Graphene (0002)/Diamond (111) Heterojunction with High Piezoresistive Response
Paper details
Section titled “Paper details”| Field | Value |
|---|---|
| Journal | Advanced Science |
| Published | 2026-07-06 |
| Authors | Xueyu Zhang, Kun Guo, Zhigang Gai, Tianxiao Guo, Yuan Gao |
| DOI | 10.1002/advs.76096 |
A covalently bonded graphene (0002)/diamond (111) heterojunction fabricated by thermal electron irradiation produced a gauge factor of −1149 under the reported loading configuration. The type-P structure covered the (111) face of a 3 × 3 × 0.3 mm single-crystal diamond chip. Its AFM roughness parameter Sa was 0.72 nm, and its tested pressure range was 0 to 112 MPa.
A parallel graphene layer on diamond
Section titled “A parallel graphene layer on diamond”The type-P heterojunction places graphene (0002) parallel to diamond (111). Type-V structures instead have graphene planes perpendicular to diamond. Diamond and graphene bond lengths differ by less than 2%, with 1.45 Å for diamond and 1.42 Å for graphene. Raman spectra from both configurations showed the diamond sp³ peak at 1332 cm⁻¹ and graphene G and 2D peaks at 1580 and 2700 cm⁻¹. The type-P spectrum had an I₂D/I_G ratio of about 0.7, assigned to 4 to 6 layers. The type-V ratio was about 1, assigned to 1 to 2 layers.
Type-V had a thicker surface graphene region, but its layer number decreased upward because the sheets were parallel to the incident laser. SEM showed flakes standing perpendicular to the diamond for type-V. Type-P had a flat 2D surface under SEM. AFM measured Sa values of 21.16 nm for type-V and 0.72 nm for type-P. Bare single-crystal diamond measured 0.65 nm. Optical microscopy found only slight transparency loss in the type-P sample.
Hot-electron irradiation creates a subsurface active layer
Section titled “Hot-electron irradiation creates a subsurface active layer”The authors used a home-built hot-electron system rather than a focused keV electron beam. Molybdenum filaments thermionically emitted hot electrons over the processing area. A positive substrate bias directed the electrons toward diamond in low-pressure hydrogen, while radiation from the filaments heated the sample.
The substrate was HPHT-synthesized (111)-oriented single-crystal diamond with a 3° misorientation. The type-P recipe used 6 kW hot-wire power, 850 to 860 °C surface temperature, and 8 to 10 Pa chamber pressure. The substrate-to-filament distance was 2 cm, the substrate bias was 200 V, and heating lasted 30 min. Before irradiation, samples were ultrasonically cleaned in anhydrous ethanol for 10 min. They were then cleaned in deionized water for 5 min and dried with high-purity nitrogen.
The controlled field activated the nanoscale subsurface while limiting cleavage of outer carbon bonds. The authors attribute this balance to a 2D active layer below the surface. Exceeding the critical electron energy cleaved surface bonds, which propagated inward and favored type-V growth. The type-P processing window was therefore narrow. High-quality substrates were required because grain boundaries and defects acted as preferential graphitization sites. They knew of no reports of covalently bonded type-P G/D heterojunctions.
Bonding and the proposed diamond-to-graphene pathway
Section titled “Bonding and the proposed diamond-to-graphene pathway”Cross-sectional TEM, prepared by focused ion beam milling, resolved cubic diamond, an interface, and graphene. HRTEM measured a 2.06 Å spacing with ABC stacking in cubic diamond. The interfacial region matched hexagonal diamond (100) planes with 2.18 Å spacing and A’B’A’ stacking. Graphene above the interface had ABA stacking and 3.5 Å interplanar spacing.
EELS measured a 292 eV σ* peak and a 302 eV dip in the diamond regions. The 285 eV π* peak became prominent in the graphene regions. Mixed bonding signatures crossed the interface, supporting covalent sp² and sp³ coupling. The type-V interface was more abrupt, with its bonding transition confined to about 1 to 2 atomic layers.
DFT calculations placed the relative atomic energy of cubic diamond above that of a CD and hexagonal diamond composite. The G/D composite had the lowest energy. The authors propose hexagonal diamond as an intermediate. In their route, continued bombardment breaks σ bonds perpendicular to the hexagonal diamond (100) plane. Partial sp³ to sp² conversion then forms the ABA-stacked graphene layers.
Piezoresistive response under defined loading
Section titled “Piezoresistive response under defined loading”Four indium electrodes sat at the corners of the graphene surface. Pressure was applied to the diamond side, while probe contacts reached the graphene from below. Stepwise loading ran from 0 to 28 N in 2 N intervals. The effective indenter area was 0.25 mm², giving a maximum pressure of 112 MPa.
Finite element analysis supplied the strain because the 0.3 mm single-crystal diamond chip dominated the elastic deformation. The calculation used a Young’s modulus of 1050 GPa. Resistance was measured along two diagonal electrode pairs. Their relative resistance change and simulated strain gave a gauge factor of −1149.
Cyclic loading at 28 N produced response times of 33 ms during pressurization and 56 ms during depressurization. The authors reported good short-term stability and reproducibility over the cyclic tests.
Comparison with other piezoresistive systems
Section titled “Comparison with other piezoresistive systems”Gauge factors are not interchangeable across devices because strain range, loading geometry, electrodes, and strain calculation differ. The paper’s comparisons place the heterojunction among rigid semiconductor pressure sensors rather than flexible graphene composites.
| System | Gauge factor | Reported context |
|---|---|---|
| Type-P G/D in this work | −1149 | Rigid chip under small elastic strain |
| Metal strain gauge | about 2 | Geometric deformation |
| Single-crystal silicon | Single digits to 200 | Crystal orientation and doping dependent |
| SiC | about 30 | Piezoresistive material |
| Polycrystalline BDD | about 283 | Film response; consistent with the authors’ prior work |
| Intra-grain BDD | 4000 | Individual grains |
| Graphene composites | 10⁴ to 10⁶ | Often large tensile or bending strain |
| ZnO nanowires | 1250 | Piezoelectric effect |
The G/D magnitude exceeds the cited silicon, SiC, and polycrystalline BDD values. It is close to the ZnO nanowire result. Flexible graphene systems can report larger apparent values because wrinkles, cracks, contact resistance, and conductive networks change during loading. The cited examples often had measurement ranges below 60 kPa. The G/D result therefore combines a large negative response with a wider test range, although it remains specific to its present configuration.
Electronic origin of the negative response
Section titled “Electronic origin of the negative response”The authors modeled the interface under 2% biaxial tensile strain. DFT relaxation elongated carbon bonds in both the graphene and diamond layers. It also shifted the density of states near the Fermi level. A charge-density difference map showed electron depletion on the graphene side and accumulation at the diamond surface.
The calculations indicate electron transfer from graphene into the diamond subsurface. The local density of states for a diamond-side carbon atom, labeled D1, rose by about a factor of three under strain. The distribution also shifted toward higher energy. The authors associate the negative gauge factor with an electron-rich diamond surface and a strain-driven rise in interfacial density of states. Monocomponent graphene and boron-doped diamond normally have positive gauge factors, so the measured sign belongs to the coupled interface.
The 1050 GPa diamond substrate keeps elastic strain small during loading. The authors argue that this limits geometric contributions and carrier-mobility degradation. TEM, EELS, DFT, and the measured resistance change then form a consistent evidence chain. The calculations support this mechanism, but they do not directly image carrier redistribution during the macroscopic loading test.
Limits for pressure-sensor engineering
Section titled “Limits for pressure-sensor engineering”The reported device uses one 3 × 3 × 0.3 mm chip and one defined loading geometry. Wafer-scale processing and systematic device-to-device statistics remain unresolved. The narrow irradiation window also makes substrate quality and surface temperature important process controls. For material selection, the paper supports high-crystallinity single-crystal diamond because grain boundaries and defects become preferential graphitization sites. Device design must retain a defined loading geometry because the gauge factor depends on force, contact area, electrode arrangement, and strain calculation.
Temperature drift was measured from 25 to 120 °C in 5 °C steps. Resistance decreased approximately linearly, with a total change of about 284 Ω. Practical pressure sensing will need temperature compensation. Humidity effects on exposed graphene were not measured. The authors expect packaging or sealing to reduce humidity influence, but they call for tests on unsealed and packaged devices.
Radiation tolerance and operation above the measured temperature range were not demonstrated. Claims about high-temperature, radiation-tolerant, wide-range, and fast-response sensors remain conditional on device optimization and validation. Interfacial thermal transport is also a proposal rather than a measured result. The authors call for direct measurements of interfacial thermal resistance and device-level thermal management.
Materials and support
Section titled “Materials and support”The piezoresistive pressure-sensing chip needs a high-quality (111)-oriented single-crystal diamond substrate with a controlled hot-electron-irradiated surface. 6CCVD’s diamond materials catalog lists MPCVD-grown single-crystal and polycrystalline diamond, plates and wafers up to 125 mm, and thickness from 0.1 to 500 micrometers for both material types. The paper’s 0.3 mm chip thickness falls within that range, but its HPHT growth and (111) orientation require review with engineering support before process compatibility is assumed.
View Original Abstract
Diamond and graphene are emerging as promising successors to silicon-based materials for semiconductor applications, particularly when integrated into graphene/diamond heterostructures. However, fabricating high-quality graphene (0002) atomic layers on diamond (111) with intimate contact-mostly desired for device integration-remains a great challenge. Here, we report a strategy to realize a covalently bonded, graphene (0002)/diamond (111) heterojunction using thermal electron irradiation. Combining structural characterization with theoretical calculations, the formation mechanism and energy band characteristics of this junction structure were clarified, and the microscopic mechanism of hexagonal diamond (lonsdaleite) acting as an intermediate state before its transformation into graphene layers was proposed. The heterostructure exhibits a high piezoresistive response under the present testing configuration, with a gauge factor of -1149. This performance is attributed to two key factors: first, the heterojunction-induced formation of an electron-rich layer on the diamond surface; and second, a significant stress-induced increase in the density of states of carbon C 2p orbitals within the diamond layer. Leveraging diamond’s intrinsic properties, piezoresistive chips based on this structure may offer opportunities for high-temperature, radiation-tolerant, wide-range, and fast-response sensing applications after further device-level optimization and validation. These results provide a useful basis for developing diamond-based semiconductor and sensing devices.
Boron Doped Diamond On Titanium Bdd Electrode
View Specs →
Boron Doped Diamond Electrode On Niobium Bdd Nb Electrode
View Specs →Boron Doped Diamond Electrode On Silicon
View Specs →References we cover
Section titled “References we cover”- Diamond (111) surface reconstruction and epitaxial graphene interface
- Laser-Induced Graphitization of Diamond Under 30 fs Laser Pulse Irradiation
Original source
Section titled “Original source”References
Section titled “References”- 2023 - van der Waals‐Interface‐Dominated All‐2D Electronics [Crossref]
- 2023 - Materials Innovations for Quantum Technology Acceleration: A Perspective [Crossref]