Geochemical anatomy of a spheroidally weathered diabase

    Geochemical anatomy of a spheroidally weathered diabase Anupam Banerjee, Ramananda Chakrabarti, Sourav Mandal PII: DOI: Reference:

S0009-2541(16)30340-0 doi: 10.1016/j.chemgeo.2016.07.008 CHEMGE 17996

To appear in:

Chemical Geology

Received date: Revised date: Accepted date:

24 March 2016 29 June 2016 10 July 2016

Please cite this article as: Banerjee, Anupam, Chakrabarti, Ramananda, Mandal, Sourav, Geochemical anatomy of a spheroidally weathered diabase, Chemical Geology (2016), doi: 10.1016/j.chemgeo.2016.07.008

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Geochemical anatomy of a spheroidally weathered diabase

Materials Research Center, Indian Institute of Science, Bangalore, India

NU

SC

2

Center for Earth Sciences, Indian Institute of Science, Bangalore, India

RI

1

PT

Anupam Banerjee1, Ramananda Chakrabarti1*,and Sourav Mandal2

* Corresponding Author: [email protected]/[email protected]

TE

D

MA

+91-80-2293-3003

Keywords: spheroidal weathering; preferential weathering of plagioclase and pyroxene; major

Highlights 

AC CE P

and trace elements; Nd, Sr isotopes; X-Ray micro-CT; timing of weathering

X-Ray micro-CT imaging of a spheroidally weathered diabase shows distribution of fractures and their relative timing of formation inside the sample



Significant variability in elemental composition in hand-specimen sized spheroidally weathered sample



Preferential weathering of plagioclase and pyroxene controls the values of weathering indices



Nd, Sr isotopic variability are related to modification of the Rb/Sr and Sm/Nd ratios during a major episode of weathering ~1.2-1.3 Ga ago

1

ACCEPTED MANUSCRIPT Abstract Major, trace element concentrations and Nd, Sr isotope ratios were measured in micro-drilled

PT

samples of a 2.37 Ga-old, hand-specimen sized spheroidally weathered diabase from southern India. A sample of the un-weathered diabase dike was also analyzed. X-Ray micro-CT imaging

RI

of the weathered sample shows three dominant mineral phases which are plagioclase, pyroxene, and a Fe-bearing phase (possibly hematite and/or ilmenite). This imaging documents the

SC

pervasive nature of two generations of ribbon-like, cross-cutting fractures. The older fracture is sealed while the more recent fracture is open without any in-filling. The values of the Chemical

NU

Index of Alteration (CIA) of the samples show a wide range but are less than 50. Despite being a relatively less weathered rock, we observe that concentrations of major, minor and trace

MA

elements vary significantly with the percentage relative standard deviation (%RSD) for the elements ranging from 10.2 - 41.8. The CIA of the samples do not show any trend with the position of the sample in the hand-specimen. Barring Ca and Li, whose concentrations decrease

TE

of the elements.

D

from the core to the rim of the sample, there is no significant spatial trend in the concentrations

AC CE P

Concentrations of Na, Al, and Sr increase with increasing CIA values while concentrations of Mg, Fe, and Sc decrease with increasing CIA. The strong positive correlations of Na and Al, as well as Na and Sr indicates preferential weathering of plagioclase in the diabase. Na/Ca increases while Mg/Al, Mg/Na, Mg/Ca, Fe/Al and Sc/Sr decrease with increasing CIA values and the unweathered rock plots in the middle of these trends. Such variations are explained in terms of differential weathering of plagioclase (in samples with lower CIA than the un-weathered rock, W1-type) and pyroxene (in samples with higher CIA than the un-weathered rock, W2-type) which have varying resistance to weathering. At the hand-specimen scale, the variability in the weathering indices like CIA are controlled by differential weathering of minerals and might not accurately reflect the intensity of weathering. Chondrite-normalized La/Sm and Gd/Lu co-vary with CIA values indicating mobility of the REEs during spheroidal weathering even at the handspecimen scale. The Eu anomaly also increases with increasing CIA values which is explained by differential weathering of pyroxene and plagioclase. We observe large percentage deviations of the Nb-normalized concentrations of elements from the un-weathered rock in specific samples but no spatial trend is observed. Overall, the variations in element concentrations can be 2

ACCEPTED MANUSCRIPT explained by varying fluid mobility of the elements, selective weathering of the minerals in the diabase, ambient environmental conditions, and heterogeneity inherent to the rock.

PT

Considerable Nd and Sr isotopic variability is observed at the hand-specimen scale and is explained in terms of weathering-related fractionation of parent/daughter ratios. This elemental

RI

fractionation must have happened long time ago to allow for radiogenic decay of the long-lived isotopes of 87Rb and 147Sm. The spread (%RSD) in the initial Sr and Nd isotope compositions of

SC

the weathered samples reach a minimum value around 1.2-1.3 Ga which we interpret as the timing of the peak weathering event which led to fractionation of the parent/daughter ratios. For

NU

Nd isotopes, the average Nd (1.2 Ga) of the weathered samples coincides with the Nd (1.2 Ga) of the un-weathered rock. The timing of the weathering event coincides with the timing of the

MA

breakup of the Columbia supercontinent and follows wide-spread alkaline volcanism in the Indian subcontinent. This is the first such attempt to determine the timing of a weathering event

AC CE P

TE

D

in rocks using long-lived radioactive isotopes.

3

ACCEPTED MANUSCRIPT 1. Introduction Silicate weathering is a fundamental process on Earth which consumes atmospheric CO2 and

PT

hence, plays an important role in the regulation of long-term (> 1 Ma) climatic conditions (Berner, 1995; Berner and Lasaga, 1989, Navarre-Sitchler and Brantley, 2007). During chemical

RI

weathering, rock-forming minerals interact with fluids and are converted to secondary minerals leading to loss and/or redistribution of elements. The weathering products of rocks are eventually

SC

transported by rivers (as dissolved and suspended loads) and are a major source of cations to the oceans (e.g., Whitfield, 1982). A common type of weathering process in rock outcrops is

NU

spheroidal weathering which is used to describe the characteristic formation of concentric layers or shells around a less weathered central core stone; this onion-skin like weathering pattern has

MA

been observed on a wide range of rock compositions like basalt, granite, andesite, sandstone, and schist in diverse modern-day climatic conditions (Fletcher et al., 2006; Jamtveit et al., 2011; Ollier, 1971; Royne et al., 2008). Spheroidal weathering is most commonly seen in rocks which

D

are dominantly composed of plagioclase, olivine, and pyroxene, and is initiated by fluids which

TE

percolate through fractures and joints thereby enhancing chemical reactions. Change in chemical compositions, like the hydration of alumina and iron in silicate minerals, cause an increase in

AC CE P

volume and expansion of the outer shells, which in turn applies stress causing fractures, and the outer layers to peel off (Buss et al., 2008; Jamtveit et al., 2009; Royne et al., 2008). A coupling between physical and chemical weathering leads to hierarchical fracturing and ultimately, layerby-layer spalling (Royne et al., 2008). Chemical weathering advance rates vary with equilibrium solubility and pore fluid velocities and are not affected by reaction kinetics in the case of transport-limited weathering (Brantley et al., 2011). Basalts, which are primarily composed of minerals like plagioclase and pyroxene, readily weather and have consequently been the focus of many studies on element mobility and changes in isotopic composition during chemical weathering at the outcrop scale (e.g., Aiuppa et al., 2000; Hill et al., 2000; Ma et al., 2006; Ma et al., 2007; Ma et al., 2010; Nesbitt and Wilson, 1992; Taylor and Lasaga, 1999; Wimpenny et al., 2007). Most of these studies have focussed on geologically recent weathering profiles that have developed on Phanerozoic-age rocks. In this study,we investigated the anatomy of a spheroidally weathered diabase of early Proterozoic age at the hand-specimen scale. We performed detailed imaging of the hand4

ACCEPTED MANUSCRIPT specimen sized sample to trace the fractures inside the rock, followed by careful micro-sampling and a detailed geochemical and isotopic investigation to evaluate the spatial extent of geochemical and isotopic heterogeneity that may be introduced due to spheroidal weathering.

PT

We show that elemental concentrations as well as radiogenic Nd, Sr isotopic compositions can vary significantly over small spatial scales of a few centimeters in weathered rocks. Our results

RI

provide deeper insight into the process of spheroidal weathering and how it affects the

SC

distribution of elements as well as radiogenic isotopic ratios over small spatial scales. Using the Nd, Sr isotopic data we attempt to constrain the timing of the peak weathering event.

NU

2. Sampling

A spheroidally weathered diabase (synonymously used as dolerite or microgabbro) sample,

MA

approximately 30 cm in diameter, and part of a dike, was collected from a quarry (12001.668'N, 77003.815'E) near the village of Yeragumballi, WSW of Hennur, Karnataka, India. An unweathered sample of the dike was also collected for geochemical comparison. These dikes are

D

well studied in terms of their mineralogy and petrography, bulk chemistry, and paleo-magnetism

TE

and have a U-Pb baddeleyite age of 2367 + 1 Ma (French and Heaman, 2010; Halls et al., 2007; Kumar et al., 2012). Major element data indicate that these dikes are iron-rich tholeiites,

AC CE P

dominantly consisting of clinopyroxene and plagioclase with minor amounts of amphibole, orthopyroxene, biotite and magnetite (Halls et al., 2007). Plagioclase in these rocks shows a brown to black colored clouding which was first reported by Pichamuthu (1959) and is explained by exsolution of magnetite from feldspar due to slow cooling at great crustal depths (Halls et al., 2007; Halls and Zhang, 1995). No intrusion of other rock types is found in this locality and hence, variations in elemental and isotopic compositions of the diabase are expected to be solely due to spheroidal weathering.

3. Analytical Methods The spheroidally weathered diabase sample was cut to a dimension of 6.3 cm x 6.0 cm x 2.5 cm such that multiple weathering layers could be sampled from the surface (rim) to the interior (core). Micro-computed tomography imaging of a representative portion of this sample was carried out using an X-Ray microscope (VersaXRM-510, Xradia, Zeiss, USA). A parallelepipedshaped sample of dimension ~3.0 cm x ~1.7 cm x ~1.4 cm, was scanned with isotropic voxel size 5

ACCEPTED MANUSCRIPT of 2.5 µm. A total of 3201 projections with a field of view (FOV) 2532.5 µm x 2532.5 µm were acquired, where the exposure time for each projection was 5 seconds. The source voltage and power were 140 kV and 10 W with transmission and intensity count of 18 - 27% and 3000 -

PT

8000, respectively. Reconstruction of collected radiographs was done in XMReconstructor with a Gaussian filter (kernel size 0.7). The reconstructed tomogram was further processed and

RI

visualized using Avizo Fire 8.1 (FEI, Bordeaux, France). A non-local-means filter was applied to

SC

the raw tomogram, which preferentially removes noise from the image while preserving the edges. The representative 2D images (Fig. 1 d-g), were taken as one voxel thick slices at three

NU

orthogonal planes denoted as XY, XZ and YZ.

Samples from 21 layers, from the surface to the interior of the sample were obtained using a

MA

hand-held micro-drill with a diamond drill bit (Figs. 1a, b). The samples are numbered based on their average radial distance from a reference point placed towards the core of the sample (Fig. 1c, Supplementary Table 1);sample number 1 is the outermost sample (closest to the rim) and

D

sample number 21 is the innermost (closest to the core) (Fig. 1b, c). Approximately 30-50 mg of

TE

sample powder was drilled from each layer. One gram of the un-weathered diabase was crushed using an agate mortar and pestle. For elemental concentration determination, ~25 mg of each

AC CE P

sample was dissolved in 15 ml screw-cap Teflon vials from Savillex, USA, using a mixture of 1 ml concentrated HF and 1 ml concentrated HNO3 acids at 120 0C. After 24 hours of reaction, the acid mixture was evaporated and a mixture of 1 ml concentrated HNO3 and 1 ml concentrated HCl was added to the sample. After another 24 hours of reaction at 120 0C, the acids were evaporated and the residue was dissolved in 2 ml of HNO3 and transferred to 125 ml pre-cleaned HDPE bottles (Tarsons, India). One ml of 1 ppm Be, In, Cs and Bi (internal standards, prepared from 1000 ppm Spec pure ICP solutions, Alfa Aesar) was added to the bottle and the solution was brought up to 100 ml by adding 18.2 MΩ water such that all sample, standard and blank solutions had ~ 10 ppb internal standards. All sample powders and solutions were weighed such that the dilution factor was ~4000 for all standards and samples. Sample preparation was done in a class 100 laminar flow work area in a class 10000 clean room at the Center for Earth Sciences (CEaS), Indian Institute of Science (IISc), Bangalore. We used Suprapur grade Hydrofluoric acid from Merck and Emsure brand (ACS, Reag. Ph Eur) nitric and hydrochloric acids, also from Merck. The nitric and hydrochloric acids were further distilled using a Savillex DST-1000 subboiling still before use. Our acid blanks are <1 ppt for the REEs, <1 ppb for most other elements 6

ACCEPTED MANUSCRIPT and <10 ppb for some major elements like Na. The 18.2 MΩ water was prepared using a Sartorious Stedim - Arium Pro VF filtration unit. The supply-water was filtered using additional

PT

coarse filters (5, 3, and 1 micron) prior to introduction into the above unit. Element concentrations were measured using a quadrupole Inductively Coupled Plasma Mass

RI

Spectrometer (ICPMS, Thermo Scientific X-Series II) at CEaS, IISc equipped with Nickel sample and skimmer cones. Samples were introduced using a 100 ml/min PFA nebulizer

SC

connected to a peristaltic pump running at 30 rpm into an ESI-PC3 Peltier cooled spray chamber. A CETAC ASX-520 auto-sampler was used. Uptake time for samples and standards was 60

NU

seconds while the rinse time (in 2% HNO3) was 90 seconds. A 10 ppb internal standard with Be, Cs, In, and Bi was used for drift correction. The concentrations of most elements (< 23 amu; >

MA

86 amu) were analyzed in the standard mode. The CeO+/Ce+ ratio was kept below 2% to minimize oxide-related interferences which affect the measurements of some of the middle to

D

heavy-REEs. The collision-cell mode (using a mixture of H2 and He gases) was used for certain

TE

elements (between masses 40 amu to 86 amu) prone to Ar-related molecular isobaric interferences. All raw counts were corrected for the contribution from acid blanks. USGS rock

AC CE P

standards BHVO-2 (Hawaiian basalt) and BCR-2 (Columbia River basalt) were used as calibration standards for determining the concentrations of most elements. The accuracy of the data was tested by analyzing the USGS rock standard AGV-2 (andesite) and BIR-1 (Icelandic basalt) interspersed with the samples. All USGS standards were processed the same way as the rock samples. The accuracy for most elements is better than 5%. For the REEs the accuracy is better than 2%. Internal precision (%RSD), based on three repeat measurements is better than 2% for most elements. Some samples were analyzed twice in two different analytical sessions and the (external) reproducibility of the data is better than 5% for all elements. For Nd and Sr isotopic measurements, approximately 10 mg of powdered sample from selected layers, showing large variations in Sr and Nd concentrations, were dissolved using concentrated HF, HNO3 and HCl acid mixtures following a protocol similar to that followed for concentration determination. Strontium and Nd were separated from the rock matrix using ion-exchange chromatography. Using BioRad AG50W X8 (100-200 mesh) resin, Sr was eluted in 3N HCl while the REEs were eluted in 6N HCl. The Sr fraction was subsequently processed in a smaller column using BioRad AG50W X8 (100-200 mesh) resin. The REE fraction was processed using

7

ACCEPTED MANUSCRIPT BioRad AG50W X8 (200-400 mesh) where Nd was separated from the other REEs using 0.2 M 2-Hydroxyisobutyric acid. The Nd fraction was subsequently processed in a smaller column using BioRad AG50W X8 (100-200 mesh) resin. The procedural blanks for Nd and Sr are

PT

estimated to be <400 pg.

RI

Strontium and Nd isotope ratio measurements were performed using a Thermal ionization mass spectrometer (TIMS, Thermo Triton Plus) at CEaS, IISc. Neodymium was loaded on degassed 143

SC

Re filaments using 0.1 N H3PO4 and analyzed using a double filament assembly. Measured Nd/144Nd ratios were normalized to 146Nd/144Nd = 0.7219 to correct for instrumental mass

NU

fractionation. Uncertainties in the measured 143Nd/144Nd ratios for the samples correspond to 3 in the sixth decimal place (internal precision), representing 2 sigma of the mean. JNdi-1 Nd

MA

isotopic-standard analyzed during the course of the study yielded 143Nd/144Nd = 0.512118 ± 9 (2SD, n= 5). Strontium was loaded on oxidized Ta filaments using 2N H3PO4 and analyzed using a single filament assembly. During the measurements, the filament temperature was continuously

D

monitored and the raw 87Sr/86Sr ratios are normalized to 86Sr/88Sr = 0.1194 to correct for

TE

instrumental mass fractionation. Uncertainties in the measured 87Sr/86Sr are less than 9 in the sixth decimal place (internal precision) representing 2 sigma of the mean. SRM-987 Sr isotopic-

4. Results

AC CE P

standard analyzed during this measurement yielded 87Sr/86Sr = 0.710265 ± 10 (2SD, n=5).

Large variations are observed in major, minor and trace element concentrations (Table 1) in the 21 samples of the spheroidally weathered diabase, which are much larger than the analytical uncertainties of the ICPMS measurements (less than 5% for most elements and less than 2% for REEs). Silicon concentration was not measured as hydrofluoric acid was used during sample dissolution. This dissolution protocol results in the formation of volatile SiF4, which is mostly lost. When compared to the un-weathered rock, the concentrations of elements in the 21 subsamples show both higher and lower values (Table 1). To assess the degree of weathering in the samples, we have utilized the following frequently used weathering indices: WIP = [(2Na2O/0.35)+(MgO/0.9) + (2K2O/0.25) + (CaO/0.7)]*100 (Parker, 1970) (1) CIA = [Al2O3/(Al2O3+CaO+Na2O+K2O)]*100 (Nesbitt and Young, 1982) (2) 8

ACCEPTED MANUSCRIPT CIW = [Al2O3/(Al2O3+CaO+Na2O)]*100 (Harnois, 1988) (3) MIA(O) = [Al2O3+Fe2O3/(Al2O3+Fe2O3+MgO+ CaO+Na2O+ K2O)]*100 (Babechuk et al., 2014) (4)

PT

MIA(R) = [Al2O3 /(Al2O3+Fe2O3+MgO+ CaO+Na2O+ K2O)]*100 (Babechuk et al., 2014) (5)

RI

Using the elemental concentration data (Table 1), which were converted to molecular

SC

proportions of oxides, we calculated the values for the above weathering indices for our samples (Table 2) and the variations are shown in Figure 2. The CIA value for the un-weathered rock

NU

(UW) is 37.8 and values for the spheroidally weathered samples range from 32.2-44.1 (Table 2). Samples with CIA values less than UW are referred to as W1-type (weathered - type 1) and the

MA

samples with CIA values more than UW are referred to as W2-type (weathered - type 2). There is no relationship between the CIA values of the samples and their position in the specimen; the CIA variations are almost identical to the CIW and MIA(R) variations and are broadly similar to

D

the MIA(O) variations (Fig. 2, Supplementary Fig. 2). Sample 19 shows the lowest CIA as well as

TE

CIW, MIA(R) and MIA(O)values while sample 9 shows the highest CIA as well as CIW, MIA(R) and MIA(O) values. The WIP index, however, shows a weak decreasing trend from the core

AC CE P

towards the rim of the sample (Fig. 2).

Concentrations of Ca and Li increase from the rim to the core (Fig. 3). Similar but weaker trends (R2< 0.5, not shown) are also observed for Mg, Fe, Zn and Lu concentrations. Spatial trends are not apparent for other elements. To understand the relative mobility of different elements during spheroidal weathering, we compared the concentration variations with respect to the CIA index. Concentrations of Al, Na and Sr increase with increasing CIA while Mg, Sc and Fe concentrations decrease with increasing CIA values (Fig. 4). With increasing CIA values, Na/Ca increases while Mg/Al, Mg/Na, Mg/Ca, Fe/Al and Sc/Sr decrease (Fig. 4). The W1-type and W2-type samples show distinct elemental ratios while in the same plots, the un-weathered sample plots in the middle of the trend (Fig. 4). Strong positive correlations are observed between the concentrations of Na and Al (R2 = 0.99), Na and Sr (R2 = 0.88), Rb and K (R2 = 0.82), Li and Ca (R2 = 0.80), Co and Mn (R2 = 0.85), Hf and Zr (R2 = 0.93) and Ta and Nb (R2 = 0.97) (Fig. 5). Moderate positive correlations are also observed between concentrations of Sc and Mg (R2 = 0.70), Mg and Ca (R2 = 0.64), as well as V and Fe (R2 = 0.72) (Fig. 5). Th/Nb and Hf/Nb ratios show a strong positive correlation (R2 = 0.95) (Fig. 5).To evaluate the mobility of 9

ACCEPTED MANUSCRIPT different elements and the extent of mobilization during weathering, we compared the changes in concentrations with respect to that of the fluid-immobile element Nb. We calculated the percentage deviation of the element/Nb ratio in all 21 samples from the same ratio in the un-

PT

weathered rock following the approach of Ma et al. (2007) and Nesbitt (1979). We do not observe any increasing or decreasing trend in the percentage deviation with sampling depth but a

RI

consistent variability is observed for most elements (Fig. 6).

SC

Chondrite-normalized REE patterns of the samples are shown in Figure 7a. All samples show a light-REE (LREE) enriched pattern and a relatively flat heavy-REE (HREE) trend. Significant

NU

variations are observed in the REE concentrations; the total REE concentrations vary between 43-81 ppm (Table 1). The average REE pattern of the 21 samples closely mimics the un-

MA

weathered rock composition (Fig. 7b). The extent of LREE fractionation, expressed by the La/Sm(N) ratio varies between 1.60 and 2.13 (Table 1) and shows a positive correlation with the CIA value (Fig. 7c). The extent of HREE fractionation, expressed by the Gd/Lu(N) ratio varies

D

between 0.93 and 1.24 (Table 1); the Gd/Lu(N) ratios also co-vary with the CIA values (Fig. 7d).

TE

Europium anomaly (Eu/Eu*) in the samples was calculated as: Eu/Eu* = Eu(N) / [Sm(N)*Gd(N)]0.5 and varies between 0.77 and 1.39 (Table 1) and shows a rough positive correlation with the CIA

AC CE P

values of the sample (Fig. 7e). Cerium anomaly (Ce/Ce*), calculated as: Ce/Ce* = Ce(N) / [LaN)*Pr(N)]0.5 shows less variability. Barring samples 19 and 2, with Ce/Ce* of 1.16, and 1.27, respectively, Ce/Ce* ranges from 0.96-1.1 (Table 1) and is not correlated with the CIA values (Fig. 7f). Samples with high Ce/Ce* also show high Ce/Nb values (Supplementary Fig. 3). We analyzed six samples with significant variations in Nd and Sr concentrations, and the unweathered rock for Nd and Sr isotopic compositions (Table 3). The six samples include three W1-type and three W2-type samples. Measured 143Nd/144Nd for the spheroidally weathered samples ranges from 0.511907 to 0.512204 which translates to Nd(0) values ranging from -14.26 to -8.47 while measured 87Sr/86Sr for these samples ranges from 0.715481 to 0.738538 (Table 3). The un-weathered rock shows 143Nd/144Nd = 0.512099 (Nd(0) = -10.51) and 87Sr/86Sr = 0.722528. Values of Nd(0) for the spheroidally weathered samples show a weak positive correlation with 147

Sm/144Nd (R2 = 0.31) while a comparatively stronger positive correlation (R2 = 0.62) is

observed between 87Sr/86Sr and 87Rb/86Sr (Fig. 8a, b). The correlations between Nd(0) and

10

ACCEPTED MANUSCRIPT 147

Sm/144Nd (R2 = 0.88) and between 87Sr/86Sr and 87Rb/86Sr (R2 = 0.91) are stronger if sample#

4, which shows anomalously radiogenic 87Sr/86Sr and unradiogenic Nd(0), is not considered.

