Skip to content

Many-Electron Atoms: Helium, the Orbital Approximation and SCF

CSIR NET chemistry: the N-electron atomic Hamiltonian, helium by perturbation and variation, orbital approximation, Hartree–Fock, Koopmans, correlation.

  • 3 min read
  • Updated
On this page · 21 sections
  1. Prerequisites
  2. The Hook: One Extra Electron, No Exact Answer
  3. Atomic Units: the Language of This Chapter
  4. The Hydrogen-Like Atom: the One Exact Case
  5. The Hamiltonian of an N-Electron Atom
  6. Helium, Step 1: The Independent-Electron (Zeroth-Order) Picture
  7. Helium, Step 2: First-Order Perturbation Theory
  8. Helium, Step 3: The Variation Method with an Effective Charge
  9. The Virial Theorem: a Free Consistency Check
  10. The Orbital Approximation
  11. The Self-Consistent Field (SCF) Method
  12. Koopmans’ Theorem
  13. Electron Correlation
  14. Beyond Helium: Lithium and the Need for Antisymmetry
  15. Memory Tricks
  16. Don’t Confuse
  17. Common Mistakes
  18. Practice
  19. Quick Revision Table
  20. Key Takeaways
  21. What’s Next

Prerequisites

The Hook: One Extra Electron, No Exact Answer

Atomic Units: the Language of This Chapter

Quantum chemistry is written in atomic units (a.u.), where \( m_e = e = \hbar = 4\pi\varepsilon_0 = 1 \). Equations lose their constants and numbers become easy to compare.

QuantityAtomic unitSI / common value
Lengthbohr, \( a_0 \)52.92 pm
Energyhartree, \( E_h \)27.211 eV = 2625.5 kJ/mol = 219 474.6 cm⁻¹
Mass\( m_e \)9.109 × 10⁻³¹ kg
Charge\( e \)1.602 × 10⁻¹⁹ C
Angular momentum\( \hbar \)1.055 × 10⁻³⁴ J s

Two conversions do most of the work: 1 \( E_h \) = 27.211 eV and 1 eV = 8065.5 cm⁻¹. Half a hartree is the familiar 13.6 eV ionisation energy of hydrogen (with infinite nuclear mass).

The Hydrogen-Like Atom: the One Exact Case

For one electron around a nucleus of charge \( Z \), the Schrödinger equation separates exactly:

$$E_n = -\frac{Z^2}{2n^2}\,E_h = -13.6\,\frac{Z^2}{n^2}\ \text{eV}$$
PropertyResult
Energydepends on \( n \) only; degeneracy \( n^2 \) (\( 2n^2 \) with spin)
Radial nodes\( n - l - 1 \)
Angular nodes\( l \)
Total nodes\( n - 1 \)
1s wavefunction\( \psi_{1s} = (Z^3/\pi)^{1/2} e^{-Zr} \) (a.u.)
\( \langle r \rangle \)\( \tfrac{1}{2Z}\left[3n^2 - l(l+1)\right] \) bohr; 1s: \( 3/(2Z) \)
Most probable radius, 1s\( 1/Z \) bohr

The \( l \)-degeneracy (2s and 2p equal in energy) is special to the pure \( 1/r \) potential. As soon as a second electron is present, the potential seen by each electron is no longer pure Coulomb and the degeneracy is lifted.

The Hamiltonian of an N-Electron Atom

For a nucleus of charge \( Z \) (treated as infinitely heavy) and \( N \) electrons, in atomic units:

$$\hat{H} = \sum_{i=1}^{N}\left(-\frac{1}{2}\nabla_i^2 - \frac{Z}{r_i}\right) + \sum_{i
  • The first sum is a set of one-electron operators \( \hat{h}(i) \): kinetic energy plus nuclear attraction. Each depends on one electron’s coordinates only.
  • The second sum is the electron–electron repulsion, with \( r_{ij} = \lvert \mathbf{r}_i - \mathbf{r}_j\rvert \). It couples every pair of electrons. There are \( N(N-1)/2 \) such terms.
  • Spin–orbit coupling and other relativistic terms are left out here; they are small for light atoms and are added later to split terms into levels (see Term Symbols, Coupling Schemes and Atomic Spectra).
  • Without the \( 1/r_{ij} \) terms the Hamiltonian would be a sum of hydrogen-like pieces, the wavefunction a product of hydrogen-like orbitals and the energy a sum of hydrogen-like energies. With them, the variables do not separate. Every method below is a way to deal with that.

    Helium in particular

    $$\hat{H}_{\text{He}} = \underbrace{-\tfrac{1}{2}\nabla_1^2 - \tfrac{Z}{r_1}}_{\hat{h}(1)} \;\underbrace{-\tfrac{1}{2}\nabla_2^2 - \tfrac{Z}{r_2}}_{\hat{h}(2)} \;+\; \underbrace{\tfrac{1}{r_{12}}}_{\text{repulsion}}, \qquad Z = 2$$

    The same Hamiltonian, with a different \( Z \), describes every two-electron (helium-like) ion: H⁻ (\( Z = 1 \)), Li⁺ (3), Be²⁺ (4), B³⁺ (5) and so on. CSIR NET likes these because the arithmetic changes but the method does not.

    Helium, Step 1: The Independent-Electron (Zeroth-Order) Picture

    Drop \( 1/r_{12} \). Then \( \hat{H}^{(0)} = \hat{h}(1) + \hat{h}(2) \), and for the ground state both electrons sit in a hydrogen-like 1s orbital:

    $$\psi^{(0)} = \psi_{1s}(1)\,\psi_{1s}(2) = \frac{Z^3}{\pi}\,e^{-Z(r_1 + r_2)}$$$$E^{(0)} = 2 \times \left(-\frac{Z^2}{2}\right) = -Z^2\ E_h$$

    For helium, \( E^{(0)} = -4\,E_h = -108.8 \) eV. The measured total energy (minus the sum of both ionisation energies) is about \( -2.9037\,E_h = -79.0 \) eV. Ignoring repulsion overbinds by about 38%.

    Helium, Step 2: First-Order Perturbation Theory

    Treat \( \hat{H}' = 1/r_{12} \) as a perturbation. The first-order correction is the expectation value of the perturbation over the zeroth-order function:

    $$E^{(1)} = \left\langle \psi^{(0)} \right| \frac{1}{r_{12}} \left| \psi^{(0)} \right\rangle = \frac{5Z}{8}\ E_h$$

    This integral is the Coulomb repulsion between two 1s charge clouds. Its evaluation (expanding \( 1/r_{12} \) in Legendre polynomials) is a standard textbook derivation; for the exam you need the result \( 5Z/8 \).

    $$E \approx E^{(0)} + E^{(1)} = -Z^2 + \frac{5Z}{8}$$

    For helium: \( -4 + 1.25 = -2.75\,E_h = -74.8 \) eV. Now the estimate is too high by about 4 eV. First-order perturbation theory gives no guarantee of being above or below the true energy in general, but here \( E^{(0)} + E^{(1)} \) is the expectation value of the full Hamiltonian over a normalised trial function, so the variation theorem applies and it must be an upper bound.

    Helium, Step 3: The Variation Method with an Effective Charge

    The zeroth-order function lets each electron feel the full charge \( Z \). In reality each electron partly screens the nucleus from the other. So use the same form but with an adjustable exponent \( \zeta \) (an “effective nuclear charge”):

    $$\phi = \frac{\zeta^3}{\pi}\,e^{-\zeta(r_1 + r_2)}$$

    Evaluate \( \langle \phi | \hat{H} | \phi \rangle \) term by term, in atomic units:

    TermExpectation valueReason
    Kinetic, both electrons\( 2 \times \tfrac{\zeta^2}{2} = \zeta^2 \)for a 1s function of exponent \( \zeta \), \( \langle T \rangle = \zeta^2/2 \)
    Nuclear attraction, both\( 2 \times (-Z\zeta) = -2Z\zeta \)\( \langle 1/r \rangle = \zeta \), multiplied by the true \( Z \)
    Repulsion\( +\tfrac{5\zeta}{8} \)same integral as before, with \( \zeta \) for \( Z \)
    $$E(\zeta) = \zeta^2 - 2Z\zeta + \frac{5}{8}\zeta$$

    Minimise:

    $$\frac{dE}{d\zeta} = 2\zeta - 2Z + \frac{5}{8} = 0 \quad\Rightarrow\quad \boxed{\zeta_{\text{opt}} = Z - \frac{5}{16}}$$

    Substituting back:

    $$\boxed{E_{\min} = -\left(Z - \frac{5}{16}\right)^2\ E_h}$$

    For helium: \( \zeta = 27/16 = 1.6875 \), \( E = -2.8477\,E_h = -77.49 \) eV. That is within 1.5 eV (about 2%) of experiment, and above it, as the variation theorem requires.

    The whole ladder for helium

    MethodEnergy (\( E_h \))Energy (eV)Error vs exact
    Zeroth order (no repulsion)−4.0000−108.829.8 eV too low
    First-order perturbation−2.7500−74.84.2 eV too high
    Variation, \( \zeta = 27/16 \)−2.8477−77.51.5 eV too high
    Hartree–Fock limit−2.8617−77.91.1 eV too high
    Exact (non-relativistic)−2.9037−79.0—

    The last two rows define the correlation energy, discussed below.

    Ionisation energy from the variational result

    The first ionisation energy is \( E(\text{He}^+) - E(\text{He}) \). He⁺ is hydrogen-like, so \( E(\text{He}^+) = -Z^2/2 = -2\,E_h \) exactly.

    $$IE_1 \approx -2 - (-2.8477) = 0.8477\ E_h = 23.07\ \text{eV}$$

    The measured value is 24.59 eV. The variational error in the atom (1.5 eV) passes straight into the ionisation energy, because He⁺ is exact.

    When the simple trial function fails: H⁻

    For H⁻, \( Z = 1 \): \( \zeta = 11/16 \) and \( E = -(11/16)^2 = -0.4727\,E_h \). But a hydrogen atom plus a free electron has \( E = -0.5\,E_h \). The simple product function predicts that H⁻ is unbound. In reality H⁻ is bound (its electron affinity is about 0.75 eV). The failure is a property of the trial function: a single product of identical 1s functions cannot describe one electron held close and one held far away. More flexible trial functions (different exponents for the two electrons, or explicit \( r_{12} \) dependence) do bind H⁻.

    The Virial Theorem: a Free Consistency Check

    For any system bound only by Coulomb forces, the exact wavefunction (and any variationally optimised scaled trial function) satisfies

    $$2\langle T\rangle = -\langle V\rangle \quad\Rightarrow\quad \langle T\rangle = -E, \qquad \langle V\rangle = 2E$$

    Check the helium variational result: \( \langle T \rangle = \zeta^2 = 2.8477 \), \( \langle V \rangle = -2Z\zeta + 5\zeta/8 = -6.75 + 1.0547 = -5.6953 \). Then \( \langle V \rangle / \langle T \rangle = -2.000 \). The optimum exponent automatically satisfies the virial theorem. For the exact He atom, \( \langle T \rangle = +2.9037\,E_h \) and \( \langle V \rangle = -5.8074\,E_h \).

    The Orbital Approximation

    Helium could be handled with one adjustable number. For a general atom we need a systematic scheme. The orbital approximation writes the N-electron wavefunction as a product (properly, an antisymmetrised product) of one-electron functions, the orbitals:

    $$\Psi(1, 2, \ldots, N) \approx \phi_1(1)\,\phi_2(2)\cdots\phi_N(N)$$

    The simple product is the Hartree product. It is the reason we can talk about “electron configurations” such as \( 1s^2 2s^2 2p^6 3s^1 \): a configuration is a statement about which orbitals are occupied in this approximation. The approximation is not exact (electrons do not move independently), but it is the language of all of chemistry.

    The Hartree product has one fatal flaw: it does not change sign when two electrons are swapped, so it violates the antisymmetry principle. Repairing that, with a Slater determinant, is the subject of The Antisymmetry Principle, Pauli Exclusion and Slater Determinants.

    The central-field idea

    Each electron moves in the field of the nucleus plus the average field of all the others. If that averaged field is taken to be spherically symmetric (the central-field approximation), the angular part of each orbital is still a spherical harmonic \( Y_l^{m_l} \), so orbitals keep their s, p, d, f labels and their \( l, m_l \) quantum numbers. Only the radial functions change.

