show that a power of 3 can never be a perfect number


First I asked myself what does a number to a power of 3 look like: so i said let $n^{3^k}$

Next, what would a perfect number of this form have to look like. Well if this was to be a perfect number it should look like: $$n^{3^k} = 1 + n^{0^1} + n^{2^1} + n^{3^1} +...+ n^{3^2} + n^{3^3} +...+ n^{3^{k-1}}$$ In this form all of the possible divisors of $n^{3^k}$ are captured.

When i wrote it I thought it was a geometric sequence so with some fiddling I obtained $$\frac{(n^{3})^k - 1}{n^3 - 1}$$ Looking at the expression you see that $n^3 - 1$ actually is a divisor of $(n^{3})^k - 1$. So it can be expressed as $$ (n^{3})^k - 1 = q(n^3 - 1)$$ Which then would mean $$n^{3k} = q(n^3 - 1) + 1$$ But this would mean that $n^{3^k}$ always has a remainder, so cannot be a perfect number.

notes: 1) i am concerned that the $n^{3^k}$ in my original expression and the $n^{3k}$ obtained at the end cannot be manipulated in the way I did it eventhough I used the properties of exponents, so my solution is off.

2) the solution i have seen in the book states through a string of inequalities : $$ \sigma(p^k) = \frac{p^{k+1} - 1}{p - 1} < \frac{p^{k+1}}{p - 1} = (\frac{p}{p - 1})p^k$$ where the quantity $(\frac{p}{p - 1})$ is less than 2 for any number $p > 2$, where $\sigma(n)$ = {sum of all divisors of n}

My issue with the second note is that they assumed that the quantity $\sigma(p^k)$ can never be as large as $2p^k$. but to be able to assume that you must be assuming that every perfect number is even, which although the existence of odd perfect numbers has not been established, it hasn't been disproven either. So how can this be used as an assumption?

  • 1
    $\begingroup$ Do you mean $n^{3^k}$ or $3^n$? That is, "a number to a power of 3" or "a power of 3"? $\endgroup$
    – GoodDeeds
    Commented Feb 26, 2016 at 19:17
  • $\begingroup$ Exactly: I think a power of $3$ means $3^k$ for $k = 0, 1, 2, \ldots$ $\endgroup$
    – Brian Tung
    Commented Feb 26, 2016 at 19:20
  • $\begingroup$ the question phrased it "a power of 3".....so i may have just interpreted the phrase wrong.....oh god $\endgroup$ Commented Feb 26, 2016 at 19:20
  • $\begingroup$ you guys are right. I misinterpreted the question.....all of that time writing out an attempt in latex all for nothing sigh......any points for "creativity"? $\endgroup$ Commented Feb 26, 2016 at 19:27

1 Answer 1


In case you meant "a power of $3$":

The sum of the factors of $3^n$ is $\frac{3^{n+1}-1}{3-1}$. For $3^n$ to be perfect, we would need $$ \frac{3^{n+1}-1}{3-1}=2\cdot3^n\implies-3^n-1=0 $$

  • $\begingroup$ What does the last implication imply? $\endgroup$ Commented Feb 26, 2016 at 19:26
  • $\begingroup$ $-3^n-1=0$ is a contradition...there is no $n$ that would ever make this true... $\endgroup$ Commented Feb 26, 2016 at 19:27
  • $\begingroup$ i don't get $-3^n$ when i work it out: $$3^{n+1} - 2(3^n) - 1 = 3^n(3-2) - 1 = 0$$ , but what about n = 0? $\endgroup$ Commented Feb 26, 2016 at 19:37
  • $\begingroup$ Multiplying both sides by $2$ gives $3\cdot3^n-1=4\cdot3^n$. Subtracting $4\cdot3^n$ from both sides gives $-3^n-1=0$. There is no $n$ that satisfies this since $3^x\gt0$ for all real $x$. $\endgroup$
    – robjohn
    Commented Feb 26, 2016 at 19:46

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .