Scanning Tunneling Energy Dispersion Spectroscopy

• Scanning Tunneling Energy Dispersion Spectroscopy (STEDS) maps electronic band structures using STM conductance measurements.
• STEDS enables precision mapping of energy-momentum dispersion in low-dimensional quantum systems with superior local sensitivity.
• Applications include graphene, spin chains, and superconductors, revealing constant-energy contours and quasiparticle excitations.

Scanning Tunneling Energy Dispersion Spectroscopy (STEDS) is a spectroscopic technique that utilizes the spatially resolved differential conductance measurements obtained via scanning tunneling microscopy (STM) to reconstruct the energy–momentum dispersion relations in low-dimensional quantum systems. STEDS couples high-resolution, local STM-derived spectroscopy with rigorous momentum-space analysis, enabling detailed mapping of electronic (and, in some cases, spin) band structures, constant-energy contours, quasiparticle interference (QPI), and collective excitations in surfaces, nanostructures, and chains. Its principal strength lies in atomically resolved access to both filled and empty states, as well as the ability to probe electronic structure modifications introduced by impurities, defects, or finite geometry—all with local sensitivity surpassing ensemble-averaging techniques such as angle-resolved photoemission spectroscopy (ARPES).

1. Experimental Protocols and Data Acquisition

STEDS initiates with acquisition of dense spatial grids of the tunneling differential conductance, $dI/dV(x,y;E)$ (or for spin excitations, $d^2I/dV^2$), while scanning over a region containing point defects, step edges, boundaries, or within fabricated finite systems such as nanoribbons or chains. The sample under study is prepared under ultra-high vacuum conditions and cooled—temperatures of 5–15 K are typical for electronic systems (Söde et al., 2014), with specific requirements for spin excitations (Henriques et al., 19 Feb 2025).

In the case of two-dimensional materials or surfaces, a small AC modulation (typically 10–30 mV) is superimposed on the DC bias, and lock-in detection is used to record the energy-resolved local density of states (LDOS) proportional to $dI/dV$ on a fine real-space grid (Simon et al., 2011). For one-dimensional chains or nanoribbons, the tip is scanned along the system edge or axis, recording $dI/dV(V,x)$ as a function of position. In spin systems, the relevant observable is $d^2I/dV^2(x,\omega)$ via inelastic electron tunneling spectroscopy (IETS) protocols (Henriques et al., 19 Feb 2025).

The resultant set of $dI/dV(x,y;E)$ images at multiple energies constitutes the essential input for subsequent Fourier analysis.

2. Fourier Analysis and Momentum-Space Mapping

The core operation in STEDS is the discrete or continuous Fourier transform of the spatially resolved conductance map at fixed energy:

$$\text{FT}(q;E) = \int\int dI/dV(x,y;E)\;e^{-i\mathbf{q}\cdot\mathbf{r}}\ d^2r$$

For chain-like systems, the transformation reduces to:

$$S(k,\omega) = \sum_{x=0}^{L-1} e^{-ikx} \frac{d^2 I}{dV^2}(x,\omega)$$

This transformation reveals features in $q$-space (or $k$-space) that correspond to scattering-induced standing waves or interference patterns produced by electronic or spin excitations reflecting off system boundaries or defects.

$d^2I/dV^2$0 (or $d^2I/dV^2$1) manifests ridges, rings, or contour features whose location and shape are direct signatures of constant-energy contours (CEC), Fermi surfaces, or excitation dispersions. Key interpretation steps involve:

1. Identifying the dominant wavevector(s) $d^2I/dV^2$2 (or $d^2I/dV^2$3) at each energy.
2. Mapping $d^2I/dV^2$4 or $d^2I/dV^2$5 versus $d^2I/dV^2$6, frequently via $d^2I/dV^2$7 for simple standing waves (Söde et al., 2014, Simon et al., 2011).
3. Tracking the evolution and dispersion of these features to reconstruct $d^2I/dV^2$8 relations, effective masses, or excitation gaps.

Isotropic Gaussian averaging (IGA) and other filtering techniques are applied to enhance S/N ratios and suppress high-frequency noise or artifacts in the Fourier-transformed data (Lee, 2012).

3. Theoretical Frameworks for Interpretation

The theoretical interpretation of STEDS data draws on multiple levels of approximation:

• Lindhard Susceptibility: In itinerant electron systems, the momentum-dependent susceptibility $d^2I/dV^2$9 shows singularities at $dI/dV$0 for a 2D electron gas, producing the hallmark circular QPI rings observed in metallic surfaces (Simon et al., 2011).
• Joint Density of States (JDOS): The JDOS approximation simplifies wavevector analysis to geometric self-correlations of the constant-energy contour:

$dI/dV$1

• Stationary Phase Condition: Dominant scattering vectors are those connecting points on the CEC with parallel group velocities, refining the JDOS with selection rules (Simon et al., 2011).
• T-matrix Approximation: Incorporates all-order impurity scattering, phase effects, and matrix element structure:

$dI/dV$2

The T-matrix framework is essential for treating sublattice, pseudospin, and chirality effects (e.g., graphene intervalley vs. intravalley scattering) and yields quantitative QPI intensity maps (Simon et al., 2011, Lee, 2012).

• Derivative Rule and Tip-Orbital Decomposition: For quantitative STS simulations, revised derivative rules enable decomposition of $dI/dV$3 into orbital-resolved and interference contributions, identifying tip–sample orbital specificity and distinguishing direct from interference fingerprints in $dI/dV$4 (Abilio et al., 15 Apr 2025).

4. Application Domains and Case Studies

Graphene and Nanoribbons

STEDS provides direct, in-situ access to the E–k dispersion, effective masses, and subband structure in armchair graphene nanoribbons (7-AGNRs), with energy resolution and precision superior to conventional single-point spectroscopy (Söde et al., 2014). The methodology enables mapping of multiple bands (VB, CB, CB+1), gap extraction with $dI/dV$50.06 eV precision, and unambiguous subband assignment through density functional theory (DFT) comparison.

Spin Chains and Magnetic Excitations

In finite spin systems, spatially resolved STM-IETS and STEDS yield the magnon or triplon band dispersion when excitations form standing waves—the measured $dI/dV$6 exhibits sharp peaks or ridges tracing $dI/dV$7 (Henriques et al., 19 Feb 2025). However, in cases where excitations are fractionalized (e.g., spinon continua in uniform Heisenberg $dI/dV$8 chains), the Fourier transform produces a broad continuum rather than a well-defined dispersion. The method is validated by agreement with experimental data for dimerized nanographene chains.

Quasi-2D Electron Systems and Spin-Orbit Coupling

STEDS applied to a Cs-induced two-dimensional electron system (2DES) on InSb(110) quantitatively recovers non-parabolic $dI/dV$9 relations and captures Rashba spin splitting via analysis of Landau level beating and spatially resolved standing-wave FFTs (Becker et al., 2010). Quantitative extraction of spin–orbit parameters and disorder scales is possible.

Strongly Correlated Superconductors

STEDS, with high-resolution cross-sectional FTSTS and comprehensive noise treatment, robustly reconstructs Bogoliubov quasiparticle dispersions and diagnoses competing orders or phase transitions in cuprate and pnictide high-Tc superconductors, even with incomplete information on tunneling and scattering matrix elements (Lee, 2012).

5. Artifacts, Mode Dependence, and Mitigation Strategies

Measurement artifacts are substantial in STEDS, primarily due to feedback-induced modulation of tip–sample distance and set-point protocols. Major acquisition modes include:

• Grid Mode: Feedback engaged at a single stabilization bias, followed by open-loop spectroscopy. Minimizes dispersing artifacts—preferred for accurate dispersion extraction (Macdonald et al., 2016).
• Constant-Current Map Mode: Feedback engaged at each energy, leading to dispersing, artificial Fourier features that can mimic real QPI bands. These artifacts are most prominent near $dI/dV(V,x)$0 and may cross $dI/dV(V,x)$1 (Macdonald et al., 2016).
• Constant-Height Mode: Feedback disengaged during entire spectroscopy, best reproducing the intrinsic dispersion features.

Set-point artifacts, alias peaks induced by position-dependent, energy-independent factors, can be identified and excluded via cross-sectional line cuts, energy-symmetrized or ratio mapping ($dI/dV(V,x)$2 or $dI/dV(V,x)$3), and careful protocol comparison (Lee, 2012, Macdonald et al., 2016).

6. Advantages, Scope, and Limitations

STEDS offers unique capabilities:

• Locality: Atomically and spatially resolved access to quantum states.
• Dual-state Sensitivity: Probes unoccupied as well as occupied states, in contrast to ARPES.
• Band Structure Mapping: Quantitative extraction of $dI/dV(V,x)$4, constant-energy contours, Fermi velocities, effective masses, and gap energies with high precision (sub-0.1 eV for GNRs (Söde et al., 2014)).
• Orbital Selectivity: Decomposition of QPI and $dI/dV(V,x)$5 into orbital-resolved channels, enabling chemical/structural identification and tip dependence studies (Abilio et al., 15 Apr 2025).
• Versatility: Applicable to electronic bands, collective modes, and magnon/triplon dispersions.

Principal limitations derive from finite-size effects (discrete $dI/dV(V,x)$6-points), window selection, tip–sample distance, orbital or matrix element selectivity, and sensitivity to acquisition protocol. JDOS and stationary phase approximations fail in systems dominated by strongly phase-mixed or fractionalized excitations, as in spinon continua (Henriques et al., 19 Feb 2025). The spatial resolution is counterbalanced by the need for careful statistical treatment and noise/artifact rejection.

7. Comparative Table: STEDS Modalities and Target Systems

Material System	Modalities/Features	STEDS Output/Limitations
2D Metals, Graphene	dI/dV maps, QPI analysis	$dI/dV(V,x)$7, CEC, Fermi surface, pseudospin, JDOS vs. T-matrix comparison (Simon et al., 2011)
1D GNRs, Chains	Line-scan dI/dV FT	Band structure, effective mass, gap (precision: 0.06 eV), DFT comparison (Söde et al., 2014)
Spin Chains	d²I/dV²(x,ω) FT (IETS)	Magon/triplon dispersion if standing waves form, failure for spinon continua (Henriques et al., 19 Feb 2025)
Strongly correlated SC	2D grid dI/dV, cross-section FTSTS	Extraction of octet and non-octet dispersions, phase transition diagnostics, full QPI comparison (Lee, 2012)
2DES with Rashba	dI/dV(x,y;E), FFT, LL/QPI	Non-parabolic $dI/dV(V,x)$8, Rashba parameter, Landau quantization, percolation (Becker et al., 2010)

• “Fourier Transform Scanning Tunneling spectroscopy: the possibility to obtain constant energy maps and the band dispersion using a local measurement” (Simon et al., 2011)
• “On determining the energy dispersion of spin excitations with scanning tunneling spectroscopy” (Henriques et al., 19 Feb 2025)
• “Electronic Band Dispersion of Graphene Nanoribbons via Fourier-Transformed Scanning Tunneling Spectroscopy” (Söde et al., 2014)
• “Scanning tunneling spectroscopy of a dilute two-dimensional electron system exhibiting Rashba spin splitting” (Becker et al., 2010)
• “A high-resolution cross-sectional analysis for Fourier-transform scanning tunneling spectroscopy and fully-phased Green-function-based quasiparticle scattering theories” (Lee, 2012)
• “Dispersing artifacts in FT-STS: a comparison of set point effects across acquisition modes” (Macdonald et al., 2016)
• “Energy-resolved tip-orbital fingerprint in scanning tunneling spectroscopy based on the revised Chen's derivative rule” (Abilio et al., 15 Apr 2025)