Amicable Pair with Smallest Common Prime Factor 5
Theorem
An amicable pair whose smallest common prime factor is greater than $3$ has the elements:
- $m_1 = 5 \times 7^2 \times 11^2 \times 13 \times 17 \times 19^3 \times 23 \times 37 \times 181 \times 101 \times 8693 \times 19 \, 479 \, 382 \, 229$
and:
- $m_2 = 5 \times 7^2 \times 11^2 \times 13 \times 17 \times 19^3 \times 23 \times 37 \times 181 \times 365 \, 147 \times 47 \, 307 \, 071 \, 129$
This was at one point the smallest known counterexample to the observation that:
- most amicable pairs consist of even integers
- most of the rest, whose elements are odd, have both elements divisible by $3$.
Proof
It is to be demonstrated that these numbers are amicable.
From Divisor Sum of Integer:
- $\ds \map {\sigma_1} n = \prod_{1 \mathop \le i \mathop \le r} \frac {p_i^{k_i + 1} - 1} {p_i - 1}$
where:
- $\ds n = \prod_{1 \mathop \le i \mathop \le r} p_i^{k_i} = p_1^{k_1} p_2^{k_2} \cdots p_r^{k_r}$
is the prime decomposition of $n$.
When $k_i = 1$ the individual factor becomes $\dfrac {p_i^2 - 1} {p_i - 1} = \dfrac {\paren {p_i + 1} \paren {p_i - 1} } {p_i - 1} = p_i + 1$.
We establish the contributions to the divisor sum values of $m_1$ and $m_2$ by taking the prime factors in turn, and extracting the prime factors of each result.
First, the elements common to both:
\(\ds \map {\sigma_1} 5\) | \(=\) | \(\ds 6\) | Divisor Sum of Prime Number | |||||||||||
\(\ds \) | \(=\) | \(\ds 2 \times 3\) |
\(\ds \map {\sigma_1} {7^2}\) | \(=\) | \(\ds \dfrac {7^3 - 1} {7 - 1}\) | Divisor Sum of Power of Prime | |||||||||||
\(\ds \) | \(=\) | \(\ds \dfrac {343 - 1} 6\) | ||||||||||||
\(\ds \) | \(=\) | \(\ds 57\) | ||||||||||||
\(\ds \) | \(=\) | \(\ds 3 \times 19\) |
\(\ds \map {\sigma_1} {11^2}\) | \(=\) | \(\ds \dfrac {11^3 - 1} {11 - 1}\) | Divisor Sum of Power of Prime | |||||||||||
\(\ds \) | \(=\) | \(\ds \dfrac {1331 - 1} {10}\) | ||||||||||||
\(\ds \) | \(=\) | \(\ds 133\) | ||||||||||||
\(\ds \) | \(=\) | \(\ds 7 \times 19\) |
\(\ds \map {\sigma_1} {13}\) | \(=\) | \(\ds 14\) | Divisor Sum of Prime Number | |||||||||||
\(\ds \) | \(=\) | \(\ds 2 \times 7\) |
\(\ds \map {\sigma_1} {17}\) | \(=\) | \(\ds 18\) | Divisor Sum of Prime Number | |||||||||||
\(\ds \) | \(=\) | \(\ds 2 \times 3^2\) |
\(\ds \map {\sigma_1} {19^3}\) | \(=\) | \(\ds \dfrac {19^4 - 1} {19 - 1}\) | Divisor Sum of Power of Prime | |||||||||||
\(\ds \) | \(=\) | \(\ds \dfrac {130 \, 321 - 1} {18}\) | ||||||||||||
\(\ds \) | \(=\) | \(\ds 7240\) | ||||||||||||
\(\ds \) | \(=\) | \(\ds 2^3 \times 5 \times 181\) |
\(\ds \map {\sigma_1} {23}\) | \(=\) | \(\ds 24\) | Divisor Sum of Prime Number | |||||||||||
\(\ds \) | \(=\) | \(\ds 2^3 \times 3\) |
\(\ds \map {\sigma_1} {37}\) | \(=\) | \(\ds 38\) | Divisor Sum of Prime Number | |||||||||||
\(\ds \) | \(=\) | \(\ds 2 \times 19\) |
\(\ds \map {\sigma_1} {181}\) | \(=\) | \(\ds 182\) | Divisor Sum of Prime Number | |||||||||||
\(\ds \) | \(=\) | \(\ds 2 \times 7 \times 13\) |
This gives a common factor of both $\sigma_1$ values of:
\(\ds \) | \(\) | \(\ds 6 \times 57 \times 133 \times 14 \times 18 \times 7240 \times 24 \times 38 \times 182\) | ||||||||||||
\(\ds \) | \(=\) | \(\ds 2^{11} \times 3^5 \times 5 \times 7^3 \times 13 \times 19^3 \times 181\) |
The remaining prime factors of $m_1$:
