Building the machines

Range

vacuum to 1 bar; to 1,600 °C (p. 18)

Precision

about 1% in P; 1 to 15 °C (p. 18)

Sample

about 0.1 ml, hung on a Pt wire loop (pp. 19-20)

Built from

ceramic muffle tube, resistance windings, insulation, two gas flowmeters, a mixing chamber, Pt wire

Original design

Darken & Gurrey (1945); applied to magmas by Osborn (1959) (p. 19)

Suited to

liquidus relations in iron-bearing systems, mineral and glass synthesis, element partitioning (p. 20)

In the book

entry pp. 18-22; schematic Fig. 3.3, p. 21

What it does. Two feed gases, CO-CO₂ or H₂-CO₂, are mixed at set flow ratios and flowed over the sample, fixing ƒO₂ through the gas equilibrium (pp. 21-22, 172). A current pulse melts the fine quench wire and the sample drops into a bath, quenching in seconds, with nearly zero iron loss to containers and several samples per run (pp. 19-21).

Range

below about 10⁻⁶ bar; to silicate liquidus temperatures and beyond

Built from

a vacuum chamber, a pumping stack, W or Mo heater and fixtures, a thermocouple or pyrometer, power feedthroughs

Suited to

evaporation and condensation kinetics, volatile loss from melts and regolith, kinetic isotope fractionation

Provenance

my bench

What it does. Free (Langmuir) evaporation into vacuum. With no returning gas flux, the evaporating surface sets the chemistry, and the escaping vapor carries a kinetic isotope fractionation that the condensed residue records.

Range

1 to 12,000 bar; to 950 °C, inversely linked (pp. 25, 27)

Precision

0.1 to 1% in P; ±1 to 30 °C (p. 25)

Sample

0.1 to 0.5 ml welded capsules (p. 25)

Built from

one superalloy rod (Stellite-25 or Rene-41) drilled nearly through, a cone seal, a water line, a gauge and hand pump, a Kanthal-wound tube furnace (p. 27)

Original design

Tuttle (1949); 12 kbar version Luth & Tuttle (1963) (p. 25)

Suited to

granite and metamorphic phase equilibria, hydrothermal mineral synthesis, diffusion and fluid-phase chemistry (p. 27)

In the book

entry pp. 25-27; schematic Fig. 3.5, p. 26

What it does. The book ranks it the simplest and least expensive of the common high-pressure systems (p. 27). The seal sits outside the furnace, hence cold-seal, so a simple cone seal survives. This is the machine of Tuttle and Bowen's granite melting experiments and the long-standing workhorse of metamorphic and hydrothermal synthesis work (p. 27).

Range

1 to 2,000 bar; to 600 °C (p. 22)

Precision

0.02% in P; ±10 °C (p. 22)

Sample

about 250 ml of fluid plus rock powder (p. 22)

Built from

an externally heated steel vessel, a flexible gold (now titanium) sample bag, a gold-filtered sampling line and valve, a rocking drive (pp. 23-24)

Original design

Dickson et al. (1963); current form Seyfried et al. (1979); fixed-volume variant Barnes (1963) (pp. 22, 24)

Suited to

hydrothermal alteration of rocks, mineral solubilities, reaction kinetics (p. 24)

In the book

entry pp. 22-25; schematic Fig. 3.4, p. 23

What it does. The rocking keeps fluid and rock stirred, and fluid aliquots are drawn at pressure and temperature through the gold filter, so a single run yields a time series of fluid chemistry. An H₂ membrane reads or sets ƒO₂ (p. 24). Quench takes hours and runs last weeks to months, so it is a machine for fluids, not crystals (p. 24). The Barnes fixed-volume variant samples both liquid and vapor, built for aqueous sulfide work (pp. 24-25).

Range

0.1 to 10 kbar; to 1,500 °C (p. 27)

Precision

0.1% in P; ±2 °C (p. 27)

Sample

0.6 to 4 ml, up to 30 ml (p. 27)

Built from

a thick-walled steel vessel with closure heads, an internal double-wound furnace, an argon intensifier chain, optionally a Shaw H₂ membrane (pp. 28-29)

Original design

Yoder (1950); large volume Burnham et al. (1969) (p. 27)

Suited to

melting of hydrous magmas, fluid-melt and fluid-crystal equilibria, synthesizing large volumes of starting material (p. 29)

In the book

entry pp. 27-29; schematic Fig. 3.6, p. 28

What it does. The furnace sits inside the vessel, so pressure and temperature limits are not inversely linked the way they are in a cold-seal. Argon at 5 kbar and 1,000 °C is about as dense as room-temperature water, and its convection is tamed with double windings and vertical operation (p. 29). The machine for volatile solubility in melts and for fluid-melt equilibria needing large fluid volumes (pp. 27-29). The honest caveat, this is the one entry on this page whose pumping system the book concedes is relatively complicated and expensive (p. 27).

