Is it true that $R=k[x,y]$ is a free $R$-module ?
I think that it isn't true. Natural candidate for the base is $\{x^{\alpha}y^{\beta}\}_{\alpha,\beta}$, but :
$x\cdot (xy) +(-y)\cdot x^2 =0$
and $x\neq 0,\ -y \neq 0$.
Is it true that $R=k[x,y]$ is a free $R$-module ?
I think that it isn't true. Natural candidate for the base is $\{x^{\alpha}y^{\beta}\}_{\alpha,\beta}$, but :
$x\cdot (xy) +(-y)\cdot x^2 =0$
and $x\neq 0,\ -y \neq 0$.
You have confused $k$-vector space basis as $R$-module basis, hence this issue.
ANy commutative ring is a free module overitself with basis the singleton $\{1\}$, consiting of the multiplicative identity.
Given two elements $a,b\in $R, by commutativity we have $a.b -b.a=0$ and so a basis has to have less than 2 elements.