Chlorine isotope fractionation associated with volcanic activity at the Kusatsu-Bandaiko hot spring...

$: Chlorine isotope fractionation associated with volcanic activity at the Kusatsu-Bandaiko hot spring in Japan$
PLEASE SCROLL DOWN FOR ARTICLE

This article was downloaded by: [Musashi, Masaaki]On: 2 September 2008Access details: Access Details: [subscription number 902091254]Publisher Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,37-41 Mortimer Street, London W1T 3JH, UK

Isotopes in Environmental and Health StudiesPublication details, including instructions for authors and subscription information:http://www.informaworld.com/smpp/title~content=t713643233

Chlorine isotope fractionation associated with volcanic activity at the Kusatsu-Bandaiko hot spring in JapanMasaaki Musashi a; Takao Oi b; Hans G. M. Eggenkamp c; Motoyuki Matsuo a

a Graduate School of Arts and Sciences, The University of Tokyo, Tokyo, Japan b Department of Chemistry,Faculty of Science and Technology, Sophia University, Tokyo, Japan c Department of Geochemistry, Facultyof Geosciences, University of Utrecht, Utrecht, The Netherlands

Online Publication Date: 01 September 2008

To cite this Article Musashi, Masaaki, Oi, Takao, Eggenkamp, Hans G. M. and Matsuo, Motoyuki(2008)'Chlorine isotope fractionationassociated with volcanic activity at the Kusatsu-Bandaiko hot spring in Japan',Isotopes in Environmental and Health Studies,44:3,305— 313

To link to this Article: DOI: 10.1080/10256010802310051

URL: http://dx.doi.org/10.1080/10256010802310051

Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf

This article may be used for research, teaching and private study purposes. Any substantial orsystematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply ordistribution in any form to anyone is expressly forbidden.

The publisher does not give any warranty express or implied or make any representation that the contentswill be complete or accurate or up to date. The accuracy of any instructions, formulae and drug dosesshould be independently verified with primary sources. The publisher shall not be liable for any loss,actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directlyor indirectly in connection with or arising out of the use of this material.

http://www.informaworld.com/smpp/title~content=t713643233

http://dx.doi.org/10.1080/10256010802310051

http://www.informaworld.com/terms-and-conditions-of-access.pdf

Isotopes in Environmental and Health StudiesVol. 44, No. 3, September 2008, 305–313

Chlorine isotope fractionation associated with volcanic activityat the Kusatsu-Bandaiko hot spring in Japan

Masaaki Musashia*, Takao Oib, Hans G.M. Eggenkampc† and Motoyuki Matsuoa

aGraduate School of Arts and Sciences, The University of Tokyo, Meguro-ku, Tokyo, Japan; bDepartmentof Chemistry, Faculty of Science and Technology, Sophia University, Chiyoda-ku, Tokyo, Japan;

cDepartment of Geochemistry, Faculty of Geosciences, University of Utrecht, Utrecht, The Netherlands

(Received 13 September 2007; final version received 28 May 2008 )

Stable chlorine isotope compositions (δ37Cl, per-mil: ‰, vs. a standard sample of sea water) of Kusatsu-bandaiko hot water samples, taken regularly in the years between 1974 and 1995 in the Kusatsu-Shiranevolcanic region, Japan, were measured mass-spectrometrically. The results show that the δ37Cl values ofthe waters taken before 1984 were at around −0.12‰, whereas those after 1984 were at around +0.18‰.The δ37Cl values are thus distinct across 1984, which is consistent with the classification by the Cl toS molar ratio (Cl/S): the higher the Cl/S ratio, the larger the δ37Cl value. The δ37Cl value increasedas much as 0.30‰ during 5 years between 1980 and 1984. This isotopic enrichment is likely correlatedwith increasing Cl/S ratios, suggesting that the heavier isotope (37Cl) may have preferentially increasedin the original Cl source of the hot spring across 1984 when volcanic activity likely increased at MtKusatsu-Shirane.