Gas mixing

at 1 atm, ƒO₂ = (P_CO₂ / P_CO·K)², set by flow ratio (p. 172)

Solid buffers

MH, Ni-NiO, QFM, IW assemblages plus H₂O, sealed in an outer capsule (pp. 172-173)

Membrane

Shaw Pd-Ag tube fed from an external H₂ reservoir (pp. 173-174)

Original designs

gas mixing Darken & Gurrey (1945); buffers Eugster (1957); membrane Shaw (1963, 1967) (pp. 172-173)

Suited to

fixing redox state in any sealed-capsule or gas-flow experiment

In the book

Appendix A.3, pp. 171-174

What it does. In the double capsule, H₂ generated by the buffer diffuses through the Pd-Ag or Pt inner wall and fixes ƒO₂ inside through the water reaction (p. 173). Permeabilities run PdAg > Pt >> Au > Ag (p. 174). Buffers are consumed in long runs and sit at widely spaced ƒO₂ values; the membrane varies ƒO₂ continuously and never runs out (pp. 173-174).

Range

1 atm; 700 to 1,200 °C (p. 43)

Precision

±0.05 log units in oxygen activity (p. 43)

Sample

small chips or pellets; equilibration in hours (p. 43)

Built from

a Y₂O₃- or CaO-doped ZrO₂ plug sealed into an alumina tube, two Pt-Rh thermocouples, a high-impedance millivoltmeter (pp. 43-44)

Original design

Kiukkola & Wagner (1957); geological use Sato (1971) (p. 43)

Suited to

intrinsic ƒO₂ of rocks and minerals, free energies of redox equilibria (p. 43)

In the book

entry pp. 43-45; schematic Fig. 3.14, p. 44

What it does. Doped zirconia conducts oxygen ions, so the emf across the plug reads the oxygen activity difference between sample and reference through the Nernst relation, E = (RT/4F) ln(a_O₂^sample/a_O₂^ref) (p. 45). It measures the intrinsic ƒO₂ of rocks and minerals directly. Watch for iron poisoning of the cell and electronic conduction at very low ƒO₂ (p. 43).

Range

internal pressure set by the charge; usable to the devitrification limit of silica glass near 1,250 °C (Fig. A.1, p. 175)

Built from

a silica tube, a torch, a vacuum line

Suited to

metal-sulfide equilibria, sulfidation of metal and silicates, sealed-tube vapor transport

In the book

one sentence on p. 180, reacting an oxide pellet in an evacuated sealed silica-glass tube when no controlled-atmosphere furnace is available

Extension

the sulfidation arrangement is standard cosmochemistry practice beyond this book, flagged constructed

What it does. Weld one end shut, load the charge, neck the tube, evacuate, and seal with the torch. Sulfur fugacity can be buffered by pyrite-pyrrhotite at one end of a thermal gradient while the sample reacts at the other, and the sealed volume doubles as a vapor-transport cell.

Thermocouples

welded from spooled wire; chromel-alumel calibrated at the NaCl melting point, 800.5 °C, Pt-Rh at the Au point; W-Re for reducing atmospheres to 2,200 °C (pp. 169-170)

Capsules

chosen by melting point and H₂ permeability (Fig. A.3, p. 176; p. 174)

Starting materials

coprecipitated gels fired at 800 °C (p. 178); glasses fused and reground, up to eight cycles for homogeneity (p. 179); freeze-dried mixes (p. 180); oxide pellets (p. 180); natural end-member minerals (p. 181)

In the book

Appendix A.1 to A.5, pp. 169-181

What it is. Every machine above reduces to these parts. The Appendix teaches the parts themselves, which wire to weld, which ceramic survives which atmosphere, which metal passes hydrogen and which blocks it, and how to make a starting material more homogeneous than anything you can buy.