Stoichiometry FAQ

How do you achieve precise stoichiometry in Alloying?

Generally alloys are provided with a nominal composition, that is, it is assumed what is put in the melt is what the ends up in the alloy. For arc-casting of elements, alloy constituents are weighed to ±0.0005 gram accuracy, and then the cast weight is compared to the input masses.  The table below demonstrates a mass balance:

Calculated Mass
Actual Mass IN
Percent Difference
31.5521 g
31.5521  g
15.6200  g
15.6205  g
2.9363  g
2.9362  g
50.1084  g
50.1008  g
50.0917  g

We assume that the element is 100% pure in the calculations. This is the first error since there is no such thing as 100% purity. The second error is that the input masses are not equal to the calculated mass. In the example above the alloy lost 0.0091 grams of mass. Until further tests are made the effect of this loss is unknown. To compound the problem, when using an element with a high vapor pressure at its melting point, volatility becomes an issue.

The only way to achieve a high degree of confidence in compositional accuracy for bulk processing is to make a material, measure mass losses, analyze the material for some key factors (ICP, XRD, metallographically, etc. - see next section), make an educated adjustment to the mass balance if necessary, and then make the material again. This is why materials research can quickly get expensive.  


Will the true stoichiometry please stand up?

We may never know the true value of the elements in an alloy; however, one can use an array of other tools to conclude the stoichiometry is correct. XRD demonstrates the expected structure and lattice constants, physical property measurements are consistent with expectations (magnetism, heat capacity, and so on), and metallography demonstrates expected microstructures, etc. The degree to which the collected evidence is self-consistent will determine confidence in the stoichiometry of the alloy.

The above casting example is from a hydrogen absorption alloy for the ESA Planck Mission.  Mass losses were specified to not exceed 0.1%; microprobe line scans were use to check homogeneity following annealing; XRD was used to check for the expected structure and lattice parameters, and finally isotherm data on H absorption was used to qualify the alloy performance for acceptance.

Wow, that is a lot of work! Yes, but if stoichiometry is critical the additional work is necessary.