This is a question about (possible unbounded) operators. We know that $\mathcal{D}(T^*)=\{0\}$ iff $\mathcal{G}(T)$ is dense in $\mathcal{H}\times\mathcal{H}$, where $\mathcal{H}$ is a separable Hilbert space and $\mathcal{G}(T)$ the graph of an operator $T$ with domain $\mathcal{D}(T)$. My task is to show that this equivalence really can happen by contsruction of an operator $T$.
I have done this so far: $\mathcal{H}$ is separable thus there exists an orthonormal countable base $(e_n)_{n\in\mathbb{N}}$. We can also find $(x_n)_n$ an countable dense subset of $\mathcal{H}$ by assumption. Define $T(e_n)=x_n$ and extend this by linearity to a domain $\mathcal{D}(T)$ which is defined to be the set of all finite linear combinations of $(e_n)_n$. I can show the following: $\mathcal{D}(T)$ is dense in $\mathcal{H}$. Thus we can apply the theory of unbounded operators to conclude the existence of the adjoint $T^*$ on $\mathcal{D}(T^*)$. But can we know conclude from out this directly that $\mathcal{G}(T)$ is dense in $\mathcal{H}$, that $\mathcal{D}(T^*)=\{0\}$ and thet $T^*$ is the zero-operator (the last to facts are my main problem)? Can we also conclude the other direction from out this?
I would be happy about help and solutions $:)$
Thank you