Gold colloids, octahedral platelets, and foils, directly and indirectly formed from the reduction of soluble Au(I) thiosulfate and Au(III) chloride complexes by iron-oxidizing bacteria, cyanobacteria, and sulfate-reducing bacteria, were combined in an experimental system. This system represented simplified biogeochemical conditions occurring within a fluvial environment in which placer Au could occur. In this study, biofilm formation and physical aggregation (i.e., sedimentation processes) were critical for the accumulation of nanometer- to micrometer-sized Au particles into grains 4–5 mm in size. Characterization of grain surface textures by scanning electron microscopy in association with monitoring soluble Au concentrations over time suggested that dissolution and reprecipitation processes were occurring at the Au grain-fluid interface. This laboratory model demonstrates that the biogeochemical cycling of Au can contribute to the formation of anomalous enrichments such as placer Au deposits.


The formation of Au grains within placer deposits has been attributed to either detrital or accretionary processes and has therefore been a topic of critical scientific study (Giusti and Smith, 1984; Bowles, 1988; Clough and Craw, 1989; Youngson and Craw, 1993; Knight et al., 1999; Hough et al., 2007). Whether Au grains are derived from primary sources or produced from the reduction of soluble Au complexes within surface and groundwater systems (Liversidge, 1893; Goldschmidt, 1937), their occurrence in nature is important for resource estimation because the nugget effect (e.g., discontinuity in semivariograms) can overestimate or underestimate the amount of Au in geochemical assays (Stanley, 2008). Gold grains can be mechanically reshaped and altered under physical weathering conditions within fluvial environments; therefore, grain size and shape can be difficult to relate to their primary gold source(s) (Yeend, 1975; Giusti, 1986; Watterson, 1992; Knight et al., 1999). Physical alterations of Au grains include smooth surface textures with round and flat morphologies that are commonly attributed to flow rate and bedload within fluvial environments. Secondary Au structures (e.g., colloids, octahedral platelets, and bacteriomorphic Au) occurring on the nanometer to micrometer scale are generally considered invisible from an exploration perspective, yet are likely ubiquitous in natural environments (Hough et al., 2008). More important, these structures are characteristic features commonly found on grain surfaces, even though they are not always indicative of transport, because abiotic Au precipitation coupled with silver leaching can occur and lead to enriched Au rims on grains (Yeend, 1975; Giusti, 1986; Watterson, 1992; Youngson and Craw, 1993; Southam, 1998; Marquez-Zavalia et al., 2004; Reith et al., 2006; Hough et al., 2008). Freise (1931) first proposed that biological material could act as catalysts for secondary Au formation. More recently, studies involving the exposure of bacteria, archaea, and bacterially produced metabolites to various soluble Au complexes have demonstrated the direct and indirect effects of the biosphere on the geochemical cycling of Au in surface and near-surface environments (see Southam and Beveridge, 1994, 1996; Southam and Saunders, 2005; Lengke and Southam, 2006, 2007; Reith and McPhail, 2006; Lengke et al., 2006a, 2006b, 2007; Reith et al., 2006, 2007, 2009, 2011; Kenney et al., 2012; Johnston et al., 2013). In these studies, biogeochemical processes catalyze the reduction of soluble Au forming secondary Au structures. A model advocating a continuum of primary Au source weathering, secondary Au mineralization, and aggregation can be attributed in part to biogeochemical cycling (Reith et al., 2010). In creating an in-vitro biogeochemical system, the purpose of this study was to grow a gold grain. Bacterial cultures, organic material, host sediments, and secondary Au were combined in an experiment. Here we show that it is possible that the accumulation of bacterially reduced Au could lead to the growth of grains and potentially nuggets in fluvial placer deposits over relatively short geological time scales.


An experimental system was constructed in a Chicago Electric rotary tumbler to represent a simplified biogeochemical model of a fluvial environment in which placer Au deposits could occur. The constituents of this system included a 500 mL fluid phase, 2.7 × 107 bacteria/mL consortium composed of iron-oxidizing bacteria and dissimilatory sulfate-reducing bacteria (SRB), 100 g SiO2 sand as the bedload, and 525.78 mg of Au obtained from previous bacteria-gold experiments. The majority of Au added to the experiment occurred as invisible secondary structures produced from the microbially catalyzed reduction of soluble Au(III) chloride and Au(I) thiosulfate complexes (Fig. DR1.1 in the GSA Data Repository1; see Lengke and Southam, 2006; Lengke et al., 2006a, 2006b, 2007; Shuster et al., 2013, 2014). A 20 mL aliquot of a cyanobacterial biofilm was added every 4 weeks as a source of reactive organic carbon for the growth of aerobic heterotrophs and SRB. Collectively, this bacterial consortium represented a simplified microbial ecosystem. The sand grains ranged from 100 to 200 µm in diameter and were obtained from the supergene Au experiments described in Lengke and Southam (2006). The experimental system was rotated 60 rpm for 24 h once per week for 14 weeks to mimic sedimentation during flood events. The primary objective was to monitor the biogeochemical cycling of Au over time and determine whether Au grains could grow from biogenic secondary Au particles. Solid materials (i.e., biofilms and Au grains) were sampled at four week intervals. The structure of these representative samples was characterized using light microscopy and scanning and transmission electron microscopy. Aqueous 10 mL samples were collected at 2 week intervals and analyzed for soluble Au using inductively coupled plasma–atomic emission spectroscopy (ICP-AES). Note that Au and bacterial concentrations were amplified in this experimental system so that biogeochemical changes could be analyzed in a reasonable timeframe under controlled laboratory conditions. See the supplementary methods in the Data Repository for full details on experimental constituents, set up, and sample processing.


Gold occurred as nanometer-scale colloids, octahedral platelets, and micrometer-scale foils that were not visible to the unaided eye at the initial construction of the experimental system. The pH of the fluid phase remained circumneutral and ICP-AES analysis indicated that an average 109 ± 12 ppm Au remained dissolved over 14 weeks. Soluble Au represented 10.7% of the total amount of Au added to the system (Table DR2 in the Data Repository).

A black biofilm developed and coated the inside surface of the experimental system (Fig. DR2.1). Surfaces of some bacterial cells from this biofilm were covered with nanometer- to micrometer-scale Au particles; these particles also occurred in suspension within the fluid phase (Fig. DR2.2). As the biofilm developed into a more structurally cohesive network over subsequent weeks it accumulated an increasing amount of Au and formed globular grains that ranged from 100 to 500 µm along the longest axis (Fig. DR2.3). The frequency of these grains gradually decreased over time as millimeter-scale grains began to appear within the quartz sand after four weeks. At the end of the experiment, we counted 23 millimeter-scale elliptical disk-shaped grains with average mass of 1.1 ± 0.2 mg (Fig. DR2.4), and 7 cylindrical-shaped grains, 4–5 mm, with average mass of 11.8 ± 0.2 mg (Fig. 1A). The mass of solid Au recovered by gravity at the end of the experiment represented 46% of the total amount of Au added to the fluvial system. Globular, elliptical, and cylindrical grains represented 55.7, 10.4, and 33.9%, respectively, of the solid Au.

The cylindrical grains were composed of multiple elliptical grains that were folded (Fig. 1B) and loosely aggregated (Fig. DR2.5). Topographically high regions on the outer surface of grains contained smooth and rounded surface textures (Fig. 1C). Crevices (i.e., topographically low regions) occurred between the smooth and rounded areas and contained articulated surfaces composed of a wide range of micrometer-scale Au particles, including bacteriomorphic gold structures (Fig. 2A) and clustered euhedral crystals (Figs. 2A and 2B). Extracellular polymeric substances appeared to be attached to surfaces within the crevices and were closely associated with colloidal and crystalline Au (Fig. 2C).


Bacteria and their metabolic byproducts caused the precipitation of elemental Au from soluble Au complexes prior to initiation of this experiment; however, ∼10.7% of the Au was measured in solution (comparable to Southam and Beveridge, 1994, 1996) despite the abundance of cells; i.e., organic carbon–reducing agents. Studies by Baker (1978) and Korobushkina et al. (1974, 1976) demonstrated that humic acids and bacterially produced amino acids can contribute to the dissolution of elemental Au under circumneutral pH conditions, resulting in the formation of soluble complexes. The death and decomposition of cyanobacteria, due to a lack of sunlight for photosynthesis, could have resulted in the release of amino acids that subsequently solubilize Au as organogold complexes. Alternatively, SRB produce hydrogen sulfide as a byproduct of their active metabolism (Reaction 1). Partial oxidation of hydrogen sulfide in the experimental system due to intermittent additions of oxygen (during sampling) could have resulted in the formation of thiosulfate ligands (Reaction 2) that could have leached Au and formed soluble Au(I) thiosulfate complexes in solution (Reaction 3): 

The growth of the biofilm was attributed to the addition of a continuous supply of nutrients from the decomposition of cyanobacteria. The black color was attributed to iron sulfide and the biofilm was interpreted as a bacterial consortium dominated by SRB, since these microorganisms are responsible for low-temperature iron sulfide formation (Trudinger et al., 1985; Shuster et al., 2013). Bacterial cell surfaces typically have a net negative charge and are capable of binding metal (Beveridge and Fyfe, 1985). In this study, the extensive Au mineralization of bacterial cell surfaces could be attributed to either reduction of soluble Au complexes from solution or the accumulation of previously formed secondary Au structures from the fluid phase. From a gold biogeochemical cycling perspective, gold dissolution via amino acid or thiosulfate leaching was counterbalanced by biomineralization. The cycling of gold is important because it provides an explanation for the formation of the secondary gold cement contributing to gold grain formation.

Nanometer- and micrometer-scale secondary Au particles were dispersed throughout the system at the beginning of the experiment. The combination of biofilm formation and episodic sedimentation was responsible for the accumulation of these invisible gold particles and agglomeration of globular grains. When the mass of these grains exceeded the holding capacity of the biofilm, the grains detached and were mechanically reshaped into elliptical-shaped grains in the sand. Continued sedimentation transformed (i.e., rolled, folded, and aggregated) the elliptical-shaped grains into the cylindrical grains. Although topographically high regions of these grains had smooth and rounded surface textures, crevices were protected from physical abrasion and preserved the delicate structure of secondary nanometer-scale bacteriomorphic structures, colloids, and octahedral platelets along with organic material (see Lengke and Southam, 2006; Shuster et al., 2013, 2014). It is possible that abiotic precipitation of Au also occurred within crevices and resulted in the formation of euhedral crystals (Hough et al., 2008). While these larger grains reached dimensions comparable to that of nuggets, their masses were too small relative to their sizes to be classified as nuggets (Fig. 1A; see Hough et al., 2009). Focused ion beam milling demonstrated that they had not compacted during aggregation; however, continued physical compaction (and/or destruction) of these grains would likely occur in natural systems with larger grain sizes and total mass of bedload sediment. The interstitial environments in these grains provide ideal microenvironments for bacterial colonization. The biogeochemical implications of these observations suggest that the growth of bacteria within crevices could lead to the development of biofilms that could potentially lead to additional secondary Au enrichment or dissolution of Au at the biofilm-grain interface.


The biosphere can have a profound influence on the cycling of Au in near-surface environments, including a role in Au grain formation. In this study, a simplified experimental system highlighted the importance of Au biogeochemical cycling and provided evidence of how Au accretion might occur in natural fluvial environments. Initial secondary Au occurred as dispersed nanometer-scale colloids, bacteriomorphic structures, and octahedral platelets and micrometer-scale foils derived from the reduction of soluble Au complexes by bacteria. As the biofilm continued to develop, so did the accumulation of these secondary Au structures. Sedimentation within the quartz sand provided physical partitioning and aggregation of micrometer-sized Au grains into millimeter-sized grains. Characterization of cylindrical Au grains revealed weathered textures on outer surfaces and retention of secondary structures occurring within creviced regions. Remarkably, a large fraction of Au occurred in solution and suggests that bacteria contribute not only to grain formation, but possibly to elemental Au dissolution. This experiment provides a basic perspective of how Au grains and possibly nuggets could form from secondary Au sources within placer deposits.

Electron microscopy and geochemical analyses were performed at the Nanofabrication Laboratory, Biotron Imaging & Data Analysis Module and Biotron Analytical Chemistry Module at Western University, Canada. Funding was provided through Natural Sciences and Engineering Research Council of Canada Discovery (NSERC) and Accelerator grants to Southam.

1GSA Data Repository item 2015038, Item DR1 (supplemental methods including Figure DR1.1) and Item DR2 (supplemental results, Figs. DR2.1–DR2.2, and Table DR2), is available online at www.geosociety.org/pubs/ft2015.htm, or on request from editing@geosociety.org or Documents Secretary, GSA, P.O. Box 9140, Boulder, CO 80301, USA.