Samples collected in the field during exploration must be analized to determine how much gold they contain. There is currently no real-time gold analysis technique available* so samples must be sent to a laboratory for analysis. The time taken to receive results from a laboratory is usually 1 to 3 weeks though during peak times can be more than a month. The term ‘assay’ is usually used in place of analysis and the results called ‘assays’.
There are many different gold assay methods available. The method used by the gold exploration company company will depend on how much gold they are expecting to be in the samples, because any assay method is accurate only within a certain range of values.
The expected level of gold in a sample is a function of what stage of exploration the sample is from. For example, samples from a stream sediment or soil survey will be at levels of parts per billion (ppb) gold. Conversely, samples from a resource drill program will be in parts per million (ppm, 1 ppm = 1000ppb) gold.
While the amount of sample sent by the company to the laboratory may be between 0.2 to 3 kg, the amount that actually gets tested will usually only be 30 to 50g. To get a representative portion of the original sample, care must be taken to carefully prepare the sample into the final ~50 g portion.
Sample preparation involves first drying it in large ovens to drive off moisture. It is then crushed in a Jaw Crusher to a grit sized material (~2 mm grains).
At this stage the sample will be split using a Jones Riffle splitter which allows un-biased division of the material into equal parts.
One part of this split, called a coarse reject or duplicate, will be either stored by the company or disposed of. The remaining split will be pulverized in a rod or ring mill down to the consistency of talcum powder (0.07 mm).
From this powder sample, called a ‘pulp’, a subsample of 30 to 50 g will be used for analysis.
It is important that through this process the sample is not divided with any kind of bias that would allow a particularly gold rich or barren part of the sample to be selected over another. The aim is to get a result that is representative of the entire sample that was submitted by only analyzing a small part of it. While it is possible to analyze the entire original sample of several kilograms this is prohibitively expensive.
The cost of sample analysis will, along with drilling, form the vast majority of exploration costs for a junior explorer.
The most routine gold analysis method is a Fire Assay. This involves several stages. The ‘Fusion’ and ‘Cupelling’ stages use high temperatures, while the finishing technique is a more passive stage usually involving dissolution and ionization of the sample.
The Fusion stage involves taking 30 or 50 g of pulp in a small clay cup, the ‘crucible’, together with chemicals to help melt the pulp and liberate gold when it is heated in a furnace at temperatures of around 1150 degrees celsius. Lead is also added to the mix to collect any molten gold.
After being fired in the furnace, the molten contents of the crucible are poured out and cooled leaving a lead ‘button’ at the base of a silica slag. This lead button is then placed in a cup or ‘cupel’ of magnesium oxide for cupellation. This cupel with lead button is again heated but the lead this time is absorbed into the cupel, leaving a small ‘prill’ of gold and other metals (most commonly silver) that may have been in the sample.
The next stage is where the actual measuring of gold concentration happens, called the ‘finish’.
There are various finishing methods but the most common is Inductively Coupled Plasma (ICP) with Optical Emission Spectrometry (ICP-OES) or Mass Spectrometry (ICP-MS). In both methods the prill is first dissolved and mixed with water, which is then ionized.
The mass of the ions can be measured (as in ICP-MS), or the ions illuminated and the light coming from them measured (as in ICP-OES) which in turn indicates the gold content.