Structure and microstructure of Mg-vermiculite

Z. Kristallogr. Suppl. 30 (2009) 429-434 / DOI 10.1524/zksu.2009.0063 © by Oldenbourg Wissenschaftsverlag, München

429

Structure and microstructure of Mg-vermiculite A. Argüelles1, M. Leoni2,*, J. A. Blanco1, C. Marcos3 1

Dpto. Física, Universidad de Oviedo, C/ Calvo Sotelo, s/n, 33007, Oviedo, Spain Department of Materials Engineering and Industrial Technologies, University of Trento, via Mesiano, 77, 38100, Trento, Italy 3 Dpto. Geología e Instituto de Organometálica Enrique Moles, Universidad de Oviedo, C/ Jesús Arias de Velasco, s/n, 33005, Oviedo, Spain * [email protected] 2

Keywords: Mg-vermiculite, clay structure, stacking faults, DIFFaX+ Abstract. A specimen of Mg-vermiculite from Santa Olalla (Spain) has been investigated. Powder diffraction data has been simulated and an experimental pattern refined by means of the DIFFaX+ software. The best modelling is obtained assuming the mineral to possess bidimensional periodicity instead of a tridimensional one: the structure can be described as a disordered sequence of two different types of layers, differing in the arrangement of the interlayer cations.

Introduction Mg-vermiculite has been the subject of mineralogical, physical and chemical studies for several decades. This hydrous phyllosilicate is nowadays finding new interesting applications in environmental science [1] and nanoscience [2]. From a structural point of view, vermiculite is a 2:1 clay: a T-O-T unit (formed by the assembly of an octahedral and two tetrahedral layers) is present, with an interlayer filled with Mg2+ hydrated cations (see figure 1). The crystalline structure of vermiculite has been theoretically and experimentally investigated by several groups (see e.g. [3-5]), but the most widely cited model is still that proposed by Shirozu and Bailey [6], which does not deal with disorder, because of its simplicity. The presence of defects such as stacking faults makes the diffraction analysis of this material a hard task. Traditionally, the refinement of a structure from powder diffraction data is accomplished by means of the Rietveld method. Existing refinement software, however, assumes the structure to be periodic and the presence of defects is at most considered through a variational approach. Extensive amount of faults, leading to strong diffuse effects on the pattern, cannot be handled. An alternative approach has been proposed, where the recursive description of the stacking proposed by Treacy et al. [7] is employed.

430

European Powder Diffraction Conference, EPDIC 11

Hydrated Mg cations Tetrahedral layer Octahedral layer Tetrahedral layer

Figure 1. Schematic drawing of Mg-vermiculite structure.

In the present paper, a preliminary structural refinement by using the DIFFaX+ software [8], implementing a refinement routine based on the formula of Treacy et al., has been undertaken. The literature model proposed by de la Calle et al. [9] has been taken as the starting point. Experimental data have been obtained by X-ray diffraction (XRD) in transmission mode. As a result, DIFFaX+ software has allowed deal with such faulted structure in an easy and fast way, refining the structure of Mg-vermiculite from Santa Olalla with that model proposed by de la Calle et al.

Experimental results and discussion A powder sample was prepared by milling some flakes of natural Mg-vermiculite from Santa Olalla (Huelva, Spain). A morphological study was conducted both by Scanning Electron Microscopy (SEM) and by Laser Diffraction using, respectively, a FEI Quanta 400 microscope and a FRITSCH MicroTec XT laser particle sizer. X-ray powder diffraction data were collected on an INEL XRD RG3000 diffractometer operating in Debye-Scherrer capillary transmission mode. The system is equipped with a cobalt tube and a primary beam Johannson-type monochromator. The crystalline and defect structure refinement was performed by means of the DIFFaX+ software v. 2.400. Morphological study Figure 2 shows the morphology of the particles in the milled sample of natural Mgvermiculite sample: the maximum size is around 80 μm. Although particles show a heterogeneous morphology, a platelet-like habit is observed, as expected for a lamellar phyllosilicate.

Z. Kristallogr. Suppl. 30 (2009)

431

Figure 2. SEM micrograph (500x) of Mg-vermiculite powder sample from Santa Olalla (Spain).

The particle size distribution was determined by laser diffraction. An effective particle diameter was calculated, as the diameter of the sphere that gives the same diffraction signal as the actual particle [10]. The granulometric curve obtained by laser scattering is shown in figure 3. A multimodal distribution is apparent, the main fractions having an average of about 10 and 55 μm, respectively.

Differential volume (%)

6 5 4 3 2 1 0 0,1

1

10 100 Particle size (μm)

1000

Figure 3. Particle size distribution of Mg-vermiculite obtained by laser diffraction.

Structural refinement The most accredited reference for vermiculite structure is probably the work of Shirozu and Bailey [6] where the mineral is described as being monoclinic and ordered. It seems, however, that this model can just partly justify the observed features in the X-ray powder diffrac-

432


tion pattern of the mineral. An alternative, not widely accepted by the scientific community, was proposed by de la Calle et al. [9], who described vermiculite as semi-ordered. According to those authors, the structure is made of a stacking of two types of layers, A and B (see below), differing in the position of the interlayer hydrated cation. Moreover, the relative translations of those layers along the crystallographic y-axis are not unique. In the present work, the model of de la Calle et al. was assumed as the starting point for the analysis. Owing to the irregular nature of the stacking, the Rietveld method is not appropriate for the structural refinement. Moreover, due to the layered character of this material, preferred orientation (in particular 00l fibre texture, i.e. enhancement of the signal from basal reflections) is highly probably in a powder specimen, further contributing to render a fullprofile analysis quite difficult [11]. For this reason, techniques specially devised for the analysis of faulted structures need being used. In this case, the DIFFaX+ software, based on a recursive description of the stacking has been employed [8]. This methodology has been already tested on clay systems and proven to be effective for the modelling of complex layered systems [12]. XRD data collected in capillary transmission mode was undertaken, to limit the effect of preferred orientation. To test the relative efficiency of the ordered and semi-ordered models to describe the structure of vermiculite, different stacking configurations were simulated and compared (cf. table 1 and figure 4). Two types of layers (named, respectively, A and B) were considered, differing only in the position of the interlayer hydrated cation relative to the silicate surface layer. One modelling attempt has been made supposing a translation of ±1/2 along the y-axis (translation mode in table 1). Table 1. Information of the different stacking models for Mg-vermiculite: type of stacking mode, spatial translations, probabilities of the sequences and reliability factor.

Description of the stacking mode

Translation from one layer to the next, in the crystallographic y-axis tyA→

Probabilities of the sequence

tyA→B

tyB→A

tyB→B

αAA

αAB

αBA

αBB

Rwp (%)

A

Random

+1/3

+1/3

-1/3

-1/3

1/2

1/2

1/2

1/2

6.6

Segregated

+1/3

+1/3

-1/3

-1/3

0.75

0.25

0.25

0.75

7.2

Translation

+1/2

+1/2

-1/2

-1/2

0.416

0.584

0.584

0.416

12.3

Alternated

0.34

0.34

-0.32

-0.33

0.416

0.584

0.584

0.416

6.5

Z. Kristallogr. Suppl. 30 (2009)

433

3000 a)

b)

Intensity (Arb. units)

2500 2000 1500 1000 500 0 2500 c)

d)

Intensity (Arb. units)

2000 1500 1000 500 0 20

25

30

35 2θ (degrees)

40

45

50 20

25

30

35

40

45

50

2θ (degrees)

Figure 4. Experimental (dots) and calculated (solid line) X-ray diffraction patterns: random mode (4a), segregated mode (4b), translation mode (4c) and alternated mode (4d). Refined probabilities corresponding to the case presented in 4d are listed in table 1. See the text for more details.

The other models consider a translation of ±1/3, as usually admitted in vermiculites. Different sequences of the stacking have been considered by changing the probabilities of layer to layer transition. In particular, the random mode corresponds to a random alternation of A and B type layers, while the segregated one corresponds to a sequence favouring one type of layer being followed by a layer of the same type. In all cases a refinement was tried. The best match between experimental and calculated diffraction patterns was obtained for the alternated model. In that case, the translation and probability parameters describing the sequence were both refined, obtaining a Rwp factor of 6.45% (see table 1). The probability of a layer being followed for other of the same type has been found to be 0.416, a bit lower than in a totally random model. In figure 4, all calculated patterns obtained for the different stacking modes described in table 1, are shown and compared to the experimental pattern. It should be noted that the (02) reflection (aka (02l) diffraction band) between 21º and 25º 2θ is the most affected by the stacking mode parameters. The best refinement is obtained with a value of approximately ±1/3 for translations along the y-axis and almost equal transition probabilities for the various sequences. A defective structure, with a small tendency to alternation of two similar types of layers seems to be the most suitable model for describing the stacking mode of this Mg-vermiculite from Santa Olalla (Spain) as has been obtained from DIFFaX+ refinements. Further investi-

434


gations coupling powder X-ray diffraction and 1D Fourier analysis are in progress and the results will be presented elsewhere [13].

Concluding remarks A powder specimen of Mg-vermiculite from Santa Olalla (Spain) was analysed by X-ray powder diffraction. The structure was modelled by means of the DIFFaX+ software. A semiordered arrangement is confirmed to be the more plausible for this Mg-vermiculite. The parameters describing such stacking mode have been refined leading to a model that confirms that proposed by de la Calle et al., a preliminary difference being the uneven layer transition probabilities. The refined model is not totally random but certain tendency to alternation has been found. Further investigations using synchrotron powder X-ray diffraction are in progress to further validate this picture.

References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

Huang, H.C., Lee, J.F., Chao, H.P., Yeh, P.W., Yang, Y.F. & Liao, W.L, 2005, J. Colloid Interface Sci., 286, 127. Matějka, V., upová-Křístková, M., Kratoová, G. & Valáková, M., 2006, J. Nanosci. Nanotechnol., 6, (8), 2482. Weiss, Z. & Durovic, S., 1980, Acta Crystallogr., A36, 633. de la Calle, C., Dubernat, J., Suquet, H. & Pezerat, H., 1980, Bull. Mineral., 103, 419. de la Calle, C., Suquet, H. & Pezerat, H., 1985, Clay Miner., 20, 221. Shirozu, H. & Bailey, S.W., 1966, Am. Mineral., 51, 1124. Treacy, M.M.J., Newsam, J.M. & Deem, M.W., 1991, Proc. R. Soc. Lond., A 433, 499. Leoni, M., Gualtieri, A.F. & Roveri, N., 2004, J. Appl. Crystallogr., 37, 166. de la Calle, C., Suquet, H. & Pons, C.H., 1988, Clays Clay Miner., 36, (6), 481. Konert, M. & Vandenberghe, J., 1997, Sedimentology, 44, (3), 523. Martínez Blanco, D., Gorría, P., Blanco, J.A., Pérez, M.J. & Campo, J., 2008, J. Phys. Condens. Mater., 20, 335213. Gualtieri, A.F., Ferrari, S., Leoni, M., Grathoff, G., Hugo, R., Shatnawi, M., Paglia, G. & Billinge, S., 2008, J. Appl. Crystallogr., 41, 402. Argüelles, A., Leoni, M., Pons, C.H., de la Calle, C., Blanco, J.A. & Marcos, C., 2009, J. Appl. Crystallogr., in preparation.

Acknowledgements. The authors wish to acknowledge the Institut des Sciences de la Terre dOrléans (France) for diffraction measurements in transmission geometry. The financial support from FEDER and the Spanish MICINN (FROMER MEC) grant number MAT200806542-C04-03 is greatly acknowledged. Special thanks to both C. de la Calle and C. H. Pons for helping us in the X-ray experiments and for stimulating discussions.