Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Vertex representations of quantum N-toroidal algebras for type C

Vertex representations of quantum N-toroidal algebras for type C Naihuan Jing, Zhucheng Xu, and Honglian Zhang ABSTRACT. Quantum N-toroidal algebras are generalizations of quantum affine algebras and quan- tum toroidal algebras. In this paper we construct a level-one vertex representation of the quantum N-toroidal algebra for type C. In particular, we also obtain a level-one module of the quantum toroidal algebra for type C as a special case. 1. Introduction ±1 The N-toroidal Lie algebra is the universal central extension of the loop algebra g ⊗ C[t , ±1 ··· , t ] associated to the simple Lie algebra g. The most important subclasses are the 1-toroidal or the affine Kac-Moody algebra g and the 2-toroidal Lie algebra or simply the toroidal Lie algebra. It is well-known that the affine Kac-Moody algebra is the one-dimensional universal central exten- sion of the loop algebra, while the N-toroidal Lie algebra (N ≥ 2) is an infinite-dimensional cen- tral extension of the loop algebra. Extensive studies have been done on representations of toroidal Lie algebras (cf. [26], [27] etc.), for instance, the vertex representations of the toroidal Lie alge- bras T (g) were constructed by Moody-Rao-Yokonuma [26] in simply laced types, generalizing the Frenkel-Kac construction [8]. As generalization of the quantum affine algebras, quantum toroidal algebras were introduced [12] in the study of geometric Langlands conjecture to algebraic surfaces. As quantum affiniza- tion, representations and structures of the quantum affine algebras have been studied first for finite- dimensional representations (cf.[2],[3],[14] etc.) and then the vertex representations (cf. [6], [16], [17], [18], [19], [20], [21] etc.). Varagnolo and Vasserot [30] proved a Schur-Weyl duality between the quantum toroidal algebras U (g ) and Cherednik’s double affine Hecke algebras. The vertex q tor representations of the quantum toroidal algebras in type A were constructed [28] as analogues of the Frenkel-Jing construction [6] of the quantum affine algebras, and then the basic (vertex) rep- resentations of the quantum toroidal algebras in the simply laced types were constructed [7] via the McKay correspondence. In a series of papers [23, 24, 25], Miki studied the structures of the quantum toroidal algebra U (g ) exclusively in type A. Vertex operator realization of the quantum q tor toroidal algebra U (g ) of type A was given in terms of the basic module for the affine Lie algebra q tor gl [9]. Much in depth work was done for the quantum toroidal algebras of type A (including the two parametric deformation) [4] [5], Yangians of toroidal types [11] [13] and also in the viewpoint of quantum affinization [15]. In spite of all these, it is still some away from full understanding of 2010 Mathematics Subject Classification. 17B37, 17B67. Key words and phrases. quantum N-toroidal algebra, vertex operator, Fock space. arXiv:2006.00597v2 [math.QA] 15 Aug 2020 2 JING, XU, AND ZHANG the quantum toroidal algebras in type A, and even less is known for the representation theory of other types. Recently, quantum N-toroidal algebras have been introduced by Gao, Jing, Xia, and Zhang [10] as natural generalization of the quantum toroidal algebras by replacing the 2-torus with the N-torus. The new quantum algebras can be realized as quotient algebras of the quantum GIM algebras [22], which generalize Slodowy’s GIM algebras in connection with algebraic geometry [29, 1]. This offers a new perspective to understand quantum toroidal algebras. As nontrivial realization, vertex representations of the quantum N-toroidal algebras for simply-laced types have been constructed in [10], generalizing [6]. The goal of this paper is to construct a level-one module for the quantum N-toroidal algebra of type C , which provides new example of realization of the algebra for non-simply laced types and a model to understand enlargement of the torus. Unlike the simply laced types [6], the vertex operators coming from the internal quantum Heisenberg subalgebra do not satisfy the Drinfeld commutation relations in the non-simply laced types. As in quantum affine algebras [20], we need to incorporate an auxiliary bosonic field to the vertex operator associated to each simple short root of the underlying Lie algebra. In our construction, one sees that N − 1 copies of the affine Heisenberg algebra stand at the common underlying lattice to represent additional dimension in the torus and similarly the auxiliary Heisenberg subalgebra also bears similar property. In the proof, we try to offer more details on those special for non-simply laced situations. The paper is organized as follows. In section 2, we recall the definition of the quantum N- toroidal algebra for type C via generating functions. We construct the Fock space and vertex operators in section 3. In the last section, we state our main result of the construction and prove in detail that the module realizes the quantum N-toroidal algebra in type C at level one. 2. Quantum N-toroidal algebra U (g ) q N,tor In the paper, we always assume that g is the simple Lie algebra of type C (n ≥ 3). We review the notion of the quantum N-toroidal algebra U (g ) of type C in this section. For this, we q N,tor n start with some notations and data of the Lie algebra g. Let I = {0,··· , n}, I = {1,··· , n} and I = {1, . . . , n − 1}. Set R as the real Euclidean space with the inner product ( | ) and fix an orthogonal basis {e , . . . , e } such that (e |e ) = δ . The root system of g is spanned by the simple roots α , 1 n i j ij i where α = e − e (i = 1, . . . , n − 1), α = 2e . So (α |α ) = 1 for 1 ≤ i ≤ n − 1 and i i i+1 n n i i (α |α ) = 2. Then the weight lattice and root lattice of g are respectively n n P = { k e |k ∈ Z}, i i i i=1 Q = { k α |k ∈ Z}. i i i i=1 ˆ ˆ Denote by g the affine Lie algebra associated to g, let δ be the null root of g satisfying (α |δ) = 0 = (δ|δ), i ∈ I , where the invariant bilinear form ( | ) on g extends that of g. The set of simple (1) roots for the root system of the affine Lie algebra ˆg (in type C ) is given by Π = {α , α ,··· , α }, 0 1 n where α = δ − 2α − 2α − ··· − 2α − α . The affine root and weight lattices are 0 1 2 n−1 n n n X X Q = Zα , P = ZΛ , i i i=0 i=0 VERTEX REPRESENTATIONS OF QUANTUM N-TOROIDAL ALGEBRAS FOR TYPE C 3 2α ∨ ∨ where Λ are defined by Λ (α ) = δ and Λ (d) = 0 for the gradation d. Here α = i i ij i j j (α ,α ) j j for j ∈ I. Then the affine Cartan matrix A = (a ) is given by a = (α |α ) for i, j ∈ I. So ij ij j d a = (α |α ) = d a and A is symmetrized by the diagonal matrix D = diag(d |i ∈ I), where i ij i j j ji i (1) 1 1 (d , d ,··· , d ) = (1, ,··· , , 1) for type C . 0 1 n 2 2 As Lie algebra, the affine Lie algebra g takes the form of the central extension of the loop ±1 −1 a b a+b algebra g⊗C[t ], that is, g = g⊗C[t, t ]⊕Cc with the bracket [x⊗t , y ⊗t ] = [x, y]⊗t + a(x|y)δ c, where x, y ∈ g, a, b ∈ Z and c is a nonzero central element. a+b,0 For generic q, we set q = q , i ∈ I. Fix N ≥ 2, let J = {1,··· , N − 1} and e = i s N−1 (0,··· , 0, 1, 0,··· , 0) the s-th standard unit vector of the (N − 1)-dimensional lattice Z . N−1 The general vector of Z will be denoted by k = (k , k ,··· , k ), and 0 is the (N − 1)- 1 2 N−1 N−1 k k dimensional zero vector. For the formal variable z = (z ,··· , z ), denote z = z . 