    What changes, crucially, is the energy ordering. In a screened potential an electron that penetrates the inner shells feels more of the nucleus:

    $$\varepsilon_{ns} < \varepsilon_{np} < \varepsilon_{nd} < \varepsilon_{nf} \quad (\text{same } n, \text{ many-electron atom})$$

    This is the penetration and shielding story told quantitatively by Slater’s rules in Electronic Configuration, Shielding and Effective Nuclear Charge.

    The Self-Consistent Field (SCF) Method

    Hartree’s idea

    Hartree (1928) proposed: guess the orbitals, use them to build the average repulsive potential felt by each electron, solve a one-electron Schrödinger equation in that potential to get new orbitals, and repeat until the orbitals stop changing. The final field is self-consistent: the orbitals generate the field that reproduces the same orbitals.

    graph TD
        A["Guess orbitals
    (e.g. hydrogen-like or Slater-type)"] --> B["Build the average field
    each electron feels from the others"] B --> C["Solve one-electron equations
    in that field"] C --> D["New orbitals and
    orbital energies"] D --> E{"Same as the
    previous cycle?"} E -- "No" --> B E -- "Yes" --> F["Self-consistent field:
    orbitals, energies,
    total energy"]

    Hartree–Fock

    Fock and Slater (1930) added antisymmetry: the wavefunction is a single Slater determinant, and the variation theorem is used to find the best orbitals of that form. The result is the Hartree–Fock equations:

    $$\hat{f}(1)\,\phi_i(1) = \varepsilon_i\,\phi_i(1)$$

    with the Fock operator for a closed-shell atom

    $$\hat{f}(1) = \hat{h}(1) + \sum_{j}^{\text{occ}}\left[2\hat{J}_j(1) - \hat{K}_j(1)\right]$$
    PieceMeaning
    \( \hat{h}(1) \)kinetic energy plus nuclear attraction of electron 1
    \( \hat{J}_j \), Coulomb operatorclassical repulsion from the charge cloud \( \lvert\phi_j\rvert^2 \)
    \( \hat{K}_j \), exchange operatorno classical analogue; arises only from antisymmetry; acts between electrons of the same spin
    \( \varepsilon_i \)orbital energy: energy of one electron in \( \phi_i \) moving in the average field of all others

    The Fock operator depends on its own solutions (through \( \hat{J} \) and \( \hat{K} \)), which is why the equations must be solved iteratively.

    In practice: basis sets

    For atoms the HF equations can be solved numerically. For molecules (and routinely for atoms too) each orbital is expanded in a fixed set of basis functions and the equations become matrix equations (the Roothaan–Hall equations).

    Basis functionRadial formStrengthWeakness
    Slater-type orbital (STO)\( r^{n-1}e^{-\zeta r} \)correct cusp at the nucleus and correct exponential tailmulti-centre integrals are hard
    Gaussian-type orbital (GTO)\( e^{-\alpha r^2} \)integrals are fast (product of two Gaussians is a Gaussian)no cusp, tail decays too fast; several GTOs are combined to mimic one STO

    As the basis is enlarged the HF energy converges from above to the Hartree–Fock limit, the best single-determinant energy.

    Koopmans’ Theorem

    Koopmans' theorem

    In Hartree–Fock theory, the energy needed to remove an electron from occupied orbital \( \phi_i \), assuming the other orbitals do not change (“frozen orbitals”), is minus the orbital energy:

    $$IE_i \approx -\varepsilon_i$$

    Why it is only approximate:

    1. Orbital relaxation. When an electron leaves, the remaining electrons are less screened from the nucleus and contract. Relaxation lowers the energy of the ion, so the true ionisation energy is smaller than \( -\varepsilon_i \).
    2. Correlation. The neutral atom has more electrons and more correlation energy than the ion. Ignoring correlation makes the HF ionisation energy too small.

    The two errors have opposite signs and partly cancel, which is why Koopmans estimates of valence ionisation energies are often reasonable. This is the theoretical basis for reading photoelectron spectra as orbital-energy diagrams.

    A companion statement for electron affinities (\( EA \approx -\varepsilon \) of the lowest unoccupied orbital) is much less reliable, because relaxation and correlation now add instead of cancelling.

    Electron Correlation

    Hartree–Fock lets each electron see only the average position of the others. Real electrons avoid each other instantaneously. The energy missed is defined as

    $$\boxed{E_{\text{corr}} = E_{\text{exact}} - E_{\text{HF limit}}}$$

    where “exact” means the exact non-relativistic energy. \( E_{\text{corr}} \) is always negative (by the variation theorem, the HF energy lies above the exact one).

    For helium: \( E_{\text{corr}} = -2.9037 - (-2.8617) = -0.042\,E_h \approx -1.1 \) eV, about 1.4% of the total energy. Small as a fraction, but comparable to chemical bond energies and reaction energies, which is why correlation cannot be ignored in quantitative chemistry.

    Two kinds of avoidance are worth distinguishing:

    Same-spin electronsOpposite-spin electrons
    In Hartree–Fockalready kept apart by antisymmetry (the Fermi hole)not kept apart at all
    Correlation energy mainly from—the missing Coulomb hole

    Methods that go beyond HF include configuration interaction (CI, mixing several determinants), Møller–Plesset perturbation theory (MP2 and higher) and coupled-cluster theory. Density functional theory approximates exchange and correlation together through a functional of the electron density. You need to recognise the names and know that all of them aim at \( E_{\text{corr}} \).

    Beyond Helium: Lithium and the Need for Antisymmetry

    Lithium has three electrons. If the orbital approximation allowed it, the lowest configuration would be \( 1s^3 \). It is not; lithium is \( 1s^2 2s^1 \), and its first ionisation energy (5.39 eV) is that of a weakly held 2s electron. The rule that forbids \( 1s^3 \) is the Pauli exclusion principle, which in turn is a consequence of the antisymmetry principle. That is the next topic.

    Memory Tricks

    Don’t Confuse

    ThisvsThat
    Hartree product: simple product of orbitals, not antisymmetricHartree–Fock: single Slater determinant, antisymmetric, includes exchange
    Coulomb integral \( J \): classical repulsion of two charge clouds, always positiveExchange integral \( K \): quantum effect of antisymmetry, acts only between same-spin electrons
    Orbital energy \( \varepsilon_i \)Total energy: not \( \sum \varepsilon_i \); pair repulsions would be double-counted
    HF limit: best single determinant (infinite basis)Exact energy: includes correlation
    Effective charge \( \zeta = 27/16 \) for He (variational)Slater \( Z_{\text{eff}} = 1.70 \) for He 1s (empirical rule)
    \( E(\text{He}^+) = -2\,E_h \): exact\( E(\text{He}) \): no exact analytical solution

    Common Mistakes

    Practice

    All questions below are original practice questions written for this page. They are not from any CSIR NET paper. Use 1 \( E_h \) = 27.211 eV.

    Level 1: Foundation

    1. In the independent-electron approximation (no electron–electron repulsion), what is the ground-state energy of Li⁺ in \( E_h \) and in eV?
    2. What is the optimum variational exponent \( \zeta \) for Li⁺ with the trial function \( e^{-\zeta(r_1 + r_2)} \)?
    3. Define the correlation energy and state its sign.
    Answers
    1. \( -9\,E_h = -244.9 \) eV. \( E^{(0)} = -Z^2 = -9 \); \( 9 \times 27.211 = 244.9 \).
    2. \( \zeta = 3 - 5/16 = 2.6875 \).
    3. \( E_{\text{corr}} = E_{\text{exact}} - E_{\text{HF limit}} \), always negative, because the HF energy is a variational upper bound to the exact non-relativistic energy.

    Level 2: Part B Style

    Practice Q1. First-order perturbation theory, with the electron–electron repulsion as the perturbation, gives the ground-state energy of the He atom as:

    (a) \( -4.00\,E_h \) (b) \( -2.75\,E_h \) (c) \( -2.85\,E_h \) (d) \( -3.375\,E_h \)

    Answer: (b) −2.75 E_h

    \( E^{(0)} = -Z^2 = -4 \); \( E^{(1)} = 5Z/8 = 1.25 \). Sum \( -2.75\,E_h \) (−74.8 eV).

