Characterisation of pillared clays by contrast-matching small-angle neutron scattering Th. A. Steriotis,* a K. L. Stefanopoulos, a U. Keiderling, b A. De Stefanis c and A. A. G. Tomlinson c a NCSR ‘DEMOKRITOS’, Institute of Physical Chemistry, Aghia Paraskevi Attikis, 153 10 Athens, Greece. E-mail: tster@chem.demokritos.gr; Fax: +30 010 6511766; Tel: +30 010 6503973 b Hahn Meitner Institut, BENSC, Glienicker Strasse 100, D-14109 Berlin, Germany. E-mail: keiderling@hmi.de; Fax: +49 030 80693059; Tel: +49 030 80692339 c Istituto per lo Studio dei Materiali Nanostrutturati, Area della Ricerca di Roma del CNR, C.P.10, 00016 Monterotondo Staz, (RM), Italy. E-mail: gus@mlib.cnr.it; Fax: +39 06 90672322; Tel: +39 06 90672322 Received (in Cambridge, UK) 29th July 2002, Accepted 30th August 2002 First published as an Advance Article on the web 19th September 2002 The contrast-matching SANS technique has been utilised to determine inter-pillar distances (and surface texture) in montmorillonite and beidellite pillared smectite clays; they lie in the range 1.40–1.80 nm, reflecting different inter-pillar orderings. Interest in pillared inter-layered clays (PILCs) has over the past decade been centred on their prospective industrial utilisation in catalysis (de-NOx), sorption and separations (air–gas mixtures, small hydrocarbons, multi-component hydrocarbon mixtures, large organic molecules). 1 The materials (which have a highly heterogeneous structure because they contain both nano-oxide pillars cross linked to aluminosilicate clay layers) are produced after intercalating large inorganic cations between smectite clay layers but their unique micro/meso pore system does not allow application of single crystal X-ray techniques. 2 However, their pore network structure is being unravelled with the aid of non- conventional micropore characterisation techniques 3 as well as X-ray absorption spectroscopy. 2,4 On the other hand, Small- Angle Neutron Scattering (SANS) is an excellent technique for characterising porous materials on a scale covering a range from 1 nm to > 200 nm. 5 The scattered intensity depends on the contrast in the neutron scattering length density of the different phases in the sample. The neutron scattering length density, r, of a molecule of i atoms can be readily calculated by the expression: (1) where b i is the scattering length of the individual atoms in the molecule, d is the bulk density of the scattering object, M w is its molecular weight and N A is the Avogadro number. Neutrons have the advantage over X-rays that their scattering length varies completely irregularly with the atomic number, even with isotopes of the same element. The fact that hydrogen and deuterium have scattering lengths of opposite sign means that, unlike X-rays, neutrons can not only ‘see’ hydrogen isotopes but they can also differentiate between them. It is then possible to match the scattering density of one phase in a multiphase material with an appropriate mixture of H 2 O–D 2 O. As a result, contrast matching is achieved and the structural characterisation of the unmatched phases is feasible. 6 We have employed contrast-matching small-angle neutron scattering (SANS) to independently resolve the structure of each phase in alumina-pillared and Mg/alumina-pillared sam- ples. Using this technique, pores are filled with appropriate ratios of H 2 O–D 2 O that match the neutron scattering length density of each phase (clay, pillars) such that scattering from samples with filled pores results only from the non-contrast matched phase. The PILCs studied were: the parent clay, EFW (Extra Fine White) a montmorillonite kindly provided by IKO-Erbslöh (Germany) and B4, a Greek beidellite provided by Silver & Baryte Ores Mining CO. SA. Additionally, pillared Al 2 O 3 EFW (Al-EFW) and Mg/Al oxide pillared EFW (MgAl-EFW) as well as Al 2 O 3 pillared B4 (Al-B4) and its Fe/Al oxide pillared analogue (FA-B4) were used. The Al 2 O 3 PILCs were prepared as reported previously 7 and MgAl-EFW (a new material) by the usual procedure of adding an aqueous solution of poly- hydroxyoxyaluminium Keggin ion (15 meq Al and 1 meq Mg per g clay) to a colloidal dispersion of EFW. The intercalated precursor was allowed to flocculate, decanted, dialysed to remove Cl 2 and freeze dried. The solid recovered was calcined at 450 °C under flowing N 2 to bring about cross-linking. SANS measurements were carried out on V4 instrument at BENSC, Hahn-Meitner Institut, Berlin. A neutron wavelength of 0.457 nm and three sample-detector distances were used (1, 4 and 16 m) in order to cover a Q-range between 4.77 and 6.83 nm 21 . Scattering curves obtained from (i) EFW, B4, Al-EFW, Al-B4, MgAl-EFW and FA-B4 dry samples, (ii) Al-EFW and MgAl-EFW samples soaked in two different H 2 O–D 2 O mix- tures, which contrast match either the clay layers or the pillars. The raw data were further corrected for background and empty cell scattering and converted to cross section units after calibration with water by using the BerSANS software developed at BENSC. 8 The left part of the figure shows the SANS spectra from dry and contrast matched pillared Al-EFW and MgAl-EFW sam- ples. Within the region of high Q values, the characteristic feature for all the pillared samples is the appearance of well- resolved peaks centred at approximately the same Q position and corresponding to real space lengths between 1.40–1.80 nm (Table 1). Further, when the clay aluminosilicate layers are matched structural information of the pillars is obtained (p samples), revealing that the peaks are still present in both pillared samples that of the MgAl-EFW being more pronounced (inset Fig. 1 left). However, on contrast matching the pillars (samples c) the peaks completely vanish suggesting that the corresponding spacing is characteristic of the distance between pillars. XRD analysis of EFW itself reveals a (001) reflection corresponding to a basal spacing of 1.27 nm, characteristic of the distance Table 1 Peak position, real space length and surface fractal dimensions Sample Q peak (1/nm) w (nm) D s EFW — — 2.60 Al-EFW 3.59 1.75 2.98 MA-EFW 3.57 1.76 2.99 Al-EFW_p Weak — 2.93 MA-EFW_p 3.57 1.76 2.98 Al-EFW_c — — 2.53 MA-EFW_c — — 2.56 B4 — — 2.60 Al-B4 4.37 1.44 2.74 FA-B4 3.50 1.80 3.09 4.22 1.49 This journal is © The Royal Society of Chemistry 2002 2396 CHEM. COMMUN. , 2002, 2396–2397 DOI: 10.1039/b207392g