Porous Pillared Clays and Layered Phosphates

Porous Pillared Clays and Layered Phosphates

5.08 Porous Pillared Clays and Layered Phosphates A Clearfield, HP Perry, and KJ Gagnon, Texas A&M University, College Station, TX, USA ã 2013 Elsev...

6MB Sizes 1 Downloads 44 Views

5.08

Porous Pillared Clays and Layered Phosphates

A Clearfield, HP Perry, and KJ Gagnon, Texas A&M University, College Station, TX, USA ã 2013 Elsevier Ltd. All rights reserved.

5.08.1 Clay Minerals 5.08.1.1 Introduction to Clay Minerals 5.08.1.1.1 Surface properties of clay minerals 5.08.1.1.2 Surface acidity 5.08.1.2 Pillared Interalyer Clays: Pillar Chemistry 5.08.1.2.1 PILCs: Synthesis 5.08.1.3 Pore Structure and Porosity 5.08.1.4 Structure of the Aluminum PILCs 5.08.1.5 Thermal Stability 5.08.1.6 Mixed Cation PILCs 5.08.1.6.1 Gallium substitutions 5.08.1.6.2 Lanthanide substitutions 5.08.1.6.3 Silicon–aluminum system 5.08.1.7 Zirconium-Pillared Clays 5.08.1.8 Synthesis of Porous Clay Heterostructures 5.08.1.9 Pillaring of Vermiculites and Micas 5.08.1.10 Adsorption and Ion-Exchange Properties of PILCs 5.08.1.11 Chromia and Titania PILCs: Additional Porosity Considerations 5.08.1.12 Theories of N2 Sorption–Desorption Isotherm Analyses 5.08.1.13 Acidic Properties of PCH 5.08.1.14 PILCs as Catalysts 5.08.1.15 Anionic Clays 5.08.1.16 Conclusion 5.08.2 Layered M(IV) Phosphates and Phosphonates 5.08.2.1 In the Beginning 5.08.2.1.1 General structure and properties of a- and g-M(IV) phosphates 5.08.2.2 Preparation and Characterization of M(IV) Phosphates 5.08.2.2.1 Exfoliation of M(IV) phosphates 5.08.2.3 Porous Phosphates and Phosphonates Synthesized by Templating and Self-Assembly 5.08.2.3.1 Nonionic templates 5.08.2.4 Pillaring of M(IV) Phosphates by Inorganic Oligomeric Cations 5.08.2.4.1 Transition metal oxides 5.08.2.4.2 Alumina and related inorganic clusters 5.08.2.4.3 Silica and other mixed oxides 5.08.2.4.4 Conclusions and general outlook for inorganically pillared phosphates 5.08.2.5 Pillaring of g-M(IV) Phosphates by Topotactic Exchange 5.08.2.5.1 Structure of topotactically exchanged g-forms of Zr and Ti phosphate 5.08.2.5.2 Examples of pillaring g-M(IV) phosphates by bisphosphonic acids 5.08.2.5.3 Pillared derivatives of g-ZrP with rigid and flexible bisphosphonic acids 5.08.2.6 Direct Pillaring Using Phosphonic Acids with a-M(IV) Phosphates, UMOFs 5.08.2.6.1 M(IV) bisphosphonates 5.08.2.6.2 Mixed ligand-pillared materials 5.08.2.6.3 Functionalization of UMOFs 5.08.2.6.4 Selective ion exchange for nuclear waste streams 5.08.2.6.5 Future outlook for organically pillared phosphonates, UMOFs Acknowledgments References

170 170 170 171 171 172 173 174 175 175 175 175 176 176 178 180 180 182 183 185 186 188 188 190 190 191 192 193 193 194 196 197 197 199 200 200 200 201 201 202 203 203 205 206 207 207 207

Comprehensive Inorganic Chemistry II

169

http://dx.doi.org/10.1016/B978-0-08-097774-4.00509-X

170

Porous Pillared Clays and Layered Phosphates

5.08.1 5.08.1.1

Clay Minerals Introduction to Clay Minerals

Clay minerals are layered phyllosilicates in which octahedral sheets are bonded to tetrahedral sheets in either a 1:1 layer or a 2:1 layer. The different types of clay minerals are listed in Table 1. Because the smectites have a low positive charge, the layers are easily spread apart or exfoliated allowing for ease of pillaring. Vermiculites can also be spread apart but with more difficulty. Micas are usually in the Kþ form and very difficult, if not impossible, to pillar. The smectites have structures that consist of aluminum or magnesium–oxygen octahedra sandwiched between layers of silica tetrahedra as shown in Figure 1. In a unit cell formed from 20 oxygen atoms and four hydroxyl groups, there are eight tetrahedral sites and six octahedral sites. The composition of the principal smectite clays is given in Table 1. If no Liþ were present in the hectorite, that is, x ¼ 0, the formula would be a magnesium talc of formula Mg6Si8O20(OH)4 with no imbalance of charge. The smectites are distinguished by the type and location of their cation substitutions.1 Substitution may occur in either the octahedral or tetrahedral layers. In some clay minerals, both types of substitutions occur. Typically the charge deficiency ranges from 0.4 to 1.2 eq per Si8O20 unit (Table 1) and is balanced by exchangeable ions held between the layers. These values translate to ion-exchange capacities (IECs) of 0.5–1.5 meq g1. The different clay minerals are distinguished by the extent of isomorphous substitution of lower-valent ions for the Mg2þ, Al3þ, and Si4þ as shown by the increasing charge in Table 1. If the charge deficiency arises from octahedral substitution, then the excess charge is distributed over all the oxygen atoms in the framework. These clays are said to be turbostratic, that is, the layers are randomly rotated about an axis perpendicular to the layers. Substitution in the tetrahedral layers provides a more localized charge distribution and, as a consequence, the structure tends to exhibit greater three-dimensional (3D) order.2 IECs for the remainder of the clays are provided in Table 1. The low-layer charge of smectites allows them to swell through hydration of the interlayer cations. Swelling may be carried to exfoliation by large water additions and sonication. The vermiculites also swell, but not to exfoliation. Micas for which Kþ is the interlayer cation, which is the case for most micas, do not swell or hydrate at all. The potassium ion is held so tightly in the hexagonal cavities of the layers that it cannot Table 1

be removed except by special means. We have demonstrated that in certain micas the Kþ may be removed by successive treatment with strong NaCl solutions.3 Also, some synthetic micas that can swell in water are available commercially.4 All of the clays that can be swollen – smectites, vermiculites, sodium micas, and synthetic micas – may be pillared by polymeric cations to obtain porous pillared clays. An excellent introduction to clay minerals which serves as a good preparation for the present discourse is available.5

5.08.1.1.1 Surface properties of clay minerals The particles of many clay minerals, especially smectites, are in the low micron range. The thickness of a single layer in the structure is about 0.9–1.0 nm. Based on the size of these particles, the internal surface is estimated to be about 750– 800 m2 g1.6 This estimate assumes that the clay is completely exfoliated. For smectites the exterior surface area is less than 20% of the total. However, by opening the layers very high surface areas are accessible. Vermiculites have about the same internal surface area as smectites but the external area is only about 10 m2 g1. This difference is the result of the tighter bonding between the layers and interlayer cations because of the higher interlayer cation content. Surface areas are best measured by N2 sorption–desorption isotherms by the Brunauer–Emmett–Teller (BET) method. Smectites may have surface areas of 30–100 m2 g1 depending upon the degree of crystallinity and structural defects of the layers. The outgassing conditions need to be controlled to make sure that all the water is removed from the sample. Much of the early data were determined by a three-point or six-point isotherm which may contain considerable errors.4 More complete isotherms are given by Rutherford et al.7 who found that the layer dimensions had an effect on the porosity and surface area. A very large effect results from the type of ion and especially tetramethylammonium ion (TMA). Large ions such as Csþ cannot fit into the hexagon-shaped openings of the silicate layers (Figure 1). Thus the layers are further apart and some of the internal surface is registered as part of the total surface area.7 Furthermore, insertion of a large molecule such as TMA between the layers allows access of N2 between the layers. An isotherm for TMA-montmorillonite is shown in Figure 2.5 The steep rise at P/P0 ¼ 0 is the filling of the small pores (i.e., below 10 A˚). The desorption curve that is slightly above the physisorption curve is termed adsorption hysteresis and almost always present in clay N2 isotherms and is generally due to capillary condensation in mesopores.8

Useful information about select clay minerals

Group

Charge

IEC (meq g1)

Clay

Idealized structural formula

Kaolin Smectite

0 0.4–1.2

0 0.5–1.5

Kaolinite Montmorillonite Hectorite Beidellite Saponite

Vermiculite Mica

1.2–1.8 2

1.5 2–2.5

Brittle Mica

4

4–4.5

Al2Si2O5(OH)4 MX/NNþ[Al4 – xMgx](Si8)O20(OH)4nH2O MX/NNþ[Mg6 – xLix](Si8)O20(OH)4nH2O MX/NNþ[Al4](Si8 – xAlx)O20(OH)4nH2O MX/NNþ[Mg6](Si8 – xAlx)O20(OH)4nH2O Mg0.45(Mg2.6Fe3þ0.24Fe2þ0.04Ti0.1)(Si2.72Al1.28)O10(OH)2 KMg3(Si3Al)O10(OH)2 K(Mg2Fe)(Si3Al)O10(OH)2 Ca(Mg2Al)(SiAl3)O10(OH)2

Phlogopite Biotite Clintonite

Porous Pillared Clays and Layered Phosphates

171

120 TMA 100

cm3 gm-1, N2

80

60

40

[M(H2O)x]n+

20

0 0

0.2

0.4

0.6

0.8

1

P/Po

Figure 1 Schematic drawing of a smectite clay.

Figure 2 A N2 sorption–desorption isotherm of a TMA-montmorillonite clay. Reproduced from Carrado, K. A., In Handbook of Layered Materials, Auerbach, S. M.; Carrado, K. A.; Dutta, P. K., Eds. M. Dekker: New York, 2004; pp 1–37, with permission.

5.08.1.1.2 Surface acidity There are several types of acidities inherent in clays and especially smectites. The interlayer cations are generally hydrated which impart some Brønsted acidity to the coordinated water. Hydroxyl groups exist at the edges of the layers and some hydronium ions may be present. Aluminum, magnesium, or transition elements that may exist in the clay layer may be coordinatively unsaturated, thereby creating Lewis acid sites.9 Because of this acidity, clays have been used as industrial catalysts for many years.9 In order to increase the porosity of the clays, treatment with H2SO4 can induce the formation of pores in the structures. This treatment exposes the central layer and access to the different cations in that layer. The Si–O(H)–Si groups can occasionally have substitution of Al for Si. When this occurs, strong Brønsted acid groups can be generated, which on drying can become Lewis acid sites. All of these features result in high catalytic activity for altered clay catalysts.

5.08.1.2

Pillared Interalyer Clays: Pillar Chemistry

The driving force for the preparation of pillared interlayered clays (PILCs) was the desire to synthesize thermally stable porous materials with large pores. For this purpose, inorganic polymeric species, with high positive charges, were chosen to exchange between the clay layers.9 The high charge was expected to place the pillars relatively far apart to obtain the desired large pores. Clays and layered silicates have a long history of use as petroleum cracking catalysts. As mentioned previously, they have been generally treated with sulfuric acid to create higher surface areas, which is not a highly controlled procedure.10 Zeolites are commonly used as fluid cracking catalysts, but they have relatively small pore openings (8 A˚).11 Thus, there is significant interest in obtaining similar

acidic catalysts with larger pores to enable cracking of the petroleum hydrocarbons. It was thought that pillaring clays with large inorganic polymers such as the aluminum Keggin ion, Al13O4(OH)24(H2O)12, with a 7þ charge would be an ideal pillar. Placed between the clay layers, followed by calcinations, they should form oxide pillars fused to the layers with large spaces between pillars. McGee and Vaughan initiated such studies 12,13 as well as Shabtai et al.14 These early alumina PILCs had surface areas of 200–300 m2 g1 and a range of pore sizes. The bulk of the pores, 60%, were below 14 A˚ and 14% were in the range of 14–20 A˚. The remainder were mesopores, which are greater than 20 A˚. Details concerning synthesis, properties, and catalysis were provided in several review articles.8,15–18 Explanation of N2 sorption isotherms is provided in Section 5.08.1.12. Aluminum pillaring has been carried out with two types of solutions. The first is an aluminum chloride solution to which was added up to 2.33 mol of NaOH per mol of Al. The other solution, known as chlorohydrol, consists of an aluminum chloride solution reacted with Al(OH)3 or Al metal to the point that the OH:Al ratio was equal to or slightly less than 2.5. Such solutions have been reported to contain polymeric species but their exact nature was unknown in the early work. Since then, the hydrolysis of the aluminum solutions has been studied by a variety of methods – magic angle spinning (MAS) nuclear magnetic resonance (NMR), powder X-ray diffraction (PXRD), and gel permeation chromatography. Lowmolecular-weight species polymerize to form the Keggin ion; both 27Al NMR and small angle X-ray scattering confirmed the formation of this ion in solution.19–21 Nazar et al. have also shown that a defect structure exists, in which one octahedron is missing from the Keggin ion so that the central Al is coordinated to 11 instead of 12 octahedra.22,23 This defect structure can combine with another like species to form an Al24 unit. Figure 3

172

Porous Pillared Clays and Layered Phosphates

O3

5.08.1.2.1 PILCs: Synthesis

O3 O1

O4

O4

O3

O3 O4

O4

O3

O4

O1

O4

O3

O1 O1

O1

O3

O3

O4 O4

O4 O3

Figure 3 Schematic representation of the [Al13O4(OH)24(H2O)12]7þ Keggin ion. Reproduced from Clearfield, A. In Advanced Catalysts and Nanostructured Materials; Moser, W. R., Ed.; Chapter 14, Academic Press: San Diego, 1996, pp 345–394, with permission.

7+

(AI13 )n m » 2.8

AI3+

m»1

AIolig

m » 2.5

(Aggregates)

7+ Aging and seeding

AI13

m » 3.0

(II)

AI(OH)3 (Solid)

AI(OH)3 or AIOOH (Solid)

Aging Aging Aging AIP1 AIP2 AIP3 (64.5 ppm) (70.2 ppm) (75.6 ppm)

Figure 4 Geometries of Al13 and Al24 and hydrolysis pathways for aluminum Keggin ion. Reproduced from Tomlinson, A. A. G., J. Porous Mater. 1998, 5 (3), pp. 259–274, with permission.

is a representation of the Keggin ion and Figure 4 is a roadmap of the several hydrolysis polymerization reactions.24 The type of species that forms depends largely on m, (m ¼ OH:Al ratio) and aging time. At m ¼ 2.5, the Al13 Keggin ion predominates and slow aging can produce the Al24 species. If m ¼ 2.8, the tendency is to create aggregates of the Al137þ species, and for m ¼ 3 the polymeric species form Al(OH)3. In solution with m ¼ 2.5 aging can lead to the Al24 species which on long aging polymerizes further.

In all of the early pillaring studies, irrespective of the type of aluminum solution used, the clay (smectite) interlayer spacing increased from 9.3 to 18–19 A˚. This result is consistent with the incorporation of the Al13 Keggin ion which has the shape of a prolate spheroid with the long axis of 9.5 A˚. Hydrolysis may lower the 7þ charge increasing the loading of the Al pillars into the clay mineral. On heating, the Keggin ion loses protons to the layers to balance the negative layer charge as the pillar is transformed into an oxide particle. Surface areas of 200–300 m2 g1 have been observed for PILCs; however, Vaughan et al.25 obtained values of 477 and 373 m2 g1. In both the cases, 62–64% of the pores were smaller than 14 A˚, 14–18% are in the range of 14–20 A˚, and the remainder are a smattering of larger pores. However, it has been shown that many factors affect the end result. Whether using the hydrolyzed AlCl3 solution or the chlorohydrol, it is well to use NMR as a guide to the extent of hydrolysis. For example, the Al(H2O)63þ solution provides a resonance of 0 ppm, whereas the Keggin ion resonance is at 62.8 ppm.26 The value of solution pH should be between 2 and 2.4 with the higher value preferred and the base should be added dropwise with stirring. For the chlorohydrol solution, a concentration of 0.06 M is considered optimum. An aging time of 10 days is also recommended.16 This solution usually contains polymers larger than the Keggin ion as evidenced by NMR peaks at 64.5 and 70.2 ppm.16,22–24 A colloidal suspension of the clay in water is added dropwise to the aluminum solution with stirring. An excess of the amount of aluminum to the IEC of the clay, usually 15 mmol of Al3þ per 1 mol equivalent of clay, is recommended.16 After addition is complete, stirring is continued for half an hour and then the mixture is repeatedly centrifuged and resuspended in water until free of chloride ions. Without the presence of the salt, the clay composite flocculates. As an example of synthesis of an aluminum-pillared montmorillonite,15,27 the clay, a well-known STX-1, was obtained and purified by wet sedimentation. The aluminum Keggin ion solution was prepared by dissolving 15 g of AlCl36H2O (0.062 mol) in 250 ml of water. To this solution was added 350 ml of 0.043 M NaOH (0.150 mol) to make OH:Al ¼ 2.43. An NMR spectrum of 27Al showed a single peak at 63.3 ppm indicating the presence of Al13. Both dilute (7.95  103 M) and moderately concentrated (0.0347 M) solutions of the Al13 Keggin ion solution were kept at room temperature to which was added a suspension of the clay. The ratio of Keggin ion to clay was 5:1 but even a ratio of 1.5:1 gave similar results. The recovered solid was washed free of chloride ion and dried at 100  C. The surface area after heating to 400  C was 414 m2 g1 and the material had a pore volume of 0.25 cm3 g1. The N2 isotherm was similar to the one shown in Figure 5. The very steep rise in the curve at very low P/P0 values is indicative of a high level of micropores that are less than 15 A˚ in diameter. Drying is an important step in controlling the nature of the PILC product. Two main conditions for doing so are air drying and freeze drying. The latter technique, being rapid, traps the layers in a more or less flocculated, random, largely edge to face orientation (Figure 6(a)). Air drying produces a more regular arrangement of face to face orientation (Figure 6(b)).

Porous Pillared Clays and Layered Phosphates

The drying procedure is followed by calcinations. The alumina is now bonded to the layers and can no longer be removed by ion exchange. What is most important in terms of catalysis is the resultant acidity of the PILC. This characteristic of the PILC has been determined by means of Fourier transform infrared (FTIR) spectra of chemisorbed pyridine28,29 and NH3 temperature-programmed desorption (TPD).30 Protons that split out from the Keggin ion are driven toward the sites of negative charge density. The most likely point of attack is at an Al–O–Si bridge with formation of a silanol group. This could be where the alumina bonds to the silica of the layer. The acidity is fairly constant to 350–400  C, but at higher temperature the Brønsted acidity is reduced, perhaps converted to Lewis acidity. All Brønsted acidity is lost by 500  C as this correlates with the dehydroxylation which occurs between 400 and 500  C.29,31 By 700  C all trace of acid character is lost. It should be kept in mind that these measures of acidity are relative. As pointed out by Tomlinson,24 one needs to compare the measured acidity to a standard scale. Usually this is a reaction, an example of which for PILCs is the cracking/ isomerization of hexane.32,33 Another procedure is that due to Haw et al. 34–36 in which NMR shifts are used to place the

1.46 1.31

Volume (x 103)

1.17 1.02 0.88 0.73 0.58 0.41 0.28 0.14 0.00 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00

P/P0 Figure 5 BET N2 sorption–desorption isotherm for Al13-pillared saponite preheated to 550  C (SA ¼ 328 m2 g1).

173

acidity on a standard scale. An example of the use of this technique to determine the acid strength of a sulfonated layered zirconium 4,40 -biphenylenebisphosphonate was reported by Wang et al.37

5.08.1.3

Pore Structure and Porosity

The pore size of a PILC is determined by the interlayer distance (height) and density of the pillars within the layer (width). The interlayer distance of a PILC depends upon the size of the pillaring agent, while the lateral distance is regulated by the charge density of the layers and effective charge on the pillars. A typical X-ray powder pattern of an aluminum-pillared saponite is shown in Figure 7(a) and after heating at 500  C (Figure 7(b)).8,27 In Figure 7(a) the initial layer peak at 19.4 A˚ is the interlayer distance 001. Lower orders present are 9.76 (002), 4.85 (004), and 3.18 A˚ (006). The other peaks are probably in-plane reflections. Upon heating the sample to 500  C the interlayer distance reduces to 18.5 A˚ and 002 and 004 peaks are still present. Both X-ray patterns suffer from intense preferred orientation. However, even if this factor was eliminated, the X-ray pattern would still be inadequate for any meaningful structure determination. However, it is possible that electron diffraction taken perpendicular to the layers may reveal some in-plane dimensions. The N2 sorption–desorption isotherm shown in Figure 5 is typical of those obtained from a variety of PILCs.38 The surface area is 328 m2 g1. The steep rise at near zero P/P0 indicates that most of the N2 was adsorbed at pressures lower than 103 torr. In fact, application of the BET method with t-plot and as-plot procedures is not very accurate in determining pore sizes of less than  10 A˚ diameter. All the small pores are filled at 103 torr. What is needed is to begin measurements at a much lower pressure such as 106 or 107 torr. The more recent application of the density functional theory (DFT) calculation still cannot accurately determine the size of the small pores. This theory works best with pores larger than 15 A˚. The actual curve in Figure 5 slopes upward beyond P/P0 ¼ 0.05 indicating the filling of larger pores with N2. The steep rise at P/P0 ¼ 0.85 is due to the filling of interparticle spaces. This isotherm exhibits a fairly large hysteresis as do most of the

Regions of zeolitic microporosity

Macroporosity (a)

(b)

Figure 6 A schematic representation of (a) delaminated clay and (b) pillared clay. Reproduced from Butruille, J. R.; Pinnavaia, T. J., Solid-state supramolecular chemistry: Two-and three-dimensional inorganic networks. In Comprehensive supramolecular chemistry, Lehn, J.-M.; Alberti, G.; Bein, T., Eds. Pergamon-Elsevier: Oxford, 1996, with permission.

174

Porous Pillared Clays and Layered Phosphates

19.4

PILCs. This is attributed to the slit-like nature of the mesopores and the strong attraction of the particle surfaces for N2. It should be remembered that there are significant differences between the clay minerals with substitution in the outer layer silica tetrahedra (saponite and beidellite) and substitution in the octahedral layers (montmorillonite and hectorite).

52.31 41.85

10.46

-2.578

-9.76

20.92

-3.18

31.38

-4.85 -4.59

Counts ´ 102

62.77

0.00 5

10

15

20 25 30 35 2–Theta (deg)

40

45

50

73.85

5.08.1.4

61.54

What is missing from our knowledge, even after 40 years of research, is a quantitative description of the PILC structure. Attempts have been made in specific cases to calculate the details of the structure from chemical data. A case in point was provided by Chevalier et al.41,42 The composition of the clay, a Ballart US saponite, was Na0.3Ca0.014 K0.003[Mg2.96 Fe2þ0.052Ti0.003Mn0.001][Si3.63Al0.369]O10(OH)2nH2O. Its IEC was 0.90 meq g1, equivalent to 1.67 mmol-Al g1. However, the amount of Al exchanged was larger than that for a 7þ charge. In fact, the amount of aluminum taken up was 2.8 mmol g1 of saponite. This places the average Keggin ion charge at 4.2þ. One unit cell contains one O20(OH)4 unit of molecular weight 777.1 g equivalent to a charge of 0.662 e. The number of unit cells per Al13 unit is 4.2/0.662 ¼ 6.5. The unit cell dimensions are a ¼ 5.29, b ¼ 9.17 A˚ and the area of 6.5 unit cells 315.3 A˚2. This area represents a square of 17.75 A˚ on a side. The Keggin ion can be considered to be a cylinder with a radius of 5.4 A˚. If the pillars are placed at the

49.23 36.92

12.31

-2.5950

-9.1458

24.62

-4.5948

Counts ´ 102

18.500

(a)

Malla and Komarneni supplied surface area data for both – a montmorillonite and a saponite.39 The former clay had an ion exchange capacity (IEC) of 1.13 meq g1 and the latter of 0.79 meq g1. Table 2 lists a higher surface area for montmorillonite, whereas we expect the opposite. The expectation is that the higher the IEC the more Keggin ion that should be intercalated into the clay. However, both clays took in almost the same amount of aluminum, 0.08 mol for saponite and 0.083 mol for montmorillonite. This phenomenon appears to be general. The average charge on the Keggin ion is 5.66þ for the montmorillonite and 4.38 þ for the saponite even though both clays were pillared the same way. Thus, the pillar can adjust itself based on other factors than charges of the clay and the pillar. Given the indeterminancy of the pore sizes it is well to turn to molecule sorption for pore-size estimates. Some early data by Vaughan and Lussier25 on sorption of molecules of different size are collected in Table 3. Sorption is high for molecules up to a size of 6.9 A˚, reasonable for 1,3,5-trimethylbenzene and not observed for molecules larger than about 7.9 A˚. This is rather disappointing as what is desired are pores larger than those of zeolites. However, it does agree with the predictions and surface area measurements that indicate micropore sizes. Sorption data presented by Occeli et al.40 for n-alkanes for both H-ZSM-5 and an alumina-pillared bentonite are shown in Table 4. The amount sorbed decreases as the length of the chain increases for both sorbents but 30% less for the PILC. However, the PILC sorbed mesitylene, whereas the zeolite did not.

0.00 5

10

(b)

15

20 25 30 35 2–Theta (deg)

40

45

50

Figure 7 X-Ray diffraction powder patterns of pillared saponite: (a) as prepared and (b) calcined at 500  C. Reproduced from Clearfield, A. In Advanced Catalysts and Nanostructured Materials; Moser, W. R., Ed.; Chapter 14, Academic Press: San Diego, 1996, pp 345–394 with permission.

Table 2

Nitrogen adsorption properties of Al13 and mixed silica–titania (Si–Ti)-pillared montmorillonite (M) and saponite (S) Surface area (m2 g1)

A13 M A13 S SiTi M SiTi S

Structure of the Aluminum PILCs

Surface area (m2 g1)

Volume (cm3 g1)

300  C

400  C

500  C

600  C

Micro 400  C

Meso 400  C

Micro 400  C

Meso 400  C

427 373 490 481

414 324 483 454

386 320 461 463

280 268 367 394

382 310 441 379

32 14 42 75

0.157 0.129 0.238 0.217

0.217 0.153 0.299 0.326

Porous Pillared Clays and Layered Phosphates

Table 3 Pore size of an Al13-pillared montmorillonite as determined by sorption of probe molecules Probe molecule

Size (A˚)

Sorption pressure (atm)

Amount sorbed (wt%)

n-Butane Cyclohexane Carbon tetrachloride 1,3,5-Trimethylbenzene 1,2,3,5-Tetramethylbenzene Perfluorotributylamine

4.6 6.1 6.9 7.6 8 10.4

0.79 0.079 0.139 0.012 0.009 0.041

8 8.4 11.5 5.3 0 0

Table 4 Comparison of the sorptive properties of H-ZSM-5 and ACH bentonite Sorbate size (A˚)

Sorbate

n-Hexane n-Heptane n-Octane n-Decane Mesitylene

Sorbate uptake (mmol g1)

Breadth

Thickness

Length

H-ZSM-5

ACH bentonite

4.9 4.9 4.9 4.9 9

4 4 4 4 4

10.3 11.51 12.3 15.24 8.6

0.981 0.774 0.669 0.65 <0.10

0.835 0.558 0.492 0.401 0.542

corners of this sphere, the free space between them is 7 A˚. Dewatering the pillars at 500  C may add 1–2 A˚ to this distance. By comparison Occelli et al.43 estimated the distance between pillars from atomic force microscopy studies to be 6.6–8.7 A˚. Chevalier et al.42 calculated the surface area to be 432 m2 g1 based upon these data, not far from the actual measured surface area of 425 m2 g1. Because the surface area depends upon so many factors, it may not always be possible to account for the surface area and pore size accurately. Pore shapes may vary leading to errors when a cylindrical or slit-type pore is chosen as a model. Some of these problems with BET and pore shape are discussed by De Stefanis and Tomlinson.44 A technique that may place the pillaring reactions on a more quantitative basis is small-angle neutron scattering. Without going into the mathematics of the procedure, the scattering of neutrons from particles in a fluid can be used to map the size and shape of the particle as the particles are much denser than the fluid. In the dried PILCs the fluid is air and its scattering density can be determined. Then the pores can be filled with a H2O–D2O mix that has the same scattering density as the solid PILC. Because hydrogen has a negative scattering length and deuterium a positive scattering length, a scattering length density can be obtained by mixing the proper quantities of H2O and D2O. The difference in the scattering curves of the dry PILC and the pore-filled PILC provides a map of the pore.45 Another technique that may prove to be very useful is positron annihilation.46,47 A beam of positrons is directed at the PILC in a vacuum and enters the pores. As the positrons strike the walls of the PILC, they are annihilated. In this way, the structure of the pore may be mapped. Finally, a theoretical approach is also required, in which a model of the pores can be used to predict the surface area and pore structure.