1 N−1 s=1 We also need formal generating functions in the variables z and z, for instance, we set for each s ∈ J X X X k ± ± −k ± ± −k δ(z) = z , x (z) = x (k)z , x (z ) = x (k e )z , s s s i i i,s i N−1 k∈Z k ∈Z k∈Z s X X (s) (s) (s) (s) −m n φ (z) = φ (m)z , ϕ (z) = ϕ (−n)z . i i i i m∈Z n∈Z + + Two versions of quantum numbers will be used in the paper. For a real number a, the q-analog n −n ∞ q −q of a is defined by (a; q) = (1 − aq ). The quantum integer is defined as [n] = ∞ q −1 k=0 q−q h i [m]! and the q-binomial numbers are = . n [m−n]![n]! DEFINITION 2.1. [10] The quantum N-toroidal algebra U (g ) is the associative algebra q N,tor ± ± with unit 1 generated by γ , s ∈ J and the coefficients of the generating functions x (z), (s) (s) φ (z ), ϕ (z ), i ∈ I satisfying the relations below: s s i i 1 1 1 ± − 2 2 2 (2.1) γ are central such that γ γ = 1, s s s ′ ′ (s) (s ) (s ) (s) (2.2) φ (0)ϕ (0) = ϕ (0)φ (0) = 1, i j j i ′ ′ ′ ′ (s) (s ) (s ) (s) (s) (s ) (s ) (s) (2.3) φ (z )φ (w ′) = φ (w ′)φ (z ), ϕ (z )ϕ (w ′) = ϕ (w ′)ϕ (z), s s s s s s s i j j i i j j i δ δ ′ ′ ′ ′ s,s s,s (s) (s ) (s ) (s) −1 −1 −1 (2.4) g z w γ ϕ (z )φ (w ′) = g z w γ φ (w ′)ϕ (z ), ij s s s s ij s s s s s s i j j i ±1 z ∓ (s) (s) s ± −1 ± (2.5) ϕ (z )x (w)ϕ (z ) = g γ x (w), s s ij s i j i j ∓1 w ∓ (s) (s) s ± −1 2 ± (2.6) φ (z )x (w)φ (z ) = g γ x (w), s s ij s i j i j ± ± ′ (2.7) lim [x (z), x (w)] = 0, for s 6= s , i,s i,s z→w ±(α |α ) ± ± ±(α |α ) ± ± i j i j (2.8) (z − q w) x (z)x (w) = (q z − w) x (w) x (z), i,s j,s j,s i,s −1 1 1 δ wγ wγ ij (s) s (s) + − s 2 2 (2.9) [ x (z), x (w) ] = φ (wγ )δ( ) − ϕ (wγ )δ( ) , s s i,s j,s −1 i i z z q − q i 4 JING, XU, AND ZHANG (s) (s) (s) (s) −1 where φ (r) and ϕ (−r) (r ≥ 0) such that φ (0) = K and ϕ (0) = K are defined by: i i i i i ∞ ∞ X X (s) (s) −r −1 −ℓ φ (r)z = K exp (q −q ) a (ℓ)z , i i i i i r=0 ℓ=1 ∞ ∞ X X (s) (s) r −1 −1 ℓ ϕ (−r)z = K exp −(q −q ) a (−ℓ)z , i i i i r=0 ℓ=1 m=1−a ij h i k ± ± ± ± ± (2.10) Sym (−1) x (z )··· x (z )x (w)x (z )x (z ) = 0, 1 k k+1 m i,s i,s j,s i,s i,s k i z ,...,z 1 m k=0 h i k ± ± ∓ ± ± ′ (2.11) lim (−1) x (z ))··· x (z )x (w)x (z )··· x (z ) = 0, s 6= s ∈ J, 1 k k+1 3 i,s i,s i,s i,s i,s z →w k i k=0 ij zq −1 ± ± ± n i ±1 where g (z) := c z is the Taylor series expansion of g (z) = ( a ) at z = 0 in ij ij n∈Z ijn ij z−q C. Alternatively, the quantum algebra has the following equivalent definition [10]. DEFINITION 2.2. The quantum N-toroidal algebra U (g ) is a complex unital associative q N,tor (s) ± ± 2 N−1 algebra with generators x (k), a (r), K and γ (i ∈ I, s ∈ J, k ∈ Z , r ∈ Z/{0}) and i i i the following relations: (s) ±1 (2.12) γ are central, K and a (r) commute with each other, i j r −r [ r a ] γ − γ (s) (s ) ij i s s (2.13) [ a (r), a (l) ] = δ δ , s,s r+l,0 i j −1 q − q ±a ij ± −1 ± (2.14) K x (k)K = q x (k), j i i j |r| [ r a ] (s) ij i ± ± (2.15) [ a (r), x (k) ] = ± γ x (re +k), s s i j j ± ± ′ (2.16) [x (ke ), x (le ) ] = 0, for s 6= s and kl 6= 0, s s i i ± ± ± ± (2.17) [ x ((k + 1)e ), x (l e ) ] + [ x ((l + 1)e ), x (ke ) ] = 0, s s s ±(α ,α ) s s s ±(α ,α ) i j i j i j j i q q k−l l−k (s) (s) 2 2 γ φ ((k + l)) − γ ϕ ((k + l)) s s + − i i (2.18) [ x (ke ), x (le ) ] = δ , s s ij i j −1 q − q (s) (s) where φ (r) and ϕ (−r) (r ≥ 0) are defined by i i ∞ ∞ X X (s) (s) −r −1 −ℓ φ (r)z = K exp (q −q ) a (ℓ)z , i i i s i i r=0 ℓ=1 ∞ ∞ X X (s) (s) r −1 −1 ℓ ϕ (−r)z = K exp −(q −q ) a (−ℓ)z , i s i i i r=0 ℓ=1 VERTEX REPRESENTATIONS OF QUANTUM N-TOROIDAL ALGEBRAS FOR TYPE C 5 m=1−a ij h i t ± ± ± ± ± (2.19) Sym (−1) x (k e )··· x (k e )x (ℓe )x (k e )··· x (k e ) = 0, t s t s s t+1 s m s i i j i i t i k ,...,k 1 m t=0 h i k ± ± ∓ ± ± (2.20) (−1) x (e m )··· x (e m )x (e ′ℓ)x (e m )··· x (e m ) = 0, s 1 s k s s k+1 s 3 i i i i i k i k=0 for i ∈ I and m m m ℓ 6= 0, s 6= s ∈ J. 0 1 2 3 Here [m] = [m] and the q-bracket is defined by [a, b] = ab − uba. i q u REMARK 2.3. Relation (2.10) or (2.19) is void for i = j, as m = −1 in this case. The defining relations (2.17) to (2.20) are written in terms of the components x (ke ). Using relation (2.15), we can obtain the relations for the generators x (k). REMARK 2.4. By relations (2.13), (2.15) and (2.18), we have more identities as follows. It follows from (2.18) that (s) (s ) −1 1/2 + − −1/2 + − a (1) = K γ [ x (0), x (e ) ], a (−1) = K γ ′ [ x (−e ′), x (0) ]. s j s s i i s i i i i i (s) (s ) Using the first relation to expand the bracket [a (1), a (−1)] we obtain for s 6= s i j ′ ′ (s) (s ) (s ) −1 + − 0 = [ a (1), a (−1) ] = K γ [ [x (0), x (e )], a (−1) ] s s i j i i i j ′ ′ (s ) (s ) −1 + − + − = K γ [ [x (0), a (−1)], x (e ) ] + [ x (0), [x (e ), a (−1)] ] s s s i i j i i i j 1 1 1 −1 + − + − 2 2 2 ′ ′ = [−a ] K γ γ [x (−e ), x (e )] − γ [x (0), x (e − e )] . ij i s ′ s s ′ s s i s i i s i i Therefore, + − −1 + − ′ ′ [x (0), x (e − e )] = γ [x (−e ), x (e )]. s s s s i i s i i Similarly expanding the bracket by using the second relation, we also get that (s) (s ) (s) 2 + − 0 = [ a (1), a (−1) ] = K γ [ a (1), [x (−e ′), x (0)] ] j s i j i j j (s) (s) 2 + − + − = K γ [ [a (1), x (−e ′)], x (0) ] + [ x (−e ′), [a (1), x (0)] ] j s s i j j j i j 1 1 1 − − + − + − 2 2 2 = [a ] K γ γ [x (e − e ′), x (0)] − γ [x (−e ′), x (e )] . ij i j ′ s s s s s s s j j j j So, + − −1 + − ′ ′ [x (e − e ), x (0)] = γ [x (−e ), x (e )]. s s s s s j j j j 3. Fock space and vertex operators We construct the Fock space and vertex operators for the quantum N-toroidal algebra U (g ) q N,tor for type C in this section. n 6 JING, XU, AND ZHANG 3.1. Fock space. Let Q be an identical copy of the root lattice of type A , or the lattice of n−1 n−1 short roots in g, so Q = ⊕ Zα ˜ . We equip Q[A ] with the inner product (α ˜ |α ˜ ) = (α |α ) = i n−1 i j i j i=1 δ − δ . ij |i−j|,1 We also adopt the nondegerate bilinear form ( | ) on the lattice ZJ defined by −1 i 6= j (3.1) (i|j) = , 0 i = j where the elements of J are written as {1, 2,··· , N − 1}. (s) Let H be the quantum Heisenberg Lie algebra generated by a (n) (n ∈ Z , i ∈ I, s ∈ (s) J), a (0) and b (n), b (0) (n ∈ Z , i ∈ I , s ∈ J) satisfying: i i a ′ m −m [a m] q − q (s) (s ) ij i [a (m), a (l)] = δ ′δ , m,−l i j s,s −1 q − q [(α ˜ |α ˜ )m] (s) (s ) i j [b (m), b (l)] = δ δ [m], m,−l i j s,s (s) (s ) [a (m), b (l)] = 0, i j ′ ′ (s ) (s ) [a (0), a (m)] = [b (0), b (m)] = 0, i i j j (s) (s) × ′ where m, l ∈ Z , s, s ∈ J and i, j ∈ I or I depending on a (m) or b (m). i i (s) (s) Let S(H ) be the symmetric algebra generated by {a (−m), b (−m)|m ∈ N, i ∈ I, j ∈ i j I , s ∈ J}. We define the Fock space V = S(H ) ⊗ C[Q] ⊗ C[Q] ⊗ C[ZJ], ˜ ˜ where C[Q] = C[ ⊕ Za ˜ ] is the group algebra of Q, and similarly for C[Q] and C[ZJ]. i∈Ia (s) (s) The space V becomes a natural H-module by letting a (−m), b (−m) (m > 0) act as i i (s) (s) multiplication operators, and a (m), b (m) (m > 0) as differentiation operators subject to the i i Heisenberg relations. α α ˜ s Define the operators a (0), b (0), s(0) (i ∈ I, j ∈ I , s ∈ J) and e , e , e (α ∈ C[Q], α ˜ ∈ i j a C[Q], s ∈ ZJ) act on V by ˜ ˜ ˜ ˜ ˜ ˜ λ λ t λ λ t λ λ t λ λ t λ λ t λ λ t a (0)e e e = (α |λ)e e e , b (0)e e e = (α ˜ |λ)e e e , s(0)e e e = (s|t)e e e , i i j j ˜ ˜ ˜ ˜ ˜ ˜ α λ λ α+λ λ t α ˜ λ λ λ α ˜+λ t s λ λ α λ s+t e e e = e e e , e e e = e e e , e e e = e e e , ˜ ˜ where λ ∈ C[Q], λ ∈ C[Q], t ∈ C[ZJ]. Let z be a complex variable, the above implies that ˜ ˜ ˜ ˜ ˜ ˜ ˜ a (0) λ λ t (α |λ) λ λ t b (0) λ λ t (α ˜ |λ) λ λ t s(0) λ λ t (s|t) λ λ t i i j j z e e e = z e e e , z e e e = z e e e , z e e e = z e e e . We introduce the sign operators ε on V such that 2(α |α ) i j ε ε = (−1) ε ε , for all 1 ≤ i, j ≤ n − 1, i j j i (α |α ) i j ε ε = (−1) ε ε , i or j = 0, n, i j i j (ε ) = 1, ∀ i. i VERTEX REPRESENTATIONS OF QUANTUM N-TOROIDAL ALGEBRAS FOR TYPE C 7 n+1 We remark that these operators can be constructed from a cocycle on the abelian group Z . 3.2. Normal order. The normal order : : is defined as usual: (s) (s) a (m)a (n), m ≤ n, (s) (s) i j : a (m)a (n) : = i j (s) (s) a (n)a (m), m > n, i j α α α : e a (0) : = : a (0)e := e a (0), i i i α ˜ α ˜ α ˜ : e b (0) : = : b (0)e := e b (0), i i i s ′ ′ s s ′ : e s (0) : = =: s (0)e := e s (0), (s) (s) (s ) and similarly for b (m). For commuting operators a (m), a (m) the normal order product is i i i the usual product. α α ˜ s It is easy to see that a (0), b (0), s(0), e , e , e commute with each other except that i j ′ ′ α α α ˜ α ˜ s s j j k k [a (0), e ] = (α |α )e , [b (0), e ] = (α ˜ |α ˜ )e , [s(0), e ] = (δ ′ − 1)e . i i j j j k ss 3.3. Vertex operators. Following [20], we introduce the main vertex operators of U (g ) q N,tor as follows. First we define the normal order product : : of the vertex operators by moving the (s) (s) annihilation operators like a (k), b (k) (k > 0) etc. to the right and keep the sign operators ε i j in the order inside (for : XY : and like). For i ∈ I, j ∈ I , s ∈ J and ǫ = ±, we define ∞ (s) ∞ (s) X X a (−k) k a (k) k ± i ∓ k i ∓ −k ±α ±a (0)+d i i i 2 2 Y (z) = exp ± q z exp ∓ q z e z ε i,s [k/d ] [k/d ] i i i i k=1 k=1 ∞ (s) |k| a (k) i ∓ −k ±α ±a (0)+d i i i =: exp ∓ q z : e z ε , [k/d ] i i k6=0 N−1 ∞ (s) X X Y |k | a (k ) s ± ∓ −k ±α ±a (0)+d i s i i i Y (z) =: exp ∓ q z : e z ε , i s s [k /d ] s i i s=1 k 6=0 s∈J |k| −k −k 2 2 q z (q z) (s) (s) X (z) =: exp ∓ a (k) − ǫ b (k) : jǫ,s j j [k/d ] [k] j j k6=0 b (0) ±α +ǫα ˜ ±s ±a (0)+ǫb (0)+1 ±s(0) (1∓ǫ)b (0) j j j j j ·e z q (−1) ε , |k | s ǫ N−1 ∞ −k −k X X s s 2 2 q z (q z ) (s) s (s) X (z) =: exp ∓ a (k ) − ǫ b (k ) : s s jǫ j j [k /d ] [k ] s j j s s=1 k 6=0 N−1 Y b (0) ±a (0)+ǫb (0)+1 j j (s) ±α +ǫα ˜ ±s ±s(0) (1∓ǫ)b (0) j j j ·e z q (−1) ε , s=1 8 JING, XU, AND ZHANG (s) (s) a (0) −1 −ℓ Φ (z) = q exp (q −q ) a (ℓ)z , i i i ℓ=1 (s) (s) −a (0) −1 ℓ Ψ (z) = q exp −(q −q ) a (−ℓ)z . i i i ℓ=1 (s) Note that we have used ε to mean the action of ε on the subspace indexed by s. Therefore |k| ′ ǫ −ǫ −k −k 2 2 ′ q z (q z) (s) (s) ǫ ′ X (z) =: exp − ǫ a (k) + ǫ b (k) : iǫ,s i i [k/d ] [k] j j k6=0 b (0) ′ ′ ′ ′ ′ ǫ α +ǫα ˜ +ǫ s ǫ a (0)+ǫb (0)+1 +ǫ s(0) (1−ǫ ǫ)b (0) i i i i i ·e z q (−1) ε . PROPOSITION 3.1. The operator product expansions (OPE) are given for i, j ∈ I, k, l ∈ {1,··· , n − 1}, ǫ, ǫ ∈ {±} and s ∈ J as follows. ′ ′ ǫ ǫ ǫ ǫ Y (z)Y (w) =: Y (z)Y (w) : i,s j,s i,s j,s 1, (α |α ) = 0, i j ′ ′ −(ǫ+ǫ )/2 ±ǫǫ (z − wq ) , (α |α ) = ±1, i j ′ ′ ′ ′ (3.2) 1−(ǫ+ǫ )/2 ǫǫ −1−(ǫ+ǫ )/2 ǫǫ (z − wq ) (z − wq ) , (α |α ) = 2,  i j ǫ+ǫ − 1 ǫǫ f(z − wq ) , (α |α ) = − . i j ′ ′ ǫ ǫ ǫ ǫ X (z)X (w) =: X (z)X (w) : kǫ ,s lǫ ,s kǫ ,s lǫ ,s 1 2 1 2  1, (α |α ) = 0, k l ′ ′ −(ǫ+ǫ )/2 ǫǫ ǫ /2 ǫ /2 ǫ ǫ 1 2 1 2 (z − q w) (q z − q w) , (α |α ) = 1, (3.3) k l  ǫ ǫ ǫ+ǫ ′ 1 2 − 1 ǫǫ ǫ ǫ (1−ǫǫ )/2 1 2 1 2 2 2 f(z − q w) f(q z − q w) (−1) , (α |α ) = − k l ′ ′ ′ ǫ ǫ ǫ ǫ ǫǫ +ǫ ǫ ′ 1 2 (3.4) X (z)X (w) =: X (z)X (w) : z , for s 6= s , ′ ′ kǫ ,s kǫ ,s kǫ ,s kǫ ,s 1 2 1 2 3/2 2 (xq /a;q ) ∞ −1/2 −1/2 where f(a − x) = a a q-analog of the infinite series (a − x) at x = 0 [17]. 1/2 2 (xq /a;q ) ′ ′ ′ ′ ǫ ǫ ǫ ǫ −(ǫ+ǫ )/2 −ǫǫ (3.5) Y (z)X (w) =: Y (z)X (w) : (z − q w) , (α |α ) = −1, i,s i,s k kǫ,s kǫ,s ′ ′ ′ ′ ǫ ǫ ǫ ǫ −(ǫ+ǫ )/2 −ǫǫ (3.6) X (z)Y (w) =: X (z)Y (w) : (z − q w) , (α |α ) = −1. i k kǫ,s i,s kǫ,s i,s 1−ǫǫ ′ 1 b (0) ǫ ǫ REMARK 3.2. The normal order product : X (z)X (w) : contains the term (−1) kǫ ,s lǫ ,s 1 2 1−ǫ ǫ b (0) 2b (0) l k (−1) . Also (−1) = 1. PROOF. It is enough to check (3.2), as (3.3) can be obtained similarly. Five cases need to be considered: (α |α ) = 0, 1,−1, 2 and − . Since all verifications are similar, we just give the case i j of (α |α ) = 1. It is clear that i j ′ ′ [ǫǫ n] w ′ ǫ ǫ ǫ ǫ n ǫǫ Y (z)Y (w) =: Y (z)Y (w) : exp − ( ) z i,s i,s i,s i,s ′ (ǫ+ǫ )/2 n[n] zq n≥1 ′ ′ ′ ǫ ǫ −(ǫ+ǫ )/2 ǫǫ =: Y (z)Y (w) : (z − wq ) . i,s i,s VERTEX REPRESENTATIONS OF QUANTUM N-TOROIDAL ALGEBRAS FOR TYPE C 9 4. Vertex representation of U (g ) q N,tor The purpose of this section is to give the main result and the proof. THEOREM 4.1. The map π from U (g ) to End(V) defined by q N,tor (s) (s) a (0) π(K ) = q , π(a (m)) = a (m), π(γ ) = q, i s i i (s) (s) (s) (s) π(φ (z)) = Φ (z), π(ϕ (z)) = Ψ (z), i i i i X X ± ± ± ± ± ± π(x (z)) = X (z) = X (z), π(x (z)) = X (z) = X (z), j,s j,s jǫ,s j j jǫ ǫ∈{±} ǫ∈{±} ± ± ± ± ± ± π(x (z)) = X (z) = Y (z), π(x (z)) = X (z) = Y (z), k,s k,s k,s k k k i ∈ I, j ∈ I = {1,··· , n − 1}, k ∈ {0, n} gives a representation of U (g ). a q N,tor REMARK 4.2. Since our construction already obeys the relations (2.1)-(2.4), we only need to check the relations (2.5)-(2.9). 4.1. Proof of relations (2.5) and (2.6). As relation (2.6) is similar to relation (2.5), it is (s) enough to check relation (2.5). In fact, by moving Ψ (z ) across Y (w) we have that i j (s) Ψ (z )Y (w) i j −la la ij ij q − q q z s (s) −(α |α ) i i l + i j = q exp( ( ) )Y (w)Ψ (z ) j i l w l>0 s (s) = g q Y (w)Ψ (z ). ij s j i 4.2. Proof of relation (2.7). We just show relation (2.7) for the short roots, i.e. i ∈ I . For s 6= s , we have that + + + + 1+ǫ ǫ 1 2 X (z)X (w) = : X (z)X (w) : z . ′ ′ i,s iǫ ,s i,s iǫ ,s 1 2 ǫ ,ǫ 1 2 + + Therefore lim [X (z), X (w)] = 0. z→w ′ i,s i,s 4.3. Proof of relation (2.8). There are five cases to consider: (α |α ) = 0,− ,±1, and 2. i j The first and the last are trivial. Now let (α |α ) = −1, by (3.2)-(3.3) it follows that i j ± ± ± ± ∓1 −1 X (z)X (w) =: X (z)X (w) : (z − q w) . n,s n,s n−1,s n−1,s − − − − Since : X (z)X (w) := − : X (w)X (z) :, we get that n,s n,s n−1,s n−1,s ∓1 ± ± ∓1 ± ± (z − q w) X (z)X (w) = (q z − w) X (w) X (z). n,s n,s n−1,s n−1,s 10 JING, XU, AND ZHANG Take the case (α |α ) = − , i, j ∈ I . Note that f(a − x) −1 = f(a − x) and i j a q→q 3 1 2 2 2 2 (xq /a; q ) (xq /a; q ) 1 ∞ ∞ −1 −1 − −1 (4.1) f(a − x)f(a − q x) = a = (a − q x) , 1 1 2 2 2 2 (xq /a; q ) (xq /a; q ) ∞ ∞ 1 3 1 −1 2 2 2 2 2 f(a − q x) (xq /a; q ) (xq /a; q ) a − q x ∞ ∞ (4.2) = = . 1 5 1 − − 2 2 f(a − qx) 2 2 2 (xq /a; q ) (xq /a; q ) a − q x ∞ ∞ It follows from (3.2)-(3.3) that ǫ ǫ 1−ǫ 1 2 1 + + + + −1 ǫ ǫ 1 2 2 2 2 X (z)X (w) = : X (z)X (w) : f(z − q w)f(zq − wq ) (−1) . i,s j,s iǫ ,s jǫ ,s 1 2 ǫ ,ǫ 1 2 Using (4.1)-(4.2) we have that 1 1 1 − + + + + − + + 2 4 4 (z − q w)X (z)X (w) =: X X : q − : X X : q i,s j,s i+ j+ i− j− 1 1 1 1 + + − + + − 4 4 4 4 + : X X : (q z − q w)− : X X : (q z − q w) i+ j− i− j+ − + + = (q z − w)X (w)X (z), j,s i,s + + + + where we have used the fact that : X (z)X (w) := − : X (w)X (z) : or ε ε = −ε ε . i j j i iǫ jǫ jǫ iǫ 1 2 2 1 4.4. Proof of relation (2.9). To prove (2.9) is to check that −1 δ 1 wq 1 wq ij (s) (s) + − − 2 2 [ X (z), X (w) ] = Φ (wq )δ( ) − Ψ (wq )δ( ) . i,s j,s i i −1 z z q − q Note that the case of (α |α ) = 2 was known in type A, so we will check the cases (α |α ) = 1 i i i i and (α |α ) = − to show the new situation. i j + − + − For i ∈ I , writing : X X :=: X (z)X (w) : etc., by (3.2)-(3.3) we have that ′ ′ iǫ iǫ,s jǫ jǫ ,s 1 1 1 + − + − + − −1 − −1 2 2 2 X (z)X (w) =: X X : q + : X X : (z − w) (q z − q w) i,s i,s i+ i+ i+ i− 1 1 1 + − −1 − −1 + − − 2 2 2 + : X X : (z − w) (q z − q w) + : X X : q . i− i+ i− i− ± ± Modifying the normal order to •X (z)X (w)• by excluding zw, we get that i j −1 1 w q w + − + − [X (z), X (w)] = •X (z)X (w) • (δ( ) − δ( )) 1 1 i,s i,s i+,s i−,s z z 2 2 q − q 1 qw w + − + • X (z)X (w) • (δ( ) − δ( )) 1 1 i−,s i+,s z z 2 2 q − q −1 1 1 qw 1 q w (s) (s) 2 2 = Φ (wq )δ( ) − Ψ (wq )δ( ) , −1 i i z z q − q where we have used the property of the δ-function and the fact that (s) (s) + − 1/2 + − −1 −1/2 •X (z)X (qz)• = Φ (q z), •X (z)X (q z)• = Ψ (q z), i+,s i−,s i i−,s i+,s i + − + − •X (z)X (z)• = •X (z)X (z) • . i+,s i−,s i−,s i+,s VERTEX REPRESENTATIONS OF QUANTUM N-TOROIDAL ALGEBRAS FOR TYPE C 11 For the case (α |α ) = − , recalling (3.3) and (4.1) we obtain that i j 1 1 1 1 + − + − − + − − 4 4 4 4 X (z)X (w) = : X X : (q z − q w)− : X X : (q z − q w) i,s j,s i+ j− i+ j+ 1 1 + − − + − 4 4 + : X X : q − : X X : q . i− j− i− j+ 1 1 1 1 − + − + − − + − 4 4 4 4 X (w)X (z) = : X X : (q w − q z)− : X X : (q w − q z) j,s i,s j− i+ j+ i+ 1 1 − + − + − 4 4 + : X X : q − : X X : q . j− i− j+ i− + − Thus for (α α ) = −1/2, we have [X (z), X (w)] = 0 due to ε ε = −ε ε . i j i j j i i,s j,s 4.5. Proof of relation (2.10). We will check some cases of a = −1 = a , as the other 0,1 n,n−1 cases are similar. By (3.2)-(3.3) we have that + + + X (z )X (z )X (w) 1 2 0,s 0,s 1ǫ,s −2 (z − z )(z − q z ) 1 2 1 2 + + + =: X (z )X (z )X (w) : . 1 2 0,s 0,s 1ǫ,s −1 −1 (z − q w)(z − q w) 1 2 Subsequently + + + −1 + + + + + + X (z )X (z )X (w) − (q + q )X (z )X (w)X (z ) + X (w)X (z )X (z ) 1 2 1 2 1 2 0,s 0,s 1ǫ,s 0,s 1ǫ,s 0,s 1ǫ,s 0,s 0,s −1 −1 −1 −1 −1 −1 = F · (w − q z )(w − q z ) + [2](z − q w)(w − q z ) + (z − q w))(z − q w) 1 2 2 1 1 2 −1 −2 = F · (q − q )w(z − q z ), 2 1 + + + :X (z )X (z )X (w): 1 2 0,s 0,s 1ǫ,s −2 −2 where F = (z − z )(z − q z ). Observe that F · (z − q z ) is −1 −1 1 2 1 2 2 1 (z −q w)(w−q z ) i i antisymmetric in z , so the Serre relation holds in this case. When a = −1 = a (i, j ∈ I ), it follows from (3.2)-(3.3) that ij ji a + + + + + + + + + X (z )X (z )X (w) − [2] X (z )X (w)X (z ) + X (w)X (z )X (z ) 1 2 i 1 2 1 2 iǫ ,s iǫ ,s jǫ,s iǫ ,s jǫ,s iǫ ,s jǫ,s iǫ ,s iǫ ,s 1 2 1 2 1 2 + + + −1 ǫ /2 ǫ /2 ǫ ǫ 1 2 1 2 =: X (z )X (z )X (w) : (z − q z )(z q − z q ) 1 2 1 2 1 2 iǫ ,s iǫ ,s jǫ,s 1 2 ǫ ǫ ǫ +ǫ ǫ ǫ 1 2 1 2 −1 ǫ ǫ −1 ǫ ǫ 1− 1 2 2 2 2 2 2 f(z − q w)f(q z − q w) · f(z − q w)f(q z − q w) (−1) 1 1 2 2 ǫ ǫ ǫ +ǫ ǫ ǫ 1 2 1 −1 ǫ ǫ −1 ǫ ǫ 1− 1 2 2 2 2 2 2 +[2] f(z − q w)f(q z − q w) · f(w − q z )f(q w − q z ) (−1) i 1 1 2 2 ǫ ǫ ǫ+ǫ ǫ ǫ 1 2 1 −1 ǫ ǫ −1 ǫ ǫ 1− 1 2 2 2 2 2 2 (4.3) +f(w − q z )f(q z − q w) · f(w − q z )f(q w − q z ) (−1) . 1 1 2 2 We claim that the above is antisymmetric with respect to the indices, i.e. under the action: z → z , ǫ → ǫ , which will immediately imply the Serre relation in this case. In fact, there are 1 2 1 2 8 subcases to verify. We check one complicated subcase (ǫ , ǫ , ǫ) = (+,−, +). Then the above 1 2 expression (4.3) is simplified to 1 1 −1 − −1 2 2 1 1 (z − q z )(q z − q z ) 1 2 1 2 + + + − −1 2 2 : X (z )X (z )X (w) : (q − q )w(z − q z ), 1 2 Q 2 1 1 1 i+,s i−,s j+,s − − 2 2 (z − q w) · (w − q z ) i 1 i 12 JING, XU, AND ZHANG where we have used the following identity [20]: −1 (w − az )(w − az ) + (a + a )(z − aw)(w − az ) + (z − aw)(z − aw) 1 2 1 2 1 2 −1 2 = (a − a)w(z − a z ). 