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

Why is there hybridization of orbitals?

  1. Dec 9, 2009 #1
    Hi, we often learn that orbital hybridization helps us in explaining the bond angle and bond length for simple molecules, e.g. CH4.

    But why is it that s and p orbitals undergo hybridization? Is it because the resulting hybridized orbitals are more stable?
     
  2. jcsd
  3. Dec 9, 2009 #2

    DrDu

    User Avatar
    Science Advisor

    Hybridization is a concept peculiar to valence bond theory. The two electron bonds will be the stronger the higher the overlap between the orbitals. In Methane the overlap is much stronger between the s orbitals on the hydrogens and the sp3 hybrids on the carbon than between the s orbitals on hydrogen and either a p or an s orbital.
    It is not true that the hybridization explains the bond angles etc. Rather both the geometry and the hybridization will adapt in such a way as to lead to the most stable description of the molecule. Btw, there are not only sp, sp2 and sp3 hybrids, as often told in oversimplified introductory textbooks. Rather, one can have continous spx with x ranging from 0 even up to 6 ( the last value being usefull e.g. in the description of SF6).
     
  4. Dec 9, 2009 #3
    Yup, I know about the basics of hybridization and the VSEPR theory and stuff.

    But the question is why does hybridization occur? Is it because of the lowering of energy of the chemical system?
     
  5. Dec 9, 2009 #4

    alxm

    User Avatar
    Science Advisor

    There's only one true description of the system, which is the actual exact wave function. Which, as we all know, isn't an easy thing to calculate for many-electron systems. So pretty quickly they got to work on finding models based on QM that could describe (if only qualitatively) chemical bonding. (after all, they didn't have computers with which to get the exact solution!)

    The result was Valence-Bond theory from Pauling & Co, and Molecular Orbital theory from Mulliken & Co. Those two models have existed side-by-side ever since, since they both do their job reasonably well in most cases. (One case where they don't work, the beryllium dimer, finally got some solid experimental results in the latest issue of Science, and earlier this year. A very important piece of data to test methods against!)

    So it's a model, not a physical process. It doesn't 'occur'.

    Although I guess you could say that hybridization "occured" when Pauling and Slater figured out that chemical bonding energy could be estimated and bonding described through the spatial overlaps of single-atom wave functions. Around the mid-late 1930's. (I'll never cease to be amazed at the speed at which quantum theory transformed the theory of chemistry from a handful of ad-hoc empirical rules to a rigorously derived branch of applied physics)

    I'm sure this hand-waving description I just gave (and the ones in most chemistry textbooks) is giving some of the physicists here jitters. Rest assured it doesn't really do VBT justice; there is a more proper mathematical derivation - and in fact, been put on a such a solid ground that it's now ('modern' VBT) becoming viable as a quantitative method. But for those curious about that I'll refer to Shaik's books and publications.
     
  6. Dec 9, 2009 #5
    Sorry but I disagree ....It does explain bond angles.Think of any three sp3, sp2 or sp hybridised compounds and compare their angles.

    yep...hence it determines the angle! Most variations in similar hybridisations( e.g sp3's) occur as a result of the presence of lone pairs plus other factors...

    Anyone with Me? hehe
     
  7. Dec 9, 2009 #6
    I think so.
     
  8. Dec 9, 2009 #7
    In addition...you may want to check any chemistry forum becuase this is what they do(from the moment they begin their university degrees)...reactions and hybridisations. It may help.
    Cheers.
     
  9. Dec 9, 2009 #8

    DrDu

    User Avatar
    Science Advisor

    Dear mccoy1,
    there simply are no such things as sp3, sp2 or sp hybridised compounds. I can describe even e.g. ethylene as well with sp3 banana bonds as with sp2 hybrids.
     
  10. Dec 9, 2009 #9
    okayyy mate.
     
  11. Dec 9, 2009 #10
    I'm a chemistry graduate. And yes, I sent the same question to a chemistry forum at the same time. But I am curious as to what chemists and physicists think of the concept of orbitals hybridization. :)
     
  12. Dec 9, 2009 #11
    Well bond description sits in the heart of chemistry....you can't get any better understanding anywhere than 'there'.
    And to be honest with you, we bash chemists and biologists here. You may have seen the evident here and elsewhere in the forum.
    By the way, I'm wondering why a chemistry graduate has to ask this question! I'm BS undergraduate and i seem to have a solid understanding of how bonding occurs.
    For chemistry graduate student to ask 'this'(hybridisation) question is just like a physics graduate asking whether operators concerned with position and momentum commute.
    Anyway.......
     