5.08.1.5

175

Thermal Stability

Occelli examined the thermal stability of an alumina-pillared montmorillonite in detail.48 The PILC kept at 500  C for 10 days had a 5% decrease in surface area, but larger decreases at 600  C (18%) and 700  C (37%). However, in steam, more pronounced loss of microporosity occurred. One of the principal reasons for preparing the PILCs was their potential as petroleum cracking catalysts. But it was found that the alumina PILC was susceptible to coking reducing its catalytic activity. Coking is burned off at high temperature in the presence of steam, just the conditions leading to collapse of the alumina-PILC structure. It was found that addition of a second cation stabilized the alumina PILCs and this development is treated in the next section. An extensive survey has been provided by Gil et al.49

5.08.1.6

Mixed Cation PILCs

5.08.1.6.1 Gallium substitutions A highly studied mixed cation system involves use of gallium substitution into the aluminum Keggin ion. Hydrolysis of a mixed GaCl3–AlCl3 solution produced an aluminum Keggin ion with Ga3þ in the central tetrahedral site.50,51 However, an all gallium Keggin ion isomorphous with the Al13 Keggin ion was also produced. In terms of improved stability, the order was GaAl12 > Al13 > Ga13. The Ga13-pillared montmorillonite exhibited a high dehydrogenation activity in cumene conversion and propane dehydrocyclodimerization reactions.52 In contrast, the Al13 and GaAl12 PILCs favored acid-catalyzed reactions. Coelho and Poncelet prepared Ga–Al PILCs by varying the ratio of Ga to Al 3:1, 1:1, 1:3 for both beidellite and montmorillonite.53,54 The pillaring hydroxyl-gallium and hydroxyl-aluminum solutions were prepared in 0.5 M NaOH to 0.2 M solutions of gallium or aluminum nitrate in order to achieve the desired OH/M3þ molar ratios. Required volumes of the pillaring solutions were slowly added to stirred clay suspensions (2 wt%) so as to introduce 20 meq of Al (or Ga) per g of clay. The suspensions were then dialyzed against distilled water. At the end of the washing step, the suspensions were freeze-dried. Surface areas were in the range of 200–235 m2 g1. Heating to 700  C reduced the surface area to 107–150 m2 g1 for most of the preparations. Only the sample that contained Ga in the tetrahedron of the Keggin ion retained its porosity after this high-temperature treatment. A pillared montmorillonite of low IEC was prepared by Gonzalez et al.55,56 with different ratios of Ga–Al. They were able to prepare a pillared clay with 70–85% Ga in the pillars. They found that these mixed cation-pillared clays were more stable at 700  C than the pure Al or Ga Keggin ion-pillared products. The sample prepared with a molar ratio of OH/(Al Ga) ¼ 2 possessed the best surface area and thermal stability at 700  C.57

5.08.1.6.2 Lanthanide substitutions The most widely used mixed cation systems are those involving lanthanides; however, the results have been mixed. Shabtai et al. incorporated La3þ and Ce3þ into aluminum solutions and obtained products containing about 10% lanthanide.58,59 They obtained surface areas of 220–280 m2 g1 for a hectorite

176

Porous Pillared Clays and Layered Phosphates

and high thermal stability while a fluorohectorite had an even higher surface area 300–380 m2 g1. Sterte prepared mixed Al–La chloride solutions of different ratios and obtained a pillared montmorillonite with a 26 A˚

basal spacing and 493 m2 g1 surface area.60 Although a similar result was obtained for Ce, similar results were not achieved with Nd or Pr. The best solutions had a ratio Al:Ln of 5:1. Trillo et al.61,62 obtained only a slight improvement in thermal stability with pillared La–Al montmorillonite. They suggested that the La–Al polymer consisted of four Al13 units linked by a tetrahedral La ion. Gonzalez et al.55,63 tried three different methods of preparing La–Al-pillared montmorillonite with only one of the methods yielding satisfactory results. Zhao et al.64 used pillaring solutions of La:Al up to 1. They claim that the La3þ cations are incorporated into the octahedral sites of the Al13 species resulting in polycations of general formula Al(Al12 – xLax)O4(OH)24. Booij et al.65 pillared bentonite, saponite, and hectorite with mixed La3þ or Ce3þ and aluminum polyoxocations. They obtained basal spacings of 24.8–25.7 A˚ after calcinations at 500  C. The OH:Al molar ratio was 2.5 and with high concentrations of Al ( 3.7 mol l1). The surface areas were generally high, up to 430 m2 g1. This brief survey of Al–Ln pillaring highlights a lack of agreement as to the outcomes. In some cases no improvement in thermal stability results, and in other cases the opposite is reported. Several reports of an enlarged interlayer spacing accompanied by high surface areas were obtained but no credible reason for these results is offered. We have seen how difficult it has been to come to an understanding of the aluminum species resulting from hydrolysis of the Al(H2O)63þ solutions. It appears that a similar understanding of the effect of Ln3þ additions to the aluminum solutions is not forthcoming. Thus, reproducibility of the results, especially catalytic behavior, may also not be achieved. Apparently large changes in acid character are noted, but explanations for such behavior is again not on sure footing. The same is true for the silica–alumina pillaring solutions.

5.08.1.6.3 Silicon–aluminum system Two types of Si–Al solutions were prepared by early workers.66,67 Either Si(OC2H5) (TEOS) was added to a solution of the Keggin ion, or was simultaneously hydrolyzed in a mixture of AlCl3 and TEOS. Depending upon the Si/Al ratio, more or less Si(OH)3 was affixed to the Keggin ion as shown in Figure 8. It is doubtful that this structure is based upon sufficient experimental evidence to be more than conjecture. Nevertheless, pillared montmorillonite and fluorohectorite with the mixed Si–Al solutions gave solids with 17–19.5 A˚ basal spacings that were stable to 600  C. In addition, surface areas as large as 499 m2 g1 were reported. Gil et al. varied the ratio of Si:Al over a wide range and were able to increase the amount of silica in the pillaring agent.68 The change in structure of the pillaring agent reduced the basal spacing and the surface area; however, Zhao et al.64 obtained PILCs with Si:Al ratios up to 10 and basal spacings of 26.5 A˚. On heating at 500  C, this spacing reduced to 17.5 A˚. These investigators proposed that [Si(OH)2]n polymeric species were binding to the Al13 Keggin ion. These high silica–alumina materials appeared to be more reactive catalytically for cumene cracking and alkylation reactions.

AI3+

OH O Atoms –Si(OH)3 group

Water molecule

Figure 8 Proposed molecular model of a hydroxyl-SiAl oligocation by partial substitution of –OH with –Si(OH)3 groups. Reproduced from Gil, A.; GandI´a, L. M.; Vicente, M. A., Catal. Rev. 2000, 42 (1–2), 145–212.

Several alternative pillaring methods have been described.69,70 Intercalation with Al13 and an oligosilsesquioxane produced a more homogenous distribution of the species yielding a montmorillonite with 21.4 A˚ spacing that reduced to 17.4 A˚ at 600  C. MAS NMR revealed that two types of pillars were present, silica and alumina. Atkins71 tried to coat Al2O3 pillars with silica by reaction with Si(OC2H5)4. It is not clear whether this objective was attained. Other elements that have been added to alumina are Fe, Cr, Zr, Cu, Mo, and Mn. Non-alumina combinations are Si–Ti, Si–Fe, Si–Cr, and Cr–Zr, Cr–Fe–Zr, and Ce–Zr–Al. The reader is referred to the Catalysis Reviews paper49 for details on these additional pillar types.

5.08.1.7

Zirconium-Pillared Clays

The use of zirconium compounds for pillaring of clays appeared in a number of early pillaring reports.72,73 In these studies, ZrOCl28H2O was used as the source of Zr. The solution was used at either room temperature72 or reflux temperature.73,74 The results were summarized by Bartley.73 Interlayer spacings were on the order of 16–22 A˚ with surface areas of 260–380 m2 g1. These surface areas decrease by half on heating to 300  C but complete collapse did not occur until 700  C. It was also noted that the reflux technique led to more disordered structures. Ohtsuka et al.75 repeated the reactions of unheated and refluxed zirconyl chloride solutions, ranging in concentration from 0.1 to 1.5 M, with a tetrasilica fluormica as host. In this case the unheated solution produced the less

Porous Pillared Clays and Layered Phosphates

OH2

H2O

OH2

H2O

OH2

H O

H2O Zr

Zr O H

H2O HO

OH

H2O

OH2 HO

OH OH2

H O Zr

Zr O H

H2O H2O

177

OH2

OH2 H2O

OH2

Figure 9 Representation of the structure of the misnamed zirconyl ion [Zr4(OH)8(H2O)16]8þ.

stable PILCs. The pillaring with refluxed solutions 0.1–0.5 M in Zr yielded PILCs with 23 A˚ basal spacings with 300 m2 g1 surface area and stability to  700  C. This group utilized the polymerization theory of Clearfield to explain the high interlayer spacing of their PILC.76,77 The layers tend to stack on top of each other rather than grow laterally. The thickness of a layer is  5.8 A˚ so that three layers amounts to 17.4 A˚ and four layers to 23.2 A˚, close to the observed values. Heating would split out water with formation of ZrO2 pillars. The principles upon which this reasoning rests are given in what follows. Solid zirconyl chloride does not contain a zirconyl ion, ZrO2þ, but is actually a tetrameric species [Zr(OH)24H2O]48þ as illustrated in Figure 9.78 The same species exists in acidic solution of ZrOCl28H2O.79 The chloride ions are not bonded to Zr but are present as Cl ions. Refluxing such acid solutions for prolonged periods of time leads to the formation of monoclinic zirconia.80 Addition of NaOH precipitates an amorphous zirconium oxohydroxide but refluxing in excess base produces cubic zirconia.76,80 Slow polymerization of the tetramer is postulated to form layered species as shown in Figure 10(a).76,77 These layers then come together by condensation of hydroxyl groups to form oxo-groups with stacking of the layers on top of each other.81 Rapid addition of base may create defects in the layers by more random condensation of the tetramers, as depicted in Figure 10(b), preparing more disordered PILCs. Farfan-Torres et al.82 used zirconyl chloride solutions to pillar montmorillonite and explored the effect of many variables on the product. Refluxing the solution at 100  C produced amorphous products but more crystalline PILCs were obtained if the solutions were kept at 40–60  C. Proper washing of the reacted clay was necessary to achieve good pillaring. Johnson et al. used zirconium acetate at room temperature to pillar a fluormica.83 The reaction was conducted at room temperature with an excess of the acetate solution. A washed calcined sample steamed at 700  C for 17 h had a surface area of 260 m2 g1 and a basal spacing of 20 A˚. Toranzo et al.84 carried out an extensive study of zirconia pillaring of three clays, two saponites, and a montmorillonite. They used both zirconyl chloride and zirconium acetate as

(a)

(b)

Figure 10 Two-dimensional representation of polymeric species formed by the aqueous Zr(IV) complex. The solid lined squares represent the original tetrameric units Zr4(OH)8. Each dashed line represents an – OH group formed by hydrolysis. A bent dashed line connected to two squares represents an OH bridge. (a) Ordered sheet polymer produced by refluxing solution. (b) Randomly formed polymer such as obtained by base addition to solution.

pillaring agents with ratios of 3.0 and 20.0 mmol of Zr to 1 g of clay. The solids were characterized by PXRD, thermal and elemental analyses, FTIR, N2 sorption–desorption isotherms, and surface acidity. The authors showed that the zirconium acetate produced solids with higher basal spacings and more stable textural properties than those PILCs obtained using ZrOCl28H2O solutions at reflux temperature. This chapter also discusses most of the earlier studies on zirconia-pillared clays in the relationship to their own data. Our own work,8 showed that the interlayer spacing depended upon the pH of the zirconyl chloride solution. Dilute solutions with a low ratio of Zr to clay (saponite) and pH 1.3–1.5 produced PILCs with basal spacings of 19.7 A˚ and surface areas of 300–325 m2 g1. With more concentrated solutions and pH values below 1, the materials had basal spacings of 24.5–28.5 A˚. When heated to 750  C, these spacings were reduced to 20.5–23.5 A˚ but the PILCs maintained high surface areas of  300 m2 g1. Cool and Vansant85 carried out Zr pillaring of a laponite and hectorite. The hectorite platelets are about 2 mm, similar to

178

Porous Pillared Clays and Layered Phosphates

that of other clays, but laponite layers are small, of the order of 30 nm. Because of the small size of the laponite layers, they stack predominantly edge to face and edge to edge. This arrangement produces a high surface area ( 350 m2 g1) and a micropore volume of 0.243 cm3 g1. In pure laponite, the micropores are about 1.7–2 nm, very close to being mesopores. Intercalation of ethylenediamine between the laponite layers tends to order them in a face-to-face orientation. The amines are then exchanged for the Zr pillars resulting in an increase in surface area and micropore volume. The laponite and hectorite pillaring was carried out by first stirring 3 g of clay for 16 h in 500 ml of 0.3 M ethylenediamine solution.85 The initial pH of the ethylenediamine solution was 10.8. After centrifugation, the amine intercalated clay was airdried. ZrOCl28H2O (0.1 M) was refluxed for 24 h after which the pH of the solution was 0.95. Each of the clays was added to 250 ml of the pillaring solution (10 mmol Zr per g clay) and stirred for 24 h. During this time, the solution pH increases and the solution becomes slightly yellow. The clay samples were collected by centrifugation, washed until free of Cl, and air-dried. Uncalcined samples and those heated for 2 h at 500  C (heating rate 10  C min1) were further investigated. Evidence for ethylenediamine uptake was obtained by thermogravimetric analysis (TGA) and FTIR. The Zr-laponite calcined at 500  C had a surface area (BET) of 482 m2 g1 and a micropore volume of 0.340 cm3 g1. These same quantities for the hectorite were 171 m2 g1 and 0.064 cm3 g1. The pore size range for laponite was 0.71–2.1 nm with the majority ( 85%) of the pore volume being due to pores ranging in size from 1.42 to 2.1 nm. Hectorite exhibited the same pore range with about half the pore volume. These materials have been utilized for separation of gases as will be described in Section 5.08.1.10.86,87 We have seen that the PILCs described above involved the use of ethylenediamine as a precursor to pillaring. The use of organic templates has become common as a way to increase the porosity of PILCs and modify their adsorption behavior. Preadsorption of alkylammonium ions covers some of the ionexchange sites so that the pillaring agent then only occupies a portion of the total IEC. The organic material is then removed by the calcination process. Some results of employing this procedure are presented in Table 5.86,87 Aluminum Keggin ion-pillared laponite and saponite have been grown layer by layer on a gold support.88 4-Aminothiophenol was used to anchor the first clay layer onto the amino (ammonium ion) groups that were anchored to the gold surface via the thiol groups. Layers of clay and Al pillars were absorbed by dipping the substrate into Table 5

suspensions of these materials. The material growth was monitored by ellipsometry, X-Ray photoelectron spectroscopy (XPS), and PXRD. The free interlayer space was estimated to be 6 A˚. The pillared laponite was used as films on surface acoustic wave devices to measure the adsorption capacity of volatile organic compounds (VOCs).

5.08.1.8

Synthesis of Porous Clay Heterostructures

Porous clay hetrostructures (PCHs) are formed by an intragallery, in situ assembly involving intercalation of quaternary ammonium cations, and/or neutral amines into the clay layers. This intercalation is followed by addition of TEOS that hydrolyzes between the layers to form silica pillars. After calcinations or acid treatment to remove the amines, a mesoporous pillared structure remains. A pictorial representation is provided in Figure 11.89 The PCHs are characterized by having larger pores, 1.4–2.2 nm, higher surface areas (450–750 m2 g1), and larger interlayer spacings. An example of a method of preparation was provided by Pinnavaia et al.90–92 In order to avoid the effect that impurities in natural saponite give rise to, these workers prepared a series of saponites of composition NaxMg6(Si8xAlx)O20(OH)4nH2O with x ¼ 1.2, 1.5, and 1.7.91,92 The corresponding IEC values were 0.9, 1.0, and 1.1 meq g1, respectively. The saponites were prepared by hydrothermal crystallization of gels (Si/ Al ¼ 5.67–3.57) in the temperate range of 175–200  C according to Vogels et al.93 The products were thoroughly washed with

+TEOS SiO2-x(OH)2x

(b)

T

Alkylammonium ion Amine cosurfactant

SiO2

(a)

(c)

Figure 11 Synthesis mechanism of the formation of PCH. Reproduced from Cool, P.; Vansant, E. F., In Handbook of Layered Materials, Auerbach, S. M.; Carrado, K. A.; Dutta, P. K., Eds. M. Dekker: New York, 2004, with permission.

Porosity and adsorption characteristics of pillared clays, prepared without and with templates

Type of PILC

Al-montmorillonite BuA-Al-montmorillonite Fe-montmorillonite BuA-Fe-montmorillonite Zr-laponite en-Zr-laponite

Temp (K)

273 273 194 194 273 273

SA(BET) (m2 g1)

340 361 134 233 425 482

mPV (cc g1)

0.112 0.131 0.037 0.121 0.227 0.34

Amount adsorbed (mmol g1) N2

O2

0.058 0.11 0 0.23 0.203 0.286

0.057 0.105 0.03 0.171 0.175 0.219

Porous Pillared Clays and Layered Phosphates

Table 6

179

Si/Al ratios, cation-exchange capacities, and unit cell formulas for synthetic saponite clays Si/Al ratio

#

Theory

Si29 NMR

EA

CEC (mequiv/100 g)

Unit cell formula

SAP1.2 SAP1.5 SAP 1.7

5.67 4.34 3.57

6.43 5.22 4.28

6.62 5.29 4.5

90 104 112

Na1.11 Mg6.1Al1.05Si6.95O20(OH)4 Na1.33 Mg6.05Al1.28Si6.77O20(OH)4 Na1.48 Mg5.97Al1.45Si6.53O20(OH)4

Table 7

Properties of Qþ and PCH intercalates of synthetic saponites d-Spacing (A)

Sample

Qþ saponite

PCH as synthesized

PCH calcined

BET surface area (m2 g1)

Pore volume (cm3 g1)

Acidity (mmol-CHA g–1)

SAP1.2-PCH SAP1.5-PCH SAP1.7-PCH

20.5 24.2 25.8

34.5 35.9 37.1

32.9 34.2 35.3

921 877 797

0.44 0.42 0.38

0.64 0.73 0.77

700

600

SAP1.7-PCH

500 SAP1.5-PCH 400 SAP1.2-PCH 300 Horvath–kawazoe plot SAP1.5-PCH

0.015

21

200

dW/dR

N2 volume adsorbed (cm3 g-1, STP)

deionized water, filtered, and air-dried over night at 120  C. Selected information about these materials is shown in Table 6. PCH synthesis:94 Cetyltrimethylammonium (CTMA) was used as the clay-exchange cation and decylamine as the cosurfactant. The CTMA bromide in twofold excess of the clay IEC was added to a 1.0 wt% suspension of the saponite and allowed to react at 50  C. After a 24-h reaction time the product was washed with ethanol and water to remove excess surfactant and air-dried. The organoclay, denoted Qþ-clay, was then added to the decylamine at a ratio Qþ-clay:amine of 1:20. The resultant suspension was stirred for 20 min at room temperature. The amine swells the clay gallery and participates in the gallery assembly process. TEOS was added to achieve a molar ratio of amine: TEOS of 1:6. After a reaction time of 3 h at room temperature, the gel was recovered by centrifugation and exposed to a controlled relative humidity of 50% for a period of 72 h. The resulting white powder was calcined at 650  C at a temperature ramp rate of 1  C min1. The actual Si:Al ratios were 6.95:1.05, 6.77:1.28, and 6.53:1.45. The properties of the resultant PCH materials are collected in Table 7, and the BET isotherms together with pore sizes are provided in Figure 12. It is observed that the surface areas and pore volumes are much higher than those for the PILCs. Several trends are observed as a function of increasing Al content of the PCH-modified saponite samples. The basal spacings increase, the surface areas and pore volumes decrease, and the acidity increases as does the IEC. The N2 isotherms are type IV with narrow hysteresis loops. The pore sizes derived using the Horvath–Kawanzoe theory95 range from 15 to 25 A˚. However, given the steep rise in the curves at near zero pressure, it is likely that pores smaller than 10 A˚ are also present.96 Figure 13 shows a very high resolution transmission electron micrograph (TEM) of the PCH particles. The particles as prepared are turbostratic, so direct imaging was difficult. The TEM in Figure 13 was obtained by thinning the samples so that the layers are more clearly seen. The high resolution reveals the pores which, according to the authors, are wormhole like in the ab-plane. The acid sites of the SAP-PCH were determined by cyclohexylamine desorption which occurs at 230  C and

0.01

0.005

100

0 10

0

0

0.2

0.4 P/Po

0.6

20

30 40 50 Pore size (Å)

0.8

60

1

Figure 12 N2 sorption/desorption isotherms for PCH derived from synthetic saponites. The isotherms for SAP1.5 and 1.7-PCH are vertically displaced by 150 and 300 cm3 g1, respectively. Insert: HorvathKawazoe pore-size distribution curve. Reproduced from Polverejan, M.; Pauly, T. R.; Pinnavaia, T. J. Chem. Mater. 2000, 12(9), 2698–2704.

directly correlates to the Si:Al ratio. The stronger acid site at 410  C is independent of the Si:Al ratio and is thought to be due to external surface sites. Undoubtedly these PCHs represent an interesting class of PILCs for their larger surface area, pore sizes, acidity, and stability. However, as synthesized here it was wasteful of reagents, particularly decylamine and TEOS. Subsequently, it was shown that a high-charge clay was not necessary to achieve PCH-type pore structure. A montmorillonite was treated in the same way as a saponite

180

Porous Pillared Clays and Layered Phosphates

50 nm

determined to be 0.55 mmol g1 for the montmorillonite and 0.37 mmol g1 for the saponite.100

5.08.1.9

(a)

17 nm

Normally these higher-charged clay minerals do not accept pillaring species such as the Al13 Keggin ion. However, it is possible to lower the overall layer charge to affect pillaring.101,102 This is done by heating the clay minerals with dilute nitric acid. The treatment is then followed by calcinations at 600  C for 4 h. This layer charge is consequently reduced by about one-third. A second acid treatment is carried out to remove extra-framework elements released by the first acid treatment. A complexing acid may be added to this second treatment to remove any liberated ions. These ions may block the interlayer space with prevention of pillaring if not removed. Finally, exchange with Naþ or Ca2þ is carried out. After pillaring the basal spacing expands to about 18 A˚ and maintains this distance on heat treatment at 400–500  C. Calcination at 700–800  C reduces the basal spacing to 17.5– 17.0 A˚. The pillared vermiculite and phlogopite are thus more stable than the pillared smectites; however, very few additional studies have been reported.

5.08.1.10

(b)

Figure 13 TEM of a thin sectioned SAP1.5-PCH. The arrow points to a domain of three layers in which only the middle clay layer and the two adjacent gallery pore structures are clearly evident. Reproduced from Polverejan, M.; Pauly, T. R.; Pinnavaia, T. J. Chem. Mater. 2000, 12(9), 2698–2704.

and achieved similar results.87,97 The silica pillars were further treated with aluminum acetylacetate (acac) followed by heat treatment to form Al–O–Si bonds with increased acidic character of the pillars. A significant portion of the pores were in the mesopore regime 20–30 A˚ and also present were small pores below 10 A˚. Control of the porosity and pore diameter can be affected by changing the chain lengths of the surfactant and template as well as their ratio. For example, changing the amine chain length from C6H13NH2 to C12H25NH2 alters the pore structure as shown in Figure 14. This is further illustrated in Table 8 where 1/2/15 is the ratio of Qþ-clay: amine:TEOS. Increasing the chain length of the amine results in greater surface area than simply adding more amine or TEOS. Vansant’s group also showed that the templates could be removed by use of acidified methanol.98 This procedure produced a very stable structure that could be heat-treated to 700  C without collapse of the pillars.99 The methanol PCHs had higher surface areas and pore sizes, 896 m2 g1 and 2.46–2.74 nm for the montmorillonite and 1122 m2 g1 and 2.45–2.94 nm for the saponite. The IEC that remains after calcinations can be determined by the amount of NH3 taken up by the PCH. The NH4þ was replaced by Kþ and then

Pillaring of Vermiculites and Micas

Adsorption and Ion-Exchange Properties of PILCs

A variety of gases have been shown to be taken up by PILCs. Early work was reported by Gameson et al.103 They examined the uptake of inert gases Xe and Kr on an aluminum PILC. Baksh et al.104,105 examined the uptake of gases, O2 and N2, on a ZrPILC at ambient temperatures. Their isotherms are shown in Figure 15 and compared to the isotherm of zeolite 5A. This zeolite is widely used for separation of O2 from N2. The recovered oxygen, which still contains some N2, is used for medical purposes. It was shown that even though the PILC sorbs more N2, the separation factor is greater for 5A as much less oxygen is taken up. In further attempts to improve the separation of N2 and O2, Molinard and Vansant106,107 prepared a series of PILCs with each PILC containing an alkaline earth cation. It was found that the one containing Sr2þ created the best separation of O2 from N2. In a second set of trials, it was found that an Al-PILC with 2.2 wt% Ca2þ performed much better than the same PILC without the added calcium ion. Zhu et al.108 examined the sorption of water by an Al-PILC before and after exchanging Ca2þ into the PILC. The presence of the calcium ion increased the uptake of H2O as might be expected from the ease of hydrate formation of Ca2þ in aqueous media. The authors suggest that this procedure may be useful in developing PILC-based desiccants and adsorbents for dehumidification and cooling applications. An example of size selectivity due to ion content of an AlPILC was presented by Zhu and Lu.109 Sodium ion in different amounts was added to the PILC to change the size of the pores. The separation of meta- and para-xylene on these Na-doped PILCs was determined from sorption isotherms. At low sodium ion contents there was almost no distinction in uptake of the isomers. However, at addition of 0.066 mmol g1 there was large decrease in uptake of the meta-isomer relative to the para-isomer. It was speculated that the reduced pore size caused by the presence of Naþ ions resulted in the hindrance of the slightly bulkier meta-xylene, but did not significantly impede

181

Porous Pillared Clays and Layered Phosphates

1 0.4 1/7/50

0.8 C6H13NH2

C12H25NH2

dV/dr (cc g-1.nm)

dV/dr (cc g-1.nm)

0.3

0.2

0.1

0.6

1/2/15

0.4

0.2

0

0 0.5

1

1.5

2

2.5

3

3.5

4

4.5

Pore radius (nm)

(a)

0.5

5

1

1.5

2

2.5

3

Pore radius (nm)

(b)

Figure 14 Pore-size distributions of (a) P-Mont-H using different cosurfactant lengths. (b) P-Mont-H prepared with Qþ-montmorillonite/ cosurfactant/TEOS concentrations of 1/2/15 and 1/7/50. Reproduced from Cool, P.; Vansant, E. F., In Handbook of Layered Materials, Auerbach, S. M.; Carrado, K. A.; Dutta, P. K., Eds. M. Dekker: New York, 2004, with permission. Table 8 Survey of the most important characteristics of the porous montmorillonite heterostructure (P-Mont-H) and porous laponite heterostructure (P-Lap-H)

P-Mont-H C6H13NH2 C12H25NH2 C12H25NH2 C12H25NH2 P-Lap-H C12H25NH2a C12H25NH2b C12H25NH2c

MesoPV (cc g1)

Total PV

Ratio clay/ surfactant/ TEOS

SA(BET) (m2 g1)

1/2/15 1/2/15 1/7/50 1/20/150

451 690 1093 1170

0.14 0.3 0.5 0.6

0.36 0.7 0.16 0.14

0.5 1 0.66 0.74

1/2/15 1/2/15 1/2/15

845 1083 652

0.41 0.52 0.29

0.11 0.07 0.16

0.52 0.59 0.45

mPV (cc g1)

T= 298 K Amount adsorbed (mmol g-1)

PCH Cosurfactant

0.6

N2/Zr–PILC 0.4 N2/5A

O2/5A 0.0 0.00

para-xylene. Further addition of Naþ ions lowers the uptake of both isomers making their adsorption more equal as both isomers are now too large for many of the micropores. The selectivity of methane relative to N2 was determined for five different PILCs: Zr, Al, Cr, Fe, and Ti.110 The isotherms were determined at 25  C. The equilibrium selectivity of CH4: N2 was much greater than 5 on the Al-PILC, as compared to the other PILCs. Another example of improved adsorption was obtained by Heylen et al.111 for cyclohexane, CCl4, and CO2 on Fe/Crmontmorillonite PILC compared to the pure Fe-pillared montmorillonite. This was attributed to specific absorption sites created by the addition of Cr. Improved absorption properties can be obtained in several ways. We have already discussed the incorporation of a second metal into the pillars as part of the synthesis procedure. The resulting mixed oxide pillars have been shown to exhibit enhanced absorption of CCl4, CO2, and cyclohexane by an Fe–Cr-pillared montmorillonite as compared to the adsorption of the Fe-pillared montmorillonite.106 Given the very large number of combinations of metals for the pillars and clay type what

O2/Zr–PILC

0.2

0.25

0.50 P (atm)

0.75

1.00

Figure 15 Equilibrium isotherms of N2 and O2 on Zr-PILC compared with 5A zeolite. Reproduced from Cool, P.; Vansant, E. F., In Handbook of Layered Materials, Auerbach, S. M.; Carrado, K. A.; Dutta, P. K., Eds. M. Dekker: New York, 2004, with permission.

is needed is a working theory which could be used to predict which combinations would be best for a given separation. Another technique that can be applied has also been described in the adsorption of organic templates such as an alkylammonium ion which decreases the density of pillars and increases the adsorption capacity. Vansant et al.107,108 utilized butylammonium ions as templates on both an Al and Fe montmorillonite PILC. The surface area and micropore volumes were 2.5 times higher on the Fe-PILC compared to the unmodified PILC. Pires and coworkers112 compared the sorption of VOCs on zeolites and PILCs by a gravimetric method. Both aluminum and zirconium oxide-pillared clays, one a smectite and another a synthetic laponite, were compared to zeolite-Y in different

182

Porous Pillared Clays and Layered Phosphates

cationic forms. The VOCs were 1,1,1-trichloroethane, trichloroethylene, methanol, and propanone. The isotherms of the zeolites and the pillared smectite were type I, whereas those of the laponites were type II. The zeolites sorbed a much higher level of the VOCs than the pillared smectites but the pillared laponites sorbed as much and sometimes more VOCs than the zeolites.

