Greisen-Hosted Lithium Resources of the Erzgebirge/Krušné Hory Province (Germany and Czech Republic)

The Erzgebirge/Krušné hory in the Bohemian Massif (Fig. 1) hosts significant hydrothermal ore deposits, which have been mainly exploited for Sn, W, Ag, U, Co, and fluorite since the Medieval times (Baumann et al., 2000). Over the course of time, human interest in various commodities constantly changed as society evolved. The recent increase in the demand of battery raw materials has led to the reassessment of classic granite-related Sn-W systems of the Erzgebirge/Krušné hory as a domestic source of lithium within the European Union. Up to date, the total Li endowment of the Erzgebirge/Krušné hory is estimated to about 2 million metric tonnes [Mt] Li metal (Weinhold, 2002; European Metals, 2021; Leopardi et al., 2024; Zinnwald Lithium Plc, 2024). The bulk of these resources is related to the Zinnwald-Cínovec Li mica system, which spans across the border of Germany and the Czech Republic (Fig. 1B). Recent drilling campaigns of Geomet/European Metals and Zinnwald Lithium expanded the total resources estimates (measured + indicated + inferred) to >1.9 Mt of Li, with 708.2 Mt at 0.2% Li + 0.05% Sn + 0.02% W related to the Czech Cínovec deposit (European Metals, 2021) and 226.8 Mt at 0.22% Li related to the German Zinnwald deposit (Zinnwald Lithium Plc, 2024). This makes them the largest and second-largest hard-rock Li deposits in Europe, respectively (Gourcerol et al., 2019; Jamasmie, 2024). In addition to Zinnwald-Cínovec, there are several other satellite Li-(Sn-W) systems in the eastern Erzgebirge that are currently under exploration (Fig. 1C). These encompass the Sadisdorf, Falkenhain, Sachsenhöhe, Hegelshöhe, and Altenberg prospects, which are all geologically related to the Altenberg-Teplice caldera and within a few kilometers from Zinnwald-Cínovec (Fig. 1B, C; App. 1). All these occurrences display an intimate spatial and genetic association of Li, Sn, and W with the potential of coproduction of several critical metals.

The economically significant lithium mica concentrations in the Erzgebirge/Krušné hory are mainly hosted by greisen alteration affecting the cupolas of high-F Li mica-bearing leucogranite intrusions of late Paleozoic age (Štemprok, 1967; Breiter et al., 1999; Leopardi et al., 2024) (Fig. 2). Metasomatic greisen alteration is a product of intense subsolidus reactions between highly reactive, acidic magmatic-hydrothermal fluids and feldspar-rich rock types, commonly recognized in the roof of highly fractionated intrusions and associated breccia pipes and veins (Štemprok, 1987; Lehmann, 1990; Qiao et al., 2024) (Fig. 3). Greisen alteration is characterized by extensive (often complete) replacement of mica and feldspars by quartz, Li mica (mainly zinnwaldite; Fig. 4), muscovite, and/or topaz and may affect granitic rocks (endocontact greisen; Fig. 2E-H) as well as surrounding feldspar-bearing host rock types (exocontact greisen; Fig. 2K, L) (Štemprok, 1987; Halter et al., 1998). Counterintuitively, endocontact sodic ± potassic alteration is often recognized alongside greisen alteration. In contrast to feldspar-destructive greisen alteration, sodic ± potassic alteration is characterized by the replacement of mainly mica and quartz by albite and alkali-feldspar (Dolejš and Štemprok, 2001; Breiter et al., 2017a; Leopardi et al., 2024).

Similar to the economic interest, the scientific interest in the magmatic-hydrothermal greisen systems of the Erzgebirge/Krušné hory varied over time. During the 19^th and 20^th centuries there was increasing scientific interest in granite-related greisen systems. Some of those early studies significantly contributed to shaping economic geology as its own scientific discipline (e.g., von Cotta, 1859). Extensive exploration and associated research during the 1950s and 1960s resulted in a significant advance in the understanding of greisen deposits (Štemprok, 1967, 1987; Štemprok and Šulcek, 1969; Kühne et al., 1972; Štemprok et al., 1994). These studies provided an important data basis for follow-up studies and eventually the development of a general metallogenetic model for Sn granites (Lehmann, 1990). Although most of the mines were closed shortly after the fall of the Iron Curtain in 1989, scientific interest remains until today. Previous work includes geochemical and mineralogical characterization of the granites (Förster et al., 1999; Förster and Romer, 2010; Dolejš et al., 2016; Štemprok, 2016), the origin of Sn granites in the context of the European Variscides (Romer and Kroner, 2016; Franke and Żelaźniewicz, 2023), petrography and geochemistry of greisen and sodic ± potassic alteration (Dolejš and Štemprok, 2001; Monecke et al., 2011; Breiter et al., 2017a), chemistry of igneous and hydrothermal minerals (Johan et al., 2012; Müller et al., 2018; Breiter et al., 2019), geochronology of granites and hydrothermal mineralization (Romer et al., 2007; Tichomirowa et al., 2019b; Leopardi et al., 2023; Meyer et al., 2023), melt inclusions (Thomas, 1994b; Webster et al., 2004; Müller et al., 2006), and fluid inclusions (Thomas and Baumann, 1980; Korges et al., 2018; Leopardi et al., 2024). Although these systems were intensively studied over the last decades, the previous studies mainly focused on Sn and W, whereas systematic studies on Li are scant. Consequently, the controlling factors for Li enrichment in greisen systems remain insufficiently understood. With some exceptions, such as the Zinnwald-Cínovec deposit, the Li resources of most prospects/showings in the Erzgebirge/Krušné hory are poorly constrained because of its previously low economic significance.

The last comprehensive review paper on greisen- and vein-hosted mineralization in the Erzgebirge/Krušné hory dates back to 1967 (Štemprok, 1967), is in the German language, and strongly focuses on Sn and W, but the included information on Li is limited. In the following, we provide a review of greisen-hosted Li resources of the Erzgebirge/Krušné hory with emphasis on the Zinnwald-Cínovec deposit as the largest and most advanced Li-(Sn-W) project in the area. A general part summarizes the anatomy of greisen-hosted Li mica systems, the origin and ore-forming mechanisms for Li enrichment, and lithium mica compositions, as well as regional characteristics of Li mica occurrences in the Erzgebirge/Krušné hory. This is followed by a specific section on the economic geology of the Zinnwald-Cínovec and associated satellite systems. The metallogeny of Li mica deposits is discussed in the context of regional exploration as well as in the global context of other Variscan rare metal provinces.

The Erzgebirge/Krušné hory region is situated in the northern part of the Bohemian Massif and belongs to the Saxo-Thuringian zone of the Variscan orogenic belt (Franke and Żelaźniewicz, 2023) (Fig. 1A). The Variscan orogenic belt formed through the late Paleozoic continent-continent collision of Laurussia and Gondwana, resulting in the amalgamation of the supercontinent Pangea (Romer and Kroner, 2016). Because of the enormous scale of this orogenic event, metamorphic and igneous units of the Variscan orogenic belt are widespread and encompass, e.g., Cornwall (UK), the Erzgebirge-Krušné hory (Germany/Czech Republic), the Harz Mountains and Schwarzwald (Germany), the Massif Central, the Vosges and the Armorican Massif (France), and the Iberian Massif (Portugal/Spain) (Franke and Żelaźniewicz, 2023).

The Erzgebirge/Krušné hory forms a SW-NE–striking ~140- × 80-km erosional window, which is constrained by the Elbe zone to the east, the Fichtelgebirge to the west, and the Cenozoic Ohře/Eger graben to the south and is dipping below younger lithological units to the north (Fig. 1B). It comprises a series of metamorphic nappes that developed during the Variscan collision, including sedimentary and igneous protoliths of early Paleozoic to Neoproterozoic age (Tichomirowa et al., 2012). Metamorphic rock units generally decrease in metamorphic grade from southeast to northwest with ortho- and paragneisses prevalent in the southeast and east and mica schist and phyllite prevalent in the northwest and west. Peak metamorphism for the different tectonometamorphic units occurred between 360 and 340 Ma (Kröner and Willner, 1998; Collett et al., 2020). Metamorphism was followed by rapid exhumation and emplacement of syn- and postkinematic felsic intrusions (Förster and Romer, 2010). The postcollisional stage, which is often referred to as Late Variscan in older literature, is constrained to ~330 to 310 Ma (Romer et al., 2007; Tichomirowa et al., 2019b, 2022). The postcollisional stage encompasses plutonic rocks of mostly syeno- and monzogranite composition, lamprophyres, and volcanic rocks ranging from rhyolitic to dacitic composition (Förster et al., 1999; Förster and Romer, 2010; Casas-García et al., 2019). Postcollisional intrusive and related volcanic rocks are grouped into three NW-SE–striking batholiths: the western pluton, the central pluton, and the eastern pluton (Štemprok and Blecha, 2015). The most extensive outcrops of late-collisional igneous rocks are the Nejdek/Eibenstock composite granite (western pluton) and the Altenberg-Teplice caldera (eastern pluton), which were dated to 315 to 314 and 314 to 313 Ma, respectively (Tichomirowa et al., 2019b, 2022). Conversely, most of the central pluton is not exposed at the surface, except for a few smaller granitic stocks.

Early classification schemes attempted to group the granitic rock suites into an older and a younger intrusive complex (Tischendorf and Förster, 1994), which is inconsistent with geochronological data and therefore has been replaced by a widely accepted classification introduced by Förster et al. (1999), which follows an earlier approach of Breiter et al. (1991), who proposed seven chemically distinct granite types. Förster et al. (1999) proposed five groups instead, which are defined as follows: (1) low-F biotite granites (e.g., Niederbobritzsch), (2) low-F two-mica granites (e.g., Bergen), (3) high-F high-P₂O₅ Li mica granites (primarily S-type; e.g.; Eibenstock), (4) high-F low P₂O₅ Li mica granites (primarily A-type; e.g., Zinnwald), and (5) medium-F biotite granites (e.g., Eichigt-Schönbrunn). Greisen-hosted Li mica mineralization is invariably associated with high-F Li mica (mainly lithian annite) granites with both low and high P₂O₅ contents. Förster et al. (1999) proposed that the contrasting P₂O₅ contents of the two Li granite series are caused by variations in the aluminum saturation index: highly peraluminous melts of the western and central Erzgebirge increase the apatite solubility, which allows buildup of P₂O₅ along the magmatic fractionation trend. On the other hand, apatite solubility in the weakly peraluminous melts of the eastern Erzgebirge is low, which did not allow buildup of P₂O₅ (Wolf and London, 1994). Variation in aluminum saturation of different magma series is likely related to heterogeneities of the crustal protoliths. The emplacement depth of Li mica granites in the Erzgebirge/Krušné hory was constrained by H₂O contents in melt inclusions to 1.5 to 2.6 km (Thomas, 1994b), with the eastern Erzgebirge/Krušné hory Li mica granites of the Altenberg-Teplice caldera seemingly having the shallowest emplacement depths in the range of 1.5 to 2 km (Dolejš and Štemprok, 2001; Korges et al., 2018; Leopardi et al., 2024). The late-collisional stage was followed by postorogenic extensional tectonics associated with felsic magmatism such as the Rochlitz and Chemnitzer Becken volcanic centers (Förster and Romer, 2010; Hoffmann et al., 2013), whereas their plutonic counterparts are poorly documented. The available geochronological data constrain the postorogenic magmatic activity in the Erzgebirge/Krušné hory to ~305 to 286 Ma (Förster and Romer, 2010; Hoffmann et al., 2013). Subsequently, extensional tectonics and rifting prevailed until the Upper Cretaceous as a far-field effect of the Pangea breakup along the Tethys-Atlantic-Caribbean rift system (Ziegler, 1990). This was associated with basin formation and burial of the Erzgebirge/Krušné hory units below Mesozoic sediments (Ziegler, 1990). A shift in the overall stress field related to the collisional tectonics that led to the formation of the European Alps is associated with localized rifting and graben formation during the Cenozoic (Ziegler and Dèzes, 2007). Cenozoic rifting of the Ohře/Eger graben eventually caused the exhumation of the Erzgebirge/Krušné hory and its hydrothermal ore deposit spectrum (Ziegler and Dèzes, 2007).

The Erzgebirge is a well-endowed metallogenetic province hosting various styles of hydrothermal ore deposits that are mainly related to three distinct tectonic periods: (1) the late Paleozoic collision of Gondwana and Laurussia forming the supercontinent Pangea (Romer and Kroner, 2016; Reinhardt et al., 2022), (2) the Mesozoic breakup of Pangea along the Caribbean-Atlantic-Tethys rift systems (Burisch et al., 2022), and (3) local Cenozoic rifting of the Ohře/Eger graben (Burisch et al., 2021).

The late Paleozoic period comprises magmatic-hydrothermal greisen, skarn, stockwork, and vein-hosted Sn ± W ± Li as well as vein-style epithermal Ag-Zn-Pb mineralization related to postcollisional magmatism (Baumann et al., 2000). Examples are the Zinnwald/Cínovec (Li-Sn-W greisen; Breiter et al., 2017a), the Hämmerlein-Tellerhäuser/Zlatý Kopec (Sn-In skarn; Schuppan and Hiller, 2013), the Gottesberg (Sn vein/breccia/greisen; Wasternack et al., 1995), and the Freiberg (Ag-Zn-Pb vein; Swinkels et al., 2021) deposits (Fig. 1B). Skarn and greisen-related cassiterite and garnet laser ablation-inductively coupled plasma-mass spectrometry (LA-ICP-MS) U-Pb dating constrained the timing of ore formation to ~330 to 295 Ma in the Erzgebirge/Krušné hory (Burisch et al., 2019a; Reinhardt et al., 2022; Leopardi et al., 2023; Meyer et al., 2023). The Mesozoic period comprises low-temperature (<200°C) fluorite-barite ± base metal, U, and native metal-arsenide (Bi-Co-Ni-Ag) veins that are related to extension and continental rifting in conjunction with the breakup of Pangea (Romer and Cuney, 2018; Guilcher et al., 2021a; Burisch et al., 2022). These include, e.g., the Niederschlag/Kovářská fluorite deposit (Haschke et al., 2021), the Schlema U-Co-Ni deposit (Guilcher et al., 2024) and the Annaberg-Buchholz polymetallic vein district (Guilcher et al., 2021b). The timing of Mesozoic veins was constrained by U-Pb geochronology on gangue carbonates to 130 to 85 Ma (Guilcher et al., 2021a). There were many attempts to directly date the historically important vein- and shear-zone-related U mineralization of the Erzgebirge/Krušné hory: U-Pb age dates on uraninite range from 280 to a few Ma and indicate that U was repeatedly remobilized since the Permian opening of brittle structures, with a peak in U mineralization between 160 to 100 Ma (Förster and Haack, 1995; Romer and Cuney, 2018). The latest Cenozoic stage comprises quartz + Mn-Fe oxyhydroxide veins with high concentrations of Co, Ni, and Li (Burisch et al., 2021). This mineralization stage occurs as late-stage vein infill and as individual veins (e.g., Eichigt). Mineralization was dated to 41 to 34 Ma, which relates it to the Ohře/Eger graben rifting and associated low-temperature processes (Burisch et al., 2021).

Greisen-hosted Li mica deposits of the Erzgebirge are intimately related to late Paleozoic highly fractionated leucogranites (rare metal granites) with A- (e.g., Zinnwald-Cinovec) and S-type (e.g., Eibenstock) affinity (Štemprok et al., 1994; Förster and Romer, 2010; Leopardi et al., 2024). The granite stocks commonly comprise several intrusions, which follow a trend of progressive magmatic fractionation with time (Seltmann, 1984; Leopardi et al., 2024). Exsolution of magmatic-hydrothermal fluids from the crystallizing magmas resulted in metasomatism of the intrusions and their surrounding metamorphic and volcanic host rocks (Štemprok, 1987; Breiter et al., 2017a). For each individual location, the main hydrothermal event seems to be temporally related to the youngest and most fractionated intrusion (Weinhold, 2002; Leopardi et al., 2024). Associated alteration varies in terms of mineralogy (sodic ± potassic, greisen, and muscovite) and style (pervasive, stockwork, and veins) as a function of time and space during the complex magmatic and hydrothermal evolution (Leopardi et al., 2024). The main Li ore mineral is zinnwaldite, which is most abundant in metasomatic greisen alteration overprinting the apex and flanks of the granites (endocontact), as well as their immediate host-rock units (exocontact).

The size and geometry of greisen-hosted Li mica systems varies, but most share certain characteristic components that define the general anatomy (Fig. 3). The following paragraph provides a qualitative description of this anatomy, form deep to shallow, including the unaltered source/host intrusion, endocontact alteration zones in the cupola, stockscheider, exocontact alteration zones, breccias, stockwork zones, and veins (Fig. 3). Depending on the degree of exposure/preservation, some of the described components, however, may be missing at specific occurrences.

Causative/host intrusion: Available information on the host/source intrusions related to greisen systems is somewhat limited, because most of the exploration boreholes went not deep enough to intersect unaltered granite (Štemprok et al., 1994; Leopardi et al., 2024). Greisen-hosted Li deposits are usually associated with Li mica-bearing monzo- to syenogranites (Fig. 2A, B), which are strongly enriched in Li, F, Sn, W, Nb, Ta, Cs, and Rb, and have significant water contents of ~3.5 to 9 wt % (Štemprok et al., 1994; Thomas, 1994a; Müller et al., 2006). The main minerals are quartz (~30–50 vol %), orthoclase (20–50 vol %), albite (up to 25 vol %), lithian annite (2–10 vol %), and in some cases muscovite (0–6 vol %) (Dolejš et al., 2016; Štemprok, 2016). Accessories encompass zircon, topaz, ilmenite, rutile, fluorite, columbite-tantalite minerals, monazite, xenotime, uraninite, thorite, and sometimes apatite (Rub et al., 1998; Štemprok, 2016). Texturally, the intrusions range from thin localized pegmatitic zones to medium-grained equigranular to fine-grained porphyritic microgranites (Štemprok, 2016; Neßler, 2017; Leopardi et al., 2024). In the uppermost portion of the unaltered granite, a fine-grained microgranitic variety is commonly predominant.

Sodic ± potassic alteration zone: The apical portion of the host intrusion is strongly metasomatized with deeper sodic ± potassic alteration (Fig. 2I, J) in contact to the unaltered biotite granite and shallower greisen alteration in the uppermost part of the cupola (Dolejš and Štemprok, 2001; Štemprok, 2016; Breiter et al., 2017a; Leopardi et al., 2024) (Fig. 3). The transition from unaltered granite to the sodic ± potassic alteration zone is gradual (Štemprok, 2016). In addition to the distinct increase in the abundance of feldspar, the transition is marked by a change from lithian annite to zinnwaldite (Fig. 3), which is associated with an increase in the bulk-rock Li content from less than 0.1 wt % to 0.15 to 0.35 wt % (Štemprok, 2016; Leopardi et al., 2024). The sodic ± potassic-altered granite may contain some disseminated cassiterite in addition to accessory topaz, fluorite, synchesite, bastnäsite, sphalerite, pyrite, wolframite, and scheelite (Rub et al., 1998; Štemprok, 2016; Leopardi et al., 2024). Particularly in the older literature, some of the features associated with sodic ± potassic alteration (Fig. 2I, J) were misinterpreted as magmatic. However, geochemical and textural arguments support a metasomatic origin of the feldspar-rich units, with zinnwaldite, albite, and lesser K-feldspar replacing the primary igneous mineral assemblage (Dolejš and Štemprok, 2001; Štemprok, 2016; Breiter et al., 2017a; Leopardi et al., 2024).

Endocontact greisen alteration zone: Greisen alteration (Fig. 2E-H) affects the uppermost 50 to 350 m of the host intrusion and typically increases in intensity toward the intrusive contact (the contact between the intrusive stocks and the wall rock) (Fig. 3). The transition from sodic ± potassic to greisen alteration may be sharp or gradual. Sodic ± potassic alteration appears to be paragenetically older than greisenization, which may, however, be a result of a progressive downward migration (telescoping) of the interface between feldspar and greisen alteration as the systems cool down (Leopardi et al., 2024). Within one occurrence, greisen composition is highly variable and may vary locally on a relatively small scale (Fig. 2H). Generally, quartz and quartz-mica greisen are the most common varieties, whereas mica-quartz greisen and topaz-quartz greisen are less abundant. Quartz-mica and mica-quartz greisen are the primary hosts for economically significant concentrations of Li mica (Fig. 2E, G, H). Quartz-mica greisen typically consists of 40 to 75 vol % quartz, 20 to 50 vol % zinnwaldite, and 5 to 10 vol % topaz, whereas mica-quartz greisen has a higher modal abundance of zinnwaldite of typically 50 to 80 vol % (Štemprok et al., 1994; Neßler, 2017; Leopardi et al., 2024). Both greisen varieties may be associated with cassiterite (<5 vol %), fluorite (<5 vol %), and wolframite (<2 vol %), as well as minor/accessory molybdenite, chalcopyrite, sphalerite, scheelite, monazite, zircon, and columbite (Rub et al., 1998; Neßler, 2017; Hreus et al., 2021). Endocontact greisen alteration may overprint the entire roof zone of the intrusive stock (e.g., Sadisdorf) or forms irregularly shaped bodies (Zinnwald-Cínovec). In addition to pervasive alteration, greisen alteration is often associated with stockwork and vein zones. Endocontact greisen veins typically occur in subvertical and/or subhorizontal orientation (Štemprok and Šulcek, 1969; Breiter et al., 2017a; Leopardi et al., 2024). The thickness of greisen alteration around veins varies from a few centimeters to several meters. The vein infill usually comprises quartz, zinnwaldite, muscovite, and fluorite as well as cassiterite, wolframite, molybdenite, and other minor sulfides. Late-stage replacement of zinnwaldite by muscovite is documented for several Li mica occurrences, (Neßler, 2017; Leopardi et al., 2024) (Fig. 4).

Stockscheider: The host intrusion is commonly separated by a stockscheider (pegmatitic border unit; Figs. 2D, 3) from the intruded host rock (Thomas et al., 2000). Stockscheider thickness varies from a few centimeters to a few meters (<5 m) with a variable mineralogy encompassing mainly quartz and feldspar, which are often completely or partly replaced by quartz, zinnwaldite, and topaz (Hösel, 1994; Leopardi et al., 2024).

Exocontact greisen alteration zone: The wall rock in the hanging wall of the apical portion of the granites is commonly strongly altered by exocontact greisen (Štemprok, 1967; Leopardi et al., 2024) (Figs. 2K, L, 3). Depending on the wall-rock composition, which may encompass granites, rhyolites, and metasedimentary rocks, the most common exocontact greisen alteration encompasses medium- to fine-grained quartz-mica and mica-quartz varieties. Exocontact greisen is mainly related to stockwork, vein, and breccia zones (Fig. 3), in contrast to mostly pervasive endocontact greisen alteration. Explosive magmatic(-hydrothermal) breccia zones are associated with most deposits that were not completely exhumed and form chimney-like subvertical bodies (Fig. 3). The vertical extent of exocontact greisen alteration varies significantly from several tens of meters to more than 900 m (Wasternack et al., 1995). The occurrence of Li mica is, however, restricted to the most proximal exocontact greisen zones, whereas distal greisen is typically quartz-muscovite rich. Conversely, Sn and W are abundant in both the proximal portion and more distal stockwork and veins. (“Zwitter” is the historical term for distal greisen veins.) A general trend of increasing Pb-Zn-Ag-Cu sulfide abundances with increasing distance to the contact is documented for some greisen systems (Seltmann, 1984; Lehmann, 2021; Leopardi et al., 2024) (Fig. 3).

The formation of economic concentrations of greisen-hosted Li mica involves enrichment of Li through magmatic, transitional, and magmatic-hydrothermal processes. Fertile Li-rich granitic melts of the Erzgebirge/Krušné hory were the product of anatectic melting of mainly crustal protoliths (Förster and Romer, 2010). Their ε_Nd(t) and ⁸⁷Sr/⁸⁶Sr values plot, however, in between mantle-derived rocks and regional Paleozoic sedimentary rocks (Förster and Romer, 2010). Förster and Romer (2010) and Romer et al. (2014) proposed metasedimentary units with mafic to felsic volcano-sedimentary intercalations as the protoliths, whereas Štemprok and Blecha (2015) and Dolejš et al. (2016) also proposed a predominant sedimentary source, but with a significant contribution of a mantle-derived melt. Conversely, Hf and O isotope compositions of zircon from various granites of the Erzgebirge/Krušné hory were used to constrain high-grade para- and orthogneisses as the dominant source of the Li granites (Tichomirowa et al., 2019a). The protoliths of these high-grade gneiss units range from Ordovician volcanic and plutonic rocks to Cadomian graywackes (Tichomirowa et al., 2012). Although there is still no consensus on the precise source rocks, bulk and mineral isotopic compositions and elemental signatures consistently indicate that the protoliths of Li(-Sn) granites were primarily metasediments. Bulk-rock lithium isotope values of high-F Li mica granites (δLi –2.0 to 0.4) overlap with Li isotope signatures of the metasedimentary rocks of the Erzgebirge (δLi –6.0 to 0.5), which is consistent with other isotopic signatures (Romer et al., 2014).

Preenrichment of Li in the protolith and a low degree of melting of the protolith are favorable to produce Li-rich melts (Förster and Romer, 2010; Romer et al., 2014). High concentrations of Li, F, Sn, W, Nb, Ta, and Rb and their systematic correlation with, e.g., increasing SiO₂ and decreasing TiO₂ of various high-F Li mica granites in the Erzgebirge/Krušné hory, suggest significant ore concentration through fractional crystallization (Dolejš and Štemprok, 2001; Lehmann, 2021; Leopardi et al., 2024). High F and H₂O concentrations in Li mica granite melts acted as fluxes and likely extended the duration of fractional crystallization through undercooling (Dolejš and Zajacz, 2018; McCaffrey and Jowitt, 2023). This is consistent with estimated crystallization temperatures that vary for different approaches (using melt inclusions and calculated zircon saturation temperatures) but are invariably lower for high-F Li mica granites compared to low- to intermediate-F granites in the area (Thomas, 1994b; Tichomirowa et al., 2019a). At the end of the crystallization sequence, the exsolution of significant amounts of fluid marks the magmatic to hydrothermal transition. Partition coefficients determined by analysis of coexisting melt and fluid inclusions in granitic melts constrain moderate to strong partitioning of Li into the fluid phase (Li fluid/melt partition coefficients = 2–30; Zajacz et al., 2008). Thus, element partitioning between fluids and melts at the magmatic-hydrothermal transition seems to be a decisive factor for Li enrichment of the fluid. LA-ICP-MS analyses of individual fluid inclusions hosted by greisen-related cassiterite and wolframite of the Zinnwald/Cínovec deposit revealed Li concentrations in the thousands of parts per million range (up to 8,000 ppm; Korges et al., 2018), providing direct evidence for the strong enrichment of Li in the hydrothermal fluid. Lithium mica is typically paragenetically older than the bulk of cassiterite and wolframite. Therefore, actual Li concentrations during Li mica precipitation might have been even higher, possibly in the weight percent range (Webster et al., 2004).

Homogenization temperatures of fluid inclusions in cassiterite, quartz, and topaz associated with zinnwaldite are typically in the range between 400° and 500°C (Thomas and Baumann, 1980; Korges et al., 2018); thus, fairly similar or higher formation temperatures can be assumed for zinnwaldite. The precise precipitation mechanism for Li mica is insufficiently constrained, but as most of the Li mica is related to pervasive alteration with zinnwaldite replacing igneous plagioclase (often pseudomorphously), water-rock interaction is assumed to be a decisive process (Štemprok, 1987; Leopardi et al., 2024). Moreover, cooling, phase separation, and dilution with meteoric fluids were proposed as ore-forming mechanisms for cassiterite precipitation and could also be relevant for Li mica formation (Audétat et al., 2000b; Korges et al., 2018; Harlaux et al., 2021; Schmidt et al., 2021; Leopardi et al., 2024). Lithium partitioning into the vapor phase during phase separation (boiling) may concentrate Li in the vapor (Fiedrich et al., 2020). Although phase separation is documented for several greisen systems (Korges et al., 2018; Leopardi et al., 2024), the role of fluid-vapor partitioning for the enrichment of Li remains, in greisen systems, entirely speculative.

Lithium micas (Fig. 4) are mainly hosted by biotite-alkali feldspar granites (Fig. 2A, B) and associated pervasive endocontact sodic ± potassic and greisen alteration (Fig. 2E-J) and to a lesser extent by exocontact greisen alteration related to stockwork and veins (Figs. 2K, L, 3). Magmatic and hydrothermal Li mica compositions across the Erzgebirge range from lithian siderophyllite/annite in unaltered biotite-alkali feldspar granites to zinnwaldite (Fig. 4) and rarely lepidolite in greisen and sodic ± potassic alteration zones (Johan et al., 2012; Breiter et al., 2017c, 2019). All these mica compositions are solid solutions, which lie on the siderophyllite-polylithionite and the annite-trilithionite compositional trends (Breiter et al., 2019). Hydrothermal zinnwaldite typically forms together with topaz as one of the paragenetically earliest minerals (Fig. 4D). Both zinnwaldite and topaz are intimately intergrown with quartz. The grain size of greisen-hosted zinnwaldite is usually in the range of millimeters to a few centimeters, whereas pegmatite- and vein-hosted zinnwaldite may reach grain sizes up to several tens of centimeters. Lithium, F, and Rb concentrations correlate positively with each other and systematically increase toward the apical intrusive contact as a function of increasing degrees of fractional crystallization and metasomatism (Breiter et al., 2019). Concentrations of those elements in Li micas typically range from ~0.5 to 2.3 wt % Li, 2.5 to 9.5 wt % F, and 0.15 to 1.7 wt % Rb (Johan et al., 2012; Breiter et al., 2019, 2023). Conversely, Mg and Ti in mica show an opposite trend and systematically decrease with progressive crystallization and metasomatism toward the roof of the granite. Tin, W, Nb, and Ta increase along the fractionation trend in lithian annite until other phases such as cassiterite and columbite-tantalite coprecipitate, which causes a depletion in those elements in Li mica within the most evolved or metasomatized portions of the granite. As a consequence, zinnwaldite hosted by sodic ± potassic and greisen alteration has generally lower Sn (~100–300 ppm), Nb (~10–500 ppm), and Ta (~10–100 ppm) concentrations compared to igneous Li mica (Sn = 100–700 ppm; Nb = 400–1,000 ppm; Ta = 60–200 ppm; Breiter et al., 2019). The Nb/Ta ratio systematically decreases with increasing fractionation and metasomatism, irrespective of other coprecipitating phases (Breiter et al., 2017c, 2019).

In addition to the critical elements Li, Sn, and W, greisen-hosted Li mica systems are enriched in Nb, Ta, and Sc (Rub et al., 1998; Breiter et al., 2017b; Hreus et al., 2021). Zinnwaldite from Cínovec contains 32 to 169 ppm Sc, with the highest concentrations reported from quartz-zinnwaldite veins (Hreus et al., 2021). Columbite group minerals are the main host for Nb and Ta and are frequently intergrown with zinnwaldite (Rub et al., 1998). The abundance of columbite minerals increases toward the apical intrusive contact (Rub et al., 1998). Their Nb/(Nb + Ta) ranges from 0.6 to 0.9 (Rub et al., 1998; Breiter et al., 2017b; Hreus et al., 2021). Whole-rock concentrations of Nb and Ta range from about 10 to 200 ppm and also increase progressively toward the apical intrusive contact (Rub et al., 1998). Columbite from Zinnwald-Cínovec may host up to 3.04 wt % Sc₂O₃ (Hreus et al., 2021). Higher concentrations of Sc are typically associated with Ta-rich rims of columbite grains (Hreus et al., 2021). Cassiterite is mainly intergrown with quartz, zinnwaldite, and muscovite and is prevalent in sodic ± potassic alteration, greisen, and veins (Breiter et al., 2017b; Neßler, 2017). Cassiterite from Zinnwald-Cínovec may contain up to ~7 wt % Nb₂O₅ and ~6 wt % Ta₂O₅. Greisen-hosted cassiterite may also host up to 0.16 wt % In₂O₃ (Breiter et al., 2017b). LA-ICP-MS analyses of Sc in cassiterite are usually below the detection limit, which conflicts with high Sc (100–10,000 ppm) concentrations in reported bulk-mineral analyses of cassiterite from Zinnwald, Altenberg, and Krupka (Kempe and Wolf, 2006). The high Sc concentrations in the mineral separates reported in Kempe and Wolf (2006) are therefore assumed to be caused by Sc-rich inclusions of other minerals in the analyzed cassiterite mineral separates. Wolframite mainly occurs in greisen and veins and has Sc₂O₃ concentrations ranging from 0.17 to 0.9 wt %, with the highest concentrations reported from veins. Wolframite may also contain significant Nb concentrations (~0.5–3 wt % Nb₂O₅; Hreus et al., 2021). Mineral chemical and bulk-rock data imply that zinnwaldite, columbite, and wolframite are the main hosts for Sc in greisen systems. Bulk-rock concentrations of greisen zones are typically in the range of 10 to 40 ppm of Sc, whereas they range from 1 to 9 ppm in the unaltered igneous units (Breiter et al., 2017a; Hreus et al., 2021).

Greisen-hosted Li mica systems occur in several clusters spread across the entire Erzgebirge/Krušné hory region (Fig. 1B; App. 1). They are traditionally divided into the western, central, and eastern Erzgebirge/Krušné hory, with separate batholiths, respectively (Štemprok, 1967; Štemprok and Blecha, 2015). The eastern and western batholiths form parallel northwest-southeast trends perpendicular to the general strike of the belt (Fig. 1B, C). A regional overview of greisen-hosted mineralization, irrespective of Li mica content, is provided in Štemprok (1967), whereas the following description mainly focuses on Li mica-bearing greisen occurrences.

Eastern Erzgebirge/Krušné hory: Recent lithium exploration activity focuses on the Zinnwald-Cínovec deposit and to a lesser extent on the nearby Sadisdorf and Falkenhain prospects (Fig. 1C; App. 1). Further occurrences of Li mica are the Sachsenhöhe and Altenberg occurrences on the German side as well as Loupežník, Preisselberg, and Knöttel (Krupka district) on the Czech side (Štemprok, 1967) (Fig. 1C; App. 1). Most of the Li mica is hosted by endocontact greisen and sodic ± potassic alteration in the cupola of high-F, low-P Li mica alkali-feldspar granites. They form pipe- or dome-shaped discrete stocks that intruded into the metamorphic basement units as well as (sub)volcanic rhyolites (Štemprok, 2016; Breiter et al., 2017a; Leopardi et al., 2024). Lithium mica granites of the eastern Erzgebirge/Krušné hory and their rhyolitic host rocks are related to the Altenberg-Teplice caldera complex (Casas-García et al., 2019; Tomek et al., 2019) (Fig. 1 B, C). The timing of Li mica granites within the evolution of the Altenberg-Teplice caldera has been controversially debated (Romer et al., 2007; Zhang et al., 2017), yet recent geochronological high-precision chemical abrasion-isotope dilution-thermal ionization mass spectrometry (CA-ID-TIMS) data constrain the extrusion of the rhyolites to 313 to 312 Ma (Tichomirowa et al., 2022). Uranium-lead LA-ICP-MS ages of greisen-related cassiterite from Sadisdorf are around 313 Ma (ranging from 315.1 ± 3.3 to 311.0 ± 4.0 Ma; Leopardi et al., 2023), and Ar-Ar ages of zinnwaldite from Zinnwald/Cínovec range from 314.9 ± 2.3 to 312.6 ± 2.1 Ma (Seifert et al., 2011), indicating that emplacement of Li mica granites and associated greisen and sodic ± potassic alteration coincide with the late-stage collapse of the Altenberg-Teplice caldera (App. 1; Leopardi et al., 2023). Cassiterite and zinnwaldite age data are consistent with crosscutting relationships of Li mica granites and the Teplice rhyolite and suggest that Li-Sn-W mineralization is significantly younger than previously proposed (Romer et al., 2007; Ackerman et al., 2017; Zhang et al., 2017). The emplacement of granite stocks seems to be structurally controlled by mainly two parallel northwest-southeast lineaments. The cupolas of the Li granites are either partly exposed or not deeper than 200 m below the surface.

Central Erzgebirge/Krušné hory: The exposure of igneous rock units is generally low in the central Erzgebirge/Krušné hory (Fig. 1B). Exceptions are several small- to medium-sized intrusions such as the Geyersberg, Greifensteine, and Ziegelberg high-F, high-P alkali-feldspar biotite granite stocks (Hösel, 1994; Hösel et al., 1996; Meyer et al., 2023). Some of these stocks comprise zinnwaldite associated with stockwork-related greisen alteration. Previous exploration and mining operations focused mainly on the exocontact greisen, stockwork, and skarn-hosted Sn mineralization of Ehrenfriedersdorf and Geyer (Hösel, 1994; Hösel et al., 1996) (Fig. 1B; App. 1). The prevailing host rocks are mica schists and paragneisses that are underlain by a laterally continuous granite massif. The contact between the granite and the metasedimentary rocks is subhorizontal and was drill intersected at depths between 150 and 400 m below the surface (Hösel, 1994). Several flat domes and ridges locally elevate the intrusive contact by 100 to 200 m and are commonly spatially associated with exocontact Sn mineralization. The granite domes host abundant Li mica related to greisen and sodic ± potassic alteration (Hösel, 1994; Hösel et al., 1996). Most of the known Li mica-bearing greisen zones are poorly constrained in their spatial extent, except for the Ostgreisenkörper and the Westgreisenkörper at Ehrenfriedersdorf. They form 500-m-long, 40-m-wide, and 2- to 60-m-thick greisen bodies, which are related to NE-SW–trending granite ridges (Hösel, 1994). Uranium-lead LA-ICP-MS ages of greisen-hosted cassiterite (associated with Li mica) constrain the timing of hydrothermal mineralization to 314 ± 1.5 and 305.7 ± 3.7 Ma for Ehrenfriedersdorf and Geyer, respectively (Meyer et al., 2023).

Western Erzgebirge/Krušné hory: Farther to the west, the exposed large-volume Nejdek/Eibenstock (Fig. 1B) high-F, high-P Li mica alkali-feldspar composite granite (Wasternack et al., 1995; Förster et al., 1999; Dolejš et al., 2016) is the most prominent geologic feature (Štemprok, 1967). The exposed granite forms a N-S–elongated ellipsoid that spans across the German-Czech border (Fig. 1B). Although the Nejdek/Eibenstock composite intrusion has a strong background enrichment of Li (~200–1,100 ppm; Förster and Romer, 2010; Romer et al., 2022) and contains Li-bearing micas ranging from siderophyllite to lithian annite in composition (Tischendorf et al., 1969; Štemprok et al., 1996), it is not associated with significant Li mineralization (i.e., zinnwaldite greisen). The small Podlesi stock located at the southeast flank of the Nejdek/Eibenstock pluton presents an exception and locally hosts some zinnwaldite (Breiter et al., 2006). Endocontact greisen alteration of the Nejdek/Eibenstock massif and other intrusions in the western Erzgebirge/Krušné hory is characterized by a quartz-muscovite ± topaz assemblage, which causes localized depletion of Li from ~200 to 1,100 ppm to less than ~100 ppm in the whole rock because of the replacement of lithian annite by Li-poor muscovite (Wasternack et al., 1995; Romer et al., 2022). The eastern flank of the Nejdek/Eibenstock pluton dips gently below metasedimentary rocks that host a series of significant Sn ± W ± In skarn deposits (Schuppan and Hiller, 2013; Burisch et al., 2019a) (Fig. 1B) and major U deposits with a total endowment of more than 270 kt of Sn and 150 kt of U (German and Czech part combined; Hiller and Schuppan, 2008). Mineralization is related to the NW-trending and about 20-km-wide Gera-Jáchymov lineament that separates the central from the western Erzgebirge/Krušné hory (Fig. 1B). This area has only a few exposed granite stocks, but deep drilling indicates larger granite intrusions (Hiller and Schuppan, 2008).

