On modules whose dual is of finite Gorenstein dimension: The -torsion, homological criteria, and applications in modules of differentials of order
Abstract.
In this paper, we aim to obtain some results under the condition that the dual of a module has finite Gorenstein dimension. In this direction we derive results involving vanishing of Ext as well as the freeness or totally reflexivity of modules. For instance, we provide a generalization of a celebrated theorem by Evans and Griffith, obtain criteria for the totally reflexivity of modules over Cohen-Macaulay rings as well as of locally totally reflexive modules on the punctured spectrum, and recover a result by Araya. Moreover, we prove that the Auslander-Reiten conjecture holds true for all finitely generated modules over a Noetherian local ring such that and . Additionally, we derive Gorenstein criteria under the condition that the dual of certain modules is of finite Gorenstein dimension. Furthermore, we explore some applications in the theory of the modules of Kähler differentials of order , specifically concerning the -torsionfreeness of these modules and the Herzog-Vasconcelos’s conjecture.
2020 Mathematics Subject Classification:
Primary: 13D45, 13D07, 13C10, 13C14; Secondary: 13D05, 13D02, 13H10, 14B15, 14B05.1. Introduction
The Gorenstein dimension is a well-developed topic in commutative algebra, as evidenced by various works such as [5, 14, 44, 50, 18, 52]. Despite the extensive research in this area, there are relatively few results concerning the condition of the Gorenstein dimension of the dual of a certains modules being finite. The aim of this work is to obtain results in this direction, including some that involve vanishing of Ext and homological criteria.
We prove the following theorem, which is a generalization of a result of Evans and Griffith ([15, Theorem 3.8]) and is an analogous version of [14, Proposion 2.4] when .
1.1 Theorem.
Let be a Noetherian local ring and be a non-negative integer. Let be a finitely generated non-zero -module such that has locally finite Gorenstein dimension on . Then:
is -torsionfree is -syzygy satisfies .
This theorem provides an answer to [37, Question 1.1]. Additionally, it allows us to recover one of the results described in the main theorem of [52].
1.2 Corollary ([52, Theorem 5.8(1)]).
Let be a Noetherian local ring, be a finitely generated -module such that and be a non-negative integer. If satisfies , then is -torsionfree.
As an application of Corollary 1.2 and [50, Theorem 42], we characterize under certain conditions and in terms of vanishing of Ext modules, the total reflexivity of a finitely generated -module with . We obtain the following.
1.3 Theorem.
Let be a Noetherian local ring of depth , and let be a finitely generated -module such that . Then:
-
(1)
satisfies if and only if is totally reflexive if and only if for all .
-
(2)
Let be a Cohen-Macaulay local ring and is non-zero. If is a non-negative integer such that and for all then is totally reflexive.
-
(3)
If is non-zero and locally totally reflexive on the punctured spectrum of , then with equality if and only if is totally reflexive.
Additionally, part of this work is motivated by the celebrated Auslander-Reiten conjecture ([6]), which states the following.
1.4 Conjecture.
(Auslander-Reiten) Let be a Noetherian local ring and let be a finitely generated -module. If for all , then is free.
This conjecture is still open but is true in several cases. For a list of some of them we refer the reader to [30], it introduction and to the recent work [12]. In [2], Araya proved the following result that provided an positive answer to Auslander-Reiten conjecture, and that implies that it is valid over normal Gorenstein local rings of dimension at least two.
1.5 Theorem ([2, Corollary 10]).
Let be a Gorenstein local ring of dimension . Let be a maximal Cohen-Macaulay -module. If is locally free on the punctured spectrum of and , then is free.
Subsequently, results concerning the context of Gorenstein rings that recover this theorem have appeared. For instance, the last theorem in [42, Section 1] and [3, Corollary 1.6]. It is important to mention that in a more general context, [30, Proposition 2.10(4)] and [8, Proposition 1.7] are improvements of these results. In terms of the finiteness of the Gorenstein dimension of the dual of a module, we achieve the following results that recover the last theorem in [42, Section 1], [3, Corollary 1.6] and of course Theorem 1.5.
1.6 Theorem.
Let be a Noetherian local ring of depth and let be a finitely generated -module. Suppose that is locally free on the spectrum punctured such that . If and , then is free.
1.7 Theorem.
Let be a Noetherian ring satisfying and a subset of containing . Let
and .
Let be a finitely generated -module and suppose that is locally free on and that . If for all and , then is projective.
On the other hand, in [12, Corollary 6.9(2)], Dey and Ghosh proved that the Auslander-Reiten conjecture holds for finitely generated -modules such that and . Motivated by this, we analyze if the conjecture holds if we consider instead of . Consequently, we get the following result that affirms that it holds.
1.8 Theorem.
Let be a Noetherian local ring of depth , and let be a finitely generated -module such that . Then the following conditions are equivalent:
-
(1)
is free.
-
(2)
is free and for all .
-
(3)
has finite projective dimension and for all and .
Additionally, we also derive some Gorenstein criteria involving the condition that the dual of certain -modules are of finite Gorenstein dimension. We do this essentially by discussing of some questions about that. We obtain the following.
1.9 Theorem.
Let be a Noetherian local ring.Then is Gorenstein in each one of the following cases:
-
(1)
be a Cohen-Macaulay local ring and there exists an -module such that and it satisfies one of the following:
-
(1.1)
is Cohen-Macaulay of positive rank and .
-
(2.2)
is an Ulrich module and .
-
(1.1)
-
(2)
for all finitely generated -modules such that .
-
(3)
has positive depth, and every finitely generated -modules whose dual is non-zero and of finite Gorenstein dimension, is also of finite Gorenstein dimension.
-
(4)
has depth at most one and .
In the last part of this paper, we focus on the Kähler differential modules and the derivation modules defined over locally Noetherian schemes and analytic spaces, where is an integer. The motivation for studying these modules arises from the Zariski-Lipman conjecture [33], which states: Let be a complex variety such that the tangent sheaf is locally free. Then is smooth. This conjecture, proposed about 50 years ago, is still open, although it has been resolved in some special cases, see, for instance, [39, Section 4]. Following the same direction as the Zariski-Lipman Conjecture, there is also a homological version of it, independently proposed by Herzog and Vasconcelos, which predicts that if , then is free. Unlike the former, this problem seems to be widely open (with exceptions in specific cases; see, for example, [26, Section 4] and [40]). There is also another conjecture called the Strong Zariski-Lipman Conjecture: If , then is regular. Due to these conjectures, the following questions arise:
1.10 Questions.
-
(i)
(Generalizations of Herzog-Vasconcelos’s conjecture (GHVC)) For some integer , we have . Under what assumptions on and , and for which values of , does this imply that is locally free?
-
(ii)
(Strong generalizations of Herzog-Vasconcelos’s conjecture (SGHVC)) For some integer , we have . Under what assumptions on and , and for which values of , does this imply that is locally smooth?
It is worth noting that these questions and others appear in the works of Graf [22] (in the case of algebraic varieties and analytic spaces) and in the case of affine schemes in [35]. Historically, to answer these questions, it is necessary to study the concepts of torsion, cotorsion, torsion-free, and reflexive of the Kähler differential module (see, for example, a summary in [39, p. 133] and [22]). For this reason,in the Sections 8 and 9, we obtain some direct applications connecting these concepts with the concepts of -torsion-free and -syzygy. Consequently, we also obtain some special answers for GHVC and SGHVC.
The organization of this paper is as follows. In Section 2 we provide definitions, notations, and some results that are considered in this paper. In Section 3 we prove Theorem 1.1 and explore some of its consequences. In Section 4, we prove Theorem 1.3, and in Section 5 we provide freeness criteria including Theorem 1.7 and 1.8. In Section 6, we discuss some questions about of the Gorensteiness of a ring involving Gorenstein dimension of the dual of modules, and prove Theorem 1.9. In Section 7, we introduce notation and review established results regarding Kähler differential modules and differential modules. Then, in Section 8, we investigate the properties of the Kähler differential module , examining its local freeness, reflexivity, -torsion-freeness, and -syzygy properties, specially when . Finally, in the last section, we provide a partial response to the SGHV and SGHV conjectures.
2. Setup and preliminaries
Throughout this paper, is a Noetherian ring, and all -modules are considered to be finitely generated. Whenever is local, we denote by and its maximal ideal and its residual field respectively.
2.1 Definition.
Let be two -modules, and be a non-negative integer.
-
(1)
For an -module , we set and .
-
(2)
Let
be a projective resolution of resolution of .
-
(2.a)
For , the -syzygy of , denoted by , is defined as the kernel of the map . When , we set .
-
(2.b)
The Auslander Transpose of , denoted by , is defined as the cokernel of the induced map When , we set
-
(2.a)
-
(3)
We write if there exist projective -modules such that .
-
(4)
is a -syzygy if there exists an exact sequence
where each is of projective dimension.
-
(5)
is -torsionfree if for all .
-
(6)
We set (resp. ) to the set of all prime ideals of such that (resp. .
-
(7)
We say that satisfies (resp. ) if (resp. .
-
(8)
We say that satisfies if the local ring is Gorenstein for all .
The -modules and () are uniquely determined up to projective summands. It is easy to see that . Note that is a -syzygy if and only if there exists an -module such that . Besides, for every -module , there exists an exact sequence of functors (see [5, Theorem 2.8]).
(2.1) |
Whenever is local, we consider the Auslander tranpose and the syzygies (of an -module) defined using minimal free resolutions. In this case, these -modules are defined uniquely up to isomorphism (rather than projective equivalence).
The notion of Gorenstein dimension was introduced by Auslander [36] and developed by Auslander and Bridger in [5].
2.2 Definition.
Let be an -module.
-
(1)
We say that is totally reflexive if is reflexive and for all .
-
(2)
The Gorenstein dimension of , denoted by , is defined to be the infimum of all non-negative integers such that there exists an exact sequence
where each is totally reflexive.
We can observe that if and only if is totally reflexive. Below, we collect some facts related to the Gorenstein dimension.
2.3 Facts.
Let be an -module.
-
(1)
([9, Theorem 1.2.7]) If and , then
-
(2)
([5, Lemma 3.19(1)]) is totally reflexive if and only if is totally reflexive.
-
(3)
([9, Corollary 1.2.9]) Let be an exact sequence of -modules. If two -modules of the sequence have finite Gorenstein dimension, then so has the third.
-
(4)
([9, Proposition 1.2.10]) with equality if .
Now, assume that is local.
- (5)
-
(6)
([9, Theorem 1.4.9]) The following conditions are equivalent:
-
(6.1)
is Gorenstein.
-
(6.2)
for all (finitely) -modules .
-
(6.3)
.
-
(6.1)
The formula given in Fact 2.3(5) is known as the Auslander-Bridger formula. From the definition of the Auslander transpose and Fact 2.3(3) we can derive the following remark.
2.4 Remark.
Let be a ring and let be an -module. Then if and only if .
3. A generalization of a theorem of Evans and Griffith
The main goal of this section is to provide a generalization of a celebrated result by Evans and Griffith ([15, Theorem 3.8]), which characterizes a module to be -torsionfree, -syzygy, involving the condition .
Before presenting such a generalization, we provide the following result, which will be crucial for the proof of our main theorem in this section.
3.1 Lemma.
Let be a local ring of depth , be a non-zero -module and be an integer. If is a -syzygy of an -module of depth , then .
Proof.
It is sufficient to consider . We proceed by induction on . Suppose . Then and there exists an exact sequence where is free -module and has of depth . The depth lemma gives us the inequalities
(3.1) |
and
(3.2) |
Since , it follows from (3.1) that . Thus (3.2) say us that .
Now suppose . Since is a -syzygy, there exists an exact sequence
where . Then it induces exact sequences
(3.3) |
and
(3.4) |
If , then (3.4) says that and, by the Auslander-Buchsbaum formula, which does not occurs by assumption. So, . Then by induction, we have that . Applying the depth formula in (3.3) we get that
and
Thus, if , then , which contradicts our assumption. Therefore, we must have . Consequently, , and thus , which implies . Therefore, . ∎
Below, we provide a characterization of a module to be -torsionfree and -syzygy, involving the condition . This characterization generalizes a celebrated result by Evans and Griffith ([15, Theorem 3.8]).
3.2 Theorem.
Let be a ring and let be a non-negative integer. Let be a non-zero -module such that has locally finite Gorenstein dimension on . Then the following conditions are equivalent:
-
(1)
is -torsionfree.
-
(2)
is -syzygy.
-
(3)
satisfies .
Proof.
The implications follow from [50, Proposition 38]. Now let us prove . We prove this by induction on . If , there is nothing to prove. Assume that . By induction, is -torsionfree, that is for all , and so it remains to conclude that .
By contradiction, assume that . Let . Then . From (2.1), we have an exact sequence
which implies that and hence
Since for all , we see from [52, 5.6] that
We claim that . Indeed, if , then by assumption , and from the Auslander-Bridger formula, . But since for all , we get that is a totally reflexive. This shows that , which does not occurs.
Now, as , it follows from Lemma 3.1 that . But since satisfies , we get that , a contradiction. ∎
Observe that Theorem 3.2 is the version with the dual of an -module being of finite Gorenstein dimension of [14, Proposition 2.4] when . Now, we explore some consequences of Theorem 3.2.
3.3 Corollary.
Let be a ring and let be an -module. Suppose that for all . Then the following conditions are equivalent:
-
(1)
is -torsionfree.
-
(2)
is -syzygy.
-
(3)
satisfies .
3.4 Corollary.
Let be a ring and let be an -module. Suppose that for all . Then the following conditions are equivalent:
-
(1)
is -torsionfree.
-
(2)
is -syzygy.
-
(3)
satisfies .
-
(4)
is reflexive.
This theorem allows us to recover one of the results described in the main result of [52].
3.5 Corollary ([52, Theorem 5.8(1)]).
Let be a local ring, be an -module such that and be a non-negative integer. If satisfies , then is -torsionfree.
3.6 Question ([37, Question 1.1]).
When -syzygy modules are -torsionfree?
The following result is an analogous version of [44, Theorem 4.5], which is one of the main results of that paper.
3.7 Theorem.
Let be a ring, and let be an -module such that has locally finite Gorenstein dimension on . Then is -torsionfree if and only if the following hold:
-
(1)
is -torsionfree for all .
-
(2)
for all with .
Proof.
If is -torsionfree, then it is clear that (1) is true. Besides, by Theorem 3.2, satisfies , which shows that (2) holds.
Reciprocally, suppose that (1) and (2) hold. By Theorem 3.2, it is sufficient to show that satisfies that is for all . By (2), it holds if . Now, let . Then by the Auslander-Bridger formula, . Thus, by (1), is totally reflexive, and hence as well. In particular, . ∎
4. Characterization of totally reflexives
In this section, is a local ring. Next, we derive some criteria of totally reflexive modules, assuming the condition , and involving the vanishing of the Ext functor. We first provide a general criterion via the vanishing of , and subsequently, we provide criteria over Cohen-Macaulay local rings and later for locally totally reflexive modules on the punctured spectrum.
4.1 Theorem.
Let be a local ring of depth . Let be an -module such that . Then the following are equivalent:
-
(1)
satisfies .
-
(2)
is totally reflexive.
-
(3)
for all .
Proof.
. According to Corollary 3.5, is -torsion-free, meaning that for all . Since it follows that is totally reflexive. Hence, is totally reflexive.
. This follows from definition of totally reflexives modules.
Over Cohen-Macaulay local rings
As an immediate application of Theorem 4.1, we obtain the following characterization for a module over a Cohen-Macaulay ring to be totally reflexive, involving the vanishing of the Ext functor.
4.2 Theorem.
Let be a Cohen-Macaulay local ring of dimension and let be a non-zero -module such that . Then the following are equivalents:
-
(1)
is maximal Cohen-Macaulay.
-
(2)
is totally reflexive.
-
(3)
for all .
Proof.
Since is Cohen-Macaulay, it is clear that is maximal Cohen-Macaulay if and only if satisfies . Thus, the result follows immediately from Theorem 4.1. ∎
4.3 Remark.
It is known that if is a Gorenstein local ring and is maximal Cohen-Macaulay, then for all . We can observe from Theorem 4.2, that in this fact instead of being Gorenstein, we can consider the weaker condition of .
We recall that the grade of an -module is defined as:
4.4 Corollary.
Let be a Cohen-Macaulay local ring and let be a Cohen-Macaulay -module such that and . Then is totally reflexive.
Proof.
Now, as an application of Theorem 4.2, we extend the formula presented in [13, Lemma 4.1] to all -modules such that . But before, we recall the definition of reduced grade. The reduced grade of an -module is defined as
4.5 Remark.
Let be a ring. If then
4.6 Theorem.
Let be a Cohen-Macaulay local ring of dimension , and let be an -module such that . Then the equality
(4.1) |
holds in .
Proof.
We may assume that is complete, and hence has a canonical module . We prove the equality by considering the following cases:
-
(1)
is not maximal Cohen-Macaulay and .
-
(2)
is not maximal Cohen-Macaulay and .
-
(3)
is maximal Cohen-Macaulay.
The equality in the case (1) was proved in [13, Lemma 4.1].
4.7 Corollary.
Let be a Cohen-Macaulay local ring with a canonical module , and let be an -module such that . Then:
Proof.
4.8 Theorem.
Let be a Cohen-Macaulay local ring of dimension and let be an -module such that . Let be a non-negative integer such that . If for all then is totally reflexive.
Proof.
We may assume that and (passing to the completion of if necessary) that has a canonical module .
Since for all and , by Corollary 4.7, for all . Since [7, Exercise 3.1.24] say us that
then or . The second case does not occur because by assumption. Thus and hence is maximal Cohen-Macaulay. Thus, by Theorem 4.2, is totally reflexive.
∎
Over locally totally reflexive modules on the punctured spectrum
The following result provides a reflexivity criterion for -modules with of locally totally reflexives on the punctured spectrum of .
4.9 Theorem.
Let be a local ring of depth and let be a non-zero -module such that . Suppose that is locally totally reflexive on the punctured spectrum of . Then with equality if and only if is totally reflexive.
Proof.
First we prove that . Set and suppose that . Then satisfies since is locally totally reflexive on the punctured spectrum of . In particular, satisfies . By Theorem 4.1, is totally reflexive, which contradicts the fact of .
Now, we prove the second part of the result. It is clear that if is totally reflexive, then . Suppose now that . Since is locally totally reflexive on the punctured spectrum of and , note that satisfies . Hence, by Theorem 4.1, is totally reflexive. ∎
4.10 Corollary.
Let be a local ring of depth and let be a non-zero module. Suppose that is locally totally reflexive on the punctured spectrum of . Suppose that . Then the following are equivalent:
-
(1)
.
-
(2)
.
-
(3)
is totally reflexive.
-
(4)
for all .
In this section, we have derived a number of criteria for an -module with to be totally reflexive. Thus, we close this section with the following question.
4.11 Question.
Let be a ring and let be an -module. When does imply that ?
5. Freeness Criteria
In this section, we aim to obtain freeness criteria by developing the following topics:
Freeness criteria from Theorem 4.8
As an application of Theorem 4.8, we obtain the following result:
5.1 Theorem.
Let be a Cohen-Macaulay local ring of dimension , and let be an -module such that . Let be a non-negative integer such that and for all . Then is free.
Proof.
As , then . By Theorem 4.8, is totally reflexive. Thus is reflexive and . Therefore, is free. ∎
5.2 Corollary.
Let be a Cohen-Macaulay local ring and let be an -module such that . If is maximal Cohen-Macaulay, then is free.
