PEDOT, or poly(3,4-ethylenedioxythiophene), is one of the workhorses of organic electronics. It conducts well, is partially transparent, and bends without breaking, which makes it attractive for flexible devices. But PEDOT is hydrophobic, so it must be processed in polar solvents such as water, and that requires stabilizing additives. The most common partner is polystyrene sulfonate, PSS, an insulating polymer that keeps PEDOT dispersed in water and enables the commercial conductive inks used in solar cells, light-emitting diodes, batteries, and supercapacitors. The same ink also serves as the active layer in organic gas sensors, where it adsorbs specific gases or vapors and converts those interactions into an amplified electrical signal.

The Brazilian-Irish collaboration prepared conductive inks by adding precise volumes of commercial PEDOT:PSS to an aqueous graphene oxide dispersion, producing blends labeled GO:PEDOT:PSS with PEDOT concentrations ranging from 40 to 88 percent. Graphene oxide was synthesized by oxidizing graphite and exfoliating the resulting sheets in water. The team deposited thin films of each blend onto flexible acetate substrates fitted with interdigitated chromium-gold electrodes, then characterized their morphology with scanning electron microscopy and their electrical properties with four-probe measurements. The films were roughly 250 nanometers thick, and the sensors were tested at 21 degrees Celsius with 1,000 parts per million of methanol in dry air.

The electrical measurements confirmed a long-suspected optimum. The lowest resistivity appeared in the GO:PEDOT:PSS 78 percent blend, corresponding to a GO-to-PEDOT mass ratio of 5.8 times 10 to the minus 2 that previous studies had identified as ideal for conductivity and optical transmittance. Earlier work attributed this optimum to improved alignment of the thiophene groups within PEDOT and to additional doping facilitated by graphene oxide. The new study replicated an ethyleneglycol treatment that selectively removes PSS from the dried films, and the resistivity minimum persisted at the same 78 percent concentration, suggesting the optimum is a robust property of the blend rather than an artifact of processing.

When the same films were exposed to methanol vapor, the sensing optimum shifted. The highest response, approximately 18 percent, was achieved with GO:PEDOT:PSS 68 percent, not the 78 percent blend that conducted best. The flexible devices reproduced the qualitative behavior previously reported for sensors on glass substrates, with response values falling within the same order of magnitude. The researchers attribute the slight differences to the flexible nature of the acetate substrates, where surface interactions between the inks and the substrate alter film morphology, thickness, and exposed active area, all of which influence how methanol interacts with the active layer.

To explain why the conductivity and sensing optima occur at different mass ratios, the team turned to a theoretical model based on Kirchhoff’s circuit approach. They represented the blend as a 100 by 100 square lattice in which every node is labeled GO, PEDOT, or PSS, and every link between nodes carries a resistance encoding the interactions between those constituents. The equivalent resistance between two electrode sites is calculated from Kirchhoff’s matrix, which encodes the connectivity and interactions of the network. Because graphene oxide and PSS are insulators while PEDOT conducts, the network exhibits a competition of resistances. Below the optimum, there is not enough PEDOT to overcome the insulating pathways; above it, the sample approaches the conductivity of pure PEDOT. The calculated resistivity minimum landed at approximately 78 percent PEDOT, in agreement with the experimental data.

The same lattice model was extended to sensing by modifying the link resistances to represent methanol exposure, with the alcohol distributed throughout every component as in a saturated atmosphere. The calculated response reached its maximum at approximately 65 percent PEDOT, closely matching the 68 percent blend that performed best experimentally. The logic mirrors the resistivity case: although both graphene oxide and PSS interact with methanol, the active layer must conduct the signal to the external circuit, and PEDOT is the only conductive constituent. Graphene oxide increases the number of free charges available, and methanol molecules consume those polarons, raising the device resistance and enhancing the sensing signal at specific concentrations.

The study also compared commercial PEDOT:PSS inks with chemically synthesized aqueous PEDOT produced without PSS, using iron(III) chloride to polymerize the EDOT monomer in acetonitrile and water. The polymerized GO:PEDOT blends showed an average particle size of about 480 nanometers and a zeta potential of plus 45.8 millivolts, indicating a metastable state that can be stored at 4 degrees Celsius for up to four months without coagulation. Their resistivity values, approximately 0.35 and 0.45 ohm centimeters for GO:PEDOT and PEDOT respectively, remain competitive with the commercial ink. Scanning electron microscopy revealed that PEDOT and GO:PEDOT:PSS films are homogeneous, while GO:PEDOT films contain clusters of graphene oxide sheets coated by polymeric PEDOT, a morphology that limits cluster connectivity and keeps the sensing efficiency close to that of pure PEDOT despite additional doping.

Perhaps the most striking finding concerns the direction of the resistance change during methanol exposure. Commercial samples containing PSS showed an increase in resistance when exposed to methanol, while polymerized samples without PSS showed a decrease. The team attributes the increase to a swelling effect: PSS is hygroscopic, so polar solvents like methanol infiltrate the active layer, separate the polymeric chains, reduce connectivity, and consume free charges, raising resistance. In the absence of PSS, the decrease in resistance is associated with a reorganization of charges within the doped polymer, which can delocalize charge carriers and create new conduction pathways, temporarily increasing conductivity. Methanol is a reducing agent, so resistance changes in these devices are typically linked to doping or dedoping processes, and the polymerized samples appear to rely on charge redistribution rather than charge exchange.

The theoretical model reproduced both behaviors qualitatively, capturing the resistance increase in PSS-containing networks and the decrease in PSS-free ones. The researchers note that graphene oxide occupies a unique position in the network: it is a component with high electrical resistance but also features regions that can conduct electricity, creating the interplay between resistances that produces the contrasting responses. The model, simple as it is, addressed the three central questions of the study: the position of the resistivity minimum, the position of the response maximum, and the role of PSS in determining how methanol interacts with the active layer. The square lattice representation proved a powerful yet tractable method for studying such disordered systems, yielding results consistent with larger and more computationally demanding network models.

The implications reach beyond methanol. Organic gas sensors operate at room temperature, unlike inorganic counterparts that often require temperatures exceeding 100 degrees Celsius, which makes them appealing for confined spaces such as grain silos and fuel tanks where methanol accumulates as a toxic and flammable vapor. Lower operating temperatures demand less energy and reduce accident risk. The sensors in this study were characterized at 21 degrees Celsius with 0 percent humidity, since water vapor can hinder methanol detection through a swelling effect. The polymerized PEDOT inks, produced exclusively in water without stabilizers or surfactants, offer enhanced ecological appeal and reduced production costs, and the sensors made from them showed greater stability than commercial counterparts, with faster saturation times and no observable drift in the resistance curves.

By combining experimental characterization with a circuit-based model, the study provides a coherent picture of how methanol detection works in graphene oxide and conductive polymer active layers. The resistivity minimum and the response maximum arise from the same competition of resistances, but at different mass ratios because sensing requires both interaction with the vapor and conduction of the signal. PSS, the insulating polymer that enables water dispersibility, turns out to be essential for the doping effect that graphene oxide provides, amplifying the disruption of PEDOT doping caused by methanol molecules. The work demonstrates that a simple square lattice model, grounded in Kirchhoff’s laws, can quantitatively and qualitatively explain the behavior of real nanostructured sensing films, offering a theoretical foundation for designing next-generation flexible gas sensors.

Subject of Research: Materials Science

Article Title: Mechanisms of methanol detection in graphene oxide and conductive polymer active layers for gas sensing devices

Article References: das Neves, M. F. F., Mukim, S., Ferreira, M. S., & Roman, L. S. (2024). Mechanisms of methanol detection in graphene oxide and conductive polymer active layers for gas sensing devices. Frontiers in Carbon, 3, Article 1352122. https://doi.org/10.3389/frcrb.2024.1352122

Image Credits: AI Generated

DOI: 10.3389/frcrb.2024.1352122

Keywords: graphene oxide, PEDOT:PSS, methanol sensor, gas sensing, flexible electronics, conductive polymers, Kirchhoff model, organic electronics, Mechanisms, methanol, detection, graphene