I suggest, you go through the derivation - and point out which step will not be valid for complex numbers. It is valid for sin(Θ), cos(Θ), ln(x), ex - why not for complex coefficients.
(a + b)2 = a2 + 2ab + b2
is valid for all a & b - real and complex. You can prove that by using FOIL.
Thus I do not see any step in derivation that will exclude complex coefficients.
Exactly, nothing in the standard proof uses a property of the real numbers which does not hold for complex numbers (there are some properties for elements in R which do not hold in C).
Here is the proof I mean, assume a is not 0 then:
\(\displaystyle ax^2+bx+c \,\,=\,\, 0 \iff x^2 + \frac{b}{a}x \,\,=\,\, -\frac{c}{a}\,\, \iff \)
\(\displaystyle x^2 + \frac{b}{a}x + \frac{b^2}{4a^2} \,\,=\,\, \frac{-4ac + b^2}{4a^2} \,\,\iff \)
\(\displaystyle (x+\frac{b}{2a})^2 \,\,= \,\,\frac{-4ac + b^2}{4a^2} \,\,\iff\)
\(\displaystyle x+\frac{b}{2a} \,\,= \,\,\pm \frac{\sqrt{-4ac + b^2}}{2a}\)
Then finish from there.
edit: I'm not so sure that simplifying the answer in the complex case is an easy task. It might be that the other method could be easier.