Characterisation of novel sodalites by neutron diffraction and solid-state NMR

Characterisation of novel sodalites by neutron diffraction and solid-state NMR

Solid Slate lonics 32/33 (1989) 430-435 North-Holland, Amsterdam C H A R A C T E R I S A T I O N O F N O V E L S O D A L I T E S BY N E U T R O N D I...

360KB Sizes 1 Downloads 11 Views

Solid Slate lonics 32/33 (1989) 430-435 North-Holland, Amsterdam

C H A R A C T E R I S A T I O N O F N O V E L S O D A L I T E S BY N E U T R O N D I F F R A C T I O N AND SOLID-STATE NMR

M.T. W E L L E R and G. WONG Department ( f Chemisto'. The University, Southampton. SO9 5NH. tlants.. UK

Received 31 May 1988: accepted for publication 8 August 1988

A range of sodalites of general formula Li,Na~ ,(A1SiO4)~,C12have been synthesised by structure conversion and ion exchange reactions. High resolution -'~SiNMR spectra of these compounds demonstrate a single resonance, the frequency of which varies smoothly with lithium content and the geomel~, of the Si-(OAI )4 tetrahedron determined by powder neutron diffraction. Incorporation of lithium in the sodalite structure results in contraction of the Si-O-AI bond angle to accommodate the preferred coordination geomet~' of the smaller cavity species. In partially exchanged materials the chloride ion is displaced awa~ from the cavity centre.

1. Introduction Sodalites are an unusual group of aluminosilicates in that they contain species such as C I - , SO?,- and N O 2 within the framework. They can be described by the general formula Ms(A1SiO4)6X, where M may be Na +, Li +, Ag + or more rarely other m o n o v a l e n t and divalent cation. Examples include noselite, N a~ ( A1SiO4 ) 6SO4, ultramarine, Naa (A1SiO4) 6S> and the parent sodalite, Naa(A1SiO4)6C12. Structurally the c o m p o u n d s are based upon a simple cubo-octahedral cage linked in three dimensions [ 1 ], where anions reside on sites at cage centres surrounded normally by four cations: in noselites, M ~ ( A I S i O 4 ) , X O 4 - , only half of the cages are occupied by anionic species. Both anions and cations can be substituted giving unparalleled compositional flexibility within the same aluminosilicate framework. Such changes in the nature of the cage species can be achieved readily such as through simple ion exchange in either solution or solid state, and occur with concomitant changes in the framework achieved by variations in the S i - O - A I bond angle. This change in T - O - T angle ( T = S i , A1) has been related by previous workers to changes in 2~Si N M R frequency for a n u m b e r of aluminosilicate materials [2,16]. In particular Newsam [15] has discussed -~Si N M R shifts in some sodalites similar to those studied in

the present work while Beagley et al. [ 17 ] h a v c also worked on such materials; in both cases powder Xray diffraction was used as the principal technique yielding cell constant values o f less accuracy than those d e t e r m i n e d here. The current work aimed to establish T-O-T/2'~Si correlations for a series of partially substituted sodalites. A major part of the work involved the detailed study of the sodalites by powder neutron and powder X-ray diffraction as well as by "~'~SiN M R .

2. Experimental Two main methods were used for sodalite synthesis in this work. Sodalite itself, Na~(AISiO4)~CI> was synthesised by the structure conversion method of Chang [3]: zeolite Linde 4A (BDH; 5-50 gm powder) was heated with excess NaCI at 800 C with several regrinds to produce a single phase product. This was then used as the starting material for a number of lithium exchanged derivatives. Four materials, Li3 ~sNa4 i~ ( AISiO4 )~,CI> Li t 41INa~,.6o( AISiO4 ) ~,C12, Li i ,/IN a~, 4n ( AISi()4 )~,CI_and Li~(AISiO4)~CI=,, were synthesiscd by heating Nas(AISiO4)~,CI2 with excess LiCI at 800~C for varying lengths of time to achieve ion exchange [ 4 ]. The stoichiometry of the products was determined

0 167-2738/89/$ 03.50 © Elsevier Science Publishers B.V. ( North-Holland Physics Publishing Division )

M. 72 Weller. G. Wong / Novel sodalites

by dissolution in dilute HC1 and then analysis by flame photometry. To check the purity of the synthesised materials powder X-ray diffraction using an INEL position sensitive detector-based system was performed (Ni filtered CuKcq radiation, 2 = 1.5406 ); a least squares method was used to refine a cell parameter value from each data set. The variation of lattice constant with composition for these materials has been studied by Taylor [4]; compositions determined from the refined cell parameter using this published data were in excellent agreement with our chemical analyses. Introduction of lithium into the parent sodalite material results in a marked reduction in lattice constant which falls from 8.8812(3) A. for Nas(AISiO4)6C12 to 8.4440(3) ~, for Li~ (A1SiO4)6CI2. The materials, Nas (AISiO4) 6C12, LiB(A1SiO4) 6C12 and Li385Na4 ~s(A1SiO4) 6C12 were then studied further by nuclear magnetic resonance and powder neutron diffraction. The former were carried out on the SERC national facility VXR300 spectrometer at Durham typically using 0.5 g sample masses with spin rates of 4 kHz; such rates were sufficient to give high resolution spectra with low intensity spinning side bands. Linewidths obtained varied from about 2.0 to 3.5 ppm. High resolution powder neutron diffraction data were collected using either the D2B instrument at the Institut Laue-Langevin (ILL), Grenoble or HRPD on ISIS at the Rutherford Appleton Laboratory (RAL). For D2B data typically 3-5 g of powder were loaded into a 12 mm diameter vanadium sample can. Data ranging from 7.525 ° to 160.000 ° were collected over periods of about 6-8 h at a wavelength of 1.5946/~. Samples were run in high intensity/medium resolution mode. Estimation of the background was achieved by averaging data points over areas of the spectra where no Bragg peak occurred. Data were treated using a full profile Rietveld refinement technique [ 5-7 ] : in the case of HRPD data, lhe latter was run as part of the Cambridge Crystallographic Library Subroutine package. Full profile refinement was used despite the materials being of a primitive cubic structure since significant peak overlap existed at high 20 angles in the PND data. The profile technique provided good refinement of all neutron data, yielding accurate cell constants in excellent agreement with the literature values quoted

431

by Beagley et al. [17]; comparison with Newsam's [ 15 ] values showed dissimilarities probably due to slight differences in stoichiometry e.g. Na6Si6AI6024' 1.6NaC1 c.f. NaB (A1SIO4)6C12. Oxygen positions were also obtained from which accurate Si-O-AI angles were calculated in good agreement with the less accurate XRD-derived values of Newsam [ 15 ]. Scattering lengths used were 0.960, 0.360, -0.214, 0.580, 0.350 and 0.420× 10 -~2 cm for C1, Na, Li, O, A1, and Si respectively. For each compound refinements of cell constant, zeropoint, scale factor, peak half-width parameters, thermal vibration parameters and atomic positions were carried out.

3. Structure refinement results The space group P43n, which has previously been employed successfully in other structural studies of sodalites by X-ray and neutron diffraction, was used in this work [8-11 ]. The powder neutron diffraction profiles obtained showed Gaussian Bragg peaks which were well separated from each other, confirming the cubic symmetry of the sodalites; because of this, Rl,t~,s,y or R/, where R/= ( 100. SU M [ ABS ( I ( OBS ) - I ( CALC ) * SCALE ) ] / SUM [ A B S ( I ( O B S ) )]), was taken as a good indication of the quality of fit in the Rietveld refinements. For sodalite, Nas(A1SiO4)6CI2, coordinates for framework atoms were taken from USns and Shulz's X-ray work on the same compound [12]. As expected the cavity Na + cations are found to exist in a tetrahedral configuration round the C1- (0, 0, 0) anion [13 ] and are directed towards 6-rings of the framework wall with fractional coordinates x = y = z = O . 1777(5). AI and Si were fixed at (1/4, 1/2, 0) and ( 1/4, 0, 1/2) respectively with C1 at (0, 0, 0); shifts of Na and O were allowed until final convergence to R/ (=5.39%) was achieved. The fit to the profile is shown in fig. 1. The final atomic positions are summarised in table 1 where the values determined by L6ns and Schulz, in excellent agreement, are given for comparison. Calculations using these atomic positions gave a value for the Si-O-AI bond angle in sodalite of 138.06°; other bond angles and distances are summarised in table 2. For lithium/chlorosodalite, Li8 ( A1SiO4 ) 6C12, time-of-flight

M 71 14'eller,

432

(;. I~bng/ N o v e l

,s'~dalite,s

1800

IGO0 I ~00 ~200 1000

80O GO0 CO0 200 0

i i i

I

I I I !lil II If;Ill[ I~ll: I::lil :ill ii I:ii I i l l l i l : !!l II!

: i

Li3.ssNa4.15(AtSi04)6[[2 F i g . I. P o w d e r

neutron

diflYacfion

,ll

i

Two T.ET^ profiles:

data from R A L was refined using the atom positions of its s o d i u m / c h l o r o - a n a l o g u e as a starting point. Refinements were based on the sodalite model structure obtained above but with lithium replacing sod i u m on the x, x, x site. A final R~ of 4.53% was reached with Li at (0.1748, 0.1748, 0.1748) and O at (0.1329, 0.4089, 0.1423), giving an S i - O - A 1 angle of 124.54 °. Comparison with the Na~(A1SiO4)~CI2 refinement results shows that coordinates of the atom on the 8 (e) site stay relatively unchanged; the smaller size o f cation in the cavity species is rather a c c o m m o d a t e d for by m o v e m e n t o f oxygen, i.e. by reduction in the S i - O - A 1 b o n d angle and reduction in lattice parameter. The material Li~ ~sNa4 ~ (A1SIO4)6C12, was again refined initially using atom positions from Na~ (AISiO4) ~CI> The large difference in scattering lengths for sodium and lithium makes it possible to distinguish these atoms easily despite their occupancy of similar sites. With the presence of the extra cation in the structure an a d d i t i o n a l position along (x, x, x ) was allowed in the refinement. Lithium was observed to prefer a site further from the cavity centre

calculated

(solid

l i n e ): o b s e r ~ , e d

( d o t s ) a n d fit.

than sodium and this could be explained by the former's requirements o f a site closer to the oxygen atoms of the framework. Naturally only one of the two cation sites along (x, x, x ) is occupied. Refinement of the isotropic temperature factor for chlorine lead to a very high value, a p p r o x i m a t e l y 8 A~, and so the position of this atom was allowed to vary away from (0, 0, 0), initially along (x, x, x ) : this removal of constraint for the chloride site enabled the refinement to converge readily to an R, of 7.04% with Na on (0.1675, 0.1675, 0.1675), Li on (0.1879, 0.1879, 0.1879), and O on (0.1369, 0.4250, 0.1476). Further relaxation in the position of the chlorine atom p r o d u c e d no significant i m p r o v e m e n t in the refinement. No evidence was found for the ordering of cations in the cavities and no refinement in a lower s y m m e t r y space group was attempted,

4. N M R results Solid state 2~Si M A S - N M R for the sodalites showed a single resonance confirming the ordered

M. T. Weller, G. Wong / Novel sodalites

433

Fable 1 Atomic coordinates, temperature factors, lattice constants, N M R chemical shifts. Position

Na~ (AISiO4) 6C12 ao=8.8812(3) A. :gSi = - 85.34 ppm R~=5.39%

Na~ (AISiO4)6C12 ref. [12]

Lis (AISiO4) 6C12 a~ = 8.4440 ( 3 ) ~ 2°Si = - 76.73 ppm Rt=4.53%

li38sNa4 i s (AISiO4)6Cl2 ") a~=8.7101 (1) A. z~'Si= - 8 1 . 6 8 ppm R~= 7.04%

x ), z B~o

x .V z B~so

x

y z B,so

x y z B,~o

2(a)

8(e)

6(c)

6(d)

24(i)

2C1

8Na

6AI

6Si

240

0 0 0 2.15(14)

0.1777 ( 5 ) 0.1777(5) 0.1777 ( 7 ) 1.84(16)

0.25 0.50 0.00 1.09(26)

0.25 0.00 0.50 0.27(20)

0.1395 ( 3 ) 0.4382(2) 0.1494 ( 3 ) 1.17(7)

2C1

8Na

6A1

6Si

240

0 0 0 1.90(4)

0.1777(4) 0.1777(4) 0.1777(4) 1.49(4)

0.25 0.50 0.00 0.80(7)

0.25 0.00 0.50 0.35(4)

0.1401 (4) 0.4385(3) 0.1487(4) 0.89(2)

2C1

8Li

6A1

6Si

240

0 0 0 1.77(7)

0.1748 ( 5 ) 0.1748 ( 5 ) 0.1748 (5) 1.96(14)

