Materials Research Bulletin
Volume 36, Issues 1-2
January 2001
Pages 315-322 

PII: S0025-5408(00)00472-4
Copyright (C) 2001 Elsevier Science Inc. All rights reserved.

The ionic size of metal atoms in correlation with acidity by the conversion of cyclohexanol over MeAPO-5

Hadi Nur, , a and Halimaton Hamdana

a Department of Chemistry, Faculty of Science, Universiti Teknologi Malaysia, 81310 Skudai, Johor, Malaysia

Received 5 May 2000; accepted 8 May 2000. Refereed. Available online 9 April 2001.

Abstract

The conversion of cyclohexanol was investigated over MeAPO-5 [Me = Mg, Co, Mn and Zn]. The acidity of the catalysts was correlated with the ionic radii of Me atoms incorporated into the structure of the catalysts. Based on this correlation, it is suggested that the distortion of structure by the larger Me atoms was associated with the T-O-P angle and hence the acid strength.

Author Keywords: A. Microporous materials; D. Catalytic properties

Article Outline

1. Introduction
2. Experimental
3. Results and discussion
3.1. Physical properties
3.2. Catalytic activity of MeAPO-5
4. Conclusions
References


1. Introduction

Crystalline microporous aluminophosphates (AlPO4-5) [1] contain frameworks of alternating AlO2- and PO2+ tetrahedra which are electrically neutral; showing no Bronsted acidity. In order to produce Bronsted acidity, an anionic framework has to be generated by substituting an n-valent tetrahedral framework cation, with an (n-1) valent tetrahedral cation. In our previous paper [2], it was shown that the conversion of cyclohexanol is a good pointer to evaluate the Bronsted acidity and basicity of metal-substituted AlPO4-5 (MeAPO-5). Despite the numerous reports on the correlations between structure and activity of AlPO by theoretical calculations [3, 4 and 5], experimental evidence of the correlation has never been reported. Furthermore, reports on the incorporation of the various divalent metals into AlPO framework are often varied and inconsistent. Previously, it was demonstrated that metal-substituted AlPO4-5 is a potential catalyst for dehydration of alcohol [6]. In this work, the conversion of cyclohexanol over MeAPO-5 [Me = Mg, Co, Mn and Zn] was studied in order to correlate the structure, ionic size of Me atoms with acidity of MeAPO-5 by XRD, solid-state 31P MAS NMR, XRF and NH3-TPD.

2. Experimental

AlPO4-5 and all MeAPO-5 [Me = Mg, Co, Mn and Zn] prepared as described in [1] were characterized after chemical treatment with methanolic HCl in order to remove the template molecules and the extraframework Me [7]. MeAPO-5 was synthesized with a gel composition of 1.0 Al2O3:1.0 P2O5:0.25 MeO:1.0 Pr3N:40 H2O. The samples resulted from the incorporation of manganese, cobalt, zinc and magnesium in the synthesis of AlPO4-5 are labeled as MnAPO-5, CoAPO-5, ZAPO-5, and MAPO-5 respectively. AlPO4-5 was used as a reference material for all measurements and calculations discussed hereafter.

The MeAPO-5 samples were characterized by x-ray diffraction (XRD) analysis using Siemens D5000 Difractometer in the scanning range of 2t between 2 and 40o using copper Ka (q = 0.1542 nm at 35 kV and 35 mA) as the source of radiation. Silicon powder was used as an internal standard in the measurement of unit cell parameters in order to correct for any instrumental error which might occur. In all cases the peak at 2q  = 28.4o of silicon (hkl=111) was taken as a reference. Unit cell parameters were calculated from peak position 2 < 8o in order to minimize errors at low diffraction angle, using the least square refinement techniques. Chemical analysis data were collected by X-ray fluorescence (XRF) analysis using Siemens SRS 303. The MAS NMR experiments were performed using Varian UnityINOVA 400 MHz 9.4T spectrometer. The 31P MAS NMR spectra were recorded at 161.87 MHz using 2.0 ms radio frequency pulses, a recycle delay of 1.0 s and a spinning rate of 4.5 kHz. Chemical shifts for 31P was referred to 85% H3PO4. The samples were characterized after calcination at 400 oC overnight by MAS NMR and XRD.

