I am doing single and binary adsorption experiments (adsorption of liquids). I was wondering how to fit binary adsorption data to non-modified Langmuir isotherm model? Also how to fit binary adsorption data to the different multi-adsorption isotherm models in general?
For single adsorption systems, it was easily to model the data following non-linear method using Solver function in Microsoft Excel. But I couldn't model my data for the binary systems.
My question in other words: how to model binary adsorption data and calculate R^2?
Muzafar Khalifa, Usually we fit an equation/function to the data and not the other way around. This is a typical mistake, mainly in student work. I once wrote in a review that this is the way to a perfect fit. Regarding your question, you can use for example Statistica, and instead of R2 value it is better to determine ARE or both measures.
Thank you Mirosław. I am using Excel to calculate R^2. However, for single adsorption systems the idea is clear, while it is not in the multiple adsorption systems. In binary adsorption systems there are two equations, one for each adsorbates, which have common parameters.
If I follow similar steps of single adsorption in modeling binary adsorption systems then I have to follow the same steps for each equations in the binary adsorption systems, which doesn't give write answers and different values for the parameters!!
What steps should I follow to model experimental data of binary adsorption systems?
Notes:
- I am using non-linear method.
- I am using SSE to calculate the error.
- I am using Solver function in Excel to maximize R^2.
Muzafar Khalifa, In a word, R2 is your objective function which you maximize using the internal Excel optimization procedure. There is no problem to find R2 for each of the two approximating functions separately and then, for example, calculate the mean. Instead of R2, you can use SSE or ARE as a measure of fit and look for the minimum instead of maximum.