Minimal representation-infinite coil algebras (Q910473)

There is the vague consense that the infinite radical \(rad^{\infty}(mod A)\) of the module category of a finite dimensional algebra A should behave ``well'' if A is tame and conversely. Along this line the authors consider cycles in mod A that is sequences of non-isomorphisms \[ (*)\quad M_ 0\to^{f_ 0}M_ 1\to^{f_ 1}...\to M_ m\to^{f_ m}M_ 0 \] between indecomposables and they call an algebra A cycle- finite, if for any cycle none of the maps \(f_ i\) is in \(rad^{\infty}(mod A)\). By Proposition 1.4 cycle-finite algebras are tame. If for any cycle (*) in mod A the modules \(M_ i\) lie in one standard coil \(\Gamma\) of the Auslander Reiten quiver \(\Gamma\) (A) of A, then A is said to be a coil algebra. Here a connected translation quiver \(\Gamma\) is a coil if (1) the homotopy group \(\pi_ 1(| \Gamma |)\) is infinite cyclic, (2) if x is in \(\Gamma_ 0\) not projective-injective, then x has at most 2 immediate predecessors (or successors) which are not projective-injective; (3) \(\Gamma\) contains a full, cofinite translation subquiver \(\Gamma '\) which is a quasi-tube such that every point \(x\in \Gamma \setminus \Gamma '\) is directing. The authors show that a full convex subcategory B of a coil algebra A is again a coil algebra, provided B has sincere tubes (Theorem 2.3, Corollary 2.4). So it is natural to ask for a characterization of minimal representation infinite coil algebras. This task is done in the main theorem of the paper: Theorem: Let A be a basic and connected finite dimensional algebra over an algebraically closed field. The following conditions are equivalent: (i) A is a tame concealed algebra. (ii) A is a representation-infinite coil algebra, and, for every non-zero idempotent e of A the quotient A/AeA is representation-finite. (iii) A is a representation-infinite coil algebra, and every proper full convex subcategory of A is representation- finite.