Author's personal copy Adsorption-and desorption-controlled magnesium isotope fractionation during...

$: Author's personal copy Adsorption-and desorption-controlled magnesium isotope fractionation during extreme weathering of basalt in Hainan Island, China$
(This is a sample cover image for this issue. The actual cover is not yet available at this time.)

This article appeared in a journal published by Elsevier. The attachedcopy is furnished to the author for internal non-commercial researchand education use, including for instruction at the authors institution

and sharing with colleagues.

Other uses, including reproduction and distribution, or selling orlicensing copies, or posting to personal, institutional or third party

websites are prohibited.

In most cases authors are permitted to post their version of thearticle (e.g. in Word or Tex form) to their personal website orinstitutional repository. Authors requiring further information

regarding Elsevier’s archiving and manuscript policies areencouraged to visit:

http://www.elsevier.com/copyright

http://www.elsevier.com/copyright

Author's personal copy

Adsorption- and desorption-controlled magnesium isotope fractionationduring extreme weathering of basalt in Hainan Island, China

Kang-Jun Huang a,b,n, Fang-Zhen Teng b,n, Gang-Jian Wei c, Jin-Long Ma c, Zheng-Yu Bao a

a State Key Laboratory of Geological Processes and Mineral Resources & Faculty of Earth Sciences, China University of Geosciences, Wuhan 430074, Chinab Isotope Laboratory, Department of Geosciences & Arkansas Center for Space and Planetary Sciences, University of Arkansas, Fayetteville, AR 72701, USAc State Key Laboratory of Isotope Geochemistry, Guangzhou Institute of Geochemistry, Chinese Academy of Sciences, Guangzhou 510640, China

a r t i c l e i n f o

Article history:

Received 18 June 2012

Received in revised form

28 September 2012

Accepted 6 October 2012

Editor: B. Marty

Keywords:

magnesium

isotope fractionation

chemical weathering

basalt

adsorption

desorption

a b s t r a c t

Magnesium isotopic compositions of a set of clay-rich saprolites developed on the Neogene tholeiitic

basalt from Hainan Island in southern China have been measured in order to document the behavior of

Mg isotopes during continental weathering. Compared with unaltered basalts (d26Mg¼�0.36%), the

overlying saprolites are strongly depleted in Mg (i.e., tTh,Mg¼�99.1% to �92.9%), and display highly

variable d26Mg, ranging from �0.49% to þ0.40%. Magnesium concentration and d26Mg value of the

saprolites display a general increasing trend upwards in the lower part of the profile, but a decreasing

trend towards the surface in the upper part. The variations of Mg concentration and isotopic

composition in this weathering profile can be explained through adsorption and desorption processes:

(1) adsorption of Mg to kaolin minerals (kaolinite and halloysite), with preferential uptake of heavy Mg

isotopes onto kaolin minerals; and (2) desorption of Mg through cation exchange of Mg with the

relatively lower hydration energy cations in the upper profile. Evidence for adsorption is supported by

the positive correlation between d26Mg and the modal abundance of kaolin minerals in saprolite of the

lower profile, while negative correlations between d26Mg and concentrations of lower hydration energy

cations (e.g., Sr and Cs) in the upper profile support the desorption process. Our results highlight that

adsorption and desorption of Mg on clay minerals play an important role in behavior of Mg isotopes

during extreme weathering, which may help to explain the large variation in Mg isotopic composition

of river waters.

& 2012 Elsevier B.V. All rights reserved.

1. Introduction

The chemical weathering of silicate rocks results in drawdownof atmospheric CO2 through precipitation of Ca and Mg bearingcarbonate minerals, and hence regulates the earth’s climate on ageological time scale (Berner et al., 1983; Gaillardet et al., 1999).As one of the most easily weathered Ca–Mg-rich silicate rocks onthe earth’s surface, basalts play a major role in the global carboncycle, with about 30% of atmospheric CO2 consumed by theweathering of continental basalts (Dessert et al., 2003).

Reconstruction of the global CO2 cycle over long-term from thegeological archives requires suitable geochemical proxies ofsilicate weathering. The stable isotopes of Mg have severaladvantages as a promising tracer of silicate weathering. Magne-sium is a fluid-mobile, major element in both the mantle and thecrust, and has three isotopes with relative mass difference of �8%

between 24Mg and 26Mg, which can potentially lead to largemass-dependent Mg isotope fractionation (Young and Galy,2004). Previous studies have revealed that Mg isotope fractiona-tion is small during magmatic differentiation (Teng et al., 2007,2010a; Liu et al., 2010), but large during biological processes suchas plant growth (Black et al., 2008; Bolou-Bi et al., 2010, 2012)and during silicate weathering process (Tipper et al., 2006a, 2008;Pogge von Strandmann et al., 2008a, 2008b, 2012; Li et al., 2010;Teng et al., 2010b; Wimpenny et al., 2010, 2011; Opfergelt et al.,2012). However, the magnitude and mechanisms of Mg isotopefractionation during chemical weathering of silicate rocks are stillpoorly understood.

Chemical weathering of silicate rocks involves the dissolutionof primary minerals and the formation of secondary minerals.Studies of Mg isotopes during primary mineral dissolution showconflicting results. Wimpenny et al. (2010) carried out dissolutionexperiments and found that light Mg isotopes are preferentiallyreleased into the dissolved phase during dissolution of basaltglass and forsterite. This conclusion agrees with studies of weath-ering profile developed on a diabase dike, which shows that Mgisotopic compositions of saprolites evolve towards heavy valueswith increased degree of weathering, suggesting that light Mg

Contents lists available at SciVerse ScienceDirect

journal homepage: www.elsevier.com/locate/epsl

Earth and Planetary Science Letters

0012-821X/$ - see front matter & 2012 Elsevier B.V. All rights reserved.

http://dx.doi.org/10.1016/j.epsl.2012.10.007

n Corresponding authors at: State Key Laboratory of Geological Processes and

Mineral Resources & Faculty of Earth Sciences, China University of Geosciences,

Wuhan 430074, China.

E-mail addresses: [email protected] (K.-J. Huang), [email protected]

(F.-Z. Teng).

Earth and Planetary Science Letters 359–360 (2012) 73–83


isotopes were preferentially released into the hydrosphere duringcontinental weathering (Teng et al., 2010b). By contrast, anotherexperimental study found limited Mg isotope fractionation duringgranite mineral dissolution and the large Mg isotopic variation inthe output solution during granite dissolution mainly reflectspreferential dissolution of isotopically distinct minerals (Ryuet al., 2011). The different behaviors of Mg isotopes duringmineral dissolution are uncertain and may reflect the differencein Mg crystalline sites in different minerals.

When compared to primary mineral dissolution, the behavior ofMg isotopes during secondary mineral formation is more complex.Soil and clay fractions are isotopically heavier than their parentrocks (Tipper et al., 2006a, 2010; Brenot et al., 2008; Teng et al.,2010b; Opfergelt et al., 2012; Pogge von Strandmann et al., 2012),suggesting that heavy Mg isotopes are preferentially incorporatedinto the structure of secondary mineral or adsorbed onto the soilexchange complex. By contrast, allophane, a secondary mineralformed during the weathering of basalt, appears to prefer light Mgisotopes to heavy ones (e.g., Pogge von Strandmann et al., 2008a).Experimental studies also show that some phyllosilicates have an

affinity for light Mg isotopes during water-rock interactions (e.g.,Wimpenny et al., 2010). These discrepancies are still unclear andmore studies are needed to address them.

To improve our knowledge of behaviors of Mg isotopes duringcontinental weathering, especially to understand the role ofsecondary minerals on Mg isotope fractionation, we studied awell-characterized saprolite profile that developed on Neogenebasalt in a tropical climate on Hainan Island in southern China.This saprolite profile is ideally suited for investigating behaviorsof Mg isotopes during secondary mineral formation, because(1) the weathering input of Mg is derived predominantly from asingle silicate mineral (pyroxene); (2) primary minerals havealmost completely broken down and secondary minerals dom-inate the saprolites due to extreme weathering (Ma et al., 2007).Our results demonstrate that Mg isotope fractionation duringextreme weathering of basalt involves a two-step process: pre-ferential adsorption of heavy Mg isotopes onto the surface ofsecondary minerals and preferential desorption of previouslyadsorbed heavy Mg isotopes from the secondary minerals intopore water, due to high rain infiltration in the upper profile.

HK06-01

Topsoil

Gravel layer

Homogeneous fine laterite

Color turnsyellowish a bit

Lateritewith corestone

Unalteredbasalts

Depth(cm)

30

0

50

100

200

250

320

400

450

HK06-02

HK06-03

HK06-04

HK06-05

HK06-06

HK06-07

HK06-08

HK06-09HK06-10

HK06-R1

HK06-12HK06-13

HK06-14

HK06-15HK06-16HK06-17HK06-18HK06-19HK06-20

HK06-R2

K+F

K+F

K+F

K+G+F

K+G+F

K+G+F

K+G+I+F

K+G+F

K+G+FH+G+F

H+FH+G

H+G+FH+G+FH+G+FH+G+FH+G+FH+G+FK+G+F

Fig. 1. (Color online) Sketch section of the basalt weathering profile in the northern region of Hainan Island, China (after Ma et al., 2007). According to the texture and

color, six different horizons are identified within the profile, but all boundaries between these horizons are essentially gradational. The numbers represent the sample ID

and their sampling position in the profile. The letters within the profile represent clay mineral distribution in the different horizons (F¼Fe-oxy-hydroxides; H¼halloysite;

G¼gibbsite; I¼ illite; K¼kaolinite).

