Textural characterization of gold in the
Björkdal gold deposit, northern Sweden.
Fredrik Westberg
Natural Resources Engineering, master's
2021
Luleå University of Technology
Abstract
The Björkdal gold deposit is located in the eastern part of the Skellefte district, northern Sweden. Twenty thin sections from four production areas in the open pit and four drifts from the underground mine were analysed for mineral association and grain size distribution of gold. In addition, the texture of gold was investigated in order to find out how that affects the recovery of gold.
The overall gold grain size distribution shows an interval from very fine-grained (2 µm) to coarse-grained (856 µm) while the overall median size is 7 µm. Gold from the Quartz Mountain production area displays the smallest median size of 4 µm, whereas gold from the sampled drifts at 340m- and 385m- level has the largest median size of 14 µm. Gold at grain boundary is the dominant textural mode of gold from all sampled locations and varies from 62% to 92%. This is followed by intergrown which ranges between 8% and 29%. Of the sulfides, pyrite, chalcopyrite and pyrrhotite are the most common. Galena and was also present in the samples.
Gold is significantly and positively correlated with tellurium (Appendix 10.1.1), and weakly positive correlated to silver and mercury. Gold show a close association to bismuth-tellurides in the samples. Apart from native gold, which is the dominant mineral phase of gold, two additional gold-bearing tellurium minerals were detected with SEM-EDS, a Au-Te-mineral and a Ag-Au-Te-mineral. One additional bismuth-telluride mineral aside from the most commonly occurring tsumoite (BiTe) was also detected with SEM, with a elemental composition of Bi-Te-S.
Liberated gold in the tailings was optically identified in two thick sections, TB1-02feb-1 and TB1-07feb-1 (Fig. 32A and B), where the flotation circuit failed to float the free gold. One grain of gold was also identified intergrown with bismuth-telluride as an inclusion in silicate (Fig. 33), where the flotation properties of the larger silicate grain likely dominated in the flotation process.
This thesis highlights the importance of further quantitative analysis utilizing SEM/QEMSCAN/MLA to retrieve representative mineralogical data to benefit the mineral processing of the ore from the active mine.
Keywords: Björkdal gold deposit, gold, gold-telluride, SEM, mineral association, grain size,
Abstract i
1. Introduction - 1 -
2. Geological setting - 2 -
2.1 Regional geology - 2 -
2.2 Björkdal gold deposit - 5 -
3. Method - 8 -
3.1 Sampling the underground and the open pit mine - 8 -
3.2 Sampling of tailings from the processing plant - 9 -
3.3 Sample preparation - 11 - 3.4 Petrography - 11 - 4. Results - 12 - 4.1 Petrography - 12 - 4.1.1 West pit - 12 - 4.1.2 Quartz Mountain - 15 -
4.1.3 South East Wall - 18 -
4.1.4 East pit - 18 - 4.1.5 285-1365-eb - 21 - 4.1.6 325-1420W - 24 - 4.1.7 340-1320E - 27 - 4.1.8 385-539W - 30 - 4.2 Tailings - 34 -
4.3 Scanning Electron Microscopy - 35 -
4.4 Whole rock geochemistry - 39 -
5. Discussion - 41 -
5.1 Mineral association of gold - 41 -
5.2 Textures of gold - 43 - 5.3 Geochemistry - 44 - 6. Conclusion - 46 - 7. Recommendations - 47 - 8. Acknowledgements - 47 - 9. References - 48 - 10. Appendix - 51 - 10.1 Correlation matrix - 52 -
10.2.2 Tellurium correlation coefficients and significance for selected trace elements - 54 - 10.2.3 Bismuth correlation coefficients and significance for selected trace elements - 55 - 10.2.4 Silver correlation coefficients and significance for selected trace elements - 56 - 10.2.5 Silicon dioxide correlation coefficients and significance for selected trace elements - 57 - 10.2.6 Sulfur correlation coefficients and significance for selected trace elements - 58 -
- 1 -
1. Introduction
The Björkdal gold deposit is located in the eastern part of the Skellefte district, in northern Sweden. The deposit was discovered by Terra Mining in 1983 after an intense period of exploration in the Skellefte district. Terra Mining discovered anomalous gold concentrations in glacial till sampling. Follow-up work discovered anomalous gold values also in the bedrock 1985 (Weihed et al., 2003). The Björkdal gold deposit went into production in 1988, as an open-pit mine, and expanded in 2008 with the opening of an underground mine. Throughout the history of the mine, several companies have taken ownership of the mine, and it is currently owned and operated by Mandalay Resources. Gold ore can be divided into free-milling and refractory with regards to mineral processing, where gold bound to quartz generally is to be considered free-milling (Prasad et al, 1991; Vaughan, 2004; Zhou and Cabri, 2004). Quartz vein gold ores can often be effectively recovered by gravity separation coupled with flotation/cyanidation after grinding due to the high specific gravity of gold (Zhou and Cabri, 2004). Gold texture, grain size of gold and gangue mineralogy affects the recovery (Harris, 1990). The gold recovery of the Björkdal processing plant is generally over 88% (Pressacco et al., 2018).
The purpose of this study is to investigate the grain size distribution and mineral association of gold, to aid in the mineral processing of the ore. The study identifies which textures dominate for gold and how these textures affect metal recovery.
- 2 -
2. Geological setting
2.1 Regional geology
The Fennoscandian Shield (Fig. 1) makes up the northwestern part of the East European craton and consists of bedrock ranging from Archean rocks in the northeast, to Paleoproterozoic rocks covering large parts of Sweden, Finland and the northwestern part of Russia (Gaal and Gorbatschev, 1987; Weihed et al., 2005).
Figure 1. Major geological domains of the Fennoscandian Shield with significant gold deposits, location of the Skellefte district represented in Fig.2 is indicated by the rectangle. Map modified after Bark and Weihed, 2007.
The Fennoscandian Shield is host to various ore districts where the Skellefte district, located in the northern part of Sweden (Fig. 1) is one of the most important mining districts in Europe (Kathol and Weihed, 2005).
- 3 -
the most well-known deposit (Kathol and Weihed, 2005). The Björkdal gold deposit is located in the eastern part of the Skellefte district (Fig. 2).
Figure 2. Bedrock map over the Skellefte district, modified after Weihed, (2001).The Björkdal gold deposit is indicated.
