The Schrödinger equation can be solved exactly only for one-electron systems. For every real molecule we need an approximate method that a computer can handle. The oldest and most important of these is the Hartree-Fock (HF) method, and every HF or DFT calculation needs a basis set, a set of mathematical functions from which the molecular orbitals are built. This module explains the approximations that lead to HF (Born-Oppenheimer, orbital approximation, Slater determinant, variation principle), the self-consistent field procedure, what HF gets right and wrong, and then the types of basis sets, their names and how to choose one. It maps to course outcome CO6. Counting basis functions and explaining the SCF loop are common exam questions.
Summary
The molecular Hamiltonian: For a molecule, the Hamiltonian contains the kinetic energy of the nuclei and the electrons, the attraction between electrons and nuclei, and the repulsion between electrons and between nuclei. In atomic units, \(\hat{H} = -\sum_A\dfrac{1}{2M_A}\nabla_A^2 - \sum_i\dfrac12\nabla_i^2 - \sum_{i,A}\dfrac{Z_A}{r_{iA}} + \sum_{i<j}\dfrac{1}{r_{ij}} + \sum_{A<B}\dfrac{Z_AZ_B}{R_{AB}}\). The electron-electron repulsion term \(1/r_{ij}\) couples the motion of all electrons, which is why the equation cannot be solved exactly.
Born-Oppenheimer approximation: Nuclei are thousands of times heavier than electrons (a proton is about 1836 times heavier than an electron), so they move much more slowly. The electrons adjust almost instantly to any position of the nuclei. Born and Oppenheimer (1927) therefore separated the two motions: the nuclei are treated as fixed while the electronic problem is solved, \(\Psi_{\text{total}} \approx \psi_{\text{el}}(\mathbf{r};\mathbf{R})\chi_{\text{nuc}}(\mathbf{R})\). The nuclear kinetic energy is dropped from the electronic Hamiltonian and the nuclear repulsion becomes a constant: \(\hat{H}_{\text{el}}\psi_{\text{el}} = E_{\text{el}}(\mathbf{R})\psi_{\text{el}}\). Solving this for many nuclear arrangements R gives the potential energy surface \(U(\mathbf{R}) = E_{\text{el}}(\mathbf{R}) + V_{\text{NN}}\), on which the nuclei move. This approximation is the basis of the ideas of molecular structure, geometry optimization and the PES used throughout computational chemistry. It fails only when two electronic states come very close in energy.
Orbital approximation and Slater determinant: Hartree's idea was to write the many-electron wave function as a product of one-electron functions (orbitals), one for each electron. But electrons are fermions: the wave function must change sign when two electrons are exchanged (antisymmetry principle, a general form of the Pauli principle). A simple product does not do this. The correct antisymmetric form is a Slater determinant of spin orbitals χ (a spatial orbital times α or β spin): \(\Psi = \dfrac{1}{\sqrt{N!}}\det|\chi_1(1)\chi_2(2)\cdots\chi_N(N)|\). Exchanging two electrons swaps two rows, which changes the sign of a determinant; putting two electrons in the same spin orbital makes two columns equal, so the determinant is zero, which is the Pauli exclusion principle.
Variation principle: For any trial wave function, the expectation value of the energy is never lower than the true ground-state energy: \(E_{\text{trial}} = \dfrac{\int\Psi^*\hat{H}\Psi\,d\tau}{\int\Psi^*\Psi\,d\tau} \ge E_0\). So the best wave function of a given form is the one with the lowest energy. HF uses this to find the best orbitals.
