Decomposition Theorem for Product of Fundamental Crystals in Monomial Realization

of type · 2025

Read Decomposition Theorem for Product of Fundamental Crystals in Monomial Realization by of type (2025) 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-2507-21444.txt.

Introduction

Manal Alshuqayr and Toshiki Nakashima study what happens when two basic crystals of type C are multiplied as monomials rather than combined as tensors, in a 2025 paper. Crystals are combinatorial skeletons of quantum group representations, written here as Laurent monomials. Taking the plain product of two fundamental crystals, the authors show the result still carries a crystal structure. They describe how it breaks into irreducible crystals. The break-up rule differs from the tensor product rule, so the monomial product is a genuinely different operation.

Catalog record

CREATORof type
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE110 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2507-21444

Browse the full library · About the Tower of Babel

Full text

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

Decomposition Theorem for Product of Fundamental Crystals in Monomial Realization of type Abstract We consider a product of fundamental crystals of type C n C_{n} in monomial realization, where the product means a natural product of Laurent monomials, not a tensor product. Then we shall show that the product still holds a crystal structure and describe how it is decomposed into irreducible crystals, which is, in general, different from the decomposition rule for the tensor product of the fundamental crystals. 1 Introduction For a symmetrizable Kac-Moody algebra \mathfrak{g} , let U\hskip-1.49994pt_{q}(\mathfrak{g}) be the corresponding quantum group. It is known that any integrable highest weight U\hskip-1.49994pt_{q}(\mathfrak{g}) -module V has a crystal base (L,B,wt,\{\varepsilon_{i}\},\{\varphi_{i}\},\{{\tilde{e}}_{i}\},\{{\tilde{f}}_{i}\})_{i\in I} [ 5 , 6 ] , where I is the index set of simple roots of \mathfrak{g} , B is called a crystal of V , \varepsilon_{i},\varphi_{i} are integer valued-functions on B and \tilde{e}_{i} , \tilde{f}_{i} are called the Kashiwara operators on B . Roughly, we can say that crystal base is a basis for a U\hskip-1.49994pt_{q}(\mathfrak{g}) -module at q=0 and holds a remarkable property for the tensor product of modules, namely, if U\hskip-1.49994pt_{q}(\mathfrak{g}) -modules M_{i} ( i=1,2 ) have crystal bases (L_{i},B_{i}) ( i=1,2 ), then the tensor product M_{1}\otimes M_{2} holds a crystal base (L_{1}\otimes L_{2},B_{1}\otimes B_{2}) . The explicit actions of the Kashiwara operators on tensor product of crystal bases are given as follows (see Sect.2 for details): for b_{1}\in B_{1},\,b_{2}\in B_{2} , we have \displaystyle\tilde{e}_{i}(b_{1}\otimes\,b_{2}) \displaystyle=\left\{{\begin{array}[]{rcl}\tilde{e}_{i}b_{1}\otimes\,b_{2}&\mbox{ if}&\varphi_{i}(b_{1})\geq\varepsilon_{i}(b_{2}),\\ b_{1}\otimes\tilde{e}_{i}b_{2}&\mbox{ if}&\varphi_{i}(b_{1})<\varepsilon_{i}(b_{2}),\end{array}}\right. \displaystyle\tilde{f}_{i}(b_{1}\otimes\,b_{2}) \displaystyle=\left\{{\begin{array}[]{rcl}\tilde{f}_{i}b_{1}\otimes\,b_{2}&\mbox{ if}&\varphi_{i}(b_{1})>\varepsilon_{i}(b_{2}),\\ b_{1}\otimes\tilde{f}_{i}b_{2}&\mbox{ if}&\varphi_{i}(b_{1})\leq\varepsilon_{i}(b_{2}).\end{array}}\right. Indeed, using these rules, we consider a decomposition of tensor product of crystal bases into irreducible crystal bases (see e.g., [ 9 ] ), which describes the decomposition of tensor product of corresponding modules, that is, for any dominant weight \lambda,\,\mu , there exist dominant weights \lambda_{1},\mathellipsis,\lambda_{k} such that B(\lambda)\otimes B(\mu)\cong B(\lambda_{1})\oplus\mathellipsis\oplus B(\lambda_{k}), and then we obtain the decomposition V(\lambda)\otimes V(\mu)\cong V(\lambda_{1})\oplus\mathellipsis\oplus V(\lambda_{k}). There are several types of realization for crystal bases, e.g. , tableaux realization, path realization, geometric realization, polyhedral realization, etc . Here we will treat the “monomial realization” which is introduced by Nakajima [ 10 ] and refined by Kashiwara [ 3 ] . Let \mathcal{M} be the set of Laurent monomials in the variables Y_{i}(n)\,(i\in I,\,n\in\mathbb{Z}, see Sect.3): \mathcal{M}\coloneqq\left\{{\left.\prod\limits_{i\in I,n\in\mathbb{Z}}{Y_{i}(n)}^{{y}_{i}(n)}\,\right|}\,y_{i}(n)\in\mathbb{Z}\,{\hbox{vanish except finitely many }}(i,\,n)\right\}. Then we can define the crystal structures on this set of Laurent monomials. This {\mathcal{M}} or its subcrystals are called a monomial realization of crystals. In Sect.3, it is introduced that if a monomial Y satisfies the highest weight condition, then the irreducible component in \mathcal{M} including Y is isomorphic to the crystal B(wt(Y)) corresponding to the module V(wt(Y)) [ 3 ] as we mentioned above. In fact, by Corollary 3.6 if \mathfrak{g} is semi-simple, then \mathcal{M} is an infinite direct sum of irreducible crystal B(\lambda) ’s, namely, any irreducible component in \mathcal{M} is isomorphic to the crystal B(\lambda) for some dominant weight \lambda . Let us denote an irreducible component corresponding to a highest monomial Y by \mathcal{M}(Y) . The tensor product \mathcal{M}(Y)\otimes\mathcal{M}(Y^{\prime}) has a crystal structure. But it is not clear whether the set of products \mathcal{M}(Y)\cdot\mathcal{M}(Y^{\prime})\coloneqq\left\{M_{1}\cdot M_{2}|\,M_{1}\in\mathcal{M}(Y),M_{2}\in\mathcal{M}(Y^{\prime})\right\}, holds a crystal structure, where M_{1}\cdot M_{2} means a natural product of Laurent monomials. Let \mathcal{M}(Y_{i}(m)) (i\in I,\,m\in\mathbb{Z}) be one of monomial realizations for the fundamental crystal B(\Lambda_{i}) (in [ 2 ] the explicit forms are given for classical type \mathfrak{g} ), where note that the weight of Y_{i}(m) is equal to the fundamental weight \Lambda_{i} and \tilde{e}_{j}Y_{i}(m)=0 for any j\in I . In this article, for type C_{n} we consider the following problems (see [ 1 ] for type A_{n} ): (1) Dose the product \mathcal{M}(Y_{p}(m))\cdot\mathcal{M}(Y_{q}(1)) (p,q\in I,\,m\in\mathbb{Z}) hold a crystal structure? (2) If the answer for ( (1) ) is affirmative, describe the decomposition of the crystal \mathcal{M}(Y_{p}(m))\cdot\mathcal{M}(Y_{q}(1)) (m\geq 1 ). Indeed, the answer for ( (1) ) is affirmative, which is shown in Proposition 5.1 , and more general results are given in Corollary 5.3 ; for any m_{1},m_{2},\mathellipsis m_{l}\in\mathbb{Z} , p_{1},p_{2},\mathellipsis p_{l}\in I\,(l>0) , \mathcal{M}(Y_{p_{1}}(m_{1}))\cdot\mathcal{M}(Y_{p_{2}}(m_{2}))\cdot\mathellipsis\cdot\mathcal{M}(Y_{p_{l}}(m_{l})), holds a crystal structure. Here note that there might be some monomials M_{1},M_{2}\in\mathcal{M}(Y) and M^{\prime}_{1},M^{\prime}_{2}\in\mathcal{M}(Y^{\prime}) such that M_{1}\neq M^{\prime}_{1} , M_{2}\neq M^{\prime}_{2} and M_{1}\cdot M_{2}=M^{\prime}_{1}\cdot M^{\prime}_{2} . Therefore, it may happen on the number of monomials: \#(\mathcal{M}(Y)\cdot\mathcal{M}(Y^{\prime}))<\#(\mathcal{M}(Y)\otimes\mathcal{M}(Y^{\prime})), and then one can deduce that the decomposition for \mathcal{M}(Y)\cdot\mathcal{M}(Y^{\prime}) may differ from the one for the tensor product \mathcal{M}(Y)\otimes\mathcal{M}(Y^{\prime}) . Indeed, the answer for (2) is given in Theorem 6.4 . \displaystyle\mathcal{M}(Y_{p}(m))\cdot\mathcal{M}(Y_{q}(1)) \displaystyle\cong B(\Lambda_{p}+\Lambda_{q}) \displaystyle\oplus\bigoplus\limits_{\begin{subarray}{c}(a,c)\neq(p,q),(q,p)\\ 0\leq a\leq p,\,\,a+p\leq c+q,\,\,(p+q)-(a+c)\in 2\mathbb{Z}_{>0.