Ceramic injection-molding and Multi-anvil research

Injection-molding is used to create complex shapes from ceramics. The ceramic powder is combined with molten wax and injected into a mold under pressure. After molding, the wax is removed and the ceramic sintered at 1500 degrees C (the size of the mold is corrected for shrinkage of the ceramic). The cost benefit of injection-molding is greater when the final shape is more complex, because the price of injection is independent of the shape of the piece. For multi-anvil work, octahedra can be injection-molded at between 5 and 10 dollars per piece, giving a substantial savings over the machined pieces that are normally used. Thus the octahedron is the most important part to be made by this process. Sleeves can also be made, but at a similar price, so that there is less benefit over machining.

For the octahedra, a clever choice of ceramic must be made. It should not have high-pressure phase transitions that damage the pressure generation. The porosity has to be near 20-30 percent. It helps if it is a good thermal insulator, so that the heater does not draw too much power.

Composition MgO-ASU: more details

We tried pure MgO first, but the sintering density was only 50 percent, too porous for multi-anvil work. In an attempt to duplicate the Aremco formula (502-1550 Old Formula) we mixed alumina and MgO; during the sintering, all the alumina reacted with MgO to form spinel, increasing the density to 80 percent. This ceramic is now being used for multi-anvil research, and has been made in the 14 mm, 10 mm and 8 mm size, with holes already in them (see below for more details). Another composition that is potentially of interest is porous mullite MUL6 (see Ceramco web page); however, this has a different shrinkage coefficient and no correct molds have been built for this composition yet.

The starting ceramic powder for MgO-ASU is 1/3 alumina, 2/3 magnesia by weight, with a little talc for sintering. The result is a mix of spinel (MgAl2O4) and MgO, in the ratio appropriate for the starting material. The ratio has been checked by Rietveld refinement of the final product.

The benefits of this ceramic are good x-ray transparency compared to zirconia, and a somewhat lower thermal conductivity than pure MgO. It has been shown to give a fairly good heater efficiency even in the absence of a zirconia sleeve (Kung). It also has no phase transitions until some 30 GPa.

The ceramic has not been fully tested for blowout frequency and other difficult-to measure properties; this work is ongoing at various labs, along with the pressure calibration of the octahedra.

Because of fairly hard sintering and the presence of spinel (Mohs hardness 8), diamond tooling is needed to cut thermocouple grooves in the octahedra made from MgO-ASU. Wheels and disks such as those made by Pfingst are recommended.

Composition MUL6

This ceramic is a stock ceramic from Ceramco. COMPRES has purchased some test blocks of it, but octahedra have not been made yet. It is a potential target ceramic because of its nice porosity (30 percent) and low thermal conductivity. However, there are phase transitions in mullite, and their effect on pressure efficiency are unknown.