The Diversity of Rare-Metal Pegmatites Associated with Albite-Enriched Granite in the World-Class Madeira Sn-Nb-Ta-Cryolite Deposit, Amazonas, Brazil: A Complex Magmatic-Hydrothermal Transition

Abstract

This study investigates pegmatites with exceptionally rare mineralogical and chemical signatures, hosted by the 1.8 Ga peralkaline albite-enriched granite, which corresponds to the renowned Madeira Sn-Nb-Ta-F (REE, Th, U) deposit in Pitinga, Brazil. Four distinct pegmatite types are identified: border pegmatites, pegmatitic albite-enriched granite, miarolitic pegmatite, and pegmatite veins. The host rock itself has served as the source for the fluids that gave rise to all these pegmatites. Their mineral assemblages mirror the rare-metal-rich paragenesis of the host rock, including pyrochlore, cassiterite, riebeckite, polylithionite, zircon, thorite, xenotime, gagarinite-(Y), genthelvite, and cryolite. These pegmatites formed at the same crustal level as the host granite and record a progressive magmatic–hydrothermal evolution driven by various physicochemical processes, including tectonic decompressing, extreme fractionation, melt–melt immiscibility, and internal fluid exsolution. Border pegmatites crystallized early from a F-poor, K-Ca-Sr-Zr-Y-HREE-rich fluid exsolved during solidification of the pluton’s border and were emplaced in contraction fractures between the pluton and country rocks. Continued crystallization toward the pluton’s core produced a highly fractionated melt enriched in Sn, Nb, Ta, Rb, HREE, U, Th, and other HFSE, forming pegmatitic albite-enriched granite within centimetric fractures. A subsequent pressure quench—likely induced by reverse faulting—triggered the separation of a supercritical melt, further enriched in rare metals, which migrated into fractures and cavities to form amphibole-rich pegmatite veins and miarolitic pegmatites. A key process in this evolution was melt–melt immiscibility, which led to the partitioning of alkalis between two phases: a K-F-rich aluminosilicate melt (low in H₂O), enriched in Y, Li, Be, and Zn; and a Na-F-rich aqueous melt (low in SiO₂). These immiscible melts crystallized polylithionite-rich and cryolite-rich pegmatite veins, respectively. The magmatic–hydrothermal transition occurred independently in each pegmatite body upon H₂O saturation, with the hydrothermal fluid composition controlled by the local degree of melt fractionation. These highly F-rich exsolved fluids caused intense autometasomatic alteration and secondary mineralization. The exceptional F content (up to 35 wt.% F in pegmatite veins), played a central role in concentrating strategic and critical metals such as Nb, Ta, REEs (notably HREE), Li, and Be. These findings establish the Madeira system as a reference for rare-metal magmatic–hydrothermal evolution in peralkaline granites.

1. Introduction

Pegmatitic rocks are very coarse-grained crystalline rocks, often containing giant crystals of feldspar, quartz, or mica. Their striking textures and concentration of rare-metal minerals have long attracted the interest of the mining industry [1]. Pegmatitic textures typically occur in igneous rocks of granitic (calc-alkaline) or syenitic (alkaline) composition. Granitic pegmatites, found in orogenic geodynamic regimes, are enriched in elements such as Li, U, Ta, and B, whereas alkaline pegmatites, associated with anorogenic settings, are richer in Nb, Zr, Th, and Mo [1].

There is currently no unified model to explain the origin of granitic pegmatites. Early models proposed that they represent the final stage of crystallization from low-viscosity granitic magmas [2,3]. Other models, supported by extensive mineralogical, chemical, and field data from “Variscan-type” pegmatites in the German basement, favor an anatectic origin [4,5]. Additionally, a line of thought emphasizes the role of fluxing elements (e.g., F, P, B, Li) and subsolidus and hydrothermal fluids in pegmatite formation [6,7,8], while the model proposed by Thomas et al. [9,10,11] suggests that pegmatites result from melt–melt immiscibility and metasomatic reactions. In this regard, pegmatites register important chemical and physical variations that take place during the transition from the magmatic to the hydrothermal phases in volatile-rich magmatic systems.

The enrichment of rare elements in pegmatites appears to occur in an essentially closed system, stemming from a minor fraction of highly evolved, volatile-rich residual silicate liquid derived from a much larger magmatic body. Despite this, only a small fraction of pegmatites (<1%) exhibits economically significant concentrations of rare metals such as Sn, Nb, Ta, REEs, Li, Be, Rb, Cs, P, and W [12].

The magmatic–hydrothermal transition is fundamental to the enrichment of rare metals and often marks the onset of ore-forming processes in pegmatitic systems [13,14]. Although pegmatite melts are typically interpreted as residual products of granitic crystallization, experimental data highlight the role of fluid immiscibility as an additional mechanism for rare-metals enrichment [15,16,17].

The geochemical signatures of rare-metal minerals in pegmatites offer valuable insights into the evolution of elements and volatiles (e.g., REEs, Nb, Ta, Li, F, H₂O) in magmatic and hydrothermal fluids. These chemical tracers help reconstruct crystallization conditions and processes at the magmatic–hydrothermal transition in highly evolved igneous systems [18,19,20,21].

This study focusses on rare-metal pegmatites exhibiting unusual mineralogical and chemical compositions hosted within the remarkable albite-enriched granite of the Madeira Sn-Nb-Ta-F (REE, Th, U) word-class deposit (164 Mt), located in Amazonas, Brazil. This 1.8 Ga peralkaline A-type granite underwent intense alteration by aqueous, F-rich hydrothermal fluids [22]. At its core lies a hydrothermal massive cryolite deposit (10 Mt, 37 wt.% Na₃AlF₆) [23]. This is an occurrence of global uniqueness due to the association of cryolite with Sn, Nb, and various other rare metals within a single peralkaline granite body.

The research team from the Universidade Federal do Rio Grande do Sul conducted an extensive investigation of the pegmatites associated with this albite-enriched granite. The work encompasses a series of targeted studies, some previously unpublished, now integrated with new mineralogical data. This study combines structural, mineralogical (including detailed compositions of key minerals), and geochemical data to pursue three main objectives: (i) to characterize the different pegmatite types associated with the albite-enriched granite; (ii) to propose a genetic model and determine their source; and (iii) to assess implications for the evolutionary history of the albite-enriched granite system. In addition, this study offers new perspectives on pyrochlore alteration, aiming to distinguish hydrothermal processes based on differences in alteration products. Special emphasis is placed on two key aspects: the magmatic–hydrothermal transition and the role of fluorine in both magmatic and hydrothermal environments.

2. Geological Setting and the History of Geoscientific Studies

2.1. Geological Setting from the Oldest to the Youngest Lithologies

The Pitinga Province is located (Figure 1) in the southern portion of the Guyana Shield [24], within the Tapajós-Parima Tectonic Province [25]. The region is predominantly composed of volcanic rocks of the Iricoumé Group [26], which yield zircon ²⁰⁷Pb/²⁰⁶Pb ages ranging from 1881 ± 2 to 1890 ± 2 Ma [27]. These rocks comprise mainly of effusive and hypabyssal rhyolites, highly welded ignimbrites, ignimbritic tuffs, and surge deposits formed in a subaerial environment with cyclic effusive and explosive activities [28,29,30]. The Iricoumé Group hosts the Madeira Granite (Figure 1).

The Madeira granite (Figure 1 and Figure 2) comprises four facies [23,32,33,34,35]. The oldest facies is a mostly metaluminous porphyritic amphibole–biotite granite (1824 ± 2 Ma) [31], characterized by plagioclase-mantled K-feldspar mega crystals and occasionally by reverse-zoned K-feldspar-mantled plagioclase ovoids. It is overlain by a metaluminous biotite granite (1822 ± 2 Ma) [31]. The younger facies are a hypersolvus porphyritic alkali feldspar granite (1818 ± 2 Ma) [31] and an albite-enriched granite (Figure 2), the latter being the host of the pegmatites studied here. The age of the albite-enriched granite is only loosely constrained at 1822 ± 22 Ma [22] due to metasomatic alteration of its zircons. According to Costi [31], the two younger facies were emplaced simultaneously. The hypersolvus granite is characterized by K-feldspar phenocrysts within a fine- to medium-grained matrix dominated by K-feldspar and quartz.

2.2. The Albite-Enriched Granite Host Rock

The albite-enriched granite (Figure 2) forms an oval-shaped body with an aerial extent of approximately 2 × 1.3 km. It is subdivided into two subfacies: the core albite-enriched granite (CAG) and the border albite-enriched granite (BAG). The CAG is a peralkaline subsolvus granite, porphyritic to seriate in texture, fine- to medium-grained, and composed of quartz, albite, and K-feldspar in roughly equal proportions (25–30% p.vol.). Accessory minerals include cryolite (4% p.vol.), polylithionite (4% p.vol.), annite (3% p.vol.), zircon (2% p.vol.), and riebeckite (2% p.vol.). Minor minerals include pyrochlore, cassiterite, xenotime, columbite, thorite, magnetite, and galena. The BAG is peraluminous and shares similar textures and essential mineralogy with the CAG. However, it is richer in zircon, contains fluorite instead of cryolite, and lacks iron-rich silicate minerals, which have been almost entirely removed by an autometasomatic process [31,35]. A massive cryolite deposit is located in the central part of the CAG, consisting of several bodies of massive cryolite intercalated with both the CAG and the hypersolvus granite (Figure 2). These bodies reach up to 300 m in length and 30 m in thickness and are composed mainly of cryolite crystals (~87 vol%), along with quartz, zircon, and feldspar.

2.3. The Madeira Rare-Metal Deposit

The Pitinga Province is the largest Sn producer in Brazil. The alluvial ore deposits were discovered in 1979 [26] and are now nearly depleted. The primary ores are predominantly associated with the Madeira Granite (Figure 1). The Madeira deposit, which has been actively mined since 1989, corresponds to the albite-enriched granite (Figure 2). The average grade of the disseminated ore (CAG + BAG) is 0.17 wt.% Sn (cassiterite), 0.20 wt.% Nb₂O₅ and 0.024 wt.% Ta₂O₅ (both hosted in pyrochlore and columbite). Potential by-products from the disseminated ore include F (4.2 wt.% cryolite), Y and HREE (xenotime), Zr and Hf (zircon), Th (0.07 wt.% ThO₂, thorite), and U (0.03 wt.% UO₂, pyrochlore). The pegmatites examined in this study have been mined indistinctly along with the disseminated ore at the Pitinga Mine. The massive cryolite deposit contains approximately 10 million tons with an average grade of 31.9% Na₃AlF₆ [23,37].

3. Materials and Methods

For this study, a collection of more than 500 rock samples from the research group at Universidade Federal do Rio Grande do Sul (UFRGS) was reviewed. The pegmatite samples were initially examined using a binocular loupe. Subsequently, 50 thin sections were selected for detailed petrographic analysis under optical microscopy to identify the mineralogy and paragenesis of the pegmatites associated with the albite-enriched granite. To obtain detailed textural data, the thin sections were examined using scanning electron microscopy (SEM), with qualitative analysis performed using an energy-dispersive X-ray detector (Zeiss, model EVO MA10) at the Center for Microscopy and Microanalysis at UFRGS, Porto Alegre, Brazil.

Most of the mineral chemistry data for the albite-enriched granite and associated pegmatites were obtained by the UFRGS research group and are available in Pires et al. [38], Bastos Neto et al. [39], Schuck [40], Stolnik [41], Lengler [42], Paludo et al. [43], and Hadlich et al. [37]. Mineral compositions of pyrochlore and associated secondary minerals were obtained using electron probe microanalysis (EPMA) with a JEOL JXA-8230 at the EPMA Laboratory of the Universidade de Brasília (UnB). Operating conditions were as follows: 15 kV accelerating voltage and 10 nA beam current for F, Mg, Zn, Al, Si, Hf, Nb, P, Cl, S, Bi, Ti, Mn, Y, Ta, Sn, Ca, Zr, Fe, V, and Rb; and 20 kV and 20 nA for Na, K, Pb, REE, Sr, Th, Ba, and U. A beam diameter of 1 μm was used, and interference corrections were applied in all cases of peak overlap. The Wavelength Dispersive X-ray Spectrometer (WDS) crystals used were: TAP (Si, Zn, Na, Al), PETJ (Nb, P, Hf, Cl, S, K, Bi, Sr, Y, Ta, Sn, Th, Pb), PETH (Rb, Zr, U), LIF (Ti, Mn, Sm, Eu, Gd, Dy, Er, Ho, Tb, Tm, Yb, Lu), LIFH (Ca, Fe, Ba, V, La, Ce, Pr, Nd), and LDE1 (F). Counting times for peaks measurements were 10 s for all elements, with background counts taken on both sides of the peaks for half that time. Natural and synthetic standards used included microcline (Si, K, Al), albite (Na), apatite (P, Ca), andradite (Fe), topaz (F), forsterite (Mg), vanadinite (V, Pb, Cl), pyrite (S), MnTiO₃ (Mn), YFe₂O₁₂ (Y), LiNbO₃ (Nb), LiTaO₃ (Ta), MnTiO₃ (Ti, Mn), ZnS (Zn), Bi₂O₃ (Bi), RbSi (Rb), BaSO₄ (Ba), baddeleyite (Zr), PbS (Pb), HfO₂, SrSO₄ (Sr), SnO₂, ThO₂, UO₂, and synthetic REE-bearing glasses. Additional data on polylithionite were obtained from Costi [31].

Whole-rock geochemical data (268 analyses) for the albite-enriched granite and associated pegmatites were also acquired by the UFRGS research group and are available in Bastos Neto et al. [23,44], Minuzzi et al. [36,45,46,47], Pires [48,49], Paludo et al. [43], Stolnik [41], and Lengler [42]. Samples were collected from drill cores and fresh outcrops, and analyses were performed at Actlabs, Ancaster, Ontario, Canada. Major elements were determined by ICP-AES, minor and trace elements by ICP-MS, and F by ISE.

4. Results

4.1. Miarolitic Pegmatites

4.1.1. Structure

Pegmatites hosted in miarolitic cavities are common [22,43,50]. These cavities range from a few centimeters to decimeters in size and exhibit irregular to rounded shapes (Figure 3). They are more frequently found near the boundary between the core albite-enriched granite (CAG) and the border albite-enriched granite (BAG). In most cases, the cavities lack structural control [50]; however, in some instances, they are aligned along the filling fractures [22]. Additionally, geodes ranging from 10 to 50 cm in diameter are also present.

