Direct and Indirect Band Gaps

Introduction

Two semiconductors can have similar gap energies but very different light-emission behaviour. The difference often lies in where the valence-band maximum and conduction-band minimum occur on a graph of energy against crystal wave vector k. A photon carries energy but relatively little crystal momentum on the scale of a typical Brillouin zone. The geometry of the band edges therefore controls whether an electron can cross the gap by exchanging only a photon or also needs a lattice vibration.

Core explanation

In a direct-gap semiconductor, the conduction-band minimum and valence-band maximum occur at the same k. Near-edge absorption or radiative recombination can make a nearly vertical transition on an E–k plot: the photon provides energy while crystal momentum is approximately conserved. In an indirect-gap semiconductor, the band edges occur at different k values. A near-edge transition needs a phonon to supply or remove the missing crystal momentum in addition to the photon. MIT's optical-semiconductor lecture distinguishes direct and phonon-assisted indirect processes.

The gap type is a property of the electronic band structure, not of whether a drawn crystal bond is “straight.” Silicon's common diamond-cubic phase has an indirect fundamental gap; gallium arsenide has a direct gap under ordinary conditions. In silicon, light emission from band-edge electron–hole recombination is relatively inefficient because the phonon-assisted radiative event competes with nonradiative paths. This does not mean silicon never emits light, nor does “direct gap” guarantee a good LED. Crystal defects, surfaces, carrier confinement, composition and competing nonradiative recombination still matter.

For absorption, a direct allowed transition commonly has a stronger onset near its edge than an indirect phonon-assisted transition. An indirect material can still absorb strongly at higher photon energies through other direct transitions; the fundamental gap is not a cutoff beyond which every photon is absorbed with the same probability. Optical selection rules also depend on symmetries of states and light polarization. A photon energy hν must at least match the energy difference for the relevant transition, with phonon absorption or emission modifying the threshold for an indirect process. The phonon's energy is often smaller than electronic gap energies but cannot always be neglected in careful spectroscopy.

Momentum here is crystal momentum associated with periodic-lattice states, defined modulo a reciprocal-lattice vector. A phonon changes the electron's k while conserving total momentum of electron, photon and lattice. This explanation assumes a sufficiently ordered crystal for E(k) to be meaningful. In amorphous materials, disorder relaxes the simple k-selection picture; optical-gap analysis uses different approximations. MIT's compound-semiconductor course places band structures, direct/indirect gaps and light emission together.

Step-by-step reasoning

1. Sketch conduction and valence energy as functions of k. 2. Locate the absolute conduction minimum and valence maximum, not arbitrary points. 3. Compare their k coordinates: same is direct, different is indirect. 4. For a near-edge optical transition, decide whether a phonon is needed for momentum balance. 5. Discuss likely optical strength while checking nonradiative recombination and defects.

Visual explanation

Draw two E–k panels. In the direct panel, the valence maximum and conduction minimum align vertically; a vertical arrow labelled photon joins them. In the indirect panel, the minimum is displaced horizontally; an oblique two-step annotation labels both photon energy and phonon momentum. The arrows are schematic: momentum conservation concerns the full interacting system.

Real-world analogy

Imagine moving from a lower platform to an upper one. If they align horizontally, a vertical lift is enough. If the upper platform is sideways, a separate sideways move is required. A photon acts mainly like the lift; a phonon supplies the sideways crystal-momentum change. The analogy should not be taken as real-space motion across a room: k is a reciprocal-space coordinate.

Real-world example

LED materials often use direct-gap III–V compounds because radiative recombination can occur efficiently near the band edges. Silicon dominates many electronic circuits for other reasons, but its indirect gap makes a simple silicon band-edge LED inefficient. Engineers can still build silicon photonic systems using external light sources or engineered structures; the band-gap classification describes the bulk transition pathway, not every device possibility.

Why?

Why is a phonon important only for the indirect near-edge transition? Energy conservation alone does not determine whether a transition is allowed. The initial and final Bloch states have different crystal momenta. The photon alone changes momentum too little on the relevant scale, whereas a lattice vibration can carry the difference. Direct transitions align the band-edge momenta and avoid this extra requirement.

Common misconception

An indirect gap is not “larger than a direct gap” by definition; directness describes k alignment, not numerical E g. A direct semiconductor is not guaranteed to glow brightly if defects provide fast nonradiative paths. A photon with energy greater than the indirect gap does not automatically use only the phonon-assisted fundamental transition; higher-energy direct transitions may become available.

Worked example

In a schematic E–k calculation, the valence maximum is at k = 0 and the conduction minimum at k = 0.8 nm⁻¹. Their minimum energy separation is 1.1 eV. This is an indirect 1.1 eV fundamental gap because the k coordinates differ. Suppose at k = 0 the conduction band lies 1.6 eV above the valence maximum. Then a vertical direct transition at k = 0 can start near 1.6 eV even though the fundamental gap is 1.1 eV. Both numbers describe distinct transitions and must not be interchanged.

Quick check

1. What extra excitation usually supplies the crystal-momentum difference in a near-edge indirect optical transition? Answer: A phonon, which is a quantised lattice vibration.

Exam focus

Always label the axes E and k. Identify the extrema before classifying the gap. Say “photon plus phonon” for a near-edge indirect optical transition, and distinguish fundamental gap from the lowest strong direct transition. If discussing device performance, mention defects and nonradiative recombination in addition to gap type.

Advanced insight

Strain, alloying or reduced dimensionality can move band extrema and even change whether the lowest gap is direct. Temperature can alter band energies and phonon populations, affecting absorption edges. Excitons—bound electron–hole pairs—can produce features near an optical onset; their presence means a measured onset need not equal a single-particle gap without correction.

Summary

Direct and indirect gaps differ by the crystal wave-vector alignment of band-edge states. Photons can readily support near-vertical direct transitions; indirect transitions usually need a phonon for momentum balance. This strongly influences absorption and light emission, though defects and device design remain essential.

Practice questions

1. Valence maximum and conduction minimum are both at k = 0. Classify the gap. Answer: Direct, because the extrema occur at the same crystal wave vector. 2. Does a 2 eV direct gap necessarily emit light more efficiently than every 1 eV indirect material in every device? Answer: No. Defects, surfaces, competing recombination and device structure also control efficiency. 3. Why can an indirect material show strong absorption above its fundamental gap? Answer: Higher-energy direct transitions may become available and absorb strongly. 4. Is k on an E–k plot a position inside the crystal? Answer: No. It labels crystal wave vector in reciprocal space.