About chemistry of tourmaline from Mnogovershinnoe ore deposit (Khabarovsk Krai, Far East)

Relevance of the work is due to the need to study the mineralogy of gold deposits in the Russian Far East, information about which is extremely scarce. Purpose of the work: study of the chemical composition of tourmaline from Mnogovershinnoe ore deposit, Khabarovsk Krai (Far East). Methodology of research: The chemical composition of minerals and BSE images were obtained using a Jeol JSM-6480 electron microscope equipped with an Inca Energy-350 EDS (analyst is N. N. Koshlyakova, Department of Petrology, Lomonosov Moscow State University). Electron microscope shooting environment: accelerating voltage is 15 kV, measuring current for the sample is 30 ± 0.1 nA. XPP corrections were used for the adjustment procedure (INCA program, version 17a). Results. The obtained data show that tourmalines of the Mnogovershinnoe deposit differ in their chemical composition and type of substitution. All studied tourmalines by these parameters can be divided into two groups. Group 1 includes schorl, foitite, and pegmatoids feruvite, as well as schorl and foitite of the first generation, tourmaline-muscovite-quartz veinlet in sandstones. Group 2 includes schorl of the second generation, tourmaline-muscovite quartz veinlet, schorl and foitite of quartz-tourmaline metasomatites, and cement dravite of quartz breccia. Conclusions. Tourmalines of the post-ore mineral associations of the Mnogovershinnoe gold deposit are divided into two groups characterized by different chemical composition and substitutions. Tourmalines of the first group with substitutions Fe ↔ Mg and X-vacancy + Al ↔ Na + R are confined to pegmatoids and were formed in reducing or weakly oxidative conditions. Later tourmalines of the second group with substitutions Fe ↔ Al and Al + O ↔ R+ OH indicate a possible porphyry-style mineralization and its formation during lowering oxidative potential.

Tourmalines are widespread in some types of gold and gold-polymetallic deposits: porphyry gold (Darasun, Kariyskoye), intrusion related, where they are constitute of barren propylite (Berezovskoye, the middle Urals; Berezitovoe, Transbaikal Krai), hydrothermal-metamorphic or orogenic (Hutti, India; Loulo, Mali; Sigma, Canada). Tourmaline is not typical for igneous gold deposits. However, it is found in some sites where its formation is associated with pre-ore or post-ore processes. is type of deposits includes Mnogovershinnoe in the Khabarovsk Krai, in which the formation of tourmaline is associated with hydrothermal activity due to the intrusion of post-mineral granitoids [5].
Despite the history of half a century for studying the eld, information about tourmaline is scarce, and data on the chemical composition are not available. erefore, the purpose of this study is to characterize the chemical composition of tourmaline to determine its formational identity and the assumption on this basis about the presence of possible new types of mineralization for the deposit.
Brief geology e Mnogovershinnoe gold deposit is located in the Nikolaevskiy region of the Khabarovsk Krai, 135 km north of Nikolaevsk-on-Amur and 17 km from the coast of the Sea of Okhotsk. It was discovered in 1959; the detailed exploration began in 1963, and acquisition -in 1991. e deposit is a part of the ore eld having the same name located at the northern termination of the East Sikhote-Alin volcanic belt at the joint of the Amgunsky and Gorinsky synclinoria of the Sikhote-Alin folded region [6,7]. e ore eld is con ned to the Ula volcano-plutonic structure. e ore eld is bounded by the Middle Ula and the Kulibinsky latitudinal deep faults from the north and south; the Malakhtinskaya tectonic depression from the northwest; the compound contact of the Bekchi-Ula granitoid pluton from the southeast. e structure of the ore eld is two-staged. e lower stage is formed by folded Lower Cretaceous sandstones, mudstones and siltstones. e upper stage is the Paleocene volcanic rocks dominated by vent and subvolcanic facies. e rocks of overlapped facies are almost completely eroded. Volcanic rocks are represented by lavas and tu s of andesites and andesidacites. Both structural complexes are intruded by the large Bekchi-Ula massif of granitoids, at the northwestern contact of which the Mnogovershinnoye deposit is located. e massif here is the Early Eocene porphyry monzogranodiorites of the rst phase. e second intrusive phase is leucocratic and pegmatoid granites. Sandy and clay rocks contacted with the pluton are transformed into hornfelses. In the southwestern part of the ore eld, subjacent intrusives of Eocene porphyritic quartz diorite and granodiorite-porphyries were eroded. e volcanic and intrusive rocks are intruded by the Eocene-Oligocene dikes of porphyritic diorite, andesite, basalt, and granite porphyry of north-west and latitudinal strike [5,8,9]. Photo: a -cross polars, without nicols (b-f) back scattered electrons. Рисунок 1. Взаимоотношения минералов пегматоидов участка Заманчивый. a-c -брекчированный оптически зональный кристалл шерла, отдельные секторы в центральной части которого обогащены Ca (до увита) и обеднены Na (до фойтита), трещины залечены альбит-мусковитовым агрегатом; d -альбит-мусковитовый прожилок в калишпате; e -срастание альбита и калишпата, в альбите отмечаются вростки калишпата; f -увеличенный фрагмент e, показывающий скопление мусковита. Образец МН-2. Фото. a -в проходящем свете, без николей; b-f -в отраженных электронах.     Note: bdl -below detection limits by the electron micrpoprobe. fe = Fe/(Fe + Mg), ca = Ca/(Ca + Na), vac = X-vacancy/(X-vacancy + Na). The range of content is given in parentheses. Примечание: bdl (н.п.о.) -ниже предела обнаружения электронно-зондовым методом; fe = Fe/(Fe + Mg), ca = Ca/(Ca + Na), vac = X-вакансия/(X-вакансия + Na). В скобках приведены пределы содержания.
