The results show a good agreement between experimental data
and the
fittings accomplished by two of the three models adopted
in this work: the pseudo-second-order (PSO) model and the
Elovich model. This conclusion was made based on the values of
the coefficient of determination, R2, obtained for the different
models (PFO: 0.885–0.988; PSO: 0.950-0.997; Elovich: 0.964–
0.999). The goodness of the
fits was also compared by the Akaike’s
Information Criterion (AIC) and by the evidence ratio [9], which
confirm that, excepting for one experimental condition, the Elovich
model is likely to be the most appropriated model to describe the
kinetics of Hg2+ removal by the sorbents (lower AIC and higher
evidence ratio), corroborating the initial conclusion achieved from
R2 comparison. In most of the cases the Elovich model was able to
fit the data in the transition between the steep ascendant part of
each curve and the horizontal branch, where the kinetic profile is
usually difficult to
fit or simulate and the pseudo-second order
model was able to estimate accurately the equilibrium qA values
(relative errors lower than 9%).