Introduction

Humidity sensors have gained vital significance in the food production, food storage, medical field, agriculture, museums, libraries, electronic industry and nuclear power plants due to the increase in demand for their use in monitoring the controlled environments.[1] Recently, humidity sensors fabricated from materials such as metal oxides, ceramics, thin layers of aluminium oxide films have attracted considerable attention in the field of sensor research.[2] But most of these sensors have encountered many limitations, such as complex device fabrication, high operating temperature, high-power consumption, and high cost.[3] To overcome these disadvantages, conducting polymer-based humidity sensors have emerged as a competitive class of alternating material with a number of advantages, such as ease of fabrication, hygroscopicity and ambient temperature operability.[4] In addition, the sensing dependent electrical and related mechanical properties of the conducting polymers can be tuned in various ways; including doping, production of blends and composites, and use of various preparation procedures.[5]

Among all the conducting polymers, polyaniline (PANI) has emerged as an excellent polymer due to its chemical expediency and ease of synthesis along with the wide range of attractive applications in sensors, EMI shielding, corrosion protection, supercapacitors, etc.[6] PANI alone has exhibited very less sensitivity to humidity due to its less hygroscopicity. To fabricate efficient humidity sensors, PANI has been composited with diverse materials like metals, semiconductors, non-metals, metal oxides, and carbon materials such as fullerene, carbon nanotubes, graphene, graphene oxide.[7] To further enhance the scope of PANI based composites, in terms of water adsorption capacity, improved sensitivity, response, and recovery times, limit of detection and linearity, in the present work we have chosen PANI and composited with neodymium oxide (Nd2O3). Since it is a rare earth oxide with a hexagonal crystal system, which has a high dielectric constant and insolubility in water and favorable band offset with semiconductors, it finds applications in many microelectronic devices such as sensors.[8]

In this perspective, PNO composites with different wt% of Nd2O3 in polyaniline were synthesized by employing the in situ chemical polymerization, and then morphologically and structurally characterized by FESEM, TEM, XRD, FTIR and Raman studies before humidity sensing studies.

Experimental

Materials

Aniline (C6H5NH2), Hydrochloric acid (HCl), Ammonium persulphate [(NH4)2S2O8], Neodymium oxide (Nd2O3) and de-ionized water all these analytical reagents (AR) grade chemicals were procured from SD fine chemicals, Mumbai, India.

Synthesis of PANI and PNO composites

PANI was prepared by in situ chemical polymerization technique. By adding 2 mL aniline with 0.1 mol L−1 to 100 mL HCl (8.58 mL of 1 M HCl added in 100 mL distilled water) and continuously subjected to magnetic stirring for 6 h to get a homogeneous mixture. To the reactant mixture, ammonium persulphate of 0.1 mol L−1 (2.28 g was dissolved in distilled water of 100 mL) was added to the mixture of reactants and left for polymerization. The product obtained after polymerization was separated and washed with 1N HCl and de-ionized water. Further, the dark-green product was kept in an oven for 24 h to obtain a dry powder. Under the similar conditions, PNO composites, namely PNO-1, PNO-2 and PNO-3 were prepared by adding crushed Nd2O3 powder of 10, 30 and 50 wt% with respect to PANI yield during the polymerization reaction respectively.

Characterization

Pestled samples of PANI, Nd2O3 and PNO composites were used to record the X-ray diffraction (XRD) spectra in the 2θ range 10–80° using D8 Advance Bruker X-ray diffractometer at a scan rate of 1.1° min−1 with Cu-Kα radiation as X-ray source of wavelength λ = 1.54 Å. To record the Fourier-transform infrared (FTIR) spectra, KBr pellets of the synthesized samples were prepared first, under a hydraulic pressure of 4–5 tons, and then pellets were introduced in to Perkin Elmer spectrum 65 instrument at a resolution of 4 cm−1. Field emission scanning electron microscopy (FESEM) was used to examine the surface morphology of the samples. FESEM images were obtained by Quanta 3D FEG and FEI Nova Nano SEM 600 instruments. JEOL-3010 instrument (fitted with a Gatan CCD camera) was used to capture the Transmission electron microscopy (TEM) image of PNO-3 composite. To capture TEM images, an accelerating voltage of 300 kV was applied. Raman studies were conducted using Horiba JobinYvon LabRam HR 800 spectrometer, employing Ar laser of power 20 mW and wavelength 514 nm in the wavenumber range 200 to 2000 cm−1.

