1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Coefficients of Secular Equations

  1. Feb 14, 2017 #1
    • Moved from a technical forum, so homework template missing
    Dear Everyone,

    I would like to ask a question regarding the evaluation of the coefficients in Linear Combination of Atomic Orbitals (LCAO). In Molecular Quantum Mechanics book authored by Peter Atkins and Ronald Friedman (Fourth Edition ISBN 0199274983), we are trying to calculate the coefficients of the LCAO of H2+ (First atom: HA and Second atom: HB) molecule-ion.

    After solving the 2x2 secular determinant of this system,
    |α-E β-ES|
    |β-ES α-E|
    the energies are obtained as:
    E+ = (α+β)/(1+S)
    E- = (α-β)/(1-S)
    Note: α and β are the expectation values of the Hamiltonian.

    Solving the secular equations:
    ∑cr{Hrs - ESrs} = 0
    We will obtain the real coefficients of the normalized wavefunctions:
    cA = cB, ca = 1/{2(1+S)}½ for E+ = (α+β)/(1+S)
    cA = -cB, ca = 1/{2(1-S)}½ for E- = (α-β)/(1-S)

    1. I have proved that cA = cB and cA = -cB by matrix multiplication of the secular equations and inserting E+ and E- in the equations. But may anyone please advise me how can we obtain ca = 1/{2(1+S)}½ and ca = 1/{2(1-S)}½ ?

    2. Another question regarding the overlap integral S in the above secular determinant. Solving this integral leads to the following relation: S = <A|B> = {1 + (ZR/ao) + 1/3(ZR/ao)2}e-ZR/ao
    How can we calculate Z? Is it the Zeff for a compound or mixture found in the following link https://en.wikipedia.org/wiki/Effective_atomic_number ?

    Thanks a lot for the help and support...
  2. jcsd
  3. Feb 14, 2017 #2


    User Avatar

    Staff: Mentor

    Write down the condition for normalization.

    Z is the actual atomic number, so Z=1 if you are considering H##_2^+##.
  4. Feb 14, 2017 #3
    1. ∫Ψ*Ψ=1
    ∫(cAΦA + cBΦB)(cAΦA + cBΦB) = 1

    Let: ∫ΦAΦB = ∫ΦBΦA = S
    If: cA = cB then: cA = 1/{2(1+S)}½

    2. But in the case of heteronuclear diatomic molecule where we have different atomic numbers, shouldn't we calculate the Zeff of the compound as a whole as given in the link https://en.wikipedia.org/wiki/Effective_atomic_number in second section "For a compound or mixture"?

    Thank DrClaude for your advice.
  5. Feb 14, 2017 #4


    User Avatar

    Staff: Mentor

    The secular equation you are using is valid for single-electron diatomic molecules. The equation for S given above is for a homonuclear diatomic molecule of hydrogenic atoms, such that wave functions ##\Phi_A## and ##\Phi_B## are identical except for their origin. For a heteronuclear molecule, you will have to recalculate that overlap integral.

    An effective atomic number is only necessary if one of the atoms is only approximately a single-electron atom, such as an alkali atom.

    What system do you want to apply this to, and what for?
  6. Feb 14, 2017 #5
    There is no specific system I want to apply this. I just wanted to ask about the case of heteronuclear compounds. The idea is much clearer now. Thanks again DrClaude...
Know someone interested in this topic? Share this thread via Reddit, Google+, Twitter, or Facebook

Have something to add?
Draft saved Draft deleted

Similar Threads - Coefficients Secular Equations Date
Phys Orgo Secular Determinant question Sep 21, 2013