J. MATER. CHEM., 1991, 1(5), 739-746 Porous Chromia-pillared a-Zirconium Phosphate Materials prepared via Colloid Methods Pedro Maireles-Torres," Pascual Olivera-Pastor,*" Enrique Rodriguez-Castellon," Antonio Jimenez- Lopez" and Anthony A. G. Tomlinson*b a Departamento de Quimica lnorganica, Cristalogra fia y Mineralogia, Universidad de Malaga, Apartado 59,29071 Malaga, Spain 1. T.S.E., Area delta Ricerca di Roma, C. N.R., Monterotondo Staz., 00016 Rome, Italy The reaction of Cr(CH3C0,), [Cr(OAc),] with colloidal n-propylammonium a-zirconium phosphate and subsequent calcination of the products have been investigated. Depending on [Cr(OAc),] :[initial phosphate] ratios and Cr3+ concentrations, a series of polyhydroxy acetato-Cr3+ intercalated precursor materials can be obtained, in which topotactic interface reactions have occurred to give materials with interlayer distances (doo2)ranging from 13.0 to 39.0A.These precursors show higher layer expansions than the analogous pillared clays (PILCS; doo,= 16.8-27.6A).A model invoking ordered in situ polymerisation of the Cr(OAc), on the phosphate surfaces is put forward. Calcination of these precursors under N, (400 "C) leads to a series of chromia-pillared materials in which the interlayers do not collapse to a single (much lower) value, as found previously for most PILCS, but instead provide a wide range of interlayer distances (10-27A).These correspond to free heights of 3.5-20.5k the widest ranging and highest yet found for such materials. These nanoscale oxide-pillared materials have N, surface areas (B.E.T., 77 K) of 250-330 m2 g-', with pore radii (cylindrical pore method) ranging from 8.5 to 13.8A, and very narrow pore-size distributions.Calcination conditions are crucial for obtaining porous solids. If calcination is carried out in air at 400 "C, although pillared powders and films are again obtained, surface areas are only ca. 40 m2 g-' (B.E.T., N,, 77 K). Furthermore, higher calcination temperatures (500 "C, under N,) give rise to X-ray amorphous materials, again having high surface areas and narrow pore-size distributions. All the materials can be processed in thin- film form without loss of textural characteristics. Keywords: a-Zirconium phosphate; Pillaring; Porosity Insertion of chemical moieties between the layers of swellable layered materials to give a fixed separation between the layers, and ultimately 'tuning' of the pores after thermal processing, is an attractive alternative method for inducing holes into materials' to the hydrothermal chemistry underlying zeolite synthesk2 It may prove of practical importance, given the limitations of (small-pore) zeolites3 for many types of applications (including, but not confined to: catalytic cracking').This strategy has to date been restricted to smectite clays, and especially use of the Keggin ion [A1,304(0H)12(0H2)24]7+(a very naive representation of the species present in the starting precursor solutions6) as a precursor for producing alumina pillar^.^ We have recently found that methods previously successful for preparing chrom- ia- and iron-oxide-pillared montmorillonite* (prior formation of polyhydroxy-Cr"' precursors or use of polynuclear com- plexes) instead give amorphous materials when applied to colloidal suspensions of the base material 'NMe,-SnP' (a-Sn[NMe4]o,9-l. 'H1, 1-o.9(P04)204H,0).