Bergaya F. Handbook of Clay Science

Подождите немного. Документ загружается.

chain standing at an average angle of 521 with respect to the layer plane. In the case

of ODTMA, the invariance in NEXAF S amplitude with respect to measurement

angle was interpreted in terms of complete disorder of the alkyl chain, although the

opposite possibility of perfect or der with a tilt angle at 54.71 could not be ruled out.

The difference in orientation between the di- and mono-alkyl ammonium ions was

probably a result of the packing order and density of alkyl chains on the external

mica surface with DODMA having twice the chain density per formula unit than

ODTMA. Given that the layer charge of mica can be nearly twice that of low-charge

vermiculites, these results are not surprising. In the case of intercalated hexadecyltri-

methylammonium (HDTMA) in vermiculite (X ¼ 1:23 and 1.5e

–

) the alkyl chains on

243

R* (Å)

(k)

Fig. 12.3.14. Fourier transforms calculated from FEFF showing the FeyM

shell of

smectites in which the ﬁrst cation shell is varied between Fe and Al. Non-centrosymmetric

structure assumed for N

¼ 0, N

¼ 3 and N

¼ 1, N

¼ 2 shells; centrosymmetric struc-

ture assumed for N

¼ 2, N

¼ 1 and N

¼ 3, N

¼ 0 shells. FeyFe and FeyAl inter-

atomic distances were held constant at d ¼ 307 pm and d ¼ 301 pm, respectively.

Chapter 12.3: X-ray Absorption Spectroscopy826

opposing layers will protrude into the inter layer space, doubling the chain packing

density as compared with a single surface (Slade and Gates, 2004b ). Using scaled

models of intercalated HDTMA-vermiculites based on powder and single-crystal

X-ray transmission diffraction data, Slade and Gates (2004b) deduced an angle of

repose of about 501. By analogy the octadecyl chains (of ODTMA) on the external

surfaces of mica (Brovelli et al., 1999) might have a highly close-packed two-

dimensional order with the chains making an angle of 54.71 to the surface. Slade and

Gates (2004b) also found that at loadings satisfying the vermiculite layer charge,

HDTMA assembled into a 3 a b super cell commensurate with the dimens ions of

the silicate surface. When uptake of HDTMA-Br was in excess of the layer charge,

the alkyl tails of the HDTMA-Br ion pairs caused closest packing and rotation ,

destroying the commensurate structure, although transmis sion diffraction still in-

dicated a highly ordered organic interlayer structure. Obviously, this is a potenitally

fruitful area of research with implications for nanotechnology and self-assembly, and

for which synchrotron-based techniques will be very useful.

SAz-1

SWy-2

Fig. 12.3.15. Models for the two-dimensional distribution of cations in the octahedral sheet

of montmorillonites SWy-2 and SAz-1 based on the XAFS analyses of Vantelon et al. (2003),

but modiﬁed using chemical and IR data (Gates et al., 2002; Gates, 2004).

12.3.3. Orientation of Intercalated Organic Molecules 827

Lin et al. (1997) studied the location of Br



counterions sorbed during the self-

assembly of organic mono- and multi-layers on Mo/Si surfaces, using X-ray standing

wave techniques. The standing wave coherence indicated a narrow distribution

of positions for the Br

–

anions, located about 1 nm above the surface. These inner

–

positions remained unchanged when multilayers of organic compounds were

sorbed, and the distances from the surface of additional layers of Br

–

were found to

be consistent with the distances between positively charged nitro groups of the or-

ganic molecules (bromopropyl trichlorosilane derivatives). These results strongly

suggest that self-assembly produces ordered superlattice structures, as Slade and

Gates (2004b) observed for HDTMA-vermiculites. Sorption of the halide anion

along with the organic cati on within the structures of organo-clays clearly inﬂuence

many of their practical applications, as shown by the thermal stability of organo-clay

complexes (Davis et al., 2004).

Finally, the studies by Fisher et al. (1998) and Ha

hner et al. (1996b) provided

information on the attachment of dye molecules to mineral surfaces. Using C and N

K NEXAFS Ha

hner et al. (1996b) showed that the fused aromatic ring plane of

methylene blue resides at steep angles (651–701) with respect to the mica surface.

