Exploring magmatic gas and subvolcanic hydrothermal system interactions: Mineralogy and sulfur isotope characteristics of the Holocene volcanic products at Mt. Tangkuban Parahu, Indonesia.

A subvolcanic-hydrothermal system involves complex interaction between magma, magmatic uids, and hydrothermal system at stratovolcanoes in subduction setting. These interactions are responsible for magmatic-hydrothermal eruption associated with rapid injection of magmatic gas into hydrothermal system at a certain depth of volcanic edice. However, capturing these interactions is challenging due to inaccessibility to the crater conduit within the edice. Therefore, we selected a method to analyze the volcanic products from several episodic phreatic and phreatomagmatic eruptions during the Holocene at Tangkuban Parahu, Indonesia. In this context, Holocene volcanic products are one of the best examples to understand an interplay between magma, magmatic uids, and hydrothermal system in producing violent eruptions. In this study, we carried out petrological and sulfur isotope analysis only for the hydrothermally altered lithic ash particles, a part of proximal volcanic products. Mineral assemblages mostly exhibit a typical acid-sulfate and advanced argillic alteration, consisting of alunite, kaolinite, and silica minerals. Acid-sulfate and advanced argillic alteration indicates that those mineral assemblages were formed under the formation temperature ranging from ~100 to ~260 . The calculated temperature from sulfur isotopic fractionation of sulfate-sulde shows 230-240 , which is almost identical with assigned temperature from mineral assemblages. Sulfur isotope and jarosite occurrence indicate the supergene alteration associated with oxygen entrainment to the hydrothermal system that oxidize pyrite to jarosite. Sulfur isotopic variation throughout the studied stratigraphy represents inux of magmatic gas to the hydrothermal system. Moreover, zoned P-bearing alunite represents repetitive injection of magmatic gas to the active acidic hydrothermal system, which also indicates the magmatic-hydrothermal interaction below the crater. Occurrence of enargite and chalcopyrite represents the nature of upper-level high suldation system at shallow volcanic edice of the Tangkuban Parahu volcano. Furthermore, we showed that coupled petrological and sulfur isotope analysis has paramount importance to evaluate the conditions of the subvolcanic hydrothermal system, magmatic-hydrothermal interaction, and the origin of steam-blast eruptions at volcanoes that contain subvolcanic-hydrothermal systems. Backscattered electron images showing textural opal-cristobalite±alunite±kaolinite in ash particles, including the Raman Spectral from selected point analyses from several ash particles. A) Opal alteration was found on vesiculated ash particles. Vesicular textures were lled with silica minerals. Magmatic minerals were replaced by opal, where the relict of magmatic minerals (A, B) remains at the core of crystals. C). Graph showing the two representatives Raman spectrum of cristobalite and opal in ash particles (B, D). D) BSE image shows opal-cristobalite after plagioclase, whereas silica-alunite±kaolinite occurs as a groundmass alteration. F) Opal-cristobalite-silica-alunite sporadically replaced the magmatic minerals in dense ash particles. Most of the hydrothermal minerals occur as pseudomorphic textures.


Introduction
A subvolcanic-hydrothermal system involves complex interactions between magma, magmatic uid, and hydrothermal uid at high-terrain stratovolcanoes in subduction settings (Hedenquist and Lowernstern, 1994). A hydrothermal system is responsible for a magmatic-hydrothermal eruption associated with a rapid injection of magmatic gas into the hydrothermal system from magmatic intrusion at a certain depth of volcanic edi ce (Mastin et al., 1995;Browne and Lawless, 2000;Stix and Moore, 2018). The magmahydrothermal interaction leads fragmentation of various rocks along the conduits in the stratovolcano, including cognate magmatic rocks and hydrothermally altered country rocks (Browne and Lawless, 2001;Ohba and Kitade, 2005;Ohba et al., 2009). As a result, the volcanic products of the eruption are highly diverse in component (e.g., Ohba and Kitade, 2005;Pardo et al., 2014). The diversity of the components provides an opportunity to understand the dynamics of the subvolcanic-and magmatic-hydrothermal systems (e.g., Hedenquist and Henley, 1985;Ohba and Kitade, 2005;Ohba et al., 2007;Minami et al., 2016) as well as the contribution of magmatic activities resulting in violent eruptions.
A crucial question is how the mineralogy in subvolcanic-hydrothermal system varies by the interaction with magmatic components along the active volcanic conduit. Answering this question is challenging because of the inaccessibility to the underground system within the volcanic edi ce. Chemical and physical variations in volcanic-hydrothermal system can be recorded as mineralogical characteristics in eruptive products (e.g., Ohba and Kitade, 2005;Ohba et al., 2007;Takahashi and Yahata, 2018). The eruptive products frequently preserve the assemblages from hydrothermal mineral zones (e.g., John et al., 2008;Minami et al., 2016;Imura et al., 2019) which are similar to those developed in ancient hydrothermal ore deposits (e.g., Hedenquist and Lowernstern, 1994;Sillitoe, 2010). In addition to the similarity to the ancient hydrothermal systems, volcanic products can provide information of on-going interaction between magma and the hydrothermal system (e.g., Miyagi et al., 2020).
Sulfur is one of important elements in sub-volcanic hydrothermal systems. Sulfur species in volcanic gas derived from magma can mix with ground water to make sulfuric acid that promotes chemical reactions between the uid and rocks (e.g., Sakai, 1968;Giggenbach, 1997;Landis and Rye, 2005). To understand magma-hydrothermal interactions under a volcano based on observations on volcanic products, sulfurbearing minerals should be carefully examined. Sulfur isotope analysis of the minerals is a potential prime method to understand the magma-hydrothermal interactions (e.g., Imai et al., 2007;Zimbelman et al., 2005;Maruoka, 2016, 2019). Therefore, we apply the sulfur isotopic analysis to the volcanic products which are derived from magma-hydrothermal interactions.
The present work focused on volcanic products from the proximal area of Tangkuban Parahu volcano, Indonesia, which has an established tephra-stratigraphy (cf. Angkasa et al., 2019). The volcano has been active within the Holocene epoch. During Holocene, the activities are magmatic-hydrothermal and phreatic explosive eruptions, along with several magmatic intrusion in contact with water bodies (e.g., crater lake) to explode as a common phreatomagmatic eruption (CVGHM, 2016;Angkasa et al., 2019). In this context, Holocene volcanic products from this volcano are good examples to understand interactions between magma and a hydrothermal system. In this study, we carried out petrological observation on ash particles to document hydrothermal mineral assemblage and overprinting relationship (e.g., veinlets, alteration type) referring the fossil and active geothermal systems at the stratovolcanoes worldwide (e.g., Hedenquist and Henley, 1985;Wohletz and Heiken, 1992;Hedenquist and Taran, 2013). We also performed sulfur isotopic analysis to understand formation conditions of sulfate and sul de, contribution of magma to the eruptions, and magma-hydrothermal interaction in the subvolcanic hydrothermal system.

Holocene volcanic activities
Holocene volcanic activities of Tangkuban Parahu are mostly still understudied in the sense of the eruption process as well as the characteristic of volcanic products. The Holocene eruptive products are composed of ne-coarse fall-out and base surge deposits (Kartadinata et al., 2002;Angkasa et al., 2019), which covered a total of 20 km 2 area of the summit area of Tangkuban Parahu volcano (CVGHM, 2016). The volcanic products consist of various proportions of lithics, magmatic juveniles, and isolated free crystals in the tephra layers from several episodic eruptions (Angkasa et al., 2019). Although the phreatic explosion is thought to be a predominant eruption type during the Holocene epoch (Kartadinata et al., 2002), the recent observation of ash components suggests the occurrence of several phreatomagmatic pulses within the 10,000 years (Angkasa et al., 2019), indicating a vital role of the magmatic intrusion and degassing in producing the explosion during the Holocene epoch.

Hydrothermal activities
The Tangkuban Parahu geothermal system widely extends to Ciater, Maribaya, and Batugede which is one of the most extensive geothermal systems associated with volcanic-magmatic activities among West Java, Indonesia (Hochstein and Sudarman, 2015). Hydrothermal manifestation currently develops as active fumaroles, hot springs, and solfataras with a fumarolic temperature of 90-100 ºC (Sriwarna, 1984;Saputra and Suryantini, 2015). At Ratu Crater, the temperature of fumaroles occasionally increased higher than 300 ºC in 1952300 ºC in , 1961300 ºC in , and 1969300 ºC in (Suryo, 1981300 ºC in and 1985. Hydrothermal alteration covered the area of ~1 km 2 (Fig. 1B, C, and D), scattered close to the proximity of Ratu Crater. Hydrothermal alteration minerals consist of a zonation of silica and native sulfur at Ratu, Upas, and Baru Craters (Fig. 1C), whereas silica and native sulfur occur together with pyrite at Domas Crater (Saputra and Suryantini, 2015).