4.1.2. Texture and Mineral Assemblage

The miarolitic cavities are filled with fine- to medium-grained aggregates composed of cryolite, polylithionite, zircon, and xenotime, along with riebeckite, albite, cassiterite, and opaque minerals [23]. In some cavities, well-marked zonation is observed [50]. The marginal zone is a few centimeters thick and composed of quartz and albite. The wall zone contains quartz and albite crystals with hypidiomorphic textures. The intermediate zone is characterized by a preferential orientation of minerals perpendicular to the cavity walls and an abrupt increase in crystal size. The core is typically composed exclusively of cryolite. The geodes are filled with quartz, cryolite, fluorite, and chlorite [23].

4.2. Pegmatite Veins

4.2.1. Structure

Pegmatite veins (PEG) are most commonly found in the central, northern, and northwestern portions of the CAG. Structural investigation [50] were conducted between the altimetric levels of 210 m and 140 m. Within this range, two types of pegmatite veins are visible across the entire mine front. The most common type consists of meter-scale tabular bodies, typically less than 1 m thick, emplaced along horizontal extensional fractures (Figure 4A). The second type comprises tabular bodies emplaced along subvertical reverse fault planes (Figure 4B). Locally, aplite dykes are observed cutting through pegmatite veins (Figure 4C). The pegmatite veins range from centimeters to decimeters in thickness and can be discontinuous along the same fault plane.

The geometric arrangement of the pegmatites is controlled by contractional brittle structures within the CAG, including reverse faults trending ~N320/60SW, imbricate fans, and horse blocks). These fractures and faults acted as conduits for fluid migration, with transport directed from SW to NE under a compressive regime, characterized by horizontal tension and at low solidus temperature. The horizontal pegmatite bodies are oriented perpendicular to the minimum stress axis (σ3), while the subvertical pegmatite veins are emplaced along reverse fault planes positioned at approximately 60° from σ3. The pegmatite veins preferentially occur in fault planes dipping toward the SW. Horizontal and subvertical structures are physically interconnected (Figure 4D), and the textural features of the pegmatites in these zones suggest that the fault planes functioned as fluid conducts facilitating the formation of the horizontal veins [50].

4.2.2. Texture

Both the subvertical and horizontal pegmatite veins have a thin, well-marked border zone, typically a few centimeters thick (Figure 5). From the margins toward the center of the bodies, there is a systematic increase in mineral size; however, no distinct mineral zoning is observed. The interior of the veins is homogeneous, composed of anhedral to subhedral minerals, with mineral sizes ranging from 0.1 to 10 cm.

In most veins, the pegmatitic texture (Figure 6) is well marked by crystals up to 10 cm in size of polylithionite, quartz, cryolite, microcline, and albite; crystals up to 7 cm of riebeckite, xenotime, and genthelvite; and crystals up to 3 cm of thorite, galena, and, more rarely, zircon, cassiterite, and gagarinite. Both the horizontal and subvertical pegmatite veins exhibit the same mineral assemblage (see below). However, differences in modal proportions allowed Paludo et al. [43] to classify the pegmatites into three groups: (i) amphibole-rich, typically with well-developed xenotime and genthelvite, and intermediate K and Na contents (Figure 6A); (ii) polylithionite-rich, typically containing abundant xenotime and genthelvite, with high K content (Figure 6B,C); and (iii) cryolite-rich, often containing quartz, galena, and well-developed xenotime, with high Na content (Figure 6D).

4.2.3. Mineral Assemblage

Albite, orthoclase, and quartz are the main constituents of the pegmatite matrix (Figure 7A). Within the matrix, these minerals may occur as anhedral, subhedral, or euhedral crystals, most commonly ranging in size from 0.3 mm to 1 cm. Albite also occurs as inclusions in microcline, quartz, polylithionite, cryolite, xenotime, or gagarinite, likely due to the inheritance of albite crystals from the host rock. The main amphibole (Figure 7D) is riebeckite, while F-arfvedsonite and F-eckermanite are very subordinate [43]. Riebeckite occurs either as disseminated anhedral grains or as aggregates of acicular crystals (0.6 mm to 7 cm), and it is frequently altered to chlorite (Figure 6A). In the amphibole-rich PEG, riebeckite crystals are often associated with polylithionite (Figure 7D). In the polylithionite-rich PEG, polylithionite most frequently forms aggregates (up to 15 cm) composed of anhedral crystals (0.2 mm–3 cm), and more rarely subhedral ones. Polylithionite crystals exhibit corrosion features in contact with cryolite II (Figure 7E). Total or partial pseudomorphosis of polylithionite by hydrothermal cryolite is common; hematite present along the cleavage planes of the original polylithionite remains preserved in alignment within the cryolite (Figure 7F).

Cryolite I (Figure 7D) is magmatic and, together with polylithionite, is one of the most abundant minerals in the pegmatites. It may form massive cryolite zones or crystal aggregates, especially in the cryolite-rich PEG, or occur disseminated throughout the rock, interstitially with other minerals and displaying sharp contacts with all magmatic phases. Macroscopically, the crystals are anhedral and black or caramel-colored. Under the optical microscope, cryolite I is colorless and exhibits first-order birefringence. Under crossed polarizers, it presents variable appearances (Figure 7A,D,G–I) and is predominantly twinned, with complex twinning commonly observed (Figure 7G,H). Cryolite I may also occur as oriented inclusions (Figure 8H) within several minerals and, in rare cases where it is strongly oriented (Figure 8I), constitutes the main component of the matrix, showing no corrosion features with any other mineral. Cryolite II is hydrothermal in origin, occurs interstitially among the aforementioned minerals or within fractures, exhibits corrosion features with nearly all minerals, and very often contains abundant microinclusions (Figure 7I).

Gagarinite-(Y) occurs as anhedral crystals ranging from 0.2 mm to 4.0 cm in size, pink in natural light, and is more frequent in cryolite-rich pegmatites. It is intensely affected by corrosion at the contact with cryolite II (Figure 8C). Xenotime is abundant, appearing as brown prismatic crystals (0.4 mm–7 cm, Figure 6A), scattered throughout the rock either as isolated grains or intergrown aggregates. It may contain inclusions of zircon, pyrochlore, cassiterite, albite, microcline, or quartz (Figure 7A). Intergrowths with thorite and zircon are common (Figure 8A), and less frequently with gagarinite (Figure 8B). Thorite occurs as dark, fully opaque, elongated crystals, usually ranging from 2 to 5 mm but occasionally reaching up to 4 cm in length. It is commonly found in the matrix and frequently as inclusions within large polylithionite (Figure 8A) and xenotime (Figure 8B) crystals. The association of thorite with xenotime and zircon is evidenced by their intergrowths, particularly with zircon. Thorite contacts with primary pegmatitic minerals—such as polylithionite, riebeckite, and pyrochlore—are generally abrupt; however, its contact with cryolite II shows characteristic corrosive features. Zircon is present in minor amounts, as euhedral to subhedral crystals (0.2 to 2.0 mm), sometimes zoned (Figure 8A), and consistently showing corrosion features at contacts with cryolite II. Cassiterite (Figure 8D,E) usually occurs as disseminated subhedral grains up to 0.8 mm in size, though larger crystals reaching up to 3 cm may also be found. Galena, commonly associated with minor sphalerite and inclusions of native Bi (Figure 8F), is generally observed as ~0.3 mm crystals, although larger crystals up to 10 cm may occur. Galena also appears as small crystals associated with the alteration of pyrochlore.

Genthelvite occurs predominantly as massive crystals (up to 7 cm), surrounding polylithionite and quartz I phenocrysts, and often includes pyrochlore, thorite, and zircon (Figure 9A). Less commonly, genthelvite is found interstitially in the matrix, associated with quartz I and orthoclase (Figure 9B). Its contacts with polylithionite, quartz I, pyrochlore, zircon, and thorite are reactive. Genthelvite is marked by corrosion features, such as cavities and microfractures, which are commonly filled by cryolite II (Figure 9C).

Pyrochlore occurs as single crystals with dimensions ranging from 0.1 to 0.7 mm. It predominantly appears as incipient to moderately opaque grains, due to alteration to columbite, while the remaining translucent portions of pyrochlore are brownish-orange (Figure 10A,B). Pyrochlore is commonly found in the matrix (Figure 10A,B) and is also observed as inclusions in large crystals of quartz, polylithionite (Figure 10C,D), xenotime (Figure 10D), and orthoclase (Figure 10E). Polylithionite and matrix quartz fill cavities within pyrochlore (Figure 10C,F). Crystals surrounded by cryolite II (Figure 10B,C,E) display highly reactive contact and are often associated with secondary pyrochlore, columbite, galena, LREE-rich fluorides, and HREE-Y-U-Th-rich silicates. Pyrochlore from these pegmatites is similar to that from the CAG and BAG [23,52] in terms of composition, size, shape, color, occurrence, and alteration. Therefore, it is suggested that the pyrochlore in the pegmatite veins is inherited from the magmatic phase of the albite-enriched granite.

Magnetite is primary and occurs locally in small amounts as euhedral crystals, associated with heterogeneous cryptocrystalline masses resulting from the alteration of other pegmatite minerals. The alteration of amphiboles and polylithionite (Figure 8G) generated chlorite. The alteration of amphiboles, Fe-Li-rich annite, polylithionite, and feldspars by late hydrothermal fluids [53] produced clay minerals and quartz II (microcrystalline), which also occurs as very thin veinlets that cut through various minerals of the pegmatite veins.

4.3. Pegmatitic Core Albite-Enriched Granite

4.3.1. Structure and Texture

The dykes and lenses of pegmatitic CAG (Figure 11A) are primarily found in the central and northern parts of the CAG. They extend up to 10 m in length and up to 50 cm in width. Drilling suggests that they are thicker (up to 15 m) at greater depths. These granitic dykes are typically oriented along the N70° E/40° N plane, nearly parallel to the strike of the magmatic foliations. Locally, the relationship between dyke margins indicates right-lateral displacement [54]. The contact with the host CAG is usually abrupt, although gradational contacts may also be observed. A characteristic feature of these pegmatites is the association of pegmatitic xenotime with megacrystals and polylithionite clusters (Figure 11B,C). More rarely, pegmatitic xenotime occurs in a quartzfeldspathiv zone (Figure 11D). The most common type of pegmatitic CAG has a matrix coarser than that of CAG but is quite similar to this rock in terms of mineralogical composition.

4.3.2. Mineral Assemblage

The pegmatitic minerals include quartz, riebeckite, Fe-Li-rich annite, polylithionite (typically ~2 cm but can reach 10 cm), xenotime (up to 5 cm), and thorite (up to 3 cm). Pegmatitic crystals of cryolite and albite mega crystals are uncommon. The matrix is mainly composed of albite, quartz, orthoclase, and microcline, with dispersed riebeckite, Fe-Li-rich annite, polylithionite, zircon, and subordinate xenotime, thorite, cassiterite, pyrochlore, and cryolite.

Quartz in the matrix (Figure 12A) also occurs as primitive poikilitic phenocrysts, up to 5 mm, anhedral to rounded, showing corrosion features in contact with albite and cryolite II. K-feldspar pegmatitic crystals are cloudy, often containing inclusions of polylithionite. Albite occurs mainly in the matrix as euhedral to subhedral, limpid crystals, with diffuse to clear twinning, ranging in sizes from 0.04 mm to 0.4 mm. Albite can also be found as inclusions or along the edges of K-feldspar. Fe-Li-rich annite in the matrix is commonly altered to polylithionite (Figure 12B).

Disseminated cryolite occurs in two generations, both with a low refractive index, nearly isotropic, and rarely twinned. Cryolite I appears as subhedral to anhedral (often rounded) crystals ranging from 0.02 mm to 1.0 mm, disseminated in the matrix and showing no corrosion features with surrounding minerals. Cryolite II forms irregular to rounded aggregates that fill spaces at the edges of other minerals, displaying corrosion features.

Zircon occurs in two distinct forms. Early zircon is predominantly skeletal and mostly enclosed within other minerals, suggesting it is inherited from the host rock. Second-generation zircon (Figure 12C) appears as euhedral to subhedral individual crystals (ranging from 0.1 mm to 1.5 mm) or as aggregates up to 1 cm. Thorite in the matrix (Figure 12D) is present as dispersed individual crystals (up to 0.40 mm), while in pegmatitic zones, larger crystals (up to 4 cm) form irregular concentrations. Cassiterite occurs in aggregates (Figure 12E) and as individual crystals (Figure 12F), ranging up to 0.5 cm in size. Crystals are usually subhedral to euhedral, though more rarely anhedral, with a reddish-brown color and commonly intense fracturing. Almost all crystals exhibit zoning, identified by color variation (white rims, transitioning to red, and brown cores). Contacts with cryolite and albite in the matrix often show corrosion features, although the original shape of larger crystals is usually preserved.

Xenotime is found both in the matrix (Figure 12E) and as pegmatitic crystals (Figure 11B–D). In both settings, it has a brown to pink color, prismatic habit, and varies from euhedral to subhedral, with sizes ranging from 0.05 mm to 5 cm. Pegmatitic crystals can occur individually or in clusters, while smaller matrix crystals are dispersed. Xenotime commonly contains inclusions, mainly of pyrochlore, thorite, and matrix crystals, which may also have corrosive contact with xenotime crystals. In the matrix, xenotime is typically found in areas enriched in zircon, cassiterite, thorite, and polylithionite. Xenotime crystallized after pyrochlore and early zircon, concurrently with thorite, and partially synchronous with late zircon.

4.4. Border Pegmatites

4.4.1. Structure and Texture

The border pegmatites are positioned between the BAG and the host rock—either amphibole-biotite granite or biotite granite—depending on their position relative to the pluton (Figure 2). In the case of the eastern border pegmatite, the BAG has no mappable thickness, and its subsurface existence was confirmed through drilling. These pegmatites can reach up to 400 m in length, with thicknesses commonly ranging from 0.5 to 4 m, and locally reaching up to 20 m. The contact with the external host rock is typically abrupt, while the interface with the BAG may be either abrupt or gradational.

