Skip to content

Accurate determination of the valence band edge in hard x-ray photoemission spectra using GW theory

MetadataDetails
Publication Date2016-04-27
JournalJournal of Applied Physics
AuthorsJohannes Lischner, SlavomĂ­r NemĆĄĂĄk, Giuseppina Conti, Andrei Gloskovskii, Gunnar Karl PĂĄlsson
InstitutionsDeutsches Elektronen-Synchrotron DESY, Thomas Young Centre
Citations5
AnalysisFull AI Review Included

Technical Documentation and Analysis: Accurate Valence Band Determination in Boron-Doped Diamond

Section titled “Technical Documentation and Analysis: Accurate Valence Band Determination in Boron-Doped Diamond”

This paper introduces a robust, high-accuracy methodology for determining the Valence Band Maximum (VBM) in semiconductors, validated specifically on synthetic Hydrogen-Terminated Boron-Doped Diamond (HBD), a critical material for solar cell substrates.

  • Methodological Advance: A new technique combining Hard X-ray Photoemission Spectroscopy (HXPS) data with high-level ab initio GW theory calculations is employed to extract VBM values.
  • Mitigation of Matrix Effects: The method successfully addresses the primary challenge in high-energy HXPS—strong matrix element effects that render standard VBM extraction methods (Kraut/Chambers) inaccurate, especially for diamond.
  • VBM Determination: The sharpest peak in the experimental Valence Band Density of States (VB DOS) is aligned directly with the corresponding feature from the GW theoretical calculation, requiring only a single fitting parameter.
  • Material Validation: The methodology was applied to high-purity, synthetic HBD doped at ≈1020 cm-3 (570 ppm), confirming its utility for wide-bandgap semiconductor analysis.
  • Key Achievement: The energy separation between the C 1s core level and the VBM (ECL - EVBM) was determined to be extremely consistent (283.74 eV @ 5.0 keV and 283.71 eV @ 2.5 keV), validating the robustness against varying photoemission escape depths.
  • Application Relevance: Accurate VBM and core-level separation data are essential prerequisites for calculating heterojunction band discontinuities and Schottky barrier heights in diamond-based devices.

The table below summarizes the critical experimental parameters and quantified results from the HXPS study on hydrogen-terminated boron-doped diamond.

ParameterValueUnitContext
Substrate MaterialHBD (Synthetic Diamond)N/ASubstrate for solar cell devices.
Boron Concentration≈1020 (570 ppm)cm-3Doping level used for HXPS analysis.
Primary Photon Energy5.0keVUsed for primary core-level analysis.
Validation Photon Energy2.5, 5.9keVUsed to verify stability and matrix element effects.
Core Level - VBM Separation (ECL - EVBM)283.74eVMeasured at 5.0 keV photon energy.
Core Level - VBM Separation (ECL - EVBM)283.71eVMeasured at 2.5 keV photon energy (Robust result, ±0.03 eV variation).
VBM Relative to Fermi Level0.45eVMeasured at 5.0 keV photon energy.
Occupied Valence Bandwidth (GW Theory)23.06eVExcellent agreement with experimental data (23.0 ± 0.2 eV).
Sharpest VB Peak Binding Energy12.56eVThe reference point used for aligning experimental and GW theoretical spectra.
Total Energy Resolution (High-Res)0.2eVUsed during 5.9 keV variable polarization measurements.
Penetration Depth (HXPS Regime)> 100ÅDeep subsurface probing achieved with multi-keV photons.

The following steps outline the crucial experimental and theoretical procedures utilized to achieve the accurate VBM determination in HBD:

  1. Material Preparation: High-quality, synthetic Boron-Doped Diamond (BDD) samples were prepared and hydrogen-terminated (HBD).
  2. Spectroscopy Choice: Hard X-ray Photoemission Spectroscopy (HXPS) was selected, utilizing photon energies in the multi-keV range (2.5, 5.0, 5.9 keV) to enable deep probing of bulk and interface properties, circumventing surface sensitivity issues.
  3. Data Acquisition Geometry: All experiments were conducted in grazing incidence/normal emission geometries to maximize photoelectron yield.
  4. Variable Polarization: Measurements were taken using both p-polarized and s-polarized light (via a diamond phase retarder) to systematically modulate matrix element effects and selectively enhance/suppress contributions from C 2s and C 2p orbitals.
  5. Band Structure Calculation (DFT): Initial Valence Band Density of States (VB DOS) was calculated using Density Functional Theory (DFT) (PBE approximation) as a baseline comparison.
  6. Quasiparticle Correction (GW Theory): Accurate quasiparticle energies and VB DOS were computed using the ab initio GW method (accounting for electron self-energy and many-electron effects) to correctly predict the true occupied valence bandwidth.
  7. VBM Alignment (New Method): The experimental VBM was determined by aligning the energy of the sharpest peak in the experimental VB DOS (observed at 12.56 eV binding energy) with the corresponding theoretical peak from the GW calculation. This single-parameter alignment is robust against high noise and matrix element variations near the band edge.

The findings presented underscore the need for extremely precise, high-quality Boron-Doped Diamond (BDD) wafers for advanced semiconductor research and device engineering. 6CCVD specializes in delivering MPCVD diamond materials tailored to meet the exacting requirements of synchrotron and laboratory spectroscopy, supporting the replication and expansion of this critical research.

Requirement in Paper6CCVD SolutionEngineering Specification & Relevance
Boron-Doped Diamond (BDD)Heavy Boron Doped Diamond (BDD)We offer custom doping control, capable of matching or exceeding the 1020 cm-3 concentration utilized here, ideal for studying metallic conductivity and band bending.
High Quality SubstrateSCD or PCD DiamondMaterials available for both fundamental physics (high-purity Single Crystal Diamond, SCD) and large-scale application studies (Polycrystalline Diamond up to 125mm, PCD).
Controlled ThicknessCustom Thin Films & SubstratesSCD and PCD available from 0.1 ”m films up to 500 ”m wafers, and substrates up to 10 mm. This control is crucial for HXPS studies that rely on specific photoelectron escape depths.
Critical Surface PreparationUltra-Smooth PolishingGuaranteeing spectroscopic quality surfaces: Ra < 1 nm (SCD) and Ra < 5 nm (inch-size PCD), ensuring optimal conditions for required surface terminations (e.g., H-termination) prior to photoemission.
  • Doping Spectrum Analysis: Researchers seeking to expand the work can order BDD wafers with a defined gradient of boron concentrations, enabling systematic studies on how VBM position, band bending, and core-level offsets evolve as a function of doping.
  • Custom Metalization for Heterojunctions: Calculating heterojunction band alignment (which requires knowledge of VBM and core levels) often leads to studying device interfaces (e.g., Schottky diodes). 6CCVD offers internal capabilities for depositing electrode materials, including Ti/Pt/Au, Pd, W, and Cu, directly onto BDD wafers, streamlining device fabrication for subsequent electrical or spectroscopic testing.
  • Precise Dimensions: To fit specific HXPS or X-ray absorption setups (like those at DESY or ALS), we offer custom laser cutting and shaping of diamond plates up to 125 mm diameter (PCD).
  • Expert Consultation: 6CCVD’s in-house PhD team provides authoritative engineering support, assisting researchers in selecting the optimal diamond type (SCD vs. PCD), thickness, and doping level required to accurately replicate complex band alignment measurements and semiconductor transport physics projects.
  • Global Research Supply Chain: We ensure reliable and secure global shipping (DDU default, DDP available) of sensitive diamond materials directly to synchrotron facilities and academic research centers worldwide.

For custom specifications or material consultation, visit 6ccvd.com or contact our engineering team directly.

View Original Abstract

We introduce a new method for determining accurate values of the valence-band maximum in x-ray photoemission spectra. Specifically, we align the sharpest peak in the valence-band region of the experimental spectrum with the corresponding feature of a theoretical valence-band density of states curve from ab initio GW theory calculations. This method is particularly useful for soft and hard x-ray photoemission studies of materials with a mixture of valence-band characters, where strong matrix element effects can render standard methods for extracting the valence-band maximum unreliable. We apply our method to hydrogen-terminated boron-doped diamond, which is a promising substrate material for novel solar cell devices. By carrying out photoemission experiments with variable light polarizations, we verify the accuracy of our analysis and the general validity of the method.