Absorbance response of a graphene oxide coated U-bent optical fiber sensor for aqueous ethanol detection

Abstract

In this work, the dynamic change of the absorbance for aqueous ethanol detection based on a graphene oxide (GO) coated U-bent optical fiber sensor (UOFS) was investigated. The U-bent probe was fabricated by bending an unclad multimode plastic cladding silica fiber (nominally at 62.5/125 μm) into a U-shaped structure. GO film was coated on the surface of the U-bent probe using a dip-coating technique. Transmission electron microscopy (TEM), Raman spectroscopy and scanning electron microscopy (SEM) were performed to characterize the GO film. A comparison of absorbance spectra and the dynamic absorbance response of the UOFS with and without GO film towards aqueous ethanol with a concentration from 5% to 100% was carried out in a wavelength range from 500 to 800 nm. The UOFS coating with GO film exhibits higher resolution and sensitivity compared with that of the UOFS without GO film. In addition, the UOFS with GO film shows excellent repeatability and reversibility, and the response and recovery times are as short as 1–2 s.

---

1. Introduction

Refractive index (RI) is a basic datum in optics and it is an optical property of materials, especially in liquids. Several methods have been developed by many scholars to measure RI, such as high-performance liquid chromatography, mass spectrometry, liquid chromatography/mass spectrometry, and infrared spectroscopy in past decades.¹ However, these traditional ways of RI measurement are time consuming and require large samples, complex operations, and experienced operators. Optical fiber, owing to high sensitivity corresponding to an external disturbance, has been widely studied as a sensor for chemicals, biological samples and in physical quantity detection in the past few decades, especially measuring RI changes in small volumes of liquid.^2–5 Besides, an immunity to electromagnetic interference and temperature, low attenuation, low cost, and small size are some advantages of optical fiber sensors over conventional sensors. Optical fiber sensors have become an emerging field in RI measurement, such as RI sensors based on evanescent wave^6–11 and surface plasmon resonance (SPR) determination.^12–14 Nonetheless, the high penetration depth of the evanescent field (particularly for the UV region) and the accurate thickness (nanometer scale) of thin metal films (such gold or silver) are also crucial for a sensitive signal.¹⁵ An alternative to this is to introduce geometric bends to increase the sensitivity in the sensing region. Such bends lead to the coupling of energy between diverse guided modes and leaky modes, and as a result the output intensity reduces and more power interacts with the analyte in the sensing region. At the same time, the output is still quite sensitive to RI changes in the surrounding medium in the sensing region. Different optical fiber shape designs including straight, U-bent, tapered tip, and biconical tapers have been employed in development of absorbance based sensors.^16–19 The tapered fibers can also achieve high sensitivity compared with normal fibers. However, U-bent fibers have some advantages over tapered fibers. First, the U-bent fibers can be prepared in a reproducible and facile way; we can obtain a U-bent fiber just by folding both ends of the fiber at high temperature, while it is difficult to control the shape of a taper with the same cone angles in a reproducible manner. Second, U-bent fibers are much more suitable for detection in narrow and small spaces; on the contrary, tapered fibers are easily destroyed in narrow spaces. Besides, U-bent fibers have better compatibility with instrument configurations, unlike tapered fibers which require a Y-type coupler.^20,21 What’s more, U-bent probes have been demonstrated to exhibit a 10-fold improvement in absorbance sensitivity over straight probes.²¹ Owing to these factors, our sensor was constructed using a U-bent optical fiber.

Graphene oxide (GO) is one of the graphene derived nanomaterials, endowed with extraordinary characteristics and benefits, including high surface-to-volume ratio, excellent transparency, high thermal conductivity and so on. Functional groups such as epoxy, carbonyl, and carboxylic acid endow GO with a better affinity than graphene in aqueous solution or organic solvents. Recently, some articles have demonstrated that GO is an attractive material for the detection of chemical molecules, such as ethanol and benzene,^22–24 that is, GO film can capture ethanol or benzene molecules effectively. Flexibility and high light transmission allow GO to effectively combine with optical fiber.²⁵ A tapered optical fiber sensor with GO for aqueous ethanol detection in a concentration range from 5% to 80% has been demonstrated by Yaacoba’s group.²⁶ However, to the best of our knowledge, no articles have been published that combine GO with U-bent optical fiber probes.

In this paper, we present a novel GO coated UOFS for aqueous ethanol detection in a concentration range from 5% to 100%. The U-bent probe was fabricated by bending an unclad multimode plastic cladding silica fiber into a U-shaped structure and coated with GO using a dip-coating technique. Importantly, the developed sensor has high resolution, perfect sensitivity, a wide measurement range and a fast response and recovery towards aqueous ethanol in a concentration range from 5% to 100%. The second linear fitting relationship between concentration and absorbance makes the developed sensor an effective platform for aqueous ethanol detection.

2. Experimental methods

2.1. U-bent probe fabrication

A standard multimode plastic cladding silica fiber with a core and cladding diameter of 62.5 and 125 μm respectively was used in this experiment. Briefly, optical fibers were cut into 40 cm long pieces using an optical fiber cleaver with 2 cm unclad in the middle area. After the fiber was washed with ethanol and acetone to remove possible residual polymers and contaminants, the middle area of the fiber was heated using a butane candle flame and bent into a U-bent structure. It should be noted that an optimum temperature and bending force should be maintained to ensure that the bending diameter remains the same.

2.2. Preparation of GO thin film

The GO was prepared using the modified Hummers method.²⁷ In a typical synthesis, graphite powder was dissolved in a cold mixture of H₂SO₄ and NaNO₃. KMnO₄ was added slowly with magnetic stirring and the mixture was cooled in an ice water bath to keep the reaction temperature below 20 °C. The mixture was stirred at 35 °C for 2 h, and then diluted with water. Additional water and H₂O₂ were added. Finally, the mixture was washed and purified with HCl aqueous solution and water. The resulting GO solid product was dried under vacuum at 60 °C. A GO suspension was prepared by suspending the GO solid in water with a concentration of 0.5 mg ml⁻¹. The GO suspension was coated on the surface of the U-bent probe using a dip-coating method. Before and after the coating, the fiber was annealed on a heating platform at 50 °C for 10 minutes to enhance the combination and deposition of GO film on the surface.

2.3. Optical set-up

The UOFS contains four portions, as shown in Fig. 1(a). In the order from left to right is the light source, U-bent fiber, spectrometer (PG2000, Ideaoptics Instruments) and the PC. A white light emitting diode generating 470–850 nm wavelength light was used as the light source. In order to obtain a steady light intensity, a constant current source was used as the power source. One end of the U-bent optical fiber was coupled to the light source and another end was coupled to the spectrometer using an optical fiber patch cord. The intensity of light coupled into the fiber changed when the UOFS was immersed in aqueous ethanol at different concentrations. The output light was collected and the full absorbance spectra were recorded in-time by a spectrometer. At last, Morpho software installed on the computer was used to process the data recorded by the spectrometer.

Fig. 1 Schematic diagram for (a) the experimental setup and sensing scheme used and (b) the U-bent sensing probe. (c) The guiding mechanism of light in the U-bent optical fiber.

2.4. Sensing mechanism

In the U-bent region as shown in Fig. 1(b), the total reflection law of light transmission is broken due to the bending, which leads to leaky modes, a greater number of light reflections and deeper penetration depth of the evanescent field. However, the transmittance of light depends on the leaky modes of light and absorption of the evanescent field penetrating into the sensing region formed by the absorbing sample fluid.²⁰ In this case, the absorbance would change significantly due to the imperceptible RI variation compared to the straight fiber.

As shown in Fig. 1(c), the light transmitting in the core would reflect and refract at the interface of the core and cladding. Then, the light radiating into the cladding would also reflect and refract at the interface of the cladding and GO film. Partial light would leak out of the cladding, and the RI of the GO film would increase more significantly than the RI increases with the increase of aqueous ethanol concentration by effectively captured ethanol molecules. More light would leak out of the fiber at the interface of the cladding and GO film compared to the interface of the cladding and aqueous ethanol. The GO coated U-bent fiber would further increase the sensitivity compared to the bare U-bent fiber. The concentration of ethanol can be determined by observing the light absorbance through the U-bent fiber.

3. Results and discussion

3.1. U-bent optical fiber properties

The U-bent probe of the sensor was observed under a SEM. The inset in Fig. 2(a) shows the SEM image of the U-bent probe. The U-bent probe is uniform and symmetrical with a 0.75 mm bend radius. The sensitivity and resolution of the U-bent fibers depend on their bend radius. The relationship between the sensitivity and the bend radius was investigated through the increase in sensitivity with the bend radius decreasing from 1.7 to 0.75 mm, and the decrease in sensitivity with the bend radius decreasing from 0.75 to 0.5 mm.²⁸ The sensitivity of the U-bent probe can be improved by applying its optimum bend radius. Hence, a U-bent probe with a 0.75 mm bend radius was selected for our experiment.

Fig. 2 SEM images of the side surface of (a) the unclad bare fiber and (c) after being coated with GO. (b) and (d) are magnified images of the rectangular areas marked in (a) and (c). Compared with the surface of the bare fiber in (b), GO film is clearly observed in (d). The inset in (a) is the SEM image of the unclad U-bent probe and its bend radius is 0.75 mm.

3.2. Characterization of the GO film

The images of the U-bent probe without and with GO film are shown in Fig. 2(a) and (c) respectively. It is obvious that the fiber has been coated with a GO film tightly and completely and the GO film has a uniform structure under high magnification. To further demonstrate the existence of GO, a contrastive U-bent probe with GO film is shown in Fig. 2(c). One can see that the surface appearance of the U-bent probe with and without GO film are obviously different, particularly the wrinkles that can be seen on the surface of the GO film. Fig. 2(b) and (d) are the magnified images of the areas marked with a yellow rectangle in Fig. 2(a) and (c) respectively; it is obvious that there are wrinkles on the surface of the GO film. In order to further characterize the GO film, Raman spectra were obtained from the GO coated U-bent fiber.

The Raman spectra of the GO film on the fiber were obtained using a 532 nm laser. As shown in Fig. 3(a), the D, G, 2D and S3 bands of GO are clearly seen. The D band (∼1360 cm⁻¹) is assigned to the ring vibration symmetrical breathing mode and associated with the defects caused by the attachment of hydroxyl and epoxide groups. The G band (∼1595 cm⁻¹) is assigned to the first-order scattering of the in-plane optical phonon E_2g mode and the 2D band (∼2722 cm⁻¹) is assigned to the second-order process involving two phonons with opposite momentum. The S3 band (∼2930 cm⁻¹) is caused by the imperfect activated grouping of phonons. The Raman spectra clearly demonstrate that high-quality GO films have coated on the U-bent probe.

Fig. 3 (a) Raman spectra of the GO film on the fiber excited with a 532 nm laser. (b) TEM image of the GO film.

The TEM image of the GO film is shown in Fig. 3(b), an edge can be seen at the point marked with a black arrow and a darker color with the increase of layers of GO film is also observed at the point marked with a red arrow. The TEM results indicate that the GO film has a uniform thickness and few defects.

3.3. The performance of the U-bent optical fiber sensor

The performance of the UOFS was studied by monitoring the absorbance using a spectrometer when both U-bent probes were put into aqueous ethanol with concentration from 5% to 100% at room temperature. The influence of background light was removed by setting the software pattern, and the light intensity with the GO coated and uncoated U-bent probes exposed to atmospheric air was used as a baseline. The absorbance spectra and dynamic absorbance response of UOFS without and with GO in a wavelength range of 500–800 nm are shown in Fig. 4(a) and (b). No matter with or without GO film, the absorbance increases with the increase of aqueous ethanol concentration. However, the absorbance increase of the UOFS without GO is imperceptible and indistinguishable. The absorbance variation for the UOFS with GO is significant: 5 times larger than that without GO. This phenomenon can be attributed to the tunable RI of the GO film with the concentration increase of the aqueous solution, which leads to a more significant absorbance variation. The sensitive RI change of the sensing area with GO film is introduced by the molecular enrichment properties of GO film. Based on these results, we can draw a conclusion that GO has excellent sensing ability due to its molecular enrichment role and the large-scale changes in RI in the sensing area. Importantly, the minimum and maximum detection limit of the developed sensor can reach 5% and 100% respectively, which indicates the developed sensor possesses high resolution and sensitivity.

Fig. 4 Absorbance spectra of the U-bent fiber sensor (a) uncoated and (b) coated in GO exposed to aqueous ethanol with a concentration from 5% to 100%. Dynamic absorbance response of sensor (c) uncoated and (d) coated in GO exposed to aqueous ethanol with a concentration from 5% to 100%.

Dynamic absorbance responses of the UOFS without and with GO are shown in Fig. 4(c) and (d) respectively. The absorbance vs. time changes with different concentration from 5% to 100% were obtained at room temperature. The detection time for each concentration is 10 s and the time interval between detections is 20 s. The UOFS with and without GO film all exhibit visible sensitivity and fast response and recovery when U-bent probes are put in the aqueous ethanol. The dynamic absorbance for the UOFS without GO film are distributed almost uniformly, while the dynamic absorbance for the UOFS with GO film gradually increases. It is obvious that the absorbance variation of the UOFS without GO film is indistinguishable with different concentrations of aqueous ethanol. The absorbance variation of the UOFS with GO film is 5% to 25% with the concentration variation of 5% to 100%. These results agree well with the absorbance spectra in Fig. 4(a) and (b). It also demonstrates that the UOFS with GO film can continuously detect aqueous ethanol with different concentration.

Table 1 presents the sensitivity of the developed sensor between different concentrations at the wavelength of 655 nm. It is important that the developed sensor fabricated in this work exhibits high sensitivity. It was found that the sensor exhibits a considerable sensitivity between a concentration range of 0% and 40%, which is superior to the reported aqueous ethanol detection or other methods for aqueous ethanol detection.²⁶ With increasing aqueous ethanol concentration, the sensitivity of the sensor declines. This verifies that the prepared GO can act as competitive sensing layer for aqueous ethanol detection once again.

Table 1 Sensitivity (defined as the slope of the curve given by S = ΔA/ΔC, where S is the sensitivity, A is the sensor absorbance in %, and C is the aqueous ethanol concentration in %) of the GO coated U-bent optical fiber sensor towards different aqueous ethanol concentrations. A wavelength of 655 nm was selected for the measurements

Aqueous ethanol concentration (%)	Absorbance (%)	ΔC	ΔA	Sensitivity = (ΔA/ΔC)
0	0	—	—	—
20	8.77	20	8.77	0.44
40	16.8	20	8.03	0.4
60	19.5	20	2.7	0.135
80	23.45	20	3.95	0.1975
100	25.3	20	1.85	0.0925

---

The response–recovery behavior is an important parameter for estimating sensing performance. The response and recovery times of the UOFS coated with GO film towards aqueous ethanol with concentration of 10% and 100% at room temperature were obtained, as shown in Fig. 5(a) and (b). For the concentration of 10%, the response time and recovery time of the developed sensor is 2 s and 1 s respectively. For a concentration of 100%, it is 2 s and 1 s respectively. The fast response may be due to ethanol molecule enrichment and the high surface-to-volume ratio of GO film.

Fig. 5 Typical response–recovery characteristic curves of the developed sensor exposed to aqueous ethanol with a concentration of (a) 10% and (b) 100% at room temperature. (c) The repeatability and reversibility test towards aqueous ethanol of different concentrations and air.

The repeatability and reversibility of the UOFS coated with GO film were investigated by continuously detection in two cycles. As shown in Fig. 5(c), the U-bent probe was continually exposed to aqueous ethanol with different concentrations under an air atmosphere. Obviously, the repeatability and reversibility of the developed sensor is very stable. Besides, the repeatable and steady results signals support that the GO film adhered tightly to the surface of the unclad U-bent probe. Based on this work, we present a cost-effective platform for aqueous ethanol detection with high sensitivity and repeatability.

In order to further investigate the relationship between absorbance and aqueous ethanol concentration, the correlation curves between the concentrations of aqueous ethanol are fitted at the selected wavelengths of 550, 600, 650 and 750 nm, as shown in Fig. 6. In all cases, the absorbance increases with the increase of the aqueous ethanol concentration from 5% to 100% in a logarithmic model. The coefficient of determination (R²) of the fit calibration curve for the four selected wavelengths reached 0.9846, 0.9858, 0.9881 and 0.9936, respectively. However, by focusing on the concentration range from 10% to 80%, as shown in the insets in Fig. 6, we can find a good linear relationship between the absorbance of the U-bent sensor and the concentration. The linear fitting curve can be expressed as y = 0.23x + 6.05, y = 0.223x + 5.93, y = 0.233x + 5.393 and y = 0.251x + 2.068 at the four selected wavelengths, respectively. To compare the sensing performance of the present sensor, the detecting range, linear response range and the limit of detection are compared with that of other aqueous ethanol sensors utilizing different approaches reported in the literature in the Table 2. By comparing the performance data in Table 2, besides the higher sensitivity, our sensor also has a wider detecting range and linear response range.

Fig. 6 The relationship between absorbance and aqueous ethanol concentration at a wavelength of (a) 550 nm, (b) 600 nm, (c) 650 nm and (d) 750 nm. A linear response range of 10–80%, with the absorbance as a function of concentration, is shown in the inset figures.

Table 2 Comparison of some basic parameters for aqueous ethanol sensors

Sensor type	Detection range	Linear response range	Limit of detection	Reference
Electronic	1.5–7.9 mM	1.5–7.9 mM	32 μM	29
Electronic	0–40%	0–40%	1%	30
Electronic	82.95–497.72 μM	82.95–497.72 μM	—	31
Optical	5–80%	—	5%	26
Optical	5–100%	10–80%	5%	Present study

---

Fig. 7(a) presents the stability testing of the developed sensor. The stability of the sensor was measured over 10 days. As can be seen, the absorbance of the sensor for aqueous ethanol of 50% concentration did not significantly change for at least 10 days, which indicates the good stability of the sensor, also for the GO film on the U-bent fiber. Besides, in order to investigate the feasibility of our sensor for a real sample, we applied our sensor to detect the ethanol content of Chinese liquor with a certified value of 42%. Then, we calculated the concentration using a linear fitting formula according to the absorbance value as shown in Fig. 7(b) and the result of calculation is 41.6%. The result indicates the potential of the GO coated UOFS for real sample detection.

Fig. 7 (a) The absorbance distribution of detection for aqueous ethanol with 50% concentration for a period of 10 days. (b) The real sample detection of Chinese liquor.

4. Conclusions

A novel UOFS coated with GO film for aqueous ethanol detection with high sensitivity, repeatability and stability was designed and demonstrated in this work. Comparing the sensing performance of the UOFS coated in GO film and uncoated UOFS, the higher precision of the sensor coated with GO film can be attributed to the excellent affinity of GO film to ethanol molecules. The detecting results showed that the U-bent fiber coated GO films exhibited higher sensitivity, wider detection range and faster response–recovery time over the tapered fiber coated GO films for aqueous ethanol detecting. The absorbance increases with the increase of aqueous ethanol concentration, and an excellent quadric relationship allows the application of a cost-effective platform for aqueous ethanol detection in a range from 5% to 100%. The response and recovery times are as short as 1–2 s, which endow the developed sensor with the ability of fast detection.

Acknowledgements

The authors are grateful for financial support from the National Natural Science Foundation of China (61205174, 61401258, 11504209 and 61377043), Shandong Province Natural Science Foundation (ZR2013HL049, ZR2013EMM009 and ZR2014FQ032) and the Excellent Young Scholars Research Fund of Shandong Normal University.

