Genesis and Evolution of Hydrothermal Fluids in the Formation of the High-Grade Hishikari Gold Deposit: Carbon, Oxygen, and Sulfur Isotopic Evidence

Abstract

The Hishikari low-sulfidation epithermal gold (Au) deposit in Kyushu, Japan, is world-famous for its premium ore. It has been hypothesized that magmatic contributions to the hydrothermal fluid during early stages of mineralization is possible, even if the hydrothermal fluids for many Au occurrences near the Hishikari deposit are of meteoric origin and are influenced by basement sedimentary rocks. The purpose of this study is to obtain constraints on the genesis and evolution of hydrothermal fluids in the formation of the high-grade Hishikari Au deposit by carbon and oxygen isotope ratios of calcite-bearing samples. Since the microanalysis of carbon and oxygen isotope ratios in every 12 μm of the calcite-bearing sample along the growth direction (corresponding to 10 years of the Hishikari mineralization) scatter in a particular range, the fluid evolution might not be a gradual change from a magmatic to a meteoric origin. Alternatively, a rapid turnover of two fluids might be happening locally. The average sulfur isotope ratio of hydrothermal pyrite is similar to that of the adjacent magma. However, according to the secondary ion mass spectrometry (SIMS) microanalysis, local pyrite with extremely low sulfur isotope ratios may interact with basement sedimentary rocks. Unlike other epithermal Au deposits in the vicinity, rapid local mixing of the magmatic-origin deep fluid and meteoric-origin fluid reacted with organic matter containing basement sedimentary rocks might cause gold precipitation at the Hishikari deposit.

1. Introduction

It is crucial to research the Hishikari gold (Au) deposit in Kyushu, Japan, as an example of vein-type high-grade Au deposits because of its highly regarded high-grade ore. The Sumitomo Metal Mining Co., Ltd. has been mining the Hishikari deposit since 1985. The Hishikari mine has produced 260 metric tons of Au, and the ore reserve amounts to 160 metric tons. The Au production in 2021 is 6 metric tons [1]. The average grade of ore is 34 g/t Au, which is significantly higher than those of neighboring and average mines worldwide. An important question to ask is why the Hishikari deposit is of a higher grade among Au deposits in the same district. The early-stage hydrothermal fluid in the Hishikari deposit was enriched in ¹⁸O over local meteoric water, despite its low-sulfidation epithermal deposit [2]. Several studies [2,3,4,5,6] have pointed out the possibility of magmatic contribution to the hydrothermal fluid in the earlier stage of mineralization. These previous studies focus on the genesis and evolution of hydrothermal fluids in forming the high-grade Hishikari Au deposit.

Fluid-rock reactions play a vital role in the carbon and oxygen isotope budgets of geologic settings of ore-forming hydrothermal systems in the Earth’s crust. Although many studies on isotopes in calcites have been undertaken, few researchers have expressed an interest in utilizing the isotope data of hydrothermal calcite. This is partly because calcite often occurs in the last stage of hydrothermal mineralization; therefore, a calcite vein is considered a barren vein in mining districts. However, fertile calcite veins may also exist in the exact location of barren calcite veins. Therefore, this study proposes using calcite’s isotope data. Although oxygen isotope ratios of quartz from the Hishikari veins have been reported [2,3,4,5,6], the veins contain silicate minerals such as adularia other than quartz. Therefore, merely analyzing pure quartz is not an easy task. However, calcite isotope data can be obtained using acid from quartz veins containing other silicate minerals.

Furthermore, the calcite isotope measurements still include specific details about the fluid that coexisted with it. Therefore, a set of calcite carbon and oxygen isotope ratios has better information than the oxygen isotope ratio of vein quartz. Consequently, isotope ratio mass spectrometry (IRMS) was used to examine the carbon and oxygen isotope ratios of samples that contained calcite.

On the other hand, pyrite is ubiquitous in the Hishikari veins. Therefore, Morishita et al. [7,8] conducted a microanalysis of Au and As concentrations in pyrite grains to indicate that the Au concentration positively correlates with the As concentration in a small area. Furthermore, sulfur isotope ratios of pyrite help constrain the source of sulfur. Therefore, sulfur isotope ratios of pyrite were measured using secondary ion mass spectrometry (SIMS).

The purpose of this study is to obtain constraints on the genesis and evolution of the hydrothermal fluids in the formation of the high-grade Hishikari Au deposit. Additionally, the isotopic characteristics of the fluid in the Hishikari deposit will be discussed.

2. Geology and Mineralogy in the Hishikari Mining Area

2.1. Geologic Setting and Au Deposits

Kyushu is located at the eastern margin of the Eurasian plate facing the Philippine Sea plate. Numerous epithermal Au deposits from the Pliocene to Pleistocene are found in Late Cenozoic subaerial andesitic to dacitic volcanic rocks that resulted from the subduction of the Philippine Sea plate beneath the Eurasian plate (Figure 1). The Hokusatsu district of southwestern Kyushu (Figure 1) has the highest Au production in Japan. Most of the ore deposits in this district, such as the Hishikari and Kushikino deposits, sit above a typical basement of the Shimanto Group.

The Hishikari low-sulfidation epithermal Au deposit (Figure 1) is the largest in Japan. Hosono et al. [9] determined the Sr–Nd–Pb isotope compositions of Late Pliocene to Pleistocene volcanic rocks around the Hishikari deposit to elucidate their source characteristics of magma related to gold mineralization. The basement rocks in the Hishikari mining area are sedimentary rocks from the Lower Shimanto Group (Cretaceous accretionary wedge), which are visible underground but do not outcrop close to the deposit. The Hishikari Lower Andesites consist mainly of andesitic pyroclastic rocks. They cover the basement sedimentary rocks unconformably and have been dated by the K–Ar method to be from 1.62 ± 0.09 to 0.98 ± 0.10 Ma for 15 whole-rock samples from within 2 km of the Hishikari mine [10,11]. Several studies have previously described the Hishikari deposit’s geology (e.g., [9,10,11,12,13,14,15,16]).

