Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Geology and Isotope Systematics of Gold Deposits in the Abansuoso Area of the Sefwi Belt, Southwestern Ghana

Geology and Isotope Systematics of Gold Deposits in the Abansuoso Area of the Sefwi Belt,... GEOLOGY, ECOLOGY, AND LANDSCAPES INWASCON https://doi.org/10.1080/24749508.2022.2142100 RESEARCH ARTICLE Geology and Isotope Systematics of Gold Deposits in the Abansuoso Area of the Sefwi Belt, Southwestern Ghana a,c b,c d e f Raymond Webrah Kazapoe , Olugbenga Okunlola , Emmanuel Arhin , Olusegun Olisa , Chris Harris , g h i Daniel Kwayisi , Sam Torkorno and Ebenezer Ebo Yahans Amuah a b Department of Geological Engineering, School of Engineering, University for Development Studies, Nyankpala, Ghana; Department of Geology, University of Ibadan, Ibadan, Nigeria; Pan African University Life and Earth Sciences Institute (PAULESI), University of Ibadan, Ibadan, Nigeria; Department of Geological Sciences, School of Geoscience, University of Energy and Natural Resources, Sunyani, Ghana; e f Department of Earth Science, Olabisi Onabanjo University, Ago-Iwoye, Nigeria; Department of Geological Sciences, University of Cape g h Town, Rondebosch, South Africa; Department of Earth Science, University of Ghana, Accra, Ghana; Exploration Department, Pelangio Gold Ltd, Ghana; Environmental Science Department, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana ABSTRACT ARTICLE HISTORY Received 6 July 2022 This study utilised lithogeochemistry and stable isotope geochemistry to assess the gold Revised 17 October 2022 mineralisation in the host rocks of the Abansuoso area. It also studied the isotopic signature Accepted 26 October 2022 of hydrothermally altered and unaltered rocks as a tool for the preliminary prospecting of gold in Ghana’s Birimian belt. The research identified a close association between Au and As, Mo, KEYWORDS and W, which is characteristic of orogenic gold deposits. The mineralised and unmineralised Stable isotopes; mineral rocks show similar content of major and trace elements which suggests that the mineralisation exploration; geochemistry; did not significantly affect the primary composition of the mineralised rocks. The concentra- gold; Birimian; Ghana tions of δ O (8.9‰ to 17.4‰, n = 22, mean = +13.7‰) and δD (−117 to −33‰, n = 22, mean = −57.9‰) in the rocks are comparable to those from other Birimian regions in Ghana and West Africa. The lack of variation in δ O suggests a system buffered by rocks, with metamorphic fluid originating from rocks with identical δ O values. However, the effect of mineralisation on the δD values of rocks is substantially more apparent than its effect on the δ O values of the rocks. 1. Introduction et al., 2017). According to Chudasama et al. (2016), the Prior to the late 1980s, when gold exploration was Abore and Dynamite Hill gold deposits are also extre- rekindled in the Birimian metallogenic region of mely important. Ghana, two key types of gold deposits were studied Researchers continue to debate the distinctions in depth; the first was auriferous quartz-pebble con- between orogenic gold-bearing granitoids and intru- glomerates. The second was lode-quartz veins and sion-related gold deposits due to the similarities in disseminated sulphides in shear zones (Kesse, 1985; their wall-rock alteration, elemental relationships, Hirdes and Leube, 1989; Leube et al., 1990; Schmidt structural controls, and ore fluids (Goldfarb et al., Mumm et al., 1997; Smith et al., 2016). Extensive 2001; Groves & Phillips, 1987). Consequently, research in Ghana’s Birimian terrain has led to the a number of granitoid-hosted gold deposits, including discovery of a third and more recent type of gold the True North deposit (Canada) and the Muruntau mineralisation (Amponsah et al., 2016; Yao et al., gold deposits (Uzbekistan), have been categorized as 2001). This deposit type is related to granitic rocks. both orogenic and intrusion-related deposits (Kempe A good example is the Ayanfuri deposit located within et al., 2001; Hart et al., 2002). It is therefore essential to the Ashanti belt, which contains multiple gold occur- understand each model, as each classification, i.e., rences within intermediate to felsic granitoids. intrusion-related or orogeny-related, may necessitate The Birimian gold deposits in granitoids are emer- distinct exploration strategies (Goldfarb et al., 2005). ging as increasingly promising exploration targets, This study employs lithogeochemistry and stable with the aim of increasing the supply of gold ore and isotope geochemistry to evaluate the gold-bearing boosting economic output (Bouabdellah and Slack, granitoid and metasedimentary rocks that have 2016). The gold deposits in the western portion of received relatively less regional and local research Ghana’s Sefwi and Ashanti Belts (Chirano, Nhyiaso, attention (Yao et al., 2001; Agbenyezi et al., 2020). In and Ayankyerim) are prime examples of gold deposits addition, the nature of hydrothermal fluid migration hosted by granitoids (Allibone et al., 2004; Fougerouse and interaction with the country rocks in the study CONTACT Raymond Webrah Kazapoe rkazapoe@yahoo.com Department of Geological Engineering, School of Engineering, University for Development Studies, Nyankpala, Ghana This article has been corrected with minor changes. These changes do not impact the academic content of the article. © 2022 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group on behalf of the International Water, Air & Soil Conservation Society(INWASCON). This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 2 R. W. KAZAPOE ET AL. area is evaluated. This study also aims to assess the (Dzikunoo et al., 2021; Perrouty et al., 2015). stable isotopic signature of the rocks in order to deter- The second orogenic event or the Eburnean II is mine the difference in isotopic composition between characterised by regional NW-SE extension and hydrothermally altered and unaltered rocks. This will formation of the Birimian metasedimentary basins serve as a preliminary investigation tool for identifying between 2150 and 2116 Ma (Allibone et al., 2004). gold mineralization in the Paleoproterozoic rocks of The study area is underlain by the Sefwi Ghana and, by extension, West Africa. Greenstone Belt which is one of the six known Birimian Greenstone Gold Belts in Ghana (Figure 1). These belts are known to host different 2. Geological setting types of gold mineralisation (Perrouty et al., 2015). Out of the six Greenstone Gold Belts, two Two discrete orogenic events proposed by have the most productive mines operating on Allibone et al. (2002) have been recognised in them whilst the others have advanced exploration Ghana. The earliest of these is termed the and minimal mining works. The two most prolific Eburnean I and is believed to be associated with belts are the Sefwi and Ashanti Greenstone Gold the eruption of the Birimian volcanic rocks and Belts with the Sefwi Belt lying Northwest of the the intrusion of granites, followed by regional Ashanti Belt. Similar to the other greenstone belts metamorphism between 2200 and 2150 Ma Figure 1. Geological map of the study area (modified after Allibone et al., 2004). GEOLOGY, ECOLOGY, AND LANDSCAPES 3 in Ghana, the Sefwi Belt is composed of metavol- a thermogravimetric analyser (TGA). Forty-one (41) canic, metasedimentary and volcano-sedimentary selected minor and trace elements were analysed. rocks which have been intruded by granitoid and Thirty-one (31) minor and trace elements were ana- mafic units (Galipp et al., 2003; Senyah et al., lysed using lithium borate fusion ICP-MS. 2016). It is bordered to the west by the Sunyani Concentrations of the other ten (10) elements (Ag, Basin and to the east by the Kumasi Basin (Griffis Co, Ni, Zn, As, Cu, Pb, Cd, Mo and Sc) were obtained & Agezo, 2000). Rocks found in the Sefwi Belt by ICP-AES from a separate four-acid digestion fol- include metamorphosed tholeiitic lavas containing lowing standard procedure. The detection limit of the basalts, andesites, dacites, rhyolites, and volcano- elements was 0.05 mass percent. Precision and accu- clastic units comprising pyroclastics and epiclas- racy were better than 5% and 3% respectively. tics as well as minor metasedimentary rocks and The concentrations of major oxides, minor ele- syn-volcanic tonalitic to granodioritic granitoids ments, and trace elements were measured to classify (Jessell et al., 2012). the lithogeochemical signatures of the rocks found in the study area, which represent varying degrees of alteration. Various statistical techniques, including correlation matrix, binary diagrams, biplots, and 3. Materials and methods R-mode factor analysis, were utilised to present and evaluate the data. This allowed for the identification of Fieldwork consisted of lithological and structural log- metal associations and the delineation of potential ging and geological mapping. The drill cores were mineralising and/or geological signatures. In addition, logged in order to provide structural, lithological, scatter plots were created to examine the linearity of alteration, and mineralisation data on the deposit. the dataset, a standard assumption made when calcu- After logging, samples from two (2) drill cores lating correlation coefficients. The purpose of the fac- (SFDD 79 and SFDD 81) were extracted. The samples tor analysis was to clarify the correlation structure of were chosen based on their gold content to accurately the variables based on a smaller number of factors. To represent the altered zone, mineralized zone, and enable data computation, analysed samples with unmineralised zone in order to characterise the miner- values below their respective analytical detection lim- alisation of the area. Samples with gold concentrations its were assigned a value equal to half the analytical greater than 0.3 g/t are considered mineralised, detection limit. Before performing population statis- whereas samples with gold concentrations less than tics on the data, the analysed data were evaluated to 0.3 g/t are considered unmineralised. This grade clas- eliminate elements with insufficient variance and sification is based on the typical gold grades found in duplication. The data was log-transformed (base 10) the Birimian terrain of Ghana and has been deemed to decrease its wide distribution. suitable for academic research (Allibone et al., 2004; The oxygen and hydrogen isotope ratios of twenty- Griffis et al., 2002; Mumin et al., 1996; Oberthuer et al., two (22) rock samples taken from the mineralised (8), 1996). Additionally, surface rock samples were also hydrothermally altered (7) and unmineralised zones collected. Thin section preparation was done at the (7) were analysed in the Stable Isotope Laboratory of Department of Earth Science, University of Ghana of the Department of Geological Sciences, University of both oriented and un-oriented rock specimens to per- Cape Town, South Africa. Isotope ratios were mea- mit the identification of constituent minerals, and sured using a Thermo DeltaXP mass spectrometer. deformation study of textures and fabric elements Aliquots of ~10 mg of the rock powders were dried using an optical microscope. Scanning electron micro- overnight at 50°C and then within a vacuum in nickel scopy (SEM) images were taken using the Zeiss EVO reaction vessels. The powders were then reacted with MA 15 SEM. All images were taken using the back- 30 kPa of CIF for ~4 hours to extract the oxygen from scatter detector (BSD). A Bruker Quatax EDS system silicates (Borthwick and Harmon, 1982). The with XFlash Detector 610 M was employed as the extracted oxygen was then converted to CO by pas- energy dispersive X-ray detector (EDS). SEM micro- sing it over a high-temperature platinised carbon rod. graphs of the samples were collected using an accel- Full analytical details are described by Vennemann erating voltage of 18–20 kV with a working distance and Smith (1990) and Harris and Vogeli (2010). that ranged between 10 mm to 14 mm. Unknowns were run with duplicates of the internal Twenty-two (22) rock samples were sent to the ALS quartz standard (MQ), which was used to calibrate the geochemical laboratory in Vancouver for whole-rock, raw data to the SMOW (Standard Mean Ocean Water) major, minor and trace elements. The rocks were scale, using a δ O value of 10.1 for MQ (calibrated taken from the mineralised zone (8 samples), and the against NBS-28). The results are reported in standard unmineralised zone (14 samples). Whole-rock major δ-notation, where δ = (R /R – 1) × 1000, element analysis was carried out using ICP-AES from sample standard 18 16 R is O/ O in the sample, and R is the ALS laboratory in Australia. Loss on Ignition sample standard 18 16 O/ O relative to SMOW. The analytical error is (LOI) at 1000°C was determined using 4 R. W. KAZAPOE ET AL. estimated as ± 0.2 ‰ (2 sigmas), based on long-term alongside the unknowns. The measured water content repeated analysis of MQ. of Serina kaolinite gave 12.42 wt% (± 0.5 wt% Hydrogen isotopes and water contents of the same 2σ, n = 3). powdered separates were determined using the method of Vennemann and O’Neil (1993). Samples 4. Results were melted in quartz glass tubes using a propane torch. The raw data was normalised to SMOW, and 4.1 Local geology corrected for compression, using the water standards RMW (δD = – 131.4‰) and CTMP2010 (δD = – The petrographic investigations reveal the two (2) 7.4‰) with the in-house Serina kaolinite standard main types of rocks in the area as gneisses and analysed with the unknowns (Serina bulk kaolinite, schist with the presence of quartz veins. δD = – 57‰, H O = 12.4 wt%; Harris et al., 1999). The measured δD values of the standard gave an 4.1.1 Gneisses average of – 59.5 ± 3.6‰ (2σ, n = 3) after corrections The gneisses are of two types based on mineralogical to the raw data. All data were adjusted to the accepted composition, the biotite and hornblende-biotite Serina kaolinite value. Water abundances (as H O ) gneisses. The biotite gneisses outcrop in the north- were measured from the voltage on the mass 2 collec- western part of the study area. They are medium to tors of the mass spectrometer. This was calibrated coarse-grained and weakly to strongly foliated (Figure 2 against measured volumes of standard waters analysed (a)). The dominant minerals are quartz, plagioclase, Figure 2. (a) xpl photomicrograph of biotite gneiss showing microcline (Mc), biotite (Bio) and quartz (Qtz) (b) xpl photomicrograph of biotite gneiss showing garnet (Grt) and hornblende (Hbl) (c) xpl photomicrograph of schist showing sericite (Ser) and calcite (Cal) alteration (d) xpl photomicrograph of sericite schist showing calcite (Cal) and quartz (Qtz) (e) xpl photomicrograph of sericite schist showing sericite alteration (Ser) and calcite alteration (f)) xpl photomicrograph of schist showing muscovite (Ms). GEOLOGY, ECOLOGY, AND LANDSCAPES 5 Figure 3. Photomicrographs of quartz showing (a) strong elongation and pressure shadow and (b) chess board extinction. microcline, and biotite (Figures 2(a)). Under thin sec- The sericite-quartz schist is medium-grained and tion, cubic to slightly elongated opaque minerals sus- composed of sericite and quartz with a trace amount pected to be sulphide can be observed. The foliation and of muscovite (Figure 2(e)). Fine to medium-grained mineral lineation are defined by the preferred orienta- texture and strong foliation characterise the musco- tion of biotite and elongated opaque minerals. The vite-quartz schist (Figure 2(f)). This rock is domi- quartz minerals exhibit an anhedral shape with undu- nantly composed of quartz and muscovite together lose extinction. The quartz minerals also appear recrys- with minor feldspar. Rare occurrence of disseminated tallised and occur as either coarse crystals or medium- cubic sulphides can be observed in the rocks. The rock grained and elongated aggregates. Plagioclase and is cut by massive quartz veins. microcline are subhedral to anhedral and altered into sericite. The biotite minerals show very weak hydro- 4.1.3 Quartz vein thermal alterations mostly into chlorite. Quartz veins are very common in the study area, The hornblende-biotite gneiss on the other hand is usually filling fractures. The veins are associated with coarse-grained and foliated. The rock is composition- all the rock types but are more common in the biotite ally banded with felsic bands alternating with mafic gneiss. Both syntaxial and antitaxial veins are observed bands. The dominant minerals are hornblende, biotite, in the study area. The veins are composed mainly of garnet, plagioclase, quartz, and microcline (Figure 2 quartz and carbonates and have thicknesses varying (b)) with the mafic bands comprised of the first three between 0.1 cm and 20 cm. Quartz in the veins is minerals whereas the latter three comprise the felsic coarse-grained and subrounded to subangular. bands. The garnet and microcline minerals are very Although this quartz is deformed, they show only coarse and poikilitic with inclusions of quartz and weak recrystallisation without evidence of grain occasional biotite (Figure 2(b)). Associated with the boundary migration. The quartz grains show undulose garnet and often occurring within the foliation are extinction with some exhibiting chessboard extinction subhedral to cubic opaque minerals suspected to be (Figure 3(a-b)). In some samples, the quartz is also sulphides. Hornblende has partially altered to epidote elongated. whereas the feldspars and biotite are partially altered to sericite and chlorite, respectively. In the hand sam- 4.2 Gold mineralisation ple, the biotite appears to dominate in composition with quartz also prominent. 