By means of ab initio MD.

Hello ma'am,

can you please elaborate your answer.

Looking for the global energy minimum is hard since there are many possibilities (local minimum). Thus, this problem could be solved if you run a molecular dynamic simulation in your system and submit some of the screenshots to an ab initio energy's optimization.

I agree with Jose Xavier Lima Neto. Obtaining global minima of the cluster is somewhat tricky. You could attain the local minimum using molecular dynamics. In dynamics, you will get number of user defined snapshots say 1000. You can pick one of them depending on your needs and calculate ab initio energy.

Dear Keshav,

I think to find the global minima of cluster you need to use different package along with Gaussian. Good options are CALYPSO (http://www.calypso.cn/) or ABCluster (http://www.zhjun-sci.com/software-abcluster-download.php) by Prof. Jun Zhang. Both this package will generate all possible structures based on Swarm Based structure prediction technique and each structure will be optimized by Gaussian09 . At the end you'll get a list of all possible isomers with their energy. Choose the lowest one.

For your structure, I think the "LEGO" scheme of ABCluster will be suitable.

Try a Genetic Algorithm (see for example 10.1016/j.chemphys.2005.10.034) or you can write your own code for shuffling the hydrogen atoms for a given oxygen framework.

Thank you Viorel Chihaia , José Xavier Lima Neto Karthick Thangavel Turbasu Sengupta for your informative answer. I am not good in programming so I cannot write my own code. However I have downloaded ABcluster and learning it.