Rich Tourmaline Research Articles

Structural nature of pyroelectric effect revisited: Experimental and theoretical studies of synthetic Ni,Al - rich tourmaline

Tourmaline, the first pyroelectric material discovered, is also a promising borosilicate material with non-linear optical and luminescence properties, high thermal stability, high hardness and variable chemical composition. The structural nature of pyroelectricity in tourmaline is a subject of scientific debate since it requires precise structural studies and measurements at elevated temperatures. Here, an in situ high temperature single-crystal and powder X-ray diffraction studies (up to ∼1000 °C) and the direct pyroelectric measurements (up to ∼200 °C) of synthetic Ni, Al-rich tourmaline were performed. The pyroelectric coefficient was calculated from the structural data. The results demonstrate that polyhedra XO9, YO6 and ZO6 exert the greatest influence on the pyroelectric effect. The pyroelectric coefficient of tourmaline is practically independent of boron-oxygen triangles and silicon-oxygen tetrahedra. The primary pyroelectric coefficient of tourmalines could be calculated from single crystal structural data. This study contributes to the synthesis of tourmalines with the required pyroelectric properties and may be used for the study of the nature and properties of other pyroelectric materials.

Fluor-schorl, a new member of the tourmaline supergroup, and new data on schorl from the cotype localities

Fluor-schorl, NaFe^(2+) _3Al_6Si_6O_(18)(BO_3)_3(OH)_3F, is a new mineral species of the tourmaline supergroup from alluvial tin deposits near Steinberg, Zschorlau, Erzgebirge (Saxonian Ore Mountains), Saxony, Germany, and from pegmatites near Grasstein (area from Mittewald to Sachsenklemme), Trentino, South Tyrol, Italy. Fluor-schorl was formed as a pneumatolytic phase and in high-temperature hydrothermal veins in granitic pegmatites. Crystals are black (pale brownish to pale greyish-bluish, if distance (r^2 = 0.93). This correlation indicates that Fe^(2+)-rich tourmalines from the investigated localities clearly tend to have a F-rich or F-dominant composition. A further strong positive correlation (r^2 = 0.82) exists between the refined F content and the Y–W (F,OH) distance, and the latter may be used to quickly estimate the F content.

Crystal structure and stability of Ni-rich synthetic tourmaline. Distribution of divalent transition-metal cations over octahedral positions

AbstractThe crystal structure of synthetic Ni-rich tourmaline with a nickel content of 18.96 wt.% NiO (a = 15.890(2), c = 7.1815(8) Å) has been refined to R = 3.1% using single-crystal X-ray diffraction data. It has been shown that Ni is distributed not only over the Y, but also over the Z sites according to the ideal formula Na(Ni22+Al)(Al5Ni2+)(Si6O18)(BO3)3(OH)4. Based on bond valences and charge balance analysis it has been shown that this composition corresponds to the stable disordered member of the solid solution series NaAl3Al6(Si6O18)(BO3)3(O)3(OH) NaNi32+Al6(Si6O18)(BO3)3(OH)4.Taking into consideration available structural data on Me2+ (Ni, Cu, Co, Fe, Mn)-rich tourmalines, cation-size mismatch and bond-valence calculations we assume that there are no structural constraints precluding occurrence of these cations in both octahedral sites. Divalent transitional metal-rich tourmalines have not been found in Nature probably due to the insufficient concentration of these elements in B-rich geological systems.

A critical comment on Ertl et al. (2012): "Limitations of Fe2+ and Mn2+ site occupancy in tourmaline: Evidence from Fe2+- and Mn2+-rich tourmaline"

In this paper we have presented a detailed response to Ertl et al. (2012a) who, in a paper in volume 97 (year 2012), pages 1402–1416, this journal, claim evidence for limitations of Fe2+ and Mn2+ occupancy at the Z site of the tourmaline structure. They also propose a model by which the distance of tourmaline varies as a function of its and bond lengths. We have examined their conclusions and find that a different distribution of cations over the Y and Z sites gives better agreement with the extensive experimental information available. In fact, on the basis of crystal-structure refinements, Mossbauer spectroscopy, optical absorption spectroscopy, bond-valence theory, ionic radius concept and literature, the occurrence of Fe2+ at the Z site of tourmaline is well supported. Conversely, existing experimental data does not provide indisputable evidence for the occurrence of Mn2+ at the Z site. Despite this, there is no evidence for inductive effects of YMn2+ on , and the proposed effects must be regarded as speculative. Statistical analysis shows that the average value is 1.906(2) A, which is consistent with the observed values of at the 99% confidence limit (within 3σ) in tourmalines with the Z site fully occupied by Al. Consequently, the proposed inductive effect of and on can be ruled out.

A Fluid Inclusion and Stable Isotope Study at the Loulo Mining District, Mali, West Africa: Implications for Multifluid Sources in the Generation of Orogenic Gold Deposits

The world-class Loulo mining district (15.5 Moz resource), in the Birimian terrane of western Mali, contains a range of mineralogically diverse styles of orogenic gold mineralization. The district is distinguished by As-rich orebodies as at Yalea, as well as tourmaline-bearing, Fe-rich, orebodies as at Gara. New fluid inclusion and stable isotope data presented here constrain the nature of the hydrothermal fluids responsible for these different types of mineralization, and point towards the role of multifluid sources (metamorphic and magmatic) in the formation of orogenic gold deposits. Microthermometric and laser Raman studies from Yalea Main and two other similar orebodies (Loulo-3 and Baboto) reveal the dominance of coexisting CO 2 -N 2 ±CH 4 (type 1) and H 2 O-NaCl (type 2) fluid inclusions. These inclusions show evidence of fluid unmixing from reduced (quartz-faylite-magnetite, QFM, buffered), low-salinity (≤10 wt % NaCl equiv), CO 2 -rich-H 2 O-NaCl-N 2 ± CH 4 primary ore fluids. The combination of microthermometric data and geothermometry based on ore and alteration assemblages indicate trapping temperatures and pressures of 270° to 340°C at 1.4 to 1.8 kbar. The P-T-X state of the Yalea-style mineralizing fluids, along with the δ 18 O fluid compositions of 8.8 to 10.7‰, is consistent with the derivation of auriferous fluids during greenschist facies regional metamorphism of the host terrane. Similar fluid compositions were previously reported elsewhere in the Birimian crust of West Africa and in other orogenic gold districts worldwide. The precipitation of gold from the H 2 S-rich metamorphic fluid is primarily linked to phase separation of the ore fluid, which is controlled by P-T fluctuations and/or fluid-rock interaction with carbonaceous host sedimentary rocks (confirmed by low δ 13 C values of −21.7 to −15.8‰). Fluid inclusion investigations from Gara and a similar style orebody (Yalea North) indicate the presence of coexisting CO 2 ± N 2 ± CH 4 (type 1) and mixed-salinity (5–21 wt % NaCl equiv) CO 2 -rich-H 2 O-NaCl inclusions (type 3). Inclusion assemblages also contain common oxidized (hematite-magnetite, HM, buffered), high temperature (>400°C), hypersaline (∼35–50 wt % total dissolved solids), metalliferous (Na+Fe+Ca+Cu+Ni+W+ Pb+Zn), multiphase H 2 O-rich-CO 2 -NaCl-FeCl 2 inclusions (type 4). This inclusion type has not been previously reported in other Birimian terranes. The composition of the brines, along with carbon isotope data ( δ 13 C of −14.4 to −4.5‰), suggests a magmatic input to the Gara-style hydrothermal system. The coexistence in the fluid inclusion assemblages of the magmatic brines with Yalea-style, CO 2 -rich, metamorphic fluids and the positive correlation between salinity and homogenization temperatures suggest mineralization was locally controlled by fluid mixing. The interaction of these two chemically contrasting fluids explains the distinctive petrographic characteristics of the Gara-style orebodies. This includes the growth of widespread multi-stage Fe 3+ -rich tourmaline (B-rich granite source) and sodic alteration, and ore assemblages consisting of abundant nickeloan pyrite, (REE)-phosphates, Ni ± Co ± Pb ± Zn minor/trace sulfides, and scheelite. Gold deposition in the Gara-style hydrothermal system is related to physical and chemical changes of the two fluids during mixing (e.g., decreases in f O2 and T in the brines and retrograde boiling of the CO 2 component in the metamorphic fluids, a “salting out effect”).

