Chapter 6 Geochemistry of Light Hydrocarbons in Subduction-Related Volcanic and Hydrothermal Fluids TARAN

Society of Economic Geologists Special Publication 10, 2003, p. 6174

Chapter 6

Geochemistry of Light Hydrocarbons in Subduction-Related

Volcanic and Hydrothermal Fluids

YURI A. TARAN,

Institute of Geophysics, Universidad Nacional Autnoma de Mxico 04510, Mexico

AND WERNER F. GIGGENBACH*

Institute of Geological and Nuclear Sciences, P.O. Box 31312, Lower Hutt, New Zealand

Abstract

During the last 20 years of his outstanding career, Werner Giggenbach collected and analyzed hundreds of samples of volcanic and hydrothermal gases from White Island volcano and geothermal systems in New Zealand, as well as many volcanoes and geothermal systems over the world. Hundreds of samples were ana- lyzed for C1-C6 hydrocarbons, including benzene. On the basis of the data set obtained for the White Island volcano, together with other available data, several general trends in the behavior of CH4, C2-C4 hydrocar- bons, and benzene are apparent, based on application of techniques developed by W. Giggenbach for the interpretation of crustal fluid composition in a high-temperature environment. The trends can be divided into two main types involving temperature-dependent equilibrium and mixing of carbon from magmatic and sedimentary sources.

The common statement that the CH4 concentration in volcanic gases decreases with increasing temper- ature is not true, as there are no temperature-dependent trends in the CH4/CO2 behavior until magmatic temperatures are reached. The concentrations of methane in hydrothermal fluids are controlled mainly by the source output, comprising organic matter buried with sedimentary rocks. Thermal decomposition of this organic matter at upper crust levels produces CH4 and light hydrocarbons as well as nitrogen accom- panied by a very high N2/Ar ratio. Therefore, the CH4-rich end member of hydrothermal fluids tends to have a high N2/Ar ratio. By contrast, subduction-related magmatic fluids have almost no methane despite having a high N2/Ar ratio due to degradation of subducted organic-rich oceanic sediments. Hence, vol- canic gases and hydrothermal fluids are characterized by two different relationships between CH4 concen- tration and N2/Ar ratio.

Two systems show a good correlation with sampling temperature in volcanic gases: alkane-alkene pairs with the same number of carbon atoms and ethene-benzene. Their concentration ratios (C2H6/C2H4, C3H8/C3H6, C4H10/C4H8, C2H4/C6H6) in volcanic gases are strongly dependent on the temperature of the fumarole, and these ratios change with temperature along metastable equilibrium paths. This means that the chemical redox reactions alkane + H2 = alkane and C6H6 + 3H2 = 3C2H4 are fast but kinetically con- trolled, probably through catalysis by oxides and sulfur species. Variations in alkane-alkane ratios in terms of the 2Cn = Cn 1 + Cn + 1 equilibrium, either for volcanic gases or for hydrothermal fluids, show no sys- tematic trends, and even at magmatic temperatures (>800C), the observed Cn/Cn 1 ratios often corre- spond to negative equilibrium temperatures.

The reaction CO2 + 4H2 = CH4 + 2H2O, mistakenly called the Fischer-Tropsch reaction by some geo- chemists, is unlikely to apply under redox and temperature conditions in the hydrothermal and magmatic environment. The only natural inorganic process in which reduction of CO2 (and carbon) could be pos- sible is the serpentinization of Mg-rich mafic rocks. Under conditions prevailing in the crust, the only process that results in equilibration within the CH4-CO2 system is the oxidation of methane. This can be facilitated by natural catalysts, which are usually oxides, but not native metals as in the case of the reduc- tion of CO2.

Introduction

WERNER GIGGENBACH collected and analyzed an enormous number of gas samples from volcanoes and geothermal sys- tems from New Zealand and the rest of the world. At the

Corresponding author: e-mail, [email protected]

*Deceased

center of his research was the study of gases from the White Island volcano. Apart from the major components (e.g., CO2, H2S, SO2, N2, Ar, CH4, H2), Giggenbach also analyzed a large set of light gaseous hydrocarbons starting with methane up to the light aromaticsbenzene and toluene. Only some of the hydrocarbon data for geothermal systems of New Zealand and other locations have been published (Giggenbach et al., 1990, 1992, Giggenbach and Corrales,

1

2 TARAN AND GIGGENBACH

1992; Giggenbach, 1995), whereas none of the hydrocar- bon data on volcanic samples are published. The first detailed thermodynamic and kinetic analysis of processes controlling the content of hydrocarbons higher than CH4 in hydrothermal fluids was presented in Giggenbach (1997a). The following review is based mainly on this paper as it discusses the essential problems of the geochemistry of carbon species in the high-temperature hydrothermal envi- ronment.

There are two main approaches to the description of hydrothermal hydrocarbon geochemistry. The first one deals with the so-called Fischer-Tropsch process, which at least theoretically in nature and practically in industry allows synthesis or reequilibration of hydrocarbons in a C- O-H system and assumes both reversible reduction of CO2 and oxidation of CH4 and hydrocarbons (Robinson, 1963; Kiyosu et al., 1992; Kiyosu and Asada, 1995). The second one is based on numerous observations, which show that the nature and distribution of hydrocarbon species in hydrothermal vapors is more consistent with the thermal degradation of kerogen rather than inorganic production and was first proposed by Gunter (1978). Nehring and Truesdell (1981) showed that hydrocarbons are a common constituent of geothermal vapor discharges. Des Marais et al. (1981) experimentally confirmed that thermal decom- position of organic matter is the most important source of hydrocarbons in subaerial geothermal systems. Thermody- namic calculations suggested that the proportions of higher hydrocarbons forming in equilibrium under natural conditions are extremely low (e.g., Taran, 1984; Takachi et al., 1987). Kiyosu and Asada (1995) and Capaccioni et al. (1993, 1995, 2001) have presented a large data set on hydrothermal hydrocarbons with some analyses from high- temperature magmatic fumaroles. Capaccioni et al (1995, 2001) ascribed formation of alkanes, alkenes, and aromat- ics in volcanic and hydrothermal gases mainly to cracking and reforming processes.

Even in the case of the completely thermogenic origin of CH4 and higher hydrocarbons one can expect that under

hydrothermal and magmatic conditions, the reequilibra- tion in a natural C-H-O system may occur due to increasing rates of reactions at elevated temperatures and possible catalysis by minerals. Darling (1998) has proposed a C2H6/CH4 geothermometer based on a data set for hydrothermal hydrocarbons. Sugisaki and Nagamine (1995) took advantage of the following simple reaction:

CH4 + C3H8 = 2C2H6. (1)

This reaction allows the pressure-independent internal equilibrium to be expressed in terms of the CH4/C2H6 and C2H6/C3H8 ratios. Giggenbach (1987, 1997a) demon- strated with many examples that the redox potential in hydrothermal systems is controlled by interaction with diva- lent and trivalent Fe in the rock contacted by the fluids, and in turn, the CH4/CO2 ratio in hydrothermal fluids is appar- ently close to equilibrium if the hydrogen fugacity is gov- erned by the Fe system of the hydrothermally altered rock.

Taran (1986, 1988) and Chiodini and Marini (1998), using large data sets from many hydrothermal systems, showed that CH4 is not in chemical equilibrium under natural hydrothermal conditions or that a heterogeneity exists in redox conditions of geothermal reservoirs.

High-temperature volcanic gases originate essentially from shallow magma degassing, and at the surface they rep- resent a mixture of magmatic, hydrothermal, and atmos- pheric end members. Their redox potential (hydrogen fugacity) is controlled by water vapor and oxidized and reduced volatile sulfur species, such as SO2 and H2S

(Giggenbach, 1987; Giggenbach et al., 1987). Even very high temperature (>600C) volcanic gases contain

detectable amounts of methane and higher hydrocarbons (Capaccioni et al., 1995; Kiyosu and Asada, 1995). Until now a detailed analysis of the sources and reactions of hydrocarbons in volcanic gases has been difficult because of the few data available. In this work, discussion of the sources, equilibria, and chemical kinetics of hydrothermal hydrocarbons, based on a large analytical data set obtained by W. Giggenbach in New Zealand and other geothermal regions, is presented. Some new data on the Kamchatkan volcanoes are also presented, and these are discussed along with published analyses, mainly from Japan and Italy. One of the main goals of this work is to constrain important aspects of the hydrocarbon geochemistry of hydrothermal and volcanic fluids, using simple thermodynamic calcula- tions and mixing relationships but without application of isotopic data. The importance of the so-called Fischer- Tropsch synthesis for geochemistry is discussed, using a standard approach for interpreting the distribution of reac- tion products in the context of catalytic chemistry.