Yet it is important to note that greisen- and skarn-hosted Sn mineralization associated with the Nejdek/Eibenstock pluton postdates the emplacement of the granite by up to 18 m.y. (Dolejš et al., 2016; Burisch et al., 2019a; Reinhardt et al., 2022). This observation indicates that mineralization is related to a younger magmatic-hydrothermal event with associated source intrusions not exposed (Burisch et al., 2019a). Such deeper causative intrusions are likely enriched in Sn, as they are associated with Sn mineralization. Whether they are also associated with Li, however, remains completely unconstrained.

The Zinnwald-Cínovec Li ± Sn ± W deposit is ~30 km south of Dresden and ~90 km northwest of Prague. The greisen-hosted Li ± Sn ± W deposit spans across the German-Czech border with about one-third on Saxonian/German and two-third on Bohemian/Czech ground, which are referred to as Zinnwald and Cínovec, respectively (Figs. 1C, 5A). Mining of alluvial cassiterite in the Zinnwald-Cínovec district goes back to the 13^th century, and the earliest record of underground mining dates back to 1378 on the Czech side and to 1557 on the Saxonian side (Schilka, 1994). For the following 400 years, mining focused on Sn and minor Ag and Cu mainly hosted by endocontact quartz-zinnwaldite veins. From 1876 onward, the primary focus shifted to W production. The extraction of some Li as a by-product started in 1890 (Schilka, 1994). The demand of metals during World Wars I and II was associated with an increase in the production of mainly W and Sn for relatively short periods. With the end of the World War II in 1945, production ceased on the German side but continued on the Czech side until 1991 (Schilka, 1994). After World War II, extensive exploration campaigns, including diamond core drilling, led to the discoveries of new orebodies in the German Democratic Republic and Czechoslovak Socialist Republic. The exploration campaigns during the 1950s and 1960s particularly provided detailed information on the geometry, geochemistry, and mineralogy of the greisen bodies (V. Grunewald, unpub. report, 1978; Štemprok et al., 1994). On the German side, many of the drill holes from the 1950s and 1960s were analyzed for Li, as Li was planned to be mined as a by-product. However, Li resource estimation was not completed. Whereas exploration activities ceased in 1967 on the German side, it continued on the Czech side. After a geophysical survey, an extensive drilling campaign was carried out in the 1970s that also led to the discovery of several hidden mineralized intrusive stocks south and southeast of Zinnwald-Cínovec (Štemprok et al., 1994). Recent Li exploration has been ongoing since 2010 on the Czech side (Geomet/ European Metals) and since 2011 on the German side (Zinnwald Lithium).

The Li mica deposit of Zinnwald-Cínovec is mainly hosted by endocontact greisen and sodic ± potassic alteration within the strongly metasomatized Zinnwald granite cupola (Figs. 5, 6). The Zinnwald granite forms a N-S–trending ellipsoid-shaped cupola, which intruded volcanic rocks of the Teplice rhyolite (Figs. 1C, 6). The granitic cupola is exposed as an ~1.4- × 0.3-km large N-S–elongated lens. The cupola is characterized by gently dipping (10°–30°) intrusive contacts to the north, east, and south, whereas the western intrusive contact is steeper with a dip of 50° to 80°. The covered southern flank steepens and forms a dome-like structure (Vrchlík sector; Fig. 6). Above this dome-like structure an up to 10-m-thick sill emanates from the intrusion in a south direction, which is almost parallel to the contact of the intrusion and the host rhyolite (Fig. 6). The uppermost ~700 m of the granitic cupola is with variable intensity overprinted by sodic ± potassic (mainly albitization) and greisen alteration (Štemprok, 2016; Breiter et al., 2017a). Therefore, the unaltered Zinnwald granite is only intersected in the 1,600-m-deep scientific borehole CS-1 (Štemprok, 2016). Most of the intersected unaltered granite is a medium-grained monzo- to syenogranite with porphyritic to equigranular texture (Štemprok, 2016) (Fig. 2B). It is composed of quartz (32–52 vol %), orthoclase (19–33 vol %), albite (20–27 vol %), and lithian annite (2–6 vol %) with accessory zircon, xenotime, monazite, thorite, rutile, columbite, pyrochlore, synchesite, and fluorite (Johan et al., 2012; Štemprok, 2016). Within the uppermost ~300 m of the unaltered granite, several zones with fine-grained textural varieties are recognized (Štemprok, 2016). At about 740 m below surface, the lithian annite granite transitions gradually within an ~30-m-zone into zinnwaldite-bearing sodic ± potassic alteration, which is prevalent within the uppermost 700 to 300 m of the granitic cupola (Štemprok, 2016) (Fig. 6). Within the transitional zone, lithian annite is only partially replaced by zinnwaldite (Štemprok, 2016). The fine- to medium-grained granite overprinted by sodic ± potassic alteration has on average 35 vol % albite, 33 vol % quartz, 23 vol % alkali-feldspar, 6 vol % zinnwaldite, and 2 vol % muscovite (Štemprok, 2016; Neßler, 2017). Accessory minerals include topaz, fluorite, cassiterite, zircon, columbite, monazite, xenotime, titanite, and scheelite (Rub et al., 1998). Lens- and chimney-shaped albitite zones, which gradually transition into the more common sodic ± potassic alteration, are frequently documented but typically do not exceed 50 cm in thickness. These albitite units may reach up to 98 vol % albite.

Most of the economic Li mica resource is associated with mica-rich greisen alteration that is prevalent in the uppermost ~350 m of the granitic cupola and along the flanks, forming irregularly and lens-shaped bodies that are parallel to the intrusive contact (Figs. 6, 7). The greisenized portion of the cupola forms an approximately 2.5- × 1- × 0.25-km Li mica orebody (Figs. 6, 7). The transition between sodic ± potassic and greisen alteration is gradual. The uppermost portion (20–100 m thickness) of the greisen zone (upper greisen) is characterized by intense greisenization, which is also associated with the highest Li grades (Figs. 6, 7). Recent drill core assays (drilled between 2017 and 2023) include 30.85 m at 0.78 wt % Li₂O and 61 m at 0.66 wt % Li₂O (European Metals, 2021; Kühn and Schultheis, 2022). The upper high-grade greisen zone gradually transitions into a lower greisen zone (100–250 m thickness), which is characterized by mainly partial greisenization (Fig. 6). Recent (2017–2022) assays of the lower greisen zone include 162.28 m at 0.44 wt % Li₂O and 122.0 at 0.46 wt % Li₂O (European Metals, 2021; Kühn and Schultheis, 2022).

Each of the two zones is composed of a variable number of vertically discontinuous and laterally extensive greisen bodies, which are mostly parallel to the intrusive contact. At the southern flank (Vrchlík sector), the greisen bodies have an orientation parallel to the intrusive contact, dipping south but bending toward a N-dipping orientation about 400 m farther to the north, which is the opposite direction as the contact (Figs. 6, 7). This offset of the greisen orientation relative to the contact coincides with the dome-like shape of the intrusion at the Vrchlík sector, which implies that the southern part of the deposit presents an individual (slightly younger) intrusion (Figs. 6, 7).

Thicknesses of individual greisen bodies typically range from 1 to 50 m; they may laterally extend for several hundred meters. The greisen bodies may or may not be associated with a subhorizontal vein with coarse-grained/pegmatitic infill (Fig. 5), which comprises mainly quartz, zinnwaldite, fluorite, cassiterite, wolframite, and scheelite. The flat veins seem to have formed coeval to the pervasive greisen bodies and show highest abundances in the central apical portion of the intrusion. Their thickness is typically 20 to 50 cm thick (rarely up to 2 m). These flat veins are the main hosts for high-grade Sn and W mineralization and were thus the prime target of historical mining operations. Some of the fracture-related greisen bodies continue beyond the granite contact into the host rhyolite. Nevertheless, exocontact greisen alteration at Zinnwald-Cínovec is generally subordinate and economically insignificant. In addition to greisen bodies and mainly subhorizontal veins parallel to the intrusive contact, steep veins with variable strike directions are documented. Such veins have a similar vein infill as the flat veins and are associated with a greisenized alteration halo that is a few centimeters to 2 m wide.

About 90% of greisen alteration at Zinnwald is quartz-mica greisen (Figs. 2E, G, 5) with local varieties of mica and quartz-topaz greisen (Fig. 2F, H). Quartz-mica greisen consists of 60 to 65 vol % quartz, 20 to 25 vol % zinnwaldite, 5 vol % topaz, and minor cassiterite, fluorite, kaolinite, scheelite, wolframite, monazite, zircon, and columbite (V. Grunewald, unpub. report, 1978; Neßler, 2017). The abundance of sulfides (arsenopyrite, loellingite, tetrahedrite, sphalerite, chalcopyrite, galena) is low at Zinnwald-Cínovec. Sulfides are typically more common in exocontact greisen and distal veins (Lehmann, 2021; Leopardi et al., 2024), which are poorly preserved at Zinnwald-Cínovec. Moreover, late-stage barite veins occur at Zinnwald-Cínovec (Neßler, 2017), which, based on their mineralogical similarity to other barite occurrences in the Erzgebirge, likely formed during the Mesozoic (Haschke et al., 2021; Burisch et al., 2022).

Recent exploration began in 2011 with the acquisition of the Zinnwald license by Solarworld. In February 2017, Bacanora Lithium Plc acquired 50% of Solarworld, and the resulting joint venture was Deutsche Lithium GmbH. In 2020 to 2021 Zinnwald Lithium Plc bought out all shares of Bacanora Minerals and Solarworld and is since then the sole owner of the Zinnwald license (Dittrich et al., 2020).

The initial exploration work focused on reevaluation of 46 historical bore holes (11,718 m) and related reports from the 1950s and 1960s exploration campaigns. Between 2012 and 2017, 24 diamond core drill holes (DDHs) and one reverse circulation (RC) drill hole with a total length of 6,942.9 m were drilled (Dittrich et al., 2020). The first two drill holes in 2012 were drilled as twins of historical holes and confirmed the validity of historical assays. Underground sampling included 88 channel samples (collected in 2012). Moreover, two bulk ore samples, 20 and 100 t, were collected for processing test work (pilot plant scale). With the takeover by Zinnwald Lithium Plc in 2020 to 2021 the project was reassessed and restructured, including an extension of exploration activity toward depth. Between July 2022 and September 2023, 84 additional DDHs with a total length of 26,969 m (avg depth of ~300 m) were drilled to improve the confidence level and to expand the resources to depth. Lithium resources were calculated to 226.8 Mt at an average Li grade of 0.22% applying a cutoff grade of 0.11% (Table 1; Zinnwald Lithium Plc, 2024).

In 2010, the company Geomet, a subsidiary of European Metals and ČEZ Group, claimed an exploration license that covers the entire Czech part of the Zinnwald-Cínovec intrusion and extends significantly farther to the south-southeast. Geomet started with a compilation and reevaluation of almost 90,000 m of historical drill holes and related reports. Between 2014 and 2021, 60 DDHs and seven geotechnical holes were drilled in several stages with a total length of 21,312.5 m (avg depth of 345 m). A significant amount of that drill core material was assayed, resulting in 10,860 multielement and 1,775 whole-rock analyses. In combination with the historical drill core material and related assays, the recent database encompasses a total number of 111,101 m of drill core with 78,257 geochemical analyses.

The latest resource calculation (European Metals, 2021) estimates a total resource of 708.2 Mt (7.39 Mt lithium carbonate equivalent) at an average grade of 0.2% Li applying a cutoff grade of 0.1% (Table 2).

Since most of the Li mica was considered as a waste rock, some of the tailings dumps related to Sn and W mining contain considerable amounts of Li. On the German side, all mine dumps were reprocessed for W and Sn during the early 20^th century. Significant amounts of the waste material from reprocessing were then used for construction and are thus not available anymore.

On the Czech side, the company Cínovecká deponie, owned by RSJ Investment, holds a mining permit for the largest of the remaining tailings dumps. The dump mainly consists of waste from a gravity treatment plant and some flotation residues, which were produced between 1959 and 1990. The total mass of the dump was estimated to 860 kt with a total metal content of 2.3 kt of Li (Paterová et al., 2013).

The following provides a concise summary of the geology and exploration stage of each of the greisen-hosted Li mica prospects and occurrences related to the Altenberg-Teplice caldera (from north to south). Emplacement of Li granites in the eastern Erzgebirge/Krušné hory occurred along two parallel northwest-southeast trends that are thought to be related to pre-Variscan fault zones that facilitated the ascent of the Variscan magmas (Štemprok et al., 1994; Štemprok and Blecha, 2015). The resource estimates and the geology of the Li mica deposits and prospects are summarized in Appendix 1. A comprehensive compilation of available geochronological data, including intrusive and mineralization ages, is provided in Leopardi et al. (2023).

The Sadisdorf Li-Sn-(W-Cu) prospect is located ~16 km north of Zinnwald-Cínovec in the northern part of the Altenberg-Teplice caldera (Fig. 1C). The deposit was episodically mined between 1505 and 1954 for Sn, Cu, and W (Müller, 1867; Seltmann, 1984). An extensive drilling campaign (about 30,000 drill m) was executed by the Soviet-German Wismut SDAG (Sowjetisch-Deutsche Aktiengesellschaft) company during the 1980s. In 2017, Lithium Australia received an exploration permit and repeated three of the previously drilled holes, confirming the historical results. In 2021, the license was acquired by Zinnwald Lithium. Recent resource estimates include inferred resources of 25 Mt at an average grade of 0.15% Li and 3.36 Mt at an average grade of 0.44% Sn (Leopardi et al., 2024) (App. 1).

Lithium-tin mineralization is spatially associated with the shallowly emplaced Sadisdorf composite stock, which intruded the metamorphic basement (Seltmann, 1984; Leopardi et al., 2024) (Fig. 6). The Sadisdorf composite stock is exposed in a N-S–trending, 200- × 100-m outcrop. The intrusion widens at depth forming an ~600- × 600-m-wide cupola (Fig. 6). The Sadisdorf stock is composed of at least four magmatic stages, which include pre-ore biotite syeno- to monzogranites (G1-3) followed by a fine-grained granitic intrusion (G4) that is assumed to be the fertile intrusion (Seltmann, 1984; Leopardi et al., 2024). A quartz-(mica)–rich stockscheider separates the G4 intrusion from earlier intrusions and the metasediments. The stockscheider becomes progressively alkali feldspar rich toward the flanks of the intrusion. A pipe-like magmatic-(hydrothermal) breccia body extends above the apex of the G4 granite toward the present-day surface (Seltmann, 1984; Leopardi et al., 2024).

The granite cupola is strongly metasomatized with a deeper sodic ± potassic alteration and a shallower greisen zone. Lithium mineralization is mainly hosted by endocontact greisen alteration that affects the uppermost 100 to 200 m of the apical part of the intrusion and laterally continues along the flanks (Leopardi et al., 2024). The predominant greisen types are quartz-mica with some local zones of mica-quartz and minor topaz-quartz greisen. Additional Li resources are hosted by the most proximal part of the apical exocontact greisen zone and endocontact sodic ± potassic alteration zone (Leopardi et al., 2024).

The Altenberg Sn-W-Mo deposit is about 3.5 km north of Zinnwald-Cínovec and occupies a central eastern position within the Altenberg-Teplice caldera (Fig. 1C). The deposit was discontinuously mined from 1446 until 1991 and is arguably the most important Sn deposit of the Erzgebirge/Krušné hory (Weinhold, 2002). Despite its more than 500 years of mining history, the remaining resources were estimated to 27 Mt at 0.26% Sn and 0.12% Li (Weinhold, 2002). The original geology and geometry of the Altenberg deposit is only partly documented because the upper levels of the historical underground mine collapsed in 1620 (Weinhold, 2002). The Altenberg composite stock consists of several magmatic stages comprising medium- to fine-grained syeno- to monzogranite with equigranular to porphyritic textures. A magmatic breccia pipe occurs at the west side of the granite stock (Weinhold, 2002). The main host rocks are the Teplice rhyolite and the porphyritic ring dike rimming the Altenberg-Teplice caldera. Widespread endo- and exocontact greisen alteration has on average 24 vol% zinnwaldite (Weinhold, 2002). The greisen zone transitions into sodic ± potassic alteration at depth. Sodic ± potassic alteration was exposed at levels 6 and 7 of the historical mine (Weinhold, 2002). The Tiefenbachhalde tailings dump was assessed for its remining potential; however, Sn was the only commodity considered in detail. The total mass of tailings material was estimated to 5.6 Mt with 6.9 wt % zinnwaldite (Büttner et al., 2018).

The Falkenhain Li-Sn-W prospect is situated at the north-east rim of the Altenberg-Teplice caldera, ~11 km north of Zinnwald-Cínovec, and is 100% owned by Zinnwald Lithium (Fig. 1C; App. 1). Shallow veins and exocontact greisen zones were exploited for Sn, Ag, and Cu in several small-scale operations related to the historical Gottes Gabe mine camp (1554–1857; Baumann et al., 2000). During the 1950s and 1960s Sn and W exploration campaigns, 138 DDHs with a total length of 32,086 m were drilled, but a resource estimation according to modern standards was never completed. In 2022, Zinnwald Lithium drilled one DDH, which confirmed the validity of historical assays. This drill hole LiSH-001 intersects mainly endocontact greisen mineralization including 80 m averaging 0.29 wt % Li, 0.049 wt % Sn, and 0.027 wt % W (Zinnwald Lithium, 2023).

The Falkenhain prospect encompasses two hidden granite stocks, the Schenkenshöhe and the Hegelshöhe sites, which intruded into the eastern porphyritic ring dike of the Altenberg-Teplice caldera (Zentrales Geologisches Institut Berlin, 1976; Trischler et al., 2022). The apical portions of the granite stocks are intersected at about 100 to 150 m below the surface. At least the uppermost 300 m of the stocks are strongly overprinted by sodic ± potassic and greisen alteration, hosting considerable amounts of Li mica. Mica-quartz and quartz-mica greisen varieties are both recognized. A centimeters- to meters-thick stockscheider pegmatite occurs at the apical intrusive contact and is strongly greisenized. Exocontact greisen alteration is associated with a dense stockwork of quartz-cassiterite veinlets that extends from the igneous contact to the surface (Geologische Landesuntersuchung GmbH Freiberg, 1990; Trischler et al., 2022).

The Sachsenhöhe prospect is about 9 km northeast of Zinnwald-Cínovec and is located at the eastern flank of the Altenberg-Teplice caldera, outside of the ring dike that delineates the caldera (Fig. 1C; App. 1; 100% owned by Zinnwald Lithium). Historical mining activity occurred between 1449 and 1877 and targeted cassiterite-bearing veins at the flanks of the exposed granitic stock. Historical mining operations did not go deeper than 60 m below surface.

The granitic stock is exposed at surface in a 200- × 200-m area (GKZ, 2008). The composite stock consists of three phases of syeno- to monzogranite intrusions (G1-G3) and a magmatic breccia that envelops the stock. The composite intrusion is hosted by mainly paragneiss and minor orthogneiss. Greisen alteration mainly occurs within the endocontact. At the southern edge of the Sachsenhöhe intrusion, an approximately 250-m-long and 100-m-wide, W-E–trending zone with intense greisen alteration and Sn-W–bearing quartz veins is recognized (GKZ, 2008). This zone is mainly hosted by monzogranite (G2) and magmatic breccia. To date, the prospect has not yet been systematically explored for its Li potential.

