Petrography and Mineral Chemistry of Magmatic and Hydrothermal Biotite in Porphyry Copper-Gold Deposits: a Tool for Understanding Mineralizing Fluid Compositional Changes During Alteration Processes

DOI: 10.17014/ijog.5.1.47-64This study aims to understand the petrography and chemistry of both magmatic and hydrothermal biotites in porphyry copper-gold deposits, and to evaluate the fluid compositional changes during alteration processes. A total of 206 biotite grains from selected rock samples taken from the Batu Hijau porphyry Cu-Au deposit was analyzed. Detailed petrography and biotite chemistry analysis were performed on thin sections and polished thin sections, respectively, representing various rocks and alteration types. A JEOL JXA-8900R electron microprobe analyzer (EMPA) was used for the chemistry analysis. The biotite is texturally divided into magmatic and hydrothermal types. Ti, Fe, and F contents can be used to distinguish the two biotite types chemically. Some oxide and halogen contents of biotite from various rocks and alteration types demonstrate a systematic variation in chemical composition. Biotite halogen chemistry shows a systematic increase in log (XCl/XOH) and decrease in log (XF/XOH) values from biotite (potassic) through chlorite-sericite (intermediate argillic) to actinolite (inner propylitic) zones. The y-intercepts on the log (XCl/XOH) vs. XMg and log (XF/XOH) vs. XFe plotted for biotite from potassic and intermediate argillic zones are similar or slightly different. In contrast, the y-intercepts on the log (XCl/XOH) vs. XMg and log (XF/XOH) vs. XFe plotted for biotite from inner propylitic zone display different values in comparison to the two alteration zones. Halogen (F,Cl) fugacity ratios in biotite show a similar pattern: in the potassic and intermediate argillic zones they show little variation, whereas in the inner propylitic zone they are distinctly different. These features suggest the hydrothermal fluid composition remained fairly constant in the inner part of the deposit during the potassic and intermediate argillic alteration events, but changed significantly towards the outer part affected by inner propylitic alteration. High halogen content, particularly Cl, in hydrothermal biotite may portray that copper and gold were transported in mineralizing fluids in the form of chloride complexes CuCl2- and AuCl2-, respectively.


Introduction
Biotite, a trioctahedral mica with the generalized formula (K, Na, Ca, Ba) (Fe +2 , Fe +3 , Mg, Ti +4 , Mn, Al) 3 (Al, Si) 4 O 10 (OH, F, Cl) 2 is a common constituent in many porphyry copper deposits. In such environment, most biotite has formed during igneous crystallization of the host intrusive I J O G or during hydrothermal alteration of the plutonic host. Since many thermodynamic variables can control the complex chemistry of biotite, its composition is potentially useful in understanding some of the physical and chemical conditions associated with igneous and hydrothermal events leading to the formation of the porphyry copper deposits.
The majority of previous studies of biotite chemical composition in porphyry Cu deposits have concentrated on the F and Cl mineral contents, with the objective of distinguishing between mineralized and barren plutons. Few studies of porphyry Cu deposits have reported full elemental chemistry of biotite. A notable exception is the study by Parry and Jacobs (1975) who reported variations in biotite chemistry from potassic, phyllic, propylitic, and argillic alteration zones in Santa Rita deposit, and estimated the HF, HCl, and H 2 O fugacity of hydrothermal fluids. Selby and Nesbitt (2000) published the chemical composition of biotite from Casino porphyry Cu-Au-Mo mineralization and an evaluation of magmatic and hydrothermal fluid chemistry. Our knowledge of the HF, HCl, and H 2 O fugacity of hydrothermal fluids associated with porphyry Cu mineralization and hydrothermal alteration is based on limited published thermodynamic data (Parry and Jacobs, 1975;Parry et al., 1978;Munoz and Swenson, 1981;Munoz, 1984;Loferski and Ayuso, 1995). Presently, there are several studies dealing with mineral chemistry of hydrothermal biotite from numerous porphyry copper deposits worldwide, such as Kahang deposit, Iran (Afshooni et al., 2013), Lar deposit, Iran (Moradi et al., 2016), Dexing deposit, China (Bao et al., 2016), and Sisson Brook deposit, New Brunswick, Canada . Halogen chemistry of magmatic biotite within the mineralization-bearing tonalite porphyry in Batu Hijau deposit, Indonesia, was described by Idrus et al. (2007).
This study comprises detailed petrographic examination and microprobe analysis of the major and minor element chemistry of various biotite types, particularly hydrothermal biotite, together with evaluation of hydrothermal fluid changes during alteration processes of the porphyry copper-gold deposits based on a case study of the Batu Hijau deposit.
The Batu Hijau porphyry Cu-Au deposit is located in the southern part of Sumbawa Island, Indonesia ( Figure 1). It contains 914 million metric tonnes of ore at an average grade of 0.53% Cu and 0.40 g/t Au (Clode et al., 1999). The wall rock units of the deposit consist of interbedded andesitic lithic breccia and finegrained volcaniclastic rocks as well as intrusions of porphyry andesite and quartz diorite bodies. These units are intruded by multiple phases of tonalite porphyry stocks. Copper and gold mineralization is directly related to the emplacement of the tonalite porphyries. Hydrothermal alteration and mineralization developed in four temporally and spatially overlapping stages include (1) early alteration stage divided into biotite (potassic), actinolite (inner propylitic), and chlorite-epidote (outer propylitic) zones, (2) transitional alteration stage typified by a chlorite-sericite (intermediate argillic) zone, (3) late alteration stage characterized by the destruction of feldspar and the formation of pyrophyllite-andalusite (advanced argillic) and sericite-paragonite (argillic) zones, and (4) very late alteration stage consisting of illite-sericite replacement of feldspar (Idrus, 2005).

