Types of graded tensor products of graded von Neumann algebras

Jumpei Tanaka · 2025

Read Types of graded tensor products of graded von Neumann algebras by Jumpei Tanaka (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-2508-10084.txt.

Introduction

Jumpei Tanaka filled in part of a multiplication table for operator algebras in a 2025 paper. Von Neumann algebras come in types that describe their projection structure, and a classic table records the type of a tensor product of two such algebras. Tanaka studies the graded version, where algebras carry an extra sign grading and the tensor product respects it, and works out the resulting types in special cases.

Catalog record

CREATORJumpei Tanaka
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE64 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2508-10084

Browse the full library · About the Tower of Babel

Full text

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

Types of graded tensor products of graded von Neumann algebras Jumpei Tanaka Abstract There is a famous multiplication table of types of tensor product of two von Neumann algebras. We filled out the multiplication table of graded tensor product of two graded von Neumann algebras in special cases. 1 introduction It is famous fact that, the types of tensor product of two von Neumann algebras are as in the multiplication table below. In this paper, we consider the types of graded tensor product of graded von Neumann algebras. In section 2, we check the definitions of some kinds of graded von Neumann algebras, and introduce some theorems we use in this paper. In section 3.1 , we prove that if at least one graded von Neumann algebra is of type III, the types of graded tensor product is equal to that of normal cases. In section 3.2 , we consider the case both (\mathcal{R}_{1},\ \text{Ad}_{\Gamma_{1}}) , (\mathcal{R}_{2},\ \text{Ad}_{\Gamma_{2}}) are central and both \mathcal{R}_{1} , \mathcal{R}_{2} are not factors. Consequently, we proved they are equal to types of normal tensor product except the case both is of type I. In the case both is of type I, the index is twice as large as the index of normal tensor product. In section 3.3 , we study the case (\mathcal{R}_{2},\ \text{Ad}_{\Gamma_{2}}) is central, \mathcal{R}_{1} is a factor, and \mathcal{R}_{2} is not a factor. We get multiplication table which is the same as that of normal tensor product. In section 3.4 , we introduce other facts. type of \mathcal{S} type of \mathcal{R} I n , n:finite I n , n:infinite II 1 II ∞ III I m , m:finite I mn I mn II 1 II ∞ III I m , m:infinite I mn I mn II ∞ II ∞ III II 1 II 1 II ∞ II 1 II ∞ III II ∞ II ∞ II ∞ II ∞ II ∞ III III III III III III III 2 Graded von Neumann algebra and Graded tensor product In this section, we check fundamental facts about graded von Neumann