TL;DR
This paper provides a direct experimental protocol to measure the "spread" (spatial localization) of valence band electrons in solids using standard optical spectroscopy. By linking the geometric properties of electron wavefunctions (the quantum metric) to optical sum rules, the authors demonstrate how to extract this fundamental material characteristic from dielectric functions in 3D materials and absorbance data in 2D materials.
What Problem Does This Paper Address?
In condensed matter physics, we often describe electrons in crystals using Wannier functions, which are essentially localized "packets" of charge. The spread of these functions—how much they leak out from one atom to its neighbors—is a fundamental measure of how insulating a material is.
While theoretically vital, calculating this spread usually requires complex first-principles simulations. There has been a lack of a simple, "universal" experimental method to measure it. The authors address this by connecting the Wannier spread to quantum geometry. Specifically, they look at the "quantum metric," a mathematical description of how electron wavefunctions overlap in momentum space. Since this metric also dictates how electrons transition between bands when hit by light, it should be hidden within the material's optical response.
Methods
The authors derive a theoretical bridge between the gauge-invariant part of the Wannier spread ($\Omega_I$) and the momentum-integrated quantum metric (which they term the fidelity number, $\mathcal{G}_{\mu\nu}$).
They propose two distinct experimental protocols based on optical sum rules:
- For 3D Materials: The spread is proportional to the frequency-integral of the imaginary part of the dielectric function ($\text{Im}[\epsilon]$).
- For 2D Materials: The spread is obtained by dividing the absorbance (opacity) by the frequency and integrating over the frequency range.
The researchers validated these protocols using a mix of existing experimental data for semiconductors (Si, Ge) and topological insulators ($Bi_2Te_3$), alongside tight-binding numerical models for 2D systems like graphene, twisted bilayer graphene (TBG), and transition metal dichalcogenides (TMDs).
Datasets
The study utilized several data sources:
- Experimental Dielectric Data: For 3D materials, they used established experimental values for Silicon (Si), Germanium (Ge), and Bismuth Telluride ($Bi_2Te_3$) from classic literature (Refs 15-17).
- New Experimental Data: The authors synthesized centimeter-scale $WS_2$ on fused silica via chemical vapor deposition (CVD) and measured its absorbance using a UV-Vis-NIR spectrophotometer.
- Numerical Models: They employed tight-binding models to simulate the electronic structures of monolayer graphene (with spin-orbit coupling), twisted bilayer graphene, and six different TMDs ($MoS_2$, $WS_2$, etc.).
Key Insights and Ideas
- Optical Sum Rules as Geometric Probes: The most significant "aha moment" is that the total area under an optical absorption curve (when weighted correctly) directly reveals the ground-state geometry of the valence electrons.
- The "Insulatingness" Ratio: The authors propose a dimensionless ratio ($\Omega_I^{3/2}/V_{cell}$) as a figure of merit. A smaller ratio indicates a more "perfect" insulator. They found $Bi_2Te_3$ has a much higher ratio (~15.08) than Si (~3.51), reflecting the more extended, "leaky" nature of electrons in topological materials.
- Detecting Flat Bands in TBG: In twisted bilayer graphene, the formation of "flat bands" at the magic angle (~1.1°) causes a massive spike in the quantum metric. The authors show this should manifest as a distinct absorbance peak in the millimeter-wave (microwave) range, providing a new way to "see" magic-angle physics.
- Substrate Influence: By comparing models to their $WS_2$ experiment, they found that the substrate (fused silica) actually reduces the Wannier spread, suggesting that the environment can "squeeze" the electronic wavefunctions.
Key Results
- 3D Spread Values: They successfully extracted absolute spread values: Si ($\sim 68 \text{ \AA}^2$), Ge ($\sim 83 \text{ \AA}^2$), and $Bi_2Te_3$ ($\sim 407 \text{ \AA}^2$).
- Graphene & ISOC: The model predicts that intrinsic spin-orbit coupling (ISOC) in graphene—usually considered negligible—actually cures a mathematical divergence in the quantum metric and creates a finite Wannier spread detectable in the microwave range.
- TMD Consistency: For TMDs like $MoS_2$, the ratio of Wannier spread to unit cell area was consistently between 0.6 and 0.8, showing that these materials have fairly localized electrons.
- Experimental Agreement: Their $WS_2$ experimental data matched their low-energy theoretical models up to ~2.5 eV, validating the protocol for real-world 2D materials.
Limitations and Future Work
- Exciton "Poisoning": The theory is based on a "Fermi sea" of independent particles. In real experiments, excitons (bound electron-hole pairs) create sharp absorption peaks that aren't related to the ground-state Wannier spread. The authors note that these peaks must be manually "excluded" or subtracted from the integral to get accurate results.
- High-Energy Contributions: Simple tight-binding models miss the contributions of higher-energy orbitals (like the $p$-orbitals in sulfur). A more complete picture would require integrating over a wider spectral range.
- Correlated Systems: A promising future direction is applying this to strongly correlated materials where the relationship between quantum geometry and phenomena like superconductivity is still a hot topic of debate.
Summary updated