Some functors preserving exceptionality

1. Anhui Polytechnic University, Wuhu 241000 Anhui, China;

   E-mail: [email protected]

2. School of Mathematical Sciences, Anhui University, Hefei 230601, Anhui, China;

   E-mail: [email protected]

3. School of Mathematics and Statistics, Guizhou University, Guiyang 550025, Guizhou, China;

   E-mail: [email protected] / [email protected]

1. Corresponding author

1. ORCID: 0009-0001-6073-7587

2. ORCID: 0000-0001-7002-4153

3. ORCID: 0009-0005-1110-386X

1. Abstract: We constructed some tensor functors that send each exceptional sequence in a module category to another exceptional sequence in another module category by using split extensions and recollements.

1. 2020 Mathematics Subject Classification: 16E30, 16G10, 16S70.

1. Keywords: Exceptional sequences; split-by-nilpotent extensions; recollements

1 Introduction

The notion of exceptional sequences was introduced by Gorodentsev and Rudakov [12] in the study of exceptional vector bundles over $\mathbb{P}^{2}$. Later, this concept was adapted to the quiver representation framework established by Crawley–Boevey [9] and Ringel [21]. In recent years, topics on exceptional sequences have remained highly popular. In triangulated categories, Aihara and Iyama [1] show that there is a bijection between full exceptional sequences and silting objects. For any finite-dimensional algebra, Buan and Marsh [7] introduce the notions of $\tau$-exceptional and signed $\tau$-exceptional sequences and prove that for a fixed algebra of rank $n$, and for any positive integer $t\leqslant n$, there is a bijection between the set of signed $\tau$-exceptional sequences of length $t$, and ordered support $\tau$-rigid objects with $t$ indecomposable direct summands. The notion of weak exceptional sequences was introduced by Sen [23], which can be viewed as another modification of the standard case, different from the works of Buan and Marsh [7]. Westin and Thuresson [24] provide a classification of generalized tilting modules and full exceptional sequences for a certain family of quasi-hereditary algebras, specifically dual extension algebras. The theory of tilting over split-by-nilpotent extensions and recollements have garnered significant attention. For example, Assem and Marmaridis [3], as well as Assem and Zacharia [5], along with Liu and Wei [13], and Suarez [17] have investigated numerous interesting results concerning the relationship between tilting modules, tilting pairs and support $\tau$-tilting modules in $\mathsf{mod}(A)$ and induced objects in $\mathsf{mod}(R)$.

A Recollements of Abelian categories, denoted by $\mathcal{R}(\mathcal{A},\mathcal{B},\mathcal{C})$, is a diagram

(1.3)

such that

1. $(i^{*},i_{*})$, $(i_{*},i^{!})$, $(j_{!},j^{*})$, $(j^{*},j_{*})$ are adjoint pairs;

2. the functors $i_{*}$, $j_{!}$, and $j_{*}$ are fully faithful;

3. $\mathrm{Im}(i_{*})=\mathrm{Ker}(j^{*})$.

Recollements of Abelian and triangulated categories were introduced by Beilinson, Bernstein, and Deligne [6] in connection with derived categories of sheaves on topological spaces, with the idea that one triangulated category may be “glued together” from two others. This concept plays an important role in the representation theory of algebras, and many algebraists and authors have researched on the recollements, for example, weighted projective line [8], Abelian and triangulated categories [26, 10, 18, 30], Homological theory [19, 20, 25, 28], Gorenstein-Homological theory [29, 27], and tilting theory [14, 16, 15], etc. Psaroudakis provided an important recollement of module categories which is the form $\mathcal{R}(\mathsf{mod}(A/A\varepsilon A),A,\varepsilon A\varepsilon)$ in [19, Example 2.7] (see Example 2.4 in this paper), where $\varepsilon$ is an idempotent of $A$, such that $i^{*}$ and $j_{!}$ are two tensor functors

$$i^{*}=-\otimes_{A}{{}_{A}(A/A\varepsilon A)}:\ \mathsf{mod}(A)\to\mathsf{mod}(A/A\varepsilon A)$$

$$\text{and~{}}j_{!}=-\otimes_{\varepsilon A\varepsilon}\varepsilon A:\ \mathsf{mod}(\varepsilon A\varepsilon)\to\mathsf{mod}(A)$$

of module categories.

On the other hand, we know that split extensions $\xi:R\to A$ of an algebra $A$ by a nilpotent bimodule, introduced by Assem and Zacharia in [5], is also an important concept providing tensor functors

$$-\otimes_{A}{{}_{A}R_{R}}:\ \mathsf{mod}(A)\to\mathsf{mod}(R)$$

$$\text{and~{}}-\otimes_{R}~{}{{}_{R}A_{A}}:\ \mathsf{mod}(R)\to\mathsf{mod}(A)$$

of module categories. Then the following question is natural:

Question 1.1.

Let $F$ be a tensor functor of module categories from $\mathsf{mod}(A)$ to $\mathsf{mod}(B)$. Then for each exceptional sequence $(M_{1},\ldots,M_{r})$ in $\mathsf{mod}(A)$, when $F$ satisfies what conditions, $(F(M_{1}),\ldots,F(M_{r}))$ is an exceptional sequence?

We want to answer the above question by using recollements and split extensions. Then we obtain two main results of this paper.

Theorem 1.2.

Let $\xi:R\to A$ be a split extension of $A$ by a nilpotent bimodule $Q$ and $\mathcal{R}(A^{\prime},A,A^{\prime\prime})$ be a recollement of module categories.

• (Theorem 3.5) For any exceptional sequence $(M_{1},\ldots,M_{r})$ in $\mathsf{mod}(A)$, if:

  • $R={{}_{A}R}$ is a projective left $A$-module;

  • $\mathrm{Hom}_{A}(M_{k},M_{k}\otimes_{A}Q)=0$ $(1\leqslant k\leqslant r)$ and $\mathrm{Hom}_{A}(M_{j},M_{i}\otimes_{A}Q)=0$ $(1\leqslant i<j\leqslant r)$ hold;

  • $\mathrm{Ext}^{n}_{A}(M_{k},M_{k}\otimes_{A}Q)=0$ $(1\leqslant k\leqslant r)$ and $\mathrm{Ext}^{n}_{A}(M_{j},M_{i}\otimes_{A}Q)=0$ $(1\leqslant i<j\leqslant r)$ hold,

  then $(M_{1}\otimes_{A}R,\ldots,M_{r}\otimes_{A}R)$ is an exceptional sequence in $\mathsf{mod}(R)$.

• (Theorem 3.10) For any exceptional sequence $(X_{1},\ldots,X_{t})$ in $\mathsf{mod}(A^{\prime})$ and any exceptional sequence $(Y_{1},\ldots,Y_{u})$ in $\mathsf{mod}(A^{\prime\prime})$, if $i^{*}$ and $i^{!}$ are exact functors, then $(i_{*}(X_{1})$, $\ldots$, $i_{*}(X_{t}))$ and $(j_{!}(Y_{1})$, $\ldots$, $j_{!}(Y_{u}))$ are exceptional sequences in $\mathsf{mod}(\mathit{\Lambda})$.

2 Preliminaries

In this section, we will give some terminologies and some preliminary results. Throughout this article, $\mathbb{K}$ denotes an algebraically closed field, and $n$ represents a positive integer. Let $A$ be a finite-dimensional $\mathbb{K}$-algebra. We consistently assume that the category $A$-mod consisting of finitely generated right $A$-modules. Denote the projective dimensions of module $M_{A}$ by $\mathrm{proj.dim}(M_{A})$.

2.1 Exceptional Sequences

Recall, as seen in [12], that an indecomposable left (or right) $A$-module $M$ is called exceptional provided that

1. $\mathrm{End}_{A}M$ is isomorphic to $\mathbb{K}$;

2. $\mathrm{Ext}_{A}^{n}(M,M)=0$, for any positive integer $n$.

A module pair $(M_{i},M_{j})$ of exceptional modules in $\mathsf{mod}(A)$ is called an exceptional pair if

1. Hom$(M_{j},M_{i})=0$;

2. Ext${}_{A}^{n}(M_{j},M_{i})=0$, for any positive integer $n$.

Furthermore, a module sequence $(M_{1},M_{2},\ldots,M_{r})$ consisting of exceptional modules is deemed an exceptional sequence if every pair $(M_{i},M_{j})$ with $1\leqslant i<j\leqslant r$ satisfies the above criteria for being exceptional. It is said to be complete if $r=r_{A}$, the rank of the Grothendieck group $K(A)$.