Hartree-Fock method: In HF, the wave function is a single Slater determinant, and the orbitals are chosen to minimize the energy. This leads to the Hartree-Fock equations, \(\hat{f}\chi_i = \varepsilon_i\chi_i\), one for each orbital. The Fock operator \(\hat{f}\) contains the kinetic energy of one electron, its attraction to all nuclei, and the average repulsion from all other electrons: a Coulomb part J (classical repulsion from the average charge cloud) and an exchange part K (a purely quantum effect from antisymmetry, acting only between electrons of the same spin). So each electron moves in the average (mean) field of the others. Since the Fock operator depends on the orbitals that we are trying to find, the equations are solved by iteration: the self-consistent field (SCF) method. The orbital energy \(\varepsilon_i\) is approximately the energy needed to remove an electron from that orbital (Koopmans' theorem): \(IE \approx -\varepsilon_{\text{HOMO}}\).
Roothaan-Hall equations: For molecules, Roothaan and Hall (1951) expanded each MO as a linear combination of known basis functions φμ, \(\psi_i = \sum_\mu c_{\mu i}\phi_\mu\) (LCAO). The HF equations then become a matrix equation, \(\mathbf{FC} = \mathbf{SC}\boldsymbol{\varepsilon}\), where F is the Fock matrix, S the overlap matrix, C the coefficients and ε the orbital energies. This is what computer programs actually solve. The number of two-electron integrals grows as about N⁴/8 for N basis functions, so HF formally scales as N⁴.
Restricted and unrestricted HF: For closed-shell molecules (all electrons paired) the α and β electrons share the same spatial orbitals: restricted Hartree-Fock (RHF). For open-shell systems (radicals, O2) unrestricted Hartree-Fock (UHF) gives α and β electrons separate orbitals; restricted open-shell HF (ROHF) is another option.
Achievements and limitations of HF: HF gives about 99 percent of the total energy, good equilibrium geometries for many molecules, reasonable dipole moments and charge distributions, and a set of orbitals that is the starting point for better methods. But each electron feels only the average of the others, so the instantaneous avoidance of electrons (electron correlation) is missing. The error is the correlation energy, \(E_{\text{corr}} = E_{\text{exact}} - E_{\text{HF}}\) (negative, and HF energy is always above the exact energy by the variation principle). Although small in total, it is comparable to chemical bond energies, so HF gives poor bond dissociation energies, fails to describe bond breaking properly with RHF, and misses dispersion forces. Post-HF methods (MP2, CI, CCSD(T)) and DFT add correlation. Even with an infinite basis set HF reaches only the HF limit, not the exact energy.
Basis sets and their importance: A basis set is a set of known functions, usually centred on the atoms, used to build molecular orbitals by LCAO. A finite basis set cannot describe the orbitals perfectly, so the basis set is a source of error along with the method; the result is always reported as method/basis set. A larger basis set gives a lower (better) energy by the variation principle and approaches the complete basis set (CBS) limit, but the cost rises steeply (about N⁴ for HF). The choice of basis set is therefore as important as the choice of method.
Slater-type and Gaussian-type orbitals: Slater-type orbitals (STOs), \(\phi^{\text{STO}} = Nr^{n-1}e^{-\zeta r}Y_{lm}(\theta,\phi)\), look like real hydrogen-like orbitals: they have a cusp (a sharp point) at the nucleus and the correct slow exponential decay at large r. But integrals involving STOs on several centres are very hard to calculate. Gaussian-type orbitals (GTOs), \(\phi^{\text{GTO}} = Nx^ay^bz^ce^{-\alpha r^2}\), have the wrong shape (no cusp, decay too fast), but the product of two Gaussians on different centres is another Gaussian (Gaussian product theorem), which makes integrals fast. The solution is to add several Gaussians (primitives) with fixed coefficients to imitate one STO: a contracted Gaussian function, \(\phi^{\text{CGF}} = \sum_k d_k g_k(\alpha_k)\). Almost all molecular programs use contracted GTOs. Here a + b + c = 0, 1, 2 gives s, p and d type functions.
Types of basis sets: (1) Minimal basis set: one function for each occupied atomic orbital (H: 1s; C to Ne: 1s, 2s, 2px, 2py, 2pz). STO-3G uses 3 primitive Gaussians per STO. Cheap but crude. (2) Double-zeta and split-valence: two functions per orbital (double zeta) let the orbital grow or shrink in different directions; in split-valence sets only the valence orbitals are doubled, since the core changes little in bonding. Pople's notation X-YZG: the core orbital is one contracted function of X primitives, and each valence orbital is split into an inner part of Y primitives and an outer part of Z primitives. So 3-21G: core 3 primitives, valence 2 + 1; 6-31G: core 6, valence 3 + 1; 6-311G is triple-split valence. (3) Polarization functions: functions of higher angular momentum (d on heavy atoms, p on H) let orbitals distort away from the atom, which is needed for correct bond angles and for bonds in strained or hypervalent molecules. Notation 6-31G(d) or 6-31G* adds d on non-hydrogen atoms; 6-31G(d,p) or 6-31G** also adds p on H. (4) Diffuse functions: very spread-out s and p functions with small exponents, needed for anions, lone pairs, excited states and weak interactions. Notation 6-31+G (on heavy atoms) and 6-31++G (also on H). (5) Correlation-consistent sets of Dunning, cc-pVXZ with X = D, T, Q, 5 (double, triple, quadruple, quintuple zeta), designed to converge systematically to the CBS limit in correlated calculations; aug-cc-pVXZ adds diffuse functions.
Choosing a basis set: Use a minimal basis only for quick, rough work; 6-31G(d) is a common choice for geometries of organic molecules with HF or DFT; add diffuse functions for anions and hydrogen bonds; use triple-zeta (6-311+G(d,p), cc-pVTZ, def2-TZVP) for accurate energies; correlated methods need larger, correlation-consistent sets. Heavy elements often use effective core potentials, which replace core electrons with a potential.
Mind map
Computational Methods in Quantum Chemistry
- Approximations
- Born-Oppenheimer: fixed nuclei
- Orbital approximation
- Slater determinant (antisymmetry)
- Variation principle
- Hartree-Fock
- Mean field: Fock operator f̂χ = εχ
- Coulomb J and exchange K
- SCF iteration
- Roothaan-Hall FC = SCε
- RHF, UHF
- HF results and limits
- ~99% of energy, good geometries
- Koopmans: IE ≈ −ε(HOMO)
- No electron correlation
- E_corr = E_exact − E_HF
- Basis functions
- STO: e^(−ζr), cusp, hard integrals
- GTO: e^(−αr²), fast integrals
- Contracted Gaussians
- Basis set types
- Minimal: STO-3G
- Split valence: 3-21G, 6-31G
- Polarization: * or (d,p)
- Diffuse: + or ++
- cc-pVXZ, CBS limit
The self-consistent field (SCF) procedure
- Specify the molecule: nuclear coordinates (fixed, Born-Oppenheimer), charge, multiplicity and basis set.
- Calculate all one-electron integrals (overlap S, kinetic and nuclear attraction) and two-electron repulsion integrals over the basis functions.
- Make an initial guess of the MO coefficients C (for example from a semi-empirical or core Hamiltonian calculation).
- From C build the electron density (density matrix) for the occupied orbitals.
- Use the density to build the Fock matrix F, which includes the average Coulomb and exchange repulsion.
- Solve the Roothaan-Hall equations FC = SCε to get new coefficients C and orbital energies ε.
- Fill the lowest orbitals with electrons (two per orbital in RHF) and compute the new density and total energy.
- Compare with the previous cycle. If energy and density have stopped changing (within a set threshold), the field is self-consistent; otherwise go back to step 5 with the new density.
- Use the converged orbitals and energy to compute properties: dipole moment, charges, orbital energies, or as a start for geometry optimization or post-HF methods.
