Crystal Chemistry and Genetic Implications of Pink Tourmalines from Distinct Pegmatite Provinces

Abstract

Borosilicate minerals of the tourmaline supergroup are valuable both for collectors and for geological research, as their chemical composition reflects the growth-medium conditions and their evolution. Tourmalines show a wide compositional variability, with pink tourmalines being particularly prized as gemstones. This study examines the crystal chemistry of pink tourmalines from Cruzeiro (Brazil), Nuristan (Afghanistan), and Malkhan (Russia) using Electron Microprobe Analysis, Micro Laser Induced Breakdown Spectroscopy (LIBS), and Single Crystal X-ray Diffraction. The results show that the pink tourmalines are Mn-rich elbaite, with the pink coloration linked to Mn at the Y site, indicating crystallization from Mn-rich pegmatitic fluids. LIBS spectra suggest a Li-rich pegmatite origin. The samples show differences: Cruzeiro exhibits strong chemical zoning, Nuristan has a uniform composition, and Malkhan shows slight zoning with high F content. A comparison with a pink tourmaline from Anjanabonoina (Madagascar) reveals that it is Ca-rich, belonging to the calcic group and crystallizing in an open system influenced by external Ca-rich fluids, contrasting with the closed system of the samples from Cruzeiro and Nuristan. The sample from Malkhan shows an anomalous chemical variation of Ca and requires further investigation.

1. Introduction

The tourmaline supergroup minerals are complex borosilicates belonging to the cyclosilicate class, characterized by a wide chemical and structural variability. These minerals can be considered not only collectible minerals or precious gemstones but also important minerals for geological and petrogenetic research. Actually, a wide range of information can be extracted from tourmaline, such as the composition of the host rock, fluid composition, and the pressure–temperature conditions under which it formed [1]. The crystallization of tourmalines occurs generally in late-stage igneous and hydrothermal rocks as well as evolved granitic rocks, chiefly pegmatites, leucogranites, and associated veins, characterized by an enrichment of light elements, like boron and lithium. Tourmaline can crystallize also in metamorphic rocks as a recrystallized tourmaline or due to boron metasomatism [1].

The multiple occurrences of tourmaline in different geologic environments reflect a large stability range in both pressure and temperature and the occurrence of zoned crystals can be considered as petrogenetic indicators of the evolution of growth environment [2,3].

In most cases, minerals belonging to the tourmaline supergroup have an isostructural arrangement (space group R3m).

The general formula for the tourmaline supergroup can be written as follows:

XY₃Z₆(T₆O₁₈)(BO₃)₃V₃W

where the letters indicate the different crystallographic sites in which the following elements can be accommodated [1]:

^IXX = Na, Ca, K or □ (vacancy)

^VIY = Al, Li, Fe, Mg, Mn, V, Cr, Ti, etc.

^VIZ = Al, Mg, Cr, V, Fe

^IVT = Si, Al, B

^IIIB = B

V (≡ O3) = OH, O

W (≡ O1) = F, O, OH

In particular, the octahedrally-coordinated Y and Z sites can accommodate different elements, such as Fe, Mn, Ti, V, Cr, and Cu, that influence the color of the crystals. Tourmalines display a remarkable range of colors, including red, pink, yellow, orange, green, blue, violet, brown, and black. They are often chemically zoned, and a single crystal may exhibit a wide range of appealing colors [4] Among the various mineral species belonging to the tourmaline supergroup, elbaite, fluor-liddicoatite, and rossmanite, generally found in pegmatites, are distinguished by their well-defined crystal forms and economic importance in the gemological market.

Specifically, two varieties, corresponding mainly to elbaite or fluor-liddicoatite, can be considered as the most appreciated colored gemstones: the blue-neon “Paraiba” and the pinkish–red “rubellite”. The beautiful color of “Paraiba” is due to a high Cu content (several hundred ppm) [5]; while the more pink–violet hues are characterized, in addition to Cu, by the presence of Mn [6]. Conversely, the pink–red color of rubellite is due to a content of Mn, with Fe³⁺ sometimes also playing a role. The deep pink to reddish color is commonly enhanced through natural or artificial irradiation or by heat treatment above 700 °C [7,8,9]. It is worth noting that the pink color is due to the presence of Mn³⁺, while the content of Mn²⁺ + Ti⁴⁺, commonly occurring in tsilaisite and fluor-tsilaisite tourmaline [10,11], causes a greenish-yellow color in tourmaline.