5.08.1.11 Chromia and Titania PILCs: Additional Porosity Considerations In preparing PILCs the amount of pillaring material incorporated can be altered by changing the IEC of the clay. By heat treatment cations of small ionic radii migrate to vacant octahedral sites in the clay. Sychev et al.113 heated samples of lithium bentonite (montmorillonite), having an IEC of 0.85 meq g1, at 135 and 150  C for 24 h. Their respective IECs were then determined by ammonium ion uptake and analyzed by the microKjeldahl method. Three samples with IEC of 0.85, 0.40, and 0.28 meq g1 were pillared by Cr2O3 and TiO2. The chromium solution was prepared by heating a 1.0 M solution of Cr(NO3)3 and Na2CO3 at 95  C for 36 h. The Na2CO3 was added to the chromium nitrate solution at room temperature to neutralize

Adsorbed volume (cm3 g-1)

140 120

two-thirds of the nitrate. The clay was dispersed in an acetone/ H2O mixed solvent to the level of 1 wt% and added with stirring to a 50 mmol Cr3þ solution per meq of clay. The titania PILC was prepared by a procedure due to Yamanaka114 with Ti isopropoxide as the source of Ti. After completion of the pillaring process, the solids were recovered by centrifugation, washed thoroughly with deionized water, air-dried, and heated at 200  C for 5 h. The porosity of the samples was determined using N2 sorption but the starting value of P/P0 was 107. The isotherms are a mixture of types I and IV as shown in Figure 16. The fact that some N2 is taken up at P/P0 values below 104 is indicative of pores of less than 7 A˚ diameter, termed ultramicropores. A second group of pores are in the 7–20 A˚ range and are designated supermicropores. The measured surface areas according to the Langmuir and BET methods are presented in Table 9. It is observed that the amount of Cr or Ti taken up as pillars depends upon the IEC of the host clay and the surface area decreases as less pillar ions are taken up. Also the ratio of supermicropores to micropores increases as the pillars are spaced further apart. The authors applied the Langmuir, BET, and as theories with the results given in Table 9. The Langmuir

30

30 1

20

20

1

2

2

100

3

10

3

10 1

80

0 10-7

10-6

10-5

0 10-7

10-4

10-6

10-5

10-4

2

60

1 2

40 20 3

(a)

0

10-7

10-6

10-5

10-4

10-3

3

(b)

10-2

10-1 100 10-7 10-6 Relative pressure (p/p0)

10-5

10-4

10-3

10-2

10-1

100

Figure 16 Nitrogen adsorption isotherms at low relative pressures: (a) (1) Cr-PM(0.85), (2) Cr-PM(0.40), (3) Cr-PM(0.28) and (b) (1) Ti-PM(0.85), (2) Ti-PM(0.40), (3) Ti-PM(0.28). Reproduced from Sychev, M.; Shubina, T.; Rozwadowski, M.; Sommen, A. I. B.; De Beer, V. H. J.; Van Santen, R. A., Microporous Mesoporous Mater. 2000, 37 (1–2), 187, with permission.

Table 9

Specific surface areas and specific total pore volume (Vt) for chromia and titania-pillared clays Surface area (m2 g1)

Sample

Cr/Ti content (mol kg1)

Langmuir

BET

as

a external

Vt (ml g1)

Cr-PM(0.85) Cr-PM(0.40) Cr-PM(0.28) Ti-PM(0.85) Ti-PM(0.40) Ti-PM(0.28)

5.1 3.11 1.72 5.69 2.68 1.87

363 324 285 440 402 356

308 285 210 290 287 254

366 301 225 332 298 267

21 35 42 46 54 75

0.191 0.233 0.227 0.21 0.274 0.302

Porous Pillared Clays and Layered Phosphates

surface area is larger than obtained by the BET method and probably overestimates the surface area, whereas the BET method probably underestimates the surface area. The as method gives values between the two and is probably the more accurate.95

5.08.1.12 Theories of N2 Sorption–Desorption Isotherm Analyses The reader may not be well versed with the intricacies of sorption–desorption isotherms and their relationship to porosity. Pore sizes are classified into three categories – 20 A˚ or less are considered as micropores, 21–500 A˚ are mesopores, and those larger than 500 A˚ are macropores. The sample to be measured needs to be heated at 200–400  C for several hours under vacuum to remove any gases clinging to the walls of the pores. The temperature is lowered to 77 K, the temperature at which N2 condenses and is allowed to enter the chamber holding the sample. The uptake of N2 as a function of the fractional pressure P/P0 is measured usually from a value of 103 torr as zero on the scale to 1. The International Union of Pure and Applied Chemistry (IUPAC) has classified the shapes of the isotherms into four categories using roman numerals.115 The ideal type I isotherm displays a sharp rise in adsorption to a high value of gas uptake at very low P/P0 and then is quite flat at higher relative pressures. Such a curve implies an ultramicropore of less than 7 A˚ as in zeolites. If this curve has a rounded feature, it indicates more than one pore size. The type II curve has an S shape and indicates a range of pores into the mesopore range. Type III is concave from zero upward and type IV has features of both types I and II, usually with hysteresis upon desorption (Figure 17). For ultramicroporous materials lower pressures are required to obtain the complete isotherm as illustrated in Figure 16 for the chromium and titanium PILCs where the relative pressure is 107. For the chromium PILCs there is a distinct change in slope at about P/P0 ¼ 104. The uptake of N2 to this pressure represents the filling of ultramicropores. At higher relative pressures, the larger pores, those between 8 A˚ and 20 A˚, are being filled. The surface area is generally

Amount adsorbed

obtained by the BET method or the Langmuir method.95

183

However, modern practice is to use the t-plot or the as-plot methods to determine the surface area.95 The idea is to compare the adsorption isotherm of a porous solid with a standard adsorption isotherm of the same adsorbate on a nonporous solid. For PILCs the reference surface might be the clay or the PILC heated to collapse so that no internal porosity is evident. The isotherm of this nonporous standard is then compared to the isotherm of the porous solid. In the t-plot method, the adsorbed volume of N2 is plotted as a function of the thickness of the adsorbed layer on the reference substance in nanometers at the same pressure. Generally, two straight lines of different slope are obtained, as illustrated for the as method in Figure 18. The points at low values of t, when extended, should extrapolate to the origin. The slope of this line is an equivalent surface area. The relevant equation connecting the slope to the surface area is S ¼ sKbt , where S is the specific surface area, s is the thickness of a single layer of N2, 0.354 nm, K is a constant (where K ¼ NA =VN2    sN2 ¼ 4:37 m2 cm3 , with sN2 being the average area occupied by a molecule of N2 in the monolayer taken to be 16.2  1020 m2 molecule1 and VN2 being the molar volume), and bt is the slope of the line through the origin. The straight line at the higher values of t can be extrapolated back to t ¼ 0. This extrapolated line at zero t provides the microporous STP volume and its slope provides the mesopore and macropore surface area. The as technique is similar to the t-plot except that as is derived from the isotherm of a reference material using the amount adsorbed at a relative pressure P/P0 equal to 0.4 as the normalization factor. It has been shown that the t-plot and as methods differ by only a constant factor when based on the same standard isotherm but yield the same estimation of the micropore volume and external surface.116 The MP method was proposed by Mikhail117 as a method for constructing a pore-size distribution based on t-plot data. The surface area of a group of pores of similar pore size can be calculated from the difference in the slopes of the tangents drawn at two adjacent points in the t-plot. The mean size of the pores in such a group is calculated from the average of the t-values of the two points. To obtain the volume of this pore group, a pore shape must be assumed. Zhu et al.116 modified this procedure to take into account a more correct value for the

I

II

III

IV

V

VI

Relative pressure Figure 17 The six main types of gas physisorption isotherms, according to the IUPAC classification. Reproduced from Sing, K. S. W.; Everett, D. H.; Haul, R. A. W.; Moscou, L.; Pierotti, R. A.; Rouquerol, J.; Siemieniewska, T., Pure Appl. Chem. 1985, 57 (4), 603, with permission.

184

Porous Pillared Clays and Layered Phosphates

Cr-PM(0.40)

Cr-PM(0.85)

Cr-PM(0.28)

Adsorbed volume (mmol g-1)

4

2

0

Ti-PM(0.85)

Ti-PM(0.40)

Ti-PM(0.28)

4

2

0

0

1

2

0

1

2

0

1

2

as Figure 18 as-Plots of chromia and titania PILCs. Reproduced from Sychev, M.; Shubina, T.; Rozwadowski, M.; Sommen, A. I. B.; De Beer, V. H. J.; Van Santen, R. A., Microporous Mesoporous Mater. 2000, 37 (1–2), 187, with permission.

0.4

0.92–0.96

Table 10 Surface area (SA m2 g1) and pore volume (PV ml g1) resulting from nitrogen porosimetry data for saponites

1.79–1.81

1.10

Preparation

Method

m SA

Total SA

m PV

Total PV

As prepared

DFT BET BJH t-Plot D-R DFT BET BJH t-Plot D-R DFT BET BJH t-Plot D-R DFT BET BJH t-Plot D-R

265

360 229 147 227

0.066

0.201

dV/dc

0.3 1.39 0.2

0.1

Expanded, air-dried 0.0 0.5

1.0

1.5

2.0

2.5

Micropore width (nm) Cr-PM(0.85)

Cr-PM(0.40)

Ti-PM(0.85)

Figure 19 Micropore width distribution of chromia and titania PILCs determined from the DS equation. Reproduced from Sychev, M.; Shubina, T.; Rozwadowski, M.; Sommen, A. I. B.; De Beer, V. H. J.; Van Santen, R. A., Microporous Mesoporous Mater. 2000, 37 (1–2), 187, with permission.

thickness of the t-layer which is statistical rather than a real pore. Distribution of pores in the Cr and Ti PILCs, as determined by the as method, is shown in Figure 19. Occelli et al.118 compared the surface areas of a pillared saponite calculated by the BET, t-plot, Barret–Joyner–Halinda (BJH),119 the Dubinin–Radushkevich (D–R)120 methods, and a hybrid DFT. As seen in Table 10, the total surface area of the DFT calculation is always much higher than those calculated by other methods. However, the micropore surface area calculated by the D–R method is always close to that of the DFT values. The authors conclude that the BET and BJH methods underestimate the surface area and only the DFT method yields reliable surface area and pore volume measurements over the entire micro-mesoporosity range investigated. However, one needs to have data taken at a pressure of 107 torr in order

Expanded, supercritical CO2 dried

Expanded, supercritical CO2 dried, steam aged for 5 h at 760 C

94 191 365

92 333 260

30 211 185

a

196

574 398 308 396 549 391 364 391 457 345 441 345

0.139 0.046 0.068 0.09

0.407 0.317

0.045 0.118 0.097

2.16 2.07

0.017 0.075 0.072

1.56 1.57

a

0.07

a

Negative values.

for the DFT method to characterize the complete micropore region. A more extended treatment of porosity in PILCs is given by Gil et al.121 The determination of the porosity of PILCs depends upon the interpretation theory used. All of them have shortcomings. The more troublesome is the lack of structural knowledge of the PILCs. We do not know with certainty the distance between the pillars, the shape, and thickness of the pillars or their uniformity. As we have seen, the range of pores may extend

Porous Pillared Clays and Layered Phosphates

5.08.1.13

Acidic Properties of PCH

We have already mentioned that clays are inherently acidic but now wish to be more specific as to the changes in acidity accompanying the synthesis of PCH. A good general introduction to the acidity of clays is given by Carrodo.5 For the most part, acidity in solids is measured by FTIR pyridine sorption. The region of interest is 1400–1650 cm1. The band at 1456 cm1 indicates the presence of Lewis acid sites and that at 1547 cm1 Brønsted acid sites. There is also a strong band at 1491 cm1 that is a combination of the Brønsted–Lewis acid site. An excellent coverage of amine interactions was presented by Yariv.6 Another common technique is temperature programmed ammonia or amine desorption (TPD). The higher the temperature required to release the amine, the greater is the acid strength. More complex is the ability to place the acidity on a standard scale such as NMR shifts induced by contact of the catalyst in question with a series of bases.34,35 As an example of acidity determination, Pinnavaia et al.124 quantitatively measured the acidity of a series of saponites by TPD of chemisorbed cyclohexylamine. You will remember that the IEC values of these saponites were respectively 0.9, 1.0, and 1.1 meq g1. Two desorption temperatures were observed, one at 220  C and the other at 410  C. The temperature of the first or weak acid desorption peak is attributed to the Si:Al ratio and largely associated with the pillars and inner surfaces of the PCH. The stronger acid sites (desorption at 410  C) are assigned to those on the outer surfaces of the PCH. A most interesting fact is that the total acidity was correlated with the IEC of the saponites. The amount of cyclohexylamine desorbed was 0.64, 0.73, and 0.77 mmol g1 for the PCHs with 0.9, 1.0, and 1.1 meq g1 IEC, respectively. This correlation between the acidity and the IEC of the saponite PILCs indicates that

the protons that remained or formed after calcinations are responsible for the acidity. Prior to Pinnavaia’s study it had been shown that the acidity of PCHs could be enhanced by the grafting of aluminum onto the surfaces of montmorillonite and saponite PCHs.87,97 A twostep procedure was employed. First Al acac was allowed to react with the PCH silanol groups. Then the absorbed aluminum acac was decomposed in an oxygen atmosphere at elevated temperatures. This procedure allowed further reaction of the aluminum acac with the large distribution of hydroxyl groups on the surfaces of the PCH. This technique was referred to as the molecular-designed dispersion procedure.125,126 It was surmised that the saponites, because they have tetrahedral Al in the silicate layers, undergo a chemical reaction with the Al acac. In contrast, only low uptake of the acac complex indicates a hydrogenbonding mechanism. The acidity of the Al(acac) clay was probed by NH3 and CH3CN sorption. The ammonia uptake is a measure of the total acidity as shown in Figure 20. It is observed that the amount of NH3 sorbed increases with the amount of Al fixed to pillared clay. The weakly basic nature of the nitrile group will only respond to sorption by strong Brønsted acid sites. Thus, its use allows determination of these strong sites and by difference the weaker Lewis acid sites.127,128 The band for CD3CN at 2309 cm1 represents the Brønsted acid site, a band at 2268 cm1 corresponds to physisorbed CD3CN, and one at 2116 cm1 is attributed to the stretching vibration ds(CD3). On heating, the band for physisorbed nitrile disappears but the band for Brønsted acidity persists. It is thought that this acidity arises from protons formed by the grafting of AlOx to SiOH to form zeolite like Si–(OH)–Al bonds. It is also possible to enhance the acidity of PCHs by using acid-treated clay minerals as starting materials for preparation of PILCs.15,129 These materials are referred to as porous acidactivated clay heterostructures (PAACHs). A natural montmorillonite was subjected to a mild acid treatment and then converted to heterostructures by the procedure of Ahenach et al.97 The results are shown in Table 11 for the acid-treated material compared to those of PCH with no prior acid treatment.130 The acidity was determined by cyclohexylamine desorption. It is obvious that the PAACH materials possess higher surface areas, pore volumes, and total acidity than the PCH, nonacid pretreated. Another technique that is gaining interest for acidity characterization is that of microcalorimetry. Heats of adsorption of 0.56

NH3 / mmol g-1 PCH

from very small into the early mesopore region. A greater effort is needed on mapping the pore sizes by using sorption of molecules of increasing size, or structural tools such as XeNMR,122 positron annihilation,46,47 X-ray and neutron small angle scattering,44 and AFM. Additionally, theoretical studies based upon the best structural analysis need to be pursued. Every theory of the nature of the pores makes assumptions about pore shape that may be inaccurate. Even the size of the N2 molecules and their orientation on the surface are contested.44 A helpful way to help interpret the N2 sorption–desorption isotherms is consideration of the types of hysteresis loops that are present in the isotherms. The most common hysteresis loops are those classified by the IUPAC.123 One would expect that the N2 sorption–desorption isotherm follows classical thermodynamics in that given the same forces involved, the isotherm should exhibit reversibility. According to Sing et al.123 the appearance of reproducible and stable hysteresis therefore implies the existence of certain well-defined metastable states. The type H1 loops are given by materials with a narrow distribution of uniform pores. For example, the H1type curve would apply to uniform cylindrical mesopores. The type H3 loops are usually a sign of slit-shaped pores or platy materials. The type H2 loops indicate a complex network of pores with both micropores and possibly mesopores connected to each other. More detailed information is available in the book by Rouquerol et al.95

185

0.54 0.52 0.5 0.48 0.46 0.44 0.42 0.4 0.09

0.38

0.58

0.77

AI-content / mmol g-1 Figure 20 Quantitative evaluation of the acidity of Al-modified P-Sap-H as a function of the Al content by NH3 adsorption. Reproduced from Cool, P.; Vansant, E. F., In Handbook of Layered Materials, Auerbach, S. M.; Carrado, K. A.; Dutta, P. K., Eds. M. Dekker: New York, 2004, with permission.

186

Porous Pillared Clays and Layered Phosphates

Table 11 Physical properties and acidity of porous clay heterostrucures derived from montmorillonite (PCH) and acid-activated montmorillonite (PAACH) clays Samples

Surface area (m2 g1)

Pore volume (ml g1)

Acidity (mmol-CHA g1)

PCH(C8) PCH(C10) PAACH(C8) PAACH(C10)

795 782 915 951

0.75 0.82 0.71 0.78

0.54 (0.14) 0.56 (0.15) 0.71 (0.25) 0.74 (0.27)

a gas, usually NH3, are determined by using a heat flow system and the amounts adsorbed determined in a volumetric chamber coupled to the microcalorimeter. The differential enthalpy of adsorption is a measure of the acid strength and the amount of NH3 sorbed provides their number. Occelli et al.118 were able to show that their air-dried saponite pillared with SiO2– TiO2 colloidal particles gave the highest heats of NH3 adsorption as opposed to a supercritical CO2 dried sample and the original saponite. The International Zeolite Association recently recommended the use of a particular reaction, the catalytic disproportionation of ethylbenzene, following a particular protocol, to assess the strength of the acid sites and the stereoselectivity of the zeolite microporous network.131 Jeronimo et al.132 prepared Al- and Zr-PILCs from a natural clay that was mostly montmorillonite with some beidellite character. Their intention was to examine the disproportion of ethylbenzene by these PILCs as a means of determining their acidity characterization.131 The PILC basal spacings were Al 17.7 and Zr 15.9 A˚. The number of pillars per nm2 of each PILC was calculated from a knowledge of the Al and Zr contents, respectively, based upon methods developed earlier.133 The pillar density of the Zr-PILC was calculated to be almost twice that of the Al-PILC. Both PILCs had about the same surface areas as measured by N2 adsorption isotherms but the Al-PILC sorbed much more ethylbenzene and n-hexane at 42  C. It was felt that the higher density of pillars in the Zr-PILC might result in some of the pores being blocked to admission of these hydrocarbons. Microcalorimetry showed that for sorption of both ethylbenzene and n-hexane, the initial heats of adsorption were greater for the Zr-PILC but became roughly equal in the middle acid strength region. Also when pyridine was used as the sorbate, it failed to show the differences in the acid character of the two PILCs. However, the Zr-PILC was the more reactive catalyst with higher conversion rates and yields of the complex reaction products. Belkhadem et al.134 prepared a number of PILCs from a montmorillonite and beidellite using as pillars Al, Ti, and Fe. Ammonia adsorption measurements found high heats of adsorption of  160 kJ mol1. These values are comparable to those of zeolites, heteropolyacids, or PILCs prepared as cracking catalysts. It was found that aging the pillared products added to the stability of the PILCs. However, the acid-treated PILCs exhibited slightly lower acidity than the non-acid-treated materials. This is in contradiction to the earlier findings of Mokaya and Jones135 and Pichowicz.130 The authors attribute this difference as being related to the procedure used for acid activation which was found to be highly specific to the type of clay.

5.08.1.14

PILCs as Catalysts

As already mentioned, clays have been used as cracking catalysts in the petroleum industry since 1938.136 In the 1990s zeolites began to be utilized as petroleum cracking catalysts.137 However, the largest pore openings in this class of catalysis are 7–8 A˚. Zeolites with larger pores have been prepared but their use in petroleum refining is limited. Therefore, one of the primary reasons for preparing PILCs was the supposition that these materials would possess larger pores than the zeolites. This is in fact so. Although PILCs also contain pores less than  7 A˚ diameter, they also contain pores in the 7–20 A˚ range. This is shown graphically for the Cr and Ti PILCs in Figure 18.113 Such pores were expected to be large enough to crack larger molecules of the heavy oil fraction into desired transportation fuels. In fact, the first results were promising,13 and similar results were reported by other workers.28,138 These PILCs yielded more light cycle gas oil than conventional catalysts.138 However, two problems inherent to PILCs are that they readily formed coke and they were not stable to steam at temperatures ( 760  C) required to remove the coke, as already indicated in Section 5.08.1.5. Subsequently, it was demonstrated that aluminum-pillared rectorite is hydrothermally more stable than zeolite-based catalysts and more effective for cracking of the heavy oil fraction. Rectorite is a mixed layer clay with both smectite and mica-type layers.139 It suffers from having only half the layers pillared such that twice as much catalyst is required for the rate of cracking to be acceptable. However, while rectorite cokes over easily, it has been found to be stable to steaming temperatures as high as 800  C.16,139–141 It also produces high levels of light oil fractions suitable as jet fuel. Gyftopoulou et al.142 claimed to prepare PILCs that continue to carry out cracking reactions even though coated with coke. Their study is based on somewhat earlier work143 that dealt with preparation of a tin-oxide-pillared laponite and a chromiapillared Wyoming Bentonite. In these preparations, both conventional and microwave heating were carried out. It is claimed that the microwave method reduces the time drastically from days to 5–30 min. We present here an example of the microwave preparation.142 Solid sodium carbonate was added to a chromium nitrate solution to attain a desired OH:Cr ration. This solution was slowly added to the clay, bentonite, or laponite, without any thermal treatment and the resulting suspension was microwave irradiated at different times and temperatures in a CEM Discover Model for focused synthesis. Alternatively the Cr–Na2CO3 solution was first microwaved at 150 W and 105  C for 15 min and then added to the clay suspension under vigorous stirring for 1.5 h. The solid suspension was repeatedly centrifuged and washed with de-ionized water until the wash water was colorless. The slurry was then filtered under vacuum, airdried, and calcined under argon for 1 h at 500  C. The exact quantities were presented in a large table.142 A micro-bomb reaction consisting of a 1/2 inch bore through Swagelok Union tee connected to a control head via a pressurization line was used for the catalytic reactions. Liquid fuel, 1 g, was introduced together with 250 mg of catalyst and pressurized with helium to detect any leaks. It was then depressurized and repressurized with H2. The reactor was then immersed into a preheated fluidized sand bath heater and

Porous Pillared Clays and Layered Phosphates

connected to a stepper-motor-driven shaker. It was kept at 440  C and 190 bar for 2 h. An elaborate description of the isolation of the products and coke formation is presented. The highest liquid conversion, 64%, was obtained with the Cr-pillared montmorillonite followed by the tin-oxide-pillared laponite. The coke was removed from the catalyst by heating from room temperature to 600  C at 10  C min1. During the second run with the decoked catalyst, the coke deposition decreased while there was no significant deactivation of the catalysts. Many other details of the PILCs are presented in this chapter.142 It has been reported in a patent144 that PCHs in combination with zeolite-Y are good catalysts for hydrocracking reactions. The advantage of hydrocracking over catalytic cracking is to yield middle distillates and light gas oils. These reactions required an acid catalyst and also a catalyst with a group 8 metal (Pt, Pd, and Ni); however, more recently, group 6 metal sulfides have been used. The combination of the PCH and the zeolite supplies high acidity and the PCH can accept higher boiling compounds in its larger pores. These catalysts also have excellent thermal stability and withstand temperatures of 800  C without degradation so they can be decoked. Additional studies are required to improve reproducibility, lower cost, and improve performance. Parulekar and Hightower145,146 impregnated a PILC with a mixture of Pt or Re and Fe3þ as the transition metal to use as a hydrocracking catalyst. In the process the Fe3þ was reduced to Fe2þ increasing the layer protons resulting in increased Brønsted acidity. This increase was determined by adsorption of pyridine. These changes produce a reactive hydroisomerization catalyst for n-C6 and n-C7 compounds. Additional examples are provided by Corma.17 PCHs may be useful for carrying out reactions with large molecules that cannot enter zeolite cavities. Polverejan et al.147 used a PCH-saponite and were able to convert 2,4-ditert-butyl phenol by a Friedel–Crafts alkylation with cinnamyl alcohol to a large flavan. While the yield was only 15%, a zeolite (H-Y) and an H*-saponite PILC produce less than 2% yields. An interesting catalyst was prepared by Mori et al.148 They ionexchanged Ru into an aluminum–montmorillonite PILC that was selective for the production of C4–C10 hydrocarbons by hydrogenation of CO. The yields were low. The literature dealing with PILCs describes many types of catalysts and diverse reactions. Rather than diminishing, the number of papers is increasing. Therefore, we briefly describe some of the diverse reaction types to help readers through this literature. Singh et al.149 synthesized pyridine and 3-picoline by the dehydrocyclization reaction of an aldehyde and ammonia. They used Al, Zr, and mixed Al–Zr PILCs and compared the product yield with the types of acid sites on the PILCs. Benito et al.150 utilized Al13 and GaAl12 –montmorillonite PILCs for methylation of toluene at 300 and 400  C. The main reaction products were xylenes, 69% for Al-PILC, and 74% for GaAlPILC. Trimethylbenzenes were formed as secondary products. A V2O5/Ti-PILC that had been delaminated by freeze-drying exhibited a high level of NO reduction by NH3. This was attributed to the highly acidic nature of the catalyst.151 The reduction of NOx to N2 by use of NH3, including kinetic studies, has been carried out by Long and Yang.152 The catalyst was an Fe–TiO2 PILC. The NO adsorbed onto the pillars and

187

then oxidized by O2 to NO2 which remained adsorbed. The NOx was then reduced by NH3 at temperatures in the range of 300–350  C to levels of 90–98. Mechanisms for the reactions were proposed. The use of NH3 for the reduction of NOx has the disadvantage of escape of some ammonia in the product stream. An alternative is to use a hydrocarbon such as methane with O2. Zirconia monoliths were impregnated with Pd and an Al-montmorillonite was impregnated with Pd, Pt, and Rh. While surface areas were low, the pore volumes were large due to the presence of mainly macro- and mesopores. Reduction of NOx with the zirconium catalysts was poor but the RhPILC exhibited a 60% NOx conversion at 400  C.153 Hectorite, montmorillonite, and saponite Al-PILCs were prepared as well as unpillared saponite and were impregnated with 2.3% Pt to make combustion catalysts. The inclusion of Pt and subsequent heating to 500  C reduced the surface area considerably and the Pt was poorly dispersed. Nevertheless, 100% combustion of methyl-ethylketone and acetone was achieved in the range of 250–350  C by the PILCs.154 Hydrogenation of benzene was carried out with Ni added to an aluminum-pillared montmorillonite (bentonite) from Greece known as AZA.155 A description of this clay mineral is given in references in this chapter.24,156 Its nickel-impregnated form is referred to as NiAZA and was examined for the hydrogenation of benzene. The most active catalysts resulted from Al3þ ion-exchanged form of the pillared NiAZA containing 12 (w/w)% Ni. Recently, more stringent regulations have been applied to reduce pollution from VOCs. Catalytic oxidation of benzene, a known carcinogen, was carried out utilizing an Al-pillared clay and an Al/rare earth-pillared clay. These PILCs supported 2 wt% of Pd for complete oxidation of benzene at about 280  C.157 The dehydrogenation of ethylbenzene to styrene was carried out with a Ti natural clay and an Al-PILC of the same clay.158 Both materials were impregnated with Co2þ and their activity at 400  C obtained. The total conversion was less than 20%. Four TiO2-pillared clays were prepared from four natural clays, montmorillonite, saponite, fluorine hectorite, and fluormica. They were utilized for the photocatalytic degradation of phthalate esters in aqueous media.159 It was shown that the best performance was obtained with the most hydrophobic PILC, the fluormica. Chlorinated organic compounds are hazardous pollutants. Cafarelli et al. compared the behavior of a series of transition metal AZA PILCs with the zeolites H-ZSM5/35 and H-ZSM-5/235 for the catalytic pyrolysis of the halocarbons.160 All the PILCs performed better than the zeolites, presumably because of their larger pores, but surprisingly the AZA performed the best. It is well known that sulfonated zirconia may act as a strong acid catalyst.38 This has prompted work to add sulfate to PILCs.161 A Fe3þ–Cr3þ PILC was heated with 1–5% of sulfate ion. In the case of methanol conversion, high levels of dimethyl ether and higher hydrocarbons were obtained at 400  C. Its use in alkylation of benzene with isopropanol yielded high formation (90%þ) of isopropylbenzene at 215  C. Additions of sulfate up to 2% were preferred as larger amounts decreased the surface area considerably. The added sulfate does improve the acidity of the PILC.

