What experiments should we be doing

1
What experiments should we be doing?

• Dion L. Heinz
• University of Chicago

2
Physical Properties

• Thermal equation of state
• Melting (noted as important by CSEDI)
• Thermal conductivity
• Heat capacity, enthalpy of melting (really need
  to know optical absorption at experiment
  conditions)
• Optical properties of material at experimental
  conditions at laser wavelength(absorption) and
  over the wavelength range of the temperature
  measurements (emissivity).

3
Chemical Properties

• Phase diagram
• Multi-component phase equilibria
• Chemical diffusion
• Partitioning of major, minor and trace elements
• In other words petrology in the laser-heated
  diamond anvil cell

4
What do we need to do to accomplish these
experiments?

• High stable laser power
• Stable sample temperatures
• Accurate temperature measurement
• Measurement of wavelength dependent optical
  properties
• Accurate pressure measurement
• Measurement of temperature gradients
• Control of temperature gradients
• Modulation of temperature as a function of time

5
What do we need to do to accomplish these
experiments?

• High stable laser power (technology)
• Stable sample temperatures (technology, but can
  do today)
• Accurate temperature measurement (can do with
  below)
• Measurement of wavelength dependent optical
  properties (can do)
• Accurate pressure measurement (should be able to
  do)
• Measurement of temperature gradients (technology)
• Control of temperature gradients (technology,
  need excess laser power)
• Modulation of temperature as a function of time
  (can do now)

6
Inter-laboratory comparisons

• We need a fixed point scale of some sort that can
  be compared between laboratories for temperature
  and pressure measurement
• Andy Jephcoats suggestion of crossing melting
  curves is the best that I have heard of so far
• High P-T triple points in single component
  systems may also be a possiblity

7
Central to good inter-laboratory comparison

• Pressure scales
• Temperature calibration, with verification!

8
Why is temperature calibration and verification
so important?

• Calibration removes system response
• Verification proves that the system works or that
  the physics of the system is correct
• Wavelength dependent emissivity is an important
  part of the verification procedure

9
Plancks Radiation Law

• I(l,T) e(l)C1l-5expC2/lT - 1-1

10
What is the physical interpretation of emissivity?

• For the solid to give up a photon, there has to
  be a physical mechanism of that energy in the
  solid e.g. a vibration, electron, etc
• Metals typically have electrons that can give up
  the appropriate energy, but the emissivity can
  easily have a wavelength dependence

11
Emissivity

• Assume absorption is equal to emissivity
  (Kirchoffs Law)
• Absorption is proportional to the complex part of
  the index of refraction (actually 4pi/lambda)

12
Fe
13
(No Transcript)
14
Fe Emissivity

• In Iron the d electron band is right at the Fermi
  level, thus there are occupied electronic states
  below unoccupied states
• Therefore, as you go to shorter wavelengths of
  light, you can excite more and more electrons
  with photons

15
(No Transcript)
16
(No Transcript)
17
How much does emissivity matter?
18
What about transparent materials?

• Take MgO for instance, it does not have any
  mechanisms in the visible region to give up
  photons
• The band gap is large (i.e. in the UV)
• The lattice vibrations are
• in the infrared

19
MgO
20
SiO2
21
Why does MgO glow?

• The answer has to be surface states (in a
  polycrystal), defects or impurities.
• The problem is what does the absorption spectrum
  of an impurity or defect look like?
• Impurities are also very sample dependent!

22
How do we measure the wavelength dependence of
the absorption?

• Use three lasers to measure reflectivity or
  transmission of the sample in situ (you have to
  give up these wavelength bands for temperature
  determination)
• Some Problems with doing the above
• Reflectivity is highly surface dependent
• Transmission could be affected by scattering

23
Temperature gradients

• We calibrate our systems with a spatially flat
  source (in terms of intensity)
• What we need is a portable temperature gradient
  that can reproduce gradients at the scale of
  laser-heated diamond anvil hot spots

24
Conclustions

• I introduced the directions that I think we need
  to go in the next five years, and outlined where
  I think the rough spots are that need to be
  addressed to make progress