Efficiency of compaction and compositional convection during mafic crystal mush solidification: the Sept Iles layered intrusion, Canada
To be submitted to Contributions to Mineralogy and Petrology
Olivier Namur1,2*, Bernard Charlier3
1.Département de Géologie, Université de Liège, Belgium
2.Department of Earth Sciences, University of Cambridge, Cambridge, UK
3.Department of Earth, Atmospheric and Planetary Sciences, Massachusetts Institute of Technology, Cambridge, USA
*Corresponding author:
Analytical methods
Whole rock analyses
All the samples were carefully cleaned prior to crushing. They were manually crushed with a hammer and milled in agate mortars. Whole-rock compositions were presented in Namur et al. (2010). Here, we present newly acquired accurate analyses of whole-rock trace elements (P, Zr, Cr and V). Measurements were performed by XRF using an ARL 9400XP spectrometer at the University of Liège on pressed powder pellets. Eight successive measurements were realized for each sample and the values reported here correspond to the average of the 8 measurements. Data were corrected for matrix effects by Compton peak monitoring. The spectrometer was calibrated for phosphorous analyses using 40 international standards ranging from 0.01 to 5 wt.% P2O5. Repeated measurements (n=17) of 15 international standards (0.01-1.39 wt.% P2O5), some of them being not included in the calibration curve, indicate an accuracy relative to the reference values in the range between 0.3 and 10.5 %, with an average of 3.1 %. For other trace elements, the spectrometer was calibrated using 64 international standards and analytical errors are estimated to be lower than 10 % in most samples.
Mineral analyses
One or two thin sections for each sample were carefully examined under the microscope to select samples for electron microprobe analyses of plagioclase. Plagioclase compositions from 38 samples were obtained with a Cameca SX100 electron probe micro-analyzer (EPMA) at the University of Clermont-Ferrand (France). Analytical conditions were 15 kV for the accelerating voltage and 15 nA for the beam current. A focused beam was used for the analyses. The following standards were used for K X-ray lines calibration: wollastonite for Si and Ca; albite for Na; forsterite for Mg, fayalite for Fe, orthoclase for K, synthetic Al2O3 for Al and synthetic TiMnO3 for Ti. Standard ZAF corrections have been made on all analyses.
In situ mineral analyses of trace elements were performed by LA-ICP-MS at the Research School of Earth Sciences (RSES; the Australian National University, Australia). A pulsed 193 nm ArF Excimer laser with 100 mJ energy at a repetition rate of 5 Hz coupled with an Agilent HP7500 quadrupole ICP-MS system was used for ablation (Eggins et al., 1998). Laser sampling was performed in a He-Ar atmosphere with a beam diameter of 150 µm. Analyses were calibrated using 29Si and 47Ti as internal standard isotopes based on SiO2and TiO2concentrations measured by EPMA and XRF. During time resolved analyses of minerals, possible contamination from inclusions and fractures was detected by monitoring several elements and only the “clean” part of the signals was integrated. The glasses NIST-612 (Pearce et al., 1997) and BCR-2G (Norman et al., 1998) were employed as external and secondary standards, respectively. The reproducibility of trace element results for the BCR glasses using the RSES analytical protocols are between 0.5 % and 4 % relative (1 sigma) for the majority of elements (Eggins, 2003).
Bulk rock densities
Bulk-rock densities (g/cm3) were determined by weighing the samples (5-15 cm in length and 4.6 cm diameter) in air and then in water. Replicate measurements (n=15) for 5 samples indicated an uncertainty between 0.001 and 0.012 g/cm3, with an average value of 0.007 g/cm3. Density of the high-temperature crystal matrix was also calculated at the liquidus temperature using mineral mode and mineral compositions given in Namur et al. (2010) and thermal expansion coefficients of Niu and Batiza (1991). Liquidus temperatures were estimated using the linear relationship between temperature and the An-content of plagioclase experimentally observed by Toplis and Carroll (1995).
Compaction and compositional convection: Theoretical formulations
Compaction of the crystal mush
In a crystal mush that accumulates at the floor of a magma chamber, the lower part is represented by a compacting layer, while the upper part is not compacted (Fig. A1). The porosity decrease in the compacting layer results from the expulsion of the intercumulus melt and its progressive crystallization. With time, the front of compaction moves upwards through the crystal mush. Original equations for the compaction process have been formulated by McKenzie (1984; 1985) and Sparks et al. (1985) and assume textural equilibrium between the crystals and the liquid in the crystal mush, i.e. there is no occlusion of the pores when the residual porosity becomes low (Hunter, 1987; Mathez et al., 1997). The theoretical thickness of the compacting layer is given by the compaction length scale parameter, (m) i.e. the height over which the compaction rate decreases by a factor e (2.718), that can range from as much as km to as little as a few cm, depending on the values of the relevant physical parameters of the crystal mush and the liquid:
(1)
where and are the bulk and shear viscosities of the crystal mush (Pa.s), øis the permeability of the crystal matrix (m2) and is the melt viscosity (Pa.s). Following Maaloe and Scheie (1982) and McKenzie (1985), the permeability of the crystal mush can be calculated as follows:
(2)
where a is the average grain size (grain radius, m) and is the porosity, i.e. the fraction of intercumulus melt by volume unit (ranging from 0 to 1).
In large layered intrusions, the thickness of the cumulate pile where compaction occurs is generally significantly different than the length scale parameter, (Sparks et al., 1985; Tharp et al., 1998; Mathez et al., 1997). In a partially molten cumulate layer of thickness h (m), the velocity of the expelled intercumulus melt is zero at the base of the layer and increases upwards. The relative velocity of the liquid at the top of the compacting layer (; m/s) is considered as the expression of the compaction rate (Sparks et al., 1985) and can be calculated as follows:
(3)
where (m/s) is the velocity of the expelled liquid, W (m/s) isthe velocity of the sinking crystal matrix and 0 (m/s) is the relative velocity between the melt and the crystal matrix of initially constant porosity, i.e. when no compaction occurs, and is defined by:
(4)
where sand l are the densities (g/cm3) of the crystal matrix and the liquid, respectively, and g is the standard acceleration constant.
Fig. A1: Theoretical model for compaction of a crystal mush (crystal matrix + liquid) beneath the liquid of the main magma body in a magma chamber. a. Density distribution between liquid and crystal matrix necessary for the onset of compaction. b. Representation of the effect of compaction on the fluid flow, the stratigraphy of the crystal mush and its porosity. It should be noted that the porosity of the not compacted part of the crystal mush could be reduced by crystallization of the intercumulus melt c. Physical properties of the liquid and the crystal mush that enhance the compaction efficiency (rate of compaction). d. Physical properties of the liquid and the crystal mush that decrease the compaction efficiency.
Compositional convection
Crystallization and fractionation of the intercumulus melt may locally decrease its density down to values lower than the density of the overlying magma (Fig. 2a). In this case, the intercumulus melt may be continuously expelled from the crystal mush through convective separation of the melt from growing crystal, and replaced in the pore space by the overlying liquid, generally coming from the main magma body. This process called compositional convection enables the pore melt to maintain a constant composition (Sparks and Huppert, 1984; Sparks et al., 1984; 1985; Tait et al., 1984). The original mathematical expression of compositional convection was given by Tait et al. (1984), Sparks et al. (1985) and Kerr and Tait (1986).
Instability occurs in the porous medium and leads to upwards intercumulus melt convection when the dimensionless local solutal Rayleigh number (Ra) of the compositional porous boundary layer ahead of the advancing crystallization front exceeds a critical value (25-80; Lapwood, 1948; Nield, 1968; Sparks et al., 1985; Tait and Jaupart, 1992; Tait et al., 1992). Following Tait et al., (1984), Racan be calculated as follows:
(5)
where is the permeability of the crystal matrix (m2), is the compositional density difference across the boundary layer (g/cm3), h is the vertical thickness of the porous medium (m), g is the standard acceleration constant, D is the melt diffusivity of the chemical species involved in the convection process (m2/s) and µ is the viscosity of the melt (Pa.s).
Following Sparks et al. (1985), the velocity at which the liquid is expelled from the pore space (characteristic convective velocity; vc) can be calculated as follows:
(6)
Liquid expelling trough compositional convection will occur if vc is higher than the rate of crystal accumulation (Kerr and Tait, 1986). If vcis lower than the rate of crystal accumulation, convection can occur but will be not efficient because the pore fluid will be frozen before it can move.
Fig. A2 Theoretical model for compositional convection within a crystal mush. a. Density distribution between the intercumulus melt and the overlying melt (here the main magma body) necessary for the onset of compositional convection. b. Representation of the effect of compositional convection on the fluid flow. The downward decrease of crystal mush porosity is due to the crystallization of the intercumulus melt of uniform composition. c. Physical properties of the liquid and the crystal mush that enhance the compositional convection efficiency (rate of compaction). d. Physical properties of the liquid and the crystal mush that decrease the compositional convection efficiency.
References
Eggins, S.M., 2003. Laser ablation ICP-MS analysis of geological materials prepared as lithium borate glasses. Geost. Newslett. 27, 147-162.
Eggins, S.M., Rudnick, R.L., McDonough, W.F., 1998. The composition of peridotites and their minerals: a laser-ablation ICP-MS study. Earth Planet. Sci. Lett.154, 53-71.
Hunter, R.H., 1987. Textural equilibrium in layered igneous rocks. In: Parsons I (Ed). Origins of igneous layering. Dordrecht, Reidel, pp 473-503.
Kerr, R.C., Tait, S.R., 1986. Crystallization and compositional convection in a porous medium with application to layered igneous intrusions. J. Geophys. Res. 91, 3591-3608.
Lapwood, E.R., 1948. Convection of a fluid in a porous medium. Proceed. Cambridge Phil. Soc. 44, 508-521.
Maaloe, S., Scheie, A., 1982. The permeability controlled accumulation of primary magma. Contrib. Mineral. Petrol. 81, 350-357.
Mathez, E.A., Hunter, R.H., Kinzler, R., 1997. Petrologic evolution of partially molten cumulate: The Atok section of the Bushveld Complex. Contrib. Mineral. Petrol. 129, 20-34.
McKenzie, D.P., 1984. The generation of compaction of partially molten rock. J. Petrol. 25, 713-765.
McKenzie, D.P., 1985. The extraction of magma from the crust and mantle. Earth Plant. Sci. Lett. 74, 81-91.
Namur, O., Charlier, B., Toplis, M.J., Higgins, M.D., Liégeois, J.P., Vander Auwera, J., 2010. Crystallization sequence and magma chamber processes in the ferrobasaltic Sept Iles layered intrusion, Canada. J. Petrol. 51, 1203-1236.
Nield, D.A., 1968. Onset of thermohaline convection in a porous medium. Wat. Resour. Res. 4, 553-560.
Niu, Y., Batiza, R., 1991. In situ densities of MORB melts and residual mantle: Implications for buoyancy forces beneath mid-ocean ridges. J. Geol. 99, 767-775.
Norman, M.D., Griffin, W.L., Pearson, N.J., Garcia, M.O., O’Reilly, S.Y., 1998.Quantitative analysis of trace element abundances in glasses and minerals: a comparison of laser ablation inductively coupled plasma mass spectrometry, solution inductively coupled plasma mass spectrometry, proton microprobe and electron microprobe data. J Anal. Atom. Spectrom.13, 477–482.
Pearce, N.J., Perkins, W.T., Westgate, J.A., Gorton, M.J., Jackson, S.E., Neal, C.R., Chenery, S.P., 1997. A compilation of new and published major and trace element data for NIST SRM 610 and NIST SRM 612 glass reference materials. Geost. Newslett.21, 115-144.
Sparks, R.S., Huppert, H.E., 1984. Density changes during the fractional crystallization of basaltic magmas: Fluid dynamic implications. Contrib. Mineral. Petrol. 85, 300-309.
Sparks, R.S., Huppert, H.E., Turner, J.S., 1984. The fluid dynamics of evolving magma chambers. Phil. Trans. Roy. Soc. Lond. A310, 511-534.
Sparks, R.S., Huppert, H.E., Kerr, R.C., McKenzie, D.P., Tait, S.R., 1985. Postcumulus processes in layered intrusions. Geol. Mag. 122, 555-568.
Tait, S.R., Huppert, H.E., Sparks, R.S., 1984. The role of compositional convection in the formation of adcumulate rocks. Lithos 17, 139-146.
Tait, S., Jaupart, C., 1992. Compositional convection in a reactive crystalline mush and melt differentiation. J. Geophys. Res. 97, 6735-6756.
Tait, S., Jahrling, K., Jaupart, C., 1992. The planform of compositional convection and chimney formation in a mushy layer. Nature 359, 406-408.
Tharp, T.M., Loucks, R.R., Sack, R.O., 1998. Modeling compaction of olivine cumulates in the Muskox intrusion. Am. J. Sci. 298, 758-790.
Toplis, M.J., Carroll, M.R., 1995. An experimental study of the influence of oxygen fugacity on Fe-Ti oxide stability, phase relations, and mineral-melt equilibria in ferro-basaltic systems. J. Petrol. 36, 1137-1170.