Rare Earth Elements in Chile: A Critical Review of Primary, Secondary and Marine Resources, Metallurgical Recoverability and Future Supply Chain Opportunities

Abstract

The economic value of rare earth elements (REE) is concentrated in Nd, Pr, Dy and Tb. The supply chain for these elements is dominated by a single country, China, which accounts for about 60% of mined production, over 90% of refinement, and nearly 95% of permanent magnet manufacturing. This shifts the critical question from geological availability to the capacity to convert resources into products. Chile hosts several domain types for REE: ion-adsorption regoliths, primary mineralization associated with skarn, iron oxide–apatite (IOA) and iron oxide copper–gold (IOCG) systems, pegmatites and placers, and an Exclusive Economic Zone with marine occurrences. The inventory is organized using a five-level evidence scale, compiled through structured searches of indexed databases and institutional repositories under explicit inclusion criteria. Only one record reaches level 1, Penco, with 27.5 Mt Measured and Indicated at 2292 ppm TREO and 62.9 kt contained TREO. Its grade is two to three times higher than that of ion-adsorption deposits abroad, but its tonnage is one to two orders of magnitude lower. Recoveries vary by element, from 19.53% for Nd to 43.23% for Dy, meaning that the REE oxide it contains does not equate to fully recoverable product. Nahuelbuta reaches 2000 ppm exchangeable, Cerro Carmen contributes 19.81 Mt containing 8203 t of rare earth materials and 1811 t of U, and El Buitre contains 6.6 Mt at 337 mg/kg. Reported concentrations span four orders of magnitude, but only two of the five measurement scales include an associated tonnage. Metallurgical recoverability follows the same pattern: saline desorption is mature and retains Th and U in the solid residue, acid leaching of copper residues reaches 64.5% at laboratory scale, the monazite–xenotime and apatite routes are proven but untested on Chilean material, and no domain has demonstrated separation capability. Opportunities lie in by-product recovery from iron and copper operations, modular development of the central-southern corridor, and domestic separation. Converting the available evidence into comparable figures would reshape the national hierarchy more than additional exploration.

1. Introduction

Rare earth elements (REE) have been established as strategic raw materials for energy transition, transport electrification, renewable power generation, and digital and high-efficiency technologies. The group consists of the fifteen lanthanides plus Y and Sc, and although several have crustal abundances comparable to or higher than those of conventional metals, their economic availability is limited by the scarcity of exploitable concentrations, the complexity of their mineralogical associations, and the inherent difficulty of individual separation [1,2,3]. These limitations are not evenly distributed across the seventeen elements: permanent magnets account for about 95% of global consumption in value terms, so demand is governed by Nd, Pr, Dy, and Tb, with Dy and Tb as the most critical despite representing a smaller fraction of the magnet mass [4,5]. The strategic importance of a deposit therefore depends not only on its total content but also on what elements it contains and whether it is realistic to convert them into marketable products.

The vulnerability of this value chain is intensified by its high geographic and industrial concentration, the magnitude of which depends on the stage examined [6]. China produced approximately 270,000 t of REO in 2025, making up the vast majority of the close to 390,000 t of REO mined globally [7]; for the four magnet-associated REE, the country continues to dominate, accounting for about 60% of mined production, over 90% of refinement and nearly 95% of permanent magnet manufacturing globally [4]. The export controls introduced in 2025 on heavy rare earth elements, compounds, and magnetic products have turned that imbalance into a concrete operational constraint for consuming industries, definitively shifting the discussion from geological availability to the construction of complete value chains [8]. The exploration boom after 2010 revealed that, of the numerous initiatives that had been put in place outside of those by the dominant producer, only a minimal fraction reached production. Those that did reach production encountered a bottleneck in separation and in the commercial placement of individual oxides [7]. Consequently, hosting significant REE occurrences does not guarantee effective market participation without quantified resources, demonstrated metallurgical routes, and the capacity to produce high-purity compounds [9].

Within this context, Chile is a particularly interesting case. Its mining tradition, dominated by Cu and Fe, has generated extensive geological information, infrastructure, industrial capabilities, and large volumes of waste, while the metallogenic diversity of the country includes weathered granitoids, iron oxide–apatite and Fe-Ti-U systems, IOCG systems, pegmatites, sedimentary accumulations, and an extensive Exclusive Economic Zone with environments favorable for ferromanganese nodules and crusts [10,11,12]. Recent studies have also demonstrated REE mineralization in a regolith developed on granitoids of Chile’s Coastal Cordillera. This has expanded the range of environments that are considered prospective for ion-adsorption deposits, a type of deposit that supplies most of the world’s heavy rare earth elements despite its moderate grades [13,14,15].

The degree of knowledge about these domains, however, is uneven. On the best-documented end, the Penco Module has an estimate reported under an international code of 27.5 Mt of Measured and Indicated resources at an average grade of 2292 ppm TREO (equivalent to 62.9 kt contained); it received a favorable environmental qualification in June 2026 [16,17]. At the intermediate level, Cerro Carmen has a historical estimate of 19.81 Mt containing 8203 t of REE and 1811 t of U, predating current reporting codes, and the El Buitre tailings deposit was modeled in three dimensions from 28 drill holes and 755 samples, yielding 6.6 Mt at an average grade of about 337 mg/kg [10,18]. Further behind are systems with a well-demonstrated carrier mineralogy but no tonnage, for example, the Nahuelbuta regoliths, with up to 2000 ppm of exchangeable REE in profiles up to 60 m thick; the zoned apatite of the iron oxide deposits of the Chilean Iron Belt, hosted in orebodies exceeding hundreds of millions of tons but without a published average REE grade; and the pegmatites of the Paleozoic batholith, with point values between 321 and 10,070 ppm [13,19]. Finally, the national tailings cadastre includes more than six hundred deposits with multielement ICP-MS determinations, and the submarine occurrences in the Exclusive Economic Zone are known for their Co, Cu, and Ni content but lack REE analyses [20,21].

This summary reveals the central problem: the existence of mineralization does not equate to the existence of an economically recoverable resource, and the available figures are not mutually comparable. National information combines resources estimated under international codes, historical estimates, academic geostatistical models, point analyses of individual minerals, regional sediment anomalies, surface sampling of tailings, and poorly characterized marine occurrences. Magnitudes are also expressed using different indicators (elemental REE, TREO, TREY, exchangeable fractions, or DREO), whose conversions are non-trivial and whose differences can reach a factor of five for the same material [13]. Extrapolating a concentration measured in an apatite crystal to the full tonnage of an iron deposit, or a surface anomaly to an undelimited volume, produces inflated inventories devoid of technical meaning. Conducting a more rigorous evaluation requires distinguishing among occurrence, characterized mineralization, resource, and reserve and linking each figure to the level of evidence that actually supports it [22].

A second distinction of a mineralogical and metallurgical nature overlays the first. The total REE content is only a first approximation of the potential value of a material: two resources with similar grades may have very different prospects if one is dominated by La and Ce and the other contains higher proportions of Nd, Pr, Dy, and Tb [23,24]. Likewise, a high grade loses relevance if the elements are hosted in refractory phases or if their release requires disproportionate energy and reagent consumption [25]. Carrier mineralogy, deportment during beneficiation, the leachable or desorbable fraction, associated Th and U, the selectivity of separation stages, and the valorization of by-products are therefore as decisive as grade and tonnage [15]. The difference is quantifiable in Chilean materials themselves: leaching tests on copper processing residues in Atacama reached 64.5% recovery of total REE, equivalent to 0.355 kg per ton treated, while in Penco, the average recoveries used in the estimate vary by element, from 19.53% for Nd to 43.23% for Dy [26]. In low-grade resources and secondary materials, differences of that magnitude alone define process viability.

Despite growing interest in critical minerals in Chile, information on REE remains scattered across the scientific literature, institutional cartography and geochemistry, technical reports, tailings cadastres, and studies focused on individual deposits. The existing background addresses parts of the problem (tailings valorization, the global industry context, or characterization of a specific district) but it does not integrate primary, secondary, and marine domains under a common evidence criterion, nor does it benchmark the national inventory against the deposits that actually sustain world production [27,28]. This fragmentation prevents establishing which part of the national potential corresponds to quantified resources, what areas constitute only exploration targets, and which ones have sufficient metallurgical evidence to consider future recovery. The inclusion of REE in the National Critical Minerals Strategy makes this ambiguity especially costly because exploration, research, and public policy decisions require differentiating geological opportunity from technically demonstrable potential [29,30].

The objective of this work is to critically evaluate mineral and alternative rare earth resources in Chile, integrating geology, mineralogy, resource quality, metallurgical recoverability, and the level of available evidence. Background information is organized using an evidence scale that distinguishes formally defined resources, historical estimates or non-standard three-dimensional models, characterized mineralization without tonnage, localized occurrences, and regional anomalies. This is applied transversally to five domains: deposits hosted in regolith and ion-adsorption clays; primary rock-hosted mineralization associated with Fe-Ti-U, IOA-IOCG, and pegmatite systems; sedimentary accumulations and drainage anomalies; tailings and mining waste; and potential marine resources linked to nodules, ferromanganese crusts, and metalliferous sediments. For each domain, carrier phases, elemental composition, inventory scale, recovery evidence, and information gaps are examined together with the conditioning factors that mediate between geological availability and effective value-chain development: association with Th and U, environmental and licensing restrictions, processing of low-grade materials, and the need to prioritize elements of greater technological relevance. The purpose is not to estimate an aggregate figure of national reserves, which the available evidence does not support, but to establish what resources are demonstrated, which ones show recovery potential, where uncertainties are concentrated, and what information must be generated to convert geological occurrences into technically defensible resources.

Four contributions distinguish this work from existing reviews on Chilean critical minerals, rare earth deposits, tailings valorization and global supply. First, the five domains that concentrate rare earth elements in the country are examined jointly and under a single criterion, whereas the available background treats them separately. Second, a five-level evidence scale is applied to every published figure, separating quantified resources from occurrences and showing that only one national record reaches the first level. Third, the Chilean resource is benchmarked in grade and contained metal against the ion-adsorption deposits that sustain world production. Fourth, geological evidence is coupled with metallurgical recoverability, identifying the measurements that would modify the national hierarchy more than any exploration campaign.

Information Sources, Search Strategy and Assignment of Evidence Levels

The evidence base combines two streams. The peer-reviewed literature was retrieved through structured searches in Scopus, Web of Science (Core Collection) and ScienceDirect, with Google Scholar used only to trace citations and locate non-indexed documents. Institutional information was obtained from various organizing bodies that generate the data: the National Geology and Mining Service of Chile, or SERNAGEOMIN (cartography, regional surveys and the tailings cadastre); the Chilean Copper Commission, or COCHILCO, and the Chilean Ministry of Mining (statistics and the National Critical Minerals Strategy); the Chilean Environmental Assessment Service (studies and resolutions). For international benchmarks, data was obtained from the United States Geological Survey (USGS), the International Energy Agency (IEA), and the International Seabed Authority, as well as from technical reports filed under the NI 43-101 and JORC codes.

Queries combined a commodity block with deposit-type, process and geographic blocks in title, abstract and keyword fields, for example, (“rare earth*” OR REE OR TREO) AND (Chile OR “Coastal Cordillera” OR Atacama OR Biobío OR “Exclusive Economic Zone”), with equivalent strings for ion-adsorption regoliths, tailings and leaching, monazite–xenotime and apatite, and ferromanganese nodules. Spanish equivalents (tierras raras, arcillas iónicas, relaves, nódulos polimetálicos) were applied to national repositories whose documents are not indexed internationally.

The search window depended on content—2015–2026 for the global context (demand, supply, trade measures and environmental performance)—with priority given to post-2020 and no lower bound for geology, mineralogy and metallurgy, since several deposit models and the only quantitative records of some districts are older. Searches were last updated in July 2026, which is the cutoff date for the documents cited.

