On the synthesis of erionite—offretite intergrowth zeolites

On the synthesis of erionite—offretite intergrowth zeolites

On the synthesis of erionite-offretite intergrowth zeolites Karl P. Lillerud and Johan H. Raeder* University of Oslo, Department of Chemistry, P O Box...

22MB Sizes 0 Downloads 10 Views

Recommend Documents

No documents
On the synthesis of erionite-offretite intergrowth zeolites Karl P. Lillerud and Johan H. Raeder* University of Oslo, Department of Chemistry, P O Box 1033, B lindern, N-0315 Oslo 3, Norway *Present address: Det norske Verita.L PO Box 300, N-1322, HCvik, No)way (Received I October 1985)

A partial exploration has been made of factors affecting crystallization of offretite-erionite type zeolites from aluminosilicate gels. The effect of the type of template-ion, the Si/AI ratio and the Na/K ratios in the gel, on offretite crystallization and on erionite inter-growth formation in offretite has been studied. The ionic contribution from the aluminium cation interaction is important for the stability of the structure. Therefore, offretite, which can adapt to more cations and crystallize with a higher Si/AI ratio than erionite, forms more easily. The zeolites have been characterized using X-ray diffraction, scanning electron microscopy, EDX-analysis and chemical analysis. Keywords: Synthesis;offretite; erionite; cation sites; XRD; SEM; intergrowth

INTRODUCTION

EXPERIMENTAL

The offretite-erionite series of zeolites are selective catalyst candidates for industrial processes such as C2-C,1-01efin p r o d u c t i o n f r o m m e t h a n o l or dimethylether 1'2, as well as for selective cracking of n-paraffins in a hydrocarbon mixture 3. Offretite and erionite are closely related zeolites of the chabazi.te group 4. Their structure and structural relationship have been extensively described elsewhereS.~ It has been shown by transmission electron microscopy (TEM) that offretite crystals completely free from erionite intergrowth are very rare TM. One single erionite layer in an offretite crystal blocks the 12-ring channels along the c-axis of the crystal. The amount and distribution of erionite in the crystals will, therefore, affect and alter the catalyst selectivity and the diffusion properties of the zeolite. It has been demonstrated that selectivity and product distribution are altered by the a m o u n t of erionite intergrowth 9. Also, the lifetime as a catalyst may be affected by this intergrowth ~°. However, little has been done to make clear which factors control the formation of these intergrowths during the synthesis of zeolites. The aim of this work is to gain a better understanding of these factors and thereby to be able to synthesize zeolites with a controlled amount and distribution of intergrowths.

Synthesis of zeolites The chemicals used for the zeolite preparations are listed in Table 1. The parent gels were prepared by dissolving sodium aluminate, sodium and potassium hydroxide and template-ion (tetramethylammonium hydroxide (TMA) or cholinchloride) in the appropriate quantity of water. This solution was then mixed with colloidal silica to give the resultant gel. The gel was homogenized with a turbo mixer working at 10,000 rev. rain -l. This final mixing was performed either at room temperature or at 0°C. The gels were then: filled in polyethylene bottles and heated at 80°C for seven days in sealed bottles; or, filled in Teflon-lined autoclaves and heated for 4 h at 190°C. The samples were then decanted, washed with hot water and finally dried at 120°C.

Zeolite characterization In this work X-ray diffraction is used for identification of crystalline phases formed as well as for quantitative determination of the ratio between the components formed and the remaining amorphous by-products. It is also used for determination of the fraction of erionite intergrowths in the offretite. A Philips powder diffractometer with a PW 1710 and a PW 1700 computer facility and a Guinier-Hegg type camera have been used in the diffraction studies.

