An optical fiber SERS sensor based on GO/AgNPs/rGO sandwich structure hybrid films

Abstract

In this work, we present a novel optical fiber SERS (OF-SERS) sensor based on a sandwich structure of GO/AgNPs/rGO. Rhodamine 6G (R6G) was selected as the probe molecule to compare the SERS ability of different films composed of the bare optical fiber, GO, AgNPs, rGO/AgNPs, GO/AgNPs and GO/AgNPs/rGO. Raman spectra, scanning electron microscopy (SEM), atomic force microscopy (AFM), energy dispersive spectroscopy (EDS) and X-ray photoelectron spectroscopy (XPS) were performed to characterize the sandwich structure hybrid films. Besides, the stability of GO/AgNPs and GO/AgNPs/rGO was compared on the optical fiber end face. The OF-SERS sensor based GO/AgNPs/rGO hybrid films had stable SERS performance even when exposed to ambient conditions for a prolonged time of 30 days. This work may provide a new strategy for an efficient stable OF-SERS sensor due to its simplicity, low-cost and long-term stability.

---

1. Introduction

Since the discovery of huge Raman signal enhancement on the surface of a rough silver electrode in 1974, surface enhanced Raman scattering (SERS) has attracted extensive attention.^1,2 Up to now, two types of the SERS substrate: traditional SERS substrates based on a flat plate and novel optical fiber SERS (OF-SERS) substrates have been proposed and demonstrated. Compared with the traditional SERS substrates, recently, OF-SERS substrates have been paid much attention due to their irreplaceable advantages such as in situ and remote detection, good flexibility and easy fabrication.^3–10 In the past years, OF-SERS substrates with noble metal had been fabricated for SERS detections.^5,11,12 However, the stability of these OF-SERS substrate is poor due to the signal variations of metal–molecule contact. What's more, the low adsorption capacity of metal for probe molecules also limits their applications. Thus, an ultrathin inert coating layer such as SiO₂ and Al₂O₃ has been demonstrated to obtain a SERS substrate with good stability and high sensitivity on traditional SERS substrate.¹³ In these cases, the coating layer should be thin enough to avoid the loss of electromagnetic enhancement.

Graphene, a single layer of sp² carbon network arranged in a perfect honeycomb lattice,¹⁴ is well known for its unique electrical performance and the amazing applications in nanoscale electronics.^15,16 GO is one of the graphene derivative nanomaterials, endowed with extraordinary characteristics and benefits, including high surface-to-volume ratio, excellent transparency, high thermal conductivity and so on. Furthermore, the functional groups of epoxy, carbonyl, and carboxylic acid endow GO better affinity than graphene in aqueous solution and organic solvents.¹⁷ Such single atomic and imporous structure of graphene and GO make them the natural candidate materials for metal nanostructures coating. Not only that, but graphene and GO can also act as molecule enricher to enhance the SERS signal due to the perfect bio-compatibility and restrain the back noise.^15,18 Many efforts have been done to employ graphene or GO as coating layer on traditional SERS substrate.^19–26 However, to the best of our knowledge, no relevant papers have been reported on the OF-SERS with graphene or GO as coating layer.

Herein, we provide an improved OF-SERS substrate based on a sandwich structure with GO/AgNPs/rGO. The GO film can effectively capture the probe molecules, the rGO film can act as passivation layer protecting AgNPs from oxidation in ambient condition and AgNPs can provide hot spots for the SERS due to its perfect SERS activity. The sandwich structure hybrid films were characterized by Raman spectra, scanning electron microscope (SEM), atomic force microscope (AFM) and energy dispersive spectroscopy (EDS) analyses. By optimizing the deposition time of AgNPs, we obtained the optimal parameter for the fabrication of OF-SERS substrate with well SERS performance. Using R6G as probe molecule, we characterized the SERS behavior of the prepared GO/AgNPs/GO OF-SERS substrate. The SERS results demonstrated that the proposed OF-SERS substrate is excellent in terms of the sensitivity, bio-compatibility and stability.