A document was included when it reported original data (analytical determinations, resource estimates, block models, cadastres or process results); addressed one of the five domains, or the global chain allowed traceability of its figures, that is, technique, samples and reporting basis; and was peer-reviewed or issued by a competent institution.

Records were excluded when they reproduced a figure already retrieved from its primary source, consisted of abstracts without accessible data, reported concentrations without stating the methodology or reporting basis, dealt with applications or prices unrelated to resources, or could not be traced to a retrievable document. Press releases were not used as quantitative sources.

Gray literature and institutional reports were admitted conditionally because much of the geological, geochemical and cadastral information on Chile exists only in that form. Three rules applied to such material: the issuing body had to be identifiable and competent and the document publicly retrievable; a quantitative figure was used only when the method, sample support and reporting basis could be established from the document itself; and corporate information was accepted as a resource figure only when backed by a technical report under a recognized code. Historical estimates were retained and labeled as such.

Each record reporting a concentration or tonnage was assigned a level by applying a fixed sequence of questions to the datum itself, not to the magnitude of the figure: tonnage and grade under a recognized code, along with categories and effective date—E1; tonnage from pre-code methods or academic modeling—E2; grade and mineralogy on a delimited body without tonnage—E3; carrier phases without geometry—E4; and drainage geochemistry, surface sampling of residues or occurrences without REE determinations—E5. In borderline cases, the level is assigned according to the most demand that the source can sustain, and an unretrievable primary document moves the record one level down.

This is not a systematic review, nor does it aim to be. Much of the relevant evidence is institutional, non-indexed and heterogeneous. What it provides is traceability, since the sources, the criteria, and the rules that place each figure on a certain level are explicit and can be reapplied.

2. Mineral and Alternative Rare Earth Resources in Chile

2.1. Criticality, Nomenclature and Evidence Levels of National Inventory

2.1.1. Position of Rare Earth Elements in Global Supply Chain

Rare earth elements (REE) include the fifteen lanthanides together with yttrium and scandium, occur in more than 250 minerals, and underpin electronics, automotive, renewable energy, medical and defense technologies [2]. Their strategic relevance does not stem from crustal scarcity (several REE are more abundant than copper or lead) but from the geographic concentration of production and, above all, processing. China has historically contributed 80% to more than 90% of the world’s primary production and controls most of the separation capacity [1,2]. This asymmetry explains why rare earth materials appear on practically all critical raw material lists and why exploration outside China has intensified following the past decade.

Demand is not evenly distributed among the seventeen elements. Growth is dominated by NdFeB permanent magnets, which require Nd and Pr as major components and Dy and Tb as additives that preserve coercivity at high temperature, while Ce and La (the most abundant REE in most deposits) face comparatively saturated markets [1]. Consequently, two deposits with identical total REE grades can have very different economic values depending on their elemental distribution, and a substantial fraction of the reported grade may correspond to low-price elements [31]. Ion-adsorption deposits hosted in regoliths occupy a unique position: they contribute about 80% of the heavy rare earth resources produced worldwide despite representing a smaller fraction of global resource tonnage because they combine moderate grades with profiles enriched in heavy REE and recovery by saline desorption that avoids grinding and intensive acid leaching [14,15].

2.1.2. Nomenclature and Comparability of Published Figures

The Chilean information on REE comes from sources that use different conventions; therefore, a direct comparison of figures requires explicitly stating what each one measures. Industry resource estimates usually report TREO (the sum of oxides of the fourteen stable lanthanides) and TREY or TREO + Y when yttrium is included, expressed in ppm or mass percent; institutional geochemical reports provide elemental concentrations, not oxides, in ppm or mg/kg; and regolith studies distinguish between total REE and exchangeable REE, that is, the fraction effectively desorbable by cation exchange. The difference is not semantic: conversion from element to oxide introduces factors of 1.15 to 1.25 depending on the element, and the exchangeable fraction can represent anything from less than one-fifth to more than half of the total REE in the same profile [13,15]. Recent industrial practice formalizes this distinction through the concept of DREO (desorbable rare earth oxides), estimated from metallurgical recoveries of the ammonium sulfate desorption process and used as the basis for block economic value instead of total grade [17].

A second convention affecting comparison is the cutoff between light and heavy REE. The most common grouping assigns La-Nd to light rare earth elements (LREE), Sm-Gd to intermediate rare earth elements, and Tb-Lu plus Y to heavy rare earth elements (HREE), but part of the literature shifts the boundary to Eu or classifies Y as light based on its ionic radius, which can change the reported HREE/LREE ratio for the same material by several points. Finally, tonnage figures respond to different reporting frameworks—an estimate under CIM or JORC standards, a historical estimate predating those codes, an academic geostatistical model, and a surface sampling campaign of residues are not equivalent categories, although all are expressed in tons.

Accordingly, no value reported here was converted by the authors: every concentration and tonnage is reproduced on the basis stated by its source (TREO, TREY, DREO, exchangeable REE, or individual oxides); that basis is indicated wherever the figure appears, and comparisons between deposits are made only between values expressed on an equivalent basis. The exchangeable fraction is treated throughout as a distinct quantity and is never equated with, or converted into, total REE content.

2.1.3. Domains of Occurrence and Evidence Scale

Chile’s REE potential does not correspond to a single deposit type, nor can it be expressed as an aggregate national figure. Available information describes a continuum ranging from formally quantified mineral resources to historical prospects, mineralization hosted in accessory minerals, sediment anomalies, mine tailings, and submarine occurrences whose economic importance remains to be demonstrated [32]. This distinction is fundamental because a given grade in a block model, a point concentration in apatite or monazite, a drainage sediment anomaly, and an average concentration in a tailings deposit represent different levels of evidence and are not directly comparable. The National Critical Minerals Strategy recognizes REE as critical minerals and highlights the national potential associated with both primary mineralization and weathering profiles capable of concentrating them by adsorption, pointing to Penco and Cerro Carmen among the most relevant expressions currently known [28,30].

On this basis, national resources and occurrences can be organized into five main domains: (i) deposits hosted in regoliths and ion-adsorption clays; (ii) primary mineralization in rock, including Fe-Ti-U-REE, IOA-IOCG, and pegmatite systems; (iii) sedimentary accumulations and geochemical anomalies derived from erosion of enriched sources; (iv) secondary resources contained in tailings and other mining wastes; and (v) potential marine resources associated with ferromanganese nodules and crusts (see Figure 1). This classification allows for the ranking of potential according to tonnage, grade, carrier mineral phase, recoverability and confidence level, avoiding interpretation of every geochemical occurrence as a mineral resource [11,12,33]. Superimposed on these domains is a transversal scale of knowledge maturity, summarized in Table 1, which allows each published record to be placed according to the type of data supporting it and not according to the magnitude of the reported figure.

Figure 1. The distribution of rare earth resources and occurrences in Chile by geological domain and evidence level, from the Penco Module, the only categorized resource for unquantified submarine occurrences. Includes detail of the Atacama belt, with primary rock-hosted mineralization, and an inset of the oceanic territory with ferromanganese nodules and the crust of the Exclusive Economic Zone.

Table 1. Rubric of different levels of evidence applied to Chilean rare earth supply.

Level	Designation	Type of Supporting Data	Legitimate Use of Figure	Frequent Associated Error	References	Criterion Separating It from Level Above
E1	Mineral resource under reporting code	Systematic drilling, block model, Measured–Indicated–Inferred categories, metallurgical recoveries by element	Direct comparison with international resources; basis for feasibility studies	None if effective date and applied standard are declared	[17,34]	Top of scale: requires recognized code, resource categories and declared effective date; absence of any moves record to E2
E2	Historical estimate or non-standard three-dimensional model	Tonnage and grade obtained with methods prior to current codes or academic geostatistical modeling	Order-of-magnitude sizing; prioritization of targets for reevaluation	Presenting it as a resource or reserve comparable to E1	[10,18]	Differs from E1 in reporting standard, not in amount of data: tonnage and grade exist but without code, categories or effective date
E3	Characterized mineralization without tonnage	Carrier mineralogy identified, point or profile grades, genetic control established	Definition of deposit model and exploration guide	Extrapolating point grade to an undelimited volume	[13,19,35]	Differs from E2 in absence of tonnage: grade, mineralogy and genetic control established but no mass associated with delimited volume
E4	Localized mineralogical or geochemical occurrence	Identification of carrier phases in mapping or multielement analysis of targeted samples	Evidence of lithological fertility; area selection	Converting occurrence into quantified potential	[36,37]	Differs from E3 in absence of geometry: carrier phases are identified in samples or mapping, without body or profile of known extent
E5	Regional anomaly	Drainage sediment geochemistry, surface sampling of residues, submarine occurrences without REE analysis	Basin ranking and first-pass target selection	Multiplying concentration by unrelated surface or tonnage	[11,33,38]	Differs from E4 in sampling target: dispersion product or occurrence without REE determination analyzed, not mineralized material itself

Placement on this scale depends on the type of datum supporting a figure, not on its magnitude: E1 requires systematic drilling and a block model under a recognized code, with categories and an effective date; E2 requires a tonnage and grade from pre-code methods or academic modeling; E3 requires a mineralogy and grade on a delimited body without tonnage; E4 requires carrier phases without defined geometry; and E5 requires drainage geochemistry, surface sampling of residues or occurrences lacking rare earth determinations. Where the primary document could not be verified, the record was moved down one level. The levels describe the maturity of the information, not the quality of the source.

The scale is not a new classification of mineral resources but rather an ordering of the available evidence adapted from three established families of criteria. From the CRIRSCO codes (JORC, NI 43-101 and equivalents), it requires that any tonnage and grade statement be supported by drilling, a three-dimensional model, categories and an effective date. From the United Nations Framework Classification, it distinguishes between geological knowledge and project maturity. From exploration practice, it borrows a graded sequence from regional anomaly to characterized mineralization. What is specific here is the joining of both ends into a single scale that is applied uniformly to indexed and institutional sources.

Adjacent levels are separated by one discriminating attribute each, so the assignment does not depend on judgement about the source. E1 and E2 differ in the reporting standards: both state tonnage and grade, but only E1 does so under a recognized code. E2 and E3 differ in the existence of a tonnage. E3 and E4 differ in the geometry of the sample body. E4 and E5 differ in the sampling target: E4 analyzes the mineralized material, while E5 analyzes a dispersion product or an occurrence without rare earth determination. Table 1 states each criterion alongside the level it delimits.

The boundary that requires the most explicit treatment is between a well-characterized occurrence with metallurgical testing but no tonnage and a resource with a tonnage estimate outside current codes. The first is E3 and the second is E2 because the scale measures how far a figure may be extrapolated: a measured recovery indicates what fraction of the metal is accessible but does not authorize multiplication by volume, whereas a historical tonnage does support order-of-magnitude sizing. Metallurgical demonstration is recorded as an independent qualifier: E3(M) when recovery was measured on the material itself, E3(m) when inferred by analogy, and E3 when no test exists. El Buitre is E2, the Atacama residues are E3(M) and Penco is E1(M).

Each record was assigned using the boundary questions, and the results were compared with the wording used by its corresponding source, which identifies documents that call a resource a figure that does not meet a code. Consequentially, for each level, Table 1 provides the legitimate use of the figure and the error that it does not support; any such assignment can therefore be considered invalid.

The scale indicates how well a figure can be supported by the evidence, but it does not in and of itself separate three notions that the literature about Chilean rare earth materials often uses interchangeably. This is kept distinct throughout this work. An analysis reports the content of a material: a concentration in a sample, a carrier phase or a regional anomaly. It does not indicate how much material it carries. Recoverability of a mineral is a fraction of the content that a proven process can release and separate from the feed. This is measured on the feed material itself and expressed by the element. Economic recoverability provides the conditions under which that fraction can be produced and placed. This includes cost, infrastructure, licensing and environmental performance. These three factors form five successive levels; each level is meaningful only once the previous level has been satisfied (Figure 2).

Figure 2. Successive levels from geochemical anomaly to economic recoverability, grouped into content, recoverability and economic recoverability. Each evidence level corresponds to the next. Chilean records are currently placed at each level.

Read as a sequence, the framework explains why the Chilean national inventory cannot be summarized. Levels 1 to 3 are within the domain of geology, and only the third yields a tonnage by which a concentration may legitimately be multiplied; level 4 is in the domain of metallurgy and is recorded here as the (M) qualifier of the evidence scale; and level 5, the only one that defines a reserve, has never been reached by any Chilean resource (even the two cases with published economics remain conditional). Each domain examined in the following sections is therefore evaluated against the level it has reached, and the next tables also apply the same sequence quantitatively to the entire inventory.

2.1.4. Lithological Fertility Documented by Regional Surveys

Recent geological surveys by SERNAGEOMIN broaden this view by showing REE signals in very different lithological contexts along the country. In Concepción-Talcahuano, the potential of regoliths derived from calc-alkaline granitoids has been recognized. Here, alteration of allanite, monazite and other accessory minerals can generate REE-enriched clays. In Coquimbo, granitoids with relative LREE enrichment and late pegmatitic dikes are documented. In Los Ríos-Los Lagos, granitoids and glaciofluvial deposits capable of participating in sedimentary dispersion are described. Finally, in Magallanes, xenotime, REE-rich epidote and REE-Ti-Th-U associations in mineralized rocks have been identified. This background constitutes evidence of fertility and mineralogical occurrence but does not alone equate to quantified resources [36,37,39,40].

The distribution of these signals has a metallogenic reading: it covers the Coastal Cordillera from Atacama to Magallanes, crosses magmatic arcs of very different ages, and appears both in intrusive domains and in sedimentary covers derived from them. This is consistent with the international observation that formation of ion-adsorption deposits depends less on an exceptional magmatic composition than on the combination of a parental rock with carrier accessory minerals, weathering capable of releasing them, and a geomorphology that preserves the resulting profile [14,41]. The present work integrates that institutional background with the recent scientific literature to provide an order for the Chilean rare earth inventory, according to geological domain and evidence level.

2.2. Deposits Hosted in Regoliths and Ion-Adsorption Clays

2.2.1. Penco Module: The Best-Defined Chilean REE Resource

The Penco Module, in the Biobío Region, represents the Chilean REE resource with the highest degree of quantification to date. Mineralization is hosted in a regolith developed on granitic rocks and comprises six zones: Liuna, Maite, Victoria Norte, Victoria Sur, Alexandra Oriente and Alexandra Poniente. The estimate with an effective date of 13 October 2022 reported 27.5 Mt of Measured + Indicated resources and 1.7 Mt Inferred resources, containing 62.9 and 3.4 kt of TREO, respectively. The Measured + Indicated category presents average grades of 2292 ppm TREO and 1932 ppm TREY, with 380 ppm Nd + Pr and 66 ppm Dy + Tb [17,28].

Elemental composition is particularly relevant because it confirms that potential value is not determined solely by Ce and La. In Measured + Indicated resources, 650 ppm Ce, 326 ppm Y, 318 ppm La, 302 ppm Nd and 78 ppm Pr are recorded, while among heavy REE, there are 50 ppm Dy, 49 ppm Gd, 36 ppm Er, 32 ppm Yb, 12 ppm Ho and 8 ppm Tb. This Nd-Pr and Dy-Tb association favorably differentiates the deposit with respect to resources dominated by Ce-La, although metallurgical response remains heterogeneous: for the Measured + Indicated resource, average recoveries of 22.8% TREY, 43.23% Dy and 19.53% Nd were used. Therefore, total grade must be interpreted jointly with exchangeable fraction and the individual recovery of each REE [28].

That gap between contained grade and recoverable grade is precisely what subsequent estimation practice has incorporated explicitly. In the sister module developed by the same company in Goiás, resources are estimated on TREO grades, but block value is calculated from desorbable grades derived from recoveries of the ammonium sulfate desorption process, so the grade governing economic decision-making represents only a fraction of the analytical grade [16]. Applied to Penco, this criterion implies that the 62.9 kt of TREO contained in the Measured + Indicated category does not constitute the productive target but rather the universe from which the desorbable portion on which revenues depend is drawn, dominated by Nd, Pr, Dy and Tb [17,34].

Mineralogical characterization of South American ion-adsorption clays carried out with combined XRD, SEM-EDX, XPS, ToF-SIMS, EPMA and LA-ICP-MS techniques provides a relevant nuance for interpreting these materials: the matrix is dominated by kaolinite and micas, and REE occur effectively as adsorbed species and inner-sphere complexes, but monazite is identified as the primary carrier phase, so part of the analytical inventory resides in a mineral refractory to saline desorption [42,43]. In the same vein, a study of the granulometry effect demonstrates that physiosorbed and chemisorbed fractions vary with particle size and associated mineralogy. This conditions both the sampling design and extrapolation of laboratory results to the industrial scale [44]. This dual residence (adsorbed and incorporated in phosphates) explains much of the dispersion observed among recoveries of different elements in Penko itself.

A distinctive attribute of this deposit type, relevant both to its environmental performance and its commercial position, is its radiological load. In ion-adsorption regoliths, the rare earths of economic interest are not held in a phosphate lattice but adsorbed on clay surfaces, while Th and U remain in the residual accessory minerals (monazite, xenotime, thorite and zircon) that weathering has not destroyed. Since the recovery route displaces the adsorbed fraction with a dilute salt solution and does not dissolve those phases, thorium and uranium report mostly to the solid residue, and the mixed carbonate obtained carries a radiological load well below that of concentrates produced from monazite or bastnäsite, which is precisely the burden that governs licensing and the commercialization of intermediate products in conventional routes [13,39]. The contrast within the national inventory is explicit: Cerro Carmen, a primary rock resource, has an estimate of 1811 t of U associated with its 8203 t of rare earths, and the Vertientes pegmatites carry thorite and Th-U silicates accompanying monazite and xenotime, whereas no equivalent association has been reported for the Penco regolith [17,45]. The public documentation of the project, however, does not report Th and U determinations in the ore, in the desorbable fraction or in the product, so this advantage rests at present on the extraction mechanism and on international analogy rather than on a published figure. Reporting the Th and U content in the ore and in the concentrate, in the same way as grades by element are reported, would turn an attribute of the deposit type into a verifiable attribute of the Chilean resource and would strengthen its position relative to hard-rock deposits.

2.2.2. International Position of Chilean Resource

A comparison with ion-adsorption deposits documented in other countries enables a more precise placement of the Chilean case. The south Chinese deposits, which defined the type, have modest individual tonnages and grades close to 1000 ppm: the Zudong deposit, the largest HREE regolith-hosted deposit, contains approximately 17,600 t of oxides at an average grade close to 0.1 wt% [41]. Table 2 summarizes the main Chilean sectors with REE potential in regoliths and sediments, comparing their geologies, contents, recoverabilities and evidence levels. Later discoveries in Brazil, Madagascar, Uganda and Australia inverted this relationship: Sierra Verde, Ambohimirahavavy, Makuutu, Koppamurra and Deep Leads record tonnages of tens to hundreds of millions of tons with grades of 640 to 980 ppm [14]. Penco lies at the opposite end of this space, with a tonnage that is one or two orders of magnitude lower than Brazilian or Ugandan deposits but with grades two to three times higher, and with the particularity of having formed in a temperate climate rather than a tropical climate (see Table 3).

Table 2. Chilean REE deposits and occurrences hosted in regoliths.

Sector	Region	Protolith	Profile Thickness	Reported REE	Exchangeable Fraction	Distinctive Feature	Evidence Level	References
Penco Module (six zones)	Biobío	Granitic rocks	Regolith developed on granite	2292 ppm TREO and 1932 ppm TREY (M + I); 380 ppm Nd + Pr; 66 ppm Dy + Tb	Average recoveries of 22.82% TREY, 43.23% Dy and 19.53% Nd	27.5 Mt M + I and 1.7 Mt Inferred; 62.9 and 3.4 kt TREO contained	E1	[28,34]
Nahuelbuta Range	Biobío–La Araucanía	Biotite and amphibole–biotite tonalites, Late Carboniferous Permian	Up to 60 m	Up to 2000 ppm exchangeable REE	Zones of 350–500 ppm with extensions of 25–50 m; sectors above 500 ppm up to ~100 m	HREE up to ~56% of total REE in preserved regoliths	E3	[13]
Concepción-Talcahuano	Biobío	Calc-alkaline granitoids	Not quantified	Not quantified	Not determined	Allanite and monazite as carrier phases susceptible to alteration	E4	[40]
Sediments of Los Ríos	Los Ríos	Granitic and possibly pegmatitic sources	Not applicable	Enrichment of 4.4 to 5.7 times relative to upper continental crust (La-Gd)	Not applicable	Possible extension of prospective belt by more than 200 km to south	E5	[33,39]

Table 3. Comparison of Penco Module with international reference ion-adsorption deposits.

Deposit	Country	Tonnage	Average Grade	Formation Climate	References
Penco (Measured + Indicated)	Chile	27.5 Mt	2292 ppm TREO (0.229 wt%)	Temperate	[17,34]
Zudong	China	Resource of ~17,600 t oxides	~0.1 wt%	Subtropical	[41,46]
Serra Verde	Brazil	458 Mt	0.098 wt%	Tropical	[14]
Ambohimirahavavy	Madagascar	628 Mt	0.089 wt%	Tropical	[14]
Makuutu	Uganda	532 Mt	0.064 wt%	Tropical	[14]
Koppamurra	Australia	186 Mt	0.071 wt%	Temperate	[14]
Deep Leads	Australia	52 Mt	0.081 wt%	Temperate	[14]

This combination has direct consequences on the business model and exploration strategy. A high-grade, low-tonnage deposit sustains modular operations of limited life, in which supply continuity depends on the successive incorporation of new modules rather than the expansion of a single pit, and it shifts the focus of exploration from the search for large flat surfaces towards the identification of preserved profiles on fertile protoliths. A review of mineral systems for ion-adsorption deposits supports this reading by concluding that the REE source is more diverse than traditionally assumed, that adsorption is controlled by clay mineralogy and pH, and that preservation requires a low-erosion environment, with excessive precipitation being a factor that can destroy the profile by clay dissolution and saprolite collapse [14,45]. A spectroscopic study of speciation confirms that heavy REE are adsorbed as hydrated outer-sphere complexes, with coordination 8 to 9, predominantly on kaolinite, which supports not only the leachability of the resource but also its sensitivity to the physicochemical conditions of the profile [15].

Figure 3 compares the grade–tonnage relationship of the Penco Module with that of the ion-adsorption deposits that sustain the world production of heavy rare earth materials. Penco occupies the high-grade, low-tonnage end: its 27.5 Mt of Measured and Indicated resources at 2292 ppm TREO exceed the ceiling of the characteristic range of this type of deposit (0.05 to 0.2 wt% REO). It is two to three times the grades of Sierra Verde, Ambohimirahavavy, Makuutu, Koppamurra and Deep Leads, but its tonnage is one to two orders of magnitude lower than those deposits, which range between 52 and 628 Mt. Metal content isolines show that the grade advantage does not compensate for that difference: Penco lies on the isoline of about 60 kt contained oxides, versus 100 to 1000 kt in the Brazilian, Malagasy and Ugandan deposits. That combination sustains modular operations with a limited life rather than a single large-scale exploitation. Cerro Carmen and El Buitre, in contrast, fall below the characteristic grade band of ion-adsorption clays, with 19.81 Mt at about 414 ppm and 6.6 Mt at approximately 337 mg/kg, expressed as elemental rare earth resources and not as TREO.

Figure 3. Grade–tonnage relationship of Chilean rare earth resources compared with international reference ion-adsorption deposits. Dashed diagonal lines are isolines of rare earth contents, where content in tons equals tonnage in Mt multiplied by grade in ppm. The shaded band marks the typical grade range of ion-adsorption deposits (0.05–0.2 wt% REO).

2.2.3. Nahuelbuta and the Regional Formation Model

The work [13] substantially modifies the interpretation of Penco because it demonstrates that REE mineralization hosted in a regolith should not be considered a local singularity. In the Nahuelbuta Range, between the Biobío and La Araucanía regions, the authors studied a regolith developed on biotite tonalites and amphibole–biotite tonalites from the Late Carboniferous–Early Permian age. Allanite and monazite constitute the main REE-carrier minerals of the parental rock, while the weathered profiles extend to depths of up to 60 m and the exchangeable REE content reaches to 2000 ppm.

Mineralization also shows vertical and geomorphological fractionation. Preserved regoliths can be dominated by HREE, reaching approximately 56% of total REE, while partially eroded profiles show the relative predominance of LREE. The progressive alteration of allanite and monazite and the advance of the pedolith–saprolite interface favor the release, migration, and subsequent retention of REE in new clay phases [13,47]. Three-dimensional modeling has identified zones with 350–500 ppm exchangeable REE with approximate extensions of 25–50 m, while sectors with more than 500 ppm can extend to about 100 m; mineralization above 350 ppm is normally concentrated between a 5 and 10 m depth [13].

These results provide a guide that is particularly important for Chile. The existence of granitic or tonalitic rock enriched in REE constitutes only the first requirement. The generation of a deposit also depends on the primary mineralogy, the weathering intensity, the depth of the pedogenic front and, decisively, the geomorphological preservation of the regolith. This explains why profiles developed on similar protoliths can present very different types of contents and HREE/LREE proportions [40]. The consistency of these observations with internationally established controls is noteworthy: the hillslope dynamics, topographic position and erosion rate govern the formation and conservation of the deposit, with a weight that is comparable to protolith composition [14].

Figure 4 summarizes the genetic model of regolith-hosted mineralization and separates its two independent controls. Panel A provides an order to the sequence along a profile: the progressive alteration of allanite and monazite releases REE³⁺ in the upper horizons, which migrate with pore water and are retained by adsorption on kaolinite as hydrated outer-sphere complexes. This generates an accumulation zone with up to 2000 ppm exchangeable rare earth materials and grades above 350 ppm concentrated between 5 and 10 m depth. Meanwhile, the alteration front advances downward and renews input from a tonalitic protolith [13,15]. Panel B illustrates that this sequence is not enough: geomorphological preservation decides both the grade and elemental signature. Therefore, the preserved profiles on low-slope surfaces consist of a proportion of heavy rare earth materials that can be up to 56% of the total amount, while truncated profiles on incised slopes lose the precisely highest-value fraction and remain dominated by light rare earth minerals [14]. The implication for exploration is direct: the search criterion is not lithology alone but its coincidence with low-erosion surfaces.

Figure 4. Conceptual model of rare earth mineralization hosted in regolith developed on tonalite in Coastal Cordillera. (A) Weathering profile from soil to parental rock with progressive alteration of allanite and monazite, downward migration of REE³⁺ and retention by adsorption on kaolinite in upper saprolite. (B) Geomorphological control on profile preservation and elemental signature: preserved profiles reach HREE proportions close to 56% of total REE, while partially eroded profiles remain dominated by LREE.

2.2.4. The Central-Southern Prospective Corridor

The regional continuity of this system receives independent support from the Los Ríos Region. SERNAGEOMIN identified that there are sediments enriched between 4.4 and 5.7 times relative to the upper continental crust for REE between La and Gd, associating them with the erosion of granitic sources and possibly pegmatitic levels. The study proposes that the previously recognized Paleozoic prospective belt between Biobío and La Araucania could extend more than 200 km to the south. Together, Penco, Nahuelbuta and Los Ríos configure a central-southern prospective corridor in which the search follows the protolith → weathering → release from allanite/monazite → adsorption on clays → preservation of the regolith relationship [33,39].

The latitudinal extension of this corridor poses a question that the current data does not resolve and that conditions the size of the aggregate resource: to the south, increased precipitation and the imprint of Quaternary glaciation operate in conflict with one another in the preservation of profiles, favoring chemical weathering but also resulting in the stripping of regoliths and the dissolution of clay [14,39]. Delimitation of the corridor therefore requires integrating mapping of low-slope paleosurfaces with sampling of the complete profile and not only the geochemistry of the surface cover.

2.3. Primary Rock Resources: Cerro Carmen and the Iron Belt

2.3.1. Cerro Carmen

Cerro Carmen represents the main Chilean example of primary REE mineralization with a historical tonnage estimate. Work by the mining company Empresa Nacional de Minería (ENAMI) and the Chilean Nuclear Energy Commission (CChEN) carried out in the Coastal Cordillera of the Atacama Region estimated that there was approximately 19.81 Mt of ore with 8203 t of REE and 1811 t of U contained, integrating the Hypothetical, Inferred and Indicated categories. Within this, one category contained 3.88 Mt with 2944 t of REE, equivalent to an approximate grade of 760 ppm plus 595 t of U [28,30].

This deposit is related to a skarn-type system in Cretaceous intrusive and volcanic rocks and contains REE, Fe, U and Th oxides, together with ilmenite, davidite and sphene. Its composition is of particular interest because of the presence of HREE, with approximately 140 ppm Y, 20 ppm Dy, 5 ppm Ho, 21 ppm Er and 36 ppm Yb. Historical tests have also enabled the production of concentrates that are strongly enriched in REE oxides. The figures, however, predate current reporting standards and must be considered historical estimates that require updates rather than directly comparable with the estimate in the CIM standard used in Penco [19,28].

Association with uranium and thorium, which in the historical estimate reaches 1811 t of contained U, is both an attribute and a restriction. It is an attribute because systems in which U and REE share carrier phases usually admit joint recovery schemes, and because davidite and sphene documented in the deposit are minerals with a demonstrated capacity to incorporate REE. It is also a restriction because handing materials with natural radionuclides needs authorization, transport, occupational exposure and waste disposal requirements that do not apply to an ion-adsorption regolith. Any modern reevaluation of Cerro Carmen must therefore quantify the specific activity of the ore and intermediate products together with the REE grade [10,19].

2.3.2. IOA-IOCG Deposits and Apatite as Host

The literature about the Chilean Iron Belt considerably broadens the potential of the northern part of the country, where numerous iron oxide–apatite (IOA) deposits and related IOCG systems are concentrated. The importance of these deposits lies less in a currently demonstrated global REE grade than in the existence of extraordinary volumes of iron ore containing apatite and other phases capable of incorporating REE [35,48,49,50].

At Los Colorados, the largest IOA deposit in the belt, approximately 900 Mt of ore at 34.6% Fe has been reported. This figure corresponds to the iron resource and cannot be used as REE tonnage; however, it establishes the scale of the available host material. Apatite from Los Colorados reveals REE enrichment, and LA-ICP-MS analyses demonstrate differential concentrations of La, Ce, Nd, Dy, Yb, Lu and Y among crystalline domains. Evolution from F-rich cores toward domains related to Cl-rich magmatic-hydrothermal fluids also demonstrates that brines could mobilize P and REE during deposit formation. Internal zones of certain apatite crystals show relative enrichment in La, Ce, Nd, Dy and Yb, confirming that REE concentrations can vary substantially even within the same mineral phase [35].

The same association is recognized in other deposits. In El Romeral, Rojas et al. [51] describe IOA systems as an Fe source that potentially also contains REE, U, Ag and Co within an Iron Belt of approximately 1000 km, grouping more than 50 important deposits. The mineralogical relevance comes from that substitution of Ca²⁺ in the structure of apatite that allows for the inclusion of REE³⁺, plus U, Th, Sr and Ba [48]. As with the example of Los Colorados, the tonnage figures available for El Romeral correspond to iron resources and reserves; no bulk REE grade and no REE resource have been reported for the deposit.

In El Laco, the authors of [52] identified monazite and thorite inclusions both in magnetite and apatite. These textures are interpreted as evidence of metasomatic redistribution and the local supersaturation of REE-Th-P-Si during fluid–mineral interaction, showing that REE can pass from apatite to new phosphate phases during hydrothermal evolution. At Cerro Negro Norte, [49], a similar magmatic-hydrothermal evolution has been confirmed, and magnetite bodies with actinolite and subordinate apatite within the Cretaceous segment of the Iron Belt have been documented. In both cases, the evidence is mineralogical: the tonnages published for these deposits are iron resources, and no bulk REE grade or REE resource has been reported for El Laco or Cerro Negro Norte.

The economic significance of these results must be kept separate from the mineralogical evidence. The studies convincingly demonstrate that apatite, monazite and other accessory phases constitute REE reservoirs in Chilean IOAs but do not provide an average REE grade directly multipliable by hundreds of millions of tons of iron ore [53]. A future priority should therefore be quantifying the modal abundance of apatite, the elemental content of apatite, the distribution of REE among apatite, monazite, xenotime and silicates, and the partitioning of these phases among Fe concentrates, tailings and other process flows. This information will enable the transformation of the enormous geological inventory of the Iron Belt into a technically defensible REE inventory [35,49,52].

2.3.3. An International Reference: Rare Earth Materials as Iron By-Products

A Swedish example helps gauge what is missing in the Chilean case. This example began from an analogous geological configuration (iron oxide–apatite mineralization) and went from a mineralogical observation to the formal report of an REE resource. In the Kiruna district, apatite mineralization associated with iron was declared in 2023 as the largest rare earth deposit known in Europe, with more than one million tons of oxides, reported according to the PERC 2021 standard [54]. The relevant point is not the figure but the sequence: a declaration was possible because a previously modal abundance of apatite had been quantified, its REE content characterized, and the behavior of the material in the beneficiation circuit demonstrated.

The corresponding metallurgical chain is also documented. Iron tailings from Kiruna, Sweden, contain 1500 mg/kg REE, and after apatite beneficiation, the concentrate reaches approximately 5000 ppm, with REE hosted in apatite and monazite. Leaching with HNO₃ allows for the simultaneous recovery of phosphorus as H₃PO₄ and the dissolution of REE, with extraction ratios of 75% to 100% for heavy rare earth material and removal of Ca as Ca(NO₃)₂ by cooling the liquor [55]. On this basis, an industrial scheme was developed that recovers phosphorus, rare earth material, fluorine and gypsum from apatite-rich tailings of iron operations [54].

Transferring this model to the Chilean Iron Belt is plausible in geological and mineralogical terms, given that apatite and monazite are also the dominant carrier phases there. However, it requires three measurements that have not yet been published for Chilean deposits: the modal abundance of apatite in the iron ore, the average REE content of that apatite at deposit scale (and not only in selected crystalline domains), and the deportment of carrier phases between iron concentrate and tailings. The second measurement is critical because the available data for Chile comes from LA-ICP-MS on individual crystals, a technique that resolves internal zonation with excellent detail but does not provide in and of itself a representative average for the deposit [35,48].

2.3.4. The Regional Geochemistry of the Belt

The regional geochemistry of Vallenar reinforces this prospective line. The SERNAGEOMIN map recognizes a La-Ce-Pr-Nd-Eu association in the northwestern sector of the sheet, accompanied by HREE and Y, and the identified mineralogy (amphibole, calcite, albite and pyrite) led to proposing an IOCG affinity. However, the cartographic intervals, which reach upper classes of 513.4 ppm Ce, 306.1 ppm La, 169 ppm Nd and 51.47 ppm Pr, are sediment values and do not constitute grades of a mineral resource [38]. Table 4 summarizes the main Chilean deposits and systems with REE associated with primary mineralization, comparing tonnages, contents, carrier phases and evidence levels, together with Kiruna as an international point of reference.

Table 4. Primary REE mineralization in rock documented in Chile and international point of reference.

Deposit or System	Region	Type	Reported Tonnage (Basis)	REE Content	Carrier Phases	Evidence Level	References
Cerro Carmen	Atacama	Skarn in Cretaceous intrusive and volcanic rocks	19.81 Mt (Hypothetical, Inferred and Indicated categories); 3.88 Mt indicated	8203 t REE and 1811 t U contained; 2944 t REE in Indicated category (~760 ppm); ~140 ppm Y, 36 ppm Yb, 21 ppm Er, 20 ppm Dy, 5 ppm Ho	REE-Fe-U-Th, oxides, ilmenite, davidite, sphene	E2	[10,28,30]
Los Colorados	Atacama	IOA	~900 Mt with 34.6% Fe (iron resource; not convertible to REE tonnage)	Enrichment in La, Ce, Nd, Dy, Yb, Lu and Y by crystalline domains; without published global grade	F-Cl zoned apatite	E3	[35]
El Romeral	Coquimbo	IOA	Iron resource (active CAP operation); no REE resource or grade reported	REE, U, Ag and Co as elements associated with Fe; no bulk REE grade published	Apatite (substitution of Ca2+ by REE3+, U, Th, Sr, Ba)	E3	[51]
El Laco	Antofagasta	IOA	Iron resource; no REE resource or grade reported	Local supersaturation of REE-Th-P-Si; no bulk REE grade published	Monazite and thorite as inclusions in magnetite and apatite	E3	[52]
Cerro Negro Norte	Atacama	IOA	Iron resource (CAP operation); no REE resource or grade reported	REE not quantified; no bulk grade published	Magnetite with actinolite and subordinate apatite	E3	[49]
Hoja Vallenar (geochemistry)	Atacama–Coquimbo	Regional anomaly of IOCG affinity	Not applicable	Upper classes of 513.4 ppm Ce, 306.1 ppm La, 169.2 ppm Nd, 51.47 ppm Pr in sediments	Association with amphibole, calcite, albite and pyrite	E5	[38]
Kiruna (international reference)	Sweden	IOA	More than 1 Mt REE oxides reported under PERC 2021	Tailings ~1500 mg/kg; apatite concentrate ~5000 ppm	Apatite and monazite	E1	[54,55]

Compiled from sources indicated in reference column. For all IOA entries, reported tonnages correspond to iron resources or reserves. They are not REE resources and must not be multiplied by apatite-scale REE concentrations to derive REE contained within them.

2.4. Pegmatites, Placers and Sedimentary Dispersion

2.4.1. Pegmatites of the Paleozoic Batholith

Pegmatites of the Paleozoic batholith of the Coastal Cordillera represent a third expression of primary mineralization. In the Biobó Region, occurrences of Vertientes, Poñén and Coyanmahuida demonstrate the presence of specific REE and U-Th minerals. At Vertientes [19], monazite and xenotime associated with thorite and Th ± U silicates have been identified, while surrounding granitic rocks contain monazite and allanite. The average concentration reported for Vertientes reaches approximately 610 ppm REE, versus 321 ppm in Poñén. Previous studies compiled for Coyanmahuida report much higher values, on the order of 10,070 ppm, although these occurrences still lack tonnages allowing for conversion into mineral resources.

Their importance may be more related to the role of these pegmatites as sources of resistant minerals than to eventual direct exploitation: monazite, xenotime, allanite, magnetite, ilmenite and zircon can be released during weathering and subsequently concentrated by hydraulic selection [19,56]. Researchers have precisely proposed that proximity of Paleozoic pegmatites to the Biobío coast could feed heavy sand deposits [19]. This dual function (source of refractory ore and source of adsorbable REE after alteration of its carrier phases) is what confers an exploratory value of pegmatites that is higher than suggested by their point grades, since the differential susceptibility of allanite, monazite and xenotime to weathering controls which elements remain available for adsorption and which remain in resistant phases [14].

2.4.2. Placers and Heavy Sands

The Fe-Ti placer of Playa Blanca, located north of Concepción, constitutes a concrete example of this process. The deposit extends approximately 1500 m in length and 30–50 m in width and contains black sand horizons enriched with titanomagnetite, magnetite, ilmenite, hematite and sphene. Although the historical study was mainly oriented to Fe-Ti, demonstration of monazite, xenotime and allanite in nearby source rocks justifies systematically evaluating heavy mineral fractions for their REE contents. In this case, the relevant concept is not yet an “REE resource” but a verifiable genetic route between the primary source, erosion, transport and sedimentary concentration [19].

Evaluating this type of material consists of an internationally standardized procedure: separation of the heavy mineral fraction by density and magnetic susceptibility, modal quantification of monazite, xenotime and zircon, and analysis of the concentrated fraction instead of the total sample [57,58]. A systematic measurement of the sand underestimates the interest in the deposit because it dilutes carriers in a quartz and feldspar matrix; measurement on concentrate, in contrast, directly reveals the product that an operation would deliver [2].

2.4.3. Magmatic Fertility in Coquimbo and Magallanes

In the Coquimbo Region, a map of the Cordillera de Doña Rosa-Paso de Los Azules has documented granitoids with REE enrichment patterns relative to HREE and late pegmatite dikes up to approximately 1 m composed of feldspar, quartz and white mica. These features are useful for recognizing magmatic fertility and a possible primary REE source, but available data are mainly petrogenetic and do not allow for assigning tonnages or ore grades [37].

Further south, in the Bahía Blanca-Caleta Benavente area of Magallanes, the Svea Granite contains xenotime and REE-rich epidote as accessory minerals, while mineralized sectors present REE, Ti, Th and U associations linked to titanite–leucoxene. The study included a multielement analysis of mineralized samples and confirms the existence of REE occurrences in southernmost Chile; however, it does not provide a resource estimate and must be interpreted as mineralogical and geochemical evidence for exploration [36].

2.4.4. Sedimentary Dispersion and Drainage Anomalies

Geochemical evidence of fluvial sediments broadens this perspective to the national scale [59], has determined Ce, La, Nd, Pr and Y in surface sediments of rivers of northern, central and southern Chile, and found higher concentrations of Ce, La, Nd and Pr in northern and central sectors. These data are valuable as baseline and provenance information, but caution must be taken in interpretation: REE can be found adsorbed on fine particles and subsequently remobilized towards the water column. Consequently, a fluvial anomaly may reflect the erosion of an enriched source, the sedimentary concentration or surface geochemical processes and does not by itself demonstrate the presence of an ore.

The study of Los Ríos provides an example of how to advance from regional anomaly toward a source model. The integration of national geochemical data has identified anomalous REE zones, while in the Los Ríos Region, spatial distribution, clast mineralogy and normalized REE patterns help relate anomalies to granitic and potentially pegmatite materials. This approach (anomalous sediment → provenance basin → source lithology → weathered profile) should constitute one of the main tools used to select exploration targets in regions where specific drilling for REE does not yet exist [33,39]. Table 5 synthesizes the main pegmatite, coastal, granitic and sedimentary occurrences with REE in Chile, comparing contents, carrier phases, exploratory contributions and evidence levels.

Table 5. Pegmatites, placers and sedimentary anomalies of REE in Chile.

Occurrence	Region	Type	Reported REE Content	Carrier Phases	Exploratory Contribution	Evidence Level	References
Vertientes	Biobío	Pegmatite of Paleozoic batholith	~610 ppm REE	Monazite, xenotime, thorite, Th ± U silicates	Primary source of resistant minerals	E3	[19]
Poñén	Biobío	Pegmatite of Paleozoic batholith	~321 ppm REE	Monazite, allanite in surrounding granites	Primary source of resistant minerals	E3	[19]
Coyanmahuida	Biobío	Pegmatite of Paleozoic batholith	~10,070 ppm REE (values compiled from previous studies)	Not detailed	High-grade occurrence without tonnage	E4	[28]
Playa Blanca	Biobío	Coastal Fe-Ti placer	Not determined for REE	Titanomagnetite, magnetite, ilmenite, hematite, sphene	Verifiable source–erosion–concentration genetic route	E4	[19]
Cordillera de Doña Rosa-Paso de Los Azules	Coquimbo	Granitoids and late pegmatitic dikes	Relative enrichment of LREE over HREE	Feldspar, quartz, white mica in dikes up to ~1 m	Magmatic fertility	E4	[37]
Bahía Blanca–Caleta Benavente (Svea Granite)	Magallanes	Granite with REE accessories and mineralized sectors	Multielement analysis without resource estimation	Xenotime, REE-rich epidote, titanite–leucoxene with REE-Ti-Th-U	Southern extension of occurrence inventory	E4	[36]
National fluvial sediments	North, central and south	Geochemical anomaly	Ce, La, Nd and Pr higher in northern and central sectors	REE adsorbed on fine particles	Baseline and provenance	E5	[59]

Compiled from sources indicated in reference column.

2.5. Tailings as a Secondary Resource

2.5.1. Magnitude and Distribution of Inventory

Tailings probably represent the most directly accessible alternative REE resource for Chilean mining, although they are also among the most heterogeneous [60,61,62,63]. Their main advantage is that the material has already been extracted, crushed and ground, eliminating an important part of the operations required by a primary deposit [64]. Their disadvantages are that grades are usually low and REE can be distributed among apatite, phyllosilicates and other phases of varying metallurgical behavior [20,27,65].

National cadastres are an exceptionally broad base from which to study this potential [20]. Integrated data from 642 Chilean deposits within a multinational base of 2976 samples have indicated that the SERNAGEOMIN information contains more than 2000 geochemical results generated since 2015, including fourteen REE determined mainly by ICP-MS. However, the same study concludes that these databases are still insufficient to declare secondary resources because they normally lack systematic information on mineralogy, grain-size distribution, three-dimensional continuity, and metallurgical response.

The most recent analysis of 653 deposits carried out by the authors of [21] grouped Chilean tailings into four main geochemical profiles. Zn-Pb-Cd-As polymetallic deposits present the highest relative concentrations of heavy metals and REE, so they combine economic potential with high environmental interest. The authors emphasize that chemical composition alone is not enough to establish reprocessing feasibility and must be complemented with mineralogy and physical properties.

Historical background also shows relevant values in tailings from Cu mining of north-central Chile. COCHILCO reported approximately 525 ppm REE on average in the Tranque Carola, with a sample reaching 868 ppm, while Planta Vallenar and Tranque El Salado presented approximately 370 and 350 ppm, respectively. These grades are within the same order of magnitude as several materials investigated internationally for secondary recovery, but their real interest depends on the tonnage and the mineral phase in which REE are hosted [10].

Technical manuals developed by JRI Ingeniería and EcoMetales provide an additional criterion by which to select tailings with valorization potential. The reprocessing guide suggests, as a preliminary filter, considering a total REE content above approximately 300 ppm together with tonnage and the presence of other valuable elements. In the candidate base presented, records close to 1692 ppm REE in Taltal, 4323 ppm in Copiapó and 812 ppm in Coquimbo are included. These values correspond to a screening stage and do not constitute resources but rather demonstrate that there are individual deposits justifying more detailed mineralogical and metallurgical characterization [66].

Mineralogical characterization is particularly important because the total concentration does not by itself identify the carrier phase [67]. In tailings studies of the “Tailings with Value” program, positive relationships between Ce-La and Al₂O₃ were observed, as well as associations between Ce-La and sericite/muscovite, suggesting that part of REE may be linked to phyllosilicates. The authors themselves warn that statistical correlation does not demonstrate a petrogenetic relationship and must be verified through mineralogical evidence, an essential criterion before designing a recovery route [68].

2.5.2. El Buitre: A Secondary Resource That Is Three-Dimensionally Quantified

The El Buitre tailings deposit in Tierra Amarilla is one of most complete cases for evaluating this category [18]. It hosts 28 drill holes from which 755 samples were analyzed, enabling the construction of a three-dimensional geostatistical model and the study of the deposit to depths far greater than what is normally considered in surface inventories.

