Center atoms in box

[Ray Matsumoto]

The carbon atoms should be centered inside the box, particularly in the direction normal to the pore surface. The current implementation works for fixed N simulations, but will cause problems particularly for MC simulations with insertion/ deletion moves.