Unit fractions with semiprime denominators: an elementary proof of Erdős Problem #306
Shisheng Li
math.NT
Sep 26, 2026 · v1
math.CO
TL;DR
The elementary proof of Erdős Problem #306 is formalised in Lean 4 with Mathlib, apart from a cited Ramanujan inequality assumed as hypothesis.
Abstract
We give an elementary proof that every positive rational number $a/b$ with $b$ squarefree is a finite sum of distinct unit fractions $1/n$, where each $n$ is a product of two distinct primes (Erdős Problem #306). After a reduction to small targets, we take a single complete bipartite graph between the primes in $(y^2,2y^2]$, together with $2$ and the primes of $b$, and a tuned initial segment of the primes in $(y^8,y^9]$, and show that some subgraph has reciprocal sum congruent to $a/b$ modulo $1$; the small total mass then forces equality. Writing the number of such subgraphs as a finite Fourier sum, we sort the frequencies into three cases using a table indexed by the two sides of the graph. The small integer frequencies give a positive main term, and all other frequencies are negligible by a divisor-counting argument and a no-wrap-around form of the Chinese remainder theorem. The only inputs about primes are Chebyshev-type bounds. The circle-method framework comes from Tang's Lean development, which gave the first proof; our construction removes its anchor-synchronisation step. The proof has been formalised in Lean 4, apart from a cited inequality of Ramanujan. This work is a human-AI collaboration: AI tools contributed substantially to the construction, the experiments and the writing.
Problem
Erdős Problem #306 asks whether every positive rational a/b with b squarefree can be written as a finite sum of distinct unit fractions 1/n where each n is a product of two distinct primes.
Approach
After reducing to small targets, a single complete bipartite graph is taken between primes in (y^2,2y^2] together with 2 and the primes of b, and a tuned initial segment of primes in (y^8,y^9]. Some subgraph is shown to have reciprocal sum congruent to a/b modulo 1, and the small total mass forces equality. The count of such subgraphs is written as a finite Fourier sum and sorted into three cases, using only Chebyshev-type bounds. The proof was formalised in Lean 4 with Mathlib, one declaration per numbered statement.
Results
The theorem is proved elementarily, and the Lean formalisation (about 3000 lines) contains no sorry, depending only on standard axioms plus a cited Ramanujan inequality supplied as an explicit hypothesis to the main theorem erdos_306_of_ramanujan.