Volcanic controls on ash iron solubility: New insights from high-temperature gas–ash interaction modeling

Journal of Volcanology and Geothermal Research 286 (2014) 67–77

Contents lists available at ScienceDirect

Journal of Volcanology and Geothermal Research journal homepage: www.elsevier.com/locate/jvolgeores

Volcanic controls on ash iron solubility: New insights from high-temperature gas–ash interaction modeling G. Hoshyaripour a,⁎, M. Hort a, B. Langmann a, P. Delmelle b a b

Institute of Geophysics, University of Hamburg/CEN, Bundesstr. 55, 20146 Hamburg, Germany Earth and Life Institute, Universitè catholique de Louvain, Croix du Sud, 2 bte L7.05.10, B-1348 Louvain-la-Neuve, Belgium

a r t i c l e

i n f o

Article history: Received 6 May 2014 Accepted 5 September 2014 Available online 19 September 2014 Keywords: Volcanic ash Thermodynamic equilibrium Iron release Eruption plume

a b s t r a c t Recent studies strongly suggest that volcanic ash can fertilize the surface ocean by releasing soluble iron. However, the volcanic and atmospheric processes that solubilize ash iron during its transport from the volcano to the ocean are poorly understood. Using thermodynamic equilibrium calculations, we investigate the influence of gas–ash interaction within the hot core (T N 600 °C) of the volcanic plume and the consequences of this for ash iron solubility. Simulations are performed by considering the plume hot core as a box model in which 1000 °C magmatic gas, ash and 25 °C ambient air are mixed together. We show that mixing and the resulting cooling of the gas–ash–air mixture affect the mineralogy and oxidation state of iron in the ash surface rim. Iron mineralogy in the ash surface layer after high-temperature plume processing is primarily governed by the ratio of the H2 and H2S content of the magmatic gas to the amount of entrained O2 into the hot plume (Xmix). The model results indicate that most of the iron in the ash surface layer is oxidized to ferric iron (Fe(III)) when log Xmix drops below −3.5 in the hot core. Such conditions may be encountered at convergent plate volcanoes, which release H2O-rich magmatic gases. In contrast, high temperature gas–ash interaction at divergent plate and hot spot volcanoes, which tend to be associated with CO2-rich and SO2-rich magmatic gases, respectively, may produce ash surfaces where iron mostly occurs as ferrous (Fe(II)). These volcanoes seem to be more favorable for iron fertilization because log Xmix does not fall below −3.5 and N 80% of the iron in the ash surface remains ferrous (Fe(II)), which is more soluble in water than Fe(III). © 2014 Elsevier B.V. All rights reserved.

1. Introduction After the 1991 eruption of Mt. Pinatubo, Philippines, it was hypothesized that ash in contact with seawater releases iron and other nutrients in sufficient amounts to the surface ocean to stimulate marine primary productivity (MPP) and in turn, global atmospheric CO2 drawdown (Sarmiento, 1993; Watson, 1997). Frogner et al. (2001) found that the ash from the eruption of Hekla in 2000, Iceland, released significant amounts of dissolved iron, silicon, and manganese together with sulfate, chloride and fluoride upon exposure to seawater. Subsequent studies have confirmed that volcanic ash affects MPP through rapid soluble iron release upon contact with seawater (Duggen et al., 2007). The first direct evidence of a phytoplankton bloom following fertilization by volcanic ash deposition was reported by Langmann et al. (2010) and later, Hamme et al. (2010) in the wake of the 2008 eruption of Kasatochi volcano in the Aleutian Islands. Achterberg et al. (2013) also reported a significant perturbation in the biogeochemistry of the Iceland Basin of the North Atlantic through the dissolved iron release from the ash erupted from Eyjafjallajökull, Iceland, in 2010.

⁎ Corresponding author. Tel.: +49 40 42838 5053. E-mail address: [email protected] (G. Hoshyaripour).

http://dx.doi.org/10.1016/j.jvolgeores.2014.09.005 0377-0273/© 2014 Elsevier B.V. All rights reserved.

Volcanic ash refers to tephra with a diameter of b 2 mm (Rose and Durant, 2009) and is typically composed of silicate glass and crystalline materials generated during an explosive eruption through magma fragmentation and to some extent, through erosion of the conduit wall rock (Heiken and Wohletz, 1992). Iron in volcanic ash produced through magma fragmentation is essentially found in non-soluble forms, i.e., in silicate glass and in primary Fe-bearing silicates and Fe-oxide minerals (Heiken and Wohletz, 1992; Schmincke, 2004). However, the source of bio-available iron involved in the alteration of the surface ocean's biogeochemistry is believed to be soluble iron species on the ash surface (Duggen et al., 2010; Hamme et al., 2010; Achterberg et al., 2013). Bio-availability of iron is suggested to be strongly linked to its solubility, which is influenced by chemical speciation (Fe(II) is more soluble), mineralogy (amorphous phases are more soluble), and Al substitution in Fe(III) oxides (Al-rich Fe phases are less soluble) (von der Heyden et al., 2012). Volcanic and atmospheric processes that modulate these properties and consequently, ash iron solubility are poorly constrained so far. Ayris and Delmelle (2012) emphasized that both high and low temperature reactions within the eruption plume can significantly alter the ash surface composition, and hence iron mineralogy and speciation. These reactions are expected to modify the surface reactivity of the ash, thus potentially influencing further (photo)chemical reactions during transport of the ash in the atmosphere.

68

G. Hoshyaripour et al. / Journal of Volcanology and Geothermal Research 286 (2014) 67–77

Within the hot core of the eruption plume (named as the effective source region in earlier studies; Martin et al., 2006; Roberts et al., 2009), where temperatures above 600 °C prevail, scavenging of magmatic volatiles (SO2, HCl and HF) can take place through gas–ash interaction (Rose, 1977; Óskarsson, 1980; Ayris et al., 2013). The reactions between these gases and the ash surface are thought to be partly responsible for the deposition of alkali and alkaline-earth sulfate and chloride salts on the ash surface (Delmelle et al., 2007). Recent modeling studies show that admixture of air with magmatic gases in the hot core of a volcanic eruption plume leads to a wide range of reactions and significant alteration of the gas composition (Gerlach, 2004; Martin et al., 2006, 2009; Bobrowski et al., 2007; Roberts et al., 2009). In the hot core (T N 600 °C), the reaction rates are sufficiently fast to ensure that the system is close to thermodynamic equilibrium conditions (Symonds and Reed, 1993). Here we investigate gas–ash interaction in the hot core of the volcanic plume in order to explore the effect of this zone of the plume on iron mineralogy and oxidation state in the ash's surface. The model framework and assumptions are presented in Section 2. High temperature (here after referred to as high-T) gas–ash interaction is then simulated for three distinct volcanic gas compositions to represent the geochemical patterns of main volcano-tectonic settings. A sensitivity analysis with respect to initial conditions (e.g., magma oxidation state, ash composition, etc.) is performed in Section 3. Finally, the influence of ash processing within the hot core on the redox state of iron in ash is discussed in Section 3. We hasten to emphasize that this study does not intend to reproduce any data or specific situation at one volcano but to explore the role of high-T volcanic gas–ash interaction in iron speciation and mineralogy in different tectonic settings in general. 2. Methodology 2.1. Conceptual model We consider the hot core of sub-plinian and plinian eruption plumes as a box model which covers the chemical processes in the temperature range 1000 °C N T N 600 °C (Fig. 1). Assuming thermodynamic

equilibrium in this zone of the plume, we use GASWORKS (Symonds and Reed, 1993), to simulate the interaction among magmatic gases, ash and atmospheric gases. GASWORKS computes heterogeneous equilibria among gases, solids and liquids during the processes of cooling, gas–gas mixing, pressure changes and gas–rock reactions. For more details about GASWORKS we refer the readers to Symonds and Reed (1993). In this study we assume that only surface layer of the ash (the 1–100 nm thick rim shown in Fig. 1) interacts with the surrounding gases according to thermodynamic equilibrium (kinetics is negligible). The theory and calculations behind this assumption are explained in details in Appendix A.

2.2. Gas and ash mixture Olgun et al. (2011) observed that ash from different volcanic settings release different amounts of iron upon contact with seawater. Therefore, we use the high-T gas composition corresponding to the three main volcanic settings (convergent plate (CP) or H2O-rich, divergent plate (DP) or CO2-rich and hot spots (HS) or SO2-rich) (Table 1). Although these average compositions do not rigorously cover all the observed high-T magmatic gases, they satisfactorily represent the geochemical differences between tectonic settings (e.g., water, sulfur, carbon and halogen contents; Symonds et al., 1994). The solid mass associated to particle sizes b64 μm is considered in this study, which may represent a substantial contribution (N50%) to tephra deposits from explosive volcanic eruptions (Rose and Durant, 2009). Particles in this size range not only have more specific surface area for interaction with the gas phase (Delmelle et al., 2005) but also can be lifted to high altitudes (Sparks et al., 1997). In order to determine the initial ash bulk composition, we use the method recommended by Symonds and Reed (1993) where magma of a given composition (for the whole rock composition used here see Table 2) is titrated into the gas phase step by step until a certain gas/rock ratio is reached. GASWORKS then determines the composition of minerals being in equilibrium with the given gas composition at a prescribed temperature, oxygen fugacity (fO2) and pressure (P = 1 atm).

Fig. 1. Sketch of the box model used for simulating gas–ash interaction within the volcanic plume hot core.

G. Hoshyaripour et al. / Journal of Volcanology and Geothermal Research 286 (2014) 67–77 Table 1 Average high-T volcanic gas composition of convergent plate (CP), divergent plate (DP) and hot spot (HS) tectonic settings which are associated with H2O, CO2 and SO2-rich gases, respectively (concentrations are in mol%) (Hoshyaripour et al., 2012). Volcano

CP

DP

HS

log fO2 T (°C) H2O CO2 H2 H2S SO2 HCl HF CO

−10.94 1000 91.05 4.50 1.44 0.29 1.76 0.76 0.06 0.11

−10.89 1000 75.78 13.38 1.12 0.85 7.39 0.42 0.42 0.32

−10.39 1000 75.94 3.25 0.64 0.39 19.16 0.17 0.18 0.04

Iron in volcanic ash occurs mainly as fayalite/forsterite ((Fe,Mg)2SiO4), magnetite (Fe3O4), ulvöspinel (Fe2TiO4) and ilmenite (FeTiO3)) (Ayris and Delmelle, 2012). For simplicity we consider minerals containing Si, Fe, Ti and Mg oxides only and assume all other phases to be inert. This assumption facilitates the convergence of the GASWORKS during calculation. We did carry out all calculations using the full mineral assemblage down to about 800 °C and did not find any significant differences between the full calculations and the ones using the reduced mineral assemblage, which we take as an indication that the reduction of the phase is valid. The equilibrium mineral assemblages are assumed to be uniformly distributed in the ash. Therefore, their relative ratios are also representative for the composition one would find on the surface of ash samples. In this study, we only simulate the crystalline part of the volcanic ash (glass is not considered) produced during near surface magma fragmentation (Symonds and Reed, 1993). These limitations (formation of glass, feldspars, pyroxene, etc.) are further discussed in Section 4.4. Since this study focuses on vent-near processes, we can assume that about 1–10 wt.% of the plume is gas and 90–99 wt.% is ash (Sparks et al., 1997) which converts to a gas to ash ratio (G/A) of approximately 0.01–0.1. A built-in assumption in GASWORKS is that the ash as a whole takes part in the reactions. We have argued earlier that only the surface rim (1–100 nm) of ash particle is available for reactions and therefore, the G/A ratio has to be adjusted accordingly. Previous studies suggest that the specific surface area of volcanic ash is in the range 1.1–2.1 m2/g (Delmelle et al., 2005; Mills and Rose, 2010). Assuming the mixture of 1 mol pure magmatic gas with mean molar weight of approximately 25 g and 3 wt.% gas in the plume, the total amount of ash coexisting with 1 mol of volcanic gas is ~ 830 g. Considering an average specific surface area of 1.6 m2/g, a density of 2500 kg/m3 for the volcanic ash and assuming the reaction surface thickness range of 1–100 nm, we calculate the mass of the ash surface layer involved in the reactions to be 4– 350 g. This amount would be correct in case the material is totally crystallized. In explosive volcanic eruptions, however, the crystal fraction is in the range of 20–60% (Blundy et al., 2006). Therefore, in all simulations we assume that only 40% of the calculated mass is crystallized. Hence, the mass of the ash available for chemical reactions would be 2–140 g, which corresponds approximately to G/A = 0.1–10. We select G/A = 0.3 (corresponding to ~ 80 g solid

Table 2 The magma (whole rock) composition used in this study (Moune et al., 2009). Oxides

wt.%

Oxides

wt.%

SiO2 FeO Fe2O3 MgO TiO2

54.5 8.1 3.5 2.5 2.2

Al2O3 CaO Na2O MnO K2O

14.3 6.5 4.1 3.1 1.2

69

phase) for our reference calculations. The sensitivity of the results to the G/A ratio (changing G/A) in the range of 0.1–10, which reflects variations in the assumptions above (degree of crystallization, ash surface rim thickness, etc.), is discussed in detail in Section 4. Volcanic plume (1000 °C mixture of magmatic gases and ash) is then mixed incrementally with the 25 °C ambient air (78% N2, 21% O2, O.1% Ar) at constant pressure (1 atm), thereby the system is cooled. The equilibrium temperature (Teq) of the mixture at each step is: Xk ni C i T i T eq ¼ X1k nC 1 i i

ð1Þ

where ni is the mass fraction, Ci is the heat capacity in J/kg K and Ti is the temperature in K for the component i in the system. Values of 1000, 1100 and 1840 J/kg K are used here as the heat capacity of air, ash and water vapor, respectively (water vapor represents the magmatic gas here since it occupies N70 wt.% of the erupted gas) (Sparks et al., 1997). We modified the mixture temperature calculation in GASWORKS in order to reflect the nonlinearities in the heat capacities according to Eq. (1). The simulation starts by assuming nH2 O ¼ 0:03, nash = 0.97 and nair = 0. Air is mixed incrementally with the magmatic gas–ash mixture by increasing nair until a temperature of 600 °C is reached. Changing the initial n values (gas 1–10 wt.% and ash 90– 99 wt.%) slightly influences the cooling rate but the results presented below are not sensitive to the cooling rate for a wide range of variation. We note that these cooling regimes are independent from G/A ratio (the amount of solid material available for chemical interaction) discussed above. In other words, while the whole ash is considered in the cooling rate calculations, only the surface rim is counted in ash–gas chemical interactions. Inherent to the mixing procedure described above are two different physical processes: cooling of the plume and mixing of the plume with air (oxidation). In order to understand the effect of each process we perform two distinct simulations: cooling without air entrainment and cooling with air entrainment. 3. Results 3.1. Initial ash mineralogy and composition Following the procedure outlined above the minerals in equilibrium with the gas phase are determined for three selected gas compositions. The results of the calculations are shown in Fig. 2. In all three settings, fayalite (Fe2SiO4) and ilmenite are the main Fe-carrying species. Due to the slightly higher oxygen fugacity (fO2 = −10.39), small amounts of magnetite and ulvöspinel precipitate in the HS setting (S-rich gas). While in CP and DP (H and C-rich, respectively) cases the fO2 is close to FMQ buffer (fayalite–magnetite–quartz) which is about − 11.00 at 1000 °C (Lindsley, 1991), the HS scenario is nearly 0.6 log units above FMQ allowing the formation of more oxidized minerals. 3.2. Cooling without air entrainment The processes discussed hereafter are assumed to influence only the surface rim of the ash (1–100 nm-thick layer), leaving the inner part of the ash (bulk composition) unaffected. The 1000 °C mixture of magmatic gas and ash simulated in the previous section is cooled to 600 °C without adding air into the system (no additional oxygen). Results are shown in Table 3 where magnetite is the main iron species for all settings considered. In the absence of atmospheric oxygen (O2), magnetite is produced by oxidation of ferrous ions (Fe2+) present in the crystal lattice of the Fe-endmember fayalite by the protons (H+) of water according to the following reaction: 3Fe2 SiO4 þ 2H2 O → 2Fe3 O4 þ 3SiO2 þ 2H2 :

ð2Þ

70

G. Hoshyaripour et al. / Journal of Volcanology and Geothermal Research 286 (2014) 67–77

CP

DP The surface rim 60−200 nm

Fayalite 19%

HS Magnetite 2% Ulvospinel Fayalite 1% 16%

Fayalite 19%

Forsterite 6%

Forsterite 6% Ilmenite 5%

Forsterite 6% Ilmenite 6%

Ilmenite 6% Tridymite 69%

Tridymite 69%

Tridymite 70%

Fig. 2. Mineral assemblages found in our starting material at 1000 °C and 1 bar (in wt.%). We assume that these minerals are uniformly distributed in the ash. However, cooling and mixing processes only affect those mineral phases exposed at the surface and in a small layer below the surface (the 1–100 nm rim shown with dotted line and discussed in the Appendix A), which is only few wt.% of total ash not the bulk composition. Hereafter, only Fe-carrying minerals are reported in tables.

