Potassium metasomatism of Precambrian paleosols Alexey A Novoselov

Potassium metasomatism of Precambrian paleosols Alexey A. Novoselov and Carlos Roberto de Souza Filho Institute of Geosciences, University of Campinas (UNICAMP) 47 th Brazilian Geological Congress Salvador 2014

RESEARCH OBJECTIVES Precambrian paleosols (or weathering regoliths) were formed under the direct influence of ancient rainfalls. They incorporated, in their compositions, chemical features corresponding with the Precambrian atmosphere and climate. The environmental conditions under which paleosols were formed have been reconstructed by a numbers of researchers. Paleosols have been used to understand the Great Oxidation Event (e. g. , Rye and Holland, 1998, 2000; Holland, 2009) and to constrain the CO 2 levels during the Neoarchean and Proterozoic (e. g. , Sheldon, 2006; Mitchell and Sheldon, 2010; Driese et al. , 2011). Modern weathering profile formed on the basaltic substratum near Campinas, SP Introduction

RESEARCH OBJECTIVES Independently of substratum composition, ancient regoliths formed by subaerial weathering are characterized by accumulation of K and other alkaline elements. Introduction

GEOCHEMICAL MODELING This is a research technique allowing numerical simulations of chemical reactions passing in the minerals-solution-gas system; includes itself thermodynamic, balance, kinetic, transport and other calculations. There a number of simulation software that combine kinetic and thermodynamic calculations: e. g. , KINDISP (Made et al. , 1994), PHREEQC (Parkhurst & Appelo, 1999), Crunch. Flow (Steefel, 2001), GEMS (Kulik et al. , 2004), Geochemist's Workbench (Bethke & Yeakel, 2012), GEOCHEQ (Mironenko and Zolotov, 2012). At each timestep, these codes firstly calculate the quantity of dissolved minerals and then the chemical equilibrium composition of the system, yielding precipitated solid phases and solution composition. Powerful universal codes, such as Geochemist's Workbench, Crunch. Flow and PHREEQC, combine this procedure with transport models and account of precipitation kinetics. Solution Minerals Gas n tio S IC on ti ta pi ci ET N u ol iss D KI e Pr Reactor Calculation of equilibrium Dissolved matter during Δt [System composition](t+ t) = W/R [Solution composition]t + Δt Σ(Ratei Si) (Zolotov and Mironenko, 2007) Δt is determined by the used algorithm of timestep calculation and constrained by the transport model. Also the transport model impacts the geochemical calculations The method description through the water-rock ratio (W/R).

Δt Σ(Ratei Si) Equation of mineral dissolution-precipitation rate Rate = f 1(p. H)·f 2(T)·f 3(Q/K) (Palandri & Kharaka, 2004), (Zolotov & Mironenko, 2007), (Brantley, 2008) The method description

Δt Σ(Ratei Si) THE REACTIONARY SURFACE The specific surface area (SSA) of the most rocks is 0. 01 -1000 m 2/g (Brantley et al. , 1999). Si = νi SSA Smk, νi – the volume portion of mineral j, Smk - is a sum of primary or secondary minerals' weights. Index and type of Observed Weight, g surface, m 2/g samples 6005 granite 1. 241 0. 451 ± 0. 064 22105 basalt 1. 272 1. 432 ± 0. 005 2906 granite 1. 102 0. 264 ± 0. 031 Кс-1 clay 0. 857 53. 04 ± 2. 06 SEM microphotographs illustrate the olivine dissolution (Lazaro and Brouwers, 2010) The method description

Calculation of equilibrium composition of the system Example of the calculating system i Component G°i T=25°C, P=1 bar, cal aji O H Si z p. H xi Condition (bj) Content (bj), mol 1 H 2 O, aq -56690 1 2 0 0 0 55. 509999821 O 53. 5102 2 H+ 0 0 1 1 1. 778286344 e-7 H 111. 02 3 H 2, aq 4236 0 2 0 0 0 2. 000002 Si 0. 0001 4 HSi. O 3 - -242801 3 1 1 -1 0 1. 219906905 e-7 z 0 5 O 2, aq 3954 2 0 0 3. 41762567 e-90 14 6 OH- -37595 1 1 0 -1 1 5. 623391324 e-8 -log 10(H+)log 10(OH-) 7 Si. O 2, aq -199190 2 0 1 0 0 9. 987805954 e-5 Gibbs-Duhem equation Constraints The minimum of the system is looking with the None-linear Programming methods (Lagrange Multipliers, Penalty Functions and Conjugated Gradients methods). The method description

THE SCHEME OF THE CRONO SOFTWARE OPERATION The CRONO is developed as a free software with open code. The software is intended for simulation of weathering and hydrothermal processes. Currently the preliminary version of the code is available in the internet: http: //www. ige. unicamp. br/crono/ Capabilities of the approach: - Long-term geochemical processes or ones proceeding at difficult of access places can be reconstructed. This makes geochemical simulations the most powerful technique to study the forming of ground waters, hydrothermal fluids, processes with duration Kyr and Myr, ect. - Impact of distinct factors to a complex phenomenon can be understood, e. g. impact of chemical weathering to the forming of soil. Limitations of the approach: - The composition of modeling systems should be fully determined: T, P, duration of interaction, rate of passing solution, composition of minerals, solution and gaseous phase. - The construction of models is limited by existed constants, account of distinct phenomena, e. g. model of activity of solution components. - Currently it is impossible to estimate the accuracy of the approach and estimate the impact of constants’ precision to the certainty of results. Written in DELPHI The method description

MODEL DESIGN In order to provide new insights to the K metasomatism , we simulated the burial of a modern weathering profile, its compaction and alteration by deep diagenetic fluids. As an object of this modeling the regolith formed on Parana basalts was considered. FG - fine-grained CG - coarse-grained M - monolithic Model design

MODEL DESIGN Model design

DISSOLUTION AT CLOSE-TO-EQUILIBRIUM CONDITIONS For plagioclase at 50°C and 180 bar: This plot evidences that magmatic plagioclase and K-feldspar are very resistant to dissolution. Pyroxenes are less stable and could be potentially dissolved in the course of interaction with more diluted solutions. Olivine can be dissolved throughout all the subsidence profile. Secondary albite and K-feldspar are close to equilibrium with percolating solutions and their dissolution is controlled by the affinity term and negligible variations of the cation activities. The most important impact to the activity quotient and affinity term is caused by Si. O 2, Ca+2, Na+ and K+. Kr = 6. 74 E - 60 Results

SOIL BASALT Results

CHEMICAL CHANGES The bulk chemical composition of basaltic horizons varies in a minor proportion only, excluding gains of calcium (21 %) and potassium (41 %). In contrast, the uppermost levels of the weathering profile are much more liable to fluctuations of chemical composition. The loss of silica reaches 13 % (final ~10 %), gains of Fe 2 O 3 - 2 -3 %, Mg. O 140 % (final 33 %), Ca. O - 1200 % (final ~550 %), Na 2 O - 600 % (final gains do not exceed 140 %) and K 2 O - ~190 %. The elemental composition formed at the subsidence level corresponding to 350°C does not change any further. Results

IMMOBILE ELEMENTS The fluctuation of P 2 O 5 content during the burial process doesn't exceed 0. 01 %, Ti. O 2 - 0. 11 % and Al 2 O 3 - 1. 42 % by weight. The nature of the immobility of these elements is distinct. Metasomatic fluids show respectively high contents of Al and P, whereas their Ti content is near zero. Results

POROSITY The actual porosity determines the water-rock (W/R) ratio, rates of fluid penetration and rock permeability. In case of decreased pore volume, the access of fluids ceases and mineral transformations are inhibited. Results

DEEP FLUIDS In contrast with deep fluids from sedimentary basins (Kharaka and Hanor, 2003), the modeled solutions show an elevated content of potassium and a reduced content of magnesium, iron and bicarbonate. Such differences can be explained by the modeling conditions: the simulated solutions interact with basaltic rocks, whereas the deep fluids dataset presents the formation waters from sedimentary rocks. The content of other solution components fully corresponds to the composition of natural fluids. Results

. The mineralogical transformations goes through the following sequence: The deposition of illite, muscovite and biotite is guided mostly by the content of Si, Al, Mg and Fe in the fluids, rather than the abundance of K. The simulated solutions are rich with K+-ion and, simultaneously, highly mineralized, which inhibits the dissolution of magmatic pyroxenes, K-feldspar and plagioclase. Such minerals are excluded from the bulk mineral exchange during the subsidence process. This allows to formulate a paradoxical conclusion: the maximal levels of K gains are caused by interactions with less mineralized fluids and generally independent of their K content. Results

Results

Fe(II)/Fe(III) RATIO The yielded calculations also show that Fe(III)-bearing minerals are not prone to endure subsidence. The resulted Fe(II)/Fe(III) ratios only occasionally reproduce the initial values of those ratios and should not be used exclusively during reconstructions of oxidation levels of Precambrian atmosphere. Results

EMPIRICAL INDICES Results

CONCLUSIONS • Potassium enrichment of Precambrian paleosols was not triggered by distinct weathering conditions in the Precambrian, but rather by their burial and long-term metasomatism. • The duration of those processes is about 500 Myr and more recent weathering profiles did not develop through enough time to produce comparable alterations. • The preservation of primary basaltic minerals, the higher alteration of the uppermost soil levels and the immobility of a few elements were also explained by the model. • The numerical experiment demonstrates the secondary nature of the Fe(II)/Fe(III) ratio and a high mobility of Ca, Mg, Na, in contrast to traditionally applied approaches that may understate the gains of those elements. This investigation was financially supported by FAPESP, grant No. 2011/12682 -3. Conclusions

Thanks a lot for your attention!!!