In this study, a structural and chemical characterization of pink tourmalines from different mines is carried out in order to define if the pink color could be related to similar growth conditions. Representative crystals from Cruzeiro mine (Brazil), Nuristan district, Kunar (Afghanistan), and Malkhan pegmatite field (Russia) were chemically investigated using electron microprobe. The light element concentrations were detected by micro laser-induced breakdown spectroscopy and structural information was obtained through single-crystal X-ray diffraction and structural refinement.

The data acquired have been compared with the data previously acquired on the pink portion of a tourmaline from Anjanabonoina (Madagascar) [4].

2. Materials and Methods

2.1. Samples

The gem-quality pink tourmalines analyzed in this study come from three localities: the Cruzeiro mine (Brazil), the Nuristan district, Kunar (Afghanistan), and the Malkhan pegmatite field (Russia) (Figure 1).

The sample from Cruzeiro (CRUZ1) shows an outer saturated pink color and an elongated prismatic morphology deeply striated parallel to the c-axis, with a size of 2.5 cm in length and a basal section of 1.5 cm in diameter. A basal thin section, cut and polished to perform a multi-analytical investigation, shows color zoning with a colorless core and a pink rim.

The sample from Afghanistan (AF1) shows a pale pink color and an elongated prismatic morphology deeply striated parallel to the c-axis, with a length of 1.5 cm and a basal section of 0.8 cm in diameter. From this raw sample, one thin section parallel to the c-axis was cut and polished to perform a multi-analytical investigation.

The sample from Russia (RUSS1) shows a saturated pink color and an elongated prismatic morphology deeply striated parallel to the c-axis with a size of 1.2 cm in length and a basal section of about 0.7 cm in diameter. From this raw sample, two thin sections parallel and perpendicular to the c-axis were cut and polished to perform a multi-analytical investigation.

2.2. Occurrence

2.2.1. Cruzeiro Mine (Brazil), Sample CRUZ1

This mine is located near the city of Governador Valadares, Minas Gerais (Brazil), and is economically the most important mine for colored tourmaline, which is often found in gem-bearing dykes. The Cruzeiro pegmatite is a Li-, Be-, B-, Nb-, Ta-, and Zn-enriched miarolitic pegmatite located near São José da Safira, Governador Valadares, Minas Gerais, Brazil. It originated from a granitic pluton and is included in a metamorphic series of Proterozoic age known as the “Serra da Safira” schist–quartzite sequence [12,13,14]. This sequence may be described, from base to top, as layers of meta-ultramafic schists, metapelitic schists, and a coarse-grained quartzite several hundred meters thick [14]. The Cruzeiro mine dikes are rich in rare elements and are classified as complex pegmatites. The dikes, which vary in width from 8 to 60 m, display symmetrical mineralogical, textural, and chemical zoning around a quartz core [14]. The mineralogical and textural zonation comprises a border zone, wall zone, intermediate zone, pockets, and core zone [15]. Tourmaline crystals are ubiquitous but are mainly present in the wall and intermediate zones and in the miarolitic pockets that are casually disseminated around the quartz core [12,13,14]. In particular, in the border zone of the pegmatite tourmaline is typically black-schorl, and it is in paragenesis with quartz, K-feldspar, muscovite, prismatic blue beryl, almandine-rich garnet, and columbite–tantalite (Nb  >  Ta); the intermediate zones and miarolitic pockets are characterized by the presence of multicolored elbaite and Li-bearing mica, quartz, platy albite (cleavelandite), tabular pink beryl, spodumene, spessartine-rich garnet, amblygonite, Mn-apatite, tantalite–columbite (Ta  >  Nb), gahnite, and cassiterite [14].

These pockets, in particular, are characterized by the presence of alkali-deficient and vacancy-rich elbaite [3].

2.2.2. Nuristan, Kunar (Afghanistan), Sample AF1

In Afghanistan, particularly in the northeastern region, there are rare metallic granite pegmatites [16]. These pegmatites are known for their substantial reserves and combinations of rare elements and minerals, including gemstones with a significant economic value. Rare-metal pegmatites in the Nuristan and Badakhshan regions are spatially and genetically linked to the granites of the Lagman complex. The Lagman complex massifs formed in three primary phases, including the development of granodiorites, plagio-granites, and quartz diorites in the first phase, followed by the strongly porphyritic biotite granites in the second phase. The third phase involved the creation of biotite, micaceous, and sillimanite granite, which are spatially and genetically associated with rare-metal pegmatites. The dimensions of the pegmatite bodies range from 1 to 60 m in width and from 10 m up to 2–5 km in length, and they appear poorly zoned.

