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

Gas Of Dipoles

  1. Aug 19, 2007 #1
    Suppose we have a gas of randomly oriented (and in random motion) electric dipoles. Obviously the dipoles do not behave as free particles. How do we describe it's properties? Can we define potential etc. ?
    I am looking for a solution without the large volume approximation. Is it possible to get an expression for the mean kinetic energy etc. of a partical given its total energy?
  2. jcsd
  3. Aug 19, 2007 #2
    will get some answers here shreyas.
    though i strictly beleive this isseroius mixture of eveidence and thermodynamic arguements
  4. Aug 19, 2007 #3


    User Avatar

    Staff: Mentor

  5. Aug 19, 2007 #4
    you are talking about a truly many-body problem, where you must (usually) include the induced dipole-induced dipole interactions. if they don't matter, then why bother?

    there are ways of doing this numerically, but they are difficult. you can start with an analytic expression by considering a dipole, mu=dq, interacting with a point charge, you'll then get

    \nabla_\alpha \nabla_\beta \frac{1}{r}

    as the induced field contribution, which needs to be solved for numerically for anything but the simplest of systems. thats why most molecular simulation techniques ignore induced dipoles.
    Last edited: Aug 19, 2007
  6. Aug 20, 2007 #5
    Could you clarify that a bit? I am not used to the standard notations. In my original attempt, I calculated the potential energy of two interacting dipoles in terms of their spacing and orientation. I want to average this out in some way (integrating over the angle obviosly gives net potential energy 0)

    Thanks for the help.
  7. Aug 20, 2007 #6
    If you're looking for a simple solution, the only viable one is to treat it like a semi-ideal gas. The dipoles have translational, rotational and vibrational degrees of freedom, each of which is 1/2 kT per molecule in this simple approximation.
  8. Aug 20, 2007 #7
    the potential energy of interaction between a dipole and a point charge (if you draw out the two charges seperated) that is far away is [tex]\nabla_a (\frac{1}{r})[/tex]

    since [tex]E = -\nabla V[/tex] then the dipoles contribution to the induced field is the expression i gave.

    do a scholar google search for "molecular polarization" if you are more interested. there are review articles out there that summarize the field. also, Jackson's E&M book may interest you.
  9. Sep 4, 2007 #8
    I found in this paper(page no. 4)

    Shouldn't there be a term to take care of direction, [tex]U=\frac{\mu_r^2\cos\alpha}{4\pi \epsilon_0}[/tex]?, [tex]\alpha[/tex] is angle between field and the axis
  10. Sep 4, 2007 #9
    You can also consider a mean field approximation -- that will allow you to calculate the potential energy due to the orientation of the dipoles. In fact, with that approximation, I guess you'd get a hybrid free-gas / Curie ferromagnet effect, with a phase change due to the orientation. As far as dipole-dipole forces beyond merely torque, perhaps use the Van de Waals approximation?
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?

Similar Discussions: Gas Of Dipoles
  1. Electric dipole (Replies: 6)

  2. Magnetic dipole (Replies: 5)

  3. Dipoles and heat (Replies: 15)

  4. Dipole radiation (Replies: 1)