Ammonia was used as the sorbed molecule in NH3-TPD experiment. TPD data was acquired from Micromeritics 2900 TPD/TPR. The sample was heated at 500oC for 40 minutes and cooled down to 100oC before passing ammonia. After saturating with NH3, the sample was flushed with N2 in order to remove the physically adsorbed NH3. Temperature for adsorption was increased at a rate of 5oC/min.

The dehydration and dehydrogenation of cyclohexanol over AlPO4-5 and MeAPO-5 and characterization of their products were carried out as described in [2].

3. Results and discussion

3.1. Physical properties

XRD patterns of AlPO4-5, MnAPO-5, CoAPO-5, ZAPO-5 and MAPO-5 show that the samples are pure and highly crystalline as reported [6]. The framework of MeAPO-5 by bulk chemical analysis as tabulated in Table 1 shows a strong evidence of isomorphous substitution of Me atoms by Al into the framework of MeAPO-5. Figure 1 shows the trend in the unit cell parameters of the a-axis and c-axis of the hexagonal lattice of AFI structure of MeAPO-5 samples respectively. It shows that an increase in parameter a and a decrease in parameter c with the incorporation of divalent metals into the framework. The decrease in parameter c was due to the influence of the size and amount of Me atoms that was incorporated into the framework of AlPO4-5 structure.

Table 1. Framework composition of MeAPO-5 and AlPO4-5 by bulk chemical analysis 


 


 


Fig. 1. The a-axis and c-axis unit cell parameters of AlPO4-5 and MeAPO-5.

Figure 2 shows the relationship between the mean T-O-P angle and ionic radii of metal atom in AlPO4-5 and MeAPO-5 as calculated from the 31P MAS NMR spectra. The T-O-P angle was calculated using the following relation [8]:
mean T-O-Pangle/degree=(47-a)/0.51 (1)

where a = chemical shift ppm. It is evident that manganese having the largest ionic radius (0.81A), imposes the largest effect to the increase of the T-O-P angle, followed by zinc (0.73A), cobalt (0.72A), magnesium (0.71A) and aluminum (0.53A) respectively. This suggests that the ionic size of the incorporated divalent metal influences the T-O-P angle in the structure of AlPO4-5.


Fig. 2. Relationship between mean T-O-P angle and ionic radii of metal atom in AlPO4-5 and MeAPO-5 (ppm from AlPO4-5).

One expects that the higher is the T-O-P angle, the bigger is the unit cell volume. Calculation of the unit cell volume of AlPO4-5 (1384A), MAPO-5 (1388A), CoAPO-5 (1387A), ZAPO-5 (1382A) and MnAPO-5 (1387A) show that the unit cell volumes do not follow the trend in the T-O-P because the unit cell volume is not solely affected by the T-O-P angle but by other factors such as the T-O and/or P-O bond lengths, the hydrogen bond and the cation coordination [9 and 10].

If a large divalent Pb atom (covalent radii = 1.54A and ionic radii = 1.20A [11]) is incorporated into the framework of AlPO4-5, the a-axis of the unit cell of Pb-AlPO4-5 is smaller than in MeAPO-5 and its c-axis is larger than in MeAPO-5. It is calculated from the AlPO4-5 phase that the lattice parameters of Pb-AlPO4-5 are, a = 13.704A and c = 8.509A. Attempt to incorporate a larger atom such as Pb (1.20A) atom into AlPO4-5 was not successful as pure product could not be obtained. Successful incorporation may be achieved only if the microporous framework is flexible. Referring to the ideal radius ratio for tetrahedral coordination of 0.225-0.414 as suggested by Flanigen et al. [12], the radius ratio for Pb2+ is obviously not ideal for tetrahedral coordination in AlPO4-5. This result reinforces our previous observation that the distortion of the structure of MeAPO-5 is caused by the size of the divalent metal atom.

