We demonstrate an ammonia sensor composed of a tapered multimode fiber coated with polyaniline nanofibers that operates at room temperature (26°C). The optical properties of the polyaniline layer changes when it is exposed to ammonia, leading to a change in the absorption of evanescent field. The fiber sensor was tested by exposing it to ammonia at different concentrations and the absorbance is measured using a spectrophotometer system. Measured response and recovery times are about 2.27 minutes and 9.73 minutes, respectively. The sensor sensitivity can be controlled by adjusting the tapered fiber diameter and the highest sensitivity is achieved when the diameter is reduced to 20 µm.
© 2015 Optical Society of America
Ammonia is an important chemical that is used in many industrial applications such as fertilizing agent in agriculture, refrigerant in refrigerator industries and food processing, precursor for nitrogenous compound in chemical industries as well as household cleaning agents. Exposure to high concentration of ammonia can cause lung irritation and pulmonary edema . Therefore, it is essential to monitor ammonia concentrations especially in a manufacturing plant that uses highly concentrated ammonia gas.
Polyaniline (PANI) is a conducting polymer, which has been proposed as a sensing layer in gas sensors [2–4]. The major advantage of conducting polymer-based sensors as compared to inorganic materials such as metal oxide based sensors is their sensitivity at room temperature. The ease in synthesis and modification of the molecular chain adds to the sensor’s advantage. PANI is an excellent candidate for ammonia sensor because of its variation in electrical conductivity and optical absorption when they are exposed to ammonia. The properties change with the state of oxidation and protonation of the polymer. When the acid form (emeraldine salt) PANI is exposed to ammonia, it will be deprotonated and becomes non-conducting (emeraldine base) PANI .
Electrical-based ammonia sensors employing PANI are well-established as reported in [6–9]. On the other hand, optical-based sensors are still new and there is more room for further investigation. As compared to electrical sensors, optical sensors offer several advantages such as immunity to electromagnetic interference, electrically passive and safer, smaller size, nondestructive and environmental ruggedness [10–12]. Optical-based ammonia sensors can be developed by depositing a thin PANI film on various optical transducing platforms such as a glass substrate , a planar waveguide  and optical fibers [15, 16]. Optical fiber sensors promote an extra feature of remote monitoring possibility and more practical as compared to the other transducing platforms. Recently, evanescence-based fiber sensors have attracted great attention due to its simple fabrication method, which involves either cladding removal or tapering of the conventional optical fiber. In [17, 18], ammonia sensors were developed by depositing PANI on the exposed core region after the fiber cladding is removed through chemical etching. The drawbacks of the chemical etching method are difficulty in controlling the etching process and low repeatability.
Advancement in fiber tapering method enables controllable process to produce desired fiber taper with specific geometrical profiles. Reproducibility of the tapers is important because the performance of tapered fiber sensor is highly dependent on its geometry; specifically the waist diameter and length [19, 20]. The smooth transition between the original fiber and the tapered waist makes the tapered fiber less fragile as compared to cladding-removed fiber. In , a fast hydrogen detector made of 1.3 µm diameter tapered single mode fiber (SMF) coated with 4 nm of palladium is reported. The detection is based on the attenuation of the evanescent field as the palladium becomes palladium hydride when exposed to hydrogen. A fast humidity detector made of 680 nm diameter fiber taper coated with gelatin layer has also been reported . The diameter is made extremely small in order to produce a fast sensor. However, the challenge in using such small taper will be in terms of handling and also its high susceptibility to environmental perturbations. In , they have shown that lossy mode resonance can be detected by using tapered SMF with waist diameter of 20 µm - 40 µm coated with very thin polymeric materials. A tapered SMF with waist diameter of 48 µm, coated with gold nanoparticles has shown a response based on surface plasmon resonance excitation .
Here, we report an ammonia sensor which consists of a tapered multimode optical fiber (MMF) coated with thin layer of PANI nanofibers. Since MMF has larger core size as compared to SMF, the diameter of the taper does not have to be very small to ensure large fraction of the evanescent field to be present at the taper surface. In this paper, we show that tapered MMFs with diameter size of 20 µm - 40 µm coated with PANI nanofiber produce good absorbance response to ammonia at room temperature.
2.1 Preparation of the tapered multimode fiber
The tapered fiber was fabricated from standard graded-index MMF using a Vytran GPX-3400 optical glass fiber processor. The Vytran machine uses a filament heater made of graphite to heat up a bare MMF while both sides of the fiber is pulled simultaneously. The tapering process is fully automated and the desired taper profiles are specified on a graphical user interface. In this work, MMF with cladding/core size of 125 µm/62.5 µm was tapered to produce biconical tapers with waist diameters of 20 µm, 30 µm and 40 µm. The waist length and transition region length were fixed at 10 mm and 2 mm, respectively. The tapered MMF was examined using Hitachi S-3400N Scanning Electron Microscope (SEM) as shown in Fig. 1.
2.2 Preparation of the tapered fiber ammonia sensor
Emeraldine base PANI nanofiber was synthesized through a chemical polymerization process based on the method in . The PANI was then re-doped with camphorsulfonic acid (CSA) and dispersed in chloroform to produce CSA-doped PANI solution with concentration of 3.75 mg/ml. The doping of PANI with CSA is essential to increase the solubility of the PANI in chloroform. The solution was stirred using magnetic stirrer at room temperature for 1 hour and then sonicated for another 1 hour. The solution was then sprayed using airbrush onto the tapered fibers and microscope glass slide that were heated at 50°C on a hot plate. The heat is applied to improve adhesivity and ensure homogeneous PANI distribution. The samples on the microscope glass slide were used for material micro-characterization purposes. The tapered MMF coated with PANI was left to dry at room temperature for 1 hour. The PANI deposited on the glass appeared as a transparent green layer, indicating that the PANI is in emeraldine salt form. The surface morphology of the thin PANI coating was examined using the SEM and their images are shown in Fig. 2.
The surface morphology of PANI exhibits a “cluster” morphology with various cluster sizes and random distribution across the surface. The average cluster size is 9.13 µm2, which is estimated from Fig. 2(a). The cluster consisted of non-uniform nanofibers that clumped together to form the “cluster” morphology observed in Figs. 2(a) and 2(b). The average nanofiber size of the PANI is 150 nm based on Fig. 2(b). The “cluster” morphology is more pronounced on the tapered MMF as shown in Figs. 2(c) and 2(d). Since the fiber surface is cylindrical, the gravitational force and surface tension during the deposition process caused the nanofibers to conglomerate at certain area of the surface.
In order to measure the thickness of the PANI coating, a section of the glass slide was covered with an aluminium tape during the spray-coating process to obtain the boundary area between coated and uncoated glass. The boundary area was then examined using an atomic force microscope (NT-MDT Solver NEXT AFM). The scan boundary area for the AFM analysis was set to be a square area of 100.005 µm2 with 65536 sampling. The AFM images are depicted in Fig. 3. From the 3D image of the boundary area shown in Fig. 3(a), the average thickness of the PANI coating is estimated to be 0.7 ± 0.15 µm. The scan area is shown in Fig. 3(b). From the AFM analysis, the average roughness of the surface is found to be 12 nm. For gas sensing application, a rough sensing surface is needed to enhance the sensitivity. This is because a rougher surface increases the active interaction sites between the gas molecules and the sensing layer . An optical ammonia sensor uses PANI film with surface roughness of 17 nm has shown good sensing performance in . This shows that the impact of light scattering to the sensing performance is minimal for such level of roughness.
2.3 Ammonia sensing experimental setup
The tapered fiber sensor was fixed in a customized gas chamber and connected to a light source and a spectrophotometer through optical fiber cables as illustrated in Fig. 4. A tungsten-halogen lamp (Ocean Optics HL2000) with wavelength range of 360–2400 nm was used as the light source. Meanwhile, a miniature spectrophotometer (Ocean Optics USB4000) with spectral range of 200 nm to 1100 nm was used for the absorbance measurement. The spectrophotometer was connected via a USB port to a computer installed with the Spectrasuite software. The absorbance Aλ, is calculated by the software, using the following equation:
The gas chamber was attached to a gas calibration system with a computer-controlled mass flow controller, which regulates gas flow at 200 sccm. The ammonia concentration in the chamber was varied by controlling the dilution of 1% ammonia gas with high purity synthetic air. Absorbance measurement was conducted while the sensor was exposed to ammonia with concentration varied from 0.125% to 1%. The dynamic response of the sensor was observed through the change in cumulative absorbance as the system alternately purged the ammonia gas into the chamber with 100% purified air. All experiments were carried out at room temperature (26°C).
3. Results and discussions
3.1 Sensor absorbance response
Figure 5(a) gives the normalized absorbance spectra of the sensor as it is exposed to purified synthetic air and ammonia with concentrations ranging from 0.125% to 1%. The waist diameter of the sensor under test is 30 µm. It shows that the absorbance of the PANI coated fiber taper increases in tandem with ammonia concentration in the wavelength range of 570 nm – 730 nm with the peak absorbance seen at 649 nm. As the emeraldine salt PANI (green) is exposed to ammonia, it is dedoped through deprotonation process. This reaction modifies the PANI from emeraldine salt to emeraldine base form (blue). It is presumed that some amount of evanescent wave that propagate at the surface of the taper waist region is absorbed by the layer when it is in the emeraldine base form. The increase in ammonia concentration has increased the number of dedoped molecules, resulting in higher absorbance value.
To estimate the response and the recovery times of the sensor, cumulative absorbance is recorded as the sensor is exposed to ammonia for 5 minutes, followed by purified air for 15 minutes. The exposure is carried out for a few cycles for different ammonia concentrations. Figure 5(b) shows the dynamic response of the sensor in terms of the normalized cumulative absorbance, by integrating the absorbance over the visible range (500 nm to 800 nm). The figure gives the absorbance response of the sensor during the adsorption-desorption process of the ammonia from the PANI layer. The response time and the recovery time are found to be 2.27 minutes and 9.73 minutes, respectively. This response is faster than the response of sensor reported by Airoudj et al., which is around 4 minutes . The recovery of the developed sensor is considered slow and known to be one of the drawbacks for polymer-based sensors . However, the use of tapered fiber sensor improves the recovery to 98% as compared to the waveguide-based sensor developed by Airoudj et al. which is 33% after 15 minutes of regeneration .
3.2 Effect of tapered MMF waist diameter
From previous studies [19, 20], it is known that the waist diameter determines the sensitivity of tapered fiber-based sensor. In order to study the effect of waist diameter on the sensing performance of our sensor, tapered MMF with waist diameter sizes of 20 µm, 30 µm, and 40 µm were prepared and tested. The normalized cumulative absorbance for these sensors is shown in Fig. 6(a). Based on , reducing the waist diameter increases the light energy propagating in the evanescent field at the surface of the waist region. Hence, larger fraction of energy will be absorbed by the emeraldine base PANI layer, which contributes to higher cumulative absorbance. However, it is also observed that the sensor with smaller waist diameter has a larger baseline drift. The elevated sensitivity of sensors with smaller diameter led to the detection of even miniscule amount of molecules that had not fully regenerated into to emeraldine salt form after 15 minutes exposure to purified air. The change in the cumulative absorbance (Δ cumulative absorbance) as the sensor is exposed to ammonia with different concentrations is plotted in Fig. 6(b). This graph represents the sensitivity of the sensor that relates to its absorbance unit (AU). The average sensitivities are 5.56 AU/% concentration, 5.25 AU/% concentration and 4.25 AU/% concentration for waist diameter of 20 µm, 30 µm and 40 µm, respectively. As expected, the result shows that smaller waist diameter increases the sensitivity of the sensor. At ammonia concentration of 0.5%, the sensitivity increases for about 42% as the waist diameter was reduced from 40 µm to 30 µm. However, the sensitivity increment is only 3% as the waist diameter was reduced from 30 µm to 20 µm. At ammonia concentration of 1%, reducing the waist diameter from 40 µm to 30 µm increases the sensitivity by 46%. Meanwhile, reducing the waist diameter from 30 µm to 20 µm gives sensitivity increment of 15%. The large increment in sensitivity as the diameter was reduced from 40 µm to 30 µm can be attributed to the elevation of evanescent field energy that penetrates into the PANI layer. When the sensor is exposed to the lowest concentration of ammonia (0.125%), the change in the PANI layer occurs only at the top surface of the layer and it fully recovers after 15 minutes exposure to air.
3.3 Sensor repeatability
In order to observe the sensor repeatability, it was tested with three cycles of 5 minutes of 1% ammonia and 15 minutes of purified air. The sensor under test is the tapered MMF with 30 µm waist diameter. The response depicted in Fig. 7 shows good repeatability and reversibility. The difference in response between the first and the second cycle is 7%. Meanwhile, the second and the third cycle response differs by only 2%. The slight baseline drift occurs because the gas molecules are not fully desorbed from the PANI layer after 15 minutes exposure to air. This problem can be solved by exposition to the doping agents or thermal treatment .
In this paper, we have demonstrated for the first time a room temperature ammonia sensor based on tapered multimode fiber coated with polyaniline nanofibers. The sensor is able to distinguish different ammonia concentrations and it is reversible. The response and recovery time are 2.27 minutes and 9.73 minutes, respectively. This response is faster than a reported work based on planar waveguide sensor. The sensitivity of the sensor increases as the taper diameter decreases due to stronger evanescent field penetration beyond the fiber cladding. It is found that the 20 µm tapered fiber diameter gives the highest sensitivity compared with the larger diameter. Based on our investigation, the sensitivity increment is 61% when the taper diameter is reduced from 40 µm to 20 µm. The developed sensor also shows good repeatability and reversibility. The experiment exhibits the immense potential of designing other gas sensors based on tapered optical fiber coated with suitable sensing layer.
This work was supported in part by the Universiti Putra Malaysia, and Ministry of Education, Malaysia under research grant #05-02-12-2015RU, and in part by the University of Malaya under research grant UM.C/625/1/HIR/MOHE/SCI/01.
References and links
1. B. Timmer, W. Olthuis, and A. Berg, “Ammonia sensors and their applications—a review,” Sens. Actuator B-Chem. 107(2), 666–677 (2005). [CrossRef]
2. D. Verma and V. Dutta, “Role of novel microstructure of polyaniline-CSA thin film in ammonia sensing at room temperature,” Sens. Actuat. B-Chem. 134(2), 373–376 (2008). [CrossRef]
3. D. Nicolas-Debarnot and F. Poncin-Epaillard, “Polyaniline as a new sensitive layer for gas sensors,” Anal. Chim. Acta 475(1–2), 1–15 (2003). [CrossRef]
4. P. Stamenov, R. Madathil, and J. M. D. Coey, “Dynamic response of ammonia sensors constructed from polyaniline nanofibre films with varying morphology,” Sens. Actuat. B-Chem. 161(1), 989–999 (2012). [CrossRef]
5. H. Bai and G. Shi, “Gas sensors based on conducting polymers,” Sensors (Basel Switzerland) 7(3), 267–307 (2007). [CrossRef]
6. A. L. Kukla, Y. M. Shirshov, and S. A. Piletsky, “Ammonia sensors based on sensitive polyaniline films,” Sens. Actuat. B-Chem. 37(3), 135–140 (1996). [CrossRef]
7. A. D. Aguilar, E. S. Forzani, L. A. Nagahara, I. Amlani, R. Tsui, and N. J. Tao, “A breath ammonia sensor based on conducting polymer nanojunctions,” IEEE Sens. J. 8(3), 269–273 (2008). [CrossRef]
8. K. Crowley, A. Morrin, A. Hernandez, E. Omalley, P. Whitten, G. Wallace, M. Smyth, and A. Killard, “Fabrication of an ammonia gas sensor using inkjet-printed polyaniline nanoparticles,” Talanta 77(2), 710–717 (2008). [CrossRef]
9. A. L. Sharma, K. Kumar, and A. Deep, “Nanostructured polyaniline films on silicon for sensitive sensing of ammonia,” Sens. Actuat. A-Phys. 198, 107–112 (2013). [CrossRef]
10. M. H. Yaacob, M. Breedon, K. Kalantar-zadeh, and W. Wlodarski, “Absorption spectral response of nanotextured WO3 thin films with Pt catalyst towards H2,” Sens. Actuat. B-Chem. 137(1), 115–120 (2009). [CrossRef]
11. M. Yang and J. Dai, “Review on optical fiber sensors with sensitive thin films,” Photon. Sens. 2(1), 14–28 (2012). [CrossRef]
12. P. Wang, G. Brambilla, M. Ding, Y. Semenova, Q. Wu, and G. Farrell, “High-sensitivity, evanescent field refractometric sensor based on a tapered, multimode fiber interference,” Opt. Lett. 36(12), 2233–2235 (2011). [CrossRef] [PubMed]
13. J. Castrellon-Uribe, M. E. Nicho, and G. Reyes-Merino, “Remote optical detection of low concentrations of aqueous ammonia employing conductive polymers of polyaniline,” Sens. Actuat. B-Chem. 141(1), 40–44 (2009). [CrossRef]
15. L. Ai, J. C. Mau, W. F. Liu, T. C. Chen, and W. K. Su, “A volatile-solvent gas fiber sensor based on polyaniline film coated on superstructure fiber Bragg gratings,” Meas. Sci. Technol. 19(1), 017002 (2008). [CrossRef]
16. M. El-Sherif, L. Bansal, and J. Yuan, “Fiber optic sensors for detection of toxic and biological threats,” Sensors (Basel Switzerland) 7(12), 3100–3118 (2007). [CrossRef]
17. M. A. El-Sherif, “Fiber-optic chemical sensor using polyaniline as modified cladding material,” IEEE Sens. J. 3(1), 5–12 (2003). [CrossRef]
18. S. K. Mishra, D. Kumari, and B. D. Gupta, “Surface plasmon resonance based fiber optic ammonia gas sensor using ITO and polyaniline,” Sens. Actuator B-Chem. 171–172, 976–983 (2012). [CrossRef]
19. A. B. Socorro, I. Del Villar, J. M. Corres, I. R. Matias, and F. J. Arregui, “Lossy mode resonances dependence on the geometry of a tapered monomode optical fiber,” Sens. Actuat. A-Phys. 180, 25–31 (2012). [CrossRef]
20. A. Leung, K. Rijal, P. M. Shankar, and R. Mutharasan, “Effects of geometry on transmission and sensing potential of tapered fiber sensors,” Biosens. Bioelectron. 21(12), 2202–2209 (2006). [CrossRef] [PubMed]
22. L. Zhang, F. Gu, J. Lou, X. Yin, and L. Tong, “Fast detection of humidity with a subwavelength-diameter fiber taper coated with gelatin film,” Opt. Express 16(17), 13349–13353 (2008). [CrossRef] [PubMed]
23. H.-Y. Lin, C.-H. Huang, G.-L. Cheng, N.-K. Chen, and H.-C. Chui, “Tapered optical fiber sensor based on localized surface plasmon resonance,” Opt. Express 20(19), 21693–21701 (2012). [CrossRef] [PubMed]
24. N. Abdul Rahman, M. Gizdavic-Nikolaidis, S. Ray, A. J. Easteal, and J. Travas-Sejdic, “Functional electrospun nanofibres of poly(lactic acid) blends with polyaniline or poly(aniline-co-benzoic acid),” Synth. Met. 160(17–18), 2015–2022 (2010). [CrossRef]
25. B.-K. Min and S.-D. Choi, “SnO2 thin film gas sensor fabricated by ion beam deposition,” Sens. Actuat. B-Chem. 98(2–3), 239–246 (2004). [CrossRef]