Skip to main content
Log in

Fused Braids and Centralisers of Tensor Representations of Uq(glN)

  • Published:
Algebras and Representation Theory Aims and scope Submit manuscript

Abstract

We present in this paper the algebra of fused permutations and its deformation the fused Hecke algebra. The first one is defined on a set of combinatorial objects that we call fused permutations, and its deformation is defined on a set of topological objects that we call fused braids. We use these algebras to prove a Schur–Weyl duality theorem for any tensor products of any symmetrised powers of the natural representation of Uq(glN). Then we proceed to the study of the fused Hecke algebras and in particular, we describe explicitly the irreducible representations and the branching rules. Finally, we aim to give an algebraic description of the centralisers of the tensor products of Uq(glN)-representations under consideration. We exhibit a simple explicit element that we conjecture to generate the kernel from the fused Hecke algebra to the centraliser. We prove this conjecture in some cases and in particular, we obtain a description of the centraliser of any tensor products of any finite-dimensional representations of Uq(sl2).

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Similar content being viewed by others

Data Availability

This manuscript has no associated data.

References

  1. Andersen, H.H., Lehrer, G., Zhang, R.: Cellularity of certain quantum endomorphism algebras. Pac. J. Math. 279(1–2), 11–35 (2015)

    Article  MathSciNet  MATH  Google Scholar 

  2. Chari, V., Pressley, A.: Quantum affine algebras and affine Hecke algebras. Pac. J. Math. 174(2), 295–326 (1996)

    Article  MathSciNet  MATH  Google Scholar 

  3. Crampe, N., Poulain d’Andecy, L.: Baxterisation of the fused Hecke algebra and R-matrices with gl(N)-symmetry. Lett. Math. Phys. 111, 92 (2021)

    Article  MathSciNet  MATH  Google Scholar 

  4. Crampe, N., Poulain d’Andecy, L., Vinet, L.: Temperley–lieb, Brauer and Racah algebras and other centralizers of su(2). Trans. Am. Math. Soc. 373(7), 4907–4932 (2020)

    Article  MathSciNet  MATH  Google Scholar 

  5. Curtis, C., Reiner, I.: Methods of Representation Theory with Applications to Finite Groups and Orders, vol. 1. Wiley, New York (1981)

    MATH  Google Scholar 

  6. Fulton, W.: Young Tableaux. With Applications to Representation Theory and Geometry. London Mathematical Society Student Texts, 35. Cambridge University Press, Cambridge (1997). x + 260 pp

    Google Scholar 

  7. Geck, M., Pfeiffer, G.: Characters of Finite Coxeter Groups and Iwahori-Hecke Algebras. London Mathematical Society Monographs, 21. Oxford University Press, Oxford (2000)

    MATH  Google Scholar 

  8. Goodman, R., Wallach, R.: Symmetry, Representations, and Invariants, vol. 255. Springer, Berlin (2009)

    Book  MATH  Google Scholar 

  9. Green, J.A.: Polynomial Representations of GLn. Lecture Notes in Mathematics, vol. 830. Springer, Berlin (1980)

    Google Scholar 

  10. Hoefsmit, P.: Representations of Hecke algebras of finite groups with BN-pairs of classical type. Ph.D. thesis, University of British Columbia (1974)

  11. Hong, J., Kang, S.-J.: Introduction to Quantum Groups and Crystal Bases. Graduate Studies in Mathematics, 42. American Mathematical Society, Providence (2002). xviii+ 307

    Google Scholar 

  12. Isaev, A., Ogievetsky, O.: On Baxterized solutions of reflection equation and integrable chain models. Nucl. Phys. B 760([PM]), 167–183 (2007)

    Article  MathSciNet  MATH  Google Scholar 

  13. Jones, V.F.R.: Baxterisation. Int. J. Mod. Phys. B 4, 701. Proceedings of “Yang–Baxter equations, conformal invariance and integrability in statistical mechanics and field theory”, Canberra, 1989 (1990)

  14. Kang, S. -J., Kashiwara, M., Kim, M.: Symmetric quiver Hecke algebras and R-matrices of quantum affine algebras. Invent. Math. 211(2), 591–685 (2018)

    Article  MathSciNet  MATH  Google Scholar 

  15. Kassel, C., Turaev, V.: Braid Groups. Graduate Texts in Mathematics, vol. 247. Springer, Berlin (2008)

    Google Scholar 

  16. Kuan, J.: Two Dualities: Markov and Schur–Weyl. International Mathematics Research Notices (2021), rnaa333

  17. Klimyk, A., Schmüdgen, K.: Quantum Groups and Their Representations. Springer, Berlin (2012)

    MATH  Google Scholar 

  18. Knuth, D.: Permutations, matrices, and generalized Young tableaux. Pac. J. Math. 34(3), 709–727 (1970)

    Article  MathSciNet  MATH  Google Scholar 

  19. Langlois-Rémillard, A., Saint-Aubin, Y.: The representation theory of seam algebras. arXiv:1909.03499 (2019)

  20. Lehrer, G., Zhang, R.: Strongly multiplicity free modules for Lie algebras and quantum groups. J. Algebra 306(1), 138–174 (2006)

    Article  MathSciNet  MATH  Google Scholar 

  21. Lehrer, G., Zhang, R.: A Temperley–Lieb analogue for the BMW algebra. In: Representation Theory of Algebraic Groups and Quantum Groups, pp 155–190. Birkhäuser, Boston (2010)

  22. Matsui, C.: Multi-state asymmetric simple exclusion processes. J. Stat. Phys. 158, 158–191 (2015)

    Article  MathSciNet  MATH  Google Scholar 

  23. Poulain d’Andecy, L.: Fusion formulas and fusion procedure for the Yang–Baxter equation. Algebr. Represent. Theor. 20, 1379 (2017)

    Article  MathSciNet  MATH  Google Scholar 

  24. Poulain d’Andecy, L.: Young tableaux and representations of Hecke algebras of type ADE. J. Comb. Algebra 1 4, 371–423 (2017)

    Article  MathSciNet  MATH  Google Scholar 

  25. Ram, A.: Skew shape representations are irreducible. In: Combinatorial and Geometric Representation Theory (Seoul, 2001), Contemporary Mathematics, vol. 325, pp 161–189. AMS, Providence (2003)

  26. Stanley, R.: Enumerative Combinatorics, vol. 2. Cambridge Stud. Adv. Math. (1999)

  27. Zinn-Justin, P.: Combinatorial point for fused loop models. Commun. Math. Phys. 272, 661–68 (2007)

    Article  MathSciNet  MATH  Google Scholar 

Download references

Acknowledgements

Both authors are partially supported by Agence National de la Recherche Projet AHA ANR-18-CE40-0001. N.Crampé warmly thanks the university of Reims for hospitality during his visit in the course of this investigation.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Nicolas Crampé.

Additional information

Presented by: Michela Varagnolo

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Appendices

Appendix A: Artin–Wedderburn Decompositions and Bratteli Diagrams

1.1 A.1 Semisimple Algebras and Algebras of the Form PAP

Artin–Wedderburn Decomposition of a Semisimple Algebra

Let A be a finite-dimensional semisimple algebra over \(\mathbb {C}\). Let S be an indexing set for a complete set of pairwise non-isomorphic irreducible representations of A. Then Artin–Wedderburn theorem asserts that we have the following isomorphism of algebras:

$$ A\cong \bigoplus_{\lambda\in S} \text{End}(V_{\lambda})\ , $$
(44)

where Vλ is a realisation of the irreducible representation corresponding to λ. The isomorphism is given naturally by sending aA to the endomorphism in End(Vλ) corresponding to the action of a on Vλ.

Algebras of the Form PAP and Their Representations

Let A be a \(\mathbb {C}\)-algebra with an idempotent P. Then the subset PAP = {PxP | xA} is an algebra with unit P. We recall very basic and classical facts on the algebra PAP which can be found for example in [9, §6.2].

Let ρ : A↦End(V ) be a representation of A and W := ρ(P)(V ) the image of the operator ρ(P). The subspace W is naturally a representation of the algebra PAP. Indeed W is obviously invariant under the action of any element the form ρ(PxP), and thus the action of PAP on W is given simply by restriction:

$$ \begin{array}{@{}rcl@{}} PAP & \to & \text{End}\left( W\right) \\ PxP & \mapsto & {\rho(PxP)}_{|_{W}}. \end{array} $$
(45)

From now on, we will always remove the map ρ from the notation, and keep the same notation for an element of an algebra and its action in a given representation.

