This is an interesting problem. This paper discusses the problem at length and contains a strong result that implies, in particular, the following:
Let $g:\mathbb{R}\longrightarrow\mathbb{R}$ be defined by $g(x)=ax^2+(b+1)x+c$ and let $\Delta(g)=b^2-4ac$. Then for each positive integer $n$, a solution $f:\mathbb{R}\longrightarrow\mathbb{R}$ to $f^n=g$ exists if and only if $\Delta(g)\leq 1$.
In your case, we have $g(x)=x^2+1$, so $\Delta(g)=1-4=-3$, and hence existence is guaranteed.
Further interesting readings on the topic might include Schröder's equation, or this paper on solutions of $f^2=g$ for $g$ that fail to be injective.