Humidity sensing studies

An experiment to examine the ambient temperature humidity sensing characteristics was carried out by preparing pellets of PANI and the PNO composites with a thickness 1.5 mm and of diameter 10 mm applying a hydraulic press. Pellets were coated with silver on both sides to provide ohmic contacts. A hygrometer (Mextech-DT-615) was used to monitor the specific relative humidity, which was maintained in glass chambers by saturated solutions of various salts, viz. LiCl (11.3% RH), K(CH3COO) (22.7% RH), MgCl2 (33.1% RH), K2CO3 (43.2% RH), Mg(NO3)2 (53.0% RH), COCl2 (64.0% RH), NaCl (75.7% RH), KCl (85.1% RH), K2SO4 (97.6% RH). Electrodes were placed inside the glass chamber, which also holds the pellet through the rubber plugs fixed on the surface of the glass chambers. An Impedance analyzer (Hioki, Japan, Model: IM 7587) was connected to one end of the electrodes to measure impedance before and after exposure to a specific relative humidity. This meter was preferred to measure the change in impedance as a function of relative humidity for each sample at a chosen frequency of 100 Hz while maintaining a constant room temperature of 28°C in the RH range of 11 to 97%, avoiding polarization effects, if any. A particular frequency of 100 Hz was chosen, since these samples are known to exhibit the best linearity in impedance and a quick increase in capacitance over the tested RH range. The detailed description of humidity sensing measurements and the schematic diagram of experimental set up are given in our earlier papers.[9]

The formula to calculate the sensing response of the sample at ambient temperature is given by Eq. (1).

$${\text{SR = }}\left[ {\frac{{Z_{{{\text{LH}}}} - Z_{{{\text{HH}}}} }}{{Z_{{{\text{LH}}}} }}} \right] \times 100$$
(1)

where SR—sensing response, ZLH—impedance at lower % RH, ZHH—impedance at various higher % RH. Further, the sensing performance of the composites has been evaluated by the method of least squares regression.[13] Then,

Limit of detection (LOD)[10] is computed using Eq. (2),

$${\text{LOD}} = 3.3 \times \left[ {\frac{{\text{Standard deviation of the regression line}}}{b}} \right]$$
(2)

where b—slope of the least square fitted line, obtained by the graph impedance versus %RH, %Hysteresis in the humidity sensing behavior is calculated using Eq. (3)

$${\text{\% Hysteresis = }}\left[ {\frac{{Z_{{{\text{mn}}}} - Z_{mp} }}{{Z_{{{\text{max}}}} - Z_{\min } }}} \right] \times 100$$
(3)

where Zmn and Zmp are the maximum and minimum values of impedances at the mean value of %RH in the hysteresis humidity graph. Zmax and Zmin are the maximum and minimum values of measured impedances in the range 11–97% RH.

Results and discussion

XRD studies

Figure S1 shows the XRD patterns of PANI, Nd2O3 and PNO-3 composite with identified hkl values. XRD spectrum of PANI [Fig.S1(a)] has exhibited two major peaks at 2θ = 21.32° and 25.49°, which corresponds to (100) and (111) planes respectively, and can be correlated to the periodicity of lattice, perpendicular and parallel to the polymer chains as per the literature.[9] Fig. S1(b) shows the XRD pattern of Nd2O3 for different values of 2θ, related to the crystal planes (110), (200), (211), (220), (310) and (321) respectively, which confirm its cubic crystal structure. The pattern matches well with the JCPDS file No. 83–1356.[11] XRD pattern of PNO-3 composite is shown in Fig. S1(c), has all the Nd2O3 peaks but with a reduced intensity, because Nd2O3 particles are embedded within the polymer matrix.

In addition, the XRD pattern of the composite (PNO-3) show that the distinctive peaks of Nd2O3 are retained with minor changes, whereas those of PANI have disappeared, showing that the crystalline structure of Nd2O3 in the composite remain unaffected.[12] Crystallite size of the composite computed using Debye–Scherrer formula was found to be 27.92 nm and that of Nd2O3 was 30.74 nm. As the crystallite size has reduced, increase in the surface to volume ratio, which may lead to the enhanced the water adsorption capacity in the composite. Another noteworthy finding is that the intense peak of Nd2O3 in the composite has shown suppressed intensity, implying that the presence of metal oxide in the composites might have enhanced surface active sites for water adsorption.[13]

FTIR characterization

FTIR spectra of PANI, Nd2O3 and PNO-3 composite are depicted in Fig. S2. FTIR spectrum of PANI (Fig. S2(a)) reveals its characteristics absorption bands at 507, 862, 1150, 1307, 1488, 1582, 2320 and 3233 cm−1.[14] The absorption band at 507 cm−1 is due to the coupling of Cl− ions with bending vibrations of PANI.[15] Bands at 862, 1150 and 1307 cm−1 correspond to C–H in plane stretching. Bands at 1488 cm−1 represent C=C stretching in cyclic benzenoid and 1582 cm−1 related to C=C stretching in the quinoid rings respectively. FTIR spectrum of Nd2O3 (Fig. S2(b)) shows its intense absorption band at 666 cm−1, represent the characteristic stretching vibration of metal and oxygen (Nd-O).[15] The presence of bands of both Nd2O3 and PANI bands in the composites [Fig. S2(c)] indicates that the conjugated structure of PANI chains interacts with Nd2O3. Moreover, the existence of shifted bands of PANI and Nd2O3 in the composite spectrum with greater depth confirms their interfacial contact. And this contact could be due to the formation of hydrogen bonds between N–H of PANI and the oxygen of Nd2O3.

FESEM and TEM studies

FESEM images of PANI, Nd2O3 and PNO-3 composite are depicted in Fig. 1. FESEM image of PANI shows clusters of agglomerated globular structures [Fig. 1(a)]. This kind of morphology is common in conducting polymers.[16] In the FESEM image of Nd2O3, uniformly distributed, densely packed clusters of grains can be observed [Fig. 1(b)]. FESEM image of PNO-3 composite [Fig. 1(c)] shows that, clusters of PANI granules deposited on Nd2O3 with pores. Such morphology readily facilitates water adsorption. Average grain size of the composite was calculated using image-J software by employing normal distribution [Fig. 1(d)]. Average grain size of the composite particles was found to be 0.35 µm, these grain clusters helpful for diffusion of water molecules through pores, which in turn enhances the humidity sensing performance of the composite.[12]

Figure 1
figure 1

FESEM micrographs of (a) PANI, (b) Nd2O3 and (c) PNO-3 composite and (d) histograms of size distribution of the PNO-3 composite particles (e) TEM image of PNO-3 composite and (f) SAED pattern PNO-3 composite.

Full size image

TEM image of PNO-3 composite is shown in Fig. 1(e). The image clearly illustrates how Nd2O3 particles are uniformly incorporated in to the PANI matrix.[17] Sizes of such composite particles were found to be in the range of 90–100 nm. SAED (selected area electron diffraction) pattern of the PNO-3 composite is shown in Fig. 1(f), which shows bright spots arising from Bragg reflection from a single crystallite, indicating the crystalline nature of the composite. Such crystalline composites helpful for fabrication of humidity sensors reported recently.[18]