~ Instead, when Cr(OAc), (OAc- =CH3C0,) reacts with NMe,-SnP under varying conditions (temperature, differing precursor :base material ratios, Cr3 + concentration) it is possible to obtain intercalate precursors having well defined polyhydroxyaceta- to-Cr3+ species inserted and highly expanded interlayers.On calcination in N,, chromia-pillared materials are obtained having a narrow pore-size distribution (>80% of pores with rp <20 A).These colloid manipulation methods have now been extended to a-zirconium phosphate, a base material exten- sively investigated in the past for ion-exchange purposes,'O membrane technology, and materials design strategies.' The objective is again to find new general methods for obtaining thermally stable, large-pore materials. Experimental Materials a-ZrP was prepared and stored by well established methods.12 n-Propylamine (n-PrA), Cr3 + acetate [Cr(OAc),] and all other chemicals used were best available analytical grade reagents. In initial experiments, NMez -exchanged a-ZrP phases were used. These proved unsuccessful, because considerable amounts of Me4N+ remained intercalated and could not be completely removed during the reaction.The use of n-PrA colloids instead provided pure phases of chromium oligomers on a-ZrP. The reasons for these differences lie in details of surface reactivity, which are being investigated. l3 Preparation of n-PrAH-a-ZrP Suspensions a-ZrP (5 g) was dispersed in water and a solution of 200 cm3 of 0.1 mol dm -n-PrA [60% cation-exchange capacity (c.e.c.) of a-ZrP] was slowly added under vigorous stirring for 2 h. The suspension was centrifuged and the solid resuspended in 1 dm3 of water. Gravimetric determination of the concen- tration of n-PrA-a-ZrP in this suspension gave a value of 4.84 g dm-3. The X-ray diffraction pattern (XRD) of a film prepared from this suspension had a 001 progression (rel.height in parentheses): 17.00A (100); 14.69 A (66); 10.64 A (21). There was no evidence for the presence of hkO reflections and the material appears to be the same as that reported by Alberti et all4 Adsorption Isotherm of Cr(OAc)3 on 60% n-PrAH-ZrP The above suspension (100 cm3) was contacted with increasing quantities of Cr(OAc), dissolved in 100 cm3 of H20: from 8.8 to 351 mequiv. Cr3+ added per gram of ZrP207, i.e. [Cr3+] varying from 0.005 to 0.194 mol dm-3. The effect of Cr3+ concentration was then studied using two-fold diluted solutions. In all cases, each suspension (green in colour) was refluxed for 4 days, cooled, centrifuged, and the dirty-green solids separated, washed well with water (first washing green, second washing colourless), air-dried, and analysed (C, H, N, TG, Cr analysis).Chemical and Physical Measurements XRD, on both powders and cast films (which usually gave narrower 001 reflexions), were recorded on a Siemens D501 diffractometer (graphite monochromator, Cu-Ka radiation). TG and DTA were measured on a Rigaku Thermoflex instru- ment (calcined Al,03 reference, 10 "C min-' heating rate). +Cr3 was analysed colorimetrically using the chromate method (A=372 nm) on alkaline solutions, after treatment with NaOH-H,02. Electronic spectra were registered on a Shimadzu MPC 3 100 spectrophotometer (BaSO, reference) and IR spectra on a Perkin-Elmer 883 spectrometer as KBr disks. Adsorption-desorption of N2 was measured on a conventional volumetric apparatus (77 K, degassing at 200 "C and lo-, mbar overnight).Results and Discussion Inspection of the uptake curve for Cr(OAc), on n-PrAH-ZrP (Fig. 1) immediately shows that uptake does not occur in a simple fashion, but involves complex Cr3 + species distributed between solution and solution/solid interface. Within a narrow range of [Cr3+] :[phosphate] ratios, the material takes up large amounts of Cr3+. At higher dilution (curve B) this range is extended, but in both cases addition of very high amounts of Cr(OAc), leads to materials with much lower Cr3+ loadings. As for the a-tin phosphate analogue, we ascribe this to the setting up of solution equilibria between small and large oligomers, which at equilibrium eventually give rise to higher oligomers no longer capable of inserting into the matrix through those already attached." The analysis of the materials separated at the points of J.MATER. CHEM., 1991, VOL. 1 50 100 150 200 250 300 mequiv. Cr3+ added/g ZrP,O, Fig. 1 Uptake of Cr(OAc), by colloidal cr-nPrAH-Zr(HPO,), *H20; each batch refluxed for 4 days. Conditions as detailed in Table 1 Table 1 Experimental conditions for the uptake of polyhydroxyaceta-to-Cr3+ by nPrA-a-ZrP (as in Fig. 1) sample mequiv. Cr3+ added/g ZrP207 mequiv. Cr3+ taken up/g ZrP207 [Cr3 +Ira 8.78 8.20 0.005 26.34 24.60 0.015 43.89 38.22 0.024 87.78 36.66 0.049 181.27 35.71 0.100 351.13 20.86 0.194 43.89 27.80 0.012 87.78 39.47 0.024 181.27 4 1.43 0.050 351.13 2 1.09 0.097 " Total molar concentration. curve A is shown in Table2.