I would like to find exemples to show and demonstrate that each of the statements of the definition of:
-measure
$\mu\left(\emptyset \right)=0$
$\mu \left( \bigcup A_n\right)=\sum \mu \left( A_n\right)$
$\mu$ defined from a subset of partition of a given set to $\left[0;+\infty\right]$
are not reduntant.
Edit1: I mean Im looking for "applications" which can fit the finite additivity but not that associates the empty set to zero. Or the opposite.
$\mu(\emptyset)=0$ is not redundant because we can have a trivial "measure" which is identically infinite. It is redundant in that it is equivalent to $\exists x \mu(x)<\infty$ (because then $\mu(x)=\mu(\emptyset)+\mu(x)$ and we can subtract $\mu(x)$ to obtain $\mu(\emptyset)=0$).
Countable additivity is not redundant, either. For example, consider the example of a measure $\mu$ defined for all subsets of $\bf R$: $\mu(A)=\infty$ if $0\in \operatorname {cl} A$, $\mu(A)=0$ otherwise. You can see that it is finitely additive (because closure of a finite union is the union of closures), but not countably additive (because the union of singletons $\{1/n\}$ has $0$ in its closure).
For a finite example, see this question.