A proof of Perrin-Riou's Heegner point main conjecture (Q2239238)

Let \(E\) be an elliptic curve over \(\mathbb Q\) of conductor \(N\) and let \(p>3\) be a prime where \(E\) has good ordinary reduction. Let \(K\ne \mathbb Q(\sqrt{-3})\) be an imaginary quadratic field of odd discriminant relatively prime to \(Np\). Assume \(N=N^+ N^-\), where \(N^+\) is a product of primes that split in \(K\) and \(N^-\) is the squarefree product of an even number of primes, each of which is inert in \(K\). Let \(\overline{\rho}\) be the representation of \(\text{Gal}(\overline{\mathbb Q}/\mathbb Q)\) on the \(p\)-torsion of \(E\). Assume \(\overline{\rho}\) is ramified at all primes dividing \(N^+\) to exactly the first power, and at all primes dividing \(N^-\) that are congruent to \(\pm 1\pmod p\). Moreover, if \(N\) is not squarefree, assume that either \(\overline{\rho}\) is ramified at a prime dividing \(N^-\) or that there are at least two primes dividing \(N^+\) to exactly the first power. Finally, assume \(\overline{\rho}\) is surjective and \(p\) is not anomalous. The first main theorem of the paper is that under the above assumptions, the Heegner point main conjecture holds. This says the following. Let \(K_n\) be the \(n\)th level of the anticyclotomic \(\mathbb Z_p\)-extension of \(K\). Let \[X_{\infty}=\text{Hom}_{\mathbb Z_p}(\lim_{\rightarrow} \text{Sel}_{p^{\infty}}(E/K_n), \mathbb Q_p/\mathbb Z_p), \qquad S_{\infty} = \lim_{\leftarrow} S_p(E/K_n), \] where \(S_{\infty}\) and \(S_p\) are certain Selmer groups. Heegner points yield a compatible system of classes \(\kappa_{\infty} \in S_{\infty}\). The now-proved conjecture states that both \(S_{\infty}\) and \(X_{\infty}\) have rank one over the Iwasawa algebra \(\Lambda\), there is a finitely generated torsion \(\Lambda\)-module \(M_{\infty}\) such that \(X_{\infty}\) is pseudoisomorphic to \(\Lambda\oplus M_{\infty}\oplus M_{\infty}\), and the characteristic ideal of \(M_{\infty}\) satisfies \[ \text{Char}_{\Lambda}(M_{\infty}) = \text{Char}_{\Lambda}(M_{\infty})^{\iota} = \text{Char}_{\Lambda}(S_{\infty}/\Lambda\kappa_{\infty}). \] Here \(\iota\) is the involution of \(\Lambda\) given by \(\gamma\mapsto \gamma^{-1}\) for \(\gamma\in\text{Gal}(K_{\infty}/K)\). For the second main theorem of the paper, assume in addition that \(p\) splits as \(\mathfrak p\overline{\mathfrak p}\) in \(K\). \textit{M. Bertolini} et al. [Duke Math. J. 162, No. 6, 1033--1148 (2013; Zbl 1302.11043)] and \textit{E. H. Brooks} [Int. Math. Res. Not. 2015, No. 12, 4177--4241 (2015; Zbl 1378.11069)] constructed a \(p\)-adic \(L\)-function \(\mathcal{L}_{\mathfrak p}^{BDP}\). The Iwasawa-Greenberg main conjecture for \(\mathcal{L}^{BDP}\) asserts that the Pontryagin dual of a certain Selmer group is \(\Lambda\)-torsion with characteristic ideal equal to \((\mathcal{L}_{\mathfrak p}^{BDP})^2\), after an unramified extension of scalars of \(\Lambda\). The present paper deduces this conjecture from the previous theorem. The proofs use the theory of bipartite Euler systems of \textit{B. Howard} [J. Reine Angew. Math. 597, 1--25 (2006; Zbl 1127.11072)] and the work of \textit{W. Zhang} [Camb. J. Math. 2, No. 2, 191--253 (2014; Zbl 1390.11091)] on Kolyvagin's conjecture.