Keywords: chlorine-37; secular change; sulphur-34; thermal water; volatile; volcanic gases

1. Introduction

Chlorine (Cl), known as a volatile element [1,2], is found in magmatic derivatives such as volcanicgases and hydrothermal waters. In a high-temperature gas phase, Cl mainly exists in the form ofHCl or NH4Cl, which is readily dissolved into the local circulating water and makes it thermalacidic water in a volcanic region. Thus, assuming Cl is simply transferred from magma to the hotspring water, the secular change of the Cl content in a volcanic gas may be documented to that inthe water.

The change in the chemical composition of the volcanic gas related to volcanic activities havebeen reported in literature [3–6]. Several studies reported possible correlation between chemicalchange in the water and volcanic activities [7–10]. It is probable that Cl in thermal water monitorsthe secular change of the gaseous Cl emitted from magma. Therefore, to monitor secular changes in

*Corresponding author. Email: [email protected]†Present address: Centro de Petrologia e Geoquimica, Instituto Suporior Tecnico, Avenida Rovisco Pais, Lisboa, Portugal

ISSN 1025-6016 print/ISSN 1477-2639 online© 2008 Taylor & FrancisDOI: 10.1080/10256010802310051http://www.informaworld.com

Downloaded By: [Musashi, Masaaki] At: 01:12 2 September 2008

306 Masaaki Musashi et al.

Cl incorporated into the hydrothermal water would be important to assess the correlation betweenthe behaviour of Cl and volcanic activity.

Chlorine has two stable isotopes, 35Cl and 37Cl, and their isotopic composition, notated as δ37Cl(a per-mil deviation relative to the standard mean ocean chloride, SMOC), significantly variesover the range of 10‰ in nature [11]. Concerning magmatic derivatives such as gas condensatestaken from island-arc volcanos, the highly positive value of +1.13‰ was found in a samplefrom the Ijen volcano (Indonesia) [12], and the highest δ37Cl value was found at the highest-temperature fumaloric vent among various fumarolic vents, i.e. +0.29‰ (860 ◦C) at SatsumaIwo-jima (Japan) [13], +1.9‰ (920 ◦C) at Kudriavy (Iturup Island, Kuriles, Russian Federation),and+3.9‰ (504 ◦C) atVulcano (Sicily, Italy) [14]. Notably, the last three volcanic systems showeda linear correlation between δ37Cl value and sampling temperature; the δ37Cl value increased withincreasing temperature, suggesting that the magmatic Cl was originally enriched in 37Cl. Recently,the magnitude of the fractionation between gaseous HCl and aqueous chloride was theoreticallyestimated as 2‰ at room temperature [15].

In spite of the relatively large isotopic deviation in the gas phases, Cl isotope fractionationbetween liquid and condensed phases obtained experimentally is in general very small. Forinstance, our chromatographic experiments showed that Cl isotope fractionation between hydratedCl− molecules in liquid and resin phases, such as Cl− (H2O)n and resin-bound Cl−(H2O)m,respectively, was as small as 0.23 or 0.30‰ at 25 ◦C of a single-stage separation [16]. Also,chlorine isotope fractionation between coexisting vapour and liquid in the H2O–NaCl systemperformed at high pressure and high temperature was generally within 0.2‰. It was impliedthat the evolved vapour–liquid samples such as volcanic fumarolic gases and degassed glassesshould have faithfully preserved the δ37Cl value of the fluid responsible for magma productionat depth [17]. Furthermore, our previous paper showed a very narrow range of the Cl iso-topic variation (−0.04 < (δ37Cl/‰) < +0.07), except for a unique sample (δ37Cl = +0.20‰,Kusatsu-Bandaiko, KB) among hot spring water samples taken at different geological settings inthe Kusatsu-Shirane volcano area, Japan [18].

KB is one of the major strongly acidic hot springs, and its temperature is the highest among hotsprings in the Kusatsu spring area located at the foot of Kusatsu-Shirane volcano (Gunma, Japan).The KB water was characterised as a volcanic thermal fluid, based on the δD-δ18O systematicsand chemical matrix [19]. Chemical compositions of the KB water have been analysed since1976 [10]. The analysis showed that the Cl− content has increased since 1982 (Table 1). TheCl/S molar ratio has been the highest among the hot springs in the area since 1975 when theanalysis started. The increase in the ratio during the recent 30 years was, in addition, only foundat KB.