0.25 0.50 0.00 1.06(16)

0.25 0.00 0.50 0.59(12)

0.1329 (2) 0.4089 (2) 0.1423 (2) 0.73(3)

2C1

4.15Na

3.85Li

240

0.0242(4) 0.0242(4) 0.0242(4) 3.84(16)

0.1675(5) 0.1675(5) 0.1675(5) 0.95(14)

0.1879(9) 0.1879(9) 0.1879(9) 2.80(36)

0.1372( 1 ) 0.4256(8) 0.1475( 1 ) 1.18(10)

"~ Silicon and aluminium on (0.25, 0, 0,5) and (0.25, 0.5, 0) as above; temperature factors 0.73(5 ) and 1.12(6) respectively.

array of silicon and aluminium and the sole existence of Si-(OA1)4 environments expected in these compounds with a Si:AI ratio of 1.0. (Fig. 2 ). A more extensive range of sodalites including lithium and potassium exchanged sodalites, noselites and other halide derivatives have been studied using 295i NMR by us [ 14]. Chemical shift (3) values have been observed over the range - 76 to - 91 ppm, much larger than has been recorded previously for silicon in an Si-(OA1)4 environment [13]. Fig. 3 shows the smooth correlation observed between resonance frequency and Si-O-A1 bond angle for the series of compounds studied in [ 14].

5. Discussion The major change which occurs when sodium is

substituted for by lithium in sodalite is a reduction in the lattice parameter. This occurs through a marked reduction in the Si-O-A1 bond angle; this displaces the framework oxygen towards the centre of the sodalite cavity and reduces the lithium-oxygen distance to a typical value of approximately 2.10

A. For the material Li3 8sNa4 ~s(AISiO4)6C12 the halide ion was displaced off its normal (0, 0, 0) site at the cavity centre. As this is a disordered material the atom positions represent average unit cell contents. In such partially exchanged compounds the cavity species is of the form (Li,Na4_,)C1. Li and Na atoms coordinate to 6-membered oxygen rings of the cage wall with the smaller lithium atom occupying a site closer to these oxygens and thus further from the cavity centre. This results in a movement of Cl along the (x, x, x) direction away from (0, 0, 0), presum-

M. 7~ H 'd/Io', G. 11 }m j.,/,%'o~c1 sodulm's

434

Table 2 Bond lengths (&) with ESD's in parentheses and bond angles ( degrees ).

29

: ' ) ' P~,rr, i ?c,,

Na~ ( AISiO4).CI: Na CI Na-OL Na-( ): Si-() AI-()

2.73318) 2.352(4 ) 3.085 ( 4 ) 1.623(6) 1.740(6)

Si-O-AI O-Na-CI O-Na-O

138.06 115.04 103.38

jj.I ./

. J 1

Li~ s,Na4 ~ ( AISiO4 ),£712 12C

Na-CI Li CI Na-() Na-O~ Na-O: ki-(), Li-O, Si-O M-O

2.163(8) 2.469(8) 2.269(4) 3.151(41 2.145(5 ) 3.098 ( 5 ) 1.627( 5 ) 1.742(5)

Si-O-AI

132.1 1

2.557 ( 8 ) 2.(127( 5 ) 3.172 ( 5 ) 1.637(5) 1,736(5)

AI-()

Or,,#¢r °}

Fig. 3. > S i d versus Si ()-AI angle, rcf. [ 14 ].

Li~ ( A1SiO4 )(,C12 Li-('l Li-Ot ki O, Si-O

S ~ ,1' AI

Si-O-AI O-Li-CI O-Li-O

124.54 112.59 106.18

all three materials of two different M - O distances ( M = N a , Li). Three, O( 1 ), sit below the Al/Si-plane directed into the cavity, while the other three. O( 2 ), sit above the A1/Si-plane directed slightly outside the cavity that is into an adjoining cage. Thus two sets of M - O distances are observed, M - O 1) and M -

r 0¢4j~ 043 t

ably towards a site occupied by lithium and not sodium, to maintain reasonable Li-C1, Na-CI. L i - O and N a - O distances. Bond length calculations showed the existence in

047 i

i

/

C¢+I ]

/

/

./

y coordlnafe

O@i

i? 0'16 " x

coordinate ,i

015 • 0% 8!3

z coor'dl~fe

012 3 il . 0 •

,

,:

8':

93

30

,-f,

-tic

.

. I

. 2

.

. 3

. 4

. :}

6

7

,q

L It ~ .lm( or>ter~b x ) 120

Fig. 2. "% NMR spectrum o f L ~(AISi()4),,( I,.

PP~

t2C

tqg. 4. (P%gcn position as a funclion of lithium ¢'ontcnt ( \ ) Ibl Li, ['Nra~ , ( AttiC )4 },f J:.

3I. I: l~ "eller, G. ~t "ong / Novel sodaliws

0 ( 2 ) . In sodalite, Nas(SiA10~)6C12, the s o d i u m coordination geometry consists o f 1×C1+3× O ( 1 ) + 3 × O ( 2 ) , though the d i s t a n c e to O (2) are s o m e w h a t longer t h a n those to O ( l ); in the fully lithiated m a t e r i a l the L i - O ( 2 ) distance is very long for a l i t h i u m - o x y g e n i n t e r a c t i o n a n d it is p r o b a b l y better to c o n s i d e r the l i t h i u m g e o m e t r y as being app r o x i m a t e l y tetrahedral. P l o t t i n g oxygen c o o r d i n a t e s as a f u n c t i o n o f lithi u m c o n t e n t ( x in Li,Na~_,(A1SiO4)6CI2) (fig. 4) illustrates the m a n n e r in which oxygen moves. The greatest v a r i a t i o n occurs in the ), p a r a m e t e r which can be interpreted in t e r m s o f a twisting o f the oxygen a r o u n d the s i l i c o n - a l u m i n i u m direction in order to p r o v i d e a better c o o r d i n a t i o n geometry for lithium. The S i - O a n d A I - O distances show only a very slight v a r i a t i o n with degree of l i t h i u m occupation. The structure r e f i n e m e n t of p o w d e r n e u t r o n diffraction data yields accurate S i - O - A I b o n d angles which correlate very well with 29Si N M R r e s o n a n c e frequencies (fig. 3 ). In less crystalline c o m p o u n d s o f the sodalite type where diffraction m e t h o d s are of l i m i t e d use it is n o w possible to predict reasonably accurate f r a m e w o r k positions from the position o f the silicon-29 resonance. The a u t h o r s would like to t h a n k the S E R C for use o f the V X R 3 0 0 N M R facility at D u r h a m a n d for

435

grants associated with this work, i n c l u d i n g a stud e n t s h i p for G.W. T h a n k s also to the SERC, R A L a n d the ILL ( G r e n o b l e ) for the p r o v i s i o n o f n e u t r o n b e a m facilities.

References [ 1] A.F. Wells, Structural inorganic chemist~', 5th Ed. (Oxford Universily Press, Oxford, 1984). [2] S. Ramdas and J. Klinowski, Nature 308 (1984) 521. [3] I.F. Chang, J. Eleclrochem. Soc. 121 (1974) 815. [41D. Taylor, Contrib. Mineral. Petrol. 51 (1975) 39. [5] H.M. Rietveld. J. Appl. Cryst. 2 (1969) 65. [6] A.W. Hewat, Harwell Report, AERE-R7350 (1973). [ 7 ] J.C. Matthewman, P. Thompson and P.J. Brown, J. Appl. C~'stallogr. 15 (1982) 167. [8] S. Luger, J. Felsche and P. Fischer, Acta Cryst. C3 (1987) 1. [9] R.R. Neurgaonhar and F.A. Hummel, Mater. Res. Bull. 11 (1976) 61. [ 10] I. Hassan and H.D. Grundy, Acta C~'st. B40 (1984) 6. [ 11 ] M.S. Perlmut'ter, L.T. Todd and E.F. Farrell, Mater. Res. Bull. 9 (1974) 65. [ 12 ] V.J. L6ns and H. Schulz, Acta Cryst. 23 ( 1967 ) 435. [ 13 ] J. Klinowski. Prog. NMR Spectrosc. 16 (1984) 237. [ 14 ] M.T. Weller and G. Wong, J. Chem. Soc., Chem. Commun. (1988) 1103. [15] J.M. Newsam, J. Phys. Chem. 91 (1987) 1259. [ 16] R.H. Jarman, J. Chem. Soc., Chem. Commun. (1983) 512. [ 17 ] B. Beagley, C.M.B. Henderson and D. Taylor, Mineral. Mag. 46 (1982) 459.