Problems of iron gain and loss during experimentation on natural rocks: the experimental crystallization of five lunar basalts at low pressures

It is extremely difficult to conduct melting experiments on iron-bearing silicate compositions under conditions where the oxygen fugacity and iron oxide content of the charges are controlled precisely at the relevant values, due to reactions between the charge, the container and the adjacent atmosphere. Possible effects are illustrated by discussion of the experimental data for five lunar basalts. At low oxygen fugacities the techniques using molybdenum capsules in an atmosphere whose oxygen fugacity is controlled by passage of a CO$\_{2}$/H$\_{2}$ mixture, and that of enclosing (better, sealing) the charge inside a high purity iron capsule inside a sealed, evacuated silica glass tube yield results which are relatively close to the desired run conditions.