Proof.
This follows immediately from Theorem 5.1 by taking . ∎
5.3 Corollary.
Let be a Cohen-Macaulay local ring and let be an -module such that and . If is Cohen-Macaulay, then is free.
Proof.
As an application of Corollary 5.2, we will provide an answer to the Generalization of Herzog-Vasconcelos’s conjecture (see Proposition 9.4). Another application is given below.
5.4 Theorem.
Let be a local Cohen-Macaulay. The maximal ideal is Cohen-Macaulay if and only .
Proof.
Assume that is Cohen-Macaulay and that . Consider the exact sequence
(5.1) |
It induces an exact sequence
Since then and . Therefore, by Corollary 5.3, is free, and hence . From the exact sequence (5.1), we obtain an exact sequence
As then . Consequently, which does not occurs.
Conversely, assume that . If it is clear that all -modules are Cohen-Macaulay. Suppose . Since is Cohen-Macaulay, then there exists an -regular element . Note that is also an -regular element. Therefore which shows that is maximal Cohen-Macaulay. ∎
Some generalizations of a result of Araya
As mentioned in the introduction, after the publication of the Araya’s result referenced in Theorem 1.5, results have appeared that generalize or recover such theorem, first in the context of Gorenstein rings and later in more general contexts. Motivated by this, we provide some generalizations of that celebrated result in terms of the dual of an -module is of finite Gorenstein dimension.
5.5 Proposition.
Let be a local ring of depth . Let be a locally free -module on the spectrum punctured such that . If for all , then is free.
Proof.
5.6 Corollary.
Let be a local ring of depth . Let be a locally free -module on the spectrum punctured such that . If and , then is free.
5.7 Theorem.
Let be a ring satisfying and a subset of containing . Let
and .
Suppose that is locally free on and that . If for all and , then is projective.
Proof.
We may suppose that . We show that is free for all . By assumption this is true if . So, suppose that . Therefore and since satisfies . By definition of and , note that , so for all and . Hence, by Proposition 5.5, is free. ∎
Now, we explore some consequences of Theorem 5.7 for Cohen-Macaulay local rings.
5.8 Remark.
([30, Remark 2.11]) Let be a Cohen-Macaulay local ring of dimension and be an integer such that . Then
5.9 Theorem.
Let be a Cohen-Macaulay local ring of dimension , be an -module and be a positive integer. Suppose that is locally of finite projective dimension on and that for all . Then the following hold:
-
(1)
If and for all , then is free.
-
(2)
If then .
Proof.
(1) We claim that is locally free on . In fact, given , we have by hypothesis that and for all . Hence, by [38, p. 154, Lemma 1(iii)], is free as -module. Thus item (1) follows immediately from Theorem 5.7 and Remark 5.8.
(2) We separate the proof in two cases. Let . We prove the result by considering two cases.
(2.a) Suppose . In this case, is totally reflexive. Hence is totally reflexive and for all . So, it follows from (1) that is free. In particular, .
(2.b) Suppose . We have for all . By [30, Lemma 2.6], we get
for all . Hence for all . Note that and that is locally of finite projective dimension on . Hence, by (1), is free and therefore . ∎
5.10 Corollary.
Let be a Cohen-Macaulay local ring of dimension , and let be a maximal Cohen-Macaulay -module such that . Let be an integer such that is locally of finite projective dimension on . If for all . Then is free.
Proof.
As and is maximal Cohen-Macaulay, it follows from the Auslander-Bridger Formula that . Then, by Theorem 5.9(2), , and hence . This yields that is free. ∎
5.11 Corollary.
Let be a Cohen-Macaulay normal local ring of dimension , and let be an -module such that . Suppose that for all and . Then is free.
Proof.
As is normal, then is a regular local ring for all . Hence, is locally of finite projective dimension on . Thus, the result follows from Theorem 5.9. ∎
On the Auslander-Reiten conjecture, and a result of Dey and Ghosh
In [12, Corollary 6.9(2)], Dey and Ghosh demonstrated that the Auslander-Reiten conjecture holds for (finitely generated) -modules satisfying and . Motivated by this, we investigate whether the conjecture remains valid when we consider instead of in the referenced case.
5.12 Theorem.
Let be a local ring of depth , and let be an -module such that . Then the following conditions are equivalent.
-
(1)
is free.
-
(2)
is free and for all .
-
(3)
has finite projective dimension and for all and .
Proof.
is trivial.
5.13 Corollary.
Let be local ring. The Auslander-Reiten conjecture holds true for all (finitely generated) -modules such that and .
An -module is said to be semidualizing if the natural map is an isomorphism and for all .
5.14 Corollary.
Let be a local ring, and let be a semidualizing -module. If , then .
6. Gorenstein criteria and related questions
Throughout this section, let be a local ring. In this section, we aim to provide a number of criteria for a local ring to be Gorenstein in terms of the dual of certain modules having finite Gorenstein dimension. In this sense, we discuss some related questions to this subject.
The first criterion that we present in this section is an application of Theorem 4.2. Before stating it, first recall that if is a ring with total quotient ring , then an -module is said to have a (generic) rank, denoted by , if is a free -module of rank . Let denote the Hilbert-Samuel multiplicity of and denote the minimum number of generators of . Moreover, if is Cohen-Macaulay and , then it is said to be an Ulrich module ([21, Definition 2.1]).
6.1 Proposition.
Let be a Cohen-Macaulay local ring, and let be an -module such that . Then is Gorenstein in each one of the following cases.
-
(1)
is Cohen-Macaulay with positive rank and .
-
(2)
is an Ulrich module and .
Proof.
Due to Proposition 6.1, is natural to ask the following.
6.2 Question.
Let be a local ring. Suppose there exists a non-free (finitely generated) -module such that and . Then, is Gorenstein?
6.3 Remark.
On the other hand, it is known that if is a local ring such that for all -modules, then is Gorenstein ([9, Theorem 1.4.9]). Thus, by considering the dual of -modules it is natural to ask the following.
6.4 Question.
Let be a local ring. If for all (finitely generated) -modules , then is Gorenstein?
Then motivated by Question 6.4, we obtain the following result:
6.5 Theorem.
Let be a local ring. If for all (finitely generated) -modules such that then is Gorenstein.
Proof.
First, suppose that . Then and by assumption . Since for some , note that . Hence is Gorenstein.
Now, suppose that . We may write . Let the homomorphism defined by . Since , we see that is injective. Thus we have an exact sequence
Dualizing this exact sequence we obtain a sequence
(6.1) |
If we see from (6.1) that which shows that is regular, and consequently Gorenstein. If by assumption, , so from (6.1), concluding that is Gorenstein. ∎
As an immediate consequence we obtain the following corollary that answer positively Question 6.4
6.6 Corollary.
Let be a local ring such that for all (finitely generated) -modules . Then is Gorenstein.
6.7 Remark.
Corollary 6.6 can be obtained in other ways. For instance, it could be derived from [44, Corollary 3.5]. Additionally, this corollary can be obtained as an application of Theorem 3.2. Indeed, setting , by the assumption and Theorem 3.2, all (finitely generated) -modules satisfy the following equivalences:
is -torsionfree is -syzygy satisfies
Then by [37, Theorem 1.4], satisfies , concluding that is Gorenstein.
For a local ring of depth at most one, the finiteness of the Gorenstein dimension of the dual of its maximal ideal characterizes its Gorensteiness.
6.8 Proposition.
Let be a local ring of depth at most one. If , then is Gorenstein.
Proof.
First, suppose . Since then . Since , the Auslander-Bridger formula shows that is totally reflexive, and hence as well. Now, by considering the exact sequence we see that . Thus, we conclude that is Gorenstein.
Now, suppose that . Then the exact sequence induces an exact sequence
Since , we see that and for some . Thus, we obtain an exact sequence
In view of that we see that . This shows that is Gorenstein. ∎
The next question that we will present is motivated by Question 4.11. Consider the following condition:
- (GDUAL):
-
Every (finitely generated) -module whose dual is finite Gorenstein dimension, is also of finite Gorenstein dimension.
6.9 Question.
Let be a local ring. If satisfies (GDUAL), then is Gorenstein?
6.10 Remark.
Let be a local ring. Suppose . By the Auslander-Bridger formula, all -modules of finite Gorenstein dimension are totally reflexive. Moreover of for any -module we have that the following implications hold:
In particular, satisfies (GDUAL). Hence the existence of non-Gorenstein local rings of depth zero implies that Question 6.9 is false in general.
Motivated by condition , we define a weaker condition than it, and prove that if is of positive depth and satisfies such condition, then is Gorenstein. This will show that Question 6.9 is positive for local rings of positive depth.
- (GDUAL*):
-
Every (finitely generated) -module whose dual is non-zero and of finite Gorenstein dimension, is also of finite Gorenstein dimension.
6.11 Proposition.
Let be a local ring of depth at least one satisfying . Then is Gorenstein.
Proof.
Since has positive depth, note that . Let . Then and hence . Thus by assumption. Hence, is Gorenstein. ∎
Now, motivated by Corollary 5.14, we will discuss about a question that was addressed by Holanda and Miranda-Neto in [28]. The question is as follows:
6.12 Question.
([28, Question 5.24]) Let be a Cohen-Macaulay local ring with a canonical module . If , then must be Gorenstein?
Firstly, we observe that the answer to Question 6.12 is positive and follows directly from the Corollary 5.14. However, it is worth highlighting that this question had already been answered affirmatively, albeit through a different proof, by Asgharzadeh in [4, Theorem 11.3]. Additionally, it is worth noting that long before Holanda and Miranda-Neto posed this question, the same result could be obtained directly from the Foxby equivalence (see [9, Theorem 3.4.11]).
We finish this section by proposing the following question, which is motivated by a celebrated result of Foxby [17] that asserts that the existence of a non-zero (finitely generated) -module of projective dimension and injective dimension both finite implies that is Gorenstein.
6.13 Question.
If there exists a (finitely generated) -module such that and , then is Gorenstein?
7. Preliminary facts on Kähler differentials
In this section, we establish our notation and, for the reader’s convenience, we recall some known results concerning Kähler differential modules and differential modules, both in the context of algebras and locally ringed spaces.
Affine derivation modules and Kähler fifferentials -th order
Let be a -algebra and be an -module. Recall that a -linear map is said to be a derivation if for any two elements , the following identity holds:
A derivation of order can be defined generalizing the previous identity as follows.
A -linear map is said to be a -th order derivation if for any , the following identity holds:
where means that this element does not appear in the product. The set of -th order derivations of an -algebra into an -module over will be denoted by . When , we shall use the notation in place of .
The module of derivations of order can be represented as follows. Let denote the kernel of the homomorphism , . Giving structure of -module to by multiplying on the left, we define the -module
Define the map , . This map is a derivation of order , and its image generates as an -module (see [41, Chapter II-1]).
The -module is called the module of Kähler differentials of order of over . The map is called the canonical derivation of in . It comes equipped with a universal derivation with the property that composition with yields an isomorphism ([43, Proposition 1.6]). To see more properties regarding modules of derivations and Kähler differentials, we recommend [43, 51, 41].
7.1 Remark.
It is important to note that the differential module and Kähler differentials may not be finitely generated. However, they are finitely generated in certain cases. For instance: If is essentially of finite type over . If is a field with a valuation and is an analytic -algebra, meaning is module-finite over a convergent power series ring . If is a field, is a complete local ring, and is a finite extension of . For this reason, in this article, for each , we will consider to be finitely generated as an -module. In particular, by the universal property, is also finitely generated as an -module.
Next, we collect some definitions and properties that will play a central role in this paper.
7.2 Definition.
A ringed space consists of a topological space paired with a sheaf of rings , known as the structure sheaf of . If is a local ring for every , is termed a locally ringed space. Let denote the maximal ideal of ; is referred to as the residue field of at . We consider the following locally ringed spaces: locally Noetherian schemes (i.e., a scheme is locally Noetherian if it has a covering by open affine subsets , where each is a Noetherian ring) and analytic spaces (the ringed space is called an analytic space if every point has a neighborhood such that is isomorphic to some analytic model space, where for all , and is Noetherian and reduced; for further details, e.g., see [29], [23]).
7.1. Global Kähler differentials
Next, we will recall the globalized definition of the constructions given above. To do this, instead of gluing together the modules , we provide a global definition, and then point out that it reduces to the original definition given above. Before defining it, it is worth noting that the fiber product exists in our context (see, for example [10, Corollary 0.32] and [20, Theorem 4.18]). Let be a morphism of space. This induces a morphism , called the diagonal morphism. The map comes with a map of sheaves on . Although really a map of sheaves of rings, we will regard it as a map of sheaves of -modules. Let be its kernel, again regarded as an -module.
7.3 Definition.
The sheaf relative Kähler differential of -module is defined as:
Thus, more generally, the sheaf of relative Kähler -differentials of over is defined by
Note that if , where is a field, is usually denoted by . For further details and properties, see for example, [1, Section I.6.6, p. 34].
7.4 Facts.
Let and be (locally) Noetherian spaces.
- (1)
-
(2)
([34, Proposition 1.17]). If is a morphism of spaces, then for there is a canonical isomorphism of -modules
Since the stalk commutes with exterior power, more generally, one obtains .
-
(3)
([1, Corollaire (I6.5.5)]). There is also a definition of the sheaf of -derivations over , denoted by . Moreover, the universal property of comes down to saying that there is an isomorphism of -modules Naturally, is defined the sheaf of -th order -derivations over .
-
(4)
[46, Lemma 28.20.2, Section 05P1] Let be a locally Noetherian scheme, and let be a coherent -module. Then is locally free if and only if for all , the stalk is a free -module.
Based on Remarks 7.1 and Facts 7.4, the schemes considered in this section will be locally Noetherian, and the -module will also always be considered a non-zero coherent sheaf for each integer . Additionally, we assume that the sheaves have non-zero stalks, implying that the chosen elements belong to the support. For simplicity, the stalk of , as defined in Facts 7.4(2), will be denoted by , where .
7.5 Lemma.
Let and be locally Noetherian spaces, then the following properties hold:
-
(i)
is a -coherent.
-
(ii)
for there is a canonical isomorphism of -modules
Thus, will be denoted by
8. Applications on the -torsion of the modules of differentials
Next, we study when the Kähler differential module of order is locally free, reflexive, -torsion-free and -syzygy, specially when . The study of these properties in the modules of Kähler differential of order concerning the regularity questions of affine rings has a long history; when and is a Spec of a field (see Kunz-Waldi’s book [31]). Next, we provide a summary given by Milher in [39], which is relevant for motivating the study of these concepts. Let and (definition for and see 8.6). Lipman (see also Suzuki [47]) showed that for an affine complete intersection germ over a field of characteristic zero, is torsion-free if and only if is normal at , while is reflexive if and only if (see [33]). Vetter proved in [49] that for a non-smooth reduced complete intersection singularity and for , the module is torsion-free (resp., reflexive) if (resp., ) and has torsion (resp., cotorsion) if (resp., if ). Lebelt strengthened Vetter’s result by showing that if is torsion-free (resp., reflexive), then (resp., ) (see [32]). Greuel in [24] show that has torsion for any ; (for any reduced complex analytic space, [16]). This fact implies that since and consequently results in for . Also, new results on the connection between torsion and cotorsion and reflexivity of Kähler differential modules can be found in [22, 39]. Due to it and the results mentioned above, we obtain the following results:
8.1 Theorem.
Let and be locally Noetherian spaces and let be a non-negative integer, , such that has locally finite Gorenstein dimension on . Then the following conditions are equivalent:
-
(1)
is -torsionfree.
-
(2)
is -syzygy.
-
(3)
satisfies .
Proof.
As an immediate consequence of Theorem 8.1 and Corollaries 3.3 and 3.4, we have the following corollaries.
8.2 Corollary.
Let and be locally Noetherian spaces and let , such that has locally finite Gorenstein dimension on . Then the following conditions are equivalent:
-
(1)
is -torsionfree.
-
(2)
is -syzygy.
-
(3)
satisfies .
8.3 Corollary.
Let and be locally Noetherian spaces and let such that has locally finite Gorenstein dimension on . Then the following conditions are equivalent:
-
(1)
is -torsionfree.
-
(2)
is -syzygy.
-
(3)
satisfies .
-
(4)
is reflexive.
8.4 Remark.
Let be a morphism of schemes which is locally of finite type. If we consider and as affine Noetherian schemes, that is, and where and are local Noetherian rings. By Fact 7.4(2) and Lemma 7.5(ii), if we consider and closed points respectively, then we have and . Moreover, these are finitely generated -modules.
From Remark 8.4, Theorem 8.1, and Corollaries 8.2 and 8.3, we obtain similar results in the affine context, that is, the derivation modules and Kähler differential modules defined over algebras.
Some application in algebraic variety and analytic variety
Before establishing our significant results that connect algebraic geometry and analytic geometry, let’s recall the following definitions: Let be a (Noetherian) ring, be an (finitely generated) -module and let be an -homomorphism. We say that is torsionless, if the map is injective, and is reflexive if is an isomorphism. If fails to be surjective, we say that has cotorsion and we denote the cotorsion by . Note that, by Definition 2.1(5), torsionless is equivalent to 1-torsionfree, and reflexive is equivalent to 2-torsionfree ([50, Proposition 5]).
The torsion submodule of is defined as the kernel of the natural map where is the total quotient ring of . One says that is torsion-free if , and is a torsion module if .
8.5 Remark.
Clearly the torsion submodule is contained in . If is a Noetherian domain and is finitely generated, one can show that . Thus, the concepts of torsionless and torsion-free are equivalent ([23, p.70]). If is only a reduced, Noetherian ring and is finitely generated, still the concepts are equivalent ([45]). Therefore, in the same situations as above, torsionless, torsion-free, and 1-torsionfree are equivalent.
We shall apply these notions to , ringed spaces locally Noetherian (or analytic space), and the stalk at a point of a coherent sheaf viewed as an -module. For a coherent sheaf , define the torsion and cotorsion sheaves, and , as the kernel and cokernel sheaves of the natural map . At the stalk level, this gives an exact sequence of -modules
where the middle map is the evaluation map upon the identification .
8.6 Definition.
Let be a scheme(or analytic space), and let . Then is a vector space over , called Zariski cotangent space to at . The Zariski tangent space of in is, by definition, the dual vector space
Recall that a locally Noetherian is said to be nonsingular (or regular) at if the Zariski tangent space to at has dimension equal to , that is, is regular; otherwise, we say that is singular at . We say that is regular (or smooth) if it is regular at all of its points. Denote by and the regular locus and the singular locus of , respectively. For , the codimension of the singular locus of at is defined as .
We say that has isolated singularity if has dimension at most zero (i.e., equal to or ).
8.7 Definition (Cohen-Macaulay Sheaf).
A coherent sheaf over a scheme is said to be (maximal) Cohen-Macaulay if, for every point , the stalk is a (maximal) Cohen-Macaulay module over the local ring . is a complete intersection at , if the stalk is complete intersection ring. A point in is said to be normal if is a normal local ring.
In the last part, we considered either one of the following two possible settings, one analytic and one algebraic. More specifically, we consider a reduced complex analytic variety with the structure sheaf , such that, at , the stalk is isomorphic to a reduced local -algebra, which is a quotient of a convergent power series ring . We also consider an algebraic variety over an algebraically closed field of characteristic with the structure sheaf , such that, at , the stalk is isomorphic as a -algebra, reduced to a quotient of a regular local ring by an ideal.
