[Paper Review] On the quantum sl_2 invariants of knots and integral homology spheres

Kazuo Habiro|ArXiv.org|Nov 4, 2002

Geometric and Algebraic Topology9 references17 citations

TL;DR

This paper introduces a new invariant $ I(M) $ for integral homology spheres with values in the completed Laurent polynomial ring $ \widehat{\mathbb{Z}[q]} $, which specializes at roots of unity to the Witten-Reshetikhin-Turaev (WRT) invariants. The construction provides a new, independent definition of the WRT invariant via quantum $ \mathfrak{sl}_2 $ invariants, unifying and generalizing previous invariants like the Ohtsuki series and colored Jones polynomials.

ABSTRACT

We will announce some results on the values of quantum sl_2 invariants of knots and integral homology spheres. Lawrence's universal sl_2 invariant of knots takes values in a fairly small subalgebra of the center of the h-adic version of the quantized enveloping algebra of sl_2. This implies an integrality result on the colored Jones polynomials of a knot. We define an invariant of integral homology spheres with values in a completion of the Laurent polynomial ring of one variable over the integers which specializes at roots of unity to the Witten-Reshetikhin-Turaev invariants. The definition of our invariant provides a new definition of Witten-Reshetikhin-Turaev invariant of integral homology spheres.

Motivation & Objective

• To define a new invariant $ I(M) $ for integral homology spheres using quantum $ \mathfrak{sl}_2 $ invariants.
• To provide a new, independent definition of the Witten-Reshetikhin-Turaev (WRT) invariant without relying on existing constructions.
• To unify the colored Jones polynomial and the Ohtsuki series under a single algebraic framework.
• To establish integrality and $ p $-adic convergence properties of quantum invariants at roots of unity.
• To prove that $ I(M) $ captures the full information of WRT invariants across all roots of unity.

Proposed method

• Define a universal $ \mathfrak{sl}_2 $ invariant of tangles in the $ h $-adic completion of the quantized enveloping algebra $ U_h(\mathfrak{sl}_2) $.
• Construct an invariant $ I(L) $ for algebraically split framed links $ L $ with $ \pm 1 $ framings in $ S^3 $, valued in $ \widehat{\mathbb{Z}[q]} $.
• Use the $ h $-adic completion of $ \mathbb{Z}[q] $ to define $ I(M) $ for integral homology spheres $ M $ via surgery presentations.
• Prove invariance of $ I(L) $ under Hoste moves, establishing that $ I(M) $ is well-defined for integral homology spheres.
• Specialize $ I(M) $ at roots of unity $ \zeta $ to recover the WRT invariant $ \tau_\zeta(M) $, showing $ I(M)|_{q=\zeta} = \tau_\zeta(M) $.
• Use the injectivity of the map $ \iota_1: \widehat{\mathbb{Z}[q]} \to \mathbb{Z}[[q-1]] $ to show $ I(M) $ is equivalent to the Ohtsuki series $ \tau(M) $.

Experimental results

Research questions

• RQ1 Can a single invariant unify the WRT invariants at all roots of unity for integral homology spheres?
• RQ2 Is there a new, independent definition of the WRT invariant that does not rely on prior constructions?
• RQ3 How does the quantum $ \mathfrak{sl}_2 $ universal invariant relate to the colored Jones polynomial and integrality?
• RQ4 Can the Ohtsuki series be reconstructed from a completion of $ \mathbb{Z}[q] $, and what is its $ p $-adic behavior?
• RQ5 What is the relationship between $ I(M) $, the WRT invariants, and the Ohtsuki series?

Key findings

• The invariant $ I(M) $ is well-defined for all integral homology spheres $ M $, with values in the completed Laurent polynomial ring $ \widehat{\mathbb{Z}[q]} $.
• Specializing $ I(M) $ at roots of unity $ \zeta $ yields the WRT invariant $ \tau_\zeta(M) $, providing a new definition of the WRT invariant independent of prior constructions.
• The invariant $ I(M) $ is as strong as the full set of WRT invariants across all roots of unity, and also as strong as the Ohtsuki series.
• The Ohtsuki series $ \tau(M) $ is recovered as $ \iota_1(I(M)) $, where $ \iota_1 $ is the natural inclusion into $ \mathbb{Z}[[q-1]] $, and $ \iota_1 $ is injective.
• For $ \zeta $ a primitive $ 2^m $-th root of unity, $ \tau_\zeta(M) \in \mathbb{Z}_2[\zeta] $, confirming Lawrence’s $ p $-adic convergence conjecture for $ p=2 $.
• The conditions $ I(M) = I(M') $, $ \tau(M) = \tau(M') $, $ \tau_\zeta(M) = \tau_\zeta(M') $ for all roots of unity $ \zeta $, and for infinitely many $ \zeta $ of prime power order are all equivalent.