The algebraic chromatic splitting conjecture for Noetherian ring spectra (Q1794590): Difference between revisions

In chromatic homotopy theory, it is now understood that for a finite \(p\)-local spectrum \(X\) of type \(n\) there is a homotopy cartesian diagram \[ \begin{tikzcd} X\arrow[r]\arrow[d] & L_{K(n)}X\arrow[d]\\ L_{n-1}X\arrow[r,"\iota_n(X)"] & L_{n-1}L_{K(n)}X \end{tikzcd} \] where the \(L_n\) are Bousfield localization with respect to the \(n\)th Johnson-Wilson theory. Hopkins' celebrated chromatic splitting conjecture asks whether or not the map \(\iota_n\) is split. If true, this conjecture would provide a fundamental local-to-global principle in stable homotopy theory. It has been computationally verified for \(n=0,1,2\). In the case of an \(E_\infty\)-ring \(R\), one can consider the category \(\mathrm{Mod}_R\) of module spectra over \(R\). In [\textit{T. Barthel} et al., J. Pure Appl. Algebra 222, No. 2, 433--463 (2018; Zbl 1384.55008)] the present authors constructed a `local duality context' \((\Gamma_\mathcal{V}, L_{\mathcal{V}}, \Delta_{\mathcal{V}}, \Lambda_{\mathcal{V}})\) which give rise to homotopy Cartesian squares \[ \begin{tikzcd} M\arrow[r]\arrow[d] & \Lambda_{\mathfrak{p}}M\arrow[d]\\ L_{\mathcal{V}(\mathfrak{p})}M\arrow[r,"\iota_\mathfrak{p}(M)"] & L_{\mathcal{V}(\mathfrak{p})}\Lambda_{\mathfrak{p}} M \end{tikzcd} \] The paper under review formulates and proves the analogue of Hopkins' splitting conjecture when \(R\) is a commutative Noetherian ring spectrum. The main theorem of the paper shows that for any pair of adjacent primes (\(\mathfrak{p'\subseteq p}\)) and for \(M\) compact and of type \(\mathfrak{p'}\), the map \(\iota_{\mathfrak{p}}\) is split. The final section of the paper gives examples and applications of this algebraic analogue of the chromatic splitting conjecture. Of particular interest is when \(R = C^*(BG,k)\), the chains on the classifying space of a finite \(p\)-group. In this case, \(\mathrm{Mod}_R\) is equivalent to Krause's category of stable \(kG\)-modules. The proof of the theorem relies on the showing the non-existence of phantom maps in \(\mathrm{Mod}_R\). The authors provide a notion of Brown-Comenetz duality for \(\mathrm{Mod}_R\) and prove several results about it which may be of independent interest. The paper is extremely well written with proofs that are motivated and understandable. This paper raises several new avenues of research and exhibits the usefulness of the growing field of tt-geometry.