2.2. Gold-Bearing Veins and Mineral Paragenesis

The Hishikari deposit consists of the Sanjin, Honko, and Yamada ore zones (Figure 2). Each ore zone has multiple vein swarms that extend roughly 1 km from NE to SW. The Hishikari veins strike N50° E and dip steeply north (70–90°). The Sanjin and Honko ore zones comprise high-grade Au-bearing quartz veins in the basement sedimentary rocks of the Cretaceous Shimanto Group and in the andesitic rocks of the Hishikari Lower Andesites that overlie the basement sedimentary rocks. However, relatively low-grade Au-bearing quartz veins in the Yamada ore zone occur in the andesitic rocks. K–Ar ages show that the Hishikari deposit’s mineralization started at 1.15 Ma and ended at 0.60 Ma [11]. Sanematsu et al. [18] determined the ⁴⁰Ar/³⁹Ar ages of the Hishikari adularia–quartz veins. The mineralization ages in the Honko and Sanjin zones range from 1.04 to 0.75 Ma, whereas the majority of results are focused in the range of 1.01–0.88 Ma and those in the Yamada zone range from 1.21 to 0.64 Ma [18].

The Hishikari veins consist mainly of quartz, adularia, and smectite. The metallic minerals in the Hishikari veins are pyrite, marcasite, electrum, naumannite, aguilarite, pyrargyrite, and chalcopyrite [10], of which pyrite is the most abundant. The host rocks also contain hydrothermal pyrite [7,8]. Additionally, the veins contain minor amounts of calcite and truscottite [10]. The homogenization temperatures for quartz in veins within the basement sedimentary rocks average 213 °C, with most values in the range of 195–230 °C [10]. The strontium and carbon isotopic ratios of calcite reveal that the Hishikari ore fluid was affected by the sedimentary basement rocks [19]. Bladed quartz (lamellar quartz) is found everywhere in the Hishikari deposit [10,20]. When initially produced, calcite reacts with later low-temperature fluids; and it dissolves because calcite’s solubility rises as the hydrothermal fluid’s temperature falls, leaving cavities for later quartz to precipitate as bladed quartz [10,20]. Bladed quartz signifies that bladed calcite was once deposited by boiling fluid and then dissolved during cooling. Boiling occurred during the Hishikari mineralization based on the observations of fluid inclusions [10,14,20]. However, other studies [4,21] do not recognize the evidence of boiling in the fluid inclusion observations, suggesting that not all areas of boiling were subject to it. Calcite precipitation from the hydrothermal fluid occurs either with increasing temperature or pH or with decreasing fugacity of CO₂ [22]. Among them, loss of CO₂ because of boiling might happen in the Hishikari hydrothermal fluid, and cause calcite precipitation.

Additionally, mixing fluids might increase temperature or pH, resulting in calcite precipitation. Therefore, calcite might initially form by boiling or mixing fluids. Ore fluids that formed the banding structure of veins, have passed through the veins and they might dissolve most of the calcite precipitated from earlier fluids, because the solubility of calcite increases as the temperature drops. Because of this, calcite is rarely observed in the veins of the Hishikari deposit, and there haven’t been many investigations on the carbon and oxygen isotopes of calcite. In this study, we attempted to find as much calcite as possible in the Hishikari deposit and we analyzed the calcite-bearing samples’ carbon and oxygen isotope ratios. The presence of ubiquitous pyrite shows that the Hishikari hydrothermal system was under reducing conditions. The mixing of deep fluids with shallow groundwater is thought to have considerably altered the temperature and oxidation conditions of the hydrothermal fluids at the boundary between the basement rocks and the overlying andesites [10].

The Hishikari veins frequently have symmetrical formations, which suggests that vein minerals near the wall rock precipitate earlier than the minerals inside the vein. The outer band close to the wall rock consists mainly of adularia and quartz, and the innermost band in the center consists mainly of quartz [14]. However, in the Hishikari veins, asymmetrical formations produced by cutting with later-stage veins are not uncommon. According to the observed crosscutting connections of the veins, Sekine et al. [16] divided the vein system of the Hishikari deposit into early and late veins according to the observed crosscutting relationships of veins. The late veins commonly display symmetrical banding, whereas banding in the early veins is often obscured.

3. Materials and Methods

3.1. Materials

3.1.1. Calcite-Bearing Hand Specimens

Calcite-bearing vein samples obtained were investigated for their mineral compositions using an X-ray diffractometer (XRD) SmartLab (Rigaku, Tokyo, Japan) at Shizuoka University. Figure 2 and

Table 1. Descriptions and isotope data of calcite-bearing vein samples from the Hishikari deposit.

Sample No.	Locality *1	Vein	Level *2	Mineral Composition (%) *3	δ13C	δ18O
					(‰)	(‰)
31001a	70W111SE-6E	Seisen-6	70 mL	Calcite (100)	−10.3	6.1
31001b	70W111SE-6E	Seisen-6	70 mL	Quartz (62), Adularia (12), Calcite (26)	−4.3	10.6
31001c	70W111SE-6E	Seisen-6	70 mL	Quartz (59), Adularia (6), Calcite (35)	−7.9	8.1
31002c	25E57S42HW	Shosen S42	25 mL	Quartz (5), Calcite (95)	−9.2	3.8
31002d	25E57S42HW	Shosen S42	25 mL	Quartz (32), Calcite (68)	−8.6	5.2
31003-1e	20W112SE-2W	Seisen-6	20 mL	Quartz (93), Calcite (7)	−3.6	12.4
31003-2a	20W112SE-2W	Seisen-6	20 mL	Quartz (90), Calcite (10)	−4.7	10.5
31003-2b	20W112SE-2W	Seisen-6	20 mL	Quartz (6), Calcite (94)	−2.6	11.8
31003-2c	20W112 SE-2W	Seisen-6	20 mL	Quartz (45), Adularia (21), Calcite (34)	−3.3	11.8
31506b	20W112SE-2W	Seisen-6	20 mL	Quartz (66), Calcite (34)	−6.5	11.1
31506c	20W112SE-2W	Seisen-6	20 mL	Quartz (57), Calcite (43)	−23.4	17.9
31004a	40 No.2 Vent		40 mL	Quartz (75), Calcite (25)	−4.8	7.9
31004b	40 No.2 Vent		40 mL	Quartz (71), Calcite (29)	−5.5	9.8
31004c	40 No.2 Vent		40 mL	Quartz (35), Calcite (65)	−5.6	8.2
31005c	70E53 cross cut	Kinsen-1-1	70 mL	Quartz (36), Calcite (64)	−4.5	10.6
31615c	70E53 cross cut	Kinsen-1-1	70 mL	Quartz (69), Calcite (31)	−25.1	17.1
31610b	100E28RY-6E	Ryosen-6	100 mL	Calcite (100)	−6.2	5.7
31611b	100E28RY-6E	Ryosen-6	100 mL	Quartz (70), Adularia (13), Calcite (17)	−5.2	14.6
20100317	25E60KE-3-1E	Keisen-3-1	25 mL	Quartz, Calcite: See Table 2	See Table 2, Table 3 and Table 4.

The vein width is not shown because each calcite-containing sample is from a branch veinlet of the main vein. Several calcite-bearing samples are obtained by the mine geologists. Branch signs like a, b, etc. show that they are from the same hand specimen. Different sample numbers with the same locality were collected at different times. *1: E or W indicates the direction of the drift. *2: “mL” corresponds to the altitude above sea level. *3: Approximate mineral composition was obtained from the XRD calibration curve.

show calcite-bearing samples’ locality and mineral composition. The approximate abundance of calcite was obtained from the XRD calibration curves. Electrum is very fine-grained for microscopic observations. Photos of representative vein samples are shown in Figure 3.

3.1.2. A Calcite-Dominant Vein Sample for Microanalysis

The Hishikari deposit’s high-grade vein sample 20100317 (Figure 4a), taken from a 25 mL E60 Keisen 3-1E, contains continuous calcite in the growth direction. The vein width is about 50 cm. The Keisen No. 3 vein swarms are representative fertile veins in the Sanjin ore zone [8,16]. Therefore, vein sample 20100317 is suitable for investigating the fluid evolution in the Hishikari ore-forming hydrothermal system, and no other appropriate sample has been found so far. The vein was expanding from right to left in Figure 4a because its proper end is close to the wall rock, and its left end corresponds to the vein’s center. This sample was cut into 10 mm widths and named for each from 1743 to 1749 (10 mm-width sample set), as shown in Figure 4b and Table 2. Then 1761, 1762, and 1763 samples were further cut to a width of 0.9 mm (Figure 4c) with a cutting machine, Minitom (Struers, Denmark) at Shizuoka University and obtained 176-1 to 176-21 sections (0.9 mm-width sample set; Table 3). The 1763 sample corresponds to the same stage as 1743 (Figure 4b). Therefore, compared to the 10 mm-width sample set, the 0.9 mm-width sample set (1761–1763) indicates earlier stages than the 10 mm-width sample set. The blue lines in Figure 4c represent space, which corresponds to the thickness of the blade, 0.3 mm, so the sample interval in the growth direction is 1.2 mm. The quartz-adularia vein’s approximate calcite abundance is determined via the XRD calibration curve (Table 3). The 176-1 section consists of 69% of quartz and 31% of adularia (determined by XRD calibration curve) with a minor amount of fine-grained pyrite and clay minerals, which causes the gray color in the early-stage mineral precipitations.

Table 2. Mineral composition and isotope data of calcite-bearing 20100317 vein sample (10 mm-width sample set; Figure 4b) from 25 mL E60 KE-3-1E.

Sample No.	Mineralization Order	Mineral Composition (%) *	δ13C (‰)	δ18O (‰)
1743	(Early)	Quartz (83), Calcite (17)	−7.7	4.3
1744		Quartz (19), Calcite (81)	−6.5	3.5
1745		Quartz (29), Calcite (71)	−6.2	3.4
1746		Quartz (86), Calcite (14)	−6.3	3.9
1747		Quartz (94), Calcite (6)	−4.8	8.2
1748	(Late)	Quartz (100)	-	-
1749	Center of the vein	Quartz (100)	-	-

*: Approximate mineral composition was obtained from the XRD calibration curve.

Table 2. Mineral composition and isotope data of calcite-bearing 20100317 vein sample (10 mm-width sample set; Figure 4b) from 25 mL E60 KE-3-1E.

Sample No.	Mineralization Order	Mineral Composition (%) *	δ13C (‰)	δ18O (‰)
1743	(Early)	Quartz (83), Calcite (17)	−7.7	4.3
1744		Quartz (19), Calcite (81)	−6.5	3.5
1745		Quartz (29), Calcite (71)	−6.2	3.4
1746		Quartz (86), Calcite (14)	−6.3	3.9
1747		Quartz (94), Calcite (6)	−4.8	8.2
1748	(Late)	Quartz (100)	-	-
1749	Center of the vein	Quartz (100)	-	-

*: Approximate mineral composition was obtained from the XRD calibration curve.

Finer samples in the subsequent work were obtained to increase the time resolution. Although cutting the vein sample on a micron scale is impractical, Shizuoka University’s biological microtome (FX-801N, Yamato Kohki Industrial Co., Ltd., Saitama, Japan) can be used to scrape off a small portion of the calcite-rich soft sample. The 176-20 and 176-21 sections (Figure 4c) contain more than 70% of calcite, and they are soft enough to scrape off calcite-dominant powder from the samples with a stainless steel blade of the microtome. The average scraped sample size is roughly 0.4 mg, corresponding to a sample width in the original sample’s growth direction of about 12 μm. The continuous samples (12 μm-width sample set) in the growth direction obtained from 176-20 and 176-21 sections are listed in Table 4. The scraped thickness is calculated from the weight of the powder scraped, and Table 4′s distance from the microtome cutting’s starting point (the ‘origin’ that is the line separating 176-19 and 176-20) is displayed. The number attached to the original sample number (176-19 and 176-20) indicates the order of crystallization; namely, 176-20-1 is the earliest end of the 176-20 section, and 176-20-12 is the latest grown sample, about 0.14 mm from the origin. 176-20 section with a thickness of about 0.76 mm remained unscraped since the sample was broken during the subsequent cutting. 176-21 section was divided into 17 samples, from 176-21-1 to 176-21-17. Each scraped size was similar to that for 176–20, and the distance from the same ‘origin’ is shown in Table 4. The 176-21 section was broken during the 18th cutting after 17 sample powders were obtained. Notably, 29 mineral samples were collected, despite the difficulty of cutting vein samples with a biological microtome. Be aware that the 176-20-12 and 176-21-1 sample differ by approximately 1.06 mm (the blade thickness and the remainder of the 176-20 sample).

Table 3. Calcite abundance and isotope data of calcite-bearing 20100317 vein sample (0.9 mm-width sample set; Figure 4b,c).

Sample No.	Mineralization Order	Calcite Abundance *	δ13C (‰)	δ18O (‰)
176-1	Wall rock side	0%
176-2	(Early)	0%
176-3		0%
176-4		0%
176-5		4%	−5.0	6.0
176-6		0%
176-7		0%
176-8		16%	−6.9	4.7
176-9		5%	−6.3	3.7
176-10		3%	−7.2	3.8
176-11		9%	−7.7	4.4
176-12		17%	−7.5	3.0
176-13		31%	−5.4	4.2
176-14		25%	−6.3	4.5
176-15		17%	−6.5	5.6
176-16		11%	−5.6	5.8
176-17		6%	−4.7	7.0
176-18		33%	−8.4	5.5
176-19		73%	−7.4	5.2
176-20		70%	−7.2	4.8
176-21	(Late)	72%	−6.1	4.5

*: Approximate calcite abundance in quartz-adularia vein was obtained from the XRD calibration curve.

Table 3. Calcite abundance and isotope data of calcite-bearing 20100317 vein sample (0.9 mm-width sample set; Figure 4b,c).

Sample No.	Mineralization Order	Calcite Abundance *	δ13C (‰)	δ18O (‰)
176-1	Wall rock side	0%
176-2	(Early)	0%
176-3		0%
176-4		0%
176-5		4%	−5.0	6.0
176-6		0%
176-7		0%
176-8		16%	−6.9	4.7
176-9		5%	−6.3	3.7
176-10		3%	−7.2	3.8
176-11		9%	−7.7	4.4
176-12		17%	−7.5	3.0
176-13		31%	−5.4	4.2
176-14		25%	−6.3	4.5
176-15		17%	−6.5	5.6
176-16		11%	−5.6	5.8
176-17		6%	−4.7	7.0
176-18		33%	−8.4	5.5
176-19		73%	−7.4	5.2
176-20		70%	−7.2	4.8
176-21	(Late)	72%	−6.1	4.5

*: Approximate calcite abundance in quartz-adularia vein was obtained from the XRD calibration curve.

Table 4. Microanalyses of 176-20 and 176-21 cutoff (12 μm-width sample set; Figure 4c) from 20100317 sample.

Sample No.	Distance (μm) from the ‘Origin’ *	Mineralization Order	δ13C (‰)	δ18O (‰)
176-20-1	7	Early	−6.3	8.4
176-20-2	15		−8.4	5.9
176-20-3	26		−9.1	5.0
176-20-4	37		−7.7	4.8
176-20-5	42		−4.5	9.9
176-20-6	57		−5.5	7.4
176-20-7	76		−8.2	5.2
176-20-8	87		−7.7	6.5
176-20-9	107		−9.5	5.1
176-20-10	121		−8.0	5.7
176-20-11	137		−6.9	6.5
176-20-12	144		−2.5	10.9
176-21-1	1207		−9.1	5.0
176-21-2	1212		−7.9	3.6
176-21-3	1222		−7.8	5.3
176-21-4	1234		−8.5	4.5
176-21-5	1241		−8.5	4.5
176-21-6	1247		−8.4	4.8
176-21-7	1255		−9.1	5.4
176-21-8	1274		−9.1	5.0
176-21-9	1285		−8.2	4.3
176-21-10	1301		−7.2	5.7
176-21-11	1315		−8.0	4.9
176-21-12	1319		−6.9	4.9
176-21-13	1324		−6.9	4.7
176-21-14	1343		−7.4	6.0
176-21-15	1361		−7.6	5.0
176-21-16	1381		−5.6	7.5
176-21-17	1388	Late	−7.4	5.2

*: The origin is the boundary between 176-19 and 176-20.

Table 4. Microanalyses of 176-20 and 176-21 cutoff (12 μm-width sample set; Figure 4c) from 20100317 sample.

Sample No.	Distance (μm) from the ‘Origin’ *	Mineralization Order	δ13C (‰)	δ18O (‰)
176-20-1	7	Early	−6.3	8.4
176-20-2	15		−8.4	5.9
176-20-3	26		−9.1	5.0
176-20-4	37		−7.7	4.8
176-20-5	42		−4.5	9.9
176-20-6	57		−5.5	7.4
176-20-7	76		−8.2	5.2
176-20-8	87		−7.7	6.5
176-20-9	107		−9.5	5.1
176-20-10	121		−8.0	5.7
176-20-11	137		−6.9	6.5
176-20-12	144		−2.5	10.9
176-21-1	1207		−9.1	5.0
176-21-2	1212		−7.9	3.6
176-21-3	1222		−7.8	5.3
176-21-4	1234		−8.5	4.5
176-21-5	1241		−8.5	4.5
176-21-6	1247		−8.4	4.8
176-21-7	1255		−9.1	5.4
176-21-8	1274		−9.1	5.0
176-21-9	1285		−8.2	4.3
176-21-10	1301		−7.2	5.7
176-21-11	1315		−8.0	4.9
176-21-12	1319		−6.9	4.9
176-21-13	1324		−6.9	4.7
176-21-14	1343		−7.4	6.0
176-21-15	1361		−7.6	5.0
176-21-16	1381		−5.6	7.5
176-21-17	1388	Late	−7.4	5.2

*: The origin is the boundary between 176-19 and 176-20.

3.1.3. Pyrite Samples

Pyrite samples were collected from the main veins of the three ore zones of the Hishikari deposit. The grain size of pyrite is fine, it is mostly between 10 and 30 µm. Figure 2 and

Table 5. Locality and the sulfur isotope ratio of pyrite from the Hishikari deposit.