Table 1 Chemicals Element/Compound

Quality

Manufacturer

Si-source

'Ludox' colloidal silica type: SM and TM NaAIO2 (Technical) (purity 98%) (purity 87.9%)

E.I. du Pont de Nemours & Co. Wilmington, Delaware 19898, USA Kebo AB, Stockholm, Sweden EKA Kemi, Stockholm, Sweden EKA Kemi, Stockholm, Sweden

(P.A.,>98%)

FLUKA AG CH-9470 Bucks, Switzerland

(P.A.,>97%)

FLUKA AG CH-9470 Bucks, Switzerland

AI-source NaOH KOH Tetramethylammonium hydroxide pentahydrate (TMA) (CH3)4N(gH)5H20 Cholinchloride (CHa)N(CI)CH2CH2OH 0144-2449/861060474-10 $03.00 (~) 1986 Butterworth & Co. (Publishers) Ltd 474

ZEOLITES, 1986, Vol 6, November

Synthesis of erionite-offretite intergrowth: K.P. Lillerud and J.H. Raeder Table 2

Gel compositions in molar ratios Si/AI series

Na/K series

Min-max

Min-max

ZSM-34 type exl ~

T-type

II°

exl"

T/1-2 ~

Jenkins'7

AIzO3

0.3-4.3

1

1

1

1

1

1

SiO2 Na20 KzO Template H20

20-27.8 8.0 1.7 1.3 400

15 0.6-4 0-4 4 400

27 4.8 0.8 3.4 414

15 2.4 1.6 4 400

20 6.3 2.0 ~ 252

28 5.8 2.9 -470

17.5 6.7 2.3 0.08 276

aThis nomenclature refers to Tab/es 4 and 5

When the film technique was employed, quantitative line intensities were obtained by a photometer on-line with a micro-computer. This was used for correction of the intensities in an iterative manner to gain quantitative information about crystalline phases present. The particle size and shape of" the solid products were determined by scanning electron microscopy (JEOL JSM-35). The chemical composition of individual particles was determined by energy dispersive X-ray analysis (PGT-1000) in the scanning microscope. ZAF corrections were performed to gain quantitative resuhs. These routine analyses were performed on zeolite powders. This is an improper geometry for a true ZAF correction, but the Si/AI ratio is not significantly affected. The aluminium and silicon K-lines are so close in energy that small errors in the corrections will be identical for the two elements and will not affect the Si/AI ratio, but the corrections made for the sodium K-line may be uncertain. Some of the largest crystals were also analysed as polished cross-sections. The average chemical compositions were analysed by ICP and atomic absorption techniques on hydrofluoric acid solutions of the zeolites.

the Na/K ratio in the gels and in the corresponding zeolites. For Na/K ratios > 0.1, the ratio in the zeolite is always lower than in the corresponding gel. At lower ratios zeolites were formed with Na/K ratio equal to that in the gels. T h e crystal shapes are as r e p o r t e d in the literature 11,12. Figure 3 shows micrographs of typical crystals resulting from gels with varying Na/K ratio. The crystals are elongated in the c-axis direction, the length being between 0.5 and 2 Ixm. The length/ thickness ratio is typically 3/1. The only exception is the specimen crystallized from the gel with Na/K=3, 100

I

!

I

/J

!

--80i 60

u 40-

RESULTS I n f l u e n c e o f Na/K ratio in gel on c r y s t a l l i z a t i o n Aiello and Barrer II have investigated zeolite crystallization in the presence of mixed bases (NaOH, KOH and TMA-OH). Among other zeolites, they synthesized erionite and offretite ;fronl gels that differed only in Na/K ratio. In this work two series of zeolites are synthesized: one at 80°C from gels with continually changing Na/K ratio: the other at 190°C from gels with continually changing Si/Al ratio. The composition of gels is listed in Table 2. T h e synthesis procedure is described above and follows mainly the method described by Aiello and Barrer II. All gels containing KOH produced pure offretite, free from erionite intergrowths and other zeolites (i.e. amounts not detectable with X-ray diffraction). From gels completely free from KOH no crystalline zeolite was formed. (Aiello and Barrer I~ reported the formation of erionite from these gels.) The X-ray diffractogram contains broad lines that may be interpreted as poorly crystalline erionite. All gels containing KOH produced offretite but the crystallinity changed significantly as a function of the Na/K ratio, (illustrated in Figure 1). The yield in a synthesis follows the crystallinity. Figure 2 compares

20

0

~ t

I 1

I 2

I 3

// [/

-

I 9

Na/K (in gel) Figure 1 Crystallinity of offretite as a function of Na/K ratio in the parent gel. Crystallization temperature 80°C

3

3

Na/K (in zeolite) 5

5 Na/K (in gel)

7

9

7

9

m

Figure 2 Comparison of the Na/K ratio in the '~]els and in the resulting solid products

ZEOLITES, 1986, Vol 6, N o v e m b e r

475

Synthesis of erionite-offretite intergrowth: K.P. Lillerud and J.H. Raeder

therefore, synthesized with gradually increasing Si/AI ratio in the gel, the other conditions being kept as constant as possible. T h e gel compositions are given in Table 2. T h e m e t h o d is described above and follows mainly that of Inui el al. i.t. Tile a m o u n t of offretite and other zeolitic species changes smoothly with the Si/AI ratio in the gel. Figure 4 shows tile a n m u n t of offretite, phillipsite and sodalite tornled as a f t , nction of Si/AI ratio in tile gel. T h e r e is a sharp nmximunl in Ihe a m o u n t of offretite at Si/AI= 7. In the gel with the lowest Si content tile main crystalline product is cubic zeolite P (zeolite p,)m. To facilitate comparison with other X-ray instruments, tile intensities of tile strongest X-ray reflections from tile most crystalline offretite s,ltnl)le, relative to o¢-AI,_,O:+, is given in Tabh, 3. Fig'ure 5 compares tile Si/AI ratio in the gels and in the corresponding zeolites. Pure offretite is [ol'nled only in tile limited Si/AI range between 2.7 ,rod 3.0. "l'he n u m b e r of cations per unit cell in each sample has been calculated fronl chemical analysis on solulions. Widfin the Si/A1 range from 3.1 to 2.3 the nund)ers of K+-ions and TMA+-ious are constant (1.5 K+-ions and 1 'f'MA+-ion per unit cell). Hmvever, the nt, mber of Na+-ions varies consideral)lv ( 1.4 Na+-ions at Si/AI=3.1 increasing to 2.5 at Si/AI=2.3) and this accounts for tile difference in charge t o n i pells;.ltion. At Si/AI ratios > 3.1 the situation is less clear with both the numhers of Na+-ions and K+-ions ch,inging. T h e lowest observed K+-ion COlacentration is one ion per unit cell. T h e morphology of the crystals formed changes dramatically with changing Si/AI ratio in the gels. Fiune 6 shows representative micrographs of the crystals fiwmed through this series, the most ci'vstal -` line zeolite particles being fornled froin the gel with Si/AI=9.5, Fiffure 6e. These crystals glmv with facets and the hexagonal appearance is clearly visible. T h e largest crystals are 3 0 × 6 I.Un. When tile Si/AI ratio is decreased relative to this optinnun ratio, Figure 6a-d, the crystal shapes become more elongated with increasing aluminium conlenl. ()ffretite synthesized with tile highest possible alunliniunl content (Figure 6a) ['orins fibrous crystals like natural erionite. When the Si/AI ratio is increased to > {).5, tile optimunl ratio, tile particles are still crystalline but the shapes are less regular and the particle size decreases.

Figure 3 Micrographs of zeolites resulting from gels with varying Na/K ratio. (a) Gel composition Na/K=3.0; (b) Gel composition Na/K= 1.8; (c) Gel composition Na/K= 1.0

which formed honlogeneous spheres 0.3 Iml in diameter.