PT

5. Discussion Much of our knowledge of the early evolution of the Earth is derived from the study of

RI

Precambrian rocks. These rocks, however, have been exposed to mechanical and chemical

SC

weathering over time and in weathered rocks, much of the original chemical and isotopic signatures are modified and reflect the composition of the fluid responsible for the weathering

NU

process, secondary minerals and/or precipitates, or the residue after fluid-mobile elements have been leached out. The extent and nature of this geochemical modification depends on the

MA

water/rock ratio, the time available for weathering, as well as ambient redox conditions. Mobility of an element is related to the rate and extent of breakdown of minerals in rocks as weathering progresses (Aiuppa et al., 2000; Eggleton et al., 1987; Hill et al., 2000; Ma et al., 2006; Nesbitt

D

and Wilson, 1992).

TE

To assess the extent of weathering in rocks, several weathering indices have been proposed which utilize the relative proportions of fluid-mobile (e.g., Na, K, Ca, etc.) and fluid-immobile

AC CE P

(e.g., Al) major elements e.g., weathering index (WIP, Parker, 1970), chemical index of alteration (CIA, Nesbitt and Young, 1982) and chemical index of weathering (CIW, Harnois, 1988) and more recently, the mafic index of alteration under oxiding and reducing conditions (MIA(O), MIA(R)) which also includes Fe and Mg-oxides (Babechuk et al., 2014) (see equations 1-5). Chemical weathering indices have limitations as they can be affected by the amount of clay minerals and sesquioxides in the rock (Duzgoren-Aydin et al., 2002) and their predictive performance is influenced by the composition of the un-weathered parent rock (Ohta and Arai, 2007) as well as rock heterogeneity (Price and Velbel, 2003). As discussed later, an added limitation of these chemical weathering indices arises from preferential weathering of minerals over small spatial scales. Degree of weathering can also be expressed in terms of trace element mobility with respect to the immobile trace elements (Hill et al., 2000; Kurtz 2000; Nesbitt, 1979) or changes in bulk density (c.f., Patino, 2003). Although powerful, these geochemical proxies do not provide any information on how pervasive the weathering is in a particular rock sample and high-resolution imaging of weathered rocks and minerals are rare (Jamtveit et al., 2011; Velbel, 2009). 11

ACCEPTED MANUSCRIPT

5.1. 3D imaging of spheroidal weathering features In order to investigate how pervasive the effects of weathering are, we imaged a small

PT

representative portion of our sample with visible fractures using X-Ray microscopy and constructed a tomogram of the sample (Fig. 1). Micro computed tomography imaging offers a

RI

better visualization of the interior of sample and is non-destructive. However, this technique has

SC

rarely been applied to spheroidally weathered rocks (e.g., Jamtveit et al., 2011). The tomogram of the weathered diabase shows three dominant mineral phases represented by varying shades of

NU

gray (Fig.1d-g). The darkest phase is plagioclase, the slightly brighter phase is pyroxene and the brightest is represented by a Fe-bearing phase (possibly hematite and or ilmenite) which exhibits

(Halls et al., 2007; Pichamuthu, 1951).

MA

exsolution features (Fig. 1g), consistent with the clouding of plagioclase reported in these rocks

The tomogram shows the spatial distribution of the fracture which appears as a thin dark ribbon

D

in the middle of the sample (Fig. 1d) as well as in the XZ and YZ sections (Fig. 1f, g). In the

TE

planar XY section (Fig. 1e), the fracture appears wider. Surprisingly, another ribbon-like layer with a lighter gray shade appears intertwined with the darker fracture. The darker ribbon-like

AC CE P

fracture cross-cuts the lighter gray colored ribbon-like deposit indicating its formation at a later stage (Fig. 1d, f, g). The exact composition of the ribbon-like deposit is unclear. These could be secondary clay-minerals although, in petrographic thin-sections (not shown), clay minerals are not observed. In the micro-CT image, two stages of weathering seem apparent whereas only a single fracture is visible to the naked eye. The secondary deposit represented by the lighter gray ribbon-like layer is likely to have contributed to the chemical variability observed in our data. This weathering feature cross-cuts all the three mineral phases including the Fe-bearing phase, consistent with an early origin of the Fe-bearing phase at great crustal depths before the rock was exposed to weathering (Halls and Zhang, 1995; Pichamuthu, 1959). It is suggested that mineralogical (chemical) changes in rocks induce the physical changes during spheroidal weathering (Buss et al., 2008; Jamtveit et al., 2009; Royne et al., 2008). This argument is consistent with our observation that the darker ribbon-like fracture cross-cuts the lighter ribbonlike deposit. However, the morphological appearance of the lighter gray ribbon-like deposit suggests that it occupies an older generation fracture. This would imply that multiple generations of fracture occurred during spheroidal weathering of the sample. The older fracture could also 12

ACCEPTED MANUSCRIPT predate the episode of spheroidal weathering although, in the absence of any in-situ analysis, it is difficult to test these hypotheses.

PT

5.2. Elemental concentration variability and extent of weathering

The spheroidally weathered samples are relatively 'fresh' given that the CIA values are less than

RI

50 (Nesbitt and Wilson, 1992; Nesbitt and Young, 1982); the CIA value of the un-weathered

SC

rock is 37.8 (Table 2). Despite being a relatively less weathered rock, we observe that concentrations of major, minor and trace elements vary significantly over a small sampling area

NU

(Table 1). The percentage relative standard deviation (%RSD, 100*SD/mean) for all the elements range from 10.2-41.8. Some of this variability could be explained by heterogeneity

MA

inherent to the rock due to varying partitioning of elements into the different minerals but for a fine grained rock, we believe most of this variation is weathering-induced. In addition, given that the sampling spot-size, which is of the order of a cm (Fig. 1b), is much larger than the grain size

D

of the sample, the heterogeneity inherent to the rock would be averaged out during sampling.

TE

In weathered horizons, samples closer to the core-stone preserve the original composition while the ones away from it show more saprolitization (e.g., Fletcher et al., 2006). We wanted to

AC CE P

investigate whether a similar trend was observed in a hand-specimen scale. WIP values show an increasing trend from the rim to the core consistent with less weathering towards the core (Fig. 2). However, CIA, CIW, MIA(R) and MIA(O) values of the same samples do not show any such trend (Fig. 2). There is no significant trend between the concentration of the elements and their position within the hand-specimen barring Ca and Li. Concentrations of Ca and Li increase from the rim to the core suggesting preferential removal of these elements from outer portions of the sample (Fig. 3). A similar but weaker trend (R2 = 0.4) is observed for Mg concentrations (not shown) of the samples. The trend observed for Ca (and Mg) is similar to the increasing trend in the WIP values which is a function of the Mg, Ca, K, and Na concentrations in a sample. Both Na and K are fluid-mobile elements but their concentrations do not show any trend with the position of the samples. This contrasting behavior of Ca (and Mg) and Na, K, all considered to be fluid-mobile, is surprising and is explained in terms of differential weathering of minerals in the following sections.

13

ACCEPTED MANUSCRIPT If CIA is a robust measure of the intensity of weathering, variations in major, minor and trace element concentrations should correlate with it. During chemical weathering, major elements like Ca, Na, and K are considered to be mobile whereas Ti and Al are generally considered to be

PT

immobile (e.g., Aiuppa et al., 2000; Chesworth et al., 1981; Eggleton et al., 1987; Gardner, 1981; Kurtz 2000; Marsh, 1991; Nesbitt et al., 1980; Nesbitt and Wilson, 1992; Patino, 2003).

RI

Interestingly, concentrations of both Na, a fluid-mobile element, and Al, considered to be a fluid-

SC

immobile element, increase with increasing CIA values (Fig. 4) and have a positive correlation with each other (Fig. 5). This is consistent with differential weathering of plagioclase feldspar in

NU

the diabase. Sodium concentraion shows a strong positive correlation with Sr concentration (Fig. 5) which also increases with increasing CIA (Fig. 4). In contrast to Sr, Ca concentrations show a

MA

weak decreasing trend with increasing CIA (not plotted) thereby ruling out the role of any carbonate-bearing fluids or precipitates. The contrasting behavior of Na and Ca suggests that these elements are hosted in different mineral phases in the rock (Na in plagioclase versus Ca,

D

dominantly in clinopyroxene) with varying resistance to weathering. The moderate correlation

TE

between Ca and Mg and the strong correlation between Li and Ca concentrations (Fig. 5) are consistent with pyroxene-weathering in the diabase.

AC CE P

The values of the different weathering indices (CIA, CIW, MIA(O), MIA(R)) depend on the relative concentrations of fluid-immobile elements like Al and Fe (ferric form) and fluid-mobile elements like Na, K, Ca and Mg. However, variations in the concentrations of these elements can be affected by selective weathering of plagioclase (e.g., for Na, Al, as well as Ca) and pyroxene (e.g., for Ca, Mg, Fe as well as Al). The ratio of Na/Ca increases with increasing CIA values with the un-weathered rock plotting exactly between the W1- and W2-type samples (Fig. 4). The low Na/Ca values in the W1-type samples reflect plagioclase weathering in the rock leading to loss of fluid-mobile Na while the high Na/Ca in the W2-type samples can be explained by selective weathering of clino-pyroxene leading to preferential removal of Ca. The decreasing trends of the Mg/Al, Mg/Na, Mg/Ca and Fe/Al ratios with increasing CIA can also be explained similarly. High Mg/Al, Mg/Na, Mg/Ca and Fe/Al in the W1-type samples reflect preferential weathering of plagioclase leading to removal of Al, Na, and Ca, while the low Mg/Al, Mg/Na, Mg/Ca and Fe/Al in the W2-type samples is consistent with selective weathering of a mafic mineral like pyroxene leading to preferential removal of Mg and Fe. In the diabase, Fe is also be hosted in other oxides (Fig. 1d-g) and their weathering can explain the relatively large spread in 14

ACCEPTED MANUSCRIPT the Fe/Al ratio (Fig. 4). The same explanation holds true for variability in trace element ratios like Sc/Sr which decreases with increasing CIA (Fig. 4). Strontium is relatively enriched in plagioclase while Sc is relatively enriched in mafic minerals like pyroxene and hence,

PT

plagioclase weathering leads to high Sc/Sr as observed in the W1-type samples while pyroxene weathering can explain the low Sc/Sr in the W2-type samples. In case of all the element ratios

RI

discussed above, the composition of the un-weathered rock plots between the W1- and W2-type

SC

samples which can be explained by mass balance in a dominantly plagioclase and pyroxenebearing rock like diabase. While weathering indices have proved to be very useful in quantifying

NU

the intensity of chemical weathering of rocks in an outcrop scale, our study suggests that in a hand-specimen scale, variations in the values of such indices are affected by selective weathering

MA

of minerals and might not reflect the intensity of weathering. This also explains why the CIA and CIW values of the un-weathered diabase (UW) is higher than eleven out of the twenty-one samples of the spheroidally weathered diabase (Table 2, Fig. 2). The MIA(R) values for the W2-

D

type samples are also affected by preferential pyroxene weathering (Supplementary Fig. 2).

TE

Fluid mobile trace elements like Rb, Pb, U, and Ba do not show any significant correlation with CIA values although, variations in the concentrations of these elements in the 21 samples is

AC CE P

significant (Table 1). Concentrations of K and Rb, Nb and Ta, as well as Hf and Zr are, however, strongly correlated (Fig. 5), consistent with their similar behaviour in crystal structures as well in fluids (mobile versus immobile) during weathering.. The strong correlation between Th/Nb and Hf/Nb (Fig. 5) is consistent with a similar extent of mobility of Th and Hf during weathering of basalts (Fig. 6) (c.f., Kurtz 2000). The weathered profiles could be sites of secondary deposits and elemental concentration variations could reflect the composition of these secondary deposits. Micro-CT images of our sample suggest the presence of secondary precipitates (Fig. 1d-g) although, the exact composition of these deposits have not been characterized. In order to understand the extent of leaching and/or redistribution of a wider suite of elements during spheroidal weathering, we investigated the variability of the element concentrations after normalizing them to immobile element concentrations and then comparing them to the unweathered rock. Analyses of soils developed on Hawaiian basalts show that during weathering Nb and Ta are virtually immobile while Al, Th, Zr, Hf show indications of mobility within the 15

ACCEPTED MANUSCRIPT soil column (Kurtz et al., 2000). Following Kurtz et al. (2000), we have selected Nb as the least mobile element and all element concentrations are normalized to Nb both in the weathered samples and the un-weathered rock. The choice of the immobile element can vary from one rock

PT

to another as the order of leaching of different elements during weathering is controlled by the mineralogical sites of the elements in the parent rock (Nesbitt and Wilson, 1992). For example,

RI

Ma et al. (2007) used Th-normalized concentrations in a study on weathering of basalts. We

to that in the un-weathered rock (Nesbitt, 1979) as:

SC

calculated the percentage change in the Nb-normalized concentration of an element in the sample

NU

% change in ratios =100*[(Rw-Ruw)/Ruw] (6)

MA

where Rw and Ruw are the elemental ratios in the weathered sample and the un-weathered diabase, respectively. The assumption for this type of calculation is that there is no volume change between the weathering product and the un-weathered rock (Nahon and Merino, 1996).

D

Negative values in the percentage change with respect to the un-weathered rock are indicative of

TE

leaching out of the elements from that particular sample while positive values indicate the reprecipitation/addition of the element. The above calculations yield similar information to the

AC CE P

mass balance calculation of Anderson et al. (2002). The -value (Anderson et al., 2002) can be calculated by dividing equation 6 by 100. Our samples do not show monotonically increasing or decreasing trends in the percentage deviation for any of the elements (Fig. 6). This is unlike what is seen in outcrop-scale studies (e.g., Kurtz et al., 2000; Ma et al., 2007). However, large percentage deviations from the un-weathered rock composition are seen in specific samples with no relationship with their relative position in the sample. Samples 1, 6, 11, 12, 15 and 17 show large deviations in the negative direction while samples 2, 10, 13, 19 show positive deviations for most elements compared to the un-weathered rock. The samples showing these deviations are mostly W1-type samples which have CIA values less than the un-weathered rock highlighting the limitations of the traditional weathering indices. During basaltic weathering Zr, Hf, Ta, Th, Ti, and Fe are considered to be relatively immobile or conservative (c.f., Ma et al., 2007, Kurtz et al., 2000, Nesbitt and Wilson, 1992, Hill et al., 2000). In our samples, Ta/Nb shows less than 20% deviation from the un-weathered rock which could possibly be considered as the natural variation range in the rock. Titanium and Fe behave coherently for most samples suggesting a similar host oxide mineral (e.g., ilmenite) in the rock 16

ACCEPTED MANUSCRIPT while Th and the other HFSE show similar behavior. Vanadium, Cr, Co, Fe, and Mn show similar patterns in the Nb-normalized profile (Fig. 6). Large positive deviations are observed in samples 2, 10, 13, 16 and 19 for most of these redox-sensitive elements. A strong correlation is

PT

observed between Co and Mn concentrations (Fig. 5) consistent with their similar redoxsensitive geochemical behavior (Koppi et al., 1996). The variability of the Mn/Nb ratio in the

RI

weathered samples (Fig. 6) might reflect changes in redox conditions during spheroidal

SC

weathering. Vanadium and Fe are also redox sensitive elements and these two elements show similar trends (except sample 2) in the Nb-normalized plot (Fig. 6) and a relatively strong

NU

correlation is observed in figure 5). This co-variation is possibly inherent to the rock as V has a similar charge and ionic radius as Fe and could substitute Fe in magnetite and in ferromagnesian

MA

silicates in an igneous rock. Compared to the above-mentioned redox-sensitive elements, percentage deviation of U/Nb is much less and is primarily depleted with respect to the unweathered rock (Fig. 6). The efficient removal of U from the weathered samples is indicative of

D

weathering under oxidizing conditions. Uranium can also be mobilized in reducing conditions in

TE

the presence of biogenic siderophores (Kraemer et al., 2015); however, large LREE depletion, characteristic of the presence of biogenic siderophores, is not observed in our samples.

AC CE P

Magnesium and Sc concentrations are typically lower than that of the un-weathered rock, but are positively correlated (Fig. 5), suggesting preferential leaching of clinopyroxene. Sodium, K, Rb, Ba, Sr, and Pb are considered to be mobile during chemical weathering. They are primarily hosted in plagioclase in a diabase which is easily altered by fluid infiltration. In the Nbnormalized plot (Fig. 6), all of these elements show almost similar variability (except for samples 8 and 9). The Ba/Nb and Pb/Nb patterns are different from the other fluid-mobile elements and their percentage deviation from the un-weathered rock is relatively less. Barium and Pb are mostly enriched in the weathered samples compared to the un-weathered rock. Aluminum is generally considered to be an immobile element during weathering. However, our data shows that Al has a similar behavior to that of Na, K, Rb, and Sr in the Nb normalized plot. A strong positive correlation is also seen in the concentration plot of Na versus Al (Fig. 5) suggesting preferential leaching of plagioclase feldspar. All Rare Earth Elements (REE), except Eu, show consistent deviations in the Nb-normalized plot (Fig. 6). A large positive deviation for Ce is observed for sample 2, similar to that observed for 17

ACCEPTED MANUSCRIPT Mn suggesting a strong redox control. Chondrite-normalized REE pattern of the samples show LREE-enriched and flat-HREE patterns with both positive and negative Eu anomalies. The average REE pattern of the 21 samples overlaps with that of the un-weathered rock (Fig. 7b) and

PT

is consistent with published trace element patterns of diabase dike samples of similar ages from southern India (Kumar et al., 2012). During aqueous weathering, REEs can be mobilized from

RI

different rock types but LREEs are less mobile than HREEs (Babechuk et al., 2014; Condie et

SC

al., 1995; Duddy, 1980; Ma et al., 2007; Ma et al., 2006; Nesbitt, 1979; Price et al., 1991; Prudencio, 1995). The total concentration of REEs as well as that of individual LREEs like La,

NU

Sm, and Nd do not show any correlation with the extent of weathering as reflected by CIA values. HREEs like Yb and Lu, however, show a decreasing trend with increasing CIA values

MA

(not shown). However, chondrite-normalized La/Sm and Gd/Lu ratios show moderate positive correlations with CIA (Fig. 7c, d). Gd/Lu(N) of most of the samples is higher than that of the unweathered rock. These observations suggest that both LREEs and HREEs can be mobilized

D

during chemical weathering of rocks, even at the hand-specimen scale. During basalt weathering,

TE

chondrite-normalized La/Yb ratio shows a decreasing trend with increasing weathering under increased drainage conditions (Prudencio, 1995) while Sm/Nd ratios show a decreasing trend

AC CE P

with increased basalt weathering (Babechuk et al., 2014). Our samples show an opposite trend where the LREE enrichment increases with increasing CIA values (Fig. 7c). The W2-type samples show higher La/Sm(N) compared to the W1-type samples and the un-weathered rock plots in the middle of this trend (Fig. 7c). Plagioclase shows higher La/Sm(N) compared to pyroxene (e.g., Rollinson, 1993) and hence, preferential weathering of plagioclase could result in low La/Sm(N) in the residual rock as seen in the W1-type samples. Preferential weathering of pyroxene would increase the La/Sm(N) of the residual rock which is consistent with our hypothesis for the formation of the W2-type samples. Negative Eu anomaly (Eu/Eu*) can be produced during chemical weathering of granitoids due to breakdown of plagioclase (Condie et al., 1995). Progressively lower Eu/Eu* with increasing CIA has also been observed in highly weathered basalts (Babechuk et al., 2014). In basaltic melts, the partition co-efficient of Eu is highest in hornblende followed by clinopyroxene and plagioclase and is low in orthopyroxene (Rollinson, 1993). However, in plagioclase, Eu is preferentially partitioned compared to Sm and Gd resulting in a positive Eu anomaly. Our samples exhibit both positive and negative Eu anomalies (Eu/Eu*) and Eu/Eu* shows a moderately positive 18

ACCEPTED MANUSCRIPT correlation (R2 =0.57) with CIA (Fig. 7e). This variability can also be explained by preferential weathering of plagioclase resulting in lower Eu/Eu* in the residual W1-type samples and preferential weathering of pyroxene resulting in higher Eu/Eu* in the W2-type samples while the

PT

un-weathered shows intermediate Eu/Eu* due to mass balance. (Fig. 7e).

RI

Cerium has a tendency to fractionate during weathering due to its multiple redox states. Trivalent REEs are more fluid-mobile whereas tetravalent Ce is not. Under oxidizing conditions, Ce3+

SC

readily forms Ce4+ and forms CeO2 which has very limited solubility in aquatic systems (Ma et al., 2007; Marsh, 1991; Melfi, 1990; Middelburg et al., 1988; Patino, 2003). Most of the

NU

analyzed samples do not have Ce anomaly (Ce/Ce*) (Fig. 7a). Four samples, two with the lowest CIA values (W1-type samples) and two others with relatively high CIA values (W2-type

MA

samples) show positive Ce/Ce* (Fig. 7f). There is no relationship between the observed Ce/Ce* and the position of the sample in relation to the rindlets. The positive Ce/Ce* suggests oxidizing fluid conditions during weathering. It has been suggested that positive Ce/Ce* can be induced by

D

siderophore-bearing solutions even under reducing environments (Kraemer et al., 2015).

TE

However, siderophore-bearing solutions also result in LREE depletions, which is not observed in the samples of this study. Large positive Ce anomalies in lateritic profiles have been explained

AC CE P

by deposition of secondary Ce-bearing minerals (Braun et al., 1990). A similar explanation could be applicable for the four samples of our study which also show high Ce/Nb (Supplementary Fig. 2).

5.3. Sr and Nd isotopic variability

Variations in radiogenic isotopes in weathered horizons and rocks can be explained by changes in the parent/daughter ratio due to weathering and the amount of time allowed for the decay of the radioactive parent. For example, preferential enrichment of Os over Re in Fe-rich laterites resulted in less radiogenic 187Os/188Os in the laterites compared to the un-weathered Deccan basalts (Wimpenny et al., 2007). Strontium and Nd isotopic composition of natural waters reflect the composition of the minerals that are most readily weathered in rocks (Aubert et al., 2001). In the presence of organic matter, Sm/Nd ratios are readily fractionated (e.g., Chakrabarti et al., 2007) and the Nd isotopic composition of rivers do not directly reflect the composition of the bedrock (Andersson et al., 2001). Selective weathering of minerals can fractionate the Sm/Nd ratio of the bulk rock and subsequently change the Nd isotopic composition of the weathered 19

ACCEPTED MANUSCRIPT rock (Ohlander et al., 2000) provided the weathering happened long-time ago to allow for the decay of 147Sm to 143Nd. In young weathering profiles, variations in radiogenic Sr and Nd are explained by mixing with extraneous sources e.g., rainwater and aeolian dust (Chabaux et al.,

PT

2013; Ma et al., 2010).

Rubidium and Sr are both fluid mobile, Rb more so than Sr (Aiuppa et al., 2000). Hence,

RI

chemical weathering can alter the Rb/Sr ratio of rocks. The strong positive correlation between Sr/86Sr and 87Rb/86Sr for five out of the six samples analyzed for Sr isotopes and the higher

SC

87 87

Sr/86Sr of the un-weathered rock compared to these five samples (Fig. 8a) suggests that the

NU

large range in the measured 87Sr/86Sr within a hand-specimen sized sample must have been caused by weathering-induced variations in the Rb/Sr ratio. Importantly, given the long half-life of 87Rb, the mobilization of Rb and/or Sr due to weathering must have occurred a long time ago,

MA

on the order of billion years, to produce perceptible variations in 87Sr/86Sr of the sample. In addition, the above-mentioned correlation suggests that the overall system must not have been

D

affected by subsequent episodes of fluid mobilization, which would have erased the primary

TE

weathering signatures. In contrast to the other five samples analyzed for Nd and Sr isotopes, sample# 4 shows signatures of Rb-addition. The Rb/Nb ratio of sample#4 (8.36) is much higher

AC CE P

than the un-weathered rock (7.37) while the other five samples show lower Rb/Nb (5.68-7.13) consistent with Rb-loss (Table 3). Neodymium is a moderately mobile element during weathering (e.g., Chakrabarti et al., 2007; Ma et al., 2010; Nesbitt and Markovics, 1988; Nesbitt and Wilson, 1992). Variation in Nd isotope composition in weathering profiles developed over a 45 Ma old quartz diorite intrusion has been explained by mixing with aeolian dust (Chabaux et al., 2013) while in the case of Neogene basalts, the reason for a positive correlation between Nd and Hf isotopes remains unclear (Ma et al., 2010). We report large variations in Nd(0), from -8.47 to -14.26, in a hand-specimen-sized spheroidally weathered ~2.37 Ga old diabase indicating finescale mobilization of the REEs due to spheroidal weathering. The Nd(0) of the un-weathered rock shows an intermediate value of -10.51. Nd(0) values show a strong positive correlation with 147

Sm/144Nd when sample# 4 is excluded (Fig. 8b) suggesting that the Nd isotopic variability was

most likely caused by a single REE mobilization event. Given the long half-life of 147Sm, the mobilization must have occurred > 1 Ga ago, which is consistent with the Sr isotopic data for the same samples. The un-weathered rock plots within this trend with two W1-type samples showing higher Sm/Nd and more radiogenic Nd(0) and three W2-type samples showing lower Sm/Nd and 20

ACCEPTED MANUSCRIPT less radiogenic Nd(0) . The Nd(0) and Sm/Nd of the W1-type and W-2 type samples are consistent with preferential plagioclase and pyroxene weathering, respectively. The relative high Sm/Nd and Nd(0) of the W1-type samples are consistent with a pyroxene-dominated signature in the

PT

residual rock (plagioclase weathering) while the relatively low Sm/Nd and Nd(0) in the W2-type samples is indicative of a plagioclase dominated signature in the residual weathered rock

RI

(pyroxene weathering). The relative position of the un-weathered rock between the W1 and W2-

SC

type samples is consistent with mass balance considerations.

The 87Sr/86Sr and Nd(0) of the samples are, however, not correlated. This is consistent with the

NU

expectation that the impact of weathering on the Rb/Sr ratio and Sm/Nd ratio is different. The Rb/Sr ratio was affected by fluid-mobilization of Rb while the Sm/Nd ratio was affected by

MA

differential weathering of plagioclase and pyroxene.

D

5.4. Estimating the timing of spheroidal weathering

TE

Weathering advance rates can vary by one to two orders of magnitude depending on lithology, fluid composition, climate as well as biological activity (Navarre-Sitchler and Brantley, 2007).

AC CE P

Short-lived radionuclides of the U and Th decay series suggest that the weathering front for quartz diorites in the Rio Icaco watershed could have progressed at rates of 45 m/Ma while modeling results for the same rock suggest that saprolite formation over 20 rindlets could have formed in as little as 5000 years (Chabaux et al., 2013; Fletcher et al., 2006). The most intense phases of rock weathering maybe associated with certain geologic events e.g., continental breakup, volcanism, etc. It is hard to date a particular weathering episode unless the preservation of the weathered rock provides clues. For example, the red boles (Wilkins et al., 1994) found in the Deccan basalts (inter-trappean sediments) are synchronous with the eruption of the Deccan basalts ~65 Ma ago. If the variation in the Rb/Sr and Sm/Nd ratios and the Sr and Nd isotopic compositions in the samples of the spheroidally weathered diabase are due to a single event of weathering, the initial Sr and Nd isotopic compositions of the samples should converge to a common value at the time of initiation of weathering which, should be identical to the isotopic composition of the unweathered rock at the same time. The percentage relative standard deviation (%RSD) of the initial isotopic composition of the samples at any time would indicate the extent of convergence 21

ACCEPTED MANUSCRIPT and hence, the lower the value of %RSD, the closer it is to the time of the weathering event. The initial Nd and Sr isotope ratios were calculated at different ages back to the formation age of the rock approximately 2.37 Ga ago (French and Heaman, 2010; Halls et al., 2007; Kumar et al.,

PT

2012).

RI

When all 6 samples are considered, RSD (%) of the calculated initial 143Nd/144Nd varies with a concave upward pattern (Fig. 8d). The lowest RSD (%) is reached at 1.1 Ga before present. If

SC

sample# 4 is excluded, the lowest value of RSD (%) (0.00548) is reached at 1.2 Ga before present (Fig. 8d). It must be noted that the calculated values of standard deviation for the initial Nd/144Nd does not reach zero. However, the 28 ppm (1SD) variation for 143Nd/144Nd

NU

143

(sample#4 excluded) could be explained by sample heterogeneity in the un-weathered sample. In

MA

addition, the average initial Nd(0) of all 6 samples at 1.2 Ga ago overlaps with the initial Nd(0) of the un-weathered rock to within 0.04 epsilon units (Table 3). This indicates that the weathering event which fractionated the Sm/Nd ratio leading to different Nd isotopic compositions of the

D

spheroidally weathered samples took place ~1.2 Ga ago. When all 6 samples are considered,

TE

RSD (%) of the calculated initial 87Sr/86Sr decreases linearly with time. If sample# 4 is excluded,

AC CE P

the RSD (%) shows a concave upward pattern and reaches a minimum value of 0.07228 at 1.3 Ga before present (Fig. 8c). This suggests that the episode of weathering which fractionated the Rb/Sr ratio took place ~1.3 Ga ago, similar to the timing of the weathering event that fractionated the Sm/Nd ratio of the rock. However, the average initial

87

Sr/86Sr (0.704692, 1SD

500 ppm) of the weathered samples (excluding sample #4) at 1.3 Ga ago is distinctly less radiogenic compared to the initial 87Sr/86Sr of the un-weathered rock at 1.3 Ga (0.709714). The Rb-Sr system is much more prone to disturbance compared to the Sm-Nd system given the relatively coherent behavior of the REEs (Sm, Nd) compared to Rb and Sr. The Nd and Sr isotopic data suggest that Nd and Sr isotopes were affected by a major episode of weathering that took place 1.2-1.3 Ga ago; the Sr isotopic composition was affected by additional, possibly older, episodes of mobilization of Rb and/or Sr. Sample# 4 possibly reflects such a weathering event. The timing of spheroidal weathering coincides with the final breakup of the supercontinent Columbia, as inferred from wide-spread 1.2-1.3 Ga old dike swarms distributed globally (Hou et al., 2008). In India, there is wide-spread evidence of alkaline volcanism in the Mesoproterozoic between ~1.5 to ~1.3 Ga ago (c.f., Upadhyay et al., 2006; Bhowmik et al., 2012). Such wide22

ACCEPTED MANUSCRIPT spread volcanism could have resulted in increased chemical weathering as observed in the samples of this study. However, in the absence of additional evidence, it is difficult to test such a

PT

hypothesis. 6. Conclusions

RI

High-resolution sampling and analyses of a ~2.37 Ga old spheroidally weathered diabase shows

SC

significant elemental and Nd, Sr isotopic variability in hand-specimen scale. This is in spite of the relatively low CIA values of the samples suggesting that the sample is not extensively

NU

weathered. X-Ray micro CT imaging, documents the pervasive nature of fractures of at least of two generations inside the rock whereas only a single fracture is visible to the naked eye. The

MA

older fracture is filled by a secondary deposit which, along with the more recent fracture, crosscuts the dominant silicate minerals of the rock as well as the exsolved Fe-oxide. The elemental concentrations of the un-weathered rock typically overlap with that of the

D

weathered samples which show a significant range (%RSD ranging from 10.2-41.8). We

TE

investigated whether the variations in major and trace element concentrations in the 21 samples correlate with chemical indices of alteration like CIA. In addition, to understand the extent of

AC CE P

element mobility during spheroidal weathering we calculated the percentage deviation of Nbnormalized elemental concentrations in the samples from that in the un-weathered rock. Broadly, the variations in elemental concentrations and ratios are explained in terms of fluid mobility of the element, selective weathering of plagioclase and pyroxene in the diabase as well as ambient environmental conditions. The results show that values of weathering indices can be affected by preferential weathering of minerals and might not accurately reflect the intensity of weathering at a hand-specimen scale. Considerable Nd and Sr isotopic variability is observed in the samples and is explained in terms of weathering-related fractionation of the Sm/Nd and Rb/Sr ratios which must have happened long time ago to allow for radiogenic decay of the long-lived isotopes of 87Rb and 147Sm to 87Sr and 143Nd, respectively. Initial isotope ratio calculations suggest that the peak episode(s) of weathering occurred about 1.2-1.3 Ga ago. This is the first such attempt to determine the timing of a weathering event in rocks using long-lived radioactive isotopes. Our estimated time of spheroidal weathering coincides with the timing of the final breakup of the supercontinent 23

ACCEPTED MANUSCRIPT Columbia and follows wide-spread alkaline volcanism in the Indian subcontinent although, in the absence of additional evidence, it is difficult to correlate the weathering episode with such wide-

PT

spread geologic events. Acknowledgements

RI

Funding for the TIMS and ICPMS laboratories was provided by IISc. Funding for the clean

SC

laboratory was obtained from DST and MoES. Dr. R. Srinivasan is thanked for guidance during fieldwork. Ratikanta Sikdar helped with trace element analyses. RC would like to thank Prof.

NU

Bikramjit Basu for help with the micro-CT analyses and two anonymous reviewers for their comments and suggestions.

MA

Figure Captions

Figure 1. (a) A slice of a spheroidally weathered diabase used in the present study before micro-

D

sampling. (b) Post micro-drilling image of the sample showing the sampling locations and

TE

sample numbers. (c) The samples are numbered based on their relative average radial distance from a reference point placed towards the core of the sample (see Supplementary Table 1).

AC CE P

Sample numbers 1 and 21 represent the outermost and innermost shells, respectively. (d) A small portion of this sample, represented by the white box in figure (a) was imaged using a X-Ray microscope; three dominant minerals, plagioclase (Pl), pyroxene (Px) and a Fe-bearing phase (Fe-rich) with different shades of gray are seen. The most recent fracture (dark ribbon) is located in the central portion along with an earlier light gray ribbon-like deposit. The pervasive nature of the weathering is seen in the XY section (e) while the cross-cutting relationship of the fracture (dark), earlier deposit (ribbon-like lighter shade) and the minerals are observed in the XZ (f) and YZ (g) sections. Figure 2. Variations in CIA, CIW, MIA(R), MIA(O) and WIP values of the different samples of the spheroidally-weathered diabase from the rim (sample 1) to progressively inner samples (see Fig. 1 and Supplementary Table 1). WIP values show an overall increasing trend from the outermost sample inward. CIA, CIW, MIA(R), and MIA(O) values do not show any spatial trend and their variations are similar (see Supplementary Fig. 2). The vertical lines show the values of the different indices for the un-weathered rock (UW).

24

ACCEPTED MANUSCRIPT Figure 3. Concentrations of Ca (R2 =0.74) and Li (R2 = 0.63) decrease from the reference point (core) to the rim (See Fig. 1c). The horizontal lines show the Ca and Li concentrations for the unweathered rock (UW). The weathered samples show both higher and lower values compared to

PT

the un-weathered rock.

RI

Figure 4. When plotted against CIA values of the samples, concentrations of major elements like Na and Al are positively correlated while Mg and Fe concentrations are negatively correlated.

SC

Among the trace elements, Sr is positively correlated while Sc is negatively correlated with the CIA values. Na/Ca increases while Mg/Al, Mg/Na, Mg/Ca, Fe/Al and Sc/Sr decrease with

NU

increasing CIA values indicating preferential weathering of plagioclase and pyroxene. The gray square indicates the un-weathered rock (UW), open circles indicate W1-type samples with CIA

MA

less than UW and filled circles indicate W2-type samples with CIA more than UW. See text for more details.

D

Figure 5. Strong positive correlations are observed between concentrations of Na and Al, Na and

TE

Sr, Rb and K, Li and Ca, Co and Mn, Hf and Zr, Ta and Nb as well as between Th/Nb and Hf/Nb. Moderate correlations are observed between concentrations of Sc and Mg, Mg and Ca,

AC CE P

and V and Fe. The samples show both higher and lower values compared to the un-weathered rock (UW, gray square). Symbols are same as in Figure 4. Figure 6. Percentage deviation of Nb-normalized multi-element concentrations in the samples compared to the un-weathered rock plotted versus the sample number (Fig. 1c). By definition, the value of zero on the x-axis shown by the vertical line defines the composition of the unweathered rock. The vertical gray bar represents a + 20% deviation in composition from the unweathered rock. Samples 1, 6, 11, 12, 15 and 17 show deviations in the negative direction (depletion compared to the un-weathered rock) while samples 2, 10, 13, 19 show enrichments for most elements (positive deviation) compared to the un-weathered rock. Figure 7. (a) Chondrite-normalized REE pattern of the samples show LREE-enriched and flatHREE patterns with both positive and negative Eu anomalies. (b) The average REE pattern of the 21 samples overlaps with that of the un-weathered rock and does not show any significant Eu anomaly. The REE concentration pattern is consistent with published data for the southern Indian dikes of this age. Chondrite normalized La/Sm (c) as well as Gd/Lu (d) ratios increase with 25

ACCEPTED MANUSCRIPT increasing CIA values of the samples. Gd/Lu(N) of the un-weathered rock (UW, gray square) is lower than most of the samples unlike La/Sm(N) and other elemental concentrations and ratios where the un-weathered rock plots within the cluster of the weathered samples. (e) The samples

PT

show both positive and negative Europium anomaly (Eu/Eu*) which is positively correlated with CIA values of the samples. (f) Cerium anomaly (Ce/Ce*) values show less variation and do not

RI

show any correlation with CIA values of the samples. Symbols are same as in Figure 4.

SC

Figure 8. Measured 87Sr/86Sr shows a strong positive correlation with 87Rb/86Sr (a) and measured Nd(0) of the samples also show a strong positive correlation with 147Sm/144Nd (b) when sample#

NU

4 is not considered. Also shown are the Sr and Nd isotopic compositions of the un-weathered rock (UW). Symbols in (a) and (b) are same as in Figure 4. Initial Sr and Nd isotopic

MA

compositions were calculated at different times before present and at every time, the percentage relative standard deviation was calculated as shown in (c) and (d). When sample# 4 is not considered, the minimum value of %RSD for Sr and Nd isotopes is attained at 1.2-1.3 Ga before

AC CE P

References

TE

D

present which possibly reflects the timing of the peak weathering event. See text for details.

Aiuppa, A. et al., 2000. Mobility and fluxes of major elements, minor and trace metals during basalt weathering and groundwater transport at Mt. Etna volcano (Sicily). Geochimica Et Cosmochimica Acta, 64(11): 1827–1841. Anderson, S. P., Dietrich, W. E., Brimhall Jr., G. H., 2002. Weathering profiles, mass-balance analysis, and rates of solute loss: Linkages between weathering and erosion in a small, steep catchment. Geological Society of America Bulletin, 114(9): 1143-1158. Andersson, P.S., Dahlqvist, R., Ingri, J., Gustafsson, O., 2001. The isotopic composition of Nd in a boreal river: A reflection of selective weathering and colloidal transport. Geochimica Et Cosmochimica Acta, 65(4): 521-527. Aubert, D., Stille, P., Probst, A., 2001. REE fractionation during granite weathering and removal by waters and suspended loads: Sr and Nd isotopic evidence. Geochimica Et Cosmochimica Acta, 65(3): 387–406.

26

ACCEPTED MANUSCRIPT Babechuk, M.G., Widdowson, M., Kamber, B.S., 2014. Quantifying chemical weathering intensity and trace element release from two contrasting basalt profiles, Deccan Traps, India. Chemical Geology, 363: 56-75.

PT

Berner, R.A., 1995. Chemical weathering and its effect on atmospheric CO2 and climate. Chemical Weathering Rates of Silicate Minerals, 31: 565-583.

RI

Berner, R.A., Lasaga, A.C., 1989. Modeling the Geochemical Carbon-Cycle. Scientific

SC

American, 260(3): 74-81.

Bhowmik, S. K., Chattopadhyay, A., Gupta, S., Dasgupta, S., 2012. Proterozoic Tectonics: An

NU

Indian Perspective, Proc. Indian. Natn. Sci. Acad, 78(3): 385-391. Brantley, S.L., Buss, H.L., Lebedeva, M., Fletcher, R.C., Ma, L., 2011. Investigating the

MA

complex interface where bedrock transforms to regolith. Applied Geochemistry, 26: S12S15.

Braun, J-J., Pagel, M., Muller, J-P, Bilong, P., Michard, A., Guillet, B., 1990. Cerium anomalies

D

in lateritic profiles. Geochimica Et Cosmochimica Acta, 54: 781-795.

TE

Buss, H.L., Sak, P.B., Webb, S.M., Brantley, S.L., 2008. Weathering of the Rio Blanco quartz diorite, Luquillo Mountains, Puerto Rico: Coupling oxidation, dissolution, and fracturing.

AC CE P

Geochimica Et Cosmochimica Acta, 72: 4488-4507. Chabaux, F. et al., 2013. Regolith formation rate from U-series nuclides: Implications from the study of a spheroidal weathering profile in the Rio Icacos watershed (Puerto Rico). Geochimica Et Cosmochimica Acta, 100: 73-95. Chakrabarti, R., Abanda, P.A., Hannigan, R.E., Basu, A.R., 2007. Effects of diagenesis on the Nd-isotopic composition of black shales from the 420 Ma Utica Shale Magnafacies. Chemical Geology, 244(1-2): 221-231. Chesworth, W., Dejou, J., Larroque, P., 1981. The Weathering of Basalt and Relative Mobilities of the Major Elements at Belbex, France. Geochimica Et Cosmochimica Acta, 45(7): 1235-1243. Condie, K.C., Dengate, J., Cullers, R.L., 1995. Behavior of rare earth elements in a paleoweathering profile on granodiorite in the Front Range, Colorado, USA. Geochimica Et Cosmochimica Acta, 59: 279 – 294.

27

ACCEPTED MANUSCRIPT De Souza, G.F., Reynolds, B.C., Bourdon, B., 2007. Evidence for stable strontium isotope fractionation during chemical weathering. Geochimica et Cosmochimica Acta, 71(15): A220. weathering profile. Chemical Geology, 30: 363 – 381.

PT

Duddy, I.R., 1980. Redistribution and fractionation of rare earth elements and other elements in a

RI

Duzgoren-Aydin, N.S., Aydin, A., Malpas, J., 2002. Re-assessment of chemical weathering

SC

indices: case study on pyroclastic rocks of Hong Kong. Engineering Geology, 63: 99119.

NU

Eggleton, R.A., Foundoulis, C., Varkevisser, D., 1987. Weathering of basalt: changes in rock chemistry and mineralogy. Clays and Clay Minerals, 35(3): 161 – 169.

MA

Fletcher, R.C., Buss, H.L., Brantley, S.L., 2006. A spheroidal weathering model coupling porewater chemistry to soil thickness during steady-state denudation. Earth and Planetary Science Letters, 244: 444–457.

D

French, J.E., Heaman, L.M., 2010. Precise U-Pb dating of Paleoproterozoic mafic dyke swarms

TE

of the Dharwar craton, India: Implications for the existence of the Neoarchean supercraton Sclavia. Precambrian Research, 183(3): 416-441.

AC CE P

Gardner, L.R., Kheoruenromme, I., Chen, H.S., 1981. Geochemistry and mineralogy of an unusual diabase saprolite near Columbia, South Carolina. Clays Clay Miner., 29: 184 – 190.

Halls, H.C., Kumar, A., Srinivasan, R., Hamilton, M.A., 2007. Paleomagnetism and U-Pb geochronology of easterly trending dykes in the Dharwar craton, India: feldspar clouding, radiating dyke swarms and the position of India at 2.37 Ga. Precambrian Research, 155(1-2): 47-68. Halls, H.C., Zhang, B., 1995. Tectonic implications of clouded feldspar in Proterozoic mafic dyke swarms. Geological Society of India Memoir, 33: 65-80. Harnois, L., 1988. The CIW index: a new chemical index of weathering. Sediment. Geol., 55: 319–322. Hill, I.G., Worden, R.H., Meighan, I.G., 2000. Yttrium: The immobility – mobility transition during basaltic weathering. Geology, 28(10): 923 – 926.

28

ACCEPTED MANUSCRIPT Hou, G., Santosh, M., Qian, X., Lister, G. S., Li, J., 2008. Tectonic constraints on 1.3-1.2 Ga final breakup of Columbia supercontinent from a giant radiating dyke swarm. Gondwana Research, 14: 561-566.

PT

Jamtveit, B., Putnis, C.V., Malthe-Sorenssen, A., 2009. Reaction induced fracturing during replacement process. Contributions to Mineralogy and Petrology, 157: 127-133.

RI

Jamtveit, B., Kobchenko, M., Austrheim, H., Malthe-Sorenssen, A., Royne, A., Svensen, H.,

SC

2011. Porosity evolution and crystallization-driven fragmentation during weathering of andesite. Journal of Geophysical Research, 116: B12204, doi:10.1029/2011JB008649.

NU

Koppi, A.J. et al., 1996. Rare earth element trends and cerium-uranium-manganese associations in weathered rock from Koongarra, northern territory, Australia. Geochimica Et

MA

Cosmochimica Acta, 60(10): 1695-1707.

Kraemer, D., Kopf, S., Bau, M., 2015. Oxidative mobilization of cerium and uranium and enhanced release of "immobile" high field strength elements from igneous rocks in the

D

presence of biogenic siderophore desferrioxamine B. Geochimica Et Cosmochimica

TE

Acta, 165: 263-279.

Kumar, A., Hamilton, M.A., Halls, H.C., 2012. A Paleoproterozoic giant radiating dyke swarm

AC CE P

in the Dharwar Craton, southern India. Geochemistry Geophysics Geosystems, 13. Kurtz , A.C., Derry, L. A., Chadwick, O. A., Alfano, M. J. , 2000. Refractory element mobility in volcanic soils. Geology, 28(8): 683 – 686. Ma, J.-L., Wei, G.-J., Xu, Y.-G., Long, W.-G., Sun, W.-D., 2007. Mobilization and redistribution of major and trace elements during extreme weathering of basalt in Hainan Island, South China. Geochimica et Cosmochimica Acta, 71(13): 3223-3237. Ma, J., Wei, G., Xu, Y., Long, W., 2010. Variations of Sr–Nd–Hf isotopic systematics in basalt during intensive weathering. Chemical Geology, 269(3–4): 376-385. Ma, J.L., Wei, G.J., Xu, Y.G., Long, W.G., 2006. Behavior of major and trace elements in an extreme weathering profile on basalt in Hainan Island, South China. Geochimica et Cosmochimica Acta, 70(18, Supplement): A379. Marsh, J.S., 1991. REE fractionation and Ce anomalies in weathered Karoo dolerite. Chem. Geol., 90: 189 – 194.

29

ACCEPTED MANUSCRIPT Melfi, A.J., Figueiredo, A.M., Kronberg, B.I., Dohert, W.D., Marques, L.S., 1990. REE mobilities during incipient weathering of volcanic rocks of the Parana basin, Brazil. Chemical Geology 84: 375 – 376.

PT

Middelburg, J.J., Vanderweijden, C.H., Woittiez, J.R.W., 1988. Chemical Processes Affecting the Mobility of Major, Minor and Trace-Elements during Weathering of Granitic-Rocks.

RI

Chemical Geology, 68(3-4): 253-273.

SC

Nahon, D., Merino, E., 1996. Pseudomorphic replacement versus dilation in laterites: Petrographic evidence, mechanisms, and consequences for modelling. Journal of

NU

Geochemical Exploration, 57(1-3): 217-225.

Navarre-Sitchler, A., Brantley, S., 2007. Basalt weathering across scales. Earth and Planetary

MA

Science Letters, 261: 321-334.

Nesbitt, H.W., 1979. Mobility and fractionation of rare earth elements during weathering of a granodiorite. Nature, 279: 206 – 210.

D

Nesbitt, H.W., Markovics, G., 1988. Weathering of granodioritic crust, long-term storage of

TE

elements in weathering profiles, and petrogenesis of siliciclastic sediments. Geochimica et Cosmochimica Acta, 61(8): 1653–1670.

AC CE P

Nesbitt, H.W., Markovics, G., Price, R.C., 1980. Chemical processes affecting alkalis and alkaline earths during continental weathering. Geochimica Et Cosmochimica Acta, 44: 1659 – 1666.

Nesbitt, H.W., Wilson, R.E., 1992. Recent chemical weathering of basalts. American Journal of Science, 292: 740 – 777.

Nesbitt, H.W., Young, G.M., 1982. Early Proterozoic climates and plate motions inferred from major element chemistry of lutites. Nature, 299: 715-717. Ohlander, B., Ingri, J., Land, M., Schoberg, H., 2000. Change of Sm-Nd isotope composition during weathering of till. Geochimica Et Cosmochimica Acta, 64(5): 813-820. Ohta, T., Arai, H., 2007. Statistical empirical index of chemical weathering in igneous rocks: A new tool for evaluating the degree of weathering. Chemical Geology, 240: 280-297. Ollier, C.D., 1971. Causes of spheroidal weathering. Earth-Sci. Rev., 7: 127 – 141. Parker, A., 1970. An index of weathering for silicate rocks. Geol. Mag. , 107: 501 – 504.

30

ACCEPTED MANUSCRIPT Patino, L.C., Velbel, M.A., Price, J.R., Wade, J.A., 2003. Trace element mobility during spheroidal weathering of basalts and andesites in Hawaii and Guatemala. Chemical Geology, 202((3-4)): 343–364.

PT

Pichamuthu, C.S., 1951. Clouded feldspars in basic dykes of charnockitic areas. Current Science, 18: 230-231.

RI

Pichamuthu, C.S., 1959. The significance of clouded plagioclase in the basic dykes of Mysore

SC

State, India. Journal of Geological Society of India, 1: 68-79.

Price, J.R., Velbel, M.A., 2003. Chemical weathering indices applied to weathering profiles

NU

developed on heterogeneous felsic metamorphic parent rocks. Chemical Geology, 202: 397-416.

MA

Price, R.C., Gray, C.M., Wilson, R.E., Frey, F.A., Taylor, S.R., 1991. The effects of weathering on rare-earth element, Y and Ba abundances in Tertiary basalts from southeastern Australia. Chemical Geology, 93: 245 – 265.

D

Prudencio, M.I., Gouveia, M.A., Squeira-Braga, M.A., 1995. REE distribution in present-day 239 – 248.

TE

and ancient surface environments ofbasaltic rocks (Central Portugal). Clay Miner., 30:

AC CE P

Rollinson, H.R., 1993. Using geochemical data: evaluation, presentation, interpretation. Longman Group UK Limited. Royne, A., Jamtveit, B., Mathiesen, J., Malthe-Sorenssen, A., 2008. Controls on rock weathering rates by reaction-induced hierarchical fracturing. Earth and Planetary Science Letters, 275: 364–369.

Taylor, A.S., Lasaga, A.C., 1999. The role of basalt weathering in the Sr isotope budget of the oceans. Chemical Geology, 161(1-3): 199-214. Upadhyay, D., Raith, M. M., Mezger, K., Bhattacharya, A., Kinny, P. D., 2006. Mesoproterozoic rifting and Pan-African continental collision in SE India: evidence from the Khariar alkaline complex. Contrib. Mineral. Petrol., 151: 434-456. Velbel, M.A., 2009. Dissolution of olivine during natural weathering. Geochimica Et Cosmochimica Acta, 73(20): 6098-6113. Whitfield, M., 1982. Sea-Water under Extreme Pressure. Nature, 297(5867): 538-538. Wilkins, A., Subbarao, K. V., Ingram, G., Walsh, J. N., 1994. Weathering regimes within the Deccan basalts. In Subbarao, K. V. (Ed.) Volcanism, Wiley, New Delhi, pp. 217-231. 31

ACCEPTED MANUSCRIPT Wimpenny, J. et al., 2007. Rhenium and osmium isotope and elemental behaviour accompanying laterite formation in the Deccan region of India. Earth and Planetary Science Letters,

AC CE P

TE

D

MA

NU

SC

RI

PT

261(1-2): 239–258.

32

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Figure 1. (a) A slice of a spheroidally weathered diabase used in the present study before micro-

sampling. (b) Post micro-drilling image of the sample showing the sampling locations and sample numbers. (c) The samples are numbered based on their relative average radial distance from a reference point placed towards the core of the sample (see Supplementary Table 1). Sample numbers 1 and 21 represent the outermost and innermost shells, respectively. (d) A small portion of this sample, represented by the white box in figure (a) was imaged using a X-Ray microscope; three dominant minerals, plagioclase (Pl), pyroxene (Px) and a Fe-bearing phase (Fe-rich) with different shades of gray are seen. The most recent fracture (dark ribbon) is located in the central portion along with an earlier light gray ribbon-like deposit. The pervasive nature of the weathering is seen in the XY section (e) while the cross-cutting relationship of the fracture (dark), earlier deposit (ribbon-like lighter shade) and the minerals are observed in the XZ (f) and YZ (g) sections.

33

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC CE P

Figure 2. Variations in CIA, CIW, MIA(R), MIA(O) and WIP values of the different samples of the spheroidally-weathered diabase from the rim (sample 1) to progressively inner samples (see Fig. 1 and Supplementary Table 1). WIP values show an overall increasing trend from the outermost sample inward. CIA, CIW, MIA(R), and MIA(O) values do not show any spatial trend and their variations are similar (see Supplementary Fig. 2). The vertical lines show the values of the different indices for the un-weathered rock (UW).

34

2

2

SC

RI

PT

ACCEPTED MANUSCRIPT

AC CE P

TE

D

MA

NU

Figure 3. Concentrations of Ca (R =0.74) and Li (R = 0.63) decrease from the reference point (core) to the rim (See Fig. 1c). The horizontal lines show the Ca and Li concentrations for the unweathered rock (UW). The weathered samples show both higher and lower values compared to the un-weathered rock.

35

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Figure 4. When plotted against CIA values of the samples, concentrations of major elements like Na and Al are positively correlated while Mg and Fe concentrations are negatively correlated. Among the trace elements, Sr is positively correlated while Sc is negatively correlated with the CIA values. Na/Ca increases while Mg/Al, Mg/Na, Mg/Ca, Fe/Al and Sc/Sr decrease with increasing CIA values indicating preferential weathering of plagioclase and pyroxene. The gray square indicates the un-weathered rock (UW), open circles indicate W1-type samples with CIA less than UW and filled circles indicate W2-type samples with CIA more than UW. See text for more details.

36

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Figure 5. Strong positive correlations are observed between concentrations of Na and Al, Na and Sr, Rb and K, Li and Ca, Co and Mn, Hf and Zr, Ta and Nb as well as between Th/Nb and Hf/Nb. Moderate correlations are observed between concentrations of Sc and Mg, Mg and Ca, and V and Fe. The samples show both higher and lower values compared to the un-weathered rock (UW, gray square). Symbols are same as in Figure 4.

37

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Figure 6. Percentage deviation of Nb-normalized multi-element concentrations in the samples compared to the un-weathered rock plotted versus the sample number (Fig. 1c). By definition, the value of zero on the x-axis shown by the vertical line defines the composition of the unweathered rock. The vertical gray bar represents a + 20% deviation in composition from the unweathered rock. Samples 1, 6, 11, 12, 15 and 17 show deviations in the negative direction (depletion compared to the un-weathered rock) while samples 2, 10, 13, 19 show enrichments for most elements (positive deviation) compared to the un-weathered rock.

38

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

Figure 7. (a) Chondrite-normalized REE pattern of the samples show LREE-enriched and flatHREE patterns with both positive and negative Eu anomalies. (b) The average REE pattern of the 21 samples overlaps with that of the un-weathered rock and does not show any significant Eu anomaly. The REE concentration pattern is consistent with published data for the southern Indian dykes of this age. Chondrite normalized La/Sm (c) as well as Gd/Lu (d) ratios increase with increasing CIA values of the samples. Gd/Lu(N) of the un-weathered rock (UW, gray square) is lower than most of the samples unlike La/Sm(N) and other elemental concentrations and ratios where the un-weathered rock plots within the cluster of the weathered samples. (e) The samples show both positive and negative Europium anomaly (Eu/Eu*) which is positively correlated with CIA values of the samples. (f) Cerium anomaly (Ce/Ce*) values show less variation and do not show any correlation with CIA values of the samples. Symbols are same as in Figure 4.

39

AC CE P

TE

D

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

87

86

87

86

Figure 8. Measured Sr/ Sr shows a strong positive correlation with Rb/ Sr (a) and measured 147 144 εNd(0) of the samples also show a strong positive correlation with Sm/ Nd (b) when sample# 4 is not considered. Also shown are the Sr and Nd isotopic compositions of the un-weathered rock (UW). Symbols in (a) and (b) are same as in Figure 4. Initial Sr and Nd isotopic compositions were calculated at different times before present and at every time, the percentage relative standard deviation was calculated as shown in (c) and (d). When sample# 4 is not considered, the minimum value of %RSD for Sr and Nd isotopes is attained at 1.2-1.3 Ga before present which possibly reflects the timing of the peak weathering event. See text for details.

40

ACCEPTED MANUSCRIPT

21

2.1

UW

1.8 7

0.6 4

7

1.9 1.8 4 1.8 4

8

2.1

9

13

2.5 1.4 3 1.6 2 1.8 0 1.8 1

14

2.3

15

1.9

16

2.0 1.6 6

5 6

10 11 12

17 18

Ti

Mn

wt%

wt%

6.2

3.4

6.6

13. 3

1.16

7.5

2.3

6.1

8.4

0.23

6.9

2.6

6.0

9.6

0.58

6.4

3.1

6.5

9.1

0.32

7.7

2.3

6.2

0.49

7.5

3.2

6.8

7.3

3.1

6.7

7.2 10. 5 11. 5

8.4

6.8

9.4

0.60

9.6

2.6 1.7 4

6.6

0.70

5.8

3.4

6.6

6.6

3.5

6.9

7.3

3.3

6.9

7.4

3.8

7.7

8.0 13. 8 13. 8 12. 0 11. 4

9.2

2.4

6.9

0.50

7.5

3.4

7.3

8.1

4.0

8.1

7.0

4.0

7.6

8.8

4.1

8.6

6.1

4.8

8.0

8.0

3.9

8.0

8.1

3.7

7.7

7.9 13. 9 13. 6 10. 2 13. 0 13. 0 13. 7 16. 3

7.3

3.3

7.0

9.9

Sc pp m

7.9

37

7.0

29

V pp m

8.3

36

8.0

31

7.5

25

6.9

37

8.5

39

7.8

38

8.4

41

7.9

32

9.2 10. 4

35

8.2 10. 8

44

9.2 10. 7 10. 3

47

1.23

0.22 0.17 3 0.17 8

33

56 4 14 3 32 8 21 0 25 4 33 5 35 9 27 9 23 9 45 0 28 0 36 1 22 9 16 8 55 2 47 0 27 2 39 6 46 7 42 7 60 7

0.68

0.15 1

9.8

42

36 8

0.78

0.16 6 0.18 2 0.13 3 0.13 5 0.10 6 0.16 8 0.14 6 0.13 2 0.09 2 0.17 4 0.17 0 0.14 9 0.15 5 0.13 8 0.16 7 0.18 6 0.16 4 0.17 9

Li pp m

6.5 7.1

27

8.3

36

7.1

0.63

0.73 0.94 0.78 0.19 0

1.34 0.80 0.63 0.80 0.68 0.86

28 33

41

37

33

34

Cr pp m

Co pp m

Rb pp m

Sr pp m

Y pp m

Zr pp m

Nb pp m

96 12 3 11 1

26

87

4.7

27

77

3.1

21

58

2.9

29

85

3.4

24

83 12 0 13 1

3.3

27 17. 2

90

4.3

58

3.8

74 13 6 10 4

2.9

62 10 2 11 3

2.1

3.6

PT

2.0

4

Fe wt %

RI

20

3

Ca wt %

SC

19

2.2 1.4 9

2

Mg wt %

NU

1.5 2 1.8 6 1.7 4 1.5 6

0.4 2 0.5 3 0.4 7 0.5 5 0.6 2 0.7 9 0.7 1 0.6 5 0.4 7 0.4 1 0.5 5 0.5 9 0.5 4 0.6 7 0.5 3 0.5 5 0.7 5 0.8 3 0.4 5 0.5 3 0.5 8

1

Al wt %

MA

K wt %

TE

Na wt %

AC CE P

Samp le

D

Table 1: Major, minor and trace element concentrations of the spheroidally-weathered diabase samples and the un-weathered rock (UW).

45

63

23

25

83

25

37

51

20

35

53

28

22

42

30

34

66

44

32

62

37

27

54

24

45

30 15. 5

96 12 1 11 9 10 8 13 0 14 7

54

83

21

92

29

57

76

31

34

30

57

33

59 18. 1

63

23

66

30

35

71

30

40

78

24

97 11 5 10 5 13 9 11 0 11 6

40

59

42

34

52

75

41

99 12 1

31

91 12 8 13 1

59

98

22

30

83

2.7

46

69

23

29

71

26

32

85 11 4

4.1

41

91 11 6 12 2

41

49

32

13 0

32

10 0

33 33

30 25 28 27 33

6.6 4.1

9.8 5.3

4.2 6.6

8.1 5.3

4.6

4.3

ACCEPTED MANUSCRIPT

Ce pp m

Pr pp m

Nd pp m

S m pp m

Eu pp m

Gd pp m

Tb pp m

Dy pp m

Ho pp m

Er pp m

21

16 7 37 4 22 8 20 7 24 3 38 5 28 9 29 6 20 9 17 4 25 2 23 5 22 4 31 6 25 9 22 8 26 8 28 9 19 9 30 1 22 9

7. 3 11 .0 7. 5 9. 2 8. 9 11 .9 10 .8 9. 5 6. 4 8. 3 11 .4 9. 7 6. 9 11 .0 7. 7 9. 7 10 .0 10 .2 8. 0 8. 2 10 .5

15 .7 29 .6 16 .5 20 .5 18 .9 26 .7 23 .4 20 .6 13 .7 20 .1 24 .6 21 .0 15 .2 25 .3 16 .9 21 .7 22 .0 23 .1 20 .8 18 .5 23 .4

2. 18 2. 99 2. 15 2. 72 2. 48 3. 34 3. 10 2. 72 1. 83 2. 41 3. 28 2. 78 2. 04 2. 97 2. 37 2. 99 2. 93 3. 15 2. 42 2. 65 3. 20

9. 5 12 .8 9. 2 11 .8 10 .5 14 .3 13 .3 11 .7 7. 8 10 .6 14 .3 11 .9 9. 0 12 .7 10 .3 12 .9 12 .7 13 .8 10 .7 11 .8 13 .8

2. 72 3. 32 2. 45 3. 23 2. 80 3. 77 3. 59 3. 12 2. 04 2. 94 3. 85 3. 23 2. 55 3. 33 2. 92 3. 62 3. 47 3. 79 3. 10 3. 30 3. 72

0. 89 1. 14 0. 93 0. 98 1. 03 1. 26 1. 11 1. 15 1. 00 0. 90 1. 05 1. 06 0. 94 1. 22 1. 03 1. 11 1. 06 1. 25 0. 94 1. 09 1. 13

3. 13 3. 99 2. 87 3. 71 3. 21 4. 52 4. 18 3. 61 2. 36 3. 57 4. 47 3. 87 3. 02 3. 96 3. 44 4. 28 4. 18 4. 31 3. 83 3. 83 4. 39

0.6 0 0.6 8 0.5 1 0.6 7 0.5 7 0.7 9 0.7 5 0.6 5 0.4 14 0.6 6 0.7 9 0.7 0 0.5 5 0.6 8 0.6 3 0.7 9 0.7 6 0.8 1 0.7 0 0.7 1 0.7 9

4. 16 4. 53 3. 48 4. 60 3. 87 5. 3 5. 2 4. 38 2. 72 4. 60 5. 4 4. 8 3. 90 4. 57 4. 46 5. 5 5. 3 5. 6 5. 1 4. 9 5. 4

0. 96 1. 00 0. 78 1. 03 0. 86 1. 19 1. 16 0. 98 0. 61 1. 03 1. 22 1. 09 0. 89 1. 03 1. 00 1. 24 1. 19 1. 24 1. 15 1. 12 1. 23

2. 73 2. 80 2. 22 2. 93 2. 44 3. 35 3. 28 2. 77 1. 70 2. 93 3. 44 3. 11 2. 51 2. 90 2. 83 3. 52 3. 37 3. 51 3. 27 3. 18 3. 45

UW

23 5

10 .0

22 .0

2. 84

12 .2

3. 34

1. 07

4. 9

1. 12

3. 16

5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

4. 00

NU

4

MA

3

D

2

TE

1

0.7 1

42

Tm pp m

Yb pp m

RI

La pp m

Lu pp m

Hf pp m

Ta pp m

Pb pp m

Th pp m

U pp m

0.4 32 0.4 27 0.3 36 0.4 53 0.3 75 0.5 1 0.5 2 0.4 22 0.2 62 0.4 55 0.5 3 0.4 8 0.3 95 0.4 45 0.4 44 0.5 5 0.5 2 0.5 6 0.5 2 0.4 9 0.5 4

2. 85 2. 76 2. 19 2. 93 2. 44 3. 37 3. 32 2. 80 1. 72 3. 03 3. 43 3. 19 2. 58 2. 96 2. 95 3. 67 3. 52 3. 74 3. 41 3. 27 3. 58

0.4 16 0.3 99 0.3 18 0.4 27 0.3 58 0.4 9 0.4 9 0.4 11 0.2 44 0.4 42 0.5 1 0.4 70 0.3 75 0.4 32 0.4 31 0.5 4 0.5 2 0.5 3 0.5 0 0.4 69 0.5 3

2. 0 2. 00 1. 55 2. 3 2. 2 2. 6 2. 9 2. 3 1. 52 2. 0 3. 0 2. 6 1. 69 2. 4 2. 4 2. 4 2. 9 3. 1 2. 1 2. 2 2. 9

0.3 4 0.2 6 0.2 3 0.2 8 0.2 7 0.4 9 0.3 4 0.3 3 0.2 8 0.2 4 0.7 1 0.4 0 0.1 78 0.3 4 0.4 8 0.3 2 0.6 0 0.4 9 0.2 3 0.3 5 0.3 9

5. 4 7. 4 4. 2 3. 7 4. 6 6. 7 6. 5 5. 2 3. 2 4. 9 5. 7 5. 3 3. 3 5. 8 4. 8 5. 8 4. 4 5. 7 4. 9 4. 4 5. 3

1. 98 2. 1 1. 32 2. 4 2. 3 2. 5 2. 9 2. 2 1. 28 1. 95 2. 6 2. 5 1. 55 2. 4 1. 73 2. 5 2. 5 2. 9 2. 1 1. 87 3. 1

0. 41 0. 42 0. 30 0. 43 0. 42 0. 46 0. 59 0. 41 0. 32 0. 35 0. 58 0. 61 0. 29 0. 51 0. 40 0. 47 0. 50 0. 55 0. 35 0. 54 0. 93

0.4 9

3. 30

0.4 9

2. 5

0.4 1

4. 6

2. 4

0. 66

SC

Ba pp m

AC CE P

Sam ple

PT

Table 1. Continued

PT

ACCEPTED MANUSCRIPT

(La/Sm)N

(Gd/Lu)N

Eu/Eu*

Ce/Ce*

REE

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

1.74 2.13 1.98 1.84 2.05 2.04 1.94 1.96 2.04 1.83 1.90 1.94 1.74 2.13 1.71 1.72 1.86 1.73 1.66 1.60 1.83

0.93 1.24 1.12 1.08 1.11 1.14 1.06 1.09 1.19 1.00 1.09 1.02 0.99 1.13 0.99 0.99 1.00 1.01 0.94 1.01 1.02

0.93 0.96 1.07 0.86 1.05 0.93 0.88 1.05 1.39 0.85 0.77 0.92 1.04 1.03 0.99 0.86 0.85 0.94 0.83 0.93 0.85

0.96 1.27 1.01 1.01 0.99 1.04 0.99 0.99 0.98 1.10 0.99 0.99 1.00 1.09 0.97 0.99 1.00 1.00 1.16 0.97 0.99

54 77 51 65 59 81 74 65 43 62 78 67 51 73 57 72 72 75 64 63 76

UW

1.87

1.00

0.89

1.00

AC CE P

TE

D

MA

NU

SC

Sample

RI

Table 1. Continued

70

Table 2: CIA, CIW, MIA(O), MIA(R) and WIP values for spheroidally-weathered diabase samples and the un-weathered rock (UW). Sample

CIA

CIW

MIA(O)

MIA(R)

WIP

1

36.1

36.7

40.5

20.0

62504

2

41.1

41.9

42.0

27.3

61023

43

40.3

41.4

24.7

59926

4

37.0

37.8

37.8

22.4

62279

5

41.4

42.4

41.2

28.5

62320

6

38.7

39.8

39.9

23.8

69776

7

38.5

39.5

40.9

23.3

68209

8

40.8

41.7

41.9

27.1

69348

9

44.2

44.8

45.7

32.6

67238

10

35.1

35.7

40.4

18.9

61122

11

36.4

37.2

40.8

20.3

66163

12

38.2

39.0

40.6

22.7

68104

13

36.4

37.1

37.6

21.5

72854

14

42.4

43.3

42.2

29.8

71264

15

37.7

38.4

41.5

21.9

16

37.1

37.7

39.4

21.7

17

35.6

36.5

35.7

21.0

18

37.4

38.3

38.7

19

32.2

32.7

34.4

20

37.0

37.6

39.6

21

38.1

38.8

42.7

UW

37.8

38.7

NU 70865 78172

MA

73774 85103

17.0

73750

21.6

77103

TE

D

22.6

21.7

76274

23.3

69600

AC CE P

38.5

RI

39.5

SC

3

PT

ACCEPTED MANUSCRIPT

44

ACCEPTED MANUSCRIPT

Table 3. Sr and Nd isotopic composition and Rb/Nb of selected samples of the spheroidallyweathered diabase as well as the un-weathered rock (UW).

3.2 3

11.8 2

0.273 2

10

40. 8 35. 1

3.1 2 2.9 4

11.6 6 10.5 7

0.267 5 0.278 0

12

38. 2

3.2 3

11.9 3

0.271 0

42. 4 37. 0

3.3 3 3.3 0

12.6 9 11.7 7

0.262 3 0.280 7

37. 8

3.3 4

12.1 6

0.274 5

20 Avera ge

UW

(0)

0.511907

0.512075 0.512204

0.512103

0.511996 0.512158

TE

14

εNd

14.2 6 10.9 8 8.47 10.4 4 12.5 2 9.36

0.512099

εNd (1.2Ga)

Rb

-35.72

28.3 8

-31.36 -30.82

-31.48

10.5 1

-31.93 -32.24 -32.26 (n=6)

-32.22

AC CE P

8

Nd/144 Nd (0)

Sr

Rb/ Sr

87

Sr/86S r (0)

PT

4

37. 0

143

RI

Sm/ Nd

0.29 5

0.73853 8

92.3

0.23 3 0.22 3

0.71732 9 0.71551 8

32.8 1

115. 4

0.28 4

0.71973 5

29.7 9 23.3

139. 0 115. 9

0.21 4 0.20 1

0.71617 3 0.71548 1

31.5 8

129. 6

0.24 4

0.72252 8

96.3

SC

Nd

30.2 20.5 7

NU

S m

MA

CI A

D

Sampl e#

129. 5

Rb/Sr, Rb/Nb, and Sm/Nd ratios are calculated from the ICP-MS trace element concentration data in Table 1 and CIA are values are from Table 2. εNd (0) is calculated using the present day bulk earth (CHUR) value of 143Nd/144Nd = 0.512638. The εNd (0) values represent the deviation of 143Nd/144Nd ratios in parts per 104 from the present day CHUR value.

45

87

Sr/86Sr

(1.3Ga)

Rb/ Nb

0.723047

8.36

0.703798

7.07

0.704903

7.13

0.704909

6.22

0.704784

7.04

0.705065 0.704692(n =5)

5.68

0.709714

7.37