2. Experimental

The preparation procedure of the GO/AgNPs/rGO OF-SERS substrate is shown in Fig. 1. One end of standard multimode silica fiber (∼7 cm in length and 125 μm in diameter) was treated with acetone to remove the plastic cladding and set as sensing end. To guarantee the surface relatively uniform, both ends were cleaved carefully using optical fiber cleaver. As shown in Fig. 1(b), the sensing end of the prepared fiber was modified with GO thin film (prepared by the modified Hummers method²⁷) via a dip-coating method. Then, the modified sensing end was immersed into the mixture of reaction and R6G solution for 2 min to make R6G molecules absorbed on the GO film. Typically, reaction solution was prepared by mixing the diluted AgNO₃ solution (2 × 10⁻³ mol L⁻¹) and Na₃C₆H₅O₇ solution (2 × 10⁻³ mol L⁻¹) in a molar ratio of 1 : 1. Next, quantitative R6G powder was added into the mixed solution to prepare the mixed solution with certain concentration of R6G. Photochemical deposition process of AgNPs on the sensing end was carried out using a Raman spectrometer (Horiba HR-800, laser wavelength 532 nm and laser power 50 mW) after the adsorption of R6G. As shown in Fig. 1(a), the laser beam was introduced into the fiber using a 50× objective lens by focusing the laser beam on the core of the upper end and the power is 50 mW for the photochemical deposition of AgNPs. AgNPs will deposit on the surface of the sensing end when the laser beam introduced into the fiber.^28,29 Finally, a thin film of rGO (converted from GO by chemical and thermal treatment) was dip-coated on the GO/AgNPs film, eventually forming the sandwich structure of GO/AgNPs/rGO hybrid film as shown in Fig. 1(c). As comparisons, bare optical fiber, GO, AgNPs rGO/AgNPs and GO/AgNPs were also fabricated. Then, all fabricated OF-SERS substrates were kept in N₂ atmosphere before using. SERS spectra were collected using the Raman spectrometer mentioned above as shown in Fig. 1(a). A 10× objective lens was used and the laser output power was 5 mW for the remote detection on the distal end. All the SERS spectra were obtained for a scanning time of 4 seconds and accumulations of 2. It should be noted that the mixed solution in photochemical deposition process was removed in process of SERS measurement due to that the R6G has been already adsorbed on the prepared OF-SERS substrate before the photochemical deposition process of AgNPs.

Fig. 1 (a) Schematic diagram of the photochemical deposition process of AgNPs and the detection of SERS spectra using a Raman spectrometer. Schematic illustration of (b) the process for fabricating GO/AgNPs/rGO hybrid film and (c) sandwich structure of GO/AgNPs/rGO hybrid film on the optical fiber end face.

3. Results and discussion

In order to optimize the photochemical deposition time of AgNPs, we performed the SEM on the sensing end to investigate the morphological features of AgNPs deposited with different time. As can be seen in the SEM images in Fig. 2(a)–(e), also in the insert figures of the size distribution histograms of Ag nanoparticles obtained at different photochemical deposition times, with the increase of photochemical deposition time from 1 to 5 min, the number and size of AgNPs gradually increase. Only few punctate AgNPs are obtained with deposition time of 1 and 2 min and the gap between AgNPs is large, which is adverse for the SERS due to the loss of electromagnetic enhancement activity. With the increase of deposition time, as shown in Fig. 2(c), many spherical AgNPs appear with average diameter of ∼80 nm and the gaps between AgNPs decrease. However, with the further increase of deposition time from 4 to 5 min, AgNPs will bulky aggregate and form large particles with diameter from 200 to 500 nm which leading to the number decrease of hot spots and is also adverse for SERS activity. Based on the SEM analysis results, it is expected that the AgNPs obtained with 3 min are much more suitable for the optimum SERS activity.

Fig. 2 SEM images of AgNPs deposited on the sensing end with different photochemical deposition times of (a) 1, (b) 2, (c) 3, (d) 4 and (e) 5 min, the insert figures show the size distribution histograms of AgNPs obtained at different photochemical deposition times respectively.

To demonstrate the SERS behaviors of the obtained AgNPs with different photochemical deposition time and verify our expect, corresponding Raman spectra of R6G (10⁻⁴ mol L⁻¹) on OF-SERS substrates were obtained. One can see from Fig. 3(a) that the typical Raman peaks of the R6G at 613, 771, 1190, 1361 and 1647 cm⁻¹ are observed. The peak at 613 cm⁻¹ is assigned to the C–C–C ring in-plane vibration mode; the peak at 771 cm⁻¹ is assigned to the C–H out-of-plane bend mode; the peak at 1190 cm⁻¹ is assigned to the C–H in-plane bend mode; the peaks at 1361 and 1647 cm⁻¹ are assigned to the C–C stretching modes.³⁰ The intensity of the characteristic peak changed with the change of the photochemical deposition time, which could reflect the difference in the SERS activity for these deposition films. To compare the intensity evolution of Raman spectra, we measured the peak intensities at 613, 771 and 1361 cm⁻¹. As shown in Fig. 3(b), the intensities change as a function of photochemical deposition time in the same trend. The intensities at 613, 771 and 1361 cm⁻¹ both increase with the increase of photochemical deposition time from 1 to 3 min and saturates at the point of 3 min. Then, the intensities decrease gradually with the further increase of deposition time. These results are well agreed with the analysis results of SEM. Undoubtedly, we can conclude that the OF-SERS substrate modified with the AgNPs film possesses optimum SERS activity with a photochemical deposition time of 3 min.

