Manganese oxide-based solids having channels, layers, and pillars can be made to specification for applications ranging from detergents to sensors.

Steven L. Suib

Octahedral molecular sieves and octahedral layered manganese oxide materials exhibit good conductivity, high porosity, and good thermal stability. The dimensions of the channels and interlayer spacings in these materials can be adjusted to fit various applications, and chemical reactions at the active sites can be varied by substituting different metals into the structure or by changing the valence of the manganese ions. This adaptability opens up possibilities for several potential applications, including adsorbents, batteries, catalysts, catalyst supports, detergents, protective coatings, and sensors.

Octahedral manganese oxides
Much of our research at the University of Connecticut has focused on the synthesis of porous octahedral molecular sieves (OMSs) and octahedral layered (OL) materials (1–15). Molecular sieves comprise infinite 3-D crystalline frameworks with molecule-sized tunnels similar to those found in the naturally occurring zeolites. Layered materials have interlayer spacings that are similar to clay-type materials. Octahedral solid materials are built from six-coordinate metal atoms surrounded by an octahedral array of anions (oxygen, in this case). OMS and OL materials having various tunnel sizes, compositions, and structures are shown in Figure 1. OMS-1, OMS-2, and OL-1 are all mixed-valence materials. Synthetic feitknechtite, OL-2, is a precursor to OL-1 and is a Mn³⁺ oxyhydroxide phase that is isostructural with Mn(OH)₂. The interlayer spacing of OL-1 is ~7 Å across, and it contains exchangeable cations and water molecules.

Synthesis and stability of these phases are highly dependent on the nature of the exchangeable cations (Mg²⁺ for OMS-1 and K⁺ for OMS-2). Many synthesis methods rely on novel sol–gel routes to tunnel (4) and layered (5) materials. In syntheses of large-tunnel 3 × 3 MnO₆ OMS-1 systems that have the todorokite structure (Figure 1), framework incorporation of Mg²⁺ is essential for obtaining high thermal stability (up to 600 °C) (6).

The conductivity of such systems is unique because it is difficult to produce conducting molecular sieve materials. Figure 2 shows the effects on conductivity of ion-exchanging various cations into synthetic cryptomelane, OMS-2. Note that the ion-exchange of OMS-2 with Ni²⁺, Cu²⁺, and Cr³⁺ into the channels leads to an overall difference in conductivity of more than one order of magnitude. The presence of small dopant levels of various transition-metal ions in all of these OMS and OL materials markedly influences the conductivity of such materials. This kind of behavior may be important for sensor applications (discussed later).

We have extended these syntheses to the preparation of semiconducting mixed-valence mesoporous manganese oxide materials (7). We also have reported the mechanisms of formation of mesoporous manganese oxide materials from oxidation, reduction, and layer conversions (9). These systems are clearly mesoporous (20–500 Å), in contrast to OMS materials that also have pores in the micropore (8–20 Å) range (13).

Argon adsorption data (10) clearly show that the larger (by a ratio of = 2:1) diagonal distances of the tunnels are accessible, meaning that 3 × 3 tunnels that are 6.9 Å on a side actually have a diagonal tunnel distance of 9.6 Å. This larger distance is observed in pore size distribution data. On the other hand, the organic molecule adsorption data for OMS-1 shown in Table 1 indicate that adsorbates having any dimension >6.9 Å are too large to fit in the tunnels. The data in Table 1 clearly show that access along the diagonal distance is hindered, which is probably caused by the presence of cations. These trends are observed for all tunnel systems and are critical for designing shape-selective catalytic materials. A broad micropore size distribution for Mg–OMS-1 that has been dehydrated at 500 °C in air is shown in Figure 3. These adsorption data show that the small argon gas molecule can be used to measure a broad pore size distribution that is centered around 6.9 Å, but there are a limited number of pores with sizes up to the diagonal distance of 9.6 Å in OMS-1.

