The Origin and Enrichment of Sedimentary Basin Lithium Brines: A Case Study From the Upper Devonian Leduc Formation, Alberta Basin, Canada Open Access

Global lithium (Li) demand tripled between 2017 and 2022 (International Energy Agency, 2023). Approximately 80% of Li chemicals produced are used in batteries pivotal to electric vehicles and low-carbon energy grid storage systems (U.S. Geological Survey, 2023). Lithium is mined from two main ore types: hard-rock pegmatites (e.g., Bikita, Zimbabwe; Goodenough et al., 2025) and closed-basin lithium brines (e.g., the Salar de Atacama, Chile; Munk et al., 2025). The United States and Canada are net importers of Li chemicals due to the limited domestic occurrences of traditional Li deposits (U.S. Geological Survey, 2023; Government of Canada, 2024). Establishing a domestic supply chain of Li is critical to national security (White House, 2022). In response, billions of dollars are being invested in the development of North American Li resources (e.g., Lithium Americas, 2024). Sedimentary basin brines (sensu Munk et al., 2025) are quickly emerging as a readily available and more environmentally friendly domestic Li resource type (Dugamin et al., 2021, 2023; Marza et al., 2024). Although significantly lower in grade than traditional hard-rock and closed-basin brine deposits, sedimentary basin brines are present in large volumes. Key sedimentary basin brines are estimated to be similar in Li resource scale to closed-basin brines projects in North and South America (Table 1; Dugamin et al., 2021). With several promising sedimentary basin brine resources across the United States and Canada (e.g., the Devonian Leduc Formation, Alberta basin; Jurassic Smackover Formation, Gulf Coast basin; Devonian Marcellus Shale, Appalachian basin; Devonian-Mississippian Duperow and Bakken Formations, Williston basin), it is imperative to exploration and exploitation of these resources that we understand how they form and where they are likely to occur. Fundamental outstanding questions on the generation of these potential Li resources persist: (1) Is lithium concentration linked to brine genesis? (2) What are the sources of lithium to sedimentary basin formation waters? (3) How and when does lithium become concentrated in these systems?

The Upper Devonian Leduc Formation brines (Alberta basin, Canada) host >16 million metric tonnes (Mt) measured and indicated lithium carbonate equivalent (Table 1; E3 Lithium, 2024c). This is one of just a few North American projects in the field pilot phase of Li production from hydrocarbon-related brines by direct lithium extraction (e.g., E3 Lithium, 2024c; Standard Lithium, 2024). Elevated lithium concentrations (>40 ppm) in the Leduc brines have been documented since the early 1980s (Hitchon, 1984; Connolly et al., 1990a, b; Hitchon et al., 1993; Underschultz et al., 1994; Bachu, 1995). Conflicting hypotheses regarding the origin of these brines and the sources of Li to the basin have confounded a fundamental understanding of the Li resource scale and grade. Previous workers have suggested the brines formed from evapoconcentrated seawater (Hitchon and Friedmann, 1969; Hitchon et al., 1971; Spencer, 1987; Connolly et al., 1990a, b; Eccles and Berhane, 2011), the dissolution of Devonian evaporites (Michael et al., 2003; Huff, 2016, 2019), or crustal hydrothermal fluids introduced along faults (Machel et al., 1996; Machel and Cavell, 1999; Buschkuehle and Machel, 2002; Huff et al., 2019; Stacey et al., 2020). These brines may have been subsequently modified by diagenesis; rock-water interactions; and mixing with freshwater, seawater, or relict brines (Billings et al., 1969; Spencer, 1987; Michael and Bachu, 2002; Michael et al., 2003; Gupta et al., 2012). The source of Li has been attributed to subsurface rock-water interactions with Cambrian-Precambrian metasediments and siliciclastic rocks (Eccles and Berhane, 2011), Li-enriched seawater and marine evaporite minerals (Huff, 2016, 2019), and crustal hydrothermal fault-derived fluids (Huff, 2016, 2019). To unravel brine generation and migration history and explain the spatial distribution of Li concentration across the Leduc Formation brines, we integrate new (85 samples) and historical (57 samples) brine sample geochemistry, including major and minor cations and anions, and key trace-metal concentrations and multiple isotopic systems (oxygen, hydrogen, radiogenic strontium, and lithium).

