Étale contractible varieties in positive characteristic (Q399382)

In this nice and short paper, the authors discuss a fundamental question in étale homotopy theory. If we think of the étale homotopy type of a scheme as the finest invariant that étale homotopy theory and étale cohomology can provide, we may ask how fine this invariant actually is. In particular, we could ask if there are schemes which are nontrivial but whose étale homotopy type is trivial. The authors give an answer to this question for smooth varieties over an algebraically closed field \(k\) of positive characteristic. In fact, they show that, for smooth varieties over such fields, the étale homotopy type is as fine as we one might wish (or fear, depending on your perspective). More precisely, the main theorem of the paper states that a smooth variety \(U\) over \(k\) is étale contractible if and only if \(U = \mathrm{Spec}\,(k)\) is the point. The situation is very different in characteristic \(0\) where there are nontrivial smooth varieties which are étale contractible. The authors provide the reader with a list of such examples and also mention the situation in motivic homotopy theory. Nevertheless, if we contemplate for a while on the situation for varieties in positive characteristic, we would have expected the result of the paper. For Abhyankar's conjecture which is a theorem by the work of \textit{M. Raynaud} [Invent. Math. 116, No. 1--3, 425--462 (1994; Zbl 0798.14013)] states that the étale fundamental group even of the affine line over \(k\) is highly nontrivial. But even though we expected the result, the paper offers a very nice and informative proof, which reveals some interesting insights to the geometry in positive characteristic. In fact, the authors prove a more general statement and show that for a normal variety \(U\) over an algebraically closed field \(k\) of positive characteristic, the conditions \(H^1(U; \mathbb{F}_p)=0\), \(H^2(U; \mu_{\ell})=0\) for some prime \(\ell \neq p\), and the existence of a big Cartier divisor imply that \(U\) has dimension \(0\).