High pressure kma

This content has been downloaded from IOPscience. Please scroll down to see the full text.

Download details:

IP Address: 201.9.158.81

This content was downloaded on 21/03/2014 at 08:16

Please note that terms and conditions apply.

Development of the Kawai-type Multi-anvil Apparatus (KMA) and Its Application to High

Pressure Earth Science

View the table of contents for this issue, or go to the journal homepage for more

2012 J. Phys.: Conf. Ser. 377 012001

(http://iopscience.iop.org/1742-6596/377/1/012001)

Home Search Collections Journals About Contact us My IOPscience

iopscience.iop.org/page/terms

http://iopscience.iop.org/1742-6596/377/1

http://iopscience.iop.org/1742-6596

http://iopscience.iop.org/

http://iopscience.iop.org/search

http://iopscience.iop.org/collections

http://iopscience.iop.org/journals

http://iopscience.iop.org/page/aboutioppublishing

http://iopscience.iop.org/contact

http://iopscience.iop.org/myiopscience

Development of the Kawai-type Multi-anvil Apparatus (KMA) and Its Application to High Pressure Earth Science

E Ito

Institute for Study of the Earth’s Interior Okayama University Misasa Tottori-ken 682-0193 Japan

E-mail: eiito645@gmail. com

Abstract. Since Birch’s prediction on the structure of the Earth’s interior high pressure Earth science has rapidly been grown by the Kawai-type multi-anvil apparatus (KMA) and the diamond anvil cell (DAC). An important advantage of the KMA is its large specimen volume which makes it possible to conduct experiments under precisely controlled P-T conditions. A typical application of the KMA to the Earth science might be determination of the post-spinel phase equilibria in the system Mg2SiO4-Fe2SiO4. By adopting sintered diamond (SD) as the anvil material accessible pressure of the KMA has substantially increased. Melting experiments of mantle materials up to 35 GPa opened a new paradigm on the mantle fractionation in early Earth. Combining the KMA equipped with SD anvils with the synchrotron radiation phase equilibria of Fe GaN and Fe2O3 were determined by means of in situ X-ray diffraction. Special attention was paid to define the phase boundaries between perovskite and post-perovskite in MgGeO3 and MnGeO3 as analogues of MgSiO3. We also observed the spin transition of Fe2+ in (Mg0.87Fe0.17)O at 300 and 700 K. Recently the maximum attainable pressure is reaching 100 GPa and high P-T experiments up to 90 GPa are our ordinary jobs. In order to produce still higher pressure however innovation of SD such as NPD is indispensable.

1. Beginning of the high pressure Earth science In the early 20th century the Earth’s interior became an object of scientific research due to the information provided by seismology [1] which made it possible to determine the density and thereby pressure as functions of depth from the surface to the center [2]. Following these pioneering works Birch [3] analyzed homogeneity of the subdivided layers of the Earth’s interior [2] using the equation of state based on finite strain theory and Bridgman’s compression data and concluded that minerals constituting the uppermost mantle such as olivine pyroxenes and garnet successively transform into closed-packed oxides similar to corundum rutile spinel or perovskite in structure. As his analyses were so concrete and convincing that the prediction had been the guiding principle for exploration of the Earth’s interior for long time and strongly stimulated the high pressure experiments.

Figure 1 shows a standard Earth’s model PREM [4]. The mantle is characterized by two sharp increases in seismic velocities at depths of 410 and 660 km which divides the mantle into three parts the upper mantle (B) the transition zone (C) and the lower mantle (D). Following Birch [3] high pressure phase equilibria of the system Mg2SiO4-Fe2SiO4 were extensively studied by many workers represented by Ringwood and Major [5] and Akimoto and Fujisawa [6] because olivine (Mg0.9Fe0.1)2SiO4 is the most dominant (> 60%) constituent of the upper mantle. And the 410 km discontinuity was reasonably assigned to the transformation of the olivine (α) into the modified spinel

23rd International Conference on High Pressure Science and Technology (AIRAPT-23) IOP PublishingJournal of Physics: Conference Series 377 (2012) 012001 doi:10.1088/1742-6596/377/1/012001

Published under licence by IOP Publishing Ltd

1

(β) structure up to the early 1970s. As capability of the apparatuses employed in these studies were limited to ca. 15 GPa neither synthesis of the end member Mg2SiO4 spinel (γ) nor acquisition of insight into the post-spinel transformation were successful at that time.