The Alberta basin is situated within the greater Western Canada sedimentary basin (Fig. 1A). It is bordered to the northwest by the Peace River arch, to the southeast by the Sweetgrass-Bow Island arch, and to the southwest by the Canadian Rocky Mountains. The Devonian units comprise a passive margin sequence deposited in a partially restricted epeiric seaway (Wendte, 1994). These units were subsequently deformed and buried during several orogenic stages, including the Devonian-Mississippian Antler orogeny, the Jurassic-Lower Cretaceous Columbian orogeny, and the Upper Cretaceous-Paleogene Laramide orogeny (Price, 1990; Root, 2001). Hydrocarbon generation was linked to Laramide deformation and the deepest burial stage (Creaney et al., 1994). Petroleum production from the Leduc Formation began in 1957 and continues today (Atchley et al., 2006).

The Upper Devonian Leduc Formation (~180–300 m thick) comprises a series of dolomitized reef complexes across central and southern Alberta, Canada (Fig. 1A; Drivet and Mountjoy, 1997). In the northwest, the Leduc reefs built conformably on older reef systems of the Middle-Upper Devonian Beaverhill Lake Group and unconformably on the Cambrian-Precambrian units of the emergent Peace River arch (Fig. 1B). In the east and south, the Leduc reef complexes were constructed over the Upper Devonian Cooking Lake carbonate platform. The Leduc reefs are encased by the heterogeneous Upper Devonian Ireton and Duvernay Formation shales across the West Shale and East Shale basins (Knapp et al., 2017, 2019; Venieri et al., 2021). Dolomitization produced significant vuggy and moldic porosity, enhancing the permeability and porosity of the Leduc reservoirs (Drivet and Mountjoy, 1997). Basin-scale reflux dolomitization occurred during early burial catalyzed by evolved Devonian seawater. Subsequent saddle dolomite cementation is linked to Laramide deformation and fluids introduced along faults that variably interacted with deeper basinal and crustal units (Stacey et al., 2020).

The Alberta basin has been the focus of numerous geochemical and hydrodynamical modeling studies (Garven, 1985, 1989; Bekele et al., 2002; Adams et al., 2004; Gupta et al., 2012, 2015) that are focused on understanding the origin of brines and resource development. Of particular relevance to the Leduc Li resource are models that simulate brine genesis and transport times. Gupta et al. (2012) in a review of the chloride and bromide compositions of brines in the Alberta basin develop 2-D flow and transport models of basin evolution. They used available geology and hydrostratigraphic information to constrain numerical models and make predictions of brine concentration evolution in the basin. Their work challenged the single brine evolution model that implied the preservation of in situ brines for over hundreds of millions of years (i.e., very long fluid residence times). The Gupta et al. (2012) models simulated conditions that required a four-end-member hydrogeochemical system, including primary and secondary brines and seawater and freshwater infiltration due to basin uplift and hydrodynamic forcing. The physically based hydrologeologic model showed that dissolution of evaporite units is an important control on basin salinity distribution. In a follow-up paper using the same modeling framework, Gupta et al. (2015) established that the age of the brines in the basin varied considerably in stratigraphic units and basin position. They show that no strong correlation exists between brine formation time and groundwater age due to dispersive mixing and diffusion. Together these studies have several implications for the origin of Li brines and their concentrations. First, since the brines have multiple origins, it follows that the Li is derived from multiple sources. Li concentrations in halite, seawater, and evaporated brines are likely to be distinct. Freshwater inflow will decrease the Li concentrations in the mixed fluid. Further, the strong differences in the age of the brines will impact the time scale of water-rock interaction, perhaps modifying the observed present-day Li concentrations.

Variable terminology has been used in the literature in reference to Li-enriched waters associated with hydrocarbon reservoirs; these include “confined saline aquifer” brines, “lithium-rich groundwaters,” “sedimentary formation waters,” “basinal brines,” “basinal (saline) fluids,” “oil field waters,” and “oil field brines” (e.g., Macpherson, 2015; Dugamin et al., 2021, 2023; Pugh et al., 2023; Darvari et al., 2024; Marza et al., 2024). We adopt the classification scheme of Munk et al. (2025) and use “sedimentary basin lithium brines.”

Since 2017, E3 Lithium has sampled and analyzed produced water obtained from third-party operating wells. The company has implemented internal best practices for measuring Li from the wellhead sampling and for laboratory analytical procedures. The Leduc Formation brines are especially complicated to measure, with high total dissolved solids (56,000–290,000 mg/L; avg 190,000 mg/L) content compared to low Li concentrations (4–140 mg/L; avg 62 mg/L; App. Table A1). Methodologies to unmask the Li from the surrounding matrix require dedicated instrument calibration techniques to measure Li accurately and consistently. The two instruments most used to measure Li in brines are inductively coupled plasma-optical emission spectroscopy (ICP-OES) and inductively coupled plasma-mass spectrometry (ICP-MS). After removal of H₂S gas from the brines, the brines are prepped for geochemical analyses by undergoing a digestion step where an acid is added to the brine to dissolve potential solids. A methodology is then set up to match the Li detection. E3 Lithium uses an ICP-OES where wavelet, calibration curve range, and dilution factors are all important criteria for attaining consistency in Li measurements from the brines (E3 Lithium, pers. commun., 2024). Much of the publicly available Li concentration data from the Alberta basin has been measured by industrial labs using ICP-OES or ICP-MS instruments and is reported in milligrams per liter. For a regional perspective on brine heterogeneity, we present the published brine data (Alberta Energy Regulator databases: Eccles and Jean, 2010; Huff et al., 2011, 2012, 2019) compiled and screened for a ± 10% cation-anion charge balance cutoff together with E3 Lithium’s internal brine analytical results.

New major cation, anion, and trace-metal concentration data were procured by E3 Lithium and provided to the authors. These measurements were conducted at SGS Canada, Lakefield, Ontario, on ICP-OES and ICP-MS instruments. Brine lithium isotope measurements were provided to the authors by E3 Lithium and made by ALS Scandinavia (Aurorum, Sweden) on a ThermoScientific Neptune Plus multicollector (MC)-ICP-MS. The data provided by E3 Lithium are summarized in Appendix Table A1. Brine samples were analyzed for δ2H and δ18O on a Picarro L-1102i WS-CRDS analyzer in the ENRI Stable Isotope Laboratory at the University of Alaska, Anchorage. Analytical results can be found in Appendix 1 (App. Table A1). Brine ⁸⁷Sr/⁸⁶Sr ratios were measured at Brown University on a ThermoScientific Neptune Plus MC-ICP-MS; analytical results can be found in Appendix Table A2.

Published geochemical data sets (n = 57 samples) are compared with new Leduc brine analyses (n = 85 samples) that include lithium concentrations (n = 141; n = 85 this study), δD and δ¹⁸O signatures (n = 41; n = 10 this study), ⁸⁷Sr/⁸⁶Sr ratios (n = 41; n = 10 this study), δ⁷Li values (n = 10 this study), and major cation and anion and key trace-metal concentrations (n = 113; n = 83 this study).

Lithium concentrations are variable across the Leduc reef complexes (5–140 mg/L; Fig. 2), clustering in geographical groups corresponding to brines from the northwestern (Fig. 1), northeastern, and southeastern regions. The northwestern brines have the highest average Li concentrations (91 mg/L, range of 40.5–130 mg/L), the southeastern brines have on average 74 mg/L Li (17–140 mg/L), and the northeastern brines show the lowest Li concentrations (29 mg/L, range of 5–50 mg/L). The highest Li concentrations (97–140 mg/L) are noted in Eccles and Jean (2010), the oldest public data set, which did not also report the paired cation-anion concentrations (Fig. 2). Lithium concentrations show strong positive correlation with total dissolved solids (TDS), depth, K, δ¹⁸O, Ca, Sr, and Cl and weak to no correlation with Mg, Na, Br, δ⁷Li, and ⁸⁷Sr/⁸⁶Sr (Fig. 3).

The Leduc brines cluster into three distinct δD versus δ¹⁸O compositional groups: The northwestern brines are –44 to –42‰ δD and –0.1 δ¹⁸O to 4.8‰ δ¹⁸O, the northeastern brines are –104 to –54.3‰ δD and –7.2 δ¹⁸O to 2.4‰ δ¹⁸O values, and the southeastern brines are –68 to –49.8‰ δD and 1.7 to 13‰ δ¹⁸O (Fig. 4). All brines plot below modern seawater (Vienna standard mean ocean water, VSMOW), below Devonian seawater, and along the expected compositional path of evaporatively concentrated waters. The northwestern and southeastern brines are enriched in δ¹⁸O with respect to VSMOW and Devonian seawater composition.

The Leduc brines cluster into two distinct groups of more radiogenic (0.72244–0.72898; northwestern brines) and less radiogenic (0.70869–0.71040; northeastern and southeastern brines) ⁸⁷Sr/⁸⁶Sr isotope composition (Fig. 5A). The less radiogenic brines can be further subdivided, with the northeastern brines showing a broad range of ⁸⁷Sr/⁸⁶Sr values (0.70869–0.71040; Fig. 5B) and relatively low Sr concentrations (104–1,190 mg/L, avg 418 mg/L) and the southeastern brines have a narrower range of ⁸⁷Sr/⁸⁶Sr values (0.70872–0.70885) and a broad range of Sr concentrations (528–1,380 mg/L, avg 924 mg/L). All brine values show higher ⁸⁷Sr/⁸⁶Sr ratios than Devonian seawater composition (0.7078–0.7090; Prokoph et al., 2008) and the Middle Devonian Prairie evaporite (0.70781–0.70789; Horita et al., 1996). All brines are significantly less radiogenic than the underlying Cambrian sediments and Precambrian basement units (0.74307–1.3097; Connolly et al., 1990b). The northwestern brines ⁸⁷Sr/⁸⁶Sr values overlap the late-stage calcite and dolomite cements (0.7103–0.7323; Machel and Cavell, 1999; Buschkuehle and Machel, 2002), whereas the northeastern and southeastern brines overlap with the ⁸⁷Sr/⁸⁶Sr value range of matrix dolomites and basinal shales (0.7079–0.7120; Amthor et al., 1993; Machel et al., 1996; Machel and Cavell, 1999; Green and Mountjoy, 2005; Buschkuehle and Machel, 2002).

Lithium isotope values of the Leduc brines range from 14 to 16‰ δ⁷Li (avg 14.76‰; Fig. 6), significantly lower than modern seawater (31‰; Misra and Froelich, 2012), overlapping Devonian seawater composition derived from carbonates (avg carbonate values of 16 ± 12‰; Kalderon-Asael et al., 2021; Wang et al., 2023), and higher than global marine shales (avg –0.2‰; Macpherson et al., 2014).

Sedimentary formation waters are commonly “brines,” characterized as having >100,000 mg/L TDS. Basinal brines can originate from, and be a mixture of, evaporatively concentrated seawater, dissolution of evaporites, groundwater, meteoric water, seawater, and hydrothermal or crustal fluids (Carpenter, 1978). The deconvolution of brine generation and evolution is nontrivial and requires a detailed understanding of the brine geochemistry, tectonic history, diagenesis, rock-water interactions, and basin-scale hydrodynamics. We document distinct brine geochemistry across the Leduc reef complexes (northwestern, northeastern, and southeastern brines; Fig. 1A) corresponding to variable brine origin and postformational modification processes.