The model obtained an average grade of 324 mg/kg LREE, with a range of 209.1–524.2 mg/kg, and 21.5 mg/kg HREE, with a range of 14.6–29.1 mg/kg. LREE demonstrate a relatively homogeneous distribution, although there are layers above 350 mg/kg; HREE present more marked spatial variation. Considering the complete deposit of approximately 6.6 Mt, the authors used the total REE grade of around 337 mg/kg, equivalent to an in situ inventory on the order of 2.2 kt REE. This amount must be understood as an exploratory estimate obtained from a tailings model and not as a mineral reserve under an international reporting code [18].

Mineralogy also explains why contained tonnage does not equal recoverable tonnage. In deep zones, P presents positive correlations with LREE and HREE, indicating apatite contribution, while HREE also show an important association with Al, compatible with their hosting or adsorption in silicate and phyllosilicate phases. This distribution implies that the same tailings may require different metallurgical routes to mobilize REE present in phosphates and those retained in silicates [18]. The LREE/HREE ratio of the deposit, close to 15:1, constitutes data as relevant as the total grade because the highest unit value fraction—including Dy, Tb and Y—here represents little more than 6% of the inventory. Tailings with 337 mg/kg REE dominated by Ce and La do not economically equal a regolith with the same grade and 56% HREE, even when both figures are expressed in the same units [13,18].

2.5.3. Recoverability and Economic Viability

The possibility of effectively recovering these elements has been demonstrated experimentally in Chilean mining residues [69]. Solid residues associated with processing copper minerals in Atacama have been studied, revealing that they obtain a maximum recovery of 64.5% total REE through leaching with 3 M HCl at 40 °C and a liquid/solid ratio of 4. Under these conditions, La reached 75.7% recovery and Ce reached 70.0% recovery. The reported consumption was 11 kg HCl/t, while HNO₃ required 29 kg/t and reached a lower total recovery. The authors obtained 0.355 kg REE of residue recovered per ton using HCl and also pointed out the need to remove Fe from solution before subsequent separation stages.

This observation about iron anticipates the central problem of reprocessing tailings from IOCG and IOA systems: the gangue is as soluble as the mineral of interest. In tailings rich in iron oxides and silicates, REE recovery simultaneously faces low grades, incomplete liberation and reagent consumption dominated by valueless phases. Therefore, the viable route usually passes through physical preconcentration (flotation or magnetic separation) before leaching and through process mineralogy characterization that identifies which carrier is hosted in each grain-size fraction [69,70,71,72,73]. Sweden’s experience with iron tailings points in the same direction, recovering apatite as an intermediate concentrate before attacking REE, with phosphorus as a co-product absorbing part of the cost [55,74,75].

Technical feasibility, however, does not guarantee profitability [27,76]. The data evaluated from 477 Chilean installations have modeled the production of Ce, La, Nd, Y, Sm, Gd, Pr and Dy oxides. The scenario calculated has produced a positive NVP but with only approximately US$ 0.67 million, with an IRR close to 10.03%, which is practically equivalent to the discount rate used; the authors concluded that profitability was marginal and especially sensitive to the REO price and discount rate. These figures correspond to assumptions and prices used in the study and should not be used as a current economic valuation of the Chilean inventory [77].

Subsequent analysis through real options reached corresponding conclusions: some elements contained in tailings, particularly Sc, could attain economically attractive conditions, but CAPEX constitutes one of the variables with the greatest impact on viability. Therefore, the opportunity of tailings does not necessarily reside in developing mines exclusively for REE but rather in integrating these elements within polymetallic schemes in which Cu, Fe, Co, V, Sc or other products contribute to absorbing reprocessing costs [78,79].

The case of El Buitre adequately summarizes this situation: there is a quantifiable and three-dimensionally modeled inventory, but LREE and HREE are distributed between apatite and silicates, so effective recovery depends on selective mineral liberation and leaching. The transition from “tailings with REE” to an economically recoverable secondary resource therefore requires simultaneously demonstrating tonnage, continuity, speciation, liberation, recovery and separation costs [18,66,68,80]. Table 6 summarizes the potential of tailings and mining residues as secondary sources of REE.

Table 6. REE content reported in Chilean tailings and mining residues, with comparable international reference.

Deposit or Set	Region or Country	Database	REE Contents	Inferred Carrier Phases	Demonstrated Recovery	Evidence Level	References
El Buitre	Atacama	28 drill holes, 755 samples, 3D geostatistical model; ~6.6 Mt	324 mg/kg LREE (209.1–524.2) and 21.5 mg/kg HREE (14.6–29.1); ~337 mg/kg total REE; ~2.2 kt contained	Apatite (correlation with P) and silicates/phyllosilicates (correlation with Al)	Not tested in same study	E2	[18]
Tranque Carola	Atacama	COCHILCO sampling	~525 ppm average; max 868 ppm	Not determined	Not tested	E5	[10]
Planta Vallenar	Atacama	COCHILCO sampling	~370 ppm	Not determined	Not tested	E5	[10]
Tranque El Salado	Atacama	COCHILCO sampling	~350 ppm	Not determined	Not tested	E5	[10]
Screening candidates (Taltal, Copiapó, Coquimbo)	Antofagasta–Atacama–Coquimbo	Reprocessing base with threshold ~300 ppm	~1692 ppm (Taltal); ~4323 ppm (Copiapó); ~812 ppm (Coquimbo)	Not determined	Not tested	E5	[66]
National cadastre	Chile	642 deposits within base of 2976 samples; more than 2000 geochemical results since 2015	Fourteen REE determined mainly by ICP-MS	Not determined	Not applicable	E5	[20]
National geochemical profiles	Chile	653 deposits grouped in four profiles	Highest relative REE concentrations in Zn-Pb-Cd-As deposits	No determined	No applicable	E5	[21]
Cu processing residues	Atacama	Leaching tests	64.5% total recovery; La 75.7%; Ce 70.0%; 0.355 kg REE/t	Not determined	HCl 3 M, 40 °C, L/S = 4; 11 kg HCl/t	E3	[69]
Kiruna iron tailings (international reference)	Sweden	Industrial operation	~1500 mg/kg in tailings; ~5000 ppm in apatite concentrate	Apatite and monazite	75–100% HREE and >99% P by leaching with HNO3	E1	[54,55]

2.6. Marine Resources: Ferromanganese Nodules and Crusts

2.6.1. Recognized Inventory in Exclusive Economic Zone

A category that broadens the traditional perception of Chilean resources is the seabed of the Exclusive Economic Zone [11], identified as having the most relevant non-energy mineral resources: Fe-Mn nodules and crusts, phosphorites, Au-Ti placers and environments, and potentially massive sulfides. Ferromanganese crusts are particularly relevant for REE because, on a global scale, they present important enrichments in Co, Te, Mo, Bi, Pt, W, Zr, Nb, Y and rare earth materials and can be more enriched in REE than polymetallic nodules [81,82].

In Chilean waters, there are occurrences of nodules at depths of nearly 2900–4300 m, with Co contents locally reaching approximately 0.53% and Cu + Ni combinations up to 1.38%. These values illustrate the polymetallic richness of materials but should not be confused with the REE content [83], indicating that available chemical information is still insufficient to quantify critical elements such as Y and REE in Chilean samples. Therefore, these deposits must currently be classified as prospective REE resources, not as quantified national inventories.

The review by the authors of [12] confirms and expands this prospectivity. The Co-rich crust formed on seamounts typically contains Co, Ni, Te, Pt and REE, and favorable environments are recognized within the Chilean platform and oceanic margin. Occurrences of crusts and metalliferous sediments have been described around Salas y Gómez, Rapa Nui, San Félix and San Ambrosio, while the seamount systems of Salas y Gómez-Nazca and Juan Fernández provide extensive and potentially favorable surfaces. The O’Higgins seamount and other guyots of the Juan Fernández ridge also present morphologies compatible with crust formation.

2.6.2. Global References of Content and Enrichment Mode

The enrichment mechanism explains why these materials concentrate REE and why their grade is relatively stable on a global scale. Metals are sorbed from seawater onto two carrier phases of opposite surface charges—FeO(OH) with a positive surface charge and MnO₂ with a negative surface charge—so hydrogenetic crusts preferentially accumulate Co, Te, Mo, Bu, Pt, V, Zr, Nb, Y and REE while nodules become relatively more enriched in Ni, Cu and Li; in both cases, rare earth materials appear to be a potential by-product of the main metals and not the primary target [84,85,86,87]. A third marine material, pelagic muds rich in REE and Y described in the Pacific, operates with a different logic by concentrating elements in biogenic phosphates and presenting grades locally exceeding several thousand ppm [88], which has motivated systematic national characterization programs in other jurisdictions [89,90].

This contrast defines the minimum agenda in the Chilean case. Figures published for national waters correspond to Co, Cu and Ni, not to REE, and come from scattered samplings, so there is still no base that allows for a comparison of Chilean crusts with global ranges, nor that evaluates whether margin sediments contain accumulations analogous to those described in the western Pacific [11,12,83,91].

2.6.3. Environmental Restrictions

The geological potential is accompanied, however, by much greater environmental uncertainty than terrestrial resources [92,93,94]. Much of the “Salas y Gómez-Nazca” ridges constitute ecosystems of high biodiversity, and its sectors within the Chilean jurisdiction are protected. Therefore, before any economic discussion, it is necessary to establish an oceanographic and ecological baseline and obtain mineralogical and geochemical samples capable of effectively quantifying REE, Y, Co, Ni and other elements [95,96,97]. In the current state of knowledge, submarine resources are probably the category with the largest potential surface and the lowest characterization level within the Chilean REE inventory [12,98].

Along with the ecological restriction, a governance restriction operates independently of the degree of geological knowledge. Extraction of nodules and crusts is subject to unresolved international debate and is subject to a precautionary pause, as sufficient environmental regulations are not yet available. Chile’s public position in that discussion has been favorable to precaution [82,98,99]. Consequently, the immediate value of research on Chilean crusts and metalliferous sediments does not reside in enabling short-term exploitation but in building the geochemical and ecological baseline that any future decision (of exploitation or protection) will equally require [83]. Table 7 synthesizes the main submarine resources with REE potential in Chile.

Table 7. Marine resources with REE potential in Chilean oceanic territory.

Material	Recognized Location	Depth	Reported Contents	Quantified REE	Main Restriction	Evidence Level	References
Polymetallic nodules	Seabed of EEZ	~2900–4300 m	Local Co ~0.53%; Cu + Ni up to 1.38%	No; insufficient chemical information for Y and REE	Absence of REE data	E5	[11]
Ferromanganese crusts	Salas y Gómez, Rapa Nui, San Félix, San Ambrosio; guyots of Juan Fernández and O’Higgins seamount	Seamounts	Co, Ni, Te, Pt and REE as typical global association	No FOR Chilean samples	Biodiversity and protected areas	E5	[12,83,91]
Metalliferous sediments	Chilean margin and platform	Variable	Occurrences described without systematic characterization	No	Absence of baseline	E5	[12]
REE-rich crusts and muds (international reference)	Pacific	Seamounts and abyssal plains	Enrichment in Co, Te, Mo, Bi, Pt, W, Zr, Nb, Y and REE; muds with thousands ppm of REE + Y	Yes	Not applicable	E1–E2	[83,89]

2.7. Associated Radioactive Elements and Development Constraints

2.7.1. Thorium and Uranium as Transversal Restrictions

The association between the rare earth materials thorium and uranium is not incidental but structural. Trivalent REE share crystallographic sites with Th⁴⁺ and U⁴⁺ in monazite, thorite, davidite, xenotime and apatite, so the same phases that confer economic interest in a deposit determine its radiological load [1,2]. In the Chilean inventory, this association is documented in three primary domains: Cerro Carmen contains 1811 t of U estimated together with Th oxides; Vertientes pegmatites present thorite and Th ± U silicates accompanying monazite and xenotime; and in El Laco, inclusions of monazite and thorite in magnetite and apatite have been described [19,52].

The design of any recovery route must define, early on, the destination of Th and U, since these concentrate in leaching residues and can restrict the commercialization of intermediate concentrates [100]. Here, ion-adsorption deposits present an intrinsic and frequently underestimated advantage: as REE are recovered by the saline desorption of the adsorbed fraction and not by dissolution of phosphates, thorium contained in residual monazite remains mostly solid, which reduces the radiological load of the product with respect to routes based on monazite concentrates [15,42]. This difference constitutes a ranking criterion as legitimate as grade or tonnage and should be explicitly incorporated into comparisons among domains [101].

2.7.2. Licensing and Social Acceptance

Development of REE resources in Chile has been conditioned, to date, less by geological availability than by timelines and acceptance associated with environmental assessment [102,103]. The trajectory of the Penco Module illustrates this. Its environmental impact study entered the Environmental Impact Assessment System in June 2024, was processed with thirty agencies with environmental competence, required three addenda responding to more than two thousand technical and citizen observations channeled through the maximum number of rounds contemplated institutionally, and was unanimously approved by the Environmental Assessment Commission of the Biobío Region on June 8, 2026, with a consequent Environmental Qualification Resolution authorizing the development and operation of a project associated with an investment of US$ 130 million [31]. Approval did not end its controversy. Communities, organizations and municipalities have made claims before the Committee of Ministers, and regulatory certainty remains pending [34,104].

Two lessons emerge from this trajectory for the rest of the national inventory. First, the permitting cycle can have a timeframe comparable to exploration and engineering, so planning new modules must take this into account. Second, technical attributes reduce the footprint (absence of blasting, crushing and grinding, absence of tailings deposits, and water recirculation in desorption processes) and are verifiable arguments in environmental assessments and not only cost advantages, which reinforces the convenience of characterizing the desorbable fraction from the early stages of exploration [34].

2.8. Metallurgical Recoverability: Comparative Synthesis of Recovery Routes

2.8.1. Physical Preconcentration and Its Unequal Role Between Domains

The first decision of any route is whether the material must be concentrated before chemical attack. The answer separates the domains more distinctly than grade does. In hard-rock resources, gravity, magnetic and electrostatic separation and hydroxamate flotation raise the head grade by one to two orders of magnitude, and in iron systems, magnetic separation leaves apatite in the non-magnetic fraction from which a phosphate concentrate can be floated [105]. In ion-adsorption regoliths, the logic is reversed: the elements are adsorbed on clay surfaces; thus, grinding and flotation destroy the advantage of the deposit. In tailings, the answer is intermediate, since desliming concentrates the fine fraction but low liberation limits physical separation [106].

2.8.2. Saline Desorption of Ion-Adsorption Resources

Desorption is the only route with a commercial precedent that applies directly to the best-documented Chilean resource. Rare earth resources adsorbed on kaolinite and halloysite are displaced by monovalent cations; thus, a dilute salt solution releases them under ambient conditions [107,108]: standardized tests with ammonium sulfate report 80 to 90% total extraction at pH 3 to 4 and below 50 °C [109]. Because the residual phosphates are not dissolved, thorium and uranium remain largely in the solid, relieving the product of the burden that dominates monazite routes. Recovery is bounded by the desorbable fraction and not by the total content, so the exchangeable fraction must be measured explicitly. The Penco recoveries, between 19.53% for Nd and 43.23% for Dy, show that the relevant figure is elemental.

2.8.3. Acid and Alkaline Attack of Phosphate, Silicate and Oxide Carriers

When elements sit inside a crystal lattice, the structure has to be broken. Monazite and xenotime concentrates admit alkaline cracking with sodium hydroxide at 140 to 150 °C or sulfuric acid baking followed by water leaching [110]. Both require fine grinding and dissolve thorium and uranium with the rare earth material, so the gap in the Chilean pegmatite and placer domains is not technological but rather the absence of tonnage and concentrate tests. For apatite, the attack is milder, with nitric or sulfuric acid. For tailings and copper residues, direct acid leaching is the only route tested on Chilean material: 3 M hydrochloric acid at 40 °C gave 64.5% extraction and 0.355 kg of rare earth per ton treated. The limitation of this family is not extraction but selectivity, since iron, aluminum and calcium dissolve with the elements of interest.

2.8.4. Purification, Separation and Form of Final Product

Every route leads to the same problem: turning a loaded solution into a saleable product. The first stage is the removal of impurities through the hydrolytic precipitation of iron and aluminum, a prerequisite identified in the Chilean tests on copper residues. The second is the grouping of rare earth materials according to precipitation with oxalic acid, carbonate or double sulfates. The third is separation into individual products, dominated by countercurrent solvent extraction with acidic organophosphorus extractants [111,112], with ion exchange or resin-in-pulp for dilute or high-solid liquors [113,114]; separation factors between adjacent lanthanides are small, so tens to hundreds of stages are required, and most of the capital cost lies here [115,116]. No Chilean domain has demonstrated this segment; what has been shown is extraction, not separation.

2.8.5. Comparative Synthesis by Resource Type

Table 8 organizes the routes applicable to each Chilean domain by carrier phase, preconcentration, reagent and operating conditions and distinguishes figures obtained on national material from international references applicable by analogy. Reagent aggressiveness increases with the strength of the bond between element and host.

Table 8. Recovery routes applicable to Chilean rare earth resources: carrier phases, pretreatments, reagents and operating conditions.

Resource Type (Chilean Case)	Main Carrier Phase	Preconcentration or Pretreatment	Leaching or Desorption Reagent	Reported Operating Conditions	References
Ion-adsorption regolith (Penco, Nahuelbuta)	REE adsorbed on kaolinite and halloysite; residual allanite, monazite and xenotime	No crushing or grinding; screening and declaying; percolation in heaps or in situ	Ammonium sulfate 0.2–0.5 M; magnesium or sodium salts as low-nitrogen alternatives	Ambient temperature, below 50 °C; pH 3–4; solid-to-liquid ratio about 1:2; contact of about 1 h; barren solution recirculated	[15,117,118]
IOA-IOCG ore and iron tailings (Los Colorados, El Romeral, Cerro Negro Norte, El Laco)	Apatite; monazite and thorite inclusions in magnetite	Magnetic separation of magnetite followed by apatite flotation to a phosphate concentrate	Nitric or sulfuric acid on apatite concentrate	Acid digestion at moderate temperature with phosphoric acid or gypsum as co-product; not tested on Chilean material	[26,52]
Pegmatites and hard-rock monazite–xenotime (Vertientes, Poñén, Cerro Carmen)	Monazite, xenotime, allanite and thorite	Grinding below 45 micrometers; gravity, magnetic and electrostatic separation; flotation with hydroxamate collectors	Sodium hydroxide 45–50% (alkaline cracking) or sulfuric acid baking	Alkaline digestion at 140–150 °C for about 3 h; sulfuric baking between 200 and 800 °C followed by water leaching	[26,119,120]
Tailings and copper processing residues (El Buitre, Atacama residues)	Apatite (correlation with P) and phyllosilicates (correlation with Al)	Classification and desliming; magnetic separation to reject iron oxides; process mineralogy by size fraction	Hydrochloric acid 3 M (demonstrated); nitric and sulfuric acid with lower performance	40 °C; liquid-to-solid ratio of 4; 11 kg HCl per ton; iron removal required before separation	[18,54,69]
Placers and heavy mineral sands (Magallanes and coastal deposits)	Monazite, xenotime, zircon and ilmenite	Spirals, magnetic and electrostatic separation to a heavy mineral concentrate	Same alkaline or acid attack as monazite route	Not tested on Chilean material; rare earths handled as by-product of titanium and zirconium circuits	[26,119]
Ferromanganese nodules and crusts (Exclusive Economic Zone)	REY adsorbed on Fe-Mn oxyhydroxides and carbonate fluorapatite	Drying and grinding; no beneficiation route demonstrated	Reductive acid leaching, or dilute hydrochloric or nitric acid on deep-sea mud	Laboratory scale only; Mn-Fe matrix dissolves together with REY	[89,98]

Table 9 completes the comparison with recovery, selectivity, impurities, maturity and the limitation that governs each route. Only saline desorption combines commercial maturity with a Chilean resource under a reporting code, only direct acid leaching of residues has been demonstrated on national material, and the remaining four rest on international analogies: what is missing is not the process but the measurement of how the Chilean material behaves in it.

Table 9. Comparative performances of recovery routes: recoveries, selectivities, impurities, maturities and principal limitations.

Resource Type	Reported REE Recovery	Selectivity	Main Impurities in Solution	Technological Maturity	Principal Limitations	References
Ion-adsorption regolith	80–90% of total REE in standardized desorption tests; recoveries adopted at Penco between 19.53% for Nd and 43.23% for Dy	High with respect to Th and U, which remain in the solid; Ce partly excluded by its negative anomaly	Al, Ca, Mg and ammonium; low iron	Commercial abroad; in Chile, a permitted project with resources under code (E1)	Only the desorbable fraction is recoverable; co-extraction of Al and management of ammonium salts	[15,34,117,118]
IOA-IOCG ore and iron tailings	Above 99% of P and 75–100% of HREE from apatite concentrate in the Swedish case; not determined in Chilean deposits	The apatite concentrate rejects most silicates, but Fe and P report to solution	Fe, Ca, P, F and Th	Industrial demonstration abroad; evidence at E3-E4 in Chile	The REE grade of the ore at deposit scale is unknown; viability depends on iron and phosphate co-products	[26,52,54]
Pegmatites and hard-rock monazite–xenotime	90% or more for monazite concentrates under alkaline cracking; about 25% with dilute acid leaching without prior cracking	High for the REE group, but Th and U dissolve together and require a dedicated stage	Th, U, P and Fe	Mature and commercial worldwide; no test on Chilean material	Radiological management and absence of tonnage; fine grinding and high reagent consumption	[119,120]
Tailings and copper processing residues	64.5% of total REE (La 75.7%, Ce 70.0%); 0.355 kg REE per ton of residue	Low: Fe, Al and Ca dissolve together with the REE	Fe, Al, Ca and Cu	Laboratory scale (E3)	Marginal economics (net present value of about US$ 0.67 million and internal rate of return of 10.03% in the national assessment); viable only in polymetallic schemes	[27,69,78,79]
Placers and heavy mineral sands	Not determined in Chile; the physical concentration of heavy minerals is efficient and chemical attack follows the monazite route	High in the physical stage, low in the chemical stage	Ti, Zr, Th and Fe	Commercial abroad in mineral sands; E4 in Chile	No tonnage estimated; rare earths only as a by-product of titanium and zirconium	[26,119]
Ferromanganese nodules and crusts	Not tested on Chilean material	Low: the Mn and Fe matrix dissolves together with the REY	Mn, Fe, Co, Ni and Cu	Laboratory and conceptual (E5)	Environmental restrictions and absence of a declared resource or an environmental baseline	[89,98]

Three measurements would change this picture: the exchangeable fraction of the central-southern regolith corridor, which converts total grades into recoverable grades; the rare earth grade and deportment of the elements in the existing circuits of the Iron Belt, which establishes whether apatite can be recovered as a by-product; and the liberation and speciation of the carriers in tailings above the screening threshold. None of these require new exploration.

2.9. Ranking of National Potential

Integrating institutional reports and the scientific literature allows for Chilean REE resources to be organized according to the available evidence. At the first level of evidence is Penco because it has drilling, three-dimensional modeling, Measured–Indicated–Inferred classification, elemental grades, and metallurgical recoveries. It is therefore the only case reviewed that can be directly compared with international mineral resources under a formal standard, and since 2026, it has also had favorable environmental support [30,34].

The second level includes materials that have substantial quantification, but they cannot be directly equated with the Penco resource. Cerro Carmen has historical tonnages and REE contents, but it requires updating to modern standards. El Buitre, for its part, has 28 drill holes, 755 samples, and a three-dimensional model estimating approximately 337 ppm REE in 6.6 Mt of tailings, equivalent to about 2.2 kt contained, but it still corresponds to an exploratory secondary inventory and not to a formal mineral reserve [10,18].

