Where does gold come from?

Quartz with an exceptionally rich deposit of macroscopic
gold grains
This image from Nevada Outback Gems web page at
http://nevada-outback-gems.com/prospect/gold_specimen/Gold_ores.htm

In general, gold is sparsely distributed in the earths crust, having a typical concentration of about 2 parts per billion (ppb). How does it become concentrated to form deposits in which it is more than a thousand times as abundant, the deposits that we mine? In a new paper* in Nature Geoscience, Dion Weatherley and Richard Henley address this question.

A typical fault jog cavity, now filled with quartz
From the referenced article.

     The observation that underlies the theory is that there is a typical geometry of quartz veins found in fault zones formed during times of mountain building ranging as far back as 3 billion years. Faults are not simple planar surfaces. Rather, they jig and jag forming irregular and complex surfaces. When a fault ruptures, causing an earthquake, some parts of the fault slide past each other smoothly; other parts bump into each other; and, importantly for the gold-story, some parts separate to form open cavities. At the instant of their formation, these cavities have extremely low pressure, causing any fluids in the vicinity to vaporize (so-called "flash vaporization"). When the fluid vaporizes, silica and its dissolved trace elements, including gold, are rapidly deposited.  Nearby fluids flow toward the cavity, and vaporization continues, until the pressure is restored to ambient (presumably lithostatic) conditions. For  case study, the authors took ambient conditions before a fault rupture appropriate to a depth of 11 km: P=290 MPa, T=390 C.  They used a simple piston model for the formation of the cavity during strike-slip motion of a vertical fault, and obtained the change of volume of the cavity in the jog by looking at the seismic moment. They assumed that the cavity initially contained water, and had a width of 100 microns. Using the moment magnitude of an earthquake to calculate a volume change for the cavity, the authors could then, using the perfect gas law linear relationship between volume and pressure change, calculate the pressure drop. They concluded that during a magnitude 4 displacement with a slip of 0.13 m, the pressure would drop from 290 MPa to only 0.2 MPa.
     Quartz solubility is strongly dependent on the density and pressure of the water in which it is dissolved. Such large pressure drops cause extreme supersaturation of SiO2--up to factors of 100,0000--which the authors propose nucleates by forming grains of less than 100 nanometers in diameter.  Silica deposition occurs through a succession of metastable polymorphs including hydrated silica. These phases evolve during the pressure recovery around the initial cavity. The authors propose, but do not develop quantitatively, that these large drops in fluid pressure cause a complicated "cascade of coupled, nonlinear chemical responses" that result in the complex assemblages that are found in veins. Precious metal solubilities depend strongly on pressure.
     Isolated slips do not deposit huge amounts of gold, but rather result in the deposition of only a thin coating of silica-gold vein material. The mass of this material is increased during the recovery stage.  The authors suggest that it takes tens of thousands of years, but less than 100,000 years to form a high-grade deposit.

*Weatherley, D.K. and Henley, R.W., "Flash vaporization during earthquakes evidenced by gold deposits," Nature Geoscience, online March 17, 2013, DOI: 10.1038/NGE01759.