The Skellefte district is defined by the extent of the Skellefte Group volcanics (yellow in the map, Fig. 2), but the district also hosts sedimentary and intrusive rocks. The district has a general trend of north-west to south-east (Fig. 2). The Skellefte district is bordered by the Arvidsjaur Group rocks in the north, which consists of felsic to intermediate volcanic rocks and volcaniclastic sedimentary rocks (Allen et al., 1997). Towards the south, the district is bordered by the Bothnian Supergroup,
dominated by greywackes with minor mafic and felsic rocks (Fig. 2) (Kathol and Weihed, 2005). The rocks in the Skellefte district have been metamorphosed in greenschist to lower amphibolite facies and the metamorphic grade generally increases in the direction of the Bothnian Basin in the south (Weihed, Bergman, & Bergström, 1992).
- 4 -
Figure 3. Stratigraphic view of the Skellefte district modified from Kathol and Weihed, 2005.
The Skellefte Group volcanics is separated into a lower volcanic formation made up mainly by felsic volcaniclastic rocks, and an upper volcanic formation containing felsic and mafic rocks displaying variations in composition around the district (Weihed et al., 1992; Allen et al., 1997). The Skellefte Group is covered by a sedimentary unit of mostly fine-grained greywackes coarsening to conglomerates higher up in the stratigraphy, this unit (Vargfors Group) is laterally equivalent with the sedimentary Bothnian basin to the south (Weihed et al., 1992).
The Skellefte Group is overlain by the Vargfors Group (Fig. 3), and the contact varies in different parts of the Skellefte district, from conformable to unconformable as well as interfingering (Allen et al., 1997). The Vargfors Group mostly contain near-shore coarse epiclastic rocks as well as turbiditic greywackes (Weihed et al., 1992; Allen et al., 1997). The Vargfors Group is suggested by (Weihed et al., 1992) to have formed coeval with the more juvenile Arvidsjaur Group, situated to the north of the Skellefte district.
The Skellefte district has been intruded locally by several generations of granites where the calc-alkaline I-type Jörn GI dated at 1888−14+20 Ma (Wilson et al., 1987) is of the same age as the Skellefte
- 5 -
is of bimodal composition with a felsic centre and an outer mafic margin dated at 1875 Ma (Skiöld et al., 1993). The Skellefte-Härnö granitic suite is of S-type composition and dated at 1.82 ─ 1.80 Ga ( Claesson and Lundqvist, 1995; Billström and Weihed, 1996). The Revsund granitoids of A- and I-type composition is between 1.81 and 1.77 Ga (Claesson and Lundqvist, 1995; Billström and Weihed, 1996; Eliasson and Sträng, 1997).
2.2 Björkdal gold deposit
Surrounding the Björkdal intrusion (Fig. 2), large quantities of sedimentary rocks of the Bothnian Supergroup occur, estimated to be older than 1.95 Ga (Wasström, 1993). The Björkdal intrusion is dome-shaped and bordered to the west and north by felsic volcanic units, to the east, south and north by mud-, silt- and sandstones with bedding surfaces mainly dipping away from the intrusion (Weihed et al., 2003).
U-Pb dating of zircons yielded 1.90 Ga as the age of the intrusion, which would make it co-eval with lower and upper volcanics of the Skellefte Group (Billström et al., 2009). The intrusion constitutes medium-grained quartz-monzodiorite to tonalite with plagioclase, quartz, biotite and amphibole as main minerals (Åberg and Weihed, 1999). Accessory minerals are mainly calcite, epidote, apatite and zircon (Åberg and Weihed, 1999).
- 6 -
Figure 4. Structural geological map over the Björkdal area (Weihed et al., 2003). The Björkdal pluton is the grey area in the centre, with the open pit outlined in red.
As suggested by Broman et al. 1994 and Billström et al. 2009, three generations of fluids affected the Björkdal deposit between 1.9 and 1.8 Ga, based on oxygen and hydrogen isotope analyses of CH4, CO2 and H2O in fluid inclusions; an early fluid phase, an ore phase and a remobilization phase. The early phase was mainly dominated by quartz and minor amounts of scheelite, sulfides, gold and likely biotite (Billström et al., 2009). The main ore fluid phase deposited the majority of sulfides, alteration-related silicates, bismuth-tellurides and gold, the final fluid phase remobilized the earlier deposited minerals (Billström et al., 2009).
The auriferous quartz veins of the Björkdal deposit are near vertical with the main set generally striking north-northeast with a subset striking east-northeast (Weihed et al., 2003). The quartz veins host varying amounts of sulfides, biotite, actinolite, chlorite, tourmaline, scheelite and bismuth-tellurides, and elevated gold grades are usually found in association with scheelite, bismuth-tellurides, sulfides and tourmaline (Broman et al., 1994; Åberg and Weihed, 1999; Billström et al., 2009). The sulfides in the quartz veins mainly consist of pyrrhotite, pyrite and chalcopyrite, with limited amounts of bornite, covellite, copper and galena (Broman et al., 1994).
- 7 -
- 8 -
3. Method
3.1 Sampling the underground and the open pit mine
Grab samples were taken from the active mine area, both from the underground mine and the open pit. The sample locations were selected from previous and currently active productive zones. For each sample, between 2 and 5 kg of rocks were collected for geochemistry (whole rock analysis) and petrography. A total of 20 samples were collected, 12 samples from four production areas in the open pit (Fig. 5) as well as 8 samples from four locations in the underground mine (Fig. 6). Samples from the open pit were labelled by blast number, underground samples were labelled according to level and drift.
Figure 5. Simplified topographical map of the open pit. Black lines mark the border of the open pit with main access roads in red. Four production areas coloured, and sample locations marked, north arrow indicates mine grid north.
- 9 -
Figure 6. Simplified map showing the 285m-, 325m-, 340m- and 385m-levels of the underground mine outlined in dark grey with sample locations indicated, north arrow indicates mine grid north.
3.2 Sampling of tailings from the processing plant
The ore is hauled and stored in ore stockpiles (Fig. 7) where it is crushed and screened, the process is repeated until the material is under 8 mm in size (Pressacco et al., 2018). The crushed ore is fed into the primary grinding circuit consisting of a ball mill and a rod mill (Fig. 7)(Pressacco et al., 2018). The material from the primary grinding circuit is screened and oversized material is re-circulated to the primary grinding circuit, while undersized material with particle sizes of 80% (P80) passing 560 µm is pumped to hydro-cyclones for further size separation (Pressacco et al., 2018).
The underflow from the cyclone (P80 800 µm) is fed through rougher spiral concentrators, the tailings are pumped to the secondary ball mill while the concentrate is fed to the cleaner spirals (Fig.7) (Pressacco et al., 2018). The concentrate from the cleaner spirals is pumped to the cleaning table, the rejected material is mixed with the tailings from the rougher spirals and pumped to the secondary mill for re-grinding (Pressacco et al., 2018).
- 10 -
The major concentrate products of the processing plant are formed from the final cleaner shaking table, where concentrate from the Knelson concentrator and concentrate from the shaker table are processed for gravity and middlings concentrate (Fig. 7) (Pressacco et al., 2018).
Figure 7. Simplified schematic overview of the processing circuit (Pressacco et al., 2018).
Overflow from the cyclone after the classifying screen (P80 230 µm) is again classified in the flotation cyclones, underflow (P80 410 µm) from the flotation cyclone is fed to a Knelson concentrator, and the concentrate is pumped to a SkimAir flash flotation cell (Pressacco et al., 2018). The concentrate from the SkimAir cell is pumped to final de-watering, whereas the tailings are mixed with the overflow (P80 125 µm) of the cyclone and pumped to the rougher flotation cells (Pressacco et al., 2018).
Tailings from the rougher flotation circuit are pumped to scavenger flotation circuit, whereas the concentrate from the scavenger flotation is sent to the rougher flotation (Pressacco et al., 2018). The rougher flotation concentrate is pumped to the first cleaner flotation circuit, tailings are sent back to the feed of the rougher flotation circuit, while tailings from the second cleaner flotation are pumped back to the first cleaner flotation unit (Pressacco et al., 2018).The second cleaner flotation
concentrate is combined with concentrate from the SkimAir flotation, de-watered and filtered to the final flotation product (Fig. 7) (Pressacco et al., 2018).
- 11 -
The final tailings (P80 230 µm) from the processing plant (Fig. 7) consists of tailings from the
scavenger flotation circuit (Pressacco et al., 2018). The gravity concentrate carries approximately 50% Au, the middlings concentrate approximately 1500g/t Au and the flotation concentrate about 100g/t (Pressacco et al., 2018).
Tailings are sampled daily by the company staff and analyzed for gold grade twice per day, samples are taken from the tailings stream (Fig. 7) automatically every fifteen minutes for 6 hours. The slurry is fed through a vacuum filter to remove most of the water and dried in an oven at 105°C. The dried material is sub-sampled using a riffle splitter to get approximately 500g of dried tailings sand. Roughly 300 ─ 400g is required for analysis, and the remaining material is archived and stored for at least 3 months. Ore from GP-130b3005, CHU-385-539W, CHU-340-1320E and ore from two additional drifts as well as one blast in the open pit was blended for a larger ore campaign. The ore campaign was processed between 1st-9th of February 2018, the first tailings sample taken in this study is from the 2 of February to allow for steady-state conditions to occur in the processing plant.
For this study, approximately 40─50g of tailings sand per sample was collected from the archived samples dated between 2 and 7 of February 2018. Samples from February 4 and 5 were combined into a single composite sample due to limited material availability. In total ten samples of tailings sand were collected in five days.
3.3 Sample preparation
Twenty rock samples were cut down to a size of approximately 50 x 30 x 10 mm with a diamond rock saw at Luleå University of Technology. The cut rock samples, as well as the tailings samples (sand), were sent to Vancouver Petrographics Ltd, Vancouver, Canada for preparation into polished thin sections and polished thick sections respectively.
A portion of the remaining material from each rock sample was sent from Luleå University of Technology to ALS Scandinavia in Piteå for whole-rock geochemistry, starting with fine crushing of the samples to 70% passing 2 mm and grinding to 85% passing 75 µm. The samples were then forwarded for processing at ALS Loughrea, Ireland where aqua regia digestion with Inductively Coupled Plasma Mass Spectrometry (ICP-MS) was used to analyze trace elements. Major oxides were analyzed using X-Ray Fluorescence (XRF).
3.4 Petrography
Twenty thin sections with dimensions of 46 x 27 mm with a thickness of 30 µm, as well as ten thick sections (tailings samples), of which five had the dimensions of 46 x 27 mm with a thickness of 200 µm, and five thick sections with a 25 mm diameter with a thickness of 200 µm were analyzed using a Nikon Eclipse LV 100POL polarizing petrographic microscope with an attached digital camera at Luleå University of Technology.
- 12 -
4. Results
4.1 Petrography
The total number of gold grains identified in this study, from both open pit and underground mine is 3329. The overall gold grain size distribution shows an interval from very fine-grained (< 2 µm) to coarse-grained (856 µm) with a median size of 7 µm. The percentage of gold grains larger than 100 µm accounts for 2.4% of the total gold grain population. Histogram plots of gold grain size distributions by mineral association have a resolution of 2 µm and are truncated to display 0─100 µm for improved legibility. The remaining size distribution for gold grains larger than 100 µm is inserted as a subplot in the upper right corner as required from the grain size data. The gold grain texture was classified into four groups, grain boundary (Fig. 8A), intergrown (Fig. 8B), inclusion (Fig. 8C) and fracture infilling (Fig. 8D).
Figure 8. Schematic illustrations of gold textures. A. Grain boundary. B. Intergrown. C. Inclusion. D. fracture infilling.
4.1.1 West pit
- 13 -
Figure 9. Textural relation of gold grains from the West pit production area, data from 119 grains of gold.
The mineral association of gold from the West pit (Fig. 10) is dominated by quartz and silicates followed by tsumoite (labelled Te in the legend), minor amounts of calcite and sulfide were present. Actinolite is the most common mineral of the silicates and pyrite is most common for sulfides.
- 14 -
The grain size distribution of gold associated with silicates (Fig. 11A) has a median size of 7 µm. Gold associated with sulfide (Fig. 11B) range between 18 µm and 134 µm in size (inserted subplot). Gold associated with quartz (Fig. 11C) has a median size of 9 µm. Gold associated with calcite (Fig. 11D) has sizes of 16 µm, 33 µm and 367 µm (inserted subplot). Gold associated with tsumoite (labelled Te) (Fig. 11E) has a median of 9 µm.
Figure 11. Grain size distribution by mineral association in the West pit production area, bin width is 2 µm. A) Gold associated with silicates. B) Gold associated with sulfide, a subplot in the upper right corner displays 100─200 µm. C) Gold associated with quartz, a subplot in the upper right corner displays 100─400 µm. D) Gold associated with calcite, a subplot in
- 15 -
4.1.2 Quartz Mountain
Four polished thin sections were inspected from the Quartz Mountain production area (Fig. 5), GP-130b3005-2, GP-130b3002-3a, GP-130b3002-3b and GP-130b3005-4. Gold grains are present in all thin sections except GP-130b3005-2. A total of 1568 gold grains were identified, with a minimum gold grain size of 2 µm, a maximum of 150 µm and a median size of 4 µm. Grain boundary is the dominant textural relation of gold grains (Fig. 12), followed by intergrown with minor inclusion and fracture infilling.
