Topological T-Duality for Twisted Tori

We apply the *-algebraic formalism of topological T-duality due to Mathai and Rosenberg to a broad class of topological spaces that include the torus bundles appearing in string theory compactifications with duality twists, such as nilmanifolds, as well as many other examples. We develop a simple pr...

Повний опис

Збережено в:
Бібліографічні деталі
Опубліковано в: :Symmetry, Integrability and Geometry: Methods and Applications
Дата:2021
Автори: Aschieri, Paolo, Szabo, Richard J.
Формат: Стаття
Мова:Англійська
Опубліковано: Інститут математики НАН України 2021
Онлайн доступ:https://nasplib.isofts.kiev.ua/handle/123456789/211176
Теги: Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
Назва журналу:Digital Library of Periodicals of National Academy of Sciences of Ukraine
Цитувати:Topological T-Duality for Twisted Tori. Paolo Aschieri and Richard J. Szabo. SIGMA 17 (2021), 012, 51 pages

Репозитарії

Digital Library of Periodicals of National Academy of Sciences of Ukraine
_version_ 1859994728149286912
author Aschieri, Paolo
Szabo, Richard J.
author_facet Aschieri, Paolo
Szabo, Richard J.
citation_txt Topological T-Duality for Twisted Tori. Paolo Aschieri and Richard J. Szabo. SIGMA 17 (2021), 012, 51 pages
collection DSpace DC
container_title Symmetry, Integrability and Geometry: Methods and Applications
description We apply the *-algebraic formalism of topological T-duality due to Mathai and Rosenberg to a broad class of topological spaces that include the torus bundles appearing in string theory compactifications with duality twists, such as nilmanifolds, as well as many other examples. We develop a simple procedure in this setting for constructing the T-duals starting from a commutative *-algebra with an action of ℝⁿ. We treat the general class of almost abelian solvmanifolds in arbitrary dimension in detail, where we provide necessary and sufficient criteria for the existence of classical T-duals in terms of purely group theoretic data, and compute them explicitly as continuous-trace algebras with non-trivial Dixmier-Douady classes. We prove that any such solvmanifold has a topological T-dual given by a *-algebra bundle of noncommutative tori, which we also compute explicitly. The monodromy of the original torus bundle becomes a Morita equivalence among the fiber algebras, so that these *-algebras rigorously describe the T-folds from non-geometric string theory.
first_indexed 2026-03-18T10:20:00Z
format Article
fulltext Symmetry, Integrability and Geometry: Methods and Applications SIGMA 17 (2021), 012, 51 pages Topological T-Duality for Twisted Tori Paolo ASCHIERI † 1†2†3 and Richard J. SZABO †1†2†4†5†6 †1 Dipartimento di Scienze e Innovazione Tecnologica, Università del Piemonte Orientale, Viale T. Michel 11, 15121 Alessandria, Italy E-mail: paolo.aschieri@uniupo.it †2 Arnold–Regge Centre, Via P. Giuria 1, 10125 Torino, Italy †3 Istituto Nazionale di Fisica Nucleare, Torino, Via P. Giuria 1, 10125 Torino, Italy †4 Department of Mathematics, Heriot-Watt University, Colin Maclaurin Building, Riccarton, Edinburgh EH14 4AS, UK E-mail: R.J.Szabo@hw.ac.uk †5 Maxwell Institute for Mathematical Sciences, Edinburgh, UK †6 Higgs Centre for Theoretical Physics, Edinburgh, UK Received June 30, 2020, in final form January 22, 2021; Published online February 05, 2021 https://doi.org/10.3842/SIGMA.2021.012 Abstract. We apply the C∗-algebraic formalism of topological T-duality due to Mathai and Rosenberg to a broad class of topological spaces that include the torus bundles appearing in string theory compactifications with duality twists, such as nilmanifolds, as well as many other examples. We develop a simple procedure in this setting for constructing the T-duals starting from a commutative C∗-algebra with an action of Rn. We treat the general class of almost abelian solvmanifolds in arbitrary dimension in detail, where we provide necessary and sufficient criteria for the existence of classical T-duals in terms of purely group theoretic data, and compute them explicitly as continuous-trace algebras with non-trivial Dixmier– Douady classes. We prove that any such solvmanifold has a topological T-dual given by a C∗- algebra bundle of noncommutative tori, which we also compute explicitly. The monodromy of the original torus bundle becomes a Morita equivalence among the fiber algebras, so that these C∗-algebras rigorously describe the T-folds from non-geometric string theory. Key words: noncommutative C∗-algebraic T-duality; nongeometric backgrounds; Mostow fibration of almost abelian solvmanifolds; C∗-algebra bundles of noncommutative tori 2020 Mathematics Subject Classification: 46L55; 81T30; 16D90 Dedicated to Giovanni Landi on the occasion of his 60th birthday Contents 1 Introduction 2 1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.2 Summary and outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 2 Crossed products and duality 5 2.1 Dynamical systems and their crossed products . . . . . . . . . . . . . . . . . . . . . . . . 5 2.2 Semi-direct products and group algebras . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9 2.3 Pontryagin duality and Fourier transform . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 2.4 Morita equivalence and Green’s theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12 This paper is a contribution to the Special Issue on Noncommutative Manifolds and their Symmetries in honour of Giovanni Landi. The full collection is available at https://www.emis.de/journals/SIGMA/Landi.html mailto:paolo.aschieri@uniupo.it mailto:R.J.Szabo@hw.ac.uk https://doi.org/10.3842/SIGMA.2021.012 https://www.emis.de/journals/SIGMA/Landi.html 2 P. Aschieri and R.J. Szabo 3 Topological T-duality and twisted tori 16 3.1 Twisted tori and their T-duals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16 3.2 T-duality in the category KK . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17 3.3 Computational tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 3.4 Topological T-duality for the torus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20 3.5 Topological T-duality for orbifolds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 4 Topological T-duality for almost abelian solvmanifolds 22 4.1 Mostow bundles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22 4.2 Almost abelian solvmanifolds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 4.3 C∗-algebra bundles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25 4.4 Rn-actions on Mostow bundles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27 4.5 Ry-actions: Circle bundles with H-flux . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29 4.6 Rz-actions: noncommutative torus bundles . . . . . . . . . . . . . . . . . . . . . . . . . . 31 5 Three-dimensional solvmanifolds and their T-duals 35 5.1 Mostow bundles and SL(2,R) conjugacy classes . . . . . . . . . . . . . . . . . . . . . . . . 35 5.2 Parabolic torus bundles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 5.3 Elliptic torus bundles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 5.4 Hyperbolic torus bundles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 References 49 1 Introduction 1.1 Background T-duality is a symmetry of string theory which relates distinct spaces that describe the same physics. It has presented a challenge to mathematics in finding a rigorous framework in which these ‘equivalences’ of spaces is manifest. It was realized early on that noncommutative geometry provides such a framework [18, 30], at least in the simplest cases of tori endowed with a trivial gerbe, where subsequently it was shown that T-duality is realised as Morita equivalence of noncommutative tori [5, 29, 38, 47, 48]. T-duality of spaces which are compactified on tori, or more generally torus bundles, can be explained topologically in terms of correspondence spaces which implement a smooth analog of the Fourier–Mukai transform [24]. In the correspondence space picture, T-duality transfor- mations are realised as homeomorphisms in the mapping class group of the fibres of doubled torus bundles. This gives rise to an isomorphism of K-theory groups, which are the groups of D-brane charges on the pertinent space; as this only concerns how topological data of the space change under T-duality, it is commonly refered to as ‘topological T-duality’, to distinguish it from the more physical notion of T-duality which also dictates how geometric data on the space should transform. It was shown by [32] that this can be reformulated in terms of the C∗-algebra of functions on the space by considering its crossed product by an action of the abelian Lie group Rn, leading to a general T-duality formalism that can be regarded as a noncommutative version of the topological aspects of the Fourier–Mukai transform; this version of T-duality is often called the ‘C∗-algebraic formulation’ of topological T-duality. The story becomes more interesting for spaces that are endowed with a non-trivial gerbe, which in string theory typically comprise torus bundles with ‘H-flux’. The gerbe can be en- coded in the data of a continuous-trace C∗-algebra with a non-trivial Dixmier–Douady class, which is a noncommutative algebra to which the formalism of topological T-duality was applied originally in [3, 32, 33]. In addition to relating spaces with different topologies, T-duality in string theory for such instances predicts the existence of ‘non-geometric’ spaces, called T-folds [25], which cannot be viewed as conventional Hausdorff topological spaces. In these instances the Topological T-Duality for Twisted Tori 3 correspondence space picture ‘geometrizes’ the action of T-duality. It was shown by [4, 32, 33] that the T-folds of [25] have a rigorous incarnation in noncommutative geometry as C∗-algebra bundles of noncommutative tori; necessary and sufficient conditions for the existence of ‘classical’ T-dual Hausdorff spaces were developed in terms of topological data, and explicit constructions of ‘non-classical’ T-duals as noncommutative torus bundles were given. These points of view were harmonised in [6], and in [7] a C∗-algebraic version of the correspondence space construc- tion was given. The explicit connections of these noncommutative torus bundles to the T-folds of [25] in the setting of noncommutative gauge theories on D-branes in T-folds was elucidated in [16, 19, 28, 31]. Topological T-duality and T-folds have also been studied rigorously from other approaches based on homotopy theory [8, 9] and on higher geometry [36]. In string theory, the simplest examples of torus bundles are sometimes called ‘twisted to- ri’ [13]; although this name is a misnomer, we continue to use it as it is convenient for our pur- poses. These are fibrations of n-dimensional tori Tn over a circle T which do not carry the extra data of a gerbe; they have monodromy in the mapping class group SL(n,Z) of the torus fibers. The simplest examples of these, the Heisenberg nilmanifolds, are T-dual to torus bundles with H-flux and also to T-folds, and they arise in the C∗-algebraic constructions of [32, 33]. However, there are other examples which do not have any classical dual with H-flux, and these are missed by the usual C∗-algebraic framework which starts from continuous-trace algebras. The simplest examples of these with n = 2 were studied in [28] in the language of noncommutative gauge theo- ries, where it was shown that the monodromy of the original torus bundles becomes a non-trivial Morita equivalence of the fiber noncommutative tori of the dual C∗-algebra bundle. As far as we are aware, these are new examples of noncommutative torus bundles which have not been rigorously studied in the mathematics literature, and the primary purpose of this paper is to fill this gap: starting from the C∗-algebra of functions on a twisted torus in any dimension, we give a rigorous construction of the topological T-duals in the C∗-algebraic framework and precisely describe the non-classical C∗-algebra bundles with their Morita equivalence monodromies. This includes some of the examples from [32, 33] based on topological T-duality applied to continuous- trace algebras, and the examples of [28] based on T-duality in noncommutative gauge theory, while at the same time it produces many new examples. In particular, we give a unified descrip- tion of the noncommutative torus bundles which are T-dual to twisted tori in any dimension. The noncommutative gauge theory on a D-brane comes with other moduli, in addition to the noncommutativity parameters, which also generally transform in a non-trivial way under the monodromies so as to leave the physics unchanged [28]. In the absence of other moduli, as in topological T-duality, the non-trivial Morita equivalences of the fibres of the C∗-algebra bundles require an interpretation akin to the topological monodromies, which act as homeomor- phisms in the mapping class group SL(n,Z) of the fibres of the original twisted torus. This is naturally achieved by considering our C∗-algebras as objects in a category where both the usual ∗-isomorphisms as well as Morita equivalences are realised as isomorphisms. This category is well-known to experts and all of our considerations of topological T-duality in this paper will take place therein. This perspective will also be advantageous for eventual rigorous consider- ations of noncommutative gauge theories on these C∗-algebra bundles in terms of projective modules, as well as for the treatments of D-branes in terms of their K-theory, though we do not pursue these further aspects in the present paper. 1.2 Summary and outline In this paper our starting point is a very general definition of a twisted torus TΛG as the quotient of a locally compact group G by a lattice ΛG in G; this definition encompasses the Tn-bundles over T discussed above, along with many other known examples from string theory. We re- gard TΛG as a ‘torus bundle without H-flux’, which is captured simply by the C∗-algebra of 4 P. Aschieri and R.J. Szabo functions C(TΛG ). This is ultimately the novelty of our approach, which leads to a simpler perspective on topological T-duality as compared to the approach of [32, 33] based on the more complicated continuous-trace algebras. Our approach uses similar techniques as those of [32, 33] for evaluating Morita equivalences of cross products by actions of Rn, though with a much sim- pler C∗-algebraic structure. In particular, in this paper we do not develop any new C∗-algebraic machinery as such, but instead we gather a fortuitously existing collection of results that enable us to explicitly identify both classical and non-classical T-duals of twisted tori with relatively straightforward algebraic techniques. On the other hand, the tradeoff for the simplicity of our framework is the absence of some key constructions from [4, 32, 33]. We have endeavoured throughout to provide a fairly self-contained, and at times pedagogical, presentation. For this reason we have collected all the key concepts and tools involving cross products of C∗-algebras and Morita equivalence in Section 2. Experts versed in C∗-algebra theory may safely skip this section. In Section 3 we give our definition of twisted tori TΛG and discuss the C∗-algebraic formula- tion of topological T-duality that we employ in this paper. We describe how T-dual C∗-algebras are naturally isomorphic when regarded as objects of the additive category KK that underlies Kasparov’s bivariant K-theory, and we adapt the construction of noncommuative correspon- dences from [7] as diagrams in this category. We spell out some simple techniques that we use to compute classical T-duals with H-flux, i.e., the T-dual algebra is a certain continuous-trace C∗- algebra with non-trivial Dixmier–Douady class, and more general techniques based on Green’s symmetric imprimitivity theorem which enable the computation of noncommutative T-duals. We illustrate our scheme on two well-known examples which have classical T-duals: we repro- duce the standard rules for T-duality of tori as well as the topology changing mechanism for T-duality of orbifolds of compact Lie groups G. In Section 4 we come to the main class of examples and results of this paper. We review the definition and topology of the special class of twisted tori provided by almost abelian solvman- ifolds, which are Tn-bundles over a circle T. Accordingly, we regard the algebra of functions C(TΛG ) as an object in the category RKKT of C∗-algebra bundles over T, where in particular fibrewise Morita equivalences are isomorphisms. T-duality in this category requires fibrewise actions of Rn, and in particular Rn-actions which act non-trivially on the base T would take the algebra out of the category RKKT. This means that the ‘essentially doubled spaces’ of [28], which arise from T-duality along the base circle and require a completely doubled formalism, are not considered in this paper; they require working in a different category, which we do not discuss here. We give necessary and sufficient criteria for the existence of classical T-duals with H-flux in this case which are based on simple algebraic data of the underlying group G, and we explicitly compute the corresponding continuous-trace C∗-algebras dual to any almost abelian solvmanifold TΛG satisfying these conditions. We further show that any such solvmanifold has a non-classical T-dual that is a C∗-algebra bundle of noncommutative n-tori over T, which we also compute explicitly; this rigorously confirms, in particular, arguments from string theory suggesting that non-geometric solutions result from T-duality on some six-dimensional almost abelian solvmanifolds [1]. Finally, Section 5 is devoted to explicit examples of the general formalism of Section 4. We apply our results to all three classes of three-dimensional solvmanifolds. We recover in this way a new perspective on the well-known T-duals of the Heisenberg nilmanifolds: the three-torus T3 with H-flux, and the basic noncommutative principal T2-bundle over T given by the group C∗- algebra of the integer Heisenberg group. Our general formalism also rigorously reproduces the noncommutative torus bundles from [28] which are T-dual to Euclidean solvmanifolds for the Z4 elliptic conjugacy class of SL(2,Z). We extend these results to give new examples of noncommu- tative torus bundles dual to Euclidean solvmanifolds for the Z2 and Z6 elliptic conjugacy classes, as well as to the Poincaré solvmanifolds. In particular, our formalism extends the C∗-algebraic Topological T-Duality for Twisted Tori 5 formulation of topological T-duality to the case of non-principal torus bundles, which have also been previously considered in [22]. 2 Crossed products and duality In this section we summarise some of the mathematical tools that we will use in this paper. A good reference for the material covered in the following is the book [52]. Throughout this paper, all topological spaces are assumed to be second countable (hence separable), locally compact and Hausdorff. 2.1 Dynamical systems and their crossed products Let G be a locally compact group and let X be a G-space, i.e., a topological space which is acted upon homeomorphically by G; we denote the G-action G×X → X by (γ, x) 7→ γ · x. The pair (X,G) is called a transformation group. A related concept is that of a dynamical system, which is a triple (A,G, α) consisting of an algebra A and a locally compact group G acting on A via a group homomorphism α : G → Aut(A), denoted γ 7→ (αγ : A → A) for γ ∈ G. In topological T-duality one usually requires A to be a C∗-algebra, in which case (A,G, α) is called a C∗-dynamical system. Two dynamical systems (A,G, α) and (B,G, β) are equivalent if there is an algebra isomorphism ϕ : A → B which intertwines the G-actions: ϕ ◦ αγ = βγ ◦ ϕ for all γ ∈ G. If A is a commutative C∗-algebra, then we call (A,G, α) a commutative dynamical system. In that case, by Gelfand duality A = C0(X) is the algebra of continuous functions vanish- ing at infinity on a topological space X equipped with the G-action α†γ ∣∣ X for γ ∈ G under the identification of points x ∈ X with irreducible representations of C0(X), which are one- dimensional and given by the point evaluation maps evx : A → A with evx(f) = f(x); then (X,G) is a transformation group. Conversely, given a transformation group (X,G), there is an associated commutative dynamical system (C0(X),G, α), where αγ(f)(x) = f ( γ−1 · x ) for γ ∈ G, f ∈ C0(X) and x ∈ X. In other words, there is a one-to-one correspondence between transformation groups and commutative C∗-dynamical systems. If the C∗-algebra A is not commutative, then we call (A,G, α) a noncommutative C∗- dynamical system. As usual, it is more useful to work with a representation rather than the abstract dynamical system itself. A covariant representation of a dynamical system (A,G, α) in a C∗-algebra B with multiplier algebra M(B) is a pair (Π,U) consisting of a homomorphism Π : A → M(B) and a unitary representation U : G→ M(B), γ 7→ Uγ which satisfy the compatibility condition Π ( αγ(a) ) = UγΠ(a)U−1 γ , for all γ ∈ G and a ∈ A. A natural choice is to take B = K(H) to be the C∗-algebra of compact operators on a separable Hilbert space H, which gives a representation Π : A → B(H) of the algebra A by bounded operators B(H) on H and a unitary representation U : G→ B(H) of the group G on H; in this case we call (Π,U) a covariant representation of (A,G, α) on H. When a group G acts on a space X, one is naturally interested in considering the quotient space X/G of G-orbits on X. When G acts freely and properly on X, this is described alge- braically by the algebra of functions C0(X/G). More generally, the subalgebra of G-invariant elements AG ⊆ A of a G-algebra A can be used to represent the quotient, even for G-actions with fixed points. A more general and systematic way of dealing with the effective algebraic “quotient” is through the crossed product algebra Aoα G for a dynamical system (A,G, α). For a transformation group (X,G), this description is particularly powerful in the cases where the 6 P. Aschieri and R.J. Szabo quotient X/G is not a Hausdorff space, while for a free and proper G-action it gives an algebra with the same spectrum X/G as the algebra C0(X/G) = C0(X)G of G-invariant functions on X. In order to define the crossed product algebra, we first define ‖f‖univ := sup (Π,U) ∥∥(Π oα U)(f) ∥∥ for compactly supported functions f ∈ Cc(G,A), where the supremum is taken over (possibly degenerate) covariant representations (Π,U) of (A,G, α) with (Π oα U)(f) := ∫ G Π ( f(γ) ) Uγ dµG(γ), and µG denotes the left invariant Haar measure on G. This defines a norm, called the universal norm, on the space Cc(G,A). Then the crossed product algebra A oα G is the completion (in the universal norm) of the algebra Cc(G,A) equipped with the convolution product (f ? f ′)(γ) := ∫ G f(γ′)αγ′ ( f ′ ( γ′−1γ )) dµG(γ′), (2.1) for all f, f ′ : G→ A. In general this is a noncommutative multiplication, even for commutative dynamical systems. Since A is a C∗-algebra, there is a ∗-structure on the convolution algebra defined by f †(γ) := ∆G(γ)−1αγ ( f ( γ−1 )∗) , where ∆G : G→ R+ is the modular function of G defined through ∆G(γ′) ∫ G f(γ) dµG(γ) = ∫ G f(γγ′) dµG(γ) for f ∈ Cc(G,A) and γ′ ∈ G. By the uniqueness of the left invariant Haar measure µG up to a positive constant, ∆G(γ′) is independent of f and ∆G is easily proven to be a continuous group homomorphism from G to the multiplicative group R+; it is trivial for abelian groups and for compact groups. When A = Cc(X) is the algebra of a commutative dynamical system, the convolution algebra Cc(G×X) consists of functions f : G×X → C and the convolution product reads as (f ? f ′)(γ, x) = ∫ G f(γ′, x)f ′ ( γ′−1γ, γ′−1 · x ) dµG(γ′), while the ∗-algebra structure is given by f †(γ, x) = ∆G(γ)−1f ( γ−1, γ−1 · x ) . The crossed product is a generalization of the usual group algebra C∗(G), the completion (in the universal norm) of Cc(G) which is recovered in the caseA = C (the C∗-algebra of a point) wherein αγ = idA : A → A for all γ ∈ G and (2.1) recovers the usual convolution product of functions on the group G. As explained in Section 2.2 below, the group C∗-algebra description illustrates the relation between crossed products and semi-direct products of groups (see Theorem 2.5). We also note that if a group G acts trivially on an algebra A, then Ao G ' A⊗ C∗(G). The crossed product can be thought of as a universal object for covariant representations of the dynamical system (A,G, α), in the following sense: Define the universal covariant represen- tation (Π,U) of (A,G, α) in Aoα G by( Π(a)f ) (γ) = af(γ) and ( Uγ′f ) (γ) = αγ′ ( f ( γ′−1γ )) , Topological T-Duality for Twisted Tori 7 for a ∈ A, f ∈ Cc(G,A) and γ, γ′ ∈ G. Then the universal property defining the crossed product implies that any covariant representation (Π,U) of (A,G, α) in a C∗-algebra B fac- tors through the universal covariant representation: There exists a unique homomorphism ϕ : M(Aoα G)→ M(B) such that Π = ϕ ◦Π and Uγ = ϕ(Uγ) for all γ ∈ G. If (Π,U) is a covariant representation of the dynamical system (A,G, α) on a Hilbert space H, then Φ(Π,U)(f) := (Π oα U)(f) defines a representation Φ(Π,U) : Cc(G,A) → B(H) of the crossed product A oα G as bounded operators on H. This is called the integrated form of the covariant representation (Π,U). In particular, it maps the convolution product (2.1) onto the operator product in the algebra B(H), Φ(Π,U)(f ? g) = Φ(Π,U)(f)Φ(Π,U)(g), and it is covariant in the sense that Φ(Π,U) ( iG(γ)(f) ) = UγΦ(Π,U)(f), where ( iG(γ)(f) ) (γ′) := αγ ( f ( γ−1γ′ )) for each γ, γ′ ∈ G and f ∈ Cc(G,A). Example 2.1 (noncommutative two-tori). The noncommutative torus is a fundamental example of a noncommutative space in both physics and mathematics. Its original incarnation [43] is a nice example of a crossed product construction, which will play a fundamental role later on in this paper. Let ( C(T),Z, τ θ ) be the commutative C∗-dynamical system where τ θ is induced through pullback by rotations of the circle T through a fixed angle θ ∈ R/Z: τ θn(f)(z) = f ( e 2πinθz ) , for n ∈ Z, f ∈ C(T) and z ∈ T. The resulting crossed product Aθ := C(T) oτθ Z is a called a rotation algebra, and for irrational values of θ it can be identified as a noncommu- tative two-torus T2 θ in the following way. By definition, the algebra Aθ is the universal norm completion of the convolution algebra Cc(Z× T), whose elements f = {fn}n∈Z can be regarded as sequences (with only finitely many nonvanishing terms) of functions fn : T→ C. The convolution product is given by (f ?θ g)n(z) := ∑ n′∈Z fn′(z)gn−n′ ( e 2πin′θz ) , and the ∗-algebra structure is f †n(z) := f−n ( e 2πinθz ) . Via the Fourier transformation f(z, w) := ∑ n∈Z fn(z)wn 8 P. Aschieri and R.J. Szabo for w ∈ T, we may regard the convolution algebra Cc(Z × T) as a subspace of the space of functions C ( T2 ) equipped with the star-product (f ?θ g)(z, w) = ∑ n∈Z (f ?θ g)n(z)wn. (2.2) After a further Fourier transformation fn(z) = ∑ m∈Z fm,nz m and some simple redefinitions of the Fourier series involved, the star-product (2.2) may be written in the form (f ?θ g)(z, w) = ∑ (m,n)∈Z2 ( ∑ (m′,n′)∈Z2 fm′,n′gm−m′,n−n′ e 2πi(m−m′)n′θ ) zmwn. This recovers the usual commutative pointwise multiplication of functions in C ( T2 ) for θ = 0. For θ 6= 0 it realizes the irrational rotation algebra Aθ as a deformation of the algebra of functions C ( T2 ) on a two-dimensional torus T2; it is equivalent to the usual strict deformation quantization of T2 whose star-product is a twisted convolution product on C ( T2 ) . In the language of covariant representations of the dynamical system (C(T),Z, τ θ), the crossed product Aθ is the universal C∗-algebra generated by two unitaries U and V satisfying the rela- tion [52, Proposition 2.56] UV = e−2πiθV U. (2.3) A concrete representation of Aθ on the Hilbert space H = L2(T) is given by defining U(f)(z) = zf(z) and V (f)(z) = f ( e 2πiθz ) . Example 2.2 (Noncommutative d-tori). The natural higher-dimensional generalization of Example 2.1 involves a skew-symmetric real d×d matrix Θ = (θij), see [45]. Then the non- commutative d-torus AΘ = TdΘ is the universal C∗-algebra generated by d unitaries U1, . . . , Ud satisfying the relations UiUj = e−2πiθijUjUi for i, j = 1, . . . , d. By [37, Lemma 1.5], every noncommutative torus TdΘ can be obtained as an iterated crossed product by Z in the following way. Let Θ|d−1 = (θij)1≤i,j≤d−1, and let U1, . . . , Ud−1 be the standard generators of AΘ|d−1 = Td−1 Θ|d−1 . Define a group homomorphism τ ~θ : Z→ Aut(AΘ|d−1 ) by τ ~θ n(Ui) = e 2πinθidUi, for n ∈ Z, where ~θ := (θid) ∈ Rd−1. Then there is an isomorphism of C∗-algebras AΘ ' AΘ|d−1 o τ~θ Z. (2.4) In the particular case where Θ|d−1 = 0d−1, we denote the corresponding noncommutative d-torus by A~θ = Td~θ , and (2.4) shows that it can be obtained by a crossed product of the commutative algebra of functions on a d−1-torus by an action of Z: A~θ ' C ( Td−1 ) o τ~θ Z. Topological T-Duality for Twisted Tori 9 2.2 Semi-direct products and group algebras Most of our considerations later on will focus on spaces that can be obtained from semi-direct products of groups. We will now explain the relation between crossed products and semi-direct products which will be useful for these examples. There are two ways to think about the semi-direct product construction: (1) Let G be a group with two subgroups N and H such that N is normal. If N∩H = {e} ⊂ G and every element of G can be written as a product of an element of N with an element of H, then we say that G is a semi-direct product of its subgroups N and H and we write G = NH. (2) Let N and H be two groups together with a left action ϕ : H → Aut(N) of H on N by automorphisms, which we denote by ϕh(n) = hn for h ∈ H and n ∈ N; in particular h(nn′) = hn hn′. We write HN to indicate that H acts on N from the left. The semi-direct product of N and H is the group Noϕ H defined to be the set N× H with the product (n, h) (n′, h′) = ( n hn′, hh′ ) . The inverse is then (n, h)−1 = ( h−1 n−1, h−1 ) . These two definitions are equivalent: Given subgroups N,H ⊂ G as in point (1), it follows that G ' N oAd H where Ad is the adjoint action: Adh(n) = hnh−1. On the other hand, every element of the group G = NoϕH defined in (2) can be written as (n, h) = (n, eH)(eN, h) and the subgroups N× {eH} and {eN} × H intersect only in the identity of G. If the action of H on N is trivial, i.e., ϕh = idN for all h ∈ H, then the semi-direct product reduces to the direct product Noϕ H = N× H. Later on we will need to consider the interplay between semi-direct products and quotient groups, which is provided by the simple Lemma 2.3. Let G = N oϕ H be a semi-direct product, and let V ⊂ N be a subgroup which is normal in G. Then the quotient group G/V is the semi-direct product (N/V)oϕV H, where ϕV is the action ϕ of H induced on the quotient group N/V. If N is a group, we write C∗(N) for the corresponding group C∗-algebra, i.e., for the crossed product C o N. If N is finite, then C∗(N) = C[N] is the linear space freely generated over C by the group elements, made into an algebra by linearly extending the product from N to C[N]; equivalently it is the algebra of continuous functions on N with the convolution product. Given a left H-action ϕ : H → Aut(N), there is an induced action ϕ∗ : H → Aut(C∗(N)) via pullback. For H and N finite the vector spaces C∗(N)×H and C∗(N×H) are canonically isomorphic, and it is straightforward to show that the corresponding crossed product and semi-direct product are related by Proposition 2.4. If N and H are finite groups, then C∗(N) oϕ∗ H ' C∗(Noϕ H). Proof. Note that C∗(N)oϕ∗ H = C[N]oϕ∗ H is the vector space of functions f : H→ C[N] with convolution product( f ?C[N]oϕ∗H f ′)(h) = ∑ h′∈H f(h′) ?C[N] ϕ ∗ h′ ( f ′ ( h′−1h )) , and using the convolution product in C[N] this can be written as( f ?C[N]oϕ∗H f ′)(n, h) = ∑ h′∈H ∑ n′∈N f(n′, h′)f ′ ( h′−1( n′−1n ) , h′−1h ) , which is easily seen to coincide with the convolution product in C∗(Noϕ H) = C[Noϕ H]. � 10 P. Aschieri and R.J. Szabo A more general result holds if N and H are locally compact groups with ϕ : H → Aut(N) a continuous action of H on N via automorphisms (i.e., (h, n) 7→ ϕh(n) is a continuous map from H × N to N). In this case the semi-direct product N oϕ H is a locally compact group (in the product topology on N×H) with N a closed normal subgroup and H a closed subgroup (see [52, Proposition 3.11] with A = C). The analogue of item (1) above also holds in the context of locally compact groups if G is σ-compact, and N and H are closed subgroups of G. The action β defining the C∗-dynamical system (C∗(N),H, β) and hence the crossed product C∗(N) oβ H is the composition of the pullback ϕ∗ of the action ϕ : H→ Aut(N) with the action σH : H→ R+ ⊂ Aut(C∗(N)) that enters the definition of the Haar measure on Noϕ H in terms of the Haar measures on N and H: If µN is a (left invariant) Haar measure on N, then the integral Ih(F ) = ∫ N F ( hn ) dµN(n) for F ∈ C∗(N) is left invariant, i.e., Ih(λn′F ) = Ih(F ) where (λn′F )(n) := F ( n′−1n ) for all n′ ∈ N (use invariance of the Haar measure under n → h−1 n′). Uniqueness of the Haar measure up to a positive constant then implies there exists a function σH : H→ R+ such that σH(h) ∫ N F ( hn ) dµN(n) = ∫ N F (n) dµN(n). (2.5) It is straighforward to see that σH is a group homomorphism and that it is continuous [52, Section 2]. The Haar measure µNoϕH on Noϕ H is then given by∫ NoϕH f(n, h) dµNoϕH(n, h) := ∫ H ∫ N f(n, h)σH(h)−1 dµN(n) dµH(h). This is trivially invariant under the left N-action, and it is also invariant under the left H-action (n, h) 7→ (1, h′)(n, h) = ( h′n, h′h ) , using (2.5) with F ( h′n ) := f ( h′n, h′h ) and recalling that h is fixed in (2.5). Theorem 2.5. Let N and H be locally compact groups and ϕ : H→ Aut(N) a continuous action of H on N. Define β : H → Aut(C∗(N)) by (βh′`)(n) = σH(h′)−1` ( h′−1 n ) for all h′ ∈ H and ` ∈ Cc(N). Then C∗(N) oβ H ' C∗(Noϕ H). For a full proof of Theorem 2.5 that takes into account the topological and C∗-algebraic aspects, see [52, Proposition 3.11]. Here we shall just show that under the canonical injection Cc(N oϕ H) ↪→ Cc(N) oβ H, given by f(n, h) 7→ f(h) where f(h)(n) = f(n, h), the convolution product in Cc(N oϕ H) is mapped to the convolution product in Cc(N) oβ H. Let f, f ′ ∈ Cc(Noϕ H), then( f ?Cc(NoϕH) f ′)(n, h) = ∫ H ∫ N f(n′, h′)f ′ ( (n′, h′)−1(n, h) ) σH(h′)−1 dµN(n′) dµH(h′). (2.6) On the other hand, for the images of f , f ′ in Cc(N) oβ H we have( f ?Cc(N)oβH f ′)(h) = ∫ H f(h′) ?C∗(N) βh′ ( f ′ ( h′−1h )) dµH(h′). Using the convolution product in C∗(N) this can be written as( f ?Cc(N)oβH f ′)(h)(n) = ∫ H ∫ N f(h′)(n′)βh′ ( f ′ ( h′−1h ))( n′−1n ) dµN(n′) dµH(h′), which from the definition of the action β is easily seen to equal the image (f ?Cc(NoϕH) f ′)(h)(n) in Cc(N) oβ H of the product (f ?Cc(NoϕH) f ′)(n, h) in Cc(Noϕ H) from (2.6). More generally we have [52, Proposition 3.11] Topological T-Duality for Twisted Tori 11 Theorem 2.6. Let (A,Noϕ H, α) be a C∗-dynamical system for the semi-direct product group Noϕ H. Then (Aoα|N N,H, β) is a C∗-dynamical system, where β : H −→ Aut(Aoα|N N), h 7−→ βh is defined by ( βh(f) ) (n) = σH(h)−1αh ( f (h−1 n )) for all f ∈ Cc(N,A) ⊂ Aoα|N N, with σH : H→ R+ defined by (2.5) and h−1 n = ϕh−1(n). Moreover, the canonical injection Cc(N oϕ H,A) ↪→ Cc ( H, Cc(N,A) ) extends to a C∗-algebra isomorphism Aoα ( Noϕ H ) ' ( Aoα|N N ) oβ H. Theorem 2.5 is then recovered by setting A = C. In the spirit of Theorem 2.5, which shows that crossed products are a generalization of semi-direct products, let us mention the semi-direct product construction behind Theorem 2.6. Consider three groups M, N and H with group actions HN and NoHM; then there are also group actions NM and HM. The associativity of the triple semi-direct product construction is then easily established through Proposition 2.7. Let M, N and H be groups with group actions HM, HN and NM satisfying the compatibility conditions h ( nm ) = (hn) ( hm ) for all m ∈ M, n ∈ N and h ∈ H. Then there exists a group action NoHM defined by (n,h)m = n ( hm ) , and a group action H(Mo N) defined by h(m,n) = ( hm, hn ) , which together satisfy the associativity property Mo (No H) = (Mo N) o H. 2.3 Pontryagin duality and Fourier transform If N is a locally compact abelian group we denote by N̂ its Pontryagin dual, i.e., the set of characters χ : N→ U(1), which is also a locally compact abelian group (with the compact-open topology and with the pointwise multiplication). For example, if N = Rd then N̂ = Rd and the characters are given by χp(x) = e 2πi〈p,x〉 for x ∈ N and p ∈ N̂. The Pontryagin duality theorem states that there is a canonical isomorphism ̂̂ N ' N, where n ∈ N is associated to the character χ 7→ χ(n) on N̂. The Fourier transform shows that the group C∗-algebra C∗(N) is isomorphic to C0 ( N̂ ) : Given a Haar measure µN on N, the Fourier transform F(f) of f ∈ Cc(N) is defined by F(f)(χ) := ∫ N f(n)χ(n) dµN(n) for χ ∈ N̂. It sends the convolution product of functions in C∗(N) to the pointwise product of functions in C(N̂): F(f ? f ′) = F(f)F(f ′), and extends to an isomorphism [52, Proposition 3.1] F : C∗(N) '−−→ C0 ( N̂ ) , (2.7) where C0 ( N̂ ) is the algebra of functions on N̂ vanishing at infinity. For N separable, Hausdorff and locally compact, C0 ( N̂ ) is indeed a C∗-algebra. 