Inclusion Chemistry in Periodic Mesoporous Hosts

Inclusion Chemistry in Periodic Mesoporous Hosts

(Parte 1 de 5)

Inclusion Chemistry in Periodic Mesoporous Hosts

Karin Moller and Thomas Bein* Department of Chemistry, Purdue University, West Lafayette, Indiana 47907

Received April 3, 1998. Revised Manuscript Received July 7, 1998

This review provides an overview of different aspects of inclusion chemistry in ordered mesoporous host materials such as MCM-41 or MCM-48 (Mobil codes). A rich field of inclusion chemistry has been explored in this context, including sorption, ion exchange, imbibition followed by reduction, grafting of reactive metal alkoxides, halides etc., grafting of silane coupling agents (sometimes followed by subsequent reactions), grafting of reactive metal complexes, and polymerization in the channels. Finally, co-condensation of reactive species during the mesopore synthesis is a method to incorporatefunctionalityinto the walls of the channel system. Important applications of these modified and functionalized systems are heterogeneous catalysis and photocatalysis involving bulky grafted catalysts and/or the conversion of large substrates. Other potential applications include ion exchange and separations, removal of heavy metals, chromatography, stabilization of quantum wires, stabilization of dyes, and polymer composites.

1. Introduction

A new era in inclusion chemistry began with the discovery of periodic mesoporous materials by Beck et al. in 1992.1 The narrow, controlled pore size distributionin the orderedhexagonalMCM-41and cubicMCM- 48 materials (Mobil codes) and the large pore openings haveaddeda newdimensionto intraporechemistrythat had previously focused on microporous zeolitic materials. This discovery has stimulated research in areas that include fundamental studies of sorption and phase transitions in confined spaces, ion exchange, the formation of intrachannel metal, metal oxide, and semiconductor clusters, and inclusion of various metal complexes and other guests. It is interesting to note that a processfor the formationof “low-bulkdensitysilica”had been patented by Chiola et al. in 1971.2 In this process tetraethyl orthosilicate (TEOS) is hydrolyzed in the presence of cationic surfactants. Although the aforementioned patent does not describe the porosity or periodicity of the resulting materials, Di Renzo et al.3 showed recently that the reaction products are similar to MCM-41. Another related family of mesoporous compounds has been explored by Kuroda et al.,4 who showedthationexchangeof thelayeredpolysilicateNa+ kanemite with alkyltrimethylammoniumions, followed by calcination, gives a hexagonal mesoporous material.

The internal surface reactivity of the mesoporous hosts has been utilized by covalently anchoring a number of functional groups to the channel walls, including attachment of ligands that are used for the formation of bulky metal complexes. An alternative approach to grafting the mesoporous internal walls is beginning to develop; that is, co-condensation of framework precursors with metal sources that are covalently bound to a functional group. Finally, intrachannel reactions even include polymerization of preadsorbed monomers, thus leading to confined filaments of common polymers as well as conducting materials such as carbon.

Intriguing applications of these systems include catalysis with large substrate molecules, formation of novel nanocomposites, and separation.

2. Review of Inclusion Chemistry

An overview of inclusion chemistry in periodic mesoporous hosts, including literature until early 1998, will be provided in this review, according to the following subject headings: (1) sorption and phase transitions; (2) ion exchange and complexation; (3) metal and semiconductor clusters and wires; (4) oxide and sulfide clusters; (5) inclusion of metal complexes and other guests; (6) covalent grafting of ligands and functional groups; (7) hybrid materials obtained by in situ cocondensation; and (8) polymerization in mesoporous channels. 2.1. Sorption and Phase Transitions. With the invention of mesoporous MCM structures with defined, yet variable pore sizes it has become possible to study the physical properties of adsorbates in confined spaces as a function of pore size, including fundamental subjects such as the process of nucleation and crystallization of ice. Baker et al.5 studied this process with poroussilicasand orderedaluminosilicatesin which the water crystallized in a hybrid form having cubic as well as hexagonal characteristics of ice. A behavior of “frustrated nucleation” was observed in the pores of MCM. Very gradual freezing was found by Morishige et al.6 when studying the crystallization in MCMs with a pore size of 24 Å. However, MCM hosts with 42 Å pores gave rise to abrupt freezing and crystallization into cubic ice at the drastically reduced freezing point of 232 K. The adsorption of the mono- and diatomic gases Ar,

* Author to whom correspondence should be sent. FAX: 765-494- 0239. E-mail:

10.1021/cm980243e C: $15.0 © 1998 American Chemical Society Published on Web 10/02/1998

Downloaded by UNIV EST PAULISTA UNESP on October 30, 2009 | Publication Date (Web): October 2, 1998 | doi: 10.1021/cm980243e a function of temperature in MCM hosts with four different pore radii ranging between 12 and 21 Å. The capillary critical temperature was far below the bulk critical temperature, and depended on the ratio of the molecular diameter to the pore radius.7 The capillary melting of methane in the channels of MCM-41, when studied with neutron inelastic scattering, showed similar deviations from bulk behavior. Here, the rotational melting is observed 10 degrees above the bulk value, whereas translational melting occurs 30 degrees below the bulk melting point.8

The hexagonal arrangement of the pore system in

MCM-41givesriseto an X-raydiffraction(XRD)pattern of at least five diffractionpeaks in high qualitysamples. However, as demonstrated by Marler et al.,9 the intensity of these peaks is not necessarily an indication of the order of the hexagonal arrangement in these materials. They reported on the inclusion of different sorbates such as bromoform into the channels of boroncontaining B-MCM-41 to show that inclusion compounds with X-ray scattering power similar to that of the matrix may even result in complete extinction of XRD reflections. Peak intensity was nearly fully regained after removal of these sorbates.

2.2. IonExchangeandComplexation. Largepore

MCM supports are of great interest for numerous catalytic applications, some of which require the introduction of accessible ions into the pores. Ion exchange has long been studied for the microporouszeolite family and is now extended toward aluminum-containing and other MCM materials with charged walls.

Nickel. Nickel(I) ions have been intoduced by liquid ion exchange and via a solid-state reaction into MCM hosts by Kevan et al.10 to generate Ni(I) species. The reductionof these ions to Ni(I) was afforded by thermal, hydrogen reduction, or by gamma irradiation in Al- MCM-41 when the ions were introducedfrom the liquid phase. Ni(I) was also introduced into the synthesis mixture, and the different Ni-containing mesoporous materials were compared in their activity to catalyze ethylene dimerization.1 The activity decreased in the order Ni-MCM-41 (Ni in synthesis mixture) > Ni- AlMCM-41> Ni-MCM-41(bothcontainingNi fromion exchange),which was correlatedwith the concentration of catalytically active Ni(I) species.

Copper. Kevan and co-workers12 further studied the copper-exchange capacity of different mesoporous materials. The ion-exchange capacity of aluminosilicate MCM-41 for copper(I) was highest when compared to zeolite Y, alumina, and MCM with extraframework aluminum and was found to be correlated with the concentration of tetrahedral aluminum in the framework.12 In aluminosilicate MCM-41, the cupric ion remainedcoordinatedto theframeworkuponadsorption and adduct formation with small ligands such as water, methanol, or ammonia. In contrast, large ligands, such as pyridine, did not coordinate with the copper, suggesting that it was located within the MCM walls.13 In siliceous MCM-41, the ion-exchanged copper was removed from its ion-exchange sites upon adsorption of polar ligands such as methanol or pyridine that could form coordination complexes. In contrast, adsorbates such as ethylene and benzene only distorted the copper location upon adduct formation but did not remove the cations from their framework oxygen coordination.14 Similarly, complex formation with Cu(I) was observed with water (with a distorted octahedral coordination) and ammonia in siliceous MCM-41, whereas in aluminosilicate-MCM, square-pyramidal coordination took place with the cation still attached to the MCM walls.15

Other Ions. Al-MCM-41 was ion exchanged with

Na+,K +,C a2+, andY3+. Thermaltreatmentwithwatersaturatedoxygengas at differenttemperaturesresulted in highly thermally stable materials that exceeded the stability of nonexchanged Al-MCM-41.16

Cesium in the form of CsCl and cesium acetate was introduced into Al-MCM-41 by ion exchange and studied in base-catalyzed Knoevenagel condensation and acid-catalyzed acetalization and aldol condensation.17

An alternative route to the commonly used ion exchange has been explored by “planting” cations into siliceous MCM-41. In this case, Mn(I) cations are introduced through partial ion exchange for the template ions into the as-synthesized MCM host. Subsequent calcination in air results in Mn2+-containing MCM-41. This method allows one to vary the Mn2+ concentration over a wide range without loss of crystallinity.18

2.3. Metal and Semiconductor Clusters and

Wires. High surface area inorganic supports have a long history of applications in heterogeneous catalysis. Crystallinezeoliteshave been widelyused as stabilizing matrix for the preparation of highly dispersed metal particles. Their microporosity and molecular sieving behavior can provide selectivity in reactions involving small molecules. If reactionsinvolvinglargermolecules or absence of shape selectivity are desired, large pore hosts are more appropriate supports. Thus, known pathways for the preparation of small metal clusters have been adapted for mesoporous supports.

Platinum. Conventionalion-exchangeroutes, includ- ing the exchange with Pt(NH3)4 2+ and subsequent activation in oxygen followed by reduction in hydrogen was described by Ryoo et al.,19 who aimed to create small platinum particles supported in MCM-41. The resulting clusters showed a high catalytic activity for the hydrogenolysis of ethane. The authors grew platinum clusters and wires inside of the channels of MCM- 41, MCM-48, and KIT-1 (a less ordered analog) through consecutive ion exchange, impregnation, and reduction of Pt2+ to image the internal morphology and dimensionality with transmission electron microscopy (TEM).20,21 A variety of pathways, such as incipient wetness, ion exchange, or direct introduction of a Ptsalt in the synthesis mixture were used by Schuth et al.2 for the introduction of platinum into MCM-41. These authors studied the catalytic activity of these catalysts in the low-temperature CO oxidation.2

Corma et al.23 compared the catalytic activity of platinum supported on MCM-41 for the hydrogenation of aromaticswiththe activityof platinumon amorphous supports such as alumina and silica as well as in zeolite USY (ultrastable Y; faujasite structure). The highest dispersion, overall hydrogenation activity, and sulfur tolerance was found for Pt in MCM-41, but the highest turnover number was observed in USY, which also contains strong Bronsted sites.

Reviews Chem. Mater., Vol. 10, No. 10, 1998 2951

Downloaded by UNIV EST PAULISTA UNESP on October 30, 2009 | Publication Date (Web): October 2, 1998 | doi: 10.1021/cm980243e

Palladium. The catalytic activity of small palladium clusters supported on MCM materials was tested in several reactions. Schuthe ta l.24 used a novel method to support monodispersed, preformed Pt clusters on MCM-41. The clusters were synthesized by reduction of Pd(I) acetate in the presence of stabilizing phenantroline ligands, and the clusters with seven and eightshells of Pd, Pd7/8(phen), were isolated. These clusters were subsequentlyincorporatedinto the MCM-41 channels either directly during synthesis of the support or by incipientwetnessimpregnationintothe calcinedhost from a water/pyridinemixture. The clustersintroduced in situ were converted into PdO particles during template removal through calcination at 873 K. However, the clusters impregnated in the MCM host remained intact and were significantly more stable than on other supports. They showed a high activity in the oxidation of CO.

Siliceous MCM-41 was also used as support for the preparation of 20-25 Å sized palladium particles.25 These particles were studied as catalysts for the hydrogenation of hexene and benzene. A high activity for hexene conversion at 298 K was found, but the catalyst was inactive for the selective hydrogenationof benzene.

Other researchers observed a substantial increase in activity with MCM-supported platinum and palladium particles in the complete hydrogenation of naphthalene when compared with catalysts with the same metals on amorphous alumina.26

Other Metals. Several other metal and metal-oxide clusters were introduced into mesoporous supports and used in catalysis. For example, the formation of small cobalt clusters was achieved in MCM-41 through the stabilizing effect of the alkylammonium bromide template.27 The preparation was achieved by direct addi- tion of CoCl2 to the synthesis mixture. This approach did not result in the inclusion of cobalt into the MCM walls, but the presence of the template prevented unwanted sintering of the clusters during calcination or sulfidation. The structure of the MCM-41 host was retained.

Hydrogenation of hexene was tested with bimetallic ruthenium/silver clusters derived from the thermolytic decomposition of a new [Ag3Ru10C2(CO)28Cl]2- complex that was adsorbed intact to the MCM walls and char- acterized by extended X-ray absorption fine structure (EXAFS) spectroscopy and microscopy.28

Ruthenium clusters in the presence of alkali and alkaline earth promoters were studied for the synthesis of ammonia on zeolite X, magnesia,and MCM-41.29 The ammonia synthesis is known to be structure sensitive. Attempts to synthesize Ru clusters with a similar size of 1 nm in all different supports were only successful in zeolite X and magnesia, but not in the mesoporous supportwherethe dispersionwas lower. Theseclusters and the promoters were supported on Si-MCM-41 and magnesiaby impregnationwith and subsequentdecom- position of Ru3(CO)12 clusters. Ruthenium was introduced into zeolite X through ion exchange with the

Ru(I)amminecomplexsalt and convertedby reduction in hydrogen.

Metal-containing MCM supports were also claimed in a patent of Pelrine and co-workers30 for the oligomerization of olefins with respect to the production of lubricants. The authors introduced group VIB metals such as Cr(I) acetate into the pores from aqueous solution. The preparation of the catalysts was completed by subsequentcalcinationand reductionin CO.30

(Parte 1 de 5)