Big Heegner Points in Iwasawa Theory

• Big Heegner points are p-adic families of Heegner-type cohomology classes attached to Hida families that interpolate classical Heegner points and generalized cycles.
• They satisfy precise Euler-system norm relations and reciprocity laws, linking their specializations to p-adic L-functions and controlling Selmer groups.
• Their construction on modular, quaternionic, and totally real curves advances Iwasawa theory by enabling effective Kolyvagin descent and exceptional zero analyses.

Big Heegner points are $p$-adic families of Heegner-type cohomology classes attached to the ordinary deformation theory of modular forms. In the foundational formulation, one starts from a Hida family, its rank-two big ordinary Galois representation, and the critical self-dual twist $T^\dagger$; one then constructs a single Iwasawa-theoretic class whose arithmetic specializations recover classical Heegner point Kummer classes in weight $2$ and, in higher weights, generalized Heegner cycle classes up to explicit Euler factors. Their defining features are interpolation in weight, norm compatibility in anticyclotomic towers, and reciprocity laws relating their localizations to $p$-adic $L$-functions (Howard, 2012, Castella, 2014).

1. Foundational construction in Hida families

Howard’s construction begins with a $p$-ordinary, $p$-stabilized modular form and the corresponding local branch $R$ of Hida’s ordinary Hecke algebra, finite and flat over the Iwasawa algebra. Hida theory supplies a free rank-two $R$-module $T$ with continuous $T^\dagger$0-action, together with the critical character $T^\dagger$1, and the basic coefficient object is the critical twist

$T^\dagger$2

which is self-dual in the appropriate sense. For arithmetic primes $T^\dagger$3, specialization yields ordinary modular forms $T^\dagger$4 and Galois representations

$T^\dagger$5

This is the representation-theoretic framework in which big Heegner points are defined (Howard, 2012).

The geometric input is a tower of modular curves $T^\dagger$6 with $T^\dagger$7-level structure and CM points $T^\dagger$8 attached to orders $T^\dagger$9. After ordinary projection, twisting by the critical character, and applying a twisted Kummer map, Howard constructs classes $2$0 and then passes to the inverse limit in the $2$1-power level to obtain the big Heegner point of conductor $2$2,

$2$3

In the anticyclotomic Iwasawa-theoretic notation used later, compatible systems are written as

$2$4

with $2$5 denoting the conductor-$2$6 system (Howard, 2012, Castella, 2014).

The interpolation property is central. In weight $2$7, the specialization of the big class recovers the classical Heegner point Kummer class on the corresponding modular abelian variety. In higher even weights, the specializations are compared with étale Abel–Jacobi images of generalized Heegner cycles on Kuga–Sato varieties, again via a precise specialization map in the Hida family. This gives the basic dictionary between a single “big” class and the classical Heegner objects attached to individual arithmetic points (Howard, 2012, Castella, 2014).

2. Euler-system structure, Selmer conditions, and Kolyvagin descent

Big Heegner points are not merely interpolating classes; they form an Euler system. Howard’s classes satisfy the expected norm relations

$2$8

and, for inert primes $2$9,

$p$0

They lie in strict Greenberg Selmer groups, with local conditions unramified away from $p$1 and Greenberg-ordinary at $p$2. In the anticyclotomic tower, Howard’s system $p$3 also satisfies a functional equation under complex conjugation of the form $p$4 with $p$5 (Howard, 2012, Castella, 2015).

A basic arithmetic consequence is nontriviality upon specialization. Howard proves that for any arithmetic prime, the specialization of the big Heegner point is nonzero for sufficiently deep ring class fields, extending results of Cornut–Vatsal from weight two and trivial character to all ordinary modular forms in the Hida family. This nontriviality is the first step toward the expected rank behavior of Selmer groups in anticyclotomic towers and across Hida families (Howard, 2012).

Kolyvagin descent for big Heegner points was developed by Büyükboduk. The main input beyond Howard’s Euler system is a family-level control of Tamagawa factors at bad primes. The paper constructs a Tamagawa element $p$6 interpolating the specialized Tamagawa factors and, under the stated hypotheses, proves that these factors are prime to $p$7 for every specialization. With this local control in place, the big Heegner point Euler system yields a big Kolyvagin system

$p$8

whose initial class is the anticyclotomic Heegner class $p$9. The resulting machinery gives the standard divisibility for characteristic ideals of big Selmer groups and removes the extra correction factors that appeared in specialization-by-specialization arguments (Buyukboduk, 2013).