4.2.1 Sulphide characterisation The principal ore minerals are arsenopyrite and 4.1.2 Schists a small amount of pyrite, both of which are found in The schists are of two types: the carbonate-sericite the mineralised zones of the altered host rock. The schist and sericite-quartz schist. The carbonate- mineralised rocks contain an average of 20% sulphides sericite schist is medium-grained and weakly foliated by volume, with arsenopyrite accounting for 50%, (Figure 2(c)). It is composed of quartz, plagioclase, pyrite for 30%, and galena for the remaining 20%. muscovite and calcite. The muscovite is strongly ser- Both primary sulphide ore minerals form in sizes icitised. The calcite occurs as very coarse crystals in the ranging from 1 μm to approximately 8 μm. Gold veins and also as impregnations in the rock (Figure 2 mineralisation occurs as micron-sized gold inclusions (c-d)). Quartz occurs as recrystallised grains and may of approximately 2 μm (Figure 4(a-b)) and larger gold be associated with the calcite or as impregnations, and grains within fractures of euhedral arsenopyrite/pyr- exhibits undulose extinction. ite, and predominantly as free gold associated with 6 R. W. KAZAPOE ET AL. Figure 4. Scanning Electron Microscopy (SEM) images of (a) sulphide phases associated with chlorite schist showing micron-sized gold (2 μm). (b) sulphide phases associated with chlorite schist showing inclusions of galena (Gn), pyrite (Py) and arsenopyrite (Apy). (c) ore zone showing gold associated with sericite (d) ore zone showing relationship of Au with dolomite (Dol). Table 1. A paragenetic scheme in the Abansuoso deposit showing the relative abundance of minerals during the regional metamorphic stage, pre-ore-forming stage and ore-forming stage. Mineral Regional Metamorphism Stage One Stage Two Plagioclase Quartz Muscovite Biotite Hornblende Garnet Chlorite Hematite Sericite Epidote Dolomite Pyrite Arsenopyrite Galena Gold sericite alteration in the host lithologies (Figure 4 infiltration, chlorite, epidote, hematite, sericite, (c-d)). Gold associated with sericite alteration is also and opaque mineralisation. The ore-forming pro- accompanied by dolomite (Figure 4(d)). A minor cess was a later event that appears to be largely amount of galena is also found as small inclusions hydrothermal alteration, but regional metamorph- within the arsenopyrite (Figure 4(b)). ism might have contributed to some fluid genera- tion or P-T-X within the greenschist zone 4.2.2 Mineral paragenesis of gold deposit The mineral assemblage of sericite, quartz, mus- The paragenetic sequence in the Abansuoso deposit covite, biotite and hornblende is suggestive of is summarised in Table 1. The silicate minerals greenschist to lower amphibolite metamorphic con- such as quartz, feldspar, micas, hornblende, pyrox- ditions. Although garnet was observed in some of ene as well as garnet formed earlier in the para- the gneissic rocks, they have been resorbed by genetic sequence, and are linked to regional quartz and thus may be pre-kinematic grains. The metamorphism. This was followed by carbonate alteration pattern within the area is chlorite- GEOLOGY, ECOLOGY, AND LANDSCAPES 7 4.2.3 Mineralisation The concession lies within a ductile sinistral shear zone with a brittle component. The mineralisation occurs in biotite gneiss of intermediate composition and schist of varying composition both of which have been affected by various stages of metamorphism and deformation. The sequence of alteration invades the rocks via stringers along weak zones caused by the shearing. The alteration is predominantly quartz- sericite, and pyrite with hematite at the fringes where low-grade gold is reported (Figure 6). The main mineralised zones are characterised by sericite altera- tion coupled with fine-grain pyrite and associated with dolomite. The preliminary sequence of alteration has Figure 5. SEM image showing elongated euhedral arsenopyr- been deduced as chlorite-carbonate-hematite-sericite- ite aligned with a fracture. chlorite-quartz/pyrite with minor albite. carbonate-hematite-sericite-epidote-sulphide. Since 4.3 Lithogeochemistry the sulphide (arsenopyrite) is occasionally elon- The major and trace element concentrations of the gated and follows the foliation (Figure 5), they schist and gneisses are presented in Tables 2 and 3 may have formed during the same deformational respectively. Histogram plots for the unmineralised event that developed the foliations, thus, syn- and mineralised rocks (Figures 7(a-d)) from the study kinematic. Figure 6. Schematic representation of drillcore SPDD 079. 8 R. W. KAZAPOE ET AL. Table 2. Stable isotopes, major and trace elements composition of the schist. 020/79 009/79 013/79 019/79 011//79 012/79 014/79 001/81 OS/A 001/OS 008/PK Schist Schist Schist Schist Schist Schist Schist Schist Schist Schist Schist δ O‰ 14.3 13.3 13 13.9 13.9 14.1 12.7 13.1 14.4 14.4 14.1 δD‰ −45 −42 −50 −62 −34 −33 −49 −47 −60 −51 −51 SiO 72 74.9 65.2 74.4 64.9 74.4 67.1 60.4 76.9 75.7 79.5 Al O 10.05 9.38 12.1 10.65 10.6 10.7 11.75 9.6 11.15 10.55 12.2 2 3 Fe O 7.91 4.87 6.92 5.92 6.68 4.49 7.54 7.49 2.39 2.49 1.4 2 3 CaO 0.54 1.66 2.19 0.56 3.81 1.27 2.29 4.83 1.31 1.56 0.03 MgO 0.24 0.61 0.57 0.2 0.99 0.4 0.92 2.33 0.44 0.5 0.2 Na O 0.31 2.03 3.89 4.13 1.73 3.55 4.04 3.96 3.06 2.85 0.12 K O 3.06 1.93 1.74 1.14 2.4 1.49 2.18 0.99 1.91 1.86 3.79 TiO 0.53 0.45 0.64 0.58 0.66 0.51 0.68 0.67 0.25 0.22 0.26 MnO 0.02 0.13 0.16 0.06 0.23 0.12 0.15 0.13 0.05 0.06 0.01 SiO /Al O 7.16 7.99 5.39 6.99 6.12 6.95 5.71 6.29 6.90 7.18 6.52 2 2 3 Na O/K O 0.10 1.05 2.24 3.62 0.72 2.38 1.85 4.00 1.60 1.53 0.03 2 2 LOI 5.28 4.6 6.26 2.29 8.19 3.82 3.83 8.96 2.75 2.93 2.07 Total 100.19 100.76 100.06 100.21 100.58 101.05 100.8 99.5 100.4 98.9 99.69 Au 1.66 2.1 1.04 2 1.4 0.9 0.15 0.22 0.11 0.01 0.11 Ba 732 303 527 216 1430 525 796 158.5 583 591 748 Ce 52.5 38.9 56.7 50.5 57.9 49.7 50.3 12.6 53.8 47.4 56.5 Cr 140 190 170 180 160 190 180 170 200 140 140 Cs 1.64 0.9 0.97 0.57 1.42 0.91 1.11 0.96 1.13 1.17 1.1 Dy 7.12 7.21 10.8 8.43 9.36 6.97 8.15 2.25 6.49 5.71 10.25 Er 4.28 4.48 6.52 5.48 6.11 4.34 5.06 1.33 4.3 3.67 6.66 Eu 1.62 1.52 2.31 1.75 2.04 1.77 1.82 0.61 1.08 0.91 1.65 Ga 15.3 16 17.4 11.7 12.7 15.7 19.9 9.8 14.4 13.1 19.6 Gd 7.82 7.25 10.5 8.47 9.27 7.79 8.54 2.38 6.86 5.96 9.54 Hf 5.9 4.9 7 6.6 7.4 5.6 6.1 1.5 6.8 6.2 8.5 Ho 1.42 1.45 2.19 1.77 1.88 1.41 1.6 0.44 1.27 1.14 2.1 La 24.4 18.9 27.1 23.8 27.1 24.2 23.6 5.8 25.8 23.5 28.5 Lu 0.65 0.62 0.87 0.74 0.96 0.65 0.71 0.21 0.69 0.57 0.94 Nb 5.9 5.3 7.3 5.7 7.6 5.7 6.3 1.7 7.4 5.1 7.7 Nd 30.6 23.5 33.4 28.2 33.3 28.6 29.8 7.7 28.1 24.7 32.7 Pr 7.53 5.69 8.3 7.2 8.32 7.25 7.27 1.91 7.38 6.54 8.2 Rb 67.6 41.4 36.6 24.9 48.9 34.7 45 26.1 42.8 42.1 79.5 Sm 7.51 6.6 9.1 7.41 8.72 7.52 7.7 2.21 6.61 5.82 8.38 Sn 3 1 2 1 3 1 1 1 2 1 2 Sr 38.3 93.8 161.5 82.1 266 92.5 208 225 102.5 117 9.1 Ta 0.4 0.3 0.4 0.4 0.5 0.3 0.4 0.1 0.4 0.4 0.5 Tb 1.2 1.17 1.68 1.34 1.49 1.24 1.32 0.38 1.07 0.9 1.57 Th 3.2 2.5 3.42 3.13 3.72 3.11 2.99 0.69 3.99 3.54 4.05 Tm 0.64 0.63 0.89 0.79 0.92 0.63 0.71 0.19 0.65 0.56 0.93 U 1.31 1.02 1.54 1.14 1.19 1.22 1.17 0.3 1.13 1.21 1.62 V 33 16 25 21 68 23 19 150 27 19 10 W 40 36 72 46 55 48 3 8 14 12 18 Y 41.1 41.6 62.7 49 55.2 39.1 46.3 12.6 38 33.1 61.7 Yb 4.32 4.24 6.15 5.38 6.33 4.28 5 1.41 4.62 3.78 6.57 Zr 214 176 251 227 260 198 215 51 243 217 310 Ag 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 As 261 36 281 27 114 40 5 5 38 28 24 Cd 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.6 0.5 0.5 0.5 Co 6 3 4 4 2 3 6 25 3 3 1 Cu 42 14 9 8 40 6 12 176 9 10 4 Li 10 10 10 10 10 10 10 10 10 10 10 Mo 3 3 3 3 3 3 3 2 3 3 2 Ni 4 1 1 4 1 3 1 6 4 3 1 Pb 21 4 7 5 10 2 8 3 6 3 3 Sc 7 12 17 12 15 12 15 15 7 7 3 Tl 10 10 10 10 10 10 10 10 10 10 10 Zn 209 100 109 94 122 81 86 82 39 44 60 area show that Pb is relatively depleted in the miner- associated with the arsenopyrite occurs with minerali- alised samples which is characteristic of orogenic gold sation (Amponsah et al., 2015). deposits (Wang et al., 2021). However, this differs from reports on the Birimian at the Julie deposit and Obuasi 4.3.1 Major elements deposits in the Northeast and Southwest of Ghana The variations in MgO versus SiO show the schist respectively (Dzigbodi-Adjimah, 1993; Amponsah and the gneiss both have similar SiO content with the et al., 2015). Additionally, elements such as Co, Cu gneisses having marginally higher MgO content and Ni are also more enriched in the unmineralised (Figure 8(a)). The relatively high SiO content of the samples as compared to the mineralised samples. This rocks suggests a felsic source for the rock in the study differs from what has been reported in the Bekpong area. The graph shows a negative correlation between deposit on the Lawra belt of Ghana, where Ni and Co GEOLOGY, ECOLOGY, AND LANDSCAPES 9 Table 3. Stable isotopes, major and trace elements composition of the gneiss and quartz. 040/PK 032/TMB 027/KS 007/81 017/81 018/81 014/OKS 013/N 027/QZ 006/N 006/81 Gneiss Gneiss Gneiss Gneiss Gneiss Gneiss Gneiss Gneiss Quartz Quartz Quartz δ O‰ 13.2 8.9 12 14.4 14.2 14.1 13.6 14.4 12.7 17.4 14.7 δD‰ −58 −43 −68 −53 −48 −50 −91 −53 −71 −117 −98 SiO 65.2 66.9 75.9 54.3 80.8 74.1 76.9 73 97.6 100 99.3 Al O 12.1 10.5 11.4 12.6 8.91 11.55 11.35 11.6 0.27 0.39 0.05 2 3 Fe O 6.92 6.58 3.01 11.15 2.37 2.96 3.01 5.9 0.83 0.82 0.7 2 3 CaO 2.19 3.61 1.83 4.54 1.4 1.34 0.43 1.87 0.02 0.02 0.01 MgO 0.57 0.99 0.54 2.8 0.48 0.56 0.18 0.24 0.01 0.02 0.01 Na O 3.89 1.76 3.18 5.07 3.68 4.87 4.13 3.22 0.06 0.03 0.01 K O 1.74 2.6 3.76 0.78 0.71 1.67 2.84 3.07 0.03 0.09 0.01 TiO 0.64 0.66 0.29 1.18 0.28 0.33 0.23 0.37 0.01 0.01 0.01 MnO 0.16 0.23 0.5 0.19 0.06 0.06 0.03 0.13 0.01 0.01 0.01 P O 0.3 0.19 0.04 0.23 0.08 0.08 0.02 0.05 0.01 0.03 0.01 2 5 SiO /Al O 5.39 6.37 6.66 4.31 9.07 6.42 6.78 6.29 361.48 256.41 1986.00 2 2 3 Na O/K O 2.24 0.68 0.85 6.50 5.18 2.92 1.45 1.05 2.00 0.33 1.00 2 2 LOI 6.26 8.19 0.5 7.45 2.1 2.41 0.57 0.65 0.04 0.14 0.03 Total 100.06 100.58 100.59 100.38 100.93 100.05 99.81 100.24 98.93 101.61 100.15 Au 1.04 1 0.001 0.01 0.001 0.001 0.01 0.001 0.11 0.001 0.01 Ba 527 1470 1205 443 206 597 1040 1050 18.6 131 2 Ce 56.7 58.7 67 23.1 35.4 38.5 36.6 44.2 0.6 38.9 0.3 Cr 170 158 120 60 150 200 60 110 350 320 350 Cs 0.97 1.24 0.19 1.02 0.79 0.58 0.41 1.81 0.01 0.04 0.01 Dy 10.8 8.96 10.25 4.05 3.87 4.11 8.42 7.81 0.07 1.28 0.05 Er 6.52 6.14 6.49 2.42 2.31 2.59 5.79 5.06 0.04 0.29 0.03 Eu 2.31 2.1 1.6 1.25 0.7 0.97 1.13 1.44 0.02 0.89 0.02 Ga 17.4 11.8 19.8 16.3 10.1 12 19.1 18.7 0.5 0.7 0.1 Gd 10.5 8.72 10.3 4.21 4.11 4.59 7.07 6.98 0.09 2.86 0.05 Hf 7 6.7 9 2.7 4.1 4.4 8.8 7.2 0.1 0.1 0.1 Ho 2.19 1.71 2.03 0.81 0.75 0.8 1.78 1.56 0.01 0.16 0.01 La 27.1 26.1 30.5 11.2 17.2 18.1 15.5 19.5 0.2 29.6 0.2 Lu 0.87 1.1 1.01 0.35 0.36 0.43 0.9 0.74 0.01 0.03 0.01 Nb 7.3 8.6 8 3 4.9 4.6 8.6 7.1 0.1 0.1 0.1 Nd 33.4 32.1 36.8 14.4 18.4 20.3 24.6 24.9 0.4 24.6 0.2 Pr 8.3 8.62 9.47 3.47 4.77 5.3 5.84 6.16 0.09 7.31 0.05 Rb 36.6 51.9 61.3 22 20.2 28.4 55.8 64.2 0.9 1.9 0.2 Sm 9.1 6.78 9.36 3.89 4.08 4.36 6.12 6.5 0.09 5.05 0.06 Sn 2 2 3 1 1 1 1 2 1 1 1 Sr 161.5 252 136 280 187.5 225 56.4 144 2 20.7 0.6 Ta 0.4 0.5 0.5 0.1 0.3 0.3 0.4 0.4 0.1 0.1 0.1 Tb 1.68 1.52 1.63 0.65 0.62 0.68 1.24 1.19 0.01 0.29 0.01 Th 3.42 4.72 4.22 1.32 2.69 3.03 4.38 3.14 0.05 0.05 0.05 Tm 0.89 0.94 0.95 0.34 0.34 0.38 0.86 0.75 0.01 0.04 0.01 U 1.54 1.22 1.62 0.57 1.33 1.12 1.62 0.85 0.05 0.1 0.05 V 25 52 5 302 17 19 6 5 5 5 5 W 72 65 1 2 10 1 1 1 1 2 1 Y 62.7 53.2 58.3 22.7 21.2 23.4 51 44.2 0.5 3.5 0.1 Yb 6.15 5.43 6.54 2.3 2.43 2.7 6.25 5.16 0.08 0.19 0.03 Zr 251 256 321 93 146 155 317 254 2 2 2 Ag 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 As 281 114 260 5 5 5 5 5 5 5 5 Cd 0.5 0.5 0.5 0.6 0.5 0.5 0.5 0.5 0.5 0.5 0.5 Co 4 2 1 26 3 4 1 1 1 1 1 Cu 9 40 8 51 12 3 9 7 3 4 4 Li 10 10 10 30 10 10 10 10 10 10 10 Mo 3 3 2 1 3 2 3 2 3 3 5 Ni 1 1 1 4 7 2 3 1 3 3 4 Pb 7 10 8 2 7 7 10 10 2 9 2 Sc 17 15 6 24 6 7 5 12 1 1 1 Tl 10 10 10 10 10 10 10 10 10 10 10 Zn 109 122 134 102 61 57 17 148 5 4 3 SiO and MgO which is characteristic of such rocks. correlation matrix shows that Al O has a strong 2 2 3 Binary plots and regression coefficients were positive correlation with TiO (R = 0.912), K 2 2 2 2 used to determine the relative mobility among the O (R = 0.945), and Fe O (R = 0.705). Figure 8 2 3 elements from the samples in the study area. (b) shows that TiO and Al O are the most immo- 2 2 3 According to de Almeida (2009), immobile ele- bile pair, regardless of the hydrothermal alteration ments are insoluble and will therefore result in and mineralisation. This strong positive correlation a regression line which passes through the origin between these oxides indicates that they were prob- of a binary diagram and will also record the highest ably immobile during the hydrothermal event. regression coefficient in a correlation matrix. The Additionally, both mineralised and unmineralised 10 R. W. KAZAPOE ET AL. Figure 7. Histogram plots of (a) trace elements concentrations of mineralized schist samples (b) trace elements concentrations of unmineralized schist samples (c) trace elements concentrations of mineralised gneiss samples (d) trace elements concentrations of unmineralized gneiss samples. Figure 8. (a) a plot of MgO versus SiO of the schist and the gneiss from the Abansuoso area (b) Plot of TiO versus Al O following 2 2 2 3 an approximate linear trend which indicates that they are the most immobile pair, regardless of the alteration and mineralization. GEOLOGY, ECOLOGY, AND LANDSCAPES 11 Figure 9. The plot of principal component I against II for the trace elements from the unmineralized samples. Figure 10. The plot of principal component I against II for the trace elements from the mineralised samples. samples align in an approximately linear trend, 4.3.3 Factor analysis suggesting that these rocks may have originated Factor analysis has been used to determine element from a shared homogenous source. association with gold from whole-rock samples from the Atacora structural unit in Benin successfully (Adjo 4.3.2 Element association et al., 2021). Additionally, Yigit and Hofstra (2003) The close association of As, Mo and SiO shows that As applied the multivariate analysis of geochemical data is linked with the felsic minerals abundant in rocks such on whole-rock samples to assess favourable geochem- as the schist. MgO is associated with MnO and CaO, ical attributes of the host lithologies as well as to indicating they originated from the mafic mineral- characterise alteration and associated element enrich- bearing rocks such as the biotite-rich gneiss (Figure 9). ments and depletion. The biplot for the mineralised samples strongly Four principal components with eigenvalues > 1 suggests the influence of hydrothermal fluids in the were extracted which accounted for a total variance element association of the rocks. Au is associated with of 81.04% (Table 4). PC1 with a total variance of Zn, P O , As, W and Ba (Figure 10). These are the 32.08%, has K O, Ga, Ba, U, Al O , Zn, TiO , MgO, 2 5 2 2 3 2 array of elements typical of orogenic gold deposits of CaO and Cr strongly loaded as the components and the Birimian (Klemd and Hirdes, 1997). This particu- these factors are plotted in rotated space applying R – lar array shows that these elements are the pathfinders mode factor analysis. PC2 accounts for 28.82% of the for Au in the study area. Additionally, the association total variance with TiO , V, Co, Cd, MgO, Fe O , 2 2 3 of Mo with Ga, K O and Cr shows the influence of CaO, Cu and Ni with negative loading for SiO as 2 2 alteration. the factors loading strongly (Table 4). PC3 which 12 R. W. KAZAPOE ET AL. a. Table 4. Rotated factor Matrix Factor 1 2 3 4 Log(K O) 0.952 0.207 0.097 −0.122 Log(Ga) 0.944 0.266 0.090 0.079 Log(Ba) 0.934 0.104 0.020 0.032 Log(U) 0.934 −0.082 0.225 0.088 Log(Al O ) 0.917 0.345 0.080 0.060 2 3 Log(Zn) 0.776 0.411 0.311 0.062 Log(TiO ) 0.730 0.601 0.225 0.146 Log(Pb) 0.418 −0.055 0.069 −0.050 Log(V) 0.079 0.955 0.101 −0.023 Log(Co) 0.032 0.951 −0.010 −0.086 Log(Cd) −0.115 0.871 −0.292 −0.164 Log(SiO ) −0.403 −0.838 0.091 0.277 Log(MgO) 0.578 0.777 −0.058 −0.087 Log(Fe O ) 0.492 0.723 0.283 0.120 2 3 Log(CaO) 0.594 0.695 −0.009 0.069 Log(Cu) 0.065 0.599 0.203 0.496 Log(As) 0.242 −0.102 0.871 −0.047 Log(W) 0.244 0.111 0.824 −0.245 Au 0.098 0.028 0.805 0.024 Log(Mo) −0.332 −0.489 0.523 0.135 Log(Na O) −0.161 −0.038 −0.187 0.653 Log(Cr) −0.512 0.153 0.161 −0.619 Log(Ni) −0.209 0.558 −0.232 −0.607 Extraction Method: Principal Axis Factoring. Rotation Method: Varimax with Kaiser Normalization. Rotation converged in 7 iterations. comprises As, Au, W and Mo as the strongly loaded altered samples in this study. The altered samples factors accounts for 12.73% of the total variance. were selected to reflect the main phases of hydrother- While PC4 has a total variance of 7.41 with Na O, Cr mal alteration identified in the study area (i.e., sericite and Ni as the strongly loaded factors. and hematite alterations). Average compositions for the least altered zones were plotted against samples from the average composition for the altered zones by 4.3.4 Mass balance: isocon diagrams using the EASYGRESGRANT MS-Excel sheet pro- The application of isocon diagrams in mass balance vided in Lopez-Moro (2012). The least altered samples studies developed by Grant, 1985) has been success- contain anomalous values of Cd, Cu, Mo, Ni, Se, fully applied to several studies in different deposit U and V, which are interpreted to be introduced in types around the world (e.g., Hofstra, 1994; Emsbo this unit during diagenesis, pre-dating the mineralisa- et al., 2003; Yigit and Hofstra, 2003; Almeida, 2009; tion event. Amponsah et al., 2015 among others). The isochron After the selection of the samples, the identification diagrams have been primarily used to evaluate major of the immobile elements was done. Binary plots and element gains or losses in hydrothermal processes by regression coefficients were used to identify that TiO comparing barren or the least altered samples to and Al O are the most immobile pair, regardless of 2 3 altered samples in a particular deposit. De Almeida the alteration and mineralisation (Figure 8(b)). These (2009) applied this technique to evaluate major ele- two oxides were therefore considered