Injective Rings

Abstract

The purpose of this article is to determine the injective objects in some complete categories of rings. All rings are assumed to have identities and it is assumed that the homomorphisms preserve these identities. We recall that an object Q in a category is called injective if for every diagram where A′ → A is a monomorphism, there is a map A → Q making the triangle commute. The zero ring belongs to all the categories discussed and it is easy to see that it is an injective object. For the categories of commutative rings, strongly regular and commutative regular rings we show that the zero ring is the only injective by using the fact that an injective object must be a retract of any extension. We include in this section the known results which characterize the injective rings and p-rings. The second part of the paper discusses injectivity with respect to regular monomorphisms. Some necessary categorical background is given and it is then shown that results analagous with those of the first section hold (including the known Boolean and p-ring cases). In an abelian category all monomorphisms are regular, so in the study of the injective objects, for example injective modules, there are not two separate cases.