In this study, the δ37Cl values of five water samples from KB, taken regularly between 1975and 1995, were examined in order to understand the relationship between the δ37Cl value and theCl/S ratio. We also discussed the influence of magmatic Cl on the KB water source using the Clisotope tracer.

2. Experimental

2.1. Water samples

This study employed the KB samples collected in 1970, 1976, 1980, 1984, 1990, and 1994 byDrs. Tomoko and Joyo Ossaka, and their co-workers at Sophia University, Japan. The samples,stored carefully in polypropylene bottles, have already been analysed for their chemical con-stituents. Water temperatures, pH values, and concentrations of Cl and S are referred in Table 1,and the other results are available in Ossaka et al. [10].


Isotopes in Environmental and Health Studies 307

Table 1. Analytical results on the stable isotopic composition of the Kustau-Bandaiko spring waters, 1976–1995.

Sampling date Sample water δ37Cl (‰)

Year Date Temperature (◦C) pH Cl (mmol l−1) S (mmol l−1) Cl/S First Second Average

1975 July 31 91.2 1.69 20.11 14.99 1.341976 July 31 93.9 1.69 21.75 15.51 1.40 −0.07 −0.14 −0.111977 August 02 93.4 1.75 19.32 13.43 1.441978 July 31 95.5 1.72 18.73 12.70 1.471979 July 24 95.4 1.70 18.28 12.49 1.461980 July 25 95.1 1.69 17.63 12.70 1.39 −0.08 −0.17 −0.131981 July 28 95.8 1.72 16.67 12.28 1.361982 December 23 92.9 1.81 16.05 11.87 1.351983 July 28 93.9 1.46 17.01 11.87 1.431984 August 01 92.9 1.57 18.59 11.87 1.57 0.22† 0.17† 0.20†

1985 July 25 97.6 1.60 19.69 12.07 1.631986 August 02 94.6 1.70 21.04 12.18 1.731987 July 27 93.4 1.66 23.24 13.12 1.771988 July 30 93.2 1.5† 22.06 13.19 1.671989 July 27 92.8 1.54 22.79 13.54 1.681990 July 26 93.7 1.43 22.79 13.52 1.69 0.19 0.10 0.151991 July 25 95.2 1.68 22.68 13.75 1.651992 August 04 95.2 1.68 22.48 13.71 1.641993 July 27 92.6 1.56 24.03 13.97 1.721994 July 25 95.2 1.54 25.59 15.02 1.70 0.27 0.14 0.211995 August 02 94.2 1.54 24.65 14.41 1.71

Note: The water temperature, pH, and Cl and S concentrations of the water samples were cited from [10].†The value was cited from [18].

2.2. Analytical procedures

This study employed the analytical technique of Cl extraction for δ37Cl measurements developedby Langvad [20] and others [12,21,22]. The procedure is summarised as follows.

An aliquot of a water sample containing at least 1.9 mg Cl was buffered and AgNO3 wasadded in order to extract Cl as AgCl precipitate. The precipitate, separated with a fibre-glassfilter paper, was dried in an oven. The Cl yield was >97% and was 100% in most runs. Thedried AgCl was reacted with CH3I to form CH3Cl. The CH3Cl thus formed was purified by gaschromatography and collected in a vacuum sample tube provided for mass spectrometry. Duringthe process ofAgCl-to-CH3Cl conversion to gas chromatographic purification, the recovery of Clwas >87% and was about 95% in most runs. In the case where Cl yield was <80%, a significantisotopic fractionation could have occurred during purification.

The purified CH3Cl was introduced into an isotope ratio mass spectrometer, VG SIRA 24EM, with a system of double-inlets and multi-collectors. The Cl isotope ratio was measuredas the peak height ratio of the beams of mass 52, CH37

3 Cl+, and mass 50, CH353 Cl+. The Cl

isotopic composition of a sample was expressed as a value of δ37Cl (‰)(=[RSample/RReference gas −RSMOC/RReference gas] × 1000), where R = 37Cl/35Cl, the reference gas is the analytical gradeCH3Cl gas, and SMOC is a standard sea water sample for Cl isotope analysis. The sea water wastaken from the Atlantic Ocean (33◦53′0′′N, 24◦20′2′′W) in December 1989. The δ37Cl value ofSMOC was defined to be zero per-mil relative to that of the sample. A day-to-day reproducibilityof the mass spectrometer was verified by measuring δ37Cl values of two isotopically independentCH3Cl gases. A difference between the two has been constant at +10.60‰(σ = 0.02‰) sinceJanuary 1995 [23]. At least three individual extractions of Cl from a water sample and from astandard seawater sample were carried out in this study. The overall precision (σ ) of Cl isotopeanalysis was better than ±0.06‰.



