If we compare these to Tutorial 2 then we can see that theCO* coverage is suppressed and there is more O* in the bottom left ofthe plot. This is what we would expect to happen when we require anadsorbate to have an extra free site to adsorb.
The cross interaction terms are very costly to calculate, since theyrequire many DFT calculations (two per adsorbate per adsorbate, orNadsorbates2). For this reason it is common to use some approximations.The most common approximations are:
First, lets assume that we already know the self-interaction parametersand want to include coverage dependent adsorbate interactions on top ofthe model discussed in Tutorial 2. In order to do this we need toadd the following to the CO_oxidation.mkm setup file:
At creation, an isotherm asks for an adsorbateparameter, a string which pyGAPS looks up in an internal list(pygaps.ADSORBATE_LIST). If any known adsorbate name/alias matches,this connects the isotherm object and the existing adsorbate class. This globallist is populated as import-time with the adsorbates stored in the internaldatabase. The user can also add their own adsorbate to the list, or upload it tothe database for permanent storage.
The Adsorbate class has methods which allow theproperties of the adsorbate to be either calculated using the CoolProp orREFPROP backend or retrieved as a string from the internal dictionary. Theproperties which can be calculated are:
Each method also accepts a bool parameter calculate, True by default. Ifset to False, the property will not be calculated by the thermodynamicbackend. Instead, the value from the properties dictionary will be returned.This is static and supplied by the user, but can be useful for adsorbateswithout a thermodynamic backend.
Adds an adsorbate at a particular coordinate. Adsorbate representedby a Molecule object and is translated to (0, 0, 0) if translate isTrue, or positioned relative to the input adsorbate coordinate iftranslate is False.
Function that generates all adsorption structures for a givenmolecular adsorbate on both surfaces of a slab. This is useful forcalculating surface energy where both surfaces need to be equivalent orif we want to calculate nonpolar systems.
This method constructs the adsorbate site finder from a bulkstructure and a miller index, which allows the surface sites to bedetermined from the difference in bulk and slab coordination, asopposed to the height threshold.
Metal-organic frameworks (MOFs) have a high internal surface area and widely tunable composition(1,2), which make them useful for applications involving adsorption, such as hydrogen, methane or carbon dioxide storage(3-9). The selectivity and uptake capacity of the adsorption process are determined by interactions involving the adsorbates and their porous host materials. But, although the interactions of adsorbate molecules with the internal MOF surface(10-17) and also amongst themselves within individual pores(18-22) have been extensively studied, adsorbate-adsorbate interactions across pore walls have not been explored. Here we show that local strain in the MOF, induced by pore filling, can give rise to collective and long-range adsorbate-adsorbate interactions and the formation of adsorbate superlattices that extend beyond an original MOF unit cell. Specifically, we use in situ small-angle X-ray scattering to track and map the distribution and ordering of adsorbate molecules in five members of the mesoporous MOF-74 series along entire adsorption-desorption isotherms. We find in all cases that the capillary condensation that fills the pores gives rise to the formation of 'extra adsorption domains'-that is, domains spanning several neighbouring pores, which have a higher adsorbate density than non-domain pores. In the case of one MOF, IRMOF-74-V-hex, these domains form a superlattice structure that is difficult to reconcile with the prevailing view of pore-filling as a stochastic process. The visualization of the adsorption process provided by our data, with clear evidence for initial adsorbate aggregation in distinct domains and ordering before an even distribution is finally reached, should help to improve our understanding of this process and may thereby improve our ability to exploit it practically.
We present a summary of theoretical results documenting changes in surface stress and surface phonon frequencies for selected light gas adsorbates on Ni(001) and Cu(001) which, when compared with experimental data, provide critical information on surface geometry and electronic structure. Our calculations of the surface electronic structure are based on density functional theory, using the pseudopotential method, while the evaluation of surface dynamics relies on the density functional perturbation theory. A c(2 x 2) overlayer of C on Ni(001) and of N on Cu(001), causes a large change in the surface stress (approximate to 5 N m(-1)) turning it from tensile to compressive in both cases. We find that while adsorbate induced surface stress change depends on the height at which the adsorbate sits on the surface, it is not a direct measure of the propensity of the substrate to reconstruct. We also consider examples of changes in surface electronic structure, and surface force fields and hence the characteristics of surface phonon dispersion curves, brought about by chemisorption, and compare them with experimental data where available. 041b061a72