I'm reading Lieb's book "Stability of matter". On page 66 he states that for any arbitrarily small negative $V$, for
$V\in L^{1+\epsilon}(\mathbb{R}^2)+L^{\infty}(\mathbb{R}^2)$ (case $d=2$) OR
$V\in L^{1}(\mathbb{R}^1)+L^{\infty}(\mathbb{R}^1)$ (case $d=1$)
the operator $H=-\Delta+V$ has at least one negative eigenvalue. Any tips how to see this?