Next Article in Journal
3D Hierarchical Composites of Hydrotalcite-Coated Carbon Microspheres as Catalysts in Baeyer–Villiger Oxidation Reactions
Next Article in Special Issue
In-Depth Characterization of Natural Clays from Southeast Albania
Previous Article in Journal
Effects of Radiation Damage on Metal-Binding Sites in Thermolysin
Previous Article in Special Issue
The Enamelled Tiles of Olite’s Castle (Spain): Characterization, Provenance, and Manufacture Technology
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Crystals
Volume 14
Issue 10
10.3390/cryst14100877
crystals-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Chemical Composition and Spectral Variation in Gem-Quality Blue Iron-Bearing Tourmaline from Brazil

by
Yifang Chen
1,
Duo Xu
1,*,
Zhengyu Zhou
2,*,
Dietmar Schwarz
3,
Junhao Zheng
1 and
Lingmin Zhang
1
1
School of Ocean & Earth Science, Tongji University, Shanghai 200092, China
2
State Key Lab of Marine Geology, School of Ocean & Earth Science, Tongji University, Shanghai 200092, China
3
Zmpery Gemology Lab, Shanghai 201403, China
*
Authors to whom correspondence should be addressed.
Crystals 2024, 14(10), 877; https://doi.org/10.3390/cryst14100877
Submission received: 9 September 2024 / Revised: 28 September 2024 / Accepted: 29 September 2024 / Published: 4 October 2024
(This article belongs to the Collection Topic Collection: Mineralogical Crystallography)
Browse Figures
Versions Notes

Abstract

This study, conducted a spectroscopic analysis of 10 gem-quality blue tourmaline samples from Minas Gerais, Brazil, focused on detailed variations in their infrared, Raman, and UV-VIS spectra. Conventional gemological tests, electron-probe microanalysis, infrared spectroscopy (mid- and near-infrared), Raman spectroscopy, and UV-visible spectroscopy were used to systematically analyze the chemical composition and spectral characteristics of the samples. The infrared spectra revealed vibrations of [YO6], [TO4], [BO3], [OH], and H2O groups, indicating different bonding profiles, with the [OH] vibrational frequency showing a direct correlation with FeO and MnO content. The Raman spectra primarily reflected the stretching vibrations of metal–oxygen bonds and hydroxyl groups, indicating the complexity of the local environment in the crystal structure. The UV-VIS spectra showed that the broad absorption band around 725 nm was due to intermetallic charge transfer between Fe2+ and Fe3+. This work provides new insights into the local bonding environment within the crystal structure by providing precise spectral data of natural blue tourmaline, and a more accurate classification and evaluation of blue tourmaline through fine spectral change characteristics related to crystal chemistry has important implications for both academic research and the gemstone industry.

1. Introduction

The complexity of the tourmaline supergroup comes mainly from its general chemical formula: XY3Z6[T6O18][BO3]3V3W, where X = □ (vacancy), Na, K, Ca, Y = Li, Mg, Fe2+, Mn2+, Cu2+, Al, V3+, Cr3+, Fe3+, Mn3+, Ti4+, Z = Mg, Fe2+, Al, V3+, Cr3+, Fe3+, T = Si, B, Al, B = B, V = OH, O, and W = OH, F, O. The letters in the general formula correspond to the cationic or anionic groups that occupy this position in the crystal structure. The complex crystal chemistry of tourmaline requires a multi-method investigation, and the occupancy has been widely determined [1]. According to the International Mineralogical Association (IMA) classification of the tourmaline supergroup (as of September 2024), 50 tourmaline mineral species have been identified. This classification is based on chemical composition definitions, and the main criterion is the place of a particular chemical composition in a given crystal structure.
Tourmaline belongs to the subclass of cyclosilicate, which consists of rings of six [TO4] tetrahedra. Six of the more regular [TO4] tetrahedrons form a compound trigonal ring [Si6O18]12− by sharing the two angular tips of each [SiO4] tetrahedron, with the X cation in its central void, which is connected to two classes of octahedrons. The two types of octahedrons are three central [YO6] large octahedrons containing Y ions and six surrounding [ZO6] small octahedrons containing Z ions. Boron atoms are distributed between the major octahedrons, and B forms a [BO3] triangular sheet with an adjacent O2 and two O8, forming a total of three [BO3] atomic groups [2,3,4]. The crystal structure (Figure 1) is quoted in the American Mineralogist Crystal Structure Database [5].
The most common variety of gem-quality tourmaline is elbaite, which has the widest range of colors, including green, blue, pink, red, orange, yellow, colorless, and multicolor [6]. Due to the complex crystal chemistry, the color of tourmaline is not limited to a specific variety. Blue tourmaline can be elbaite, cuprian (copper-bearing) liddicoatite, schrol, dravite, etc. [7,8,9]. Brazil is known for producing tourmaline in almost all gemstone colors. These tourmalines, commonly referred to as “Paraíba-type,” became popular after their discovery in the late 1980s in the state of Paraíba, Brazil [10,11]. Their unique colors are attributed primarily to copper ions (Cu2+) and their interactions with other elements, such as manganese (Mn), iron (Fe), and titanium (Ti) [12]. But in addition to this particular tourmaline, the market is also focusing on gem-quality blue tourmaline from other sources in Brazil, and some even show a “sapphire blue”.
Previous studies have extensively investigated the crystal chemistry, spectroscopic properties, and gemological characteristics of copper-bearing tourmalines from various localities, including Brazil, Nigeria, and Mozambique [12,13,14]. Vereshchagin et al. (2013) refined the crystal structures of natural and synthetic copper-bearing tourmalines and analyzed their crystal chemistry, highlighting the ordered distribution of copper and other elements in the Y site [15]. Similarly, Watenphul et al. (2016) employed Raman spectroscopy to study the potential of this technique for species identification and site-occupancy analysis in complex hydrous silicates, including tourmalines. Their work emphasized the correlation between the positions and intensities of Raman vibrational modes and the compositional variations within tourmaline species [16,17]. Moreover, Bronzova et al. (2019) studied the short-range order in Li–Al tourmalines using infrared spectroscopy and X-ray single-crystal diffraction, revealing the stability of atomic arrangements coordinating W- and V-sites and proposing a unified model for assigning absorption bands in IR spectra. Their findings underscored the importance of considering both the local and long-range order for understanding the structural variations in tourmalines [18]. Despite these extensive studies, there remains a need to explore the unique properties of tourmalines from Brazil, especially those exhibiting distinct compositions and colors. In this study, we focus on iron-rich blue tourmalines, which differ significantly from the copper-bearing blue tourmalines previously studied. These iron-rich tourmalines offer a unique opportunity to investigate the relationship between chemical composition, spectroscopic properties, and coloration mechanisms.
This project is dedicated to the study of the characteristics of different shades of blue tourmaline using a variety of methods such as general gemological methods, electron microprobe, infrared spectroscopy, Raman spectroscopy, and UV-visible spectroscopy. Previous studies on the spectra of iron-bearing tourmaline were mostly normalized results, and detailed spectral data from natural mineral samples were lacking. This project analyzed 10 natural blue tourmaline samples and comprehensively compared the changes in infrared, Raman, and UV-visible spectra, which were directly related to the crystal chemistry changes. Our objectives include the analysis of the chemical composition and the study of its crystal structure and spectral properties using Raman and infrared spectroscopy. By integrating these techniques, this project aims to contribute to the body of knowledge on the chemical changes and coloring mechanisms of tourmaline crystals, thereby deepening our understanding of these fascinating gems.

2. Materials and Methods

2.1. Materials

The samples were 10 blue tourmalines of different shades (Figure 2). The samples were obtained from the areas of Governador Valadares and Aracuaí in the state of Minas Gerais, Brazil, both of which are tourmaline-bearing pegmatite deposits. The samples are numbered from S01 to S10, and additional information is given in Table 1.

2.2. Methods

2.2.1. General Gemological Analysis

A gemological microscope, refractometer, spectroscope, single-plate balance (hydrostatic weighing method), and ultraviolet fluorescent lamps were used to observe the general gemological characteristics of the samples. Among them, the gemological microscope model is GM143/168, the refractometer model is NZS-03, the spectroscope model is MG7206, and the UV fluorescent lamp model is FUV-4.

2.2.2. Electron Microprobe Analysis (EPMA)

The electron-probe microscope analyzer with a JEOL JXA-8230 from Japan Electronics Company (Tokyo, Japan) was used to analyze the mineral major elements of the samples quantitatively. The test conditions were an acceleration voltage of 15 kV, an electron beam current of 10 nA, and an electron beam diameter of 5 μm. Natural and synthetic minerals or oxides were used as standards, and the data were processed using the ZAF correction method. The data result is the average of 3 points of analysis for each sample.
Lithium is an important component of tourmaline, but like H and B, it cannot be accurately analyzed by electron probes. In this project, the method was based on a multiple linear regression model proposed by Pesquera et al. (2016) to estimate Li from the electron-probe data of tourmaline, i.e., Li2O = 2.356 + 0.124 SiO2 − 0.121Al2O3 − 0.178 FeO − 0.162 MnO. The model uses statistically significant wt.% oxide data and does not rely on the normalization procedure [19]. In addition, the model is highly accurate (r2 = 0.95) for the Li2O content data derived from low-Mg tourmaline (<2 wt.% MgO) for the case of this project, whereas the previously widely used expression for approximating Li content on a stoichiometric basis Li (atoms per formula unit(apfu)) = 15 − (T + Y + Z) cations assumes (OH + F) = 4 is less accurate (r2 = 0.79).

2.2.3. Infrared Spectroscopy (IR)

The samples were tested by a Fourier transform infrared (FTIR) spectrometer TENSOR27 from BRUKER OPTICS, Ettlingen, Germany. The test surfaces of the samples have been polished. The test conditions were a resolution of 4 cm−1, 32 scans, grating setting of 6 mm, scanning speed of 10 KHz, mid-infrared spectral scanning range of 400~4000 cm−1, and near-infrared spectral scanning range of 3000~9000 cm−1. The mid-infrared spectral data were converted to absorbance by k-k, and the near-infrared spectral data were used as the transmittance. Because the S10 sample was too dark in color, it was partially ground into powder to obtain more accurate mid-infrared spectral data.

2.2.4. Laser Raman Spectroscopy (RS)

A LabRAM HR Evolution high-resolution Raman spectrometer from Horiba, France, was used for Raman spectroscopy. The test conditions were an excitation light source of 532 nm, a power of 50 mW, a grating inscription density of 600 g·mm−1, a confocal pinhole of 120 μm, a measurement range of 100~4000 cm−1, an acquisition time of 10 s, an accumulation number 3, and in silico calibration before testing.

2.2.5. Ultraviolet-Visible Absorption Spectroscopy (UV-VIS)

The UV-VIS absorption spectra of the samples were tested by an Ocean Optics GEM-3000 ultraviolet-visible spectrometer. The test method was the reflectance method. The test conditions were an integration time of 120 ms, an average number of times 20 times, a smooth width of 5, and a wavelength band of 220~980 nm.

3. Results

3.1. Gemological Characteristics

The 10 test samples selected for this project were orientated slices cut and ground parallel to their crystallographic C-axis direction. All 10 tourmaline samples were vitreous and transparent, and the colors from S01 to S10 were greyish blue, light blue, bluish green, dark greyish blue, and dark blue. All exhibit strong dichroism except for S01/02, which has a lighter body color. The relative densities of the samples are 2.93~3.10. The refractive indices are 1.620~1.642. The birefringence is 0.019~0.020. The samples are inert under the long-wave and short-wave UV fluorescent lamps, and the details are listed in Table 2. The gas–liquid two-phase inclusions, fibrous or long tubular inclusions parallel to the C-axis, healed fractures, and other common features of tourmaline inclusions can be seen when the sample is enlarged (Figure 3).

3.2. Chemical Composition

For the EPMA data of 10 blue tourmalines (Table 3), it was possible to estimate point occupancy by using stoichiometric H2O and B2O3 content, as well as to determine the IMA-recognized alkali-, calcic-, and vacant-group tourmaline species and their end-member composition [1]. The results of the classification are presented in Figure 4 The classification of the measured tourmaline samples according to the cations of Na, K, Ca, or vacancy in the X-point position, all of them belong to alkali tourmaline (Figure 4a), and based on the occupancy of OH1− in the W-point position (Figure 4b) and then further classified according to the alkali tourmaline in the Y-point position of Li, Fe, Mg (Figure 4c), it shows that all the measured samples are lithium tourmaline (alkali-subgroup 2). Considering that the sample also showed F1− occupying the W position, the final result was S01/02/03/04/06/09 as fluor-elbaite and S05/07/08/10 as elbaite.

3.3. Spectroscopic Characteristics

3.3.1. Infrared Spectroscopy (IR)

The infrared spectra of the blue tourmaline samples S01 to S10 are shown in Figure 5. The infrared spectra of the 10 tourmaline samples were similar overall, with the mid-infrared spectra mainly showing peak displacement and splitting (Figure 5a), while the near-infrared spectra mainly showed obvious differences in the absorption intensity of the peaks (Figure 5b).
The mid-infrared spectra of tourmaline (Figure 5a) are mainly generated by the vibrations of IR-active groups, including [YO6], [TO4], [BO3], [OH], and H2O. The characteristic peak positions in the range of 500–1400 cm−1 are generally shifted to the low-frequency region from S01 to S10, while the absorption peaks in the range of 3400–3800 cm−1 are shifted to the high-frequency region.
For [YO6] octahedral vibrations, most of the vibrations of the Y-O cationic bonds are located close to or below 400 cm−1. Based on the vibrational frequencies of the various ligand polyhedral [21,22]: [NaO8] 270 cm−1, [CaO8] 380 cm−1, [FeO6] 400 cm−1, and [MgO6] 470 cm−1, the absorption peaks at 431 cm−1 were attributed to νFe-O, and the 457 cm−1 absorption peaks were attributed to νMg-O. The absorption intensities of the above peaks of S01–04 and S08–09 are significantly lower than those of the other samples, corresponding to the relatively low Fe and Mg contents of these samples in the electron microprobe analyses. (FeO < 0.20 wt.% and MgO < 0.02 wt.%).
For [TO4] tetrahedral vibrations, the infrared vibrational modes of [SiO4] tetrahedra can be classified into: νs Si-O-Si (νs-symmetric telescopic vibration), νs Si-O-Si (νas-asymmetric telescopic vibration), νs O-Si-O, νs O-Si-O, and δSi-O (δ-bending vibration) [23]. δSi-O is in the range of 600 cm−1 or less. S01 to S09 all exhibit corresponding two peak positions, while S10 undergoes spectral peak splitting to form three absorption peaks. νs Si-O-Si in the range of 600 to 830 cm−1, S01 to S10 all exhibit four absorption peaks, of which there are two strong absorption peaks (e.g., 717 and 795 cm−1 for S06) and two weak absorption peaks (e.g., 629 and 755 cm−1 for S06). νs O-Si-O, νas O-Si-O, and νas Si-O-Si in the range of 900 to 1200 cm−1, S01 to S10 all showed three strong absorption peaks, of which two peaks of low frequency corresponded to νs and νas of O-Si-O. The high-frequency peaks correspond to νas Si-O-Si.
For [BO3] group vibrations, studies of [BO3] are often referred to as borates, which can be divided into νas [BO3], νs [BO3], and δ [BO3], based on the delineation of the IR vibrational modes of the borate ion [BO3] 3−. νas [BO3] is in the range of 1300 to 1500 cm−1 [24], and so, the two symmetrical absorption peaks of S01 to S10 in this range were attributed to νas [BO3].
For [OH] hydroxyl vibration, in the crystal structure of tourmaline, three OH groups are connected to two cations at the Z-point site and one cation at the Y-point site, which are called external hydroxyl OH(3), and one OH group is connected to three cations at the Y-point site, which is called internal hydroxyl OH(1). The stretching vibration of the O-H bond corresponds to the range from 3400 to 3800 cm−1, and the absorption peaks below 3600 cm−1 represent the stretching vibration of hydroxyl OH(3) [25,26,27,28]. So, the two or three vibrational bands (e.g., S06: ν1 = 3588 cm−1, ν2 = 3558 cm−1, and ν3 = 3485 cm−1) appearing in this range for S01 to S10 are attributed to OH(3), and the bonding environment is (Fe, Li, Al)YAlZAlZ [29]. Compared with ν1 and ν2, ν3 exhibits a stronger and broader band. From S01 to S10, ν3 gradually shifts to a higher wave number with the increasing FeO + MnO content (Figure 6), and the two show a significant positive correlation (r2 = 0.97).
The near-infrared spectra of tourmaline (Figure 5b) show mainly multiplicative and combinatorial frequencies of hydroxyl groups in different forms of existence. The peak positions of the 10 samples are in good agreement, except for the differences in the absorption intensities of the spectral peaks.
A series of sharp peaks in the interval of 4000–4800 cm−1 are attributed to the stretching and bending vibrations of cationic hydroxyl units (Y-OH, Y = Al, Mg, Fe, Mn, etc.) [30]. The 4164 and 4206 cm−1 peaks are attributed to the [Fe2+, Mn]-OH units, with the 4164 cm−1 as the dominant peak in S01 to S04 and the 4206 cm−1 peak in S05 to S10. The peaks in the interval from 4300 to 4500 cm−1 are attributed to Mg-OH and Fe-OH, with 4344 cm−1 corresponding to the [Mg, Mg]-OH unit, 4444 cm−1 to the [Fe2+, Fe2+]-OH unit, and 4544 cm−1 to the [Fe2+, Mg]-OH unit [31]. The absorption peaks at 4444 and 4544 cm−1 show the most pronounced increase in absorption intensity from S01 to S10, which is consistent with the change in Fe2+ content in the electron microprobe data. The 4598 cm−1 corresponds to the Al-OH unit. The 4873 cm−1 is the tertiary octave of νas Si-O-Si. Spectral peaks in the interval from 4900 to 5400 cm−1 are attributed to the combination frequencies of the bending vibration of the water and the stretching vibration, with 5180 and 5365 cm−1 implying that tourmaline contains water molecules [32]. The spectral peaks in the interval from 6800 to 7200 cm−1 are attributed to the first-octave frequency of the hydroxyl [OH] telescopic vibration, which corresponds to the [OH] fundamental frequency region of the mid-infrared spectrum from 3400 to 3800 cm−1, where the absorption peaks of S10 disappear in this range. Among them, 6783 cm−1 corresponds to the hydroxyl group replacing the O2− position in [BO3], 6995, 7132, and 7185 cm−1 correspond to the first multiplicative frequency of the [OH] vibration residing at the center of the complex tripartite ring, the relatively strong 6995 cm−1 corresponds to the pure hydroxyl group, and the weak peaks of 7132 and 7185 cm−1 correspond to the hydroxyl group in the water molecule.

3.3.2. Laser Raman Spectroscopy (RS)

The Raman spectra of blue tourmaline samples S01 to S10 are tested in Figure 7.
The Raman spectra of tourmaline are mainly generated by the vibration of anionic groups with Raman activity. The Raman shifts at lower frequencies from 200 to 1200 cm−1 are related to the metal–oxygen bonding of the mineral (16A1 + 17E Raman modes), and the Raman shifts at higher frequencies from 3400 to 3800 cm−1 are related to the hydroxyl stretching vibrations in different lattice positions (A1 Raman modes) in different lattice positions [17,33,34].
For the [YO6] and [ZO6] octahedral vibration, the 200–240 cm−1 range is dominated by the [YO6] vibration, and the strongest Raman peak in the 360–375 cm−1 range is generated by the [ZO6] vibration [16]. The peak at 223 cm−1 of S01 to S07 and S10 is attributed to the Mg-O stretching vibration at the Y-point position [35], and 252/241 cm−1 of S08/09 (MgO = 0 wt.%) corresponds to the O-Al-O bending vibration. The strong peak near 373 cm−1 of S01 to S07 and S10 is attributed to the Al-O stretching vibration at the Z-point position [30]. Meanwhile, the Al-O stretching vibrations of S08/09 are shifted to 400/392 cm−1.
For the [Si6O18] complex tripartite ring vibrations, the symmetric stretching vibrational peaks of the rings are located in the 400–570 cm−1 interval. The 408 cm−1 peaks of S05/06/10 correspond to the O-Si-O bending vibrations in the rings, and the Raman peaks near 510 cm−1 of S01–S07 correspond to the oxygen vibrations in Si-O rings. The antisymmetric stretching peaks of the rings are located in the 600–700 cm−1 interval, and the Raman peaks near 634 cm−1 of S05–S07 correspond to Si-O (bridging oxygen) bond rocking. The Raman peaks near 1059 and 1087 cm−1 of S01–S07 and S10 correspond to Si-O (non-bridging oxygen) stretching vibrations, and the peaks near 1110/1104 cm−1 of S08/09 correspond to Si-O (bridging oxygen) stretching vibration. The above four peaks of S08/09 are shifted to higher frequencies by about 30 cm−1, while two of the peaks of S01 to S04 and S10 have weak or missing absorption.
For the [BO3] atomic cluster vibrations, the peaks of B-O stretching vibration and O-B-O bending vibration are located in the interval of 700–800 cm−1 [36]. The Raman peaks of S01 to S04 are located at 710, 752 cm−1. The Raman peaks of S05 to S07 are located at 705 and 752 cm−1. The Raman peaks of S08/S09 are located at 731/725 and 776/769 cm−1, and the Raman peaks of S10 are located at 728 and 757 cm−1.
For the [OH] vibrations, three [OH] stretching vibrational peaks appear in the interval from 3400 to 3700 cm−1 for all of S01 to S10, and the spectral peaks of S01 to S04 are located in the vicinity of 3480 (ν3), 3588 (ν1), and 3655 (ν2) cm−1. The peaks at 3480 and 3588 cm−1 correspond to the OH(3) groups ((Li, Al, Fe, Mn) YAlZAlZ), while 3655 cm−1 corresponds to the OH(1) group (LiYAlY(Fe, Mn)Y) [33,34]. Meanwhile, the ν1 spectral peaks of S05 to S10 split into double peaks located near 3565 and 3600 cm−1 and ν2 disappears [32]. This feature is consistent with the performance in the mid-infrared spectra. A corresponding short-range occupation assignment was performed, and the results are shown in Table 4 [35].