\(\ds \map {\sigma_1} {101 \times 8693 \times 19 \, 479 \, 382 \, 229}\) | \(=\) | \(\ds \paren {101 + 1} \paren {8693 + 1} \paren {19 \, 479 \, 382 \, 229 + 1}\) | Divisor Sum of Square-Free Integer | |||||||||||
\(\ds \) | \(=\) | \(\ds \paren {2 \times 3 \times 17} \paren {2 \times 3^3 \times 7 \times 23} \paren {2 \times 3 \times 5 \times 7 \times 11^2 \times 37 \times 20 \, 719}\) | ||||||||||||
\(\ds \) | \(=\) | \(\ds 2^3 \times 3^5 \times 5 \times 7^2 \times 11^2 \times 17 \times 23 \times 37 \times 20 \, 719\) |
The remaining prime factors of $m_2$:
\(\ds \map {\sigma_1} {365 \, 147 \times 47 \, 307 \, 071 \, 129}\) | \(=\) | \(\ds \paren {365 \, 147 + 1} \paren {47 \, 307 \, 071 \, 129 + 1}\) | Divisor Sum of Square-Free Integer | |||||||||||
\(\ds \) | \(=\) | \(\ds \paren {2^2 \times 3^4 \times 7^2 \times 23} \paren {2 \times 3 \times 5 \times 11^2 \times 17 \times 37 \times 20 \, 719}\) | ||||||||||||
\(\ds \) | \(=\) | \(\ds 2^3 \times 3^5 \times 5 \times 7^2 \times 11^2 \times 17 \times 23 \times 37 \times 20 \, 719\) |
and so:
\(\ds \map {\sigma_1} {m_1} = \map {\sigma_1} {m_2}\) | \(=\) | \(\ds \paren {2^{11} \times 3^5 \times 5 \times 7^3 \times 13 \times 19^3 \times 181} \paren {2^3 \times 3^5 \times 5 \times 7^2 \times 11^2 \times 17 \times 23 \times 37 \times 20 \, 719}\) | ||||||||||||
\(\ds \) | \(=\) | \(\ds 2^{14} \times 3^{10} \times 5^2 \times 7^5 \times 11^2 \times 13 \times 17 \times 19^3 \times 23 \times 37 \times 181 \times 20 \, 719\) |
It remains to be shown that:
- $\map {\sigma_1} {m_1} = \map {\sigma_1} {m_2} = m_1 + m_2$
So:
\(\ds m_1 + m_2\) | \(=\) | \(\ds 5 \times 7^2 \times 11^2 \times 13 \times 17 \times 19^3 \times 23 \times 37 \times 181 \times \paren {101 \times 8693 \times 19 \, 479 \, 382 \, 229 + 365 \, 147 \times 47 \, 307 \, 071 \, 129}\) | ||||||||||||
\(\ds \) | \(=\) | \(\ds 6921698310331805 \times \paren {17102761241386397 + 17274035101540963}\) | ||||||||||||
\(\ds \) | \(=\) | \(\ds 6921698310331805 \times 34376796342927360\) | ||||||||||||
\(\ds \) | \(=\) | \(\ds 237945813161460881075326612684800\) | ||||||||||||
\(\ds \) | \(=\) | \(\ds 2^{14} \times 3^{10} \times 5^2 \times 7^5 \times 11^2 \times 13 \times 17 \times 19^3 \times 23 \times 37 \times 181 \times 20719\) |
$\blacksquare$
Historical Note
In $1986$, David Wells reported in his Curious and Interesting Numbers of $1986$ that "numerous mathematicians" had conjectured that all amicable pairs both of whose elements are odd are multiples of $3$.
This amicable pair, a counterexample, was communicated by letter dated $15$th May $1987$ from Herman te Riele to Richard K. Guy.
The latter published it, with mistakes, in his Unsolved Problems in Number Theory, 2nd ed. of $1994$:
- $5 \cdot 7^2 \cdot 11^2 \cdot 13 \cdot 17 \cdot 19^3 \cdot 23 \cdot 37 \cdot 181 \begin{cases} 101 \cdot 8643 \cdot 1947938229 \\ 365147 \cdot 47303071129 \end{cases}$
David Wells repeated the mistakes in his Curious and Interesting Numbers, 2nd ed. of $1997$.
Guy finally published the corrected version in his Unsolved Problems in Number Theory, 3rd ed. of $2004$.
It was for some time believed to be the smallest such amicable pair, but smaller ones have since been discovered.
Sources
- 1997: David Wells: Curious and Interesting Numbers (2nd ed.) ... (previous) ... (next): $220$