Introduction

Dawsonite, a hydrated sodium aluminum carbonate (NaAlCO3(OH)2), was considered a rare mineral up to the twentieth century (Loughnan and Goldbery 1972). However, today it is regarded as an unusual mineral on Earth's surface, having been identified in several parts of the world. Examples include Argentina (Comerio et al. 2014), Australia (Golab et al. 2006), Belarus (Limantseva et al. 2008), Brazil (Teles et al. 2022), China (Liu et al. 2011; Li et al. 2024), Italy (Wopner and Höcker 1987), Japan (Okuyama 2014), Mongolia (Dong et al. 2011), Poland (Rybak-Ostrowska et al. 2020), Romania (Cseresznyes et al. 2024), Tanzania (Hay and Reeder 1991), United States (Burnham et al. 2015), and Yemen (Worden 2006). Dawsonite belongs to the orthorhombic crystal system and was first identified by Harrington (1875). It is a whitish mineral with a silky luster and a fine fibrous habit (Golab et al. 2006), occurring mainly in an authigenic subsurface context. It is substantially more unstable and therefore rarer on the surface (Saldanha et al. 2023; Cseresznyes et al. 2024).

Mineral dawsonite forms at temperatures between 25 and 200 ºC (Li et al. 2017; Qu et al. 2022), while synthetic dawsonite can be produced between 60 and 180 ºC (Li et al. 2022; Knorpp et al. 2023) under high partial pressure of CO2 (Marinos et al. 2021; Li et al. 2023) and remains stable in alkaline pH environments (Hellevang et al. 2010). Dawsonite is mostly found in rocks at depths between 1000 and 2200 m (Qu et al. 2022), although there are records of its occurrence at depths shallower than 200 m (Limantseva et al. 2008; Comerio et al. 2014 and this work) and at depths greater than 3000 m (Worden 2006). It is mainly found in clastic rocks, accounting for approximately 75% of its occurrence (Qu et al. 2022), including feldspathic (Li and Li 2016) and quartz sandstones (Gao et al. 2009), pyroclastic rocks (Dong et al. 2011) and sedimentary tuffs (Zalba et al. 2011). Dawsonite is also recorded in igneous rocks (Sirbescu and Nabelek 2003), limestones (Goldbery and Loughnan 1977), oil shales (Palayangoda and Nguyen 2015), coal (Ming et al. 2017) and soils (Reynolds et al. 2012).

Dawsonite has recently gained prominence due to its CO2 mineral trapping potential (Hellevang et al. 2005, 2011, 2013; Kaszuba et al. 2011; Lu et al. 2022). Initiatives to mitigate greenhouse gas (GHG) atmospheric concentrations by safely storing GHGs in the subsurface for long periods (Carbon Capture and Storage—CCS) have been gaining notoriety as carbon neutrality and circular economy policies become popular (Nobre et al. 2021, 2022a). Mineral trapping is a type of CCS strategy discussed since the 1990s (Lohuis 1993), involving the injection of CO2 into geological reservoirs with suitable compositions, porosities, permeabilities, fluids, and thermodynamic conditions to cause the precipitation of carbonate mineral phases (e.g., dawsonite), thereby immobilizing the CO2 in the formation (Bachu et al. 1994). Geochemical (Gaus et al. 2005) and CO2 injection models (Johnson et al. 2004) for mineral trapping with dawsonite crystallization have demonstrated higher potential for long-lasting CCS than strategies such as hydrodynamic or dissolution capture (Moore et al. 2005). However, if pCO2 in the formation is not kept high, dawsonite can destabilize and release CO2 (Hellevang et al. 2005; Ketzer et al. 2005; Lu et al. 2022). This indicates that the rock package must meet specific prerequisites to behave as an effective CO2 reservoir.

A good geological reservoir for mineral trapping must present a permeability high enough to allow the mobility and dissemination of CO2 in the subsurface, in addition to high porosity to accommodate a significant volume of gas. Furthermore, the reservoir must not be associated with freshwater aquifers due to the huge importance of this resource for human life (Xu et al. 2004; Lu et al. 2022). Computational models vary in their conclusions but generally indicate that ideal reservoirs are found at depths greater than 800 m, comprising a layer at least 20 m thick and sealed by cap rock at least 10 m thick (Soong et al. 2004; Xu et al. 2005; André et al. 2007; Qu et al. 2022). The Rio Bonito Formation of the Paraná Basin has demonstrated the greatest potential for CCS actions in South America, with packages of porous, quartz, and feldspathic sandstones and thick coal seams and carbonaceous shales with a high organic matter content, which are strategic rocks due to its high CO2 adsorption capacity (Ketzer et al. 2009; Abraham-A and Tassinari 2023; de Oliveira et al. 2023; Abraham-A et al. 2024a; 2024b).

This study unveils the first finding of natural dawsonite within the Rio Bonito Formation, occurring in quartz sandstones sampled from cores associated with wells drilled for coal exploration along the eighties. The dawsonite identification in this formation improves its potential for mineral trapping (CCS), confirming that the Rio Bonito Formation provides the required conditions for dawsonite crystallization. The characterization of dawsonite, its textures, and associated microstructures was carried out using petrographic microscopy, scanning electron microscopy (SEM) with coupled energy dispersive x-ray spectroscopy (EDX) system and x-ray diffraction (XRD).

Geological background

The Rio Bonito Formation is part of the Gondwana I Supersequence (Carboniferous-Lower Triassic) of the Paraná Basin (Milani et al. 2007). The Rio Bonito Formations is part of the transgressive portion of the Permo-Carboniferous transgressive–regressive cycle recorded in this supersequence. It comprises conglomerate, very fine- to very coarse-grained sandstone, claystone, and coal seams, some of which have economic significance. The Rio Bonito Formation contains significant reserves of methane adsorbed in the coal layers, which are preserved due to adequate sealing (Kalkreuth et al. 2008; 2013). Its deposition is related to tidal-dominated fluvial and estuarine environments and wave-dominated shoreface environments (Perinotto and Castro 2000; Lopes and Lavina 2001; Holz 2003; Cagliari et al. 2014; Bicca et al. 2020; Kern et al. 2021).

The Rio Bonito Formation may be up to 350 m thick, with an average thickness exceeding 170 m. Positioned in the central-southern region of the Paraná Basin, it occurs at a target depth exceeding 800 m, meeting the requirements for CO2 storage. It is overlain by the Palermo Formation, which serves as a proposed sealing rock, with a minimum thickness of 20 m and an average thickness surpassing 120 m. The Palermo Formation is composed of fine- to very fine-grained sandstones and siltstones interspersed with thinly laminated shales, mudstones, and occasionally limestones (Ramos et al. 2015; de Oliveira et al. 2023; Abraham-A 2023; 2024a). The Rio Bonito Formation thins out towards the south of the basin. In the Rio Grande do Sul State, where dawsonite was found (Fig. 1), the greatest thicknesses of the formation (ranging from 150 to 200 m) are related to paleo valleys distributed along the basin’s edge. However, these thicknesses may significantly diminish over the basement highs (Ketzer et al. 2003; Jasper et al. 2006).

Fig. 1
figure 1

Geological map of the area where well 5-CA-91-RS was drilled

Full size image

The Well 5-CA-91-RS (Fig. 2 and 3) drilled by the Geological Survey of Brazil (SGB-CPRM) provided the study samples. This well cuts into the Pirambóia (Fig. 1), Rio do Rasto and Palermo formations before reaching the Rio Bonito Formation. Dawsonite was identified in quartz sandstones at depths from 541.20 to 541.05 m (Fig. 3A) and 566.25 to 566.00 m (Fig. 3B), herein referred to as intervals A and B (Fig. 2).

Fig. 2
figure 2

Stratigraphic sequence of well 5-CA-91-RS between 500 and 600 m depth, showing the Palermo and Rio Bonito Formations. A and B represent the dawsonite finding zones. A between 541.20 and 541.05 m, and B between 566.25 and 566.00 m

Full size image
Fig. 3
figure 3

Cores from well 5-CA-91-RS. A Core from 541.20 and 541.05 m and B core from 566.25 to 566.00 m. Both intervals (A and B) contain dawsonite

Full size image

Porosity ranges between 10 and 20% in the sandstones and 1.6 to 4.3% in the coal seams of the Rio Bonito Formation (Milani et al. 2007; Ketzer et al. 2003; Lourenzi and Kalkreuth 2014). In addition to sandstones, coal seams, and carbonaceous shales also have potential for CO2 adsorption due to their relevant contents of organic matter, ranging from 5 to 25% (Lourenzi and Kalkreuth 2014; Costa et al. 2016; Bicca et al. 2020). De Oliveira et al. (2023) conducted stratigraphic correlations between drilling cores in the Paraná Basin, mapping an area of 383,951 km2 where the Rio Bonito Formation fully meets the requirements for onshore mineral trapping.

