Targeting research to meet commercial pressures is hardly a new scientific malaise. The medieval alchemists sought the conversion of base metals to gold as a fundamental research objective, albeit one that it is difficult for people brought up on the atomic theory of matter to understand. But although nature may not change lead into gold, it can produce the precious metal from the most unpromising starting materials 鈥� ordinary rocks. Geochemical research has begun to show how this masterful piece of alchemy is performed. The key is the recognition of a remarkable link between some active volcanoes, the hot springs and geysers that flank them and the deposition of gold. And this new search for gold has brought a deeper understanding of the interaction of chemical and physical processes involving rocks and fluids, with wide-ranging applications.
Gold is a rare element. The Earth, as a whole, contains about 0.29 grams of gold for every tonne of rock, but this drops to only 0.004 grams per tonne for the continental crust, about 10 milligrams in a cubic metre of rock. Most gold exists as impurities in rocks, as isolated atoms within mineral structures occupying sites that would normally be filled by ions of common metals that are about the same size. There are trace amounts of gold in sulphide minerals, oxides and certain silicates, such as pyroxenes. Ores are rocks that contain far more gold, either as a metal, the form known as native gold, or in other minerals such as gold tellurides, gold alloys and sulphides. A high grade 鈥榖onanza鈥� deposit consists of rock containing 10 000 times as much gold as average.
Most of the world鈥檚 gold has been mined from rocks more than 2500 million years old, in Canada, western Australia, and especially southern Africa. But the new insights into the origin of gold have come from some of the youngest rocks on Earth, still forming in areas of volcanic activity. Here, gold ore bodies are growing today.
Advertisement
The first clues to this surprising source of gold came from the development of geothermal energy as a power source. Superheated steam escaping from the ground near recently active volcanoes is a potent source of energy, first used to make electricity at Larderello in Italy early this century. Exploration in the 1950s and 1960s around hot springs and geysers led to widespread development of geothermal energy in countries with active or recent volcanoes. Larderello turned out to be unusual in supplying steam; most geothermal areas that have been explored contain liquid water at depth. The water fills the pore spaces of rocks such as sandstone or volcanic ash, or lies in fractures within units such as lavas.
In the most active areas, the water is at a temperature close to its boiling point, irrespective of depth. Pressure increases with depth because of the increasing weight of the water above, and water boils at higher temperatures when the pressure is higher. Deeper in the geothermal field, the water becomes hotter, but without boiling; it can reach 300 掳C a little more than 1 kilometre underground. When water escapes upwards, either naturally or along a borehole, the drop in pressure causes it to boil spontaneously or 鈥榝lash鈥�. The steam produced this way can be used to generate power and then condensed back to water. The spent water from some boreholes is very pure, but geothermal discharges elsewhere precipitate mineral scale with concentrations of heavy elements high enough to be ores.
In 1962, drilling at the Salton Sea geothermal area in the Imperial Valley, southern California, produced discharges of very salty brines that had dissolved as much as 35 per cent solids by weight. Over three months, they precipitated several tons of opaline silica containing fine-grained sulphide minerals rich in copper, silver and iron. Springs and discharges elsewhere contain mercury, antimony and arsenic. In the light of this nearly 30 years ago, Donald White of the US Geological Survey realised that the active geothermal systems of today are the ore bodies of tomorrow. One class of ore body, the epithermal deposits, is now recognised as having developed within ancient geothermal systems.
The North Island of New Zealand has been the key area for understanding the link between geothermal activity and the formation of gold ore bodies. The volcanoes of the North Island arise from the convergence of the Pacific plate with the Australian plate. This type of volcano erupts sporadically and explosively. The thick deposits of volcanic ash and debris laid down by such eruptions provide excellent aquifers for water heated by the volcanoes. There are many separate geothermal fields along the length of the Taupo volcanic zone, which extends northeast from Lake Taupo to the Bay of Plenty. Most are associated with volcanic calderas, the sites of craters formed during massive volcanic eruptions over the past few hundred thousand years. Stuart Wilson, of New Zealand鈥檚 Department of Scientific and Industrial Research at Lower Hutt, pioneered chemical investigations of these fields in the 1930s, and a major research effort began when some of them were developed for commercial power in the 1950s.
In 1983, routine maintenance of wells at the Ohaaki-Broadlands field (see map), revealed that the pipes and joints that carried water at high pressure were lined with deposits of scale, very rich in heavy elements and copper, silver, gold, zinc and lead. Concentrations of the metals in the water discharged at atmospheric pressure were no higher than typical ground water, but these deposits showed that the fluid reaching the well head under pressure nevertheless carried abundant metals. Kevin Brown, of the DSIR鈥檚 Geothermal Research Centre at Taupo, experimented with the piping system at one of the wells. He substituted a new back-pressure plate parts in the high pressure part of the system, taking it out after 44 days and analysing the deposits that had built up. He estimated that more than 5 kilograms of scale was precipitated through the piping in this time, containing 150 grams of gold. Brown also established that the sudden flashing that converts hot water to steam releases the gold.
Going for gold
The pipes carrying geothermal waters at Ohaaki-Broadlands established that gold can precipitate quickly. The process happens naturally as well; traces of metallic gold have been found in cores drilled from some deep geothermal wells. But there is a problem in finding out where the gold in the water comes from, because the original igneous rocks that form the geothermal field do not carry free gold. And what makes the gold precipitate as the fluid flashes at the well head or rises naturally through fractures?
The answers lie in understanding more about the chemistry of both gold and the rocks in which it is found. Usually, gold scarcely dissolves in water, which is one reason why it is such a precious metal. But its solubility improves greatly if water contains certain negatively charged ions or groups of ions (known as ligands), which can cluster around gold to form complexes, as described in the Box. A range of ligands that make soluble metal complexes contain reduced sulphur, usually combined with hydrogen.
Reduced sulphur is an important constituent of many geothermal fields. Small steam vents are often encrusted with native sulphur and hydrogen sulphide discharging with the steam gives these areas a memorable smell. The breakthrough in understanding the transport and deposition of gold came from experiments carried out by Terry Seward, then of the DSIR鈥檚 Chemistry Division. In the early 1970s, Seward showed that gold sulphide complexes were much more soluble than gold alone, in the conditions prevailing in geothermal fields. Seward pinned down three types of complex that enhance the solubility of gold. If the water is acid, gold dissolves as the neutral complex Au(HS); in neutral water Au(HS)2- predominates and under alkaline conditions Au2(HS)2S 2鈥� is the dominant gold species in solution.
What appears to happen deep in the geothermal field is that hot water attacks the original minerals of the igneous host rocks. Some constituents dissolve, while the rest are reconstituted into new minerals that are stable in the warm, wet geothermal environment rather than the hot, dry igneous one. Clays, chlorite, calcite and epidote are common products of this alteration process. There are few suitable sites for gold in the lattices of any of these new minerals, so if reduced sulphur is present, any gold in the primary igneous rock dissolves.
The next step is to explain why this gold should precipitate out of solution as native metal later on. Brown鈥檚 experiment on the precipitation of gold at the well head in the Ohaaki-Broadlands field shows that precipitation takes place at the point where the pressure drops and water at 260 掳C flashes to steam and liquid water. At the same time, some of the components dissolved in the water separate into the steam phase, while others remain in solution. Dissolved gases, including hydrogen sulphide, tend to separate into the steam, with a catastrophic effect on the solubility of the gold. The drop in the levels of hydrogen sulphide in the water makes the gold sulphide complexes break down. They release hydrogen sulphide in an attempt to compensate for the gas that has been lost to the steam. This in turn overloads the solution with gold ions; it becomes supersaturated and native gold forms.
Precisely how the gold precipitates depends on what else is going on nearby. At Lake Rotokawa, to the south west of the Ohaaki-Broadlands fields in New Zealand, Seward and another member of the DSIR group, Ralf Krupp, found a surface froth of the clay mineral kaolinite and sulphur, together with particles of arsenic and antimony sulphides. But the froth also contains a number of metals, including silver, mercury, tellurium, lead, zinc as well as gold, probably adsorbed onto the arsenic and antimony compounds. The wind blows the froth so that it accumulates on mud flats on one shore of the lake. Krupp and Seward estimate that 250 kilograms of gold have accumulated over the past 1800 years in surface deposits alone 鈥� truly an ore body of the future in the making.
Once researchers realised that gold ore bodies are forming today in the Taupo volcanic zone, gold exploration gained new impetus. During the past decade, prospectors have found many sites associated with recent vulcanism, especially around the western Pacific, and many other ancient gold deposits are now recognised as having formed under comparable conditions. Some of these epithermal deposits are proving to be as rich and as extensive as some of the Precambrian ore bodies that hitherto provided most of the world鈥檚 gold.
The Ladolam ore body on Lihir Island, Papua New Guinea, developed in rocks where the waning stages of hydrothermal activity still go on. It formed over the past 1 million years in the collapsed caldera of a large volcano; the published reserves of 600 million grams of gold put it in the first division of gold ore bodies. Of course, not all volcanic geothermal fields develop ore bodies; the volcanoes of the Pacific margin are presently the most successful.
Understanding of the role of sulphide complexes in transporting gold has also been applied to the origin of many other types of gold deposit. At Cloggau, near Dolgellau, in Wales, gold is found in quartz veins, but only where they cut through Cambrian black shales. The shales are black because organic debris in the original mud broke down to form fine particles of graphite spread throughout the rock. Within the vein quartz, there are minute bubbles, a few tens of micrometres in diameter, that preserve samples of the original fluid that formed the ore. Most are largely water, but some contain methane as well.
