Introduction

Minerals of the tourmaline supergroup are cyclosilicates with essential B. They are the most common and the earliest boron minerals formed on Earth and are widespread in the Earth’s crust, occurring typically in granites and granitic pegmatites but also sedimentary and metamorphic rocks (Henry and Dutrow 1992, 1996; Grew et al. 2016). In accordance with Henry et al. (2011), the general formula of tourmaline, XY3Z6T6O18(BO3)3V3W, represents many constituents: X = Na+, K+, Ca2+, □ (= vacancy); Y = Al3+, Fe3+, Cr3+, V3+, Mg2+, Fe2+, Mn2+, Li+; Z = Al3+, Fe3+, Cr3+, V3+, Mg2+, Fe2+; T = Si4+, Al3+, B3+; B = B3+; V = (OH)–, O2–; W = (OH)–, F–, O2–. Note that the not italicized letters X, Y, Z, and B represent groups of cations at the [9]X, [6]Y, [6]Z, [4]T, and [3]B crystallographic sites (letters italicized). The letters V and W represent groups of anions accommodated at the [3]-coordinated O3 and O1 crystallographic sites, respectively.

Minerals of the tourmaline supergroup are commonly used for understanding crustal evolution (e.g., Slack and Trumbull 2011). Because tourmaline has wide chemical and physical stability fields and its composition generally reflects the local environment of formation, it is of major petrogenetic interest (e.g., Dutrow and Henry 2011; Lussier et al. 2008, 2011; Hinsberg and Schumacher 2011; van Hinsberg et al. 2011; Altieri et al. 2022, 2023). In this regard, it is of interest to perform high-temperature experiments on well-characterized tourmaline single crystals containing important elements, such as Mg, Al, Mn, and Fe, to reveal their thermally induced crystal-chemical behavior over the Y and Z sites, changes in the oxidation states, and color variations. Recently, Fe3+-bearing schorl, oxy-dravite, Fe-free dravite, Fe-rich fluor-elbaite, fluor-tsilaisite, and Mn-bearing elbaite were thermally treated in air or hydrogen atmospheres (Afonina et al. 1993; Pieczka and Kraczka 2004; Bačίk et al. 2011; Filip et al. 2012; Ertl et al. 2012; Bosi et al. 2016a, b, 2019; Liu et al. 2019). However, no similar study has addressed Fe-rich tourmalines with dominant Mn. The aim of this study is to understand the crystal-chemical and color variations in a princivalleite, ideally Na(Mn2Al)Al6(Si6O18)(BO3)3(OH)3O, as a function of temperature under oxidizing conditions. The Fe2+-rich princivalleite crystal from princivalleite holotype material from Veddasca Valley, Luino area, Varese, Lombardy, Italy (Bosi et al. 2022) was used for the present study. Untreated and treated samples were structurally and chemically characterized by X-ray single-crystal diffraction, electron microprobe analysis, micro-laser induced breakdown spectroscopy, Mössbauer, infrared, and optical absorption spectroscopy.

Analytical methods

Thermal treatment

A crystal fragment of princivalleite was cut parallel to the c-axis and doubly polished to a thickness of 270 μm. Heat treatments were performed at 700 °C in air, with the sample placed in a gold container and inserted into a pre-heated horizontal tube furnace equipped with a quartz-glass tube. The experiments were performed stepwise, with heating times of 0.33, 1.7, 5, 21, and 86 h, and were ended by pushing the sample container out to the cold zone of the quartz tube, which caused the sample to cool to below 100 °C within 1 min. Spectroscopic data (FTIR and OAS) were acquired on the untreated sample and after each heating experiment. After the 21-h heating experiment, some of the bands in the OAS spectrum showed too high intensities, and the sample was further thinned to a thickness of 117 μm. Throughout this work, the untreated and heat-treated princivalleite samples are indicated as Upva and Tpva (respectively).

Microprobe analysis

Electron microprobe analysis was obtained using a wavelength dispersive spectrometer (WDS mode) with a Cameca SX50 instrument at the “Istituto di Geologia Ambientale e Geoingegneria (Rome, Italy), CNR”, operating at an accelerating potential of 15 kV, a sample current of 15 nA and 10 μm beam diameter. Minerals and synthetic compounds were used as standards: wollastonite (Si and Ca), magnetite (Fe), rutile (Ti), corundum (Al), vanadinite (V) fluorophlogopite (F), periclase (Mg), jadeite (Na), orthoclase (K), sphalerite (Zn), rhodonite (Mn), and metallic Cr and Cu. The PAP routine was applied (Pouchou and Pichoir 1991). The composition (mean of seven spot analyses) is given in Table 1. Titanium, V, Cr, and K were below detection limits (< 0.03 wt.%).

