Silica from Steam Condensate Alteration at Tikitere, New Zealand

Silica residue is the main product of the alteration of a rhyolitic, tuff-rich, volcanic substrate by acid sulfate condensate formed, in a major steam field at Tikitere, near Rotorua. Here steam and accompanying H2S discharge in an area characterized by large standing pools of acid water and mud pots. As in silica sinters, that precipitate from hot, near-neutral, alkali chloride waters, the first formed silica phase in the residue consists of noncrystalline opal-A that subsequently crystallizes to paracrystalline opal-CT. Residue and sinter can superficially resemble each other and there is a need to clearly differentiate both types of deposit in order to correctly interpret the hydrology or paleohydrology of a geothermal field. Both the opal-AN and opal-CT in the Tikitere residues are more disordered than those in typical sinters. Tikitere residue also lacks mesoscopicallyrecognizable, biologically mediated textures, vascular plant remains, and unaltered tuffaceous clasts, commonly present in silica sinters. However, once crystallized to opal-CT lepispheres (1-3 μm across) and, subsequently, to quartz, both residue and sinter are less readily distinguished as earlier textures become overprinted. INTRODUCTION Tikitere is a hot water geothermal field in which a large amount of steam discharges. It is one of the most active thermal areas in New Zealand. It is remarkable in that alkali chloride waters are not known to have ever discharged at the surface in the most visibly active portion of the field. The surface alteration products are soley those resulting from acid-sulfate thermal activity, including the action of acidified steam condensate, gases and steam upon the formerly pristine country rock. One of the distinctive effects of the acid-sulfate surface activity is the accumulation of silica residues that result from the surface and near-surface leaching of silicate rock by sulfuric acid, derived from oxidation of ascending H2S (White et al., 1956). Silica residues contrast with silica sinters deposited from near-neutral, alkali chloride waters and with which they have sometimes been confused. In both types of deposits the first-formed silica phase is opaline silica and Herdianita et al. (2000a) have stressed the importance of clearly differentiating residue from sinter if either is to be used to interpret hydrology or paleohydrology. Although a large body of literature exists as to the nature and origins of silica sinters, no systematic characterization of any residue deposit has been undertaken. The results presented here include a mineralogical and textural examination of three typical residue samples collected from the surface of the Tikitere field. They illustrate how alike both residue and sinter can be, and the need for caution in appraising any geothermal-derived silica deposit in the field. SETTING The Tikitere geothermal field is situated within the Rotorua Volcanic Centre of New Zealand's Taupo Volcanic Zone (TVZ), 18 km east of Rotorua City (Fig. 1). The country rock consists of late Pleistocene (~60,000 BP), largely rhyolitic, Rotoiti Breccia. Minor amounts of Holocene, basaltic Rotokawau Ash may be present. Thermal activity occurs over an area 3.2 x 2.4 km. Two adjacent zones at the southern extent of the field, due north of the Rotorua-Whakatane Road, comprise the Tikitere tourist area known as Hell's Gate. The output of H2S from this southern part of the field was regarded by Grange (1937) as probably the highest in the RotoruaTaupo area. The confusion of alkali-chloride-derived sinter with silica residue is seen in Grange's 1937 description of the Tikitere field. He reported the presence of encrusting masses of white sinter throughout much of the field and cited their presence as evidence that alkali-chloride waters had once discharged at the surface there in the past. All such deposits within the main body of the field are now interpreted as silica residue, consistent with the presence of acid sulfate waters and the now-known absence of any discharge of alkali-chloride waters at the surface, both at present and in the past (Glover, 1974). Four large acid-sulfate pools occupy the valley floor, in the southern portion of the Hell's Gate area (Fig.1). Small deposits of both sulfur and pyrite have formed on different areas of the valley's surface and Grange (1937, p.87) noted that "white sinter to a small extent encrusts much of the area." The northern portion of Hell's Gate is dominated by three, large, acid, shallow pools with diameters that vary up to 100 m across, and about which Grange (1937, p.87) observed "some big blocks of white sinter." The overflow from these pools feeds Mariwai Creek. All springs discharging in the field today are fairly close to present groundwater level but the water supply is probably small. Fig. 1. Locality map, Tikitere steam field, modified slightly after Grange (1937, p.88), showing locations from which silica residue was collected for the present study. Insert: location of Tikitere within the Taupo Volcanic Zone (TVZ). SAMPLES Environmental concerns restricted sampling and only three specimens were collected. Each represented a different morphological and textural aspect of the upstanding silica blocks described by Grange (1937). All such masses proved to be sub-recent erosional remnants such that no examples of residue forming today were collected. Sample numbers prefixed AU are those of the University of Auckland, Department of Geology, petrology collection. The block-like mounds of residue are typically 400-1200 mm across and from 250 to 750 mm high (Fig. 2a,b). In many instances they are sited in and around the edges of acid pools where they outcrop amongst moderately sorted, medium to coarse silica gravel. Elsewhere, they protrude above masses of white, kaolin-rich clay. In color, the residue mounds range from a bleached white, through a pale cream to light grey. Most surfaces are heavily corroded and/or abraded, but in protected nooks and crannies throughout the area, porous grey crusts and irregular lumps of weakly-cemented, friable, fine opaline sand encrust brown, clay-rich altered tuff (e.g. AU49877). About the edge of ponds in the northern end of the field, residue mounds of low relief have smooth, if somewhat irregular abraded surfaces (Fig. 2a). Higher-standing coralloidal-like mounds in both the northern and southern portions of the field typically display extremely rough, gnarled surfaces (Fig. 2b). In many respects this contorted appearance resembles that of sub-aerial coralline algal masses subject to differential solution and endolithic microbial attack on tropical atolls. Spiculose, knobby masses, 2-10 mm high and 2-5 mm across, protrude from the surface and coalesce to form irregular ridges and stony prominences. Broken knobs show crudely concentric structure. Elsewhere, undulose ridges, 2-5 mm high, wander erratically over the surface. The depressions between knobs and/or ridges are commonly lined with a porous layer of silica-cemented, well sorted, medium sand-sized clasts. Broken masses of residue commonly show a clasts of massive to finely laminated, pale incarnadine to sallow, opaline silica. Color differences highlight the laminae, that are generally <<1 mm thick and display small sedimentary structures consistent with a water-laid origin (Fig 2c). Typically these laminae truncate abruptly against the opaline matrix and indicate the clasts are derived from an earlier-formed deposit. Clusters of diatom skeletons are also common as are those of other aquatic microorganisms and confirm the residue originated in an environment that contained standing water. The general impression in the field is that all blocks and silica masses have been, and continue to be, subject to cycles of silica deposition, erosion and cementation on both macroscopic and microscopic chemical scales; presently destructional processes exceed constructional among the upstanding residue masses. MINERALOGY Determinative techniques used were those of Herdianita et al. (2000b) so that direct comparison could be made of the residue results with those obtained from sinters (e.g. Herdianita et al., 2000a, Sannazzaro et al., 2001). The x-ray powder diffraction patterns of three silica phases were found in the three samples analyzed (Fig. 3): opal-A, opal-CT and quartz. Other diffraction lines present were those of a feldspar and possibly alunogen. Opal-A was the principal phase in the grey, porous friable Fig. 3. Typical x-ray powder diffraction and scattering traces of silica residue, Tikitere, showing variation in silica species and other minerals. Broad scattering band with intense fine structure and maximum at ~4 Å arises from opal AN. Q = quartz; CT = opal-CT; al = alunogen: (a) finely-laminated, massive opaline residue, AU49876; (b) pale-grey, gnarly residue, AU49875; (c) weaklycemented, friable, fine opaline sand, AU49877. opaline crust (AU49877). Superimposed upon the opal-A x-ray scattering broadband were the weakly developed principal diffraction lines of quartz, and ill-defined broad diffraction lines of opal-CT at ~4.07 and ~2.5 Å, respectively. The width at half maximum intensity (FWHM) of the opal-A scattering band is 1.52 Å ∆d (9.30° ∆2θ). Such a value is exceptional among opal-A silicas. It is found only at the higher end of the range reported by Teece (2000) from silicas deposited from highly acid waters at Rotokawa, ∆d = 1.29-1.52 Å (∆2θ = 7.40-9.30°). FWHM values among low temperature (sinter apron), alkali chloride-sourced sinters are typically lower: ∆d = 0.951.25 Å (∆2θ = 5.0-7.0°). Higher values have been found among only very young sinters from the Wairakei discharge drain (∆2θ = 7.20-7.80°), the high temperature facies sinter of the Crow's Nest geyser vent (∆2θ = 7.10°), and surface samples of porous and friable sinters that have been subjected to weathering (∆2θ = 7.00-8.20°) (Herdianita, 1996; Holland, 2000; Pastars, 2000). The fine structure of the Tikitere scattering band closely resembles that of opal-AN, rather than opal-AG, the latter of which is the typical, first-deposited phase of low to moderate temperature alkali-chloride sinters (cf. Flörke et al., 1991; Smith, 1998). In both the gnarly residue (AU49875) and the finely laminated, massive opaline residue (AU49876), paracrystalline opal-CT is the dominant silica phase. Weak quartz lines are present in the diffraction scans of both deposits, and in AU49875 a subdued scattering broadband, with an FWHM of ~1.35 Å ∆d (~8.00° ∆2θ), similar to that in the porous crust (AU49877) underlies the opal-CT pattern. The maxima of the principal opal-CT line ranges from 4.04 to 4.08 Å and its FWHM from 0.33 to 0.42 Å ∆d (1.70-2.25° ∆2θ). The FWHM values are considerably larger than those found among cristobaliticopals reported by Sannazzaro et al. (2001) from recrystallized, Late Quaternary, low-temperature facies sinter from Umukuri. In samples in which opal-CT was dominant, these authors reported FWHM values of between 0.54 and 0.69° ∆2θ. The marked differences between Tikitere and Umukuri result, in large part, from the ill-defined nature of the principal opal-CT diffraction line in the Tikitere samples. This line is a composite. It consists of a main reflection at ~4 Å, arising from small, coherently-scattering domains that mimic the arrangement of atoms in cristobalite (Smith, 1998), and a prominent satellite at ~4.3, that arises from tridymite interlayers within the structure. The ratio of the relative intensities of the cristobalitic and tridymitic components of this composite line in the Tikitere residues (tridymite/cristobalite ~0.6) is considerably greater than that of Umukuri opal-CT (~0.3). Presumably, the difference in this ratio implies a greater proportion of the tridymitic lattice component in the Tikitere opal-CT samples. However, it also means that measurement of FWHM in the Tikitere samples occurs in that part of the band that is broadened by overlap of the tridymite and cristobalite diffraction responses (Fig. 3). In the Umukuri sinter the measurement is made within that part of the composite line ascribed to diffraction from cristobalitic stacking alone. Hence the two sets of measurements are not directly comparable. Nonetheless the general structure, shape and line width of the Tikitere samples 0 100 200 300 10 15 20 25 30 35 40 degrees 2θ co u n ts (c) Q Q