TWO SLIP SYSTEMS IN OLIVINE ... 123
TABLE 4-3: CHOICE OF RELATIVE CRSS VALUES USED FOR VARIOUS MODELS IN ORDER TO SYNTHETICALLY GENERATE THE POLE FIGURE FOR
SPECIMEN DD455 ............................................................................................................................................................................................ 131
TABLE 4-4: CHOICE OF RELATIVE CRSS VALUES USED FOR VARIOUS MODELS IN ORDER TO SYNTHETICALLY GENERATE THE POLE FIGURE FOR
SPECIMEN DD456 ............................................................................................................................................................................................ 135
TABLE 4-5: VSH /VSV ANISOTROPY FOR VARIOUS OLIVINE FABRICS AS A FUNCTION OF MANTLE FLOW DIRECTION (FROM KARATO, 2008)
............................................................................................................................................................................................................................. 137
[VII]
wi d a gumen ie , dass de Rückgang de seismischen Aniso opie, de in den obe s en
300 km des obe en Man els beobach e wi d, nich du ch eine d uckinduzie e Ände ung
des dominan en Glei sys ems in Oli in e u sach wi d. S a dessen wi d o geschlagen,
dass Ände ungen im H2O-Gehal des Oli ins mi de Tie e eine Ve schiebung de Tex u
on Typ A nach Typ C bewi ken, mi einem möglichen Umweg übe die Tex u des Typs E
(cha ak e isie du ch das Glei sys em (001) [100]), die in diese S udie jedoch nich
beobach e wu de. Modellie ungen wu den du chge üh , um zu zeigen, dass diese
Ände ung im dominan en Glei sys em die Ve inge ung de seismischen Aniso opie mi
de Tie e im obe en E dman el e u sachen kann. De H2O-Gehal on Oli in s eig on
un e 100 ppm bei 50 km Tie e au 250 ppm bei 300 Kilome e n Tie e an. Alle dings wi d
nich a gumen ie , dass de gesam e H2O-Gehal des Man els mi de Tie e ans eig ,
sonde n, dass diese Ände ung im H2O-Gehal des Oli ins du ch die Ände ung des
Ve eilungskoe izien en on H2O zwischen Oli in und Py oxen mi de Tie e bei einem
kons an en H2O Konzen a ion on 200 ppm im obe en E dman el e u sach we den
kann.
Ähnliche Ve o mungsexpe imen e wu den mi eine P obe pe ido i ische
Zusammense zung bei 8.5 GPa und 1300 ° C du chge üh und e gaben iden ische
Oli in ex u en wie die monomine alischen Expe imen e. Tex u en ü Diopsid und Ens a i
in diesen Expe imen en sind ähnlich mi denen, die zu o in Expe imen en bei nied ige em
D uck p oduzie wu den.
Expe imen e mi einem piezoelek ischen Eink is all aus GaPO4 wu den in de D-DIA und
de 6-S empel MAVO P esse bei hohen D ücken du chge üh , um die elek ische Ladungen
zu messen, die du ch die de ia o ischen Spannungen e zeug we den. Die elek ische
Ladung des K is alls wu de mi hil e eines Ope a ions e s ä ke s gemessen. In
Expe imen en bei Raum empe a u wu den un e Ve wendung de kubischen Zelle
e olg eich quan i izie ba e elek ische Ladungen gemessen, die bei absolu en
Bewegungen de De o ma ionss empel on wenige als 0,5 µm e zeug wu den. Obwohl

[VIII]
die piezoelek ische Kons an e ü GaPO4 bei hohen D ücken noch nich kalib ie is ,
konn en aus den gemessenen elek ischen Ladungen mechanische Spannungen im Be eich
on 4-350 MPa abgeschä z we den.
[1]
1 In oduc ion
Only ocks o he ea h’s ou e c us a e di ec ly accessible o analysis bu his su ace
ese oi accoun s o less han 1% o Ea h's o al olume. Clues o he chemical and
physical s a e o he Ea h’s unde lying man le can be ob ained h ough he s udy o
xenoli hs i.e. ocks b ough o Ea h's su ace in basal lows o by mo e exo ic magmas
such as diamond-bea ing kimbe li e pipes. In addi ion la ge sec ions o he oceanic
li hosphe ic man le can become ec onically a ached o he con inen al c us du ing
moun ain building episodes and me eo i es also p o ide some clues abou he likely
chemical composi ion o he bulk Ea h and i s me allic co e. The deep in e io o he Ea h
emains di ec ly inaccessible, howe e , and i can only be s udied indi ec ly, using ools
p o ided by geophysics – such as analysis o seismic wa es and he measu emen o
g a i y, hea low, and magne ism.
Seismology has been he mos impo an ool o he de e mina ion o he Ea h’s deep
s uc u e and likely chemical composi ion. Seismic body wa e da a has been employed o
p oduce one dimensional global model o S and P wa e eloci ies in he Ea h [Dziewonski
and Ande son, 1981]. These one-dimensional p o iles ha e been compa ed wi h mine al
seismic eloci ies de e mined as a unc ion o p essu e, empe a u e and composi ion in
he labo a o y in o de o de e mine he likely mine alogy and composi ion o he man le
as a unc ion o dep h [F os , 2008; S ix ude and Li hgow-Be elloni, 2005]. Such
compa isons a e gene ally consis en wi h an ul ama ic uppe man le composed o oli ine,
o ho- and clinopy oxene and ga ne o dep hs o app oxima ely 410 km. The eloci ies o
he man le a g ea e dep hs a e consis en wi h a bulk man le o simila composi ion o he
uppe man le, al hough unde going phase ans o ma ions o dense mine al polymo phs
wi h inc easing dep h. Seismic discon inui ies a 410, 520 and 660 km, whe e wa es a e
e lec ed and con e ed a sha p bounda ies in mine al elas ic p ope ies, a e consis en
wi h expe imen al s udies ha show phase ans o ma ions o oli ine o highe p essu e
[2]
s uc u es a p essu e co esponding o hese dep hs. Seismic obse a ions a e he e o e
consis en wi h he uppe man le being comp ised dominan ly o oli ine, ∼60 olume %
(Fig 1-1), which is also in ag eemen wi h he majo i y o man le xenoli hs which show he
uppe man le o be oli ine domina ed.
Figu e 1-1: Py oli ic man le mine alogy as a unc ion o mine al olume ac ion and dep h
a ia ion. (Figu e cou esy: Dan F os )
Veloci ies o seismic body wa es in he Ea h inc ease wi h dep h as a esul o bo h he
e ec o p essu e on mine al elas ic p ope ies and phase ans o ma ions o dense
mine al s uc u es. While he o me e ec causes a g adual inc ease in eloci y wi h dep h
he la e in oduces discon inuous changes in eloci y. Howe e , seismic wa e eloci ies
[3]
also change as a unc ion o he di ec ion o p opaga ion in some egions o he man le. This
aniso opic beha iou is a phenomenon o g ea signi icance because i esul s om he
de elopmen o ab ic in man le ocks caused by con ec i e low d i en by hea low in he
ea h.
Con ec ion in he ea h’s man le has a di ec bea ing on he mo ion o ec onic pla es,
which in u n go e ns he e olu ion o he mos p ominen geological ea u es on he
ea h’s su ace, such as moun ains, oceans, olcanoes e c. The s udy o seismic aniso opy
can he e o e be used o unde s and he di ec ion o man le con ec ion, o in es iga e he
coupling be ween he li hosphe e and he as henosphe e and o delinea e s uc u es in he
in e io such as con inen al oo s.
1.1 Seismic aniso opy in he ea h
Seismic aniso opy a ises om aniso opic elas ic p ope ies o ocks and mine als and
can esul om wo main p ocesses in he Ea h, bo h ela ed o de o ma ion. Shape
p e e ed o ien a ion (SPO) a ises om laye ing, caused, o example, by mine al banding
o mel channelling, while c ys allog aphic p e e ed o ien a ion (CPO) a ises om he
alignmen o elas ically aniso opic mine als. Seismic aniso opy due o SPO would in e
banding o mine als in o laye s wi h s ongly di e en elas ic p ope ies. Mos man le
mine al, howe e , do no ha e su icien ly di e en elas ic p ope ies o cause s ong
seismic aniso opy h ough SPO [Shea e , 1999]. The excep ion would be banding in ol ing
mel laye s, al hough hese could only occu in he e y op o he man le, po en ially
benea h idges, o possibly in he D’’ laye . Aniso opy in he bulk o he man le is gene ally
a ibu ed o CPO.
Seismic aniso opy is measu ed using wo main echniques:
1. Azimu hal aniso opy is he a ia ion in wa e eloci y, bo h S and P wa es, wi h he
di ec ion o wa e p opaga ion. This was i s de ec ed om he azimu hal
dependence o wa es (a P wa e ha a els along he bounda y be ween he
c us and man le) in he oceanic li hosphe e, benea h he Paci ic Ocean [Hess, 1964].
This ype o aniso opy is measu ed by de e mining he di ec ional dependence o
wa e eloci ies h ough a egion o he ea h. Fo his pu pose a se ies o di e en
[4]
sou ces (e.g. ea hquakes) a e equi ed o send wa es o one o mo e ecei e s so
ha sui able di ec ional co e age o he egion in ques ion is ob ained. The
di e ences in eloci y as a unc ion o he di ec ion ha he wa e a elled h ough
he egion o in e es a e hen analyzed. The main d awbacks in s udying azimu hal
aniso opy a e ha many sou ce- ecei e pai s a e equi ed and he e ogenei ies
can also cause appa en aniso opy because ays a el h ough di e en egions on
hei way o he egion o in e es . Azimu hal aniso opy can also be s udied using
Su ace wa es (Rayleigh and Lo e wa es).
2. Pola iza ion aniso opy is simila o bi e ingence in op ical mine alogy and leads o
shea -wa e spli ing in seismog ams. S-wa e pa icle mo ion is no mal o he
p opaga ion di ec ion and he eloci y is he e o e a unc ion o he ma e ial elas ic
p ope ies in di ec ions no mal o he p opaga ing di ec ion. In an aniso opic
medium S-wa es he e o e become pola ized wi h a as S-wa e di ec ion
o hogonal o a slow one. The eloci y di e ence be ween he pola ized S-wa es
causes shea wa e spli ing, whe e he wo pola ized wa es de elop a delay ime. A
a single ecei e s a ion he pola iza ion di ec ion (φ) o he as shea wa e and he
ime delay (𝛿𝑡) be ween he wo pulses can be measu ed (Fig. 1-2).
Figu e 1-2: A wa e a elling
h ough a elas ically aniso opic
media spli s in o wo
o hogonally pola ized wa e.
Magni ude o he shea wa e
spli ing is gi en by he ime
delay (δ ) be ween he as wa e
and he slow wa e. Figu e
sou ce: - Ed Ga ne o -
h p://ga ne o.asu.edu/ esea ch
_images
In addi ion o hese
di e en measu emen
echniques, seismologis s
o en model aniso opy in he
man le assuming ans e se iso opy, which conside s an elas ic medium o ha e a

[5]
symme y axis no mal o he p opaga ion di ec ion. In mos ins ances he symme y axis is
conside ed o be e ical and his ype o aniso opy is he e o e o en e med adial
aniso opy. Di e ences in p ope ies a e implied in he ho izon al and e ical di ec ions. P
wa es a e hen esol ed in o PH and PV componen s in he ho izon al and e ical
di ec ions and SH and SV a e he co esponding pola ized S-wa es. The i s e e ence
model o seismic s uc u e o ea h, PREM (Dziewonski and Ande son, 1981), includes his
kind o aniso opy in he op 220 km o he Ea h (Fig 1-3). The alue VSH/VSV o VSH2/VSV2
o inequali ies such as VSH>VSV a e equen ly used o quan i y aniso opy in he man le as
shown in igu e 1-3.
Figu e 1-3: Physical and chemical s uc u e and Radial seismic aniso opy obse ed in he ea h.
Sou ce o signi ican aniso opy in he uppe man le is belie ed o be he c ys allog aphic p e e ed
o ien a ion o man le mine al, mainly oli ine (Cou esy: D. Mainp ice).
Signi ican seismic aniso opic is obse ed in he uppe man le and he D’’ laye
(be ween lowe man le and he ou e co e) (Fig 1-3). Majo sou ce o he seismic
aniso opy in he uppe man le is he CPO o majo mine al phases e.g. oli ine and
[6]
py oxene. Ou o hese wo mine al phases, oli ine has much la ge con ibu ion o he
o e all aniso opy in he uppe man le because o i s la ge in insic aniso opy and he
olume. Aniso opy in he c us al egions is gene ally caused by he shaped p e e ed
o ien a ion and p esence o mel s and luids.
1.2 CPO in Oli ine
C eep wi hin he ea h’s con ec ing man le esul s in he non- andom dis ibu ion o
c ys allog aphic o ien a ions o majo man le mine als such as oli ine, py oxene and
ga ne because hese mine als ha e aniso opic mechanical p ope ies (Ka a o, 1989). As
hese mine als, pa icula ly oli ine, also ha e in insic elas ic aniso opy he de elopmen
o CPO makes he man le seismically aniso opic in places. The e o e, seismic aniso opy in
he man le e lec s he s ain ield p e ailing in he pas ( ozen-in aniso opy) wi hin he
li hosphe e o p esen con ec i e p ocesses in he as henosphe e and deepe man le. As
shown in Figu e 1-3, he bulk o man le below 200 km appea s seismically iso opic, which
could be in e p e ed as a esul o di usion c eep o supe plas ic low (Ka a o, 1995),
because such di usi e p ocesses do no lead o CPO de elopmen . CPO esul s om
disloca ion glide which ac s o o a e he o ien a ion o a c ys al o ma ch he imposed
de o ma ion egime. CPO depends on he de o ma ion geome y and he ac i e disloca ion
slip sys ems o he c ys al. A slip sys em is ela ed o he mo ion o he dominan
disloca ions ac i e in he mine al. Na u al oli ine –bea ing ocks show a de o ma ion ab ic
domina ed by alignmen o he [100] axis pa allel wi h he di ec ion o appa en shea
de o ma ion, abb e ia ed as a-slip, while he (010) plane is pa allel o he shea plane (Fig
1-4).
[7]
Figu e 1-4: C ys allog aphic p e e ed o ien a ion de elopmen in oli ine due o shea ing na u e
o he man le low. CPO o elas ically aniso opic mine als is he p incipal cause o seismic
aniso opy obse ed in he uppe man le.
The oli ine ab ic da abase compiled by (Ben Ismaїl and Mainp ice 1998) indica es ha
up o 50% o na u ally de o med samples possess his (010)[100] CPO oli ine ab ic. Apa
om he mos common (010)[100] slip sys em, o he known slip sys ems in oli ine a e
(001)[100], (010)[001] and (100)[001] [Ca e and A e'lalleman , 1970]. In he pas ,
expe imen al s udies ha e been pe o med on na u al and syn he ic samples in o de o
unde s and he e ec o de o ma ion condi ions on he CPO de elopmen in oli ine and
o he mine als (Ca e and A e'Lalleman 1970, Nicolas e al. 1973, Zhang & Ka a o 1995).
Ini ial wo ks on expe imen al de o ma ion on San Ca los oli ine indica ed ha slip
sys ems in oli ine change wi h changing s ess and empe a u e [Ca e and A e'lalleman ,
1970; Goe ze, 1978]. A lowe s esses (o highe empe a u es) (010)[100] slip sys em
was ound o be mos ac i e whe e as highe s esses (o lowe empe a u es) (010)[001]
slip sys em was mos ac i e. Unde in e media e s ess and empe a u e condi ions,
{0kl}[100] slip sys em was he easies (Fig. 1-5). La e wo ks by Ka a o and co-wo ke s
[8]
also p oposed a change in slip sys em a om (010)[100] o (010)[001] slip sys em a
highe s esses [Jung and Ka a o, 2001a]. Change in slip sys em has also been obse ed
wi h change in wa e con en o oli ine [Jung and Ka a o, 2001a; Ka ayama e al., 2004]. A
mode a e wa e con en (less han 60 w . ppm), (001)[100] slip sys em was ound o be
dominan slip sys em while i changed o (100)[001] slip sys em a highe wa e con en s
(Fig. 1-5).
Figu e 1-5: Dominan slip sys ems in oli ine as a unc ion o s ain a e and empe a u e (a P =
1.5 GPa) ( om Ca e & A ´e Lalleman 1970). Resul s shown he e sugges s ess-induced ansi ions
in he dominan slip sys ems. (b) A compa ison o c eep s eng h o di e en o ien a ions o single
c ys al and polyc ys al a ˙ε ≈ 10−5 s−1 ( om Goe ze 1978). The [110]c ac i a es he [100] (010) slip
sys em, he [101]c o ien a ion, he [100] (001) and [001] (100) slip sys ems, and he [011]c and [001]
(010) slip sys ems. Figu e Sou ce: Ka a o (2008)
P e ious wo ks a Baye isches Geoins i u [Cou y e al., 2004] indica ed ha a highe
p essu es (11 GPa), he de o ma ion was mainly caused by slip o disloca ion wi h [001]
bu ge s ec o s, indica ing ha c-slip was easie unde hese condi ions. De o ma ion o
single c ys al oli ine specimens also ea i med his conclusion whe e [001]-slip o e
[100]-slip p og essi ely becomes easie wi h inc easing p essu e [Ra e on e al., 2007].
Theo e ical modelling o disloca ion co e s uc u e by i s p incipal calcula ions indica ed
ha his change in slip sys em wi h p essu e can be explained by change in disloca ion co e
s uc u e wi h p essu e [Du inck e al., 2005].
[15]
1.5 Mine al desc ip ion - Oli ine
Mg- ich oli ine is a common mine al o ma ic and ul ama ic ocks, and is gene ally
conside ed o be he majo cons i uen o he Ea h’s uppe man le (40-80% in olume)
wi h a composi ion close o
(𝑀𝑔0.9𝐹𝑒0.1)2𝑆𝑖𝑂4 based on analyses o oli ine in man le
xenoli hs.
1.5.1 C ys al-chemis y
Oli ine is an o hosilica e [Dee e al., 1997], and a solid solu ion be ween he wo end-
membe s, 𝑓𝑜𝑟𝑠𝑡𝑒𝑟𝑖𝑡𝑒,(𝑀𝑔)2𝑆𝑖𝑂4 (𝐹𝑜100𝐹𝑎0) and 𝑓𝑎𝑦𝑎𝑙𝑖𝑡𝑒,(𝐹𝑒)2𝑆𝑖𝑂4(𝐹𝑜0𝐹𝑎100). Oli ine
c ys al la ice has an o ho hombic symme y (space g oup: Pbnm). The impo an
c ys allog aphic pa ame e s o oli ine a e shown in Table 1-5. The s uc u e consis s o
independen 𝑆𝑖𝑂4 e ahed a linked by di alen ca ions (M1 and M2) in six old co-
o dina ion (Fig 1-8). The oxygen anions a e a anged in shee s nea ly pa allel o he (001)
plane and closely esemble a hexagonal close-packed s uc u e (Fig. 1-9). Each oxygen
a om is bonded o one silicon ca ion and h ee di alen ca ions (e.g. Mg2+, Fe2+) wi h
oc ahed al co-o dina ion. Howe e , since he oxygen a oms a e no pe ec ly close-packed,
he M1 and M2 polyhed a a e i egula shaped, and in such a way ha he M2 si e is
sligh ly la ge han he M1 si e. The adjacen M1 si es sha e edges o o m bands pa allel o
he [001] axis. These bands a e connec ed o he nex M1 bands in he uppe laye (o
lowe laye ) by he M2 oc ahed al si es. The e is appa en ly no comple e o de ing in he
Mg/Fe2+ dis ibu ion be ween he M1 and M2 si es, bu Fe2+ has a p e e ence o he M1
si e. M1 and M2 may also be occupied by o he ca ions such as Ni2+, Mn2+, Ca2+, C 3+ [Dee e
al., 1997] o B3+ [Sykes e al., 1994]. In pa icula , oli ines om Xenoli hs o en con ain a
small amoun o nickel (F ey and P inz, 1976).

