A fibre optic surface plasmon resonance (SPR) sensor system for smartphones is reported, for the first time. The sensor was fabricated by using an easy-to-implement silver coating technique and by polishing both ends of a 400 µm optical fibre to obtain 45° end-faces. For excitation and interrogation of the SPR sensor system the flash-light and camera at the back side of the smartphone were employed, respectively. Consequently, no external electrical components are required for the operation of the sensor system developed. In a first application example a refractive index sensor was realised. The performance of the SPR sensor system was demonstrated by using different volume concentrations of glycerol solution. A sensitivity of 5.96·10−4 refractive index units (RIU)/pixel was obtained for a refractive index (RI) range from 1.33 to 1.36. In future implementations the reported sensor system could be integrated in a cover of a smartphone or used as a low-cost, portable point-of-care diagnostic platform. Consequently it offers the potential of monitoring a large variety of environmental or point-of-care parameters in combination with smartphones.
© 2015 Optical Society of America
Currently, many different approaches for point-of-care-tests (POCT) for smartphones and mobile phones are investigated . For instance, digital microscopes for smartphones or mobile phones have been developed for bioanalytical applications [2–6]. Another technique relies on adaptive lenses for mobile phone cameras to read out commercially available external POCT . In contrast to such applications, Ruano-Lopez et al. developed a concept of a SmartBioPhoneTM . The SmartBioPhoneTM consists of a usual smartphone with external electrical read-out system for POCT based on lab-on-a-chip (LOC) devices. Also a spectrometer based on a mobile phone has been proposed for e.g. the measurement of the absorption spectrum of human skin .
Moreover, optical sensors based on photonic crystals , evanescent fields  or surface plasmon resonances (SPR) [12, 13] have been explored in combination with a smartphone in order to develop sensors for chemical/biochemical applications. The photonic crystal and evanescent field based optical sensor systems rely on a free space optical light path that guides light from an external light source via the optical sensor, i.e. the photonic crystal or a glass prism in case of the evanescent field sensor, to the camera of the smartphone. The free space optical light paths of both sensor systems consist of several external optical components such as a pinhole and various lenses, for example. Furthermore, in both sensor systems a diffraction grating is applied in front of the smartphone camera in order to disperse the light into a line spectrum. The obtained line spectrum allows the measurement of the transmission spectrum of the photonic crystal or the absorption spectrum of the evanescent field. However, since an external light source and a free space optical light path are required the systems reported so far for these approaches are relatively bulky.
Optical sensors based on SPR have been applied in the past in order to measure various biological and chemical analytes . Surface plasmon waves (SPW) are electron density oscillations where the associated transverse-magnetic (TM) polarised electromagnetic waves are guided parallel to metal-dielectric interfaces. They can be resonantly excited by light when the propagation constants of the incident light and the SPW are equal, which becomes obvious by e.g. a dip in the angular or wavelength spectrum of the transmitted light. The propagation constant of the SPW strongly depends on the refractive index (RI) of the surrounding medium and thus SPR sensors can be used as highly sensitive refractometers.
The SPR sensors which have been designed for smartphones are based on discrete lenses  or diffraction gratings  in order to excite SPWs at a gold-polymer or gold-glass interface. For the excitation and interrogation of the SPR sensors the display as well as the front camera of the smartphone have been utilised so far and the change of the SPR due to environmental changes has been measured by resolving the angle of the incident light.
In this paper, a fibre optic SPR sensor system for smartphones for concentration measurements on liquids is reported, for the first time. SPR sensors based on optical fibres have the advantages of being miniature in size, of enabling simple optical coupling and optical alignment as no external prisms, optical lenses etc. are required as well as that low-cost optoelectronic components can be applied for interrogation . A schematic of the fibre optic SPR sensor developed here is shown in Fig. 1. For the sensor system the LED and the camera at the back side of the smartphone are used as light source and detector. Both optoelectronic components of the smartphone are coupled to the fibre optic SPR sensor by polishing both end-faces of the optical fibre to 45°. Furthermore, the SPR sensor was realized by coating approx. 10 mm of the optical fibre core with a thin silver layer and a diffraction grating is applied in order to disperse the light into a line spectrum, which allows the tracking of the wavelength shift of the SPR due to a change of the surrounding RI.
2. Fibre optic SPR smartphone sensor
The fibre optic SPR smartphone sensor was fabricated by initially establishing the approx. 10 mm long sensing part. For this, the polymer coating of a 400 µm (core diameter) plastic cladded silica (PCS) fibre (Thorlabs BFL48-400, NA = 0.48) was removed first by using a razor blade and then the bare glass fibre was cleaned by using acetone and distilled water subsequently. A 400 µm PCS fibre was applied for the sensor system due to the ease of handling (less rigid than e.g. a 600 µm PCS fibre) and the light coupling efficiency (higher surface area compared to 100 µm or 200 µm PCS fibres and hence higher light coupling efficiency).
Next the sensor part was coated with silver by using a chemical method adapted from Pal et al.  and Zhao et al. . In order to fabricate the silver coating the bare fibre was carefully prepared to achieve an appropriate outcome. First the bare fibre was cleaned with a stannous chloride solution (0.2%) to sensitise the coating area and then rinsed with distilled water. A dextrose solution (0.05 mol/L) was poured into a beaker containing 1:5 diluted Tollens’ reagent and the bare fibre section. The silver film was thus formed very rapidly on the surface of the bare fibre. Finally, the coated fibre section was entirely cleaned with distilled water. Due to the dilution of the Tollens’ reagent the rate and strength of deposition was controlled and hence the silver coating process was very repeatable. The Tollens’ reagent was achieved beforehand by: (1) Pouring an aqueous solution of silver nitrate (2 ml, 0.1 M) in a tube. (2) Pouring ammonium hydroxide (25%) to the solution dropwise with stirring (first a brown precipitate will from). Continuing adding ammonium hydroxide until the solution becomes clear again. (3) Adding potassium hydroxide (1.4 ml, 0.8 M) so that the solution becomes brown again. (4) Pouring ammonium hydroxide (25%) dropwise until the solution becomes clear again.
After the silver coating process was finished the outcome and the performance of the fibre optic SPR sensor was verified by using the interrogation system shown in Fig. 2(a). The system consists of a white light source (Energetiq Eq. (-99)X) and a spectrometer (StellarNET EPP2000). The fabricated fibre optic SPR sensor was connected to the devices by using bare fibre SMA connectors. As shown in Fig. 2(b) the fibre optic SPR sensor shows a clear resonance which shifts towards the red wavelength range for increasing RI of the surrounding liquid. This result is consistent with the results reported in the literature for SPR sensors based on silver coated optical fibres . The RI of the liquid was changed by using different glycerol solutions as explained in section 3.
Following this, the length of the optical fibre was reduced to 25 cm and both end-faces were polished in order to obtain 45° angled end-faces. The polishing was performed by using the polishing machine Presi “le Cube” and a 45° angled fibre holder. The polishing process was started with P600 and P1200 polishing paper and finished with diamond (3 µ) and alumine (0.04 µ) suspensions. After the polishing process both 45° end-faces were rinsed in distilled water in order to clean the polished surfaces.
The diffraction grating of the sensor system is based on a Thorlabs holographic reflective diffraction grating with 1200 Lines/mm (Thorlabs GH13-12V). Since a transmission diffraction grating is more suitable for the sensor system a negative copy of the Thorlabs grating was made by using PDMS (Sylgard 184). The negative copy was obtained by coating the microstructure surface of the Thorlabs grating first with a layer of PDMS followed by a subsequent curing at room temperature for 48 hours. At the end the PDMS was removed from the Thorlabs grating and both the fibre optic SPR sensor and the PDMS diffraction grating were glued on top of a protective cover for a smartphone (Huawei Ascend Y300 (Android 4.1.1)), as illustrated in Fig. 3.
The fibre optic SPR sensor was illuminated and the resulting sensor spectrum was acquired by using the standard camera app of the smartphone and turning the flashlight on. For each measurement an mp4-video (480 x 640 pixels) was taken and subsequently processed by using Matlab. The signal processing included the segmenting of the mp-4 video into several images which are then converted into greyscale and normalised by using standard Matlab functions. Since the transmission grating disperses the light of the sensor into a line spectrum the resulting grey-scaled pictures can be interpreted as sensor spectrum. An example of an RGB picture obtained is shown in Fig. 4(a). In the picture the zero, first and second order diffraction can be identified. The corresponding grey-scaled intensity distribution of row 200 and, thus, the sensor spectrum is illustrated in Fig. 4(b). In order to analyse the SPR sensor spectrum the first order diffraction was used.
3. Experimental results
In order to evaluate the fibre optic SPR sensor system the active sensing region was immersed in different glycerol solutions and the resulting sensor spectrum was analysed. The different solutions were obtained by changing the volume concentrations of glycerol . The corresponding refractive indices of the glycerol solutions, summarised in Table 1, were determined by measuring the Fresnel reflections that occur at the interface between the solutions and the end-face of an optical fibre by using a powermeter (FiboTec dB-meter). The power readings of the powermeter have been calibrated to known RI (air, distilled water and ethanol) beforehand. For the designed sensor system the measureable shift of the SPR and hence the measureable RI range is limited by the spectral range of the smartphone-camera. Since the smartphone-camera only operates in the wavelength range from 400 - 680 nm  the excitation of the SPR was investigated for glycerol solutions with concentrations between 0 and 20% . Moreover, to avoid any measurement errors due to external light sources, the experiments were conducted at room temperature in darkroom environments.
In Fig. 5(a) the obtained sensor spectra (intensity distribution versus camera pixel number, i.e. wavelength) at different glycerol solutions are illustrated. In order to identify the SPR more effectively, the measured sensor spectrum was normalised to the spectrum when the sensor was exposed to air. As shown in Fig. 5(a), a clear wavelength shift of the SPR was obtained when the glycerol solutions were altered. From Fig. 5(a) it follows that for increasing volume concentrations the SPR shifts towards the red wavelength range. This result is consistent with the results obtained in Fig. 2(b). The rather broad SPR in the spectral domain results from the modal dispersion of the 400 µm optical fibre. In terms of sensing applications a narrow SPR spectral response is usually preferred. However, with respect to the developed fibre optic SPR smartphone sensor system a broadband SPR can be detected more efficiently, since the spectral resolution of the measured line spectrum is relatively low due to the beam divergence and the relatively large beam diameter of the light coupled out of the 400 µm optical fibre. Note that we use non-polarised light of the smartphone LED for illumination which does not deteriorate the sensor performance. Since the optical fibre and the deposited silver coating are axially symmetric the only component of the reported sensor system which is sensitive to polarisation is the absolute diffraction efficiency of the diffraction grating. Consequently, the measured line spectrum of the developed sensor system represents a superposition of both light polarisations. In Fig. 5(b) the shift of the SPR in the spectral domain as function of the glycerol volume concentrations is shown. A linear dependence of the SPR was obtained with a sensitivity of 1.83·10−3 RIU/ pixel. The sensitivity of the sensor system can be further enhanced by e.g. optimising the dispersion of the diffraction grating and the beam profile of the light coupled out of the optical fibre as well as by increasing the pixel number of the smartphone camera.
Another possibility to enhance the sensor sensitivity is to increase the resolution of the captured picture. Instead of recording an mp4-video a 5 MP (2592 x 1944 pixels) JPEG picture was taken for each glycerol solution. Following this, each picture was converted into greyscale and processed as explained above. In Fig. 6 the pixel position of the SPR is illustrated depending on the refractive index of the glycerol solution. Since the resolution of the pictures is now 2592 x 1944 pixels the sensitivity of the sensor system increased to 5.96·10−4 RIU/ pixel.
In this paper a simple and low-cost fibre optic SPR sensor for smartphones has been successfully realised and its applicability for refractive index measurement on liquids has been demonstrated. The sensor system shows a linear response to refractive index changes between 1.33 and 1.36 with a sensitivity of 5.96·10−4 RIU/ pixel. The measureable refractive index range of the sensor system is limited by the spectral range of the smartphone-camera. In the future, the sensitivity of the sensor system can be further enhanced by e.g. optimizing the dispersion of the transmission diffraction grating used to obtain the line spectrum, by optimizing the beam profile of the light coupled out of the fibre by e.g. using smaller-diameter fibers and by increasing the resolution of the smartphone camera. Compared to SPR sensors for smartphones reported so far, our waveguide based SPR sensor system has the advantage that it can be implemented into a planar optical waveguide structure and thus be integrated into the protective cover of the smartphone, for example, or applied as low-cost and disposal lab-on-a-chip device. Hence, depending on the functionalization of the sensor surface, the SPR waveguide sensor system could be used to detect trace gases or applied as a pregnancy test as well as to monitor diabetes, for instance.
This work was performed under the support of the “Wege in die Forschung”- grant of the Leibniz Universität Hannover. We also acknowledge support by Deutsche Forschungsgemeinschaft and Open Access Publishing Fund of Leibniz Universität Hannover.
References and links
8. J. M. Ruano-López, M. Agirregabiria, G. Olabarria, D. Verdoy, D. D. Bang, M. Bu, A. Wolff, A. Voigt, J. A. Dziuban, R. Walczak, and J. Berganzo, “The SmartBioPhone™, a point of care vision under development through two European projects: OPTOLABCARD and LABONFOIL,” Lab Chip 9(11), 1495–1499 (2009). [CrossRef] [PubMed]
9. S. X. Wang and X. J. Zhou, “Spectroscopic sensor on mobile phone,” US Patent 7420663 B2 (September 2, 2008).
11. S. Dutta, A. Choudhury, and P. Nath, “Evanescent wave coupled spectroscopic sensing using smartphone,” IEEE Photon. Technol. Lett. 26(6), 568–570 (2014). [CrossRef]
12. P. Preechaburana, M. C. Gonzalez, A. Suska, and D. Filippini, “Surface plasmon resonance chemical sensing on cell phones,” Angew. Chem. Int. Ed. Engl. 51(46), 11585–11588 (2012). [CrossRef] [PubMed]
13. C. A. de Souza Filho, A. M. N. Lima, and H. Neff, “Smartphone based, portable optical biosensor utilizing surface plasmon resonance,” in Proceedings of IEEE Conference on Instrumentation and Measurement Technology (IEEE, 2014), pp. 890–895. [CrossRef]
14. J. Homola, Surface Plasmon Resonance Based Sensors (Springer, 2006)
15. W. Peng, Y. Liu, P. Fang, X. Liu, Z. Gong, H. Wang, and F. Cheng, “Compact surface plasmon resonance imaging sensing system based on general optoelectronic components,” Opt. Express 22(5), 6174–6185 (2014). [CrossRef] [PubMed]
16. A. Pal, R. Sen, K. Bremer, S. Yao, E. Lewis, T. Sun, and K. T. V. Grattan, “All-fiber’ tunable laser in the 2 μm region, designed for CO2 detection,” Appl. Opt. 51(29), 7011–7015 (2012). [CrossRef] [PubMed]
17. Y. Zhao, Z. Deng, and Q. Wang, “Fiber optic SPR sensor for liquid concentration measurement,” Sens. Actuators B Chem. 192, 229–233 (2014). [CrossRef]