Raman studies

Raman spectra of PANI, Nd2O3 and PNDO-50% is shown in Fig. S3. Raman spectrum of PANI as in Fig. S3(a), exhibited its significant peaks at 1126, 1351 and 1577 cm−1 which can be attributed to in-plane bending of C–H, stretching vibrations of C–N• + and stretching vibrations in quinoid rings of C=C respectively. Raman spectrum of Nd2O3 is depicted in Fig. S3(b)has shown four phonon related bands of (2A2u + 2Eu) at room temperature. Raman peaks of the oxide are found at 295, 352, 401, 466 cm−1. Among these signature peaks of Nd2O3, the prominent peak can be seen at 352 cm−1. Also, peak at 466 cm−1, can be assigned to A2u mode. Apart from these peaks two shoulders can be seen at 568 and 773 cm−1. Raman spectrum of the composite PNO-3, as in Fig. S3(c)exhibited the peaks of both PANI and Nd2O3, interestingly, the prominent peak 352 cm−1 of Nd2O3, has been shifted to 325 cm−1, indicating a strong interaction between PANI and Nd2O3. The peak shift towards the smaller wavenumber indicating the formation of the composite has exerted a tensile stress on the crystalline Nd2O3, which in turn has increased its bond length. This structural deformation in the crystalline composite has facilitated favorably for the water adsorption and there by contributing to the humidity sensing.[19]

Humidity sensing studies

Variation in impedance of PANI and PNO composites in response to a vivid environment in the RH range 11–97% is depicted in Fig. S4. Impedance value decreases with increase in %RH for PANI and PNO composites as shown in Fig. S4. Decrease in the impedance of the composites with an increase in %RH exhibited a linear behavior, and therefore for the three composites, after applying the least square method, the value of linear correlation R2 was calculated and tabulated (Table I) and the related plot is shown in Fig. 2(a). The point sensitivity, real sensitivity, limit of detection (LOD), and humidity hysteresis are the other sensing parameters computed using the Eqs. (1, 2 and 3) for PANI and the PNO composites and are listed in Table I.[9] Noteworthy observation from Table I, is that, for PNO-3 composite, point sensitivity of 91% at 97%RH, real sensitivity of 6.37 MΩ/% RH with a LOD of 4%RH, and the linearity R2 of 0.998 have excelled over other composites. These favorable parameters of PNO-3 composite are deciding factors in the design of a competitive humidity sensor.

Table I Point sensitivity, real sensitivity, limit of detection and linearity of PNO composites.
Full size table

Timing behavior

Timing behavior is an important aspect in the design of any sensing material. Figure 2(b) shows the characteristic response and recovery curve of the PNO-3 composite. Response time and recovery time characteristics are very important parameters for evaluating the performance of any sensor. Initially, the sample was kept in a chamber containing 11% RH and then transferred to a chamber with a 97% RH, with a switching time of one second, which recorded a response time of 28 s. A recovery time of 29 s was recorded when the sample was brought back from the chamber with 97%RH to the 11%RH. The response and recovery times are almost same indicating that the exothermic and endothermic processes occur at the same rate.[23] Good response and recovery times of the composite may be due to the presence of smaller grain size as evidenced by the FESEM image. Therefore, this composite has best humidity sensing properties, making it a truly competent humidity sensing material. The response-recovery times for the various composites are listed in Table I.

Figure 2
figure 2

(a) Least square fits of the impedance of each composite for varied humid environments, (b) Response-recovery curves of PNO-3 composite, (c) Humidity hysteresis characteristic curves of PNO-3 composite, (d) Humidity sensing stability of PNO-3 composite.

Full size image

Humidity hysteresis and stability

Figure 2(c) shows the humidity hysteresis curve for PNO-3 composite. Impedance values were recorded by subjecting the composite sample at first in increasing and later in decreasing order of RH between 11 and 97%, the humidity hysteresis loss of PNO-3 composite was evaluated experimentally. In the ascending order of humid environment, as prevailed in different humidity chambers of the experimental set up, the adsorption of water molecules was detected with a simultaneous decrease in the impedance of the composite. This impedance change, exhibited a linear behavior with an increase in %RH. Whereas, during the descending order of %RH environments, it is interesting that the impedance variation with decreasing %RH, followed a slightly different path as shown in Fig. 2(c), which indicates a desorption process.[24] Also, at an optimum value of 55% RH, the maximum difference in humidification and dehumidification was observed. In this humid environment, hysteresis loss was calculated to be 2%.

Humidity sensing stability test for PNO-3 composite was performed once in every five days for one month as shown in Fig. 2(d). The observed sensing responses in two different relative humidity environments particularly at 55%RH and 97%RH were found to be nearly stable, implying that the composite acts as a stable humidity sensor.

Humidity sensing mechanism

The three steps of the humidity sensing mechanism are chemisorption, physisorption and capillary condensation. At the lower RH level, chemical adsorption of water molecules occurs on active sites on the surface of the sensing material, which then dissociates into H+ and OH− ions due to self-ionization.[25] When PNO-3 composite was exposed to ambient humidity, the water vapour self-ionizes to form proton H+ and hydroxyl ions OH− as depicted in Eq. (4). Because metal cations have a high density of charge carriers with a strong electric field, these ions cause protonation and condensation on the surface of PNO-3 composite.

$${\text{H}}^{ + } + {\text{ OH}}^{ - } \; \rightleftharpoons \; {\text{H}}_{{2}} {\text{O}}$$
(4)

With increase in the relative humidity, a chemisorbed layer of ions is formed due to the chemical adsorption of water vapour to the active sites of the composite surface. At the same time, the dissociated hydroxyl ions bind to the highly charged surfaces of the metal cations and the released H+ ions flow freely on the surface of the composite, thereby slightly increasing its conductivity. As this process proceeds, with the increase in relative humidity, two OH− ions form hydrogen bonds with a water molecule, generating H3O+ and OH− ions, as shown in Eq. (5) and the resulting layer is known as a physisorbed layer and this layer becomes stationary due to hydrogen bonding.[26]

$${\text{2H}}_{{2}} {\text{O}} \; \to \;{\text{H}}_{{3}} {\text{O}}^{ + } + {\text{ OH}}^{ - }$$
(5)

As %RH further increases, huge water molecules accumulate on the top of the already formed physisorbed layers. These physisorbed layers would pile up on the top of each other; creating an extremely unstable mass of H3O+. As a result, it rapidly splits into a water molecule and a proton, as shown in Eq. (6) and shown in Fig. 3(b). This process is called Grotthus reaction.[27]

$${\text{H}}_{{3}} {\text{O}}^{ + } \; \to \;{\text{H}}_{{2}} {\text{O }} + {\text{ H}}^{ + }$$
(6)
Figure 3
figure 3

(a) Schematic diagram of sensing mechanism, (b) Grotthus reaction.

Full size image

Later, when the humidity is extremely high, more water molecules are adsorbed on the surface of the composite, eventually leading to the formation of poly-physisorbed layers that release many protons. These charged particles migrate from one water molecule to another, simultaneously making and breaking bonds with nearby molecules, releasing a huge number of protons leading through various water molecules, resulting in a drastic reduction in impedance.[21] As a result, in a particularly humid environment, protons are the main charge carriers. At the end of the process water molecules begin to condense in the capillary pores formed due to the reduced grain size as evident from XRD, which lead to the electrolytic conduction, which in turn resulted in further decrease in impedance. Figure 3(a) depicts the predicted condensation mechanism.

Overall, the PNO-3 composite, exhibited a good sensitivity, fast timing behavior, minimal humidity hysteresis and good humidity sensing stability as compared to the PANI and other composites at room temperature. All these features are worth considering, when designing a reliable humidity sensor.

Conclusion

PANI composites using rare earth neodymium oxide were prepared by in situ chemical polymerization technique by changing the composition of the oxide in the PANI. Composites were investigated for their humidity sensing behavior. Of all the composite samples, PNO-3 composite prevailed with appreciable response, recovery times of 28 and 29 s, minimal humidity hysteresis, in addition to stability in sensing for a period of one month. From these obtained results, it follows that the PNO composite is an effective material in the fabrication of the humidity sensor that can be operated at the ambient temperature.