12 P. Aschieri and R.J. Szabo Given a continuous left group action ϕ : H → Aut(N), which we also denote as before by ϕh(n) = hn, consider the induced action β : H→ Aut ( C∗(N) ) as defined in Theorem 2.5. There is also an induced left action ϕ̂ : H → Aut ( N̂ ) defined by pullback: ( ϕ̂hχ ) (n) := χ ( h−1 n ) , together with its pullback ϕ̂ ∗ : H → Aut ( C0 ( N̂ )) defined by ( ϕ̂ ∗h f̂ ) (χ) = f̂ ( ϕ̂h−1χ ) for all h ∈ H, f̂ ∈ C0 ( N̂ ) and χ ∈ N̂. The Fourier transform isomorphism (2.7) then extends to the isomorphism Proposition 2.8. If N is a locally compact abelian group and ϕ : H → Aut(N) is a continuous action of a locally compact group H on N, then C∗(N) oβ H ' C0 ( N̂ ) oϕ̂ ∗ H. Proof. We show that the triples (C∗(N),H, β) and ( C0 ( N̂ ) ,H, ϕ̂ ∗ ) are equivalent dynamical systems, see [52, Example 3.16]. For this, we prove that the Fourier transform (2.7) is H- equivariant with respect to the H-actions β and ϕ̂ ∗. For h ∈ H, f ∈ Cc(N) and χ ∈ N̂ we compute F ( βh(f) ) (χ) = ∫ N βh(f)(n)χ(n) dµN(n) = σH(h)−1 ∫ N f ( h−1 n ) χ(n) dµN(n) = σH(h)−1 ∫ N f ( h−1 n ) χ (h(h−1 n )) dµN(n) = ∫ N f(n)χ ( hn ) dµN(n) = F(f) ( ϕ̂h−1χ ) = ( ϕ̂ ∗hF(f) ) (χ), where in the fourth equality we used (2.5) with F ( h−1 n ) = f ( h−1 n ) χ (h(h−1 n )) . � Replacing N with N̂ in Proposition 2.8 we also obtain the isomorphism C∗ ( N̂ ) o β̂ H ' C0 (̂̂ N ) ô̂ϕ ∗ H ' C0(N) oϕ∗ H, (2.8) where β̂ : H → Aut ( C∗(N̂) ) is defined by β̂h(f̂ )(χ) = σ̂H(h)−1f̂ ( ϕ̂h−1χ ) for h ∈ H, f̂ ∈ Cc(N̂) and χ ∈ N̂, with σ̂H : H→ R+ defined as in (2.5) but using the dual group N̂ instead of N, and in the final isomorphism we used Pontryagin duality ̂̂ N ' N. Another important property of crossed products is Takai duality [52, Section 7.1]. If G is a locally compact abelian group and (A,G, α) is a C∗-dynamical system, then (Aoα G, Ĝ, α̂) is a C∗-dynamical system, where α̂ : Ĝ −→ Aut(Aoα G), χ 7−→ α̂χ is defined by α̂χ(f)(γ) := χ(γ) f(γ) for all f ∈ Cc(G,A). Theorem 2.9 (Takai duality). Let (A,G, α) be a C∗-dynamical system where G is a locally compact abelian group. Then there is an isomorphism of C∗-algebras (Aoα G) oα̂ Ĝ ' A⊗K ( L2(G) ) . 2.4 Morita equivalence and Green’s theorem Crossed products of algebras provide a host of examples of dualities which come in the form of various levels of strong and weak equivalences of algebras, see, e.g., [6]. The most primitive form of such dualities is provided by (strong) Morita equivalence [41]. A bimodule for a pair of algebras A and B is a vector space M which is simultaneously a left A-module and a right B-module, where the left action of A commutes with the right action of B: (a · ξ) · b = a · (ξ · b) Topological T-Duality for Twisted Tori 13 for all a ∈ A, b ∈ B and ξ ∈ M. If A and B are C∗-algebras, we say that a bimodule M is an A–B Morita equivalence bimodule (or imprimitivity bimodule) if it is equipped with an A-valued inner product A〈 · , · 〉 and a B-valued inner product 〈 · , · 〉B satisfying the associativity condition A〈ψ, φ〉 · ξ = ψ · 〈φ, ξ〉B, for all ψ, φ, ξ ∈ M, under which M is complete in the norm closures, and such that the ideal A〈M,M〉 is dense in A and 〈M,M〉B is dense in B. The bimodule M establishes a Morita equivalence between the algebras A and B, and in this case we write A ∼M B. Morita equivalent C∗-algebras have equivalent categories of nondegenerate ∗-representations: If ΠB : B → B(HB) is a representation of B on a Hilbert space HB, then we can construct another Hilbert space HA :=M⊗B HB which is the quotient of the tensor productM⊗HB by the relation (ξ · b)⊗ψ− ξ⊗ΠB(b)ψ = 0 identifying the B-actions for ξ ∈M, b ∈ B and ψ ∈ HB. The inner product on HA is given by〈 ξ ⊗B ψ ∣∣ξ′ ⊗B ψ′〉HA := 〈 ψ ∣∣ΠB(〈ξ, ξ′〉B)ψ′〉HB , and a representation ΠA : A → B(HA) of the algebra A is defined by ΠA(a)(ξ ⊗B ψ) = (a · ξ)⊗B ψ for a ∈ A and ξ ⊗B ψ ∈ HA; this representation is unitary equivalent to the representation ΠB. Conversely, starting with a representation of A, we can use a conjugate B–A equivalence bimod- uleM to construct a unitary equivalent representation of B; then there are surjective bimodule homomorphisms M⊗BM → A and M⊗AM → B which satisfy a certain transitivity law. As a particular consequence of this equivalence, Morita equivalent algebras have homeomorphic spectra and isomorphic K-theory groups. Example 2.10 (noncommutative two-tori). A famous example of Morita equivalence in both mathematics and string theory is provided by the noncommutative tori Aθ = T2 θ from Exam- ple 2.1. Firstly, notice from (2.3) that changing the coset representative θ ∈ R/Z yields an identical algebra: Aθ+m = Aθ for all m ∈ Z. Secondly, there is an obvious C∗-algebra isomor- phism A−θ ' Aθ obtained by interchanging the two generators U and V . The converse is also true [43, 44]: Aθ′ ' Aθ if and only if θ′ = θ mod 1. More generally, two irrational rotation C∗-algebras Aθ and Aθ′ are Morita equivalent if and only if θ and θ′ lie in the same orbit under the action of GL(2,Z) by fractional linear transformations θ′ = M[θ] := aθ + b cθ + d for M = ( a b c d ) ∈ GL(2,Z). The explicit Morita equivalence bimodules can be found in [43]. On the other hand, the rational rotation algebras Aθ are all Morita equivalent to the commutative algebra C ( T2 ) of functions on the two-torus [44]. Example 2.11 (noncommutative d-tori). The Morita equivalences of Example 2.10 generalize to the higher-dimensional noncommutative tori AΘ = TdΘ from Example 2.2 in the following way [46]. First of all, the algebra AΘ is unchanged if the matrix Θ is written in another basis of Zd: if B ∈ GL(d,Z) with transpose Bt, then there is a C∗-algebra isomorphism ABt ΘB ' AΘ. More generally, consider the set of real skew-symmetric d×d matrices Θ whose orbits M [Θ] are defined for all M ∈ SO(d, d;Z), where M [Θ] = (AΘ +B) (C Θ +D)−1 for M = ( A B C D ) ∈ SO(d, d;Z), 14 P. Aschieri and R.J. Szabo and A, B, C and D are d×d block matrices satisfying AtC + CtA = 0 = BtD +DtB and AtD + CtB = 1d. The set of all such matrices is dense in the space of all skew-symmetric real d×d matrices, and there is a Morita equivalence AM [Θ] ∼M AΘ. However, for d > 2 the converse is not generally true: In fact, there are algebras AΘ and AΘ′ that are isomorphic (and so Morita equivalent) but for which the matrices Θ and Θ′ do not belong to the same SO(d, d;Z) orbit [46]. We will also need an equivariant version of Morita equivalence in order to show that Morita equivalent algebras induce Morita equivalent crossed products according to [14, Section 5.4] Theorem 2.12. Let (A,G, α) and (B,G, β) be C∗-dynamical systems such that A and B are Morita equivalent. Then the crossed product C∗-algebras AoαG and BoβG are Morita equivalent if there exists a G-equivariant A–B Morita equivalence bimodule M, i.e., if there is a strongly continuous action U : G→ Aut(M) of G on an A–B Morita equivalence bimodule M such that Uγ(a · ξ) = αγ(a) · Uγ(ξ) and Uγ(ξ · b) = Uγ(ξ) · βγ(b), and A〈Uγ(ξ), Uγ(ξ′)〉 = αγ ( A〈ξ, ξ′〉 ) and 〈Uγ(ξ), Uγ(ξ′)〉B = βγ ( 〈ξ, ξ′〉B ) , for all γ ∈ G, ξ, ξ′ ∈M, a ∈ A and b ∈ B. In this paper, our main application of Morita equivalence will involve Green’s symmetric imprimitivity theorem. Let X be a locally compact space, and let H and K be locally compact groups with commuting free and proper actions on the right and on the left on X, respectively. We can lift these actions to left actions on C0(X) by defining (hf)(x) = f ( h−1 ·x ) and ( kf ) (x) = f(x · k) for all f ∈ C0(X), x ∈ X, h ∈ H and k ∈ K. Commutativity of the actions of H and K implies that there are well-defined induced actions of H and K respectively on the quotient spaces K\X and X/H, and hence respectively on the algebras C0(K\X) and C0(X/H) which we denote rt and lt. Green’s symmetric imprimitivity theorem then reads as [52, Corollary 4.10] Theorem 2.13. There is a Morita equivalence of C∗-algebras C0(K\X) ort H ∼M C0(X/H) olt K (2.9) implemented by the Morita equivalence (or imprimitivity) bimodule M which is the completion of Cc(X) with the actions (a · ξ)(x) = ∫ K a(k, x · H)ξ ( k−1 · x ) ∆K(k)1/2 dµK(k), (ξ · b)(x) = ∫ H ξ ( x · h−1 ) b ( h,K · x · h−1 ) ∆H(h)−1/2 dµH(h), for all x ∈ X, a ∈ Cc(K×X/H), b ∈ Cc(H× K\X) and ξ ∈ Cc(X), and the inner products A〈ξ, ξ′〉(k, x · H) = ∆K(k)−1/2 ∫ H ξ(x · h)ξ′ ( k−1 · x · h ) dµH(h), 〈ξ, ξ′〉B(h,K · x) = ∆H(h)−1/2 ∫ K ξ ( k−1 · x ) ξ′ ( k−1 · x · h ) dµK(k), for all x ∈ X, h ∈ H, k ∈ K and ξ, ξ′ ∈ Cc(X). Topological T-Duality for Twisted Tori 15 Theorem 2.13 has several useful applications and corollaries, see, e.g., [42]. A particularly relevant special case that we shall use below is when K is the trivial group, in which case (2.9) reduces to the Morita equivalence C0(X) ort H ∼M C0(X/H), illustrating the use of crossed products in describing quotients. This equivalence can in fact be strengthened to a stable isomorphism [42] C0(X) ort H ' C0(X/H)⊗K ( L2(H) ) , (2.10) where K denotes the algebra of compact operators. Example 2.14 (tori). A particularly relevant example for us is the case X = Rd with H = Zd acting by translations (n, x) 7→ x + n for n ∈ Zd and x ∈ Rd, which realizes the d-dimensional torus Td = Rd/Zd as a crossed product: C ( Td ) ∼M C0 ( Rd ) ort Zd. Let us illustrate the construction explicitly. The convolution algebra Cc ( Zd×Rd ) ⊂C0 ( Rd ) ortZd can be identified with the space of sequences f = {fn}n∈Zd of functions fn : Rd → C with the convolution product (f ? g)n(x) = ∑ m∈Zd fm(x)gn−m(x−m). Consider the algebra A := { f ∈ C0 ( Rd,K ( `2 ( Zd ))) ∣∣ f(x+m) = Umf(x)U−1 m } , where Um is the unitary shift operator on `2 ( Zd ) defined by (Uma)n = an−m for each m ∈ Zd and a = {an}n∈Zd . Define a map Φ : Cc ( Zd × Rd ) → Cc ( Rd,K ( `2 ( Zd ))) by( Φ(f)(x) ) mn = fm+n(x+ n). It is easy to see that Φ(f)(x+m) = UmΦ(f)(x)U−1 m for all f ∈ Cc ( Zd×Rd ) , and if a = (amn) ∈ A then defining fn := a0n gives Φ(f) = a. It is also easy to check that Φ(f ? g) = Φ(f)Φ(g), and consequently Φ gives an algebra isomorphism Φ : C0 ( Rd ) ort Zd '−−→ A. The explicit Morita equivalence bimodule is now obtained from the completion of M = { ξ ∈ Cc ( Rd, `2 ( Zd )) ∣∣ ξ(x+m) = Umξ(x) } . The left action of the algebra A = Φ ( Cc ( Zd × Rd )) is by left matrix multiplication on M: A×M −→M, (a, ξ) 7−→ a · ξ, (a · ξ)n = ∑ m∈Zd anmξm, while the right action of the algebra C ( Td ) is by right pointwise multiplication on M: M× C ( Td ) −→M, (ξ, b) 7−→ ξ · b, (ξ · b)n = ξnb. 16 P. Aschieri and R.J. Szabo The left and right inner products are respectively given by A〈ξ, η〉 = ξ ⊗ η∗, 〈ξ, η〉C(Td) = ∑ n∈Zd ξnηn, for ξ, η ∈ M. Together with the isomorphism Φ, this establishes a Morita equivalence between the algebras C0 ( Rd ) ort Zd and C ( Td ) . Another important class of examples is provided by taking X = G to be a locally compact group with closed subgroups K and H acting respectively by left and right multiplication on G. In particular, in the special case K = G, so that C0(K\G) = C, Theorem 2.13 gives a Morita equivalence between the commutative dynamical system (C0(G/H),G, lt), with K = G acting by left multiplication on the homogeneous space G/H (so that (ltγf)(x) = f ( γ−1x ) for all f ∈ C0(G/H), x ∈ G/H and γ ∈ G), and the group C∗-algebra C∗(H) = Co H: C0(G/H) olt G ∼M C∗(H). This equivalence also follows from the C∗-algebra isomorphism [52, Theorem 4.29] C0(G/H) olt G ' C∗(H)⊗K ( L2(G/H) ) . 3 Topological T-duality and twisted tori In this section we shall apply the results of Section 2, and in particular Green’s theorem, to present a scheme that will be employed in our study of topological T-duality. We shall then illustrate how our scheme works to reproduce some standard (commutative) examples of T-dual spaces. 3.1 Twisted tori and their T-duals We are interested in formulating a notion of T-duality for “torus bundles without H-flux”, which for our purposes can be characterised by the following general class of spaces. Definition 3.1 (twisted tori). Let G be a locally compact group which admits a cocompact discrete subgroup ΛG, i.e., a lattice in G, which we let act on G by left multiplication. The quotient space TΛG := ΛG\G is a twisted torus. By [34, Lemma 6.2], only unimodular groups can contain lattices, i.e., groups G whose mod- ular function ∆G is identically equal to 1. Example 3.2 (tori). Every lattice in the abelian Lie group G = Rd is isomorphic to ΛG = Zd, acting by translations. Then TZd = Zd\Rd =: Td is the d-dimensional torus. Example 3.3 (nilmanifolds). Generalizing Example 3.2, let G be a connected and simply- connected nilpotent Lie group. Then a theorem of Malcev [40, Theorem 2.12] establishes the existence of a lattice ΛG in G if and only if G can be defined over the rationals, i.e., there exists a basis for its Lie algebra which has rational structure constants, and in this case TΛG = ΛG\G is a nilmanifold. Topological T-Duality for Twisted Tori 17 Example 3.4 (orbifolds). Let G be a compact Lie group. Then the lattices in G are precisely the finite subgroups Γ of G, and TΓ = Γ\G is a smooth orbifold. For instance, for G = SU(2) the twisted tori are precisely the three-dimensional ADE orbifolds TΓ = Γ\S3 of the three-sphere for a finite subgroup Γ ⊂ SU(2); for Γ = Zn a cyclic subgroup of order n ≥ 2, this recovers the familiar lens spaces TZn = Zn\S3 =: L(n, 1). Following [32], we come now to a central concept of this paper. Definition 3.5 (topological T-duality). Let TΛG be a twisted torus which admits a non-trivial right action of the abelian Lie group Rn for some n ≥ 1. The crossed product C(TΛG ) ort Rn is a C∗-algebraic T-dual of the twisted torus. If the spectrum of the crossed product algebra C(TΛG ) ort Rn is a Hausdorff topological space X (for instance if it is Morita equivalent to a commutative C∗-algebra C(X)), then we say that X is T-dual to the twisted torus TΛG and call X a ‘classical T-dual’; otherwise we say that the T-dual of TΛG is a noncommutative space. For Definition 3.5 to be a ‘good’ notion of T-duality, we should first explain (a) in what precise sense TΛG and C(TΛG ) ort Rn are ‘equivalent’, and (b) how T-duality applied twice returns the original twisted torus. The answers to both of these points turns out to be provided by working in a suitable category tailored to our treatment of topological T-duality. 3.2 T-duality in the category KK The terminology ‘topological T-duality’ refers to a coarse equivalence at the level of topology; for C∗-algebras the topology is measured by K-theory. A more powerful refinement is provided by Kasparov’s bivariant K-theory which constructs groups KK(A,B) for any pair of separable C∗-algebras A and B; when A = C, the group KK(C,B) ' K(B) is the K-theory group of B. The cycles in Kasparov’s groups KK(A,B), called Kasparaov A–B bimodules, are triples (H, φ, T ) where H is a right Hilbert B-module, φ is a ∗-representation of A on H, and T ∈ EndB(H) is a B- linear operator onH, subject to certain compactness conditions; we do not provide further details of the definition here and instead refer to [6] for a concise review of KK-theory in the context that we shall use it in this paper. Kasparov bimodules may be thought of as generalizations of morphisms between C∗-algebras, in the sense that any algebra homomorphism φ : A → B determines a class [φ] ∈ KK(A,B), represented by the A–B-bimodule (B, φ, 0). A key feature of Kasparov’s KK-theory is the composition product ⊗B : KK(A,B)×KK(B, C) −→ KK(A, C), which is bilinear and associative. This product is compatible with the composition of morphisms φ : A → B and ψ : B → C of C∗-algebras: [φ] ⊗B [ψ] = [ψ ◦ φ]. It also makes KK(A,A) into a ring with unit 1A = [idA]. We say that an element α ∈ KK(A,B) is invertible if there exists an element β ∈ KK(B,A) such that α⊗B β = 1A and β ⊗A α = 1B. An important special instance of Kasparov bimodules comes from Morita equivalence: Any Morita equivalence A–B bimodule M is also a Kasparov bimodule (M, φ, 0), with φ : A → End(M) the left action of A, which defines an invertible class [M] ∈ KK(A,B) with inverse [M ] ∈ KK(B,A) given by the conjugate B–A bimodule M. Generally, if there exists an invertible element α ∈ KK(A,B), then the algebras A and B are said to be KK-equivalent, and 18 P. Aschieri and R.J. Szabo we write A ∼KK B. Thus Morita equivalence implies KK-equivalence, but the converse is not generally true. KK-equivalent algebras have isomorphic K-theory groups, but not necessarily homeomorphic spectra. This refinement naturally suggests an approach to T-duality where the category of separable C∗-algebras with ∗-homomorphisms is replaced with an additive category KK , whose objects are again separable C∗-algebras but whose morphisms between any two objectsA and B are given by the classes in KK(A,B) (see, e.g., [7]). The composition product defines the composition law, and isomorphic algebras in this category are precisely the KK-equivalent algebras; in particular, Morita equivalent algebras are isomorphic as objects in KK . Our formulation and computations of topological T-duality will always take place in this category, and in this setting we can easily provide answers to points (a) and (b) below Definition 3.5 through Theorem 3.6. If TΛG is a twisted torus with a non-trivial right action of Rn, then there are isomorphisms in the category KK given by the equivalences (a) C(TΛG ) ∼KK C(TΛG ) ort Rn (up to a shift of degree n mod 2), and (b) ( C(TΛG ) ort Rn ) or̂t R n ∼M C(TΛG ). Proof. The KK-equivalence (a) follows from the Connes–Thom isomorphism, formulated in the language of KK-theory [17]. The Morita equivalence (b) follows from Takai duality (Theo- rem 2.9). � Another virtue of the categorical setting is that it enables a general algebraic reformulation of the correspondence space construction, which for topological spaces ‘geometrizes’ the action of topological T-duality. In [7, Proposition 5.3] it is proven that, if A and B are separable C∗- algebras, then any class in KK(A,B) can be represented by a ‘noncommutative correspondence’. For this, we first recall, following [6, 7], that KK-theory provides a definition of Gysin or “wrong way” homomorphisms on K-theory for C∗-algebras. If φ : A → B is a morphism of separable