Bergaya F. Handbook of Clay Science

Подождите немного. Документ загружается.

Handbook of Clay Science

Edited by F. Bergaya, B.K.G. Theng and G. Lagaly

Developments in Clay Science, Vol. 1

755

Chapter 12.1

SSBAUER SPECTROSCOPY OF CLAYS AND CLAY

MINERALS

E. MURAD

Bayerisches Landesamt fu

¨r

Umwelt, D-95603 Marktredwitz, Germany

12.1.1. HISTORICAL BACKGROUND

The earliest Mo

ssbauer data on clay minerals sensu stricto (clay-size phyllosil-

icates) were published ( Malden and Mead s, 1967; Weaver et al., 1967) less than a

decade after the appearance of the ﬁrst report on the Mo

ssbauer effect (Mo

ssbauer,

1958a, 1958b). Since then Mo

ssbauer spectroscopy was widely used to study and

characterise clay-related materials. The Mo

¨s

sbauer Mineral Handbook (Stevens et al.,

1998) lists 68 publications that refer to montmorillonite, 46 to chlorite, and 44 each

to glauconite and nontronite. In addition, the handbook contains 132 references to

hematite and 112 references to goethi te, two of the most common accessory con-

stituents of soils and raw clays. Almost all of these entries were concerned with

Fe,

and the present discussion will be conﬁned to this nuclide.

Although there are still some discrepancies on speciﬁc details such as data eval-

uation, Mo

ssbauer spectroscopy is now routinely used for the characterisation of the

oxidation state of iron in clays and clay minerals. In favourable cases the coordination

of iron can also be deduced and, where applicable, the magnetic properties induced by

the presence of iron in the structure. For some minerals (including the majority of iron

oxides), Mo

ssbauer spectroscopy can also be used to identify the actual species.

12.1.2. BASIC PRINCIPLES

ssbauer spectroscopy involves the recoil-free emission and absorption of

gamma rays. A principal feature of the technique is that it only ‘sees’ the nuclide

under survey. Thus the sole effects arising from the presence of other elements in

samples of complex chemical composition are dilution and absorption. The vast

majority of Mo

ssbauer spectra are taken in the transmission mode. Here a source

(

Co for iron spectra) emitting gamma rays of the appropriate energy is period-

ically moved over a succession of velocities, and the radiation transmitted by the

sample (‘absorber’) is registered as a function of the source velocity.

DOI: 10.1016/S1572-4352(05)01027-5

ssbauer spectra result from the so-called hyperﬁne interactions between the res-

onant nuclei and their electric and magnetic environments. The principal hyperﬁne

interactions are (1) the electric monopole interaction, giving rise to the isomer shift (d);

(2) the electric dipole interaction, leading to the quadrupole splitting (D); and (3) the

magnetic hyperﬁne interaction when a magnetic hyperﬁne ﬁeld (B

) acts at the nuclei of

theresonantatoms.Theisomershiftistheshift of the centroid of the spectrum from

zero velocity, and is given relative to either the source or some standard material—in the

case of

Fe usually metallic iron. The quadrupole splitting is the separation of the two

lines of a

Fe doublet. Both isomer shift and quadrupole splitting are customarily given

in terms of the source velocity in mm/s. Isomer shifts are related to the oxidation state of

iron and may provide information on iron coordination, whereas the quadrupole split-

ting provides a measure for Fe

site distortion. An intrinsic (‘hyperﬁne’) or extrinsic

magnetic ﬁeld splits an Fe

ssbauer spectrum into a sextet, the spread of which is

proportional to the ﬁeld, and is usually expressed in kilooersteds or tesla (1 T ¼ 10 kOe).

In complex spectra the relative intensities of individual components are often taken

as proportional to the corresponding site population. This relationship, however, only

applies as a ﬁrst approximation; for example, Fe

exhibits a lower recoil-free fraction

than Fe

(De Grave and Van Alboom, 1991). Additional information may be ob-

tained from the widths and shapes of the lines. The ideal Mo

ssbauer line shape is the

Lorentzian, and experimental data are often computer-ﬁtted on this basis. However,

deviations from Lorentzian shape may occur because of variations of local environ-

ments or ﬂuctuations of parameters, to name just two factors. In such cases, the data

may have to be ﬁtted using other functions, for example, the Voigtian (a convolution

of Gaussian and Lorentzian functions) or distributions of Lorentzians.

Numerous textbooks on the Mo

ssbauer effect and its applications were published

(for example Gonser (1975) an d Gibb (1976) to name two of the better known

‘classics’, and more recently a book by Murad and Cashion (2004) that focuses on

the Mo

ssbauer spectra of materials formed on the earth’s surface), and for more

information on the Mo

ssbauer effect than can be included here, the reader is referred

to these and similar sources.

Depending upon sample structure and composition,

Fe Mo

ssbauer spectra may

consist of one or more singlet(s), quadrupole-split doubl et(s), and magnetically split

sextet(s). Singlets develop only in the case of cubic symmetry around Fe

, and

hence are not observable in the spectra of phyllosilicates. Sextets, arising from iron in

magnetically ordered materials, are only observed for extremely iron-rich phyllo-

silicates at low temperatures (p10 K), and for iron oxides of sufﬁciently good crys-

tallinity and chemical purity. The Mo

ssbauer spectra of phyllosilicates thus generally

consist of one doublet for iron in each oxidation state on every structural site.

12.1.3. MO

SSBAUER SPECTRA OF SELECTED CLAY MINERALS

Both divalent and trivalent iron can substitute for aluminium and magnesium in the

structures of phyllosilicates. The extent of substitution depends on both the availability

Chapter 12.1: Mo

ssbauer Spectroscopy of Clays and Clay Minerals756

of iron during mineral formation and the speciﬁc mineral structure (i.e., the numbers

and geometries of structural sites into which iron can be incorporated).

Although kaolinites, in general, show little variation of chemical composition,

some substitution of iron for aluminium in octahedral sites can occur. As a result,

the Mo

ssbauer spectra of kaolinite are deceptively simple: a single Fe

doublet or,

as observed in the early study by Malden and Meads (1967), each one Fe

and

doublet. Besides conﬁrming this observation, subsequent studies of kaolinite—

the Mo

ssbauer Mineral Handbook (Stevens et al., 1998) lists a total of 20 references—

indicated that octahedral substitution by Fe

is more common than that by Fe

A detailed investigation of 22 kaolinites from different genetic environments and

with widely varying iron contents (Murad and Wagner, 1991), however, indicates

that it is not uncommon to have Fe

substituting for Al

. Indeed, in one sample

(KGa-1) the concentration of Fe

even exceeds that of Fe

. This study also

conﬁrms the work by Fysh et al. (1983) that slow paramagnetic relaxation can have a

marked inﬂuence on the Mo

ssbauer spectra of kaolinite as indicated by a non-

speciﬁc broadening of the paramagnetic doublet at room temperature and the

development of a magnetic component at 4.2 K (Fig. 12.1.1).

The 2:1 layer silicates offer a larger variety of possibilities for the incorporation of

iron than their 1:1 counterparts. Not only can iron occupy two distinctly different

octahedral sites (cis–andtrans–OH) in 2:1 structures, but Fe

can sometimes also

substitute for silicon in tetrahedral sites. The problem of distinguishing iron between

cis and trans sites, and also between tetrahedral and octahedral Fe

by Mo

ssbauer

spectroscopy was hotly debated. In many instances, the spectra were ﬁtted with an

increasing number of doublets until the ﬁtted and experimental data matched

adequately well, after which the resulting spectral components were assigned to struc-

tural sites. High-quality spectra of well-crystallized phyllosilicates, however, showed

that this procedure of assigning parameters to cis and trans sites is often untenable

(Murad and Wagner, 1994; Rancourt, 1994). A debate between Dyar (1993) and

Rancourt (1993) illustrates the problem of characterising tetrahedral Fe

.Theisomer

shift of tetrahedral Fe

is known to be lower than that of octahedral Fe

by about

0.1 mm/s. Using computer-ﬁtted data, Dyar (1993) assigns doublets with a corre-

spondingly low isomer shift to tetrahedrally coordinated Fe

. On the other hand,

Rancourt (1993) contends that tetrahedral Fe

is not present unless the spectra show

a visible shoulder from the high-velocity line of the tetrahedral Fe

doublet.

The iron contents of illites range fromo1 to over 10 wt.% (S

rodon

and Eberl,

1984), and their Mo

ssbauer spectra were reported in a number of papers. Unfor-

tunately, data comparison is complicated by the large variation in chemical com-

position as well as a multiplicity of (often not comparable) models used to ﬁt the

spectra. In an attempt to rationalise the data, Murad and Wagner (1994) have taken

the Mo

ssbauer spectra of 8 illites with iron contents between 0.8 and 8.4 wt.% Fe.

The main conclusions of this study are that tetrahedral Fe

is present only in the

most iron-rich sample, and that many illites, like kaolinite, show distinct effects of

paramagnetic relaxation at 4.2 K.

12.1.3. Mo

ssbauer Spectra of Selected Clay Minerals 757

Nontronite is the iron-rich end-member of the dioctahedral beidellite–nontronite

smectite group. Nontronites contain much more iron than kaolinite and illite; in

some instances, the iron content may exceed 24 wt.% (Ko

ster et al., 1999). Room

temperature Mo

ssbauer spectra of nontroni te show a rather small average quad-

rupole splitting ( 0.4 mm/s), consistent with a moderate distortion of the octahedral

sites. The broad Fe

resonance has to be ﬁtted with at least two Lorentzian dou-

blets of similar isomer shift but different quadrupole splittings. There was some

controversy regarding the origin of these doublets. In early papers these are inter-

preted in terms of cis and trans oc tahedral site occupancies, although electron mi-

croscopy (e.g., Me

ring and Oberlin, 1967 ) indicates that iron occupies only the cis

sites. The most probable cause for the line broadening is the existence of many

different environments at the atomic scale, causing the Mo

ssbauer parameters to

Fig. 12.1.1. Mo

ssbauer spectra of a kaolinite (total Fe content 1.29 wt.%) from the Jari

deposit in Brazil, taken at room temperature (RT) and 4.2 K. The effects of slow paramagnetic

relaxation at these temperatures are evidently different. Nevertheless, the non-relaxing com-

ponents resulting from Fe

in the octahedral sites, indicated by the broken subspectra, have

identical parameters (d=Fe ¼ 0:35, D ¼ 0:51 mm=s). Adapted from Murad and Wagner (1991).

Chapter 12.1: Mo

ssbauer Spectroscopy of Clays and Clay Minerals758

‘smear out’. Such cases are best accounted for by ﬁtting the spectra with distributions

of quadrupole splittings (Murad, 1987).

The ratios of tetrahedral Fe

to total iron, derived from Mo

ssbauer spectra of

selected nontronites (e.g., Garﬁeld H33a), also showed a ‘disconcertingly’ large

variation (Stucki, 1988). These variations may be due to differences between

subsamples, incompatible ﬁtting techniques (such as the use of different line proﬁles

or constraint of parameters), and the presence of other minerals, in particular iron

oxides. In deed, the association of ﬁnely divided iron oxides with nontronites seems

to be the rule rather than the exception (Murad, 1987). However, when spectra are

taken at room temperature, the contributions of iron oxides (see next section) cannot

be separated from the lines due to nontronite. Such a separation would require

taking Mo

ssbauer spectra at significantly low er temperatures (Fig. 12.1.2).

Variations in the Fe

/Fe

ratios of smectites as a result of chemical treatments

can be readily monitored by Mo

ssbauer spectroscopy. Mo

ssbauer spectra can also

provide evidence for non-reversible changes in mineralogy when some nont ronites

are reoxidised following reduction with sodium dithionite (e.g., Rozenson and Hell-

er-Kallai, 1976; Russell et al., 1979), showing that problems can occur if such treat-

ments are not carried out with sufﬁcient care.

12.1.4. MO

SSBAUER SPECTRA OF IRON (HYDR)OXIDES

Iron oxides and/or oxyhydroxides, collectively referred to as ‘iron (hydr)oxides’

(see Chapters 1 and 5), are common constituents of soils and clays. As already

remarked on, these ‘associated minerals’ can have signi ficant effects on the

ssbauer spectra of phyllosilicates.

Iron (hydr)oxides have relatively high magnetic ordering temperatures, ranging from

955 K for hematite and 400 K for goethite (both of which should therefore be

magnetically ordered at room temperature) to 77 K for lepidocrocite. In practice,

however, the Mo

ssbauer spectra of iron (hydr)oxides, associated with soils and clays,

frequently deviate from ideal behaviour. One reason for this is that other elements,

notably aluminium, can substitute for iron in the iron (hydr)oxide structure. As a

result, the magnetic ordering temperatures decrease and the magnetic hyperﬁne ﬁelds

are reduced at all temperatures. Moreover, small-particle effects may cause ﬂuctuations

of the magnetic ﬁeld (‘superparamagnetic relaxation’), leading to line broadening and

hyperﬁne ﬁeld reduction. In extreme cases, magnetic order can be completely sup-

pressed. In order to differentiate iron (hydr)oxides from phyllosilicates, Mo

ssbauer

spectra must be taken at temperatures that are adequately low to minimise these effects

and ensure the establishment of magnetic order in the iron (hydr)oxides, yet high

enough to avoid magnetic order in the phyllosilicates (generally in the range 120–20 K).

ssbauer spectroscopy is extremely sensitive to the presence of magnetically

ordered phases in samples of complex composition. Murad and Wagner (1991,

1994), for example, showed that as little as 1% ferrihydrite associated with illite

12.1.4. Mo

ssbauer Spectra of Iron (Hydr)Oxides 759

(Fig. 12.1.3), and 0.1% goethite in association with kaolinite, can be clear ly detected

in Mo

ssbauer spectra taken at 4.2 K.

12.1.5. MO

SSBAUER SPECTRA OF FIRED CLAY MINERALS

ssbauer spectroscopy is ideally suited for monitoring the thermal reactions of

clays and clay minerals because it can reveal variations in Fe

/Fe

ratios as well

Fig. 12.1.2. Low-temperature Mo

ssbauer spectra of three ‘pure’ nontronites: H33a and

SWa–1 (taken at 77 K) and H’ANG (taken at 120 K). The central doublet is assigned to Fe

in the nontronite structure, whereas sextets are due to ancillary goethite. Analyses of the

spectra indicate that goethite makes up 2.9% of the iron in H33a, 4.6% in SWa–1, and 13% in

H’ANG. Adapted from Murad (1987).

Chapter 12.1: Mo

ssbauer Spectroscopy of Clays and Clay Minerals760

as changes in coordination polyhedron symmetry. The samples used in many

published investigations, however, are not well characterised, and their mineralogical

purity is open to question. Furthermore, the resul ts are often difﬁcult to compare

because of variations in experimental conditions (specifically with respect to the

duration of heating) and spectral ﬁtting techniques. In an attempt to resolve this

problem, Murad and Wagner (1991, 1996) recorded the Mo

ssbauer spectra of a

kaolinite from the Jari deposit in Brazil (see Fig. 12.1.1) and a reference illite

(OECD #5) after heating the kaolinite in steps of 50 1C up to 1350 1C and the illite up

to 1300 1C.

When the illite is heated under free access of air, the ﬁrst change that occurs is a

decrease in the proportion of Fe

, from an initial value of  7% to essentially zero

Fig. 12.1.3. Mo

ssbauer spectra of an iron-poor illitic clay (total Fe content 0.80 wt.%) from

Upper Silesia, Poland, taken at room temperature (RT) and 4.2 K. Doublets arising from

and Fe

are indicated with dotted and broken lines, respectively. Note the different

velocity ranges of the spectra. The sextet in the latter spectrum shows that the iron mineralogy

of the sample is dominated by ferrihydrite (although this amounts to just 1% of the sample),

whereas no more that 25% of the iron content are bound in the phyllosilicates. Adapted from

Murad and Wagner (1994).

12.1.5. Mo

ssbauer Spectra of Fired Clay Minerals 761

at 300 1C(Fig. 12.1.4). However, the most striking result for both kaolinite and illite

is the development of regions of high quadrupole splitting between  450 and

900 1C. In the case of kaolinite, the quadrupole splitting increases from 0.52 mm/s

(for the unheated material) to a maximum of 1.63 mm/s, while the illite spectra show

a less pronounced increase from 0.63 to 1.53 mm/s. This observation is indica tive of

an extensive Fe

site distortion in metakaolin. The average quadrupole splittings

for the paramagnetic components decreases somewhat after ﬁring to 900 1C and

above, but the complex appearance of the spectra indicates the presence of iron in

more than one (high-temperature) phase.

The room temperature spectra of the ﬁred kaolinite show only minor indications

of a magnetically ord ered phase, whereas the corresp onding illite spectra exhibit

distinct hematite sextets after ﬁring to temperatures between 950 and 1250 1C(Fig.

12.1.4). At 4.2 K, the spectra of the ﬁred illite show poorly deﬁned, broad magnetic

components from 650 1C upwards, whereas a well-deﬁned magnetic component that

can be attributed to hematite ﬁrst appears at 900 1C. It would therefore appear that

the iron oxides initially formed are of small particle size, and their magnetisation is

subject to superparamagnetic relaxation. As the ﬁring temperature is raised, both the

proportion and particle size of the hematite formed increase. The disappearance of

hematite at X1250 1C coincides with the formation of glass, capable of incorporating

more iron than the crystalline high-temperature phases.

ssbauer spectroscopy was often used to characterise archaeological ceramics.

Provided that the nature and origin of the clay material used for their manufacture

are known, it is possible to reconstruct details of the processing technique, such as

Fig. 12.1.4. Room temperature Mo

ssbauer spectra of the OECD #5 illite (total Fe content

5.18 wt.%) ﬁred in 50 1C steps for 48 h. Note the disappearance of Fe

at 300 1C, the

appearance of a magnetically ordered component (hematite) at 950 1C, and its subsequent

disappearance 1250 1C. From Murad and Wagner (1996).

Chapter 12.1: Mo

ssbauer Spectroscopy of Clays and Clay Minerals762

the tempe rature and duration of ﬁring, and whether oxidising or reducing conditions

were used (Wagner et al., 1998). Given that such potsherds may be the sole remnants

of some former civilisations, this information can be of inestimable value to

archaeological research.

REFERENCES

De Grave, E., Van Alboom, A., 1991. Evaluation of ferrous and ferric Mo

ssbauer fractions.

Physics and Chemistry of Minerals 18, 337–342.

Dyar, M.D., 1993. Mo

ssbauer spectroscopy of tetrahedral Fe

in trioctahedral micas—

Discussion. American Mineralogist 78, 665–668.

Fysh, S.A., Cashion, J.D., Clark, P.E., 1983. Mo

ssbauer effect studies of iron in kaolin.

I. Structural iron. Clays and Clay Minerals 31, 285–292.

Gibb, T.C., 1976. Principles of Mo

ssbauer Spectroscopy. Chapman & Hall, London.

Gonser, U. (Ed.), 1975. Mo

ssbauer Spectroscopy. Springer-Verlag, Berlin.

ster, H.M., Ehrlicher, U., Gilg, H.A., Jordan, R., Murad, E., Onnich, K., 1999. Miner-

alogical and chemical characteristics of ﬁve nontronites and Fe-rich smectites. Clay Min-

erals 34, 579–599.

Malden, P.J., Meads, R.E., 1967. Substitution by iron in kaolinite. Nature 215, 844–846.

ring, J., Oberlin, A., 1967. Electron-optical study of smectites. Clays and Clay Minerals 15,

3–25.

ssbauer, R.L., 1958a. Kernresonanzabsorption von Gammastrahlung in Ir

191

Naturwissenschaften 45, 538–539.

ssbauer, R.L., 1958b. Kernresonanzﬂuoreszenz von Gammastrahlung in Ir

191

. Zeitschrift

r Physik 151, 124–143.

Murad, E., 1987. Mo

ssbauer spectra of nontronites: structural implications and character-

ization of associated iron oxides. Zeitschrift fu

r Pﬂanzenerna

hrung und Bodenkunde 150,

279–285.

Murad, E., Cashion, J., 2004. Mo

ssbauer Spectra of Environmental Materials and their In-

dustrial Utilization. Kluwer, Boston.

Murad, E., Wagner, U., 1991. Mo

ssbauer spectra of kaolinite, halloysite and the ﬁring prod-

ucts of kaolinite: new results and a reappraisal of published work. Neues Jahrbuch fu

Mineralogie, Abhandlungen 162, 281–309.

Murad, E., Wagner, U., 1994. The Mo

ssbauer spectrum of illite. Clay Minerals 29, 1–10.

Murad, E., Wagner, U., 1996. The thermal behaviour of an Fe-rich illite. Clay Minerals 31,

45–52.

Rancourt, D.G., 1993. Mo

ssbauer spectroscopy of tetrahedral Fe

in trioctahedral micas—

Reply. American Mineralogist 78, 669–671.

Rancourt, D.G., 1994. Mo

ssbauer spectroscopy of minerals. II. Problem of resolving cis and

trans octahedral Fe

sites. Physics and Chemistry of Minerals 21, 250–257.

Rozenson, I., Heller-Kallai, L., 1976. Reduction and oxidation of Fe

in dioctahedral

smectites – 1: reduction with hydrazine and dithionite. Clays and Clay Minerals 24,

271–282.

Russell, J.D., Goodman, B.A., Fraser, A.R., 1979. Infrared and Mo

ssbauer studies of reduced

nontronites. Clays and Clay Minerals 27, 63–71.

References 763

rodon

, J., Eberl, D.D., 1984. Illite. In: Bailey, S.W. (Ed.), Micas. Reviews in Mineralogy,

vol. 13. Mineralogical Society of America, Washington, DC, pp. 495–544.

Stevens, J.G., Khasanov, A.M., Miller, J.W., Pollak, H., Li, Z., 1998. Mo

ssbauer Mineral

Handbook. Mo

ssbauer Effect Data Center, Asheville, N.C., p. 527.

Stucki, J.W., 1988. Structural iron in smectites. In: Stucki, J.W., Goodman, B.A., Schwert-

mann, U. (Eds.), Iron in Soils and Clay Minerals. D. Reidel, Dordrecht, pp. 625–675.

Wagner, U., Gebhard, R., Grosse, G., Hutzelmann, T., Murad, E., Riederer, J., Shimada, I.,

Wagner, F.E., 1998. Clay: an important material for prehistoric man. Hyperﬁne Interac-

tions 117, 323–335.

Weaver, C.E., Wampler, J.M., Pecuil, T.E., 1967. Mo

ssbauer analysis of iron in clay minerals.

Science 156, 504–508.

Chapter 12.1: Mo

ssbauer Spectroscopy of Clays and Clay Minerals764