I posted this question on MSE two days ago, but did not receive any responses. I have cross-posted it on MO, hoping it gets more attention here and that it is appropriate for this site.

A positive integer $N$ is said to be perfect if $\sigma(N) = 2N$, where $\sigma(x)$ is the sum of the divisors of $x$.

An odd perfect number $N$ is said to be given in Eulerian form if $N = {q^k}{n^2}$, where $q$ is prime, $q \equiv k \equiv 1 \pmod 4$ and $\gcd(q,n)=1$.

Broughan, Delbourgo and Zhou (2013) defines the perfect number index of $N$ at prime $r$ to be the integer $$m := \frac{\sigma(N/{r^{\alpha}})}{r^{\alpha}},$$ where $r^{\alpha} || N$. They also show that $m \ge 315$. (Chen and Chen (2014) extend these results in "On the index of an odd perfect number".)

Since $\sigma$ is weakly multiplicative, we have $$\sigma(q^k)\sigma(n^2)=\sigma(N)=2N=2{q^k}{n^2}.$$ Because $\gcd(q^k,\sigma(q^k))=1$, this means that $$\sigma(n^2)/q^k = 2{n^2}/\sigma(q^k) = s \in \mathbb{N},$$ where $s \ge 315$.

In particular, we have the simultaneous equations $$\sigma(n^2) = s{q^k}$$ and $$2{n^2} = s{\sigma(q^k)}.$$

We obtain $$2{n^2} - \sigma(n^2) = s{\sigma(q^k)} - s{q^k} = s{\sigma(q^{k-1})}.$$

UPDATE - September 27 2016 In fact, we know that $s = \gcd(n^2,\sigma(n^2))$.

It follows that $$\sigma(q^{k-1}) \mid (2{n^2} - \sigma(n^2)).$$

Now, here is my question:

What are the divisors of $\sigma(q^{k-1})$?

Notice that $4 \mid (k-1)$.

By simple congruence considerations: $$\sigma(q^{k-1}) \equiv \sum_{i=0}^{k-1}{1} \equiv 1 + (k-1) \equiv k \equiv 1 \pmod 4.$$

Additionally, I checked using WolframAlpha and found only the following factorization (for $k = 9$): $$\sigma(q^8) = 1 + q + q^2 + q^3 + q^4 + q^5 + q^6 + q^7 + q^8 = (q^2 + q + 1)(q^6 + q^3 + 1) = \sigma(q^2)(q^6 + q^3 + 1)$$


In general there is a factorization $\sigma(q^{k-1}) = 1+q+\ldots+q^{k-1} = \frac{q^k-1}{q-1} = \prod_{d\mid k, d>1} \Phi_d(q)$. Here $\Phi_d$ is the $d$'th cyclotomic polynomial, defined as the minimal polynomial of a primitive $d$'th root of unity. Your example is a special case of this since $\Phi_3(x)=1+x+x^2$ and $\Phi_9(x)=x^6+x^3+1$.

In certain cases it possible to factor $\Phi_d(q)$ further. See "aurifeuillean factorization" on Wikipedia.

| cite | improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.