Figure 1. A standard Earth’s model (PREM) by Dziewonski and Anderson (1981). Layers A-G are correspond to division by Bullen [2].

2. Devices employed in the high pressure Earth sciences Beginning of the high pressure Earth science Soon after however γ–Mg2SiO4 was synthesized by Suito [7] in the Kawai-type multi-anvil apparatus (KMA) and the decomposition of γ-Fe2SiO4 into an assemblage of FeO (wüstite) and SiO2 (stishovite) was found by Mao and Bell [8] by adopting the diamond anvil cell (DAC). Hereafter the high pressure Earth science has been extensively developed by adopting both the devices. In the DAC it is possible to compress a small amount of sample to multi Mbar heat it to thousands of Kelvin and observe the state of the sample under these extreme conditions by means of X-ray optical and other measurements. A marked advantage of the KMA over the DAC on the other hand is the much larger sample volume which makes it possible to conduct experiments under precisely controlled P-T conditions. Therefore the KMA has been adopted in various researches determination of phase equilibria synthesis of high pressure phases and measurements of physical properties. Development of the KMA was carried out mainly during 1965-1973 at Osaka University under the direction of the late N. Kawai [9]. Photos of the KMA are shown in Figure 2. The cubic assemblage of eight cubes of tungsten carbide (WC) an octahedral pressure medium and gaskets so-called the Kawai-cell (left) is squeezed along the [111] direction in the split sphere guide blocks with the aid of a uniaxial press (right).


2

FK

presqu

igure 2.The KKawai-cell cossure medium

ueezed in the

Kawai-type momposed of cum pyrophyllitsplit sphere g

press (l

multi anvil appubic WC anvite pre-gasketsguide block wlower photo).

paratus (KMAils MgO octas etc. (upper

with the aid of

A). The ahedral photo) is

f uniaxial


3

The egrown sinthan 23 Gmantle ma

3. The poto phase eSynthesis assemblagwith the lasystem Mand 1600 the mant(Mg0.97Fe0

pressures density anIt is realizthe high p

The pquite narrby makingtemperatuthrough th

Seconrevealed t

easiest way tongle crystals oGPa and coulaterial thus sy

per

ost-spinel traequilibrium sof MgSiO3

ge of MgSiOaser heating s

Mg2SiO4-Fe2Si℃ determinetle spinel w0.03)SiO3 peroof 23.1-24.5

nd seismic vezed that the sipressure transfpost-spinel diow pressure ig the dissocia

ure fixed poinhe transition zndly the phasthat some of

o convince theof MgSiO3 perld be the mosynthesized hav

Figure 3. “Hrovskite which

ansformationstudy perovskite (P3Pv and MgOsystem. Later O4 by Ito and

ed by quenchiwith compo

ovskite (Pv) anGPa and at locities [13] ilicate Pv’s anformations inissociation pointerval less tation boundarnt of 1600 ±zone [14]. se boundary hdescending s

e performancrovskite [10] st abundant mve been used

Huge” (over 1mh is stable onl

Af

n in the syste

Pv) and confOpericlase we the post-spin

d Takahashi ing method a

osition (Mg0nd (Mg0.83Fe0

1100-1600 ℃the dissociat

nd Fp are by n pyroxenes anossesses two than 0.15 GPry correspond± 100 ℃ at

has a negativeslabs stagnate

ce of the KMA(see Figure 3

material in outo measure v

mm) single crly at pressurefter [10].

m Mg2SiO4-F

firmation of ere first carrinel transform[12] using th

are reproduced0.9Fe0.1)2SiO4

0.17)O magnes℃. As the disstion must be rfar dominant

nd garnet intoimportant ch

Pa (or 4 km ind to the depththe depth an

e slope dP/dTe around the

A may be to s3) which is staur planet. Thevarious kinds o

rystals of Mges higher than

Fe2SiO4: An

the dissociatied out by Li

mation was exe KMA. Pseud in Figure 4

dissociates siowüstite (Msociation brinresponsible fot constituents o consideratioharacteristics.n depth) for th of the 660 knd can constr

T~ -3 MPa/de660 km disc

show “huge (oable only at pe single crystof physical pr

gSiO3 n 23 GPa.