The result obtained, indicates that the incorporation of the Me atoms distorts the structure of MeAPO-5 in the long-range. Short-range order of 31P in MeAPO-5 suggests that the distortion of MeAPO-5 crystals is correlated with the T-O-P angle.

3.2. Catalytic activity of MeAPO-5

It was observed that cyclohexene and cyclohexanone were the major products in the conversion of cyclohexanol over MeAPO-5. Cyclohexene was the product from the reaction of cyclohexanol in the presence of Bronsted acid sites and cyclohexanone from the reaction of cyclohexanol in the presence of the base sites [2, 13 and 14]. The role of the acid and base sites and the mechanism of conversion of cyclohexanol over MeAPO-5 has been discussed and reported [2 and 15]. The percentages of cyclohexene and cyclohexanone are given in Figure 3. The results show the overall trend of the acidity of MeAPO-5 in decreasing order as follows:
MnAPO-5 > ZAPO-5 >CoAPO-5 > MAPO-5
most acidic                             least acidic

Interestingly, this trend is in agreement with its T-O-P angle (see Figure 2. It is generally accepted that the larger is the T-O-P angle, the stronger is the acid [9 and 16]. Evidently, in the case of MnAPO the T-O-P angle contributes to the highest conversion of cyclohexanol to cyclohexene.


Fig. 3. The trend in percentage of cyclohexanone and cyclohexene after conversion of cyclohexanol over AlPO4-5 and MeAPO-5.

The only credible, though complex, relationship between activity and acidity of the MeAPO-5 was found when we considered the acid strength. MnAPO-5, being the strongest acid in the group studied, gave the highest conversion of cyclohexanol and is therefore considered as the most reactive catalyst. However, in our view, acid strength itself is not the only cause of the high activity; the number of acid sites also affects the activity.

Counting of acid sites seems to be more difficult. Prior studies have shown that there is a good correlation between the catalytic activity of a solid in certain reaction with the variation in the number of acid centers. The correlation between the amount of ammonia desorbed from the TPD curves with the number of acid sites from the overall composition of MeAPO-5 is shown in Table 2. It shows that the amount of ammonia desorbed among the MeAPO-5 samples is almost the same and is in agreement with the hypothetical concentration of H atom as Bronsted acid. It should be noted that the expected concentrations of the acid sites are estimated from the overall composition. The calculation is based on the assumption that Me atoms occupy the framework position.

Table 2. The hypothetical concentration of H atom as Bronsted acid in AlPO4-5 and MeAPO-5 molecular sieves and the conversion of cyclohexanol over these catalysts 

The results in Table 2 indicates that the activity of the MeAPO-5 increase in the following order: MnAPO-5>ZAPO-5>CoAPO-5>MAPO-5. Since the ionic radii of Me atom affects the acid strength, it is considered as a measure of activity of the MeAPO-5. The effect of the acid strength of catalyst is clearly observed in Figure 4. From this result, the synergy effect of acid strength and number of acid sites to activity is clearly demonstrated.

Fig. 4. Comparing the trend of percentage of cyclohexene with the amount of acid sites and activity of MeAPO-5.

The rank of the acidity obtained above contradicts the result observed by Pozas et al. [17]. They observed that the dehydration of ethanol increases over catalysts in the following order CoAPO-5>ZAPO-5>MnAPO-5. Calculation of lattice parameter by Pozas et al. shows that the incorporation of Mn, Co and Zn into AlPO4-5 implies a lattice expansion compared to the original AlPO4-5. However, they observed that the cell parameters of the hexagonal lattice of AFI structure type of CoAPO-5, ZAPO-5 and MnAPO-5 are similar. Our observations are different because the distortion of MeAPO-5 follow the trend of the ionic radii of Mn, Co and Zn. The Me concentration used by us in the synthesis of MeAPO-5 differs from those used by Pozas et al. Elemental ratio of Me reported by Pozas et al. was 0.010-0.014. Instead we increased the Me content threefold. Clearly these changes influence the distortion of the hexagonal lattice of the AFI structure.

