The hydrothermal diamond anvil cell (HDAC)

for Raman spectroscopic studies of geologic fluids

at high pressures and temperatures "

Christian Schmidt, I-Ming Chou*"

GFZ German Research Centre for Geosciences, Potsdam, Germany"*U.S. Geological Survey, Reston, Virginia, U.S.A. "

crucial for element cycling"!

properties of interest" - density! - viscosity! - sound velocity! - electrical conductivity! - phase transitions! - complexation,! speciation! - solubility, partitioning! - kinetics of mineral-fluid! and fluid-fluid interaction!"

Introductionhydrous fluids/melts in crust and mantle"

many of these properties need or should be studied "in situ at high P and T"

•  designed to study fluids in situ at lithospheric P-T conditions!

!!!•  ~ 180 citations in scopus !!

•  HDACs used particularly for experiments with hydrous fluids to 23 GPa at 750 °C (Lin et al., 2004) and 1025±10 °C at ~2 GPa (Audétat and Keppler, 2005)!

IntroductionBassett et al. (1993): HDAC"

Constructioncentral portion with sample chamber"

pressuresensor(e.g.,

quartz)

metal gasket(Ir, Re ± Au coating)

heater wires(Mo, Pt, PtRh, NiCr)

diamondanvils

sample chamber, fluid thermocouples(K type)

objective lensof microscope

pressuregeneration

tungstencarbide

seat

ceramic support

Constructionupper and lower platen"

design minimizes temperature gradient in sample chamber and permits accurate measurement + control of sample temperature!

Smith and Fang (2009) "

Constructionalignment of culet faces of anvils"

interference color pattern indicates parallelism!of culet faces!

•  rocker for rotation!•  sliding disk for translation!

Constructionassembled HDAC on xyz-stage of Raman spectrometer"

Bassett"et al. (1993) "

gas inlet

heater wiresto thyristor-driven power supplies

thermocouple wiresto temperature controllers

driver screwwith Belleville

washer

temperature measurementcalibration using melting points"

melting points at atmospheric pressure, e.g. of

•  NaCl (halite) (800.7 °C) •  CsCl (645 °C) •  K2Cr2O7 (398 °C) heating to >500 °C damages culet •  NaNO3 (306.8 °C) heating to >500 °C damages culet •  S (112.8 °C) •  azobenzene (68 °C)

triple points of water for calibration at low temperature

•  ice I + liquid + vapor at 0.01 °C, 0.6 kPa •  ice I + ice III + liquid at -21.985 °C, 209.9 MPa

temperature measurementcalibration using α-β quartz transition"

573 °C 574 °C

•  displacive, very little hysteresis •  574 °C upon heating at 0.1 MPa •  optical observation under crossed polars •  should be cut parallel to c axis, section ~75 µm thick

pressure determinationoverview"

indirectly: optical microscopy, measurement of phase-transition temperatures in

•  solid calibrant •  fluid in sample (with application of appropriate EoS)

directly: X-ray diffractometry or optical spectrometry, measurement of P-dependent property of a standard •  angle or energy positions of Bragg reflections of e.g. Au, Pt, NaCl, MgO. Rarely applied in studies on fluids •  frequency shifts of Raman or fluorescence lines of optical pressure sensors - fluorescence sensors: ruby (α-Al2O3:Cr3+), Sm:YAG, SrB4O7:Sm2+ - Raman spectroscopic sensors (work often better at high T): α-quartz, berlinite (AlPO4), zircon, c-BN, 13C-diamond

pressure determinationRaman bands: α-quartz"

ν206: <2.5 GPa at RT, but high resolution

ν464: to ~600 °C, to ~3 GPa (10 GPa at RT)

0

1

2

3

4

5

6

100 200 300 400 500 600wavenumber (cm-1)

Inte

nsity

(103 c

ount

s)

Raman spectra of -quartz

23 C, 2130 MPa

23 C, 0.1 MPa

EE E

A1 A1 A1

128

206

265 35

539

440

2

464

510

355

139

247

482

E E

276

394

405

520

206

P ~20 cm-1GPa-1

464P

~9 cm-1GPa-1

ν464"

Schmidt and Ziemann (2000) "

Pressure determination Raman bands: ν3(SiO4) band of zircon"

Schmidt et al. (in revision)"980

990

1000

1010

1020

700 °C, 5

.87 cm-1 /G

Pa

600 °C, 5

.88 cm-1 /G

Pa

500 °C, (6.08 cm

-1 /GPa)

200 °C, (5.83 cm-1 /GPa)

300 °C, (6.12 cm-1 /GPa)

400 °C, (6.31 cm

-1 /GPa)

0 10.5 1.5 2Pressure (GPa)

Ram

an

sh

ift,

3(S

iO4),

(cm

-1)

27 °C, 5

.81 cm-1 /G

Pa

0

1

2

3

4

5

6

7

940 960 980 1000 1020 1040wavenumber, cm-1

inte

nsit

y (

10

3 c

ou

nts

)

700 °C,

2.36 GPa

700 °C,

0.1 MPa

27 °C,

1.95 GPa

23 °C,

0.1 MPa

3(SiO

4)

1(SiO

4)

T P

ν1008: to ~1000 °C, to ~10 GPa

pressure determinationRaman bands: ν1055 (= νTO) of c-BN"

c-BN Raman spectr. pressure scales •  to 900 K, 80 GPa (Datchi et al. 2007) •  to 3300 K, 70 GPa (Goncharov et al. 2007)

Datchi and Canny (2004) "

inert

nearly linear dependence of ν on P, but small (∂ν/∂P ~3 cm-1GPa-1)

pressure determinationRaman bands: ν1111-ν462 of berlinite"

•  ν1111 and ν462: shift in opposite direction with P and T •  ∂(ν1111-ν462)/∂P ~10 cm-1GPa-1 •  reacts with aqueous fluids at elevated T

Watenphul and Schmidt (2012) "

pressure determinationfluorescence bands: ruby"

most common technique to measure pressure in DACs

doublet, at ambient P-T: •  R1 at ~14404 cm-1 (~694.25 nm) •  R2 at ~14433 cm-1 (~693.85 nm) •  ∂ν/∂P ~-7.5 cm-1GPa-1

(=0.365 nmGPa-1 ) •  has been used to 0.55 TPa •  recent ruby pressure scales to

~300 GPa (e.g., Dorogokupets and Oganov 2007)

•  original calibration by Piermarini et al. (1975) still valid to ~10 GPa

Piermarini et al. (1975) "

pressure determinationfluorescence bands: ruby"

Datchi et al. (2007) "

with increasing T: •  broadening •  R1 and R2 merge •  intensity decreases •  strong and nonlinear

shift in wavenumber (∂νR1/∂T ~ -0.14 cm-1K-1)

accurate pressure determination becomes difficult at T >300 °C, particularly at relatively low P to a few GPa

pressure determinationfluorescence bands: SrB4O7:Sm2+"

Datchi et al. (2007) "

single, very intense, fluorescence line at ~685.41 nm at ambient PT

•  ∂ν/∂P ~-5.5 cm-1GPa-1

(=0.26 nmGPa-1) •  ∂ν/∂T ~ 0

•  still detectable at 900 K •  calibrated to 130 GPa

(Datchi et al. 1997) •  drawback: quite soluble in

aqueous fluids

in neon medium"

pressure determinationphase transitions and isochores in fluid"

modified from Schmidt and Ziemann (2000) "

liquid-vaporcurve

Schematicplan viewof samplechamber

Temperature

Pressure

critical point

1 Vapor

Crystal

2 3

4

Fluid

5

isochore

1 2 3

4

56

Pressure medium(aqueous fluid)

liquid-vaporhomogenizationbefore andafter heating

P-T pathof experiment

Liquid

pressure determinationdilute aqueous fluids: EoS of water"

equation of state: Wagner and Pruß (2002) "

0

100

200

300

400

500

600

700

800

900

1000

1100

1200

1300

1400

1500

-100 0 100 200 300 400 500 600 700 800 900

P, M

Pa

T, °C

Ice VI

Ice V

Ice III

Ice Iliquid-vapor curve

Th(L+V=L

) = 50

°C

Th(crit) = 373.95 °C

Th(L+V=V) = 300 °C350(V)360(V)370(V)

370(L)360(L

)

340(L)330(L)320(L)310

(L)