Recall that the following definitions are given in [2].

• $S\in\mathsf{mod}(A)$ is called a brick if $\mathrm{Hom}_{A}(S,S)$ is isomorphic to $\mathbb{K}$. The set of isoclasses of bricks in $\mathsf{mod}(A)$ will be denoted by $\mathsf{brick}(A)$.

• A subset $\mathcal{S}\subseteq$ $\mathsf{brick}(A)$ is called a semibrick if $\mathrm{Hom}_{A}(S_{1},S_{2})=0$ for any $S_{1}\neq S_{2}\in\mathcal{S}$. The set of semibricks in $A\mathsf{mod}$ will be denoted by $\mathsf{sbrick}(A)$.

It is evident that every simple module is a brick, and a set of isoclasses of simple modules forms a semibrick. However, while every exceptional module is indeed a brick, a sequence $(S_{1},S_{2},\cdots,S_{r})$ of isoclasses of simple modules does not necessarily constitute an exceptional sequence. As demonstrated in [11], there exists a bijection between $\tau$-tilting modules and sincere left finite semibricks, facilitating the construction of (sincere) semibricks over split-by-nilpotent extensions.

2.2 Split-by-nilpotent extensions

Let $A$, $R$ be two finite dimensional algebras.

Definition 2.1 (​​[5, Definition 1.1]).

We say that $R$ is a split extension of $A$ by the nilpotent $(A,A)$-bimodule $Q$, or simply a split-by-nilpotent extension if there exists a split surjective algebra morphism $\xi:R\to A$ whose kernel $Q:=\operatorname{Ker}(\xi)=\{r\in R\mid\xi(r)=0\}$ is contained in the radical $\mathrm{rad}R$ of $R$.

Remark 2.2.

Let $R$ be a split-by-nilpotent extension of $A$ by the nilpotent bimodule $Q$. Clearly, the short exact sequence of $(A,A)$-bimodules

$0\to~{}_{A}Q_{A}\to~{}_{A}R_{A}\to~{}_{A}A_{A}\to 0$

splits. Therefore, there exists an isomorphism ${}_{A}R_{A}\cong A\oplus{{}_{A}Q_{A}}$. The module categories $\mathsf{mod}(A)$ and $\mathsf{mod}(R)$ are related by the following four functors

$$-\otimes_{A}R:\mathsf{mod}(A)\to\mathsf{mod}(R),$$

$$-\otimes_{R}A:\mathsf{mod}(R)\to\mathsf{mod}(A),$$

$$\mathrm{Hom}_{A}({{}_{A}R},-):\mathsf{mod}(A)\to\mathsf{mod}(R),$$

$$\mathrm{Hom}_{R}({{}_{R}A},-):\mathsf{mod}(R)\to\mathsf{mod}(A).$$

Moreover, we have

$$(-\otimes_{A}R)\otimes_{R}A\simeq 1_{\mathsf{mod}(A)},$$

$$\mathrm{Hom}_{R}(A_{R},\mathrm{Hom}_{A}(R_{A},-))\simeq 1_{\mathsf{mod}(R)}.$$

2.3 Recollements of Abelian categories

We list some properties of recollements (see [14, 19, 18]) in this section, which will be used in the sequel.

Lemma 2.3.

Let $(\mathcal{A},\mathcal{B},\mathcal{C})$ be a recollement of abelian categories. Then we have

• $i^{*}j_{!}=0=i^{!}j_{*}$;

• the functors $i_{*}$ and $j^{*}$ are exact, $i^{*}$ and $j_{!}$are right exact, and $i^{!}$and $j_{*}$ are left exact;

• all the natural transformations

  $$i^{*}i_{*}\rightarrow 1_{\mathcal{A}},1_{\mathcal{A}}\rightarrow i^{!}i_{*},1_{\mathcal{C}}\rightarrow j^{*}j_{!},j^{*}j_{*}\rightarrow 1_{\mathcal{C}}$$

  are natural isomorphisms;

• $\text{ if }i^{*}\text{ is exact, then }i^{!}j_{!}=0\text{, and if }i^{!}\text{is exact, then }i^{*}j_{*}=0\text{. }$

The following lemma is widely studied, which plays a crucial role in the sequel.

Lemma 2.4 (​​[19, Example 2.7]).

Let $\varepsilon$ be an idempotent of an algebra $A$. Denote by $\overline{A}$ the quotient $A/A\varepsilon A$ of $A$ and $\widetilde{A}$ the algebra $\varepsilon A\varepsilon$. Then we have a recollement of module categories:

(2.3)

where $\mathfrak{e}$ is an embedding functor.

3 Functors preserving exceptionality

3.1 Projective split extensions and functor $-\otimes_{A}R_{R}$

A split extension $\xi:R\to A$ of $A$ by the nilpotent $(A,A)$-bimodule $Q$ is said to be projective if $R={{}_{A}R}$, as a left $A$-module, is projective. In this section, we show that $-\otimes_{A}R$ preserves the exceptionality of exceptional sequence, see Theorem 3.5.

Lemma 3.1.

Let $\mathcal{A}$, $\mathcal{B}$ be two Abelian categories with enough projective objects, and $F:\mathcal{A}\to\mathcal{B}$, $G:\mathcal{B}\to\mathcal{A}$ be two functors. If $(F,G)$ is an adjoint pair and $F$ is an exact functor preserving projective objects, then

$\mathrm{Ext}^{n}_{\mathcal{B}}(F(M),F(N))\cong\mathrm{Ext}^{n}_{\mathcal{A}}(M,GF(N))$

holds for each objects $M$, $N$ in $\mathcal{A}$.

Proof.

Assume that the projective resolution of ${M}$ is

$$\cdots\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits P_{1}\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits P_{0}\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits M\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits 0. \spadesuit$$

Then we obtain that

$$\cdots\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits F(P_{1})\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits F(P_{0})\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits F(M)\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits 0 \clubsuit$$

is a projective resolution of $F(M)$, since $F$ is an exact functor and preserves projective modules. By applying the functor $\mathrm{Hom}_{\mathcal{B}}(-,GF(N))$ to sequence $\spadesuit$ and applying the functor $\mathrm{Hom}_{\mathcal{B}}(-,F(N))$ to sequence $\clubsuit$, we obtain that the following diagram

commutes since $(F,G)$ is an adjoint pair. Thus, $\mathrm{Ext}^{n}_{\mathcal{B}}(F(M),F(N))\cong\mathrm{Ext}^{n}_{\mathcal{A}}(M,GF(N))$ holds for all $n\in\mathbb{N}$. ∎

Corollary 3.2.

Let $\xi:R\to A$ be a projective split extension of a finite-dimensional algebra $A$ by the nilpotent $(A,A)$-bimodule $Q$. Then

$\mathrm{Ext}^{n}_{R}(M\otimes_{A}R,N\otimes_{A}R)\cong\mathrm{Ext}^{n}_{A}(M_{A},\mathrm{Hom}_{R}({{}_{A}R_{R}},N\otimes_{A}R))$

holds for all $n\in\mathbb{N}$.

Proof.

First of all, since ${{}_{A}}R$ is projective, then there exists a family of idempotents $(e_{i})_{1\leqslant i\leqslant t}$ such that

$${{}_{A}R}\cong A\oplus{{}_{A}Q}\cong A\oplus\bigoplus\limits_{i=1}^{t}Ae_{i},$$

where $Q\cong\bigoplus\limits_{i=1}^{t}Ae_{i}$. Then, for each projective right $A$-module $eA$, the following formula

$$eA\otimes_{A}R \cong(eA\otimes_{A}A)\oplus\Big{(}eA\otimes_{A}\Big{(}\bigoplus\limits_{i=1}^{t}Ae_{i}\Big{)}\Big{)}$$

$$\cong eA\oplus\Big{(}\bigoplus\limits_{i=1}^{t}(eA\otimes_{A}Ae_{i})\Big{)}$$

$$\cong eA\oplus\bigoplus\limits_{i=1}^{t}eAe_{i}$$

$$=e\Big{(}A\oplus\bigoplus\limits_{i=1}^{t}Ae_{i}\Big{)}$$

$$\cong eR$$

admits that $-\otimes_{A}R$ preserves projective modules. Note that $(-\otimes_{A}R,\mathrm{Hom}_{R}(R,-))$ is an adjoint pair, and since $R={{}_{A}R}$ is projective we have $-\otimes_{A}R$ is an exact. Then, by Lemma 3.1, we obtain that

$\mathrm{Ext}^{n}_{R}(M\otimes_{A}R,N\otimes_{A}R)\cong\mathrm{Ext}^{n}_{A}(M,\mathrm{Hom}_{R}({{}_{A}R_{R}},N\otimes_{A}{{}_{A}R_{R}}))$

holds for all $n\in\mathbb{N}$. ∎

Example 4.1 given in Section 4 illustrates that the condition “${{}_{A}}R$ is projective” is indispensable, yet it is not uncommon. We provide an instance in Example 4.2 for it.

Lemma 3.3.

Let $\xi:R\to A$ be a split extension of $A$ by the nilpotent bimodule $Q$ and $M_{A}$ be an exceptional module in $\mathsf{mod}(A)$. If

1. $\xi$ is projective;

2. $M_{A}$ satisfies $\mathrm{Hom}_{A}(M,M\otimes_{A}Q)=0$;

3. and $M_{A}$ satisfies $\mathrm{Ext}_{A}^{n}(M,M\otimes_{A}Q)=0$,

then $M\otimes_{A}R$, as a right $R$-module, is exceptional.

Proof.

First of all, we have the following isomorphism

$$\begin{split}\mathrm{Hom}_{R}(M\otimes_{A}R,M\otimes_{A}R)&\cong\mathrm{Hom}_{A}(M_{A},\mathrm{Hom}_{R}(R,M\otimes_{A}R))\\ &\cong\mathrm{Hom}_{A}(M_{A},M\otimes_{A}R)\\ &\cong\mathrm{Hom}_{A}(M_{A},M\otimes_{A}(A\oplus Q))\\ &\cong\mathrm{Hom}_{A}(M_{A},M_{A})\oplus\mathrm{Hom}_{A}(M_{A},M\otimes_{A}Q).\end{split}$$

Since $M_{A}$ is an exceptional module in $\mathsf{mod}(A)$, we have $\mathrm{Hom}_{A}(M_{A},M_{A})=\mathbb{K}$ by (E1). Moreover, by the condition (2) of this lemma, $\mathrm{Hom}_{A}(M_{A},M\otimes_{A}Q)=0$. Then we obtain

$$\mathrm{Hom}_{R}(M\otimes_{A}R,M\otimes_{A}R)=\mathbb{K},$$

i.e., the condition (E1) holds for $M\otimes_{A}R$.

Second, by the condition (1) of this lemma and Corollary 3.2, we have the following isomorphism

$$\begin{split}\mathrm{Ext}^{n}_{R}(M\otimes_{A}R,M\otimes_{A}R)&\cong\mathrm{Ext}^{n}_{A}(M_{A},\mathrm{Hom}_{R}(R,M\otimes_{A}R))\\ &\cong\mathrm{Ext}^{n}_{A}(M_{A},M\otimes_{A}R)\\ &\cong\mathrm{Ext}^{n}_{A}(M_{A},M\otimes_{A}(A\oplus Q))\\ &\cong\mathrm{Ext}^{n}_{A}(M,M)\oplus\mathrm{Ext}^{n}_{A}(M_{A},M\otimes_{A}Q),\end{split}$$

By the condition (3) of this lemma and (E2), we have

$$\mathrm{Ext}^{n}_{R}(M\otimes_{A}R,M\otimes_{A}R)=0,$$

i.e., the condition (E2) holds for $M\otimes_{A}R$. Thus, $M\otimes_{A}R$ is an exceptional module in $\mathsf{mod}(R)$. ∎

Lemma 3.4.

Let $\xi:R\to A$ be a split extension of $A$ by the nilpotent bimodule $Q$ and $M_{1}$ and $M_{2}$ be two modules in $\mathsf{mod}(A)$. If

1. $(M_{1},M_{2})$ is an exceptional pair in $\mathsf{mod}(A)$;

2. $\xi$ is projective;

3. $\mathrm{Hom}_{A}(M_{k},M_{k}\otimes_{A}Q)=0$ $(k\in\{1,2\})$ and $\mathrm{Hom}_{A}(M_{2},M_{1}\otimes_{A}Q)=0$ hold;

4. $\mathrm{Ext}^{n}_{A}(M_{k},M_{k}\otimes_{A}Q)=0$ $(k\in\{1,2\})$ and $\mathrm{Ext}_{A}^{n}(M_{2},M_{1}\otimes_{A}Q)=0$ hold for all $n\in\mathbb{N}$,

Then $(M_{1}\otimes_{A}R,M_{2}\otimes_{A}R)$ is an exceptional pair in $\mathsf{mod}(R)$.

Proof.

First of all, by Lemma 3.3, the conditions (2), (3) ($\mathrm{Hom}_{A}(M_{k},M_{k}\otimes_{A}Q)=0$ $(k\in\{1,2\})$) and (4) ($\mathrm{Ext}^{n}_{A}(M_{k},M_{k}\otimes_{A}Q)=0$ $(k\in\{1,2\})$) admit that $M_{1}\otimes_{A}Q$ and $M_{2}\otimes_{A}Q$ are exceptional modules.

Second, by using adjoint isomorphism, we have

$$\begin{split}\mathrm{Hom}_{R}(M_{2}\otimes_{A}R,M_{1}\otimes_{A}R)&\cong\mathrm{Hom}_{A}(M_{2},\mathrm{Hom}_{R}(R,M_{1}\otimes_{A}R))\\ &\cong\mathrm{Hom}_{A}(M_{2},M_{1}\otimes_{A}R)\\ &=\mathrm{Hom}_{A}(M_{2},M_{1}\otimes_{A}(A\oplus Q))\\ &\cong\mathrm{Hom}_{A}(M_{2},M_{1})\oplus\mathrm{Hom}_{A}(M_{2},M_{1}\otimes_{A}Q)\\ \end{split}$$

By the condition (1) of this lemma and the condition (E1^′), we have $\mathrm{Hom}_{A}(M_{2},M_{1})=0$. By the condition (3), we have $\mathrm{Hom}_{A}(M_{2},M_{1}\otimes_{A}Q)=0$. Thus, $\mathrm{Hom}_{R}(M_{2}\otimes_{A}R,M_{1}\otimes_{A}R)=0$, i.e., it follows that (E1^′) holds for $(M_{1}\otimes_{A}R,M_{2}\otimes_{A}R)$.

Third, by Corollary 3.2, we have

$$\begin{split}\mathrm{Ext}^{n}_{R}(M_{2}\otimes_{A}R,M_{1}\otimes_{A}R)&\cong\mathrm{Ext}^{n}_{A}(M_{2},\mathrm{Hom}_{R}(R,M_{1}\otimes_{A}R))\\ &\cong\mathrm{Ext}^{n}_{A}(M_{2},M_{1}\otimes_{A}R)\\ &\cong\mathrm{Ext}^{n}_{A}(M_{2},M_{1}\otimes_{A}(A\oplus Q))\\ &\cong\mathrm{Ext}^{n}_{A}(M_{2},M_{1})\oplus\mathrm{Ext}^{n}_{A}(M_{2},M_{1}\otimes_{A}Q)\\ \end{split}$$

By (E2^′) and the condition (4) of this lemma, we have $\mathrm{Ext}^{n}_{R}(M_{2}\otimes_{A}R,M_{1}\otimes_{A}R)=0$, i.e., (E2^′) holds for $(M_{1}\otimes_{A}R,M_{2}\otimes_{A}R)$. The above arguments follows that $(M_{1}\otimes_{A}R,M_{2}\otimes_{A}R)$ is an exceptional pair in $\mathsf{mod}(R)$. ∎

Now, we show the first main result of this paper.

Theorem 3.5.