K.-J. Huang et al. / Earth and Planetary Science Letters 359–360 (2012) 73–8374


2. Samples

The weathering profile studied here is well exposed at a small hillin the northeastern region of Hainan Island in southern China(11013807100E and 1913407400N). This region has a tropical, moist,monsoonal style climate, with a mean annual air temperature of25 1C, and mean annual precipitation of �1500 mm (Ma et al., 2007).

The sampling profile consists of an uninterrupted progressionfrom unaltered basalts to an extremely weathered laterite residuetowards the surface (Fig. 1). The topsoil and the gravel layer at theupper 50 cm were not sampled to avoid the disruption of farmingactivities. Beneath the gravel layer, a set of fine laterite with ahomogeneous red color has developed. Seven samples (HK06-01to HK06-07) were collected at intervals of 30 cm. Three samples(HK06-08 to HK06-10) were sampled in the section from 250 to320 cm, where the soil has a slight yellowish color. Below 320 cm,the soil color becomes pistachio with unaltered core stones, andnine samples were collected at intervals of 10–15 cm. In addition,two unaltered tholeiitic basalt samples (HK06-R1 and HK06-R2)were collected from 5 m below (Fig. 1).

The tholeiitic basalts in this region erupted during the Neo-gene, with a K–Ar age of 4.0570.33 Ma (Zhu and Wang, 1989).The unaltered tholeiitic basalts contain 10% of pyroxene in thephenocryst, and 60% of plagioclase, 25% of clinopyroxene and fewopaque minerals in the groundmass (Ma et al., 2007). Thesaprolites are dominated by secondary minerals, such as kaolinite,halloysite, gibbsite and Fe-oxy-hydroxides, while primary

minerals are absent (Fig. 1). A transition in clay mineralogy existsat �3 m depth of the weathering profile (Fig. 2A). Beneath 3 m,the clay mineralogy is dominated by halloysite with modalabundances ranging from 30.7% to 87.4%, and kaolinite is absentexcept in the sample HK06-20 (20%). Above 3 m, kaolinitedominants the secondary mineral assemblage with modal abun-dances ranging from 28.3% to 82.0%, and only one sample (HK06-7) contains halloysite (53.4%). Other clay minerals, such as illite,are absent except in the sample HK06-07 (�16%) (Ma et al.,2007).

To quantitatively evaluate the relative depletion or enrichmentof an element during chemical weathering, we calculate thepercentage changes of elemental ratios to Th relative to parentrock (tTh,j¼100� [(Cj/CTh)s/(Cj/CTh)p�1], where Cj and CTh repre-sent the concentration of element j and Th, respectively, and ‘‘s’’and ‘‘p’’ refer to sparolite and parent rock, Nesbitt et al., 1980).Thorium is used as a conservative element since it has beendemonstrated to be more resistant to an acid environment thanother elements (Ma et al., 2007). Positive and negative tTh,j valuesindicate the enrichment and depletion of element j relative to theparent rock, respectively, while a zero value means that elementj is as immobile as Th during weathering. Ma et al. (2007) foundthat the abundances of most elements, including major elements(Fig. 2), rare earth elements (REEs) and immobile elements(Ti, Zr, Hf, Nb, and Ta, Fig. 3) are significantly depleted in thesection above 3 m and gradually become enriched or lessdepleted in the section below 3 m. The depletion of REEs and

0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

Dep

th (m

)

Abundance of kaolin mineral (%)

kaolinite

halloysite

0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

4.5

5.0

Dep

th (m

)

τTh,Si(%)

TOC (%) pH

τTh,Al (%)

0 20 40 60 80 100

-90 -80 -70 -60 -50

0 0.05 0.10 0.15 0.20 5.0 5.2 5.4 5.6 5.8 6.0

-60 -40 -20 0 20 40 60 80 -60 -40 -20 0 20

τTh,Fe(%)

A B C

D E F

Fig. 2. (Color online) Modal abundance of kaolin mineral, total organic carbon (TOC), pH and percentage changes of selected major elements (Si, Al, Fe) to Th ratios relative

to unaltered basalts (tTh,j) as a function of depth in the basalt weathering profile. In this and subsequent figures, open circles represent samples taken above 3 m depth,

solid circles are samples from below 3 m depth.

K.-J. Huang et al. / Earth and Planetary Science Letters 359–360 (2012) 73–83 75


immobile elements increases towards the surface (Fig. 3A and B),suggesting that this weathering profile has been subjected toextreme weathering and that the weathering intensity increasestowards the surface (Ma et al., 2007, 2010). In addition, the redox-sensitive elements such as Mn, Co, Ce, Cr and U are enriched inthe middle profile, with maximum enrichment occurring at 3 m(Fig. 3C). These enrichments are accompanied by significantdepletion of total organic carbon (TOC, Fig. 2B) and the absenceof organic nitrogen, as well as higher water content in the middleprofile, reflecting that organic colloids and redox conditionsplayed an important role in transferring these elements duringweathering (Ma et al., 2007).

3. Analytical methods

Magnesium isotopic analyses were carried out at the IsotopeLaboratory of the University of Arkansas, Fayetteville. Details forsample dissolution, column chemistry and instrumental analysishave been reported in previous studies (Yang et al., 2009; Li et al.,2010; Teng et al., 2010a). Only a brief description is given below.

Depending on Mg concentration, 1 to 25 mg of sample pow-ders were dissolved in Savillex screw-top beakers in a mixture ofOptima-grade HF-HNO3-HCl, in order to obtain �50 mg Mg forhigh-precision isotopic analysis. Magnesium was separatedfrom the sample matrix by cation exchange chromatographywith Bio-Rad AG50W-X8 resin (200–400 mesh) in 1 N HNO3

media following the established procedures (Yang et al., 2009;

Teng et al., 2010a). At least two standards (seawater, olivine, ormeteorite) were processed with samples for each batch of columnchemistry. The same column procedure was then repeated inorder to obtain a pure Mg solution for mass spectrometry.

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0

Dep

th (m

) Rare Earth Elements

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0

Dep

th (m

)

τTh,j (%)

-100 400 900 1400 1900 2400

-100 0 100 200 300 400

-100 0 100 200 300 400 500 -100 -95 -90 -85 -80

τ Th,j (%)

-100 -50 0 50

A

C

B

D

-40 -20 0 20

Immobile Elements

Zr TabN iT fH

Ce

U

Mn

Co

Sr K Rb Cs

Alkalis and Alkaline Earth Elements

Fig. 3. The percentage changes of (A) rare earth elements (exclude Ce), (B) immobile elements (Ti, Zr, Hf, Nb, Ta), (C) redox-sensitive elements (Mn, U, Co, Ce) and

(D) alkalis and alkaline earth elements (K, Rb, Cs, Sr) to Th ratios (tTh,j) relative to unaltered basalts as a function of depth in the basalt weathering profile.

Table 1Magnesium isotopic compositions of reference materials analyzed in this study.

Standard Type d26Mg 2SD d25Mg 2SD

Hawaii seawater Seawater �0.82 0.09 �0.44 0.04

Replicate �0.85 0.07 �0.40 0.05

Wt. Average �0.84 0.05 �0.43 0.03

Kilbourne Hole Olivine �0.25 0.07 �0.14 0.05

Replicate �0.19 0.07 �0.10 0.07

Replicate �0.26 0.08 �0.13 0.07

Replicate �0.26 0.10 �0.14 0.04

Wt. Average �0.23 0.06 �0.13 0.03

Allende Chondrite �0.31 0.10 �0.19 0.05

Replicate �0.31 0.09 �0.16 0.06

Wt. Average �0.31 0.07 �0.18 0.04

BHVO-2 Basalt �0.24 0.08 �0.12 0.05

dxMg¼[(xMg/24Mg)sample/(xMg/24Mg)DSM3 �1]�1000, where x¼25 or 26 and

DSM3 is Mg solution made from pure Mg metal (Galy et al., 2003).

2SD¼2 times the standard deviation of the population of n (n420) repeat

measurements of the standards during an analytical session.

‘‘Replicate’’ refers to repeat column chemistry and instrumental measurement of

different aliquots of the same stock solution.

Wt. Average¼weighted average.