3. Results

Analytical results are summarised in Table 1 and Figure 1. In the table, the δ37Cl values ofduplicate analyses of each sample are shown. The δ37Cl values of all the analyses vary between−0.17 and +0.27‰. The significance of this variation is safely ignored, judging from the result ofa statistical test on a hypothesis that the population corresponding to the differences has a mean of0‰ under the following condition: the significance level of 1% and the variance of the populationof 0.062‰ [23]. This implies that the mean value of each sample represents the δ37Cl value ofthe sample within the analytical error.

Analytical results show that the mean δ37Cl value was −0.11‰ in 1976, stayed at around−0.12‰ between 1976 and 1980, and then increased to +0.19‰ in 1984, and stayed at around+0.18‰ for the next 11 years between 1984 and 1994. During 20 years between 1976 and 1994,the δ37Cl values thus all of a sudden increased by 0.30‰ in 1984. The magnitude of this gainwas nearly equal to the degree of Cl isotopic fractionation observed in the experiments of NaClprecipitation (0.26‰ [12]) and ion-exchange chromatography (0.30‰ s [15]).

During 1975–1995, profiles of secular changes in Cl and S concentrations showed slow v-shapesreaching minimum values in 1982: Cl (=16.1 mmol l−1) and S (=11.9 mmol l−1) (Table 1). Incontrast, as to the Cl/S molar ratio (Figure 1), the secularly taken samples are highly distinctiveacross 1984: Student’s t-distribution with 19 degrees of freedom under the significant level of5% showed that the Cl/S ratios taken before 1984 (n1 = 9, x̄1 = 1.41, and s2

1 = 0.00244) weredistinct from those taken after 1984 (n2 = 12, x̄2 = 1.68, and s2

2 = 0.00291).

4. Discussion

4.1. Drastic increase of Cl and S contents in the waters across 1984

On the basis of the published data [10] (Table 1), we show relationships between Cl and S contentsin the secular water samples in Figure 2. Their Cl/S ratios are statistically distinctive and dividedinto two groups A and B. The regression line (Line A) belonging to A is f (Cl/S) = 1.29 ×(Cl/S) + 1.53 (R2 = 0.88) and that (Line B) belonging to B is f (Cl/S) = 1.97 × (Cl/S) − 3.84(R2 = 0.90).

Figure 1. Secular changes of the δ37Cl values and the Cl/S molar ratio of the KB waters. The filled circles with errorbars indicate the δ37Cl , and the open circles represent the Cl/S molar ratios. The Cl/S secular change was drawn basedon the data [10]. The vertical bar tacked on the sample indicates the range between the duplicate analyses (Table 1).



Figure 2. Correlation between Cl and S contents of the secular samples taken from KB, based on the data [10]. Thenumber tacked on the circles indicates the year when the sample was taken: for instance, ’76 means 1976. The filledcircles indicate the samples taken before 1984, and the open circles represent those taken after 1984. The two solid lines(Lines A and B) are regression lines described in the text. The broken line (Line C) shows a regression line on the basisof the data belonging to the five samples taken in the period between 1982 and 1986.

Both Cl and S contents tended to monotonously decrease along Line A between 1975 and 1983,whereas both tended to monotonously increase along Line B between 1984 and 1995. Yamamotoet al. [24] analysed similar samples taken regularly from KB between 1989 and 1994 and reportedno significant trend in both the values (Table 2), although the variance of their values was largerthan that of the data we used [10].

The interesting points found in Figure 2 are as follows: (1) Cl and S contents were increaseddrastically from 1982 to 1986, and (2) the monotonous decrease in both Cl and S contents turnedover to a monotonous increase across 1984.

