Endomorphisms of the projective plane and the image of the Suslin-Hurewicz map (Q6040623)

This paper has two main theorems. The first main theorem is a calculation of the endomorhpism ring \([\mathbb{P}^2, \mathbb{P}^2]\) in the motivic stable homotopy category over a field of characteristic different from \(2\) or \(3\). The underlying group of this ring is isomorphic to \(\mathbb{Z} \oplus \mathbb{Z} \oplus F^\times/(F^\times)^6\), and the multiplication is determined. The second is a proof, again in the case of a field of characteristic different from \(2\) and \(3\), of the \(n=4\) case of a conjecture attributed to \textit{A. A. Suslin} [Lect. Notes Math. 1046, 357--375 (1984; Zbl 0528.18007)]: If \(A\) is an essentially smooth local \(F\)-algebra, then the image of a homomorphism \(K_n(A) \to K_n^M(A)\) consists precisely of the multiples of \((n-1)!\). This conjecture is trivial when \(n=1\), amounts to Matsumoto's theorem when \(n=2\) and follows from the Merkurjev-Serre form of the norm-residue isomorphism theorem when \(n=3\). In [\textit{A. Asok} et al., Invent. Math. 219, No. 1, 39--73 (2020; Zbl 1444.19004)]], it was shown that unstable motivic homotopy could establish the result when \(n=5\), skipping the case \(n=4\). That case is settled here. The main ideas involved in the proof of the first theorem are these: the object \(\mathbb{P}^2\) is the cofibre of the motivic Hopf map \(\eta\), which means that the ring \([\mathbb{P}^2, \mathbb{P}^2]\) admits a description, at least up to untangling exact sequences, in terms of various groups derived from the algebra of the endomorphisms of the sphere spectrum \(\1\) and the element \(\eta\). The various algebraic problems that arise can be solved using techniques from [\textit{O. Röndigs}, Contemp. Math. 745, 199--215 (2020; Zbl 1442.14082)] and [\textit{O. Röndigs} et al., Ann. Math. (2) 189, No. 1, 1--74 (2019; Zbl 1406.14018)]. In particular, comparison with the very-effective cover of the hermitian \(K\)-theory spectrum plays a substantial role. Along the way, in Lemma 3.4, a detailed description of the stable motivic homotopy sheaf \(\pi_2(\mathbb{P}^2)\) is obtained: a short exact sequence is given and the extension problem is resolved. For the second main theorem, using [Asok et al., loc. cit.], one can restate Suslin's conjecture in degree \(4\) as the assertion that the map \(\pi^{\mathbb{A}^1}_3(\mathrm{SL}_4) \to \pi^{\mathbb{A}^1}_3(\mathbb{A}^4 \setminus \{0\})\) is contained in the subsheaf of \(K^{MW}_4\) generated by \(6\) and the image of \(\eta\). By explicit geometric considerations in this paper, the image is shown to be contained in the kernel of a map \(\pi^{\mathbb{A}^1}_2({\mathbb{G}_m} \wedge \mathbb{P}^2) \to \pi^{\mathbb{A}^1}_2(\mathrm{SL}_3)\), which in turn is contained in the kernel of a connecting map \(\pi^{\mathbb{A}^1}_{2-(1)} (\mathbb{A}^4 \setminus \{0\} )\to \pi^{\mathbb{A}^1}_{2-(1)} (\mathbb{P}^2)\) -- here the \(-(1)\) denotes \(1\)-fold contraction. The target group has been well studied in the first part of the paper, and the connecting map is calculated essentially by étale or complex realization, which appears in this paper as Remark 3.5. There are two appendices that contain useful results about modules over \(K^{MW}\) and about motivic Toda brackets respectively.