[16]
Table 1-5: la ice cons an s and densi ies o Oli ines (Dee e al., 1997)
Fo s e i e
Man le oli ine
Fayali e
Chemical Fo mula
Mg2SiO4
(Mg0.9,Fe0.1)2SiO4
Fe2SiO4
a (Å)
4.754
4.755
4.8211
b Å
10.197
10.21
10.4779
c (Å)
5.9806
5.985
6.0889
Densi y (g/cm
3
)
3.222
3.4
4.392
Figu e 1-8: Idealized o s e i e
s uc u e p ojec ed on (100) plane
(Red awn om Dee e al., 1997). Si
a oms a e a he cen e o he
e ahed ons. Small black ci cle, Si; la ge
g ay ci cle, oxygen; black ci cle, M1;
diagonally ha ched ci cle, M2
Figu e 1-9: Fo s e i e s uc u e
pe pendicula o (100) showing he
app oxima ely hexagonal close-
packing s uc u e ( ed awn om Dee
e al., 1997)
[17]
1.6 Aim o he hesis
P incipal goal o his PhD wo k is o de e mine he o igin o changes in he dominan
slip sys ems a high p essu e and high empe a u e en i onmen p e alen in he uppe
man le. Fo his pu pose, we plan o conduc simple shea de o ma ion expe imen s on San
Ca los oli ine and Pe ido i e modal composi ion, using De o ma ion-DIA. Expe imen s a e
designed o delinea e he in luences o p essu e, empe a u e, s ain a e and H2O con en
on slip sys ems in oli ine.
A he ini ial s ages o his PhD wo k, a new high p essu e assembly will be designed
which should allow us o eliable conduc de o ma ion expe imen s a p essu es abo e 8
GPa unde high empe a u e condi ions.
De o med specimens a e o be analyzed using a a ie y o analy ical ools ha includes
SEM & EBSD, TEM, FTIR and NMR.
1. E ec o s ess and p essu e on slip sys ems in oli ine
A ange o expe imen s has been conduc ed a a ious p essu e and
empe a u es o explo e he e ec o s ess on he slip sys ems in oli ine. In his
way, we also asce ained he ole played by p essu e on changes in oli ine slip
sys ems. P essu e has been a ied be ween 3 o 8.5 GPa a empe a u es be ween
1300 o 1500°C.
De e mina ion o s ess in he sample, o some expe imen s, has been
pe o med Ex-Si u using disloca ion densi ies ob ained by T ansmission elec on
mic oscopy. Al e na i ely, s esses ha e also been measu ed using ec ys allized
g ain size da a ob ained by EBSD echnique and compa ing hem wi h known
ec ys allized g ain size e sus s ess calib a ion. TEM s udies we e also be
employed o de ailed s udy o disloca ion mic os uc u e.
2. E ec o wa e on slip sys ems in oli ine
A simila ange o expe imen s ha e been pe o med on “we ” oli ine specimens
o de e mine he e ec o H2O on slip sys ems in oli ine. Wa e con en o he
hyd ous specimens has been analyzed using FTIR spec oscopy. Some o hese
[18]
specimens ha e also been s udied by TEM o unde s and he ole o wa e on
disloca ions.
3. Oli ine is he mos oluminous phase in he uppe man le ocks. O he impo an
mine al phases in he uppe man le a e Py oxenes, Spinel and Ga ne . De o ma ion
s udies ha e been pe o med on Pe ido i e modal composi ion, which ep esen s
he uppe man le ock composi ion, o unde s and he ole o o he mine al phases
e.g. py oxenes on he o e all ab ic de elopmen in a uppe man le.
4. S ess is he mos c ucial pa ame e in a con olled de o ma ion expe imen . Ye , i
is one o he mos poo ly cons ained. In-si u measu emen o s ess, in mos cases,
is ei he pe o med using ex e nally (ex e nal o he p essu e cell) placed load cell
o using x- ay adiog aphy. The o me me hod o en gi es a measu ed alue o
s ess ha can be a o om he ac ual alue expe ienced by he sample ma e ial.
This di e ence is caused by he ic ional o ces ac i e h oughou he p essu e cell.
O he me hod o s ess measu emen using he x- ay adiog aphy elies upon
synch o on based adia ion sou ce. This app oach is no sui able o day- o-day use
in a s anda d labo a o y en i onmen . The e o e, a emp has been made o de elop
a s ess senso based upon piezoelec ic p ope y o GaPO4 c ys al. Once his
echnique has ully de eloped, i should p o ide us a mean o make in-si u s ess
measu emen s.
[19]
2 Me hodology
2.1 De o ma ion expe imen s unde ex eme condi ions
Many ypes o equipmen ha e been used o de o m mine al assemblages a ambien
and high p essu e and empe a u e condi ions. Fo many yea s he gas media Pa e son ig
and he G iggs appa a us we e he main de ices used o de o m samples bu could only be
employed a condi ions equi alen o he c us and e y shallow man le (see able 1).
Howe e he need o in es iga e heological p ope ies a highe p essu es esul ed in
ini ial a emp s using he 6-8 Kawai- ype mul ian il, DIA and diamond an il cell [Ka a o,
2008a]. These de ices we e no designed o applying con olled de o ma ion a high
p essu e bu hey could expose samples o high de ia o ic s esses as a esul o
aniso opic comp ession. In he diamond cell his occu s qui e no mally when a poo ly
hyd os a ic p essu e medium is employed, while in mul ian il de ices de ia o ic s esses
can be applied by placing ha de ma e ials in he sample column di ec ion compa ed o he
pe pendicula di ec ion. In hese de ices high de ia o ic s esses hen de elop du ing
comp ession. Once high p essu es a e eached he applica ion o empe a u e so ens he
p essu e medium and he s esses elax.
The need o p o ide a be e con ol o e he sample de o ma ion en i onmen ,
howe e , lead o he de elopmen o new high p essu e de o ma ion de ices such as he
Ro a ional D ickame de ice (Ka a o, 2008) and he de o ma ion DIA appa a us. In o de
o s udy he de elopmen o C ys allog aphic p e e ed o ien a ion (CPO) o mine al
assemblages a man le condi ions, de o ma ion expe imen s mus be able o sa is y he
c i e ion gi en below:
1. The me hod mus p o ide he abili y o gene a e and main ain ex eme s a es o
p essu e ( o uppe man le 1-14 GPa) and empe a u e (abo e 1000-1600°C)
condi ions p e alen in he ea h’s man le, o ime du a ions ha allow ab ics o
de elop a sui able s ain a es (1-48 hou s).
[20]
2. Such a me hod should also p o ide con ol o e s ain a e (o s ess) which is essen ial
o any meaning ul unde s anding o he ab ic de elopmen .
3. The abili y o measu e s ain a e and s ess.
4. P o ide con ol o e chemical en i onmen e.g. wa e and oxygen ugaci y is also
equi ed.
5. Allow eco e y o he sample o ab ic and ex u al analysis. And,
6. Ensu e ha he sample does no de o m du ing comp ession and decomp ession phase
o he expe imen .
2.1.1 High p essu e de o ma ion appa a us
Unde s anding he ole o de ia o ic s ess (and se e al o he physical & chemical
pa ame e s) on he de elopmen o ex u e in ma e ial is a p e equisi e o he
unde s anding o aniso opic beha iou o mine als. Ma e ial scien is s also use such
s udies o es ablishing ela ions be ween a ious manu ac u ing p ocesses and he
mechanical pe o mance o he p oduc ma e ial. The immense scope o such s udies has
led o he de elopmen o a se ies o high p essu es appa a uses wi h each ha ing a unique
se o ad an ages and d awbacks.
Simple dead weigh loading expe imen s a e an e ec i e way o pe o m p ecise
de o ma ion expe imen s unde ambien p essu e condi ion [Ca e e al., 1980] whe eas
a he o he ex eme o p essu e diamond-an il cells can be used o de o med ma e ials up
o 200 GPa p essu e bu wi h poo con ol o e s ain a e and s ess dis ibu ion [Kinsland
and Basse , 1977]. A in e media e p essu es, a ange o de ices exis which all end
owa ds a comp omise in e ms o maximum p essu e and he con ol o e de o ma ion
and sample en i onmen . Some salien ea u es o a ious de o ma ion appa a us a e gi en
below in he able 2-1. Fo expe imen s o add ess oli ine ab ic de elopmen p essu es o
be ween 1-15 GPa a e equi ed. Though, he 6-8 mul ian il con igu a ions can achie e
hese p essu es, only s ess elaxa ion expe imen s can be pe o med and no con ol o e
s ain a e is possible [Bussod e al., 1993; Ka a o and Rubie, 1997]. The Ro a ional
D ickame de ice (e.g. [Yamazaki and Ka a o, 2001] can p o ide con ol o s ain a e o e
he p essu es o in e es , howe e , i is e y limi ed in sample size and con ol o e

[21]
p essu e and empe a u e is limi ed accep when used in conjunc ion wi h in si u X- ays.
Consequen ly he de o ma ion-DIA appea s o embody a sui able comp omise be ween
allowing con ol o e s ain a e, p essu e and empe a u e and also achie ing a ange o
p essu e o in e es o he s udy o oli ine ab ic de elopmen in he uppe man le.
Table 2-1: Lis o de o ma ion de ices and p ope ies (Modi ied a e Ka a o 2008)
Type o appa a us
Max.
P (GPa)
Max.
T(K)
In-si u
S ess
measu emen
Commen
Dead-weigh c eep
appa a us
10-4
2000
F om applied
load
Low ƒH
2
O
Gas-medium
appa a us (e.g.
Pa e son ig)
0.5
1600
In e nal load cell
Limi ed ƒH
2
O
G iggs- ype appa a us
3
1600
Ex e nal load
cell
Limi ed s ain
De o ma ion-DIA
23
1700
X- ay di ac ion
Limi ed s ain
Ro a ional D ickame
appa a us
18
2000
X- ay di ac ion
Unlimi ed s ain
6-8 Mul ian il s ess-
elaxa ion
23
2000
X- ay di ac ion
Non-s eady s a e,
elaxa ion expe imen s
only
Diamond an il
200
1000
X- ay di ac ion
Non-s eady s a e, e y
high s ess, S udy o LPO
di icul due o small
sample size
6- am cubic p ess
25
2000
X- ay di ac ion
Limi ed s ain
De o ma ion-DIA
The De o ma ion-DIA (o D-DIA) is a modi ied o m o DIA cubic-an il appa a us [Osugi
e al., 1964]. The D-DIA inco po a es wo addi ional hyd aulic ac ua o s; he e e e ed o as
di e en ial ams, which p o ide independen con ol o he displacemen o wo e ically
opposing an ils [Wang e al., 2003]. The o iginal DIA consis s o uppe and lowe guide
blocks, ou wedge-shaped side wedges, and six ungs en ca bide an ils. The six an ils o
he D-DIA de ine a cubic olume due o hei squa e shaped unca ion. Fou ou o six
an ils a e a ached o he side wedges whe eas one each o he o he wo an ils is moun ed
o he uppe and lowe guide blocks.
[22]
Di e en ial
Ram
Side
Wedge
Displacemen
T ansduce
Main load
F om P ess
Hyd aulic
Oil line
Sample
Guide
Block
An il
Figu e 2-2: A e ical c oss-sec ion o D-DIA showing he wo side wedges and he di e en ial am.
P esence o di e en ial ams p o ides con olled de o ma ion o he cubic sample a a cons an
p essu e.
O iginal DIA
De o ma ion DIA
Figu e 2-1: Schema ic diag ams o o iginal DIA and De o ma ion-DIA. A) O iginal DIA consis s o
uppe and lowe guide blocks, ou wedge shape side wedges and six ungs en ca bide an ils. B) A
de o ma ion-
DIA has wo addi ional hyd aulic ac ua o s called de o ma ion ams which p o ides a
mean o achie e con olled de o ma ion. (sou ce: Y. Wang)
[23]
As he main guide block o he DIA is comp essed he e ically opposing an ils a e
ad anced. The 45° su aces o he side wedges, howe e , ensu e ha a componen o he
e ical displacemen is con e ed o a ho izon al displacemen which ad ances he 4
ho izon al an ils. Once high p essu e has been achie ed h ough he ad ancemen o he
main am, he e ical di e en ial ams can be ad anced hus applying a p inciple s ess on
he cubic sample assembly.
D-DIA’s unique design p o ides a way o ad ance he di e en ial ams in o he sample
assembly wi hou aising he con ining p essu e in he p ocess. A ypical expe imen al un
in he D-DIA is illus a ed below in Figu e 2-2. One s a s by hyd os a ically comp essing
he sample assembly, wi h di e en ial ams ully wi hd awn o main ain he oil p essu e in
he main am a a cons an alue. The ac ion o he guide block ans e s he e ical
comp essional o ce in o ho izon al comp ession ia he side wedges. Then he sample is
hea ed up o he equisi e empe a u e and kep in his s a e o app oxima ely 30 min,
wi h a iew o achie e he mal equilib ium. Then he di e en ial ams a e ad anced so as
o b ing a non-cubic shape change o he assembly while simul aneously wi hd awing he
main am a an app op ia e a e. This ensu es ha o al o ce exe ed by he main am
s ays cons an and hence he olume o he sample assembly is also conse ed (Wang e al.,
2003).
Figu e 2-3: P essu e-
Tempe a u e-S ain p o ile o
ypical expe imen al un in D-
DIA p ess. A e comp essing
he p essu e cell o he
equisi e p essu e, sample is
hea ed up o he desi ed
empe a u e and i is allowed
o hea o a leas 30 min o
elease he ini ial s ess build-
up, i any p esen in he
sample. Then, he sample is
de o med a a cons an s ain a e. Once he a ge amoun o s ain is achie ed, de o ma ion is
s opped and he sample is quenched igh a e ha . The ea e he p essu e is eleased slowly.
[24]
The onse o mo emen o he di e en ial ams occu s only when he di e en ial am
p essu e is su icien enough o o e come he con ining o ce and he ic ion. Hence, he
highe he con ining p essu e, highe is he ini ial di e en ial am p essu e equi ed o
mo e he di e en ial ams. This imposes a limi o he maximum con ining p essu e a
which de o ma ion can be achie ed wi hou b eaking he an ils. Using he 500 onne
Voggen ei e D-DIA p ess a ailable a he Baye isches Geoins i u wi h 4 mm squa e
unca ions ungs en ca bide (WC) an ils, i is possible o pe o m de o ma ion
expe imen s up o 10 GPa con ining p essu e, a empe a u es as high as 1500°C.
The displacemen o he wo di e en ial an ils is measu ed ela i e o he guide block
using displacemen ansduce s. The ansduce s employ a magne ic esponse o measu e
displacemen o a p ecision o 0.2 µm.
Sample assembly design
As s a ed ea lie , in he D-DIA a wide ange o p essu e and empe a u e condi ions can
be achie ed (up o 10 GPa and 1700°C). In o de o access highe p essu es he unca ion
size on he ungs en ca bide an ils can be educed, howe e , his also educes he sample
size as he dimensions o he cubic p essu e assembly mus also be educed. Fo he wo k
pe o med in his hesis wo assembly con igu a ions we e de eloped. The 6/8 assembly
employs 6 mm edge leng h squa e ace ed ungs en ca bide an il unca ions wi h a cubic
sample assembly ha is 8 mm in edge leng h. I was used o achie e p essu es up o 3.5
GPa. The 4/6 assembly, used be ween 3.5 and 10 GPa, employs 4 mm an il unca ions and
a 6 mm edge leng h cube.
The expe imen al se -up o de o ma ion expe imen s consis s o a cubic p essu e
medium composed o i ed o un i ed py ophylli e. The cube is d illed ou along one axis
o he inse ion o a u nace sepa a ed by a he mally insula ing slee e. The esis i e
ma e ials g aphi e and henium a e used as u naces, while zi conia is used as a he mally
insula ing slee e a ound he u nace o make he expe imen al se -up he mally e icien . A
numbe o a emp s we e made o employ lan hanum ch omi e as a u nace ma e ial,
[31]
P essu e and empe a u e calib a ion o he Sample assembly
P essu e calib a ion a oom empe a u e
The sample p essu e mus be calib a ed in each D-DIA assembly agains he oil
p essu e using phase ansi ions ha occu a well-de e mined p essu es. Ini ially oom
empe a u e calib a ions we e employed o iden i y he app oxima e p essu e ange o
u he high empe a u e calib a ions and o de ine he g adien o he p essu e calib a ion.
Bismu h unde goes phase ansi ions a 2.54 GPa (Bi I-II), 2.7 GPA (Bi II-III) and 7.7 GPa
(Bi III-IV) which causes a change in he esis i i y o bismu h [Ge ing, 1998; Lloyd, 1971],
whe eas he esis i i y o Manganin wi e (Cu86/Mn12/Ni2) changes linea ly wi h
p essu e. In o de o calib a e he p essu e achie able wi h he 6 mm py ophylli e
p essu e cell, a hin wi e o Bismu h
was placed
be ween wo AgCl disks as shown in Fig
2-8. Thin coppe oils which se ed he
pu pose o elec ode, we e placed a he
wo ends o he Bismu h wi e. A
cons an cu en was supplied h ough
he bismu h wi e and he ol age ac oss
he Cu-elec odes was measu ed du ing
he comp ession.
Bi I-II ansi ion was obse ed a a ound 17 ba oil p essu e. Fo ob aining he
p essu e- esis i i y ela ionship o Manganin, he same se up was used bu wi h Manganin
wi e eplacing he Bismu h wi e. The change in Manganin
esis i i y wi h he con ining p essu e can be exp essed as he p essu e coe icien o he
esis ance change gi en by �𝛿𝑅𝑅0
���1𝑃
� �=(2.322 ± 0.008)×10−2 𝐺𝑃𝑎−1 whe e 𝛿𝑅 is
he change in esis ance and R is he esis ance a any gi en p essu e
P
measu ed in GPa
(Robe J. Ze o and H. B. Van lee ). The p essu e dependence o esis i i y mus be
Figu e 2-8: Schema ic o assembly used o p essu e
calib a ion using Bismu h and Manganin