Let $\xi:R\to A$ be a split extension of $A$ by the nilpotent bimodule $Q$, and $M_{1}$, $M_{2}$, $\ldots$, $M_{r}$ be right $A$-modules. If:

1. $(M_{1},M_{2},\ldots,M_{r})$ is an exceptional sequence;

2. $\xi$ is projective;

3. $\mathrm{Hom}_{A}(M_{k},M_{k}\otimes_{A}Q)=0$ $(1\leqslant k\leqslant r)$ and $\mathrm{Hom}_{A}(M_{j},M_{i}\otimes_{A}Q)=0$ $(1\leqslant i<j\leqslant r)$ hold;

4. $\mathrm{Ext}^{n}_{A}(M_{k},M_{k}\otimes_{A}Q)=0$ $(1\leqslant k\leqslant r)$ and $\mathrm{Ext}^{n}_{A}(M_{j},M_{i}\otimes_{A}Q)=0$ $(1\leqslant i<j\leqslant r)$ hold,

then $(M_{1}\otimes_{A}R,M_{2}\otimes_{A}R,\ldots,M_{r}\otimes_{A}R)$ is an exceptional sequence in $\mathsf{mod}(R)$. In addition, if $(M_{1},M_{2},\ldots,M_{r})$ is complete, then so is $(M_{1}\otimes_{A}R,M_{2}\otimes_{A}R,\ldots,M_{r}\otimes_{A}R)$.

Proof.

All modules $M_{k}\otimes_{A}Q$ ($1\leqslant k\leqslant r$) are exceptional by Lemma 3.3 and the conditions (2), (3) and (4) of this theorem. Moreover, for each $1\leqslant i<j\leqslant r$, (E1^′) and (E2^′) hold for $(M_{i}\otimes_{A}R,M_{j}\otimes_{A}R)$ by Lemma 3.4. Thus, $(M_{1}\otimes_{A}R,M_{2}\otimes_{A}R,\ldots,M_{r}\otimes_{A}R)$ is an exceptional sequence. ∎

3.2 Functors $i_{*}$ and $j_{!}$

Lemma 3.6 (Ma–Xie–Zhao [15]).

Let $\mathcal{R}(\mathit{\Lambda}^{\prime},\mathit{\Lambda},\mathit{\Lambda}^{\prime\prime})$ be a recollement as shown in (1.3). If $i^{*}$ and $i^{!}$ are exact functors, then $i_{*}$ and $j_{!}$ are exact functors preserving projective modules.

Proof.

[15, Proposition 2.6 (2)] has showed that $i_{*}$ functor preserving projective modules whenever $i^{!}$ is exact. In addition, $i_{*}$ is an exact functor by [15, Proposition 2.2 (2)]. Since $\mathsf{mod}(\mathit{\Lambda}^{\prime\prime})$ has enough projective objects, then $j_{!}$ preserves projective modules by [15, Proposition 2.5 (2)]. Moreover, by using [15, Lemma 2.4 (1)], we obtain that $j_{!}$ is exact since $i^{*}$ is exact, then this completes the proof of this lemma. ∎

Lemma 3.7.

Let $\mathcal{R}(\mathit{\Lambda}^{\prime},\mathit{\Lambda},\mathit{\Lambda}^{\prime\prime})$ be a recollement of module categories. If $i^{*}$ and $i^{!}$ are exact functors, then for any $M\in\mathsf{mod}(\mathit{\Lambda}^{\prime})$, $N\in\mathsf{mod}(\mathit{\Lambda}^{\prime\prime})$, and $n\geqslant 1$, we have

1. $\mathrm{Ext}_{\mathit{\Lambda}}^{n}(i_{*}(M),i_{*}(M))\cong\mathrm{Ext}_{\mathit{\Lambda}^{\prime}}^{n}(M,M)=0$;

2. $\mathrm{Ext}_{\mathit{\Lambda}}^{n}(j_{!}(N),j_{!}(N))\cong\mathrm{Ext}_{\mathit{\Lambda}^{\prime\prime}}^{n}(N,N)=0$.

Proof.

By the definition of recollement $\mathcal{R}(\mathit{\Lambda}^{\prime},\mathit{\Lambda},\mathit{\Lambda}^{\prime\prime})$, we have that $(i_{*},i^{!})$ is an adjoint pair, see (R1), and since $i^{*}$ is exact, we obtain that $i_{*}$ is an exact functor preserving projective modules by Lemma 3.6. Then, by Lemma 3.1,

$\mathrm{Ext}_{\mathit{\Lambda}}^{n}(i_{*}(M),i_{*}(M))\cong\mathrm{Ext}_{\mathit{\Lambda}^{\prime}}^{n}(M,i^{!}(i_{*}(M)))=0$

holds for all $n\geqslant 1$. Notice that $i^{!}i_{*}=1_{\mathsf{mod}(\mathit{\Lambda}^{\prime})}$ (see Lemma 2.3 (3)), then the above follows that

$\mathrm{Ext}_{\mathit{\Lambda}}^{n}(i_{*}(M),i_{*}(M))\cong\mathrm{Ext}_{\mathit{\Lambda}^{\prime}}^{n}(M,M)=0$

holds for all $n\geqslant 1$. One can check that $\mathrm{Ext}_{\mathit{\Lambda}}^{n}(j_{!}(N),j_{!}(N))\cong\mathrm{Ext}_{\mathit{\Lambda}^{\prime\prime}}^{n}(N,N)=0$ holds for all $n\geqslant 1$ by a similar way. ∎

Lemma 3.8.

Let $M$ and $N$ be exceptional modules in $\mathsf{mod}(\mathit{\Lambda}^{\prime})$ and $\mathsf{mod}(\mathit{\Lambda}^{\prime\prime})$, respectively. If $i^{*}$ and $i^{!}$ are exact functors, then $i_{*}(M)$ and $j_{!}(N)$ are exceptional right $\mathit{\Lambda}$-modules in $\mathsf{mod}(\mathit{\Lambda})$.

Proof.

We need show that (E1) and (E2) hold for $i_{*}(M)$ and $j_{!}(N)$ in this proof.

First of all, by (R2), $i_{*}$ is fully faithful, then we have the following isomorphism

$$\mathrm{End}_{\mathit{\Lambda}}(i_{*}(M))\cong\mathrm{End}_{\mathit{\Lambda}^{\prime}}(M)\cong\mathbb{K}$$

since $M$ is exceptional in $\mathsf{mod}(\mathit{\Lambda}^{\prime})$. Thus, (E1) holds for $i_{*}(M)$. Similarly, one can check that (E1) holds for $j_{!}(N)$ by using $j_{!}$ being fully faithful and $N$ being exceptional in $\mathsf{mod}(\mathit{\Lambda}^{\prime\prime})$.

Second, for any $n\geqslant 1$, we have

$\mathrm{Ext}_{\mathit{\Lambda}^{\prime}}^{n}(M,M)\cong\mathrm{Ext}_{\mathit{\Lambda}}^{n}(i_{*}(M),i_{*}(M))=0$

by Lemma 3.7, then (E2) hold for $i_{*}(M)$. One can check that (E2) holds for $j_{!}(N)$ by a similar way. ∎

Lemma 3.9.

Let $(M_{1},M_{2})$ and $(N_{1},N_{2})$ be exceptional pairs in $\mathsf{mod}(\mathit{\Lambda}^{\prime})$ and $\mathsf{mod}(\mathit{\Lambda}^{\prime\prime})$, respectively. If $i^{*}$ and $i^{!}$ are exact functors, then $(i_{*}(M_{1}),i_{*}(M_{2}))$ and $(j_{!}(N_{1}),j_{!}(N_{2}))$ are exceptional pairs in $\mathsf{mod}(\mathit{\Lambda})$.

