Stable Isotopic, Micro-FTIR, and Geochemical Characteristics of the Permian Madzaringwe Shale of Tuli Basin, South Africa: Implications for Organic-Rich Shale Provenance

: The paleo-environmental setting of an organic-rich shale remains an essential controlling factor for shale reservoir distribution. The scarcity of generalised data on paleo-environment settings has been spurred using a simple investigative approach to decipher the provenance of organic-rich shale in various regions. This study investigates the organic-rich Madzaringwe shale of the Tuli Basin to reconstruct the provenance of the organic material for shale gas generation potential. Representative shale core samples were analysed for the stable isotopic fractions, functional groups, and major and trace compositions. The carbon isotopic composition, δ 13 C value, ranging from − 21.01 to − 24.0‰, averaging at − 22.4‰. Inference from the stable isotopic compositions and functional group analysis indicate Type-III kerogen prone to gas generation in the studied Madzaringwe shale. The micro-Fourier transformed infrared (micro-FTIR) analysis reveals infrared absorption peaks between 2800 and 3300 cm − 1 wavelengths corresponding to gaseous hydrocarbon. The x-ray ﬂuorescence (XRF) result reveals major elements comprising Al 2 O 3 (29.25–29.11%), CaO (0.29–0.28%), Fe 2 O 3 (1.16–1.09%), K 2 O (0.97–0.98%), MgO (0.13–0.12%), Na 2 O (0.12–0.09%), P 2 O 5 (0.22–0.21%), SiO 2 (52.50–52.30%), and TiO 2 (1.20–1.18%). The major element ratio of Al 2 O 3 /TiO 2 values ≥ 25 indicates felsic and intermediate provenance from a terrigenous paleo-environment. In addition, laser ablation inductively coupled plasma mass spectrometry (LAICP-MS) reveals the trace elements in which elemental proxy of V/(V + Ni) with a value greater than 0.5 represent reducing environments. Furthermore, the geochemical proxies and isotopic compositions have revealed an anoxic paleoenvironment for the non-marine-derived organic matter in the studied carbonaceous shale.


Introduction
The provenance of organic-rich shale is critical to determining a favourable target for shale gas exploration, among other controlling factors. Controlling factors such as thermal maturity are important parameters for organic matter alteration to gas-prone kerogen [1,2]. However, the type and amount of organic matter in a shale vary due to geologic conditions and paleo-depositional environments [3]. Depending on the prevailing geological settings and paleo-climatic overprints, sedimentation of an organic-rich shale may occur in a marine, terrestrial, or transitional setting. Under favourable geologic and thermal conditions, organic matter from each depositional environment has various compositions, resulting in different source rock types.
Stable isotopic fractionations have demonstrated that interactions among elements of carbon, oxygen, and rocks are a significant indicator of a paleo-depositional environment for organic-rich shale rock. The organic materials that form kerogen prone to unconventional hydrocarbon have heterogeneous sources, depending on the paleo-depositional Formation is of interest in this study due to its high estimated thickness varying from 60 to 150 m [15]. The study area is situated between 21°00′00″ S and 22°30′00″ S and longitude 28°28′00″ E and 30°15′00″ E, as shown in Figure 1. The Madzaringwe shale appears to have been largely laid down by meandering rivers flowing from the northwest in the Permian period [15]. The plant materials accumulated in flood basins under cool, reducing conditions, giving rise to coal seams [16]. The Madzaringwe Formation consists of the coal-bearing strata of the Tuli Basin [17]. Although recent studies have reported varied thicknesses of coal seams, no distinct consistency is less than 20 m with six continuous coal seams interbedded with grey to black shale in a rhythmic manner [18,19]. It is not unlikely that the plant-organic-rich shale has potential for natural gas resources because the evolutionary pathway for coal and shale gas has a similar origin but different endpoints.

Materials and Methods
The boreholes penetrating the organic-rich Madzaringwe shale at different depths as shown in Figure 2 were analysed for stable isotopic fractions and FTIR spectra to determine organic and mineral content at various infrared wavelengths and geochemical compositions.