Important definitions
| Term | Meaning |
|---|---|
| Born-Oppenheimer approximation | Separation of nuclear and electronic motion: electrons are solved for fixed nuclear positions because nuclei are much heavier. |
| Potential energy surface | Electronic energy plus nuclear repulsion as a function of nuclear positions, obtained from repeated Born-Oppenheimer calculations. |
| Spin orbital | A one-electron function that is the product of a spatial orbital and a spin function α or β. |
| Slater determinant | A determinant of spin orbitals that gives an antisymmetric many-electron wave function, satisfying the Pauli principle. |
| Variation principle | The energy calculated from any trial wave function is greater than or equal to the true ground-state energy. |
| Hartree-Fock method | An ab initio method using a single Slater determinant whose orbitals are optimized in the mean field of all other electrons. |
| Fock operator | The effective one-electron Hamiltonian in HF: kinetic energy, nuclear attraction and average Coulomb and exchange repulsion. |
| Self-consistent field (SCF) | The iterative procedure in which orbitals are improved until the field they produce no longer changes. |
| Exchange energy | A quantum effect arising from antisymmetry that lowers the repulsion between electrons of the same spin. |
| Electron correlation | The correlated, instantaneous avoidance of electrons that HF misses; its energy is E_exact − E_HF. |
| Koopmans' theorem | The ionization energy is approximately equal to minus the HF orbital energy of the electron removed. |
| Basis set | A set of known functions used to expand molecular orbitals. |
| Contracted Gaussian function | A fixed linear combination of primitive Gaussians used as one basis function. |
| Split-valence basis set | A basis with one function per core orbital and two or more per valence orbital, e.g. 6-31G. |
| Polarization functions | Functions of higher angular momentum (d on C, p on H) that allow orbitals to distort. |
| Diffuse functions | Widely spread functions with small exponents for anions, lone pairs and weak interactions. |
| HF limit / CBS limit | The result obtained with a complete (infinite) basis set. |
Formulas
| Formula | Meaning |
|---|---|
| \( \Psi_{\text{total}}(\mathbf{r},\mathbf{R}) \approx \psi_{\text{el}}(\mathbf{r};\mathbf{R})\,\chi_{\text{nuc}}(\mathbf{R}) \) | Born-Oppenheimer separation. |
| \( \hat{H}_{\text{el}} = -\sum_i\tfrac12\nabla_i^2 - \sum_{i,A}\dfrac{Z_A}{r_{iA}} + \sum_{i<j}\dfrac{1}{r_{ij}} \) | Electronic Hamiltonian in atomic units (fixed nuclei). |
| \( \hat{H}_{\text{el}}\psi_{\text{el}} = E_{\text{el}}(\mathbf{R})\psi_{\text{el}},\quad U(\mathbf{R}) = E_{\text{el}} + V_{\text{NN}} \) | Electronic Schrödinger equation and potential energy surface. |
| \( \Psi = \dfrac{1}{\sqrt{N!}}\begin{vmatrix}\chi_1(1) & \chi_2(1) & \cdots & \chi_N(1)\\ \chi_1(2) & \chi_2(2) & \cdots & \chi_N(2)\\ \vdots & & & \vdots\\ \chi_1(N) & \chi_2(N) & \cdots & \chi_N(N)\end{vmatrix} \) | Slater determinant for N electrons. |
| \( E_{\text{trial}} = \dfrac{\int\Psi^*\hat{H}\Psi\,d\tau}{\int\Psi^*\Psi\,d\tau} \ge E_0 \) | Variation principle. |
| \( \hat{f}\chi_i = \varepsilon_i\chi_i,\quad \hat{f} = \hat{h} + \sum_j(\hat{J}_j - \hat{K}_j) \) | Hartree-Fock equations; ĥ is the one-electron part (kinetic + nuclear attraction). |
| \( \psi_i = \sum_\mu c_{\mu i}\phi_\mu \) | LCAO expansion of MOs in basis functions. |
| \( \mathbf{FC} = \mathbf{SC}\boldsymbol{\varepsilon} \) | Roothaan-Hall equations. |
| \( E_{\text{corr}} = E_{\text{exact}} - E_{\text{HF}} \) | Correlation energy (negative). |
| \( IE \approx -\varepsilon_{\text{HOMO}} \) | Koopmans' theorem. |
| \( \phi^{\text{STO}} = Nr^{n-1}e^{-\zeta r}Y_{lm}(\theta,\phi) \) | Slater-type orbital. |
| \( \phi^{\text{GTO}} = Nx^ay^bz^ce^{-\alpha r^2} \) | Cartesian Gaussian-type orbital; l = a + b + c. |
| \( \phi^{\text{CGF}} = \sum_{k=1}^{K}d_k\,g_k(\alpha_k, \mathbf{r}) \) | Contracted Gaussian of K primitives. |
| \( \text{number of unique two-electron integrals} \approx \dfrac{N^4}{8} \) | Reason for N⁴ scaling of HF. |
Key points to remember
- Born-Oppenheimer: nuclei fixed, electrons solved; gives the idea of a potential energy surface.
- A Slater determinant is antisymmetric and automatically obeys the Pauli principle.
- Variation principle: E_HF ≥ E_exact; a lower energy means a better wave function.
- HF is a mean-field method; each electron feels the average of the others.
- HF equations are solved iteratively (SCF) because the Fock operator depends on its own solutions.
- Roothaan-Hall: FC = SCε; HF scales as N⁴.
- HF misses correlation energy; post-HF (MP2, CI, CC) and DFT recover it.
- STOs are physically better; GTOs are computationally faster; contracted GTOs combine both advantages.
- Minimal STO-3G; split valence 3-21G, 6-31G; * or (d) = polarization; + = diffuse; cc-pVXZ for correlated work.
- Bigger basis set: lower energy, higher cost; results should be reported as method/basis.
Solved examples
Example 1. How many basis functions does H2O have in (a) STO-3G, (b) 6-31G, (c) 6-31G(d), (d) 6-31G(d,p)? (Use 6 Cartesian d functions.)
(a) STO-3G: O has 1s, 2s, 2px, 2py, 2pz = 5; each H has 1s = 1. Total = 5 + 2 = 7. (b) 6-31G: O core 1s = 1, valence 2s and 2p split in two: 2 × 4 = 8, so O = 9; each H 1s split in two = 2. Total = 9 + 4 = 13. (c) 6-31G(d): add 6 d functions on O: 13 + 6 = 19 (18 if 5 spherical d functions are used). (d) 6-31G(d,p): also add 3 p functions on each H: 19 + 6 = 25.
Example 2. How many primitive Gaussians are in the STO-3G and 6-31G basis sets for H2O?
STO-3G: each of the 7 functions has 3 primitives: 7 × 3 = 21. 6-31G: O core 1s: 6; O valence inner functions (2s, 2px, 2py, 2pz): 4 × 3 = 12; O valence outer functions: 4 × 1 = 4; so O = 22. Each H: inner 3 + outer 1 = 4, two H = 8. Total = 30 primitives.
Example 3. Count the basis functions of CH4 in STO-3G and 6-31G.
STO-3G: C = 5, 4 H = 4, total 9. 6-31G: C = 1 + 8 = 9, 4 H = 4 × 2 = 8, total 17.
Example 4. For H2 at its equilibrium distance, the HF limit energy is −1.1336 Eh and the exact non-relativistic energy is −1.1745 Eh. Find the correlation energy in Eh and kJ/mol.
\(E_{\text{corr}} = E_{\text{exact}} - E_{\text{HF}} = -1.1745 - (-1.1336) = -0.0409\) Eh. In kJ/mol: \(-0.0409\times2625.5 = -107\) kJ/mol. This is only about 3.5 percent of the total energy but is a quarter of the H-H bond energy (about 432 kJ/mol), which shows why correlation matters for chemistry.
Example 5. The HOMO energy of a molecule from an HF calculation is −0.500 Eh. Estimate its first ionization energy.
By Koopmans' theorem \(IE \approx -\varepsilon_{\text{HOMO}} = 0.500\) Eh \(= 0.500\times27.211 = 13.6\) eV.
Example 6. How many unique two-electron integrals are there for H2O with 13 basis functions, and roughly how much more work is needed with 25 basis functions?
With 8-fold symmetry the number of unique integrals is M(M + 1)/2 where M = N(N + 1)/2. For N = 13: M = 91, integrals = 91 × 92/2 = 4186 (close to 13⁴/8 = 3570). Work ratio for N⁴ scaling: \((25/13)^4 = 13.7\), about 14 times more.
Important questions and answers
Q1. Explain the Born-Oppenheimer approximation and its importance. (5 marks)
The molecular Hamiltonian contains the kinetic energies of nuclei and electrons and all Coulomb interactions. Since nuclei are at least 1836 times heavier than electrons, they move far more slowly, and electrons adjust almost instantly to any nuclear arrangement. So the wave function is approximated as a product \(\psi_{\text{el}}(\mathbf{r};\mathbf{R})\chi_{\text{nuc}}(\mathbf{R})\); the electronic Schrödinger equation \(\hat{H}_{\text{el}}\psi_{\text{el}} = E_{\text{el}}(\mathbf{R})\psi_{\text{el}}\) is solved with the nuclei fixed (nuclear kinetic energy dropped, nuclear repulsion a constant). Importance: it makes the problem much smaller; it gives the concept of a potential energy surface U(R) on which nuclei move; it gives meaning to molecular geometry, equilibrium structure, transition states and vibrations; geometry optimization, MM and MD all rely on it. It breaks down when electronic states are nearly degenerate (some photochemistry).
Q2. Describe the Hartree-Fock method and the SCF procedure. (8 marks)
Approximations: Born-Oppenheimer; the wave function is a single Slater determinant of spin orbitals (antisymmetric, obeys Pauli); each electron moves in the average field of all others (mean field); finite basis set. Applying the variation principle gives the HF equations \(\hat{f}\chi_i = \varepsilon_i\chi_i\), with Fock operator = one-electron kinetic and nuclear attraction terms + Coulomb (J) − exchange (K) operators. With the LCAO expansion they become the Roothaan-Hall matrix equations FC = SCε. Since F depends on C, they are solved iteratively: guess C, build density, build F, solve for new C, repeat until energy and density converge (self-consistency). Write the flow chart. Output: total energy, orbital energies, MOs, and properties. Limitation: no electron correlation; E_HF lies above E_exact.
Q3. What is electron correlation? Why does the Hartree-Fock method not include it? (5 marks)
Electrons repel each other and so tend to avoid each other at every instant; their motions are correlated. In HF each electron interacts only with the average charge cloud of the others, not with their instantaneous positions; only same-spin electrons are kept apart through exchange (Fermi correlation). The missing part is the correlation energy, \(E_{\text{corr}} = E_{\text{exact}} - E_{\text{HF}}\), always negative, about 1 percent of the total energy but similar in size to bond and reaction energies. Consequences: poor bond energies, wrong dissociation curves with RHF, no dispersion. It is recovered by MP2, CI, coupled cluster and (approximately) DFT.
Q4. What is a basis set? Explain STO and GTO and why contracted Gaussians are used. (7 marks)
A basis set is a set of mathematical functions, centred on the atoms, used to build MOs as \(\psi_i = \sum c_{\mu i}\phi_\mu\). STO: \(Nr^{n-1}e^{-\zeta r}Y_{lm}\); correct cusp at the nucleus and correct exponential decay, so few are needed, but multi-centre integrals are very difficult and slow. GTO: \(Nx^ay^bz^ce^{-\alpha r^2}\); no cusp and too rapid decay, so more are needed, but the product of two Gaussians on different centres is a single Gaussian, so integrals are fast and analytic. Contracted Gaussian: a fixed linear combination of several primitive GTOs fitted to look like an STO (e.g. STO-3G uses 3 per function). This keeps the speed of GTOs while approaching the shape of STOs, and reduces the number of variational coefficients.
Q5. Explain minimal, split-valence, polarized and diffuse basis sets with examples. Explain the notation 6-31G**. (8 marks)
Minimal: one basis function per occupied atomic orbital; STO-3G; cheap, rigid, qualitative only. Split valence: core described by one function, valence by two or more of different size, so orbitals can expand or contract with the bonding environment; 3-21G, 6-31G (double), 6-311G (triple). Polarization: adds higher angular momentum functions (d on heavy atoms, p on H) so that orbitals can bend and polarize; important for geometries and strained or hypervalent molecules; 6-31G(d) = 6-31G*, 6-31G(d,p) = 6-31G. Diffuse: adds wide s and p functions for electrons far from the nucleus; anions, lone pairs, excited states, hydrogen bonding; 6-31+G, 6-31++G. Notation 6-31G: core orbitals are one contracted Gaussian of 6 primitives; each valence orbital is split into an inner contracted function of 3 primitives and an outer function of 1 primitive; the first * adds d functions on non-hydrogen atoms; the second * adds p functions on hydrogen. Also mention correlation-consistent cc-pVDZ, cc-pVTZ for correlated methods and convergence to the CBS limit.
Q6. Why are basis sets important in computational chemistry? (3 marks)
MOs are built from the basis functions, so the basis set limits how well the orbitals and the energy can be described; a poor basis gives errors even with an excellent method. A larger basis gives lower energy (variation principle) and better properties, approaching the basis set limit, but cost grows roughly as N⁴. So the basis must suit the property and system (polarization for geometries, diffuse for anions), and the result must always be quoted as method/basis.
Q7. State Koopmans' theorem and its use. (3 marks)
In HF theory the ionization energy for removing an electron from orbital i is approximately \(-\varepsilon_i\), assuming the other orbitals do not relax after ionization. So the first ionization energy ≈ −ε(HOMO). It allows photoelectron spectra to be interpreted with HF orbital energies. Errors from orbital relaxation and missing correlation partly cancel, so agreement is often fair.
Exam tips
- Start the HF answer by listing the approximations (Born-Oppenheimer, single determinant, mean field, finite basis).
- A flow chart of the SCF cycle is almost always expected; draw it with a "converged?" decision box.
- Write the Slater determinant for two electrons, \(\Psi = \tfrac{1}{\sqrt2}[\chi_1(1)\chi_2(2) - \chi_2(1)\chi_1(2)]\), to show antisymmetry quickly.
- Practise counting basis functions: s = 1, p = 3, Cartesian d = 6 (spherical 5).
- Decode Pople names digit by digit: core primitives, then inner and outer valence primitives, then *, ** or +.
- Always state that E_HF ≥ E_exact and define E_corr with its sign.
FAQs
Is Hartree-Fock still used if DFT is better?
Yes. HF is the starting point for accurate correlated methods (MP2, coupled cluster), its exact exchange is part of hybrid DFT functionals, and it is a clean model for understanding orbitals. For final energies, DFT or correlated methods are usually preferred.
Why can the SCF calculation fail to converge?
A poor initial guess, nearly degenerate orbitals or a strange geometry can make the density oscillate between cycles. Programs use damping, level shifting and DIIS extrapolation to help convergence.
Does a bigger basis set always give better results?
It always gives a lower or equal energy for the same method, and usually better properties, but not always better agreement with experiment, because errors of the method may stop cancelling. Diffuse functions can also cause numerical problems. Match the basis to the method and property.
What is the difference between 6-31G* and 6-31G(d)?
Nothing; they are two names for the same basis set. Similarly 6-31G** equals 6-31G(d,p).
These notes follow the University of Mumbai first-year engineering syllabus (NEP 2020, Introduction to Computational Chemistry) and were prepared with AI help and checked.