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

Derivation of E=mc2

  1. Oct 28, 2009 #1
    I have looked at the Einsteins Derivation of the Relation E=mc2.

    But I don't if this shorthand derivation is O.k. or not.
    like the derivation, I use maxwells law and say the momentum p is given by
    p = E / c ---> 1

    Now from newtonian mechanics, we have
    p= m* v
    or, p=mc (here v=c) --->2

    Combining 1 and 2
    mc = E / c
    therfore, E = mc2

    What's wrong here?
    If its wrong, why is the result correct?
    If its correct, why didn't Einstein use it?
  2. jcsd
  3. Oct 29, 2009 #2

    Well, for one, photon mass is zero!

    Einstein's point was to figure out how much a photon would shift the center of mass of the box and then ask how big a massive object would have to be to have a similar effect.
  4. Oct 29, 2009 #3


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    Here you're using the formula [tex]E^2=\vec p^2c^2+m^2c^4[/tex] with m=0. Nothing wrong with that.

    As Clamtrox mentioned, m=0, so this formula doesn't work. The correct formula for a massive particle (which has [itex]|\vec v|<c[/itex]) is [itex]\vec p=\gamma m \vec v[/itex]. If you square both sides and do a little algebra, you recover the formula I posted above, which actually holds for massless particles too. So instead of p=mc, you have [tex]\vec p^2=E^2/c^2-m^2c^2=E^2/c^2[/tex].

    If you combine 1 and the correct version of 2 (which is actually the same formula as 1), all you get is


    which tells you exactly nothing.

    See this post for a better derivation.
  5. Oct 29, 2009 #4
    I question if this can really be rigorously derived. In relativity, mass and rest energy are exactly the same thing, the reason why there is a c^2 is simply because we choose to use inconsistent units for the same physical quantity.

    The nontrivial statement is really that (E, p) transforms under Lorentz transformations as a four-vector.
  6. Oct 29, 2009 #5
    Actually, the fact that dx^{mu}/dtau transforms as a four-vector is trivial. So, the non-trivial statement is that four-momentum is conserved.
  7. Oct 29, 2009 #6
    Four-momentum is conserved because the laws of physics are invariant under arbitrary space-time translations. So, the whole matter is trivial.

    Then, given the fact that the heuristic derivation (that use intuitive concepts from classical mechanics in some ways) looks non-trivial, suggests that the derivation of classical mechanics from relativity is less trivial than is presented in most textbooks.

    In classical mechanics, mass is an independent physical quantity from energy, while in relativity mass and (rest) energy are the same thing. A derivation of classical equations like
    E_kin = 1/2 m v^2 is then not just a simple matter of doing a Taylor expansion around v = 0.

    The non-relativistic limit involves rescaling physical variables in a certain way and then studying the infinite scaling limit. In this limit, the (trivial) relation between mass and energy breaks down (it becomes singular). So, you then have a new independent physical quantity.
  8. Oct 29, 2009 #7


    User Avatar
    Staff Emeritus
    Science Advisor
    Gold Member

    I agree. It isn't possible to derive [itex]E=mc^2[/itex] without making some assumptions. Some of the details about this didn't become 100% clear to me until I wrote about it the Science Advisor forum (hidden for normal users). I quoted myself from that thread here. (Start reading at "How are you going to...").
    Last edited: Oct 29, 2009
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook