Studies in Surface Science and Catalysis 130 A. Corma, F.V. Melo, S.Mendioroz and J.L.G. Fierro (Editors) 9 2000 Elsevier Science B.V. All rights reserved.
B e c k m a n n R e a r r a n g e m e n t over A n i o n i c Clays L. Forni, G. Fornasari, R. Trabace, F. TrifirS, A. Vaccari a and L. Dalloro b a Dip. di Chimica Industriale e dei Materiali, Universith di Bologna, viale Risorgimento 4, 40136 Bologna, Italy. Fax +39-51-6443680 e-mail: [email protected]
ms. unibo.it b Enichem S.p.A., Centro Ricerche di Novara, via G. Fauser 4, 28100 Novara, Italy Beckmann Rearrangement reaction in gas phase has been tested over hydrotalcite - type (HT) compounds containing boron as active phase. Furthermore these materials are claimed to own basic properties which could help the desorption of heavy by-products . Main target of the work was to synthesise anionic clays keeping high quantities of borate. Many studies have been made in the attempt to introduce in the HT structure high boron quantities. It was important, most of all, to carefully adjust the pH values during the preparation of the catalyst in order to select the kind of anion to introduce in the structure of the HT compound. The effects of washing conditions were also studied. The results emphasised that catalysts washed in different conditions show a decrease in the content of boron and sodium, due to the increase of both temperature and volume of water, while the quantities of Mg and A1 remain nearly the same. To reduce the quantity of sodium and to increase the content of boron, the precipitates were then washed with different solutions of ammonium tetraborate. The catalytic behaviour of the samples was thus compared. 1. I N T R O D U C T I O N Due to the growing concern about environmental problems new technologies in the Beckmann Rearrangement reaction focused on utilising heterogeneous acid catalysts in gas phase are being investigated. These processes have not been applied yet because of the lower values of caprolactam selectivity, the fast deactivation rate of the catalyst and its hard regeneration conditions. The short catalyst lifetime can be due both to the deposition of tars on its surface and to the partial leaching of the active phase. The solid acids investigated are hydrotalcite - type (HT) anionic clays, which are characterised by a layered structure, where the positive charge in the Mg/A1 hydroxide sheets is neutralised by boron oxides present in the interlayers
3514 together with water molecules. It is also claimed that borates own the acid sites of intermediate Bronsted strength which promote caprolactam selectivity . Therefore on one hand the introduction of borates into the HT structure should: lock the active phase, decrease its partial loss by leaching phenomena and provide the catalyst with the right acidity for the rearrangement to the desired product. On the other hand it seems that cationic sheets could help in some way the desorption of the main product from the surface of the catalyst and consequently reduce the formation of tars . Finally the structural stability of the mixed oxide phase is another reason for its use for regeneration processes. In the attempt of introducing in the HT structure high boron quantities, the effect of mixing, pH values, and washing of the precipitate by water or ammonium tetraborate, was investigated. It was found that all the above mentioned parameters are important to achieve the expected results, but the most important is the pH regulation which allows to select the most suitable anion. The catalytic activity of the samples was then compared and related to the boron content. 2. E X P E R I M E N T A L Catalysts were prepared by coprecipitation of a solution of Mg(NO3)2 . 6H20 and Al(NO3)3 . 9H20 (Mg/A1 = 1.9) in an excess of aqueous solution containing boric acid. NaOH 0.3 N was utilised to adjust and select the pH of the solution. The precipitate was kept under mixing 15 hours long, then filtered and washed by water or a solution of ammonium tetraborate, dried under vacuum and calcined at 400~ for 4 hours. Boron elemental analysis was made by mean of a spectrophotometer Uvikon 860, while magnesium and aluminium elemental analysis by a spectrophotometer for atomic absorption Philips PU 9100. XRD was carried out by a Philips diffractometer PW 1050/81. Catalysts surface area was measured by Sorpty Carlo Erba 1700. Beckmann Rearrangement was carried out in a glass fixed bed reactor having an internal diameter of 4 mm. Reaction conditions were: temperature = 350~ atmospheric pressure, WHSV =1,21 h -1. A solution of cyclohexanone oxime (4% mol) in toluene (44% mol) was supplied by a micro-feeder at a constant flow rate. Nitrogen (52% mol) was utilised as carrier. The products were determined by a GCD gascromatograph with electron ionisation detector then analysed by a GC with flame ionisation detector. 3. R E S U L T S AND D I S C U S S I O N S In the attempt to introduce in the HT structure high boron quantities many studies have been made. First of all it was noticed that an efficient mixing brought to higher boron contents. In fact the preparation of a catalyst was repeated in different mixing conditions; the materials obtained by mixing the solutions with magnets showed a very low boron content (B/A1 = 0.20) while the materials obtained using a mechanical glass arm showed B/A1 ratios equal to
3515 0.90 ca. To further increase the amount of boron, it is possible to introduce in the structure of the material anions containing 2 or 3 boron atoms for each negative charge. Each aluminium ion present in the HT introduces a positive charge in the cationic sheet, indeed. It should, then, be neutralised by the negative charges of borates. During the preparation, the control of pH values should allow to determine the boron anion present in the solution. Thus the desired anion should be intercalated between the growing sheets. According to a previously described method  a scheme of the dependence of the kind of anion on the pH is described in Table 1. Table1 Dependence of the anion on pH value Anion pH B303(OH)47 . 5 - 9.5 B405(OH)42" 8 . 5 - 9.5 B(OH)4>10.5
T e m p e r a t u r e (~ 70 80 85
The influence of pH of the solution on the obtained compound both by structural and compositional point of view was studied. With the purpose of introducing in the HT structure different kind of anions, several preparations were carried out at different pH, in a range between 7.5 and 11 and maintaining a constant value during the precipitation, pH regulation showed a relevant effect.
Fig. 1. HT compounds with different anions in the interlayers depending on the pH of the reaction. A: pH= 11 (*) HT like phase with CO32- in the interlayer. B: pH=9 (.) HT like phase with B303(OH)4-in the interlayer, (.)HT like phase with B303(OH)4- or B(OH)4- in the interlayer. C: pH=7.5 (.) HT like phase with B303(OH)4-in the interlayer.
0. . . . . . .
Fig. 2. (A): HT compounds with borates in the interlayer washed with 600 cc of a solution of ammonium tetraborate (.) HT like phase with B4Os(OH)42-; (,) phase (NH4)2B4OT.4H20. (B): catalyst A washed with 100 cc of water. (.)HT like phase with B4Os(OH)42- in the interlayer, (.) HT like phase with B303(OH)4- or B(OH)4 in the interlayer.
3516 As a m a t t e r of fact at pH = 7.5 XRD analysis showed a prevalently amorphous m a t e r i a l with a m i n i m u m a m o u n t of HT (Figure 1). By elemental analysis a little a m o u n t of boron was present, even though it was not possible to define which kind of species. At higher pH values the obtained materials showed a HT phase containing borates. At pH = 9 a multiphase sample was identified. It seems t h a t two different anions such as B303(OH)4- or B(OH)4 are present within the structure. An increase of p H up to 11 gave rise to the formation of a HT compound containing carbonates. This phenomenon was due to the contamination by CO2 present as impurity in NaOH, which is added in large quantities to adjust the pH. However the prevalent phase contains borates as anions, like B(OH)4- or B303(OH)4-. Furthermore concentration changes of the NaOH solution did not lead to a significant decrease of the HT carbonates content. The calcined samples showed an evolution of the HT phase to a low cristallinity phase. The presence of anions other t h a n carbonates yields the HT phase more stable and, as a consequence, the formation of the oxide phase is more difficult. The surface area of all these samples was relatively low (about 20 m2/g). Nevertheless on the basis of elemental analysis it was evidenced that, if only B303(OH)4- anion was present, the theoretical B/A1 ratio of 3 was not achieved. As a m a t t e r of facts the values varied between 1.8 (pH=9) and 1.6 (pH=11), except for the prevalently amorphous material, in which the a m o u n t of boron was very low B/A1 = 0.1. In order to see whether the washing step shows any effect on boron content, HT compounds were washed by water. In fact one of the cons of the preparation method was the use of NaOH to control the pH range. The presence of Na ions in the catalyst is dangerous for the reaction, because it brings to tars formation and, consequently, to a quick catalyst deactivation. Catalysts washed with different methods showed t h a t increasing both the t e m p e r a t u r e and the volume of water, boron and sodium content decreased, while the quantities of Mg and A1 remained nearly the same. Table 2 shows how the composition and sodium content vary with t e m p e r a t u r e and volume of water. Table 2. Dependence of the composition and sodium content on the t e m p e r a t u r e and volume of w a t e r Temp Water Vol. Mg/AI B/A1 Na/AI Na Surf. Area (~ (cm 3) 10~mol/l (m2/g) / / 1.67 1.87 1.76 576.3 32 60 200 1.86 1.88 0.08 26.1 69 60 400 1.90 1.99 0.12 38.0 56 60 600 1.96 1.80 0.04 13.7 37 60 1000 1.99 1.42 0.009 2.7 10 90 400 1.93 1.77 0.06 18.6 38
3517 As it can be seen, by increasing both the volume and the temperature of water, sodium content substantially decreased, nevertheless at least a decrease in boron content is observed. Furthermore by XRD analysis it was noticed that by increasing the volume of water, owing to their highest affinity to Mg +§ ions in the cationic sheet, anions such as CO32- were introduced into the HT structure. Therefore, catalysts need to be washed with an adequate volume of water, by this procedure it is possible to reduce significantly the sodium content without decreasing too much the amount of boron. Thus in order to reduce the quantity of sodium and to try to increase the content of boron, the precipitates were washed with different solutions of ammonium tetraborate. It was found that the amount of boron was higher than that forecast for a structural insertion (see Table 3). Table 3 Dependence of the composition and ammonium tetraborate Water (NH4)2B407 Mg/A1 (cm 3) 0,5 M (cm 3) 600 / 1.43 / 600 1.45 100 600 1.45
boron content on the washing with B/A1
1.05 5.38 3.33
0.0011 0.0051 0.0030
B 10 .3 mol/g 4.02 12.0 8.54
Surf. Area (m2/g) 62 27 122
In this case elemental analysis was carried out by ICP. It can be seen that by washing a portion of the sample with 600 cc of water the B/A1 ratio is only 1.05: quite far from the theoretical value for a B303(OH)4- anion. Actually by XR diffraction it was evidenced that the sample is characterised by a low cristallinity so it is not possible to exclude the presence of small quantities of other anions. When the sample is washed by 600 cc of a solution of ammonium tetraborate the ratio (B/Al=5.38) is much higher than that it can be theoretically introduced in the layers. This result was confirmed by XRD spectra. In figure 2 the patterns of the two samples washed with ammonium tetraborate are reported and two different phases are evidenced: HT with borates in the interlayer and crystals of tetraborate. In the sample washed also by a small amount of water, a substantial decrease in boron content was observed (B/A1 = 3.33), confirming that not all the boron was stabilised in the HT structure. Catalytic behaviour of coprecipitated samples washed with ammonium tetraborate and of the other ones washed only with water were compared. The first kind of catalyst has higher values of conversion of cyclohexanone-oxime which starts from 100% and gradually ends within 9 hours at 20%. The catalyst washed only with water shows a starting value of conversion of 70% and quickly decreases to 10% (Figures 3 and 4). Caprolactam selectivity starts for the first sample at 80% and slowly decreases to about 35% while the second sample has a starting selectivity of 15%. For what concerns the cyclohexanone selectivity it is lower when the content of boron is higher. This means that the presence of a higher quantity of boron and of easier access improves not only the conversion of
0 3 5 7 9 T i m e on s t r e a m (h)
Fig.3 Oxime conversion for: (A) sample washed with ammonium tetraborate; (B) sample washed with the same volume of water.
3 5 7 Time on s t r e a m (h)
Fig. 4 Caprolactam selectivity for: (A) sample washed with ammonium tetraborate; (B) sample washed with the same volume of water.
the catalyst but also its selectivity and is also responsible for the decrease in the formation of by-products. 4. C O N C L U S I O N S In order to prepare, by coprecipitation, HT compounds with high boron contents, pH regulation was found to be the most important parameter in catalyst preparation. In fact it was possible to select the kind and the amount of boron species to be introduced in the structure, even if it was difficult to introduce a single species of anion in the interlayer. The best results in boron content were obtained by adjusting the pH at intermediate values (pH = 9). The materials needed to be washed with an adequate volume of water, in order to significantly reduce the sodium content, but not to decrease too much the amount of boron. A high boron content leads both to higher caprolactam selectivity and cyclohexanone-oxime conversion. Borates deposited on the surface of the catalyst were easily attainable by washing with ammonium tetraborate and gave more satisfactory results in terms of catalytic activity than those of the HT structure catalyst.
REFERENCES 1. Y. Izumi, S.Sato, K. Urabe, Chem. Lett. (1983) 1649. 2. Aucejo, M.C. Burguet, A. Corma, V. Fornes; Appl. Cat. 22 (1986) 187. 3. A. Bhattacharyya et al., E.P.0 536 879 (1993) assigned to Amoco Corporation.