  13. Dec 9, 2009 #12

    DrDu

    User Avatar
    Science Advisor

    Dear opuktun,

    if you seem to be already well trained you should have a look at the book by Roy McWeeny "Methods of molecular quantum mechanics", maybe the only modern textbook on quantum chemistry which describes also VB theory and its modern guises thoroughly.
    I did VB calculations myself long time ago. A kind of hybridization, which is not mentioned in introductory texts, but is extremely usefull is the hybridization of e.g. the sp3 orbitals on C with a little bit of the s orbitals on the adjacent H atoms in case of Methane. Like this, you can avoid most of the ionic structures in basically covalent compounds.
     
  14. Dec 9, 2009 #13
    Yes, the concept of bonding is easy to understand - energetically driven to reach the stable state, i.e. one with lowest energy. But I felt there was this gap in my understanding of the chemical modeling and quantum mechanic calculation. That is why I think physicists would be best in enlightening my doubts and hence my question here.

    I don't think bashing followers of different scientific discipline is appropriate. Afterall, science is the rationalization of the physical world. In my opinion, if one is narrow minded enough to do so, he is just a fool who fails to see the different perspectives of the world.

    If this question is inappropriate, please feel free to close the thread.

    Thanks for the recommendation. I would check out the book tomorrow.
     
  15. Dec 9, 2009 #14

    alxm

    User Avatar
    Science Advisor

    Not at all. It's true that basic VB and MO theory are taught in first-year chemistry. But it's also true that most chemists never delve much deeper into the topic of bonding than that. Some chemists know more than others, but most know relatively little apart from the aforementioned models. (as well as perhaps Hückel and VSEPR) It's it's own set of (heavily overlapping) fields and those fields are named: Physical chemistry, theoretical chemistry, quantum chemistry, atomic and molecular physics and chemical physics.

    Any chemical physicist is far better qualified to answer a question on bonding than your average chemist.

    I haven't seen that, no. And that's just stupid amateurish group-identity nonsense.
    I wouldn't. I've got an M.S. in physical chemistry and a PhD in chemical physics. I do quantum chemistry and roughly half the department started out as chemists and half as physicists. If anyone is, we're the experts on bonding and orbitals, and we certainly don't make any distinction. Chemistry does not 'own' bonding, nor does any of those other aforementioned fields have a monopoly on their subject.

    And just to add yet another dimension, there's the distinction between solid-state QM and molecular QM. No knowledgeable person would agree with the superficial idea that the Tight-Binding model is certainly "solid-state physics" but Valence Bond theory is "chemistry" - They're essentially the same thing.

    Based on your responses in this thread, I can't agree with that.
    Stating "Two orbitals overlap and that lowers the total energy" is not understanding. It's regurgitating a finished result you were taught. If you know the justification (even a non-rigorous one.. Pauling wasn't really) for that in terms of quantum mechanics, then that qualifies as a deeper understanding of that model at least.
    (This is not terribly hard to justify if you know Hartree-Fock theory)

    No, because most chemists don't actually know the answer.

    Although in this case, the question is badly phrased. As I said, hybridization doesn't exist in reality. It's not a measurable thing. Quoting Charles Coulson, hybridization is "a feature of a theoretical description".

    The question of why and how atoms bond is not an easy one. Even if you did understand VB and MO models very well, there are plenty of cases where they fail. What kind of bond does H2+ have? What about the aforementioned beryllium dimer?
    (it's unusual enough that there is no name. I'd call it 'covalent-ish')
     
    Last edited: Dec 9, 2009
  16. Dec 9, 2009 #15
    Thank you alxm for your advice.

    Recently, I have been trying to figure out what I have been learning for over the past eight years. And I realised that I have taken some theories I have learned for granted.
     
  17. Dec 10, 2009 #16

    berkeman

    User Avatar

    Staff: Mentor

    Thread locked pending moderation -- bickering posts deleted. C'mon guys.
     
  18. Dec 10, 2009 #17

    berkeman

    User Avatar

    Staff: Mentor

    I'm re-opening this thread for now. Please, no "bashing" or other negative remarks, and certainly no bickering. alxm has the high ground at the moment (and an impressive background that would appear to apply to this thread) -- I'd advise the others to stay on-topic as alxm has. Thank you.

    [added by jtbell nearly a year later: closed again due to necro-bickering]
     
    Last edited by a moderator: Nov 24, 2010
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook