Skip to main content

Welcome to a Grasp lesson

Create a free account to keep studying and create your own study path.

Lesson 21

Exciton Transitions in Cubic Crystals: A Group Theory Approach

Recommended resources

Hello! Welcome back to our course on exciton physics.

In our last lesson, we discovered that while the exchange interaction correctly predicts the energy ordering of ortho- and para-excitons, a simple spin model isn't enough to explain their optical properties, especially in a material like Cu₂O. We saw that the specific symmetries of the crystal lattice and the electronic bands—designated by labels like Γ⁺₆ and Γ⁺₇—are what truly govern the selection rules. This observation perfectly sets the stage for today's topic.

Our goal for this lesson is to learn how to apply group theory and character tables to determine dipole-allowed and forbidden exciton transitions in crystals with cubic symmetry (e.g., Cu₂O). We will develop a systematic and powerful method to predict which excitonic states can be created by light, finally explaining the puzzling optical spectrum of the yellow excitons in cuprous oxide.

1. The Language of Symmetry: Group Theory in Crystals

Your physics background has undoubtedly exposed you to group theory, particularly in quantum mechanics with rotation groups like SO(3). In solid-state physics, we apply the same principles to the discrete symmetries of a crystal lattice.

A crystal's structure is invariant under a set of symmetry operations (rotations, reflections, inversion), which form a mathematical group called a point group. For Cu₂O, which has a cubic lattice with an inversion center, the point group is Oₕ. The quantum states of any particle or quasiparticle within this crystal must transform according to the irreducible representations (irreps) of this group.

To familiarize yourself with the core concepts and terminology we'll be using, please start by reading an introductory section from a doctoral thesis on Rydberg excitons.

Rydberg excitons in external fields

This reading from F. Krüger's dissertation, 'Rydberg excitons in external fields', provides a concise but formal introduction to the group theory concepts relevant to solid-state physics. It defines key terms and outlines the connection between symmetry and quantum mechanics.

Please read Section 2.1, 'Group theory' (pages 4-10). Focus on: The definitions of a group and its representations. The idea that eigenstates of a symmetric Hamiltonian can be classified by the irreps of the symmetry group. The concept of the Wigner-Eckart theorem (Eq. 2.10), which provides the fundamental basis for selection rules.

The Master Key: Selection Rules from Group Theory

The central principle for determining if a transition is allowed is elegantly simple. A transition from an initial state to a final state driven by an interaction operator is allowed if, and only if, the matrix element is non-zero.

In the language of group theory, this matrix element is guaranteed to be zero unless the final state's irrep, , is contained within the direct product of the operator's irrep, , and the initial state's irrep, :

If the initial state is the crystal ground state, which is always totally symmetric (), the rule simplifies: a transition to state is allowed if its irrep is the same as the operator's irrep .

Our task, therefore, breaks down into three steps:

  1. Identify the irrep of the initial state (the crystal ground state).
  2. Identify the irrep of the interaction operator (the photon).
  3. Identify the irrep of the final state (the exciton).
  4. Apply the selection rule.

2. Assigning Symmetries in Cu₂O (Point Group Oₕ)

We'll use the Oₕ character table as our guide. This table is the "dictionary" that translates physical functions and states into the language of irreps. We will use the Koster notation (e.g., Γ⁻₄) which is common in semiconductor physics literature, but the table also shows the equivalent Schoenflies notation (e.g., T₁ᵤ).

Oh Character Table
The character table for the Oₕ point group. The first column lists the irreducible representations (irreps). The columns on the right indicate how basis functions like coordinates (x, y, z) or their products transform, allowing us to assign irreps to physical operators and wavefunctions.

Step 1 & 2: Symmetry of the Ground State and the Photon

  • Crystal Ground State (): The state with no excitons is the vacuum. It is invariant under all symmetry operations of the crystal. It therefore always belongs to the totally symmetric irrep, which is Γ⁺₁ (also A₁g).

  • Interaction Operator ():

    • Electric Dipole Transition: The interaction operator is proportional to the position vector . Looking at the basis functions in the character table, we see that transforms as the Γ⁻₄ irrep (T₁ᵤ). The parity is odd, as indicated by the minus sign (or 'u' for ungerade).
    • Electric Quadrupole/Magnetic Dipole Transition: These higher-order transitions involve operators that transform like second-rank tensors, such as or . From the table, these functions correspond to the even-parity irreps Γ⁺₃ (E_g) and Γ⁺₅ (T₂g).

Step 3: Symmetry of the Exciton State ()

The total symmetry of the exciton is given by the direct product of the symmetries of its components:

where CB is the conduction band, VB is the valence band, and 'env' is the exciton envelope function.

  • Band Symmetries in Cu₂O:

    • The lowest conduction band is formed from Cu 4s orbitals and has Γ⁺₆ symmetry.
    • The highest valence band originates from Cu 3d orbitals, split by spin-orbit coupling, and has Γ⁺₇ symmetry.
    • Crucially, both bands have even parity (indicated by the '+'). This means they are symmetric under inversion.
  • Envelope Function Symmetry: This depends on the relative angular momentum () of the electron-hole pair.

    • s-like envelope (L=0): The wavefunction is spherically symmetric (like a hydrogen 1s orbital). It is invariant under all rotations and has even parity. It transforms as Γ⁺₁.
    • p-like envelope (L=1): The wavefunction has three components that transform like the vector . It therefore has the symmetry of the dipole operator: Γ⁻₄. It has odd parity.

