I would like to illustrate my lecture on p-adic numbers with some elementary results. I proved that the series $e^p=\sum_{n\ge0}\frac{p^n}{n!}$ converges in $\mathbb Q_p$ for every prime $p$.

Now I would like to teach that $e^p$ is irrational in $\mathbb Q_p$ by elementary methods (this is true by Mahler result on transcendence of $e^x$ in $p$-adic fields).

Do you know such an elementary proof. I can't find one in the literature...

Good catch! I meant $\frac{p^n}{n!}$ instead of $\frac1n$. I edited it.

  • 5
    $\begingroup$ The sum of $1/n$ doesn't converge $p$-adically, since the $n$'th term doesn't go to $0$. Also, you certainly don't want to start with $n=0$. The series $\sum_{n\ge1} x^n/n$ converges if and only if $|x|_p<1$. And for $e^x$, the series $\sum_{n\ge0}x^n/n!$ converges for $|x|_p<1$ if $p$ is odd, otherwise you need $|x|_2<\frac12$. $\endgroup$ – Joe Silverman Feb 12 '14 at 3:25
  • 14
    $\begingroup$ You should call the constant $e^p$. $\endgroup$ – S. Carnahan Feb 12 '14 at 3:37
  • 8
    $\begingroup$ Please make sure you tell the students that this $p$-adic $e^p$ has nothing at all to do with the real number $e$. But then it also means that this application is less exciting than others, since the $p$-adic $e^p$ (or $e^4$ when $p = 2$) is some new, unfamiliar number in the student's experience. Why not use $p$-adics to show any prime factor in the denominator of $\binom{r}{n}$, for $r \in {\mathbf Q}$, must be a prime in the denominator of $r$? That's a nice use of $p$-adic limits and continuity to say something about rational numbers. $\endgroup$ – KConrad Feb 12 '14 at 9:16
  • 3
    $\begingroup$ @Gerald Edgar: just out of curiosity: how would you go about proving that, for $p$ a prime such that $p\equiv 1\pmod{4}$, the $p$-adic expansion of $\sqrt{-1}$ is aperiodic? Personally, I would prove this as a corollary to the fact that $\sqrt{-1} \notin \mathbb{Q}$. However, you seem to be saying that there is a more direct proof. $\endgroup$ – RP_ Feb 13 '14 at 0:04
  • 1
    $\begingroup$ Does it converge for $p=2$? It seems that no. $\endgroup$ – Fedor Petrov Apr 26 '16 at 17:25

Maybe I harp on this too much, but I think the $p$-adic exponential is not nearly as interesting as the logarithm. It’s defined on a much larger subgroup than the exponential, it’s a homomorphism (unlike the complex logarithm), and its roots are the $p$-power roots of unity. As a function on $1+D$, where $D$ is the open unit disk in $\Bbb C_p$, it’s onto $\Bbb C_p$, so that even though for $z\in\Bbb C_p$ you won’t usually be able to define $e^z$, you can find $\alpha\in\Bbb C_p^*$ such that $\log\alpha=z$.

  • 1
    $\begingroup$ But do you have a simple/elementary proof of the fact that $log(1+p)$ is irrational? Or of the fact that $log(1+x)$ is irrational for every rational $p$-adic $x$ such that $|x|<1$? (the latter would of course answer OP's question) I'd love to see a short and simple proof of the irrationality of just about any $p$-adic "of this sort" (and that is, in fact, transcendental). $\endgroup$ – Gro-Tsen Apr 28 '16 at 18:29
  • $\begingroup$ Indeed I don’t. The burden of my argument was that it makes better sense to ask about $\log(1+p)$ than $\exp p$. $\endgroup$ – Lubin Apr 29 '16 at 5:23

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.