Proof.

since modules $M_{1}$ and $M_{2}$ are exceptional in $\mathsf{mod}(\mathit{\Lambda}^{\prime})$ and $i^{*}$ and $i^{!}$ are exact, then $i_{*}(M_{1})$ and $i_{*}(M_{2})$ are exceptional in $\mathsf{mod}(\mathit{\Lambda})$ by Lemma 3.8. By Lemma 2.3 (3), $i_{*}$ is fully faithful, then

$$\mathrm{Hom}_{\mathit{\Lambda}}(i_{*}(M_{2}),i_{*}(M_{1}))\cong\mathrm{Hom}_{\mathit{\Lambda}^{\prime}}(M_{2},M_{1})=0,$$

i.e, (E1^′) holds for $(i_{*}(M_{1}),i_{*}(M_{2}))$. Moreover, for any $n\geqslant 1$, we have

$$\mathrm{Ext}_{\mathit{\Lambda}}^{n}(i_{*}(M_{2}),i_{*}(M_{1}))\mathop{\cong}\limits^{\spadesuit}\mathrm{Ext}_{\mathit{\Lambda}}^{n}(M_{2},i^{!}(i_{*}(M_{1})))\mathop{\cong}\limits^{\clubsuit}\mathrm{Ext}_{\mathit{\Lambda}}^{n}(M_{2},M_{1})\mathop{=}\limits^{{\color[rgb]{1,0,0}\definecolor[named]{pgfstrokecolor}{rgb}{1,0,0}\boldsymbol{\heartsuit}}}0$$

by using Lemmas 3.6 and 3.1 (see the isomorphism marked by $\spadesuit$), Lemma 2.3 (3) (see the isomorphism marked by $\clubsuit$), and $(M_{1},M_{2})$ being an exceptional pair (see the equation marked by ${\color[rgb]{1,0,0}\definecolor[named]{pgfstrokecolor}{rgb}{1,0,0}\boldsymbol{\heartsuit}}$). It follows that (E2^′) holds for $(i_{*}(M_{1}),i_{*}(M_{2}))$. Therefore, $(i_{*}(M_{1}),i_{*}(M_{2}))$ is an exceptional pairs in $\mathsf{mod}(\mathit{\Lambda})$.

One can prove that $(j_{!}(N_{1}),j_{!}(N_{2}))$ is an exceptional pair in $\mathsf{mod}(\mathit{\Lambda})$ by a similar way. ∎

Theorem 3.10.

Let $(X_{1},\ldots,X_{s})$ and $(Y_{1},\ldots,Y_{t})$ be two exceptional sequences in $\mathsf{mod}(\mathit{\Lambda}^{\prime})$ and $\mathsf{mod}(\mathit{\Lambda}^{\prime\prime})$, respectively. If $i^{*}$ and $i^{!}$ are exact functors, then $(i_{*}(X_{1}),\ldots,i_{*}(X_{s}))$ and $(j_{!}(Y_{1}),\ldots,j_{!}(Y_{t}))$ are exceptional sequences in $\mathsf{mod}(\mathit{\Lambda})$.

Proof.

All modules $X_{1}$, $\ldots$, $X_{s}$ are exceptional in $\mathsf{mod}(\mathit{\Lambda}^{\prime})$ by the definition of exceptional sequence, then Lemma 3.8 admits that $i_{*}(X_{1})$, $\ldots$, $i_{*}(X_{s})$ are exceptional in $\mathsf{mod}(\mathit{\Lambda})$. Moreover, for each $1\leqslant i<j\leqslant s$, we have that $(X_{j},X_{i})$ is an exceptional pair in $\mathsf{mod}(\mathit{\Lambda}^{\prime})$ then, by the definition of exceptional sequence, $(i_{*}(X_{j}),i_{*}(X_{i}))$ is also an exceptional pair in $\mathsf{mod}(\mathit{\Lambda})$ by using Lemma 3.9. Thus, $(i_{*}(X_{1}),\ldots,i_{*}(X_{s}))$ is an exceptional sequence in $\mathsf{mod}(\mathit{\Lambda})$.

One can prove that $(j_{!}(Y_{1}),\ldots,j_{!}(Y_{t}))$ is an exceptional sequence in $\mathsf{mod}(\mathit{\Lambda})$ by a similar way. ∎

4 Examples

4.1 Two examples for Corollary 3.2

First of all, we provide two examples for Corollary 3.2, see Examples 4.1 and 4.2. In Example 4.1, the instance illustrates that the condition “${{}_{A}}R$ is projective” is indispensable. In Example 4.2, we provide an instance for projective split extension of a finite-dimensional algebra $A$ by the nilpotent $(A,A)$-bimodule $Q$.

Example 4.1.

Let $A=\mathbb{K}\mathcal{Q}_{A}/\mathcal{I}_{A}$ be a finite-dimensional algebra given by

$$\mathcal{Q}_{A}=\lx@xy@svg{\hbox{\raise 0.0pt\hbox{\kern 5.5pt\hbox{\ignorespaces\ignorespaces\ignorespaces\hbox{\vtop{\kern 0.0pt\offinterlineskip\halign{\entry@#!@&&\entry@@#!@\cr&&\crcr}}}\ignorespaces{\hbox{\kern-5.5pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\raise 0.0pt\hbox{$\textstyle{1\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\lx@xy@droprule}}\ignorespaces\ignorespaces\ignorespaces{\hbox{\kern 12.26105pt\raise-4.50694pt\hbox{{}\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\hbox{\hbox{\kern 0.0pt\raise-1.50694pt\hbox{$\scriptstyle{\alpha}$}}}\kern 3.0pt}}}}}}\ignorespaces{\hbox{\kern 29.5pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1}\lx@xy@tip{-1}}}}}}{\hbox{\lx@xy@droprule}}{\hbox{\lx@xy@droprule}}{\hbox{\kern 29.5pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\raise 0.0pt\hbox{$\textstyle{2\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\lx@xy@droprule}}\ignorespaces\ignorespaces\ignorespaces{\hbox{\kern 47.52031pt\raise-6.1111pt\hbox{{}\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\hbox{\hbox{\kern 0.0pt\raise-1.75pt\hbox{$\scriptstyle{\beta}$}}}\kern 3.0pt}}}}}}\ignorespaces{\hbox{\kern 64.5pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1}\lx@xy@tip{-1}}}}}}{\hbox{\lx@xy@droprule}}{\hbox{\lx@xy@droprule}}{\hbox{\kern 64.5pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\raise 0.0pt\hbox{$\textstyle{3}$}}}}}}}\ignorespaces}}}}\ignorespaces~{}\text{and}~{}\mathcal{I}_{A}=\langle\alpha\beta\rangle \textstyle{1\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{\alpha} \textstyle{2\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{\beta} \textstyle{3} \textstyle{1\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{\alpha} \textstyle{2\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{\beta} \textstyle{3}$$

and $R=\mathbb{K}\mathcal{Q}_{R}/\mathcal{I}_{R}$ be a finite-dimensional algebra given by

$$\mathcal{Q}_{R}=\lx@xy@svg{\hbox{\raise 0.0pt\hbox{\kern 5.5pt\hbox{\ignorespaces\ignorespaces\ignorespaces\hbox{\vtop{\kern 0.0pt\offinterlineskip\halign{\entry@#!@&&\entry@@#!@\cr&&\crcr}}}\ignorespaces{\hbox{\kern-5.5pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\raise 0.0pt\hbox{$\textstyle{1\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\lx@xy@droprule}}\ignorespaces\ignorespaces\ignorespaces{\hbox{\kern 12.26105pt\raise-4.50694pt\hbox{{}\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\hbox{\hbox{\kern 0.0pt\raise-1.50694pt\hbox{$\scriptstyle{\alpha}$}}}\kern 3.0pt}}}}}}\ignorespaces{\hbox{\kern 29.5pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1}\lx@xy@tip{-1}}}}}}{\hbox{\lx@xy@droprule}}{\hbox{\lx@xy@droprule}}{\hbox{\kern 29.5pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\raise 0.0pt\hbox{$\textstyle{2\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\lx@xy@droprule}}\ignorespaces\ignorespaces\ignorespaces{\hbox{\kern 47.52031pt\raise-6.1111pt\hbox{{}\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\hbox{\hbox{\kern 0.0pt\raise-1.75pt\hbox{$\scriptstyle{\beta}$}}}\kern 3.0pt}}}}}}\ignorespaces{\hbox{\kern 64.5pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1}\lx@xy@tip{-1}}}}}}{\hbox{\lx@xy@droprule}}{\hbox{\lx@xy@droprule}}{\hbox{\kern 64.5pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\raise 0.0pt\hbox{$\textstyle{3\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{}{}{{}{}{}{}{}{}{}{}{}{}{}{}{}{}}{}\ignorespaces\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{{}{}{{}}{{}{}{}}{}}}}\ignorespaces{}\ignorespaces{}{}{}{{}{}}\ignorespaces\ignorespaces{\hbox{\kern 30.18794pt\raise 26.78754pt\hbox{{}\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\hbox{\hbox{\kern 0.0pt\raise-0.8264pt\hbox{$\scriptstyle{\gamma}$}}}\kern 3.0pt}}}}}}\ignorespaces{}{}{}{{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}{}}{\hbox{\kern 5.50293pt\raise 6.25827pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1}\lx@xy@tip{-1}}}}}}\ignorespaces\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{{}{}{}{{}}{{}{}{}\lx@xy@spline@}{}}}}\ignorespaces{}\ignorespaces\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{{}{}{{}}{{}{}{}}{}}}}\ignorespaces{}\ignorespaces}}}}\ignorespaces~{}\text{and}~{}\mathcal{I}_{R}=\langle\alpha\beta,\beta\gamma,\gamma\alpha\rangle \textstyle{1\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{\alpha} \textstyle{2\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{\beta} \textstyle{3\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{\gamma} \textstyle{1\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{\alpha} \textstyle{2\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{\beta} \textstyle{3\ignorespaces\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{\gamma}$$

