Current Issue

Different types of electrospinning techniques and instruments are available. However, they differ in the type of collector and spinneret. Some of them use an electrode material as a spinneret, whereas some others are needleless (Afshari, 2016). Co-axial electrospinning, co-electrospinning ( Jiang et al., 2005a), multi-needle electrospinning (Tomaszewski and Szadkowski, 2005) and needleless electrospinning (Wang et al., 2009a) are the main types of electrospinning processes. Multi-needle and needleless electrospinning methods are used to improve the efficiency of the traditional electrospinning method. Co-axial electrospinning is utilized to prepare multi-layer and core-shell composite NFs structures with more functionality and better quality (Zhang et al., 2004). In this technique, two discrete NF building blocks are provided via coaxial narrow channels and are synthesized into core-shell NFs (Lim, 2017). Co-electrospinning is also capable of producing single layer and bilayer NFs (Starr and Andrew, 2013). Multi-layer electrospinning can also be utilized to prepare hierarchical NF morphologies (Kayaci et al., 2014; Li et al., 2014; Tran and Kalra, 2013).

Structure, porosity, diameter, and surface area play important roles in the final features of the NFs, and fortunately they can be controlled during the electrospinning process (Tan et al., 2005). The parameters involved in the electrospinning method should be carefully optimized to synthesize NFs with better performance in gas sensors. We discuss electrospinning parameters in the following sections. More details can be found in Tan et al. (2005).

In resistive-based gas sensors, the electrical resistance greatly depends on the ambient gases in terms of type and concentration (Mirzaei et al., 2016). Such gas sensors are particularly suitable for detecting different gases and concentration variations (Neri and Donato, 1999). Fabrication of the sensor includes the deposition of a sensing layer (CPs) on an alumina, SiO₂/Si or flexible substrate that has interdigitated electrodes (e.g., Au, Pt) for reading the sensor resistance. A heater is also used on the back side of the sensor to heat it to the desired sensing temperature.

The electrospun CP-based gas sensors discussed in this review can be used to detect many toxic, VOC, and flammable gases such as H₂, NO₂, CO₂, triethylamine, methanol, and liquefied petroleum gas (LPG). In Table 1, the main source of emission, hazardous effects, and the threshold limit value (TLV) of the most important gases are presented.

The sensing capabilities of CPs are due to interactions of the target gas with the CP sensing materials, which can result in changes in conductivity, the doping level and electron transport in the conjugated backbone. It has been suggested that there are five possible interactions between the CP NF sensing layer and the target gas that can alter the conductance (Neri and Donato, 1999), as shown in Fig. 4. First, such a response may be caused by how the target gas affects the charge transfer between the electrode and CP. Second, this response may be due to direct creation or elimination of the charge carriers, e.g., via oxidation or reduction of the polymer chain. In particular, this interaction may be significant for NH₃, although this seems less likely for weakly interacting organic vapors such as alcohols (Wetchakun et al., 2011; Gardner and Bartlett, 2000). Third, the charge carrier mobility in a polymer chain can be altered via interactions between the mobile charge carriers and the target gas across the polymer chains. Fourth, the target gas can interact with counter ions (X^-) in the sensing layer, thus altering the mobility of the charge carriers along the polymer chain. Counter ions with relatively low mobility are generally connected in the polymer chains. The target gas diffuses into the polymer film and acts as a solvent for the counter ions, resulting in ionic conduction (Dixit et al., 2005). Finally, the target gas can modulate the resistance of a CP film by affecting the rate of interchain hopping.

Fig. 4. 
Possible interaction sites for a target gas on a CPs sensing layer. Reproduced with permission from (Neri and Donato, 1999), copyright 2020.