Ulvöspinel can substitute ilmenite: 2FeTiO3 þ Fe2 SiO4 → 2 Fe2 TiO4 þ SiO2 :

ð3Þ

And finally, pyrrhotite (Fe0.8S) forms according to the following reaction between magnetite and H2 and H2S (Kishima, 1989): 0:8 Fe3 O4 þ 0:2H2 þ 3H2 S → 3Fe0:8 S þ 3:2H2 O:

ð4Þ

Conversion of magnetite to pyrrhotite also leads to scavenging of H2S by ash. The model predicts that 75%, 32% and 11% of the erupted H2S is scavenged by the ash surface during cooling in the CP, DP and HS scenario, respectively. 3.3. Cooling with air entrainment Mixing of hot magmatic gas with ambient air (T = 25 °C) in subplinian and plinian eruption plumes results in concurrent oxidation and cooling, which modifies both the gas phase and ash surface composition (Fig. 3). At T N 900°C magnetite forms due to fayalite oxidation: 3Fe2 SiO4 þ O2 → 2Fe3 O4 þ 3SiO2 :

ð5Þ

Then, at 750 °C b T b 900 °C hematite (Fe2O3) becomes stable and replaces magnetite: 2Fe3 O4 þ 0:5O2 → 3Fe2 O3

ð6Þ

Finally, at 600 °C b T b 700 °C sulfuric acid reacts with hematite and produce iron sulfate: Fe2 O3 þ 3H2 SO4 → Fe2 ðSO4 Þ3 þ 3H2 O:

ð7Þ

Table 3 Fe-carrying minerals at the ash surface (1–100 nm rim) for different gas compositions at initial conditions (1000 °C) and after cooling to 600 °C (without and with airentrainment). Volcano type

CP

T (°C)

Mineral

wt.%

Mineral

wt.%

Mineral

wt.%

1000

Fayalite Ilmenite

76 24

Fayalite Ilmenite

76 24

600 (w/o air)

Fayalite Magnetite Ulvöspinel Pyrrhotite Ilmenite Hematite

Fayalite Magnetite Ulvöspinel Pyrrhotite

20 43 9 29

Fayalite Ilmenite Magnetite Ulvöspinel Fayalite Magnetite Ulvöspinel Pyrrhotite

67 21 7 6 19 52 8 21

Hematite Fe3+ sulfate

98 2

Hematite Fe3+ sulfate

72 28

600 (with air)

DP

29 32 13 19 7 100

Vertical lines in the gas phase graphs in Fig. 3 show the transition stage, which corresponds to the depletion of reduced species (e.g., H2S and H2) at the expense of oxidized compound (e.g., SO2 and H2SO4) production and thus, transition of the system to an oxidized state. It is suggested that during mixing with ambient air in the hot core, magmatic gases are completely oxidized through a single-step transition (Martin et al., 2006; Bobrowski et al., 2007; Roberts et al., 2009). However, Fig. 3, shows that the transition involves two steps when ash is present as an additional sink for oxygen. The first step (dashed vertical lines) starts around 840–880 °C in all scenarios whereby H2S and H2 rapidly decrease and H2SO4 increases. The concentrations become stable prior to the second step (dotted vertical line) after which the rapid changes in species concentrations take place and the system becomes completely oxidized. Minerals at the ash surface can react with oxygen (e.g., reactions (5) and (6)) in the hot core and affect the transition. These reactions become more important as soon as the system passes the first transition step (dashed vertical line in Fig. 3). Consumption of oxygen by the solid phase stabilizes the concentration of gas species until the depletion of reduced solid species whereby the second transition step (dotted vertical line) occurs. This controls the iron oxidation state in the solid phase (e.g., hematite becomes stable as a product of reaction (6)). The final ash surface composition (at T = 600 °C) is shown in Table 3. Hematite is the main iron mineral in the ash surface in all scenarios but Fe(III) sulfate (Fe2(SO4)3) also forms in DP and HS volcanoes. This is due to the fact that DP and HS magmatic gases are more sulfur rich (see Table 1) resulting in higher sulfuric acid concentrations and progress of the reaction (7) accordingly. However, reactions (5) and (6) result in complete oxidation of Fe(II) to Fe(III), which is the less soluble oxidation state of iron. It is reported that magmatic gas systems exhibit two clear compositional regimes, divided by a compositional discontinuity (Gerlach and Nordlie, 1975; Martin et al., 2006; Roberts et al., 2009). Martin et al. (2006) and later, Hoshyaripour et al. (2012) showed that the oxidation state of pure magmatic gas during mixing with ambient air in the hot core is controlled by the ratio of its H2 and H2S content to the amount of entrained oxygen which is defined as: X mix ¼ ðnðH2 Þ þ nðH2 SÞÞ=ð0:21nðairÞÞ

ð8Þ

HS

where n corresponds to the mole number of each species and the coefficient 0.21 represents the entrained oxygen from ambient air. We note that Martin et al. (2006) used 0.5 and 1.5 as the coefficients for H2 and H2S in Eq. (8), respectively, according to the stoichiometry of their reactions in the gas phase. We omit these coefficients in order to generalize the definition and encompass the possible reactions in the solid phase (e.g., reaction 6). It is reported that when Xmix reaches a critical value (log Xmix = −2.5 to − 3.0), the system becomes completely oxidized (Hoshyaripour et al., 2012). However, as mentioned above, the transition mechanisms in ash-free cases are different from the transition in this study due to presence of ash. Fig. 4 shows the iron oxidation state versus

G. Hoshyaripour et al. / Journal of Volcanology and Geothermal Research 286 (2014) 67–77

0

H2O

0

HO 2

log(moles)

log(moles)

HS 2

H

2

H SO 2

−5

HS 2

−10

H2

H2SO4

DP; Gases

700

800

0

900

−15 600

1000

700

0

Faya

Magn

Hema

900

−10

H

H2SO4

2

−15 600

1000

700

800

0

Faya Ilme

Magn

CP; Fe minerals

Fe2(SO4)3

DP; Fe minerals

900

1000

−15 600

TemperatureoC

700

1000

Faya

Ulvo

−5

−10

900

Hema

log(moles)

log(moles)

log(moles)

−10

800

H2S

Ulvo

−5

700

−5

HS; Gases

800

Magn

Hema

Ilme Ulvo

−15 600

SO

2

4

CP; Gases

−15 600

2

2

−5

−10

HO

SO

SO2

log(moles)

0

71

−5 Ilme Fe2(SO4)3

−10

HS; Fe minerals

800

900

−15 600

1000

TemperatureoC

700

800

900

1000

TemperatureoC

Fig. 3. Gas–ash interaction during cooling with air entrainment for different tectonic settings; upper part: major gas species (vertical lines show the transition from reduced to oxidized state), lower part: Fe-carrying minerals at the ash surface.

Xmix. When the mixture of magmatic gas and ash is exposed to ambient air and undergoes cooling and oxidation, iron in the ash surface starts to oxidize as described above, resulting in a decrease in Fe2+/Fetotal. This continues until log Xmix = − 2.5 to − 3.5 where the transition in the iron oxidation state occurs for all studied volcanic settings. 4. Sensitivity study 4.1. Gas/ash ratio

Fe2 O3 þ SO2 þ SO3 → 2FeSO4 :

As discussed in Section 2.2 the G/A value is used to parameterize the portion of the ash surface available for reaction with the gas species. The G/A is affected by varying the ash content of the plume, ash surface rim thickness and degree of crystallization and therefore, it is necessary to assess the sensitivity of the model to different G/A values. Table 4 shows the Fe-carrying minerals at the ash surface at 600 °C with very high (G/A = 0.01 or 99 wt.%) and very low (G/A = 10 or 1 wt.%) ash content. While varying the G/A does not affect the final iron speciation predicted for the CP volcano case, it modifies significantly the relative proportions of the Fe-bearing minerals that form upon gas–ash interaction in the DP and HS settings. Since the solid phases are nearly similar in all cases, the difference in relative amounts is related to the gas composition as discussed in Section 3.3. According to our results, more hematite can form and less sulfur scavenging occurs at lower G/A ratio

100

DP HS

60

CP 40

Higher ash content shifts the occurrence of the transition point to colder temperatures and thus, can control the oxidation state of the system. Therefore, in an ash- and sulfur-rich volcanic plume, formation of soluble iron in the hot core appears to be more likely. Sulfur scavenging is found to be more important in HS settings. While at low ash content about 66% of the erupted sulfur is scavenged by ash iron, at mid and high ash contents it is 48% and 23%, respectively. However, in CP and DP settings sulfur scavenging by ash iron is negligible. 4.2. Oxidation state Hoshyaripour et al. (2012) demonstrated that the initial oxidation state of the magmatic gas plays a key role in the sulfur speciation in

Volcano type

CP

DP

T (°C)

Mineral

wt.%

Mineral

wt.%

Mineral

wt.%

600 (G/A = 0.1)

Hematite

100

Hematite

100

600 (G/A = 10)

Hematite

100

1000 (FMQ-1)

Fayalite Ilmenite

71 29

600 (FMQ-1)

Hematite

100

Hematite Fe3+ sulfate Fayalite Ilmenite Pyrrhotite Fayalite Magnetite Pyrrhotite Ulvöspinel

Hematite Fe2+ sulfate Hematite Fe3+ sulfate Fayalite Ilmenite Pyrrhotite Fayalite Magnetite Pyrrhotite Ulvöspinel

80 20 4 96 23 20 57 16 26 53 5

20 0 10−6

ð9Þ

Table 4 Fe-carrying minerals at the ash surface (1–100 nm rim) for different gas compositions under different initial conditions: ash-rich mixture (G/A = 0.1), ash-poor mixture (G/A = 10) and reduced magma (FMQ-1).

80

Fe2+/Fetotal

values. This can be explained by considering that iron in the solid phase acts as a sink for oxygen parallel to H2 and H2S in the gas phase. With low ash content, more rapid oxidation of sulfur species results in sulfuric acid formation and production of iron sulfate accordingly. At high ash contents, iron consumes the oxygen and inhibits the rapid oxidation of sulfur in the gas phase. This effect is more pronounced in the HS setting, with the highest sulfur content, where ferrous sulfate (FeSO2), a soluble iron compound, can precipitate:

10−4

10−2

Xmix Fig. 4. Iron oxidation state versus Xmix for different volcanoes.

100

HS

88 12 39 23 37 24 51 14 11

72

G. Hoshyaripour et al. / Journal of Volcanology and Geothermal Research 286 (2014) 67–77

100

the hot core of volcanic plumes. They reported that the transition stage in reduced magmas could be shifted to colder temperatures, thereby increasing the abundance of reduced sulfur species in the system. Thus, we next constrain the effect of reduced conditions on iron speciation in the hot core for different tectonic settings. The initial fO2 of the magmatic gases are reduced by −1 (~FMQ-1) and three reduced scenarios are defined: CP-R, DP-R and HS-R. Fig. 5 shows the gas–ash interaction in these scenarios. The transition is shifted to colder temperatures in CP-R (in comparison with Fig. 3) and no longer occurs in DP-R and HS-R cases. In these scenarios reduced gas species (H2 and H2S) as well as ferrous iron (pyrrhotite, ulvöspinel, fayalite and magnetite) are present at 600 °C. Therefore, under reduced conditions, not only injection of H2S into the atmosphere is more likely (Hoshyaripour et al., 2012) but also preservation of Fe2+ (the more soluble iron oxidation state) is more expected. Table 4 shows the initial (at 1000 °C) and final (at 600 °C) Fecarrying minerals at the ash surface under reduced conditions. Changing the initial oxidation state does not affect iron speciation in CP-R volcanoes and it is completely oxidized to Fe3+ in hematite. However, ash iron in DP-R and HS-R settings remains in the ferrous form. These results are comparable with studies on igneous rocks in different tectonic settings. The reported oxidizing conditions of igneous rocks are not far from those of the FMQ redox buffer (Lindsley, 1991). Nonetheless, there are systematic differences that correlate with tectonic setting. Igneous rocks erupted in arc volcanoes typically record an oxygen fugacity of 1 or more log units above those of the NNO (nickel-nickel oxide) buffer (oxidation of Fe2+ to Fe3+). In contrast, in non-arc settings rocks typically record oxygen fugacities from about those of the FMQ buffer to a log unit or so more reducing than that buffer (some amount of iron remains in Fe2+) (Lindsley, 1991). The presence of iron sulfide (pyrrhotite, Fe0.8S) is also consistent with the stability of this species under reducing conditions (Scaillet et al., 1998). Fig. 6 shows the changes in ferrous to total iron ratio for different scenarios. Reduced conditions postpone the transition from ferrous to ferric iron. In CP, DP and HS scenarios iron is in ferric form even before 700 °C. In CP-R case although the transition is shifted by about 100 °C, iron is completely oxidized at 600 °C too. But in DP-R and HS-R, 82% and 90% of the iron remains ferrous, respectively. Under reduced conditions, the higher initial H2 and H2S content increases the sink for the oxygen and shifts the transition point in the solid phase to colder temperatures. Therefore, at the output of the box model (T = 600 °C) iron remains in ferrous form given the reduced initial conditions.

0

H2O H2

−5

log(moles)

H2S

−10

Fe2+/Fetotal

0 600

800

900

Highly viscous magmas (rhyolitic) are expected to trigger most of the explosive eruptions, which is usually the case in CP setting (Schmincke, 2004). So far we have studied basaltic composition (Table 2). For considering rhyolitic magma in arc volcanoes, first the magmatic gas is cooled to 850 °C because such magmas usually have T b 900 °C (Schmincke, 2004). Then a Si-rich composition (SiO2 75%, FeO 2.5%, TiO2 1.0%, MgO 1.0%) is titrated into the magmatic gas and this mixture is mixed with ambient air. The model predicts that the initial ash surface in equilibrium with this gas phase contains fayalite (65%), magnetite (27%) and ulvöspinel (8%) as Fe-carrying minerals. However, in the final composition at 600 °C hematite is the only iron species at the ash surface as it is the case in the standard scenarios (CP in Table 3). 4.4. Limitations The solid phases, which have been considered so far in this paper, neglected some major species (e.g., Al, Ca, Mn, Na) in order to reduce the complexity of the model as well as computational burdens. However, these species may play an important role during different processes. For instance, where alkali cations are not available, formation of discrete Fe3 +-bearing phases like titanomagnetite, spinels and pyroxenes is more likely (Ayris and Delmelle, 2012). Therefore, existence of alkali

0

H2O

−2

−10

−4 −6 CP−R; Fe minerals

700

900

TemperatureoC

1000

2

−5

−10

H2SO4

HS−R; Gases

700

800

900

−15 600

1000

700

800

0

−2

Ulvo

Magn

−6

700

1000

Faya

Ilme

−4

−8 600

900

Pyrr

Pyrr

DP−R; Fe minerals

800

HO

2

H2

4

Faya

log(moles)

Ulvo

−8 600

SO HS

0 Ilme

1000

2

Faya

Magn

900

4.3. Rhyolitic magmas and arc volcanism

−5

−15 600

1000

800

Fig. 6. Changes in iron oxidation state in the ash surface as function of temperature for different tectonic settings and oxidation states under reduced conditions (FMQ-1).

2

2

0 Hema

700

Temperature oC

DP−R; Gases

700

CP

20

H SO

CP−R; Gases

log(moles)

CP−R

40

H2

H2SO4

−15 600

60

log(moles)

log(moles)

2

HS

SO H2S

SO

DP

DP−R

80

log(moles)

0

HS−R

−2

Ilme Magn

Ulvo

−4 −6 HS−R; Fe minerals

800

900

TemperatureoC

1000

−8 600

700

800

900

1000

TemperatureoC

Fig. 5. Gas–ash interaction during cooling with air entrainment for different tectonic settings under reduced initial conditions (~FMQ-1); upper part: major gas species, lower part: Fecarrying minerals at the ash surface.

G. Hoshyaripour et al. / Journal of Volcanology and Geothermal Research 286 (2014) 67–77

metals in the solid phase can change the mineralogy of the ash iron. In addition, such metals can interact with the gas phase (provide further sinks for oxygen) and affect the iron oxidation reactions. Thus, we explore their impact on the iron speciation on the ash surface. As the number of input solid and gas compounds increases, GASWORKS become more sensitive to the initial compositions because it has to solve a much larger system of thermodynamic equations. The convergence is difficult to achieve specially in case of simultaneous cooling and mixing (Symonds and Reed, 1993). In our calculations we successfully achieved temperatures of 800 °C for the full composition but for colder temperatures GASWORKS does not converge. In order to assess the differences between a calculation including the full composition and the results presented in Section 3, Fig. 7 shows the results of a complete ash composition calculation for CP setting. In Fig. 7a most of the iron is contained in olivine (fayalite) but ilmenite (which we find in the simplified case in Fig. 2) does not precipitate when considering the full composition. This is due to the absence of alkali metals in the simplified scenario that allows formation of ilmenite. Some Fe-carrying pyroxene (e.g., hedenbergite (CaFeSi 2 O 6 ), jadite (Na(Al,Fe)Si2 O 6 ) form too but their amount is negligible. During cooling without air entrainment (Fig. 7b) fayalite remains the main iron species but small amount of spinels (magnetite and ulvöspinel) and pyrrhotite are stable too. This is in good agreement with the results presented in Section 3.2 for the simplified solid phase composition (CP in Fig. 2). Finally, after mixing with ambient air (Table 3), hematite remains the major iron species. During mixing with air, hedenbergite and jadite decompose to separate Ca2 + , Na + and Fe 2 +-bearing phases (e.g., CaSO4, Fe3O4). Again, these results are only valid for the ash surface rim. Inner parts of the particle remain unchanged and most likely carry iron similar to the composition in Fig. 7a. Alkali and alkali-earth metals can provide more reaction sites at the ash surface (Farges et al., 2009) and affect iron–gas interaction. For instance in the DP scenario Ca2 +-bearing phases scavenge sulfur (up to 30% of the erupted sulfur) and form anhydrite (CaSO4). In this case, sulfur scavenging by other minerals is negligible. Another limitation of this study is that GASWORKS assumes 100% crystallization (no melt remains and no glass is produced). Thus, the glass–gas interaction cannot be simulated in this model. Ayris et al. (2013) found that CaSO4 is a major cause of sulfur scavenging through experimental studies on high temperature volcanic glass but the mechanisms proposed is different from the one detailed above (their study focuses mainly on in-conduit processes in absence of ambient air while our approach concentrates on high-T in-plume processes and mixture of volcanic emissions and ambient air). Both the presence of surface defects (glass versus crystal) and network modifiers (alkali

a

b

Other 5%

Quartz 27%

Feldspar 44%

and alkali-earths) are proposed to be important for sulfur to bond onto the ash surface (Farges et al., 2009). Therefore, such processes overcome Fe–S reactions and restrain the formation of iron sulfide and sulfate. Moreover, in volcanic eruptions with low sulfur content in the magmatic gas (CP and DP) formation of iron sulfates may not take place, whereas in sulfur-rich gas (HS), Fe2(SO4)3 can precipitate on the ash in the hot core. 5. Discussion 5.1. The hot core as an oxidizing reactor The initial bulk composition of volcanic ash reflects the composition of the magma from which it derives. The oxidation state of magma, which is conventionally expressed with reference to a known oxygen buffer (e.g., FMQ, NNO), dictates the speciation of iron within it. However, the oxidation ratio measured in freshly-deposited ash may not necessarily reflect that of the source magma due to oxidation processes occurring within the eruption plume (Horwell et al., 2003; Moriizumi et al., 2009). In this study we have shown that high-T equilibrium can significantly change iron mineralogy and speciation in the ash surface. Fig. 8 summarizes the effects of such processes on ash iron. In Fig. 8a, solid black lines show the common buffers (Giggenbach, 1987; Lindsley, 1991): Q IF ðQuartz−Ilmenite−FayaliteÞ : log f O2 ¼ −29520:8=T þ 7:492 ð10Þ

FMQ ðFayalite−Magnetite−QuartzÞ : log f O2 ¼ −25096:3=T þ 8:735

ð11Þ

MH ðMagnetite−HematiteÞ : log f O2 ¼ −25700:6=T þ 14:558

ð12Þ

Fe2−Fe3 ð FeðIIÞ−FeðIIIÞÞ : log f O2 ¼ −25414=T þ 10736

ð13Þ

where T is the absolute temperature in K. It can bee seen that the initial ash oxidation states in all scenarios are not far from FMQ which is consistent with studies on igneous rocks (e.g., Lindsley, 1991). Mixing of the 1000 °C magmatic ash–gas mixture with 25 °C ambient air (simultaneous cooling and oxidation) affects the ash surface compositions differently depending on gas composition (or geochemistry of the tectonic setting) and initial oxidation state. In CP, DP, HS and CP-R scenarios the system moves roughly parallel to FMQ during cooling and mixing until the first sharp increase near Fe2–Fe3 (first transition), which leads the

c

Other 4% Spinel 4%

Feldspar 43%

Quartz 30%

Pyroxene 6% Olivine 18%

73

Olivine 12%

Pyroxene 7%

Feldspar 39%

Other 5% Hematite 11% Spinel < 1%

Olivine 2% Pyroxene 8%

Quartz 34%

Fig. 7. Ash surface composition with complete rock (containing all major species from Table 2); a) initial composition at 1000 °C, b) after cooling down to 800 °C without air entrainment, c) after simultaneous cooling and mixing down to T = 800 °C. Precipitated mineral groups are feldspar: albite (NaAlSi3O8), anhortite (CaAl2Si2O8), sanidine ((K,Na)(Si,Al)4O8); olivine: fayalite (Fe2SiO4), forsterite (Mg2SiO4), tephroite (Mn2SiO4); pyroxene: diopside (CaMgSi2O6), hedenbergite (CaFeSi2O6), jadite (Na(Al,Fe)Si2O6), spinel: magnetite (Fe3O4), ulvöspinel (Fe2TiO4), ilmenite (FeTiO3), other: pyrrhotite (FexS1-x), anhydrite (CaSO4).

74

G. Hoshyaripour et al. / Journal of Volcanology and Geothermal Research 286 (2014) 67–77

b

−5

MH Fe2−Fe3

log fO2

−10 −15 FMQ

QIF

−20 −25 −30 600

700

800

900

Temperature(oC)

1000

c 2

2

0

0

−2

−2

log Xmix=−1.70

log Xmix

0

log Xmix

a

log Xmix=−3.50

−4 −6

Oxidized zone

−6

−8

−8 −10 700

800

900

1000

Temperature(oC)

Transition zone

−4

−10 −12 600

Reduced zone

−12 600

700

800

900

1000

Temperature(oC)

Fig. 8. Changes in the iron oxidation state at the ash surface as a function of temperature during the mixing of the plume with ambient air. Colored lines show: green: CP, blue: DP, red: HS. Dashed lines represent reduced scenarios. a) fO2 of the system and common mineral buffers in solid black lines. b) Xmix of the system and controlling Xmix values in dashed black lines. c) The schematic changes in oxidation state of the iron as a function of temperature and Xmix. Solid line: mixing of plume with ambient air; dotted line: mixing of plume with ambient air under reduced conditions; dash-dotted line: cooling of the system without air entrainment.

