Is this proof for the divisibility constraint $\sigma(q^k)/2 \mid n$ correct, where $q^k n^2$ is an odd perfect number with special prime $q$?

41 Views Asked by At

Let $N = q^k n^2$ be an odd perfect number given in Eulerian form, i.e. $q$ is the special prime satisfying $q \equiv k \equiv 1 \pmod 4$ and $\gcd(q,n)=1$.

Define the GCDs $$G = \gcd(\sigma(q^k),\sigma(n^2))$$ $$H = \gcd(n^2,\sigma(n^2))$$ and $$I = \gcd(n,\sigma(n^2)),$$ where $\sigma(x)=\sigma_1(x)$ is the classical sum of divisors of the positive integer $x$.

Note that $$G \times H = I^2$$ $$G \mid I$$ $$I \mid H$$ and that $$H = \frac{\sigma(n^2)}{q^k} = \frac{n^2}{\sigma(q^k)/2}.$$

Let $$J = \frac{I}{G} = \frac{H}{I} = \sqrt{\frac{H}{G}}.$$

It can be shown that $$J = \frac{n}{\gcd(\sigma(q^k)/2, n)}.$$


Suppose that $$J = \frac{n}{\gcd(\sigma(q^k)/2, n)} = \frac{n}{d}, \tag{1}$$ where $d$ is a divisor of $n$.

Note that $(1)$ is true if and only if $\gcd(\sigma(q^k)/2, n) = d$.

Notice that $$\frac{Jnd}{\sigma(q^k)/2} = \frac{n^2}{\sigma(q^k)/2} = H.$$

But we also know that $$H = \frac{\sigma(n^2)}{q^k} = \frac{n^2}{\sigma(q^k)/2}.$$

Hence, under the given assumption, we get $$\sigma(n^2) = \frac{Jnd}{\sigma(q^k)/2}\cdot{q^k}$$ $$n = Jd.$$

Consequently, we obtain $$\gcd(n,\sigma(n^2)) = \gcd\left(Jd,\frac{Jnd}{\sigma(q^k)/2}\cdot{q^k}\right)$$ Since $\gcd(\sigma(q^k)/2, n) = d$, then $d \mid \sigma(q^k)/2$ holds.

Since $\gcd(q^k,\sigma(q^k)/2)=1$, we consider three cases:

Case (1): $\sigma(q^k)/2 \mid d$

This case implies that $d = \sigma(q^k)/2$, and therefore that $\sigma(q^k)/2 \mid n$ holds.

Case (2): $\sigma(q^k)/2 \mid n$

No problem.

Case (3): $\sigma(q^k)/2 \mid J$

$$\sigma(q^k)/2 \mid J \implies J = e\cdot\sigma(q^k)/2 \text{ for some } e \implies \frac{n}{\gcd(\sigma(q^k)/2, n)} = e\cdot\sigma(q^k)/2$$ $$\implies n = e\cdot{\sigma(q^k)/2}\cdot{\gcd(\sigma(q^k)/2, n)} \implies \sigma(q^k)/2 \mid n$$


Here is my:

QUESTION: Does this proof suffice? I have doubts if the list of three cases is complete, as I am aware that it may happen that either $\sigma(q^k)/2 \mid Jn$, $\sigma(q^k)/2 \mid Jd$, or $\sigma(q^k)/2 \mid nd$ also hold.


Update (Added on March 26, 2023 - 15:22 PM (Manila time))

The condition $\sigma(q^k)/2 \mid Jd$ is equivalent to $\sigma(q^k)/2 \mid n$ since $Jd = n$.

That leaves us (???) with the following three further cases to consider: $\sigma(q^k)/2 \mid Jn$, $\sigma(q^k)/2 \mid nd$, or $\sigma(q^k)/2 \mid Jnd$.

Alas, this is where I get stuck!