    (a) is zeroth order only. (c) is the variational result. (d) would come from using \( 5/8 \) instead of \( 5Z/8 \), that is, forgetting the factor \( Z \): \( -4 + 0.625 = -3.375 \).

    Practice Q2. Using the variational trial function \( \phi = (\zeta^3/\pi)e^{-\zeta(r_1 + r_2)} \), the minimum energy of Be²⁺ (Z = 4) is closest to:

    (a) \( -16.00\,E_h \) (b) \( -13.60\,E_h \) (c) \( -13.50\,E_h \) (d) \( -14.06\,E_h \)

    Answer: (b) −13.60 E_h

    \( \zeta = 4 - 0.3125 = 3.6875 \); \( E = -(3.6875)^2 = -13.598\,E_h \approx -13.60\,E_h \) (−370.0 eV).

    (a) is zeroth order. (c) is first order: \( -16 + 5(4)/8 = -13.50 \). (d) is \( -(3.75)^2 \), using \( Z - 1/4 \) by mistake.

    First order and variation differ by \( (5/16)^2 = 25/256 \approx 0.098\,E_h \) for every \( Z \), so the relative difference shrinks as \( Z \) grows: the repulsion becomes a smaller fraction of the total and perturbation theory improves.

    Practice Q3. Which statement about Koopmans’ theorem is correct?

    (a) It equates the total Hartree–Fock energy with the sum of orbital energies. (b) It approximates an ionisation energy by the negative of a Hartree–Fock orbital energy, neglecting orbital relaxation. (c) It is exact for two-electron atoms. (d) It gives electron affinities more accurately than ionisation energies.

    Answer: (b)

    Koopmans: \( IE_i \approx -\varepsilon_i \) with frozen orbitals. (a) is false: the total energy is not the sum of orbital energies. (c) is false: even for He, relaxation and correlation are ignored. (d) is false: for electron affinities the two errors add.

    Practice Q4. The exact non-relativistic ground-state energy of helium is \( -2.9037\,E_h \). The expectation values of its kinetic and potential energies are:

    (a) \( \langle T \rangle = 2.9037 \), \( \langle V \rangle = -5.8074 \) (b) \( \langle T \rangle = 1.4519 \), \( \langle V \rangle = -4.3556 \) (c) \( \langle T \rangle = -2.9037 \), \( \langle V \rangle = 0 \) (d) \( \langle T \rangle = 5.8074 \), \( \langle V \rangle = -8.7111 \)

    Answer: (a)

    Virial theorem for Coulomb systems: \( \langle T \rangle = -E = 2.9037\,E_h \), \( \langle V \rangle = 2E = -5.8074\,E_h \). Check: \( T + V = -2.9037 \). Kinetic energy is never negative, which rules out (c) at once.

    Level 3: Part C Style

    Practice Q5. The following energies (in \( E_h \)) refer to the helium atom ground state.

    ApproximationEnergy
    A−2.8477
    B−2.8617
    C−2.9037

    A is the variational result with one effective charge, B is the Hartree–Fock limit and C is the exact non-relativistic energy. He⁺ has \( E = -2.0000\,E_h \). Which set of conclusions is correct?

    (a) Correlation energy −0.042 \( E_h \); IE from A 23.07 eV; IE from C 24.59 eV (b) Correlation energy −0.056 \( E_h \); IE from A 23.07 eV; IE from C 24.59 eV (c) Correlation energy −0.042 \( E_h \); IE from A 77.49 eV; IE from C 79.01 eV (d) Correlation energy +0.042 \( E_h \); IE from A 23.45 eV; IE from C 24.59 eV

    Answer: (a)

    Correlation energy: \( E_{\text{exact}} - E_{\text{HF}} = -2.9037 - (-2.8617) = -0.0420\,E_h \) (−1.14 eV). (b) subtracts A from C instead of B from C; that difference includes the restriction of the trial function as well as correlation.

    Ionisation energies:

    • From A: \( -2.0000 - (-2.8477) = 0.8477\,E_h \times 27.211 = 23.07 \) eV.
    • From C: \( -2.0000 - (-2.9037) = 0.9037\,E_h \times 27.211 = 24.59 \) eV, the experimental value.

    (c) forgets that He⁺ still has one electron at \( -2\,E_h \) (those are total binding energies). (d) has the wrong sign for the correlation energy.

    Practice Q6. For the two-electron ion Li⁺ (Z = 3), consider the following statements, using the trial function \( e^{-\zeta(r_1 + r_2)} \).

    1. The first-order perturbation energy is \( -7.125\,E_h \).
    2. The variational energy is \( -7.2227\,E_h \), lower than the first-order result.
    3. The predicted second ionisation energy of lithium (Li⁺ → Li²⁺) is about 74.1 eV.
    4. The optimum \( \zeta \) is larger than \( Z \) because the electrons attract each other.

    The correct statements are:

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

    Answer: (b) 1, 2 and 3 only
    1. Correct. \( -Z^2 + 5Z/8 = -9 + 1.875 = -7.125\,E_h \).
    2. Correct. \( \zeta = 2.6875 \); \( E = -(2.6875)^2 = -7.2227\,E_h \), lower (better) than −7.125, as the variation method must give when it optimises over a family that contains the first-order function (\( \zeta = Z \)).
    3. Correct. \( E(\text{Li}^{2+}) = -Z^2/2 = -4.5\,E_h \). \( IE = -4.5 - (-7.2227) = 2.7227\,E_h = 74.09 \) eV. The measured value is about 75.6 eV; the shortfall is the variational error.
    4. Wrong. Electrons repel; each screens the nucleus from the other, so \( \zeta = Z - 5/16 < Z \).

    Practice Q7. Which of the following statements about the Hartree–Fock method for atoms are correct?

    1. The Fock operator contains the operators it is used to find, so the equations are solved iteratively.
    2. The exchange operator acts only between electrons of the same spin.
    3. Hartree–Fock includes the Coulomb hole between electrons of opposite spin.
    4. As the basis set is enlarged, the HF energy approaches the exact non-relativistic energy.

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

    Answer: (a) 1 and 2 only
    1. Correct. \( \hat{J} \) and \( \hat{K} \) are built from the occupied orbitals, the very functions being solved for: hence self-consistent field.
    2. Correct. Exchange arises from antisymmetry and appears only for pairs of electrons with parallel spin (it is the Fermi hole).
    3. Wrong. The Coulomb hole is exactly what HF lacks; opposite-spin electrons move independently in HF.
    4. Wrong. An infinite basis takes HF to the Hartree–Fock limit, which lies above the exact energy by the correlation energy.

    Quick Revision Table

    IdeaFact or formula
    Atomic units1 \( E_h \) = 27.211 eV = 2625.5 kJ/mol; 1 bohr = 52.92 pm
    Hydrogen-like energy\( E_n = -Z^2/(2n^2)\,E_h = -13.6\,Z^2/n^2 \) eV
    Atomic Hamiltonian\( \sum_i(-\tfrac{1}{2}\nabla_i^2 - Z/r_i) + \sum_{i
    He zeroth order\( -Z^2 = -4\,E_h \) (−108.8 eV)
    Repulsion integral (1s²)\( 5Z/8 \)
    He first order\( -Z^2 + 5Z/8 = -2.75\,E_h \) (−74.8 eV)
    Variation\( \zeta = Z - 5/16 \); \( E = -(Z - 5/16)^2 \); He −2.8477 \( E_h \)
    HF limit / exact, He−2.8617 / −2.9037 \( E_h \)
    He IE24.59 eV measured; 23.07 eV from simple variation
    H⁻simple product trial function fails to bind it
    Virial theorem\( \langle T \rangle = -E \), \( \langle V \rangle = 2E \)
    Fock operator\( \hat{h} + \sum_j(2\hat{J}_j - \hat{K}_j) \)
    Koopmans\( IE_i \approx -\varepsilon_i \) (frozen orbitals)
    Correlation energy\( E_{\text{exact}} - E_{\text{HF}} < 0 \); He −0.042 \( E_h \)
    Many-electron ordering\( ns < np < nd < nf \) for the same \( n \)

    Key Takeaways

    What’s Next