Limitations of Fe2+ and Mn2+ site occupancy in tourmaline: Evidence from Fe2+- and Mn2+-rich tourmaline

Fe 2+ - and Mn 2+ -rich tourmalines were used to test whether Fe 2+ and Mn 2+ substitute on the Z site of tourmaline to a detectable degree. Fe-rich tourmaline from a pegmatite from Lower Austria was characterized by crystal-structure refinement, chemical analyses, and Mössbauer and optical spectroscopy. The sample has large amounts of Fe 2+ (~2.3 apfu), and substantial amounts of Fe 3+ (~1.0 apfu). On basis of the collected data, the structural refinement and the spectroscopic data, an initial formula was determined by assigning the entire amount of Fe 3+ (no delocalized electrons) and Ti 4+ to the Z site and the amount of Fe 2+ and Fe 3+ from delocalized electrons to the Y-Z ED doublet (delocalized electrons between Y-Z and Y-Y): X (Na 0.9 Ca 0.1 ) Y (Fe 2+ 2.0 Al 0.4 Mn 2+ 0.3 Fe 3+ 0.2 ) Z (Al 4.8 Fe 3+ 0.8 Fe 2+ 0.2 Ti 4+ 0.1 ) T (Si 5.9 Al 0.1 )O 18 (BO 3 ) 3 V (OH) 3 W [O 0.5 F 0.3 (OH) 0.2 ] with a = 16.039(1) and c = 7.254(1) Å. This formula is consistent with lack of Fe 2+ at the Z site, apart from that occupancy connected with delocalization of a hopping electron. The formula was further modified by considering two ED doublets to yield: X (Na 0.9 Ca 0.1 ) Y (Fe 2+ 1.8 Al 0.5 Mn 2+ 0.3 Fe 3+ 0.3 ) Z (Al 4.8 Fe 3+ 0.7 Fe 2+ 0.4 Ti 4+ 0.1 ) T (Si 5.9 Al 0.1 )O 18 (BO 3 ) 3 V (OH) 3 W [O 0.5 F 0.3 (OH) 0.2 ]. This formula requires some Fe 2+ (~0.3 apfu) at the Z site, apart from that connected with delocalization of a hopping electron. Optical spectra were recorded from this sample as well as from two other Fe 2+ -rich tourmalines to determine if there is any evidence for Fe 2+ at Y and Z sites. If Fe 2+ were to occupy two different 6-coordinated sites in significant amounts and if these polyhedra have different geometries or metal-oxygen distances, bands from each site should be observed. However, even in high-quality spectra we see no evidence for such a doubling of the bands. We conclude that there is no ultimate proof for Fe 2+ at the Z site, apart from that occupancy connected with delocalization of hopping electrons involving Fe cations at the Y and Z sites. A very Mn-rich tourmaline from a pegmatite on Elba Island, Italy, was characterized by crystal-structure determination, chemical analyses, and optical spectroscopy. The optimized structural formula is X (Na 0.6 □ 0.4 ) Y (Mn 2+ 1.3 Al 1.2 Li 0.5 ) Z Al 6 T Si 6 O 18 (BO 3 ) 3 V (OH) 3 W[F 0.5 O 0.5 ], with a = 15.951(2) and c = 7.138(1) Å. Within a 3σ error there is no evidence for Mn occupancy at the Z site by refinement of Al ↔ Mn, and, thus, no final proof for Mn 2+ at the Z site, either. Oxidation of these tourmalines at 700-750 °C and 1 bar for 10-72 h converted Fe 2+ to Fe 3+ and Mn 2+ to Mn 3+ with concomitant exchange with Al of the Z site. The refined Z Fe content in the Fe-rich tourmaline increased by ~40% relative to its initial occupancy. The refined Y Fe content was smaller and the <Y-O> distance was significantly reduced relative to the unoxidized sample. A similar effect was observed for the oxidized Mn 2+ -rich tourmaline. Simultaneously, H and F were expelled from both samples as indicated by structural refinements, and H expulsion was indicated by infrared spectroscopy. The final species after oxidizing the Fe 2+ -rich tourmaline is buergerite. Its color had changed from blackish to brown-red. After oxidizing the Mn 2+ -rich tourmaline, the previously dark yellow sample was very dark brown-red, as expected for the oxidation of Mn 2+ to Mn 3+ . The unit-cell parameter a decreased during oxidation whereas the c parameter showed a slight increase.

Iron redox reactions in the tourmaline structure: High-temperature treatment of Fe3+-rich schorl

We present a detailed study of thermally-driven oxidation and reduction of Fe in the structure of schorl (as the most widespread tourmaline), Fe2+-bearing olenite and fluor-schorl. The principal Fe3+-rich tourmaline investigated in this study is a natural schorl sample from a unique occurrence in peralkaline rocks near Cancrinite Hill, east of Bancroft, southern Ontario. Tourmaline samples were thermally-treated in air and hydrogen at temperatures of 700, 800 and 900°C to oxidize or reduce the structural Fe. High-temperature changes were continuously monitored using 57Fe Mössbauer and infrared spectroscopy. Proportions of Fe2+ and Fe3+ vary as a function of the heat treatment. An increase in Fe3+ up to 100% after heating in air at 700°C was observed, whereas only small changes in the Fe3+/Fetot ratio after heating under hydrogen at 700°C was revealed. Partial deprotonation/protonation represents charge compensation for the oxidation/reduction of Fe at the Y and Z sites.Critical samples of Cancrinite Hill tourmaline were investigated in detail by means of 57Fe Mössbauer and infrared spectroscopy, single-crystal X-ray diffraction, electron microprobe and magnetometry. The optimized structural formulae are:X(Na0.93K0.02□0.05)Y(Ti0.13Al0.20Fe1.263+Fe1.102+Mn0.022+□0.30)Z(Al5.16Fe0.532+Mg0.31)BB3T(Si5.88Al0.12)O27V(OH)3W(O0.12OH0.88) – untreated tourmaline (schorl);X(Na0.93K0.02□0.05)Y(Ti0.13Al0.52Fe1.143+Fe0.842+Mg0.05Mn0.022+□0.30)Z(Al4.85Fe0.902+Mg0.25)BB3T(Si5.88Al0.12)O27V(OH)3W(OH) – tourmaline reduced in hydrogen (schorl);X(Na0.93K0.02□0.05)Y(Ti0.13Al0.40Fe2.003+Mg0.15Mn0.022+□0.30)Z(Al4.99Fe0.873+Mg0.14)BB3T(Si5.90Al0.10)O27V(O1.05OH1.95)W(O0.70OH0.30) – tourmaline oxidized in air (H+-rich “buergerite”).There is evident disorder of Al over the Y, Z and T sites as well as disorder of Fe2+ over the Y and Z sites, and ordering of Fe3+ at the Y site and Mg at the Z site. The fully oxidized tourmaline shows disorder of Fe3+ and Mg over the Y and Z sites. The occurrence of “extra” Fe2+ in the Z site of the reduced tourmaline relative to the untreated sample demonstrates the intracrystalline cation-exchange YR2++ZR3+⇆YR3++ZR2+ driven by elevated temperature. Increased disorder of Fe cations over the edge-shared YO6 and ZO6 octahedra enhances the antiferromagnetic exchange interactions in the tourmaline structure, which are stronger in reduced samples than in oxidized and untreated ones. This suggests that OH groups at the W site could mediate Y–Z inter-site exchange interactions.Complete oxidation of Fe within the tourmaline structure is possible. However, significant reduction of Fe cannot occur because an excess of H+ may not be incorporated within the tourmaline structure. Further reduction of Fe3+ will not occur until the breakdown of the tourmaline structure above its temperature of reductive decomposition, where metallic Fe immediately appears as a separate phase (α-Fe) together with cristobalite and Na–Al–Fe-borosilicate glass. The scarcity of Fe3+-rich tourmalines in nature and the separate existence of “buergerite” as the only tourmaline with almost all Fe as Fe3+ support its specific origin, such as from an HT-LP, high fO2 overprint.

Ferric-iron-rich tourmaline from the Darasun gold deposit, Transbaikalia, Russia

We have studied tourmaline of the schorl – dravite – povondraite series in a sample from the Darasun porphyry-style gold deposit, Chita region, Transbaikalia, Russia. A combination of analytical procedures has been used to determine composition in terms of major elements (electron microprobe), Fe 2+ and Fe 3+ (Mossbauer spectroscopy), the probable position of Fe 3+ in the Fe-rich core of a separated grain (FTIR), the Li content (ICP–MS), and conditions of formation (fluid inclusions in coeval quartz). A significant amount of Fe 3+ has been directly determined by Mossbauer spectroscopy (83% of iron is Fe 3+ over the averaged sample). The evolution of chemical composition of the tourmalines studied here can be generalized in the following model. The first generation of low-Al and high-Fe “oxy-dravite” – povondraite (Al~3, Fe~4, Ca~0.2, □ apfu ) is followed by a second generation of higher-Al and lower-Fe “oxy-dravite” (Al~6, Fe~1.5, Ca~0.1, □~0.2 apfu ), and the third generation consists of dravite having the highest Al and the lowest Fe contents (Al > 6, Fe apfu ). Depletion of the second and third generations in Fe is caused by deposition of early sulfides. The evolution in chemical composition of the tourmalines studied here is different from that of redeposited tourmalinites in Western Carpathians, where tourmaline evolves from schorl–dravite to povondraite. The chemical composition of the Darasun tourmalines is different from Fe 3+ -rich tourmalines from evaporite formations in low K, and is similar in Mg content. At the same time, the major type of homovalent substitution (AlFe −1 ) in the tourmalines examined is similar to that in tourmalines from Sn- and Cu-porphyry deposits, and may be accepted as a prospecting guide for the porphyry-style deposits. This finding of povondraite is the third in the world and the first in hydrothermal gold deposit. The fluid inclusions in the associated quartz indicate that that first-stage tourmaline precipitated from boiling fluid at low pressure. Boiling favored the increasing oxygen activity, and resulted in the increase of Fe 3+ in that tourmaline.