Magnesium isotopic compositions were measured by the standardbracketing method using a Nu Plasma MC-ICP-MS. A ‘‘wet’’ plasmaintroduction system, which consisted of a quartz cyclonic spraychamber (Elemental Scientific Inc.) and a MicroMist micro-uptakeglass concentric nebulizer (Glass Expansion), was utilized for theanalysis of purified Mg sample solution (�300 ppb Mg in 3% HNO3

solution). The three Mg isotopes (24, 25 and 26) were measuredsimultaneously in separate Faraday cups (H5, Ax and L4) under low-resolution mode. No molecular interferences or double charge inter-ferences were observed. A typical 26Mg signal for a 300 ppb Mgsolution was between 2 and 4 V. The background Mg signals for 24Mg(o 10�4 V) were negligible relative to the sample signals (2–4 V).Variations in Mg isotopic compositions are typically expressed interms of d26Mg or d25Mg, which represent the per mil (%) differencein the ratio 26Mg/24Mg or 25Mg/24Mg between the sample and thestandard DSM3 (Galy et al., 2003): dxMg¼[(xMg/24Mg)sample/(xMg/24Mg)DSM3�1]�1000, where x refers to 25 or 26.

At least one standard was routinely analyzed during the courseof an analytical session. The long-term precision based on replicateruns of the IL-Mg-1, a synthetic multi-element standard solutionwith concentration ratios of Mg:Fe:Al:Ca:Na:K:Ti¼1:1:1:1:1:1:0.1,over three years is 70.07% (2SD) (Teng et al., 2010a). Results forsolution, rock and mineral reference materials analyzed during thecourse of this study (Table 1) agree with previously published valuesfor seawater (Foster et al., 2010; Ling et al., 2011, and referencestherein), Allende chondrite (Teng et al., 2010b), Kilbourne Holeolivine (KH olivine) (Teng et al., 2010b), and BHVO-2 (Pogge vonStrandmann et al., 2011), confirming that our data are accurate.

4. Results

Magnesium isotopic compositions are reported in Table 1 forreference materials (KH olivine, seawater, Allende chondrite, andBHOV-2), and Table 2 for saprolite samples and unaltered basalts,along with major and trace elemental data that are presented inMa et al. (2007).

The unaltered basalts have the highest MgO content with anaverage value of 7.34 wt%. Saprolites have much lower MgO con-tents ranging from 0.21 wt% to 0.87 wt% (Table 2). tTh,Mg values ofsaprolites vary from �99.1% to �92.9%, indicating that about 93%to 99% Mg in the parent rock has been removed from the profile.Like other major elements (Si, Al, Fe) (Fig. 2), Mg also displays aconcentration discontinuity at 3 m depth (Fig. 4A). Below 3 m depth,tTh,Mg values are more variable, and less negative than those in theupper profile. Above the 3 m depth, tTh,Mg values gradually becomemore negative towards the surface.

The unaltered basalts have d26Mg values ranging from�0.38% to �0.33%, with an average value of �0.36%, whichfalls in the range of d26Mg values of the typical continental basalts(�0.60% to �0.26%, Yang et al., 2012). Relative to the unalteredbasalts, the overlying saprolite is heterogeneous with d26Mgranging from �0.49% to þ0.40%. Similar to Mg concentration,d26Mg values of saprolites from the profile below 3 m becomeheavier towards the surface, but changes less systematically thantTh,Mg (Fig. 4B). d26Mg values of saprolites above 3 m graduallybecome lighter towards the surface, and the three uppermostsamples are lighter than the unaltered basalts (Fig. 4B).

5. Discussion

The large variations in Mg concentration and Mg isotopiccomposition in this weathering profile likely reflect adsorptionand desorption processes associated with clay minerals duringthe extreme continental weathering. Heavy Mg isotopes are first

retained in the saprolite through adsorption and are later des-orbed from the saprolite in the upper profile by cation exchange.Below, we will discuss these processes in detail.

5.1. Adsorption and desorption control on Mg concentrations

of saprolites

During continental weathering, primary minerals progressivelydissolved and secondary minerals precipitated from pore water, orderived from simple alterations to existing primary mineral structure(Wilson, 1999). In these processes, like other alkalis and alkaline earthelements, Mg is liable to leach from the Mg-bearing primary mineralsduring incipient-to-intermediate weathering, and tends to beadsorbed onto or incorporated into secondary minerals duringadvanced weathering (Nesbitt et al., 1980; Nesbitt and Wilson,1992). In the weathering profile studied here, the tTh,Mg values areclose to �100%, suggesting that nearly all Mg has been removed dueto the extreme weathering. Indeed, no primary mineral is identifiedin this profile (Ma et al., 2007) and saprolites are mainly composed oftypical secondary minerals formed by weathering of basalts undertropical conditions, such as kaolin minerals (kaolinite and halloysite)and Al- and Fe-oxy-hydroxides (e.g., gibbsite, goethite and hematite)(Trolard and Tardy, 1987; Ma et al., 2012). No Mg-rich secondaryphase is present in these saprolite samples except sample HK06-07,which contains about 16% illite. All these suggest that Mg in thesesaprolites is a ‘‘trace’’ element, not controlled by Mg-rich mineralsand redistributed in secondary minerals.

Previous studies have demonstrated that kaolin minerals andAl- and Fe-oxy-hydroxides can absorb certain ions and retain themin an exchangeable state (Carroll and Starkey, 1960; Weaver andPollard, 1973; Kinniburgh et al., 1976; Bleam and McBride, 1985;Ma and Eggleton, 1999; Olu-Owolabi and Ajayi, 2003). Adsorption ofions onto these secondary minerals strongly depends on the surface-charge properties of minerals, which can be described by the pH ofthe point of zero charge (pHpzc). The pHpzc for kaolinite andhalloysite is relatively low, ranging from 2.53 to 3.56 (Scroth andSposito, 1997; Kosmulski, 2009). By contrast, the pHpzc of Al- and Fe-oxy-hydroxides is higher, ranging from 6.12 to 8.81 (Kosmulski,2006). At the pH of saprolite in this profile (5.09 to 5.78, Table 2), thesurface charge of kaolinite and halloysite is negative, whereas that ofAl- and Fe-oxy-hydroxides is positive. Due to its positive charge,Mg2þ , one of the most common exchangeable cations, is mainlyadsorbed to the surfaces of kaolin minerals, and not to Al- and Fe-oxy-hydroxides. This is consistent with relatively higher MgOcontent in natural kaolinite and halloysite (up to 2.51 wt%, Weaverand Pollard, 1973) than Al- and Fe-oxy-hydroxides (o 0.37 wt%,Manceau et al., 2000; Peskleway et al., 2003). Since the cationexchange capacity (CEC) depends on the amount of clay mineralspresent (Bas-c-etin and Atun, 2006), the positive correlation betweenthe abundance of kaolin minerals and the tTh,Mg values of saprolitesin the profile below 3 m (Fig. 5A) indicates that Mg is mainly presentin kaolin minerals due to adsorption through equilibrium exchangebetween water and kaolin minerals.

By contrast, tTh,Mg value decreases towards the surface in theprofile above 3 m (Fig. 4A), negatively correlates with the modalabundance of kaolin minerals (Fig. 5A), and is overall lower thanthose of saprolites in the lower profile. Since kaolinite and halloysitehave the same layer structure and similar cation exchange capacity(Ma and Eggleton, 1999; Joussein et al., 2005), this negative trend andlower tTh,Mg values in the upper profile are not likely due to themineralogical difference between kaolinite and halloysite, but morelikely result from desorption of Mg from kaolinite through cationexchange during extreme weathering in the upper profile (e.g., Whiteet al., 2009). The more intense weathering in the upper profile issupported by the greater depletion of most elements includingimmobile elements and REEs (Fig. 3A and B), where large amount



Table 2Major, trace element concentration, and Mg isotopic composition of saprolite and unaltered basalt from Hainan Island, China.

Sample # Depth pH MgO Sr Cs Th tTh,Mg tTh,Sr tTh,Cs d26Mg 2SD d25Mg 2SD D25Mg0

(m) (wt %) (ppm) (ppm) (ppm) (%) (%) (%) (%) (%) (%)