The sudden turnover of Cl and S contents across 1984 may have caused an increase in Na+ andK+ contents in the water samples after 1984, as reported by Ossaka et al. [10]. Acidic volatilessuch as HCl and SO2 may readily leach alkali cations from rocks during alteration. This fact waswell documented in the secular changes of the pH value and both Na+ and K+ contents in theKB waters. For instance, the KB water drastically became highly acidic (from pH 1.81 in 1982to 1.46 in 1983) and gradually recovered to pH 1.70 in 1986 (Table 1). This pH drop was likelyassociated with the increase in Na+ and K+ contents from 77 and 22.8 mg l−1 in 1983 to 122 and37 mg l−1 in 1986, respectively [10].

4.2. Isotopic shift across 1984 found in several isotopic compositions in the water samples

As seen in Figure 1, the δ37Cl values of KB water samples are constant at −0.12‰ before 1984and at +0.20‰ after 1984, indicating that the δ37Cl value was shifted across 1984. A similaroccurrence can be found in a study by Yamamoto et al. [24]. They reported that the δ34S, δ18O,and δD values in B were nearly constant at +18.6, −17.0, and −78‰, respectively. Unfortunately,none of these isotopic data in A was reported (Table 2). Instead, we examined the isotopic dataof hot spring waters taken at Kusatsu-yubatake (KY) located at a downstream site of KB. Itis probable that the KY waters were contaminated by the KB waters [10]. Excluding the databetween 1960 and 1965, the rest of the data taken between 1970 and 1979 were nearly constantat δ34S = +15‰, δ18O = −10.5‰, and δD = −75‰. The data between 1988 and 1994 werealso nearly constant at δ34S = +11.5‰, δ18O = −11.8‰, and δD = −78‰. Obviously, across1984, the δ34S, δ18O, and δD values were shifted or depleted isotopically. As far as the δ18O and



Tabl

e2.

Che

mic

alan

dis

otop

icda

taof

the

KB

and

KY

spri

ngw

ater

sta

ken

regu

larl

ydu

ring

1965

–199

4[2

4].

Sam

plin

gda

teSa

mpl

ew

ater

Yea

rD

ate

Tem

pera

ture

(◦C

)pH

Cl−

(mg

l−1)

SO2− 4

(mg

l−1)

Cl(

mm

oll−

1)

S(m

mol

l−1)

Cl/

Sδ

34S

(‰)

δD

(‰)

δ18

O(‰

)

Ban

daik

o19

89A

ugus

t17

94.0

1.7

862

1440

24.3

214

.99

1.62

18.3

1990

Aug

ust1

894

.81.

781

513

1022

.99

13.6

41.

6918

.6−8

2.4

−11.

219

92Se

ptem

ber

1396

.01.

781

612

7023

.02

13.2

21.

7418

.6−7

7.7

−11.

119

93A

ugus

t17

94.0

1.6

733

1320

20.6

813

.74

1.50

−77.

6−1

1.0

1994

Aug

ust3

098

.11.

689

212

3025

.16

12.8

01.

9718

.6

Yub

atak

e19

65A

ugus

t11

66.9

1.7

584

1470

16.4

715

.30

1.08

20.0

−75.

2-1

0.1

1970

June

1366

.51.

851

312

1014

.47

12.5

91.

1518

.3−7

5.8

−10.

719

75Ju

ne22

64.8

2.0

437

963

12.3

310

.02

1.23

13.3

−74.

0−1

0.6

1979

July

2464

.02.

039

191

411

.03

9.51

1.16

15.9

−71.

2−1

0.5

1984

July

2961

.31.

929

480

58.

298.

380.

9914

.9−7

7.8

−10.

619

88N

ovem

ber

2856

.32.

132

681

79.

208.

501.

0812

.619

89A

ugus

t19

56.9

2.1

317

829

8.94

8.63

1.04

11.1

1990

Aug

ust1

956

.72.

129

482

58.

298.

590.

9711

.5−8

1.0

−11.

819

91A

ugus

t20

55.8

2.0

314

840

8.86

8.74

1.01

11.2

−77.

1−1

1.7

1992

Sept

embe

r13

56.0

2.0

294

752

8.29

7.83

1.06

12.5

−78.