Sampling and sample preparation
Volcanic products were sampled during the eldwork campaign in August 2017 and October 2018 ( Fig.  2A, B, and C). More than 50 samples were collected from several outcrops at the proximal area of the Tangkuban Parahu volcano (Fig. 1B). Most of the volcanic products are white and yellow to reddish in color at a wet condition. We cleaned the outcrops to avoid contamination from surrounding organic material. From the outcrop, volcanic products were carefully collected in the eld by using a small shovel and placed into a plastic container. A total of 26 representative samples (Fig. 2) were manually sieved into different fractions. For this study, we selected volcanic products with a size of particles smaller than 500 micrometers (µm). Selected particles were bathed with distilled water along with an ultrasonic cleaner to remove clay and dust from the samples and dried with an oven at 40 C for 12 hours.
X-ray diffraction X-ray diffraction (XRD) analysis was performed to identify hydrothermal and magmatic mineral phases in the volcanic products on a total of 21 sieved samples. Volcanic products were manually crushed into a ne-grained powder using iron and agate mills. Powder samples were used to produce randomly oriented and oriented samples. Ethylene glycol treatment was optionally applied to some of the oriented samples to verify the presence of 15 Å-smectite group minerals. The XRD instrument was a Rigaku Multi-Flex installed at Faculty of International Resource Sciences, Akita University. The operation conditions are at 30 kV and 16 mA, using Cu target anode for CuKα radiation source. Randomly oriented samples were scanned with a step width of 0.01 and a counting time of 1 min -1 , whereas the oriented samples were scanned with a step width of 0.02 and a counting time of 0.5 min -1 . Moreover, synthetic quartz was analyzed for optimization of X-ray intensity.
Optical and electron microscopes A total of 15 thin and polished sections from different ash layers ( Fig. 2) was observed under optical and electron microscopes, using a Nikon ECLIPSE LV100N and a JEOL JSM-IT300 scanning electron microscope (SEM) equipped with energy dispersive spectrometer (EDS, an Oxford instrument Inca X-act). Microprobe analysis was operated at an accelerating voltage of 15 kV, a beam current 2.2 nA, and a working distance 10 mm. Cobalt was analyzed for an optimization of X-ray intensity. To avoid sample damage, we applied different live times during microprobe analysis of hydrothermal and magmatic minerals. Silica and magmatic minerals (e.g., plagioclase, augite, Fe-Ti oxide) were analyzed at the live time for 80 s. Alunite and kaolinite were analyzed at the live time of 30 s.
Most of the minerals were determined based on their general stoichiometry acquired from EDS spot chemical analysis. Alunite ((K, Na, Ca) Al 3 (SO 4 ) 2 (OH) 6 ) and kaolinite (Al Si O (OH) ) were evaluated by the general formula of DG 3 (TX 4 ) 2 X 6 (Bayliss et al., 2010) and X 4 Z 4 O 10 (OH) 8 with a total oxygen number of 11 and 7, respectively. Moreover, pristine Si-Al clay mineral (kaolinite) was evaluated by using Al/Si ratio of kaolinite stoichiometry close to ~1 because kaolinite commonly occur as ne crystal aggregates and interlocked together with silica minerals.

Raman spectroscopy
Raman spectroscopy analysis using a Renishaw Invia Raman Microspectrometer at Faculty of International Resource Sciences, Akita University was performed to determine some phases (silica and Tioxide polymorph minerals). Silica minerals are speci ed after a near-pure SiO 2 phase is identi ed with EDS, except for pristine quartz crystals. The same method is applied to Ti-oxide minerals to distinguish rutile and anatase. We operated at a laser power of 150 mW and a slit diameter of 1 µm, and an argon excitation laser of 532 nm-1. Silica standard was employed to calibrate the Raman spectrum. A CrystalSleuth software (Laetsch and Down, 2006) was employed to correlate the Raman spectra with speci c crystal reference in the American Mineralogist Crystal Structure Database (www.rruff.info).