Recent results on the connection between torsion and cotorsion and reflexivity of Kähler differential modules on algebraic and analytic varieties can be found in the papers by Graf and Miller-Vassiliadou ([22, 39]). Next, we establish a connection between the results of Graf and Miller-Vassiliadou with -torsionfree, -syzygy, and where .
8.8 Proposition.
Let be either a complex analytic variety or an algebraic variety over an algebraically closed field of characteristic zero. Assume that is a complete intersection at a normal point of dimension . Suppose is a singular point of . Then, for , the following conditions are equivalent:
-
(1)
-
(2)
is not -torsionfree.
-
(3)
is not -syzygy.
-
(4)
does not satisfy .
Proof.
8.9 Proposition.
Let be either a complex analytic variety or an algebraic variety over an algebraically closed field of characteristic zero. Assume that is a complete intersection at a normal point of dimension . Suppose is a singular point of . Then, for , the following conditions are equivalent:
-
(1)
.
-
(2)
is not -torsionfree.
-
(3)
is not -syzygy.
-
(4)
does not satisfy .
-
(5)
is not reflexive.
9. Applications on Herzog-Vasconcelos’s Conjecture
One of the main conjectures, which may have motivated the study of properties of the Kähler differential modules and the regularity of schemes or analytic spaces, is the famous Lipman-Zariski conjecture [33], which states the following: Let be a complex variety such that the tangent sheaf is locally free. Then is smooth. This conjecture remains widely open. Following the same direction as the Zariski-Lipman Conjecture, there is also a homological version proposed by Herzog and Vasconcelos as:
9.1 Conjecture (Herzog-Vasconcelos’s conjecture).
Let be a local Noetherian ring. If , then is a free -module.
9.2 Conjecture (Strong Zariski-Lipman Conjecture).
Let be a local Noetherian ring. If , then is regular.
These conjectures, particularly the Lipman-Zariski conjecture when , have been extensively studied and resolved in special cases. References of Lipman-Zariski Conjecture can be found in Miller-Vassiliadou [39, Section 4]. While progress has been made for , it has been to a lesser extent compared to . Further investigations can be found in Graf [22] and Miller-Vassiliadou [39, Section 4]. It’s worth noting that these conjectures are generally false for positive characteristic. For instance, for , consider the surface over a perfect field of characteristic defined by the equation , where divides . Then one can see that is free (see [33, p. 892]). Example for , see [35]. Due to this, from now on our spaces or rings will be equicharacteristic and of characteristic zero.
Next, we present conjectures which are generalizations of 9.1 and 9.2, where SGHVC was introduced by Ludington [35] and GHVC by Graf [22] in the analytical context. Le be an -module. Since, Then, the projective dimension of a coherent sheaf over is defined as
9.3 Questions.
-
(i)
(Generalizations of Herzog-Vasconcelos’s conjecture (GHVC)) For some integer , . Under what assumptions on and , and for which values of , does this imply that is locally free?
-
(i)
(Strong generalizations of Herzog-Vasconcelos’s conjecture (SGHVC)) For some integer , . Under what assumptions on and , and for which values of , does this imply that is locally smooth?
-
(iii)
In particular, consider and as in Remark 8.4. Assume that for some integer , . Under what assumptions on , and for which values of does this imply that is free (resp. smooth)?
The following results provide a partial answer to the SGHV conjecture when is Cohen-Macaulay and the Kähler differential module is Cohen-Macaulay or maximal Cohen-Macaulay.
9.4 Proposition.
Let and be locally Noetherian spaces, with being Cohen-Macaulay, and suppose that for some . Then is locally free if one of the following conditions holds:
-
(1)
If is maximal Cohen-Macaulay.
-
(2)
If is Cohen-Macaulay.
Proof.
It is suffices to show it at the level of stalks. Note that, if and only if for all . Moreover, is Cohen-Macaulay (maximal) if and only if is Cohen-Macaulay (maximal) for all .
Note that in Proposition 9.4 for , if we consider , as in Remark 8.4, the assumption implies that is Gorenstein (see [25]). Thus, by Theorem 4.2 and Corollary 4.4, is totally reflexive, hence is free.
The following results provide a partial answer to the SGHV conjecture for complete intersections.
9.5 Proposition.
Let be either a complex analytic variety or an algebraic variety over an algebraically closed field of characteristic zero. Assume that is a complete intersection at a point of dimension and is (maximal) Cohen-Macaulay for some . Let the codimension of the singular locus of at be at least three. If at , then is a smooth point of .
Proof.
In the proof of Theorem 9.4, it is noted that such a result holds at the level of stalks. Additionally, in the same proof or by Corollary 5.2, Corollary 5.3, we show that is free as an -module. Therefore, is a -free module, because . Now by hypothesis and from [39, Theorem 4.1], we get the result. ∎
Acknowledgement.
We thank Rafael Holanda for his valuable comments on the manuscript. The first author was supported by grant 2022/03372-5, São Paulo Research Foundation (FAPESP). The second author was supported by grant 2019/21181-0, São Paulo Research Foundation (FAPESP)
References
- [1] J. Dieudonné A. Grothendieck, Éléments de géométrie algébrique. IV: Étude locale des schémas et des morphismes de schémas Quatrième partie, vol. 32, 1967, pp. 5–361.
- [2] T. Araya, The Auslander-Reiten conjecture for Gorenstein rings, Proc. Amer. Math. Soc. 137 (2009), no. 6, 1941–1944.
- [3] T. Araya, O. Celikbas, A. Sadeghi, and R. Takahashi, On the vanishing of self extensions over Cohen-Macaulay local rings, Proc. Amer. Math. Soc. 146 (2018), no. 11, 4563–4570.
- [4] M. Asgharzadeh, Reflexivity revisited, arXiv preprint arXiv:1812.00830v6 (2022).
- [5] M. Auslander and M. Bridger, Stable module theory, Mem. of the AMS, No. 94, American Mathematical Society, Providence, R.I., 1969.
- [6] M. Auslander and I. Reiten, On a generalized version of the Nakayama conjecture, Proc. Amer. Math. Soc. 52 (1975), no. 1, 69–74.
- [7] W. Bruns and J. Herzog, Cohen-Macaulay rings, Cambridge University Press, New York, 1998.
- [8] O. Celikbas, S. Dey, T. Kobayashi, H. Matsui, and A. Sadeghi, Two theorems on the vanishing of Ext, arXiv preprint arXiv 2308.08999v2 (2023).
- [9] L. W. Christensen, Gorenstein dimensions, vol. 1747, Springer, 2000.
- [10] T. de Jong and G. Pfister, Local analytic geometry: Basic theory and applications, Advanced lectures in mathematics, Vieweg, 2000.
