Modified Zeolites Part 1 .-Dealuminated Mordenites and their Silanation BY RICHARD M. BARRER" AND JEAN-CHRISTIAN TROMBEt Physical Chemistry Laboratories, Chemistry Department, Imperial College, London SW7 2AY Received 23rd March, 1978 A series of partially dealuminated H-mordenites having different ratios Si01/A1203 has been prepared from Na-mordenite by various treatments with acid. From these a series of dealuminated Na-mordenites has been made by treating the corresponding H-mordenites with dilute NaOH. HO The loss of water, including that from the groups [ +A1 S i t ] and the " nests " of hydroxyls, H404, has been investigated as a function of temperature for the dealuminated forms. This showed at 360°C the existence of some OH groups not originating from [ +Al S i r ] .Such groups were attributed to -OH originating from nests. The silanation of the dealuminated H- and Na- mordenites was investigated quantitatively. In the mordenites free of zeolitic water the ratios H2 (evolved)/SiH4 (chemisorbed) were always between 1.5 and 2 and each ratio usually remained reasonably constant over the course of a run. This suggested a rather constant ratio of primary chemisorption to secondary reaction during many runs. The ratio increased above 2 only if zeolitic water was also present. No evidence was found for a significant amount of the reaction \ HO \ -OH HO- [-OH HO-] +SiH4 -+ in which SiH4 interacts with intact " nests ", H404, to regenerate defect-free framework. The Elovich equation appeared able to represent the kinetics of chemisorption of SiH4.Since Keough and Sand reported the stability of mordenite in acids and Barrer and Makki completely dealuminated clinoptilolite with mineral acid without destroying the essential crystal structure, considerable further work on acid and dealuminated zeolites has been undertaken. 3-7 Dealumination with acids necessarily involves also the formation of the H-zeolite, the two reactions being formally represented as 0 M+ 0 H30+ + A1 S i r +H30+ + +A1 S i e +M+ (1) \ / 0 \ / 0 and 0 *.*' \ H' H I 0 0- + 4HzO + A13+. (2) / le H 3 0 + -0-Al-0-+ 3H30f -+ -0 / H H I 0 t Present address : Institut National Polytechnique de Toulouse, Laboratoire de Physico-Chimie des Solides et des Hautes Temperatures, 38 rue des 36 Ponts, Toulouse, France. 2786R . M.BARRER AND J.-C. TROMBE 2187 Both reactions (1) and (2) may be followed by other reactions. On outgassing, and perhaps also slowly at room temperature,* one may have HO \ H30+ 3 A1 S i e + [+A1 Sie]+H,O. \ / (3) 0 The silanol OH groups have been found to react with SiH4 to give silanated zeolites 9 y through such processes as SiOH + H-SiH3 + + Si-0-SiH3 + H2 3 Si-O-SiH3 + HO-Si 4 + Z- Si-O-SiH2-O-Si f + H2 + Si-0-SiH2-0-Si f + HO-Si 4 -+ + Si-0-SiH-0-Si 4 + H2. (4) ( 5 ) (6) I 0-Si 4 Reactions ( 5 ) and (6) imply either mobility of protons from more distant -OH groups or else pairs or groups of three -OH groups close enough together to react with the same SiH4 molecule. The stability of the nests, H404, formed in reaction (2) may be low, due to loss of water on heating.Thus Thackur and Weller ’ reported no nests stable above 100°C in various dealuminated H-mordenites. However the reaction of a dealuminated H-mordenite with SiH4 required a number of -OH groups in excess of those formed in reaction (3),9 so that some hydroxyls originating from H404 nests appeared to have survived the prior outgassing at 360°C. The reaction of intact nests with SiH4 would, if complete, replace the A1 removed by dealumination by Si, and would yield crystalline silicas having the topology of the