Open Access
Add to BibSonomyAbstract
We demonstrated a novel selective chemical sensing approach by incorporating a poly(dimethylsiloxane) (PDMS)-coated fiber Bragg grating (FBG) structure for optically detecting various volatile organic compounds (VOC’s). When the proposed structure is exposed to a nonpolar solvent, a tensile stress is induced between the coated PDMS and the optical fiber by a VOC-dependent swelling effect of the PDMS, which results in a Bragg wavelength shift dependent on the concentration and the type of VOC’s. Because of no need of an etching process of a fiber cladding, the proposed PDMS-coated FBG can be used as a simple, convenient, and durable chemical sensing element with a high sensitivity, compared with conventional FBG sensors requiring an evanescent wave coupling.
©2010 Optical Society of America
1. Introduction
In the past few years, fiber Bragg grating (FBG) sensors have been studied extensively for optical sensing of several types of physical parameters such as strain, bending, pressure, temperature, and chemicals [1–5]. In using FBG sensors, the environmental perturbations on the FBG sensors, such as strain, temperature, and pressure, can be easily measured by monitoring the degree of the Bragg wavelength shift induced by the amount of the FBG deformation. However, for chemical sensing applications, normal FBGs cannot be used because optical fibers do not react to chemical solutions. Also, conventional FBGs are intrinsically insensitive to a surrounding-medium refractive index (SRI) variation because optical fields are well-bound within a fiber core and a light coupling with the SRI is screened by a thick cladding layer. For this reason, long-period fiber gratings (LPFGs) are generally used for chemical sensing applications, where the light couplings between the core and the cladding modes exist, accompanying an optical interaction between the cladding and an external medium [3,6–9]. However, the broad transmission resonance, the multiple resonance peaks, and the leaky waveguiding properties of the cladding modes of the LPFGs lower the measurement accuracy and make multipoint measurements difficult although the ability of the multipoint measurements, by using a wavelength-division multiplexing scheme, is one of important merits in using optical fiber sensing methods.
To resolve these problems, several types of modified FBG chemical sensors have been proposed, which enable the optical core mode to interfere with the SRI variation. Meltz et al. and Usbeck et al. demonstrated the SRI measurements by using a one-side etched D-shaped fiber or a side-polished fiber to increase the evanescent field interaction with the surrounding medium [10,11]. After that, the researchers have been trying to increase the sensitivity of the SRI measurement by optimizing the etched structures because their light couplings are highly dependent on the selected optical modes determined by the degree of the etching [12–14]. However, such devices required additional optical equipments in systems to consider polarization dependence induced by asymmetric light coupling behaviors. In addition, as for low-cost fabrication methods and sensitive SRI measurements, symmetrically etched microstructure FBGs (MSFBGs) were developed by using a hydrofluoric acid (HF)-based wet chemical etching or/and an electric arc-discharge process [15–17]. Nevertheless, the proposed methods remarkably reduced the durability and stability of the FBG sensor, which was caused by the etching process and the tapered fiber diameter. It was also difficult to get desired transmission spectra with high reproducibility by controlling the tapering process of the optical fiber.
In this paper, we proposed a novel approach for an optically selective chemical sensing method using a poly(dimethylsiloxane) (PDMS)-coated fiber Bragg grating without any additional etching process. The schematic illustration of the proposed device structure is presented in Fig. 1 . The principle mechanism is that the PDMS coated onto the bare FBG swells by the exposure to the volatile organic compounds (VOC’s), whereafter the swelled PDMS provides a tensile force to the bare fiber. Because the amount of this tensile force applied to the fiber is highly dependent on the swelling ratio of the PDMS to the nonpolar solvents, the concentration and the type of the exposed VOC’s can be precisely and selectively determined by monitoring the amount of the Bragg wavelength shift of the PDMS-coated FBG. Here, we provide the analytic model to describe the Bragg wavelength shift of the proposed structure in response to the exposed VOC’s and their experimental results, which show that the proposed method can be widely applied to chemical sensing areas with high selectivity, durability and stability.
