Abstract
In this paper, we investigate the notions of -projective, -injective, and -flat modules and give some characterizations of these modules, where is a class of left modules. We prove that the class of all -projective modules is Kaplansky. Further, if the class of all -injective -modules is contained in the class of all pure projective modules, we show the existence of -projective covers and -injective envelopes over a -hereditary ring. Further, we show that a ring is Noetherian if and only if -injective -modules coincide with the injective -modules. Finally, we prove that if every module has a -injective precover over a coherent ring, where is the class of all pure projective -modules and is the class of all -modules.
1. Introduction
Throughout this paper, denotes an associative ring with identity and all -modules, if not specified otherwise, are left -modules. - denotes the category of left -modules.
The notion of -injective modules over arbitrary rings was first introduced by Stenström in [1]. An -module is called -injective (absolutely pure) if for all finitely presented -modules . Let be a class of left -modules. Mao and Ding in [2] introduced the concept of -injective modules (see Definition 6). Selvaraj et al. developed -injective and -flat -modules and studied covers and envelopes of modules with Goresntein properties in [3–5].
A pair is a cotorsion theory (see Definition 4). In our article, is a cotorsion theory generated by the class [6], that is, and If is a hereditary ring (that is, every ideal is projective), then is closed under submodules and containing all injective -modules in the cotorsion theory . In this case, results are trivial. For this reason, we introduced -hereditary ring (see Definition 16), and we restrict the setting to -hereditary rings.
The notions of (pre)covers and (pre)envelopes of modules were introduced by Enochs in [7] and, independently, by Auslander and Smalø in [8]. Since then, the existence and the properties of (pre)covers and (pre)envelopes relative to certain submodule categories have been studied widely. The theory of (pre)covers and (pre)envelopes, which play an important role in homological algebra and representation theory of algebras, is now one of the main research topics in relative homological algebra.
Salce introduced the notion of a cotorsion theory in [9]. Enochs showed the important fact that closed and complete cotorsion pairs provide minimal versions of covers and envelopes. Eklof and Trlifaj [10] proved that a cotorsion pair is complete when it is cogenerated by a set. Consequently, many classical cotorsion pairs are complete. In this way, Bican et al. [11] showed that every module has a flat cover over an arbitrary ring. These results motivate us to define -projective -modules (see Definition 14) and -pure projective -modules (see Definition 21), and we prove the existence of -projective cover and -injective envelope. In particular, we denote by the class of all pure injective -modules, and we prove the following main result.
Theorem 1. Let be a -hereditary ring and . Then, every -module has a -projective cover and an -injective envelope.
Self-injective rings were introduced by Johnson and Wong in [12]. A ring is said to be self injective if over itself is an injective -module. In this paper, we introduce self--injective ring (see Definition 27). Recall that an -module is called reduced [13] if it has no nonzero injective submodules. An -module is said to be coreduced [14] if it has no nonzero projective quotient modules. Mao and Ding [15] proved that an -injective -module decomposes into an injective and a reduced -injective -module over a coherent ring. Similarly, we prove the following result.
Theorem 2. Let be a self -injective and -hereditary ring and . Then, an -module is -projective if and only if is a direct sum of a projective -module and a coreduced -projective -module.
Pinzon [16] proved that every module has an -injective cover over a coherent ring. We prove the following result that provides the existence of -injective cover, where is the class of all pure projective -modules and the class of all --modules (see Definition 41).
Theorem 3. Let be a coherent ring and Then, every -module has a -injective cover.
This paper is organized as follows: In Section 2, we recall some notions that are necessary for our proofs of the main results of this paper.
In Section 3, we investigate the notions of -injective and -flat modules and give some characterizations of these classes of modules.
In Section 4, we introduce -hereditary ring. Further, we investigate -projective module and give some characterizations. Further, we prove Theorems 1 and 2.