Table 1 Chemical data (seven spots) for the princivalleite untreated (Upva) and treated (Tpva)
Full size table

Micro-laser-induced breakdown spectroscopy

The micro-Laser-Induced Breakdown Spectroscopy (µ-LIBS) instrument used for Li measurement integrates a dual-pulse Q-Switched laser (Nd-YAG, fundamental λ = 1064 nm), which emits two collinear laser pulses with energy of 110 mJ per pulse at 10 Hz, and the reciprocal delay was set at 1 μs.

The coupling of the laser with a petrographic microscope with a 10×  objective lens (NA = 0.25 and WD = 14.75 mm) allows to obtain a very small spot size (7–10 μm). The μ-LIBS spectra were acquired using an AvaSpec Fiber Optic Spectrometer (390–900 nm with 0.3 nm resolution) with a delay of 2 μs after the second pulse and were integrated for 1 ms. Quantitative data were obtained by generating a linear regression using the main Li emission line intensity (670.706 nm corresponding to resonance transition 1s2 2s > 1s2 2p) particularly sensitive to Li amounts. The linear fit was made according to Bosi et al. (2021) and revealed amounts of Li2O = 0.11(4) wt.% (Table 1) in line with that estimated by Pesquera et al.’s (2016) method (0.08 Li2O wt.%).

Single-crystal structure refinement

Crystal fragments of samples Upva and Tpva (700 °C) were selected for X-ray diffraction measurements on a Bruker KAPPA APEX-II single-crystal diffractometer (at Sapienza University of Rome, Earth Sciences Department), equipped with a CCD area detector (6.2 × 6.2 cm2 active detection area, 512 × 512 pixels) and a graphite-crystal monochromator, using MoKα radiation from a fine-focus sealed X-ray tube. The sample-to-detector distance was 4 cm. A total of 1621 exposures were measured (step = 0.4°, time/step = 20 s) covering the full reciprocal sphere with a redundancy up to 12 and a completeness of about 98%. Final unit-cell parameters were refined with the Bruker AXS SAINT program using reflections with I > 10 σI in the range 5° < 2θ < 75°. The intensity data were processed and corrected for Lorentz, polarization, and background effects with the APEX2 software program of Bruker AXS. The data were corrected for absorption using the multi-scan method (SADABS), leading to a significant improvement in Rint. No violations of R3m symmetry were noted.

Structure refinement was done with the SHELXL-2013 program (Sheldrick 2013). Starting coordinates were taken from Bosi et al. (2022). Variable parameters were: scale factor, extinction coefficient, atom coordinates, site-scattering values, and atom-displacement factors. In detail, the X site was modeled by considering the presence of the Na scattering factor; the Y site was modeled considering Al–Mn for Upva and Al–Fe for Tpva; the occupancy of the Z site was modeled considering Al and Fe. The T, B, and anion sites were modeled, respectively, with 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 position of the H atom bonded to the oxygen at the O3 site in the structure was taken from the difference-Fourier map and incorporated into the refinement model; the O3–H3-bond length was 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 O3 site. Three full-matrix refinement cycles with isotropic-displacement parameters for all atoms were followed by anisotropic cycles until convergence was attained. There were no significant correlations over 0.7 between the parameters at the end of refinement. Table 2 lists crystal data, data-collection information, and refinement details; Table 3 gives the fractional atom coordinates and site occupancies; Table 4 gives selected bond distances. The crystallographic information file has been deposited with the Principal Editor of Physics and Chemistry of Minerals and is available as Supplementary material (see below).

Table 2 Single-crystal X-ray diffraction data details for princivalleite untreated (Upva) and treated (Tpva)
Full size table
Table 3 Fractional atom coordinates, isotropic (*) or equivalent-isotropic-displacement parameters (in Å2), and site occupancies (s.o.) for untreated (Upva) and treated (Tpva) princivalleite
Full size table
Table 4 Selected bond lengths (Å) for untreated (Upva) and treated (Tpva) princivalleite
Full size table

Mössbauer spectroscopy

After the final thermal treatment, sample Tpva was analyzed by Mössbauer spectroscopy (MS) using a conventional spectrometer system equipped with a 10 mCi point source and operated in constant acceleration mode. The absorber was prepared by grinding small amounts of sample material together with a thermoplastic resin, which were then shaped into a 1-mm sized cylinder under mild heating (ca. 100 °C). The spectrum was collected at room temperature over the velocity range – 4.2 to + 4.2 mm−1 and was calibrated against α-Fe foil before folding and spectral fitting with the software MossA (Prescher et al. 2012).

Infrared spectroscopy

Fourier-transform infrared (FTIR) spectroscopy was used to characterize the (OH) vibrational absorption bands of the sample before and after each thermal treatment. The spectrometer system consisted of a Bruker Vertex 70 spectrometer equipped with a halogen-lamp source and a CaF2 beam-splitter coupled to a Hyperion 2000 microscope with a ZnSe wire-grid polarizer and an InSb detector. Polarized spectra were collected parallel and perpendicular to the c-axis over the wavenumber range 2000–12000 cm−1 with a resolution of 4 cm−1. The measuring areas were masked by a ~ 100 μm rectangular aperture to avoid cracks and inclusions.

Optical absorption spectroscopy

Polarized optical absorption spectra were measured at room temperature on the same doubly polished single-crystal sections used for the FTIR measurements. Spectra in the range 270–1000 nm (37,037–10,000 cm–1) were recorded at a spectral resolution of 1 nm using an AVASPEC-ULS2048X16 spectrometer (at the Swedish Museum of Natural History) connected with a 400 µm UV fiber cable to a Zeiss Axiotron UV-microscope. A 75 W Xenon arc lamp was used as a light source, and Zeiss Ultrafluar 10× lenses served as objective and condenser. The diameter of the circular aperture was 40 μm. A UV-quality Glan–Thompson prism with a working range from 250 to 2700 nm (40,000 to 3704 cm−1) was used as a polarizer. The wavelength scale of the spectrometers was calibrated against Ho2O3-doped and Pr2O3/Nd2O3-doped standards (Hellma glass filters 666F1 and 666F7). Spectral data in the range 10,000–5000 cm–1 (1000–2000 nm) were taken from the FTIR measurements. Recorded spectra were fitted using the Jandel PeakFit 4.12 software, assuming Gaussian peak shapes.

Results and discussion

The color of the untreated crystal is very pale blue (azure). During the steps of heat treatment, it darkens, and the color changes via shades of gray to very dark brown (Fig. 1).

Fig. 1
figure 1

Images of the same Fe-rich princivalleite crystal untreated (Upva) and heat-heated (700 °C) for different times (1.7 h, 5 h and 21 h) showing color variation from faint blue (azure) to gray to dark brown. The polished plate is approximately 1 mm in length and 270 μm thick

Full size image

Mössbauer spectra: oxidation state of Fe

The spectrum of the heat-treated sample (Tpva) was fitted using Lorentzian doublets. In agreement with Andreozzi et al. (2008), a model with three doublets was adopted (Fig. 2). The spectrum was fitted with two doublets assigned to Fe3+ and one doublet assigned to Fe2+, confirming the presence of both Fe2+ and Fe3+ in the sample Tpva. The relatively broad absorption related to Fe3+ indicates that Fe3+ occurs at both the Y and Z sites. However, a unique site distribution cannot be achieved due to lack of resolution. The resulting hyperfine parameters and absorption intensity areas are shown in Table 5. The spectrum of the heat-treated sample, compared to the spectrum of the untreated sample (see Fig. 2 in Bosi et al. 2022), shows a strongly increased absorption in the velocity range, typical for Fe3+. The results from the fitting procedure indicate the Fe3+/ΣFe ratio of 92% for the treated sample, whereas no indication of absorption due to Fe3+ was observed for untreated princivalleite (Bosi et al. 2022).

Fig. 2
figure 2

Mössbauer spectrum obtained at room temperature for the fully heat-treated princivalleite sample (Tpva). Fitted absorption doublets assigned to Fe2+ and Fe3+ are indicated in blue and red, respectively. Diamonds denotes measured spectra, and black curve represents summed fitted spectra

Full size image
Table 5 Mössbauer parameters for treated princivalleite sample (Tpva) obtained at room temperature
Full size table

Optical spectra: oxidation state of Mn

The recorded spectra show two broad absorption bands at 13,500 and 8900 cm−1. In the spectra of the untreated sample, these bands are moderately polarized in E⊥c (Bosi et al. 2022). In addition to these two strong bands, there are weak and relatively narrow bands at ∼18,000, ∼22,500, ∼24,000, and ∼27,500 cm–1, which are assigned to spin-forbidden electronic transitions in [6]-coordinated Mn2+ (Bosi et al. 2022).

During continued heat treatment up to 5 h, the 13,500 and 8900 cm−1 bands in spectra become progressively more intense and stronger polarized (E⊥c >  > E||c). However, at 21 h of heat treatment, a decrease in their intensities is observed (Fig. 3). The lower intensity and the moderate polarization of these two absorption bands in spectra of the untreated princivalleite crystal are explained by its very low or non-existing Fe3+ concentration, and consequently, these two bands mark pure, or nearly so, Fe2+ d–d transitions in spectra of the untreated sample. Intensity increases, and a higher degree of E⊥c-polarization of the two bands during continued heat treatment of the sample is explained by progressive oxidation of Fe2+ to Fe3+ and consequently increasing Fe3+-enhancement of band intensities in E⊥c. In agreement with previous optical studies of tourmaline (e.g., Mattson and Rossman 1984, 1987), the absorption bands at 13,500 and 8900 cm−1 in the heat-treated sample are assigned to Fe3+-enhanced spin-allowed d–d transitions in six-coordinated Fe2+. The intensity loss for the 13,500 and 8900 cm−1 bands recorded after 5 h of heat treatment is explained by continued oxidation of Fe2+, which results in a decrease in [Fe2+][Fe3+] concentration product when [Fe3+] >  > [Fe2+].

Fig. 3
figure 3

Optical absorption spectra polarized in E⊥c of untreated and stepwise heat-treated (700 °C) princivalleite. The main absorption bands are indicated

Full size image

After 5 h of heat treatment, an additional broad absorption band is observed at 17,900 cm−1, and this band gain in intensity after continued heat treatment. In accordance with band assignments of spectra recorded on Mn-rich tourmaline samples (Reinitz and Rossman 1988; Taran et al. 1993; Ertl et al. 2005; Novák et al. 2013; Bosi et al. 2021), this band is ascribed to a Mn3+ spin-allowed transition.

The observed intensity increases of the UV-edge absorption marking O–M3+ charge transfer processes, in combination with a strong initial increase in intensity of the 13,500 and 8900 cm−1 bands in E⊥c-spectra followed by a subsequent decrease in their intensities during further heat treatment reflects a progressive Fe2+ to Fe3+ oxidation of the sample. The occurrence of a Mn3+ absorption band after several hours of heat treatment demonstrates that oxidation of Mn2+ starts subsequent to the initial oxidation of Fe2+ to Fe3+. The Mn2O3 content in the princivalleite sample subject to different heating times was calculated from the intensity of the Mn3+ band at 17,900 cm−1 in tourmaline spectra using the molar absorption coefficient (e) proposed by Bosi et al. (2021). Values of 0.27, 0.73, and 1.49 wt% were estimated in the sample subjected to 1.7, 5, and 21 h of heat treatment, respectively (Table 6).

Table 6 Oxidation of MnO in the princivalleite sample at different heating times
Full size table

Due to an extremely strong UV-edge absorption in the spectrum of the sample after 86 h treatment, it was not possible to record the Mn3+ band at 17,900 cm−1, but its Mn2O3content is expected to be > 1.5 wt%. It is interesting to point out that the oxidation of MnO follows a hyperbolic behavior, with an estimated maximum of Mn2O3 content of ⁓2.3 wt%, corresponding to a Mn3+/ΣMn ratio of ⁓0.4.

Site populations

In agreement with the structure-refinement results, the boron content was assumed to be stoichiometric (B3+ = 3.00 apfu). Both the site-scattering results and the bond lengths of B and T are consistent with the B site being fully occupied by boron and no amount of B3+ at the T site (e.g., Bosi and Lucchesi 2007). Lithium was determined by μ-LIBS for sample Upva. The iron oxidation state was determined by Mössbauer spectroscopy. In accordance with the optical absorption spectroscopy results mentioned above, Mn is considered as Mn2+ for sample Upva, whereas it cannot be considered only Mn2+ for sample Tpva. Consequently, the Mn2O3 contents were fixed to 1.49 wt% for the treated sample based on bond-valence analysis and the direct correlation between bond-valence sum (BVS) at the O3 site, BVS(O3), and O3O contents (Bosi et al. 2018). The latter, in fact, suggests that Mn2O3 amount >  > 1.5 wt% would produce Mn3+ apfu values that deviate too much from the expected trend (see Fig. 4). The (OH) content and the atoms per formula unit (apfu) were then calculated by charge balance with the assumption (T + Y + Z) = 15 apfu and 31 anions. Note that we assume that no compositional change occurred in the treated sample with respect to the untreated princivalleite, except for the contents of FeO, Fe2O3, MnO, and Mn2O3, which changed in accord with the Fe3+/ΣFe and Mn3+/ΣMn ratios (see above) and for the contents of O2– and (OH)– needed to maintain charge balance. The very good agreement between the number of electrons per formula unit (epfu) derived from chemical data and X-ray single-crystal structure refinement (SREF), respectively, 224.4 and 224.7 epfu for Upva, and 223.0 and 224.3 epfu for Tpva, supports the stoichiometric assumptions.

Fig. 4
figure 4

Plot of bond-valence sum (BVS) incidents at the O3 site versus oxygen content at the O3 site (after Bosi et al. 2018). Solid line is the linear regression showing the empirical trend, whereas dashed line is the expected trend. Plot obtained using 11 data sets accompanied by SREF. Black circles = samples from Bosi et al. (2018), including sample BLSH2 (Ertl et al. 2012). Blue circles = samples from this study

Full size image

The site populations were optimized on the procedure described in Bosi et al. (2017). The robustness of this optimized cation distribution was successively confirmed by the procedure of Wright et al. (2000). The final structural formulae for the studied samples are as follows.

  • Sample Upva

X(Na0.54Ca0.13□0.33)Σ1.00 Y(Mn2+0.73Fe2+0.37Al1.78Li0.07Zn0.03Mg0.02)Σ3.00 Z(Al5.85Fe2+0.15)Σ6.00 [T(Si5.67Al0.37)Σ6.00O18] (BO3)3 O3(OH2.71O0.29)Σ3.00 O1[O0.69F0.21(OH)0.10]Σ1.00

  • Sample Tpva (700 °C)

X(Na0.54Ca0.13□0.33)Σ1.00 Y(Mn2+0.40Fe2+0.04Fe3+0.33Mn3+0.19Al1.93Li0.07Zn0.03Mg0.02)Σ3.00 Z(Al5.70Fe3+0.15 Mn2+0.15)Σ6.00 [T(Si5.63Al0.37)Σ6.00O18] (BO3)3 O3[(OH)2.14O0.86]Σ3.00 O1(O0.79F0.21)Σ1.00.

A comparison between the values of refined site-scatterings and mean bond lengths and those calculated from this site population is reported in Table 7. The agreement between the refined and calculated values is very good, and validates the distribution of cations over the Y and Z sites in the empirical structural formula of samples Upva and Tpva. This site population is also supported by the comparison of weighted BVS and mean formal charge (weighted atomic valence) calculated from the empirical structural formula (Table 8).

Table 7 Optimized Y and Z site-populations (apfub), site-scattering (epfua), and mean bond-length (Å) values for untreated (Upva) and treated (Tpva) princivalleite
Full size table
Table 8 Weighted bond valences (valence units) for untreated (Upva) and treated (Tpva) princivalleite
Full size table

FTIR spectra in the (OH)-stretching region and band assignment

The infrared spectra of the untreated sample (Upva) recorded in the polarized mode both perpendicular and parallel to the c direction show (OH) band positions very similar to those observed in the holotype material described by Bosi et al. (2022). In detail, spectra recorded in polarized mode parallel to the crystallographic c-axis (E||c) show a significant band at 3365 cm–1, a very intense band around 3500 cm–1, two weaker bands at 3632 and 3644 cm–1, and two very weak bands at 3662 and 3671 cm–1 (Fig. 3 of Bosi et al. 2022). As observed typically for tourmaline spectra in the (OH) range, the main band is off-scale for the E||c direction due to excessive absorption, whereas spectra obtained perpendicular to the c-axis (E⊥c) show considerably weaker bands (Fig. 3 of Bosi et al. 2022). In accordance with Watenphul et al. (2016), Bosi et al. (2022) assigned these bands as follows: the band at 3365 cm–1 is consistent with the presence of minor Al along with Si in [4]-fold coordination, and the main band FTIR bands at ∼3500 cm–1 are likely caused by the occurrence of the atomic arrangements 3[Y(Fe, Mn, Al)ZAlZAl]–O3(OH)3, whereas the bands above 3600 cm–1 may be caused by the arrangements Y[(Li, Fe, Mn)(Fe, Mn, Al)Al]–O1(OH)–X(□). However, alternative assignments of atomic arrangements are possible by considering the model proposed by Fuchs et al. (2022) and Balan et al. (2023). The E||c-spectrum of the Upva sample reported in Fig. 5a shows only the two weak bands at 3662 and 3671 cm–1, while the other bands are off-scale.