Simon Bottrell, working with me at the University of Leeds, and Tom Shepherd of the British Geological Survey at Keyworth, suggested the gold was transported upwards as sulphide complexes in water. When this hot water met the black shale at around 300 掳C, it produced methane. The methane then separated out from the water, taking the hydrogen sulphide with it in the same way that hydrogen sulphide separated into steam when the geothermal fluids boiled. As in the geothermal fields, loss of reduced sulphur made the gold sulphide complexes unstable and triggered the precipitation of gold.
Sulphide complexing may also explain some of the rich goldfields of southern Africa. Bob Foster and colleagues at the University of Southampton have pointed out that gold in Precambrian iron formations 鈥� predominantly iron oxides and silica 鈥� is present where the iron oxides are replaced by sulphides along fractures or faults. They suggest that the cracks mark the pathways of fluids which carried gold in solution, again as sulphides. Sulphide left the fluid in this case because it reacted with iron oxides to form pyrite 鈥� iron sulphide. As with boiling, removal of sulphide complexes broke up the complexes and precipitated the gold.
Our ideas about how gold is extracted from rocks, redistributed and concentrated by natural processes, are being revolutionised by simple, elegant ideas from solution chemistry, combined with astute geological observations. Understanding the origin of many other types of metal deposits is more complex since they cannot be observed directly as they form. But these geothermal systems have provided an illustration of the chemical processes that form ores.
There is no doubt that a major factor in this advance has been the collaborative work by the geologists and chemists of the New Zealand DSIR in investigating natural geothermal systems and experimenting on mineral solubilities. The response of the New Zealand government to this unique centre of excellence has been emphatic and unambiguous; swingeing budget cuts in the late 1980s severely decreased the work the group could take on. The nucleus of key scientists is now scattered around the world, and the programme of fundamental research that put New Zealand at the forefront of modern geochemistry is finished.
Bruce Yardley is reader in metamorphic geochemistry in the Department of Earth Sciences at the University of Leeds.
* * *
Complex chemistry in the Earth鈥檚 crust
Natural water in the Earth鈥檚 crust tends to be a saturated solution of the rocks which host it, for the simple reason that there is very little fluid. Only the uppermost crust fails to reach this ideal state, because there low temperatures and high rates of fluid flow mean that chemical reactions cannot keep pace. But minerals are complex substances; the way in which they dissolve depends on the composition of the water itself.
An example of a simple reaction to dissolve a mineral is the reaction of the ferrous silicate mineral olivine with hydrogen ions in solution to give aqueous silica and ions of divalent iron:
Fe2SiO4 + 4H+ = 2Fe2+ + H4SiO4(aq)
The standard way to assess how much olivine dissolves in a reaction like this is through its solubility product: this is a measure of the concentration of the dissolved species, iron and silica, in a solution saturated with olivine. But there is a complication. The total amount of dissolved iron or silicon in the solution need not correspond directly to the amount of these species present, because additional iron or silicon may be present in other forms.
Many metals become much more soluble if they are also able to combine in solution with negatively charged ions or ions groups, known as ligands. In the resulting clusters of atoms, known as complex aqueous species, the metal cation is joined to one or more ligand groups instead of occurring in isolation. Of course, this is possible only if the right ligands are in the water. If this is the case, a metal can exist in solution in these complexes in addition to being present as simple cations.
The species that carry the metal may have a positive or a negative charge overall, or be electrically neutral. Some are groups of ions held together by purely electrostatic interactions; positive and negative ions attract and together they are known as ion pairs. Others, known as complexes, are held together by transfer of electron pairs to form dative bonds. These are covalent bonds with a significant ionic character, formed by donation of a pair of electrons from the ligand to the central cation.
Many ore metals form stable chloride complexes, and since chloride is the dominant anion in most deep crustal waters, metals such as iron and copper often dissolve primarily as chloride complexes, such as the neutral FeCl2 complex. Gold is a heavier atom than many common metals and its solution chemistry differs appreciably.
Gold can exist in a number of oxidation states, but in the very reducing conditions that prevail inside the Earth鈥檚 crust away from our oxygen-rich atmosphere, only monovalent gold is important. This gold ion is an example of a 鈥榮oft acid鈥�. The outer electron sheath of the cation deforms readily under the influence of the electric field of an adjacent negatively charged ligand. Such soft, polarisable cations prefer to coordinate with 鈥榮oft bases鈥�, ligands that are similarly easy to polarise, including iodide, cyanide and sulphides, including HS 鈥�, H2S and S 2-.
Gold dissolves effectively as cyanide complexes; this reaction forms the basis of the most common process used to extract gold from ores in solution. But in nature, cyanide is an insignificant component of natural waters, and it is unlikely to be responsible for the transport and precipitation of gold. In contrast, reduced sulphur is an important constituent of natural water from deep underground, and holds the key to gold solubility.