Skip to content

Boron-doped diamond solution-gate field-effect transistor (BDD-SGFET) biosensor for gene mutation detection

FieldValue
JournalMicrosystems & Nanoengineering
Published2026-03-10
AuthorsZelong Lin, Yun Zheng, Yisen Chen, C. Y. Wang, Zhishan Yuan
DOI10.1038/s41378-026-01184-6
Citations1

A microwire boron-doped diamond solution-gate field-effect transistor detected DNA at a reported limit of 10 pM. It also separated DNA with two, four, and eight base-pair mismatches in a 21-mer EGFR sequence. The device reached a maximum transconductance of 289 μS/mm and a maximum output current of 3.3 mA/mm at VGS = −0.4 V.

The authors modeled the device in Silvaco TCAD with DevEdit for geometry and Atlas for electrochemistry. They varied microwire width at a 100 μm length, then length at a 30 μm width.

Output curves were nearly linear as VDS moved from 0 to −1 V. At VDS = −0.5 V, increasing width from 5 to 20 μm raised simulated transconductance from 242.1 to 444.3 μS.

With a 30 μm width, increasing length reduced transconductance. The authors therefore chose a target wire of 100 μm length and 10 μm width before fabrication.

The final SEM geometry was 100.8 μm long and 7.68 μm wide. The microwire raises sensing area and gate capacitance relative to a planar channel, according to the design rationale.

A 200 nm BDD epitaxial layer was grown on a 10 × 10 mm single-crystal diamond substrate by MPCVD. Growth used 1050 to 1060 °C, 6500 W microwave power, and 90 torr pressure.

The methane/hydrogen ratio was 6%, and a 5000 ppm B₂H₆/CH₄ mixture supplied boron. Hall measurements gave 7.7 × 10¹⁹ cm⁻³ hole concentration and 4.52 × 10³ Ω/sq sheet resistance.

Profilometry gave Ra of 5 nm and Rmax of 6 nm.

Photolithography defined 100 μm-long, 10 μm-wide channels in the BDD. Plasma etching used 400 eV ions, 80 sccm O₂, 5 sccm Ar, and 2 min 30 s.

Electrodes used a 20 nm Cr layer under 200 nm Au and were annealed at 420 °C for 3 min. A 500 nm SiO₂ layer insulated the device while exposing the sensing and electrode regions.

Epoxy resin then encapsulated the assembly.

Electrical performance and electrolyte conditions

Section titled “Electrical performance and electrolyte conditions”

The measured transfer curve gave a maximum transconductance of 289 μS/mm. The threshold voltage was 0.19 V, placing the device in normally-on operation.

Maximum output current reached 3.3 mA/mm at VGS = −0.4 V, with current peaking near VDS = −0.6 V.

At equal gate length, the authors report output current over 40 times that of cited MESFETs and over 60 times that of cited MOSFETs.

The oxygen-terminated surface reached peak transconductance at pH 7.4 in 10 mM PBS. Protonation of oxygen groups at lower pH reduced hole density. Deprotonation at higher pH increased hole attraction and device activation.

At pH 7.4, increasing PBS concentration from 0.1 to 10 mM shifted current upward but flattened the transfer curves. The authors attribute this to a shorter Debye length and stronger electrostatic screening.

DNA surface chemistry and concentration response

Section titled “DNA surface chemistry and concentration response”

Oxygen plasma treated the sensing area at 70 W, minus 100 kPa, for 90 s to raise hydroxyl content. APTES at 2% reacted for 1 h at room temperature, followed by NHS and EDC activation for 1 h.

Activation used a 1:1 mixture of 0.1 M NHS and 0.2 M EDC for 1 h. Probes and targets were diluted in 3× SSC.

A 10 nM probe was immobilized for 4 h. A 10 nM target was hybridized for 1 h at 25 °C. Target DNA was applied directly to the functionalized channel without a reporting label.

After hybridization, the sensor was washed in 10 mM PBS for 7 min. The target sequence came from EGFR exon 21. The paper notes that L858R accounts for approximately 40% of EGFR mutations.

The change in drain-source current, ΔIDS, was measured at fixed bias. This change was the difference between post-hybridization current and the baseline after probe immobilization.

ΔIDS increased with complementary DNA concentration and was linear with log concentration from 1 pM to 10 nM. The blank noise was 11.2 μA. The signal was 8.4 μA at 1 pM and 12.3 μA at 10 pM.