the most immo- ment gains and losses in the Carlin type deposit in bile pair and used for the isocon diagrams, following Nevada. Amponsah et al. (2015) also applied the tech- the method of Grant, 1985, Grant, 1985). The isocon nique in a study of the Julie deposit in the north of diagrams (Figures 11(a-c)). Ghana in a similar geological setting as the study area. The least altered samples used in this study were taken from the hills around Nfante and South of Subriso 4.4 Stable Oxygen (O) and Hydrogen (H) isotope away from the mining pits or deposits. The relative data level of hydrothermal alteration of the samples was evaluated based on the petrographic studies as well as The δ O values of the rocks found in the Abansuoso the whole-rock composition of the samples. deposit range from 8.9‰ to 17.4‰ (n = 22, According to Querubin and Yumul (2005) samples mean = +13.7‰). The δ O values in the schist are which are the least altered tend to characteristically between 12.7‰ – 14.4‰ (n = 11, mean = 13.7‰). have low Loss on Ignition (LOI) values of less than 5% These values are similar to those determined in the and have Na O values ranging between 3–5% for felsic gneisses which range between 8.9‰ – 14.4‰ (n = 8, samples and CaO + Na O falling between 5 and 10% mean = 13.1‰; Table 5). The range of δ O values for mafic gneisses. Samples which displayed these determined in the gneisses of the Abansuoso area are characteristics were therefore classified as the least similar to those recorded in the granitic rocks found GEOLOGY, ECOLOGY, AND LANDSCAPES 13 Figure 11. (a) Isocon diagrams constructed for pairs of unaltered and altered samples from the outer alteration zone (hematite alteration) (b) Isocon diagrams constructed for pairs of unaltered and altered samples from the inner alteration zone (Sericite alteration) (c) Isocon diagrams constructed for pairs of unaltered and altered samples from the inner alteration zone (Sericite alteration). Table 5. δ O and δD composition of rocks from the Abansuoso Area of Southwestern Ghana. SN Sample ID Rock Type δ O‰ δD‰ Au Grade (g/ton) H O (wt%) 1 020/79 Schist 14.3 −45 1.66 0.98 2 009/79 Schist 13.3 −42 2.1 0.91 3 013/79 Schist 13 −50 1.04 0.50 4 019/79 Schist 13.9 −62 2 0.66 5 011//79 Schist 13.9 −34 1.4 0.94 6 012/79 Schist 14.1 −33 0.9 0.60 7 014/79 Schist 12.7 −49 0.15 0.65 8 001/81 Schist 13.1 −47 0.22 0.81 9 OS/A Schist 14.4 −60 0.11 0.62 10 001/OS Schist 14.4 −51 0.22 0.83 11 008/PK Schist 14.1 −51 0.11 1.51 12 040/PK Gneiss 13.2 −58 1.3 0.56 13 032/TMB Gneiss 8.9 −43 1 2.53 14 027/KS Gneiss 12 −68 0.11 0.27 15 007/81 Gneiss 14.4 −53 0.01 0.74 16 017/81 Gneiss 14.2 −48 0.1 0.44 17 018/81 Gneiss 14.1 −50 0.1 0.78 18 014/OKS Gneiss 13.6 −91 0.01 0.30 19 013/N Gneiss 14.4 −53 0.1 0.34 20 027/QZ Quartz vein 12.7 −71 0.11 0.10 21 006/N Quartz vein 17.4 −117 0.1 0.10 22 006/81 Quartz vein 14.7 −98 0.01 0.12 around the Lake Bosomtwi area of the Ashanti Region δ O values determined in the quartz veins, although of Ghana (12.7‰ to 12.9‰, Koeberl et al., 1998). from a limited number of samples, show However, they are higher than the δ O values of the a comparatively wider range relative to the values Pepiakese granites of Southwestern Ghana (8.6‰ to (15.6‰ – 18.5‰) determined in quartz in the 9.0‰, Koeberl et al., 1998). The δ O values of the Prestea deposit by Hammond and Shimazaki (1994). quartz veins were higher, ranging from 12.7‰ – The δ O values determined for the rocks of the 17.4‰ (n = 3, mean = 14.9‰). The range of study area are within a narrow range of 12.0 to 14.7‰ 14 R. W. KAZAPOE ET AL. 8 10 12 14 16 18 -20 -40 -60 Schist Gneiss -80 Quartz vein -100 -120 -140 δ O‰ Figure 12. A plot of δ O against δD of the rocks from the study area. with the notable exception of two samples from the suggests that the variation of δD is not controlled schist and quartz veins which recorded values of 8.9‰ by depth. and 17.4‰, respectively. This generally narrow range of δ O values is similar to those previously recorded 4.4.1 Controls on oxygen and hydrogen isotope in the Birimian in Ghana (Hammond & Shimazaki, composition of the rocks in the Abansuoso area 1994; Oberthuer et al., 1996; Mumming et al., 1996). The H O vs δD shows a negative correlation between The δD values of the rocks for the Abansuoso area the water content and the δD for the samples from the range between −117 to −33‰ (n = 22, study area (Figure 12). This may reflect a greater mean = −57.9‰). The values in the schist range extent of water-rock reaction while the increasing δD from −62 to −33‰ (n = 11, mean = −47.6‰), while values suggest an influence from mineralogy and tem- the δD values of the gneisses however range from −91 perature of alteration. to −43‰ (n = 8, mean = −58‰). The δD values within The factors which control the final δD values of the unmineralised quartz vein have a wider range from minerals and as a result the bulk rock are tempera- −117 to −71‰ (n = 3, mean = −95.3‰). An inverse ture of crystallisation of the minerals in the rocks, relationship exists between δD and δ O values in the the δD values of the fluid present and interaction rocks from the study area (Figure 12). This is consis- with hydrothermal fluids (e.g., Harris et al., 1997). tent with the relationship between δD and δ O in this According to Taylor & South, 1985), hydrother- instance. mally altered rocks with significant chlorite content Samples collected from the mining pits present tend to have low δ O values as compared to other similar isotopic composition to the drill core samples. alteration minerals because chlorite has low 18 18 The δ O values of the surface samples range from 8.9 δ O. Some hydrothermally altered rocks are also to 17.4‰ (n = 10, average = 13.51‰) in comparison to characterised by elevated δD values as compared to the drillcore samples which range from 12.7–14.7‰ clay-rich rocks altered by the same fluids at lower (n = 12, average = 13.81‰). The values are similar to temperatures (Friedman and O’Neil, 1977). Casey values recorded in the Pepiakese granites of and Taylor (1982) determined δ O values for Southwestern Ghana (8.6‰ – 9.0‰) but higher than chlorite within vein chlorites to be between 1.5‰ 18 18 the quartz veins δ O values ranging from 12.7–17.4‰ to 2.0‰. The sample with the lowest δ O value from the same area (Koeberl et al., 1998). identified in the Abansuoso area is TMB-32 which The δD values of the surface samples are between is rich in chlorite. The influence of quartz and −43 and −117‰ (n = 10, mean = −66.3‰). The δD sericite on the mineralised samples which repre- values of the drillcore samples are slightly lower, sents the later stages of alteration in the area between −98 and −33‰ (n = 12, tends to have higher δ O values which may serve average = −50.92‰). to mitigate the effect of chlorite to an extent. This 18 18 The variation of δD and δ O values of the rocks is reflected in the variation in δ O values in the sampled across the two drill holes from different analysed samples (Table 5). Furthermore, chlorite depths studied is relatively constant and is similar alteration is minimal across the area, which sug- to those of the surface samples. The δ O values of gests a greater influence of isotopic exchange the rocks vary averagely by ±1‰ across the drill- between the mineralising fluid and the host rock. core. The δD values show more variation but do not This may account for the lower δ O values of the appear to follow a systematic pattern. This also mineralised samples. This assertion is corroborated δD‰ GEOLOGY, ECOLOGY, AND LANDSCAPES 15 Figure 13. (a) Distribution of δ O among the various rock types in the study area compared with average values from the Birimian (b) Distribution of δD among the various rock types in the study area compared with average values from the Birimian as reported by Kazapoe et al. (2021). by the depleted H O which is usually high in rocks values which correlates poorly. The erratic relation with chlorite alteration and tends to have an may be a result of the limited samples used in this inverse relationship with δ O as described by study. Although it is possible to identify a relationship Larson (1984). between δ O and δD and mineralisation, the erratic As shown in Figures 13(a-b), the general distribu- distribution among the sample renders it difficult to tion of δ O and δD among the rocks do not appear to contour. be controlled by lithology. The distribution of The δ O value of the mineralised samples ranges δ O among the schist, which has a higher number between 8.9 and 14.3‰ with a mean value of 13.08‰. of mineralised samples in this study, is relatively more Aside from the aforementioned sample TMB which depleted than the gneiss. The unmineralised quartz has a value of 8.9‰, most of the samples vary by only samples are relatively more enriched than the schist ± 1‰. This restricted range of values is similar to and the gneiss (Figure 13(a)). A similar situation is samples located in the Birimian of Ghana, the majority seen in the distribution of the δD concentration values of which fall between 12 and 16 ‰. These values are in the samples where the schist is the most relatively similar to those determined from the silicate rocks enriched, with the quartz being the most depleted from the Ashanti Gold Belt, which are between 13– (Figure 13(b)). The distribution suggests an influence 15‰ (Mumming et al, 1996; Hammond & Shimazaki, of the mineralising fluid. 1994). Similarly, these values are comparable to those determined for mineralised samples from other parts of the Birimian in West Africa which range between 4.4.2 Relationship between the stable isotopic 13–18‰ (Fouillac et al., 1993; Lambert-Smith et al., concentration in rocks and mineralisation of the 2016; Lawrence et al., 2013). Abansuoso area The δ O values are consistent across the various The relationship shown in Figures 14(a-b) do define rock types as well. The schist shows a range of values a clearer correlation between δD and gold values in from 13 to 14.3‰ (n = 6, mean = 13.75‰) and 12.7 comparison to the relationship between δ O and gold 16 R. W. KAZAPOE ET AL. Figure 14. (a) Isotopic compositions of hydrogen (δD) versus its corresponding gold values (g/t) (b) Isotopic compositions of oxygen (δ O) versus its corresponding gold values (g/t). to 14.4‰ (n = 5, mean = 14.4‰). The gneisses also across the samples. The composition of δ O is show values between 12 and 14.1‰. The mineralised affected by fluid temperature as well as the δ O of and altered samples taken from the schist averagely the fluid from which it precipitated, especially with differ with a value of ±1‰ with the mineralised quartz (Taylor & South, 1985). samples being relatively depleted. The δ O from Figure 15 shows the bulk of the sample points dis- the unmineralised quartz samples from the study play the typical inverse relationship between δ O and area falls between 12.7 and 17.4‰ which is slightly δD values. Additionally, the majority of both δ O and more enriched than the mineralised samples. This δD values correlate with the gold values positively and may be due to the δ O isotopic exchange between inversely, respectively. This suggests a relationship the hydrothermal fluid-bearing quartz and the coun- between gold and the isotopic values. try rock. The vein quartz which represents the gangue The distribution of δD values among the samples mineral of the mineralising fluid, as a result, becomes shows a clearer differentiation between the minera- subsequently relatively richer in δ O compared to lised, altered and unmineralised samples (Figure 16 the wall rock. (a)). The values of δD within the mineralised samples Unmineralised samples from Ghana, Mali, and range from −62 to −33‰ (n = 8, mean = −15.88‰). Senegal had consistently high values, ranging from This is similar to values determined in the mineralised 12 to 20‰. The unmineralised samples from the zone from Prestea, Ghana by Hammond and study area fall within the range of 12.70–17.40‰ Shimazaki (1994) (−65 to −29‰) and Oberthuer (n = 7, mean = 14.49‰). However, when compared et al. (1996; −60 to −43‰) from the Birimian of with results from unmineralised samples from the Ghana. study area, there is only a little relative enrichment. The altered samples from the study area show The unmineralised samples taken from the gneisses a relative average depletion of −8‰ as compared to show a consistently narrow range between 13.6‰- the mineralised samples which does not show much 14.4‰ (n = 4, mean = 14.15) which varies with 1‰ distinction between the mineralised and altered GEOLOGY, ECOLOGY, AND LANDSCAPES 17 Figure 15. Schematic representation of drillcore SPDD 079 showing variations of δ O and δD with Au across the zones. Figure 16. (a) Variation in δ O values between the 3 main zones (b) Variation in δD values between the 3 main zones. 18 R. W. KAZAPOE ET AL. 18 18 Figure 17. Plots of (a) δ O against as of the samples (b) δD against As of the samples (c) δ O against W of the samples (d) δD against W of the samples. samples (Figure 16(b)). However, the unmineralised influence and relationship with the mineralising fluid samples are comparatively more depleted and show in relation to the δ O value. a wide range of δD values between −117 to −48‰ (n = 7, mean = −75.86‰). Although the differentia - 5. Discussion tion between δD values from the mineralised and altered samples is not distinguishable, the unminera- The mineral assemblage of sericite, quartz, mus- lised samples are much more distinct from the miner- covite, biotite and hornblende is suggestive of alised samples and altered samples, varying averagely greenschist to lower amphibolite metamorphic around −30 and −22‰ for the mineralised and altered conditions. The rocks in the area have quartz samples respectively. grains that show undulose extinction with some The isotopic content of the rocks was plotted exhibiting chessboard extinction which indicates against As and W. This is because As and W were dislocation slip activity during high-temperature determined from the lithogeochemistry to strongly deformation, probably linked to high strain within correlate with the gold and is characteristic of the the shear zone (Figure 3(a-b)). The introduction suite of elements associated with gold in orogenic of gold possibly occurred during hydrothermal deposits such as the area. This was done to further alteration, with the gold deposited together with examine the relationship between the isotopic concen- the sulphide phases. The occurrence of arsenopyr- tration of the rocks with the mineralisation in the area. ite and pyrite, within the altered and mineralised A plot of δ O and δD values against As show a poor rocks of the Abansuoso area is widespread. The correlation between δ O and As whiles δD shows two principal sulphide minerals are similar to a more consistent relationship with As (Figure 17 those associated with the gold mineralisation in (a-b)). Similarly, the plot of δ O and δD values the Obuasi deposit except for the absence of mar- against W also shows a poor correlation between casite or pyrrhotite (Oberthuer et al., 1996; Osae δ O and W whiles δD shows a relatively consistent et al. 1999). No silver was found in association relationship with As (Figure 17(c-d)). This further with the gold grains by EDS-SEM. This suggests supports the assertion that the δD shows a clearer a uniqueness of the gold mineralisation of the GEOLOGY, ECOLOGY, AND LANDSCAPES 19 Abansuoso deposit which differs from similar mean = −16‰) which shows an average depletion deposits within the Birimian of Ghana. of −8‰ relative to the altered samples. The unmi- From the factor analysis, PCA1 has elemental asso- neralised samples range between −117 and −48‰ ciations of heavy elements characteristic of the mafic (n = 7, mean = −76‰) which are lower than the minerals associated with the gneisses in the study area. mineralised and altered samples between −30 and PCA 2 also has elemental groups which correlate with −22‰ respectively. The general distribution of the hematite alteration seen in the mineralised zone. The δ O among the mineralised rocks does not appear presence of Fe O and TiO is further corroborated by to be controlled by lithology. Although the 2 3 2 the SEM result for hematite alteration. δ O of the mineralised samples show The elemental associations in factor three (Au, As, a systematic variation which suggests an influence Mo and W) are characteristic of the elemental associa- of the mineralising fluid, the values are not differ - tions linked with the generally accepted supracrustal entiated from the altered samples or to a lesser metamorphic models of gold origin proposed to be extent the unmineralised samples. This makes the sourced from around the brittle-ductile transition δ O on its own not diagnostic of gold mineralisa- zone (~10 km) (Groves and Santosh, 2016). The pre- tion. The effect of mineralisation on the δD values sence of As associated with the sulphide minerals of the rocks is however relatively more pro- (arsenopyrite) associated with mineralisation in the nounced. Therefore, a combination of the two area further confirms this. This confirms the results isotopes with conventional lithogeochemistry will of the biplot and suggests that these elements are the be the most suitable tool for the exploration of pathfinder elements for gold mineralisation in the gold resources in the area. The lack of variation in area. The isocon diagrams (Figures 11(a-c)) suggest δ O suggests a rock-buffered system, with an that mass was commonly conserved during alteration influx of metamorphic fluid derived from rocks with a minor mass gain observed in the inner altera- with similar δ O comparable to many orogenic tion envelope (sericite). The result shows the outer gold deposits. The δD values seem to be lower in alteration envelope representing hematite alteration the mineralised samples suggesting an influence of recorded a slight mass gain in oxides of MgO, CaO the fluid. and Fe O (Figure 11(a)). There were mass gains in 2 3 Co, Pb, Ni and Ca with a slight loss of mass in W, Cu, 6. Conclusion Pr, Mo and K O. Results for the inner alteration zone, representing the sericite alteration and mineralised The study found a close association between Au and zone show that As, Zn, Ba, Au, Co and W were As, Mo and W which is characteristic of the element added. Oxides such as K O, P O , Fe O , Na O and associations linked with the generally accepted supra- 2 2 5 2 3 2 MnO were lost. Additionally, Mo, Cr, CaO and MgO crustal metamorphic models of gold origin proposed remained constant. to be sourced from around the brittle-ductile transi- The uniform distribution of δ O values within tion zone. It was also determined that mass was com- the shear zone and outside the shear zone also monly conserved during alteration with a minor mass indicates that the mineralising fluid invades the gain observed in the inner alteration envelope (sericite rock in a pervasive manner. Additionally, the dis- alteration). 18 18 tribution of δ O and δD across the depth of the The concentrations of δ O (8.9‰ to 17.4‰, drillholes suggests no significant variation of iso- n = 22, mean = +13.7‰) and δD (−117 to −33‰, topic values with depth. The lack of variation in n = 22, mean = −57.9‰) in the rocks are comparable δ O suggests a rock-buffered system, with an to those from other Birimian regions in Ghana and influx of metamorphic fluid derived from rocks West Africa. The lack of variation in δ O suggests of similar δ O like many orogenic gold deposits. a system buffered by rocks, with metamorphic fluid The δD values seem to be lower in the mineralised originating from rocks with identical δ O values. samples which suggest an influence of the fluid However, the effect of mineralisation on the δD values which probably had a δD value close to zero. of rocks is substantially more apparent than its effect 18 18 Recorded δ O values among the mineralised sam- on the δ O values of the rocks. ples (8.9–14.3‰, n = 8, mean = 13.08‰) fall in The differentiation between the isotopic values a similar range to samples from the other parts of of the unmineralised samples on one hand and the the Birimian in Ghana and West Africa. This is altered and mineralised samples on the other hand lower than in the unmineralised samples with an points to the utility of the methodology in targeting average of +1‰ (12.70–17.40‰ give n = 7, gold mineralisation in general. However, the lack of mean = 14.49‰). The δD however shows more a clear distinction between the altered and miner- variation between the mineralised samples and the alised samples shows that the methodology requires unmineralised samples. The mineralised samples an additional layer of data to be able to pick up the record values between −62 and −33‰ (n = 8, mineralised zones alone. The authors recommend 20 R. W. KAZAPOE ET AL. that a larger scale study using more samples on the Amponsah, P. O., Salvi, S., Béziat, D., Siebenaller, L., Baratoux, L., & Jessell, M. W. (2015). Geology and geo- scale of conventional lithogeochemical surveys be chemistry of the shear-hosted Julie gold deposit, NW carried out. This will further investigate the viabi- Ghana. Journal of African Earth Sciences, 112, 505–523. lity of this approach as a gold exploration https://doi.org/10.1016/j.jafrearsci.2015.06.013 technique. Amponsah, P. O., Salvi, S., Didier, B., Baratoux, L., Siebenaller, L., Jessell, M., Nude, P. M., & Gyawu, E. A. (2016). Multistage gold mineralisation in the Wa-Lawra greenstone belt, NW Ghana: The Bepkong deposit. Acknowledgments Journal of African Earth Sciences, 120, 220–237. https:// This research paper is part of the first author’s doi.org/10.1016/j.jafrearsci.2016.05.005 Ph.D. dissertation, and he would like to thank the African Borthwick, J., & Harmon, R. S. (1982). A note regarding Union and the Pan African University for awarding him CIF3 as an alternative to BrF5 for oxygen isotope analysis. a Ph.D. Scholarship and the University for Development Geochimica et cosmochimica acta, 46(9), 1665–1668. Studies for granting him study leave. Supercare Gold and Bouabdellah, M., & Slack, J. F. (Eds.). (2016). Mineral depos- Pelangio Exploration Company are greatly appreciated for its of North Africa. Springer. granting access to their concession and assisting with the Bowman, J. R., Parry, W. T., Kropp, W. P., & Kruer, S. A. fieldwork. The authors are also grateful to the anonymous (1987). Chemical and isotopic evolution of hydrothermal reviewers whose insightful comments, suggestions, and solutions at Bingham, Utah. Economic Geology, 82(2), reviews significantly improved the paper’s quality. 395–428. https://doi.org/10.2113/gsecongeo.82.2.395 Casey, W. H., & Taylor, B. E. (1982). Oxygen, hydrogen, and sulfur isotope geochemistry of a portion of the West Shasta Cu-Zn district, California. Economic Geology, 77(1), 38–49. Disclosure statement Chudasama, B., Porwal, A., Kreuzer, O. P., & Butera, K. (2016). Geology, geodynamics and orogenic gold pro- No potential conflict of interest was reported by the spectivity modelling of the Paleoproterozoic Kumasi author(s). Basin, Ghana, West Africa. Ore Geology Reviews, 78, 692–711. Dzigbodi-Adjimah, K. (1993). Geology and geochemical pat- Funding terns of the Birimian gold deposits, Ghana, West Africa. Journal of Geochemical Exploration, 47(1–3), 305–320. This research received partial funding from the African Dzikunoo, E. A., Kazapoe, R. W., & Agbetsoamedo, J. E. Union Commission through the Pan African University. (2021). An integrated structural and geophysical approach to defining the structures of part of the Nangodi greenstone belt, northeastern Ghana. Journal ORCID of African Earth Sciences, 180, 104238. https://doi.org/ 10.1016/j.jafrearsci.2021.104238 Raymond Webrah Kazapoe http://orcid.org/0000-0002- Emsbo, P., Hofstra, A. H., Lauha, E. A., Griffin, G. L., & 0307-7834 Hutchinson, R. W. (2003). Origin of high-grade gold ore, source of ore fluid components, and genesis of the Meikle and neighboring Carlin-type deposits, northern References Carlin trend, Nevada. Economic Geology, 98(6), 1069– Adjo, F. B., Nude, P. M., Adissin, L. G., Bolarinwa, A. T., & Fougerouse, D., Micklethwaite, S., Ulrich, S., Miller, J., Anagonou, B. (2021). Paleoweathering, Provenance and Godel, B., Adams, D. T., & McCuaig, T. C. (2017). tectonic setting of gold-bearing neoproterozoic rocks Evidence for two stages of mineralization in West from the Atacora Structural Unit, Pan-African Africa’s largest gold deposit: Obuasi, Ghana. Economic Dahomeyide belt, northwestern Bénin. Journal of Geology, 112(1), 3–22. African Earth Sciences, 183, 104327. Fouillac, A. M., Dommanget, A., & Milési, J. P. (1993). Agbenyezi, T. K., Foli, G., & Gawu, S. K. (2020). A carbon, oxygen, hydrogen and sulfur isotopic study of Geochemical Characteristics of Gold-Bearing Granitoids the gold mineralisation at Loulo, Mali. Chemical Geology, at Ayanfuri in the Kumasi Basin, Southwestern Ghana: 106(1–2), 47–62. https://doi.org/10.1016/0009-2541(93) Implications for the Orogenic Related Gold Systems. 90165-F Earth Sciences Malaysia (ESMY), 4(2), 127–134. Friedman, I., & O'Neil, J. R. (1977). Compilation of stable Allibone, A., Hayden, P., Cameron, G., & Duku, F. (2004). isotope fractionation factors of geochemical interest (Vol. Paleoproterozoic gold deposits hosted by albite-and 440). US Government Printing Office. carbonate-altered tonalite in the Chirano District, Galipp, K., Klemd, R., & Hirdes, W. (2003). Metamorphism Ghana, West Africa. Economic Geology, 99(3), 479–497. and geochemistry of the paleoproterozoic Birimian sefwi https://doi.org/10.2113/gsecongeo.99.3.479 volcanic belt (Ghana, West Africa). Geologisches Jahrbuch Allibone, A., Teasdale, J., Cameron, G., Etheridge, M., Uttley, D, 111, 151–191. http://dx.doi.org/10.1016/S0301-9268 P., Soboh, A., & Lamb, E. (2002). Timing and structural (99)00024-8 controls on gold mineralization at the Bogoso gold mine, Goldfarb, R. J., Baker, T., Dubé, B., Groves, D. I., Hart, C. J., Ghana, West Africa. Economic Geology, 97(5), 949–969. & Gosselin, P. (2005). Distribution, character, and genesis Almeida, C. (2009). Mineral paragenesis, geochemistry and of gold deposits in metamorphic terran. geochronology investigations of the Carlin-type gold Goldfarb, R. J., Groves, D. I., & Gardoll, S. (2001). Orogenic deposits at the Goldstrike property, northern Nevada: gold and geologic time: A global synthesis. Ore Geology Implications for ore genesis, igneous petrogenesis and Reviews, 18(1–2), 1–75. mineral exploration (Doctoral dissertation). GEOLOGY, ECOLOGY, AND LANDSCAPES 21 Grant, F. S. (1985). Aeromagnetics, geology and ore envir- (Uzbekistan): Implications for the age and sources of onments, I. Magnetite in igneous, sedimentary and meta- Au mineralization. Mineralium Deposita, 36(5), 379–392. morphic rocks: An overview. Geoexploration, 23(3), Kesse, G. O. (1985). The mineral and rock resources of 303–333. https://doi.org/10.1016/0016-7142(85)90001-8 Ghana. A.A. Balkema. Griffis, R. J., & Agezo, F. L., 2000, Mineral occurrence and Klemd, R., & Hirdes, W. (1997). Origin of an unusual fluid exploration potential of northern Ghana. Mineral composition in Early Proterozoic palaeoplacer and lode- Commission Report. 132–1133 (Accra, Ghana: Minerals gold deposits in Birimian greenstone terranes of West Commission). Africa. South African Journal of Geology, 100(4), 405–414. Griffis, J., Barning, K., Agezo, F. L., & Akosa, F., 2002, Gold Koeberl, C., Reimold, W., Blum, J., & Chamberlain, C. P. deposits of Ghana prepared on behalf of Ghana mineral (1998). Petrology and geochemistry of target rocks from commission. 432 (Accra, Ghana: Minerals Commission). the Bosumtwi impact structure, Ghana, and comparison Groves, D. I., & Phillips, G. N. (1987). The genesis and with Ivory Coast tektites. Geochimica et Cosmochimica tectonic control on Archaean gold deposits of the western Acta, 62(12), 2179–2196. https://doi.org/10.1016/S0016- Australian shield—a metamorphic replacement model. 7037(98)00137-9 Ore Geology Reviews, 2(4), 287–322. https://doi.org/10. Lambert-Smith, J. S., Lawrence, D. M., Vargas, C. A., 1016/0169-1368(87)90009-6 Boyce, A. J., Treloar, P. J., & Herbert, S. (2016). The Groves, D. I., & Santosh, M. (2016). The giant Jiaodong gold Gounkoto Au deposit, West Africa: Constraints on ore province: The key to a unified model for orogenic gold genesis and volatile sources from petrological, fluid inclu- deposits?. Geoscience Frontiers, 7(3), 409–417. sion and stable isotope data. Ore Geology Reviews, 78, 606–622. https://doi.org/10.1016/j.oregeorev.2015.10.025 Hammond, N. Q., & Shimazaki, H. (1994). Geology and Larson, S. Å., & Tullborg, E. L. (1984). Stable isotopes of geochemical aspects of ore formation at the Prestea fissure-filling calcite from Finnsjön, Uppland, Sweden. mesothermal vein gold deposit in the Birimian system Lithos, 17, 117–125. of Ghana. International Geology Review, 36(8), 715–731. Lawrence, D. M., Treloar, P. J., Rankin, A. H., Boyce, A., & https://doi.org/10.1080/00206819409465484 Harbidge, P. (2013). A fluid inclusion and stable isotope Harris, C., Faure, K., Diamond, R. E., & Scheepers, R. study at the Loulo mining district, Mali, West Africa: (1997). Oxygen and hydrogen isotope geochemistry of Implications for multifluid sources in the generation of S-and I-type granitoids: The Cape Granite suite, South orogenic gold deposits. Economic Geology, 108(2), Africa. Chemical Geology, 143(1–2), 95–114. 229–257. https://doi.org/10.2113/econgeo.108.2.229 Harris, C., Oom, B. M., & Diamond, R. E. (1999). A pre- Leube, A., Hirdes, W., Mauer, R., & Kesse, G. O. (1990). The liminary investigation of the oxygen and hydrogen iso- early Proterozoic Birimian Supergroup of Ghana and tope hydrology of the greater Cape Town area and an some aspects of its associated gold mineralization. assessment of the potential for using stable isotopes as Precambrian Research, 46(1–2), 139–165. tracers. Water S. A, 25(1), 15–24. Lopez-Moro, F. J. (2012). EASYGRESGRANT—A Harris, C., & Vogeli, J. (2010). Oxygen isotope composition of Microsoft Excel spreadsheet to quantify volume changes garnet in the Peninsula Granite, Cape Granite Suite, South and to perform mass-balance modeling in metasomatic Africa: Constraints on melting and emplacement mechan- systems. Computers & Geosciences, 39, 191. isms. South African Journal of Geology, 113(4), 401–412. Mumin, A. H., Fleet, M. E., & Longstaffe, F. J. (1996). Hart, C. J., McCoy, D. T., Goldfarb, R. J., Smith, M., Roberts, Evolution of hydrothermal fluids in the Ashanti gold P., Hulstein, R., & Bundtzen, T. K. (2002). Geology, belt, Ghana; stable isotope geochemistry of carbonates, exploration, and discovery in the Tintina Gold Province, graphite, and quartz. Economic Geology, 91(1), 135–148. Alaska and Yukon. Special Publication-Society of https://doi.org/10.2113/gsecongeo.91.1.135 Economic Geologists, 9, 241–274. Mumm, A. S., Oberthür, T., Vetter, U., & Blenkinsop, T. G. Hirdes, W., Leube, A., Für Geowissenschaften, B., & , R. (1997). High CO2 content of fluid inclusions in gold miner- (1989). On gold mineralization of the Proterozoic Birimian alisations in the Ashanti Belt, Ghana: A new category of ore supergroup in Ghana/West Africa. BGR. forming fluids?. Mineralium Deposita, 32(2), 107–118. Hofstra, A. H. (1994). Reconnaissance Stream-Sediment Oberthuer, T., Mumm, A. S., Vetter, U., Simon, K., & Geochemistry of Twentynine Palms Marine Corps Base, Amanor, J. A. (1996). Gold mineralisation in the California, 95(13). US Geological Survey. Ashanti Belt of Ghana; genetic constraints of the stable Hrdy, F., & Kyser, T. K. (1995). Origin, timing, and fluid isotope geochemistry. Economic Geology, 91(2), 289–301. characteristics of an auriferous event; the Proterozoic https://doi.org/10.2113/gsecongeo.91.2.289 Jasper lode gold deposit, Saskatchewan, Canada. Osae, S., Kase, K., & Yamamoto, M. (1999). Ore Mineralogy Economic Geology, 90(7), 1918–1933. https://doi.org/10. and Mineral Chemistry of the Ashanti Gold Deposit at 2113/gsecongeo.90.7.1918 Obuasi, Ghana. Resource Geology, 49(1), 1–11. Jessell, M. W., Amponsah, P. O., Baratoux, L., Perrouty, S., Jessell, M. W., Bourassa, Y., Miller, J., Apau, D., Asiedu, D. K., Loh, G. K., & Ganne, J. (2012). Crustal- Siebenaller, L., Salvi, S., Baratoux, L., Aillères, L., scale transcurrent shearing in the paleoproterozoic Béziat, D., & Salvi, S. (2015). The Wassa deposit: A poly- Sefwi-Sunyani-Comoe region, West Africa. deformed orogenic gold system in southwest Ghana– Precambrian Research, 212, 155–168. https://doi.org/ implications for regional exploration. Journal of African 10.1016/j.precamres.2012.04.015 Earth Sciences, 112, 536–547. https://doi.org/10.1016/j. Kazapoe, R.W., Olugbenga, O., Arhin, E. and Olisa, O. jafrearsci.2015.03.003 (2021). Isotope geochemistry as a tool in the exploration Querubin, C. L., & Yumul, G. P. (2005). Mass change calcu- of gold occurrences in Ghana: a review. Arabian Journal lations as applied to the search for volcanogenic massive of Geosciences, 14(19), pp.1–15. sulfide deposits: an example from Malusok, Siocon, Kempe, U., Belyatsky, B., Krymsky, R., Kremenetsky, A., & Zamboanga del Norte, Mindanao, Philippines. Resource Ivanov, P. (2001). Sm–nd and Sr isotope systematics of Geology, Vol. 55(2), pp. 61–72. scheelite from the giant Au (–W) deposit Muruntau 22 R. W. KAZAPOE ET AL. Senyah, G. A., Dampare, S. B., & Asiedu, D. K. (2016). Vennemann, T. W., & Smith, H. S. (1990). The rate and Geochemistry and tectonic setting of the Paleoproterozoic temperature of reaction of CIF3 with silicate minerals, metavolcanic rocks from the Chirano Gold District, Sefwi and their relevance to oxygen isotope analysis. belt, Ghana. Journal of African Earth Sciences, 122, 32–46. Chemical Geology: Isotope Geoscience Section, 86(1), https://doi.org/10.1016/j.jafrearsci.2015.07.022 83–88. Smith, A. J., Henry, G., & Frost-Killian, S. (2016). A review Wang, C., Deng, J., Bagas, L., He, X., & Zhang, J. (2021). of the Birimian supergroup-and Tarkwaian group-hosted Origin and classification of the Late Triassic Huaishuping gold deposits of Ghana. Episodes, 39(2), 177–197. https:// gold deposit in the eastern part of the Qinling-Dabie doi.org/10.18814/epiiugs/2016/v39i2/95775 Orogen, China: Implications for gold metallogeny. Taylor, B. E., & South, B. C. (1985). Regional stable isotope Mineralium Deposita, 56(4), 725–742. systematics of hydrothermal alteration and massive sul- Yao, Y., Murphy, P. J., & Robb, L. J. (2001). Fluid fide deposition in the West Shasta District, California. characteristics of granitoid-hosted gold deposits in Economic Geology, 80(8), 2149–2163. https://doi.org/10. the Birimian Terrane of Ghana: A fluid inclusion 2113/gsecongeo.80.8.2149 microthermometric and Raman spectroscopic study. Tsiboah, T., & Grant, T., 2009, Application of geophysics to Economic Geology, 96(7), 1611–1643. https://doi.org/ gold exploration in Ghana: Examples from Newmont. 10.2113/gsecongeo.96.7.1611 ASEG Extended abstracts 2009: 20th Geophysical Yigit, O., & Hofstra, A. H. (2003). Lithogeochemistry of Conference, pp.1–5, ASEG 20th Geophysical Conference Carlin-type gold mineralization in the Gold Bar district, and Exhibition. Adelaide (CSIRO Publishing). https:// Battle Mountain-Eureka trend, Nevada. Ore Geology doi.org/10.1071/ASEG2009ab089 . Reviews, 22(3–4), 201–224. Vennemann, T. W., & O'Neil, J. R. (1993). A simple and Yigit, O., & Hofstra, A. H. (2003). Lithogeochemistry of inexpensive method of hydrogen isotope and water ana- Carlin-type gold mineralization in the Gold Bar district, lyses of minerals and rocks based on zinc reagent. Battle Mountain-Eureka trend, Nevada. Ore Geology Chemical Geology, 103(1–4), 227–234. Reviews, 22(3–4), 201–224. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Geology Ecology and Landscapes Taylor & Francis

Geology and Isotope Systematics of Gold Deposits in the Abansuoso Area of the Sefwi Belt, Southwestern Ghana

Download PDF
22 pages

Loading next page...
