Thermal stability of hydroxy groups in dealuminated mordenites

Thermal stability of hydroxy groups in dealuminated mordenites

Thermal stability of hydroxy groups in dealuminated mordenites P. Fejes, I. H a n n u s and I. Kiricsi Applied Chemistry Department,Jdzsef Attila Uni...

303KB Sizes 2 Downloads 66 Views

Thermal stability of hydroxy groups in dealuminated mordenites P. Fejes, I. H a n n u s and I. Kiricsi

Applied Chemistry Department,Jdzsef Attila University, Szeged, Hungary and H. Pfeifer, D. Freude and W. O e h m e

Sektion Physik der Karl-Marx-Universitiit Leipzig Leipzig German Democratic Republic (Received 6June 1984) Results obtained by proton magnetic resonance and derivatographic measurements permitted the following conclusions regarding the thermal stability of OH groups in the framework of dealuminated mordenites: (1) after vacuum pretreatment at room temperature, the adsorbed water molecules are removed. OH groups exist either in completely filled OH nests at the sites of aluminium vacancies, or as structural and terminal OH groups; (2) the thermal stability of the OH nests is low, for after pretreatment at about 500 K, such nests could be detected neither with n.m.r., nor with derivatographic methods. Keywords: Derivatography, nuclear magnetic resonance; dealuminated mordenites; OH groups; OH nests

INTRODUCTION A l u m i n i u m ions removed from the zeolite framework leave behind a zeolite with a defect structure. It is assumed that an a l u m i n i u m vacancy can be stabilized by the formation of hydroxylated nests or new ----Si--O - - S i ~- linkages, depending on the dealumination procedure applied I. From the results of i.r. studies the occurrence of both = S i - - O H and - - = S i - - O - - S i ~bonds at former sites of a l u m i n i u m in the framework can be substantiated 2-s. It is generally accepted that dealumination is accompanied by an increase of acidity 9'1°, though a detailed explanation of this is at present unknown. T h e proton magnetic resonance method allows determination of the absolute n u m b e r of O H groups and may yield valuable information concerning their local arrangement n- 14. For study of the thermal stability of zeolitic O H groups, thermal analysis is the most widely used method; by this means the dehydroxylation process can be followed quantitatively '5. With a combination of the two methods, i.e.n.m.r. spectroscopy and derivatography, more detailed information can be obtained about the properties of O H groups in zeolitic structures. This paper deals with experimental results of both proton magnetic resonance and derivatographic m e a s u r e m e n t s concerning the thermal stability of O H groups in mordenites dealuminated with phosgene.

A Derivatograph-Q. instrument was used for the derivatographic measurements, which were carried out in the temperature range 298-1273 K. Details of the proton magnetic resonance m e a s u r e m e n t s can be found in ref. 16.

RESULTS

Figure 1 shows the t.g. and d.t.g, curves of untreated N a - m o r d e n i t e (no. 1), H - m o r d e n i t e (no. 2) and dealuminated mordenite (no. 3). For sample no. 2, two weight loss processes can be observed. T h e first process reaches its m a x i m u m at about 370 K, whereas the second occurs at higher temperatures. Dehydroxylation o f d e a l u m i n a t e d mordenite starts at lower temperature, as can be seen by comparing the t.g. and d.t.g, curves of Figure lb and c. In the latter case there is no weight loss at 800 K, in contrast with Hmordenites. This behaviour is characteristic for dealuminated mordenites, as observed first by Beyer 17. Proton magnetic resonance spectra of dealuminated mordenite (no. 4), measured at a t e m p e r a t u r e of 120 K using a resonance frequency of 90 M H z , are shown in Figure 2. T h e peak doublet which gives rise to the Table 1 Aluminium contents of the zeolite samples used in the experiments

No.

Sample

EXPERIMENTAL Dealuminated synthetic mordenites were used in the experiments. T h e starting material were N a - m o r d e n i t e and H - m o r d e n i t e originating from the Norton Co. After dealumination with phosgene at selected temperatures 7'8, the samples were washed free of C1- ions, dried and analysed. T h e a l u m i n i u m contents of the samples are listed in Table1. 0 1 4 4 - 2 4 4 9 / 8 5 / 0 1 0 4 5 - 0 4 $03.00 © 1985 Butterworth & Co. (Publishers) Ltd.

1 2 3 4 5 6 7

Aluminium content (mmol g -1 )

Na-mordenite H-mordenite Dealuminted H-mordenite Dealuminated H-mordenite Dealuminted Na-mordenite Dealuminated Na-mordenite Dealuminated Na-mordenite

ZEOLITES, 1985, Vol 5, January

2.80 2.80 2.25 2.00 2.70 2.30 1.50

45

OH groups in dealuminated mordenite: P. Fejes et al.

10

200

Table 2 Proton concentration in the dealuminated H-mordenite specimen no. 4 measured by n.m.r.. Except the specimen denoted by an asterisk (293*) all other samples were evacuated for 24 h at the pretreatment temperature shown in the first column. For comparison the results of t.g. measurements (performed under normal pressure) are collected in the last column

°o

II



o

." \

Pretreatment temperature

lOO

-05

.

a

~,,

J

200 -

.

.

.

./'~.

.

293* 293 323 373 428 473 523

.

-1 0

lOO

0.5

E L~ <3

E v 01

b

I

I

I

I

i i

i i i i I

2OO

"

100

0

i I

!

).5

-

ci

473

673 T(K)

873

Figure 1 T.g. ( ) and d.t.g. ( - - . - - ) curves of (a) Na-mordenite, (b) H-mordenite and (c) dealuminated mordenite

46

ZEOLITES, 1985, Vol 5, January

Proton per AI vacancy

(mmol g - l )

"\\..,.,.

,

Proton conc.

16.7 6.3+0.3 5.5±0.3 4.2+0.3 3.7+0.3 2.7±0.3 2.7±0.3

4.3±0.3 3.9±0.3 2.3±0.3 1.6±0.3 0.4±0.3 0.4±0.3

Water content (retool g - l ) n.m.r,

t.g.

8.3 3.2 2.7 2.1 1.8 1.3 1.3

9.2 8.8 6.2 3.6 2.6 1.9

shoulders (arrows) in curve 'a' is due to adsorbed water molecules firmly bound to the framework at 120 K. Its intensity decreases strongly after evacuation ( = 10 -2 Pa) at room temperature (curve 'b'), and it disappears completely at 323 K (curve 'c'). This behaviour complements the finding of Bosa~ek et al. l~ in that dealuminated Y-zeolites become completely dehydrated following evacuation for 12 h at 333 K. Table 2 contains values of the concentrations of protons in dealuminated mordenite (no. 4) following treatments at increasing temperatures. Pretreatment involved evacuation for 24 h at temperatures between 293 and 523 K (first column), except the sample marked with an asterisk. T h e values in the second column were derived from the second moments (M2) of the n.m.r, signals. Table 3 reflects the effect of dealumination on the proton concentration in the case of Na-mordenites dealuminated to various extents: the second column gives the second moments of the n.m.r, signals (M2), and the last one the computed proton concentrations. Measurements o f m 2 at room temperature and the temperature of liquid nitrogen (77 K) resulted in the same values. An additional experiment with Hahn's echo =6 showed that in sample no. 1, half of the protons can be attributed to bridging O H groups, like the structural O H groups in decationated zeolites, characterized by a strong proton-aluminium dipole-dipole interaction. The other half of the protons in sample no. 1 ( - 0 . 4 mmol g-t) can be assigned to terminal O H groups and to O H groups at lattice defects. T h e values M 7 = 0 . 8 • 10-ST e measured for sample nos. 5 and 6 are comparable with the value .442=1.0' 10-~T 2 measured for structural O H groups in decationated zeolites. If we subtract the concentration of nonstructural O H groups (0.4 mmol g-i) in the Namordenite sample no. 1 from the proton concentrations given in Table 3, and divide the difference by the aluminmm content given in Table 1, we obtain the n u m b e r of protons per lattice aluminium. T h e resulting values for sample nos. 5 and 6 are 0.92 and 0.74, respectively, and support the assumption that these protons are in structural O H groups, for which the value should be one. Only sample no. 7, the most strongly dealuminated Namordenite, contains significantly more than one proton per aluminium. Hence, there should be non-negligible n u m b e r of protons due to lattice defects in this sample.