- [11] S. Dey and D. Ghosh, Complexity and rigidity of Ulrich modules, and some applications, Math. Scand. 129 (2023), no. 2, 209–237.
- [12] S. Dey and D. Ghosh, Finite homological dimension of Hom and vanishing of Ext, arXiv preprint arXiv 2310.10607v2 (2023).
- [13] M. T. Dibaei and A. Sadeghi, Linkage of finite Gorenstein dimension modules, J. Algebra 376 (2013), 261–278.
- [14] M. T. Dibaei and A. Sadeghi, Linkage of modules and the Serre conditions, J. Pure Appl. Algebra 219 (2015), no. 10, 4458–4478.
- [15] E. G. Evans and P. Griffith, Syzygies, London Mathematical Society Lecture Note Series, Cambridge University Press, 1985.
- [16] A. Ferrari, Coomologia e forme differenziali sugli spazi analitici complessi, Annali della Sc. Norm. Super. di Pisa 25 (1971), no. 3, 469–480.
- [17] H.-B Foxby, Isomorphims between complexes with applications to the homological theory of modules., Math. Scand. 40 (1977), 5–19.
- [18] D. Ghosh and T. J. Puthenpurakal, Gorenstein Rings via homological dimensions, and symmetry in vanishing of Ext and Tate cohomology, Algebr. Represent. Theory 27 (2024), no. 1, 639–653.
- [19] D. Ghosh and M. Samanta, Auslander-reiten conjecture for modules whose (self) dual has finite complete intersection dimension, arXiv preprint 2405.01497v1 (2024).
- [20] U. Görtz and T. Wedhorn, Algebraic geometry: Part I: Schemes. With examples and exercises, Advanced Lectures in Mathematics, Vieweg+Teubner Verlag, 2010.
- [21] S. Goto, R. Takahashi, and N. Taniguchi, Almost gorenstein rings – towards a theory of higher dimension, J. Pure Appl. Algebra 219 (2015), no. 7, 2666–2712.
- [22] P. Graf, The generalized Lipman-Zariski problem, Math. Ann. 362 (2015), 241–264.
- [23] H. Grauert and R. Remmert, Coherent analytic sheaves, Springer-Verlag, 1984.
- [24] G-M. Greuel, Der gauss-manin zusammenhang isolierter singularitäten von vollständigen durchschnitten, Math. Ann. 214 (1975), 235–266.
- [25] J. Herzog, The module of differentials, Lecture notes, Workshop on Commutative Algebra and its Relation to Combinatorics and Computer Algebra, 16–27 May 1994, Trieste, Italy.
- [26] J. Herzog and A. Martsinkovsky, Gluing Cohen-Macaulay modules with applications to quasihomogeneous complete intersections with isolated singularities, Comment. Math. Helv. 68 (2093), 365–384.
- [27] R. G. Heyneman and M. E. Sweedler, Affine hopf algebras I, J. Algebra 13(2) (1969), 192–241.
- [28] R. Holanda and C.B. Miranda-Neto, Vanishing of (co)homology, freeness criteria, and the Auslander-Reiten conjecture for Cohen-Macaulay Burch rings, arXiv preprint arXiv:2212.05521 (2022).
- [29] S. Ishii, Introduction to singularities, Springer-Verlag, Tokyo, 2014.
- [30] K. Kimura, Auslander–Reiten conjecture for normal rings, arXiv preprint arXiv:2304.03956 (2023).
- [31] E. Kunz and R. Waldi, Regular differential forms, Contemporary Mathematics 79 (1988).
- [32] K. Lebelt, Torsion äuerer Potenzen von Moduln der homologischen Dimension 1, Math. Ann. 211 (1974), 183–197.
- [33] J. Lipman, Free derivation modules on algebraic varieties, Amer. J. Math.. 87 (1965), 874–898.
- [34] Q. Liu, Algebraic geometry and arithmetic curves, Oxford Graduate Texts in Mathematics, 2002.
- [35] A. L. Ludington, A counterexample two conjectures about high order derivations and regularity, Osaka J. Math. 14 (1977), 159–163.
- [36] M. Mangeney, C. Peskine, and L. Szpiro, Anneaux de Gorenstein, et torsion en algèbre commutative, Séminaire Samuel. Algèbre commutative 1 (1966-1967), 2–69.
- [37] H. Matsui, R. Takahashi, and Y. Tsuchiya, When are -syzygy modules -torsionfree?, Arch. Math. 108 (2017), no. 4, 351–355.
- [38] H. Matsumura, Commutative ring theory, Cambridge Studies in Advanced Mathematics, Cambridge University Press, 1987.
- [39] C. Miller and S. Vassiliadou, (Co)torsion of exterior powers of differentials over complete intersections, J. of Singularities 19 (2019), 131–162.
- [40] G. Müller and D. P. Patil, The Herzog-Vasconcelos conjecture for affine semigroup rings, Comm. Algebra 27 (2099), 3197–3200.
- [41] Y. Nakai, High order derivations I, Osaka J. Math. (1970), 1–27.
- [42] M. Ono and Y. Yoshino, An Auslander–Reiten principle in derived categories, J. Pure Appl. Algebra 221 (2017), no. 6, 1268–1278.
- [43] H. Osborn, Modules of differentials I, Math. Ann. 170 (1967), 221–244.
- [44] M. Salimi, E. Tavasoli, and S. Yassemi, k-torsionless modules with finite Gorenstein dimension, Czechoslov. Math. J. 62 (2012), 663–672.
- [45] G. Scheja, Differential modules of analytic rings, Lectures given at University of Genova, 1968.
- [46] Stacks-project, https://stacks.math.columbia.edu/tag/05p1.
- [47] S. Suzuki, On torsion of the module of differentials of a locality which is a complete intersection, J. Math. Kyoto Univ. 4-3 (1965), 471–476.
- [48] B. Ulrich, Gorenstein rings and modules with high numbers of generators, Math. Z. 188 (1984), 23–32.
- [49] U. Vetter, Äuere Potenzen von Differentialmoduln reduzierter, vollständiger Durchschnitte, Manuscripta Math. 2 (1970), 67–75.
- [50] M. Vladimir, Gorenstein dimension and torsion of modules over commutative Noetherian rings, Commun. Algebra 28 (2000), no. 12, 5783–5811.
- [51] K. Kosaki Y. Nakai and Y. Ishibashi, High order derivations, II, J. Sci. Hiroshima. Univ Ser. A-I. 34(1) (1970), 17–27.
- [52] M. R. Zargar, O. Celikbas, M. Gheibi, and A. Sadeghi, Homological dimensions of rigid modules, Kyoto J. Math. 58 (2018), no. 3, 639 – 669.