Table of Contents
The Bureau of Mines’ Intermountain Field Operations Center, Denver, CO; and the Bureau’s Salt Lake City Research Center, Salt Lake City, UT, jointly studied domestic sources of tantalum and niobium as part of the Bureau’s program to reduce national dependency on imported minerals and metals that have key uses in industry and defense.
Presently the United States imports nearly 90% of its annual consumption of tantalum while the balance comes from domestic recycling, and 100% of its annually consumed niobium according to Cunningham (1989). Tantalum, an important component in electronic capacitors, is valued also as a ductile, durable alloying metal that has high temperature and corrosion resistance required for superalloys as also noted by Kock (1989), U.S. Bureau of Mines (1987). Niobium, a significant steel strengthener, possesses high heat resistance and superconductivity properties. Niobium thus finds importance in the aerospace, energy and transportation industries which is noted by Gupta (1984), U.S. Bureau of Mines (1987). Neither metal was mined commercially in the United States from the late 1950’s through the 1980’s. (One domestic site recently reported limited production.)
This study focuses on tantalum recovery from two high-tonnage pegmatite dikes located at Copper Mountain, Fremont County, Wyoming [see Chatman (1989, 1990)]. Surface samples indicate a low grade resource with Ta2O5 concentration about one-third to one-half that of current producing hardrock pegmatite ores as noted by Burt (1979, 1988).
The tantalite assays of the surface samples are not representative of an economic resource. Literature surveys indicate that primary hardrock tantalum deposits generally have grades in the range of 0.05 to 0.15% Ta2O5. Prices in the range of $35 to $60/lb of contained Ta2O5 are necessary to meet the costs of mining and beneficiation of tantalum from these active deposits (see Jones, 1989). Current prices (Har. 1990) are $25 to $28/lb of contained Ta2O5 in tantalite concentrates. This resource could possibly be placed into production in the event of foreign supply interruption. This report describes in general the deposit makeup and deals with the results of beneficiation studies conducted on selected samples. The goal of the study was to determine what grade and recovery of tantalum and niobium could be achieved through a simple processing scheme.
Area Description and Geologic Setting
Copper Mountain is in the Bridger (eastern Owl Creek) Mountains, a part of the Middle Rocky Mountains that adjoin the northern Wind River physiographic basin [see Blackstone (1988)]. The semi-arid southern slope of the mountain is drained by Tough Creek, West Fork of Dry Creek, and uppermost Hoodoo Creek, which are tributaries to the Wind River. Elevations range from 1860 m in the southeast along the West Fork of Dry Creek, to about 2500 m on ridges in the north. The town of Shoshoni is 29 km by road to the south. The Burlington Northern Railroad line is accessible at the Bonneville stop, 4.8 km north of Shoshoni (Fig. 1).
Precambrian-age crystalline rocks that comprise the study area are mostly metasedimentary schist and gneiss with quartz-scheelite veins and replacement zones, quartz veins, and amphibolite dikes indicated by Thaden (1980) on geologic maps. Archean granite intruded those rocks, mostly in the form of granitic dike swarms stated McLaughlin (1940). The granitic rocks are medium- to coarse-grained and composed predominately of quartz, microcline, albite, and muscovite, with accessory biotite and garnet, see Thaden (1980).
Copper Mountain pegmatites formed as a late-stage magmatic differentiation from the Archean granitic intrusion. The pegmatites contain economic minerals from the feldspar and mica groups, beryl, and tantalite-columbite. Other tantalum-bearing minerals present are tapiolite and pyrochlore-microlite Chatman (1989), Harris (1987). Many pegmatites underwent albitization, a process now thought to be important for the formation of economic tantalum concentrations McLaughlin (1940), Holler (1987).
Sampling and Mineralogy
Reconnaissance work during 1987-1989 resulted in 494 rock samples being collected mainly from surface exposures of the pegmatite, granitic rock, and adjoining schist. Samples were analyzed for tantalum and niobium using neutron activation, X-ray fluorescence and atomic absorption spectrometry. Predominant minerals in the pegmatite are quartz, feldspar, and muscovite. Accessory minerals include beryl, garnet, tourmaline, fluorapatite, and lepidolite. Specific identified tantalum-niobium minerals found in the area are tantalite-columbite [(Fe, Mn) (Ta, Nb2)O6], tapiolite (FeTa2O6), and pyrochlore-microlite (Ca2Ta2O7). The tantalum resource sections of the pegmatites have no zonation of a scale that could be segregated during mining. Traditional hand cobbing methods of pegmatite beneficiation generally are not applicable to domestic pegmatites due to high labor cost; these methods would not be applied to tantalum recovery at Copper Mountain due to low tantalum grade there, and the microscopic tantalum mineral grain size. Many of the highest grade samples collected have no tantalum mineral grains visible in field exposures of pegmatite; specimens examined in the laboratory have tantalite and tapiolite grains in the size range of 50 to 100 microns.
