Retrograde fluid infiltration is an important process associated with base metal mineralisation in metamorphic rocks. In the SW Highlands of Scotland, retrograde infiltration resulted in the propagation of dolomitisation reaction fronts (or sides) and oxygen isotope fronts into and along calcite-rich marble layers adjacent to faults cutting the Loch Tay Limestone (Fein et al.,1994). Dolomite is isotopically heavy (+25 to + 28 ‰) compared with host marbles (+15 to +18 ‰). An initial ion probe study (Guest et al., 2000) of calcite spatially associated with dolomitisation has revealed extreme heterogeneity in a succession of grain boundary alteration domains and cements, with δ18O values ranging from + 7 to +33 ‰. These exciting results suggest that metamorphic basement rocks were infiltrated by a succession of surface-derived basinal and meteoric fluids, probably at relatively low temperature (100 + 50 oC), and not by greenschist facies metamorphic fluids as previously thought.
It was originally intended to include analysis of Fe-dolomites in the above study, but given the unexpected range in calcite values, time did not allow this. Dolomites show a wide range of textural types and Fe-content, and in view of the calcite results above it seems likely that they are heterogeneous isotopically as well. In this new proposal, analyses will be concentrated on replacive dolomite zones and veins, and on ankerites growing in pelitic schist. All these dolomites are strongly zoned in Fe content, and we will investigate whether δ18O correlates either with Fe content or position relative to calcite cement zones. These data will help constrain mechanisms of propagation of the reaction and isotopic fronts within the marble layers, direction and mechanism of fluid flow, fluid sources, and the relative timescale of infiltration of isotopically variable fluids.
These results will be important in the wider context of models of fluid flow in basement rocks under retrograde conditions, the mechanism/s of permeability generation, pathways of fluid flow, and limits on the composition of the migrating fluid. Similar mechanisms appear to operate during the formation of ore deposits and the generation of secondary porosity in petroleum reservoirs.
This proposal is a continuation of ion probe project IMP/156/1099, results of which are summarised below. It was originally intended to analyse both calcite and Fe-dolomite in the course of this project. However, unexpectedly large variations in δ18O of calcite were obtained, indicating a surprisingly varied fluid infiltration history. As a result the time allocated to stable isotope analysis was not sufficient both to characterise these variations, and undertake the more complex calibrations required to analyse dolomite with variable Fe contents. This proposal aims to complete the project as originally proposed, with some modifications in the light of the calcite data already obtained. Rather than repeat the project proposal as approved by the Steering Committee only 6 months ago, we concentrate here on briefly presenting the results of the study so far, while re-emphasising the nature of the dolomite targets.
In the Dalradian Loch Tay Limestone of the SW Highlands of Scotland both dolomitisation reaction fronts and oxygen isotopic fronts have propagated the same distance (typically several metres) into and along the calcite marble layers in wall rock away from fault-controlled veins during retrograde fluid infiltration (Fein et al., 1994). Dolomites have unusually high δ18O values, up to +28‰ compared with +15-17 ‰ in the metamorphic calcite marbles. Similar isotopically heavy dolomites and ankerites are found in veins and as isolated rhombs in pelites over a wide area of the SW Highlands (Fein et al., 1994). The coincidence of the isotopic and reaction fronts in marbles suggests that the replacement of calcite by dolomite has facilitated oxygen isotopic exchange between the fluid and the host rock.
Infiltration of dolomitising fluids has previously been assumed to have occurred under greenschist facies conditions (Graham et al., 1983; Fein et al., 1994), largely on the basis of fluid inclusion data from quartz-dolomite veins. However, new stable isotope work on the quartz which hosts the inclusions has shown that it is not in isotopic equilibrium with coexisting dolomite, and it is unlikely that the fluid inclusion temperatures are related to the dolomitisation event. Our current working hypothesis is that the dolomite was precipitated by basinal brines, possibly of Permo-Triassic age, at relatively low temperatures (100 + 50 °C). Our new ion probe studies show that a huge range in isotopic composition (from +7 to +33 ‰) is present in calcite close to the reaction front, with compositions being closely related to a series of textural domains. Table 1 summarises textural and ion probe data from domains which are related to dolomitisation. Additional data from sample 3 (Fig. 1) has been presented elsewhere (Guest et al., 2000).