Electron transfer is the general sensing mechanism in CP-based gas sensors. Gases with an electron acceptor nature like NO₂ can eliminate electrons from CPs. It is well-known that at low temperatures (<100°C), O₂^- is the dominant chemisorbed oxygen type on the surfaces of most sensing materials. First, oxygen gas is adsorbed on the surface of the sensor and then (by taking of electrons from the surfaces of the sensor), it becomes adsorbed in the form of a molecular ion: O₂+e → O₂^-. By exposing CP-based gas sensors to a reducing target gas, the target gas molecules can react with the oxygen species and transfer electrons back to the conduction bands of the sensor. Such electrons transferred in the conduction band can be combined with holes from the valence band, thus leading to higher resistances and lower carrier concentrations for the CP gas sensors. In addition, the resistance decreases for oxidizing gases. However, a signal can appear in both cases. At room temperature, this type of mechanism may contribute less to the sensor signal in CP-based gas sensors, as the concentration of O₂^- is generally very low at room temperature (Densakulprasert et al., 2005).

Occasionally, a noble metal catalyst like Pd, Au, or Pt is added to the CP film, which allows for the detection of various less reactive gases (Shirsat et al., 2009). Noble metals can catalytically decompose and increase the adsorption of target gases via a spill over process. Furthermore, since the work functions of noble metals and CPs are different, heterojunctions can form between a noble metal and a CP, which can change the resistance of the sensor by modulating the height of the potential barrier formed in the presence of target gases. Hybrid composites of CPs and n-type metal oxides show synergic interactions between the CPs and metal oxides, resulting in the generation of p-n junctions. For example, in a hybrid consisting of PPy (which is a p-type material) and WO₃ (which is an n-type material), the p-n junctions formed in the PPy-WO₃ p-n heterojunctions can enhance the depletion barrier height, resulting in changes in the sensor response (Sun et al., 2017). Swelling is another mechanism present in CP-based gas sensors (Thenmozhi et al., 2017); as the gas molecules diffuse into the polymer chains, electron hopping becomes difficult because of the increased distance between polymer chains resulting from swelling. Swelling can disrupt conductive pathways in the CP film and increase the sensor resistance when the desired gas is injected, thus enhancing the gas response of the sensor (Guernion et al., 2002).

In the next sections, we analyze how the above gas sensing mechanisms can explain the sensing behavior of CP-based gas sensors. The sensing parameters (e.g., sensing range, sensitivity, DL, response time, and selectivity) of all the covered CPs-based sensors are shown in Table 2.

Among many CPs, PANI has been the most studied material due to its various characteristics such as reversible doping, dedoping, tunable electrical conductivity, high environmental stability and easy preparation (Sen et al., 2016). PANI can be synthesized using the oxidizing polymerization of aniline from two structural units, i.e., a reduced (B-NH-B-NH) unit and an oxidized (B-N=Q=N^-) unit, where B is denoted as a benzenoid and Q is denoted as a quinoid ring.

PANI has the ability to undergo reversible doping under acidic (proton doping) or basic conditions. When PANI is doped with an acid, the imine nitrogen of the polymer backbone is protonated to induce charge carriers. In this situation, the electrical conductivity of PANI increases (σ>1 S cm^-1), and this doped form of PANI is known as an emeraldine oxidation state or emeraldine salt (Fig. 5). The un-doped polymer chain is known as an emeraldine base (σ<1×10^-10 S cm^-1) (Fratoddi et al., 2015; Huang et al., 2004).

Fig. 5. 
Repeat unit of the emeraldine oxidation state of polyaniline in the un-doped base form (top) and the fully doped, acid form (bottom). Reproduced with permission from (Huang et al., 2004) from Springer, copyright 2020.

Various nanostructures of PANI including NFs, nanotubes, and nanoparticles have been synthesized by numerous of methods (Kumar et al., 2017, 2016; Gu et al., 2014; Ayad et al., 2013; Long et al., 2011; Rahy and Yang, 2008). These morphologies have high exposed surface area (e.g., PANI NFs with 50 nm diameter have 49.3 m²/g surface area), which can facilitate gas molecules adsorption and greater penetration into the pores (Huang et al., 2004). Introducing a secondary component to the primary PANI nanostructures further enhances its sensing properties due to the synergistic effect between both the components. Various materials such as metallic nanoparticles, metal-oxides, carbonbased materials (CNT, Graphene), and polymers have been added to PANI to tune the sensitivity and selectivity properties (Park et al., 2017). In this section, we discuss gas sensors based on PANI NFs as a sensing layer in their pristine or composite form.

For the sensing of H₂, PANI NFs were fabricated as a resistive sensor on Pt or Au electrodes (Fowler et al., 2009). The responses (R_g/R_a) of the sensors with Pt and Au electrodes were 1.65 and 1.03, respectively. The reason for the low sensor response on an Au electrode is the interaction between hydrogen and PANI NFs. The high response of the sensor on Pt electrodes is because of interactions with Pt at the PANI-Pt interface. Hydrogen helps to form a Schottky barrier between Pt and PANI via a variation in the work function as Pt was changed to Pt hydride. The formation of a hydride between the PANI NF layer and the platinum electrode resulted in non-ohmic behavior in the Pt-PANI NFs upon exposure to a H₂ environment. Pt is introduced to make a platinum hydride in contact with H₂ gas.

The mechanism of hydride formation is shown in Fig. 6. The sensor includes a porous mesh of PANI NFs on Pt electrodes (Fig. 6a). When this sensor is in contact with H₂, the gas passes through PANI and reaches the Pt surfaces (Fig. 6b-d). However, some of the hydrogen can also react with PANI before reaching the Pt. Platinum hydride (PdH_x) is produced at the interface of the PANI NFs and the platinum metal. When the PANI NFs mesh is exposed to oxygen, it reacts with the PdH_x surface. Water and Pt metal form due to the elimination of hydride. Because the work function for PANI NFs is between the work functions for Pt and PdH_x, H₂ results in the formation of a Schottky barrier, which is the main reason for the sensitivity of the Pt-PANI NFs gas sensor. The hydrogen exposure results in large changes in the work function of the Pt contact. Metal hydride cannot form on Au electrodes, and ohmic contacts with the PANI NFs are formed, and only the interaction between PANI NFs and H₂ are observed. The small change in resistance is due to binding between H₂ and PANI NFs. In fact, platinum hydride forms via the Pt-PANI NF-hydrogen interactions, and Pt hydride formation is the main reason for the improved sensor response.

Fig. 6. 
Interaction of PANI-coated platinum-based sensor with the H₂ gas. (a) Sensor structure and (b-d) interaction of hydrogen and oxygen on the surface of sensor. (e) Work functions of Pt, Au, PtH_x. Reprinted with permission from (Fowler et al., 2009). Copyright (2020) American Chemical Society.

In another work, camphorsulfonic acid (CSA) doped PANI NF thin films were utilized for sensing of hydrogen (Virji et al., 2006). The resistance change of PANI NFs was greater than conventional PANI due to their porous structure backbone that leads to enhanced diffusion of gas into the film. Hydrogen gas sensing studies showed that the response (R_a/R_g) of the gas sensor was 1.03 for 10000 ppm H₂ gas. Fig. 7 shows a gas sensing mechanism for the interaction of H₂ with PANI. H₂ interacts with doped PANI at the charged amine nitrogen sites. Formation of new N-H bonds to the amine nitrogen of the PANI chains lead to dissociation of H₂ bonds. This reaction is quite reversible because charge transfer among adjoining amine nitrogen groups brought the PANI back to its doped, polaronic, emeraldine salt state free of H₂. This mechanism just occurs in the emeraldine salt form of PANI.

Fig. 7. 
Possible mechanism for hydrogen interaction with doped PANI, where A represents the counteranion. Reprinted with permission from (Virji et al., 2006), copyright (2020) American Chemical Society.