C∗- algebras, a K-orientation is a functorial assignment of a corresponding element φ! ∈ KK(B,A). If a K-orientation exists, we say that φ is K-oriented and call φ! the associated Gysin element. The Gysin homomorphism on K-theory is now defined by φ! := (−)⊗B φ! : K(B) → K(A). We then slightly adapt the definition from [7] to the present context of Theorem 3.6. Definition 3.7 (noncommutative correspondences). Let TΛG be a twisted torus which admits a non-trivial right action of Rn, and let C C(TΛG ) [φ] ;; C(TΛG ) ort Rn [ψ] ff (3.1) be a diagram in KK whose arrows are induced by homomorphisms φ : C(TΛG ) → C and ψ : C(TΛG ) ort Rn → C of separable C∗-algebras. Assume that ψ is K-oriented, and let ψ! ∈ KK ( C, C(TΛG )ortRn ) be its corresponding Gysin element. The separable C∗-algebra C is a non- commutative correspondence if the associated element [φ]⊗C ψ! ∈ KK ( C(TΛG ), C(TΛG ) ort Rn ) is a KK-equivalence between the twisted torus and its C∗-algebraic T-dual. Analogously to [4], we obtain a noncommutative correspondence by restricting the Rn-action to the lattice Zn ⊂ Rn. Topological T-Duality for Twisted Tori 19 Proposition 3.8. The crossed product C = C(TΛG ) ort|Zn Zn is a noncommutative correspondence in the sense of Definition 3.7. Proof. We need to construct a diagram (3.1) in KK for the crossed product. For this, note that for any dynamical system of the form (A,Λ, α) where Λ is a discrete group, there is a natural injection j of the algebra A into the crossed product Aoα Λ: given a ∈ A, define the sequence j(a) ∈ Cc(Λ,A) by j(a)γ = aδγ,e for γ ∈ Λ. It is easy to check, using the explicit formula for the convolution product, that the map a 7→ j(a) is an algebra monomorphism: j(a) ? j(b) = j(a b) for a, b ∈ A. In particular, there is a C∗-algebra injection j : C(TΛG ) −→ C(TΛG ) ort|Zn Zn. (3.2) Next we apply [20, Corollary 2.8] with Rn acting on Tn = Rn/Zn by (right) translation and the diagonal action of Rn on Tn × TΛG to obtain an isomorphism C ( Tn × TΛG ) ort Rn ' ( C(TΛG ) ort|Zn Zn ) ⊗K ( L2 ( Tn )) . The projection Tn×TΛG → TΛG induces an injection Cc(Rn×TΛG ) ↪→ Cc(Rn×Tn×TΛG ) which preserves the convolution product, and we obtain a C∗-algebra monomorphism ψ′ : C(TΛG ) ort Rn −→ ( C(TΛG ) ort|Zn Zn ) ⊗K ( L2(Tn) ) , which is easily checked to be K-oriented since it is induced by a projection. This gives algebra morphisms φ′ := ι ◦ j : C(TΛG )→ C ⊗K and ψ′ : C(TΛG ) ort Rn → C ⊗K, where C = C(TΛG ) ort|Zn Zn, K denotes the C∗-algebra of compact operators on a separable Hilbert space, and ι : C → C⊗K is the usual stabilization map. Taking the composition products of [φ′] and [ψ′] with the Morita equivalence C⊗K∼M C then yields the required maps in (3.1). � When the spectrum of the C∗-algebraic T-dual is a Hausdorff space, we identify the corre- spondence space with the spectrum of C = C(TΛG ) ort|Zn Zn; otherwise C is a noncommutative space. 3.3 Computational tools Let us now explain how to compute these C∗-algebraic T-duals in some special instances that will appear throughout the remainder of this paper. For certain actions of R, we may compute the C∗-algebraic T-dual via Proposition 3.9. Let TΛG be a twisted torus equipped with an action of R for which every point has isotropy subgroup Z. Let T = TΛG /R, and denote the corresponding principal circle bundle by p : TΛG → T . Then the C∗-algebraic T-dual C(TΛG ) ort R ' CT(T × T, δ) is a continuous- trace algebra with spectrum T × T and Dixmier–Douady class δ = ζ ^ c1(p) ∈ H3(T × T,Z), where c1(p) ∈ H2(T,Z) is the Chern class of the circle bundle and ζ is the standard generator of H1(T,Z) ' Z. Proof. This is just a straightforward adaptation of the statement of [39, Proposition 4.5]. � In these instances, the T-dual of TΛG is the Hausdorff space X = T× (TΛG /R) with a three- form ‘H-flux’ whose cohomology class is represented by [H] = ζ ^ c1(p). 20 P. Aschieri and R.J. Szabo More generally, suppose that the Rn-action on TΛG is induced by a free and proper right action of Rn on the covering group G which commutes with the left action of the lattice ΛG on G. We can then apply Green’s theorem (Theorem 2.13) to get the Morita equivalence C(TΛG ) ort Rn ∼M C0 ( G/Rn ) olt ΛG. (3.3) In this special case, we obtain an easy proof of Proposition 3.8: The inclusion Zn ↪→ Rn of groups induces a monomorphism C0 ( G/Rn ) olt ΛG −→ C0 ( G/Zn ) olt ΛG ∼M C(TΛG ) ort|Zn Zn, where in the last step we replaced Rn by its subgroup Zn in (3.3). This gives monomor- phisms (3.2) and C(TΛG ) ort Rn → C(TΛG ) ort|Zn Zn in the category KK . 3.4 Topological T-duality for the torus Let us now describe how our considerations reproduce the standard T-duality for tori. The simplest example of the T-duality scheme (3.3) is the case where G = R, ΛG = Z ⊂ R, and G/R = {0} with the obviously trivial ΛG-action. Then TZ = Z\R = T is a circle, and (3.3) with n = 1 reads C(T) ort R ∼M Co Z = C∗(Z) ' C ( T̃ ) , where in the last passage we used the Fourier transform isomorphism F : C∗(Z) → C ( T̃ ) ; ex- plicitly, if a = {an}n∈Z ∈ C∗(Z), then F(a)(χ) = ∑ n∈Z an e 2πinχ so that F(a) is a function on the dual circle T̃ = R∗/Z∗. The generalization to T-duality along a single direction i of a d-dimensional torus is straight- forward. Let Λ ' Zd be the lattice in Rd given by Λ = {∑d i=1 ai~ei | a1, . . . , ad ∈ Z } , where ~e1, . . . , ~ed is the standard basis of Rd; this is the direct sum Λ = ⊕d i=1 Z~ei. Let Ri be the subgroup of Rd linearly spanned by ~ei and let Zi ⊂ Ri be the corresponding lattice; we write Ti = Ri/Zi and decompose the d-torus Td = Rd/Λ as Td = Td−1 ı̂ × Ti, where Td−1 ı̂ is the (d−1)-dimensional torus defined by omitting the i-th factors of Rd and Λ. Then Ri acts trivially on Td−1 ı̂ and we have C ( Td ) ort Ri = C ( Td−1 ı̂ × Ti ) ort Ri ' ( C ( Td−1 ı̂ ) ⊗ C(Ti) ) oid⊗rt Ri ' C ( Td−1 ı̂ ) ⊗ ( C(Ti) ort Ri ) ∼M C ( Td−1 ı̂ ) ⊗ C ( T̃i ) ' C ( Td−1 ı̂ × T̃i ) = C ( Tdı̃ ) , (3.4) where Tdı̃ := Td−1 ı̂ × T̃i. This is the expected action of the i-th factorized T-duality, and in this way we have thus reproduced the standard rules for T-duality of tori. In fact, in this case we can use Proposition 3.9 to strengthen the statement of topological T-duality: Every point of Td has isotropy group Z under the action of Ri, and the corresponding circle bundle p : Td → Td−1 ı̂ is trivial, so the C∗-algebraic T-dual of Td is a continuous-trace algebra with spectrum Tdı̃ and trivial Dixmier–Douady class. Hence the Morita equivalence in (3.4) can be replaced by a stable isomorphism. By iterating these T-duality transformations one can perform T-dualities along multiple directions of a d-dimensional torus. In particular, iterating the procedure d times and using Theorem 2.6 we end up with the full T-duality C ( Td ) ort Rd ∼M C ( T̃d ) , (3.5) Topological T-Duality for Twisted Tori 21 where T̃d = ( Rd )∗ /Λ∗ is the dual torus with Λ∗ the dual lattice in the dual vector space ( Rd )∗ . This can also be obtained directly by setting G = Rd, ΛG = Λ ⊂ Rd and G/Rd = {0} in (3.3) with n = d, and by using the Fourier transforms in all directions ~e1, . . . , ~ed. Finally, let us consider the correspondence space construction. For the i-th factorized T- duality, this is obtained by restricting the action of Ri to the lattice Zi ⊂ Ri. The action of the group Zi on the algebra of functions C ( Td ) is trivial and so we get isomorphisms C ( Td ) ort Zi ' C ( Td ) ⊗ C∗ ( Zi ) ' C ( Td ) ⊗ C ( T̃i ) ' C ( Td × T̃i ) . This results in the noncommutative correspondence induced by the diagram C ( Td × T̃i ) C ( Td ) pr∗ 99 C ( Tdı̃ ) π∗i ee C ( Td−1 ı̂ ) j 99 j ee where pr : Td × T̃i → Td is the projection to the first factor and πi : Td × T̃i → Tdı̃ omits the i-th factor of Td. The algebra inclusions j of C ( Td−1 ı̂ ) are induced by the trivial circle bundle projections Td → Td−1 ı̂ and Tdı̃ → Td−1 ı̂ . By either iterating this construction using Theorem 2.6 or by direct calculation, the corre- spondence space for a full T-duality is obtained by restricting the action of Rd from (3.5) to the lattice Λ ⊂ Rd, and we analogously find C ( Td ) ort Λ ' C ( Td × T̃d ) . Thus the crossed product with the lattice of periods Λ defining the d-torus Td recovers the doubled torus Td × T̃d which is the correspondence space for the smooth Fourier–Mukai trans- form, wherein a full or factorized T-duality has a geometric interpretation as an element of its automorphism group GL(2d,Z). 3.5 Topological T-duality for orbifolds Let G be a compact connected semisimple Lie group of rank r, and let Γ ⊂ G be a finite subgroup. The maximal torus T = U(1)r ' Rr/Zr of G carries a natural action of Ri by translation along the i-th direction for i = 1, . . . , r, and we can apply a fibrewise T-duality to the principal torus bundle G→ G/T. Under this R-action every point of G has isotropy subgroup Z, and the action descends to the smooth orbifold TΓ = Γ\G. Then the quotient map pi : TΓ → TΓ/Ri is the circle fibration TΓ → TΓ/U(1)i, where T = U(1)r−1 ı̂ × U(1)i, and by Proposition 3.9 the C∗-algebraic T-dual C(Γ\G) ort Ri ' CT ( Γ\G/U(1)i × T̃i, δi ) (3.6) of the orbifold TΓ is a continuous-trace algebra with spectrum TΓ/U(1)i × T̃i and Dixmier– Douady class δi = c1(pi) ^ ζi. In the rank one case, this T-duality is well-known (see, e.g., [3]): Then G = SU(2) which we regard as the three-sphere S3, and for Γ = Zn ⊂ T = U(1) the twisted torus is the lens space L(n, 1). The quotient map p : L(n, 1) → S2 is a circle bundle whose Chern class c1(p) is equal to n times the standard generator of H2 ( S2,Z ) ' Z, and applying (3.6) we find that the 22 P. Aschieri and R.J. Szabo C∗-algebraic T-dual of L(n, 1) is a continuous-trace algebra whose spectrum is the trivial circle bundle L(0, 1) = S2×T̃ and whose Dixmier–Douady class δ is n times the standard generator of H3 ( S2×T̃,Z ) ' Z. Generally, the correspondence space construction is obtained by noting that, since the isotro- py subgroup for any point of the Ri-action is Zi ⊂ Ri, the group Zi acts trivially on the algebra C(TΓ) and there are isomorphisms C(TΓ) ort Zi ' C(TΓ)⊗ C∗(Zi) ' C(TΓ)⊗ C ( T̃i ) ' C ( TΓ × T̃i ) . Let pr : TΓ × T̃i → TΓ be the projection to the first factor. Since H2(G,Z) = 0, Künneth’s theorem implies( pi × idT̃i )∗( c1(pi) ^ ζi ) = 0 ∈ H3 ( TΓ × T̃i,Z ) , and hence the algebra CT ( TΓ × T̃i, (pi × idT̃i) ∗δi ) is isomorphic to C ( TΓ × T̃i ) ⊗K. Then there is the noncommutative correspondence C ( TΓ × T̃i ) C(TΓ) [pr∗] 88 CT ( TΓ/U(1)i × T̃i, δi ) [(pi×idT̃i )∗] ii C ( TΓ/U(1)i ) [j] 55 [p∗] ff as a diagram in the category KK . 4 Topological T-duality for almost abelian solvmanifolds A large class of twisted tori of interest as string compactifications come in the form of fibrations over tori. These are the solvmanifolds which are based on solvable groups G and generalize the nilmanifolds discussed in Example 3.3. The fibrations underlying these twisted tori are called Mostow bundles [35], and we are particularly interested in the cases where the Mostow bundle is a torus bundle. A good source for the material used in this section is [2] (see also [10, 50]). 4.1 Mostow bundles Let G be a connected and simply-connected solvable Lie group. Recall that its nilradical N is the maximal connected nilpotent normal subgroup. It has dimension dimN ≥ 1 2 dimG. We first consider the case dimN = dimG. Then N = G and the group G is nilpotent. In this case, under the conditions discussed in Example 3.3, there exists a lattice ΛG ⊂ G and the twisted torus TΛG is a nilmanifold. If G is abelian then TΛG is a torus. If G is non-abelian then there is a group extension 1 −→ [G,G] −→ G π−−→ Gab −→ 1 of its commutator subgroup [G,G], and both ΛG ∩ [G,G] and π(ΛG) are lattices in the nilpotent Lie group [G,G] and the abelianization Gab := [G,G]\G of G, respectively [11]. This exhibits the twisted torus TΛG = ΛG\G as a fibration over the torus π(ΛG)\Gab with nilmanifold fibres,( ΛG ∩ [G,G] ) \[G,G] −→ TΛG −→ π(ΛG)\Gab. If [G,G] is an abelian Lie group then the twisted torus is a torus bundle over a torus. Topological T-Duality for Twisted Tori 23 Suppose now that the group G is not nilpotent. Then N\G is a non-trivial abelian Lie group. If G admits a lattice ΛG, then ΛN := ΛG∩N is a lattice in N and ΛGN = NΛG is a closed subgroup of G, so ΛGN\G is a torus. The twisted torus TΛG = ΛG\G is then a fibration over this torus with fibre the nilmanifold ΛN\N = ΛG\ΛGN. This bundle is called the Mostow bundle [35]. We summarise these statements as Theorem 4.1 (Mostow bundles). Let ΛG be a lattice in a connected and simply-connected solvable Lie group G and TΛG = ΛG\G the associated solvmanifold. Let N be the nilradical of G. Then ΛGN is a closed subgroup of G, ΛN := ΛG ∩ N is a lattice in N, and ΛGN\G is a torus. It follows that the twisted torus TΛG is a fibration over this torus with nilmanifold fibre: ΛN\N = ΛG\ΛGN −→ TΛG −→ ΛGN\G. Remark 4.2. The structure group of the Mostow bundle is ΛG0\ΛGN, where ΛG0 is the largest subgroup of ΛG which is normal in ΛGN (cf. [2]). In particular, if ΛG = ΛG0 then the Mostow bundle is a principal ΛG\ΛGN-bundle. In this case there is a well-defined left ΛGN-action on TΛG = ΛG\G and each point has isotropy subgroup ΛG, so that the induced ΛG\ΛGN-action is principal. If the solvable Lie group G admits an abelian normal subgroup V, then ΛGV = VΛG is a subgroup of G; if ΛG is normal in ΛGV, then the Mostow bundle construction can be refined via an intermediate step involving a principal torus bundle over a second solvmanifold. Adapting [2, Theorem 3.6] we have Proposition 4.3. Let G be a connected and simply-connected solvable Lie group and ΛG a lattice in G. Let V be a closed normal abelian Lie subgroup of G such that ΛG is normal in VΛG. If VZ := ΛG ∩V is a lattice in V, then ΛG\ΛGV is a torus and the solvmanifold TΛG = ΛG\G is the total space of the principal torus bundle ΛG\ΛGV −→ TΛG −→ ΛGV \ G, with base the solvmanifold TΛ GV = ΛGV\GV := ( VZ\ΛG )∖( V\G ) = ΛGV \ G. There is moreover a double fibration Tn // TΛG �� ΛNV\NV // TΛ GV �� Tm (4.1) where n = dimV, m = dim ( NV\GV ) , NV is the nilradical of GV and ΛNV = NV∩ΛGV the associated lattice. Proof. Let p: G → V\G be the canonical projection. Since V is normal in G, and ΛG and ΛG ∩ V are lattices in G and V, respectively, by [11, Lemma 5.1.4(a)] it follows that p(ΛG) is a lattice in V\G. Hence p−1(p(ΛG)) = VΛG = ΛGV is closed in G, and π : G → ΛGV\G is a bundle. By [49, Section 7.4] (adapted to the smooth case), since ΛG is a closed normal subgroup of ΛGV, it follows that ΛG\G → ΛGV\G is a principal ΛG\ΛGV-bundle (or in other words, Remark 4.2 holds as well under the present hypotheses). The fiber is a torus because V is abelian, ΛG\ΛGV = VZ\V = Tn, with n = dimV. Moreover, because V is normal in G, there is a canonical action of the group VZ\ΛG on the connected and simply-connected solvable Lie group V\G (given by ( VZλ ) (Vg) = V(λg) for λ ∈ ΛG and g ∈ G), so that ΛGV\GV := ( VZ\ΛG )∖( V\G ) is a solvmanifold. It is then easily proven that ( VZ\ΛG )∖( V\G ) = ΛGV\G. The double fibration (4.1) follows immediately from Theorem 4.1 applied to TΛ GV = ΛGV\GV. � 24 P. Aschieri and R.J. Szabo Remark 4.4. An equivalent form for the double fibration (4.1) is given by Tn // TΛG �� ΛNV\NV // TΛG /V �� Tm which is obtained by observing that ΛGV\G = ΛG\G / V since V is normal in G. Notice also that G/V = V\G and ΛG/V Z = VZ\ΛG (because VZ = ΛG ∩ V is normal in ΛG), and moreover ΛGV\GV = ( VZ\ΛG )∖ (V\G) = ( ΛG/V Z)∖(G/V). 4.2 Almost abelian solvmanifolds In contrast to the case of nilpotent groups, there is no simple criterion for the existence of a lattice in a general connected and simply-connected solvable Lie group, as is required to define a corresponding twisted torus. To formulate such a criterion, we specialise to almost abelian solvable groups: these are the solvable Lie groups G of dimension d whose nilradical N has codimension one and is abelian: N ' Rd−1. Then G has the structure of a semi-direct product G = Noϕ R for a continuous one-parameter left group action ϕ : R→ Aut(N); concretely, ϕ is given by the adjoint action of the one-dimensional subgroup H = R on N in the group G (cf. Section 2.2). This exhibits G as a nontrivial group extension 1 −→ N −→ G −→ R −→ 1. We can regard the one-parameter group action ϕ as a matrix ϕx ∈ GL(d−1,R) for each x ∈ R acting on the vector space N, which we identify with Rd−1 via a choice of basis ~e1, . . . , ~ed−1. Since ϕ0 = 1d−1 and ϕx is always non-singular, it follows from the continuity of ϕ and the determinant that detϕx > 0 for all x ∈ R. Then G admits a lattice ΛG if and only if there exists x0 ∈ R× such that ϕx0 is conjugate to an integer matrix M ∈ SL(d− 1,Z): Σ−1ϕx0Σ = M (4.2) for some Σ ∈ GL(d−1,R). In this case the twisted torus TΛG = ΛG\G is called an almost abelian solvmanifold . The condition (4.2) strongly restricts the homomorphisms ϕ : R → Aut(N); in particular, it requires that the characteristic polynomial of ϕx0 has integer coefficients. In this case the lattice (in the standard basis ~e1, . . . , ~ed−1) is given by ΛG = Σ · Zd−1 oϕ|x0 Z x0Z, which correspondingly sits as a nontrivial group extension 1 −→ Σ · Zd−1 −→ ΛG −→ x0Z −→ 1. Then ΛN \ N ' Td−1, and the corresponding Mostow bundle realises the twisted torus TΛG as a torus bundle over a circle ΛGN\G ' T, whose monodromy is specified by the matrix M in the Topological T-Duality for Twisted Tori 25 mapping class group SL(d − 1,Z) of orientation-preserving automorphisms up to homotopy of the torus fibres Td−1. For an almost abelian solvmanifold we can make the twisted torus construction more concrete by choosing the global coordinates (z, x) ∈ Rd−1 × R on the group manifold (associated with the basis ~e1, . . . , ~ed−1). The group multiplication of the semi-direct product G = Rd−1 oϕ R is (z, x)(z′, x′) = (z + ϕx · z′, x+ x′), (4.3) where we used ϕxϕx′ = ϕx+x′ , and the inverse of a group element is (z, x)−1 = (−ϕ−x · z,−x), where we used ϕ−1 x = ϕ−x. The twisted torus is defined as the quotient TΛG = ΛG\G which is generated by the equivalence relation (z, x) ∼ (Σ · γ, x0 α) (z, x) for all (γ, α) ∈ Zd−1 × Z. The global structure of TΛG is generated by the simultaneous local coordinate identifications under the action of the elements (Σ · γ, x0α) of ΛG given by (z, x) 7−→ (z + Σ · γ, x), (z, x) 7−→ (ΣMα Σ−1 · z, x+ x0 α). (4.4) These identifications explicitly exhibit the twisted torus as a torus bundle over a circle, with local fiber coordinates z ∈ Td−1 and base coordinate x ∈ T, whose monodromy is specified by the matrix M ∈ SL(d− 1,Z), and whose periods are given respectively by Σ ∈ GL(d− 1,R) and x0 ∈ R×. 4.3 C∗-algebra bundles Mostow bundles and their C∗-algebraic T-duals can be grouped together under the general heading of ‘C∗-algebra bundles’, which encompasses the notions of C0(X)-algebras and C∗- bundles, as we now explain; see Section 8.1 and Appendix C of [52] for further details. Let X be a locally compact Hausdorff space. There are two equivalent notions for the C∗-algebra analogue of a fibre bundle over X. A C0(X)-algebra is a C∗-algebra A equipped with a nondegenerate injection ι of C0(X) into the centre of its multiplier algebra, called the structure map. For f ∈ C0(X) and a ∈ A, we abbreviate ι(f)a by f · a. This endows A with a C0(X)-bimodule structure. Given a family B = (Bx)x∈X of C∗-algebras, a section of B is a map s : X → B such that s(x) ∈ Bx for all x ∈ X; we denote the space of sections of B which vanish at infinity by Γ0(B). The family B is then called a C∗-bundle over X with fibres Bx if the following conditions are satisfied: � Γ0(B) is a C∗-algebra under pointwise operations and the supremum norm; � Bx = {s(x) | s ∈ Γ0(B)} for each x ∈ X; � Γ0(B) is closed under multiplication by C0(X); and � For each s ∈ Γ0(B), the function x 7→ ‖s(x)‖ is upper semi-continuous, i.e., the set {x ∈ X | ‖s(x)‖ < ε} is open in X for all ε > 0. In this paper we will only be concerned with C∗-bundles that have non-zero fibres. If B is a C∗-bundle over X, then its section algebra Γ0(B) is a C0(X)-algebra: its structure map ι from C0(X) is defined by ι(f)s = fs. Conversely, if A is a C0(X)-algebra, then the fibre Ax of A over x ∈ X is Ax := A/Ix, where Ix = {f · a | f ∈ C0(X), f(x) = 0, a ∈ A} is 26 P. Aschieri and R.J. Szabo identified as the ideal in A of sections vanishing at x. If a ∈ A, we write a(x) = a + Ix for its image in Ax. The function x 7→ ‖a(x)‖ is upper semi-continuous and vanishes at infinity with ‖a‖ = sup x∈X ∥∥a(x) ∥∥ for all a ∈ A. The elements a ∈ A can in this way be viewed as sections of a C∗-bundle (Ax)x∈X . We will sometimes use the notation ∐ x∈X Ax for A when we wish to emphasise its structure as a C∗-bundle over X with fibre C∗-algebras Ax. These definitions do not require local triviality of the bundle nor the fibres of the bundle to be isomorphic to one another. C∗-algebra bundles over X are objects of a category whose morphisms are fibrewise ∗-homomorphisms, i.e., C0(X)- linear morphisms ψ : A → B: ψ(f · a) = f · ψ(a) for all f ∈ C0(X) and a ∈ A; then ψ induces ∗-homomorphisms ψx : Ax → Bx such that ψx ( a(x) ) = ψ(a)(x) for all a ∈ A. Example 4.5 (trivial C∗-algebra bundles). If D is any C∗-algebra, then A = C0(X,D) ' C0(X)⊗D is naturally a C0(X)-algebra with structure map( ι(f)a ) (x) := f(x) a(x) for f ∈ C0(X), a ∈ A and x ∈ X. In this case each fibre Ax is canonically identified with D and elements of A are obviously identified with sections. Example 4.6 (continuous maps). Let X and Y be locally compact spaces and σ : Y → X a continuous surjective map. Then C0(Y ) is a C0(X)-algebra with structure map ι(f)g := (f ◦ σ) g, for f ∈ C0(X) and g ∈ C0(Y ), and fibers C0(Y )x ' C0 ( σ−1(x) ) . In this paper we are particularly interested in crossed products of C∗-algebra bundles. Let A be a C0(X)-algebra, and denote by AutX(A) the group of fibrewise automorphisms of A. A fibre- wise action of a locally compact group G on A is then a group homomorphism α : G→ AutX(A). This implies that α induces an action αx on each fiber Ax for x ∈ X, and in this case we say that the dynamical system (A,G, α) is C0(X)-linear. Theorem 4.7. Let X be a locally compact Hausdorff space, and let (A,G, α) be a C0(X)-linear C∗-dynamical system. Then the crossed product Aoα G is again a C0(X)-algebra with fibres (Aoα G)x ' Ax oαx G, where αxγ ( a(x) ) = αγ(a)(x) for each x ∈ X, γ ∈ G and a ∈ A. Proof. The structure map of A oα G is given by precomposing the structure map of A with the natural injection of the center of the multiplier algebra of A into the center of the multiplier algebra of Aoα G; it satisfies (f · f)(γ) = f · ( f(γ) ) for all f ∈ C0(X), f ∈ Cc(G,A) and γ ∈ G. See [52, Theorem 8.4] for further details. � Example 4.8 (transformation groups). Let (X,G) be a second countable transformation group whose quotient X/G is a Hausdorff space. By Example 4.6, C0(X) is a C0(X/G)-algebra whose fiber over G · x is isomorphic to C0(G/Gx), where Gx = {γ ∈ G | γ · x = x} is the stabilizer subgroup at x ∈ X. Then the crossed product C0(X)oαG is the section algebra of a C∗-algebra bundle over X/G whose fiber over G · x is isomorphic to C∗(Gx) ⊗ K ( L2(G/Gx) ) [51]. In the special case where G = R and Gx = Z for all x ∈ X, this is contained in the statement of Proposition 3.9. Topological T-Duality for Twisted Tori 27 Example 4.9 (principal torus bundles). Let E → X be a principal Tr-bundle. By Exam- ple 4.8, C0(E) is a C0(X)-algebra with fibers C0(E)x ' C(Tr), and by (2.10) there is a stable isomorphism C0(E) ort Tr ' C0(X)⊗K ( L2(Tr) ) . More generally, a noncommutative principal Tr-bundle on X is a C0(X)-linear C∗-dynamical system (A,Tr, α) with an isomorphism Aoα Tr ' C0(X,K) of C∗-algebra bundles over X. For further details and a classification of noncommutative prin- cipal torus bundles, see [15, 21]. Example 4.10 (noncommutative correspondences). The noncommutative correspondence C = C(TΛG ) ort Zn from Proposition 3.8 is a noncommutative principal Tn-bundle on X = TΛG in the sense of Example 4.9: The C∗-algebra C is naturally equipped with the dual action of Tn = Ẑn, and the Takai duality theorem implies that there is an isomorphism C or̂t T n ' C(TΛG )⊗K ( `2(Zn) ) of C∗-algebra bundles over the twisted torus TΛG . Remark 4.11. There is a natural notion of Morita equivalence of C∗-algebra bundles over X, similar to the notion of equivariant Morita equivalence from Theorem 2.12, which uses the C0(X)-bimodule structures: a C0(X)-linear Morita equivalence between two C0(X)-algebras is a Morita equivalence which is compatible with the C∗-bundle structures over X. More gener- ally, there is a category RKKX of C∗-algebra bundles over X whose morphisms are elements of Kasparov’s groups RKK(X;A,B), see, e.g., [15]: the cycles are the usual cycles (H, φ, T ) for Kasparov’s bivariant K-theory KK(A,B) (cf. Section 3.2) with the additional requirement that φ : A → EndB(H) is C0(X)-linear. There is an obvious faithful functor RKKX → KK which forgets the C0(X)-algebra structures. Isomorphic C∗-bundles in the category RKKX are precisely the RKK-equivalent C∗-bundles. If A and B are isomorphic in RKKX , i.e., there exists an invertible class α ∈ RKK(X;A,B), then they are also isomorphic as C∗-algebras in the category KK . 4.4 Rn-actions on Mostow bundles We can now apply the results of Section 3 to the class of twisted tori given in Section 4.2. The Mostow fibration of any almost abelian solvmanifold identifies TΛG as a torus bundle over a circle, hence the algebra of functions C(TΛG ) is a C(T)-algebra. In other words, C(TΛG ) is an object of the category RKKT, and we are interested in the T-duality isomorphisms of C(TΛG ) in this category. In particular, given a fibrewise right action of the abelian Lie group Rn on TΛG , it follows from Theorem 4.7 that the C∗-algebraic T-dual C(TΛG )ortRn is also a C(T)-algebra, and by [15, Theorem 3.5] the C∗-bundles C(TΛG ) and C(TΛG ) ort Rn are isomorphic as C∗-algebras in the category RKKT. In order to have sensible definitions of T-duality, we need to identify the homologically non- trivial one-cycles of the twisted torus TΛG , which are determined in [2, Proposition 4.7]. Write M = (mij) for the integer matrix elements mij ∈ Z of the monodromy matrix. Since G = Rd−1 oϕ R is simply-connected, the fundamental group of the twisted torus is π1(TΛG ) ' ΛG whose abeliani- sation ΛG/[ΛG,ΛG] gives the first homology group via the presentation H1(TΛG ,Z) = Z⊕ 〈 ê1, . . . , êd−1 ∣∣∣ d−1∑ j=1 mjiêj = êi for i = 1, . . . , d− 1 〉 , (4.5) 28 P. Aschieri and R.J. Szabo where the first factor of Z corresponds to the base circle of the torus fibration and the generators ê1, . . . , êd−1 correspond to the torus fibres; they are given by êi := d−1∑ k=1 Σki~ek, (4.6) where ~e1, . . . , ~ed−1 is the standard basis of Zd−1 giving the group law (4.3). We can then apply the structure theorem for finitely-generated Z-modules by appealing to some classical matrix algebra. From the presentation (4.5), H1(TΛG ,Z) = Z⊕coker(ϕx0− idZd−1) where (ϕx0−idZd−1) : Zd−1 → Zd−1 in the basis ê1, . . . , êd−1 is given by the integer relation matrix A := M− 1d−1 . (4.7) Let r be the rank of A. This matrix can be brought into its Smith normal form D by finding invertible integer matrices L, R ∈ GL(d− 1,Z) such that D = LAR is diagonal with entries mi ∈ Z for i = 1, . . . , d− 1. The integers mi are the elementary divisors of A. They have the properties that mi divides mi+1, for 0 < i < d− 1, and in particular mi = 0 for i > r; they can be computed explicitly (up to sign) as mi = di(A) di−1(A) , where the i-th determinant divisor di(A) is the greatest common divisor of all i×i minors of the relation matrix A, with d0(A) := 1. The matrices L, R ∈ GL(d − 1,Z) are found by reducing the matrix A to its Smith normal form D through a sequence of elementary row and column operations over Z, see, e.g., [23]. Given the Smith normal form, we set ẽi := d−1∑ j=1 ( L−1 ) ji êj (4.8) and observe that the image of A, which is generated over Z by the vectors d−1∑ j=1 Ajiêj = d−1∑ k,l=1 ( R−1 ) ki Dlkẽk, is equivalently generated by the vectors mkẽk with k = 1, . . . , r. Hence coker(ϕx0 − idZd−1) = 〈 ê1, . . . , êd−1 〉/〈 d−1∑ j=1 Aj1êj , . . . , d−1∑ j=1 Ajd−1êj 〉 = 〈 ẽ1, . . . , ẽd−1 〉/〈 m1ẽ1, . . . ,mrẽr 〉 and H1(TΛG ,Z) ' Z⊕ Zd−1−r ⊕ r⊕ i=1 Zmi . (4.9) We are exclusively interested in the natural Rn-actions on TΛG which descend from actions of abelian subgroups Rn ⊂ G, acting on G by right multiplication. They can be organised into Topological T-Duality for Twisted Tori 29 three classes associated with the different types of summands in the Z-module presentation of the homology group (4.9), and we only retain those which are fiberwise actions on the Mostow bundle. The first summand Z in (4.9) corresponds to the subgroup Rx = { (0, ξ) ∈ G } acting on G by right multiplication: (z, x)(0, ξ) = (z, x+ ξ) for all (z, x) ∈ G and ξ ∈ R. Clearly this does not descend to a fiberwise action on the twisted torus TΛG , and the crossed product C(TΛG ) ort Rx is no longer a C(T)-algebra. Thus an Rx- action takes us out of the category RKKT, and we will henceforth discard Rn-actions where Rn contains the subgroup Rx. For the remaining types of summands in (4.9), we can give explicit descriptions of the C∗- algebraic T-duals of an almost abelian solvmanifold. We consider both classes in turn. As the only solvmanifolds in one and two dimensions are tori, which are already treated by our analysis from Section 3.4, we assume d ≥ 3 for the remainder of this paper. 4.5 Ry-actions: Circle bundles with H-flux Let us consider the second summand Zd−1−r in (4.9), which corresponds to the lattice ΛG ∩ ker ( ϕx0 − idRd−1 ) in ker ( ϕx0 − idRd−1 ) . In terms of the generators e′i := d−1∑ j=1 Rjiêj of ΛG ∩N, the sublattice ΛG ∩ ker ( ϕx0 − idRd−1 ) is generated by the vectors e′r+1, e ′ r+2, . . . , e ′ d−1 in the kernel of the relation matrix (4.7). We begin with some elementary observations. Firstly, the subgroups ker ( ϕx0 − idRd−1 ) and ΛG commute in G: (−v, 0)(0, x0)(v, 0) = (0, x0) ∈ ΛG for all v ∈ ker ( ϕx0− idRd−1 ) . Secondly, ker ( ϕx0− idRd−1 ) is a closed abelian normal subgroup of G: (z, x)(v, 0)(z, x)−1 = (ϕx(v), 0) ∈ ker ( ϕx0 − idRd−1 ) since ϕx0(ϕx(v)) = ϕx(ϕx0(v)) = ϕx(v). In the following we consider a subgroup V ' Rn ⊂ ker ( ϕx0 − idRd−1 ) which is normal in G such that VZ := V∩ΛG is a lattice in V. This is the case, for example, if V is the span of a subset of the generators e′r+1, e ′ r+2, . . . , e ′ d−1. As an immediate consequence of Lemma 2.3 we then have Proposition 4.12. The quotient group GV := G/V is a (d−n)-dimensional almost abelian solv- able Lie group GV ' Rd−n−1 oϕV R, where ϕV : R → GL(d − n − 1,R) is defined by ϕV x[z] = (1d−1 − prV)ϕx(z) for all x ∈ R and z ∈ Rd−1, with prV the projection of Rd−1 to V and [z] = (1d−1 − prV)(z) ∈ Rd−1/V. We are particularly interested in the induced action of V on the twisted torus TΛG . Proposition 4.13. Let V, as above, be normal in G and let ΛG be normal in ΛG V. Then the quotient map pV : TΛG → TΛG /V is a principal torus bundle of rank n = dimV over an almost abelian solvmanifold TΛ GV of dimension d−n. Its Chern class c1(pV) ∈ H2(TΛ GV ,Z) can be computed by Chern–Weil theory from the curvature of the connection κV ∈ Ω1(TΛG ,V) given by κV = −(prVϕ−x) · dz in the notation of Proposition 4.12. 30 P. Aschieri and R.J. Szabo Proof. The first statement follows from Proposition 4.3. For the Chern class, we note that the left-invariant Maurer–Cartan one-forms on the Lie group G are given by dx and P ∈ Ω1 ( G,Rd−1 ) where P = ϕ−x · dz, and this descends to the twisted torus TΛG . The desired principal Tn-connection on TΛG is then given by κV = −prVP = −(prVϕ−x) · dz and the result follows. � By virtue of the fibration pV : TΛG → TΛ GV , the algebra of functions C(TΛG ) is also a C(TΛ GV )- algebra. We are particularly interested in the case n = 1, whereby we can explicitly apply our framework of topological T-duality. Combining Propositions 4.13 and 3.9, we immediately arrive at Theorem 4.14. Let y0 ∈ ker ( ϕx0 − idRd−1 ) and let Ry0 := R(y0, 0) be the corresponding one-dimensional subgroup. Suppose that Ry0 is normal in G and ΛG is normal in ΛGRy0 (this is the case, for example, if (y0, 0) is in the center of G). Let Gy0 = G/Ry0 be the almost abelian solvable Lie group constructed by Proposition 4.12, and py0 : TΛG → TΛGy0 the principal circle bundle constructed by Proposition 4.13. Then the C∗-algebraic T-dual C(TΛG ) ort Ry0 ' CT ( TΛGy0 × Ty0 , δy0 ) is a continuous-trace algebra with spectrum TΛGy0 × Ty0 and Dixmier–Douady class δy0 = c1(py0) ^ ζy0 , where ζy0 is the standard generator of H1(Ty0 ,Z) ' Z and the Chern–Weil representative of c1(py0) ∈ H2(TΛGy0 ,Z) is the curvature of the connection κy0 ∈ Ω1(TΛG ) on this circle bundle given by κy0 = −(pry0ϕ−x) · dz. Thus in the case of an action of R due to a normal subgroup of G that is in ker ( ϕx0−idRd−1 ) in the setting of Theorem 4.14, which we collectively refer to as Ry-actions, the T-dual of an almost abelian solvmanifold TΛG is the Hausdorff space X = TΛGy0 × Ty0 with a three-form ‘H-flux’ whose cohomology class is represented by [Hy0 ] = c1(py0) ^ ζy0 . The associated correspondence space construction proceeds analogously to Section 3.5, which we can give explicitly as Proposition 4.15. The topological T-duality of Theorem 4.14 is implemented by the noncom- mutative correspondence C ( TΛG × T̃y0 ) C ( TΛG ) [pr∗] 88 CT ( TΛGy0 × Ty0 , δy0 ) [(py0×idT̃y0 )∗]ii C ( TΛGy0 ) [j] 55 [p∗y0 ] gg as a diagram in the category RKKT. Topological T-Duality for Twisted Tori 31 Proof. Since the subgroup Zy0 := Ry0 ∩ ΛG acts trivially on the algebra of functions C(TΛG ), there is an isomorphism C(TΛG ) ort Zy0 ' C ( TΛG × T̃y0 ) , where T̃y0 is the circle dual to Ty0 ' Ry0/Zy0 . Proposition 4.13 shows that the Chern–Weil representative of c1(py0) ∈ H2 ( TΛGy0 ,Z ) pulls back to the exact two-form dκy0 ∈ Ω2(TΛG ) under the bundle projection py0 : TΛG → TΛ y0 G . Hence p∗y0c1(py0) = 0 which implies( py0 × idT̃y0 )∗( c1(py0) ^ ζy0 ) = 0 ∈ H3 ( TΛG × T̃y0 ,Z ) . Thus the algebra CT(TΛ y0 G × Ty0 , δy0) pulls back to an algebra isomorphic to C ( TΛG × T̃y0 ) ⊗K by py0 × idT̃y0 , and the result follows. � 4.6 Rz-actions: noncommutative torus bundles Let us now come to the torsion summands Zmi in the Z-module decomposition (4.9). Pick a non-trivial elementary divisor mi > 1 for some i ∈ {1, . . . , r}. The corresponding homology generator ẽi is constructed as a Z-linear combination (4.8) of the generators (4.6). It defines a fixed element z0 in the image of ϕx0 − idRd−1 and a corresponding one-dimensional subgroup of G given by Rz0 := R(z0, 0). We further assume that Zz0 := Rz0 ∩ ΛG is a lattice in Rz0 . The choice of z0 ∈ Rd−1 is not unique, and any change of basis of Zd−1, represented by a matrix B ∈ GL(d− 1,Z), defines an equally good element B · z0 ∈ Rd−1 as long as B · z0 ∈ im ( ϕx0 − idRd−1 ) . We write prz0 for the linear projection of Rd−1 to Rz0 , and denote by 〈z0, z〉 ∈ R the component of z ∈ Rd−1 in Rz0 , i.e., prz0 · z = 〈z0, z〉 z0. In the basis êi, the lattice of G is given by ΛG = Zd−1 oϕ̂|x0 Z x0Z, where ϕ̂x := Σ−1 ϕx Σ for all x ∈ R. Then the fibres of the underlying Mostow bundle are ‘square’ tori Td−1 with unit periodicities z ∼ z + ~ei for i = 1, . . . , d − 1, where as before ~ei denotes the standard basis of Rd−1. The action of the subgroup Rz0 on elements (z, x) ∈ G by right multiplication is given by (z, x)(ζ z0, 0) = ( z + ζΣ−1ϕx · z0, x ) (4.10) for ζ ∈ R, where Σ−1ϕx · z0 lies in the image of the relation matrix A = M− 1d−1. Our principal tool to compute the C∗-algebraic T-dual for such an action of R in the image of ϕx0 − idRd−1 , which we collectively refer to as Rz-actions, will be Green’s theorem in the form (3.3): C(TΛG ) ort Rz0 ∼M C0(G/Rz0) olt ΛG. (4.11) 32 P. Aschieri and R.J. Szabo By appealing to Theorem 4.7, we may apply (4.11) fibrewise. For fixed x ∈ R, the fibre Gx of the semi-direct product G = Rd−1oϕ̂R is the subgroup Rd−1, and the corresponding fibre of the solvmanifold TΛG over x ∈ R/x0 Z is the torus Td−1 = Rd−1/Zd−1. The Morita equivalence (4.11) is C(T)-linear and the fibres of the corresponding T-dual C(T)-algebra are given by the fibrewise Morita equivalence( C(TΛG ) ort Rz0 ) x ' C ( Rd−1/Zd−1 ) ortx Rz0 ∼M C0 ( Rd−1/Rz0 ) oltx Zd−1. (4.12) The action of the subgroup Zd−1 ⊂ ΛG on the coset space Rd−1/Rz0 is induced by left multipli- cation in the group G. After a basis transformation, we can decompose the discrete group Zd−2 into a direct sum Zd−1 ' Zd−2 v ⊕ Zz0 , where Zd−1 v = (1d−1 − prz0) · Zd−1. Let F∗ := { x ∈ R ∣∣ 〈z0,Σ −1ϕx · z0 〉 = 0 } . Lemma 4.16. Over any x ∈ F∗, the fiber ( C(TΛG ) ort Rz0 ) x is Morita equivalent to the com- mutative C∗-algebra C ( Td−1 ) . Proof. If x ∈ F∗, then w0 := Σ−1 ϕx · z0 only shifts the corresponding component of z in (1d−1−prz0) ·Rd−1 in the Rz-action (4.10). In this case any element (z, x) ∈ G may be factorized as (z, x) = ( prz0 · z + (1d−1 − prz0 − prw0 ) · z , x ) ( 〈w0, z〉 z0 , 0 ) . Hence the coset space Rd−1/Rz0 ' Rd−2 can