I n f l u e n c e o f gel Si/AI ratio o n crystallization Naturally occurring erionite has a higher Si/AI ratio than natural offretite ~:. A series of zeolites was,

476

ZEOLITE& 1986, Vol 6, November

I n f l u e n c e o f t e m p l a t e - i o n o n crystallization In the literature offretite-erionite type zeolites ;ire synthesized with two different org,mic template-ions, T M A and cholinchloride, and also with only the inorganic ions, Na + and K +, as tenlplates (T-type zeolites). In this work the structure directing role of the template ion is demonstrated by crystallization from gels where T M A is substituted with cholinchloride or Na + and K +. Z S M - 3 4 type zeolites In the synthesis of these zeolites cholinchloride is employed as template ion. Only two zeolites are synthesized, as indicated in Table 4. One is synthesized according to Example i in the

Synthesis of erionite-offretite intergrowth: K.P. Lillerud and J.H. Raeder I

I

I

I

IO0

1

10

1

10

Si/AI (in zeolite) 20

30

80

'~ 60

20 Si/AI (in gel)

30

Figure 5 C o m p a r i s o n of Si/AI ratio in gels and the corresponding zeolites, from the synthesis series w h e r e the Si/AI ratio is changed systematically

20 0

b

I

i

l

I

l

I

I

I

60-

.E = 40

Qr

20 0

C

,T~""~I~I

IT

I

I

I

I

I

A

60

= 40

0

i 0

10

I 20

30

40

50

Si/AI (in gel) Figure 4 Crystallinity as a function of Si/AI ratio in the parent gels. (a) Ratio of offretite relative to the most crystalline species; (b) ratio of phillipsite relative to a mineral standard; (c) ratio of sodalite relative to the sodalite content in the species where sodalite is the main c o m p o n e n t . In this species the sodalite content is set to 50%

"C r~

!.

,

,,.--t~.'

original patent ~'l~ a n d the o t h e r with a gel composition c o r r e s p o n d i n g to the most crystalline offretite o b t a i n e d with T M A as a template ion and 8 0 ° C as the synthesis t e m p e r a t u r e . O f these two zeolites, the latter has the most d e v e l o p e d crystals, but both are 50% crystalline. T h e particle a p p e a r a n c e s are shown in Figure 7. Synthesis o f zeolites with cholinchloride at I90°C was not successful. T-type zeolites T h e T - t y p e zeolites are synthesized without any organic t e m p l a t e - i o n a c c o r d i n g to the description in E x a m p l e 1 in the p a t e n t Ref. 16. T h e gelcompositions are given in Tab& 2 a n d the composi-

Figure 6 Micrographs of representative zeolites f o r m e d in a synthesis series w h e r e the Si/AI ratio is changed systematically. (a) Si/AI in gel = 2.3, in zeolite = 2.0; (b) Si/AI in gel = 3.4, in zeolite = 2.3; (c) Si/AI in gel = 5.5, in zeolite = 3.2; (d) Si/AI in gel = 7.0, in zeolite = 3.0; (e) Si/AI in gel = 9.5, in zeolite = 3.2; (f) Si/AI in gel = 11.2, in zeolite = 3.2; (g) Si/AI in gel = 13.5, in zeolite = 3.5; (h) Si/AI in gel = 32.0, in zeolite = 4.1

Table 3 Absolute X-ray intensities of the most crystalline TMA-offretite Line (d-values in A)

IIIo"

11.5 (2.98)

7.58 0.61

6.65 0.94

6.63 0.31

3.77 1.92

3.60 1.48

alo is the intensity of the d=3.48 line of c~-AI203

ZEOLITES, 1986, Vol 6, November

477

Synthesis of erionite-offretite intergrowth: K.Po Lillerud and J.H. Raeder Table 4

ZSM-34 type zeolites

Sample

Si/AI (in gel)

Si/AI (in zeolite)

Na/K (in gel)

Na/K (in zeolite)

Crystallized offretite (%)

Erionite intergrowth (%)

ZSM-34 exl ZSM-34 II

13.3 7.5

3.7 4.0

5.0 3.0

0.8 0.4

45 50

>5 >5

tions of tile resuhing zeolites are given in Table 5. The samples termed T/I and T/2 are grown at 190°(;. Typical crystal shapes are shown in Figure 8. The low-temperature synthesis (80°C) gave particles that were agglomerates of very small plate shaped crystals. Jenkins j7 reported a method for synthesis of TMA-offi'etite with only 1% of the TMA content normally used (Tabh, 2). Zeolites synthesized according to tiffs patent are very similar to the T-type zeolites with respect to particle shapes (Figure 9) as well as to the anaount of erionite intergrowth. Therefore, variations in TMA content in the gel within this range are not likely to affect the offretite crystallization path.