At the same time, the δ 13 C and δ 18 O values of organic-rich shale samples were measured using the Finnigan Delta plus isotope ratio mass spectrometer equipped with Gas Bench II system for water equilibrium. An analytical precision ±0.2‰ was determined by repeated analyses of a standard. The dried bulk samples were dissolved in 99% H3PO4, and the resulting CO2 gas was analysed by the spectrometer [20]. The measured isotopic composition is presented as delta notation (δ) in relation to the Vienna Pee Dee Belemnite (V-PDB) standard and expressed in per mil (‰). Delta notation is defined as δ = [(Rsample − Rstandard)/Rstandard] × 1000, where R is the stable isotope ratio of 13 C/ 12 C or 18   The Madzaringwe shale appears to have been largely laid down by meandering rivers flowing from the northwest in the Permian period [15]. The plant materials accumulated in flood basins under cool, reducing conditions, giving rise to coal seams [16]. The Madzaringwe Formation consists of the coal-bearing strata of the Tuli Basin [17]. Although recent studies have reported varied thicknesses of coal seams, no distinct consistency is less than 20 m with six continuous coal seams interbedded with grey to black shale in a rhythmic manner [18,19]. It is not unlikely that the plant-organic-rich shale has potential for natural gas resources because the evolutionary pathway for coal and shale gas has a similar origin but different endpoints.

Materials and Methods
The boreholes penetrating the organic-rich Madzaringwe shale at different depths as shown in Figure 2 were analysed for stable isotopic fractions and FTIR spectra to determine organic and mineral content at various infrared wavelengths and geochemical compositions.
At the same time, the δ 13 C and δ 18 O values of organic-rich shale samples were measured using the Finnigan Delta plus isotope ratio mass spectrometer equipped with Gas Bench II system for water equilibrium. An analytical precision ±0.2‰ was determined by repeated analyses of a standard. The dried bulk samples were dissolved in 99% H 3 PO 4 , and the resulting CO 2 gas was analysed by the spectrometer [20]. The measured isotopic composition is presented as delta notation (δ) in relation to the Vienna Pee Dee Belemnite (V-PDB) standard and expressed in per mil (‰). Delta notation is defined as δ = [(Rsample − Rstandard)/Rstandard] × 1000, where R is the stable isotope ratio of 13 C/ 12 C or 18 The functional group of the studied organic-rich shale was determined by the Fourier Transform Infrared (FTIR) Alpha Bruker spectrophotometer at the University of Venda, South Africa.
The pulverised samples were placed in the measurement position of the spectrometer and scanned between the minimum and maximum wavelength range of 500 to 4000 cm −1 , The functional group of the studied organic-rich shale was determined by the Fourier Transform Infrared (FTIR) Alpha Bruker spectrophotometer at the University of Venda, South Africa. The pulverised samples were placed in the measurement position of the spectrometer and scanned between the minimum and maximum wavelength range of 500 to 4000 cm −1, respectively. The absorbance spectral depicts the crystalline phases, reflecting their functional group.
The geochemical analyses of the major elements of the shale samples were quantitatively estimated using PANalytical X-ray fluorescence (XRF) at Stellenbosch University Central Analytical facilities unit, Cape Town. It is equipped with 3 kWatt Rhodium (Rh) tube and inserted glass disks containing 1 g calcined sample, 8 g flux composed of 35% alkali borate (LiBO2) and 64.71% lithium tetraborate (Li2B4O7) as oxidant at 1000°C. At this temperature, the weight loss or gain on ignition (LOI) includes the total volatiles content of the rock samples, including the water combined with the lattice of silicate minerals and the gain on ignition related to the oxidation of the rock mostly due to Fe. Pressed powder pellets for XRF analyses were prepared using 3 g of sample powder and 6 g of boric acid as a binder. The mixture was fused into a steel cup and pressed at a pressure of 30 tons in a hydraulic set.