HK06-01 0.50 5.18 0.27 25.52 1.71 9.20 �99.0 �97.6 27.7 �0.50 0.07 �0.27 0.05 �0.01

Replicate �0.48 0.07 �0.23 0.05 0.02

Wt. Average �0.49 0.05 �0.25 0.04 0.01

HK06-02 0.90 5.16 0.25 25.44 1.51 9.19 �99.1 �97.6 12.6 �0.48 0.07 �0.24 0.05 0.01

Replicate �0.46 0.07 �0.19 0.05 0.05

Wt. Average �0.47 0.05 �0.21 0.04 0.03

HK06-03 1.30 5.30 0.23 21.54 1.09 8.58 �99.1 �97.8 �12.7 �0.44 0.07 �0.23 0.05 0.00

Replicate �0.44 0.07 �0.22 0.05 0.01

Wt. Average �0.44 0.05 �0.23 0.04 0.00

HK06-04 1.60 5.26 0.21 18.97 0.85 7.90 �99.1 �97.9 �26.2 �0.32 0.07 �0.13 0.05 0.04

Replicate �0.27 0.07 �0.16 0.05 �0.02

Wt. Average �0.29 0.05 �0.15 0.04 0.00

HK06-05 1.90 5.38 0.26 13.35 0.76 7.83 �98.9 �98.5 �33.1 �0.13 0.07 �0.09 0.05 �0.02

Replicate �0.07 0.07 �0.05 0.05 �0.01

Wt. Average �0.10 0.05 �0.07 0.04 �0.02

HK06-06 2.20 5.33 0.33 10.40 0.69 7.10 �98.4 �98.7 �33.5 �0.21 0.07 �0.13 0.05 �0.02

Replicate �0.18 0.07 �0.13 0.05 �0.04

Wt. Average �0.19 0.05 �0.13 0.04 �0.03

HK06-07 2.50 5.09 0.40 7.71 0.55 5.86 �97.7 �98.9 �35.8 0.01 0.07 �0.01 0.05 �0.02

Replicate �0.04 0.07 �0.01 0.05 0.01

Wt. Average �0.02 0.05 �0.01 0.04 0.00

HK06-08 2.80 5.14 0.57 3.96 0.41 5.02 �96.2 �99.3 �43.2 �0.15 0.10 �0.09 0.05 �0.01

Replicate �0.12 0.08 �0.07 0.07 �0.01

Wt. Average �0.13 0.06 �0.08 0.04 �0.01

HK06-09 3.00 5.53 0.49 4.25 0.35 3.72 �95.5 �99.0 �34.4 �0.13 0.10 �0.03 0.05 0.04

Replicate �0.25 0.08 �0.14 0.07 �0.01

Wt. Average �0.20 0.06 �0.07 0.04 0.03

HK06-10 3.20 5.52 0.60 2.39 0.31 5.99 �96.6 �99.7 �64.7 �0.14 0.07 �0.07 0.06 0.00

Replicate �0.20 0.09 �0.09 0.06 0.01

Wt. Average �0.16 0.06 �0.08 0.05 0.00

HK06-12 3.50 5.64 0.84 3.20 0.44 4.90 �94.2 �99.4 �38.9 0.10 0.09 0.03 0.06 �0.02

Replicate 0.11 0.08 0.06 0.07 0.00

Wt. Average 0.10 0.06 0.04 0.05 �0.01

HK06-13 3.65 5.70 0.81 1.84 0.30 3.87 �92.9 �99.6 �46.0 0.44 0.10 0.20 0.05 �0.03

Replicate 0.37 0.09 0.19 0.06 0.00

Wt. Average 0.40 0.07 0.20 0.04 �0.01

HK06-14 3.80 5.43 0.64 1.24 0.35 5.21 �95.8 �99.8 �53.8 0.10 0.09 0.04 0.06 �0.01

Replicate 0.13 0.10 0.10 0.04 0.03

Wt. Average 0.12 0.07 0.08 0.03 0.02

HK06-15 3.95 5.74 0.87 1.43 0.38 4.74 �93.8 �99.7 �44.2 0.04 0.07 0.05 0.06 0.03

Replicate 0 0.09 0 0.06 0.01

Wt. Average 0.02 0.06 0.03 0.04 0.02

HK06-16 4.10 5.76 0.63 1.36 0.44 6.37 �96.7 �99.8 �52.0 �0.17 0.07 �0.08 0.06 0.01

Replicate �0.14 0.09 �0.09 0.06 �0.02

Wt. Average �0.16 0.06 �0.09 0.04 �0.01

HK06-17 4.20 5.72 0.68 2.57 0.46 5.97 �96.1 �99.6 �47.0 0.05 0.07 0 0.06 �0.03

Replicate 0.06 0.09 0.01 0.06 �0.02

Wt. Average 0.05 0.06 0.02 0.04 �0.02

HK06-18 4.30 5.78 0.76 2.99 0.54 6.81 �96.2 �99.6 �45.1 �0.09 0.07 �0.02 0.06 0.03

Replicate �0.03 0.10 �0.02 0.04 0.00

Wt. Average �0.07 0.07 �0.02 0.04 0.02

HK06-19 4.40 5.42 0.50 5.22 0.42 6.70 �97.5 �99.3 �57.2 0.04 0.07 �0.01 0.06 �0.03

Replicate 0.08 0.10 0.02 0.04 �0.02

Wt. Average 0.05 0.06 0.01 0.03 �0.02

HK06-20 4.50 5.70 0.50 2.43 0.28 6.58 �97.4 �99.7 �70.6 �0.20 0.11 �0.10 0.05 0.00

Replicate �0.15 0.07 �0.08 0.07 0.00

Wt. Average �0.16 0.06 �0.09 0.04 �0.01

HK06-R1 7.26 286 0.36 2.46 �0.34 0.07 �0.18 0.05 0.00

Replicate �0.33 0.09 �0.18 0.04 �0.01

Wt. Average �0.34 0.06 �0.18 0.03 0.00

HK06-R2 7.41 275 0.31 2.63 �0.38 0.09 �0.19 0.04 0.01

Replicate �0.37 0.07 �0.18 0.07 0.01

Wt. Average �0.37 0.05 �0.19 0.04 0.00

Depth, major and trace elemental data are from Ma et al. (2007). Samples HK06-R1 and HK06-R2 are unaltered basalts and other samples are saprolites.

tTh,j¼[(Cj/CTh)saprolite/(Cj/CTh)protolith�1]�100, where j refers to mobile element.

dxMg¼[(xMg/24Mg)sample/(xMg/24Mg)DSM3�1]�1000, where x¼25 or 26 and DSM3 is Mg solution made from pure Mg metal (Galy et al., 2003).

2SD¼2 times the standard deviation of the population of n (n420) repeat measurements of the standards during an analytical session.

D25Mg0 is defined as d25Mg0 �0.521� d26Mg0 following Young and Galy (2004).

‘‘Replicate’’ refers to repeat column chemistry and measurement of different aliquots of a stock solution.

Wt. Average¼weighted average.



of leaching occurs due to rainwater infiltration (Ma et al., 2007). It isimportant to note that the values of tTh,K, tTh,Cs, tTh,Rb, and tTh,Sr insaprolites from the upper profile all increase towards the surface(Fig. 3D). These trends are consistent with the increasing trend ofmodal abundance of kaolinite (Fig. 2A), but opposite to the graduallydecreasing trend of tTh,Mg (Fig. 4A). The distinct behavior of Mg

relative to other alkalis and alkaline earth elements reflects theexchange of Mg with these cations in kaolin minerals. BecauseMg2þ has a relatively higher hydration energy (Evangelou, 1998;Driesner et al., 2000), it is less strongly fixed on the surface of clayminerals and can be exchanged with those competing cations withlow hydration energy, such as Kþ , Csþ , Rbþ , and Sr2þ (Sawhney,1972; Cornell, 1993; Tatsuya and Komarneni, 1999; Bas-c-etin andAtun, 2006; Teppen and Miller, 2006). Additional evidence fordesorption of Mg through cation exchange comes from the pH valuesof saprolite samples, which gradually decrease towards the surface(Fig. 2C). The lower pH values in the upper profile can also causeexchange of Hþ for Mg adsorbed to kaolin minerals (Nesbitt et al.,1980), and hence result in the depletion of Mg in the saprolite.

To summarize, a two-step process controls the distribution ofMg in this weathering profile. Magnesium is first retained in thesaprolite through adsorption and later removed from the saproliteby desorption due to the extreme weathering in the upper profile.

5.2. Adsorption and desorption control on Mg isotopic

compositions of saprolites

Compared to the relatively uniform Mg isotopic compositionof unaltered basalts (d26Mg¼�0.36%), Mg isotopic compositions

R2 = 0.79

R2 = 0.77

-0.7

-0.5

-0.3

-0.1

0.1

0.3

0.5

0.7

0 10 20 30 40 50 60 70 80 90 100

δ26M

g (‰

)

Abundance of kaolin mineral (%)

R2 = 0.49

R2 = 0.64

-100

-99

-98

-97

-96

-95

-94

-93

-92

3 23 43 63 83 103

τ Th,

Mg(

%)

Above 3 mBelow 3 m

Unaltered Basalts

Fig. 5. (A) tTh,Mg and (B) d26Mg as a function of the modal abundance of kaolin

mineral in the weathering profile. The solid line represents a linear regression

fitted through the data. Horizontal gray bar represents d26Mg of unaltered basalts.

-0.7

-0.5

-0.3

-0.1

0.1

0.3

0.5

0.7

-100 -99 -98 -97 -96 -95 -94 -93 -92

δ26M

g (‰

)

τTh,Mg (%)

Above 3 mBelow 3 m

Unaltered Basalts

R2 = 0.65

Fig. 6. tTh,Mg vs. d26Mg for the weathering profile. The solid line represents

a linear regression fitted through the data. Horizontal gray bar represents d26Mg

of unaltered basalts.