Ore Zone	Sample Name	Vein *1	Level *2	Sample Locality	Host Rock	Analysis Spot	Analysis No.	δ34S (‰)
Sanjin	NS-10347	Keisen No.3	40 mL	E21	andesite	13-52	SMB3*76.ais	−0.4
Sanjin	NS-10570	Keisen No.3	25 mL	E45	shale	6-26	SMB3*73.ais	2.1
Sanjin	NS-10813	Keisen No.3	10 mL	E45	shale	1-2	SMB3*69.ais	−10.2
						1-3	SMB3*70.ais	−15.1
						1-1	SMB3*71.ais	−16.4
						2-5	SMB3*72.ais	−10.1
Sanjin	NS-11302	Keisen No.3	−5 mL	E45	shale	1-1	SMB2*47.ais	−2.1
						2-1	SMB2*48.ais	−16.5
						2-2	SMB2*49.ais	−12.1
						3-1	SMB2*50.ais	1.7
						4-1	SMB2*51.ais	3.5
Honko	98121705	Zuisen No.1	25 mL + 2sl *3	W33	andesite	1-18 *4	SMB3*80.ais	−1.4
						1-1	SMB3*81.ais	0.5
						2-4	SMB3*82.ais	−3.4
						2-6	SMB3*83.ais	0.2
						2-19 *4	SMB3*84.ais	1.0
						4-13	SMB3*86.ais	−0.4
						4-15	SMB3*87.ais	−2.0
Honko	98121807	Hosen No.1	55 mL + 2sl	W1	shale	1-1	SMB1*30.ais	0.4
				(E1) *5		1-2	SMB1*31.ais	0.1
						1-3	SMB1*32.ais	−1.4
						2-2	SMB1*33.ais	−1.6
						2-3	SMB1*34.ais	0.6
						2-4	SMB1*35.ais	−0.7
						2-5	SMB1*36.ais	−1.2
Honko	98121806	Hosen No.1	55 mL + 2sl	W1	shale	1-1	SMB3*92.ais	0.3
				(E8) *5		1-2	SMB3*93.ais	0.6
						1-4	SMB3*94.ais	−0.1
						1-6	SMB3*95.ais	1.9
						1-9	SMB3*96.ais	1.5
						1-10	SMB3*97.ais	2.1
Honko	98121805	Shale	55 mL + 2sl	W1		3-2	SMB1*21.ais	3.2
		Host rock for the Hosen No.1 vein		(E8) *5		3-3	SMB1*22.ais	2.2
						3-5	SMB1*23.ais	2.7
						1-1	SMB1*24.ais	2.0
						4-1	SMB1*25.ais	0.7
						5-1	SMB1*26.ais	1.3
Yamada	98121810	Yusen No.1-2	30 mL	W85	andesite	2-23	SMB3*101.ais	1.2
						2-22	SMB3*102.ais	0.7
						2-18	SMB3*103.ais	2.3
						2-17	SMB3*104.ais	1.7
						2-13	SMB3*105.ais	−1.0
						3-30	SMB3*106.ais	4.5
						3-36	SMB3*107.ais	3.1
						3-41	SMB3*108.ais	4.3
						3-47	SMB3*109.ais	1.5
						3-48	SMB3*110.ais	4.4
Yamada	NS-11167	Yusen No.1-5	40 mL	W95	andesite	2-1	SMB2*58.ais	4.0
						2-2	SMB2*59.ais	2.9
						2-3	SMB2*60.ais	3.4
						6-2	SMB2*64.ais	0.8
						7-1	SMB2*65.ais	−2.0

*1: Samples are from quartz veins except for one sample (98121805) from the basement shale. *2: “mL” corresponds to the altitude above sea level. *3: “sl”, which corresponds to +3.75 m, stands for the slice level for mining. *4: The two grains are chalcopyrite that were confirmed by electron probe microanalysis (EPMA). *5: W1 is a working name. E1 or E8 is the actual location.

show the locality of analyzed veins, which are Keisen veins (Sanjin ore zone), Zuisen and Hosen veins (Honko ore zone), and Yusen veins (Yamada ore zone). Additionally, Table 5 includes two chalcopyrite grains for analysis. Figure 5 shows backscattered electron (BSE) images of representative pyrite and chalcopyrite grains. Additionally, pyrite grains in the host rock (shale) adjacent (within 2 cm from the vein boundary) to the Hosen No. 1 vein were also used for analysis. Au and As concentrations were analyzed using SIMS [8,23] on the same pyrite samples used in this study (Table 5).

3.2. Analytical Methods

3.2.1. Carbon and Oxygen Isotope Ratios of Calcite

IRMS determined calcite-bearing samples’ carbon and oxygen isotope ratios using the method described in [24]. Several aliquots of the powder sample of 0.15 mg each are kept in steel thimbles and dropped down one-by-one into phosphoric acid at 60 °C in an online reaction chamber under vacuum conditions [24]. Evolved gas is cryogenically purified to retain CO₂. The reaction chamber is connected to the inlet system of a Finnigan MAT-250 mass spectrometer (Thermo Finnigan, Bremen, Germany) at Shizuoka University. Isotope ratios are reported in standard δ notation in per mil (‰) relative to the Vienna standard mean ocean water (SMOW) for δ¹⁸O and relative to the Vienna Pee Dee Belemnite (PDB) for δ¹³C. Reproducibility was approximately ±0.1‰ (2σ) for both the δ¹³C_PDB and δ¹⁸O_SMOW values of calcite. Using measurements on a laboratory working standard and the NBS 19, measured isotope ratios were normalized to limestone reference material [25].