be parameterized by prz0 ·z+ ( 1d−1−prz0−prw0 ) ·z with z ∈ Rd−1, which we decompose correspondingly into a direct product Rz0 ×Rd−3 (with the second factor absent for d = 3). By (4.4) the discrete group Zd−2 v acts trivially on the line Rz0 and by translations on Rd−3, while Zz acts by translations on Rz0 and trivially on Rd−3. Then the crossed product on the right-hand side of (4.12) may be unravelled to get C0 ( Rd−1/Rz0 ) oltx Zd−1 ' C0 ( Rz0 × Rd−3 ) oltx ( Zd−2 v × Zz0 ) ' [( C0(Rz0)⊗ C0 ( Rd−3 )) oid⊗ltx Zd−2 v ] oβx Zz0 ∼M ( C0(Rz0)⊗ C∗(Z)⊗ C ( Td−3 )) oltx⊗id⊗id Zz0 ' ( C0(Rz0) oltx Zz0 ) ⊗ C∗(Z)⊗ C ( Td−3 ) ∼M C(T)⊗ C∗(Z)⊗ C ( Td−3 ) ' C ( Td−1 ) . In the second line we applied Theorem 2.6. In the third line we used Example 2.14 together with the fact that the homomorphism σZz0 from (2.5) is trivial since the groups Zz0 and Zd−2 v are discrete, and so the action of Zz0 on the crossed product is induced by left multiplication on Rz0 , the trivial action on Rd−3, and the trivial action on Zd−2 v . In the fifth line we used Example 2.14 again. � The central result of this paper is Theorem 4.17. The C∗-algebraic T-dual of any almost abelian solvmanifold TΛG with respect to an Rz-action is Morita equivalent to a C∗-algebra bundle of noncommutative tori over the circle T: C(TΛG ) ort Rz0 ∼M ∐ x∈R/Z Td−1 ~θz0 (x) , Topological T-Duality for Twisted Tori 33 where the noncommutativity parameters ~θz0(x) ∈ Rd−2 are given by ~θz0(x) =  0 for x ∈ F∗, (1d−1 − prz0)Σ−1ϕx0x · z0〈 z0,Σ−1ϕx0x · z0 〉 for x ∈ R \ F∗. (4.13) Proof. That the fibres over x ∈ F∗ are just ordinary tori Td−1 is established by Lemma 4.16, so we may assume that x ∈ R \F∗. Then a simple calculation shows that any element (z, x) ∈ G can be factorized as (z, x) = (v, x) ( prz0 · z〈 z0,Σ−1ϕx · z0 〉 , 0) , where v := ( 1d−1 − prz0 ) · z − 〈z0, z〉〈 z0,Σ−1ϕx · z0 〉(1d−1 − prz0 ) Σ−1ϕx · z0. (4.14) Thus the coset space Rd−2 x,v := Rd−1/Rz0 may be parameterized by the coordinates v ∈ Rd−2 over any x ∈ R \ F∗, and we explicitly retain the fibre index in the notation for convenience. We now need to unravel the crossed product on the right-hand side of (4.12). From (4.4) and (4.14) it follows that the action of elements (γv, γz0) ∈ Zd−1 ' Zd−2 v ⊕ Zz0 on the coset is given by (γv, 0, 0) · (v, x) = (v + γv, x), (4.15) (0, γz0 , 0) · (v, x) = ( v − γz0〈 z0,Σ−1ϕx · z0 〉(1d−1 − prz0 ) Σ−1ϕx · z0, x ) . (4.16) Applying Theorem 2.6 as in the proof of Lemma 4.16 gives C ( Rd−1/Zd−1 ) ortx Rz0 ∼M ( C0 ( Rd−2 x,v ) oltx Zd−2 v ) oβx Zz0 , where the action of Zd−2 v on the coset Rd−2 x,v is given by (4.15), while the action of Zz0 on the crossed product C0 ( Rd−2 x,v ) oltx Zd−2 v is induced by the action on Rd−2 x,v given in (4.16) and the trivial action on Zd−2 v . Next, Example 2.14 yields C0 ( Rd−2 x,v ) oltx Zd−2 v ∼M C ( Td−2 ) . (4.17) Since the homomorphism σZz0 : Zz0 → R+ is trivial, it is not difficult to see that the Morita equivalence bimoduleM implementing this equivalence is Zz0-equivariant (in the sense of Theo- rem 2.12): the Zz0-action U : Zz0 → Aut(M) is given by Uγv(ξ)(v) = ξ(v+ γv), for all γv ∈ Zz0 , ξ ∈ M and v ∈ Rd−2 x,v . Applying Theorem 2.12 we conclude that the Morita equivalence (4.17) induces the Morita equivalence( C0 ( Rd−2 x,v ) oltx Zd−2 v ) oβx Zz0 ∼M C ( Td−2 ) oltx Zz0 , (4.18) where the action of the group Zz0 on C ( Td−2 ) is the pullback of the action on Td−2 = Rd−2 x,v /Zd−2 v induced from (4.16). After rescaling x by x0 to give it unit period, this shows that the corresponding algebra of functions on the fiber at e 2πix is that of a noncommutative d−1-torus Td−1 ~θz0 (x) with noncom- mutativity parameter ~θz0(x) given by (4.13), see Example 2.2. Thus the C∗-algebraic T-dual C(TΛG ) ort Rz0 is a C∗-algebra bundle of noncommutative tori A~θz0 (x) = Td−1 ~θz0 (x) over the circle T = { e 2πix |x ∈ R/Z } . � 34 P. Aschieri and R.J. Szabo What becomes of the non-trivial monodromy (4.4) of the Td−1 fibers parameterized by z in the original Mostow bundle? To answer this question, we write the monodromy matrix M = (mij) ∈ SL(d− 1,Z) in the block form M = ( M|d−2 ~m ~m′ md−1 d−1 ) , (4.19) where M|d−2 = (mij)1≤i,j≤d−2, while ~m = (mi d−1)i=1,...,d−2 and ~m′ = (md−1 i)i=1,...,d−2 are respectively column and row vectors in Zd−2. Below we denote the usual Euclidean inner product on Rd−2 by 〈 · , · 〉. Proposition 4.18. The noncommutativity parameter ~θ~ed−1 (x) from (4.13) varies with a change of coset representative x ∈ R/Z according to its SL(d − 1,Z) orbit under the action of the monodromy matrix (4.19) by (d−2)-dimensional linear fractional transformations ~θ~ed−1 (x+ 1) = M [ ~θ~ed−1 (x) ] := M|d−2 · ~θ~ed−1 (x) + ~m〈 ~m′, ~θ~ed−1 (x) 〉 +md−1 d−1 . (4.20) Under a change of basis of Zd−1 given by a matrix B ∈ GL(d−1,Z), the corresponding noncom- mutativity parameter varies according to ~θB·~ed−1 (x+ 1) = ( BtMB )[ ~θB·~ed−1 (x) ] . Proof. The key feature stems from the definition (4.2) of the monodromy matrix M: Σ−1 ϕx+x0 = Σ−1ϕx0ϕx = MΣ−1ϕx. The statements then follow from straightforward calculations in components. � Remark 4.19. The right-hand side of (4.20) is an example of a linear fractional transformation in higher dimensions, known from complex analysis, see, e.g., [12]. We can show that it defines a Morita equivalence of noncommutative d−1-tori from Example 2.11. For this, we introduce the skew-symmetric matrix Θ corresponding to the vector ~θ ∈ Rd−2 as in Example 2.2: Θ = ( 0d−2 ~θ −~θ t 0 ) . We denote by gM the element of SO(d − 1, d − 1;Z) corresponding to the monodromy matrix M ∈ SL(d− 1,Z): gM = ( M 0d−1 0d−1 ( Mt )−1 ) ∈ SO(d− 1, d− 1;Z). Introduce matrices Td−1 of order 2 with determinant −1 by Td−1 = ( 1d−1 − Ed−1 Ed−1 Ed−1 1d−1 − Ed−1 ) ∈ O(d− 1, d− 1;Z), where Ed−1 is the matrix unit whose only non-zero element is (Ed−1)d−1 d−1 = 1. Finally, define the element M ∈ SO(d− 1, d− 1;Z) by M = Td−1gMTd−1. Topological T-Duality for Twisted Tori 35 Using the adjugate formula for matrix inverses, a straightforward if tedious calculation then shows that the corresponding SO(d − 1, d − 1;Z) orbit of Θ reproduces the (d−2)-dimensional linear fractional transformation of (4.20): M [Θ] = ( 0d−2 M [ ~θ ] −M [ ~θ ]t 0 ) with M [ ~θ ] = M|d−2 · ~θ + ~m〈 ~m′, ~θ 〉 +md−1 d−1 . It follows that, while the fibre noncommutative tori of Theorem 4.17 are not generally identical under a change of representative x ∈ R/Z, by Example 2.11 they are always Morita equivalent: Td−1 ~θ~ed−1 (x+1) ∼M Td−1 ~θ~ed−1(x) . Recalling that our formulation of topological T-duality takes place in the additive category KK from Section 3.2, this has a natural interpretation: The non-trivial monodromy in the automor- phism group of the Td−1 fibres of the twisted torus TΛG is manifested as a (generally non-trivial) isomorphism in the Morita automorphism group of the fibre noncommutative tori in KK . 5 Three-dimensional solvmanifolds and their T-duals The goal of this final section is to give some explicit examples in low dimensions of the general formalism we have developed in Section 4, recovering some previously known results in the litera- ture from a new perspective, as well as providing several new examples. Note that tori Td in any dimension d are covered by our general framework of Section 4.4 for Ry-actions: when ϕx = 1d−1 for all x ∈ R, the torus bundles Td → Td−n of Proposition 4.13 are trivial. Theorem 4.14 then shows that any C∗-algebraic T-dual C ( Td ) ort Ry0 is isomorphic to CT ( Td−1 × Ty0 , 0 ) ' C ( Td−1 × Ty0 ) ⊗ K. Below we apply our formalism to the well-known Mostow fibrations of three-dimensional solvmanifolds. 5.1 Mostow bundles and SL(2,R) conjugacy classes In three dimensions, solvmanifolds are completely classified, see, e.g., [2]. In particular, by [2, Proposition 5.1] all solvmanifolds in this dimension which are based on connected and simply- connected Lie groups G are almost abelian. If G = R2oϕR is abelian, i.e., ϕx = 12 for all x ∈ R, then TΛG ' T3 is a torus, while the remaining cases correspond to non-trivial one-parameter group actions ϕ : R −→ SL(2,R). Elements of SL(2,R) are classified up to conjugacy by trace, so there are three classes, de- termined by which of the three types of conjugacy classes of SL(2,R) that the image of the homomorphism ϕ lands in: parabolic (|Trϕx| = 2 for all x ∈ R), elliptic (|Trϕx| < 2 for all x 6= k π/2 with k ∈ Z), or hyperbolic (|Trϕx| > 2 for all x 6= 0), see, e.g., [26]. The existence of cocompact discrete subgroups ΛG of G is highly restrictive on the allowed homomorphisms ϕ; a necessary condition is that there exists x0 ∈ R× such that Trϕx0 ∈ Z. The monodromy matrices M = Σ−1 ϕx0Σ = ( a b c d ) ∈ SL(2,Z) (5.1) are integer matrices that live in the corresponding conjugacy classes of SL(2,Z). As automor- phisms of the torus fibres of the Mostow bundle, in the mapping class group SL(2,Z), they act on T2 in the following way. Let ( λ, λ−1 ) be the eigenvalues of M, which are the roots of the characteristic polynomial t2 − (Tr M)t+ 1. Then there are three possibilities: 36 P. Aschieri and R.J. Szabo � Parabolic: In this case Tr M = ± 2 and λ = λ−1 = ± 1. Then M has an integral eigenvector corresponding to a closed curve on T2 which is invariant under the associated automor- phism, and homeomorphisms of this type are Dehn twists of T2. � Elliptic: In this case Tr M = 0,±1, and the eigenvalues ( λ, λ−1 ) are complex of modulus 1. Then the associated automorphism has finite order equal to 2, 3, 4 or 6, and corresponds to a periodic homeomorphism of T2. � Hyperbolic: In this case |λ| > 1 > ∣∣λ−1 ∣∣. The associated automorphism has no invariant closed curve on T2; it ‘stretches’ the eigenspace corresponding to λ and ‘contracts’ the eigenspace corresponding to λ−1. These are called Anasov homeomorphisms of T2. In this section we shall apply Theorem 4.17 with z0 = ~e2 = (0, 1), which for convenience we combine with Proposition 4.18 and Remark 4.19 to reformulate it as Theorem 5.1. The C∗-algebraic T-dual of a three-dimensional solvmanifold TΛG with respect to an Rz-action with z0 = (0, 1) is Morita equivalent to a C∗-algebra bundle of noncommutative two-tori over the circle T: C(TΛG ) ort R(0,1) ∼M ∐ x∈R/Z T2 θ(x), where the noncommutativity parameters θ(x) ∈ R are given by θ(x) =  0 for x ∈ F∗,( Σ−1ϕx0x ) 12( Σ−1ϕx0x ) 22 for x ∈ R \ F∗, where ( Σ−1ϕx0x ) ij for i, j ∈ {1, 2} denote the matrix elements of Σ−1ϕx0x. The noncommu- tativity parameter θ(x) varies under a change of coset representative x ∈ R/Z according to its SL(2,Z) orbit under the corresponding monodromy matrix (5.1): θ(x+ 1) = M [ θ(x) ] = aθ(x) + b cθ(x) + d . (5.2) By Example 2.10 the corresponding fibre noncommutative tori are Morita equivalent: T2 θ(x+1) ∼M T2 θ(x), and so isomorphic in the category KK . The C∗-algebraic T-dual in these instances has more structure in general than the original algebra of functions C(TΛG ), as a consequence of Proposition 5.2. The noncommutative torus bundle ∐ x∈R/Z T2 θ(x) is a C ( T2 ) -algebra. Proof. We can also describe these noncommutative torus bundles via (strict) deformation quan- tization: Performing an identical calculation to that of Example 2.1, we may describe the fibre- wise crossed product (4.18) (with d = 3) explicitly in terms of a star-product, and hence the C∗-algebra bundle of Theorem 5.1 as a deformation of the algebra of functions C ( T × T2 ) on the trivial torus bundle over T by regarding the convolution algebra Cc(T×T×Z) as the space of functions C ( T× T2 ) equipped with the star-product (f ?θ g)(x, v1, v2) = ∑ (p1,p2)∈Z2 e 2πi(p1v1+p2v2) × ∑ (q1,q2)∈Z2 fq1,q2(x)gp1−q1,p2−q2(x) e 2πiθ(x)(p1−q1)q2 , (5.3) Topological T-Duality for Twisted Tori 37 where we used the fibrewise Fourier transformation f(x, v1, v2) = ∑ (p1,p2)∈Z2 fp1,p2(x) e 2πi(p1v1+p2 v2), with functions fp1,p2 : T→ C. In this formulation of the C∗-algebraic T-dual, it is immediately evident that there is an injection C ( T2 ) ↪→ ∐ x∈R/Z T2 θ(x) as the space of functions which are independent of the coordinate v1 (or v2); for such functions, the sums in (5.3) truncate to p1 = q1 = 0 (or p2 = q2 = 0) and the noncommutative star-product reduces to the commutative pointwise product of functions in C ( T2 ) . This defines C∗-algebra monomorphisms making∐ x∈R/Z T2 θ(x) into a C ( T2 ) -algebra. � The remainder of this paper is devoted to providing illustrations of Theorem 5.1, through explicit calculations in each of the three conjugacy classes of SL(2,R). The features differ for each conjugacy class so we consider them individually in turn. 5.2 Parabolic torus bundles We start with the best studied example in the literature, which is based on the nilpotent Heisen- berg group. In string theory it is T-dual to the three-torus T3 with H-flux by the standard Buscher rules (see, e.g., [27]), and in topological T-duality it gives the basic example of a non- commutative principal torus bundle [15, 21, 32]. Here we shall give a new algebraic perspective on both these T-duals by applying our formalism of topological T-duality directly to the Heisenberg nilmanifold. This class has several features that make it special among the three-dimensional solvmanifolds, which we explain in detail. Heisenberg nilmanifolds The three-dimensional Heisenberg group Heis(3) is the nilpotent Lie group whose Mostow bundle structure is based on the semi-direct product Heis(3) = R2 oϕ R with ϕx = ( 1 x 0 1 ) . Since |Trϕx| = 2, the matrix ϕx parameterizes a parabolic conjugacy class of SL(2,R). Here we denote the coordinates on the group manifold of N = R2 by (y, z). The group multiplication on Heis(3) is then given by (x, y, z)(x′, y′, z′) = (x+ x′, y + y′ + xz′, z + z′), and the inverse of a group element is (x, y, z)−1 = (−x, xz − y,−z). In this case it is clear that for x0 = m with m ∈ Z×, the matrix ϕx0 is integer-valued and hence may be taken as monodromy matrix Mm of infinite order with Σ = 12: Mm = ϕm = ( 1 m 0 1 ) . (5.4) 38 P. Aschieri and R.J. Szabo The Heisenberg nilmanifold THeism(3;Z) is the compact space obtained as the quotient of Heis(3) with respect to the lattice given by the discrete Heisenberg group Heism(3;Z) := { (mα, β, γ) ∈ Heis(3) ∣∣α, β, γ ∈ Z } . The equivalence relation is given by the left action of Heism(3;Z), which leads to the local coordinate identifications under the action of the generators of Heism(3;Z) given by (x, y, z) 7−→ (x+m, y +mz, z), (x, y, z) 7−→ (x, y + 1, z), (x, y, z) 7−→ (x, y, z + 1). (5.5) Geometrically, this exhibits the Heisenberg nilmanifold as a non-trivial principal T2-bundle THeism(3;Z) → T. Using the algorithm described in Section 4.4, the relation matrix Am = Mm − 12 has rank r = 1 with elementary divisors m1 = m and m2 = 0, and the first homology group of THeism(3;Z) can thus be presented as the Z-module H1(THeism(3;Z),Z) ' Z⊕ (Z⊕ Zm), with respective free generators denoted ex and ey, and torsion generator ez of order m, mez = 0. Dually, the topology of the nilmanifold is captured by its algebra of functions, which can be computed as the subalgebra of invariant functions on the Heisenberg group Heis(3) with respect to the left action of the lattice Heism(3;Z): C ( THeism(3;Z) ) = C0 ( Heis(3) )Heism(3;Z) . Harmonic analysis on the nilmanifold can be used to determine the Fourier decomposition of any function f ∈ C(THeism(3;Z)). Invariance under the generators (5.5) of Heism(3;Z) forces the expansion to take the form f(x, y, z) = ∑ k∈Zm ∑ p,q∈Z fp,k(x+mq) e 2πipy+2πi(k+mpq)z, (5.6) for functions fp,k : R→ C vanishing at infinity. From this expression we can easily see the C(T)- algebra structure of C(THeism(3;Z)): Restricting (5.6) to y = z = 0 determines a function on the circle T = {( e 2πix/m, 1, 1 ) ∈ THeism(3;Z) } and defines a C∗-algebra monomorphism C(T) ↪→ C(THeism(3;Z)). On the other hand, the evaluation of (5.6) at z = 0 yields a function on the two-torus T2 = {( e 2πix/m, e 2πiy, 1 ) ∈ THeism(3;Z) } and defines a C∗-algebra monomorphism C ( T2 ) ↪→ C(THeism(3;Z)) making C(THeism(3;Z)) into a C ( T2 ) -algebra in this case. Following the prescriptions of Sections 4.5 and 4.6, we shall now study the known T-duals of the Heisenberg nilmanifold through the formalism of topological T-duality discussed in Section 3. Ry-action: T3 with H-flux The one-parameter subgroup ker(Am) = R(1,0) := { (0, λ, 0) ∈ Heis(3) } is the center of Heis(3). The quotient group Heis(3)/R(1,0) can be parameterized by equivalence classes [ (x, 0, z) ] , and the multiplication law[ (x, 0, z) ] [ (x′, 0, z′) ] = [ (x+ x′, 0, z + z′) ] Topological T-Duality for Twisted Tori 39 is that of the abelian Lie group Heis(3)/R(1,0) ' R2. It follows that the quotient THeism(3;Z)/R(1,0) is the two-dimensional torus T2 with coordinates ( e 2πix/m, e 2πiz ) . The corresponding principal circle bundle p(1,0) : THeism(3;Z) → T2, with fibre coordinate e 2πiy, is the standard realization of the Heisenberg nilmanifold as a T-fibration over T2 of degree m: the connection κ(1,0) = −dy + xdz of Theorem 4.14 has curvature dκ(1,0) = dx ∧ dz, and so the first Chern class c1(p(1,0)) is m times the standard generator of H2 ( T2,Z ) ' Z. By Theorem 4.14 it follows that the C∗-algebraic T-dual C ( THeism(3;Z) ) ort R(1,0) ' CT ( T3,m ) is a continuous-trace algebra with spectrum T3 and Dixmier–Douady class equal to m times the standard generator of H3 ( T3,Z ) ' Z. Thus in this case we recover the standard T-duality between the three-torus T3 with H-flux and the Heisenberg nilmanifold viewed as a circle bundle. The correspondence space construction is given by Proposition 4.15 with TΛ G(1,0) = T2. Rz-action: Noncommutative principal torus bundles We shall now recover the known noncommutative principal torus bundle which is T-dual to THeism(3;Z) from a new perspective, by directly working with the algebraic description of the nilmanifold. For this, we follow our prescription for Rz-actions from Section 4.6 and consider the one-parameter subgroup R(0,1) of Heis(3) given by R(0,1) = { (0, 0, ζ) ∈ Heis(3) } , whose right action by multiplication generates (x, y, z)(0, 0, ζ) = (x, y + xζ, z + ζ). The quotient of this action to THeism(3;Z) fixes every point of the form ( e 2πin/m, e 2πiy, e 2πiz ) ∈ T2 ⊂ THeism(3;Z) for n ∈ Z with isotropy subgroup Z(0,1) = R(0,1) ∩ Heism(3;Z). In this case F∗ = ∅, and Theorem 5.1 thus identifies the C∗-algebraic T-dual through the Morita equivalence C ( THeism(3;Z) ) ort R(0,1) ∼M ∐ x∈R/Z T2 mx of C(T)-algebras. We can explicitly check the anticipated monodromy (5.2): Under a change of coset rep- resentative x ∈ R/Z of the fibres of the noncommutative torus bundle ∐ x∈R/Z T2 θm(x), the noncommutativity parameter changes θm(x+ 1) = θm(x) +m = Mm [ θm(x) ] (5.7) by the SL(2,Z) orbit of θm(x) = mx under the monodromy matrix Mm given by (5.4), and by Example 2.10 the fiber noncommutative tori are identical: T2 θm(x+1) = T2 θm(x). Hence in this 40 P. Aschieri and R.J. Szabo case the non-trivial monodromy in the automorphism group of the T2 fibres of the twisted torus is implemented trivially as an equality in its T-dual C∗-algebra bundle. It is well known that these C∗-bundles are isomorphic to the convolution C∗-algebras of the corresponding integer Heisenberg groups, and indeed this is precisely the original description of these T-dual noncommutative torus bundles from [32]. For later comparison, it is instructive to explicitly demonstrate this result using our scheme for topological T-duality, which results in Proposition 5.3. The C∗-algebraic T-dual C ( THeism(3;Z) ) ort R(0,1) is Morita equivalent to the group C∗-algebra C∗ ( Heism(3;Z) ) . Proof. The crux of the proof hinges on the fact that the parabolic torus bundles are the only class in three dimensions for which a fibrewise analysis is not necessary and the Morita isomorphisms can be implemented directly in the category KK . For this, we observe that any element (x, y, z) ∈ Heis(3) can be uniquely factorized as (x, y, z) = (x, y − xz, 0)(0, 0, z) =: (x, v, 0)(0, 0, z), with the action of the subgroup R(0,1) on the normal abelian subgroup R2 (x,v) := {(x, v, 0) ∈ Heis(3)} given by Adz(x, v, 0) := (0,0,z)(x, v, 0) = (0, 0, z)(x, v, 0)(0, 0,−z) = (x, v − xz, 0). It follows that the Heisenberg group can alternatively be presented as the semi-direct product Heis(3) = R2 (x,v)R(0,1). Correspondingly, its lattice is also a semi-direct product Heism(3;Z) = Z2 (x,v) oAd Z(0,1), where Z2 (x,v) = R2 (x,v) ∩ Heism(3;Z). Thus the Morita equivalence from (3.3) can be expressed as C ( THeism(3;Z) ) ort R(0,1) ∼M C0 ( Heis(3)/R(0,1) ) olt Heism(3;Z) ∼M C0 ( R2 (x,v) ) olt ( Z2 (x,v) oAd Z(0,1) ) ' ( C0 ( R2 (x,v) ) olt Z2 (x,v) ) oβ Z(0,1), (5.8) where in the last line we have further applied Theorem 2.6. Since the groups Z2 (x,v) and Z(0,1) are discrete, the homomorphism σZ(0,1) from (2.5) is trivial, and so the Z(0,1)-action on C0 ( R2 (x,v) ) olt Z2 (x,v) is the canonical action obtained from the diagonal action of Z(0,1) on Z2 (x,v) × R2 (x,v): βζ(f) ( (mα, ν), (x, v) ) = f ( (mα, ν +mαζ), (x, v + xζ) ) for all ζ ∈ Z(0,1), (mα, ν) ∈ Z2 (x,v), (x, v) ∈ R2 (x,v) and f ∈ Cc ( Z2 (x,v) × R2 (x,v) ) ⊂ C0 ( R2 (x,v) ) olt Z2 (x,v). By Example 2.14 there is a Morita equivalence C0 ( R2 (x,v) ) olt Z2 (x,v) ∼M C ( T2 ) . (5.9) Since the homomorphism σZ(0,1) : Z(0,1) → R+ is trivial, it follows that the Morita equivalence bimodule M implementing this equivalence is Z(0,1)-equivariant: the Z(0,1)-action U : Z(0,1) → Aut(M) is given by Uζ(ξ)(x, v) = ξ(x, v + xζ), for all ζ ∈ Z(0,1), ξ ∈ M and (x, v) ∈ R2 (x,v). Applying Theorem 2.12 we conclude that the Morita equivalence (5.9) induces the Morita equiv- alence ( C0 ( R2 (x,v) ) oltZ2 (x,v) ) oβZ(0,1) ∼M C ( T2 ) oAd∗Z(0,1), where the action of the group Z(0,1) on C ( T2 ) is the pullback of the action on T2 = R2 (x,v)/Z 2 (x,v) induced from the left multiplication of Z(0,1) on Heis(3). Topological T-Duality for Twisted Tori 41 Substituting into (5.8) we thus obtain C(THeism(3;Z)) ort R(0,1) ∼M C ( T2 ) oAd∗ Z(0,1). (5.10) Now we implement Pontryagin duality via Fourier transform and apply (2.8) to the right-hand side of (5.10) to obtain C ( T2 ) oAd∗ Z(0,1) ' C∗ ( Z2 (x,v) ) o Âd∗ Z(0,1), where we used the fact that σ̂Z(0,1) : Z(0,1) → R+ is trivial since Z(0,1) is discrete. Applying Theorem 2.5, we then arrive at C ( THeism(3;Z) ) ort R(0,1) ∼M C∗ ( Z2 (x,v) oAd Z(0,1) ) = C∗ ( Heism(3;Z) ) , as required. � Remark 5.4. The star-products (5.3) with θ(x) = θm(x) = mx are equivalent to the star- products of [21, 28, 31]. In [15, 21], it is shown that the description of Proposition 5.3 defines a noncommutative principal T2-bundle (cf. Example 4.9). It is also shown in [15] that the monodromy (5.7), while acting trivially at the purely algebraic level, has a non-trivial action on the K-theory group of the C∗-algebra bundle ∐ x∈R/Z T2 mx, viewed as a bundle of abelian groups over T; a physical picture of this action in terms of monodromies of fiber D-branes is discussed in [28]. Noncommutative correspondences We come now to the noncommutative correspondences underlying this topological T-duality. For this, we recall that both C ( THeism(3;Z) ) and C ( T2 ) oAd∗ Z(0,1) are C ( T2 ) -algebras, with T2 parameterized by the local coordinates (x, y). We will show that the noncommutative corre- spondence is given by the balanced tensor product C ( THeism(3;Z) ) ort Z(0,1) ' C ( THeism(3;Z) ) ⊗C(T2) ∐ x∈R/Z T2 mx over C ( T2 ) . Note that here the subgroup Z(0,1) acts non-trivially on the algebra of functions C(THeism(3;Z)), in contrast to our previous examples. We consider elements of the convolution algebra Cc(THeism(3;Z) × Z(0,1)) as sequences f = {fq̃}q̃∈Z(0,1) of functions fq̃ : THeism(3;Z) → C with the convolution product (f ? g)q̃(x, y, z) = ∑ p̃∈Z fp̃(x, y, z)gq̃−p̃(x, y − p̃mx, z). We write the Fourier transformation f(x, y, z, z̃) := ∑ q̃∈Z fq̃(x, y, z) e 2πiq̃z̃ for functions on THeism(3;Z) × T̃z, and define the star-product (f ? g)(x, y, z, z̃) = ∑ q̃∈Z (f ? g)q̃(x, y, z) e 2πiq̃z̃. We further use the harmonic expansion of functions on the nilmanifold given from (5.6) by fq̃(x, y, z) = ∑ k∈Zm ∑ p,q∈Z fq̃;p,k(x+ q) e 2πiqy+2πi(k+mpq)z. 42 P. Aschieri and R.J. Szabo After some elementary manipulations, we can then write the star-product on the space of func- tions C ( THeism(3;Z) × T̃z ) as (f ? g)(x, y, z, z̃) = ∑ l∈Zm ∑ r,s,p̃∈Z e 2πiry+2πi(l+mrs)z e 2πip̃z̃ × ∑ k∈Zm ∑ p,q,q̃∈Z fq̃;p,k(x+ q)gp̃−q̃;r−p,l−k(x+ q − s) e 2πimr(q−s)z e−2πimx(r−p)q̃. (5.11) We can define an injection C ( THeism(3;Z) ) ↪→ C ( THeism(3;Z) × T̃z ) as the space of functions which are independent of the coordinate z̃. For these functions the sums in (5.11) truncate to p̃ = q̃ = 0, and one recovers the pointwise product of functions in C ( THeism(3;Z) ) . On the other hand, we can include the noncommutative algebra ∐ x∈R/Z T2 mx ↪→ C ( THeism(3;Z) × T̃z ) as the space of functions which are independent of the coordinate z, and one sees that (5.11) truncates for l = k = 0 and s = q = 0 to the star-product (5.3) with θ(x) = θm(x) = mx. Altogether, the noncommutative correspondences are induced by the diagram C ( THeism(3;Z) ) ⊗ C ( T2 ) ∐ x∈R/Z T2 mx C ( THeism(3;Z) ) 55 ∐ x∈R/Z T2 mx ii C ( T2 ) 55jj where all arrows are ∗-monomorphisms of C∗-algebras. 5.3 Elliptic torus bundles We shall now illustrate Theorem 5.1 for the three-dimensional solvmanifolds based on the Eu- clidean group in two dimensions, which have no classical Hausdorff T-duals and whose noncom- mutative T-dual fibration was first discussed in [28] for the case of the Z4 elliptic monodromy. Here we shall recover this noncommutative geometry rigorously from our algebraic framework, and moreover extend it to the Z2 and Z6 elliptic monodromies which have not been considered previously. The Euclidean group ISO(2) in two dimensions is the three-dimensional almost abelian solv- able Lie group ISO(2) = R2 oϕ R with ϕx = ( cosx sinx − sinx cosx ) . Since |Trϕx| < 2 for x /∈ πZ, the rotations ϕx parameterize an elliptic conjugacy class of SL(2,R). Here we shall denote coordinates on the group manifold of N = R2 by z = (z1, z2). The group multiplication on ISO(2) is then (x, z1, z2) (x′, z′1, z ′ 2) = ( x+ x′, z1 + z′1 cosx+ z′2 sinx, z2 − z′1 sinx+ z′2 cosx ) and the inverse of a group element is (x, z1, z2)−1 = ( −x,−z1 cosx+ z2 sinx,−z1 sinx− z2 cosx ) . The existence of lattices in ISO(2) is highly restrictive [26, 28]. They require angles x0 ∈ R× for which 2 cosx0 ∈ Z, i.e., cosx0 = 0,±1 2 ,±1. For our purposes, there are three inequivalent Topological T-Duality for Twisted Tori 43 choices of non-trivial elliptic monodromies, which are characterized by matrices M ∈ SL(2,Z) of finite order in the cyclic subgroups Z2, Z4 and Z6. They lead to three types of Euclidean solvmanifolds, according to the order of the monodromy in the fibre of the associated Mostow bundle. A crucial distinction from the case of the Heisenberg nilmanifold is that here the Mostow fibrations are not principal T2-bundles. We consider each of the three cases separately in turn. Euclidean Z2-solvmanifolds We observe that for x0 = mπ with m ∈ Z×, the matrix ϕx0 is integer-valued, and so may be set equal to M with Σ = 12. For m even, the monodromy M = 12 is trivial and the corresponding Mostow bundle is simply the three-torus T3, which we have already considered in Section 3.4. For m odd, which we now assume, the nontrivial monodromy matrix is given by M(1) = ϕmπ = −12. (5.12) The Euclidean Z2-solvmanimanifold T ISO (1) m (2;Z) is the compact space obtained as the quotient of ISO(2) with respect to the lattice given by the discrete Euclidean group ISO(1) m (2;Z) := { (mπα, γ1, γ2) ∈ ISO(2) ∣∣α ∈ Z, γ = (γ1, γ2) ∈ Z2 } . The quotient is taken by the left action of ISO(1) m (2;Z), which leads to the local coordinate identifications under the action of the generators of ISO(1) m (2;Z) given by (x, z1, z2) 7−→ (x+mπ,−z1,−z2), (x, z1, z2) 7−→ (x, z1 + 1, z2), (x, z1, z2) 7−→ (x, z1, z2 + 1). The relation matrix A(1) = M(1) − 12 = −212 has maximal rank r = 2 with elementary divisors m1 = m2 = 2, and the first homology group of T ISO (1) m (2;Z) can thus be presented as the Z-module H1 ( T ISO (1) m (2;Z) ,Z ) ' Z⊕ (Z2 ⊕ Z2). It follows that the Z2-solvmanifolds do not possess any classical T-duals. By symmetry the two possible noncommutative torus bundles corresponding to the two torsion generators ez1 and ez2 of order two are the same. Consider the one-parameter subgroup R(0,1) of ISO(2) given by R(0,1) = { (0, 0, ζ) ∈ ISO(2) } , which acts on the Euclidean group ISO(2) by right multiplication (x, z1, z2)(0, 0, ζ) = (x, z1 + ζ sinx, z2 + ζ cosx). In this case F∗ is the subset of x ∈ R where cosx = 0, so that F∗ = π 2 (2Z + 1). (5.13) Applying Theorem 5.1, it follows that the C∗-algebraic T-dual of the Euclidean Z2-solvmanifold is Morita equivalent to the C(T)-algebra C ( T ISO (1) m (2;Z) ) ort R(0,1) ∼M ∐ x∈R/Z T2 θ (1) m (x) , (5.14) 44 P. Aschieri and R.J. Szabo where θ(1)m (x) = { 0 for x ∈ 1 m ( Z + 1 2 ) , tan(mπx) for x ∈ R \ 1 m ( Z + 1 2 ) . (5.15) Recalling that the integer m is odd here, it is easily seen that the noncommutativity para- meter (5.15) is invariant under changing coset representative x ∈ R/Z, which is consistent with its SL(2,Z) orbit under the corresponding monodromy matrix (5.12): θ(1)m (x+ 1) = θ(1)m (x) = M(1) [ θ(1)m (x) ] , so that fibrewise A θ (1) m (x+1) = A θ (1) m (x) . Thus the non-trivial monodromy in the automorphism group of the T2 fibres of the twisted torus T ISO (1) m (2;Z) becomes a trivial identity action on its T- dual C∗-algebra bundle. This is analogous to what we found in the parabolic case. Moreover, by Proposition 5.2 the C∗-algebra bundle (5.14) can be described as a deformation of the algebra of functions C ( T×T2 ) on a trivial T2-bundle over the circle via a star-product f? θ (1) m g given by (5.3). In marked contrast to the parabolic case, however, the algebra of functions C ( T ISO (1) m (2;Z) ) is not itself a C ( T2 ) -algebra, and the noncommutative torus bundle cannot be identified with the group C∗-algebra of the integer Euclidean group ISO(1) m (2;Z). Euclidean Z4-solvmanifolds We next observe that for x0 = mπ 2 with m an odd integer, the matrix ϕx0 is again integer-valued, and so may be set equal to M with Σ = 12: M(2) = ϕmπ 2 = ± ( 0 1 −1 0 ) . (5.16) Without loss of generality, we shall fix the positive sign by assuming that m ∈ 4Z+1 (the choice of negative sign for m ∈ 4Z + 3 simply corresponds to a reflection (z1, z2) 7→ (−z1,−z2) below). The Euclidean Z4-solvmanifold T ISO (2) m (2;Z) is the compact space obtained as the quotient of ISO(2) with ISO(2) m (2;Z) := {( mπ 2 α, γ1, γ2 ) ∈ ISO(2) ∣∣α ∈ Z, γ = (γ1, γ2) ∈ Z2 } . The quotient is taken by the left action of ISO(2) m (2;Z), which leads to the local coordinate identifications under the action of the generators of ISO(2) m (2;Z) given by (x, z1, z2) 7−→ ( x+ mπ 2 , z2,−z1 ) , (x, z1, z2) 7−→ (x, z1 + 1, z2), (x, z1, z2) 7−→ (x, z1, z2 + 1). The relation matrix A(2) = M(2) − 12 has maximal rank r = 2 with elementary divisors m1 = 1 and m2 = 2, and the first homology group of T ISO (2) m (2;Z) can thus be presented as the Z-module H1 ( T ISO (2) m (2;Z) ,Z ) ' Z⊕ Z2, where the torsion generator of order two is ez2 = ez1 . Thus again there is no classical T-dual, as is well-known in this case (see, e.g., [28]). Topological T-Duality for Twisted Tori 45 Proceeding as in the previous case, the subset F∗ ⊂ R is again given by (5.13) and we arrive at the Morita equivalence C ( T ISO (2) m (2;Z) ) ort R(0,1) ∼M ∐ x∈R/Z T2 θ (2) m (x) , where θ(2)m (x) = { 0 for x ∈ 1 m(2Z + 1), tan ( mπ 2 x ) for x ∈ R \ 1 m(2Z + 1). (5.17) The corresponding star-product f ? θ (2) m g on C ( T × T2 ) is given by (5.3) and was originally written down in [28], where it was also pointed out that the noncommutative torus fibration is not isomorphic to the group C∗-algebra of ISO(2) m (2;Z). Compared to the previous cases, the new feature here is that a change of coset representa- tive x ∈ R/Z generally has a non-trivial action on the fibers of the C∗-bundle ∐ x∈R/Z Aθ(2)m (x) according to (5.2). We check this explicitly here: Recalling that m ∈ 4Z + 1 in this case, the fibre over any x ∈ 1 mZ is preserved as then the form of (5.17) implies θ(2)m (x+ 1) = 0 = θ(2)m (x) for x ∈ 1 mZ. However, over any x ∈ R \ 1 mZ the noncommutativity parameter (5.17) changes non-trivally according to the trigonometric identity tan ( x+ mπ 2 ) = − cotx, but in a way which is consistent with its SL(2,Z) orbit under the corresponding monodromy matrix (5.16): θ(2)m (x+ 1) = − 1 θ(2)m (x) = M(2) [ θ(2)m (x) ] for x ∈ R \ 1 mZ. By Theorem 5.1, the fibre noncommutative tori are Morita equivalent, and so are isomorphic in the category KK . A physical picture of the action of the monodromy on the K-theory group of this C∗-bundle in terms of fiber D-branes is given in [28]. Euclidean Z6-solvmanifolds Finally, we observe that for x0 = mπ 3 with m /∈ 3Z, the matrix ϕx0 takes four possible forms ϕmπ 3 = ε 2 ( ε′ √ 3 − √ 3 ε′ ) with independent signs ε, ε′ = ±1. It conjugates to an integer matrix via the element Σ ∈ SL(2,R) given by Σ = √ 2√ 3 ( 1 1 2 0 √ 3 2 ) . (5.18) The precise form of ϕx0 and hence of the corresponding monodromy matrix M depends on the congruence class of the integer m in Z3: Σ−1 ϕmπ 3 Σ = { ± ( 1 1 −1 0 ) , m ∈ 3Z + 1, ± ( 0 1 −1 −1 ) , m ∈ 3Z + 2. 46 P. Aschieri and R.J. Szabo Without loss of generality, we shall fix m ∈ 6Z + 2 and hence take M(3) = ( 0 1 −1 −1 ) (5.19) in the following (with the other three possibilities obtained simply by reflection (z1, z2) 7→ (−z1,−z2) and interchange (z1, z2) 7→ (z2, z1) below). The Euclidean Z6-solvmanifold T ISO (3) m (2;Z) is the compact space obtained as the quotient of ISO(2) with respect to the lattice given by the discrete subgroup ISO(3) m (2;Z) := {( mπ 3 α, √ 2√ 3 γ1 + √ 1 2 √ 3 γ2, √√ 3 2 γ2 ) ∈ ISO(2) ∣∣∣α ∈ Z, γ = (γ1, γ2) ∈ Z2 } . The quotient is taken by the left action of ISO(3) m (2;Z), which leads to the local coordinate identifications under the action of the generators of ISO(3) m (2;Z) given by (x, z1, z2) 7−→ ( x+ mπ 3 , 1 2 (√ 3z2 − z1 ) ,−1 2 (√ 3z1 + z2 )) , (x, z1, z2) 7−→ ( x, z1 + √ 2√ 3 , z2 ) , (x, z1, z2) 7−→ ( x, z1 + √ 1 2 √ 3 , z2 + √√ 3 2 ) . The relation matrix A(3) = M(3) − 12 has maximal rank r = 2 with elementary divisors m1 = 1 and m2 = 3, and the first homology group can thus be presented as the Z-module H1 ( T ISO (3) m (2;Z) ,Z ) ' Z⊕ Z3, where the torsion generator of order three is ez2 = ez1 . Here too there is no classical T-dual. Applying Theorem 5.1 in this case, with the non-trivial period matrix Σ from (5.18) and the subset F∗ ⊂ R again given by (5.13), gives the Morita equivalence C ( T ISO (3) m (2;Z) ) ort R(0,1) ∼M ∐ x∈R/Z T2 θ (3) m (x) , where θ(3)m (x) = { 0 for x ∈ 1 m ( 3Z + 3 2 ) , −1 2 + √ 3 2 tan ( mπ 3 x ) for x ∈ R \ 1 m ( 3Z + 3 2 ) . (5.20) As in the previous case, changing coset representative x ∈ R/Z generally has a non-trivial action on the fibers of the C∗-bundle over T: Recalling that m ∈ 6Z+ 2 here, the fiber over any x ∈ 1 m ( 3Z + 3 2 ) is preserved by the form of (5.20): θ(3)m (x+ 1) = 0 = θ(3)m (x) for x ∈ 1 m ( 3Z + 3 2 ) . On the other hand, over any x ∈ R\ 1 m ( 3Z+ 3 2 ) the noncommutativity parameter (5.20) changes according to the trigonometric identity tan ( x+ mπ 3 ) = tanx− √ 3 1 + √ 3 tanx , but consistently with its SL(2,Z) orbit under the corresponding monodromy matrix (5.19): θ(3)m (x+ 1) = − 1 θ(3)m (x) + 1 = M(3) [ θ(3)m (x) ] for x ∈ R \ 1 m ( 3Z + 3 2 ) . By Theorem 5.1 the corresponding fibre noncommutative tori are Morita equivalent, and so isomorphic in the category KK . Topological T-Duality for Twisted Tori 47 5.4 Hyperbolic torus bundles Our last application of Theorem 5.1 is to the final class of three-dimensional solvmanifolds, which are hyperbolic analogues of the Euclidean solvmanifolds based on the Poincaré group, and have also not been previously studied in the present context. The Poincaré group ISO(1, 1) in two dimensions is the three-dimensional almost abelian solvable Lie group which can be presented as ISO(1, 1) = R2 oϕ R with ϕx = ( coshx sinhx sinhx coshx ) . Since |Trϕx| > 2 for all x ∈ R×, the hyperbolic rotation ϕx parameterizes a hyperbolic conjugacy class of SL(2,R). Again we denote coordinates on the group manifold of N = R2 by z = (z1, z2), so that the group multiplication on ISO(1, 1) is given by (x, z1, z2)(x′, z′1, z ′ 2) = ( x+ x′, z1 + z′1 coshx+ z′2 sinhx, z2 + z′1 sinhx+ z′2 coshx ) and the inverse of a group element is (x, z1, z2)−1 = ( −x,−z1 coshx+ z2 sinhx, z1 sinhx− z2 coshx ) . Clearly there is no x0 ∈ R× for which ϕx0 is an integer matrix in this case. However, there is an infinite family of discrete points x0 = log ( m 2 ± √(m 2 )2 − 1 ) (5.21) labelled by an integer m > 2 with 2 coshx0 = m, at which the matrix ϕx0 takes the form ϕx0 = 1 2 ( m ± √ m2 − 4 ± √ m2 − 4 m ) . This matrix has eigenvalues ( λ, λ−1 ) where λ±1 = 1 2 ( m± √ m2 − 4 ) = e x0 , and it conjugates to the integer matrix Mm = Σ−1ϕx0Σ = ( m 1 −1 0 ) (5.22) by the element Σ ∈ SL(2,R) given by Σ =  λ±1 2 ( λ− λ−1 ) − λ∓1 λ 2 ( λ2 − 1 ) − 1 λ±1 2 ( λ− λ−1 ) + λ∓1 λ 2 ( λ2 − 1 ) + 1  . (5.23) For definiteness, we choose the positive square root in (5.21) (with the choice of negative square root obtained from the interchange (z1, z2) 7→ (z2, z1) below). The Poincaré solvmanifold TISOm(1,1;Z) is the compact space obtained as the quotient of ISO(1, 1) by the lattice ISOm(1, 1;Z) := { (x0 α,Σ · γ) ∈ ISO(1, 1) ∣∣α ∈ Z, γ = (γ1, γ2) ∈ Z2 } . 48 P. Aschieri and R.J. Szabo The quotient by the left action of this lattice leads to the local coordinate identifications under the action of the generators given by (x, z1, z2) 7−→ ( x+ cosh−1 ( m 2 ) , 1 2 ( mz1 + √ m2 − 4z2 ) , 1 2 (√ m2 − 4z1 +mz2 )) , (x, z1, z2) 7−→ ( x, z1 + λ2 2(λ2−1) − λ−1, z2 + λ2 2(λ2−1) + λ−1 ) , (x, z1, z2) 7−→ ( x, z1 + λ 2(λ2−1) − 1, z2 + λ 2(λ2−1) + 1 ) . The relation matrix Am = Mm − 12 has maximal rank r = 2 with elementary divisors m1 = 1 and m2 = m− 2, so the first homology group of the Poincaré solvmanifold may be presented as the Z-module H1 ( TISOm(1,1;Z),Z ) ' Z⊕ Zm−2, with torsion generator ez2 = −ez1 of order m− 2. Consider the one-parameter subgroup R(0,1) = { (0, 0, ζ) ∈ ISO(1, 1) } acting on the Poincaré group ISO(1, 1) by right multiplication (x, z1, z2)(0, 0, ζ) = (x, z1 + ζ sinhx, z2 + ζ coshx). In this case the set F∗ consists of a distinguished point x∗ ∈ R on the base of the C∗-algebra bundle given by x∗ = tanh−1 ( 2 ( λ2 − 1 ) − λ3 2 ( λ2 − 1 ) + λ3 ) . Applying Theorem 5.1, with the periods (5.21) and (5.23), then identifies the Morita equivalence C ( TISOm(1,1;Z) ) ort R(0,1) ∼M ∐ x∈R/Z T2 θm(x) of C(T)-algebras, where θm(x∗) = 0 and θm(x) = 2λ ( λ2 − 1 ) − λ2 + λ ( 2 ( λ2 − 1 ) + λ ) tanh ( cosh−1 ( m 2 ) x ) λ3 − 2 ( λ2 − 1 ) − ( λ3 + 2 ( λ2 − 1 )) tanh ( cosh−1(m2 )x ) (5.24) for x 6= x∗. This noncommutative torus bundle, along with its topological T-duality with C ( TISOm(1,1;Z) ) , has formally the same properties as the C(T)-algebras described in Section 5.3, so we refrain from repeating the details here. We only mention the fibre monodromy behaviour anticipated from (5.2): Using the hyperbolic identity tanh ( x+ cosh−1 ( m 2 )) = m tanhx+ √ m2 − 4 m+ √ m2 − 4 tanhx , we find θm(x∗ + 1) = 0 = θm(x∗) Topological T-Duality for Twisted Tori 49 and we see explicitly here that the noncommutativity parameter (5.24) changes consistently with its SL(2,Z) orbit under the monodromy matrix (5.22): θm(x+ 1) = − 1 θm(x) −m = Mm [ θm(x) ] for x 6= x∗. Thus the fiber noncommutative tori are Morita equivalent by Theorem 5.1, and so are isomorphic C∗-algebras in the category KK . Acknowledgments We thank Ryszard Nest and Erik Plauschinn for helpful discussions. We thank the anonymous referees for their detailed suggestions. This research was supported by funds from Università del Piemonte Orientale (UPO). P.A. acknowledges partial support from INFN, CSN4, and Iniziativa Specifica GSS. P.A. is affiliated to INdAM-GNFM. R.J.S. acknowledges a Visiting Professorship through UPO Internationalization Funds. R.J.S. also acknowledges the Arnold–Regge Centre for the visit, and INFN. The work of R.J.S. was supported in part by the Consolidated Grant ST/P000363/1 from the UK Science and Technology Facilities Council. References [1] Andriot D., New supersymmetric vacua on solvmanifolds, J. High Energy Phys. 2016 (2016), no. 2, 112, 43 pages, arXiv:1507.00014. [2] Bock C., On low-dimensional solvmanifolds, Asian J. Math. 20 (2016), 199–262, arXiv:0903.2926. [3] Bouwknegt P., Evslin J., Mathai V., T -duality: topology change from H-flux, Comm. Math. Phys. 249 (2004), 383–415, arXiv:hep-th/0306062. [4] Bouwknegt P., Pande A.S., Topological T -duality and T -folds, Adv. Theor. Math. Phys. 13 (2009), 1519– 1539, arXiv:0810.4374. [5] Brace D., Morariu B., Zumino B., Dualities of the Matrix model from T -duality of the type II string, Nuclear Phys. B 545 (1999), 192–216, arXiv:hep-th/9810099. [6] Brodzki J., Mathai V., Rosenberg J., Szabo R.J., D-branes, RR-fields and duality on noncommutative manifolds, Comm. Math. Phys. 277 (2008), 643–706, arXiv:hep-th/0607020. [7] Brodzki J., Mathai V., Rosenberg J., Szabo R.J., Non-commutative correspondences, duality and D-branes in bivariant K-theory, Adv. Theor. Math. Phys. 13 (2009), 497–552, arXiv:0708.2648. [8] Bunke U., Rumpf P., Schick T., The topology of T -duality for Tn-bundles, Rev. Math. Phys. 18 (2006), 1103–1154, arXiv:math.GT/0501487. [9] Bunke U., Schick T., On the topology of T -duality, Rev. Math. Phys. 17 (2005), 77–112, arXiv:math.GT/0405132. [10] Console S., Macr̀ı M., Lattices, cohomology and models of 6-dimensional almost abelian solvmanifolds, Rend. Semin. Mat. Univ. Politec. Torino 74 (2016), 95–119, arXiv:1206.5977. [11] Corwin L.J., Greenleaf F.P., Representations of nilpotent Lie groups and their applications. Part I. Basic theory and examples, Cambridge Studies in Advanced Mathematics, Vol. 18, Cambridge University Press, Cambridge, 1990. [12] Cowen C.C., MacCluer B.D., Linear fractional maps of the ball and their composition operators, Acta Sci. Math. (Szeged) 66 (2000), 351–376. [13] Dabholkar A., Hull C., Duality twists, orbifolds, and fluxes, J. High Energy Phys. 2003 (2003), no. 9, 054, 25 pages, arXiv:hep-th/0210209. [14] Echterhoff S., Crossed products and the Mackey–Rieffel–Green machine, Oberwolfach Sem. 47 (2017), 5–79, arXiv:1006.4975. [15] Echterhoff S., Nest R., Oyono-Oyono H., Principal non-commutative torus bundles, Proc. Lond. Math. Soc. 99 (2009), 1–31, arXiv:0810.0811. https://doi.org/10.1007/JHEP02(2016)112 https://arxiv.org/abs/1507.00014 https://doi.org/10.4310/AJM.2016.v20.n2.a1 https://arxiv.org/abs/0903.2926 https://doi.org/10.1007/s00220-004-1115-6 https://arxiv.org/abs/hep-th/0306062 https://doi.org/10.4310/ATMP.2009.v13.n5.a6 https://arxiv.org/abs/0810.4374 https://doi.org/10.1016/S0550-3213(99)00009-7 https://doi.org/10.1016/S0550-3213(99)00009-7 https://arxiv.org/abs/hep-th/9810099 https://doi.org/10.1007/s00220-007-0396-y https://arxiv.org/abs/hep-th/0607020 https://doi.org/10.4310/ATMP.2009.v13.n2.a4 https://arxiv.org/abs/0708.2648 https://doi.org/10.1142/S0129055X06002875 https://arxiv.org/abs/math.GT/0501487 https://doi.org/10.1142/S0129055X05002315 https://arxiv.org/abs/math.GT/0405132 https://arxiv.org/abs/1206.5977 https://doi.org/10.1088/1126-6708/2003/09/054 https://arxiv.org/abs/hep-th/0210209 https://doi.org/10.1088/1126-6708/2003/09/054 https://arxiv.org/abs/1006.4975 https://doi.org/10.1112/plms/pdn050 https://arxiv.org/abs/#2 50 P. Aschieri and R.J. Szabo [16] Ellwood I., Hashimoto A., Effective descriptions of branes on non-geometric tori, J. High Energy Phys. 2006 (2006), no. 12, 025, 22 pages, arXiv:hep-th/0607135. [17] Fack T., Skandalis G., Connes’ analogue of the Thom isomorphism for the Kasparov groups, Invent. Math. 64 (1981), 7–14. [18] Fröhlich J., Gawȩdzki K., Conformal field theory and geometry of strings, in Mathematical Quantum The- ory. I. Field Theory and Many-Body Theory (Vancouver, BC, 1993), CRM Proc. Lecture Notes, Vol. 7, Amer. Math. Soc., Providence, RI, 1994, 57–97, arXiv:hep-th/9310187. [19] Grange P., Schäfer-Nameki S., T -duality with H-flux: non-commutativity, T -folds and G × G structure, Nuclear Phys. B 770 (2007), 123–144, arXiv:hep-th/0609084. [20] Green P., The structure of imprimitivity algebras, J. Funct. Anal. 36 (1980), 88–104. [21] Hannabuss K.C., Mathai V., Noncommutative principal torus bundles via parametrised strict deformation quantization, in Superstrings, Geometry, Topology, and C∗-Algebras,Proc. Sympos. Pure Math., Vol. 81, Amer. Math. Soc., Providence, RI, 2010, 133–147, arXiv:0911.1886. [22] Hannabuss K.C., Mathai V., Parametrized strict deformation quantization of C∗-bundles and Hilbert C∗- modules, J. Aust. Math. Soc. 90 (2011), 25–38, arXiv:1007.4696. [23] Havas G., Majewski B.S., Integer matrix diagonalization, J. Symbolic Comput. 24 (1997), 399–408. [24] Hori K., D-branes, T -duality, and index theory, Adv. Theor. Math. Phys. 3 (1999), 281–342, arXiv:hep- th/9902102. [25] Hull C.M., A geometry for non-geometric string backgrounds, J. High Energy Phys. 2005 (2005), no. 10, 065, 30 pages, arXiv:hep-th/0406102. [26] Hull C.M., Reid-Edwards R.A., Flux compactifications of string theory on twisted tori, Fortschr. Phys. 57 (2009), 862–894, arXiv:hep-th/0503114. [27] Hull C.M., Reid-Edwards R.A., Non-geometric backgrounds, doubled geometry and generalised T -duality, J. High Energy Phys. 2009 (2009), no. 9, 014, 79 pages, arXiv:0902.4032. [28] Hull C.M., Szabo R.J., Noncommutative gauge theories on D-branes in non-geometric backgrounds, J. High Energy Phys. 2019 (2019), no. 9, 051, 40 pages, arXiv:1903.04947. [29] Landi G., Lizzi F., Szabo R.J., String geometry and the noncommutative torus, Comm. Math. Phys. 206 (1999), 603–637, arXiv:hep-th/9806099. [30] Lizzi F., Szabo R.J., Duality symmetries and noncommutative geometry of string spacetimes, Comm. Math. Phys. 197 (1998), 667–712, arXiv:hep-th/9707202. [31] Lowe D.A., Nastase H., Ramgoolam S., Massive IIA string theory and matrix theory compactification, Nuclear Phys. B 667 (2003), 55–89, arXiv:hep-th/0303173. [32] Mathai V., Rosenberg J., T -duality for torus bundles with H-fluxes via noncommutative topology, Comm. Math. Phys. 253 (2005), 705–721, arXiv:hep-th/0401168. [33] Mathai V., Rosenberg J., T -duality for torus bundles with H-fluxes via noncommutative topology. II. The high-dimensional case and the T -duality group, Adv. Theor. Math. Phys. 10 (2006), 123–158, arXiv:hep- th/0508084. [34] Milnor J., Curvatures of left invariant metrics on Lie groups, Adv. Math. 21 (1976), 293–329. [35] Mostow G.D., Factor spaces of solvable groups, Ann. of Math. 60 (1954), 1–27. [36] Nikolaus T., Waldorf K., Higher geometry for non-geometric T-duals, Comm. Math. Phys. 374 (2020), 317–366, arXiv:1804.00677. [37] Phillips N.C., Every simple higher-dimensional noncommutative torus is an AT algebra, arXiv:math.OA/0609783. [38] Pioline B., Schwarz A., Morita equivalence and T -duality (or B versus Θ), J. High Energy Phys. 1999 (1999), no. 8, 021, 16 pages, arXiv:hep-th/9908019. [39] Raeburn I., Rosenberg J., Crossed products of continuous-trace C∗-algebras by smooth actions, Trans. Amer. Math. Soc. 305 (1988), 1–45. [40] Raghunathan M.S., Discrete subgroups of Lie groups, Ergebnisse der Mathematik und ihrer Grenzgebiete, Vol. 68, Springer-Verlag, New York – Heidelberg, 1972. [41] Rieffel M.A., Induced representations of C∗-algebras, Adv. Math. 13 (1974), 176–257. [42] Rieffel M.A., Strong Morita equivalence of certain transformation group C∗-algebras, Math. Ann. 222 (1976), 7–22. https://doi.org/10.1088/1126-6708/2006/12/025 https://arxiv.org/abs/hep-th/0607135 https://doi.org/10.1007/BF01393931 https://doi.org/10.1090/crmp/007/03 https://arxiv.org/abs/hep-th/9310187 https://doi.org/10.1016/j.nuclphysb.2007.02.003 https://arxiv.org/abs/hep-th/0609084 https://doi.org/10.1016/0022-1236(80)90108-1 https://doi.org/10.1090/pspum/081/2681761 https://arxiv.org/abs/0911.1886 https://doi.org/10.1017/S1446788711001170 https://arxiv.org/abs/#2 https://doi.org/10.1006/jsco.1996.0141 https://doi.org/10.4310/ATMP.1999.v3.n2.a5 https://arxiv.org/abs/hep-th/9902102 https://arxiv.org/abs/hep-th/9902102 https://doi.org/10.1088/1126-6708/2005/10/065 https://arxiv.org/abs/hep-th/0406102 https://doi.org/10.1002/prop.200900076 https://arxiv.org/abs/hep-th/0503114 https://doi.org/10.1088/1126-6708/2009/09/014 https://arxiv.org/abs/0902.4032 https://doi.org/10.1007/jhep09(2019)051 https://doi.org/10.1007/jhep09(2019)051 https://arxiv.org/abs/1903.04947 https://doi.org/10.1007/s002200050839 https://arxiv.org/abs/hep-th/9806099 https://doi.org/10.1007/s002200050468 https://doi.org/10.1007/s002200050468 https://arxiv.org/abs/hep-th/9707202 https://doi.org/10.1016/S0550-3213(03)00547-9 https://arxiv.org/abs/hep-th/0303173 https://doi.org/10.1007/s00220-004-1159-7 https://doi.org/10.1007/s00220-004-1159-7 https://arxiv.org/abs/hep-th/0401168 https://doi.org/10.4310/ATMP.2006.v10.n1.a5 https://arxiv.org/abs/hep-th/0508084 https://arxiv.org/abs/hep-th/0508084 https://doi.org/10.1016/S0001-8708(76)80002-3 https://doi.org/10.2307/1969700 https://doi.org/10.1007/s00220-019-03496-3 https://arxiv.org/abs/1804.00677 https://arxiv.org/abs/math.OA/0609783 https://doi.org/10.1088/1126-6708/1999/08/021 https://arxiv.org/abs/hep-th/9908019 https://doi.org/10.2307/2001039 https://doi.org/10.2307/2001039 https://doi.org/10.1016/0001-8708(74)90068-1 https://doi.org/10.1007/BF01418238 Topological T-Duality for Twisted Tori 51 [43] Rieffel M.A., C∗-algebras associated with irrational rotations, Pacific J. Math. 93 (1981), 415–429. [44] Rieffel M.A., The cancellation theorem for projective modules over irrational rotation C∗-algebras, Proc. London Math. Soc. 47 (1983), 285–302. [45] Rieffel M.A., Noncommutative tori – a case study of noncommutative differentiable manifolds, in Geometric and Topological Invariants of Elliptic Operators (Brunswick, ME, 1988), Contemp. Math., Vol. 105, Amer. Math. Soc., Providence, RI, 1990, 191–211. [46] Rieffel M.A., Schwarz A., Morita equivalence of multidimensional noncommutative tori, Internat. J. Math. 10 (1999), 289–299, arXiv:math.QA/9803057. [47] Schwarz A., Morita equivalence and duality, Nuclear Phys. B 534 (1998), 720–738, arXiv:hep-th/9805034. [48] Seiberg N., Witten E., String theory and noncommutative geometry, J. High Energy Phys. 1999 (1999), no. 9, 032, 93 pages, arXiv:hep-th/9908142. [49] Steenrod N., The Topology of Fibre Bundles, Princeton Mathematical Series, Vol. 14, Princeton University Press, Princeton, N.J., 1951. [50] Tralle A., Oprea J., Symplectic manifolds with no Kähler structure, Lecture Notes in Mathematics, Vol. 1661, Springer-Verlag, Berlin, 1997. [51] Williams D.P., The structure of crossed products by smooth actions, J. Austral. Math. Soc. Ser. A 47 (1989), 226–235. [52] Williams D.P., Crossed products of C∗-algebras, Mathematical Surveys and Monographs, Vol. 134, Amer. Math. Soc., Providence, RI, 2007. https://doi.org/10.2140/pjm.1981.93.415 https://doi.org/10.1112/plms/s3-47.2.285 https://doi.org/10.1112/plms/s3-47.2.285 https://doi.org/10.1090/conm/105/1047281 https://doi.org/10.1090/conm/105/1047281 https://doi.org/10.1142/S0129167X99000100 https://arxiv.org/abs/math.QA/9803057 https://doi.org/10.1016/S0550-3213(98)00550-1 https://arxiv.org/abs/hep-th/9805034 https://doi.org/10.1088/1126-6708/1999/09/032 https://arxiv.org/abs/hep-th/9908142 https://doi.org/10.1007/BFb0092608 https://doi.org/10.1017/S1446788700031657 https://doi.org/10.1090/surv/134 https://doi.org/10.1090/surv/134 1 Introduction 1.1 Background 1.2 Summary and outline 2 Crossed products and duality 2.1 Dynamical systems and their crossed products 2.2 Semi-direct products and group algebras 2.3 Pontryagin duality and Fourier transform 2.4 Morita equivalence and Green's theorem 3 Topological T-duality and twisted tori 3.1 Twisted tori and their T-duals 3.2 T-duality in the category KK 3.3 Computational tools 3.4 Topological T-duality for the torus 3.5 Topological T-duality for orbifolds 4 Topological T-duality for almost abelian solvmanifolds 4.1 Mostow bundles 4.2 Almost abelian solvmanifolds 4.3 C*-algebra bundles 4.4 Rn-actions on Mostow bundles 4.5 Ry-actions: Circle bundles with H-flux 4.6 Rz-actions: noncommutative torus bundles 5 Three-dimensional solvmanifolds and their T-duals 5.1 Mostow bundles and SL(2,R) conjugacy classes 5.2 Parabolic torus bundles 5.3 Elliptic torus bundles 5.4 Hyperbolic torus bundles References
id nasplib_isofts_kiev_ua-123456789-211176
institution Digital Library of Periodicals of National Academy of Sciences of Ukraine
issn 1815-0659
language English
last_indexed 2026-03-18T10:20:00Z
publishDate 2021
publisher Інститут математики НАН України
record_format dspace
spelling Aschieri, Paolo
Szabo, Richard J.
2025-12-25T13:22:52Z
2021
Topological T-Duality for Twisted Tori. Paolo Aschieri and Richard J. Szabo. SIGMA 17 (2021), 012, 51 pages
1815-0659
2020 Mathematics Subject Classification: 46L55; 81T30; 16D90
arXiv:2006.10048
https://nasplib.isofts.kiev.ua/handle/123456789/211176
https://doi.org/10.3842/SIGMA.2021.012
We apply the *-algebraic formalism of topological T-duality due to Mathai and Rosenberg to a broad class of topological spaces that include the torus bundles appearing in string theory compactifications with duality twists, such as nilmanifolds, as well as many other examples. We develop a simple procedure in this setting for constructing the T-duals starting from a commutative *-algebra with an action of ℝⁿ. We treat the general class of almost abelian solvmanifolds in arbitrary dimension in detail, where we provide necessary and sufficient criteria for the existence of classical T-duals in terms of purely group theoretic data, and compute them explicitly as continuous-trace algebras with non-trivial Dixmier-Douady classes. We prove that any such solvmanifold has a topological T-dual given by a *-algebra bundle of noncommutative tori, which we also compute explicitly. The monodromy of the original torus bundle becomes a Morita equivalence among the fiber algebras, so that these *-algebras rigorously describe the T-folds from non-geometric string theory.
We thank Ryszard Nest and Erik Plauschinn for helpful discussions. We thank the anonymous referees for their detailed suggestions. This research was supported by funds from Università del Piemonte Orientale (UPO). P.A. acknowledges partial support from INFN, CSN4, and Iniziativa Speci ca GSS. P.A. is affiliated with INdAM-GNFM. R.J.S. acknowledges a Visiting Professorship through UPO Internationalization Funds. R.J.S. also acknowledges the ArnoldRegge Centre for the visit and INFN. The work of R.J.S. was supported in part by the Consolidated Grant ST/P000363/1 from the UK Science and Technology Facilities Council.
en
Інститут математики НАН України
Symmetry, Integrability and Geometry: Methods and Applications
Topological T-Duality for Twisted Tori
Article
published earlier
spellingShingle Topological T-Duality for Twisted Tori
Aschieri, Paolo
Szabo, Richard J.
title Topological T-Duality for Twisted Tori
title_full Topological T-Duality for Twisted Tori
title_fullStr Topological T-Duality for Twisted Tori
title_full_unstemmed Topological T-Duality for Twisted Tori
title_short Topological T-Duality for Twisted Tori
title_sort topological t-duality for twisted tori
url https://nasplib.isofts.kiev.ua/handle/123456789/211176
work_keys_str_mv AT aschieripaolo topologicaltdualityfortwistedtori
AT szaborichardj topologicaltdualityfortwistedtori