I am trying to generalize the example of the Atiyah-Singer index theorem for the Dirac operator given in the appendix of http://xxx.lanl.gov/pdf/0802.0634v1.pdf.
Please consider the generalized two-dimensional $U(1)$ instanton configuration given by
$$A={\frac {m{n}^{2}{k}^{2}y{\it dx}}{{x}^{2}{k}^{2}+{y}^{2}{n}^{2}+{c}^{2}{n}^{2}{k}^{2}}}-{\frac {m{n}^{2}{k}^{2}x{\it dy}}{{x}^{2}{k}^{2}+{y}^{2}{n}^{2}+{c}^{2}{n}^{2}{k}^{2}}}$$
where $m$, $n$ and $k$ are integers.
The corresponding Yang-Mills field takes the form
$$F= dA = -2\,{\frac {m{n}^{4}{k}^{4}{c}^{2}{\it dx}\wedge {\it dy}}{ \left( {x}^{2}{k}^{2}+{y}^{2}{n}^{2}+{c}^{2}{n}^{2}{k}^{2} \right) ^{2}}}$$
Then, given that $F = \frac{1}{2}F_{ij}dx^i \wedge dx^j$ we have that
$$F = \frac{1}{2}F_{ij}dx^i \wedge dx^j = \frac{1}{2}F_{12}dx \wedge dy +\frac{1}{2}F_{21}dy \wedge dx $$
$$F = \frac{1}{2}F_{12}dx \wedge dy -\frac{1}{2}F_{12}dy \wedge dx $$
$$F = \frac{1}{2}F_{12}dx \wedge dy +\frac{1}{2}F_{12}dx \wedge dy $$
$$F = F_{12}dx \wedge dy $$
and then we obtain
$$ F_{12}dx \wedge dy = -2\,{\frac {m{n}^{4}{k}^{4}{c}^{2}{\it dx}\wedge {\it dy}}{ \left( {x}^{2}{k}^{2}+{y}^{2}{n}^{2}+{c}^{2}{n}^{2}{k}^{2} \right) ^{2}}}$$
which implies that
$$ F_{12} = -2\,{\frac {m{n}^{4}{k}^{4}{c}^{2}}{ \left( {x}^{2}{k}^{2}+{y}^{2}{n}^
{2}+{c}^{2}{n}^{2}{k}^{2} \right) ^{2}}}$$
The instantonic number for such configuration is
$$-\frac{1}{2\pi}\int_{\mathbb{R}^2}F=-\frac{1}{2\pi}\int_{\mathbb{R}^2}F_{12}dxdy$$
which is reduced to
$$-\frac{1}{2\pi}\int_{\mathbb{R}^2}F=-\frac{1}{2\pi}\int _{-\infty }^{\infty }\int _{-\infty }^{\infty } -2\,{\frac {m{n}^{4}{k}^{4}{c}^{2}}{ \left( {x}^{2}{k}^{2}+{y}^{2}{n}^{2}+{c}^{2}{n}^{2}{k}^{2} \right) ^{2}}}dxdy$$
and the computation gives
$$-\frac{1}{2\pi}\int_{\mathbb{R}^2}F= mkn$$
Now, the Dirac operator in such background gauge field is given by
$$\not{D}=\left[ \begin {array}{cc} 0&\partial_{{x}}+i{\partial}_{{y}}-{\frac {im{n}^{2}{k}^{2}y+m{n}^{2}{k}^{2}x}{{x}^{2}{k}^{2}+{y}^{2}{n}^{2}+{c}^{
2}{n}^{2}{k}^{2}}}\\\partial_{{x}}-i{\partial}_{{y}}-
{\frac {im{n}^{2}{k}^{2}y-m{n}^{2}{k}^{2}x}{{x}^{2}{k}^{2}+{y}^{2}{n}^
{2}+{c}^{2}{n}^{2}{k}^{2}}}&0\end {array} \right]$$
A positive chirality zero-mode satisfies
$$\not{D}\left[ \begin {array}{c} \chi \left( x,y \right) \\0\end {array} \right]=0$$
it is to say
with the explicit form
These solutions are square normalizable and then there are $mnk$ linearly independent zero-modes of positive chirality.
On the other hand, a negative chirality zero-mode satisfies
$$\not{D}\left[ \begin {array}{c} 0 \\ \eta \left( x,y \right) \end {array} \right]=0$$
it is to say
with the explicit form
These solutions are not square normalizable and then there are $0$ linearly independent zero-modes of negative chirality.
Then, we obtain that the index of $\not{D}$ , which is the number of linearly independent normalizable
positive chirality zero-modes minus the number of linearly independent normalizable negative chirality zero-modes is $mnk-0 = mnk$. Formally, the index theorem in this case reads:
$$index(i\not{D}) = mnk = -\frac{1}{2\pi}\int_{\mathbb{R}^2}F= -\frac{1}{4\pi}\int_{\mathbb{R}^2}\epsilon^{\mu\nu}F_{\mu\nu} $$
My questions are:
1. Do you agree with the computations in this example?
2. Do you know other example in two-dimensions?