Research Methods
A suite of representative thin section and polished thin section samples from the Batu Hijau porphyry copper-gold deposit was petrographically observed. A total of 206 biotite grains from selected rock samples was analyzed chemically. All elemental analysis of biotites was obtained from polished thin section using a JEOL JXA-8900R electron microprobe at Institute of Mineralogy and Economic Geology, RWTH Aachen University, Germany. Each analysis was classified according to the textural type of the biotite and to the kind of rock and the type of alteration of which the biotite is associated. (Si, Ti, Al, Fe Total , Mn, Mg, Ca, Na, K, F, and Cl) were carried out using a beam size of 1.6 µm, an accelerating potential voltage of 15 kV, a probe current of 23.5 nA, and a counting time of 10 s for each element analyzed. Ba was not analyzed due to very low concentration of the element. The analytical procedure for Si, Ti, Al, Fe, Mn, Mg, Ca, Na, K, F, and Cl used synthetic standards of plagioclase (Pl 64 ), rutile (Ru 83 Ti 3 P), spinel (Sp 761 T), fayalite (Fa 66 Mn), jadeite (J 107 ), orthoclase (Or 79 ), fluorite (F l80 TAP), and tugtupite (Tug 34 ). Matrix effects were corrected using the ZAF software provided by JEOL. The accuracy of the reported values for the analysis is 0.2%-15% (1σ) depending on the abundance of the element. A microprobe analysis was defined as the value of one spot analysis of biotite grain. The OH values were calculated on the basis of eleven oxygen formula units. The X phl (mole fraction phlogopite), X Mg, and X Fe values were determined from cation fractions and defined as Mg/(sum of octahedral cations), Mg/ (Fe+Mg), and (Fe/Al IV )/(Mg+Fe+Al VI ), respectively (Zhu and Sverjensky, 1992;Loverski and Ayuso, 1995). The X F , X Cl , and X OH are the mole fractions of F, Cl, and OH in the hydroxyl site. The halogen fugacity ratios consisting of log (fH 2 O)/(fHF), log (fH 2 O)/(fHCl), and log (fHF)/ (fHCl) were computed using the equations of Munoz (1992), which were based on the revised coefficients for the halogen-hydroxyl exchange Sverjensky, 1991, 1992

