PAPER

www.rsc.org/pccp | Physical Chemistry Chemical Physics

A new route for local probing of inner interactions within a layered double hydroxide/benzene derivative hybrid material S. Fleutot,a J. C. Dupin,a I. Baraille,a C. Forano,b G. Renaudin,b F. Leroux,b D. Gonbeaua and H. Martinez*a Received 22nd October 2008, Accepted 2nd February 2009 First published as an Advance Article on the web 2nd March 2009 DOI: 10.1039/b818730d This paper presents the preparation and characterization of hybrid hydrotalcite-type layered double hydroxides (Zn1 xAlx(OH)2HBSx
nH2O, with x = 0.33) where HBS is the 4-phenol sulfonate, with a detailed analysis of the grafting process of this organic entity onto the host lattice. As a set of the usual techniques (XRD, TG-DT/MS, FTIR and 27Al MAS NMR) was used to characterize the hybrid materials, this work focuses on a joint study by X-ray photoelectron spectroscopy and some quantum-calculation modeling in order to highlight the nature of the interactions between the organic and the mineral sub-systems. For the as-prepared hybrid material, the main results lead to a quasi-vertical orientation of the organic molecules within the mineral sheets via H-bond stabilization. By heating the hybrid material up to 200 1C, the structure shrinks with the condensation of the organics; the different theoretical modeling done gives an energy-stable situation when a direct attachment of the HBS sulfonate group sets up with the mineral layers, in agreement with the recorded XPS experimental data.

1. Introduction In the past two decades, layered double hydroxides (LDHs) have been considered as materials with large capabilities for intercalation particularly of organic anions or molecules. These peculiarities result from the specific clay-like network, which offers extensible spaces for such guest entities. Some prolific workers have mentioned such organic–inorganic systems with potential applications as ionic conductors, catalysts, electrodes, biomaterials and materials for optics. The use of the anion-exchange properties of LDHs and the adsorption capacity of the positively charged layers has been extended to the removal of organic pollutants as acid organic species, for example. The properties and applications of these hybrid materials have been the subject of a number of papers1–8 but the understanding of the nature of the interaction between the mineral support and the organic molecules has never been deeply investigated. The various intercalated organic anions may be so different in term of molecular structure, dimension and charge that they are orientated in the interlayers in order to maximize the interactions with the surrounding host layers and lead to specific intralamellar ordering and stacking sequences. The orientations of organic anions depend also on the anionic concentration which can provide self interactions (e.g. aggregation effects, etc.), and the experimental parameters of the reaction (e.g. temperature, pH).9–13 The electrostatic interactions and hydrogen bonds between the a

Universite´ de Pau et des Pays de l’Adour, IPREM CNRS UMR 5254, Helioparc Pau Pyre´ne´es, 2 avenue du pre´sident Angot, 64053 Pau cedex 9, France. E-mail: [email protected]; Fax: +33 (0)5 59407622; Tel: +33 (0)5 59407599 b Laboratoire des Mate´riaux Inorganiques, UMR CNRS 6002, Universite´ Blaise Pascal and Ecole Nationale Supe´rieure de Chimie de Clermont-Ferrand, Clermont Universite´, 63177 Aubie`re Cedex

3554 | Phys. Chem. Chem. Phys., 2009, 11, 3554–3565

layers and the contents (water molecules and compensating anions) of the gallery are known to hold the layers together, forming the three-dimensional stacking structure.14 Beyond the intercalation phenomenon, grafting of organic functions to the mineral support is one of the interesting ways to prepare very-stable pillared or functional systems by combining the chemical activity of the organics with the physical specificities of the support. The functionalization of layers generally goes with an ionocovalent link between the organic and inorganic systems, and initiates new properties for both. Some workers have achieved the grafting of organics onto hydrotalcite substrates9,15–19 but few data report a study devoted to the nature of the chemical bonds built up between the sub-systems. In this paper, an aromatic acid salt, the sodium 4-phenol sulfonate dihydrate salt (HOC6H4SO3Na,2H2O), called HBS salt for short, was intercalated into an hydrotalcite-type layered double hydroxide lattice [Zn2Al(OH)6], either by a coprecipitation method at constant pH or by a hydrothermal mode (thermal treatment after the coprecipitation of reagents). For a couple of years benzene sulfonate derivatives have played an important role in some industrial processes. They are used as intermediates in the production of chemicals such as azo dyes, optical brighteners, detergents, etc.20 Their presence, if not controlled, may lead to a horrendous perturbation of an ecosystem. From the simple organic intercalation state to their condensation/grafting state, a better understanding of the evolution of the considered forces and/or bonds both connecting the subsystems is envisaged. The original way to reach some initial information in this direction is, beside the use of some classical analytical techniques (XRD, FTIR, DTA, NMR), to couple the X-ray photoelectron spectroscopy (XPS) with quantum calculations. The XPS is a surface technique (5–10 nm probed) and allows analysis of the This journal is

c

the Owner Societies 2009

chemical environment of atoms and the electronic evolution in relation with some local changes (e.g. bond formation). To best understand the XPS data, a modeling of the interactions between the organic and inorganic sublattices has been developed by an approach combining some molecular and periodical quantum calculations (for optimizing geometries and analyzing the Mulliken charges).

2. Experimental and computational details 2.1

Materials

2.1.1 LDH precursors and LDH-derivatives materials. The hybrid material, with a general formula [Zn1 xAlx(OH)2)]x+[HBS x,nH2O] with x = MIII/(MII + MIII) (x = 0.33), has been synthesized by the standard coprecipitation method.21,22 The sample was obtained by the coprecipitation of zinc nitrate hexahydrate (Acros organics reagent, 98% pure), aluminium nitrate nonahydrate (Acros organics reagent, 99+% pure) and sodium 4-phenol sulfonate dihydrate (HBS) (Sigma Aldrich organics reagent, 98% pure, CAS: 10580-19-5). A mixture (to obtain a host network with a M(II)/M(III) composition of 2) of Zn(NO3)2
6H2O and Al(NO3)3
9H2O (S[M(II)]+[M(III)]) = 1 M) was dissolved in 200 ml of deionised water. The aqueous solution was slowly added (0.13 mL min 1) under magnetic stirring to 100 mL of HBS solution (0.1 M) at room temperature. Experimental conditions of pH were continuously maintained to a 9.5  0.2 value with controlled addition of NaOH solution (1 M) to ensure the correct building of the host LDH phase. The resulting suspension was stirring for 24 h at 70 1C. The product was thereafter filtered, washed with distilled water and finally kept in a drier at 80 1C for 48 h. Under similar conditions, the parent LDH [Zn1 xAlx(OH)2)]x+[X x,nH2O] (with x = 0.33 and X = Cl ) was also prepared in order to get a starting reference system. The different ICP-AES analyses done on the hybrid material gave Zn2+/Al3+ = 1.98 and HBS /Al3+ = 1.0. In the second step, from an amount of these coprecipitated materials (hybrid and parent), two new samples were prepared after a hydrothermal treatment in order to optimize the structural properties. In the case of the coprecipitation/hydrothermal treatment mode, typically, 50 mg of the coprecipitated product was suspended in 25 mL deionised water in a 30 mL Teflon inner vessel within a stainless steel outer autoclave and left at 120 1C for 72 h under autogeneous pressure. The as-treated sample was then washed with deionised water and dried under room atmosphere. For more convenience, hybrid phases were labelled, according their chemical composition, Zn2Al-HBScop and Zn2Al-HBShyd with index ‘‘cop’’ or ‘‘hyd’’ stating, respectively for the coprecipitation or the coprecipitation/hydrothermal treatment modes of preparation. Parent LDH follow same indexations and were then called Zn2Al-Clcop and Zn2Al-Clhyd. In the following parts, results of the coprecipitation/ hydrothermal treatment mode of preparation will mainly be discussed as this way of synthesis led to better crystalline products with well-defined analytical data. This journal is

c

the Owner Societies 2009

2.1.2 Aluminium complex material. Considering the few data concerning the inorganic–organic interactions, the synthesis of a template aluminium complex was done in order to get some initial informations about data processing of some Al–O–S (iono)covalent bonds sequences (see chapter 3.2). The template material was prepared in a aqueous medium according to the simple acid–base reaction as following: Al(OH)3 (c) + 3 C6H5SO3H (c) - Al(C6H5SO3)3 (c) + 3 H2O (l) A solution of aluminium hydroxide was prepared with 0.3125 g (0.0040 moles) of powder Al(OH)3, nH2O (Sigma Aldrich, 99% pure) in 100 mL of deionised water and was then blended with an acid solution, prepared from 1.9014 g (0.0120 moles) of benzene sulfonic acid (Fluka, 98% pure) in 100 mL of deionised water. The mixed solution was stirred under a 60 1C for a thermal activation during 4 h and maintained afterwards for 2 days long at ambient temperature. The aluminium complex was then filtered and washed with deionised water and held 24 h at 80 1C in a drier. The obtained material was checked up by XRD which has revealed a well-defined crystalline phase different from the parent LDH one. 2.2 Analytical techniques Powder X-ray diffraction (PXRD). After the preparation, all the samples have been systematically analysed by the PXRD technique to monitor the phase composition, the crystalline state and the possible structural changes after organics incorporation. PXRD patterns were obtained with an X-Pert Pro X-ray diffractometer using Cu Ka radiation (1.5418 A˚) equipped with a graphite back-end monochromator and an Argon-filled proportional counter. Powder patterns were first recorded at room temperature for the LDH-precursor and the Zn2Al-HBS hybrid material and afterwards, an in-situ thermal acquisition was achieved on the hybrid material from room temperature up to 1000 1C. Measurements conditions were: diffraction interval 21 o 2y o 1051 (patterns are represented in the papers in a 2–701 range and 4–381 range in Fig. 1 and 5, respectively), step size D(2y) = 0.041 and 0.081 for room temperature and thermal acquisition, respectively; counting time per step 20 s and 4 s for room temperature and thermal acquisition, respectively. X-Ray photoelectron spectroscopy (XPS). XPS analyses were performed on a Kratos Axis Ultra photoelectron spectrometer which employed a magnetic immersion lens to increase the solid angle of photoelectrons collection from small analysis areas to minimise the aberrations of the electron optics. A monochromatic and focused (spot dimensions of 700 mm by 300 mm) Al Ka radiation (1486.6 eV) was operated at 450 W under a residual pressure of 8  10 9 mbar. The spectrometer was calibrated using the photoemission lines of Au (Au4f7/2 = 83.9 eV, with reference to the Fermi level) and Cu (Cu2p3/2 = 932.5 eV); for the Au4f7/2 line, the full width at half maximum (FWHM) was 0.86 eV in the recording conditions. Charge effects were compensated by the use of a charge neutralisation Phys. Chem. Chem. Phys., 2009, 11, 3554–3565 | 3555

shifts are not corrected from the second order quadrupolar effect, which induces shift to lower frequency. Spectra were calibrated against AlCl3. Thermal analysis (TGA-MS). For studying the thermal behaviour of the LDH and derivative materials some thermogravimetric experiments were carried out on a TG model 2950 (TA Instruments) using about 5 mg of sample, from 303 to 923 K (5 K min 1), under a nitrogen flow. The samples were measured under air atmosphere at a heating rate of 51 min 1. The TG-DT apparatus was coupled with a Thermostar 300 mass spectrometer from Balzers Instruments for analysis of evolved gas. FTIR analysis. Attenuated total reflectance Fourier transform infrared (FT-IR) spectra were measured in the range 400–4000 cm 1 on a FTIR Nicolet 5700 (Thermo Electon Corporation) spectrometer equipped with a Smart Orbit accessory. For the recording procedure, a diamond crystal was used and some advanced corrections were performed on the raw data. Grafting procedure. To initiate the grafting process, the hybrid sample was activated by a thermal treatment operated in a tubular oven at 200 1C. The sample was then kept in the oven in a sealed air-tight tube for 24 h which afterwards was opened in a continuously purified nitrogen atmosphere glove box with low rates of O2 and H2O (around 5 ppm for both), directly connected with the X-ray photoelectron spectrometer. This precaution against wet surroundings is necessary to avoid the dehydrated LDH readsorbs water while cooling down and then expanding a little bit. 2.3 Computational details

Fig. 1 Rietveld plots of Zn2Al-Cl (top), Zn2Al-HBS coprecipitated (middle) and Zn2Al-HBS hydrothermally treated (bottom). Observed (crosses), calculated (solid line) and difference (solid line below) are shown with Bragg peak positions for the HDL-chloride (a), HDL-HBS (b) and HDL-carbonated (c) phases. Super-structure peaks, observed for Zn2Al-HBS hydrothermally treated, are indicated by stars.

system (low energy electrons [typically 1.85 eV]) which had the unique ability to provide consistent charge compensation. All the neutraliser parameters remained constant during analysis. High resolution regions were acquired at a constant pass energy of 40 eV. The XPS signals were analysed by using a least squares algorithm and a non-linear baseline. The fitting peaks of the experimental curves were defined by a combination of Gaussian (70%) and Lorentzian (30%) distributions. All the samples were ground prior to analysis to avoid effects due to the surface texture. Solid state NMR. 27Al MAS NMR spectra were recorded on a 300 MHz Bruker spectrometer at 78.204 MHz using a 4-mm diameter zirconia rotor operating at 10 kHz spinning rate in magic angle spinning (MAS) condition. Short radio frequency pulses associated to a recycling time of 2 s were used. Chemical 3556 | Phys. Chem. Chem. Phys., 2009, 11, 3554–3565

The geometries of the Zn2Al-Cl LDH phase were performed using the biperiodical approach within the periodic LCAOB3LYP approximation23,24 developed in the periodic ab initio code CRYSTAL06.25 All electrons basis sets (8-41G* for oxygen,26 8-31G for aluminium,27 86-411d4G for zinc28 and 2-11G* for H29) are adopted. In optimizing geometry, we allowed the relaxation of all atoms. All the molecular and aggregated (mineral cluster in interaction with the HBS molecule) structures were performed at the B3LYP level using the GAUSSIAN98 package30 (basis set: 6-31G*).

3. Results 3.1 Intercalation process 3.1.1 PXRD results. Powder pattern analyses were performed by using the Rietveld method with the program Fullprof.2k.31 The refined cell parameters of the [Zn2Al-Clhyd] parent LDH and the [Zn2Al-HBShyd], [Zn2Al-HBScop] hybrid systems are given in Table 1. The observed, calculated and difference XRD patterns are shown in Fig. 1. While refinements of the Zn2Al-Clhyd pattern were realised by taking into account the whole structural parameters (Fig. 1 top), those of Zn2Al-HBS patterns (Fig. 1 middle and bottom) were perfomed using the profile matching procedure. Refinement This journal is

c

the Owner Societies 2009

convergences were satisfactory by using a hexagonal unit cell with the space group R3m, resulting from a three layer polytype. Nevertheless supplementary peaks are clearly observed in the case of the hydrothermally treated Zn2Al-HBShyd sample (indicated by stars in Fig. 1 bottom). The temperature dependence of these supplementary peaks indicates that they come from a lowering of symmetry and not from impurities (reversible disappearing on heating and cooling around 75 1C). Up to now attempts to find the corresponding correct low symmetry of Zn2Al-HBShyd failed. In the case of the parent LDHs (‘‘cop’’ and ‘‘hyd’’), the XRD patterns have indicate a single phase which correspond to the expected hydrotalcite-like mineral (JPSCD file no. 38-0487). For the most crystallized compound, [Zn2Al-Clhyd], the calculated lattice parameter a = 3.0713(2) A˚ (Table 1) is in good agreement with the experimental Zn2+/Al3+molar ratio of 1.99. The calculated basal spacing (d003 = c/3 = 7.722(3) A˚) corresponds to the presence of the chloride anions between the sheets32 excluding any impurity like carbonate anions.33 The height of the interlayer space has been estimated to be ca. 2.942 A˚ considering the thickness of the brucite-like octahedral hydroxide layer made with Zn and Al (about 4.780 A˚.33,34). For the Zn2Al-HBS materials (‘‘cop’’ and ‘‘hyd’’), the powder patterns clearly indicate the efficiency of the intercalation process (Fig. 1 middle and bottom). It is noticeable that the coprecipitation method leads to a less crystallized phase with larger (001) diffraction peaks and wide and asymmetric (101) reflections due to a turbostratic behaviour as already mentioned in previous reports.34 A clear shift to the small angles (relate to the associated parent LDHs) is noted for the (001) peaks (e.g., a 5.7151 and 5.6561 displacement for the (003) peak, respectively for Zn2Al-HBScop and Zn2Al-HBShyd). Considering these changes in the XRD pattern, it is readily to consider a real enlargement of the interlayers space with a small change in the planar sheets structure (Da o 0.013 A˚). The basal spacing of the layered structure of LDHs is mainly influenced by the orientation of the inserted anions in the interlayer space. To best interpret the intercalation process, a first simulation of the organics orientation within the host system has been performed considering XRD powder data. The height of the gallery (7.912 A˚ and 7.757 A˚, respectively for Zn2Al-HBScop and Zn2Al-HBShyd , see Fig. 2) can be deduced by subtracting from the basal spacing (15.392 A˚ and 15.238 A˚, respectively for Zn2Al-HBScop and Zn2Al-HBShyd) the value of a single sheet thickness increased by the hydrogen bond domain (ca. 2.700 A˚ thick in the hydrotalcite case35) on one side of the inorganic layer (7.480 A˚). The resulting value gives the space available for the organic and the hydrogen bonds network (Fig. 2). Table 1

Refined lattice parameters of the synthesized products

Samples

a/A˚

Zn2Al-Cl Zn2Al-HBScop Zn2Al-HBShyd Zn2Al-HBShyd 200 1C

3.0713 3.0640 3.0579 3.0560

This journal is

c

c/A˚ (2) (1) (2) (1)

23.1670 46.1760 45.7130 33.7800

the Owner Societies 2009

Symmetry (3) (6) (5) (3)

R3m R3 R3 R3

From those considerations, the available interlayer space for the organic entities is comprised between 7.912 A˚ (‘‘cop’’) and 7.757 A˚ (‘‘hyd’’) depending on the preparation conditions. The basal spacing for the highly ordered Zn2Al-HBShyd compound is shorter due to the optimisation of the interactions and distances between both guest and host partners. Taking into account the result of an optimized structure calculation of the HBS molecule giving a 7.90 A˚ size, two possible assumptions for the HBS molecules orientations in the inter-layers space were done: an oblique orientation (possible for both modes of preparation) or a vertical positioning (only possible for the ‘‘hyd’’ mode). In both cases, the sulfonate groups of HBS interact with the hydroxyl functions of the brucite-like layers (via H-bonds as reported in the literature8,9,15,36). On the basis of these whole results, it seems then appropriated to propose a first scheme (Fig. 2) of an arrangement of HBS molecules via H-bonds stabilization for these systems. 3.1.2 XPS results. In addition to these experiments, a XPS examination has been performed to primarily determine the homogeneity of the LDH-intercalate. The Zn/Al atomic ratio was found around 1.9 which is in good agreement with the expected value of 2. XPS, via the analysis of the binding energies, has also revealed the precise local environment of the atoms. To get some coherent results, energetic regions were chosen that related closely to same free mean-path electrons (Zn3p3/2–1/2, Al2p3/2–1/2, O1s and C1s core peaks). For these energetic regions, the obtained binding energies (Table 2) correspond to the values of the metal hydroxides environments37,38 as previously noted for the parent host systems. With the intercalation of HBS entities, no significant changes are observed for these binding energies and their relative full width at half maximum [FWHM] (except for the O1s peak); the chemical structure of the mineral layers is then kept unchanged and not chemically sensitive to the insertion of the organic molecules. Moreover, the synthesis reaction was controlled by monitoring the nitrogen N1s signal to avoid any residual nitrate amounts from reagents which could disrupt the characterization of the intercalated products. For any prepared hybrid material, the N1s signal was void then attesting of a single hybrid phase as already suggested by the XRD patterns. To get a direct overview of the intercalation process, it was useful to run some close up surveys on small restricted energetic areas. As the concerned hydrated HBS salt gets a sulfur atom in its formula, it was appropriated to use the S2p signal as a specific probe of the electronic evolutions of the organic molecules during the intercalation process. The advantage of the S2p core level ionization is the significant sensitive factor of photoemission39 and the large scale of chemical shifts associated with different oxidation states. Indeed, the binding energy scale of sulfur atoms spreads over 17 eV, from sulfur atoms of type 2 [BES2p3/2 = 160.0 eV] (as encountered in Na2S for example) up to an oxidation state +6 [BES2p3/2 = 177.0 eV] (for SF6). The main conclusion deduced from the XPS analysis is the observation of the same 2p3/2–1/2 binding energies (168.0 eV–169.2 eV) for the sulfur doublet of the HBS salt reference powder (Fig. 3a) and the Zn2Al-HBS prepared Phys. Chem. Chem. Phys., 2009, 11, 3554–3565 | 3557

Fig. 2 Table 2

Scheme of a possible conformation of the incorporated HBS salt in the hydrotalcite host matrix.

XPS data: binding energies (eV) with FWHM (eV) in parentheses of the hybrid, pillared materials and their precursors

Compound BE region

Zn2Al-Cl LDH

Hydrated HBS salt

Zn2Al/HBShyd

Zn2Al/HBShyd calcined (200 1C)

Zn3p3/2–1/2 C1s

89.2–92.2 (2.6) 284.9 (1.4) 286.6 (1.4) 289.3 (2.1) 531.8 (1.6)

— 284.6 (1.0)

89.3–92.2 (2.7) 284.6 (1.3) 286.0 (1.3) 288.9 (1.4) traces 531.8 (1.8)

89.3–92.3 (2.8) 284.6 (1.4) 285.9 (1.9)

O1s a

Al2p S2p3/2–1/2 a

74.6 (1.5) —

531.6 (1.3) 532.8 (2.0) — 168.0–169.2 (0.9)

74.6 (1.6) 168.0–169.2 (1.5)

530.5 (1.6) 531.8 (1.9) 74.6 (2.5) 168.6–169.9 (1.9)

XPS accuracy could not allow differentiation of the two spin orbit energetic levels (2p3/2 et 2p1/2) of the Al2p peak.

material (Fig. 3b). The only remarkable evolution is the increase of the full-width-at-half-maximum (FWHM) for the hybrid sample (1.5 eV instead of 0.9 eV for the HBS salt). This indicates a less-ordered chemical-surrounding for the sulfur atoms for the intercalate material, in agreement with the widening of the O1s peak. At this stage of the study, the intercalation of the 4-phenol sulfonate in the Zn2Al(OH)6 hydrotalcite type matrix is easy to monitor with the XRD analysis considering the clear shift of the (00l) diffraction peaks. The weak interactions between the two sub-systems after the insertion step do not allow any clear analytical evolution, despite the high sensitivity of the XPS technique. 3.2

Grafting process

According to the first results obtained for the intercalation phenomenon, thermal treatment was carried out to get pillared materials which have micro and/or mesoporous specificities and present a potential interest for various applications.40–45 The possibility of grafting HBS anions onto the hydrotalcite layers may be an opportunity to best appreciate the complementary nature of the XPS analysis and some quantum calculations, which both ought to differentiate electrostatic type interactions (the intercalation step) from more covalent ones (the grafting step). To this aim, an initial TG experiment was initiated to investigate the behaviour of the composite material with temperature, which could be a first sign of sub-lattice interactions. Moreover, to reach the grafting state, 3558 | Phys. Chem. Chem. Phys., 2009, 11, 3554–3565

Fig. 3 XPS S2p peak of: (a) HBS salt as received, (b) Zn2Al-HBScop hybrid material, (c) the same Zn2Al-HBScop hybrid material after thermal treatment at 200 1C.

This journal is

c

the Owner Societies 2009

a precise control of the temperature is necessary in order to avoid structural collapse of the material. 3.2.1 Thermal treatment. Generally, in the literature, combustion of organic-LDH derivatives is often cited as a well-known thermal event.46,47 In the present work, the comparative TG-DT analysis concerns three samples: the hydrated HBS precursor salt; the Zn2Al-Cl LDH hydrotalcite matrix; and the Zn2Al-HBShyd hybrid material. Note that the thermal study of the coprecipitated phase has been also investigated and gives similar results than for the hydrothermally treated materials. Through the following quick DT-TG survey, we could readily value the adapted temperature range for the grafting procedure without collapsing the hybrid material architecture. In Fig. 4d, the host matrix material presents a three endothermic main thermal events, corresponding to well-known differentiated weight losses. According to the literature,41 the first peaks between ambient and 200 1C are due to physically adsorbed water on the external surface of the crystallites and to the LDH interlamellar water. Beyond 200 1C, the dehydroxylation of brucite-like layers occurs. This last event generally depends on several parameters: the cation ratio M(II) : M(III) in the brucite-like layer; the nature of charge compensation anions; and the degree of crystallinity of the LDH lattice. Considering the combustion of the pure HBS salt (Fig. 4c and f), the peaks just above 500 1C (decomposition of HBS molecules, ca 35% of weight loss) are shifted by about 50 1C when the salt is trapped into the mineral matrix (Fig. 4b and e). This indicates some real interactions between the organic and the brucite like layers and an enhancement of the organic thermal stability. For the hybrid material, the two first weight losses (9.25% and 12.46%) are completed at 110 1C and 293 1C, respectively. At this stage, the organic molecule is retained in the calcined structure. These events only represent the release of the interlayer water molecules and the dehydroxylation of the mineral sheets. It can be noticed that the dehydroxylation is well-observed at a lower temperature for the parent LDH for which compensating anion– mineral sheet interactions should be less-important than for

Fig. 4 TG and DT analysis curves of the host Zn2Al-Clhyd hydrotalcite (a) and (d), the hybrid Zn2Al/HBShyd material (b) and (e), and the HBS salt (c) and (f).

This journal is

c

the Owner Societies 2009

the hybrid material. In these conditions, from this DT analysis, it appears more suitable to study the grafting process below a temperature of 230 1C, at which the hybrid material starts to degenerate under dehydroxylation. 3.2.2 In situ PXRD results versus temperature (HTK PXRD). Once the temperature domains of combustion determined, an in situ PXRD analysis was performed by gradually increasing the temperature from room temperature up to 1000 1C in the analysis chamber as reported in some previous works.15,21 The different recorded powder patterns are superimposed and displayed for temperature up to 600 1C in Fig. 5. The general profile-evolution follows a three step process. For a thermal treatment lower than 80 1C in the controlled temperature chamber, the hybrid material retains its original crystalline layered structure while the (001) and (h01) lines are still well-defined and do not shift. Then, a structural transition is observed in a very narrow temperature-range around 80 1C, which corresponds to a pronounced shift of the (001) peaks to the high 2y angle side corresponding to a 2.05 A˚ contraction of the inter-layer space. This compression in the stacking planes direction is simultaneous with the first loss of water molecules (DT data, Fig. 4). A special observation was made at 80 1C with a stronger contraction than for the following temperature set (see inset in Fig. 5). The Zn2Al-HBShyd refined lattice parameters readily traduce the contraction with a 6.15 A˚ variation of the c lattice parameter value (c = 39.62(3) A˚); the unchanged a lattice parameter (a = 3.057(9) A˚) testifies to the preserved LDH structure (Table 1). Beyond 200 1C, the powder pattern drastically changes with the disappearance of most of the reflection lines (collapse of the HDL-type structure), a single diffraction line being retained at a low angle. In this range of temperatures, the inter-layer contraction gets its top-level value (contraction of 3.978 A˚) before the collapse of the lamellar lattice, leading to the shortest basal

Fig. 5 XRD powder patterns (lKa Cu = 1.5418 A˚) with in situ temperature measurements of the Zn2Al-HBS hybrid from 25 1C to 600 1C. Evolution of the Bragg peaks (003), (006), (009) and (101) are indicated (using the rhombohedral R3 symmetry, usual for HDL compounds). Insert, at the top of the Figure, shows the variation of the interlamallar distance, the refined d003 distance, as a function of temperature.

Phys. Chem. Chem. Phys., 2009, 11, 3554–3565 | 3559

spacing d003 = 11.26 A˚. Thus, the temperature of 200 1C appears adequate for grafting without destroying the whole LDH network. Indeed, as observed, there is a temperature delay for the HTK XRD static experiment compared to dynamical TGA analysis. 3.2.3 FTIR and NMR analysis. The change under thermal treatment in the local environment of the Al3+ intra-layer cations was investigated by 27Al NMR spectroscopy. There is clear evidence to show that the local surroundings of Al are not modified during the calcination either at 80 or 200 1C, as all the spectra are superimposable (Fig. 6): the chemical shift as well as the width of the resonance line are both unchanged. This means that the 27Al nuclei presents in the three samples the same local environment and symmetry as the strength of the interaction dipolar and quadripolar. All the cations keep the initial octahedral environment Al(OH)6, as no conversion toward tetrahedral site, which is usually observed in such cases, is seen. This corresponds to a minor phase. Most of the layer structure is still retained at 200 1C under the so-called layered double oxide (LDO) state. Fig. 7 shows the ATR FTIR spectra of HBS, Zn2Al-HBShyd and pillared Zn2Al-HBShyd (treated at 200 1C). The F letter has been chosen to represent the aromatic cycle of HBS molecules. At room temperature, the spectrum of the hybrid Zn2Al-HBShyd material (Fig. 7b) displays the characteristics bands of an hydrotalcite with mainly the H2O bending vibration of interlayer water (1620 cm 1) and the metal– oxygen–metal bending in the range 400–750 cm 1.48 The changes in the headgroup regions can be studied through the inspection of the evolutions in the S–O and F–O stretching bands of the HBS molecules. The bands at 1448 cm 1 and 1370 cm 1 can be assigned to the vibration modes of the bond between the aromatic ring and this hydroxyl group. The symmetric S–O stretching vibration bands are observed at 1005, 1030 cm 1 and 1125 cm 1 (ns) while the asymmetric bands of the SO3 group arise at 1167 cm 1 and 1196 cm 1 (nas).49,50 By heating the sample up to 200 1C, some of the free vibration modes of the mineral sheet’s hydroxyl groups (e.g., dM–OH at 774 cm 1) modify when the HBS hydroxyl group (e.g., dF–OH at 1460 cm 1) clearly disappear (Fig. 7c). Those of the S–O group in the 1000–1200 cm 1 become more complex and broaden strongly; this is significant of a change in

Fig. 6 27Al MAS NMR of the Zn2Al-HBShyd samples, the temperature of treatment is indicated. The asterisks represent the spinning side bands.

3560 | Phys. Chem. Chem. Phys., 2009, 11, 3554–3565

Fig. 7 FTIR spectra: (a) HBS salt; (b) Zn2Al/HBShyd (intercalation); (c) Zn2Al/HBShyd treated at 200 1C (grafting).

the local oxygen environment around the sulfur atoms. The group of bands at 1167 cm 1 and 1196 cm 1 shifts to 1132 cm 1, which can be definitely assigned to a vibration mode of more-negative sulfur surroundings.51 Through these results, the grafting process seems to be effective with the involvement of the mineral sheet (via their hydroxyl groups) and both the headgroups of the HBS molecules, the hydroxyl and sulfonate functionals. At the first level of interpretation the FTIR spectrum evolution indicates a pillaring process, by heating up the hybrid material with a stabilization of the organics (HBS) into the LDH interlayer space. 3.2.4 XPS results. As done previously to investigate the intercalation process, an XPS study was run onto the pillared LDH systems. The quantitative survey has shown for the first first time a clear variation in the atomic oxygen content (around 34% decrease of the atomic O/Al ratio) after the calcination at 200 1C of the hybrid system. Considering the fact that LDH interlayer water molecules (corresponding initially to 25% of the whole oxygen content in the hybrid material) have been released from the host matrix, a real dehydration of the mineral sheets can be assumed. Even if a partial dehydration process can be considered, the close-up XPS spectra of the metal atoms constituting the mineral sheets (zinc and aluminium) underline a pretty well-preserved hydroxide environment (Table 2). For the aluminium peak, a clear widening is observed, showing the possible involvement of aluminium atoms in the grafting process. The O1s profile changes with a shoulder on the low binding energy side at about 530.5 eV, revealing a more negative character for part of the oxygen atoms. With the same approach used for the examination of the intercalation process, the specific S2p peak area was also recorded as a direct probe of the interactions between sublattices. If any ionocovalent link is formed with the thermal treatment, one might naturally think about the role of the sulfonate group of the HBS salt molecules. The S2p spectrum at 200 1C reveals a significant shift (+0.6 eV) on the high binding-energy side and an increase of the FWHM (Fig. 3c). This journal is

c

the Owner Societies 2009

Table 3 studied

XPS data, DEL (O1s–S2p3/2) (eV) of the different samples

Compounds

DEL (O1s–S2p3/2)/eV

Al(C6H5SO3)3 complex Zn2Al/HBShyd—200 1C Zn2Al/HBShyd—room T HBS salt

363.2 363.3 363.8 363.7

This corresponds to more positive sulfur-atoms and then to a change in the S–O bond character (ionocovalent bonds). A new electronic distribution between the aluminium atoms of the mineral sheets and the sulfur of HBS via some more negative oxygen atoms seems to occur. In order to avoid some calibration artefacts, a set of samples were compared on the basis of the DBE (O1s–S2p) binding energy difference. The DBE determination is a sensitive way to get an overview of the electronic density between the two elements with a more-or-less ionic (or covalent) character. Generally, if a O1s peak shift occurs to the low binding energy side (oxygen atoms more negative) simultaneously with a high-energy side-shift of the S2p (sulfur atoms more positive), the direct interaction between the two associated atoms gets more ionic. Four samples were monitored, the template Al-complex, the Zn2Al-HBShyd hybrid material before and after calcination and the pure hydrated HBS salt. The results for the DBE(O1s–S2p) deduced from the XPS data are reported in Table 3. The obtained data have shown two different trends in the recorded values. A same value (363.2  0.1 eV) for the template Al complex and the pillared LDH (hybrid system at 200 1C) is observed. In the other hand, an identical value (363.7  0.1 eV) is noticed for the hybrid system and the pure HBS salt. These observations lead us to consider similar S–O bonds for the hybrid material at 200 1C and for the template Al sample, which arranges structurally by ionocovalent links. In regard to these results, XPS has shown the involvement of the sulfonate group of the HBS molecules in the grafting process but the data could not reveal the precise role of the HBS hydroxyl group. The O1s spectrum becomes very complex with the grafting phenomenon (–OH groups and –S–O–R functions having their peaks overlap) and it cannot clearly indicate the sub-lattice interactions via the organic hydroxyl group, if existing.

4. Theoretical calculations and discussion Considering the intricate problem of simulating the interactions between organic and mineral entities (no such interest being reported in the literature for a LDH-based hybrid material), the considered quantum calculations reported in this paper were performed with a molecular approach. As the aluminium atoms are positive centers, we consider the aluminium hydroxide fragment as the active agent in the organic–inorganic interference and we reduce the mineral network to the Al(OH)3 cluster. Then, in a first step, for the interacting crystallized systems (mineral matrix and HBS salt) a set of periodical calculations were undertaken to extract the This journal is

c

the Owner Societies 2009

Fig. 8 (a) Zn2Al(OH)6 slab, (b) local optimised geometry surrounding aluminium atoms.