 
/lp/taylor-francis/geology-and-isotope-systematics-of-gold-deposits-in-the-abansuoso-area-uA7XdmpxK9

References (37)

Publisher
Taylor & Francis
Copyright
© 2022 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group on behalf of the International Water, Air & Soil Conservation Society(INWASCON).
ISSN
2474-9508
DOI
10.1080/24749508.2022.2142100
Publisher site
See Article on Publisher Site

Abstract

GEOLOGY, ECOLOGY, AND LANDSCAPES INWASCON https://doi.org/10.1080/24749508.2022.2142100 RESEARCH ARTICLE Geology and Isotope Systematics of Gold Deposits in the Abansuoso Area of the Sefwi Belt, Southwestern Ghana a,c b,c d e f Raymond Webrah Kazapoe , Olugbenga Okunlola , Emmanuel Arhin , Olusegun Olisa , Chris Harris , g h i Daniel Kwayisi , Sam Torkorno and Ebenezer Ebo Yahans Amuah a b Department of Geological Engineering, School of Engineering, University for Development Studies, Nyankpala, Ghana; Department of Geology, University of Ibadan, Ibadan, Nigeria; Pan African University Life and Earth Sciences Institute (PAULESI), University of Ibadan, Ibadan, Nigeria; Department of Geological Sciences, School of Geoscience, University of Energy and Natural Resources, Sunyani, Ghana; e f Department of Earth Science, Olabisi Onabanjo University, Ago-Iwoye, Nigeria; Department of Geological Sciences, University of Cape g h Town, Rondebosch, South Africa; Department of Earth Science, University of Ghana, Accra, Ghana; Exploration Department, Pelangio Gold Ltd, Ghana; Environmental Science Department, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana ABSTRACT ARTICLE HISTORY Received 6 July 2022 This study utilised lithogeochemistry and stable isotope geochemistry to assess the gold Revised 17 October 2022 mineralisation in the host rocks of the Abansuoso area. It also studied the isotopic signature Accepted 26 October 2022 of hydrothermally altered and unaltered rocks as a tool for the preliminary prospecting of gold in Ghana’s Birimian belt. The research identified a close association between Au and As, Mo, KEYWORDS and W, which is characteristic of orogenic gold deposits. The mineralised and unmineralised Stable isotopes; mineral rocks show similar content of major and trace elements which suggests that the mineralisation exploration; geochemistry; did not significantly affect the primary composition of the mineralised rocks. The concentra- gold; Birimian; Ghana tions of δ O (8.9‰ to 17.4‰, n = 22, mean = +13.7‰) and δD (−117 to −33‰, n = 22, mean = −57.9‰) in the rocks are comparable to those from other Birimian regions in Ghana and West Africa. The lack of variation in δ O suggests a system buffered by rocks, with metamorphic fluid originating from rocks with identical δ O values. However, the effect of mineralisation on the δD values of rocks is substantially more apparent than its effect on the δ O values of the rocks. 1. Introduction et al., 2017). According to Chudasama et al. (2016), the Prior to the late 1980s, when gold exploration was Abore and Dynamite Hill gold deposits are also extre- rekindled in the Birimian metallogenic region of mely important. Ghana, two key types of gold deposits were studied Researchers continue to debate the distinctions in depth; the first was auriferous quartz-pebble con- between orogenic gold-bearing granitoids and intru- glomerates. The second was lode-quartz veins and sion-related gold deposits due to the similarities in disseminated sulphides in shear zones (Kesse, 1985; their wall-rock alteration, elemental relationships, Hirdes and Leube, 1989; Leube et al., 1990; Schmidt structural controls, and ore fluids (Goldfarb et al., Mumm et al., 1997; Smith et al., 2016). Extensive 2001; Groves & Phillips, 1987). Consequently, research in Ghana’s Birimian terrain has led to the a number of granitoid-hosted gold deposits, including discovery of a third and more recent type of gold the True North deposit (Canada) and the Muruntau mineralisation (Amponsah et al., 2016; Yao et al., gold deposits (Uzbekistan), have been categorized as 2001). This deposit type is related to granitic rocks. both orogenic and intrusion-related deposits (Kempe A good example is the Ayanfuri deposit located within et al., 2001; Hart et al., 2002). It is therefore essential to the Ashanti belt, which contains multiple gold occur- understand each model, as each classification, i.e., rences within intermediate to felsic granitoids. intrusion-related or orogeny-related, may necessitate The Birimian gold deposits in granitoids are emer- distinct exploration strategies (Goldfarb et al., 2005). ging as increasingly promising exploration targets, This study employs lithogeochemistry and stable with the aim of increasing the supply of gold ore and isotope geochemistry to evaluate the gold-bearing boosting economic output (Bouabdellah and Slack, granitoid and metasedimentary rocks that have 2016). The gold deposits in the western portion of received relatively less regional and local research Ghana’s Sefwi and Ashanti Belts (Chirano, Nhyiaso, attention (Yao et al., 2001; Agbenyezi et al., 2020). In and Ayankyerim) are prime examples of gold deposits addition, the nature of hydrothermal fluid migration hosted by granitoids (Allibone et al., 2004; Fougerouse and interaction with the country rocks in the study CONTACT Raymond Webrah Kazapoe rkazapoe@yahoo.com Department of Geological Engineering, School of Engineering, University for Development Studies, Nyankpala, Ghana This article has been corrected with minor changes. These changes do not impact the academic content of the article. © 2022 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group on behalf of the International Water, Air & Soil Conservation Society(INWASCON). This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 2 R. W. KAZAPOE ET AL. area is evaluated. This study also aims to assess the (Dzikunoo et al., 2021; Perrouty et al., 2015). stable isotopic signature of the rocks in order to deter- The second orogenic event or the Eburnean II is mine the difference in isotopic composition between characterised by regional NW-SE extension and hydrothermally altered and unaltered rocks. This will formation of the Birimian metasedimentary basins serve as a preliminary investigation tool for identifying between 2150 and 2116 Ma (Allibone et al., 2004). gold mineralization in the Paleoproterozoic rocks of The study area is underlain by the Sefwi Ghana and, by extension, West Africa. Greenstone Belt which is one of the six known Birimian Greenstone Gold Belts in Ghana (Figure 1). These belts are known to host different 2. Geological setting types of gold mineralisation (Perrouty et al., 2015). Out of the six Greenstone Gold Belts, two Two discrete orogenic events proposed by have the most productive mines operating on Allibone et al. (2002) have been recognised in them whilst the others have advanced exploration Ghana. The earliest of these is termed the and minimal mining works. The two most prolific Eburnean I and is believed to be associated with belts are the Sefwi and Ashanti Greenstone Gold the eruption of the Birimian volcanic rocks and Belts with the Sefwi Belt lying Northwest of the the intrusion of granites, followed by regional Ashanti Belt. Similar to the other greenstone belts metamorphism between 2200 and 2150 Ma Figure 1. Geological map of the study area (modified after Allibone et al., 2004). GEOLOGY, ECOLOGY, AND LANDSCAPES 3 in Ghana, the Sefwi Belt is composed of metavol- a thermogravimetric analyser (TGA). Forty-one (41) canic, metasedimentary and volcano-sedimentary selected minor and trace elements were analysed. rocks which have been intruded by granitoid and Thirty-one (31) minor and trace elements were ana- mafic units (Galipp et al., 2003; Senyah et al., lysed using lithium borate fusion ICP-MS. 2016). It is bordered to the west by the Sunyani Concentrations of the other ten (10) elements (Ag, Basin and to the east by the Kumasi Basin (Griffis Co, Ni, Zn, As, Cu, Pb, Cd, Mo and Sc) were obtained & Agezo, 2000). Rocks found in the Sefwi Belt by ICP-AES from a separate four-acid digestion fol- include metamorphosed tholeiitic lavas containing lowing standard procedure. The detection limit of the basalts, andesites, dacites, rhyolites, and volcano- elements was 0.05 mass percent. Precision and accu- clastic units comprising pyroclastics and epiclas- racy were better than 5% and 3% respectively. tics as well as minor metasedimentary rocks and The concentrations of major oxides, minor ele- syn-volcanic tonalitic to granodioritic granitoids ments, and trace elements were measured to classify (Jessell et al., 2012). the lithogeochemical signatures of the rocks found in the study area, which represent varying degrees of alteration. Various statistical techniques, including correlation matrix, binary diagrams, biplots, and 3. Materials and methods R-mode factor analysis, were utilised to present and evaluate the data. This allowed for the identification of Fieldwork consisted of lithological and structural log- metal associations and the delineation of potential ging and geological mapping. The drill cores were mineralising and/or geological signatures. In addition, logged in order to provide structural, lithological, scatter plots were created to examine the linearity of alteration, and mineralisation data on the deposit. the dataset, a standard assumption made when calcu- After logging, samples from two (2) drill cores lating correlation coefficients. The purpose of the fac- (SFDD 79 and SFDD 81) were extracted. The samples tor analysis was to clarify the correlation structure of were chosen based on their gold content to accurately the variables based on a smaller number of factors. To represent the altered zone, mineralized zone, and enable data computation, analysed samples with unmineralised zone in order to characterise the miner- values below their respective analytical detection lim- alisation of the area. Samples with gold concentrations its were assigned a value equal to half the analytical greater than 0.3 g/t are considered mineralised, detection limit. Before performing population statis- whereas samples with gold concentrations less than tics on the data, the analysed data were evaluated to 0.3 g/t are considered unmineralised. This grade clas- eliminate elements with insufficient variance and sification is based on the typical gold grades found in duplication. The data was log-transformed (base 10) the Birimian terrain of Ghana and has been deemed to decrease its wide distribution. suitable for academic research (Allibone et al., 2004; The oxygen and hydrogen isotope ratios of twenty- Griffis et al., 2002; Mumin et al., 1996; Oberthuer et al., two (22) rock samples taken from the mineralised (8), 1996). Additionally, surface rock samples were also hydrothermally altered (7) and unmineralised zones collected. Thin section preparation was done at the (7) were analysed in the Stable Isotope Laboratory of Department of Earth Science, University of Ghana of the Department of Geological Sciences, University of both oriented and un-oriented rock specimens to per- Cape Town, South Africa. Isotope ratios were mea- mit the identification of constituent minerals, and sured using a Thermo DeltaXP mass spectrometer. deformation study of textures and fabric elements Aliquots of ~10 mg of the rock powders were dried using an optical microscope. Scanning electron micro- overnight at 50°C and then within a vacuum in nickel scopy (SEM) images were taken using the Zeiss EVO reaction vessels. The powders were then reacted with MA 15 SEM. All images were taken using the back- 30 kPa of CIF for ~4 hours to extract the oxygen from scatter detector (BSD). A Bruker Quatax EDS system silicates (Borthwick and Harmon, 1982). The with XFlash Detector 610 M was employed as the extracted oxygen was then converted to CO by pas- energy dispersive X-ray detector (EDS). SEM micro- sing it over a high-temperature platinised carbon rod. graphs of the samples were collected using an accel- Full analytical details are described by Vennemann erating voltage of 18–20 kV with a working distance and Smith (1990) and Harris and Vogeli (2010). that ranged between 10 mm to 14 mm. Unknowns were run with duplicates of the internal Twenty-two (22) rock samples were sent to the ALS quartz standard (MQ), which was used to calibrate the geochemical laboratory in Vancouver for whole-rock, raw data to the SMOW (Standard Mean Ocean Water) major, minor and trace elements. The rocks were scale, using a δ O value of 10.1 for MQ (calibrated taken from the mineralised zone (8 samples), and the against NBS-28). The results are reported in standard unmineralised zone (14 samples). Whole-rock major δ-notation, where δ = (R /R – 1) × 1000, element analysis was carried out using ICP-AES from sample standard 18 16 R is O/ O in the sample, and R is the ALS laboratory in Australia. Loss on Ignition sample standard 18 16 O/ O relative to SMOW. The analytical error is (LOI) at 1000°C was determined using 4 R. W. KAZAPOE ET AL. estimated as ± 0.2 ‰ (2 sigmas), based on long-term alongside the unknowns. The measured water content repeated analysis of MQ. of Serina kaolinite gave 12.42 wt% (± 0.5 wt% Hydrogen isotopes and water contents of the same 2σ, n = 3). powdered separates were determined using the method of Vennemann and O’Neil (1993). Samples 4. Results were melted in quartz glass tubes using a propane torch. The raw data was normalised to SMOW, and 4.1 Local geology corrected for compression, using the water standards RMW (δD = – 131.4‰) and CTMP2010 (δD = – The petrographic investigations reveal the two (2) 7.4‰) with the in-house Serina kaolinite standard main types of rocks in the area as gneisses and analysed with the unknowns (Serina bulk kaolinite, schist with the presence of quartz veins. δD = – 57‰, H O = 12.4 wt%; Harris et al., 1999). The measured δD values of the standard gave an 4.1.1 Gneisses average of – 59.5 ± 3.6‰ (2σ, n = 3) after corrections The gneisses are of two types based on mineralogical to the raw data. All data were adjusted to the accepted composition, the biotite and hornblende-biotite Serina kaolinite value. Water abundances (as H O ) gneisses. The biotite gneisses outcrop in the north- were measured from the voltage on the mass 2 collec- western part of the study area. They are medium to tors of the mass spectrometer. This was calibrated coarse-grained and weakly to strongly foliated (Figure 2 against measured volumes of standard waters analysed (a)). The dominant minerals are quartz, plagioclase, Figure 2. (a) xpl photomicrograph of biotite gneiss showing microcline (Mc), biotite (Bio) and quartz (Qtz) (b) xpl photomicrograph of biotite gneiss showing garnet (Grt) and hornblende (Hbl) (c) xpl photomicrograph of schist showing sericite (Ser) and calcite (Cal) alteration (d) xpl photomicrograph of sericite schist showing calcite (Cal) and quartz (Qtz) (e) xpl photomicrograph of sericite schist showing sericite alteration (Ser) and calcite alteration (f)) xpl photomicrograph of schist showing muscovite (Ms). GEOLOGY, ECOLOGY, AND LANDSCAPES 5 Figure 3. Photomicrographs of quartz showing (a) strong elongation and pressure shadow and (b) chess board extinction. microcline, and biotite (Figures 2(a)). Under thin sec- The sericite-quartz schist is medium-grained and tion, cubic to slightly elongated opaque minerals sus- composed of sericite and quartz with a trace amount pected to be sulphide can be observed. The foliation and of muscovite (Figure 2(e)). Fine to medium-grained mineral lineation are defined by the preferred orienta- texture and strong foliation characterise the musco- tion of biotite and elongated opaque minerals. The vite-quartz schist (Figure 2(f)). This rock is domi- quartz minerals exhibit an anhedral shape with undu- nantly composed of quartz and muscovite together lose extinction. The quartz minerals also appear recrys- with minor feldspar. Rare occurrence of disseminated tallised and occur as either coarse crystals or medium- cubic sulphides can be observed in the rocks. The rock grained and elongated aggregates. Plagioclase and is cut by massive quartz veins. microcline are subhedral to anhedral and altered into sericite. The biotite minerals show very weak hydro- 4.1.3 Quartz vein thermal alterations mostly into chlorite. Quartz veins are very common in the study area, The hornblende-biotite gneiss on the other hand is usually filling fractures. The veins are associated with coarse-grained and foliated. The rock is composition- all the rock types but are more common in the biotite ally banded with felsic bands alternating with mafic gneiss. Both syntaxial and antitaxial veins are observed bands. The dominant minerals are hornblende, biotite, in the study area. The veins are composed mainly of garnet, plagioclase, quartz, and microcline (Figure 2 quartz and carbonates and have thicknesses varying (b)) with the mafic bands comprised of the first three between 0.1 cm and 20 cm. Quartz in the veins is minerals whereas the latter three comprise the felsic coarse-grained and subrounded to subangular. bands. The garnet and microcline minerals are very Although this quartz is deformed, they show only coarse and poikilitic with inclusions of quartz and weak recrystallisation without evidence of grain occasional biotite (Figure 2(b)). Associated with the boundary migration. The quartz grains show undulose garnet and often occurring within the foliation are extinction with some exhibiting chessboard extinction subhedral to cubic opaque minerals suspected to be (Figure 3(a-b)). In some samples, the quartz is also sulphides. Hornblende has partially altered to epidote elongated. whereas the feldspars and biotite are partially altered to sericite and chlorite, respectively. In the hand sam- 4.2 Gold mineralisation ple, the biotite appears to dominate in composition with quartz also prominent. 4.2.1 Sulphide characterisation The principal ore minerals are arsenopyrite and 4.1.2 Schists a small amount of pyrite, both of which are found in The schists are of two types: the carbonate-sericite the mineralised zones of the altered host rock. The schist and sericite-quartz schist. The carbonate- mineralised rocks contain an average of 20% sulphides sericite schist is medium-grained and weakly foliated by volume, with arsenopyrite accounting for 50%, (Figure 2(c)). It is composed of quartz, plagioclase, pyrite for 30%, and galena for the remaining 20%. muscovite and calcite. The muscovite is strongly ser- Both primary sulphide ore minerals form in sizes icitised. The calcite occurs as very coarse crystals in the ranging from 1 μm to approximately 8 μm. Gold veins and also as impregnations in the rock (Figure 2 mineralisation occurs as micron-sized gold inclusions (c-d)). Quartz occurs as recrystallised grains and may of approximately 2 μm (Figure 4(a-b)) and larger gold be associated with the calcite or as impregnations, and grains within fractures of euhedral arsenopyrite/pyr- exhibits undulose extinction. ite, and predominantly as free gold associated with 6 R. W. KAZAPOE ET AL. Figure 4. Scanning Electron Microscopy (SEM) images of (a) sulphide phases associated with chlorite schist showing micron-sized gold (2 μm). (b) sulphide phases associated with chlorite schist showing inclusions of galena (Gn), pyrite (Py) and arsenopyrite (Apy). (c) ore zone showing gold associated with sericite (d) ore zone showing relationship of Au with dolomite (Dol). Table 1. A paragenetic scheme in the Abansuoso deposit showing the relative abundance of minerals during the regional metamorphic stage, pre-ore-forming stage and ore-forming stage. Mineral Regional Metamorphism Stage One Stage Two Plagioclase Quartz Muscovite Biotite Hornblende Garnet Chlorite Hematite Sericite Epidote Dolomite Pyrite Arsenopyrite Galena Gold sericite alteration in the host lithologies (Figure 4 infiltration, chlorite, epidote, hematite, sericite, (c-d)). Gold associated with sericite alteration is also and opaque mineralisation. The ore-forming pro- accompanied by dolomite (Figure 4(d)). A minor cess was a later event that appears to be largely amount of galena is also found as small inclusions hydrothermal alteration, but regional metamorph- within the arsenopyrite (Figure 4(b)). ism might have contributed to some fluid genera- tion or P-T-X within the greenschist zone 4.2.2 Mineral paragenesis of gold deposit The mineral assemblage of sericite, quartz, mus- The paragenetic sequence in the Abansuoso deposit covite, biotite and hornblende is suggestive of is summarised in Table 1. The silicate minerals greenschist to lower amphibolite metamorphic con- such as quartz, feldspar, micas, hornblende, pyrox- ditions. Although garnet was observed in some of ene as well as garnet formed earlier in the para- the gneissic rocks, they have been resorbed by genetic sequence, and are linked to regional quartz and thus may be pre-kinematic grains. The metamorphism. This was followed by carbonate alteration pattern within the area is chlorite- GEOLOGY, ECOLOGY, AND LANDSCAPES 7 4.2.3 Mineralisation The concession lies within a ductile sinistral shear zone with a brittle component. The mineralisation occurs in biotite gneiss of intermediate composition and schist of varying composition both of which have been affected by various stages of metamorphism and deformation. The sequence of alteration invades the rocks via stringers along weak zones caused by the shearing. The alteration is predominantly quartz- sericite, and pyrite with hematite at the fringes where low-grade gold is reported (Figure 6). The main mineralised zones are characterised by sericite altera- tion coupled with fine-grain pyrite and associated with dolomite. The preliminary sequence of alteration has Figure 5. SEM image showing elongated euhedral arsenopyr- been deduced as chlorite-carbonate-hematite-sericite- ite aligned with a fracture. chlorite-quartz/pyrite with minor albite. carbonate-hematite-sericite-epidote-sulphide. Since 4.3 Lithogeochemistry the sulphide (arsenopyrite) is occasionally elon- The major and trace element concentrations of the gated and follows the foliation (Figure 5), they schist and gneisses are presented in Tables 2 and 3 may have formed during the same deformational respectively. Histogram plots for the unmineralised event that developed the foliations, thus, syn- and mineralised rocks (Figures 7(a-d)) from the study kinematic. Figure 6. Schematic representation of drillcore SPDD 079. 8 R. W. KAZAPOE ET AL. Table 2. Stable isotopes, major and trace elements composition of the schist. 020/79 009/79 013/79 019/79 011//79 012/79 014/79 001/81 OS/A 001/OS 008/PK Schist Schist Schist Schist Schist Schist Schist Schist Schist Schist Schist δ O‰ 14.3 13.3 13 13.9 13.9 14.1 12.7 13.1 14.4 14.4 14.1 δD‰ −45 −42 −50 −62 −34 −33 −49 −47 −60 −51 −51 SiO 72 74.9 65.2 74.4 64.9 74.4 67.1 60.4 76.9 75.7 79.5 Al O 10.05 9.38 12.1 10.65 10.6 10.7 11.75 9.6 11.15 10.55 12.2 2 3 Fe O 7.91 4.87 6.92 5.92 6.68 4.49 7.54 7.49 2.39 2.49 1.4 2 3 CaO 0.54 1.66 2.19 0.56 3.81 1.27 2.29 4.83 1.31 1.56 0.03 MgO 0.24 0.61 0.57 0.2 0.99 0.4 0.92 2.33 0.44 0.5 0.2 Na O 0.31 2.03 3.89 4.13 1.73 3.55 4.04 3.96 3.06 2.85 0.12 K O 3.06 1.93 1.74 1.14 2.4 1.49 2.18 0.99 1.91 1.86 3.79 TiO 0.53 0.45 0.64 0.58 0.66 0.51 0.68 0.67 0.25 0.22 0.26 MnO 0.02 0.13 0.16 0.06 0.23 0.12 0.15 0.13 0.05 0.06 0.01 SiO /Al O 7.16 7.99 5.39 6.99 6.12 6.95 5.71 6.29 6.90 7.18 6.52 2 2 3 Na O/K O 0.10 1.05 2.24 3.62 0.72 2.38 1.85 4.00 1.60 1.53 0.03 2 2 LOI 5.28 4.6 6.26 2.29 8.19 3.82 3.83 8.96 2.75 2.93 2.07 Total 100.19 100.76 100.06 100.21 100.58 101.05 100.8 99.5 100.4 98.9 99.69 Au 1.66 2.1 1.04 2 1.4 0.9 0.15 0.22 0.11 0.01 0.11 Ba 732 303 527 216 1430 525 796 158.5 583 591 748 Ce 52.5 38.9 56.7 50.5 57.9 49.7 50.3 12.6 53.8 47.4 56.5 Cr 140 190 170 180 160 190 180 170 200 140 140 Cs 1.64 0.9 0.97 0.57 1.42 0.91 1.11 0.96 1.13 1.17 1.1 Dy 7.12 7.21 10.8 8.43 9.36 6.97 8.15 2.25 6.49 5.71 10.25 Er 4.28 4.48 6.52 5.48 6.11 4.34 5.06 1.33 4.3 3.67 6.66 Eu 1.62 1.52 2.31 1.75 2.04 1.77 1.82 0.61 1.08 0.91 1.65 Ga 15.3 16 17.4 11.7 12.7 15.7 19.9 9.8 14.4 13.1 19.6 Gd 7.82 7.25 10.5 8.47 9.27 7.79 8.54 2.38 6.86 5.96 9.54 Hf 5.9 4.9 7 6.6 7.4 5.6 6.1 1.5 6.8 6.2 8.5 Ho 1.42 1.45 2.19 1.77 1.88 1.41 1.6 0.44 1.27 1.14 2.1 La 24.4 18.9 27.1 23.8 27.1 24.2 23.6 5.8 25.8 23.5 28.5 Lu 0.65 0.62 0.87 0.74 0.96 0.65 0.71 0.21 0.69 0.57 0.94 Nb 5.9 5.3 7.3 5.7 7.6 5.7 6.3 1.7 7.4 5.1 7.7 Nd 30.6 23.5 33.4 28.2 33.3 28.6 29.8 7.7 28.1 24.7 32.7 Pr 7.53 5.69 8.3 7.2 8.32 7.25 7.27 1.91 7.38 6.54 8.2 Rb 67.6 41.4 36.6 24.9 48.9 34.7 45 26.1 42.8 42.1 79.5 Sm 7.51 6.6 9.1 7.41 8.72 7.52 7.7 2.21 6.61 5.82 8.38 Sn 3 1 2 1 3 1 1 1 2 1 2 Sr 38.3 93.8 161.5 82.1 266 92.5 208 225 102.5 117 9.1 Ta 0.4 0.3 0.4 0.4 0.5 0.3 0.4 0.1 0.4 0.4 0.5 Tb 1.2 1.17 1.68 1.34 1.49 1.24 1.32 0.38 1.07 0.9 1.57 Th 3.2 2.5 3.42 3.13 3.72 3.11 2.99 0.69 3.99 3.54 4.05 Tm 0.64 0.63 0.89 0.79 0.92 0.63 0.71 0.19 0.65 0.56 0.93 U 1.31 1.02 1.54 1.14 1.19 1.22 1.17 0.3 1.13 1.21 1.62 V 33 16 25 21 68 23 19 150 27 19 10 W 40 36 72 46 55 48 3 8 14 12 18 Y 41.1 41.6 62.7 49 55.2 39.1 46.3 12.6 38 33.1 61.7 Yb 4.32 4.24 6.15 5.38 6.33 4.28 5 1.41 4.62 3.78 6.57 Zr 214 176 251 227 260 198 215 51 243 217 310 Ag 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 As 261 36 281 27 114 40 5 5 38 28 24 Cd 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.6 0.5 0.5 0.5 Co 6 3 4 4 2 3 6 25 3 3 1 Cu 42 14 9 8 40 6 12 176 9 10 4 Li 10 10 10 10 10 10 10 10 10 10 10 Mo 3 3 3 3 3 3 3 2 3 3 2 Ni 4 1 1 4 1 3 1 6 4 3 1 Pb 21 4 7 5 10 2 8 3 6 3 3 Sc 7 12 17 12 15 12 15 15 7 7 3 Tl 10 10 10 10 10 10 10 10 10 10 10 Zn 209 100 109 94 122 81 86 82 39 44 60 area show that Pb is relatively depleted in the miner- associated with the arsenopyrite occurs with minerali- alised samples which is characteristic of orogenic gold sation (Amponsah et al., 2015). deposits (Wang et al., 2021). However, this differs from reports on the Birimian at the Julie deposit and Obuasi 4.3.1 Major elements deposits in the Northeast and Southwest of Ghana The variations in MgO versus SiO show the schist respectively (Dzigbodi-Adjimah, 1993; Amponsah and the gneiss both have similar SiO content with the et al., 2015). Additionally, elements such as Co, Cu gneisses having marginally higher MgO content and Ni are also more enriched in the unmineralised (Figure 8(a)). The relatively high SiO content of the samples as compared to the mineralised samples. This rocks suggests a felsic source for the rock in the study differs from what has been reported in the Bekpong area. The graph shows a negative correlation between deposit on the Lawra belt of Ghana, where Ni and Co GEOLOGY, ECOLOGY, AND LANDSCAPES 9 Table 3. Stable isotopes, major and trace elements composition of the gneiss and quartz. 040/PK 032/TMB 027/KS 007/81 017/81 018/81 014/OKS 013/N 027/QZ 006/N 006/81 Gneiss Gneiss Gneiss Gneiss Gneiss Gneiss Gneiss Gneiss Quartz Quartz Quartz δ O‰ 13.2 8.9 12 14.4 14.2 14.1 13.6 14.4 12.7 17.4 14.7 δD‰ −58 −43 −68 −53 −48 −50 −91 −53 −71 −117 −98 SiO 65.2 66.9 75.9 54.3 80.8 74.1 76.9 73 97.6 100 99.3 Al O 12.1 10.5 11.4 12.6 8.91 11.55 11.35 11.6 0.27 0.39 0.05 2 3 Fe O 6.92 6.58 3.01 11.15 2.37 2.96 3.01 5.9 0.83 0.82 0.7 2 3 CaO 2.19 3.61 1.83 4.54 1.4 1.34 0.43 1.87 0.02 0.02 0.01 MgO 0.57 0.99 0.54 2.8 0.48 0.56 0.18 0.24 0.01 0.02 0.01 Na O 3.89 1.76 3.18 5.07 3.68 4.87 4.13 3.22 0.06 0.03 0.01 K O 1.74 2.6 3.76 0.78 0.71 1.67 2.84 3.07 0.03 0.09 0.01 TiO 0.64 0.66 0.29 1.18 0.28 0.33 0.23 0.37 0.01 0.01 0.01 MnO 0.16 0.23 0.5 0.19 0.06 0.06 0.03 0.13 0.01 0.01 0.01 P O 0.3 0.19 0.04 0.23 0.08 0.08 0.02 0.05 0.01 0.03 0.01 2 5 SiO /Al O 5.39 6.37 6.66 4.31 9.07 6.42 6.78 6.29 361.48 256.41 1986.00 2 2 3 Na O/K O 2.24 0.68 0.85 6.50 5.18 2.92 1.45 1.05 2.00 0.33 1.00 2 2 LOI 6.26 8.19 0.5 7.45 2.1 2.41 0.57 0.65 0.04 0.14 0.03 Total 100.06 100.58 100.59 100.38 100.93 100.05 99.81 100.24 98.93 101.61 100.15 Au 1.04 1 0.001 0.01 0.001 0.001 0.01 0.001 0.11 0.001 0.01 Ba 527 1470 1205 443 206 597 1040 1050 18.6 131 2 Ce 56.7 58.7 67 23.1 35.4 38.5 36.6 44.2 0.6 38.9 0.3 Cr 170 158 120 60 150 200 60 110 350 320 350 Cs 0.97 1.24 0.19 1.02 0.79 0.58 0.41 1.81 0.01 0.04 0.01 Dy 10.8 8.96 10.25 4.05 3.87 4.11 8.42 7.81 0.07 1.28 0.05 Er 6.52 6.14 6.49 2.42 2.31 2.59 5.79 5.06 0.04 0.29 0.03 Eu 2.31 2.1 1.6 1.25 0.7 0.97 1.13 1.44 0.02 0.89 0.02 Ga 17.4 11.8 19.8 16.3 10.1 12 19.1 18.7 0.5 0.7 0.1 Gd 10.5 8.72 10.3 4.21 4.11 4.59 7.07 6.98 0.09 2.86 0.05 Hf 7 6.7 9 2.7 4.1 4.4 8.8 7.2 0.1 0.1 0.1 Ho 2.19 1.71 2.03 0.81 0.75 0.8 1.78 1.56 0.01 0.16 0.01 La 27.1 26.1 30.5 11.2 17.2 18.1 15.5 19.5 0.2 29.6 0.2 Lu 0.87 1.1 1.01 0.35 0.36 0.43 0.9 0.74 0.01 0.03 0.01 Nb 7.3 8.6 8 3 4.9 4.6 8.6 7.1 0.1 0.1 0.1 Nd 33.4 32.1 36.8 14.4 18.4 20.3 24.6 24.9 0.4 24.6 0.2 Pr 8.3 8.62 9.47 3.47 4.77 5.3 5.84 6.16 0.09 7.31 0.05 Rb 36.6 51.9 61.3 22 20.2 28.4 55.8 64.2 0.9 1.9 0.2 Sm 9.1 6.78 9.36 3.89 4.08 4.36 6.12 6.5 0.09 5.05 0.06 Sn 2 2 3 1 1 1 1 2 1 1 1 Sr 161.5 252 136 280 187.5 225 56.4 144 2 20.7 0.6 Ta 0.4 0.5 0.5 0.1 0.3 0.3 0.4 0.4 0.1 0.1 0.1 Tb 1.68 1.52 1.63 0.65 0.62 0.68 1.24 1.19 0.01 0.29 0.01 Th 3.42 4.72 4.22 1.32 2.69 3.03 4.38 3.14 0.05 0.05 0.05 Tm 0.89 0.94 0.95 0.34 0.34 0.38 0.86 0.75 0.01 0.04 0.01 U 1.54 1.22 1.62 0.57 1.33 1.12 1.62 0.85 0.05 0.1 0.05 V 25 52 5 302 17 19 6 5 5 5 5 W 72 65 1 2 10 1 1 1 1 2 1 Y 62.7 53.2 58.3 22.7 21.2 23.4 51 44.2 0.5 3.5 0.1 Yb 6.15 5.43 6.54 2.3 2.43 2.7 6.25 5.16 0.08 0.19 0.03 Zr 251 256 321 93 146 155 317 254 2 2 2 Ag 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 As 281 114 260 5 5 5 5 5 5 5 5 Cd 0.5 0.5 0.5 0.6 0.5 0.5 0.5 0.5 0.5 0.5 0.5 Co 4 2 1 26 3 4 1 1 1 1 1 Cu 9 40 8 51 12 3 9 7 3 4 4 Li 10 10 10 30 10 10 10 10 10 10 10 Mo 3 3 2 1 3 2 3 2 3 3 5 Ni 1 1 1 4 7 2 3 1 3 3 4 Pb 7 10 8 2 7 7 10 10 2 9 2 Sc 17 15 6 24 6 7 5 12 1 1 1 Tl 10 10 10 10 10 10 10 10 10 10 10 Zn 109 122 134 102 61 57 17 148 5 4 3 SiO and MgO which is characteristic of such rocks. correlation matrix shows that Al O has a strong 2 2 3 Binary plots and regression coefficients were positive correlation with TiO (R = 0.912), K 2 2 2 2 used to determine the relative mobility among the O (R = 0.945), and Fe O (R = 0.705). Figure 8 2 3 elements from the samples in the study area. (b) shows that TiO and Al O are the most immo- 2 2 3 According to de Almeida (2009), immobile ele- bile pair, regardless of the hydrothermal alteration ments are insoluble and will therefore result in and mineralisation. This strong positive correlation a regression line which passes through the origin between these oxides indicates that they were prob- of a binary diagram and will also record the highest ably immobile during the hydrothermal event. regression coefficient in a correlation matrix. The Additionally, both mineralised and unmineralised 10 R. W. KAZAPOE ET AL. Figure 7. Histogram plots of (a) trace elements concentrations of mineralized schist samples (b) trace elements concentrations of unmineralized schist samples (c) trace elements concentrations of mineralised gneiss samples (d) trace elements concentrations of unmineralized gneiss samples. Figure 8. (a) a plot of MgO versus SiO of the schist and the gneiss from the Abansuoso area (b) Plot of TiO versus Al O following 2 2 2 3 an approximate linear trend which indicates that they are the most immobile pair, regardless of the alteration and mineralization. GEOLOGY, ECOLOGY, AND LANDSCAPES 11 Figure 9. The plot of principal component I against II for the trace elements from the unmineralized samples. Figure 10. The plot of principal component I against II for the trace elements from the mineralised samples. samples align in an approximately linear trend, 4.3.3 Factor analysis suggesting that these rocks may have originated Factor analysis has been used to determine element from a shared homogenous source. association with gold from whole-rock samples from the Atacora structural unit in Benin successfully (Adjo 4.3.2 Element association et al., 2021). Additionally, Yigit and Hofstra (2003) The close association of As, Mo and SiO shows that As applied the multivariate analysis of geochemical data is linked with the felsic minerals abundant in rocks such on whole-rock samples to assess favourable geochem- as the schist. MgO is associated with MnO and CaO, ical attributes of the host lithologies as well as to indicating they originated from the mafic mineral- characterise alteration and associated element enrich- bearing rocks such as the biotite-rich gneiss (Figure 9). ments and depletion. The biplot for the mineralised samples strongly Four principal components with eigenvalues > 1 suggests the influence of hydrothermal fluids in the were extracted which accounted for a total variance element association of the rocks. Au is associated with of 81.04% (Table 4). PC1 with a total variance of Zn, P O , As, W and Ba (Figure 10). These are the 32.08%, has K O, Ga, Ba, U, Al O , Zn, TiO , MgO, 2 5 2 2 3 2 array of elements typical of orogenic gold deposits of CaO and Cr strongly loaded as the components and the Birimian (Klemd and Hirdes, 1997). This particu- these factors are plotted in rotated space applying R – lar array shows that these elements are the pathfinders mode factor analysis. PC2 accounts for 28.82% of the for Au in the study area. Additionally, the association total variance with TiO , V, Co, Cd, MgO, Fe O , 2 2 3 of Mo with Ga, K O and Cr shows the influence of CaO, Cu and Ni with negative loading for SiO as 2 2 alteration. the factors loading strongly (Table 4). PC3 which 12 R. W. KAZAPOE ET AL. a. Table 4. Rotated factor Matrix Factor 1 2 3 4 Log(K O) 0.952 0.207 0.097 −0.122 Log(Ga) 0.944 0.266 0.090 0.079 Log(Ba) 0.934 0.104 0.020 0.032 Log(U) 0.934 −0.082 0.225 0.088 Log(Al O ) 0.917 0.345 0.080 0.060 2 3 Log(Zn) 0.776 0.411 0.311 0.062 Log(TiO ) 0.730 0.601 0.225 0.146 Log(Pb) 0.418 −0.055 0.069 −0.050 Log(V) 0.079 0.955 0.101 −0.023 Log(Co) 0.032 0.951 −0.010 −0.086 Log(Cd) −0.115 0.871 −0.292 −0.164 Log(SiO ) −0.403 −0.838 0.091 0.277 Log(MgO) 0.578 0.777 −0.058 −0.087 Log(Fe O ) 0.492 0.723 0.283 0.120 2 3 Log(CaO) 0.594 0.695 −0.009 0.069 Log(Cu) 0.065 0.599 0.203 0.496 Log(As) 0.242 −0.102 0.871 −0.047 Log(W) 0.244 0.111 0.824 −0.245 Au 0.098 0.028 0.805 0.024 Log(Mo) −0.332 −0.489 0.523 0.135 Log(Na O) −0.161 −0.038 −0.187 0.653 Log(Cr) −0.512 0.153 0.161 −0.619 Log(Ni) −0.209 0.558 −0.232 −0.607 Extraction Method: Principal Axis Factoring. Rotation Method: Varimax with Kaiser Normalization. Rotation converged in 7 iterations. comprises As, Au, W and Mo as the strongly loaded altered samples in this study. The altered samples factors accounts for 12.73% of the total variance. were selected to reflect the main phases of hydrother- While PC4 has a total variance of 7.41 with Na O, Cr mal alteration identified in the study area (i.e., sericite and Ni as the strongly loaded factors. and hematite alterations). Average compositions for the least altered zones were plotted against samples from the average composition for the altered zones by 4.3.4 Mass balance: isocon diagrams using the EASYGRESGRANT MS-Excel sheet pro- The application of isocon diagrams in mass balance vided in Lopez-Moro (2012). The least altered samples studies developed by Grant, 1985) has been success- contain anomalous values of Cd, Cu, Mo, Ni, Se, fully applied to several studies in different deposit U and V, which are interpreted to be introduced in types around the world (e.g., Hofstra, 1994; Emsbo this unit during diagenesis, pre-dating the mineralisa- et al., 2003; Yigit and Hofstra, 2003; Almeida, 2009; tion event. Amponsah et al., 2015 among others). The isochron After the selection of the samples, the identification diagrams have been primarily used to evaluate major of the immobile elements was done. Binary plots and element gains or losses in hydrothermal processes by regression coefficients were used to identify that TiO comparing barren or the least altered samples to and Al O are the most immobile pair, regardless of 2 3 altered samples in a particular deposit. De Almeida the alteration and mineralisation (Figure 8(b)). These (2009) applied this technique to evaluate major ele- two oxides were therefore considered the most immo- ment gains and losses in the Carlin type deposit in bile pair and used for the isocon diagrams, following Nevada. Amponsah et al. (2015) also applied the tech- the method of Grant, 1985, Grant, 1985). The isocon nique in a study of the Julie deposit in the north of diagrams (Figures 11(a-c)). Ghana in a similar geological setting as the study area. The least altered samples used in this study were taken from the hills around Nfante and South of Subriso 4.4 Stable Oxygen (O) and Hydrogen (H) isotope away from the mining pits or deposits. The relative data level of hydrothermal alteration of the samples was evaluated based on the petrographic studies as well as The δ O values of the rocks found in the Abansuoso the whole-rock composition of the samples. deposit range from 8.9‰ to 17.4‰ (n = 22, According to Querubin and Yumul (2005) samples mean = +13.7‰). The δ O values in the schist are which are the least altered tend to characteristically between 12.7‰ – 14.4‰ (n = 11, mean = 13.7‰). have low Loss on Ignition (LOI) values of less than 5% These values are similar to those determined in the and have Na O values ranging between 3–5% for felsic gneisses which range between 8.9‰ – 14.4‰ (n = 8, samples and CaO + Na O falling between 5 and 10% mean = 13.1‰; Table 5). The range of δ O values for mafic gneisses. Samples which displayed these determined in the gneisses of the Abansuoso area are characteristics were therefore classified as the least similar to those recorded in the granitic rocks found GEOLOGY, ECOLOGY, AND LANDSCAPES 13 Figure 11. (a) Isocon diagrams constructed for pairs of unaltered and altered samples from the outer alteration zone (hematite alteration) (b) Isocon diagrams constructed for pairs of unaltered and altered samples from the inner alteration zone (Sericite alteration) (c) Isocon diagrams constructed for pairs of unaltered and altered samples from the inner alteration zone (Sericite alteration). Table 5. δ O and δD composition of rocks from the Abansuoso Area of Southwestern Ghana. SN Sample ID Rock Type δ O‰ δD‰ Au Grade (g/ton) H O (wt%) 1 020/79 Schist 14.3 −45 1.66 0.98 2 009/79 Schist 13.3 −42 2.1 0.91 3 013/79 Schist 13 −50 1.04 0.50 4 019/79 Schist 13.9 −62 2 0.66 5 011//79 Schist 13.9 −34 1.4 0.94 6 012/79 Schist 14.1 −33 0.9 0.60 7 014/79 Schist 12.7 −49 0.15 0.65 8 001/81 Schist 13.1 −47 0.22 0.81 9 OS/A Schist 14.4 −60 0.11 0.62 10 001/OS Schist 14.4 −51 0.22 0.83 11 008/PK Schist 14.1 −51 0.11 1.51 12 040/PK Gneiss 13.2 −58 1.3 0.56 13 032/TMB Gneiss 8.9 −43 1 2.53 14 027/KS Gneiss 12 −68 0.11 0.27 15 007/81 Gneiss 14.4 −53 0.01 0.74 16 017/81 Gneiss 14.2 −48 0.1 0.44 17 018/81 Gneiss 14.1 −50 0.1 0.78 18 014/OKS Gneiss 13.6 −91 0.01 0.30 19 013/N Gneiss 14.4 −53 0.1 0.34 20 027/QZ Quartz vein 12.7 −71 0.11 0.10 21 006/N Quartz vein 17.4 −117 0.1 0.10 22 006/81 Quartz vein 14.7 −98 0.01 0.12 around the Lake Bosomtwi area of the Ashanti Region δ O values determined in the quartz veins, although of Ghana (12.7‰ to 12.9‰, Koeberl et al., 1998). from a limited number of samples, show However, they are higher than the δ O values of the a comparatively wider range relative to the values Pepiakese granites of Southwestern Ghana (8.6‰ to (15.6‰ – 18.5‰) determined in quartz in the 9.0‰, Koeberl et al., 1998). The δ O values of the Prestea deposit by Hammond and Shimazaki (1994). quartz veins were higher, ranging from 12.7‰ – The δ O values determined for the rocks of the 17.4‰ (n = 3, mean = 14.9‰). The range of study area are within a narrow range of 12.0 to 14.7‰ 14 R. W. KAZAPOE ET AL. 8 10 12 14 16 18 -20 -40 -60 Schist Gneiss -80 Quartz vein -100 -120 -140 δ O‰ Figure 12. A plot of δ O against δD of the rocks from the study area. with the notable exception of two samples from the suggests that the variation of δD is not controlled schist and quartz veins which recorded values of 8.9‰ by depth. and 17.4‰, respectively. This generally narrow range of δ O values is similar to those previously recorded 4.4.1 Controls on oxygen and hydrogen isotope in the Birimian in Ghana (Hammond & Shimazaki, composition of the rocks in the Abansuoso area 1994; Oberthuer et al., 1996; Mumming et al., 1996). The H O vs δD shows a negative correlation between The δD values of the rocks for the Abansuoso area the water content and the δD for the samples from the range between −117 to −33‰ (n = 22, study area (Figure 12). This may reflect a greater mean = −57.9‰). The values in the schist range extent of water-rock reaction while the increasing δD from −62 to −33‰ (n = 11, mean = −47.6‰), while values suggest an influence from mineralogy and tem- the δD values of the gneisses however range from −91 perature of alteration. to −43‰ (n = 8, mean = −58‰). The δD values within The factors which control the final δD values of the unmineralised quartz vein have a wider range from minerals and as a result the bulk rock are tempera- −117 to −71‰ (n = 3, mean = −95.3‰). An inverse ture of crystallisation of the minerals in the rocks, relationship exists between δD and δ O values in the the δD values of the fluid present and interaction rocks from the study area (Figure 12). This is consis- with hydrothermal fluids (e.g., Harris et al., 1997). tent with the relationship between δD and δ O in this According to Taylor & South, 1985), hydrother- instance. mally altered rocks with significant chlorite content Samples collected from the mining pits present tend to have low δ O values as compared to other similar isotopic composition to the drill core samples. alteration minerals because chlorite has low 18 18 The δ O values of the surface samples range from 8.9 δ O. Some hydrothermally altered rocks are also to 17.4‰ (n = 10, average = 13.51‰) in comparison to characterised by elevated δD values as compared to the drillcore samples which range from 12.7–14.7‰ clay-rich rocks altered by the same fluids at lower (n = 12, average = 13.81‰). The values are similar to temperatures (Friedman and O’Neil, 1977). Casey values recorded in the Pepiakese granites of and Taylor (1982) determined δ O values for Southwestern Ghana (8.6‰ – 9.0‰) but higher than chlorite within vein chlorites to be between 1.5‰ 18 18 the quartz veins δ O values ranging from 12.7–17.4‰ to 2.0‰. The sample with the lowest δ O value from the same area (Koeberl et al., 1998). identified in the Abansuoso area is TMB-32 which The δD values of the surface samples are between is rich in chlorite. The influence of quartz and −43 and −117‰ (n = 10, mean = −66.3‰). The δD sericite on the mineralised samples which repre- values of the drillcore samples are slightly lower, sents the later stages of alteration in the area between −98 and −33‰ (n = 12, tends to have higher δ O values which may serve average = −50.92‰). to mitigate the effect of chlorite to an extent. This 18 18 The variation of δD and δ O values of the rocks is reflected in the variation in δ O values in the sampled across the two drill holes from different analysed samples (Table 5). Furthermore, chlorite depths studied is relatively constant and is similar alteration is minimal across the area, which sug- to those of the surface samples. The δ O values of gests a greater influence of isotopic exchange the rocks vary averagely by ±1‰ across the drill- between the mineralising fluid and the host rock. core. The δD values show more variation but do not This may account for the lower δ O values of the appear to follow a systematic pattern. This also mineralised samples. This assertion is corroborated δD‰ GEOLOGY, ECOLOGY, AND LANDSCAPES 15 Figure 13. (a) Distribution of δ O among the various rock types in the study area compared with average values from the Birimian (b) Distribution of δD among the various rock types in the study area compared with average values from the Birimian as reported by Kazapoe et al. (2021). by the depleted H O which is usually high in rocks values which correlates poorly. The erratic relation with chlorite alteration and tends to have an may be a result of the limited samples used in this inverse relationship with δ O as described by study. Although it is possible to identify a relationship Larson (1984). between δ O and δD and mineralisation, the erratic As shown in Figures 13(a-b), the general distribu- distribution among the sample renders it difficult to tion of δ O and δD among the rocks do not appear to contour. be controlled by lithology. The distribution of The δ O value of the mineralised samples ranges δ O among the schist, which has a higher number between 8.9 and 14.3‰ with a mean value of 13.08‰. of mineralised samples in this study, is relatively more Aside from the aforementioned sample TMB which depleted than the gneiss. The unmineralised quartz has a value of 8.9‰, most of the samples vary by only samples are relatively more enriched than the schist ± 1‰. This restricted range of values is similar to and the gneiss (Figure 13(a)). A similar situation is samples located in the Birimian of Ghana, the majority seen in the distribution of the δD concentration values of which fall between 12 and 16 ‰. These values are in the samples where the schist is the most relatively similar to those determined from the silicate rocks enriched, with the quartz being the most depleted from the Ashanti Gold Belt, which are between 13– (Figure 13(b)). The distribution suggests an influence 15‰ (Mumming et al, 1996; Hammond & Shimazaki, of the mineralising fluid. 1994). Similarly, these values are comparable to those determined for mineralised samples from other parts of the Birimian in West Africa which range between 4.4.2 Relationship between the stable isotopic 13–18‰ (Fouillac et al., 1993; Lambert-Smith et al., concentration in rocks and mineralisation of the 2016; Lawrence et al., 2013). Abansuoso area The δ O values are consistent across the various The relationship shown in Figures 14(a-b) do define rock types as well. The schist shows a range of values a clearer correlation between δD and gold values in from 13 to 14.3‰ (n = 6, mean = 13.75‰) and 12.7 comparison to the relationship between δ O and gold 16 R. W. KAZAPOE ET AL. Figure 14. (a) Isotopic compositions of hydrogen (δD) versus its corresponding gold values (g/t) (b) Isotopic compositions of oxygen (δ O) versus its corresponding gold values (g/t). to 14.4‰ (n = 5, mean = 14.4‰). The gneisses also across the samples. The composition of δ O is show values between 12 and 14.1‰. The mineralised affected by fluid temperature as well as the δ O of and altered samples taken from the schist averagely the fluid from which it precipitated, especially with differ with a value of ±1‰ with the mineralised quartz (Taylor & South, 1985). samples being relatively depleted. The δ O from Figure 15 shows the bulk of the sample points dis- the unmineralised quartz samples from the study play the typical inverse relationship between δ O and area falls between 12.7 and 17.4‰ which is slightly δD values. Additionally, the majority of both δ O and more enriched than the mineralised samples. This δD values correlate with the gold values positively and may be due to the δ O isotopic exchange between inversely, respectively. This suggests a relationship the hydrothermal fluid-bearing quartz and the coun- between gold and the isotopic values. try rock. The vein quartz which represents the gangue The distribution of δD values among the samples mineral of the mineralising fluid, as a result, becomes shows a clearer differentiation between the minera- subsequently relatively richer in δ O compared to lised, altered and unmineralised samples (Figure 16 the wall rock. (a)). The values of δD within the mineralised samples Unmineralised samples from Ghana, Mali, and range from −62 to −33‰ (n = 8, mean = −15.88‰). Senegal had consistently high values, ranging from This is similar to values determined in the mineralised 12 to 20‰. The unmineralised samples from the zone from Prestea, Ghana by Hammond and study area fall within the range of 12.70–17.40‰ Shimazaki (1994) (−65 to −29‰) and Oberthuer (n = 7, mean = 14.49‰). However, when compared et al. (1996; −60 to −43‰) from the Birimian of with results from unmineralised samples from the Ghana. study area, there is only a little relative enrichment. The altered samples from the study area show The unmineralised samples taken from the gneisses a relative average depletion of −8‰ as compared to show a consistently narrow range between 13.6‰- the mineralised samples which does not show much 14.4‰ (n = 4, mean = 14.15) which varies with 1‰ distinction between the mineralised and altered GEOLOGY, ECOLOGY, AND LANDSCAPES 17 Figure 15. Schematic representation of drillcore SPDD 079 showing variations of δ O and δD with Au across the zones. Figure 16. (a) Variation in δ O values between the 3 main zones (b) Variation in δD values between the 3 main zones. 18 R. W. KAZAPOE ET AL. 18 18 Figure 17. Plots of (a) δ O against as of the samples (b) δD against As of the samples (c) δ O against W of the samples (d) δD against W of the samples. samples (Figure 16(b)). However, the unmineralised influence and relationship with the mineralising fluid samples are comparatively more depleted and show in relation to the δ O value. a wide range of δD values between −117 to −48‰ (n = 7, mean = −75.86‰). Although the differentia - 5. Discussion tion between δD values from the mineralised and altered samples is not distinguishable, the unminera- The mineral assemblage of sericite, quartz, mus- lised samples are much more distinct from the miner- covite, biotite and hornblende is suggestive of alised samples and altered samples, varying averagely greenschist to lower amphibolite metamorphic around −30 and −22‰ for the mineralised and altered conditions. The rocks in the area have quartz samples respectively. grains that show undulose extinction with some The isotopic content of the rocks was plotted exhibiting chessboard extinction which indicates against As and W. This is because As and W were dislocation slip activity during high-temperature determined from the lithogeochemistry to strongly deformation, probably linked to high strain within correlate with the gold and is characteristic of the the shear zone (Figure 3(a-b)). The introduction suite of elements associated with gold in orogenic of gold possibly occurred during hydrothermal deposits such as the area. This was done to further alteration, with the gold deposited together with examine the relationship between the isotopic concen- the sulphide phases. The occurrence of arsenopyr- tration of the rocks with the mineralisation in the area. ite and pyrite, within the altered and mineralised A plot of δ O and δD values against As show a poor rocks of the Abansuoso area is widespread. The correlation between δ O and As whiles δD shows two principal sulphide minerals are similar to a more consistent relationship with As (Figure 17 those associated with the gold mineralisation in (a-b)). Similarly, the plot of δ O and δD values the Obuasi deposit except for the absence of mar- against W also shows a poor correlation between casite or pyrrhotite (Oberthuer et al., 1996; Osae δ O and W whiles δD shows a relatively consistent et al. 1999). No silver was found in association relationship with As (Figure 17(c-d)). This further with the gold grains by EDS-SEM. This suggests supports the assertion that the δD shows a clearer a uniqueness of the gold mineralisation of the GEOLOGY, ECOLOGY, AND LANDSCAPES 19 Abansuoso deposit which differs from similar mean = −16‰) which shows an average depletion deposits within the Birimian of Ghana. of −8‰ relative to the altered samples. The unmi- From the factor analysis, PCA1 has elemental asso- neralised samples range between −117 and −48‰ ciations of heavy elements characteristic of the mafic (n = 7, mean = −76‰) which are lower than the minerals associated with the gneisses in the study area. mineralised and altered samples between −30 and PCA 2 also has elemental groups which correlate with −22‰ respectively. The general distribution of the hematite alteration seen in the mineralised zone. The δ O among the mineralised rocks does not appear presence of Fe O and TiO is further corroborated by to be controlled by lithology. Although the 2 3 2 the SEM result for hematite alteration. δ O of the mineralised samples show The elemental associations in factor three (Au, As, a systematic variation which suggests an influence Mo and W) are characteristic of the elemental associa- of the mineralising fluid, the values are not differ - tions linked with the generally accepted supracrustal entiated from the altered samples or to a lesser metamorphic models of gold origin proposed to be extent the unmineralised samples. This makes the sourced from around the brittle-ductile transition δ O on its own not diagnostic of gold mineralisa- zone (~10 km) (Groves and Santosh, 2016). The pre- tion. The effect of mineralisation on the δD values sence of As associated with the sulphide minerals of the rocks is however relatively more pro- (arsenopyrite) associated with mineralisation in the nounced. Therefore, a combination of the two area further confirms this. This confirms the results isotopes with conventional lithogeochemistry will of the biplot and suggests that these elements are the be the most suitable tool for the exploration of pathfinder elements for gold mineralisation in the gold resources in the area. The lack of variation in area. The isocon diagrams (Figures 11(a-c)) suggest δ O suggests a rock-buffered system, with an that mass was commonly conserved during alteration influx of metamorphic fluid derived from rocks with a minor mass gain observed in the inner altera- with similar δ O comparable to many orogenic tion envelope (sericite). The result shows the outer gold deposits. The δD values seem to be lower in alteration envelope representing hematite alteration the mineralised samples suggesting an influence of recorded a slight mass gain in oxides of MgO, CaO the fluid. and Fe O (Figure 11(a)). There were mass gains in 2 3 Co, Pb, Ni and Ca with a slight loss of mass in W, Cu, 6. Conclusion Pr, Mo and K O. Results for the inner alteration zone, representing the sericite alteration and mineralised The study found a close association between Au and zone show that As, Zn, Ba, Au, Co and W were As, Mo and W which is characteristic of the element added. Oxides such as K O, P O , Fe O , Na O and associations linked with the generally accepted supra- 2 2 5 2 3 2 MnO were lost. Additionally, Mo, Cr, CaO and MgO crustal metamorphic models of gold origin proposed remained constant. to be sourced from around the brittle-ductile transi- The uniform distribution of δ O values within tion zone. It was also determined that mass was com- the shear zone and outside the shear zone also monly conserved during alteration with a minor mass indicates that the mineralising fluid invades the gain observed in the inner alteration envelope (sericite rock in a pervasive manner. Additionally, the dis- alteration). 18 18 tribution of δ O and δD across the depth of the The concentrations of δ O (8.9‰ to 17.4‰, drillholes suggests no significant variation of iso- n = 22, mean = +13.7‰) and δD (−117 to −33‰, topic values with depth. The lack of variation in n = 22, mean = −57.9‰) in the rocks are comparable δ O suggests a rock-buffered system, with an to those from other Birimian regions in Ghana and influx of metamorphic fluid derived from rocks West Africa. The lack of variation in δ O suggests of similar δ O like many orogenic gold deposits. a system buffered by rocks, with metamorphic fluid The δD values seem to be lower in the mineralised originating from rocks with identical δ O values. samples which suggest an influence of the fluid However, the effect of mineralisation on the δD values which probably had a δD value close to zero. of rocks is substantially more apparent than its effect 18 18 Recorded δ O values among the mineralised sam- on the δ O values of the rocks. ples (8.9–14.3‰, n = 8, mean = 13.08‰) fall in The differentiation between the isotopic values a similar range to samples from the other parts of of the unmineralised samples on one hand and the the Birimian in Ghana and West Africa. This is altered and mineralised samples on the other hand lower than in the unmineralised samples with an points to the utility of the methodology in targeting average of +1‰ (12.70–17.40‰ give n = 7, gold mineralisation in general. However, the lack of mean = 14.49‰). The δD however shows more a clear distinction between the altered and miner- variation between the mineralised samples and the alised samples shows that the methodology requires unmineralised samples. The mineralised samples an additional layer of data to be able to pick up the record values between −62 and −33‰ (n = 8, mineralised zones alone. The authors recommend 20 R. W. KAZAPOE ET AL. that a larger scale study using more samples on the Amponsah, P. O., Salvi, S., Béziat, D., Siebenaller, L., Baratoux, L., & Jessell, M. W. (2015). Geology and geo- scale of conventional lithogeochemical surveys be chemistry of the shear-hosted Julie gold deposit, NW carried out. This will further investigate the viabi- Ghana. Journal of African Earth Sciences, 112, 505–523. lity of this approach as a gold exploration https://doi.org/10.1016/j.jafrearsci.2015.06.013 technique. Amponsah, P. O., Salvi, S., Didier, B., Baratoux, L., Siebenaller, L., Jessell, M., Nude, P. M., & Gyawu, E. A. (2016). Multistage gold mineralisation in the Wa-Lawra greenstone belt, NW Ghana: The Bepkong deposit. Acknowledgments Journal of African Earth Sciences, 120, 220–237. https:// This research paper is part of the first author’s doi.org/10.1016/j.jafrearsci.2016.05.005 Ph.D. dissertation, and he would like to thank the African Borthwick, J., & Harmon, R. S. (1982). A note regarding Union and the Pan African University for awarding him CIF3 as an alternative to BrF5 for oxygen isotope analysis. a Ph.D. Scholarship and the University for Development Geochimica et cosmochimica acta, 46(9), 1665–1668. Studies for granting him study leave. Supercare Gold and Bouabdellah, M., & Slack, J. F. (Eds.). (2016). Mineral depos- Pelangio Exploration Company are greatly appreciated for its of North Africa. Springer. granting access to their concession and assisting with the Bowman, J. R., Parry, W. T., Kropp, W. P., & Kruer, S. A. fieldwork. The authors are also grateful to the anonymous (1987). Chemical and isotopic evolution of hydrothermal reviewers whose insightful comments, suggestions, and solutions at Bingham, Utah. Economic Geology, 82(2), reviews significantly improved the paper’s quality. 395–428. https://doi.org/10.2113/gsecongeo.82.2.395 Casey, W. H., & Taylor, B. E. (1982). Oxygen, hydrogen, and sulfur isotope geochemistry of a portion of the West Shasta Cu-Zn district, California. Economic Geology, 77(1), 38–49. Disclosure statement Chudasama, B., Porwal, A., Kreuzer, O. P., & Butera, K. (2016). Geology, geodynamics and orogenic gold pro- No potential conflict of interest was reported by the spectivity modelling of the Paleoproterozoic Kumasi author(s). Basin, Ghana, West Africa. Ore Geology Reviews, 78, 692–711. Dzigbodi-Adjimah, K. (1993). Geology and geochemical pat- Funding terns of the Birimian gold deposits, Ghana, West Africa. Journal of Geochemical Exploration, 47(1–3), 305–320. This research received partial funding from the African Dzikunoo, E. A., Kazapoe, R. W., & Agbetsoamedo, J. E. Union Commission through the Pan African University. (2021). An integrated structural and geophysical approach to defining the structures of part of the Nangodi greenstone belt, northeastern Ghana. Journal ORCID of African Earth Sciences, 180, 104238. https://doi.org/ 10.1016/j.jafrearsci.2021.104238 Raymond Webrah Kazapoe http://orcid.org/0000-0002- Emsbo, P., Hofstra, A. H., Lauha, E. A., Griffin, G. L., & 0307-7834 Hutchinson, R. W. (2003). Origin of high-grade gold ore, source of ore fluid components, and genesis of the Meikle and neighboring Carlin-type deposits, northern References Carlin trend, Nevada. Economic Geology, 98(6), 1069– Adjo, F. B., Nude, P. M., Adissin, L. G., Bolarinwa, A. T., & Fougerouse, D., Micklethwaite, S., Ulrich, S., Miller, J., Anagonou, B. (2021). Paleoweathering, Provenance and Godel, B., Adams, D. T., & McCuaig, T. C. (2017). tectonic setting of gold-bearing neoproterozoic rocks Evidence for two stages of mineralization in West from the Atacora Structural Unit, Pan-African Africa’s largest gold deposit: Obuasi, Ghana. Economic Dahomeyide belt, northwestern Bénin. Journal of Geology, 112(1), 3–22. African Earth Sciences, 183, 104327. Fouillac, A. M., Dommanget, A., & Milési, J. P. (1993). Agbenyezi, T. K., Foli, G., & Gawu, S. K. (2020). A carbon, oxygen, hydrogen and sulfur isotopic study of Geochemical Characteristics of Gold-Bearing Granitoids the gold mineralisation at Loulo, Mali. Chemical Geology, at Ayanfuri in the Kumasi Basin, Southwestern Ghana: 106(1–2), 47–62. https://doi.org/10.1016/0009-2541(93) Implications for the Orogenic Related Gold Systems. 90165-F Earth Sciences Malaysia (ESMY), 4(2), 127–134. Friedman, I., & O'Neil, J. R. (1977). Compilation of stable Allibone, A., Hayden, P., Cameron, G., & Duku, F. (2004). isotope fractionation factors of geochemical interest (Vol. Paleoproterozoic gold deposits hosted by albite-and 440). US Government Printing Office. carbonate-altered tonalite in the Chirano District, Galipp, K., Klemd, R., & Hirdes, W. (2003). Metamorphism Ghana, West Africa. Economic Geology, 99(3), 479–497. and geochemistry of the paleoproterozoic Birimian sefwi https://doi.org/10.2113/gsecongeo.99.3.479 volcanic belt (Ghana, West Africa). Geologisches Jahrbuch Allibone, A., Teasdale, J., Cameron, G., Etheridge, M., Uttley, D, 111, 151–191. http://dx.doi.org/10.1016/S0301-9268 P., Soboh, A., & Lamb, E. (2002). Timing and structural (99)00024-8 controls on gold mineralization at the Bogoso gold mine, Goldfarb, R. J., Baker, T., Dubé, B., Groves, D. I., Hart, C. J., Ghana, West Africa. Economic Geology, 97(5), 949–969. & Gosselin, P. (2005). Distribution, character, and genesis Almeida, C. (2009). Mineral paragenesis, geochemistry and of gold deposits in metamorphic terran. geochronology investigations of the Carlin-type gold Goldfarb, R. J., Groves, D. I., & Gardoll, S. (2001). Orogenic deposits at the Goldstrike property, northern Nevada: gold and geologic time: A global synthesis. Ore Geology Implications for ore genesis, igneous petrogenesis and Reviews, 18(1–2), 1–75. mineral exploration (Doctoral dissertation). GEOLOGY, ECOLOGY, AND LANDSCAPES 21 Grant, F. S. (1985). Aeromagnetics, geology and ore envir- (Uzbekistan): Implications for the age and sources of onments, I. Magnetite in igneous, sedimentary and meta- Au mineralization. Mineralium Deposita, 36(5), 379–392. morphic rocks: An overview. Geoexploration, 23(3), Kesse, G. O. (1985). The mineral and rock resources of 303–333. https://doi.org/10.1016/0016-7142(85)90001-8 Ghana. A.A. Balkema. Griffis, R. J., & Agezo, F. L., 2000, Mineral occurrence and Klemd, R., & Hirdes, W. (1997). Origin of an unusual fluid exploration potential of northern Ghana. Mineral composition in Early Proterozoic palaeoplacer and lode- Commission Report. 132–1133 (Accra, Ghana: Minerals gold deposits in Birimian greenstone terranes of West Commission). Africa. South African Journal of Geology, 100(4), 405–414. Griffis, J., Barning, K., Agezo, F. L., & Akosa, F., 2002, Gold Koeberl, C., Reimold, W., Blum, J., & Chamberlain, C. P. deposits of Ghana prepared on behalf of Ghana mineral (1998). Petrology and geochemistry of target rocks from commission. 432 (Accra, Ghana: Minerals Commission). the Bosumtwi impact structure, Ghana, and comparison Groves, D. I., & Phillips, G. N. (1987). The genesis and with Ivory Coast tektites. Geochimica et Cosmochimica tectonic control on Archaean gold deposits of the western Acta, 62(12), 2179–2196. https://doi.org/10.1016/S0016- Australian shield—a metamorphic replacement model. 7037(98)00137-9 Ore Geology Reviews, 2(4), 287–322. https://doi.org/10. Lambert-Smith, J. S., Lawrence, D. M., Vargas, C. A., 1016/0169-1368(87)90009-6 Boyce, A. J., Treloar, P. J., & Herbert, S. (2016). The Groves, D. I., & Santosh, M. (2016). The giant Jiaodong gold Gounkoto Au deposit, West Africa: Constraints on ore province: The key to a unified model for orogenic gold genesis and volatile sources from petrological, fluid inclu- deposits?. Geoscience Frontiers, 7(3), 409–417. sion and stable isotope data. Ore Geology Reviews, 78, 606–622. https://doi.org/10.1016/j.oregeorev.2015.10.025 Hammond, N. Q., & Shimazaki, H. (1994). Geology and Larson, S. Å., & Tullborg, E. L. (1984). Stable isotopes of geochemical aspects of ore formation at the Prestea fissure-filling calcite from Finnsjön, Uppland, Sweden. mesothermal vein gold deposit in the Birimian system Lithos, 17, 117–125. of Ghana. International Geology Review, 36(8), 715–731. Lawrence, D. M., Treloar, P. J., Rankin, A. H., Boyce, A., & https://doi.org/10.1080/00206819409465484 Harbidge, P. (2013). A fluid inclusion and stable isotope Harris, C., Faure, K., Diamond, R. E., & Scheepers, R. study at the Loulo mining district, Mali, West Africa: (1997). Oxygen and hydrogen isotope geochemistry of Implications for multifluid sources in the generation of S-and I-type granitoids: The Cape Granite suite, South orogenic gold deposits. Economic Geology, 108(2), Africa. Chemical Geology, 143(1–2), 95–114. 229–257. https://doi.org/10.2113/econgeo.108.2.229 Harris, C., Oom, B. M., & Diamond, R. E. (1999). A pre- Leube, A., Hirdes, W., Mauer, R., & Kesse, G. O. (1990). The liminary investigation of the oxygen and hydrogen iso- early Proterozoic Birimian Supergroup of Ghana and tope hydrology of the greater Cape Town area and an some aspects of its associated gold mineralization. assessment of the potential for using stable isotopes as Precambrian Research, 46(1–2), 139–165. tracers. Water S. A, 25(1), 15–24. Lopez-Moro, F. J. (2012). EASYGRESGRANT—A Harris, C., & Vogeli, J. (2010). Oxygen isotope composition of Microsoft Excel spreadsheet to quantify volume changes garnet in the Peninsula Granite, Cape Granite Suite, South and to perform mass-balance modeling in metasomatic Africa: Constraints on melting and emplacement mechan- systems. Computers & Geosciences, 39, 191. isms. South African Journal of Geology, 113(4), 401–412. Mumin, A. H., Fleet, M. E., & Longstaffe, F. J. (1996). Hart, C. J., McCoy, D. T., Goldfarb, R. J., Smith, M., Roberts, Evolution of hydrothermal fluids in the Ashanti gold P., Hulstein, R., & Bundtzen, T. K. (2002). Geology, belt, Ghana; stable isotope geochemistry of carbonates, exploration, and discovery in the Tintina Gold Province, graphite, and quartz. Economic Geology, 91(1), 135–148. Alaska and Yukon. Special Publication-Society of https://doi.org/10.2113/gsecongeo.91.1.135 Economic Geologists, 9, 241–274. Mumm, A. S., Oberthür, T., Vetter, U., & Blenkinsop, T. G. Hirdes, W., Leube, A., Für Geowissenschaften, B., & , R. (1997). High CO2 content of fluid inclusions in gold miner- (1989). On gold mineralization of the Proterozoic Birimian alisations in the Ashanti Belt, Ghana: A new category of ore supergroup in Ghana/West Africa. BGR. forming fluids?. Mineralium Deposita, 32(2), 107–118. Hofstra, A. H. (1994). Reconnaissance Stream-Sediment Oberthuer, T., Mumm, A. S., Vetter, U., Simon, K., & Geochemistry of Twentynine Palms Marine Corps Base, Amanor, J. A. (1996). Gold mineralisation in the California, 95(13). US Geological Survey. Ashanti Belt of Ghana; genetic constraints of the stable Hrdy, F., & Kyser, T. K. (1995). Origin, timing, and fluid isotope geochemistry. Economic Geology, 91(2), 289–301. characteristics of an auriferous event; the Proterozoic https://doi.org/10.2113/gsecongeo.91.2.289 Jasper lode gold deposit, Saskatchewan, Canada. Osae, S., Kase, K., & Yamamoto, M. (1999). Ore Mineralogy Economic Geology, 90(7), 1918–1933. https://doi.org/10. and Mineral Chemistry of the Ashanti Gold Deposit at 2113/gsecongeo.90.7.1918 Obuasi, Ghana. Resource Geology, 49(1), 1–11. Jessell, M. W., Amponsah, P. O., Baratoux, L., Perrouty, S., Jessell, M. W., Bourassa, Y., Miller, J., Apau, D., Asiedu, D. K., Loh, G. K., & Ganne, J. (2012). Crustal- Siebenaller, L., Salvi, S., Baratoux, L., Aillères, L., scale transcurrent shearing in the paleoproterozoic Béziat, D., & Salvi, S. (2015). The Wassa deposit: A poly- Sefwi-Sunyani-Comoe region, West Africa. deformed orogenic gold system in southwest Ghana– Precambrian Research, 212, 155–168. https://doi.org/ implications for regional exploration. Journal of African 10.1016/j.precamres.2012.04.015 Earth Sciences, 112, 536–547. https://doi.org/10.1016/j. Kazapoe, R.W., Olugbenga, O., Arhin, E. and Olisa, O. jafrearsci.2015.03.003 (2021). Isotope geochemistry as a tool in the exploration Querubin, C. L., & Yumul, G. P. (2005). Mass change calcu- of gold occurrences in Ghana: a review. Arabian Journal lations as applied to the search for volcanogenic massive of Geosciences, 14(19), pp.1–15. sulfide deposits: an example from Malusok, Siocon, Kempe, U., Belyatsky, B., Krymsky, R., Kremenetsky, A., & Zamboanga del Norte, Mindanao, Philippines. Resource Ivanov, P. (2001). Sm–nd and Sr isotope systematics of Geology, Vol. 55(2), pp. 61–72. scheelite from the giant Au (–W) deposit Muruntau 22 R. W. KAZAPOE ET AL. Senyah, G. A., Dampare, S. B., & Asiedu, D. K. (2016). Vennemann, T. W., & Smith, H. S. (1990). The rate and Geochemistry and tectonic setting of the Paleoproterozoic temperature of reaction of CIF3 with silicate minerals, metavolcanic rocks from the Chirano Gold District, Sefwi and their relevance to oxygen isotope analysis. belt, Ghana. Journal of African Earth Sciences, 122, 32–46. Chemical Geology: Isotope Geoscience Section, 86(1), https://doi.org/10.1016/j.jafrearsci.2015.07.022 83–88. Smith, A. J., Henry, G., & Frost-Killian, S. (2016). A review Wang, C., Deng, J., Bagas, L., He, X., & Zhang, J. (2021). of the Birimian supergroup-and Tarkwaian group-hosted Origin and classification of the Late Triassic Huaishuping gold deposits of Ghana. Episodes, 39(2), 177–197. https:// gold deposit in the eastern part of the Qinling-Dabie doi.org/10.18814/epiiugs/2016/v39i2/95775 Orogen, China: Implications for gold metallogeny. Taylor, B. E., & South, B. C. (1985). Regional stable isotope Mineralium Deposita, 56(4), 725–742. systematics of hydrothermal alteration and massive sul- Yao, Y., Murphy, P. J., & Robb, L. J. (2001). Fluid fide deposition in the West Shasta District, California. characteristics of granitoid-hosted gold deposits in Economic Geology, 80(8), 2149–2163. https://doi.org/10. the Birimian Terrane of Ghana: A fluid inclusion 2113/gsecongeo.80.8.2149 microthermometric and Raman spectroscopic study. Tsiboah, T., & Grant, T., 2009, Application of geophysics to Economic Geology, 96(7), 1611–1643. https://doi.org/ gold exploration in Ghana: Examples from Newmont. 10.2113/gsecongeo.96.7.1611 ASEG Extended abstracts 2009: 20th Geophysical Yigit, O., & Hofstra, A. H. (2003). Lithogeochemistry of Conference, pp.1–5, ASEG 20th Geophysical Conference Carlin-type gold mineralization in the Gold Bar district, and Exhibition. Adelaide (CSIRO Publishing). https:// Battle Mountain-Eureka trend, Nevada. Ore Geology doi.org/10.1071/ASEG2009ab089 . Reviews, 22(3–4), 201–224. Vennemann, T. W., & O'Neil, J. R. (1993). A simple and Yigit, O., & Hofstra, A. H. (2003). Lithogeochemistry of inexpensive method of hydrogen isotope and water ana- Carlin-type gold mineralization in the Gold Bar district, lyses of minerals and rocks based on zinc reagent. Battle Mountain-Eureka trend, Nevada. Ore Geology Chemical Geology, 103(1–4), 227–234. Reviews, 22(3–4), 201–224.

Journal

Geology Ecology and Landscapes – Taylor & Francis

Published: Nov 21, 2022

Keywords: Stable isotopes; mineral exploration; geochemistry; gold; Birimian; Ghana

There are no references for this article.