Zonal Spherical Functions of Quantum Symmetric Pairs

Philip Schlösser · 2026

Read Zonal Spherical Functions of Quantum Symmetric Pairs by Philip Schlösser (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2606-27901.txt.

Introduction

Philip Schlösser identifies the spherical functions of quantum symmetric pairs in 2026. Quantum symmetric pairs generalise symmetric spaces into quantum group theory. Their zonal and character spherical functions are special functions of representation theory. Schlösser shows they coincide with the symmetric Koornwinder–Macdonald polynomials. Those polynomials are a classical family in several variables. The identification gives the abstract functions a concrete polynomial form.

The proof reworks Letzter's 2004 methods. They are translated into modern conventions and right coideal subalgebras. The treatment extends beyond earlier work in two directions. Non-standard cases are covered. Cases with non-reduced restricted root systems are covered too. One Satake type resisted the method, the type called FII. For that last elusive case, a conjecture is provided. The paper thus nearly completes the identification programme for these spherical functions. It also aligns the notation of the older literature with current practice.

Catalog record

CREATORPhilip Schlösser
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE112 KB
LICENSEhttp://creativecommons.org/licenses/by-sa/4.0/
RIGHTSCC BY SA
SOURCEarXiv

Source: https://arxiv.org/abs/2606.27901

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Zonal Spherical Functions of Quantum Symmetric Pairs Thanks: My thanks to Stein Meereboer whom I could always bounce ideas off of. This work is funded by grant OCENW.M20.108 of the Dutch Research Council. Philip Schlösser Abstract . We identify the zonal and character spherical functions for quantum symmetric pairs with the symmetric Koornwinder–Macdonald polynomials. To this end, the methods of Letzter’s 2004 paper are translated to modern conventions and right coideal subalgebras, and extended to non-standard cases and cases with non-reduced restricted root systems. For the last elusive Satake type, 𝖥𝖨𝖨 \mathsf{FII} , a conjecture is provided. 2020 Mathematics Subject Classification Primary 33D80; Secondary 33D52, 17B37 1. Introduction The study of quantum symmetric pairs (QSP) and their spherical functions has advanced vastly in conventions, generality, and knowledge since one of its most foundational papers, Gail Letzter’s [ Let04 ] , was written. This paper’s aim is to translate the contents of loc.cit. into modern notation, provide more details in the proofs and a better understanding of the techniques that Letzter used, and also deliver on Letzter’s promise from loc.cit.’s introduction: show that these techniques work in the general case as well. We then use these techniques and combine them with results on rank-1 cases from the literature to identify the zonal spherical functions of QSPs with twisted Macdonald–Koornwinder polynomials for all finite-type QSPs except for type \mathsf{FII} for which we provide a conjecture. This happens in Theorem 7.2 (and Conjecture 6.4 ). In particular, we determine the zonal spherical functions for all Hermitean QSPs, i.e. those (\mathbf{U},\mathbf{B}) where \mathbf{B} has characters besides the counit \epsilon . This allows us to extend the results from [ Mee25 ] to non-standard parameters, to different parameters on the left and the right, and to different characters on the left and the right (Theorem 8.7 ). For specific choices of characters, the resulting functions can also be interpreted as \epsilon -symmetric Macdonald–Koornwinder polynomials (Proposition 8.9 ). Lastly, we provide a survey about which Koornwinder polynomials (i.e. Macdonald for a (\mathsf{C}_{n}^{\vee},\mathsf{C}_{n}) affine root system) can be realised as character spherical functions of a quantum symmetric pair. 1.1. Structure In Section 2 we introduce standard notation around QSPs, their root systems, and their spherical functions. In particular, we introduce the restriction procedure that produces symmetric polynomials from spherical functions. In Section 3 we study the centraliser Z_{\widetilde{\mathbf{U}}}(\widetilde{\mathbf{B}}) and use it to construct a commutative algebra of symmetric difference operators (the radial parts of Z_{\widetilde{\mathbf{U}}}(\widetilde{\mathbf{B}}) ) that are diagonalised by the restrictions of zonal spherical functions. Our goal is to show that these symmetric difference operators are Macdonald operators. In Section 4 we establish the existence of a power series p that determines the leading terms of the radial parts of Z_{\widetilde{\mathbf{U}}}(\widetilde{\mathbf{B}}) in a uniform manner. In Section 5 we establish that p can be computed from the p -functions of rank-1 QSPs. In Section 6 we compute these rank-1 p -functions from cases that are known from literature. In Section 7 we consider small cases (that always exist) in which the leading term of a radial part actually determines the entire radial part, so that we can compute these radial parts using our knowledge of p . This knowledge is sufficient to identify the joint eigenfunctions, i.e. the restrictions of ZSF, as symmetric Macdonald–Koornwinder polynomials. In the next section (Section 8 ), we complete the computation of character spherical functions from [ Mee25 ] by also considering non-standard cases and by considering different parameters and characters on the left and right. For specific non-standard cases, this also gives us an interpretation of character spherical functions as \epsilon -symmetric Koornwinder polynomials. For specific standard cases, this also allows us to classify the zonal spherical functions for quantisations of several non-symmetric spherical pairs. Finally, in Section 9 , we classify which parameter values of rank- n Koornwinder polynomials are “geometric” , i.e. can be realised as restrictions of character spherical functions. 2. General Definitions and Notations 2.1. Root System Let I:=\left\{1,\dots,n\right\} and let A=(a_{ij})_{i,j=1,\dots,n} be a symmetrisable Cartan matrix of finite type and let X,Y be two lattices equipped with a perfect pairing \langle\cdot,\cdot\rangle:Y\times X\to\mathbb{Z} , and with linearly independent elements \alpha_{1},\dots,\alpha_{n}\in X,\qquad h_{1},\dots,h_{n}\in Y such that \langle h_{i},\alpha_{j}\rangle=a_{ij}\qquad(i,j\in I). Let there be a symmetric bilinear form \cdot on X such that \forall i,j\in I\colon\quad\langle h_{i},\alpha_{j}\rangle=2\frac{\alpha_{i}\cdot\alpha_{j}}{\alpha_{i}\cdot\alpha_{i}}. This corresponds to a choice (I,\cdot) of Cartan datum ( [ Lus10 , §1.1.1] ), and a choice (X,Y) of X - and Y -regular root datum ( [ Lus10 , §2.2.1] ), where the embeddings of I\to X,Y are denoted by i\mapsto\alpha_{i},h_{i} , respectively. Furthermore, we write h:X\to Y\otimes_{\mathbb{Z}}\mathbb{Q} for the linear map mapping \lambda\in X to the unique element h\in Y\otimes_{\mathbb{Z}}\mathbb{Q} satisfying \forall\mu\in X\colon\quad\lambda\cdot\mu=\langle h,\mu\rangle. In particular, h_{\alpha_{i}}=\frac{\absolutevalue{\alpha_i}^{2}}{2}h_{i} . Assume that \absolutevalue{\alpha_i}^{2}\in 2\mathbb{Z} , so that h_{\alpha_{i}}\in Y . Write Q,Q^{\vee} for the sublattices of X,Y generated by \alpha_{i},h_{i} ( i\in I ), respectively. Write P^{\vee},P for the corresponding