Dehydration process and transient channel deformations of slightly hydrated boron leucite: An “in situ” time-resolved synchrotron powder diffraction study

Microporous and Mesoporous Materials 142 (2011) 570–576

Contents lists available at ScienceDirect

Microporous and Mesoporous Materials journal homepage: www.elsevier.com/locate/micromeso

Dehydration process and transient channel deformations of slightly hydrated boron leucite: An ‘‘in situ’’ time-resolved synchrotron powder diffraction study A. Martucci ⇑, P. Pecorari, G. Cruciani Dipartimento di Scienze della Terra, University of Ferrara, Ferrara, Italy

a r t i c l e

i n f o

Article history: Received 19 October 2010 Received in revised form 24 December 2010 Accepted 31 December 2010 Available online 6 January 2011 Keywords: Borosilicate Leucite Dehydration Rietveld refinement In situ X-ray powder diffraction

a b s t r a c t The step by step thermal dehydration process of slightly hydrated boron leucite K16[B16Si32O96], was studied in situ by synchrotron radiation powder diffraction. A time-resolved experiment was performed using a translating imaging plate system. Rietveld refinements were carried out on 41 consecutive powder patterns in the 25–940 °C temperature range. Results show that temperature-induced transformations can be schematized into two main steps. In the 25–565 °C temperature range, the symmetry remained cubic I-43d and the unit cell parameter increased with an increase in temperature. The migration of H2O molecules through the [1 1 1] channels during dehydration determined an opening of the six-member ring apertures which was as wide as possible and, at the same time, a narrowing in the eight-ring along [1 1 1]. This process was accomplished by a twisting in the tetragonal prism, constituting a leucite framework, which led to an opposed tilting in the tetrahedra connecting the prisms. Above 565 °C, a continuous structural transformation led to a displacive polymorphic transition. Rietveld structure refinement of the unit cell parameters showed a remarkable change, thus indicating an I-43d ? Ia-3d change in symmetry. This was associated with a relaxation in the continuous structural distortions of the leucite framework and the T–O–T and O–T–O angles indicated the formation of more regular apertures. Ó 2011 Elsevier Inc. All rights reserved.

1. Introduction The response of zeolites to heating is not only of academic significance but also of potential industrial importance. A number of factors contribute to macroscopically observable thermal effects such as the evolution of H2O and encapsulated organic species, variations in unit-cell volume and structural breakdown or modification [1,2]. These effects, which modify pore and channel geometry, affect the absorption and diffusion of molecules in zeolites and microporous catalysts, and consequently the catalytic properties of materials [3]. When boron ions are incorporated into the tetrahedral framework sites of zeolites, selective heterogeneous catalysts are produced (which are referred to as ‘‘boralites’’ or ‘‘borosilicates’’). In comparison to zeolites, boralites display lower thermal stability and framework deboronation often occurs during the calcination steps which are required to eliminate organic molecules trapped within their pores [3]. The structure of leucite, which shows an ANA-type zeolitic framework, has been the subject of extensive research, due to its ⇑ Corresponding author. Address: Dipartimento di Scienze della Terra, Università di Ferrara, Via G. Saragat, 1, 44100 Ferrara, Italy. Tel.: +39 0 532 974730; fax: +39 0 532 974767. E-mail address: [email protected] (A. Martucci). 1387-1811/$ - see front matter Ó 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.micromeso.2010.12.046

displacive tetragonal to cubic transition, Al and Si ordering on T-sites and its ferroelastic and dielectric properties [4,5]. The structure of analcime, leucite and analogue materials have been extensively studied by several authors [6–11]; comprehensive summaries of structural investigations have been reported by Henderson and Taylor [12], Hazen and Finger [13], and Palmer et al. [11]. Leucite frameworks can be built using six-membered tetrahedra rings connected through distorted four-rings to form chains along the b axis [14] which are common to the ANA group of zeolites [14,15]. Two different types of channels exist: larger channels which are made of highly distorted eight-membered rings along [1 1 0], and smaller channels which are defined by six-membered rings along the [1 1 1] direction of the cubic lattice. Cavities in ANA consist of three distorted prismatic cages which are delimited by 4-, 6- and 8-member tetrahedral rings. The major differences between the analcime and leucite stems are due to the type and distribution of extraframework ions. In analcime Na, cations occupy the centre (S site) of each prismatic cage and are coordinated by two water molecules (W site) and four framework oxygens. The water molecules in analcime occupy the same structural positions in the [1 1 1] direction as the K ions in leucite, whereas the analcime Na positions fill sites that are vacant in leucite. Analcime is a zeolite-type hydrous mineral; its

A. Martucci et al. / Microporous and Mesoporous Materials 142 (2011) 570–576

stoichiometric water content is >8.0 wt.%, whereas, leucite is essentially anhydrous (a so-called ‘nominally anhydrous mineral’, NAM), in agreement with its origin as an early crystallizing volcanic mineral. However, minor but significant amounts of water have been detected in leucite from Vesuvius and Roccamon-fina volcanic centres [16]. Recently, Della Ventura et al. [17] reported on a single-crystal FTIR spectroscopic study of many leucite samples deriving from different occurrences and localities in the volcanic Alban Hills area (Latium, Italy). Their FTIR spectra clearly indicated that almost all the examined samples contained hydrous components in the form of structurally-bound water molecules. Leucite structures are extremely tolerant towards ionic substitutions [12,18,19], and it is its considerable structural flexibility with respect to changing composition, temperature, and pressure that makes leucite of particular interest from a crystal–chemical viewpoint. The hydrothermal synthesis of a fully B-substituted ANA-type zeolite structure was achieved by Millini et al. [20] who demonstrated the possibility of completely replacing Al with B in the analcime framework by hydrothermal synthesis. The same authors performed a Rietveld refinement on their sample in the Ia-3d topological symmetry, concluding that boron-substituted analcime was actually isostructural with leucite rather than with analcime. This result was at variance with the previous structural analysis reported by Ihara and Kamei [21] who solved and refined the structure of K-boroleucite crystals in the I-43d space group. The latter space group was also used by Mikloš et al. [22] for the refinement of their K(BSi2O6) boroleucite. Concerning the effects of extraframework species on the stability of B-ANA-type structures, Millini et al. [20] also showed that in the presence of K+ ions in the synthesis batch, the stoichiometry of the B-ANA-type structure is consistently close to the theoretical K16[B16Si32O96] structure. Upon replacement of K+ by other alkali metal ions, the formation of amorphous products was observed, while the addition of tetralkylammonium hydroxide did not influence the behaviour of the system [3]. The hydrothermal synthesis of boron-containing zeolites with a leucite structure and large

571

alkali cations, like potassium, rubidium or cesium, has been reported in the presence of organic cations by Di Renzo and co-workers [23]. These authors stated that the easy formation of borosilicates in the presence of the largest alkali cations was related to the organic-templated synthesis of borosilicate zeolites and confirmed the relevance of a hydrophobic effect in the insertion of boron in tetrahedral lattice sites. As far as the thermal stability of these systems is concerned, Millini et al. [20] pointed out the lack of any signal as a result of trigonal boron species in the B MAS NMR spectra of the calcined material. This revealed the high thermal stability of their B-substituted K-leucite, which is unusual for borosilicates with zeolite structures. Mazza et al. [24] examined different samples prepared using a sol– gel preparation along the K[Al1-xBxSi2O6] compositional junction with 0 6 x 6 1 using XRD, DTA, FTIR, and dilatometry techniques. This work showed that by increasing the x value from 0 to 1 the samples crystallize, in a temperature range from 900 to 1250 °C, with tetragonal (0 6 x 6 0.25), cubic (0.50 6 Ix 6 0.80) and again tetragonal (0.90 6 Ix 6 1.00) symmetry. The above-mentioned authors concluded that all the phases derived from a cubic hightemperature form by means of a displacive phase transition. The aim of the present work is to study the thermal behaviour of boron-substituted leucite, using Rietveld structure analysis of temperature-resolved powder diffraction data collected using synchrotron radiation. Such experimental conditions are ideal for the rapid collection of the diffraction data necessary to monitor each step of the dehydration process in detail. 2. Experimental 2.1. Materials A powder sample of (B-LEU), K16[B16Si32O96] B-leucite which was the same as was used for the X-ray powder diffraction study of Ref. [20], was selected for this experiment. As reported by Millini et al. [20], the B and K content were determined by ICP-AES using a Jobin Yvon 38II Plus spectrophotometer. TG and differential thermal analysis (DTA) measurements of the as-synthesized

Fig. 1. Evolution of the XRPD patterns in the selected 5–20° 2h interval as a function of the temperature during the in situ time-resolved experiment.

572

A. Martucci et al. / Microporous and Mesoporous Materials 142 (2011) 570–576

sample were performed in air up to 900 °C using an STA 409 PC LUXXÒ – Netzch at 5 °C/min heating rate. 2.2. Methods A powder pattern of the as-synthesized B-LEU sample was measured on a Bruker D8 Advance diffractometer equipped with an

Si(Li) solid state (Sol-X) detector, using Cu Ka1,a2 radiation in the 3–110° 2h range and a counting time of 12 s/step. Rietveld structure refinement was therefore performed, using the GSAS package [25] with an EXPGUI interface [26]. Time-resolved diffraction data were collected on the GILDA beamline at ESRF (Grenoble). The powder sample was loaded and packed in a 0.3 mm diameter Lindemann capillary, open at both ends, and heated in situ up to

Fig. 2. Observed (dotted upper line), calculated (solid upper line), and difference (solid lower line) powder diffraction patterns of B-LEU at room temperature 25° (a), 350° (b), 540° (c), 700° (d), 830° (e) and 910° (f) from in situ time-resolved data.

Table 1 Crystallographic data from the Rietveld refinement of boron leucite K16[B16Si32O96], at 25°, 350°, 540°, 700°, 830° and 910° from in situ time-resolved data.

Space group a0 (Å) V (Å)3 2h range (°) Rwp (%) Rp (%) R2F (%) No. of diffraction lines No. of observations No. of variables

25 °C

350 °C

540 °C

700 °C

830 °C

910 °C

I-43d 12.641(1) 2019.9(1) 3–45 6.55 5.0 6.00

I-43d 12.675(1) 2036.5(3) 3–45 5.73 4.39 6.04

I-43d 12.725(2) 2060.6(1) 3–45 5.15 4.01 6.06

Ia-3d 12.778(4) 2086.5(3) 3–45 9.31 7.2 7.75

Ia-3d 12.811(4) 2102.4(3) 3–45 10.40 7.77 8.54

Ia-3d 12.824(4) 2108.9(4) 3–45 12.19 9.15 14.65

167 2909 40

147 2609 40

149 2609 40

136 2772 34

136 2772 34

136 2772 34

k = 0.68888(1) Å. Rp ¼ R½Y io  Y ic R=Y io ; Rwp ¼ ½Rwi ðY io  Y ic Þ2 =Rwi Y 2io 0:5 ;

R2F ¼ RjF 2o  F 2c j=RjF 2o j.

573

A. Martucci et al. / Microporous and Mesoporous Materials 142 (2011) 570–576

940 °C using a hot air stream. The heating rate was 5 °C/min. The capillary sample was mounted on a standard goniometer head and kept spinning during data collection in parallel beam Debye– Scherrer geometry, using an Si(1 1 1) monochromatized wavelength of 0.95337(1) Å. During the heating process, powder diffraction patterns were recorded on the 4 mm slit-delimited portion of a translating flat image plate [27], which had a translation rate of 2.5 pixel/°C in comparison to the temperature increase. An external standard LaB6 was used to calibrate the wavelength, as well as to determine the zero-shift position, to sample the detector distance, and tilting angle of the image plate detector. A total of 41 one-dimensional powder patterns, 2h ranging from 3 to 58.5°, were extracted from the image plate by the integration of constant temperature slices (temperature variable ±5 °C and temperature step separation 20 °C), using a locally adapted routine (Fig. 1). The structure refinements were performed by full-profile Rietveld analysis in the space group I-43d by the GSAS package [26] using the site positions of framework atoms from Mikloš et al. [22]. Evidence was found to support a change in symmetry in the powder pattern at 565 °C, thus indicating an I-43d ? Ia-3d transition. The Bragg peak profile was modelled using a pseudo-Voigt function with three refinable coefficients (Gw, Lx, Ly) and a 0.01% cut-off in peak intensity. The instrumental background was empirically fitted using a Chebyschev polynomial of the first kind with 16 variable coefficients. The 2h-zero shift was accurately refined in all the data set patterns. The scale factor and unit cell parameters were allowed

to vary for all the histograms. The refined structural parameters for each data histogram were the following: fractional coordinates for all atoms, isotropic displacement and occupancy factors for the K extraframework site and two isotropic displacement factors, one for all the tetrahedral and the other for all the framework oxygen sites. Occupancy factors and isotropic displacement factor coefficients for the extraframework position were initially varied in alternative cycles to avoid correlations. Soft constraints were imposed on T–O distances during the initial cycles, and then released in the final refinement cycles. Final observed and calculated powder patterns at 25, 350, 540, 700, 830 and 910° C are shown in Fig. 2a–f, respectively. Refinement parameters are reported in Table 1, the refined coordinates in Table 2 and bond distances and angles in Table 3.

3. Results and discussion 3.1. Structure refinement at room temperature The Rietveld structure refinement of the as-synthesized B-ANA sample was performed using the room temperature laboratory powder diffraction data. The initially-tested crystal structure model was the one in the Ia-3d space group as reported by Millini et al. [20] for the same sample as used in our study. The same authors suggested that the real symmetry of their B-analcime (it turned out to be B-leucite) could be different from the topological one.

Table 2 Atomic coordinates and thermal parameters of framework atoms for boron leucite K16[B16Si32O96], at 25°, 350°, 540°, 700°, 830° and 910° from in situ time-resolved data.

T1

O1

O2

K

25 °C

350 °C

540 °C

700 °C

830 °C

910 °C

x/a y/b z/c Uiso

0.1249(5) 0.1670(5) 0.4106(4) 0.042(4)

0.1214(4) 0.1617(5) 0.4086(3) 0.045(3)

0.1245(4) 0.1643(4) 0.4096(3) 0.052(3)

0.6602(1) 0.4102(1) 1/8 0.062(3)

0.6590(1) 0.4090(1) 1/8 0.064(3)

0.6566(1) 0.4066(1) 1/8 0.066(4)

x/a y/b z/c Uiso

0.2259(3) 0.1240(7) 0.3577(8) 0.062(2)

0.2245(2) 0.1195(6) 0.3604(7) 0.066(2)

0.2248(2) 0.1186(6) 0.3598(6) 0.075(3)

0.1050(2) 0.1328(2) 0.2217(1) 0.077(3)

0.1055(2) 0.1332(2) 0.2229(1) 0.084(3)

0.1040(3) 0.1344(3) 0.2246(1) 0.093(4)

x/a y/b z/c Uiso

0.1232(6) 0.2891(5) 0.4152(7) 0.062(2)

0.1317(6) 0.2825(4) 0.4121(6) 0.066(2)

0.1289(6) 0.2853(4) 0.4097(6) 0.075(3)

x/a y/b z/c Uiso

0.1282(4) 0.1282(4) 0.1282(4) 0.056(2)

0.1282(4) 0.1282(4) 0.1282(4) 0.065(2)

0.1278(4) 0.1278(4) 0.1278(4) 0.087(3)

1/8 1/8 1/8 0.132(4)

1/8 1/8 1/8 0.153(4)

1/8 1/8 1/8 0.157(4)

Note: Estimated standard deviations in parentheses refer to the last digit.

Table 3 Selected bond distances (Å) and angles (°) for boron leucite K16[B16Si32O96], at 25°, 350°, 540°, 700°, 830° and 910° from in situ time-resolved data.

T–O1 T–O1 T–O2 T–O2 T–O1[x2] T–O1[x2] T–O1-T T–O2-T K1–O1 [X3] K1–O1 [X3] K1–O2 [X3] K1–O2 [X3] K1–O1 [X6] K1–O1 [X6]

25 °C

350 °C

540 °C

700 °C

830 °C

910 °C

1.541(2) 1.542(1) 1.544(1) 1.534(1) – – 146.5(1) 133.4(1) 3.152(10) 3.424(9) 2.844(7) 3.447(8) – –

1.541(2) 1.541(8) 1.540(2) 1.542(6) – – 150.6(8) 132.9(8) 3.191(11) 3.375(9) 2.927(7) 3.382(10) – –

1.539(2) 1.540(6) 1.541(2) 1.539(5) – – 147.6(8) 136.4(8) 3.202(10) 3.374(8) 2.951(7) 3.404(9) – –

– – – – 1.526(2) 1.525(1) 146.8(2) – – – – – 3.189(3) 3.342(3)

– – – – 1.529(2) 1.528(1) 147.5(2) – – – – – 3.211(3) 3.353(3)

– – – – 1.535(2) 1.537(1) 146.9(2) – – – – – 3.206(3) 3.352(4)

Note: Estimated standard deviations in parentheses refer to the last digit.

574

A. Martucci et al. / Microporous and Mesoporous Materials 142 (2011) 570–576

Since the structure of synthetic boron-substituted leucite was determined by a single-crystal X-ray diffraction analysis in the I43d [21,22] space group, we proceeded with our refinement in the I-43d space group, using atomic coordinates for the framework as reported by Mikloš et al. [22]. It is noteworthy that the same space group was also reported for the Rb[BSi2O6] crystal structure and its silica-rich solid solutions [27,28] and for (Rb,Cs)[BSi2O6] [29]. The lower R-factors obtained in our full-profile Rietveld analysis using I-43d compared to Ia-3d symmetry confirmed that the model based on subgroup symmetry was the best description for our B-leucite at room temperature. Our refined model was in perfect agreement with that given by Mikloš et al. [22] confirming evidence for the tetrahedral coordination of B atoms. Overall, the B-leucite framework is similar to that of cubic (Ia3d) high-leucite [11] consisting of tetrahedra linked into four and six-membered rings. B and Si atoms in B-leucite are statistically placed on one crystallographic position the same way as Al and Si are in high-leucite. In our B-leucite refinement, the occupation factors for the K, Si and B atoms yielded values close to the ideal chemical formula. The average bond length (1.542 Å) was

an intermediate between typical Si–O and B–O bond distances (1.61 and 1.43 Å, respectively). The mean T–O–T angle (139.45°) was wider than in the K-boroleucite as reported by Mikloš et al. [22] (135.5°), and in K-leucite (137.8°, Peacor, [30]). This can be explained by the variable degree of framework distortion related to the different interaction between the framework oxygens and the extraframework species. In leucite, the K atoms occupy the centre of the channel along [1 1 1] delimited by sixmembered rings (W site) (Fig. 3). Potassium ions at this site can be replaced by other large alkali metal ions (e.g. Rb or Cs in pollucite) and are normally occupied by water in natural analcime. Our refinement of K-boroleucite at RT (see Table 3) shows that K ions are bonded to three framework oxygens at a short distance (2.846 Å) and to another three oxygens at a significantly longer distance (3.158 Å). For the sake of comparison, the HT form of leucite (cubic, Ia-3d) has six equal K–O distances (3.350 Å) while lowleucite (tetragonal, I41/a) shows six unequal K–O distances in the range of 2.895–3.091 Å [11]. In the case of our K-boroleucite sample, thermogravimetric analysis (Fig. 7) revealed the occurrence of very low water content (2.45% weight loss). Part of the water is lost

Fig. 3. Projection along [1 1 1] of the six-membered ring, as refined at room temperature. The 6-fold coordinated K ion is shown at the centre of the cavity.

