1 Injective Modules in Module Theory
1.1 Definition via Extension Property
1.1.1 Extension along Monomorphisms
Let \(R\) be a ring and consider left \(R\)-modules. An \(R\)-module \(E\) is injective if it has the following extension property: for every monomorphism \(i\colon A \hookrightarrow B\) of \(R\)-modules and every homomorphism \(f\colon A \to E\), there exists a homomorphism \(g\colon B \to E\) such that \(g\circ i=f\). Intuitively, maps into an injective module behave like “flexible targets”: any partial map defined on a submodule extends across an inclusion.
1.1.2 Equivalent Formulations
The extension property admits several standard reformulations. One common equivalent statement is that \(E\) is injective precisely when \(\operatorname{Hom}_R(-,E)\) turns monomorphisms into epimorphisms, i.e., for each short exact sequence \(0\to A\to B\to C\to 0\), the sequence \[ 0\to \operatorname{Hom}_R(C,E)\to \operatorname{Hom}_R(B,E)\to \operatorname{Hom}_R(A,E)\to 0 \] is exact at the right two terms. Another closely related characterization uses lifting properties with respect to injective envelopes and, more categorically, the notion of injective objects in an abelian category.
1.2 Basic Examples
1.2.1 Injective Modules Over a Field
If \(R\) is a field \(k\), then every \(k\)-vector space is injective. This follows because every monomorphism of vector spaces splits, and an extension across a split inclusion is always possible by defining the extension arbitrarily on a complementary summand.
1.2.2 Baer’s Criterion (Special Cases)
A fundamental test for injectivity over more general rings is Baer’s criterion, which reduces injectivity to checking extensions from ideals. In the most familiar special case (commonly stated for left modules over a ring \(R\)), \(E\) is injective iff every homomorphism \(I\to E\) from any left ideal \(I\subseteq R\) extends to a homomorphism \(R\to E\). This criterion is especially effective because ideals are often manageable subobjects compared with arbitrary submodules.
2 Relationship with Projective Modules
2.1 Duality Perspective
2.1.1 Projective vs. Injective Definitions
Projective modules are characterized by a lifting property with respect to epimorphisms: \(P\) is projective if homomorphisms \(P\to C\) lift along surjections \(B\twoheadrightarrow C\). Injective modules reverse the direction: homomorphisms defined on submodules extend along monomorphisms into the injective module. Thus, injectivity is best understood as a categorical dual notion to projectivity.
2.1.2 Categorical Duals and Opposite Categories
In categorical terms, the duality is realized by passing to opposite categories and reversing morphisms. The functor \(\operatorname{Hom}_R(P,-)\) behaves covariantly in its second argument, while \(\operatorname{Hom}_R(-,E)\) is contravariant in its first argument; injectivity corresponds to exactness properties of the latter.
2.2 Core Properties and Contrasts
2.2.1 Splitting Behavior in Short Exact Sequences
Injectivity yields splitting statements. If \(0\to A\to B\to C\to 0\) is short exact and \(A\) embeds into \(B\) while \(A\) is injective, then the sequence splits: \(B \cong A\oplus C\). Dually, projectivity forces splitting when the quotient is projective. This parallel is central in many structural arguments.
2.2.2 Hom Functor Behavior
A defining practical contrast appears in the behavior of Hom. For an injective module \(E\), the functor \(\operatorname{Hom}_R(-,E)\) is exact on the left and preserves exactness with respect to extensions, aligning with the extension property. For projective modules \(P\), \(\operatorname{Hom}_R(P,-)\) has an analogous favorable behavior with respect to surjections and lifting.
3 Homological Characterizations
3.1 Ext and Injectivity
3.1.1 Vanishing of Ext¹
Injectivity is characterized by the vanishing of certain Ext groups. Specifically, an \(R\)-module \(E\) is injective if and only if \[ \operatorname{Ext}^1_R(-,E)=0 \] as a functor, equivalently \(\operatorname{Ext}^1_R(M,E)=0\) for every \(R\)-module \(M\). At a conceptual level, \(\operatorname{Ext}^1\) measures the obstruction to splitting extensions; injectivity removes such obstructions for all modules.
3.1.2 Injectivity Tests Using Ext
Because \(\operatorname{Ext}^1\) detects extension classes, one can test injectivity by computing or proving vanishing results. In many settings (notably with noetherian conditions or special classes of rings), checking Ext against a smaller collection of modules suffices to conclude injectivity for \(E\). These reductions are frequently implemented using standard homological tools such as long exact sequences and devissage.
3.2 Exactness and Derived Functors
3.2.1 How Injectives Affect Computations
Injective modules allow one to compute derived functors defined via contravariant Hom. When resolving a second argument by injectives, derived functors such as \(\operatorname{Ext}^n_R(M,E)\) can be computed from homology of complexes obtained by applying \(\operatorname{Hom}_R(M,-)\) to an injective resolution of \(E\). This approach is often more efficient than using projective resolutions in the dual direction, depending on the problem.
3.2.2 Connections to Cohomological Functors
In broad terms, injectivity connects to cohomology theories in algebra: the Ext functors behave like cohomological invariants, and injective resolutions provide a systematic mechanism for defining them. The extension property ensures that short exact sequences induce the expected long exact sequences in cohomology.
4 Injective Resolutions and Derived Constructions
4.1 Injective Resolutions
4.1.1 Existence and Construction Outline
In many module categories, there are “enough injectives,” meaning every module admits an injective resolution. Concretely, one constructs a chain complex \[ 0\to M \to I^0 \to I^1 \to I^2 \to \cdots \] where each \(I^n\) is injective and the complex is exact. Existence is guaranteed in standard module categories over rings, often via injective hulls and iterative embedding.
4.1.2 Minimality in Specific Settings
While injective resolutions are not unique, they can sometimes be chosen to satisfy minimality conditions—particularly in special algebraic settings such as certain noetherian or artinian categories. Minimal injective resolutions are valuable because their structure reflects invariants of the module and supports finer computations (for example, in local cohomology contexts).
4.2 Dimension-Shifting
4.2.1 Computing Ext via Resolutions
Dimension-shifting is a technique that relates Ext groups in different degrees using short exact sequences from an injective resolution. If \(0\to M\to I^0\to C^1\to 0\) is the start of an injective resolution, then \(\operatorname{Ext}^{n}_R(M,E)\) can often be identified with \(\operatorname{Ext}^{n-1}_R(C^1,E)\) for \(n\ge 2\). This reduces higher-degree calculations to lower-degree ones, which may be easier to manage.
4.2.2 Consequences for Long Exact Sequences
Dimension-shifting interacts with long exact sequences derived from short exact sequences of modules. By repeatedly “shifting” degrees, one can transform statements about \(\operatorname{Ext}^1\) and splitting into statements about \(\operatorname{Ext}^n\) for larger \(n\), or conversely. This method is frequently used to prove vanishing theorems or to propagate exactness properties across a chain of derived functors.
5 Structural Results and Closure Properties
5.1 Direct Sums, Products, and Summands
5.1.1 Closure Under Direct Sums
Injective modules are closed under direct sums: if \(\{E_\alpha\}\) is a family of injective modules, then \(\bigoplus_\alpha E_\alpha\) is injective. This follows from the extension property and the fact that homomorphisms out of submodules into a direct sum correspond to families of maps into each summand.
5.1.2 Closure Under Direct Products
Injective modules are also closed under direct products: \(\prod_\alpha E_\alpha\) is injective whenever each \(E_\alpha\) is injective. The proof uses the ability to lift maps componentwise and the compatibility of Hom with products in the appropriate variance.
5.2 Essential Submodules and Envelopes
5.2.1 Injective Hulls
An injective hull (or injective envelope) of a module \(M\) is an injective module \(E\) containing \(M\) such that the inclusion is essential: every nonzero submodule of \(E\) intersects \(M\) nontrivially. Injective hulls are unique up to isomorphism and are typically constructed by taking a maximal injective submodule with respect to inclusion.
5.2.2 Existence of Injective Envelopes
In categories with enough injectives, every module has an injective envelope. The existence relies on general arguments using Zorn’s lemma and essentiality, together with injective objects being “large enough” to embed any module. Injective envelopes serve as the building blocks for injective resolutions.
5.3 Behavior Under Functors
5.3.1 Restriction and Extension of Scalars (Overview)
When changing the ring via homomorphisms \(R\to S\), injectivity may be preserved under certain functorial operations. For example, restriction of scalars can preserve injectivity under appropriate conditions, while extension of scalars may not. These behaviors depend on how adjoint functors interact with Hom and on whether the relevant module categories have enough injectives.
5.3.2 Compatibility with Module Homomorphisms
Injectivity is stable under isomorphism and interacts well with homomorphisms: if \(E\) is injective and \(E'\cong E\), then \(E'\) is injective. Moreover, embeddings into injective modules can be refined using injective envelopes, and maps between injective resolutions can be constructed so that the resulting derived constructions are consistent.
6 Injective Modules in Categories Beyond Modules
6.1 Abelian Category Viewpoint
6.1.1 Injective Objects
In an abelian category \(\mathcal{A}\), an object \(I\) is injective if every monomorphism \(A\hookrightarrow B\) and every morphism \(A\to I\) admits a lift \(B\to I\). This reproduces the module definition when \(\mathcal{A}\) is a module category, but it is formulated purely in terms of categorical morphisms.
6.1.2 Enough Injectives
An abelian category has enough injectives if every object embeds into an injective object. Module categories over rings are standard examples with enough injectives, enabling systematic construction of injective resolutions and derived functors.
6.2 Grothendieck Categories and General Theorems
6.2.1 Existence of Injective Resolutions
Grothendieck categories—abelian categories satisfying certain completeness and exactness properties—are known to have enough injectives, and every object admits an injective resolution. This broadens the applicability of Ext, derived functors, and cohomological methods far beyond classical module categories.
6.2.2 Applications to Derived Categories (Conceptual)
In derived-category language, injective resolutions provide models for derived functors and cohomology objects. While the derived category can be defined abstractly, injective resolutions supply computational tools: they let one replace objects by injective complexes without changing derived invariants.
7 Special Classes and Notable Theorems
7.1 Enochs–Jenda Theory (If Applicable)
7.1.1 Cotorsion Pairs and Injectives
A broader framework uses cotorsion pairs, which organize modules into two classes that are orthogonal with respect to \(\operatorname{Ext}^1\). In such settings, injective modules form one cornerstone class (or appear as trivial objects within it), and relative homological algebra studies modules that behave injectively only with respect to a chosen class.
1.1.2 Relative Injectivity
Relative injectivity generalizes absolute injectivity by restricting the extension property to certain morphisms or certain tests. This is useful in contexts where one works with a specified subcategory (for example, modules of a particular size or satisfying a given property), producing refined invariants and more targeted resolutions.
7.2 Baer Criterion (Ring-Theoretic Criteria)
7.2.1 Conditions for Module Injectivity
Baer-type criteria reduce injectivity to extension along maps from specific submodules, most often left ideals of the ring. Depending on the ring and module side conventions, variants exist that test injectivity via homomorphisms from particular “building” subobjects.
7.2.2 Typical Use in Computations
In practice, Baer’s criterion is applied by verifying that every map from a relevant ideal into \(E\) extends to the whole ring. Because ideals are simpler than arbitrary submodules, this gives a streamlined method to prove injectivity for explicitly described modules, such as those constructed from ring-theoretic data.