Analytical Results

This study is based on detailed analyses of >200 gas sam- ples from high-temperature fumaroles, most of them from White Island and geothermal wells in New Zealand and other geothermal areas. A selection was made of partial analyses from White Island and other volcanoes contain- ing hydrocarbon abundances data in addition to informa- tion on the total gas content, N2/Ar ratios, and CO2 and H2 concentrations. Volcanic and hydrothermal gases were collected in preevacuated bottles containing 40 ml of a 5N NaOH solution for the absorption of acid gases such as CO2, H2S, SO2, HCl, and HF. The samples were analyzed

by well-established techniques, as described by Giggen- bach (1975).

Table 1 shows representative data for hydrocarbon abun- dances, total gas content, CO2, H2, and composition from

fumaroles of the White Island volcano, in order of increas- ing sampling temperature. The whole data set contains more than 90 analyses of hydrocarbons in dry gas. These data are not presented here but are used for diagrams where only the ratios among hydrocarbon concentrations are needed. The available analyses of hydrocarbons together with the total gas contents, CO2, H2, and N2/Ar from other volcanoes are shown in Table 2. Unfortunately the data set for hydrocarbons and the data set for major gas

3 GEOCHEMISTRY OF LIGHT HYDROCARBONS

TABLE 1. Hydrocarbons in Fumarolic Gases of White Island

Year ST Xg CO2 H2 CH4 C2H4 C2H6 C3H6 C3H8 C4 C6H6

1985 103 21 873 0.15 6.5 0.003 2.7 0.25 0.06 1.04 1984 106 49 911 0.4 1.3 0.13 2.6 0.26 1.9 1983 109 55 907 2.1 1.4 0.07 2.9 0.95 0.68 2.9 1979 110 22 790 0.31 4.3 0.01 2.3 0.18 0.04 1.7 1979 119 12 740 0.74 11 2.0 0.17 0.044 1.0 1978 120 26 877 0.7 5.7 0.001 1.9 0.004 0.11 0.021 1.5 1983 182 13 661 5.0 6.2 0.08 2.6 0.003 0.28 0.038 0.4 1983 182 13 665 4.8 3 0.06 2.5 0.003 0.16 0.5 1984 191 89 841 2.1 2.5 0.05 2.2 0.16 1.6 1984 193 11 489 1.9 5.2 0.07 2.2 0.11 1.3 1983 198 10 557 10 7.4 0.06 2.4 0.10 1.0 1979 208 25 241 9.7 2.7 0.06 1.5 0.003 0.08 0.03 1.6 1978 281 76 648 0.5 0.003 26 79 24 28 8 21 1979 295 71 683 1.1 0.02 0.12 1.2 2.5 1979 327 114 783 12.1 0.07 0.12 1.2 0.02 1.6 4.1 1.8 1983 370 79 579 1.6 0.02 4.2 2.2 0.5 0.5 1.0 1.8 1984 394 72 170 7.7 0.1 1.2 2.4 0.5 0.6 1.2 1983 416 99 600 3.1 0.01 1.1 2.0 1984 426 49 539 4.6 0.18 2.9 4.5 2.6 1.7 7.3 3.3 1983 430 52 558 35.8 0.26 2.3 3.3 3.1 1.5 7.0 0.9 1983 440 69 517 4.5 0.02 1.2 1.4 2.5 1.2 9.7 1.5 1984 457 59 493 9.2 0.05 3.3 3.6 1984 509 700 206 7.9 0.13 2.3 5 0.7 0.9 0.23 2 1979 510 52 331 14.2 0.94 0.41 0.9 0.007 0.003 0.002 1.1 1979 512 426 862 2.1 0.006 1.4 2.1 9.2 1978 520 301 805 12.5 0.009 7.1 2.3 2.0 1978 540 119 824 20 0.03 5.3 3.8 17 1978 540 120 786 18 0.018 5.8 2.3 20 7 27 0.8 1983 590 166 776 5.1 0.012 2.9 2.2 1983 611 177 720 26.1 0.04 5.6 7.3 3.0 1.4 1983 630 114 748 18.7 0.13 3.4 2.6 5.7 2.4 7.3 2.7 1983 655 133 754 17.2 0.10 2.7 2.8 0.7 0.9 1984 664 808 147 22.8 0.16 1.3 1.6 0.16 1983 685 253 831 9.4 0.02 5.8 7.7 9.2 1984 694 59 748 25.8 0.14 9.8 4.8 3.4 1.7 0.8 2.6 1983 745 120 719 15.6 0.01 10.9 8.8 1984 760 73 724 41.4 0.14 8.6 5.2 1984 772 56 684 41.6 0.22 3.1 2.6 3.8 1.8

Notes: ST = sampling temperature (C); Xg (mmol/mol) = the total gas content (mmol gas/(mol gas + mol H2O); CO2, H2, and CH4 (mmol/mol) = dry gas content; C4 = all butanes and butenes; all hydrocarbons (mmol/mol) normalized to CH4

components in White Island fumaroles were decoupled. Table 3 therefore shows the available data on CH4 concen-

trations in the total discharge and the N2/Ar ratios for White Island fumaroles separately from the hydrocarbons data. Most of the gas analyses containing hydrocarbons in hydrothermal fluids from New Zealand were published in Giggenbach (1995). The hydrocarbon concentrations in Kamchatkan hydrothermal fluids were published in Taran (1984, 1988). However, analytical data with concentrations of CH4 in samples from hydrothermal wells together with the N2/Ar ratio for different hydrothermal systems are dis- tributed in the literature. Table 4 is a compilation of these analyses.

Methane in Volcanic and Hydrothermal Gases

From the very beginning of the geochemical exploration of geothermal fields, methane was always used as one of the principal gas species for the estimation of reservoir tem- perature by using the so-called Fischer-Tropsch geother- mometer based in the following equilibrium (Gunter and

Musgrave, 1971; Ellis and Mahon, 1977; DAmore and Panichi, 1980; Giggenbach, 1980):

CO2 + 4H2 = CH4 + 2H2O. (2)

Giggenbach (1980) developed a technique that takes into account the distribution of gases between saturated vapor and liquid water, which became a common geo- chemical tool (DAmore and Celati, 1983; DAmore and Truesdell, 1984; Taran, 1986). For liquid-dominated reser- voirs in subaerial geothermal systems the direct use of this geothermometer (comparing values of the quotient and the equilibrium constant) gives equilibration temperatures that generally matched measured ones. In combination with another gas reaction (e.g., 2NH3 = N2 + 3H2), one can

use a grid discriminating temperature and the steam frac- tion dependences (DAmore and Celati, 1983). This approach assumes the attainment of full chemical equilib- rium between CO2, H2, CH4, and H2O, and a good agree-

ment between measured and calculated temperatures can

4 TARAN AND GIGGENBACH

TABLE 2. Hydrocarbons in Gases from Subduction-Related Volcanoes

Codes

(see

T

CO2

H2

CH4

Fig. 4) Volcano (C) (mol%) (mol %) (ppm) C2H4 C2H6 C3H6 C3H8 C4 C6H6 N2/Ar Ref.