age 15 kV, current intensity 30 ± 0.1 nA. XPP corrections were used for the correction procedure (INCA program, version 17a). The profile lines of characteristic X radiation are optimized and normalized using natural silicate standards [11].
Tourmalines formulas were calculated on the basis 15 cations, excluding Na, Ca, and K, which implies the absence of vacancies at tetrahedral and octahedral sites and a small amount of Li [12]. The amounts of OHand O 2 -at sites V and W are estimated from the charge balance constraints. It is assumed that oxygen ions O 2preferably occupies position W along with Fions [1]. The vacancy rate () in the X position is calculated by the stoichiometric ratio based on the equation 1 = Na + Ca + K + . The B 2 O 3 content is also calculated by stoichiometry. The muscovite formula is calculated for 22 negative charges.
Results Samples of tourmaline were selected within prospecting sites of Zamanchivy (sam. MN-2, MN-4, MN-7) located in the continuation of the ore body Valunistoe and Kulibinsky (MN-33). The sites are situated in the eastern and western flanks of the ore field, respectively. Samples were taken from pit bings and rock outcrops.
Pegmatoids (sample MN-2) are massive large-crystalline muscovite-quartz-feldspar rocks with tourmaline bunches. The size of the last ones is 2-5 mm.
Tourmaline assemblages interstices between quartz grains and cracks in them and forms relatively large crystals of up to several mm in size (Fig. 1a). It is pleochroic; from light mossy or mossy to dark mossy color. Optical zoning is very weak. Tourmaline crystals are sometimes brecciated and exhibit complex zoning in chemical composition. Rare sectors in the central part of such crystals are enriched in Ca (0.44-0.62 apfu) and Fe (1.75-2.48 apfu) (Fig. 1b, c). In addition, calculations show that the content of Fe 3+ in them is 0.23-0.27 apfu. Other sectors in the central part of crystals have a lower Ca content (0.05-0.16 apfu) and contain less Fe (1.45-1.54 apfu).
The crystal rims have a composition close to the low-calcium sectors in the central part differing slightly in Fe content (1.70 apfu). It is impossible to calculate Fe 3+ concentration in them and in low-calcium sectors. A special feature of tourmaline is a fairly high Zn (up to 0.12 wt.% ZnO); there are other impurities such as Mn (up to 0.50 wt.% MnO) and Ti (up to 1.0 wt.% TiO 2 ) ( Table 1).
In the triangular diagram in terms of X-vacancy -Ca -Na (+ K) (Fig. 2a), the tourmaline compositions fall into all three fields; taking into account Fe and Mg contents, it makes it possible to attribute these compositions to feruvite, schorl and foitite. On the other hand, the determination of the content of ferric iron can change classification.
Tourmaline-muscovite-quartz veinlet in hydrothermally altered sandstones (sam. . Tourmaline occurs as relatively large crystals with a thickness of 1-5 mm and a length of up to 12 mm, the fractures in which are healed with later tourmaline (Fig. 3a-c). The crystals are pleochroic from light orange to dark orange. The BSE images of such crystals show slightly osscillatory zoning (Fig. 3c, d) due to variable Fe and Mg contents (Fig. 1a, b, d; Table 1). Chemically, large crystals belong to the vacancy and alkali series (Fig. 2a). The Fe/(Fe + Mg) versus X / ( X + Na) diagram (Fig. 2b) shows that tourmalines are classified as foitite and schorl; just one composition -to dravite. The relationship between schorl and foitite is unclear.   Similar to pegmatoids, the described schorl is enriched in Zn (up to 0.17 wt.% ZnO). The Mn concentration (up to 0.33 wt.% MnO) and Ti (up to 1.2 wt.% TiO 2 ) in schorl and foitite (Table 1) is also comparable. The calculated formulas of foitite and veinlet schorl turned out to be balanced in charge; therefore, it is impossible to estimate Fe 3+ content in these minerals using calculations.
Blue-green tourmaline, which heals the fractures, can be considered as a second-generation tourmaline. Alongside, quartz and carbonate are found in the cracks. Chemically, tourmaline falls into the alkali field (Fig. 2a). The diagram (Fig. 2b) shows that the mineral is classified as schorl. Its iron content varies from 0.60 to 0.78; it contains 0.03-0.12 apfu Ti and 0.07-0.21 apfu. Сa (Table 1). Calculations indicate that tourmaline II is enriched in Fe 3+ (up to 0.19 a.f.) compared with earlier schorl and foitite. This may indicate more oxiditive conditions of its formation.
For early tourmaline, a significant negative correlation was established between the pairs of NaMg and X Al (r = -0.84) and between Fe and Mg (r = -0.82). This fact shows that the leading types of substitutions in the early schorl and foitite are X Na + Mg ↔ X + Al and Fe Mg with the minor substitution Al + O 2-↔ R + OH -, which is reflected in the diagrams shown in Fig. 2e, f. White mica composes colorless separations of two types differing in size: rosettes with a diameter up to several tenths of mm and small scales which rosettes are coated with (Fig. 3f, g) and replace primary plagioclase and potassium feldspar. Chemically, mica refers to muscovite ( Table 2, an. 5-8). The Fe and Mg concentrations range from 0.07 to 0.09 apfu and from 0.10 to 0.15 apfu, respectively. Muscovite does not contain F.
Breccia with muscovite-tourmaline cement (sam. MN-7). The rock is composed of fragments of vein quartz, which are cemented by tourmaline and a small amount of white mica. The size of fragments of quartz is up to several cm.
Tourmaline is represented by small crystals of several dozens of microns (Fig. 4), which are pleochroic from light green to green. The BSE images show that the crystals are weakly zoned (Fig. 4a). The content of Fe in the mineral is low (0.61-1.03 apfu); Ca content is also low (0.07-0.21 apfu). Tourmaline contains minor Mn (up to 0.01 apfu); Ti concentration ranges from 0.02 to 0.09 apfu (Table  1). In the triangular diagram in terms of X-vacancy -Ca-Na (+ K) (Fig. 2a), the compositions fall in the alkali field; together with the results of calculation of the formulas, it makes it possible to refer the described tourmaline to dravite. The calculated formulas of dravite turned out to be balanced in charge; therefore, it is impossible to estimate Fe 3+ content in the mineral by means of calculations. The excess charge versus Mg + Fe + X-vacancy plot (Fig. 2c) indicates that the chemical substitution of Al + O 2-↔ R + OHis exposed in dravite, where R = Fe 2+ , Mg.
White mica forms aggregates of small scales between tourmaline crystals (Fig. 4b). Chemically, mica refers to muscovite (Table 2, an. 9). The Fe and Mg contents are 0.11 and 0.09 apfu, respectively.
Tourmaline-bearing assemblages of the Kulibinsky site are quartz-tourmaline metasomatites (sample MN-33); it appears to form a linear zone in volcanic rocks. Metasomatites are dissected by veinlets of late quartz with 1-4 cm thick. The veinlets are composed of coarse-grained parallel-columnar quartz, the color of which varies from lacteous to lilaceous.
Tourmaline forms dense, almost monomineralic aggregates of acicular crystals and is represented by two generations. The first generation tourmaline is composed of relatively large, complexly zoned crystals ranging in size from 100 μm to a few mm (Fig. 5a, b). It contains fine inclusions of W-bearing rutile and REE phosphates. The second generation tourmaline is represented by small individual crystals up to 20 μm in diameter (Fig. 5b, c); as an aggregate of split crystals, it molds on tourmaline I (Fig. 5d) or forms the outer zones on tourmaline crystals of the first generation (Fig. 5b).
Large crystals of tourmaline-I are pleochroic from brown or dark blue to black. The BSE images (Fig. 5a, b) show their complex zonality due to the variable Fe content. Zonality is especially well exhibited along the long axis (Fig. 5a). Crystallization begins with the formation of Fe-rich tourmaline (Fe/(Fe + Mg) 0.55-0.77 (Table 1). Then ferriferous acquires a rhythmic character varying in wide range from 0.41 to 0.71. Chemically, tourmaline-I is confined to the alkali series (Fig. 2a) and can be referred to schorl, but composition -to dravite. However, it should be noted that establishing the concentration of Fe 3+ in the mineral can change its classification. Schorl contains up to 0.16 apfu Ti; Ca concentration is 0.10-0.37 apfu. The Mn content in most compositions is less-than-detection limits; sometimes, it is 0.01-0.02 apfu. Second-generation tourmaline crystals are also pleochroic from brown or dark blue to black. In contrast to the first-generation schorl crystals, they have the usual growth zonality with a darker central part and a lighter outer zone (Fig. 5b, c). The central parts of crystals are composed of foitite, the edge parts can be confined to schorl. However, establishing the concentration of Fe 3+ in the mineral can change its classification. Such a change is more likely to take place because the tourmaline edge zones are characterized by a significant deficit of positive charges (when calculating a formula for 15 cations, where iron is considered as ferrous). The calculated amount of Fe 3+ in the edge zones of tourmaline-II crystals is 0.43-1.78 apfu. In the diagram in terms of of the excess charge -Mg + Fe + X (Fig. 2c), two imaging points of compositions are located near the bosiite point, tourmaline with the ideal formula NaFe 3+ 3 (Mg 2 Al 4 ) [Si 6 O 18 ] (BO 3 ) (OH) 3 O [13]. The same graph as well as diagrams in Fig. 2c, e, g show that tourmaline of both generations is characterized by substitution of Fe 3+ ↔ Al. Indirectly, the location of a number of imaging composition points below schorl-dravite line on the triangular diagram in terms of Fe-Al-Mg (Fig. 2d) indicate the enrichment of Fe 3+ tourmaline.
White mica composes colorless scales of several hundred microns in size. Chemically, it refers to muscovite ( Table 2, an. 10). The Fe and Mg contents are 0.18 and 0.05 apfu, respectively. Muscovite does not contain F.

Discussion
The obtained data show that tourmalines of the Mnogovershinnoe deposit differ in their chemical composition and type of substitution. All studied tourmalines by these parameters can be divided into two groups.
For tourmaline of the first group, substitutions Fe ↔ Mg and X-vacancy + Al ↔ Na + R 2 + are common, with the secondary role of replacing Al + O 2-↔ R 2+ + OH -, except for single sectors of crystals enriched in Ca. One more fact that allows us to combine these tourmalines is the increased Zn and Mn content compared with group 2. The large size of quartz-muscovite veinlets of tourmaline-I crystals in sandstones and the chemical composition identical to tourmaline of pegmatoids implies that these veinlets are pegmatoids as well. The charge-balanced chemical formulas of tourmaline do not even allow an approximate estimate the ratio of Fe 3+ /Fe total in tourmaline, which indicates a low Fe 3+ grade in the mineral or its absence. In turn, this indicates reducing or slightly oxidative conditions for the formation of tourmaline and pegmatoids in general. According to [14], the ratio Fe 3+ /Fe tot in tourmaline of pegmatites is 0.05-0.22.
For tourmaline of the second group, substitutions Fe 3+ ↔ Al and Al + O 2-↔ R 2+ + OHare common, which is typical of tourmalines from porphyry-style deposits [15]. The formulas of tourmalines of quartz-tourmaline metasomatites calculated using the electron probe analysis are not balanced in charges, which makes it possible to approximately estimate the ratio Fe 3+ /Fe tot . In tourmaline of quartz-tourmaline metasomatites, it varies from 0.01 to 0.48 indicating the enrichment of Fe 3+ in a number of compositions and the increased oxidative potential of mineral formation environment. The location of points below the line of schorl-dravite on a triangular diagram (Fig. 2d) indirectly indicates the enrichment of tourmaline compounds with ferric iron. The calculated ratio Fe 3+ /Fe total in tourmaline-II quartz veinlets in sandstones and tourmaline cement quartz breccia does not exceed 0.08 giving evidence of more reducing conditions of formation than in the case of quartz-tourmaline metasomatites. The imaging points of these tourmalines are located above the schorl-dravite line on the triangular diagram indirectly indicating a low content of Fe 3+ in minerals. The obtained data suggest that the formation of tourmaline of the second group took place against the background of a decrease in the oxidative potential of the mineral formation environment.
Conclusion Tourmalines of the post-ore mineral associations of the Mnogovershinnoe gold deposit are divided into two groups characterized by different chemical composition and substitutions. Tourmalines of the first group with substitutions Fe ↔ Mg and X-vacancy + Al ↔ Na + R 2+ are confined to pegmatoids and were formed in reductive or slightly oxidative conditions. Later tourmalines of the second group with substitutions Fe 3+ ↔ Al and Al + O 2-↔ R 2+ + OHindicate the possible presence of mineralization of the porphyry type and the formation of a mineralization while reducing the oxidation potential of mineral formation environment.