OH groups in dealuminated mordenite: P. Fejes et al. Table 3 Values for the second moment M2 of the n.m.r, signals and proton concentration in Na-mordenites dealuminated to various extents, All samples were evacuated (--- 10 -2 Pa) at 673 K for 24 h

Sample

1 5 6 7

M2 10 -8T2

Proton conc. (mmol g -1 )

0.42±0.05 0.82±0.05 0.74±0.05 0.65±0.05

0.78±0.08 2.9±0.3 2.1 + 0 . 2 2.5±0.2

This view is supported by the silanation experiments of Barrer et al. as well 2°. It was found that partially dealuminated, wetted and outgassed mordenites never reacted with Sill4, as might have been expected on the basis of the nests having 4 O H groups III Si

III

I

Si I

OH

0

------Si--OH

I

HO--Si~

+SiHc'*~Si--O--Si--O--Si--

+4H 2

I

T h e proton concentration in sample no. 4 is given in Table 2 as a function of the pretreatment temperature. Sample no.4 was derived from the original Namordenite (sample no. 1) by decationization and dealumination. In order to find the proton concentration following dealumination, 0.4 m m o l g-t must first be subtracted from the concentrations given in column 2 for the non-structural O H groups, followed by the concentration corresponding to one proton per aluminium, which takes account of the structural O H groups caused by decationization (2.0 m m o l g - i). T h e result divided by the concentration of a l u m i n i u m vacancies in this sample (0.8 m m o l g - l , cf. Table 1) is shown in column 3, Table 2. We find about 4 O H groups per vacancy in the sample evacuated at room temperature, half of this value after vacuum pretreatment at 373 K and an insignificant amount of O H groups in O H nests alter vacuum pretreatment at 473 K. DISCUSSION In the t.g. curve of the dealuminated mordenite (see Figure lc) the weight loss processes relating to dehydration and dehydroxylation are not separated. It is therefore difficult to find out how m a n y protons form an O H nest near to the a l u m i n i u m ion removed. It is even more difficult to judge whether these O H nests exist at all in the zeolite after heat treatment at temperatures above 700 K, i.e. those generally used for the activation of zeolite catalysts. In spite of various speculations concerning the origin, structures and features of the O H nests presumed to be formed in the dealumination process, structures exhibiting 4 O H groups/nest have never been observed in proton magnetic resonance experiments, although this is the best method of determining the absolute n u m b e r and a r r a n g e m e n t of O H groups in zeolites 14. If the O H nests are built up from 4 O H groups formed in the dealumination process, they should be of low thermal stability and should decompose by dehydroxylation at relatively low temperatures. According to Beyer, the O H nests dehydroxylate at a t e m p e r a t u r e lower than 473 K 17. Water and new = S i - - O - - S i ~ bonds are formed as a result of this process. T h o u g h under quite different conditions (at 873 and 1473 K), Moulson and Roberts 19, studying the penetration of water in silica glass, found that the equilibrium H 2 0 + =Si--O--Si----- ~2=---SiOH lies well to the left, which again shows the limited stability of hydrated nests.

OH

O

I

I

Si

Si

III

III

T h e low ratio of 2.28 O H groups/nest found by monitoring the H2 evolved strongly suggests that the hydrated nests have already reacted to give at least one new - - ~ S i - - O - - S i ~ bond/nest and one molecule of water during the outgassing, which at 630K is in agreement with our n.m.r, data. As can be seen in Figure 3, the d.t.g, curve of the dealuminated mordenite sample no. 4 runs parallel to the t e m p e r a t u r e axis between about 570 and 1270 K, indicating a constant weight loss in this t e m p e r a t u r e range. This also follows from the t.g. curve, which decreases linearly between these temperatures. T h e t e m p e r a t u r e relating to the m a x i m u m of the d.t.g, curve of the dealuminated mordenite ( ~ 3 7 0 K) is smaller than that for N a - m o r d e n i t e (400 K) or H - m o r d e n i t e (390 K). From this result we conclude that the water is more loosely bound in dealuminated mordenite structures. From the proton concentration data determined by n.m.r, m e a s u r e m e n t s it was possible to determine the water content, including the O H groups (2 O H groups are equivalent to one water molecule), for sample no. 4. T h e data are shown in Figure 3. T h e agreement between

Q

b C

Figure 2 Proton magnetic resonance signals of a dealuminated mordenite (no. 4) measured at 120 K: (a) before evacuation at room temperature; (b) after an evacuation (= 1 O-2 Pa) for 24 h at room temperature; (c) after an evacuation (--- 1 0 - 2 Pa) for 24 h at 323 K

ZEOLITES, 1985, Vol 5, January

47

OH groups in dealuminated mordenite: P. Fejes et al.

loss of adsorbed water and dehydration of O H nests up to a temperature of about 500 K. At about 500 K and above, water is released by the loss of structural and terminal O H groups. In other words, this means that no O H nests remain in the framework after heat treatment at 500 K.

1oo- j

_~ ,I

50

2oo o E

- p

lO0

"o

I

I

I 473

t 673

L ~'"t"--~ I 873 1073 127: T(K) Figure 3 C o m p a r i s o n of t.g. ( ), d.t.g. (---) and p r o t o n m a g n e t i c resonance (e, ; ) data for a d e a l u m i n a t e d m o r d e n i t e (no.

4)

the results of the derivatographic and n.m.r, investigations is quite good for the non-evacuated sample (filled circle in Figure 3). From the line shapes of the n.m.r, signals in Figure 2 it follows that after evacuation ( ~ 10-" Pa) for 24 h at room temperature the zeolite sample contains only negligible amounts of molecular water. It is evident from Table2 that the water contents of the evacuated samples, determined by n.m.r. (open circles in Figure 3), are lower than the water contents of the non-evacuated samples (filled circle and t.g. curve in

Figure 3). From the n.m.r, intensities due to aluminium vacancies (column 3 in Table 2) it must be concluded that each aluminium vacancy is filled with about 4 O H groups/OH nest at room temperature. Heat treatment above 373 K causes their destruction, and no O H nests exist above 473 K. Hence, on the basis of the n.m.r. measurements, the t.g. curve can be explained by the

48

ZEOLITE& 1985, Vol 5, January

REFERENCES 1 8arrer, R. M. 'Zeolites and Clay Minerals as Sorbents and Molecular Sieves', Academic Press, London, 1978, Ch. 7 2 Ha, B. H., Guidot, J. and Barthomeuf, D. J. Chem. Soc. Faraday Trans. 1, 1979, 75, 1245. 2366 3 Mirodatos, C., Ha, B. H., Otsuka, K. and Barthomeuf, D. "Proc. 5th Intern. Conf. Zeolites, Naples; Heyden, 1980, 13-382 4 Chen, N. Y. J. Phys. Chem. 1976, 80, 60 5 Ghos, A. K. and Curthoys, G. ,I. Chem. Soc. Faraday Trans. 1, 1983, 79, 805 6 Kerge, H. G. Z. Phys. Chem. N.F. 1980, 122, 103 7 Fejes, P., Kiricsi, I. and Hannus, I. Acta Phys. Chem. Szeged 1982, 28, 121 8 Fejes, P., Hannus, I. and Kiricsi, I. Zeolites 1984, 4, 73 9 Wendlandt, K. P., Weigel, W., Hofmann, F. and Bremer, H. Z. anorg, allg. Chem. 1978, 445, 51, 59 10 Mishin, I. V., Rubinstein, A. M. and Wendlandt, K:P. Z. anorg, allg. Chem. 1980, 467, 17 11 Freude, D., Hunger, M. and Pfeifer, H. Chem. Phys. Le~. 1982, 9 1 , 3 0 7 12 Freude, D. and Behrens, H.*J. Cryst. Research Techn. 1981, 16, 36 13 Freude, D., Pfeifer, H., Ploss, W. and Staudte, 8. J. Mol. Catal. 1981, 12, 1 14 Freude, D. and Pfeifer, H. 'Proc. 5th Intern. Conf. Zeolites, Naples', Heyden, 1980. p. 732 15 Beyer, H. K., Mihalyfi, J., Kiss, A. and Jacobs, P. A. J. Thermal Anal. 1981, 20, 351 16 Freude, D., Oehme, W., Schmiedel, H. and Staudte, 8. J. Catal. 1977, 49, 123 17 Beyer, H. K., Belenkaja, I. M., Dubinin, M. M. and Hange, F. Izv. Akad. Nauk, Set. Khim. 1982, 1457 18 Bosaeek, V., Patzelova, V., Tvaruzkova, Z., Freude, D., Lohse, U., Schirmer, W., Stach, H. end Thamm, H. J. Cata/. 1980, 6 1 , 4 3 5 19 Moulson, A. J. and Roberts, J. P. Trans. Faraday Soc. 1961, 57, 1208 20 Barrer, R. M. 'Zeolites and Clay Minerals" Academic Press, London, 1978, p. 358