3. Explicit reciprocity and higher-weight specialization

A decisive advance was the construction of a two-variable anticyclotomic $L$0-adic $L$1-function

$L$2

attached to the Hida family $L$3 and the imaginary quadratic field $L$4. This $L$5-adic $L$6-function interpolates the anticyclotomic $L$7-adic Rankin $L$8-series of Bertolini–Darmon–Prasanna and Casazza–Hsieh as the weight varies. On the cohomological side, a two-variable Perrin–Riou regulator $L$9 is constructed using Ochiai’s theory for nearly ordinary deformations, the two-variable formalism of Loeffler–Zerbes, and Yager modules for the unramified tower (Castella, 2014).

The central reciprocity law identifies these two constructions: $p$0 In other words, the image of the big Heegner point under the Perrin–Riou logarithm is exactly the two-variable $p$1-adic $p$2-function, up to the explicit twist $p$3. The paper presents this as the Hida-family and anticyclotomic analogue of a $p$4-adic Gross–Zagier formula: the cohomology class generated by Heegner points is measured by a $p$5-adic $p$6-function (Castella, 2014).

This reciprocity law is then combined with classical formulas for generalized Heegner cycles to derive higher-weight specialization results. Under the stated irreducibility, ramification, and $p$7-splitting hypotheses, the specialization of Howard’s big Heegner point at an arithmetic prime of weight $p$8 matches the $p$9-adic class built from generalized Heegner cycles in the strict Greenberg Selmer group, and at finite level the specialized class recovers the étale Abel–Jacobi image of the generalized Heegner cycle up to an explicit Euler factor. The phrase “big Heegner points specialize to classical Heegner cycle classes” is thus literal, not heuristic (Castella, 2014).

4. Exceptional zero phenomena and derivative formulas

One of the most delicate aspects of big Heegner points appears at exceptional specializations. Castella’s paper on exceptional specializations treats a weight-$p$0 newform

$p$1

with $p$2, $p$3, ordinary at $p$4, and split multiplicative reduction at $p$5, so that

$p$6

The modular form lies on a Hida family $p$7, and Howard’s big Heegner points provide a system

$p$8

At weights $p$9, the localization of the specialization is related to generalized Heegner cycles by an explicit factor, but at weight $R$0 that factor vanishes precisely because $R$1 (Castella, 2015).

The resulting phenomenon is an exceptional zero. The naive specialization forces the localized class at $R$2 to vanish, even though the classical Heegner point class $R$3 is expected to be nonzero when $R$4. The remedy is to take an Iwasawa-theoretic derivative. If $R$5 satisfies $R$6, then for a topological generator $R$7 one writes

$R$8

and the normalized specialization

$R$9

is independent of $R$0. This is the precise derivative at the trivial character used in the theorem (Castella, 2015).

The main exceptional-zero formula is

$R$1

where $R$2 is the classical Heegner point class and

$R$3

The proof extends the $R$4-adic Gross–Zagier formula of Bertolini–Darmon–Prasanna to the semistable non-crystalline setting, which is necessary because the representation $R$5 is semistable but non-crystalline at $R$6 in the split multiplicative case. Conceptually, the theorem is presented as the exact analogue, for big Heegner points, of a Mazur–Tate–Teitelbaum exceptional zero phenomenon (Castella, 2015).

5. Quaternionic, Shimura-curve, and totally real generalizations

The theory has been extended from modular curves to quaternionic Shimura curves. In the indefinite quaternionic setting, one starts with a quaternion algebra $R$7, Shimura curves $R$8 of level $R$9, and a primitive ordinary Hida family. Compatible Heegner points on the Jacobians $T$0 yield cohomology classes

$T$1

and after anticyclotomic corestriction one obtains

$T$2

This is explicitly described as the quaternionic analogue of Howard’s big Heegner points and of Castella’s interpolation and specialization results (Longo et al., 5 Oct 2025).