algebra. We define some objects we study and introduce some lemmas we use later. Definition 2.1 . Let \mathcal{R} be a von Neumann algebra on a Hilbert space \mathcal{H} . Let \Gamma be a self-adjoint unitary on \mathcal{H} satisfies \text{Ad}_{\Gamma}(\mathcal{R})=\mathcal{R} . We set \theta=\text{Ad}_{\Gamma}|_{\mathcal{R}} , then we say (\mathcal{R},\ \theta) is a (spatially) graded von Neumann algebra with grading operator \Gamma . In general, graded von Neumann algebra is a pair (\mathcal{R},\theta) with \mathcal{R} a von Neumann algebra and \theta an involutive automorphism on \mathcal{R} . Given a graded von Neumann algebra (\mathcal{R},\ \theta) , we set \mathcal{R}^{(\sigma)}=\{x\in\mathcal{R}:\theta(x)=(-1)^{\sigma}x\},\>\sigma=0,\>1. Then \mathcal{R} is a direct sum of two self-adjoint \sigma- weakly closed linear subspaces as \mathcal{R}=\mathcal{R}^{(0)}\oplus\mathcal{R}^{(1)} . An element of \mathcal{R}^{(\sigma)} is said to be homogeneous of degree \sigma , or even/odd for \sigma=0 / \sigma=1 . For a homogeneous x , its degree is denoted by \partial x . For x\in\mathcal{R} , we set x^{(0)}:=\frac{x+\theta(x)}{2} , x^{(1)}:=\frac{x-\theta(x)}{2} . Let (\mathcal{R}_{1},\ \theta_{1}) and (\mathcal{R}_{2},\ \theta_{2}) be graded von Neumann algebras. A * -homomorphism \varphi:\mathcal{R}_{1}\to\mathcal{R}_{2} is a graded homomorphism if \varphi(\mathcal{R}_{1}^{(\sigma)})\subseteq\mathcal{R}_{2}^{(\sigma)} , ( \sigma=0 , 1 ). Balanced or Central graded von Neumann algebras are main object we study in this paper. Definition 2.2 . Let (\mathcal{R},\ \theta) be a graded von Neumann algebra. If \mathcal{R} has an odd self-adjoint unitary, we say (\mathcal{R},\ \theta) is balanced. If Z(\mathcal{R})\cap\mathcal{R}^{(0)}=\mathbb{C}I for the center Z(\mathcal{R}) of \mathcal{R} , we say (\mathcal{R},\ \theta) is central. Next, we define graded tensor product of graded von Neumann algebras. Let (\mathcal{R}_{1},\ \text{Ad}_{\Gamma_{1}}) and (\mathcal{R}_{2},\ \text{Ad}_{\Gamma_{2}}) be graded von Neumann algebras acting on Hilbert spaces \mathcal{H}_{1} , \mathcal{H}_{2} respectively with grading operator \Gamma_{1} , \Gamma_{2} . We define a product and involution on the algebraic tensor product \mathcal{R}_{1}\odot\mathcal{R}_{2} by \displaystyle(A_{1}\hat{\otimes}B_{1})(A_{2}\hat{\otimes}B_{2}) \displaystyle=(-1)^{\partial B_{1}\partial A_{2}}(A_{1}A_{2}\hat{\otimes}B_{1}B_{2}) (1) \displaystyle(A\hat{\otimes}B)^{*} \displaystyle=(-1)^{\partial A\partial B}A^{*}\hat{\otimes}B^{*} (2) for homogeneous simple tensors. The algebraic tensor product with this product and involution is denoted by \mathcal{R}_{1}\hat{\odot}\mathcal{R}_{2} . This is a * -algebra. For homogeneous A\in\mathcal{R}_{1} , B\in\mathcal{R}_{2} we set \pi(A\hat{\otimes}B)=A\Gamma_{1}^{\partial B}\otimes B then we get faithful representation of \mathcal{R}_{1}\hat{\odot}\mathcal{R}_{2} on \mathcal{H}_{1}\otimes\mathcal{H}_{2} . A von Neumann algebra generated by \pi(\mathcal{R}_{1}\hat{\odot}\mathcal{R}_{2}) is said to be a graded tensor product of (\mathcal{R}_{1},\ \mathcal{H}_{1},\ \Gamma_{1}) and (\mathcal{R}_{2},\ \mathcal{H}_{2},\ \Gamma_{2}) , and denote it by \mathcal{R}_{1}\hat{\otimes}\mathcal{R}_{2} . \mathcal{R}_{1}\hat{\otimes}\mathcal{R}_{2} is a graded von Neumann algebra with grading operator \Gamma_{1}\otimes\Gamma_{2} . The next four lemmas are Lemma A.1, A.2, A.4, A.5 of [ BO21 ] respectively. Lemma 2.3 . Let (\mathcal{R},\ \theta) be a balanced graded von Neumann algebra. Assume that \mathcal{R} is of type \mu and \mathcal{R}^{(0)} is of type \lambda , with some \mu,\lambda =I, II, III, and both of \mathcal{R} , \mathcal{R}^{(0)} have finite-dimensional centers. Then \mu=\lambda . Lemma 2.4 . Let (\mathcal{R},\ \theta) be a central graded von Neumann algebra. Then either Z(\mathcal{R})=\mathbb{C}I or Z(\mathcal{R}) has a self-adjoint unitary b\in Z(\mathcal{R})\cap\mathcal{R}^{(1)} such that Z(\mathcal{R})\cap\mathcal{R}^{(1)}=\mathbb{C}b. Lemma 2.5 . Let (\mathcal{R}_{1},\ \text{Ad}_{\Gamma_{1}}) and (\mathcal{R}_{2},\ \text{Ad}_{\Gamma_{2}}) be graded von Neumann algebras acting on \mathcal{H}_{1} , \mathcal{H}_{2} respectively with grading operator \Gamma_{1} , \Gamma_{2} . Let \mathcal{R}_{1}\hat{\otimes}\mathcal{R}_{2} be the graded tensor product of (\mathcal{R}_{1},\ \mathcal{H}_{1},\ \Gamma_{1}) and (\mathcal{R}_{2},\ \mathcal{H}_{2},\ \Gamma_{2}) . Then commutant (\mathcal{R}_{1}\hat{\otimes}\mathcal{R}_{2})^{\prime} of the graded tensor product is generated by (\mathcal{R}_{1}^{\prime})^{(0)}\odot\mathcal{R}_{2}^{\prime},\>(\mathcal{R}_{1}^{\prime})^{(1)}\odot\mathcal{R}_{2}^{\prime}\Gamma_{2}. Lemma 2.6 . Let (\mathcal{R}_{i},\ \text{Ad}_{\Gamma_{i}}) , (\mathcal{L}_{i},\ \text{Ad}_{W_{i}}) , i=1 , 2 , be graded von Neumann algebras on \mathcal{H}_{i} , and \mathcal{K}_{i} respectively. Let \alpha_{i}:\mathcal{R}_{i}\to\mathcal{L}_{i} , i=1 , 2 , be graded * -isomorphisms. Suppose that \mathcal{R}_{2} is either (hence \mathcal{L}_{2} as well) balanced or trivially graded. Let \mathcal{R}_{1}\hat{\otimes}\mathcal{R}_{2} be the graded tensor product of (\mathcal{R}_{1},\ \mathcal{H}_{1},\ \Gamma_{1}) and (\mathcal{R}_{2},\ \mathcal{H}_{2},\ \Gamma_{2}) . Let \mathcal{L}_{1}\hat{\otimes}\mathcal{L}_{2} be the graded tensor product of (\mathcal{L}_{1},\ \mathcal{K}_{1},\ W_{1}) and (\mathcal{L}_{2},\ \mathcal{K}_{2},W_{2}) . Then there exists a unique * -isomorphism \alpha_{1}\hat{\otimes}\alpha_{2}:\mathcal{R}_{1}\hat{\otimes}\mathcal{R}_{2}\to\mathcal{L}_{1}\hat{\otimes}\mathcal{L}_{2} such that (\alpha_{1}\hat{\otimes}\alpha_{2})(A\hat{\otimes}B)=\alpha_{1}(A)\hat{\otimes}\alpha_{2}(B) for all A\in\mathcal{R}_{1} , B\in\mathcal{R}_{2} . 