J. CHEM. SOC. FARADAY TRANS., 1994, 90(9), 1323-1328 Conformational Properties of Monosubstituted Cyclohexane Guest Molecules Constrained within Zeolitic Host Materials A Solid-state NMR Investigation Abil E. Aliev and Kenneth D. M. Harris* Department ofChemistry, University College London, 20 Gordon Street, London, UK WCIH OAJ Raphael C. Mordi Department of Chemistry, University of St. Andrews, St. Andrews, Fife, UK KY16 9ST The conformational properties of monosubstituted cyclohexane guest molecules (C6H1, X with X = CH, , OH, CI, Br and I) included within microporous solid host materials (silicalite-I, H-ZSM-5, NH,-mordenite and zeolite NH,-Y) have been elucidated via high-resolution solid-state '3C NMR spectroscopy. For all of the inclusion compounds investigated, the fraction of monosubstituted cyclohexane molecules in the equatorial conformation is similar to that in solution, suggesting that these host materials do not impose any significant constraints upon the conformational properties of the monosubstituted cyclohexane guest molecules.For the mono-halogenocyclohexane guest molecules (C6H11X with X = CI, Br and I), this result is in marked contrast to the situation for the same guest molecules in the thiourea host structure, for which the conformational properties of the guest molecules are substantially different from those of the same molecules in solution. For cyclohexanol (C6H1 OH) in H-ZSM-5, some amount of dicyclohexyl ether (C6H1 OC6H1 1) is observed, and is analogous to the proposed production of dimethyl ether in the first stage of methanol-to-gasoline conversion on this zeolite.The comparatively low temperature (ambient temperature) at which this conversion from cyclohexanol to dicyclo- hexyl ether occurs is noteworthy. In addition to our high-resolution solid-state 13C NMR studies of these materials, 'H MAS and 27AI MAS NMR spectra have also been recorded, and are discussed. One major impetus underlying current research on solid inclusion compounds is the desire to investigate the proper- ties that may be conferred upon an organic 'guest' molecule by virtue of embedding it within a crystalline 'host' material and to understand the extent to which the properties of the guest molecule may be altered from those of the same mol- ecule in dispersed phases or in its 'pure' crystalline phase.Many solid host materials are known, encompassing aluminosilicates, aluminophosphates, organic solids and many other classes of material. These host materials possess a wide variety of different inclusion topologiesy1v2 such as linear tunnels, isolated cages, networks of intersecting tunnels and/or cages, and two-dimensional regions within layered hosts. Monohalogenocyclohexanes (C6HllX, with X = C1, Br, I) generally exist as an equilibrium between axial and equato- rial conformations (with a chair conformation of the cyclo- hexane ring). In the liquid and vapour phases there is a considerable excess of the equatorial conf~rmer,~-~ whereas in the solid state it has been reported6 that only the equato- rial conformation exists.However, when included as guest molecules within the thiourea host structure (the inclusion topology of which comprises uni-directional tunnels'), C6HllC1, C6Hl,Br and C6H111 have been shown to exist predominantly in the axial conformation. These results have been established from IR,8 Raman' and high-resolution solid-state 13C NMR1'*' techniques. High-resolution solid-state 13C NMR investigations' of thiourea inclusion compounds containing monosubstituted cyclohexane guest molecules (C6H1 ,X) have shown that these guest molecules can be subdivided into two classes: those with X = C1, Br and I have a predominance of the axial con- former (fraction of equatorial conformer CQ.0.05-0.15), whereas those with X = CH,, NH,, and OH have a pre- dominance of the equatorial conformer (fraction of equatorial conformer ca. 0.82-0.97). The fact that the conformational properties of the C6HllX guest molecules within their thio- urea inclusion compounds depend critically upon the