Postdoctoral research carried out at
The Royal Institution of Great Britain
Davy-Faraday Research Laboratory
(1995-96)

1. Characterisation of silico-alumino-phosphate molecular sieves (SAPO). Use of forcefield simulations intended to understand the catalytic properties of these materials.


AlPO-5 structure

Initially, the structure of these (crystalline) materials can be considered as an AlPO (alumino-phosphate) framework...

Some considerations about AlPOs...

AlPO frameworks contain tetrahedrically coordinated Al and P (TO4 units, where T=Al,P), edged linked forming -T1-O-T2- units in which there is a strict structural alternance between Al & P: that means if T1=Al, then T2=P; or to put it another way the first T environments of Al and P are: P(4Al), and Al(4P). This fact is frequently refered to as 'ordered material', because no Al-O-Al, or P-O-P ever appear, as it should happen in a 'random' distribution for Al & P in the network. This has important consequences in the configurational entropy of the material about which I can tell you privately because I don't want to bother you too much.

Two more things about AlPOs: they don't have cations to compensate the structure (unlike zeolites) and therefore they don't have either Bronsted acid sites. Consequence of that: they are not very reactive in catalysis. Well, they can have Lewis acidity but you have much better materials in this regard. The second thing is: what about the structure of these materials?. Well, they also microporous like zeolites...

Structure of zeolites. Flexibility of TOT (T=Si,Al) angles allow formation of different rings with 4, 5, 6, 8, 10, etc members which gives raise to a virtually infinite different structures with micropores and cavities in the range 4-20 Angstroms, making zeolites microporous materials.



Some AlPOs have the same structures than some zeolites (vg. AlPO-5 & SSZ-24; or SAPO-37 & Y; or SAPO-34 & chabazite), but normally they have unique structures, microporous structures like zeolites but with different pore systems. Therefore they are potentially atractive as acid catalysts if we manage to create Bronsted acidity in the structure...
...but how to do that? (That's one of the topics of my research!)

...end of considerations about AlPOs

P-->Si,H and Al,P-->Si,Si substitutions in the framework. Let's see a scheme ...

Mechanisms 2 and 3 have been investigated with computational techniques (Mechanism 1 is not experimentally observed).


The extent of both substitution mechanisms has a relation with the number and location of the Bronsted acid centers in the material, which are responsible of the use of the SAPO in acid catalysis.
A better understanding of these mechanisms can provide a good insight in the selection of appropiate catalysts for industrial processes, according to their structural and chemical properties.


2. Diffusion of C8 aromatics in a 10 and 12 Membered Rings Zeolite (CIT-1).

Use of forcefield Molecular Dynamics aimed to calculate self-diffusivities, activation energies, and diffusion features: shape selectivity effects caused by the two-channel microporous structure and the different size of the guest molecules.

It is well known that two of the most iportant properties of zeolites are their microporosity and acidity. The microporososity of these crystalline materials is controlled in the synthesis process according to the pore and channel size distribution of the structure. Channels are window openings formed by 4,6,8,10,12, etc T(Si,Al) atoms. The more T atoms are forming the channel the bigger it is. Channels of a given number of T atoms are slightly different in size according to the structure in which they occur: this is due to the different flexibility of the framework in the different cases. In the figure above (AlPO-5 structure) you can see a view across a 12 MR channel.

CIT-1 structure.


The MD simulations have carried out with the general purpose DL_POLY code in its paralell version implemented in a 512 PE CRAY-T3D. The results show how para-xylene and ortho-xylene diffuse through the 12MR and 10MR channels systems in [001] and [110] in CIT-1. It is seen how the slightly bigger ortho isomer can not diffuse through the smaller 10MR channels although some penetration into the channels is appreciated. The penetration into the 10MR channels produces a decrease in the diffusivity of ortho-xylene which blocks itself in these incursions. On the other hand the para isomer diffuses through both 12MR and 10MR channels in CIT-1. In particular, diffusion through the 10MR channel proceeds quicker due to the impossibility to rotate in these narrower 'corridors'. Curiously, the 10MR channels are slowing down the diffusion of ortho-xylene and -at the same time- they are fastening the diffusion of para-xylene. CIT-1 presents new and interesting features for the diffusion processes of C8 aromatics and provides a possible scenario for industrial applications.

Xylene (centre of mass trajectories) diffusing mainly through the 12MR channels of the CIT-1 structure.


Activation energies have also been calculated, the results showing that the para isomer diffuses with very similar activation energies through both channels, and in the case of the ortho-xylene the activation energy for the diffusion through 12MR is a bit higher than in the case of 'para' but still allowing diffusion. In the case of ortho-xylene diffusing through 10MR channels a very high activation energy, more than 100kcal/mol, was obtained showing that no diffusion through these channels is possible.



Para-xylene trajectories in a 2-D projection through the CIT-1 structure. It can be observed in both graphs how there are preferential diffusion in the bigger 12MR channel but there is also some diffusion in the narrower 10MR channel. Similar trajectories obtained for the bigger ortho-xylene isomer show selective diffusion in the 12MR channel only.


This Project was funded by Ministerio de Educacion y Ciencia (Spain)

This postdoctoral project was conducted under Professor C. R. A. Catlow