3.2.2. Sulfur Isotope Ratios of Pyrite

Sulfur isotope ratios of pyrite were determined using an ims-1270 SIMS (Cameca, France) with a multi-collector at the Geological Survey of Japan (GSJ), National Institute of Advanced Industrial Science and Technology (AIST). Analysis was conducted using positive primary ions and negative secondary ions. A 10 kV defocused Cs+ primary beam (Köhler illumination) was restricted to 20 µm in diameter by a circular aperture to obtain a homogeneous primary beam of about 1 nA. Additionally, an electron flood gun was deployed on the sample surface for charge compensation to achieve better stability of the secondary ions. A 20 µm-diameter circle that served as the primary beam bombardment’s target area is the area that was examined on the sample surface. Sputtered negative secondary ions with an accelerating voltage of −10 kV were extracted from the carbon-coated sample surface. To guarantee that the same energy band of secondary ions was consistently chosen for experiments, an energy window of 50 eV was adjusted to the energy distribution curve. The secondary ³²S⁻ and ³⁴S⁻ ions were detected simultaneously without energy filtering using Faraday cups of the multi-collection system. The primary beam intensity was adjusted to around 1 nA to generate a high intensity of secondary ions (³²S = 2 × 10⁸ cps) for high-accuracy sulfur isotope studies. The measured data were normalized using a working standard sample (Ak pyrite, δ³⁴S = +5.1‰) and are reported utilizing the standard δ notation in per mil (‰) relative to Canyon Diablo Troilite for δ³⁴S. The accuracy is less than ±0.2‰ (2σ).

4. Results

4.1. Carbon and Oxygen Isotope Ratios of Calcite-Bearing Hand Specimens

Table 1 lists calcite’s δ¹³C and δ¹⁸O values from the Hishikari deposit. The δ¹³C and δ¹⁸O values of calcite range from −25.1‰ to −2.6‰ and from +3.8‰ to +17.9‰, respectively. The ranges of the δ¹³C and δ¹⁸O values of calcite, when two samples with extraordinarily low δ¹³C values are excluded, are −10.3‰ to −2.6‰ and +3.8‰ to +14.6‰, respectively.

4.2. Microanalysis of Calcite from a Calcite-Dominant Vein

Continuous carbon and oxygen isotope analysis of the calcite-dominant vein (20100317 sample) was determined. Samples from 1743 to 1749 (10 mm-width sample set; Figure 4b) were first assessed to understand the isotopic trend through the growing direction. The most central portion of the vein, the 1748 and 1749 samples, have just 100% quartz and no calcite. The δ¹³C and δ¹⁸O values of calcite range from −7.7‰ to −4.8‰ and from +3.4‰ to +8.2‰, respectively (Table 2). The value of δ¹³C tends to increase in the growth direction, although the value of δ¹⁸O does not significantly change except for the latest 1747 sample having a higher value.

Sections from 176-1 to 176-21 (0.9 mm-width sample set; Figure 4c) provide every 1.2 mm interval isotope data along the growth direction. The earlier stages of samples contain no calcite (Table 3). The δ¹³C and δ¹⁸O values of the sections range from −8.4‰ to −4.7‰ and from +3.0‰ to +7.0‰, respectively (Table 3).

Every 12 μm of the sample made using a microtome (12 μm-width samples) is micro-analyzed for δ¹³C and δ¹⁸O values in the direction of growth. The δ¹³C and δ¹⁸O values range from −9.5‰ to −2.5‰ and from +3.6‰ to +10.9‰, respectively (Table 4). Additionally, the distance in μm from the boundary between 176-19 and 176-20 is shown to understand each sample’s positional relationship (Table 4).

4.3. Sulfur Isotope Ratios of Pyrite

Sulfur isotope ratios of pyrite in quartz veins from three ore zones were determined using SIMS. The δ³⁴S values of pyrite from the Sanjin ore zone range from −16.5‰ to +3.5‰ (Table 5). The δ³⁴S values of pyrite from veins and sedimentary rocks in the Honko ore zone range from −3.4‰ to +2.1‰ and from +0.7‰ to +3.2‰, respectively (Table 5). The δ³⁴S values from the Yamada ore zone range from −2.0‰ to +4.5‰ (Table 5). The δ³⁴S values of two chalcopyrite grains from the Zuisen No. 1 vein are −1.4‰ and +1.0‰. The average δ³⁴S value of the whole data is −0.7‰. The average δ³⁴S values of pyrites in veins and sedimentary rocks are −1.1‰ and +2.0‰, respectively. As the grain size of pyrite is small (mostly between 10 and 30 µm) relative to the analysis area (20 µm), separate analytical values for the rim and core of pyrite are unavailable.

5. Discussion

5.1. Role of Basement in the Epithermal Au Deposits in Kyushu

Fluid-rock dynamics significantly impact the carbon and oxygen isotope budgets of the geologic contexts of ore-forming hydrothermal systems. First, let’s consider the geologic settings of the Kushikino deposit adjacent to the Hishikari deposit (Figure 1). The Kushikino deposit occurs in Pliocene andesitic rocks, and the basement rocks of the deposit are sedimentary rocks of the Lower Shimanto Group (a Cretaceous accretionary wedge). On the other hand, the Noya deposit is found in Pleistocene andesitic rocks in northern Kyushu (Figure 1). The deposit’s basement rocks are high-temperature metamorphic and granitic rocks of the Cretaceous period. The influence of basement rocks on the ore-forming fluid is examined in the Kushikino and Noya deposits. The δ¹³C values in the fluid for the Noya and Kushikino gold deposits in Kyushu were estimated to be −6.5‰ [26] and −10.8‰ [19,27], respectively. The calculated δ¹³C value of about −7‰ for the Noya deposit is regarded as average crustal carbon (e.g., [28,29]) and as deep-seated sources (e.g., [30,31]). Both deposits have andesitic host rocks, although the basement rocks are mostly granitic for the Noya deposit and sedimentary accretionary rocks for the Kushikino deposit. The difference in basement rocks might cause the difference in δ¹³C value between the two deposits. The δ¹³C value in the Kushikino fluid might be lowered by about 4‰ by reactions with organic carbon in the basement sedimentary rocks [19].

The isotopic equilibrium fractionation lines for calcite from the Kushikino and Noya deposits are shown in Figure 6. The δ¹³C and δ¹⁸O values of most calcite from the Kushikino deposit lie on the isotopic equilibrium fractionation lines, the Kushikino trend (Figure 6), which suggests that the reaction with the basement sedimentary rocks occurred not locally but in a vast reservoir. There is a discrepancy between the two trends in Figure 6 because the basement rocks for the two deposits are different. Calcites from the Lower Shimanto Group sedimentary basement rocks that are part of the epithermal Au deposits in the Hokusatsu district (Figure 1) are located on the Kushikino trend [19,32]. The widely distributed basement rocks of the Shimanto Group may have exerted regional control over the district’s relatively low and homogeneous δ¹³C values of ascending ore-forming fluids [32]. The Sr isotope ratios (⁸⁷Sr/⁸⁶Sr) of the Naya’s veins are the same as adjacent andesite (0.704). However, the ratios of the Kushikino and Hishikari veins are both high (higher than 0.706), which is attributed to reactions with basement Cretaceous sedimentary rocks with high Sr isotope ratios [26]. So, the δ¹³C depletion and ⁸⁷Sr/⁸⁶Sr enrichment show the basement influence on Kyushu’s epithermal ore-forming fluid.

5.2. Hydrothermal Fluids for the Hishikari Deposit

Figure 7 shows calcite’s δ¹³C and δ¹⁸O values from the Hishikari deposit. The figure excludes two calcite samples with very low δ¹³C values of −23.4‰ and −25.1‰ (Table 1). The average global δ¹³C value of C3 plants is approximately −28.5‰ [36]. Local pyrolysis or microbial decomposition of these organic materials in the basement sedimentary rocks can result in the production of CO₂. The fluid temperature at the precipitation of the very low δ¹³C calcite is calculated to be around 45 °C using the δ¹⁸O values of the calcites and calibration curve of O’Neil et al. [35], assuming that the δ¹⁸O value of −7‰ for the fluid, which is similar to that of the Kushikino ore fluid (δ¹⁸O = −7‰) and regional meteoric water (δ¹⁸O = −7.1‰) [19,37]. Therefore, the calcites with very low δ¹³C values might precipitate after the ore mineralization.

Since the Hishikari and Kushikino deposits both occur in the same Hokusatsu district and share the same basement sedimentary rocks, it is expected that the calcite’s δ¹³C and δ¹⁸O values will follow the Kushikino pattern. Although several isotope data are along the Kushikino trend, the others lie between the Noya and Kushikino trends (Figure 7). These particular characteristics of the Hishikari fluid will be investigated next.

First, the δ¹³C and δ¹⁸O values of the 10 mm-width sample set of the calcite-dominant vein (Section 3.1.2) are obtained to get a rough idea of the isotope evolution from the early to late stages of vein formation. In Figure 8, the δ¹³C–δ¹⁸O combination moves slightly from 1743 through 1746 samples. The last sample, 1747, has a high δ¹⁸O value, probably because the temperature drops at the later stage of calcite mineralization, since the oxygen isotope fractionation between calcite and water increases with decreasing temperature. Next, we take finer samples (0.9 mm-width sample set: from 176-1 to 176-21) from the early mineralization stage than the 10 mm-width sample set (Figure 4b,c). A more precise and definite evolutional isotopic trend of the fluid is expected to appear. However, no such trend is found. The δ¹³C and δ¹⁸O values of the 0.9 mm-width samples move irregularly over time, but the range is limited (Figure 8).

Finally, maximum efforts to increase the time resolution were made. The δ¹³C and δ¹⁸O values of calcite from 176-20 and 176-21 sections (12 μm-width sample set; Table 4) are shown in Figure 9a,b, respectively. The range of δ¹³C and δ¹⁸O values for the 176-20 section is slightly bigger than that for the 176-21 section. However, they are similar to the range of 10 mm-width and 0.9 mm-width samples (Figure 9a,b). The change in δ¹⁸O values might reflect temperature changes. However, isotope ratios and growth order do not correlate with one another. The δ¹³C and δ¹⁸O values of calcite jump around in a particular range (Figure 9a,b), and it is difficult to explain the fluid evolution by a simple change of the fluid in temperature or pH.

5.3. Sulfur Isotopes

A histogram of sulfur isotope ratios of pyrite from the Hishikari deposit is shown in Figure 10. The analysis area using the SIMS microanalysis is 20 µm. The whole data’s average δ³⁴S value is −0.7‰. The δ³⁴S values of pyrites in the host rock are similar to those in the adjacent ore vein at the same location of the host rock (E8) and Hosen-1 (E8) in Figure 10, which indicates that the same fluid forms both hydrothermal pyrites. The pyrite grains in the host rock (shale), within 2 cm from the vein, occur near the vein by the ore fluid that flows into the host rock. Late Cenozoic volcanic rocks in the Hokusatsu district are of the magnetite series (similar to I-type granitoids; [38]) that are defined by Ishihara [39]. Ishihara et al. [40] measured δ³⁴S values of three vein ores and six basement sedimentary rocks from the Hishikari deposit by the conventional method, and the average δ³⁴S value is around +0.3‰. The δ³⁴S value of stibnite from the Hishikari deposit is −0.2‰ [41], and the δ³⁴S values of seven stibnite samples, which are from the latest stage of the Hishikari mineralization, are narrow-ranged (from −0.2‰ to +0.7‰) among several different veins [42]. The near-zero values of δ³⁴S from the Hishikari hydrothermal system are consistent with that from magnetite series magma [43].

Very low δ³⁴S values from −16.5‰ to −10.1‰ of pyrites are found only in the high-grade Keisen–3 vein of the Sanjin ore zone (Figure 10; Table 5). The very low δ³⁴S values, first obtained by the SIMS microanalysis, are apart from the pile of other values (Figure 10). Sanematsu et al. [18] proposed that the fractures corresponding to the Keisen veins were a significant channel for the ascending hot hydrothermal fluids. The δ³⁴S values of unaltered Shimanto sedimentary rocks are from −21.5‰ to −1.4‰ [43]. Therefore, the pyrite with a very low δ³⁴S value might form locally that reacted with the Shimanto sedimentary rocks. The reaction with the sedimentary rocks might occur in the microscopic region because relatively high δ³⁴S values are also obtained from the same hand specimen of the Keisen–3 vein (Figure 10).

5.4. Genesis and Evolution of Hydrothermal Fluids in the Formation of the Hishikari Deposit

The Hishikari deposit’s ore-forming fluids may mix meteoric and magmatic fluids. Particularly along the unconformity between the basement sedimentary rocks and overlying andesites [44]. The fluids are formed from meteoric water undergoing isotopic exchange with basement rocks [6]. The isotope results in this study also suggest that the ore-forming fluid for the Hishikari deposit is a mixture of two fluids. We refer to the two fluids as Fluid 1 (along the Noya trend), which the basement sedimentary rocks may not affect, and Fluid 2 (along the Kushikino trend), which they may constrain. Fluid 2 is common to the deposits in the Hokusatsu district (Figure 1); that is, the δ¹³C and δ¹⁸O values of calcite from most Au deposits in the Hokusatsu district are along Fluid 2 [19,32]. According to past research [2,3,4,5,6], the Hishikari fluid may have received some magmatic fluid input, particularly during the earliest stages of mineralization. The genesis of the Hishikari hydrothermal fluid is the most of meteoric origin. However, some sporadically input of magmatic origin to the fluid has occurred. The Hishikari deposit also has Fluid 2 [19,21,32], like any other deposit in the district, but it should be emphasized that Fluid 1 is peculiar to the Hishikari deposit. Calcite samples from Fluids 1 and 2 are everywhere in the deposit.

Figure 11 shows the δ¹³C and δ¹⁸O values of vein calcite with/without electrum [21]. Although the δ¹³C values of samples with electrum are relatively high compared to those without electrum, the δ¹³C and δ¹⁸O values of calcites with electrum lie either near the Noya trend or Kushikino trend. Additionally, since vein calcite samples without electrum in Figure 11 are from Au-rich veins, they are also recognized as fertile. Therefore, at the Hishikari deposit, Fluids 1 and 2 are both created from fluids capable of forming ore (Figure 11). The 20100317 sample is from the Au-rich Keisen 3-1 vein, and the earliest formed calcite in the 20100317 sample (176-5; Table 3) lies on the Noya trend (Fluid 1), which is apart from the Kushikino trend (Fluid 2 in Figure 11). Fluids 1 and 2 are both fertile ore-forming fluids. Fluid 2 is thought to have reacted well with basement sedimentary rocks that contain organic matter, and Fluid 1, on the other hand, is not affected on a large scale by sedimentary rocks.

The previous studies [2,3,4,5,6] have suggested that some input from magmatic fluid was found in the Hishikari hydrothermal fluid in the earlier stages of mineralization. In other words, the genesis of the Hishikari fluid is mostly of meteoric origin, but some sporadical input of magmatic fluid to the hydrothermal system has occurred. Additionally, the unusually low δ³⁴S readings in the Keisen veins suggest that magmatic fluid only locally and in a small volume interacted with sedimentary rocks. The microanalysis results of δ¹³C and δ¹⁸O values in every 12 μm of the calcite-bearing sample along the growth direction scatter in a particular range (Figure 9a,b). The duration of 12 μm-thickness mineralization is about 10 years of the Hishikari mineralization according to the ⁴⁰Ar/³⁹Ar ages of adularia [45]. Since the very high temporal resolution cannot find the isotope evolution of the hydrothermal fluid, the evolution is not a gradual transition from magmatic to meteoric origin. Still, a rapid turnover of two fluids might be locally happening.

The average pyrite δ³⁴S value of −0.7‰ from veins and sedimentary rocks is similar to that of the adjacent magma, consistent with previous studies. However, the SIMS microanalysis reveals that very low δ³⁴S values from −16.5‰ to −10.1‰ and positive δ³⁴S values of pyrites are found in the same hand specimen from the high-grade Keisen-3 vein of the Sanjin ore zone (Figure 10). Therefore, the pyrite having a very low δ³⁴S value might form locally that was reacted with the Shimanto sedimentary rocks. The evidence from sulfur isotope data proves that the reaction of the hydrothermal fluid with basement sedimentary rocks occurs locally in the microscopic region.

Fluid 2, the mainstream in the Hokusatsu district, is supposed to flow continuously in the Hishikari deposit during the mineralization. However, Fluid 1, a magmatic-origin deep fluid with little reaction with the basement sedimentary rocks, might inject intermittently into the Fluid 2-filled vein system. Intermittent opening of the top of a vein lowers the pressure in the vein, resulting in the introduction of deep Fluid 1 to the epithermal system and subsequent mixing with Fluid 2 and boiling in some cases. Au was probably transported as Au(HS)₂⁻ [3] and might be deposited in response to changes in the temperature, pH, log fO₂, and total sulfur concentrations [46]. Boiling would fractionate H₂S strongly into the vapor, lower HS⁻ activity, and increase pH by losing H₂S, which causes the deposition of calcite and Au [47].

The characteristics of the high-grade Hishikari deposit, compared to the other Au deposits in the Hokusatsu district, might be caused by Fluid 1. Unlike other epithermal Au deposits in the district, the rapid local mixing of Fluid 1 and 2 might cause the Au-rich ore deposition at the Hishikari deposit.

6. Concluding Remarks

The Hishikari gold deposit in Kyushu, Japan, is world-famous for its high-grade ore. The Hishikari deposit is one of the numerous epithermal gold occurrences in the Hokusatsu district. It is located in the basement rocks of the Lower Shimanto Group (Cretaceous accretionary wedge). The average pyrite δ³⁴S value of −0.7‰ from veins and sedimentary rocks is similar to that of the adjacent magma, consistent with previous studies. However, the SIMS microanalysis reveals that very low δ³⁴S values and positive δ³⁴S values of pyrites are found in the same hand specimen from the high-grade Keisen-3 vein. Therefore, sulfur isotope measurements demonstrate that the hydrothermal fluid’s reaction with the basement sedimentary rocks occurs locally because the pyrite with a very low δ³⁴S value might form locally.

Although the basement rocks regulate the feature of the hydrothermal fluids for the deposits, it has been inferred that the incorporation of magmatic-origin deep fluid to the meteoric-origin hydrothermal fluid might cause the gold and calcite precipitation at the Hishikari deposit. Although this scenario applies to the Hishikari deposit generally, the microanalysis of δ¹³C and δ¹⁸O values of vein calcite reveals that the fluid evolution was not gradually occurring, and a rapid turnover of two fluids might be locally happening. Thus, the δ¹³C and δ¹⁸O values of hydrothermal calcite can provide insights into the genesis and evolution of the epithermal deposits.