Irreducible Representations and Semisimplicity

Let {Vλ}λS be a complete set of pairwise non-isomorphic irreducible representations of A. Then the set

$$\{P(V_{\lambda})\ |\ \lambda\in S \text{ and } P(V_{\lambda})\neq 0\}$$

is a complete set of pairwise non-isomorphic irreducible representations of the algebra PAP.

Moreover, if the algebra A is a finite-dimensional semisimple algebra then the algebra PAP is also a finite-dimensional semisimple algebra and its Artin–Wedderburn decomposition is

1.2 A.2 Primitive Central Idempotents and Ideals

Let A be a finite-dimensional semisimple algebra over \(\mathbb {C}\) and S an indexing set for a complete set of pairwise non-isomorphic irreducible representations of A.

Primitive Central Idempotents

Let λS. We define Eλ as the element of A corresponding under the Artin–Wedderburn decomposition of A to \(\text {Id}_{V_{\lambda }}\) in the component corresponding to λ and 0 in all other components. The set {Eλ}λS is a complete set of primitive central orthogonal idempotents of A, meaning that they are central, they sum to 1, they satisfy \(E_{\lambda }E_{\lambda ^{\prime }}=\delta _{\lambda ,\lambda ^{\prime }}E_{\lambda }\) and they cannot be written as the sum of two non-zero central idempotents.

In any representation W of A, the action of Eλ projects onto the isotopic component of W corresponding to λ. More precisely, if the decomposition of W into irreducible is then the action of E is the projection onto V m corresponding to this decomposition, that is:

Ideals and Quotients

From the decomposition (1), one sees immediately that ideals (and equivalently quotients) of A are in correspondence with subsets SS as follows:

$$I_{S'}:=\bigoplus_{\lambda\in S'} \text{End}(V_{\lambda})\ \ \ \ \ \ \text{and}\ \ \ \ \ \ A/I_{S'}\cong\bigoplus_{\lambda\in S\backslash S'} \text{End}(V_{\lambda})\ .$$

One set of generators of the ideal \(I_{S'}\) consists of the elements \(E_{\lambda }\) with λS.

Since we use it in Section ??, we recall that the ideal generated by an element xA is \(I_{S'}\) where S is the subset of irreducible representations such that x acts as a non-zero element.

1.3 A.3 Bratteli Diagram of a Chain of Algebras

Let {An}n≥ 0 be a family of algebras and assume that, for any n ≥ 0, there is a given injective map from An to An+ 1. We call these maps “inclusion maps” and we say that {An}n≥ 0, sometimes denoted as follows

$$A_{0}\subset A_{1}\subset A_{2}\subset\dots\ldots\subset A_{n}\subset A_{n+1}\subset\dots\ ,$$

forms a chain of algebras. The inclusion maps allow to consider elements of An as elements of An+ 1, and more generally of An+k for k ≥ 1, and in turn to consider An as a subalgebra of An+ 1, and more generally of An+k for k ≥ 1.

From the inclusion of An into An+ 1, any representation V of An+ 1 can be seen as a representation of An by restriction; we denote this representation of An by \(\text {Res}_{A_{n}}(V)\).

Bratteli Diagram of a Chain of Semisimple Algebras

Let {An}n≥ 0 be a chain of algebras and we assume here that the algebras An are finite-dimensional semisimple algebras, since this will always be our setting in this paper.

For an irreducible representation Vi of An+ 1, the restriction to An decomposes by semisimplicity into a direct sum of irreducible representations, namely,

$$\text{Res}_{A_{n}}(V_{i})\cong\bigoplus m_{j} W_{j}\ ,$$

where Wj are non-isomorphic irreducible representations of An and the numbers mj are called the multiplicities. The knowledge of such decomposition for any irreducible representation of An+ 1 for any n ≥ 0 are called the branching rules of the chain of algebras. Then the Bratteli diagram of the chain of algebras {An}n≥ 0 is the following graph:

  • The set of vertices is partitioned into subsets indexed by n ≥ 0. We call n the level. The vertices of level n are indexed by the (isomorphism classes of) irreducible representations of the algebras An.

  • The edges express the branching rules of the chain of algebras and they only connect vertices of adjacent levels. Let V be an irreducible representation of An and V an irreducible representation of An+ 1. Then there are m edges connecting the vertices indexed by V and V if and only if V appears in the decomposition of \(\text {Res}_{A_{n}}(V')\) with multiplicity m.

