You have in the background a universal set $E$ and some collection $\mathcal{C}$ of subsets of $E$. The collection $\mathcal{D}$ is the one containing any set whose complement is a member of $\mathcal{C}$ (provided you stay inside the universal set $E$, of course).
Another way to imagine this is to form $\mathcal{D}$ by taking the complement of each set belonging to $\mathcal{C}$ and placing it in $\mathcal{D}$, since $X^{\prime\prime} = X$ for any set $X$.
The collection $\mathcal{D}$ is only related to DeMorgan's laws because they both reference the collection of all set complements of another collection. For example,
$$
\bigcup_{X \in \mathcal{C}} X^\prime
$$
is the same as
$$
\bigcup_{X \in \mathcal{D}} X.
$$
The latter hides the complementation in the indexing set.