Extracting a prime pair from a quantitative von Mangoldt lower bound
ProvedGoldbach.vonMangoldt_convolution_extract_prime_pairfinite-sumsgoldbachnumber-theoryprime-powers
Let be the von Mangoldt function and let be a natural number. Put and , with at zero. If
then there are primes such that .
The strict quantitative hypothesis excludes contributions that arise solely from proper prime powers. This implication is useful for a binary circle-method or correlation argument. The hypothesis is left explicit: establishing it uniformly for all sufficiently large even would require additional analytic information not supplied here. No parity assumption is needed for this extraction implication itself.
Formalization note: the endpoint convention includes and , with natural subtraction. Nat.sqrt and Nat.log 2 supply the integer parameters.
Preamble
import Mathlib.NumberTheory.ArithmeticFunction.VonMangoldt import Mathlib.Data.Nat.Sqrt import Mathlib.Algebra.BigOperators.Intervals open scoped BigOperators set_option autoImplicit false
Formal statement
theorem Goldbach.vonMangoldt_convolution_extract_prime_pair (N : ℕ)
(hlarge : 2 * (Nat.sqrt N * Nat.log 2 N : ℕ) * (Real.log N)^2 <
∑ m ∈ Finset.range (N+1), ArithmeticFunction.vonMangoldt m * ArithmeticFunction.vonMangoldt (N-m)) :
∃ p q : ℕ, Nat.Prime p ∧ Nat.Prime q ∧ N = p+q := by sorrySource
Elementary finite-support estimate developed for the Goldbach mission, https://prove2.me/missions/The_Goldbach_Conjecture. Uses Mathlib ArithmeticFunction.vonMangoldt and Nat.sqrt/Nat.log APIs; no claim of a new prime-distribution estimate or literature novelty.