Among the abundant metals and minerals discovered in the pegmatites of Eastern Afghanistan, lithium-rich minerals are the most interesting ones. Lithium deposits are represented by large and numerous veins of spodumene pegmatites on the roof of granite massifs. The most productive sources of gemstones are albitized microcline pegmatites that include lepidolite, spodumene, and polychrome tourmaline [17].

2.2.3. Malkhan Pegmatite Field (Russia), Sample RUSS1

The Malkhan pegmatite field, located about 250 km southeast of Ulan-Ude and Lake Baikal, is currently the most important producing pegmatite district in Siberia. The Malkhan pegmatite district is located in the metamorphosed rocks of the Malkhanseries (amphibole to biotite schists) and the Malkhan complex–metamorphosed intrusive bodies including metagabroids (amphibolite), amphibole metadiorites, and amphibole–biotite–quartz–gneiss metagranites. These rocks show the intrusion of two Mesozoic granite plutons: the Bolsherech and Oreshnymassifs. Many of the pegmatite dikes are hosted by metamorphosed rocks of the Malkhan complex, and minor dikes are within leucogranites of the Oreshny Massif [18]. The pegmatites and granites have been dated with ages between 123.8 and 127.6 Ma [19]. The district is significant for the presence of gem-bearing pegmatite dikes, historically, the Mokhovaya, Oreshnaya, Oktyabrskaya, and Sosedka pegmatites have been among the most important producers of gem-quality tourmalines, especially the rubellite variety. In particular, the high-productive bodies are characterized by the presence of residual vugs containing quartz–lepidolite–tourmaline–albite (+/–petalite, beryl, pollucite, and Ta–Nb minerals) assemblages of variable composition. The miarolitic cavities are sources not only of tourmaline, but also of quartz, topaz, beryl, danburite, and hambergite [19].

2.3. Methods

2.3.1. Electron Microprobe Analysis (EMPA)

EMPA equipped with wavelength dispersive spectrometers (WDS) is a model Cameca SX50 located at the Istituto di Geologia Ambientale e Geoingegneria (Rome, Italy), CNR, operating at an accelerating potential of 15 kV and a sample current of 15 nA. The beam diameter is about 10 µm. The following minerals and synthetic compounds were used as standards: wollastonite (Si, Ca), magnetite (Fe), rutile (Ti), corundum (Al), vanadinite (V), fluorphlogopite (F), periclase (Mg), jadeite (Na), K-feldspar (K), rhodonite (Mn), sphalerite (Zn), and metallic Cr and Cu. The counting time was 20 s for all elements. For quantitative micro-analyses, the PAP correction procedure was used [20]. The relative error of the data was ≤5% and the detection limits were ≤0.03%. The polished slices for each tourmaline sample were investigated by performing traverses (see below).

2.3.2. Micro-Laser Induced Breakdown Spectroscopy (μ-LIBS)

The Laser-Induced Breakdown Spectroscopy technique was employed to conduct traverses on slices of the samples, with an average of five laser shots taken at each designated point. The apparatus utilized for these analyses is located within the Department of Earth and Geo-environmental Sciences at the University of Bari “Aldo Moro”. It features a double pulse Q-switched laser (Nd-YAG, wavelength: 1064 nm), which operates with a 1 microsecond delay between successive pulses, delivering an energy output of 20 millijoules per pulse. The small diameter of the laser spot, approximately 10 μm, was achieved by employing a petrographic optical microscope (OL 10 × NA 0.25 WD 14.75 mm). The spectral data obtained through LIBS were recorded using an AvaSpec Fiber Optic Spectrometer, which operates across the wavelength range of 390 to 900 nanometers (resolution ≈ 0.3 nm). The timing of the spectral acquisition was set to 2.28 microseconds following the second pulse, with an integration period of 1 millisecond. Quantitative analysis of the data was performed utilizing linear regression methods, focusing on the principal lithium emission line observed at 670.706 nanometers. This line corresponds to the resonance transition from the electronic state 1s²2s to 1s²2p, which is particularly sensitive to variations in lithium concentration [4,7,21,22]. Calibration was meticulously executed using two specific standard glasses, namely NIST610 and FLX-SLAG1, which contain lithium oxide (Li₂O) at concentrations of 0.105% and 5.0%, respectively. Furthermore, a tsilaisite mineral sample, with the chemical formula NaMn²⁺₃Al₆(Si₆O₁₈)(BO₃)₃(OH)₃, was incorporated into this calibration process [9]; this sample exhibited a Li₂O content of about 2.04% as measured by secondary ion mass spectrometry (SIMS).

2.3.3. Single-Crystal X-Ray Diffraction (SC-XRD) and Structural Refinement (SREF)

Single-crystal structure refinement was performed on four fragments extracted from each tourmaline sample. For the color-zoned sample from Cruzeiro mine (CRUZ1), one fragment was extracted from the rim and one from the core region.

The fragments were mounted on a glass fiber for X-ray diffraction acquisition. X-ray intensity data were acquired with a Bruker KAPPA APEX-II single-crystal diffractometer (Sapienza University of Rome, Earth Science Department) equipped with a charge-coupled device (CCD) area detector and a graphite crystal monochromator using a MoKα X-ray source (0.70930 Å). Crystal structure refinement was performed in SHELXL-2013 software [23]. All crystal structures were refined in the space group R3m.

3. Results

The chemical analyses were performed on thin sections of samples, while the structural analyses were conducted on fragments extracted from the slices.

3.1. Chemical Analyses

For each tourmaline sample, EPMA was performed through multiple traverses along the polished sections, following careful optical examination aimed at identifying potential complex zoning patterns. EPMA data, integrated with complementary µ-LIBS and SREF analyses (see below), enabled the accurate determination of the chemical composition of each sample.

3.1.1. EPMA Data

Optical observations on the basal slice of the Cruzeiro sample (CRUZ1) reveal a concentric color-zoning, with the rim displaying a darker pink color compared to the lighter pink core (Figure 2). To verify if the color zoning corresponds to a chemical variation, an EPMA traverse of 40 analysis points was performed along a 750 µm profile, with a step size of 19 µm, from rim (A) to core (B) (Figure 2). The results show that the saturated pink color of the rim is due to a higher content of MnO (in the form of Mn³⁺) with respect to the core, which contains a significant decrease of Mn. The Al₂O₃ content appears almost constant across the entire sample. The amount of Na₂O is approximately 1.76 wt%, while CaO is around 0.18 wt%. FeO is present in very small quantities and absent in some point analysis (Figure 2).

All Mn was measured as MnO, but unquantifiable amounts of Mn³⁺, responsible for the pink–red coloration, are present.

Likewise, the RUSS1 sample exhibits a pink color. The analyses were performed on a section cut parallel to the c-axis (AB in Figure 3). EPMA traverse shows a significant variation in some elements: CaO is present in high amounts in the upper part of the sample (3.15 wt%) and decreases notably toward the lower zone of the sample, reaching a value of 0.62 wt%.

An opposite trend is observed for Al₂O₃, which is present at approximately 39.70 wt% in the upper part of the sample, increases to 41.65 wt% in the lower part, and then decreases again to 40.18 wt%. The Na₂O content increases toward the lower part of the sample, starting at 0.95 wt% and reaching a maximum of 1.71 wt%. Notably, around analysis points 14–15, the Na₂O content exceeds that of CaO (1.71–1.64 wt% vs. 0.62–0.73 wt%, respectively). The upper part of the slice shows a discrete amount of MnO (0.56 wt%), which correlates with the observed color. The MnO content decreases slightly in the lower part to 0.10 wt%. The FeO content remains negligible throughout the sample.

Optical observations of a slice from the AF1 sample, cut parallel to the c-axis, reveal an almost uniform pink color, without an appreciable color-zoning. EPMA spot was conducted across the entire sample to investigate the variation of the main elements (Figure 4).

In accordance with the optical observation, the traverse does not reveal significant variations in the major elements. The initial analysis points (from point 20 to point 23) show a very small MnO content (0.04 wt%), which gradually increases across the sample, reaching a maximum of 0.24 wt%. The first three analysis points also show a slight increase in CaO, ranging from a maximum of 0.17 wt% to a minimum of 0.05 wt%, while Na₂O remains relatively constant with an average value of 1.85 wt%. The FeO content is negligible throughout the sample and absent in some analysis points. The Al₂O₃ content remains stable at approximately 41.17 wt%.

