Abstract
Let Q be a finite acyclic valued quiver. We define a bialgebra structure and an integration map on the Hall algebra associated to the morphism category of projective representations of Q. As an application, we recover the surjective homomorphism defined in [12], which realizes the principal coefficient quantum cluster algebra \(\mathcal {A}\,_q(Q)\) as a sub-quotient of the Hall algebra of morphisms. Moreover, we also recover the quantum Caldero–Chapoton formula, as well as some multiplication formulas between quantum Caldero–Chapoton characters.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
1 Introduction
The Hall algebra of a finitary abelian category is defined to be an associative algebra with a basis indexed by the isomorphism classes of objects and with a multiplication which encodes information about extensions of objects (cf. [25, 30]). A typical example of such categories is provided by the category of finite dimensional representations of a finite acyclic quiver over a finite field. For a Dynkin quiver Q, Ringel [26] realized the positive part of the quantum group associated to Q via the so-called Ringel–Hall algebra of Q. Later on, for any hereditary algebra A, Green [17] introduced a bialgebra structure on the Ringel–Hall algebra of A, and he showed that the composition subalgebra generated by simple modules provides a realization of the positive part of the corresponding quantum group.
Cluster algebras were introduced by Fomin and Zelevinsky [14] with the aim to set up a combinatorial framework for the study of total positivity in algebraic groups and canonical bases in quantum groups. The quantum versions of cluster algebras, called the quantum cluster algebras, were later introduced by Berenstein and Zelevinsky [3]. Acyclic (quantum) cluster algebras associated with acyclic valued quivers are an important class of (quantum) cluster algebras, which has a close relation with the representation theory of acyclic valued quivers, see [4,5,6,7, 23, 27, 28] for instance. Among others, the combinatorial structure of an acyclic cluster algebra \(\mathcal {A}(Q)\) can be categorified by the combinatorial structure of the category of representations of Q (cf. [4]). Caldero and Chapoton [5] proved that the cluster algebra \(\mathcal {A}(Q)\) associated with a Dynkin quiver Q can be recovered from the data of the corresponding quiver representation category. Caldero and Keller [6] further proved the cluster multiplication formulas for \(\mathcal {A}(Q)\).
Let Q be a finite acyclic valued quiver and \(\mathcal {A}\) be the category of finite dimensional representations of Q. The work of Caldero and Keller [6] suggested that there should be a deep connection between the quantum cluster algebra \(\mathcal {A}_q(Q)\) (with coefficients) and the dual Hall algebra \(\mathcal {H}(\mathcal {A})\). In [11], Ding and Xu showed that the comultiplication of \(\mathcal {H}(\mathcal {A})\) implies the cluster multiplication formulas in \(\mathcal {A}_q(Q)\) (cf. also [13]). The expected connection has been established by Berenstein and Rupel [2]. In particular, Berenstein and Rupel [2] proved that there is a homomorphism of algebras from \(\mathcal {H}(\mathcal {A})\) to the quantum torus associated with the quantum cluster algebra \(\mathcal {A}_q(Q)\), which maps the indecomposable rigid modules to the non-initial quantum cluster variables of \(\mathcal {A}_q(Q)\). Such a homomorphism is called a quantum cluster character of the Hall algebra. The homomorphism and its generalization have been further investigated in [8, 16, 29]. Let us emphasize that the bialgebra structure of \(\mathcal {H}(\mathcal {A})\) has also played a central role in [2, 8, 13, 16].
In general, the initial quantum cluster variables do not belong to the image of a quantum cluster character of the Hall algebra \(\mathcal {H}(\mathcal {A})\). In order to overcome this shortcoming, Ding, Xu and Zhang [12] introduced the (localized) Hall algebra \(\mathcal{M}\mathcal{H}(\mathcal {A})\) of morphisms associated to the representation category of Q, and established a homomorphism of algebras from a twisted version \(\mathcal{M}\mathcal{H}_\Lambda (\mathcal {A})\) of \(\mathcal{M}\mathcal{H}(\mathcal {A})\) to the quantum torus associated with the quantum cluster algebra \(\mathcal {A}_q(Q)\) with principal coefficients. Along this way, the quantum cluster algebra \(\mathcal {A}_q(Q)\) has been realized as a subquotient of the Hall algebra \(\mathcal{M}\mathcal{H}_\Lambda (\mathcal {A})\).
The aim of this paper is to pursue a bialgebra approach to understand the quantum cluster character introduced in [12]. To do this, we introduce a comultiplication \(\Delta \) for \(\mathcal{M}\mathcal{H}(\mathcal {A})\) and prove that \(\mathcal{M}\mathcal{H}(\mathcal {A})\) is a bialgebra without counit. We also introduce an integration map \(\int \) from \(\mathcal{M}\mathcal{H}(\mathcal {A})\) to a suitable torus. Our main result yields an explicit and simple factorization of the quantum cluster character \(\Psi \) in [12] via the comultiplication \(\Delta \) and integration map \(\int \). This gives an alternative way to recover the quantum cluster character \(\Psi \). Compared with [12], we obtain the quantum cluster character \(\Psi \) without using cluster multiplication formulas, on the contrary, cluster multiplication formulas (cf. [10,11,12,13]) are the outcome of our main result. Moreover, we also recover the quantum Caldero–Chapoton formula, and present its intrinsic interpretation via the Hall algebra of morphisms. As an appendix, we also show how to apply a certain subalgebra of the derived Hall algebra of the bounded derived category \(\mathcal {D}^b(\mathcal {A})\) to realize the whole acyclic quantum cluster algebras. By the way, the three-step homomorphism approach used in this paper has been further developed in [9] to deduce the quantum version of the cluster multiplication formulas in the classical cluster algebra proved by Caldero-Keller [6] for finite type, Hubery [19] for affine type and Xiao-Xu [34] for acyclic quivers.
The paper is organized as follows: In Section 2 we recollect the definitions and properties of the morphism category and the construction of the Hall algebra of morphisms. We introduce a bialgebra structure for \(\mathcal{M}\mathcal{H}(\mathcal {A})\) in Section 3 and an integration map in Section 4. After certain computations involving compatible pairs in Section 5, we prove the main result in Section 6. As a byproduct, we give a categorical interpretation of g-vectors via the morphism category in Section 7. In Section 8, we generalize the idea of Section 6 to show that a certain subalgebra of the derived Hall algebra can also be used to realize acyclic quantum cluster algebras.
Let us fix some notations used throughout the paper. For a finite set S, we denote by |S| its cardinality. Let \(k=\mathbb {F}_q\) be a finite field with q elements, and set \(v=\sqrt{q}\). Let \(\mathbb {Z}[v,v^{-1}]\) be the ring of integral Laurent polynomials. Let A be a finite dimensional hereditary k-algebra, and denote by \(\mathcal {A}\,\) the category of finite dimensional left A-modules; let \(\mathscr {P}=\mathscr {P}_{\mathcal {A}\,}\subset \mathcal {A}\,\) be the subcategory consisting of projective objects. For an essentially small exact or triangulated category \(\mathcal {E}\), the Grothendieck group of \(\mathcal {E}\) and the set of isomorphism classes [X] of objects in \(\mathcal {E}\) are denoted by \(K(\mathcal {E})\) and \(\hbox {Iso}\,(\mathcal {E})\), respectively; let \(\hbox {Ind}\,(\mathcal {E})\) be a complete set of indecomposable objects in \(\mathcal {E}\). For each object M in \(\mathcal {E}\), its automorphism group and image in \(K(\mathcal {E})\) are denoted by \(\hbox {Aut}\,(M)\) and \(\hat{M}\), respectively; we set \(a_M=|\hbox {Aut}\,(M)|\) and denote by aM the direct sum of a copies of M for a positive integer a. For a module \(M\in \mathcal {A}\,\), we also use \({{\textbf {dim}}}\,M\) to denote its dimension vector. We always assume that all the vectors are column vectors, and all tensor products are taken over \(\mathbb {Z}[v,v^{-1}]\).
2 Preliminaries
2.1 Hall algebras
Given objects \(L,M,N \in \mathcal {A}\), let \(\hbox {Ext}\,_\mathcal {A}^1(M,N)_L \subset \hbox {Ext}\,_\mathcal {A}^1(M,N)\) be the subset consisting of those equivalence classes of short exact sequences with middle term isomorphic to L.
Definition 2.1
The Hall algebra \(\mathcal {H}(\mathcal {A})\) of \(\mathcal {A}\) is the free \(\mathbb {Z}[v,v^{-1}]\)-module with basis elements \([M] \in \hbox {Iso}\,(\mathcal {A})\), and with the multiplication defined by
Remark 2.2
Given objects \(L,M,N\in \mathcal {A}\,\), set
By the Riedtmann–Peng formula [21, 24],
Thus, in terms of the rescaled generators \([[M]]=\frac{[M]}{a_M}\), the product takes the form
which is the definition used, for example, in [26, 30]. The associativity of Hall algebras amounts to the following identity
for any objects \(L,X,Y,Z\in \mathcal {A}\,\).