The δD versus δ¹⁸O values of the Leduc brines show an evaporative trend that could reflect evaporatively concentrated Devonian seawater or the dissolution of Devonian marine halite (Fig. 4). The Cl/Br ratio is commonly used to distinguish between brines originating from evapoconcentrated seawater or “primary brines” versus halite dissolution or “secondary brines” (Carpenter, 1978). The Cl/Br ratio of progressively evaporated seawater is constant until reaching halite saturation when Cl consumption is initiated, thereby lowering the Cl/Br ratio (Fig. 7; Fontes and Matray, 1993). Bromine remains in the fluid phase up to potash saturation. Conversely, during halite dissolution Cl is released, and the Cl/Br ratio increases. The relationship of formation water Cl/Br ratio to Cl/Br seawater evaporation trajectory (SET) is therefore a useful indicator of brine origin. The majority of northeastern and southeastern Leduc brines parallel SET (Fig. 7A), indicating a possible seawater evapoconcentration or “primary” origin for these brines. The northwestern brines plot to the right of SET, indicating a possible halite dissolution or “secondary” origin. The spread of values in the brines of primary origin may be explained by variable mixing with meteoric water (Fig. 7B; Connolly et al., 1990a). The δD versus δ¹⁸O values of the Leduc brines show an evaporative trend that is offset parallel with SET, which may be explained by a combination of initial Devonian seawater δD versus δ¹⁸O composition, evaporative concentration, rock-water interactions, and meteoric water dilution (Fig. 4; Hitchon and Friedman, 1969; Hitchon et al., 1971). The southeastern brines show an offset parallel trend with SET that could be explained by the evaporative concentration of Devonian seawater modified by rock-water interactions. The northeastern brines continue this trend and extend toward the global meteoric water line (GMWL), suggesting mixing with meteoric waters. Neither Cl/Br ratios nor δD versus δ¹⁸O composition unambiguously decipher the fluid origin for these brines as marine. With the additional observations that brine compositions overlap with Devonian δ⁷Li carbonate compositions, plot between modern seawater and global shale δ⁷Li compositions (Fig. 6), and approach Devonian seawater and marine evaporite ⁸⁷Sr/⁸⁶Sr ratios (Fig. 5), we interpret that the Leduc brines originate as primary (southeastern and northeastern) and secondary (northwestern) brines associated with Devonian seawater and Devonian marine evaporites (e.g., Middle Devonian Prairie evaporite).

Strontium isotopes have a well-documented utility in tracking brine origin, mixing, and fluid-rock interactions. As the radiogenic Sr isotope signature of a fluid is not influenced by fractionation effects, it reflects the ⁸⁷Sr/⁸⁶Sr ratio of the rocks and other fluids it has interacted with (Frost and Toner, 2004). ⁸⁷Sr/⁸⁶Sr values of the Leduc brines indicate distinct interactions with basinal rocks and fluids for the northwestern, northeastern, and southeastern brines. The northwestern brines are more radiogenic, falling within the ⁸⁷Sr/⁸⁶Sr composition of late-stage dolomite cements, which formed from Laramide deformation-related fluids that interacted with deep basinal siliciclastic sediments (Fig. 5A; Stacey et al., 2020). Contrastingly, the less radiogenic northeastern and southeastern brines do not exceed the Devonian maximum strontium isotope ratio of basinal shales (MASIRBAS) (Machel and Cavell, 1999), suggesting that brines did not interact with more radiogenic fluids or rocks and are instead defined by the composition of Devonian seawater, evaporites, shales, and the early burial–stage dolomitizing fluids. The Leduc brine δ⁷Li values reflect the significance of shale pore waters in the southeastern brines, with brine compositions plotting between modern seawater and global shale δ⁷Li compositions and overlapping with Devonian δ⁷Li carbonate compositions (Fig. 6). The fractionation between water and carbonate for δ⁷Li is heterogeneous between 0 and –9‰, depending on temperature, salinity, mineralogy, and depositional setting (e.g., Marriott et al., 2004; Pogge von Strandmann et al., 2019; Day et al., 2021), but this overlapping agreement between the brine samples and Devonian seawater carbonates (Kalderon-Asael et al., 2021; Wang et al., 2023) broadly supports the radiogenic Sr isotope results, suggesting a Devonian seawater origin for the brines with potential modification associated with early-stage reflux dolomitizing fluids (Stacey et al., 2020) and interaction with the Devonian marine shales.