Figure 12. Textural relation of gold grains from the Quartz Mountain production area, data from 1568 grains of gold.
- 16 -
Figure 13. Mineral association of gold from the Quartz Mountain production area.
- 17 -
Figure 14. Grain size distribution by mineral association in the Quartz Mountain production area, bin width is 2 µm. A) Gold associated with silicates, a subplot in the upper right corner displays 30─160 µm. B) Gold associated with sulfide. C) Gold
- 18 -
4.1.3 South East Wall
Two polished thin sections were inspected from the South-East Wall production area (Fig. 5), GP-105g4016-1 and GP-105g4016-3. No gold grains were identified. The main gangue minerals are quartz with lesser amounts of actinolite as well as calcite veinlets. Minor amounts of pyrite and pyrrhotite are also present.
4.1.4 East pit
Two polished thin sections were inspected from the East pit production area (Fig.5), GP-140b5030-1 and GP-140b5030-2 and gold is present in both thin sections. A total of 66 gold grains were identified from the East pit production area, with a minimum gold grain size of 2 µm, a maximum of 100 µm and a median size of 8 µm. Grain boundary is the dominant textural relation of gold grains (Fig.15), followed by intergrown, fracture infilling with minor inclusion.
Figure 15. Textural relation of gold grains from the East pit. Data from 66 grains of gold.
- 19 -
Figure 16. Mineral association of gold in the East pit.
- 20 -
Figure 17. Grain size distribution by mineral association, East pit production area, bin width is 2 µm. A) Gold associated with silicates. B) Gold associated with sulfide. C) Gold associated with quartz. D) Gold associated with calcite. E) Gold associated
- 21 -
4.1.5 285-1365-eb
Two polished thin sections were inspected from the 285m level 1365eb drift, CHU-285-1365eb-1 and CHU-285-1365eb-2 (Fig. 6). Gold is present in both thin sections. A total of 115 gold grains were identified, with a minimum gold grain size of 2 µm, a maximum size of 362 µm and a median size of 12 µm. Grain boundary is the dominant textural relation of gold grains (Fig. 18), followed by intergrown, fracture infilling with minor inclusion.
Figure 18. Textural relation of gold grains from 285-1365eb drift. Data from 115 grains of gold.
- 22 -
Figure 19. Mineral association of gold from 285-1365eb drift.
- 23 -
Figure 20. Grain size distribution by mineral association in 285-1365eb drift, bin width is 2 µm. A) Gold associated with silicates, subplot displays 100─200 µm. B) Gold associated with sulfide. C) Gold associated with quartz, subplot displays
- 24 -
4.1.6 325-1420W
Two polished thin sections were inspected from the 325m level 1420W drift, CHU-325-1420W-1 and CHU-325-1420W-2 (Fig. 6). Gold was present in both thin sections. A total of 149 gold grains were identified, the minimum gold grain size is 2 µm, the maximum is 195 µm with a median size of 6 µm. Grain boundary is the dominant textural relation of gold grains (Fig. 21), followed by intergrown.
Figure 21. Textural relation of gold grains from 325-1420W drift, data from 149 grains of gold.
- 25 -
Figure 22. Mineral association of gold from 325-1420W drift.
- 26 -
Figure 23. Grain size distribution by mineral association in 325-1420W drift, bin width is 2 µm. A) Gold associated with silicates. B) Gold associated with quartz, subplot displays 100─200 µm. C) Gold associated with Te, subplot displays 100─200
- 27 -
4.1.7 340-1320E
Two polished thin sections were inspected from the 340 level 1320E drift, 340-1320E-1 and CHU-340-1320E-2 (Fig. 6). Gold was present in both thin sections. A total of 102 gold grains were identified, the minimum gold grain size is 2 µm, the maximum is 393 µm with a median size of 14 µm. Grain boundary is the dominant textural relation of gold grains (Fig. 24), followed by intergrown with minor inclusion.
Figure 24. Textural relation of gold grains from 340-1320E drift, data from 102 grains of gold.
- 28 -
Figure 25. Mineral association of gold from 340-1320E drift.
- 29 -
Figure 26. Grain size distribution by mineral association in 340-1320E drift, bin width is 2 µm. A) Gold associated with silicates. B) Gold associated with sulfide. C) Gold associated with quartz, subplot displays 100-400 µm. D) Gold associated
with calcite. E) Gold associated with Te, subplot displays 100-300 µm.
- 30 -
Figure 27. Centre-left part of image, larger gold grains intergrown with bismuth-tellurides in a matrix of quartz. Smaller grain of gold with a rim of chalcopyrite in centre-right of the image with gold at grain boundary with quartz upper right
corner.
4.1.8 385-539W
- 31 -
Figure 28. Textural relation of gold grains from 385-539W drift, data from 1210 grains of gold.
- 32 -
Figure 29. Mineral association of gold from 385-539W drift.
- 33 -
Figure 30. Grain size distribution by mineral association in 385-539W drift, bin width is 2 µm. A) Gold associated with silicates, a subplot in the upper right corner displays 100─875 µm. B) Gold associated with sulfide, a subplot in the upper right corner displays 100─200 µm. C) Gold associated with quartz, a subplot in the upper right corner displays 100─600 µm.
D) Gold associated with Te, a subplot in the upper right corner displays 100─875 µm.
From part of thin section 385-539W-1, in (Fig. 31) fifty-six gold grains identified ranging in size from 3 µm to 154 µm, gold mainly at grain boundary to assemblages of tourmaline with actinolite and minor biotite in a matrix of quartz.
- 34 -
4.2 Tailings
Ten polished thick sections of tailings were analysed, and gold grains were identified in two thick sections. The tailings mainly consist of silicate gangue minerals with quartz, biotite and plagioclase/feldspar. Minor amounts of sulfides, chalcopyrite, pyrite and pyrrhotite are present. Liberated gold grains were identified in two thick sections (Fig. 32A and 32B), one grain of gold in thick section TB1-02feb-1 and two grains of gold in TB1-07feb-1 with sizes of 23 µm, 8 µm and 4 µm respectively.
Figure 32: A. Liberated grain of gold in TB1-02feb-1 thick section. B. Grains of liberated gold in TB1-07feb-1.
In thick section TB1-02feb-1 one grain of gold (Fig. 33) were also identified intergrown with a bismuth-telluride as inclusion in larger silicate.
- 35 -
4.3 Scanning Electron Microscopy
Five thin sections and four thick sections were analysed with SEM; CHP-484, 130b3005-3a, 285-1365eb-1, 385-539W-1, TB1-06feb-2. In thin section CHP-484, gold was observed to be intergrown with grey-coloured mineral (Fig. 34A) and possibly one additional mineral as inclusion in a larger bismuth-telluride mineral grain. Scanning electron microscopy (SEM-EDS) (Fig. 34B) was utilized to confirm gold intergrown with galena, and a mineral phase of Bi-Te-S as inclusion in tsumoite (BiTe).
Figure 34. A. Optical microscopic image from thin section CHP-484 of bismuth-telluride mineral with gold as inclusion, reflected light. B. Scanning electron image displaying gold intergrown with galena with a mineral phase of Bi-Te-S as
inclusion in tsumoite. Abbreviations: Au-gold, gn- galena, tsu- tsumoite.
- 36 -
In thin section CHP-484, an opaque mineral with high reflectance and bright creamy colour was observed (Fig. 35A), intergrown with a darker mineral. Scanning electron microscopy (SEM-EDS) (Fig. 35B) was used to confirm native bismuth intergrown with galena in a quartz matrix.
Figure 35. A. Optical microscopic image from CHP-484 in reflected light. B. Scanning electron image of native Bismuth (light grey) intergrown with galena (darker grey). Abbreviations: Bi- bismuth, gn- galena.
- 37 -
Optical microscopy (Fig. 36A) of thin section 130b3005-3a displays gold intergrown with bismuth-telluride mineral in quartz matrix, slight variances in colour in the bismuth-telluride mineral were observed. Four points (Fig. 36B) were selected for further study with scanning electron microscopy utilizing EDS for determination of the composition of elements.
Figure 36. A. Optical microscopy image reflected light of gold as inclusion in Bismuth-telluride mineral in quartz matrix. B. SEM image (secondary electrons) of the magnified upper section from A, with selected points for spot analysis.
- 38 -
Figure 37. Spectrum intensity (cps / keV) with respect to energy (keV) from scanning electron microscopy. A. Spectra of gold-telluride mineral. B. Spectra of gold. C. Spectra of silver-gold-gold-telluride. D. Spectra of Bismuth-gold-telluride mineral.
- 39 -
Scanning electron microscope element mapping of Au (Fig. 38A), Ag (Fig. 38B), Te (Fig. 38C) and Bi (Fig. 38D) further supports the intensity spectra of four minerals displayed above in (Fig. 37A-D) and indicates Au-Te mineral in spectrum 95, gold in spectrum 96, Ag-Au-Te mineral in spectrum 97 and bismuth-telluride mineral in spectrum 98.
Figure 38. Scanning electron microscope with points marked for intensity spectrum and mapping displaying the distribution of: A. Detected gold. B. Detected silver. C. Detected tellurium. D. Detected bismuth.
4.4 Whole rock geochemistry
A correlation matrix was constructed (Appendix 10.1) from analyzed trace elements received from ALS, with a perfect correlation having the value of 1, no correlation with a value of 0 and a perfect negative correlation with -1. Statistically significant correlations in (Appendix 10.1) are highlighted in green while marginally significant correlations were highlighted in yellow.
- 40 -
Table 1. Statistically significant correlations of elements
Correlation r Elements
0.95+ Bi-Ag, Bi-Sb, Bi-Se
0.85-0.95 Ag-Sb, Ag-Se
0.75-0.85 Bi-Hg, Ag-Hg, Ag-Pb 0.65-0.75 Au-Te, Ag-Te, Bi-Pb, 0.55-0.65 Te-Bi, Te-Pb, Te-Sb, Te-Se, S-Cu
0.45-0.55 Te-Hg, S-Pb, S-Fe
- 41 -
5. Discussion
Differences in the median grain size of gold from the production areas and drifts sampled varies, gold from the Quartz Mountain production area has the smallest median size of 4 µm, whereas gold from 340m- and 385m- level has the largest median size of 14 µm. Gold from the western and eastern production areas in the open pit shows a median size of 8 µm, the median size of gold from 285m- and 325m-level has a median size of 12 µm and 6 µm respectively.
The maximum size of gold grains was identified from the 385-539W drift and measured 856 µm, and the size distribution is considerably coarser-grained (Fig. 30) compared to the size distribution of gold from other sampled localities (Fig. 11, 14, 17, 20, 23, 26) and accounts for 74% of the total grains identified larger than 100 µm.
5.1 Mineral association of gold
Mineral association of gold in samples from the open pit (Fig. 39) shows quartz as the most common association in the western and eastern pit, while quartz accounts to 38% from the Quartz Mountain production area. Silicates (dominantly actinolite and biotite) are the second most common association in the open pit samples. Bismuth-tellurides (labelled Te) is the third most common. Sulfide association accounts for a smaller percentage of 1%, 3% and 5% for West pit, Quartz Mountain and East pit respectively. Calcite was identified in minor amounts in the West and East pit.
- 42 -
Mineral association from underground mine samples displays a wider range of associations (Fig. 40) compared to samples from the open pit. Quartz is the most common association in samples from the 325-1420W and the 340-1320E drifts, whereas silicates are the most common association from the 285-1365eb and the 385-539W drift. Quartz is the second most common association from 285-1365eb and 385-539W drifts, whereas silicates are the second most common from 325-1420W and bismuth-tellurides (labelled Te) is the second most common from 340-1320E drifts. Bismuth-bismuth-tellurides (labelled Te) constitutes the third most common association, replaced by calcite in the 340-1320E drift. Sulfides are the fourth most common association in the 285-1365eb and 385-539W drifts, replaced by silicates in the 340-1320E drift. The least common association for the 285-1365eb and 340-1320E drifts is calcite and sulfides respectively.
Figure 40. Mineral association of gold from samples taken from the 285m-, 325m-, 340m-, 385m drifts. Pie chart of 285-1365eb drifts based on 115 grains of gold, 325-1420W chart based on 149 grains of gold, 340-1320E chart based on 102
- 43 -
5.2 Textures of gold
The texture of gold from the open pit (Fig. 41) is dominated by grain boundary ranging from 70-73%. The intergrown texture is the second most common with 15%, 29% and 20% for West, Quartz Mountain and East pit respectively. The texture inclusion accounts for 12% from samples taking in the west pit, whereas it accounts for 1% in samples from Qtz mountain and east pit. Fracture filling, as texture is less common with 6% in the east pit and 1% in the West pit.
Figure 41. Texture of gold in the open pit with samples locations marked from West pit, Quartz Mountain and East pit. Pie chart of west pit based on data from 119 grains of gold, quartz mountain based on data from 1568 grains of gold and east
pit based on data from 66 grains of gold.
- 44 -
Figure 42. Texture of gold in the underground mine from 285m-, 325m-, 340m-, 385m levels. Pie chart of 285-1365eb drifts based on 115 grains of gold, 325-1420W chart based on 149 grains of gold, 340-1320E chart based on 102 grains of gold
and 385-539W chart based on 1210 grains of gold.
5.3 Geochemistry
Gold is significantly and positively correlated with tellurium (Appendix 10.1.1), and marginally correlated to silver and mercury. Gold was observed with a close association to bismuth-tellurides in the samples, from the production areas in the open pit and sampled drifts in the underground mine. In addition, gold was identified in two telluride minerals from SEM analysis, Au-Te and Ag-Au-Te mineral, as possibly calaverite and petzite, respectively.
- 45 -
this texture of gold may account for some of the loss since it might not be efficiently recovered by gravity and flotation circuits and therefore considered refractory .
With the detection of gold-telluride and silver-gold-telluride minerals from SEM analysis, this minor gold fraction of the ore can be considered to be refractory (Zhang et al, 2010), where the recovery of gold from refractory ores varies extensively from below 50% to 90% (La Brooy et al., 1994; Vaughan, 2004).
Liberated gold was identified in two tailings samples, TB1-02feb-1 and TB1-07feb-1 (Fig. 32A and B), where the flotation circuit failed to float the free gold. One possible explanation is that flaky gold grains have a lower rate of recovery by flotation compared to grains having crystal axis of more equal lengths due to flaky grains exhibiting less susceptible surfaces for bubbles to collect on (Knipe et al., 2004). This grain shape results in slower flotation kinetics and require higher collector loadings to float the flaky grains of gold (Chryssoulis et al., 2003). In addition, rejected flaky gold in flotation tailings is pumped for regrinding, with a risk of further flattening the gold resulting in lower chances to float the gold (Knipe et al., 2004).
Gold grains from the sampled localities are most commonly located at grain boundaries to quartz and silicate, although it is difficult to make a more general determination of the gold grain shape of the active mine from the samples, it may be possible that a small fraction of elongated gold grains could be flattened at the milling stage to show the behaviour of flaky gold suggested by (Knipe et al., 2004). In the tailings samples, gold was also identified intergrown with bismuth-telluride as inclusion in silicate (Fig. 33), where the flotation properties of the larger silicate grain likely dominated in the flotation process. Finer grinding (currently P80 230 µm) could potentially liberate the gold, with the potential drawback of lower throughput and higher treatment costs.
Gold substituted into the structure of common sulfide minerals and gold existing as inclusion smaller than 0.1 µm is termed invisible gold and is considered refractory (Cook and Chryssoulis, 1990). Sulfide minerals arsenopyrite and arsenic-rich pyrite generally contains higher concentrations of gold, while pyrrhotite, chalcopyrite and galena often contains smaller amounts of invisible gold (Cook and Chryssoulis, 1990). Invisible gold situated in the structure of pyrite could possibly be present at the Björkdal deposit and may be analysed utilizing LA ICP-MS, however this was not a focus of this thesis.
- 46 -
6. Conclusion
A total of 3329 gold grains was identified in this study and the overall gold grain size distribution shows a range between very fine-grained (2 µm) and coarse-grained (856 µm). The percentage of gold grains larger than 100 µm accounts for 2.4%. The median grain size of gold from the sampled production areas and drifts varies, gold from the Quartz Mountain production area has the smallest median size of 4 µm, whereas gold from the sampled drifts at 340m- and 385m- level has the largest median size of 14 µm.
Grain boundary is the dominant textural mode of gold from all sampled locations and varies from 62% to 92%, followed by intergrown which ranges between 8%-29%. Gold as inclusion and fracture filling occur less frequently. Quartz is the most common mineral association of gold in thin sections from the western and eastern production area in the open pit, as well as the 325-1420W and 340-1320E drifts, with silicates (actinolite, biotite and tourmaline) being the second most common association except for the 340-1320E where bismuth tellurides is more common. Silicates are the most common association of gold in thin sections from the Quartz Mountain production area in the open pit, and in the 285-1365eb and 385-539W drifts, with quartz as the second most common association. Bismuth-tellurides is the third most common association of gold, sulfides and calcite is also present to a lesser degree in varying amounts. Gold identified in samples from 385-539W drift was most commonly associated with tourmaline and was notably more coarse-grained, accounting for 74% of all gold grains larger than 100 µm.
Apart from the native gold, which is the dominant mode of gold, two additional gold-bearing minerals were identified during SEM analysis, one Au-Te mineral and one Ag-Au-Te mineral. Bismuth-tellurides mainly consist of tsumoite (BiTe), and during SEM analysis an additional phase with Bi-Te-S was identified as well as native bismuth.
Gold is significantly and positively correlated to tellurium, and gold show a close association with bismuth-tellurides in the samples. In addition, gold is marginally associated with silver and mercury. Three gold grains were identified in the tailings samples (Fig. 32 and Fig. 33), of which two occurred as liberated gold where the flotation circuit likely failed to float the free gold. One grain of gold was identified intergrown with bismuth-telluride as inclusion in silicate (Fig. 33), where the flotation properties of the larger silicate grain likely dominated in the flotation process.
- 47 -
7. Recommendations
Due to a lack of detected gold grains in the investigated tailings samples (only 3 identified grains) no real conclusion can be made about which gold texture that is typically lost in the process. Further studies utilizing quantitative analysis of SEM/QEMSCAN/MLA to determine the occurrence and distribution of gold-telluride minerals. As well as more sample locations and analysed drill cores for better representability of data.
8. Acknowledgements
I wish to thank Mandalay Resources, in particular, Lena Printzell for making this thesis possible, and Peder Hoppstadius for his vast knowledge of the processing plant. I would also like to thank mine geologists Sören Vikström, Niklas Nygård and Thomas Eriksson as well as mine engineer Nils Lindberg for their help with sampling and sampling locations.
I would also like to thank Ph. D students Reginald Fettweis and Joel Andersson for showing how to prepare and send the samples to Vancouver Petrographics Ltd, and help with ordering chemical assay from ALS Scandinavia.
