Recently I opened a question about what it might be a new property of Euler's Totient function.
I am still studying the Totient function and I found another interesting relationship, it is very different from the ones I wrote in my former question, so I would like to ask about it apart because this might be another interesting topic. As far as I have checked, I did not find any reference to this property, but if I am wrong I will remove the question. This is my new proposed statement:
$\forall n\ge3$, if $(n-\phi(n)) = \sqrt{n}\ $,$\ $then $(n-\phi(n)) \in \Bbb P$
Meaning that $n=p^2$ is the perfect square of a prime $p$ if and only if the distance between $n$ and $\phi(n)$ is exactly $p$.
I have also tested this using Python, and no counterexamples are found in the interval $[1..7000]$.
Usually when I find these kind of possible properties I test them, but I do not have enough theoretical knowledge to prove them, so any comment, hint or explanation about it is very appreciated. So as usual my questions are:
Is my proposed statement about the Totient function already known or trivial?
Is there a counterexample of it? Thank you!
Your statement needs to be more precise.
The implication $\Rightarrow$ is obvious, if $n=p^2$ then it follows immediately from the definition or formula pf $\phi(n)$ that $n-\phi(n)=p$.
For the converse, you need to be more precise. The way in which is stated, is correct but trivial: you are saying $n=p^2$ if and only if $n=p^2$ and $n=(n-\phi(n))^2$.
If you mean $n=p^2 \Leftrightarrow n-\phi(n)$ is prime, this is not true. $n=15$ fails the converse.
Added
If $p=n-\phi(n)=\sqrt{n}$ it follows that $n=p^2$.
Now, it is easy to see that for all integers $m$ we have $\phi(m^2)=m \phi(m)$. Therefore
$$p=p^2-p\phi(p) \Rightarrow \phi(p) = p-1 $$
This implies $p \in P$.