4.4.2. Mineral Assemblage

The border pegmatite is typically red (Figure 13), though locally it may appear yellow or gray due to hydrothermal alteration. The main pegmatitic minerals are K-feldspar (microcline and orthoclase), which are predominant and occur in crystals up to 7 cm in size, and quartz (up to 5 cm). Fe-Li-rich annite crystals (up to 2 cm) are sparsely distributed. Clusters of zircon, up to 1 cm in size, are common. The matrix is medium- to coarse-grained and mainly composed of microcline and quartz, with minor albite, zircon, and thorite. Disseminated within the matrix are Fe-Li-rich annite, polylithionite, riebeckite, fluorite, pyrite, cassiterite, and pyrochlore. Fe-Li-rich annite and riebeckite are intensely chloritized. Zircon is subhedral, consistently highly altered, and significantly more abundant than in the pegmatite veins.

Pyrochlore in the border pegmatites occurs as single crystals (0.2 to 0.5 mm) disseminated in the quartz-feldspathic matrix. In the eastern border pegmatites, euhedral light-yellow crystals (Figure 14A) with incipient alteration to columbite are predominant, alongside subhedral orange crystals with moderate alteration (Figure 14B,C). In contrast, the northern border pegmatites mainly feature anhedral, dark, opaque grains showing advanced alteration to columbite (Figure 14D–F), frequently associated with secondary U-Th-HREE-Y-rich silicates and LREE-rich fluorides. Pyrochlore is commonly associated with zircon and, more rarely, with polylithionite (Figure 14B). A distinctive feature is the overgrowth of late zircon at pyrochlore surface, showing abrupt, rectilinear contacts (Figure 14D). Pyrochlore-matrix contacts are reactive and typically surrounded by halos of columbite or iron oxide. Hydrothermal fluorite is consistently present around pyrochlore and columbite grains (Figure 14A–E), with characteristic corrosion textures at the interface. These petrographic relationships indicate the magmatic origin of pyrochlore, followed by alteration by F-rich hydrothermal fluids. The features observed are similar to those of the pegmatite veins, suggesting that the pyrochlore in both eastern and northern border pegmatites is inherited from the BAG.

Thorite crystals ranges from 1 to 2 mm in size and frequently display halos formed by galena (Figure 15). In the eastern border pegmatites, thorite is typically translucent, while in the northern and southern sectors it is more often opaque. Geminated thorite crystals (Figure 15A) and associations with zircon (Figure 15B) are common, but associations with xenotime are absent. Contacts with fluorite are frequent, though they do not corrode or alter the thorite crystals.

Fluorite, pyrite (commonly with galena and molybdenite along grain edges), and a second generation of quartz (anhedral crystals up to 15 mm containing inclusions of matrix minerals and reddish-brown Fe-oxides needles) are late-stage minerals that occur disseminated in the matrix and in veins cutting the host rock. Hematite is associated with fluorite and also occurs finely disseminated throughout the rock. The northern border pegmatite shows significantly more intense hydrothermal alteration compared to both the eastern border pegmatite and the pegmatite veins.

4.5. Chemical Studies of Selected Minerals

4.5.1. Thorite

In the CAG, BAG, pegmatite veins, and border pegmatites, thorite (ThSiO₄) is highly hydrated, with low average Th concentrations (48 wt.% ThO₂), and high contents of Fe (0.11 to 29.56 wt.% Fe₂O₃) and F (up to 6.02 wt.% F) [37]. The most common variety in the CAG and BAG is a Zr-Fe-rich thorite. Primary thorite from the pegmatite veins is systematically richer in Y and REEs than that from the CAG and BAG. Additionally, a primary Y-U-(Fe)-rich thorite was identified exclusively in the northern border pegmatite, while a hydrothermal Y-Al-Fe-rich thorite was observed only in the pegmatite veins.

4.5.2. Xenotime

Xenotime (YPO₄) was analyzed in the CAG and pegmatitic CAG by Bastos Neto et al. [22], and in the polylithionite-rich PEG by Paludo et al. [43]. Xenotime from the CAG displays the highest average REE content (38.65 wt.% HREE₂O₃ and 4.14 wt.% LREE₂O₅), whereas xenotime from the polylithionite-rich PEG has the highest Y content (30.93 wt.% Y₂O₃). The highest average F content is also found in CAG xenotime (2.83 wt.% F), and the lowest in xenotime from the pegmatitic CAG (1.35 wt.%). The LREE/HREE and Th/U ratios of xenotime decrease progressively from CAG > pegmatitic CAG > polylithionite-rich PEG. Zircon, coffinite, and thorite components in xenotime are minor and subordinated.

4.5.3. Genthelvite

Genthelvite from the pegmatite veins shows homogeneous compositions, with limited substitutions within the helvine–genthelvite–danalite solid solution system. It is characterized by high Zn contents (36.96 to 49.45 wt.% ZnO), low Mn (0.61 to 3.03 wt.% MnO), and variable Fe (2.10 to 10.94 wt.% FeO). Remarkable features include elevated U (0.13 to 0.25 wt.% UO2) and REE concentrations (up to 0.40 wt.% REE₂O₃), with LREE prevailing over HREE on average [51].

4.5.4. Gagarinite

Gagarinite-(Y) [(NaCaY(F,Cl)₆] was first identified in the CAG by Minuzzi [36] and later studied in detail by Pires et al. [38]. Crystals occur in the central CAG, associated with fluocerite-(Ce) exsolutions. In the cryolite-rich PEG, gagarinite-(Y) was also reported by Paludo et al. [43], but without exsolutions. Average compositions and structural formula of these minerals are presented in

Table 1. EPMA average compositions (wt.%) of exsolved fluocerite-(Ce) and host gagarinite-(Y) from the core albite-enriched granite (CAG), and gagarinite-(Y) from the cryolite-rich pegmatite vein (PEG).

	CAG Fluocerite a		CAG Gagarinite a		Cryolite-Rich PEG Gagarinite b
	Range	2σ	Range	2σ	Range	2σ
	n c = 24		n = 16		n = 25
U	d.l.		d.l.		0.20	0.05
Th	d.l.		d.l.		0.17	0.10
Y	0.36	0.53	31.12	1.32	25.31	1.78
HREE	0.48	2.31	12.15	1.37	15.66	0.82
LREE	66.15	3.21	9.03	3.05	7.12	1.66
Ca	0.14	0.71	8.10	0.48	7.61	1.70
Pb	d.l.		d.l.		0.25	0.08
Sr	d.l.		d.l.		0.14	0.10
Na	d.l.		1.90	1.01	3.19	1.38
F	35.67	3.59	38.22	3.27	42.29	1.42
Total	102.80	4.22	100.52	3.03	101.95	2.59
Structural formula in a.p.f.u.
U					0.003	0.001
Th					0.002	0.001
Y	0.010	0.010	1.009	0.051	0.917	0.087
HREE	0.003	0.029	0.210	0.030	0.302	0.023
LREE	0.977	0.070	0.139	0.057	0.158	0.034
Ca	0.007	0.037	0.584	0.028	0.609	0.094
Pb					0.004	0.002
Sr					0.005	0.004
Na	0.001	0.012	0.238	0.127	0.450	0.226
F	3.883	0.448	5.808	0.574	5.840	0.288
LREE/HREE	289.08	243.86	0.75	0.40	0.45	0.10

^a Pires et al. [38], ^b Paludo et al. [43]. ^c n = number of analyses. Fluocerite structural formula calculated based on one cation. Gagarinite structural formula calculated based on Y + REE + Ca = 2. Abbreviation: d.l. = below detection limit.

. In the CAG, gagarinite-(Y) presents higher average contents of Y (31.12 wt.%), LREE (9.03 wt.%), and Ca (8.10 wt.%), whereas in the cryolite-rich PEG it is richer in HREE (15.66 wt.%), Na (3.19 wt.%), and F (42.29 wt.%). Fluorine and Na, and Na and HREE exhibit positive correlations (Figure 16A,B), while Ca has a strong negative correlation with F and with Y+REE (Figure 16C,D). Moderate negative correlations exist between Y and LREE (Figure 16E), and between HREE and LREE (Figure 16F) in the CAG gagarinite-(Y).

The chondrite-normalized REE pattern [55] (Figure 17) of gagarinite-(Y) from the CAG shows a strong LREE depletion, especially in La and Ce, compared to gagarinite from the cryolite-rich PEG. The exsolved fluocerite-(Ce) presents HREE depletion relative to LREE. The calculated composition of the earliest gagarinite-(Y) in the CAG (prior to fluocerite exsolution), based on densities and modal proportions of gagarinite-(Y) and fluocerite-(Ce), exhibit a flat REE pattern (Figure 17). In contrast, gagarinite-(Y) from the cryolite-rich PEG is enriched in HREE and has lower LREE contents than the earliest gagarinite-(Y) in the CAG.

4.5.5. Riebeckite

Average compositions and structural formulas for riebeckite from the amphibole-rich PEG and the CAG are presented in

Table 2. Average EPMA data (in wt.%) for riebeckite from the core albite-enriched granite (CAG) and the amphibole-rich pegmatite vein (PEG).

	CAG a		Amphibole-Rich PEG b
	Mean	2σ	Mean	2σ
	n c = 43		n = 19
SiO2	49.38	1.27	51.14	1.20
TiO2	0.16	0.39	0.11	0.25
Al2O3	0.75	0.42	1.03	0.54
Fe2O3	-	-	4.01	8.60
FeO	-	-	25.95	3.89
FeOT	34.17	3.20	-	-
MnO	0.47	0.54	0.78	0.34
ZnO	2.09	2.64	2.31	0.95
Na2O	7.43	0.48	7.63	1.51
K2O	0.27	0.25	1.02	0.79
F	0.67	0.27	2.12	2.27
Cl	d.l.	-	0.01	0.02
H2O *	1.58	0.14	0.83	1.12
O = F, Cl	−0.28	0.12	−0.90	0.96
Total	96.68	1.93	96.42	3.20
Structural formula based on 23 oxygens (a.p.f.u.)
Si4+	7.930	0.110	8.343	0.226
IVTi4+	0.001	0.014	0.000	0.000
IVAl3+	0.070	0.097	0.000	0.000
IVFe3+	0.001	0.010	0.000	0.000
SumT	8.002	0.015	8.343	0.226
Ti4+	0.018	0.044	0.013	0.030
VIAl3+	0.073	0.100	0.199	0.105
Fe3+	1.585	0.285	0.486	1.034
Fe2+	2.996	0.334	3.543	0.609
Mn2+	0.064	0.073	0.108	0.048
Zn2+	0.255	0.308	0.278	0.119
SumC	4.991	0.092	4.614	0.278
Na+B	2.000	0.000	1.998	0.017
Na+	0.313	0.162	0.417	0.530
K+	0.054	0.047	0.213	0.168
SumA	0.376	0.223	0.631	0.647
F−	0.339	0.142	1.102	1.195
Cl−	0.000	0.000	0.002	0.005
OH− *	1.661	0.142	0.895	1.196

^a Schuck (2015), ^b Paludo et al. [43]. ^c n = number of analysis. * OH calculated after Hawthorne et al. [56]. Abbreviation: d.l. = below detection limit.

. Riebeckite from the amphibole-rich PEG has significantly higher averages of F (2.12 wt.%) than that from the CAG (0.67 wt.% F). It also shows slightly higher average contents of Si, Al, K, Na, and Zn, while the CAG riebeckite is richer in total Fe. In the riebeckite from the amphibole-rich PEG, fluorine has strong positive correlation with Si (Figure 18A) and Na (Figure 18B), correlations absent in the CAG riebeckite. Fluorine also shows a positive correlation with K (Figure 18C) in both cases. The Fe³⁺ shows a strong negative correlation with F (Figure 18D) and a moderate one with Al (Figure 18E). Manganese and Zn display a good positive correlation, likely reflecting substitution for Fe²⁺.

4.5.6. Polylithionite

EPMA data and structural formulas for polylithionite from the pegmatite veins and various elevations (altitude quota) of the CAG are presented in

Table 3. EPMA data (in wt.%) of average polylithionite from the core albite-enriched granite (CAG) at different altimetric quotas (120, 140, and 160 m) and from surface samples (~200–220 m) of the pegmatite veins (PEG): amphibole-rich, polylithionite-rich, and cryolite-rich.

	CAG 120 m a		CAG 140 m a		CAG 160 m a		Amphibole-Rich PEG b		Polylithionite-rich PEG		Cryolite-Rich PEG
	Mean	2σ	Mean	2σ	Mean	2σ	Mean	2σ	Mean	2σ	Mean	2σ
	nc = 11		n = 14		n = 20		n1 = 32; n2 = 15 d		n = 13		n = 10
SiO2	53.32	1.71	52.79	1.70	52.34	1.71	55.82	3.77	56.16	2.63	55.05	1.24
TiO2	0.10	0.06	0.12	0.06	0.13	0.05	0.10	0.23	0.09	0.16	0.13	0.22
UO2	n.a.	-	n.a.	-	n.a.	-	1.76 *	4.14	1.04	0.06	1.01	0.09
Al2O3	12.49	0.58	12.63	0.71	12.40	0.65	12.22	1.35	12.61	0.42	12.42	0.38
HREE2O3	n.a.	-	n.a.	-	n.a.	-	0.07	0.13	0.12	0.15	0.08	0.07
LREE2O3	n.a.	-	n.a.	-	n.a.	-	0.09	0.17	0.14	0.13	0.10	0.12
FeO	7.55	2.02	6.37	1.88	6.08	1.48	5.86	3.83	7.31	2.97	8.50	1.24
MnO	0.22	0.09	0.34	0.13	0.24	0.38	0.15	0.19	0.12	0.18	0.16	0.11
ZnO	1.14	0.33	1.21	0.43	2.51	0.72	0.54	0.61	0.25	0.18	0.72	0.20
Li2O e	5.75	0.49	5.60	0.49	5.47	0.49	6.47	1.08	6.57	0.75	6.25	0.36
Na2O	0.02	0.04	0.04	0.13	0.01	0.04	0.06	0.09	0.09	0.07	0.10	0.05
K2O	8.57	0.41	8.06	0.38	8.24	0.24	7.48	2.74	9.27	0.48	9.01	0.20
Rb2O	4.62	0.65	5.57	0.00	5.46	0.04	3.92	3.63	n.a.	-	n.a.	-
F	6.40	0.68	6.70	0.00	7.53	0.65	9.18	1.10	9.26	0.48	8.92	0.37
F=O2	2.70	0.29	2.82	0.00	3.17	0.28	−3.86	0.46	3.90	0.20	3.76	0.16
Total	97.51	1.71	96.60	1.19	97.23	1.82	97.73	4.32	103.03	1.22	98.69	1.39
Structural formula based on 11 Oxygens (a.p.f.u.)
Si4+	3.877	0.035	3.874	0.034	3.860	0.023	3.951	0.071	3.919	0.049	3.904	0.029
Ti4+	0.006	0.004	0.006	0.004	0.007	0.003	0.006	0.012	0.005	0.009	0.007	0.012
IVAl3+	0.123	0.035	0.126	0.034	0.140	0.023	0.052	0.060	0.081	0.049	0.096	0.029
∑IV	4.006	0.004	4.006	0.004	4.007	0.003	4.008	0.022	4.005	0.009	4.007	0.012
U4+							0.028	0.066	0.016	0.001	0.016	0.001
VIAl3+	0.947	0.027	0.966	0.035	0.938	0.028	0.968	0.066	0.957	0.018	0.942	0.020
HREE3+							0.002	0.004	0.003	0.005	0.002	0.003
LREE3+							0.002	0.004	0.003	0.003	0.003	0.003
Fe2+	0.207	0.060	0.176	0.056	0.169	0.045	0.157	0.108	0.193	0.083	0.227	0.035
Mn2+	0.014	0.006	0.021	0.008	0.015	0.024	0.009	0.012	0.007	0.011	0.009	0.006
Zn2+	0.062	0.019	0.066	0.025	0.136	0.040	0.026	0.035	0.013	0.010	0.038	0.011
Li+	1.681	0.099	1.652	0.100	1.621	0.094	1.838	0.193	1.842	0.145	1.782	0.071
∑VI	2.911	0.046	2.881	0.046	2.880	0.053	3.031	0.098	3.034	0.057	3.019	0.046
Na+	0.003	0.006	0.005	0.018	0.001	0.005	0.008	0.012	0.012	0.009	0.014	0.007
K+	0.795	0.030	0.755	0.030	0.775	0.024	0.675	0.231	0.825	0.036	0.815	0.023
Rb+	0.216	0.031	0.263	0.007	0.259	0.008	0.085	0.213
∑XII	1.014	0.039	1.023	0.026	1.035	0.032	0.767	0.079	0.837	0.038	0.829	0.025
OH− f	0.527	0.159	0.446	0.043	0.244	0.160	0.009	0.041	0.001	0.010	0.012	0.040
F−	1.473	0.159	1.554	0.043	1.756	0.160	2.053	0.167	2.044	0.072	2.001	0.069
Mn/Mn + Fe	0.063	0.015	0.107	0.026	0.076	0.116	0.050	0.057	0.034	0.045	0.039	0.026
K/Rb	1.69	0.24	1.31	0.06	1.37	0.04	1.62	0.97
LREE/HREE							2.73	7.34	2.96	9.51	1.82	3.32

^a Costi [31]. ^b Paludo et al. [43]. ^c n = number of analysis. ^d Analyses with Rb₂O determination. ^e LiO₂ calculated after Tindle and Webb [57], ^f Calculated. Abbreviations: n = number of samples, n.a = not analyzed. * The mean value of 1.76 wt.% UO₂ includes two main groups, one averaging 0.61 wt.% UO₂ (n = 24) and the other with an average of 5.52 wt.% UO₂ (n = 8).

. Lithium average content ranges from 5.47 to 5.75 wt.% in the CAG and 6.25 to 6.57 wt.% in the pegmatite veins. A strong negative correlation between Li and Fe (Figure 19A) indicates Fe substitution for Li (~0.2 apfu) in the octahedral site, and a moderate negative correlation exists between Li and Al (Figure 19B). In the polylithionite from the CAG, F content increases with altitude (6.40 at 120 m to 7.53 wt.% F at 160 m), while in the pegmatite veins it becomes significatively higher (8.92 to 9.26 wt.% F), where F fully occupies the (OH, F)-site (~2 apfu). In contrast, the F-OH ratio in the CAG polylithionite is near 1.5–0.5 apfu. No F-Fe avoidance effect is observed (Figure 19C). The highest Zn (up to 2.51 wt.%) is found in the CAG samples, with Zn and Mn showing a positive correlation (Figure 19D), increasing towards shallower levels of the CAG but decreasing in the pegmatite veins. The ^IVAl is present in all samples, with Al decreasing and Si increasing in polylithionite from the pegmatite veins compared to the CAG (Figure 19E). The highest K (9.27 wt.% K₂O) is observed in the polylithionite from the polylithionite-rich PEG, while the lowest K (7.48 wt.% K₂O) occurs in the polylithionite from the amphibole-rich PEG, where a set of low-K crystals is found (Figure 19F). Rubidium was measured only in the low-K group from the amphibole-rich PEG, averaging 3.92 wt.% Rb₂O. Polylithionite from the CAG has significantly higher Rb₂O, ranging from 4.62 to 5.57 wt.%.

Uranium and REE contents were determined only in polylithionite from the pegmatite veins. The lowest and highest U contents occur in the polylithionite from the amphibole-rich PEG: the predominant group (n = 24) has an average of 0.61 wt.% UO₂, while a smaller group (n = 8) shows an anomalous high U content, averaging 5.52 wt.% UO₂. The highest REE averages are found in polylithionite from the polylithionite-rich PEG, with 0.12 wt.% HREE₂O₃ and 0.14 wt.% LREE₂O₃. The REE normalized pattern (Figure 20A) for polylithionite from all pegmatite vein types displays a strong M-type tetrad effect [58] and a positive Eu anomaly. Generally, the LREE are more abundant, with LREE/HREE ratios (Figure 20B) of 1.82 in the cryolite-rich PEG, 2.73 in the amphibole-rich PEG, and 2.96 in the polylithionite-rich PEG.

4.5.7. Pyrochlore

Representative compositions and structural formulas of pyrochlore are presented in

Table 4. EPMA data (in wt.%) for pyrochlore: (1) U-Pb-LREE-rich pyrochlore; (2) LREE-U-Pb-rich pyrochlore; (3) Fe-U-rich pyrochlore; (4) LREE-Pb-rich pyrochlore; (5) U-Pb-rich pyrochlore; (6) Na-LREE-Pb-rich pyrochlore; (7) Na-Pb-LREE-rich pyrochlore; (8) Fe-U-Pb-rich pyrochlore; (9) HREE-Y-U-Pb-rich pyrochlore; (10) Ca-Fe-U-Pb-rich pyrochlore; (11) Ca-Fe-Pb-U-rich pyrochlore.

	CAG 1		BAG 1	Amphibole-Rich PEG				Northern Border Pegmatite		Eastern Border Pegmatite
Crystal	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Nb2O5	46.20	40.10	31.93	47.45	33.20	43.53	43.66	34.36	24.32	34.29	33.83
Ta2O5	06.13	02.47	01.69	03.44	04.04	16.99	15.17	12.86	09.23	05.43	05.04
SiO2	00.21	00.42	13.82	01.09	04.89	00.68	00.58	09.37	16.92	14.61	15.27
SnO2	01.16	00.72	d.l.	01.53	00.78	01.12	00.62	00.23	00.24	00.57	00.61
TiO2	01.05	00.92	01.45	00.84	d.l.	00.36	00.28	02.24	02.03	00.00	00.62
UO2	02.81	06.97	12.64	00.04	08.86	01.13	01.13	06.40	09.76	04.79	05.72
ThO2	01.77	00.49	00.84	00.73	00.35	00.90	00.98	00.00	00.66	01.44	00.91
Y2O3	01.00	00.13	00.25	00.73	00.24	00.68	00.65	01.72	03.53	00.18	00.29
HREE2O3	00.91	00.20	01.63	00.38	00.07	01.21	01.91	00.24	02.72	00.27	00.26
LREE2O3	07.02	04.08	01.88	03.87	00.26	07.14	08.05	00.66	00.89	02.32	03.28
FeO (2)	00.69	01.70	03.24	00.38	00.00	00.23	00.04	02.70	01.90	03.27	03.76
CaO	01.42	01.04	00.34	00.96	00.00	00.77	01.48	00.61	00.19	02.48	01.78
MnO	00.11	00.23	00.49	02.60	01.33	00.08	00.00	00.12	00.29	00.45	00.41
PbO	07.23	13.98	00.02	23.72	28.87	09.01	05.27	22.00	14.91	05.53	04.93
Na2O	00.76	00.18	00.31	00.24	00.21	02.80	04.42	d.l.	00.07	00.75	00.27
F	02.73	00.85	00.21	01.51	00.31	04.35	04.47	00.00	00.00	00.14	00.08
F=O2	−01.15	−00.36	−00.09	−00.64	−00.13	−01.83	−01.88	−00.00	−00.00	−00.06	−00.03
Total	80.08	74.18	69.34	91.00	83.29	89.18	86.86	93.29	85.49	76.22	76.70
Structural formula based on a sum of 2 a.p.f.u. in the [6]B site
U4+	0.052	0.154	0.189	0.033	0.185	0.020	0.020	0.094	0.135	0.067	0.078
Th4+	0.034	0.011	0.013	0.036	0.008	0.016	0.018		0.009	0.021	0.013
Y3+	0.044	0.007	0.009	0.033	0.012	0.028	0.028	0.061	0.117	0.006	0.009
HREE3+	0.024	0.006	0.034	0.01	0.002	0.03	0.049	0.005	0.053	0.005	0.005
LREE3+	0.213	0.149	0.046	0.117	0.009	0.203	0.235	0.016	0.021	0.054	0.073
Pb2+	0.162	0.373		0.531	0.730	0.189	0.114	0.393	0.251	0.094	0.081
Fe2+	0.048	0.141	0.182	0.027		0.015	0.003	0.150	0.099	0.172	0.193
Mn2+	0.008	0.020	0.028	0.183	0.106	0.005		0.007	0.015	0.024	0.021
Ca2+	0.127	0.111	0.024	0.086		0.064	0.127	0.044	0.013	0.167	0.117
Na+	0.123	0.035	0.041	0.039	0.039	0.422	0.689		0.009	0.092	0.033
Ʃ [8]A	0.835	1.006	0.565	1.094	1.090	0.992	1.284	0.769	0.722	0.701	0.624
Nb5+	1.739	1.795	0.968	1.781	1.408	1.531	1.585	1.029	0.685	0.973	0.936
Ta5+	0.139	0.067	0.031	0.078	0.103	0.360	0.332	0.232	0.157	0.093	0.084
Si4+	0.017	0.042	0.929	0.091	0.460	0.053	0.047	0.622	1.057	0.919	0.937
Sn4+	0.039	0.028		0.051	0.029	0.035	0.020	0.006	0.006	0.014	0.015
Ti4+	0.066	0.069	0.073			0.021	0.017	0.111	0.095		0.029
Ʃ [6]B	2.000	2.000	2.000	2.000	2.000	2.000	2.000	2.000	2.000	2.000	2.000
O2−	4.878	5.328	3.580	5.305	5.061	4.855	5.320	4.068	3.759	3.617	3.504
F−						0.071	0.137
OH−	1.122	0.672	2.420	0.695	0.939	1.074	0.543	1.932	2.241	2.383	3.498
ƩX	6.000	6.000	6.000	6.000	6.000	6.000	6.000	6.000	6.000	6.000	6.000
F−	0.721	0.268	0.046	0.398	0.091	1.000	1.000			0.027	0.016
OH−	0.279	0.732	0.954	0.602	0.909	0		1.000	1.000	0.973	0.984
ƩY	1.000	1.000	1.000	1.000	1.000	1.000	1.000	1.000	1.000	1.000	1.000
Nb/Ta	12.519	26.957	31.336	22.914	13.633	4.257	4.780	4.440	4.378	10.482	11.152
Fe/Mn	6.064	7.169	6.567	0.145	0	2.949		22.838	6.524	7.219	9.019
LREE/HREE	8.875	24.833	1.352	11.700	4.500	6.766	4.795	3.200	0.396	10.800	14.600

¹ [52]. ⁽²⁾ Total Fe as FeO. Abbreviations: d.l. = below detection limit.

. In the CAG and BAG, the least altered varieties are U-Pb-LREE-rich pyrochlore (Table 4, crystal 1). The predominant variety in the CAG is U-Pb-rich pyrochlore (Table 4, crystal 2), while the Fe-U-rich pyrochlore (Table 4, crystal 3) is most common in the BAG and in the central CAG, where alteration was stronger near the massive cryolite deposit [52].

In the amphibole-rich PEG, remnants of LREE-Pb-rich pyrochlore (Table 4, crystal 4) are surrounded by U-Pb-rich pyrochlore (Table 4, crystal 5), showing reactive contacts. The loss of LREE is accompanied by decreasing F (Figure 21A) and a relative increase in Pb (Figure 21B). As alteration progresses, Pb is gradually lost, and U becomes relatively enriched (Figure 21C). This Pb-U-Si-enriched hydrothermal phase is the main pyrochlore variety in the amphibole-rich PEG. Additionally, grains with advanced alteration retain zones of compositionally zoned pyrochlore, grading from Na-Pb-LREE-rich (Table 4, crystal 6) at the core to Na-LREE-Pb-rich (Table 4, crystal 7) at the rim. Despite its occurrence under intense hydrothermally alteration, these pyrochlore varieties show high LREE (up to 8.05 wt.% LREE₂O₃), F (up to 4.47 wt.% F), and increasing Na contents up to 4.42 wt.% Na₂O (Figure 21D).

In highly altered pyrochlore grains from the northern border pegmatite, relicts of hydrothermal Fe-U-Pb-rich pyrochlore (Table 4, crystal 8) and HREE-Y-U-Pb-rich pyrochlore (Table 4, crystal 9) occur, the latter notable for elevated HREE (up to 2.72 HREE₂O₃) and Y contents (up to 3.53 wt.% Y₂O₃) (Figure 21E). In the eastern border pegmatite, incipient to moderately altered pyrochlore grains include the high Si Ca-Fe-U-Pb-rich pyrochlore (~14 wt.% SiO₂, Table 4, crystal 10) and Ca-Fe-Pb-U-rich pyrochlore (~15 wt.% SiO₂, Table 4, crystal 11). These varieties have CaO concentrations ranging from 1.78 to 2.48 wt.% higher than in pyrochlore from the amphibole-rich PEG (up to 1.48 wt.% CaO) and similar to the less altered grains in the CAG. At grain margins, in microfractures, and dispersed in the matrix, a hydrothermal, non-stoichiometric Ca-U-rich pyrochlore (

Table 5. EPMA data (in wt.%) for hydrothermal phases associated with pyrochlore alteration: (1) Ca-U-rich pyrochlore; (2) U-HREE-Y-Th-rich silicate; (3) LREE-rich fluoride; (4) HREE-Y-U-rich silicate; (5) HREE-U-Y-rich silicate; and (6) LREE-rich fluoride.

Facies	Eastern Border Pegmatite	Amphibole- Rich PEG		Northern Border Pegmatite
Crystal	(1)	(2)	(3)	(4)	(5)	(6)
Nb2O5	22.26	01.48	02.62	03.19	02.79	00.26
Ta2O5	00.27	00.38	01.05	00.00	01.17	00.00
P2O5	00.00	01.82	00.00	01.07	06.80	00.00
SiO2	00.93	09.51	00.10	14.76	16.24	00.03
UO2	34.56	01.57	00.30	29.23	12.48	00.31
ThO2	00.16	35.69	00.49	00.25	07.69	00.05
ZrO2	00.00	00.40	d.l.	00.00	00.00	00.00
Y2O3	00.17	03.02	00.22	15.09	13.34	00.28
HREE2O3	00.00	02.59	00.29	10.88	07.92	00.18
LREE2O3	00.30	00.59	55.40	00.33	00.40	56.26
FeO 1	01.75	01.28	00.17	00.50	00.27	00.00
CaO	04.71	00.88	d.l.	00.32	00.48	00.27
MnO	00.41	d.l.	d.l.	d.l.	00.26	00.16
PbO	00.31	01.16	d.l.	02.71	01.09	00.00
Na2O	00.37	d.l.	d.l.	00.00	d.l.	00.00
F	00.00	04.33	08.65	02.61	02.80	07.13
F=O2	−00.00	−01.82		−01.10	−01.18
Total	67.10	63.03	69.67	79.87	72.62	64.93

¹ Total Fe as FeO. Abbreviations: d.l. = below detection limit.

, crystal 1) is observed, with up to 34.56 wt.% UO₂ and 4.71 wt.% CaO (Figure 21C,D,F).

The average Nb/Ta ratio in pyrochlore (Figure 22A) is lowest in the northern border pegmatite (10.2), CAG (12.9), and amphibole-rich PEG (13.7), and highest in the BAG (23.5) and eastern border pegmatite (42.5). Nb/Ta ratios show a strong negative correlation with Ta content (Figure 21G), but no significant correlation with Nb. Extremely low Nb/Ta values (<5) are restricted to Na-enriched hydrothermal pyrochlore in the amphibole-rich PEG. Niobium correlates strongly and negatively with Si (Figure 21H) in all samples. The average Fe/Mn ratio (Figure 22B) is highest in the pyrochlore grains from the northern border pegmatite (14.9) and lowest in those from the amphibole-rich PEG (2.0), following a direct relationship with Mn content (Figure 21I). The average LREE/HREE ratio (Figure 22C) is lowest in pyrochlore grains from the northern border pegmatite (1.0) and highest in the CAG (10.2). Consistently, normalized REE patterns [55] (Figure 23) show the CAG pyrochlore with the highest absolute LREE and lowest HREE content, while pyrochlore from the eastern border pegmatite displays the lowest overall REE contents.

4.5.8. Columbite

In the CAG and BAG, pyrochlore breakdown led to the formation of columbite. Representative compositions and structural formulas of columbite are shown in

Table 6. EPMA data (in wt.%) for columbite: (1) Mn-Fe-rich columbite; (2) U-Mn-Fe-rich columbite; (3, 4) Mn-Fe-rich columbite; (5) Fe-Mn-rich columbite; and (6) U-Fe-Mn-rich columbite.

Facies	CAG 1	BAG 1	Amphibole- Rich PEG		Northern Border Pegmatite
Crystal	(1)	(2)	(3)	(4)	(5)	(6)
Nb2O5	66.74	65.61	73.87	73.32	66.51	65.04
Ta2O5	03.32	05.72	01.46	04.67	06.65	08.08
SiO2	00.51	00.57	00.05	00.13	00.20	00.25
SnO2	d.l.	d.l.	00.30	00.00	00.00	00.00
TiO2	02.57	02.36	00.84	00.43	00.15	01.61
UO2	01.15	03.64	00.73	00.28	00.32	01.28
ThO2	d.l.	00.18	00.04	00.00	d.l.	00.03
Y2O3	00.12	00.07	d.l.	00.09	00.00	d.l.
HREE2O3	00.00	00.62	00.15	00.23	00.26	00.45
LREE2O3	00.30	00.41	00.47	00.13	00.27	00.23
FeO 2	15.33	16.13	13.38	14.27	03.62	08.53
CaO	00.40	d.l.	00.00	00.00	d.l.	00.51
MnO	06.70	04.92	07.84	07.07	17.54	11.74
PbO	00.81	00.06	00.81	00.00	00.00	01.23
Na2O	d.l.	00.04	00.03	00.00	00.02	00.03
F	d.l.	d.l.	00.00	00.00	00.00	00.00
F=O2	−00.00	−00.00	−00.00	−00.00	−00.00	−00.00
Total	97.89	99.74	99.99	100.63	95.44	99.04
Fe2+	0.725	0.771	0.632	0.676	0.181	0.415
Mn2+	0.321	0.238	0.375	0.339	0.886	0.578
Ʃ [8]A	1.046	1.009	1.036	1.015	1.067	0.993
Nb5+	1.704	1.693	1.886	1.875	1.792	1.709
Ta5+	0.051	0.089	0.023	0.072	0.108	0.128
Si4+	0.029	0.032	0.003	0.007	0.012	0.015
Sn4+			0.007
Ti4+	0.109	0.101	0.036	0.018	0.007	0.070
U4+	0.014	0.046	0.009	0.004	0.004	0.017
Th4+		0.002	0.001
Y3+	0.004	0.002		0.003
HREE3+		0.011	0.003	0.004	0.005	0.008
LREE3+	0.006	0.008	0.010	0.003	0.004	0.005
Pb2+	0.012	0.001	0.012			0.019
Ca2+	0.025					0.032
Na+		0.004	0.004		0.002	0.004
Ʃ [8]B	1.954	1.991	1.964	1.985	1.933	2.007
O2−	5.576	5.694	5.837	5.895	5.738	5.693
OH− *	0.424	0.306	0.163	0.105	0.262	0.307
ƩX	6.000	6.000	6.000	6.000	6.000	6.000
Nb/Ta	33.413	19.041	83.732	26.087	16.617	13.366
Fe/Mn	2.260	3.236	1.685	1.993	0.204	0.717
LREE/HREE		0.756	3.746	0.672	0.772	0.602

¹ [52]. ² Total Fe as FeO. * Calculated. Abbreviations: d.l. = below detection limit.

. In the CAG, the predominant species is Mn-Fe-rich columbite (Table 6, crystal 1). In the BAG, U-Mn-Fe-rich columbite is relatively common (Table 6, crystal 2), with up to 3.64 wt.% UO₂. In the amphibole-rich PEG, only Mn-Fe-rich columbite occurs (Table 6, crystals 3, 4), filling cavities and microfractures in hydrothermal pyrochlore or surrounding its remnants. Columbite was not observed in the eastern border pegmatite, whereas in the northern border pegmatite, Fe-Mn-rich columbite (Table 6, crystal 5) and U-Fe-Mn-rich columbite (Table 6, crystal 6), with up to 1.28 wt.% UO₂, were found surrounding hydrothermal pyrochlore remnants.

The average Fe/Mn ratio (Figure 24A) in columbite decreases from the BAG (2.8) to the CAG (2.7), amphibole-rich PEG (2.0), and northern border pegmatite (0.5). The Fe/Mn distribution correlates better with Mn (Figure 25A) than with Fe (Figure 25B), with manganese showing negative correlation with Fe (Figure 25C). The average Nb/Ta ratio in columbite (Figure 24B) is lowest in the northern border pegmatite (14.8) and highest in the amphibole-rich PEG (41.7), with Nb/Ta variations linked to Ta content (Figure 25D). Silicon substitutes for Nb (Figure 25E) up to 0.015 apfu in the B-site, with lower Si content in pegmatite columbite than in the CAG and BAG. The LREE/HREE ratio is lowest in columbite from the northern border pegmatite (0.8) and BAG (1.9), and highest in the CAG (3.10) and amphibole-rich PEG (3.14), positively correlating with LREE content (Figure 25F). REE normalized patterns (Figure 26) show the lowest REE content in columbite from the CAG, followed by the amphibole-rich PEG, BAG, and northern border pegmatite.

4.5.9. Other Products of Pyrochlore Alteration

In the amphibole-rich PEG, besides columbite and galena, the most frequent secondary minerals associated with pyrochlore alteration are as follows: (i) U-HREE-Th-rich silicate (Table 4, crystal 2), with up to 35.68 wt.% ThO₂, 9.51 wt.% SiO₂, and 4.33 wt.% F, disseminated in columbite grains; and, (ii) LREE-rich fluorides (Table 4, crystal 3), partially replacing pyrochlore, surrounding hydrothermal pyrochlore remnants, and included in columbite. In the northern border pegmatite, columbite is associated with silicates enriched in HREE, Y, U, and Th in varying proportions (Table 4, crystals 4, 5), with up to 29.23 wt.% UO₂, 30.71 wt.% ThO₂, 15.09 wt.% Y₂O₃, and 10.88 wt.% HREE₂O₃. These silicates contain 14.76–19.60 wt.% SiO₂, 2.79–9.10 wt.% Nb₂O₅, and 1.45–2.80 wt.% F. Galena and LREE-rich fluorides (Table 4, crystal 6) also occur.

The REE normalized patterns (Figure 27A) of (HREE-Y-U-Th)-rich silicates from the northern border pegmatite and the amphibole-rich PEG are similar to those of the CAG and BAG, with flat HREE distributions. Silicates from the northern border pegmatite show the highest HREE contents, and those from the amphibole-rich PEG, the lowest. The REE-normalized patterns of LREE-rich fluorides (Figure 27B) are remarkably similar across all rocks.

Secondary pyrochlore, columbite, LREE-rich fluorides, silicate phases, and galena formed during the early hydrothermal stage. In the amphibole-rich PEG, columbitized grain borders are intensely dissolved, and columbite shows irregular, reactive contacts with hydrothermal cryolite, quartz, and iron oxides, which also fill internal cavities. These features and minerals are products of the late hydrothermal stage. In the border pegmatites, the late hydrothermal stage overprinted all earlier-formed minerals, corroded grain margins, and led to the crystallization of fluorite and quartz in grain borders and cavities of pyrochlore and columbite.

4.6. Whole Rock Geochemical Data

4.6.1. Trends of Compositional Variation

Whole-rock data are presented for the CAG, BAG, pegmatitic CAG, and border pegmatites, with emphasis on strategic major (

Table 7. Major element analyses (wt.%) of the CAG, BAG, border pegmatites, and pegmatite veins (PEG): amphibole-rich, polylithionite-rich, and cryolite-rich.

	CAG		BAG		Border Pegmatite 1		Amphibole-Rich PEG 2		Polylithionite-Rich PEG 2		Cryolite-Rich PEG 2
	Mean	2σ	Mean	2σ	Mean	2σ	Mean	2σ	Mean	2σ	Mean	2σ
	na = 64		n = 57		n = 5		n = 23		n = 11		n = 10
SiO2	69.95	5.81	72.31	11.60	73.39	3.98	66.92	6.82	58.35	18.22	12.41	24.55
TiO2	0.03	0.11	0.03	0.05	0.06	0.06	0.03	0.03	0.06	0.06	0.02	0.03
Al2O3	12.80	1.98	12.19	3.93	11.92	1.37	11.29	4.39	10.90	4.26	18.34	11.62
CaO	0.28	1.39	0.73	1.86	0.79	0.86	0.06	0.10	0.14	0.27	0.60	3.17
FeO 3	2.21	1.12	2.67	4.49	1.95	1.14	3.63	2.74	4.79	2.87	0.24	0.61
MgO	0.02	0.10	0.03	0.12	0.03	0.01	0.01	0.01	0.04	0.02	0.05	0.11
MnO	0.06	0.08	0.06	0.12	0.02	0.01	0.09	0.08	0.16	0.16	0.02	0.02
K2O	4.26	1.14	4.33	2.86	5.92	1.35	2.87	2.98	5.95	3.15	0.11	0.25
Na2O	5.55	3.23	3.87	3.20	2.95	1.05	6.62	3.15	3.13	4.26	33.01	21.06
P2O5	0.03	0.07	0.05	0.26	0.03	0.01	0.27	0.65	1.03	3.48	0.09	0.33
LOI	1.70	1.27	1.39	2.30	1.15	0.37	2.30	1.73	3.64	2.86	14.40	10.02
F	2.31	4.49	0.59	1.49	0.32	0.36	3.09	4.63	5.69	6.04	35.00	19.31
F=O	−0.97	1.89	−0.25	0.63	−0.13	0.15	−1.30	1.95	−2.39	2.54	−14.74	8.13
Total	97.91	2.84	97.88	4.56	98.39	0.67	95.88	4.93	91.48	9.38	99.54	29.91
Fe/Mn	46.64	37.61	53.59	49.62	79.58	31.55	42.66	20.94	34.44	31.90	10.91	28.61
A/CNK	1.28	0.24	1.39	0.36	1.24	0.13	1.19	0.37	1.22	0.41	0.54	0.11
A/NK	1.33	0.41	1.53	0.57	1.34	0.06	1.20	0.37	1.24	0.42	0.55	0.06

¹ Lengler [42], ² Paludo et al. [43]. ³ Total Fe as FeO. ^a n = number of analysis.

) and trace elements (

Table 8. Trace element analyses of the CAG, BAG, border pegmatites, pegmatitic CAG, and pegmatite veins (PEG): amphibole-rich, polylithionite-rich, and cryolite-rich.