Biotite Petrography
At the Batu Hijau deposit, biotite grains occur in the central biotite (potassic), transitional chlorite-sericite (intermediate argillic), and proximal actinolite (inner propylitic) alteration zones. The biotites are petrographically classified as magmatic and hydrothermal types. The term "magmatic" denotes biotite grains inferred to have crystallized directly from silicate melt. In the case of the Batu Hijau deposit, the magmatic biotites are observed in both least-altered and potassic-altered tonalite samples. The magmatic biotites are texturally euhedral to subhedral, with the length ranging from 1 to 4 mm ( Figure   2). Some magmatic biotites are slightly ragged, splintery, or frayed.
The term "hydrothermal" describes biotites inferred to have precipitated from the hydrothermal fluids and biotites that have partially completely replaced magmatic biotite and hornblende. The hydrothermal biotites are petrographically distinct from magmatic biotites, commonly occurring as aggregates of fine-grained flakes (typically 10-100 µm) with a crude orientation ( Figure 2).
Biotite grains that occur in the early veinlets/ stringers are also termed "hydrothermal". This biotite type is identified in all altered rock types, including tonalite porphyries, equigranular quartz diorite, and andesitic volcaniclastic rocks.

Biotite Chemistry
A total of 172 electron microprobe analyses of 860 spot analyses of magmatic and hydrothermal biotite grains from the central biotite (potassic), transitional chlorite-sericite (intermediate argillic), and proximal actinolite (inner propylitic) alteration zones were made. In the context used here, a microprobe analysis was defined as the arithmetic mean of 5 -10 spot analyses of a biotite grain or an aggregate of fine-grained biotites, a b c d The representative electron microprobe data indicate that the magmatic and hydrothermal biotites are virtually identical and plotted in the field of phlogopite (cf. Rieder et al., 1998; Figure 3). However, there are some variations in elemental In contrast, the concentration of F tends to be higher in hydrothermal biotites than in the magmatic type. Compositional variations of the two biotite types are shown graphically in Figure 4, involving mole fraction phlogopite (X Phl = Mg/ (sum of octahedral cations) (Zhu and Sverjensky, 1991) and weight percent of TiO 2 , FeO, and F. Separate plots (not shown here) also indicate that the magmatic biotites have relative lower SiO 2 The X Mg of magmatic biotite is lower than that of hydrothermal biotite. With respect to the rock types, X Mg of secondary biotites in andesitic volcaniclastic rocks from various alteration zones tends to be lower (X Mg ≤ 0.70) than in intrusive rocks (X Mg > 0.70). Secondary biotites in tonalite porphyries (intermediate and young tonalites) contain the highest X Mg with an average value of 0.75, relative to those in equigranular quartz diorite (X Mg = 0.73). Compositional trends between the secondary biotites in equigranular quartz diorite and andesitic volcaniclastic rocks associated with different zones of alteration are shown in a series of variation diagrams ( Figure   4). Hydrothermal biotite hosted by equigranular quartz diorite shows an increase in Al 2 O 3 , FeO, and Cl, but a decrease in TiO 2 , MnO, and F from the central biotite and transitional chlorite-sericite to proximal actinolite alteration zones. SiO 2 , MnO, and F display a positive correlation with X Phl , whereas the Al 2 O 3 , FeO, and Cl are negatively correlated with X Phl .
Hydrothermal biotite in the andesitic volcaniclastic rocks shows an increase in TiO 2 , Al 2 O 3 , FeO, MnO, and Cl, and a decrease in SiO 2 and F from the inner part (central early   noz (1984). It clearly demonstrates that biotites with high Mg tend to incorporate more F and less Cl in comparison to biotites with lower Mg; a crystal-chemical effect referred to as "Fe-F and Mg-Cl avoidance rules" (Munoz, 1984;Zhu and Sverjensky, 1991;Munoz, 1992). This suggests that the octahedral Mg and Fe of biotite control the halogen-hydroxyl exchange of hydrothermal fluids within the Batu Hijau deposit.

