The Naked Scientists

The Naked Scientists Forum

Author Topic: Predicting the color of substances?  (Read 6664 times)

Offline Supercryptid

  • Hero Member
  • *****
  • Posts: 606
    • View Profile
    • http://www.angelfire.com/sc2/Trunko
Predicting the color of substances?
« on: 19/01/2005 03:50:53 »
Every material absorbs light in a particular fashion. This is what gives materials their specific colors. The wavelength of light that is absorbed is determined by the electron shells/energy levels of the molecules that make up that substance. Therefore, shouldn't the color of a smooth surface of a substance be directly determined by its molecular structure?

I know some materials get their colors from minor facets or indentations along their surface (butterfly wings, for instance), and that some substances even have a different color when they are divided up as a powder (gold is shiny yellow as a solid mass, but when it is ground into a powder, it looks purple).

Taking this aside, let's assume that the substance in question has a smooth surface. Shouldn't it be theoretically predictable what color the substance will be, since the molecular structure would directly determine its color? There should be a way to predict a substances color based on the orbitals and energy levels in its molecules, right?

I've never heard of any equations that predict this. Do any exist? Some trends have been discovered (conjugated alkenes tend to have certain absorption patterns based on their pi bonds), but I've not heard of such things for molecules without pi bonds (like nitrogen triiodide).

Is there any computer software that can predict a substances color in this way? If so, I'd really like to have it. For instance, if I were to draw the azulene molecule (C10H8) in the program, it should be able to tell me what wavelengths of light it absorbs, and from there tell me that it is blue.

Any ideas on this?


 

Offline Ylide

  • Moderator
  • Hero Member
  • *****
  • Posts: 905
    • View Profile
    • http://clem.mscd.edu/~mogavero
Re: Predicting the color of substances?
« Reply #1 on: 19/01/2005 20:27:18 »
I don't know about software, but if you know the molecular structure of a conjugated alkene you can predict the wavelength of color that will be absorbed using quantum mechanics.  I did this last year in a Phys. Chem. experiment with various dyes.  It's a simple particle-in-the-box problem using the net bond lengths across the conjugated system of the molecule.  

As for more simple molecules or atoms, I'm sure there's a way to predict their color as well by the energies and orbitals of the valence electrons, I'm just not recalling exactly how.  



This message brought to you by The Council of People Who Are Sick of Seeing More People
 

Offline anthony

  • Full Member
  • ***
  • Posts: 65
    • View Profile
Re: Predicting the color of substances?
« Reply #2 on: 21/05/2005 15:01:00 »
Yep, it is possible to arrive at a colour for most materials based on first principles and quantum mechanics, the Hamiltonian Equation. Most of these models are complex using a mixture of theoretical and numerical approaches, so the people using them aren't interested in anything as trivial as colour, but these models generate the energy levels for the material and hence, by the differences between the energy levels, it would be possible to predict colour.
 

Offline rosy

  • Neilep Level Member
  • ******
  • Posts: 1018
  • Chemistry
    • View Profile
Re: Predicting the color of substances?
« Reply #3 on: 21/05/2005 15:41:02 »
quote:
Originally posted by Ylide

I don't know about software, but if you know the molecular structure of a conjugated alkene you can predict the wavelength of color that will be absorbed using quantum mechanics.  I did this last year in a Phys. Chem. experiment with various dyes.  It's a simple particle-in-the-box problem using the net bond lengths across the conjugated system of the molecule.  

As for more simple molecules or atoms, I'm sure there's a way to predict their color as well by the energies and orbitals of the valence electrons, I'm just not recalling exactly how.  



Hm, I, too, did this sort of experiment in a chemistry practical in the last year... particle-in-a-box calculations are pretty approximate (not really a great model of the conjugated system) and I suspect you'd have trouble predicting what colour they'd be exactly (red or orange, or green or blue, or whatever).

Also, you can make some predictions about the colours of compounds/complexes of transition metals based on what ligands are complexing the metal atom and how strongly they're going to interact with the d-orbitals of the metal. But I think that mainly it's a case of measuring and rationalising spectra rather than actually making predictions.

Basically colour of a material is usually dependent on the transfer of an electron between molecular orbitals and the colour of light absorbed therefore depends on the energy gap between those orbitals. To predict colours we therefore have to know the energy gap between the orbitals and MO calculations are complicated and involve heavy computing.
 

Offline chimera

  • Sr. Member
  • ****
  • Posts: 475
    • View Profile
Re: Predicting the color of substances?
« Reply #4 on: 21/05/2005 16:03:35 »
quote:
Originally posted by Supercryptid
Is there any computer software that can predict a substances color in this way? If so, I'd really like to have it.



Wow, that's one cool question topic. Found this, called Chembrain, hope it's any use:

'Nevertheless, not only can it store, collate, search and retrieve molecules, as well as predicting the properties of unknowns, PiSystems allows the user to generate their electronic spectra and work out their light absorption properties, i.e. the colour of organic molecules.'
from:

http://www.scientific-computing.com/scwnovdec03chembrain.html

The living are the dead on holiday.  -- Maurice de Maeterlinck (1862-1949)
 

Offline Supercryptid

  • Hero Member
  • *****
  • Posts: 606
    • View Profile
    • http://www.angelfire.com/sc2/Trunko
Re: Predicting the color of substances?
« Reply #5 on: 11/06/2005 20:13:23 »
Ah, yes. I've been looking for that. Thanks for pointing it out to me!

PiSystems correctly predicted that benzene (C6H6) is colorless, but it also predicted that coronene (C24H12) is colorless too, whereas it is actually yellowish. I supposed that might be due to imperfections in crystal structure, or impurities, though. I think it is supposed to predict a molecule's color in solution anyway, and not necessarily its solid-state color. Strangely, it gives a large number of colors for colored molecules, so I'm not sure what color they are trying to tell me the substance is.
 

The Naked Scientists Forum

Re: Predicting the color of substances?
« Reply #5 on: 11/06/2005 20:13:23 »

 

SMF 2.0.10 | SMF © 2015, Simple Machines
SMFAds for Free Forums