1 2 Similarly, (4.3) at (ǫ , ǫ , ǫ) = (−, +, +) is equal to 1 2 1 1 −1 − −1 2 2 (z − q z )(q z − q z ) 1 1 1 2 1 2 + + + − −1 2 2 : X (z )X (z )X (w) : (q − q )w(z − q z ). 2 1 1 2 Q 1 1 i−,s i+,s j+,s − − 2 2 (z − q w) · (w − q z ) i 2 This means that (4.3) is indeed antisymmetric under z → z , ǫ → ǫ . Therefore we have proved 1 2 1 2 that for i, j ∈ I , (α |α ) = −1/2 a i j + + + + + + Sym X (z )X (z )X (w) − [2] X (z )X (w)X (z ) 1 2 i 1 2 iǫ ,s iǫ ,s jǫ,s iǫ ,s jǫ,s iǫ ,s 1 2 1 2 1↔2 + + + +X (w)X (z )X (z ) = 0. 1 2 jǫ,s iǫ ,s iǫ ,s 1 2 Now we consider the case of a = −2 = a . By Prop. 3.1 we have that 1,0 n−1,n + + + + X (z )X (z )X (z )X (w) 1 2 3 1ǫ ,s 1ǫ ,s 1ǫ ,s 0,s 1 2 3 −1 ǫ /2 ǫ /2 ǫ ǫ i j i j (z − q z )(q z − q z ) i j i j i<j + + + + =: X (z )X (z )X (z )X (w) : . 1 2 3 1ǫ ,s 1ǫ ,s 1ǫ ,s 0,s 1 2 3 −1 (z − q w) Pulling out the common factor we get that h i k + + + + + (−1) X (z )··· X (z )X (w)X (z )··· X (z ) 1 k k+1 3 1ǫ ,s 1ǫ ,s 0,s 1ǫ ,s 1ǫ ,s 1 k k+1 3 k i k=0 = F (z − qw)(z − qw)(z − qw) + [3] (z − qw)(z − qw)(w − qz ) 1 2 3 1 1 2 3 +[3] (z − qw)(w − qz )(w − qz ) + (w − qz )(w − qz )(w − qz ) . 1 1 2 3 1 2 3 −1 2 2 3 2 3 = F · (q − q) w (z − (q + q )z + q z ) + w(z z − (q + q )z z + q z z ) , 1 2 3 1 2 1 3 2 3 ǫ /2 ǫ ǫ −1 ǫ /2 j i j (z −q z )(q z −q z ) i j i j 1≤i<j≤3 + + −3 where F =: X (z )··· X (w) : q . Now consider the ac- 1 −1 −1 1ǫ 0 (z −q w)(w−q z ) i i 1≤i≤3 −1 tion of S defined by σ(z , ǫ ) = (z , ǫ ). The factor F excluding (z − q z ) is 3 i i σ(i) σ(i) i j 1≤i<j≤3 antisymmetric under the action of S . Recall the following identity [20]: l(σ) 2 3 −1 −1 −1 (−1) (z − (q + q )z + q z )(z − q z )(z − q z )(z − q z ) = 0. 1 2 3 1 2 1 3 2 3 σ∈S Replacing z by (z z z )/z and factoring out an appropriate factor, we also have the following i 1 2 3 i identity l(σ) 2 3 −1 −1 −1 (−1) (z z − (q + q )z z + q z z )(z − q z )(z − q z )(z − q z ) = 0. 2 3 1 3 1 3 1 2 1 3 2 3 σ∈S 3 VERTEX REPRESENTATIONS OF QUANTUM N-TOROIDAL ALGEBRAS FOR TYPE C 13 Using these two identities, we have proved the following equation: m=3 h i k ± ± ± ± ± Sym (−1) X (z )··· X (z )X (w)X (z )··· X (z ) = 0, 1 k k+1 m 1ǫ ,s 1ǫ ,s 0,s 1ǫ ,s 1ǫ ,s 1 m k k+1 k 1 k=0 where Sym runs through the action of S on (z , ǫ ), and this then implies (2.10). m i i For the last relation, without loss of generality, we take i ∈ I in the “+” case as an example. Note that for s 6= s + + + − + + + − X (z )X (z )X (z )X (w) =: X (z )X (z )X (z )X (w) : ′ ′ 1 2 3 1 2 3 iǫ ,s iǫ ,s iǫ ,s iǫ,s iǫ ,s iǫ ,s iǫ ,s iǫ,s 1 2 3 1 2 3 −1 ǫ /2 ǫ /2 ǫ ǫ −1 ǫ ǫ ǫ ǫ i j i j 1 2 3 · (z − z q )(z q − z q ) (z z z ) (z z z ) . i j i j 1 2 3 1 2 3 1≤i<j≤3 − + When we move X (w) across X (z ) to the left, the contraction function only changes in the ′ i iǫ,s iǫ ǫ ǫ ǫ −1 1 2 ǫ ǫ 2 −1 1 ǫ +ǫ ǫ 3 −1 ǫ +ǫ +ǫ ǫ 3 2 3 1 2 3 last three factors respectively to (z z w) (z z w ) , (z w ) (z w ) , and (w ) (w ) . 1 2 1 1 2 1 Then we have that + + + − lim X (z )X (z )X (z )X (w) 1 2 3 iǫ ,s iǫ ,s iǫ ,s iǫ,s 1 2 3 z →w ǫ,ǫ ∈{±} + + − + −[3] X (z )X (z )X (w)X (z ) i 1 2 ′ 3 iǫ ,s iǫ ,s iǫ,s iǫ ,s 1 2 3 + − + + +[3] X (z )X (w)X (z )X (z ) i 1 ′ 2 3 iǫ ,s iǫ,s iǫ ,s iǫ ,s 1 2 3 − + + + −X (w)X (z )X (z )X (z ) ′ 1 2 3 iǫ,s iǫ ,s iǫ ,s iǫ ,s 1 2 3 X Y + + + − ǫ /2 ǫ /2 ǫ ǫ i j i j = lim : X (z )X (z )X (z )X (w) : (z q − z q ) 1 2 3 i j iǫ ,s iǫ ,s iǫ ,s iǫ,s 1 2 3 z →w 1≤i<j≤3 ǫ,ǫ ∈{±} ǫ ǫ ǫ ǫ ǫ −1 −1 1 2 3 ǫ −1 1 2 ǫ ǫ ×(z − z q ) (z z z ) (z z z ) − [3] (z z w) (z z w ) i j 1 2 3 i 1 2 1 2 3 1 2 2 −1 1 ǫ +ǫ ǫ 3 −1 ǫ +ǫ +ǫ ǫ 2 3 1 2 3 +[3] (z w ) (z w ) − (w ) (w ) = 0. i 1 ACKNOWLEDGMENT N. Jing would like to thank the support of Simons Foundation grant 523868 and NSFC grant 11531004. H. Zhang would like to thank the support of NSFC grant 11871325. References [1] S. Berman and R. V. Moody, Lie algebras graded by finite root systems and the intersection matrix algebras of Slowdowy, Invent. Math. 108 (1992), 323–347. [2] V. Chari and A. Pressley, A guide to quantum groups, Cambridge Univ. Press, Cambridge, 1994. [3] V. Chari and A. Pressley, Quantum affine algebras and their representations. Representations of groups (Banff, AB, 1994), CMS Conf. Proc., 16, Amer. Math. Soc. Providence, RI (1995), 59–78. [4] B. Feigin, M. Jimbo, T. Miwa and E. Mukhin, Branching rules for quantum toroidal gl(n), Adv. Math. 300 (2016), 229-274. 14 JING, XU, AND ZHANG [5] B. Feigin, M. Jimbo, T. Miwa and E. Mukhin, Representations of quantum toroidal gl , J. Algebra 380 (2013), 78–108. [6] I. Frenkel and N. Jing, Vertex representations of quantum affine algebras, Proc. Nat’l. Acad. Sci. USA. 85 (1998), 9373–9377. [7] I. B. Frenkel, N. Jing and W. Wang, Quantum vertex representations via finite groups and the McKay correspon- dence, Comm. Math. Phys. 211 (2000), 365–393. [8] I. Frenkel and V. G. Kac, Basic representations of affine Lie algebras and dual resonance models, Invent. Math. 62 (1980), 23-66. [9] Y. Gao and N. Jing, U (gl ) action on gl -modules and quantum toroidal algebras, J. Algebra 273 (2004), 320– N N [10] Y. Gao, N. Jing, L. Xia and H. Zhang, Quantum N-toroidal algebras and quantized GIM algebras of N-fold affinization, 2019, arXiv: 1907.06301. [11] S. Gautam and V. Toledano-Laredo, Yangians and quantum loop algebras, Selecta Math. (N.S.) 19 (2013), 271–336. [12] V. Ginzburg, M. Kapranov and E. Vasserot, Langlands reciprocty for algebric surfaces, Math. Res. Lett. 2 (1995), 147–160. [13] N. Guay and X. Ma, From quantum loop algebras to Yangians, J. Lond. Math. Soc. 86 (2012), 683–700. [14] D. Hernandez, Drinfeld coproduct, quantum fusion tensor category and applications, Proc. London Math. Soc. 95 (2007), 567–608. [15] D. Hernandez, Quantum toroidal algebras and their representations, Selecta Math. (N.S.) 14 (2009), 701–725. [16] N. Jing, Twisted vertex representations of quantum affine algebras, Invent. Math. 102 (1990), 663–690. [17] N. Jing, Higher level representations of the quantum affine algebra U (sl(2)), J. Algebra 182 (1996), 448–468. [18] N. Jing, Quantum Kac-Moody algebras and vertex representations, Lett. Math. Phys. 4 (1998), 261–271. [19] N. Jing, S.-J. Kang and Y. Koyama, Vertex operators between level one irreducible representations of the quantum (1) affine algebra U (D ), Comm. Math. Phys. 174 (1995), 367–392. (1) [20] N. Jing, Y. Koyama and K. C. Misra, Level one representations of quantum affine algebras U (C ), Sel. Math. q n New Ser. 5 (1999), 243–255. [21] N. Jing and K. C. Misra, Vertex operators for twisted quantum affine algebras, Trans. Amer. Math. Soc. 351 (1999), 1663–1690. [22] S. Kolb, Quantum symmetric Kac-Moody pairs, Adv. Math. 267 (2014), 395–469. b b [23] K. Miki, Toroidal and level 0 U (sl ) actions on U (gl )-modules J. Math. Phys. 40 (1999), 3191–3210. n+1 q q n+1 [24] K. Miki, Representations of quantum toroidal algebra U (sl , tor)(n > 2), J. Math. Phys. 41 (2000), 7079– q n+1 [25] K. Miki, Quantum toroidal algebra U (sl , tor) and R matrices, J. Math. Phys. 42 (2001), 2293–2308. q 2 [26] R. V Moody, S. E. Rao and T. Yokonuma, Toroidal Lie algebras and vertex representations, Geom. Ded. 35 (1990), 283-307. [27] S. Rao and R. Moody, Vertex representations for N-toroidal Lie algebras and a generalization of the Virasoro algebra, Comm. Math. Phys. 159 (1994), 239–264. [28] Y. Saito, Quantum toroidal algebras and their vertex representations, Publ. RIMS. Kyoto Univ. 34 (1998), 155–177. [29] P. Slodowy, Beyond Kac-Moody algebras, and inside. Lie algebras and related topics (Windsor, Ont., 1984), pp. 361–371, CMS Conf. Proc., 5, Amer. Math. Soc., Providence, RI, 1986. [30] M. Varagnolo and E. Vasserot, Schur duality in the toroidal setting, Comm. Math. Phys. 182 (1996), 469–484. DEPARTMENT OF MATHEMATICS, SHANGHAI UNIVERSITY, SHANGHAI 200444, CHINA DEPARTMENT OF MATHEMATICS, NORTH CAROLINA STATE UNIVERSITY, RALEIGH, NC 27695, USA, E-mail address: jing@math.ncsu.edu DEPARTMENT OF MATHEMATICS, SHANGHAI UNIVERSITY, SHANGHAI 200444, CHINA DEPARTMENT OF MATHEMATICS, SHANGHAI UNIVERSITY, SHANGHAI 200444, CHINA E-mail address: hlzhangmath@shu.edu.cn http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Mathematics arXiv (Cornell University)

Vertex representations of quantum N-toroidal algebras for type C

Mathematics , Volume 2022 (2006) – May 31, 2020

Loading next page...
