Sa. Sande et J. Karlsen, EVALUATION OF REACTION ORDER - SOFTWARE IN PHARMACEUTICS .3., International journal of pharmaceutics, 98(1-3), 1993, pp. 209-218
A computer program for the simulation of experiments intended to deter
mine the order of a reaction is presented along with a discussion and
evaluation of the important parameters for the reaction order analysis
. The source code is written in C. The program calculates the maximum
acceptable amount of reactant remaining at the last sampling point for
the estimation of the correct reaction order with a level of signific
ance of 0.95. Based upon the simulations, the precision of the analyti
cal method was shown to be more important than the number of samples i
n determining the correct order of the reaction at an early stage. The
simulations also showed that even with good analytical methods (SD =
1.0%). the reaction had to be followed for about one half-life (t1/2)
of the reactant. A sampling scheme with equal times between sampling w
as found to be as good as a scheme with equal differences between the
measured concentrations. This is due to the lack of correlation betwee
n the rate of reaction and differences between the course of reaction
predicted by the two reaction orders. When the experimental errors are
expected to be normally distributed, the use of integer orders is pre
ferred over a model employing decimal orders due to the increase in sp
eed of the simulations.