Skip to content

LCAO-MO Theory: H₂⁺, Secular Equations and Heteronuclear Bonds

CSIR NET physical chemistry: Born–Oppenheimer, LCAO for H2+, overlap, Coulomb and resonance integrals, secular determinant, and polar two-orbital bonds.

  • 20 min read
  • Updated
On this page · 15 sections
  1. Prerequisites
  2. The Hook: One Electron, Two Protons
  3. The Born–Oppenheimer Approximation
  4. The LCAO Trial Function
  5. Solving the Homonuclear Case
  6. Where the Electron Goes: Density and Symmetry
  7. Two Different Atoms: The Polar Bond
  8. From H₂⁺ to H₂
  9. Memory Tricks
  10. Don’t Confuse
  11. Common Mistakes
  12. Practice
  13. Quick Revision Table
  14. Key Takeaways
  15. What’s Next

Prerequisites

The Hook: One Electron, Two Protons

The Born–Oppenheimer Approximation

Nuclei are at least about 1836 times heavier than an electron, so electrons adjust almost instantly to any nuclear arrangement. The Born–Oppenheimer approximation separates the problem:

  1. Clamp the nuclei at a separation \( R \).
  2. Solve the electronic Schrödinger equation for that \( R \).
  3. Add the nuclear repulsion. The total electronic energy \( U(R) \), repeated for many \( R \), is the potential energy curve on which the nuclei vibrate.

For H₂⁺ in atomic units (energies in hartree \( E_h \), lengths in bohr \( a_0 \)), with \( r_A \) and \( r_B \) the electron’s distances from nuclei A and B:

$$\hat H_{\text{el}} = -\tfrac{1}{2}\nabla^2 - \frac{1}{r_A} - \frac{1}{r_B} + \frac{1}{R}$$

The \( 1/R \) term is a constant for fixed nuclei; it shifts the energy but not the wavefunction.

Two numbers from U(R)
  • \( R_e \), the equilibrium bond length: the position of the minimum of \( U(R) \).
  • \( D_e \), the dissociation energy measured from the minimum: \( U(\infty) - U(R_e) \). The experimentally measured \( D_0 \) is smaller by the zero-point vibrational energy: \( D_0 = D_e - \tfrac{1}{2}h\nu \) (harmonic approximation).

The LCAO Trial Function

Near nucleus A the electron should look like a hydrogen 1s electron on A; near B, like one on B. So try

$$\psi = c_A\,\phi_A + c_B\,\phi_B$$

where \( \phi_A \) and \( \phi_B \) are real, normalised 1s orbitals. This is a linear combination of atomic orbitals (LCAO). The variation theorem says the best \( c_A, c_B \) minimise

$$E = \frac{\int \psi \hat H \psi\, d\tau}{\int \psi^2\, d\tau} = \frac{c_A^2 H_{AA} + 2c_Ac_B H_{AB} + c_B^2 H_{BB}}{c_A^2 + 2c_Ac_B S + c_B^2}$$

The three integrals

IntegralSymbolDefinitionMeaningSign
Overlap\( S \)\( \int \phi_A \phi_B\, d\tau \)how much the two orbitals occupy the same space0 to 1 for two 1s orbitals; 1 at \( R = 0 \), 0 at \( R \to \infty \)
Coulomb\( \alpha = H_{AA} \)\( \int \phi_A \hat H \phi_A\, d\tau \)energy of an electron confined to \( \phi_A \) in the moleculenegative; close to the atomic orbital energy
Resonance (exchange)\( \beta = H_{AB} \)\( \int \phi_A \hat H \phi_B\, d\tau \)energy of the “shared” or interference partnegative near \( R_e \); the source of bonding

\( H_{AA} = H_{BB} = \alpha \) for a homonuclear molecule by symmetry. \( \beta \) has no classical analogue: it exists only because the wavefunction can be in two places at once.

The secular equations

Setting \( \partial E/\partial c_A = \partial E/\partial c_B = 0 \) gives two linear equations:

$$\begin{aligned} (H_{AA} - E)\,c_A + (H_{AB} - ES)\,c_B &= 0\\ (H_{AB} - ES)\,c_A + (H_{BB} - E)\,c_B &= 0 \end{aligned}$$

A non-trivial solution needs the secular determinant to vanish:

$$\begin{vmatrix} H_{AA} - E & H_{AB} - ES \\ H_{AB} - ES & H_{BB} - E \end{vmatrix} = 0$$

This recipe, write the determinant, solve for \( E \), back-substitute for the coefficients, is exactly the recipe of Hückel theory later in this chapter. Learn it once here.

Solving the Homonuclear Case

With \( H_{AA} = H_{BB} = \alpha \), \( H_{AB} = \beta \):

$$(\alpha - E)^2 = (\beta - ES)^2 \quad\Rightarrow\quad \alpha - E = \pm(\beta - ES)$$
LCAO energies and orbitals for a homonuclear diatomic
$$E_+ = \frac{\alpha + \beta}{1 + S}, \qquad \psi_+ = \frac{1}{\sqrt{2(1+S)}}\left(\phi_A + \phi_B\right) \quad (\sigma_g,\ \text{bonding})$$

$$E_- = \frac{\alpha - \beta}{1 - S}, \qquad \psi_- = \frac{1}{\sqrt{2(1-S)}}\left(\phi_A - \phi_B\right) \quad (\sigma_u^*,\ \text{antibonding})$$

Normalisation check. \( \int \psi_+^2 d\tau = N^2(1 + 1 + 2S) = 2N^2(1+S) = 1 \), so \( N_+ = [2(1+S)]^{-1/2} \). For the minus combination the cross term is \( -2S \).

Because \( \beta < 0 \), \( E_+ \) is the lower root. The coefficients are equal in magnitude because the atoms are equivalent: the electron spends equal time on each.

Why antibonding is pushed up more

Measure each orbital from \( \alpha \):

$$E_+ - \alpha = \frac{\beta - \alpha S}{1 + S}, \qquad E_- - \alpha = -\frac{\beta - \alpha S}{1 - S}$$

The numerators are equal and opposite; the denominators differ. Since \( 1 - S < 1 + S \),

$$\boxed{\frac{\text{destabilisation of }\psi_-}{\text{stabilisation of }\psi_+} = \frac{1+S}{1-S} > 1}$$

This is the quantitative reason He₂ (two bonding, two antibonding electrons) is not bound. If overlap is neglected (\( S = 0 \)) the splitting becomes symmetric, \( \alpha \pm \beta \), which is the Hückel approximation.

What the integrals look like for H₂⁺

Write \( \hat H = \hat h_A - 1/r_B + 1/R \), where \( \hat h_A \phi_A = E_{1s}\phi_A \). Then (atomic units)

$$\alpha = E_{1s} + \frac{1}{R} - j, \qquad \beta = E_{1s}S + \frac{S}{R} - k$$

with

  • \( j = \int \phi_A^2 / r_B\, d\tau \): the attraction of nucleus B for the electron cloud sitting on A (a classical Coulomb term);
  • \( k = \int \phi_A\phi_B / r_A\, d\tau \): the attraction of a nucleus for the overlap density \( \phi_A\phi_B \) (no classical analogue).

Substituting into \( E_\pm \):

$$E_\pm = E_{1s} + \frac{1}{R} - \frac{j \pm k}{1 \pm S}$$

For 1s orbitals with exponent 1, these integrals have closed forms: \( S = e^{-R}(1 + R + R^2/3) \), \( j = 1/R - e^{-2R}(1 + 1/R) \), \( k = e^{-R}(1 + R) \).

How good is it?

Treatment of H₂⁺\( R_e \)\( D_e \)
LCAO, 1s exponent fixed at 1about 2.49 \( a_0 \) (132 pm)about 1.76 eV
Exact (fixed nuclei)2.00 \( a_0 \) (106 pm)about 2.79 eV

The LCAO function is a variational trial function, so its energy lies above the exact energy at every \( R \): it underbinds. Letting the orbital exponent vary (so the 1s orbitals contract as the protons approach) and adding some 2p character (polarisation) both move the answer towards the exact one. That is the general lesson of basis-set improvement.

Where the Electron Goes: Density and Symmetry

Square the orbitals:

$$\psi_\pm^2 = N_\pm^2\left(\phi_A^2 + \phi_B^2 \pm 2\phi_A\phi_B\right)$$

The overlap density \( 2\phi_A\phi_B \) is positive in the region between the nuclei. In \( \psi_+ \) it is added: charge accumulates between the nuclei and attracts both. In \( \psi_- \) it is subtracted: a nodal plane appears halfway between the nuclei, perpendicular to the axis, and charge is pushed to the outsides.

graph TD
    A[phi_A + phi_B] --> B[overlap density added]
    B --> C[charge between nuclei]
    C --> D[sigma_g bonding, E+ = alpha+beta over 1+S]
    E[phi_A - phi_B] --> F[overlap density subtracted]
    F --> G[nodal plane at bond midpoint]
    G --> H[sigma_u antibonding, E- = alpha-beta over 1-S]

Symmetry labels. Both orbitals are cylindrically symmetric about the axis (\( \sigma \)). Under inversion through the bond centre, \( \phi_A \leftrightarrow \phi_B \): \( \psi_+ \) is unchanged (gerade, g), \( \psi_- \) changes sign (ungerade, u). The H₂⁺ ground state is \( 1\sigma_g \); the first excited orbital is \( 1\sigma_u \). The g/u label exists only when the molecule has a centre of inversion, so it applies to H₂, N₂ and O₂ but not to HF or CO.