 
/lp/arxiv-cornell-university/vertex-representations-of-quantum-n-toroidal-algebras-for-type-c-RspoTn6UbE

References (31)

ISSN
0219-4988
eISSN
ARCH-3343
DOI
10.1142/S0219498821501851
Publisher site
See Article on Publisher Site

Abstract

Naihuan Jing, Zhucheng Xu, and Honglian Zhang ABSTRACT. Quantum N-toroidal algebras are generalizations of quantum affine algebras and quan- tum toroidal algebras. In this paper we construct a level-one vertex representation of the quantum N-toroidal algebra for type C. In particular, we also obtain a level-one module of the quantum toroidal algebra for type C as a special case. 1. Introduction ±1 The N-toroidal Lie algebra is the universal central extension of the loop algebra g ⊗ C[t , ±1 ··· , t ] associated to the simple Lie algebra g. The most important subclasses are the 1-toroidal or the affine Kac-Moody algebra g and the 2-toroidal Lie algebra or simply the toroidal Lie algebra. It is well-known that the affine Kac-Moody algebra is the one-dimensional universal central exten- sion of the loop algebra, while the N-toroidal Lie algebra (N ≥ 2) is an infinite-dimensional cen- tral extension of the loop algebra. Extensive studies have been done on representations of toroidal Lie algebras (cf. [26], [27] etc.), for instance, the vertex representations of the toroidal Lie alge- bras T (g) were constructed by Moody-Rao-Yokonuma [26] in simply laced types, generalizing the Frenkel-Kac construction [8]. As generalization of the quantum affine algebras, quantum toroidal algebras were introduced [12] in the study of geometric Langlands conjecture to algebraic surfaces. As quantum affiniza- tion, representations and structures of the quantum affine algebras have been studied first for finite- dimensional representations (cf.[2],[3],[14] etc.) and then the vertex representations (cf. [6], [16], [17], [18], [19], [20], [21] etc.). Varagnolo and Vasserot [30] proved a Schur-Weyl duality between the quantum toroidal algebras U (g ) and Cherednik’s double affine Hecke algebras. The vertex q tor representations of the quantum toroidal algebras in type A were constructed [28] as analogues of the Frenkel-Jing construction [6] of the quantum affine algebras, and then the basic (vertex) rep- resentations of the quantum toroidal algebras in the simply laced types were constructed [7] via the McKay correspondence. In a series of papers [23, 24, 25], Miki studied the structures of the quantum toroidal algebra U (g ) exclusively in type A. Vertex operator realization of the quantum q tor toroidal algebra U (g ) of type A was given in terms of the basic module for the affine Lie algebra q tor gl [9]. Much in depth work was done for the quantum toroidal algebras of type A (including the two parametric deformation) [4] [5], Yangians of toroidal types [11] [13] and also in the viewpoint of quantum affinization [15]. In spite of all these, it is still some away from full understanding of 2010 Mathematics Subject Classification. 17B37, 17B67. Key words and phrases. quantum N-toroidal algebra, vertex operator, Fock space. arXiv:2006.00597v2 [math.QA] 15 Aug 2020 2 JING, XU, AND ZHANG the quantum toroidal algebras in type A, and even less is known for the representation theory of other types. Recently, quantum N-toroidal algebras have been introduced by Gao, Jing, Xia, and Zhang [10] as natural generalization of the quantum toroidal algebras by replacing the 2-torus with the N-torus. The new quantum algebras can be realized as quotient algebras of the quantum GIM algebras [22], which generalize Slodowy’s GIM algebras in connection with algebraic geometry [29, 1]. This offers a new perspective to understand quantum toroidal algebras. As nontrivial realization, vertex representations of the quantum N-toroidal algebras for simply-laced types have been constructed in [10], generalizing [6]. The goal of this paper is to construct a level-one module for the quantum N-toroidal algebra of type C , which provides new example of realization of the algebra for non-simply laced types and a model to understand enlargement of the torus. Unlike the simply laced types [6], the vertex operators coming from the internal quantum Heisenberg subalgebra do not satisfy the Drinfeld commutation relations in the non-simply laced types. As in quantum affine algebras [20], we need to incorporate an auxiliary bosonic field to the vertex operator associated to each simple short root of the underlying Lie algebra. In our construction, one sees that N − 1 copies of the affine Heisenberg algebra stand at the common underlying lattice to represent additional dimension in the torus and similarly the auxiliary Heisenberg subalgebra also bears similar property. In the proof, we try to offer more details on those special for non-simply laced situations. The paper is organized as follows. In section 2, we recall the definition of the quantum N- toroidal algebra for type C via generating functions. We construct the Fock space and vertex operators in section 3. In the last section, we state our main result of the construction and prove in detail that the module realizes the quantum N-toroidal algebra in type C at level one. 2. Quantum N-toroidal algebra U (g ) q N,tor In the paper, we always assume that g is the simple Lie algebra of type C (n ≥ 3). We review the notion of the quantum N-toroidal algebra U (g ) of type C in this section. For this, we q N,tor n start with some notations and data of the Lie algebra g. Let I = {0,··· , n}, I = {1,··· , n} and I = {1, . . . , n − 1}. Set R as the real Euclidean space with the inner product ( | ) and fix an orthogonal basis {e , . . . , e } such that (e |e ) = δ . The root system of g is spanned by the simple roots α , 1 n i j ij i where α = e − e (i = 1, . . . , n − 1), α = 2e . So (α |α ) = 1 for 1 ≤ i ≤ n − 1 and i i i+1 n n i i (α |α ) = 2. Then the weight lattice and root lattice of g are respectively n n P = { k e |k ∈ Z}, i i i i=1 Q = { k α |k ∈ Z}. i i i i=1 ˆ ˆ Denote by g the affine Lie algebra associated to g, let δ be the null root of g satisfying (α |δ) = 0 = (δ|δ), i ∈ I , where the invariant bilinear form ( | ) on g extends that of g. The set of simple (1) roots for the root system of the affine Lie algebra ˆg (in type C ) is given by Π = {α , α ,··· , α }, 0 1 n where α = δ − 2α − 2α − ··· − 2α − α . The affine root and weight lattices are 0 1 2 n−1 n n n X X Q = Zα , P = ZΛ , i i i=0 i=0 VERTEX REPRESENTATIONS OF QUANTUM N-TOROIDAL ALGEBRAS FOR TYPE C 3 2α ∨ ∨ where Λ are defined by Λ (α ) = δ and Λ (d) = 0 for the gradation d. Here α = i i ij i j j (α ,α ) j j for j ∈ I. Then the affine Cartan matrix A = (a ) is given by a = (α |α ) for i, j ∈ I. So ij ij j d a = (α |α ) = d a and A is symmetrized by the diagonal matrix D = diag(d |i ∈ I), where i ij i j j ji i (1) 1 1 (d , d ,··· , d ) = (1, ,··· , , 1) for type C . 0 1 n 2 2 As Lie algebra, the affine Lie algebra g takes the form of the central extension of the loop ±1 −1 a b a+b algebra g⊗C[t ], that is, g = g⊗C[t, t ]⊕Cc with the bracket [x⊗t , y ⊗t ] = [x, y]⊗t + a(x|y)δ c, where x, y ∈ g, a, b ∈ Z and c is a nonzero central element. a+b,0 For generic q, we set q = q , i ∈ I. Fix N ≥ 2, let J = {1,··· , N − 1} and e = i s N−1 (0,··· , 0, 1, 0,··· , 0) the s-th standard unit vector of the (N − 1)-dimensional lattice Z . N−1 The general vector of Z will be denoted by k = (k , k ,··· , k ), and 0 is the (N − 1)- 1 2 N−1 N−1 k k dimensional zero vector. For the formal variable z = (z ,··· , z ), denote z = z . 