Large-scale simulations of diffusion-limited n-species annihilation

We present results from computer simulations for diffusion-limited n-species annihilation, A(i)+A(j)-->0 (i,j=1,2, em leader,n;i not equal j), on the line, for lattices comprising of up to 2(28) sites, and where the process proceeds to completion (no further reactions possible), involving up to 10(15) time steps. These enormous simulations are made possible by the renormalized reaction-cell method. Our results suggest that the concentration decay exponent for n species is alpha(n)=(n-1)/2n instead of (2n-3)/(4n-4), as previously believed, and are in agreement with recent theoretical arguments of Deloubrière et al. We also propose an expression for Delta, the correction-to-scaling exponent for the concentration decay, defined by c(t) approximately t(-alpha)(A+Bt-Delta).