Two Different Atoms: The Polar Bond

For a heteronuclear diatomic AB, \( H_{AA} = \alpha_A \ne H_{BB} = \alpha_B \). Take A as the more electronegative atom, so \( \alpha_A < \alpha_B \) (lower, more negative). To keep the algebra clean, neglect overlap (\( S = 0 \)), as Hückel theory does:

$$\begin{vmatrix} \alpha_A - E & \beta \\ \beta & \alpha_B - E \end{vmatrix} = 0$$
Two-orbital energies, S = 0
$$E_\pm = \frac{\alpha_A + \alpha_B}{2} \pm \sqrt{\left(\frac{\alpha_A - \alpha_B}{2}\right)^2 + \beta^2}$$

The lower root (bonding) lies below \( \alpha_A \); the upper (antibonding) lies above \( \alpha_B \), by the same amount when \( S = 0 \).

Coefficients from either secular equation: \( \dfrac{c_B}{c_A} = \dfrac{E - \alpha_A}{\beta} \).

Two limits tell you everything:

CaseConditionBonding orbitalCharacter
Homonuclear\( \alpha_A = \alpha_B \)\( \tfrac{1}{\sqrt 2}(\phi_A + \phi_B) \), split \( \pm\lvert\beta\rvert \)pure covalent
Small difference\( \lvert\alpha_A - \alpha_B\rvert \sim \lvert\beta\rvert \)larger coefficient on Apolar covalent
Large difference\( \lvert\alpha_A - \alpha_B\rvert \gg \lvert\beta\rvert \)almost pure \( \phi_A \)nearly ionic, A⁻B⁺

In the large-difference limit, expand the square root: the bonding orbital is stabilised below \( \alpha_A \) by only about \( \beta^2/\lvert\alpha_A - \alpha_B\rvert \). A big energy mismatch makes a weak covalent interaction. That is the quantitative form of the “comparable energies” rule.

HF as the standard example

In HF the F 2p orbitals lie well below H 1s. Only F \( 2p_z \) (and, to a small extent, F 2s) has σ symmetry to mix with H 1s. The result:

  • a bonding σ orbital mostly F \( 2p_z \), with a smaller H 1s contribution: the electron pair is drawn toward F, giving the H\( ^{\delta+} \)–F\( ^{\delta-} \) dipole;
  • an antibonding σ* mostly H 1s;
  • F \( 2p_x \) and \( 2p_y \) have the wrong symmetry: they stay non-bonding (the two π lone pairs);
  • F 2s is so low in energy that it is essentially a non-bonding lone pair.

The full heteronuclear MO diagrams (HF, CO, NO, CN⁻) are drawn in Bonding in Heteronuclear Molecules.

From H₂⁺ to H₂

Put two electrons into \( 1\sigma_g \) with opposite spins:

$$\Psi_{\text{MO}}(1,2) = \sigma_g(1)\,\sigma_g(2)\cdot\frac{1}{\sqrt 2}\left[\alpha(1)\beta(2) - \beta(1)\alpha(2)\right]$$

(here \( \alpha \) and \( \beta \) are spin functions, not integrals). The spatial part is symmetric and the spin part antisymmetric, so the total is antisymmetric, as the Pauli principle requires. Bond order is 1 and the state is a singlet, \( ^1\Sigma_g^+ \).

What this simple product hides, its 50 % ionic character at every distance, is the subject of the next page, where it is compared with the valence bond wavefunction.

Memory Tricks

Don’t Confuse

ThisvsThat
Coulomb integral \( \alpha = H_{AA} \): energy of an electron on one atomResonance integral \( \beta = H_{AB} \): interaction between atoms, the source of bonding
Overlap integral \( S \): dimensionless, 0 to 1\( \beta \): an energy, negative
Coulomb integral \( j \) (classical attraction of B for A’s cloud)Exchange-type integral \( k \) (attraction for overlap density; quantum)
\( D_e \), from the potential minimum\( D_0 = D_e - \tfrac12 h\nu \), from the lowest vibrational level
g/u: inversion symmetry, centrosymmetric molecules onlyσ/π: symmetry about the bond axis, any linear molecule

Common Mistakes

Practice

All questions below are original practice questions written for this page. They are not from any CSIR NET paper.

Level 1: Foundation

  1. Write the normalised bonding LCAO-MO of H₂⁺ including overlap.
  2. Which integral is responsible for the splitting of the bonding and antibonding levels?
  3. Can a 1s orbital on atom A and a \( 2p_x \) orbital on atom B combine if z is the bond axis?
Answers
  1. \( \psi_+ = [2(1+S)]^{-1/2}(\phi_A + \phi_B) \).
  2. The resonance integral \( \beta = H_{AB} \). With \( \beta = 0 \) and \( S = 0 \) both roots equal \( \alpha \).
  3. No. The overlap is zero by symmetry: reflection through the yz-plane leaves 1s unchanged but changes the sign of \( 2p_x \), so the positive and negative contributions cancel exactly.

Level 2: Part B Style

Practice Q1. For a homonuclear diatomic, \( \alpha = -10 \) eV, \( \beta = -4 \) eV and \( S = 0.2 \). The bonding and antibonding MO energies are:

(a) −14 eV and −6 eV (b) −11.67 eV and −7.50 eV (c) −11.67 eV and −5.00 eV (d) −12.50 eV and −7.50 eV

Answer: (b)

\( E_+ = (-10 - 4)/(1 + 0.2) = -14/1.2 = -11.67 \) eV.

\( E_- = (-10 + 4)/(1 - 0.2) = -6/0.8 = -7.50 \) eV.

Relative to \( \alpha \): bonding lowered by 1.67 eV, antibonding raised by 2.50 eV; the ratio is \( 1.2/0.8 = 1.5 = (1+S)/(1-S) \). Option (a) neglects \( S \).

Practice Q2. In the LCAO treatment of H₂⁺, the ratio of the destabilisation of \( \sigma_u^* \) to the stabilisation of \( \sigma_g \), both measured from \( \alpha \), is:

(a) 1 (b) \( (1-S)/(1+S) \) (c) \( (1+S)/(1-S) \) (d) \( 1/(1-S^2) \)

Answer: (c)

\( E_+ - \alpha = (\beta - \alpha S)/(1+S) \) and \( E_- - \alpha = -(\beta - \alpha S)/(1-S) \). The ratio of magnitudes is \( (1+S)/(1-S) \), greater than 1 for any positive overlap.

Practice Q3. Which statement about the LCAO-MO description of H₂⁺ with fixed 1s exponents is correct?

(a) It gives a bond energy greater than the exact value. (b) It gives a bond length shorter than the exact value. (c) It gives a bond energy less than the exact value and a bond length longer than the exact value. (d) It gives the exact bond length but too small a bond energy.

Answer: (c)

About 1.76 eV at about 2.49 bohr, against about 2.79 eV at 2.00 bohr. The variation theorem guarantees the energy is too high, so the binding is too small; the fixed, too-diffuse 1s orbitals also give too long a bond. Optimising the exponent fixes the bond length but still underbinds.

Practice Q4. For a heteronuclear diatomic AB with \( \alpha_A < \alpha_B < 0 \) and \( \beta < 0 \), neglecting overlap, the bonding MO:

(a) has equal coefficients on A and B (b) has the larger coefficient on A and lies below \( \alpha_A \) (c) has the larger coefficient on B and lies below \( \alpha_A \) (d) has the larger coefficient on A and lies between \( \alpha_A \) and \( \alpha_B \)

Answer: (b)

The lower root of the 2 × 2 determinant lies below the lower atomic level \( \alpha_A \), and \( c_B/c_A = (E - \alpha_A)/\beta \) has magnitude less than 1 because \( \lvert E - \alpha_A\rvert < \lvert\beta\rvert \) whenever \( \alpha_A \ne \alpha_B \). The bonding pair spends more time on the more electronegative atom A.

Level 3: Part C Style

Practice Q5. A model heteronuclear diatomic has \( \alpha_A = -12 \) eV, \( \alpha_B = -8 \) eV and \( \beta = -2 \) eV; neglect overlap. Two electrons occupy the bonding MO. Which set is correct?

(a) \( E_{\text{bonding}} = -12.83 \) eV; probability of finding the bonding electron on A = 0.85 (b) \( E_{\text{bonding}} = -14.00 \) eV; probability on A = 0.75 (c) \( E_{\text{bonding}} = -12.83 \) eV; probability on A = 0.50 (d) \( E_{\text{bonding}} = -12.00 \) eV; probability on A = 1.00

Answer: (a)

Energy. \( E_\pm = (\alpha_A + \alpha_B)/2 \pm \sqrt{((\alpha_A - \alpha_B)/2)^2 + \beta^2} = -10 \pm \sqrt{4 + 4} = -10 \pm 2.83 \) eV. Bonding: \( -12.83 \) eV; antibonding: \( -7.17 \) eV.