3.1.2. µ-LIBS

EPMA data were integrated with LIBS point analyses to determine the Li₂O content in each sample. Focusing on the main lithium emission line intensity at 670.706 nm acquired on CRUZ1, a slightly variable Li content is observed from core to rim (Figure 5). The Li₂O content was quantified with a value ranging from an average of 2.16 wt% on the rim to 2.27 wt% at the core. Analyses of the AF1 and RUSS1 samples show nearly homogeneity of Li₂O variation throughout the sample, with values of 2.20 (±0.15)% and 2.02 (±0.14)%, respectively.

3.2. Crystal Structure Refinement

Crystal structure refinement was performed on four fragments extracted from the samples. For the AF1 sample, only one fragment was extracted as the sample appears homogeneous. For the RUSS1 sample, the fragment was extracted from the lower zone of the sample rich in Na. Conversely, for the CRUZ1 sample, the crystal structure refinement was performed on two fragments: one from the rim and one for the core region.

Structural refinements were performed with the SHELXL-2013 program [23]. The initial coordinates are the same used by Bosi et al. [7]. The variable parameters were scale factor, extinction coefficient, atom coordinates, site scattering values, and atom displacement factors. Based on crystallographic and mineral chemical information, the X site was modeled with Na, except for sample RUSS1, where it was modeled with Na (fixed to the value of the empirical formula; see below) vs. Ca. The Y site was modeled with Li and Al, whereas the Z, T, B, and O2–O8 anion sites were modeled, respectively, with Al, Si, B, and O scattering factors and with a fixed occupancy of 1, because refinement with unconstrained occupancies showed no significant deviations from this value. The O1 site was successfully modeled with O vs. F fixed to the value obtained from the empirical formula. The position of the H atom bonded to the oxygen at the O1 and O3 sites in the structure was taken from the Fourier difference map and incorporated into the refinement model; the O1–H1 and O3–H3 bond lengths were restrained (by DFIX command) to be 0.97 Å, with the isotropic displacement parameter constrained to be equal to 1.2 times that obtained for the O1 and O3 sites. It should be noted that very weak hydrogen bonds may occur in the tourmaline structure, involving H1···O4 and O5, as well as H3···O5 interactions. Three full matrix refinement cycles with isotropic displacement parameters for all atoms were followed by anisotropic cycles until convergence was attained.

Table 1. Single-crystal X-ray diffraction data for the tourmalines studied.

	CRUZ1 Core	CRUZ1 Rim	RUSS1	AF1
Crystal data
Size (mm)	0.24 × 0.28 × 0.32	0.20 × 0.25 × 0.40	0.04 × 0.40 × 0.40	0.26 × 0.30 × 0.30
Space group; Z	R3m, 3
a (Å)	15.84606(19)	15.8344(3)	15.8279(2)	15.8317(2)
c (Å)	7.10555(9)	7.1005(2)	7.1014(1)	7.1034(1)
V (Å3)	1545.15(4)	1541.78(7)	1540.71(4)	1541.88(4)
Data collection
Data collection temperature (K)	293
Range for data collection, 2θ (°)	5–75
Radiation, wavelength (Å)	MoKα, 0.71073
Reciprocal space range, hkl	−25 ≤ h ≤ 21	−26 ≤ h ≤ 26	−23 ≤ h ≤ 27	−26 ≤ h ≤ 26
	−26 ≤ k ≤ 26	−27 ≤ k ≤ 26	−26 ≤ k ≤ 26	−23 ≤ k ≤ 26
	−11 ≤ l ≤ 12	−9 ≤ l ≤ 11	−8 ≤ l ≤ 9	−10 ≤ l ≤ 12
Measured reflections	11,595	11,502	11,573	11,543
Unique reflections, Rint	1846, 0.0211	1760, 0.0157	1663, 0.0205	1811, 0.0233
Redundancy	12
Absorption correction method	Multi-scan (SADABS)
Refinement
Refinement method	Full-matrix last-squares on F2
Structural refinement program	SHELXL-2013
Extinction coefficient	0.0073(4)	0.0040(3)	0.0051(3)	0.0014(3)
Flack parameter	0.02(10)	0.09(8)	0.07(8)	0.01(10)
wR2	0.0401	0.0358	0.0334	0.0403
R1 all data	0.0159	0.0135	0.0138	0.0160
R1 for I > 2σ(I)	0.0158	0.0134	0.0137	0.0158
GooF	1.124	1.129	1.148	1.174

and

Table 2. Average bond lengths (Å) for the tourmalines studied.

	CRUZ1 Rim	CRUZ1 Core	RUSS1	AF1
< X–O >	2.664	2.662	2.650	2.664
< Y–O >	2.006	2.011	2.012	2.008
< Z–O >	1.905	1.906	1.905	1.905
< B–O >	1.375	1.375	1.375	1.374
< T–O >	1.617	1.618	1.616	1.616

show single-crystal X-ray diffraction data details for the analyzed samples.

The crystallographic information (CIF) files showing all structural data are available as Supplementary Materials.

3.3. Determination of Site Population and Formulae

The boron content was assumed to be stoichiometric (B³⁺ = 3.00 apfu), in accordance with the crystallographic model elaborated by Bosi et al. (2021) [7]. Both the site scattering results and the bond lengths of B and T are consistent with the B sites being entirely occupied by boron. B³⁺ at the T site was not found. µ-LIBS analyses were used to determine the Li amounts. The OH content and the atoms per formula unit (apfu) were then calculated by charge balance, assuming a total of 31 anions and that ^V+W(OH + F + O) = 4. To simplify the formula calculations, Fe and Mn were considered as divalent ions. The oxidation state changes of transition elements span a few hundred ppm that cause a distinct color change but cannot be quantified through structural or chemical analyses [24]. Therefore, even if the Mn³⁺ represents the chromophore element for these samples, in the determination of the formulae, the Mn oxidation state was assumed only as +2.

These assumptions are supported by the reasonable accord for all crystals, of the number of electrons per formula unit (epfu) calculated by site-scattering refinement and electron-microprobe analysis.

The chemical composition of each tourmaline sample, calculated from the average of multiple analysis points using EPMA, µ-LIBS data, and SREF information (e.g., B = 3.00 atoms per formula unit, apfu), is reported in

Table 3. Chemical composition, based on EPMA, µ-LIBS, and SREF data, for the tourmalines studied.

	Average of 4 Spots (wt%)	Average of 36 Spots (wt%)	Average of 10 Spots (wt%)	Average of 10 Spots (wt%)
	CRUZ1 rim	CRUZ1 Core	RUSS1	AF1
SiO2	39.07(21)	39.06(29)	39.09(38)	39.18(17)
TiO2	0.02	0.01(2)	0.01	0.00
B2O3 (a)	11.52	11.20	11.05	11.20
Al2O3	40.72	40.56	40.67(41)	41.17(10)
FeOtot	0.03	0.02	0.01(1)	0.03(2)
MnOtot (b) tot	0.63(39)	0.12(4)	0.23(10)	0.18(7)
CaO	0.12	0.25(6)	1.36(42)	0.09(4)
Na2O	1.74(1)	1.77(5)	1.51(13)	1.85(4)
Li2O (c)	2.16(2)	2.27(2)	2.02(14)	2.20(15)
F	0.56(2)	0.66(12)	1.01(16)	0.58(12)
H2O (a)	3.45	3.63	2.87	3.59
O = F	−0.236	−0.28	−0.42	−0.29
Total	99.84	99.30	99.43	99.87
Atoms normalized to 31 anions
Si (apfu)	6.08	6.09	6.15	6.09
Ti4+	0.00	0.00	0.00	0.00
B	3.00	3.00	3.00	3.00
Al	7.47	7.46	7.54	7.54
Fe2+	0.01	0.00	0.00	0.00
Mn2+	0.08	0.02	0.03	0.02
Ca	0.02	0.04	0.23	0.02
Na	0.53	0.54	0.46	0.56
Li	1.35	1.42	1.28	1.37
F	0.28	0.33	0.50	0.28
OH	3.58	3.67	3.01	3.72

^(a) Calculated by stoichiometry and SREF information. ^(b) All Mn was measured as MnO, but unquantifiable amounts of Mn³⁺, responsible for the pink–red coloration, are present. ^(c) Determined by µ-LIBS. Errors for oxides and fluorine are standard deviations (in brackets).

.

In detail, the tourmalines’ allocation procedure follows the standard site preference suggested by Henry et al. [25]. Na, Ca, and K are assigned to the X site with any site deficiency assumed to represent X-site vacancy (^X□). Silicon is assumed to be exclusively located at the T site [26], in accordance with the observed mean bond length <T-O> (Table 2) that ranges between 1.616–1.618 Å consistent with the expected value for <^TSi-O> [27]. Similarly, the <B-O> bond lengths, ranging from 1.374 to 1.375 Å, support full occupancy of the B site by boron [27]. The anion distribution between the V and W positions is well established: F exclusively to W, while O tends to preferentially occupy it [28], with any surplus O being allocated to V. Evidence from bond angle distortion of the ZO₆ octahedron and Y-O distances, along with bond valence sums at the O3 site, suggests that most tourmaline species hold approximately ^V(OH)₃ [1,27,29,30]. The Y and Z site populations can be more difficult: typically Li, Mn, Fe, and Al occupy the Y site. In all tourmaline samples in this work, the average distance <Z-O> varies between 1.905–1.907Å, consistent with the site exclusively occupied by Al [31].

Following this standard site preference suggested for tourmaline, the resulting empirical formulae for each sample are as follows:

CRUZ1 rim

^X(Na_0.53□_0.45Ca_0.02)_Σ1.00^Y(Al_1.47Li_1.35Mn_0.08Fe_0.01)_Σ2.91^ZAl₆[^TSi_6.08O₁₈](BO₃)₃^V(OH)₃^W[F_0.28(OH)_0.58O_0.14]_Σ1.00

CRUZ1 core

^X(Na_0.54□_0.42Ca_0.04)_Σ1.00^Y(Al_1.46Li_1.42Mn_0.02)_Σ2.90^ZAl₆[^TSi_6.09O₁₈](BO₃)₃^V(OH)₃^W[F_0.33(OH)_0.67]_Σ1.00

AF1 sample

^X(Na_0.56□_0.42Ca_0.02)_Σ1.00^Y(Al_1.54Li_1.37Mn_0.02)_Σ2.93^ZAl₆[^TSi_6.09O₁₈](BO₃)₃^V(OH)₃^W[F_0.28(OH)_0.72]_Σ1.00

RUSS1

^X(Na_0.46Ca_0.23□_0.31)_Σ1.00^Y(Al_1.54Li_1.28Mn_0.03)_Σ3.00^ZAl₆[^TSi_6.15O₁₈](BO₃)₃^V(OH)₃^W[F_0.50(OH)_0.01O_0.49]_Σ1.00

4. Discussion and Conclusions

The detailed structural and chemical analyses performed on pink tourmalines from Cruzeiro mine (Brazil), Nuristan district, Kunar (Afghanistan), and Malkhan pegmatite field (Russia) show that all the samples investigated are Mn-rich elbaite species, ideally Na(Li_1.5Al_1.5)Al₆(Si₆O₁₈)(BO₃)₃(OH)₃OH. All the studied samples are characterized by ^XNa > ^XCa, with (Li + Al) dominant in the Y position. For each sample, the pink color and its nuance are intimately related to the amount of ^YMn [32], suggesting that, in all cases, the pegmatitic fluid had relatively high Mn concentrations. The LIBS spectra confirm their provenance from Li-rich pegmatites if compared with spectra acquired on tourmalines from different deposits (e.g., Li-poor pegmatites and silicic igneous rocks, calcareous metamorphic rocks, pelitic metamorphic rocks, and hydrothermal-related deposits) as found by McMillan et al. [33].

However, although these samples are quite similar, they differ in some aspects that can be associated with their distinct provenances. To further characterize the pink tourmalines from different pegmatite provinces, the data obtained in this study were also compared with those previously collected from the pink region of multicolored tourmaline from Anjanaboniona (Madagascar), another important source of multicolor and giant crystal tourmalines [4].

The ternary diagram shown in Figure 6 illustrates the compositional differences among the samples.

This diagram, based on the dominant occupancy of the X site, shows that the samples from Cruzeiro (CRUZ1) mine and Afghanistan (AF1) belong to the alkali group. The RUSS1 sample also belongs to the same group, although it is richer in Ca. In contrast, the pink tourmaline from Anjanabonoina (M1CR) falls under the Calcic group, having ^XCa > ^XNa, ^WF > ^who, and ^Y(2Li + Al) dominant. These chemical features correspond to Mn-bearing fluor-liddicoatite Ca(Li₂Al)Al₆(Si₆O₁₈)(BO₃)₃(OH)₃F [4].

In particular, CRUZ1 and AF1 appear quite similar on the basis of the occupancy of X site, even if they show some important differences. CRUZ1 shows an evident chemical zoning reflecting the evolution of the growth medium. In tourmaline, the chemical composition variation from core to rim represents an excellent example of a petrogenetic indicator of the growth environment [34,35,36]. In the case of the tourmalines from Cruzeiro mine, several studies have been carried out on these famous pegmatites [3,12,13,14]. These studies have shown that the contacts between the pegmatite bodies and the host quartzite appear well-defined, and several pieces of evidence support the hypothesis of a pegmatitic melt–fluid system crystallized virtually without any chemical interaction with the surrounding host rock. Specifically, very recently, Andreozzi et al. [3] studied a multi-colored, concentrically zoned tourmaline crystal from the intermediate zone of Cruzeiro dikes that exhibits chemical composition variations, ranging from an F-bearing schorl core to an OH-bearing fluor-elbaite rim. The authors hypothesized that the observed chemical zoning reflects the chemical evolution of the pegmatite bodies, corresponding to decreasing temperature and increasing fractionation of the melt. In CRUZ1, the color zoning corresponding to a chemical zoning consists of a nearly colorless core crystallized from an early-stage magma and a dark pink rim formed in the late stage. The covariation of cation distribution at the tourmaline Y-site indicates a progressive enrichment of Mn and a depletion of Li from core to rim. Likewise, the decrease in F and increase in ^WO could be due to crystallization under water-rich conditions during the late stage of pegmatite solidification [3,32].

In contrast, the sample from Afghanistan (AF1) exhibits no zoning. Instead, it is characterized by a nearly uniform pink hue and a corresponding uniformity in chemical composition. This feature suggests that the crystallization of the tourmaline occurred during a stage of a relatively low degree of fractionation within the pegmatite.

RUSS1 shows an intermediate situation with slight chemical zoning. Even if its Ca content is lower than that of M1CR, this tourmaline shows a high content of F, similar to the tourmaline from Anjanabonoina. It is worth noting that the sample from Anjanabonoina also exhibits a coexistence of Mn-rich fluor-liddicoatite and Mn-rich fluor-elbaite in the same crystal. The valuable enrichment of Ca, observed in the pink tourmaline from Anjanabonoina, characterizes the transition from these two different mineralogical species and can be related to the access of the Ca-rich external fluids coming from the marble unit that forms the host rocks [4].

The observed Ca-enrichment in the RUSS1 sample, with respect to the other pink tourmalines, could be related to the origin of the Malkhan granite–pegmatite system. This system consists of an injection of chemically heterogeneous magma with the independent fractionation of each magmatic body after its emplacement [37]. In particular, in the Malkhan granite–pegmatite system, Zagorsky et al., in a previous study, described the presence of some miarolitic cavities characterized by a high content of Ca also evidenced by the presence of danburite (CaB₂Si₂O₈) [37]. Peretyazhko et al. hypothesized that the heterogeneity of magma found in the Malkhan granite–pegmatite system could be related to the presence of aqueous fluids which caused an increase in fluid pressure in pockets with subsequent fracturing of pockets and heterogenization and mixing of pocket fluids [38]. The zoning characterizing the RUSS1 sample and its Ca- and F- enrichment could be due not only to decreasing temperature and increasing fractionation of the melt during the evolution of fluid during the pegmatite crystallization, but could also be ascribed to the contamination with external fluids of different composition during pocket fracturing and fluid mixing.

Based on the crystal-chemical data obtained from the pink tourmalines from Cruzeiro, Nuristan, and Malkhan, and in comparison with the previous data from a pink tourmaline from Anjanabonoina, it can be concluded that these samples crystallized under different conditions. The tourmaline from Anjanabonoina records crystallization in an open system influenced by external Ca-rich fluids, whereas the pink tourmalines from Cruzeiro and Nuristan suggest crystallization in a closed system. The sample RUSS1 from Malkhan shows evidence of an intermediate scenario and requires further investigation.