Find $n$ such that the polynomial function $1+X+X^2$ divides $P_n = (1+X^4)^n-X^n$.
What I've written is $$ \left(1+X^4\right)^n - X^n = \left(1+X+X^2\right)Q\left(X\right) + R\left(X\right) $$ where $R\left(X\right) = aX+b$. Finding $a$ and $b$ as function of $n$ 'd lead me to the result by putting $a=b=0$.
What I've tried is to evaluate the equation in $e^{2i\pi/3}$ and $e^{-2i\pi/3}$ which are the roots of $1+X+X^2$ then $$ P_n\left(e^{2i\pi/3}\right) = R\left(e^{2i\pi/3}\right) \text{ and also }P_n\left(e^{-2i\pi/3}\right) = R\left(e^{-2i\pi/3}\right) $$ However my problem 'd be to simplify as far as I can $P_n\left(e^{2i\pi/3}\right)$ : $$P_n\left(e^{2i\pi/3}\right) = \left(1+e^{8 i \pi/3}\right)^n-\left(e^{2i\pi/3}\right)^n=\left(1+e^{2 i \pi/3}\right)^n-\left(e^{2i\pi/3}\right)^n = \sum_{k=0}^{n-1}\left(1+e^{2i\pi/3}\right)^{n-k}\left(e^{2i\pi/3}\right)^k $$ Then, yes I can make the sum of a geometrical sequence terms appear, but it looks still not friendly to manipulate. Is there a better way to go to solve the problem ? I think the result is $n$ being a multiple of $6$
Thanks for the help