Here's a proof of the general case that I got from Courant's Introduction to Calculus and Analysis. This is one of my favourite proofs:
"If a function y = f(x) satisfies an equation of the form [tex]y' = \alpha y[/tex] where [tex]\alpha[/tex] is a constant, then y has the form [tex]y = f(x) = ce^{\alpha x}[/tex] where c is also a consant; conversely, every function of the form [tex]ce^{\alpha x}[/tex] satisfies the equation [tex]y' = \alpha y[/tex].
It is clear that [tex]y = ce^{\alpha x}[/tex] satisfies this equation for any arbitrary constant c. Conversely, no other function satisfies the differential equation [tex]y' - \alpha y = =[/tex]. For if y is such a function, we consider the function [tex]u = ye^{-\alpha x}[/tex]. We then have
[tex]u' = y'e^{-\alpha x} - \alpha y e^{-\alpha x} = e^{-\alpha x}(y' - \alpha y}) .[/tex]
However, the right-hand side vanishes, since we have assumed that [tex]y' = \alpha y[/tex]; hence [tex]u' = 0[/tex] so that u is a constant c and [tex]y = ce^{\alpha x}[/tex] as we wished to prove."