1

GSECARS, University of Chicago, Chicago

prakapenka@cars.uchicago.edu

Dynamics of Crystallization and Melting under PressureVitali Prakapenka

LVP

DAC

R.Wenk, 2005

•Most of what we know about the structure of the Earth's deep interior comes from the study of seismic wave velocities

•In order to interpret such measurements in terms of mineralogical/compositional models of the Earth's interior, data on the physical and chemical properties of minerals at high pressures and temperatures are essential

•Knowledge of thermodynamics, phase equilibria, crystal chemistry, crystallography, rheology, diffusion and heat transport are required to characterize the structure and dynamics of the Earth's deep interior

The world’s deepest borehole in Sakhalin is 12,345 m (40,502 ft)

At pressures above 225 gigapascals andtemperatures over 3400 kelvin, Fe0.9Ni0.1 adoptsa body-centered cubic structure

Science 316, 1880 (2007), Times Cited: 53

6 8 10 12 14

5 6 7 8 9 10 11 12 13 14 15

molten Ge

~1280K

Inte

nsity

, a.u

.

2Θ, degrees

1280K 1150K 1050K RT

MgOGe II

Ge, 19 GPa

0 5 10 15 20 25 30 35 40

400

800

1200

1600

2000 Voronin Ge-I Ge-I/Ge-II Ge-II melt

Tem

pera

ture

, K

Pressure, GPa

Ge-I Ge-II

Liquid

Prakapenka et al., High Pressure Research, 2008 , 28:3,225

High P-T phase diagram of Germanium

hot spot in the confined space

10 microns

Diamond anvil

Diamond anvil

Pressure medium

Sample

Laser

X-ray

Fe : SiO2, 38 GPa,

30 µm

Carbon transport in diamond anvil cells

Prakapenka et al, High Temperatures High Pressures, 2003/2004, volume 35/36, pages 237-249

Dewaele et al, PRL, 2010

RT

1400K

1520K

1900K

1700K

T quenched

Fe:C(d) + NaCl, 1:1, P ~ 92 GPa

^ diamond

Fe3C

Fe7C3

Prakapenka et al, 2011

SEM images of FeO sample after laser heating

Leonid Dubrovinsky, BGI

11

10 15 20

10 15 20

10 15 20

10 15 20

10 15 20

10 15 20

101 hcp

101 hcp

111 fcc

MeltT ~ 3250 K

10 15 20

10 15 20

Melting of Fe-Ni alloy at 60 GPa

12

Ice VII, 7 GPa, Tm ~ 600K

(110)

13

HT RT

Ice VII, 7 GPa, Tm ~ 600K

Laser heating

RT, starting

14

15

60 s, 100 mm, 3 p/s

(110)

Laser effect?

Ice VII, 7 GPa, Tm ~ 600K

16

Pt foil: melting at ambient pressure with double sided laser heating

2100 K 1900 K

solid Pt fluid

17

Grain boundary misorientation maps of electron back scattered diffraction

Kim et al, METALS AND MATERIALS International, Vol. 8, No. 1 (2002), pp. 7~13

Images of electron back scattered diffraction at 1000”C and 0.5/sec under various strains; (a) 10%, (b) 50% and (c) 500%

18

Dynamic recrystallization

http://www.ndt-ed.org

•Chemistry •Sample diffusion•Crystallization

fs ps ns µs ms s min

Atomicvibrations

Dislocation nucleation

recrystallization

Thermally activated reaction dynamics

Phase Transformation diffusion

Static experiment

o phase transition kineticso structural dynamics & deformation o chemical reaction dynamicso transport properties (diffusion)o electronic properties

Probing fundamental ultrafast processes

High P-T-B-εLaserDynamic load

XESInelastic scattering

SAXSSynchrotron

Finite-element calculations of the temperature profiles in the DAC cavity

8 ns pulse width

Goncharov, JSR, 2009

CW

0.16 0.20 0.24 0.28 0.32 0.36 0.400.0

0.3

0.6

Inte

nsity

, a.u

.

Time, ms

up to 50 kHz

Pulse laser heating

1 μs

Goncharov, Prakapenka et al, Rev. Sci. Instrum. 81, 113902 (2010)

Prakapenka et al., High Pressure Research, 2008 , 28:3,225

Optical schema of the on-line, flat top, double-sided laser heating system at GSECARS

500 – 10,000 K

Standard Operating Mode, top-up102 mA in 24 singlets (single bunches) with a nominal current of 4.25 mA and a spacing of 153 nanoseconds between 40 ps singlets.

Special Operating Mode - hybrid fill, top-upTotal current is 102 mA. A single bunch containing 16 mA isolated from the remaining bunches by symmetrical 1.594 microseconds gaps. The remaining current is distributed in 8 group of 7 consecutive bunches with a maximum of 11 mA per group. The total length of the bunch train is 500 ns.

