Approximation Methods: Variational Principle and Perturbation Theory

In quantum mechanics, obtaining exact solutions to the Schrödinger equation is possible only for very simple systems, such as the hydrogen atom or the harmonic oscillator. For more complex systems, like multi-electron atoms or molecules, the Schrödinger equation becomes too difficult to solve analytically. In such cases, we rely on approximation methods to find solutions that are close to the exact ones. The two most fundamental and widely used approximation methods are the Variational Principle and Perturbation Theory.

The Need for Approximation Methods

The Hamiltonian operator (Ĥ), which represents the total energy of a system, becomes increasingly complex with the number of particles and interactions. For instance, the Hamiltonian for a helium atom includes kinetic energy terms for both electrons, electron-nucleus attraction terms for both electrons, and electron-electron repulsion terms. This latter term, the electron-electron repulsion, makes the Schrödinger equation non-separable and analytically intractable. Approximation methods allow us to tackle these complex problems by building upon the solutions of simpler, solvable systems or by making educated guesses about the true wave function.

1. Variational Principle

The Variational Principle is a powerful method used to estimate the ground state energy of a system. It is based on the idea that if you use a trial wave function (Ψ_trial) that is a reasonable approximation to the true ground state wave function (Ψ₀), the expectation value of the energy calculated using this trial function will be greater than or equal to the true ground state energy (E₀).

The Mathematical Formulation

The expectation value of the energy, denoted by ⟨E⟩, is calculated using the formula:

⟨E⟩ = ∫ Ψ_trial^* Ĥ Ψ_trial dτ / ∫ Ψ_trial^* Ψ_trial dτ

Here, Ĥ is the Hamiltonian operator of the system, Ψ_trial is the trial wave function, Ψ_trial^* is its complex conjugate, and the integral is taken over all space (dτ represents the volume element).

The Variational Principle states that:

⟨E⟩ ≥ E₀

where E₀ is the true ground state energy.

How it Works

The core idea is to choose a trial wave function that depends on one or more adjustable parameters. We then calculate the expectation value of the energy ⟨E⟩ as a function of these parameters. To obtain the best possible approximation to the ground state energy, we minimize ⟨E⟩ with respect to these parameters. The minimum value obtained will be the closest possible approximation to the true ground state energy, and the corresponding trial wave function will be the best approximation to the true ground state wave function among all functions of that chosen form.

Steps for Applying the Variational Method

1. Choose a Trial Wave Function (Ψ_trial): Select a function that has the correct symmetry properties and boundary conditions of the system. It often incorporates adjustable parameters (e.g., α, β). For example, for a particle in a box, one might try a Gaussian function instead of the sine function.
2. Calculate the Hamiltonian (Ĥ): Write down the Hamiltonian operator for the system you are studying.
3. Calculate the Expectation Value of Energy (⟨E⟩): Compute the integral ⟨E⟩ = ∫ Ψ_trial^* Ĥ Ψ_trial dτ / ∫ Ψ_trial^* Ψ_trial dτ. This often involves integrating over all spatial coordinates.
4. Minimize ⟨E⟩: Treat ⟨E⟩ as a function of the adjustable parameters. Take the partial derivative of ⟨E⟩ with respect to each parameter and set it to zero. Solve the resulting equations to find the optimal values of the parameters.
5. Calculate the Variational Energy: Substitute the optimal parameter values back into the expression for ⟨E⟩. This minimum value is your best estimate for the ground state energy (E₀).

Example: Particle in a 1D Box

Consider a particle of mass 'm' in a one-dimensional box of length 'L' (from x=0 to x=L). The true ground state wave function is Ψ₁(x) = √(2/L) sin(πx/L) and the ground state energy is E₁ = h² / (8mL²).

Let's try a variational approach with a simple trial function, for instance, a parabolic function that is zero at the boundaries:

Ψ_trial(x) = C * x * (L - x)

where C is a normalization constant. First, normalize Ψ_trial:

∫₀^L |C * x(L-x)|² dx = 1

C² ∫₀^L (x²L² - 2x³L + x⁴) dx = 1

C² [ L²x³/3 - 2Lx⁴/4 + x⁵/5 ]₀^L = 1

C² [ L⁵/3 - L⁵/2 + L⁵/5 ] = 1

C² L⁵ [ (10 - 15 + 6) / 30 ] = 1

C² L⁵ (1/30) = 1 => C² = 30/L⁵ => C = √(30/L⁵)

So, Ψ_trial(x) = √(30/L⁵) * x(L-x).

The Hamiltonian for a free particle is Ĥ = - (ħ² / 2m) d²/dx².

Now calculate ⟨E⟩ = ∫₀^L Ψ_trial^* Ĥ Ψ_trial dx (since it's a real function, Ψ_trial^* = Ψ_trial).

First, calculate ĤΨ_trial:

dΨ_trial/dx = C * (L - 2x)

d²Ψ_trial/dx² = C * (-2)

ĤΨ_trial = - (ħ² / 2m) * C * (-2) = (ħ²C / m)

Now, the integral for ⟨E⟩:

⟨E⟩ = ∫₀^L [C * x(L-x)] * (ħ²C / m) dx

⟨E⟩ = (ħ²C² / m) ∫₀^L (Lx - x²) dx

⟨E⟩ = (ħ²C² / m) [ Lx²/2 - x³/3 ]₀^L

⟨E⟩ = (ħ²C² / m) [ L³/2 - L³/3 ]

⟨E⟩ = (ħ²C² / m) [ L³/6 ]

Substitute C² = 30/L⁵:

⟨E⟩ = (ħ² / m) * (30/L⁵) * (L³/6)

⟨E⟩ = (ħ² / m) * (5/L²) = 5ħ² / (mL²)

Since ħ = h / 2π, ħ² = h² / (4π²).

⟨E⟩ = 5 * (h² / 4π²) / (mL²) = 5h² / (4π²mL²)

The true ground state energy is E₁ = h² / (8mL²). Our variational estimate is ⟨E⟩ = 5h² / (4π²mL²). Note that π² ≈ 9.87. So, 4π² ≈ 39.48.

⟨E⟩ ≈ 5h² / (39.48 mL²) ≈ 0.127 h² / (mL²).

The true energy is E₁ = 0.125 h² / (mL²).

As expected, ⟨E⟩ > E₁. This trial function gives a reasonable approximation, but not perfect.

2. Perturbation Theory

Perturbation theory is used when the Hamiltonian of a system can be split into two parts: a solvable part (Ĥ₀) and a small, non-solvable part (Ĥ'). This small part is called the "perturbation." The method allows us to find approximate solutions for the energy levels and wave functions of the perturbed system by starting from the known exact solutions of the unperturbed system.

The Mathematical Formulation

Let the total Hamiltonian be Ĥ = Ĥ₀ + Ĥ'.

The unperturbed Schrödinger equation is Ĥ₀ Ψ_n⁽⁰⁾ = E_n⁽⁰⁾ Ψ_n⁽⁰⁾, where Ψ_n⁽⁰⁾ and E_n⁽⁰⁾ are the known wave functions and energies of the unperturbed system.

The perturbed Schrödinger equation is Ĥ Ψ_n = E_n Ψ_n.

We express the perturbed energy and wave function as a power series in a dimensionless parameter 'λ', which represents the strength of the perturbation (λĤ'). As λ approaches 0, the system becomes unperturbed.

E_n = E_n⁽⁰⁾ + λE_n⁽¹⁾ + λ²E_n⁽²⁾ + ...

Ψ_n = Ψ_n⁽⁰⁾ + λΨ_n⁽¹⁾ + λ²Ψ_n⁽²⁾ + ...

where E_n^(k) and Ψ_n^(k) are the k-th order corrections to the energy and wave function, respectively.

First-Order Perturbation Theory (Non-Degenerate Case)

Assuming the energy levels of the unperturbed system are non-degenerate (each energy level is unique).

Energy Correction: The first-order energy correction is given by the expectation value of the perturbation Hamiltonian with respect to the unperturbed wave function:

E_n⁽¹⁾ = ⟨Ψ_n⁽⁰⁾ | Ĥ' | Ψ_n⁽⁰⁾⟩ = ∫ Ψ_n^(0)* Ĥ' Ψ_n⁽⁰⁾ dτ

This means the first-order corrected energy is E_n ≈ E_n⁽⁰⁾ + E_n⁽¹⁾.

Wave Function Correction: The first-order wave function correction is given by:

Ψ_n⁽¹⁾ = Σ_m≠n [ ⟨Ψ_m⁽⁰⁾ | Ĥ' | Ψ_n⁽⁰⁾⟩ / (E_n⁽⁰⁾ - E_m⁽⁰⁾) ] Ψ_m⁽⁰⁾

The term ⟨Ψ_m⁽⁰⁾ | Ĥ' | Ψ_n⁽⁰⁾⟩ is the matrix element of the perturbation between states m and n. The denominator (E_n⁽⁰⁾ - E_m⁽⁰⁾) is the energy difference between the states.

Second-Order Perturbation Theory (Non-Degenerate Case)

The second-order energy correction is:

E_n⁽²⁾ = Σ_m≠n [ |⟨Ψ_m⁽⁰⁾ | Ĥ' | Ψ_n⁽⁰⁾⟩|² / (E_n⁽⁰⁾ - E_m⁽⁰⁾) ]

The second-order correction to the wave function, Ψ_n⁽²⁾, is more complex and involves terms from both the first-order correction and the perturbation.

Perturbation Theory for Degenerate Systems

When energy levels are degenerate (E_n⁽⁰⁾ = E_m⁽⁰⁾ for n ≠ m), the denominators in the perturbation formulas become zero, and the above formulas are invalid. In such cases, a different approach is needed, typically involving solving a secular determinant within the degenerate subspace to lift the degeneracy and find the correct first-order energy corrections.

Example: Helium Atom (as a conceptual illustration)

The helium atom has two electrons and a nucleus with charge Z=2. The Hamiltonian can be written as:

Ĥ = - (ħ² / 2m) ∇₁² - (ħ² / 2m) ∇₂² - (2e² / 4πε₀r₁) - (2e² / 4πε₀r₂) + (e² / 4πε₀r₁₂)

If we ignore the electron-electron repulsion term (e² / 4πε₀r₁₂), the Hamiltonian becomes separable into two hydrogen-like atom Hamiltonians (one for each electron), which is solvable. This is our unperturbed Hamiltonian Ĥ₀.

Ĥ₀ = [ - (ħ² / 2m) ∇₁² - (2e² / 4πε₀r₁) ] + [ - (ħ² / 2m) ∇₂² - (2e² / 4πε₀r₂) ]

The unperturbed energy would be the sum of two hydrogen-like energies for Z=2. For example, the ground state energy (1s²) would be 2 * E_1s(Z=2) = 2 * [-13.6 eV * Z² / 1²] = 2 * [-13.6 * 4] = -108.8 eV.

The perturbation is Ĥ' = (e² / 4πε₀r₁₂).

Using first-order perturbation theory, the correction to the ground state energy is:

E_ground⁽¹⁾ = ⟨Ψ_1s(1)Ψ_1s(2) | Ĥ' | Ψ_1s(1)Ψ_1s(2)⟩

This integral is challenging but can be calculated. It represents the average repulsion energy between the two electrons when they are in the 1s orbitals. The calculated value is approximately +34 eV.

So, the approximate energy of the helium atom is E_ground ≈ E_ground⁽⁰⁾ + E_ground⁽¹⁾ = -108.8 eV + 34 eV = -74.8 eV.

The experimentally observed ionization energy of Helium is about -79 eV. First-order perturbation theory gives a reasonable approximation, but to get closer, higher-order corrections or more sophisticated methods like the Variational method with a better trial function (e.g., including screening) are needed.

Comparison and Complementarity

Both the Variational Principle and Perturbation Theory are essential tools in quantum chemistry, but they are applied in different scenarios and have different strengths:

• Variational Principle: Best for estimating the ground state energy when the exact Hamiltonian is known but too complex to solve. It provides an upper bound and can be used to improve approximate wave functions.
• Perturbation Theory: Best when a system can be divided into a solvable part and a small perturbing part. It's good for understanding the effect of small changes or external influences and can be extended to excited states.

Often, these methods are used together. For example, a variational method might be used to get a good initial guess for the ground state wave function, which can then be used as the unperturbed wave function in perturbation theory to study the effect of further small perturbations.

Applications in Spectroscopy

These approximation methods are fundamental to understanding spectroscopic phenomena:

• Stark Effect: The splitting and shifting of spectral lines of atoms and molecules when placed in an external electric field. This is a classic example of perturbation theory, where the electric field is the perturbation (Ĥ').
• Zeeman Effect: The splitting and shifting of spectral lines in the presence of an external magnetic field. Again, the magnetic field acts as a perturbation.
• Vibrational and Rotational Spectroscopy: While basic models for molecular vibrations and rotations are solvable, real molecules have anharmonic vibrations and non-rigid rotations. Perturbation theory is used to account for these deviations from ideal models, leading to more accurate predictions of spectral transitions.
• Fine Structure and Hyperfine Structure: Interactions like spin-orbit coupling (fine structure) or spin-spin coupling (hyperfine structure) are often treated as small perturbations to the electronic structure calculated without these interactions.

In essence, approximation methods are the workhorses that allow quantum mechanics to be applied to the vast majority of real-world chemical systems, bridging the gap between theoretical models and experimental observations, particularly in spectroscopy.