Isotopic and Chemical Compositions Volcanic Gases from Satsuma-Iwojima

Geochemical Journal, Vol. 8, pp. 165 to 173, 1974

Isotopic and chemical compositions of volcanic gases from Satsuma-Iwojima, Japan

SADAO MATSUO1, TETSURO SUZUOKI2, MINORU KUSAKABE1, HIROAKI WADA3 and MASARU SUZUKI4

Department of Chemistry, Tokyo Institute of Technology, O-okayama , Megur.. xu, T okyo 1521, Meteorological College, Kashiwa-shi, Chiba-ken 2772 , National Institute for R esearches in Inorganic Materials, Sakura-mura , Niihari-gun, Ibaragi-ken 300-313, Tanabe High School, Tanabe-shi, Wakayama-ken 6464, Japan.

(Received November 13, 1974)

Abstract-D/H and 180/160 ratios of fumarolic condensate, hot spring water and surface water collected from a volcanic island, Satsuma-Iwojima were determined together with some of chemical components. 6D and 6180 of fumarolic condensates range from -27 to -17%o (SMOW) and from +7.3 to +9.5 %o (SMOW), respectively. The high value of 5180 was concluded to be the result of thorough oxygen isotope exchange with ambient andesitic rocks. 6D values are also higher than that of the local surface water, but we could not find a positive evidence that supports the assumption of mixing of the local surface water and sea water. On the basis of the relationship between the concentration of chemical components and isotope ratios of fumarolic condensates, it was concluded that water vapor and chemical components behave independently in an individual fumarole.

INTRODUCTION

Satsuma-Iwojima is a small volcanic island (12 km2) situated about 50km south of Kyushu, Japan (Fig. l ). The island constitutes a part of the somma of the sub merged Kikai caldera of which area is estimated to be 230km2 (MATSUMOTO, 1943). There is an active volcano, Iwodake (703m) on the island. Iwodake is one of the central cones (most of which have been submerged) , and is close to the caldera wall which runs NE-SW across the island. The lava of Iwodake consists mostly of glassy matrix with minor amounts of phenocryst composed of augite , hypersthene, and plagioclase (An 48 58%) (MATSUMOTO,1951). A submarine eruption with intense earthquakes in 1934 formed a tiny volcanic island (Showa-Iwojima, 0.15 km') about 3 km east of Satsuma-Iwojima , where the water depth was about 300m. (TANAKADATE, 1935). Since a research group of Satsuma-Iwojima was organized in 196 1, various phases 165 166 S. MATSUO et al. of this volcanic island have been studied such as the chemistry of fumarolic gases (IWASAKIet al., 1962, 1966), volcanic sublimates (YOSHIDA et al., 1966), alteration products (YOSHIDA et al., 1969) and hot spring waters (KAMADA, 1964). Even before the organization of the research group, observations of earthquakes during the 1934 eruption (YASUI, 1962), measurements of fumarolic temperatures (KAMADA and OZAWA, 1958), and a petrological study on the lava (MATSUMOTO,1951, 1954) have been made. Temperatures of fumarole outlets have been intermittently measured since 1935. Through 1961 to 1962, the highest temperature was recorded to be 945°C that is comparable with that of Showa-Shinzan at its earliest stage. The hydrological situation of Satsuma-Iwojima is possibly similar to that of Surtsey Island formed in 1964 some 30km off the south coast of Iceland, and to that of White Island, an andesitic volcano 50km from the coast of Bay of Plenty of New Zealand, since these are solitary volcanic islands in the sea with limited supply of meteoric water. ARNASONand SIGURGEIRSSON(1968) continuously measured the D/H ratio of fu marolic condensates from Surtsey to be about -5 3 ° o (ED, SMOW) with rather narrow spread. This value is quite close to -55 %o of the surface water of the south coast of Iceland. ARNASONand SIGURGEIRSSON(1968) were prudent enough to draw a definite conclusion on the origin of the-water. STEWARTand HULSTON(1974) suggested the contribution of sea water to the source water of White Island fumarolic vapor for which they had measured the D/H and "0/"0 ratios. It has been widely believed that fluids associated with geothermal systems are of the local surface water origin (CRAIG, et al., 1956; CRAIG, 1963). Supporting evidences based also on isotopic studies were given by McDONALD(1966) for Ngawha, New Zealand, and KUSAKABEet al. (1970) for Nasudake, Japan. However, for many of Japanese thermal waters along the ocean coast, MIZUTANI and HAMASUNA(1972) and MATSUBAYAet al. (1973) concluded that coastal thermal waters are isotopically intermediate between oceanic and local meteoric waters and are regarded as a mixture of the two types of water. We collected samples of volcanic gases and various types of water from Satsuma Iwojima and analyzed for D/H, 180/160, and chemical components of the samples to see how the volcanic vapor correlates with sea water and the local surface water.

SAMPLE COLLECTION AND ANALYTCAL METHODS

Sampling was made in July, 1967 and the sampling localities are shown in Fig. 1. Fumarolic gas condensate samples (AF-1, AF-2 and AF-3) from the crater of Iwodake were collected by introducing the steam from a fumarole through a Pyrex glass tubing into a water-cooled trap to condense. Hot spring waters (EH and FH-3), a Isotpic and chemical compositions of volcanic gases 167 ground water (G, being used as tap water on the island) and a rain water (P) were also sampled. The other three fumarolic condensate samples (BF and CF from the crater and DF from the flank of Iwodake), and some hot spring water samples (DH-1 , DH-2, FH-1 and FH-2) were donated by T. OZAWA and M. YOSHIDAof Tokyo Institute of Technology. All these samples were subjected to isotopic analysis . The fumarolic gases (AF-1, AF-2 and AF-3) were collected in a water-cooled alkaline solution by means of a method modified from that Of MIZUTANI (1962) for chemical analysis.

Kagoshima

i 31o30'N i 0) D

AO u rc IL ~ 5000

30°3O'N ~ ~ 300 OG LV ) O 100

1 1 km 1

130°E 131°E

Fig. 1. Satsuma-Iwojima and sampling localities on the island. Broken line shows the rim of the Kikai caldera. A; Kuromoe, B; Omaebira, C; Kamanokuchi , D; Kitabira, E; Sakamoto; F; Higashi, G; Nagahama, where rain water was sampled. 0; meteoric waters, v; hot spring waters and •; fumarolic condensates .

To determine hydrogen isotope ratios, about 10mg of water sample was reduced to hydrogen on heated metallic uranium as described by BIGELEISENet al. (1952) and FRIEDMAN(1953), and the D/H ratio of the hydrogen was measured on a mass spectrometer. For 180/160 ratio determinations, the carbon dioxide gas which was equilibrated with a water sample at 25°C (EPSTEINand MAYEDA,1953) was run in a mass spectrometer. The isotopic ratios were expressed as S values relative to SMOW standard (CRAIG, 196 1 a). Overall accuracies are estimated to be ± 1 %o for S D and ±0.3 %o for S 180. Chemical analyses of fumarolic gas condensates in alkaline solutions followed the method developed by MIZUTANI(1962) and OZAWA(1966) except for CO-2 determina tions. The carbon dioxide was determined by the following manner. To an aliquot of the oxidized (by H202) alkaline sample solution , Ba(OH)2 solution was added in ex cess to form BaCO3 and BaSO4 precipitate. The excess barium in the solution was determined by chelate-titration. The carbon dioxide concentration was worked out by subtraction of sulfate concentration determined by the conventional gravimetry . The residual gases such as hydrogen, nitrogen and so on were analyzed on a gaschromatograph. 168 S. MATSUO et al.

The outlet temperature of fumaroles was measured by a calibrated alumel chromel thermocouple.

RESULTS AND DISCUSSION

According to our observation at Nasudake volcano, Japan (WADA and SUZUOKI, 1967), the concentrations of some of the components in volcanic gases vary to a great extent in a very short period of time. The one-shot analysis could give inadequate information on the chemical nature of volcanic gases from a particular fumarole. In order to check the constancy of chemical composition of volcanic gases with time, continuous sampling of condensed water was made on the fumarole AF-3 (487°C) for chemical analysis and on two fumaroles AF-1 (659°C) and AF-3 for isotopic analysis at the time interval of 10 minutes for 2 hours. The results of analyses are presented in Tables 1 and 2. In the case of AF-3 fumarole, the maximum variation of CO2 content was about 17% around the averaged value. Those of total-S, HC1 and HF were 6, 4 and 7%, respectively. In the case of Nasudake volcano, however, the variation of CO, content under the same sampling condition as that of Satsuma-Iwojima was 30%, and those of Total-S and HCl were 40 and 50%, respectively. SD and S'$O values of AF-1 and AF-3 fumaroles were roughly constant, while S's0 of Nasudake volcano varied almost 10 times more than the variation range of AF-1 and AF-3 (SuzuoKl, 1967). From these results, we may safely conclude that the chemical and isotopic compositions of the volcanic gases of Satsuma-Iwojima are rather constant within the time period of at least an order of hour. When we compare the results of the one

Table 1. Results of chemical and isotopic analyses of continuously collected condensates at AF-3 fumarole, Satsuma-Iwojima

Sample Chemical composition (Q/1000Q) Isotopic composition (%0) No. H20 CO2 SO2 H2S HC1 HF SD 518o 1 983 2.81 6.39 1.11 5.81 0.59 -20.1 +8.0 2 983 3.89 6.44 0.78 5.48 0.55 +7.8 3 983 3.66 6.40 0.55 5.42 0.57 +7.8 4 983 3.90 6.71 0.68 5.55 0.58 -20.3 +7.6 5 983 3.64 6.46 0.81 5.59 0.59. +7.5 6 983 3.31 6.68 0.78 5.36 0.54 +7.5 7 983 3.50 6.77 0.50 5.53 0.57 -19.5 +7.4 8 984 3.05 6.51 0.88 5.53 0.53 +8.1 9 983 2.97 6.63 1.05 5.73 0.52 +7.6 10 984 3.32 6.47 0.39 5.60 0.55 +7.9 11 983 3.13 6.63 0.81 5.65 0.5 3 +7.4 12 983 3.45 6.66 0.67 5.61 0.57 -21.S +7.4

Average 983 3.38 6.56 0.75 5.57 0.56 -20 .4 +7.6 Isotopic and chemical compositions of volcanic gases 169

Table 2. Results of isotopic analyses of various waters from Satsuma-Iwojima

Sample SD (%o) 8180(%0) Cl-** (ppm) P meteoric precipitate -48 .9 -7.7 G ground water -33 .9 -6.0 82 FH-1 Higashi hot spring water -31.0 -5.0 1570 FH-2 Higashi hot spring water -31.5 -5.2 2180 FH-3 Higashi hot spring water -35.0 -2.0 1550 EH Sakamoto hot spring water -26.8 -4.3 5000 DH-1 Kitabira hot spring water -26.9 -2.4 1390 DH-2 Kitabira hot spring water -25 .8 -3.4 1070 DF Kitabira fumarolic condensate -28.1 -0 .4 BF Omaebira fumarolic condensate -17 .1 +7.8 CF Kamanokuchi fumarolic condesnate -18 .1 +9.5 AF-2 Kuromoe fumarolic condensate -27 .0 +7.8 +0.8 +1.1 AF-1 * Kuromoe fumarolic condensate -25 .3 +7.3 -0.6 -0 .6 +0.9 +0.5 AF-3 * Kuromoe fumarolic condensate -20 .4 +7.6 -1.1 -0.5 * Average value of the continuously collected samples with maximum deviation from the average. Results of the individual samples are shown in Table 1 for AF-3, but not for AF-1. ** Analyzed by M . KolzuMI.

shot analyses in 1962 and 1967 shown in Table 3, an excellent agreement is found between these two sets of data. Even the one-shot collection takes time no shorter than 30 minutes, the quick change, if any, in the composition can be smoothed out . On this basis, we see that the composition of volcanic gases from an active fumarol.c area of this volcano is preserved for several years. The relationship between 5D and 5180 of various types of water collected in this volcanic island is shown in Fig. 2. The pattern found in Fig . 2 is similar to those of the other geothermal areas. The common features are; a) surface waters are on the meteoric water line (CRAIG, 1961 b), b) fumarolic condensates are slightly high in iD and quite high in 5180 as compared with those of the surface water , and c) 6D and 5180 values of hot spring waters are distributed between those of the surface water and fumarolic condensates. The values of 5180 of the condensates (+7 +9 %o) are definitely higher than those of other Japanese volcanic condensates such as -1 0%o of Showa-Shinzan (KUSAKABE,unpublished data) and -7 -5 %o of Nasudake (KUSAKABEet al., 1970) and close to +6 +9%o of Kuju (SUGISAKI, personal communication) . The lave of the central volcano (Iwodake) of this island consists of two-pyroxene andesite (MATSU MOTO, 1951) of which the bulk 5180 value may be close to +6 +8%o of Japanese andesitic rocks (MATSUHISA, 1973). The 5180 value of the condensates is almost the 170 S. MATSUO et al.

Table 3. Chemicalcomposition of volcanicgases from Satsuma-Iwojima,Nasudake and Showa-shinzan(Q/1000Q)

Volcano Satsuma-Iwojima Nasudake Showa-shinzan Fumarole AF-1 AF-2 AF-3 AF-4* AF-5* BF* CF* M-3** A-1*** Date 1967 1967 1967 1962 1962 1958 1958 1966 1959 Temp.(°C) 659 570 487 360 585 120 200 332 750 H2O 979 981 983 982 983 967 953 996 993 C02 4.2 4.7 3.4 4.1 4.3 3.2 4.5 2.3 4.7 H2S 0.5 1.0 1.1 1.4 3.6 2.8 1.5 7 X 10-3 8.3 SO2 r 8.2 6.3 6.2 7.5 9.5 16.1 0.2 0.12 HCl 7.1 4.9 5.3 5.3 2.3 13.8 20.6 0.11 0.39 HF 0.35 0.40 0.22 0.45 0.42 1.8 2.5 0.20 H2 1.20 0.72 0.41 0.9 0.13 0.19 0.05 1.81 N2 0.05 0.04 0.34 0.2 0.04 0.26 Ar 3 X 10' 9 X 10'4 CH4 8 X 10'6 8 X 10-6 2 X 10-3 6 X 10-3 Co 6 X 10-3 1 X 10-3 7 X 10-3 * KAMADA(1964) , ** KUSAKABE(1969), *** MATSUO(1961). same as that of the lava. This indicates that the vapor coming out of the fumarole is mostly of the water that has been subjected to a thorough interaction with the rocks in regard to oxygen isotope exchange at or close to the "cross-over" tempera ture. Most of the condensate samples were Collected from the crater of Iwodake, of which diameter is 400 500m and the depth from the rim is about 50m. There are a variety of temperature, chemical composition and isotope ratios of fumarolic gases even in such a small area. In AF series (3 fumaroles), 5180 is almost identical (+7.3 +7.8%o) while 5D changes from -27 to -20%o. When we include 5180 and 5D values of BF and CF (also inside the crater), the range of 6180 expands from +7.3 to +9.5%o and 5D ranges from -27 to -17%o. Both SD and 5180 values of the fumarolic con densates are clearly higher than those of the local surface water. As already mentioned the high 5180 of the fumarolic condensates can be attributed to "oxygen isotopic shift". On the other hand, we can not find any definite evidence of the contribution of sea water to the fumarolic condensates that results in high 5D values of the con densates. At the present time, we have no account for the high 5D values of the fumarolic condensates. There is a slight inverse correlationship between 5D values and concentrations of F, total-S, CO2 in AF series. However, when BF and CF are included, this regularity is no longer reproduced. This suggests that water and other chemical components behave independently in individual fumarolic gases. There is isotopically an intermediate group (hot spring water) between two distinct groups, i.e., fumarolic condensates and the local surface water as seen in Fig.2. On the basis of the relationship between liD and chloride concentration, the inter Isotopic and chemical compositions of volcanic gases 171

6180 (°i°°) -8 -6 -4 -2 0 +2 +4 +6 +8 +10

i , 6D = 86180 + 10 , -10

-20

, 0 -30

0 , -40

O Ground water & Rain water 0/ -50 0 Hotspring water • Fumarolic condensate 6D

Fig. 2. Isotopic composition of waters from Satsuma-Iwojima. mediate group can not be concluded to be formed by the mixing of sea water and the local surface water. Since there is no unique value of chloride concentration for the fumarolic condensates, we can not assume that the intermediate group is a mixture of fumarolic condensate and the local surface water. On the other hand , the possi bility that the intermediate group consists of the water which originates from the local surface water and has been subjected to oxygen isotopic shift is suggested . If this is the case, the chemistry of the water is not preserved in the process of oxygen isotopic shift. It is to be noted that the water of EH can be regarded as a mixture of the local surface water and sea water on the basis of both SD and chloride concentration . EH is a hot spring situated at the shore line just outside the rim of Kikai caldera . In this respect EH is a singular hot spring in Satsuma-Iwojima . In this particular hot spring, the volcanic activity seems to furnish only heat and not to affect the isotopic and chemical characteristics of the water.

ACKNOWLEDGMENT

Authors are grateful to Prof. M. KAMADA of Kagoshima University for his invaluable informa tion on various phases of Satsuma-Iwojima and his guidance and help during the field work . Mr. Y. NIWANO of Tokyo Kyoiku University helped us throughout the field work . Authors' thanks are due to Drs. T. OZAWA and M. YOSHIDA of Tokyo Institute of Technology for the cooperative field work and discussion on this subject. They donated us some of water samples from their own collection. Mr. M. KOIZUMI of Tokyo Kyoiku University kindly analyzed chloride in some of water samples. Prof. H. S AKAI of Okayama University critically read the manuscript and gave us many valuable and useful comments. A part of the expenses of this study was defrayed by the Grant in Aid for Fundamental Scientific Research of the Ministry of Education . 172 S. M ATSUO et al.

REFERENCES

ARNASON,B. and SIGURGEIRSSON,T. (1968) Deuterium content of water vapour and hydrogen in volcanic gas at Surtsey, Iceland. Geochim. Cosmochim. Acta 32, 807-813. BIGELEISEN,J., PERLMAN,M. L. and PROSSER,H. C. (1952) Conversion of hydrogenic mate rials to hydrogen for isotope analysis. Anal. Chem. 24, 1356-1357. CRAIG, H. (1961a) Standard for reporting concentrations of deuterium and oxygen-18 in natural waters. Science 133, 1833-1834. CRAIG, H. (1961b) Isotopic variations in meteoric waters. ibid. 133, 1702-1703. CRAIG, H. (1963) The isotope geochemistry of water and carbon in geothermal area. Nucl. Geol. on Geothermal Areas, 17-53, Spoleto, 1963. . CRAIG, H., BOATO, G. and WHITE, D. E. (1956) Isotopic geochemistry of thermal waters. Proc. Second Conf Nucl. Proc. in, Geol. Settings, Publ. 400, Natl. Acad. Sci.-Natl. Res. Council, 29-38. EPSTEIN,S. and MAYEDA,T. (1953) Variation of 180 content of waters from natural sources. Geochim. Cosmochim. Acta 4, 213-224. FRIEDMAN,I. (1953) Deuterium content of natural waters and other substances. ibid. 4, 89-103. IWASAKI,I., OZAWA, T., YOSHIDA,M., KATSURA,T., IWASAKI,B., KAMADA,M. and HIRAYA MA, M. (1962) Volcanic gases in Japan. Bull. Tokyo Inst. Tech. No. 47, 1-54. IWASAKI,I., OZAWA,T., YOSHIDA,M., KATSURA, T., IWASAKI,B. and KAMADA, M. (1966) Differentiation of magmatic emanation. ibid. No. 74, 1-57. KAMADA,M. (1964) Volcanoes and geothermy of Satsuma-Iwojima, Kagoshima Prefecture. J. Japan Geothermal Energy Assoc. No. 3, 1-23 (in Japanese). KAMADA,M. and OZAWA,T. (1959) Volcanic gases from Tokara archipelago. Bull. Volcanol. Soc. Japan, 2nd Ser. 3, 164 (in Japanese). KUSAKABE,M. (1969) Atomic composition and chemical equilibrium of volcanic gases. Geochem. J. 3, 141-151. KUSAKABE,M., WADA,H., MATSUO,S. and HORIBE, Y. (1970) Oxygen and hydrogen isotope ratios of monthly collected waters from Nasudake volcanic area, Japan. J. Geophys. Res. 75, 5941-5951. MATSUBAYA,0., SAKAI, H., KUSACHI,I. and SATAKE,H. (1973) Hydrogen and oxygen iso topic ratios and major element chemistry of Japanese thermal water systems. Geochem. J. 7, 123-151. MATSUHISA,Y. (1973) Oxygen isotope variations in magmatic differentiation processes of the volcanic rocks in Japan. Contr. Mineral. Petrol. 39, 277-28&. MATSUMOTO,H. (1951) Lavas of Iwo Island, Kagoshima Prefecture. J. Japan. Assoc. Min. Pet. Econ. Geol. 35(6), 182-188 (in Japanese). MATSUMOTO,H. (1954) Petrological study on rocks of Iwozima, Kagoshima Prefecture. Kuma moto J. Sci. Ser. B. No. 4, 96-107. MATSUMOTO,T. (1943) The four gigantic caldera volcanoes of Kyushu. Japan. J. Geol. Geogr. 19, 1-57. MATSUO,S. (1961) On the chemical nature of fumarolic gases of volcano Showashinzan, Hokkaido, Japan. J. Earth Sci. Nagoya Univ. 9, 80-100. MCDONALD,D. C. (1966) Hydrogen and oxygen isotope ratios in the waters of the Ngawha hydrothermal area, North Auckland. Bull. Volcanol. 29, 691-708. MIZUTANI,Y. (1962) Chemical analysis of volcanic gases. J. Earth Sci. Nagoya Univ. 10, 125 134. Isotopic and chemical compositions of volcanic gases 173

MIZUTANI,Y. and HAMASUNA,T. (1972) Origin of the Shimogamo geothermal brine, Izu. Bull. Volcanol. Soc. Japan, 2nd Ser. 17, 123-134 (in Japanese with English abstract) . OZAWA,T. (1966) Chemical analysis of volcanic gases containing water vapor, hydrogen chloride, sulfur dioxide, hydrogen sulfide, carbon dioxide etc. Nippon Kagaku Zasshi 87, 848-853 (in Japanese). STEWART,M. K. and HULSTON,J. R. (1974) Stable isotope ratios of volcanic steam from White Island, New Zealand. Bull. Volcanol. (in press). SUZUOKI,T. (1967) Continuous observation of 180/160 and 13C/12C ratios of fumarolic gases from Nasudake volcano. Read at 1967 annual meeting of Geochem. Soc. Japan (in Japanese). TANAKADATE,H. (1935) A preliminary report on the eruption of Iwo-zima, Kagoshima Pre fecture. Bull. Volcanol. Soc. Japan, 1st Ser. 2, 188-209 (in Japanese). YASUI, Y. (1962) A survey of swarm earthquakes at southern Kyushu, Japan. Kenshin-ziho 27, 109-124-(in Japanese). YOSHIDA,M., OZAWA,T. and OSSAKA,J. (1966) A singular silica sublimate mineral found in Satsuma-Iwo-zima Volcano (I) and (II). J. Japan. Assoc. Min. Pet. Econ. Geol. 55, 201-211 (I) and 262-271 (II) (in Japanese). YOSHIDA,M., OZAWA,T. and OSSAKA,.J. (1969) Fluorine and chlorine contents of the rock alteration products by volcanic gases obtained at Satsuma-Iwo-zima Volcano. Nippon Kagaku Zasshi 90, 159-162 (in Japanese with English abstract). WADA,H. and SUZUOKI,T. (1967) Continuous observation on the chemical composition of fumarolic gases from Nasudake volcano. Read at 1967 annual meeting of Geochem. Soc. Japan (in Japanese).