	CAG		BAG		Border Pegmatite 1		Amphibole-Rich PEG 2		Polylithionite-Rich PEG 2		Cryolite-Rich PEG 2		Pegmatitic CAG
	Mean	2σ	Mean	2σ	Mean	2σ	Mean	2σ	Mean	2σ	Mean	2σ	Mean	2σ
	n1 a = 64; n2 = 133		n1 = 57; n2 = 72		n = 5		n = 23		n = 11		n = 10		n = 75
Nb5+	1578.88	2101.09	1354.75	1092.53	693.20	111.04	998.48	14.60	913.00	385.23	159.60	365.31	1979.65	1443.57
Ta5+	444.69	2656.00	231.34	311.97	90.88	22.68	237.57	129.92	194.95	224.58	3.37	8.60	441.72	390.08
Sn4+	1722.63	3336.91	1445.47	2203.54	92.20	20.25	988.35	111.76	751.27	708.93	262.50	658.59	2459.14	2353.19
U4+	293.04	382.26	311.34	321.27	261.32	127.98	290.67	232.01	71.41	104.82	1.80	4.39	511.91	518.60
Th4+	831.09	2670.89	714.90	1734.86	386.80	220.56	1779.22	936.88	1223.45	1867.51	193.48	560.56	5026.85	7318.83
Zr4+	5218.30	4153.28	4676.56	6759.94	5624.00	1505.69	5886.09	4853.42	939.45	1991.83	25.90	42.98	6753.56	7356.04
Hf4+	317.01	305.31	306.42	324.07	242.20	88.42	635.74	484.63	158.27	314.14	18.27	65.11	n.a.	-
Y3+	1546.98	6076.98	1129.73	5962.74	1617.00	2573.24	2121.13	4978.48	3773.36	7840.26	1690.30	6238.30	1870.29	2779.06
HREE3+	352.52	1082.90	746.28	5149.99	1098.22	1506.67	2110.42	3601.88	2915.28	4824.19	1449.73	4761.78	n.a.	-
LREE3+	498.29	2909.07	377.03	1837.77	679.28	443.23	320.21	392.97	688.30	1535.92	1062.06	6339.04	n.a.	-
Bi3+	39.08	192.07	14.49	39.74	n.a.	-	10.83	12.21	53.76	165.80	49.48	166.43	n.a.	-
Zn2+	942.00	1068.84	1036.32	4424.25	838.00	1385.34	1860.43	3924.70	3675.45	4609.66	1060.00	1528.86	n.a.	-
Pb2+	1133.67	2715.57	994.68	3204.83	345.60	377.67	1100.87	2496.62	1928.27	4976.10	5360.90	9795.87	n.a.	-
Sr2+	34.96	59.09	25.75	81.31	27.00	11.27	42.57	31.35	271.64	239.91	185.40	213.31	n.a.	-
Be2+	30.74	64.93	18.71	47.31	21.80	12.15	118.74	794.42	591.27	2234.32	11.60	33.23	n.a.	-
Li+	668.46	518.77	226.49	1096.46	6.00	3.39	880.65	711.35	7938.18	4234.05	192.50	415.71	n.a.	-
Rb+	6184.52	3821.08	4456.58	5374.03	1000.00	0.00	1000.00	0.00	1000.00	0.00	316.80	693.24	6192.30	5680.93
Cs+	92.61	147.28	25.66	78.12	13.70	10.27	112.82	134.08	275.00	234.08	6.79	15.59	n.a.	-
S−	120.89	376.29	256.07	393.27	n.a.	-	417.39	1361.35	827.27	2895.64	7730.00	26,155.27	n.a.	-
Nb/Ta	9.69	33.06	7.93	3.16	7.75	0.72	4.45	1.97	7.18	10.42	40.41	18.60	4.88	3.22
Th/U	3.90	8.71	1.79	3.94	1.51	0.35	6.61	4.55	17.96	32.57	152.96	507.82	17.87	98.71
LREE/HREE	1.20	1.79	1.17	1.62	1.09	0.68	0.32	0.61	0.25	0.22	0.39	0.83	n.a.	-

¹ Lengler [42]. ² Paludo et al. [43]. ^a n = number of analysis. For the pegmatite veins and border pegmatite, maximum detection limit is 1000 ppm for Nb, Sn, Rb, and REE, 2000 ppm for Th, and 10,000 ppm for Pb, Y and Zr. Abbreviation: n.a. = not analyzed.

). Analyses of amphibole-rich, polylithionite-rich, and cryolite-rich pegmatite veins are also included, though interpreted with caution due to the inherent challenges in obtaining representative chemical data from pegmatite samples.

Three general compositional trends are observed among the analyzed subfacies and associated pegmatites. The Type-A trend (Figure 28) is characterized by decreasing average Ca and K contents and increasing Na, F, S, and Pb contents in the sequence: border pegmatite > BAG > CAG > amphibole-rich PEG > polylithionite-rich PEG > cryolite-rich PEG. The border pegmatite shows the highest average Ca (0.78 wt.%) and K (5.91 wt.%), while the cryolite-rich PEG exhibits extremely high average F (~35 wt.%), Na (~34 wt.%), S (~7730 ppm), and Pb (4845 ppm). Exceptions to this trend include the polylithionite-rich PEG, which has relatively high K (5.94 wt.%) and low Na (3.1 wt%), and the BAG, which has higher average S (207 ppm) than the CAG (120 ppm).

In the Type-B trend (Figure 29), an increase in Y, Li, Be, and Zn is observed in the sequence: CAG > amphibole-rich PEG > polylithionite-rich PEG, followed by a decrease in the cryolite-rich PEG. This pattern highlights the significant enrichment of the polylithionite-rich PEG in Y (3773 ppm), Li (7938 ppm), Be (591 ppm), and Zn (3675 ppm). The border pegmatite has higher concentrations of Y and Be compared to the BAG.

The Type-C trend (Figure 30) shows an increase in Nb, Ta, U, Th, Zr, and Sn in the sequence: border pegmatite > BAG > CAG > pegmatitic CAG, followed by a decrease in the direction amphibole-rich PEG > polylithionite-rich PEG > cryolite-ich PEG. The pegmatitic CAG presents the highest average values for Nb (1978 ppm), Ta (451 ppm), Rb (6192 ppm), U (511 ppm), Th (5026 ppm), Zr (6753 ppm), and Sn (2459 ppm). The cryolite-rich PEG has the lowest concentration of these elements. An exception is the Zr average content of the border pegmatites (5624 ppm Zr), which is higher than the albite-enriched granite averages (4708 ppm Zr in the BAG and 4606 ppm Zr in the CAG). The border pegmatites also present the lowest Sn content (92 ppm Sn). Given the analytical constraints, it can be cautiously stated that the Rb contents align with the Type-C trend, with the highest average in the pegmatitic CAG (6026 ppm) and a significant decline in the cryolite-rich PEG (240 ppm).

4.6.2. REE Contents and Patterns

All the investigated rocks show varying degrees of fractionation in the chondrite-normalized REE distribution pattern (Figure 31A). The CAG exhibits the lowest fractionation, with high LREE (average 598 ppm) and the lowest HREE content (average 397 ppm). The BAG and border pegmatite have similar REE signatures, but the border pegmatite contains slightly higher concentrations (679 ppm LREE, 1098 ppm HREE) compared to the BAG (473 ppm LREE, 1015 ppm HREE). The pegmatite veins have lower LREE and higher HREE relative to the host rock. Among the pegmatite veins, there is a decrease in both LREE and HREE from the polylithionite-rich PEG (688 ppm LREE, 2915 ppm HREE) > amphibole-rich PEG (320 ppm LREE, 2110 ppm HREE) > cryolite-rich PEG (59 ppm LREE, 833 ppm HREE). The general REE pattern for the amphibole-rich PEG and polylithionite-rich PEG shows a well-defined M-type tetrad effect [58]. Despite variations in LREE and HREE concentrations, the LREE/HREE average ratio (Figure 31B) remains remarkably similar in the BAG (1.09), border pegmatite (1.10), and CAG (1.13). The pegmatite veins have significantly lower LREE/HREE ratios, with 0.32 in the amphibole-rich PEG, 0.25 in the polylithionite-rich PEG, and 0.27 in the cryolite-rich PEG.

4.6.3. Chemical Correlations

In the amphibole-rich PEG and polylithionite-rich PEG, LREE, and U show a good positive correlation (Figure 32A). In the cryolite-rich PEG, LREE content correlates with Ca (Figure 32B), but not with U, similar to the amphibole-rich PEG and polylithionite-rich PEG. The HREE elements correlate strongly with P (Figure 32C) in the amphibole-rich PEG and polylithionite-rich PEG, but not in the cryolite-rich PEG. In both the amphibole-rich PEG and the border pegmatite, FeO correlates negatively with SiO₂ + Al₂O₃ (Figure 32D) and positively with MnO (Figure 32E). The polylithionite-rich PEG shows no correlation with Fe content. In the cryolite-rich PEG, FeO correlates positively with K₂O (Figure 32F), and K₂O correlates with Rb (Figure 32G). In the polylithionite-rich PEG, Be and Zn show a strong correlation (Figure 32H). Better correlations are seen in the polylithionite-rich PEG, amphibole-rich PEG, and cryolite-rich PEG for Zn + Pb + Zn versus S (Figure 32I). Fluorine content shows a negative correlation with S (Figure 32J) and Si (Figure 32K), especially in the cryolite-rich PEG. No correlation was found between Na and Ca (Figure 32L).

5. Discussion

5.1. Paragenetic Evolution in the Magmatic and Hydrothermal Stages of the Studied Pegmatites

Several key minerals record the behavior of trace elements in the melt. With fractionation, the grades of Li, Rb, and Cs in K-feldspar and muscovite increase, while the Nb/Ta ratio in columbite-group minerals decrease [59]. Competition for Al and alkalis between the HFSE and fluxing elements depends on the melt’s overall composition [60]. In the studied pegmatites, 28 minerals were identified, and a crystallization order (Figure 33) was established for the magmatic and hydrothermal phases.

5.1.1. Magmatic Phases

Early magmatic stage: During the initial magmatic stage of the albite-enriched granite, LREE and U contents were incorporated into the primary U-Pb-LREE-rich pyrochlore [52]. This early magmatic phase was inherited by the pegmatite veins and border pegmatites, where most of the Ca, LREE, and U were concentrated. The Ca enrichment in the pyrochlore from the eastern border pegmatite likely reflects the Ca fractionation during the early magmatic crystallization of the albite-enriched granite, rather than the influence of external host rock. The crystallization of zircon, thorite, and cassiterite was delayed due to the high F content, with their formation occurring later in a melt depleted in LREE and U [37]. In the pegmatite veins, large crystals of primary thorite enriched in HREE and Y indicate their formation from a melt enriched in these elements during the pegmatitic stage. In contrast, the occurrence of U-rich primary thorite in the border pegmatite suggests earlier formation.

Early to late magmatic stage: The crystallization of riebeckite, Fe-Li-rich annite, and polylithionite buffered Fe in the albite-enriched granite melt from its early magmatic stage. In the BAG, these Fe-rich silicates were almost entirely replaced by hematite through autometasomatism [35,61]. Riebeckite also helped buffer Na and F, especially in the BAG and border pegmatites. With increasing F in the residual melt, riebeckite in the amphibole-rich PEG shows significantly higher F content (2.12 wt.%) than in the CAG (0.67 wt.% F). Due to the Fe-F avoidance effect [62,63], this F enrichment limited Fe incorporation, favoring Na, K, and Si substitution. Comparable F-rich sodic amphiboles are rare, but examples include the Katugin cryolitic deposit (Transbaikalia, Russia), where amphibole reaches 2.5 wt.% F [64].

Polylithionite buffered K, Li, and F from early to late magmatic stages. As the melt became progressively enriched in HREE, Li, Si, and F, polylithionite composition also evolved, reaching peak enrichment in the polylithionite-rich PEG. According to Breiter [65], mica chemistry is more influenced by earlier-crystallizing minerals than by the mineral structure. Thus, earlier polylithionite and cryolite may have reduced Li and F availability in the cryolite-rich PEG fluid. The K/Rb ratios in polylithionite from the CAG (1.31–1.69) and amphibole-rich PEG (1.62) are among the lowest reported. According to Costi [31], these ratios reflect extreme melt fractionation, similar to Rb-Cs-rich pegmatites from Tanco [66] (Figure 34). Continuous crystallization of Rb-bearing K-feldspar and polylithionite [31] in the CAG (and possibly pegmatitic CAG) may explain the reduced Rb availability during polylithionite crystallization in the amphibole-rich PEG.

Additionally, abundant albite, microcline, and quartz formed in all studied rocks. In the CAG, pegmatitic CAG, and pegmatite veins, cryolite I continuously buffered Na and F, reaching peak modal values in the cryolite-rich PEG.

Late magmatic stage: The elements Y, HREE, and P concentrated in the residual melt and were later incorporated into xenotime. In the CAD, these elements formed disseminated xenotime, while in the amphibole-rich and polylithionite-rich PEG they crystallized as large xenotime crystals. In CAG xenotime, F substitutes O to form PO₃F tetrahedra [39], shortening the structure and promoting incorporation of larger cations (e.g., Er, Yb) at the expense of Y, hindering LREE incorporation. This is supported by the lower F and higher Y content in xenotime from the polylithionite-rich PEG. In the cryolite-rich PEG, HREE were mainly incorporated into gagarinite-(Y), whereas no HREE- or Y-bearing primary phases were identified in the border pegmatites.

In the CAG, residual LREE were buffered by gagarinite-(Y), which contains fluocerite-(Ce) inclusions, attributed to exsolution of larger LREE during cooling as the gagarinite structure contracted [38]. This process produced HREE-rich gagarinite-(Y) and the exsolved LREE-rich fluocerite-(Ce). No such exsolution was observed in the gagarinite-(Y) from the cryolite-rich PEG, likely due to prior LREE depletion in the melt. Instead, this gagarinite shows higher HREE, Na, and F, indicating crystallization from a more evolved melt progressively enriched in these elements. Genthelvite crystallized after polylithionite and early quartz I, but before hydrothermal cryolite II. Along with galena, it was among the last magmatic minerals to form, incorporating Zn, Be, Pb, and S from the residual melt of the pegmatite veins.

5.1.2. Hydrothermal Phases

Early hydrothermal stage: Riebeckite alteration extensively generated chlorite, with residual Na likely incorporated into secondary albite. In the CAG and pegmatitic CAG, Fe-Li-rich annite altered to polylithionite and hematite. Hydrothermal fluids redistributed Fe throughout the albite-enriched granite system, precipitating abundant hematite in all rocks. The abundance of Fe, Pb, Zn, Bi, and S in the fluid led to the precipitation of pyrite, sphalerite, galena, native lead, and native bismuth. Early hydrothermal albite, microcline, and quartz indicate a Na-, K-, and Si-rich fluid.

In the amphibole-rich PEG, the hydrothermal phases cryolite and Na-LREE-Pb-rich pyrochlore reflects a fluid with greater Na and F activity. The absence of U-enriched silicates as product of primary pyrochlore alteration, as observed in the CAG and BAG, suggests U was incorporated in other pegmatitic minerals, as polylithionite (average 1.72 wt.% UO₂). In the border pegmatites, abundant hydrothermal fluorite and Ca-enriched hydrothermal pyrochlore indicates Ca-rich fluids. In the eastern border pegmatite, limited alteration prevented columbite formation, whereas in the northern sector, higher HREE, Y, and Mn availability in the hydrothermal fluid led to secondary HREE-Y-enriched pyrochlore, (U)-Fe-Mn-rich columbite, and (Th, U, Y, HREE)-rich silicates.

Late hydrothermal stage: Both the magmatic and early hydrothermal phases were affected by residual fluids enriched in Na, F, and Si in the CAG and pegmatite veins, precipitating cryolite II and quartz II. In the BAG and border pegmatites, the residual fluid was enriched in Ca instead of Na, precipitating abundant hydrothermal fluorite. Upon cooling, these fluids triggered oxidation, silicification, and clay mineral transformation of earlier paragenesis [53].

5.2. The Parental Rock

During fractional crystallization, if trace elements behave as perfectly incompatible in all crystalline phases, then the pegmatites inherit and amplify the source’s trace element signature. This allows the origin of pegmatites to be traced back to granites in which the source characteristics themselves are known and distinguishable [67]. In the pegmatitic CAG and pegmatite veins, the anomalous concentration of key rare metals (Na, F, S, Pb, Y, Li, Be, Zn, Sr, Nb, Ta, HREE, Th, U) reflects the amplified geochemical signature of the CAG, which is consistent with the Type-A, -B, and -C trends. In contrast, the border pegmatites show the lowest F and highest Ca contents, suggesting they represent less evolved melts derived from the BAG, as also indicated by their pyrochlore chemistry.

Although pegmatites typically lodge in structures external to their parent rock [1], in Pitinga, the geochemical evolution and paragenetic similarities indicate that the host albite-enriched granite is also the parental rock of all pegmatite types studied: miarolitic pegmatites, pegmatite veins, pegmatitic CAG, and border pegmatites.

5.3. Role of Fluorine in Magmatic-Hydrothermal Systems

Parental magmas of highly evolved granites and pegmatites are typically enriched in fluxing elements (e.g., F, Cl, Li, P, and B), which reduce melt viscosity and solidus, and increase H₂O solubility up to 30 wt.% [68,69,70,71]. These elements also enhance trace metal solubility and promote rapid growth of large, well-formed silicate crystals [72]. However, their abundance is generally low in pegmatites: even the most fractionated bodies contain <1 wt.% total B, P, and F [73], with even lower contents in simple pegmatites.

Notable F-rich granitic pegmatites include Quartz Creek, Colorado (100–6000 ppm F) [74], Pohjanma, Finland (2000 ppm F) [75]; Bernic Lake, Manitoba (5000 ppm F) [76], Mongolia (7700 ppm F) [77]; Mora, New Mexico (9000 ppm F) [78], and Ivigtut, Greenland (5000–30,000 ppm F) [79]. In contrast, the albite-enriched granite at Pitinga exhibits unprecedented F levels: 2.31 wt.% F in the CAG; 3.09 wt.% F in amphibole-rich PEG; 5.69 wt.% in polylithionite-rich PEG; and up to 35.00 wt.% in cryolite-rich PEG. In the CAG and associated pegmatites, F was key in concentrating Group I elements (Li, Na, K, Rb), and to a lesser extent Cs, along with anomalously high REE, U, Th, Be, Zr, Nb, and Ta contents compared to the localities above.

The REE tend to form complexes with alkalis and F, migrating to the apical zones of granitic intrusions [80]. The HREE show stronger complexation with F than LREE [81], explaining the LREE-rich early phases in the BAG and CAG (e.g., LREE-rich pyrochlore), and the HREE enrichment in later paragenesis of the CAG and pegmatites (e.g., xenotime, gagarinite), as well as in residual F-rich hydrothermal fluids precipitating secondary HREE-bearing phases.

Williams-Jones and Vasyukova [82] showed that neutral weathering fluids alter pyrochlore by replacing Na and Ca with Ba, Sr, Ce, and K, while acidic fluids leach pyrochlore to Nb₂O₅ skeletons. In the albite-enriched granite, hydrothermal fluid acidity, and thus intense pyrochlore alteration, is primarily controlled by F content, a relationship likely applicable to other altered minerals in the studied rocks.

5.4. Nb/Ta Ratio Behavior in Magmatic-Hydrothermal Systems

Magmatic-hydrothermal processes involving fluids and hydrosaline melts are key to Nb-Ta fractionation and HFSE enrichment [21,83]. In the Nb/Ta versus Nb and Ta diagrams (Figure 35A,B), the data from this study partially overlaps the lower field of rare-metal A1-type granitoids compiled by Ballouard et al. [83], including associated metasomatic rocks (e.g., greisens, albitites, skarns).The samples also show similar Nb/Ta ratios and Nb and Ta contents to the Ririwai arfvedsonite albite granite (Nigeria) [84] and albitized portions of the Ivigtut alkali granite (Greenland) [85].

However, several samples from the CAG, BAG, and pegmatitic CAG extend beyond this field, falling into the magmatic–hydrothermal domain (Nb/Ta < 5) defined by Ballouard et al. [86] for peraluminous granites. The Pitinga albite-enriched granite displays Nb and Ta concentrations comparable to, or exceeding, those of the most fractionated peralkaline A1-type granitoids globally, but with a significantly lower average Nb/Ta ratio—especially in the pegmatitic CAG. The pegmatitic CAG and amphibole-rich PEG enhance this fractionation trend, whereas the formation of polylithionite-rich PEG, cryolite-rich PEG, and border pegmatites likely involved additional mechanisms beyond simple fractionation in determining their Nb/Ta ratios.

5.5. Emplacement of the Host Rock and the Studied Pegmatites

Pegmatites are placed in response to space created by geological processes and are intricately linked, both temporally and spatially, to structural disturbances—often associated with orogenic events or broader geodynamic evolution of the crust and subcrustal regions [1].

5.5.1. Host Rock

Bastos Neto et al. [22,23] propose that the A-type magmatism in Pitinga developed in a post-collisional extensional setting, likely in a within-plate scenario dominated by extensional and transtensional tectonics. The albite-enriched granite magma would have formed during the third step of isotherm rise, when mantle-derived fluids ascended further into the crust, inducing fenitization-type reactions [89] in Sn-enriched rocks and introducing anomalous concentrations of F, Nb, Y, REE, and Th.

5.5.2. Border Pegmatites

Border pegmatites, or stockscheiders, are interpreted as rocks emplaced in fractures formed by the contraction of the main stock during cooling. These fractures develop parallel to the intrusion walls, allowing the injection of late-stage residual pegmatitic magma along the contact between the intrusive body and its older host rock [90,91]. Examples of such pegmatites are found in Alaska [92], Algeria [93], Germany, and the Czech Republic [90,94,95], Brazil [96], China [97], Finland [91,98], and Greenland [99].

In the case of the border pegmatite of the Black Pearl albitite [100], the orientation of crystals perpendicular to the contact (pointing toward the center of the stock) along with the presence of pegmatitic autoliths within the albitite, led the authors to interpret the pegmatite not as a late-stage dyke intruded along the margin, but rather as a product of early crystallization, synchronous with or shortly after pluton emplacement.

Given its limited surface area (2 × 1.5 km), the albite-enriched granite under study likely cooled relatively quickly. During its early magmatic stage, fluids derived from the BAG melt likely migrated into contraction fractures at the cooler margins of the pluton, which solidified faster than the interior. This interpretation, supported by geochemical trends, is important because it establishes a relative chronology: the border pegmatites formed during the initial stages of the albite-enriched granite’s evolution, rather than as late-stage intrusions.

5.5.3. Pegmatite Veins

In contrast to the border pegmatites, the pegmatite veins formed later, from residual fluids derived from the crystallizing CAG melt. These fluids migrated into spaces opened by reverse faults and horizontal extension fractures, which developed when CAG crystallization was already well advanced. Such brittle structures likely formed during the final amalgamation stages of juvenile terrains. The NE vergence of contractional structures in the CAG is consistent with foreland tectonic features expected in the Ventuari-Tapajós orogeny [50]. Emplaced in the cold upper crust and characterized by a low solidus temperature, the granite provided favorable conditions for pegmatite formation. Structural features of both the granite and pegmatites indicate that they crystallized at the same crustal level.

The presence of reverse fault planes and extensional fractures—both with and without pegmatites—suggests that the fractures hosting the veins were not formed solely by fluid pressure. Located above the critical crustal depth, the CAG’ reverse faults did not primarily host pegmatites. Instead, the associated horizontal extension fractures were more significant for pegmatite emplacement, while the reverse faults mainly served as fluid conduits.

5.5.4. Miarolitic Pegmatites and Pegmatitic CAG

The crystallization of miarolitic pegmatites follows a similar pattern to that of pegmatite veins [101]. Differences between pegmatite types can be partly attributed to the efficiency of drainage networks for residual magmatic fractions rich in volatiles. Miarolitic pegmatites reflect inefficient drainage, whereas larger pegmatites tend to form in favorable nodes within efficient networks. In Königshain, for instance, the size and distribution of miarolitic pegmatites are controlled by the percolation of residual magma through grain-boundary pathways along late-magmatic fractures [101].

In the albite-enriched granite, once supercritical aqueous fluid separated from the pegmatitic melt, they migrated through grain-scale pathways toward the transition zone between the CAG and BAG. The advanced cooling of the BAG in this zone prevented further fluid ascent, leading to the emplacement of miarolitic pegmatites. Siachoque et al. [54] described miarolitic cavities with pegmatitic texture filling fractures that are typically subparallel—and occasionally perpendicular—to the orientation of granitic dykes. In this study, the granitic dykes correspond to the pegmatitic CAG.

5.6. Classification of the Studied Pegmatites

The classification proposed by Černý et al. [67] distinguishes pegmatite classes based on characteristics of their host environment (abyssal class), mineralogy (muscovite class), elemental composition (rare-element class), and texture (miarolitic class). According to this scheme, both the border pegmatites and the pegmatite veins correspond to the moderate-depth rare-element class, specifically the REE subclass within the NYF family. This classification is supported by their marked enrichment in REE, Y, U, Th, Be, Nb > Ta, Zr, and F. However, the pegmatite veins also show notable enrichment in Li, Rb, and Sn, and contain typical minerals such as riebeckite, polylithionite, xenotime, thorite, pyrochlore-columbite, genthelvite, and cryolite (fluorite in the border pegmatite). These mineralogical assemblages differ from the typical NYF family minerals such as allanite–monazite, euxenite, and gadolinite. Miarolitic pegmatites are classified under the miarolitic class, which encompasses shallow-level pegmatites, geode-bearing pegmatite facies, and intrusive pegmatites emplaced within granites and schists. These pegmatites crystallize under relatively low pressures, often as low as 1 kbar [67]. In the albite-enriched granite, both the pegmatite veins and the miarolitic cavities are interpreted to have formed at the same shallow crustal level, within the cold upper crust, above the critical crustal depth and under low solidus temperatures.

The alternative classification proposed by Dill [1], known as the CMS system—standing for Chemical composition, Mineral assemblage, and Structural geology—categorizes pegmatites according to elemental groups and mineral assemblages, which are further organized by associated commodities. These commodities are then analyzed in relation to their geological and geodynamic context, both temporally and spatially. Following the CMS classification, the pegmatites associated with the albite-enriched granite are categorized as follows: (i) amphibole-rich PEG: cm-sized, unzoned vein-type REE-Y-Sn-U-Zr-Hf-Zn-Pb-Be-Li-F granite pegmatite (riebeckite); (ii) polylithionite-rich PEG: cm-sized, unzoned vein-type REE-Y-Sn-Th-Zr-K-Zn-Pb-Be-Li-F granite pegmatite (polylithionite-genthelvite); (iii) cryolite-rich PEG: cm-sized, unzoned vein-type REE-Y-Zn-Pb-Na-F granite pegmatite (cryolite); (iv) miarolitic pegmatite: cm-sized zoned miarolitic cryolite-albite granite pegmatite; and (v) border pegmatite: m-sized, unzoned border-type REE-Y-Ca-U-Zr-F granite pegmatite (fluorite). All these pegmatite types fall within the 24dE subtype of Dill’s classification, due to their significant REE-Y mineralization and their emplacement in an alkaline igneous context within an intra-plate rift setting.

5.7. Pegmatite Genesis

5.7.1. Border Pegmatites

Border pegmatites or stockscheiders are commonly interpreted as late-stage products of fluid activity [90,102]. However, alternative genetic models have been proposed. The border pegmatite of the Black Pearl albitite is interpreted as having formed in the presence of low-viscosity aqueous fluids, with magma reaching fluid saturation at or shortly after emplacement. The abrupt transition to fine-grained magmatic albitite was attributed to a rapid drop in confining pressure and volatile loss [100].

In the studied border pegmatites, the coarse grain size and enrichment in K—relative to Sn, Na, and F in the host BAG—support crystallization from a low-viscosity, F-poor aqueous fluid. This suggests that the BAG melt reached fluid saturation shortly after emplacement. The resulting fluid likely migrated through the rock matrix, accumulating at the intrusion apexes and selectively concentrating elements such as K, Ca, Y, Zr, Sr, and Be, while depleting Sn, Li, Na, F, Nb, Ta, U, Th, and Pb. The abrupt pegmatite–host rock contact suggests a pressure quench triggered by rapid decompression during the ascent of the albite-enriched granite. Early fluid saturation may also explain the virtually absence of Fe-rich silicates in the BAG, likely removed by autometasomatism [35,61].

5.7.2. Pegmatitic CAG and Amphibole-Rich PEG

Early models [2] proposed that granite pegmatites form through continuous fractional crystallization of a low-viscosity granitic melt, progressively concentrating rare elements (Li, Be, Ta, etc.), fluxes (B, P, F, etc.), and volatiles (H₂O, Cl, etc.) in the residual melt. This extensive fractionation process is well documented in granitic pegmatites worldwide, such as the Tashidaban Li-rich pegmatites derived from the Kumudaban granite pluton [3].