References

1. P. L. Jia and H. S. Wang, Sens. Actuators, B, 2013, 177, 1035–1042.
2. P. Dutta, I. Matias, C. Aramburu, A. Bakas, M. Lopez-Amo and J. M. Oton, Microw. Opt. Tech. Lett., 1996, 11, 93–95.
3. J. Villatoro, D. Luna-Moreno and D. Monzon-Hernandez, Sens. Actuators, B, 2005, 110, 23–27.
4. N. Nath, S. R. Jain and S. Anand, Biosens. Bioelectron., 1997, 12, 491–498.
5. R. Bharadwaj, V. V. R. Sai, K. Thakare, A. Dhawangale, T. Kundu, S. Titus, P. K. Verma and S. Mukherji, Biosens. Bioelectron., 2001, 26, 3367–3370.
6. P. H. Paul and G. Kouchakoff, Appl. Phys. Lett., 1987, 51, 12–14.
7. V. Ruddy, Fiber Integr. Opt., 1990, 9, 142–150.
8. V. Ruddy, B. D. MacCraith and J. A. Murphy, J. Appl. Phys., 1990, 67, 6070–6074.
9. S. K. Girei, A. A. Shabana, K. T. Arasu, S. Painam and M. H. Yaacob, IEEE Int. Conf. Photonics, 4th, 2013, 275–277.
10. F. B. Xiong and D. Sisler, Opt. Commun., 2010, 283, 1326–1330.
11. M. Yang, Y. Sun, D. Zhang and D. Jiang, Sens. Actuators, B, 2010, 143, 750–753.
12. M. Iga, A. Seki and K. Watanabe, Sens. Actuators, B, 2004, 101, 368–372.
13. D. Monzoon-Hernandez, V. Joel, D. Talavera and D. Luna-Moreno, Appl. Opt., 2004, 43, 1216–1220.
14. S. K. Srivastava, R. Verma and B. D. Gupta, Sens. Actuators, B, 2011, 153, 194–198.
15. A. K. Mishra, S. K. Mishra and B. D. Gupta, Opt. Commun., 2015, 344, 86–91.
16. A. Leung, P. M. Shankar and R. Mutharasan, Sens. Actuators, B, 2007, 125, 688–703.
17. Y. M. Wong, R. J. Scully, H. J. Kadim, V. Alexiou and R. J. Bartlett, J. Opt. A: Pure Appl. Opt., 2003, 5, S51–S58.
18. A. Kumar, T. V. B. Subrahmonium, A. D. Sharma, K. Thyagarajan, B. P. Pal and I. C. Goyal, Electron. Lett., 1984, 20, 534.
19. H. W. Qiu, S. S. Gao, P. X. Chen, Z. Li, X. Y. Liu, C. Zhang, Y. Y. Xu, S. Z. Jiang, C. Yang, Y. Y. Huo and W. W. Yue, Opt. Commun., 2016, 366, 275–281.
20. D. Littlejohn, D. Lucas and L. Han, Appl. Spectrosc., 1999, 53, 845–849.
21. B. D. Gupta, H. Dodeja and A. K. Tomar, Opt. Quantum Electron., 1996, 28, 1629–1639.
22. Y. Liu, X. Dong and P. Chen, Chem. Soc. Rev., 2012, 41, 2283.
23. I. V. Lightcap and P. V. Kamat, Acc. Chem. Res., 2013, 46, 2235–2243.
24. H. Ma, D. Wu, Z. Cui, Y. Li, Y. Zhang, B. Du and Q. Wei, Anal. Lett., 2013, 46, 1–17.
25. G. Eda, G. Fanchini and M. Chhowalla, Nat. Nanotechnol., 2008, 3, 270–274.
26. A. A. Shabaneha, S. H. Gireia, P. T. Arasua, W. B. W. A. Rahmanc, A. A. A. Bakarc, A. Z. Sadekd, H. N. Lime, N. M. Huangf and M. H. Yaacoba, Opt. Commun., 2014, 331, 320–324.
27. N. M. Huang, H. N. Lim, C. H. Chia, M. A. Yarmo and M. R. Muhamad, Int. J. Nanomed., 2011, 6, 3443–3448.
28. S. K. Khijwania and B. D. Gupta, Opt. Commun., 2000, 175, 135–137.
29. E. T. Hayes, B. K. Bellingham, H. Mark and A. Galal, Electrochim. Acta, 1996, 41, 337–344.
30. X. Xu, A. V. Goponenko and S. A. Asher, J. Am. Chem. Soc., 2008, 130, 3113–3119.
31. Y. C. Weng, J. F. Rick and T. C. Chou, Biosens. Bioelectron., 2004, 20, 41–51.

---