Fig. 5
figure 5

Polarized FTIR spectra (E||c) of untreated and stepwise heat-treated (700 °C) princivalleite: a fundamental (OH) region and b (OH) overtone region. Spectra have been vertically adjusted for clarity. Peak positions are indicated. Sample thickness 270 μm. * = sample thinned to 117 μm, but spectrum normalized to the original thickness of 270 μm

Full size image

Spectra acquired after the stepwise heating experiments at 700 °C show significant changes (Fig. 5a). For spectra polarized parallel to the c-axis (E||c), the relatively large thickness of the sample obscures the strong fundamental (OH) bands. However, the weaker bands that occur at 3644, 3662, and 3671 cm−1 show a continuous decrease in intensity after each heating step and are chiefly absent after the last treatment. The (OH) overtone bands in the range 6700–7200 cm−1 also indicate substantial dehydrogenation during the heat treatments (Fig. 5b). It is apparent that the bands at higher wavenumbers, related to O1(OH), decrease faster than bands occurring at lower wavenumbers related to O3(OH).

Based on the result from Mössbauer spectroscopy as well as optical absorption spectroscopy with the Fe3+-enhanced Fe2+ band about 9000 cm−1 (Fig. 3) showing an increase in intensity after the sequential heating experiments, the following redox reaction occurred:

$$\left( {{\text{Fe}},{\text{Mn}}} \right)^{{{2} + }} + \left( {{\text{OH}}} \right)^{-} \, \to \,\left( {{\text{Fe}},{\text{Mn}}} \right)^{{{3} + }} \, + \,{\text{O}}^{{{2}{-}}} \, + \,{\raise0.7ex\hbox{$1$} \!\mathord{\left/ {\vphantom {1 2}}\right.\kern-0pt} \!\lower0.7ex\hbox{$2$}}{\text{H}}_{{2}} \left( {\text{g}} \right).$$
(1)

It is worth noting that the weak bands above ~ 3600 cm–1 observed for the untreated sample are consistent with low concentrations of O1(OH) obtained from the other methods. Similarly, the lack of bands above ~ 3600 cm–1 after the final treatment is consistent with the absence of O1(OH). Moreover, with respect to the untreated sample, the strongly decreased band intensities in the (OH) overtone region (6700–7200 cm–1) in the spectra of the treated sample are consistent with substantial deprotonation of the O3 site accompanied by oxidation of Fe and Mn from + 2 to + 3.

Crystal chemistry

Compositional differences between the untreated and treated samples occur in the Fe3+/ΣFe and Mn3+/ΣMn ratios (zero for untreated and, respectively, 0.92 and 0.25 for Tpva heated at 700 °C), and in (OH) contents (2.71 apfu for untreated, 2.14 apfu for Tpva) as a result of the redox reaction (1).

The structural data show that the treated Tpva sample is somewhat different from the untreated Upva: for instance, the unit-cell parameter a decreased from 15.9162(2) Å in Upva to 15.8323(2) Å in Tpva, whereas the c-parameter increased from 7.11485(9) Å in Upva to 7.13120(10) Å in Tpva (Table 2). Such a difference is also found by the variation in the refined site-scattering and mean bond-length values for the untreated–treated samples: e.g., the site-scattering of Y 51.2(2)–50.8(2) epfu and of Z 79.3(2)–82.2(3) epfu, < Y–O > 2.012–1.972 and < Z–O > Å 1.911–1.914 Å. All these variations reflect the almost total Fe oxidation from Fe2+ to Fe3+ and the oxidation of one-third amount of Mn from Mn2+ to Mn3+ along with a minor degree of disorder of Al–Fe–Mn over the Y and Z sites. On this basis, the observed decrease in the a-parameter is interpreted as a result of the Fe and Mn oxidation from + 2 to + 3, which occurred during the heating experiments, whereas the increase in the c-parameter is interpreted as a result of the Al–Fe–Mn disorder. This interpretation is consistent with other thermally treated tourmalines (Pieczka and Kraczka 2004; Bačίk et al. 2011; Bosi et al. 2016a, 2018, 2019; Liu et al. 2019; Celata et al. 2021; Ballirano et al. 2022a, b, 2023; Hovis et al. 2023).