3.3.3. Ultraviolet–Visible Absorption Spectroscopy (UV-VIS)

The UV-VIS spectra of the blue tourmaline samples S01 to S10 are shown in Figure 8. To avoid the influence of sample dichroism on the spectral test, all analyzed samples were cut parallel to the C-axis orientation.
In the raw UV-VIS spectra (Figure 8a), all samples show strong absorption in the UV region below 370 nm. S01 to S9 have transmission windows in the violet-to-yellow wavelength region (400–600 nm), and all show symmetrically broad absorption bands centered at 725 nm in the deep-red region. These samples lack absorption in the blue–violet region, thereby transmitting a disproportionate amount of light. These samples lack absorption in the blue–violet region and, thus, transmit disproportionate light, ultimately leading to a blue coloration. S10 is unusual because it is fully absorbed in the visible region, black to the naked eye, and “sapphire blue” in transmitted light.
As the absorption peaks within the visible region of 400–800 nm are difficult to calibrate, second-order derivative spectra (Figure 8b) were used to separate them for more accurate localization [37]. Second-order derivative spectroscopy improves the resolution of the raw spectral signals but also makes the spectral noise increase, and the peaks were smoothed to facilitate analysis and comparison. It should be noted that all samples show false absorption peaks around 400 nm, which is due to the discontinuous switching of the light source. Although the derivative spectra of S10 show a series of wave-like symmetric peaks, only the absorption below the baseline at 417 nm is a reliable absorption peak.
The weak absorption peaks were carefully located and separated in the 400–600 nm region of the second-order derivative spectrum (Figure 8b). S01/02 exhibited absorption peaks at 379, 415, and 500 nm, S03/04 at 377, 416, 461, and 498 nm, and S05/06 at 377, 417, 460, 498, and 555 nm. S07/08/09 exhibited absorption peaks at 377, 416, 456, 498, and 555 nm, and S10 exhibited absorption peaks at 378 and 417 nm.
The strong absorption peak at 378 nm reflects the Fe3+ outer electron leaps (6A1 → 4E). The absorption peak at 416 nm is attributed to Fe2+- Ti4+ interval charge transfer [38]. The absorption peaks at 456/460 nm reflect the Fe3+ outer electron leaps (6A1 → 4E + 4A1). The absorption peak at 498 nm is believed to be attributed to the Fe2+ spin forbidden leaps. The weak absorption peak at 555 nm is thought to be attributed to the forbidden leaps of Mn2+ (6A1 → 4T1), which is consistent with the presence of a small amount of Mn (0.82 wt% < MnO < 2.56 wt.%) in the electron microprobe results as a chromogenic cation.

4. Discussion

4.1. Crystal Structure and Spectral Variation

Combining the differences in the spectral peaks of the different kinds of spectra and relating them to the differences in the chemical composition content of FeO and MgO, the 10 samples can be classified into four categories, i.e., S01 to S04 (0.44 wt.% < FeO < 1.51 wt.%, MgO < 0.02 wt.%), S05 to S07 (3.30 wt.% < FeO < 4.05 wt.%, 0.02 wt.% < MgO < 0.05 wt.%), S08 to S09 (2.25 wt.% < FeO < 2.37 wt.%, MgO = 0 wt.%), and S10 (FeO = 4.9 wt.%, MgO = 0.03 wt.%).
In the mid-infrared spectra, the vibrational peak shapes of the [SiO4] tetrahedra and [BO3] triangles in the range of 500–1400 cm−1 are the same for all tourmaline samples, but the peak positions of the S01 to S10 in this range as a whole are regularly shifted to the lower frequencies with the increase in Fe content, and this pattern is typical of the disordered distribution of cations in non-equivalent positions. When the Fe2+ content is low, the larger Fe2+ preferentially occupies the Y site, and the cation distribution shows a clear tendency to be ordered. As the Fe2+ content continues to increase, more Y sites are occupied, and Fe2+ enters the Z site. The disorder in the cation distribution increases, leading to the splitting of the spectral peaks of S10 in the small range of 500–600 cm−1. The [YO6] octahedral vibrational peaks below 500 cm−1 and the vibrational peaks in the [OH] fundamental frequency region from 3400 to 3800 cm−1 are more different. From S01 to S10, the absorption intensity of the νFe-O peak in the octahedral vibrational region of the [YO6] octahedron below 500 cm−1 increases with the increase in FeO content and the increase in the radius and mass of mass-like isomorphic replacement ions, which finally shows the sharpest expression in the S10 sample. In the [OH] fundamental frequency region from 3400 to 3800 cm−1, the ν3 of elbaite and Fe-elbaite is more sensitive to hydrogen bonding than ν1 and ν2 [29]. The ν3 vibrational band moves towards a higher wave number, and this shift can also be explained by the presence of the Y-point cations Fe2+ and Mn2+, which represent the enhanced stretching vibration of the chemical bond between the hydroxyl group and the environment, i.e., 3(Li, Fe2+, Mn2+, Al)YAlZAlZ-OH (3) [17,39]. It can be observed in Figure 6 that the position of ν3 is controlled by the content of (Fe2+, Mn2+)Y, and 3465 cm−1 is assigned as Fe*YAlZAlZ-2AlYAlZAlZ with Fe*Y = (Fe2+ +Mn2+) Y. It is not possible to distinguish individual spectral peaks associated only with divalent iron or manganese because these elements are very similar in terms of mass, electronegativity, and ionic radius, resulting in nearly identical vibration frequencies [17]. In addition, the strongest absorption peaks in the absorption spectra of the IR spectra of the 10 samples are all located in the range of 1000–1100 cm−1, which belong to the [SiO4] tetrahedral vibrational generation, and are also believed to belong to the Si-O-Al stretching generation, which is presumed to be due to the increased intensity of the IR absorption spectra due to the enhanced Al-Si substitution caused by Al enrichment in the samples [24]. The vibrational frequency of νas Si-O-Si is larger than that of νas O-Si-O, being in the range of 900–1200 cm−1, which is due to the larger bond angle of hexagonal cyclic silicates Si-O-Si. All samples show a strong absorption peak of νas Si-O-Si near 1110 cm−1, a characteristic IR spectral absorption peak that is common to many different layered silicate minerals containing OH groups [21]. In the NIR spectrum, the 7132 cm−1 peak shows hydroxyl groups in the water molecules in the center of the complex tripartite ring, suggesting a high water fugacity and the presence of water molecules in the tunnel structure during the formation of tourmaline. The hydrogen endowment state in tourmaline is summarized in two forms, namely hydroxyl and water molecules. In addition, none of the 10 tourmaline samples showed significant H2O stretching vibrational bands (broad absorption bands from 3200 to 3550 cm−1) with its bending vibrational bands (1600 to 1660 cm−1), indicating that there is no significant presence of adsorbed water.
In the Raman spectrum, the peak at 223 cm−1 can quickly and effectively distinguish elbaite from uvite/Mg-foitite/dravite, as the latter would exhibit the characteristic two Mg-O vibrational peaks due to the substitution of Al and Li for Mg at the Y point position [35]. It is also possible to distinguish elbaite and liddicoatite based on the position of the main and weak peaks in the [OH] vibrational region. Overall, the samples have the same number of peaks in the Raman spectra, but the symmetry of the peaks tends to decrease from S01 to S10. The splitting of the B-O vibrational peak in [BO3] into two peaks in the range of 700 to 800 cm−1 is thought to be caused by the inconsistency of the B-O2 and B-O8 bond lengths, making the position group of [BO3]3− drop from C3 to C2v [32]. The [ZO6] vibration peak depends on the metal cation content of the octahedral coordination [16]. In general, the Raman peak assignment model of [OH] is affected by the chemical properties of the three YZZ octahedral position atoms that share the O3 atom with OH(3). The Raman spectra of YZZ coordination in 3300–3900 cm−1 should contain only two Raman peaks, OH(3)-ν3 and OH(3)-ν1. But when at least two different elements occupy one or more crystalline cation sites, the significant deviation from the expected number of only two OH stretching peaks is due to the presence of different chemical compositions around the hydroxyl group [17]. It can be interpreted as the decrease in the symmetry of the Raman vibrational peaks of [OH] because the increase in FeO content suppresses the ν2 vibrational peaks, and the entry of Fe2+ into the octahedra makes the lattice dilated, leading to complex coordination ions, and lattice distortion of the Y-O octahedra, leading to the appearance of ν1 double peaks and the disappearance of the ν2 peaks. Both hydroxyl groups, corresponding to the OH(3) group and OH(1) group, can be produced in case of substitution of oxygen by fluorine [36].
This study presents detailed spectroscopic data that demonstrate how subtle chemical variations directly influence the vibrational spectroscopy (IR, Raman) of natural blue tourmaline samples. The observed shifts in spectral peaks, especially in the OH and Si-O vibrational regions, offer new insights into the local bonding environments of these minerals, showcasing nuances that were previously unreported in the literature with similar sample sets.

4.2. Causes of Color

The diversity of tourmaline colors is essentially related to the different transition metal elements and is caused by several different mechanisms [40,41], including (1) ligand field (LF, or d-d) transitions, (2) matrix-to-metal charge transfer (CT) transitions, and (3) metal-to-metal intervalence charge transfer (IVCT) transitions. For example, the green color of elbaite is mainly caused by the d-d charge transfer of Fe2+, while the green color of uvite–dravite mixtures is green due to the presence of V and small amounts of Cr [42,43]. The brownish-yellow color of dravite and the yellow color of manganese-rich elbaite is mainly due to Mn2+-Ti4+ charge transfer [44,45]. The pink and red colors of elbaite and liddicoatitic are due to d-d electron leaps of Mn3+ [23,46,47,48], and the yellow hue produced when artificially irradiating pink tourmaline to deepen its red hue is due to O-hole color centers [49]. In contrast, colorless tourmaline does not contain or only contains small amounts of excess metal ions.
The blue tourmaline study samples S01/02 in this paper are greyish blue. S03/04 are light blue. S05/06 are bluish green. S07/08/09 are dark greyish blue, and S10 is dark blue.
Because EPMA data for tourmaline do not determine the content of Fe2+ and Fe3+ and their distribution at Y and Z positions, it is common practice to identify total iron as Fe2+ at Y-positions. However, the visible spectrum does demonstrate a distribution of Fe2+ and Fe3+ between the Y-and Z-sites (0.4wt.% < FeO < 5 wt.%). All the samples exhibit a broad band of absorption centered at 725 nm (Figure 8a), which is due to YFe2+-ZFe3+ interaction [9,50]. The crystal structure of tourmaline has several co-apical and co-prismatic octahedra at different positions. Fe2+ and Fe3+ accommodated at this position are prone to IVCT, and this absorption broadband is more pronounced with the increasing iron content. The samples with a distinct green hue have a lower absorption intensity in the 725 nm band and show wider asymmetric spectral bands compared to the blue samples. Whereas the S10 sample looks too dark in appearance, which may be due to the presence of excess Fe2+-Fe3+, the quasi-homogeneous substitution of cations at the Y-point position leads to structural distortion of the octahedra and the increase in the Y-O distance, thereby decreasing the octahedral field-splitting coefficient value, resulting in a shift of the tourmaline color towards darker colors.

5. Conclusions

Ten blue tourmaline samples of varying shades from Brazil were systematically analyzed both gemologically and spectroscopically, revealing differences in their chemical composition and spectral characteristics. All samples exhibit a glassy luster, transparency, and strong dichroism. Common tourmaline inclusions are visible. Chemical composition analyses identified the tourmaline samples as elbaite, with significant variations in FeO, MnO, and MgO content among them. Infrared spectral analysis indicated vibrations of groups such as [YO6], [TO4], [BO3], [OH], and H2O, reflecting different bonding profiles within the crystal structure. Notably, the [OH] vibrational shift is directly related to the wt.% (FeO + MnO), with the overall spectral patterns primarily influenced by FeO content. Raman spectral analysis highlighted metal–oxygen bonding and hydroxyl group stretching vibrations, revealing the complexity of the crystal structure’s local environment. Based on the Raman frequencies of [OH], short-range crystal occupancy assignments of metal cations were inferred for different samples. The UV-visible spectral analysis demonstrated that the symmetric broad absorption band centered at 725 nm in the blue samples results from intermetallic charge transfer between Fe2+ and Fe3+, with Fe content directly affecting the depth of the blue hue.
This study represents an advancement in the understanding of the spectroscopic properties of iron-bearing blue tourmaline. By providing detailed spectral data from natural mineral samples, it bridges a critical gap left by previous studies. The fine-scale spectral variations correlated with chemical composition have important implications for both academic research and the gemstone industry, allowing for more accurate classification and evaluation of blue tourmalines based on their detailed spectroscopic signatures.

Author Contributions

Conceptualization, Y.C. and D.X.; methodology, Z.Z. and L.Z.; formal analysis, Y.C. and D.X.; investigation, Y.C.; resources, D.S.; data curation, Y.C. and J.Z.; writing—original draft preparation, Y.C.; writing—review and editing, D.X. and Z.Z.; visualization, Y.C. and J.Z.; supervision, Z.Z.; project administration, Z.Z.; funding acquisition, Z.Z. All authors have read and agreed to the published version of the manuscript.

Funding

This research was supported by special funding from the Shanghai Science and Technology Commission (15DZ2283200, 18DZ2281300) and the Gem Discipline Development Fund.

Data Availability Statement

The original contributions presented in the study are included in the article, further inquiries can be directed to the corresponding authors.

Acknowledgments

This project is grateful to D.S. for providing research samples, to L.Z. for her help in the EPMA experiments, and to Z.Z. for his valuable comments and financial support.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Henry, D.J.; Novák, M.; Hawthorne, F.C.; Ertl, A.; Dutrow, B.L.; Uher, P.; Pezzotta, F. Nomenclature of the Tourmaline-Supergroup Minerals. Am. Mineral. 2011, 96, 895–913. [Google Scholar] [CrossRef]
  2. Bosi, F. Tourmaline Crystal Chemistry. Am. Mineral. 2018, 103, 298–306. [Google Scholar] [CrossRef]
  3. Bosi, F. Crystal Chemistry of the Elbaite-Schorl Series. Am. Mineral. 2005, 90, 1784–1792. [Google Scholar] [CrossRef]
  4. Pasetti, L.; Borromeo, L.; Bersani, D.; Andò, S.; Schnellrath, J.; Hennebois, U.; Karampelas, S. Identification of Some Gem Quality Blue to Green Li-Tourmalines. Minerals 2023, 14, 44. [Google Scholar] [CrossRef]
  5. Downs, R.T.; Hall-Wallace, M. The American Mineralogist Crystal Structure Database. Am. Mineral. 2003, 88, 247–250. [Google Scholar]
  6. Sun, Z.; Palke, A.C.; Breeding, C.M.; Dutrow, B. A New Method for Determining Gem Tourmaline Species by LA-ICP-MS. Gems Gemol. 2019, 55, 2–17. [Google Scholar] [CrossRef]
  7. Katsurada, Y.; Sun, Z. Cuprian Liddicoatite Tourmaline. Gems Gemol. 2017, 53, 34–41. [Google Scholar] [CrossRef]
  8. Nasdala, L.; Wildner, M.; Giester, G.; Chanmuang, N.C.; Scicchitano, M.R.; Hauzenberger, C. Blue Dravite (‘Indicolite’) from the Elahera Gem Field, Sri Lanka. J. Gemmol. 2021, 37, 618–630. [Google Scholar] [CrossRef]
  9. Li, M. Characterization of Blue Tourmaline from Madagascar for Exploring Its Color Origin. Adv. Condens. Matter Phys. 2022, 2022, 1–7. [Google Scholar] [CrossRef]
  10. Okrusch, M.; Ertl, A.; Schüssler, U.; Tillmanns, E.; Brätz, H.; Bank, H. Major- and Trace-Element Composition of Paraíba-Type Tourmaline from Brazil, Mozambique and Nigeria. J. Gemmol. 2016, 35, 120–139. [Google Scholar] [CrossRef]
  11. Katsurada, Y.; Sun, Z.; Breeding, C.M.; Dutrow, B.L. Geographic Origin Determination of Paraiba Tourmaline. Gems Gemol. 2019, 55, 648–659. [Google Scholar] [CrossRef]
  12. Laurs, B.M.; Zwaan, J.C.; Breeding, C.M.; Simmons, W.B.; Beaton, D.; Rijsdijk, K.F.; Befi, R.; Falster, A.U. Copper-Bearing (Paraíba-Type) Tourmaline from Mozambique. Gems Gemol. 2008, 44, 4–30. [Google Scholar] [CrossRef]
  13. Shigley, J.E.; Cook, B.C.; Laurs, B.M.; Bernardes, M.O. An Update on “Paraíba” Tourmaline from Brazil. Gems Gemol. 2001, 37, 260–276. [Google Scholar] [CrossRef]
  14. Abduriyim, A.; Kitawaki, H.; Furuya, M.; Schwarz, D. “Paraíba”-Type Copper-Bearing Tourmaline from Brazil, Nigeria, and Mozambique: Chemical Fingerprinting by LA-ICP-MS. Gems Gemol. 2006, 42, 4–21. [Google Scholar] [CrossRef]
  15. Vereshchagin, O.S.; Rozhdestvenskaya, I.V.; Frank-Kamenetskaya, O.V.; Zolotarev, A.A.; Mashkovtsev, R.I. Crystal Chemistry of Cu-Bearing Tourmalines. Am. Mineral. 2013, 98, 1610–1616. [Google Scholar] [CrossRef]
  16. Watenphul, A.; Schlüter, J.; Bosi, F.; Skogby, H.; Malcherek, T.; Mihailova, B. Influence of the Octahedral Cationic-Site Occupancies on the Framework Vibrations of Li-Free Tourmalines, with Implications for Estimating Temperature and Oxygen Fugacity in Host Rocks. Am. Mineral. 2016, 101, 2554–2563. [Google Scholar] [CrossRef]
  17. Watenphul, A.; Burgdorf, M.; Schlüter, J.; Horn, I.; Malcherek, T.; Mihailova, B. Exploring the Potential of Raman Spectroscopy for Crystallochemical Analyses of Complex Hydrous Silicates: II. Tourmalines. Am. Mineral. 2016, 101, 970–985. [Google Scholar] [CrossRef]
  18. Bronzova, Y.; Babushkina, M.; Frank-Kamenetskaya, O.; Vereshchagin, O.; Rozhdestvenskaya, I.; Zolotarev, A. Short-Range Order in Li–Al Tourmalines: IR Spectroscopy, X-ray Single Crystal Diffraction Analysis and a Bond Valence Theory Approach. Phys. Chem. Miner. 2019, 46, 815–825. [Google Scholar] [CrossRef]
  19. Pesquera, A.; Gil-Crespo, P.P.; Torres-Ruiz, F.; Torres-Ruiz, J.; Roda-Robles, E. A Multiple Regression Method for Estimating Li in Tourmaline from Electron Microprobe Analyses. Mineral. Mag. 2016, 80, 1129–1133. [Google Scholar] [CrossRef]
  20. Yavuz, F.; Karakaya, N.; Yıldırım, D.K.; Karakaya, M.Ç.; Kumral, M. A Windows Program for Calculation and Classification of Tourmaline-Supergroup (IMA-2011). Comput. Geosci. 2014, 63, 70–87. [Google Scholar] [CrossRef]
  21. Šontevska, V.; Jovanovski, G.; Makreski, P. Minerals from Macedonia. Part XIX. Vibrational Spectroscopy as Identificational Tool for Some Sheet Silicate Minerals. J. Mol. Struct. 2007, 834–836, 318–327. [Google Scholar] [CrossRef]
  22. Li, W.; Wu, R.; Dong, Y. Study on Infrared Spectra and Infrared Radiation Characteristics of Tourmaline. Geol. J. China Univ. 2008, 14, 426–432. [Google Scholar]
  23. Li, M. Spectroscopic Characteristics and Color Origin of Red Tourmaline from Brazil. J. Spectrosc. 2022, 2022, 1–6. [Google Scholar] [CrossRef]
  24. Makreski, P.; Jovanovski, G. Minerals from Macedonia. Spectrochim. Acta Part A Mol. Biomol. Spectrosc. 2009, 73, 460–467. [Google Scholar] [CrossRef]
  25. Fuchs, Y.; Lagache, M.; Linares, J. Oxydation expérimentale de Fe-tourmalines et corrélation avec une déprotonation des groupes hydroxyle. Comptes Rendus Geosci. 2002, 334, 245–249. [Google Scholar] [CrossRef]
  26. Prasad, P.S.R. Study of Structural Disorder in Natural Tourmalines by Infrared Spectroscopy. Gondwana Res. 2005, 8, 265–270. [Google Scholar] [CrossRef]
  27. Ertl, A.; Rossman, G.R.; Hughes, J.M.; London, D.; Wang, Y.; O’Leary, J.A.; Dyar, M.D.; Prowatke, S.; Ludwig, T.; Tillmanns, E. Tourmaline of the Elbaite-Schorl Series from the Himalaya Mine, Mesa Grande, California: A Detailed Investigation. Am. Mineral. 2010, 95, 24–40. [Google Scholar] [CrossRef]
  28. Maneewong, A.; Seong, B.S.; Shin, E.J.; Kim, J.S.; Kajornrith, V. Color Change of Tourmaline by Heat Treatment and Electron Beam Irradiation: UV-Visible, EPR, and Mid-IR Spectroscopic Analyses. J. Korean Phys. Soc. 2016, 68, 83–92. [Google Scholar] [CrossRef]
  29. Castañeda, C.; Oliveira, E.F.; Gomes, N.; Soares, A.C.P. Infrared Study of OH Sites in Tourmaline from the Elbaite-Schorl Series. Am. Mineral. 2000, 85, 1503–1507. [Google Scholar] [CrossRef]
  30. Reddy, B.J.; Frost, R.L.; Martens, W.N.; Wain, D.L.; Kloprogge, J.T. Spectroscopic Characterization of Mn-Rich Tourmalines. Vib. Spectrosc. 2007, 44, 42–49. [Google Scholar] [CrossRef]
  31. Li, X.; Zu, E. Near-Infrared Spectrum Analysis of Cyclosilicates Gem Minerals. Bull. Chin. Ceram. Soc. 2016, 35, 1318–1321. [Google Scholar] [CrossRef]
  32. Peng, M.; Wang, H. A Study on the Vibrational Spectra of Water in Tourmaline. Acta Mineral. Sin. 1995, 15, 372–377. [Google Scholar] [CrossRef]