Sulfur isotopes
Sulfur isotope analysis was performed to constrain the formation conditions of the sulfate and sul de minerals. A total of 21 samples was handpicked from different tephra layers to produce the temporal variation of the sulfur isotopic composition of volcanic products. Alunite crystals were handpicked from the ash particle of 1 ≤ ≤ 2 in size and collected from yellow and pinkish ash particles, according to macroscopic detailed characteristics alunite bearing hydrothermally altered rocks (Rye, 1993). Jarosite was sampled from a highly oxidized tephra layer in Unit 8 (Fig. 2C). Pyrite in silica-rich ash particles was extracted from gently crushed ash particles because isolated pyrite crystal is absent from the volcanic products. In addition, XRD analysis was performed for all samples to con rm the presence of sulfate and sul de minerals.
Sulfur in alunite and jarosite were dissolved with 0.5 N NaOH at 80 ºC (Wesserman et al., 1992;Manalo et al., 2018), whereas pyrite was decomposed with 16 N HNO 3 and Br 2 solutions at 90 ºC. Leached sulfur solutions were extracted as BaSO 4 precipites by adding 10% BaCl 2 • 2H 2 O. The BaSO 4 precipites were packed with V 2 O 5 in tin foils (modi ed from Yanagisawa and Sakai, 1983). Sulfur isotope measurements were performed with a mass spectrometer (Thermo Fisher Scienti c Delta V at Faculty International Resource Science, Akita University). Seawater sulfate (NBS-127) and barium sulfate (SO-5 and SO-6) were employed as standards, which were provided by the International Atomic Energy Agency (IAEA). The isotopic ratios are presented with an expression of δ 34 S per mil (‰) relative to the sulfur concentration of Troilite in Canyon Diablo Meteorite. The precision of analysis is better than ±0.2 ‰.

Results
Mineral phase identi cation X-ray diffractogram and a summary of identi ed mineral phases are presented in Figures 3 and 4, respectively. Most of the X-ray spectrums indicate the occurrence of quartz, cristobalite, and tridymite as dominant phases in all analyzed samples. These minerals occur together with alunite, 7Å kaolin mineral, and anatase. However, 7Å kaolin mineral and anatase are not necessarily contained in the volcanic products. Although the spectrums of 7Å kaolin were carefully examined, there is no indication of dickite ( Fig. 4). It is noteworthy that 15 Å-smectite group minerals are absent. Major magmatic minerals are sanidine, plagioclase, and augite. Orthopyroxene randomly occurs at the middle and upper portion of composite stratigraphy (Fig. 4). The XRD spectrums also indicate the presence of jarosite and goethite in several tephra layers.

Volcanic ash petrography
Total petrographic observation indicates the abundance of altered ash particles (n= 351 ash particles, observation = 23 ash particles/sample). From the petrographic observation, ash particles can be categorized into two types of hydrothermally altered ash particles and unaltered pyroclasts. The hydrothermally altered ash particles are classi ed into two: 1) partly altered glassy pyroclasts and 2) intensely altered ash particles. Here, we focus on detailed petrographic observation of the hydrothermally altered ash particles. In this section, we will describe the petrography focusing on the mineral assemblages and textural variation of hydrothermally altered lithic ash particles.

Intensely altered lithics
In the intensely altered lithics, the phenocrysts and microlites of magmatic minerals and the glassy groundmass are intensely altered. There are two main hydrothermal mineral assemblages in the intensely altered ash particles (Fig. 4). The association of alunite alteration represents the most abundant alteration type, with approximately 57% of observed ash particles. It is associated with several hydrothermal minerals, including opal, cristobalite, quartz, kaolinite, and jarosite. Quartz alteration is the second abundant alteration type, about 30% of the total petrographic observation, which occurs with ore minerals. Furthermore, these two hydrothermal mineral associations are further subclassi ed into the following ve speci c alteration types and two veinlet feeders.
Pristine cristobalite displays Raman peaks at 226 and 411 cm -1 (Fig. 6C). Altered groundmass consists of alunite, kaolinite, anatase, and mixture of silica minerals. In Figure 6E, the same mineral assemblage appears in angular and dense ash particles, although it shows different occurrence. In these particles, opal, cristobalite, and alunite have entirely replaced the plagioclase crystal. Alunite occurs as zoned crystals with >50 µm in diameters, which commonly associated with kaolinite.

