I have begun to study instantons and I have the following difficulty:


I am considering theory with $SU(2)$ gauge group: $S=\frac{1}{2g^{2}}\int \tr F_{\mu\nu}^{2} $.

I've obtained the following expression for topological number $Q$ as a degree of mapping $S^{3}\longrightarrow S^{3}$ (based on differential volume form: $deg(f)=\int_{\Omega} f^{*}\omega$):

$$ Q=\frac{1}{24\pi^{2}}\int d\sigma_{\mu}\epsilon^{\mu\nu\lambda\rho}\tr\left(\omega\partial_{\nu}\omega^{-1}\cdot\omega\partial_{\lambda}\omega^{-1}\cdot\omega\partial_{\rho}\omega^{-1}\right)\tag{1} $$ Here I integrate over the sphere $S^{3}$.

Now I want to show that this expression is equal to the following: $$ Q=-\frac{1}{16\pi^{2}}\int d^{4}x \tr\left(F_{\mu\nu}{F_{\mu\nu}^{*}}\right)\tag{2} $$

The first equation I proved strictly, but for the second one I have no ideas except Stoke's theorem, but i cant do it explicitly.

Could you please explain me how to obtain second expression from the first one?


1 Answer 1


It's Stokes's theorem. Consider a field $F = dA + A \wedge A$ such that $A$ is pure gauge at infinity, that is, $\lim_{x\to\infty} A(x) = \omega\, d \omega^{-1}$ for some $\omega : S^3 \to SU(2) \sim S^3$ where $\omega$ is a function on the 3-sphere because the limit can depend on the direction out to infinity.

In differential forms the first expression is $\newcommand{tr}{\operatorname{tr}}\tr A \wedge A \wedge A.$ Since at infinity, $F = dA + A \wedge A$ vanishes, we can add $0$ in the form of $-3\tr F \wedge A$ to this expression.1

We have $$d \tr (F\wedge A) = d \tr (dA \wedge A + A \wedge A \wedge A) = \tr dA \wedge dA + 3\tr dA \wedge A \wedge A. $$ Now apply Stokes's theorem where we regard infinity as a 3-sphere bounding spacetime. The last part follows from the cyclic property of the trace. Thus we obtain $$\int_{S^3} \tr A \wedge A \wedge A = \int_{S^3} \tr A\wedge A \wedge A - 3 F\wedge A = \int_M \tr \big[3 dA\wedge A \wedge A - 3 dA \wedge dA - 9 dA\wedge A\wedge A\big].$$ Except for the term $A^{\wedge 4} = A\wedge A \wedge A \wedge A$, we are taking the trace of $-3 F\wedge F$. But $A^{\wedge 4}$ is traceless (use the cyclic property of the trace). Hence $$\int_{S^3} A\wedge A \wedge A = -3 \int_M F\wedge F.$$

Now $F\wedge F = 2 F^{\mu\nu} F^*_{\mu\nu}$ because the $F^*_{\mu\nu}$ tensor is usually defined with a $\frac{1}{2}$ factor, so this establishes the relative $-\frac{3}{2}$ in the two expressions for $Q$.

1 This seems pulled out of thin air because it is. I add it because I know what I want to get. Usually, one goes in the opposite direction starting with $F \wedge F = d(dA \wedge A + \frac{2}{3}A\wedge A \wedge A)$. The term we added is the part of this that vanishes at infinity. Putting it back in is a little less natural.

  • $\begingroup$ Thank you for your answer! Also I've learned how to write "trace" in the right way from it ;) $\endgroup$
    – xxxxx
    Oct 6, 2014 at 21:26
  • $\begingroup$ Btw, I think you meant $-9 dA\wedge A\wedge A$ instead of $-9 A\wedge A\wedge A$, right? (: $\endgroup$
    – xxxxx
    Oct 8, 2014 at 19:53
  • $\begingroup$ Yes, that is correct. $\endgroup$ Oct 8, 2014 at 21:12

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.