Tunneling probe-based identification of the sp${}^3$ dangling bond on the H-C(100) -$2times1$ surface
Paper details
Section titled “Paper details”| Field | Value |
|---|---|
| Journal | Open MIND |
| Published | 2026-01-29 |
| Authors | Lachlan M. Oberg, Yi‐Ying Sung, Cedric Weber, Marcus W. Doherty, Christopher I. Pakes |
| DOI | 10.48550/arxiv.2601.21154 |
Scanning tunneling spectroscopy identified the sp³ dangling bond on hydrogen-terminated C(100) through two spectral features. Their tunneling onsets were -1.9 V and +3.0 V. The positive-bias onset remained nearly fixed as band bending changed, making it a stable reference for defect identification.
A two-feature identification signature
Section titled “A two-feature identification signature”High-resolution STM first located one sp³ dangling bond as a bright, asymmetric, chestnut-shaped feature among reconstructed dimer rows. The authors selected an isolated site because neighboring defects and step edges could affect its spectrum or cause unwanted tunneling. A targeted voltage sweep after spectroscopy re-passivated the site with hydrogen, confirming the defect assignment.
The defect forms when one terminating H atom leaves a carbon site on the dimerized H-C(100):2 × 1 surface. The bare carbon retains an unsatisfied electron in an sp³ bond and supports localized states within the diamond band gap. Those states can trap charge and produce magnetic and electric noise, reducing near-surface NV coherence and screening electric fields.
Hydrogen capping can neutralize these effects, but the procedure has poor control, yield, and spatial resolution on the order of nanometres.
Surface preparation and tunneling spectra
Section titled “Surface preparation and tunneling spectra”CVD-grown C(100) pieces measuring 3.5 × 3.5 mm² were outgassed at 823 K for 1 h. This treatment removed airborne surface impurities and produced a small number of dangling bonds.
STM and STS used a SPECS Aarhus ultra-high-vacuum instrument at room temperature and a base pressure of 10^-9 mbar. Images were recorded in constant-current mode at +2.5 V sample bias and a 0.3 nA setpoint to image empty states.
Single-point dI/dV spectra placed the tip on the dangling bond or a dimer row. A lock-in amplifier applied 583 Hz modulation with 50 mV peak-to-peak amplitude. Bias was ramped from +4 V or +2.0 V to -2.5 V while tip-sample distance remained fixed. The fixed distance corresponded to I = 0.30 nA and V = 4 V.
Subtracting the dimer-row spectrum isolated features associated with the dangling bond. The measured peak centers, tunneling onsets, and widths were:
| Feature | Peak center | Tunneling onset | FWHM |
|---|---|---|---|
| Negative-bias peak | -2.1 V | -1.9 V | approximately 0.5 V |
| Positive-bias peak | +3.4 V | +3.0 V | approximately 1 V |
Onset is the half-maximum voltage of the difference between the dangling-bond and dimer-row dI/dV curves. Neither peak appeared in the dimer-row spectrum, which supported its assignment to defect states.
Calculated defect states
Section titled “Calculated defect states”Earlier first-principles work had determined the defect electronic structure without aligning it to an experimental STS spectrum. VASP calculations in this study represented the surface with a 419-ion slab containing eight carbon layers and dimensions of 17.7 Å × 15.2 Å.
PBE geometry optimization used a 5 × 10^-3 eV/Å force tolerance, a 600 eV cutoff, and a 4 × 4 × 1 Γ-centered mesh. Total energy was optimized to 10^-4 eV. HSE06 Γ-point calculations then included self-consistent electric fields between ±0.2 V/Å.
Charge-density analysis identified three occupied defect states at -0.7 eV, -0.55 eV, and -0.2 eV relative to E_f. One unoccupied state lay at +2.9 eV relative to E_f. For the modeled boron-doped diamond, E_f sat about 0.2 eV above the bulk valence-band maximum.
The authors numbered the occupied states 1 through 3 in ascending energy order. The unoccupied state’s charge density extended farther into the vacuum, so it underwent the largest Stark shift. The calculation supplied candidate energies, but it did not predict the observed onset voltages directly.
Band bending controls the voltage scale
Section titled “Band bending controls the voltage scale”STS peak positions cannot be read directly as defect energies because diamond screens the tip electric field poorly. Without intrinsic surface carrier accumulation, the applied bias does not drop entirely across the vacuum region. The tip Fermi level translates by qV_d rather than qV_a, with V_d = V_a - V_bb. V_bb is the surface band-bending potential, and V_d is the voltage drop from tip to surface.
Holes accumulate under positive sample bias and screen the field strongly, so V_a and V_d are close. At negative bias, screening comes from ionized boron acceptors with an activation energy of approximately 0.4 eV. Boron ionization remains incomplete, so large negative biases produce poor screening and a smaller V_d.
The calculated +2.9 eV unoccupied state was numerically close to the +3.0 V onset. Occupied orbital 3 at -0.2 eV was far from the -1.9 V onset. Stark shifts and band bending prevent direct energy equivalence.
COMSOL finite-element solutions of Poisson’s equation varied acceptor concentration N_A and absolute tip height t_H. They also used V_BI, measured from I(t_H), as a boundary condition. Built-in voltage was -0.9 eV, derived from a 3.1 eV tip work function and an approximately 4.0 eV sample work function. Current-distance measurements at 1.5 V gave κ = 0.9 Å^-1 for an Ag/W tip.
The estimated acceptor concentration was constrained to 10^16 cm^-3 ≤ N_A ≤ 10^20 cm^-3. Tip height was expected within 5 to 10 Å, but its absolute value was unknown. Lower N_A reduced ionized-acceptor screening, while higher t_H increased screening through a larger vacuum voltage drop. This polarity dependence explains why the positive onset changed only weakly with N_A and t_H.
Peak assignment and parameter extraction
Section titled “Peak assignment and parameter extraction”The authors assigned the positive-bias feature to the unoccupied defect orbital. Its onset remained near 3 V across the modeled N_A, t_H, and V_BI changes. The negative-bias feature was assigned to occupied orbital 3. Orbitals 1 and 2 were considered below the measured range for V_BI = -0.9 eV.
Stark shifts and electrostatic band bending were fitted together to the measured onsets. Using -1.9 V and +3.0 V gave t_H = 6.2 Å and N_A = 1.3 × 10^19 cm^-3. The fitted N_A falls within the estimated sample range, and t_H falls within the expected tunneling range.
The fit is not unique because many combinations of t_H and N_A reproduce the same two onset voltages. A ±0.05 V change in the positive peak allowed t_H solutions between 5 Å and 9 Å. That spread follows from asymmetric band bending at positive and negative bias.
The positive onset is therefore a practical reference, while the negative feature carries material and geometry information. Peak position alone does not give a reliable orbital energy for this wide-bandgap semiconductor.
Identification limits and device implications
Section titled “Identification limits and device implications”The positive onset cannot identify the defect by itself because many other diamond defects have unoccupied mid-gap states near 3 eV. Identification therefore requires the positive onset, the negative-bias feature, and STM topography.
The reported spectrum came from one isolated dangling bond, so the signature was not tested across many defect types. STM morphology can also resemble benign surface defects, which complicates selective hydrogen capping. STS can check the local electronic structure before and after a manipulation. The paper does not show that STS makes voltage-pulse capping deterministic or improves its fabrication yield.
For hydrogen desorption lithography, the proposed route removes H from the surface with STM-tip voltage pulses, then adsorbs an N-based gas at reactive sites. CVD overgrowth would then produce bulk NV centers, and coupled NV centers are a proposed scaled device outcome. The complete process was not demonstrated.
STS could validate many such sites before adsorption and overgrowth in a scaled process. The demonstrated result is spectroscopic identification and model-based parameter extraction, not deterministic lithography at device scale.
Materials and support
Section titled “Materials and support”For probe-based identification of sp³ dangling bonds, boron-doped diamond is the material family for surface-state spectroscopy in diamond quantum devices. The process needs a hydrogen-terminated C(100) surface, and 6CCVD’s diamond materials catalog lists boron-doped diamond alongside MPCVD-grown single-crystal and polycrystalline diamond. The linked engineering support page can be used to define the substrate and surface preparation needed for this work.
View Original Abstract
The sp${}^3$ dangling bond on the diamond surface plays a critical role in the performance and fabrication of diamond quantum technologies. For the former, the magnetic and electric properties of this defect can impede the performance of quantum sensors and computers. For the latter, the chemical properties of the dangling bond are integral to proposed methods for bottom-up fabrication of scalable diamond quantum devices. In pursuit of high performance and scalable diamond quantum technology, tunneling probe-based techniques offers the ability to create and modify the sp${}^3$ dangling bond with atomic-scale precision. However, these capabilities cannot be realised either deterministically or at scale without a means for identifying the sp${}^3$ dangling bond amidst the myriad of other defects on the diamond surface. Consequently, in this work we provide a comprehensive experimental and theoretical framework for STS-based characterisation of the sp${}^3$ defect on the H-terminated (100) diamond surface. This capability provides the foundation for future tunneling probe studies in the modification of dangling bonds.
Electronic Grade Intrinsic Diamond Epitaxial Wafer H Terminated Superior 100 Mosfet Ready Substrate
View Specs →
Boron Doped Single Crystal Diamond Plate
View Specs →