C h apter 9 : Conclusions a n d fu tu re directions
9.1 Conclusions
This thesis has detailed w ork studying the atom ic structure o f the surfaces of three different crystalline systems, siliceous faujasite, zinc oxide and sodium chloride. In each case, new inform ation has been predicted about the m aterials and for the case o f zinc oxide, the inform ation can be used to further understand previous research.
9.2 Conclusions from the siliceous faujasite study
The study o f the 111 surface o f siliceous faujasite has provided several key pieces of inform ation. Firstly, the structure o f the surface term ination has, together with experim ental evidence from other workers, been shown to be able to adopt two particular conform ations, one with “double six ring” units presented on the surface and a related one which is similar, but w ithout the double six ring units. The m icroscopy evidence more often shows the surface with the double six rings present, but it is not clear w hether such features merely m ake that particular surface more resolvable or w hether there is a definite preference. These surfaces both present a tw elve m em bered ring as the surface aperture, suggesting that transport between the exterior and interior o f the crystal is governed by the sam e sized aperture as the supercage-to-supercage w indow s in the bulk structure, albeit with the surface modifications. Secondly, our understanding of the differences betw een the interior and the exterior o f the crystal have been im proved, with the location o f surface adsorption positions for the tw o favoured surfaces and the energetic findings that the surface adsorption energy in the surface regions is low er than that for the interior adsorption positions, although the difference is very sm all w hen the surface adsorption positions are not in close proxim ity to the surface hydroxyl groups. Thirdly, the study o f the energetic barriers for the penetration o f benzene molecules have shown that a transition from a surface adsorption point to an interior adsorption point has an energy barrier o f 18 kJ mol % which is less than the equivalent m ovem ent between two interior surface sites and less than the reverse transition. This finding suggests that the m olecules are likely to becom e incorporated w ithin the zeolite fram ew ork m ore readily than they are to leave it and also suggests that
C hapter 9 : C onclusions an d fu tu re directions
diffusion into the upperm ost supercages in a structure will be subject to different diffusion rates com pared to diffusion in the deeper parts o f the crystal.
9.3 Future directions for the siliceous faujasite study
The findings for the siliceous faujasite are not fully represeentative o f real systems of course, and the logical progression for this w ork is to re-investigate the m odel using a m ore realistic version o f the zeolite fram ew ork to represent the m aterials w hich are used in catalytic environm ents in real processes. Such m aterials will necessarily feature a low er Si:Al ratio and will be likely to include a variety o f cations and an array o f defects. If such a m odel were to be made, it could be usefully used to predict transport rates for various m olecules and also to investigate possible surface m odifications which could prom ote the transport o f reagents through the surface layer of the crystal in order to make a faster catalytic process.
9.4 Conclusions from the zinc oxide study
The study o f the prim ary surfaces o f zinc oxide has involved a process o f m aking a new model for the material which was both m ore flexible and m ore robust than previous m odels. The main point o f creating such a m odel was to allow study of usefully large pieces o f the m aterial to be made, for m odels w here the com putational cost required for a quantum m echanical study w ould be too great. T he resulting param eterisation o f interatom ic potentials gives a useful m odel w hich can be applied to structures with undercoordinated atoms with m ore confidence than any previous model. The application of this new model has been shown to be able to bridge the gap betw een experim ental observations o f m aterial w hich contains defects and quantum m echanical studies o f perfectly coordinated surfaces.
9.5 Future directions for the zinc oxide study
The further uses o f the new m odel are twofold. Firstly, a m ore detailed com parison of the differences between experim ent and quantum m echanical studies can now
C hapter 9 : Conclusions a n d fu tu re directions
defective structures, both in the bulk and at surfaces can now be investigated com putationally for this material. Secondly, the interatom ic potential m odel can now be em ployed in conjunction with quantum m echanical m ethods using an “em bedded cluster” m ethod to investigate the fine detail o f larger features o f surface structures w hich w ould be too large to investigate by quantum m echanical m ethods alone.
9.6 Conclusions from the sodium chloride growth inhibition study
The study o f the docking of the growth inhibitor ion, (Fe°^(CN)6)^', into the 100 surface o f N aCl has shown that previous findings about the m ost energetically favourable way to dock the ion into a surface m ight need re-exam ining. D ocking the m olecule into the surface and arranging the appropriate counter ions to neutralise the charge was shown to give a m ore favourable structure than docking the ion into either a surface site or a step site and using vacant sites to neutralise the overall charge in the model. This finding could challenge the current thoughts about how this particular inhibitor m odifies the growth o f surfaces, but is not in itself a com plete picture.
9.7 Future directions for the sodium chloride growth inhibition study
The further w ork needed to conclude the findings o f the docking and inhibition procedures is to treat all of the possible docking configurations in a m odel which includes a representation o f the solvent in which the real system crystallises from. The w ork w ithout solvent can readily give clues to favourable configurations, but convincing evidence can only be obtained by m odelling the solvent as well as the grow ing crystal.