v Vulcano 313 11.0 0.04 0.10 0.13 0.17 0.4 0.30 818 1 v Vulcano 329 9.32 0.003 0.30 505 2 v Vulcano 604 12.1 0.20 0.11 975 2 g Galeras 642 2.73 0.11 11 0.15 0.21 0.31 0.31 0.20 47 1,3 sw Showashinzan 750 0.47 0.18 60 200 4 sw Showashinzan 637 0.02 0.06 0.9 15 5.8 1.7 5.1 5 sw Showashinzan 617 0.27 0.21 0.4 39 1.1 3.3 6.7 5 k Kuju 350 0.29 0.005 0.3 0.67 0.60 0.24 0.24 6 k Kuju 143 5.36 0.001 2.5 65 7 m Merapi 803 2.60 0.24 21 16 5.3 21 13 4.3 54 8 m Merapi 575 3.79 0.16 2.2 55 33 12 12 29 80 8 p Papandayan 356 2.20 0.008 0.3 5.9 4.2 1.2 2.2 0.7 907 8 kr Koryak 220 4.97 0.07 600 6.7 40 8.3 88 8 a Avacha 473 2.26 0.05 3 15 7.1 5.5 2.3 0.5 275 8

ch Chornyi 344 0.88 0.08 0.3 19 15 5.6 3.1 1.2 826 8 gr Groznyi 174 0.72 0.001 20 2.6 8.1 1.8 3.0 4.9 200 8 ku Kudryavy 825 1.81 1.2 10 2.4 0.5 1.2 0.2 0.8 280 8 ku Kudryavy 705 2.11 0.24 6.3 11 7.5 4.3 5.2 198 8 md Mendeleev 114 0.57 0.0001 150 7.0 140 0.6 2.1 0.15 8 e Ebeko 144 1.13 0.00001 10 32 1.0 16 21 180 8,9

mt Mutnovsky 360 0.51 0.02 13 12 8 0.3 1.0 1.0 115 8,10 ki Kawah Ijen 187 11.4 0.002 0.1 331 11 ki Kawah Ijen 169 9.41 0.002 0.19 370 11 si S-Iwojima 870 0.34 0.54 5.0 275 12 ng Ngauruhoe 640 1.61 0.26 0.3 810 13 sh Mt. St Helens 540 0.91 0.24 0.2 618 13

Notes: Hydrocarbon concentrations (mmol/mol) are normalized to methane; the concentrations of CO2, H2, and CH4 in the total discharge (H2O + gases) and N2/Ar ratios are also presented

References: 1 = Capaccioni et al. (1995), 2 = Giggenbach and Matsuo (1991), 3 = Fischer et al. (1997), 4 = Mizutani and Sugiura, 1982, 5 = Kiyosu and Asada (1995), 6 = Mizutani et al. (1986), 7 = Saito et al. (2001), 8 = this work, 9 = Menyailov et al. (1986), 10 = Taran et al. (1992), 11 = Delmelle et al. (2000), 12 = Shinohara et al. (1993), 13 = Giggenbach et al. (1990)

TABLE 3. Methane Mole Fraction and N2/Ar Mole Ratio

ST XCH4 10

6 N2/Ar

100 230 156 101 110 30 104 79 471 106 430 213 107 500 87 110 94 411 110 82 34 120 150 190 120 76 38 139 62 110 156 96 251 229 110 47 327 120 306 350 26 633 355 66 37 356 4.9 630 365 27 122 365 42 397 396 19 164 513 33 91 515 14 970 520 2.7 1,220 540 2.2 530 540 3.3 920 550 2.3 650

ST = sampling temperature (C)

be considered as reasonable experimental proof for such equilibrium. However, if equilibrium is attained, and if the redox conditions are controlled in a uniform way, the equi- librium fugacity ratios CH4/CO2 should have monotonic temperature dependence along an appropriate buffer line. For geothermal systems, redox conditions of fluids, as shown by DAmore and Gianelli (1984) and Giggenbach (1980, 1987), are governed by the interaction of aqueous fluids with rock containing di- and trivalent Fe. Redox potential in terms of RH = log (H2/H2O) can be calculated by use of fayalite and hematite as thermodynamic proxies for mineral phases containing iron in these two oxidation states (Giggenbach, 1987). In this case the H2/H2O ratio remains practically constant with temperature, and a value RH = 2.8 can be taken to adequately describe the equilib- rium redox potential over a hydrothermal range of tem- perature, up to ~400C, buffered by typical altered volcanic host rocks. Another important assumption is that the fugac- ity of water vapor in volcanic hydrothermal systems is con- trolled by an NaCl brine. In this case, log H2O = 4.9 1,820/T, and hence, using other thermodynamic data (e.g., Stull et al., 1969; Barin and Knake, 1973): log (CH4/CO2) log (PCH4/PCO2) =

log (XCH4/XCO2)v = log Kc + 4RH + 2log H2O = (3) 5,280/T 11.12.

5 GEOCHEMISTRY OF LIGHT HYDROCARBONS

TABLE 4. Methane Mole Fraction and N2/Ar Mole Ratio in Gases from Wells and Steam Vents in Some Geothermal Fields

(2003). The Taran (1988) line represents the function (eq

Codes Geothermal

XCH4 fluids from wells at various aquifer temperatures (from

(see Fig. 3) system 106 N2/Ar Ref.

ru El Ruiz 1,680 295 1 ru El Ruiz 1.6 86 1 cf Campi Flegrei 33 738 1 kr Krafla 1.9 69 1 mt Mutnovsky 90 220 2 pz Pauzhetka 150 260 2 su Sumikawa 2 73 3 ni Nigorikawa 70 155 3 ni Nigorikawa 80 200 3 kw Kaverau 29 135 4 kw Kaverau 28 107 4 mo Mokai 18 68 4 mo Mokai 8 65 4 mo Mokai 2 70 4 br Broadlands 24 416 4 br Broadlands 26 429 4 br Broadlands 56 428 4 br Broadlands 15 192 4 br Broadlands 100 454 4 ro Rotorua 110 140 4 ro Rotorua 51 429 4 wr Wairakei 0.2 48 4 wr Wairakei 2.7 48 4 wr Wairakei 0.04 79 4 ld Larderello 230 632 5 ld Larderello 300 527 5 ld Larderello 960 913 5 G1 The Geysers 45 165 6 ge The Geysers 3,700 597 6 ge The Geysers 2.7 44 6

Taran, 1984, 1988), geothermal systems of New Zealand (Giggenbach, 1997a), and volcanic gases including White Island data from Tables 1 and 2. Two clusters of data points can be seen: one below 300C, which can be called hydrothermal, with the average CH4/CO2 ratio close to 102, and another >300C, which can be called volcanic, with the average CH4/CO2 near 10

4. The scattering of data

points is of three to four orders of magnitude for the same temperature, and there are no trends along any buffer, though the data for high-temperature ( 300C) hydrother- mal fluids are close to the intersection between the FeO- FeO1.5 and the Taran (1988) buffer lines. This may indicate a partial equilibration of CO2 and CH4 under hydrother- mal conditions, and the scatter of data points can be attrib- uted to the redistribution of CO2 and CH4 between steam

and water, resulting from their different solubilities (Giggenbach, 1980, 1987; Taran, 1986; Chiodini and Marini, 1998). However, large deviations of data from the buffer lines of more than 1.5 log units, or 100C, pre- cludes the use of CH4/CO2 ratios as a geothermometer, as discussed by Chiodini and Marini (1998).