The Krupka pluton is a mostly concealed intrusive body situated southeast of Zinnwald-Cínovec. It forms a 5-km-long and 1- to 2-km-wide igneous body at the southeastern margin of the Altenberg-Teplice caldera. Several shallower stocks and domes emanate from the pluton and are associated with Li-Sn-W-Mo mineralization. These include the Loupežný, Presisselberg, and Knöttel deposits (from northwest to southeast; (Fig. 1C; App. 1).

Loupežný: The Loupežný dome is just 2 km south from the covered southern flank of the Zinnwald-Cínovec pluton and is still within the exploration license of Geomet. The cupola was intersected by several historical drill holes, which constrain its extent to about 6 km² in the subsurface. The apex of the dome is about 200 m below the surface. The main host rocks are the Teplice rhyolites. Lithium mica related to greisen and sodic ± potassic alteration at its northwest flank was confirmed by drilling (e.g., borehole KV-16; Štemprok et al., 1994).

Preisselberg: The historical mine camp of Preisselberg is located about 1 km southwest of the town Horní Krupka, 13 km southeast of Zinnwald-Cínovec. Previous applications for exploration licenses were denied, and the area became a protected UNESCO world heritage site in 2019. The Preisselberg stock is a composite intrusion with a diameter of about 400 m (Štemprok et al., 1994). Its apex is situated ~50 m below the surface. The host rocks encompass gneiss, intrusive rocks, and minor volcanic rocks. The apical portion of the granite is strongly greisenized and hosts abundant zinnwaldite. Mica-quartz and quartz-topaz are the most common greisen varieties. The greisen alteration zone transitions into fine-grained sodic ± potassic alteration toward depth (Štemprok et al., 1994).

Knöttel: The hidden Knöttel stock is ~1 km southeast of the town Horní Krupka. The upper ~200 m of the stock are exposed in four historical underground galleries. The stock is small, with a diameter of 100 to 150 m. A pegmatite cap is only developed in the apex of the intrusion and consists of quartz and K-feldspar (Peterková and Dolejš, 2019). The pegmatite transitions into a shallower fine-grained granite and a deeper medium-grained granite, which are overprinted by sodic ± potassic and greisen alteration that hosts some zinnwaldite. Greisen alteration extends only 10 to 40 m into the surrounding gneiss (Peterková and Dolejš, 2019).

The overall time span in which magmatic-hydrothermal mineralization (greisen-, skarn-, and vein-hosted) occurred in the Erzgebirge/Krušné hory is constrained to 335 to 295 Ma by LA-ICP-MS dating of garnet and cassiterite (Burisch et al., 2019a; Reinhardt et al., 2022; Leopardi et al., 2023). However, the timing of Li mineralization within this time window is poorly constrained, since direct geochronological data of zinnwaldite is scant. Greisen-hosted zinnwaldite at Zinnwald has Ar-Ar plateau ages of 314 to 312 Ma (Seifert et al., 2011; Neßler, 2017), and cassiterite intergrown with zinnwaldite at Sadisdorf has U-Pb ages of 315 to 311 Ma (Leopardi et al., 2023). These ages coincide with zircon ages of the western Nejdek/Eibenstock pluton, which was dated to 315 to 314 Ma (Tichomirowa et al., 2019b), implying that there are no obvious regional differences in terms of the timing of Li mineralization. Timing of Li mica formation in the central Erzgebirge/Krušné hory is indirectly constrained by U-Pb ages related to different generations of greisen-hosted cassiterite ranging from 314 to 305 Ma (Meyer et al., 2023). Their paragenetic relationship to Li mica is however unclear, since they are hosted in mica-poor quartz-topaz ± muscovite greisen. Judging from this limited available data, it seems that Li mineralization did not occur prior to about 315 Ma.

Conspicuously, all prolific Li mica prospects and deposits occur in the eastern Erzgebirge/Krušné hory, where most of the known greisen systems host abundant zinnwaldite. This is in stark contrast to the western Erzgebirge/Krušné hory, where greisen alteration is abundant but mainly contains muscovite as the prevailing mica (Štemprok, 1967). In some cases muscovite-rich greisen alteration may cause depletion of Li in igneous host rocks (Wasternack et al., 1995). Concomitantly, the metal tenor changes from Li-Sn-W in the east to Sn-W or W-Sn in the west (Štemprok, 1967; Baumann et al., 2000).

In the eastern Erzgebirge/Krušné hory, most of the mineralized stocks occur very close to the present-day land surface (0–200 m) and are fully (e.g., Sadisdorf and Krupka) or mostly preserved (e.g., Zinnwald-Cínovec and Sachsenhöhe). Sadisdorf presents the northernmost known greisen-hosted Li mica system in the east. However, farther to the northwest, magmatic-hydrothermal mineralization continues but transitions into several intermediate-sulfidation epithermal Ag-Zn-Pb vein systems of the Freiberg district (Burisch et al., 2019b; Swinkels et al., 2021), which may be related to a progressively shallower level of erosion toward the northwest (Fig. 1B). Cassiterite and stannite in the deepest vein intersection of the Freiberg district support that the epithermal veins may be related to Sn(± Li?) granites at depth (Bauer et al., 2019; Swinkels et al., 2021). Whether the transition from Li mica greisen to epithermal veins in the eastern Erzgebirge/Krušné hory reflects a temporal or spatial continuum remains, however, speculative because of limited availability of geochronological data.

The central Erzgebirge/Krušné hory is characterized by a low abundance of exposed igneous rocks, except for some small Li mica-bearing granite stocks such as the Greifensteine and Geyersberg intrusions (Fig. 1B). Most of the few known greisen-hosted Li mica zones are associated with flat granitic domes that underlie exocontact greisen stockwork zones and skarn alteration. Their apex is typically 200 to 400 m below the present-day surface, which suggests slightly less erosion relative to the emplacement depth of Li granites compared to those of the eastern Erzgebirge.

The Nejdek/Eibenstock granite in the western Erzgebirge/Krušné hory (Fig. 1B) shows many similarities to the deep-seated unaltered biotite granite of the Zinnwald-Cínovec deposit; e.g., major and trace element compositions (except P₂O₅) as well as lithian annite as the predominant mica. This suggests that the exposed Nejdek/Eibenstock granite likely presents a deeply eroded Li mica system with its shallower greisen zone mainly eroded. This is consistent with the large area (~20 × 40 km) of exposure of the Nejdek/Eibenstock pluton, which suggests an erosional level typical of deeper plutons/batholiths, whereas Li greisen alteration is related to shallower and smaller stocks. The assumption that the Nejdek/Eibenstock presents a deeply eroded equivalent of an Li granite is furthermore cemented by comparing it to the Slavkovský les region. The Slavkovský les, south of the Cenozoic Ohře/Eger graben (Fig. 1B), is the southeastern extension of the western Nejdek/Eibenstock pluton (Štemprok and Blecha, 2015). Large parts of the Slavkovský les consist of the same lithian annite granite that dominates the Nejdek/Eibenstock composite intrusion. Additionally, greisen and sodic ± potassic alteration-hosted zinnwaldite is abundant at the southwestern flank (Krásno district) of the Slavkovský les. Hubstock and Schnödenstock are two occurrences within the Krásno district with abundant zinnwaldite (Štemprok, 1967). They show many similarities to Sadisdorf in the eastern Erzgebirge/Krušné hory in terms of their mineralogy and fluid inclusion systematics (Štemprok, 1967; Štemprok et al., 1994; Dolníček et al., 2012; Leopardi et al., 2024). In contrast to the eastern Erzgebirge/Krušné hory, greisen alteration in the Nejdek/Eibenstock widely lacks zinnwaldite and is invariably muscovite rich instead (e.g., Gottesberg, Tannenberg, Auersberg; Štemprok, 1967). Furthermore, it shows characteristics typical of distal exocontact vein systems. For example, at the Gottesberg Sn deposit, which is hosted within the Nejdek/Eibenstock, drilling confirmed continuation of vein- and breccia-hosted greisen alteration to a depth of at least 900 m below surface, without intersecting the causative intrusive stock (Wasternack et al., 1995). This distal exocontact mineralization conflicts with a deep emplacement of the Nejdek/Eibenstock granite and thus suggests telescoping associated with significant exhumation of the Nejdek/Eibenstock in between its emplacement and the superimposition of younger Sn mineralization. The absolute timing of younger Sn systems hosted by the Nejdek/Eibenstock pluton and the nature of their causative source intrusions remain unconstrained. Knowledge on whether those deeper source intrusions are endowed in Li will have important implications for the metallogeny and the exploration potential of the western Erzgebirge/Krušné hory.

Since there seems to be no clear age difference of Li granite emplacement across the Erzgebirge/Krušné hory, the present-day distribution of Li mica deposits and prospects is likely primarily controlled by postmineralization exhumation relative to the emplacement depth of the granites. Rapid exhumation of the Erzgebirge/Krušné hory after the main collisional stage adds another layer of complexity (Schmädicke et al., 1995) but would support the hypothesis of telescoping, meaning that younger intrusions occupy a deeper crustal position relative to the present-day surface than those that formed at 315 Ma and earlier (Burisch et al., 2019a). On a smaller scale, the structural control is one of the most decisive factors, since NW-SE–trending (and north-south) pre-Variscan structures facilitated emplacement of shallow stocks and related stockwork and vein systems (Štemprok and Blecha, 2015).

Despite the extensive drilling campaigns in the second half of the 20^th century on both the German and the Czech side, there is still significant potential for exploration on the regional and deposit scale. A combination of detailed geophysical surveys and confirmation drilling has led to the discovery of the mineralized hidden Loupežný dome, northwest of Krupka. Available subsurface data from the German side is summarized in Tischendorf et al. (1965), which is however not detailed enough to be useful for targeting relatively small individual concealed stocks. Therefore, the resource potential is open and widely untested to the north and northwest of the Altenberg-Teplica caldera. However, potential Li resources are likely to occur progressively deeper toward north. Toward the south, the eastern Erzgebirge trend is constrained by the Ohře/Eger graben boundary fault (Krušné hory fault). Mineralization continues beyond the extent of the Altenberg-Teplice caldera but seems to be limited in an east-west direction to an ~6-km-wide zone, which is constrained by two NW-SE–trending structures (Štemprok et al., 1994). On the district and deposit scale, there is significant potential to increase the known Li resources, as the recent increase in mineral resources at the Zinnwald and Cínovec projects demonstrates. The lateral extent of the Zinnwald-Cínovec orebody along the flanks of the cupola is still widely unconstrained. The area southeast of Cínovec, including Loupežný, provides further potential to increase the resource and to make new discoveries. Systematic exploration of some of the satellite prospects, such as Falkenhain and Sachsenhöhe, recently started or is planned for the near future.

