Mineralogy and Geochemistry of the Kinnikinic Quartzite at the Arco Hills Silica and Gold Project in Butte County, Idaho: Results of an Ore Quality Spot Check and Implications for Potential Plasma Furnace Processing

Abstract

Plasma furnace processing has the potential to transform solar cell production. If informed decisions regarding silicon ore and mineral exploration can be made such that waste streams are also of high economic value, then production is ultimately more environmentally integrated. This study presents results from a spot check of the Kinnikinic Quartzite, ~4.5 km east of Arco, Butte County, Idaho (43.639091°, −113.243295°), for ore quality. The mineralogical and geochemical characteristics are explored within the context of a planned plasma furnace project at the sampled site and are compared to previous consulting reports. X-ray diffraction analysis detected only quartz, while scanning electron microscopy identified quartz grains, secondary quartz cement, trace amounts of potassium feldspar, minor iron oxides, and secondary illite. The bulk chemical characterization of 20 samples (including repeats) reports several wt. % variation in SiO₂ from 96.47 to 99.66. Other notable chemical components include Al₂O₃, K₂O, CaO, and Rb, all consistent with the presence of potassium feldspar (and illite). Gold concentrations vary from below detection (n = 12 out of 20) to a maximum concentration of 0.086 ppm. Total sum REE concentrations vary from 13 to 143 ppm. Conservatively and optimistically, assuming ideal extraction and recovery in plasma furnace operation, a resulting waste stream would have approximately 15.2 ppm (0.488 oz./metric ton) gold and 3400 ppm REE in the average waste. Gold (and REE extraction) may, however, be complicated by the presence of Fe and Cu if cyanide approaches were implemented. Gold concentrations are significantly lower than reported in previous work, warranting further characterization of this unit locally and regionally in order to characterize ore potential. This study works to demonstrate the possibility of evaluating other potential silicon ore units, such as the St. Peter Sandstone in Illinois and Missouri, for the co-production of materials in support of an emerging green economy.

1. Introduction

Valuable resources including precious metals and rare earth elements are key to the green economy as it relates to energy production, waste reduction, and pollutant degradation [1,2,3,4,5]. Accordingly, the maximization of exploration efforts and the assessment of non-traditional sources of ores, REEs, and precious metals are increasingly being investigated [6,7,8,9,10]. For example, recycling mine waste for metals and other associated resources has become much more targeted, owing to growing economic interest over the last two decades [11,12,13,14,15]. Ideally, however, multiple green economy resources would be extracted synchronously, i.e. during the same mining process, in order to minimize waste and support the mitigation of economic risk owing to commodity fluctuation. For example, wide price fluctuations have occurred over the past decade for both REEs [16,17,18] and gold [19,20,21]. The study presented here stems from efforts to investigate a planned plasma furnace operation using the Kinnikinic quartzite of central Idaho [22,23]. This once proposed facility would have a primary production volume of silicon metal and several potential byproduct streams, with Au and REEs being of primary interest.

The stratigraphy of the Middle Ordovician Kinnikinic quartzite in Idaho is summarized in Figure 1 [22,23] and has been described as a fine- to medium-grained, silica-cemented, supermature quartz arenite [24].

The depositional environment is associated with a marine, shallow-shelf, high energy setting. Within the context of the broader paleogeography, these regionally extensive arenites record a time period of siliciclastic sedimentation in an otherwise predominantly carbonate dominated regime associated with the Western Laurentian passive margin [24,25,26,27].

