D. This fails because if $x_0\ne (0,\dots, 0),$ then the map $x\to x + x_0$ satisfies the condition.
C. Suppose $Df(x_0)$ is singular at some $x_0.$ Then $Df(x_0)(v) = 0$ for some nonzero $v\in \mathbb R^n.$ Thus as $t\to 0$ in $\mathbb R,$
$$f(x_0 + tv)- f(x_0)= Df(x_0)(tv) + o(|tv|) = 0 + o(|tv|) .$$
This violates the given condition on $f.$ Therefore $Df$ is nonsingular everywhere. Hence by the inverse function theorem, $f$ is an open map. Thus $f(\mathbb R^n)$ is open.
B. Suppose $y_k \in f(\mathbb R^n)$ and $y_k \to y_0.$ Each $y_k = f(x_k)$ for some $x_k\in \mathbb R^n.$ The condition on $f$ shows that if $\{x_k\}$ is unbounded, then so is $\{f(x_k)\}.$ That is a contradiction, and therefore $\{x_k\}$ is bounded. Thus there exists a subsequence $x_{k_j}$ converging to some $x_0\in \mathbb R^n.$ It follows by continuity that $f(x_{k_j}) \to f(x_0).$ Since $f(x_{k_j}) \to y_0,$ we have $y_0= f(x_0).$ Thus $y_0 \in f(\mathbb R^n),$ proving $f(\mathbb R^n)$ is closed.
A. We have shown $f(\mathbb R^n)$ is both open and closed in $\mathbb R^n,$ a connectd metric space. Since $f(\mathbb R^n)$ is nonempty, it must be everything. Thus $f$ is onto.