Fig. 1 The schematic illustration of the PDMS-coated FBG device exposed to the nonpolar solvents. (Media 1)
2. Principle and experiments
2.1 Selective swelling effect of PDMS by VOC
The crosslinked PDMS, composed of repeating units of -OSi(CH3)2- groups, is an elastomeric polymer, which is widely used for several applications such as mirofluidic biochips and patterned molds for a soft-lithography process due to its superior flexibility in fabricating complex patterned structures [18]. The surface energy of the PDMS is very low, thus polar solvents like water do not wet on the surface [19]. However, nonpolar solvents, as listed in Table 1 , can be easily infiltrated into the crosslinked PDMS matrix, which accompanies the effective swelling effect of the PDMS due to its elasticity [20,21]. The covalent bonding between the building unit of the crosslinked PDMS is such highly stable that the infiltrated solvents do not dissolve the crosslinked PDMS. Therefore, the swelled structure is recovered into initial dimensions due to its elasticity after vaporization of the solvents. The degree of the partitioning behaviors between the PDMS and the solvents can be predicted by the solubility parameter (δ) of the solvents, where δ is expressed by a sum of the dispersion force (δd2), the polar force (δp2), and the hydrogen-bonding forces (δh2) of the material, Eq. (1) is expressed as follows [21].
When we measure the degree of the PDMS swelling after the solvent infiltration is saturated, the swelling ratio (S) is expressed as follows,where SVOC is the swelling ratio of the PDMS at each VOC, L is the initial length of the PDMS, and ΔLp is the increased length of the PDMS after saturating its swelling behavior. As shown in Table 1, the swelling ratio is monotonically increasing with decrease of the solubility parameter due to the polarity of the solvent, which means that SVOC is selectively determined by the solubility parameter of the VOC. The other factor determining SVOC is the elasticity of the PDMS block.
Table 1. Swelling ratios of PDMS and Hansen solubility parameters.
2.2 Experiments
The FBG (SMF-28) utilized in our experiments presented the central wavelength of 1554.9 nm, the 3 dB bandwidth of about 0.2 nm, and the peak reflectivity of about 90%. The longitudinal length of the FBG was 10 mm. Before coating the PDMS block structure, the FBG was prepared by the O2 Plasma treatment with 30 W for 30 seconds to enhance the interfacial adhesion between the PDMS and the optical fiber [22]. And then, the PDMS-coated FBG was fabricated with a molding process by using PDMS with a 10:1 mixing ratio of precursor of elastic polymer (Sylgard 184-A) and hardener (Sylgard 184-B). The crosslinked PDMS block was made by baking at 60 °C for 12 hours. Two kinds of the PDMS blocks with the different cross section areas (Ap = 1.5 × 1.5 mm2 and Ap = 2 × 2 mm2) were prepared to compare the sensitivity variation of the PDMS-coated FBG sensor depending on the elasticity of the PDMS block. The length of the PDMS block coated onto the FBG was L = 30 mm.
A broadband light source (Opto-Link Corporation Ltd. OLS15CLGB-15-FA) was connected to the proposed FBG sensor with an optical isolator. The transmission spectrum of the PDMS-coated FBG was monitored using an optical spectrum analyzer (OSA, Ando AQ-6315A) with a resolution bandwidth of 0.05 nm. During the experiments, temperature was kept constantly at 25°C by using a pyrostat. For characterizing our PDMS-coated FBG sensor in terms of the selectivity of VOC types, four kinds of pure nonpolar solvents were used, which were methanol (99.9%), 1-propanol (99.7%), 2-propanol (99.5%), and acetone (99.5%). For measuring the Bragg wavelength shift depending on the relative concentration, a solution, prepared by a binary mixture with these pure nonpolar solvents, was used. After every measurement, the infiltrated VOC was naturally volatilized in air for checking the recovery of the Bragg wavelength. Before immersing the other VOC, our device was additionally thermally annealed at 40 °C for 10 minutes to remove the VOC completely which is possible to remain in the PDMS matrix.
When the PDMS-coated FBG (Ap = 4.5 mm2) is immersed into acetone, the Bragg wavelength of the FBG shows a red-shift due to the tensile effect induced by swelling of the PDMS as acetone is absorbing into the PDMS block, as shown in Fig. 2(a) . There is no noticeable signal distortion in the transmission spectra of the FBG during the wavelength shift resulting from the PDMS swelling, which means that the distribution of the tensile force on the FBG, induced by the swelled PDMS, is uniform. When the PDMS-coated FBG is exposed to the air, the Bragg wavelength returns to its original wavelength location also without any deformation in the transmission spectra shape as shown in Fig. 2 (c), as the acetone evaporates from the PDMS. Figures 2(b) and (d) show the temporal evolutions of the Bragg wavelength shift when our sensing device is immersed into acetone and then the acetone, infiltrated into the PDMS matrix, is naturally evaporated. It is observed that the Bragg wavelength red-shifts by 0.36 nm. In comparison, the amount of the Bragg wavelength shift of the proposed device is comparable to those (~tens of pm) in previously reported works using the SRI measurement [10–14,16,17]. However, the required times for saturation of the swelling and de-swelling behaviors of the PDMS show slow responses, which are about 12 minutes and 23 minutes, respectively. The response time is determined by the saturated amount of the infiltrated solvent, which is related to the volume of the PDMS block. To improve the response time, the cross-section area of the PDMS should be small.
Fig. 2 (a) Time evolution of the transmission spectra of the PDMS-coated FBG at the exposure of the acetone; (b) Temporal evolution of the Bragg wavelength shift of the PDMS-coated FBG sensor during the acetone infiltration; (c) time evolution of the transmission spectra during the acetone volatilization; (d) temporal evolution of the Bragg wavelength shift during the acetone volatilization. (Media 2)
3. Results and discussion
3.1 Selectivity of the PDMS-coated FBG chemical sensor
In our PDMS-coated FBG chemical sensor, the relative Bragg resonance wavelength is obtained by variation of Bragg grating periodicity from the tensile force induced by the swelled PDMS. The tensile force relation between the fiber and the swelled PDMS can be express as
where Fp, Ff Δ L, and Fp Δ L are the swelling force of the PDMS by the VOC immersion considering the PDMS block only, the tensile force induced on the FBG, and the net tensile force of the PDMS considering Ff Δ L as the force from the reciprocal actions. To get the periodicity variation of the FBG, the increased length of the fiber by the swelled PDMS can be obtained by modification of Eq. (3) as follows, where kp, kf, and ΔL are the elastic constants of the PDMS and the bare fiber, and the increased length of the PDMS-coated FBG. ΔLp is given by the swelling ratio as expressed in Eq. (2). Considering the cross-section area (Ap, Af), length (Lp, Lf), and Young’s modulus (Ep, Ef) of the PDMS and the bare fiber [23], Eq. (4) can be expressed as follows,In our device structure, it can be assumed that the longitudinal lengths affected by the tensile force are the same (Lp = Lf = L) in both the PDMS and the fiber. Then, the relative length variation of Eq. (5) can be expressed in terms of SVOC by using Eq. (2) as follows,As a result, the shift in the Bragg wavelength of the PDMS-coated FBG can be described by as follows,
where ΔλB, λB, , and ε CH are the relative Bragg wavelength, the Bragg wavelength, the effective photoelastic coefficient of the fiber, and the strain induced on the fiber, respectively [24]. Finally, with Eqs. (6) and (7), the Bragg wavelength shift can be obtained as follows,Figure 3(a) shows the transmission spectra of the PDMS-coated FBG (Ap = 2.25 mm2) at the different VOS’s. In our measurement, the saturated values of ΔλB’s were used although the saturation time was different due to the different absorption rate originated from the solubility parameter of each VOC [18]. The saturated ΔλB’s increase with increasing the swelling ratios of the VOC’s listed in Table 1. The shapes as well as the 3 dB bandwidths of the transmission spectra are the same in all cases. In addition, the shifted