In this quaternionic setting, the algebraic anticyclotomic $T$3-adic $T$4-function is defined by applying a big Perrin–Riou regulator to the localization of the big Heegner class, while an analytic anticyclotomic $T$5-adic $T$6-function is constructed from Serre–Tate expansions and quaternionic modular forms. The main reciprocity statement identifies them up to an explicit factor: $T$7 This equality implies, in particular, non-$T$8-torsion of the big Heegner class under a generic root-number $T$9 hypothesis and supplies the bridge to higher-weight specialization results (Longo et al., 5 Oct 2025).

A parallel development appears in the paper on big Heegner points, generalized Heegner classes, and $T^\dagger$00-adic $T^\dagger$01-functions in the quaternionic setting. There the geometric anticyclotomic $T^\dagger$02-adic $T^\dagger$03-function obtained from the big Perrin–Riou map is identified with an analytically defined family of quaternionic BDP-type $T^\dagger$04-adic $T^\dagger$05-functions, and higher-weight specializations of the big Heegner point are compared with generalized Heegner classes. On sufficiently small affinoids in weight space, the family of big Heegner points and the independently constructed family of generalized Heegner classes are shown to coincide (Longo et al., 2024).

The framework also extends beyond $T^\dagger$06. Over totally real fields, one can work simultaneously with totally definite and indefinite quaternion algebras. In the definite case, compatible Heegner points on Gross curves produce theta elements and a two-variable $T^\dagger$07-adic $T^\dagger$08-function; in the indefinite case, the same construction, followed by a twisted Kummer map, yields big Heegner classes

$T^\dagger$09

This is presented as a totally real quaternionic analogue of the Longo–Vigni and Howard picture (Jiménez, 30 Oct 2025). A related but analytically oriented formulation over totally real fields constructs anticyclotomic rigid-analytic families of Heegner classes

$T^\dagger$10

whose specializations are the classical Heegner point Kummer classes, and proves anticyclotomic Gross–Zagier and Waldspurger formulas on Shimura curves (Disegni, 2015).

6. Terminological scope and related Heegner-type families

In the strict sense used by Howard and its later Hida-family descendants, “big Heegner points” are Iwasawa cohomology classes attached to a big Galois representation and equipped with Euler-system norm relations. Some nearby literatures use the phrase more loosely, and the distinction matters. For example, the paper on congruent numbers treats a family of Heegner points indexed by square-free integers with many prime factors and calls this the “big Heegner point aspect,” but the construction there is not a Hida-family class in a big Galois representation; it is a systematic generalization of classical CM-point constructions on $T^\dagger$11 and their Gross–Zagier–Kolyvagin applications (Tian, 2012).

A second neighboring line concerns higher Heegner points varying in the conductor rather than in Hida weight. In the strengthening of Mazur’s conjecture, higher Heegner points

$T^\dagger$12

on modular and Shimura curves are studied through the comparison of Galois orbits with Hecke orbits. The resulting vertical and horizontal non-torsion theorems are quantitative and concern eventual non-torsion of traces for large powers $T^\dagger$13 and sufficiently large primes $T^\dagger$14. This belongs to the broader Heegner-family philosophy, but it is not Howard’s big Heegner point construction (Zhang, 6 Feb 2026).

A third neighboring theory is the Stark–Heegner setting over real quadratic fields. There the objects are initially local points at a prime $T^\dagger$15 inert in the real quadratic field, and weighted combinations

$T^\dagger$16

are conjectured to be global and to satisfy

$T^\dagger$17

The paper “Stark-Heegner points and diagonal classes” places these points in the same broad family-valued philosophy as big Heegner points by showing that the relevant combinations arise from global Selmer classes and are governed by the derivative of a weight-variable $T^\dagger$18-adic $T^\dagger$19-function through

$T^\dagger$20

This suggests a conceptual unification at the level of $T^\dagger$21-adic families of arithmetic classes, even though the underlying geometry is different from the classical imaginary quadratic CM construction (Darmon et al., 2022).

In the narrow technical sense, then, big Heegner points are Hida-family Euler-system classes interpolating Heegner points and generalized Heegner classes, controlled by Perrin–Riou regulators and $T^\dagger$22-adic $T^\dagger$23-functions. In a broader arithmetic sense, the term sits inside a larger landscape of Heegner-type families—quaternionic, totally real, higher-conductor, and Stark–Heegner—whose common theme is the packaging of Heegner phenomena into $T^\dagger$24-adic variation and Iwasawa-theoretic control (Howard, 2012, Longo et al., 2024).