The pegmatites, therefore, were evaluated as potential bulk-minable, high-tonnage, low-grade resources. Any mining would take the entire width of the pegmatite structure wherein tantalum concentrations were both elevated and persistent.
When mapping and defining the tantalum resources within the pegmatites, persistent elevated tantalum grade (0.01% Ta2O5 or more) was the primary consideration. Two pegmatite dikes containing tantalum resources were thus outlined: one within the area of the Bonneville 1 and 3 unpatented mining claims, while the other is within the Bonneville 7 and 8 unpatented mining claims as explained by Chatman (1990). These two sources of pegmatite are estimated to contain approximately 0.32 million kg of contained Ta2O5 in 1.27 mega tonne of rock at grades of 0.02 to 0.04% Ta2O5. Estimates are based on a weighted average of sample assays and dimensional measurements of pegmatite exposures. No drilling has been done to date.
Samples obtained for beneficiation studies from the two pegmatite dikes consisted of a mica-rich bulk sample from the western end of the pegmatite on the Bonneville 1 claim (sample A), a composite of surface chip samples at 3 m intervals along a 34.5-m linear section of the pegmatite on the Bonneville 8 claim (sample B), a composite of chip samples collected at several intervals across the pegmatite on the Bonneville 1/3 claim (sample C), and a similar composite of chip samples collected across the pegmatite within the Bonneville 7/8 dike (sample D). Assays of these samples are reported in Table 1.
Sample A, was recovered from an area containing megascopically visible tantalite-columbite crystals, and consisted of coarse-grained quartz, potassium feldspar, beryl, and coarse-grained muscovite mica sheets. The mica represented from 35 to 45% by weight of the total sample which is atypically high for pegmatites in the region. Samples covering a wide region averaged less than 5% mica. Tantalite-columbite crystals occur as single or radiating bladed crystals surrounded by coarse feldspar and quartz. Tantalite-columbite crystals were also noted between mica sheets and at junctions of intersecting books of mica. Fine-grained garnet is also associated with the tantalite-columbite. Individual tantalite crystals examined by electron microscope varied in composition from 23.4% to 37.4% Ta2O5 and from 36.9% to 57.3% Nb2O5.
Sample B, combined from 10 chip samples collected along 30.5 m of strike of the exposed pegmatite along the Bonneville 88 dike, consisted of very fine-grained albite, quartz, and some potassium feldspar. The tantalum minerals, tapiolite, and microlite were widely dispersed in grain sizes of 50 to 100 microns. No beryl was observed, and the niobium assay was less than 0.002% Nb2O5. Fine-grained mica, accounting for only a small percentage of the overall sample, was dispersed typically in the 100 to 200 micron range. Scanning electron microprobe examination of tapiolite crystals showed tantalum as high as 83% Ta2O5.
Samples C and D, assaying 0.04 and 0.025% Ta2O5, were composites of several reconnaissances samples of the Bonneville 7/8 pegmatite and the Bonneville 1/3 pegmatite, respectively. They were used to test tantalum recovery from low-grade material. The mineralogy for the chip samples from the Sample C composite parallels that noted in Sample B, while Sample D’s mineralogy is similar to that noted in Sample A, with the exception of less mica.
Tantalite Upgrading and Recovery
Development of a beneficiation process for tantalum and niobium recovery was initiated on Sample A. Although the mica and Ta2O5 content of the sample was higher than the pegmatite average, sample A offered a good initial evaluation of process requirements. Three general areas of process development were studied: coarse crushing and mica removal, gravity jigging for fine mica removal followed by table beneficiation to separate and recover tantalum minerals from the quartz substrate, and upgrading of tantalite by magnetic separation. The general flow scheme is outlined in Figure 2.