Microwave and thermal preparations
Both 2 × 2 (cryptomelane type) (11) and 3 × 3 (OMS-1 todorokite type) (14) tunnel structure materials (Figure 1) can be prepared using microwave heating methods (11, 14). The OMS-1 materials prepared via microwave heating have different average oxidation states of manganese than materials made by conventional heating, which leads to different catalytic activity. To demonstrate this process, we prepared two sets of materials, using the same starting reagents and conditions, but we heated one set with microwaves and the other thermally. The thermal profiles were designed to mimic temperature profiles measured in the microwave apparatus.

We found that the microwave materials are much better total oxidation catalysts (e.g., converting benzene to CO₂). The only difference that can be measured between catalysts prepared by thermal and microwave synthesis methods is the average oxidation state of manganese, which seems to be ~3.7 for microwave materials and ~3.5 for thermally produced catalysts. Currently, we do not understand why microwave-prepared materials have less reduction of Mn⁴⁺ to Mn³⁺ and Mn²⁺, but this effect may be related to the heating mechanism. In microwave preparations, heating occurs from the inside to the outside of the particle, and active water in the preparation mixture may be heated preferentially. The thermally produced catalysts convert ethylbenzene to styrene, instead of oxidizing it totally to CO₂, as occurs when using materials made with microwave syntheses.

Synthesis and properties of small particles
The thermodynamic properties of such small-particle systems are of considerable interest. Calorimetry studies done in collaboration with Alex Navrotsky of the University of California, Davis, have shown that the layered materials (synthetic birnessite-structure materials, referred to as OL-1) have the lowest heats of formation among all manganese oxide systems (16). These data suggest that this phase is quite stable, which is in agreement with natural birnessite being the most abundant manganese oxide mineral and with our observation that birnessite usually crystallizes before other phases (17). Birnessite can be transformed from a layered structure into tunnel structure materials. We have exploited these thermodynamic data in the synthesis of quantum-size–effect manganese oxide systems (18) and other novel materials.

Our synthetic work has led to phase-transfer preparations of small particles (<200 Å) of colloidal layered materials of mixed-valence manganese oxides (18). These systems have shown excellent long-term stability at low temperatures, and the size of the particles can be controlled by aging conditions. An example of such novel quantum size effects is shown in Figure 4. The UV–vis absorbance maxima shift from 288 nm after 1 day aging to 362 nm after 3 weeks, compared with bulk commercial MnO₂, which absorbs at 425 nm. Such behavior is related to absorbance properties of small clusters of manganese oxide materials. Absorbances for small nanocluster materials are typically blue-shifted with respect to absorbances for their bulk counterparts. Because of this, UV–vis data can be used to obtain some idea of the relative sizes of the nanoclusters and can show how specific chemical and physical treatments influence particle sizes. Surface area, which is enhanced by producing small cluster sizes, can directly influence the rate of a reaction, such as in certain catalytic reactions, which are said to be structure-sensitive. The ability to control the size of such clusters is also important in the synthesis of novel materials.

These studies clearly show that it is possible to produce extremely small particulates of nanosize porous manganese oxide materials. We have studied the particle sizes and shapes with small-angle neutron scattering (SANS) methods in collaboration with several groups at Argonne National Laboratory. Our data suggest that these particles are as small as 60 Å during initial aging of the gel. Most recently, we have studied the structural properties of these and other materials with X-ray absorption methods in collaboration with Thorsten Ressler and Joe Wong of Lawrence Livermore Laboratories (19, 20). We have also carried out several in situ SANS, extended X-ray absorption fine-structure (EXAFS), and X-ray absorption near-edge spectroscopy (XANES) experiments. The X-ray absorption studies provided valuable information on the oxidation stats and the nature of the bonding in these solids.

Doping and pillaring
In some of our recent synthetic studies, we focused on finding concentration limits for doping metals in the framework and tunnel sites, especially for copper substitution (21). We also have begun to study the use of nonaqueous solvents in synthesis of these materials, for example, using alcohols to produce layered OL-1 (22). Another new direction involves phase transformations induced by microwave radiation, as in the production of OMS from OL systems via oxygen evolution (23). We have reported many of the synthetic aspects of our OMS and OL materials, as well as a survey of the literature in Reference 24.

Another recent synthetic avenue has been the preparation of pillared OL (POL) materials, in which the layers are held apart by molecules or molecular clusters that span the interlayer spaces. To pillar OL-1 systems, it is necessary to initially modify the charge density of the layered system by incorporating hexylamine and treating it with acid. Various cation oligomers can then be used to pillar OL-1 readily. Evidence of pillaring comes from X-ray diffraction (XRD) data and X-ray absorption studies (25). Examples of XRD data for pillared chromium oxide OL-1 compared with the unpillared OL-1 are shown in Figure 5. Such studies have shown that pillaring can lead to new porous materials that show enhanced thermal stability.

Engineering accessibility: Modifying the tunnels
It is possible to produce materials having asymmetric tunnel structures with dimensions of 1 × 2, 2 × 3, 2 × 4, and similar arrangements. The structures in Figure 6 clearly show that the diagonal distances across the rectangular pores are larger than those for the square pores. This dimension depends on the angle between the two intersecting manganese oxide layers, which may not be 90°. As the size and dimensions of the pore change, it may be possible to fine-tune the fit of a reactant into the tunnel and the accessibility of this reactant to specific tunnel sites.

In addition to modification of the general dimensions of such tunnels, it is also clear from Figure 6 that there are cations and water molecules present in these tunnels that may either facilitate or hamper diffusion of reactants, into, through, and out of these tunnels. The overall dimensions of these materials can be modified by substituting different cations into the framework structure. This is shown clearly by electron diffraction data for many of these systems. High-resolution electron microscopy data also suggest that these tunnels are somewhat unusual because they are not very long. These short tunnels provide rapid access of reactants to catalytically active sites.

Putting the materials to work
Most applications of OMS and OL systems rely on their outstanding porosity and potential for use in catalytic oxidations (2). Other applications include adsorbents, batteries, protective coatings, and sensors.

Adsorbents. OMS and OL materials are synthetic counterparts of natural manganese nodules, which form in oceanic environments. Because of the extremely large adsorption capacity of these nodules, the United States, The Republic of Korea, and Japan continue to mine them at various oceanic sites. Various researchers have observed the ability of porous manganese oxides to adsorb metals in high concentrations (26). One problem with these natural materials is that they contain many impurities and mixed phases, and it is often difficult to obtain uniform properties from materials mined from different oceanic sites. With the ability to selectively substitute framework and tunnel sites of OMS systems, selective adsorbent materials might readily be produced.

Batteries. Coin-shaped batteries are typically lithium-doped manganese oxides. There is considerable interest in secondary nonaqueous rechargeable batteries that are charged by intercalating lithium and discharged by de-intercalation. The enhanced porosity of OMS, OL, manganese oxide mesoporous structures (MOMS), and similar systems may allow faster intercalation, which might be useful in such battery systems.

Catalysts and catalyst supports. All of these systems are being studied in catalytic reactions. We are looking for correlations between the pore size of the tunnels and product distribution. We have a variety of microporous tunnel, layer, and mesoporous materials to choose from in such studies. We have also varied the organic adsorbates in many of these catalytic oxidation reactions to study shape selective effects.

Cryptomelane (OMS-2) materials are the most active of all the catalysts we have prepared for selective oxidations of cyclohexane (27, 28) and other adsorbates (29) that can fit into the 2 × 2 tunnel structure (diagonal distance ~6.4 Å). Product distributions such as those shown in Figure 7 clearly show that selective oxidations occur with these materials and that extensive oxidation (e.g., to benzene) and total oxidation do not occur. Several products, such as cyclohexanol, cyclohexanone, and cyclohexyl peroxide, are formed, as shown in Figure 7. These and other data clearly point to a shape-selective effect and to selective oxidations (30). However, the reaction mechanisms are quite complicated, because they are also influenced by radical and Brønsted acid pathways. Our goal is to design new OMS, OL, POL, and mesoporous catalysts in which shape-selective mechanisms can predominate.

Oxidations of CO with OMS-2 catalysts at low temperature (<25 °C) are very efficient. the catalysts show higher activity and stability than other materials reported in the literature (31). The photo-assisted catalytic decompositions of phosphine and phosphine oxide materials with amorphous manganese oxide materials also show promise. However, poisoning of the catalyst is a severe problem (32). The conversion of ethylbenzene to styrene can be carried out over doped OMS catalysts with good selectivity and activity (33). A recent fundamental study of OMS and OL systems has shown that OMS-2 is quite hydrophobic and can be used for total oxidations of aromatic hydrocarbons at temperatures (<200 °C) significantly lower than for other materials (34).

Our most recent studies have focused on the use of OMS and OL materials in other catalytic reactions, such as in H₂O₂ decomposition (34), selective oxidation of cyclohexane to cyclohexanol and cyclohexanone (35), and in the oxidative dehydrogenation of ethanol (36). The mesoporous MOMS-1 systems are good selective oxidation catalysts, producing cyclohexanol and cyclohexanone in high yields (7). The studies with cyclohexane are quite promising, with yields, conversions, and selectivities as high as any reported homogeneous catalyst. Some modeling studies of hydrocarbons in zeolites were also recently reported (8). We are currently extending these synthetic studies to other transition-metal systems that can form OMS-type structures such as tantalum defect pyrochlores (37) and others. In addition, four patents on synthesis and catalytic applications of these materials were issued in 1997 (38–41). A summary of various catalytic applications in the box Catalytic applications of OMS and OL materials clearly shows that different structures and compositions markedly influence catalytic activity of these systems.

The primary focus of catalytic reactions with manganese oxides has been in the area of oxidations. The overall average oxidation state of manganese and the composition (primarily the presence of other transition-metal ions) in these systems control the relative rates of these reactions. In addition, small-tunnel materials are usually more active and selective for such oxidations than large-tunnel materials. For example, CO oxidation is more efficient over the small-tunnel OMS-2 than over the large-tunnel OMS-1.

Another key feature of such catalytic reactions is that for different adsorbates, different transition-metal dopants are most active and selective. For example, in ethanol oxidation, cobalt ions are most active, whereas in isopropanol oxidation, copper ions are most active. Partially oxidized systems are the products of most of the oxidation reactions listed in the box. Mg–OMS-1 made by microwave methods is an exception because it effects the total oxidation of CO to CO₂. For many years, Mars–van Krevelen mechanisms have been proposed to explain the weak Mn–O bonds in oxidizing small adsorbates on the surfaces of manganese oxides. Loss of oxygen leads to sites for adsorption. The catalysts can be regenerated after oxidation by using oxygen in the feed.

Other reactions catalyzed by manganese oxides include decompositions of several reactants. Amorphous manganese oxide materials have been found to photochemically and thermally decompose phosphine oxide materials that are model complexes for chemical warfare agents. Various metal oxide OMS materials have also been shown to decompose H₂O₂. This method, often used for in situ oxygen generation in submarines and in space vehicles, can also be used for selective oxidations.

Some other reactions that are catalyzed by metal OMS systems are reductions of NO, hydrogenolysis of cyclopropane, and oxidation of CH₄. Each of these reactions is optimized by choosing a specific transition-metal dopant in OMS. Note that the small-tunnel OMS-2 catalysts are most effective for reduction and hydrogenolysis reactions; however, oligomerization reactions are favored by using large-tunnel OMS-1 materials. Although Fe–OMS-2 is active for reduction of NO, the activity does not reach that of other more active catalysts such as the zeolitic metal–ZSM-5 systems. We believe the hydrogenolysis of cyclopropane occurs over metal sites because hydrogen atoms must be provided to form the CH₄ product. Note that with this catalyst, any C₂ product initially formed after ring opening is totally cracked to CH₄. This behavior is somewhat unusual compared with other catalyst systems.

Detergents. Another application is in the degradation of dye molecules such as in stain removal by detergents (12). In collaboration with researchers at Clorox, we have shown that OMS and OL systems can be used to degrade a variety of dye molecules that are representative of stains. The degradation of stains is a difficult task because there are often a several types of stains with varying hydrophilic and hydrophobic properties. The pursuit of a material that can universally attack stains is of considerable practical importance.

