The BET 'c' constant (which as we all know is not a constant... it varies from surface to surface) should always be positive, even when analyzing microporous materials.  A negative 'c' value arises from a negative intercept on the BET plot, and (ignoring some instrument problem) is due to using data point at too high P/Po values. The stronger the adsorption, as in microporous materials, the "sooner" the monolayer is formed and the data point range must be adjusted downward. In fact, the monolayer in very small microporous materials can be correlated with the filling of the micropores.

To find the appropriate range, you should employ what is often referred to as the "Rouquerol" method as described in the attached (section 5.2.2). If your instrument software does not have the calculation, you can easily replicate the method by calculating the single point BET value, point by point, say from 0.01 to 0.15. the values will go through a maximum. That maximum represents the highest P/Po value you should use. Simple select that one, plus a few immediately below it, and these are your best BET points.

If, by shifting the P/Po range down, you cannot achieve a positive intercept it means that you either (i) don't have measured points at low enough P/Po (repeat analysis including lower range, start at 0.005 to be sure) or (ii) have an instrument problem (e.g. leak).

Dear Saeed

I know that it is relevant to heat of adsorption; however, just in case for micropore materials it becomes negative since this equation is suitable for mesopore  materials (multilayer adsorption) it is not able to have good prediction for monolayer adsorption.

Dear Saeed

Thank you. However, there shouldn't be any meaningful difference between the linear and non-linear form of an equation in which one of the parameters is positive and in the another one become negative. 

Q1-QL can be positive; but on certain surfaces where gas phase molecules interact like hydrogen bonding etc QL can be greater than 1 in those cases one will observe type III or Type V isotherms.

The c value is related to the heat of adsorption and must always give a positive value. With microporous materials the BET equation will probably be linear in the 0.02 to 0.1 p/p° región. Thus, displaced to lower values tan those for non-porous materials and much narrower. However, there is one added problema in that since the heat of adsorption is stronger in microporous materials the intercept will be much closer to zero and so for almost the same value or Specific Surface Area this intercept can sometimes appear to be negative. Maybe you could avoid this by taking more data points in the narrow relative pressure range that I mentioned above.

Dear Dr. Yates

Thank you very much for your advice.

The BET 'c' constant (which as we all know is not a constant... it varies from surface to surface) should always be positive, even when analyzing microporous materials.  A negative 'c' value arises from a negative intercept on the BET plot, and (ignoring some instrument problem) is due to using data point at too high P/Po values. The stronger the adsorption, as in microporous materials, the "sooner" the monolayer is formed and the data point range must be adjusted downward. In fact, the monolayer in very small microporous materials can be correlated with the filling of the micropores.

To find the appropriate range, you should employ what is often referred to as the "Rouquerol" method as described in the attached (section 5.2.2). If your instrument software does not have the calculation, you can easily replicate the method by calculating the single point BET value, point by point, say from 0.01 to 0.15. the values will go through a maximum. That maximum represents the highest P/Po value you should use. Simple select that one, plus a few immediately below it, and these are your best BET points.

If, by shifting the P/Po range down, you cannot achieve a positive intercept it means that you either (i) don't have measured points at low enough P/Po (repeat analysis including lower range, start at 0.005 to be sure) or (ii) have an instrument problem (e.g. leak).

Dear Dr. Thomas

Thank you for your comprehensive explanation.

It is possible to obtain negative value and it could be accounted for by the BET equation not fitting the experimental data for relative pressure greater than 0.2, perhaps you should try relative pressure lower than 0.2

If the C constant value is more than 1000, does it mean that the surface area is unreliable? What should be the typical C constant value?

The constant in BET equation involves the difference in adsorption heat in the first layer and the condensation heat and this can have a lower limit to zero and this will give rise to approximately C value to one and hence this may be the limit of the value of C

Balasubramanian Viswanathan For one of my measurements I am getting C constant 368 or 405 depending on the number of points I choose between 0.01 to 0.13 (material is microporous), i.e. if I remove the last point close to 0.13. Which C value of these two will give me correct value of the surface area?

for microporous systems you should choose points at low values of relative pressure then the value may be more meaningful

thanks

Balasubramanian Viswanathan Okay. But the question is still, of the two values of C , which one should I report for surface area , lower one or the higher one?

This has to be decided on the basis of the adsorbent since C is related to the difference between the heat of adsorption in the first layer and the condensation heat. It is my simple request you can chose the lower value of 368 since it includes observation at the low relative pressure

thanks

Balasubramanian Viswanathan Thanks for the answer. I will have to check it. I am using activated carbon.