Cementation of gold onto iron

One of the problems associated with the Anglo elution procedure has been the occasional cementation of gold on the iron pipes and elution vessels. Recently, Kenna [46] has largely elucidated the reasons for this phenomenon, using electrochemical techniques. Initially, it was thought that iron dissolved to form ferrocyanides. In which case, Pourbaix diagrams indicate that geld should cement out over a wide range of pH and potentials. However, Kenna found that cementation occurred only at high temperatures when the pH was allowed to drop below 9 or rise above 14, and was accelerated by the presence of Cl” m the water. The mechanism depends on the corrosion of ircn to Iron oxides. Between pH 9-14 a passive layer of Fe(OHh prevents cementation, but outside this range, soluble iron species exist allowing the iron metal to reduce the gold from solution. Chloride ion interferes with this passive film giving pit corrosion. Thus for most operators, the simple solution is to rubber-line the columns.

Cementation versus eiectrowinning for gold recovery

Because Anglo elution is a batch process a multi-compartment flow-through cell has been designed to treat each batch of eluate. Gold plates onto the steel wool in the normal way but the several banks of steel wool cathodes ensure that the final eluant contains < 1 ppm gold. However although most gold plants in South Africa use these cells, it still takes much time and effort to finally recover the gold from the steel wool. In Australia this problem is being tackled in various ways. Baxter has shown that a pressurised eiectrowinning cell coupled to a pressure Zadra Unit efficiently electrowins gold at 130 degrees leading to high loadings of gold on the steel wool. Currently Costello is examining refining the gold from loaded stainless steel wool by making it the anode of a second cell and plating out gold foil on a steel cathode plate, whilst Biegler is developing a forced circulation cell to produce gold foil directly from dilute eluates.


Meanwhile, in South Africa, zinc cementation is often used to recover gold from batch eluates because it is quick and simple. Recent electrochemical studies have highlighted the important controlling factors and concluded that for this application there was no need for additives of Pb(N03>2 or to de-aerate solutions. Nevertheless a strict range of conditions was still required to avoid passivation or too high consumption of Zn.


Degradation of cyanide

The trend towards using carbon-in-leach, and high temperature elution procedures, has prompted a recent study at Murdoch University on the degradation of cyanide. The extent of the problem is indicated by the 60-70% cyanide loss during the Anglo elution process reported by Davidson. In summary, cyanide can be lost by a variety of mechanisms. For plant operators, cyanide decomposition by oxygen to cyanate and carbonate; and hydrolysis with water to ammonium formate and carbonate, are of most concern. Hydrolysis is important at the high temperatures used for elution.

Initial results from the laboratory show that half the cyanide is decomposed by hydrolysis in 8 hours at 100° whilst at 120 deg the rate is much faster.

Degradation by oxygen is catalysed by carbon and takes place during leaching. As shown the oxidation of cyanide with air and carbon takes place quite rapidly even at 18’C. Half the CN- is lost in 24 hrs; but in the presence of copper the rate is even faster. These results question the desirability of operating a carbon-in-leach circuit and rationalise the significant build-up of CaC03 on carbon as well as evolution of ammonia experienced by plant operators.

Fouling of carbon, regeneration and Elution

In all plant circuits, the carbon activity quickly deteriorates from stage to stage but the reasons why it should do so are complex. It is known that the carbon activity is enhanced by lower pH, higher temperature; the presence of Ca2*, Mg2*, Na*; and by small particle size. On the other hand, its activity is inhibited by excess CN~, build-up of CaCOa and silica, adsorption of xanthates, oils, frothers, huraic acids; and degradatin of active sites [33]. When inorganic foulants are considered, it was found that CaC03 and Mg(OH>2 in particular are adsorbed at high pH, and that copper is adsorbed when the CN~ concentration is low [28]. Silica and iron are invariably present as fine quartz, clay or calcine, which can block the pores of the carbon.

Acid washing with 5-10% HC1 removes many of these foulants but does not restore the carbon activity entirely. As recently shown by La Brooy et al [33] the organic foulants are particularly important, especially the flotation agents, frothers and humic acids. Fortunately most of these get burnt off by heating the carbon, but some require temperatures of 750° to remove them completely.

Carbon regeneration

There is no doubt that the temperature of carbon during reactivation is critical in removing organics and restoring carbon activity and that most plant carbons are not properly regenerated. Reactivation not only removes organics, but also reams out pores, and regenerates surface oxide sites. But as recent studies show, it is important to control temperature, added water/steam, and salts [34]. The key reaction is the water-gas reaction which is rapid above 800° and catalysed by Ca2-, Fe2-. However the type of functional groups produced are also determined by temperature.

Through a proper understanding of key parameters, improved Rotary Kiln designs are evolving, as well as the Electrical Resistive furnace (Rintoul), the Vertical Tubular furnace. Most gold plants currently use rotary kilns with poor control over the carbon temperature. Carbon temperatures vary according to how much water is fed into the furnace with the carbon. It is therefore of interest to compare the various kiln performances.carbon

Elution of gold from carbon

Over the last 10 years there has been much research and development into elution procedures, all trying to improve on the traditional Zadra Process [37] which takes days to complete. Recent studies by Adams and Nicol on the kinetics of elution [38] show that NaCN is more effective than NaOH as an eluant and that high ionic strength solutions retard gold desorption. The optimum NaCN concentration proves to be about 1% w/v

To date, the pressurised Zadra, Anglo, Duval c;nd Micron processes have been commercialised whilst an improved solvent elution system devised at Murdoch University is yet to be developed. All elute the carbon in less than 12 hours using a variety of eluants.

The Anglo process is slowly gaining acceptance over the Zadra process because of its lower overall cost, but the quality of water used is critical to the process [43] which is an important aspect for arid regions. On the other hand the W. Australian Micron process does offer cost advantages in the gold recovery step (see below) and is suited to organically fouled carbons. In South Africa, for example, it is being installed at gold plants which operate a solvent extraction circuit for uranium. The unique advantage of the Micron system (Figure 3) is that less than 1 bed volume of refluxing solvent is used to extract the gold from the carbon charge, and the gold is concentrated in the distillation pot.

The reason why organic solvents are so good at eluting gold is that they increase the reactivity of CN- and stabilise the auro-cyanide in solution.It was the more detailed fundamental knowledge about solvents at Murdoch led to the discovery that acetonitrile elutes gold more efficiently than methanol.

Because the gold is concentrated in the eluant, the Micron process lends itself to elegant gold recovery techniques which give 99.9% pure gold directly. The process selectively discards copper and silver as insoluble cyanides by acidification to pH 3 in the presence of thiourea. The aurocyanide/ thiourea complex is oxidised by chlorine to auric chloride, which is then readily reduced to gold sand by SO2 or sulfites leaving other impurities in solution. Figure 4 shows how a decrease in pH, precipitates out Zn, Ni and Cu cyanides around pH 2-6 leaving gold in solution.

Adsorption of gold onto carbon

Once the gold is leached into solution, there is the problem of selectively concentrating it onto carbon. Again the metallurgists have shown how it works and the scientists are beginning to understand why. Over the last five years the AMIRA Gold Group has been looking at the practical properties of carbon for gold adsorption with postgraduate students looking at some of the fundamental characteristics of carbon.

The essential features of the carbons used today are that they are have and have a graphitic structure with a blend of macro and meso pores. Activated carbon possesses surface oxide sites which ion exchanges with aurocyanide to release OH-, and which also adsorbs cations like Ca2+ and H-.

It is known that more active sites for Au(CN)2- are formed by regenerating carbon at 750“ rather than 550°, and that the carbon potential drops after adsorption of reducible species such as CN- and I-. Carbon also catalyses the air oxidation of CN* into carbonate as discussed below.

Some of the functional groups believed to be formed during activation include carboxylic acid, phenolic, lactone, quinone, hydro-peroxide and chromenol [29] . These groups provide the ion exchange or redox properties of carbon. Aurocyanide is believed to adsorb via ion-exchange with OH” at chromenol sites, Ca2’ adsorb at phenolic sites, whilst CN- is believed to be degraded by peroxide sites.

Significantly the pH of the solution affects both the load ng or au(CN>2′ and associated cations, As shown by Tsuchida less aurocyanide but more cations are loaded at pH < 10. It is believed that adsorption of Ca2* and Mg2* provides excess positive change which allows more aurocyanide to adsorb as an ion-pair.

These ores, in which the gold is locked into pyrite or arsenopyrite, or present as tellurides, are becoming more important. The Kalgoorlie ores in particular, contain a significant amount of gold tellurides. Thus current studies at Murdoch University on the reactivity of gold tellurides are very pertinent [10]. Gold tellurides prove to be unreactive to CN-, though reactive to CI2, and again show evidence of gold and tellurium oxide passivating films. Their electrochemistry proves to be quite challenging [10] . Normally, however, roasting decomposes the tellurides but there is some evidence which suggests that the volatility of telluriates leads to gold plating in the roaster.

In general, refractory gold ores may be treated in a number of ways and over the past few years, 5 approaches have been taken, all with varying degrees of success depending on the ore.

For some ores liberation is no problem and direct cyanidation works well if the pyrite is finely ground. But traditionally, these ores are simply roasted to oxidise the iron and arsenic and liberate the finely dispersed gold for cyanidation. However, the oxidation roast may not liberate the gold fully, or give unexpectedly low gold recoveries. In Western Australia, investigating this problem which is believed to depend upon the roasting temperature, salinity and degree of sintering or porosity of the calcine. It has been found that the calcine becomes less porous as the roasting temperature increases. This is backed up by recent work which showed that regrinding the calcine led to a greater % Au recovered by cyanidation. This aspect is being followed up on local calcines. Furthermore, recent work indicates that the salinity of the ore or ground water may lead to the formation of volatile gold chlorides in the roaster and gold losses in the fume or on the roaster walls.


The problem with oxidation roasting is that it is exothermic and therefore difficult to control the temperature. Thus another approach currently under study is the thermal decomposition of pyrite to pyrrhotite under controlled

conditions. This produces S and avoids the SO2 problem from the roasters. A well developed porous structure is readily observed when arsenopyrite and pyrite crystals are thermally decomposed to remove sulfur.

Much overseas interest has focussed on pressure leaching and bacterial leaching as a means of oxidising the pyrite under controlled conditions to liberate the free gold. Pressure leaching, developed largely by Sherritt in Canada, is no doubt fast and efficient and leaves an insoluble iron oxide or iron arsenate residue. However, bacterial leaching, whilst slower, is a rapidly developing new technology which is promising for certain ores. Initial bacterial leaching studies has now lead to the establishment of a local company, to test and evaluate the process commercially on Australian concentrates.

A comparison of these processes carried out on the Porgera deposit [26] showed pressure leaching to give the highest gold recovery.

In general, the amount of gold leached by cyanidation after oxidation is directly proportional to the degree of oxidation of the sulfide.

Oxidation roasting often gives low values due to sintering of the calcine, whilst bacterial leaching sometimes gives low values because of the mineralogy of the ore. Pressure leaching proves to be most efficient – but it is usually the most expensive option.

As shown, pressure leaching gives high gold recoveries from a variety of ore types, whilst direct leaching variable poor recoveries.

Optimum conditions: 170–190CC 1500-2000 kPa O2

However as pointed out the process of choice may be determined not only by direct capital and operating costs but other factors such as environmental pollution, technological risk, by-product values, and safety – when dealing with arsenical ores.

Leaching free gold with Cyanide

The introduction of carbon-in-pulp technology to the gold industry has rekindled the dormant research and development interests in gold metallurgy. Over the past decade many new developments have emerged which have capital’ :ed on the advances in science and technology in other disciplines. Five years ago, Australian producers were catching-up with the new C.I.P. technology and learning of its idiosyncracies. They quickly realised that Australian conditions did not match those of South Africa and America and that each plant or ore body needed its own refinement Since 1982 several workshop courses on C.I.P. Technology have been presented in South Africa and advances have been made both overseas and in Australia which broaden the scope of investigations as producers struggle to cope with the decreasing grades and increasing complexity of gold ores [4]. Through the Australian Minerals Industry Research Association (AHIRA) and the Western Australian Mining and Petroleum Research Institute (WAMFRI), the industry has funded various applied research projects into various areas of processing – including carbon fouling and regeneration, and gold elution.

Gold metallurgy covers many facets including leaching, concentration on carbon, elution and recovery. This paper attempts to briefly cover each aspect in turn with particular regard to Australian contributions and future directions and developments.

Although leaching with cyanide is not a recent advance, it illustrates the importance of linking fundamental research with plant practice which is essential to all facets of the process. Leaching with cyanide is one aspect which metallurgists have successfully applied for over 100 years yet it is only in the last 20 years that scientists have been able to explain why leaching with CN- is so slow and varies from ore to ore. It is now established that the gold leaching rate is usually diffusion controlled, but depending on conditions, different factors such as CN~ and Oz diffusion, or Oz reduction on the mineral may limit the rate. It is not possible to increase the rate significantly because gold is passivated by cyanide or oxide films.


The mechanism established by Cathro and Koch at CSIRO over 20 years ago involves separate anodic and cathodic processes where gold dissolves via passivating films at one site, and oxygen is reduced at other sites (Scheme 1). Of particular interest was their observation that certain heavy metal impurities can enhance the rate by disrupting the film. This mechanism has been further refined by Nicol in recent years.

The electrochemists can show this heavy metal effect by looking at the corrosion currents of gold at particular potentials. Only trace concentrations of Pb and Hg {~10-6 M) are required to enhance the rate of gold dissolution when typical operating potentials are applied. (Figure 1).

The metallurgist now knows that the use of more CN~, H2O2, O2 or pressure may not give much improvement in rate due to the passive film on gold and that heavy metal additives may help. However many pulps already contain trace heavy metals or may scavange them from solution; thus each ore has to be tested. At the Kambalda plant, for example, Pb(NO3)2 is added to the leach tanks, but it is not clear how effective it is because the presence of sulphate in the water leads to precipitation of PbSO4.