Protective coatings. Bacteria have been found to assist manganese ions to migrate to the surface of sandstones to form a protective rock varnish coating. Such coatings are very thin, on the order of micrometers, and can take thousands of years to form. Depending on the amount of manganese that migrates or is present in the material, the color of the coating can change, going from brown to black as the manganese concentration increases. If OMS or OL materials could be made to imitate this bacterial action in a controlled fashion over a reasonably short time period, with predictable colors, it might be possible to use these coatings to protect surfaces such as those on sandstone buildings from acid rain and erosion.

Sensors. We have produced OMS and OL materials as thin films (3). These films are conductive and can be used as sensing devices for halogenated hydrocarbons. For example, trace levels of CHCl₃, CH₂Cl₂, and CCl₄ can be distinguished by the magnitude of changes in conductivity they induce in the sensor materials. Such selectivity is difficult to achieve with sensors, which points to the promise of finding other uses for the inherent conductivity of these OMS and OL systems. Still other applications are indicated by the influence of adsorbates on the conductivities of these films.

Manganese oxides: Materials by design
We have been very successful in synthesizing OMS and OL materials that have a variety of structures and compositions. Various substituents can be incorporated into framework, tunnel, and interlayer sites to produce novel porous systems. The ability to adjust pore sizes and interlayer spacings makes these materials useful as molecular sieves for separations and sorption applications. Fine-tuning the valence states of the active sites gives us the ability to perform selective catalytic reactions and to take advantage of conductivity changes for use in sensor applications. Ion diffusion through the channels and pores opens up potential applications for battery materials and self-generated coatings. Some OMS materials have been produced on a commercial scale by Engelhard Corp., which can be obtained by contacting Ralph Goshe (42).

Acknowledgments
This research has been supported by the U.S. Department of Energy, Office of Basic Energy Sciences, Division of Chemical Sciences. Much of the early research in this area was done in collaboration with C. L. O'Young of Texaco. I acknowledge many collaborators and helpful discussions with Abraham Clearfield, Mark E. Davis, Kathleen A. Carrado, Lennox E. Iton, Alexandra Navrotsky, T. J. Pinnavaia, Thorsten Ressler, and Joe Wong.

1. Suib, S. L.; O'Young, C. L. In Synthesis of Porous Materials; Occelli, M. L., Kessler, H., Eds.; Marcel Dekker: New York, 1997; pp 215–231.
2. O'Young, C. L.; Suib, S. L. Stud. Surf. Sci. Catal. 1997, 105, 189–196.
3. Segal, S. R.; Park, S. H.; Suib, S. L. Chem. Mater. 1997, 9, 98–104.
4. Ching, S.; Roark, J. L.; Duan, N.; Suib, S. L. Chem. Mater. 1997, 9, 750–754.
5. Ching, S.; Petrovay, D. J.; Jorgensen, M. L.; Suib, S. L. Inorg. Chem. 1997, 36, 883–890.
6. Tian, Z. R.; Yin, Y. G.; Suib, S. L.; O'Young, C. L. Chem. Mater. 1997, 9, 1126–1133.
7. Tian, Z. R.; Tong, W.; Wang, J. Y.; Duan, N.; Krishnan, V. V.; Suib, S. L. Science 1997, 276, 926–930.
8. Krishnan, V. V.; Bennett, C. O.; Suib, S. L. Appl. Catal., A 1997, 151, 267–287.
9. Luo, J.; Suib, S. L. Chem. Commun. 1997, 1031–1032.
10. O'Young, C. L.; Sawicki, R. A.; Suib, S. L. Microporous Mesoporous Mater. 1997, 11, 1–8.
11. Zhang, Q.; Luo, J.; Vileno, E.; Suib, S. L. Chem. Mater. 1997, 9, 2090–2095.
12. Segal, S. R.; Suib, S. L.; Foland, L. Chem. Mater. 1997, 9, 2526–2532.
13. Suib, S. L. Curr. Opin. Solid State Chem. 1997, 3, 63–70.
14. Vileno, E.; Ma, Y.; Zhou, H.; Suib, S. L. Microporous Mesoporous Mater. 1997, 20, 3–15.
15. Ching, S.; Suib, S. L. Comments Inorg. Chem. 1997, 19, 263–282.
16. Fritsch, S.; Post, J. E.; Suib, S. L.; Navrotsky, A. Chem. Mater. 1998, 10, 474–479.
17. Luo, J.; Suib, S. L. J. Phys. Chem., B. 1997, 101, 10403–10413.
18. Brock, S. L.; Sanabria, M.; Suib, S. L.; Urban, V; Thiyagarajan, P.; Potter, D. I. J. Phys. Chem., in press.
19. Ressler, T.; Brock, S. L.; Wong, J.; Suib, S. L. J. Phys. Chem., B 1999, 103, 6407–6420.
20. Ressler, T.; Brock, S. L.; Wong, J.; Suib, S. L. Synchrotron Radiat. 1999, 6, 728–730.
21. Nicolas-Tolentino, E.; Tian, Z. R.; Zhou, H.; Xia, G. G.; Suib, S. L. Chem. Mater. 1999, 11, 1733–1741.
22. Ma, Y.; Suib, S. L. Chem. Mater. 1999, 11, 1972–1979.
23. Zhang, Q.; Suib, S. L. Chem. Mater. 1999, 11, 1306–1311.
24. Brock, S. L.; Duan, N.; Tian, Z. R.; Giraldo, O.; Zhou, H.; Suib, S. L. Chem. Mater. 1998, 10, 2619–2628.
25. Ma, Y.; Suib, S. L.; Ressler, T.; Wong, J.; Lovallo, M.; Tsapatsis, M. Chem. Mater., in press.
26. Varentsov, I. M. Manganese Ores of the Supergene Zone: Geochemistry of Formation; Kluwer: NY, 1996.
27. Xia, G. G.; Wang, J. Y.; Yin, Y. G.; Suib, S. L. In Catalysis of Organic Reactions; Herkes, F. E., Ed.; Marcel Dekker: New York, 1998; pp 615–620.
28. Wang, J. Y.; Xia, G. G.; Yin, Y. G.; Suib, S. L. In Catalysis of Organic Reactions; Herkes, F. E., Ed.; Marcel Dekker: New York, 1998; pp 621–638.
29. Wang, J. Y.; Xia, G. G.; Duan, N.; Ma, Y.; Suib, S. L. In Shape Selective Catalysis, Chemicals Synthesis and Hydrocarbon Processing. ACS Symposium Series 738; C. Song, J. M. Garces, Y. Sugi, Eds.; American Chemical Society: Washington, DC, 1999; pp 80–94.
30. Xia, G. G.; Yin, Y. G.; Wang, J. Y.; Suib, S. L. J. Catal., in press.
31. Segal, S. R.; Suib, S. L.; Tang, X.; Satyapal, S. Chem. Mater. 1999, 11, 1687–1695.
32. Vileno, E.; Zhou, H.; Suib, S. L.; Corbin, D. R.; Koch, T. J. Catal., in press.
33. Luo, J.; Zhang, Q.; Huang, A.; Suib, S. L. Microporous Mesoporous Mater., in press.
34. Zhou, H.; Shen, Y. F.; Wang, J. Y.; Chen, X.; O'Young, C. L.; Suib, S. L. J. Catal. 1998, 176, 321–328.
35. Wang, J. Y.; Xia, G. G.; Yin, Y. G.; Suib, S. L.; O'Young, C. L. J. Catal. 1998, 176, 275–284.
36. Zhou, H.; Wang, J. Y.; Chen, X.; O'Young, C. L.; Suib, S. L. Microporous Mesoporous Mater. 1998, 21, 315–324.
37. Duan, N.; Tian, Z. R.; Willis, W. S.; Suib, S. L.; Newsam, J. M.; Levine, S. M. Inorg. Chem. 1998, 37, 4697–4701.
38. O'Young, C. L.; Sawicki, R. A.; Yin, Y.; Xu, W. Q.; Suib, S. L. U.S. Patent 5,597,944, 1997.
39. Suib, S. L.; O'Young, C. L. U.S. Patent 5,635,155, 1997.
40. O'Young, C. L.; Suib, S. L. U.S. Patent 5,702,674, 1997.
41. O'Young, C. L.; Suib, S. L. U.S. Patent 5,695,618, 1997.
42. Goshe, R. Engelhard Corp. Personal communication (216-360-5005; ralph_goshe@engelhard.com).

Return to Top