On K-absolutely pure complexes (Q6142851)

Resolutions of unbounded complexes were investigated by \textit{N. Spaltenstein} [Compos. Math. 65, No. 2, 121--154 (1988; Zbl 0636.18006)] where the classes of K-projective, K-injective and K-flat complexes over an ambient ring \(R\) were introduced, denoted by \(\mathbf{K}\mathrm{-Proj}\), \(\mathbf{K}\mathrm{-Inj}\) and \(\mathbf{K}\mathrm{-Flat}\) respectively. It follows from Spaltenstein's classic results that there exist two Bousfield localizing pairs (equivalently, stable t-structures) \((\mathbf{K}\mathrm{-proj},\mathbf{K}_{\mathrm{ac}}(R))\) and \((\mathbf{K}_{\mathrm{ac}}(R),\mathbf{K}\mathrm{-inj})\) in the homotopy category of complexes \(\mathbf{K}(R)\). Neeman in his highly influential work [\textit{A. Neeman}, Invent. Math. 174, No. 2, 255--308 (2008; Zbl 1184.18008)] proved the existence of a Bousfield localizing pair \((\mathbf{K}(\mathrm{Proj}R),\mathbf{K}_{\mathrm{ac}}(\mathrm{Flat} R))\) in the homotopy category \(\mathbf{K}(\mathrm{Flat}R)\) (here the categories involved are homotopy categories of degreewise projective or flat complexes, and should not be confused with K-projective and K-Flat complexes). It seems that Neeman's motivation had its origins in algebro-geometric contexts, however, his work led to interesting advances in fields such as ring theory and representation theory. The reason is the close connection between flatness and the concept of purity. Indeed, from a relative homological algebra perspective, Neeman's localizing pair takes the form \((\mathbf{K}(\mathrm{PProj}R),\mathbf{K}_{\mathrm{pac}}(R))\) in the homotopy category \(\mathbf{K}(R)\). Here on the left hand side is the homotopy category of degreewise \textit{pure} projective modules while the right hand side denotes the homotopy category of \textit{pure} acyclic complexes. In addition, there is a dual Bousfield localizing pair given by \((\mathbf{K}_{\mathrm{pac}}(R),\mathbf{K}(\mathrm{PInj}R))\) in the homotopy category \(\mathbf{K}(R)\). For proofs of these results see [\textit{J. Stovicek}, ``On purity and applications to coderived and singularity categories'', Preprint, \url{arXiv:1412.1615}]. In the same spirit it is proved in [\textit{I. Emmanouil}, Isr. J. Math. 255, No. 1, 201--230 (2023; Zbl 07720417)] that \((\mathbf{K}\mathrm{-Flat},\mathbf{K}_{\mathrm{ac}}(\mathrm{PInj}))\) is a Bousfield localizing pair in the homotopy category \(\mathbf{K}(R)\). This is interesting and new since it involves Spaltenstein's K-flat complexes on the left hand side of the pair. On the right hand side we find the homotopy category of acyclic complexes with pure injective components. A main result of the paper under review is the dual of this, namely, the authors prove the existence of a Bousfield localizing pair \[(\mathbf{K}_{\mathrm{ac}}(\mathrm{PProj}),\mathbf{K}\mathrm{-Abspure})\] in the homotopy category \(\mathbf{K}(R)\). The class on the right hand side of this pair is what the authors (formally) call the homotopy category of \textit{K-absolutely pure complexes}. Absolutely pure (a.k.a. \textit{fp-injective}) modules play a role dual to that of flat modules and the above result wishes to take this analogy further to the level of homotopy categories. The authors also prove various facts on K-absolutely pure complexes such as the following. Theorem 3.5: A complex is K-absolutely pure if it is isomorphic in the pure derived category with a K-injective complex. Corollary 4.4: Any left-bounded complex of strongly fp-injective modules is K-absolutely pure.