(Note that very small amounts of starting n-PrA were still present in several materials pre- pared from more concentrated solutions; this does not signifi- cantly change the discussion below.) The empirical formulations are also given in Table 2. They are based on the assumption that six-co-ordinate geometry is present for the Table 2 Analyses of the materials of Table 1 material empirical formulation doo2lA c ("/.I N (Yo) H (Yo) 13.2 1.17 0.3 1 2.39 19.1 1.72 0.10 2.28 32.0 39.3 27.3 20.0 3.48 4.25 4.4 1 4.38 0.10 0.15 0.19 0.20 2.45 2.47 2.50 2.5 1 32.9 4.38 0.02 2.37 33.7 36.6 20.6 2.15 3.24 3.82 0.02 0.02 0.02 2.28 2.42 2.63 wt.loss at 200 "C(Toy total H20 (YO) Cr3+:OAc- ratio exothermic effect/ "C" 9.7 10.4 11.9 19.6 10.7 13.4 13.1 28.1 28.8 15.1 14.3 13.4 11.5 19.9 19.5 18.4 17.2 27.9 27.8 18.0 -3.0 2.0 2.5 3.0 1.5 3.0 2.0 2.0 1.5 388 318 280 283 290 335 275 275 273 305 (I From TG/DTA. J. MATER. CHEM., 1991, VOL. 1 Cr3+ throughout and that the oligomers are mixed polyhyd- roxy-acetato Cr3+ moieties (apart from material 1, there was no evidence from ion-exchange experiments that OAc- was ever present as free anion or as an insalation compound). Local probes of structure (optical spectra, EPR, EXAFS, etc.) can give only indirect evidence to support the presence of trimeric, tetrameric, and pentameric Cr3 + species. Neverthe- less, we note that the 4T2g+4A2g(vl) and 4Tl,+4A2,(v2) transitions in chromium hydroxides move to higher energy with polymerisation degree,16 as found here. Secondly, much work on hydroxy-Cr3 oligomers in solution has suggested + the criterion E 1/~21.17- 1.20 (monomer, dimer), 1.6 (trimer), 1.95 (tetramer) and 1.5-1.6 (pentamer and hexamer), i.e.this intensity ratio increases as the connectivity of the chromium cluster increase^.'^ Applying this criterion, the data of Table 3 and Fig. 2 point to: (i) the presence of single species (bands are not broadened), and (ii) formation of extended structures, rather than the compact closed structures believed to exist in solution (but see ref. 18). The IR spectra are relatively well defined for this type of material (see Fig.3) and all show a difference v,,~~~(CO;)-~,~,,(CO;) in the range 91-101 cm-', which is characteristic of bidentate OAc- groups." Furthermore, the v3(P04)region is shifted by ca. 50cm-' with respect to the starting a-ZrP, is unusually distinct, and split, with two bands (at 1137 and 1009 cm-') rather than a single very broad one. This is evidence that the oligomers are attached to the pendant P-OH groups. Unfortunately, there is ambi- guity between a symmetry-lowering of the v3 vibration and the presence of Cr-0-Cr bridges (Cr-0 is expected to have a strong band at 1100-1200 cm-' with h(Cr-O-Cr) vibrations in a wide range, 450-950 cm-1).20 The XRD also favour the presence of well defined oligomers within the layers; all are characteristic of highly crystalline materials, although only a single 001 reflexion is observed, i.e.there is no interstratification or phase segregation (see Fig. 4). This contrasts with the much less defined XRD found in polyhydroxy precursors with smectite clays.I6 These highly expanded precursors re-acquire water after dehydration, as expected for zeolite-like materials. More important is the fact that, when all the zeolitic water is Table 3 Electronic spectra (diffuse reflectance) material v 1/nm v2b other/nm AIIA2 precursors -599 429 1.24 594 420 259 1.02 594 426 263 1.10 59 1 425 -1.10 587 41 1 262 0.96 583 417 259 1.08 597 428 254 1.27 590 425 -1.09 590 426 259 1.10 585 415 256 1.06 calcined at 400 "C N, 61 lsh 454sh 357sh, 323sh, 273 6 17sh 454sh 349, 272 623sh 453sh 351sh, 272 620sh 456sh 366, 320sh, 270 619 450sh 346sh, 273 61 lsh 456sh 332,277 calcined at 400 "C air 61 Ish, 577 454 366, 322sh, 273 598 460 320, 274 617sh, 580 45 Ish 309,272 60 1 453 330sh, 274 741 A '\ B IlIIII1 1IIIIlIII IIIIIIIIIIIIII 400 600 800 Ilnrn Fig.2Optical spectra (as diffuse reflectance) of selected materials. A, Precursors; B, materials calcined at 400 "C under N2; C, materials calcined at 400°C in air (Cr,03 shown for comparison). Numbers refer to materials of Fig. 1 J. MATER. CHEM., 1991, VOL. 1 hi 1'37\ IV 1009 #. .*I,?. 700 1400 1100 , 800 500 2 0 v/cm -' Fig. 3 IR spectra: (a) material 3, as prepared; (b) material 3, 400 "C 24 h N,; (c) Cr(OAc), removed (at 200 "C) they do not collapse to give materials with a restricted range of interlayer distances, but instead provide a range spreading from 9.7 to 35 A (see Fig.5). This behaviour is different from that in the a-tin phosphate ana- logues, which even at this low calcination temperature do collapse considerably' (as do all the PILCS reported to date).16 The electronic and IR spectra do not undergo any marked changes at this stage. The TG/DTA of the materials are very similar to those of the tin phosphate analogues (see Fig. 6) and a similar decomposition scheme operates,' apart from the presence of a plateau at 350-420 "C in TG, which is much more defined than in the tin phosphate analogues.Nevertheless, all water molecules are removed only at much higher temperatures (>600 "C). A further point of interest is that the temperature at which the acetate ion is decomposed decreases more or less linearly as the interlayer distance increases. The difference is large: 113 "C between material 1 and material 3 (Table 2), and it directly reflects the ease with which decomposition products leave the solids. Note also that there is no clear evidence for the presence of a-ZrP,O, phase (as confirmed by XRD) expected at 450-600 "C, constant weight being achieved at 950 "C. As before,' we take this to be indirect evidence that cross-layer condensation between -P-OH groups is not possible because of the pillaring (condensation between adjacent OHs in the layers may also contribute to the TG plateau at 420 "C).A possible mechanism can now be put forward, assuming that as for hexaaquo-Cr3+ ion, Cr3 + acetate forms hydroxo- bridged polymers in solution and polymerisation degree 39.3 Fig. 4 XRD powder patterns of material 4 (thermal treatment under N2) 40 30 * (N*) 5 8 20 t? 10 I I 100 200 300 Tl"C Fig. 5 Changes in interlayer distances on calcination in air. The numbers refer to the materials separated along curve A of Fig. 1. Ir (N,): material 4 calcined under N, increases with Cr3+ concentration. This is shown as (i) in Scheme 1, and addition of more Cr3+ (region 3-4 of Fig. 1) increases oligomer nuclearity and charge, with simultaneous buffering action of AOc-.At low amounts of added Cr3+, competition between oligomers in solution is pushed towards low-nuclearity clusters by electrostatic interaction with the J. MATER. CHEM., 1991, VOL. 1 + high -polymersH20 \ 461 + n -PrAH OAc higher polymer -2 H20 intercalates (iii) Scheme 1 Mechanism (schematic) of the forced polymerisation of Cr(OAc), on colloidal surfaces of a-nPrA-a-ZrP -283 "C0 10 EX0 h s 4 vr 20 DTA u) c -0 v 3c V ENDO 85 4c r.t. 250 500 750 T/"C Fig. 6 TG/DTA of material 4 TC phosphate layer. The higher negative charge on the layer now accommodates the higher positive charge of the oligomer (ii). After attachment of the first, trinuclear, oligomers, olation in the usual manner for hydroxy-bridged Cr3 + complexes then occurs (iii).OAc -must be involved in this oligomerisation stage, and in such a way as to preclude formation of high- connectivity cluster^,'^ as shown. As even more Cr3+ is added, both Cr3+ and negative charge on the phosphate layer increase. An effective way for such a highly charged species to neutralise a high-charge layer is by tilting. This well known 'covering' effect becomes more important as polymerisation proceeds, leaving progressively less space for other species. This nicely rationalises both the uptake curve, the decrease in the total amount of oligomer and the drop in dOo2between materials 4 and 5. The evidence thus points to different orderings of the precursors in a-tin phosphate and a-zirconium phosphate, the former probably containing monolayers and considerable amounts of water on re-forming from the colloidal solution (as suggested previously'), and the latter a bilayer (Scheme 2).Furthermore, we suggest that oligomer chain-endings are different in the two cases, with specific binding possible in the a-zirconium phosphate precursors. Calcination and Formation of Oxidic Pillars Calcination of the precursors with complete removal of organics and formation of chromia pillars supports the above suggestions. Calcining in air at 400 "C in all cases gave pillared products with only low B.E.T. surface areas: ca. 40m2 g-' (after thorough washing). (This is as found for chromia- pillared smectites,8 but different from the a-tin phosphate analogue^.^) Calcination under N2 at 400 "C instead leads to a series of materials (with still recognisable basal spacings in the XRD ranging between 9.5 and 26.9 A) and surface areas in the range 250-330 m2 g-'.The basal spacings of all mater- ials did not change after treatment with 0.2 mol dm-3 HCl or CuCl, (in other words, there are no extractable Cr3+ species remaining between the layers). The most highly pillared are those derived from precursor materials 3 and 4 of the uptake isotherm, i.e. those with the highest Cr3+ content (in these cases, broad dOo2peaks were observed at low angle). The reflectance spectra of the final porous materials show clear differences from that of Cr203 itself, the 4T2, band lying J.MATER. CHEM., 1991, VOL. 1 Table 4 Textural parameters for chromia-pillared a-zirconium phosphates" ' material"/calcination temp SB.E.T.lm2 g- CB.E.T./m2 g- Sb/m2 g-' V,"/cm g - 31400 "C 272 75 31 1 0.224 33 41400 "C 308 87 344 0.199 26 41500 "C 33 1 77 369 0.254 31 5(b)/400"C 321 102 379 0.378 47 'All samples calcined under nitrogen. Parameters as defined and calculated in Cranston and Inkley." m Scheme 2 Illustrating straight (i) and tilted (ii) orderings of oligomers in precursors 1 , , , . , , , , . 0.2 0.4 0.6 0.8 0.2 0.4 0.6 0.8 PIP0 at higher wavelengths (Table 3) by 20-30 nm. This is in agreement with an oxy-hydroxy formulation for the 'chromia' pillar, rather than simply an oxide one.Conversely, the non- porous materials obtained via calcination in air clearly involve mixed-oxidation species, including Cr2OS, and, possibly, Cr5 + (very-low-energy broad band in the near-IR region (Table 3). The pore radii of the porous series from the plateau of the uptake curve [i.e. materials 3-5, and 4(b) and 5(b)] were calculated from the adsorption branches of the adsorption- desorption curves of Fig. 7 using the cylindrical pore model.21 They are characteristic of orderly pillared materials, and the distributions are narrow (Fig. 8). The best derive from mater- ials 3 and 4, in which ca. 75% of the pores have fp<15 A) that from material 5(b)has only ca. 55% of pores <15 A.We underline the fact that these materials are predominantly mesoporous with a microporous contribution (i.e. the iso- therms are Type IV isotherms in the BDDT classification),22 and that the cylindrical pore method has limitations (in particular, it is unable to provide a reliable estimate of srnall- pore contributions less than the Kelvin limit of 8 It is therefore not obvious just what pore is being accessed by N, (vertical or horizontal to the layers?) nor how this is correlated with the free height deduced from the XRD. Despite this uncertainty, the pore-size distributions represent some of the most homogeneous yet obtained for a pillared material.". 