Fischer et al. (1998) observed similarly for crystal violet and malachite green. These

results indicate that the speciﬁc surface areas of smectites, determined by sorption of

methylene blue, are overestimated if the dye is assumed to lie with its aromatic rings

ﬂat to the silicate surface. The steep angle of inclination may be due to the high

packing density of methylene blue on high-charge micas. Visible spectroscopic stud-

ies ( Bujda

k and Komadel, 1997; Bujda

k et al., 2001) showed that the inter action of

methylene blue molecules with reduced-charge smectites (with a layer charge o0.25

that of micas) is more complex. When the layer charge is very low, methylene blue is

intercalated as oligomers by van der Waals interactions with a residence time pro-

portional to layer charge. Eventually these oligomers are protonated and degrade to

isolated monomers at the surface. A steeply inclined orientation of the oligomers,

with the aromatic rings lying parallel to one another, would favour dye–dye electron

transfer.

12.3.4. SPATIAL RESOLUTION WITH MICROPROBE XAFS

Micro-XAFS (m-XAFS, or speciﬁc applications such as m-XRF, m-EXAFS,

m-XANES or m-XRD) can now achieve spatial resolution of o1 mm

with respect to

chemical interactions between metal sorbates on minerals (Bajt et al., 1993, 1994,

1995; Hunter and Bertsch, 1998; Manceau et al., 2002a), plant material (Schulze et

al., 1995a, 1995b; Punshon et al., 2003) and animal material (Hunter et al., 1997).

Spatially resolved XRF studies showed that the distribution of trace metals at

mineral surfaces inﬂuences their incorporation during crystallisat ion (Hunter and

Bertsch, 1998; Bertsch and Hunter, 2001; Manceau et al., 2002a). Trace metals

also inﬂuence mineral formation during pedogenesis (Manceau et al., 2002b;

Chapter 12.3: X-ray Absorption Spectroscopy828

Strawn et al., 2002). The reviews by Schulze et al. (1999), Bertsch and Hunter (2001),

Fenter et al. (2002) and Hirshchmugl (2002) provide excellent background on the

various applications, as well as the pros and cons, of synchrotron-based microprobe

XAFS, X-ray and infrared spectromicroscopy.

12.3.5. XAFS STUDIES OF REACTIVITY OF CLAYS AND CLAY

MINERALS

Synchrotron-based XAFS made an important contribution to environmental sci-

ence in the past 20 years because of its ability to assess structural, compositional and

reaction mechanisms under in situ conditions. Such applications include the study of

phosphates (Hest erberg et al., 1999) or biogenic silica (Gehlen et al., 2002). Because

the EXAFS phenomenon occurs at a much faster time-scale than many processes

such as diffusion, exchange, sorption, precipitation and even electron transfer,

EXAFS can provide nearly instantaneous ‘snapshots’ of these reactions, especially

those involving chemically unstable intermediate minerals. Recent applications in-

clude assessing the structure of hydrated mineral surfaces (Eng et al., 2000), and that

of the diffuse electric double layer (Sturchio et al., 1997). Studies of the sorption of

heavy metals by phyllosilicates are summarised in Table 12.3.4. The interested reader

is referred to Hayes and Katz (1991), Schulze et al. (1999) and Fenter et al. (2002)

for comprehensive reviews on the application of various synchrotron-based tech-

niques (ﬂuorescence, reﬂectivity, etc.) to study the reactivity of minerals and mineral

surfaces.

Table 12.3.4 shows that the application of P-EXAFS to the sorption of various

metals by smectites provided an insight into the types of complexes formed. The

examples described below illustrate that unambiguous interpretation of XAFS ex-

periments is very difﬁcult. They also show the importance of a solid understanding of

the mineral phase involved.