[32]
calib a ed using a known p essu e poin , o which we used he Bi I-II ansi ion. As shown
in Fig 2-8 he p essu e a oom empe a u e de e mined using Manganin wi e eaches a
pla eau jus o e 7 GPa, consis en wi h he obse a ion ha we we e unable o obse e he
Bi III-IV ansi ion a 7.7 GPa. This pla eau is ypical o mos mul ian il assemblies and
esul s om he gaske suppo ing an e e -inc easing p opo ion o he load.
Figu e 2-9 shows ha da a om calib a ion expe imen s a oom empe a u e and
1000°C. The assembly is mo e e icien a gene a ing p essu e a 1000°C han a oom
empe a u e as shown by he Coesi e/S isho i e ansi ion, which occu s a app oxima ely
110 ba s oil p essu e (78.6 onnes o applied load).
P essu e calib a ion a high empe a u e
Phase ansi ion in qua z o i s high p essu e polymo phs, coesi e and s isho i e, is
p essu e and empe a u e dependen . A 1000°C, qua z ans o ms o coesi e a ~ 2.95
GPa; whe eas he coesi e o s isho i e ansi ion p essu e a 1000°C is a ~9.25 GPa
(Akaogi & Na o sky, 1984).
A ine g ained mix u e o ib ous qua z was placed in he s anda d assembly (bu in
pu e shea con igu a ion). The sample assembly was b ough up o he equisi e p essu e
and hen i hea ed a 1000°C o ~5 hou s. A e quenching and decomp ession, he end
p oduc was sec ioned and polished and analyzed using Raman spec oscopy o iden i y
he SiO2 polymo ph in he un p oduc s.
Table 2-2: Lis o expe imen s and he end p oduc s - Calib a ion o cell p essu e a 1000°C using
phase ansi ion in Qua z
Expe imen Run
Oil p essu e(ba s)
Obse a ion
DD451
17.1
Qua z
DD452
19.6
Qua z and Coesi e coexis
DD387
78.6
Coesi e
DD400
85.7
S isho i e
[33]
Figu e 2-9: Calib a ed cell p essu e has been plo ed as a unc ion o oil p essu e. Room
empe a u e calib a ion has been done by using phase ansi ions in Bismu h and Manganin
esis i i y me hod. High empe a u e (1000°C) p essu e calib a ion was done using phase ansi ion
in Qua z (qua zcoesi e and coesi es isho i e). 700 ba oil p essu e is equi alen o 500 onne
load o D-DIA p ess a BGI.
Tempe a u e calib a ion o he sample assembly
The sample empe a u e dis ibu ion wi hin he 4/6 de o ma ion cell assembly was
measu ed using a wo-py oxene he mome e assemblage, employing he calib a ion o
Nickel & B ey, (1984). An equimola powde ed mix u e o Al- ee ens a i e and diopside
we e placed be ween he alumina de o ma ion pis ons a 45° o he axial di ec ion. 4 w %
PbO was added o he sample as a lux. The sample powde was compac ed be o e he
second de o ma ion pis on was placed on op, bu he sample was no ho p essed. The
expe imen was hea ed a 1300°C ( he mo-couple empe a u e) and 8 GPa p essu e o
app ox. 6 h s. No de o ma ion was applied. Chemical composi ions o ully eac ed
neighbou ing pai s o ens a i e and diopside g ains we e measu ed using a JEOL JXA – 8200
elec on mic op obe a BGI.
[34]
The eco e ed sample slice was app oxima ely 200 µm hick in he axial di ec ion. The
calib a ion employs he dis ibu ion o Ca and Mg be ween he wo-py oxene mine als. The
de e mined mean empe a u e in he cen al po ion o he sample slice was
app oxima ely 150°C highe han he empe a u e measu ed a he he mocouple. The
di e ence be ween he mean empe a u e a he middle o he sample and a he ex eme
poin s o he sample was ound o be ~75°C (Fig 2-10). This esul s in a empe a u e
g adien o 84° C/mm along he sample leng h. The a ia ion in empe a u e likely e lec s
he he mal g adien along he u nace; wi h he cen e o he sample, he e o e, placed will
in he ho es pa o he u nace, while he he mocouple and sample ex emi ies nex o
he u nace a e sligh ly ou o he ho spo .
Figu e 2-10: Measu ed empe a u e along he sample leng h. Cen e o he sample eco ded he
highes empe a u e wi h app ox. 85°C/mm empe a u e g adien as we mo e owa ds he
ex emi ies.
Oxygen ugaci y o he expe imen s
The oxygen ugaci y is an impo an pa ame e in de o ma ion expe imen s as i
con ols he concen a ion and mobili y o poin de ec s in Fe-bea ing mine als [Demouchy
and Mackwell, 2006; Kohls ed and Mackwell, 1998]. The oxygen ugaci y is no con olled
1300
1350
1400
1450
1500
1550
-0.85 -0.65 -0.45 -0.25 -0.05 0.15 0.35 0.55 0.75
Tempe a u e (°C)
Dis ance om he cen e o he sample (mm)
Tempe a u e (°C) along he sample leng h
S anda d De ia ion = 43°
[35]
in D-DIA expe imen s bu i is in luenced by he ype o u nace employed and by he
na u e o he s a ing ma e ial. The use o g aphi e o me al u naces gene ally leads o
mo e educing condi ions because he u naces can only be oxidized du ing he expe imen
and has no mechanism o elease oxygen o oxidized species. The oxygen ugaci y can be
measu ed in he expe imen howe e pa icula ly in he egion o he P capsule o P
s ain ma ke s whe e he p opo ion o Fe alloying wi h he P can be measu ed.
𝐹𝑒2𝑆𝑖𝑂4 = 𝐹𝑒𝑆𝑖𝑂3 + 𝐹𝑒 + 𝑂2
𝑂𝑙𝑖𝑣𝑖𝑛𝑒 𝑒𝑛𝑠𝑡𝑎𝑡𝑖𝑡𝑒 𝐴𝑙𝑙𝑜𝑦
The equilib ium can be used o de e mine he oxygen ugaci y om he he modynamic
ela ionship
24 3
2
log log log 2log
ln(10)
oli ine o hopy oxene me al
Fe SiO FeSiO Fe
o
G
o RT
aa a
−∆
=+− −
Whe e,
24
oli ine
Fe SiO
a
,
a
me al
Fe
and
3
o hopy oxene
FeSiO
a
a e he ac i i ies o he Fe2SiO4 componen in
oli ine, Fe in he P -Fe alloy and FeSiO3 in py oxene espec i ely. ∆G0 is he ee ene gy o
he end-membe equilib ium, which was aken om he modynamic da a epo ed by
[S agno e al., 2011]S agno and F os (2011). In some de o ma ion expe imen s whe e
py oxene was obse ed and he Fe con en o P oil o s ain ma ke was measu ed i was
possible o calcula e he oxygen ugaci y in he expe imen . Fo his ac i i y composi ion
ela ions o Fe in P alloy we e equi ed, which we e aken om Mann e al. (submi ed).
The de e mined oxygen ugaci y a 8 GPa and 1500°C was -0.8±0.5 log uni s o
app oxima ely a he ayali e-magne i e-qua z (QFM) oxygen bu e . This alue is wi hin
he ange ound o man le ocks, al hough i may be conside ed sligh ly mo e oxidised
han many samples om he deep man le, >100 km, which a e in gene al close o QFM –2
[F os and Mccammon, 2008].
[36]
2.2 Sample P epa a ion
2.2.1 Ho p essing San Ca los oli ine
A ine g ained powde o
(𝑀𝑔0.9,𝐹𝑒0.1)2𝑆𝑖𝑂4 San Ca los oli ine wi h g ain size below
10 µ𝑚 was loaded in o a 12 𝑚𝑚 long and 5 𝑚𝑚 ou e Diame e cylind ical Pla inum
capsule. Using ¾” Talc-Py ex glass assembly and 200 pis on cylinde p ess a Baye isches
Geoins i u , he assembly was cold p essed o 10𝑘𝑏 con ining p essu e. The ea e i was
hea ed o a empe a u e o 1100°𝐶 and kep as such o 30 𝑚𝑖𝑛. Then he assembly was
slowly decomp essed o e 12 ℎ𝑜𝑢𝑟𝑠 pe iod along wi h simul aneously cooling o oom
empe a u e. Slow decomp ession wi h simul aneous cooling was pe o med in o de o
a oid decomp ession c acks.
Hyd ous oli ine samples we e p epa ed by adding an equimola mix u e o B uci e and
Silica o he powde ed oli ine and hen ho p essed in he 6-8 mul ian il de ice. In some
ins ance, B uci e-Silica agg ega e was added o he p e-ho p essed samples while
assembling p ocess be o e he expe imen .
2.2.2 Placing pla inum Shea s ain ma ke
A e ho p essing he sin e ed samples we e ca e ully emo ed om he high-p essu e
assembly and he op and bo om o he capsule emo ed. 1.2mm diame e cylind ical
co es we e p epa ed om hese samples in he hin sec ion labo a o y by H. Schulze. Using
a diamond wi e saw, 200µm hick ellip ical slices o ho p essed oli ine sample we e hen
cu om he co e each o which we e o ien ed a 45° o he co e axis. Each oli ine slice was
hen cu in o wo symme ical hal es, wi h he di ec ion o cu ing o ien ed pa allel o he
o iginal cylind ical co e (Fig 2-11). The exposed su aces o his cu we e hen spu e
coa ed wi h a ~100nm P laye . Pu ing a P ma ke in his way, on a c oss-sec ion cu
pa allel o he cylind ical axis minimizes he ma ke o a ion due o axial comp ession o
he sample.

[37]
0.2 mm
Sample
P S ain
Ma ke
Alumina
Pis on
P capsule
Figu e 2-11: Emplacemen o pla inum s ain ma ke o shea s ain measu emen .
App oxima ely 100 nm hick Pla inum-laye is spu e coa ed on he sides o he wo cu hal es o he
ho p essed sample. Ro a ion o he s ain ma ke is di ec ly ela ed o he shea s ain
2.3 Analy ical Me hods
As he cen al heme o his hesis has been o s udy o he e ec s o physical and
chemical pa ame e s on he c ys allog aphic p e e ed o ien a ion o oli ine, elec on
backsca e di ac ome e (EBSD) has been he p ima y analysis echnique. I is also c ucial
howe e o ela e he de e mined CPO, o lack he eo , o he unde lying de o ma ion
mechanism and ac i e slip sys em in he c ys al and o his pu pose T ansmission elec on
mic oscope (TEM) has been a i al analysis echnique.
Wa e con en in he s a ing sample and he eco e ed sample has been measu ed
using Fou ie T ans o m In a ed (FTIR) spec oscopy. A b ie in oduc ion o hese h ee
impo an ins umen s is p o ided in he nex pa ag aphs.
The Elec on p obe mic o-analyze (EPMA) was used o analyzing he chemical
composi ion o eco e ed samples and o examine he equilib ium dis ibu ion o Ca and
Mg be ween ens a i e and diopside agg ega es employed in expe imen s o calib a e he
empe a u e in he p essu e cell. Raman spec oscopy has also been u ilized o phase
[38]
iden i ica ion in he un p oduc s om expe imen s pe o med o calib a e cell p essu e
using qua z-coesi e and coesi e-s isho i e phase ans o ma ions.
2.3.1 Measu emen o c ys allog aphic p e e ed o ien a ion using Elec on
backsca e di ac ion echnique (EBSD)
Elec on backsca e ed di ac ion (EBSD); some imes also e e ed o as backsca e
Kikuchi di ac ion (BKD) is a echnological add-on o a scanning elec on mic oscope. I
p o ides an SEM wi h a mic os uc u al-c ys allog aphic analysis capabili y. P ima ily,
EBSD is used o s udy ex u e o p e e ed o ien a ion o any c ys alline o polyc ys alline
ma e ial. This is achie ed by indexing and iden i ying he c ys al sys ems. Apa om
s uc u e and o ien a ion in o ma ion, EBSPs (Elec on back sca e pa e ns; See igu e
No. 2-12 o an example o such a pa e n in mine al oli ine) con ain addi ional in o ma ion
on c ys al la ice pe ec ion, local s ain, de o ma ion, and g ain bounda ies.
T adi ionally hese ypes o s udies ha e been ca ied ou using x- ay di ac ion (XRD),
neu on di ac ion and/o elec on di ac ion in a TEM.
[39]
Figu e 2-12: Fo ma ion o backsca e ed Kikuchi pa e ns by EBSD in he SEM. (a) O igin o
Kikuchi lines om he EBSD (i.e., il ed specimen) pe spec i e. (b) EBSD pa e n om oli ine
(accele a ing ol age 20 kV).
EBSD sys em consis s o a phospho sc een, compac lens and low ligh CCD came a
a ached o a Scanning Elec on Mic oscope (SEM) (Fig. 2-13). A polished sample specimen
is placed in o he no mal posi ion in he specimen chambe , and is il ed o ~70° om he
no mal posi ion. Doing so boos s he con as o EBSPs.
EBSPs a e gene a ed i a s a iona y beam in e ac s wi h he su ace o a c ys al [Alam e
al., 1954],[Venables and Ha land, 1973]. The elec ons while in e ac ing wi h an a om
unde go inelas ic sca e ing. I esul s in a ac ion o he elec ons losing a small pa o
hei ene gy. This p ocess c ea es a di e gen sou ce o elec ons close o he su ace o he
sample. Some o hese elec ons a e inciden on a omic planes a angles which sa is y he
B agg equa ion. Fo each gi en plane, hese elec ons emana e in di ac ion cones om
bo h he on and back su ace o he plane. When hese cones in e sec he phospho
sc een, he Kikuchi lines a e o med. The Kikuchi lines appea as almos s aigh lines
because he cones a e e y shallow as he B agg angle is o he o de o 1°. Small B agg
angle esul s om he ac ha inciden elec ons ha e e y high ene gy and hence e y
small wa eleng h (λ ≈ 8 pm o a 25 KeV elec on beam). Hence, Kikuchi bands a e
e ec i ely he ace o he plane om which hey a e o med and he EBSD pa e n is
he e o e a gnomonic p ojec ion o he c ys al s uc u e.
[40]
Figu e 2-13: Schema ic se up o an EBSD sys em showing i s p incipal componen s
Thus, he whole Kikuchi pa e n consis s o pai s o pa allel lines whe e each pai , o
“band,” has a dis inc wid h and co esponds o a dis inc c ys allog aphic plane. The
in e sec ion o bands co esponds o a zone axis (pole), and majo zone axes a e
ecognized by in e sec ion o se e al bands. The Kikuchi pa e n he e o e essen ially
embodies all he angula ela ionships in a c ys al—bo h he - and in e -plana angles—
and hence implici ly con ains he c ys al symme y. Figu e 2-12 shows an EBSD Kikuchi
pa e n om mine al San-Ca los oli ine. The o ien a ion o he pa e n and hence o he
olume om which i has a isen is e alua ed by “indexing,” ha is, iden i ying he poles and
bands in he pa e n, and calcula ing he ela ionship be ween hese and some chosen
e e ence axes.
F om Kikuchi bands o pole igu e
Au oma ed iden i ica ion o a c ys al o ien a ion in ol es iden i ying he kikuchi lines in
he g ay scale image con aining EBSPs. Fi s s ep owa ds his p ocess in ol es he p e-
p ocessing s age o edge de ec ion. This is a non- i ial ask because con as no mal o he
kikuchi lines a ely change ab up ly. The pa e n o Kikuchi lines on he phospho sc een is
elec onically digi ized and p ocessed o ecognize he indi idual Kikuchi lines. These da a
a e used o iden i y he phase, o index he pa e n, and o de e mine he o ien a ion o he
[47]
In an A ion-milling de ice he sample is loaded on a s age and lowe ed in o a acuum
chambe . In his acuum he e a e wo a gon ion guns, on opposing sides o he sample.
Each gun consis s o an anode inside o a ca hode ube wi h small hole in i . The anode is
connec ing o an A supply, and since he e is a se e al kV di e ence be ween he anode
and ca hode, A is ionized and accele a ed h ough a hole in he ca hode ube and di ec ed
a he specimen. As he accele a ed A ions hi he su ace o he specimen, hey spu e he
op su ace laye o hen specimen away, by which he specimen is hinned. Ion milling
usually esul s in wo kinds o damage o he specimen, c ea ion o opog aphy on a ini ially
la su ace and secondly he c ea ion o an amo phous laye o ma e ial on he su ace
(Ba na e al., 1999). The i s kind o damage can be educed by o a ing he sample while
hinning and using a low incidence angle o he ion beam on he specimen, howe e due o
he geome y o he holde , he lowes incidence angle ha can be used o hinning in he
ion milling de ice is 12 deg ees. Ro a ion o he specimen educes he c ea ion o
opog aphy since he spu e ing a e is dependen on he o ien a ion o he inciden ion
beam ela i e o he c ys al la ice o he c ys alli es in an agg ega e [Ba na and Menyha d,
1994]. Ro a ion o he specimen hus a e ages ou (o a leas educes) he o ien a ion
dependen spu e ing a e. C ea ion o a damaged o amo phous laye is ha de o educe.
Typical ope a ing condi ions o ion mille a e 2 – 5 kV, in his ange ioniza ion o he a ge
can happen by elec on exchange be ween he inciden elec on and specimen, and hus
modi ying he di ec su ace laye o he specimen (Malhe be, 1994). As he inciden ions
impinge on he su ace, he ion will lose i s ene gy in wo di e en way, ei he h ough
elec on in e ac ions wi h a oms o he specimen, o by nuclea in e ac ion (momen um
ans e ) wi h specimen a oms be o e i inally becomes apped (implan ed) wi hin he
specimen. Elec onic in e ac ion esul s in he ioniza ion o specimen a oms. Nuclea
in e ac ion, o collision, be ween he inciden ion and a oms o he specimen esul in a
momen um o ene gy ans e om he impinging ion o a specimen a om. Momen um
ans e se s he a oms o he specimen in mo ion and leads o spu e ing o he specimen
a oms i he a oms is eed om he su ace, o o he wise may lead o a cascade o
collisions inside he specimen. I he impinging ion ans e s enough ene gy, no a single
cascade o collision will occu , bu a g oup o a oms will be se simul aneously in mo ion, a
so-called 'spike', which may comple ely amo phize a po ion o he la ice (Malhe be,

[48]
1994). The abo e men ioned p ocesses a e dependen on he inciden angle, cha ge, mass
and ene gy o he inciden ion, and he p ope ies o he a ge (Ba na and Menyha d, 1994;
Ba na e al., 1999). The hickness o he damaged o amo phized laye hus will also a y
wi h hese a iables. A highe ene gy o he inciden ion inc eases pene a ion dep h and
hus he hickness o he damaged laye , whe eas as highe mass o he inciden ion
dec eases i pene a ion dep h. A lowe inciden angle will dec ease he hickness, hough
below an inciden angle o 10° he dependence on inciden angle becomes e y weak
(Ba na e al., 1999). A mul iply ionized inciden ion will ha e a highe ene gy and hus
c ea e a hicke damage laye . Nex o his, ene gy ans e o he inciden ion o he
specimen may also esul in hea ing o he specimen, which may also damage o modi y he
specimen, cooling he e o e is equi ed o some specimens.
Sc ew and edge disloca ions
S ain is de ined by he displacemen o an a om om i s posi ion ha would be
expec ed om he o dina y pe iodici y o he c ys al. Causes o such a s ain ield can be
disloca ions, plana de ec s o o he impe ec ions in he c ys al la ice. In he case o
disloca ions wo di e en ypes, o end ypes since mos disloca ions ha e a mixed
cha ac e , o disloca ions can be dis inguished, i.e. sc ew and edge disloca ions. Edge
disloca ions can be seen as he inse ion o an addi ional la ice hal plane in he o he wise
egula c ys al la ice. A he place whe e his addi ional hal plane ends, he la ice will be
dis o ed a ound he end o he edge disloca ion, called he disloca ion co e, due o
elaxa ion o he la ice a ound he co e (Figu e 2-16a). Sc ew disloca ion he la ice is
shea ed on a plane such ha pa o he la ice abo e his plane has an o se o he la ice
below he plane, he di ec ion o he o se also lies in his plane ( igu e 2b). Disloca ion a e
desc ibed by he Bu ge s ec o b, he la ice ec o ha closes he ci cui a ound he
disloca ion co e ( igu e 2-16), and he disloca ion line (di ec ion DL) a ound which he
la ice is de o med s onges . Fo an edge disloca ion he disloca ion line and he Bu ge s
ec o a e pe pendicula , o a sc ew disloca ion hey a e pa allel.
[49]
The dis o ion o he la ice a ound he disloca ion co e may al e he la ice is such a
way ha di ac ed beams ha a e in he undis o ed la ice no exci ed, become exi ed in
he dis o ed la ice.
Figu e 2-16: Illus a ion o edge and sc ew disloca ions in a hypo he ical c ys al. Bu ge s ec o
“b”, he la ice ec o ha closes he ci cui a ound he disloca ion co e and disloca ion line has been
ep esen ed by “DL”, a). In case o edge disloca ion, bu ge s ec o is no mal o he disloca ion line. In
his case, slip plane is de ined as he plane con aining he disloca ion line and he bu ge s ec o , b).
In case o sc ew disloca ion, he disloca ion line and bu ge s ec o a e pa allel.
2.3.3 FTIR
The hyd oxyl dis ibu ion wi hin he samples we e analyzed wi h unpola ized Fou ie
ans o m in a ed (FTIR) using a B uke ™ IFS 120 HR high- esolu ion FTIR spec ome e
wi h a B uke ™ IR mic oscope.
A schema ic pic u e o he in e e ome e is shown in ig 2-17. The spec ome e is
coupled wi h a B ucke IR mic oscope con aining all e lec ing Casseg ain op ics ha
allows measu emen s o small a eas wi h ape u es down o 10 µm. Measu emen s in he
nea in a ed egion we e ca ied ou using ungs en ligh sou ce, CaF2 beam/spli e and
high sensi i i y, na ow-band MCT de ec o . The s anda d op ic se ings o he di e en
equency anges o in e es o analyzing wa e species a e also shown in able 2-3.
[50]
1. IR beam in ansmission mode
2.
Condense mi o
3.
Sample holde
4.
Casseg anian objec i e
5.
Objec i e
6.
Ape u e
7.
Mo ing mi o
8.
Ocula
9.
MCT de ec o
Figu e 2-17: De ails o he FTIR mic oscope (Red awn om Bol an-Casano a, 2000)
Table 2-3 : Op ics se ings o di e en equency anges used o analyze wa e species
Op ic
pa ame e
F equency
Range
Nea in a ed
Mid In a ed
Beam spli e se ing
CaF2
KB
De ec o se ing
MCT mic oscope
MCT mic oscope
Sou ce se ing
W lamp
Globa
Se e al hund ed scans we e accumula ed o each spec um wi h 1 o 4 cm-1 esolu ion.
Du ing he measu emen s, he op ics o he spec ome e was e acua ed and he
mic oscope was pu ged wi h a s eam o H2O and CO2— ee pu i ied ai . Pola ized in a ed
[51]
adia ion was gene a ed using a me al-s ip pola ize on a KRS-5 subs a e. Backg ound
co ec ions o abso bance spec a we e ca ied ou by a piecewise con inuous cubic i o
he baseline de ined by poin s ou side he OH-s e ching egion.
Calcula ion o he wa e con en
The measu emen o wa e con en by IR spec oscopy is based on he Bee Lambe
law:
𝐴= 𝜀 × 𝑐 × 𝑡
Whe e,
𝐴 is he abso bance,
𝜀 is he ex inc ion coe icien , in cm-1 /(mol/L),
𝑐 is he concen a ion o he abso be , in mol/L, and
𝑡 is he sample hickness, in cm.
Thus, o ob ain quan i a i e da a he ex inc ion coe icien s o he OH bands mus be
known. These a e de e mined by he calib a ion o he in a ed da a wi h an independen
analy ical me hod. Fo mos o he man le phases, howe e , 𝜀 is no p ecisely known
because i s calib a ion is di icul o wo easons:
Analy ical echniques, such as Ka l Fishe i a ion and gas ex ac ion manome y, a e
limi ed o high wa e con en s and hus equi e a la ge amoun o homogeneous ma e ial,
which, o high p essu e syn he ic samples, is ex emely di icul o ob ain. 1H MAS NMR
(Magic-Angle-Spinning Nuclea magne ic esonance) spec oscopy is an in insically
quan i a i e echnique which has ecen ly been used o he s udy o wa e in man le NAMs
(Kohn, 1996). The main disad an age o his echnique, howe e , is ha he samples mus
be i on- ee.
H2O bound o mine al su aces and con ained in inclusions can lead o inaccu a e
alues. Thus he wa e con en o he sample can easily be o e es ima ed by such bulk
echniques.
[52]
In he p esen s udy, because o he absence o speci ic calib a ion o mos high-
p essu e phases, he concen a ion o hyd oxyl g oups we e de e mined by in eg a ing he
abso p ion using he calib a ion o ex inc ion coe icien s by Pa e son (1982).
𝐶=𝑋𝑖
150𝜉�𝐾(𝜈)
(3780 −𝜈)𝑑𝜈
Whe e,
𝐶 is he concen a ion o hyd oxyl (in H/106 o ppm w H2O),
𝜉 is an o ien a ion ac o , equal o
13
� o unpola ized measu emen on oli ine
[Mackwell and Kohls ed , 1990],
𝐾(𝜈) is he abso p ion coe icien (in cm-1) o a gi en wa enumbe 𝜈 , and
𝑋𝑖 is a densi y ac o . I s alue is chemical composi ion dependen . 𝑋𝑖= 4.39 ×
104 𝐻106𝑆𝑖
⁄ o 2695 w ppm H2O o oli ine (Fo90).

[53]
2.3.4 Piezoelec ic measu emen s o s ess in he Mul ian il appa a us
Quan i a i e heological measu emen s and he de e mina ion o mine al and ock low
laws ely on he abili y o measu e de ia o ic s esses in ma e ials unde well-de ined
condi ions. In oom o low p essu e de ices s ess measu emen s a e achie ed h ough he
use o a load cell which mus be mechanically coupled o he loaded sample bu which
esides ou side o he sample en i onmen o high p essu e chambe . A load cell uses he
s ain esponse o a calib a ed ma e ial o measu e o ce, wi h s ain con e ed in o
a ia ions in elec ical conduc i i y using a s ain gauge. In he G iggs solid media
de o ma ion appa a us, which ope a es o p essu es o app oxima ely 3 GPa, he s ess in
he de o ma ion pis on is measu ed by means o a load cell in mechanical con ac wi h he
sample h ough he pis on and ha d alumina ods [Holyoke and K onenbe g, 2010].
Howe e , a p essu es highe han 3 GPa whe e mul ian il de ices a e employed s ess
measu emen s a e ex emely challenging. Ex-si u measu emen s can be made o de e mine
a e age s esses using sample speci ic calib a ions o disloca ion densi ies o
ec ys alliza ion g ain size. Howe e , such me hods can only be calib a ed a lowe
p essu es wi h de ices ha use load cells and hei applica ion a high p essu e is
unce ain and limi ed. In si u x- ay di ac ion measu emen s o examine he dis o ion o
di ac ed Debye Sche e ings can be used o de e mined la ice s ain om which
s esses can be de e mined. Howe e , he accu acy o such measu emen s is cu en ly o
he o de o a leas 100 MPa and a igo ous in e nally consis en ea men o such
di ac ion da a has ye o be demons a ed [Du ham e al., 2009].
A load cell ha can be used in e nally in solid media p essu e assemblies would be a
signi ican ad an age, pa icula ly i i could p ese e he accu acy in s ess demons a ed
by low-p essu e de ices. Mechanical s ain gauges placed in e nally in a solid media de ise
would be ex emely di icul o calib a e and o sepa a e changes in p essu e om hose o
de ia o ic s ess. The calib a ion o he cha ge de eloped on he su ace o piezoelec ic
c ys als as a unc ion o o ce, howe e , may be one p omising al e na i e.
A piezoelec ic cha ge de elops on he su ace o a non-cen osymme ic c ys al in
esponse o he applica ion o mechanical s ain. Piezoelec ic c ys als and ce amics ind
[54]
uses in an eno mous ange o de ices such as ansduce s, mic ophones, igni e s and
mic o-ac ua o s. One o he mos common piezoelec ic c ys als employed is qua z, which
is used o example as a piezoelec ic esonan oscilla o o p oduce an elec ic signal wi h
a p ecise equency.
The cha ge pola iza ion o a c ys al caused by he applica ion s ess is ela ed o he
applied o ce h ough he piezoelec ic cha ge cons an , dab, whe e “a” is he di ec ion o
he pola iza ion and b is he di ec ion o he applied s ess. I is con en ion o desc ibe he
c ys allog aphic di ec ions x, y, z wi h subsc ip s 1, 2, 3, wi h shea abou one o hese axes
being e e ed o by 4, 5, 6. Piezoelec ic c ys als ha e di e en esponses depending on
he di ec ion o s ess wi h espec o he di ec ion o cha ge pola iza ion as shown in
igu e 2-18.
Figu e 2-18: Piezoelec ic c ys al con igu a ions showing di e en o ien a ions o he applied
o ce wi h espec o he cha ge pola iza ion.
The ela ionship be ween change Q and o ce F a ies wi h he con igu a ion bu only
o ans e se ope a ion is i a unc ion o c ys al shape.
Longi udinal
𝑄 = 𝐹𝑑11
2.a
T ans e se
Q= 𝐹𝑑13𝐿𝑇
�
2.b
Shea
Q = 𝐹𝑑14
2.c
[55]
d11 o qua z, o example, is –2.3x10-12 C/N. The cha ge de eloped on a c ys al can be
de e mined by measu ing he ol age; howe e , as he cha ge is e y small he discha ge
ime would be o he o de o nano-seconds. A me hod o ampli ica ion is equi ed in o de
o con e he cha ge on he c ys al in o a measu able ol age. This can be pe o med by
cons uc ing a cha ge ampli ie o in eg a o ampli ie . An in eg a o ampli ie uses an
ope a ional ampli ie ci cui wi h a esis o and capaci o in se ies, a so called RC ne wo k.
The ci cui p oduces an ou pu ol age, which is p opo ional o he in eg al o he inpu
ol age as a unc ion o ime. A cha ge o equal magni ude o ha on he piezoelec ic
c ys al builds up on he ange capaci o and he ou pu ol age is a unc ion o he ange
capaci ance and he cha ge.
Op
amp
V
C = 10 nF
Piezoelec ic
C ys al Cables
Cc
Rc
Ri
Vo
Cha ge Ampli ie
Figu e 2-19: A simpli ied ci cui diag am o he cha ge ampli ie p oduced by combining an
ope a ional ampli ie wi h an RC ne wo k.
In igu e 2-19, C is he ange capaci o , Ri is an insula ing inpu esis ance and Rc and
Cc a e he esis ance and capaci ance o he cables connec ing he c ys al o he cha ge
ampli ie . The ou pu ol age 𝑉𝑜 is
Shea
𝑉0=−𝑄𝐶𝑟
�
2.d
[56]
An impo an aspec in he use o a cha ge ampli ie in he measu emen o small
cha ges is he elimina ion o d i , which is an undesi able change in he ou pu signal o e
ime. D i o igina es om leakage o cu en o cha ge h ough he cabling (Cc,Rc), he
c ys al i sel o he ope a ional ampli ie (al hough mode n MOSFET op amps ha e an
ex emely high gain ha elimina es d i ). I a piezoelec ic cha ge is o be measu ed om
wi hin a high p essu e mul ian il assembly, hen he cabling wi hin he cell assembly needs
o ha e a e y high Rc. As many ce amics can con ain H2O o C ex eme cau ion has o be
aken o ensu e he esis ance ac oss he cables emains ex emely high in o de o
elimina e d i .
Choice o Piezoelec ic c ys al
Fo high p essu e and po en ially high empe a u e measu emen s in a mul ian il
assembly a piezoelec ic c ys al needs o be selec ed wi h a sui able he modynamic
s abili y. Al hough qua z is s able o a leas 15 GPa a oom empe a u e, a 573°C i
ans o ms o β-qua z and he piezoelec ic e ec is los . GaPO4 is isomo phic wi h qua z
bu has a piezoelec ic coe icien (𝑑11 =– 4.5𝑥10 −12 𝐶/𝑁) ha is wice as la ge [K empl
e al., 1997]; [Damjano ic, 1998]. The compa able β-qua z high empe a u e phase
ansi ion occu s a 930°C a oom p essu e, enabling GaPO4 o be used as a piezoelec ic
ma e ial o much highe empe a u es han qua z. High p essu e s udies ha e indica ed
ha a oom empe a u e GaPO4 is s able in he qua z s uc u e o a leas 9 GPa [Sowa,
1994]; [Bad o e al., 1998] Two x-pla es o GaPO4 single c ys al (a pla e wi h he hickness
in x-di ec ion o measu e d11 in he longi udinal con igu a ion) we e kindly dona ed by
Piezoc ys GmBH. The pla es we e 0.4 and 1 mm hick and could be co ed o any diame e .
High p essu e sample assembly o piezoelec ic measu emen s
The main conside a ion in he de elopmen o a high p essu e cell assembly o
piezoelec ic measu emen s is elimina ing o minimizing d i by ensu ing ha no cu en
can leak ac oss he piezoelec ic c ys al h ough he assembly ma e ial. D i causes he
ou pu ol age o ise s eadily and o en apidly wi h ime. Once he sa u a ion ol age o
10 V is eached C is ully cha ged and is discha ged au oma ically by he closing o he
swi ch ac oss C . Many di e en assembly and cable con igu a ions we e es ed in o de o
[63]
Figu e 3-3: Ro a ion o pla inum s ain ma ke θ and amoun o shea ∆l o a s ain ma ke
ini ially o ien ed a 45° o he base o he specimen. Do ed pa allelog am depic s he ini ial
o ien a ion o a hypo he ical plana elemen o hickness “ ” ha unde goes shea ing due o he
sidewise mo emen o he alumina pis ons. Solid lines indica e he new o a ed posi ion o he same
elemen a e he shea s ain o γ.
𝐶𝑎𝑠𝑒 𝐼:𝑖𝑓 𝜃 ≤45°
𝛾=�∆𝑙
𝑡�= 1 − an (45°−𝜃)
3.a.
𝐶𝑎𝑠𝑒 𝐼𝐼∶𝑖𝑓 𝜃>45°
𝛾=�∆𝑙
𝑡�= 1 + an (𝜃−45°)
3.b.
𝐻𝑒𝑛𝑐𝑒,𝑖𝑛 𝑔𝑒𝑛𝑒𝑟𝑎𝑙,𝑠ℎ𝑒𝑎𝑟 𝑠𝑡𝑟𝑎𝑖𝑛
𝛾=�∆𝑙
𝑡�= 1 − an(|45°−𝜃|)
3.c.
𝑤ℎ𝑒𝑟𝑒 |𝑥| 𝑖𝑠 𝑚𝑜𝑑𝑢𝑙𝑢𝑠 𝑜𝑓 𝑥.
In a e e ence ame a ached o he alumina pis on, o a ional componen o he s ain
ma ix is absen . Hence, he equi alen s ain ma ix can be ep esen ed as:
�𝜖
𝑖𝑗
�=�0𝜀12 0
𝜀
21
0 0
0 0 0�
3.d.
𝑤ℎ𝑒𝑟𝑒 𝜀12 =−𝜀21 = 𝛾2
⁄

[64]
Figu e 3-4: Va ia ion in s ain expe ienced by he sample DD402 along i s hickness. Top-Le : The
pa s close o he alumina pis on a e s ained mo e han hose a e close o he neu al line N´N. Local
o ien a ion o he P s ain ma ke is shown using a solid whi e line whe eas o iginal o ien a ion o
P -ma ke is shown using a do ed ed line. Sense o shea is as indica ed by he wo ed a ows on he
op and bo om. Top-Righ : Local inc ease in he shea s ain in he sample nea alumina pis on has
been ma ked by a cu ly b acke . Bo om (Le and Righ ): These images show he di e ence in he
o a ion angle as we mo e away om he neu al line owa ds he alumina pis on.
The pla inum s ain ma ke also p ese es e idence ha in mos ins ances he sample
expe iences non-uni o m s ain along i s hickness. The shea s ain close o he alumina
pis on is gene ally sligh ly la ge han he s ain nea he cen e o neu al line o he
[65]
sample, N´N in Figu e 3-4. This occu s because he polyc ys alline specimen, unlike a single
c ys al, does no beha e as a pe ec igid body. T ans e ence o shea o ce be ween wo
adjacen low laye s occu s ia he in e laye ic ion, simila o a luid column. This
a ia ion in shea s ain should be mo e ma ked when g ains a e polygonal shaped and
ha e s aigh edges. In his case, g ains can easily slip pas each o he wi h lesse e ec o
in e -g anula ic ion han in he case o i egula shaped g ains (as is he case wi h
hyd ous oli ine sample).
Employing he P s ain ma ke and equa ion 3.c he shea s ain de e mined o he
edge o he sample is calcula ed om,
𝛾𝑚𝑎𝑥 = 1 − an(45°−26.7°)= 0.67
Howe e in he cen e o he sample he angle o he s ain ma ke is only 16.4° and he
de e mined shea s ain is 0.44. Inhomogenei ies in shea s ain o he o de o 30% occu
h ough mos samples and he shea s ain epo ed is he maximum alue eco ded.
Figu e 3-5: Reac ion o oli ine wi h
alumina o ms a laye o spinel and
ga ne a hei in e ace. This may
enhance he coupling be ween he pis on
and he specimen ma e ial (oli ine)
I is also possible o es ima e
shea s ain om he sidewise
displacemen o he alumina pis ons.
This assumes ha he pis ons a e
mechanically coupled o he sample because slip a he sample pis ons in e ace would
esul in e oneously la ge maximum s ain es ima es. The maximum shea s ain
expe ienced by sample DD402 can be calcula ed om he pis on displacemen om:
𝛾𝑚𝑎𝑥 =𝐷𝑖𝑠𝑝𝑙𝑎𝑐𝑒𝑚𝑒𝑛𝑡 𝑜𝑓 𝑡ℎ𝑒 𝑎𝑛𝑣𝑖𝑙
𝑆𝑎𝑚𝑝𝑙𝑒 𝑡ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠 ≅140 µ𝑚
200 µ𝑚= 0.7
[66]
This is in excellen ag eemen wi h he s ain ma ke es ima e and implies good
mechanical coupling be ween he sample and pis ons.
The shea s ain impa ed o he sample dec eases as he dis ance om he pis on
inc eases e en hough he e is no app eciable sliding be ween he alumina pis on and he
sample ma e ial, unlike in o he s udies ( [Zhang e al., 2000] ) whe e loss o s ain has
been obse ed qui e equen ly. We ha e made no special a emp s a enhancing he
coupling be ween he pis on and he specimen ma e ial. All alumina pis ons we e cu using
diamond wi e saw.
In ou case, he be e coupling could be esul o highe con ining p essu es (3 o 8.5
GPa) which should ensu e a much be e ic ional con ac by localized de o ma ion o he
alumina pis ons and he oli ine sample along hei in e ace. Reac ion be ween oli ine and
alumina leading o he o ma ion o laye o spinel and ga ne a he oli ine-alumina
in e ace could be ano he ac o ha migh be esponsible o a be e coupling (Figu e 3-
5).In some ins ances i was no possible o place a P s ain ma ke in he sample due o he
sample being oo agile. In hese cases i was also possible, o es ima e he shea s ain by
measu ing he la e al displacemen o he op and bo om alumina pis ons (Figu e 3-2B).
[67]
3.4 SEM and EBSD cha ac e iza ion
Reco e ed samples we e cu and polished o SEM obse a ions o de e mine he g ain
size dis ibu ion and la ice-p e e ed o ien a ion. 5° clus e size and 15° Gaussian
smoo hing has been applied o gene a ing he pole igu es om EBSD da a, unless
o he wise speci ied.
3.4.1 LPO de e mina ions o d y San Ca los oli ine samples
3 𝑮𝑷𝒂 p essu e and 1300°C
S ain- a e: 2.5x10-5; No. O g ains: 1835; Shea s ain ≈ 1.5
Figu e 3-6
: D y samples de o med a 3 GPa and 1300°C. Sample de o med a lowe s ain a e
(Top) shows dominan slip sys em o be (𝟎𝟏𝟎)[𝟏𝟎𝟎]. Oli ine a-axes a e p e e en ially aligned sub-
pa allel o he shea di ec ion whe eas b-axes a e aligned subno mal o he slip plane. (Bo om)
Sample de o med unde highe s ain a e also show he p esence o (𝟎𝟏𝟎)[𝟏𝟎𝟎] slip sys em along
wi h
(𝟎𝟏𝟎)[𝟎𝟎𝟏]
slip sys em.
EBSD pa e ns om samples eco e ed om expe imen s pe o med a 3 GPa and
1300°C a e shown in Figu e 3-4 o as and slow s ain a es. The expe imen al shea
di ec ion is indica ed ho izon al o he page. The sample de o med a a slowe s ain a e o
2.5x10-5 s-1 exhibi s alignmen o oli ine a-axes sub-pa allel o he shea di ec ion whe eas
he oli ine (010) is aligned sub-pa allel o he shea plane, as e iden om he alignmen o
oli ine [010] axes no mal o he shea di ec ion. The specimen de o med a a highe s ain
a e o 50x10-5 s-1 also shows a s ong ab ic o igina ing om he (010)[100] slip sys em.
[68]
Howe e , he e is also an e idence o a weake ex u e o igina ing om slip on
(010)[001] slip sys em as indica ed by he pa ial alignmen o [001] axes sub pa allel o
he shea di ec ion. Obse a ion o (010)[001] along wi h (010)[100] is in line wi h he
obse a ion o Jung and Ka a o (2001) whe e hey ound (010)[001] slip sys em o be
dominan a highe s esses.
5 𝑮𝑷𝒂 p essu e and 1300°C
The sample de o med a 5 GPa and 1300°C a a s ain a e o 2.5x10-5 has (010)[100] as
he dominan slip sys em, simila o expe imen s a 3 GPa. Whe eas he sample de o med
unde simila condi ion bu wi h a as e s ain a e o 50x10-5 appea s o ha e he
(010)[001] slip sys em also making an impo an con ibu ion o he o e all de o ma ion.
In his case, he o e all LPO in he sample becomes weake . This weakness may esul om
he compe ing ac ions o mo e han one slip sys em.
S ain- a e: 2.5x10-5; No. O g ains: 1675; Shea s ain ≈ 1.1
S ain- a e: 40x10-5; No. O g ains: 1520; Shea s ain ≈ 0.7
Figu e 3-7 : D y samples de o med a 5 GPa and 1300°C. Sample de o med a lowe s ain a e (Top) shows
dominan slip sys em o be
(𝟎𝟏𝟎)[𝟏𝟎𝟎]
. Oli ine a-axes a e p e e en ially aligned sub-pa allel o he shea
di ec ion whe eas b-
axes a e aligned subno mal o he slip plane. (Bo om) Sample de o med unde highe
s ain a e also has bo h
(𝟎𝟏𝟎)[𝟏𝟎𝟎]
slip and
(𝟎𝟏𝟎)[𝟎𝟎𝟏]
slip sys em
ac i e. 20° Gaussian smoo hing was
applied o he pole igu e o specimen DD350.

[69]
5 𝑮𝑷𝒂 p essu e and 1400°C
Samples de o med a 5 GPa, bu a a sligh ly highe empe a u e o 1400°C show a
simila mix o he wo slip sys ems as ound a lowe empe a u es. While a he slowe
s ain a e bo h (010)[100] and (010)[001] slip sys ems a e sub equally ac i e, a he
highe s ain a e o 50x10-5, he (010)[001] slip sys em appea s o dominan de o ma ion.
In e es ingly, he LPOs a his empe a u e appea o be s onge han a 1300°C. This may
be a esul o sligh ly highe s ains in hese samples.
8.5 𝑮𝑷𝒂 p essu e and 1300°C
The LPO o he sample de o med a 8.5 GPa, 1300°C and a s ain a e o 2.5x10-5 shows
e idence o con ibu ion om (010)[100], (010)[001] and (100)[001] slip sys em.
P esence o (010)[100] and (010)[001] slip sys em is consis en wi h de o ma ion unde
mode a ely high s ess (285 MPa). P esence o (100)[001] may be esul o he ac i a ion
o ha de slip sys em as pe he Von Mises c i e ia.
S ain- a e: 4x10-5; No. O g ains: 2320; Shea s ain ≈ 1.4
S ain- a e: 50x10-5; No. O g ains: 1980 ; Shea s ain ≈ 1.5
Figu e 3-8: D y samples de o med a 5 GPa and 1400°C. Sample de o med a lowe s ain a e (Top) shows
has an LPO esul an o signi ican s ain con ibu ion om bo h
(𝟎𝟏𝟎)[𝟏𝟎𝟎]
and
(𝟎𝟏𝟎)[𝟎𝟎𝟏]
slip sys em.
(Bo om) Sample de o med unde highe s ain a e has
(𝟎𝟏𝟎)[𝟎𝟎𝟏]
slip sys em dominan .
[70]
Specimen DD335, which was also de o med a 8.5 GPa and 1300°C bu a a highe s ain
a e (50x10-5) indica es ha he (010)[001] slip sys em was p edomina ely ac i e. This
obse a ion is also consis en wi h he epo s o (010)[001] being easy slip sys em unde
highe s esses (395 MPa).
Specimen DD455 and ac i e slip sys ems
HKL Channel™ 5 p og am which has been used o EBSD da a analysis allows o
selec ion o subse s o a ew da a poin s. A new da a ile is c ea ed by using an ellip ical
subse selec ion ool shows by ed-do ed like in igu e 3-10-A. Such a subse can be
c ea ed in a way o include g ains wi h a pa icula o ien a ion. Figu e 3-10-A shows
selec ion whe e only he g ains wi h hei [010]-axes o ien ed sub-pa allel o he specimen
Y0 axis. The selec ed da a poin s ma ked by ed-do ed ellipse a e used o d awing a new
EBSD pole- igu e as shown in igu e 3-10-B. The pu pose o analyzing such subse s is o
es ablish he simul aneous ac i i y o mo e han one slip sys em.
S ain- a e: 2.5x10-5; No. O g ains: 1850 ; Shea s ain ≈ 1.4
S ain- a e: 5x10-4; No. O g ains: 2500 ; Shea s ain ≈ 1.2
Figu e 3-9: D y samples de o med a 8.5 GPa and 1300°C. Sample de o med a slowe s ain a e (Bo om)
shows dominan slip sys em o be
(𝟎𝟏𝟎)[𝟏𝟎𝟎]
and
(𝟎𝟏𝟎)[𝟎𝟎𝟏]
. Oli ine a-axes and c-axes a e p e e en ially
aligned sub-pa allel o he shea di ec ion whe eas b-axes a e aligned subno mal o he slip plane. (Bo om)
Sample de o med unde highe s ain a e show he p esence o
(𝟎𝟏𝟎)[𝟎𝟎𝟏]
slip sys em. 20° Gaussian
smoo hing was applied o he pole igu e o DD335.
[71]
Subse selec ion – Only g ains
o ien ed sub-pa allel o b-axis
[010] || y0
No. o da a poin s in he
subse ;
N = 46918
[001] || x0
N = 43663
[100] || x0
N = 22546
[100] || z0
N = 27094
[001] || z0
N = 28081
[100] || y0
N = 2733
Figu e 3-10: Subse s o pole igu es indica ed a pa icula c ys allog aphic axis pa allel o a
selec ed specimen axis. [100] || x0 implies ha he subse con ains only he da a poin s such ha
oli ine [100] axes a e aligned pa allel (o sub-pa allel) o x-axis o he specimen.
X
Y
A
B
C
D
E
F
G
[72]
Taking he example o DD455, he [010] || y0 subse i.e. a subse consis ing o g ains
wi h hei b-axes aligned (sub-)pa allel o he y0 specimen axes, we obse e ha some o
hese g ains ha e hei a-axes aligned sub-pa allel o x0 –specimen di ec ion o shea
di ec ion. This pa icula subse also includes g ains wi h hei c-axes aligned sub-pa allel
o shea di ec ion. This obse a ion can be cons ued o be indica i e o compa a i e
ac i i y o he wo slip sys ems – (010)[100] and (010)[001]. A mo e e ined unde s anding
o slip sys em ac i i y can be ob ained by analyzing se e al such subse s and coun ing he
numbe o g ains (o da a poin s) p esen in hose subse s. Figu e 3-10-E shows mos ly he
g ains wi h dominan slip sys em being (010)[001]. We can make such an assump ion
because all such g ains wi h hei a-axes aligned pa allel o z0-axis o he specimen also
ha e hei b-axis aligned sub-no mal o he shea di ec ion and c-axis aligned sub-pa allel
o he shea di ec ion. Such a con igu a ion de elops due o he dominan ac i i y o
(010)[001] slip sys ems wi h numbe o da a poin s being 27094. On he o he , igu e 3-10-
B shows he pole igu e o hose g ains who a e nei he de o med p e e ably ei he in
(010)[100] slip sys em o (010)[001] slip sys em and he numbe o da a poin s in his case
is 46918. Hence, he numbe o da a poin s indica i e o (010)[100] slip sys em is (46918 –
27094) = 19824. Hence, we can say ha he ela i e ac i i y o (010)[100] o (010)[010] is
19824/27094 = 0.73:1. Simila ly, om igu e 3-10-D he ela i e ac i i y o (001)[100] o
(010)[100] is (22546 – 19824)/19824 = 0.137:1. F om igu e 3-10-G, we can de i e he
ela i i y ac i i y o (100)[001] slip sys em wi h espec o (010)[100] which is equal o
(2733 / 19824 ) = 0.138:1. Hence, he ac i i y o 4 majo slip sys ems is as ollows:
Slip sys em
(010)[100]
(010)[001]
(100)[001]
(001)[100]
Ac i i y
1
1.37
0.138
0.137
[79]
Figu e 3-14: TEM mic og aphs o he d y sample DD455 de o med slowly a 1300°C. Ac i e slip
sys ems a e (010)[100], (100)[001] and (010)[001]. Figu e on he le side shows la ge numbe o b =
[100] disloca ions p esen in one g ain. Whi e double-a ows in he pic u e indica e ha sense o
shea o he bulk sample.
3.4.4 Measu emen o sample s ess
Disloca ion densi y me hod
In his wo k, we ha e aken wo di e en app oaches o measu ing low s ess in he
sample. The i s app oach is based upon he ela ionship be ween disloca ion densi y and
low s ess, whe eas he second app oach elies upon he ac ha he ec ys allized g ain
size in ma e ials ha ha e de o med plas ically, is a unc ion o s ess as has been
discussed in he las sec ion.
Whe e e , TEM mic og aphs a e a ailable, s ess has been es ima ed using he Taylo ’s
equa ion (Eq. 3. ), which ela es low s ess o he disloca ion densi y in he specimen
[Kohls ed e al., 1976b].
𝜎1−𝜎3=𝛼𝑏𝜇𝜌12
�
3. .
Whe e; 𝛼≈3
𝑏=𝑏𝑢𝑟𝑔𝑒𝑟′𝑠 𝑣𝑒𝑐𝑡𝑜𝑟
𝜇=𝑠ℎ𝑒𝑎𝑟 𝑚𝑜𝑑𝑢𝑙𝑢𝑠
𝜌=𝑑𝑖𝑠𝑙𝑜𝑐𝑎𝑡𝑖𝑜𝑛 𝑑𝑒𝑛𝑠𝑖𝑡𝑦

[80]
Figu e 3-15: Disloca ion densi y e sus s ess ela ionship [Jung and Ka a o, 2001a]. The solid
line is he s ess e sus disloca ion densi y ela ionship o a single c ys al wi h he Schmid ac o =
0.5 [Kohls ed e al., 1976b].
The c i ical s ep in adop ing his p ocedu e is he es ima ion o disloca ion densi y, ρ,
which in u n is de ined as he o al leng h o disloca ion in a uni olume o he specimen.
Se e al me hods ha e been p oposed o he measu emen o o al leng h o disloca ion
which in ol e manual p ocessing o he mic og aphs [Bailey and Hi sch, 1960; Ham, 1961].
I is also no ewo hy ha he disloca ion densi y in polyc ys alline agg ega e may di e
om a single c ys al because o he e ogeneous de o ma ion nea g ain-bounda ies [De
B esse , 1996] and g ain bounda y mig a ion [Jung and Ka a o, 2001b]. Jung and Ka a o
2001) ha e p oposed a new calib a ion cu e o he ela ionship be ween disloca ion
densi ies and s ess (Fig 3-16).
Measu emen o he sample hickness: As desc ibed in he las chap e , ou TEM
specimens ha e been hinned using A gon milling p ocess. This p ocess esul s in a wedge
shaped g ain wi h a pla eau op as shown in Figu e 3-15 o he specimen DD384 de o med
[81]
a 8.5 GPa and 1300°C. Wedge shaped egion can be iden i ied easily by he p esence o
hickness inges a he edge o he g ain.
Figu e 3-16: Le -Du ing a gon milling p ocess, a gon s eam bomba ds he sample om op and
bo om (only op s eam is shown in he igu e). This gi es he milled g ain shape o a wedge (ma ked
by he p esence o hickness inges) wi h hal -angle being equal o he angle o incidence o a gon
s eam (~5°). App oxima e hickness o he pla eau o he g ain can be calcula ed om his simple
model. Righ -Wedge shaped pa and pla eau op ( egion enclosed by whi e ec angle) o a gon-milled
oli ine g ain o specimen DD384 is shown he e. This sample was de o med a 8.5 GPa and
1300°CNo e han base (b) o he wedge pa is app oxima ely 2 µm.
A gon milling has been done a a ying ol age wi h an angle o incidence o A gon
s eam being ~5°.
F om igu e 3-14, we ha e:
𝑇ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠 𝑜𝑓 𝑡ℎ𝑒 𝑝𝑙𝑎𝑡𝑒𝑎𝑢,𝑡= 2𝑏 an𝜃
𝐹𝑜𝑟 𝑏≅2µ𝑚 𝑎𝑛𝑑 𝜃= 5°; 𝑠𝑎𝑚𝑝𝑙𝑒 𝑡ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠≈350 𝑛𝑚
To al leng h o disloca ions: P ojec ed leng h o disloca ion in he a ea o in e es
(whi e ec angula egion ma ked in Figu e 3.14- igh ) can be es ima ed manually o using
an image p ocessing p og am. Assuming ha disloca ions a e a s aigh line and a e
inclined o he su ace o he mic og aph (Figu e 3.14-le ), he ue leng h o a
[82]
disloca ion, 𝑙 = �(𝑙𝑝)2+ (𝑡)2 whe e 𝑙𝑝 is he p ojec ed leng h o a disloca ion and 𝑡 is he
hickness o he g ain. To al leng h in his case, as measu ed om he 2-D mic og aph,
u ns ou o be ~17.34 µ𝑚.
Disloca ion Densi y: Now he disloca ion densi y in he olume co esponding o he
whi e ec angula a ea in he Figu e 3.14 is
𝑙(𝑎𝑟𝑒𝑎×𝑡ℎ𝑖𝑐𝑘𝑛𝑒𝑠𝑠)=17.34 µ𝑚
12.92 µ𝑚2×0.35 µ𝑚= �3.84 µ𝑚−2= 3.84 ×1012 𝑚−2
𝐻𝑒𝑛𝑐𝑒,𝑓𝑟𝑜𝑚 𝐸𝑞.3.𝑓; 𝑭𝒍𝒐𝒘 𝒔𝒕𝒓𝒆𝒔𝒔,𝜎1−𝜎2=𝟐𝟕𝟐 𝑴𝑷𝒂 𝑎𝑡 8.5 𝐺𝑃𝑎 𝑎𝑛𝑑 1300°𝐶.
Measu emen o s ess using ec ys allized g ain size piezome e
As has been discussed in he p e ious sec ions, mean ec ys allized g ain size in a de o med
specimen a ies wi h low s ess.
Figu e 3-15 shows his ela ionship o d y and we specimens [Jung and Ka a o, 2001b]. Figu e
3-17: S ess e sus ec ys allized g ain-size
ela ionship om Jung and Ka a o 2001. S ess
magni udes in he samples om his s udy we e
es ima ed om disloca ion densi ies. The solid
lines indica e he esul s o he leas squa e i
o he ‘d y’ and ‘we ’ condi ion. The size o
ec ys allized oli ine de o med unde ‘we ’
condi ions is signi ican ly la ge han ha unde
‘d y’ condi ions a he same s ess.
Re e ing back o he same sample
DD384 (See igu e 3-14 o a TEM
mic og aph o he same sample),
ec ys allized g ain size in his sample has been ound o be ~7.6 𝜇𝑚 . The s ess
co esponding o his alue o ec ys allized g ain size is ~230 𝑀𝑃𝑎 . This alue is 40 𝑀𝑃𝑎
lowe han he calcula ed alue om he disloca ion densi y me hod in he las sec ion. The
sou ce o his disc epancy in he low s ess alue could be he e o in es ima ing g ain
sizes om EBSD measu emen o due o he e o in he hickness measu emen o he
g ain om TEM mic og aph. In his case, a s ep size o 4 µ𝑚 has been used o da a
collec ion whe eas he mean g ain size is 7.6 𝜇𝑚.
[83]
3.5 Expe imen s unde we condi ion
De o ma ion expe imen s on we oli ine samples we e pe o med a 3, 5 and 8.5 GPa
p essu es wi h he maximum alue o he p essu e co esponding o ~250 KM o dep h in
he uppe man le (Table 3-2). A leas wo expe imen s we e pe o med a each p essu e-
empe a u e condi ion wi h he s ain a e di e ing by an o de o magni ude. Fo each o
he h ee p essu e poin s, expe imen s we e ca ied ou a 1300°C. Addi ionally,
expe imen s we e also pe o med a 1400°C a 5GPa and 1500°C a 8.5GPa.
The s a ing ma e ial o each de o ma ion expe imen consis ed o polyc ys alline San
Ca los oli ine powde mixed wi h equimola mix u e o b uci e and silica as he sou ce o
wa e (Table 3-3). A e achie ing he desi ed p essu e, he assembly was hea ed up and
le o anneal o 2-3 hou s a 1150°C, a empe a u e alue less han he a ge
empe a u e, o a oid excessi e g ain g ow h du ing annealing. This pe iod also p o ides
su icien ime o hyd a ion o he oli ine sample because o he b eakdown o b uci e
du ing hea ing. The ea e , he empe a u e was aised o he inal a ge le el o ha
expe imen and de o ma ion commenced h ough he ad ancemen o he de o ma ion
an ils.
Table 3-3: Lis o expe imen s and expe imen al condi ions unde we condi ion
Run ID
P essu e
(GPa)
Tempe a u e
(°C)
S ain
a e
(x10
-5
s
-1
)
Shea 1
S ain
Wa e Con en 2
(w . ppm)
S ess3
(MPa)
DD430
3
1300
3.2
1.2
90
250
DD477
3
1300
55
1.3
74
340
DD463
5
1300
4
1.1
279
210
DD461
5
1300
50
0.9
214
325
DD462
5
1400
5
2.4
168
310
DD466
5
1400
50
1.0
189
365
DD457
8.5
1300
2.5
1.1
419
310
DD456
8.5
1300
50
1.2
461
370
DD473
8.5
1500
15
1.1
401
300
DD460
8.5
1500
60
1.3
340
325
1 Shea s ain has been de e mined using sidewise displacemen o he pis ons.
2 Es ima ion o wa e con en is based upon calib a ion by Pa e son (1982).
3S ess has been de e mined using ec ys allized g ain size e sus s ess ela ion excep o DD456 which has been
calcula ed using he disloca ion-densi y ob ained om TEM mic og aph.
[84]
The shea s ain o each expe imen was calcula ed om he sidewa d displacemen o
he op and bo om alumina shea -pis ons om pos -mo em obse a ion o he de o med
assembly. This me hod had o be employed because ho p essed H2O-bea ing oli ine
agg ega es p o ed o be a oo agile o cu in o slices o added P s ain ma ke s. The
o ally s ain applied o he cubic assembly, measu ed using he ansduce s a ached o he
independen de o ma ion an ils, was always g ea e han ha de e mined om he
sidewa d displacemen o he alumina shea pis ons. This implies ha he s ain a e
ini ially applied o he cubic assembly was ac ually mode a ely as e han hose
expe iences by he sample. This di e ence esul s om he mino de o ma ion o assembly
pa s o he han he sample ma e ial. Hence, he nominal a ge s ain on he cubic
assembly was acco dingly eplaced by he de e mined sample s ain and hen he ue
s ain a e expe ienced by he sample was calcula ed and gi en in Table 3-1. The
co ec ions in s ain be ween hose measu ed on he en i e cubic assembly and hose
ac ually expe ienced by he sample whe e gene ally wi hin 10% o each o he . This is in
be e ag eemen wi h wha was obse ed o he d y samples and likely esul s om
signi ican so ening o he samples in he p esence o H2O.
Table 3-4: S a ing ma e ial o de o ma ion expe imen s on hyd ous oli ine
Sample ID
S a ing ma e ial
DD477
Ho p essed San Ca los oli ine (1 GPa) +
0.1 w % B uci e+SiO2 equimola mix u e
DD466
Ho p essed San Ca los oli ine (1 GPa) +
0.5 w % B uci e+SiO2 equimola mix u e
O he s
Polyc ys alline oli ine agg ega e +
0.5 w % B uci e+SiO2 equimola mix u e

[85]
3.5.1 Measu emen o wa e con en using FTIR
The wa e con en o each de o med specimen has been analyzed using FTIR
spec oscopy. Two hund ed scans a e collec ed o each spec um a a esolu ion o 1 cm-1.
Backg ound co ec ion has been made using a baseline ob ained by piecewise cubic
in e pola ion me hod (Fig. 3-16). Abso bance alues o each specimen a e no malized o
hypo he ical specimen o 1 cm hickness. In eg a ion has been pe o med be ween
wa enumbe s 2950 o 3780 cm-1. The calib a ion p oposed by Pa e son (1982), ins ead o
he ela i ely newe one p oposed by Bell e a. (2003), has been employed o ela e he
o al in eg a ed abso bance wi h wa enumbe o he wa e con en . This choice o
calib a ion makes i possible o compa e ou wa e con en da a wi h esul s om
Kohls ed e al. (1996) and also di ec ly compa e wi h esul s om p e ious de o ma ion
s udies (Jung e al. 2001; Ka a o e al., 2008) which ha e ou inely employed he Pa e son
calib a ion.
Figu e 3-18: Backg ound co ec ion o he aw FTIR da a. A baseline was c ea ed using piecewise
cubic in e pola ion me hod. Wa e solubili y alues a e sensi i e o he choice o he baseline and
ange o wa enumbe used o in eg a ion (2950 o 3780 cm-1 in ou case).
The Bell e al. (2003) calib a ion mos likely p o ides a be e es ima e o he H2O
con en o oli ine because i was pe o med speci ically on oli ine. The Bell calib a ion
would imply H2O con en s ha a e app oxima ely 3 imes highe in compa ison o ha
250030003500400045005000
-0.4
-0.35
-0.3
-0.25
-0.2
-0.15
-0.1
-0.05
0
0.05
Wa enumbe (cm
-1
)
Abso bance ( o 100 µm hick specimen)
Backg ound co ec ed da a
Baseline
Raw da a (DD461)
[86]
measu ed by he Pa e son calib a ion o H20 con en . Howe e , a he han co ec ing
da a p e ious s udies using an a bi a y co ec ion pa ame e he H2O con en s a e simply
epo ed using he olde calib a ion. Ul ima ely, he absolu e H2O con en s a e less
impo an han he deg ee o sa u a ion, which can be de e mined om compa ison wi h
he s udy o Kohls ed e al. (1996), which also used he olde calib a ion. Measu emen o
wa e con en by ei he o he abo e-men ioned app oaches, esul s in solubili y alues
ha a e sensi i e o bo h he choice o baseline and he ange o wa enumbe selec ed o
he in eg a ion.
As shown in he igu e 3-17, s ong IR abso bance peaks can be obse ed a
wa enumbe s (in cm-1) 3612, 3599, 3579, 3568 and 3568 o 8.5 GPa expe imen s (DD456,
DD457 and DD460). Mino peaks a 3504, 3475 and 3450 cm-1 can also be seen. In he 5
GPa un p oduc s (DD461, DD463 and DD462), peaks a 3579 cm-1 and 3568 cm-1 a e no
so well esol ed. IR spec a o nei he o he samples exhibi any ecognizable p esence o
g oup II bands ha has been epo ed o occu below 3450 cm-1 [Bai and Kohls ed , 1993;
Ma ee e al., 2001].
Figu e 3-19: FTIR spec a o hyd ous oli ine specimen a e he expe imen s. Abso bance o he
spec a was no malized o 1 cm hick specimen.
3400
3450
3500
3550
3600
3650
0
50
100
150
200
250
Wa enumbe (cm-1)
Abso bance (cm-1)
DD456
DD457
DD460
DD461
DD462
DD463
DD466
DD430
DD473
DD477
3568
3504
3475 3450
3612
3548
3599
3579
[87]
Figu e 3-20: Wa e solubili y in San Ca los oli ine (modi ied a e Kepple and Bol an-Casano a
[2006]). Ou esul s a e shown along wi h he expe imen al da a om Mosen elde e al. (2006; blue
diamond) and Kohls ed e al. (1996; ed squa es). The H2O con en s om his s udy employ he
Pa e son calib a ion so as o compa e hem di ec ly wi h he wo k o Kohls ed e al., 1996 whe e
oli ine was sa u a ed wi h excess H2O.This compa ison indica es ha he oli ine om his s udy had
H2O con en s less han he sa u a ion le el (25-35%). The S udy o Mosen elde e al (2006) epo ed
highe H2O con en s mainly because o using he newe Bell e al calib a ion.
Compa ison o ou wa e solubili y esul s wi h p e ious wo ks indica es ha solubili y
alues obse ed in ou samples a e lowe han he sa u a ion le el epo ed by Kohls ed
e al. (1996) om 1100°C (Fig. 3-18). The wa e con en in oli ines eco e ed om his
s udy a ies be ween 25 and 35 % o he sa u a ion le el. Sa u a ed H2O oli ine con en s
a e expec ed o ise sligh ly wi h empe a u e bu hen d op a highe empe a u es due o
he p esence o silica e mel ing. Some idea o how empe a u e may in luence H2O
sa u a ion le els can be gained by examining expe imen s pe o med by Smy h e al.
(2006) on o s e i e. The H2O sa u a ion limi in pu e o s e i e a 1250°C is almos double
he alue a 1100°C bu i d ops o almos hal o he 1100°C alue by 1500°C. The H2O
con en s epo ed o San Ca los oli ine in his s udy om 1500°C a e lowe han he
sa u a ed alues epo ed by Kohls ed e al. (1996) a 1100°C bu may s ill be close o he
[88]
sa u a ion limi , which may be lowe a his highe empe a u e. Da a om Mosen elde e
al. (2006) yield sa u a ed H2O con en s o empe a u es be ween 1100-1300°C a e 2-4
imes highe han he wa e con en s om Kohls ed e al. (1996), bu mos o his inc ease
can be a ibu ed o he use o he newe FTIR H2O calib a ion p oposed by Bell e al.
(2003).
3.5.2 NMR spec oscopy on hyd ous Fo s e i e
Posi ions o he esonance peaks ob ained om 1H MAS (magic angle spinning) NMR
a e ex emely sensi i e o he minu e di e ences in he chemical en i onmen a ound a
nucleus. Di e ences in he posi ion o he esonance peaks a e e e ed as chemical shi ,
which is measu ed in ppm. A known weigh o powde ed sample ma e ial is loaded in a
ce amic o o o leng h 1.2 cm and inne diame e 1.5 mm. A e being placed in he NMR
p obe, such ha he angle be ween he di ec ion o he ex e nal magne ic ield and o o
axis is 54.7° ( he magic angle), he o o is spun a e y high speed (30 KHz). This choice o
angle coupled wi h high o a ion speed, minimizes he b oadening o esonance peaks.
In nominally anhyd ous mine als, H should be s ongly bonded o he adjacen oxygen
[Kohn, 2006]. Resul s om hyd oxyl con aining mine als and o he ma e ials indica e ha
a s ong co ela ion exis s be ween no only be ween chemical shi (𝛿) and 𝑂−𝐻 dis ance
(𝑟𝑂𝐻) [B unne and S e nbe g, 1998] bu also be ween 𝛿 and 𝑂−𝐻. . 𝑂 dis ance (𝑟𝑂..𝑂)
[Ecke e al., 1988]. Mo eo e , he a ea unde an NMR esonance cu e is di ec ly
p opo ional o he numbe o esona ing nuclei. Hence, posi ion o he esonance peaks,
exp essed as chemical shi wi h espec o he esonance peak o e e ence ma e ial
Te ame hylsilane ((CH3)4Si, usually e e ed o as TMS), gi es us in o ma ion ega ding
he chemical en i onmen o he hyd ogen. On he o he hand, calcula ing he a ea unde
he esonance cu e and compa ing i wi h NMR spec a o a s anda d ma e ial wi h known
wa e con en (e.g. Gypsum), we can ind ou he absolu e numbe o he H nuclei p esen in
he sample. This app oach o measu ing wa e con en using 1H MAS NMR has a de ec ion
limi as small as 1 ppm o H2O by weigh [Kohn, 2006].
Hyd ous o s e i e samples we e p epa ed a 11 GPa and 1150°C using 6-8 ype mul i-
an il appa a us. In case o sample Z771, 2 w % o equimola mix u e o b uci e and silica
[95]
3.5.4 SEM and EBSD cha ac e iza ion
3 GPa p essu e and 1300°C
In he we sample de o med a 3GPa and a a slowe s ain a e o 3.2x10-5 e idence can
be seen o ela i ely equal ac i i y o bo h he (010)[100] and (100)[001] slip sys ems
(Fig.3-23: Top). In e es ingly poles o he (100) plane o he (100)[001] slip sys em a e sub
ho izon al and o a ed an iclockwise wi h espec o he Y0 axis. This o a ion is
anomalous, as he shea sense should cause an iclockwise o a ion only un il he maxima
a e aligned wi h he Y0 axis. The mos likely explana ion o his is as a esul o addi ional
comp essi e s ain expe ienced by he powde ed s a ing ma e ial. The di ec ion o he
comp essi e s ain is a 45° o he shea di ec ion.
In he d y expe imen s, epo ed in sec ion 3.3.2, (010)[100] was ound as he mos
common slip sys em. This is also obse ed o be he mos common slip sys em in na u al
samples and in p e ious expe imen s pe o med a ela i ely lowe s esses unde d y
condi ions [Ca e and A e'lalleman , 1970; Nicolas e al., 1973; Phakey e al., 1971; Zhang
and Ka a o, 1995; Zhang e al., 2000]. The (100)[001] slip sys em on he o he hand is
common in expe imen ally de o med specimens ha con ain mo e han app oxima ely 40
w . ppm wa e [Ka a o, 1995; Ka a o e al., 2008; Ka ayama and Ka a o, 2008], which is
he e o e qui e consis en wi h his esul .
The pole igu e o he specimen (DD477) de o med unde simila p essu e-
empe a u e-wa e condi ions bu pe o med a a highe s ain a e (5.5x10-4 s-1) (Fig. 3-
23: Bo om) exhibi s a ab ic de eloped only h ough he ac i i y o he (100)[001] slip
sys em. This expe imen is he e o e also consis en wi h p e ious expe imen s pe o med
a hese H2O concen a ions a p essu es below 2.2 GPa [Ka a o, 1995; Ka a o e al., 2008;
Ka ayama and Ka a o, 2008].

[96]
5 GPa p essu e and 1300°C
S ain- a e: 3.2x10-5; Wa e con en : 90 w . ppm;
No. O g ains: 3320; Shea s ain ≈ 1.2
S ain- a e: 5.5x10-4; Wa e con en : 74 w . ppm; No. O g ains: 3500; Shea s ain ≈ 1.3
Figu e 3-25
: We samples de o med a 3 GPa and 1300°C. Sample de o med a lowe s ain a e
(Top) shows wo ac i e slip sys ems – (010)[100] and (100)[001]. (Bo om) Sample de o med unde
highe s ain shows only (100)[001] slip sys em o be ac i e.
S ain- a e: 4x10-5; Wa e con en : 279 w . ppm; No. o g ains: 4010; Shea s ain ≈ 1.1
S ain- a e: 5x10-4; Wa e con en : 214 w . ppm; No. o g ains: 3800; Shea s ain ≈ 0.9
Figu e 3-26: We samples de o med a 5 GPa and 1300°C. Bo h he high s ain a e and low s ain
a e sample exhibi only one ac i e slip sys em – (100)[001].
[97]
5 GPa p essu e and 1400°C
The (100)[001] slip sys em is he only ac i e slip sys em obse ed in he specimens
de o med a 5 GPa and 1300°C i espec i e o he s ain a e a which hey we e de o med
(Fig 3-24). In his case he [001] axis poles appea in a gi dle pa e n which likely esul s
om a componen o comp essi e s ain o he powde ed agg ega e sample. Comp ession
causes alignmen in he [001] di ec ion bu as he e is no unique shea di ec ion a gi dle
de elops. De o ma ion expe imen s on powde ed agg ega es o en esul in a non-ze o
con ibu ion om comp essi e s ain o he o e all de o ma ion du ing compac ion o he
sample ma e ial.
A 5 GPa and sligh ly highe empe a u e o 1400°C he slowe s ain a e expe imen
exhibi s he same ab ic as ha obse ed a lowe empe a u e i.e. (100)[001]. A he same
condi ions bu a highe s ain a e he ab ic o he eco e ed sample is again domina ed
by he (100)[001] slip sys em (Fig 3-25). In addi ion, howe e , a weake ab ic esul ing
om ac i i y o he (010)[001] slip sys em is also p esen . While he (100)[001] slip
S ain- a e: 5x10-5; Wa e con en : 168 w . ppm; No. O g ains: 4670; Shea s ain ≈ 2.4
S ain- a e: 5x10-4; Wa e con en : 189 w . ppm;
No. O g ains: 3920;
Shea s ain ≈ 1.0
Figu e 3-27: We samples de o med a 5 GPa and 1400°C. Sample de o med a lowe s ain a e
(Top) shows mainly one ac i e slip sys ems – (100)[001]. Whe eas, (Bo om) Sample de o med unde
highe s ain has wo (010)[001] and (100)[001] slip sys ems ac i e.
[98]
sys em has been p e iously obse ed o be ac i e a high H2O con en , he (010)[001] slip
sys em has been p e iously documen ed unde highe s ess condi ions i espec i e o he
H2O con en . (Ka a o 1995; Ka a o e al. 2008; Ka ayama e al. 2008) epo he (010)[001]
slip sys em as dominan o e a ange o H2O con en s a s esses o e 300 MPa, om
expe imen s pe o med a p essu es <2.2 GPa. This would again be consis en wi h he
expe imen s pe o med a hese condi ions whe e e idence o he (010)[001] slip sys em
appea s in he expe imen pe o med wi h a as e s ain a e and he e o e unde highe
s esses.
8.5 GPa p essu e and 1300°C
Fab ics de eloped in samples de o med a 8.5 GPa and 1300°C (Fig 3-26) a e in gene al
e y simila o hose ound in he we samples a 3 and 5 GPa. The sample a low s ain a e
shows e idence o he dominan slip sys em being (100)[001], howe e ei he (010)[100]
o (010)[001] o bo h may also be ac i e, albei wi h much lowe ac i i y han he
(100)[001] slip sys em.
S ain- a e: 2.5x10-5; Wa e con en : 419 w . ppm; No. O g ains: 3850; Shea s ain ≈ 2.4
S ain- a e: 5x10
-4
; Wa e con en : 461 w . ppm; No. O g ains: 4350; Shea s ain ≈ 1.0
Figu e 3-28: We samples de o med a 8.5 GPa and 1300°C. I espec i e o he s ain a e, bo h
he specimens de o med a 8.5 GPa and 1300°C show wo ac i e slip sys ems –(010)[100] and
(100)[001]. This obse a ion is consis en wi h ac i i y o (010)[001] slip sys em a ela i ely highe
s esses and (100)[001] slip sys em unde hyd ous condi ion.
[99]
Sample DD456, which was de o med a he same p essu e and empe a u e condi ions
bu a a highe s ain a e, has also de eloped a ab ic domina ed by he (100)[001] slip
sys em bu again a weak con ibu ion om he (010)[001] slip sys em also seems o be
p esen . This wo-slip sys em combina ion appea s consis en wi h disloca ion ac i i y
unde highe s esses and hyd ous condi ion as also p oposed in p e ious s udies [Ka a o,
1995; Ka a o e al., 2008; Ka ayama and Ka a o, 2008]. TEM obse a ion on he sample
DD456 con i ms he p esence o hese wo slip sys ems (See sec ion 3.1.2).
8.5 GPa p essu e and 1500°C
Unlike he pole igu es o d y expe imen s de o med a 8.5 GPa and 1500°C, whe e no
app eciable LPO was de ec ed, mos likely as a esul o de o ma ion occu ing in he
di usion c eep egime and likely assis ed by g ain bounda y sliding, we expe imen s ha e
de eloped s ong ab ics unde bo h slow and as s ain a es.
S ain- a e: 1.5x10-4; Wa e con en : 401 w . ppm; No. O g ains: 1600; Shea s ain ≈ 1.1
S ain- a e: 6x10-4; Wa e con en : 340 w . ppm; No. O g ains: 1920; Shea s ain ≈ 1.3
Figu e 3-29
: We samples de o med a 5 GPa and 1500°C. Sample de o med a lowe s ain a e
(Top) shows wo ac i e slip sys ems – (010)[100] and (100)[001]. (Bo om) Sample de o med unde
highe s ain shows only (100)[001] slip sys em o be ac i e.
[100]
A he ela i ely slowe s ain a e o 1.5x10-4 he (010)[001] slip sys em appea s o
ha e led o a sligh ly s onge ab ic han he sub equally p esen (100)[001] slip sys em.
A he as e s ain a e o 6x10-4, howe e , he (100)[001] slip sys em domina es wi h only
a ain indica ion ha (010)[001] slip may also be p esen .
3.5.5 TEM cha ac e iza ion
In o de o ela e he obse ed ab ic wi h he disloca ion mic os uc u e a TEM
in es iga ion was ca ied ou on sample DD456, which was de o med unde a highe
s ain- a e o 5x10-4 a 8.5 GPa and 1300°C and con ained a wa e con en o 958 w . ppm.
I was conside ed unnecessa y o examine mul iple we samples om a ious condi ions
using he TEM as EBSD measu emen s show all we samples o ha e essen ially simila LPO
ab ic and he e o e should all show e idence o disloca ions wi h [001] Bu ge s ec o s
slipping on ei he he (010) o (100) planes. The LPO o sample DD456 is ypical o his
we ab ic, which is domina ed by he (100)[001] slip sys em bu wi h e idence o weake
ac i i y o he (010)[001] slip sys em.
TEM mic og aphs o DD456 show e idence o c-disloca ions wi h slip planes being
(010) and (100). Edge segmen s o he (010)[001] disloca ions a e mo e mobile whe eas
o he (100)[001] slip sys em disloca ions, sc ew segmen s a e mo e mobile. C oss-slip is
an ac i e p ocess as ma ked by he poin e 1 in he op image (Fig 3-28) and some o he
edge (100)[001] disloca ions a e kinked (poin e 2 in Fig 3-28). E idence o c oss-slip can
also be seen in he bo om-le image as indica ed by he whi e a ows. The bo om- igh
image shows long and s aigh sc ew disloca ions om (010)[001] slip sys em. The e is
ewe sub g ain bounda ies obse ed in compa ison o samples de o med unde d y
condi ions. P esence o long and s aigh sc ew segmen s could also indica e esis ance o
he glide o disloca ion.
Al hough po en ially also showing he ops o disloca ion loops, he s uc u es in he
HRTEM image shown in Fig 3-29 seem o be mo e consis en wi h being he co es o a
weakly dissocia ed c-disloca ions. The Fas Fou ie ans o med image (Fig 3-29: lowe
igh ) o he dissocia ed c-edge disloca ion when iewed along he {110} zone axis, shows a

[101]
a ia ion in con as . The image con as in he disloca ion co e egions is di e en om
ha in he su ounding bulk, indica ing he co e is expanded.
Figu e 3-30: TEM mic og aphs o he we specimen DD456. De o ma ion expe imen was ca ied
ou a 8.5 GPa and 1300°C wi h a s ain a e o 5x10-4. Top igu e shows he p esence o c-disloca ions.
(100)[001] disloca ions a e mos ly o edge na u e whe e as he [001] sc ew disloca ions a e mos
likely om (010)[001] disloca ion. E idence o c oss-slip can be also seen as indica ed by ma ke 1 in
op image and whi e a ow in he bo om-le image. Bo om- igh igu e also shows s aigh c-sc ew
disloca ion om (010)[001] slip sys em.
[102]
Figu e 3-31: A ypical HRTEM image (uppe and lowe igh ) and he Fas Fou ie ans o med
image (lowe igh ) o he dissocia ed c-edge disloca ion iewing along he {110} zone axis o a
de o med hyd ous oli ine. The image con as in he disloca ion co e egions is di e en om ha in
he su ounding bulk, which indica es ha he co e is expanded.
[103]
3.6 De o ma ion expe imen on Pe ido i e modal composi ion
De o ma ion expe imen s pe o med we e pe o med on a pe ido i e assemblage a 8.5
GPa and 1300°C. In he low s ain a e sample, he likely ac i e slip sys ems a e (010)[100]
and (010)[001] whe eas he dominan slip sys em in oli ine in he high s ain a e sample
is (010)[001]. In case o py oxene assemblage (100)[001] slip sys em is he only likely
ac i e slip sys em in bo h high and low s ain a e sample (Table 3-8).
Table 3-8: Expe imen al condi ions o Pe ido i e de o ma ion expe imen s and likely ac i e slip
sys ems
Run ID
P essu e
(GPa)
Tempe a u e
(°C)
S ain a e
(x10
-5
s
-1
)
S ess
(MPa)
Likely Ac i e slip sys ems
DD495
8.5
1300
5
230
Oli ine: (010)[100] and (010)[001]
Py oxene : (100)[001]
DD483
8.5
1300
40
325
Oli ine: (010)[001]
Py oxene : (100)[001]
S ain- a e: 5x10
-5
; Shea s ain ≈ 1.6
S ain- a e: 50x10-5; Shea s ain ≈ 1.3
Figu e 3-32: Pe ido i e samples de o med a 8.5GPa and 1300°C. Oli ine in he slowly de o med agg ega e
likely has bo h (010)[100] and (010)[001] slip sys ems ac i e whe eas in he expe imen conduc ed a highe
s ain a e he slip sys em is (010)[001]. Py oxene in bo h he cases show (100)[001] slip sys em
[104]
These esul s indica e an iden ical oli ine ab ics o hose obse ed in monomine alic
expe imen s a he same condi ions. Fab ics o diopside and ens a i e we e ound o be
simila o hose ound in p e iously pe o med lowe p essu e expe imen s.
In-si u measu emen o s ess using piezoelec ic senso
Piezoelec ic s ess measu emen s we e made using GaPO4 single c ys als wi hin a 8/6
mul ian il assembly. Many es expe imen s we e made in he D-DIA howe e he mos
success ul uns, whe e d i was minimal, we e pe o med using he 6-axis MAVO p ess.
The success o he MAVO p ess in hese expe imen s mos likely o igina es om he highly
esis i e elec ical insula ion o each indi idual an il. In he D-DIA p ess he op and
bo om guide blocks a e elec ically connec ed ia he oil lines o he de o ma ion ams.
Figu e 3-33: Ou pu ol age om
he cha ge ampli ie as a unc ion o
ime o an expe imen whe e a GaPO4
c ys al was comp essed o 2 GPa and
hen held a cons an s a ic p essu e
o 80 min.
The an ils he e o e canno be
used as pa o he ci cui and
c ys als mus , he e o e, be
connec ed ia sepa a e cables ha
pass h ough he gaske . The use o
he cables ha ha e o pass
h ough he gaske s ha o m as
he cubic assembly is comp essed p obably esul s in cu en leakage.
Figu e 3-33 shows he ou pu ol age om he cha ge ampli ie o an expe imen
comp essed o 2 GPa, which emained unde s a ic condi ions o 80 min. This ol age
change would co espond o changes in s ess o he o de o 5 GPa o e his ime pe iod
and mo e likely esul s om d i . The o igin o his d i is unclea . D i is posi i e o he
i s 30 min bu e en ually goes h ough a maximum and dec eases. Howe e , he slope o
[111]
s ain a e (and hence highe low s ess) has a la ge p opo ion o da a poin s indica ing
he B- ype ab ic.
Figu e 4-1: Summa y o ab ics obse ed in San-ca los oli ine de o med unde d y and we
condi ion a di e en s ain a es. Expe imen s we e pe o med be ween 3 o 8.5 GPa and 1300°C o
1500°C. Wid h o each colou ba is p opo ional o he app oxima e numbe o g ains ha we e
p esen in he subse con aining da a poin s o ha slip sys em. Re e o sec ion 3.4.14 o mo e
de ails. As shown in he able a op- igh co ne o he page, he lowe ow in he 2x2 ma ix con ains
esul s om d y expe imen s while uppe ow con ains esul s om we expe imen s. The le
column in 2x2 ma ices has esul s om slowly de o med samples whe eas samples de o med a
ela i ely highe s ain a e ha e hei ab ics shown in he igh column.
Hence, i can be concluded ha unde high s ess de o ma ion condi ions he B- ype
ab ic domina es i.e. he (010)[001] slip sys em con ibu es o mos o he s ain. Al hough,

[112]
he p opo ion o da a poin s wi h he B- ype ab ic inc eases a 5 GPa and 1400°C wi h
espec o 5GPa and 1300°C , his is ac ually in line wi h a sligh ly highe s ain a e and
hence highe low s esses in specimens de o med a 1400°C.
Equally impo an is he obse a ion ha a 8.5 GPa and 1300°C, he specimen
de o med a a slowe s ain a e shows a la ge p opo ion o da a poin s wi h he A- ype
ab ic. The e ha e been epo s ha a highe p essu es c-slip, slip wi h a Bu ge s ec o
[001], becomes easie han a-slip. The app oxima e ansi ion p essu e be ween a-slip and
c-slip appa en ly a ies be ween s udies wi h Jung e al. (2009) epo ing his ansi ion a
app oxima ely 3.6 GPa whe eas Ra e on e al. (2007) place he ansi ion a app oxima ely
7.6 GPa. I he e was indeed a p essu e ela ed easing o he b = [001] slip hen i would be
ha d o econcile his wi h he esul s in Fig. 4-1, which show dominan (010)[100] slip
sys em ac i i y a slowe s ain a es a 8.5 GPa. These esul s a e a he mo e in line wi h
he obse a ion ha a change om a-slip (b = [100] slip) o c-slip (b = [001] slip) can occu
a highe s esses [Jung e al., 2006]. Ano he piece o s ong e idence ha a-slip con inues
o be he dominan slip mechanism a high p essu es, comes om analysis o specimens
ha we e ho -p essed in he 6-8 mul ian il unde pseudo-hyd os a ic condi ions i.e.
wi hou he inclusion o alumina ha d pa s in he mul ian il assembly ha can lead o
s ong de o ma ion (Fig 4-2). In wo such ho -p essing expe imen s, d y oli ine powde
was ho p essed a 1400°C a p essu es o 8.5 GPa and 11 GPa. Du ing cold comp ession o
he mul ian il assembly, some disloca ions a e in oduced in he oli ine sample. Once, he
sample is hea ed up o high empe a u es, disloca ion eco e y p ocesses s a , leading o
de o ma ion o he sample and de elopmen o a weak LPO. As seen in he pole igu es on
he nex page, he A- ype ab ic a e clea ly p esen bo h o hese specimens.
[113]
Figu e 4-2: Pole igu es o wo polyc ys alline oli ine specimen ho p essed a 8.5 GPa (H3115)
and 11 GPa (H3354). Specimens we e annealed a 1400°C. Bo h pole igu es esemble A- ype ab ic
which is o en obse ed unde low s ess and d y de o ma ion en i onmen . P esence o A- ype ab ic
in hese ho p essed specimen is indica i e o (010)[100] slip sys em ac i i y.
The LPO ab ics obse ed in Fig 4-2 could ha e only de eloped in hese samples i a-slip
was dominan . In e es ingly, he i s epo o a p essu e induced ansi ion in oli ine LPO
was by Cou y e al. (2004) om de o ma ion expe imen s pe o med a BGI using a simila
mul i-an il appa a us. The assembly se up was in simple shea con igu a ion, howe e ,
unlike ha epo ed he e whe e no delibe a e a emp was made o de o m he oli ine
sample du ing ho -p essing. The peak s ess ha would ha e de eloped du ing he ini ial
s ages o annealing mus be lowe han in he expe imen s o Cou y e al. (2004). Hence,
he ho -p essing ab ic in specimens H3115 and H3354 mus esul om de o ma ion
unde low s esses and hus i can be concluded ha a-slip con inues o be he easies slip
mechanism e en up o 11 GPa p essu e unde low s ess de o ma ion condi ions.
The obse a ions in p e ious s udies whe e a ansi ion o a-slip o c-slip has been
a ibu ed o inc easing p essu e a e mo e likely o esul om highe s esses which may
ine i ably inc ease in expe imen s a highe p essu es.. Pole igu es shown by Jung e al.
(2008) do no show s ong LPOs despi e la ge amoun s o s ain ha hei samples
[114]
expe ienced. Resul s om Ra e on e al. [2007] whe e hey obse ed a p essu e ela ed
ansi ion in he slip sys em, a e based upon expe imen s pe o med a s esses a ying
be ween 300 MPa o 1800 MPa. Simila ly in he s udy o Ohuchi e al. [2011] mos
specimens exhibi ing he p esence o c-slip ha e es ima ed s esses in excess o 350 MPa.
Only in one specimen a e s esses epo ed o ha e been less han 300 MPa bu his sample
ac ually indica es dominan ac i i y o he (100)[001] slip-sys em unde d y condi ions.
Addi ionally he obse a ion ha a ious s udies ha e placed he p essu e o he ansi ion
a widely a ying alues can also be explained i his ansi ion in ac esul s om
inc easing low s esses.
Figu e 4-3 shows he ab ics obse ed in his s udy compa ed wi h hose epo ed in
p e ious s udies. La ge symbols wi h black bounda ies a e da a poin s om his s udy.
The esul s a e in e y good ag eemen wi h a ia ions in ab ic epo ed by Ka a o e al.
[2008]. A highe s esses, unde d y condi ion he B- ype ab ic was obse ed, whe eas
unde lowe s esses and d y condi ions de o ma ion esul ed in he dominance o A- ype
ab ic. Hyd ous specimens exhibi C- ype ab ic which also ag ees wi h p e ious epo s
om lowe p essu e de o ma ion s udies. None o he specimens cha ac e ized in his
s udy showed e idence o E- ype o D- ype ab ic because expe imen al condi ions in we e
ne e en e ed hese egions o s ess-wa e con en space. B oken g ey lines in he image
a e he likely ansi ion bounda ies be ween wo di e en ab ic ypes.
The main conclusion o he esul s ob ained om de o ma ion expe imen s unde d y
condi ion is ha p essu e appa en ly plays no di ec ole in he slip sys em ansi ion in
oli ine. (010)[001] slip sys em con ibu es mos o he o e all s ain a highe s esses and
he B- ype ab ic should be obse ed unde such en i onmen . A- ype ab ic should be he
mos dominan ype ab ic ha we would expec o see in he li hosphe ic man le because
s esses a e expec ed o be oo low (<10 MPa) o gene a e B- ype ab ic. Howe e , in he
man le wedge nea subduc ing slabs i is possible ha egions exis whe e B- ype ab ic
could occu as hese egions a e mos likely de o ming a ela i ely high s esses.
[115]
Figu e 4-3: De o ma ion da a om his s udy and o he s udies a e shown as a unc ion o s ess and wa e
con en s (T
∼
1470–1670 K). La ge symbols wi h black bounda ies ep esen da a om his s udy whe eas
es o da a a e om Ka ayama e al. 2004. Excep , one o he da a o D- ype ab ic is om Bys icky e al.
(2001). Wa e con en was es ima ed using he Pa e son (1982) calib a ion. B oken g ay lines indica e he
likely ansi ion line be ween wo di e en ab ic ypes (Modi ied a e Ka a o e al., 2008)
[116]
Lack o LPO in he d y samples de o med a high empe a u e a 1500°C and 8.5 GPa
has likely esul ed om di usion accommoda ed g ain bounda y sliding. High g ain
g ow h, as e iden om g ain size measu emen s made in eco e ed expe imen s, poin s
o he dominance o di usion c eep. The p esence o s aigh g ain bounda ies indica e ha
condi ions we e sui able o g ain bounda y sliding. The e we e many ou -g ain junc ions
p esen in hese samples and hese junc ions we e sligh ly diamond shaped which is
addi ional e idence ha g ain bounda y sliding may ha e been he ac i e de o ma ion
p ocess. I is well known ha di usion c eep does no esul in LPO de elopmen .
Mo eo e , g ain bounda y sliding (GBS) also does no a ou LPO de elopmen because in
he p esence o GBS, in ac ys alline de o ma ion is only weakly dependen on he
o ien a ion o g ains [D u y and Humph eys, 1988; Ka a o e al., 1986; Zhang e al., 1994].
Fab ic ypes obse ed in he oli ine de o med as a pa o he pe ido i e modal
composi ion a e iden ical o hose obse ed in monomine alic expe imen s a he same
condi ions. The dominan slip sys em in he py oxene componen o his agg ega e had
mos likely (100)[001] slip sys em ac i e.
4.2 Fab ic ypes unde wa e ich condi ions
The hyd ous specimens in his s udy all con ain wa e con en s in excess o 50 ppmw.
F om igu e 4-3, i can be seen ha he C- ype ab ic is he mos commonly obse ed ab ic
in hese specimens. Howe e , some o he high s ess wa e - ich samples also indica e he
p esence o he B- ype ab ic e.g. DD456. A TEM s udy on one o he we samples (DD456)
indica es he dominan p esence o b = [001] disloca ions. The ac i e slip sys ems in his
specimen we e obse ed h ough TEM o be (100)[001] and (010)[001]. The ac i i y o he
(010)[001] slip sys em in his sample is simila o ac i i y obse ed in he d y specimens
de o med unde highe s esses. The (100)[001] slip sys em is an unusual slip sys em
because glide on he (100) plane mos likely in ol es b eaking o SiO4- e ahed a, unlike
glide on (010) whe e no SiO4 e ahed a a e encoun e ed du ing glide. Disloca ion
mic os uc u es in he we sample (DD456) obse ed using he TEM show ha disloca ion
loops on he (010) glide plane ha e longe sc ew segmen s han hei edge componen s.
The e a e long edge segmen s o c-disloca ions isible and mos o hem a e pa o c-

[117]
disloca ions on he (100) glide plane. Whe e only sc ew segmen s o c-disloca ions a e
isible i is unclea whe he glide is also occu ing on he (100) plane because
de e mina ion o he glide plane is no possible when only sc ew segmen s a e isible. The
abundan numbe o c-disloca ion on he (010) slide plane wi h a sc ew na u e is in clea
con as o he c-disloca ions obse ed in he d y specimens which we e mos ly o edge
na u e unde simila P-T-s ess condi ions. Ano he pe inen obse a ion in he we
sample is he e idence o climb o edge disloca ions and c oss-slip o sc ew disloca ions. A
ew p isma ic loops lying in he (010) plane can also be obse ed, which ha e been
associa ed wi h he p esence o H2O in oli ine.
A u he in e es ing obse a ion is ha he ansi ion line be ween C- ype and B- ype
ab ics appea s o ha e a posi i e slope wi h espec o he oli ine H2O con en . This is in
con as o he p e ious s udy o Ka ayama e al. 2004 whe e a nega i e slope was
p oposed. In acco dance wi h Fig 4-3 i would equi e inc easingly highe s esses o push
oli ine om he C- ype ab ic egime in o he B- ype egime wi h inc easing wa e con en
o oli ine. I would also imply ha wa e a ou s slip on he (100) plane o e he (010)
plane.
4.3 Physical basis o slip sys em changes in oli ine
In he ollowing sec ion an a emp is made o iden i y a physical basis o changes in
slip sys ems as a unc ion o s ess and wa e con en . As men ioned p e iously, unde high
s ess de o ma ion condi ions de o ma ion by b = [001] slip becomes dominan and he slip
sys em changes om (010)[100] o (010)[001]. Slip wi h b = [001] also becomes dominan
in he p esence o wa e . In his case, he dominan glide plane becomes (100) which a e
no known o be an easy glide plane because o way he SiO4 e ahed a a e a anged in
oli ine. Based upon he heological and mic os uc u al da a a ailable om p e ious
s udies along wi h esul s om his s udy, an a emp is made he e o explain he
dominance o he (010)[001] slip sys em unde highe s esses and he cause o C- ype
ab ic de elopmen in we specimens due o he likely ac ion o he (100)[001] slip sys em.
[118]
4.3.1 Dominance o (010)[001] slip sys em a highe s esses
S udies on heology o oli ine single c ys als poin owa ds a empe a u e ela ed
ansi ion in he oli ine slip sys em (Fig 4-4).
Measu emen s o he c i ical esol ed shea s esses o (010)[100] and (010)[001] slip
sys ems indica es ha CRSS o he (010)[001] slip sys em is lowe han he CRSS o he
(010)[100] slip sys em a lowe empe a u e. Whe eas, he e e se is obse ed a highe
empe a u es whe e he (010)[100] slip sys em is ini ia ed a much lowe s esses. This
ansi ion be ween CRSS alues o he wo slip sys ems occu s a a ound 1200°C a a
co esponding alue o 400 MPa CRSS.
Figu e 4-4: C i ical esol ed shea s esses (CRSS) o he (010)[100] and (010)[001] slip sys ems
as a unc ion o empe a u e. Da a (co esponding o a s ain a e o 10-5 s-1) om expe imen s
pe o med on single c ys als o ien ed along [011]c (black- illed symbols) o p omo e [001](010) glide
and along [110]c (open symbols) o p omo e [100](010) glide. (Sou ce- PhD hesis – Helen Cou y,
2005)
In e es ingly, in his s udy he ab ic ansi ion om A- ype o B- ype, which is ela ed
o he ansi ion be ween he dominan slip sys ems (010)[100] and (010)[001], was also
obse ed a a ound 300 MPa, which gi en he unce ain y in bo h de e mina ions pu s i in
close p oximi y o he c oss-o e in he CRSS o he wo slip sys ems shown in Fig 5-4. In
he ollowing sec ion, i is a gued ha he s ess ela ed ansi ion in he ab ics is a logical
[119]
consequence o changes in he CRSS o he wo slip sys ems wi h empe a u e. Be o e his
a gumen is pu sued, howe e , i is wo hwhile o e iew why CRSS changes wi h s ess.
The empe a u e dependence o CRSS has been obse ed in me als and a ious ionic
and co alen compounds [Cas aing e al., 1981]. Also, di e en slip planes can ha e
di e en empe a u e dependency o CRSS as is shown o α-Al2O3 in Fig 4-5.
Figu e 4-5: Tempe a u e dependence o he c i ical shea s ess τc(T) o co alen c ys als
measu ed unde high o a mosphe ic p essu e. The da a a e aken om he e e ences: Lage lo e al.
(1994) o α-Al2O3, Cas aing e al. (1981b) o Si and Boi in e al. (1990) o GaAs o in insic and p-
ype. (Figu e sou ce: Koizumi e al., 1994)
This empe a u e dependence o he CRSS alue is gene ally explained on he basis o
in insic esis ance o disloca ion mo ion by he p ocess known as he “Peie ls p ocess”. I
has been sugges ed ha he Peie ls p ocess becomes impo an o he heology o oli ine
a highe s esses. Powe law c eep p o ides a good desc ip ion o oli ine low only a
lowe s esses, whe eas a highe s esses he oli ine low law becomes exponen ial
[Goe ze, 1978; Ka ayama and Ka a o, 2006]. An exponen ial low law can be p edic ed
based upon disloca ion mo ion by he Peie ls mechanism ia nuclea ion o double kinks.
Disloca ion mo ion by he Peie ls mechanism may be mo e impo an o b = [001] because
[120]
disloca ion lines end o be s aigh e o c-disloca ions [Ka a o e al., 2008; Phakey e al.,
1971]. Al hough, he alue o c i ical s ess a which he Peie ls p ocess s a s o domina e
he low beha iou is no ye clea .
Peie ls mechanism o de o ma ion by double kink nuclea ion
A ini e empe a u es, disloca ions do no mo e all a once in a plane s ain manne bu
mo ion occu s h ough he gene a ion and nuclea ion o kink-pai s. In his egime,
nuclea ion o kink-pai s is he a e de e mining s ep o slip. Once a kink g ows abo e a
c i ical size, u he s aining akes place by mig a ion o hese kinks.
A kink on a disloca ion line can be en isaged as a disloca ion on a disloca ion line (Fig 4-
6). Acco ding o he Peie ls mechanism, a s aigh disloca ion line has i s lowes ene gy
when i lies in a po en ial alley pa allel o lines o closes packing o a oms on he slip
plane. Du ing he mo ion o a s aigh disloca ion om one alley owa ds he nex , he
a oms in he icini y o he co e o he disloca ion change hei posi ions and bond angles,
causing he ene gy o he disloca ions o inc ease. Midway be ween wo adjacen alleys,
he disloca ion ene gy eaches a maximum alue and any addi ional displacemen will
cause he disloca ion o all down he ene gy maxima in o he nex alley. The maximum
shea s ess necessa y o p omo e such o wa d mo ion o he disloca ion is known as he
Peie ls s ess 𝜏𝑝.
Figu e 4-6: Le image shows a kink (da k line) lying ac oss a po en ial alley. B oken lines
indica e he po en ial maxima wi h minima ep esen ed by he solid lines. Righ : A kink in he
p esence o ex e nal s ess has i s equilib ium posi ion displaced away om he uns essed posi ion.
Size o he kink is ep esen ed by kink heigh h and wid h 2K+L in case o a apezoidal kink model.
(Sou ce: Suzuki e al. 1995)
[127]
segmen s unlike in d y specimens. I is a known ac ha mo ion o sc ew disloca ion is
conse a i e and i gene ally does no in ol e di usion [Naba o, 1967].
Glide on (100) plane in hyd ous specimens
As men ioned ea lie (100) plane is no an easily explicable glide o disloca ions in
oli ine because i would appea o in ol e b eaking o Si-O bonds p esen in he SiO4
e ahed a (Fig 4-10). The e a e models o glide on he (100) plane which in ol e
signi ican con ibu ion om disloca ion climb. A e obse ing equen occu ence o
(100)[001] slip sys em in na u al oli ine Olsen and Bi kelan.T [1973] p oposed ha such a
glide may be possible by pe iodic occu ence o jogs on he disloca ion line shown as he
b oken line in he igu e 4-10.
Figu e 4-10: (001) p ojec ion o he oli ine s uc u e. Only he oxygen ions a e shown, bu he
posi ions o he silicon ions a e indica ed by he Si 04 e ahed a. Pe iodic jogs in a (100) plane a e
indica ed by he b oken line. The a om posi ions a e hose o he pape by Hanke (1965). (Figu e
sou ce: Olsen and Bi keland, 1973)
Figu e 4-10 shows oli ine (001) plane pa allel o he plane o pape . I is clea ha
(010) is easie because disloca ion line gliding on his plane will expe ience no obs uc ion
om SiO4 e ahed on whe eas glide on (100) plane would be ex emely di icul . Howe e ,

[128]
model p oposed by Olsen and Bi kelan.T [1973] canno explain de elopmen o CPO in
oli ine. Acco ding o hei model a io o s ain by glide o s ain by climb 𝑙𝑔𝑙𝑐
⁄< 1
because o e e y glide s ep by a segmen o disloca ion, i needs o climb by minimum o
one climb s ep, hence he maximum 𝑙𝑔𝑙𝑐
⁄ a io in his case would he
𝑏[100]𝑏[001]= 4.76 5.99
⁄⁄ ≈0.80. This implies ha a signi ican pa o he s ain would
be accommoda ed by disloca ion climb alone. This obse a ion becomes e en mo e
pe inen because dynamic ec ys alliza ion and g ain g ow h a e e y ac i e in we
specimens and he o me is known o andomize he CPO. So, in o de o p oduce a
pe cep ible CPO, he a io 𝑙𝑔𝑙𝑐
⁄ should be conside ably highe han 1. This can happen only
i disloca ion segmen s a e ac ually able o slice h ough he SiO4 e ahed a. Ano he
ele an obse a ion in his ega d is he de o ma ion s udies on oli ine single c ys al by
Du ham and Goe ze [1977] whe e hey ha e epo ed only 20-30% s ain by disloca ion
climb.
Figu e 4-11: FTIR spec a o hyd ous samples show peaks a 3477, 3448, 3629 and 3676 cm-1.
These peaks could be a ising om hyd ogen associa ed wi h acan Silicon si es.
[129]
As b eaking o Si-O co alen bonds is ene ge ically un a ou able, he only easonable
way o achie e his glide may be o eplace some o he Si-O bonds by weak hyd ogen
bonds. Al hough mos o he wa e in hyd ous oli ine is known o be associa ed wi h
di alen me al acancies, i has been a gued ha some OH- is accommoda ed by cha ge
balancing h ough he c ea ion o acancies a Si si es [B ai hwai e e al., 2003; B odhol
and Re son, 2000]. Theo e ical modelling on o s e i e has shown ha he hyd oga ne
de ec , 4𝐻𝑆𝑖
𝑋 should p oduce IR peaks a ound 3425, 3448, and 3478cm-1 [B ai hwai e e al.,
2003]. Whe eas IR peaks a 3674 and 3624 cm-1 could also be band double s ela ed o Si
acancies [Libowi zky and Be an, 1995]. B oad peaks a 3448 and 3478cm-1 p esen in he
hyd ous specimens s udied he e migh esul om hyd oga ne subs i u ion and peaks a
3676 and 3629 cm-1 a e also p esen in hese specimens (Fig. 4-11). I hese peaks indeed
esul om OH bonds associa ed wi h silicon acancy si es and hus a small o wa e is
dissol ed a he silicon si e, hen such si es could ac as p e e ed loca ions o disloca ion
nuclea ion in (100)[001] slip sys em. Also, in such a scena io glide on he (100) plane
would become he p e e ed glide plane because i is in ac he denses close packed plane
o he oli ine oxygen anion sub-la ice. This explana ion is also in line wi h he obse a ion
ha wi h inc easing wa e con en , glide on (010) plane becomes inc easing di icul as
e iden om he posi i e slope o he ansi ion bounda y be ween C- ype and B- ype
ab ic (Fig. 4-3).
[130]
4.4 Viscoplas ic sel consis en modelling o ab ic de elopmen in oli ine
Viscoplas ic sel consis en modelling is a use ul ool o modelling ab ic de elopmen
in mine al phases and has been used pa icula ly o model oli ine ab ics [Tommasi e al.,
2000]. Inpu pa ame e s in he model a e ela i e slip sys em CRSS alues o he mine al
phase conce ned, hei elas ic cons an s and s ess exponen “n”. Based upon hese inpu
pa ame e s, he modelling p og amme can p edic he CPO de elopmen in a ious
de o ma ion geome ies. O he use ul in o ma ion ob ainable om his modelling ool is
he na u e o in e ac ion be ween di e en slip sys ems and hei ela i e con ibu ion o
he o al s ain in he sample. Fu he mo e, he ab ic da a ob ained om hese models can
be employed o p edic ing seismic aniso opy esul ing om ac i i y o he chosen slip
sys ems.
The c ucial aspec o his kind o modelling is he choice o ela i e CRSS alues. The
in o ma ion abou CRSS alues can be es ima ed om he obse a ion o CPO and he
na u e o disloca ions in na u ally o expe imen ally ob ained specimens by using a ious
ools such as EBSD and TEM.
To pe o m his kind o modelling ini ial guesses a e made o he ela i e CRSS alues
o di e en slip sys ems and hen hese choices can be imp o ed u he based upon he
p edic ed ac i i y o di e en slip sys ems by he p og am. Simila i y o he isual
appea ance and s eng h o he p edic ed pole igu e wi h he expe imen ally de e mined
pole igu es p o ides one way o cons ain he co ec choices o he ela i e CRSS alue.
This kind o modelling also p o ides in o ma ion on geome ical cons ain s in he ac i i y
o he p obable slip sys ems. In ce ain cases an easy slip sys em may be p e en ed om
making signi ican con ibu ion o he o e all s ain because o geome ical cons ain s
imposed by he choice o ac i e slip sys ems. Fo example, ac i i y o (001)[100] slip
sys em is supp essed in he p esence o (010)[001] slip sys em [Tommasi e al., 2000].
Howe e , i mus be bo ne in mind ha his kind o modelling does no ake in o
accoun he e ec o de o ma ion p ocesses o he han disloca ion glide. Dynamic
ec ys alliza ion and o he di usi e p ocesses may s abilize he ac i i y o some slip
sys ems a le els which a e ei he highe o lowe han hose p edic ed by he p og amme.
[131]
4.4.1 Modelling he pole ab ic o d y specimen DD455
Specimen DD455 was de o med unde d y condi ion a 8.5 GPa, 1300°C a slow s ain
a e o 2.5x10-5 s-1. Ac i e slip sys ems as obse ed unde TEM a e (010)[100], (010)[001]
and (100)[001]. Pole igu e o he specimens shows he p esence o A- ype and B- ype
ab ic. A mino componen o C- ype could also be seen.
Figu e 4-12: D y samples de o med a 8.5 GPa and 1300°C. Sample de o med a slowe s ain a e
shows dominan slip sys em o be (𝟎𝟏𝟎)[𝟏𝟎𝟎] and (𝟎𝟏𝟎)[𝟎𝟎𝟏].
Table 4-3: Choice o ela i e CRSS alues used o a ious models in o de o syn he ically
gene a e he pole igu e o specimen DD455
Models
(010)[100]
(010)[001]
(100)[001]
(001)[100]
Model 1
1
1
1
1
Model 2
1
1
1
5
Model 3
1
1
5
1
Model 4
1
3
1
4
Model 5
1
3
1.5
5
Model 6
1
3
1
3
In o de o syn he ically gene a e a pole igu e simila o wha is obse ed in he
specimen DD455 (Fig 4-12); a numbe o models we e chosen. S a ing by assigning all he
slip sys ems an equal alue o CRSS, he ela i e CRSS alues a e hen i e a ed un il a ai
simila i y is achie ed be ween he obse ed pole igu e in he specimen and he syn he ic
pole igu e.
The i s model, which assumes all he slip sys ems o ha e simila CRSS alue, clea ly
ails o ec ea e he pole igu e o he specimen DD455. By supp essing he slip sys em
(001)[100] wi h a choice o highe CRSS in model 2, because he ele an ab ic gene a ed
[132]
by his slip sys em shows e y weak con ibu ion in he o iginal pole igu e o DD455, a
close simila i y in he isual appea ance o he simula ed and measu ed pole igu es is
achie ed. Model 3 is again e y di e en om he o iginal pole igu e. On he o he hand,
models 4, 5 and 6 which assume (010)[100] and (010)[001] o ha e e y simila CRSS wi h
la e being sligh ly mo e di icul han he o me , p oduces a much be e ma ch in he
isual appea ance and ela i e concen a ion o di e en c ys allog aphic di ec ions. The
las 3 models also assume ha (100)[001] and (001)[100] slip sys ems a e a leas h ee
imes mo e di icul han he slip sys ems wi h (010) glide plane.
When he bes i models a e examined hey all iden i y he (010)[100] slip sys em as
he easies slip sys em a 8.5 GPa and 1300°C unde d y condi ions. Two o he success ul
models also iden i y he (100)[001] slip sys em as being o compa able s eng h o
(010)[100],howe e , any model ha impose his slip sys em o be weake han (010)[100]
esul s in a poo ma ch wi h he expe imen al pole igu e. This obse a ion is in line wi h
ou o he obse a ions ha he e is no app eciable ha dening o a-slip wi h p essu e up o
11 GPa.
Figu e 4-14 shows ha ac i i y o a ious slip sys ems may a y wi h changes in s ain
o main ain he s ain homogenei y in he agg ega e wi hou he need o g ain bounda y
mig a ion o o he seconda y s ain gene a ing p ocesses. In his sense, models 4 and 5 a e
e y s able and his is addi ional e idence ha ela i e CRSS choices ep esen ing models 4
and 5 a e ep esen a i e o he slip sys ems ac i e in specimen DD455.

[133]
Model 1
Model 2
Model 3
Model 4
Model 5
Model 6
Figu e 4-13: Pole igu es o models desc ibed in he able 5-3. Models which assume e y simila
CRSS alue o (010)[100] and (010)[001] and a leas h ee imes highe CRSS alue o o he wo slip
sys ems can mimic he expe imen al pole igu e.
[134]
Model 1
Model 2
Model 3
Model 4
Model 5
Model 6
Figu e 4-14: No malized ac i i y e sus equi alen s ain plo o a ious model. Model 1 o 6 is
shown he e. Ac i i y o slip sys ems can change wi h inc easing s ain because o geome ical
cons ain s. In his sense, model 4 and 5 appea e y s able
4.4.2 Modelling he pole ab ic o d y specimen DD456
Specimen DD456 was de o med unde we condi ion a 8.5 GPa, 1300°C a a ela i ely
as s ain a e o 5x10-5 s-1. Ac i e slip sys ems as obse ed unde TEM a e (010)[001] and
(100)[001]. The pole igu e o he specimens shows he p esence o C- ype and B- ype
ab ic.
[135]
Figu e 4-15: We samples de o med a 8.5 GPa and 1300°C. The specimens shows wo likely ac i e
slip sys ems – (010)[100] and (100)[001] which has also been con i med by TEM s udy on his sample.
Table 4-4: Choice o ela i e CRSS alues used o a ious models in o de o syn he ically
gene a e he pole igu e o specimen DD456
Models
(010)[100]
(010)[001]
(100)[001]
(001)[100]
Model 1
10
1
1
10
Model 2
10
2
1
10
Model 1
Model 2
Figu e 4-16: Pole igu es o models desc ibed in he able 5-4. Models which assume (100)[001]
o be he easies and (010)[100] as sligh ly highe han he o me along wi h e y high alue o CRSS
o (010)[100] and (001)[100] i.e. o a-slip can ep oduce well he pole igu e o he specimen
DD456.
Models ( able 4-4) ha assume he (010)[001] sys em o be he easies wi h (010)[001]
being sligh ly ha de can ec ea e he pole igu e o DD456 (Fig 4-16). These models also
equi e ha he o he wo slip sys ems in ol ing a-slip i.e. (010)[100] and (001)[100]
should be conside ably ha d o p e en he alignmen o oli ine [100] axes along he shea
di ec ion.
[136]
Seismic aniso opy in he uppe man le – Implica ions om his s udy
C ys allog aphic p e e ed o ien a ion o majo man le mine al phases e.g. oli ine,
py oxene and ga ne is belie ed o be he majo cause o seismic aniso opy in he uppe
man le. Seismic aniso opy by CPO de elopmen is he esul o in insic seismic aniso opy
o hese c ys als.
Composi ionally, man le pe ido i e consis s o up o 60% oli ine, o o hopy oxene,
clinopy oxene, ∼0%–20% o ga ne , and spinel depending on he dep h (e.g., Ringwood
[1975]. Oli ine, being he composi ionally dominan and mechanically weakes phase in
he pe ido i e, dic a es he o e all aniso opy o he agg ega e and con ibu ion o
py oxenes esul s in sligh dilu ion o he o e all P-wa e and S-wa e aniso opy [Blackman
e al., 2002; Mainp ice e al., 2000; Mainp ice e al., 2005]. Whe eas, ga ne is known o
de elop only weak LPO in he p esence o o he weake phases e.g. oli ine and py oxenes
[Mainp ice e al., 2004].
The gene al cha ac e is ic o Seismic aniso opy esul ing om LPO o oli ine, as seen
in he na u al samples, is ha he as es S-wa e pola iza ion di ec ion lies (sub-)pa allel o
he olia ion plane. The maximum shea wa e spli ing is obse ed no mal o he linea ion
di ec ion in he olia ion plane. P-wa e eloci y is he as es along he oli ine a-axis.
Symme y o S-wa e aniso opy is in luenced by bo h [100] and [001] axis and P-wa e
aniso opy is mainly dependen upon [100] axis. O e all magni ude o he seismic
aniso opy is dependen upon he o ien a ions o all h ee oli ine axis [Ben Ismail and
Mainp ice, 1998]. Magni ude o Seismic aniso opy inc eases he ab ic s eng h bu does
no inc ease beyond 20% o P-wa es and 15% o S-wa es [Ben Ismail and Mainp ice,
1998]. P-wa e p opaga ion di ec ion is also as es pa allel o i s [100] axis like in oli ine.
Howe e , unlike in oli ine whe e he maximum spli ing di ec ions occu s a an angle o
15 −20° om he c-axis di ec ion, pola iza ion di ec ion o he as es s-wa e in ens a i e
lies a ~25° om he b-axis [Blackman e al., 2002].
In case o oli ine, he a io 𝑉𝑆𝐻 𝑉𝑆𝑉
⁄ a ies om ab ic o ab ic. In gene al, A- ype
ab ics a e known o p oduce s onge 𝑉𝑆𝐻 𝑉𝑆𝑉
⁄ a io han E- ype ab ic wi h a io being