application o

ion of γ–Mgiu [11] usingamined in deudobinary dia. The diagram

into an aMw) (or ferrop

ngs about 10 or the 660 kmof the lower

on as well. First it comthe mantle spkm discontinuruct the temp

eg. Seismic tocontinuity be

over 1 mm)” pressures hightals of the deeroperties.

of the KMA

g2SiO4 into ag DAC coupletail all over thagrams at 110ms indicate thassemblage periclase Fp)

% increases m discontinuitr mantle takin

mpletes withinpinel. Therefouity we haveperature profi

omography hfore collapsin

as her ep

an ed he 00 hat of at in

ty. ng

n a ore e a ile

has ng


4

down into the lower mantle [15]. On the other hand it is known that a phase boundary with a negative slope in the mantle suppresses material transport through the boundary [16]. Therefore it is highly likely that the spinel dissociation substantially contributes to the stagnation of slabs.

Figure 4. Pseudobinary diagrams of the post-spinel transformation in the

system Mg2SiO4-Fe2SiO4 at 1200 and 1600 . After [12].

4. Adoption of sintered diamond (SD) for anvils of the KMA Decisive disadvantage of the KMA in comparison with the DAC was that the maximum attainable pressure was limited to 28 GPa so far as WC was used as the anvil material. Fortunately sintered diamond (SD) which was twice harder than WC [17] became available for anvils of the KMA in the late 1980s. Figure 5 shows SD cubes of Co-binder. We have employed cubes of an edge length 14 mm with truncated corners of 2 to 0.75 mm. We first adopted the split-sphere system [9] to squeeze the Kawai-cell of SD cubes and compared the performances with those of WC cubes. Generated pressure was calibrated using several fixed points and plotted versus oil pressure for both the SD and WC cubes as Figure 6. Superiority of SD to WC is overwhelming suggesting high potential of SD in pressure generation. Following these


5

calibration runs we carried out high P-T quenching experiments. The followings are summaries of studies which had never been achieved by using WC anvils.

Figure 5.Sintered diamond cubes with Co binder.

Figure 6.Comparison of pressure generation using SD with WC anvils.

Superiority of SD to WC anvils is overwhelming.

4.1.Synthesis of perovskite with pyrope (Mg3Al2Si3O12) As the system MgSiO3-Al2O3 represents the compositions of pyroxene and garnet in the mantle high pressure phase equilibria of the system are also indispensable to understand mineralogy of the lower


6

mantle. Sysolid solutthat the bi

We eorthorhomloop and ttiny amou

4.2MeltingIn order tomelting phare indispmaterial (fractionati

We e

liquidus p

ystematic studtions of the Minary loop betexamined the mbic perovskithe silicate pe

unt of SiO2 sti

g experimentso simulate pohase relations

pensable. TheCMM) [21] hion had been

extended melphase changed

dy by Irifune MgSiO3-rich ptween the twostability of pyte phase at 37

erovskite can ishovite sugg

s of the primoossible materis of the Earth erefore meltinhad been carrpostponed.

lting experimd from ferrop

et al. [18] haperovskite ano phases woulyrope compo7 GPa and 16accommodat

gested slight n

ordial mantle ial fractionatimaterial and

ng experimenried out up to

ment on both periclase (Fp)

ad shown thatnd Al2O3-richld persist to hsition using S600 ℃ [19]. Tte certain amononstoichiome

materials ion in a deep d element portnts of fertile po 25 GPa. Ne

the materialsto Mg-perov

t pyrope (Mg3h corundum sthigher pressurSD anvils andThe result indount of Al2O3etry of Al2O3-

magma oceationing betweperidotite [20evertheless a

Figureimagesperidot

s to 35 GPa vskite (Mg-Pv

3Al2Si3O12) brtructures at cres. d confirmed fodicates closin3. Ubiquitous-bearing pero

an formed on een liquidus p0] and CI choa clear model

e 7. Back scatts of quenchedtite at 29 31

[22]. In periv) at 31 GPa

roke-down ina. 27 GPa an

formation of thng of the bina coexistence

ovskite.

the early Earphases and meondritic mant for the mant

tered electrond charges of

and 33 GPa.

ditite the firand at 33 GP

nto nd

he ary of

rth elt tle tle

n

rst Pa


7

liquidus Mg-Pv was successively followed down temperature by Fp and Ca-perivskite (Ca-Pv) within a small temperature range. The change of liquidus phase with pressure is clearly demonstrated in a series of pictures shown in Figure 7. In the CMM Mg-Pv was the liquidus phase which was followed down temperature by Ca-Pv at pressures higher than 28 GPa and Fp was absent in super solidus conditions. Differentiation in a deep magma ocean was examined by crystal fractionation of Mg-Pv Fp and Ca-Pv for CMM and peridotitic bulk silicate Earth models. Mass balance calculation indicated that subtraction of about 40% Mg-Pv and 2% Ca-Pv from CMM yielded a residual melt with characteristics of fertile upper mantle composition. The fractionated Mg-Pv and Ca-Pv would pile up to a depth ca. 1400 km from bottom of the mantle and might be characterized as an enriched and possibly heat producing reservoir by the higher capability of Ca-Pv to accommodate large cations such as rare earth elements and alkaline elements than melt. The observed effect of pressure on element partitioning suggested that better mass balance solutions might be obtained for higher pressure liquidus composition.

5. In situ X-ray observation using the KMA In situ X-ray observation using synchrotron radiation has rapidly pushed the high pressure Earth science towards the exact sciences because the pressure value is determined continuously and uniquely from volume of a pressure marker via its equation of state. The pressure marker is usually mixed with the sample or put next to the sample. By being interfaced with synchrotron radiation versatility of the KMA has been expanded in various research fields such as phase equilibria reaction kinetics equation of state radiographic measurements etc. At the synchrotron facility SPring-8 two DIA-type presses [23] were installed on beam line BL04B1 to squeeze the Kawai-cell. One of them SPEED mkII [24] has exclusively been used to squeeze the Kawai-cell of SD cubes. Experimental methodology of in situ X-ray observation using the Kawai-cell were described to some extent elsewhere [25].

5.1.In situ X-ray diffraction studies on Fe GaN and Fe2O3 We have first studied high pressure polymorphism of iron by means of in situ X-ray diffraction. Our aim of the study was to clarify the stability of β-phase the 5th polymorph of iron which was claimed to be stable under the conditions higher than 35 GPa/1500 K by several groups from the experiments using the DAC [26]. Our experiments up to 44 GPa and 2100 K confirmed progression of the ε→γand the reverse transitions on increasing and decreasing temperature respectively and the presence of β-phase was thus excluded [27].

We successively examined phase relations in GaN [28] and Fe2O3 [29]. In GaN onset of the wurtzite→rocksalt transformation was observed at conditions of 54 GPa/300 K and 51.5 GPa/ 750 K.suggesting a negative dP/dT slope. However it was observed that the rocksalt phase persisted for 90 min at 48.9 GPa/850 K. Therefore the reverse transformation is rather sluggish and more examination is needed to decide the equilibrium phase boundary. Electrical resistance of GaN was tens of mega ohms up to ca. 62 GPa at 300 K and no remarkable change in resistance was observed on the wurtzite – rocksalt transformation.

Phase equilibria of Fe2O3 were examined up to 58 GPa and 1400 K. Hematite (phase I) transformed successively into the Rh2O3 II type structure (II) and an orthorhombic phase (III) with increasing pressure. The phase boundaries for the I - II and II - III were defined to be P(GPa) = -0.015 T(K) + 44.2 and P(GPa) = -0.005 T(K) + 48.7 respectively by observing both the forward and backward reactions in situ. It should be noted that the transformations were observed only at temperatures higher than 500 K and hematite persisted metastably at least up to 58 GPa at 300 K. The electrical resistanceof hematite on the other hand decreased monotonically from tens of mega ohms at 10 GPa to a few ohms with increasing pressure and then plateaus at ca. 60 GPa as shown in Figure 8. The isostructural electrical resistance reduction strongly suggests the occurrence of a Mott transition. The plateau point of electric resistance 60 GPa is usable as a pressure fixed point at room temperature.


8

5.2. Post-pThe loweraseismic avelocity VThese enimantle anclarify dyn

In 20MgSiO3 pGPa and important transformainsight to

As thpresent was the best

Crossmixture othe center thermocousee througfrom the v

The tdiffractionpressure bnot recogn10 (B). Wover the crelevant towas defineby T(K) =

Figure 8.

perovskite trarmost 200-30anomalies s

VS and bulk sgmatic featur

nd the core. Tnamics and ev004 Murakaperovskite (Pv

2200 K usinrole in form

ation is of urgthe thermal s

he P-T conditwe examined tt analogs of Ms section of af MgGeO3ilmof octahedro

uple whose jugh the samplvolume of goltransformation profiles in Fby reducing thnized until pr

We carried outconditions upo define the ped by passing

= 177P(GPa) –

Change in el

ansformation 00 km regionsuch as velocound velocityres would be Therefore unvolution of thami et al. [3v) into the Cang the laser-hmation and sgent need. Estructure throuions for the Pthe PPv trans

MgSiO3. a sample assemenite + 0.1Aon of MgO + 5unction was ale and the theld based on thn from Pv to

Figure 10(A).he applied loaressure went dt more few ru

p to 74 GPa aphase boundag through the – 9677 with d

300

ectrical resist3

in meta-germn of the mantcity jumps [3y VC [32] an

refection of nderstanding he Earth. 4] and OganIrO3 structureheated DAC.structure of tspecially theugh the D” layPPv transformsformation in

embly adoptedAu (in weight5% Cr2O3. Teat the center ermocouple bhe Anderson e

PPv was firs Then we triead and keepindown to 60.5uns to recognand 2200 K. Fary between P

Pv-growth PdP/dT = 5.6 M

0 K

tance of Fe2O300 K.

manates tle called th30] anisotropnd presence of

active exchaof the D” lay

nov and Onoe (the post-pe The post-pethe D” layer slope of the yer and to est

mation in MgSMgGeO3 [37

d in MgGeO3t) was put in emperature wof the sampleby the CCD et al.’s [39] Ast confirmed aed to observe ng temperatur5 GPa as clarinize growth eFigure 11 sho

Pv and PPv. TPPv-growth a

MPa/K.

O3 (hematite) w

he D” layer py [31] antf ultra-low ve

ange of energyer should be

o [35] discoverovskite (PPverovskite tranr. Therefore phase bound

timate of therSiO3 are out o7] and MnGe

3 experimentsa cylindrical as measured be. By adoptincamera. Pres

Au scale. at 63.2 GPa athe reverse tr

re constant atified by the s

either PPv or ows P-T plot

The phase bouand the coexi

with pressure

is characterizti-correlation elocity zone agy and materie of essential

vered the tranv) transformansformation sdetailed know

dary dP/dT wormal flux fromof the ability

eO3 [38] whic

s is shown inTiB2 heater

by a thin W3ng the TiB2 hssure value w

and 1325 K aransformationt 1323 K. Grosuccessive proPv from the

ts of the X-raundary betweistence points

at

zed by variobetween she

at the base [33ial between thl importance

nsformation ation) at ca. 12should play awledge on thould give som

m the core [36of the KMA

ch are regarde

n Figure 9. Finwhich is set %Re/W25%R

heater we couwas determine

as shown by thn by decreasinowth of Pv wofiles in Figucounter phas

ay observatioeen Pv and PPs and expresse

us ear 3]. he to

of 20 an he

me 6]. at ed

ne at

Re uld ed

he ng

was ure ses ns Pv ed


9

Fexp

Figure 9. A speriments for

chematic drawthe post-pero

wing of the saovskite transfo

ample assembformation in M

Figure 10M511. Prthe growtfrom the p1325 K athe bottomreverse trand 1323

bly for MgGeO3.

0.Diffraction rofiles in (A) th of post-perperovskite at and a series om to top in (Bransformation

K.

profiles of rudemonstrate

rovskite phase63.2 GPa andf profiles from

B) do the n at 60.5 GPa

un

e d m


10

The dthe D” diboundary mantle doassuming such as Tpressure bpressure s

The P

method ason Tsuchiwith a lag 5.3.High sHigh spinbeen attraand geochcompressitransition counter paspinel [12

We asample as(Mg0.83Fe0

dP/dT value iiscontinuity dand the geo

wnwellings. Tthermal cond

Tsuchiya [42]by 3.5 GPa cale is fatally

Figure 1boundar

denote th

Pv-PPv phases in MgGeO3ya’s Au scaleer dP/dT valu

spin-low spin n-low spin tracted special ahemical procion curve due[44]. We se

art of magnes].

acquired presssembly show0.17)OFp powd

is slightly smdeeper and

otherm [40] wThe slope of

ductivity of 10 instead of Aand the dP/d

y important.

11. P-T plots ory between Pvhe different ru

by the

e equilibria i. The phase be [42] which ue [38].

transition in ansition of Feattention beccesses [43]. Te to a drastic elected Fp wisian perovskit

sure (P)-voluwn Figure 9.der based on

maller than thoso called PP

would be pre5.6 MPa/K su0 Wm-1K-1 [Anderson et adT increases

of the X-ray ov and PPv (seuns. Growingarrows and c

in MnGeO3 [boundary wasis located aro

ferropericlase2+ in (Mg1-xFause the transThe transitionreduction of

ith a composte in the asse

ume (V) data . Pressure wAnderson et

ose estimated Pv lens formeesent only inuggests a larg[41]. Nevertheal.’s [38] outo 8.7 MPa

observations ree text). Circl

g phases and pcolor markers

[38] were alss determined ound ca. 10 G

se (Mg0.83Fe0.Fex)OFp occusition is consn can be det

f effective ionsition (Mg0.83

emblage produ

at 300 and 7as determine al.’s scale [3

for MgSiO3ed by the do

n relatively coge heat flux oeless if we adur phase bouna/K. Therefor

relevant to deles squares aphases presen. After [37].

o examined bto be P(GPa)

GPa lower pre

17)O urring under lidered to subtected by obnic radius of F3Fe0.17)O beuced by disso

700 K and uped from volu39]. The acqu

[36]. The smouble-crossingold region suf 16 TW acrodopt other Aundary shifts tre establishm

efine the phasand diamondsnt are indicate

by in situ X-) = 39.2 + 0.

essures than th

lower mantle stantially affe

bserving an aFe2+ accompcause the coociation of (M

p to 90 GPa ume of Au muired P-V dat

maller value seg of the phauch as beneaoss the D” layu pressure scatowards high

ment of reliab

se s d

-ray diffractio013T(K) basehat of MgGeO

conditions hect geophysicanomaly in thpanied with thomposition is Mg0.9Fe0.1)2SiO

employing thmixed with thta are shown

ets ase ath yer ale her ble

on ed O3

has cal he he a O4

he he in


11

Figure 12were with

We trdata up tozero pressvalues [45higher tha

Figure Fitting u300 and

occurrencwe fit

deviation AssumingM EoS withe insertiThe V0 anrespectivehigher preNeverthelLS state in

6. RecentBy adoptinFigure 13 observatioreaching 1ordinary j

One ospinel tran

. Errors in prhin ± 0.4 GPa ried to fit the 40 GPa whi

sure values o5] which arean 50 GPa a

12.Compressup to 40 GPa 700 K data. T

ce of the spinEoS’s with p

from the sog that the tranith K0’ = 4 foions as LS vand K0 at 300 ely in the saessure and exess it is evidn more detail.

t pressure geng SD anvils shows our re

on at SPring-100 GPa anobs. of our urgentnsformation.

ressure determand those in

e 3rd order Bich were succf volume V0 e noted as Hat 300 K and

sion data of (Massuming K0The data for h

n transition. Foparameters no

olid curves wsition comple

or the data higalues in the fK for the low

ame composixpands with ident that acqu.

eneration and the attainabl

ecent perform-8 using anvid high pressu

t objects is toIn order to ac

mination cauvolume deter

Birch-Murnagcessfully perfo

and bulk moHS values in td those highe

Mg0.83Fe0.17)O0’ = 4.0 (solidhigher pressuror low spin reted as LS in t

with increasinetes at 70 GPagher than thefigures and thw spin regimition. Our daincreasing temuisition of LS

d future persle pressure of

mance of presils with 1.0 mure and temp

o clarify the Pccomplish the

used those of rmination werhan equationormed as shoodulus K0 wethe insertionser than 55 G

OFp and fittind curves) yieldres clearly deegime of 70-9the insertions

ng pressure a at 300 K an

ese pressures. he correspond

me are larger aata suggest thmperature in

S data up to h

spectives f the KMA hassure generatimm truncatioperature expe

PPv transforme object we h

Au volume re typically w

n of state (B-Mwn by the sol

ere in good ags of the figur

GPa at 700 K

ng the 3rd ordeds reasonableeviate toward 90 GPa at 300

which are sh

indicating ond 80 GPa at

The resultanding EoS’s arand smaller that the mixed

accord with higher pressu

as increased yion carried ouon. The maxiriments up to

mation in detahave to exten

and temperatwithin ± 0.05 %M EoS) with lid curves. Angreement witres. HoweverK show grad

er Birch-Mune V0 and K0 va

lower directi0 K and 80-90hown by the b

nset of the s700 K we tri

nt V0’s and K0re shown by than those of d spin regiontheoretical p

ure is required

year by year sut by means oimum attainao 90 GPa hav

ail just like dond the accessi

ture fluctuatio%.

K0’ = 4 to thnd the resultath the literatur volume da

dual downwa

naghanEoS’s.alues for bothon indicating

0 GPa at 700 Kbroken lines.

spin transitioied to fit the B0’s are noted broken curve

f Lin et al. [4n shifts towaprediction [47d to specify th

ince 1998 [19of in situ X-raable pressure ve become o

one in the posible pressure

on

he ant ure ata ard

h g K

on. B-in

es. 46] ard 7]. he

9]. ay is

our

st-at


12

least by 20pressures possibilityimportanc(NPD) rereaches 14

Acknowle I am verytheir guidKawai S.

Reference[1] Jeffrey[2] Bullen[3] Birch F[4] Dziew[5] Ringw[6] Akimo[7] Suito K[8] Mao H[9] KawaiKawai N [10] Shats[11] Liu L[12] Ito E [13] SawaA Weidne

0 GPa. The syfor WC SD

y to producece to produce cently develo40 GPa

Figure1.0 m

edgements y grateful to

dance help aAkimoto an

es ys H 1939 Mon E 1940 Bull.F P 195) J. G

wonski A E anwood A E and oto S and FujiK 1972 J. Ph

H-K and Bell Pi N 1966 ProTogaya M an

skiy A FukuiL 176 Nature

and E. Takahamoto H Weier D J and Ito

ystematics beD and sing pressures hstill higher p

oped by Irifu

e 13. Plots of gmm and pressu

all those whoand collaborand T. Matsumo

on. Not. R. As. Seis. Soc. Am

Geophys. Res. nd Anderson D

Major A 197isawa H 1968hys. Earth 20P M 1971 Ca

oc. Jpn. Acad.nd Onodera A H Matsuzake 262 770-77hashi E 1989 idner D J Saso E 1989 Scie

etween the hale crystal diaigh than 100

pressures. In tune et al. [4

generated preure values are

o have been ation. I wouldoto and also t

str. Soc. Geopmer. 30 235. 57 227.

D L 1981 Phy70 Phys. Earth8 J. Geophys 0 225. arnegie Instit 42 385; Ka

A 1973 Proc. Jki T et al. 20072. J. Geophys. saki S and Kuence 243 787

ardness of anvamond and t0 GPa. Howthis context b48] expands o

essure vs. appe based on Ts

in contact wid like to dedito the late sup

phys. Sup. 4 4

ys. Earth Planh Planet InteRes. 73 1467

tute Year Booawai N and EnJpn. Acad. 4

07 Am. Miner

Res. 94 106umazawa M 7.

vil materials athe trends shever innovabinder-free naour dream

plied load. Trusuchiya’s Au

ith me on thecate this artic

per technician

498.

net. Inter. 25er. 3 89. 7.

ok 70 176. ndo S 1970 R9 623.

ral. 92 1744

637 1984 Science

and the maximhown in Figuation of SD ano-polycrystbecause itsK

uncation was scale [42].

e high-pressucle to the laten K. Tanaka .

297.