Disordering of Fe2+ over octahedrally coordinated sites of tourmaline

The partitioning of iron among octahedrally coordinated sites in tourmaline, and its stereochemical consequences, were investigated in a Fe-rich dravite in a skarn rock from Uto, Sweden. A multi-analytical approach using structure refinement (SREF), electron microprobe analysis (EMPA), and Mossbauer spectroscopy (MS) established the chemical and structural nature of the tourmaline. A structural formula obtained by optimization procedures indicates disordering of Al, Mg, and Fe 2+ over the Y and Z sites, and ordering of Fe 3+ at the Y site. Two Fe-rich tourmalines from the literature, reexamined with the optimizing site assignment procedure, appear to have iron partitioning comparable to that of the Uto tourmaline with Fe 2+ disordered over the octahedral sites. This is best explained by disordered Fe 2+ distributions that minimize the strain state of the Y -O bonds and provide a shielding effect reducing Y - Z repulsion. This is consistent with predictions from bond-valence theory and Pauling’s rules. An indication of Z -site occupancy by Fe 2+ in tourmaline may be signaled by a significant correlation between Z -O> and the c lattice parameter ( r 2 = 0.96). The c value for a very Fe 2+ -rich tourmaline and an ideal end-member schorl, with Fe 2+ and Al ordered at Y and Z (respectively), yielded Z -O> values larger than 1.907 A (the likely bond length for Z Al-O>). These large Z -O> lengths indicate that Fe 2+ occurs at the Z site. The hypothesis of a dragging effect from Y -O> to explain lengthening of Z Al-O> is not supported by experimental evidence.

The F-analogue of schorl from Grasstein, Trentino South Tyrol, Italy: crystal structure and chemistry

The F-analogue of schorl has been identified in samples from a pegmatite at Grasstein, Trentino-South Tyrol, Italy. The crystal chemistry of this tourmaline has been characterized by a combination of single-crystal structure refinement, chemical analysis, and Mossbauer spectroscopy, yielding the structural formula X (Na 0.78 K 0.01 □ 0.2l ) Y (Fe 2+ 1.89 Al 0.58 Fe 3+ 0.13 Mn 2+ 0.13 Ti 4+ 0.02 Mg 0.02 Zn 0.02 □ 0.21 ) Z (Al 5.74 Fe 3+ 0.26 ) T (Si 5.90 Al 0.10 O 18 ) (BO 3 ) 3 V (OH) 3 W [F 0.76 (OH) 0.24 ]; a = 15.997(2), c = 7.179(1) A, V = 1591.0(4) A 3 , R 1( F ) = 1.60 %. This F-rich and Fe 2+ -rich tourmaline, a pneumatolytic phase crystallized in the presence of a F-rich fluid (coexisting with fluorite), is very near the proposed end-member composition of the F-analogue of schorl: NaFe 3 2+ Al 6 Si 6 O 18 (BO 3 ) 3 (OH) 3 F. The relatively high amount of Fe 2+ at the Y site is consistent with the large distance of 2.056 A. Refinement ofthe F:O occupancy ratio at the W site yields F 0.8 O 0.2 pfu, consistent with the chemical data (F 0.76 apfu). Because ofthe local bonding ofthe W -site anion to three neighbouring Y -site cations and the X -site cation, the charge of the X-site cation should affect the F occupancy at the W site. The cation and anion occupancy of this tourmaline is consistent with observations that tourmalines not dominated by X -site vacancies can have high F concentrations in the W site if F is present in the coexisting fluid phase. It is thus likely that the occurrence of high amounts of F in Fe-rich tourmalines requires a significant percentage of Fe 3+ in the tourmaline structure.

Tourmaline chemistry in the Miocene and Paleozoic granites, Central Nepal Himalaya

Tourmalines are found in almost all the formations, from the Lesser Himalaya to the Tibetan-Tethys Himalaya. They are equally found in the Miocene and Paleozoic granites in Central Nepal. Aplite-pegmatite dykes and Miocene granite from the Higher Himalaya, Tibetan-Tethys Himalaya also contain the tourmaline.
 The chemical composition of tourmalines from the Miocene granite, Paleozoic granite, and aplite pegmatites is presented in schorlite (Fe)-dravite (Mg)-elbaite (Al) diagram. The tourmalines from the granites are schorlitic (rich in Fe) in composition. No compositional variation between rim and core of the tourmaline in augen gneiss of the Formation III of the Higher Himalaya could correspond to the recrystallization of the tourmaline during the Himalayan metamorphism. The tourmalines from Manaslu and Chhokang granites can be evolved from Mg rich composition in higher temperature towards the Fe rich in lower temperature during the crystallization reflected by the XFe variation. Similarly, the composition of tourmalines is found to evolve from dravite (Mg rich) in the aplite-pegmatite dykes of the Higher Himalaya to schorlite (Fe rich) and elbaite (Al) in the aplite-pegmatite dykes of the Annapurna Formation of the Tibetan-Tethys Himalaya and Manaslu granite.
 The aplite-pegmatite dykes intruded into the pelitic rocks contain Mg rich tourmalines while Fe rich tourmalines are found in calcareous host rocks. So, the composition of the tourmalines of aplite-pegmatites can not be controlled by the composition of host rocks, but it is indirectly controlled by the composition of fluid phase of magma.

Modelling of some structural parameters of tourmalines on the basis of their chemical composition. I. Ordered structure model

This study models the mean sizes of X, Y, Z, B and T sites as well as the unit-cell parameters of tourmalines on the basis of their chemical compositions, using previously determined relations between the mean bond lengths in co-ordination polyhedra and the cation populations in those sites. For the octahedral sites of Al-rich tourmalines, these relations also take into account the possible modifying influences of the IInd co-ordination shells. For Al-poor tourmalines, the following relationships were determined: Fe 3+ , Mg 2+ and Ti 4+ ions supplementing the deficit of aluminium located at the Z octahedra have an effect on the mean size of the Z site, and these ions have a modifying influence on the adjacent Y sites. For a set of 43 tourmalines from various localities, representing various members of this mineral group, the observed unit-cell parameters were compared with the calculated values. The standard errors SE of the models: a calc . = K a a obs. and C calc . = k c c obs . are 0.0 13 A and 0.008 A, respectively, and the k a , k c coefficients are close to the theoretical value of 1 within the limits of a statistical error se = 0.0001-0.0002. These values suggest a highly satisfactory agreement of the measured and modelled values and indicate that the determined model equations can be used to predict the structural parameters of tourmalines. On this basis, the model predicts the structural characteristics of all end-member species known in the tourmaline group, assuming the ordering of octahedral ions between Y and Z sites. In some (Fe 3 +,Al)-rich tourmalines (buergerites and schorls) and Al-rich dravites small differences have sometimes been noted between the observed and predicted values of and mean sizes, as well as the a and c unit-cell parameters; the differences suggest a limited disordering of Fe 3+ and Al or Mg and Al between Y and Z sites.

Ferric iron in tourmaline

Red Fe^(3+)-rich and black Fe^(3+), Fe^(2+)-rich tourmalines have been studied by optical and Mossbauer spectroscopies to determine the optical characteristics of Fe^(3+) in tourmaline. Prominent optical absorption features at 485 and 540 nm are assigned to transitions of multiple exchange-coupled Fe^(3+) pairs in several site combinations. These transitions are more intense than those of isolated Fe^(3+) and are polarized along the vector between the interacting ions, thus permitting site assignments. The 485 nm band occurs at an unusually low energy for Fe^(3+) in silicate minerals. Similar behavior has been observed in the spectrum of coalingite, Mg_(10)Fe_2^(3+) (OH)_(24)CO_(3)·2H_2O, in which Fe^(3+) occurs in related pairs in edge-shared sheets. These lower energies are proposed to result from magnetic exchange in edge-shared geometries. Antiferromagnetic exchange has been confirmed by a variable temperature magnetic susceptibility study of a Kenyan dravite with 3.36 wt percent Fe. The Mossbauer spectrum of this sample is unusual in that it shows a pronounced decrease in width of component peaks from 298 K to 5 K.