A. Exchange/Sorption Reactions at the Smectite– Water Interface

The terms ‘outer-sphere’ (OS) and ‘inner-sphere’ (IS) are often used in the non-clay

mineral literature to describe two distinctly different sorption mechanisms. OS im-

plies an indirect electrostatic bonding where a coordination sphere surrounds the

sorbed species that is not directly bound to the surface other than by hydrogen

bonding or electrostatic attr action. Hydrated exchangeable cations occupying in-

terlayer sites in smectites are examples of OS sorption complexes. Although the

exchangeable cations have more afﬁnity for the surface than the diffuse swarm of

ions in the electric double layer, both are considered OS sorption complexes because

the interaction is electrostatic. An IS sorption complex is one where direct coor-

dination bonding exists between the sorbed species and the surface, usually but not

necessarily through a surface ligand. The sorption of phosphate or citrate at the

12.3.5. XAFS Studies of Reactivity of Clays and Clay Minerals 829

(continued on next page)

Table 12.3.4. Summary of metals and metaloid ion sorption complexes at the phyllosilicate/solution interface studied by XAFS

spectroscopy. Individual references should be consulted for exact conditions of pH, ionic strength, metal concentration, surface

coverage and other experimental variables. See Brown and Sturchio (2002) for a more complete listing of sorption mechanisms

involving related minerals

Metal Layer silicate Complex reported Reference

Co(II) Kaolinite IS edge sites Charlet and Manceau (1994)

IS multinuclear, bidentate O’Day et al. (1994a, 1994b)

Chen and Hayes (1999)

OS low pH charge sites Chen and Hayes (1999),

Thompson et al. (1999, 2000)Co(II)Al LDH precipitate

Illite OS low pH charge sites Chen and Hayes (1999)

Co(II)Al LDH precipitate

Hectorite OS low pH charge sites Papelis and Hayes (1996) , Chen

and Hayes (1999)

Co(II)Al LDH precipitate Chen and Hayes (1999)

IS edge sites Schlegel et al. (1998, 1999a,

1999b)

Montmorillonite OS low pH, charge sites Papelis and Hayes (1996)

IS polynuclear (high pH)

Co(II)Al LDH precipitate

Sepiolite IS exchange site Fukushima and Okomoto (1987)

IS edge site Manceau and Decarreau (1988),

Charlet and Manceau ( 1994)

Ni(II) Kaolinite Ni(II)Al LDH precipitate Eick and Fendorf (1998)

Pyrophyllite Ni(II)Al LDH precipitate Scheidegger et al. (1996b, 1997,

1998), Elzinga and Sparks

(1999), Ford et al. (1999b),

Scheinost and Sparks (2000),

Scheckel et al. (2000)

Chapter 12.3: X-ray Absorption Spectroscopy830

Table 12.3.4. (Continued )

Metal Layer silicate Complex reported Reference

Montmorillonite Ni(II)Al LDH precipitate Scheidegger et al. (1996, 1997,

1998)

IS edge sites Elzinga and Sparks (1999) ,

hn et al. (2001, 2002b, 2003)

Pyrophyllite Smectite mix Ni(II)Al LDH precipitate & IS

edge sites

Elzinga and Sparks (1999)

Talc a-Ni(OH)

precipitate Scheinost and Sparks (2000),

Scheckel et al. (2000)

Soil clay Ni(II)Al LDH precipitate Roberts et al. (1999)

Cu(II) Muscovite IS at defects Farquhar et al. (1996, 1997)

Biotite OS and IS Farquhar et al. (1997)

Montmorillonite OS & IS edge sites monodentate Morton et al. (2001)

Zn(II) Pyrophyllite Zn(II)Al LDH precipitate Ford and Sparks (2000)

Hectorite IS edge sites Manceau et al. (2000c) , Schlegel

et al. (2001a, 2001b)OS exchange sites

Schlegel et al. (2001b)

As(III) Kaolinite Oxidation to As(V) Foster et al. (1998b)

Sr(II) Kaolinite OS/IS edge sites (low

concentration)

Chen et al. (1998), Sahai et al.