A possible explanation for the difference between our results and that of Pozas et al. is that, the incorporation of a large amount of Me atom into the framework of AlPO4-5 not only cause distortion of the hexagonal lattice of the AFI structure, but also demonstrates the effect of ionic radii of Me atom on the structure. This effect was not clearly observed for a lower concentration of Me atom. The order of activity for MeAPO-5 with the lower concentration of Me atom was in agreement with the effective charge on oxygen atom calculated using Pauling's third rule as reported earlier [2].

4. Conclusions

The correlation between the ionic radii of divalent metals incorporated into the framework with the acidity of MeAPO-5 catalysts was observed in the conversion of cyclohexanol. The acidity of catalysts was correlated with the ionic size of Me atom in MeAPO-5 according to the following order: Mn>Zn>Co>Mg. The metal-substituted AlPO4-5 influences the acid strength and number of acid sites. The number of acid sites is the controlling factor in cyclohexanol conversion. Both structural (T-O-P angle) and chemical formula are factors that affect the acidity of MeAPO-5.

References

1. S.T. Wilson, E.M. Flanigen, European Patent Application, 0043562 (1982).
2. H. Nur, H. Hamdan, React. Kinet. Catal. Lett. 66 (1999) 33.
3. J. Sauer, Modelling of Structure and Reactivity in Zeolites (Ed.) C.R.A. Catlow, Academic Press, 1992.
4. J. Sauer, J. Molec. Catal. 54 (1989) 132.
5. J. Sauer, Chem. Rev. 89 (1989) 199.
6. H. Nur, H. Hamdan, Bul. Kim. 13 (1998) 31.
7. H. Nur, H. Hamdan, in Proceedings of Regional Symposium on Chemical Engineering, Malaysia, 1997, p. 613.
8. D.E. Aporiaye, A. Andersen, I.M. Dahl, H.B. Mostad, R. Wendelbo, J. Phys. Chem. 99 (1995) 14142.
9. W.J. Mortier, in Proceedings of the Sixth International Zeolite Conference (Eds.) D. Olson, A. Bisio, Butterworths, 1984, p. 734.
10. W.F. Bleam, P.E. Pfeffer, J.S. Frye, Phys. Chem. Mineral, 16 (1989) 455.
11. J.A. Dean (Ed.), Lange's Handbook of Chemistry, 13th edition, McGraw-Hill Book Co. (1985).
12. E.M. Flanigen, R.L. Patton, S.T. Wilson, Stud. Surf. Sci. Catal. 37 (1988) 13.
13. S.P. Elangovan, V. Murugesan, J. Molec. Catal. 118 (1997) 301.
14. K.V.V.S.B.S.R. Murthy, B. Srinivas, S.J. Kulkarni, V.K. Kaushik, M. Subrahmanyam, P.K. Rao, Indian J. Chem. 35A (1996) 649.
15. S. Endud, S., H. Nur, H. Hamdan, Stud. Surf. Sci. Catal. 117 (1998) 453.
16. J.A. Rabo, G.J. Gajda, Catal. Rev. Sci. Eng. 31 (1989) 385.
17. C. de las Pozas, R. Lopez-Coerdero, J.A. Gonzales-Morales, N. Travieso, R. Roque-Malherbe, J. Molec. Catal. 83 (1993) 145.

Corresponding author. Present address: Catalysis Research Center, Hokkaido University, North 11, West 10, Sapporo 060-0811, Japan. Fax: +81-11-7094748; email: hadi@cat.hokudai.ac.jp

 
 
Materials Research Bulletin
Volume 36, Issues 1-2
January 2001 
Pages 315-322 
 
Send feedback to ScienceDirect
Software and compilation (C) 2001 ScienceDirect. All rights reserved.
ScienceDirect(R) is an Elsevier Science B.V. registered trademark.