The PANI base emeraldine form is electrically insulating. Thus, there is no transfer of charge among the nitrogen units on the PANI chain, hindering the interactions between H₂ and the polymer. In other words, the response of dedoped PANI towards hydrogen is negligible since hydrogen cannot dissociate to interact with the dedoped PANI. The presence of water can also significantly decrease the response of a sensor for hydrogen. Since water molecules have a high affinity to bond with the nitrogen atoms of PANI, the available sites for adsorption of hydrogen decrease, resulting in a low response of sensors to hydrogen. The doping and dedoping in PANI is the crucial for the sensing of H₂. In a report on conductometric H₂ sensor, it was observed that the doped form of PANI displayed higher sensitivity in comparison to the dedoped PANI. Nonetheless, the dedoped PANI-based H₂ sensor outperformed doped PANI in terms of repeatability and baseline stability (Sadek et al., 2007). The composite forms of PANI are also found excellent in improving the H₂ sensing capabilities of PANI. In particular, the composites of PANI with graphene-materials (e.g., reduced graphene oxide; rGO), metal organic frameworks (MOFs), and SnO₂ displayed good performance for the sensing of H₂ gas (Mashao et al., 2019; Zou et al., 2016; Sharma et al., 2015; Al-Mashat et al., 2010). The performance of these sensors can be accessed from Table 2.

The conductivity of PANI composites may increase by including metallic particles or acidic or oxidative doping (Ayad and Zaki, 2008). The selectivity of PANI gas sensors may increase by adding different kinds of dopants like protonic acids (Samui et al., 2001), surfactants (Stejskal et al., 2003) and metals (Wong et al., 2019). Mixing PANI with nanosized metal oxides can increase the selectivity and sensitivity of the CPs. The sensors worked at ambient temperature, and the selectivity to different gases could be improved by increasing the concentration of nanosized metal oxides. Nimkar et al. (2015) prepared TiO₂ incorporated PANI NFs using the electrospinning technique. The NFs of the PANI/TiO₂ nanocomposite showed a good response (R_g/R_a) of~2.68 towards 0.1% CO₂ gas at 48°C. Also, the response time of the PANI/TiO₂ NFs film for 0.1% CO₂ gas was~80 s. When PANI/TiO₂ NFs films were exposed to CO₂ gas, the gas penetrated into the PANI matrix, and the improved sensitivity of this sensor was related to the formation of a positively charged depletion layer. Electron transfer from TiO₂ to PANI occurs at the heterojunction. As a result, the activation energy and enthalpy of physisorbed CO₂ gas decreased. The excellent reversibility and the short response time of the PANI/TiO₂ NFs layer is due to the surface morphologies of NFs. Specifically, the gas can diffuse in a NF structure more easily, resulting in large exposure regions and high penetration depths for gas molecules. As shown in Table 2, the composites of PANI with other potential structures, e.g., cross-linked bacterial cellulose-amino graphene (CLBC-AmG) and LaFeO₃ microsphere (PLFO) are also efficient for the sensing of CO₂ (Abdali et al., 2019; Hashemi Karouei and Milani Moghaddam, 2019).

Hybridization of CPs and metal oxides can enhance the features of the resulting gas sensor, and newly fabricated materials demonstrate the synergistic results of these two materials. In an interesting study, electrospun D-camphor-10-sulphonic acid (CSA) doped PANI/polyethylene oxide (PANI/PEO) and PANI/ZnO/PEO NFs were synthesized (Patil et al., 2015). A sol-gel method was employed for the synthesis of ZnO NPs. PANI (in the form of emeraldine base) and PANI/ZnO were prepared by chemical oxidative polymerization with CSA to improve conductivity and solubility. These were mixed with PEO solutions to obtain NFs by the electrospinning method. The LPG sensing capabilities of the sensors were tested at 30-90°C, and the optimal sensing temperature was 36°C, which is a safe temperature for LPG detection due to the explosive nature of LPG gas. The LPG gas sensing results showed that PANI/ZnO/PEO sensors had a higher response to LPG relative to PANI/PEO sensors. In particular, the responses (R_g/R_a) to 1000 ppm LPG were 1.087 and 1.073 for the PANI/ZnO/PEO and PANI/PEO gas sensors, respectively. CPs are often doped/undoped by redox reactions. Thus, transferring electrons can change the doping level and alter the resistance and work function of the CPs. The transfer of electrons happens when PANI and PANI/ZnO composite NFs were in contact with LPG gas. Electron acceptors remove electrons from PANI, resulting in an increased doping level and electrical conductivity in p-type CPs. In contrast, when an electro-donating gas such as LPG is used, the doping level of PANI decreases, thus changing the resistance of the sensor. Also, in PANI/ZnO/PEO NFs, p-n heterojunctions lead to additional resistance modulation, resulting in enhanced gas sensing. The other mechanism is the alteration of physical properties of the polymers such as crystallinity and morphology, which alter the electrical resistance via an increase in electron transfer. In such a case, the sensitivity of the sensor in the presence of LPG is related to adsorption on a polymer surface via diffusion to inner domain spaces.

After PANI CPs, polypyrrole (PPy) is highly popular due to its easy synthesis via chemical or electrochemical oxidation polymerization, high electrical conductivity and high stability. The material and sensing properties of PPy can be enhanced by combining it with additional inorganic or organic components. Usually a doped PPy behaves like a p-type semiconductor, where a change in conductivity occurs through interactions with a target gas ( Joshi et al., 2011; Kim and Yoo, 2011; Yoon et al., 2006; Penza et al., 1997). Lots of PPy composite materials have been reported for gas sensing applications for a wide range of gases or vapors such as NO₂, H₂S, water, and VOCs (Su and Shiu, 2011; Geng, 2010; Wang et al., 2009b; Han and Shi, 2007; Geng et al., 2006; Tandon et al., 2006; An et al., 2004; Bhat et al., 2003).

A highly porous core-shell polyacrylonitrile-(Tos doped) PAN-PPy NF mat was fabricated as a methanol sensor through electrospinning and two-step vapor-phase polymerization (Jun et al., 2013). The results demonstrated that very thin (~10 nm) conductive PPy shell layers were deposited on electrospun PAN NFs with a median diameter of 258 nm. The PAN-PPy gas sensor revealed a response (R_g/R_a) of 1.4 to 10000 ppm methanol at room temperature. The resistance increase upon exposure to methanol was explained by the electron donating character of the CH₃OH gas and the p-type semiconducting nature of PPy. The interactions between Tos-doped PPy and CH₃OH might be described by the redox (dedoping) reaction: PPy⁺/Tos^-+CH₃OH (δ^-) ↔ [PPy-CH₃OH (δ^-)]⁺/Tos^-. Likewise, PPy-based materials (e.g., poly (vinyl alcohol) (PVA), silver nanoparticle (Ag NPs), and plastic optical fiber (POF)) are also an effective option to sense other oxygen-containing VOCs, e.g., ethanol and acetone (Table 2) (Adhikari et al., 2020; Liu et al., 2020; Das and Sarkar, 2018; Jiang et al., 2005b).

The transfer of electrons from electron-donating CH₃OH to the positively-charged PPy backbone led to the formation of neutral polymer chains, resulting in a reduction in the hole carrier density (i.e., a resistance increase). In another study (Al-Mashat et al., 2008), PPy NFs were prepared through a simple chemical method for the detection of hydrogen at different temperatures. The preparation of PPy NFs was completed using bipyrrole to accelerate the polymerization of pyrrole with iron chloride as the oxidizing agent. The effects of temperature (up to 100°C) on the H₂ sensing properties were studied. At higher working temperatures (>70°C), the response was low (R_a/R_g=1.07 for 10000 ppm H₂ at 75°C) because a lot of carriers with high energy moved freely in the polymer chain, preventing the polymer from exhibiting its inherent redox reaction behavior. In addition, the nanostructured PPy film was smoother at high temperatures and included a dense NF morphology that had less porosity relative to the film at lower temperatures, which led to improved conductivity due to enhanced charge carrier hopping between localized states in the polymer chains. The high conductivity of the nanostructured PPy layer decreased the response at high temperatures because the gas molecules could not deeply penetrate into the bulk of the sensor. Accordingly, the sensor showed its maximum response (R_a/R_g=1.12 to 10000 ppm H₂ gas) at ambient temperature. A dynamic response was produced because of the interactions between the adsorbed hydrogen gas molecules and the polymer film. This interaction led to a heterogeneous charge transfer reaction, and chemical modulation of the polymer doping level, which depends on the Fermi level of the sensing film. This can cause a shift in the work function or electrical conductivity of the organic film. In terms of the work function, the polarity of the response relates to the ability of the diffusing gas to shift charge density with the polymer matrix by reduction or oxidation reactions. Hydrogen behaves as a reducing gas, so the PPy NFs resistance decreased when in contact with H₂ gas. However, the large variation in its electrical resistance is due to the redox reaction of PPy. The modification of PPy with Pt and Pd nanoparticles have also been tested for the sensing of H₂ gas (Su and Liao, 2016). Out of both types of sensors, the sensor developed using Pd nanoparticles modified with PPy exhibited superior sensing characteristics. Such modification led to the formation of more catalytically active sites for the H₂ gas molecules. The performances of PPy-based H₂ gas sensors are displayed in Table 2.

Electrospun NFs of poly(3,4-ethylenedioxythiophene) doped with (polystyrene sulfonic acid)-PEDOT-PSSA were synthesized to sense methanol, ethanol, and 2-propanol vapors (Pinto et al., 2011). Due to the large surface to volume ratio, the sensors had higher sensitivity and faster response times compared to other PEDOT-based alcohol sensors. The sensing mechanisms were related to polymer swelling that increased the inter-chain distance, which reduced the hopping rate. The response times for sensing of methanol, ethanol and 2-propanol were 8, 10 and 20 s, respectively, which are relatively fast for room temperature sensors. Zampetti et al. (2013) prepared PEDOT-PSS/TiO₂ NFs for NO₂ gas sensing (Fig. 8). The sensor detected 20 ppb NO₂ in the presence of 40 RH% at room temperature with a response (I_g/I_a, where I_a and I_g are the electrical current of the sensor in the presence of air (reference current) and NO₂ gas, respectively) of 1.045. The sensing mechanism was mainly related to the change in electrical resistance of the sensor at the interface between CP and TiO₂, where the NO₂ gas is able to easily alter the resistance of the formed p-n heterojunction. Some other PEDOT-based hybrid structures (e.g., Ti₃C₂T_x/PEDOT : PSS) are also explored for the sensing of methanol (Wang et al., 2019).

Fig. 8. 
Scanning electron microscopic images of (a) TiO₂ electrospun fibrous on IDE and (b) PEDOT-PSS coated TiO₂ fibers. E and g in Figure A represents electrode and gap between electrodes, respectively. Reprinted with permission from Zampetti et al.(2013).

The same group used electrospinning to fabricate sensing materials via a TiO₂ NF scaffold to support a CP (PEDOT: PSS), which resulted in a sensitive membrane (Zampetti et al., 2013). It showed a response (I_g/I_a) of 0.5 ppm NO₂ gas at ambient temperature. In the presence of 8 RH% it was 1.25. This outcome is related to the competitive nature of absorption/desorption between H₂O vapor and NO₂ gas, which decreased the available adsorption sites needed for effective adsorption of NO₂ gas. The other PEDOT-based materials (e.g., WO₃- PEDOT/PSS and PEDOT-graphene) have also tested successfully for the fabrication of NO₂ gas sensors (Dunst et al., 2017, 2016; Lin et al., 2015).