Near-Room-Temperature Detection of Aromatic Compounds with Inkjet-Printed Plasticized Polymer Composites

Chemiresistive gas sensors composed of a thermoplastic polymer matrix and conductive fillers offer various advantages for detecting volatile organic compounds (VOCs), including low power consumption due to near-room-temperature operation, high sensitivity, and inherent selectivity toward VOCs. However, such sensors have a slow response time as the polymer matrix often has a glass transition temperature (Tg) higher than the sensor operating temperature slowing the analyte diffusion to/from the polymer. A plasticizer lowers polymer Tg to match the sensor operation temperature, reducing its response time. In this study, the effect of a plasticizer diethylene glycol dibenzoate (DEGDB) on the sensing properties of polystyrene (PS)-carbon black (CB) composite is investigated to obtain sensors with a fast response time and high sensitivity to VOCs. The sensors are fabricated via drop-on-demand inkjet printing, providing a high degree of control over the sensory film morphology and reproducibility. A design-of-experiment (DoE) approach is adopted to find the optimum ink and print parameters with a minimum number of experiments. As a result, sensors with 30 times faster response time and 25 times higher effective sensitivity are obtained while operating near room temperature (27 °C). Furthermore, the sensors show high sensitivity toward aromatic hydrocarbons (toluene, benzene, and ethylbenzene), with a sub-10 ppm limit of detection (LoD) and a negligible sensitivity toward humidity. Our results show the potential of PS-DEGDB-CB composite as a selective and cost-effective sensory material compatible with large-scale manufacturing techniques for selective near-room-temperature detection of toxic VOCs.


Material characterization
Figure S1 shows the thermogravimetric analysis (TGA) of PS and DEGDB under synthetic dry air flow.Figure S5 shows the sensors' dynamic responses after exposure to a contact acetone concentration.Table S3 shows the vapor pressures of the used analytes calculated using the Antoine equation.
Table S3: Vapor pressure of target analytes calculated based on the Antoine equation (log(p)=A-B/(C+T)) [1] Analyte Equations S1 and S2 below are used to calculate the HSP distance and the RED number.
Eq. S1 Eq. S2 Table S4 below shows the HSPs of the analytes used for sensor characterization.
Table S4: Hansen Solubility Parameters of target analytes and polystyrene used to calculate the HSP distance. [2]alyte/polymer Polar (MPa  Table S6 shows the HSPs of PS, DEGDB and PS-DEGDB mixture.The HSPs of DEGDB are calculated using a group contribution method proposed by Stefanis et.al. [2]as we could not find any experimental studies regarding the solubility parameters of DEGDB.The HSPs of the PS-DEGDB mixture are then calculated based on the nominal volume fractions in a mixture containing 72 vol% PS and 28 vol% DEGDB, corresponding to 70 wt% PS and 30 wt% DEGDB in the organic matrix.
Moreover, based on the computed values, the solubility distance and the RED number for the PS-DEGDB pair are 8.5 MPa 1/2 , and 1.6, respectively, indicating that PS and DEGDB are not fully miscible.The solubility limit of PS-DEGDB is presumably linked to higher polarity and hydrogen bonding numbers of DEGDB than PS.Equation S3 below is used to calculate the analyte concentration.
Table S7, below, shows a comparison between the key features of chemiresistive sensors based on their sensing materials, including metal oxides (MOX), Graphene, transitional metal dichalcogenides (TMDCs), carbon nanotube (CNT), MXenes, metal-organic frameworks (MOFs), polymers and polymer composites.The selected sensors show the performance upon exposure to acetone, ethanol, and aromatic hydrocarbons for comparison with the sensors fabricated in this work.
The main interest in conductive polymers and polymer composites for VOC sensing is their operation at room temperature, resulting in lower power consumption than MOXs.Moreover, polymers are inherently selective towards various VOCs, considering their chemical structure.
The affinity of a polymer to VOCs can be easily estimated via the application of Hansen Solubility Parameters, which allows for predicting the sensor's behavior.Moreover, polymers can be readily processed and adapted to various additive fabrication methods, which is convenient for depositing sensing materials.However, as explained in the manuscript, some

Figure S1 :Figure S2 :
Figure S1: Thermogravimetric analysis (TGA) of (a) polystyrene and (b) DEGDB.The TGA is performed under a synthetic dry air flow.The temperature range is set from 30 to 400 °C for DEGDB and 30 to 700°C for PS characterization, with heating rates of 10°C/min.Based on the TGA results, PS and DEGDB are stable up to ~350 and ~200°C, respectively.

Figure
FigureS3shows the surface profiles of inkjet-printed sensors measured by a Dektak mechanical profilometer.

Figure S3 :
Figure S3: Film thickness measured with a Bruker Dektak mechanical profilometer.The thickness of the sensory films printed with a dot spacing larger than 200 µm becomes comparable with alumina surface roughness, making it difficult to measure the thickness.

Figure
Figure S4 shows SEM images of the inkjet-printed sensory films containing different concentrations of DEGDB.The dark zones are CB-rich, and the bright ones contain little or no CB.

Figure S5 :
Figure S5: Effect of sensor exposure to acetone.Each sensor is exposed to 0.4 % acetone three times during 5 min exposure and 5 min recovery cycles.The sensors are measured at 27°C.The initial drop of the sensor resistance results from increasing the temperature from the room temperature to 27°C.Subsequently, exposing the sensors to acetone results in increasing the baseline resistance.The last figure shows the normalized sensor response during the third exposure.

Figure
FigureS6shows a comparison between the dynamic response of non-plasticized and plasticized composites.

Figure S6 :
Figure S6: Direct comparison between the dynamic response of PS-CB composites with (30 wt%) and without plasticizer.(a)Dynamic responses upon exposure to 0.4% acetone.The dotted rectangle shows the portion of the sensor response used for fitting.(b) fitting the dynamic response by a double exponent equation to estimate the time for each sensor to reach its equilibrium.The equation and fitted parameters are shown in the inset of the figure.Extrapolating the data shows that it takes the sensor 00-135 approximately 60 minutes to equilibrate, whereas the sensor 30-135 reaches the steady-state in less than 2 minutes.Moreover, 30-135 fully recovers after a few minutes, whereas 00-135 only partially recovers, resulting in a significant baseline drift after each exposure to acetone.

Figure
Figure S7 shows he micrographs of the sensory films containing different CB concentrations are shown in figure S8 below.

Figure S7 :
Figure S7: Optical microscope images of the printed sensory film containing different CB concentrations.The sensory films contain (a) 10 wt%, (b) 8 wt%, (c) 6 wt%, (d) 4 wt%, (e) 2 wt%, and (f) 1 wt% CB.It is observed that the film morphology and uniformity are affected significantly by decreasing the CB concentration

Figure S8 :
Figure S8: SEM images of sensory films containing (a) and (b) 10 wt%, (c) 8 wt%, (d) 6 wt%, (e) 4 wt%, and (f) 2 wt% CB.The microstructure of the sensory film changes by decreasing the CB concentration, becoming visibly coarser and less homogeneous when the CB concentration is decreased.

Figure
FigureS9compares the HSP distance of PS-analyte pairs versus that of the PS-DEGDB-analyte pairs.It is clear that due to the higher polarity of the PS-DEGDB mixture, the solubility distance to polar compounds decreases, whereas the distance to the non-polar compounds increases.In other words, the composite containing the PS-DEGDB mixture becomes slightly more sensitive to polar compounds than the sensor containing only PS.However, the general trend does not change, indicating that the solubility distance and RED numbers considering only PS can explain the sensors' behavior.

Figure
Figure S9: A comparison between solubility distance of PS and PS-DEGDB mixture with various analytes.Due to the higher polarity of the PS-DEGDB mixture, it has a lower solubility distance to polar analytes and a larger distance to non-polar analytes compared to PS.However, the general trend, regarding the affinity of the sensor to the target analytes does not show a considerable change.

Table S1 :
The nominal weight and volume fractions of PS, DEGDB, and CB in dry composites.The sensors printed from the following composites were used to find the optimum DEGDB concentration.

Table S2 :
The nominal weight and volume fractions PS, DEGDB, and CB in dry composites.The sensors printed from the following composites were used to study the effect of CB loading on the sensor response.

Table S5 :
HSP distance ( ) and number calculated from the HSP values of Table S4.The polystyrene solubility radius

Table S6 :
HSPs of PS, DEGDB, and PS-DEGDB mixture, The HSPs of the DEGDB are calculated using the group contribution polymers, such as PS, offer interesting sensing properties for detecting VOCs, but they are limited by the kinetics of the sensor response.Here it is shown that additives such as plasticizers can significantly improve the polymer's sensing performance without affecting the processability and sensitivity of PS-based composite.As shown in the table, the performance of PS-DEGDB-CB is comparable to examples of room temperature MOXs, 2D materials, CNT, and polymer composites.However, it lags behind more advanced materials such as hetero MOXs, operating at elevated temperatures, MOF-MOX hybrids, and MXenes.Nevertheless, PS-DEGDB-CB still offers comparable performance for detecting aromatic compounds at/near room temperature.Moreover, the composite formulated here is a suitable material for stretchable and flexible sensor applications.

Table S7 :
Comparing key features of chemiresistive sensors composed of various sensing materials