Th(L+V=L) = 350 °CTh(L+V=L) = 300 °CTh(L+V=L) = 250 °CTh(L+V=L) = 200 °C

Th(L+V=L) = 10

0 °C

290(L)280(L)270(L)260(L)240

(L)230(L)220(L)210(L)

120(L)

140(L)

160(L)

180(L)Tm(ic

e I) = 0

.01 °C

Tm(ic

e I) = -

5 °CTm

(ice III

) = -2

0 °C

Tm(ic

e I) = -

20 °C

Tm(ic

e I) = -

10 °C

Tm(ic

e I) = -

15 °C

Tm(ic

e V) =

-15 °

C

Tm(ic

e V) =

-10 °

C

Tm(ic

e V) =

-5 °C

Tm(ic

e V) =

0 °C

Tm(ic

e VI)

= 10

°C

Tm(ic

e VI)

= 20

°C

Tm(ic

e VI)

= 30

°C

Tm(ic

e VI)=

40°C50(V

I)

0.3220 g/cm3

0.0462 g/cm3

0.5747 g/cm3

0.1136 g/cm30.2018

0.1439

0.45140.5276

0.61070.64080.66710.69

070.7121 g/cm30.73

190.75030.76750.78360.7989 g/cm3

0.81340.82710.84020.85270.8647 g/cm3

0.958

4 g/cm

30.9

880 g

/cm3

0.999

8 g/cm

3

0.8870

0.90750.9

2610.9431

1.02931.05181.07021.0860

1.1037

1.13921.15521.17161.1886

1.2063

1.22431.2429

1.2621

1.2819

pressure determinationmelting curve of pressure medium"

0

1

2

-50 0 50 100 150Temperature, °C

Pres

sure

, GPa Ice VI

Ice V

Ice I

IceIII

IceII

Ice VIIIce VIII

liquidH2O

liquidH2O

33 °C

Ice VIic

e m

eltin

g cu

rves

(Wag

ner e

t al.

1994

)

pressure determinationα-β quartz transition temperature"

Bassett et al. (1996) "

•  α-β quartz transition is sensitive to pressure (~260 K GPa-1),

•  drawbacks: T too high for many HDAC experiments, high solubility in H2O

other transitions applicable at lower T: •  BaTiO3 (tetrag./cubic) •  Pb3(PO4)2

(monoclinic/trigonal) •  PbTiO3 (tetrag./cubic)

•  optical microscopy, microthermometry, falling sphere technique

•  Raman spectroscopy

•  synchrotron-radiation X-ray fluorescence and absorption spectroscopy

•  inelastic X-ray scattering

•  infrared absorption

•  Brillouin spectroscopy

•  laser-induced phonon spectroscopy

•  electrical conductance measurements

techniques that have been applied to study aqueous fluids/melts in situ using HDACs"

Application optical microscopy, microthermometry "

Thomas et al. (2011)"

Example: experiment on H2O + 4.5 molal NaHCO3 + SiO2 "!fluid-1 = aqueous fluid!fluid-2 = 2nd fluid (H2O+! carbonate+silicate)!Qtz = quartz!NaHCO3 = nahcolite!Na2CO3(s) = natrite!

Raman spectroscopy and HDAC In situ studies on geologic fluids"

permits study of molecules and molecular ions of light elements in fluids at high P and T

complexation, speciation, phase transitions, solubility, kinetics, e.g.,

•  complexation in H2O + KAlSi3O8 fluids to 900 °C, 2.3 GPa, phase diagram of H2O + 40 mass% KAlSi3O8 (Mibe et al., 2008) •  silica speciation and solubility of quartz in H2O to 900 °C, 1.4 GPa (Zotov and Keppler 2002) •  ammonium in aqueous fluids to 600 C, 1.3 GPa: silica and N speciation, silica solubility of Qz + Ky + Kfs/Ms in H2O ± NH4Cl, kinetics of Kfs to Ms reaction (Schmidt and Watenphul 2010) •  ice VII melting curve to 630 °C, 22 GPa (Lin et al. 2004)

Raman spectroscopy and HDAC In situ studies on geologic fluids"

Quantification (e.g. solubility measurement) possible in in some cases

Difficulties:

•  rather high detection limits for most species •  not all relevant species are Raman active or Raman distinguishable •  changes in the Raman scattering cross sections with P,T,X are unknown for most species

Raman spectroscopy and HDAC measurement of species concentration"

Raman spectroscopy and HDAC vertical scan of ν1-SO4

2- intensity in fluid"

Schmidt (2009)"-100

-80

-60

-40

-20

0

20

0 500 1000 1500 2000 2500

Z (

µm

)

intensity of 1-SO

4

2- band (cps)

aqueous Na2SO

4 fluid

upper diamond anvil

Raman spectroscopy and HDAC change in ν1-SO4

2- cross section with P and T!

Schmidt (2009)"

900 920 940 960 980 1000 1020 1040Raman shift, cm-1

200

400

600

800

1000

1200

1400

1600

1800

2000

inte

nsity

600 °C, 1120 MPa

100 °C, 120 MPa

500 °C, 920 MPa400 °C, 720 MPa

200 °C, 320 MPa300 °C, 520 MPa

1.1604 g/cm3 isochore1.54 molal Na2SO4 in HDAC

probably effect of bond expansion with temperature on polarizability

0.9

0.95

1

1.05

1.1

1.15

1.2

1.25

1.3

0 100 200 300 400 500 600Temperature, °C

1-SO

42-,P

T

theore

tical

liquid-vaporcurve

1.1604g/cm3

isochore

1.0587 g/cm3isochore

1.54 molal Na2SO4 in HDAC

from Bose-Einstein factor n(population of energy states)n + 1 = (1 - e-hc j/kBT )-1

Raman spectroscopy and HDAC m SiO2(aq) in Qz+Ky+Kfs/Ms+H2O±NH4Cl"

Schmidt and Watenphul (2010)"

for calibration:"mSiO2(aq) "= 0.375"at 600 °C, "970 MPa "(Manning 1994)"

mon

omer

sH 4S

iO4(a

q)

Qz + H2O

vapor970 MPa, 600 °C

800

900

700

600

500

400

300

2001000800600400

Raman shift (cm-1)

inte

nsity

(cou

nts)

* *

* *

* *

* *

* *PV, 40 °C

920 MPa, 600 °C

730 MPa, 500 °C

1290 MPa,600 °C

Qz + Ky + Kfs

+ Ms + H2O

Qz + Ky + Kfs + Ms/Tob

+ 5 m NH4Cl

Qz + Ky + Kfs

+ Ms/Tob

+ 9.6 m NH4Cl

Q0Q1

dim

ers

trim

ers background

signalfrom

diamond

Muscovite/Tobelite

100 µm

K-feldspar/Buddingtonite

KyaniteVapor

Quartz

H2O+ NH4ClFluid

after reaction at 500 °C, 730 MPa

250

300

350

400

450

500

710 730 750 770 790 810 830

550600 °C

Raman shift (cm-1)

inte

nsity

(cou

nts)

,no

rmal

ized

to d

ensi

ty o

f H2O

1290 MPa

750 MPa

970 MPa

920 MPa

575 MPa

Qz + Ky + Kfs + Ms/Tob

+ 9.6 m NH4Cl

Qz + 6.6 m NH4Cl

Qz + H2O

Qz + Ky + Kfs + Ms + H2O

Qz + Ky + Kfs + H2O

Raman spectroscopy and HDAC m SiO2(aq) in Qz+Ky+Kfs/Ms+H2O±NH4Cl"

Schmidt and Watenphul (2010)"

0.1

0.2

0.3

0.4

0.1 0.2 0.3 0.4

600°C575MPa

quartz solubility in H2O (mol SiO2 / kg H2O)

9.6 mNH4Cl +Qz+Ky+Kfs, this study

H2O+Qz+Ky+Kfs, this study

6.6 mNH4Cl +Qz, this study9.6 m NaCl +Qz, NM2000

6.6 mNaCl +Qz, NM2000

SiO

2 con

cent

ratio

n in

flui

d (m

ol /

kg H

2O)

600 °C920 MPa

(+Ms)

600 °C750 MPa

600 °C750 MPa

600 °C1290 MPa(+Ms/Tob)

500 °C1100 MPa

500 °C780 MPa

500 °C355 MPa

500 °C555 MPa

600 °C1290 MPa

500 °C1100 MPa(+Ms/Tob)

Thank you for your attention!"