Multiscale has been a buzzword for such a long time now, most of us must be genuinely tired of it. Nevertheless, when you see actual applications, and the fruits of a lot of hard work come together, I find it still exciting.
A great example I encountered last week is the work by Prof Markus Kraft and his group at Cambridge University’s Chemical Engineering Department. He was over at our Cambridge office for an Accelrys Science and Technology Seminar, talking about soot particles, the black stuff of course that’s actually used to good effect in dyes, and that engineers try and avoid in combustion engines.
The formation of these nano-particles is really and truly a multiscale process. Kraft’s research team starts the long multiscale journey at the quantum level, using DMol3 in Materials Studio to calculate transition states for oxidation reactions of polycyclic aromatic hydrocarbons (PCAH)
This information then enters into rate constant calculations, which then in turn go into Kinetic Monte Carlo simulations (see some cool and funny examples). With KMC you can see the PCAH structures grow. They are then analysed to give input to a population balance model for particles at the next scale, finally entering into engine models.
You can obviously read up the whole story much better in the Kraft group publications. The point here is that it’s a great example of how the different simulation tools through the scales fit together to solve a complex engineering problem.
Developing such a multiscale toolset is what the Nanotechnology Consortium is all about. Already its 14 Members access a module (also tested at Markus Kraft’s lab), to determine rate constants on the basis of transition state calculations. The tool was developed by Struan Robertson, Accelrys’ Simulations group manager. Incidentally he’s just got another great publication out on the topic: “Detailed balance in multiple-well chemical reactions” with guys from Sandia, Argonne, Leeds and Oxford. Great stuff about how you get a handle on calculating rate constants for complex reactions such as in combustion and atmospheric chemistry.
Transition state calculations themselves become more realistic as a result of another Consortium development, i.e. hybrid QM/MM calculations with MS QMERA, based on the well-known ChemShell environment.
In many cases, a detailed understanding of reactive processes, especially at interfaces, is required. The challenge is that quantum methods can only provide a very limited range of dynamics, while forcefield methods cannot adaequately describe reactions.
So we got together with Prof Frauenheim’s group at Bremen University and collaborators to integrate DFTB+ into the Materials Studio toolset .
Last not least of course there is Kinetic Monte Carlo. As in the work by Kraft I described above, KMC really makes the leap in scale, especially time scale, and connects the ‘science into engineering’ world. The Nanotech Consortium is moving forward in this field as well. Watch this space for more on Kinetic Monte Carlo in the Accelrys toolset.
Loading ...