I think that the only structure that can satisfy this formula is the empty one.
Consider the interpretation $I$ defined on the domain $D=\emptyset$.
We have to interpret the predicate $R$ with a subset $R^I$ of $D$; clearly : $R^I=\emptyset$.
For the RHS, we have that it is equivalent to :
$\lnot \exists x R(x)$.
The formula is true in $I$ because $R^I$ is empty; thus it is true that there are no $a \in R^I$.
Consider now the LHS; it is an open formula.
A "reasonable" semantic condition is that $R(x)$ is true in $I$ if for all $a, a \in R^I$ [i.e. $R$ holds for all objects in the domain].
But trivially, there are no $a$ such that $a \notin R^I$.
Thus we can conclude that also $R(x)$ is true in $I$.
In an empty domain both $\forall x (x=x)$ and $\forall x (x \ne x)$ holds [see this post].
Thus, if we agree on the semantics for open formulae, i.e. that $R(x)$ is true iff $\forall x R(x)$ is, we can consider a simlar formula, with $=$ in place of $R$ :
$x = x \leftrightarrow \forall x (x \ne x)$.