The most evident result of the thermal treatment of the Fe-rich princivalleite sample in the air at 700 °C is the simultaneous presence of Fe2+, Fe3+, Mn2+, and Mn3+, which suggests that the oxidation process during the heating experiments was largely controlled by kinetic factors, with an almost total oxidation of Fe2+ to Fe3+ and a fair degree of oxidation of Mn2+ to Mn3+. This process is accompanied by significant deprotonation of the (OH) groups at the O1 and O3 sites (the VO2– content increased to 0.86 apfu in sample Tpva) according to the redox reaction (1). The oxidation of Fe2+ and Mn2+, traced by the Mössbauer and optical absorption data, is further supported by the decrease of the FTIR-band intensities related to the O1 and O3 sites (see above) as well as by the bond-valence analysis (Table 8). For sample Tpva, BVS at the O3 (= V letter in the general formula) site (1.31 valence units, vu) clearly indicates that it is not fully occupied by (OH), but significant amounts of oxygen also occur at O3.

The bond-valence method may be useful to evaluate the presence of O2– at O3. Of particular relevance in this regard is the plot of O3O2– versus BVS at O3 of Bosi et al. (2018) showing an expected linear trend from BVS(O3) values of ~1.1 to ~2.0 vu. Note that the minimum BVS(O3) value of ~ 1.1 vu is consistent with the O3 site fully occupied by (OH), which forms a weak hydrogen bond with the closest O5 atom (O3–H⋯O5). The plot of data of samples Upva and Tpva in Bosi et al.’s (2018) diagram shows that they are consistent with data from other samples and the expected trend (Fig. 4). Note that in Bosi et al. (2018) was missing the data of sample BLSH2 from Ertl et al. (2012).

Conclusions

Single-crystal X-ray diffraction and spectroscopic data acquired after stepwise heating experiments at 700 °C of a natural Fe-rich princivalleite sample provide information on the induced structural and compositional response. Results show that the oxidation of Fe and Mn from + 2 to + 3 leads to variation in the color from pale blue to dark brown. A similar change in color to darker shades was also observed by Ertl et al. (2012) in a yellow Fe-free fluor-tsilaisite, ideally NaMn3Al6(Si6O18)(BO3)3(OH)3F, which, after heating in air at 750 °C for 30 h, changed color to dark brown–red due to the oxidation of Mn2+ to Mn3+.

In our sample, the oxidation process is accompanied by significant deprotonation of the O1 and O3 sites, as confirmed by the progressive decrease in intensity of the bands above ~3600 cm–1 in the FTIR spectra during the heating experiments. However, the oxidation processes that Fe2+ and Mn2+ underwent are kinetically different, faster for Fe than Mn. In fact, almost all Fe2+ is oxidized after the treatment, whereas approximately one-third of the amount of Mn changed from + 2 to + 3. Structural data revealed that the YFe and YMn oxidation yield variations in the unit-cell parameters. Specifically, the decrease of the a-parameter is consistent with the oxidation of Fe2+ and Mn2+ and the consequent change in their ionic radius. On the other hand, the increase in the c-parameter during the treatment can be interpreted as a result of the increased content in Z(Fe, Mn).

This work represents the first in-depth investigation regarding the effects of heat treatment on a Fe-rich Mn-dominant tourmaline. In a previous study on a tourmaline sample with a slight enrichment in Mn (MnO ⁓ 1 wt%), a similar thermal treatment (750 °C, 90 h) led to a complete oxidation of Mn from + 2 to + 3 (Bosi et al. 2021). Although the different oxidation rates observed for Mn and Fe can be the result of their distinctive redox standard potential, the incomplete oxidation of Mn2+ could be related either to the duration, the temperature, or to the initial content of Mn in the sample. For this reason, further investigations are required to characterize the different behavior of Mn and Fe during the oxidation process.