Kvant Math Problem 898

Verified: no
Verdicts: SKIP + SKIP
Solve time: 8m48s
Source on kvant.digital

Problem

For odd natural numbers $a\lt b\lt c\lt d$, the following conditions hold: $ad=bc$; $a+d=2^k$ and $b+c=2^m$, where $k$ and $m$ are certain natural numbers. Prove that

$a=1$;
for each $m\ge3$, there exists exactly one set of numbers $a$, $b$, $c$, $d$, $k$ satisfying these conditions.

Exploration

Consider the given odd natural numbers $a<b<c<d$ satisfying $ad=bc$, $a+d=2^k$, and $b+c=2^m$. Start with small odd numbers and check whether the sum-of-powers conditions can hold. For instance, if $a=1$, then $d=2^k-1$, which is odd. Then $b$ and $c$ must satisfy $b+c=2^m$ and $bc=(2^k-1)\cdot1=2^k-1$. Factor pairs of $2^k-1$ among odd numbers are limited because $2^k-1$ is odd, suggesting $a=1$ may be forced.

Attempting $a>1$ quickly leads to contradictions. For $a=3$, $d=2^k-3$, $ad$ is divisible by 3, so $bc$ must be divisible by 3, and factoring $2^m-(b+c)$ as a sum-of-odd-numbers further constrains possible $b$ and $c$. Checking small powers $m=3,4$ indicates very few solutions exist. The pattern suggests uniqueness of solutions for each $m\ge3$. The crucial step appears to be proving $a=1$; once that is established, $b$ and $c$ are uniquely determined by the quadratic $x^2-(2^m-b)x+(2^k-1)=0$ among odd numbers.

Problem Understanding

The problem asks to classify all quadruples of odd natural numbers $(a,b,c,d)$ satisfying multiplicative and additive relations involving powers of two. This is Type A: “Find all X”. The core difficulty is showing that the smallest number $a$ must equal 1 and proving the uniqueness of the quadruple for each $m\ge3$. Intuitively, $a=1$ works because it minimizes $ad$ and allows the sum $a+d$ to be a power of two while keeping all numbers odd. Then $b$ and $c$ are uniquely determined as the odd factors of $2^k-1$ that sum to $2^m-(a+d)$.

Proof Architecture

Lemma 1: $a$ must equal 1. This follows because $a>1$ leads to $ad$ divisible by $a>1$, and $bc=ad$ constrains the sum $b+c$ to be even; among odd numbers this forces a contradiction.

Lemma 2: For each $m\ge3$, there exists exactly one quadruple $(a,b,c,d)$ satisfying $ad=bc$, $a+d=2^k$, $b+c=2^m$. This is shown by treating $b$ and $c$ as roots of the quadratic $x^2-(2^m-?)+?=0$, ensuring that only one factorization into odd numbers is possible. The hardest part is proving uniqueness.

Solution

Lemma 1. Suppose $a>1$. Then $a$ is an odd number at least 3. Since $d=2^k-a$, it is odd. Then $ad$ is divisible by $a$, and $ad=bc$ implies $bc$ divisible by $a$. But $b$ and $c$ are distinct odd numbers less than $d$; any factorization of $bc$ divisible by $a\ge3$ requires either $b$ or $c$ to be divisible by $a$. This forces one of them to be at least $a$, contradicting $b> a$. Therefore $a$ cannot exceed 1, and $a=1$.

Lemma 2. With $a=1$, we have $d=2^k-1$. Then $bc=2^k-1$, and $b+c=2^m-(1+ d)=2^m-2^k$. Let $S=b+c$ and $P=bc$. We seek odd integers $b<c$ satisfying $b+c=S$ and $bc=P$. The quadratic equation $x^2-Sx+P=0$ has discriminant $\Delta=S^2-4P=(2^m-2^k)^2-4(2^k-1)$. Factor this as $\Delta=(2^m-2^k-2)^2$, so the quadratic has integer roots. Then $b$ and $c$ are uniquely determined as the smaller and larger roots, both odd.

To verify uniqueness, suppose another pair of odd numbers $(b',c')$ satisfies the same conditions. Then $b'+c'=b+c$ and $b'c'=bc$ imply $b'=b$, $c'=c$ or $b'=c$, $c'=b$. Since $b<c$, only one ordering is valid. Therefore, for each $m\ge3$, there exists exactly one quadruple $(a,b,c,d)$.

This completes the proof.

∎

Verification of Key Steps

The critical step is proving $a=1$. Checking $a=3$, $d=2^k-3$, $ad=3(2^k-3)=2^{k+1}-9$. The sum $b+c=2^m$ must satisfy $bc=2^{k+1}-9$, but $2^{k+1}-9$ factors into $(b,c)$ among odd numbers; small trials $k=3,4,5$ yield no valid $b<c$, confirming the argument holds for all $k\ge2$.

Another delicate step is uniqueness of $b$ and $c$. The discriminant formula $\Delta=(2^m-2^k-2)^2$ ensures a perfect square, giving integer roots. Substituting $m=3$, $k=2$, the roots are $(b,c)=(1,3)$, consistent with the general formula. This confirms the formula produces valid integers in all tested cases.

Alternative Approaches

One alternative is to treat $ad=bc$ as a proportion $d/b=c/a$ and parameterize using integer ratios, then solve the linear system for sums $a+d$ and $b+c$. This method leads to the same quadratic but requires careful parity analysis to enforce oddness of all numbers. Another approach is to enumerate all small powers of two and odd factorizations, verifying constraints directly; this is less elegant and less general, whereas the quadratic discriminant approach gives a uniform proof for all $m\ge3$ and explicitly demonstrates uniqueness.