most-realistic local geometry, especially around the Al atoms in the LDH phase. For the host matrix, the procedure consists of two steps: (i) We perform a periodic calculation (by using the CRYSTAL06 code) of the Zn(OH)2 phase, (Mg(OH)2 brucitelike structure52,53). The structure Zn(OH)2 crystallizes in the CdI2 system (P 3m1, space group no 164) and can be readily described as a simple juxtaposition of stacked layers along a crystallographic axis. The slabs of octahedrons are made of two hydroxyl layers (–OH) sandwiching a metallic plane (Zn). The hydroxyl layers describe a hexagonal close packing with the metal occupying half of the octahedral sites. We have defined an unit cell (Zn(OH)2 slab consisting of 5 layers) which is periodically repeated in two dimensions and built from the optimized lattice parameters of the bulk as a starting point (a = b = 3.278 A˚, c = 4.732 A˚). The relaxation of the surface allows us to obtain an equilibrium geometry (a = b = 3.243 A˚), in agreement with DFT calculations54 achieved on brucite-like compounds such as Zn(OH)2 or Mg (OH)2; (ii) The whole Zn2Al(OH)6 hydrotalcite sheets considered in this work were obtained by forming a 3  3 supercell with the partial substitution of some Zn atoms by some Al atoms (Fig. 8a). The new generated slab geometry was then optimized (lattice parameters a = b = 9.673 A˚) and the resulting local geometry surrounding the aluminium atoms is reported in Table 4. For the organic entity, the molecular geometry (dehydrate sodium HBS salt) was calculated (6.31 G* basis set) by referring to the crystallographic data from the literature55 (space group P 1). All the calculations have been achieved by considering two 4-phenol sulfonate ions (HBS ), each of them interfering with a sodium ion (Na+) and being encircled with three water molecules (Fig. 9). The optimization of the lattice parameters with the CRYSTAL06 code gives the following results: a = 8.737 A˚; b = 12.480 A˚; c = 6.149 A˚; a = 90.001; b = 106.121; g = 101.19. The resulting Mulliken charges analysis leads to some net charges equal to 0.9 e and +0.8 e for the anionic HBS sub-system and the sodium cations, respectively, according to the considered molecular structure HBS /Na+/H2O. Note that the population analysis of the sulfur atoms leads to qS = +1.12 e . This result, discussed in the following section, is especially suitable when compared to the main XPS Phys. Chem. Chem. Phys., 2009, 11, 3554–3565 | 3561

Table 4 Intercalation model-bond lengths (A˚), angles and energies (a.u) after optimization of the two molecular fragments (HBS salt and Al(OH)3) [first column], and the hybrid material (O1QO of the sulfonate group from HBS; O3QO from the mineral framework) [second column] HBS (molecular approach)

Hybrid system HBS –Al(OH)3 (molecular approach)

d(S–O1)/A˚ d(S–C)/A˚ O1–S–O1/1 O1–S–C/1 E/eV

1.488 1.824 114.8 103.7 25325.191

1.492 1.824 114.2 103.7 38110.301

d(Al–O3)/A˚ d(O3–H)/A˚ O3–Al–O3/1 Al–O3–H/1 E/eV

Al(OH)3 (periodical approach) 2.100 0.958 97.1 120.0 12783.710

2.084 0.958 99.5 120.9 38110.301

4.1 Intercalation

Fig. 9 Configuration considered for the calculation of the sodium HBS dihydrate salt.

The modeling of the intercalation situation considers the interaction of the anion HBS (its geometric structure is deduced from the previous calculations by removing the cation Na+ and H20 molecules) with the mineral network (reduced to Al(OH)3 cluster) (Fig. 8b). The HBS entity is positioned in front of the optimized Al(OH)3 fragment (Fig. 10). The oxygen atoms from the SO3 group are positioned below the hydrogen atoms of the mineral framework, with a d(O1–H) = 2.526 A˚ which corresponds to the interaction domain size of the hydrogen bonds31 as previously reported in the experimental section. The geometry of the as constituted hybrid system was then optimized (Table 4). A slight increase of the S–O1 bond lengths within the organic entity is noted to be simultaneous with a significant reduction of the distance between O1 and H of Al(OH)3 (1.995 A˚). At the same time, the angles O3–Al–O3 and O1–S–O1 are slightly modified (Table 4). The resulting net charge of sulfur atoms for the hybrid system [HBS/Al(OH)3], qS = +1.16 e is similar to the hydrated salt (qS= +1.12 e ). As an initial approximation, by simply considering an initial state effect, this small evolution of the charges agrees with the absence of a chemical shift for the XPS S2p3/2 core peak. Moreover, it confirms the weak interaction between sub-systems via hydrogen bonds. Note that the hydrogen O


H bond length between the oxygen

probe (sulfur peaks) studied in this paper. Indeed, it is wellestablished that the binding energies determined by XPS of the core electrons of an atom are dependent on the chemical state of this atom. Generally, as the positive character of an atom increases, so do its core level binding energies. Different relations between chemical shifts and real charge variations have been used successfully.56 At the simplest level of approximation, the correlation between binding energy (BE) and charge can be written as follows: DBE = k
Dq, where DBE is the experimental shift of binding energy, Dq is the charge variation, and k is a constant characteristic of the element. In the following, the different atoms acting in the intercalation or in the grafting processes have been indexed as following: oxygen atoms are called, O1 for the sulfonate headgroup, O2 for the hydroxyl headgroup and O3 for the mineral sheet. The F letter has been chosen to represent the aromatic cycle of HBS molecules.

Fig. 10 Model of the interaction between Al(OH)3 and HBS during the intercalation process.

3562 | Phys. Chem. Chem. Phys., 2009, 11, 3554–3565

This journal is

c

the Owner Societies 2009

atoms from the sulfonate group (O1) and H from the Al(OH)3 (1.995 A˚) is close to the hydrogen O


H bond between the same oxygen atoms (–SO3 ) and the hydrogen atom of water molecules (2.147 A˚); this last distance being deduced from the previous calculation within the HBS hydrated salt structure. 4.2

Grafting

The experimental attempts of grafting (after a thermal treatment at 200 1C) have displayed a contraction of the interlamellar distance, explained by the condensation of the anions on the partially dehydroxylated mineral layers, via the formation of Al–O–S bonds. To further investigate and model this process, two different assumptions of iono-covalent interactions could be envisaged, assuming that both head functional groups of the HBS entities (–OH and –SO3 ) may create a chemical link. Then, two different configurations were calculated and optimized considering the partial dehydration process of the mineral sheet (some hydroxyl groups of the Al(OH)3 cluster being cut). As for the intercalation process some geometrical constraints on the mineral cluster and the organic molecule should be introduced in order to preserve the structure around the aluminium atoms in agreement with the previous NMR and DRX results. In the case of the interaction via the sulfonate headgroup, an initial assumption was to consider the formation of an unique covalent link Al–O1–S (Fig. 11a). The most stable configuration was obtained by considering four degrees of freedom: the S–O1 and Al–O1 bonds, the O1–S–CF angle, and the dihedral Al–O1–S–O1. The optimization of these parameters leads to a decrease of 0.084 A˚ of the S–O bonds (1.516 instead of 1.600 A˚) and of 0.182 A˚ of the Al–O1 bond compared with the

Fig. 11 (a) Grafting model (first assumption of a one bond attachment): geometric structure with restricted degrees of freedom: interaction between Al(OH)3 and an oxygen atom O1 of SO3 ; (b) grafting model (second assumption of a one bond attachment): interaction between Al(OH)3 and the oxygen atom O2 (from the hydroxyl group).

This journal is

c

the Owner Societies 2009

one of the Al (OH)3 cluster (Table 5). Simultaneously, the O1–S–CF angle opens up (5.61) whereas the dihedral Al–O1–S–O1 closes up from 21.0 to 7.21. Note that the bond overlap population S–O1 (0.182) decreases compared with that calculated for the HBS anion (0.313), whereas it increases for the two other free S–O1 bonds. This contraction is also observed for the Al–O1 bond in the Al–O1–S sequence (0.144 instead of 0.260 for Al–O3 bonds of Al(OH)3). The Mulliken charges analysis indicates a sulfur net charge of qS = +1.25 e , in agreement with the XPS chemical shift towards high binding energies (+0.6 eV) of the sulfur core levels after the thermal treatment. This value is to be compared with that obtained for the calculation of the hydrated salt (+1.12 e ). In the case of the interaction via the hydroxyl headgroup (Fig. 11b), the same constraints as before were considered. In order to study a neutral charged system and a constant number of atoms, the hydrogen atom from the OH group was moved and connected to an oxygen atom from the sulfonate group. With this configuration, the Al–O2–CF angle and the dihedral Al–O2–CF–CF were optimized (Table 5). The bond overlap population of Al–O2 (0.170) in the Al–O2–CF sequence is weaker than the other Al–O (from Al(OH)3) bonds (0.260). The same trend is observed for the C–O2 bond in the C–O2–Al sequence. The net charge of the sulfur atom is qS = +1.11 e , which is similar to the charge observed for the HBS salt (qS = +1.12 e ). The stabilization energy of this configuration can be compared to the previous one (Fig. 11a, interaction via the sulfonate group with restricted degrees of freedom). For these two assumptions of a single link between the mineral cluster and the organics, the formation of the Al–O1–S ionocovalent link is more energetically stable than the Al–O2–CF (DE = 0.46 eV) and leads to a sulfur net charge in agreement with XPS data. Note that two other configurations could be considered, as the sulfonate group presents three oxygen atoms arranged in a tetrahedron. A double or a triple attachment via oxygen atoms may be then considered. The first step is obviously to check if the geometrical constraints imposed by this type of interaction are coherent with those of the HBS anion. We have then achieved the molecular calculation of a HBS anion with the sulfonate group, with the necessity of slightly distortion to preserve the O3–O3 distance (2.90 A˚) in the mineral matrix. As a consequence, only an interaction of the aluminium atom with two oxygen atoms is possible. The most stable configuration was obtained by considering 4 degrees of freedom: the 2 S–O1 distances; the O1–S–CF angle; and the dihedral Al–O1–O1–S. This configuration leads to a sulfur net charge equal to +1.25 e , very close to the previous single attachment with the sulfonate headgroup. Considering all these results, the grafting process can be envisaged from both headgroups of the organics with, however, from a thermodynamic point of view, a favoured interaction via the sulfonate function. In another hand, it seems difficult to define a particular method of anchorage, as the single or double link both correspond to coherent situations with the XPS experimental observations. Phys. Chem. Chem. Phys., 2009, 11, 3554–3565 | 3563

Table 5 Grafting model-bond lengths (A˚), angles, dihedral angles and overlap populations of the two configurations considered with a one bond attachment: interaction between the local geometry surrounded aluminium atom and SO3 (first column) or OH (second column). NB: O1QO of the sulfonate group from HBS; O2QO of the hydroxyl group from HBS

Al–O2–CF/1 Al–O2–CF–CF/1 d(S–O1)/A˚ O1–S–CF/1 Al–Ol–O1–S/1 d(Al–O1)/A˚ Over. pop. S–O1 Over. pop. Al–O1 Over. pop. C–O2 Over. pop. Al–O2 Over. pop. Al–O3 E/eV

Interaction with SO3 single attachment

Interaction with O (hydroxyl group)

Non optimized

Non optimized

Optimized

120.0 90.0

157.6 117.8

0.295 0.260 0.260 36043.337

0.262 0.170 0.250 36043.881

1.600 104.3 21.0 2.100 0.313 0.260 0.260 36043.502

Optimized

1.516 109.9 7.2 1.918 0.182 0.144 0.250 36044.359

5. Conclusion

Acknowledgements

In this work, taking advantage of the lamellar structure of LDHs, 4-phenol sulfonate salt is easily incorporated within the mineral interlayer space. The aim of the study is, for such an LDH/organics hybrid system, to initiate a careful consideration of the inorganic/organic subsystem interactions. Our approach consists of two steps: a close-up survey of the intercalation process and the stabilization of the as-made hybrid material; and afterwards, the examination of a pillared material configuration deduced from a thermal treatment of the hybrid material at 200 1C. The first results put into evidence strong similarities whatever the hybrid material mode of preparation (salt coprecipitation mode or hydrothermal mode; this last one leading to more crystalline products with very well-defined analytical data). After intercalation, the enlargement of the mineral host matrix is observed in quite the same proportions and the XPS data agree with H-bond stabilization of the organics. The second objective of the work is to constrain the hybrid material to a soft thermal treatment to finally get a stable pillared solid. The XRD results have affirmed a consequent contraction of the mineral interlayer space (ca. 4.0 A˚) with the condensation of the HBS organic molecules onto the LDH layers. In complement to these observations, the XPS survey has shown a clear evolution of the sulfur core peaks (+0.6 eV shift) and an oxygen content decrease (occurring with the dehydration of the mineral layers) which can be explained through the consideration of a new electronic distribution along the S–O bonds of the HBS sulfonate headgroup. The present changes have been correlated to a template aluminium complex for which the S–O–Al bonds sequences present the same XPS response. With the grafting process, the net charge on the sulfur atoms has drastically changed (+0.13e ) and an ionocovalent attachment with the mineral layers is established. The different modeling cases developed have proposed an attachment via both headgroup (sulfonate and/or hydroxyl), a bridging link to the LDH layers being more energetically stable with the sulfonate function. This work is now extended to a large variety of organic molecules with a particular interest in the nature of the headgroups and the influence of the mineral matrix composition and reactivity.

Calculations have been carried out on an IBM/P1600Power5+ computer at the Centre Informatique National de l’Enseignement Supe´rieur (CINES). We thank the scientific council of CINES for their support to this project.

3564 | Phys. Chem. Chem. Phys., 2009, 11, 3554–3565

References 1 G. R. Williams and D. O’Hare, J. Mater. Chem., 2006, 16, 3065–3074. 2 A. I. Khan and D. O’Hare, J. Mater. Chem., 2002, 12, 3191–3198. 3 C. Sanchez, B. Lebeau, F. Chaput and J.-P. Boilot, Adv. Mater., 2003, 15(23), 1969–1994. 4 P. R. Somani and S. Radhakrishnan, Mater. Chem. Phys., 2002, 77, 117–133. 5 C. Del Hoyo, Appl. Clay Sci., 2007, 36(1–3), 103–121. 6 J.-H. Choy, S.-J. Choi, J.-M. Oh and T. Park, Appl. Clay Sci., 2007, 36(1–3), 122–132. 7 P. Gomez-Romero, Adv. Mater., 2001, 13(3), 163–174. 8 (a) V. Rives, LDH: Present and Future Nova Science, New York, 2001; (b) P. S. Braterman, Z. P. Xu and F. Yarberry, in Handbook of Layered Materials, ed. S. M. Auerbach, K. A. Carrado and P. K. Dutta, Marcel Dekker, New York, 2003. 9 S. Carlino, Solid State Ionics, 1997, 98, 73–84. 10 C. Vaysse, L. Guerlou-Demourgues and C. Delmas, Inorg. Chem., 2002, 41(25), 6905–6913. 11 Y. Liu, J. Su and J. Wu, Inorg. Chem., 2005, 44(11), 3890–3895. 12 L. Benes, V. Zima, J. Svoboda and K. Melanova, J. Inclusion Phenom. Macrocyclic Chem., 2005, 53, 41–46. 13 E. Kanezaki, J. Inclusion Phenom. Mol. Recognit. Chem., 1996, 24, 341–351. 14 (a) R. Allmann, Chimia, 1970, 24, 99; (b) M. Franc¸ois, G. Renaudin and O. Evrard, Acta Crystallogr., Sect. C, 1998, C54(9), 1214–1217; (c) G. Renaudin, M. Franc¸ois and O. Evrard, Cem. Concr. Res., 1999, 29(1), 63–69; (d) G. Renaudin and M. Franc¸ois, Acta Crystallogr., Sect. C, 1999, C55(6), 835–838; (e) G. Renaudin, J.-P. Rapin, B. Humbert and M. Franc¸ois, Cem. Concr. Res., 2000, 30(2), 307–314; (f) G. Renaudin, J.-P. Rapin, E. Elkaim and M. Franc¸ois, Cem. Concr. Res., 2004, 4(10), 1845–185. 15 V. Prevot, C. Forano and J.-P. Besse, Appl. Clay Sci., 2001, 18, 3–15. 16 C. Vaysse, L. Guerlou-Demourgues, A. Demourgues and C. Delmas, J. Solid State Chem., 2002, 167, 59–72. 17 F. Malherbe and J.-P. Besse, J. Solid State Chem., 2000, 155, 332–341. 18 U. Costantino, M. Casciola, L. Massinelli, M. Nocchetti and R. Vivani, Solid State Ionics, 1997, 97, 203–212. 19 C. Depe`ge, L. Bigey, C. Forano, A. De Roy and J.-P. Besse, J. Solid State Chem., 1996, 126, 314–323.

This journal is

c

the Owner Societies 2009

20 B. Elvers, J. F. Rounsaville and G. Schulz, Ullmann’s Encyclopedia of Industrial Chemistry, ed. B. Elvers, Basel, Cambridge/New-York, 1991, vol. A17, p. 9. 21 V. Prevot, C. Forano and J. P. Besse, Inorg. Chem., 1998, 37, 4293–4301. 22 S. Miyata, Clays Clay Miner., 1983, 31(4), 305–311. 23 A. D. Becke, J. Chem. Phys., 1993, 98, 5648. 24 C. Lee, W. Yang and R. G. Parr, Phys. Rev., 1988, B37, 785. 25 R. Dovesi, V. R. Saunders, C. Roetti, R. Orlando, C. M. Zicovich-Wilson, F. Pascale, B. Civalleri, K. Doll, N. M. Harrison, I. J. Bush, Ph. D’Arco and M. Llunell, CRYSTAL06, User’s Manual, University of Turin, 2007. 26 N. M. Harrison and M. I. McCarthy, Phys. Rev. B, 1994, 49, 8574. 27 B. Montanari, B. Civalleri, C. M. Zicovich-Wilson and R. Dovesi, Int. J. Quantum Chem., 2006, 106, 1703–1714. 28 T. Homann, U. Hotje, M. Binnewies, A. Borger, K. D. Becker and T. Bredow, Solid State Sci., 2006, 8, 44–49. 29 M. Merawa, P. Labeguerie, P. Ugliengo, K. Doll and R. Dovesi, Chem. Phys. Lett., 2004, 387, 453. 30 M. J. Frisch, G. W. Trucks, H. B. Schlegel, P. M. W. Gill, B. G. Johnson, M. A. Robb, J. R. Cheeseman, T. Keith, G. A. Petersson, J. A. Montgomery, K. Raghavachari, M. A. Al-Laham, V. G. Zahrzewski, J. V. Ortiz, J. B. Foresman, J. Cioslowski, B. B. Stefanov, A. Nanayakkara, M. Challacombe, C. Y. Peng, P. Y. Ayala, W. Chen, M. W. Wong, J. L. Andres, F. S. Replogle, R. Gomperts, R. L. Martin, D. J. Defrees, J. Baker, J. P. Stewart, Head-M. Gordon, C. Gonzales and J. A. Pople, GAUSSIAN98 (Revision E.2), Gaussian, Inc., Pittsburgh, PA, 1995. 31 J. Rodriguez-Carvajal, PROGRAM FullProf.2000 Multi-Pattern Version 1.5, Laboratoire Le´on Brillouin (CEA-CNRS), France, 2000. 32 M. Meyn, K. Beneke and G. Lagaly, Inorg. Chem., 1990, 29, 5201. 33 Drits and V. Rives, Double Hydroxides: Present and Future, Nova Science Publishers Inc., New York, 2001. 34 S. Miyata, Clays Clay Miner., 1975, 23, 369–375. 35 S. Bonnet, C. Forano, A. De Roy, J.-P. Besse, P. Maillard and M. Momenteau, Chem. Mater., 1996, 8, 1962–1968. 36 H. Zhang, X. Wen and Y. Wang, J. Solid State Chem., 2007, 180, 1636–1647.

This journal is

c

the Owner Societies 2009

37 L. S. Dake, D. R. Baer and J. M. Zachara, Surf. Interface Anal., 1989, 14, 71–74. 38 N. C. Hosking, M. A. Stro¨m, P. H. Shipway and C. D. Rudd, Corros. Sci., 2007, 49, 3669–3695. 39 Handbook of X-Ray Photoelectron Spectroscopy, ed. C. D. Wagner, W. M. Riggs, L. E. Davis, J. F. Moulder and G. E. Muilenberg, Perkin-Elmer Corporation, Physical Electronics Division, 1979, pp. 56–57. 40 F. Wypych, W. H. Schreiner and R. Marangoni, J. Colloid Interface Sci., 2002, 253, 180–184. 41 J.-L. Guimaraes, R. Marangoni, L. Pereira Ramos and F. Wypych, J. Colloid Interface Sci., 2000, 227, 445–451. 42 R. A. Schoonheydt, T. Pinnavaia, G. Lagaly and N. Gangas, Pure Appl. Chem., 1999, 71(12), 2367–2371. 43 C. Mousty, Appl. Clay Sci., 2004, 27, 159–177. 44 A. Beres, I. Palinko, I. Kiricsi, J. B. Nagy Yoshimichi Kiyozumi and F. Mizukami, Appl. Catal., A, 1999, 182, 237–247. 45 M. R. Weir and R. A. Kydd, Microporous Mesoporous Mater., 1998, 20, 339–347. 46 B. Li, J. He, D. G. Evans and X. Duan, Appl. Clay Sci., 2004, 27, 199–207. 47 W. T. Reichle, Chemtech, 1986, 58–63. 48 W. Kagunya, R. Baddour-Hadjean, F. Kooli and W. Jones, Chem. Phys., 1998, 236(1–3), 225–234. 49 M. Lakraimi, A. Legrouri, A. Barroug, A. De Roy and J.-P. Besse, Mater. Res. Bull., 2006, 41(9), 1763–1774. 50 H. Zhang, X. Wen and Y. Wang, J. Solid State Chem., 2007, 180(5), 1636–1647. 51 T. Nishida and K. Kawakami, Solid State Commun., 1998, 107(9), 453–458. 52 M. Catti, G. Ferraris, S. Hull and A. Pevese, Phys. Chem. Miner., 1995, 24(2), 85–89. 53 P. Ugliengo, F. Pascale, M. Merawa, P. Labeguerie, S. Tosoni and R. Dovesi, J. Phys. Chem. B, 2004, 108(36), 13632–13637. 54 D. G. Costa, A. R. Rocha, W. F. Souza, S. Shirley, X. Chiaro and A. A. Leitao, J. Phys. Chem. C, 2008, 112, 10681–10687. 55 P. J. Squattrito, Acta Crystallogr., Sect. C, 1995, 51, 1282. 56 K. Siegbahn, C. Nordling, G. Johannson, J. Hedman, P. F. Heden, K. Hanu’in, U. Gelius, T. Bergmark, L. O. Wenne and Y. Baer, ESCA Applied to Free Molecules, North Holland, Amsterdam, 1969.

Phys. Chem. Chem. Phys., 2009, 11, 3554–3565 | 3565