Synchrotron24 singlets

4.25 mA153 ns

T=3.672 µs

Synchrotronhybrid fill

500 ns, 88 mA

65 ps, 16 mA

1.594 µsT=3.672 µs

Pt, 69 GPa

200 ns delay

600 ns delay

0 ns delay

Pilatus, 10000 pls600 ns width

2Θ, degree

Detector window: 0.6 usScan step: 0.2 usCeO2, 105 pulses, 37 keV

0.2 µs

Synchrotronhybrid fill

500 ns, 88 mA

16 mA

-0.4 0.0 0.4 0.8 1.2 1.6

0

20

40

60

80

100

120Co

unts

, a.u

Delay, microseconds

Synchrotronhybrid fill

500 ns, 88 mA

16 mA

9

100

200

300

Inte

nsity

, a.u

.

Two theta, degree

HT RT

Synchrotronhybrid fill

500 ns, 88 mA

16 mALaser

-2 -1 0 1 2 3 4 5 6 7

0

2

4

Inte

nsity

, a.u

.

Time, µs

detector laser

Wavelength (nm)

560 580 600 620 640 660

Rel

ativ

e In

tens

ity

1316(5) K

2741(6) KMeasurements

Planck fit

The time-resolved radiometric measurements of temperature

accumulation times: 0.1-10 s500 ns temporal resolution

Goncharov, Prakapenka et al, Rev. Sci. Instrum. 81, 113902 (2010)

9 10 11

50

(200

)

Inte

nsity

, a.u

2Θ, degree

300 K 2600 K 3600 K(1

11) Ir, 40 GPa

2600 K300 K 3600 K

Pulsed laser heating Ir at 40 GPa

TwoTheta (degree)

4 6 8 10 12 14 16

Inte

nsity

(arb

. uni

ts)

0

1

2

3

4 2400 K 3000 K 3350 K 3600 K

• Melting can be detected by observing a diffuse diffraction ring

dspace/aIr

0.3 0.4 0.5 0.6 0.7 0.8 0.9

Inte

nsity

(arb

. uni

ts)

IrN2 at 42 GPa- this work IrN2 at 0 GPa -Crowhurst et al., 2007

IrIr Ir Ir

• Chemical reactivity is very fast in pulsed heating experiments• New possibilities for studying of chemical reactions

IrN2

6 8 10 12 14 16 18

40

80

120In

ensi

ty, a

.u

2Θ, degree

pulse mode, 0.2s continius, 10s

Compare XRD for CeO2 collected with PILATUS detector for different exposure time:

continues 10s (divided by 50) and 0.2s averaged over 105pls, 2us window

5 10 150.00

0.01

0.02

0.03

0.04In

tens

ity, a

.u.

2Θ, degree

10 pls

Detector window: 2 us or ~12 single bunches or ~2x105 photons

Standard Operating Mode, top-up: 102 mA in 24 singlets (single bunches) with a nominal current of 4.25 mA and a spacing of 153 nanoseconds between 40 ps singlets

5 10 150.0

0.1

0.2

0.3

0.4In

tens

ity, a

.u.

2Θ, degree

10 pls 100 pls

Detector window: 2 us or ~12 single bunches or ~2x105 photons

Standard Operating Mode, top-up: 102 mA in 24 singlets (single bunches) with a nominal current of 4.25 mA and a spacing of 153 nanoseconds between 40 ps singlets

5 10 150

10

20

30

40In

tens

ity, a

.u.

2Θ, degree

10 pls 100 pls 10000 pls

Detector window: 2 us or ~12 single bunches or ~2x105 photons

Standard Operating Mode, top-up: 102 mA in 24 singlets (single bunches) with a nominal current of 4.25 mA and a spacing of 153 nanoseconds between 40 ps singlets

MAR-CCDReadout time: 3.5 sDynamic range: 16 bitsSize: Ø165 mmPixel size: 79 um

PILATUS 100KReadout time: 2.7 msDynamic range: 20 bitsSize: ~84x33 mm2

Pixel size: 172 um

Dynamic x-ray probe optimization:

Synchrotronhybrid fill

500 ns, 88 mA

65 ps, 16 mA

1.594 µsT=3.672 µs

< 1 µs:

Synchrotron24 singlets

4.25 mA153 ns

T=3.672 µs2 -100 µs:

Detector: larger area, higher efficiency above 30 keV

Sample: thick, high Z, single crystal

General:

Probing fundamental ultrafast processes

o high temperature EOSo phase transition kineticso structural dynamics & deformation o chemical reaction dynamicso transport properties (e.g., diffusion)o electronic properties

Pulse laser heating and optical spectroscopy combined with time-resolved x-ray probe

Reliable experimental conditions at higher then static T and P

CIW :A. GoncharovV. Struzhkin

GSECARS: Mark Rivers, Steve Sutton, Yanbin Wang, Peter Eng, Matt Newville, Przemek Dera, Nancy Lazarz, Fred Sopron

ESRF :I. Kantor