In Section 5, we prove that if is a submodule of an -injective -module , then, is a special -injective envelope of if and only if is an -projective essential extension of
In the last section, we assume that is the class of all pure projective -modules and we prove that every -module has a -injective preenvelope. Finally, we prove the main theorem of this section Theorem 3.
2. Preliminaries
In this section, we recall some known definitions and some terminology that will be used in the rest of the paper.
Given a class of left -modules, we write
Following [7], we say that a map with is a -precover of , if the group homomorphism is surjective for each . A -precover of is called a -cover of if is right minimal, that is, if implies that is an automorphism for each . is a precovering class (resp., covering class) provided that each module has a -precover (resp., -cover). Dually, we have the definition of -preenvelope (resp., -envelope).
A -precover of is said to be special [6] if is an epimorphism and .
A -preenvelope of is said to be special [6] if is a monomorphism and .
A -envelope is said to have the unique mapping property [17] if for any homomorphism with , there is a unique homomorphism such that
A module is said to be pure projective [18] if it is projective with respect to pure exact sequence.
An -module is called super finitely presented ([19]) if there exist a projective resolution with each projective -module is finitely generated.
Following [20, 21], an -module is called weak injective if for every super finitely presented -module . A right -module is called weak flat if for every super finitely presented -module .
A class of left -modules is said to be injectively resolving [6] if contains all injective modules and if given an exact sequence of left -modules
whenever
Definition 4. A pair of classes of modules is called a cotorsion theory [6] if and (1)A cotorsion theory is said to be perfect [22] if every module has an -cover and a -envelope
Definition 5. A cotorsion theory is said to have enough injectives [13] if for every -module there is an exact sequence with and
For an -module , denotes the flat dimension of and denotes the injective dimension of
The -coresolution dimension of , denoted by , is defined to be the smallest nonnegative integer such that for all -modules (if no such exists, set ), and is defined as
We denote by the ring of all integers and by the field of all rational numbers. For a left -module , we denote by the character module of denotes the class of all injective left -modules.
For unexplained terminology, we refer to [23, 24].
3. -Injective and -Flat Modules
In this section, we assume that is any class in We begin with the following definition:
Definition 6 (see [2]). A left -module is called -injective if for all left -modules A right -module is said to be -flat if for all left -modules
We denote by the class of all -injective modules and the class of all -projective -modules and, further, the class of all -flat -modules.
Example 7. (1)If then, is the class of all injective -modules and is the class of all flat right -modules.(2)If is the class of all finitely presented -modules, then, is the class of all -injective -modules.(3)If is the class of all left super finitely presented -module, then, is the class of all weak injective -modules and is the class of all weak flat right -modules.
Proposition 8. Let be an -module. Then, the following are true: (1)Let Then, is injective if and only if is -injective and (2) is injective if and only if is -injective and
Proof. (1)The direct implication is clear. Conversely, let be -injective and . For any -module , consider an exact sequence with an injective envelope of . We have an exact sequenceSince , , and hence, is injective. (2)The direct implication is clear by the definition of an injective module. Conversely, let be -injective and . Consider an exact sequence with an injective envelope of For any -module we have an exact sequence . Since , , and hence, is -injective. Therefore, , so that the exact sequence is split. It follows that is a direct summand of as desired.
We now give some of the characterizations of -injective module:
Proposition 9. Let The following are equivalent for a left -module (1) is -injective(2)For every exact sequence , with is an -precover of (3) is a kernel of an -precover with an injective module(4) is injective with respect to every exact sequence , with
Proof. . Consider an exact sequence
where Then, by hypothesis, is surjective for all left -modules as desired.
. Let be an injective hull of and consider the exact sequence Since is injective, it belongs to So assertions holds.
. Let be a kernel of an -precover with an injective module. Then, we have an exact sequence . Therefore, for any left -module the sequence is exact. By hypothesis, is surjective. Thus, , and hence (1) follows
. Consider an exact sequence
where Then, is surjective, as desired.
. For each left -module there exists a short exact sequence with a projective -module, which induces an exact sequence . By hypothesis, is exact. Thus, , and hence (1) follows.
The following note is useful for understanding the notations in Examples 11 and 15.
Note 10. From Introduction, We get a new cotorsion theory generated by the class
Example 11. Let be a commutative Noetherian and complete local domain. Assume that the and . Then, is an -injective -module, where is the class of all -injective -modules.
Proof. Consider the residue field and an exact sequence . If is an -module, the sequence is exact. By ([25], p 43), is an injective cover of Since then, the class of all injective -modules and the class of all -injective -modules are equal, that is, . Clearly, is an -injective cover of . Thus, is surjective for every -injective -module . We get for every -injective -module , and hence, is -injective. On the other hand, is injective and so -injective. Therefore, is -injective.
We now give some characterizations of -flat module:
Proposition 12. The following are equivalent for a right -module : (1) is -flat(2)For every exact sequence with the functor preserves the exactness(3) for all (4) is -injective
Proof. . Consider an exact sequence with Since is -flat, . Hence, the functor preserves the exactness.
. Let Then, there exists a short sequence with a projective module, which induces an exact sequence . By hypothesis, . Thus, is -flat.
It follows from the natural isomorphism ([26], p 34) .
It follows from the natural isomorphism ([27], VI 5.1) .
Proposition 13. Let be a coherent ring. Then, a right -module is flat if and only if is -flat and .
Proof. The “only if” part is trivial. Conversely, suppose that is -flat. By Proposition 12, is -injective. By ([28], Theorem 2.1), . Then since . By Proposition 8, is injective and hence is flat.
4. -Projective Cover and -Injective Envelope
Now, we introduce -projective modules.
Definition 14. An -module is called -projective if for all -modules .
Example 15. Let be a complete local ring and be the class of all -injective -modules. Assume that the . Then is -projective, where is the residue field and is an -injective envelope of .
Proof. By Note 10, we get a cotorsion theory generated by the class Consider an exact sequence where is an injective envelope of Since then, the class of all injective -modules and the class of all -injective -modules coincide, that is, . It follows that is an -injective envelope of Since the class of all -projective modules is closed under extensions, then by Lemma 2.1.2 in [25], is -projective.
We now introduce Definition 16.
Definition 16. A ring is called left -hereditary if every left ideal of is -projective.
Remark 17. If is the class of injective left -modules, then, every ring is -hereditary. It is also easy to see that a ring is left hereditary if and only if is -hereditary for every class of left -modules.
Given a class of left -modules, we denote by the class
Example 18. Let be a commutative Noetherian ring. If then is a -hereditary ring.
Proof. Let be an ideal of . We claim that is -projective. By hypothesis, for all and for all It follows that Thus, for all ideals of and for all Consider an exact sequence So for all Hence, is -projective, as desired.
Note that an -module is -projective if for all -modules . Clearly, is a cotorsion theory.
Proposition 19. A ring is left -hereditary if and only if every submodule of a -projective left -module is -projective.
Proof. Let be left -hereditary and be a left ideal of . Then, there is an exact sequence By hypothesis, for any -injective left -module Thus, Let be a submodule of an -projective left -module Then, for any -injective left -module , the sequence
is exact. Thus, since is -projective and The reverse implication is clear.
In general, is not closed under pure submodules; for example, if is the class of flat modules, then and this class is not closed under pure submodules in general. As an easy consequence of Proposition 19, we have that the class is closed under pure submodules over a -hereditary ring.
Definition 20 (see [22]). Let be a class of -modules. Then, is said to be Kaplansky class if there exists a cardinal such that for every and for each , there exists a submodule of such that and and
Definition 21. An -module is called -pure injective if extends to a homomorphism for all where (that is, is a pure submodule of ). We denote by the class of all -pure projective modules.
Remark 22. The class of all pure injective -modules is contained in the class of all -pure injective -modules.
Example 23. (1)Clearly, is a cotorsion theory. Then, the class of all -pure injective -modules is the class of all pure injective -modules.(2)If is the class of all flat -modules and is the class of all cotorsion -modules, then, is a cotorsion theory. Then, the class of all -pure injective -modules is the class of all cotorsion -modules.If is Noetherian and then, the class of all -pure injective -modules is
Proof. (1)Straightforward(2)Let be a -pure injective -module. We show that is a cotorsion -module. For any flat -module , there is a pure exact sequence where is a projective -module. Then, we get the following exact sequence:It follows that because is surjective. Hence, is cotorsion. Conversely, let be a cotorsion -module. Suppose the following sequence: where is pure exact. Then, we get the exact sequence It follows that since is flat. Hence , is -pure injective. (3)By Exercises 6 in [13], is a cotorsion theory. Let be an -pure injective -module. We show that For any -module consider the following exact sequence:where is projective. By hypothesis, is injective. Since the cotorsion theory has enough injectives, is injective, that is It follows that the above sequence is pure exact. Then, we get the exact sequence Thus, since is -pure injective. Hence, Conversely, Let be an -injective -module. Suppose the following exact sequence: where is pure exact. It follows that since the cotrosion theory has enough injectives. Then, is surjective. Hence, is -pure injective.
Proposition 24. Let be a -hereditary ring and . Then, is a Kaplansky class.
Proof. Let and . Consider the inclusion and we get by Lemma 5.3.12 in [13], a cardinal and a pure submodule such that and . We get the pure exact sequence By Proposition 19, is -projective. It follows that for any -injective -module This implies that is surjective by hypothesis. Thus, Hence, we proved the proposition.
Theorem 25. Let be a -hereditary ring and . If the class of all -projective -modules is closed under direct limits, then, every -module has a -projective cover and an -injective envelope.
Proof. By Proposition 24, is a Kaplansky class. Since all projective modules are -projective, contains the projective modules. Clearly, is closed under extensions. By hypothesis, is closed under direct limits. Then by Theorem 2.9 in [22], is a perfect cotorsion theory. Hence, by Definition 4, every module has a -cover and a -envelope.
Note that if then the class of all -projective -modules is closed under direct limits by Lemma 3.3.4 in [6]. Then, by Theorem 25 and Remark 22, we have the following.
Theorem 26. Let be a -hereditary ring and . Then, every -module has a -projective cover and an -injective envelope.
Now we introduce a self--injective ring.
Definition 27. A ring is said to be self -injective if over itself is an -injective module.
Example 28. (1)If the class is then is a self injective ring(2)If is the class of all finitely presented -modules, then is a self -injective ring(3)If is the class of all flat -modules, then is a cotorsion ring
We now give some characterizations of -projective module:
Proposition 29. Let be a self -injective ring, and let be an -module. Then, the following conditions are equivalent: (1) is -projective(2) is projective with respect to every exact sequence , with an -injective -module(3)For every exact sequence , where is -injective, is an -injective preenvelope of (4) is cokernel of an -injective preenvelope with is a projective -module
Proof. . Let be an exact sequence, where is -injective. Then, by hypothesis, is surjective.
. Let be an -injective -module. Then, there is a sequence with an injective envelope of . By (2), is surjective. Thus, , as desired.
. Clearly, for all -injective . Hence, we have an exact sequence where is projective.
. Consider an exact sequence , where projective. Since is self--injective, every projective module is -injective. Hence, is -injective. Then, by hypothesis, is an -injective preenvelope.
. By hypothesis, there is an exact sequence , where is an -injective preenvelope with projective. It gives rise to the exactness of for each -injective -module . Since is self -injective, is surjective. Hence, , as desired.
Proposition 30. Let be a self -injective and -hereditary ring and . Then the following are equivalent for an -module : (1) is coreduced -projective(2) is a cokernel of an -injective envelope with is a projective -module
Proof. . Consider an exact sequence with a projective module. Since is self -injective, is -injective. By Theorem 26, the natural map is an -injective preenvelope of . By hypothesis, has an -injective envelope . Then, there exist and such that and . Hence, . It follows that is an isomorphism, . Note that , and so is exact. But is coreduced and , and hence, , that is, So is an isomorphism, and hence, is an -injective envelope of .
. By Proposition 29, is -projective and is coreduced by Lemma 3.7 in [17].
We are now to prove the main result of this section.
Theorem 31. Let be a self -injective and -hereditary ring and . Then, an -module is -projective if and only if is a direct sum of a projective -module and a coreduced -projective -module.
Proof. The “if” part is clear.
“Only if” part. Let be a -projective -module. By Proposition 29, we have an exact sequence with a projective module, where is an -injective preenvelope of . By Theorem 26, has an -injective envelope with an -injective -module. Then, we have the following diagram (Figure 1):
Note that is an isomorphism, and so . Since , and are projective. Therefore, is a coreduced -projective module by Proposition 30. By the Five Lemma, is an isomorphism. Hence, we have , where . In addition, we get the commutative diagram (Figure 2).
Hence, .
5. Some Relation between -Projective and -Injective Modules
In this section, we deals with -injective envelope of a module and -projective module for any class in .
Theorem 32. Let be an -injective envelope. Then, is -projective, and hence, is -projective whenever is -projective.
Proof. It follows from Lemma 2.1.2 in [25].
Theorem 33. Let be a minimal generator of all -projective extensions of . Then, is an -injective envelope of .
Proof. It follows from Theorem 2.2.1 in [25].
Let be a submodule of a module . Then, is called a -projective extension of a submodule if is -projective.
Recall that among all -projective extensions of , we call one of them a generator for (or a generator for all -projective extensions of ) if for any -projective extension of , then there is a diagram (Figure 3).
Furthermore, a generator is called minimal if for all the vertical maps are isomorphisms whenever are replaced by , , respectively.
Theorem 34. Let be a -hereditary ring and Then, for an -module , there must be a minimal generator whenever has a generator.
Proof. It follows from Theorem 2.2.2 in [25].
Theorem 35. Suppose that an -module has an -injective envelope. Let be a submodule of an -injective -module . Then, the following are equivalent: (1) is a special -injective envelope is -projective, and there are no direct summands of with and (2) is -projective, and for any epimorphism such that is split, , where is the canonical map(3) is -projective, and any endomorphism of such that is a monomorphism(4) is -projective, and there is no nonzero submodule of such that and is -projective
Proof. . It follows from Corollary ~1.2.3 in [25] and Theorem 32.
. Since is split, there is a monomorphism such that . Note that , and so by (2). Thus, , and hence, .
. If with . Let be a canonical projection. Then, there is an epimorphism such that . Thus, by hypothsis, and hence, , as required.
. By Wakamatsu’s Lemma ([13], Proposition ~7.2.4), is -projective. Since and is monomorphism, is monomorphism.
. Since is -projective, is a special -injective preenvelope. Let be an -injective envelope of . Then, there exist and such that and . Hence and . Thus, is an isomorphism, and so is epic. In addition, by (4), is monic, and hence, is monic. Therefore, is an isomorphism, and hence, is an -injective envelope of .
. It is obvious that is -projective. Suppose there is a nonzero submodule such that and is -projective. Let be a canonical map. Since is -projective and is -injective, there is a in the following diagram (Figure 4)
Hence, . Note that is an envelope, and so is an isomorphism, whence is an isomorphism. But this is impossible since .
. Let be an -injective envelope of . Since is -projective, is a special -injective preenvelope. Thus, we have the following diagram shown in Figure 5.