parent zeolite : 1 0 I I H404 + SiH4 -+ -0-Si-0- + 4H2. (7) 0 I Silanation of dealuminated H-zeolites offers the opportunity to study the hydroxyls in nests, and provides means of changing permanently the sorption, molecule sieving and catalytic behaviour of dealuminated crystals. It was therefore of interest to investigate silanation of such crystals to complement studies already made on silanation of H-mordenite and H-Y 9* lo and intracrystalline reactions of silane with zeolitic water.ll EXPERIMENTAL Na-mordenite and H-mordenite were supplied (as Na-Zeolon and H-Zeolon) by the Norton Co.Dealuminated H-mordenites were prepared from the Na-mordenite by refluxing 10 g samples in 150 cm3 AnalaR hydrochloric acid of different normalities and for different times (table 1). After the products had been washed free of acid they were dried at 50°C. Parts of the products were then treated with 0.1 mol dmV3 NaOH solution at room temperature for 24 h to generate dealuminated Na-mordenites* TABLE 1 .-PREPARATION OF DEALUMINATED MORDENITES conc.HCI/ reflux time designat ions rnol dm-3 /h 0.1 rnol d m - 3 NaOH 0.1 mol dm-3 NaOH 0.1 mol dm-3 NaOH -+ Na-M1 -+ Na-M2 >11 5 M3 -+ Na-M3 2 3 M1 6 4 Mz 6 28 M42788 MODIFIED ZEOLITES HO \ 8 Na+ [+A1 Si<]+NaOH + +A1 S i r +H20. (8) \ / 0 Analyses were made by standard methods : Si02 and A1203 were determined gravi- metrically and Na by absorption spectrophotometry. The total water was found by ignition (zeolitic and hydroxyl), and the weight loss was also measured as a function of temperature both by thermogravimetry, and more accurately on a silica spring balance by heating in vacuo over 24 h at each of a series of fixed temperatures. The monosilanol -OH content OH \ (associated with groups [ +A1 Sic]) was obtained for dealuminated H-mordenites both from the A1 content and by chemisorption of NH3 l2 in a silica spring gravimetric unit.The sample, outgassed at 360"C, was exposed to NH3 at 25°C for 24 h and the physically sorbed NH3 was evacuated until the weight of the sample was constant: HO 0 NH,+ Si<. (9) \ [+A1 Sir]+NH3-+ >A1 \ / 0 The apparatus used for silanation has been described previo~sly.~ At any time, t, the amounts of H2 evolved, of SiH4 chemisorbed and of SiH4 physisorbed were determined. Before all runs (dehydration, silanation, sorption) the zeolites were stored for at least one day over saturated Ca(N03)2 at -56 % RH. For silanation the zeolites were outgassed for not less than a day, nearly always at the temperature selected for the reaction with SiH4. In all tables or figures relating to silanation the amounts of SiH4 chemisorbed or H2 evolved are expressed in mmol/g-l of zeolite stored over saturated Ca(N03)2.RESULTS AND DISCUSSION ANALYSES AND THERMAL BEHAVIOUR Crystallinity to X-rays of all the dealuminated mordenites was good, and there was only a small contraction of the unit cell on extensive removal of Al. The molar ratios SiO2/Al,O3 ranged from 9.52 in parent Na-mordenite to 61-62 in M4 (table 2). The weight losses under vacuum at 360°C (LUV) and the loss on ignition (LOI) are given in columns 2 and 3 of the table. At 360°C under vacuum virtually all zeolitic water is removed [cf. fig. 1 (a) for Na-mordenite] so that LOI-LUV represents hydroxyl water retained at 360°C either in nests or as monosilanols or both.As expected, LOI-LUV for Na-mordenite is extremely small. The H-mordenite (H-Zeolon) with SiO2/Al2O3 = 1 1.84 is already partially dealuminated if derived Ca(N03), SATURATED SOLUTION TABLE 2.-ANALYTICAL RESULTS ON MORDENITES AIR DRIED AT 50°C AND STORED OVER A mordenite sample % wt loss vacuum Na-mordenite H-mordenite Na-M 1 MI Na-Mz Mz Na-M3 Na*-M3 M3 M4 at on 36OOC ignition (LUW (LO11 mmol g-1 of OH from Al analysis silanols nests asmono- in 12.27 12.60 7.22 1.22 11.22 l.18 9S2 - - 0.37 12.53 14.87 8.26 0.0 12.37 1.05 11.84 2.09 2.05 2.6 1.81 4.71 12.14 0.917 13.24 1.83 2.87 3.93 2.07 5.25 - 12.47 13.37 7.43 0.935 11.98 0.905 13.24 - 2.83 1.0 - 13.62 16.49 9.16 0.053 12.55 0.69 18.19 1.39 5.05 3.19 1.57 4.98 10.99 13.02 7.23 0.83 12.78 0.49 26.02 - 6.81 2.26 - 14.47 16.82 9.34 - 13.01 0.50 26.02 1.00 6.93 2.61 1.20 4.84 - 14.22 17.76 9.87 11.46 13.62 7.57 0.935 12.28 0.675 18.19 - 4.92 2.40 - - - 0.66 26.02 9.33 11.56 6.42 - 14.30 0.232 61.62 0.46 10.16 2.4s 0.79 3.87R. M.BARRER AND J.-C. TROMBE 2789 from a Na-mordenite with Si02/A1203 of -9.5. On this assumption reactions 3 and 2 give ideally 2.09 mmol of OH per g as monosilanols and 2.05 mmol g-' in H404 nests. The value calculated from LOI-LUV is 2.6 mmol of OH per g still present after outgassing at 360°C. This, and all the other values recorded in column 11 of table 2, lie between the extremes of -OH ideally present as monosilanols [reaction (3)] TABLE 3.-ESTIMATED WATER LOST FROM H404 NESTS AT 360°C fraction lost for monosilanol content mordenite derived from sample A1 content NH3 chemisorption M1 0.27 0.35 Mz 0.64 0.68 M3 0.77 0.80 M4 0.80 0.83 H-mordenite 0.75 0.61 13 13 1 ternperaturel'c FIG. 1.-Dehydration under high vacuum of mordenites : (a) Na-mordenite ; (b) curve 1, Na-MI ; curve 2, Na-M2; curve 3, Na-M3; (c) curve 1, Mz, curve 2, H-mordenite, curve 3, M4.and the sum of this -OH content plus that ideally present in nests [reaction (2)], as calculated in columns 9 and 10. Thus even if at 360°C no -OH present as monosilanols had been lost it is still necessary to assume a second source of --OH groups, some of which remain at 360°C. This source is ascribed to -OH from H404 nests. From the figures in columns 9, 10 and 11 , and assuming no loss of water from2790 MODIFIED ZEOLITES monosilanols, the fractions evolved up to 360°C of the total water which the nests could produce by complete dehydroxylation are those given in table 3, column 2.In the mordenites M1 to M4 the acidic -OH associated with groups HO [+A1 Si 43, i.e. the monosilanol content, determined by chemisorption of NH3 at 25°C (table 2, column 12) exceeds, but follows the trend of, values based upon the A1 content (column 9). For the H-mordenite, on the other hand, chemisorption of NH3 gave a 13 % lower value than that based on the A1 content. If the monosilanol content at 360°C is taken as that found by chemisorption of NH3 then the fractions of total water in H404 nests lost by 360°C are those given in table 3, column 3. There is reasonable consistency between the values in columns 2 and 3, with fractions lost tending to increase with increasing dealumination among the mordenites MI to M4.When samples of M, were outgassed at 200 and 360°C respectively the acidic silanol contents found by chemisorption of NH3 were 1.5, and 1.5, mmol g-l of zeolite. This supports the view that up to 360°C there is minimal loss of water from acidic silanols. Further information was obtained by measuring weight losses of the parent Na- and dealuminated mordenites which had been air-dried at 50°C and then stored over saturated Ca(NO,),, and finally outgassed at each of a series of fixed temperatures for at least 24 h (fig. 1). With all the dealuminated mordenites (which includes the H-mordenite of table 2) an upward inflexion appeared above 200°C [fig. l(b) and (c)], which was not observed with Na-mordenite, and therefore is not due to zeolitic water.The inflexion was observed also with each of the three dealuminated Na-mordenites of table 1, in which the acidic silanols present as \ HO \ 0 [+A1 S i r ] had been converted according to reaction (8) to [aAl-0-Sie]. Since these latter three zeolites should contain no acidic silanols, and also because for M, the ammonia chemisorption indicated no loss of acidic -OH up to 360"C, the inflexions above 200°C are attributed to an accelerated loss of water from non- acidic -OH in the original nests formed in reaction (2). The nests are evidently considerably less stable to heating than are the acidic silanols. Water can be evolved from them in two stages : The first stage could occur more easily than the second, especially if distortion and resultant steric factors arising from the first stage reorient the remaining pair of hydroxyls.It is probable that some dehydroxylation occurred below 200°C. In parent Na-mordenite with SiO2/A1,O3 = 9.S5 the curve of water loss against temperature [fig. l(a)] showed that 97.2 % of the zeolitic water had been removed at 200"C, assuming that all had been lost at 360°C. Table 4 gives the unit cell compositions of mordenites M1 to M4 and of H- mordenite, all air-dried at 50°C and then stored over saturated Ca(NO,),. In evaluating the zeolitic water contents it was assumed that at 50°C there had been no irreversible loss of water from H404 nests. If this were not the case the zeolitic water contents would increase by the amounts of the loss of water from nests at 50°C.Accordingly the zeolitic water contents are minimum values. The amounts of zeolitic water and of NH3 physisorbed at 25°C and 50 Torr (table 4) both decrease inR . M . BARRER AND J.-C. TROMBE 279 1 the sequence M1 to M4. Evidently selectivity towards NH, and retentivity for water decrease with increasing dealumination. DEAL U M I N A T ED Na-M 0 R D E N I T E S When the dealuminated H-mordenites MI, M2 and M3 were soaked in 0.1 mol dm-3 NaOH and washed to give Na-MI, Na-M, and Na-M,, the Na-contents exceeded those of the framework Al, especially for Na-M, and Na-M, (table 2). However further washing of Na-M, reduced the Na-content to give Na*-M, (table 2). The extra Na could thus represent free base imbibed by the crystals in excess of amounts required for reaction (8).SI LA NATION OF DE ALUMIN ATED H-MORDENI TES The kinetics of silanation at different temperatures are shown in fig. 2(a) for Mz outgassed at the silanation temperature. Reaction begins rapidly but soon becomes slow. The ratio R = H2 (produced)/SiH, (chemisorbed) was always considerably above unity (table 5) so that, as reported earliery9. lo reactions other than reaction (4) are involved. At 280°C a light brown colour was observed on the glass wall of the sample holder. This may be caused by formation of polymers (SiHx),.13 At 200°C or below no discolouration could be seen. TABLE 4.-uNIT CELL COMPOSITIONS OF MORDENlTES AIR DRIED AT 50°C AND STORED OVER A Ca(NO& SATURATED SOLUTION calculated LO1 due to water from cell OHas OH in HzO mordenite weight cell composition monosilanols nests zeottic Na-mordenite 3505 Na8 .3 ,[&.3 ,Si3 9. ci709 6124. 