Graphically, we place all the vertices of a given level on an horizontal line, and we put the vertices of level n + 1 below the vertices of level n. We often think of the edges as going down from vertices of level n to vertices of level n + 1.

Let v,v be two vertices of the Bratteli diagram. A path from v to v is a sequence of vertices of the form \(v,v_{1},\dots ,v_{k-1},v'\), for some k ≥ 1, such that at each step of the sequence, the level increases by 1. In other words, a path from v to v, if it exists, is obtained by starting from v and following edges only in the downward direction to reach v.

The partial order ≤ on the set of vertices of a Bratteli diagram is defined by setting that vv if and only if v = v or there is a path from v to v.

Dimensions

We often add the following numerical information to the Bratteli diagram of the chain {An}n≥ 0: next to each vertex, we indicate the dimension of the corresponding irreducible representation.

Obviously, for n ≥ 1 and any vertex V of level n, this dimension can be obtained from the preceding level. Indeed it is the sum of the dimensions of the irreducible representations of level n − 1 connected to V (counted with multiplicity indicated by the number of edges).

Moreover, from the Artin–Wedderburn decomposition of the algebras An, we have that the dimension of the algebra An is the sum of the squares of the dimensions appearing at level n.

Example A.1

A standard example of a Bratteli diagram is the Young diagram, corresponding to the poset of partitions partially ordered by inclusion. It is the Bratteli diagram associated to the chain \(\{\mathbb {C}[\mathfrak {S}_{n}]\}_{n\geq 0}\) of the complex group algebras of the symmetric groups. The first levels are given in Appendix A.

1.4 A.4 Quotients of Bratteli Diagrams and Chains of Ideals

We keep our setting of a chain {An}n≥ 0 of finite-dimensional semisimple algebras. Let \(\mathcal {D}\) be the Bratteli diagram of the chain {An}n≥ 0 and let S be a subset of vertices of \(\mathcal {D}\).

Quotients of Bratteli Diagrams

We denote by 〈S〉 the set of vertices v such that there exists vS with vv (i.e. all vertices in S and all vertices connected by a path to them).

Definition A.2

We define \(\overline {\mathcal {D}}_{S}\) to be the diagram obtained from \(\mathcal {D}\) by removing all vertices v∈〈S〉 and keeping only the edges of \(\mathcal {D}\) which connect the remaining vertices.

We call the resulting diagram \(\overline {\mathcal {D}}_{S}\) the quotient of \(\mathcal {D}\) generated by S.

Obviously, the quotient \(\overline {\mathcal {D}}_{S}\) depends only on the set of vertices 〈S〉 generated by S. Hence, several choices of S can lead to the same quotient. There is a unique minimal choice Smin, which is the set of minimal elements (for the partial order ≤) in 〈S〉. We call Smin the minimal generating set for the quotient \(\overline {\mathcal {D}}_{S}\).

Example A.3

A standard example of a quotient of a Bratteli diagram is the following. Take the Bratteli diagram of the chain \(\{\mathbb {C}[\mathfrak {S}_{n}]\}_{n\geq 0}\) (see Appendix A below) and make the quotient generated by the vertex labelled by the partition . It is easy to see that the remaining vertices are the partitions with no more than two lines. The quotient is equal to the Bratteli diagram of the chain of Temperley–Lieb algebras.

Representation-Theoretic Meaning and Chains of Ideals

We will explain the name Bratteli diagram for \(\overline {\mathcal {D}}_{S}\), and its representation-theoretic meaning.

For every n ≥ 0, let Sn be the set of vertices of level n inside 〈S〉 (that is, the vertices of level n which have been removed from \(\mathcal {D}\)). To Sn corresponds an ideal In of An. We have that {In}n≥ 0 forms a chain of ideals in {An}n≥ 0:

$$ I_{0}\subset I_{1}\subset I_{2}\subset\dots\ldots\subset I_{n}\subset I_{n+1}\subset\dots\ , $$
(46)

which means that In, seen as a subset of An+ 1 using the inclusion map, is contained in In+ 1 for any n ≥ 0. This chain property follows from the fact that, by definition of 〈S〉, every edge starting from Sn ends in Sn+ 1 (see Remark A.4 below).

Now the quotient \(\overline {\mathcal {D}}_{S}\) has the following meaning:

  • The vertices of level n are in bijection with the irreducible representations of An/In.