- 48 -
9. References
Allen, R. L., Weihed, P., Svenson, S. A. (1997). Setting of Zn-Cu-Au-Ag massive sulfide deposits in the evolution and facies architecture of a 1.9 Ga marine volcanic arc, Skellefte district, Sweden. Economic Geology, 91(6), 1022–1053. https://doi.org/10.2113/gsecongeo.91.6.1022 Bark, G., Broman, C., Weihed, P. (2007). Fluid chemistry of the Palaeoproterozoic Fäboliden
hypozonal orogenic gold deposit, northern Sweden: Evidence from fluid inclusions. GFF, 129(3), 197–210. https://doi.org/10.1080/11035890701293197
Bark, G., Weihed, P. (2012). Geodynamic settings for Paleoproterozoic gold mineralization in the Svecofennian domain: A tectonic model for the Fäboliden orogenic gold deposit, northern Sweden. Ore Geology Reviews, 48, 403–412. https://doi.org/10.1016/j.oregeorev.2012.05.007 Billström, K., Broman, C., Jonsson, E., Recio, C., Boyce, A. J., & Torssander, P. (2009).
Geochronological, stable isotopes and fluid inclusion constraints for a premetamorphic development of the intrusive-hosted Björkdal Au deposit, northern Sweden. International Journal of Earth Sciences, 98(5), 1027–1052. https://doi.org/10.1007/s00531-008-0301-8 Billström, K., Weihed, P. (1996). Age and Provenance of Host Rocks and Ores in the Paleoproterozoic
Skellefte Disrict. Economic Geology, 91, 1054–1072.
Broman, C., Billström, K., Gustavsson, K., & Fallick, A. E. (1994). Fluid Inclusions, Stable Isotpes and Gold Deposition at Björkdal. Mineral Deposita, 29, 139–149.
Chryssoulis, S. L., Venter, D., Dimov, S. (2003). On the floatability of gold grains. Proceedings of the 35th Annual Meeting of the Canadian Mineral Processors, 455–472.
Claesson, S., Lundqvist, T. (1995). Origins and ages of Proterozoic granitoids in the Bothnian Basin, central Sweden; isotopic and geochemical constraints. Lithos, 36(2), 115–140.
https://doi.org/10.1016/0024-4937(95)00010-D
Coetzee, L. L., Theron, S. J., Martin, G. J., Merwe, J. D. Van Der, & Stanek, T. A. (2011). Modern gold deportments and its application to industry. Minerals Engineering, 24(6), 565–575.
https://doi.org/10.1016/j.mineng.2010.09.001
Cook, N. J., Chryssoulis, S. L. (1990). Concentrations of invisible gold in the common sulfides. Journal of the Mineralogical Association of Canada, 28, 1–16.
Eliasson, T., Sträng, T. (1997). Kartbladen 23H Stensele. In C.-H. Wahlgren (ed.): Regional
berggrundsgeologisk undersökning, Sammanfattning av pågående undersökningar. In Sveriges Geologiska Undersökning, Rapporter och meddelanden (Vol. 89).
Erneholm, M. (2017). Characterization of the Hydrothermal Alteration around the Björkdal Au Deposit , Skellefte District , Sweden. Luleå: Luleå University of Technology Department of Civil,
Environmental and Natural Resources Engineering.
- 49 -
Gonzàlez Roldàn, M. J. (2010). Mineralogy, petrology and geochemistry of syn-volcanic intrusions in the Skellefte mining district, Northern Sweden. Retrieved from
http://www.researcherid.com/rid/I-2366-2012
Groves, D., Goldfarb, R. J., Gebre-Mariam, M., Hagemann, S. ., & Robert, F. (1998). Orogenic gold deposits: A proposed classification in the context of their crustal distribution and relationship to other gold deposit types. Ore Geology Reviews, 13, 7–27.
Groves, D. I., Goldfarb, R. J., Robert, F., & Hart, C. J. R. (2003). Gold Deposits in Metamorphic Belts : Overview of Current Understanding , Outstanding Problems , Future Research , and Exploration Significance. Economic Geology, 98(1), 1–29.
Harris, D. C. (1990). The Mineralogy of gold and its relevance to gold recoveries. Mineralium Deposita, 25(1), 3–7. https://doi.org/10.1007/BF00205243
Hart, C. J. R., Goldfarb, R. J. (2005). Distinguishing intrusion-related from orogenic gold systems. New Zealand Minerals Conference Proceedings, 2703(January 2005), 125–133. Retrieved from http://www.nzpam.govt.nz/cms/pdf-library/minerals/conferences-1/125_papers_48.pdf Kathol, B., Weihed, P. (2005). Description of regional geological and geophysical maps of the Skellefte
District and surrounding areas. In Sveriges Geologiska Undersökning, Ba 57.
Knipe, S. W., Chryssoulis, S. L., Clements, B. (2004). Flaky gold: Problems with recovery and mineralogical quantification. Jom, 56(7), 58–62. https://doi.org/10.1007/s11837-004-0095-5 La Brooy, S. R., Walker, G. S., Linge, H. G. (1994). Review of gold extraction from ores. Minerals
Engineering, 7(10), 1213–1241. https://doi.org/10.1016/0892-6875(94)90114-7
Lang, J. R., Baker, T. (2001). Intrusion-related gold systems: The present level of understanding. Mineralium Deposita, 36(6), 477–489. https://doi.org/10.1007/s001260100184
Prasad, M. ., Mensah-Biney, R., Pizarro, R. . (1991). Modern trends in gold processing-overview. Minerals Engineering, 4(12), 1257–1277.
Pressacco, R., Lunnon, J., Smith, D. J. F., Robson, D., Weir, I., Altman, K. A., Evans, L. (2018). MANDALAY RESOURCES CORPORATION TECHNICAL REPORT ON THE BJÖRKDAL GOLD MINE , SWEDEN NI 43-101 Report Qualified Persons : (Vol. 1).
Roberts, S., Palmer, M. R., & Waller, L. (2006). Sm-Nd and REE characteristics of tourmaline and scheelite from the Björkdal gold deposit, Northern Sweden: Evidence of an intrusion-related gold deposit? Economic Geology, 101(7), 1415–1425.
https://doi.org/10.2113/gsecongeo.101.7.1415
Sillitoe, R. H., Thompson, J. F. H. (1998). Intrusion-Related Vein Gold Deposits: Types, Tectono-Magmatic Settings and Difficulties of Distinction from Orogenic Gold Deposits. Resource Geology, 48(4), 237–250. https://doi.org/10.1111/j.1751-3928.1998.tb00021.x
Skiöld, T., Öhlander, B., Markkula, H., Widenfalk, L., Claesson, L. Å. (1993). Chronology of Proterozoic orogenic processes at the Archaean continental margin in northern Sweden. Precambrian Research, 64(1–4), 225–238. https://doi.org/10.1016/0301-9268(93)90078-G
Skyttä, P., Bauer, T. E., Tavakoli, S., Hermansson, T., Andersson, J., Weihed, P. (2012). Pre-1.87Ga development of crustal domains overprinted by 1.87Ga transpression in the Palaeoproterozoic Skellefte district, Sweden. Precambrian Research, 206–207, 109–136.