Then $Q=\langle\gamma\rangle$ is an $(A,A)$-bimodule given by the left $A$-action

$$A\times Q\to Q,~{}(\wp,k\gamma)\mapsto\wp\gamma=\begin{cases}\gamma,&\text{if}~{}\wp=\varepsilon_{3};\\ 0,&\text{if}~{}\wp~{}\text{is another path}\end{cases}$$

(and $\varepsilon_{v}$ is the path of length zero corresponded by the vertex $v$)

and the right $A$-action

$$Q\times A\to Q,~{}(\gamma,\wp)\mapsto\gamma\wp=\begin{cases}\gamma,&\text{if}~{}\wp=\varepsilon_{1};\\ 0,&\text{if}~{}\wp~{}\text{is another path}.\end{cases}$$

By $\mathcal{Q}=\mathbb{K}\gamma\subseteq\mathrm{rad}R=\mathbb{K}\alpha+\mathbb{K}\beta+\mathbb{K}\gamma$, $R$ is a split extension of $A$ by the nilpotent bimodule $Q$. Consider the projective resolution

$$\mathbf{P}(M)=\ \ 0\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits(3)_{A}\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits({{}^{2}_{3}})_{A}\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits({{}^{1}_{2}})_{A}\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits(1)_{A}\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits{}0$$

of $M=(1)_{A}\in\mathsf{mod}(A)$. By $R=\mathbb{K}\varepsilon_{1}\oplus_{\mathbb{K}}\mathbb{K}\varepsilon_{2}\oplus_{\mathbb{K}}\mathbb{K}\varepsilon_{3}\oplus_{\mathbb{K}}\mathbb{K}\alpha\oplus_{\mathbb{K}}\mathbb{K}\beta\oplus_{\mathbb{K}}\mathbb{K}\gamma+\mathcal{I}_{R}$, we have

$$(1)_{A}\otimes_{A}R\cong\ (\mathbb{K}\varepsilon_{1}+\langle\alpha,\alpha\beta\rangle)\otimes_{A}R$$

$$\cong\ \mathbb{K}\varepsilon_{1}+\langle\alpha,\alpha\beta,\beta\gamma,\gamma\alpha\rangle$$

$$\cong\ (1)_{R},$$

$$({{}^{1}_{2}})_{A}\otimes_{A}R\cong\ (\mathbb{K}\varepsilon_{1}\oplus_{\mathbb{K}}\mathbb{K}\alpha+\langle\alpha\beta\rangle)\otimes_{A}R$$

$$\cong\ \mathbb{K}\varepsilon_{1}\oplus_{\mathbb{K}}\mathbb{K}\alpha+\langle\alpha\beta,\beta\gamma,\gamma\alpha\rangle$$

$$\cong\ ({{}^{1}_{2}})_{R},$$

$$({{}^{2}_{3}})_{A}\otimes_{A}R\cong\ (\mathbb{K}\varepsilon_{2}\oplus_{\mathbb{K}}\mathbb{K}\beta)\otimes_{A}R$$

$$\cong\ \mathbb{K}\varepsilon_{2}\oplus_{\mathbb{K}}\mathbb{K}\beta+\langle\beta\gamma\rangle$$

$$\cong\ ({{}^{2}_{3}})_{R},$$

$$(3)_{A}\otimes_{A}R\cong\ \mathbb{K}\varepsilon_{3}\otimes_{A}R$$

$$\cong\ \mathbb{K}\varepsilon_{3}\oplus_{\mathbb{K}}\mathbb{K}\gamma+\langle\gamma\alpha\rangle$$

$$\cong\ ({{}^{3}_{1}})_{R},$$

then the sequence obtained by applying $F:=-\otimes_{A}R$ to the projective resolution of $M$ is

$$\mathbf{P}(M)\otimes_{A}R=\ \ 0\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits({{}^{3}_{1}})_{R}\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits({{}^{2}_{3}})_{R}\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits({{}^{1}_{2}})_{R}\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits(1)_{R}\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits 0$$

which is not a projective resolution of $F(M)=M\otimes_{A}R=(1)_{R}$ since it is not left exact (to be precise, $R=A\oplus Q$, as a right $A$-module, is not projective). Indeed, one can check that the projective resolution of $(1)_{R}$ is

$$\mathbf{P}(M\otimes_{A}R)=\ \ \cdots\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits({{}^{2}_{3}})_{R}\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits({{}^{1}_{2}})_{R}\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits({{}^{3}_{1}})_{R}\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits({{}^{2}_{3}})_{R}\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits({{}^{1}_{2}})_{R}\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits(1)_{R}\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits 0$$

Applying $\mathrm{Hom}_{R}(-,M\otimes_{A}R)$ to $\mathbf{P}(M\otimes_{A}R)$, we get the following complex

$$\mathrm{Hom}_{R} (\mathbf{P}(M\otimes_{A}R),M\otimes_{A}R)=$$

$$0 \mathop{-\!\!\!-\!\!\!\longrightarrow}\limits\mathrm{Hom}_{R}((1)_{R},(1)_{R})\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits\mathrm{Hom}_{R}(({{}^{1}_{2}})_{R},(1)_{R})$$

$$\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits\mathrm{Hom}_{R}(({{}^{2}_{3}})_{R},(1)_{R})=0\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits\mathrm{Hom}_{R}(({{}^{3}_{1}})_{R},(1)_{R})=0\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits\mathrm{Hom}_{R}(({{}^{1}_{3}})_{R},(1)_{R})$$

$$\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits\mathrm{Hom}_{R}(({{}^{2}_{3}})_{R},(1)_{R})=0\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits\mathrm{Hom}_{R}(({{}^{3}_{1}})_{R},(1)_{R})=0\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits\mathrm{Hom}_{R}(({{}^{1}_{3}})_{R},(1)_{R})$$

$$\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits\cdots.$$

It follows that

$$\mathrm{Ext}_{R}^{3}(M\otimes_{A}R,M\otimes_{A}R)\cong\mathrm{Hom}_{R}(({{}^{1}_{3}})_{R},(1)_{R})\cong_{\mathbb{K}}\mathbb{K}\neq 0,$$

where “$\cong_{\mathbb{K}}$” represents the isomorphism of $\mathbb{K}$-vector spaces.

On the other hand, $G=\mathrm{Hom}_{R}(R,-)$ is the functor such that $(-\otimes_{A}R,\mathrm{Hom}_{R}(R,-))$ is an adjoint pair, and we have $\mathrm{Hom}_{R}(R,M\otimes_{A}R)=(1)_{A}$. Applying $\mathrm{Hom}_{R}(-,\mathrm{Hom}_{R}(R,M\otimes_{A}R))=\mathrm{Hom}_{R}(-,(1)_{A})$ to $\mathbf{P}(M)$, we have

$$\mathrm{Hom}_{R} (\mathbf{P}(M),\mathrm{Hom}_{R}(R,M\otimes_{A}R))=\mathrm{Hom}_{R}(\mathbf{P}(M),(1)_{A})=$$

$$0 \mathop{-\!\!\!-\!\!\!\longrightarrow}\limits\mathrm{Hom}_{R}((1)_{A},(1)_{A})\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits\mathrm{Hom}_{R}(({{}^{1}_{2}})_{A},(1)_{A})$$

$$\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits\mathrm{Hom}_{R}(({{}^{2}_{3}})_{A},(1)_{A})=0\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits\mathrm{Hom}_{R}((3)_{A},(1)_{A})=0\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits 0$$

$$\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits 0\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits 0\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits 0\mathop{-\!\!\!-\!\!\!\longrightarrow}\limits\cdots.$$

It follows that

$$\mathrm{Ext}_{A}^{3}(M,\mathrm{Hom}_{R}(R,M\otimes_{A}R))\cong\mathrm{Ext}_{A}^{3}((1)_{A},(1)_{A})=0.$$

Then we have

$$0\neq\mathrm{Ext}_{R}^{3}(M\otimes_{A}R,M\otimes_{A}R)\not\cong\mathrm{Ext}_{A}^{3}(M,\mathrm{Hom}_{R}(R,M\otimes_{A}R))=0$$

in this example.