X-ray diffraction

Figure 7 Micrographs of the ZSM-34 zeolites. (a) ZSM-34ex1 synthesized according to the Refs. 1 and 15, patent Example 1; (b) ZSM-34 II, gel composition corresponding to the most crystalline offretite but with cholinchloride instead of TMA as the template-ion; (c) detail of B

478

ZEOLITES, 1986, Vol 6, November

Pure offretite and erionite are easily distinguished by X-ray diffraction. Figure 10 shows diffractio,a patterns of synthetic offi'etite, natural erionite and natural phillipsite. The phillipsite pattern is included because phillipsite frequently forms in mixtures of offretite and erionite. Our erionite standard samples, one from Pine Valley, Nevada, USA and one fi'om Shoshone, California, USA contain small amounts of phillipsite (3% and 1% phillipsite, respectively). They are both commercially available*. In the erionite diffraction pattern, Figure 10, phillipsite lines are indicated as filled lines. Unfortunately the (201) and (211) odd / lines used to determine the erionite content in offretite ~, overlap with strong phillipsite reflexions. Small anaounts of phillipsite, which will be easily formed in an offi'etite synthesis, will therefore aher the ratio between the odd l lines. The use of these lines in quantitative evaluations of erionite content must, therefore, be subject to reservations. The only line where overlap is no severe problem, is the (101) peak at 9.3 /~. However, this peak is normally very broad, due to small erionite domains in the offretite crystals. The integration and comparison of these peaks are therefore doubtful. Chen et al. :~ compare simulated patterns of ordered and random 50% stacking-fauhed structures, and show that the relative intensities of the odd I lines are dependent on the amount of order in the erionite distribution. Due to these limitations the results obtained by X-ray diffraction can only be used as a rough indication of the erionite content in an offretite sample. Figure 11 shows diffraction patterns of offretiteerionite intergrowth structures. A realistic detection limit for erionite in offretite by means of X-ray diffraction is ~ 10% by weight. *Mineral Research, Clarkson, New York, USA

Synthesis of erionite-offretite intergrovvth: K.P. Lillerud and J.H. Raeder Table 5

T-type zeolites

Sample T/1 T/2 T-exl Jenkins 17

Si/AI (in gel)

Si/AI (in zeolite)

Na/K (in gel)

Na/K (in zeolite)

Crystallized offretite (%)

Erionite intergrowth (%)

7.5 7.5 13.3 9.0

3.0 3.0 3.2 2.3

3.0 3.0 5.0 3.0

0.4 0.4 0.4 0.4

30 75 45 6

30 30 30 30

DISCUSSION Factors affecting the formation of offretite or erionite The most significant difference in chemical composition between natural offretite and erionite is the Si/AI ratio l:~. This is also the composition parameter that has the strongest influence on the offretite fi)rmation. Si/AI ratios are compared with reported values in Table 6. Natural offretite has the highest alunfinium content observed, with Si/AI=2.5. In this work it is observed that the crystallinity increases with increasing alunfinium content, until a sharp drop

Figure 8 Micrograph of a T-type zeolite. (a) Zeolite T synthesized at 190°C; (b) detail of A

above Si/AI=3. This indicates that the stability of the offretite structure increases with the ionic contribution from the positive aluminium ions in the framework and from the cations at interstitial positions. The most fawmrable Si/AI ratio under these crystallization conditions is Si/AI=3, which corresponds to 4.5 aluminium atoms per unit cell. Erionite formed under similar conditions has Si/AI=4. It should be

Figure 9 (a) T-zeolite synthesized at 80°C; (b) zeolite synthesized according to Jenkins 17