The main reason for the apparent success of the Fischer- Tropsch geothermometer is the temperature dependence of the hydrogen partial pressure, which for liquid-domi- nated aquifers is very steep (Taran, 1986; Giggenbach,

1991). For volcanic gases, as shown by Giggenbach (1987), 1 Average for The Geysers References: 1 = Giggenbach et al. (1990), 2 = this work, 3 = Chiba

Lowenstern et al. (1999)

For hydrothermal fluids, H2O is controlled by coexisting steam and water and can be expressed as log H2O = log PH2O = 5.51 2,048/T (Giggenbach, 1980). In this case:

log (CH4/CO2) log (PCH4/PCO2) =

log (XCH4/XCO2)v = log Kc + 4RH + 2log PH2O = (3a) 4,625/T 10.4,

where P and X are partial pressures and mole fractions, respectively; subscript (v) stands for the vapor phase; Kc is

the equilibrium constant for reaction (2), and T is Kelvin. (Giggenbach, 1987). The calculated PCH4/PCO2 ratio is very sensitive to redox control because of the high weighting of the RH in equation (3). For example, if RH is controlled by

the function proposed by DAmore and Panichi (1980):

log PCH4/PCO2 = 0.30 853/T (3b)

(Taran, 1988; Chiodini and Marini, 1998). The redox dia- gram in Figure 1 is similar to that used by Giggenbach (1987, 1997a). Lines represent RH control by rock buffers involving the minerals fayalite, magnetite, hematite, and quartz, and the SO2/H2S gas buffer. For details see Giggen- bach (1987), and more recent discussion by Einaudi et al.

hydrogen partial pressure (concentration) is controlled by the so-called gas buffer redox reactions between reduced and oxidized sulfur species with H2S and SO2 as proxies (Fig. 2A). The XCH4/XCO2 ratio shows no obvious trend with temperature (Fig. 2B), but if log(XCH4/XCO2) 4RH is plotted versus temperature, as it is done in Figure 2C, a good correlation can be seen, with the trend close to the calculated equilibrium with the gas buffer. Such a correla- tion means that concentrations of CH4 in volcanic gases are

strongly controlled by the discharge temperature or that chemical reactions with CH4 are very fast. But this does not make sense. Methane cannot be easily produced from CO2 or oxidized to CO2. However, equilibrium hydrogen con- centrations controlled by the gas buffer attain equilibrium very fast and vary in the range of 100 to 800C by more than eight orders of magnitude. Therefore, the trend in Figure 2C is a consequence of the fast response of the water thermal dissociation to temperature variations.

If there is no tendency to equilibrate, what is the reason of the large variations of the CH4 concentrations in

hydrothermal and volcanic fluids? To answer this we can start by assuming that most of the methane comes from organic-rich sediments. One of the important signatures of natural gases produced by the degradation of organic-rich sediments is their high to very high N2/Ar ratio, because the degradation of organics produces N2 but not Ar. This was shown by Jenden et al. (1988) and Motyka et al. (1989) for natural gases from California and Alaska and used by Giggenbach (1992, 1997b) to develop the triangular plot of

6 TARAN AND GIGGENBACH

FeO

log

PC

H4

/PC

O 2

logX

CH

4 /

XC

O2

4R

H

logX

CH

4 /

XC

O2

RH

= logX

H2 /

XH

2O

B

3

2

1

0

-1

-2

-3

-4

-5

-6

-7

-8

-9

-10

White Island

Other volcanoes

NZ hydrothermal systems

Other hydrothermal systems

200 400 600 800

Temperature C

A -1

-2

-3

-4

-5

-6

-2

-3

-4

FeO

1.5

FIG. 1. Variations in observed values of log XCH4/XCO2 in hydrothermal fluids and volcanic gases (Tables 1 and 2) as a function of sampling tem- perature. Solid lines = redox control by potential mineral buffers involving fayalite-magnetite-quartz, the FeO-FeO1.5 (fayalite-hematite-quartz) buffer of Gigenbach (1987), magnetite-hematite, and the H2S/SO2 gas buffer at the brine H2O. The Taran 1988 line = empirical (average) buffer of H2, according to Taran (1988).

relative concentrations of N2, Ar, and He. If the main

source of methane in hydrothermal fluids is organic-rich sediments, a correlation should exist between the CH4 con- centration in the total discharge and the N2/Ar ratio. A high proportion of magmatic fluids released from subduc- tion-related volcanoes comes from the subducted oceanic sediments, rich in organics. High N2/Ar ratios in volcanic

gases of subduction zones have been observed at many vol- canoes, and this characteristic was interpreted as an indica- tor of the subduction-type magmatic fluid (Giggenbach, 1992, 1997b). However, in contrast to hydrothermal fluids circulating in the upper crust, magmatic fluids, if contain- ing detectable amounts of methane, should at the surface be characterized by CH4/CO2 ratios

7 GEOCHEMISTRY OF LIGHT HYDROCARBONS X

CH

4 m

ole

fra

ction

XC

H4 m

ole

fra

ctio

n

XC

H4 m

ole

fra

ction

G

1E-4

1E-5

1E-6

ru

ro

pz

ni ni

ge

ld

ld ld

ro br

rates are high enough to equilibrate the C-H-O system by oxi- dation of the excess methane. Even so, the equilibrium con- centrations of the resulting CH4 are extremely low.

Giggenbach (1997a), using a selected set of gas analyses, demonstrated an apparent equilibrium trend in CH4/CO2 ratios for a set of samples starting from low-temperature,

1E-7

1E-8

mt br kw kw br cf

mo br

mo su br

ge kr ru

wr

CH4-rich natural gases, through medium CH4, CO2-

enriched natural gases from the high-heat flow gas deposits in Thailand and ending with the high-temperature

hydrothermal systems in New Zealand. The apparent equi- librium trend has little to do with the equilibration by chem-

wr

1E-9 wr

10 100 1000

N2/Ar

FIG. 3. The observed correlation between the concentration of CH4

and the N2/Ar ratio in the fluids from geothermal wells and steam vents listed in Table 4.

These data correspond to low-temperature fumaroles and most probably are the result of a shallow subsurface boiling of a condensed mixture between magmatic brine and mete- oric water (Taran et al., 1997). A negative N2/Ar-CH4 cor- relation is clearly expressed in both plots.

In summary, the concentrations of methane in subduction- related hydrothermal and volcanic gases in general are not controlled by temperature-dependent chemical reactions, and instead probably depend on the mixing proportions from magmatic and sedimentary sources. The exception for typical magmatic redox conditions may exist at very high tem- peratures, near to magmatic environment, where reaction

ical reactions involving CH4. Rather, the decrease in the CH4/CO2 ratio is caused by the temperature-dependent CO2 production from dissolution of reservoir carbonates and oxidation of organic matter and, so, the observed trend reflects the admixture of CO2 from sedimentary sources.

Distribution of Higher Alkanes in Volcanic and Hydrothermal Fluids

Equilibrium distribution of Cn (n > 1) hydrocarbons in

an aqueous fluid is always monotonic and depends on the redox conditions, but for typical rock buffer systems and

pressures below 10 kbars, C1 > C2 >> Cn > Cn + 1... For more oxidized conditions, the Cn/Cn 1 ratios are lower than for reduced conditions, which can be seen from equa-

tions for subreactions controlling ratios Cn/Cn 1 of satu- rated hydrocarbon chains:

CnH2n + 2 + 2H2O = Cn 1H2n + CO2 + 3H2. (4)

This is an expression of the so-called hydrolytic dispropor- tionation of alkanes (Helgeson et al., 1993; Price and DeWitt, 2001), which skips the CH4 production step. Such a

presentation is convenient because it includes the redox term H2. The equilibrium Cn/Cn 1 ratio can be found from the following equation:

1E-3

A 1E-3 kr B

1E-4 1E-4

sw m

gr

1E-5

shallow boiling?

1E-5

1E-6

g mt e ku ku si

k m a

ch

n p

1E-6

White Island

1E-7 1E-8

ki sh

ki v

v v

100 1000 10 100 1000

N2/Ar N2/Ar

FIG. 4. Variations in the CH4 concentration in volcanic gases as a function of their N2/Ar ratio. A. White Island fumaroles. B. Other volcanoes. For codes see Table 2.

8 TARAN AND GIGGENBACH

log C

2H

6/C

3H

8

log C

H4/C

2H

6

log (XCn/XCn 1) =

3RH + log H2O + log CO2 log K4, (5) A 6

and the equilibrium constant for n = 2, 3, and 4 can be approximated by the following:

log K4 = 10.5 (5,150 + 310n)/T (6)

(Taran, 1984). These hydrolytic disproportionation reac- tions involve the stepwise breakage of a C-C bond of a higher carbon number alkane. They represent a final stage of oxidation of one link of the C chain to CO2, and the

reduction of the rest to the more positive nominal oxida- tion state of carbon, in the terminology proposed by Helge-

son et al. (1993). They suggested that a set of these irre- versible hydrolysis reactions at the oil-water interface produces light gaseous hydrocarbons.

It follows from equation (6) that for hydrothermal temper- atures of 250 to 300C, log K4 is close to zero, and the equi- librium Cn/Cn 1 ratio can be calculated from the value of RH and the total fluid pressure, if Ptotal = PH2O + PCO2 is assumed. The RH values for hydrothermal conditions are always nega- tive, close to 2.8, accepting the FeO-FeO1.5 buffer control. Thus, for hydrothermal fluids where total pressure is 1 kbar) conditions. The temperature relationships between temperature and CH4/C2H6 and C2H6/C3H8 ratios in vol-

canic and hydrothermal fluids are shown in Figure 5 together with equilibrium lines of redox buffers.

The fully empirical CH4/C2H6 geothermometer proposed

by Darling (1998) for geothermal systems is based on an assumption that there are temperature-controlled reactions

5

4

3

2

1

0

B 2.5

2.0

1.5

1.0

0.5

0.0

-0.5

200 400 600800

Temperature C

consuming ethane and/or producing methane as tempera- ture increases. We can suggest that this is the previously men- tioned irreversible hydrolysis of ethane, facilitated at higher temperatures. At close to magmatic temperatures one could expect that the CH4/C2H6 ratio in hot volcanic gases should show a clear temperature dependence compared to cooler hydrothermal fluids. The data plotted in Figure 5A show that this is not the case. By contrast, the CH4/C2H6 ratios in

White Island fumarolic gases (100800C) show no temper- ature dependence or correlation with redox buffers or the geothermometer line of Darling (1998). There is no indica- tion of temperature correlation with the C2H6/C3H8 ratios,

which similarly remain almost constant over the whole tem- perature range (Fig. 5B).

Following Sugisaki and Nagamine (1995), we plot analyt- ical data and calculated equilibrium lines on the CH4/C2H6 vs. C2H6/C3H8 diagram (Fig. 6A) using the pressure-independent equilibrium, as follows:

2C2H6 = CH4 + C3H8. (7)

The idea of metastable chemical equilibrium between adja-

FIG. 5. Variations in log CH4/C2H6 (A) and log C2H6/C3H8 (B) values in White Island fumaroles as a function of sampling temperature. For log CH4/C2H6 in (A) redox buffers are shown, as well as the empirical CH4/C2H6 geothermometer line proposed by Darling (1998). cent alkanes with a low carbon number (n 200C.

9 GEOCHEMISTRY OF LIGHT HYDROCARBONS

log C

2H

6/C

3H

8

log C

H4/C

2H

6

log C

H4/C

2H

6

A

4

3

B 2

4

1

3

0 2

-1 0 1 2 3 1

log C2H6/C3H8

0

Geothermal systems of

New Zealand

Dry pyrolysis of kerogen

2 -1

-1 0 1

2 3 4

1 log C2H6/C3H8

0

-1

-2

-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5

log C3H8/C4H10

FIG. 6. A. Variations in log CH4/C2H6 vs. log C2H6/C3H8 and log C2H6/C3H8 vs. log C3H8/C4H10 values in White Island fumaroles (Table 1). The lines represent expected ratios for equilibrium at 100, 400, and 1,000C. B. Variations of log CH4/C2H6 vs. log C2H6/C3H8 in hydrother- mal fluids of New Zealand (data from Giggenbach, 1995) and in the prod- ucts of the dry pyrolysis of kerogen (Tannenbaum and Kaplan, 1986).

In other words, there are some reaction pathways with fast kinetics for the reversible breaking and forming of the C-C bonds. Giggenbach (1997a) showed that this is only a coinci- dence and that the irreversible thermodegradation of long carbon chains gives the apparent equilibrium Cn/Cn 1 con- centration ratios due to random breakage of C-C bonds. Fig- ure 6 demonstrates that, in general, most of the data points for volcanic and hydrothermal alkanes lie far from the equi- librium lines, and thus there are no metastable equilibria among light alkanes under hydrothermal and shallow mag- matic conditions. It should be noted that the CH4/C2H6 ratios for hydrothermal fluids vary 3 log units and that the C2H6/C3/H8 ratio, in general, varies within 1 log unit (Fig. 6B). Dry pyrolysis of kerogen and bitumen in the presence of clay minerals at 200 to 300C as shown in Figure 6B gives

products with a much lower CH4/C2H6 ratio (Tannenbaum and Kaplan, 1985).

Unsaturated Hydrocarbons in Volcanic Fluids

As was shown by Seewald (1994, 2001) in a series of labo- ratory experiments at 300 to 350C and 350 bars pressure, low molecular weight aqueous hydrocarbons interact with water and Fe-bearing mineral assemblages under controlled redox conditions. The stability of aqueous hydrocarbons at elevated temperature and pressure reactions was found to be a function of time and temperature but depended also on the oxidation state and the presence of catalytically active aqueous sulfur species. However, the main mechanisms of this interaction are the stepwise oxidation and decomposi- tion of low molecular weight hydrocarbons and the produc-

10 TARAN AND GIGGENBACH

log

C3H

8/C

3H

6

log

C2H

6/C

2H

4

tion of CO2-rich and methane-enriched gas. An important finding from these experiments is that the alkane-alkene

4

pair and some other reduced-oxidized pairs with the same carbon number attained metastable thermodynamic equi-

3

librium states. They react reversibly according to a general simple redox reaction, as follows: 2

CnH2n + H2 = CnH2n + 2. (8)

The equilibrium constants of reaction (eq 8) for ethene- ethane and propene-propane pairs can be approximated as follows:

log K (n = 2) = 6.82 + 7,337/T, (9)

log K (n = 3) = 6.32 + 5,676/T + 222,490/T2. (10)

For the dehydrogenation (oxidation) of alkanes and the formation of alkenes more oxidized conditions, higher temperatures and lower pressures are preferable. We sug- gest that under appropriate geochemical conditions, this type of temperature-controlled reaction can be found in nature. Unsaturated hydrocarbons are very rare in hydrothermal fluids in detectable amounts, but they are always present in volcanic gases (Tables 1 and 2). From Fig- ure 7, where the ratios C2H6/C2H4 and C3H8/C3H6 are

plotted versus sampling temperature, the relative amount of light saturated and unsaturated hydrocarbons with the

same carbon number is apparently controlled by the tem- perature of a fumarolic vent. In other words, reaction (eq 8) in volcanic gases seems to be very fast, similar to the ther- mal decomposition of water that controls the H2/H2O ratio. In contrast to the experimental conditions (Seewald, 2001) where hydrocarbons interacted in the over pressured

1

0

A -1 4

3

2

1

0

-1 B

White Island Other volcanoes

200 400 600 800

Temperature C

and heterogeneous mixtures of water and mineral phases under static conditions, reactions in volcanic gases are thought to take place as the gas (vapor) phase flows through altered volcanic rocks. The mass action equation for the reaction (eq 8) can be written in the following form:

log (XCnH2n + 2/XCnH2n) = log K + RH + log H2O, (11)

where RH = log H2/H2O and K is one of the equilibrium constants (eqs 9, 10). If, following Giggenbach (1987), it is assumed that the fugacity of water vapor beneath the White Island crater is controlled by vapor saturated brine (log H2O = 4.9 1,820/T for

11 GEOCHEMISTRY OF LIGHT HYDROCARBONS

C2H4

C2H4

canic gases are too high and very far from equilibrium even at a high temperature. This may indicate that the source of hydrocarbons in volcanic gases is buried organic matter of varying maturity. Thermal degradation of these organics, probably by stepwise hydrolytic disproportionation or by some other mechanism of high-temperature thermal destruction, produces light hydrocarbons with concentra- tions much higher than at equilibrium. However, under high-temperature magmatic-hydrothermal conditions, light alkanes are reversibly disproportionate to alkenes and hydrogen but without breakage and rearrangement of the C-C bonds. Any reactions involving the breakage of C-C bonds may be fast enough to produce lower carbon num- ber hydrocarbons and eventually CH4, H2, and CO2; how- ever, within the dynamic fluid flow system of a volcano, this seems to be unlikely. The formation of new C-C bonds requires very specific conditions and catalysts, and in the natural high-temperature, magmatic-hydrothermal envi- ronment can be practically excluded.