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0

Dep

th (m

)

τTh,Mg(%)-100 -98 -96 -94 -92 -0.5 -0.3 -0.1 0.1 0.3 0.5 0.7

δ26 Mg (‰)U

naltered Basalts

A B

Fig. 4. (A) tTh,Mg and (B) d26Mg as a function of depth in the weathering profile. Vertical gray bar represents d26Mg of unaltered basalts. Error bars in this and subsequent plots

represent 2SD.



of the overlying saprolites vary significantly, ranging from�0.49% to þ0.40%. Since primary minerals are completelyweathered in this saprolite profile and secondary minerals dom-inate the saprolites, this large Mg isotopic variation likely reflectsMg isotope fractionation associated with the formation anddestabilization of secondary minerals during extreme weathering.This is supported by the significant correlations between d26Mgand modal abundance of kaolin minerals, i.e., negative correlationin the upper profile and positive correlation in the lower profile(Fig. 5B). The correlations between tTh,Mg, d26Mg and modalabundance of kaolin minerals suggest that similar processescontrol Mg elemental and isotopic distribution in this weatheringprofile, i.e., adsorption of Mg onto kaolin minerals followed bydesorption of Mg through cation exchange in the upper profile.This suggestion is further supported by the positive correlationbetween d26Mg and tTh,Mg in the whole weathering profile(Fig. 6), which reflects Mg isotope fractionation during an adsorp-tion/desorption process between soil pore water and kaolinminerals.

5.2.1. Magnesium isotope fractionation during adsorption of Mg

onto secondary minerals

Previous studies have shown that adsorption of aqueous ionsto secondary mineral surfaces can produce significant fractiona-tion of metal stable isotopes (Schwarcz et al., 1969; Palmer et al.,1987; James and Palmer, 2000; Pistiner and Henderson, 2003;Icopini et al., 2004; Pokrovsky et al., 2005; Lemarchand et al.,2007; Balistrieri et al., 2008; Wasylenki et al., 2008; Goldberget al., 2009; Brennecka et al., 2011; Fantle et al., 2012). Forexample, heavier isotopes of Fe, Cu and Zn are preferentiallyadsorbed to Fe-oxy-hydroxide surfaces (Icopini et al., 2004;Pokrovsky et al., 2005; Balistrieri et al., 2008), whereas lighterMo isotopes are preferentially incorporated into ferromanganeseoxide during the adsorption of Mo from seawater (Barling andAnbar, 2004; Kashiwabara et al., 2011). Similarly, light Li isotopesare preferentially incorporated into kaolinite and vermiculitefrom seawater (Zhang et al., 1998). These distinct behaviors ofmetal isotopes during adsorption to secondary mineral surfacesreflect the difference in coordination environment between dis-solved metal and adsorbed metal, with the heavier isotopesfavoring high-energy sites, which in most cases are lower coordi-nation sites (Hoefs, 2009; Brennecka et al., 2011; Schauble, 2011;Wasylenki et al., 2011).

The Mg2þ in soil pore water is octahedrally coordinated withH2O molecules forming an octahedral aquo ion ([Mg(OH2)6]2þ)due to its high degree of hydration (Li et al., 2011; Schauble,2011). By contrast, adsorption of Mg2þ ions onto secondaryminerals (i.e., kaolin minerals in this weathering profile) formsstronger chemical bonds and results in surface complex species(Begum et al., 1998). Formation of surface complex on theexchangeable sites (the silanol (RSiOH) and aluminol (RAlOH)hydroxyl groups) at mineral edges is the dominate pathway foradsorption of metal cations onto kaolin minerals (Ma andEggleton, 1999; Peacock and Sherman, 2005; Srivastava et al.,2005; Gu and Evans, 2008). Although detailed studies on the exactcoordination environment of Mg2þ are not available, studies ofcopper sorption onto kaolinite surface indicate that aqueous Cu2þ

mainly adsorbs onto aluminol and silanol sites by forming surfacecomplexes with the coordination number of 4 and 2 respectively(Peacock and Sherman, 2005). Since the ionic radii of Mg2þ andCu2þ are similar (Marcus, 1988), Mg2þ is expected to formsimilar surface complexes on kaolin minerals with similar coor-dination environments. Therefore, the Mg–O bonds in the kaolinminerals are stronger than those in the pore water, which leads tothe preferential adsorption of heavy Mg isotopes onto kaolin

minerals. This agrees with the fact that the Mg isotopic composi-tion of saprolite is not only heavier than that of soil pore water inbasaltic terrain (e.g., Pogge von Strandmann et al., 2012), but alsois heavier than that of parent rock with the exception of the threeuppermost samples that experienced additional processes (seenext section).

Further evidence in support of Mg isotope fractionation duringadsorption of Mg to kaolin minerals comes from the positivecorrelations between d26Mg and the modal abundance of kaolinminerals in the saprolites from the profile below 3 m (Fig. 5B), aswell as between d26Mg and tTh,Mg in the whole weathering profile(Fig. 6). Considering that Mg is mainly present in kaolin mineralsdue to their high affinity for sorbing Mg2þ (aq) from solution,these positive correlations further indicate that adsorption of Mgto kaolin minerals can cause Mg isotope fractionation, withpreferential uptake of heavy Mg isotopes onto them. This inter-pretation agrees with a recent study of behaviors of Mg isotopesin a basaltic weathering profile in Iceland (Pogge von Strandmannet al., 2012), which reveals that adsorption of heavy Mg isotopesonto the soil exchange complex drives the soil to isotopicallyheavy values.

Based on the positive correlation between d26Mg and the modalabundance of kaolin minerals in the profile below 3 m (Fig. 5B),d26Mg value of pure kaolin minerals (d26Mgkaolin) can be estimatedby extrapolating the modal abundance of kaolin minerals to 100%,which is about þ0.35% (Fig. 5B). If these pure kaolin mineralsare in isotopic equilibrium with soil pore water, as representedby local meteoric water (d26Mgfluid¼�1.59% to �0.59%,Tipper et al., 2010, 2012; Riechelmann et al., 2012), then theapparent fractionation factors between kaolin minerals and porewater [akaolin-fluid¼(1000þd26Mgkaolin)/(1000þd26Mgfluid)] varyfrom 1.00094 to 1.00194. These values are higher than those basedon a saprolite profile developed on a diabase dike (asaporolite-fluid¼

1.00005 to 1.0004, Teng et al., 2010b), suggesting more signifi-cant isotope fractionation during adsorption of Mg onto kaolinminerals.

5.2.2. Magnesium isotope fractionation during desorption in the

upper profile

d26Mg value of saprolites in the upper section negativelycorrelates with the modal abundance of kaolin minerals(Fig. 5B), and shows an overall decreasing trend towards thesurface (Fig. 4B). These trends are opposite to those observed inthe profile below 3 m, suggesting that additional processes, inaddition to adsorption of heavy Mg isotopes onto kaolin minerals,have modified the Mg isotopic distribution in the upper profile.

Biological cycling of Mg has the potential to drive the sapro-lites to isotopically lighter values because of the preferentialuptake of heavy Mg isotopes during plant growth (Black et al.,2008; Bolou-Bi et al., 2010). However, the level of vegetation onthe surface of this profile is limited, and the top 50 cm of theprofile was not sampled, hence the impact of biological processeson these saprolites is expected to be limited. Thus, the most likelyprocess controlling Mg isotopic compositions of saprolites in theupper profile is desorption of Mg from clay minerals, duringwhich process heavy Mg isotopes were preferentially removedfrom the saprolite to water.

Previous studies have found large Mg isotope fractionationassociated with ion exchange processes. For example, Oi et al.(1987) studied Mg isotope separation by using an acidic cation-exchange resin and Sr2þ as the replacement ion, and found thatheavy Mg isotopes are preferentially fractionated into solutionduring cation exchange. Chang et al. (2003) and Teng et al. (2007)also noted that Mg isotopic composition of eluent becomes lighterwith time during cation-exchange chromatography, suggesting



that light Mg isotopes are preferentially retained in the resin.In the upper section of the weathering profile, tTh,Sr and tTh,Cs

negatively correlate with d26Mg of saprolites (Fig. 7). This correla-tion likely reflects Mg isotope fractionation during cationexchange between kaolin minerals and soil pore waters, withheavy Mg isotopes preferentially portioned into soil pore waters,leaving the residual enriched in light Mg isotopes (i.e., sampleswith higher tTh,Sr and tTh,Cs values). This observation is consistentwith enrichment of heavy Mg isotopes in some pore watersrelative to soils from Iceland (Pogge von Strandmann et al.,2012), and retention of light Mg isotopes in soils from Guade-loupe through ion-exchange on the soil exchange complex(Opfergelt et al., 2012). Hence, the removal of adsorbed Mgthrough cation exchange during extreme weathering may accountfor the even lower tTh,Mg and lighter Mg isotopic compositions ofsaprolites in the upper profile.

5.3. Behavior of Mg isotopes during continental weathering and

implications

Chemical weathering of silicate rocks involves two majorprocesses: primary mineral dissolution and secondary mineralformation. During these processes, removal of Mg by primarymineral dissolution is accompanied by incorporation of Mg intosecondary minerals, coupled with large Mg isotope fractionation.

These processes have significantly modified the Mg isotopiccomposition of the upper continental crust (Li et al., 2010).

