Hi Sterlin

The model you choose to use should most closely match the real situation in your material. The different models listed in the document include slit-shaped, cylindrical, and spherical pores. Activated carbons are most often assumed to have slit-shaped pores (although obtaining definite proof of this is difficult or even impossible). So I would have thought that choosing either the cylindrical or spherical pore models would be a bit controversial. For carbons, assuming a cylindrical pore shape would probably make the most sense for nanotubes, or perhaps a templated mesoporous carbon like CMK-5. Similarly, a spherical pore shape may apply to certain templated carbons. It seems rather unlikely that an activated carbon would have cylindrical or spherical pores, although I may be wrong.

NLDFT and QSDFT, meanwhile, assume different levels of surface heterogeneity. NLDFT, I think, assumes a homogeneous surface, while QSDFT, which was developed later, assumes a heterogeneous surface. So, really, it is just a case of which model you think will better represent the case in your material. QSDFT was introduced to try to improve the results of these models, as most materials are unlikely to have energetically homogeneous surfaces. If I were to choose any of these, I would therefore go for slit pore QSDFT for an activated carbon. This doesn't necessarily mean that the results obtained are more accurate, just that this model seems to make the most sense for your material...

There is also more than one reason why you may get different results using different adsorbates. As Mohammed suggests, also using CO2 at 273 K is a good idea, although this probably depends on how narrow you think the pores are. Generally, it's best to combine as much information as you can, but if you know that, for whatever reason, your material is mostly ultramicroporous then trying CO2 adsorption makes more sense. Activated carbons, though (in contrast to carbon molecular sieves), quite often have fairly broad pore size distributions and I think that in that case there may be no great benefit to using CO2 as well; apart from the fact that it is always a good idea to obtain more information rather than less.

With regard to the possibility of measuring non-equilibrium data, you can always try lengthening the equilibration times on repeat measurements to see if this affects the calculated uptakes, and hence the results of subsequent data analysis. If you double the equilibration time, for example, and you see no change in the data, then you can be a bit more confident that you are not analysing non-equilibrium data.

Best regards,

Darren

The physical adsorption of gases is the technique most used for the characterization of porous solids, with various adsorptives such as N2, CO2, and CH4 being employed. N2 at 77 K is the more widely used adsorptive for investigating the microporous characteristic of activated carbons. However, the diffusion of N2 molecules into carbon micropores is a very slow process at low temperatures. Hence, the adsorption of N2 at 77 K over the narrow pore range (pores

Thanks Prof. Thalji for your answer

Hi Sterlin

The model you choose to use should most closely match the real situation in your material. The different models listed in the document include slit-shaped, cylindrical, and spherical pores. Activated carbons are most often assumed to have slit-shaped pores (although obtaining definite proof of this is difficult or even impossible). So I would have thought that choosing either the cylindrical or spherical pore models would be a bit controversial. For carbons, assuming a cylindrical pore shape would probably make the most sense for nanotubes, or perhaps a templated mesoporous carbon like CMK-5. Similarly, a spherical pore shape may apply to certain templated carbons. It seems rather unlikely that an activated carbon would have cylindrical or spherical pores, although I may be wrong.

NLDFT and QSDFT, meanwhile, assume different levels of surface heterogeneity. NLDFT, I think, assumes a homogeneous surface, while QSDFT, which was developed later, assumes a heterogeneous surface. So, really, it is just a case of which model you think will better represent the case in your material. QSDFT was introduced to try to improve the results of these models, as most materials are unlikely to have energetically homogeneous surfaces. If I were to choose any of these, I would therefore go for slit pore QSDFT for an activated carbon. This doesn't necessarily mean that the results obtained are more accurate, just that this model seems to make the most sense for your material...

There is also more than one reason why you may get different results using different adsorbates. As Mohammed suggests, also using CO2 at 273 K is a good idea, although this probably depends on how narrow you think the pores are. Generally, it's best to combine as much information as you can, but if you know that, for whatever reason, your material is mostly ultramicroporous then trying CO2 adsorption makes more sense. Activated carbons, though (in contrast to carbon molecular sieves), quite often have fairly broad pore size distributions and I think that in that case there may be no great benefit to using CO2 as well; apart from the fact that it is always a good idea to obtain more information rather than less.

With regard to the possibility of measuring non-equilibrium data, you can always try lengthening the equilibration times on repeat measurements to see if this affects the calculated uptakes, and hence the results of subsequent data analysis. If you double the equilibration time, for example, and you see no change in the data, then you can be a bit more confident that you are not analysing non-equilibrium data.

Best regards,

Darren