Smooth geometry of the noncommutative pillow, cones and lens spaces (Q2013490)

An affine algebra \(A\) with unit \(1\) and having Gelfand-Kirillov dimension \(n\) is called differentiably smooth if it can be endowed with an \(n\)-dimensional connected integrable differential calculus. The paper under review is concerned with interesting examples of such differentiably smooth algebras (namely those which can be viewed as noncommutative versions of orbifolds / singular spaces). A differential calculus \(\Omega A\) is a differential graded algebra \((\Omega, d)\) such that \(\Omega^0 = A\), \(\Omega^n = A dA \wedge dA \wedge \cdots \wedge dA\) and by the Leibniz rule \(\Omega^n A = dA \wedge dA \wedge \cdots \wedge dA A\) (with \(dA\) appearing \(n\)-times). The calculus is called connected if \(\ker d_{|A} = \mathbb{C} \cdot 1\). It is called integrable of dimension \(n\) if there is a complex of integral forms \((JA, \nabla)\), where \(J_n A\) is the space of all right \(A\)-linear maps \(\Omega^n A \to A\) which is endowed with the natural \(A\)-bimodule structure as the dual of the \(A\)-bimodule \(\Omega^n A\). Here \(\nabla_n\) denotes the divergence \(J_{n+1} A \to J_n A\) which fulfills the corresponding Leibniz rule. Additionally, the complex \((JA, \nabla)\) fulfills the following condition: There is an algebra automorphism \(\nu\) of \(A\) and \(A\)-bimodule isomorphisms \(\Theta : \Omega^k A \to ^\nu J_{n-k} A\), \(k = 0, \cdots, n\) (taking values in the \(A\)-bimodule twisted by the automorphism \(\nu\)) such that \(\nabla_{n-k} \circ \Theta_{k-1} = \Theta_{k} \circ d\) for \(k = 1, \cdots, n\), where \(^\nu J_0 A = ^\nu A\). The following three examples are studied: 1) \(\mathcal{O}(\mathbb{T}_{\theta}^2)_{+} =: \mathcal{O}(\mathbb{P}_{\theta})\), the non-commutative pillow, being the fixed-point subalgebra of the coordinate algebra of the non-commutative torus \(\mathcal{O}(\mathbb{T}_{\theta}^2) = \mathcal{O}(\mathbb{T}_{\theta}^2)_{+} \oplus \mathcal{O}(\mathbb{T}_{\theta}^2)_{-}\); where the grading is induced by the automorphism \(\sigma\) acting by \(V \mapsto V^{\ast}\), \(W \mapsto W^{\ast}\) on the unitary generators \(V\) and \(W\). The non-commutative pillow has Gelfand-Kirillov dimension \(2\). 2) \(\mathcal{O}(C_{q, \kappa}^{N})\) for a positive integer \(N\) is the complex \(\ast\)-algebra generated by self-dual \(a\), \(b\) and \(b^{\ast}\) subject to the relations \[ ab = q^N ba + \kappa \frac{1 - q^N}{1 - q}, \;bb^{\ast} = \prod_{l=0}^{N-1} \left(q^{-l} a + \kappa \frac{1 - q^{-l}}{1 - q}\right), \] \[ b^{\ast} b = \prod_{l=1}^{N} \left(q^l a + \kappa \frac{1 - q^l}{1 - q} \right), \] where \(q > 0\) and \(\kappa \in \mathbb{R}\) are parameters. The Gelfand-Kirillov dimension in this case is \(2\). 3) \(\mathcal{O}(L_q(N; 1, N))\) is the \(\ast\)-algebra generated by \(\xi, \zeta\) subject to the relations \[ \xi \zeta = q^l \zeta \xi, \;\xi \zeta^{\ast} = q^l \zeta^{\ast} \xi, \;\zeta \zeta^{\ast} = \zeta^{\ast} \zeta \] \[ \xi \xi^{\ast} = \prod_{m=0}^{l-1} (1 - q^{2m} \zeta \zeta^{\ast}), \;\xi^{\ast} \xi = \prod_{m=1}^l (1 - q^{-2m} \zeta \zeta^{\ast}) \] where \(q \in (0,1)\). The Gelfand-Kirillov dimension of this algebra is \(3\). The limiting case \(q = 1\) is the coordinate algebra of the singular lense space. The main result of the paper under review reads as follows: The algebras of examples 1) through 3) admit connected integrable differential calculi of dimension \(2\) (in cases 1) and 2)) and dimension \(3\). In the classical case there is a known criterion with which it is possible to distinguish between manifolds and (good) orbifolds via an orientability condition posed on a spectral triple. Namely, if there is a Hochschild cycle whose image in the differential calculus induced by the Dirac operator is the chirality operator, we are in the case of manifolds and not good orbifolds. In the paper under review the authors consider an analog of this criterion for differentiably smooth algebras. For the case of the non-commutative pillow and the Moyal cone (i.e. Moyal deformation of the unit disc) they construct suitable spectral triples and prove that there is a cycle that maps to the chirality operator, which is however not a Hochschild cycle. Therefore these cases can be viewed as non-commutative (deformations) of manifolds rather than orbifolds.