- 50 -
Vaughan, J. P. (2004). The process mineralogy of gold: The classification of ore types. Jom, 56(7), 46– 48. https://doi.org/10.1007/s11837-004-0092-8
Wasström, A. (1993). The Knaften graitoids of Västerbotten County, northern Sweden. Sveriges Geologiska Undersökning, C823, 60–64.
Weihed, P. (2001). A review of Palaeoproterozoic intrusive hosted Cu-Au-Fe-oxide deposits in northern Sweden. Economic Geology Research, 1, 4–32.
Weihed, P., Arndt, N., Billström, K., Duchesne, J. C., Eilu, P., Martinsson, O., … Lahtinen, R. (2005). 8: Precambrian geodynamics and ore formation: The Fennoscandian Shield. Ore Geology Reviews, 27(1–4), 273–322. https://doi.org/10.1016/j.oregeorev.2005.07.008
Weihed, P., Bergman, J., Bergström, U. (1992). Metallogeny and tectonic evolution of the early Proterozoic Skellefte District, northern Sweden. Precambrian Research, 58, 143–167. Weihed, P., Billström, K., Persson, P., Weihed, J. B. (2002). Relationship between 1.90-1.85 Ga
accretionary process and 1.82-1.80 Ga oblique subduction at the Karelian craton margin , Fennoscandian Shield. GFF, 124, 163–180.
Weihed, P., Weihed, J. B., Sorjonen-Ward, P. (2003). Structural evolution of the Björkdal Gold Deposit, Skellefte district, northern Sweden: Implications for Early Proterozoic Mesothermal Gold in the Late Stage of the Svecokarelian Orogen. Economic Geology, 98(7), 1291–1309. https://doi.org/10.2113/gsecongeo.98.7.1291
Wilson, M. R., Sehlstedt, S., Claesson, L. åke, Smellie, J. A. T., Aftalion, M., Joseph Hamilton, P., & Fallick, A. E. (1987). Jörn: An early proterozoic intrusive complex in a volcanic-arc environment, north sweden. Precambrian Research, 36(3–4), 201–225. https://doi.org/10.1016/0301-9268(87)90021-0
Zhang, J., Zhang, Y., Richmond, W., Wang, H. P. (2010). Processing technologies for gold-telluride ores. International Journal of Minerals, Metallurgy and Materials, 17(1), 1–10.
https://doi.org/10.1007/s12613-010-0101-6
Zhou, J. Y., Cabri, L. J. (2004). Gold process mineralogy: Objectives, techniques, and applications. Jom, 56(7), 49–52. https://doi.org/10.1007/s11837-004-0093-7
- 52 -
10.1 Correlation matrix
Figure 2. Matrix of correlation coefficients with statistically significant correlation coefficients marked in green, marginally significant levels marked in yellow.
- 53 -
10.2.1 Gold correlation coefficients and significance for selected trace elements
Correlation for Au Correlation r significance p Ag 0.4292 0.0590 As 0.1018 0.6692 Bi 0.3604 0.1185 Cu 0.0919 0.7001 Hg 0.4215 0.0642 Pb 0.3273 0.1589 Sb 0.3646 0.1140 Se 0.3341 0.1499 Te 0.7161 3.84E-04 Zn -0.2415 0.3050
- 54 -
10.2.2 Tellurium correlation coefficients and significance for selected trace elements
Correlation for Te Correlation r significance p Ag 0.7051 5.16E-04 As -0.1594 0.5019 Au 0.7161 3.84E-04 Bi 0.6123 4.11E-03 Cu 0.1041 0.6622 Hg 0.5227 0.0180 Pb 0.5656 9.35E-03 Sb 0.5979 5.36E-03 Se 0.6172 3.74E-03 Zn -0.1381 0.5615- 55 -
10.2.3 Bismuth correlation coefficients and significance for selected trace elements
Correlation for Bi Correlation r significance p Ag 0.9627 1.16E-11 As -0.1773 0.4545 Au 0.3604 0.1185 Cu -0.0106 0.9646 Hg 0.8266 7.05E-06 Pb 0.6772 1.04E-03 Sb 0.9867 1.21E-15 Se 0.9783 9.38E-14 Te 0.6123 4.11E-03 Zn -0.0467 0.8449
- 56 -
10.2.4 Silver correlation coefficients and significance for selected trace elements
Correlation for Ag Correlation r significance p As -0.1920 0.4175 Au 0.4292 0.0590 Bi 0.9627 1.16E-11 Cu -0.0184 0.9386 Hg 0.7777 5.45E-05 Pb 0.7652 8.48E-05 Sb 0.9210 8.56E-09 Se 0.9289 3.39E-09 Te 0.7051 5.16E-04 Zn -0.0209 0.9304- 57 -
10.2.5 Silicon dioxide correlation coefficients and significance for selected trace
elements
Correlation for silica
Correlation r significance p Ag -0.0004 0.9988 As -0.1749 0.4607 Au -0.1532 0.5191 Bi 0.0323 0.8924 Cu -0.1868 0.4304 Hg 0.2284 0.3327 Pb -0.3867 0.0921 Sb 0.0550 0.8177 Se -0.0191 0.9362 Te 0.0282 0.9061 Zn -0.5933 5.82E-03
- 58 -
10.2.6 Sulfur correlation coefficients and significance for selected trace elements
Correlation for Sulphur
Correlation r significance p Ag 0.0585 0.8065 As 0.1327 0.5771 Au 0.1487 0.5316 Bi 0.0656 0.7836 Cu 0.6389 2.43E-03 Hg -0.1437 0.5456 Pb 0.4778 0.0331 Sb 0.0500 0.8341 Se 0.2201 0.3512 Te 0.2016 0.3941 Zn 0.2516 0.2846 Fe 0.5063 0.0227
- 59 -
10.3 Geochemistry, major oxides and trace elements
Al2O3 BaO CaO Cr2O3 Fe2O3 K2O MgO MnO
- 60 -