Coarse Crushing and Mica Removal
Sample A material was received as rock chips as large as 15 cm. To maximize coarse mica removal, the material was crushed using the following procedure, as displayed in Figure 3. Coarse rock was dry crushed through a 15-cm jaw crusher to minus 2.5 cm, and then transferred to a hammer mill for size reduction to minus 1.25 cm. The hammer mill fractured the mica sheets and liberated attached silicates and tantalite. Flat mica sheets were removed away from the silicate on a 1.25-cm vibrating screen. The remaining material was processed through roll crushers set at 0.6 cm followed by screening on a 0.64-cm vibrating screen. The procedure was repeated through a roll setting of 1.6 mm and a 1.67-mm screen. The combined coarse mica from the three screens comprised 41% of the feed weight and contained 4.9% of the tantalum and 7.9% of the niobium. This lost tantalum and niobium was considered part of the mica lattice rather than unliberated material. Although a large portion of the mica was removed by coarse screening approximately 10% remained which subsequently caused sizing and separation problems in a gravity beneficiation circuit, so an additional step was incorporated to remove it.
Removal of mica from the minus 1.65 mm material from Sample A was investigated using a conventional jig. Screen analysis of the minus 1.65 mm material identified 90% of remaining mica was concentrated in the plus 0.29 mm material. The plus 0.29 mm material was fed to a jig containing a 1.65-mm hutch screen. The jig bed contained a bottom layer of 0.6 cm steel shot ragging to a depth of 2.5 cm covered with coarse feed material to a total depth of 10 cm. Jigging removed an additional 9.8% of the feed weight and lost an additional 1% of the total tantalum. Later testing indicated jigging only plus 0.59 mm material was equally effective in removing fine mica and reducing problems in subsequent circuits. Coarse mica removal and jigging, resulting in a 51%-feed weight reduction and a total loss of 6.6% of the tantalum. After mica removal was accomplished, testing was conducted on separation of tantalite-columbite from quartz and feldspar by other gravity beneficiation methods.
Gravity Beneficiation
Gravity separation of minerals is an effective beneficiation technique down to a partial size of 74 microns when the concentration criteria of the heavy minerals to the lighter minerals is at least 2.5 or greater said Curnie (1973).
Concentration criteria = specific gravity of heavy mineral – 1/specific gravity of light mineral – 1
Tantalite with a specific gravity of 6.8 to 7.5, and quartz-feldspar with specific gravity of 2.7 to 3.1, fit well into this category. The particular gravity separation method selected depends on the level of upgrading required. Spirals normally operate in the upgrading ratio of concentrate grade to feed grade of 3:1, while tables often have upgrading ratios greater than 200:1. Table beneficiation was thus chosen as the primary upgrading methods.
The concentration of tantalite-columbite was achieved using a 1 m by 1.3 m laboratory shaking table. Initial table testing was performed on Sample A jig concentrate rodmilled to minus 0.29 mm, microscopically determined the liberation size. The feed was split into two fractions for tabling: minus 0.29 mm plus 0.148 mm and minus 0.148 mm. Only 20% of the table feed was greater than 0.148 mm. Although 80.6% of the tantalum was recovered into the table concentrate, and 7.1% to the middlings, there was a loss of 12.2% in the minus 0.148 mm table tails and slimes compared to 0.16% loss to the plus 0.148 mm table tails, indicating that over-grinding had occurred. The coarse grained pegmatite contained a greater portion of the tantalite as larger crystals.
To counteract this problem, a second series of tests was performed by rodmill grinding only the minus 1.65 mm by plus 0.59 mm jig concentrated material and adding a third table which processed the coarser feed size of minus 0.59 to plus 0.29 mm. Table 2 contains data for Sample A table tests with material sized into three fractions below 0.59 mm. Reduced grinding time resulted in only 20.5% of the feed material being finer than 0.148 mm. The overall recovery increased to 83.8% with middling accounting for 10.4%, while the total loss of tantalum to the tails was cut from 12.3% to 5.8%. Mica contained in the coarse size fraction, minus 0.59 to 0.29 mm, did not interfere with tabling operations, but was recovered as a separate tailing and could either be discarded or added to the mica separated by coarse screening. For this study it was included in the table tailings products. In a continuous operation the recovery of tantalum and niobium should increase with the regrinding and recycling of unliberated middlings. One batch table test indicated that the concentrate from coarser tables could be selectively split to produce
high grade concentrate assaying as high as 31% Ta2O5 and a secondary concentrate containing 3.5% Ta2O5. This method was later employed in the pilot plant testing. Concentrates were initially combined for separation testing by electromagnetic methods. Several secondary minerals, i.e. magnetite and garnet, tended to be concentrated with the tantalite during tabling. Separation of these two minerals from the tantalite was achieved using magnetic separation techniques.
Magnetic Upgrading of Tantalite
Tantalite has been shown to be paramagnetic as Rosenblum states (1958) and can be separated from quartz and garnet by electrostatic or magnetic separation devices. Separation of the tantalite minerals from non-magnetic quartz and para-magnetic garnet was tested on a crossbelt magnetic separator. Using a Sample A table concentrate. Figure 4 displays the data for extraction of tantalite and garnet away from quartz as a function of the magnetic gauss applied. Magnetite was separated from the table concentrate at 200 gauss prior to processing material on the crossbelt electromagnetic unit. A minimum setting of 2300 gauss was needed before garnet was extracted from the quartz while a minimum setting of 3400 gauss was required before tantalite was separated from the quartz. Applied magnetic force above 6000 gauss extracted 95% or greater of both the garnet and tantalite. By processing table concentrates at various gauss settings a good separation between garnet, the lesser magnetic tantalite and non-magnetic quartz was achieved. Table 3 contains typical data for the upgrading of table concentrate to tantalite product suitable for feed to a leaching circuit. Table concentrate processed through two passes in the magnetic separator was upgraded from 10.8% Ta2O5 to 25.7% Ta2O5. A small percentage of the tantalite was lost in the garnet and magnetitc fractions. Tantalite lost in the quartz as unliberated material or non-magnetic particles accounted for 2.15% of the tantalum to the magnetic separation. This material could be reground and reprocessed for additional recovery.
Pilot Plant Testing
Pilot plant testing was performed on a semi-continuous basis to evaluate the entire process and gather data for a cost evaluation. The individual batch processes were combined to evaluate entire process without intermediate sampling.
High-Grade Material Sample A
Three 23 kg batch samples of Sample A which had previously been crushed to minus 1.65 mm and treated to remove coarse mica were further processed through remainder of the flow scheme presented in Figure 5. Finer mica from the minus 1.65 mm plus 0.59 mm material was then removed by processing through a 0.3 x 0.3-m laboratory jig with a 1.65-mm hutch screen. Removal of mica by coarse screening and jigging decreased the material weight by 51.2% and increased the table feed analysis from 0.15% to 0.29% Ta2O5. Jig hutch material was stageground to minus 0.59 mm and combined with the remaining feed material to recover tantalite-columbite in the wet tabling operation. Wet vibrating screens were used to separate the ground material into respective size fractions for tabling. Table operations produced high- and low-grade concentrates, middlings and tailings. Coarse middlings were reground in open circuit at 50% solids and sent to the next finer table. Rather than recycle the middling product from the minus 0.15 mm table, the material was processed through a wet magnetic separator at a field strength of 10000 gauss to recover fine magnetic particles. Table concentrates were dried and magnetite was removed by a permanent magnet at 200 gauss. The remaining concentrate was sent through three levels of magnetic separation. Garnet and tantalite were initially separated from quartz at 5000 gauss. Non-magnetic particles were then processed thorough a scavenger unit at 8000 gauss with the magnetic material recycled. A cleaner unit at 2300 gauss separated garnet from tantalite. Separate tantalite, garnet and magnetite products were produced.
Data from table beneficiation of Sample A after mica removal is contained in Table 4.
Tabling beneficiation produced a high-grade concentrate averaging 19.5% Ta2O5 and a low-grade concentrate assaying 2.8% Ta2O5. These two concentrates recovered 96.5% of the tantalum in the feed to the tables. Magnetic separation of the two table concentrates produced a tantalite concentrate averaging 22.6% Ta2O5 and 25.1% Nb2O5. The various products produced by magnetic separation are listed in Table 5. In the overall process recovery of tantalum to the high and low concentrate from Sample A material was greater than 81.7% (Table 6). This is divided between a high grade concentrate assaying 31.1% Ta2O5 and a lower grade concentrate assaying 16.6% Ta2O5. The results from the gravity beneficiation alone indicate that 88.9% of the tantalum was recovered in 2.6% of the feed at an intermediate analysis of 4.8% Ta2O5.
Low-grade Material Samples B, C, and D
The low-grade Samples B, C, and D presented in Table 2 contained little or no mica and insufficient amounts of tantalite crystals which could be liberated at coarser sizes. Samples B and C contained fine grained pegmatite while Sample D contained a coarse grained pegmatite. Liberation size for tantalum minerals was estimated at between 50 and 100 microns. Therefore, the flow scheme used for the lower grade materials was modified as outlined in Figure 6. The use of the jig for mica removal was eliminated and no separate high-grade concentrate was produced on the tables. The amount of niobium in these samples was below the economic range of consideration and thus was considered only a byproduct of tantalum recovery. Material was stage ground to minus 0.59 mm and separated into the same three size ranges as used for processing Sample A. Table concentrates from the three respective size ranges were dried and individually fed to the crossbelt magnetic set at 8000 gauss. Table middlings and the crossbelt separator non-magnetics from the two coarser sizes, (plus 0.29 mm feed, and minus 0.29 mm plus 0.15 mm material) were reground in an open circuit rodmill and sent to the next finer table. Magnetite was removed prior to the crossbelt separator by a permanent magnet at 200 gauss. The minus 0.15 mm non-magnetic material constituted a separate product for this test scheme. The magnetic fraction at 8000 gauss was then reprocessed once at an intensity setting of 5000 gauss and a second time at 2300 gauss. This produced a low- and high-grade tantalite concentrate separated from garnet.
Table 6 contains the overall results of beneficiation for Sample B, C, and D processed through the flow scheme in Figure 6. Several interesting facts were observed in the beneficiation of lower grade materials. The recovery of tantalum into the table concentrate was distinctly lower for the fine grained Samples B and C than for the coarse grained Samples A and D, 61.4 and 64.2% to 88.9 and 84.3%, respectively. The weight of magnetite separated from these lower grade samples was 10 to 15 times that removed from high grade sample A, with an equally corresponding loss of tantalum. The average grade of concentrate produced in the higher grade magnetic concentrate decreased as the feed grade decreased, from 18.8% Ta2O5 in Sample B to 9.9% Ta2O5 in Sample D, indicating a
lack of complete liberation in the lower grade material or an increase of para-magnetic gangue minerals in the concentrate. Distribution of the tantalum to the final concentrate ranged from 16.6% to 20.5% for the low-grade samples. The niobium content of these concentrates was considerably less than Sample A, assaying 1.3%, 0.8%, and 2.0% Nb2O5 in Samples B, C, and D, respectively.
In Samples B, C and D a significant amount of tantalum reported to the non-magnetic product, up to 32% for Sample D. This observation is better understood when a photomicrograph of particles reporting to the non-magnetic recycle material is studied, figure 7. As seen in the photo, a grain of magnetic tapiolite (light color) is surrounded by a non-magnetic rim of microlite (gray), thus limiting the amount of tapiolite removed by magnetic concentration. In general the lower grade materials tended to contain a greater proportion of the mineral microlite. Because it is non-magnetic, a separate gravity circuit would be required to further upgrade this material.
Conclusions
- Sampling and mapping efforts by the Bureau have outlined a low grade tantalum source in a central Wyoming pegmatite. Resource estimates are set at 0.32 million kg of contained Ta2O5 at ore grades of 0.02 to 0.04% Ta2O5. These estimates are from surface chip samples and area measurements. No drilling has been performed in the area.
- Gravity and magnetic beneficiation studies of a coarse grained tantalite-columbite pegmatite sample assaying 0.15% Ta2O5 and 0.16% Nb2O5 showed the material could be upgraded to a concentrate assaying 22.6% Ta2O5 and 25.1% Nb2O5 with recoveries of 81.2 and 79.5%, respectively.
- Beneficiation studies on three lower grade pegmatite samples resulted in tantalum recoveries in the gravity circuit of 61 to 64% for the fine grained pegmatite and 84% for the coarse grained pegmatite with concentrate grades from 0.53 to 5.26% Ta2O5. Magnetic separation proved only partially successful in upgrading these table concentrates to between 9.8 and 18.8% Ta2O5 with recoveries from 17 to 20%. A large portion of the tantalum was associated with magnetite and non-magnetic microlite (Ca2Ta2O7).
Acknowledgements
We would like to thank Mr. Ray Harris from the Wyoming Geological Survey for recommending a study of the Copper Mountain pegmatites, and Mr. Robert Heckert, claimant of the Bonneville claims, for access to the property and aid in obtaining study samples.