3−1

1.7

1993

Aug

ust1

755

.62.

029

381

28.

278.

450.

9812

.3−7

9.9

−11.

819

94A

ugus

t29

55.4

2.1

323

742

9.11

7.72

1.18

11.8



δD values taken between 1989 and 1994 are concerned, the data of KB and KY were in goodagreement, indicating that oxygen and hydrogen isotope fractionations in water samples taken attwo different geological locations (KY and KB) were negligible.

In summary, δ37Cl as well as δ34S, δ18O, and δD values in KB waters likely shifted across 1984in the 20-year period between 1975 and 1994.

4.3. Elucidation of the contamination of KB water by volatile Cl

As shown earlier, the Cl content increased drastically from A to B across 1984. If this is a cause ofthe concurrent positive shift of the δ37Cl, it meant that the heavier isotope (37Cl) was preferentiallyintroduced into a source of the KB hot spring water.

The major source of Cl to KB is most likely HCl, a magmatic volatile [25,26]. The increase ofCl in the KB water in 1984 was coincidentally synchronous to Cl increase in the Yugama craterlake water monitored during the period when the Kusatsu-shirane area became volcanically moreactive than usual [9,27]. Although the δ37Cl value of the acidic gas supplied to KB is unknown, if agas with a higher δ37Cl value was introduced into the KB source associated with volcanic activityin 1984, the synchronous increase in both Cl content and δ37Cl value across 1984 is explained asa preferential addition of the isotopically heavier volatile chlorine.

Sulphur isotopic compositions of dissolved SO2−4 in hot spring waters in the Kusatsu region

were reported as a resultant of a mixture of two sources: end member 1, a highly concentratedsulphate solution (ca. 0.4 mol l−1) of a high δ34S value at high temperature; and end member 2, adiluted sulphate solution (ca. 4 mmol l−1) of a low δ34S value at low temperature [24]. Yamamotoet al. [24] estimated that the former originated from volatile SO2 (δ34S = +29‰) and the latterwas an oxidised product of volatile H2S (δ34S = −2 to −6‰) degassed from the Kusatsu-Shiranevolcano. They interpreted that the decrease of the δ34S value in the KY water was caused byincreasing proportion of the isotopically light H2S component relative to the isotopically heavySO2 component in the water. As for the δ34S value of the KB water (Table 2), unfortunately,there was no report between 1975 and 1988. The S content in KB gradually decreased during theperiod 1975–1984 (Table 1), which may be due to the dilution of the KB water with end member1, as it was seen in KY. The drastic increase of S content in 1984 (Table 1), however, may needalternative explanation such as the addition of the end member 2 to the KB water. If this is thecase, the KB water would be contaminated by the volatile SO2 of δ34S = +29‰, suggesting thata high-temperature gas may be suddenly introduced into the KB water in 1984.

4.4. Correlation between δ37Cl and Cl/S ratio

In order to estimate the Cl/S ratio of the contaminant (C), we derive a mass-balance equationwith respect to 37Cl molar fraction, f = 37Cl/(35Cl + 37Cl), for A, B, and C:

rB × fB = rA × fA × (1 − α) + rC × fC × α (1)

where rA, rB, and rC are the Cl/S molar ratios of the different reservoirs (A, B, and C, respectively),andfA, fB, and fC are the 37Cl isotopic molar fractions of A, B, and C, respectively. The parameterα is defined as a proportion of C against A in B.

As a first approximation, we employ (δ37Cl)A = −0.11‰ (as fA = 0.242278) and rA = 1.40,the average values between 1975 and 1982, and (δ37Cl)B = +0.18‰ (asfB = 0.242331) and rB =1.70, taking the average values between 1986 and 1995. Thus, Equation (1) can be rewritten as:

rC = rB · fB − rA · fA · (1 − α)

fC · α= 0.072774 + 0.339189 · α

fC · α(2)



Figure 3. Estimation of the Cl/S molar ratio of the contaminant (C) (rC of vertical axis) at a proportion of C against Ain B (α of horizontal axis). The asymptotic curve represents a bi-mixing model between reservoirs A and B with differentdegrees of proportion (see Equation (2) and text). The broken line is an assumption of rB = 1.7 at any value of α.

Here, the δ37Cl values are converted to 37Cl isotopic molar fraction by using 0.31978 as the37Cl/35Cl ratio of the standard.

We assume that (δ37Cl)C = +0.29‰ [13], the value obtained in a gas condensate taken fromthe highest-temperature fumaloric vent at a Japanese volcano. The value of fC then becomes0.242351, and Equation (2) is drawn as an asymptotic curve representing a bi-mixing modelbetween B and C with different degrees of proportion (Figure 3).

In the Kusatsu region, chloride and sulphate dissolved in hot spring water most likely originatedfrom HCl and SO2 in magmatic volatiles [25,26]. By Equation (2), we could estimate the α value,only if we know the δ37Cl value and Cl/S ratio of the magmatic volatiles in this region, which weunfortunately do not know at present. Nevertheless, there is still an interesting coincidence thatthe synchronous increase in δ37Cl and Cl/S seems to have occurred across 1984 when volcanicactivity increased at Mt Kusatsu-Shirane [9]. This also means that δ37Cl and Cl/S may graduallydecrease over time, once the new volcanic activity stops. Future investigation is necessary to builda model explaining the secular changes in both properties over time.

5. Conclusions

Chlorine isotope compositions (δ37Cl) of five water samples, taken regularly at KB in the Kustatsuhot spring region during the recent 21 years, 1975–1995, were measured mass-spectrometrically.The main conclusions are summarised as follows:

(1) The δ37Cl values increased by 0.30‰ in 1984. The mean δ37Cl value of the samples takenbefore 1984 was −0.11‰, which was statistically distinct from that of the samples after 1984,+0.18‰.

(2) The positive shift of the δ37Cl value across 1984 was likely associated with increasing Cl/Sratio of the KB water from 1.40 to 1.70 during 1982–1986, due to a gain of Cl relative to thatof S. The δ37Cl as well as δ34S, δ18O, and δD values in KB waters likely shifted across 1984during the period between 1975 and 1994.

(3) The synchronous increase of the δ37Cl and the Cl/S ratio during 1975–1995 suggested thatthe heavier isotope (37Cl) may preferentially increase in a Cl source of the KB water.



Acknowledgements

We appreciate the referees’ critical review of the manuscript. We acknowledge Drs Tomoko Ossaka and Joyo Ossakafor kindly donating us the spring water samples. We thank Dr Rob Kreulen for his helpful comments on Cl isotopegeochemistry and Mr A.E. van Dijk for his assistance in mass spectrometry.

References

[1] W.W. Rubey, Geologic history of sea water – an attempt to state the problem, Geol. Soc. Am. Bull. 62, 1111–1147(1951).

[2] V.M. Goldschmdt, Geochemistry (Oxford University Press, London, 1954).[3] J. Hirabayashi, J. Ossaka, and T. Ozawa, Relationship between volcanic activity and chemical-composition of

volcanic gases – a case-study on the Sakurajima volcano, Geochem. J. 16, 11–21 (1982).[4] K. Noguchi and H. Kamiya, Prediction of volcanic eruption by measuring the chemical composition and amounts

of gases, Bull. Volcano. 26, 367–378 (1967).[5] R.E. Stoiber and W.I. Rose, Geochemistry of central American volcanic gas condensates, Geol. Soc. Am. Bull. 81,

2891–2911 (1970).[6] I.A. Menyailov, Prediction of eruptions using changes in composition of volcanic gases, Bull. Volcano. 39, 112–125

(1975).[7] W. Giggenbach, The chemistry of Crater Lake Mt. Ruapehu (New Zealand) during and after the 1971 active period,

N. Z. J. Sci. 17, 33–45 (1974).[8] B. Takano and K. Watanuki, Monitoring of volcanic-eruptions at Yugama crater lake by aqueous sulfur oxyanions.

J. Volcano. Geotherm. Res. 40, 71–87 (1990).[9] T. Ohba, J. Hirabayashi, and K. Nogami, Water, heat and chloride budgets of the crater lake, Yugama at Kusatsu-

Shirane volcano, Japan. Geochem. J. 28, 217–231 (1994).[10] J. Ossaka, T. Ossaka, J. Hirabayashi, T. Oi, T. Ohba, K. Nogami, Y. Kikawada, and N. Hida, Water quality changes

of Kusatsu hot springs caused by the emergence of Bandaiko hot spring, J. Balneol. Soc. Japan 47, 166–178 (1998)(in Japanese with English abstract).

