Noncommutative Ricci curvature and Dirac operator on \(\mathbb C_q[{\text SL}_2]\) at roots of unity (Q1870305)

In [\textit{S. Majid}, Commun. Math. Phys. 225, No. 1, 131-170 (2002; Zbl 0999.58004)] hereafter refered to as [1], the author developed a functional theory of noncommutative Riemannian geometry at an algebraic level. As a sequel of this research, a unique torsion free Riemannian spin connection for the natural right coaction invariant Killing metric on the quantum group \({\mathcal A}= \mathbb{C}_q[\text{SL}_2]\) is constructed. The outline of the paper is as follows: Let \(\Lambda\) be the invariant exterior algebra over \({\mathcal A}\) generated by \(e_a\), etc., where \(a\), etc., are the generators of \({\mathcal A}\). The space of left-invariant forms is denoted by \(\Omega= {\mathcal A}\otimes\Lambda\). The noncommutative differential calculus for these data, e.g. exterior differential \(d= \mu^{-1}[\theta, ]\), \(\theta= e_a+ e_d\), and so on, are reviewed in Sec. 2. In Sec. 3, according to [1], natural right coaction invariant Killing metric \(\eta\) is introduced and torsion and cotorsion for this metric are defined. Then existence of unique torsion- and cotorsion-free connection for this metric if \(q\) is generic or an odd root of unity, is proved (Th. 3.1). It is noted that the connection is not regular in general. \(\theta\) is covariant constant. \(\eta\) is not covariant constant in general but satisfy \[ (\nabla\wedge\text{id}-\text{id}\wedge\nabla)\eta=0. \] The author says naive metric compatibility \(\nabla\eta= 0\) is too strong for noncommutative geometry. Then Riemann and Ricci curvatures are calculated (Prop. 3.2, Prop. 4.2). If the Hodge \(*\) operator is a multiple of the identity on the four-dimensional space of 1-forms, \(\eta\) is a \(q\)-Minkowski space metric [\textit{S. Majid}, `Foundation of quantum group theory', Cambridge (1995; Zbl 0857.17009)], and it is shown to be an Einstein space up to a shift by \(\theta\otimes\theta\) (Sec. 4). In Sec. 5, the Dirac operator for the canonical spin connection is computed (Prop. 5.1). Its eigenvalues are computed for \(q\) an \(r\)th root of unity, \(r= 3,5,7\) (Prop. 5.2). Then the Dirac operator is fully solved for \(r= 3\) (Prop. 5.3). This paper does not give physical applications or meanings. But the author says this is a follow-up to \textit{X. Gomez} and \textit{S. Majid} [Lett. Math. Phys. 60, No. 3, 221-237 (2002; Zbl 1004.58003)], where the electromagnetic theory on \(\mathbb{C}_q[\text{SL}_2]\) was covered.