Regardless of formation as primary versus secondary brines (Fig. 7), these end-member fluid origin scenarios cannot fully explain the present-day elevated Li concentrations of the Leduc brines. The majority of the Leduc brine Li concentrations are significantly higher than concentrations that can be achieved with modern seawater evapoconcentration up to bischofite saturation (Fig. 7A; Fontes and Matray, 1993). One possibility is that average Devonian seawater Li composition was higher than it is today. However, Li concentration data from Devonian marine carbonates reported by Kalderon-Asael et al. (2021) are on average ~25% lower than modern data. That said, assuming a relatively high partition coefficient (D Li/Ca of ~0.016; Delaney et al., 1989; Marriott et al., 2004), a seawater Li concentration higher than today of ~1 mg/L is possible but unlikely. Another more plausible scenario is that the Devonian epeiric seaway was locally enriched in Li. For example, it has been demonstrated that the Cretaceous interior seaway of North America had heterogeneous δ⁷Li compared with the open marine (Pogge von Strandmann, 2013; Ibarra, 2018). Even so, it is likely that additional sources of Li are needed to explain the elevated present-day brine Li concentrations.

Marine shale dewatering during early burial–related compaction may have expelled pore fluids rich in clay mineral exchangeable cations, such as Li and K. Correspondingly, a strong correlation is observed between Li and K concentrations in the Leduc brines (Fig. 3). Due to the basin-scale dolomitization of reservoir units and a lack of constraints on the Devonian shale marine clay mineralogy, Mg concentrations cannot be used to interrogate possible cation-exchange processes associated with Mg smectite clays or the possibility that structurally bound Li releases during smectite-illite transition (e.g., Emproto et al., in press; Gagnon et al., in press). The Li isotope composition of brines is useful for tracking rock-water interactions. Lithium fractionates appreciably with the formation of secondary minerals with the lighter isotope (⁶Li) preferentially incorporated into clay minerals (Dellinger et al., 2017; Pogge von Strandmann et al., 2021). Average global seawater and meteoric, river, and hydrothermal waters are isotopically heavy compared to clay-rich shales (Fig. 6). Thus, as an isotopically heavy fluid interacts with isotopically light clay-rich sediments, the isotopic signature of the fluid approaches that of the rock (i.e., it is driven toward lighter values) (Pogge von Strandmann et al., 2021). Several studies have used Li isotopes to infer rock-water interactions as an important contributor of Li to formation waters (Chan et al., 2002; Millot et al., 2011; Macpherson et al., 2014; Phan et al., 2016; Li et al., 2021; Yan et al., 2023; Yu et al., 2024). Indeed, the Leduc brines have δ⁷Li versus Li values that approach global shale and Devonian carbonate values, and plausible Devonian seawater composition (Fig. 6). Relatively low ⁸⁷Sr/⁸⁶Sr ratios rule out Li-rich crustal fluids or brine interaction with the Cambrian-Precambrian units as a source of excess Li (Fig. 5A; Connolly et al., 1990b). However, fluids with elevated radiogenic Sr were involved in late-stage reef cementation and these may have contributed Li to the northwestern brines (Fig. 5A). The observed regional variability in the brine Li concentrations could reasonably be explained by meteoric water dilution or mixing with relict deeper Li-rich brines (Figs. 4, 7A, B).

Building on more than 30 years of previous studies, our new measurements and the synthesis of data compiled in this work highlight that the sedimentary basin lithium brines are geochemically heterogeneous across the Late Devonian Leduc Formation reservoirs. We present a conceptual model for the Leduc brine formation, brine modification, and Li sources (Fig. 8). Lithium may have been introduced from variable sources at multiple stages of the Leduc brine formation and modification. We categorize potential Li sources associated with fluid origin (Fig. 8A; pink arrows) and Li sources associated with brine modification (Fig. 8B, C; green arrows). Li sources genetically linked to brine formation involve Li delivered to the Devonian seaway and concentrated by evaporation in the seaway (Fig. 8A). We observe positive covariation between Li and various dissolved species and isotopes—indicating that evaporative processes concentrated Li in seawater—including TDS, Na, Cl, and δ¹⁸O (Fig. 3). Conceivably, subaerial sources may include continental weathering products (e.g., fluvial or aeolian), possible distal ash fall from the Devonian-Mississippian Antler magmatic arc (Root, 2001), and Li-rich spring waters. Subaqueous sources may have included submarine groundwater discharge, hydrothermal fluids, and secondary brines produced by the dissolution of older Devonian evaporite sequences (e.g., the Middle Devonian Prairie evaporite). It would follow that the elevated concentrations of dissolved Li in the Devonian seaway were sorbed to detrital clay minerals, held in shale pore waters, and bound in marine evaporite or detrital clay minerals. Previous studies have hypothesized that the Devonian marine shales influenced the composition of the Devonian brines in the Alberta basin (e.g., Machel and Cavell, 1999). It is well documented that during burial, marine shales expel pore waters in response to compaction (Fig. 8B). Thus, it is not surprising to find evidence that the Devonian marine shales had a significant influence on the brine composition and Li concentration. We observe a strong positive correlation between Li and K (Fig. 3) and that brine ⁸⁷Sr/⁸⁶Sr ratios approach the maximum strontium isotope signature of basinal shales (Fig. 5B) and δ⁷Li values of brines overlap global marine shale compositions (Fig. 6). From these observations we interpret that Li was locally concentrated in the Devonian epeiric seaway and marine shales (Fig 8A) and that during burial the shale pore waters were expelled into the Leduc reservoirs due to enhanced porosity and permeability from early-stage reflux dolomitization (Fig 8B; Stacey et al., 2020, and references therein). This is observed in other hydrocarbon brines systems such as the Appalachian basin Middle Devonian Marcellus Shale (e.g., Phan et al., 2016) and the Gulf Coast basin Jurassic Smackover Formation (e.g., Stueber et al., 1994). To further investigate the hypothesis of local Li enrichment of the Devonian epicontinental seaway, future studies focused on the Li concentrations and isotopic signature of the Alberta basin Devonian shales, carbonates, and evaporites are needed.

As discussed in previous studies, Li concentrations of the Leduc brines likely cannot be fully explained by the evaporative concentration of seawater. Seawater brines are likely significantly modified by diagenetic, mixing, and dilution processes to produce the present-day brine compositions (e.g., Gupta et al., 2012). Lithium sources may have been introduced after brine genesis during shallow (Fig. 8B) or deep (Fig. 8C) burial stages. These sources may have included marine shale pore waters released during sediment compaction or dolomitization and other diagenetic fluids, relict basinal fluids modified by subsurface water-rock interactions, and fault-derived fluids related to the Laramide orogeny. Further modification of the brine composition may include mixing with relict brines of variable Li concentration and dilution by incursion, either meteoric or seawater. A large body of work on the diagenetic history of the Devonian reef trend facies indicates that basin-scale dolomitization spans shallow and deep-burial stages, invoking the influx of connate seawater, relict basinal brines modified by rock-water interactions, and fault-derived fluids (Stacey et al., 2020, and references therein). It is challenging to disentangle when the Li-rich brines took residence in the Leduc reservoirs and to what spatial and temporal extent various diagenetic and hydrodynamic processes influenced the brine composition. We hypothesize that expulsion of Li-enriched marine shale pore waters introduced seawater-derived brines to the Leduc reef complexes after or coeval early-stage reflux dolomitization during shallow burial (Fig. 8B). These brines were locally modified by subsequent diagenetic, fluid-mixing, and dilution processes (Fig. 8C).

