Square roots and inverses in e-rings (Q2456618)

E-rings were introduced by the author in [Proc. Am. Math. Soc. 132, No. 12, 3581--3587 (2004; Zbl 1063.47003)] under the name `effect-ordered rings' and further studied by him in a still unpublished paper. An e-ring is a pair \((R,E)\) consisting of an associative ring with unit 1 and a subset \(E \subseteq R\) of elements (called effects) such that (a) \(e \in E \Rightarrow 1 - e \in E\), and (b) the set \(E^+\) of all finite sums of elements of \(E\) satisfies the following conditions for all \(a,b \in E^+\): (i) if \(-a \in E^+\), then \(a =0\), (ii) if \(1-a \in E^+\), then \(a \in E\), (iii) if \(ab = ba\), then \(ab \in E^+\), (iv) \(aba \in E^+\), (v) if \(aba = 0\), then \(ab = 0 =ba\), (vi) \((a - b)^2 \in E^+\). The *-algebra of all bounded linear operators on a Hilbert space \(H\) together with the subset of Hermitian operators on \(H\) provide an example of an e-ring. A more abstract example is a pair \((A,C)\), where \(A\) is a unital \(C^*\)-algebra and \(E := \{aa^* \mid a \in A \;\&\; \exists b \in A\colon aa^* + bb^* = 1\}\). Considering those e-rings for which the effects admit suitable square roots, provides a starting point for the study of so-called sequential products. The author's main purpose here is to deal with existence and uniqueness of square roots of effects and, more generally, positive elements in e-rings. However, the paper contains also a lot of facts concerning the structure of \(E\), of the directed group \(G := E^+ - E^+\), and of the set \(P\) of projections in \(G\). In particular, it is shown that \(P\) is an orthomodular poset, and demonstrated that existence of the ``semi-transparent effect'' (the unique effect \(s\) with \(2s = 1\)) implies that \(G\) acquires many of the features of a Jordan algebra of self-adjoint operators. Also, the question when a positive element of \(G\) has a multiplicative inverse is studied.