A note on Sobolev-Lorentz Capacity and Hausdorff measure
Daniel Campbell
TL;DR
This work gives an elementary, potential-theory-free proof that $\gamma_{p,1}(E)=0$ implies $\mathcal{H}^{n-p}(E)=0$ for $p\in(1,n]$, by leveraging Orlicz-Sobolev methods via MSZ results. It constructs a decomposed energy scheme leading to an Orlicz control that forces the $n-p$ Hausdorff measure to vanish, and it provides Cantor-type constructions showing that the bound is sharp, with $\dim_{\mathcal{H}}(E)=n-p$ yet $\mathcal{H}^{n-p}(E)=0$. The paper also demonstrates sharpness in the opposite direction: the result does not extend to $q>1$, through explicit counterexamples. Together, these results clarify the precise size properties of $p,1$-Sobolev-Lorentz null-sets and their relation to Hausdorff dimension, offering a more accessible route to key capacity–measure connections.
Abstract
In this paper we give an elementary proof that sets of zero $p,1$-Sobolev-Lorentz capacity are $\mathcal{H}^{n-p}$-null sets independently of non-linear potential theory. We further show that there exists a set of Sobolev-Lorentz-$(p,1)$ capacity equal zero with Hausdorff dimension equal $n-p$.