24 This supports the presence of a true pillaring model rather than one in which porosity derives from 'card-house' or 'book- house' models based on face-to-edge associations, both of 20 3.5 15 7 10 0) r I 45 0E z X (c1 h 1.3Gn20--.rd, 15 10 5 \ \ ,ALI-20 40 60 80 20 40 60 80 FpIAFig. 7 Adsorption-desorption isotherms for materials calcined under N, (77 K, N,) on: (a) material 3 400 "C; (b) material 4 400 "C; (c) Fig.8 Pore-size distributions of the 'chromia'-pillared materials of material 4 500 "C; (d) material 5(b)400 "C Fig. 7; same legend J. MATER. CHEM., 1991, VOL. 1 Fig. 9 SEM micrographs of material 5, (a) before and (b) after calcination in N, at 400 “C which are expected to give rise to a scattered pore distri- b~tion).~~Previously, we invoked the ‘book-house’ model to rationalise the wide distribution of pores found in alumina- pillared tin phosphates prepared by ex situ methods.26 As regards gross morphology of materials, SEM micrographs show that after calcination straight and pleated layers are still visible (Fig.9). On calcining at 500 “C under N2, amorphous chromia- pillared zirconium phosphates are obtained, with virtually the same pore size and pore-size distribution as the crystalline form (see the example of material 4 in Fig. 8). This poses the problem of whether XRD alone is a good measure of porosity induced by pillaring. SEM and SAXS investigations have shown that smectite-water, kaolinite-water and sepiolite- water systems give different particle arrangements depending Inon particle fle~ibility.~~ a similar vein, the amorphous nature of the porous ‘chromia’ zirconium phosphates may be rationalised by assuming that it arises from end-to-end sin- tered particles which are so small (<50 A) as not to preclude long-range order yet give only ill defined (amorphous) XRD patterns (Scheme 3).We suggest that these most expanded phases are close to a limit of layer expansion, which renders them more likely to be amorphous. Or, put crudely, the calcination step probably causes gross rearrangements within the pillar, which break up the layers. Although superficially similar, the materials reported here Scheme 3 ‘Break-up’ of layer structure on calcination >400 “C, giving amorphous, but still porous, materials are in reality very different from those obtained in the reflux reactions of Cr(OAc)3 with CX-Z~N~H(PO~)~ 5H20, which are non-porous with stuffed layers, and possess no clear precursor phases.28 Although also believed to involve in situ partially hydrolysed chromium oligomers, the hydrolysis occurs in the restricted space (interlayer distance 11.8 A) of NaH-ZrP 5H20particles.Ordered hydrolysis of the chromium acetate with diffusion of NaOAc out of the matrix is not possible and interlayer fragments are formed. Conclusions The forced hydrolysis of metal complexes on the surfaces of colloidally dispersed layered phosphates can be used success- fully to obtain porous solids, the final products obtained depending critically on the details of surface structure and pH.Further kinetic and structural details are being investi- gated and SAXS experiments planned on these novel, tuned- pore solids. We thank the EEC (European Project 1-0027) and CICYT (Spain, Project No. MAT90-0298) for financial support. References 1 R. M. Barrer, Pure Appl. Chem., 1989, 61, 1903; T. J. Pinnavaia, M. S. Tzou, S. D. Landau and R. H. Raythatha, J. Mol. Catal., 1984, 27, 195. 2 A. Dyer, Introduction to Zeolite Molecular Sieves, Wiley, Chichester, 1988; R. M. Barrer, The Hydrothermal Synthesis of Zeolites, Academic Press, London, 1986. 