It is well known that in curve fitting of linear and non-linear form of adsorption isotherms on a set of experimental data, different values are achieved for isotherm parameters and R2.
What "mathematical" explanation can be given for this discrepancy?
e.g. Langmuir adsorption isotherm can be written in different forms and for a special set of data there are different R2 as followings:
Non-linear form:
qe=qm.K.Ce/(1+K.Ce) , R2=0.9991
Linear form 1:
1/qe=1/qm+(1/(qm.K).1/Ce , R2=0.9715
Linear form 2:
qe=qm-(1/K).qe/Ce , R2=0.4542
Any references are appreciated.