5H20 12.60 H-mordenite 3197 H6.70[A16.7~Si39.67(oH)6.52089.4,]19.8H20 1.88 1.84 11.15 M1 3266 H5.g9[A1s.99Si39.67(0H)g.36086.6,]24.54H20 1.65 2-58 13-52 M2 3 118 H4.36[A14.3 LSi3g.67(OH)i 5.88080.12]1 8S1H20 1.26 4.58 10.68 M3 3048 H3.05[A13.0sSi3g.67(OH)2i.i2074.88]16.39H20 0.90 6.24 9.68 For all runs in fig. 2(a), even though reaction was incomplete when the runs were ended, the H2 evolved (-2 mmol g-1 at 65,100 and 200°C) exceeded the acidic silanol content obtained by chemisorption of NH3 (1.58 mmol g-l). Other -OH groups are therefore involved which arise from nests or from residual zeolitic water.However, the outgassing at 200°C removed virtually all zeolitic water [fig. l(a)] so that residual water is not expected to influence the results for MI to M4 because these are increas- ingly less retentive of molecular water, as shown in a previous section. Accordingly it is considered that some -OH groups from partially dehydroxylated H404 nests are taking part in the chemisorption of SiH4. However, at the end of each run the number of SIH, niolecules chemisorbed was less than the maximum number of sites theoretically available. This maximum is the number of acidic silanol OH groups plus the number of nests, which for M2 is 1.39 + 1.26 = 2.65 mmol of sites per g. At 200°C fig. 2(a) shows only 1.2 mmol SiH4 chemisorbed per g. Likewise the H2 evolved when the run ended (-2 mmol g-l) was less than the -OH content remaining at 360°C (3.1 mmol g-I) as given in table 2, column 1 1.Reaction was still proceeding very slowly when the run was ended [fig. 2(a) and (b) and fig. 41, but evidently a considerable number of sites of hydroxyls do not react or react only very slowly with SiH,. The results in table 6 were obtained by silanation of M2 at 200°C but with different pretreatments. In the first run unreacted SiH4 was evacuated after 162.5 h M4 2766 Hi.2,[Al1.2gSi39.6,(oH)28.1~067.8~]3.0sHz0 0.42 9.16 2.002792 MODIFIED ZEOLITES TABLE 5.-REACTION OF SiH4 WITH MORDENITE Mz AT DIFFERENT TEMPERATURES time of 65°C lOO~C0 200"c 280°C treatmentlh R % Si R A si R % Si R % Si 0.5 1 2 3.5 4 19.5 21 26.5 41.5 44.5 56 65 79 92.5 101 125 1.76 1.69 1.89 l.S3 1.93 1.92 l.S7 2.09 1.g7 1.94 2.43 1.98 1.95 1.93 1.89 1.87 1.85 1.g3 2.60 1.91 2.72 r( bo I i!! 0 .5 l-++-l- 10 dtlh+ FIG.2.-(u) Silanation of mordenite M2 : 1.73 2.24 1.71 2-73 1.g4 1.70 2.60 1.73 2.S1 l.69 2.g1 1.72 3.OS 2.07 1.70 2.75 2.10 l.67 3.23 1.74 3.39 2.41 2.48 1.77 3.38 2.62 1.69 3.32 1.74 3.6, 1.82 3.88 H2 evolved SiH4 chemisorbed 65°C X + 100°C 0 0 200°C A A 280°C 0 .. (b) Silanation of various dealuminated mordenites at 200°C : H2 evolved SiH4 chemisorbed H-mordenite + X M1 0 0 M2 n A M3 0 M4 v v.R . M. BARRER AND J . - C . TROMBE 2793 and subsequent release of hydrogen was then measured. In 100 h over 0.22 mmol H2 per g were formed through slow secondary reactions. Finally oxidation by addition of excess water at 280°C brought the ratio H2 (evolved)/SiH, (chemisorbed) very near to the theoretical maximum of 4.The kinetics of chemisorption at 200°C were followed in greater detail in a special gravimetric unit [fig. 3(a)]. Fig. 3(b) indicates that the kinetics are reasonably represented by the Elovich l4 equation for t $ to : where Qt is the silane chemisorbed at time t and k and to are constants. This equation can be interpreted in terms of a model in which the energy of activation for chemisorption varies linearly with TABLE 6.-SILANATION OF MORDENITE Mt AT 200°C outgassing at 200°C. addition of 0.85 % time outgassing at 200°C outgassing at 360°C H 2 0 by wt at 200°C treatment/h silanation ato200"C silanation at 220°C silanation at 200°C R h Si R A Si R % Si 1 1.75 2.46 1.49 1.91 4 1.51 2.04 2.28 1.70 2 1.75 2.6, 22 1.75 2.92 23 1.52 2.36 24 2.18 2.Z5 44.5 1.57 2.44 56 1.75 3.12 66.5 1.58 2.53 92.5 1-76 3.20 101 1.6, 2.60 121 1.77 3.23 162.5 1.77 3.31 SiH4 pumped off, 200°C oxidation by water at 280°C 1st dose 2nd dose 100 h at 1.96 19 h 2.50 23 h 3.94 Fig.2(b) compares silanation and hydrogen evolution for M, to M4 and H- mordenite, all outgassed and silanated at 200°C. Table 7 records ratios R=H2 (evolved)/SiH, (chemisorbed). Particularly for M, to M4 the % Si added decreased with increasing dealumination. This could mean that reaction of SiH4 with the HO \ acidic -OH of the groups [ 3 A1 -OH originating from the nests. S i t ] occurs more readily than reaction with S I LAN AT I 0 N 0 F D E A L U MI N A TED Na-M OR D E N I T E S As noted earlier acidic -OH should not occur in the dealuminated Na-mordenites. Accordingly chemisorption of SiH4 should take place primarily at -OH groups2794 MODIFIED ZEOLITES originating from the nests, provided zeolitic water is absent.Therefore the zeolites Na-M,, Na-M2, Na-M3 and Na*-M, were outgassed and silanated at 2oo"c, as in table 7. In addition Na-M, was outgassed and silanated at 100°C. The H2 evolved and SiH4 chemisorbed are shown as functions of time in fig. 4 and the ratios R and % Si added in table 8. When tables 7 and 8 are compared it is seen for MI to M3 HO \ that removal of the acidic -OH of the groups [ s A l Si 41 has substantially FIG. 3.-Silanation of mordenite Mz at 200°C: kinetic analysis. (a) The rate curve; (b) test of Elovich equation. TABLE 7.-sILANATION OF DEALUMINATED H-MORDENITES AT 200°C AFTER OUTGASSING AT time of treatmentlh 1 4 20 22 26.5 43.5 56 70 72.5 90.5 92.5 114 121 137 139 158 1 62 182 H-mordzni te R XSi 2.03 1.69 2.00 2.04 1.98 2.41 1.99 2.4, 1-96 2.60 1.99 2.72 1-96 2-86 MI R %Si 1.73 2.95 1-75 3-06 1-78 3-26 1.79 3.38 1.81 3.45 1.84 3.58 1.86 3.60 1.75 2.92 1.75 3.12 1.76 3.20 1.77 3.23 MI R %Si 1.70 0.8, l.67 1.15 1.67 1.48 1.69 l.65 1.71 1.72 1.73 l.83 1.76 1.85R . M.BARRER AND J.-C. TROMBE 2795 reduced the initial rates of chemisorption of SiH4. This supports the view given in the previous section that -OH groups originating from nests are less reactive to SiH4 HO S i r ] groups. Removal of some Na from Na-M3 to give than those in [ +A1 Na*-M, (table 2) did not change the kinetics of chemisorption appreciably (table 8, columns 5 and 6).When the runs were terminated the ratios of the numbers of SiH4 molecules chemisorbed to the ideal numbers of nests (from table 4, column 3) were as follows : Na-MI, 1.3; Na-M,, 0.g8; Na-M,, 0.5,. Reaction in Na-M1 must therefore \2796 MODIFIED ZEOLITES As noted earlier these contain excess Na (table 2) which may be present as NaOH. This aspect of silanation requires further study. RATIOS H2 (FORMED)/SiH4 (CHEMISORBED) When tables 5-8 are considered together the following properties of R = H2 (produced)/SiH, (chemisorbed) are seen : (i) R lies in the range 1.5 to 2.0 for all the mordenites outgassed at 200 or 360°C. (ii) R does not depend strongly upon temperature of silanation, although at 65 and 100°C R is larger than at 200 and 280°C.(iii) During the course of any run R usually stays nearly constant. (iv) If zeolitic water is present R increases above 2, in agreement with an earlier study 11 (table 8, column 2 ; table 6, column 6). TABLE 8.-sILANATION OF DEALUMINATED Na-MORDENITE time of treatment/h 1 2 4 4.5 20 22 25.5 26.5 42.5 44.5 48 50 52 66 68 70.5 89 91 112 114 1 20 139 186 outgassing and silanation at 100°C. Na-M2 % Si 1 .26 1.49 1.31 1.67 1.67 1.79 1.87 outgassing and silanation at 200°C Na-l'$ R /,Si 1.75 1.47 1.70 1.74 1.73 lA4 1.70 2.36 l.69 2.55 1.69 2.84 1.70 3.13 1.72 3.21 1.71 3.33 Na-Mi R /,Si l.S7 1.84 1.39 1.81 1.56 l.S7 1& 1 . 8 ~ 2.26 1.86 2.43 1.g7 2.53 1-91 2.58 Na-v3 R Si 1.75 1.49 1.79 1.74 l.81 2.13 1.84 2.36 1.86 2.42 1.86 2.53 1.87 2.73 Na*+ R /,Si 1.83 1.51 l J O l.81 1.85 2.22 1.85 2.44 l.87 2.s3 l.89 2.63 Of the mordenites outgassed and silanated at 200°C the largest value of R was 2.0, for H-mordenite, the form richest in acidic hydroxyls.The smallest value of R was obtained with mordenite M, outgassed at 360°C and silanated at 200°C. In this latter system when reaction was terminated 0.g3 mmol of SiH4 had been chemisorbed per g of zeolite and 1.53 mmol of H2 had been evolved. The zeolite outgassed at this temperature had 1 .39 mmol g-l of acidic hydroxyls and 3.1 mmol total hydroxyl g-l (table 2). The H2 evolved exceeds the total acidic OH content but represents only about half the total hydroxyl. Because for all the zeolite samples nearly free of zeolitic water R lies between 1.5 and 2.0 the primary chemisorption reaction (4) cannot occur alone.The constant values of R during the runs also suggest that during a particular run the primary and secondary reactions occur in approximately constant proportions. Because R is in the range 1.5 to 2.0 a high proportion, if any, of reaction (7) is unlikely.R . M . BARRER AND J.-C. TROMBE A possible process additional to reactions (2)-(6), which gives R = 2, is 2797 ] +2H2. -0- ]+SiH4 + [ -O-SiH2-O- It involves the hydroxyl pairs generated by the first stage of reaction (10). CONCLUSION The existence has been demonstrated in dealuminated H-mordenites and Na- mordenites of -OH groups which can react with SiH4 but which do not arise from residual zeolitic water or as monosilanols in the groups [ + A1 Sir]. These have been attributed to hydroxyls originating from the H404 nests formed in the initial dealumination by acids.The analytical evidence shows that some -OH from this source still remains in dealuminated mordenites outgassed at 360°C. The loss of water from nests is necessarily a two-step process, and the second stage is likely to be less easy than the first. However, water loss from the nests has been found to occur considerably more easily than water loss from acidic -OH in the groups [+A1 Sir]. No evidence has been obtained that Si from SiH4 can replace to any marked extent the four H-atoms of the nest to recreate a defect-free framework, at least up to 200°C. Thus in zeolites free of molecular water the ratios H2 (evolved)/ SiH4 (chemisorbed) did not exceed 2 instead of reaching 4, even for Na-M1, Na-M2 and Na-M3 in which acidic -OH should be absent, but where nests are initially present. HO \ HO \ A. H. Keough and L. B. Sand, J. Amer. Chem. SOC., 1961,83,3536. R. M. Barrer and M. B. Makki, C d . J. Chem., 1964,42,1481. R. M. Barrer and D. L. Peterson, Proc. Roy. SOC. A, 1964,280,466. R. M. Barrer and B. Coughlan, in Molecular Sieves (SOC. Chem. Ind., London, 1968), p. 141. W. L. Kranich, Y. H. Ma, L. €3. Sand, A. H. Weiss and I. Zwiebel, in Molecular Sieve Zeolites in Adv. Chem. Ser. (Amer. Chem. SOC., 1971), no. 101, p. 502. G. T. Kerr, J. Phys. Chem., 1968,72,2594. ' D. K. Thakur and S. L. Weller, in Molecular Sieves, Adv. Chem. Ser. (Amer. Chem. SOC., 1973), no. 121, p. 596. R. M. Barrer and J. Klinowski, J.C.S. Furahy I, 1975,71,690. R. M. Barrer, R. G. Jenkins and G. Peeters, in Molecular Sieves 11, Amer. Chem. SOC. Symp. Ser., 1977, 40, 258. lo R. M. Barrer, E. F. Vansant and G. Peeters, J.C.S. Faraday I, 1978, 74, 1871. I f R. M. Barrer and J . 4 . Trombe, in preparation. I2 B. K. G. Theng, E. F. Vansant and J. B. Uytterhoeven, Trms. Furadby Soc., 1968, 64, 3370. I3 E. G. Rochow, Comprehensive InorQanic Chemistry (Pergamon Press, Oxford, 1973), vol. I, l4 S. Yu. Elovich and G. M. Zhabrova, Zhur. fiz. Khim., 1939,13, 1761. Is B. M. W. Trapnell, Chemisorption (Butterworth, London, 1955), p. 104 et seq. chap. 15, p. 1364. (PAPER 8/556)