Questions & explanations
1. Compare the rules for one-electron and two-electron operators.
For one-electron operators (like kinetic energy or nuclear attraction), the Slater-Condon rules are simpler: if the determinants are the same, the matrix element is the sum of one-electron integrals over occupied spin-orbitals. If they differ by one orbital, it's the one-electron integral between the two differing orbitals. If they differ by two or more, the matrix element is zero. For two-electron operators (like electron repulsion), the rules are more complex. For the same determinant, it's the sum of Coulomb minus exchange integrals over all pairs. For one orbital difference, you sum over common orbitals Coulomb-minus-exchange integrals involving the differing orbitals. For two orbital differences, you get a single integral like (ab|cd) minus (ab|dc) with sign. The key difference is that two-electron operators allow nonzero matrix elements for up to two orbital differences, while one-electron operators allow at most one.
2. Compare the Born criterion with the Lindemann criterion for describing when a solid loses stability.
The Born criterion checks the static stiffness of the whole lattice: if any elastic constant goes to zero, the solid is unstable. It is a condition on the second derivatives of energy with respect to strain. The Lindemann criterion focuses on the motion of individual atoms: when their thermal vibration amplitude reaches about 10% of the inter‑atom distance, the solid melts. Born is often used to explain pressure‑induced solid‑solid transitions, while Lindemann is older and more common for thermal melting. Both can predict a loss of order, but Born can sometimes signal a mechanical instability even below the melting point, leading to a different crystal phase. In real materials, the two criteria often align, but they describe the onset of disorder from different angles—one from elasticity, the other from vibration size.
3. Compare internally contracted and externally contracted multireference methods in terms of accuracy.
Both internally and externally contracted methods reduce the computational cost of multireference calculations, but they differ in how they contract configurations. Internal contraction groups excitations within an internal orbital space, preserving more correlation among the active electrons. External contraction, on the other hand, contracts the external (virtual) part, often leading to larger errors for static correlation. Generally, internally contracted methods are more accurate for systems with strong static correlation because they retain more flexibility in the active space. External contraction might be faster but can miss important correlation effects. For quantitative accuracy, internal contraction is usually preferred, while external contraction is used for large systems where speed is critical.
4. Give an example using the Gibbs phase rule for a two‑component system like salt water that can form ice, liquid, and vapor.
Consider a mixture of salt (NaCl) and water. This is a two‑component system. In a closed container, if we have a single liquid phase, F = 2 minus 1 plus 2 = 3, meaning we can freely change temperature, pressure, and salt content. If ice starts to form from the salty water, we have two phases (solid ice and liquid), so F drops to 2 minus 2 plus 2 = 2; we can still change two things, like temperature and pressure, but the salt concentration in the liquid is then fixed by the temperature. If vapor also appears (three phases), F becomes 2 minus 3 plus 2 = 1. Finally, if all four possible phases—ice, salt crystals, liquid, and vapor—could exist together, F would be 2 minus 4 plus 2 = 0, meaning that state could only happen at one exact temperature, pressure, and liquid composition, which is a quadruple point.
5. Why is internal contraction especially useful for treating bond breaking or diradicals?
Bond breaking and diradicals have multiple near-degenerate electronic configurations, so a single-reference method fails. Multireference methods capture this static correlation by using several reference determinants. Internal contraction keeps the wavefunction compact while still including all important configurations from the active space. This allows calculations on molecules where bonds are stretched, like the dissociation of N2 or F2, to be done accurately but at a lower cost. Without contraction, the number of configurations explodes, making the calculation impractical. Internal contraction thus makes multireference calculations feasible for moderate-sized diradicals and bond-breaking reactions. It provides a balanced description of static and dynamic correlation needed for these challenging cases.
6. Explain the Franck-Condon principle in terms of potential energy curves for a diatomic molecule.
The Franck-Condon principle states that during an electronic transition, the nuclei hardly move because they are much heavier than electrons. So on a potential energy diagram (bond length vs energy), the transition is vertical: the molecule goes from the ground state to an excited state at the same internuclear distance. The most intense transition corresponds to the largest overlap between the vibrational wavefunctions of the initial and final states. For example, if the excited state has a longer equilibrium bond length, the vertical transition from v=0 of the ground state often ends at a high vibrational level of the excited state. This gives a vibrational progression in the absorption spectrum, with intensities following the Franck-Condon factors (square of overlap integrals).