  33. Hoang, L.H.; Hien, N.T.M.; Chen, X.B.; Minh, N.V.; Yang, I. Raman Spectroscopic Study of Various Types of Tourmalines. J. Raman Spectrosc. 2011, 42, 1442–1446. [Google Scholar] [CrossRef]
  34. Hoang, L.H.; Hien, N.T.M.; Chen, X.-B.; Yang, I.S. Annealing Effect in Raman Scattering of Various Types of Tourmalines. J. Appl. Spectrosc. 2013, 79, 881–887. [Google Scholar] [CrossRef]
  35. Fantini, C.; Tavares, M.C.; Krambrock, K.; Moreira, R.L.; Righi, A. Raman and Infrared Study of Hydroxyl Sites in Natural Uvite, Fluor-Uvite, Magnesio-Foitite, Dravite and Elbaite Tourmalines. Phys. Chem. Miner. 2013, 41, 247–254. [Google Scholar] [CrossRef]
  36. Wesełucha-Birczyńska, A.; Natkaniec-Nowak, L. A Raman Microspectroscopic Study of Organic Inclusions in “Watermelon” Tourmaline from the Paprok Mine (Nuristan, Afghanistan). Vib. Spectrosc. 2011, 57, 248–253. [Google Scholar] [CrossRef]
  37. Chenguang, Y.; Ruifeng, K.; Zhenyu, X.; Guangle, Z.; Jianguo, L. Second derivative of Voigt function. Acta Phys. Sin. 2014, 63, 142–147. [Google Scholar] [CrossRef]
  38. Mattson, S.M.; Rossman, G.R. Fe2+-Ti4+ Charge Transfer in Stoichiometric Fe2+,Ti4+-Minerals. Phys. Chem. Miner. 1988, 16, 78–82. [Google Scholar] [CrossRef]
  39. Bosi, F.; Skogby, H.; Agrosi, G.; Scandale, E. Tsilaisite, NaMn3Al6(Si6O18)(BO3)3(OH)3OH, a New Mineral Species of the Tourmaline Supergroup from Grotta d’Oggi, San Pietro in Campo, Island of Elba, Italy. Am. Mineral. 2012, 97, 989–994. [Google Scholar] [CrossRef]
  40. Rossman, G.R. Optical Spectroscopy. Rev. Mineral. Geochem. 2014, 78, 371–398. [Google Scholar] [CrossRef]
  41. Pezzotta, F.; Laurs, B.M. Tourmaline: The Kaleidoscopic Gemstone. Elements 2011, 7, 333–338. [Google Scholar] [CrossRef]
  42. Schwarzinger, C.; Wildner, M.; Ulatowski, S.; Sawyer, M. Vanadium-Bearing Tourmaline from the Commander Mine, Nadonjukin, Tanzania. J. Gemmol. 2019, 36, 534–543. [Google Scholar] [CrossRef]
  43. Cui, L.; Guo, Y.; Tang, J.; Yang, Y. Spectroscopy Characteristics and Color-Influencing Factors of Green Iron-Bearing Elbaite. Crystals 2023, 13, 1461. [Google Scholar] [CrossRef]
  44. Laurs, B.M.; Simmons, W.B.; Rossman, G.R.; Fritz, E.A.; Koivula, J.I.; Anckar, B.; Falster, A.U. Yellow Mn-Rich Tourmaline From The Canary Mining Area, Zambia. Gems Gemol. 2007, 43, 314–331. [Google Scholar] [CrossRef]
  45. Thongnopkun, P.; Naowabut, P. Effect of Heat Treatment on Madagascar Dravite Tourmaline: UV-Visible and Diffuse Reflectance Infrared Spectroscopic Characterization. J. Appl. Spectrosc. 2018, 85, 616–623. [Google Scholar] [CrossRef]
  46. Kurtz, D.; Rossman, G.; Hunter, B. The Nature of the Mn(III) Color Centers in Elbaite Tourmalines. Inorg. Chem. 2020, 59, 9618–9626. [Google Scholar] [CrossRef]
  47. Bosi, F.; Celata, B.; Skogby, H.; Hålenius, U.; Tempesta, G.; Ciriotti, M.E.; Bittarello, E.; Marengo, A. Mn-Bearing Purplish-Red Tourmaline from the Anjanabonoina Pegmatite, Madagascar. Mineral. Mag. 2021, 85, 242–253. [Google Scholar] [CrossRef]
  48. Ertl, A.; Hughes, J.M.; Prowatke, S.; Ludwig, T.; Lengauer, C.L.; Meyer, H.-P.; Giester, G.; Kolitsch, U.; Prayer, A. Alumino-Oxy-Rossmanite from Pegmatites in Variscan Metamorphic Rocks from Eibenstein an Der Thaya, Lower Austria, Austria: A New Tourmaline That Represents the Most Al-Rich End-Member Composition. Am. Mineral. 2022, 107, 157–166. [Google Scholar] [CrossRef]
  49. Suwanmanee, W.; Sutthirat, C.; Wanthanachaisaeng, B.; Utapong, T. Colour Enhancement of Pink Tourmaline from Nigeria by Electron-Beam and Gamma Irradiation. J. Gemmol. 2021, 37, 514–526. [Google Scholar] [CrossRef]
  50. Mattson, S.M.; Rossman, G.R. Fe2+-Fe3+ Interactions in Tourmaline. Phys. Chem. Miner. 1987, 14, 163–171. [Google Scholar] [CrossRef]
Crystals 14 00877 g001
Figure 1. Polyhedron model of tourmaline crystal structure.
Figure 1. Polyhedron model of tourmaline crystal structure.
Crystals 14 00877 g001
Crystals 14 00877 g002
Figure 2. Ten samples of blue tourmaline.
Figure 2. Ten samples of blue tourmaline.
Crystals 14 00877 g002
Crystals 14 00877 g003
Figure 3. Inclusions of blue tourmaline samples. (a) Fingerprint gas–liquid inclusions of S01 (×80); (b) nearly parallel fractures in the perpendicular c-axis of S05 and tubular inclusions in parallel to the c-axis (×32); (c) gas–liquid inclusions of S10 (×48).
Figure 3. Inclusions of blue tourmaline samples. (a) Fingerprint gas–liquid inclusions of S01 (×80); (b) nearly parallel fractures in the perpendicular c-axis of S05 and tubular inclusions in parallel to the c-axis (×32); (c) gas–liquid inclusions of S10 (×48).
Crystals 14 00877 g003
Crystals 14 00877 g004
Figure 4. Classification of the chemical composition of blue tourmaline samples. (a) Classification of primary tourmaline groups in the ternary system based on the dominant occupancy at the X-site; (b) classification of general tourmaline species in the ternary system based on the anion occupancy of the W-site; (c) classification of alkali-group tourmaline species in the ternary subsystem; (d) classification of alkali-group tourmaline fluorine species in the ternary subsystem. The above tourmaline composition classification figure was generated by the Wintcac program [20] with the assistance of processing.
Figure 4. Classification of the chemical composition of blue tourmaline samples. (a) Classification of primary tourmaline groups in the ternary system based on the dominant occupancy at the X-site; (b) classification of general tourmaline species in the ternary system based on the anion occupancy of the W-site; (c) classification of alkali-group tourmaline species in the ternary subsystem; (d) classification of alkali-group tourmaline fluorine species in the ternary subsystem. The above tourmaline composition classification figure was generated by the Wintcac program [20] with the assistance of processing.
Crystals 14 00877 g004
Crystals 14 00877 g005
Figure 5. IR spectra of 10 tourmaline samples: (a) mid-infrared spectrum; (b) near-infrared spectrum.
Figure 5. IR spectra of 10 tourmaline samples: (a) mid-infrared spectrum; (b) near-infrared spectrum.
Crystals 14 00877 g005
Crystals 14 00877 g006
Figure 6. Positive correlation between OH(3)-ν3 vibrational band wave number(cm−1) and increasing FeO + MnO (wt.%) content of samples.
Figure 6. Positive correlation between OH(3)-ν3 vibrational band wave number(cm−1) and increasing FeO + MnO (wt.%) content of samples.
Crystals 14 00877 g006
Crystals 14 00877 g007
Figure 7. Raman spectra of 10 tourmaline samples.
Figure 7. Raman spectra of 10 tourmaline samples.
Crystals 14 00877 g007
Crystals 14 00877 g008
Figure 8. UV-VIS absorption spectra of 10 tourmaline samples: (a) UV-VIS spectra; (b) second-order derivative spectra of UV-VIS spectra.
Figure 8. UV-VIS absorption spectra of 10 tourmaline samples: (a) UV-VIS spectra; (b) second-order derivative spectra of UV-VIS spectra.
Crystals 14 00877 g008
Table 1. Details of 10 tourmaline samples.
Table 1. Details of 10 tourmaline samples.
SampleColorSize (cm3)Weight (ct)Source
S01Greyish blue2.22 × 0.93 × 0.308.44Cruzeiro Mine—Vein 1, Governador Valadares
S02Greyish blue3.25 × 1.46 × 0.2415.46Cruzeiro Mine—Vein 1, Governador Valadares
S03Light blue2.90 × 1.20 × 0.269.12Golconda Mine, Governador Valadares
S04Light blue1.88 × 1.20 × 0.246.37Golconda Mine, Governador Valadares
S05Bluish green1.92 × 1.21 × 0.288.47Cruzeiro Mine—Vein 2, Governador Valadares
S06Bluish green1.84 × 1.56 × 0.268.25Cruzeiro Mine—Vein 2, Governador Valadares
S07Dark greyish blue2.56 × 1.91 × 0.3213.51Cruzeiro Mine—Vein 2, Governador Valadares
S08Dark greyish blue2.36 × 1.28 × 0.3510.9Rubelita mining District, Aracuaí
S09Dark greyish blue2.11 × 1.20 × 0.328.77Rubelita mining District, Aracuaí
S10Dark blue0.98 × 0.94 × 0.283.12Rubelita mining District, Aracuaí
Table 2. Gemological characteristics of 10 tourmaline samples.
Table 2. Gemological characteristics of 10 tourmaline samples.
SampleColorPleochroismSpecific
Gravity		Refractive
Index		Double
Refraction		UV Fluorescence
S01Greyish blueWeak3.071.6201.6400.020inert
S02Greyish blueWeak3.061.6201.6400.020inert
S03Light blueStrong—light blue/grey2.941.6201.6400.020inert
S04Light blueStrong—light blue/grey2.931.6201.6390.019inert
S05Bluish greenStrong—light blue/greenish blue3.061.6221.6420.020inert
S06Bluish greenStrong—light blue/greenish blue3.071.6221.6420.020inert
S07Dark greyish blueStrong—light blue/blue3.101.6201.6400.020inert
S08Dark greyish blueStrong—light green/blue3.071.6221.6420.020inert
S09Dark greyish blueStrong—light green/blue3.071.6201.6400.020inert
S10Dark blueStrong—blue/dark blue3.091.6221.6420.020inert
Table 3. Chemical composition of 10 tourmaline samples by EPMA.
Table 3. Chemical composition of 10 tourmaline samples by EPMA.
S01S02S03S04S05S06S07S08S09S10
Major oxide (wt.%) analyses
SiO2 37.462 37.537 37.666 36.652 37.119 37.345 36.467 36.669 37.692 36.083
TiO20.046 0.000 0.000 0.006 0.000 0.008 0.000 0.000 0.000 0.016
Al2O339.336 39.668 38.611 38.246 36.577 37.448 35.742 37.138 38.008 36.169
V2O30.022 0.000 0.013 0.015 0.012 0.025 0.017 0.012 0.025 0.000
Cr2O30.023 0.054 0.000 0.034 0.015 0.000 0.000 0.036 0.000 0.009
FeO0.441 0.547 1.510 1.431 3.896 3.314 4.047 2.256 2.370 4.906
MnO2.084 2.129 0.850 0.827 1.926 2.171 2.131 2.433 2.563 1.207
ZnO0.000 0.053 0.172 0.150 0.058 0.013 0.067 0.000 0.039 0.064
CuO0.000 0.000 0.021 0.000 0.000 0.024 0.000 0.006 0.060 0.006
MgO0.016 0.000 0.000 0.006 0.042 0.038 0.023 0.000 0.000 0.033
CaO0.105 0.117 0.363 0.259 0.309 0.175 0.200 0.449 0.437 0.259
PbO0.000 0.115 0.013 0.000 0.000 0.000 0.076 0.000 0.000 0.000
Na2O1.980 1.956 2.063 2.050 2.494 2.588 2.359 2.118 2.333 2.408
K2O0.019 0.054 0.004 0.013 0.047 0.031 0.023 0.017 0.016 0.060
Li2O *1.826 1.768 1.948 1.884 1.527 1.514 1.488 1.614 1.594 1.385
F0.852 0.791 0.716 0.782 0.682 0.882 0.691 0.654 0.846 0.684
H2O **2.998 3.031 3.118 3.020 3.161 3.086 3.078 3.085 3.095 3.071
B2O3 **10.815 10.868 10.786 10.574 10.625 10.739 10.424 10.555 10.838 10.414
O = F0.359 0.333 0.301 0.329 0.287 0.371 0.291 0.275 0.356 0.288
Total97.66698.35697.55495.62298.20599.03296.54396.76799.56096.486
Normalization: Cations (apfu) based on 31 anions
Si6.020 6.003 6.069 6.024 6.072 6.044 6.080 6.038 6.044 6.022
Ti0.006 0.000 0.000 0.001 0.000 0.001 0.000 0.000 0.000 0.002
Al7.450 7.476 7.333 7.409 7.051 7.143 7.024 7.207 7.183 7.114
V0.003 0.000 0.002 0.002 0.002 0.003 0.002 0.002 0.003 0.000
Cr0.003 0.007 0.000 0.004 0.002 0.000 0.000 0.005 0.000 0.001
Fe2+0.059 0.073 0.203 0.197 0.533 0.449 0.564 0.311 0.318 0.685
Mn2+0.284 0.288 0.116 0.115 0.267 0.298 0.301 0.339 0.348 0.171
Zn0.000 0.006 0.020 0.018 0.007 0.002 0.008 0.000 0.005 0.008
Cu0.000 0.000 0.003 0.000 0.000 0.003 0.000 0.001 0.007 0.001
Mg0.004 0.000 0.000 0.001 0.010 0.009 0.006 0.000 0.000 0.008
Ca0.018 0.020 0.063 0.046 0.054 0.030 0.036 0.079 0.075 0.046
Pb0.000 0.005 0.001 0.000 0.000 0.000 0.003 0.000 0.000 0.000
Na0.617 0.606 0.645 0.653 0.791 0.812 0.763 0.676 0.725 0.779
K0.004 0.011 0.001 0.003 0.010 0.006 0.005 0.004 0.003 0.013
Li1.180 1.137 1.262 1.246 1.005 0.985 0.997 1.068 1.028 0.930
F0.433 0.400 0.365 0.406 0.353 0.451 0.364 0.341 0.429 0.361
B2.940 2.942 2.936 2.937 2.949 2.950 2.949 2.946 2.948 2.953
* Li2O (wt.%) calculated based on multiple linear regression model [19]. ** H2O and B2O3 (wt.%) calculated by the Wintcac program based on stoichiometric equations [20].
Table 4. Raman frequencies (cm−1) for OH-stretching modes and their corresponding short-range occupation assignment.
Table 4. Raman frequencies (cm−1) for OH-stretching modes and their corresponding short-range occupation assignment.
SampleOH(3)-ν3OH(3)-ν1OH(1)-ν2
S013486
(Al,Ti) YAlZAlZ		3589
MgYAlZAlZ		3653
□X
S023483
AlYAlZAlZ		3588
MgYAlZAlZ		3653
□X
S033476
AlYAlZAlZ		3588
MgYAlZAlZ		3652
□X
S043480
AlYAlZAlZ		3588
MgYAlZAlZ		3655
□X
S053493
FeYAlZAlZ		3563
NaX		3595
MgYAlZAlZ
S063495
FeYAlZAlZ		3563
NaX		3596
MgYAlZAlZ
S073495
FeYAlZAlZ		3564
NaX		3595
MgYAlZAlZ
S083510
(Fe,Mg)YAlZAlZ		3583
MgYAlZAlZ		3612
(Fe,Mg)YAlZAlZ
S093503
FeYAlZAlZ		3578
MgYAlZAlZ		3605
(Fe,Mg)YAlZAlZ
S103498
FeYAlZAlZ		3566
NaX		3599
MgYAlZAlZ
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Chen, Y.; Xu, D.; Zhou, Z.; Schwarz, D.; Zheng, J.; Zhang, L. Chemical Composition and Spectral Variation in Gem-Quality Blue Iron-Bearing Tourmaline from Brazil. Crystals 2024, 14, 877. https://doi.org/10.3390/cryst14100877

AMA Style

Chen Y, Xu D, Zhou Z, Schwarz D, Zheng J, Zhang L. Chemical Composition and Spectral Variation in Gem-Quality Blue Iron-Bearing Tourmaline from Brazil. Crystals. 2024; 14(10):877. https://doi.org/10.3390/cryst14100877

Chicago/Turabian Style

Chen, Yifang, Duo Xu, Zhengyu Zhou, Dietmar Schwarz, Junhao Zheng, and Lingmin Zhang. 2024. "Chemical Composition and Spectral Variation in Gem-Quality Blue Iron-Bearing Tourmaline from Brazil" Crystals 14, no. 10: 877. https://doi.org/10.3390/cryst14100877

APA Style

Chen, Y., Xu, D., Zhou, Z., Schwarz, D., Zheng, J., & Zhang, L. (2024). Chemical Composition and Spectral Variation in Gem-Quality Blue Iron-Bearing Tourmaline from Brazil. Crystals, 14(10), 877. https://doi.org/10.3390/cryst14100877

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop