I have this problem: Probe that $\varphi:\Bbb R^3-[0] \to \Bbb R^3, \varphi=\frac{(x,y,z)}{\vert\vert(x,y,z)\vert\vert^3}$ verify that $\nabla \cdot\varphi=0$ but does not exist $G:\Bbb R^3\to \Bbb R^3$, of class $\mathcal C^1$ such that $\nabla \times G=\varphi$
It's easy to prove that $\nabla \cdot\varphi=0$ but I have a problem with the part of there is no $G$ such that $\nabla \times G=\varphi$, if is $G=(G1,G2,G3)$ I have this conditions:
$\frac{x}{\vert\vert(x,y,z)\vert\vert^3}= \frac{\partial G3}{\partial y}-\frac{\partial G2}{\partial z}$
$\frac{y}{\vert\vert(x,y,z)\vert\vert^3}= \frac{\partial G3}{\partial x}-\frac{\partial G1}{\partial z}$
$\frac{z}{\vert\vert(x,y,z)\vert\vert^3}= \frac{\partial G2}{\partial x}-\frac{\partial G1}{\partial y}$
And with $\nabla \cdot\varphi=0$ if the lateral derivatives of $G$ coincide I can find some $G$, so something is wrong. Has something to do with the fact that $\varphi$ is not $\mathcal C^1$?
If $\varphi = \nabla \times G$ on $\mathbb{R}^3 - \{0\}$ where $G \in C^1(\mathbb{R}^3)$, then by the divergence theorem, since $\nabla \cdot (\nabla \times G) = 0,$
$$\int_{\partial B_R(0)} \varphi \cdot \mathbf{n} \,dS = \int_{\partial B_R(0)} \nabla \times G \cdot \mathbf{n} \,dS = \int_{B_R(0)} \nabla \cdot (\nabla \times G)\, dV = 0$$
where $B_R(0)$ is the closed ball of radius $R$ centered at the origin.
However,
$$\int_{\partial B_R(0)} \varphi \cdot \mathbf{n} \,dS = \int_0^{2\pi}\int_0^\pi\frac{R\mathbf{e_r}}{R^3} \cdot \mathbf{e}_rR^2\sin \theta \,d\theta \, d\phi = 4\pi$$