Bragg wavelengths are recovered to the initial position irrespective of the VOC’s in our experiments. Figure 3(b) shows the saturated ΔλB’s at the different VOC’s depending on the cross section areas (Ap = 2.25 mm2, Ap = 4 mm2) of the coated PDMS, where the filled dots and the lines denote the measured results and the simulation results from Eq. (8), respectively. The material parameters used in our theoretical results are that the Young’s moduli of the PDMS and the bare fiber are Ep = 750 Kpa and 72 Gpa, respectively, and the diameter of the bare fiber is 125 μm [25,26]. The ΔλB’s of the PDMS-coated FBG sensor exposed to acetone, 2-propanol, 1-propanol, and methanol are 0.190 nm (0.306 nm), 0.084 nm (0.154 nm), 0.048 nm (0.090 nm), and 0.020 nm (0.042 nm), respectively, in Ap = 2.25 mm2 (or Ap = 4 mm2). As shown in Fig. 3(b), the simulated results using Eq. (8) match well with the measurement results, which show that the type of the VOC can be optically detected with good selectivity from the amount of ΔλB and the swelling ratio of the VOC to the PDMS. Because the amount of ΔλB is highly dependent to the cross section area of the fabricated PDMS block, we can use Ap as a variable device parameter to control the sensitivity of the PDMS-coated FBG sensor at each VOC, which means that the multiple detections from a VOC mixture can be achieved by cascading our PDMS-coated FBG sensors with different Ap’s. In a future development, we will show the multiple detections from a VOC mixture including analysis on temperature and strain variation conditions in our device. Because ΔλB increases but the response time becomes slower as Ap increases, a polymer, showing a swelling-induced high tensile force, should be explored to get a high sensitivity as well as faster response.
Fig. 3 (a) Transmission spectra of the PDMS-coated FBG (A p = 2.25 mm2) at different VOC’s; (b) ΔλB at different VOC’s depending on the cross-section area of the coated PDMS, where the dots and the lines denote the measured results and the simulation results from Eq. (8), respectively. (Media 3)
3.2 Measurement of relative VOC concentration
For chemical sensor applications, the VOC concentration as well as the VOC type should be measurable and distinguishable. For preparing a diluted VOC with various concentration conditions, a polar solvent like water cannot be used in our system because it does not infiltrate into the PDMS matrix. Here, methanol is mixed with the other nonpolar solvent for a diluent to vary concentration conditions because the swelling ratio of methanol is insignificant, as listed in Table 1, but it can be infiltrated into the PDMS matrix. When a binary mixture of nonpolar solvents infiltrates into the PDMS, the swelling ratio of the binary mixture varies linearly with the weighted average value using the swelling ratios of two solvents, where the weighting parameter is chosen with the relative volume ratio of the mixture [27]. When the relative volume ratio (01) is used as the VOC concentration, ΔλB by a binary VOC mixture can be expressed by modifying Eq. (8) as follows.
where SVOC1 and SVOC2 are the swelling ratios of two different VOCs constituting a binary mixture and ΔCH denotes the relative volume ratio of VOC1 in the mixture.Figure 4 shows ΔλB’s as a function of the relative concentration ratio of acetone (VOC1) diluted with methanol (VOC2), where the dots and the lines denote the measurement results and the simulation results from Eq. (9), respectively. As for the material parameters to obtain the simulation results, the same values with those used in Fig. 3(b) are applied. The experimental results show that ΔλB’s increase linearly with increasing the concentration ratio of acetone, which match well with our theoretic model of Eq. (9). The sensitivities of ΔλB to the concentration variation are ΔλB = 1.63×10−3 nm/% for Ap = 2.25 mm2 and ΔλB = 2.9×10−3 nm/% for Ap = 4 mm2 because the swelling force induced by the PDMS block, competing with the tensile force of the fiber, becomes higher as Ap increases. The result of Fig. 4 also shows that a simultaneous chemical sensing of the VOC types and their relative concentration from a VOC mixture can be obtained from Eq. (9) by using a cascaded PDMS-coated FBG device.
Fig. 4 ΔλB of the PDMS-coated FBG sensor as a function of the concentration ratio (the relative volume ratio) of the VOC binary mixture (acetone diluted with methanol), where ΔCH denotes the relative volume ratio of acetone in the mixture. The dots and the lines denote the experimental results and the simulation results from Eq. (9), respectively. (Media 4)
4. Conclusions
We propose the optically selective chemical sensing method using the PDMS-coated FBG sensor without any etching process of the fiber cladding. The PDMS-coated FBG sensor can selectively measure the type as well as the concentration of a VOC by monitoring the Bragg wavelength shift which is induced by the selective swelling effect of the PDMS exposed to a solvent. In particular, the experiment results as well as the theoretic model show that the amount of the effective tensile force onto the imbedded FBG, induced by the swelled PDMS, is highly dependent to the thickness of the PDMS, thus we can control the sensitivity easily by varying the cross section area of the PDMS, which means that simultaneous sensing of the types and their concentrations from a solvent mixture can be achieved by cascading our devices with two different coating thicknesses and by considering the amount of the Bragg wavelength shifts with respect to each coating condition. The proposed PDMS-coated FBG chemical sensor is expected to be widely applied to various industrial areas requiring multi-point chemical sensing tools due to its high durability and stability [28,29].
Acknowledgements
This research was financially supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) grant funded by the Korea government (MEST) (2010-0001884).
References and links
1. A. D. Kersey, M. A. Davis, H. J. Patrick, M. Le Blanc, K. P. Koo, C. G. Askins, M. A. Davis, and E. J. Friebele, “Fiber graing sensors,” J. Lightwave Technol. 15(8), 1442–1463 (1997). [CrossRef]
2. A. Othonos, and K. Kalli, Fiber Bragg Gratings Fundamentals and Applications in Telecommunications and Sensing, (Boston: Artech House, 1999).
3. Y. J. Rao, “Recent progress in applications of in-fiber Bragg grating sensors,” Opt. Lasers Eng. 31(4), 297–324 (1999). [CrossRef]
4. L. Zhang, W. Zhang, and I. Bennion, “In-fiber graing optic sensors” in Fiber Optics Sensors (New York: Dekker, Chaper 4, 2002).
5. X. Shu, K. Chisholm, I. Felmeri, K. Sugden, A. Gillooly, L. Zhang, and I. Bennion, “Highly sensitive transverse load sensing with reversible sampled fiber Bragg gratings,” Appl. Phys. Lett. 83(15), 3003–3005 (2003). [CrossRef]
6. V. Bhatia and A. M. Vengsarkar, “Optical fiber long-period grating sensors,” Opt. Lett. 21(9), 692–694 (1996). [CrossRef] [PubMed]
7. X. Shu and D. X. Huang, “High sensitive chemical sensor based on the measurement of the separation of dual resonant peaks in a 100-mm period fiber grating,” Opt. Commun. 171(1-3), 65–69 (1999). [CrossRef]
8. X. Shu, X. L. Zhang, and I. Bennion, “Sensitivity characteristics of long-period fiber gratings,” J. Lightwave Technol. 20(2), 255–266 (2002). [CrossRef]
9. J. A. Barnes, R. S. Brown, A. H. Cheung, M. A. Dreher, G. Mackey, and H.-P. Loock, “Chemical sensing using a polymer coated long-period fiber grating interrogated by ring-down spectroscopy,” Sens. Act. B Chem. 148(1), 221–226 (2010). [CrossRef]
10. G. Meltz, S. J. Hewlett, and J. D. Love, “Fiber grating evanescent-wave sensors,” Proc. SPIE 2836, Chemical, Biochemical, and Environmental Fiber Sensors VIII, 1996.
11. K. Usbeck, W. Ecke, A. Andreev, V. Hagemann, R. Mueller, and R. Willsch, “Distributed optochemical sensor network using evanescent field interaction in fiber Bragg gratings,” Proc. SPIE 3483, First European Workshop on Optical Fibre Sensors, 1998.
12. K. Schroeder, W. Ecke, R. Mueller, R. Willsch, and A. Andreev, “A fibre Bragg grating refractometer,” Meas. Sci. Technol. 12(7), 757–764 (2001). [CrossRef]
13. K. Zhou, X. Chen, L. Zhang, and I. Bennion, “High-Sensitivity optical chemsensor or based on eteched D-fiber Bragg gratings,” Electron. Lett. 40(4), 232–234 (2004). [CrossRef]
14. R. Willsch, W. Ecke, G. Schwotzer, and H. Bartelt, “Nanostructure-based optical fibre sensor systems and examples of their application,” Proceedings of SPIE 6585, International Congress on Optics and Optoelectronics, 2007.
15. W. Liang, Y. Huang, Y. Xu, R. K. Lee, and A. Yariv, “High sensitive fiber Bragg grating refractive index sensors,” Appl. Phys. Lett. 86(15), 151122 (2005). [CrossRef]
16. A. Iadicicco, S. Campopiano, A. Cutolo, M. Giordano, and A. Cusano, “Refractive Index Sensor Based on Microsturctured Fiber Bragg Grating,” IEEE Photon. Technol. Lett. 17(6), 1250–1252 (2005). [CrossRef]
17. D. Paladino, A. Iadicicco, S. Campopiano, and A. Cusano, “Not-lithographic fabrication of micro-structured fiber Bragg gratings evanescent wave sensors,” Opt. Express 17(2), 1042–1054 (2009). [CrossRef] [PubMed]
18. J. N. Lee, C. Park, and G. M. Whitesides, “Solvent compatibility of poly(dimethylsiloxane)-based microfluidic devices,” Anal. Chem. 75(23), 6544–6554 (2003). [CrossRef] [PubMed]
19. M. Morra, E. Occhiello, R. Marola, F. Garbassi, P. Humphrey, and D. Johnson, “On the aging of oxygen plasma-treated polydimethylsiloxane surfaces,” J. Colloid Interface Sci. 137(1), 11–24 (1990). [CrossRef]
20. J. Hildebrand, and R. L. Scott, The solubility of Nonelectrolytes, (New York: Reinhold, 1950).
21. C. M. Hansen, Hasen solubility parameter: a user’s hand book, (Florida: CRC Press 2000).
22. S. Bhattacharya, A. Datta, J. M. Berg, and S. Gangopadhyay, “Studies on Surface Wettability of Poly(Dimethyl)Siloxane (PDMS) and Glass Under Oxygen-Plasma Treatment and Correlation With Bond Strength,” J. Microelectromech. Syst. 14(3), 590–597 (2005). [CrossRef]
23. T. Young, and D. Hugh, University Physics, (7th Ed., Addison Wesley, 1992).
24. G. Meltz and W. W. Morey, “Bragg grating formation and germanosilicate fiber photosensitivity,” Proc. SPIE 1516, 185–199 (1991). [CrossRef]
25. F. P. Mallinder and B. A. Proctor, “Elastic constants of fused silica as a function of large tensile strain,” Phys. Chem. Glasses 2, 91–103 (1964).
26. D. Armani, C. Liu, and N. Aluru, “Re-configurable Fluid Circuits by PDMS Elastomer Micromachining,” Proc. IEEE Int. Conf. Micro-Electro Mech. Syst. (17–21 January 1999), pp. 222–227.
27. J.-H. Seo, R. Matsuno, T. Konno, M. Takai, and K. Ishihara, “Surface tethering of phosphorylcholine groups onto poly(dimethylsiloxane) through swelling--deswelling methods with phospholipids moiety containing ABA-type block copolymers,” Biomaterials 29(10), 1367–1376 (2008). [CrossRef]
28. A. Hajizadeh and M. A. Golkar, “Power flow control of grid-connected fuel cell distributed generation systems,” KIEE J. Electr. Eng. Tech. 3(2), 143–151 (2008). [CrossRef]
29. H.-S. Kang, G.-Y. Choe, B.-K. Lee, and J. Hur, “A feasibility design of PEMFC parallel operation for a fuel cell generation system,” KIEE J. Electr. Eng. Tech. 3(3), 408–421 (2008). [CrossRef]
