Ammoximation, of cyclohexanone

In this section we report a detailed summary of the experimental studies on the interaction of TS-1 with H2O and NH3. The choice of these two molecules is far from random. Interaction with water is important since the catalyst works in aqueous solution (Sect. 2). The interest in the study of NH3 is twofold ammonia is a reactant in the ammoximation of cyclohexanone to give cyclohexanone oxime and it is a stronger base than water, thus allowing a direct comparison between the effects induced by Lewis bases of increasing strength. 

The TS-1 catalysed ammoximation of cyclohexanone with NH3/H2O2 is a new process (Romano et ai, 1990) for the production of cyclohexanone oxime, the precursor of caprolactam. In the existing process, the oxime is produced by reaction of cyclohexanone... 

The oxidation of NH3 to NH2OH forms the basis of a process for the ammoximation of cyclohexanone to the oxime because the NH2OH formed in solution readily reacts with the ketone (non-catalytically) to give the oxime (231). Table XXX (165) illustrates the conversions and selectivites obtained for a few typical ketones and aldehydes. The ammoximation of aldehydes is faster than that of ketones. The oxime selectivity is also higher. The ammoximation of cyclohexanone by this method offers a more eco-friendly alternative route to the cyclohexanone oxime intermediate for the production of Nylon-6. The current route coproduces large quantities of ammonium sulfate and involves the use of hazardous chemicals such as oleum, halides, and oxides of nitrogen. 

The TS-l catalyzed hydroxylation of phenol to a 1 1 mixture of catechol and hydroquinone has already been commercialized by Enichem. Another reaction of considerable commercial importance is the above mentioned ammoximation of cyclohexanone to cyclohexanone oxime66, an intermediate in the manufacture of caprolactam. It could form an attractive alternative to the established process that involves a circuitous route via oxidation of ammonia to nitric acid followed by reduction of the latter to hydroxylamine (figure 4). 

There is, as is well known, a close similarity between the crystalline and porous structures of silicalite-1 and silicalite-2. The same similarity therefore exists between TS-1 and TS-2, and it appears logical that they should have very similar catalytic properties. TS-2 has been evaluated as a catalyst for many different reactions, such as Beckmann rearrangement of cyclohexanone oxime with vapor-phase reactants H202 oxidation of phenol, anisole, benzene, toluene, n-hexane, and cyclohexane and ammoximation of cyclohexanone. As described in detail in Section V.C.3, differences that had been claimed between the catalytic properties of TS-1 and those of TS-2 have not been substantiated. Later investigations have shown that, when all the relevant parameters are identical, the catalytic activities of TS-1 and TS-2 are also identical. The small differences in the crystalline structure between the two materials have no influence on their catalytic properties (Tuel et al., 1993a). 

The ammoximation of cyclohexanone with NH3 and 02, which had been investigated with silica catalysts (Armor et al., 1979, 1980, 1982), has been investigated with TS-1 as the catalyst (Dreoni et al., 1991). However, the results have not been considered satisfactory for industrial exploitation, especially when... 

The ammoximation of cyclohexanone had been known before the discovery of TS-1, but the performances of conventional catalysts were far below the standards required for development work. In the EniChem process, the reaction is carried out in the liquid phase, at ca. 80°C, using a suspension of TS-1 in aqueous t-butanol, with a slight excess of hydrogen peroxide over the ketone. The substrate and the oxidant undergo total conversion with selectivities close to 98% and 94%, respectively. Inorganic by-products comprise minor amounts of ammonium nitrate and nitrite, N2O, and N2 produced by the oxidation of ammonia, and O2 by the decomposition of the oxidant. 

A process based on the ammoximation of cyclohexanone and on the catalyzed rearrangement of the oxime went on stream in Japan in 2003 it allows the salt-free production of s-caprolactam, at lower investment and operating costs than by conventional routes. 

The truly innovative nature of the EniChem process over earlier ones is apparent. The preparation of hydroxylamine in the latter case necessitates multi-step operation, often ending with major co-production of inorganic salts, as in the Raschig process. The ammoximation of cyclohexanone, with its in situ generation of the intermediate, reduces significantly the investment and operahon costs while improving the environmental compahbility. The new process represents a good example of how to combine profitability and environmental concern. 

Ti, V and Sn-modified mesoporous silicates were reported to be active in a number of liquid phase oxidation reactions. Ti-containing samples were used for the selective oxidation of large organic molecules in the presence of te/t-butyl hydroperoxide (TBHP) or dilute H2O2 [71,136,137,139-141,147,186,237]. Typical data shown in Table 5 indicate that both Ti-MCM-41 and Ti-HMS are efficient cat ysts for the epoxidation of bulky olefins such as a-terpineol and norbomene in the presence of TBHP or H2O2. Comparison with H-B indicates that the accessibility of active sites plays a critical role in the liquid phase oxidation of organic molecules. Mesoporous titanosilicates also exhibited remarkable activity in the hydroxylation of 2,6-di-rerr-butyl phenol (2,6 DTBP) [142,147] and the oxidation of cyclododecanol [147], naphthol [147] aniline [237] and chloroaniline [186]. However, they were disappointingly poor catalysts for the liquid phase oxidation of n-hexane and aliphatic primary amines, as well as the ammoximation of cyclohexanone [147,238]. 

The TS-1 catalyzed hydroxylation of phenol to a 1 1 mixUne of catechol and hydroquinone has been commercialized by Enichem. Similarly, the ammoximation of cyclohexanone is being developed commercially as a low-salt alternative to the conventional process for the production of cyclohexanone oxime, the raw material for nylon-6. The reaction involves initial TS-1 catalyzed oxidation of NH3 by HjOj to give NH2OH. The fact that bulky ketones such as cyclododecanone undergo ammoximation is consistent with subsequent reaction of NHjOH with the ketone substrate taking place outside the molecular sieve. The method has been used [49] for the conversion of p-hydroxyacetophenone to the corresponding oxime which is the precursor of the analgesic paracetamol (Reaction 13). 

The presence of a silica framework with few defects makes TS-1 a highly hydrophobic material suitable for oxidations in the liquid phase with H2O2 as oxidant. Thus, TS-1 has proven to be successful in the oxidation of alcohols, epox-idation of linear olefins, hydroxylation of aromatics, ammoximation of cyclohexanone, oxidation of alkanes to alcohols and ketones, oxidation of amines, oxidation of sulfur-containing compounds, and oxidation of ethers [66-75]. 

Figure 2.30. Comparison of the experimental data for ammonia adsorption on titanium silicate and that calculated using the quasi-logarithmic adsorption isotherm (N.V. Kul kova, M.Yu.Kvyatkovskaya, D.Yu.Murzin, Liquid-phase ammoximation of cyclohexanone. 3. Ammonia adsorption on ammoximation catalysts, Khimicheskayci Promyshlennost, (1997), 28).

Heterogeneously catalyzed ammoximation of cyclohexanone with molecular oxygen in vapor phase... 

The ammoximation of cyclohexanone to cyclohexanone oxime using TS-1 was invented in 1987 (221). The limited amount ofpubHshed work focuses on the mechanism (222) and the kinetics (223) of this reaction on titanium silicate with dilute hydrogen peroxide as oxidant. There is no agreement thus far regarding the reaction pathway. The following two hypotheses have been formulated (7,90,222b,224) ... 

Several different titanium-containing zeotype structures have been invest ted for the ammoximation of cyclohexanone, for example TS-1 (221,224), TS-2 (177,226), Ti-beta (227), Ti-MOR (228), Ti-MCM-41 (229), and Ti-MCM-48 (227a). Le Bars et al. (227a) evaluated zeotype materials with various pore sizes and concluded that the best performance and highest turnover in ammoximation is achieved with TS-1 as catalyst. The superiority of this material is believed to arise from its hydrophobicity, in combination with the three-dimensional framework of straight charmels that facilitate the diffusion of reactants and products in the catalyst pores. 

The direct ammoximation of cyclohexanone has been performed with hydrogen peroxide and ammonia using a catalyst based on TS-1 (5a). Good performance was obtained, with water as the only side product. EniChem built a 12,000 t/yr demonstration plant (300), which has been running successfully since 1994. Sumitomo built a unit with a capacity of60,000 t/yr in 2003 and both companies plants have operated without any major problems. (232) Sumitomo (301) commercialized its process, in combination with a zeohte-catalyzed Beckmann rearrangement of cyclohexanone oxime... 

Thirty years after the invention of ENI s TS-1, new environmentally friendly processes have been commercialized for the epoxidation of propylene and the ammoximation of cyclohexanone that have no major by-products or coproducts. The BASF-Dow and Degussa-Uhde processes claim to be environmentally friendly, have low capital investment costs, and have no significant quantities of by-products when compared to either the chlorohydrin or the PO with styrene monomer processes. The Sumitomo PO process is also beneficial for the environment, because it needs only a cumene-to-cumene hydroperoxide reactivation process step. 

Both academic and industrial research and development, as well as cooperation between industries, has resulted in a large effort in the discovery, synthesis, and catalytic application of framework metal-containing zeotype materials. These collaborations have led to new appHcations of metal-containing zeotype materials for the large-scale production of chemicals without major by-product formation. New environmentally friendly processes such as propylene epoxidation, phenol hydroxylation, ammoximation of cyclohexanone, and aromatization of light paraffins have been commer-ciahzed. Many new developments are in the pipehne, and they will Hkely be commerciahzed when both the economics and the environmental requirements become favorable. 

---