Benzene in Volcanic Gases

Benzene and more complicated volatile aromatics like toluene and ethylbenzene were analyzed in volcanic and hydrothermal gases on the parts per million and parts per billion level in dry gas (Giggenbach and Corales, 1992; Capaccioni et al., 1995; Giggenbach, 1995; Darling, 1998). Aromatics are an important part of oil and sedimentary organic material, and the aromatic ring itself (benzene) is stable at a high temperature even under highly oxidizing conditions. It may be that there is geochemical importance in the fact that benzene can be synthesized in an empty (catalyst-free) reactor at 950C from a CH4 air mixture with a yield of 2.5 percent (York et al., 1996). This so-called oxidative aromatization of methane is the subject of many studies in the catalytic chemistry of hydrocarbons. Methane can be converted to benzene in the presence of oxides like MoO3 or WO3, and the mechanism of the aromatization

includes formation of ethane and ethene as intermediate reactants (Arutunov and Krylov, 1998, and references therein). We suggest that the presence of CH4 and unsatu-

rated hydrocarbons in volcanic gases could possibly control the concentration of benzene. The thermodynamic con- straints can be obtained from the following reaction:

3C2H4 = C6H6 + 3H2. (12)

The equilibrium constant can be approximated as log K12 = 1.36 + 3,446/T. Equilibrium (eq 12) is pressure dependent, and taking into account that the total pressure in volcanic gases is controlled mainly by the water vapor and log H2 = RH + log H2O:

ence of the saline brine or at 1 bar total pressure. In Figure 8A the data for White Island fumaroles lie between the FeO-FeO1.5 and gas buffer lines and also show a correlation

with temperature. At high temperatures, the data cluster near the gas buffer lines. In Figure 8B and C, C6H6 and C2H4 trends broadly correlate with temperature, despite considerable scatter. These trends suggest that some fast reactions occur involving unsaturated and aromatic hydro- carbons in volcanic gases. However, these reactions cannot fully equilibrate benzene as well as alkanes and alkenes. Their concentrations in volcanic (and hydrothermal, see Giggenbach et al., 1990; Darling, 1998) fluids are many orders of magnitude higher than expected for full equilib- rium. A decrease in the C6H6 concentration with an

increasing temperature suggests that benzene is irreversibly consumed (reduced?), partially converting to alkenes, and that new C-C bonds do not form under high-temperature volcanic-hydrothermal conditions. In other words, oxida- tive or reductive degradation are the only mechanisms of equilibration for hydrocarbons and CH4 in the high-tem- perature crustal environment. Relationships between con- centrations of alkanes and alkenes with the same carbon number, or between ethane and benzene, showing close to equilibrium ratios of concentrations, most probably corre- spond to metastable equilibrium; without forming or rear- ranging C-C bonds, the kinetic barriers of reactions ("waves in the metastability pond; Giggenbach, 1997a) are much lower than for the full disproportionation of hydrocarbons to CO2, H2, and H2O.

Concluding Remarks

The application of the CO2-CH4 reaction (eq 2) as a

geothermometer for hydrothermal fluids suggests that full equilibrium between members of the equation can be attained, or that methane (and other hydrocarbons) can be produced from CO2 under hydrothermal conditions simi-

lar to those in industry where the Fischer-Tropsch synthesis is the basis for the multitonnage production of various H- O-C-chemicals (Storch et al., 1951; Anderson, 1984). The original Fischer-Tropsch synthesis is a catalytic reaction of so-called synthesis gas CO/H2 mixture of different pro- portions conducted on native metal catalysts with the redox state corresponding to the pure H2-CO mixture. These con-

ditions are not generally typical for the crust as a whole. However, they may exist on a large scale along ocean-

spreading zones or locally within ophiolite formations where present-day serpentinization of Mg-rich mafic rocks can produce highly reducing conditions (Abrajano et al., 1988; Lyon and Giggenbach, 1990; Sano et al., 1993; Taran et al., 2002). Experiments confirm the possibility of reduc- ing CO2 to produce CH4 during serpentinization of olivine (Berndt at al., 1996; Horita and Berndt, 1999; McCollom

log XC6H6/X3 = log K12 3RH log H2O. (13) and Seewald, 2001). Limited bodies of coal and oil heated to a sufficiently high temperature by magmatic sources, as

Here Xi is concentrations in mole fraction of the total fluid. Lines in Figure 8 represent the equilibrium XC6H6/X

3

ratios expected for the gas and rock (FeO-FeO1.5) buffers

of the hydrogen fugacity with H2O controlled by the pres-

is the case for some geothermal fields in Kamchatka (Taran, 1988), could also be important. Whatever the real environment is, the hydrocarbons produced by the Fischer- Tropsch reaction should have a very special distribution,

12 TARAN AND GIGGENBACH

work like a normal industrial Fe-Co or Fe-Ni catalyst, yield- ing the same set of products with the same distribution, same conversion, and same selectivity. The main feature of the alkane distribution in the Fischer-Tropsch reaction is a maximum at a carbon number (n) between 3 and 5, depending mainly on two kinetic parameters, the probabil- ity of the chain growing at the end or the chain growing at the next-to-end carbon atom (Anderson, 1984). The distri- bution is called the Anderson-Shultz-Flory distribution, which for long chains does not differ from the Shultz-Flory

200

400 600 800

distribution and can be evaluated for the random chain growth or breaking. The production of ethane in the Fis- cher-Tropsch synthesis is always minimal for 1 < n < 6

logX

C6H

6

3lo

g X

C2H

4

log X

C2H

4

log X

C6H

6

Yie

ld,

mole

%

28 A 60

24 50

20 40

16

30

Basalt

Rhyolite

Dacite

250C, 50 atm

H2:CO = 1:1

FeCo

12

-6.0 20

B -6.5

10

-7.0

-7.5

-8.0

-8.5

0

1 2 3 4 5 6

Carbon Number (n)

-9.0

-6

-7

FIG. 9. Distribution of light alkanes in the Fischer-Tropsch synthesis on volcanic rock catalysts (Taran et al., 1981) and the Co-Fe industrial catalyst (one of numerous examples from Storch et al., 1951).

C

-8

-9

-10

Temperature C

FIG. 8. Variations in the quotient log XC6H6 3logXC2H6 (A), benzene (B), and ethene (C) concentrations in the total discharge (mole fractions) of fumarolic gases from White Island, as a function of the sampling tem- perature (Table 1).

quite different from those of natural CH4-rich gases or any natural gases from the Earths crust (Anderson, 1984).

Figure 9 shows the distribution of alkanes in the Fischer- Tropsch synthesis from a 1/1, CO/H2 mixture, at 250C

and 50-bar pressure in a flow-through reactor using basalt, dacite, and rhyolite as catalysts (Taran, 1980; Taran et al., 1981). The reaction has a long inductive period involving reduction of part of the iron in the rock to the native form but mainly to magnetite. Then, this natural catalyst starts to

because its next-to-end carbon is the end atom and ethane becomes an actively consuming intermediate in the Fischer- Tropsch synthesis.

It should be noted that Fischer-Tropsch distribution of the light alkanes has never been observed in natural gases, including volcanic and hydrothermal gases. Instead the dis- tribution of light alkanes in terrestrial gases is very similar to those obtained in experiments on thermal decomposi- tion (catalytic or not) of kerogen (Tannenbaum and Kaplan., 1985) and model organic compounds (Jackson et al., 1995) produced by the catalytic decomposition of oil or a long linear alkane in the presence of H2 (Mango, 2000).