Fig. 5. Variation in tetrahedral bond distances (Å), showing the phase transitions at 565 °C. T1–O1 distances (filled and empty squares, respectively) and T1–O2 distances (filled and empty triangles, respectively). Error bars are smaller than symbols.

Fig. 4. Evolution of unit cell parameter a (void circles) and volume V (full circles) with temperatures in B-LEU. Initial values a = 12.641(1) Å, V = 2019.9(1) Å3. Error bars are smaller than symbols.

Fig. 6. Variation in T–O–T angles, showing the phase transitions at 565 °C. T–O1–T angles (filled circles) and T–O2–T angles (empty circles). Error bars are smaller than symbols.

A. Martucci et al. / Microporous and Mesoporous Materials 142 (2011) 570–576

at very low temperatures (<100 °C) and should be considered as being adsorbed onto the crystal surface and/or in the intercrystalline porosity. The water content in the pores is much lower, less than 1.5 wt.%, and for this reason all attempts to locate these water molecules in our Rietveld refinement failed. Nevertheless, we suggest that in our K-boroleucite sample, the K ions might also be coordinated by water molecules. The RT structural model was used as a starting model to carry out the analysis for the ‘‘in situ’’ time-resolved synchrotron powder diffraction data. 3.2. Refinements by in situ X-ray data: variations in unit cell parameters and crystal-structure modifications upon heating The evolution of the powder diffraction patterns as a function of temperature is shown in Fig. 1. Final observed and calculated pow-

Fig. 7. Thermogravimetric (TG), differential thermogravimetric (DTG) and differential thermal (DTA) curves in B-LEU dry air atmosphere.

575

der patterns at 25, 350, 540, 700, 830 and 910° C are reported for the boron leucite studied here in Fig. 2a–e. For the same selected temperature steps, lattice parameters and refinement details are reported in Table 1. Atomic co-ordinates, occupancy and temperature factors are reported in Table 2. Bond distances and angles (°) are in Table 3. Fig. 4 shows the variations in unit cell parameters and cell volume in the 25–940 °C range as fractions of the values at room temperature. The occurrence of discontinuity at about 565 °C suggests that the temperature-induced transformations can be schematized into two main steps. In the 25–565 °C temperature range, the symmetry is I-43d cubic and the unit cell parameter (and cell volume) increases with increasing temperature. The thermal expansion coefficients in this temperature range are 1.497  105 and 4.947  103 for a0 and V0, respectively. Continuous structural distortions of the leucite framework associated with the heating process are well recorded by the changes in T–O distances (Fig. 5) and T–O–T angles (Fig. 6). The discontinuities observed in this temperature range clearly indicate that the variations observed are associated with a temperature-dependent distortion mechanism in the framework. It is reasonable to suppose that these structural modifications are associated with the migration of H2O molecules through the [1 1 1] channels during dehydration. This process, which explains the weight loss in the TG curve (Fig. 7), determines an opening of the six-member ring apertures which is as wide as possible and, at the same time, a narrowing in the eight-ring along [1 1 1] (Fig. 8). As already observed in leucite [11] and in natural tetragonal analcime [31], this was accompanied by a twisting in the tetragonal prism constituting the leucite framework, which led to the opposite tilting of the tetrahedra connecting the prisms. The position of the K cations was only slightly affected by the loss of their coordination water. Rietveld structure refinement

Fig. 8. Variations in the six-member ring and the eight-ring along [1 1 1] apertures. (a) Evolution of O1–O2 (filled squares) and O1–O1 (filled circles) distances for the sixmember, and for O1–O1 (empty squares) and O1–O1 (empty circles) distances for eight-ring as a function of temperature. All O–O distances are normalized with respect to their values at 25 °C, (b) the rotation angle around [1 1 0] vs. temperature; (c) schematic representation of six-apertures and (d) of the eight-apertures in B-LEU.

576

A. Martucci et al. / Microporous and Mesoporous Materials 142 (2011) 570–576

showed that K ions were located in a tunnel along [1 1 1] throughout all of this temperature range (Table 2), and always occupied the centre of the channel delimited by six-membered rings. Above 565 °C, a continuous structural transformation led to an I-43d ? Ia-3d displacive phase transition. The refined unit cell parameter shows a remarkable change, thus indicating drastic structural rearrangement (Fig. 4). This result has already been observed in K-Boron-leucite [32], in K-Cs-Boron-leucite [32], and RbCs-Boroleucite [29]. According to Filatov and Hazen [33], an increase in temperature induces a deformation-structural rearrangement towards higher crystal symmetry due to a rise in the vibrational symmetry of atoms and molecules which can explain the transformation to a more symmetrical high-temperature crystal phase. The structure maintained its Ia-3d symmetry until the highest investigated temperatures. The thermal expansion coefficients in the 565–940 °C temperature range are 1.780  105 and 0.568  103 for a0 and V0, respectively. The continuous structural distortions of the boron leucite framework, provoked by the passage of H2O molecules through the six-member rings during dehydration are now relaxed and the T–O–T and O–T–O angles indicate the formation of more regular apertures (Fig. 8). The formation of wide-open six-member ring apertures is likely to be associated with a relative expansion in the framework, as demonstrated by the increase in both T–O and T–O– T angles with increasing temperatures. This expansion is associated with a volume increase due to thermal tetrahedra expansion, which is expected in the anhydrous phase, and is actually observed in leucite [11]. In all of the present structure refinements, K atoms are statistically distributed over the same positions and are coordinated by 12 oxygen atoms. The K–O bond lengths (Table 3) are in the 3.20– 3.35 Å range, thus indicating that with increasing temperatures, coordination becomes more regular. In agreement with the B MAS NMR spectra of the calcined material [20], our Rietveld structure refinements found no evidence for the presence of trigonal boron in boroleucite heated ‘‘in situ’’, thus confirming the unusually high thermal stability of this zeolite-like borosilicate structure. Acknowledgements The authors would like to acknowledge the Italian MURST for its financial support under the ‘‘Zeolites at non–ambient conditions: theoretical-experimental characterisation and novel technological applications‘‘ PRIN 2006 project. We are also indebted to Carlo

Meneghini (University of Rome) for his assistance during the experiments at the BM08 (GILDA) beamline (ESRF, Grenoble, France). The Italian CNR and INFM should also be acknowledged for having provided financial support to GILDA and its associated facilities. References [1] L. Occelli, J.T. Hsu, L.G. Galya, J. Mol. Catal. 32 (1985) 377. [2] N. Kutz, in: B.L. Shapiro (Ed.), Proceedings of the Second Annual IUCCP Research Symposium in Heterogeneous Catalysis, Texas A&M University Press, College Station, TX, 1984, p. 123. [3] R. Millini, G. Perego, G. Bellussi, Top. Catal. 9 (1999) 13. [4] F. Mazzi, E. Galli, G. Gottardi, Am. Mineral. 61 (1976) 108. [5] D.M. Hatch, S. Ghose, H.T. Stokes, Phys. Chem. Miner. 17 (1990) 220. [6] W.H. Taylor, Zeischrift fuer Kristallographie 74 (1930) 1. [7] J. Wyart, Comptes Rendus de l’Académie des Sciences Paris 205 (1941) 1077. [8] M. Calleri, G. Ferraris, Atti Accademia delle Scienze di Torino 98 (1964) 821. [9] C.R. Knowles, F.F. Rinaldi, J.V. Smith, Indian Mineral. 6 (1965) 127. [10] F. Mazzi, E. Galli, Am. Mineral. 63 (1978) 448. [11] D.C. Palmer, M.T. Dove, R.M. Ibberson, B.M. Powell, Am. Mineral. 82 (1997) 16. [12] C.M.B. Henderson, D. Taylor, Mineral. Mag. 45 (1) (1982) 111. [13] R.M. Hazen, L.W. Finger, Phase Trans. 1 (1979) 1. [14] C. Baerlocher, W.M. Meier, D.H. Olson, Atlas of Zeolite Framework Types, fifth ed., Elsevier, Amsterdam, NL, 2001. pp. 302. [15] T. Armbruster, M. E. Gunter, Natural zeolites: occurrence, properties, application, in: D.L. Bish, D.W. Ming (Eds.), Reviews in Mineralogy and Geochemistry, Vol. 45, Mineralogical Society of America and Geochemical Society, Washington, USA, 2001, pp. 1–57. [16] G. Balassone, A. Beran, G. Fameli, C. Amalfitano, C. Petti, Neues Jahrbuch für Mineralogie Abhandlungen 182 (2006) 149. [17] G.D. Ventura, F. Bellatreccia, M. Piccinini, Am. Mineral. 93 (2008) 1538. [18] G. Bayer, O.W. Florke, Naturwissenschaften 60 (1973) 102. [19] E. Gottardi, G.E. Mazzi, Mineral. et Petrograph. Acta 22 (1978) 185. [20] R. Millini, L. Montanari, G. Bellussi, Microporous Mater. 1 (1993) 9. [21] M. Ihara, F. Kamei, J. Ceram. Sot. Japan 88 (1980) 32. ˇ urovicˇ, Handlovicˇ M. Gyepesová, Acta Cryst. C48 [22] D. Mikloš, L. Smrcˇok, S. D (1992) 1831. [23] F. Di Renzo, M. Derewinski, G. Chiari, J. P1evert, M.F. Driole, F. Fajula, P. Schulz, Microporous Mater. 6 (1) (1996) 151. [24] D. Mazza, M.L. Borlera, C. Brisi, B. Onida, J. Eur. Cer. Soc. 17 (1997) 951. [25] A.C. Larson, R.B. Von Dreele, Report LAUR 86–748 (2000). [26] B.H. Toby, J. Appl. Cryst. 34 (2001) 210. [27] M.G. Krzhizhanovskaya, R.S. Bubnova, S.K. Filatov, P. Paufler, D. Meyer, Abstracts of the II National Crystal Chemistry Conference, Moscow, Russia, 2000. pp. 71–72. [28] R.S. Bubnova, I.G. Polyakova, K.M.G. Rzhizhanovskaya, S.K. Filatov, P. Paufler, D. Meyer, Phys. Chem. Glasses 41 (2000) 389. [29] M.G. Krzhizhanovskaya, R.S. Bubnova, S.K. Filatov, D.C. Meyer, P. Paufler, R.S. Bubnova, S.K. Filatov, D.C. Meyer, P. Paufler, Cryst. Res. Technol. 41 (3) (2006) 285. [30] D.R. Peacor, Z. Kristallogr. 127 (1968) 213. [31] G. Cruciani, A. Gualtieri, Am. Mineral. 84 (1999) 112. [32] R.S. Bubnova, A.A. Levin, N.K. Stepanov, I.A. Belger, D.C. Meyer, I.G. Polyakova, S.K. Filatov, P. Paufler, Z. Kristallogr. 217 (2002) 55. [33] S.K. Filatov, R.M. Hazen, in: A.S. Marfunin (Ed.), Advanced Mineralogy, Vol. 1, Springer-Verlag, Berlin, 1994, pp. 76–90.