The Computing for Clean Water project was created to provide deeper insight on the molecular scale flow of water through a novel class of filter materials. Thanks to the millions of virtual experiments that the team was able to run on World Community Grid, they discovered conditions under which water can pass through tiny carbon nanotubes much more efficiently. This groundbreaking understanding of a fundamental physical process could help improve access to clean water for millions of people through more efficient water filtration and desalination, and also may have applications in clean energy and medicine.
The Value of Independent Verification
Computing for Clean Water finished 2017 on a high note, with a follow-up pair of publications [1,2] inspired by our original Nature Nanotechnology paper , which used data made possible through the efforts of volunteers.
The story behind these articles illustrates an important point in science: the value of independently verifying new results. In this case, an international team with lead author Eduardo Cruz-Chú at ETH Zurich was inspired by our results to do a series of complementary simulations. The team used a somewhat different model of the water flow, and also focused on the diffusion of oxygen atoms in the water.
These authors reproduced the main finding of our article, namely the positive impact of phonons (the vibrations of the carbon nanotube atoms induced by thermal energy) on the diffusion of water in nanotubes, and the implications this has for ways to optimize such diffusion through nanotube arrays.
These authors did, however, obtain a smaller diffusion enhancement using their model than what we had reported in our study. In the field of molecular dynamics simulations, it is quite common to see some variations depending on the details of the models used. So, we did a series of further simulations to test the robustness of our original results. What we found is that the effect of phonons on the water diffusion is always large compared with a phonon-free calculation, even allowing for considerable variation in some of the parameters used in our model.
Differences between Studies
A significant difference between the two studies concerns the type of diffusion that is being monitored – we only considered water molecules, whereas our colleagues studied also the diffusion of oxygen atoms. Their results suggest that the diffusion of other molecules or ions will be different. This difference is something that we hope to study in future, since it has implications for how effective nanotubes can be in filtering out unwanted molecules and ions, for example salt ions from seawater.
While it is great to see the main insight from our World Community Grid study validated in this new study, and corroborated by our further simulations, the two new articles are also an important reminder that experimental techniques still need to be developed to study the flow of water in individual nanotubes. In the end, the ultimate arbiter of the importance of simulated results like ours will be hard experimental data. You can read our detailed response to the new article here.
In the meantime, we thank all World Community Grid participants in Computing for Clean Water for helping to obtain the original results, which are clearly getting the attention of the scientific community.
 Eduardo R. Cruz-Chú et al., Nat. Nanotech. 12, 1106–1108 (2017).
 Ming Ma et al., Nat. Nanotech. 12, 1108 (2015).
 Ming Ma et al., Nat. Nanotech. 10, 692–695 (2015).