The main crustal process involved with the transforma- tion of organic-rich sediments, including oil and gas, is recombination and oxidation. Oil may dissipate probably by a stepwise hydrolytic disproportionation mechanism

13 GEOCHEMISTRY OF LIGHT HYDROCARBONS

(Helgeson et al., 1993; Price and DeWitt, 2001), eventually converting in the oil-water interface to CO2 and CH4 at ele-

vated temperatures and pressures. Methane in the presence of oxides can participate in a number of reactions of oxida- tive conversion (Arutunov and Krylov, 1998), producing minor amounts of more complex hydrocarbons, oxy- genated hydrocarbons, and eventually, CO2 and H2O. This is probably the most important mechanism of leading to equilibration of CH4 and CO2 in high-temperature

hydrothermal systems. Terrestrial hydrothermal systems and volcano-hydrothermal systems do not produce methane and

higher hydrocarbons except through conversion of organic material at a high temperature.

Acknowledgments

This work was supported by the Marsden project Under- standing Crustal Fluids and partially by a grant from Okayama University. I wish to thank Ian Graham for the invitation to participate in the project and Agnes Reyes for assistance with Werners data and archives. Many thanks to Doug Sheppard for supplying much of the unpublished data on the gas chemistry of White Island and for his per- mission to use them. Discussions with Jeff Hedenquist and Minoru Kusakabe were very helpful. Marino Martini and Jeff Seewald have provided constructive reviews. I am espe- cially grateful to Stuart Simmons for the effort he put into the improvement of the paper.

REFERENCES

Abrajano, T.A., Sturchio, N.C., Bohlke, J.K., Lyon, G.L., Poreda, R.J., and Stevens, C.M., 1988, Methane-hydrogen gas seeps, Zambales Ophiolite, Philippines: Deep or shallow origin?: Chemical Geology, v. 71, p. 211222.

Anderson, R.B., 1984, The Fischer-Tropsch synthesis: New York, NY, Acad- emic Press, 301 p.

Arnorsson, S., Gronvold, K., and Sigurdsson, S., 1978, Aquifer chemistry of high-temperature hydrothermal systems in Iceland: Geochimica et Cosmochimica Acta, v. 42, p. 523536.

Arutunov, V.S., and Krylov, O.V., 1998, Oxidative conversion of methane: Moscow, Nauka, 361 p. (in Russian).

Barin I., and Knake O., 1973, Thermochemical properties of inorganic substances: Berlin, Springer, 921 p.

Berndt, M.E., Allen, D.E., and Seyfried, W.E., Jr., 1996, Reduction of CO2 during serpentinization of olivine at 300C and 500 bar: Geology, v. 24, p. 351354.

Capaccioni, B., Martini, M., Mangani, F., Giannini, L., Nappi, and G., Prati, F., 1993, Light hydrocarbons in gas-emissions from volcanic areas and geothermal fields: Geochemical Journal, v. 27, p. 717.

Capaccioni, B., Martini, M., and Mangani, F., 1995, Light hydrocarbons in hydrothermal and magmatic fumaroles: Hints of catalytic and thermal reactions: Bulletin of Volcanology, v. 56, p. 593600.

Capaccioni, B., Tassi, F., and Vaselli, O., 2001, Organic and inorganic geo- chemistry of low temperature gas discharges at Baia di Levante beach, Vulcano Iasland, Italy: Journal of Volcanology and Geothermal Research, v. 108, p. 173185.

Chiba, H., 1991, Attainment of solution and gas equilibrium in Japanese geothermal systems: Geochemical Journal, v. 25, p. 335355.

Chiodini, G., and Marini, L., 1998, Hydrothermal gas equilibria: The H2O- H2-CO2-CO-CH4 system: Geochimica et Cosmochimica Acta, v. 62, p. 26732687.

DAmore, F., and Celati, R., 1983, Methodology for calculation steam qual- ity in geothermal reservoirs: Geothermics, v. 12, p. 129140.

DAmore, F., and Gianelli, G., 1984, Mineral assemblages and oxygen and sulfur fugacities in natural water-rock interaction processes: Geochim- ica et Cosmochimica Acta, v. 48, p. 847857.

DAmore F., and Panichi C., 1980, Evaluation of deep temperatures by a new gas geothermometer: Geochimica et Cosmochimica Acta, v. 44, p. 549556.

DAmore, F., and Truesdell, A.H., 1984, Helium in the Larderello geot- hermal fluid: Geothermics, v. 13, p. 227238.

Darling, W.G., 1998, Hydrothermal hydrocarbon gases: 1. Genesis and geothermometry: Applied Geochemistry, v. 13, p. 815824.

Delmelle, P., Bernard, A., Kusakabe, Fischer, T.P., and Tacano, B., 2000, Geochemistry of the magmatic-hydrothermal system of Kawah Ijen vol- cano, East Java, Indonesia: Journal of Volcanology and Geothermal Research, v. 97, p. 3153.

Des Marais, D.J., Donchin, J.H., Truesdell, A.H., and Nehring, N.L., 1981, Molecular carbon isotopic evidence for the origin of geothermal hydro- carbons: Nature, v. 292, p. 826828.

Ellis, A.J., and Mahon, W.A.J., 1977, Chemistry and geothermal systems: New York, NY, Academic Press, 392 p.

Fischer, T.P., Sturchio, N.C., Stix, J., Counce, D., Arehart, G.B., and Will- liams, S.N., 1997, The chemical and isotopic composition of fumarolic gases and spring discharges at Galeras volcano, Colombia: Journal of Volcanology and Geothermal Research, v. 77, p. 229259.

Giggenbach, W.F., 1975, A simple method for the collection and analysis of volcanic gas samples: Bulletin of Volcanology, v. 39, p. 132145.

1980, Geothermal gas equilibria: Geochimica et Cosmochimica Acta, v. 44, p. 20212032.

1987, Redox-processes governing the chemistry of fumarolic gas dis- charges from White Island, New Zealand: Applied Geochemistry, v. 2, p. 143161.

1991, Chemical techniques in geothermal exploration, in DAmore, F, ed., Application of geochemistry of geothermal reservoir develop- ment: UNITAR/UNDP, Rome, p. 119144.

1992, The composition of gases in geothermal and volcanic systems as a function of tectonic setting, in Kharaka, Y.K., and Maest, A.S., eds., Symposium on Water-Rock Interaction: 7th, Park City, Utah, Proceed- ings, Rotterdam, Balkemap. 873878.

1995, Variations in the chemical and isotopic composition of fluids discharged from the Taupo volcanic zone, New Zealand: Journal of Vol- canology and Geothermal Research, v. 68, p. 89116.

1997a, Relative importance of thermodynamic and kinetic processes in governing the chemical and isotopic composition of carbon gases in high-heat flow sedimentary basins: Geochimica et Cosmochimica Acta, v. 61, p. 37633785.

1997b, The origin and evolution of fluids in magmatic-hydrothermal systems, in Barnes, H.L., ed., Geochemistry of hydrothermal ore deposits: New York NY, John Wiley, p. 737796.

Giggenbach, W.F., and Corales, R., 1992, The isotopic and chemical com- position of water and steam discharges from volcano-magmatic- hydrothermal system of the Guanacaste geothermal province, Costa Rica: Applied Geochemistry, v. 7, p. 309322.

Giggenbach, W.F., and Matsuo, S., 1991, Evaluation of results from second and third IAVCEI field workshop on volcanic gases: Mt. Usu, Japan, and White Island, New Zealand: Applied Geochemistry, v. 6, p. 125141.

Giggenbach, W.F., Martini, M., and Corazza, E., 1987, The effects of hydrothermal processes on the chemistry of some recent volcanic gas discharges: Periodico Mineralogia, v. 55, p. 1528.

Giggenbach, W.F., Garcia, N., Londono, P.A., Rodrigez, L., Rojas, N., and Calvache, M.I., 1990, The chemistry of fumarolic vapor and thermal spring discharges from the Nevado del Ruiz volcanic-magmatic- hydrothermal system: Journal of Volcanology and Geothermal Research, v. 42, p. 1339.

Giggenbach W.F., Paniagua de Gudiel, D., and Roldan Manzo, A.R., 1992, Isotopic and chemical composition of water and gas discharges from the Zunil geothermal system, Guatemala, in Geothermal investigations with isotope and geochemical techniques in Latin America: Interna- tional Atomic Energy Agency, 641, p. 245278.