ZEOLITES, 1986, Vol 6, November

479

Synthesis of erionite-offretite intergrovvth: K.P. Lillerud and J.H. Raeder

28 (Cu-rod) I

a

I0

15

20

I

I

I

I

I

I

potassium ions while the gmelinite cages and the wide channels are filled with calcium and magnesium ions. This cation filling is illustrated in Figure 12b. If only monovalent ions are available, as in the synthesis of offretite and erionite, the same ion arrangement will give Si/AI=5. This means that with monovalent ions more of the interstitial positions must be filled with cations. Aiello et al. ~s suggest that the hexagonal prisms are filled with potassium ions. This cation arrangement, one potassium or sodium ion in every hexagonal prism and every cancrinite cage; one cation, probably a TMA + ion, in every gmelinite cage and one cation per unit cell length of the wide channels, results in a Si/AI ratio of 3.5 (Figure 12c). In offretites with Si/AI=3, there must be 4.5 cations per unit cell. From geometrical considerations tiffs can only be achieved by a denser packing of cations in the wide channels. In accordance with the chemical

0.5

0

b

28 (Cu-rad) i

0.5

IO I

I

C

a

I

15

20

I

I

0.5

I

0.5-

0

b

I

'

tr-

12

I0

8

6

5

0.5

4

d-value (~) 10 High angle parts of the following diffractograms: (a) synthetic TMA offretite; (b) natural erionite (Pine Valley, Nevada, USA), (Filled lines are interpreted as phillipsite); (c) natural phillipsite (Pine Valley, Nevada, USA)

> 4o

Figure

noted that the maximum in crystallinity coincides with the formation of the largest and most regularly shaped crystals. If the number of cations is > 4.5 per unit cell geometrical restrictions in possible cation positions and thereby repulsive electrostatic forces decrease the stability. The key to understanding the difference in crystallization between the closely related zeolites offretite and erionite is their ability to accommodate counterions in the structure. Figures 12 and 13 show schematic illustrations of the two structures with different cation densities. In natural erionite and offretite Si/AI ratios are lower than for the ecluivalent synthetic zeolites since "Z+ ~Z+ bivalent ions, e.g. Ca and Mg , are responsible for a part of the charge compensation. In natural offretite the cancrinite cages are all occupied by

480

ZEOLITES, 1986, Vol 6, November

IZ

C

I

I

I

0.5

0

0¢v% 12 10

8

6 d-value (~)

5

4

11 High angle parts of the following diffractograms. (a) Zeolite T ('!"/2) (ca. 30% erionite); (b) Si-rich TMA offretite (H-off 10) (ca. 10% erionite); (c) ZSM-34 ex 1 (<10% erionite). Figure

(Filled lines are interpreted as phillipsite)

Synthesis of erionite-offretite intergrowth: K.P. Lillerud and J.H. Raeder Table 6

Reported Si/AI ratios in offretite and erionite

Zeolite Natural offretite Natural erionite TMA-offretite TMA-offretite TMA-offretite T-zeolite ZSM-34 TMA-erionite

Si/AI

Reference

2.5a 2.9--3.7a 4.0 4.1 2.3-4.0 3.0 3.8-4.0 4.0

13 13 11 18 This Paper This Paper This Paper This Paper

aThese natural zeolites contain bivalent cations

analyses the extra cations are represented as small Na+-ions. The lowest Si/AI ratio observed in this work is 2.3, which is equivalent to 5.5 cations per unit cell. The only way to achieve this cation density is by filling o f 2.5 cations per unit cell length o f the wide channels. In the offretite and erionite structures the columns of hexagonal prisms and cancrinite cages are identical. But, for erionite the gmelinite cages and the open channels that are found in offretite are substituted with supercages. Every two offretite unit cells, which are similar in size to the erionite unit cell, contain two

a

b

\

C

/

111®)

,)0

}

!1

111® /

\@

e I (~

} }

V~/

I\ ~ . J //e l

\

Figure 12 Schematic representation of the offretite structure. The large cations represent TMA + ions, while the smaller represent Na + and K+..(a) c-axis projection; (b) a-axis projection with 3 cations per unit cell, (cation filling in natural erionite); (c) a-axis projection with 4 cations per unit cell; (d) a-axis projection with 4.5 cations per unit cell; (e) a-axis projection with 5.5 cations per unit cell

ZEOLITES, 1986, Vo16, November

481

Synthesis of erionite-offretite intergrowth: K.P. Lillerud and J.H. Raeder

a

Table 7 •

\

Offretite with erionite intergrowths

Zeolite T-type Si-rich offretite ZSM-34

Erionite (%) 30 a 10 <10 b

aAII T-zeolites bUnder the detection limit

/

\

/

\

\

/

/ \

/

+0) Figure 13 Schematic representation of the erionite structure. The large cations represent TMA ÷ ions, while the smaller represent Na + and K +. (a) c-axis projection; (b) a-axis projection with 4 cations per unit cell

supercages. This implies that to reach Si/AI=3 every supercage must accommodate 2.5 cations. The observed ratio of Si/Al=4 corresponds to 1.5 cations per supercage. ., As seen in Figure 13, the filling o f more than two cations per supercage will brin K the/Ce c~tions close together, resulting in large repulsive. ~lectrostatic forces, which in turn will destabilize the structure. This means that the geometrical restrictions on the cation distribution are more severe for erionite than for offretite. Erionite cannot crystallize with the Si/A1 ratio that seems to be most favourable for offretite. This may well be the main reason why erionite is more difficult to synthesize in the laboratory. Thus, the use of a mixture of bivalent and monovalent cations may facilitate erionite synthesis and make it possible to synthesize erionite with a higher aluminium content. However, the use of bivalent cations is limited by their low solubility in the pH-range that gives reasonable crystallization times.

Factors affecting erionite intergrowth in offretite Offretites with erionite intergrowths synthesized in this work are summarized in Table 7.

482

ZEOLITES, 1986, Vol 6, November

Gels with essentially equal concentrations of sodium, potassium and TMA give offretites almost free from erionite intergrowths. Similar gels where TMA is substituted with sodium and potassium grow random offretite-erionite structures. Thus, T M A is obviously important in regulating the formation of erionite. In order to avoid erionite intergrowths, it is necessary that TMA is present in concentrations equivalent to the inorganic cations. TMA concentrations of ~ 1% of the total cation concentration, as in the synthesis according to Jenkins 17, behave like the T-type, both in amount of erionite intergrowth and in macroscopic crystallite appearance. The TMA concentration must, therefore, exceed a certain limit before it acts in a structure regulating manner. The detailed mechanism of the TMA template action is not understood, but the fact that the size of the TMA ion coincides with the internal space of the gmelinite cages in offretite is probably of importance. To establish in detail the effect of TMA further investigations should be carried out. When offretite is forced to crystallize with a high Si/AI ratio the possibility to adapt more cations is not utilized and the stabilities of offretite and erionite become nearly equal. The formation of intergrowths and erionite domains are therefore to be expected. In accordance with this, intergrowth structures are sometimes obtained from gels with Si/AI ratios above the most favourable for erionite formation. In this work, the amount of erionite intergrowths obtained by crystallization from Si rich gels could not be reproduced. A l u m i n i u m distribution in the offretite and e r i o n i t e lattices The above discussion on cation filling in the structure may also be used for deducing information about the aluminium distribution in the Si/Ai lattice. In both offretite and erionite there are two different tetrahedral sites; i.e. in the hexagonal prism (Tl site) and in the 6-ring (T2 site). Their location relative to the large cages, and thereby the possible densest location of cations, are different. In both structures there is more intracrystalline space near the T2 sites than near the Tz sites. The cation density will, therefore, be larger near the T2 sites, and to maintain short range electrical neutrality the aluminium density must also be larger on the T2 sites. The relative displacement of aluminium from T i to T2 sites is given by the fraction RI]R2, where Rj and R2 are the Si/AI ratios for the two sites. The magnitude of Rt/R2 will depend on the cation filling and it will be larger for erionite than for offretite. An estimate of this aluminium displacement can be

Synthesis of erionite-offretite intergrowth: K.P. Lillerud and J.H. Raeder

calculated assuming that the cations in the hexagonal prisms and those in the cancrinite cages compensate for aluminium on Tl sites, cations in the gmelinite cages compensate for aluminium on Tu sites and cations in the offretite channels contribute equally to compensation on both T~ and T2 sites. Cations in erionite supercages are assumed to compensate for aluminium on T2 sites only. This non-random Si/AI distribution has been investigated by high resolution 29Si n.m.r. 2~, Rt/R2 values are within the range 0.67-0.77. This result is in good agreement with the aluminium displacement calculated from the cation distribution and charge compensation. CONCLUSIONS The aluminium cation interaction is important for the stability of the offretite and the erionite structures. The cation filling in the extra-framework sites in these structures does, therefore, regulate the Si/AI ratio in the crystallization process. The position of cations are less restricted in offretite, due to the large channels in this structure. Offretite can therefore adapt more cations than erionite. When crystallized with only monovalent cations available, offretite will allow a lower Si/AI ratio than erionite, with the extra thermodynamic stability this implies. This may explain why offretite is more easily formed in the laboratory. When formed in nature at a lower pH, polyvalent cations are available, intracrystalline space restrictions are, therefore, no longer the critical factor.

REFERENCES 1 Rubin, M,K., Rosinski, E.J. and P ank, CS; US Pat. 4 086 186 (1978) 2 Inui, T., Araki, E,, Sezume, T., Ishihara, T. and Takegami, Y. React. Kinet. CataL Lett. 1981, 18, 1 3 Chen, N.Y., Schlenker, J.L., Garwood, W.E. and Kokotailo, G.T.J. CataL 1984, 86, 24 4 Bennett, J.M. and Gard, J.A. Nature 1967, 214, 1005 5 Gard, J.A. and Tait, J.M. Adv. Chem. Ser. 1971, 101,230 6 Gard, J.A. and Tait, J.M. Acta Crystallogr. 1972, B28, 825 7 Kokotailo, G.T., Sawruk, S. and Lawton, S.L Am. Mineral 1972, 57, 439 8 Millward, G.R. and Thomas, J,M. J. Chem. Soc. Chem. Commun. 1984, 77 9 Chen, N.Y., 'Proceedings of the 5th International Congress on Catalysis' (20-26 August, 1972), Palm Beach, FL, USA 1973, p. 1343 10 Inui, T., Ishihara, T. and Takegami, Y. J. Chem. Soc. Chem. Commun. 1981, 936 11 Aiello, R. and Barrer, R.M.J. Chem. Soc. A 1970, 1470 12 Whyte, Jr., T.E., Wu, E.L., Kerr, G.T. and Venuto, P.B.J. CataL 1971, 20, 88 13 Sheppard, R.A. and Gude, A.J. Am, Mineral. 1969, 34, 875 14 Inui, T,, Morinaga, N. and Takegami, Y, Appl. Catal. 1983, 8, 187 15 Givens, E.N., Plank, C.J. and Rosinski, E.J. US Pat. 4 079 096 (1978) 16 Breck, D.W. and Acaram, N.A, US Pat. 2 950 952 (1960) 17 Jenkins, E.E. US Pat. 3 578 398 (1971) 18 Aiello, R., Barrer, R.M., Davies, J.A. and Kerr, I.S.J. Chem. Soc. Faraday Trans. I 1969, 1610 19 von Ballmoos, R., 'Collection of simulated XRD powder patterns for zeolites', Butterworths, Guildford, UK, 1984 20 Barrer, R.M., 'Hydrothermal chemistry of zeolites', Academic Press, London, UK, 1982 21 Lillerud, K.P. in press

ZEOLITES, 1986, Vel6, .Nb~,emUer

'483