3 J. M. Newsam, Science, 1986,231, 1093; J. V. Smith, Chem. Rev., 1988, 88, 149; M. Szostak, Molecular Sieves, Academic Press, London, 1988. 4 G.A. Ozin, A. Kugerman and A. Stein, Angew. Chem. Int. Ed. Engl., 1989, 28, 359; P. A. Diddams, J. M. Thomas, W. Jones, J. A. Ballantine and J. H. Purnell, J. Chem. SOC.,Chem. Commun., 1984, 1340. 5 D. E. W. Vaughan, P. K. Maher and E. W. Albers, U.S.Pat., 3 775 345, 1973; M. Occelli, Ind. Eng. Chem. Prod. Res. Dev., 1983, 22, 553. 6 J. W. Akitt, J. M. Elders, X.L. R. Fontaine and A. K. Kindu, J. Chem. Soc., Dalton Trans., 1989, 1897, and refs. therein: A. Clearfield, in Surface Organometallic Chemistry: Molecular Approaches to Surface Catalysis, ed. J. M. Bassett, Kluwer, Amsterdam, 1988, p. 271. 7 D. E. W. Vaughan and R. J. Lussier, Proc. 5th Int. Con$ Zeolites, ed. L. V. C. Rees, Heyden, London, 1980, p. 94; G. W. Brindley and R. E. Sempels, Clay Minerals, 1977, 12, 229.8 M. S. Tzou and T. J. Pinnavaia, in Catalysis Today, ed. R. Burch, Elsevier, London, 1988, p. 243; D. H. Doff, N. H. J. Gangas, J. E. M. Alan and J. M. D. Coey, Clay Minerals, 1988, 23, 367. 9 P. Maireles-Torres, P. Olivera-Pastor, E. Rodriguez-Castellon, A. Jimenez-Lopez, A. A. G. Tomlinson, in Pillared Layered Struc- tures. Current Trends and Applications, ed. I. V. Mitchell, Elsevier Applied Science, London, 1990, p. 137; J. Solid State Chem., 1991, in the press. 10 Inorganic Exchange Materials, ed. A. Clearfield, CRC Press, Boca Raton, 1982, ch. 1; G. Alberti, in Recent Developments in Ion Exchange, ed. P. A. Williams and M. J. Hudson, Elsevier Applied Science, London, 1987, p. 233. 11 A. Clearfield, in Design of New Materials, ed.D. L. Cocke and A. Clearfield, Plenum Press, New York, 1987, p. 121; G. Alberti, U. Costantino, F. Marmottini, R. Vivani and P. Zappelli, in Pil-lared Layered Structures. Current Trends and Applications, ed. I. V. Mitchell, Elsevier Applied Science, London, 1990, p. 119. 12 G. Alberti and E. Torracca, J. Znorg. Nucl. Chem., 1968, 30,317. 13 P. Maireles-Torres, P. Olivera-Pastor, E. Rodriguez-Castellon, A. Jimenez-Lopez and A. A. G. Tomlinson, in preparation. 14 G. Alberti, A. Casciola and U. Costantino, J. Colloid Interface Sci., 1985, 107, 256. 15 C.H. Giles, T.H. McEwan, S.N. Nakhwa and D.Smith, J. Chem. SOC., 1960, 3973. 16 T. J. Pinnavaia, M. S. Tzou and S.O. Landau, J. Am. Chem. Soc., 1985, 107, 4783.746 J. MATER. CHEM., 1991, VOL. 1 17 H. Stunzi, F. P. Rotzinger and W. Marty, Inorg. Chem., 1984, 23, 2160. 25 T. J. Pinnavaia, in Chemical Physics of Intercalation, NATO, AS1 Ser. B, vol. 172, p. 233; H. van Damme, P. Levitz, J. J. Fripiat, 18 L. Monsted, 0. Monsted and J. Springborg, Inorg. Chem., 1985, J. F. Alcover, L. Gatineau, and F. Bergaya, in Physics of Finely 24, 3496. 19 K. Nakamoto, Infrared and Raman Spectra, Wiley, New York, 1986, p. 233. 26 Divided Matter, ed. N. Boccara and M. Daoud, Springer, 1985, p. 24. P. Maireles-Torres, P. Olivera-Pastor, E. Rodriguez-Castellon, 20 J. R. Ferraro, R. Driver, W. R. Walker and W. Wozniak, Inorg. A. Jimenez-Lopez, L. Alagna and A. A. G. Tomlinson, J. Chem. Chem., 1967,6, 1586. SOC., Chem. Commun., 1989, 741; J. Muter. Chem., 1991, 319. 21 R. W. Cranston and F. A. Inkley, Adv. Catal., 1957, 9, 143. 27 D. Tessier, G. Pedro and L. Camara, C.R. Acad. Sci. Paris, 1980, 22 S. J. Gregg and K. S. W. Sing, Adsorption, Surface Area and 2WD, 1169; C. H. Pons, F. Rousseaux and A. Tchoubar, Clay Porosity, Academic Press, Orlando, 2nd edn., 1982, p. 4. Minerals, 1981, 16, 23. 23 F. Rodriguez-Reinoso and A. Linares-Solano, in The Chemistry 28 D. J. MacLachlan and D. M. Bibby, J. Chem. SOC., Dalton Trans., and Physics of Carbon, Marcel Dekker, New York, Vol. 21; 1989, 895. K. S. W. Sing, Pure Appl. Chem., 1982, 54, 201. 24 J. Sterte, Clays Clay Minerals, 1986, 34,658. Paper 1 /00500F; Received 4th February, 1991