Trace elements of organic-rich samples were analysed using Agilent 7700 Q ICP-MS attached to a high resolution 193 nm Excimer laser ablation system at the Central Analytical Facilities, Stellenbosch University. The pulverised samples were inserted and fixed by using a stainless-steel holder in the ablation chamber. Helium carrier gas transported the ablated material to the ICP-MS at a flow rate of 0.35 L/min, then mixed with argon (0.9 L/min) and nitrogen (0.003 L/min). For trace infusions, two spots of 104µm are ablated on each sample using a frequency of 8 Hz and fluence of ~3.5 J/cm 2 . Inductively couple plasma mass spectrometry (ICP-MS) is optimised for sensitivity and low oxide ratios of less than 0.2% by tuning both the ICP and laser parameters while ablating a line on NIST612. The geochemical analyses of the major elements of the shale samples were quantitatively estimated using PANalytical X-ray fluorescence (XRF) at Stellenbosch University Central Analytical facilities unit, Cape Town. It is equipped with 3 kWatt Rhodium (Rh) tube and inserted glass disks containing 1 g calcined sample, 8 g flux composed of 35% alkali borate (LiBO 2 ) and 64.71% lithium tetraborate (Li 2 B 4 O 7 ) as oxidant at 1000 • C. At this temperature, the weight loss or gain on ignition (LOI) includes the total volatiles content of the rock samples, including the water combined with the lattice of silicate minerals and the gain on ignition related to the oxidation of the rock mostly due to Fe. Pressed powder pellets for XRF analyses were prepared using 3 g of sample powder and 6 g of boric acid as a binder. The mixture was fused into a steel cup and pressed at a pressure of 30 tons in a hydraulic set.
Trace elements of organic-rich samples were analysed using Agilent 7700 Q ICP-MS attached to a high resolution 193 nm Excimer laser ablation system at the Central Analytical Facilities, Stellenbosch University. The pulverised samples were inserted and fixed by using a stainless-steel holder in the ablation chamber. Helium carrier gas transported the ablated material to the ICP-MS at a flow rate of 0.35 L/min, then mixed with argon (0.9 L/min) and nitrogen (0.003 L/min). For trace infusions, two spots of 104µm are ablated on each sample using a frequency of 8 Hz and fluence of~3.5 J/cm 2 . Inductively couple plasma mass spectrometry (ICP-MS) is optimised for sensitivity and low oxide ratios of less than 0.2% by tuning both the ICP and laser parameters while ablating a line on NIST612.

Results
The result of the geochemical compositions of the studied samples showed the presence of major oxides, as shown in Table 1 values exceeded the threshold value of 0.5 and ranged from 0.5 to 0.9, indicating the intermediate provenance consistent with the previous study [21]. The study [21] suggested a threshold limit > 0.60 as mixture of both mafic and felsic input. The trace and rare-earth element distributions (REE) in the studied samples are presented in Table 2. The trace element showed relatively high concentrations in Ba, Zr, Rb, Zn, Sr, V, and Cr with values above 50 ppm, while all other elements are lower. In the studied samples, the highest value of Barium, Ba (513.85-560.88 ppm), with an average of 545.77 ppm, suggests the dissolution of mineral barite in the black shale by action sulphatereducing bacteria [22,23]. Furthermore, the sulphate-reducing and methanogenesis bacteria actions are reflected by the increasing concentration of Pb, averaging 32.75 ppm. The reduction processes were observed in the studied samples but at a low rate indicated by the lower value of Ba and Pb concentration. The concentrations of Zn and Cu of the Madzaringwe showed, but zirconium, Zr (314.58−341.73 ppm) and vanadium showed a higher concentration (150.94-156.69 ppm). This higher concentration of Zr and V suggest an increasing paleo-productivity of organic matter and brackish deposit [24][25][26]. The average geochemical proxies ratio of V/(V + Ni) and V/Ni for all studied samples exceed 0.54 and 2.50 ppm, respectively, suggesting an anoxic condition of the depositional environment [8,27]. In addition, the siderophile and sulphurophile elemental proxies of the U/Th, Ni/Co, and V/(V + Ni) plot shown in Figure 3a,b depict anoxic redox conditions. This environment indicates reducing H 2 S, which is necessary for preserving organic matter during methanogenesis processes [28]. This interpretation is consistent with black shale in North America that showed V/(V + Ni) ≥ 0.54, indicating anoxic environments. ture increase depth due to bond deformation. The aromatic C=C absorption suggests the wavelength band of detrital plant lignin-derivative materials, unlike the amide band prevalent for marine-derivative materials [29]. The peak stretching from around 1000 to 1220 cm −1 indicates the C-O stretching and OH deformation of the carboxylic, COOH groups are exhibited samples suggesting decarboxylation of organic matter. The deformation of the carboxylic group releases CO2, which is recognised for its characteristic strong absorptions around 2000 to 2300 cm −1 [30].  The micro-FTIR spectra of the studied carbonaceous shale depict different organic and mineral content absorbance bands at various diagnostic wavelengths of the infrared spectrum ( Figure 3). Due to C-H stretching vibrations, hydrocarbons show infrared absorption peaks between 2800 and 3300 cm −1 . The aromatic compounds of the organic matter show absorption peaks between 550 to 1200 cm −1 . Most of the studied samples exhibit a strong characteristic absorbance peak around 650 and 900 cm −1 , corresponding to the aromatic phenolic group, C=C-OH. However, the wavelength peak reduces as temperature increase depth due to bond deformation. The aromatic C=C absorption suggests the wavelength band of detrital plant lignin-derivative materials, unlike the amide band prevalent for marine-derivative materials [29]. The peak stretching from around 1000 to 1220 cm −1 indicates the C-O stretching and OH deformation of the carboxylic, COOH groups are exhibited samples suggesting decarboxylation of organic matter. The deformation of the carboxylic group releases CO 2, which is recognised for its characteristic strong absorptions around 2000 to 2300 cm −1 [30].