systems to the MH buffer. Across a small temperature interval (~50 °C) they follow the MH buffer until the second sharp increase in fO2 occurs (second transition). While in these scenarios the system passes the MH line, which means significant decrease in Fe2+/Fetotal, in DP-R and HS-R scenarios the system remains between FMQ and MH preserving high Fe2+/Fetotal at 600 °C. It also slightly passes the Fe2–Fe3, which explains formation of magnetite in these scenarios (see Fig. 5). Giggenbach (1987) classified the geochemical redox systems based on Fe2–Fe3 (Eq. (13)), which is a non-specific buffer simply involving Fe(II) and Fe(IlI): before reaching this line, the system is under reducing conditions and after that, it becomes oxidized. However, this buffer not only poorly fits the first transition point, but also does not control the oxidation state alone (as it is followed by the second transition). Although in all scenarios there is a span of about 100 °C between the transitions, their occurrence is constrained by neither log fO2 nor T. A much better parameter is the Xmix introduced above which is depicted as a function of temperature in Fig. 8b for all scenarios during mixing of the magmatic ash–gas mixture with 25 °C ambient air. In this figure two transition stages are clear and unambiguously correlated to log Xmix. The first transition point (log Xmix = − 1.70) controls the gas phase oxidation and the second one (log Xmix = − 3.50) modulates the solid phase oxidation. Therefore, Xmix as a temperatureindependent parameter can be used to estimate the system's oxidation state. It can be seen that log Xmix = −1.70 and log Xmix = −3.50 correspond approximately to Fe2–Fe3 and MH buffers in Fig. 8a, respectively. The schematic trend of the changes in the system is shown in Fig. 8c. Three zones (reduced, transition and oxidized zones) are shown separated by transition lines. During air entrainment, the plume cools and oxidizes simultaneously and moves along the solid line. Therefore, the oxidation state of gas phase as well as the ash surface changes. However, it can remain in the reduced zone under reduced initial conditions (dotted line) as well as cooling without air entrainment (dash-dotted line) and maintain the iron in ferrous form down to lower temperatures. This explains the sensitivity of the model to initial conditions. In case of colder initial temperatures (e.g., rhyolitic magma discussed above) the starting point of the solid line moves along the dash-dotted line, which obviously shifts the whole solid line and the transition point toward the left (colder temperatures). On the other hand, reducing initial conditions means shifting the start point of the solid line up and then following the dotted line.

Gislason, 2008; Olgun et al., 2011), iron mineralogy and speciation at the ash surface are lacking. Duggen et al. (2010) report that the experimentally-measured iron release from ash produced by subduction zone and hot spot volcanism is in the ranges 100–400 and 35–107 nmol Fe/g ash, respectively. They concluded that subduction zone volcanism could be more favorable for bio-available iron production. Nevertheless, they excluded the ash from the eruption of Hekla, Iceland, in 2000, which exhibited the highest iron release in deionised water (Frogner et al., 2001; Jones and Gislason, 2008). The strong acidity and elevated fluoride concentration in the eruption plume of Hekla is assumed to have promoted deposition of readily soluble Fe-bearing phases (Ayris and Delmelle, 2012). However, these effects are not unique to this eruption. Ayris and Delmelle (2012) suggested that the origin of the soluble iron in Hekla ash relates to efficient in-plume scavenging of H2S by the ash's Fe-oxide constituents. Studies on the effect of the hot core on sulfur speciation in volcanic plumes showed that H2S could be depleted very fast via high-T oxidation reaction (Hoshyaripour et al., 2012). Absence of H2 S in a few hours old volcanic cloud of Hekla, which is studied using aircraft measurements, confirms this hypothesis (Rose et al., 2006). Therefore, additional processes may have contributed to the formation of soluble iron. As Hekla is located on a mid-ocean ridge but its geochemistry also suggests a hot spot signature (Schmincke, 2004) its magmatic gas composition can be compared to both DP and HS scenarios. In addition, petrological estimates suggest log fO2 = − 10.50 (close to DP and HS scenarios) to −11.00 (closer to DP-R and HS-R) for the basaltic Hekla magma prior to eruption (Moune et al., 2007). The different behavior of the Hekla ash can be considered in agreement with our results for DP-R and HS-R scenarios in which N 80% of the ash iron remains in the ferrous form (more soluble oxidation state). Further in-plume and incloud processes (e.g., dissolution by acid and water condensates) can then produce soluble iron species. On the other hand, in DP and HS scenarios iron is completely oxidized to hematite. In this case, mid and low temperature in-plume and cloud processes could be expected to play the main role in iron mobilization in Hekla ash and the hot core effect is negligible. However, such processes are beyond the scope of this work and more studies are needed in order to understand their effect on ash iron mobilization.

6. Conclusion 5.2. Implications for Hekla eruption Although some studies have been conducted on ash iron release in water (Frogner et al., 2001; Duggen et al., 2007; Jones and

The hot core of the volcanic plume does not solubilize the iron but significantly controls the iron mineralogy and oxidation state, which is the center to studying further in-plume and in-cloud processes. While

G. Hoshyaripour et al. / Journal of Volcanology and Geothermal Research 286 (2014) 67–77

our results may represent an extreme case, they nevertheless suggest that reduced initial magmatic conditions, high ash content and lower initial temperatures in divergent plate and hot spot volcanism appear to be more favorable for soluble iron formation because they maintain iron in a more soluble oxidation state (ferrous or Fe(II)). However, convergent plate volcanoes (subduction zones), which are thought to be more relevant to plinian and sub-plinian eruptions as well as ocean fertilization (Duggen et al., 2007), may produce iron mainly in ferric form (Fe(III) in hematite) at the ash surface. Mineral dust studies show that iron contained in hematite in such particles can be mobilized during atmospheric transport due to interaction with natural and anthropogenic gases and aerosols (Meskhidze et al., 2003; Johnson et al., 2010). By analogy, hematite from CP volcanoes can finally transform to soluble iron species. Nevertheless, chemical and physical conditions in volcanic ash clouds are quite different from dust clouds. Therefore, studying mid and low-T conditions in volcanic plume and in-cloud processes are very important in order to understand the causes of ash iron mobilization. Besides, experimental characterization of the volcanic ash surface (e.g., mineralogy, surface area, porosity, etc.) is a key to determine favorable source and atmospheric conditions for bio-available iron production. Acknowledgments We thank T. Mather for the in depth discussion of potential process in the hot core. Thanks also to M.H. Reed for providing us the source code of GASWORKS as well as helpful comments. This work is supported through the Cluster of Excellence CliSAP (EXC177), University of Hamburg. Appendix A. Thermodynamic equilibrium in high-T gas–ash interaction The thermodynamic equilibrium assumption for reactions of volcanic gases at high-T (T N 600 °C) has been validated (see Symonds et al., 1994, and the references therein) and widely used in earlier studies (e.g., Gerlach and Nordlie, 1975; Symonds et al., 1987; Symonds and Reed, 1993; Gerlach, 2004; Martin et al., 2006, 2009; Roberts et al., 2009). Here, we evaluate the validity of this assumption for gas–ash interaction at high-T. Ideally, one would calculate the relaxation time (characteristic time to reach equilibrium) in order to verify the thermodynamic equilibrium assumption (Seinfeld and Pandis, 2006). Unfortunately the data for those calculations for a high-T heterogeneous system containing minerals and magmatic gases are currently not available. Another approach is to verify the capability of thermodynamic equilibrium-based models in reproducing the observed mineralogical composition of ash samples (similar to the approach used by Symonds et al., 1994, for volcanic gases). Given the scarcity of field measurements and lab experiments on ash surface mineralogy, this approach is not applicable too. Since iron release from the ash in contact with water is an interfacial process (Delmelle et al., 2007; Duggen et al., 2010), it is suggested that iron speciation in the ash's surface layer (with a thickness up to 300 nm) governs its iron release behavior (Gislason et al., 2011; Achterberg et al., 2013). Therefore, as a first order assessment, we focus on oxidation reactions that may affect iron in the ash surface. In other words, we verify if in a given amount of time, oxygen can diffuse into the ash to a certain distance. Then we assume that once the oxygen is there, the oxidation reaction is much faster than the diffusion process (Cooper et al., 1996) and the ash–oxygen interaction most probably favors rapid attainment of thermodynamic equilibrium. It has been shown that the oxidation reaction of Fe2+-bearing amorphous silicates is controlled by three diffusion mechanisms (sometimes acting in parallel) (Moriizumi et al., 2009): l) an inward flux of

75

molecular oxygen (O2) from the surface; 2) a inward flux of oxygen ions (O2−) from the surface; and 3) an outward flux of networkmodifying cations to the surface (e.g., Ca2+; Ayris et al., 2013). For the interaction of iron-carrying minerals with the volcanic and atmospheric gases, we assume that the limiting factor in iron oxidation in such minerals is (just as described for glass above) the inward diffusion of the oxygen into the solid phase (Cooper et al., 1996). If we assume spherical solid particles, transport of oxygen in minerals is given by (Crank, 1975): ∂C ∂2 C 2 ∂C þ ¼D ∂t ∂r 2 r ∂r

! ðA:1Þ

where C is the concentration of the diffusing component and D is the diffusion constant (here for simplicity it has been assumed that D is independent of C). This equation is subject to the following boundary conditions: at r = 0, dC/dr = 0 and at r = a, C is kept at C0, respectively. a is the radius of the ash particle. The non-steady state solution to this equation is (Crank, 1975): ∞   C−C 1 2a X ð−1Þn nπr 2 2 2 sin ¼1þ exp −Dn π t=a C 0 −C 1 n π 1 a

ðA:2Þ

where C1 is the initial concentration in the sphere. To evaluate the assumption of thermodynamic equilibrium in ash–gas interaction (mineral–oxygen), we make the following assumption: if the amount of diffused substance (in this case oxygen which is the left side of the Eq. (A.2)) reaches a “certain value,” then the gas–ash interaction attains thermodynamic equilibrium. To get a sense of this “certain value” we examine the oxidation of fayalite (Fe2SiO4) to magnetite (Fe3O4) as an example: 3Fe2 SiO4 þ O2 → 2Fe3 O4 þ 3SiO2 :

ðA:3Þ

In this reaction 3 mol of fayalite (with molecular weight of 204 g/mol) react with 1 mol of O2 (with molecular weight of 32 g/mol). Thus, each 612 g of fayalite needs only 32 g oxygen for oxidation to magnetite. This means if the weight ratio of oxygen/fayalite reaches 5%, the oxidation reaction can proceed. Therefore, we use 5% as the “certain value” needed for the oxidation reaction. This means only the solid mass affected by the diffusion of oxygen (ash rim thickness in which the oxygen concentration reaches to 5% of that in the surroundings) takes part in the oxidation. Therefore, by fixing the left side of Eq. (A.2) to 0.05, we can calculate the rim thickness (a–r) for given t, D and a. Here we use five diffusion co18 efficients of O in minerals at T N 800 °C proposed in experimental stud−20 and 10−19 m2/s for olivine as a major Fe-carrying mineral ies: 10 (Brady, 1995), 10−17 m2/s for typical aluminosilicates (Brady and Cherniak, 2010; Zhang, 2010), 10−16 m2/s as an upper bound (diffusion in glass; Cooper et al., 1996) and 10−18 m2/s as the average. We also assume 10–50 s diffusion time, as this is about the length of time an ash particle spends in the hot core of an eruption column. This range is calculated using the 1D volcanic plume model Plumeria2 (Mastin, 2007) and shown in Fig. A.1-a. Several simulations have been conducted assuming the following initial conditions: gas fraction: 3%, initial temperature: 1000 °C, exit velocity: 50–250 m/s and vent diameter: 50–200 m. The results of the calculations are shown in Fig. A.1-b. It can be seen that the rim thickness for high diffusion coefficients (D = 10−16 m2/s) in very fine mineral particles (the black line) could reach N 100 nm (more or less the whole mineral) while for small D (10−20, olivine) the rim thickness is almost independent of the size and is approximately 1 nm. Longer residence time (dashed-lines) increases the diffusion depth within the ash. We conclude that the solid mass inherent to the ash rim with the thickness 1–100 nm in minerals and 10–300 nm in glass takes part in the thermodynamic equilibrium interactions with the surrounding gases.

76

G. Hoshyaripour et al. / Journal of Volcanology and Geothermal Research 286 (2014) 67–77

a

b

1200

Zone where 5% concentration is achived

100 Rim thickness [nm]

o

Temperature C

1100 1000 900 800

10

−16

10−17 −18

10 10 10−19

700 −20

10 600 0

10

20 30 Time sec

40

50

1 10

100 Particle size [nm]

1000

Fig. A.1. a) Range of the temperature versus time in eruption plumes with initial temperature of 1000 °C and gas fraction of 3%, based on several simulations with Plumeria2 by varying exit velocity: 50–250 m/s and vent diameter: 50–200. b) Ash rim thickness (in which the diffused oxygen concentration reaches 5% of that of the surrounding environment) as a function of 18 particle size, a, for different O diffusion coefficients, D, at t = 10s (solid lines) and t = 50s (dashed lines); values of D are indicated in the figure.

References Achterberg, E.P., Moore, C.M., Henson, S.A., Steigenberger, S., Stohl, A., Eckhardt, S., Avendano, L.C., Cassidy, M., Hembury, D., Klar, J.K., Lucas, M.I., Macey, A.I., Marsay, C.M., Ryan-Keogh, T.J., 2013. Natural iron fertilization by the Eyjafjallajökull volcanic eruption. Geophys. Res. Lett. 40, 921–926. http://dx.doi.org/10.1002/grl.50221. Ayris, P.M., Delmelle, P., 2012. Volcanic and atmospheric controls on ash iron solubility: a review. Phys. Chem. Earth A/B/C 45–46, 103–112. Ayris, P.M., Lee, A.F., Wilson, K., Kueppers, U., Dingwell, D.B., Delmelle, P., 2013. SO2 sequestration in large volcanic eruptions: high-temperature scavenging by tephra. Geochim. Cosmochim. Acta 110, 58–69. Blundy, J., Cashman, K., Humphreys, M., 2006. Magma heating by decompression-driven crystallization beneath andesite volcanoes. Nature 443, 76–80. Bobrowski, N., von Glasow, R., Aiuppa, A., Inguaggiato, S., Louban, I., Ibrahim, O.W., Platt, U., 2007. Reactive halogen chemistry in volcanic plumes. J. Geophys. Res. D 112, D06311. http://dx.doi.org/10.1029/2006JD007206. Brady, J.B., 1995. Mineral Physics and Crystallography: A Handbook of Physical Constants. In: Ahrens, T.H. (Ed.), Chapter: Diffusion data for silicate minerals, glasses, and liquidsAGU Reference Shelf 2. American Geophysical Union, Washington, DC, pp. 269–290. Brady, J.B., Cherniak, D.J., 2010. Diffusion in minerals: an overview of published experimental diffusion data. Rev. Mineral. Geochem. 72, 899–920. Cooper, R.F., Fanselow, J.B., Poker, D.B., 1996. The mechanism of oxidation of a basaltic glass: chemical diffusion of networkmodifying cations. Geochim. Cosmochim. Acta 60, 3253–3265. Crank, J., 1975. The Mathematics of Diffusion. Oxford University Press. Delmelle, P., Villieras, F., Pelletier, M., 2005. Surface area, porosity and water adsorption properties of fine volcanic ash particles. Bull. Volcanol. 67, 160–169. Delmelle, P., Lambert, M., Dufrene, Y., Gerin, P., Óskarsson, N., 2007. Gas/aerosol/ash interaction in volcanic plumes: new insights from surface analyses of fine ash particles. Earth Planet. Sci. Lett. 259, 159–170. Duggen, S., Croot, P., Schacht, U., Hofmann, L., 2007. Subduction zone volcanic ash can fertilize the surface ocean and stimulate phytoplankton growth: evidence from biogeochemical experiments and satellite data. Geophys. Res. Lett. 34, L01612. Duggen, S., Olgun, N., Croot, P., Hofmann, L., Dietze, H., Teschner, C., 2010. The role of airborne volcanic ash for the surface ocean biogeochemical iron cycle: a review. Biogeosciences 7, 827–844. Farges, F., Keppler, H., Flank, A.M., Lagarde, P., 2009. Sulfur K-edge XANES study of S sorbed onto volcanic ashes. Phys. Conf. Ser. 190, 012177. Frogner, P., Gislason, S.R., Òskarsson, N., 2001. Fertilizing potential of volcanic ash in ocean surface water. Geology 29, 487–490. Gerlach, T.M., 2004. Volcanic sources of tropospheric ozone-depleting trace gases. Geochem. Geophys. Geosyst. 5, Q09007. Gerlach, T.M., Nordlie, B.E., 1975. The C-O-H-S gaseous system, part II: temperature, atomic composition, and molecular equilibria in volcanic gases. Am. J. Sci. 275, 377–394. Giggenbach, W.F., 1987. Redox processes governing the chemistry of fumarolic gas discharges from White Island, New Zealand. Appl. Geochem. 2, 143–161. Gislason, S.R., Hassenkam, T., Nedel, S., Bovet, N., Eiriksdottir, E.S., Alfredsson, H.A., Hem, C.P., Balogh, Z.I., Dideriksen, K., Oskarsson, N., Sigfusson, B., Larsen, G., Stipp, S.L.S., 2011. Characterization of Eyjafjallajökull volcanic ash particles and a protocol for rapid risk assessment. Proc. Natl. Acad. Sci. 108, 7303–7312. http://dx.doi.org/10.1073/pnas.1015053108 (arXiv:http://www.pnas.org/content/early/2011/04/22/1015053108.full.pdf+html). Hamme, R.C., Webley, P.W., Crawford, W.R., Whitney, F.A., DeGrandpre, M.D., Emerson, S.R., Eriksen, C.C., Giesbrecht, K.E., Gower, J.F.R., Kavanaugh, M.T., Panea, M.A., Sabine, C.L., Batten, S.D., Coogan, L.A., Grundle, D.S., Lockwood, D., 2010. Volcanic ash fuels anomalous plankton bloom in subarctic northeast Pacific. Geophys. Res. Lett. 37, L19604. http://dx.doi.org/10.1029/2010GL044629. Heiken, G., Wohletz, K., 1992. Volcanic Ash. University of California Press, London. Horwell, C.J., Fenoglio, I., Ragnarsdottir, K.V., Sparks, R.S.J., Fubini, B., 2003. Surface reactivity of volcanic ash from the eruption of Soufrire Hills volcano, Montserrat, West

Indies with implications for health hazards. Environ. Res. 93, 202–215. http://dx. doi.org/10.1016/S0013-9351(03)00044-6. Hoshyaripour, G., Hort, M., Langmann, B., 2012. How does the hot core of a volcanic plume control the sulfur speciation in volcanic emission? Geochem. Geophys. Geosyst. 13, 100–115. Johnson, M.S., Meskhidze, N., Solmon, F., Gass, S., Chuang, P.Y., Gaiero, D.M., Yantosca, R.M., Wu, S., Wang, Y., Carouge, C., 2010. Modeling dust and soluble iron deposition to the South Atlantic Ocean. J. Geophys. Res. D 115, D15202. http://dx.doi.org/10.1029/ 2009JD013311. Jones, M.T., Gislason, S.R., 2008. Rapid releases of metal salts and nutrients following the deposition of volcanic ash into aqueous environments. Geochim. Cosmochim. Acta 72, 3661–3680. Kishima, N., 1989. A thermodynamic study on the pyrite-pyrrhotite-magnetite-water system at 300-500 °C with relevance to the fugacity/concentration quotient of aqueous H2S. Geochim. Cosmochim. Acta 53, 2143–2155. Langmann, B., Zaksek, K., Hort, M., Duggen, S., 2010. Volcanic ash as fertiliser for the surface ocean. Atmos. Chem. Phys. 10, 3891–3899. Lindsley, D.H., 1991. Oxide minerals: petrologic and magnetic significance. Mineralogical Society of America Reviews in Mineralogy volume 25. Martin, R.S., Mather, T.A., Pyle, D.M., 2006. High-temperature mixtures of magmatic and atmospheric gases. Geochem. Geophys. Geosyst. 7, Q04006. Martin, R.S., Mather, T.A., T.J.R., Pyle, D.M., 2009. The implications of H2S and H2 kinetic stability in high-T mixtures of magmatic and atmospheric gases for the production of oxidized trace species (e.g., BrO and NOx). Chem. Geol. 263, 143–150. Mastin, L.G., 2007. A user-friendly one-dimensional model for wet volcanic plumes. Geochem. Geophys. Geosyst. 8, Q03014. http://dx.doi.org/10.1029/2006GC001455. Meskhidze, N., Chameides, W.L., Nenes, A., Chen, G., 2003. Iron mobilization in mineral dust: can anthropogenic SO2 emissions affect ocean productivity? Geophys. Res. Lett. 30. http://dx.doi.org/10.1029/2003GL018035. Mills, O., Rose, W., 2010. Shape and surface area measurements using scanning electron microscope stereo-pair images of volcanic ash particles. Geosphere 6, 805–811. Moriizumi, M., Nakashima, S., Okumura, S., Yamanoi, Y., 2009. Color-change processes of a plinian pumice and experimental constraints of color-change kinetics in air of an obsidian. Bull. Volcanol. 71, 1–13. Moune, S., Sigmarsson, O., Thordarson, T., Gauthier, P.J., 2007. Recent volatile evolution in the magmatic system of Hekla volcano, Iceland. Earth Planet. Sci. Lett. 255, 373–389. Moune, S., Holtz, F., Botcharnikov, R.B., 2009. Sulphur solubility in andesitic to basaltic melts: implications for Hekla volcano. Contrib. Mineral. Petrol. 157, 691–707. Olgun, N., Duggen, S., Croot, P.L., Delmelle, P., Dietze, H., Schacht, U., Oskarsson, N., Siebe, C., Auer, A., Garbe-Schonberg, D., 2011. Surface ocean iron fertilization: the role of airborne volcanic ash from subduction zone and hot spot volcanoes and related iron fluxes into the Pacific Ocean. Glob. Biogeochem. Cycles 25, GB4001. Óskarsson, N., 1980. The interaction between volcanic gases and tephra: fluorine adhering to tephra of the 1970 Hekla eruption. J. Volcanol. Geotherm. Res. 8, 251–266. Roberts, T.J., Braban, C.F., Martin, R.S., Oppenheimer, C., Adams, J.W., Cox, R.A., Jones, R.L., Griffiths, P.T., 2009. Modelling reactive halogen formation and ozone depletion in volcanic plumes. Chem. Geol. 263, 151–163. http://dx.doi.org/10.1016/j.chemgeo.2008. 11.012. Rose, W.I., 1977. Scavenging of volcanic aerosol by ash: atmospheric and volcanologic implications. Geology 5, 621–624. Rose, W.I., Durant, A.J., 2009. Total grain size distribution of explosive volcanic eruptions. J. Volcanol. Geotherm. Res. 186, 32–39. Rose, W.I., Millard, G.A., Mather, T.A., Hunton, D.E., Anderson, B., Oppenheimer, C., Thornton, B.F., Gerlach, T.M., Viggiano, A.A., Kondo, Y., Miller, T.M., Ballenthin, J.O., 2006. Atmospheric chemistry of a 33–34 hour old volcanic cloud from Hekla Volcano (Iceland): insights from direct sampling and the application of chemical box modeling. J. Geophys. Res. 111, D20.

G. Hoshyaripour et al. / Journal of Volcanology and Geothermal Research 286 (2014) 67–77 Sarmiento, J.L., 1993. Atmospheric CO2 stalled. Nature 365, 697–698. Scaillet, B., Clemente, B., Evans, B., Pichavant, M., 1998. Redox control of sulfur degassing in silicic magmas. J. Geophys. Res. 103, 23937–23949. Schmincke, H.U., 2004. Volcanism. Springer-Verlag, Berlin Heidelberg. Seinfeld, J.H., Pandis, S.N., 2006. Atmospheric chemistry and physics: from air pollution to climate change. John Wiley and Sons, New York. Sparks, R.S.J., Bursik, M.I., Carey, S.N., Gilbert, J.S., Glaze, L.S., Siggurdsson, H., Woods, A.W., 1997. Volcanic Plumes. John Wiley and Sons, New York. Symonds, R.B., Reed, M.H., 1993. Calculation of multicomponent chemical equilibria in gas-solid-liquid systems; calculation methods, thermochemical data, and applications to studies of high-temperature volcanic gases with examples from Mount St. Helens. Am. J. Sci. 293, 758–864.

77

Symonds, R.B., Rose, W.I., Reed, M.H., Lichte, F.E., Finnegan, D.L., 1987. Volatilization, transport and sublimation of metallic and non-metallic elements in high temperature gases at Merapi Volcano, Indonesia. Geochim. Cosmochim. Acta 51, 2083–2101. Symonds, R.B., Rose, W.I., Bluth, G.J.S., Gerlach, T.M., 1994. chapter Volcanic gas studies: methods, results and applicationsVolatiles in Magma. Reviews in Mineralogy vol. 30. American Mineralogical Society, pp. 1–66. von der Heyden, B.P., Roychoudhury, A.N., Mtshali, T.N., Tyliszczak, T., Myneni, S.C.B., 2012. Chemically and geographically distinct solid-phase iron pools in the southern ocean. Science 338 (6111), 1199–1201. Watson, A.J., 1997. Volcanic Fe, CO2, ocean productivity and climate. Nature 358, 587–588. Zhang, X., 2010. Diffusion in minerals and melts: theoretical background. Rev. Mineral. Geochem. 72, 5–59.