Given objects \(M,N \in \mathcal {A}\), one defines
which descends to give a bilinear form
called the Euler form of \(\mathcal {A}\). We also consider the symmetric Euler form
defined by \((\alpha ,\beta )=\langle \alpha ,\beta \rangle +\langle \beta ,\alpha \rangle \) for all \(\alpha ,\beta \in K(\mathcal {A})\).
The twisted Hall algebra \(\mathcal {H}\,_{\hbox {tw}\,}(\mathcal {A}\,)\) is the same module as \(\mathcal {H}\,(\mathcal {A}\,)\) but with the twisted multiplication defined by
2.2 Morphism categories
Let \(C_2(\mathcal {A}\,)\) be the category whose objects are the morphisms in \(\mathcal {A}\,\), and each morphism from to is a pair (u, v) of morphisms in \(\mathcal {A}\,\) such that the following diagram
is commutative. In what follows, we also write \(M_\bullet \) as a morphism Let \(C_2(\mathscr {P}\,)\subset C_2(\mathcal {A}\,)\) be the extension-closed subcategory consisting of the morphisms in \(\mathscr {P}\,\).
For each object \(P\in \mathscr {P}\,\), define two objects in \(C_2(\mathscr {P}\,)\)
For each object \(M\in \mathcal {A}\,\), fixing a minimal projective resolutionFootnote 1
by [12] we define an object in \(C_2(\mathscr {P}\,)\)
The indecomposable objects of \(C_2(\mathscr {P}\,)\) are characterized in the following
Lemma 2.3
[12, Proposition 2.4] Each object \(M_\bullet \) in \(C_2(\mathscr {P}\,)\) has a direct sum decomposition
for some \(P, Q\in \mathscr {P}\,\) and \(M\in \mathcal {A}\,\). Moreover, the objects P, Q and M are uniquely determined up to isomorphism.
Lemma 2.4
[1, Corollaries 3.1 and 3.2] The objects \(C_P\) and \(K_P\), where \(P\in \textrm{Ind}(\mathscr {P}\,)\), provide a complete set of indecomposable projective objects in \(C_2(\mathscr {P})\); and the objects \(Z_P\) and \(K_P\) provide a complete set of indecomposable injective objects. Moreover, all \(K_P\) are exactly all the indecomposable projective-injective objects.
For each \(M_\bullet \in C_2(\mathscr {P}\,)\), we have a projective resolution and an injective resolution of \(M_\bullet \) in the following
Lemma 2.5
[1, Proposition 3.2] For each \(M_\bullet \in C_2(\mathscr {P}\,)\), we have the following short exact sequences
By Lemma 2.5, we know that the global dimension of \(C_2(\mathscr {P}\,)\) is equal to one. Thus, for any objects \(M_\bullet ,N_\bullet \in C_2(\mathscr {P}\,)\), we define
It induces a bilinear form on the Grothendieck group of \(C_2(\mathscr {P}\,)\). We use the same notation as the Euler form of \(\mathcal {A}\,\), since this should not cause confusion by the context. By [12, Lemma 2.9], for any \(M_\bullet ,N_\bullet \in C_2(\mathscr {P}\,)\),
2.3 Hall algebras of morphisms
Let \(\mathcal {H}\,(C_2(\mathcal {A}\,))\) be the Hall algebra of \(C_2(\mathcal {A}\,)\) as defined in Definition 2.1. Let \(\mathcal {H}\,(C_2(\mathscr {P}\,))\) be the submodule of \(\mathcal {H}\,(C_2(\mathcal {A}\,))\) spanned by the isomorphism classes of objects in \(C_2(\mathscr {P}\,)\). Since \(C_2(\mathscr {P}\,)\) is closed under extensions, \(\mathcal {H}\,(C_2(\mathscr {P}\,))\) is a subalgebra of the Hall algebra \(\mathcal {H}\,(C_2(\mathcal {A}\,))\). Define \(\mathcal {H}\,_{\hbox {tw}\,}(C_2(\mathscr {P}\,))\) to be the same module as \(\mathcal {H}\,(C_2(\mathscr {P}\,))\), but with the twisted multiplication
For any \(M_\bullet \in C_2(\mathscr {P}\,)\) and \(P\in \mathscr {P}\,\), by [12, Lemma 3.1], we have that in \(\mathcal {H}\,_{\hbox {tw}\,}(C_2(\mathscr {P}\,))\)
Define the localized Hall algebra \(\mathcal {M}\mathcal {H}\,(\mathcal {A}\,)\) to be the localization of \(\mathcal {H}\,_{\hbox {tw}\,}(C_2(\mathscr {P}\,))\) with respect to all elements \([K_P]\). In symbols,
For each \(\alpha \in K(\mathcal {A}\,)\), by writing \(\alpha =\hat{P}-\hat{Q}\) for some objects \(P,Q\in \mathscr {P}\,\), we define
Then for any \(\alpha ,\beta \in K(\mathcal {A}\,)\) and \(M_\bullet \in C_2(\mathscr {P}\,)\),
For each \(M\in \mathcal {A}\,\), we define
By [12], \(\mathbb {X}_M\) does not depend on the minimality of projective resolutions of M. Given \(M\in \mathcal {A}\,\) and \(P\in \mathscr {P}\,\), we define
In particular,
Given objects B, M, P, Q in \(\mathcal {A}\,\), define
Theorem 2.6
[12, Theorem 5.2] The algebra \(\mathcal{M}\mathcal{H}(\mathcal {A}\,)\) is spanned by all \(K_{\alpha }*\mathbb {X}_{M\oplus P[1]}\) (with \(\alpha \in K(\mathcal {A}\,)\), \(M\in \mathcal {A}\,\) and \(P\in \mathscr {P}\,)\), which are subject to the following relations
for any \(\alpha ,\beta \in K(\mathcal {A}\,)\), \(M,N\in \mathcal {A}\,\) and \(P,Q\in \mathscr {P}\,\).
3 Bialgebra structure on the Hall algebra \(\mathcal {M}\mathcal {H}\,(\mathcal {A}\,)\) of morphisms
In this section, we define a structure of a bialgebra without counit on the Hall algebra \(\mathcal {M}\mathcal {H}\,(\mathcal {A}\,)\) of morphisms. That is, we define a comultiplication map \(\Delta \) on \(\mathcal {M}\mathcal {H}\,(\mathcal {A}\,)\) such that \(\Delta \) is an algebra homomorphism. Let us define the multiplication \(*\) on \(\mathcal {M}\mathcal {H}\,(\mathcal {A}\,)\otimes \mathcal {M}\mathcal {H}\,(\mathcal {A}\,)\) by
for any \(\alpha ,\beta ,\gamma ,\delta \in K(\mathcal {A}\,)\), \(M,N,U,V\in \mathcal {A}\,\) and \(P,Q,S,T\in \mathscr {P}\,\). We also define a homomorphism of \(\mathbb {Z}[v,v^{-1}]\)-modules
by
for any \(\alpha \in K(\mathcal {A}\,)\), \(L\in \mathcal {A}\,\) and \(P\in \mathscr {P}\,\).
Remark 3.1
Note that \(\mathcal {M}\mathcal {H}\,(\mathcal {A}\,)\) is a \(K(\mathcal {A}\,)\)-graded algebra by defining the degree of \(K_{\alpha }*\mathbb {X}_{M\oplus P[1]}\) to be \(\hat{M}-\hat{P}\) for any \(\alpha \in K(\mathcal {A}\,), M\in \mathcal {A}\,\) and \(P\in \mathscr {P}\,\). For any elements \(x_{ij}\in \mathcal {M}\mathcal {H}\,(\mathcal {A}\,)\) with the degree equal to \(\alpha _{ij}\in K(\mathcal {A}\,)\), where \(i=1,2,3\) and \(j=1,2\), it is easy to see that \((\alpha _{12},\alpha _{21})+\langle \alpha _{11},\alpha _{22}\rangle +(\alpha _{12}+\alpha _{22},\alpha _{31})+\langle \alpha _{11}+\alpha _{21},\alpha _{32}\rangle =(\alpha _{22},\alpha _{31})+\langle \alpha _{21},\alpha _{32}\rangle +(\alpha _{12},\alpha _{21}+\alpha _{31})+\langle \alpha _{11},\alpha _{22}+\alpha _{32}\rangle \), and then \([(x_{11}\otimes x_{12})*(x_{21}\otimes x_{22})]*(x_{31}\otimes x_{32})=(x_{11}\otimes x_{12})*[(x_{21}\otimes x_{22})*(x_{31}\otimes x_{32})]\), i.e., the twisted multiplication (3.1) on \(\mathcal {M}\mathcal {H}\,(\mathcal {A}\,)\otimes \mathcal {M}\mathcal {H}\,(\mathcal {A}\,)\) is also associative.
Proposition 3.2
The map \(\Delta :(\mathcal {M}\mathcal {H}\,(\mathcal {A}\,),*)\longrightarrow (\mathcal {M}\mathcal {H}\,(\mathcal {A}\,)\otimes \mathcal {M}\mathcal {H}\,(\mathcal {A}\,),*)\) is a homomorphism of algebras.
Proof
It suffices to prove that \(\Delta \) preserves all the relations in Theorem 2.6. We only prove the relations (2.18) and (2.20), since the other relations can be easily proved.
By Theorem 2.6 (see also [12, Theorem 3.5]), the map
is an embedding of algebras. It follows from Green’s formula (cf. [17]) that the map
defined by
is a homomorphism of algebras, when the multiplication on \(\mathcal {H}\,_{\hbox {tw}\,}(\mathcal {A}\,)\otimes \mathcal {H}\,_{\hbox {tw}\,}(\mathcal {A}\,)\) is given by
Here we remark that the twisted multiplication (3.4) is also associative by considering the restriction of the twisted multiplication (3.1) to \(\mathcal {H}\,_{\hbox {tw}\,}(\mathcal {A}\,)\otimes \mathcal {H}\,_{\hbox {tw}\,}(\mathcal {A}\,)\). By the commutative diagram
we obtain that \(\Delta \) preserves the relation (2.18).
Next we prove that \(\Delta \) preserves the relation (2.20). On the one hand,
On the other hand,
Thus,
\(\square \)
Remark 3.3
(1) Let \(m: (\mathcal {M}\mathcal {H}\,(\mathcal {A}\,)\otimes \mathcal {M}\mathcal {H}\,(\mathcal {A}\,),*)\longrightarrow (\mathcal {M}\mathcal {H}\,(\mathcal {A}\,),*)\) be the multiplication map. The Hall algebra \((\mathcal {M}\mathcal {H}\,(\mathcal {A}\,),m,\Delta )\) is a bialgebra without counit. In fact, for any \(\mathbb {Z}[v,v^{-1}]\)-module homomorphism \(\epsilon :\mathcal {M}\mathcal {H}\,(\mathcal {A}\,)\rightarrow \mathbb {Z}[v,v^{-1}]\), \(m\circ (\epsilon \otimes \hbox {id}\,)\circ \Delta (\mathbb {X}_{P[1]})=\epsilon (1)\mathbb {X}_{P[1]}\), while \(m\circ (\hbox {id}\,\otimes \epsilon )\circ \Delta (\mathbb {X}_{P[1]})=\epsilon (\mathbb {X}_{P[1]})1\). Clearly, \(m\circ (\epsilon \otimes \hbox {id}\,)\circ \Delta (\mathbb {X}_{P[1]})\ne m\circ (\hbox {id}\,\otimes \epsilon )\circ \Delta (\mathbb {X}_{P[1]})\).
(2) The comultiplication defined in (3.2) is not the intrinsic comultiplication (cf. [17, 18]) of the Hall algebra of an exact category.
4 Integration map on the Hall algebra \(\mathcal {M}\mathcal {H}\,(\mathcal {A}\,)\) of morphisms
In this section, we define an integration map on the Hall algebra \(\mathcal {M}\mathcal {H}\,(\mathcal {A}\,)\) of morphisms. First of all, we give a characterization on the Grothendieck group \(K(C_2(\mathscr {P}\,))\) of the morphism category \(C_2(\mathscr {P}\,)\). Assume that the rank of \(K(\mathcal {A}\,)\) is equal to n. Using the basis consisting of simple A-modules, we identify \(K(\mathcal {A}\,)\) with \(\mathbb {Z}^n\) by sending \(\hat{M}\) to \({{\textbf {dim}}}\,M\) for any \(M\in \mathcal {A}\,\).
Given an object , we define
For any short exact sequence
in \(C_2(\mathscr {P}\,)\), it is easy to see that \({{\textbf {dim}}}\,L_\bullet ={{\textbf {dim}}}\,M_\bullet +{{\textbf {dim}}}\,N_\bullet \).
Let \(P_1,P_2,\cdots ,P_n\) be all indecomposable projective objects in \(\mathcal {A}\,\) up to isomorphism.
Lemma 4.1
The Grothendieck group \(K(C_2(\mathscr {P}\,))\) is a free abelian group having as a basis the set
and there exists a unique group isomorphism \(f: K(C_2(\mathscr {P}\,))\longrightarrow \mathbb {Z}^{2n}\) such that \(f(\hat{M_\bullet })={{\textbf {dim}}}\,M_\bullet \) for each object \(M_\bullet \) in \(C_2(\mathscr {P}\,)\).
Proof
For any object \(M_\bullet \in C_2(\mathscr {P}\,)\) with \(M_0=\oplus _{i=1}^na_iP_i\) and \(M_{-1}=\oplus _{i=1}^nb_iP_i\), where \(a_i, b_i\in \mathbb {N}\), \(1\le i\le n\), by Lemma 2.5, taking an injective resolution
we obtain that in \(K(C_2(\mathscr {P}\,))\)
This shows that \(\{\hat{Z}_{P_i},\hat{ K}_{P_i}~|~1\le i\le n\}\) generates the group \(K(C_2(\mathscr {P}\,))\).
For any objects \(M_\bullet ,N_\bullet \in C_2(\mathscr {P}\,)\), it is clear that \(M_\bullet \cong N_\bullet \) implies \({{\textbf {dim}}}\,M_\bullet ={{\textbf {dim}}}\,N_\bullet \). Thus, the additivity of \(\textbf{dim}\) implies the existence of a unique group homomorphism \(f: K(C_2(\mathscr {P}\,))\longrightarrow \mathbb {Z}^{2n}\) such that \(f(\hat{M_\bullet })={{\textbf {dim}}}\,M_\bullet \) for each object \(M_\bullet \) in \(C_2(\mathscr {P}\,)\). Since
is a basis of the free group \(\mathbb {Z}^{2n}\), we conclude that \(\{\hat{Z}_{P_i},\hat{ K}_{P_i}~|~1\le i\le n\}\) is \(\mathbb {Z}\)-linearly independent in \(K(C_2(\mathscr {P}\,))\). It follows that \(K(C_2(\mathscr {P}\,))\) is free and f is an isomorphism. \(\square \)
Let \(\mathcal {T}\) be the \(\mathbb {Z}[v,v^{-1}]\)-algebra with a basis \(\{X^{\alpha }~|~\alpha \in \mathbb {Z}^{2n}\}\) and multiplication defined by
We give integration maps on Hall algebras of morphisms in the following
Proposition 4.2
The integration map
is a homomorphism of algebras.
Proof
For any objects \(M_\bullet ,N_\bullet \in C_2(\mathscr {P}\,)\),
\(\square \)
In what follows, for each \(\alpha \in \mathbb {Z}^{n}\), we always set \(\widetilde{\alpha }:={\textbf{0}\atopwithdelims ()\alpha } \in \mathbb {Z}^{2n}\) and \(\bar{\alpha }:={\alpha \atopwithdelims ()\textbf{0}} \in \mathbb {Z}^{2n}\). For each given object in \(\mathcal {A}\,\) we will always use the corresponding lowercase boldface letter to denote its dimension vector.
Corollary 4.3
The integration map
is a homomorphism of algebras. In particular, \(\int K_\alpha =X^{\tilde{\alpha }}\) and \(\int \mathbb {X}_{M\oplus P[1]}=X^{\bar{\textbf{m}}-\bar{\textbf{p}}}\).
Proof
By definition, \(\int K_\alpha =X^{\tilde{\alpha }}\) and
By Proposition 4.2 and the relation (2.12), we can prove that \(\int \) is a homomorphism of algebras. Thus,
Therefore, we complete the proof. \(\square \)
5 Compatible pairs
Let Q be an acyclic valued quiver (cf. [27, 28]) with the vertex set \(\{1,2,\cdots ,n\}\). For each vertex i, let \(d_i\in \mathbb {N^+}\) be the corresponding valuation. Note that each finite dimensional hereditary k-algebra can be obtained by taking the tensor algebra of the k-species \(\mathfrak {S}\) associated to Q. We identify a k-species \(\mathfrak {S}\) with its corresponding tensor algebra. Let \(m\ge n\), we define a new quiver \(\widetilde{Q}\) by attaching additional vertices \(n+1,\ldots ,m\) to Q. The full subquiver Q is called the principal part of \(\widetilde{Q}\).
For \(1\le i\le m\), denote by \(S_i\) the i-th simple module for \(\widetilde{\mathfrak {S}}\) which is the k-species associated to \(\widetilde{Q}\), and set \(\mathcal {D}_i=\hbox {End}\,_{\widetilde{\mathfrak {S}}}(S_i)\). Let \(\widetilde{R}\) and \(\widetilde{R}'\) be the \(m\times n\) matrices with the i-th row and j-th column elements given respectively by
and
where \(1\le i\le m\) and \(1\le j\le n\). Define \(\widetilde{B}=\widetilde{R}'-\widetilde{R}\), and denote by \(\widetilde{I}\) the left \(m\times n\) submatrix of the \(m\times m\) identity matrix \(I_m\). Denote the principal parts of the matrices \(\widetilde{R}\), \(\widetilde{R}'\) and \(\widetilde{B}\) by R, \(R'\) and B, respectively. That is, \(B=R'-R\). Let \(D_n=diag(d_1,\cdots ,d_n)\), it is easy to see that \(D_nR'=R^{\textrm{tr}}D_n\). Thus, \(D_nB\) is skew-symmetric. Moreover, the matrix representing the Euler form associated to \(\mathfrak {S}\) under the standard basis is \((I_n-R^{\textrm{tr}})D_n=D_n(I_n-R')\).