188

Porous Pillared Clays and Layered Phosphates

A series of iron aluminum PILCs were prepared and examined for their ability to oxidize phenols with H2O2.162 Percent removal varied from 1 to 40. More complex phenols were also treated with somewhat better results. The role of PILCs, layered double hydroxides (LDHs), and zeolites as catalysts in Fenton oxidation chemistry is the subject of a recent extensive review.163 A second review covers the use of PILCs in Fenton oxidations for destruction of organic pollutants in wastewater. The process also involves the change of Fe3þ to Fe2þ by photoreduction.164 Aluminum-pillared PILCs were treated with 1 M HCl at 80  C for 2 h, then stirred at room temperature for 24 h. These PILCs outperformed the zeolite H-ZSM-5 in the reaction of propylene oxide and methanol.165

solution to precipitate the desired LDH product, followed by the exchange of the included anion for the desired anion. The simplest of procedures, and subsequently the most common, for the preparation of LDHs is coprecipitation. This involves the mixing of aqueous solutions of the desired metal salts in appropriate proportions followed by an addition of an alkaline solution to precipitate the LDH product. While this process consists of simple steps, the precise control of these steps is crucial for appropriate product development. Many factors have to be considered during this process such as temperature of reaction, pH of solutions, concentrations of metal salts, concentration of alkaline solution, rate of mixing, length of reaction, and presence of ionic contaminants. 168,169

5.08.1.16 5.08.1.15

Anionic Clays

There is another classification of materials that are closely related to the presently described cationic clay materials. These are commonly known by many names: namely LDHs, lamellar double hydroxides, anionic clays, anion-exchanging clays, layered hydroxy carbonates, hydrotalcite-like compounds, and hydrotalcite-type compounds. While this list is extensive, it is generally understood that all of these compounds may be referred to as LDHs, and thus we will use that convention. Materials that belong to this class are both natural and synthetic, and form as a result of metal salts being exposed to base. They consist of positively charged layers of metal hydroxides made from two or more metal cations which are intercalated by an appropriate anion. They have the general formula [MII1xMIIIx(OH)2]xþ(An)x/n mH2O; however, this is only a general formula as they may possess either monovalent or divalent cations or a mixture of both. LDHs are useful materials as catalysts, precursors to catalytic metal oxides, material precursors (such as ceramics), anionic-exchange or extraction materials, drug-delivery platforms and other medical applications, and additives for composite polymer materials. These materials contain many basic sites, and thus their most prominent applications serve to neutralize large quantities of acids. Unfortunately, the major application of these materials is their usefulness as-is and without the need for added porosity; subsequently, the field of making porous LDHs has not been extensively explored except to serve as precursors for porous metal oxides which are covered elsewhere in this book series. The authors would address that there have been several major reviews and books in the past two decades that do an immaculate job of describing the methods for the preparation of LDHs. For this reason, this classification of compounds will not be discussed in much detail other than to compare and contrast them to their close relatives, the cationic clays.166,167 LDHs can be found in nature or synthesized in the lab. These compounds are referred to as hydrotalcite-like or hydrotalcite-type compounds because they all have structures similar to the mineral hydrotalcite or its hexaganol analog, manasseite. The routes to synthesis of these compounds are diverse and we will discuss a few of the most common methods, their benefits, and their limitations. Most of these methods involve addition of soluble metal salts to a high pH

Conclusion

We shall complete this section of the chapter with a consideration of two important papers that point to the future of PILCs. It has been pointed out that the PILCs differ from other types of porous materials such as zeolites and MOFs, but that they have interesting properties of their own.170 For example, (1) the physical properties may be varied by isomorphous substitution, (2) the pillars become part of the inner pore surface introducing catalytic behavior or chirality, and (3) pore size and shape of the PILCs are established by both the dimensions and shape of the pillars and pillar density. These factors provide the synthetic chemist with a wide array of pillars and clay minerals from which to choose in order to engineer porous materials with the desired chemical reactivity. Furthermore, the charge density can be adjusted by redox chemistry.171 Finally, PILCs have a measure of swelling as in gas storage. It is further pointed out that natural clays have defects and suffer from structural disorder and heterogeneity of the charge density of the layers.170 These factors are responsible for the broad range of pore sizes observed in PILCs. To eliminate these deficiencies, Stocker et al.172 prepared a highly crystalline, 2D-ordered, synthetic Cs-fluorohectorite. Synthesis of Cs-fluorohectorite is as follows: High-purity (99.9%) reagents were used throughout. CsF (10 g), LiF, MgF2, MgO, and SiO2 amounts were chosen to duplicate the composition Cs[Mg5Li]Si8O20F4. The reaction was carried out in a sealed Mo crucible at 1750  C. The compound was then annealed for 7 days at 1175  C in a sealed Mo crucible that was gas tight so as to release strain and increase crystal size. 2H-DABCO (1,4-diazabicyclo[2.2.2]-octane) was used to pillar the clay. This pillaring was carried out in two steps. First, H-DABCOþ was intercalated at pH ¼ 7.5. This required 10 treatments for 24 h reflux of 500 mg with an aqueous 100 ml of H-DABCOþ Cl solution (0.1 M). After this treatment, the H-DABCOþ can be replaced rapidly and quantitatively by 2H-DABCO2þ (three treatments). A highly crystalline, large platelet product of composition Cs0.56[Mg2.44Li0.56]Si4O10F2 and unit cell a ¼ 5.2401(10), b ¼ 9.0942(10), c ¼ 10.7971(10) A˚, and b ¼ 99.21 , was obtained by this fusion procedure. The high crystallinity achieved allowed observation of the mechanism of the pillaring reactions. Weiss et al.173 propose that two different mechanisms of intercalation are possible. If the platelets of the clay are smaller than the cooperative elastic action length, then the one-sided mechanism holds. That is, through elastic deformation, all edges

Porous Pillared Clays and Layered Phosphates

immediately sense when the reaction has started on one side and nucleation on the other side is stalled. Large platelets with diameter larger than the cooperative elastic length may not sense this elastic deformation so that intercalation may proceed from all the edges at the same time. This mechanism is termed the ‘ring mechanism.’ The particles described here have sizes of > 100 mm and a series of scanning electron micrography (SEM) images shows the replacement of the Csþ by the DABCOþ from the outer edges inward. Of further interest is the fact that after all the H2DABCO2þ is inserted between the layers, the crystallinity remains strong and the powder patterns show that the 3D stacking order is maintained with new unit cell dimensions a ¼ 5.247(1), b ¼ 9.083(2), and c ¼ 14.581(5)A˚, and b ¼ 96.799 , space group C2/m. In addition to reflections giving rise to the listed unit cell, there are additional weak reflections arising from a super cell with dimensions a*  3a or 15.731(3) A˚, b*  b ¼ 9.090(2) A˚. The authors are of the opinion that this is the first example of a material for which a 3D ordering in the stacking of the layers could be observed together with a 2D superstructure of pillars.170

The authors have provided a scheme of the two lattices based on the small and supercell (Figure 21). It would be expected that placement of the pillars 15 A˚ apart results in pores larger than 7 A˚. However, the Ar adsorption–desorption isotherm is type I with the majority of pores in the 4–6 A˚ range as shown in Figure 22. There is only a smattering of pores of larger size. This lack of large pores is attributed to the presence of water in the pores as shown by 1H MAS NMR. A considerable amount of space within the pores is occupied by H3Oþ by transfer of the protons from the layer to water molecules. Upon ion exchanging half the Hþ for Naþ the isotherm shows an almost nonporous material with a type II isotherm. The Naþ is surrounded by a layer of H2O filling the cavities. This water cannot be completely removed by drying. The authors point out that the clay they prepared is not the best choice for a host in terms of developing porosity; it did show the intercalation mechanism for the pillars and the arrangement of the pillars within the layers. The second paper to consider is that of De Stefanis and Tomlinson.44 These authors point out that not a single detailed reaction mechanism for PILC-catalyzed organic reactions has been forthcoming. The problem is structural in nature, without a better knowledge of the particular PILC structure; little can be done to elucidate catalytic properties. PILCs have complex structures; the layers can deform which can lead to disordered topologies. Inserting material between layers exerts a further stress on the layers as does drying and heating. Furthermore, the distance between pillars is not known with certainty, nor is the thickness and shape of the pillars. Figure 23 shows the many ways that PILCs can be made. This is both a curse and a blessing. What is required is uniformity and reproducibility but even clays with the same name may have different compositions and impurities. What is needed are a few model clays that are structurally simple and can be synthesized in a highly crystalline state. Synthetic clays are not new.174,175 Kloprogge et al.175 obtained a fairly crystalline magnesium aluminum saponite after a 72-h hydrothermal treatment at 200  C. Surely such a preparation may be improved upon to obtain clays that meet the requirements outlined. With modern PXRD methods and other advanced structure revealing tools, the required structural details should be forthcoming. The ultimate test is to then unravel the entire process of pillaring and to relate structure properties to catalytic reactions.

a

b 3a b

b

a

78

0.07

65

0.06

Pore volume (cm3 g-1)

Volume (cm3 g-1)

Figure 21 Scheme of the 3a  b superlattice of pillars in 2H-DABCOhectorite. The unit cell of the parent fluorohectorite is shown in dotted lines. Reproduced from Stoecker, M.; Seyfarth, L.; Hirsemann, D.; Senker, J.; Breu, J., Appl. Clay Sci. 2010, 48 (1–2), 146, with permission.

52 39 26 13

0.05 0.04 0.03 0.02 0.01 0

0 0 (a)

189

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Relative pressure ( p/po)

0.9

4–6

1.0 (b)

6–8

8–10 10–12 12–14 14–16 16–18 18–20

Pore width (Å)

Figure 22 (a) Physisorption/desorption isotherm for Ar and (b) histogram of the pore width of 2H-DABCO-hectorite (gray) and Na/H-DABCO-hectorite. Reproduced from Stoecker, M.; Seyfarth, L.; Hirsemann, D.; Senker, J.; Breu, J., Appl. Clay Sci. 2010, 48 (1–2), 146, with permission.

190

Porous Pillared Clays and Layered Phosphates

Intercalation/ flocculation +

IV

III

Calcination

1

etc 2 I

Expansion -nPrAH + OAc

II Exfoliation

3 +

+ TEOS (or other alkoxide) +

Nano-particle 4

Direct insertion

Amine-solvatated Q-clay TEOS

TEOS H2O

OH OH OH

OH OH OH

TEOS-intercalated intermediate

OH

OH

calc.

Sio2

OH OH

O

5

OH

Templated heterostructure

Porous clay heterostructure

6 Oriented platelets in platelets

Figure 23 Current methods for pillaring clays. (1) Traditional route, using Al13 Keggin ion for alumina pillars. (2) A surface hydrolysis/polymerization reaction used for preparing chromia-pillars. (3) Via sol–gel generated intercalates, both homogenous and heterogenous. (4) Direct intercalation of externally generated known-size nanoparticles. (5) Assisting interpillar spacing via amines and surfactants. (6) An example of possible use of platelets themselves as pillars. Reproduced from De Stefanis, A.; Tomlinson, A. A. G. Catal. Today 2006, 114(2–3), 126, with permission.

5.08.2 5.08.2.1

Layered M(IV) Phosphates and Phosphonates In the Beginning

Zirconium phosphate gels had been synthesized and studied as early as 1925.176 These materials showed great stability in acidic and neutral solutions, as well as significant IECs. In the 1950s, concomitant with the advent of nuclear power, these amorphous ion exchangers garnered significant interest for their potential role in removal of radioactive species generated in the cooling water of nuclear reactors.177,178 Inorganic compounds were preferred for this purpose because of their stability to ionizing radiation. However, these phosphates were susceptible to hydrolysis at elevated temperatures and therefore were not utilized for this intended purpose. Up until the 1960s, all of the reported work on ion exchange by zirconium phosphate was carried out

by studying the amorphous powders or gels. The discovery of crystalline zirconium phosphate resulted from experimental work carried out by James Stynes under the direction of Clearfield.179 Further work by this group resulted in the synthesis of a second zirconium phosphate.180 It was evident that both of these compounds were layered and possessed ionexchange properties. The former compound of composition Zr(O3POH)2 H2O was termed a-zirconium phosphate (a-ZrP) and the latter of composition Zr(PO4)(H2PO4)2H2O is identified as g-zirconium phosphate (g-ZrP). Subsequently, Clearfield visited Giulio Alberti in Italy. The Alberti group had been working on the zirconium phosphate gels for a number of years and were leaders in this field of research. Upon learning of the crystalline forms, Alberti remarked ‘this changes everything.’ This group then took up

Porous Pillared Clays and Layered Phosphates

work on the crystalline materials and the two principals, Clearfield and Alberti, remained friendly rivals to this day. Shortly after, Clearfield and coworkers were able to solve the crystal structure of a-ZrP. This allowed researchers to fully understand the kinetics and thermodynamics of ion-exchange behavior of this robust material. About the time that the ionexchange behavior and exfoliation properties of ZrP had been exhaustively characterized; the field was pushed forward by the concept of replacing phosphate ions with phosphonate groups. This allowed for the inorganic layers to be separated in a prescribed manner according to the size of the R group. In addition, the functionality of the R group was now united with the stability of the inorganic layer and arranged in a uniform manner on the surface. During the development of these materials, once again Alberti and Clearfield came together in collaboration. Alberti had synthesized a zirconium phenylphosphonate derivative (Zr(O3PC6H5)2)181 and had sent a small sample of these crystals to Clearfield for analysis. The crystals were too small for single-crystal studies; however, the structure was solved from 39 reflections obtained by synchrotron powder X-ray data from the Brookhaven National Laboratory.182 From this structure and the extensive study of zirconium phosphate, an entirely new field of hybrid materials would flourish. Zirconium phosphates are not highly porous in and of themselves, but can be made so by appropriate modification. By applying the pillaring method commonly used to impart porosity to layered clays, the layers of tetravalent metal phosphates can be pillared by transition metal oxides, silica, or alumina. Not only does this incorporate the chemical attributes of the pillar, but it spaces the layers apart to allow access to the surfaces in the interior of the material. This method for pillaring ZrP is complicated by the high surface charge density of the layers, which tends to result in densely packed pillars that preclude the formation of porosity. However, there are clever strategies for dealing with this tendency, and many pillared M(IV)P derivatives have been synthesized and shown to have potential applications in ion exchange and catalysis. Another method of pillaring M(IV) phosphonates is by using organic pillars, in the form of bisphosphonates, to hold the layers apart. These pillars can be topotactically exchanged into the gamma form of ZrP, or incorporated by direct precipitation into the alpha form. Smaller, monophosphonate spacer groups can be interspersed between the pillars to increase the porosity, and the spacers or pillars themselves can be functionalized to impart reactivity to the material. Here we describe the synthetic methods used to prepare these materials, as well as properties and applications of such compounds. The principles which drive the design of these types of porous materials are also discussed. Synthetic methods and techniques are covered in the context of their application to porous phosphates and phosphonates.

191

predicted a structure in which the Zr atoms were coordinated by six oxygen atoms from six different phosphate groups, forming a sheet-like structure. The phosphate oxygens bearing the hydrogen atoms were predicted to be in two different hydrogen-bonding environments with the solvent water molecules, resulting in the different acidities observed for the two exchangeable hydrogen atoms. This structure was later confirmed184 and then improved185 by combining single-crystal XRD with knowledge gained from neutron diffraction studies of powders. The neutron studies186 allowed for an accurate description of the hydrogen bonding within the interlayer space, completing the structural characterization. Figure 24 shows the structure of a-ZrP, which can be thought of as PO4 tetrahedra capping hexagonal lattices of Zr cations on both sides, forming a sheet. The P-OH groups extend perpendicular to the layers into the interlayer space, where they hydrogen-bond with the water molecules held there. There are no hydrogen bonds between the layers themselves, which are held together only by van der Waals interactions. The inorganic layers are 7.6 A˚ apart, and the exchangeable protons are spaced at a surface density of one per every 24 A˚2. In a sense the structure is clay-like, with octahedral metal ions between layers of tetrahedral phosphate groups encasing the metal. However, unlike clays the phosphate tetrahedra are inverted with three oxygen atoms bonding to the metal and one oxygen atom projecting into the interlayer space. While the structure is claylike, the IEC is 6.64 meq g1, which is many times greater than for clay minerals. An extensive literature exists on the ionexchange properties of a-ZrP. The exact behavior depends upon the degree of crystallinity which varies from amorphous to fully crystalline.187 Figure 25 shows the structure of g-ZrP, which contains two different phosphate groups, one fully deprotonated and one which retains two protons. The fully deprotonated phosphate is trapped within the layer being bonded to four Zr ions, while the doubly protonated one is bound to the surface of the layer through an M–O–P–O–M bridge formed by the unprotonated oxygen atoms. This is in stark contrast with the phosphate groups in the a-phases, which are bound to metal atoms in the layers by three of their four oxygen atoms. The interlayer spacing is 12.4 A˚ which is also the c-axis.188 There are two water molecules between the layers that form a continuous network of hydrogen bonding with the hydroxyl groups of the layers. The dihydrogen phosphate group bridges across Zr(IV) or Ti(IV) ions held within

5.08.2.1.1 General structure and properties of a- and g-M(IV) phosphates The structure of a-ZrP was elucidated by Clearfield and Stynes in 1964, based upon the ion exchange and dehydration behavior of the crystalline material they had prepared.179 Formulated as Zr(HPO4)2H2O, the compound lost one mole of water upon heating to 110  C, and exhibited a two-stage ion exchange of Na for H.183 From these data, the researchers

Figure 24 Ball and stick structure of a-ZrP layers. Hydrogens have been removed for clarity.

192

Porous Pillared Clays and Layered Phosphates

the layers by PO4 groups. The metal ions form straight rows running parallel to the a-axis. The unprotonated oxygen atoms of the second phosphate group form bridges across these rows, making the coordination about the metal ions 6. This leaves the two hydroxyl groups hanging into the interlayer space forming hydrogen bonds with the interstitial water molecules. a- and g-M(IV) phosphates of many 4þ ions have been synthesized and characterized. Table 12 lists their compositions, unit cells, interlayer spacings, surface density, and IECs. Of the forms, not all have been structurally characterized, but we can assume that the structures are reasonably similar based on interlayer spacing and IECs. One of the most interesting properties of a-ZrP is its ability to intercalate a number of different compound types. This behavior is a result of the acidic protons forming a double layer within the interlayer space. Intercalation may be driven by a topotactic acid–base reaction as, for example, uptake of amines, oxidation–reduction reactions, and ion exchange. With special procedures, it is possible to insert guests with

Figure 25 Ball and stick structure of g-ZrP layers. Hydrogens have been removed for clarity.

Table 12

very low base strength such as alkanols, glycols, ketones, and amides, and much more.189–194 The mechanism of intercalation is similar to that of the ionexchange processes. It was observed that Naþ ion exchange occurred with formation of the half sodium-exchanged phase even when as low as 5% of the required Naþ to form the half exchanged phase was taken up. The question posed was whether 5% of the crystals were converted or did the crystallites themselves consist of two phases. It was shown conclusively that exchange begins at the intersection of two edges and diffuses inward with an advancing phase boundary; however, it may happen that the intercalate forms a number of phases that differ in composition and interlayer distance. This is particularly evident with alkyl amine intercalation. In general, the arrangement of the guest depends on its size, shape, and stoichiometry as well as cointercalated solvent. Two early reviews on a-ZrP intercalation have been provided.195,196

5.08.2.2 Preparation and Characterization of M(IV) Phosphates Zirconium phosphate was originally prepared as a gel by mixing aqueous solutions of a ZrIV salt and phosphoric acid, which resulted in a precipitate that could be collected by filtration or centrifugation. Varying the Zr source, reaction temperature, and concentration of phosphoric acid was shown to have drastic effects on the ion-exchange properties, surface area, and overall composition of these gels. P:Zr ratios could range from 1:1 to 2:1, and the compounds contained significant amounts of water. The compounds with low P:Zr ratios could be accurately considered as Zr oxophosphates. Even when the starting P:Zr ratio is 2:1, the product may have a deficiency of phosphate. The synthesis of crystalline M(IV) phosphates was first carried out by refluxing mixtures of Zr salts and phosphoric acid (3–12 M). Increasing both the reaction time and the concentration of phosphoric acid resulted in more crystalline materials with larger average particle size.187 This has been effectively used to control the aspect ratio of ZrP particles.197 X-ray patterns and particle sizes of the products obtained by the reflux method are shown in Figure 26.187 Another method, referred to as ‘direct precipitation,’ involves the slow removal of HF (which effectively solubilizes Zr compounds by forming

Some characteristics of layered phosphates of group 4 and 14 elements

Formula

a (pm)

b (pm)

c (pm)

b ( )

Interlayer distance (pm)

Free areaa (104 pm2)

Ion-exchange cap (mmol Hþ g1)

Density (g cm3)

a-Ti(HPO4)2H2O a-Zr(HPO4)2H2O a-Hf(HPO4)2H2O g-Ti(PO4)(H2PO4)22H2O g-Zr(PO4)(H2PO4)22H2O a-Si(HPO4)2 a-Ge(HPO4)2H2O a-Sn(HPO4)2H2O a-Pb(HPO4)2H2O

863.0(2) 906.0(2) 901.42(1) 518.1(1) 538.6

500.6(1) 529.7(1) 525.66(5) 634.7(1) 663.6

1618.9(3) 1541.4(3) 1547.68(2) 1188.1(1) 2480.6

110.2 101.71(2) 101.64 102.59(1) 98.7

21.6 24 23.7 16.5 17.8

496.43(5) 498.70(2)

1586.05(4) 1612.50(6)

100.03(1) 100.615(3)

7.76 6.64 5.15 7.25 6.27 8.9 7.07 6.08 4.79

2.61 2.72

861.15(3) 862.38(4)

756 756 760 1160 1220 740 775 780 795

a

Area associated with each OH group on the plane.

21.4 21.5

2.37 2.43

3.12

Porous Pillared Clays and Layered Phosphates

Intensity

ZrP(12M)

ZrP(9M)

193

ensure a highly crystalline product. The solid was recovered by filtration, washed with 2 M HCl to remove Naþ followed by washing with 0.2 M H3PO4 to remove chloride ion. A final wash with small quantities of distilled water and drying at low temperature under vacuum was required to prevent removal of the interlayer water. Only three types of g-phases are known, those of Zr, Ti, and the Zr-arsenate.201

5.08.2.2.1 Exfoliation of M(IV) phosphates ZrP(6M)

ZrP(3M) (a)

10

20

30 40 2-Theta

50

60

a-ZrP and g-ZrP can be exfoliated, meaning they can be converted to a colloidal dispersion of single layers. n-Alkylamines are taken up by a-ZrP from aqueous solutions, organic solutions, and even from the gas phase.189–191 For primary amines a total of 2 mol of amine are taken up to form a bilayer. This uptake corresponds to the number of –POH groups in the interlammelar space. When propylamine is slowly added to a-ZrP, the amine enters and lies parallel to the layers forming a phase with a 10.4 A˚ interlayer spacing. During uptake to saturation, several other phases form. However, as one approaches the midpoint of uptake slowly, swelling of the layers occurs.192 By shaking or use of sonication at this point, exfoliation occurs.189 Alberti correctly interpreted this phenomenon.192 At half exchange, the amines are  10.6 A˚ apart but the interlayer spacing is increased considerably by the bilayer. This expansion has weakened the interlayer forces and the van der Waals forces are also weak because of the small chain and the large separation distance. Exfoliation allows for insertion of larger species between the layers.

5.08.2.3 Porous Phosphates and Phosphonates Synthesized by Templating and Self-Assembly

(b)

Figure 26 (a) X-Ray powder diffraction patterns of a-ZrP prepared at different molar concentrations of phosphoric acid. (b) SEM pictures of the same samples. Reproduced from Sun, L.; Boo, W. J.; Sue, H.-J.; Clearfield, A. New J. Chem. 2007, 31(1), 39–43, with permission.

ZrF62 ion) from the solution, resulting in the gradual precipitation of large ZrP crystals.198 The use of microwaves and hydrothermal methods speeds up the process. It should be obvious that highly crystalline samples of ZrP are not porous, as the interlayer space is filled by water molecules. Upon removal of the water molecules, the layers simply move closer together. Poorly crystalline ZrP synthesized by these methods exhibits only very low surface area, which is because it consists of much smaller particles of the same layered structure. However, other techniques can be used to create nanoparticulate ZrP, which has been calculated (based on the volume and density of the nanoparticles) to have an external surface area of 250 m2 g1.199,200 There is another form of ZrP, the g-form which is more difficult to prepare, and thus a more detailed procedure is required. Preparation of g-ZrP was carried out as follows180: 100 ml of a 1 M ZrOCl28H2O solution was added dropwise to a stirred solution consisting of 276 g (2 mol) of NaH2PO4H2O in 200 ml of 3 M HCl. Refluxing of the sodium dihydrogen phosphate solution was required for complete dissolution of the salt. Refluxing was continued for 25 h to

The use of surfactants has been shown to create mesoporosity in M(IV) phosphates. This is generally done by adding the surfactant to the solution of phosphoric acid and aging this solution, which allows the surfactant to form micelles. Upon addition of the M(IV) source, which can be MOCl2, MCl4, M(iOPr)4, etc., the M(IV) phosphate precipitates in the aqueous phase around the micelles, which act as templates. The precipitate is then aged, sometimes with heating, and dried. The surfactant is then removed by extraction with a solvent or by calcination. Sometimes solvent extraction is used prior to calcination to decrease the presence of carbon within the mesopores. Early work in this area was inspired by the synthesis of the mesoporous silica MCM-41 and related compounds.202,203 CTMA bromide was used to template the hydrolysis product of Zr(SO4)24H2O. This precipitate was then treated with phosphoric acid solution to replace some of the sulfate groups, thereby stabilizing the structure. After removal of the surfactant and sulfate by calcination, the surface area of the material was determined to be 230 m2 g1. Increasing the length of the alkyl chain on the surfactant to 20 carbons resulted in a surface area for the final product of 390 m2 g1.204 This material is technically a zirconium oxyhydroxide which has had its surface modified by treatment with phosphoric acid, and not porous zirconium phosphate. However, considerable success has been achieved by using templates to synthesize porous versions of MIV phosphates. The use of long-chain trimethylammonium surfactants (CTMA bromide, chloride, and derivatives) has been effectively applied to the synthesis of porous zirconium and titanium phosphates.

194

Porous Pillared Clays and Layered Phosphates

Zirconium and titanium phosphate prepared in this manner were shown to have surface areas of 326 and 340 m2 g1, respectively, once the surfactant had been removed by solvent extraction or calcination.205,206 Later work207 showed that titanium phosphates with surface areas up to 740 m2 g1 could be obtained by careful manipulation of the reaction temperature and pH. After calcination at 540  C, the materials still showed mesoporous character and surface areas up to 290 m2 g1. The surface acidities of these compounds were found by gas-phase NH3 adsorption at 80  C to be up to 900 mmol g1. Optimization of reaction conditions for ZrP using CTMA chloride yielded materials with surface areas up to 500 m2 g1 and tunable pore-size distributions centered between 3 and 8 nm.208 The phosphate source was found to be quite influential on the final product: Zirconium phosphate precipitated with (NH4)2HPO4 was found to be highly deficient in phosphorus (P:Zr < 1.6), while that precipitated with H3PO4 or NH4H2PO4 had P:Zr ratios close to 2:1. This is attributed to the increased pH of (NH4)2HPO4 solutions which leads to increased rates of hydrolysis relative to H3PO4 or NH4H2PO4. Controlled hydrolysis can be an effective tool for manipulating the surface charge of ZrP, and this is discussed further in Section 5.08.2.3. In general, it seems that while cationic (CTMA bromide/ chloride) and neutral (hexadecylamine, alkylamines) surfactants have a positive effect on the surface area of ZrP, anionic surfactants such as sodium dodecylsulfate (SDS) and 4dodecylbenzenesulfonic acid (DBSA) result in materials with low surface area. This is not the case for Ti phosphates, which have been shown to form templated porous materials with cationic, neutral, and anionic surfactants. It should be noted that the Zr samples templated with SDS were obtained by stirring at room temperature while the SDS-templated Ti phosphate was heated at 75  C and TiCl4 was used as the Ti source. Long-chain n-alkylamines serve as good templating agents for Ti phosphate, resulting in wormlike pores.209 This material was found to be an active catalyst for the oxidation of cyclohexene. The surface areas of the mesoporous materials were found to depend on the alkyl chain length: 359 m2 g1 for C12, 430 m2 g1 for C16, and 497 m2 g1 for C18. The mean pore diameter also varied as a function of the length of the alkyl chain, ranging from 1.7 nm (for C12) to 3.3 nm (for C18). This is slightly different from the results obtained by templating zirconium phosphate with hexadecylamine, which gave materials with a mean pore diameter between 4 and 8 nm, depending on the reaction conditions.208 Trioctylamine has been used to template g-TiP, resulting in a multiphase compound that forms tubes as shown in Figure 27.210 Trioctylamine facilitates the rolling of these sheets of g-TiP into tubes with mean diameters of 50–100 nm. No surface area was reported, but it appears that the tubes could act as somewhat uniform mesopores. Titanium phosphate nanospheres templated with dioctyl sodium sulfosuccinate were found to have a surface area of 305 m2 g1. These nanoparticles are interesting because they consist of NaTi2(PO4)3 cores surrounded by a mesoporous layer of Ti(HPO4)2. They showed catalytic activity in ketalization reactions which was not decreased after five cycles.211 Mesoporous TiP was synthesized with C18TMA Br (420 m2 g1) and Cl (548 m2 g1), and also with SDS and dodecyl p-benzenesulfonic acid. Samples were shown to have not only high cation-exchange capacity due to Brønsted acid

sites, but also significant anion-exchange capacity. This was attributed to the presence of tetrahedral phosphonium ions in the amorphous pore walls. Most of the titanium atoms were also found to be tetrahedral, which gives the material great potential as an oxidation catalyst.212

5.08.2.3.1 Nonionic templates It is of interest to prepare porous phosphates using nonionic templates since these are more easily removed by calcination than quaternary ammonium salts. Polymer templates also tend to yield mesopores, since they aggregate into larger structures than ionic templates. This can be beneficial in creating multimodal pore structures which allow easy access to the microporous internal regions of the material. Titanium phosphate was precipitated in the presence of 5% polyethylene oxide (PEO), and it was found that this material had higher porosity than nontemplated titanium phosphate. The material was mesoporous with a pore-size distribution between 80 and 250 nm. Increasing the surfactant concentration to 10% resulted in a hierarchically porous system with multimodal pore-size distributions. With 10% PEO, the macropores formed as parallel channels, but with 15% PEO, the macropores formed a sponge-like 3D network as shown in Figure 28.213 A study of a series of zirconium phosphate materials templated with C16H33(EO)10 and C18H35(EO)10 revealed that pore size was slightly dependent upon the alkyl chain length on the PEO chain. The samples templated with C18H35(EO)10 also showed higher total surface area, up to  300 m2 g1. The samples retained at least some of their porosity and nearly all of their acidic P-OH groups even after calcination at 650  C.214 Preliminary work215 determined that amorphous titanium phosphates templated with PEO or CTMA chloride exhibited high affinity for aqueous Pu(IV), some with Kd values as high as 16 000. Promising results were also obtained for the sequestration of Np(V), indicating that these materials may have applications in nuclear waste processing. An interesting macroporous titanium phosphonate was reported by Ma.216 The material was synthesized by the inverse-opal method using 500 nm polystyrene spheres to template the reaction product of 1-hydroxy ethylidene-1, 1-diphosphonic acid (HEDP) and tetrabutyl titanate. b-Cyclodextrin ((C6H10O5)7) was also used to modify the packing of the polystyrene spheres, resulting in a less-ordered macroporous material. The materials were found to adsorb

200 nm

20 nm

Figure 27 High-resolution TEM images of g-TiP tubes. Reproduced from Blanco, J. A.; Khainakov, S. A.; Khainakova, O.; Garcıa´, J. R.; Garcıa´Granda, S. Phys. Status Solidi C 2009, 6(10), 2190–2194, with permission.

Porous Pillared Clays and Layered Phosphates

195

5 mm (a)

2000 nm (b)

1000 nm (c)

200 nm (c)

Figure 28 SEM images showing TiP sponge-like material prepared with 15% PEO solution. Reproduced from Ren, T.-Z.; Yuan, Z.-Y.; Azioune, A.; Pireaux, J.-J.; Su, B.-L. Langmuir 2006, 22(8), 3886–3894, with permission.

Cu2þ, Cd2þ, and Pb2þ ions from solution, which may make them useful in water purification. Interestingly, there has been some work done using yeast as a biotemplate to synthesize porous zirconium phosphate.217 The authors report that including yeast as a template yields zirconium phosphate with surface area of 217 m2 g1. However, this value was obtained for a sample which contained about 20% carbon by weight, due to the decomposition of the yeast cells when the sample was heated to 300  C. This carbon probably

contributes significantly to the surface area of the material. The maximum in the pore size distribution histogram was reported to be 5.15 nm, which is much smaller than an average yeast cell. The templated framework most likely suffers a partial collapse as the organic portion decomposes upon heating, resulting in pore sizes much smaller than the yeast template. Overall, template-assisted synthesis of phosphates seems to have dwindled, most likely overshadowed by the silicas, which form crystalline compounds and have regular pores (while

196

Porous Pillared Clays and Layered Phosphates

phosphates still show a range of pore sizes). Current interest is centered on self-assembled materials (those not requiring a template) and hierarchal porous materials. Recently, there have been some developments in the template-free synthesis of porous phosphates.218 Fei et al. found that slowing the rate of hydrolysis by increasing the ratio of n-butanol to titanium n-butoxide resulted in larger pores and higher overall surface area; however, upon calcination these materials lost a significant portion of their surface area and micropore volume. Other work has shown that the alcohol generated by the hydrolysis of M(IV) alkoxides is essential to the formation of pores.219 A zirconium phosphate synthesized by adding zirconium isopropoxide to a solution of orthophosphoric acid and then heating the solution at 80  C was shown to have a surface area of 528 m2 g1, of which only 24 m2 g1 were from micropores. Calcination at 650  C resulted in a decrease of overall surface area to 248 m2 g1 along with an increase in micropore surface area to 43 m2 g1. The authors suggest that the water–alcohol channels formed by the rapid generation of propanol upon hydrolysis of the zirconium isopropoxide are the source of the channel-like pores within the material.219 The inclusion of the surfactant C16(EO)10 (Brij 56) resulted in materials with lower total surface areas but higher percentages of micropores. The authors hypothesize that the surfactant may affect the precipitation and condensation of the zirconium phosphate nanoparticles, inhibiting the formation of the macroporous structure.220 CTMA bromide has been utilized as a template to create a mesoporous titanium HEDP with a surface area of 1052 m2 g1. The CTMA bromide forms an ordered hexagonal arrangement of channels which are the source of the mesoporosity in the final material. The surfactant was extracted from the material with acidic ethanol. Calcination resulted in a decrease in surface area, but even after being heated to 570  C the sample retained 321 m2 g1. This material was found to be an active photocatalyst for the degradation of Rhodamine B and an effective sorbent for CO2 with a capacity of 1.0 mmol g1. The CO2 uptake was reversible and the sample exhibited good reusability.221 Similar hexagonal channels were obtained when Brij 56 was used as a template in the reaction between TiCl4 and tetrasodium ethylenediamine tetra(methylene phosphonate) (EDTMPA). The polymer template was removed by Soxhlet extraction with ethanol to yield a mesoporous titanium phosphonate with a surface area of 1066 m2 g1. This material showed promise as a one-pot wastewater treatment that could remove toxic heavy metals (Cu2þ and Pb2þ) as well as catalyze the photodegradation of organic contaminants (Rhodamine B).222 Other CTMA bromide-templated titanium phosphonates have shown similar activity for photodegradation catalysts.223 Triblock copolymers F127 (EO106PO70EO106) and P123 (EO20PO70EO20) were used to template Ti HEDP, resulting in materials with surface areas of 377 and 511 m2 g1, respectively. These compounds were also evaluated for Cu2þ removal from aqueous systems and reversible CO2 adsortion, with promising results.224 The inclusion of alkylamines in the reaction between diethylenetriamine penta(methylene phosphonic acid) (DTPMPA) and tetrabutyl titanate served to increase the number of defective P–OH groups, resulting in porous materials with Hþ exchange

capacity up to 5.76 mmol g1. These materials show potential as heterogeneous acid catalysts.225 EDTMPA and DTPMPA were used by Zhang et al.226 to synthesize mesoporous titania-based materials which exhibited selective adsorption properties for Cd2þ over Pb2þ and Cu2þ. This behavior was shown to be due to the chelating ability of the aminophosphonate ligands. The authors also reported enhanced photocatalytic activity for the decomposition of Rhodamine B compared to mesoporous titania. A similar titanium phosphonate synthesized from EDTMPA and tetrabutyl titanate was shown to have a hierarchically macroporous structure and to preferentially adsorb Pb2þ over Cu2þ and Cd2þ.227 Other meso/macroporous titanias phosphonated with HEDP, EDTMPA, and DTPMPA were shown to have varying selectivity in the uptake of Cu2þ, Cd2þ, and Pb2þ.228 Sarkar et al. reported a titanium oxophenylphosphonate synthesized hydrothermally in the absence of a template.229 TEM images and the PXRD pattern support the conclusion that this material is layered, and BET analyses showed that the compound had a surface area of 340 m2 g1. 31P NMR did not show the peak at 4 ppm characteristic of layered titanium phenylphosphonate. XPS showed the presence of both tetrahedral and octahedral Ti(IV) atoms, so the structure may be a derivative of g-TiP. The synthesis of porous derivatives of M(IV) phosphates by the use of templates is a promising area of research. One of the limitations of layered phosphates is the rate of diffusion into the interlayer space, so the incorporation of mesopores into the structure could drastically improve mass-transfer properties. A convenient feature of this method is that M(IV) phosphates can be precipitated from many different solvents, so the design of the template mesostructure is the limiting factor in the creation of new porous materials. Triblock copolymers have been used with great success as templates in lithography and the design of porous silicas, and it is likely that they will be the templates for the next novel M(IV) phosphate-based porous materials.

5.08.2.4 Pillaring of M(IV) Phosphates by Inorganic Oligomeric Cations Pillaring of M(IV) phosphates with inorganic oligomeric cations is performed in a manner similar to that used to pillar clays. The process can be divided into three basic steps: preswelling or exfoliation, then ion exchange, followed by calcination. Typically, M(IV) phosphates are converted into their butylammmonium salts or exfoliated to facilitate the exchange of large oligomeric cations into the space between the layers. Much research has been done on the ion-exchange behavior of M(IV) phosphates,230 and only the aspects of ion exchange pertaining to the process of pillaring to introduce porosity are discussed in this section. The amount of oligomeric cation intercalated depends on the charge of the ion and its size. With pillared clays, the exchange is limited by the low amount of positive charge required to balance the negative charges of the layers. However, layered M(IV) phosphates have high IECs, so a greater number of cations are required for charge balance. This results in what are called ‘stuffed’ materials, which are pillared to such an extent that there remains hardly any space

Porous Pillared Clays and Layered Phosphates

between the pillars and very low porosity. The high density of pillars also inhibits access to any pores that may be formed. After the inorganic cation has been exchanged into the M (IV) phosphate, the material is calcined to chemically bind the pillar to the layers and to remove any organic ligands which may accompany the cations, such as acetate groups. Calcination also serves to convert the oligomeric cations to oxides, which often have desirable catalytic properties which were intended to be incorporated into the porous material.

5.08.2.4.1 Transition metal oxides In an effort to incorporate chromia into alumina-pillared layered phosphates,231 researchers mixed Cr(NO3)3∙9H2O and Al (NO3)3∙ 9H2O in water with n-propylammonium acetate, then refluxed that solution with colloidal suspensions of a-ZrP intercalated with n-propylamine. By changing the ratio of Al3þ to Cr3þ in the initial mixture, they were able to control the same ratio in the final product. Up to 20 meq g1 of Cr3þ and Al3þ were taken up by ZrP. ZrP exhibited a higher affinity for Al3þ than Cr3þ, requiring a 40:60 ratio of Al3þ to Cr3þ in the initial solution to produce a 50:50 ratio of Al3þ to Cr3þ in the product. In all ratios a significant increase in d-spacing was observed. The sample obtained from the 40:60 ratio of Al3þ to Cr3þ in the initial mixture had an interlayer distance of 42.1 A˚, which decreased to 35.8 and then 31.3 A´˚ upon calcination at 200 and 400  C, respectively. After calcination at 400  C, the BET surface area was 643 m2 g1. The decrease in interlayer distance upon heating was observed for all compounds. The fact that large d-spacings were maintained, indicating that the structure had not collapsed and that the layers were indeed held apart by the oxide pillars. At high concentrations of Al3þ, the products were found to have low surface area, presumably due to blocking of the pores by alumina on the surface. A similar study was carried out using Cr3þ and Fe3þ solutions.232 From XPS measurements, it was determined that the ZrP surface exhibited a higher affinity for Cr3þ at high Fe/Cr ratios. BET surface areas up to 306 m2 g1 were obtained when the samples were calcined in N2, and were lower when the samples were calcined in air. A partial oxidation of Cr3þ to Cr6þ and subsequent segregation of CrO3 were observed when the samples were calcined in air at 400  C. The Ga/Cr mixed cation system was studied by using a methodology similar to that of Olivera-Pastor et al., by adding n-propylammonium acetate, followed by addition of colloidal ZrP/n-propylamine, and then refluxing for 2 days.233 Calcination was done under N2 to prevent oxidation of Cr3þ and formation of CrO3 which segregates and causes a decrease in the surface area. After calcination the materials had moderately high surface areas (184–236 m2 g1). The ZrP exhibited a higher affinity for Ga3þ than Cr3þ, but this may be due to the types of oligomers formed, which may be greater for Ga than Cr. Gallium oligomers were also incorporated into ZrP synthesized using dodecyltrimethyl ammonium bromide as a surfactant. It was found that the cationic surfactant molecules would easily exchange for gallium oligomeric cations. These exchanged compounds were then calcined at 400  C to yield gallium oxidepillared layered ZrP with surface area up to 366 m2 g1.234 TEM images clearly showed the layered structure of the pillared materials. Pyridine adsorption was used to determine that the majority of the acid sites within the materials were Lewis type,

197

and the compounds were found to be catalytically active in the conversion of isopropanol to propene. An interesting crystalline material was obtained by refluxing propylamine intercalates of ZrP with Ni2þ salts for several days. The structure retained the layers of ZrP, but the protons had been exchanged for Ni2þ ions, which were bound to the phosphate O atoms. In between these layers was a third layer of nickel hydroxide, resulting in the formula Zr(PO4)2Ni4(OH)5 (C2H3O2)2H2O. Upon calcination, the layered structure of the ZrP is maintained as the nickel hydroxide/acetate decomposes to form a layer of nickel oxide three layers thick between the layers of ZrP. This results in a contraction of the d-spacing. A maximum surface area of 158 m2 g1 was obtained when the material was calcined at 350  C. Increasing the calcination temperature caused a decrease in surface area, and complete structural collapse was observed at temperatures above 650  C. The authors reported that the compounds showed significant catalytic activity for the aromatization of n-hexane.235 By contacting aqueous suspensions of sodium-exchanged ZrP with titania sols prepared by hydrolysis of titanium isopropoxide, titania-pillared ZrP, and TiP were prepared.236 In general, the TiP samples exhibited larger d-spacings and higher surface areas (up to 172 m2 g1) than the ZrP samples. The compounds, loaded with 2–4% TiO2 by weight, were found to have surface acid sites which enabled the materials to function as active catalysts for the photodegradation of 4-nitrophenol.

5.08.2.4.2 Alumina and related inorganic clusters Usually intercalation is achieved with aluminum salt solutions having Al/OH ratios from 1.8 to 2.2. Many species exist in solution, as determined by the effects of aging, temperature, and pH, but the system can be tuned to give predominantly Al13, which has a distinct NMR signal. This topic has been discussed at length in Section 5.08.1.4. Alumina is a highly porous, thermally robust material with a wide variety of applications in catalysis. It also forms large, highly charged oligomers under proper conditions. These Keggin-type ions are too large to be directly intercalated into ZrP, but are readily taken up by ZrP that has been preintercalated with butylamine. The hypothetical advantage to pillaring with these ions is due to their uniform size, which could result in more uniform pores in the final material, and their high charge density, which decreases the number of clusters necessary to balance the negative charge of the deprotonated ZrP layers. Fewer pillars would mean more void space between the layers, resulting in higher surface area and pore connectivity, allowing reagent molecules to easily enter into the material and access the active sites. However, the situation is complicated by the effects of aging and pH on the solutions of Al13, which cause agglomeration; therefore, pillaring is often done with larger aluminum clusters. Early work was done by Clearfield and Roberts in 1988.237 In an effort to overcome the inherent thermal instability of organic pillars in porous phosphates, they attempted to use the aluminum Keggin ion, [Al13O4(OH)2412H2O]7þ, and [Zr(OH)24H2O]48þ as pillars to separate the zirconium phosphate layers. In a manner similar to what had been done for pillared clays, butylamine was first intercalated between the layers of ZrP. Then the Al13 or Zr4-based pillar was exchanged for the butylammonium ions. The observed surface areas for

198

Porous Pillared Clays and Layered Phosphates

the pillared zirconium phosphates were only 30–37 m2 g1 and they did not absorb amines or ammonia. These results are indicative of uptake of smaller aluminum species, which produces a stuffed layered structure. Many other ions and polymerized aluminum species were present in the solutions, so a combination of Keggin ions and other smaller species could result in the stuffed character of pillared ZrPs. However, titanium phosphates (a- and g-) pillared with Al13 solution had higher surface areas 121 (g-) and 183 m2 g1 (a-) and were shown to adsorb large molecules such as cyclohexane and pentane. Titanium phosphates pillared in this way retained 70% of their surface area to 400  C. The typical method for the preparation of alumina-pillared MIV phosphates starts with a colloidal suspension of MIVP exfoliated or intercalated with an alkylamine. A solution of Al oligomer is then added, and the mixture may be refluxed or hydrothermally treated. The product is then collected by centrifugation or filtration and washed. Calcination above 400  C results in the conversion of the aluminum polycations to alumina, which presumably forms as isolated pillars that hold the MIVP layers apart, creating pores and accessible surface in the interlamellar space. There are many variables in this scheme, the major ones being the nature of the hydrolyzed aluminum species and the state of the MIV phosphate. Many of the reported studies of these compounds do not use carefully made solutions which contain only Al13, instead opting for commercially available solutions of hydrolyzed aluminum chloride or ‘chlorhydrol.’ It should be remembered that these solutions contain a number of other partially hydrolyzed aluminum chloride species, of both lower and higher nuclearities than Al13. In general, ZrP exhibits a stronger interaction with Al13 and other similar species than does SnP. This results in stuffed pillared structures as all kinds of hydrolyzed aluminum polycations are captured between the layers. To decrease the strength of this interaction and avoid the inclusion of the smaller oligomers which cause the stuffing, F ion has been used by some researchers to compete with OH. Fluoride ion has a high affinity for Al3þ, and replacement of OH ions by fluoride decreases the number of possible O–P–O bonds which can form between the phosphate layer and the hydrolyzed aluminum polycation. This results in weaker interactions between the phosphate and the aluminum cation and the exclusion of low-nuclearity species from the interlamellar space, avoiding the formation of stuffed structures. a-SnP was intercalated with Al13 ions. Calcination of this material would presumably lead to a porous layered material in which alumina pillars hold apart tin phosphate layers.238 A colloidal suspension of 1:1 a-SnP and propylamine was prepared to which was added a tenfold excess of tetramethylammonium chloride (TMAC). After centrifugation and drying, the half exchanged material was obtained (d002 ¼ 17.5 A˚). This material was suspended in aliquots of Al13 for 1 day at pH 4.4, at room temperature, and it was seen that d002 increased to 19.27 A˚. The formula was determined to be Sn([Al13O4 (OH)24(OH2)12]0.14 H(PO4)2)nH2O. However, after heating to 400  C the interlayer spacing was 27 A˚, (d002 ¼ 13.5 A˚) with structure collapse occurring when heated to 500  C. The Al13SnP had a surface area of 85 m2 g1, but a wide distribution of pores. After calcination at 400  C, the surface area was 228 m2 g1 and had pores centered at 22 A˚. Cobalt and nickel

divalent ions were taken up to the theoretical maximum (2.5 meq g1), but Cu was taken up to 4.1 meq g1, presumably due to formation of CuOHþ. The material made by using chlorhydrol solution was found to have lower surface area, only 190 m2 g1.239 Chlorhydrol intercalated products had a d-spacing of 24.2 A˚ and about 50% more Al than Al13SnP, indicative of a double layer of Al13 or partially condensed oligomers of other nuclearity, perhaps due to the pH effects of the layer. Ammoniation increased the metal uptake capacity, perhaps because NH4þ is a better leaving group, or the ion-exchange process has a cleaning effect and removes decomposition products from between the pillars, allowing greater access to the porous structure. A similar material was evaluated by the same group.240 AC conductivity measurements were performed on both the Hþ and Liþ phases of the material. The low conductivity of the Liþ exchanged sample, relative to the parent Sn phosphate, was due to the weak lateral connectivity between the small layered particles which form upon calcination of the Al13 intercalated material. NMR was used to study the Sn and Zr phosphates pillared with Al oligomers.241 It was found that either monolayers or bilayers of the inorganic cation formed, depending on the solution used. Variations in PXRD d-spacing were used to determine this feature, with the monolayered materials exhibiting a d-spacing of 17 A˚ and bilayers giving higher values. The dialyzed solutions also resulted in the incorporation of significantly higher amounts of aluminum in the products (23X). Chlorhydrol solution also resulted in the incorporation of more aluminum. Again, the ZrP samples were found to have much lower surface areas (70–75 m2 g1) than the SnP samples (180–190 m2 g1). Oligomer condensation was shown to occur when the intercalation was performed under reflux. The interaction between the inorganic phosphate layers was stronger for ZrP than in SnP. 27Al MAS NMR showed resonances indicative of aluminum ions in both octahedral and tetrahedral coordination geometries, as well as a third resonance at 30 ppm which may be attributed to pentacoordinate aluminum ions. However, this NMR evidence for the pentacoordinate aluminum is inconclusive, as distorted tetrahedral geometries may cause the same shift in resonance. High concentrations of this signal were observed. The increased strength of the interaction between the phosphate layer and the oligomer in the ZrP compounds results in the stuffed pillared structures. Peeters and VanSant described an unsuccessful attempt at intercalation of Al13 into amine-loaded ZrP.242 Their primary difficulty was in displacing the amines with the Al13 Keggin ion. The materials they obtained from the butylamine intercalate were found to be Naþ or Al(H2O)63þ loaded phases. NMR showed that Keggin ions were on the surface, so other amines were tried to increase accessibility to the interlayer space. Complete exfoliation with small amines led to amorphous products, and larger amines did not completely exchange. In the case of butylamine, which gave the best results, the butylamine/ZrP was added with stirring to the pillaring solution at different temperatures. This is perhaps not the best way compared with other methods of preparation. Fluoride ions were found to partially replace OH ions in the oligomers used to pillar ZrP with (presumably) Al2410þ cations. A colloidal suspension of propylamine-ZrP was

Porous Pillared Clays and Layered Phosphates

prepared, and the Al solution added to that, then refluxed for 2 days. The product was collected by centrifugation, washed, added to n-propylammonium fluoride again, and hydrothermally treated at 200  C for 1 day. The product was collected by centrifugation and washed before it was calcined at 400 or 600  C. Surface areas up to 184 and 139 m2 g1 were obtained by heating to 400 and 600  C, respectively. Again, 27Al MAS NMR indicated the presence of possibly pentacoordinated Al3þ ions. The acidity of the materials was probed by ammonia and pyridine sorption, and the compounds were found to have a large number of both Lewis and Brønsted acid sites. The authors suggested that Lewis acid centers associated with lowcoordinated aluminum ions were the source of the materials significant catalytic activity toward the conversion of isopropanol to propene.243 This chapter references materials used in several future papers. Subsequently, Braos-Garcia and coworkers intercalated fluorinated oligomers of Al/Ga larger than Al13 into SnP materials.244 The highest surface area was obtained for AlSnP (304 m2 g1), but the materials intercalated with the mixed Ga/Al species also had high surface areas, between 160 and 243 m2 g1. Pentacoordinated aluminum was observed in the NMR spectrum, and the materials also displayed activity and selectivity as dehydrating catalysts for isopropyl alcohol. Other groups picked up on this technique and started similar research in this area.245 Xu and coworkers found that incorporation of the alumina pillars resulted in a decrease in Brønsted acidity (fewer P–OH groups). Aluminum ions were hydrolyzed in situ, resulting in large uptake of Al3þ relative to Al13 intercalated samples. This chapter contains work strikingly similar to that performed by Jiminez-Lopez and coworkers. Alumina-pillared derivatives of g-ZrP have also been prepared.246 g-ZrP was dispersed in acetone/water at 80  C, then a solution containing fluorinated Al24 was added, enough for 75% of exchange capacity, then heated at 200  C in a Teflon liner, centrifuged, washed, and dried at 50  C or calcined at 400  C. Simultaneous presence of four-, five-, and six-coordinate aluminum ions was shown by NMR. Again, the five-coordinate signal could be due to distorted four-coordinate, but Al24 has been known to form pentacoordinate aluminum upon thermal decomposition.247 Surface areas were 116–119 m2 g1. Transition metal-loaded alumina-pillared phosphates have been shown to be effective as catalysts in a variety of applications. The AlSnP compounds discussed earlier were loaded with Cu ions after calcination. After loading, the materials were calcined at 500  C. The copper oxide supported on these materials was found to effectively catalyze the reduction of NO with propane.248 This reaction was also investigated for cobaltloaded alumina-pillared ZrP derivatives.249 The catalysts did not perform as well as zeolite-based catalysts, since the majority of the acid centers are of the lewis type, and Brønsted acid sites are needed to activate the hydrocarbon molecules.250 The authors suggested that the incorporation of Brønsted acid sites by using a physical mixture of the catalyst with an acidic material would increase the catalytic performance. The same group251 found that nickel particles (in the range of 3–42 nm) supported on the Al/Ga oxide-pillared SnP were active in the gas-phase hydrogenation of acetonitrile at atmospheric pressure, although they suffered greatly from deactivation. This deactivation was hypothesized to be the result of carbonaceous

199

deposits, which blocked access to the pores, causing a decrease of the BET surface area by  25%. Treatment under hydrogen or helium flows did not reactivate the catalyst. Interestingly, TEM of the deactivated catalyst showed that sintering was not the cause of deactivation, as an increase in metal particle size was not observed. Fluorinated alumina-pillared a-ZrP versions loaded with Ni, Ni and Rh, or Ni–Mo and Rh were evaluated as hydrogenation catalysts for tetralin. The authors observed a high yield of decaline and cracking compounds with a low yield of volatile compounds, but the catalysts were deactivated when the reaction temperature exceeded 400  C. The addition of dibenzothiophene at a concentration of 1000 ppm to the reagent flow revealed that the catalysts suffered from severe deactivation, but they could be partially regenerated by treatment with hydrogen. The conversion of the feed with 300 ppm dibenzothiophene and the feed without any sulfur compounds was similar, indicating that this concentration of dibenzothiophene did not inhibit the conversion.252

5.08.2.4.3 Silica and other mixed oxides Partially condensed silanes are an attractive choice for a pillaring agent, as their bulk reduces the overall loading between the layers, which after calcination should result in higher surface areas. The organic moiety of a silane can also include various functional groups such as amines. These functional groups can serve to bind transition metals either after the silanes have been intercalated or before intercalation; in the latter case, the silane will serve as a chaperone to bring outside transition metals with them to the interlayer space. After calcination, the organic parts of the silanes are driven off, and only TM-doped silica remains, acting as a pillar to hold apart the layers of ZrP. A typical reaction is to reflux the oligomeric siloxane with ZrP. Amine groups on silanes can coordinate metal ions which can be incorporated into the oxide pillars upon calcination. 3-Aminopropyltriethoxysilane was used by Jiao et al.253 to synthesize a crystalline layered g-titanium phosphate derivative. The compound was made hydrothermally at 180  C, which effectively increased the crystallinity so that a unit cell could be determined. The authors suggested that the structure consisted of inorganic titanium phosphate layers which were pillared by the silane. Calcination of the material resulted in the decomposition of the organic groups and a contraction of the interlayer spacing from 19.9 to 14.0 A˚. Nitrogen adsorption revealed that the silica-pillared material obtained after calcination had a BET surface area of only 51 m2 g1, which is a consequence of the dense packing of the silica pillars in the interlayer space. Roziere and coworkers also utilized octameric aminopropyltriethoxy silane to pillar MIV phosphates in alcohols.254 ZrP was found to form a fully exchanged intercalate with the octameric siloxanes, which exhibited a d-spacing of 17.69 A˚ that decreased to 12.5 A˚ upon heating to 500  C. Similar results could be obtained with TiP and SnP, but SnP was found to form two phases when intercalated with the siloxane octamers. Low ratios of Si:Sn resulted in a material with a d-spacing similar to that obtained for ZrP and TiP, but when the ratio of Si:Sn was increased to 3:1, a second phase was obtained with a d-spacing of 26.9 A˚. The high amount of Si in the compound and the large d-spacing support the conclusion that a double layer of siloxane octamer has been intercalated. This SnP

200

Porous Pillared Clays and Layered Phosphates

compound was the only material which had significant surface area (230 m2 g1). The authors suggested that hydrolysis results in the deactivation of acid sites on the surface of the SnP layers, which prevents ‘stuffed’ pillared structures from forming. The same ligand was used by Clearfield’s group255 to pillar Zr, Ti, and Sn phosphates through one of two techniques. The first was the direct intercalation of the siloxane oligomer into the metal phosphate by reflux in water or ethanol. This method was unsuccessful for titanium phosphate, which suffered from significant hydrolysis. The hydrolysis and subsequent collapse of the layer structure were also observed for SnP, but could be mitigated by performing the direct intercalation in ethanol instead of water. A fully exchanged phase with a d-spacing of 26.2 A˚ was obtained. For ZrP, only partial intercalation was achieved in ethanol, but reflux in water yielded a single phase with a d-spacing of 20.1 A˚. The second technique used to prepare the intercalated materials involved forming the halfexchanged propylamine phase of the MIV phosphate prior to reflux with the aminopropylsilane solution. This procedure was unsuccessful for TiP and SnP. However, ZrP pillared in this manner yielded a microporous material with d-spacings of 29.6 and 17.2 A˚. Heating to 500  C resulted in the formation of a single phase with a d-spacing of 21.5 A˚ and a surface area of 232 m2 g1. The hydrolysis observed for the colloidal SnP when refluxed in the presence of aminopropylsilane is in contrast with the results obtained by Roziere’s group, which showed that SnP could be successfully pillared with silane oligomers by allowing the reaction to proceed at room temperature for 5 days, instead of under reflux. Reflux significantly increases the extent of hydrolysis of the SnP, resulting in the degradation of the layer structure. While this hydrolysis may prove to be beneficial at low levels by decreasing the exchange capacity of the SnP, resulting in a more open structure, at higher levels the overall layered structure is compromised. To accomplish silica pillaring g-ZrP was synthesized by standard procedures and then exfoliated in a 1:1 mixture of water and acetone.256 Pillaring of the material was carried out by adding tetrapropylammonium hydroxide and TEOS. The mixture was coagulated with the addition of acetic acid then KCl to obtain a gel which was dried at 80  C, then calcined at 600  C for 2 h under air. Very high loadings of silica could be obtained by this method, up to 91 wt.%. The d-spacings of the materials were  35 A˚, except for the sample loaded to 64 wt% silica, which had a d-spacing of 45 A˚. The BET surface areas were 300 m2 g1, but the sample loaded to 84 wt% silica had a surface area of 639 m2 g1. The compounds were impregnated with PdCl2 and H2PtCl6 solutions as a route to obtain nanoparticles supported on the ZrP/silica matrix. PXRD studies revealed that the samples consisted of both ZrP and silica phases, as well as reduced or oxidized Pt and Pd. The samples were found to be catalytically active in the hydrogenation of toluene and naphthalene. Later work257 showed that the inclusion of lithium within the supports promoted the hydrogenation reaction by weakening the interaction between the metal particles and the support, but H2S eliminated the beneficial effect.

5.08.2.4.4 Conclusions and general outlook for inorganically pillared phosphates The difficulty with pillaring of the a-layered metal phosphonates is the inability to control the synthesis reactions to obtain predictable structures. Furthermore, the structures obtained are

noncrystalline or poorly crystalline. Therefore, the process is hit or miss as opposed to metal–organic frameworks where crystalline products are obtained.258 This knowledge of structure allows the design of a great variety of porous materials. The same problem exists for the pillared clays, but in that case we have reported more efforts at structure determination and greater understanding of the end product. It is unlikely that the exhaustive physical and theoretical efforts are made with respect to the pillared phosphates because there is no guarantee that the end result warrants it. In Section 5.08.2.6, we discuss unconventional metal–organic frameworks (UMOFs), a group of porous pillared phosphonates that are worth expending the effort to achieve structural knowledge.

5.08.2.5 Pillaring of g-M(IV) Phosphates by Topotactic Exchange There are two chemically different phosphate groups in the gforms of MIV phosphates, which have the general formula M (PO4)(H2PO4)H2O as has been described in Section 5.08.2.1.1. This feature of g-forms allows for these surface phosphates to be replaced by a process of topotactic exchange, in which another phosphate (or arsenate, etc.) group can take the place of the phosphate bound to the layer. This process can be utilized to incorporate phosphonate groups onto the layer. Often simply dissolving the phosphonic acid in a suspension of the g-MIVP is enough to facilitate the exchange; however, this only exchanges phosphates on the surface of the material. To access the phosphate groups in the interlayer space, the material must first be exfoliated, which can be done in a 1:1 water–acetone mixture.259 Contacting exfoliated g-ZrP with a solution of a bisphosphonic acid results in each phosphonate group replacing dihydrogen phosphate on two different layers, covalently bonding them together. The reaction may be considered to be an anion replacement as in eqn [1].   ZrðPO4 Þ O2 PðOHÞ2 þ O2 PðOHÞR ! ZrðPO4 Þ½O2 PðOHÞR þ O2 PðOHÞ2 [1] The resulting material now consists of layers of g-ZrP pillared by the organic group of the bisphosphonic acid. Small monophosphonate spacer groups can also be included to decrease the density of pillars between the layers and increase the porosity.260

5.08.2.5.1 Structure of topotactically exchanged g-forms of Zr and Ti phosphate Phosphonic acids can replace the two bridging atoms by donation of a proton to form phosphoric acid and bonding across the vacated four valent metal atoms. In the topotactic incorporation of phenylphosphonic acid into g-ZrP, the group being displaced is O2P(OH)2. The phosphonic acid donates a proton to the leaving group with formation of H3PO4. The displacing group is phenylphosphonic acid, and the phenyl group and one OH group remain in the interlayer space as shown in Figure 29.261 In order to access the interlayer space topotactically, the layers must be exfoliated as indicated above. As depicted in Figure 29, the moiety incorporated between the layer is the O ¼P–R group; in this case, the phenyl ring is R. We note that only every other group is occupied by the phenyl ring. The reason for this peculiar behavior results from the

Porous Pillared Clays and Layered Phosphates

201

1320 pm 664 pm

539 pm

Figure 29 Idealized schematic representation of a g-ZrP topotactically exchange with phenylphosphonate filling 50% of the available sites. Reproduced from Alberti, G. In Comprehensive supramolecular chemistry; Lehn, J.-M., Alberti, G., Bein, T., Eds. Pergamon-Elsevier: Oxford, 1996; p 111, with permission.

closeness of the space available. As shown in the figure, the distance between phosphorus atoms in the b-axis direction is 6.63 A˚ and in the a-axis direction is 5.39 A˚.261 The free space is 35.7 A˚2. Therefore if the ingoing group is bulky, steric hindrance may occur and only a portion of the sites can be occupied.261 Conversely, the composition can be controlled by the amount of ingoing groups added in the reaction.

5.08.2.5.2 Examples of pillaring g-M(IV) phosphates by bisphosphonic acids The pillaring of g-ZrP with bisphosphonic acids is usually affected by the exfoliation of ZrP which is then reacted with a solution of the bisphosphonic acid. Gradually, the surface O2P(OH)2 groups are replaced by the phosphonic acids, while the layer structure is maintained. At low concentrations of bisphosphonic acid, crosslinking of the layers occurs, and if the pillars are not too densely packed between the layers, a porous material results. However, at high concentrations of bisphosphonic acid, a bilayer forms and the layers are not crosslinked, but held together by hydrogen bonding between the pendant phosphonic acid groups. An example is provided with monophenyl diphosphonic acid to form Zr(PO4)0.56(HO3PC6H4PO3H)0.222.2H2O.262 A series of compounds with the general formula Zr(PO4) (H2PO4)1x(HO3PC12H8PO3H)x/2nH2O were prepared by Alberti and coworkers.263 They found that solid solutions occurred when x varied from 0.1 to 0.58. The maximum surface area obtained was 320 m2 g1 for the compound which was 25% pillared. The compounds with greater extent of pillaring showed decreased surface area due to the increased filling of the space between the layers. The pores are small with an average diameter of 5.8 A˚. A functional pillar, 5,50 -bis(dihydroxyphosphoryl)-2,20 bipyridine, was used to prepare two porous derivatives of g-ZrP. The first, pillared at 20% of the theoretical capacity, was predominantly mesoporous with a surface area of 60 m2 g1.

The second material, pillared at 50% of theoretical capacity, had a surface area of 380 m2 g1 with pore diameters centered at 5 A˚. The capability of the 2,20 -bipyridine groups to chelate metals was shown by the quantitative uptake of Fe2þ and Cuþ. The ability of these compounds to chelate metals may make them suitable for selective ion exchange or as supports in heterogeneous catalysis.264

5.08.2.5.3 Pillared derivatives of g-ZrP with rigid and flexible bisphosphonic acids Alberti et al. were able to prepare a series of n-alkyl derivatives of general composition ZrPO4[O2P(OH)2]1 – x[O2POH-(CH2)y-O2POH]x/2yH2O, n ¼ 2, 4, 6, 8, 10, 12, 16.265,266 Values of x varied from 0.12 to 1.00. The reactions were topotactic in which the g-ZrP (0.25 g) was dispersed 35 ml of a 1:1 water/acetone mix. The mixture was kept at 80  C for 15–20 min and then the n-alkylamine, also heated to 80  C, was added. The mixture after some time began to flocculate and was kept for 3 days to ensure completion of the reaction. The interlayer spacings of the products fell on a straight line as shown in Figure 30 and the information is outlined in Table 13.266 These compounds when produced with low values of x, for example, 0.1 for the C10 compound exhibited an elastic structure that had an interlayer distance of 19.4 A˚ when filled with solvent but on drying shrunk to a value of 14.9 A˚. This phenomenon was termed the accordion effect and pictured as a crumpling of the chains. It was suggested that these compounds can be seen as molecular vessels in which selective reactions could be carried out. However, Clearfield and Wang pictured the behavior as a sliding of the layers creating a more slanted positioning of the chains.267 In fact, Clearfield et al. had earlier prepared a series of polyether and polyimine diphosphonates of a-ZrP that exhibited the ‘accordion effect.’ This is discussed later in this chapter.

202

Porous Pillared Clays and Layered Phosphates

3.0

d (nm)

2.0

1.0

0.0 0

2

4

6

8

10

12

14

16

18

n Figure 30 Plot of d-spacing versus n in g-ZrP alkyl-pillared materials where n ¼ alkyl chain length. Reproduced from Alberti, G.; Vivani, R.; Murcia Mascaro´s, S. J. Mol. Struct. 1998, 470(1–2), 81–92, with permission.

Table 13 Composition and interlayer spacing of alkyl-pillared g-zirconium phosphonates Formula

Interlayer spacing (A˚)

Zr(PO4)(HO3PC4H8PO3H)0.5H2O Zr(PO4)(H2PO4)0.8(HO3PC10H20PO3H)3H2O Zr(PO4)(H2PO4)0.76[HO3PCH2(CH2OCH2)3CH2]0.12H2O Zr(PO4)(H2PO4)0.76[HO3PCH2(CH2OCH2)6CH2]0.12H2O

13 17.4 15.7 17.6

Exfoliated g-ZrP was treated with a polyether bisphosphonic acid, resulting in the material with the formula ZrPO4 (O2P(OH)2)0.74(O2P(OH)-R-(OH)PO2)0.13nH2O, R¼ CH2CH2 (OCH2CH2)5OCH2CH2.268,269 Treatment of this material with methylamine and then hypophosphorous acid resulted in the complete substitution of hypophosphite groups for the phosphate groups. The d-spacing was  18 A˚, indicating that the large bisphosphonate ligand was crosslinking the layers of the material. Presumably, the hypophosphite spacer groups reside in between the crosslinking bisphosphonate units to create pores between the layers, but porosity measurements were not carried out. Brunet and coworkers found similar compounds, investigated the arrangement of 1,10-decanebis(phosphonic acid) between the layers of exchanged g-ZrP.270 Porosity was not measured, but this compound picks up amines by swelling, which increases the interlayer distance. We have already described the topotactic exchange of phenylphosphonic acid by g-ZrP. It was reasonable to expect that if a rigid diphosphonic acid were exchanged in this manner that a pillared structure would result. Alberti et al. had prepared a monophenylbis(phosphonic acid) derivative which showed very little to no surface area, but was able to determine that the a and b axes were very close to that of g-ZrP and, therefore, the inorganic backbone was still g-type.271 With this knowledge, he prepared a 4,40 -biphenylbis(phosphonic acid) derivative

with varying percent loadings of the bisphosphonic acid. The topotactic insertion of this ligand resulted in a maximum of 320 m2 g1 surface area with a volume of micropores of 0.12 cm3 g1 at 25% coverage of the sites. The pore size average was about 5.8 A˚ diameter.263 Brunet et al. investigated the exchange of 4,40 -terphenylbis (phosphonic acid) into g-ZrP to create a porous material.269 The exchange was performed in a water/acetone mixture to ensure that the g-ZrP was exfoliated and the bisphosphonate was dissolved. After refluxing for 4 days, a material with a surface area of 90 m2 g1 was obtained. After exchanging the residual phosphate groups for phosphite groups by treatment with hypophosphorous acid, a dramatic increase in surface area, to 400 m2 g1, was observed. The d-spacing decreased from 24.3 to 21.1 A˚ as a result of this treatment. This decrease, along with SS MAS 31P NMR, led the researchers to the conclusion that treating the material with hypophosphorous acid caused a structural rearrangement within the layer, During this process, the layer isomerized from a gamma-type structure to an a-type. The material was found to take up hydrogen to 74 cm3 g1 at 650 torr. Brunet and coworkers further investigated this isomerization using phosphoric acid and methylphosphonic acid.272 They found that the rigidity of the biphenyl or terphenyl ligands was a key factor influencing the conversion to the a-type layer structure. Various compounds were obtained with surface areas up to 568 m2 g1. The authors prepared the Liþ exchanged versions of these compounds and evaluated their hydrogen sorption capabilities. They observed that Liþ exchange resulted in a significant decrease in surface area (as determined by N2 sorption), but it did not noticeably affect the hydrogen sorption properties. The same group has continued their investigation into these types of materials by extending the organic ligands by inserting ethynyl groups between the phenyl rings. This can be considered analogous to the isoreticular extension of ligands in MOFs.273 These materials were prepared with surface areas up to 339 m2 g1. Interestingly, the gamma to alpha isomerization does not occur in the polyphenylethynyl derivatives. This may be a consequence of the higher concentrations of pillar groups found in the polyphenylethynyl derivatives compared to the polyphenyl diphosphonates. Additional information on these types of phosphonates can be found in a recently released book.274

5.08.2.6 Direct Pillaring Using Phosphonic Acids with a-M(IV) Phosphates, UMOFs Crystalline porous phosphonates (phosphonate MOFs) have recently been the subject of a review.275 The poorly crystalline porous phosphonates are known as unconventional MOFS, or UMOFs. These materials are porous and thermally robust, but lacking in long-range order. Many M(IV) bisphosphonate compounds, which can be considered as inorganic layers pillared by organic groups, are UMOFs. In this section we describe these types of compounds. In contrast to the involved g-ZrP synthesis, the a-type derivatives can be synthesized directly by addition of the desired phosphonic acid to a soluble Zr, Ti, Sn four-valent compound followed by refluxing or hydrothermal or solvothermal

Porous Pillared Clays and Layered Phosphates

treatment. It is well at this point to describe the structure of zirconium phenylphosphonate (Zr(O3PC6H5)2). It has a layered structure very similar to that of a-ZrP.188 The difference lies in the fact that the space group is C2/c, which places the Zr4þ ion in the mean basal plane of the layer as opposed to above and below the plane as in a-ZrP. This has the effect of tilting the phenyl rings about 30 to the axis perpendicular to the layer (c-axis). The phenyl rings at 5.3 A˚ apart allowing for ring p–p interactions. It stands as a model to understand certain structural considerations relevant to the noncrystalline UMOFs. In Dines et al. methods of pillaring by bis(phosphonic acids) were described.276 They emphasized the use of adding a small ligand along with the larger pillaring ligand as being largely responsible for the porosity. However, Clearfield277 observed that their compound containing no small spacer group had a surface area of 316 m2 g1 as opposed to the best mixed derivative with a surface area of 390 m2 g1. They showed that synthesis in water resulted in a broad range of pores with almost no microporosity; however, when carrying out the reactions solvothermally in dimethylsulfoxide (DMSO) when using Zr4þ and alcohol or a mixture of alcohol and water for Sn4þ, greater regularity in pore sizes is observed, more so in the case of the tin compounds. An excellent summary of these types of reactions and products is provided in the book Metal Phosponate Chemistry: From Synthesis to Applications.274 Major features of these compounds are the high porosity ranging from 250 m2 g 1to more than 500 m2 g1 and the fact that they are natural nanoparticles. Interestingly, the particle size decreases from about 80–90 nm when only a single ligand such as biphenylbisphosphonic acid is used to about 25 nm or below when the spacer ligand such as methylphosphonic acid is incorporated into the structure. These materials have been termed UMOFs for unconventional metal–organic frameworks because of the particle size and the fact that they are almost totally amorphous.

5.08.2.6.1 M(IV) bisphosphonates In work published in the early 1980s,276 Dines et al. outlined three synthetic pathways for generating porous solids by using bisphosphonic acids to crosslink the layers of tetravalent phosphates. One of these was the topotactic exchange reaction of a bisphosphonic acid with exfoliated g-ZrP, as detailed in Section 5.08.2.5. Another method outlined was the preparation of a mixed–ligand-pillared compound using both bisphosphonate and bisphosphate ligands, which was subsequently subjected to hydrolysis to remove the organic component of the bisphosphate ligands, leaving phosphate groups. The bisphosphonate ligands are more robust and survive the hydrolysis step with P–C bonds intact, serving as pillars to hold the layers apart. This results in compounds of the general formula M(O3POH)x(O3P-R-PO3)2–x/2. The third method presented was the coprecipitation of a tetravalent metal with both phosphoric acid and a bisphosphonic acid. The authors noted that this method produced poorly crystalline products which were difficult to characterize. Although this observation is, in general, true, it should be noted that many M(IV) phosphate–bisphosphonate compounds have since been synthesized in this manner and have been shown to be useful materials due to their high surface areas and catalytic

203

properties. Some informative reviews are available.267,278,279 It is these types of compounds that we will focus on for the remainder of this section.

5.08.2.6.2 Mixed ligand-pillared materials The coprecipitation method is useful in the preparation of porous compounds from a bisphosphonate crosslinking ligand and a smaller, monophosphonate spacer ligand. The spacer ligand can be varied to control the nature of the chemical environment within the pores. For example, phosphoric acid can be used to impart acid properties to the material, or methylphosphonic acid can be used to create a more hydrophobic region. Also, phosphonate ligands with additional chemical moieties such as amines or carboxylic acids can be used to impart new functional capabilities to the material. The crosslinker bisphosphonate ligand can be tailored to influence the size of the pores obtained, for example, biphenylbis(phosphonic acid) would give a larger interlayer distance, and hence, larger pore sizes than 1,4-phenylbis(phosphonic acid). The use of rigid crosslinker groups is not required, and more flexible alkylbis(phosphonic acid) results in materials in which the interlayer distance can expand to accommodate solvent molecules in the interlayer region.280,281 The crosslinking bisphosphonates can also be functionalized to impart chemical reactivity. With any of these systems, there is a particular ratio of spacer to crosslinker which results in the maximum surface area. For the phosphite–biphenylbisphosphonate system explored by Dines et al.,282 the maximum surface area of 481 m2 g1 was obtained when 50 mol% biphenyl pillar was used. Lower amounts of pillar ligand resulted in decreased surface area. However, they reported the interesting result that even the material which contained no phosphite spacer, the pure Zr(O3PC12H8PO3), had surface area of 316 m2 g1. This was unusual because, according to the structural model, this structure would be dense and have only external surface area comparable to unpillared ZrP of similar particle size. Later, this was further explored by Clearfield et al.283 Their work will be discussed in more detail after first describing a variety of preparations by other workers. Alberti et al.284 used a clever strategy to evenly space the pillar ligands to obtain more uniform pores. The ligand 3,30 ,5,50 -tetramethylbiphenylbis(phosphonic acid) was used as a pillar ligand, and hypophosphorous acid was used as a spacer. The methyl groups on the pillar ligand do not allow it to take adjacent positions on the surface of the layer. The phosphite groups were small enough to fill in the positions between the pillars without causing steric interference. The material Zr(HPO3)1.34(O3PRPO3)0.33 was obtained. BET analysis showed that the surface area was 375 m2 g1, with a narrow pore-size distribution centered at 6 A˚. This is in good agreement with the structural model presented. Mutin and coworkers utilized Zr(OC3H7)4 to prepare a series of compounds using biphenylbisphosphonate as the pillar and phosphate as a spacer ligand.285 Surface area up to 307 m2 g1 was obtained, with pore-size distributions centered at  6 A˚. The particles were 10–20 nm in diameter, which resulted in a significant contribution of mesopores to the total porosity. Derouane and Jullien-Sardot prepared a series of alkylbisphosphonate-pillared Zr phosphonates with phosphoric acid as a spacer ligand.286 The ratio of phosphoric acid to

204

Porous Pillared Clays and Layered Phosphates

bisphosphonic acid was varied from 0 to 10. Methanol was used as a solvent, with HF as a solubilizing agent. As the amount of phosphoric acid was increased, the PXRD peaks became broader, indicating a lower degree of order and decreased particle size. The highest surface area (236 m2 g1) was obtained at a ratio of 0.75 phosphoric acid to 1 bisphosphonic acid for the C8 compound. Upon sorption of small amines the d-spacing increased from 13.9 to 15.5 A˚, indicating that the chains are tilted in the dry compound and become more perpendicular to the layers upon sorption of the amine. It is interesting to note that Alberti et al. prepared a series of alkylbisphosphonate a-ZrP derivatives in which phosphite was used as a spacer group. These compounds had surface areas of 230–400 m2 g1 but the pore sizes ranged from 4 to 14 nm. The formation of this interparticle mesoporsity was attributed to edge-to-edge (or house of cards) interactions between rigid particles containing only a few layers.287 It is interesting to note that Alberti et al. were also able to obtain a crystalline compound for the butyl derivative at almost 100% coverage of the bonding sites.266 The structure is shown in Figure 31. Similarly, Clearfield and coworkers prepared a series of polyether and polyiminediphosphonates that exhibited the ‘accordion’ effect mentioned earlier in the gamma-type materials. This is shown in Table 14 with the interlayer spacings in the anhydrous and wet states. In today’s parlance they should be considered as breathing molecules.267 The fact that the zirconium biphenylbisphosphonate prepared by Dines et al. was highly porous, even when no spacer molecule was present, prompted Clearfield to reexamine the synthesis. It was found that reaction of 4,4’-biphenylbis(phosphonic acid) with ZrOCl2∙8H2O, with or without the presence of HF, produced porous products.278 However, there was no regularity to the pore structure. The use of DMSO or DMSO-ethanol in solvothermal reactions at temperatures of 80–120  C did improve the situation.37 Interestingly, the products are nanosized particles (75–100 nm).

Electron micrographs reveal the layered nature of the particles and addition of a spacer group such as –PH, –POH, –CH3 as phosphonic acids reduces the particle size in proportion to the amount incorporated while increasing the surface area.275,288 Surface areas of 300–500 m2 g1 are common with a group of pores of less than 10 A˚ size as demonstrated by the height of the N2 uptake at 103 torr and larger pores up to 20 A˚. A typical X-ray pattern is shown in Figure 32. The first peak is at 13.6 A˚ and represents the interlayer distance. This distance is slightly smaller than the interlayer distance of zirconium phenylphosphonate, Zr(O3PC6H5)2 (14.7 A˚), where the interlayer space contains a bilayer of phenyl groups. Thus, the picture we derive is that the biphenylbisphosphonate PO3 groups each bond to three Zr ions through the phosphonate oxygen atoms in adjacent layers, thereby crosslinking the layers. However, such a structure does not account for the porosity as the pillars would fill all the space, being only 5.3 A˚ apart. The pore sizes often exceed the interlayer spacing; the use of phenyl rings as spacers does not increase the interlayer distance and other phenomena indicate that the pores result from the large numbers of defects within the structure. A tentative model is shown in Figure 33. This model attributes the defects as being derived from the uneven growth of the layers.279 The low solubility of the compounds prevents them from dissolving once they have precipitated, precluding Ostwald ripening. This results in a rapidly formed, poorly ordered compound with significant internal porosity.

Table 14 Polyethylene glycol esters of zirconium diphosphate Zr[O3PO(CH2CH2O)nPO3]x[HPO4]2x Interlayer spacing (A˚)

Elemental analysis

x (from)

n

%C

%H

EA

TGA

Anhydrous

wet

2 3 4 9 13 22

13.56 15.16 21.04 22.81 37.97 35.56

2.17 3.14 3.7 3.96 6.32 6.24

0.96 0.83 0.97 0.73 1.02 0.83

0.96 0.87 1 0.71 1 0.87

12.3 14 16.1 20.1 33.3 >44

15.5 16.1 17.2 21.3 >44 >44

600 d = 13.62 Å

P Zr 0.92 nm

O

Counts

1.32 nm C

400

d = 4.356 Å

200

d = 6.857 Å

P

0 2

Figure 31 Structure of a ZrP-crosslinked material with 1,4-butanebis (phosphonic acid). Reproduced from Alberti, G.; Murcia-Mascaro´s, S.; Vivani, R. J. Am. Chem. Soc. 1998, 120(36), 9291–9295, with permission.

10

20

30

2 Theta (degrees)

Figure 32 PXRD pattern of a zirconium biphenylbisphosphonate material. Reproduced from Clearfield, A. Chem. Mater. 1998, 10(10), 2801–2810, with permission.

40

205

Porous Pillared Clays and Layered Phosphates

O O O P H

O OH HO P H

O O O P

H

R P O O O

P O O O

HH H

H H

H H H

H HH

H

HH H H H

O OH HO P H H

H H H

H

P O O O O O O P

H

HH H H H

P O O O O O O P R

P O O O

H

P O O O

H HH H

P O O O O O O P H

P OH HO O

P O O O

H

H HH

O OH HO P H

H

H

H H

H

P O O O

HH H

H H

H R P O O O

O O O P

H

H

H

P O O O

O O O P

P O O O

H

H H H

H H

H H H

H H

P O O O

H H

H H H

H

P O O O

P O P O H O O O O O O O P H H H H

HH

H H

P O O O O O O P H

HH H

H

H H

H H H

H H

H H

H

H H H

H

H

R H

H H

H H

P O O O O O O P

H

H H H

HH H

H H

H H

H P O H P O P O O O O O O O O O O O O O O P O P O P H H H

H H

H H H HH

H

H

H H

H H

H H H

H

H H H

HH H

P O O O O O O P R

H

H

P O O O O O O P H

H HH H

P O O O

H H

O O O P

H

R

O O O P

H H

H H H

P O H P O O O O O O O O O O P O P H H

O O O P R

H

O O O P HH H

H H

H H

P O O O

H

H

H H

H H H

H

O O O P

H

O O O P

H H H H H

H H

O OH HO P H

H

H

H H H H

H

O O O P

O O O P R

H

H H

P O O O

Figure 33 Schematic representation of the formation of pores inside a biphenyl-pillared zirconium phosphonate. Reproduced from Gagnon, K. J.; Perry, H. P.; Clearfield, A. Chem. Rev. 2011, 112(2), 1034–1054, with permission.

The solubility of these compounds can be increased by adding HF to the reaction mixtures. The formation of the ZrF62 ion allows for some degree of recrystallization to occur, resulting in more crystalline products which can have narrower pore-size distribution. However, increasing the solubility in this manner also limits the composition of the materials, as high amounts of spacer ligand in the reaction mixture with HF can result in the formation of two separate phases. Without HF and at low temperature, solid solutions with the full range of compositions can typically be prepared.

5.08.2.6.3 Functionalization of UMOFs There is a great deal of potential in the development of these UMOFs. A large array of aryl or combination alkyl–aryl chains may be used as pillars. Also, a number of spacer groups may be utilized. Insertion of phosphate groups provides ion-exchange capabilities to the UMOFs. The use of 2,20 -bipyridylbis(phosphonic acid) provides the ability to complex and potential catalytic qualities.288 The phenyl rings of Zr(O3PC6H5)2 have been sulfonated by several methods.37,289,290 The phenyl groups can be sulfonated either by direct contact with fuming sulfuric acid289 or by exposure to gaseous SO3.290 The sulfonated materials were

found to have acid strength (as determined by the 13C chemical shift of acetone when exposed to the material) comparable to pure sulfuric acid. The aromatic rings of Zr(O3PC12H8PO3) have been sulfonated with more than one SO3H group per ring. This created a very strong Brønsted acid that has value as a catalyst.121 In addition, this material should serve as a facile proton conductor, if we judge from the fact that sulfonated zirconium and titanium phosphonates exhibited conductivities of 102–101 O1 cm1 at 85% relative humidity.291 The pillared sulfonated compounds are extremely hydrophilic filling the pores with water on standing in air. A TGA curve of such a compound of composition Zr(O3PC12H8PO3) (SO3)2.2∙5.8H2O is shown in Figure 34.279 Most of the water (15.3%) is lost between 25 and 140  C, but all of the sulfonate groups are still in place to beyond this temperature. It is then likely that this compound, enclosed in a membrane such as Nafion, may retain high proton conductivity above 120  C and thus be very attractive as a fuel cell membrane. A more direct route to functional M(IV) phosphonate derivatives is achieved by the use of bisphosphonate ligands which are covalently linked to additional functional groups. The technique of using a rigid diphosphonate pillar and a

206

Porous Pillared Clays and Layered Phosphates

100 90

15.49%

80

11.10%

Weight (%)

70 60 34.18%

50 40 30 20 0

100

200

300

400

500

600

700

800

900

1000

Temperature (°C) Figure 34 TGA of sulfonated zirconium biphenylbisphosphonate. Reproduced from Gagnon, K. J.; Perry, H. P.; Clearfield, A. Chem. Rev. 2011, 112(2), 1034–1054, with permission.

small phosphonate spacer group has been shown to be an effective means for enhancing porosity in M(IV) phosphonates. Massiot et al. reported the first efforts in this area synthesizing two compounds which covalently incorporated the 2,20 -bipyridyl group.292 One was a derivative of g-ZrP which was prepared by a topotactic exchange reaction of 2,20 -bipyridinediyl-5,50 -bis(phosphonic acid) with exfoliated g-ZrP. The other, an a-ZrP derivative, was synthesized solvothermally in a mixture of DMSO and HF. Both materials had a surface area of 330 m2 g1, but the g-ZrP derivative had smaller pores (5 A˚) with a narrower pore-size distribution. Upon reacting the materials with aqueous solutions of FeII or CuI, it was observed that only a fraction of the bipyridyl sites were coordinating metal ions in the purely microporous g-ZrP phase, while the a-ZrP derivative, with its larger pores, exhibited the maximum uptake. The poor loading of the former was most likely a consequence of the metal ions blocking the small pore channels, preventing access to the bipyridyl sites in the interior of the material.292 Intrigued by the potential applications of such stable, porous materials which could be used to sequester soft metals, researchers in Clearfield’s group undertook a systematic investigation of porosity in the hydrothermally synthesized a-type materials with both Zr and Sn4þ. As in the work just described, the pillar ligand was 2,20 -bipyridinediyl-5,50 -bis(phosphonic acid), but methylphosphonic acid was chosen as a spacer group to avoid the potential reactivity of the phosphite while still using a small phosphonate which would not occupy too much space between the pillar groups. It was observed that for both the Sn and Zr materials, increasing the ratio of spacer to pillar, up to 6:1, resulted in a concomitant increase in the surface area. As Dines discovered many years before, even the materials which contained no spacer groups still exhibited appreciable surface areas, in excess of 320 m2 g1. Remarkably, the Sn material with six spacer ligands for every bipyridyl pillar had a surface area of 515 m2 g1, with 77% micropores. This

20 nm Figure 35 TEM image of zirconium 2,20 -bypyridine-4,40 bisphosphonate with reduced 2–4 nm Pd nanoparticles distributed throughout the matrix. Reproduced from Perry, H. P.; Law, J.; Zon, J.; Clearfield, A. Microporous Mesoporous Mater. 2011, 149, with permission.

high ratio of spacer to pillar resulted in two separate phases of the Zr compounds, but even the 2:1 compound had a surface area of 528 m2 g1 with 93% micropores. These materials were loaded with Pd(O2CCH3) which was reduced to form 2–4 nm Pd0 particles which were trapped by in the phosphonate matrix as shown in Figure 35. This process of loading and reduction was demonstrated to be repeatable, which may be useful in the preparation of supported bimetallic or core-shell nanoparticles.288

5.08.2.6.4 Selective ion exchange for nuclear waste streams A number of years ago Clearfield et al. prepared mixed derivatives of monophenyl and biphenyl diphosphonic acids of Zr4þ with added phosphoric acid. This synthesis introduced

Porous Pillared Clays and Layered Phosphates

Table 15

207

N2 sorption surface area data for Zr and Sn hybrids determined from BET and percent microporosity by the t-plot method Zr inorganic–organic hybrid materials

Sn inorganic–organic hybrid materials

Synthesis temperature ( C)

Total surface area (cc g )

Microporous (%)

Total surface area (cc g1)

Microporous (%)

80 105 120 145 160 180 200

547.8 453.3 452.3 443.8 342.3 473.3 –

55.77% 56.25% 71.48% 44.95% 62.78% 42.76% –

– – 270.6 268.6 290.4 339.2 342.2

– – 98.62% 95.20% 96.49% 95.34% 94.36%

Table 16

1

Kd values observed at pH 3 in HNO3 and reported in ml g1 for cations of increasing charge

Sample

Naþ

Csþ

Ca2þ

Sr2þ

Nd3þ

Sm3þ

Ho3þ

Yb3þ

H-Zr Na-Zr H-Sn Na-Sn

32 27 – <1 –

450 30 3700 310 130 14 260 57

190 130 780 180 18 13 270 100

650 73 12 000 140 4 650 18

29 000 2000 1 900 000 340 000 170 000 300 000 17 000

80 000 10 000 1 300 000 380 000 320 000 16 000 300 000 13 000

110 000 10 000 450 000 110 000 320 000 14 000 220 000 14 000

90 000 13 000 450 000 12 000 160 000 91 000 220 000 13 000

phosphate groups with three oxygen atoms bonded to the metal ion and P–OH groups pointing into the attendant pores.293 Both types of compounds were porous and were ion exchangers by virtue of the phosphate proton being exchangeable. As an interesting facet of their behavior, the selectivity for ions increased as the charge on the ion increased and this was more pronounced in the monophenyl derivatives. One of the interesting applications of these materials is that involved in potential nuclear separations. In processes to recover the useable fuel from the current spent rods, the waste materials Csþ, Sr2þ, and lanthanides need to be removed from the actinides. Methods for removal of cesium and strontium are well developed but the separation of lanthanides from actinides is still problematic. While the oxidation state of lanthanides is as plus three ions, that of actinides is variable. Thus by manipulating the oxidation state of the actinides so as to leave them with a plus one or two state, it should be possible to separate the two groups. Two sets of compounds were prepared with Zr4þ and Sn4þ as the metal ions and phosphoric acid or Na3PO4 to supply the spacer ion-exchange groups. The general procedure was to dissolve a mixture containing 0.5 mmol of C6H4(PO3H2)2 and 1 mmol of H3PO4 in 3 ml of ddi H2O in a Teflon-lined steel acid digestion vessel. A solution of 0.5 M [M4þ] in ddi H2O was prepared ahead of time. While stirring, 2 ml of the metal solution was added to the phosphate/phosphonate mixture dropwise. Immediately after the metal addition, the stir bar was removed and the reaction vessel was sealed and placed in a constant temperature oven at the desired temperature for the desired time. The reactions that started from Na3PO4 were carried out using the same method, but the sodium phosphate was substituted for the phosphoric acid. The resulting product was a white paste that ground easily into a white powder. The surface areas and percent microporosity are provided in Table 15.294 Some Kd values are collected in Table 16. Due to the very great difference in the Kd values for 3þ and 1þ species, it is likely that the actinides may be separated from lanthanides using these monophenyl UMOFs.

5.08.2.6.5 Future outlook for organically pillared phosphonates, UMOFs It is our opinion that these UMOFs are worthy of a concerted effort to obtain further structural data. For example, the use of EXAFS coupled with atomic pair distribution would reveal the coordination of the metal and the outstanding inter-atomic distances. This can be followed by obtaining the same spectra with phosphate– phosphonate solid solutions followed by ion exchanging a second cation with ion charges to see how these ions arrange themselves within the compounds. Electron diffraction examination may also prove to be fruitful in fixing some short-range order and the nature of defects. Furthermore, the shapes and types of pores or cavities can be revealed by phase contrast neutron scattering, positron annihilation spectroscopy,46,47 and isotherms at lower pressures such as 107 torr. This structure will then form the basis of comprehending the chemical behavior of the UMOFs. For related chapters in this Comprehensive, we refer to Chapters 5.02, 5.09, 7.10, and 7.11.

Acknowledgments The authors would like to thank the following funding organizations for their continued support for some of the work presented herein as well as the authors themselves: Department of Engery Basic Energy Science under grant DEFG02-O3ER15420, National Science Foundation under grants DMR-0652166 and DGE-0750732 (K.J.G.), the Robert A. Welch Foundation under grant A-0673, and Savanah River National Lab under grant AC70059-0.

References 1. Brown, G.; Nadeau, P. Phil. Trans. Roy. Soc. Lond. Math. Phys. Sci. 1984, 311(1517), 221–240. 2. Mott, C. J. B. Catal. Today 2008, 2(2–3), 199–208. 3. Bortun, A. I.; Bortun, L. N.; Khainakov, S. A.; Clearfield, A. Solvent Extr. Ion Exch. 1998, 16(4), 1067–1090.

208

Porous Pillared Clays and Layered Phosphates

4. Van Olphen, H.; Fripiat, J. J. Data Handbook for Clay Materials and Other NonMetallic Minerals. Pergamon Press: New York, 1979; p. 355. 5. Carrado, K. A. In Handbook of Layered Materials; Auerbach, S. M., Carrado, K. A., Dutta, P. K., Eds.; Marcel Dekker: New York, 2004; pp 1–37. 6. Yariv, S.; Michaelian, K. H. In Organo-Clay Complexes and Interactions; Yariv, S., Cross, H., Eds.; Marcel Dekker: New York, 2002; pp 203–216. 7. Rutherford, D. Clays Clay Miner. 1997, 45(4), 534. 8. Clearfield, A. In Advanced Catalysts and Nanostructured Materials; Moser, W. R., Ed.; Chapter 14, Academic Press: San Diego, 1996 pp 345–394. 9. Ryland, L. B.; Tamele, M. W.; Wilson, J. N. In Catalysis: Oxidation, Hydration, Dehydration and Cracking Catalysts; Emmett, P. H., Ed.; Reinhold Publishing: New York, 1960. 10. Occelli, M. L. In Expanded Clays and Other Microporous Solids; Occelli, M. L., Robson, H., Eds.; Van Nostrand Reinhold: New York, 1992. 11. Davis, M. E. Acc. Chem. Res. 1993, 26(3), 111–115. 12. Vaughan, D. E. W. Catal. Today 1988, 2(2–3), 187–198. 13. Vaughan, D. E. W.; Lussier, R. J.; Magee, J. S., Jr. Pillared Interlayered Clay Materials Useful As Catalysts and Sorbents. U.S. Patent: 4,176,090, 1979 14. Lahav, N.; Shani, V.; Shabtai, J. Clays Clay Miner. 1978, 26(2), 107. 15. Clearfield, A.; Kuchenmeister, M. In: Bein, T., Ed.; Supramolecular Architecture American Chemical Society: Washington, DC, 1992; vol. 499 pp 128–144. 16. Butruille, J. R.; Pinnavaia, T. J. In Comprehensive Supramolecular Chemistry; Lehn, J.-M., Alberti, G., Bein, T., Eds.; Pergamon-Elsevier: Oxford, 1996. 17. Corma, A. Chem. Rev. 1997, 97(6), 2373–2420. 18. Yang, R.; Cheng, L. In Pillared Clays and Ion-Exchanged Pillared Clays as Gas Adsorbents and as Catalysts for Selective Catalytic Reduction of NOx Access in Nanoporous Materials; Pinnavaia, T., Thorpe, M., Eds.; Springer: US, 2002; pp 73–92. 19. Bottero, J. Y.; Cases, J. M.; Fiessinger, F.; Poirier, J. E. J. Phys. Chem. 1980, 84(22), 2933–2939. 20. Akitt, J. W.; Farthing, A. J. Chem. Soc. Dalton Trans. 1981, (7), 1606–1608. 21. Bottero, J. Y.; Axelos, M.; Tchoubar, D.; Cases, J. M.; Fripiat, J. J.; Fiessinger, F. J. Colloid Interface Sci. 1987, 117(1), 47–57. 22. Fu, G.; Nazar, L. F.; Bain, A. D. Chem. Mater. 1991, 3(4), 602–610. 23. Nazar, L. F.; Fu, G.; Bain, A. D. J. Chem. Soc. Chem. Commun. 1992, (3), 251–253. 24. Tomlinson, A. A. G. J. Porous Mater. 1998, 5(3), 259–274. 25. Vaughan, D. E. W.; Lussier, R. J. In 5th International Conference on Zeolites, Naples, Italy, Heyden, Naples, Italy, 1980. 26. Pinnavaia, T. J.; Tzou, M.-S.; Landau, S. D.; Raythatha, R. H. J. Mol. Catal. 1984, 27(1–2), 195–212. 27. Clearfield, A.; Aly, H. M.; Cahill, R. A.; Serrette, G. P. D.; Shea, W. L.; Tsai, T. Y. In Studies in Surface Science and Catalysis; Tadashi, H., Tatsuaki, Y., Eds.; Elsevier: Amsterdam, 1994; vol. 83, pp 433–442. 28. Occelli, M. L.; Tindwa, R. M. Clays Clay Miner. 1983, 31, 22–28. 29. He, M.-Y.; Liu, Z.; Min, E. Catal. Today 1988, 2(2–3), 321–338. 30. Matsuda, T.; Asanuma, M.; Kikuchi, E. Appl. Catal. 1988, 38(2), 289–299. 31. Tennakoon, D. T. B.; Thomas, J. M.; Jones, W.; Carpenter, T. A.; Ramdas, S. J. Chem. Soc. Faraday Trans. 1986, 82(2), 545–562. 32. Molina, R.; Schutz, A.; Poncelet, G. J. Catal. 1994, 145(1), 79–85. 33. Moreno, S.; Sun Kou, R.; Poncelet, G. J. Catal. 1996, 162(2), 198–208. 34. Xu, T.; Munson, E. J.; Haw, J. F. J. Am. Chem. Soc. 1994, 116(5), 1962–1972. 35. Xu, T.; Torres, P. D.; Beck, L. W.; Haw, J. F. J. Am. Chem. Soc. 1995, 117(30), 8027–8028. 36. Haw, J. F.; Nicholas, J. B.; Xu, T.; Beck, L. W.; Ferguson, D. B. Acc. Chem. Res. 1996, 29(6), 259–267. 37. Wang, Z.; Heising, J. M.; Clearfield, A. J. Am. Chem. Soc. 2003, 125(34), 10375–10383. 38. Clearfield, A.; Serrette, G. P. D.; Khazi-Syed, A. H. Catal. Today 1994, 20(2), 295–312. 39. Malla, P. B.; Komarneni, S. Clays Clay Miner. 1993, 41, 472–483. 40. Occelli, M. L.; Innes, R. A.; Hwu, F. S. S.; Hightower, J. W. Appl. Catal. 1985, 14, 69–82. 41. Chevalier, S.; Franck, R.; Suquet, H.; Marcilly, C.; Barthomeuf, D. In Expanded clays and Other Microporous Solids; Occelli, M. L., Robson, H., Eds.; Van Nostrand Reinhold: New York, 1992; p 32. 42. Lambert, J.-F.; Chevalier, S.; Franck, R.; Suquet, H.; Barthomeuf, D. J. Chem. Soc. Faraday Trans. 1994, 90(4), 675–682. 43. Occelli, M. L.; Drake, B.; Gould, S. A. C. J. Catal. 1993, 142(2), 337–348. 44. De Stefanis, A.; Tomlinson, A. A. G. Catal. Today 2006, 114(2–3), 126. 45. De Stefanis, A.; Tomlinson, A. A. G.; Steriotis, T. A.; Stefanopoulos, K. L.; Keiderling, U. Phys. Rev. B 2004, 350(1–3, Supplement), E521–E524. 46. Liu, M.; Wong-Foy, A. G.; Vallery, R. S.; Frieze, W. E.; Schnobrich, J. K.; Gidley, D. W.; Matzger, A. J. Adv. Matzger. 2010, 22(14), 1598–1601. 47. Feldblyum, J. I.; Liu, M.; Gidley, D. W.; Matzger, A. J. J. Am. Chem. Soc. 2011, 133(45), 18257–18263.

48. Occelli, M. L.; Bertrand, J. A.; Gould, S. A. C.; Dominguez, J. M. Microporous Mesoporous Mater. 2000, 34(2), 195–206. 49. Gil, A.; GandI´a, L. M.; Vicente, M. A. Catal. Rev. 2000, 42(1–2), 145–212. 50. Bradley, S. M.; Kydd, R. A.; Yamdagni, R. J. Chem. Soc. Dalton Trans. 1990, (2), 413–417. 51. Bradley, S. M.; Kydd, R. A.; Yamdagni, R. J. Chem. Soc. Dalton Trans. 1990, (9), 2653–2656. 52. Bradley, S. M.; Kydd, R. A. J. Catal. 1993, 142(2), 448–454. 53. Coelho, A. V.; Poncelet, G. In Pillared Layered Structures: Current Trends and Applications; Mitchell, I. V., Ed.; Elsevier Applied Science: London, 1990; p 185. 54. Coelho, A. V.; Poncelet, G. Appl. Catal. 1991, 77(2), 303–314. 55. Gonzalez, F.; Pesquera, C.; Benito, I.; Mendioroz, S.; Poncelet, G. J. Chem. Soc. Chem. Commun. 1992, (6), 491–493. 56. Benito, I.; Pesquera, C.; Blanco, C.; Gonzalez, F. Trends Inorg. Chem. 1996, 4, 93–105. 57. Hernando, M. J.; Pesquera, C.; Blanco, C.; Benito, I.; Gonza´lez, F. Chem. Mater. 1996, 8(1), 76–82. 58. Shabtai, J.; Rosell, M.; Tokarz, M. Clays Clay Miner. 1984, 32, 99–107. 59. Tokarz, M.; Shabtai, J. Clays Clay Miner. 1985, 33, 89–98. 60. Sterte, J. Clays Clay Miner. 1991, 39, 167–173. 61. Trillo, J. M.; Alba, M. D.; Alvero, R.; Castro, M. A.; Mun˜oz, A.; Poyato, J.; Tobı´as, M. M.; Lagaly, G. Solid State Ion. 1993, 63–65, 457–463. 62. Trillo, J. M.; Alba, M. D.; Castro, M. A.; Poyato, J.; Tobı´as, M. M. J. Mater. Sci. 1993, 28(2), 373–378. 63. Mendioroz, S.; Gonzalez, F.; Pesquera, C.; Benito, I.; Blanco, C.; Poncelet, G. In: Guczi, L.,Solmosi, F., Tetenyi, P., Eds.; Studies in Surface Science and Catalysis Elsevier: Amsterdam, 1993; vol. 75, pp 1637–1640. 64. Zhao, D.; Yang, Y.; Guo, X. Inorg. Chem. 1992, 31(23), 4727–4732. 65. Booij, E.; Kloprogge, T.; Van, V. J. A. R. Clays Clay Miner. 1996, 44, 774–782. 66. Wada, S.-I.; Wada, K. J. Soil Sci. 1980, 31(3), 457–467. 67. Sterte, J.; Shabtai, J. Clays Clay Miner. 1987, 35, 429–439. 68. Gil, A.; Guiu, G.; Grange, P.; Montes, M. J. Phys. Chem. 1995, 99(1), 301–312. 69. Fetter, G.; Tichit, D.; Massiani, P.; Dutartre, R.; Figueras, F. Clays Clay Miner. 1994, 42, 161–169. 70. Fetter, G.; Tichit, D.; de Menorval, L. C.; Figueras, F. Appl. Catal. A 1995, 126(1), 165–176. 71. Ashton, A. G.; Atkins, M. P. Pillared Interlayered Clays and Their Uses. WO 8503016A1, 1985. 72. Yamanaka, S.; Brindley, G. W. Clays Clay Miner. 1979, 27, 119–124. 73. Bartley, G. J. J.; Burch, R. Appl. Catal. 1985, 19(1), 175–185. 74. Bartley, G. J. J. Catal. Today 1988, 2(2–3), 233–241. 75. Ohtsuka, K.; Hayashi, Y.; Suda, M. Chem. Mater. 1993, 5(12), 1823–1829. 76. Clearfield, A. Rev. Pure Appl. Chem. 1964, 14(3), 91. 77. Clearfield, A. J. Mater. Res. 1990, 5(01), 161. 78. Clearfield, A.; Vaughan, P. A. Acta Crystallogr. 1956, 9(7), 555. 79. Muha, G. M.; Vaughan, P. A. J. Chem. Phys. 1960, 33(1), 194. 80. Clearfield, A. Inorg. Chem. 1964, 3(1), 146–148. 81. Fryer, J. R.; Hutchinson, J. L.; Peterson, R. J. Colloid Interface Sci. 1970, 34(2), 238. 82. Farfan-Torres, E.; Sham, E.; Grange, P. Catal. Today 1992, 15(3–4), 515. 83. Johnson, J. W.; Brody, J. F.; Alexander, R. M.; Yacullo, L. N.; Klein, C. F. Chem. Mater. 1993, 5(1), 36–42. 84. Toranzo, R.; Vicente, M. A.; Banores-Munoz, M. A.; Gandia, L. M.; Gil, A. Microporous Mesoporous Mater. 1998, 24(4–6), 173. 85. Cool, P.; Vansant, E. F. Microporous Mater. 1996, 6(1), 27. 86. Heylen, I.; Vanhoof, C.; Vansant, E. F. Microporous Mater. 1995, 5(1–2), 53. 87. Cool, P.; Vansant, E. F. In Adsorption Science and Technology, Proceedings of the 2nd Pacific Basin Conference on Adsorption Science and Technology, Brisbane, Australia, World Scientific Publishing, Brisbane, 2000; pp 633–637. 88. Cool, P.; Clearfield, A.; Mariagnanam, V.; McEllistrem, L. J.; Crooks, R. M.; Vansant, E. F. J. Mater. Chem. 1997, 7(3), 443–448. 89. Tanev, P. Science 1995, 267(5199), 865. 90. Tanev, P. T.; Pinnavaia, T. J. In Access in Nanoporous Materials, Proceedings of a Symposium on Access in Nanoporous Materials, East Lansing, Mich; Pinnavaia, T. J., Thorpe, M., Eds.; Plenum, New York, East Lansing, Mich, 1995; pp 13–27. 91. Galarneau, A.; Barodawalla, A.; Pinnavaia, T. J. Nature 1995, 374(6522), 529. 92. Galarneau, A.; Barodawalla, A.; Pinnavaia, T. J. Chem. Commun. 1997, (17), 1661–1662. 93. Vogels, R. J. M. J.; Breuklaar, J.; Kloprogge, J. T.; Jansen, J. B. H.; Geus, J. W. Clays Clay Miner. 1997, 45(1), 1. 94. Galarneau, A.; Barodawalla, A.; Pinnavaia, T. J. In Clays for Our Future, Proceedings of the 11th International Clay Conference, Ottawa, Ontario, ICC97 Organizing Committee, Ottawa, Ontario, 1997; pp 21–29.

Porous Pillared Clays and Layered Phosphates

95. Rouquerol, F.; Rouquerol, J.; Sing, K. S. W. Adsorption by Powders and Porous Solids: Principles, Methodology and Applications. p 467. 96. Gregg, S. J.; Sing, K. S. W. Adsorption, Surface Area and Porosity, 2nd ed.; Academic Press: London, 1982; p 303. 97. Ahenach, J.; Cool, P.; Vansant, E. F. Phys. Chem. Chem. Phys. 2000, 2(24), 5750–5755. 98. Benjelloun, M.; Cool, P.; Van Der Voort, P.; Vansant, E. F. Phys. Chem. Chem. Phys. 2002, 4(12), 2818–2823. 99. Mokaya, R.; Jones, W. J. Catal. 1997, 172(1), 211. 100. Benjelloun, M.; Cool, P.; Linssen, T.; Vansant, E. F. Microporous Mesoporous Mater. 2001, 49(1–3), 83. 101. del Rey-Perez-Caballero, F.; Poncelet, G. Microporous Mesoporous Mater. 2000, 41(1–3), 169. 102. del Rey-Perez-Caballero, F.; Poncelet, G. Microporous Mesoporous Mater. 2000, 37(3), 313. 103. Gameson, I.; Stead, W. J.; Rayment, T. J. Phys. Chem. 1991, 95(4), 1727–1730. 104. Cheng, L.; Yang, R. T. Microporous Mater. 1997, 8(3–4), 177. 105. Yang, R.; Baksh, M. S. A. Am. Inst. Chem. Eng. J. 1991, 37(5), 679. 106. Molinard, A.; Vansant, E. F. In Gas adsorption properties of cation modified alumina pillared montmorillonite, Separation Technology, Process Technology Proceedings, Amsterdam, Elsevier, Amsterdam, 1994; p 423. 107. Molinard, A.; Vansant, E. F. Adsorption 1995, 1(1), 49. 108. Zhu, H.; Gao, W. H.; Vansant, E. F. J. Colloid Interface Sci. 1995, 171(2), 377. 109. Zhu, H.; Lu, G. Q. J. Porous Mater. 1998, 5(3/4), 227. 110. Baksh, M. S. A.; Yang, R. T. AIChE J. 1992, 38(9), 1357–1468. 111. Heylen, I.; Maes, N.; Molinard, A.; Vansant, E. F. In Separation Technology, Process Technology Proceedings, Amsterdam; Vansant, E. F., Ed.; Elsevier, Amsterdam, 1994; p 355. 112. Pires, J.; Carvalho, A. P.; de Carvalho, M. B. Microporous Mesoporous Mater. 2001, 43(3), 277. 113. Sychev, M.; Shubina, T.; Rozwadowski, M.; Sommen, A. I. B.; De Beer, V. H. J.; Van Santen, R. A. Microporous Mesoporous Mater. 2000, 37(1–2), 187. 114. Yamanaka, S.; Nishihara, T.; Hattori, M. Mater. Chem. Phys. 1987, 17(1–2), 87. 115. Lecloux, A.; Pirard, J. P. J. Colloid Interface Sci. 1979, 70(2), 265–281. 116. Zhu, H. Y.; Lu, G. Q.; Maes, N.; Vansant, E. F. J. Chem. Soc. Faraday Trans. 1997, 93(7), 1417–1423. 117. Mikhail, R.; Brunauer, S.; Bodor, E. E. J. Colloid Interface Sci. 1968, 26(1), 45. 118. Occelli, M. L.; Olivier, J. P.; Perdigon-Melon, J. A.; Auroux, A. Langmuir 2002, 18(25), 9816–9823. 119. Barrett, E. P.; Joyner, L. G.; Halenda, P. P. J. Am. Chem. Soc. 1951, 73(1), 373–380. 120. Dubinin, M. M.; Radushkevich, L. V. Dokl. Akad. Nauk SSSR 1947, 55, 327–329. 121. Gil, A.; Korili, S. A.; Vicente, M. A. Catal. Rev. Sci. Eng. 2008, 50(2), 153–221. 122. Tsiao, C.-J.; Carrado, K. A.; Botto, R. E. Microporous Mesoporous Mater. 1998, 21(1–3), 45. 123. Sing, K. S. W.; Everett, D. H.; Haul, R. A. W.; Moscou, L.; Pierotti, R. A.; Rouquerol, J.; Siemieniewska, T. Pure Appl. Chem. 1985, 57(4), 603. 124. Polverejan, M.; Pauly, T. R.; Pinnavaia, T. J. Chem. Mater. 2000, 12(9), 2698–2704. 125. Van Der Voort, P.; Baltes, M.; Vansant, E. F.; White, M. G. Interface Sci. 1997, 5(2/3), 199–206. 126. Cool, P.; Vansant, E. F. In Handbook of Layered Materials; Auerbach, S. M., Carrado, K. A., Dutta, P. K., Eds.; Marcel Dekker: New York, 2004. 127. Busca, G. Phys. Chem. Chem. Phys. 1999, 1(5), 723–736. 128. Pelmenschikov, A. G.; van Santen, R. A.; Janchen, J.; Meijer, E. J. Phys. Chem. 1993, 97(42), 11071–11074. 129. Cool, P.; Vansant, E. F. In Handbook of Layered Materials; Auerbach, S. M., Carrado, K. A., Dutta, P. K., Eds.; Marcel Dekker: New York, 2004; p 299. 130. Pichowicz, M.; Mokaya, R. Chem. Commun. 2001, (20), 2100–2101. 131. De Vos, D.; Ernst, S.; Perego, C.; O’Connor, C. T.; Stocker, M. Microporous Mesoporous Mater. 2002, 56(2), 185. 132. Jero´nimo, D.; Guil, J. M.; Corbella, B. M.; Vasques, H.; Miranda, A.; Silva, J. M.; Lobato, A.; Pires, J.; Carvalho, A. P. Appl. Catal. A 2007, 330, 89–95. 133. Guil, J.; Perdigon-Melon, J. A.; de Carvalho, M. B.; Carvalho, A. P.; Pires, J. Microporous Mesoporous Mater. 2002, 51(2), 145. 134. Belkhadem, F.; Maldonado, A.; Siebenkaur, B.; Clacens, J. M.; Perez Zurita, M. J.; Bengueddach, A.; Figueras, F. Appl. Clay Sci. 2008, 39(1–2), 28. 135. Mokaya, R.; Jones, W. J. Catal. 1995, 153(1), 76. 136. Rupert, J.; Granquist, W.; Pinnavaia, T. J. In: Newman, A. C. D., Ed.; Chemistry of Clays and Clay Minerals Wiley Interscience: New York, 1987; vol. 6, pp 275–318.