}\\ a\leq c\leq n,\,\,a+q\leq c+p,\,\,\frac{p+q+c-a}{2}\leq n.\end{subarray}}B(\Lambda_{a}+\Lambda_{c})\cdot\mathcal{L}(m\geq\frac{q-p-a-c+2}{2}+n) \displaystyle\oplus\bigoplus\limits_{\begin{subarray}{c}(a,c)\neq(p,q),(q,p)\\ {0\leq a<p<c\leq n},\,\,a+p\leq c+q,\,\,p+q=a+c,\\ a+q\leq c+p,\,\,\frac{p+q+c-a}{2}\leq n.\end{subarray}}B(\Lambda_{a}+\Lambda_{c})\cdot\mathcal{L}(m\geq q-a+1), where \mathcal{L}(P)=1 if P is true and \mathcal{L}(P)=0 otherwise. By Proposition 4.4 , forgetting the condition in \mathcal{L} , then we get the decomposition rule for the tensor product \mathcal{M}(Y_{p}(m))\otimes\mathcal{M}(Y_{q}(1)) . Hence, by this theorem, if m is sufficiently large, we get that \mathcal{M}(Y_{p}(m))\cdot\mathcal{M}(Y_{q}(1))\cong B(\Lambda_{p})\otimes B(\Lambda_{q}) . However, if not we find that \mathcal{M}(Y_{p}(m))\cdot\mathcal{M}(Y_{q}(1)) is not necessarily isomorphic to B(\Lambda_{p})\otimes B(\Lambda_{q}) . Indeed, if m=1 , we know that \mathcal{M}(Y_{p}(m))\cdot\mathcal{M}(Y_{q}(1))\cong B(\Lambda_{p}+\Lambda_{q})\varsubsetneq B(\Lambda_{p})\otimes B(\Lambda_{q}) . The organization of the article is as follows: In Section 2, we will recall the theory of crystal bases, such as, the existence theorem, the tensor product theorem and crystal graphs. In Section 3, Nakajima’s monomial realization of crystals will be introduced ( [ 3 , 10 ] ), which is our main ingredient in the article. In Section 4, first, the decomposition theorem for a tensor product of the fundamental crystals B(\Lambda_{p})\otimes B(\Lambda_{q}) will be given in Lemma 4.1 . The explicit form of monomial realization for the fundamental crystal B(\Lambda_{i}) of type C_{n} will be described following [ 2 ] . In Section 5, we state the first main result, which is the answers for question (1). Indeed, we show that the product for fundamental crystals in monomial realization of type C_{n} holds a crystal structure and we will obtain several important results for monomial realization of fundamental crystal \mathcal{M}(Y_{p}(m)) . In particular, Theorem 5.10 is crucial to describe the decomposition of product \mathcal{M}(Y_{p}(m))\cdot\mathcal{M}(Y_{q}(1)) (m\geq 1) . In the last section, we state the second main result, which is the answers for question (2). In fact, we obtain the explicit form of the decomposition for product \mathcal{M}(Y_{p}(m))\cdot\mathcal{M}(Y_{q}(1)) (m\geq 1) , which is given in Theorem 6.3 and Theorem 6.4 . J.Kamnitzer et al. treated the similar problem in [ 8 ] , for simply-laced cases. Their approach is completely differs from ours but the result for type A_{n} might be the same. But their method may not be applied to multi-laced cases, like as type C_{n} . Acknowledgments : We appreciate M.Nakasuji for her helpful comments and suggestions. 2 Preliminaries Our basic setting is the same as the one in [ 1 , Sect.2] . Notation 2.1 . Let A=(a_{ij})_{i,j\in I} be the Cartan matrix for a simple Lie algebra \mathfrak{g} and let \mathbb{Q}(q) be the rational function field in q , where I=\{1,\mathellipsis,n\} . Let t be a Cartan subalgebra of \mathfrak{g} , \{\alpha_{i}\in t^{\ast}\}_{i\in I} the set of simple roots and \{h_{i}\in t\}_{i\in I} the set of simple coroots. We take an inner product (\,\,,\,\,) on t^{\ast} such that (\alpha_{i},\alpha_{i})\in\mathbb{Z}_{>0} and \langle h_{i},\lambda\rangle=2(\alpha_{i},\lambda)/(\alpha_{i},\alpha_{i}) for \lambda_{i}\in t^{\ast} . Let \{\Lambda_{i}\}_{i\in I} be the

Read the complete text