Benzene (AR grade) was obtained from Thai Aromatic Co., Ltd and was freshly distilled before use as the monomer. Aluminium chloride and cupric chloride (AR grades, Riedel-de Haen) were used as the oxidant and the catalyst for the synthesis of poly(p-phenylene), respectively. Hydrochloric acid (AR grade, JT Baker) was used as the washing solution. Ferric chloride (AR grade, Fisher Scientific) and ethanol (99.5%, Carlo Erba) were used in the doping of PPP. ZSM-5 zeolite (SiO₂/Al₂O₃: 23) in powder form was purchased from Zeolyst International. Sodium chloride, potassium chloride and lithium chloride, (AR grade, Ajax Finechem), were used in the cation exchange processes. Nitrogen (99%, TIG) and ammonia gas (99.99%, Poontaweporn Limited Partnership) carbon monoxide (0.1%, TIG) and hydrogen (99.999%, TIG) were used to investigate the electrical conductivity responses of the composites.

Poly(p-phenylene) was chemically synthesized via the oxidative polymerization method described elsewhere [21,22]. The reaction between benzene, anhydrous aluminum chloride, and anhydrous cupric chloride at a mole ratio of 8:2:1 was carried out at 32–37 °C under a nitrogen atmosphere for 3 h. The mixture was subsequently cooled and added to ice-cold 18% hydrochloric acid. After filtration, the PPP particles were washed several times with boiling acid solution and with boiling distilled water and then dried at 110 °C for 3 hr.

The doping process of PPP with FeCl₃ in our work was adapted from Shiga et al. [23] and performed by suspending the PPP particles in 100 mL of FeCl₃-ethanol solutions at 60 °C in a 50:1 mole ratio between the dopant and PPP monomer. After the mixture was filtered, the doped PPP was dried at ambient temperature under vacuum for 12 h. The obtained doped PPP is coded as 50:1 dPPP and the undoped PPP as uPPP.

The starting ZSM-5 material with SiO₂/Al₂O₃ ratios of 23 was in an ammonium form (NH₄ZSM-5). The cation exchange process was carried out by using the conventional method [24]: a mixture of 1 g zeolite per 100 mL of 0.3 M NaCl solution was stirred at 90 °C for 1 hr, then filtered, and washed with distilled water several times. After the cation-exchanged zeolite was dried, the zeolite was calcined at 550 °C for 3 h with a heating rate of 1 °C/min.

PPP powder was ground, sieved with 53 μm sieve, and then dried prior to being mechanically mixed with dried zeolite powder at various zeolite amounts: 10%, 20% 30% and 40% v/v, in order to investigate the effect of zeolite content. The dry mixed composites were subsequently pressed into pellets with a diameter of 15 mm and a nominal thickness of 0.4 mm, using a hydraulic press machine at a pressure of ∼5 kN.

A Fourier transform infrared spectrometer (FTIR Nicolet, Nexus 670) was used to identify the PPP's functional groups and its interaction with CO. PPP powder or PPP/zeolite powder were mixed with optical grade KBr (Carlo Erba Reagent) and compressed into a disc shape. The spectrometer was operated in the transmission mode averaging 20 scans at a resolution of 4 cm⁻¹, covering a wave number range of 400–4,000 cm⁻¹. The thermal stability of undoped and 50:1 doped PPP was investigated by using a thermogravimetric analyzer (Dupont, TGA 2950). An X-ray diffractometer (XRD Phillips, Rigaku) was used to examine the crystallinity of PPP and the crystal order of the zeolite. A scanning electron microscope (SEM JEOL, JSM 5200) was used to observe the morphology of sample materials. The surface areas and the pore volumes of ZSM-5 zeolites were measured by using a surface area analyzer (Quantachrome, Autosorb-1). An atomic absorption analyzer, AAA, (Varian, Spectre AA 300) was used to determine the exact amount of silicon-aluminium containing in the zeolite and the contents of the exchanged cations in terms of cation exchange level. The NH₃-TPD thermogram (TPDRO, 1100) of the zeolite was measured to examine the NH₃ desorption.

The bulk electrical conductivity of PPP pellets under exposure to air, N₂ and NH₃ were measured by using a custom made two-point probe which was connected to a voltage supplier (Keithley, 6517A) in which its voltage was varied and the current was measured in the linear Ohmic regime. Therefore, the electrical conductivity can be calculated from the equation: σ = (I/KVt), where I is the measured current (A), V is the applied voltage (V), t is the thickness, and K is the geometric correction factor of the two-point probe which can be determined by calibrating the probe with a silicon wafer possessing a known resistivity value. The electrical conductivity response and sensitivity of the composites were determined by following the equations: Δσ = σ_NH3 − σ_{N2 intial} and Δσ/σ_{N2 intial}, respectively. The two-point probe was located in a gas chamber connected upstream with a mixing gas chamber. NH₃-N₂ mixture of 20%v/v was initially injected into the mixing chamber at the pressure of 1.1 atm. Then it was diluted by injecting with an equal volume of N₂ to obtain the pressure of 2.2 atm. Half of the mixture was then transferred to the measurement chamber, which now contains 10% v/v NH₃-N₂ at a pressure of 1.1 atm. Other mixtures with lower NH₃ concentrations were obtained by successive dilutions with N₂ in the mixing gas chamber. In all measurements, the gas chamber temperature was maintained at 27 ± 1 °C and the gas pressure was 1.1 atm to prevent external air from leaking in.

The infrared spectrum of the synthesized undoped PPP (uPPP) shows a very intense band at 805 cm⁻¹ and moderate intensity bands at 999, 1,396, and 1,479 cm⁻¹, corresponding to the para aromatic substitution. Peaks at 763 and 695 cm⁻¹ can be assigned to the mono substitution [21,25]. For 50:1 doped PPP, additional adsorption peaks appear at 1,545 and 1,180 cm⁻¹, due to the intrinsic vibration of the polymer chain in the doped state [26-28].

UV-Vis absorption spectra of undoped poly(p-phenylene) show a reflectance peak at 350 nm corresponding to the π - π* transition of the benzenoid ring [29]. After doping, the reflectance peak of the π - π* transition of the benzenoid ring shifts to 300–400 nm [29].

Thermograms under air of both uPPP and 50:1 dPPP display single step decompositions at 569 °C and 480 °C, respectively. This demonstrates that uPPP is both thermally and thermooxidatively stable [21]. dPPP is less thermally stabile since the doping induces the defects in the polymer chain; yet our synthesized dPPP can still withstand heat up to 400 °C.

A XRD pattern of uPPP shows a high degree of crystallinity accompanied by d-spacing values of 4.53, 3.63, and 3.24 Å. The most intense d-spacing is at 4.53 Å, corresponding to the length of the phenyl unit [21,30]. This suggests that the rings are very nearly coplanar. This sharp peak representing crystallinity decreases upon further doping. The mean particle size and the density of dPPP are 33.61 ± 0.18 μm and 1.3273 ± 0.0011 g/cm³, respectively.

The mean particle size of ZSM-5 from the particle size analysis is 5.46 ± 0.01 μm, comparable to that obtained from the SEM image. The density of the zeolite is 1.9739 ± 0.0004 g/cm³. From the atomic absorption spectroscopy data (AAS), the numbers of moles of Si and Al in ZSM-5 can be calculated and reported in terms of Si/Al ratio which is equal to 12.67. The amounts of Na⁺ and K⁺ exchanged in the zeolites are tabulated in Table 1.

The specific surface areas of ZSM-5 containing various cation types (Na⁺, K⁺, NH₄⁺ and H⁺) are shown in Table 2. Thus, the surface areas and the pore volumes of the zeolites vary according to the cation size.

The NH₃-TPD profiles in Figure 1 indicate the acidic properties of the zeolites with different cations. The acidic activity follows this order: H⁺ > Na⁺ > K⁺, consistent with that of previous work [31].

Umar et al., studied the effect of cation exchange to the acid sites, the activity, and the selectivity of a dealuminated faujasite. The acidic properties were determined by NH₃/TPD measurements. The deactivating effect of individual cations was a strong function of size. The acidic activity follows the order H⁺ ∼ Mg²⁺ ∼ Li⁺ > Na⁺ > K⁺ [31].

Figure 2a-f shows SEM images of 50:1dPPP, ZSM-5(23), and 50:1dPPP/ZSM-5(23) composites of various zeolite amounts. ZSM-5 zeolite particles possess irregular crystal shapes and appear to be inhomogeneously dispersed in the conductive polymer matrix.

The specific electrical conductivity measurements of uPPP, 50:1 dPPP, and its composites under air and N₂ atmospheres were carried out at 28 ± 1 °C and at 1 atm. The specific electrical conductivity of uPPP in air is (1.17 ± 0.07) × 10⁻⁵ S/cm, as shown in Figure 3. Doping uPPP increases the electrical conductivity in air by several orders of magnitude, up to 0.87 ± 0.42 S/cm, as shown in this figure.

The increase in electrical conductivity is due to the oxidation of the π-conjugated system of the conductive polymer. Doping with FeCl₃ causes the increase in electrical conductivity, and it converts the brown insulating PPP powder into black conductive PPP. The counter ion or the dopant species of dPPP FeCl₄⁻ occurs from the reduction of Fe³⁺ to Fe²⁺ and the oxidation of the π-system of PPP [32]. After introducing the zeolite into the polymer matrix, the electrical conductivity in air decreases to 0.22 ± 0.13 S/cm (Figure 3) with increasing zeolite content from 0% to 40%, because of the reduction in the number of polarons. Under a N₂ atmosphere, it can be seen from Figure 3 that the electrical conductivity is less than that in air at any zeolite content. This is due to the presence of moisture in air which can induce an increase in electrical conductivity.

The electrical conductivity response (Δσ= σ_gas − σ_N2) is identified as the difference in the steady state electrical conductivities of the sample under target gas and under N₂ exposure at 28 ± 1 °C and at 1 atm. Because various composites possessing different initial electrical conductivity values, the sensitivity (Δσ/σ_N2), defined as the ratio of the electrical conductivity response and the electrical conductivity under pure N₂ exposure, will be used in comparing responses of various composites. The electrical conductivity sensitivities towards the target gases (CO, H₂, and 10% NH₃) are investigated and reported below.

Under CO atmosphere, dPPP shows a very small positive response with a sensitivity of 2.1 × 10⁻², as tabulated and shown in Table 3. CO behaves like an electrophilic gas, CO molecules are expected to withdraw electrons from dPPP, the p-type doped conductive polymer consisting of polarons and bipolarons, causing an increase in electrical conductivity [33,34]. Unlike the CO-polyaniline interaction, the CO-dPPP interaction is relatively weak with a much lower sensitivity. Therefore, it can be summarized there is no apparent interaction between dPPP and CO; if there is any interaction it appears to be very weak.

Yamamoto and Gu studied the effect of ammonia gas on the electrical property of a conducting polymer. The electrical conductivity of the PPP film increased by several orders of magnitude when exposed to NH₃. The change in electrical conductivity took place relatively fast and was reversible. When the partial pressure of NH₃ was reduced through evacuating the sample chamber, the electrical conductivity of the PPP film decreased in a similar pattern. NH₃ acted as an effective doping agent in n-doped PPP. Similar results were obtained when PPP films were exposed to diethylamine and triethylamine [35].

The sensitivity value of dPPP when exposed to H₂ is even smaller, with a positive value of 0.5 × 10⁻², as shown in Table 3. H₂ is a reducing gas [36] and it is weakly nucleophilic; thus, when the sample is exposed to H₂, a decrease in the electrical conductivity may be expected. In a certain case, H₂ may provide an electrical conductivity enhancement [37]. Our data in Table 3 clearly suggest that dPPP does not interact with H₂, due to the weak electron transfer between H₂ and dPPP.

Figure 4(a) shows the specific electrical conductivity of dPPP(60)/NaZ23 vs. time (min) when exposed to 5%v of NH₃ at 28 ± 1 °C and at 1 atm. Under the exposure to NH₃, the specific electrical conductivity of dPPP(60)/NaZ23 first increases and reaches a maximum value of 0.235 S/cm after a NH₃ exposure time of 2.90 min, then decreases to a steady state value of 0.03 S/cm after the NH₃ exposure time of 30 min.

Figure 4(b) shows the specific electrical conductivity of dPPP(60)/NaZ23 vs. time (min) after evacuating 5%v NH₃ and exposure to N₂. The specific electrical conductivity increases from a value of 0.01 S/cm to a steady state value of 0.03 S/cm after the N₂ exposure time of 56 min.

Figure 5 shows the electrical conductivity sensitivity of 50:1 dPPP vs. NH₃ concentration at 28 ± 1 °C and at 1 atm. The negative sensitivity of 50:1dPPP increases linearly with increasing NH₃ concentration.

The doping process and the resultant formations of polarons and biporalons induce dPPP electrons to be in a higher energy level having lesser stability [37,38]. When dPPP is treated with an electron rich molecule, such as NH₃, an electron transfer interaction occurs. Since dPPP is a p-type doped conductive polymer and NH₃ is a strong nucleophilic gas [12], NH₃ tends to give electrons to dPPP causing a decrease in the number of charge carriers, polarons and bipolarons, and the decrease in electrical conductivity; therefore, the negative response is expected and observed. A possible 50:1dPPP-NH₃ interaction is proposed in Figure 6. Here, a NH₃ molecule attaches itself to the polaron site through the electrostatic interaction, thus reducing the charge carrier mobility.

The electrical conductivity sensitivities of 50:1dPPP/ZSM-5(23) composites at various NH₃ concentrations versus ZSM-5 zeolite content are shown in Figure 7.

The sensitivities of 50:1dPPP at 0.625, 1.25, and 5%v NH₃ exposures increase from −0.140 ± 0.11 to −0.51 ± 0.02, from −0.16 ± 0.02 to −0.79 ± 0.01, and from −0.341 ± 0.03 to −0.88 ± 0.01, respectively, as NaZSM-5(23) content increases from 0 to 30%v. This is because a composite with a higher zeolite content allows more NH₃ molecules to adsorb and to diffuse deeper into the composites, thus promoting the NH₃-dPPP interaction [35]. Beyond this volume fraction of 40%v zeolite, the sensitivities decrease to −0.32 ± 0.13, −0.55 ± 0.01 and −0.73 ± 0.01 for the 0.625, 1.25 and 5%v NH₃ exposures, respectively. In the present case of mechanically mixed systems, the decrease in sensitivity with increasing zeolite content is due to the reduction in the number of available active sites on dPPP for NH₃ to interact as zeolite content increases.

We investigate next on the influence of the cation type on the sensitivity of the conductive polymer/zeolite composites. Four cations were chosen: (1) Na⁺; (2) K⁺, both cations are very common and we obtain 100% exchange by the traditional cation exchange process; (3) NH₄⁺, which is the original cation in the zeolite, and (4) H⁺, this form of ZSM-5(23) is accomplished by the decomposition of the ammonium form. To avoid the effect of zeolite content which may be predominant over the effect of cation type, we fixed the zeolite volume fraction to be at 10%v. 50:1 dPPP composites, containing 10%v of ZSM-5(23) having four different cations Na⁺, K⁺, NH₄⁺, and H⁺, are coded as 50:1 dPPP(90)/NaZ23, 50:1 dPPP(90)/KZ23, 50:1 dPPP(90)/NH₄Z23 and 50:1 dPPP(90)/HZ23, respectively. Figure 8 shows the electrical conductivity sensitivities of 50:1 dPPP, 50:1 dPPP(90)/KZ23, 50:1 dPPP(90)/NaZ23, 50:1 dPPP(90)/NH₄Z23 and 50:1 dPPP(90)/HZ23; they are −0.14 ± 0.01, −0.13 ± 0.005, −0.22 ± 0.003, −0.36 ± 0.016 and −0.36 ± 0.62, respectively.

Since Na⁺ is a common cation, it is used as a basis for the comparison. The sensitivity of composite is remarkably increased when the cation is replaced by either H⁺ or NH₄⁺ as the balancing cations in the zeolite. This is due to the adsorption property of NH₃ onto ZSM-5. The sensitivity of the composites apparently increases with the acidity of the zeolite, shown by the NH₃-TPD profiles (Figure 1), the surface area and the pore volume (Table 2). H-ZSM-5 with the highest acidity, the surface area, and the pore volume, providing a more favorable NH₃ adsorption, resulting in the highest sensitivity. NH₄-ZSM-5 has nearly the same sensitivity value to that of H-ZSM-5, with a lower acidity, surface area and pore volume. On the other hand, the addition of K-ZSM-5 into the dPPP matrix does not improve the sensitivity, instead it causes a slight sensitivity reduction. NH₃ does not prefer to adsorb on K-ZSM-5 due to its lower acidity, surface area, and pore volume. The addition of the zeolite merely serves to decrease the number of dPPP active sites available. Hence, the KZSM-5 composite has an even lesser sensitivity than that of pure dPPP.

The interaction of NH₃ and the active sites of dPPP, NaZSM-5(23) and their composites are investigated via FTIR spectroscopy under pressure 1 atm and at 28 ± 1 °C. The FTIR spectra of 50:1 dPPP before, during, and after the NH₃ exposure are shown in Figure 9.

Under NH₃ exposure, FTIR spectra were taken every five minutes for a duration of 60 min; each FTIR spectrum has the same patterns but with different intensities evolving in time. Peaks at 929.4, 964.3, and 1,624.5 cm⁻¹ belong to the vibrations of NH₃ molecules, and a peak at 1,398.5 cm⁻¹ can be assigned to the vibration of NH₄⁺ molecules [40] or the para aromatic substitution of dPPP [21]. The first bands are slightly different from those of free NH₃ (931.58, 968.08, 1,627.5 cm⁻¹). The last band is diatinct from that of the free NH₄⁺ (1,397 cm⁻¹) [40] or the para-aromatic substitution of dPPP band (1,396 cm⁻¹) [21]. The presence of the bands for NH₃ and NH₄⁺ suggest both of the C⁺…NH₃ interaction (Figure 5) and the FeCl₄⁻…NH₄⁺ interaction taking place. After replacing NH₃ with N₂, the bands at 929.4, 964.3, 1,398.5 cm⁻¹ still remain in the dPPP/NH₃ FTIR spectrum. The ammonia-dPPP interactions are thus irreversible, consistent with the irreversible conductivity response observed when replacing NH₃ with N₂.

Figure 10 shows the FTIR spectra of NaZSM-5 before, during, and after the NH₃ exposure. The bands at 916.0 and 950 cm⁻¹ appear after the NH₃ exposure; and they disappear when NH₃ is replaced by N₂. These peaks are due to the NH₃ vibrations, with shifts in the frequency from those of the free NH₃ (931.58 and 968.09 cm⁻¹), representing with the Lewis sites [40]. There is no significant difference in the band patterns before and after the exposure to NH₃; thus, the interaction between NH₃ and NaZSM-5 is reversible based on the FTIR spectrum.