At the third level are systems whose carrier mineralogy and enrichment mechanism are solidly demonstrated but where transforming mineralogical evidence into resource estimates is still lacking. Nahuelbuta regoliths contain up to 2000 ppm exchangeable REE and provide an exploration model directly applicable to the Coastal Cordillera. IOA deposits of Los Colorados, El Romeral, El Laco and Cerro Negro Norte demonstrate that apatite, monazite and other phases can concentrate REE within large-scale iron orebodies; however, the globally recoverable grade of ore is still unknown. Pegmatites of Vertientes, Poñén and Coyanmahuida demonstrate an equivalent situation: enrichments and well-defined carrier minerals exist, but there are no defined tonnages [19,35,48].

At the bottom level are drainage sediment anomalies, placers not yet specifically evaluated for REE, and submarine deposits. Their importance should not be underestimated: this data helps select the next exploration targets. However, multiplying a sediment concentration by regional surface, or a concentration measured inside an apatite crystal by the complete tonnage of an iron mine, would produce artificial inventories without any economic significance [33,38,39,59].

Figure 5 presents the scales at which rare earth concentrations have been reported in Chile, presented on a single logarithmic axis, together with the spatial support of each measurement. The comparison shows that a numerical range does not distinguish evidence levels. Grades documented at a deposit scale are concentrated between about 340 and 4300 ppm, an interval that also contains drainage sediment anomalies without a delimited volume and point values of pegmatites without tonnage, so Coyanmahuida, with approximately 10,700 ppm, exceeds any deposit grade in the country despite having a background with the least support ([33,38]). At the upper end, the carrier mineral phase is between 10² and 10³ times richer than any Chilean grade, a distance that corresponds exactly to the jump covered when a micrometric-scale measurement is extrapolated to regional volume. The right column clearly illustrates the operational consequences: only two of the five levels (the grade of the regolith profile and that of the sampled tailings deposit) have associated tonnages by which concentrations can be legitimately multiplied. Table 10 ranks the main REE resources and systems in Chile according to their levels of evidence, highlighting the available figures and gaps that are needed for better technical definitions.

Figure 5. Concentration scales of rare earth materials reported in Chile, ordered on a logarithmic axis according to spatial support for each measurement, from mineral grain to drainage basin. The right column indicates whether a level carries the tonnage by which a concentration can be legitimately multiplied.

Table 10. Synthesis of Chilean REE inventory ordered by evidence level.

Level	Resource or System	Domain	Available Figure	Gap Preventing Advancing Level	References
E1	Penco Module	Ion-adsorption regolith	27.5 Mt M + I at 2292 ppm TREO (62.9 kt) and 1.7 Mt Inferred (3.4 kt)	None regarding category; conversion to reserves after feasibility remains	[34]
E2	Cerro Carmen	Primary in rock	19.81 Mt with 8203 t REE and 1811 t U contained	Update under current reporting standard and modern metallurgical tests	[28]
E2	El Buitre	Tailings	~6.6 Mt at ~337 mg/kg; ~2.2 kt contained	Liberation and recovery tests; conversion to reportable category	[18]
E3	Nahuelbuta	Ion-adsorption regolith	Up to 2000 ppm exchangeable REE; profiles up to 60 m	Systematic drilling and tonnage estimation	[13]
E3	Iron Belt (Los Colorados, El Romeral, El Laco, Cerro Negro Norte)	IOA-IOCG	~900 Mt with 34.6% Fe in Los Colorados (iron resource, Los Colorados only; not a REE tonnage). No tonnage reported for El Romeral, El Laco or Cerro Negro Norte	Modal abundance of apatite, average REE grade and deportment in circuit	[35,51,52]
E3	Biobío Pegmatites	Pegmatitic	~610 ppm (Vertientes); ~321 ppm (Poñén); ~10.070 ppm (Coyanmahuida)	Volumetric delimitation and tonnage	[19]
E3–E5	Cu processing residues	Tailings	64.5% REE recovery in test; 0.355 kg REE/t	Tonnage associated with tested materials and continuity	[69]
E4	Coquimbo and Magallanes occurrences	Primary in rock	Relative LREE enrichment; xenotime and REE-rich epidote	Targeted sampling with sufficient density	[36,37]
E5	National tailings cadastre	Tailings	642 and 653 deposits analyzed; fourteen REE by ICP-MS	Mineralogy, granulometry, 3D continuity and test	[20,21]
E5	Drainage sediments	Sedimentary	Enrichment 4.4–5.7 times above upper crust in Los Ríos	Follow-up to source lithology and weathered profile	[33,59]
E5	Nodules and crust of EEZ	Marine	Co ~0.53%; Cu + Ni up to 1.38%	REE and Y analyses and environmental baseline	[12,83]

The conclusion that emerges from these different backgrounds is that the potential in Chile is considerably more diverse than a formal resource inventory suggests. The country has at least three independent geological systems that are capable of concentrating REE (ion-adsorption regoliths, primary IOA–IOCG–pegmatitic mineralization, and sedimentary or marine deposits); a fourth reservoir, tailings, was created by the mining industry itself. The fundamental gap is no longer just demonstrating the presence of REE but determining how much material there is, which elements the material contains, the minerals or adsorption sites where they are found, and what fraction could be economically feasible to recover.

Ordering the information by level of evidence reveals what is known about each resource but does not indicate which deserves attention first. Table 11 provides a means to judge this with a screening matrix in which every background of the inventory is scored from 1 to 5 against ten criteria, arranged along the sequence of Figure 2: evidence level (EV), tonnage confidence (TN), TREO grade (GR), proportion of Nd, Pr, Dy and Tb (MQ), recoverable fraction (RF), mineralogical simplicity (MS), low Th and U burden (RU), processing maturity (PM), infrastructure (IF) and low environmental or social constraint (EN). The criteria have been worded so that 5 always denotes the favorable condition. The scores are derived from the figures compiled in Table 2, Table 3, Table 4, Table 5, Table 6, Table 7, Table 8, Table 9 and Table 10, and they are relative to the Chilean inventory rather than absolute.

Table 11. Screening matrix of Chilean rare earth resources, in order of total scores. Scores are from 1 (unfavorable) to 5 (favorable). EV, evidence level; TN, tonnage confidence; GR, TREO grade; MQ, proportion of Nd-Pr-Dy-Tb; RF, recoverable fraction; MS, mineralogical simplicity; RU, low Th-U burden; PM, processing maturity; IF, infrastructure; EN, low environmental or social constraint.

Resource or System	EV	TN	GR	MQ	RF	MS	RU	PM	IF	EN	Total	Priority
Penco Module	5	5	4	4	3	5	5	5	5	2	43	Metallurgical optimization
Nahuelbuta corridor	2	1	4	4	4	5	5	5	4	3	37	Immediate metallurgical work
El Buitre tailings	3	4	1	1	3	3	4	3	5	5	32	By-product testing
Iron Belt (IOA-IOCG)	2	2	2	2	3	3	3	4	5	4	30	By-product testing
Cu processing residues	2	2	1	1	3	3	4	3	5	5	29	By-product testing
National tailings cadastre	1	2	1	1	2	3	4	3	5	5	27	Exploration and characterization
Biobío pegmatites	2	1	3	3	2	2	1	4	4	3	25	Exploration and characterization
Cerro Carmen	3	3	2	2	2	2	1	3	3	2	23	Exploration and characterization
Coquimbo and Magallanes	1	1	2	2	2	3	2	3	2	3	21	Exploration and characterization
Drainage sediments	1	1	1	2	1	2	3	2	3	4	20	Exploration guide
Nodules and crusts (EEZ)	1	1	1	2	1	2	3	1	1	1	14	Environmental baseline

Four groups emerge, and they do not coincide with the ranking by grade. Ranked with 37 points or more, the Penco Module and the Nahuelbuta corridor are the cases in which the next step should be metallurgical and engineering work rather than exploration: optimization of the desorbable fraction by element first and drilling together with standardized desorption tests on the already characterized profiles second. Ranked between 29 and 32 points, El Buitre, the Iron Belt and the copper processing residues share a favorable position in infrastructure and environmental terms and an unfavorable position in grade and magnet–element proportion, which is precisely the combination that justifies metallurgical testing conducted as by-product screening inside operations already in production and not as stand-alone projects. Placed between 20 and 27 points are the national tailings cadastre, the Biobío pegmatites, Cerro Carmen, the granitoids of Coquimbo and Magallanes, and the drainage sediments. These remain exploration and characterization targets since their limitation is the absence of tonnage and not the process. With under 20 points, the marine domain requires a geochemical and environmental baseline before any resource discussion can take place. The matrix also illustrates that no resource scores well on every criterion, so the question is not which is the best but rather what combination of criteria is needed to make a decision.

3. Conclusions

The potential of Chilean rare earth resources is distributed across five domains (ion-adsorption regoliths, primary rock mineralization, pegmatites and placers, tailings, and marine occurrences), but only one record is in the category of mineral resources under an international code: the 27.5 Mt Measured and Indicated of the Penco Module, at 2292 ppm TREO and 62.9 kt of contained oxides. The rest are historical estimates, non-standard models, mineralization without tonnage, and regional anomalies. Presence no longer needs to be established; rather, how much material exists and what fraction is recoverable need to be determined.

The regolith domain is the most advanced and is the primary rock domain in the inverse case. Penco doubles or triples the grades of Serra Verde, Makuutu and Koppamurra, but the metal it contains, which is close to 60 kt against 100 to 1000 kt, will sustain modular operations rather than a single large-scale exploitation, and Nahuelbuta, with up to 2000 ppm of exchangeable rare earth material, demonstrates that Penco is not a singularity. The Iron Belt offers a broad inventory without the three measurements that would convert it into a resource: the modal abundance of apatite, its grade at deposit scale and the deportment of the carrier phases.

Secondary and marine resources complete the picture. The cadastres cover between 642 and 653 tailings deposits, and El Buitre shows that they can be quantified, with 6.6 Mt at 337 mg/kg, but with light-to-heavy ratios close to 15:1 and a recovery of 64.5%, equivalent to 0.355 kg per ton treated. National assessments yield marginal profitability; therefore, the question as to whether to exploit them depends on integration into polymetallic schemes with Cu, Co, Ni or Sc. Marine occurrences still lack rare earth analyses in Chilean samples; the immediate priority is establishing a geochemical and ecological baseline.

Two conclusions cut across these figures. The numerical range does not distinguish the level of evidence: the interval between 340 and 4300 ppm contains modeled deposit grades, sediment anomalies without a delimited volume, and point values without tonnage, and only two of the five measurement scales carry a tonnage. The total content never equals value: recoveries differ between 19.53% for Nd and 43.23% for Dy, saline desorption leaves Th and U in the solid, acid leaching of copper residues has been demonstrated only at laboratory scale, and no domain has demonstrated the separation stage.

Applied to the inventory, the screening matrix separates the two resources that require metallurgical work now—the Penco Module and the Nahuelbuta corridor—from the three that justify by-product testing inside operations already in production—El Buitre, the Iron Belt and the copper processing residues—and from the remainder, which are exploration or baseline targets.

Moving from resource to industry requires four measures. The first is to generate the three missing measurements (the exchangeable fraction of the central-southern corridor, the rare earth grade and deportment of apatite in the Iron Belt, and analyses in crusts and metalliferous sediments), each of which would raise a whole domain by one level of evidence. The second is to incorporate rare earth as a by-product of the iron and copper operations already in production, the only route that makes low grades viable. The third is to install the midstream segment the country lacks, the capacity to separate a mixed carbonate into individual oxides, with a national protocol for reporting desorbable fractions, Th and U contents and code-compliant resources. The fourth is to anticipate the permitting cycle, which, for Penco, consumed two years and more than two thousand observations by characterizing the desorbable fraction and the radiological load from exploration. Chile’s position will depend less on the magnitude of its geology and more on the quality of its measurements.