No Heronian Triangle Has Catalan-Number Sides

Theorem. There is no Heronian triangle whose three sides are all Catalan numbers. Proof in four steps. (1) Growth-rate reduction. The Catalan recurrence C_{n+1}/C_n = 2(2n+1)/(n+2) satisfies C_{n+1}/C_n ≥ 2 for n ≥ 1, and C_{n+2}/C_n > 4 for n ≥ 0, so any valid Catalan-sided triangle (C_i, C_j, C_k) with i ≤ j ≤ k must satisfy k − i ≤ 1 (else C_k ≥ C_{i+2} > C_i + C_j, violating the strict triangle inequality). The candidate configurations are therefore the three families (A) equilateral (C_n, C_n, C_n), (B) (C_n, C_n, C_{n+1}), (C) (C_n, C_{n+1}, C_{n+1}). (2) Family A: area = C_n²·√3/4 is irrational. (3) Family B: for n ≥ 1, C_{n+1} ≥ 2 C_n, so C_n + C_n ≤ C_{n+1}, failing strict triangle inequality; for n = 0, (1, 1, 1) is equilateral — covered by Family A. (4) Family C: expand Heron's formula and substitute C_{n+1} = 2(2n+1)·C_n/(n+2). The difference 4 C_{n+1}² − C_n² factors as (7n+2)(9n+6) = 3·(7n+2)(3n+2), yielding 16 A² = C_n^4 · 3(7n+2)(3n+2) / (n+2)². For A integer, 3(7n+2)(3n+2) must be a perfect square. Case 4a: if n ≢ 1 (mod 3), then v_3(7n+2) = 0 and v_3(3n+2) = 0, so v_3(3(7n+2)(3n+2)) = 1 — odd, not a square. Case 4b: if n = 3t+1, 3(7n+2)(3n+2) = 9·(7t+3)(9t+5), so the square condition reduces to (7t+3)(9t+5) = s² for some integer s ≥ 0. Expanding and completing the square: 252·(7t+3)(9t+5) = (126t + 62)² − 64, giving (63t + 31)² − 63 s² = 16, a Pell-like equation with u = 63t+31. The orbits of solutions under the fundamental unit 8 + √63 (which has norm 1) are closed; searching representatives with 0 ≤ u ≤ 68 finds only {(4, 0), (32, 4)}, which lie in a single orbit (since (4,0)·(8+√63) = (32, 4)). The orbit's u-values modulo 63 alternate 4, 32, 4, 32, ..., so u ≡ 31 (mod 63) never occurs. Therefore (7t+3)(9t+5) = s² has no integer solutions t ≥ 0. Combining Cases 4a and 4b closes Family C. ∎ The result is significant because it illustrates how classical sequences defined by simple recurrences can interact non-trivially with Heronian-triangle constraints: Catalan numbers grow too fast for most index configurations to even form a triangle, and the remaining family is killed by a clean Diophantine obstruction that ultimately rests on the Pell equation for √63. A complete exhaustive check over all 1,541 Catalan triples with C_k ≤ 10^30 (55 distinct Catalan values through C_54 ≈ 4.5 × 10^29) independently confirms: zero Heronian triangles.

Catalan numbers show up everywhere in counting problems: how many ways to balance parentheses, how many triangulations a polygon has, how many paths on a grid stay below the diagonal. They form the sequence 1, 1, 2, 5, 14, 42, 132, 429, 1430, ... , each one a little less than 4 times the last. A natural question: is there any triangle whose three sides are Catalan numbers and whose area is a whole number? The answer is no — there isn't a single one, not even for enormous Catalan numbers beyond astronomical sizes. The proof has three parts. Part one: Catalan numbers grow so fast that out of all possible triples (C_i, C_j, C_k), only triples where all three indices are within 1 of each other can even form a real triangle (the rest violate the triangle inequality, meaning two of the sides can't stretch to meet the third). That narrows the possibilities to three shapes: (C_n, C_n, C_n) equilateral, (C_n, C_n, C_{n+1}) where the odd-one-out is bigger, and (C_n, C_{n+1}, C_{n+1}) where the odd-one-out is smaller. Part two: the first two shapes are easy to eliminate. Equilateral triangles have areas that are always √3/4 times an integer, never integer themselves. And the second shape (C_n, C_n, C_{n+1}) always fails the triangle inequality because C_{n+1} is at least twice C_n. Part three: the third shape is where the real work happens. Plug in Catalan's formula and push through the algebra, and the question becomes 'is 3·(7n+2)·(3n+2) ever a perfect square?' For two-thirds of all n values (those that aren't 1 more than a multiple of 3), the factor of 3 gets stuck with an odd exponent and kills the possibility. The remaining one-third reduces to a famous kind of equation called a Pell equation, specifically u² − 63·s² = 16. This equation has infinitely many integer solutions, but they all have u equal to either 4 or 32 modulo 63 — and the one value we'd need, 31, never appears. So no n works, and the last family is ruled out. The theorem is complete: no Catalan-sided Heronian triangle exists anywhere. I also ran a brute-force computer check over every Catalan triple with sides up to about 10^30 (55 distinct Catalan values) and it found none, matching the theorem exactly.

No Heronian triangle has all three sides equal to Catalan numbers.

Growth-rate reduction (k−i ≤ 1) → 3 families A, B, C → A irrational, B triangle-inequality failure, C reduces to 3(7n+2)(3n+2) square → mod-3 kills 2/3 of n → Pell equation (63t+31)² − 63 s² = 16 kills the rest (u ≡ 4 or 32 mod 63 only).

55 distinct Catalan values (C_0..C_54) · 1,541 Catalan triples tested · 0 Heronian triangles found