Table 1 Summary of textural domains and ion probe results
Textural domain [samples where observed - Fig. 1] |
Description/textural interpretation (CL characteristics) |
Oxygen isotope (ion probe) characteristics (δ18O SMOW) |
1 [1, 2] |
Dark twinned grains inherited from the metamorphic marble |
+15 to +18; mean +17.5 (n=12) |
2 [1] |
Bright irregular grain boundary zones |
+19 to +24; mean +21.2 (n=8) |
3 [1, 2, 3] |
Mottled "recrystallised" grains |
+15 to +24; mean +20.2 (n=37) |
4a [2] |
Euhedral dolomite, zoned to ankeritic compositions |
No ion probe data. +25 to +28 from whole rock data |
4b [2] |
Dark Fe-rich zones along domain 3 grain boundaries and associated with domain 4a |
+20 to +7, decreasing outwards |
5 [2] |
Discontinuous bright zones lining relict porosity |
+27 to +33 |
Our preliminary interpretation of this data is that:
Isotopically heavy dolomites also occur in veins in the Carboniferous in N. Ireland (Evans et al., 1998), and probably precipitated from Permo-Triassic basinal brines. We believe that the widespread growth of dolomite and ankerite in marbles and pelites in the SW Highlands indicates that metamorphic basement rocks were highly permeable during infiltration of basinal brines derived from overlying basins. Although fractures were important in controlling this permeability, it appears that the dolomitisation reaction played a key role in increasing porosity and permeability, and by inference in focussing fluid flow. This has considerable implications for the formation of carbonate-hosted ore deposits in basement rocks.
The marble layers alternate with layers of pelitic schist which contain dolomite with +24 to +28 ‰ δ18O. In calcite marble layers, beyond the isotopic/dolomite reaction fronts, dolomitisation of the marble layer can be observed adjacent to the pelitic schist layer. Whole rock isotopic analysis reveals an oxygen isotopic anomaly associated with this dolomitisation similar to that observed at dolomite reaction fronts propagating into and along layers, implying that fluid flow may have been controlled by pelite layers as well as the fault vein. Dolomite textures within the pelites are extremely complex (Guest et al., 1999), suggesting repeated dissolution and growth of dolomite.
In view of the low temperatures we now infer for dolomitisation, it is unlikely that diffusion through the solid phases played any significant role in isotopic redistribution (cf. Graham et al., 1998). We interpret the wide range in isotopic composition of domain 4b to reflect progressive reaction between incoming fluid and host carbonates. The fact that dolomite growth was accompanied or preceded by calcite dissolution, coupled with the presence of domains 2 and 3 which are intermediate in composition between host marble and dolomite, suggests that similar compositional variations may occur within dolomite. If dissolution of calcite is the principal way in which the fluid changes isotopic composition, then the composition of the dolomite is a measure of the extent to which the fluid has previously dissolved calcite. The dolomite is clearly heterogeneous chemically (Fig. 2) - the question is, do these chemical variations correlate with isotopic variations?
Fein et al. (1994) suggested that the isotopic and dolomitisation fronts propagated as a result of variable fluid fluxes away from the fault and along the marble layers (i.e. advection across the front). It is more likely that dolomitisation fronts formed on the sides of the fault-controlled pathway, with the unaltered marbles remaining essentially impermeable. Fronts would propagate largely because of the dissolution of calcite by the dolomitising fluid and hence increased permeability and fluid focussing. This could be an extremely important mechanism for generating and enlarging flow channels in initially impermeable meatmorphic rocks.
Aims of this project are as follows:
As detailed above, the samples have already been characterised thoroughly in terms of SEM and CL textures, and oxygen isotope composition of calcite. We would seek to analyse a few more calcite spots in order to check for extremes of composition in domain 4b, and calcites included in dolomite. Dolomite can be analysed with a lower precision than calcite, depending on Fe-content, due to matrix effects, but variations greater than 1 or 2 ‰ should be detectable using available Ca-Fe-Mg standards (Eiler et al., 1997), although further calibration is required for "normal" terrestrial compositions.
Stable isotope analysis of the carbonates will be made using the high voltage offset technique described by Valley et al., 1997a. The spot size will be about 20µm for oxygen. Sample analyses will be interspersed with analysis of carbonate standards.
Locations of samples 1, 2 and 3 are illustrated in Figure 1
Sample 1: Partially dolomitised zone adjacent to pelite layer. Domains 1, 2 and 3 calcites have already been analysed in this sample (Guest et al., 2000), and we will extend the analyses to small zoned dolomites.
