<div dir="ltr">Hello, I have a question related to the NUMBER of REPLICAS and NPROC_REP I thought the way that the NEB calculations worked was that I could provide initial and final geometries and then any other geometries that I request will be formed with interpolation. This is done through the number of replicas correct? I was trying to look at co2 dissociation on a Cu55 cluster. So I specified 4 replicas and provided my initial and final geometries as two of the 4 replicas in the &REPLICA section. Then I specified the number of NPROC_REP to be 16 and so in total I used 64 cores when I submitted the calculation. When I ran this calculation I received 4 neb1-pos-Replica-nr_1-1.xyz files back. Files 1 and 4 were what I believe were the initial geometries I provided. However in files 2 and 3 there was a significant change in the morphology of the Cu55 cluster, (it lost its icosahedral symmetry). Is this significant change a result of the interpolation process? if so is there a way prevent such a change? I did not observe significant changes in the cluster when I optimized a geometry of CO adsorbed on Cu55 so I wasn't expecting the cluster to change shape. I've provided the input file i used in the 4 replicas. Thanks, Natalie </div>