We always assume that there exists a skew-symmetric \(m\times m\) integral matrix \(\Lambda \) such that
We remark that such \(\widetilde{Q}\) and \(\Lambda \) exist for a given quiver Q (cf. [27]). We call such \((\Lambda ,\widetilde{B})\) a compatible pair. In what follows, we also denote by \(\Lambda \) the skew-symmetric bilinear form on \(\mathbb {Z}^{m}\) associated to the skew-symmetric matrix \(\Lambda \).
For simplicity of notation, we set \(\widetilde{E}=\widetilde{I}-\widetilde{R}'\) and \(\widetilde{E}'=\widetilde{I}-\widetilde{R}\). Thus, \(\widetilde{E}'-\widetilde{E}=\widetilde{B}\).
Lemma 5.1
[10, Lemma 3.1] For any \(\alpha ,\beta \in \mathbb {Z}^n\), we have that
(1) \(\Lambda (\widetilde{E}\alpha ,\widetilde{B}\beta )=-\langle \beta ,\alpha \rangle \); (2) \(\Lambda (\widetilde{B}\alpha ,\widetilde{B}\beta )=\langle \beta ,\alpha \rangle -\langle \alpha ,\beta \rangle \).
Using Lemma 5.1, we easily obtain the following
Lemma 5.2
-
(1)
For any \(\alpha ,\beta \in \mathbb {Z}^n\), we have that
$$\begin{aligned}\Lambda (\widetilde{E}'\alpha ,\widetilde{E}'\beta )=\Lambda (\widetilde{E}\alpha ,\widetilde{E}\beta ).\end{aligned}$$ -
(2)
For any \(\alpha _i,\beta _i\in \mathbb {Z}^n\), \(i=1,2\), we have that
$$\begin{aligned}\Lambda (\widetilde{E}\alpha _1+\widetilde{E}'\beta _1,\widetilde{E}\alpha _2+\widetilde{E}'\beta _2) =\Lambda (\widetilde{E}(\alpha _1+\beta _1),\widetilde{E}(\alpha _2+\beta _2))-\langle \beta _2,\alpha _1\rangle +\langle \beta _1,\alpha _2\rangle .\qquad \end{aligned}$$
Proof
(1) By definition,
(2) By definition,
\(\square \)
6 Factorization of a homomorphism
Let Q, \(\widetilde{Q}\) be the same as given in Section 5. We take \(\mathcal {A}\,\) to be the category of finite dimensional left \(\mathfrak {S}\)-modules, and \(m=2n\). Recall that for each \(\alpha \in \mathbb {Z}^{n}\), \(\widetilde{\alpha }:={\textbf{0}\atopwithdelims ()\alpha } \in \mathbb {Z}^{m}\).
By [12, Lemma 8.1], the algebra \(\mathcal{M}\mathcal{H}(\mathcal {A}\,)\) is \(\mathbb {Z}^{m}\)-graded with the degree of \(K_{\alpha }*\mathbb {X}_{M\oplus P[1]}\) defined by
for any \(\alpha \in \mathbb {Z}^n, M\in \mathcal {A}\,\) and \(P\in \mathscr {P}\,\). It induces a grading on the tensor algebra \((\mathcal {M}\mathcal {H}\,(\mathcal {A}\,)\otimes \mathcal {M}\mathcal {H}\,(\mathcal {A}\,),*)\) defined by
for any \(\alpha , \beta \in \mathbb {Z}^n\), \(M, N\in \mathcal {A}\,\) and \(P, Q\in \mathscr {P}\,\). Using these gradings, we twist the multiplication on \(\mathcal {M}\mathcal {H}\,(\mathcal {A}\,)\), and define \(\mathcal{M}\mathcal{H}_{\Lambda }(\mathcal {A}\,)\) to be the same module as \(\mathcal{M}\mathcal{H}(\mathcal {A}\,)\) but with the twisted multiplication defined on basis elements by
where \(\alpha ,\beta \in \mathbb {Z}^{n}\), \(M, N\in \mathcal {A}\,\) and \(P, Q\in \mathscr {P}\,\). We also twist the multiplication on the tensor algebra \((\mathcal {M}\mathcal {H}\,(\mathcal {A}\,)\otimes \mathcal {M}\mathcal {H}\,(\mathcal {A}\,),*)\) by defining
where \(\lambda =\Lambda (\widetilde{E}'(\textbf{m}-\textbf{p}+\textbf{n}-\textbf{q})-\widetilde{\alpha }-\widetilde{\beta },\widetilde{E}'(\textbf{u}-\textbf{s}+\textbf{v}-\textbf{t})-\widetilde{\gamma }-\widetilde{\delta })\), \(M, N,U,V\in \mathcal {A}\,\), \(P, Q,S,T\in \mathscr {P}\,\) and \(\alpha ,\beta ,\gamma ,\delta \in \mathbb {Z}^{n}\).
For use below we reformulate Theorem 2.6 in the following
Proposition 6.1
The algebra \(\mathcal{M}\mathcal{H}_{\Lambda }(\mathcal {A}\,)\) is generated by all \(K_{\alpha }\) and \(\mathbb {X}_{M\oplus P[1]}\) (with \(\alpha \in \mathbb {Z}^{n}\), \(M\in \mathcal {A}\,\) and \(P\in \mathscr {P}\,)\), which are subject to the following relations
for any \(\alpha ,\beta \in \mathbb {Z}^n\), \(M,N\in \mathcal {A}\,\) and \(P,Q\in \mathscr {P}\,\).
Since the comultiplication \(\Delta \) defined in (3.2) is homogeneous, it is easy to obtain the following
Lemma 6.2
The map \(\Delta :(\mathcal {M}\mathcal {H}\,_\Lambda (\mathcal {A}\,),\star )\longrightarrow (\mathcal {M}\mathcal {H}\,(\mathcal {A}\,)\otimes \mathcal {M}\mathcal {H}\,(\mathcal {A}\,),\star )\) is a homomorphism of algebras.
Proof
Let \(M, N\in \mathcal {A}\,\), \(P, Q\in \mathscr {P}\,\) and \(\alpha ,\beta \in \mathbb {Z}^{n}\). By (6.1), we have that
\(\square \)
We twist the multiplication on the tensor algebra of the torus \(\mathcal {T}\) defined in (4.1) by defining
for any \(\alpha _i, \beta _i, \gamma _i, \delta _i\in \mathbb {Z}^n, i=1, 2.\)
Lemma 6.3
The map \(\int \otimes \int : (\mathcal {M}\mathcal {H}\,(\mathcal {A}\,)\otimes \mathcal {M}\mathcal {H}\,(\mathcal {A}\,),\star )\longrightarrow (\mathcal {T}\otimes \mathcal {T},\star )\) is a homomorphism of algebras.
Proof
Let \(M, N,U,V\in \mathcal {A}\,\), \(P, Q,S,T\in \mathscr {P}\,\) and \(\alpha ,\beta ,\gamma ,\delta \in \mathbb {Z}^{n}\). On the one hand,
where \(a_0=\Lambda (\widetilde{E}'(\textbf{m}-\textbf{p}+\textbf{n}-\textbf{q})-\widetilde{\alpha }-\widetilde{\beta },\widetilde{E}'(\textbf{u}-\textbf{s}+\textbf{v}-\textbf{t})-\widetilde{\gamma }-\widetilde{\delta })\) and \(a_1=a_0+2(\textbf{n}-\textbf{q},\textbf{u}-\textbf{s})+2\langle \textbf{m}-\textbf{p},\textbf{v}-\textbf{t}\rangle \).
On the other hand,
\(\square \)
Define the quantum torus \(\mathcal {T}_\Lambda \) to be the \(\mathbb {Z}[v,v^{-1}]\)-algebra with a basis \(\{X^{\alpha }~|~\alpha \in \mathbb {Z}^{m}\}\) and multiplication defined by
For the need of the sequel proof, we give the following
Lemma 6.4
For any \(\alpha ,\beta \in \mathbb {Z}^n\), we have that
(1) \(\Lambda (\widetilde{E}'\alpha ,\widetilde{\beta })=\Lambda (\widetilde{E}\alpha ,\widetilde{\beta })\); (2) \(\Lambda (\widetilde{\alpha },\widetilde{E}'\beta )=\Lambda (\widetilde{\alpha },\widetilde{E}\beta )\).
Proof
By (5.1), we obtain that \((D_n~0)=\widetilde{B}^{\textrm{tr}}\Lambda =(\widetilde{E}')^{\textrm{tr}}\Lambda -(\widetilde{E})^{\textrm{tr}}\Lambda \). Thus,
Similarly, we can prove the second identity. \(\square \)
Proposition 6.5
The map \(\mu : (\mathcal {T}\otimes \mathcal {T},\star )\longrightarrow (\mathcal {T}_\Lambda ,\star )\) defined by
where \(\alpha _i,\beta _i\in \mathbb {Z}^n\), \(i=1,2\), is a homomorphism of algebras.