In the albite-enriched granite, the Type-A geochemical trend—marked by increasing F and progressive fractionation toward the pluton’s center—supports this model. During late magmatic stages, the pegmatitic CAG and amphibole-rich PEG crystallized. The pegmatitic CAG, although poor in cryolite, is notably enriched in Nb, Ta, Rb, U, Th, Zr, and Sn relative to the host granite, highlighting the roles of melt saturation and advanced fractionation.

The amphibole-rich PEG exhibits even higher concentrations of Y, Li, Be, Zn, and F, and features exceptionally coarse-grained minerals such as riebeckite, genthelvite, xenotime, polylithionite, and cryolite. These characteristics suggest the involvement of a crystal growth mechanism beyond simple fractionation. London [8] proposed the Constitutional Zone Refining model, in which a flux-rich silicate fluid layer forms at the crystallization front, undercooled by ~200 °C. This low-viscosity, undercooled fluid delays nucleation and promotes rapid crystal growth, producing pegmatitic textures in hours to days [12]—contrasting with the earlier model of slow cooling. This mechanism may also have contributed to the formation of the studied pegmatites.

5.7.3. Miarolitic Pegmatites

Miarolitic cavities form when the residual pegmatite melt becomes oversaturated with volatiles, primarily water. Černý [103] identifies three main mechanisms driving the separation of supercritical aqueous fluid from the melt: (1) decompression due to magma ascent or tectonic uplift (pressure quench); (2) volatile saturation during late-stage isobaric crystallization; and (3) chemical quenching caused by the crystallization of minerals that remove solubility-enhancing fluxes from the melt.

In the albite-enriched granite, miarolitic pegmatites share mineralogical features with amphibole-rich pegmatite veins [50], suggesting a common melt source. Their formation is primarily attributed to pressure quenching, which triggered the separation of supercritical aqueous fluids and led to undercooling of the system. However, chemical quenching remains a plausible mechanism, as the extensive crystallization of cryolite buffered fluorine levels in the melt.

5.7.4. Polylithionite-Rich PEG and Cryolite-Rich PEG

Fractional crystallization alone does not fully explain the origin of the polylithionite- and cryolite-rich pegmatite veins. Within the Type-A trend, the polylithionite-rich PEG stands out for its higher K and lower Na contents compared to all other rocks, a compositional signature consistent with the incongruent alkali partitioning proposed in the Jahns and Burnham [6] model. This model suggests a density-driven separation between a K-rich aqueous fluid and a Na-rich silicate melt. The silicate melt serves primarily as the elemental source, while the aqueous fluid extracts incompatible elements from the bottom of the magmatic body and transports them upward, forming pegmatites [7].

In the studied case, however, both the K-rich and Na-rich phases gave rise to distinct pegmatites: the polylithionite-rich PEG and the cryolite-rich PEG, respectively. This observation aligns with the pegmatite genesis model proposed by Thomas et al. [9,10,11,68,69,101], which describes a dynamic, boiling magmatic–hydrothermal system involving melt–melt immiscibility. In this model, pegmatites crystallize from H₂O-rich magmatic residues and metasomatic reactions, within a system comprising three coexisting phases: a viscous, H₂O-poor aluminosilicate melt; a low-viscosity, H₂O-rich hydrosaline melt; and a low-salinity aqueous fluid. The immiscibility-driven origin of pegmatites is further supported by fluid inclusion and isotopic data [15,16,17], and it is thought to be triggered by high concentrations of fluxing elements (F, Li, B, P), water enrichment, low oxygen fugacity, and pressure drops.

In this context, the formation of polylithionite-rich and cryolite-rich PEGs is best explained by melt–melt immiscibility, while preserving the alkali partitioning. In the final residual melt, two immiscible phases separated: a K-F-rich aluminosilicate melt (low H₂O, ~58.35 wt.% SiO₂; ~5.69 wt.% F) and a Na-F-rich aqueous melt with low Si (~12.4 wt.% SiO₂; ~35 wt.% F), which crystallized the polylithionite- and cryolite-rich PEGs, respectively.

Furthermore, Nb and Ta solubility increases in F-rich aqueous solutions at high temperatures (>100 °C), with Nb generally more mobile than Ta [104,105]. Experimental studies show that Nb preferentially partitions into fluoride–silicate melt over Ta [106], and into silicate melts over F-rich fluids [107]. Fluid inclusions in quartz and topaz from the Beauvoir rare-metal granite suggest that trapped magmatic–hydrothermal fluids are enriched in Nb relative to Ta [108]. Accordingly, the cryolite-rich PEG shows both a sharp decrease in Nb and Ta contents and an increase in the Nb/Ta ratio compared to the polylithionite-rich PEG, supporting the immiscibility model.

In the studied pegmatites, immiscibility is an additional mechanism for rare-metal enrichment, with elements such as Y, Li, Be, Zn, Sr, and HFSE preferentially partitioning into the K-rich aluminosilicate melt. Although the exact timing and mechanisms of pegmatite formation remain debated, their evolution reflects a transition from a magmatic to a magmatic–hydrothermal system [16,72].

5.8. Source of the Hydrothermal Fluids

5.8.1. Hydrothermal Fluids in the Pegmatites

The geochemical signatures of rare-metal minerals in pegmatites provide valuable insights into the behavior of elements and volatiles (e.g., REEs, Nb, Ta, Li, F, H₂O) during magmatic and hydrothermal evolution. These chemical tracers help reconstruct crystallization conditions and processes at the magmatic-hydrothermal transition in highly evolved igneous systems [18,19,20,21].

In this study, the source of the hydrothermal fluids affecting the pegmatites were inferred from secondary minerals formed during pyrochlore alteration, based on a comparison between the border pegmatite and the amphibole-rich pegmatite vein. In the border pegmatite, where magmatic HREE- and Y-rich phases are scarce, columbite exhibits a low LREE/HREE ratio (0.8), secondary silicate phases are enriched in HREE and Y, and hydrothermal pyrochlore incorporates these elements—indicating a hydrothermal fluid with high HREE and Y availability. In contrast, the amphibole-rich pegmatite vein, which contains abundant xenotime and gagarinite-(Y), shows a higher columbite LREE/HREE ratio (3.14), lower HREE and Y concentrations in secondary silicates, and hydrothermal pyrochlore that lacks these elements—suggesting reduced HREE and Y availability in the fluid.

These findings imply distinct hydrothermal fluid compositions between the two pegmatite types. In the border pegmatite, the HREE and Y content of the hydrothermal fluid was not derived from leaching of magmatic HREE-Y-rich phases but instead originated from a residual aqueous phase exsolved during pegmatite crystallization. This fluid composition reflects the degree of melt fractionation at the point of H₂O saturation.

During early magmatic evolution, the F-poor aqueous melt in the border pegmatite did not reach saturation in HREE and Y to crystallize minerals such as xenotime or gagarinite. Consequently, these elements were concentrated in the exsolved deuteric fluids and incorporated into secondary hydrothermal phases. As magmatic evolution progressed from the BAG to the CAG, HREE and Y became concentrated in the residual melt, forming complexes with F. When the amphibole-rich PEG formed, the crystallization of xenotime and minor gagarinite led to depletion of HREE and Y in the exsolved deuteric fluid.

Additionally, the presence of Ca-enriched hydrothermal pyrochlore and fluorite in the border pegmatites, in contrast with Na-rich hydrothermal pyrochlore and cryolite in the amphibole-rich PEG, further supports the link between hydrothermal fluid composition and the degree of magmatic fractionation of the host pegmatite. Altogether, these features provide strong evidence that the hydrothermal fluids were locally derived residual aqueous phases rather than externally sourced. Given that the border and vein pegmatites crystallized during different stages of the albite-enriched granite system’s evolution, their associated hydrothermal alterations also occurred at distinct times, rather than as part of a single, pluton-wide hydrothermal event.

5.8.2. Hydrothermal Fluid in the BAG and CAG

At the pluton scale, pyrochlore alteration was more intense in the BAG and in the central portion of the CAG [52]. In the thinner BAG, this alteration likely reflects autometasomatic processes, as described by Costi et al. [35,61]. In the much thicker CAG, alteration is concentrated around the massive cryolite deposit at the pluton’s center, linked to the migration and concentration of H₂O-F-rich fluids towards this region.

To explain the genesis of the massive cryolite deposit, Lenharo [109] and Costi [31] proposed that the albite-enriched granite magma evolved into an extremely Na- and F-rich residual melt. According to Costi [31], upon reaching H₂O saturation, this melt split into a low-H₂O, Na-Al-F-rich fraction—responsible for cryolite crystallization—and an aqueous, relatively F-poor fluid that formed the surrounding polylithionite-feldspar-quartz aureole. However, Bastos Neto et al. [23] argued that extreme fluorine enrichment was buffered by magmatic cryolite crystallization. Paludo et al. [43] suggested that such enrichment may have occurred locally, and that cryolite-rich PEG could represent the most evolved residual melt fractions in the system.

Fluid inclusion data [23,53] support a hydrothermal origin for the massive cryolite deposit, indicating it formed from exsolved saline deuteric fluids (0–25% eq. NaCl, 100–400 °C) derived from a volatile-rich magma. These fluids likely lowered the solidus, promoting pegmatitic textures within the granite. This study adds that pressure quenching from tectonic activity and melt–melt immiscibility played an important role to the formation of pegmatite veins and miarolitic pegmatites. Moreover, hydrothermal alteration within the pegmatite bodies is interpreted as a local response to in situ exsolution of deuteric fluids, rather than to the centralized hydrothermal system of the CAG.

Extending this interpretation, it is plausible that the magmatic-hydrothermal transition occurred independently for each body—border pegmatites, BAG, CAG, and pegmatite veins—driven by their distinct compositional and chronological crystallization histories. In the CAG, this transition is marked by the exsolution of fluids that led to the formation of the massive cryolite deposit and intense alteration in the pluton’s core. In the pegmatite veins and border pegmatites, fluid exsolution resulted in the formation of cryolite II (or fluorite, in the border pegmatites), quartz II, and the autometasomatic alteration of primary minerals.

6. Conclusions

The albite-enriched granite (BAG + CAG) hosts four distinct types of pegmatites: border pegmatites, pegmatitic CAG, miarolitic pegmatites, and pegmatite veins. All were emplaced at the same crustal level as the host granite. Border pegmatites formed within contraction fractures between the BAG and country rocks; pegmatitic CAG developed in centimetric fractures within the CAG; miarolitic pegmatites crystallized in poorly drained cavities; and pegmatite veins were emplaced along reverse faults and extensional fractures.

Except for the border pegmatites, all pegmatite types share the mineralogy of the CAG, dominated by pyrochlore, riebeckite, polylithionite, zircon, thorite, xenotime, gagarinite-(Y), genthelvite, galena, microcline, albite, quartz, and cryolite. Border pegmatites, however, reflect the mineralogy of the BAG, featuring fluorite instead of cryolite and lacking genthelvite. The fluids responsible for pegmatite formation were sourced from the host granite: BAG-derived for border pegmatites and CAG-derived for the others.

The genesis of the pegmatites reflects distinct evolutionary stages of the albite-enriched granite system. Border pegmatites crystallized early, coinciding with fluid saturation of the BAG. The pegmatitic CAG and amphibole-rich PEG, in contrast, formed during late-stage crystallization and show significant enrichment in Nb, Ta, Rb, U, Th, Zr, and Sn relative to the host granite, emphasizing the roles of advanced fractionation and melt saturation. The polylithionite-rich and cryolite-rich PEGs originated from melt–melt immiscibility, which promoted alkali partitioning and acted as a key mechanism for concentrating rare metals. Elements such as Y, Li, Be, Zn, Sr, and HFSE preferentially partitioned into the K-rich aluminosilicate melt phase.

Magmatic fractionation in this system exceeds that of most known peralkaline A1-type rare-metal granitoids worldwide, marked by unprecedent F enrichment—reaching up to 35 wt.% F in cryolite-rich PEG. Fluorine-complexes enriched the residual melt in Li, Na, K, Rb, and rare metals (REE, U, Th, Be, Zr, Nb, Ta), enhancing late-stage HREE enrichment in pegmatites.

All pegmatites and the host rock were intensely altered by highly acidic, F-rich hydrothermal fluids, which corroded magmatic minerals and generated secondary phases. These fluids represent residual magmatic aqueous phases rather than externally sourced fluids, and alteration (autometasomatism) occurred at distinct times across the system, not as a single, pluton wide hydrothermal event. In border pegmatites, Ca- and HREE-rich fluids formed fluorite and HREE-enriched hydrothermal pyrochlore, columbite, and silicates. In pegmatite veins, Na-rich, HREE-poor fluids formed cryolite and HREE-depleted secondary minerals.

This study leads to the following key conclusions regarding the magmatic–hydrothermal evolution of the system:

• The ascent of the albite-enriched granite magma triggered rapid decompression, causing separation of an F-poor, K-Ca-Sr-Zr-Y-HREE-rich aqueous phase from the BAG melt. This fluid ascended and crystallized early as border pegmatites.
• Continued crystallization in the CAG led to a highly fractionated residual melt, forming pegmatitic CAG enriched in Rb, Nb, Ta, Th, and other HFSE.
• Reverse faulting may have caused a second pressure quench, prompting the separation of a supercritical aqueous melt enriched in Y, Li, Be, Zn, and F. This melt intruded fractures and miarolitic cavities, forming amphibole-rich PEG and miarolitic pegmatites.
• During this stage, melt–melt immiscibility occurred, producing two distinct phases: a K-F-rich aluminosilicate melt (low in H₂O), enriched in Y, Li, Be, and Zn; and a Na-F-rich aqueous melt (low in SiO₂), leading to the formation of polylithionite-rich PEG and cryolite-rich PEG, respectively.
• The magmatic–hydrothermal transition occurred independently in each body—border pegmatites, BAG, CAG, and pegmatites veins—driven by their distinct compositional and chronological crystallization histories. Residual aqueous fluids exsolved upon H₂O saturation, with local melt composition controlling elemental availability.
• In border pegmatites and pegmatite veins, exsolution of F-rich hydrothermal fluids led to the formation of fluorite and cryolite II, respectively, along with intense autometasomatic alteration of magmatic minerals. In the CAG, this process generated a massive cryolite body and extensive hydrothermal alteration in the pluton’s core.

The pegmatites of the Madeira deposit show exceptional enrichment in strategic and critical metals, including Nb, Ta, REEs (especially HREE), Li, and Be, making it a highly prospective target for mineral exploration. These findings emphasize the importance of precise spatial and paragenetic control during exploration, as rare-metal mineralization varies according to pegmatite types and structural setting.