Mineralogical investigations in gold ore beneficiation

Gold has been treasured since ancient times for its beauty and permanence. Because of its superior electrical conductivity and resistance to corrosion and other desirable combinations of physical and chemical properties, gold also emerged in the late 20th century as an essential industrial metal (USGS, 2021). Gold is recovered by cyanide leaching, with various processes preceding this, depending on the complexity of the ore.

The principal gold minerals that are recovered in the processing of gold ores are native gold (Au) and electrum (AuAg), amongst a few more (Vaughan, 2004). In addition, sub-microscopic (solid solution) gold, principally in arsenopyrite and pyrite, is important. Sub-microscopic gold, and very fine-grained gold (<10 μm) locked in sulfides are other modes in which gold occurs.

From a metallurgical perspective, gold ores can be broadly subdivided into free-milling (non-refractory) and refractory types. The former are relatively easy to treat and provide good gold recoveries (> about 90%), whereas the latter are difficult to treat and yield gold recoveries below 90%, and much less than 50% in some cases (Vaughan, 2004). Figure 1 shows a simplified process flow for refractory and non-refractory ores (Youlton, 2020).

Figure 1. A simplified process flow for the treatment of refractory and non-refractory or free-milling gold ores (Youlton, 2020)

Ideally, mineralogy investigations should precede ore beneficiation attempts. Nevertheless, mineralogy is often used to troubleshoot reasons for poor recoveries. The responsibilities of a process mineralogist typically range from identifying constituent minerals as well as rock types, to characterizing ore types in either mine samples or mill feeds. A more comprehensive investigation of ore minerals includes evaluating the textural associations of precious metals within sulfides and/or siliceous gangue phases (Hausen, 2000). Defining the mode of occurrence of gold has become especially important for the reprocessing of tailings for gold recovery.

Conventional fire assay results for gold ores and products of varying complexities could report similar chemistry, however representing very different potentials for recovery. It is for this reason that mineral modal analysis and deportment of constituent gold minerals are key requirements for the prediction of possible recovery.

Gold characterisation is often carried out by means of an automated scanning electron microscope, like the mineral liberation analyser (MLA), whereby the constituent gold minerals are characterised in terms of their gangue associations, grain sizes as well as liberation, all of which dictate the ease with which the gold will potentially be recovered.  The MLA combines backscattered electron (BSE) images with X-ray analysis (Gu, 2003; Fandrich et al., 2007) to allow the mineralogist to distinguish and classify minerals present. A particle map is produced of the mineral of interest (i.e. gold minerals) based on a bright phase search mode of analysis. This allows to extract information on the identified gold mineral, as well as its mode of occurrence in the particle.

Examples of BSE images showing the observed mode of occurrence of gold grains in a sample are shown in Figure 2. With characterisation of ores, details about the minerals of interest that are not always obvious during processing can be sought and presented in order to assist with recoveries and optimisation of processes. The finer gold grains would be more difficult to recover compared with the coarser gold grain.

Figure 2. Back-scattered electron images of gold minerals: top – with spectrum identifying Ag-bearing gold grains included in arsenopyrite; bottom – dominant gold associated with smaller grains of arsenopyrite.

Gold grains are commonly classified into five arbitrary sizes: very fine; fine; medium; coarse; and very coarse, as presented in Table 1. The gold in some cases differs considerably in grain size, ranging from very fine to coarse-grained within the same deposit, whereas, in other deposits it is more homogeneous, ranging within one or two size classes (Hausen, 2000).

Table 1. Arbitrary classification of gold grain sizes

Refractoriness in gold ores may be caused by several factors including the presence of base metal sulfides, Au-Ag tellurides, cyanicides and carbonaceous matter, all of which can be characterised mineralogically. When the ore contains both sulfides and carbonaceous matter it is said to be double refractory (Amankwah et al., 2005). In refractory ores, gold grains may sometimes be occluded in the sulfide minerals, and pre-treatment becomes necessary in order to decompose the mineral structure to liberate gold for subsequent recovery by cyanide leaching (Figure 1).

In cases of sub-microscopic gold occurrences, mineralogical analyses may need to include electron-microprobe analysis (EMPA) or even laser ablation inductively-coupled mass spectrometry (LA-ICP-MS) analysis in order to determine the proportions of gold contained in sulfide phases, primarily pyrite and arsenopyrite, for deportment determinations. When probed, the gold contained within the different sulfide phases can be determined as is presented in Tables 2 and 3, and gold deported to sulfide phases can be established (Table 4). In order to establish the proportions (in %) of gold deported to sulfide phases, the modal proportion of the sulfide mineral in question is multiplied by the average Au EMPA value and then divided by a hundred. The proportions may then be normalised for the sulphides, as in Table 4. The discrete gold grade attributed to sulphide hosts may be provided as a proportion of the total gold grade of the sample to estimate potential losses in gold recovery using conventional leaching.  In the absence of EMPA or LA-ICP-MS methods, gold contained in sulfides (and oxide phases in some cases) may remain unaccounted for whilst actively contributing to “poor” or lower-than-expected recoveries.

Table 2. Compositional analysis for gold-bearing arsenopyrite (wt.%)

Table 3. Compositional analysis for gold-bearing pyrite (wt.%)

Table 4. Gold deportment in sulfide minerals

A mineralogical investigation of gold ores can also help unravel issues of nugget effects, where chemical assays return a high gold grade value for an ore whose gold mineral content is not necessarily commensurate with that suggested by the chemistry. The beneficiation of gold ores should begin by examining and classifying the types of gold occurrences and possible recovery methods. In this respect, gold grain size and textural association with gangue minerals, including possibilities of sub-microscopic gold must be investigated. The factors stated above highlight the relevance of mineralogical investigations in gold ore processing so as to predict process behaviour, address potentially problematic aspects, and thereby enhance recoveries from a variety of ore types.

REFERENCES:

Amankhwah R. K, Yen W. -T, & Ramsay J. A, (2005): A two-stage bacterial pre-treatment for double refractory gold ores, Minerals Engineering, 18, 103 – 106

Fandrich R, Gu Y, Burrows D. and Moeller K., (2007): Modern SEM-based mineral liberation analysis. International Journal of Mineral Processing 84, 310-320.

Gu, Y, (2003): Automated Scanning Electron Microscope Based Mineral Liberation Analysis, Journal of Minerals & Materials Characterisation & Engineering, Vol 2, 33-44.

Hausen D. M, (2000): Characterising the textural features of gold ores for optimizing gold extraction, JOM, Microtextural Mineralogy, 14-15.

Vaughan, J.P, (2004): The Process Mineralogy of Gold: The Classification of Ore Types, JOM, Gold Process Mineralogy Part 1, 46-48.

Youlton, K, 2020. The Study of Alternative Leaching Processes on South African Gold Ores. Unpublished thesis

Online:

https://www.usgs.gov/centers/nmic/gold-statistics-and-information, DA: 25/02/2021