Dot Physics

Some of the other blog sites have talked about physics vs. chemistry. It seems this started with The experimental Error blog. Tom at Swans on Tea added a very excellent point to the discussion and the discussion continues at Uncertain Principles.

So, here is my take on the subject. Physics essentially deals with the fundamental stuff. You know, Maxwell’s equations, the four forces, the particles, quantum mechanics. Chemistry is the study of substances and their interactions. First, let me attack chemistry. Here are some things I don’t like:

  • Photo electron. What is a photo electron? According to chemists, it is an electron released by interaction with light. We are in the department of chemistry AND physics, so many of the seminars I go to are chemistry oriented. Whenever they mention photoelectron, I lean over to a physicist and say “what’s that?”
  • Chemical potential. Ok, physicist use this also. The problem (in my book) is that chemical potential is just kind of a potential. Well, it’s not directly due to a fundamental force.
  • Van der Waals force. Again, not a fundamental force. Really, this is an electrostatic force.
  • Pressure. What? Ok, wait. Let me explain. What is pressure? I think this is a great example. Pressure is due to interactions with a gas particle, really a whole bunch of gas particles. Really each particle has some type electrostatic interaction with other particles that creates a force and a pressure.

I just wanted to point out some aspects of chemistry that are really somewhat related to fundamental stuff. But here is the point. We need chemistry. Example – take a gas. Suppose you want to model this gas. A fundamental approach (with physics) would look at the gas as a whole bunch of particles. Each particle can have an electromagnetic interaction with other particles. Great, but if you have 1 mole of this stuff, that is not very easy to model. It’s impossible. Even for a computer.

The chemist would say: “it’s not impossible. We used to bulls-eye wamprats back home and they are not much bigger than 2 meters.” The chemisty approach doesn’t always look at each atom in a gas, but instead uses things like pressure and temperature to model it. Or take some complicated molecule. How would you find the vibrational energy levels of this? You could start with shrodinger’s equation and solve this quantum mechanically, but good luck. Really, if it is not hydrogen, you can’t do much without some tricks. So the chemistry way allows use to make up some non-fundamental stuff and use it in a useful way.

One day, there will be no difference between chemistry and physics. As computers become more powerful (but before they take over the world), maybe we will be able to model larger things with fundamental principles. But for now, that is not going to happen. The world needs chemists.


  1. #1 Jaime
    May 20, 2009

    I’m afraid your line of reasoning is severely flawed: Fluid mechanics are not physics? Rigid body motion is not physics? Classical celestial mechanics are not physics? Because according to your classification they aren’t, since none of them takes care of the most fundamental particles known, and their individual interactions.

    My sister, a physicist, had an Einstein quote pinned on a corkboard in her room, that read something like: “All science is either physics, or collecting stamps.” And I think that gets closer to the real difference between them…

  2. #2 Rhett
    May 20, 2009


    You are correct. Fluid mechanics, thermodynamics and stuff could be considered physics. But still, I think that at some point we will be able to model fluids by modeling the individual particles instead of something like the Navier-Stokes equation.

  3. #3 Uncle Al
    May 20, 2009

    If you want an aspirin get a physicist? If you need an aspirin get a chemist! If you want nobody else to have an aspirin get a lawyer.

    William A. Little, Phys. Rev. 134 A1416 (1964), exciton-based ambient temperature superconductors: polyacetylenes substituted with polarizable chromophores, [-C(Ar)=(Ar)C-]n. Such were impossible to synthesize, later physicists argued it couldn’t work, and substituted polyacetylenes are utterly insoluble.

    1) Contemporary synthesis is unremarkable: Aryl aldehyde to the benzoin to the benzil, then methylenation to H2C=C(Ar)-(Ar)C=CH2. Bifunctional monomer is linearly polymerized to high MW with ethylene extrusion by Grubbs’ or Schrock’s acyclic diene metathesis (ADMET). [=C(Ar)-(Ar)C=]n is the result (Little’s puppy with the bracket viewing window shifted by one atom). ADMET living polymerization allows controlled block co-polymerization: high temp supercon diodes obtain form two blocks, or a redox ladder by adding serial blocks. Quick and dirty is a McMurry coupling of the benzoin.

    2) Physical theory applied to complex systems predicts what observation tells it to predict, re proton decay.

    3) Append long hydrocarbon, poly(propylene oxide), or poly(ethylene oxide) mer side chains to obtain non-polar, polar, or water solvent micelle formation, then spin filament re Kevlar.

    4) (Pi-bonds and hydrogens omitted for easy viewing) Look at the pi-stacking in that puppy! Add sidechains for liquid crystal ordering and solubility. It can’t work! Will it work, intrinsic or doped?

    5) This is also obtainable (on paper) as an arbitrarily extended polymer. Append sidechains for solubility. Does it do something physically (rigid rod polymer!), optically (undoped) or conductively (doped) naughty?

    Calculate all you like. You cannot extrapolate outside your model’s assumptions. No physicist can even contemplate making said compositions of matter. The knowing is only vested in the looking. (What physicist can look into a stereogram?)

  4. #4 rob
    May 20, 2009

    need i point out that it was a couple chemists that “discovered” cold fusion?


  5. #5 Excited State
    May 21, 2009

    The quote attributed to Einstein above was actually said by Ernest Rutherford.

  6. #6 Doug Natelson
    May 22, 2009

    I think that at some point we will be able to model fluids by modeling the individual particles instead of something like the Navier-Stokes equation.

    Why would we want to do that? That would be like abandoning the ideal gas law and instead trying to track all the molecular trajectories. It seems like you really don’t like statistical mechanics (and thermodynamics). Stat mech, IMO, turns out to be a phenomenally profound piece of science.

    Physics is all about finding the right language to describe problems at the desired level of approximation. Rather than solving the Schroedinger equation for the many-electron problem of the electronic structure of H2O, we’ve found that it’s reasonable shorthand to assume that the electrons live in molecular orbitals described as linear combinations of hydrogen-like wavefunctions. (That gives us the bond angle.) Chemistry (as in Lewis dot structures) is one level of abstraction that’s proven useful. Continuum mechanics (ignoring the atomic granularity of matter, as in the Navier-Stokes case) is another.

  7. #7 Rhett
    May 23, 2009


    Good point. I do like statistical mechanics, but maybe I just chose a bad example. I really don’t know why someone would model fluid dynamics at the particle level – but Bill Gates also didn’t think anyone would need more than 16 kb of memory in a computer.

    What about at the nanascale level? When do you model something with macro properties and when do you model it at the particle level?

  8. #8 Rob
    May 23, 2009