The carbon and oxygen isotopic compositions of organic materials associated with the carbonaceous shale samples of the Tuli Basin are presented in Table 3. The carbon isotopic composition, δ 13 C, of the organic-rich shales of the Madzaringwe Formation, ranged from values −21.01 to −24.0‰, averaging at −22.4‰. It implies that most carbon isotopic signatures of studied samples reflect provenance from terrigenous plant material since it exceeds the −22‰ value, as shown in Figure 4. Meanwhile, the carbon isotopic composition found with −20‰ values indicates some influx of organic materials from marine settings [31].
This interpretation is consistent with the δ 13 C values of the Whitehill Formation of the main Karoo Basin found with mixed marine and terrigenous organic carbon material. This isotopic discrimination indicates higher land plants with vascular tissues to range from −22.4 to −23.1‰ (Figure 5), contrary to marine plants such as algae which have a lower isotopic signature of −30 to −27‰. The interpretation of terrigenous plant source is consistent with isotopic fractions for carbonaceous Barakar Formation in India, ranging from −23 to −25‰ [32]. the carbonaceous shale samples of the Tuli Basin are presented in Table 3. The carbon isotopic composition, δ 13 C, of the organic-rich shales of the Madzaringwe Formation, ranged from values −21.01 to −24.0‰, averaging at −22.4‰. It implies that most carbon isotopic signatures of studied samples reflect provenance from terrigenous plant material since it exceeds the −22‰ value, as shown in Figure 4. Meanwhile, the carbon isotopic composition found with −20‰ values indicates some influx of organic materials from marine settings [31].  This interpretation is consistent with the δ 13 C values of the Whitehill Formation of the main Karoo Basin found with mixed marine and terrigenous organic carbon material. This isotopic discrimination indicates higher land plants with vascular tissues to range from −22.4 to −23.1‰ (Figure 5), contrary to marine plants such as algae which have a lower isotopic signature of −30 to −27‰. The interpretation of terrigenous plant source is consistent with isotopic fractions for carbonaceous Barakar Formation in India, ranging from −23 to −25‰ [32].
In contrast, marine plants such as phytoplankton contribute carbon enrichment in a typical euxinic depositional environment with an average δ 13 C value of around −30‰ [33]. Notably, the positive excursion of organic carbon isotopes δ 13 C suggests local anoxia deposited about 280 Ma Permian period instead of the large-magnitude negative carbon isotope excursion close to the Ediacaran-Cambrian period [34]. During this period, a prolonged high burial rate of organic material led to global δ 13 C enrichment in surface water, resulting in isotopically light CO2 in basins [34]. In contrast, marine plants such as phytoplankton contribute carbon enrichment in a typical euxinic depositional environment with an average δ 13 C value of around −30‰ [33]. Notably, the positive excursion of organic carbon isotopes δ 13 C suggests local anoxia deposited about 280 Ma Permian period instead of the large-magnitude negative carbon isotope excursion close to the Ediacaran-Cambrian period [34]. During this period, a prolonged high burial rate of organic material led to global δ 13 C enrichment in surface water, resulting in isotopically light CO 2 in basins [34].  [36]. The negative correlation of δ 18 O relate with degradation of organic matter content reflected in loss of ignition. The loss of ignition suggests combustion of the carbon content of land-derived organic matter, which is rich in vascular plant materials. Such inference is consistent with the carbon isotopic compositions of the studied Madzaringwe shales, which show Type-III kerogen [35,37] in Figure 6. This interpretation agrees with the terrigenous source depicted by the plot of δ 18 O vs δ 13 C values. In contrast, the observed carbon isotopic composition of alum shale is rich in vascular plant materials. Such inference is consistent with the carbon isotopic compositions of the studied Madzaringwe shales, which show Type-III kerogen [35,37] in Figure 6. This interpretation agrees with the terrigenous source depicted by the plot of δ 18 O vs δ 13 C values. In contrast, the observed carbon isotopic composition of alum shale in southern Scandinavia is typical of marine sources varying from −30 to −27‰, which mainly belong to the Type-I kerogen domain [31].
Meanwhile, isotopic composition, δ 18 O, tends to increase downward from 485 m to 580 m with intermittent anomalous depleted values. The increasing δ 18 O values are attributed to a significant influx of meteoric water, while the anomalously depleted oxygen values show a decrease in pore-water δ 18 O values during diagenesis of mineral-water interactions in a closed chemical microsystem of shale [36]. The negative correlation of δ 18 O relate with degradation of organic matter content reflected in loss of ignition. The loss of ignition suggests combustion of the carbon content of land-derived organic matter, which is rich in vascular plant materials. Such inference is consistent with the carbon isotopic compositions of the studied Madzaringwe shales, which show Type-III kerogen [35,37] in Figure 6. This interpretation agrees with the terrigenous source depicted by the plot of δ 18 O vs δ 13 C values. In contrast, the observed carbon isotopic composition of alum shale in southern Scandinavia is typical of marine sources varying from −30 to −27‰, which mainly belong to the Type-I kerogen domain [31].    (Figure 7) showed that the studied samples had both felsic and intermediate igneous provenances, thus explaining the higher values of alumina and silica [38].
Elemental proxies of trace elements have proven to be veritable tools for deciphering and discriminating paleo environments under which sedimentary deposits were formed due to their geochemical resilience to weathering, diagenesis, and sensitivity to redox reaction. Vanadium and Nickel exhibit similar geochemical behaviour and are considered to be highly stable organophilic and sulphur-forming elements [27,39]. The elemental ratio of V/(V + Ni) with a value greater than 0.5 has been interpreted to represent a reducing environment, with less than 0.5 as an oxidising environment [40,41].
Under this oxic condition, elements such as V, Ni, Fe, Mn, Mo, and sulphate are normally present in soluble species in the form of V 5+ , Ni 2+ , Fe 3+ , Mn 4+ , Mo 6+ , and SO 4 6+ , respectively, in the water column. At the same time, the aerobic degradation of organic matter continues until the dissolved oxygen is less than 0.2 mL/L H 2 O [42]. Deficiency of oxygen has been adduced to climate, sea-level changes, tectonic activities, and hydrographic factors, resulting in insufficient oxygen supply to oxidise the organic  [43,44]. Below this oxygen concentration, anaerobic condition and degradation of organic matter begin and reduce the oxidising elements to insoluble species of V 4+ , Ni, Fe 2+ , Mn 2+ , thiomolybdate oxyanion (MoO 4 2− ) and SO 4 2+ , which eventually concentrate in shale sediments [27].
The oxides of Al and Ti elements have been used as a veritable tool to interpret deposited sediments due to their immobility nature during transportation from weathered source rock (McLennan et al., 2006, McLennan, 2001, Hayashi et al., 1997). The elemental proxies ratio of Al2O3/TiO2 values ≥ 25 indicates sedimentary and intermediate igneous sources. The Index of Compositional Variation (ICV) of the carbonaceous shale samples has an average value of <0.1, which is comparable with the Paleogene-Neogene black shales of Arang, Nyalau Tanjong Balingian Bergih Liang Formations in Malaysia enriched with kerogen capable of generating gas (Baiyegunhi et al., 2018). The elemental ratio K2O/Al2O3 (0.03) of the studied samples showed values ≤ 0.5, which suggests a highly mature shale since a significant amount of Al2O3 is typical of immature sediments (Cox et al., 1995). The plot (Figure 7) showed that the studied samples had both felsic and intermediate igneous provenances, thus explaining the higher values of alumina and silica [38]. Elemental proxies of trace elements have proven to be veritable tools for deciphering and discriminating paleo environments under which sedimentary deposits were formed due to their geochemical resilience to weathering, diagenesis, and sensitivity to redox reaction. Vanadium and Nickel exhibit similar geochemical behaviour and are considered to be highly stable organophilic and sulphur-forming elements [27,39]. The elemental ratio of V/(V + Ni) with a value greater than 0.5 has been interpreted to represent a reducing environment, with less than 0.5 as an oxidising environment [40,41].
Under this oxic condition, elements such as V, Ni, Fe, Mn, Mo, and sulphate are normally present in soluble species in the form of V 5+ , Ni 2+ , Fe 3+ , Mn 4+ , Mo 6+ , and SO4 6+ , respectively, in the water column. At the same time, the aerobic degradation of organic matter continues until the dissolved oxygen is less than 0.2 mL/L H2O [42]. Deficiency of oxygen has been adduced to climate, sea-level changes, tectonic activities, and hydrographic factors, resulting in insufficient oxygen supply to oxidise the organic materials [43,44]. Below this oxygen concentration, anaerobic condition and degradation of organic matter begin and reduce the oxidising elements to insoluble species of V 4+ , Ni, Fe 2+ , Mn 2+ , thiomolybdate oxyanion (MoO4 2− ) and SO4 2+ , which eventually concentrate in shale sediments [27].
The anoxic setting is established as all the oxidants in the shale sediments deplete due to further reduction processes relating to burial depth, thus promoting the activities The anoxic setting is established as all the oxidants in the shale sediments deplete due to further reduction processes relating to burial depth, thus promoting the activities of sulphide-reducing bacteria [45]. Studies [46][47][48][49] highlighted compounds such as hydrogen sulphide and zinc sulphide and iron sulphide as evidence of the reduction process ensued from sulphide-driven bacteria. With further increase in burial depth below the sedimentwater interface, the methanogens begin to degrade the organic matter to generate biogenic methane. This explanation is consistent with the previous studies [27,[50][51][52][53] which argued that methane is unlikely to form unless all sulphates are depleted while vanadium and nickel are concentrated.

Hydrothermal Activity and Paleo-Productivity
Some studies [54,55] have indicated that high amounts of Zn in shale rocks may be related to syn or post-depositional hydrothermal overprint. At the same time, elements such as Ni, Co, and Cu have been used to indicate hydrothermal overprint in shale formation during the initial stage of deposition [40,56,57]. In recent work [27], the diagram of Zn-Ni-Co has been used to determine the influence of hydrothermal activities on hydrocarbon sources (Figure 8). The average concentrations of elements Zn, Ni, Co, and Cu for the Madzaringwe shale are relatively higher, suggesting heating at depth and re-emergence at upper sedimentation. This slight hydrothermal influence might have emanated from the geothermal spring's temperatures ranging from 40-100 • C [58]. In Figure 7, the plot of the Zn, Ni, and Co thermometry elements shows heat influence from hydrothermal overprint. It is unlikely that the hydrothermal fluid could have permeated since shale has extremely low permeability and porosity [59]. Still, convection can transfer heat [60].
for the Madzaringwe shale are relatively higher, suggesting heating at depth and re-emergence at upper sedimentation. This slight hydrothermal influence might have emanated from the geothermal spring's temperatures ranging from 40 °C-100 °C [58]. In Figure 7, the plot of the Zn, Ni, and Co thermometry elements shows heat influence from hydrothermal overprint. It is unlikely that the hydrothermal fluid could have permeated since shale has extremely low permeability and porosity [59]. Still, convection can transfer heat [60]. Ba as a paleo-productivity indicator in sediments exposed to hydrothermal overprint seems unviable due to their susceptibility to alteration. However, the Ba concentration of the studied shales is relatively high, suggesting a non-marine environment since it exceeds the 400 ppm threshold for the marine environment in the Madzaringwe shale [61,62]. Implicitly, high detrital influx with low marine paleo-productivity characterises the studied shale, and this interpretation is consistent with the previous study [62], suggesting accumulation of land-derived plant material. To further corroborate the paleo-productivity condition, the Sr/Rb elemental proxy ratio in shale is mostly considered a veritable tool to interpret paleo-environmental settings due to their higher anionic radii and charges [63].
Following the Sr/Rb proxy, the studied shale samples indicate a significant amount of organic matter production from brackish water rather than marine, since brackish water represents a value greater than 0.6, while marine water is less than 0.6 [63]. Ba as a paleo-productivity indicator in sediments exposed to hydrothermal overprint seems unviable due to their susceptibility to alteration. However, the Ba concentration of the studied shales is relatively high, suggesting a non-marine environment since it exceeds the 400 ppm threshold for the marine environment in the Madzaringwe shale [61,62]. Implicitly, high detrital influx with low marine paleo-productivity characterises the studied shale, and this interpretation is consistent with the previous study [62], suggesting accumulation of land-derived plant material. To further corroborate the paleo-productivity condition, the Sr/Rb elemental proxy ratio in shale is mostly considered a veritable tool to interpret paleo-environmental settings due to their higher anionic radii and charges [63].
Following the Sr/Rb proxy, the studied shale samples indicate a significant amount of organic matter production from brackish water rather than marine, since brackish water represents a value greater than 0.6, while marine water is less than 0.6 [63].

Aliphatic Hydrocarbon Potentials
Due to C-H stretching vibrations, hydrocarbons show infrared absorption peaks between 2800 and 3300 cm −1 . The sp 2 C-H hybridisation found with an absorption peak between 3000-3100 cm −1 wavelength indicates the gaseous hydrocarbon. Some studies [64,65] reported the gaseous hydrocarbon of CH 4 has infrared absorbance peaks between 2850 and 3000 cm −1 due to aliphatic C-H stretching of methyl and methylene vibration, while the low absorption peak around 3500 cm −1 corresponds to water desorption wavelength [66]. In aromatic hydrocarbon, the carbon chain voids act as trapping and binding sites for organic lipids and inorganic materials such as hydrous oxides and clay minerals [67]. At greater burial depths around 510 and 580 m, the spectra showed inconspicuous peaks of the carboxylic group suggesting the maturation of organic matter beyond the methanogenesis window to coalification.

Stable Isotopic Signatures of an Anoxic Depositional Environment
The relative amounts of carbon and oxygen isotopic compositions of organic matter are a function of paleo-environmental deposition, which is significant to oil and gas exploration. In the studied carbonaceous Madzaringwe shale, the relationship between δ 13 C and δ 18 O of organic material suggests a bacterial reduction of organic compounds to CO 2 and hydrogen through an enzymatic reaction in the absence of sulphate. The significant role of reducing-bacterial involves converting the CO 2 and degradation of organic matter into new biomass. This reduction process increases the total organic carbon contents of sedimentary rocks through a biogenic origin [68,69]. Like the Madzaringwe shale, the Santa Barbara Basin in the United States has a biogenic source from a burial depth of less than 1000 m. This basin has been reported to have high total organic carbon contents that resulted from a biogenic mediation of chemoautotrophic and Beggiatoa bacteria activities [70]. Furthermore, an organic isotopic study of the Qiongzhusi Formation in the Sichuan Basin revealed a high degree of total organic content generated by microbial bacteria [71]. Furthermore, the oxygen isotopic composition δ 18 O of the studied organic matter has revealed a depleted oxygen value from 16.15‰ to 9.82‰, indicating a mixture of organic water with hydrothermal-meteoric water conditions. This depleted oxygen favours the abundance of organic matter, indicating the redox condition of the water body from which the shale was deposited because organic matter can be preserved well in a strongly reducing environment.

Comparison of the Permian Madzaringwe Shale with Selected Permian Southern Gondwana Basin
The Madzaringwe shale of the Tuli Basin is geographically positioned in the southern Gondwana supercontinent having similar age as the Irati Formation in Parana Basin, Brazil [72,73], Bowen Formation in Australia [74,75], and Whitehill Formation in the main Karoo group, South Africa [76][77][78] during the Permian period ( Figure 9). Recently, the age of the shale gas-bearing Irati Formation has been constrained using U-Pb zircon, indicating 275.75 ± 0.29 Ma corresponding to the Permian period [79]. The rapid evolutionary changes in land plants of Gondwana are attributable to the dramatic climatic change after the deglaciation of the Carboniferous Tshidzi fluvioglacial of the Tuli Basin [15], Dywka group of Karoo group [80,81], and Itararé sub-group of Paraná Basin [82,83]. On the other hand, the Tuli Basin differs from the mentioned southern Gondwana Basin based on nonmarine sedimentation. The preservation of carbonaceous shale and coal seams is associated with a humid climate under autochthonous reducing conditions and deposited by fluvial activity in the depozones. The lithologies of the Tuli and the Gondwana Basins generally consist of diamictite with a clast ranging up to 2 m in diameter embedded in an argillaceous and sandy matrix [15,75,82] underlying the shale gas-bearing formations. Plant fossils-"Glossopteris Flora" characterised the southern Gondwana Permian Basin of the Ecca group [16], Parana, Bowen [84], Beacon [85,86], and Tuli Basin [15]. Although evaluation of the Madzaringwe shale gas potential is in progress in South Africa, the Permian Whitehill Formation (South Africa) has a technically recoverable shale gas resource of 13-23 trillion cubic feet (Tcf) [76,77,87]. Based on the total organic carbon (TOC), which ranged from 2.57 wt%-6.90 wt%, the Whitehill Formation can be considered a good to excellent shale gas potential source. However, the source rock is thermally overmatured because the vitrinite reflectance exceeds 3% Ro, but still capable of generating shale gas such as the Marcellus shale of the Appalachian Basin [88,89]. In Australia, most shales and coals within the Bowen Basin have source rock potential [75] as well as the Irati and Rio Bonito Formation in Brazil, which host shale and coal source rock, respectively. Furthermore, the gas potential in the Taroom and Denison Troughs is reported to be dominated by Type III kerogen, which is gas-prone [90].

Conclusions
Stable carbon and oxygen isotopes, functional groups and geochemical compositions have been a veritable tool to decipher the provenance of organic-rich shale with implications for potential gas generation. Organic rich shale emanating from a terrigenous anoxic paleo-environment indicates Type-III kerogen under sufficient thermal maturity. Terrigenous organic materials were preserved in a reducing condition, having a significant amount of organic matter production from brackish water. The stable isotopic and functional group of organic rich shale reveal the predominance of land-derived vascular plant materials. Meanwhile, the infrared spectra of the studied carbonaceous shale depict characteristic absorbance wavelengths corresponding to the aliphatic hydrocarbon functional group.  Plant fossils-"Glossopteris Flora" characterised the southern Gondwana Permian Basin of the Ecca group [16], Parana, Bowen [84], Beacon [85,86], and Tuli Basin [15]. Although evaluation of the Madzaringwe shale gas potential is in progress in South Africa, the Permian Whitehill Formation (South Africa) has a technically recoverable shale gas resource of 13-23 trillion cubic feet (Tcf) [76,77,87]. Based on the total organic carbon (TOC), which ranged from 2.57-6.90 wt%, the Whitehill Formation can be considered a good to excellent shale gas potential source. However, the source rock is thermally overmatured because the vitrinite reflectance exceeds 3% R o , but still capable of generating shale gas such as the Marcellus shale of the Appalachian Basin [88,89]. In Australia, most shales and coals within the Bowen Basin have source rock potential [75] as well as the Irati and Rio Bonito Formation in Brazil, which host shale and coal source rock, respectively. Furthermore, the gas potential in the Taroom and Denison Troughs is reported to be dominated by Type III kerogen, which is gas-prone [90].

Conclusions
Stable carbon and oxygen isotopes, functional groups and geochemical compositions have been a veritable tool to decipher the provenance of organic-rich shale with implications for potential gas generation. Organic rich shale emanating from a terrigenous anoxic paleoenvironment indicates Type-III kerogen under sufficient thermal maturity. Terrigenous organic materials were preserved in a reducing condition, having a significant amount of organic matter production from brackish water. The stable isotopic and functional group of organic rich shale reveal the predominance of land-derived vascular plant materials. Meanwhile, the infrared spectra of the studied carbonaceous shale depict characteristic absorbance wavelengths corresponding to the aliphatic hydrocarbon functional group.