Gunter, B.D., 1978, C1-C4 hydrocarbons in hydrothermal gases: Geochim- ica et Cosmochimica. Acta, v. 42, p. 137139.

Gunter, B.D., and Musgrave, B.C., 1971, New evidence on the origin of methane in hydrothermal gases: Geochimica et Cosmochimica Acta, v. 35, p. 113118.

Helgeson, H.C., Knox, A.M., Owens, C.E., and Shock, E.L., 1993, Petro- leum, oil field waters, and authigenic mineral assemblages: Are they in

14 TARAN AND GIGGENBACH

metastable equilibrium in hydrocarbon reservoirs?: Geochimica et Cos- mochimica Acta, v. 57, p. 32953339.

Horita J., and Berndt, M.E., 1999, Abiogenic methane formation and iso- topic fractionation under hydrothermal conditions: Science, v. 285, p. 10551057.

Jackson, K.J., Burnham, A.K., Braun, R.L., and Knauss, K.G., 1995, Tem- perature and pressure dependence of n-hexadecane cracking: Organic Geochemistry, v. 23, p. 941953.

Jenden, P.D., Kaplan, I.R., Poreda, R.J., and Craig, H., 1988, Origin of nitrogen-rich natural gases in the California Great Valley: Evidence from helium, carbon and nitrogen isotope ratios: Geochimica et Cos- mochimica Acta, v. 52, p. 851861.

Kiyosu, Y., and Asada, N., 1995, Light hydrocarbons in volcanic gases from the Japanese Island arc: Geochemical Journal, v. 29, p. 231242.

Kiyosu, Y., Asada, N., and Yoshida, Y., 1992, Origin of light hydrocarbon gases from the Matsukawa geothermal area in Japan: Chemical Geol- ogy, v. 94, p. 321329.

Lowenstern, J.B., and Janik, C.J., 2003, The origins of reservoir liquids and vapors from The Geysers geothermal field, California, Society of Eco- nomic Geologists Special Publication 10, p. 000000.

Lowenstern, J.B., Janik, C.J., Fahlquist, L.S., and Johnson, L.S., 1999, Gas and isotope geochemistry of 81 steam samples from wells in The Gey- sers geothermal field, Sonoma and Lake Counties, California, USA: U.S. Geological Survey Open-File Report OF 99-304, 28 p.

Lyon, G.L., and Giggenbach, W.F., 1990, Composition and origin of the hydrogen-rich seep, Fiordland, New Zealand: EOS, v. 71, p. 1717

Mango, F.D., 2000, The origin of light hydrocarbons: Geochimica et Cos- mochimica Acta, v. 64, p. 12631277.

McCollom, T.M., and Seewald, J.S., 2001, A reassessment of the potential for reduction of dissolved CO2 to hydrocarbons during serpentiniza- tion of olivine: Geochimica et Cosmochimica Acta, v. 65, p. 37693878.

Menyailov, I.A., Nikitina, L.P., and Shapar, V.N., 1985, Results of chemical monitoring of the activity of Ebeko volcano (Kurile islands) used for eruption prediction: Journal of Geodynamics, v. 3, p. 259274.

Mizutani, Y., and Sugiura, T., 1982, Variations in chemical and isotopic composition of fumarolic gases from Showashinzan volcano, Hokkaido, Japan: Geochemical Journal, v. 16, p. 6371.

Mizutani, Y., Hayashi, S., and Sugiura, T., 1986, Chemical and isotopic compositions of fumarolic gases from Kuju-Iwoyama, Kyushi, Japan: Geochemical Journal, v. 20, p. 264273.

Motyka, R.J., Poreda, R.J., and Jeffrey, A.W.A., 1989, Geochemistry, isotopic composition, and origin of fluids emanating from mud volcanoes in the Copper River basin, Alaska: Geochimica et Cosmochimica Acta, v. 53, p. 2941.

Nehring, N.L., and Truesdell, A.H., 1981, Hydrocarbon gases in some geothermal and volcanic systems: Transactions of Geothermal Resources Council, v. 2, p. 483486.

Price, L.C., and DeWitt, E., 2001, Evidence and characteristics of hydrolytic disproportionation of organic matter during metasomatic processes: Geochimica et Cosmochimica. Acta, v. 65, p. 37913826.

Robinson, R., 1963, Duplex origin of petroleum: Nature, v. 199, p. 113114.

Saito, G., Shinohara, H., and Kazahaya, K., 2001, Successive sampling of fumarolic gases at Satsuma-Iwojima and Kuju volcano, southwest Japan:

Evaluation of short-term variations and precision of the gas sampling and analytical techniques: Geochemical Journal, v. 36, p. 120.

Sano, Y., Urabe, T., Wakita, H., and Wushiki, H., 1993, Origin of hydrogen- nitrogen gas seeps, Oman: Applied Geochemistry, v. 8, p. 8389.

Seewald, J.S., 1994, Evidence for metastable equilibrium between hydro- carbons under hydrothermal conditions: Nature, v. 370, p. 285287.

2001, Aqueous geochemistry of low molecular weight hydrocarbons at elevated temperatures and pressures: Constraints from mineral buffered laboratory experiments: Geochimica et Cosmochimica Acta, v. 65, p. 16411664.

Shinohara, H., Giggenbach, W.F., Kazahaya, K., and Hedenquist, J.W., 1993, Geochemistry of volcanic gases and hot springs from Satsuma- Iwojima, Japan: Geochemical Journal, v. 27, p. 271285.

Stull, D.R., Westrum, E.F., and Sinke, G.G., 1969, The chemical thermody- namics of organic compounds: New York, NY, Wiley and Sons, 865 p.

Sugisaki, R., and Nagamine, K., 1995, Evolution of light hydrocarbon gases in subsurface processes: Constraints from chemical equilibrium: Earth

and Planetary Science Letters, v. 133, p. 151161. Takachi N.E., Barker, C., and Kemp, M.K., 1987, Stability of natural gas in

the deep subsurface: Thermodynamic calculation of equilibrium com- position: American Association of Petroleum Geochemists Bulletin, v. 71, p. 322333.

Tannenbaum, E., and Kaplan, I.R., 1985, Role of minerals in the thermal alteration of organic matter I: Generation of gases and condensates under dry conditions: Geochimica et Cosmochimica Acta, v. 49, p. 25892604.

Taran, Y.A., 1980, On catalytic properties of igneous rocks: Volcanology and Seismology, no. 5, p. 8083 (in Russian).

1984, Hydrocarbons C1-C4 under hydrothermal conditions: Akademiia Nauk SSR Doklady, v. 274, p. 11931196.

1986 Gas geothermometers for hydrothermal systems: Geochemistry International, v. 23, p. 111126.

1988, Geothermal gas geochemistry: Moscow, Nauka, 170 p. (in Russ- ian).

Taran, Y.A., Novak, F.A., and Bashkirov, A.N., 1981, Catalytic properties of volcanic rocks in the Fischer-Tropsch synthesis: Akademiia Nauk SSR Doklady, v. 271, p. 234237.

Taran, Y.A., Pilipenko, V.P., Rozhkov, A.M., and Vakin, E.A., 1992, A geo- chemical model for fumaroles of the Mutnovsky volcano, Kamhatka, USSR: Journal of Volcanology and Geothermal Research, v. 49, p. 269283.

Taran, Y.A., Connor, C.B., Shapar, V.N., Ovsiannikov, A.A., and Bilichenko, A.A., 1997, Fumarolic activity of Avachinsky and Koryaksky volcanoes, Kamchatka, from 1993 to 1994: Bulletin of Volcanology, v. 58, p. 441448.

Taran, Y.A., Fischer, T.P., Cienfuegos, E., Morales, P., 2002, Geochemistry of hydrothermal fluids from an intraplate ocean island: Everman vol- cano, Socorro island, Mxico: Chemical Geology, v. 188, p. 5163

York, A.P.E., Claridge, J.B., Green, M.L.H., and Tsang S.C., 1996, A study of the catalytic oxidative oligomerization of methane to aromatics: Catalysis Letters, v. 40, p. 207214.