identity of the substituent X reflects the fine and subtle energetic bal- ances that exist for these inclusion compounds, and impor- tant insights into the reasons underlying the preference for the axial conformation of the C6Hl,C1 guest molecules in thiourea have been obtainedI2 from the application of a theoretical approach that has been developed for the predic- tion and rationalization of structural properties of one-dimensional inclusion compounds.In this paper, we have extended our studies of the confor- mational properties of organic guest molecules in constrained solid-state environments to encompass monosubstituted cyclohexane guest molecules within several crystalline zeolitic host materials. From the results, direct comparisons can be drawn between the inclusion compounds of these micro- porous hosts and the inclusion compounds containing the same guest molecules within the thiourea host structure. Such comparisons are particularly interesting in view of the pro- spect that the thiourea inclusion compounds (and other solid organic inclusion compounds) may, in many respects, rep- resent model systems for structurally similar (i.e. possessing uni-directional tunnel topologies) zeolitic host materials. In this work, the following microporous host materials have been considered : silicalite-I, H-ZSM-5, NH,-mordenite and zeolite NH,-Y.ZSM-5 is a medium-pore zeolite, the structure of which consists of a set of sinusoidal tunnels inter- secting a set of straight tunnels, each with 10-membered ring openings. The diameter of the straight tunnels is CQ. 5.3-5.6 A, and the diameter of the sinusoidal tunnels is ca. 5.1-5.5 A. The framework structure of silicalite-I is the same as that of ZSM-5, and can therefore be regarded as the purely siliceous version of ZSM-5. Zeolite Y has 12-membered rings of ca. 7.4 8, diameter leading to a supercage with diameter ca. 13 A, whereas mordenite has 12-membered rings forming one-dimensional tunnels with diameter CQ.6.5-7.0 A. The conformational properties of monosubstituted cyclo- hexane guest molecules (C6H1 ,X with X = CH3, OH, Cl, Br and I) included within these host materials have been investi- gated uia high-resolution solid-state '3C NMR spectroscopy. The results are discussed in the light of our previous studies" of the conformational properties of the same guest molecules included within the thiourea host tunnel structure and in the solution state. Experimental The following host materials were used in this work: silicalite-I, H-ZSM-5 (Laporte Inorganics, RD 1136/88), NH,- mordenite (Si/Al =lO.l), and zeolite NH,-Y (Strem Chemi- cals, Inc.). All of these samples were calcined in a mume furnace at 773 K for at least 24 h before use. The mono- subs ti tuted cyclohexanes were obtained commercially and were used without further purification, with the exception of C,H,,Cl which was distilled at 415 K before use.Two different methods for including the guest molecules within the host materials were considered. In method A, the monosubstituted cyclohexanes were adsorbed into the host materials by contacting about 5 cm3 of the liquid mono- substituted cyclohexane with ca. 0.6-1.0 g of the powdered host material in a round-bottomed flask under vacuum for ca. 3 days. After this period, the excess liquid was removed under vacuum and the solid allowed to dry. The flask was sealed and removed to a dry box, in which the solid was transferred to the rotor to be used in the solid-state NMR experiments.In method B, the host material was exposed to the liquid monosubstituted cyclohexane for ca. 20-40 h in an ultrasonic bath at ca. 303 K. After this treatment, the excess liquid was removed and the solid washed with 2,2,4-tri- methylpentane and then allowed to dry. The solid was packed into the NMR rotor in the open laboratory (ie.not in a dry box). It is clear that the amount of water present within the host materials may be higher for those materials prepared via method B. In order to confirm that the monosubstituted cyclohexanes were adsorbed on the internal (rather than the external) sur- faces of the host materials, a control experiment was carried out by subjecting a sample of quartz to the same preparation procedures (with C6H1 ,C1 as the potential adsorbate).High- resolution solid-state I3C NMR spectra of the samples of quartz recovered following these preparation procedures revealed no detectable amounts of C,H,,C1 (which, if present, would necessarily have been adsorbed on the exter- nal surfaces of the quartz). On the basis of the results of these control experiments, it was concluded that the amounts of monosubstituted cyclohexanes on the external surfaces of the host materials, subjected to the same preparation procedure, would be insignificant, and that any detected amounts of these molecules must be adsorbed on the internal surfaces of the host materials. Solid-state 'H, 13Cand 27Al NMR spectra were recorded at 500.13, 125.76 and 130.32 MHz, respectively, on a Bruker MSL500 spectrometer using a standard Bruker magic-angle sample spinning (MAS) probe with double-bearing rotation mechanism.The samples were studied as polycrystalline powders in zirconia rotors (4 mm external diameter) and MAS frequencies between 2 and 12 kHz (with stability better than ca. k10 Hz) were used. Single-pulse and cross-polarization (CP) techniques were used to record the 13C NMR spectra, under conditions of MAS and with inverse-gated 'H decoupling applied during acquisition. Although the CP technique is intrinsically non-quantitative (since the efficiency of polarization transfer may vary from one carbon environment to another), our experiments have shown that single-pulse and CP (contact time =1 ms) tech- niques give the same relative intensities for the resonance lines for the cyclohexane derivatives at room temperature.I3C and 'H chemical shifts are given relative to tetra-J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 methylsilane and 27Al chemical shifts are given relative to the signal assigned to Al(H,0)63+ in the sample. The stability and accuracy of the temperature controller (Bruker B-VT1000) were ca. k2 K. Results and Discussion Fig. 1 shows I3C CP-MAS NMR spectra of C,H,,Br/H-ZSM-5 (prepared by method A), recorded at 293 and 200 K. At 200 K, there are two sets of signals with approximate inte- grated intensity ratio 4 :1. The I3C NMR resonance lines with chemical shifts 39.4, 25.0 and 27.9 ppm are assigned as carbons C(2), C(3) and C(4) in the equatorial conformer and the resonance lines at 34.7 and 21.1 ppm are assigned as carbons C(2) and C(3) in the axial conformer.These chemical shifts are in close agreement with those found for the C,H 1,Br/thiourea inclusion compound.' The resonance line for C(l) is broad (particularly at low temperature), and it is possible that second-order quadrupolar effects from bromine contribute to this broadening. l3 The I3C CP-MAS NMR spectrum of C,H, ,CI/H-ZSM-5 (prepared by method A) recorded at 200 K also contains two sets of signals, assigned to equatorial [6,(,, =38.2 ppm; SCO, =24.8 ppm; bC(,)=27.1 ppm] and axial [6c(2)=34.7 ill"l'""ll""ll'''~~i' 60 50 40 30 20 10 6 Fig.1 13C CP-MAS NMR spectra of C6H,,Br/H-ZSM-5 (pre- pared using method A) recorded (a)at 200 K and (b)at 293 K J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 ppm; 8c(3)= 20.7 ppm] conformers, with approximate inte- grated intensity ratio 4 :1. As in the case of C6H1 ,Br/H- ZSM-5, the resonance line for C(l) is broad (particularly at low temperature), and it is possible that second-order quad- rupolar effects from chlorine contribute to this broadening.' Although such low-temperature '3C NMR spectra provide direct identification of the axial and equatorial conformers, the low loadings of guest molecules within the host materials and the fact that the resonance lines are comparatively broad (as discussed in more detail below) complicate considerably the application of 13C NMR for an accurate study of the conformational properties of the guest molecules in these inclusion compounds. To determine the relative proportions of equatorial and axial conformers at ambient temperature, we have employed a technique which considers NMR param- eters averaged by the axial-equatorial exchange.This tech- nique has been applied successfully to study conformational equilibria for substituted cyclohexanes and other systems from solution-state NMR data. l4 For the conformational equilibrium between the axial and equatorial conformers in monosubstituted cyclohexanes (i.e. a simple two-site exchange process), the fraction (peq)of molecules in the equa- torial conformation is determined from : where 6,, and Sax are the chemical shifts for a given carbon in the equatorial and axial conformers, respectively, and (6) is the averaged chemical shift observed for the same carbon in the measured spectrum (recorded under fast-exchange conditions), see Table 1.Although this approach requires the intrinsic temperature dependence of the chemical shifts for the axial and equatorial conformations to be known, it has been shown14 that this technique provides acceptable results for the C(2) and C(3) carbons in monosubstituted cyclo- hexanes in the solution state; furthermore, the temperature dependence of the isotropic chemical shifts for these mol- ecules in the solid host materials considered here may be expected to be less than in solution, in view of the compara- tively small coefficients for thermal expansion of these materials.For the C(2) carbon of the C6H11X guest mol- ecules considered here, the 13C NMR chemical shift can be determined with an accuracy of ca. k0.2 ppm, leading to a percentage error of ca. 10% in the estimate of peq; although this error is comparatively large, the results obtained by this method are nevertheless sufficiently accurate to distinguish whether the conformational properties of the guest molecules I325 in these host materials resemble those of the same molecules in the thiourea host structure or in solution. Using this tech- nique, the values of peq for C6HllCl/H-ZSM-5 and for C,H, ,Br/H-ZSM-5 were estimated [from the chemical shift values for C(2)] to be ca.0.8 in both cases. Values of peq for C6HI1C1/H-ZSM-5 and for C6HllBr/H- ZSM-5 have also been determined directly from the inte- grated intensity ratios for the peaks due to C(2) in the 13C CP-MAS NMR spectra recorded at 200 K, and it is inter- esting to note that these values are in close agreement (see Table 2) with those determined at 293 K via the method described above. It should be noted that our assessment of peq from integrated peak areas in the 13C CP-MAS NMR spectra recorded at 200 K is based upon the assumption that the CP efficiency for a given carbon environment [specifically C(2)] is the same in the axial and equatorial conformations. This assumption is reasonable in view of the fact that the protons dGtly bonded to carbon will dominate the polar- ization transfer from protons to carbon in the CP experi- ment, together with the fact that the geometry of the CH, group is essentially the same for the axial and equatorial con- formations.For all the other inclusion compounds studied in this Table 2 Fraction (peq) of equatorial conformer, as a function of temperature, for monosubstituted cyclohexane (C,H ,X)molecules in solid host materials and in solution X CH3 OH I Br c1 environment thiourea solution H-ZSM-5 thiourea solution H-ZSM-5 thiourea solution H-ZSM-5 thiourea solution H-ZSM-5 H-ZSM-5 thiourea solution H-ZSM-5 H-ZSM-5 silicalite-I NH,-mordenite NH,-Y T/K Peq 208 0.97 200 0.99 293 1 198 0.82 200 0.96 303 0.9 177 0.15 220 0.76 293 0.8 208 0.05 200 0.75 200 0.83 293 0.8 200 0.08 200 0.81 200 0.76 293 0.8 293 0.8 293 0.8 293 0.8 Table 1 13C NMR chemical shifts for monosubstituted cyclohexane (C,H, ,X)guest molecules within microporous host materials CH3 H-ZSM-5 293 33.5 36.3 27 27 21.1 OH H-ZSM-5 293 72.5 36.4 25 25 I H-ZSM-5 293 34 40.9 25.0 29.0 Br H-ZSM-5 293 53 38.6 27 27 Br H-ZSM-5 200 * 39.4 (eq) 25.0 (eq) 27.9 (eq)* 34.7 (ax) 21.1 (ax) * c1 H-ZSM-5 293 60 37.5 26 26 c1 H-ZSM-5 200 * 38.2 (eq) 24.8 (eq) 27.1 (eq)* 34.7 (ax) 20.7 (ax) * c1 silicalite-I 293 60 37.5 26 26 c1 NH,-mordenite 293 63 37.6 26 26 c1 NH,-Y 293 63 37.6 25.5 27.0 ~~ Data for the equatorial and axial conformations are indicated by (eq) and (ax), respectively; entries marked * in the table refer to data that cannot be established from the spectra.1326 paper, the boundary values for a,,and 6,, in eqn. (1) were taken as the values determined from the low-temperature high-resolution solid-state '3C NMR spectrum of the appro- priate C6H ',X/thiourea inclusion compound.' 'The values of pes determined via this approach are presented in Table 2, which also contains the corresponding data (taken from ref. 11) for the C6H, ,X/thiourea inclusion compounds and for C6Hl1X in solution. From Table 2, the proportion of mol- ecules in the equatorial conformation is greater than ca. 0.8 for all C6H,,X guest molecules in the zeolitic host materials at ambient temperature, and the conformational properties of the C,H, ,X guest molecules in these inclusion compounds are thus very similar to those of the same molecules in solu- tion. The uncharacteristic conformational behaviour found for the C6H,,X guest molecules with X =c1, Br and 1 in their thiourea inclusion compounds is not reproduced for these guest molecules in the zeolitic host materials investi- gated here.Fig. 2 shows I3C CP-MAS NMR spectra, recorded at 293 K, of C6Hl,Cl/H-ZSM-5 prepared by method A [Fig. &I)] and by method B [Fig. 2(b)]. The linewidths of the reson- ances are in the range 180-400 Hz, and the chemical shifts (Table 1) are in good agreement with those expected for the equatorial conformer of C6H1 ,C1 (on the basis of substituent chemical shift parameters for the C1 substituent determined from low-temperature solution state 13C NMR studies' ',13,14 and from low-temperature solid-state 13C NMR studies of the C,H,,Cl/thiourea inclusion compound' ').These linewidths are larger than those (16-27 Hz) observed previously' 'for the C6H1 ,Cl/thiourea inclu- sion compound at 293 K.The linewidth of the resonance (at ca. 60 ppm) due to the C(1) carbon (directly bonded to C1) is particularly large in comparison with the other resonances. For the other inclusion compounds studied in this paper, the 13C NMR linewidths are also larger than for the same guest molecule in its thiourea inclusion compound. There are several possible explanations for this observation.One pos- sible explanation is that each broad resonance line observed in the 13C CP-MAS NMR spectra of C6HllC1/H-ZSM-5 comprises a superposition of several isotropic peaks (with dif- ferent isotropic chemical shifts), each representing C6H 1,X guest molecules in a different environment with respect to the H-ZSM-5 host structure. The non-Lorentzian lineshape observed (Fig. 2) for the C(1) carbon in C6HllC1/H-ZSM-5 is consistent with this proposal. Specifically, the signal due to the C(l) carbon in Fig. 2(a) and (b) can be considered as a superposition of at least two components: a 'broad' com-ponent at higher frequency and a 'narrow' component at lower frequency. In samples prepared by the two different methods, the ratio of the 'broad' and 'narrow' components is different, resulting in the different lineshapes for this signal in Fig.2(a)and (b). For C6Hl1C1/H-ZSM-5 (prepared by method B) and C6Hl,Br/H-ZSM-5 (prepared by method B), a weak signal was detected at ca. 73.1 ppm in the 13C NMR spectrum [Fig. 2(b)]; this signal is assigned to the C(1) carbon of C6Hl,0H (with C6HllX :C6H110H z7 for X =C1 and Br). A signal at ca. 74.5 ppm was also detected for C6Hl,C1/NH,-Y, and is also assigned to the C(1) carbon of C6Hl10H (with C6H11C1 C6Hl10H X 2). The presence Of C6Hl10H is attributed to the occurrence of a hydration reaction of C H -!L-6 C1 and-C6H_Br within the H-ZSM-5 and NH,-Y host materials. In this regard, it should be recalled that the amount of water present within the host materials is likely to be higher for those materials prepared uia method B.This assignment of the signal at ca. 73-75 ppm in the 13C NMR spectra of C6H1 ,Cl/H-ZSM-5, C6H1 ,Br/H-ZSM-5 and C&i1 ,Cl/NH,-Y samples was confirmed by recording the J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 !I IIII I,,, I,,, I,,/ ,Ill ,,I, ,I,, I1' 80 70 60 50 40 30 20 10 I I,,, ,I,, ,,,I I,,, ,,It I,,, ,(,,,, 80 70 60 50 40 30 20 10 s Fig. 2 13CCP-MAS NMR spectra of C,H,,Cl/H-ZSM-5 recorded at 293 K: (a)for the sample prepared using method A and (b)for the sample prepared using method B 13C NMR spectra for C6H1 ,OH/H-ZSM-5 samples prepared using methods A and B [Fig. 3(a) and (b)]; in these spectra, the chemical shift for the C(l) carbon is 73.2 ppm.For C6H, ,OH/H-ZSM-5, however, there is an additional signal at 79.2 ppm in the 13C NMR spectrum. This signal is assign- ed as the C(1) carbon of dicyclohexyl ether (C6H1,0C6H11), with C,HllOH :C6H,1OC6Hll x 4 for the sample prepared by method A and 1.3 for the sample prepared by method B. I3C NMR chemical shift increments reported" for alkyl groups R in ethers C6H110R corroborate our assignment of this signal to the c(1) carbon of C6Hl,0C6Hl,. Other signals [C(2), C(3) and C(4)] for C6H,,OC6Hl1 are in the region 24-36 ppm and overlap the signals due to C6H,,0H. It is interesting to note that an analogous conversion of methanol into dimethyl ether has been proposed as the first stage of methanol-to-gasoline conversion on H-ZSM-5.l6 As for C6H, ,X/thiourea inclusion compounds,' 'there is substantial line narrowing in the 'H MAS NMR spectra of C6H1,X/H-ZSM-5 inclusion compounds. Analogous line narrowing has also been observed for zeolite and cyclo- phosphazene inclusion compounds containing various substi- tuted benzenes as the guest species.17*'* It has been suggested' that, for such systems, all the dipolar interaction tensors have their principal axes in the same direction, either J. CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 1"''I~"'I'"'I""1"''1"''1""1'"'I""I 90 80 70 60 50 40 30 20 10 6 Fig. 3 13C CP-MAS NMR spectra of C6H,,0H/H-ZSM-5 record- ed at 293 K: (a) for the sample prepared using method B; (b)for the sample prepared using method A because of special features of the crystal stru~ture,'~ or because the whole molecule performs rotational motions which substantially average the intermolecular dipolar inter- actions.'8 In the case of C6HllC1/H-ZSM-5, the narrow 'H NMR resonance lines observed for the guest molecules (Fig. 4) suggest that 'H-'H dipoleaipole interactions (including the intermolecular dipole-dipole interactions between the 'H of the host and the 'H of the guest) are aver- aged substantially as a result of high conformational, rota- tional and translational mobility of the guest molecules at room temperature. For C6H1 ,CI/H-ZSM-5, 'H NMR reson- ances at 6 =1.7 ppm [linewidth at half-height AvIl2 x 0.6 ,..I' ,I..,. .,. .I. .,. . . 10.0 8.0 6.0 4.0 2.0 0.0 -2.0 4.0 s Fig. 4 'H MAS NMR spectrum of C6HllCl/H-ZSM-5 (prepared using method B) recorded at 293 K (MAS frequency =9.3 kHz). The sharp signal at 0.2 ppm is due to tetrakis(trimethylsi1yl)silane (added as an internal standard); this signal does not overlap with any signals from C6Hl ,Cl/H-ZSM-5. -. 100 80 60 40 20 0 -20 -Fig. 5 27Al MAS NMR spectra of C,H,,Cl/H-ZSM-5 (prepared using method A) recorded (a) at 200and (b)at 293 K kHz; assigned as H(2), H(3) and H(4) of C,H,,Cl], 6 =4.0 ppm [Avl12 x 0.4 kHz; assigned as H(l) of C6HllC1], 6 =2.2 ppm (AvlIz xO.3 kHz; assigned as 'H in H-ZSM-5) and 6 =5.6 ppm (Avl12 x 0.7 kHz; assigned as 'H in H-ZSM-5) are observed in the 'H MAS NMR spectrum recorded at 293 K (Fig.4). For unloaded zeolites,20 the line at 2.2 ppm can be assigned to the non-acidic OH groups (e.g. terminal OH groups at the outer surface of the zeolite or at structural defects). The line at 5.6 ppm can be assigned to the acidic bridging OH groups in the zeolite framework and/or to water molecules present within the structure; the comparatively large linewidth may arise from the presence of both of these types--of OH group. 27Al MAS NMR spectra were also obtained for the inclu- sion compounds with H-ZSM-5 as the host material. Fig. 5 shows 27Al MAS NMR spectra of C,HllC1/H-ZSM-5 recorded at 293 and 200 K. The broad line at 54.9 ppm (Av,,, x 690 Hz at 293 K and Avl12 x 860 Hz at 200 K) is assigned the the four-coordinated aluminium in the host framework. The narrow line at 0 ppm (Avl,, x 60 Hz) in the spectrum recorded at 293 K is assigned to the six-coordinated aluminium in Al(H,0),3+.21 The line at 0 ppm is considerably broader at 200 K (Avl,, x 1 kHz) than at 293 K, consistent with the suggestion that the Al(H20)63+ has less motional freedom at low temperature.Concluding Remarks The high-resolution solid-state '3C NMR results reported here suggest that the solid host materials silicalite-I, H-ZSM- 5, NH,-mordenite, and zeolite NH,-Y do not impose any major constraints upon the conformational properties of the monosubstituted cyclohexane guest molecules studied (C,H,,X with X = CH, , OH, C1, Br and I); the relative pro- portions of the axial and equatorial conformations of these guest molecules are the same, within experimental error, as those of the same molecules in solution (at the same temperature).It is interesting to speculate on the reasons underlying the difference in behaviour for the C,H,,X guest molecules with X = C1, Br and I in the thiourea host structure (for which the axial conformation predominates) compared with the zeolitic host structures considered here (for which the equatorial con- formation predominates). One major difference concerns the effective loading of guest molecules within these host struc- tures. For thiourea inclusion compounds, the host structure is stable only when there is a dense packing of guest mol- ecules within the tunnels, and this tunnel structure collapses to a more compact structure if the guest molecules are removed ; thus, the thiourea inclusion compound containing a particular type of guest is known at only one specific guest : host ratio (corresponding to ‘saturation’).Zeolitic hosts, on the other hand, generally remain stable if the guest molecules are removed, and, as a consequence, inclusion compounds can be formed between a particular zeolitic host and a particular guest species with a range of guest concen- trations [ranging from zero (‘empty’ host) to some maximum value (corresponding to saturation)]. For the preparation methods employed in this work, the loading of guest mol- ecules is actually rather low [in the range 0.1-1.0 guest mol- ecules per loo0 A3 of the host material, determined from elemental analysis (carbon percentage) results], which is con- siderably lower than saturation.It is, therefore, reasonable to assume that, in the zeolitic inclusion compounds investigated in this paper, the guest molecules are essentially ‘isolated’ from each other. As discussed in detail elsewhere,22 there are major funda- mental differences in considering the optimum structural properties of guest molecules for cases (such as the thiourea inclusion compounds) in which the inclusion compound can exist with only one specific guest : host ratio (corresponding to saturation), in comparison with those cases (such as the zeolitic inclusion compounds) in which the guest : host ratio is an experimental variable.If the inclusion compound is of the former type, and if the host structure is a strictly one- dimensional tunnel structure, it is possible to predict and rationalize the structural properties of the inclusion com-pound by applying a theoretical approach that has been developed For the inclusion compounds with zeolitic hosts (that can be prepared with essentially arbitrary guest : host ratio), on the other hand, this theoretical approach is not valid. Furthermore, the question of predict- ing and rationalizing the structural properties of the guest molecules on the basis of computed potential-energy func- tions would, in any case, become considerably more difficult for host Structures (such as ZSM-5) that do not consist of independent one-dimensional tunnels, and the methodology (analogous to that developed previously for the strictly one- dimensional inclusion compounds) required for such systems has not yet been developed.Nevertheless, for zeolitic hosts containing low loadings of guest molecules, it is qualitatively clear that the host-guest interaction energy and tne intramol- ecular potential energy of the guest molecule are the major determinants of the structural and conformational properties of the guest molecules (since the guest molecules in the inclu- sion compounds with low loadings of guest probably behave as essentially isolated molecules, the guest-guest interaction can be considered negligible). The constraints imposed upon the guest molecules by the host environment (quantified by a consideration of the host-guest interaction) could have a crucial influence in controlling the conformational properties J.CHEM. SOC. FARADAY TRANS., 1994, VOL. 90 of the guest molecules, and could, in principle, outweigh the effect of the intramolecular potential energy in dictating the observed conformational behaviour of the guest molecules. The presence of C6H1 ,OC,H, ,in the inclusion compound formed between C,H,,OH and H-ZSM-5 is interesting, par- ticularly in view of the proposal that the corresponding ether (dimethyl ether) is produced in the first stage of methanol-to- gasoline conversion on this zeolite. It is particularly inter- esting that, in the case of C,H,,OH/H-ZSM-5, significant amounts of C,H,,OC,H,, are produced even at ambient temperature.Control experiments were performed to prove that the C6HllOC6Hl1 is produced within the zeolite, and not in the liquid C,H,,OH phase during preparation of the inclusion compound (i.e. during adsorption of C,H1 ,OH from the liquid phase). It is interesting that the relative amount of C,H,,OC,H,, produced is higher for the inclu- sion compound prepared by method B than for the inclusion compound prepared by method A, although the exact reasons underlying this fact remain to be investigated in detail. We are grateful to the Royal Society and the SERC for the award of postdoctoral fellowships (to A.E.A.) and to the SERC for general support (to K.D.M.H.).Professor Sir John Meurig Thomas and colleagues at the Royal Institution are thanked for providing some of the zeolite samples used in this work. References 1 Inclusion compounds, ed. J. L. Atwood, J. E. D. Davies and D. D. MacNicol, Academic Press, London, 1984, vol. 1-3. 2 K. D. M. Harris, Chem. Br., 1993,29,132. 3 M. Larnaudie, Compt. Rend., 1952,235, 154. 4 P. Klaeboe, J. J. Lothe and K. Lunde, Acta Chem. Scand., 1956, 10, 1465. 5 K. Kozima and K. Sakashita, Bull. Chem. SOC. Jpn., 1958, 31, 796. 6 P. Klaeboe, Acta Chem. Scand., 1969,23,2641. 7 K. D. M. Harris and J. M. Thomas, J. Chem. SOC., Faraday Trans., 1990,86, 1095. 8 K. Fukushima, J. Mol. Struct., 1976,34, 67. 9 A. Allen, V. Fawcett and D. A. Long, J. Raman Spectrosc., 1976, 4, 285. 10 M. S. McKinnon and R. E. Wasylishen, Chem. Phys. Lett., 1986, 130,565. 11 A. E. Aliev and K. D. M. Harris, J. Am. Chem. SOC., 1993, 115, 6369. 12 P. A. Schofield, K. D. M. Harris, I. J. Shannon and A. J. 0. Rennie, J. Chem. SOC., Chem. Commun., 1993,1293. 13 R. K. Harris and A. C. Olivieri, Prog. Nucl. Magn. Reson. Spec- trosc., 1992,24,435. 14 0.A. Subbotin and N. M. Sergeyev, J. Am. Chem. SOC., 1975,97, 1080. 15 H-0. Kalinowski, S. Berger and S. Braun, Carbon-13 NMR Spectroscopy, Wiley, Chichester, 1988. 16 M. W. Anderson and J. Klinowski, J. Am. Chem. SOC., 1990,112, 10. 17 S. Sekine, A. Kubo and H. Sano, Chem. Phys. LRtt., 1990, 171, 155. 18 J. Forbes, C. Husted and E. Oldfield J. Am. Chem. SOC., 1988, 110,1059. 19 J. P. Yesinowski and M. J. Mobley, J. Am. Chem. SOC.,1983, 105, 6191. 20 H. Pfeifer, D. Freude and M. Hunger, Zeolites, 1985,5,274. 21 J. M. Thomas and J. Klinowski, Adu. Catal., 1985,33, 199. 22 A. J. 0. Rennie and K. D. M. Harris, Proc. R. SOC. London, A, 1990,430,615. 23 A. J. 0. Rennie and K. D. M. Harris, J. Chem. Phys., 1992, %, 7117. 24 I. J. Shannon, K. D. M. Harris, A. J. 0. Rennie and M. B. Webster, J. Chem. SOC.,Faraday Trans., 1993,89,2023. Paper 4/0116H;Received 10th January, 1994