Proof
Let \(\alpha _i,\beta _i,\gamma _i,\delta _i\in \mathbb {Z}^n\), \(i=1,2\). On the one hand,
where \(a_0=\Lambda (\widetilde{E}'(\alpha _1+\beta _1)-\widetilde{\alpha }_2-\widetilde{\beta }_2,\widetilde{E}'(\gamma _1+\delta _1)-\widetilde{\gamma }_2-\widetilde{\delta }_2)+2(\beta _1,\gamma _1)+2\langle \alpha _1,\delta _1\rangle \), and \(a_1=a_0-(\alpha _1+\gamma _1,\beta _1+\delta _1)-\langle \alpha _1+\gamma _1,\beta _1+\delta _1\rangle \). By simple calculation, we obtain that
On the other hand,
where \(b_0=-(\alpha _1,\beta _1)-\langle \alpha _1,\beta _1\rangle -(\gamma _1,\delta _1)-\langle \gamma _1,\delta _1\rangle \), and \(b_1=b_0+\Lambda (-\widetilde{E}\alpha _1-\widetilde{E}'\beta _1+\widetilde{\alpha }_2+\widetilde{\beta }_2,-\widetilde{E}\gamma _1-\widetilde{E}'\delta _1+\widetilde{\gamma }_2+\widetilde{\delta }_2)\).
Since
we obtain that \(a_1=b_1\). Therefore,
\(\square \)
Define \(\Psi :=\mu \circ (\int \otimes \int )\circ \Delta \). In other words, we have the following commutative diagram
Given a module \(M\in \mathcal {A}\,\), we denote by \(\textrm{Gr}_{\textbf{e}}M\) the set of all submodules V of M with \({{\textbf {dim}}}\,V= \textbf{e}\). Let us present our main result as the following
Theorem 6.6
The map \(\Psi :(\mathcal {M}\mathcal {H}\,_\Lambda (\mathcal {A}\,),\star )\longrightarrow (\mathcal {T}_\Lambda ,\star )\) is a homomorphism of algebras. Moreover, for any \(\alpha \in \mathbb {Z}^n\), \(M\in \mathcal {A}\,\) and \(P\in \mathscr {P}\,\),
Proof
By Lemma 6.2, Lemma 6.3 and Proposition 6.5, we obtain that \(\Psi \) is a homomorphism of algebras.
For any \(\alpha \in \mathbb {Z}^n\),
For any \(M\in \mathcal {A}\,\) and \(P\in \mathscr {P}\,\), we have that
\(\square \)
The quantum Caldero–Chapoton map associated with a finite acyclic valued quiver has been defined in [27] and [23]. In [27], the quantum Caldero–Chapoton map is defined for kQ-modules, while in [23] such a map is defined for coefficient-free rigid objects in the cluster category associated with Q. Let \(M\in \mathcal {A}\,\) and \(P\in \mathscr {P}\,\). We recall from [12] that the quantum cluster character associated to \(M\oplus P[1]\) is defined to be
where \(\textbf{t}_P={{\textbf {dim}}}\,(P/\hbox {rad}\,P)\).
Lemma 6.7
Let \(M\in \mathcal {A}\,\) and \(P\in \mathscr {P}\,\). If \(\widetilde{E}'\textbf{p}=\textbf{t}_P\), then \(\Psi (\mathbb {X}_{M\oplus P[1]})=X_{M\oplus P[1]}\).
Proof
Since \(\widetilde{B}=\widetilde{E}'-\widetilde{E}\) and \(\widetilde{E}'\textbf{p}=\textbf{t}_P\), we obtain that
Thus, by Theorem 6.6, we get that \(\Psi (\mathbb {X}_{M\oplus P[1]})=X_{M\oplus P[1]}\). \(\square \)
Remark 6.8
If the quiver \(\widetilde{Q}\) satisfies that there is no arrow from i to j for any vertices \(1\le i\le n\) and \(n<j\le 2n\), i.e., for each \(1\le i\le n\) the indecomposable projective \({\mathfrak {S}}\)-module associated to i is also a projective \(\widetilde{\mathfrak {S}}\)-module. Then \(\widetilde{E}'\textbf{p}=\textbf{t}_P\) for any \(P\in \mathscr {P}\,\), thus \(\Psi (\mathbb {X}_{M\oplus P[1]})=X_{M\oplus P[1]}\). In particular, we consider the quiver \(\widetilde{Q}\) associated to Q as follows: for each vertex \(1\le i\le n\) we add the arrow \(n+i\longrightarrow i\). In this case, by [28] there exists a compatible pair \((\Lambda ,\widetilde{B})\).
In what follows, let \((\Lambda ,\widetilde{B})\) be a compatible pair as mentioned in Remark 6.8. Noting that the quantum torus \(\mathcal {T}_\Lambda \) is an Ore domain, we denote by \(\mathcal {F}\) its skew-field of fractions. Let \(\mathcal {A}\,(\Lambda ,\widetilde{B})\) be the quantum cluster algebra (cf. [3]) associated to the compatible pair \((\Lambda ,\widetilde{B})\), which is defined to be the \(\mathbb {Z}[v,v^{-1}]\)-subalgebra of \(\mathcal {F}\) generated by all quantum cluster variables and coefficients. This quantum cluster algebra \(\mathcal {A}\,(\Lambda ,\widetilde{B})\) is called the principal coefficient quantum cluster algebra. We also denote \(\mathcal {A}\,(\Lambda ,\widetilde{B})\) by \(\mathcal {A}\,_q(Q)\). Let \(\mathcal {A}\,\mathcal {H}\,_q(Q)\) be the \(\mathbb {Z}[v,v^{-1}]\)-subalgebra of \(\mathcal {F}\) generated by all the quantum cluster characters \(X_{M\oplus P[1]}\) and coefficients \(X^{\widetilde{\alpha }}\), where \(M\in \mathcal {A}\,\), \(P\in \mathscr {P}\,\) and \(\alpha \in \mathbb {Z}^n\).
By Theorem 6.6, we recover the following
Corollary 6.9
[12, Theorem 8.4] There exists a surjective algebra homomorphism
defined on generators by
for any \(\alpha \in \mathbb {Z}^n\), \(M\in \mathcal {A}\,\) and \(P\in \mathscr {P}\,\).
According to [28], it is known that the quantum cluster algebra \(\mathcal {A}\,_q(Q)\) is the subalgebra of \(\mathcal{A}\mathcal{H}_{q}(Q)\) generated by
Parallelly, we define \(\mathcal{M}\mathcal{C}_{\Lambda }(\mathcal {A}\,)\) to be the subalgebra of \(\mathcal{M}\mathcal{H}_{\Lambda }(\mathcal {A}\,)\) generated by
Then we have the following
Corollary 6.10
[12, Corollary] There exists a surjective algebra homomorphism
Using Theorem 6.6, we can also recover the cluster multiplication formulas (cf. [10,11,12,13]).
Corollary 6.11
Let \(M,N\in \mathcal {A}\,\) and \(P\in \mathscr {P}\,\). Then
-
(1)
\({X}_{M}{X}_{N}=q^{\frac{1}{2}\Lambda (\widetilde{E}{} \textbf{m}, \widetilde{E}{} \textbf{n})+\langle \textbf{m},\textbf{n}\rangle }\sum \limits _{[L]}\frac{|\textrm{Ext}_{\mathcal {A}\,}^{1}(M,N)_{L}|}{|\textrm{Hom}_{\mathcal {A}\,}(M,N)|}X_L;\)
-
(2)
\(\ X_{P[1]}X_{M}=q^{\frac{1}{2}\Lambda (\widetilde{E}{} \textbf{m},\widetilde{E}{} \textbf{p})-\langle \textbf{p},\textbf{m}\rangle } \sum \limits _{[B],[Q]}|{}_Q\textrm{Hom}_{\mathcal {A}\,}(P,M)_B| X_{B\oplus Q[1]}.\)
Proof
It is proved by (6.6) and (6.8) in Proposition 6.1 together with Theorem 6.6 and Lemma 5.2. \(\square \)
Remark 6.12
We obtain more than relations given in Corollary 6.11. In fact, each relation in Proposition 6.1 corresponds to a relation in \(\mathcal{A}\mathcal{H}_{q}(Q)\).
7 An interpretation of g-vectors
Let \((\Lambda ,\widetilde{B})\) be a compatible pair as mentioned in Remark 6.8. In particular, the quantum cluster algebra \(\mathcal {A}\,(\Lambda ,\widetilde{B})\) is of principal coefficients. We endow the quantum torus \(\mathcal {T}_\Lambda \) a \(\mathbb {Z}^n\)-grading by setting
where \(\alpha =(a_1,\cdots , a_{2n})^{\textrm{tr}}\in \mathbb {Z}^{2n}\) and \(\beta _j\) is the j-th column vector of the matrix B. According to [15, Proposition 6.1], every quantum cluster variable z of \(\mathcal {A}\,(\Lambda ,\widetilde{B})\) is homogeneous with respect to the above \(\mathbb {Z}^n\)-grading and the g-vector g(z) of z is defined as the degree of z. Note that the non-frozen quantum cluster variables are precisely
We are going to give a categorical interpretation of g-vectors via the morphism category \(C_2(\mathscr {P}\,)\). Similar to Lemma 4.1, \(\hat{C}_{P_1},\cdots , \hat{C}_{P_n},\hat{K}_{P_1},\cdots , \hat{K}_{P_n}\) is also a \(\mathbb {Z}\)-basis of \(K(C_2(\mathscr {P}\,))\). For each \(M_\bullet \in C_2(\mathscr {P}\,)\), let
be the coordinate vector of \(\hat{M}_\bullet \) with respect to the basis \(\hat{C}_{P_1},\cdots , \hat{C}_{P_n},\hat{K}_{P_1},\cdots , \hat{K}_{P_n}\). We define
to be the truncation of \({\text {ind}}(M_\bullet )\).
Proposition 7.1
Let \(M\in \textrm{Ind}(\mathcal {A}\,)\) be rigid and \(P\in \textrm{Ind}(\mathscr {P}\,)\), we have that
Proof
By the definition of \(X_M\), we have that
where \(E, E'\) denote the principal parts of the matrices \(\widetilde{E}\) and \(\widetilde{E}'\), respectively.
Let \(0\rightarrow \bigoplus _{i=1}^nb_iP_i\rightarrow \bigoplus _{i=1}^n a_iP_i\rightarrow M\rightarrow 0\) be a projective resolution of M. It follows that the k-th component of \(-E'{} \textbf{m}\) is \(-\langle M, S_k\rangle =-(a_k-b_k)\). Consequently, \(g(X_M)=-(a_1-b_1,\cdots , a_n-b_n)^{\textrm{tr}}\). On the other hand, by Lemma 2.5, we have the short exact sequence
In particular, \({\text {ind}}^\circ (C_M)=(a_1-b_1,\cdots ,a_n-b_n)^{\textrm{tr}}=-g(X_M)\).
Without loss of generality, we assume that \(P=P_k\) for some \(1\le k\le n\). By definition, we have \(g(X_{P[1]})=e_k\). Again by Lemma 2.5, we have the short exact sequence
Hence, \(g(X_{P_k[1]})=-{\text {ind}}^0(Z_{P_k})\). \(\square \)
8 Appendix: Quantum cluster characters of a derived Hall subalgebra
8.1 Derived Hall algebras
The derived Hall algebra of the bounded derived category \(D^b(\mathcal {A}\,)\) of \(\mathcal {A}\,\) was introduced in [32] (see also [33]). By definition, the (Drinfeld dual) derived Hall algebra \(\mathcal {D}\mathcal {H}(\mathcal {A}\,)\) is the free \(\mathbb {Z}[v,v^{-1}]\)-module with the basis \(\{u_{X_\bullet }~|~X_\bullet \in \hbox {Iso}\,(D^b(\mathcal {A}\,))\}\) and the multiplication defined by
where \(\hbox {Ext}\,^1_{D^b(\mathcal {A}\,)}({X_\bullet },{Y_\bullet })_{Z_\bullet }\) is defined to be \(\hbox {Hom}_{D^b(\mathcal {A}\,)}({X_\bullet },{Y_\bullet }[1])_{{Z_\bullet }[1]}\), which denotes the subset of \(\hbox {Hom}_{D^b(\mathcal {A}\,)}({X_\bullet },{Y_\bullet }[1])\) consisting of morphisms \(f:{X_\bullet }\rightarrow {Y_\bullet }[1]\) whose cone is isomorphic to \({Z_\bullet }[1]\).
For any \({X_\bullet },{Y_\bullet }\in D^b(\mathcal {A}\,)\), define
it also descends to give a bilinear form on the Grothendieck group of \(D^b(\mathcal {A}\,)\). Moreover, it coincides with the Euler form of \(K(\mathcal {A}\,)\) over the objects in \(\mathcal {A}\,\). In particular, for any \(M,N\in \mathcal {A}\,\) and \(i,j\in \mathbb {Z}\), we have that \(\langle M[i],N[j]\rangle =(-1)^{i-j}\langle M,N\rangle \).
Let us twist the multiplication in \(\mathcal {D}\mathcal {H}(\mathcal {A}\,)\) as follows:
for any \({X_\bullet },{Y_\bullet }\in D^b(\mathcal {A}\,)\). The twisted derived Hall algebra \(\mathcal {D}\mathcal {H}_{q}(\mathcal {A}\,)\) is the same module as \(\mathcal {D}\mathcal {H}(\mathcal {A}\,)\), but with the twisted multiplication. Then we have the following
Proposition 8.1
( [32]) \(\mathcal {D}\mathcal {H}_q(\mathcal {A}\,)\) is an associative unital algebra generated by the elements in \(\{u_{M[i]}~|~M\in \textrm{Iso}(\mathcal {A}\,),~i\in \mathbb {Z}\}\) and the following relations
Let \(C_{\mathcal {A}\,}\) be the subcategory of \(D^b(\mathcal {A}\,)\) consisting of objects \(M\oplus P[1]\) with \(M\in \mathcal {A}\,,P\in \mathscr {P}\,\). Let \(\mathcal {D}\mathcal {H}_q^c(\mathcal {A}\,)\) be the submodule of \(\mathcal {D}\mathcal {H}_q(\mathcal {A}\,)\) spanned by all elements \(u_{X_\bullet }\) with \(X_\bullet \in C_{\mathcal {A}\,}\).
Lemma 8.2
The submodule \(\mathcal {D}\mathcal {H}_q^c(\mathcal {A}\,)\) is a subalgebra of \(\mathcal {D}\mathcal {H}_q(\mathcal {A}\,)\).
Proof
We only need to prove that the subcategory \(C_{\mathcal {A}\,}\) is closed under extensions. Let
be any triangle in \(D^b(\mathcal {A}\,)\) with \(M,N\in \mathcal {A}\,\) and \(P,Q\in \mathscr {P}\,\). By considering the long exact sequence of homology groups induced by (8.6), we obtain that the i-th homology group of \(X_\bullet \) vanishes unless \(i=-1,0\). Hence, we obtain that \(X_\bullet \cong X\oplus Y[1]\) in \(D^b(\mathcal {A}\,)\) for some \(X,Y\in \mathcal {A}\,\), we need to prove that \(Y\in \mathscr {P}\,\).
Since \(\hbox {Hom}(M[-1]\oplus P,Q[1])=0\), by [22, Lemma 2.5], we know that Q is a direct summand of Y. Let \(Y\cong Q\oplus Y'\) and \(M[-1]\oplus P\rightarrow N\rightarrow X\oplus Y'[1]\rightarrow M\oplus P[1]\) be a triangle, then we have a long exact sequence in homologies
Since \(\mathcal {A}\,\) is hereditary, we get that \(Y'\) is projective. Thus, Y is projective and \(X_\bullet \in C_{\mathcal {A}\,}\).
\(\square \)
Using Proposition 8.1, we obtain the following
Proposition 8.3
The subalgebra \(\mathcal {D}\mathcal {H}_q^c(\mathcal {A}\,)\) is generated by the elements in \(\{u_{M}, u_{P[1]}~|M\in \mathcal {A}\,, P\in \mathscr {P}\,\}\), and the following relations
for any \(M,N\in \mathcal {A}\,\) and \(P,Q\in \mathscr {P}\,\).
Proof
By Proposition 8.1, we only need to note that for any \(M, N\in \mathcal {A}\,\) and \(i\in \mathbb {Z}\), the following identity
holds in \(\mathcal {D}\mathcal {H}_q(\mathcal {A}\,)\). \(\square \)
Combining Proposition 8.3 with Theorem 2.6, we obtain the following
Corollary 8.4
There exists an embedding of algebras
defined on generators by \(u_M\mapsto \mathbb {X}_M\) and \(u_{P[1]}\mapsto \mathbb {X}_{P[1]}.\)
The comultiplication \(\Delta \) on \(\mathcal {M}\mathcal {H}\,(\mathcal {A}\,)\) defined in (3.2) induces a comultiplication
defined by
for any \(L\in \mathcal {A}\,\) and \(P\in \mathscr {P}\,\). Let us define the multiplication \(*\) on \(\mathcal {D}\mathcal {H}_q^{c}(\mathcal {A}\,)\otimes \mathcal {D}\mathcal {H}_q^{c}(\mathcal {A}\,)\) by
for any \(M, N, U, V\in \mathcal {A}\,\) and \(P, Q, S, T\in \mathscr {P}\,\). In a similar way to Proposition 3.2, we obtain the following
Proposition 8.5
The map \(\Delta :(\mathcal {D}\mathcal {H}_q^{c}(\mathcal {A}\,),*)\longrightarrow (\mathcal {D}\mathcal {H}_q^{c}(\mathcal {A}\,)\otimes \mathcal {D}\mathcal {H}_q^{c}(\mathcal {A}\,),*)\) is a homomorphism of algebras.
Remark 8.6
Let \(D^{[-1,0]}(\mathcal {A}\,)\) be the subcategory of \(D^b(\mathcal {A}\,)\) consisting of objects \(L\oplus X[1]\) with \(L,X\in \mathcal {A}\,\). Let \(\mathcal {D}\mathcal {H}_q^{[-1,0]}(\mathcal {A}\,)\) be the submodule of \(\mathcal {D}\mathcal {H}_q(\mathcal {A}\,)\) spanned by all elements \(u_{X_\bullet }\) with \(X_\bullet \in D^{[-1,0]}(\mathcal {A}\,)\). Then \(\mathcal {D}\mathcal {H}_q^{[-1,0]}(\mathcal {A}\,)\) is a subalgebra of \(\mathcal {D}\mathcal {H}_q(\mathcal {A}\,)\), which is related to the Heisenberg double Hall algebra defined in [20]. Let
be the homomorphism of modules defined by
for any \(L, X\in \mathcal {A}\,\). Consider the multiplication \(*\) which is similar to (8.13) on \(\mathcal {D}\mathcal {H}_q^{[-1,0]}(\mathcal {A}\,)\otimes \mathcal {D}\mathcal {H}_q^{[-1,0]}(\mathcal {A}\,)\). In a similar way to the proof of Proposition 3.2, we can obtain that the map \(\Delta :(\mathcal {D}\mathcal {H}_q^{[-1,0]}(\mathcal {A}\,),*)\longrightarrow (\mathcal {D}\mathcal {H}_q^{[-1,0]}(\mathcal {A}\,)\otimes \mathcal {D}\mathcal {H}_q^{[-1,0]}(\mathcal {A}\,),*)\) is also a homomorphism of algebras.
8.2 Integration map on the derived Hall subalgebra \(\mathcal {D}\mathcal {H}_q^c(\mathcal {A}\,)\)
Let Q, \(\widetilde{Q}\) be the same as given in Section 5. Let \(\mathcal {A}\,\) be the category of finite dimensional left \(\mathfrak {S}\)-modules. In this subsection, we define an integration map on the derived Hall subalgebra \(\mathcal {D}\mathcal {H}_q^c(\mathcal {A}\,)\). For each positive integer t, let \(\mathcal {T}_t\) be the \(\mathbb {Z}[v,v^{-1}]\)-algebra with a basis \(\{X^{\alpha }~|~\alpha \in \mathbb {Z}^t\}\) and multiplication defined by
It is well known that there is an isomorphism of groups
defined by \(f(\hat{X}_\bullet )=\sum \limits _{i\in \mathbb {Z}}(-1)^i{{\textbf {dim}}}\,X_i=:{{\textbf {dim}}}\,{X}_\bullet .\) Moreover, \(\langle X_\bullet ,Y_\bullet \rangle =\langle {{\textbf {dim}}}\,{X}_\bullet ,{{\textbf {dim}}}\,{Y}_\bullet \rangle .\)
Lemma 8.7
For any objects \(X_\bullet , Y_\bullet \in C_{\mathcal {A}\,}\), we have that \(\textrm{dim}_k\textrm{Hom}_{D^b(\mathcal {A}\,)}(X_\bullet ,Y_\bullet [i])=0\) if \(|i|>1\).
Proof
Let \(X_\bullet =M\oplus P[1]\) and \(Y_\bullet =N\oplus Q[1]\) with \(M,N\in \mathcal {A}\,\) and \(P,Q\in \mathscr {P}\,\).
If \(i>1\), since \(\mathcal {A}\,\) is hereditary and P is projective, we obtain that
For \(i<-1\), it is easy to see that \(\hbox {Hom}_{D^b(\mathcal {A}\,)}(X_\bullet ,Y_\bullet [i])=0.\) \(\square \)
Proposition 8.8
The integration map
is a homomorphism of algebras.
Proof
For any objects \(X_\bullet , Y_\bullet \in C_{\mathcal {A}\,}\),
8.3 Quantum cluster characters via derived Hall subalgebras
Let \(Q, \widetilde{Q}\) be the same as given in Section 5. We keep the notations as in Section 5, in particular, we have \(m\times n\) integral matrices \(\widetilde{B}\) and \(\widetilde{E}\). Let \(\mathcal {A}\) (resp. \(\widetilde{\mathcal {A}}\)) be the category of finite dimensional left \(\mathfrak {S}\) (resp. \(\mathfrak {\widetilde{S}}\))-modules. We may identify \(\mathcal {A}\) with the full subcategory of \(\widetilde{\mathcal {A}}\) consisting of modules with supports on Q. For an \(\mathfrak {S}\)-module X we also denote by \(\textbf{x}\) the dimension vector of X viewed as an \(\widetilde{\mathfrak {S}}\)-module, since this should not cause confusion by the context. Let \(R(\widetilde{Q})\) and \(R'(\widetilde{Q})\) be the \(m\times m\) matrices with the i-th row and j-th column elements given respectively by
and
where \(1\le i,j\le m\). Define \(B(\widetilde{Q})=R'(\widetilde{Q})-R(\widetilde{Q})\), \(E(\widetilde{Q})=I_m-R'(\widetilde{Q})\) and \(E'(\widetilde{Q})=I_m-R(\widetilde{Q})\). Note that \(\widetilde{B}\) is the submatrix of \(B(\widetilde{Q})\) consisting of the first n columns.
In what follows, we assume that there is a skew-symmetric \(m\times m\) integral matrix \(\Lambda \) such that
As in Section 6, we twist the multiplication on \(\mathcal {D}\mathcal {H}_q^c(\widetilde{\mathcal {A}\,})\), and define \(\mathcal {D}\mathcal {H}_\Lambda ^c(\widetilde{\mathcal {A}\,})\) to be the same module as \(\mathcal {D}\mathcal {H}_q^c(\widetilde{\mathcal {A}\,})\) but with the twisted multiplication defined on basis elements by
where \(M, N\in \widetilde{\mathcal {A}\,}\) and \(P, Q\in \mathscr {P}\,_{\widetilde{\mathcal {A}\,}}\). We also twist the multiplication on the tensor algebra \((\mathcal {D}\mathcal {H}_q^c(\widetilde{\mathcal {A}\,})\otimes \mathcal {D}\mathcal {H}_q^c(\widetilde{\mathcal {A}\,}),*)\) by defining
where \(\lambda =\Lambda (E'(\widetilde{Q})(\textbf{m}-\textbf{p}+\textbf{n}-\textbf{q}),E'(\widetilde{Q})(\textbf{u}-\textbf{s}+\textbf{v}-\textbf{t}))\), \(M, N,U,V\in \widetilde{\mathcal {A}\,}\) and \(P, Q,S,T\in \mathscr {P}\,_{\widetilde{\mathcal {A}\,}}\). In a similar way to Lemma 6.2, we obtain the following
Lemma 8.9
The map \(\Delta :(\mathcal {D}\mathcal {H}_\Lambda ^c(\widetilde{\mathcal {A}\,}),\star )\longrightarrow (\mathcal {D}\mathcal {H}_q^c(\widetilde{\mathcal {A}\,})\otimes \mathcal {D}\mathcal {H}_q^c(\widetilde{\mathcal {A}\,}),\star )\) is a homomorphism of algebras.
We twist the multiplication on the tensor algebra of the torus \(\mathcal {T}_m\) by defining
for any \(\alpha , \beta , \gamma , \delta \in \mathbb {Z}^m\). In a similar way to Lemma 6.3, we obtain the following
Lemma 8.10
The map \(\int \otimes \int : (\mathcal {D}\mathcal {H}_q^c(\widetilde{\mathcal {A}\,})\otimes \mathcal {D}\mathcal {H}_q^c(\widetilde{\mathcal {A}\,}),\star )\longrightarrow (\mathcal {T}_m\otimes \mathcal {T}_m,\star )\) is a homomorphism of algebras.
In a similar way to Proposition 6.5, we obtain the following
Proposition 8.11
The map \(\mu : (\mathcal {T}_m\otimes \mathcal {T}_m,\star )\longrightarrow (\mathcal {T}_\Lambda ,\star )\) defined by
where \(\alpha ,\beta \in \mathbb {Z}^m\), is a homomorphism of algebras.
Define \(\psi :=\mu \circ (\int \otimes \int )\circ \Delta \). In other words, we have the following commutative diagram
In a similar way to Theorem 6.6, we obtain the following
Theorem 8.12
The map \(\psi :(\mathcal {D}\mathcal {H}_\Lambda ^c(\widetilde{\mathcal {A}\,}),\star )\longrightarrow (\mathcal {T}_\Lambda ,\star )\) is a homomorphism of algebras. Moreover, for any \(M\in \widetilde{\mathcal {A}\,}\) and \(P\in \mathscr {P}\,_{\widetilde{\mathcal {A}\,}}\),
Corollary 8.13
For any \(M\in \widetilde{\mathcal {A}\,}\) and \(P\in \mathscr {P}\,_{\widetilde{\mathcal {A}\,}}\),
Proof
We only need to note that \({B}(\widetilde{Q})={E}'(\widetilde{Q})-{E}(\widetilde{Q})\) and \({E}'(\widetilde{Q})\textbf{p}=\textbf{t}_P\), thus obtain that
\(\square \)
Remark 8.14
In Corollary 8.13, we do not need to make the principal coefficient assumption as in Remark 6.8. In fact, in Corollary 8.13 we use projectives P in \(\widetilde{\mathcal {A}\,}\) and \({E}'(\widetilde{Q})\textbf{p}=\textbf{t}_P\) always holds (cf. [19, Lemma 1]), while in Lemma 6.7 we use projectives P in \({\mathcal {A}\,}\) and the matrix \(\widetilde{E}'\) is of size \(m\times n\).
Let \(\widetilde{C}_{\mathcal {A}\,}\) be the full subcategory of \({D}^b(\widetilde{\mathcal {A}})\) consisting of objects \(M\oplus P[1]\) with \(M\in \mathcal {A}\) and \(P\in \mathscr {P}_{\widetilde{\mathcal {A}}}\). Let \(\mathcal{D}\mathcal{H}_\Lambda ^{\tilde{c}}({\mathcal {A}})\) be the submodule of \(\mathcal{D}\mathcal{H}_\Lambda ^c(\widetilde{\mathcal {A}})\) spanned by all elements \(u_{X_\bullet }\) with \(X_\bullet \in \widetilde{C}_{\mathcal {A}\,}\).
Lemma 8.15
The submodule \(\mathcal{D}\mathcal{H}_\Lambda ^{\tilde{c}}({\mathcal {A}})\) is a subalgebra of \(\mathcal{D}\mathcal{H}_\Lambda ^c(\widetilde{\mathcal {A}})\).
Proof
We only need to prove that the subcategory \(\widetilde{C}_{\mathcal {A}\,}\) is closed under extensions. Let
be any triangle in \(D^b(\widetilde{\mathcal {A}\,})\) with \(M,N\in \mathcal {A}\,\) and \(P,Q\in \mathscr {P}_{\widetilde{\mathcal {A}}}\). By Lemma 8.2, we assume that \(X_\bullet =X\oplus Y[1]\) with \(X\in \widetilde{\mathcal {A}\,}\) and \(Y\in \mathscr {P}_{\widetilde{\mathcal {A}}}\), and we need to prove that \(X\in \mathcal {A}\,\). By considering the long exact sequence of homology groups induced by (8.19), we have the following exact sequence
Thus, we have a short exact sequence \(0\rightarrow \hbox {Im}\,f\rightarrow X\rightarrow M\rightarrow 0.\) Since \(\hbox {Im}\,f\) is a quotient of N and \(N\in \mathcal {A}\,\), we get that \(\hbox {Im}\,f\in \mathcal {A}\,\). Hence, we obtain that \(X\in \mathcal {A}\,\) since \(M\in \mathcal {A}\,\). \(\square \)
Let us define \(\Psi =\psi \circ l\), where l is the embedding of \(\mathcal{D}\mathcal{H}_\Lambda ^{\tilde{c}}({\mathcal {A}})\) into \(\mathcal{D}\mathcal{H}_\Lambda ^c(\widetilde{\mathcal {A}})\).
Corollary 8.16
The map \(\Psi :(\mathcal {D}\mathcal {H}_\Lambda ^{\tilde{c}}({\mathcal {A}\,}),\star )\longrightarrow (\mathcal {T}_\Lambda ,\star )\) is a homomorphism of algebras. Moreover, for any \(M\in {\mathcal {A}\,}\) and \(P\in \mathscr {P}\,_{\widetilde{\mathcal {A}\,}}\),
That is, \(\Psi (u_{M\oplus {P}[1]})\) is precisely the quantum cluster character associated to \(u_{M\oplus {P}[1]}\) for the quantum cluster algebra \(\mathcal {A}(\Lambda , \widetilde{B})\).
Proof
Clearly, \(\Psi \) is a homomorphism of algebras. By Corollary 8.13, we only need to note that for any \(X\in \mathcal {A}\,\), we have that \(B(\widetilde{Q})\textbf{x}=\widetilde{B}{} \textbf{x}\) and \(E(\widetilde{Q})\textbf{x}=\widetilde{E}{} \textbf{x}\). Moreover, for any two \(\mathfrak {S}\)-modules X and Y, the Euler forms of X and Y as \(\mathfrak {S}\)-modules and \(\widetilde{\mathfrak {S}}\)-modules are equal. \(\square \)
Notes
The notations \(P_M\) and \(\Omega _M\) will be used throughout the paper.
References
Bautista, R.: The category of morphisms between projective modules. Comm. Algebra 32(11), 4303–4331 (2004)
Berenstein, A., Rupel, D.: Quantum cluster characters of Hall algebras. Selecta Math. N. S. 21, 1121–1176 (2015)
Berenstein, A., Zelevinsky, A.: Quantum cluster algebras. Adv. Math. 195, 405–455 (2005)
Buan, A., Marsh, R., Reineke, M., Reiten, I., Todorov, G.: Tilting theory and cluster combinatorics. Adv. Math. 204, 572–618 (2006)
Caldero, P., Chapoton, F.: Cluster algebras as Hall algebras of quiver representations. Comm. Math. Helv. 81, 595–616 (2006)
Caldero, P., Keller, B.: From triangulated categories to cluster algebras. Invent. Math. 172, 169–211 (2008)
Caldero, P., Keller, B.: From triangulated categories to cluster algebras II. Ann. Sci. École Norm. Sup. 39(4), 983–1009 (2006)
Chen, X., Ding, M., Xu, F.: On homomorphisms from Ringel-Hall algebras to quantum cluster algebras. Algebr. Represent. Theor. 19, 171–180 (2016)
Chen, X., Ding, M., Zhang, H.: The cluster multiplication theorem for acyclic quantum cluster algebras, arXiv: 2108.03558 (2021)
Ding, M., Sheng, J., Chen, X.: Multiplicative properties of a quantum Caldero-Chapoton map associated to valued quivers. J. Algebra 442, 299–315 (2015)
Ding, M., Xu, F.: A quantum analogue of generic bases for affine cluster algebras. Sci China Math 55(10), 2045–2066 (2012)
Ding, M., Xu, F., Zhang, H.: Acyclic quantum cluster algebras via Hall algebras of morphisms. Math. Z. 296(3–4), 945–968 (2020)
Fei, J.: Counting using Hall algebras I. Quivers, J. Algebra 372, 542–559 (2012)
Fomin, S., Zelevinsky, A.: Cluster algebras. I. Foundations. J. Amer. Math. Soc 15, 497–529 (2002)
Fomin, S., Zelevinsky, A.: Cluster algebras. IV. Coefficients, Compos. Math 143(1), 112–164 (2007)
Fu, C.: Feigin’s map revisited. J. Pure Appl. Algebra 222(12), 4199–4222 (2018)
Green, J.A.: Hall algebras, hereditary algebras and quantum groups. Invent. Math. 120, 361–377 (1995)
Hubery, A.: From triangulated categories to Lie Algebras: A theorem of Peng and Xiao, Trends in Representation Theory of Algebras and Related Topics, Contemp. Math. 406, Amer. Math. Soc., Providence, 51–66 (2006)
Hubery, A.: Acyclic cluster algebras via Ringel-Hall algebras, preprint
Kapranov, M.: Heisenberg Doubles and Derived Categories. J. Algebra 202, 712–744 (1998)
Peng, L.: Some Hall polynomials for representation-finite trivial extension algebras. J. Algebra 197, 1–13 (1997)
Peng, L., Xiao, J.: Root categories and simple Lie algebras. J. Algebra 198, 19–56 (1997)
Qin, F.: Quantum cluster variables via Serre polynomials. J. Reine Angew. Math. 668, 149–190 (2012)
Riedtmann, C.: Lie algebras generated by indecomposables. J. Algebra 170, 526–546 (1994)
Ringel, C. M.: Hall algebras, in: S. Balcerzyk, et al. (Eds.), Topics in Algebra, Part 1, in: Banach Center Publ. 26, 433–447 (1990)
Ringel, C.M.: Hall algebras and quantum groups. Invent. Math. 101, 583–592 (1990)
Rupel, D.: On a quantum analogue of the Caldero-Chapoton Formula. Int. Math. Res. Not. 14, 3207–3236 (2011)
Rupel, D.: Quantum cluster characters for valued quivers. Trans. Amer. Math. Soc. 367(10), 7061–7102 (2015)
Rupel, D.: The Feigin tetrahedron. SIGMA 11(24), 1–30 (2015)
Schiffmann, O.: Lectures on Hall algebras, Geometric methods in representation theory II, 1–141, S\(\acute{\rm e}\)min. Congr., 24-II, Soc. Math. France, Paris (2012)
Sevenhant, B., Van den Bergh, M.: The double of the Hall algebra of a quiver. J. Algebra 221, 135–160 (1999)
Toën, B.: Derived Hall algebras. Duke Math. J. 135, 587–615 (2006)
Xiao, J., Xu, F.: Hall algebras associated to triangulated categories. Duke Math. J. 143, 357–373 (2008)
Xiao, J., Xu, F.: Green’s formula with \(\mathbb{C}^{*}\)-action and Caldero-Keller’s formula for cluster algebras, in: Representation Theory of Algebraic Groups and Quantum Groups, in: Progr. Math., vol. 284, Birkhauser/Springer, New York, 313–348 (2010)
Zhang, H.: A note on Bridgeland Hall algebras. Comm. Algebra 46(6), 2551–2560 (2018)
Acknowledgements
The authors would like to thank the anonymous reviewers for the careful reading, helpful comments and suggestions. This project was partially supported by the National Natural Science Foundation of China (No.s 11971326, 11821001, 11801273, 12271257).
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.
About this article
Cite this article
Fu, C., Peng, L. & Zhang, H. Quantum cluster characters of Hall algebras revisited. Sel. Math. New Ser. 29, 4 (2023). https://doi.org/10.1007/s00029-022-00811-0
Accepted:
Published:
DOI: https://doi.org/10.1007/s00029-022-00811-0