Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Polymer modelling

  1. Feb 26, 2009 #1
    Hi PF,

    I'm working on a semi-classical model for a group of rather strange polymers based on carbon, nitrogen, and oxygen. In this treatment, I'm approximating the bond length and bond angle to be more or less classical (ie. ball and stick). But I would also like to treat the dihedral angles quantum mechanically. But quantum mechanically, I mean that the rotations aobut dihedral angles can simultaneously be in different states, governed by some probability density p(theta). So I want to subject the model to the following experiment:

    Initial environment:

    Low pH, the polymer is a linear chain and unstable (ie. there are a huge degree of accessible microstates)

    Final environment:

    Neutral pH, the polymer is compact (ie. one or two microstates represent the ensemble average far better than all other microstates)

    Three things:
    1) I want to know the pathway, ie. the conformations that my model samples through when the environment changes from initial to final.
    2) I want to know the details of p(theta), ie. the probability density function for each angle.
    3) I want to know what physical observables I can derive from this model that I can match with experiments (eg. NMR, Small Angle Xray Scattering, etc.)

    And most importantly, is it even OK to combine classical and QM treatments like this. Also, when do quantum effects vanish? IE. I know it's generally restricted to things on a planck's constant, but then people started reporting things like buckyball having interference effects.
  2. jcsd
  3. Feb 27, 2009 #2


    User Avatar
    Science Advisor

    In short, no, you can't combine quantum mechanical and 'classical' treatments like that.
    The energy isn't separable into dihedral/angle/bond distance components classically even.

    There are QM/MM methods that combine both types of models, but then they do so for discrete parts of the molecule (and the 'cut' needs to be carefully chosen). What is it you want to do that can't be done with conventional force-field methods, such as http://ambermd.org/" [Broken]? They're used for polymers all the time.
    Last edited by a moderator: May 4, 2017
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook