Skip to content

Process–Structure–Property Relationships in Boron-Doped CVD Diamond Films on Si3N4 for Biosensor Applications

FieldValue
JournalMaterials
Published2026-07-14
AuthorsSusana Ferreira, André Costa Vieira, Miguel Neto
DOI10.3390/ma19143027

Direct hot-filament CVD growth of boron-doped diamond on Si3N4 produced one film that met the authors’ preferred electrical resistance range of 1 to 10 kΩ. Sample S7 combined a 150 mbar pressure, 2 sccm CH4 flow, a 214 nm mean crystallite size, and hydrophilic C-O-C and OH surface chemistry. The authors selected this condition as the most balanced for future implantable glucose biosensors, although they did not test enzyme immobilization, glucose detection, cytotoxicity, or hemocompatibility.

Films were grown for 3 h in a home-built hot-filament CVD reactor. Each sintered Si3N4 disc measured 10 mm in diameter and 1 mm in thickness. Polishing used 15 µm, 6 µm diamond, and 0.05 µm colloidal silica. Seeding used 0.5 to 1 µm diamond in an ethanol suspension at 0.1 g/mL.

Six tungsten filaments, each 250 µm thick and 8 cm long, sat 6 mm above the substrates. Spacing between filaments was 8 mm. The filaments ran at 2200 to 2300 °C with approximately 1000 W total power. Hydrogen flow stayed at 128 sccm, while CH4 flow was set at 2, 4, or 6 sccm.

The boron source was B2O3 dissolved in ethanol at a B/C ratio of 5000 ppm. Substrates rotated at 2 rpm, and the authors produced three samples for each condition. S1 to S3 used a Cu/Si3N4/Mo support and 8 sccm argon. S4 to S9 used a Cu/Al2O3/Si3N4/Mo support and 6 sccm argon.

Surface resistance was measured by the Van der Pauw method. Four copper wires were bonded to each film with silver glue. SEM, XRD with LIBAD, Raman spectroscopy, AFM, XPS, FTIR, and sessile-drop contact angles characterized the coatings.

Microstructure and crystallographic response

Section titled “Microstructure and crystallographic response”

XRD identified a diamond (111) reflection at 2θ = 43.91° in S1 through S7. SEM and LIBAD separated faceted microcrystalline diamond from finer nanocrystalline diamond. S1, S4, and S7 had the largest crystallites and strongest diamond Raman phases. Raman analysis also showed more non-diamond sp2 carbon as methane flow increased. The authors linked this change to lower atomic hydrogen at higher CH4/H2 ratios.

Selected structural and roughness measurements show the trade between crystallite size, methane flow, and wettability.

SamplePressure and CH4 flowMean LIBAD crystalliteAFM RqWetting result
S125 mbar, 2 sccm286 nm63.6 nmHydrophobic
S425 mbar, 2 sccm214 nm90.9 nmHydrophilic
S525 mbar, 4 sccm39 nm13.6 nm81°
S7150 mbar, 2 sccm214 nm71.4 nmHydrophilic
S8150 mbar, 4 sccm32 nm15.7 nm83°
S9150 mbar, 6 sccmMissingMissing87°, incomplete film

AFM roughness was measured over a representative 50 µm by 50 µm area. Increasing pressure while also raising methane produced irregular aggregates in S8 and S9. S9 had an incomplete film, so its LIBAD crystallite and AFM roughness values were missing. The authors inferred from S1 versus S3 and S4 versus S6 that lower methane favors (111) orientation. They cautioned that their pressure comparisons did not support fully substantiated conclusions.

XPS found carbon dominant in all films, with oxygen second and silicon minor. S7 contained 92.99% carbon and 6.93% oxygen. S8 had the highest measured oxygen content at 9.55%. Minor silicon and occasional nitrogen, sodium, sulfur, or tungsten appeared in the surface analysis. The authors attributed oxygen to reactive species formed from ethanol and B2O3 near the hot tungsten filaments.

FTIR showed pronounced oxygen-containing bands in S2, S4, and S7. Deconvolution identified C-O as dominant in S2 and S4. S7 showed stronger C-O-C contributions together with OH terminations. All samples had mixed hydrogenated and oxygen-containing terminations rather than a uniform hydrogen-terminated surface.

Contact angles used 2 µL distilled water at 25 ± 1 °C. Each result averaged five measurements taken 30 s after droplet contact. S2, S4, S6, and S7 were hydrophilic, while S1 and S3 were hydrophobic. S5, S8, and S9 measured 81 degrees, 83 degrees, and 87 degrees, respectively. The authors linked oxygen-containing groups to the hydrophilic behavior and considered them favorable for enzyme immobilization. No enzyme immobilization was measured, so that application remained unverified.

The authors set a resistance below 10 kΩ as the maximum for biosensor use and preferred a range of 1 to 10 kΩ. Van der Pauw measurements placed S1, S4, and S7 at the lowest resistance among the nine films. S7 alone met the preferred range. S4 fell outside that range but remained close to the target.

Stronger (111) orientation was associated with more effective boron incorporation and lower resistance. The authors also noted that manual application of silver paste could cause minor contact-quality variation, although they considered its effect on the measured trends minimal.

The discussion cites earlier work in which microcrystalline diamond formed at a CH4/H2 ratio of 0.016 and a substrate temperature of 850 °C. Nanocrystalline diamond formed at ratios of 0.032 to 0.064 and a substrate temperature of 750 °C. The low-methane S1, S4, and S7 films in the present study followed the earlier microcrystalline trend.

That earlier work reported hydrophobic, partly oxygenated surfaces with contact angles near 90 degrees. It also linked (111) orientation to boron incorporation and reduced dopant segregation at grain boundaries. Within the current matrix, the same 2 sccm methane flow produced hydrophobic S1 but hydrophilic S4 and S7. Methane flow alone therefore did not determine wetting.

The authors identify several limits on the results. The study did not measure cytotoxicity, hemocompatibility, or long-term stability in physiological environments. It also did not demonstrate enzyme immobilization efficiency, electrochemical glucose sensitivity, or detection in real media.

Changes in holder configuration and gas-phase composition prevented clean assignment of every result to a single parameter. Trace tungsten from filament evaporation was detected, and its biological effect remains unknown. The work used a laboratory-scale HFCVD system, so scalability, industrial process reproducibility, and economic feasibility were not tested.

The missing crystallite and roughness data for incomplete S9 also prevent a full comparison with the other films. S7 is therefore a material candidate rather than a validated implantable biosensor.

Implications for biosensor material selection

Section titled “Implications for biosensor material selection”

S7 provides the strongest starting point because it combines (111)-oriented microcrystalline diamond, 214 nm crystallites, low resistance, and hydrophilic oxygen chemistry. Surface area and electrical response must be considered together. The authors note that smaller boron-doped diamond crystallites offer more area for enzyme immobilization. Very small crystallites may also increase enzyme-enzyme interactions and blocking.

That consideration supports microcrystalline diamond for glucose biosensors when larger crystallites give better boron incorporation and conductivity. Process development should preserve the S7 balance while checking film continuity and resistance across the coated area. Device work must then add biological and glucose-response measurements before the coating is selected for implantation.

The material requirement is boron-doped CVD diamond on a Si3N4 carrier for implantable electrochemical glucose biosensor electrodes. 6CCVD’s diamond materials catalog lists boron-doped diamond and metallization with Au, Pt, Pd, Ti, W, or Cu, which match the coating and contact needs described here. The engineering support link is the contact point for custom fabrication questions around the Si3N4 electrode platform. The paper used manually applied silver paste for its Van der Pauw contacts, so any production contact metal requires separate qualification.

View Original Abstract

This study explores the direct growth of boron-doped diamond films on biocompatible silicon nitride (Si3N4) ceramic substrates using hot-filament chemical vapor deposition (HFCVD), with a view toward their use in implantable electrochemical biosensors. The focus of this work is the establishment of process-structure-property relationships relevant to biosensor performance, including microstructure, surface chemistry, wettability, and electrical behaviour. The effects of key deposition parameters, namely methane concentration, deposition pressure, and sample holder configuration, were analysed in relation to film microstructure, crystallographic orientation, surface chemistry, wettability, and electrical performance. Under low CH4/H2 ratios, microcrystalline diamond films with a pronounced (111) preferential orientation were obtained, enabling improved boron incorporation and electrical resistivity values within the range required for biosensor operation (≈1-10 kΩ). Surface analyses revealed partially hydrogen-terminated diamond layers enriched with oxygen-containing functional groups (C-O and C-O-C), which enhance surface wettability and are suitable for enzyme immobilization. Among the studied conditions, films deposited at 150 mbar and low methane flow displayed the most balanced combination of electrical conductivity, surface wettability, and microstructural stability. Overall, the results highlight the potential of boron-doped CVD diamond grown directly on Si3N4 as a robust and biocompatible material platform for future implantable biosensors, particularly for glucose monitoring applications.

  1. 2019 - The global epidemics of diabetes in the 21st century: Current situation and perspectives [Crossref]
  2. 2025 - Composites-based electrodes in enzymatic electrochemical glucose biosensors [Crossref]
  3. 2022 - IDF Diabetes Atlas: Global, regional and country-level diabetes prevalence estimates for 2021 and projections for 2045 [Crossref]
  4. 2023 - Blood glucose self monitoring [Crossref]
  5. 2025 - Wearable biosensor with high specific surface area PGA-CNTs electrode for sweat glucose detection [Crossref]
  6. 1962 - Electrode systems for continuous monitoring in cardiovascular surgery [Crossref]
  7. 1967 - Enzyme electrode [Crossref]
  8. 2015 - Roles of interstitial fluid pH in diabetes mellitus: Glycolysis and mitochondrial function [Crossref]
  9. 1945 - Chemistry of palmar sweat preliminary report: Apparatus and technics [Crossref]