Zeolite dehydroxylation

The scheme proposed for the reaction over HFAU was that PA dissociates in phenol (P) and ketene and that o-HAP, which was highly favoured over the para isomer, results partly from an intramolecular rearrangement of PA, partly from acyl group transfer from PA to P whereas p-HAP results from this latter reaction only. In these experiments, the zeolite deactivation was very fast, as a result of coke deposition and zeolite dehydroxylation. Catalyst stability can be considerably improved by operating at lower temperatures and especially by substituting equimolar mixtures of PA and water or P and acetic acid for PA. Much higher HAP yields were obtained by using the P - acetic acid mixture as reactants.[68]... 

With H-MFI, the p-/o-HAP ratio was much higher this is indicative of shape-selectivity effects. With all the catalysts, HAP selectivity was poor, phenol being the main product because of the rapid dissociation of PA [9,10]. Very fast deactivation as a result of coke deposition and zeolite dehydroxylation was also observed. Catalyst stability can, however, be considerably improved by use of equimolar mixtures of PA and water or of phenol and acetic acid (AA) instead of PA [11]. 

The present acid Y-zeolite dehydroxylated at 650°C (in vacuo), more than 100°C higher than the NH4-HY system (4), probably because of the difference in site density between the 2 Y samples. A similar difference in thermal stability (toward dehydroxylation) was reported (28) between 50 and 90% ammonium-exchanged NaY. 

Figure 3.53. IR transmission absorption spectrum of pyridine adsorbed on partly dehydroxylated HY zeolite (Van Bekkum et al, 1991) B = Bronsted acid sites L = Lewis acid sites.

Fig. 30. The ESR spectra of 7-irradiated Y-type zeolites that were dehydroxylated at 600° prior to irradiation (a) A1HY, (b) NH,Y.

Pyridine sorption studies have shown the presence of both Bronsted and Lewis acid sites in USY zeolites, although to a lesser extent than in the corresponding HY zeolite (51,53). Acidity is maintained even after strong dehydroxylation of USY-B at 820°C. Rehydration of the calcined material did not regenerate significantly Bronsted acid sites, due to irreversible changes in the zeolite framework (51). 

Lewis acid sites may be formed following dehydroxylation of zeolite surface in H-form. At sufficiently high temperatures two Bronsted acid sites can drive off a water molecule and leave behind a coordinatively unsaturated Al site, as illustrated in Figure 13.16 [32]. Here not only the resulting tri-coordinated Al but also the tri-coordinated positively charged Si can act as a Lewis acid. Furthermore dehydroxylation may be followed by framework dealumination, leading to cationic extra-framework species like AlO AlfOHij that can act as Lewis acids [33-37]. 

NHj adsorption microcalorimetry was used by Shannon et al. [225] to follow the changes in acid sites of a H Y zeolite during dehydroxylation, framework dealumina-tion, and the formation of nonframework aluminum species. 

It is known that the activation temperature can influence the acid strength distribution. For example, measurements of the differential heats of ammonia adsorbed at 150°C for a HY zeolite have led to the conclusion that stronger acid sites, in the 150-180 kJ/mol range, are formed upon increasing the activation temperature from 300 to 650°C. Dehydroxylation at high temperature resulted in the formation of strong Lewis acid sites and the disappearance of intermediate and weak Brpnsted sites [62]. 

An additional complexity arises from the dissociation of water molecules which occurs when alkaline-earth-exchanged zeolites are thermally activated since several modes of dehydroxylation are possible. This problem has been extensively investigated by IR spectroscopy, in particular by Ward (268,269) and Uytterhoeven et ai (270), and by X ray (271). They concluded that the electrostatic field associated with the cation causes dissociation of adsorbed water to produce acidic hydroxyl groups. The dissociation reaction may occur according to the following reactions ... 

Later, Cattanach, Wu, and Venuto did an elaborate thermogravi-metric study on the calcination of ammonium zeolite Y and the resulting products (19). They found that the hydrogen zeolite reacted with anhydrous ammonia to yield an ammonium zeolite identical in ammonia content with the initial ammonium zeolite. Further, these workers reported that after loss of chemical water ( dehydroxylation according to Uytter-hoeven, Christner, and Hall or decationization according to Rabo, Pickert, Stamires, and Boyle) the sample became amorphous when exposed to moisture. This observation conflicted with the statement of Rabo et al. (16) in which they emphasized the extreme stability of their decationized Y. The data of Cattanach, Wu, and Yenuto prove, beyond any doubt, that they obtained the expected normal hydrogen zeolite Y prior to the loss of chemical water above 450°. Rabo et al., however, did not prove that the material from which they removed chemical water, was in fact, the hydrogen zeolite. They probably prepared, unknown to them at the time, the ultrastable zeolite described below. 

The chemistry and structure of the hydrogen form of zeolite Y have been thoroughly investigated 82) and are not considered further. The structure of the dehydroxylated zeolite proposed by Uytterhoeven, Christ-ner, and Hall 15) remains unchanged. Recently Ward, on the basis of infrared studies, suggested that this form may be amorphous 27). The extreme instability of dehydroxylated zeolite Y to moisture complicates detailed study 19). The elucidation of the detailed nature of this material lies in the future. At present, completely dehydroxylated Y is little understood and presents a challenging void in our knowledge of the nature of ammonium zeolite Y thermal decomposition products. 