209

137. Breck, D. W. Zeolite Molecular Sieves: Structure, Chemistry, and Use. Academic Press: New York, 1974; p. 752. 138. Occelli, M. L. Ind. Eng. Chem. Prod. Res. Dev. 1983, 22(4), 553–559. 139. Guan, J.; Yu, Z.; Chen, Z.; Tang, L.; Wang, X.-Q. Chon, H., Ihm, S.-K., Uk, U. S., Eds.; Studies in Surface Science and Catalysis Elsevier: Amsterdam, 1997, p 61. 140. Occelli, M. Stud. Surf. Sci. Catal. 1991, 63(Prep. Catal. 5), 287. 141. Occelli, M. L.; Pollack, S. S.; Sanders, J. V. Grobet, P. J., Mortier, W. J., Vansant, E. F., Schulz-Ekloff, G., Eds.; Studies in Surface Science and Catalysis Elsevier: Amsterdam, 1988; vol. 37, pp 45–55. 142. Gyftopoulou, M.; Millan, M.; Bridgewater, A. V.; Dugwell, D.; Kandiyoti, R.; Hriljac, J. A. Appl. Catal. A 2005, 282(1–2), 205. 143. Fetter, G.; Hernandez, V.; Rodriguez, V.; Valnyuele, M. A.; Lara, V. H.; Bosch, P. Mater. Lett. 2003, 57(5–6), 1220. 144. Benazzi, E.; Brendle, J.; Le Dred, R.; Baron, J.; Saehr, D.; Georges-Marchal, N.; Lacombe, S. Long-Reticulate-Distance Pillared Dioctahedral Phyllosilicate 2:1, Catalyst and Conversion Products. U.S. Patent 6,139,719, 10/31/2000, 2000. 145. Parulekar, V.; Hightower, J. W. Appl. Catal. 1987, 35(2), 263. 146. Parulekar, V. N.; Hightower, J. W. Appl. Catal. 1987, 35(2), 249–262. 147. Polverejan, M.; Liu, Y.; Pinnavaia, T. J. Stud. Surf. Sci. Catal. 2000, 129, 401 (Nanoporous Materials II, Proceedings of the Conference on Access in Nanoporous Materials, 2000). 148. Mori, H.; Okuhara, T.; Misono, M. Chem. Lett. 1987, 2305–2308. 149. Singh, B.; Roy, S. K.; Sharma, K. P.; Goswami, T. K. J. Chem. Technol. Biotechnol. 1998, 71(3), 246. 150. Benito, I.; del Riego, A.; Martinez, M.; Blanco, C.; Pesquera, C.; Gonzalez, F. Appl. Catal. A 1999, 180(1–2), 175. 151. Chae, H. J.; Nam, I.-S.; Yang, H. S.; Seok, L.; do Hur, I. J. Chem. Eng. Jpn. 2001, 34(2), 148. 152. Long, R. O.; Yang, R. T. J. Catal. 2000, 190(1), 22. 153. Bahamonde, A.; Mohino, F.; Robollar, M.; Yates, M.; Avila, P.; Mendioroz, S. Catal. Today 2001, 69(1–4), 233. 154. Gil, A.; Vicente, M. A.; Lambert, J.-F.; Gandia, L. M. Catal. Today 2001, 68(1–3), 41. 155. Louloudi, A.; Papayannakos, N. Appl. Catal. A 2000, 204(1), 167. 156. Kaloidas, V.; Koufopanos, C. A.; Ganges, N. H.; Papayannakos, N. Microporous Mater. 1995, 5(1–2), 97. 157. Zuo, S.; Zhou, R.; Qi, C. J. Rare Earths 2011, 29(1), 52. 158. Gonzalez, E.; Moronta, A. Appl. Catal. A 2004, 258(1), 99. 159. Ooka, C.; Yashida, H.; Suyuki, K.; Hattori, T. Microporous Mesoporous Mater. 2004, 67(2–3), 143. 160. Cafarelli, P.; De Stefanis, A.; Perez, G.; Tomlinson, A. A. G. React. Kinet. Catal. Lett. 2005, 86(1), 163. 161. Mishra, T.; Parida, K. M. J. Colloid Interface Sci. 2006, 301(2), 554. 162. Timofeeva, M. N.; Khankhasaeva, S. T.; Chesalov, Y. A.; Tsybulya, S. V.; Panchenko, V. N.; Dashinamzhilova, E. T. Appl. Catal. B 2009, 88(1–2), 127–134. 163. Navalon, S.; Alvaro, M.; Garcia, H. Appl. Catal. B 2010, 99(1–2), 1. 164. Herney-Ramirez, J.; Vicente, M. A.; Madeira, L. M. Appl. Catal. B 2010, 98(1–2), 10. 165. Timofeeva, M.; Panchenko, V. N.; Gil, A.; Chesalov, Y. A.; Sorokina, T. P.; Likholobov, V. A. Appl Catal, B 2011, 102(3–4), 433. 166. Rives, V. Layered Double Hydroxides: Present and Future. Nova Science Publishers: Huntington, NY, 2001. 167. Braterman, P. S.; Xu, Z. P.; Yarberry, F. In Handbook of Layered Materials; Auerbach, S. M., Carrado, K. A., Dutta, P. K., Eds.; Marcel Dekker: London, 2004; pp 373–466. 168. He, J.; Wei, M.; Li, B.; Kang, Y.; Evans, D.; Duan, X. In: Duan, X., Evans, D., Eds.; Layered Double Hydroxides Springer: Berlin/Heidelberg, 2006; vol. 119, pp 89–119. 169. Ogawa, M.; Inomata, K. Clay science 2011, 15(3), 131. 170. Stoecker, M.; Seyfarth, L.; Hirsemann, D.; Senker, J.; Breu, J. Appl. Clay Sci. 2010, 48(1–2), 146. 171. Baumgarter, A.; Sattler, K.; Thun, J.; Breu, J. Angew. Chem. Int. Ed. 2008, 47(9), 1640. 172. Stocker, M.; Seidl, W.; Seyfarth, L.; Senker, J.; Breu, J. Chem. Commun. 2008, (5), 629–631. 173. Weiss, A.; Becker, H. O.; Orth, H.; Mai, G.; Lechner, H.; Range, K. J. In Proceedings of the International Clay Conference, Tokyo, Japan; Heller, L., Ed.; Israel Universities Press, Tokyo, 1969; pp 180–186. 174. Granquist, W.; Hoffman, G. W.; Boteler, R. C. Clays Clay Miner. 1972, 20(5), 323. 175. Kloprogge, J.; Breuklaar, J.; Geus, J. W.; Jansen, J. B. H. Clays Clay Miner. 1994, 42(4), 416. 176. de Boer, J. H. Z. Anorg. Allg. Chem. 1925, 144(1), 190–196.

210

Porous Pillared Clays and Layered Phosphates

177. Kraus, K. A.; Phillips, H. O. J. Am. Chem. Soc. 1956, 78(3), 694. 178. Amphlett, C. B.; McDonald, L. A.; Redman, M. J. J. Inorg. Nucl. Chem. 1958, 6(3), 220–235. 179. Clearfield, A.; Stynes, J. A. J. Inorg. Nucl. Chem. 1964, 26(1), 117–129. 180. Clearfield, A.; Blessing, R. H.; Stynes, J. A. J. Inorg. Nucl. Chem. 1968, 30(8), 2249–2258. 181. Alberti, G.; Costantino, U.; Allulli, S.; Tomassini, N. J. Inorg. Nucl. Chem. 1978, 40(6), 1113–1117. 182. Poojary, M. D.; Hu, H.-L.; Campbell, F. L., III; Clearfield, A. Acta Crystallogr. Sect. B: Struct. Sci. 1993, 49(6), 996–1001. 183. Clearfield, A.; Duax, W. L.; Medina, A. S.; Smith, G. D.; Thomas, J. R. J. Phys. Chem. 1969, 73(10), 3424–3430. 184. Clearfield, A.; Smith, G. D. Inorg. Chem. 1969, 8(3), 431–436. 185. Troup, J. M.; Clearfield, A. Inorg. Chem. 1977, 16(12), 3311–3314. 186. Albertsson, J.; Oskarsson, A.; Tellgren, R.; Thomas, J. O. J. Phys. Chem. 1977, 81(16), 1574–1578. 187. Clearfield, A. Annu. Rev. Mat. Sci. 1984, 14(1), 205–229. 188. Poojary, D. M.; Shpeizer, B.; Clearfield, A. J. Chem. Soc. Dalton Trans. 1995, (1), 111–113. 189. Clearfield, A.; Tindwa, R. M. J. Inorg. Nucl. Chem. 1979, 41(6), 871–878. 190. Tindwa, R. M.; Ellis, D. K.; Peng, G.-Z.; Clearfield, A. J. Chem. Soc. Faraday Trans. 1985, 81(2), 545–552. 191. Costantino, U. J. Chem. Soc. Dalton Trans. 1979, 2, 402–405. 192. Alberti, G.; Casciola, M.; Costantino, U. J. Colloid Interface Sci. 1985, 107(1), 256–263. 193. Clearfield, A.; Costantino, U. Layered metal phosphates and their intercalation chemistry. Pergamon: New York, 1996; pp. 107–149. 194. Kumar, C. V.; Bhambhani, A.; Hnatiuk, N. In Handbook of Layered Materials; Auerbach, S. M., Carrado, K. A., Dutta, P. K., Eds.; Marcel Dekker: New York, 2004; pp 320–332. 195. Costantino, U. In Inorganic Ion Exchange Materials; Clearfield, A., Ed.; CRC Press: Boca Raton, FL, 1982; pp 111–132. 196. Alberti, G.; Costantino, U. In Intercalation Chemistry; Whittingham, M. S., Jacobson, J. A., Eds.; Academic Press: New York, 1982; pp 147–180. 197. Sun, L.; Boo, W. J.; Sue, H.-J.; Clearfield, A. New J. Chem. 2007, 31(1), 39–43. 198. Alberti, G.; Costantino, U.; Giulietti, R. J. Inorg. Nucl. Chem. 1980, 42(7), 1062–1063. 199. Bellezza, F.; Cipiciani, A.; Costantino, U.; Marmottini, F.; Quotadamo, M. A. Colloid Polym. Sci. 2006, 285, 19–25. 200. Clearfield, A.; Berman, J. R. J. Inorg. Nucl. Chem. 1981, 43(9), 2141–2142. 201. Bortun, A. I.; Clearfield, A.; Sua´rez, M.; Llavona, R.; Rodrı´guez, J. Mater. Chem. Phys. 1998, 55(2), 152–154. 202. Beck, J. S.; Vartuli, J. C.; Roth, W. J.; Leonowicz, M. E.; Kresge, C. T.; Schmitt, K. D.; Chu, C. T. W.; Olson, D. H.; Sheppard, E. W. J. Am. Chem. Soc. 1992, 114(27), 10834–10843. 203. Kresge, C. T.; Leonowicz, M. E.; Roth, W. J.; Vartuli, J. C.; Beck, J. S. Nature 1992, 359(6397), 710–712. 204. Ciesla, U.; Schacht, S.; Stucky, G. D.; Unger, K. K.; Schueth, F. Angew. Chem. Int. Ed. Engl. 1996, 35(5), 541. 205. Jime´nez-Jime´nez, J.; Maireles-Torres, P.; Olivera-Pastor, P.; Rodrı´guezCastello´n, E.; Jime´nez-Lo´pez, A.; Jones, D. J.; Rozie`re, J. Adv. Mater. 1998, 10(10), 812–815. 206. Rodrı´guez-Castello´n, E.; Jime´nez-Jime´nez, J.; Jime´nez-Lo´pez, A.; MairelesTorres, P.; Ramos-Barrado, J. R.; Jones, D. J.; Rozie`re, J. Solid State Ion. 1999, 125(1–4), 407–410. 207. Jones, D. J.; Aptel, G.; Brandhorst, M.; Jacquin, M.; Jimenez-Jimenez, J.; Jimenez-Lopez, A.; Maireles-Torres, P.; Piwonski, I.; Rodriguez-Castellon, E.; Zajac, J.; Roziere, J. J. Mater. Chem. 2000, 10(8), 1957–1963. 208. Sun, Y.; Afanasiev, P.; Vrinat, M.; Coudurier, G. J. Mater. Chem. 2000, 10(10), 2320–2324. 209. Pan, C.; Yuan, S.; Zhang, W. Appl. Catal. A 2006, 312, 186–193. 210. Blanco, J. A.; Khainakov, S. A.; Khainakova, O.; Garcı´a, J. R.; Garcı´a-Granda, S. Phys. Status Solidi C 2009, 6(10), 2190–2194. 211. Liu, J.; Wei, X.; Yu, Y.; Song, J.; Wang, X.; Li, A.; Liu, X.-W.; Deng, W.-Q. Chem. Commun. 2010, 46(10), 1670–1672. 212. Bhaumik, A.; Inagaki, S. J. Am. Chem. Soc. 2001, 123(4), 691–696. 213. Ren, T.-Z.; Yuan, Z.-Y.; Azioune, A.; Pireaux, J.-J.; Su, B.-L. Langmuir 2006, 22(8), 3886–3894. 214. Zhao, G. L.; Yuan, Z. Y.; Chen, T. H. Mater. Res. Bull 2005, 40(11), 1922–1928. 215. Li, X. S.; Courtney, A. R.; Yantasee, W.; Mattigod, S. V.; Fryxell, G. E. Inorg. Chem. Commun. 2006, 9(3), 293–295. 216. Ma, T.-Y.; Zhang, X.-J.; Shao, G.-S.; Cao, J.-L.; Yuan, Z.-Y. J. Phys. Chem. C 2008, 112(8), 3090–3096.

217. Tian, X.; He, W.; Cui, J.; Zhang, X.; Zhou, W.; Yan, S.; Sun, X.; Han, X.; Han, S.; Yue, Y. J. Colloid Interface Sci. 2010, 343(1), 344–349. 218. Fei, H.; Zhou, X.; Zhou, H.; Shen, Z.; Sun, P.; Yuan, Z.; Chen, T. Microporous Mesoporous Mater. 2007, 100(1–3), 139–145. 219. Ren, T.-Z.; Yuan, Z.-Y.; Su, B.-L. Chem. Commun. 2004, (23), 2730–2731. 220. Yuan, Z.-Y.; Ren, T.-Z.; Azioune, A.; Pireaux, J.-J.; Su, B.-L. Catal. Today 2005, 105(3–4), 647–654. 221. Ma, T.-Y.; Lin, X.-Z.; Yuan, Z.-Y. Chem. Eur. J. 2010, 16(28), 8487–8494 S8487/1-S8487/5. 222. Ma, T.-Y.; Lin, X.-Z.; Yuan, Z.-Y. J. Mater. Chem. 2010, 20(35), 7406–7415. 223. Ma, T.-Y.; Ren, T.-Z.; Yuan, Z.-Y. Adv. Mater. Res. 2010, 132, 87–95. 224. Ma, T.-Y.; Lin, X.-Z.; Zhang, X.-J.; Yuan, Z.-Y. New J. Chem. 2010, 34(6), 1209–1216. 225. Ma, T.-Y.; Liu, L.; Deng, Q.-F.; Lin, X.-Z.; Yuan, Z.-Y. Chem. Commun. 2011, 47(21), 6015–6017. 226. Zhang, X.-J.; Ma, T.-Y.; Yuan, Z.-Y. J. Mater. Chem. 2008, 18(17), 2003–2010. 227. Ma, T.-Y.; Zhang, X.-J.; Yuan, Z.-Y. Microporous Mesoporous Mater. 2009, 123(1–3), 234–242. 228. Ma, T.-Y.; Zhang, X.-J.; Yuan, Z.-Y. J. Mater. Sci. 2009, 44(24), 6775–6785. 229. Sarkar, K.; Laha, S. C.; Mal, N. K.; Bhaumik, A. J. Solid State Chem. 2008, 181(8), 2065–2072. 230. Kumar, C. V.; Bhambhani, A.; Hnatiuk, N. In Handbook of Layered Materials; Auerbach, S. M., Carrado, K. A., Dutta, P. K., Eds.; Marcel Dekker: New York, 2004; pp 317–332. 231. Olivera-Pastor, P.; Maza-Rodriguez, J.; Maireles-Torres, P.; RodriguezCastellon, E.; Jimenez-Lopez, A. J. Mater. Chem. 1994, 4(2), 179–184. 232. Perez-Reina, F. J.; Olivera-Pastor, P.; Rodriguez-Castellon, E.; Jimenez-Lopez, A.; Fierro, J. L. G. J. Solid State Chem. 1996, 122(1), 231–238. 233. Alcantara-Rodriguez, M.; Olivera-Pastor, P.; Rodriguez-Castellon, E.; JimenezLopez, A. J. Mater. Chem. 1996, 6(2), 247–252. 234. Jime´nez-Jime´nez, J.; Maireles-Torres, P.; Olivera-Pastor, P.; Rodrı´guezCastello´n, E.; Jime´nez-Lo´pez, A. Langmuir 1997, 13(10), 2857–2862. 235. Shpeizer, B. G.; Sylvester, P.; Cahill, R. A.; Clearfield, A. Chem. Mater. 1999, 11(5), 1201–1209. 236. Parida, K.; Das, D. P. J. Photochem. Photobiol. A 2004, 163(3), 561–567. 237. Clearfield, A.; Roberts, B. D. Inorg. Chem. 1988, 27(18), 3237–3240. 238. Olivera-Pastor, P.; Jimenez-Lopez, A.; Maireles-Torres, P.; RodriguezCastellon, E.; Tomlinson, A. A. G.; Alagna, L. J. Chem. Soc. Chem. Commun. 1989, (12), 751–752. 239. Maireles-Torres, P.; Olivera-Pastor, P.; Rodriguez-Castellon, E.; JimenezLopez, A.; Alagna, L.; Tomlinson, A. A. G. J. Mater. Chem. 1991, 1(3), 319–326. 240. Criado, C.; Ramos Barrado, J. R.; Maireles-Torres, P.; Olivera-Pastor, P.; Rodrı´guez Castello´n, E.; Jime´nez-Lo´pez, A. Solid State Ion. 1993, 61(1–3), 139–142. 241. Rodriguez-Castellon, E.; Olivera-Pastor, P.; Maireles-Torres, P.; JimenezLopez, A.; Sanz, J.; Fierro, J. L. G. J. Phys. Chem. 1995, 99(5), 1491–1497. 242. Peeters, K.; Grobet, P.; Vansant, E. F. J. Mater. Chem. 1996, 6(2), 239–246. 243. Me´rida-Robles, J. M.; Olivera-Pastor, P.; Jime´nez-Lo´pez, A.; Rodrı´guezCastello´n, E. J. Phys. Chem. 1996, 100(35), 14726–14735. 244. Braos-Garcı´a, P.; Rodrı´guez-Castello´n, E.; Maireles-Torres, P.; Olivera-Pastor, P.; Jime´nez-Lo´pez, A. J. Phys. Chem. B 1998, 102(10), 1672–1678. 245. Xu, J.; Gao, Z. Microporous Mesoporous Mater. 1998, 24(4–6), 213–222. 246. Rodrı´guez-Castello´n, E.; Jime´nez-Lo´pez, A.; Maireles-Torres, P.; Me´ridaRobles, J.; Braos-Garcı´a, P.; Aguilar-Armenta, G.; Flores-Loyola, E.; Marmottini, F.; Felici, E. Langmuir 2001, 17(12), 3769–3775. 247. Engelhardt, G.; Michel, D. High-Resolution Solid-State NMR of Silicates and Zeolites. John Wiley and Sons: Chichester, 1987; p. Medium: X; Size: Pages: 499. 248. Braos-Garcia, P.; Santamaria-Gonzalez, J.; Maireles-Torres, P.; RodriguezCastellon, E.; Jimenez-Lopez, A. Green Chem. 2001, 3(6), 289–295. 249. Herna´ndez-Huesca, R.; Braos-Garcı´a, P.; Me´rida-Robles, J.; Maireles-Torres, P.; Rodrı´guez-Castello´n, E.; Jime´nez-Lo´pez, A. Chemosphere 2002, 48(4), 467–474. 250. Yan, J. Y.; Kung, H. H.; Sachtler, W. M. H.; Kung, M. C. J. Catal. 1998, 175(2), 294–301. 251. Braos-Garcı´a, P.; Maireles-Torres, P.; Rodrı´guez-Castello´n, E.; Jime´nez-Lo´pez, A. J. Mol. Catal. A: Chem. 2001, 168(1–2), 279–287. 252. Merida Robles, J.; Biondi, E.; Hernandez-Huesca, R.; Maireles-Torres, P.; Rodriguez-Castellon, E.; Jimenez-Lopez, A. Ind. Ceram. 2002, 22(1), 33. 253. Jiao, X.; Chen, D.; Pang, W.; Xu, R.; Yue, Y. J. Mater. Chem. 1998, 8(12), 2831–2834. 254. Roziere, J.; Jones, D. J.; Cassagneau, T. J. Mater. Chem. 1991, 1(6), 1081–1082. 255. Sylvester, P.; Cahill, R.; Clearfield, A. Chem. Mater. 1994, 6(10), 1890–1898. 256. Pawelec, B.; Murcia-Mascaro´s, S.; Fierro, J. L. G. Langmuir 2002, 18(21), 7953–7963.

Porous Pillared Clays and Layered Phosphates

257. Murcia-Mascaro´s, S.; Pawelec, B.; Fierro, J. L. G. Catal. Commun. 2002, 3(7), 305–311. 258. Michel, J., Zhou, H.-C., Long, J. R., Yaghi, O. M., Eds.; Special Issue on Metal-Organic Frameworks Chem. Rev. American Chemical Society: Washington, DC, 2012; vol. 112, pp 673–1268. 259. Alberti, G.; Costantino, U.; Vivani, R.; Biswas, R. K. React. Polym. 1992, 17(3), 245–253. 260. Vivani, R.; Costantino, U.; Taddei, M. In Metal Phosphonate Chemistry: From Synthesis to Applications; Clearfield, A., Demadis, K., Eds.; Royal Society of Chemistry: Cambridge, 2011; pp 45–86. 261. Alberti, G. In Comprehensive supramolecular chemistry; Lehn, J.-M., Alberti, G., Bein, T., Eds.; Pergamon-Elsevier: Oxford, 1996; p 111. 262. Alberti, G.; Casciola, M.; Biswas, R. K. Inorg. Chim. Acta 1992, 201(2), 207–212. 263. Alberti, G.; Marmottini, F.; Murcia-Mascaro´s, S.; Vivani, R. Angew. Chem. Int. Ed Engl. 1994, 33(15–16), 1594–1597. 264. Alberti, G.; Vivani, R.; Marmottini, F.; Zappelli, P. J. Porous Mater. 1998, 5(3), 205–220. 265. Alberti, G.; Vivani, R.; Murcia Mascaro´s, S. J. Mol. Struct. 1998, 470(1–2), 81–92. 266. Alberti, G.; Murcia-Mascaro´s, S.; Vivani, R. J. Am. Chem. Soc. 1998, 120(36), 9291–9295. 267. Clearfield, A.; Wang, Z. J. Chem. Soc. Dalton Trans. 2002, (15), 2937–2947. 268. Alberti, G.; Brunet, E.; Dionigi, C.; Juanes, O.; de la Mata, M. J.; Rodrı´guez-Ubis, J. C.; Vivani, R. Angew. Chem. Int. Ed. 1999, 38(22), 3351–3353. 269. Brunet, E.; Alhendawi, H. M. H.; Cerro, C.; de la Mata, M. J.; Juanes, O.; Rodrı´guez-Ubis, J. C. Angew. Chem. Int. Ed. 2006, 45(41), 6918–6920. 270. Brunet, E.; de la Mata, M. J.; Alhendawi, H. M. H.; Cerro, C.; Alonso, M.; Juanes, O.; Rodrı´guez-Ubis, J. C. Chem. Mater. 2005, 17(6), 1424–1433. 271. Alberti, G.; Casciola, M.; Vivani, R.; Biswas, R. K. Inorg. Chem. 1993, 32(21), 4600–4604. 272. Brunet, E.; Alhendawi, H. M. H.; Cerro, C.; de la Mata, M. J.; Juanes, O.; Rodrı´guez-Ubis, J. C. Chem. Eng. J. 2010, 158(2), 333–344. 273. Brunet, E.; Alhendawi, H. M. H.; Cerro, C.; de la Mata, M. J.; Juanes, O.; Rodrı´guez-Ubis, J. C. Microporous Mesoporous Mater. 2011, 138(1–3), 75–85. 274. Clearfield, A.; Demadis, K., Eds.; Metal Phosphonate Chemistry: From Synthesis to Application. Royal Society of Chemistry: Cambridge, 2012; pp 655.

211

275. Gagnon, K. J.; Perry, H. P.; Clearfield, A. Chem. Rev. 2011, 112(2), 1034–1054. 276. Dines, M. B.; Cooksey, R. E.; Griffith, P. C.; Lane, R. H. Inorg. Chem. 1983, 22(6), 1003–1004. 277. Clearfield, A. In Progress in Inorganic Chemistry; Karlin, K. D., Ed.; John Wiley and Sons: New York, 2007; pp 371–510. 278. Clearfield, A. Chem. Mater. 1998, 10(10), 2801–2810. 279. Clearfield, A. Dalton Trans. 2008, (44), 6089–6102. 280. Ortiz-Avila, C. Y.; Bhardwaj, C.; Clearfield, A. Inorg. Chem. 1994, 33(12), 2499–2500. 281. Clearfield, A.; Ortiz-Avila, C. Y. Supramolecular Architecture American Chemical Society: Washington, DC, 1992; vol. 499 pp. 178–193. 282. Dines, M. B.; Digiacomo, P. M.; Callahan, K. P.; Griffith, P. C.; Lane, R. H.; Cooksey, R. E. Chemically Modified Surfaces in Catalysis and Electrocatalysis ACS: Washington, DC, 1982; vol. 192 pp. 223–240. 283. Clearfield, A.; Wang, J. D.; Tian, Y.; Campbell, F. L., III; Peng, G. Z. In Materials Synthesis and Characterization; Perry, D. L., Ed.; Plenum: New York, 1997; pp 103–111. 284. Alberti, G.; Costantino, U.; Marmottini, F.; Vivani, R.; Zappelli, P. Angew. Chem. Int. Ed Engl. 1993, 32(9), 1357–1359. 285. Medoukali, D.; Hubert Mutin, P.; Vioux, A. J. Mater. Chem. 1999, 9(10), 2553–2557. 286. Derouane, E. G.; Jullien-Lardot, V. Stud. Surf. Sci. Catal. 1994, 83, 11. 287. Alberti, G.; Marmottini, F.; Vivani, R.; Zappelli, P. J. Porous Mater. 1998, 5(3/4), 221–226. 288. Perry, H. P.; Law, J.; Zon, J.; Clearfield, A. Microporous Mesoporous Mater. 2011, 149. 289. Yang, C.; Clearfield, A. React. Polym. Ion Exch. Sorbents 1987, 5(1), 13. 290. Clearfield, A.; Wang, Z.; Bellinghausen, P. J. Solid State Chem. 2002, 167(2), 376–385. 291. Stein, E. W., Sr.; Clearfield, A.; Subramanian, M. A. Solid State Ion. 1996, 83(1–2), 113–124. 292. Odobel, F.; Bujoli, B.; Massiot, D. Chem. Mater. 2000, 13(1), 163–173. 293. Cahill, R. A.; Shpeizer, B. G.; Peng, G. Z.; Bortun, A. I.; Clearfield, A. In Separations of f Elements; Nash, K. L., Choppin, G. R., Eds.; Plenum Press: New York, 1995. 294. Burns, J. D.; Clearfield, A.; Borkowski, M.; Reed, D. T. Radiochim. Acta 2012, 100(6), 381–387.