[R] optimize in very small values

[Troels Ring]

I hope you will forgive me this simple question on titration.
I'm trying to find very small values from the algorithm below, which I 
believe is
correctly formatted, and the constants are also correct. When SID goes over 
ATOT, fitted vales are
much too low compared to the literature. I guess I must be using optimize 
in a wrong way but cannot find out how to improve it. I'm on windows, R 
2.0. Best wishes
Troels Ring, Aalborg

ATOT <- 0.019
KA <- 3e-7
KW <- 4.4e-14
SID <- 0.01

ff <- function(H,KA,SID,ATOT)
{H^3 + (KA+SID)*H^2+(KA*(SID-ATOT)-KW)*H- KA*KW}
-log10(optimize(ff,c(0,.1),tol=.Machine$double.eps,KA=KA,SID=SID,ATOT=ATOT)$minimum)