1. Why go to high-pressures & high-temperatures?

High pressures and temperatures provide a means of synthesizing new materials that may be useful for materials science, by:

• stabilizing cations in higher co-ordinations
• increasing melting points so that high temperature phases can be synthesized
• generating high activities of normally gaseous species such as O₂ and N₂

A piston-cylinder is an apparatus, where a small piston is forced into a cylindrical area that contains a sample. A large hydraulic ram is used to transmit force to the smaller piston. Since pressure equals force/area, going from a relatively large area (ram) to a small area (piston) results in high pressures on a small sample (0.01-0.1 g). Generally, the sample is contained within a furnace assembly and so a pressure correction is required to account for the materials around the sample.

Temperature is generated by passing low-voltage, high current AC signal through a cylindrical graphite furnace. Temperature gradients along the entire length of the sample capsule are ~10°C. Temperature is measured ~0.5 mm away from the sample with W-W₇₄Re₂₆ or chromel-alumel thermocouples.

End-loaded versus non-end-loaded piston-cylinders

End-loaded piston-cylinders are designed with an additional hydraulic cylinder compared to non-end-loaded PCs. This hydraulic cylinder applies an extra vertical load to the pressure vessel, allowing it to withstand higher pressures.

Non-end-loaded piston-cylinder	End-loaded piston-cylinder

Pressure range:

Non-end-loaded: 0.5-2.5 GPa

End-loaded: up to 5.5GPa

At ASU Depths of the Earth Laboratory we have 2non-end-loaded PCs set up for both 1/2" (12.7 mm) and 3/4" (19.05 mm) size assemblies, thus our pressure range is 0.5 to 2.5 GPa.

Our laboratory recently acquired 2 end-loaded PCs. One of them is now up and running. So far, we have only been using a 3/4" (19.05 mm) pressure plate allowing us to go up to 3.5 GPa. However, we are planning on setting up a 1/2" (12.7 mm) pressure plate, which would extend our pressure range to 5.5 GPa.

Temperature range: Up to 1800°C.

3. Experimental Configurations and furnace assemblies

4. Controlling the chemical environment

Sample containers

Sample containers should protect the sample from reacting with the furnace assembly and should not react with the sample either. In some cases an unsealed container can be used (e.g. bulk synthesis of materials such as oxide ceramics that have low vapor pressure compounds and are not subject to redox reations). However, samples of geologic interest generally require sealed containers.

Noble metals or noble metal alloys are the most commonly used sample containers. These may be arc-welded shut (although care must be taken not to lose volatile samples or ignite low flash point compounds!). Alternately, capsules may be made in the shape of a bucket and lid and then pressure-welded in the piston-cylinder appartus. One method is to use oxidized transition metal capsules (e.g. NiO) that are lined with less reactive noble metal (e.g. Au; Ayers et al. 1992; Am. Min., 77, 1080-1086).

The two major restrictions of noble metal containers are:

• transition metal dissolution in the noble metal container from the sample (especially Fe)
• H diffusion through the metal.

These restrictions may be avoided largely by carefully choosing appropriate metal(s) or alloy(s), temperature conditions and salt-powder combinations in the furnace assembly.

Gas fugacities

The ambient oxygen activity in the sample container directly afffects the oxidation sate of all variable valence cations in the sample. In a completely closed chemical system, the oxygen activity of the sample is controlled by the bulk composition. However, as noted above, H diffusion through the metal can cause oxidation or reduction and dissolution of multi-valent transition metals may also affect redox conditions.

A double-capsule technique in which an inner capsule containing the sample and water is placed in an outer capsule filled with a mixture of metal, metal oxide and water in one method for fixing the chemical potential of H (e.g. Huebner, J.S., 1971, in Ulmer, G.C., Research Techniques for High Pressure and High Temperature, Springer-Verlag, 123-177). Another method is to use graphite capsule liners with a coexisting CO-CO₂, which is an effective method as long as excess C is not detrimental to the sample (Holloway, 1992, EUG).

The oxygen fugacity of the sample may be measured directly from the phase assemblage in the container (e.g. Gudmundsson and Holloway, 1994), or with a sensor capsule placed within the sample container (e.g. Taylor et al., 1992, American Mineralogist, 77, 284-295).

Page last updated April 19, 2002. Comments and Suggestions to Darren Locke