The Devonian units structurally dip toward the deformed belt of the Canadian Rocky Mountains, to the present-day southwest (Fig. 1A). The Gupta et al. (2012, 2015) model predicts old (>120 Ma) primary brines in the deepest southwestern parts of the basin, secondary halite-derived brines of mixed age in the central-eastern basin, and young (<30 Ma) brines of mixed origin in the shallowest northeastern reservoirs (Fig. 8C). We find that the southeastern Leduc brines (deep primary brines) and northeastern Leduc brines (intermediate- to shallow-depth mixture of primary, secondary (?), and diluted brines) align well with these modeled ages and origins. Based on limited compositional data, the northwestern brines (deep to intermediate-depth secondary brines) appear to compare less favorably with the model. As such, our conceptual model (Fig. 8) begins to capture and explain the observed geographic heterogeneity of the brine geochemistry and Li concentrations (Figs. 1, 2). Future work is needed on the Li concentrations and Li and Sr isotope signatures of marine shales, carbonates, and evaporites to directly test the Li sources hypothesized here. Additional brine sampling and geochemical modeling and comparison of brine proximity to subsurface fault networks may elucidate the origin and Li sources of the northwestern brines.

The Leduc brines cluster into distinct dissolved species and isotopic compositional groups across the Leduc Formation reef complexes corresponding to northwestern, northeastern, and southeastern regions. This suggests that these fluids have unique origins, migration and mixing pathways, and rock-water interaction histories. Lithium sources are likely variable across the basin and introduced at different stages of brine genesis and modification. Rock-water interactions with basinal shales and carbonates appear to have a significant and regional influence on the present-day brine composition. Specifically, we posit that the Devonian shales are a key source of lithium to the brines. Our results and conceptual model begin to capture the regional heterogeneity of the brine composition and lithium concentration. This has implications for lithium resource exploration, production, and valuation in the basin. Our results suggest the following brine formation mechanisms and lithium sources:

1. Northwestern Leduc brines: Although the data set for the northwestern brines is limited, these samples clearly show geochemical trends at odds with the northeastern and southeastern brines. The northwestern Leduc Formation brines are likely secondary in origin. These brines mixed with fluids of elevated radiogenic strontium levels not observed in the northeastern and southeastern brines and have the overall highest lithium concentrations. High radiogenic strontium and lithium levels are regionally restricted to the northwestern brines, suggesting Li-enriched fluids were introduced locally, possibly along faults associated with Laramide deformation. Future exploration should focus on reservoir proximity to fault zones that intercept the deeper basinal siliciclastic sequences.

2. Northeastern Leduc brines: These fluids are predominantly primary in origin and have a variable degree of freshwater dilution. Dilution effects are likely to account for the overall lower lithium concentrations of the northeastern brines compared with the northwestern and southeastern brines.

3. Southeastern Leduc brines: These brines are predominantly primary in origin with limited freshwater dilution. The lithium concentration and isotopic signatures of these fluids are homogeneous and appear to be intimately linked to basinal rock-water interactions, specifically with the Devonian shales. Lithium concentrations are likely to be consistent across the southeastern region as the brines do not appear to be significantly modified by mixing with fault-derived or meteoric fluids.

The authors would like to thank E3 Lithium for their contribution of data and sponsored project agreement funds to Butler. We are grateful to Kirsten Pugh, Peter Ratzlaff, and Chris Doornbos of E3 Lithium for their input that improved the manuscript. We thank Soumen Mallick for his assistance with the strontium isotope measurements made at Brown University. We would also like to thank the guest editors (Tom Benson, Adam Simon, and Simon Jowitt), M.A. McKibben, and an anonymous reviewer for their suggestions and improvement of the manuscript.

Kristina L. Butler is an assistant professor at the Department of Sustainable Earth Systems Sciences at the University of Texas at Dallas. Her research group uses sedimentological and geochemical methods to investigate sediment- and brine-hosted critical element and economic mineral deposits. She received her bachelor’s degree from the University of Alaska Anchorage and her PhD from the University of Texas at Austin Jackson School of Geosciences. Her postdoctoral research was supported by an NSF postdoctoral research fellowship with joint affiliations at Brown University of the University of Washington.