Example 4.2.

Let $A=\Bbbk\mathcal{Q}_{A}/\mathcal{I}_{A}$ be the finite dimensional algebra given in Example 4.1 and $R^{\prime}=\Bbbk\mathcal{Q}_{R}^{\prime}/\mathcal{I}_{R}^{\prime}$ be the finite dimensional algebra whose quiver $\mathcal{Q}_{R}^{\prime}$ is the quiver $\mathcal{Q}_{R}$ of the algebra $R$ given in Example 4.1 and $\mathcal{I}_{R}^{\prime}=\langle\alpha\beta\rangle$. Then the epimorphism

$h:R^{\prime}\to R^{\prime}/\langle\gamma\rangle\cong A$, $\wp+\mathcal{I}_{R}^{\prime}\mapsto(\wp+\mathcal{I}_{R}^{\prime})+\langle\gamma+\mathcal{I}_{R}^{\prime}\rangle$

(note that “$+\mathcal{I}_{R}$” can be ignore for simplicity)

is a split extension of $A$ by the nilpotent bimodule $Q=\operatorname{Ker}(h)\cong\langle\gamma\rangle\cong\mathbb{K}\gamma$, where $Q$ is an $(A,A)$-bimodule given by the left $A$-action

$$A\times Q\to Q,~{}(\wp,\gamma)\mapsto\wp\gamma=\begin{cases}\wp\gamma,&\text{if}~{}\wp\in\{\varepsilon_{3},\beta\};\\ 0,&\text{if}~{}\wp~{}\text{is another path}\end{cases}$$

and the right $A$-action

$$Q\times A\to Q,~{}(\gamma,\wp)\mapsto\gamma\wp=\begin{cases}\gamma\wp,&\text{if}~{}\wp\in\{\varepsilon_{1},\alpha\};\\ 0,&\text{if}~{}\wp~{}\text{is another path}.\end{cases}$$

Then the quiver representation of $Q$, as a left $A$-module, is

$$(\varepsilon_{v}Q,\varphi_{\alpha})_{v\in(\mathcal{Q}_{A})_{0},\alpha\in(\mathcal{Q}_{A})_{1}}=\ (\lx@xy@svg{\hbox{\raise 0.0pt\hbox{\kern 14.59804pt\hbox{\ignorespaces\ignorespaces\ignorespaces\hbox{\vtop{\kern 0.0pt\offinterlineskip\halign{\entry@#!@&&\entry@@#!@\cr&&\crcr}}}\ignorespaces{\hbox{\kern-14.59804pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\raise 0.0pt\hbox{$\textstyle{\varepsilon_{1}\cdot\mathbb{K}\gamma\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\lx@xy@droprule}}\ignorespaces\ignorespaces\ignorespaces{\hbox{\kern 21.35909pt\raise 4.50694pt\hbox{{}\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\hbox{\hbox{\kern 0.0pt\raise-1.50694pt\hbox{$\scriptstyle{\alpha}$}}}\kern 3.0pt}}}}}}\ignorespaces{\hbox{\kern 38.59804pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1}\lx@xy@tip{-1}}}}}}{\hbox{\lx@xy@droprule}}{\hbox{\lx@xy@droprule}}{\hbox{\kern 38.59804pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\raise 0.0pt\hbox{$\textstyle{\varepsilon_{2}\cdot\mathbb{K}\gamma\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\lx@xy@droprule}}\ignorespaces\ignorespaces\ignorespaces{\hbox{\kern 74.81442pt\raise 6.1111pt\hbox{{}\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\hbox{\hbox{\kern 0.0pt\raise-1.75pt\hbox{$\scriptstyle{\beta}$}}}\kern 3.0pt}}}}}}\ignorespaces{\hbox{\kern 91.79411pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1}\lx@xy@tip{-1}}}}}}{\hbox{\lx@xy@droprule}}{\hbox{\lx@xy@droprule}}{\hbox{\kern 91.79411pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\raise 0.0pt\hbox{$\textstyle{\varepsilon_{3}\cdot\mathbb{K}\gamma}$}}}}}}}\ignorespaces}}}}\ignorespaces) \textstyle{\varepsilon_{1}\cdot\mathbb{K}\gamma\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{\alpha} \textstyle{\varepsilon_{2}\cdot\mathbb{K}\gamma\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{\beta} \textstyle{\varepsilon_{3}\cdot\mathbb{K}\gamma} \textstyle{\varepsilon_{1}\cdot\mathbb{K}\gamma\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{\alpha} \textstyle{\varepsilon_{2}\cdot\mathbb{K}\gamma\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{\beta} \textstyle{\varepsilon_{3}\cdot\mathbb{K}\gamma}$$

$$\cong\ (\lx@xy@svg{\hbox{\raise 0.0pt\hbox{\kern 5.5pt\hbox{\ignorespaces\ignorespaces\ignorespaces\hbox{\vtop{\kern 0.0pt\offinterlineskip\halign{\entry@#!@&&\entry@@#!@\cr&&\crcr}}}\ignorespaces{\hbox{\kern-5.5pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\raise 0.0pt\hbox{$\textstyle{0\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\lx@xy@droprule}}\ignorespaces\ignorespaces\ignorespaces{\hbox{\kern 12.26105pt\raise 4.50694pt\hbox{{}\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\hbox{\hbox{\kern 0.0pt\raise-1.50694pt\hbox{$\scriptstyle{\alpha}$}}}\kern 3.0pt}}}}}}\ignorespaces{\hbox{\kern 29.5pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1}\lx@xy@tip{-1}}}}}}{\hbox{\lx@xy@droprule}}{\hbox{\lx@xy@droprule}}{\hbox{\kern 29.5pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\raise 0.0pt\hbox{$\textstyle{0\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\lx@xy@droprule}}\ignorespaces\ignorespaces\ignorespaces{\hbox{\kern 48.21475pt\raise 6.1111pt\hbox{{}\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\hbox{\hbox{\kern 0.0pt\raise-1.75pt\hbox{$\scriptstyle{\beta}$}}}\kern 3.0pt}}}}}}\ignorespaces{\hbox{\kern 64.5pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1}\lx@xy@tip{-1}}}}}}{\hbox{\lx@xy@droprule}}{\hbox{\lx@xy@droprule}}{\hbox{\kern 64.5pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\raise 0.0pt\hbox{$\textstyle{\mathbb{K}}$}}}}}}}\ignorespaces}}}}\ignorespaces). \textstyle{0\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{\alpha} \textstyle{0\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{\beta} \textstyle{\mathbb{K}} \textstyle{0\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{\alpha} \textstyle{0\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{\beta} \textstyle{\mathbb{K}}$$

Thus, ${{}_{A}Q}={{}_{A}(3)}$. It follows that ${{}_{A}R}\cong{{}_{A}({{}^{1}_{2}})}\oplus{{}_{A}({{}^{2}_{3}})}\oplus{{}_{A}(3)}^{\oplus 2}$ is a projective left $A$-module.

4.2 An examples for Theorem 3.5

Example 4.3.