Coefficients. \( c_B/c_A = (E - \alpha_A)/\beta = (-12.83 + 12)/(-2) = 0.414 \). Normalising (\( S = 0 \)): \( c_A = 1/\sqrt{1 + 0.414^2} = 0.924 \), \( c_B = 0.383 \).

Probability on A \( = c_A^2 = 0.854 \approx 0.85 \). The bond is polar with the electron pair drawn toward A. Each electron carries 0.85 of its charge on A, so the pair places about \( 2 \times 0.854 - 1 = 0.71 \) of an electron’s charge in excess on A (a crude, overlap-free population analysis).

Practice Q6. Consider these statements about the LCAO treatment of a homonuclear diatomic:

  1. If \( S \) is neglected, the bonding and antibonding levels are placed symmetrically about \( \alpha \).
  2. The normalisation constant of the antibonding MO is larger than that of the bonding MO.
  3. The bonding MO has a nodal plane perpendicular to the bond axis at the bond midpoint.
  4. For H₂⁺, the classical Coulomb attraction term \( j \) alone produces a bound state.

Which are correct?

(a) 1 and 2 only (b) 1, 2 and 3 only (c) 2 and 4 only (d) 1, 3 and 4 only

Answer: (a) 1 and 2 only
  1. Correct. With \( S = 0 \), \( E_\pm = \alpha \pm \beta \).
  2. Correct. \( N_- = [2(1-S)]^{-1/2} > N_+ = [2(1+S)]^{-1/2} \) because \( 1 - S < 1 + S \).
  3. Wrong. That nodal plane belongs to the antibonding MO.
  4. Wrong. With \( k \) set to zero, \( 1/R - j/(1+S) \) is positive at every \( R \); binding comes from the overlap (exchange-type) term \( k \).

Practice Q7. In a model diatomic, the energy gap between the atomic levels is increased from 0 to \( \Delta = 6\lvert\beta\rvert \) while \( \beta \) is held fixed (\( S = 0 \)). By what factor does the stabilisation of the bonding MO below the lower atomic level fall?

(a) about 2 (b) about 3.2 (c) about 6 (d) about 36

Answer: (c) about 6

At \( \Delta = 0 \) the stabilisation below \( \alpha \) is \( \lvert\beta\rvert \).

At \( \Delta = 6\lvert\beta\rvert \): the bonding MO lies at \( \bar\alpha - \sqrt{9\beta^2 + \beta^2} = \bar\alpha - 3.162\lvert\beta\rvert \), where \( \bar\alpha \) is the mean of the two levels. The lower atomic level is at \( \bar\alpha - 3\lvert\beta\rvert \). Stabilisation below it: \( 0.162\lvert\beta\rvert \).

Factor: \( 1/(\sqrt{10} - 3) = 6.16 \), about 6. The perturbation estimate \( \beta^2/\Delta = \lvert\beta\rvert/6 \) gives a factor of 6. A large energy mismatch kills covalent stabilisation.

Quick Revision Table

IdeaFact
Born–Oppenheimerfix nuclei, solve for electrons, add \( 1/R \): gives \( U(R) \)
LCAO trial function\( \psi = c_A\phi_A + c_B\phi_B \), coefficients by the variation method
Secular determinant\( \begin{vmatrix} H_{AA} - E & H_{AB} - ES \\ H_{AB} - ES & H_{BB} - E \end{vmatrix} = 0 \)
Homonuclear energies\( E_\pm = (\alpha \pm \beta)/(1 \pm S) \)
Normalisation\( N_\pm = [2(1 \pm S)]^{-1/2} \)
Asymmetrydestabilisation / stabilisation \( = (1+S)/(1-S) \)
Signs\( \alpha, \beta < 0 \); \( 0 < S < 1 \)
H₂⁺ LCAO vs exactabout 1.76 eV at 2.49 \( a_0 \) vs about 2.79 eV at 2.00 \( a_0 \)
Heteronuclear, \( S = 0 \)\( E_\pm = \bar\alpha \pm \sqrt{(\Delta/2)^2 + \beta^2} \); \( c_B/c_A = (E - \alpha_A)/\beta \)
Large mismatchstabilisation about \( \beta^2/\Delta \): nearly ionic
g/uinversion symmetry; bonding \( \sigma_g \), antibonding \( \sigma_u^* \)

Key Takeaways

What’s Next