Silicification as the result of incongruent rock dissolution

Silicification as the result of incongruent rock dissolution

S.H.J.M. van den Boorn , M.J. van Bergen and P.Z. Vroon

a b

Utrecht University, Faculty of Earth Sciences, Utrecht

Free University, Faculty of Earth and Life Sciences, Amsterdam

*E-mail: boorn@geo.uu.nl

RESULTS

References

[3] Walther et al., 1977,

[4] Gunnarsson and Arnorsson, GCA,

Acknowledgement: Financial support from the Dr.

Molengraaff Fonds is greatly appreciated [1] Delmelle et al., 2000, JVGR, , 31-53

[2] Van Hinsberg, 2001, unpubl. MSc thesis, Utrecht

[5] Takano et al., 2004, JVGR, , 299-329

[6] Van den Boorn et al., 2007, Geology, : 939-942 [7] Van den Boorn et al., 2006, JAAS, 734-741 [8] Georg et al. 2006, Chem. Geol., , 95-104

97 135 35 21: 235 Am. J. Sci, , 1315-1351 , 2295-2307 277 64 SILICON ISOTOPES ! ! ! d d 30 30 30 30 30

Si of altered rocks largely fall in the range of unaltered rocks and are close to 0‰ (Fig. 4), supporting the view that silica is a residue and isotope fractionation is minimal

Close to 0‰ values for sediments imply positive Si for the lake water since silica deposition is known to discriminate against Si [e.g. 6]

This inferred Si enrichment might be the result of continuous precipitation of

Si-depleted opal

INTRODUCTION

We studied silicification mechanisms during acid water-rock interaction to unravel the sources and sinks of silica in active hydrothermal environments. We obtained geochemical signatures including silicon isotopes from altered/silicified volcanic rocks, their unaltered parents, chemical precipitates and the strongly acidic waters from Kawah Ijen volcano (Java).

SETTING + SAMPLES

Kawah Ijen is an active volcano with a large crater lake (3x10 m ) (Fig 1a). A low pH (<0.5), strong mineralization (TDS~105 g/l) and high

contents (Table 1) reflect the continuous input of volcanic gasses and fluids at the bottom (e.g. [1], . Seepage springs on the outer flank form the acidic Banyupahit stream.

The hyperacidic waters interact with andesitic and basaltic lavas and

pyroclastics, leaving behind. Samples [this

study and [2]) come from the crater area and from the flanks where volcanics are in contact with seepage water.

7 3

SO , Cl and F Fig. 1b)

variably altered and silicified rocks

4

Fig. 1a: The Kawah Ijen crater lake is ~800m in

diameter and is filled with hyperacidic water of pH<0.5 and T ~40ºC!

Fig 1b: Schematic cross section of the Kawah

Ijen hydrothermal-volcanic complex (after [1]).

CONCLUSIONS

!

! !

!

Silicification of volcanic rocks by hyperacidic waters occurs by incongruent rock dissolution as shown by two independent approaches

Silicon isotope fractionation is limited during silicification reactions

The rate and mode of plagioclase dissolution is an important controlling factor in generating REE patterns of altered rocks

Si isotope measurement procedures for highly altered rocks should take the possible presence of newly formed S-bearing minerals into account

MAJOR AND TRACE ELEMENT CHEMISTRY

! All major elements are leached from the rock, including silica, assuming Nb as immobile (Fig. 2a, cf. [2])

! ! !

!

!

Silica is the least mobile oxide resulting in silicified residues

Si (153 ppm SiO ) is depleted relative to Mg in lake water (Fig. 2b)

Opal precipitation [1] can only partly explain this depletion; congruent silica dissolution would imply unrealistically high concentrations in input waters (Fig. 3)

Most other elements including REE show close-to-congruent dissolution behaviour (Fig. 2b, cf. [5])

Strong enrichments in volatile elements (e.g. Pb and S) are due to gaseous inputs; depletions in Ti and Ba reflect deposition of barite and anatase [1]

2

Fig. 5: Increase in with SO

content in solution expressed as SO /Si percentage (in ppm)

4 4

d30_Si

Fig 4: Si isotope ratios (relative to NBS28) of

unaltered and altered rocks and chemical sediments. Isotope measurements were corrected based on SO contents in the samples (see Fig. 5)

4

Fig. 2a: Relative losses of major oxides from altered

rocks (assuming Nb to be immobile). Bars represent 14 individual samples

Fig. 2b: Depletions and enrichments in sediments,

lake and stream waters relative to unaltered rocks

Fig. 4: Normalized REE patterns of altered

rocks and sediments

!

!

REE patterns of altered rocks (norm. to mean unaltered rock, Fig. 4) show variable Eu anomalies

The Eu anomalies seem to be controlled by the behaviour of plagioclase during the acidic attack which depends on texture and composition

Fig. 3: Solubilities of silica polymorphs in

distilled water [3, 4] ANALYTICAL ! ! ! !

Elevated sulphate contents in analytical solutions generated offsets in our MC-ICPMS Si isotope measurements following the procedure of [7]

SO -doped standards were used to correct for this effect assuming a linear correlation between SO and Si and a maximum offset of +1.4‰ (Fig. 5)

This correction decreases the accuracy of our data (Fig. 4)

The observed SO interference with Si isotope analysis is in contrast with previous work [8]