English

Bilinear embedding for divergence-form operators with negative potentials

Analysis of PDEs 2026-05-13 v2 Classical Analysis and ODEs Functional Analysis

Abstract

Let ΩRd\Omega \subseteq \mathbb{R}^d be open, AA a complex uniformly strictly accretive d×dd\times d matrix-valued function on Ω\Omega with LL^\infty coefficients, and VV a locally integrable function on Ω\Omega whose negative part is subcritical. We consider the operator L=div(A)+V\mathscr{L} = -\mathrm{div}(A\nabla) + V with mixed boundary conditions on Ω\Omega. We extend the bilinear inequality of Carbonaro and Dragi\v{c}evi\'c [15], originally established for nonnegative potentials, by introducing a novel condition on the coefficients that reduces to standard pp-ellipticity when VV is nonnegative. As a consequence, we show that the solution to the parabolic problem u(t)+Lu(t)=f(t)u'(t) + \mathscr{L} u(t) = f(t) with u(0)=0u(0)=0 has maximal regularity on Lp(Ω)L^p(\Omega), in the same spirit as [13]. Moreover, we study mapping properties of the semigroup generated by L-\mathscr{L} under this new condition, thereby extending classical results for the Schr\"{o}dinger operator Δ+V-\Delta + V on Rd\mathbb{R}^d [8,47].

Keywords

Cite

@article{arxiv.2510.05714,
  title  = {Bilinear embedding for divergence-form operators with negative potentials},
  author = {Andrea Poggio},
  journal= {arXiv preprint arXiv:2510.05714},
  year   = {2026}
}

Comments

59 pages

R2 v1 2026-07-01T06:20:53.182Z