[11] M.A. Stewart and A.J. Spivack, The stable-chlorine isotope compositions of natural and anthropogenic materials,Rev. Mineral. Geochem. 55, 231–254 (2004).

[12] H.G.M. Eggenkamp, Ph.D. thesis, University of Utrecht, 1994.[13] M. Musashi and H.G.M. Eggenkamp, presented at 10thAnnual Goldshmidt Conference, Oxford, 2000 (unpublished).[14] C. Wahrenberger, C.J. Eastoe, T.M. Seward, and V. Dietrich, presented at 7th Annual Goldshmidt Conference, 1997

(unpublished).[15] E.A. Schauble, G.R. Rossman, and H.P. Tayler Jr, Theoretical estimates of equilibrium chlorine-isotope fractiontions,

Geochim. Cosmochim. Acta 67, 3267–3281 (2003).[16] M. Musashi, T. Oi, and H.G.M. Eggenkamp, Experimental determination of chlorine isotope separation factor by

anion-exchange chromatography, Anal. Chim. Acta 508, 37–40 (2004).[17] A. Liebscher, J. Barnes, and Z. Sharp, Chlorine isotope vapor-liquid fractionation during experimental fluid-phase

separation at 400 degrees C/23 MPa to 450 degrees C/42 MPa, Chem. Geol. 234, 340–345 (2006).[18] M. Musashi, Stable chlorine isotope compositions in waters from the Kusatsu-Shirane volcanic system, Japan,

Radioisotopes 48, 100–106 (1999) (in Japanese with English abstract).[19] H. Sakai and O. Matsubaya, Isotopic geochemistry of thermal waters of Japan and its bearing on kuroko ore solutions,

Econ. Geol. 69, 974–991 (1974).[20] T. Langvad, Separation of chlorine isotopes by ion-exchange chromatography, Acta Chem. Scand. 8, 526–527 (1954).[21] J.W. Taylor and E.P. Grimsrud, Chlorine isotopic ratios by negative ion mass spectrometry, Anal. Chem. 41, 805–810

(1969).[22] R.S. Kaufmann, A. Long, H. Bentley, and S. Davis, Natural chlorine isotope variations, Nature 309, 338–340 (1989).[23] M. Musashi, G. Markl, and R. Kreulen, Stable chlorine-isotope analysis of rock samples: New aspects of chlorine

extraction. Anal. Chim. Acta 362, 261–269 (1998).[24] M.Yamamoto, T. Koike, F. Masui,A. Shiota, H. Tsurita,A. Ohtsuka, K. Nogami, and J. Ossaka, Isotope geochemistry

of hot spring waters on the eastern side of Kusatsu-Shirane volcano, Gumma Prefecture, Japan, J. Balneol. Soc. Japan47, 68–75 (1997) (in Japanese with English abstract).

[25] J. Hirabayashi, Formation of volcanic fluid reservoir and volcanic activity, J. Balneol. Soc. Japan 49, 99–105 (1999)(in Japanese with English abstract).

[26] T. Ohba, J. Hirabayashi, and K. Nogami, Volcanic hydrothermal system of Mt. Kusatsu-Shirane, Japan, J. Balneol.Soc. Japan 49, 163–175 (2000) (in Japanese with English abstract).

[27] J. Ossaka, T. Ossaka, J. Hirabayashi, T. Oi, T. Ohba, K. Nogami,Y. Kikawada, M.Yamano, M.Yui, and H. Fukuhara,Volcanic activity of Kusatsu-Shirane volcano, Gunma, and secular change in water quality of crater lake Yugama,Chikyukagaku 31, 119–128 (1997) (in Japanese with English abstract).


Chlorine isotope fractionation associated with volcanic activity at the Kusatsu-Bandaiko hot spring...

Documents

Transcript of Chlorine isotope fractionation associated with volcanic activity at the Kusatsu-Bandaiko hot spring...