3 Main result In this section, we prove main results. 3.1 The case at least one of \mathcal{R}_{1} , \mathcal{R}_{2} is of type III Proposition 3.1 . Let (\mathcal{R},\ \text{Ad}_{\Gamma}) be a graded von Neumann algebra. Then \mathcal{L}=\mathcal{R}^{(0)}+\mathcal{R}^{(1)}\Gamma is a von Neumann algebra, and (\mathcal{L},\ \text{Ad}_{\Gamma}) is a graded von Neumann algebra. Furthemore, \mathcal{L}^{(0)}=\mathcal{R}^{(0)},\>\mathcal{L}^{(1)}=\mathcal{R}^{(1)}\Gamma. < proof > It is trivial that \mathcal{L} is closed under addition, scalar multiplication, and involution * . Since (A^{(0)}+A^{(1)}\Gamma)(B^{(0)}+B^{(1)}\Gamma)=A^{(0)}B^{(0)}-A^{(1)}B^{(1)}+(A^{(0)}B^{(1)}+A^{(1)}B^{(0)})\Gamma\in\mathcal{L}, it is also closed under multiplication. Let A_{a}^{(0)}+A_{a}^{(1)}\Gamma\to B (WOT), then we get A_{a}^{(0)}=\frac{1}{2}(\text{Ad}_{\Gamma}(A_{a}^{(0)}+A_{a}^{(1)}\Gamma)+A_{a}^{(0)}+A_{a}^{(1)}\Gamma)\to\frac{B+\text{Ad}_{\Gamma}(B)}{2} (WOT) by weak-operator continuity of \text{Ad}_{\Gamma} . Accordingly A_{a}^{(1)}\Gamma\to\frac{B-\text{Ad}_{\Gamma}(B)}{2} (WOT) i.e. A_{a}^{(1)}\to\frac{B-\text{Ad}_{\Gamma}(B)}{2}\Gamma . Since \mathcal{R}^{(\sigma)} , \sigma=0 , 1 is WOT closed, \frac{B+\text{Ad}_{\Gamma}(B)}{2}\in\mathcal{R}^{(0)},\>\frac{B-\text{Ad}_{\Gamma}(B)}{2}\Gamma\in\mathcal{R}^{(1)}. Therefore B\in\mathcal{L} and \mathcal{L} is WOT closed. It is clear that \text{Ad}_{\Gamma} defines a grading of \mathcal{L} . Then it is also obvious that \mathcal{L}^{(0)}=\mathcal{R}^{(0)} , \mathcal{L}^{(1)}=\mathcal{R}^{(1)}\Gamma . \Box Proposition 3.2 . Let (\mathcal{R},\ \text{Ad}_{\Gamma}) be a graded von Neumann algebra. Then \mathcal{R} is *- isomorphic to \mathcal{R}^{(0)}\oplus\mathcal{R}^{(1)}\Gamma . < proof > Since (\mathcal{R},\ \text{Ad}_{\Gamma}) is a graded von Neumann algebra, \mathcal{R}^{(0)}\oplus\mathcal{R}^{(1)}\Gamma is a von Neumann algebra by 3.1 . Let V=\frac{1-i}{2}I+\frac{1+i}{2}\Gamma , then \displaystyle V^{*}V=VV^{*} \displaystyle=(\frac{1-i}{2}I+\frac{1+i}{2}\Gamma)(\frac{1+i}{2}I+\frac{1-i}{2}\Gamma) (3) \displaystyle=\frac{1}{2}I-\frac{i}{2}\Gamma+\frac{i}{2}\Gamma+\frac{1}{2}I (4) \displaystyle=I (5) so V is a unitary. We can define a map \mathcal{R}\twoheadrightarrow\mathcal{R}^{(0)}\oplus\mathcal{R}^{(1)}\Gamma by A\mapsto V^{*}AV for all A\in\mathcal{R} : \displaystyle V^{*}AV \displaystyle=(\frac{1+i}{2}I+\frac{1-i}{2}\Gamma)A(\frac{1-i}{2}I+\frac{1+i}{2}\Gamma) (6) \displaystyle=\frac{1}{2}A+\frac{i}{2}A\Gamma-\frac{i}{2}\Gamma A+\frac{1}{2}\Gamma A\Gamma (7) \displaystyle=\frac{1}{2}(A+\Gamma A\Gamma)+\frac{i}{2}(A\Gamma-\Gamma A) (8) \displaystyle=A^{(0)}+iA^{(1)}\Gamma\in\mathcal{R}^{(0)}\oplus\mathcal{R}^{(1)}\Gamma. (9) This map is a surjection and

Read the complete text