Let $R=\mathbb{K}\mathcal{Q}$ be a finite-dimensional $\mathbb{K}$-algebra given by

$$\mathcal{Q}=\lx@xy@svg{\hbox{\raise 0.0pt\hbox{\kern 5.5pt\hbox{\ignorespaces\ignorespaces\ignorespaces\hbox{\vtop{\kern 0.0pt\offinterlineskip\halign{\entry@#!@&&\entry@@#!@\cr&&\crcr}}}\ignorespaces{\hbox{\kern-5.5pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\raise 0.0pt\hbox{$\textstyle{1\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\lx@xy@droprule}}\ignorespaces\ignorespaces\ignorespaces{\hbox{\kern 12.26105pt\raise 4.50694pt\hbox{{}\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\hbox{\hbox{\kern 0.0pt\raise-1.50694pt\hbox{$\scriptstyle{\alpha}$}}}\kern 3.0pt}}}}}}\ignorespaces{\hbox{\kern 29.5pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1}\lx@xy@tip{-1}}}}}}{\hbox{\lx@xy@droprule}}{\hbox{\lx@xy@droprule}}{\hbox{\kern 29.5pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\raise 0.0pt\hbox{$\textstyle{2\ignorespaces\ignorespaces\ignorespaces\ignorespaces}$}}}}}}}\ignorespaces\ignorespaces\ignorespaces\ignorespaces{}{\hbox{\lx@xy@droprule}}\ignorespaces\ignorespaces\ignorespaces{\hbox{\kern 47.52031pt\raise 6.1111pt\hbox{{}\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\hbox{\hbox{\kern 0.0pt\raise-1.75pt\hbox{$\scriptstyle{\beta}$}}}\kern 3.0pt}}}}}}\ignorespaces{\hbox{\kern 64.5pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\lx@xy@tip{1}\lx@xy@tip{-1}}}}}}{\hbox{\lx@xy@droprule}}{\hbox{\lx@xy@droprule}}{\hbox{\kern 64.5pt\raise 0.0pt\hbox{\hbox{\kern 0.0pt\raise 0.0pt\hbox{\hbox{\kern 3.0pt\raise 0.0pt\hbox{$\textstyle{3}$}}}}}}}\ignorespaces}}}}\ignorespaces. \textstyle{1\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{\alpha} \textstyle{2\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{\beta} \textstyle{3} \textstyle{1\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{\alpha} \textstyle{2\ignorespaces\ignorespaces\ignorespaces\ignorespaces} \scriptstyle{\beta} \textstyle{3}$$

Take $A=R/\langle\alpha\rangle$ and $Q=\langle\alpha\rangle$, then $\xi:R\to A$ be a split extension of $A$ by the nilpotent bimodule $Q$. It is easy to check that

$$(1)_{A}\otimes_{A}R\cong\left(\begin{smallmatrix}1\\ 2\\ 3\end{smallmatrix}\right)_{R},~{}(\begin{smallmatrix}2\\ 3\end{smallmatrix})_{A}\otimes_{A}R\cong(\begin{smallmatrix}2\\ 3\end{smallmatrix})_{R},~{}(3)_{A}\otimes_{A}R\cong(3)_{R},~{}\text{and~{}}(2)_{A}\otimes_{A}R\cong(2)_{R}.$$

Moreover, we have

$$(1)_{A}\otimes_{A}Q_{A}\cong\left({{}^{2}_{3}}\right)_{A},~{}(\begin{smallmatrix}2\\ 3\end{smallmatrix})_{A}\otimes_{A}Q_{A}=0~{}(3)_{A}\otimes_{A}Q_{A}=0~{}\text{and~{}}(2)_{A}\otimes_{A}Q_{A}=0.$$

Note that ${{}_{A}Q}\cong{{}_{A}(1)}^{\oplus 2}$, which is a direct sum of two copies of an indecomposable projective left $A$-module corresponding to the vertex $1$ of the quiver of $A$, then we obtain that ${}_{A}R$, as a left $A$-module, is isomorphic to the projective module $A\oplus{{}_{A}Q}$.

One can check that there are 9 complete exceptional sequences (=CES for short) in $\mathsf{mod}(A)$ and 16 CESs in $\mathsf{mod}(R)$, see [22]. It is easy to check that $\mathrm{Ext}_{A}^{1}(-,(1)_{A}\otimes_{A}Q)=0$, then, by Theorem 3.5, we have five CESs of $\mathsf{mod}(R)$ by applying $-\otimes_{A}R$ to the sequences (a), (d), (e), (f), (i) in Table 1.

	CESs of $\mathsf{mod}(A)$	$-\otimes{{}_{A}}R$
sequence (a):	$(\begin{smallmatrix}3\end{smallmatrix},\begin{smallmatrix}2\\ 3\end{smallmatrix},\begin{smallmatrix}1\end{smallmatrix})$	$\color[rgb]{1,0,0}\definecolor[named]{pgfstrokecolor}{rgb}{1,0,0}{(\begin{smallmatrix}3\end{smallmatrix},\begin{smallmatrix}2\\ 3\end{smallmatrix},\begin{smallmatrix}1\\ 2\\ 3\end{smallmatrix})}$
sequence (b):	$(\begin{smallmatrix}3\end{smallmatrix},\begin{smallmatrix}1\end{smallmatrix},\begin{smallmatrix}2\\ 3\end{smallmatrix})$	$(\begin{smallmatrix}3\end{smallmatrix},\begin{smallmatrix}1\\ 2\\ 3\end{smallmatrix},\begin{smallmatrix}2\\ 3\end{smallmatrix})$
sequence (c):	$(\begin{smallmatrix}1\end{smallmatrix},\begin{smallmatrix}3\end{smallmatrix},\begin{smallmatrix}2\\ 3\end{smallmatrix})$	$(\begin{smallmatrix}1\\ 2\\ 3\end{smallmatrix},\begin{smallmatrix}3\end{smallmatrix},\begin{smallmatrix}2\\ 3\end{smallmatrix})$
sequence (d):	$(\begin{smallmatrix}2\\ 3\end{smallmatrix},\begin{smallmatrix}2\end{smallmatrix},\begin{smallmatrix}1\end{smallmatrix})$	$(\begin{smallmatrix}2\\ 3\end{smallmatrix},\begin{smallmatrix}2\end{smallmatrix},\begin{smallmatrix}1\\ 2\\ 3\end{smallmatrix})$
sequence (e):	$(\begin{smallmatrix}2\\ 3\end{smallmatrix},\begin{smallmatrix}1\end{smallmatrix},\begin{smallmatrix}2\end{smallmatrix})$	$(\begin{smallmatrix}2\\ 3\end{smallmatrix},\begin{smallmatrix}1\\ 2\\ 3\end{smallmatrix},\begin{smallmatrix}2\end{smallmatrix})$
sequence (f):	$(\begin{smallmatrix}1\end{smallmatrix},\begin{smallmatrix}2\\ 3\end{smallmatrix},\begin{smallmatrix}2\end{smallmatrix})$	$(\begin{smallmatrix}1\\ 2\\ 3\end{smallmatrix},\begin{smallmatrix}2\\ 3\end{smallmatrix},\begin{smallmatrix}2\end{smallmatrix})$
sequence (g):	$(\begin{smallmatrix}1\end{smallmatrix},\begin{smallmatrix}2\end{smallmatrix},\begin{smallmatrix}3\end{smallmatrix})$	$(\begin{smallmatrix}1\\ 2\\ 3\end{smallmatrix},\begin{smallmatrix}2\end{smallmatrix},\begin{smallmatrix}3\end{smallmatrix})$
sequence (h):	$(\begin{smallmatrix}2\end{smallmatrix},\begin{smallmatrix}1\end{smallmatrix},\begin{smallmatrix}3\end{smallmatrix})$	$(\begin{smallmatrix}2\end{smallmatrix},\begin{smallmatrix}1\\ 2\\ 3\end{smallmatrix},\begin{smallmatrix}3\end{smallmatrix})$
sequence (i):	$(\begin{smallmatrix}2\end{smallmatrix},\begin{smallmatrix}3\end{smallmatrix},\begin{smallmatrix}1\end{smallmatrix})$	$\color[rgb]{1,0,0}\definecolor[named]{pgfstrokecolor}{rgb}{1,0,0}{(\begin{smallmatrix}2\end{smallmatrix},\begin{smallmatrix}3\end{smallmatrix},\begin{smallmatrix}1\\ 2\\ 3\end{smallmatrix})}$

Table 1: The table of CESs.

Authors’ contributions

D. Liu, H. Gao and Y.-Z. Liu contributed equally to this work.

Conflicts of Interest

Data Availability

The authors declares that no data was used in this article.

Declarations

The authors states that there is no Conflict of interest.

Funding

• Dajun Liu is supported by the National Natural Science Foundation of China (No. 12101003), and the Natural Science Foundation of Anhui province (No. 2108085QA07);

• Hanpeng Gao is supported by the National Natural Science Foundation of China (No. 12301041);

• Yu-Zhe Liu is supported by the National Natural Science Foundation of China (Grant No. 12401042, 12171207), Guizhou Provincial Basic Research Program (Natural Science) (Grant Nos. ZD[2025]085 and ZK[2024]YiBan066) and Scientific Research Foundation of Guizhou University (Grant Nos. [2022]53, [2022]65).