Quartz-alunite±kaolinite alteration
Quartz-alunite±kaolinite alteration (Figs. 7A and B) was predominantly observed in a subrounded to rounded ash particles. Quartz occurs as subhedral to euhedral ne-grained crystals, which are associated with the alunite crystals. Alunite occurs as euhedral, bladed-lath, and zoned crystals. The crystals range from 5 to 10 µm in diameter. Most of the alunite have an intermediate composition between KAl 3 (SO 4 ) 2 (OH) 6 and NaAl 3 (SO 4 ) 2 (OH) 6 . Huangite (Li et al., 1992) are found in quartz-rich ash particles as ne-grained anhedral crystals, containing anatase, pyrite, and chalcopyrite. Kaolinite commonly occurs as circular crystal clusters and ne-grained alteration groundmass.

Alunite-jarosite alteration
Alunite-jarosite alteration was observed in highly porous ash particles, where the voids or pore zones mostly occupied by a mixture of silica minerals and anatase crystals (Fig. 7C). Jarosite is dominant in the alunite-jarosite intergrowth. Alunite occurs as subhedral to euhedral crystal with >10 µm in size, whereas jarosite mostly shows anhedral crystals, ranges from 10-20 µm in diameter.

Quartz alteration
Quartz alteration was observed in angular to subangular ash particles (Figs. 7D and 7E). Dissolution vugs are the most common texture of the quartz alteration. Quartz occupies 90-95 vol% of ash particles, indicating pervasive alteration intensities. Quartz occurs as euhedral, equant crystals with almost pure SiO 2 concentration (98-100 wt.%); such crystals commonly form an interlocked mosaic cluster (Fig. 7D).
Crystals vary in size, ranging from 10 to 50 µm. Quartz alteration is associated with magnetite, titanomagnetite, enargite, chalcopyrite (Figs. 7E and 9A), and pyrite. These minerals disseminate throughout the ash surface with a total concentration of <1 vol%, occurring as pore or vug lling minerals.

Quartz-pyrite veinlet
A quartz-pyrite veinlet was observed in the alunite vein of a porphyritic ash particle (Figs. 8C and 8D). Host rock is possibly an extremely silici ed volcanic rock. The alunite vein is mainly composed of Na-K alunite and P-bearing alunite. Within the alunite vein, a 15-20-µm-thick quartz-pyrite veinlet is developed. Pyrite occurs as centerline crystal inclusions of pyrite in quartz crystals (Fig. 8D). Quartz in the veinlets displays a uniform band in the BSE image, indicating homogenous SiO 2 concentration in the veinlets.
Selvages are primarily composed of ne-grained crystal of quartz and P-bearing alunite. The presence selvages indicate the relative timing of quartz-pyrite veinlet deposition after the alunite vein (Fig. 8D), which also provides an evidence of hydrothermal alteration overprint.

Sulfur isotopes
Sulfur isotopic compositions of sul de and sulfate minerals are presented in Table 2. The δ 34 S values of alunite are isotopically heavier compare to pyrite and jarosite, widely range from +2.8 to +20.8 ‰ (n:19).
The δ 34 S values of all sulfur-bearing minerals were plotted against the stratigraphic unit of proximal volcanic products to represent the temporal change of the sulfur isotopic composition (Fig. 10). The δ³⁴S values of alunite in the lower stratigraphy show the range between +5 to +10‰ (TPT 4, 3, 2, and 1). On the other hand, those in upper stratigraphy are lighter in the upper stratigraphy and concentrated approximately in +5‰ (TPT 6, 7, and 10). Moreover, the decreasing upward trend of the alunite δ³⁴S is illustrated even within a single unit of TPT4. The trend shows the temporal change from +10‰ in TP17 lowermost to +5‰ in TP11 uppermost.

Hydrothermal alteration
Some alteration types were identi ed from the mineralogical observations on the volcanic products as described above. Silica, alunite alterations are the major alterations and some minor alteration types were inferred from the presence of sul de, sulfosalt, jarosite, and goethite.
Cristobalite, opal, tridymite, and quartz are identi ed from the volcanic products. Cristobalite and opal commonly precipitate under low pressure and temperature, probably formed at a temperature range from 100 to 150 ºC (Fournier, 1985). The silica minerals constitute the silica alteration caused by cation leaching and dissolution of the magmatic components by acidic uids (Christenson et al., 2010, Rodriguez andvan Bergen, 2017). Si, Al, and Ti are resilient during the cation-leaching process Yoshida, 1993, 1995;Hara and Tsuchiya, 2009).
Quartz occurs pervasively in silici ed ash particles (Figs. 7A, D, and E). It commonly crystallizes at a relatively high temperature (>180 ) in the hydrothermal environment, depending on the silica solubility in hydrothermal uids (Fournier, 1985;Rimstidt, 1987). The coexistence of quartz and kaolinite in ash particles (Figs. 7B, D, and E) can constrain the crystallization temperature, by the following reaction of the Al 2 O 3 -SiO 2 -H 2 O system (Hemley et al., 1980;Chaterjee et al., 1984) : Alunite is the most abundant sulfur-bearing mineral, which is associated with silica minerals. Alunite and P-bearing alunite are the most common alunite species in the volcanic products. The P-bearing alunite and associated alunite (Figs. 8A and D) were possibly in uenced by magmatic components at hypogene environment because phosphate in alunite is related to phosphorous released from magmatic apatite (Stoffregen and Alpers, 1987). The zoned alunite with P-rich zone in the veinlets implies pulses of magmatic components into the acidic hydrothermal environment.
Jarosite is associated with goethite and alunite-jarosite intergrowth (Figs. 4 and 7C), which is mostly interpreted as pyrite-alunite intergrowth in origin Watanabe et al., 1997). The similarity in sulfur isotopic composition suggests that the jarosite was formed by the oxidation of pyrite (e.g., .

Sulfur isotope variation in alunite Temporal variation
Heterogeneity of sulfur isotopic composition is clearly shown in alunite from this studied samples. The variation of sulfur isotopic composition for alunite reasonably suggests that the volcanic products contain various alunite minerals formed at different stages and conditions. In volcanic environments, the variations of the sulfur isotopic composition are caused mainly by SO 2 disproportionation (Imai et al., 2007;Maruoka, 2016, 2019 (Ohmoto and Rye, 1987;Ohmoto and Goldhaber, 1999;Zimbelman et al., 2005). Consequently, a pair of sulfates (e.g., heavier gypsum or anhydrite) and sul de (e.g., lighter pyrite) can be formed in a magmatic-hydrothermal environment (Imai et al., 2010;Maruoka, 2016, 2019). In this study, only TP15 in TPT 4 shows a pair of alunite and pyrite and indicates a possibility of sulfate-sul de equilibrium condition (Ohmoto and Rye, 1987;Ohmoto and Goldhaber, 1997). Therefore, a pair of isotopically heavy alunite (+20 ‰) and light pyrite (-5 ‰) in TP15 of TPT 4 eruption implies the timing of development of an aqueous magmatic-hydrothermal system. In contrast with TP15 in TPT 4, most other samples contain only alunite as a sulfur mineral phase with a range of δ 34 S from +5 to +10 ‰. This range indicates the SO 2 disproportionation had not proceeded enough to form the alunite-pyrite pair with the isotopic fractionation. This case is suitable at the oxidized stage of subvolcanic hydrothermal system with much abundance of magmatic SO 2 vapor (e.g., Satsuma-Iwojima in, Hendenquist and Taran, 2013). Moreover, after TPT 4, the sulfur isotopic compositions of alunite in the layers concentrate on 5 ‰ of δ 34 S, which is lighter than the values before TP11 of TPT 4. This stratigraphic variation of sulfur isotope might indicate an increase of input uxes of magmatic vapor with their own isotopic composition (Fig. 10).

Sulfur isotope geothermometer
Sulfur isotope geothermometry Rye 1979, Ohmoto andGoldhaber, 1997) was applied for two pairs of sulfur isotopic compositions of alunite and pyrite from the TP15 in Unit 4, assuming the alunite and pyrite were in equilibrium ( Table 1). The calculated equilibrium temperatures were 232 ℃ and 240 ℃, respectively. The values are consistent with the temperatures estimated from the P-free alunite and the silica alteration (quartz and kaolinite). On the other hand, the P-bearing-alunite and huangitealunite solid solution must have precipitated at higher temperatures. The zoned texture of alunite (Figs. 7B, 8A, and B) represents repeated injections of magmatic steam or gas (Aoki, 1991), which could result in the sulfur isotopic compositions similar to that of the magmatic steam or gas.

Comparison between Tangkuban Parahu and other volcanoes
We compared the δ 34 S values of alunite with several available δ 34 S datasets from Mt. Muine, Japan (Watanabe and Ohta, 2005), and Kawah Ijen, Indonesia (Scher et al., 2013) (Fig. 10). Watanabe and Ohta (2005) argued that the magmatic-steam or gas alunite has sulfur isotopic composition below +20 ‰, whereas the δ 34 S values higher than +20 ‰ is falling into magmatic-hydrothermal alunite. At the Kawah Ijen, the alunite mostly shows sulfur isotopic composition within a range of +18 to +20 ‰ and classi es as magmatic-hydrothermal origin (Scher et al., 2013). These suggest that alunite in the Tangkuban Parahu volcanic products were mostly formed by the process of magmatic-steam or gas (Fig. 10).
Nevertheless, a sample with the highest δ 34 S value in TP15 of TPT4 is classi ed as magmatichydrothermal alunite.

Implication
Hydrothermal minerals in the volcanic products represent acid-sulfate or advanced argillic alteration (e.g., Arribas et al., 1995;Silitoe, 2010). Mineral assemblages represent the temperature conditions along the conduit. Cristobalite and opal alteration were formed at a temperature of 100-150 ºC (Fournier, 1985) which is the temperature range in the near-surface condition below the crater (e.g., Lowernstern et al., 2018). Quartz-kaolinite alteration occurs at temperatures lower than 260 ºC. The sulfur isotopic temperatures for the TPT15 are 232 and 240ºC. These temperatures indicate that most of the ash particles were ejected from the hydrothermal system at temperatures below 300 ºC. In addition, the presence of jarosite and goethite implies supergene alteration, indicating atmospheric oxygen entrainment to the hydrothermal system near the surface.
From the mineralogy of volcanic products, we can conclude that the alteration at the summit of Tangkuban Parahu volcano represents the nature of upper-level high sul dation epithermal ore deposits (Hedenquist and Lowernsten, 1994;Arribas et al., 1995;Hedenquist et al., 1998). The occurrence of chalcopyrite and enargite in the volcanic product ( Fig. 7E) is also the similarity to these ore systems.
The stratigraphic sulfur isotopic variation implies increase in in ux of magmatic vapor into the hydrothermal system. The abrupt increase in the vapor ux is consistent with the activity of magma at the TPT 4 when a series of phreatomagmatic eruptions occurred (Angkasa et al., 2019). Alunite-group minerals also indicate pulses of magmatic components into the hydrothermal system. This study showed that sulfur isotope of volcanic products is a useful method to detect the magmatic contribution on hydrothermal system.

Conclusions
In this work, the mineralogy and sulfur isotopic composition were observed and documented from the hydrothermally altered lithic ash particles, part of the Tangkuban Parahu proximal volcanic products. The followings are the conclusions: 1. Mineral assemblage exhibits a typical acid-sulfate and advanced argillic alteration with a temperature range from ~100 to ~260 ºC, which is almost identical with the temperature of 230-240 ºC calculated from the sulfate-sul de 34 S fractionation equation.
2. The sulfur isotope and occurrence of jarosite indicate the supergene alteration associated with atmospheric oxygen entrainment to oxidize pyrite to jarosite.
3. Ore minerals such as enargite and chalcopyrite represent the nature of upper-level high sul dation system at the shallow volcanic edi ce.
4. Stratigraphic sulfur isotopic variation represents in ux of magmatic gas into the hydrothermal system.
5. Sulfur isotope signature of the volcanic products has a paramount importance to evaluate the conditions of subvolcanic hydrothermal system, magmatic-hydrothermal interaction, and the origin of steam-blast eruptions at volcanoes that contain subvolcanic hydrothermal systems.

Declarations
Availability of data and materials Not applicable

Competing interests
All authors declare that we have no competing interests. Zimbelman DR, Rye RO, Breit GN (2005) Tables   Due to technical limitations, table 1, 2 is only available   Photographs are showing outcrop of the analyzed samples at Tangkuban Parahu, Indonesia. A) White to grey volcanic products are characteristics of proximal tephra succession. Units were divided by the presence of very-thin paleosols (>5 cm) and continuous successive volcanic products. B) Thicknesses range from laminae (>1cm) to bedding. C) The upper portion of studied stratigraphy (e.g., Unit 7-10) is predominantly composed of poorly sorted, heterolithics, and matrix-supported, whereas the lower stratigraphic section consisting mainly of well-sorted ne-ash to lapilli deposits.  Composite log of the proximal tephra succession, together with compiled mineral assemblages of the studied ash particles from x-ray diffraction and petrographic observation. Hydrothermal minerals are associated with silica polymorph (quartz, cristobalite, and tridymite), alunite, and kaolinite, whereas jarosite and goethite crystals were rarely found in all analyzed samples. Magmatic minerals consist of augite, plagioclase, sanidine, rutile, and orthopyroxene/enstatite. A total of 351 (see Table. 1) ash particles was observed using optical and electron microscopes. Altered and pristine magmatic pyroclast Page 24/30 were found in several volcanic products (Unit 1, 2, 4, and 6). Hydrothermal minerals are predominantly comprise of a large number of silica minerals associated with alunite, alunite±kaolinite, and alunite+jarosite+goethite together with partly altered glassy magmatic pyroclasts.

Figure 5
Backscattered electron images showing the example of partly altered glassy magmatic juveniles in Unit 3 and 4. A) Partly altered glassy magmatic juveniles mainly composed of plagioclase, clinopyroxene, and interstitial glass along with vesicle and bubble textures. Most of the magmatic minerals (e.g., plagioclase) are fractured and possibly lled by hydrothermally altered minerals (e.g., silica). B) BSE image shows colloform-like banding of thinly altered interstitial glass in contact with lithics.
Hydrothermally altered lithics are occurring as a replacement of glassy groundmass as well as microlite of magmatic minerals.

Figure 6
Backscattered electron images showing the textural features of opal and opal-cristobalite±alunite±kaolinite in ash particles, including the Raman Spectral from selected point analyses from several ash particles. A) Opal alteration was found on vesiculated ash particles. Vesicular textures were lled with silica minerals. Magmatic minerals were replaced by opal, where the relict of magmatic minerals (A, B) remains at the core of crystals. C). Graph showing the two representatives Raman spectrum of cristobalite and opal in ash particles (B, D). D) BSE image shows opal-cristobalite after plagioclase, whereas silica-alunite±kaolinite occurs as a groundmass alteration. F) Opal-cristobalitesilica-alunite sporadically replaced the magmatic minerals in dense ash particles. Most of the hydrothermal minerals occur as pseudomorphic textures.

Figure 7
Backscattered electron images showing the textural features of quartz-alunite±kaolinite, alunite-jarosite, and quartz alteration assemblages in ash particles. A) Quartz-alunite±kaolinite alteration is also associated with a cluster of quartz±pyrite±anatase assemblage along with pervasive quartz alteration. B) Alunite occurs as equant crystal shape, which is associated with circular kaolinite crystal clusters, anatase, and silica mineral. C) Alunite-jarosite was found in the subangular to subrounded ash particles.
Jarosite occurs as alunite-jarosite intergrown texture. Both minerals are associated with silica minerals that mainly occurs as pore-lling minerals. D) BSE image shows a pervasive quartz alteration of subangular to subrounded ash particles. A vuggy texture develops at the edge of ash particles, where the anhedral and elongated shape pyrite lled up the pore space of vuggy textures. E) BSE image shows an example of ash particle with pervasive quartz alteration. Quartz alteration is associated with anatase, enargite, chalcopyrite, titanomagnetite, and magnetite. F) Graph showing the representative of Raman spectrum of anatase in alunite and quartz alteration. . BSE image band also shows the colloform-like banding of early-stage veinlet deposition, which is located close in contact with holocrystalline-porphyritic textured hypabyssal rocks. C) The second veinlet consists of quartz and centreline pyrite crystal bleps within the cluster of alunite crystals. P-bearing alunite mainly occurs along the rim of quartz-pyrite veinlet.

Figure 9
Graphs showing the elemental spectrum of EDS analysis. A) EDS spectrum displays the analytical results for the occurrence of chalcopyrite in Fig. 7E. It should be noted that Si spectrum represents the quartz x-ray spectrum on host alteration mineral. B) The EDS spectrum shows the presence of P-bearing alunite in alunite vein (Fig. 8D).

Figure 10
Graph showing the plot of the sulfur isotopic composition of alunite, pyrite, and jarosite against the studied stratigraphic unit in Angkasa et al., 2019. The δ34S values are compared to δ34S values from sul de and sulfate samples at Mt. Muine (Watanabe and Ohta, 2005) and Kawah Ijen (Scher et al., 2013).

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download.