Rev. Sci. Instr

.

e 224 749; Y

mum attainabure 13 suggeis of essenti

talline diamonKnoop hardne

ure research fe professors N

r. 41 425;

Yeganeh-Haer

ble est ial nd ess

for N.

i


13

[14] Ito E and Katsura T 1989 Geophys. Res. Lett. 16 425. [15] Fukao Y Widiyantoro S and Obayashi M 2001 Rev. Geophys. 39 291. [16] Verhoogen J 1965 Phil. Trans. R. Soc. London A258 276. [17] Sung C.-M. and Sung M 1996 Mat. Chem. Phys. 43 1. [18] Irifune T Koizumi T and Ando J 1996 Phys. Earth Planet. Inter. 96 147. [19] Ito E Kubo A Katsura T Akaogi M and Fujita T 1998 Geophys. Res. Lett. 25 821. [20] Trønes R G and Frost D J 2002 Earth Planet Sci. Lett. 197 117. [21] Ohtani E and Sawamoto H 1987 Gephys. Res. Lett. 14 733. [22] E. Ito E Kubo A Katsura T and Walter M J 2004 Phys. Earth Planet. Inter. 143-144 397. [23] Osugi J Shimizu K Inoue K et al. 1964 Rev. Phys. Chem. Jpn. 34 1. [24] Katsura T Funakoshi K Kubo A et al. 2004 Phys. Earth Planet. Inter. 143-144 497. [25] Ito E 2007 Theory and Practice-Multianvil Cells and High-Pressure Experimental Methods In: Treatises on Geophysics 2 edited by G. D. Price and G. Schubert Elsevier B. V. 197. [26] Boehler R 1993 Nature 363 534; Andrault D Fiquet G Charpin T and le Bihan T 2000 Am. Mineral. 85 364; Saxena S K and Dubrovinsky L S 2000 Am. Mineral. 85 372. [27] Kubo A Ito E Katsura T et al. 2003 Geophys. Res. Lett. 30 1126 doi: 1029/200GL06394. [28] Ito E Katsura T Aizawa Y et al. 2005 In: Advanced in High-Pressure Technology for Geophysical Applications edited by J. Chen Y. Wang T. Duffy and L. F. Dobrhinetskaya Elsevier B.V. 451. [29] Ito E Fukui H Katsura T et al. 2009 Am. Mineral. 94 205. [30] Wysession M E Lay T Revenaugh J et al. 1998 In: The D” discontinuity and its implications edtied. M.E. Wysession E. Knittle and B. A. Buffet 28 273 AGU Washington D.C. [31] Pinning M and Romanowicz B 2004 Science 303 351. [32] Masters G Laske G Bolton H Dziewonski A 2000 In: Erath’s deep interior Mineral physics and tomography From the Atomic to the Global Scale edited by S. Karato A. M. Forte R. C. Liebermann G. Masters and L. Stixrude 117 Geophys. Mono.AGU Washington DC.63. [33] Lay T Williams Q and Garnero J 1998 Nature 392 641. [34] Murakami M Hirose K Kawamura K et al. 2004 Science 304 855. [35] Oganov A R and Ono S 2004 Nature 430 445. [36] van der Hilst R. D. de Hoop M V Wang P et al. 2007 Science 315 1813. [37] Ito E Yamazaki D Yoshino T et al. 2010 Earth Planet. Sci. Lett. 293 84. [38] Yamazaki D Ito E Katsura T et al. 2011 Am. Mineral. 96 89. [39] Anderson O Isaak D G and Yamamoto S 1989 J. Appl. Phys. 65 1534. [40]Herunlund J W Thomas C and Tackley P J 2005 Nature 434 882. [41] Stacey F 1992 Physics of the Earth 3rd ed. Brookfield Brisbane Australia p. 513. [42] Tsuchiya T 2003 J. Geophys. Res. 108 dio:10.1029/3002JB002446. [43] Lin J-F and Tsuchiya T 2008 Earth Planet. Sci. Lett. 170 248. [44] Burns R G Mineralogical Applications of Crystal Field Theory Cambridge Univ. Press p. 224. [45] Anderson O L and Isaak D G 1995 Elasticity of Minerals Glasses and Melts In: Mineral Physics & Crystallography A Handbook of Physical Constants edited by T. J. Ahrens AGU Reference Shelf 2 64. [46] Lin J-F Struzhkin V V Jacobsen S D et al. 2005 Nature 436 doi:10.1038/nature03825. [47] Tsuchiya T Wentzcovitch R M da Silva C R S et al. 2006 Phys. Rev. Lett. 96 198501-198504. [48] Irifune T Kurio S Sakamoto S et al. 2003 Nature 421 599; Irifune T 2011 ABSTRACTS AIRAPT-23 52.


14