dual lattices (which are then quotients of Y,X , respectively). Let W be the corresponding Weyl group; it acts on X,Y and their sub- and quotient lattices Q,Q^{\vee},P,P^{\vee} . Let R=\bigcup_{i\in I}W\alpha_{i}\subseteq Q be the root system associated to A , and let Q^{+}:=\bigoplus_{i\in I}\mathbb{N}_{0}\alpha_{i} and R^{+}:=R\cap Q^{+} (similarly, R^{\vee},Q^{\vee+},R^{\vee+} ). Write furthermore X^{+}:=\left\{\mu\in X\mathrel{}\middle|\mathrel{}\langle Q^{\vee+},\mu\rangle\subseteq\mathbb{N}_{0}\right\} for the set of dominant elements of X . Let 2\rho\in X be the sum of positive roots. 2.2. Quantum Group Let q be an indeterminate and consider k:=\overline{\mathbb{C}(q)} . Let K be an invertible element of a k -algebra A ; for a,b\in\mathbb{Z} we define [a]_{q^{b}}:=\frac{q^{ab}-q^{-ab}}{q^{b}-q^{-b}},\qquad[K;a]_{q^{b}}:=\frac{Kq^{ab}-K^{-1}q^{-ab}}{q^{b}-q^{-b}} as well as [K]_{q^{b}}:=[K;0]_{q^{b}} if this does not lead to confusion. Define q_{i}:=q^{\frac{\absolutevalue{\alpha_i}^{2}}{2}} . Definition 2.1 . The Drinfel’d–Jimbo quantum group \mathbf{U} is the associative k -algebra generated by (E_{i})_{i\in I},\qquad(F_{i})_{i\in I},\qquad(K_{h})_{h\in Y} subject to the relations \displaystyle K_{0} \displaystyle=1 \displaystyle K_{h}K_{h^{\prime}} \displaystyle=K_{h+h^{\prime}} \displaystyle K_{h}E_{i} \displaystyle=q^{\langle h,\alpha_{i}\rangle}E_{i}K_{h} \displaystyle K_{h}F_{i} \displaystyle=q^{-\langle h,\alpha_{i}\rangle}F_{i}K_{h} \displaystyle\commutator{E_i}{F_j} \displaystyle=\delta_{ij}[K_{\alpha_{i}}]_{q_{i}} \displaystyle F_{ij}(E_{i},E_{j}) \displaystyle=0\qquad(i\neq j) \displaystyle F_{ij}(F_{i},F_{j}) \displaystyle=0\qquad(i\neq j), where F_{ij}(x,y):=\sum_{p+p^{\prime}=1-a_{ij}}\frac{(-1)^{p^{\prime}}}{[p]^{!}_{q_{i}}[p^{\prime}]^{!}_{q_{i}}}x^{p}yx^{p^{\prime}}, h,h^{\prime}\in Y,i,j\in I , and where K_{\mu} is understood to mean K_{h_{\mu}} for \mu\in X with h_{\mu}\in Y . Here, we used [p]^{!}_{q}:=[p]_{q}[p-1]_{q}\cdots[1]_{q} for p\in\mathbb{N}_{0} . We also define the completion \widetilde{\mathbf{U}} of \mathbf{U} to be the algebra generated by (E_{i})_{i\in I},\qquad(F_{i})_{i\in I},\qquad(K_{h})_{h\in Y\otimes\mathbb{Q}} subject to the same relations. Write \mathcal{A} for the vector space of matrix elements of finite-dimensional representations. There is a well-established way (e.g. [ Lus10 , §3.3.4] or [ MS25 , Lemma 3.2] ) of turning \mathbf{U} and \mathcal{A} into dual Hopf algebras with comultiplication \Delta , counit \epsilon , and antipode S . Sweedler notation will be used whenever appropriate. We define \mathbf{U}^{0} (resp. \widetilde{\mathbf{U}}^{0} ) to be the commutative k -subalgebra of \mathbf{U} (resp. \widetilde{\mathbf{U}} ) generated by K_{h} for h\in Y (resp. h\in Y\otimes\mathbb{Q} ). Write \mathbf{U}^{\pm} for the subalgebras generated by the E_{i} (resp. F_{i} ) for i\in I . We have \mathbf{U}\cong\mathbf{U}^{+}\otimes\mathbf{U}^{0}\otimes\mathbf{U}^{-},\qquad\widetilde{\mathbf{U}}\cong\mathbf{U}^{+}\otimes\widetilde{\mathbf{U}}^{0}\otimes\mathbf{U}^{-}. Write \ad,\mathrm{ad}_{\mathrm{r}} for the left- and right-adjoint actions of \mathbf{U} (resp. \widetilde{\mathbf{U}} ) on itself. They are given as follows: \ad(x)(y):=x_{(1)}yS(x_{(2)}),\qquad\mathrm{ad}_{\mathrm{r}}(x)(y):=S(x_{(1)})yx_{(2)}. For any \mathbf{U} -module V and \mu\in X let V_{\mu}:=\left\{v\in V\mathrel{}\middle|\mathrel{}\forall h\in Y:K_{h}v=q^{\langle h,\mu\rangle}v\right\}, the \mu -weight space . Applying this to the left-adjoint action of \mathbf{U} on itself, we obtain a decomposition \mathbf{U}=\bigoplus_{\mu\in Q}\mathbf{U}_{\mu} into weight spaces. For a group homomorphism \xi:Q\to

Read the complete text