The Redlich Peterson isotherm model is known as: $$ \begin{aligned} q_e=\frac{AC_e}{1+BC^g_e} \end{aligned} $$ Is there any way to obtain the linear form of this equation by algebraic transformation or any ohter ways(like using software origin)?In other words,given that $q_e$ and $C_e$ as variables and A,B,g as parameters,how can I make linear fitting based on this equation?

  • $\begingroup$ What about taking logs on both sides? $\endgroup$
    – AChem
    Feb 14, 2020 at 5:12
  • $\begingroup$ @M. Farooq I am not sure of it.Could you give details? $\endgroup$
    – Chor
    Feb 14, 2020 at 5:27
  • 2
    $\begingroup$ If you are familiar with Origin (which you suggest) you can fit your data directly. To get initial values for the fitting, for A try the value of q when C is small, and A/B when C is large. $\endgroup$
    – porphyrin
    Feb 14, 2020 at 12:05
  • 1
    $\begingroup$ If you want to use a linear form rearrange to $\displaystyle \frac{C}{q}=\frac{1}{A}+\frac{B}{A}C^g$ and a plot $C^g$ vs $C/q$ which has intercept $1/a$ and slope $B/A$. $\endgroup$
    – porphyrin
    Feb 14, 2020 at 16:51
  • 1
    $\begingroup$ Directly fitting the equation, as per the first comment by @porphyrin, is the best way to proceed: it does not alter the noise structure and programs like Origin make this curve fitting trivially simple. The linearization idea is from the days before curve fitting software. Personally, I wish it would fade away. $\endgroup$
    – Ed V
    Feb 14, 2020 at 17:12


Browse other questions tagged or ask your own question.