  • The edges give the branching rules for the following restriction procedure: a representation V of An+ 1/In+ 1 can be seen as a representation of An+ 1 where In+ 1 acts as 0. Thus, from the inclusion AnAn+ 1, we can form the representation \(\text {Res}_{A_{n}}(V')\) of An. Now by the inclusion InIn+ 1, we have that In obviously acts as 0, and therefore \(\text {Res}_{A_{n}}(V')\) can be seen as a representation of An/In.

Remark A.4

  • The family of quotients {An/In}n≥ 0 does not necessarily form a chain of algebras, even if the ideals In form a chain of ideals. In fact, one can check that the quotients An/In form a chain of algebras for the natural inclusion maps x + Inx + In+ 1 if and only if we have, for any n ≥ 0, In = In+ 1An as subsets of An+ 1. This is stronger than the chain property for the ideals In.

  • The property In = In+ 1An ensuring that the quotients form a chain of algebras can be seen easily in the Bratteli diagram. We have that In = In+ 1An if and only if:

    $$v\in S_{n}\ \ \ \ \Longleftrightarrow\ \ \ \ \forall v' \text{of level} n+1, v\leq v' \text{implies} v'\in S_{n+1}.$$

    We note that the weaker property InIn+ 1 is equivalent to the single implication ⇒.

Algebraic Description of the Quotients A n/I n

To make explicit the situation we need in this paper, assume that the set Smin contains vertices of a single level, say N + 1 (the general case can be obtained by partitioning the set Smin according to the level, and applying this procedure to each part).

Denote by X a generator of the ideal IN+ 1 of AN+ 1 (for example the sum of the central idempotents corresponding to Smin). Then we have that for any nN + 1, the ideal In of An is generated by the element Xn which is the element X seen as an element of An by the inclusion (indeed, this element Xn is non-zero precisely in the correct set of irreducible representations of An, by definition of Smin and of a Bratteli diagram).

As a conclusion, we note that for all nN + 1, the algebra An/In is the quotient of An over the relation Xn = 0. We refer to Example 4.8 for the well-known example of Temperley–Lieb algebras.

Appendix B: Examples

2.1 B.1 The Chain of Hecke Algebras H n(q)

The Hecke algebra Hn(q) is the particular case of the algebra Hk,n(q) where \(\mathbf {k}=(1,1,\dots )\) is the infinite sequence of 1’s. The first levels of the Bratteli diagram for the chain of Hecke algebras {Hn(q)}n≥ 0 is shown below.

The shaded areas indicate the connections between the Hecke algebras Hn(q) and the centralisers of the representations (corresponding to \(\mathbf {k}=(1,1,1,\dots )\)) of Uq(glN). Namely, by deleting the vertices included in the shaded area labelled glN together with the edges touching them, we obtain the Bratteli diagram of the centralisers of Uq(glN). For example, if N = 2, the quotiented Bratteli diagram is the Bratteli diagram of the Temperley–Lieb algebras.

figure b

2.2 B.2 The Chain of Algebras H k,n(q) When \(\mathbf {k}=(2,2,2,\dots )\)

When \(\mathbf {k}=(2,2,2,\dots )\) is the infinite sequence of 2’s, the Bratteli diagram for the chain of algebras {Hk,n(q)}n≥ 0 begins as:

figure c

Note that there is no arrow from to even if μλ since λ/μ contains two boxes in the same column.

The shaded area indicates the connections between the fused Hecke algebras Hk,n(q) and the centraliser of the representations (corresponding to \(\mathbf {k}=(2,2,2,\dots )\)) of Uq(glN), as in the preceding example.

2.3 B.3 The Chain of Algebras H k,n(q) When \(\mathbf {k}=(3,1,1,1,\dots )\)

When \(\mathbf {k}=(3,1,1,1\dots )\), the Bratteli diagram for the chain of algebras {Hk,n(q)}n≥ 0 begins as (the shaded areas have a similar meaning as in the preceding examples):

figure f

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Crampé, N., d’Andecy, L.P. Fused Braids and Centralisers of Tensor Representations of Uq(glN). Algebr Represent Theor 26, 901–955 (2023). https://doi.org/10.1007/s10468-022-10116-7

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s10468-022-10116-7

Keywords

Mathematics Subject Classification (2010)

Navigation