That is, . So . Note that is an -injective envelope, and hence, is an isomorphism. Without loss of generality, we may assume . Write . It is clear that is epic and . We show that . Clearly, . Let . Then, . It follows that for some , and hence, . Thus, , and so , as desired. Consequently, is -projective by Wakamatsu’s Lemma. Thus, by hypothesis, and hence, is an isomorphism. So is an -injective envelope.
Theorem 36. Let be a -hereditary ring and If is a submodule of an -injective -module , then, the following are equivalent: (1) is a special -injective envelope of is a -projective essential extension of
Proof. . It follows by Proposition 35.
. By hypothesis, we have an exact sequence: with an -injective module and an -projective module. This sequence is a generator of all -projective extensions of . By Theorems 33 and 34, we have an -projective extension sequence of which gives an -injective envelope of . Then, we have the diagram shown in Figure 6.
It is easy to see that . We claim that . Since , . We define the following homomorphism , . Obviously, is well defined. By diagram chasing, we see that is injective. But both and are surjective, so is . Therefore, is -projective essential extension of . This contradicts the hypothesis that is -projective essential extension of . This implies that and so is an isomorphism.
6. -Injective Cover
In this section, we assume is the class of all pure projective -modules and we prove that all modules have -injective covers.
Proposition 37. The class of all -injective modules is closed under pure submodules.
Proof. Let be a pure submodule of a -injective module . Then there is a pure exact sequence and a functor preseves this sequence is exact whenever This implies that the sequence is also exact for all . It follows that for all , as desired.
Theorem 38. Every -module has a -injective preenvelope.
Proof. Let be an -module. By Lemma 5.3.12 in [13], there is a cardinal number such that for any -homomorphism with a -injective -module, there exists a pure submodule of such that and Clearly, is closed under direct products and by Proposition 37 is -injective. Hence, the theorem follows Proposition 6.2.1 in [13].
Proposition 39. The class of all -injective modules is injectively resolving.
Proof. Let be an exact sequence of left -modules with . Let . By Theorem 38, every module has a -preenvelope. By Lemma 1.9 in [29], has a special -preenvelope. By Lemma 2.2.6 in [6], has a special -precover. Then, there exists an exact sequence with and . We prove that is -injection, i.e., to prove that . For this, it suffices to extend any to an element of . Clearly, has -precover, where , , and . As the class is closed under extensions, . Since with and a -injective module, then there exists such that . That is, we have the diagram shown in Figure 7.
Now, we define , where is a restriction map. Then, there exists such that . Hence, we have the diagram shown in Figure 8.
The -injectivity of yields a homomorphism such that . So for each , we get . Then, there exists a map such that , and we get . Thus, we have the diagram shown in Figure 9.
Since is -injective, there exists such that . Thus, , where . Hence, is -injective.
Proposition 40. Let be a -injective -module and be a pure submodule . Then, an -module is -injective.
Proof. By Proposition 39, is injectively resolving. Let and be a pure submodule of . By Proposition 37, is -injective. From the short exact sequence , we get is -injective.
The terminology of the Definition 41 is used in [30].
Definition 41 (see [31]). An -module is if there is a projective resolution such that the first syzygy is finitely presented (small). We denote by the class of all -modules and the class of all -modules.
Theorem 42 (see [31], Theorem 2.3). Let be an -module. (1)If is an -module, then commutes with direct limits(2)If is a small--module, then commutes with direct sumsThe converses of these statements are true when an -module is finitely generated.
We recall that an -module is pure projective if and only if is a direct summand of a direct sum of finitely presented -modules [32]. Warfield proved that every pure projective -module is a direct sum of finitely presented modules over a Noetherian local ring in Corollary 4 in [33]. Further, Puninski and Rothmalar [18] proved that the general question of when all pure projective modules are direct sum of finitely presented -modules. They proved this over a hereditary Noetherian ring ([18], Corollary 6.5). Thus, every pure projective -module is finitely presented over a hereditary Noetherian ring. It follows that commutes with direct limits for all pure projective -modules By Theorem 42, every pure projective -module is over a hereditary Noetherian ring.