Sample 2: Marble layer 0.1m upstream of the front. This is one of the main samples analysed previously. The sample is 90% carbonate; the remaining 10% consists of quartz, pyrite, phengite, albite, apatite, and oxide phases. The carbonate proportion is 60% calcite and 40% dolomite. The dolomites exhibit oscillatory zoning due to changes in Mg/Fe ratio with an overall trend to more Fe-rich compositions at the rim (Fig.2). The centres of dolomitised areas often show high-Fe patches with identical Fe compositions to dolomite at the dolomite-calcite interface (Figure 2). Some additional calcite analyses will be undertaken
Sample 4 (sample 3 from the previous study contains no dolomite): Dolomite vein containing complex zoning patterns within marble 1 m downstream of front. Analyses will test for heterogeneity within the vein and possible alteration of calcite wall rocks.
Sample 5: Pelite layer from within the marble sequence. Dolomite is present in triangular zones within crenulations and as trellis-like grains apparently forcing apart mica cleavages. Complex compositional variations occur. Most zones are too small to be resolved with the ion probe, Analyses of the dolomites should establish compositional heterogeneity and how the values compare with dolomite in adjacent marble layers (Guest et al., 1999). Only a few analyses are likely to be possible due to fine grain size
Sample 6: Marble layer upstream of the front. The sample is 90% dolomite and 10% albite, potassium feldspar, quartz, phengite, apatite and oxide phases. The dolomite exhibits oscillatory zoning due to changes in Mg/Fe ratio.This suggests that the flow of fluid may have been pulsed.
The number of samples that can be characterised will depend on the isotopic homogeneity of dolomite
There is clear evidence that the dolomitising reaction created porosity and highly permeable pathways within metamorphic marbles. These pathways were subsequently exploited by other fluids. The presence of ankerite in schists over large parts of the SW Highlands suggests that fluid infiltration was regional, and that pelitic rocks were quite permeable with respect to the passage of basinal brines. This could be important for formation of ore deposits in basement rocks and the hydrogeology of the substrate of sedimentary basins. In the future we hope to develop models for fluid circulation where permeability is generated by porosity-forming reactions (cf. Balashov and Yardley, 1998). These may be able to predict patterns of mineral and isotopic zoning as a function of flow rate and extent of fluid-rock reaction.
Balashov, V.N. and Yardley, B.W.D. (1998) Modelling metamorphic fluid flow with reaction-compaction-permeability feedbacks. Am. J. Sci. 298, 441-470.
Eiler, J.M., Valley, J.W. and Graham, C.M. (1997) Standardisation of SIMS analysis of O and C isotope ratios in carbonates from ALH84001. Abstr. 28th Lunar and Planetary Science Conference, part 1, 327-328.
Fein, J.M., Graham, C.M., Holness, M.B., Fallick, A.E., and Skelton, A.D.L. (1994) Controls on the mechanisms of fluid infiltration and front advection during regional metamorphism: a stable isotope and textural study of retrograde Dalradian rocks of the SW Scottish Highlands. J Metamorphic Geol. 12, 249-260.
Graham, C.M., Greig, K.M., Sheppard, S.M.F. and Turi, B. (1983) Genesis and mobility of the H2O-CO2 fluid phase during regional greenschist and epidote-amphibolite facies metamorphism: a petrological and stable isotope study in the Scottish Dalradian. J Geol. Soc. London, 140, 577-599.
Guest, R.E, McCaig, A.M. and Graham, C.M. (1999) Mechanisms of retrogressive dolomitisation of marble, SW Highlands, Scotland. Terra Abstracts 11, 705.
Guest, R.E, McCaig, A.M. and Graham, C.M. (2000) Retrogressive dolomitisation and fluid flow in the SW Highlands, Scotland. Report on IMP/156/1099
Valley, J.W., Graham, C.M., Harte, B., Eiler, J.W. and Kinny, P.D. (1997a) Ion microprobe analysis of oxygen, carbon and hydrogen isotope ratios. Soc.Econ.Geol. Rev. Econ.Geol. 7, (in press).
Return to text
Fig. 2a: CL montage of part of sample 2 (Fig. 1) showing domain 4a dolomite rhombs growing in grain boundary areas of domain 3 recrystallised calcite. Dark, zoned calcite of domain 4b is intimately associated with dolomite rhombs. It is not clear whether domain 4b zones are truncated by dolomites or butt up to them passively. Numbers are ion probe δ18O SMOW values. Note bright, isotopically heavy domain 5 zones lining porosity.
Fig. 2b: Backscatter SEM collage of approximately the same area as 2a. Zoning of dolomite rhombs to more ankeritic compositions is clearly visible, as is replacement of rhomb cores locally by ankerite. Faint zones in Fe-rich domain 4b calcite can also be seen. Zone 4b and 5 calcite appears to be growing around dolomite at A, while calcite may be replacing dolomite at B. Porosity locally contains poorly imaged clay minerals