Material and methods

Sample collection and preparation

The studies were conducted on cores from well 5-CA-91-RS (Fig. 3) provided by the Geological Survey of Brazil (SGB-CPRM) and stored at its headquarters in Caçapava do Sul City, in the central portion of the Rio Grande do Sul State. The 4.85 cm wide well was drilled between 1976 and 1977 in the Gravataí municipality as part of a coal exploration campaign promoted by the Brazilian Ministry of Mines and Energy. The rock samples were repurposed in 2023 to evaluate mineral trapping potential. SGB-CPRM provided 2 kg (equivalent to 0.75 L) of rock samples from the Rio Bonito Formation for characterization tests.

A portion of this material was cut into 4 × 2 × 0.5 cm slivers using a cutting disc to prepare thin sections for microscopy following a method adapted from Pike and Kemp (1996) and Adams et al. (2014). The slivers were immersed in a mixture of 10 g of epoxy resin, 0.5 g of Oracet B® blue dye (to dye the rock's porous blue), and 5 g of Araldite® hardening agent dissolved in 50 mL of hydrated ethanol to liquefy the mixture. This solution was then subjected to a vacuum pump for 24 h to ensure complete percolation throughout the porosity of the quartz sandstones. After drying and subsequent resin hardening, the samples were affixed onto a glass slide and polished until they reached a thickness of 30 µm. The choice of resins was an adaptation of traditional methods to optimize costs and sample preparation time, as observed in adaptations to other sedimentary materials (see Montana 2020; Broekmans et al. 2022). The same thin sections used in petrography were utilized for SEM–EDX analysis. Thin sections for microscopy were prepared at the Geological Thin Section Laboratory of the Universidade do Vale do Rio dos Sinos (UNISINOS).

For XRD analysis, 1 kg of sample was ground to 75 µm (#200 mesh) followed by successive homogenization stages and quartering until 10 g of powder was selected. Given the homogeneous nature of the quartz sandstone, no special care was required to select any specific rock segment for grinding, in accordance with the methods of Waseda et al. (2011) and Ali et al. (2022).

Analytical methods

Petrographic analyses were conducted using a Zeiss AxioLab A1 Microscope with a Zeiss AxioCam MRc camera system from the Fluid Inclusion Lab at UNISINOS.

SEM–EDX analyses were performed at the Technological Institute of Paleoceanography and Climate Change (ITT OCEANEON) at UNISINOS in Zeiss EVO MA 15 electron microscope. The microscope operated at an acceleration voltage (EHT) of 25 kV and a working distance (WD) of 8.5 mm, with a probe current of 8 nA. Samples were gold-coated with a layer thickness of 46 nm. The SEM was coupled with an Oxford Instruments EDX spectrograph featuring an X-Max detector. Analyses were performed over seven interactions with a live time of 180 s each. This analytical technique was employed to generate false-color images for mapping strategic chemical elements (C, O, Na, Al, and Si) to identify dawsonite in quartz sandstone, following the methodology outlined by Gomes (2015).

XRD was performed at ITT OCEANEON at UNISINOS on an Empyrean PANalytical diffractometer with a reflection-transmission configuration, spinning at two revolutions per second, with a goniometric range from 2 to 75° (2θ), a step of 0.0131° with 170 s per step, and a Cu tube operating at 40 kV and 40 mA. Bragg–Brentano HD incident beam geometry was used, with a 0.02 rad Soller slit, a 20 mm fixed mask, a 1/4" fixed anti-scattering slit, and a 1/16" fixed divergent slit. A 7.5 mm anti-scattering slit and a 0.02 rad Soller slit were mounted on the diffracted beam. The diffractometer was equipped with a PIXcel 3DMedipix3 area detector with 255 channels.

Results

Dawsonite (Fig. 4B, 4C and 4E) was initially found in thin sections under polarized light optical microscopy (petrographic microscopy) within quartz sandstones containing grains ranging from 0.2 mm to 1.5 mm (Fig. 4A), alongside carbonate (identified in XRD as dolomite) and muscovite. The sandstone exhibited moderate to well-sorting, with subangular to subrounded grains. Two types of pores, measuring 0.1 to 1.0 mm, were observed: isolated intergranular primary porosity (Fig. 4D) and moldic secondary porosity (Fig. 4G), indicating mineral dissolution during diagenesis. Diagenetic processes include quartz overgrowth (Fig. 4F), carbonate cement deposition (Fig. 4H), and partial dissolution of framework grains, formingwhich form moldic pores where dawsonite precipitated during mesodiagenesis (Fig. 4B-G).

Fig. 4
figure 4

Thin sections of quartz sandstones of the Rio Bonito Formation in the well 5-CA-91-RS. Photos A and C to H correspond to cross-polarized transmitted-light images and B relates to a polarized transmitted-light image. A Quartz (Qz) sandstone with dolomite (Dol), muscovite (Ms), and dawsonite (Dws). B, C, E, and G show dawsonite formed from moldic porosity (Dws mp). D Example of an isolated pore (ip) and dawsonite-filled moldic pore (Dws-fmp). F Quartz overgrowth (Qz-og). H Carbonate cementation (dolomite) in quartz sandstone. Mineral abbreviations follow Whitney and Evans (2010) for Dol, Ms, and Qz, while the nomenclature for Dws follows Warr (2021)

Full size image

Following the identification of dawsonite as thin radiating acicula filling the moldic porosity of the quartz sandstones (Fig. 5A), the sample was analyzed by SEM for compositional imaging using EDX (Fig. 5). In the image captured by secondary and backscattered electrons (respectively Fig. 5A and 5B), the contrast in morphology and average atomic number between the unfilled pore, dawsonite, and quartz-dominated framework is discernible. The simultaneous presence of carbon (Fig. 5C), oxygen (Fig. 5D), sodium (Fig. 5E), and aluminum (Fig. 5F), alongside the complete absence of silicon (Fig. 5G), corroborates the petrographic observations.

Fig. 5
figure 5

SEM of dawsonite grown in moldic porous and its surroundings. C-G are on the same scale shown in B. A Secondary electron image. B Backscattered electron image. C Compositional image of carbon. D Compositional image of oxygen. E, F, and G correspond to compositional images of sodium, aluminum, and silicon, respectively

Full size image

Definitive confirmation of the dawsonite occurred after performing an XRD analysis on the total rock powder (Fig. 6). The XRD results confirmed the presence of dawsonite, identified the carbonate as dolomite, and confirmed the presence of muscovite.

Fig. 6
figure 6

XRD results of powder from total quartz sandstone sample of the Rio Bonito Formation. A The diffractogram of total rock powder. B Interpretation of XRD results: Quartz was identified by the main peak at 2θ = 26.5º, and the secondary peak at 2θ = 21º; dawsonite was diagnosed by its main peak at 2θ = 15.5º, and the secondary peak at 2θ = 32º; muscovite was detected by its main peak at 2θ = 9º, and the secondary peak at 2θ = 26.5º; and dolomite was verified by its main peak at 2θ = 31º, and the secondary peak at 2θ = 41º

Full size image

Final remarks

The literature presents several technological uses for dawsonite, such as in catalysts (Zumbar et al. 2021), fire retardants (Zhang et al. 2024), nanotechnology (Duan et al. 2013), sorbents (Zhao et al. 2020) and water treatment (Li et al. 2020). These applications highlight the dawsonite's potential for advanced applications (Nobre et al. 2022b, 2023). These advanced applications rely on synthetic crystals, as natural dawsonite is not typically abundant or stable enough at the surface to be mined as a commodity. However, its crystallization induced by high pCO2 levels makes dawsonite prominent as a CCS strategy.

The discovery of dawsonite in the Rio Bonito Formation represents a significant advancement in understanding the potential of this unit for CO2 trapping, particularly within quartz sandstones as investigated in this study. This paper marks the first documented occurrence of dawsonite within the Rio Bonito Formation. Dawsonite formed during mesodiagenesis and was always found filling moldic pores, indicating that some primary minerals dissolved in earlier diagenetic stages, creating the necessary chemical conditions for dawsonite formation. Thus, the documented dawsonite is not a primary mineral, aligning with descriptions in the literature that highlight its common authigenic occurrence (Saldanha et al. 2023; Cseresznyes et al. 2024).

In the studied region, the Rio Bonito Formation is closer to the surface than in the Paraná Basin depocenter. In this context, dawsonite was found at depths of 541.20 to 541.05 m and 566.25 to 566.00 m. While these depths are shallow for mineral trapping initiatives, they facilitate sample acquisition, as drilling depths exceeding 800 m are substantially more expensive. Previous studies (Ketzer et al. 2009; Abraham-A and Tassinari 2023; de Oliveira et al. 2023; Abraham-A et al. 2024a; 2024b) have demonstrated the high potential of the Rio Bonito Formation for CO2 storage in the basin's deeper portions. The finding of dawsonite reinforces this potential, as it is the most prominent mineral formed in the CO2 mineral trapping process. It is reasonable to infer that occurrences of dawsonite may exist in other sections of the Rio Bonito Formation, particularly those adjacent to layers of coal and organic matter-rich shales within the same unit. As a result, the discovery of dawsonite in the Rio Bonito Formation reinforces the potential of the Paraná Basin for onshore storage of considerable volumes of CO2 in the future.

The petrophysical characteristics of the Rio Bonito Formation such as porosity, depth, thickness, and the presence of an effective sealing layer, highlights its growing potential for future CCS initiatives. Moreover, the presence of natural dawsonite further enhances this potential.