Can a rougher, defect-ridden material actually outperform perfect crystal graphene? This study reveals that reduced graphene oxide beats pristine CVD graphene for detecting heart failure markers in saliva. Graphene-based materials are actively being investigated as sensing elements for the detection of different analytes in biosensors.

The authors compare chemical vapor deposition graphene with graphene oxide to find the most efficient material for point-of-care diagnostic devices. Both sensor types demonstrate a good dynamic range from ten femtograms per milliliter to one hundred picograms per milliliter.

NT-proBNP is considered a promising heart failure marker for point-of-care diagnostic applications because it is released when heart muscles are damaged. Detecting this biomarker in saliva is challenging because its concentration can be lower than one picogram per milliliter.

The high sensitivity of field-effect transistor-based biosensors allows for measuring even the lowest concentration of the biomarker in saliva. Two types of field-effect transistors were investigated using commercially available interdigitated electrodes.

The reduced graphene oxide device was produced by drop-casting a suspension onto gold electrodes modified with cysteamine. Monolayer graphene on copper foil was transferred onto the working electrodes using a wet transferring method with poly(methyl methacrylate) support.

Figure 1 characterizes the physical and chemical properties of the graphene films used in the sensors. The AFM images reveal that the CVD graphene forms an intact surface, while the rGO film consists of interconnected flakes with a rougher texture.

Finally, the Raman spectra confirm the high quality of the transferred graphene by showing distinct intensity peaks for the G and 2D bands, which are characteristic of monolayer structures. Raman spectra confirm that the transferred CVD graphene has a low defect density and high quality.

On the contrary, reduced graphene oxide demonstrates a high number of defects even after reduction and almost the absence of a 2D band. This confirms that the surface still contains vacancies and other defects that give an impact in the high channel resistance. Table 1 compares the electrical characteristics of two biosensor types: GFETs and rGO-FETs.

The authors report that while both devices show similar initial transconductance values around eighty-five microsiemens, the change in transconductance after assembly is notably larger for the rGO-FETs at forty-seven point one microsiemens compared to seven point five microsiemens for the GFETs.

Additionally, the table lists a Raman band intensity ratio greater than or equal to five point one one for the graphene channels versus zero point seven six for the rGO channels, indicating differences in material quality between the two substrates. For the CVD graphene field-effect transistor, the transconductance was increased only by ten percent after assembly.

However, for the reduced graphene oxide configuration, the increase in values was more than eighty percent. This transconductance increase can be associated with ionic redistribution in the Stern layer that can modulate the thickness to demonstrate high sensitivity.

This figure illustrates the electrical response of graphene and reduced graphene oxide field-effect transistors during aptasensor assembly. Panels a and b display current-voltage curves, revealing that attaching the PBASE layer causes opposite shifts in the Dirac point for the two materials: a left shift for the graphene device and a right shift for the reduced graphene oxide device.

Panel c quantifies these changes, showing how the magnitude of the Dirac point shift varies significantly depending on the buffer concentration used. Finally, panel d provides schematic diagrams explaining the underlying electrical mechanisms driving these distinct responses.

Measurements revealed different Dirac point shifts in terms of value but similar direction for both sensor types. Transconductance did not change for the CVD graphene aptasensor, confirming its intact structure and the absence of trapped states associated with defects.

For the reduced graphene oxide device, both transconductance and Dirac point shift are more pronounced due to the effect of binding on electrostatic doping. Figure 3 compares the performance of graphene and reduced graphene oxide field-effect transistors as sensors for the biomarker NT-proBNP.

The top panels show that in a standard buffer, increasing concentrations cause distinct shifts in the electrical transfer curves and Dirac points for both device types. The bottom panels demonstrate how these responses change when tested in artificial saliva, revealing significant baseline shifts due to differences in pH and ionic strength compared to the pure buffer environment.

Table 2 compares the analytical performance of two aptasensor types, CVD GFETs and rGO-FETs. Both sensors share an identical dynamic range from ten to the power of negative two to ten squared pg/mL. However, they differ in their detection limits and sensitivity; the GFET shows a limit of detection of one pg/mL with a sensitivity of approximately zero point seven mV per decade, while the rGO-FET reports a limit of detection of zero point one pg/mL and a sensitivity of roughly two point five mV per decade.

The results indicate that when choosing technology for biosensors, one should consider not only the internal properties of the sensing materials but also the analyte. To increase sensitivity and decrease noise, materials with a larger bandgap and presence of trapped states, such as reduced graphene oxide, are preferred.

The authors demonstrated that the aptamer-based assay for NT-proBNP analysis can be performed in diluted artificial saliva. While CVD graphene offers stability, reduced graphene oxide provides superior sensitivity for small peptide detection in complex fluids like saliva, making it the better choice for point-of-care heart failure diagnostics.