7. Why are local methods not as accurate as canonical methods for describing delocalized electrons?
Local methods assume that correlation is short-ranged, but delocalized electrons (e.g., in conjugated systems or metals) have long-range correlation that is not captured by local pair approximations. For example, in a polyene chain, the π-electrons are delocalized over many atoms, and their correlation extends over long distances. Local methods that truncate pairs based on distance may miss important contributions from distant but strongly correlated electrons. This can lead to errors in relative energies, barriers, and properties. Canonical methods treat all pairs exactly, so they are more accurate for delocalized systems. Therefore, local methods are best applied to molecules with localized electron pairs, like saturated hydrocarbons or systems with strong single bonds.
8. Give an example of how the Lindemann criterion is used to estimate a melting curve under high pressure.
Under high pressure, atoms are squeezed closer together, which changes the vibration frequencies and the Debye temperature. By measuring how the sound speeds change with pressure, we can track the Debye temperature. The Lindemann criterion then predicts that melting will occur when the ratio of vibration amplitude to inter‑atom distance crosses the critical value. As pressure goes up, the bonds stiffen, the Debye temperature rises, and the needed melting temperature also rises. Scientists use this method to estimate melting points deep inside planets where we cannot measure directly. For example, they apply it to iron at Earth's core conditions, helping to map out the solid‑liquid boundary without needing a melting experiment at those extreme pressures and temperatures.
9. Compare the Runge-Lenz vector with angular momentum. How are they different?
Angular momentum L describes the rotational motion and is conserved in any central potential. The Runge-Lenz vector A is conserved only in the specific inverse-square law (Coulomb or gravitational) potential. L is an axial vector (like a vector pointing perpendicular to the orbit), while A lies in the plane of the orbit and points from the force center to the point of closest approach. L and A are both constants of motion for hydrogen, but they have different commutation relations. Together they generate the full symmetry group. While L determines the shape of the orbit (elliptical), A determines its orientation and eccentricity. In quantum mechanics, the eigenvalues of L give the angular momentum quantum number l, while A is related to the runge-lenz vector operator.
10. Compare Cholesky decomposition with density fitting (RI) in terms of accuracy.
Both Cholesky decomposition and the Resolution of the Identity (RI) approximation try to compress the ERI tensor. RI uses an auxiliary basis set to expand products of orbitals, leading to errors typically on the order of microHartree per atom or less. Cholesky decomposition is a pure numerical procedure that works with the actual integrals and does not need an auxiliary basis. The accuracy of Cholesky decomposition is controlled by a threshold (e.g., 10^-6) and can be made arbitrarily high, but at the cost of more vectors. Generally, Cholesky is slightly more accurate for a given computational effort because it adapts to the specific molecule. However, RI is faster and more widely used for routine calculations. Both are excellent approximations for most applications.
11. Compare the Lindemann criterion with the Born criterion in the context of melting.
The Lindemann criterion looks at how atoms move—when the shaking gets too big, melting happens. It is a kinematic rule. The Born criterion, on the other hand, looks at the mechanical stiffness of the crystal. It says a solid melts when one of its elastic constants (measures of resistance to squeezing or shearing) drops to zero, meaning the lattice becomes unstable against a deformation. Lindemann is historically older and easier to use because it only needs vibration data. Born is more fundamental for some systems, especially when melting is driven by a sudden softening of the structure. In many cases, both criteria give similar trends, but Born can also signal transitions without melting, like a solid‑to‑solid phase change, while Lindemann is purely a melting rule.
12. What is the Maxwell construction in the study of phase changes?
The Maxwell construction is a method to fix a problem that appears in some simple theories of gases and liquids. The van der Waals equation of state, for example, gives an S‑shaped curve with a wiggly part in the pressure‑volume graph at temperatures below the critical point. That wiggly part suggests that the volume could get bigger while the pressure also goes up, which is physically impossible for a stable fluid. The Maxwell construction replaces that wiggly loop with a flat, horizontal line. The flat line gives the true pressure at which gas and liquid can live side by side in equilibrium. The rule is that the area above and below the flat line inside the loop must be equal. This way, the model correctly shows a constant pressure during boiling or condensation.