The formation’s deposition is specifically interpreted as being associated with moderate to high-energy currents which were in a predominantly southwestward-flowing direction. The Kinnikinic quartzite is characterized by structureless beds with medium to thin parallel bedding. Within this unit parallel laminae dominate and small-scale ripples and shallow scours are rare. Skolithos tubes are uncommon and largely constitute the otherwise sparse trace-fossil assemblage. Previous studies of the Kinnikinic quartzite have focused on its provenance via study of detrital zircons. Extensive investigations of time equivalent deposits have previously been conducted [27,28,29,30,31,32]. These efforts initially supported a sole sediment source for the Kinnikinic quartzite, namely the Peace River Arch of British Columbia, identified by a prominent 1.8–2.0 Ga detrital zircon age peak. Subsequent work in East-Central Idaho identified significant age peaks in the detrital zircon record at 0.9–1.3 Ga, 1.9–2.0 Ga, in addition to peaks 2.0–2.2 Ga, and 2.4–2.9 Ga from >1400 detrital zircon analyses [24]. The identification of multiple age populations precludes the single source model previously proposed and instead supports sediment derivation from multiple sources, e.g. the Trans-Hudson Arch, neighboring Paleoproterozoic-Archean domains, and potentially even the Mesoproterozoic Belt Supergroup [24]. A more recent study assessing the entire Paleozoic and early Mesozoic passive margin stratigraphy reported a dominant age peak of c. 1.8 to 1.9 Ga, with smaller age peaks between 2.0 and 2.1 Ga and between 2.6 and 2.8 Ga [33].

While extensive geochronological constraints on the origin of the Kinnikinic quartzite (and time equivalents) exist, comprehensive mineralogical and geochemical investigations of this formation are lacking in the peer-review literature. The work presented here focuses on a suite of Kinnikinic Quartzite samples from the Arco Hills Silica and Gold Project (AHSGP) in Butte County, Idaho, within the context of validating a spot check of ore quality for a proposed plasma furnace facility. The AHSGP was a 2200-acre site and was being developed by the Arco Hills Silica Company. Previous investigations by Parkinson Geologic Services in 2009 indicated that the property contained 460 million ounces of gold and 20 billion tons of silica. From 2009 pricing, the potential gold resource was valued at $276 billion US and the accompanying silica resource was valued at $500 billion US.

From April 1998, the summary report on 13 buckets of crushed silica ore by Systems Integrated Corporation/M & W Milling and Refining, silica ore is stated to contain 0.031 oz./ton Au and trace Ag. From the June 2002 summary report of Norris Labs/M & W Milling and Refining, silica ore is stated as containing 0.012 oz./ton Au and 0.22 oz./ton Ag based on fire assay. These data also appear to be from a single sample. From the July 2002 summary report of Norris Labs/M & W Milling and Refining, selected dark materials from the silica ore, presumably Fe-oxide rich material, is stated as containing 0.398 oz./ton Au and 3.89 oz./ton Ag based on fire assay. These data also appear to be from a single sample. From three separate reports from Acme Analytical Laboratories, on 19 samples total, the silica content was determined as 97.32 to 98.48 wt. % (July 2006), 98.37 to 98.66 wt. % (July 2002) and 98.54 to 99.48 wt. % (July 2001). All of these summary reports are provided in the Supplementary Materials.

This prior work established an in-place gross value estimate of the proposed site at $776 billion, as determined by Arco Hills Silica Company, based on information provided by Parkinson Geologic Services, who reviewed the above geochemical data. Calculations were checked by Darren Kenny, CPA, using data provided by Parkinson Geologic Services and Arco Hills Silica Company (see Supplementary Materials). The Arco Hills Silica Company continued to seek development and investment of the proposed site in Butte County, Idaho. One investor group, led by Mr. Franco Palma, gained interest in developing the Butte County, Idaho site, for plasma furnace technology, in order to produce high quality Si metal for US domestic solar panel production. The aim was to use the gold byproduct generated by the waste stream to mitigate and offset operational and production costs. As part of an initial validation, Mr. Palma collected 10 large spot check samples from the Kinnikinic quartzite associated with the planned development site in the Spring of 2014. Those samples were supplied for mineralogical and geochemical analysis and form the basis of this study. Analyses were carried out in order to evaluate previously reported silica and gold contents, as well as assess the possibilities for further resource exploration of this region.

2. Methods

2.1. Samples and Preparation

Sample material was provided by Mr. Franco Palma from a single site indicated to him to be representative of the lithology on the planned development site (personal communication). Quartzite samples were labeled 1–10 and small sub-samples were obtained for analysis via scanning electron microscopy (SEM). Duplicates (A and B) were prepared from each sample, and both were used for bulk geochemical analysis and powder X-ray diffraction (XRD) characterization. A Spex Certiprep 8000M Mixer/Mill with a tungsten-carbide grinding vial was used to powder each sample for ten minutes. Alumina and silica mills were avoided, as these were critical analytes.

2.2. Analytical Methods

Concentrations of major elements including silicon (Si) were measured on an Agilent 720ES axial-viewing inductively coupled plasma optical emission spectrometer (ICP-OES) instrument in the Department of Geology and Environmental Earth Science at Miami University [12,13]. The Si 288.1 nm is the largest Si peak on this instrument, being twice the intensity of all other Si peaks and was the wavelength used. A Rh 233.4 nm wavelength was used as the internal standard. Plasma conditions were 1.2 kW Rf power, 16.5 L/min argon plasma flow, 1.5 L/min auxiliary flow, and 0.7 L/min nebulizer flow rates. Samples were introduced to the axial torch using a SeaSpray pneumatic nebulizer and cyclonic spray chamber. Signals were integrated over six 10-second intervals.

Minor and trace element concentrations were determined on a Varian ICP-MS instrument housed in Miami University’s Chemistry and Biochemistry department [12,13]. Approximately ~100 mg of sample material was weighed out into 30 mL Savillex teflon beakers, to which 2 mL of twice distilled 70% HNO₃ was added. These were tightly capped and placed on a fume hood hot plate at 120 °C for one week. The entire contents of each were then transferred into 125 mL polypropylene bottles and diluted to 125 mL, using distilled water made from a sub-boiling quartz still. Bottles and caps were pre-cleaned by rinsing in 1:1 HNO₃ acid, followed by a rinse with 18.2 megohm deionized water, and then dried in an oven at 60 °C. External calibration standards were made from 10 μg/mL solution standards purchased from Inorganic Ventures, which were diluted to 500, 50, and 5 ng/mL using 1% nitric acid. A 100 ppb solution of Ge, In, Re, and Bi was used for internal standardization. Five replicates of twenty readings each were averaged for each analytical solution.

Powder X-ray diffraction (XRD) data was collected in house in the Department of Geology and Environmental Earth Science at Miami University, using a Scintag X-1 powder diffractometer equipped with a Peltier detector, using Cu Kα2 (0.15418 nm) radiation operating at 40 kV and 35 mA. Samples were scanned from 5° to 65° 2θ at 0.02 degrees steps, using a count time of 1 s per step. The detection limit for most minerals using powder diffraction is a few weight percent.

For scanning electron microscopy (SEM) work, images and analyses were acquired using a Zeiss Supra 35 VP field emission scanning electron microscope (FESEM), at the Center for Advanced Microscopy and Imaging (CAMI), Miami University [34]. Carbon adhesive tabs were used to mount sample material on aluminum stubs. Characterization was completed using a variable pressure Zeiss Supra 35VP FEG, with nitrogen (N₂) as the compensating gas. The instrument is equipped with a backscatter detector (BSD) and an energy dispersive spectrometry (EDS) detector (EDAX2000).

3. Results

3.1. Mineralogical Results

Powder XRD patterns confirm the dominance of quartz throughout the sampled suite (Figure 2). Quartz was identified using PDF card # 00-046-1045, using the major reflections and d-values of d₍₁₀₁₎ = 3.3435 Å, d₍₁₀₀₎ = 4.2550 Å, and d₍₁₁₂₎ = 1.8120 Å. Only one sample, 1B, exhibited a trace peak for potassium feldspar, at approximately 3.24 Å, consistent with microcline and other feldspars [35]. Patterns show very little variation, indicating, from the perspective of bulk mineralogy, that the Kinnikinic quartzite is homogenous.

The quartzite samples collectively consist of moderate to well-rounded detrital primary quartz grains that are held together by secondary quartz cement, as established by SEM characterization (Figure 3 and Figure 4). This explains the high purity of the quartzite samples with respect to silica content. Several impurities do occur and include potassium feldspar, illite, and iron oxides (as documented in Figure 3). Comparatively, potassium feldspar, which is rare in the samples (Figure 3B), was investigated by SEM and illite is shown to occur as a late stage pore filling mineral after quartz cement (Figure 3C and Figure 4C). From SEM-EDS (energy dispersive X-ray spectroscopy), illite chemical composition is dominated by potassium, aluminum, silica, and minor magnesium. Iron oxides (Figure 3D) appear to be formed at the same time as, or after, formation of the silica cement.

Approximately 10% of the observed iron oxide grains contain trace amounts of titanium, consistent with the phase magnetite. Areas of the studied samples that contained iron oxides generally appeared to have lesser amounts of, or no, illite associated with them. No phases of environmental concern (such as lead sulfides, arsenic-, uranium-, mercury-bearing minerals) were observed. No gold or silver was observed.

3.2. Bulk Geochemical Results

Bulk major oxide and trace element concentrations for the samples studied, including their repeats, are reported in Table 1.

The concentrations of silica, expressed here as weight percent SiO₂, varied from 96.47 wt. % to 99.66 wt. %, with an average of 99.10 wt. % and an associated standard deviation of 1.36 (n = 20, reported at 2σ). Compared to the previous Acme Analytical Laboratory reports from 2001, 2002, and 2006, these values are broadly consistent, where values, across all three reports, ranged from 97.32 wt. % to 99.48 wt. % SiO₂. Other major oxide concentrations from the samples which form the basis of this study are presented in Figure 5.

The majority of samples contain <1 wt. % of the remaining oxides, with Al₂O₃ (0.17–0.68, average: 0.38, 2σ at 0.25) and K₂O (0.028–0.188, average: 0.087, 2σ at 0.090) being the most abundant of these. This is attributed to the presence of alkali feldspar (Figure 3B). All other major oxides are present at <0.14 wt. % respectively. From Figure 5, it is shown that sample 4 contains higher total abundances of the other major oxides and correspondingly, some of the lowest wt. % SiO₂ values at 98.68 and 98.84 for replicates A and B respectively. These two replicates exhibit the highest K₂O values at 0.19 wt. % and the highest Al₂O₃ values at 0.64–0.68 wt. %. Collectively, these observations are consistent with a slightly higher abundance potassium feldspar. This is also implied by complementary trace element data. Figure 6A,B summarize the relationship between wt. % Al₂O₃ vs Rb (ppm) and wt. % K₂O vs. Rb (ppm), both of which are strongly positive (r² of 0.84 to 0.93).

First, these overall trends, where all samples are considered, are consistent with the variable presence of potassium feldspar (where Rb is incorporated into its crystal structure). Second, the highest concentrations of Rb (ppm) are found in samples 4A and 4B, at 2.41 and 2.49. Furthermore, shown in Figure 6 is the strong correlation between Zr (ppm) and Hf (ppm) at r² = 0.99. This is interpreted to reflect the presence of volumetrically trace amounts of accessory zircon which incorporates Hf into its crystal structure (D_Hf in zircon >900 [36]). Samples 4A and 4B are consistently higher, with Zr (ppm) at 32–41 and Hf at 0.82–0.97.

In Figure 6D, gold concentrations are shown vs. wt. % Fe₂O₃ (for samples in which gold was detected). Of the 20 samples analyzed, gold was detected in eight. It is also noted here that of those eight samples, seven of them are associated with the “A” split, hence, only one sample repeatedly returned detectable gold (Samples 1A and 1B). A maximum concentration of 0.086 ppm (0.003 oz./ton) is reported in sample 2A. When above detection, there appears to be weak correlation with Fe content as summarized in Figure 4D, indicating a potential association. However, this correlation is not statistically very strong (r² = 0.669, p = 0.076). No gold was detected using SEM-EDS in any sample material and no gold was found to be associated with the Fe-oxides. If present in Fe-oxides (see later), the heterogenous distribution of Fe-oxides throughout the quartzite samples could potentially account for the, albeit it statistically weak, trend in Figure 6D.

In Figure 7A–D, the compositions of the sampled quartzite are shown normalized to Earth’s crustal geochemical reservoirs: bulk, upper, middle, and lower continental crust [37]. As shown, the majority of sampled quartzites are relatively depleted in all of the trace elements shown when compared Earth’s crustal reservoirs by 1–3 orders of magnitude, with particularly notable depletions in Th and Sr. The overall normalized trace element patterns of the majority of samples (n = 18) are similar, with no distinguishing features which could otherwise be used to discern samples from one another. The only exception to these observations are samples 4A and 4B, which overall exhibit similar patterns to the rest of the sample suite, with several notable differences: relative enrichments in Th, La, Ce, Nd, Sm, and Eu. Only Nd and Sm are consistently >1 when normalized to crustal reservoirs, but never >4.

4. Discussion

Quartzites in general are not widely established as ore materials, however there are some notable exceptions and contexts. Quartzites can serve primarily as ore bodies for gold, such as the extensive Witwatersrand Basin in South Africa [38,39] and the smaller gold paleo-placers of the Moeda Formation of Brazil [40]. Quartzite is also recognized as both an important historical building stone, such as the Bargiolina Quartzite from the western Alps, northern Italy [41]. The rock type is also a current and future building stone such as that observed in Benin, West Africa [42]. Quartzite also has desirable mechanical properties for use as an aggregate [43]. There are known occurrences of quartzite mining for producing silicon product, including investigations of nanostructured silicon from quartzite ore via hydrometallurgical processes combined with low energy wet blending, in a controlled atmosphere [44]. Quartzite is also extensively mined in Labrador, Canada for export to make silicon metal [45].

4.1. Mineralogy and Geochemistry Implications for Potential Future Mineral Resource Development

The sampled Kinnikinic quartzites of central Idaho are high silica (average wt. % SiO₂ of 99.11, n = 20). Specifically, the silica is associated with quartz grains and quartz cement, both of which contribute to the prominent Si peaks on all XRD patterns (Figure 2), and are well characterized via SEM-EDS (Figure 3 and Figure 4). Minor to trace amounts of illite, potassium feldspar, and Fe-oxides are present as physically observed via SEM-EDS (Figure 3) and identified via bulk major element trends (see earlier). With respect to previous work on the Kinnikinic quartzite from the Arco Hills Silica and Gold Project, bulk wt. % SiO₂ contents reported in this study are consistent with values reported prior (97.32–99.48% SiO₂). Results presented here are, however, inconsistent with previous reports regarding the abundance of gold reported. Average gold concentrations, based on when detected, are 0.04 ppm (~0.001 oz./ton), and for all samples (60% were below detection) are 0.014 ppm (~0.0004 oz./metric ton). Commonly, gold concentrations at or above 5 ppm are sought for mine development [46]. However, some operations owing to scale and ore properties will support exploration at concentrations < 5 ppm, such as the Round Mountain operation in central Nevada with gold ore that commonly varies between 0.2 ppm and 1.2 ppm (Kinross USA Staff, Personal Communication 2015–2019). Even if gold concentrations were an order of magnitude higher in the Kinnikinic quartzite, it is highly unlikely that gold is economically recoverable given overall rock quality: low porosity and low permeability and the tentative association of gold with Fe₂O₃ content. Such ore properties are less than desirable, particularly for cyanide extraction methods [47].

Within the context of utilizing the Kinnikinic quartzite for the development of a plasma furnace operation in Butte County, Idaho, (for silicon metal), other byproducts potentially associated with waste streams are REEs. As gold has already been demonstrated as an economically challenging target within these contexts (see above and discussion later), the REEs are now considered. Chondrite-normalized REE patterns for the sampled quartzite are summarized in Figure 8, alongside Earth’s bulk crustal reservoirs. As shown, the quartzites are light-REE (LREE) enriched (La-Nd) and heavy-REE depleted (Dy-Lu).

Compared to Earth’s bulk crustal reservoirs, they are consistently depleted by at least one order of magnitude. Again, however, there are exceptions associated with samples 4A and 4B, which display concave upward LREE patterns with clear Sm and Nd enrichments compared to crustal reservoirs and overall abundance enrichments compared to the sampled suite.

Rare earth element resources are a crucial component of the world’s economy in both established and emerging markets, and in order to keep up with increasing consumer demand, new sources are going to need to be identified [6,7]. Establishing efficient approaches to recycling REEs is highly desired. Previous work has proposed that the extraction of REEs from mine-related waste could be one such approach [48]. None of the Kinnikinic quartzite samples considered in this study exhibit individual REE concentrations that could be considered economical. When considered within the context of a waste stream from plasma furnace methods, there is the potential for economically significant concentrations to be generated, however this would, of course, be significantly market dependent, and the pricing of REE is known to be volatile [16]. Ultimately, the economic viability of REE materials is likely to be tied to the viability of other resources (such as gold) in a given location. From the Kinnikinic quartzite, the total sum REE concentrations varied from 13 to 143 ppm, with an average concentration of 31 ppm and an associated standard deviation of 33 ppm. Samples 4A and 4B are enriched in total REEs relative to others. In the context of a hypothetical waste stream, assuming 100% removal of SiO₂, and 100% retention of average concentrations of total REEs, waste would have approximately 3400 ppm of total REEs. For sample material such as 4A and 4B, using a value of 143 ppm waste concentrations would be approximately 15,900 ppm. Th concentrations for these scenarios would be approximately 30 to 130 ppm and U concentrations would be approximately 13 to 35 ppm. The processes associated with extraction of REE from different host materials are, however, complex [49,50,51]. Certainly, the form and extractability of REEs in such a hypothetical waste stream would need to be comprehensively evaluated. Although REEs alone are likely not economically viable in such a process, the economics associated with potential gold may enable REEs to be a considered as a viable byproduct in the future.

It is noted here that more extensive mineralogical and geochemical exploration is needed to fully characterize and comprehend the nature of this potential resource. There are implications for plasma furnace utilization and basic economic constraints offered by the results, however, if the Kinnikinic quartzite were to be explored in the future, numerous other factors should also be considered. In particular, the quartzites’ geotechnical properties, fuel costs, and health issues surrounding silica should be exposed to workers in order to prevent or mitigate against silicosis [52].

4.2. Implications for Plasma Furnace Utilization

Prior work has demonstrated that plasma furnace methods for producing silicon metal show promise for the development of large-scale production of solar cells [53,54,55]. Recycling of waste streams that are rich in REE, Au and other metals such as platinum group elements, indium, and gallium within these contexts may be a significant impetus for the broader implementation of recycling practices in the future. In the context of the sampled Kinnikinic quartzite from Butte County Idaho presented in this study, when present, gold concentrations in a hypothetical waste fraction (assuming 100% removal of SiO₂ with 100% retention of gold in the waste stream), would produce a waste product containing between 2 ppm (0.064 oz./metric ton) and 112 ppm (3.600 oz./metric ton) gold, with an average of 38 ppm of gold. For samples studied, our best-case scenarios yield that gold was detected in 40% of the samples, yielding an average gold concentration estimate of 15.2 ppm (0.488 oz/metric ton). Theoretically, this would be the waste stream yield as a result of plasma furnace processing. These are fully acknowledged as crude and optimistic estimates here, based only on the spot check sample set provided. The potential recovery of gold may be significantly much lower, owing to the fact that Fe and Cu would likely also be present in the waste stream. While crude, such gold concentrations would be of interest to mid and large size producers, and may be viable in the context of extracting gold alone, as ore material with greater than 5 ppm concentration is generally recognized as being viable [46].

The cause of the tentative correlation of gold and Fe₂O₃ contents in the Kinnikinic quartzite (Figure 6D) is unclear, but may be density related with gold being associated with Fe-oxides as inclusions or adsorption on Fe-oxides [56]. The presence of gold and Fe-oxide minerals are broadly recognized as having strong associations in a number of other contexts, including the well-characterized Olympic Dam deposit [57,58], in addition to laterite deposits [59]. The extraction of Au from Fe-oxide rich materials is, however, complex [47,60,61]. For future exploration purposes, a reasonable working hypothesis is that higher iron contents in the Kinnikinic quartzite will be associated with higher gold contents. Although highly speculative, and assuming a linear relationship based on the existing data for when gold is present (Figure 6D), a concentration of total iron (expressed here as Fe₂O₃) of approximately 8000 ppm would have an approximate concentration of 5 ppm gold. This has the potential to be an economic ore if certain challenges can be overcome during processing, such as the tight porosity. It is noted that Norris Lab’s result of 0.398 oz./ton on select dark material (refer to Supplementary Materials) is consistent with the observations of the current study, where there is a positive correlation with iron with gold concentrations. Note that the Norris lab material was a concentrate and not bulk quartzite (see Supplementary Material).

The approach of this study could be applied to other known or potential resources. For example, China has used imported silica sand to produce polysilicon [62]. These sources could be further evaluated for potential plasma furnace processing. Moreover, of note is the St. Peter Sandstone across regions of Illinois and Missouri [63,64]. This is an ultra-mature quartz sandstone, which is poorly consolidated and may have more favorable properties for physical processing. Deposits of the St. Peter Sandstone are commonly mined for glass and proppant by minor controlled blasting, excavation and water techniques [65]. The Kinnikinic quartzite would likely require blasting, therefore creating significant dust and an associated possible risk of silicosis or lung disease for workers if not managed properly.

Not only are other geologic formations potentially more suitable for plasma furnace approaches (see above), but there are examples of mine waste recycling which may be more attractive, more economically feasible, and have positive environmental impacts. Tailings and mine waste from quartzite ores can be a significant environmental problem and this is perhaps most evident in South Africa, with extensive impacts on the Witwatersrand reefs as a result of mining operations [66,67,68,69,70]. Limited recycling efforts have been explored within these contexts [71]. Detailed investigations of Witwatersrand tailings for recycling in the context of plasma furnace silicon metal production may be warranted, as there may be gold concentrations of interest. The ore is already granulated and it would be one approach to addressing the associated environmental health issues. Of concern, however, would be the associated uranium content [39].

Data from this investigation supports the position that reserve estimates, put forth by previous consulting reports, likely overestimated the potential gold resources of the Kinnikinic quartzite associated with the proposed site. However, the detailed mineralogical and geochemical work presented here suggests that further work is justified in the context of identifying the variation of (and source of) gold, REEs, and silica resources in the region. However, it appears unlikely that there are economic resources in the immediate vicinity. A broader, systematic mineralogical and geochemical survey of the Kinnikinic quartzite throughout East-Central Idaho and the region (Figure 1B) would generate a more statistically meaningful data set and work to identify potential regions of interest that could establish whether or not the Kinnikinic quartzite is of economic interest.

5. Conclusions

Gold concentrations in the sampled Kinnikinic quartzite of Central Idaho (maximum of 0.003 oz./ton) are inconsistent with those reported previously (0.012 to 0.031 oz./ton) from consulting reports. Silica content are, however, consistent at >96 wt. % SiO₂. Exploration for gold in this system should be considered economically high risk, based on the texture of potential ore and the potential association of iron with gold, making traditional cyanide heap leaching techniques potentially inviable. The observations of gold and REEs indicate, however, that broader, more regionally oriented work, could determine if there are statistically significant trends in elemental concentrations and associations. This would work to determine if better byproduct streams are possible and should be considered. The potential implementation of plasma furnace technology may open new resources for the green economy; however, detailed, integrated studies of targeted formations must be undertaken to understand the variability of potential byproduct composition. Results of this study suggest that detailed mineralogical and geochemical studies of other quartzites formations, the St. Peter Sandstone of Illinois and Missouri for example, are warranted within these contexts.