Conclusion

We were able to reproduce the experimental behavior in the MD simulation, i.e. we observed cluster formation and growth. Unfortunately we were not able to perform the runs with the Nosé-Hoover thermostat and pressure control, yet. It would be particular interesting to study the system in the case of a true isobaric-isenthalpic ($N$,$H$,$P$) or isothermal-isobaric ($N$,$T$,$P$) ensemble. This would allow us for instance to calculate the work of expansion during cluster formation and growth. Due to constraints in time and computing power we had to restrict ourselves to a relatively small system. Especially to improve the statistics the simulation of a bigger system would be favorable.

The comparison between simulations and experiment is difficult in this case. Experiments are performed at pressures of the order $10^{-3}$kbar while simulations work with pressures tens of orders bigger.

We would like to thank Robert Averback, Yinon Ashkenazy and Pavel Krasnochtchekov for their help and support.