3. Applying the Selection Rules: Solving the Cu₂O Puzzle

Now we have all the pieces. We'll use the Oₕ multiplication table (found in Appendix B.3 of the Krueger dissertation you read, LINK) to evaluate the direct products.

Case 1: The 1s Yellow Exciton

  1. Find the Exciton Symmetry:

    From the multiplication table, .
    Since is the identity, the final symmetry is .
    (As we learned last time, corresponds to the para-exciton and to the ortho-exciton). Both states have even parity.

  2. Apply Dipole Selection Rule:

    • Initial State: Γ⁺₁
    • Operator: Γ⁻₄
    • Allowed Final States: Γ⁻₄ ⊗ Γ⁺₁ = Γ⁻₄
    • Conclusion: The dipole operator has odd parity (Γ⁻₄), but the 1s exciton states have even parity (Γ⁺₂ and Γ⁺₅). A transition between states of the same parity is forbidden by an odd-parity operator. Therefore, the 1s exciton is dipole-forbidden.
  3. Apply Quadrupole Selection Rule:

    • Initial State: Γ⁺₁
    • Operator: Γ⁺₃ ⊕ Γ⁺₅
    • Allowed Final States: (Γ⁺₃ ⊕ Γ⁺₅) ⊗ Γ⁺₁ = Γ⁺₃ ⊕ Γ⁺₅
    • Conclusion: The final state must have Γ⁺₃ or Γ⁺₅ symmetry. Our 1s ortho-exciton has Γ⁺₅ symmetry! Therefore, the 1s ortho-exciton transition is quadrupole-allowed. This explains why a very weak absorption line for this state can be measured. The 1s para-exciton (Γ⁺₂) remains forbidden.

Case 2: The 2p Yellow Exciton

  1. Find the Exciton Symmetry:

    Using the multiplication table again:

    So, the 2p excitons form a set of states with symmetries . All have odd parity.

  2. Apply Dipole Selection Rule:

    • Initial State: Γ⁺₁
    • Operator: Γ⁻₄
    • Allowed Final States: Γ⁻₄
    • Conclusion: The manifold of 2p exciton states includes one with Γ⁻₄ symmetry. Therefore, the 2p exciton is dipole-allowed.

This perfectly explains the experimental absorption spectrum of Cu₂O, where one sees a series of strong peaks starting from n=2 (the P-series), but the n=1 line is conspicuously absent.

The following resource provides a practical walkthrough of this exact process for one-photon and two-photon transitions in Cu₂O.

Nonlinear optical spectroscopy of exciton-polaritons

This section from M. Mund's dissertation, 'Nonlinear optical spectroscopy of exciton-polaritons...', applies the group theory formalism directly to the exciton selection rules in Cu₂O. It's an excellent, concrete example of the theory in practice.

Please read Section 5.3 'Exciton symmetry selection rules' and Appendix B 'Group theory'. Section 5.3: Follow the derivation of the S- and P-exciton symmetries (Eqs. 5.2, 5.3) and the symmetries of the one-photon operators (Eqs. 5.4, 5.5). This will reinforce the steps we just took. Appendix B: This provides a full, step-by-step derivation for the 1S exciton, including the spin structure. It's a great summary that connects back to our previous lesson on ortho/para excitons.

Test your understanding!

Using the method we've developed, determine if the 2s exciton in Cu₂O is dipole-allowed.

Show answer

The 2s exciton has an s-like envelope function (L=0), which has Γ⁺₁ symmetry. The calculation of the total exciton symmetry is therefore identical to the 1s case: . These are even-parity states. The dipole operator has odd parity (Γ⁻₄). Therefore, the 2s exciton is also dipole-forbidden. In general, all s- and d-envelope excitons in Cu₂O will be dipole-forbidden, while all p- and f-envelope excitons will be dipole-allowed. This is often called a "parity" selection rule.


Conclusion

Today, we have demystified the optical selection rules in Cu₂O by employing the powerful and rigorous framework of group theory. We've moved beyond simple spin models to a complete description that accounts for the crystal's symmetry.

Key Takeaways:

  • Quantum states in a crystal are classified by the irreducible representations (irreps) of the crystal's point group.
  • A transition is allowed if the final state's irrep is contained in the direct product of the operator's and initial state's irreps.
  • In Cu₂O (Oₕ symmetry), the electric dipole operator has odd parity (Γ⁻₄). Since the valence and conduction bands both have even parity, an exciton's parity is determined solely by its envelope function.
  • s-like excitons (1s, 2s, ...) have even-parity envelopes (Γ⁺₁), leading to even-parity exciton states. These are dipole-forbidden.
  • p-like excitons (2p, 3p, ...) have odd-parity envelopes (Γ⁻₄), leading to odd-parity exciton states. These are dipole-allowed.
  • The 1s ortho-exciton (Γ⁺₅) is weakly allowed via the electric quadrupole interaction (operator contains Γ⁺₅).

Preview of the Next Lesson

We have now established the "zero-order" energy levels and selection rules based on the crystal's symmetry. However, finer effects can further split these levels and modify their properties. In our next lesson, we will investigate these effects by learning how crystal field and spin-orbit coupling modify the exciton fine structure. This will add another layer of detail to our model, bringing us even closer to a complete picture of exciton physics in real materials.