(2000), O’Day et al. (2000b),

Parkman et al. (1998)

Illite OS Chen et al. (1998), Chen and

Hayes (1999)

Hectorite OS Chen et al. (1998), Chen and

Hayes (1999)

Montmorillonite OS Chen et al. (1998), Chen and

Hayes (1999)

(continued on next page)

12.3.5. XAFS Studies of Reactivity of Clays and Clay Minerals 831

Table 12.3.4. (Continued )

Metal Layer silicate Complex reported Reference

Cd(II) Muscovite OS Farquhar et al. (1997)

Biotite OS, IS Farquhar et al. (1997)

I(-I) Biotite IS (IO



sorbed) Fuhrmann et al. (1998)

Cs(I) Montmorillonite OS hydrated, interlayer Kemner et al. (1997)

Illite OS dehydrated, interlayer Kemner et al. (1997)

Phlogopite OS dehydrated, interlayer Kemner et al. (1997)

Pb(II) Montmorillonite OS low I, pHo6.5 Strawn and Sparks (1999)

OS/IS high I, pH>6.5

Th(IV) Montmorillonite IS bidentate (low concentration) Da

hn et al. (2002a)

Th(OH)

precipitate (high

concentration)

U(VI) Kaolinite IS multiple sites Thompson et al. (1998)]

Montmorillonite OS interlayer (low pH) Sylwester et al. (2000a, 2000b)

IS neutral pH

Hydrobiotite OS interlayer, IS when dry Hudson et al. (1999)

Vermiculite OS interlayer, IS when dry Hudson et al. (1999)

Muscovite U(VI) hydroxide precipitate Moyes et al. (2000)

Chapter 12.3: X-ray Absorption Spectroscopy832

broken edges of smectit e particles would be an example of an IS sorption complex

since the anion forms a coordination bond with one or more oxygen atoms exposed

at the octahedral and tetrahedral edge site. IS sorption complexes can occur as

isolated units (mononuclear) or in clusters (multinuclear) as with polymers. IS com-

plexes can be mono-, bi-, tri-, tetra- or multidentate, implying the formation of 1, 2,

3, 4 or more bonds, respect ively, connecting the sorbed species with the surface.

Two other variants of OS and IS exist (i) ‘ﬁxed’ cations in the (collapsed) in-

terlayers of illit es, micas, and some smectites are considered to be OS as the forces

holding the cation are electrostatic and physical in nature, and not covalent , even

though the ﬁxed cation lost its waters of hydration; and (ii) layer substituted cations

which enter the structure either by exchange, diffusion or coprecipitation. Layer

substitution is considered different from simple IS sorption as it generally involves

multidentate bonding.

Table 12.3.4 illustrates that sorption complexes formed by transition metals at

smectite surfaces are highly dependent on experimental conditions. Generalisations

to conditions differing strongly from those reported should be made with care, but

trends can be discerned. For example, monovalent (e.g., Cs

) and certain divalent

cations (Co

,Cd

,Sr

) form OS complexes in the interlayers under low pH or

low ionic strength (Papelis and Hayes, 1996; Farquhar et al., 1997; Kemner et al.,

1997; Chen and Hayes, 1999; Strawn and Sparks, 1999). The same divalent cations

may form IS sorption complexes at clay mineral edges under high pH or ionic

strength conditions (Papelis and Hayes, 1996; Farquhar et al., 1997; Kemner et al.,

1997; Schlegel et al., 1999a, 1999b; Strawn and Sparks, 1999). Other divalent cations,

such as Cu

and Zn

appear to form mostly IS sorption complex es, but were

shown to sorb as OS complexes for a short time (Farquhar et al., 1997; Morton et al.,

2001; Schlegel et al., 2001a, 2001b). Still other divalent cations such as Ni

only

appear to form IS sorption complexes, regardless of pH and ionic strength. Indeed,

these metal ions also react strongly with layer silicates that have no permanent (pH-

independent) charge (Scheidegger et al., 1996a, 1996b, 1997, 1998). Finally, Th

was shown to be IS at low electrolyte concentrations, but OS at high electrolyte

concentrations (Da

hn et al., 2002b).

Interestingly, these same metals and metaloids behave differently with respect to

other mineral surfaces from what was described above for smectites. Thus, the fol-

lowing divalent metals generally form isolated IS complexes at low concentra tions,

but multinuclear IS complexes at higher sorption densities: Co(II) (Chisholm-Brause

et al., 1989a, 1990a, 1991; Towle et al., 1995, 1999a, 1999b), Pb(II) (Chisholm-Brause

et al., 1989b, 1990b; Roe et al., 1991; Trainor et al., 2002; Bargar et al., 2004), Cu(II)

(Weesner and Bleam, 1997; Xia et al., 1997; Cheah et al., 1998, 1999, 2000; Bochatay

et al., 1997) and Zn(II) (Trainor et al., 1999; Trivedi et al. 2001a, 2001b; Roberts et

al., 2003). The reasons for these differences in sorption complexes are probably as

varied as the number of studies. However, it appears that the greater the tendency

for a metal or metaloid to undergo hydrolysis, the greater its tendency to form IS

sorption complexes. Metal hydrolysis reactions are strongly de pendent on pH, thus

12.3.5. XAFS Studies of Reactivity of Clays and Clay Minerals 833

the formation of sorption complexes is sensitive to pH. The following two examples

of XAFS studi es of hydrolysing and non-hydrolysing cations at the surfaces of

smectite further illustrate the complexity of OS or IS sorption.

In an EXAFS study of Cs–Ca competition at exchange sites on montmorillonite

(SWy-1), illite (Silver Hill) and phlogopite mica, Kemner et al. (1997) found that at

Cs loadings of 100% of the cation exchange capacity (CEC), Cs L EXAFS spectra

showed limited long-range order. On the other hand, at low Cs loading s (e.g., 10%

CEC), the Cs environment showed strong long-range order up to 600 pm. The 10%

CEC loadings gave similar FT amplitudes of the atomic shells for illite and mica, but

3–4 times less intense for montmorillonite. The Cs L EXAFS FT for montmorillo-

nite with 10% CEC Cs also showed a shift of the FT peaks to higher R*. These

observations were interpreted in terms of Cs occupying interlayer sites for which the

extent of interlayer collapse increases in the sequence montmorillonite>illite>mica.

Although the authors did not venture beyond that explanation, it seems reasonable

to suggest that Cs occupies hydrated OS coordination environments in the inter-

layers of hydrated montmorillonite, but dehydrated OS (or ﬁxed interlayer) sites in

illite and mica. (Note that Brown and Sturchio (2002) preferred to label ﬁxed in-

terlayer sorption as IS, but in reality, the attraction is electrostatic and the immo-

bilisation of the cation between the smectitic layers is due primarily to steric

hindrance.)

What else might the Cs L EXAFS data tell us? Recall that both illite and mica have

a higher total layer charge as well as considerable Al

for Si

substitution in the

tetrahedral sheets as compared with montmorillonite. Thus, the enhanced structural

order of Cs at low CEC coverage, observed by Cs L EXAFS, is a function of the

probability of exchangeable Cs occurring near sites of tetrahedral substitution, where

a local site speciﬁcity may exist. Comparison of the layer charges (montmorillonite

XE0.75 about 25% of which is in the tetrahedral sheet; illite XE1.5 of which 67% is

tetrahedral; mica XE2 of which 67% is tetrahedral) indicates that tetrahedral-

to-total charge ratios agree well with the reported amplitude differences in the Cs L

EXAFS spectra (Kemner et al., 1997). The greater long-range order of Cs in illite and

mica are due to Cs residing between two adjacent layers within a distinct site, prob-

ably above the ditrigonal cavity near sites of Al

substitution in the tetrahedral

sheet. With an anhydrous radius of 196 pm, Cs can partially sit within the ditrigonal

cavities of dioctahedral smectites (Gu

ven, 1988). The high layer charge density, and

hence the high ﬁeld strength at the basal surfaces, of illite and mica cause dehydration

of most monovalent cations occupying any interlayer exchange sites, and ultimately

interlayer collapse. Thus, the Cs L EXAFS results of Kemner et al. (1997) clearly

differentiate between a hydrated (in montmorillonite) and a dehydrated (in illite and

mica) exchangeable cation at low CEC coverage. This interpretation is in agreement

with previous NMR interpretations (Weiss et al., 1990). A worthwhile extension

would be to study the interlayer structures of La

-orBa

-smectites, the structures

of which, at least for vermiculites, were well characterised by conventional powder

X-ray transmission techniques (Telleria et al., 1977; Slade et al., 1998).

Chapter 12.3: X-ray Absorption Spectroscopy834

As a secon d example, Da

hn et al. (2002a) studied the sorption of Th by mont-

morillonite at under- (10

–6

–10

–4

M, pH 2–3) and super-saturated (10

–5

–10

–4

pH 5) conditions (with respect to ThO

). They precluded the formation of OS

(exchange) complexes by includ ing 0.1 M NaClO

. The Th K EXAFS of the su-

persaturated Th -mon tmorillonite most resemble d poorly crystalline Th(OH)

,but

at under-satur ated condi tion s, their spectral analysis was consistent with the for-

mation of IS edge-sharing bi dentate complexes. The analysis clearly s howed that (i)

IS complexes of Th formed at low electrolyte concentrations with Si making up the

neighbouring cation shell at R*E380–390 pm, and (ii) the Th–Si interatomic dis-

tance was consistent with edge-shar ing polyhedra. The results were also consistent

with those reported by O

stols et al. (1997) for the sorption of Th by poorly crys-

talline Si. The lack of pleochroism, however, implied that Th did not form IS

complexes at the smec tite surface, and they thus prop osed that Th sorption com-

plexes at Si/Al edges were disordered with respect to the montmorillonite surface.

They discounted the possibili ty of Th sorbing to a poorly crystalline SiO

phas e

beca use the measured Si release rates wer e no t affected by the addi tion of Th.

However, they did report that solubl e Si increased continually throu ghout the ex-

periment , whereas soluble Al remained consta nt. This was taken as an indi cation

that silica precipi tation was unlikely under the conditi ons of their s tudy. One sourc e

of Si considered by Da

hn et al. (2002a) was poorly crystalline Si O

, but it escaped

detection by XRD in the bulk montmorillonite (STx-1) used. Therefore, Da

hn et al.

(2002a) preferred the interpretation that Th for med bidentate, edge-shari ng com -

plexes with Si and Al at the edges of montmorillonite that unde rwent incongruent

dissolution. Obviously, fur ther wo rk is required to decipher the a mbiguities of Th

sorption by mont morillonite.

Of the oxyanions (e.g., As(V), As(III), Se(VI), Se(IV) and U(VI)), only U(VI)

sorption by smectite (Chisholm-Brause et al., 1992; Dent et al., 1992; Chisholm -

Brause et al., 1994; Peterson et al., 1996; Giaquinta et al., 1997a, 1997b; Sylwester

et al., 2000; Ritherdon et al., 2003), hydrob iotite and vermiculite (Hudson et al.,

1999), muscovite mica (Moyes et al., 2000) and kaolin ( Thompson et al., 1998) was

studied. This is because swelling uranyl–clay complexes are potent photo-oxidisers

(Suib and Carrado, 1985; Duff et al., 1999). The aqueous uranyl oxyanion (UO

)

forms OS complexes (exchangeable cation) at low pH. At pH 5–6, uranyl-hydroxide

species begin to form (Toth and Begun, 1981). The hydrolysis of UO

at higher pH

results in the formation of IS bidentate complexes in the presence of competing ions

(Sylwester et al., 2000). At low pH the U L EXAFS spectra of U(VI)-montmorillo-

nite resemble those of aqueous uranyl (Chisholm -Brause et al., 1994; Peterson et al.,

1996; Giaquinta et al., 1997a, 1997b; Sylwester et al. 2000), but at higher pH (6.4) the

spectra more closely resemble U(VI) sorption complexes on silica or alumina (IS

bidentate complex) (Reich et al., 1998; Allard et al., 1999; Sylwester et al., 2000). The

composite evidence indicates that U(VI)-surface complexes are commonly formed

with montmorillonite and most minerals (Thom pson et al., 1997), although Duff

et al. (2002) showed that U(VI) sorption by goethite also involved layer substitution

12.3.5. XAFS Studies of Reactivity of Clays and Clay Minerals 835