The converse part of the Proposition 43 is not necessarily true. It will be held immediately over a hereditary Noetherian ring by the argument of the previous paragraph. In this article, we proved the converse part of Proposition 43 over a Noetherian ring by Theorem 45; that is, -injective -modules are injective from ([34], Theorem 3).
Proposition 43. Every -injective -module is absolutely pure.
Proof. Every finitely presented -module is pure projective, hence the proposition.
Theorem 44. Let be a perfect ring. Then the following conditions are equivalent: (1) is coherent(2)The class of -injective -modules is closed under direct limits and closed under direct sums(3)Every absolutely pure -module is injective
Proof. . It follows by Corollary 57 (2) and (3) in [30].
. By Proposition 43 and Theorem 3.2 in [1], is coherent. A perfect coherent ring is Noetherian. Then, we get (3) immediately by Theorem 3 in [34].
Let denote the class of all absolutely pure -modules.
Theorem 45. The following conditions are equivalent: (i) is Noetherian(ii)-injective -modules coincide with the injective -modules(iii)Absolutely pure -modules coincide with injective -modules
Proof. By Proposition 43, every -injective -module is an absolutely pure -module. It follows that every absolutely pure -module is injective by Theorem 3 in [34]. Hence, every -injective -module is injective.
Clearly, By (ii), Thus
It follows Theorem 3 in [34].
By Theorem 45, -injective cover and injective cover for every -module are coincide over a Noetherian ring. Therefore, we prove Theorem 46 and Proposition 47 without Noetherian condition.
The following result establishes an analog version of Theorem 2.6 in [16].
Theorem 46. Suppose Then, every -module has a -injective precover.
Proof. Let be a set with , where is the cardinal in [11] (Theorem 5). Denote the power set of . We find all the binary operations for each element , and we get a new collection . From , find all the scalar multiplications, which are functions from the cross product into itself. This remains a set which is denoted by . Some collection of members of form a module, and we can get the class of -injective modules which is contained in the class . Clearly, is contained in the class . Since then by Theorem 42 (2) the direct sum of -injective -modules is -injective. Hence is an -injective -module because every injective module is -injective. We prove that is a -injective precover. That is, to show that if for any homomorphism with , then we have the following diagram (Figure 10):
Let be the kernel of the map , where is sufficiently large. Then, is sufficiently small since . By Bican et al.’s Theorem [11], has a nonzero submodule that is pure in . Therefore, is -injective by Proposition 37. This implies that is -injective by Proposition 40. If is still sufficiently large, then repeat the process from the map . Since is sufficiently small, has a nonzero submodule that is pure in . Thus, is -injective. But again, this may be too large. Then by continuing this process we get is -injective since is closed under direct limits, and it is sufficiently small, namely . Then the map can be factored through a -injective module . Let . We define a map such that , with . Clearly, is a linear map. Then, we have the diagram shown in Figure 11.
Hence, we get the diagram shown in Figure 12.
Thus, is a -injective precover.
By Theorem 42 (1), the class of all -injective -modules are closed under direct limits if . But we prove this result over a coherent in Proposition 47.
Proposition 47. Let be a coherent ring. Then, the class of all -injective -modules is closed under direct limits.
Proof. It follows Corollary 57(2) shown in [30].
Theorem 48 follows from Theorem 46 and Proposition 47 by Corollary 5.2.7 in [13].
Theorem 48. Let be a coherent ring and Then, every -module has a -injective cover.
Data Availability
No data were used to support the findings of the study.
Conflicts of Interest
The authors declare that they have no conflict of interests.
Acknowledgments
We thank the referee for several corrections and important comments that improved the presentation at several points. The first author was supported by Harish-Chandra Research Institute (HRI), Allahabad, Institute Postdoctoral Fellowship (Code No. M170563VF).