In order to understand the general behavior of Mg isotopesduring different stages of continental weathering, we compareour Mg isotope data with those of other weathering profiles (Tenget al., 2010b; Pogge von Strandmann et al., 2012; Tipper et al.,2012) and suspended loads (Pogge von Strandmann et al., 2008b;Wimpenny et al., 2011). Two distinct zones in the d26Mg–MgOdiagram can be defined: (1) the high Mg zone (MgO42.2 wt%)corresponds to the incipient stage of weathering, in which partialdissolution of primary minerals dominates. Magnesium isotoperatios in those weathered products with high MgO content aresimilar to their parent rocks (Fig. 8). No isotope fractionationduring congruent dissolution of primary silicate minerals in theabsence of secondary mineral formation is expected (Pistiner andHenderson, 2003; Pogge von Strandmann et al., 2010). Thus,limited Mg isotope fractionation in high Mg zone is likelyattributed to dissolution of primary mineral in the incipient stageof weathering and (2) the low Mg zone (MgOo2.2 wt%) corre-sponds to the advanced or extreme stage of weathering. In thisstage, primary minerals are exhausted and secondary mineralsdominate the weathered products (Nesbitt and Wilson, 1992).The Mg isotopic composition of weathered product in this zonedisplays a large range (�0.49% to þ0.65%), and positivelycorrelates with MgO content (Fig. 8). The positive correlationsuggests that secondary mineral formation fractionates Mg iso-topes, with heavy Mg isotopes preferring secondary minerals tofluids during adsorption and incorporation of Mg into secondaryminerals, and with light Mg isotopes preferring secondary mineralsduring desorption of Mg from secondary minerals.

Better understanding of the behaviors of Mg isotopes duringcontinental weathering has important applications and may helpto interpret the large Mg isotopic variations in river waters. Forexample, river waters draining basaltic terrains have highly

variable d26Mg values, which are often different from the bedrockover which they drain (e.g., Tipper et al., 2006a, 2006b; Pogge von

Strandmann et al., 2008a, 2008b). The large variation in d26Mgvalues of river waters has been interpreted as a result of isotopicfractionation during secondary mineral formation and mixing ofMg released from isotopically distinct minerals (Tipper et al.,2006a, 2006b; Pogge von Strandmann et al., 2008a, 2008b). Here,our results suggest that adsorption and desorption of Mg on thesurface of clay minerals, besides incorporation of Mg into

R2 = 0.78

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

-80 -60 -40 -20 0 20 40

δ 26M

g (‰

)

τ Th,Cs (%)

R2 = 0.79

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

-100 -99.5 -99 -98.5 -98 -97.5 -97

δ 26M

g (‰

)

τ Th,Sr (%)

Above 3 m Below 3 m

Fig. 7. d26Mg as a function of (A) tTh,Sr and (B) tTh,Cs in the weathering profile. The

solid line represents a linear regression fitted through the data.

-0.8

-0.6

-0.4

-0.2

0.0

0.2

0.4

0.6

0.8

1.0

0.1 1 10 100

δ26M

g(‰

)

MgO (wt%)

Hainan Island, China (This study)Iceland (Pogge von Strandmann et al., 2008b)South Carolina, USA (Teng et al., 2010b)Greenland (Wimpenny et al., 2011)Alps, Swiss (Tipper et al., 2012)Iceland (Pogge von Strandmann et al., 2012)

BSE

Basalt

Diabase

R2 = 0.55

Fig. 8. (Color online) Relationship between d26Mg and MgO contents in weathered

products from Hainan Island, South Carolina, Iceland and Alps as well as

suspended loads from Iceland and Greenland. Horizontal gray bar represents

d26Mg of the bulk silicate Earth (BSE, Teng et al., 2010a). The solid line represents a

linear regression fitted through the data. See text for details.



secondary minerals, can also fractionate Mg isotopes duringsilicate weathering and potentially change the Mg isotopic com-position of rivers. Nonetheless, more experimental and fieldstudies are needed to quantify the magnitude and direction ofadsorption and desorption on Mg isotope fractionation before wecan use it as a common mechanism to explain Mg isotopicvariations during low-temperature water–rock interactions.

6. Conclusions

This study presents Mg isotopic data for a well-characterizedsaprolite profile developed on the Neogene basalt from HainanIsland in southern China. Compared to the unaltered basalts(d26Mg¼�0.36%), the overlying saprolites are strongly depletedin Mg (tTh,Mg¼�99.1% to �92.9%) and have heterogeneous Mgisotopic composition (d26Mg¼�0.49% to þ0.40%).

tTh,Mg and d26Mg values of the saprolite in the lower profileincrease with the modal abundance of kaolin minerals, suggestingthat adsorption of Mg onto kaolin minerals fractionates Mgisotopes, with preferential uptake of heavy Mg isotopes intokaolin minerals. This fractionation is mainly driven by thedifference in coordination environment between adsorbed Mg inkaolin minerals and dissolved Mg in pore water. By contrast, bothtTh,Mg and d26Mg values decrease with the modal abundance ofkaolin minerals in the upper profile, and gradually decreasetowards the surface. These Mg concentration and isotopic varia-tions indicate that a second episode of weathering has modifiedMg elemental and isotopic distribution. The negative correlationsbetween tTh,Cs, tTh,Sr and d26Mg of saprolites in this section of theweathering profile suggest that the adsorbed heavy Mg isotopesare preferentially removed from the surface of kaolin minerals bycation exchange with lower hydration energy cations.

Compilation of Mg isotopic data of weathered products revealsthat behaviors of Mg isotopes during continental weathering can beconsidered as a two-stage process: (a) limited Mg isotope fractiona-tion during primary mineral dissolution in the incipient stage ofweathering and (b) large Mg isotope fractionation during secondarymineral formation in the advanced stage of weathering.

Acknowledgments

We are grateful to Yan Hu, Yan Xiao and Xiao-Ming Liu for helpin the lab, and Lin Ma, Shui-Jiong Wang, Johnnie Chamberlin andFatemeh Sedaghatpour for stimulating discussions. The manu-script was significantly improved by the constructive reviewsfrom Philip A.E. Pogge von Strandmann and two anonymousreviewers. The efficient editorial handling of Bernard Marty isgreatly appreciated. This work was financially supported by theNational Science Foundation (EAR-0838227 and EAR-1056713).K.J.H. is partially supported by the China Scholarship Council.

References

Balistrieri, L.S., Borrok, D.M., Wanty, R.B., Ridley, W.I., 2008. Fractionation of Cuand Zn isotopes during adsorption onto amorphous Fe(III) oxyhydroxide:experimental mixing of acid rock drainage and ambient river water. Geochim.Cosmochim. Acta 72 (2), 311–328.

Barling, J., Anbar, A.D., 2004. Molybdenum isotope fractionation during adsorptionby manganese oxides. Earth Planet. Sci. Lett. 217 (3–4), 315–329.

Bas-c-etin, E., Atun, G., 2006. Adsorption behavior of strontium on binary mineralmixtures of Montmorillonite and Kaolinite. Appl. Radiat. Isot. 64 (8), 957–964.

Begum, S., Khan, F.U., Qasiar, M.A., 1998. Effect of temperature on the adsorptionof Mg2þ on kaolin from aqueous solutions and the stability constants forcomplex formation. J. Chem. Soc. Pak. 20 (1), 19–24.

Berner, R.A., Lasaga, A.C., Garrels, R.M., 1983. The carbonate–silicate geochemicalcycle and its effect on atmospheric carbon dioxide over the past 100 millionyears. Am. J. Sci. 283, 641–683.

Black, J.R., Epstein, E., Rains, W.D., Yin, Q.-Z., Casey, W.H., 2008. Magnesium-isotope fractionation during plant growth. Environ. Sci. Technol. 42 (21),7831–7836.

Bleam, W.F., McBride, M.B., 1985. Cluster formation versus isolated-site adsorp-tion. A study of Mn(II) and Mg(II) adsorption on boehmite and goethite.J. Colloid Interf. Sci. 103 (1), 124–132.

Bolou-Bi, E.B., Poszwa, A., Leyval, C., Vigier, N., 2010. Experimental determinationof magnesium isotope fractionation during higher plant growth. Geochim.Cosmochim. Acta 74 (9), 2523–2537.

Bolou-Bi, E.B., Vigier, N., Poszwa, A., Boudot, J.-P., Dambrine, E., 2012. Effects ofbiogeochemical processes on magnesium isotope variations in a forestedcatchment in the Vosges Mountains (France). Geochim. Cosmochim. Acta 87(15), 341–355.

Brennecka, G.A., Wasylenki, L.E., Bargar, J.R., Weyer, S., Anbar, A.D., 2011. Uraniumisotope fractionation during adsorption to Mn-oxyhydroxides. Environ. Sci.Technol. 45 (4), 1370–1375.

Brenot, A., Cloquet, C., Vigier, N., Carignan, J., France-Lanord, C., 2008. Magnesiumisotope systematics of the lithologically varied Moselle river basin. FranceGeochim. Cosmochim. Acta 72 (20), 5070–5089.

Carroll, D., Starkey, H.C., 1960. Effect of sea–water on clay minerals. Clays ClayMiner. 7, 80–101.

Chang, V.T.C., Makishima, A., Belshaw, N.S., O’Nions, R.K., 2003. Purification of Mgfrom low-Mg biogenic carbonates for isotope ratio determination usingmultiple collector ICP-MS. J. Anal. At. Spectrom. 18 (4), 296–301.

Cornell, R.M., 1993. Adsorption of cesium on minerals: a review. J. Radioanal. Nucl.Chem. 171 (2), 483–500.

Dessert, C., Dupr, B., Gaillardet, D., Franc-oisc, L.M., All�egre, C.J., 2003. Basaltweathering laws and the impact of basalt weathering on the global carboncycle. Chem. Geol. 202 (3–4), 257–273.

Driesner, T., Ha, T.K., Seward, T.M., 2000. Oxygen and hydrogen isotope fractiona-tion by hydration complexes of Liþ , Naþ , Kþ , Mg2þ , F� , Cl� , and Br�:a theoretical study. Geochim. Cosmochim. Acta 64 (17), 3007–3033.

Evangelou, V., 1998. Environmental Soil and Water Chemistry: Principles andApplications. John Wiley & Sons, New York 592 p.

Fantle, M.S., Tollerud, H., Eisenhauer, A., Holmden, C., 2012. The Ca isotopiccomposition of dust-producing regions: measurements of surface sedimentsin the Black Rock Desert, Nevada. Geochim. Cosmochim. Acta 87 (15),178–193.

Foster, G.L., Pogge von Strandmann, P.A.E., Rae, J.W.B., 2010. Boron and magnesiumisotopic composition of seawater. Geochem. Geophys. Geosyst. 11 (8), Q08015,http://dx.doi.org/10.1029/2010GC003201.

Gaillardet, J., Dupre, B., Louvat, P., All�egre, C.J., 1999. Global silicate weathering andCO2 consumption rates deduced from the chemistry of large rivers. Chem.Geol. 159 (1–4), 3–30.

Galy, A., Yoffe, O., Janney, P.E., Williams, R.W., Cloquet, C., Alard, O., Halicz, L.,Wadhwa, M., Hutcheon, I.D., Ramon, E., Carignan, J., 2003. Magnesium isotopeheterogeneity of the isotopic standard SRM980 and new reference materials formagnesium-isotope-ratio measurements. J. Anal. At. Spectrom. 18, 1352–1356.

Goldberg, T., Archer, C., Vance, D., Poulton, S.W., 2009. Mo isotope fractionationduring adsorption to Fe (oxy-hydr)oxides. Geochim. Cosmochim. Acta 73 (21),6502–6516.

Gu, X., Evans, L.J., 2008. Surface complexation modelling of Cd(II), Cu(II), Ni(II),Pb(II) and Zn(II) adsorption onto kaolinite. Geochim. Cosmochim. Acta 72 (2),267–276.

Hoefs, J., 2009. Stable Isotope Geochemistry, sixth edition Springer Verlag, Berlin,Heidelberg.

Icopini, G.A., Anbar, A.D., Ruebush, S.S., Tien, M., Brantley, S.L., 2004. Iron isotopefractionation during microbial reduction of iron: the importance of adsorption.Geology 32 (3), 205–208.

James, R.H., Palmer, M.R., 2000. Marine geochemical cycles of the alkali elements andboron: the role of sediments. Geochim. Cosmochim. Acta 64 (18), 3111–3122.

Joussein, E., Petit, S., Churchman, J., Theng, B., Righi, D., Delvaux, B., 2005.Halloysite clay minerals–a review. Clay Miner. 40 (4), 383–426.

Kashiwabara, T., Takahashi, Y., Tanimizu, M., Usui, A., 2011. Molecular-scalemechanisms of distribution and isotopic fractionation of molybdenumbetween seawater and ferromanganese oxides. Geochim. Cosmochim. Acta75 (19), 5762–5784.

Kinniburgh, D.G., Jackson, M.L., Syers, J.K., 1976. Adsorption of alkaline earth,transition, and heavy metal cations by hydrous oxide gels of iron andaluminum. Soil Sci. Soc. Am. J. 40 (5), 796–799.

Kosmulski, M., 2006. pH-dependent surface charging and points of zero charge: III.Update. J. Colloid Interf. Sci. 298 (2), 730–741.

Kosmulski, M., 2009. Surface charging and points of zero charge. SurfactantScience Series. The CRC Press Taylor & Francis Group, Florida, USA, vol. 145,1092 p.

Lemarchand, E., Schott, J., Gaillardet, J., 2007. How surface complexes impactboron isotope fractionation: evidence from Fe and Mn oxides sorptionexperiments. Earth Planet. Sci. Lett. 260 (1–2), 277–296.

Li, W., Beard, B.L., Johnson, C.M., 2011. Exchange and fractionation of Mg isotopesbetween epsomite and saturated MgSO4 solution. Geochim. Cosmochim. Acta75, 1814–1828.

Li, W.-Y., Teng, F.-Z., Ke, S., Rudnick, R.L., Gao, S., Wu, F.-Y., Chappell, B.W., 2010.Heterogeneous magnesium isotopic composition of the upper continentalcrust. Geochim. Cosmochim. Acta 74 (23), 6867–6884.

Ling, M.-X., Sedaghatpour, F., Teng, F.-Z., Hays, P.D., Strauss, J., Sun, W., 2011.Homogeneous magnesium isotopic composition of seawater: an excellent



geostandard for Mg isotope analysis. Rapid Commun. Mass Spectrom. 25 (19),2828–2836.

Liu, S.-A., Teng, F.-Z., He, Y., Ke, S., Li, S., 2010. Investigation of magnesium isotopefractionation during granite differentiation: implication for Mg isotopiccomposition of the continental crust. Earth Planet. Sci. Lett. 297 (3–4),646–654.

Ma, C., Eggleton, R.A., 1999. Cation exchange capacity of kaolinite. Clays ClayMiner. 47 (2), 174–180.

Ma, J.-L., Wei, G.-J., Xu, Y.-G., Long, W.-G., Sun, W.-D., 2007. Mobilization and re-distribution of major and trace elements during extreme weathering of basaltin Hainan Island, South China. Geochim. Cosmochim. Acta 71 (13), 3223–3237.

Ma, J., Wei, G., Xu, Y., Long, W., 2010. Variations of Sr–Nd–Hf isotopic systematicsin basalt during intensive weathering. Chem. Geol. 269 (3–4), 376–385.

Ma, L., Chabaux, F., Pelt, E., Granet, M., Sak, P.B., Gaillardet, J., Lebedeva, M.,Brantley, S.L., 2012. The effect of curvature on weathering rind formation:evidence from Uranium-series isotopes in basaltic andesite weathering clastsin Guadeloupe. Geochim. Cosmochim. Acta 80, 92–107.

Manceau, A., Schlegel, M.L., Musso, M., Sole, V.A., Gauthier, C., Petit, P.E., Trolard, F.,2000. Crystal chemistry of trace elements in natural and synthetic goethite.Geochim. Cosmochim. Acta 64 (21), 3643–3661.

Marcus, Y., 1988. Ionic radii in aqueous solutions. Chem. Rev. 88 (8), 1475–1498.Nesbitt, H.W., Markovics, G., Price, R.C., 1980. Chemical processes affecting alkalis

and alkaline earths during continental weathering. Geochim. Cosmochim. Acta44 (11), 1659–1666.

Nesbitt, H.W., Wilson, R.E., 1992. Recent chemical weathering of basalts. Am. J. Sci.292, 740–777.

Oi, T., Yanase, S., Kakihana, H., 1987. Magnesium isotope fractionation in cation-exchange chromatography. Sep. Sci. Technol. 22 (11), 2203–2215.

Olu-Owolabi, B.I., Ajayi, S.O., 2003. Cations adsorption on goethite–humic acidcomplex. Sci. Iran. 10 (3), 329–333.

Opfergelt, S., Georg, R.B., Delvaux, B., Cabidoche, Y.M., Burton, K.W., Halliday, A.N.,2012. Mechanisms of magnesium isotope fractionation in volcanic soil weath-ering sequences, Guadeloupe. Earth Planet. Sci. Lett. 341–344, 176–185.

Palmer, M.R., Spivack, A.J., Edmond, J.M., 1987. Temperature and pH controls overisotopic fractionation during adsorption of boron on marine clay. Geochim.Cosmochim. Acta 51 (9), 2319–2323.

Peacock, C.L., Sherman, D.M., 2005. Surface complexation model for multisiteadsorption of copper(II) onto kaolinite. Geochim. Cosmochim. Acta 69 (15),3733–3745.

Peskleway, C.D., Henderson, G.S., Wicks, F.J., 2003. Dissolution of gibbsite: directobservations using fluid cell atomic force microscopy. Am. Mineral. 88 (1),18–26.

Pistiner, J.S., Henderson, G.M., 2003. Lithium-isotope fractionation during con-tinental weathering processes. Earth Planet. Sci. Lett. 214 (1–2), 327–339.

Pogge von Strandmann, P.A.E., Burton, K.W., James, R.H., van Calsteren, P., Gislason,S.R., Sigfusson, B., 2008a. The influence of weathering processes on riverinemagnesium isotopes in a basaltic terrain. Earth Planet. Sci. Lett. 276 (1–2),187–197.

Pogge von Strandmann, P.A.E., James, R.H., van Calsteren, P., Gıslason, S.R., Burton,K.W., 2008b. Lithium, magnesium and uranium isotope behaviour in theestuarine environment of basaltic islands. Earth Planet. Sci. Lett. 274 (3–4),462–471.

Pogge von Strandmann, P.A.E., Burton, K.W., James, R.H., van Calsteren, P., Gislason, S.R.,2010. Assessing the role of climate on uranium and lithium isotope behaviour inrivers draining a basaltic terrain. Chem. Geol. 270, 227–239.

Pogge von Strandmann, P.A.E., Elliott, T., Marschall, H.R., Coath, C., Lai, Y.-J.,Jeffcoate, A.B., Ionov, D.A., 2011. Variations of Li and Mg isotope ratios in bulkchondrites and mantle xenoliths. Geochim. Cosmochim. Acta 75, 5247–5268.

Pogge von Strandmann, P.A.E., Opfergelt, S., Lai, Y.-J., Sigfusson, B., Gislason, S.R.,Burton, K.W., 2012. Lithium, magnesium and silicon isotope behaviouraccompanying weathering in a basaltic soil and pore water profile in Iceland.Earth Planet. Sci. Lett. 339–340, 11–23.

Pokrovsky, O.S., Viers, J., Freydier, R., 2005. Zinc stable isotope fractionation duringits adsorption on oxides and hydroxides. J. Colloid Interf. Sci. 291 (1), 192–200.

Riechelmann, S., Buhl, D., Schroder-Ritzrau, A., Spotl, C., Riechelmann, D.F.C.,Richter, D.K., Kluge, T., Marx, T., Immenhauser, A., 2012. Hydrogeochemistryand fractionation pathways of Mg isotopes in a continental weatheringsystem: lessons from field experiments. Chem. Geol. 300-301, 109–122.

Ryu, J.-S., Jacobson, A.D., Holmden, C., Lundstrom, C., Zhang, Z., 2011. The majorion, d44/40Ca, d44/42Ca, and d26/24Mg geochemistry of granite weathering atpH¼1 and T¼25 1C: power-law processes and the relative reactivity ofminerals. Geochim. Cosmochim. Acta 75 (20), 6004–6026.

Sawhney, B.L., 1972. Selective sorption and fixation of cations by clay minerals: areview. Clays Clay Miner. 20 (2), 93–100.

Schauble, E.A., 2011. First-principles estimates of equilibrium magnesium isotopefractionation in silicate, oxide, carbonate and hexaaquamagnesium(2þ) crys-tals. Geochim. Cosmochim. Acta 75 (3), 844–869.

Schwarcz, H.P., Agyei, E.K., McMullen, C.C., 1969. Boron isotopic fractionationduring clay adsorption from sea-water. Earth Planet. Sci. Lett. 6 (1), 1–5.

Scroth, B.K., Sposito, G., 1997. Surface charge properties of kaolinite. Clays ClayMiner. 45 (1), 85–91.

Srivastava, P., Singh, B., Angove, M., 2005. Competitive adsorption behavior ofheavy metals on kaolinite. J. Colloid Interf. Sci. 290 (1), 28–38.

Tatsuya, T., Komarneni, S., 1999. Alkali metal and alkaline earth metal ionexchange with Na-4-mica prepared by a new synthetic route from kaolinite.J. Mater. Chem. 9, 2475–2479.

Teng, F.-Z., Li, W.-Y., Ke, S., Marty, B., Dauphas, N., Huang, S., Wu, F.-Y., Pourmand,A., 2010a. Magnesium isotopic composition of the Earth and chondrites.Geochim. Cosmochim. Acta 74 (14), 4150–4166.

Teng, F.-Z., Li, W.-Y., Rudnick, R.L., Gardner, L.R., 2010b. Contrasting lithium andmagnesium isotope fractionation during continental weathering. Earth Planet.Sci. Lett. 300 (1–2), 63–71.

Teng, F.-Z., Wadhwa, M., Helz, R.T., 2007. Investigation of magnesium isotopefractionation during basalt differentiation: implications for a chondriticcomposition of the terrestrial mantle. Earth Planet. Sci. Lett 261 (1–2), 84–92.

Teppen, B.J., Miller, D.M., 2006. Hydration energy determines isovalent cationexchange selectivity by clay minerals. Soil Sci. Soc. Am. J. 70 (1), 31–40.

Tipper, E.T., Galy, A., Bickle, M.J., 2006a. Riverine evidence for a fractionatedreservoir of Ca and Mg on the continents: implications for the oceanic Ca cycle.Earth Planet. Sci. Lett 247 (3–4), 267–279.

Tipper, E.T., Galy, A., Gaillardet, J., Bickle, M.J., Elderfield, H., Carder, E.A., 2006b.The magnesium isotope budget of the modern ocean: constraints fromriverine magnesium isotope ratios. Earth Planet. Sci. Lett. 250 (1–2), 241–253.

Tipper, E.T., Galy, A., Bickle, M.J., 2008. Calcium and magnesium isotope systema-tics in rivers draining the Himalaya–Tibetan-Plateau region: lithological orfractionation control? Geochim. Cosmochim. Acta 72 (4), 1057–1075.

Tipper, E.T., Gaillardet, J., Louvat, P., Capmas, F., White, A.F., 2010. Mg isotopeconstraints on soil pore-fluid chemistry: evidence from Santa Cruz, California.Geochim. Cosmochim. Acta 74 (14), 3883–3896.

Tipper, E.T., Lemarchand, E., Hindshaw, R.S., Reynolds, B.C., Bourdon, B., 2012.Seasonal sensitivity of weathering processes: hints from magnesium isotopesin a glacial stream. Chem. Geol. 312–313, 80–92.

Trolard, F., Tardy, Y., 1987. The stabilities of gibbsite, boehmite, aluminousgoethites and aluminous hematites in bauxites, ferricretes and laterites as afunction of water activity, temperature and particle size. Geochim. Cosmo-chim. Acta 51 (4), 945–957.

Wasylenki, L.E., Rolfe, B.A., Weeks, C.L., Spiro, T.G., Anbar, A.D., 2008. Experimentalinvestigation of the effects of temperature and ionic strength on Mo isotopefractionation during adsorption to manganese oxides. Geochim. Cosmochim.Acta 72 (24), 5997–6005.

Wasylenki, L.E., Weeks, C.L., Bargar, J.R., Spiro, T.G., Hein, J.R., Anbar, A.D., 2011.The molecular mechanism of Mo isotope fractionation during adsorption tobirnessite. Geochim. Cosmochim. Acta 75 (17), 5019–5031.

Weaver, C.E., Pollard, L.D., 1973. The chemistry of clay minerals, Development inSedimentology. Elsevier, Amsterdam 213 p.

White, A.F., Schulz, M.S., Stonestrom, D.A., Vivit, D.V., Fitzpatrick, J., Bullen, T.D.,Maher, K., Blum, A.E., 2009. Chemical weathering of a marine terracechronosequence, Santa Cruz, California. Part II: Solute profiles, gradients andthe comparisons of contemporary and long-term weathering rates. Geochim.Cosmochim. Acta 73 (10), 2769–2803.

Wilson, M.J., 1999. The origin and formation of clay minerals in soils: past, presentand future perspectives. Clay Miner. 34 (1), 7–25.

Wimpenny, J., Burton, K.W., James, R.H., Gannoun, A., Mokadem, F., Gıslason, S.R.,2011. The behaviour of magnesium and its isotopes during glacial weatheringin an ancient shield terrain in West Greenland. Earth Planet. Sci. Lett.304 (1–2), 260–269.

Wimpenny, J., Gıslason, S.R., James, R.H., Gannoun, A., Pogge Von Strandmann,P.A.E., Burton, K.W., 2010. The behaviour of Li and Mg isotopes during primaryphase dissolution and secondary mineral formation in basalt. Geochim.Cosmochim. Acta 74 (18), 5259–5279.

Yang, W., Teng, F.-Z., Zhang, H.-F., 2009. Chondritic magnesium isotopic composi-tion of the terrestrial mantle: a case study of peridotite xenoliths from theNorth China craton. Earth Planet. Sci. Lett 288 (3–4), 475–482.

Yang, W., Teng, F.-Z., Zhang, H.-F., Li, S.-G., 2012. Magnesium isotopic systematicsof continental basalts from the North China craton: implications for tracingsubducted carbonate in the mantle. Chem. Geol. 328, 185–194.

Young, E.D., Galy, A., 2004. The isotope geochemistry and cosmochemistry ofmagnesium. Rev. Mineral. Geochem. 55 (1), 197–230.

Zhang, L., Chan, L.-H., Gieskes, J.M., 1998. Lithium isotope geochemistry of porewaters from ocean drilling program Sites 918 and 919, Irminger Basin.Geochim. Cosmochim. Acta 62 (14), 2437–2450.

Zhu, B., Wang, H., 1989. Nd–Sr–Pb isotopic and chemical evidence for thevolcanism with MORB-OIB source characteristics in the Leiqiong Area. China.Geochim. 18 (3), 193–201.


Author's personal copy Adsorption-and desorption-controlled magnesium isotope fractionation during...

Documents

Transcript of Author's personal copy Adsorption-and desorption-controlled magnesium isotope fractionation during...