1 N−1 s=1 We also need formal generating functions in the variables z and z, for instance, we set for each s ∈ J X X X k ± ± −k ± ± −k δ(z) = z , x (z) = x (k)z , x (z ) = x (k e )z , s s s i i i,s i N−1 k∈Z k ∈Z k∈Z s X X (s) (s) (s) (s) −m n φ (z) = φ (m)z , ϕ (z) = ϕ (−n)z . i i i i m∈Z n∈Z + + Two versions of quantum numbers will be used in the paper. For a real number a, the q-analog n −n ∞ q −q of a is defined by (a; q) = (1 − aq ). The quantum integer is defined as [n] = ∞ q −1 k=0 q−q h i [m]! and the q-binomial numbers are = . n [m−n]![n]! DEFINITION 2.1. [10] The quantum N-toroidal algebra U (g ) is the associative algebra q N,tor ± ± with unit 1 generated by γ , s ∈ J and the coefficients of the generating functions x (z), (s) (s) φ (z ), ϕ (z ), i ∈ I satisfying the relations below: s s i i 1 1 1 ± − 2 2 2 (2.1) γ are central such that γ γ = 1, s s s ′ ′ (s) (s ) (s ) (s) (2.2) φ (0)ϕ (0) = ϕ (0)φ (0) = 1, i j j i ′ ′ ′ ′ (s) (s ) (s ) (s) (s) (s ) (s ) (s) (2.3) φ (z )φ (w ′) = φ (w ′)φ (z ), ϕ (z )ϕ (w ′) = ϕ (w ′)ϕ (z), s s s s s s s i j j i i j j i δ δ ′ ′ ′ ′ s,s s,s (s) (s ) (s ) (s) −1 −1 −1 (2.4) g z w γ ϕ (z )φ (w ′) = g z w γ φ (w ′)ϕ (z ), ij s s s s ij s s s s s s i j j i ±1 z ∓ (s) (s) s ± −1 ± (2.5) ϕ (z )x (w)ϕ (z ) = g γ x (w), s s ij s i j i j ∓1 w ∓ (s) (s) s ± −1 2 ± (2.6) φ (z )x (w)φ (z ) = g γ x (w), s s ij s i j i j ± ± ′ (2.7) lim [x (z), x (w)] = 0, for s 6= s , i,s i,s z→w ±(α |α ) ± ± ±(α |α ) ± ± i j i j (2.8) (z − q w) x (z)x (w) = (q z − w) x (w) x (z), i,s j,s j,s i,s −1 1 1 δ wγ wγ ij (s) s (s) + − s 2 2 (2.9) [ x (z), x (w) ] = φ (wγ )δ( ) − ϕ (wγ )δ( ) , s s i,s j,s −1 i i z z q − q i 4 JING, XU, AND ZHANG (s) (s) (s) (s) −1 where φ (r) and ϕ (−r) (r ≥ 0) such that φ (0) = K and ϕ (0) = K are defined by: i i i i i ∞ ∞ X X (s) (s) −r −1 −ℓ φ (r)z = K exp (q −q ) a (ℓ)z , i i i i i r=0 ℓ=1 ∞ ∞ X X (s) (s) r −1 −1 ℓ ϕ (−r)z = K exp −(q −q ) a (−ℓ)z , i i i i r=0 ℓ=1 m=1−a ij h i k ± ± ± ± ± (2.10) Sym (−1) x (z )··· x (z )x (w)x (z )x (z ) = 0, 1 k k+1 m i,s i,s j,s i,s i,s k i z ,...,z 1 m k=0 h i k ± ± ∓ ± ± ′ (2.11) lim (−1) x (z ))··· x (z )x (w)x (z )··· x (z ) = 0, s 6= s ∈ J, 1 k k+1 3 i,s i,s i,s i,s i,s z →w k i k=0 ij zq −1 ± ± ± n i ±1 where g (z) := c z is the Taylor series expansion of g (z) = ( a ) at z = 0 in ij ij n∈Z ijn ij z−q C. Alternatively, the quantum algebra has the following equivalent definition [10]. DEFINITION 2.2. The quantum N-toroidal algebra U (g ) is a complex unital associative q N,tor (s) ± ± 2 N−1 algebra with generators x (k), a (r), K and γ (i ∈ I, s ∈ J, k ∈ Z , r ∈ Z/{0}) and i i i the following relations: (s) ±1 (2.12) γ are central, K and a (r) commute with each other, i j r −r [ r a ] γ − γ (s) (s ) ij i s s (2.13) [ a (r), a (l) ] = δ δ , s,s r+l,0 i j −1 q − q ±a ij ± −1 ± (2.14) K x (k)K = q x (k), j i i j |r| [ r a ] (s) ij i ± ± (2.15) [ a (r), x (k) ] = ± γ x (re +k), s s i j j ± ± ′ (2.16) [x (ke ), x (le ) ] = 0, for s 6= s and kl 6= 0, s s i i ± ± ± ± (2.17) [ x ((k + 1)e ), x (l e ) ] + [ x ((l + 1)e ), x (ke ) ] = 0, s s s ±(α ,α ) s s s ±(α ,α ) i j i j i j j i q q k−l l−k (s) (s) 2 2 γ φ ((k + l)) − γ ϕ ((k + l)) s s + − i i (2.18) [ x (ke ), x (le ) ] = δ , s s ij i j −1 q − q (s) (s) where φ (r) and ϕ (−r) (r ≥ 0) are defined by i i ∞ ∞ X X (s) (s) −r −1 −ℓ φ (r)z = K exp (q −q ) a (ℓ)z , i i i s i i r=0 ℓ=1 ∞ ∞ X X (s) (s) r −1 −1 ℓ ϕ (−r)z = K exp −(q −q ) a (−ℓ)z , i s i i i r=0 ℓ=1 VERTEX REPRESENTATIONS OF QUANTUM N-TOROIDAL ALGEBRAS FOR TYPE C 5 m=1−a ij h i t ± ± ± ± ± (2.19) Sym (−1) x (k e )··· x (k e )x (ℓe )x (k e )··· x (k e ) = 0, t s t s s t+1 s m s i i j i i t i k ,...,k 1 m t=0 h i k ± ± ∓ ± ± (2.20) (−1) x (e m )··· x (e m )x (e ′ℓ)x (e m )··· x (e m ) = 0, s 1 s k s s k+1 s 3 i i i i i k i k=0 for i ∈ I and m m m ℓ 6= 0, s 6= s ∈ J. 0 1 2 3 Here [m] = [m] and the q-bracket is defined by [a, b] = ab − uba. i q u REMARK 2.3. Relation (2.10) or (2.19) is void for i = j, as m = −1 in this case. The defining relations (2.17) to (2.20) are written in terms of the components x (ke ). Using relation (2.15), we can obtain the relations for the generators x (k). REMARK 2.4. By relations (2.13), (2.15) and (2.18), we have more identities as follows. It follows from (2.18) that (s) (s ) −1 1/2 + − −1/2 + − a (1) = K γ [ x (0), x (e ) ], a (−1) = K γ ′ [ x (−e ′), x (0) ]. s j s s i i s i i i i i (s) (s ) Using the first relation to expand the bracket [a (1), a (−1)] we obtain for s 6= s i j ′ ′ (s) (s ) (s ) −1 + − 0 = [ a (1), a (−1) ] = K γ [ [x (0), x (e )], a (−1) ] s s i j i i i j ′ ′ (s ) (s ) −1 + − + − = K γ [ [x (0), a (−1)], x (e ) ] + [ x (0), [x (e ), a (−1)] ] s s s i i j i i i j 1 1 1 −1 + − + − 2 2 2 ′ ′ = [−a ] K γ γ [x (−e ), x (e )] − γ [x (0), x (e − e )] . ij i s ′ s s ′ s s i s i i s i i Therefore, + − −1 + − ′ ′ [x (0), x (e − e )] = γ [x (−e ), x (e )]. s s s s i i s i i Similarly expanding the bracket by using the second relation, we also get that (s) (s ) (s) 2 + − 0 = [ a (1), a (−1) ] = K γ [ a (1), [x (−e ′), x (0)] ] j s i j i j j (s) (s) 2 + − + − = K γ [ [a (1), x (−e ′)], x (0) ] + [ x (−e ′), [a (1), x (0)] ] j s s i j j j i j 1 1 1 − − + − + − 2 2 2 = [a ] K γ γ [x (e − e ′), x (0)] − γ [x (−e ′), x (e )] . ij i j ′ s s s s s s s j j j j So, + − −1 + − ′ ′ [x (e − e ), x (0)] = γ [x (−e ), x (e )]. s s s s s j j j j 3. Fock space and vertex operators We construct the Fock space and vertex operators for the quantum N-toroidal algebra U (g ) q N,tor for type C in this section. n 6 JING, XU, AND ZHANG 3.1. Fock space. Let Q be an identical copy of the root lattice of type A , or the lattice of n−1 n−1 short roots in g, so Q = ⊕ Zα ˜ . We equip Q[A ] with the inner product (α ˜ |α ˜ ) = (α |α ) = i n−1 i j i j i=1 δ − δ . ij |i−j|,1 We also adopt the nondegerate bilinear form ( | ) on the lattice ZJ defined by −1 i 6= j (3.1) (i|j) = , 0 i = j where the elements of J are written as {1, 2,··· , N − 1}. (s) Let H be the quantum Heisenberg Lie algebra generated by a (n) (n ∈ Z , i ∈ I, s ∈ (s) J), a (0) and b (n), b (0) (n ∈ Z , i ∈ I , s ∈ J) satisfying: i i a ′ m −m [a m] q − q (s) (s ) ij i [a (m), a (l)] = δ ′δ , m,−l i j s,s −1 q − q [(α ˜ |α ˜ )m] (s) (s ) i j [b (m), b (l)] = δ δ [m], m,−l i j s,s (s) (s ) [a (m), b (l)] = 0, i j ′ ′ (s ) (s ) [a (0), a (m)] = [b (0), b (m)] = 0, i i j j (s) (s) × ′ where m, l ∈ Z , s, s ∈ J and i, j ∈ I or I depending on a (m) or b (m). i i (s) (s) Let S(H ) be the symmetric algebra generated by {a (−m), b (−m)|m ∈ N, i ∈ I, j ∈ i j I , s ∈ J}. We define the Fock space V = S(H ) ⊗ C[Q] ⊗ C[Q] ⊗ C[ZJ], ˜ ˜ where C[Q] = C[ ⊕ Za ˜ ] is the group algebra of Q, and similarly for C[Q] and C[ZJ]. i∈Ia (s) (s) The space V becomes a natural H-module by letting a (−m), b (−m) (m > 0) act as i i (s) (s) multiplication operators, and a (m), b (m) (m > 0) as differentiation operators subject to the i i Heisenberg relations. α α ˜ s Define the operators a (0), b (0), s(0) (i ∈ I, j ∈ I , s ∈ J) and e , e , e (α ∈ C[Q], α ˜ ∈ i j a C[Q], s ∈ ZJ) act on V by ˜ ˜ ˜ ˜ ˜ ˜ λ λ t λ λ t λ λ t λ λ t λ λ t λ λ t a (0)e e e = (α |λ)e e e , b (0)e e e = (α ˜ |λ)e e e , s(0)e e e = (s|t)e e e , i i j j ˜ ˜ ˜ ˜ ˜ ˜ α λ λ α+λ λ t α ˜ λ λ λ α ˜+λ t s λ λ α λ s+t e e e = e e e , e e e = e e e , e e e = e e e , ˜ ˜ where λ ∈ C[Q], λ ∈ C[Q], t ∈ C[ZJ]. Let z be a complex variable, the above implies that ˜ ˜ ˜ ˜ ˜ ˜ ˜ a (0) λ λ t (α |λ) λ λ t b (0) λ λ t (α ˜ |λ) λ λ t s(0) λ λ t (s|t) λ λ t i i j j z e e e = z e e e , z e e e = z e e e , z e e e = z e e e . We introduce the sign operators ε on V such that 2(α |α ) i j ε ε = (−1) ε ε , for all 1 ≤ i, j ≤ n − 1, i j j i (α |α ) i j ε ε = (−1) ε ε , i or j = 0, n, i j i j (ε ) = 1, ∀ i. i VERTEX REPRESENTATIONS OF QUANTUM N-TOROIDAL ALGEBRAS FOR TYPE C 7 n+1 We remark that these operators can be constructed from a cocycle on the abelian group Z . 3.2. Normal order. The normal order : : is defined as usual: (s) (s) a (m)a (n), m ≤ n, (s) (s) i j : a (m)a (n) : = i j (s) (s) a (n)a (m), m > n, i j α α α : e a (0) : = : a (0)e := e a (0), i i i α ˜ α ˜ α ˜ : e b (0) : = : b (0)e := e b (0), i i i s ′ ′ s s ′ : e s (0) : = =: s (0)e := e s (0), (s) (s) (s ) and similarly for b (m). For commuting operators a (m), a (m) the normal order product is i i i the usual product. α α ˜ s It is easy to see that a (0), b (0), s(0), e , e , e commute with each other except that i j ′ ′ α α α ˜ α ˜ s s j j k k [a (0), e ] = (α |α )e , [b (0), e ] = (α ˜ |α ˜ )e , [s(0), e ] = (δ ′ − 1)e . i i j j j k ss 3.3. Vertex operators. Following [20], we introduce the main vertex operators of U (g ) q N,tor as follows. First we define the normal order product : : of the vertex operators by moving the (s) (s) annihilation operators like a (k), b (k) (k > 0) etc. to the right and keep the sign operators ε i j in the order inside (for : XY : and like). For i ∈ I, j ∈ I , s ∈ J and ǫ = ±, we define ∞ (s) ∞ (s) X X a (−k) k a (k) k ± i ∓ k i ∓ −k ±α ±a (0)+d i i i 2 2 Y (z) = exp ± q z exp ∓ q z e z ε i,s [k/d ] [k/d ] i i i i k=1 k=1 ∞ (s) |k| a (k) i ∓ −k ±α ±a (0)+d i i i =: exp ∓ q z : e z ε , [k/d ] i i k6=0 N−1 ∞ (s) X X Y |k | a (k ) s ± ∓ −k ±α ±a (0)+d i s i i i Y (z) =: exp ∓ q z : e z ε , i s s [k /d ] s i i s=1 k 6=0 s∈J |k| −k −k 2 2 q z (q z) (s) (s) X (z) =: exp ∓ a (k) − ǫ b (k) : jǫ,s j j [k/d ] [k] j j k6=0 b (0) ±α +ǫα ˜ ±s ±a (0)+ǫb (0)+1 ±s(0) (1∓ǫ)b (0) j j j j j ·e z q (−1) ε , |k | s ǫ N−1 ∞ −k −k X X s s 2 2 q z (q z ) (s) s (s) X (z) =: exp ∓ a (k ) − ǫ b (k ) : s s jǫ j j [k /d ] [k ] s j j s s=1 k 6=0 N−1 Y b (0) ±a (0)+ǫb (0)+1 j j (s) ±α +ǫα ˜ ±s ±s(0) (1∓ǫ)b (0) j j j ·e z q (−1) ε , s=1 8 JING, XU, AND ZHANG (s) (s) a (0) −1 −ℓ Φ (z) = q exp (q −q ) a (ℓ)z , i i i ℓ=1 (s) (s) −a (0) −1 ℓ Ψ (z) = q exp −(q −q ) a (−ℓ)z . i i i ℓ=1 (s) Note that we have used ε to mean the action of ε on the subspace indexed by s. Therefore |k| ′ ǫ −ǫ −k −k 2 2 ′ q z (q z) (s) (s) ǫ ′ X (z) =: exp − ǫ a (k) + ǫ b (k) : iǫ,s i i [k/d ] [k] j j k6=0 b (0) ′ ′ ′ ′ ′ ǫ α +ǫα ˜ +ǫ s ǫ a (0)+ǫb (0)+1 +ǫ s(0) (1−ǫ ǫ)b (0) i i i i i ·e z q (−1) ε . PROPOSITION 3.1. The operator product expansions (OPE) are given for i, j ∈ I, k, l ∈ {1,··· , n − 1}, ǫ, ǫ ∈ {±} and s ∈ J as follows. ′ ′ ǫ ǫ ǫ ǫ Y (z)Y (w) =: Y (z)Y (w) : i,s j,s i,s j,s 1, (α |α ) = 0, i j ′ ′ −(ǫ+ǫ )/2 ±ǫǫ (z − wq ) , (α |α ) = ±1, i j ′ ′ ′ ′ (3.2) 1−(ǫ+ǫ )/2 ǫǫ −1−(ǫ+ǫ )/2 ǫǫ (z − wq ) (z − wq ) , (α |α ) = 2,  i j ǫ+ǫ − 1 ǫǫ f(z − wq ) , (α |α ) = − . i j ′ ′ ǫ ǫ ǫ ǫ X (z)X (w) =: X (z)X (w) : kǫ ,s lǫ ,s kǫ ,s lǫ ,s 1 2 1 2  1, (α |α ) = 0, k l ′ ′ −(ǫ+ǫ )/2 ǫǫ ǫ /2 ǫ /2 ǫ ǫ 1 2 1 2 (z − q w) (q z − q w) , (α |α ) = 1, (3.3) k l  ǫ ǫ ǫ+ǫ ′ 1 2 − 1 ǫǫ ǫ ǫ (1−ǫǫ )/2 1 2 1 2 2 2 f(z − q w) f(q z − q w) (−1) , (α |α ) = − k l ′ ′ ′ ǫ ǫ ǫ ǫ ǫǫ +ǫ ǫ ′ 1 2 (3.4) X (z)X (w) =: X (z)X (w) : z , for s 6= s , ′ ′ kǫ ,s kǫ ,s kǫ ,s kǫ ,s 1 2 1 2 3/2 2 (xq /a;q ) ∞ −1/2 −1/2 where f(a − x) = a a q-analog of the infinite series (a − x) at x = 0 [17]. 1/2 2 (xq /a;q ) ′ ′ ′ ′ ǫ ǫ ǫ ǫ −(ǫ+ǫ )/2 −ǫǫ (3.5) Y (z)X (w) =: Y (z)X (w) : (z − q w) , (α |α ) = −1, i,s i,s k kǫ,s kǫ,s ′ ′ ′ ′ ǫ ǫ ǫ ǫ −(ǫ+ǫ )/2 −ǫǫ (3.6) X (z)Y (w) =: X (z)Y (w) : (z − q w) , (α |α ) = −1. i k kǫ,s i,s kǫ,s i,s 1−ǫǫ ′ 1 b (0) ǫ ǫ REMARK 3.2. The normal order product : X (z)X (w) : contains the term (−1) kǫ ,s lǫ ,s 1 2 1−ǫ ǫ b (0) 2b (0) l k (−1) . Also (−1) = 1. PROOF. It is enough to check (3.2), as (3.3) can be obtained similarly. Five cases need to be considered: (α |α ) = 0, 1,−1, 2 and − . Since all verifications are similar, we just give the case i j of (α |α ) = 1. It is clear that i j ′ ′ [ǫǫ n] w ′ ǫ ǫ ǫ ǫ n ǫǫ Y (z)Y (w) =: Y (z)Y (w) : exp − ( ) z i,s i,s i,s i,s ′ (ǫ+ǫ )/2 n[n] zq n≥1 ′ ′ ′ ǫ ǫ −(ǫ+ǫ )/2 ǫǫ =: Y (z)Y (w) : (z − wq ) . i,s i,s VERTEX REPRESENTATIONS OF QUANTUM N-TOROIDAL ALGEBRAS FOR TYPE C 9 4. Vertex representation of U (g ) q N,tor The purpose of this section is to give the main result and the proof. THEOREM 4.1. The map π from U (g ) to End(V) defined by q N,tor (s) (s) a (0) π(K ) = q , π(a (m)) = a (m), π(γ ) = q, i s i i (s) (s) (s) (s) π(φ (z)) = Φ (z), π(ϕ (z)) = Ψ (z), i i i i X X ± ± ± ± ± ± π(x (z)) = X (z) = X (z), π(x (z)) = X (z) = X (z), j,s j,s jǫ,s j j jǫ ǫ∈{±} ǫ∈{±} ± ± ± ± ± ± π(x (z)) = X (z) = Y (z), π(x (z)) = X (z) = Y (z), k,s k,s k,s k k k i ∈ I, j ∈ I = {1,··· , n − 1}, k ∈ {0, n} gives a representation of U (g ). a q N,tor REMARK 4.2. Since our construction already obeys the relations (2.1)-(2.4), we only need to check the relations (2.5)-(2.9). 4.1. Proof of relations (2.5) and (2.6). As relation (2.6) is similar to relation (2.5), it is (s) enough to check relation (2.5). In fact, by moving Ψ (z ) across Y (w) we have that i j (s) Ψ (z )Y (w) i j −la la ij ij q − q q z s (s) −(α |α ) i i l + i j = q exp( ( ) )Y (w)Ψ (z ) j i l w l>0 s (s) = g q Y (w)Ψ (z ). ij s j i 4.2. Proof of relation (2.7). We just show relation (2.7) for the short roots, i.e. i ∈ I . For s 6= s , we have that + + + + 1+ǫ ǫ 1 2 X (z)X (w) = : X (z)X (w) : z . ′ ′ i,s iǫ ,s i,s iǫ ,s 1 2 ǫ ,ǫ 1 2 + + Therefore lim [X (z), X (w)] = 0. z→w ′ i,s i,s 4.3. Proof of relation (2.8). There are five cases to consider: (α |α ) = 0,− ,±1, and 2. i j The first and the last are trivial. Now let (α |α ) = −1, by (3.2)-(3.3) it follows that i j ± ± ± ± ∓1 −1 X (z)X (w) =: X (z)X (w) : (z − q w) . n,s n,s n−1,s n−1,s − − − − Since : X (z)X (w) := − : X (w)X (z) :, we get that n,s n,s n−1,s n−1,s ∓1 ± ± ∓1 ± ± (z − q w) X (z)X (w) = (q z − w) X (w) X (z). n,s n,s n−1,s n−1,s 10 JING, XU, AND ZHANG Take the case (α |α ) = − , i, j ∈ I . Note that f(a − x) −1 = f(a − x) and i j a q→q 3 1 2 2 2 2 (xq /a; q ) (xq /a; q ) 1 ∞ ∞ −1 −1 − −1 (4.1) f(a − x)f(a − q x) = a = (a − q x) , 1 1 2 2 2 2 (xq /a; q ) (xq /a; q ) ∞ ∞ 1 3 1 −1 2 2 2 2 2 f(a − q x) (xq /a; q ) (xq /a; q ) a − q x ∞ ∞ (4.2) = = . 1 5 1 − − 2 2 f(a − qx) 2 2 2 (xq /a; q ) (xq /a; q ) a − q x ∞ ∞ It follows from (3.2)-(3.3) that ǫ ǫ 1−ǫ 1 2 1 + + + + −1 ǫ ǫ 1 2 2 2 2 X (z)X (w) = : X (z)X (w) : f(z − q w)f(zq − wq ) (−1) . i,s j,s iǫ ,s jǫ ,s 1 2 ǫ ,ǫ 1 2 Using (4.1)-(4.2) we have that 1 1 1 − + + + + − + + 2 4 4 (z − q w)X (z)X (w) =: X X : q − : X X : q i,s j,s i+ j+ i− j− 1 1 1 1 + + − + + − 4 4 4 4 + : X X : (q z − q w)− : X X : (q z − q w) i+ j− i− j+ − + + = (q z − w)X (w)X (z), j,s i,s + + + + where we have used the fact that : X (z)X (w) := − : X (w)X (z) : or ε ε = −ε ε . i j j i iǫ jǫ jǫ iǫ 1 2 2 1 4.4. Proof of relation (2.9). To prove (2.9) is to check that −1 δ 1 wq 1 wq ij (s) (s) + − − 2 2 [ X (z), X (w) ] = Φ (wq )δ( ) − Ψ (wq )δ( ) . i,s j,s i i −1 z z q − q Note that the case of (α |α ) = 2 was known in type A, so we will check the cases (α |α ) = 1 i i i i and (α |α ) = − to show the new situation. i j + − + − For i ∈ I , writing : X X :=: X (z)X (w) : etc., by (3.2)-(3.3) we have that ′ ′ iǫ iǫ,s jǫ jǫ ,s 1 1 1 + − + − + − −1 − −1 2 2 2 X (z)X (w) =: X X : q + : X X : (z − w) (q z − q w) i,s i,s i+ i+ i+ i− 1 1 1 + − −1 − −1 + − − 2 2 2 + : X X : (z − w) (q z − q w) + : X X : q . i− i+ i− i− ± ± Modifying the normal order to •X (z)X (w)• by excluding zw, we get that i j −1 1 w q w + − + − [X (z), X (w)] = •X (z)X (w) • (δ( ) − δ( )) 1 1 i,s i,s i+,s i−,s z z 2 2 q − q 1 qw w + − + • X (z)X (w) • (δ( ) − δ( )) 1 1 i−,s i+,s z z 2 2 q − q −1 1 1 qw 1 q w (s) (s) 2 2 = Φ (wq )δ( ) − Ψ (wq )δ( ) , −1 i i z z q − q where we have used the property of the δ-function and the fact that (s) (s) + − 1/2 + − −1 −1/2 •X (z)X (qz)• = Φ (q z), •X (z)X (q z)• = Ψ (q z), i+,s i−,s i i−,s i+,s i + − + − •X (z)X (z)• = •X (z)X (z) • . i+,s i−,s i−,s i+,s VERTEX REPRESENTATIONS OF QUANTUM N-TOROIDAL ALGEBRAS FOR TYPE C 11 For the case (α |α ) = − , recalling (3.3) and (4.1) we obtain that i j 1 1 1 1 + − + − − + − − 4 4 4 4 X (z)X (w) = : X X : (q z − q w)− : X X : (q z − q w) i,s j,s i+ j− i+ j+ 1 1 + − − + − 4 4 + : X X : q − : X X : q . i− j− i− j+ 1 1 1 1 − + − + − − + − 4 4 4 4 X (w)X (z) = : X X : (q w − q z)− : X X : (q w − q z) j,s i,s j− i+ j+ i+ 1 1 − + − + − 4 4 + : X X : q − : X X : q . j− i− j+ i− + − Thus for (α α ) = −1/2, we have [X (z), X (w)] = 0 due to ε ε = −ε ε . i j i j j i i,s j,s 4.5. Proof of relation (2.10). We will check some cases of a = −1 = a , as the other 0,1 n,n−1 cases are similar. By (3.2)-(3.3) we have that + + + X (z )X (z )X (w) 1 2 0,s 0,s 1ǫ,s −2 (z − z )(z − q z ) 1 2 1 2 + + + =: X (z )X (z )X (w) : . 1 2 0,s 0,s 1ǫ,s −1 −1 (z − q w)(z − q w) 1 2 Subsequently + + + −1 + + + + + + X (z )X (z )X (w) − (q + q )X (z )X (w)X (z ) + X (w)X (z )X (z ) 1 2 1 2 1 2 0,s 0,s 1ǫ,s 0,s 1ǫ,s 0,s 1ǫ,s 0,s 0,s −1 −1 −1 −1 −1 −1 = F · (w − q z )(w − q z ) + [2](z − q w)(w − q z ) + (z − q w))(z − q w) 1 2 2 1 1 2 −1 −2 = F · (q − q )w(z − q z ), 2 1 + + + :X (z )X (z )X (w): 1 2 0,s 0,s 1ǫ,s −2 −2 where F = (z − z )(z − q z ). Observe that F · (z − q z ) is −1 −1 1 2 1 2 2 1 (z −q w)(w−q z ) i i antisymmetric in z , so the Serre relation holds in this case. When a = −1 = a (i, j ∈ I ), it follows from (3.2)-(3.3) that ij ji a + + + + + + + + + X (z )X (z )X (w) − [2] X (z )X (w)X (z ) + X (w)X (z )X (z ) 1 2 i 1 2 1 2 iǫ ,s iǫ ,s jǫ,s iǫ ,s jǫ,s iǫ ,s jǫ,s iǫ ,s iǫ ,s 1 2 1 2 1 2 + + + −1 ǫ /2 ǫ /2 ǫ ǫ 1 2 1 2 =: X (z )X (z )X (w) : (z − q z )(z q − z q ) 1 2 1 2 1 2 iǫ ,s iǫ ,s jǫ,s 1 2 ǫ ǫ ǫ +ǫ ǫ ǫ 1 2 1 2 −1 ǫ ǫ −1 ǫ ǫ 1− 1 2 2 2 2 2 2 f(z − q w)f(q z − q w) · f(z − q w)f(q z − q w) (−1) 1 1 2 2 ǫ ǫ ǫ +ǫ ǫ ǫ 1 2 1 −1 ǫ ǫ −1 ǫ ǫ 1− 1 2 2 2 2 2 2 +[2] f(z − q w)f(q z − q w) · f(w − q z )f(q w − q z ) (−1) i 1 1 2 2 ǫ ǫ ǫ+ǫ ǫ ǫ 1 2 1 −1 ǫ ǫ −1 ǫ ǫ 1− 1 2 2 2 2 2 2 (4.3) +f(w − q z )f(q z − q w) · f(w − q z )f(q w − q z ) (−1) . 1 1 2 2 We claim that the above is antisymmetric with respect to the indices, i.e. under the action: z → z , ǫ → ǫ , which will immediately imply the Serre relation in this case. In fact, there are 1 2 1 2 8 subcases to verify. We check one complicated subcase (ǫ , ǫ , ǫ) = (+,−, +). Then the above 1 2 expression (4.3) is simplified to 1 1 −1 − −1 2 2 1 1 (z − q z )(q z − q z ) 1 2 1 2 + + + − −1 2 2 : X (z )X (z )X (w) : (q − q )w(z − q z ), 1 2 Q 2 1 1 1 i+,s i−,s j+,s − − 2 2 (z − q w) · (w − q z ) i 1 i 12 JING, XU, AND ZHANG where we have used the following identity [20]: −1 (w − az )(w − az ) + (a + a )(z − aw)(w − az ) + (z − aw)(z − aw) 1 2 1 2 1 2 −1 2 = (a − a)w(z − a z ). 1 2 Similarly, (4.3) at (ǫ , ǫ , ǫ) = (−, +, +) is equal to 1 2 1 1 −1 − −1 2 2 (z − q z )(q z − q z ) 1 1 1 2 1 2 + + + − −1 2 2 : X (z )X (z )X (w) : (q − q )w(z − q z ). 2 1 1 2 Q 1 1 i−,s i+,s j+,s − − 2 2 (z − q w) · (w − q z ) i 2 This means that (4.3) is indeed antisymmetric under z → z , ǫ → ǫ . Therefore we have proved 1 2 1 2 that for i, j ∈ I , (α |α ) = −1/2 a i j + + + + + + Sym X (z )X (z )X (w) − [2] X (z )X (w)X (z ) 1 2 i 1 2 iǫ ,s iǫ ,s jǫ,s iǫ ,s jǫ,s iǫ ,s 1 2 1 2 1↔2 + + + +X (w)X (z )X (z ) = 0. 1 2 jǫ,s iǫ ,s iǫ ,s 1 2 Now we consider the case of a = −2 = a . By Prop. 3.1 we have that 1,0 n−1,n + + + + X (z )X (z )X (z )X (w) 1 2 3 1ǫ ,s 1ǫ ,s 1ǫ ,s 0,s 1 2 3 −1 ǫ /2 ǫ /2 ǫ ǫ i j i j (z − q z )(q z − q z ) i j i j i<j + + + + =: X (z )X (z )X (z )X (w) : . 1 2 3 1ǫ ,s 1ǫ ,s 1ǫ ,s 0,s 1 2 3 −1 (z − q w) Pulling out the common factor we get that h i k + + + + + (−1) X (z )··· X (z )X (w)X (z )··· X (z ) 1 k k+1 3 1ǫ ,s 1ǫ ,s 0,s 1ǫ ,s 1ǫ ,s 1 k k+1 3 k i k=0 = F (z − qw)(z − qw)(z − qw) + [3] (z − qw)(z − qw)(w − qz ) 1 2 3 1 1 2 3 +[3] (z − qw)(w − qz )(w − qz ) + (w − qz )(w − qz )(w − qz ) . 1 1 2 3 1 2 3 −1 2 2 3 2 3 = F · (q − q) w (z − (q + q )z + q z ) + w(z z − (q + q )z z + q z z ) , 1 2 3 1 2 1 3 2 3 ǫ /2 ǫ ǫ −1 ǫ /2 j i j (z −q z )(q z −q z ) i j i j 1≤i<j≤3 + + −3 where F =: X (z )··· X (w) : q . Now consider the ac- 1 −1 −1 1ǫ 0 (z −q w)(w−q z ) i i 1≤i≤3 −1 tion of S defined by σ(z , ǫ ) = (z , ǫ ). The factor F excluding (z − q z ) is 3 i i σ(i) σ(i) i j 1≤i<j≤3 antisymmetric under the action of S . Recall the following identity [20]: l(σ) 2 3 −1 −1 −1 (−1) (z − (q + q )z + q z )(z − q z )(z − q z )(z − q z ) = 0. 1 2 3 1 2 1 3 2 3 σ∈S Replacing z by (z z z )/z and factoring out an appropriate factor, we also have the following i 1 2 3 i identity l(σ) 2 3 −1 −1 −1 (−1) (z z − (q + q )z z + q z z )(z − q z )(z − q z )(z − q z ) = 0. 2 3 1 3 1 3 1 2 1 3 2 3 σ∈S 3 VERTEX REPRESENTATIONS OF QUANTUM N-TOROIDAL ALGEBRAS FOR TYPE C 13 Using these two identities, we have proved the following equation: m=3 h i k ± ± ± ± ± Sym (−1) X (z )··· X (z )X (w)X (z )··· X (z ) = 0, 1 k k+1 m 1ǫ ,s 1ǫ ,s 0,s 1ǫ ,s 1ǫ ,s 1 m k k+1 k 1 k=0 where Sym runs through the action of S on (z , ǫ ), and this then implies (2.10). m i i For the last relation, without loss of generality, we take i ∈ I in the “+” case as an example. Note that for s 6= s + + + − + + + − X (z )X (z )X (z )X (w) =: X (z )X (z )X (z )X (w) : ′ ′ 1 2 3 1 2 3 iǫ ,s iǫ ,s iǫ ,s iǫ,s iǫ ,s iǫ ,s iǫ ,s iǫ,s 1 2 3 1 2 3 −1 ǫ /2 ǫ /2 ǫ ǫ −1 ǫ ǫ ǫ ǫ i j i j 1 2 3 · (z − z q )(z q − z q ) (z z z ) (z z z ) . i j i j 1 2 3 1 2 3 1≤i<j≤3 − + When we move X (w) across X (z ) to the left, the contraction function only changes in the ′ i iǫ,s iǫ ǫ ǫ ǫ −1 1 2 ǫ ǫ 2 −1 1 ǫ +ǫ ǫ 3 −1 ǫ +ǫ +ǫ ǫ 3 2 3 1 2 3 last three factors respectively to (z z w) (z z w ) , (z w ) (z w ) , and (w ) (w ) . 1 2 1 1 2 1 Then we have that + + + − lim X (z )X (z )X (z )X (w) 1 2 3 iǫ ,s iǫ ,s iǫ ,s iǫ,s 1 2 3 z →w ǫ,ǫ ∈{±} + + − + −[3] X (z )X (z )X (w)X (z ) i 1 2 ′ 3 iǫ ,s iǫ ,s iǫ,s iǫ ,s 1 2 3 + − + + +[3] X (z )X (w)X (z )X (z ) i 1 ′ 2 3 iǫ ,s iǫ,s iǫ ,s iǫ ,s 1 2 3 − + + + −X (w)X (z )X (z )X (z ) ′ 1 2 3 iǫ,s iǫ ,s iǫ ,s iǫ ,s 1 2 3 X Y + + + − ǫ /2 ǫ /2 ǫ ǫ i j i j = lim : X (z )X (z )X (z )X (w) : (z q − z q ) 1 2 3 i j iǫ ,s iǫ ,s iǫ ,s iǫ,s 1 2 3 z →w 1≤i<j≤3 ǫ,ǫ ∈{±} ǫ ǫ ǫ ǫ ǫ −1 −1 1 2 3 ǫ −1 1 2 ǫ ǫ ×(z − z q ) (z z z ) (z z z ) − [3] (z z w) (z z w ) i j 1 2 3 i 1 2 1 2 3 1 2 2 −1 1 ǫ +ǫ ǫ 3 −1 ǫ +ǫ +ǫ ǫ 2 3 1 2 3 +[3] (z w ) (z w ) − (w ) (w ) = 0. i 1 ACKNOWLEDGMENT N. Jing would like to thank the support of Simons Foundation grant 523868 and NSFC grant 11531004. H. Zhang would like to thank the support of NSFC grant 11871325. References [1] S. Berman and R. V. Moody, Lie algebras graded by finite root systems and the intersection matrix algebras of Slowdowy, Invent. Math. 108 (1992), 323–347. [2] V. Chari and A. Pressley, A guide to quantum groups, Cambridge Univ. Press, Cambridge, 1994. [3] V. Chari and A. Pressley, Quantum affine algebras and their representations. Representations of groups (Banff, AB, 1994), CMS Conf. Proc., 16, Amer. Math. Soc. Providence, RI (1995), 59–78. [4] B. Feigin, M. Jimbo, T. Miwa and E. Mukhin, Branching rules for quantum toroidal gl(n), Adv. Math. 300 (2016), 229-274. 14 JING, XU, AND ZHANG [5] B. Feigin, M. Jimbo, T. Miwa and E. Mukhin, Representations of quantum toroidal gl , J. Algebra 380 (2013), 78–108. [6] I. Frenkel and N. Jing, Vertex representations of quantum affine algebras, Proc. Nat’l. Acad. Sci. USA. 85 (1998), 9373–9377. [7] I. B. Frenkel, N. Jing and W. Wang, Quantum vertex representations via finite groups and the McKay correspon- dence, Comm. Math. Phys. 211 (2000), 365–393. [8] I. Frenkel and V. G. Kac, Basic representations of affine Lie algebras and dual resonance models, Invent. Math. 62 (1980), 23-66. [9] Y. Gao and N. Jing, U (gl ) action on gl -modules and quantum toroidal algebras, J. Algebra 273 (2004), 320– N N [10] Y. Gao, N. Jing, L. Xia and H. Zhang, Quantum N-toroidal algebras and quantized GIM algebras of N-fold affinization, 2019, arXiv: 1907.06301. [11] S. Gautam and V. Toledano-Laredo, Yangians and quantum loop algebras, Selecta Math. (N.S.) 19 (2013), 271–336. [12] V. Ginzburg, M. Kapranov and E. Vasserot, Langlands reciprocty for algebric surfaces, Math. Res. Lett. 2 (1995), 147–160. [13] N. Guay and X. Ma, From quantum loop algebras to Yangians, J. Lond. Math. Soc. 86 (2012), 683–700. [14] D. Hernandez, Drinfeld coproduct, quantum fusion tensor category and applications, Proc. London Math. Soc. 95 (2007), 567–608. [15] D. Hernandez, Quantum toroidal algebras and their representations, Selecta Math. (N.S.) 14 (2009), 701–725. [16] N. Jing, Twisted vertex representations of quantum affine algebras, Invent. Math. 102 (1990), 663–690. [17] N. Jing, Higher level representations of the quantum affine algebra U (sl(2)), J. Algebra 182 (1996), 448–468. [18] N. Jing, Quantum Kac-Moody algebras and vertex representations, Lett. Math. Phys. 4 (1998), 261–271. [19] N. Jing, S.-J. Kang and Y. Koyama, Vertex operators between level one irreducible representations of the quantum (1) affine algebra U (D ), Comm. Math. Phys. 174 (1995), 367–392. (1) [20] N. Jing, Y. Koyama and K. C. Misra, Level one representations of quantum affine algebras U (C ), Sel. Math. q n New Ser. 5 (1999), 243–255. [21] N. Jing and K. C. Misra, Vertex operators for twisted quantum affine algebras, Trans. Amer. Math. Soc. 351 (1999), 1663–1690. [22] S. Kolb, Quantum symmetric Kac-Moody pairs, Adv. Math. 267 (2014), 395–469. b b [23] K. Miki, Toroidal and level 0 U (sl ) actions on U (gl )-modules J. Math. Phys. 40 (1999), 3191–3210. n+1 q q n+1 [24] K. Miki, Representations of quantum toroidal algebra U (sl , tor)(n > 2), J. Math. Phys. 41 (2000), 7079– q n+1 [25] K. Miki, Quantum toroidal algebra U (sl , tor) and R matrices, J. Math. Phys. 42 (2001), 2293–2308. q 2 [26] R. V Moody, S. E. Rao and T. Yokonuma, Toroidal Lie algebras and vertex representations, Geom. Ded. 35 (1990), 283-307. [27] S. Rao and R. Moody, Vertex representations for N-toroidal Lie algebras and a generalization of the Virasoro algebra, Comm. Math. Phys. 159 (1994), 239–264. [28] Y. Saito, Quantum toroidal algebras and their vertex representations, Publ. RIMS. Kyoto Univ. 34 (1998), 155–177. [29] P. Slodowy, Beyond Kac-Moody algebras, and inside. Lie algebras and related topics (Windsor, Ont., 1984), pp. 361–371, CMS Conf. Proc., 5, Amer. Math. Soc., Providence, RI, 1986. [30] M. Varagnolo and E. Vasserot, Schur duality in the toroidal setting, Comm. Math. Phys. 182 (1996), 469–484. DEPARTMENT OF MATHEMATICS, SHANGHAI UNIVERSITY, SHANGHAI 200444, CHINA DEPARTMENT OF MATHEMATICS, NORTH CAROLINA STATE UNIVERSITY, RALEIGH, NC 27695, USA, E-mail address: jing@math.ncsu.edu DEPARTMENT OF MATHEMATICS, SHANGHAI UNIVERSITY, SHANGHAI 200444, CHINA DEPARTMENT OF MATHEMATICS, SHANGHAI UNIVERSITY, SHANGHAI 200444, CHINA E-mail address: hlzhangmath@shu.edu.cn

Journal

MathematicsarXiv (Cornell University)

Published: May 31, 2020

There are no references for this article.