Some considerations regarding the determination of the activation energy of solid-state reactions from a series of isothermal data