Fusion key step in fire assaying for PGEs

Fire assay is the preferred method for assaying rocks for platinum group elements (PGEs). The classical lead collection method is suitable for determining platinum, palladium, gold and silver in most samples. However, the most common laboratory errors tend to generate low metal values and underestimate grades.

The most crucial step of the fire assay process is fusion, in which a series of flux reagents (soda, borax, silica, flour and lead oxide) are added to a 20-to-50-gram sample and heated to 1,000C. The sample becomes molten and the PGEs are released from mineral grains to be collected by the lead. The lead sinks to the bottom of the molten sample and, when it cools, is separated from the remainder of the sample, called the slag. The PGEs must be transferred to the lead; otherwise, they are lost to the slag, which is discarded.

Samples that contain high levels of chromite, sulphur, arsenic and copper may not be fused properly. The flux reagents need to be adjusted properly to achieve collection of the PGEs in the lead button and some samples may require roasting pretreatment. During the fusion process under oxidizing conditions, osmium, ruthenium and iridium are wholly or partially lost by volatilization. Rhodium losses may also occur.

The next step in the fire assay process is cupellation, in which the lead button, which contains the PGEs, is heated again until the lead is lost by volatilization. At the same time, osmium and ruthenium oxides are also volatilized. The end result is a tiny precious metal bead, more likely the size of the end of a pin rather than the head of pin. The bead must be handled carefully and transferred to a test tube.

Modern analytical methods require that the bead be completely dissolved. Up to a 50% reduction in low-level palladium values has been reported in cases where palladium was readsorbed on the precipitate in test tubes. The metal content of the solutions is determined instrumentally, with the inductively coupled plasma-mass spectrometry giving the lowest detection limits. A common source of small biases is poor calibration of the instrumentation.

Several factors can contribute to underestimation of grades, including: incomplete recovery of the PGEs, volatilization of some elements, readsorption and partial dissolution of the bead. Contamination is the most likely cause of grades being overestimated, and this can be easily monitored by submitting barren test samples. Also, the fire assay method is labour-intensive and involves many steps, and so another concern is that samples could get mixed up.

Nickel sulphide fusions were introduced 30 years ago, principally to provide osmium, ruthenium and iridium assays that cannot be performed by the standard lead fire assay. The cost of a nickel sulphide fire assay is at least 10 times greater than the standard lead fire assay and is not used for routine exploration. The higher cost reflects that fact that nickel sulphide fire assaying is time-consuming and more difficult to perform.

Fluxes need to be adjusted for the presence of nickel, chromite, copper, sulphur and possibly tin. Fusion conditions are carefully controlled for temperature, time and oxygen concentration to achieve a consistent button. Losses may occur if the button is not pulverized sufficiently or if it is not completely dissolved. The nickel sulphide fusion is not suitable for gold.

Standard lead fire assays are used to perform routine platinum, palladium, gold and silver assays. Samples reporting higher assays can be confirmed using nickel sulphide fire assays that also provide ruthenium, osmium and iridium assays. In unusual samples, the PGE contents may be underestimated by either method. Grades can be confirmed by cross-checking at different laboratories, the use of both assay techniques and, if necessary, assaying slags and crucibles.

— The author is the principal of Toronto-based Analytical Solutions.