At an IDS of −15 μA, probe immobilization shifted VGS leftward by 40 mV. Target hybridization shifted VGS by 150 mV under the same current comparison.

All mismatch tests used 10 nM probes and targets. The paper reports the following offsets relative to complete complementary hybridization.

Base-pair mismatchesExperimental VGS shiftModeled VGS shift
226 mV33.9 mV
464 mV43.9 mV
883 mV54.5 mV

The model assumes an ideal, uniform probe monolayer. Normal DNA formed a compact double-stranded layer that coupled negative charge strongly to the channel.

Mismatched strands were looser, and their negative charges lay farther from the surface. The authors link weaker gate modulation to the resulting lower equivalent capacitance. At eight mismatches, current approached the probe-only baseline.

The authors also suggest that more mismatches reduce hybrid stability and stable binding. Against the cited planar diamond SGFET, which detected 3-mer mismatches at 10 pM, this device identified a 2-mer mismatch. The paper does not present a controlled cross-device comparison.

Repeatability, stability, and interference

Section titled “Repeatability, stability, and interference”

Hybridization and dissociation were repeated with DNA containing two, four, or eight mismatches. Repeated measurements produced small error bars and a smooth normalized current profile.

Residual current after dissociation differed from the probe-only baseline. The authors attribute this difference to incomplete dissociation, which did not change comparisons under controlled operation.

Storage testing occurred every 5 days at 20 °C and 10% relative humidity. Current declined over 10 days, with reductions of 17.54% after 5 days and 25.31% after 10 days.

The authors associate this drift with inactive DNA probes and rising channel resistivity.

Anti-interference testing used 1 mg/mL lysozyme in 10 mM PBS at pH 7.4. Current and signal amplitude decreased, but mismatch-dependent trends remained. Positively charged lysozyme repelled holes and reduced current without erasing the discrimination.

Engineering implications and unresolved issues

Section titled “Engineering implications and unresolved issues”

The data favor oxygen-terminated BDD for a neutral aqueous gate. They also show that pH, buffer concentration, boron doping, and surface chemistry must be controlled together. Doping sets carrier availability, while termination sets the electrochemical surface response.

Device geometry should favor wide, short microwires, but the final 7.68 μm width was below the 10 μm simulation target. Etch tolerance therefore affects the modeled performance gain.

The mutation model assumes a uniform, closely packed, ordered probe layer. Actual probe density varies, nonspecific adsorption occurs, and environmental conditions vary between runs.

The reported 10 pM detection limit needs a clearer noise calculation. The printed 12.3 μA signal at 10 pM is not three times the 11.2 μA blank noise. The noise statistic may require clarification.

Longer storage, a wider interference panel, and patient samples remain untested. The results support analytical use for selected sequences, not yet clinical validation.

For this gate-channel application, 6CCVD’s diamond materials catalog lists MPCVD-grown single-crystal and polycrystalline diamond, boron-doped diamond, Au metallization, and substrates up to 10 mm. The paper’s 200 nm boron-doped layer on a 10 × 10 mm single-crystal substrate falls within the listed 0.1 to 500 μm thickness range. Its microwire patterning and oxygen-plasma surface treatment can be discussed through engineering support.

View Original Abstract

Gene mutation is one of the core pathogenic factors in numerous major diseases, making the detection of base mismatches resulting from these mutations critically important in both biological and clinical contexts. This study presents a high-performance boron-doped diamond solution-gated field-effect transistor (BDD-SGFET) biosensor, designed with a diamond microwire structure, for the label-free detection of base mismatches associated with EGFR gene mutations. The simulations examining the impact of variations in diamond microwire dimensions on the electrical properties of BDD-SGFET reveal that increasing the microwire width while reducing its length enhances the electrical performance of the device. Utilizing microwave plasma chemical vapor deposition (MPCVD), photolithography, and plasma etching, we successfully fabricated high-performance BDD-SGFETs featuring microwire structures that demonstrate outstanding transconductance, a reduced threshold voltage, and a limit of detection of 10 pM. Notably, the enhanced performance of the fabricated BDD-SGFET enables the successful identification of DNA molecules with two base-pair mismatches. Furthermore, the device exhibits impressive anti-interference capabilities and exceptional stability in complex environments. These findings highlight the significant potential of microscale BDD-SGFETs as rapid, label-free, and robust platforms for point-of-care testing in genetic mutation analysis pertinent to cancer diagnosis.