Ambs and Flank correctly observed that variables can be introduced into the calcination of ammonium Y so that a variable series of products can be obtained 33). However, there is no doubt that the normal hydrogen zeolite can be obtained from the ammonium form by carefully controlled calcination. In addition, carefully controlled calcination of the acid yields the dehydroxylated form. The ultrastable form, which can be prepared by a number of procedures described below, differs drastically in stability and composition from the other two forms. That it may contain some sites similar to, or perhaps identical with, sites in the hydrogen and dehydroxylated forms cannot be refuted. Unquestionably, however, the ultrastable form differs significantly from the other two forms. 

The relationship between acid site density and effective acidity may account for the interesting observation of Hopkins that maximum cracking activity of n-hexane was obtained over a partially dehydroxylated hydrogen zeolite Y (45). While the normal hydrogen form would contain a greater overall concentration of acid sites, the partially dehydroxylated form may have a greater overall acid activity because of the increased effective acidity of the remaining sites. 

Cince the catalytic activity of synthetic zeolites was first revealed (1, 2), catalytic properties of zeolites have received increasing attention. The role of zeolites as catalysts, together with their catalytic polyfunctionality, results from specific properties of the individual catalytic reaction and of the individual zeolite. These circumstances as well as the different experimental conditions under which they have been studied make it difficult to generalize on the experimental data from zeolite catalysis. As new data have accumulated, new theories about the nature of the catalytic activity of zeolites have evolved (8-9). The most common theories correlate zeolite catalytic activity with their proton-donating and electron-deficient functions. As proton-donating sites or Bronsted acid sites one considers hydroxyl groups of decationized zeolites these are formed by direct substitution of part of the cations for protons on decomposition of NH4+ cations or as a result of hydrolysis after substitution of alkali cations for rare earth cations. As electron-deficient sites or Lewis acid sites one considers usually three-coordinated aluminum atoms, formed as a result of dehydroxylation of H-zeolites by calcination (8,10-13). 

Since maximum reducing and oxidizing power in the zeolite requires activation temperatures around 600° C, dehydroxylation is necessary for the formation of the active centers. Electropositive and electronegative sites produced as shown below may be responsible. On this basis the H... 

When zeolite H-Y obtained by decationation of NH4-Y is heated further, water is irreversibly lost from the framework. The dehydroxylated zeolite Y displays Bronsted and Lewis acid properties. The mechanism for this process... 

The nature of the surface acidity is dependent on the temperature of activation of the NH4-faujasite. With a series of samples of NH4—Y zeolite calcined at temperatures in the range of 200° to 800°C, Ward 148) observed that pyridine-exposed samples calcined below 450°C displayed a strong infrared band at 1545 cm-1, corresponding to pyridine bound at Brpnsted (protonic) sites. As the temperature of calcination was increased, the intensity of the 1545-cm 1 band decreased and a band appeared at 1450 cm-1, resulting from pyridine adsorbed at Lewis (dehydroxylated) sites. The Brtfnsted acidity increased with calcination temperature up to about 325°C. It then remained constant to 500°C, after which it declined to about 1/10 of its maximum value (Fig. 19). The Lewis acidity was virtually nil until a calcination temperature of 450°C was reached, after which it increased slowly and then rapidly at calcination temperatures above 550°C. This behavior was considered to be a result of the combination of two adjacent hydroxyl groups followed by loss of water to form tricoordinate aluminum atoms (structure I) as suggested by Uytterhoeven et al. 146). Support for the proposed dehydroxylation mechanism was provided by Ward s observations of the relationship of Brpnsted site concentration with respect to Lewis site concentration over a range of calcination tem-... 

The relationship of Lewis and Brpnsted acid site concentrations on H—Y zeolite was explored further in a study by Ward (156) of the effect of added water. At low calcination temperatures (<500°C) only a small increase in the Brpnsted acid site concentration occurred upon addition of water to the sample. Rehydration of samples dehydroxylated by calcination above 600°C resulted in a threefold increase in the amount of Brpnsted-bound pyridine. However, no discreet hydroxyl bands were present in the infrared spectrum after rehydration. Thus, the hydroxyl groups reformed upon hydration must be in locations different from those present in the original H—Y zeolite, which gave rise to discreet OH bands at 3650 and 3550 cm-1. 

These results strongly pointed toward the involvement of the acidic hydroxyl groups in the catalytic reaction as suggested by Benesi (157), since the maximum activity was obtained when the zeolite was completely deammoniated. In addition, catalysts which had been dehydroxylated by high-temperature calcination demonstrated low activity. Thus, Benesi proposed that the Brtfnsted acid sites rather than the Lewis acids were the seat of activity for toluene disproportionation. This conclusion was supported by the enhancement in toluene disproportionation activity observed when the dehydroxylated (Lewis acid) Y zeolite was exposed to small quantities of water. As discussed previously, Ward s IR studies (156) indicated a substantial increase in Brdnsted acidity upon rehydration of dehydroxylated Y sieve. 

Infrared spectra of alkaline earth-exchanged Y sieve containing adsorbed pyridine showed that the bands at 3650 and 3550 cm-1 arise from hydroxyl groups which function as Brpnsted acids (151, 209, 211). The zeolites undergo dehydroxylation upon treatment at temperatures in the range of 400° to 600°C. Experiments with Ca—Y zeolite by Eberly (151) indicated that extensive dehydroxylation occurred upon heating to 427°C. There was little evidence of hydroxyl groups in the 3800 to 3500 cm-1... 

Alkaline earth-exchanged samples examined by Ward (211) were more resistant to thermal dehydroxylation hydroxyl bands were present in the spectra after dehydration at 500°C. The concentration of OH groups was, however, much smaller than found in H—Y zeolite, and was dependent on the cation type. An almost linear inverse relationship was found between the alkaline earth cation radius and the concentration of acidic hydroxyl groups (210). 

---