Biotite Halogen Chemistry
The halogen contents of the hydrothermal biotites are illustrated in Figure 6. The contents in different rock types are grouped according to their related alteration zones. The biotites from the central biotite, transitional chlorite-sericite, and proximal actinolite alteration zones have average log (X Cl /X OH ) values of -1.54, -1.53, and -1.43, respectively. The log (X F /X OH ) values are -0.89, -0.96, and -1.34, whereas log (X Cl /X F ) values vary from 0.65, 0.58 to 0.09. The systematic variation of the logarithmic ratios reflects a systematic variation of the halogen content in biotites associated with these alteration zones. Zhu and Sverjensky (1992) have shown that the composition of biotite formed under similar physicochemical conditions produces linear trends on log (X F /X OH ) against X Fe and log (X Cl / X OH ) against X Mg plots. The slopes of these linear trends are a function of temperature, but they are independent from pressure and fluid composition, whereas the y-intercept value is a function of all these parameters (Zhu and Sverjensky, 1992;Selby and Nesbitt, 2000). The slopes for log (X F /X OH ) against X Fe and log (X Cl /X OH ) against X Mg plots were calculated from Equation 23 and 24 of Zhu and Sverjensky (1992). The calculation indicates that the slopes on the log (X F /X OH ) against X Fe diagram change from -2.93 at 100°C to -1.90, -1.25 and -0.858 at 300, 600, and 1000°C, respectively. In comparison, the slopes on the log (X Cl /X OH ) against X Mg diagram plots are more gentle and less dependent on the temperature, which change from -1.32 at 100°C to -0.86, -0.56, -0.39 at 300, 600 and 1000°C, respectively (Zhu and Sverjensky, 1992). Using the homogenization temperature (T h ) of 510°C from microthermometric data of fluid inclusions in the early quartz veins/veinlets associated with the early biotite (potassic) alteration zone (Turner, 1995;Garwin, 2000), and temperature stability of mineral assemblages within the transitional chlorite-sericite (intermediate argillic) and proximal actinolite (inner propylitic) alteration zones (475 °C and 290 °C, respectively) (Idrus, 2005), the slopes in this study were determined on the basis of those experimental diagram plots. These calculated slopes pass through the majority of the data points from the central biotite, transitional chlorite-sericite, and proximal actinolite alteration zones ( Figure 6).

I J O G
The slopes for the central biotite alteration zone were calculated using maximum Th (510°C) of the fluid inclusions within the early quartz veinlets (Turner, 1995;Garwin, 2000). No microther-mometric data are available for the transitional chlorite-sericite and actinolite alteration zones. Therefore, their slopes were defined on the basis of median stability temperatures of those mineral  Sverjensky (1991, 1992), and temperatures are from fluid inclusion data and stability of mineral assemblages. Notes: X F , X Cl, and X OH respectively, represent the mole fractions of F, Cl, and OH in the hydroxyl site, e.g. X F was defined as X F = F/(F+Cl+OH). N = Number of biotite grains analyzed.

I J O G
assemblages. Biotite geothermometry (Beane, 1974) is not applicable, due to the absence of Kfeldspar coexisting with biotite within the Batu Hijau deposit. The formation temperatures for the transitional chlorite-sericite and proximal actinolite zones are 475 and 290°C, respectively (Browne, 1978;Clode et al., 1999;Garwin, 2002). As a result, the calculated slopes of log (X Cl /X OH ) against X Mg and log (X F /X OH ) against X Fe plots for the central biotite alteration zone are -0.63 and -1.40, respectively. The slopes of log (X Cl /X OH ) against and X Mg log (X F /X OH ) against X Fe for the transitional chlorite-sericite zone are -0.66 and -1.46 ( Figure  6c -d), whereas those for the proximal actinolite zone are -0.88 and -1.94 (Figure 6e -f).
The y-intercepts on the log (X Cl /X OH ) against X Mg and log (X F /X OH ) against X Fe plots for the central biotite zone are -1.05 and -0.50, respectively (Figure 6a -b). The values are similar to those for the transitional chlorite-sericite zone (-1.00 and -0.55; Figure 6c -d). Some biotite data, particularly log (X F /X OH ) against X Fe for the transitional chloritesericite zone (Figure 6d), show a scatter distribution. This may suggest that the biotite compositions were strongly affected by later hydrothermal fluids during alteration. Compared to these alteration zones, the y-intercepts on the log (X Cl /X OH ) against X Mg and log (X F /X OH ) against X Fe plots for the proximal actinolite zone display different values (-0.75 and -0.70, respectively; Figure 6e -f).

Halogen Fugacity Ratios
The greater exchangeability of elements residing in the hydroxyl site relative to those in the tetrahedral and octahedral sites of biotite opens the possibility that the biotite will undergo F-Cl-OH equilibrium with the hydrothermal fluids that are responsible for the formation of the alteration zones (Selby and Nesbitt, 2000). Mg and Fe contents of the Batu Hijau biotites are directly related to F-Cl-OH exchange between the biotites and hydrothermal fluids, as indicated by a good agreement of their compositions with the "Mg-Cl and F-Fe avoidance rules". The theoretical estimates of F-Cl-OH exchange between biotite and hydrothermal fluids are based on the model equilibrium reaction (Munoz and Ludington, 1974;Gunow et al., 1980;Parry et al., 1984, andMunoz, 1984): where X stands for either F or Cl.
The equilibrium constant is calculated from log K = log (X X /X OH ) biotite +log [(fH 2 O)/f(HX)] fluid . The log (X X /X OH ) biotite portrays the halogen contents of biotite, whereas log [(fH 2 O)/f(HX)] fluid represents the fugacity ratios of the hydrothermal fluids involved. The fugacity ratios consisting of log (fH 2 O)/(fHF), log (fH 2 O)/(fHCl), and log (fHF)/(fHCl) were further computed using the equations of Munoz (1992), which were based on the revised coefficients for the halogen-hydroxyl exchange Sverjensky, 1991, and1992).  (Figure 7). The significant variation in the fugacity ratios for hydrothermal fluids related to the proximal actinolite zone has also been recorded by Selby and Nesbitt (2000) at the Casino porphyry deposit.

Fluid halogen Fugacity: A comparison to other porphyry systems
This section aims to compare the biotite halogen chemistry and halogen fugacity ratios of the hydrothermal fluid responsible for the formation of alteration zones in the Batu Hijau deposit, with those of other major porphyry systems. This allows to better understand the physicochemical conditions of the hydrothermal fluid in the deposit.
The biotite halogen chemistry and halogen fugacity ratios of the hydrothermal fluids related to the central biotite (potassic), transitional chlorite-sericite (intermediate argillic), and proximal actinolite (inner propylitic) alteration zones in the Batu Hijau deposit were compared to those of fluids related to various alteration zones of Bingham (Parry et al., 1978;Bowman et al., 1987), Santa Rita (Jacobs and Parry, 1979) and Casino (Selby and Nesbitt, 2000) porphyry copper deposits as well as those of fluids associated with the Hanover (Jacobs and Parry, 1976) and Deboullie (Loferski and Ayuso, 1995) porphyry Cu-related plutons ( Figure 8). As described in a previous section, the fugacity ratios for hydrothermal fluids associated with the central biotite, transitional chlorite-sericite, and proximal actinolite alteration zones in the Batu Hijau deposit were determined at the temperatures of 510, 475 and 290°C, respectively. The fugacity ratios for other deposits/plutons were recalculated at temperatures established from fluid inclusion or mineral pair geothermometry. The analytical data were taken from related authors above, e.g. 350°C for the Santa Rita (potassic alteration), Hanover, and Deboullie as well as 400°C for the Bingham (potassic alteration). The fugacity ratios of the Casino porphyry copper deposit were calculated at temperatures of 420, 390, and 350°C for potas-sic, phyllic, and propylitic alteration zones (cf. Selby and Nesbitt, 2000).
As outlined before, the hydrothermal fluid

I J O
G zones in the Casino porphyry copper deposit. The fugacity ratios established for potassic and phyllic alteration zones are relatively identical, but the fugacity ratios for propylitic alteration greatly contrast with those for both alteration zones above. Furthermore, log (fH 2 O)/(fHF) and log (fH 2 O)/(fHCl) values for fluid related to the Casino propylitic zone are typically lower than those for fluid of the Batu Hijau propylitic zone.
This suggests that F and Cl contents are higher in the propylitic-related fluid of the Casino than those of Batu Hijau. Interestingly, the trend of the fugacity ratios of fluids related to all alteration zones in both porphyry copper deposits is similar; they increase towards outer parts of the deposits, from central biotite (potassic), transitional chlorite-sericite (intermediate argillic)/phyllic to proximal actinolite (propylitic) alteration zones (Figure 8). This suggests that the halogen contents of hydrothermal fluids in both porphyry copper deposits decrease outward, as a result of a degree increase of mixing between magmatic fluid and meteoric water.
The fugacity ratios determined for potassic alteration at Bingham describe a large range of values (Figure 8a -b). However, the data cluster and with respect to log (fH 2 O)/(fHCl) and (fH 2 O)/ (fHF) define two populations. The large spread in the fugacity ratios at Bingham may suggest that post potassic hydrothermal fluids have modified the biotite halogen chemistry. However, the two populations of the fugacity ratios suggest that two different hydrothermal fluids forming the potassic alteration were present at Bingham. In addition, log (fHF)/(fHCl) values for fluid associated with the potassic alteration zone in the Bingham are significantly higher (-2.0 to -0.1) than those of the fluid determined for the central biotite (potassic) alteration in Batu Hijau (-2.5 to -2.0). This suggets that the magmatic-dominated fluid forming the copper-gold-bearing potassic alteration zone at Bingham has a higher chlorine content than that in fluid related to the central biotite (potassic) alteration zone at the Batu Hijau deposit. The relatively higher chlorine content in the potassic-related fluid at Bingham may reflect a higher activity of chloride complexes in the fluid, in comparison to that at the Batu Hijau deposit.
The chloride complexes have been identified as important components of hydrothermal fluids in transporting copper, gold, and other metals in porphyry deposits (e.g. Shinohara, 1994;Gammons and Williams-Jones, 1997). The higher activity of chloride complexes within the potassic-related fluid at Bingham, relative to the Batu Hijau may be expressed by higher average grades of Cu and Au (0.88 % and 0.50 g/t, respectively; Babcock et al., 1995 andKeith et al., 1997) (Note: the average Cu and Au grades of the Batu Hijau deposit are 0.53 % and 0.40 g/t, respectively; Clode et al., 1999).

Conclusions
Biotites in the tonalite porphyries are texturally distinguished into magmatic and hydrothermal types. The magmatic type contains higher TiO 2 and FeO than the hydrothermal type, while F is relatively higher in the hydrothermal biotites. The correlation between X Mg and halogen (F, Cl) contents shows a good agreement with the experimental results of Munoz's (1984). Biotite with high Mg tends to incorporate more F and less Cl compared to biotite with lower Mg; a crystal-chemical effect referred to as "Fe-F and Mg-Cl avoidance rules" (cf. Munoz, 1984;Zhu and Sverjensky, 1991;Munoz, 1992). This suggests that the octahedral Mg and Fe in biotite controls the halogen-hydroxyl exchange of the hydrothermal fluids.
The halogen chemistry of the biotite exhibits a systematic variation through the various alteration zones. The biotite shows a systematic increase in log (X Cl /X OH ) values, and a decrease in log (X F / X OH ) and log (X Cl /X F ) values from the early central biotite, transitional chlorite-sericite to the proximal actinolite alteration zones. The y-intercepts on the log (X Cl /X OH ) vs. X Mg and log (X F /X OH ) vs. X Fe plots for biotite from biotite zone (potassic) and chlorite-sericite zone (intermediate argillic) are similar or slightly different. The similarity of I J O G intercepts and the narrow scatter shown in the data for biotite and chlorite-sericite zones imply that the X F /X OH and X Cl /X OH values of the hydrothermal fluid were fairly constant during the formation of both potassic and intermediate argillic alteration, forming under broadly similar temperature conditions. In contrast, the y-intercepts on the log (X Cl / X OH ) vs. X Mg and log (X F /X OH ) vs. X Fe plots for biotite from actinolite zone (inner propylitic) display different values in comparison to both alterations above. The latter suggests that biotite chemistry of the actinolite zone may record the change in fluid composition during the alteration formation.
The compositional changes in oxide and halogen chemistry of biotite from the various rocks and alteration types at Batu Hijau deposit conform to those from other porphyry Cu deposits, although absolute concentrations vary considerably among deposits. The narrow variation of the halogen (F, Cl) fugacity ratios in the early central biotite and transitional chlorite-sericite alteration zones may suggest that the hydrothermal fluid composition was fairly constant at a broadly similar temperature. This may also explain the strong overlapping of the early central biotite zone by the transitional chlorite-sericite assemblages. The fugacity ratios increase systematically from the inner, transitional to outer parts of the deposit, corresponding to fluid composition change and relative timing for the subsequent formations of the early, transitional to late alteration zones. High halogen content particularly Cl, contained by hydrothermal biotite may imply that copper and gold in the porphyry deposit were transported in the form of chloride complexes CuCl 2 and AuCl 2 -, respectively.