Only a few endocontact Li mica greisen zones were intersected during historical Sn exploration, as the focus was mainly on exocontact skarn and greisen-hosted Sn mineralization (Hösel, 1994; Hösel et al., 1996). Geophysical gravity data suggest that large parts of the central Erzgebirge are underlain by granitic rock (Hösel, 1994; Štemprok and Blecha, 2015), which has a significant potential for new discoveries. Besides the few drill holes that confirm Li mineralization at Ehrenfriedersdorf, ambient seismic noise tomography indicates the presence of undiscovered endocontact greisen zones in the Ehrenfriedersdorf-Geyer area (Ryberg et al., 2022). From the limited information available, we conclude that Li mineralization in the central Erzgebirge/Krušné hory seems to be mainly associated with flat granitic domes instead of discrete stocks, which makes targeting more challenging.

Current knowledge suggests that the resource potential for Li is low in the western Erzgebirge/Krušné hory because of the deep exhumation and correspondingly poor preservation of the upper parts of the Nejdek/Eibenstock system. This does not however apply to the Slavkovský les, south of the Ohře/Eger graben, where Li mineralization is open to the south-west. Some exploration potential would also remain in the western Erzgebirge, if younger exocontact greisen-hosted Sn systems (e.g., Tannenberg, Auersberg) prove to be associated with Li-rich roots.

The formation of the Variscan Li-Sn-W provinces in Europe is related to the continental collision of Gondwana and Laurussia, and hence they all belong to one large mineral system (Romer and Kroner, 2016; Reinhardt et al., 2022). As a consequence, the Iberian Massif, the Massif Central, the Armorican Massif, and the Cornwall Sn ± W ± Li provinces share many similarities to but also decisive differences from the Erzgebirge/Krušné hory (Štemprok, 1995; Müller et al., 2006; Simons et al., 2017; Harlaux et al., 2023).

The provinces within the European mainland belong to the margin of the Gondwana continent (internal Variscides), whereas Cornwall belongs to Laurussia (external Variscides; Romer and Kroner, 2016). Recent geochronological studies include cassiterite, columbite group minerals, wolframite, and garnet dating in Cornwall (Moscati and Neymark, 2020), the Erzgebirge/Krušné hory (Burisch et al., 2019a; Reinhardt et al., 2022; Leopardi et al., 2023; Meyer et al., 2023), the French Massif Central (Harlaux et al., 2018, 2024; Carr et al., 2021), and the Iberian Massif (Melleton et al., 2022; Ballouard et al., 2024). Timing of hydrothermal activity and related magmatism overlaps in the internal provinces ranging from ~335 to 280 Ma (Reinhardt et al., 2022), excluding some younger outliers that were interpreted as hydrothermal overprint (Harlaux et al., 2018). Conversely, U-Pb cassiterite ages of Cornwall are significantly younger at ~290 to 275 Ma, overlapping with only the youngest hydrothermal ages of the internal zone (Moscati and Neymark, 2020). The timing of Li granites within this >40 m.y. window of magmatic(-hydrothermal) activity is for all provinces insufficiently documented.

Most of the Li resources of the Erzgebirge/Krušné hory are related to magmatic-hydrothermal greisen alteration, whereas individual Li pegmatites, excluding minor stockscheider, are absent. Conversely, abundant pegmatites occur in the Iberian Massif and the French Massif Central, which host significant Li resources in both provinces (Roda-Robles et al., 2016; Gourcerol et al., 2019). The Beauvoir (France; 141 kt Li; Gourcerol et al., 2019), Argemela (Portugal; 23 kt Li; Gourcerol et al., 2019), and Zinnwald/Cínovec (Germany/Czech Republic; 1,915 kt Li; this study) systems host the largest granite-hosted Li resources for each of the respective provinces. The ages of Zinnwald/Cínovec and Beauvoir overlap with 314 to 311 (Seifert et al., 2011) and 315 to 310 Ma (Cuney et al., 1992, 2002), respectively, whereas the Argemela stock is older (326 ± 3 Ma; Melleton et al., 2022). The main lithology of each stock is a fine- to medium-grained leucogranite, which is strongly enriched in Li, Rb, Ta, Nb, and Sn (Cuney et al., 1992; Charoy and Noronha, 1996; Štemprok, 2016). When we compare the unaltered portions of each granite, the Beauvoir (~0.45 wt % Li) and the Argemela (~0.17 wt % Li) granites show significantly stronger enrichment in Li than the Zinnwald/Cínovec (~0.1 wt % Li) granite (Štemprok, 2016; Michaud et al., 2020; Harlaux et al., 2023). The most decisive differences, however, are the degree and style of hydrothermal alteration. The Argemela stock is only minimally affected by greisen alteration (Charoy and Noronha, 1996). The Beauvoir granite shows some alteration, which is mainly restricted to veins and stockwork zones. However, even in the most altered cupola, greisen alteration does not affect more than 15 vol % of the granite, and muscovite is invariably the main greisen-related mica (Cuney et al., 1992). Therefore, greisen-style hydrothermal processes apparently did not play an important role for Li enrichment for the Argemela and Beauvoir granites; instead they even may have caused local depletion of Li in the cupola of Beauvoir (Monnier et al., 2022). Conversely, hydrothermal metasomatism is the key ore-forming process that led to the formation of the Zinnwald/Cínovec deposit. The underlying reason for the absence of alteration at Argemela and lack of hydrothermal Li enrichment at Beauvoir is poorly constrained. We speculate that the emplacement depth may be a decisive factor for hydrothermal Li enrichment, as it controls the amount and composition of the exsolved fluids (and related alteration) and associated phase separation (Candela, 1997; Audétat et al., 2000a; Driesner and Heinrich, 2007). The emplacement depth of Argemela is unconstrained, whereas emplacement depth was estimated at 3 km for Beauvoir (Cuney et al., 1992) and 1.5 to 2 km for Zinnwald/Cínovec (Korges et al., 2018). The deeper emplacement of Beauvoir reflects the crystallization of a volatile-rich magma, which experienced less permeability on fluid exsolution and hydraulic fracturing, i.e., less fluid overprint. The Li budget of Beauvoir seems to be largely controlled by extreme fractional crystallization and preservation of the magmatic composition, i.e., high Li content in a relatively small volume (tonnage), with some, but limited, Li loss to the geochemical halo around the intrusion. This is different from Zinnwald/Cínovec, where the overprint by quartz veining (flat veins) and greisen-style alteration affected a large rock volume with lower-grade Li enrichment mainly driven by hydrothermal processes (Breiter et al., 2017a). Lithium enrichment to ore grade is thus largely magmatic in the Beauvoir system, whereas the Li mineralization of Zinnwald/Cínovec is largely of a postmagmatic fluid-dominated nature, although superimposed on a magmatically enriched reservoir (Cuney et al., 1992; Raimbault et al., 1995; Štemprok, 2016; Breiter et al., 2017a).

The greisen-hosted Li mica systems of the Erzgebirge/Krušné hory present a major resource of lithium within the European Union and provide a significant opportunity for the domestic production of several critical metals (Li, Sn, W, and others). Despite the long mining and exploration history, there is significant potential for exploration, particularly in the eastern and the central Erzgebirge/Krušné hory.

The greisen-hosted Li mica systems of the Erzgebirge/Krušné hory formed through a combination of processes related to the magmatic, magmatic-hydrothermal transitional, and hydrothermal stages of highly fractionated leucogranites (rare metal granites). The bulk of the Li resource is hosted by zinnwaldite, which is mainly related to greisen and lesser sodic ± potassic alteration that affects large parts of the shallowly emplaced (1.5–3 km) intrusive stocks. Lithium enrichment during the hydrothermal stage and the intense alteration distinguish the Li mica systems of the Erzgebirge/Krušné hory from other granite-related Li mica occurrences in the European Variscides. However, the controlling factors for Li enrichment at the magmatic-hydrothermal transition and in the hydrothermal stage are yet insufficiently understood. Geochronological data related to late Paleozoic magmatic-hydrothermal systems in the Erzgebirge/Krušné hory range from 335 to 295 Ma. The timing of Li mineralization within this time window remains poorly constrained but likely did not occur prior to 315 Ma. Therefore, further geochronological studies, particularly direct dating (e.g., cassiterite and hydrothermal mica), are important to better understand the metallogenetic position of Li granites within the postcollisional stage of the European Variscides. The Zinnwald-Cínovec deposit presents by far the largest Li mica system of the Erzgebirge/Krušné hory; the factors that control its extraordinary size remain, however, unclear.

This study was supported by the Deutsche Forschungsge-meinschaft (award 441189074) and the Sächsische Landesamt für Umwelt, Landwirtschaft und Geologie (Neues Potenzial Projekt). We would like to thank Geomet and Zinnwald Lithium for their support and the permission to use and publish their data. Furthermore, we would like to thank Tom Benson, Simon Jowitt, and Adam Simon for editorial handling. We are thankful for the photomicrographs of zinnwaldite provided by Jörg Neßler. We appreciate the fruitful discussions with Jens Gutzmer on the Erzgebirge mineral system. We are grateful for constructive reviews by Matthieu Harlaux, Karel Breiter, and an anonymous reviewer, which significantly improved a former version of this manuscript.

Mathias Burisch is an associate professor and head of the Mineral Systems Analysis Group in the Department of Geology and Geological Engineering at Colorado School of Mines, Colorado, USA. His research focuses on a better understanding of geologic processes that result in the formation, transformation, and preservation of ore deposits. Burisch mainly works on magmatic-hydrothermal systems and associated critical, precious, and base metal resources, aiming to constrain ore-forming factors and exploration criteria from the regional to the deposit scale.