Fig. 3 (a) Raman spectra of R6G (10⁻⁴ mol L⁻¹) were detected on OF-SERS substrates prepared with different photochemical deposition times. (b) Intensity at 613, 771 and 1361 cm⁻¹ changed as a function of photochemical deposition time.

With the optimized photochemical deposition time of 3 min, we prepared GO/AgNPs hybrid films and GO/AgNPs/rGO hybrid films on the sensing end. Fig. 4(a) shows the SEM image of GO/AgNPs hybrid film. After the coating of GO film, a similar surface structure on GO/AgNPs is detected compared with that of directly deposited AgNPs (Fig. 2(c)), where wrinkles of the GO film in white color can also be clearly seen. The SEM image of the GO/AgNPs/rGO film in Fig. 4(b) shows that AgNPs can still be observed even after the dip-coating of rGO film, which indicates that the coated rGO film is thin enough. The insert in Fig. 4(b) shows the AgNPs are covered by rGO film and form the sandwiche structure of GO/AgNPs/rGO, which can be seen from the darker color of the AgNPs compared with that of GO/AgNPs. To characterize the roughness change of GO/AgNPs film after dip-coating of rGO film, we analyzed the GO/AgNPs film and GO/AgNPs/rGO film using AFM. By comparing the line profiles of GO/AgNPs and GO/AgNPs/rGO film as shown in Fig. 4(c) and (d), we can find that the GO/AgNPs/rGO film possesses a smoother surface compared to that of GO/AgNPs films, which is beneficial from the presence of rGO film. Besides, the roughness average (R_a) of GO/AgNPs and GO/AgNPs/rGO film is respectively 75.0 nm and 54.3 nm. The successful preparation of GO/AgNPs/rGO film was confirmed by UV-vis absorbance spectra, as shown in the Fig. 4(e). GO displays a maximum absorption peak centered at 232 nm and a shoulder peak at about 300 nm. After the deposition of AgNPs, besides the two absorption peaks of GO, there is evidently a new peak at about 435 nm. This is the character of AgNPs due to the surface plasmon absorption. After the reduction, the peak at about 232 nm of GO will shift to about 271 nm. And, there are four peaks of GO/AgNPs/rGO film: 232, 271, 300 and 435 nm. These results indicate that rGO film has been coated on the GO/AgNPs film successfully.

Fig. 4 SEM images of (a) GO/AgNPs and (b) GO/AgNPs/rGO hybrid films. AFM images and line profiles of (c) GO/AgNPs and (d) GO/AgNPs/rGO hybrid films. (e) The UV-vis absorption spectra of GO, GO/AgNPs and GO/AgNPs/rGO films.

To investigate the superior in terms of SERS activity of GO/AgNPs/rGO, SERS spectrum of R6G (10⁻⁴ mol L⁻¹) was detected on the optical fiber end face. As a contrast, the SERS spectra of R6G on the bare optical fiber, GO, AgNPs, rGO/AgNPs and GO/AgNPs are also collected. As shown in Fig. 5(a), Raman signals of R6G were enhanced on both GO, AgNPs, rGO/AgNPs, GO/AgNPs and GO/AgNPs/rGO. However, the enhancing effect is varied with different substrates. We measured the peak intensity at 613, 771 and 1361 cm⁻¹ to compare the SERS activity of different substrates. The intensity of R6G at 613, 771 and 1361 cm⁻¹ was increased by ∼1.5 fold on the hybrid film rGO/AgNPs and ∼2.5 fold both on the hybrid film GO/AgNPs and GO/AgNPs/rGO compared to that on AgNPs in Fig. 5(b). And the intensity on AgNPs is higher than that on GO. The differences on the SERS activity can be due to the various SERS mechanisms. For the GO substrate, only chemical mechanism works for the SERS signal enhancement. For the AgNPs substrate, although the better enhancement is achieved compared with the GO substrate, the enhancement is still poor than that of rGO/AgNPs, GO/AgNPs and GO/AgNPs/rGO as only electromagnetic mechanism exists in the AgNPs substrate. The rGO/AgNPs has poor SERS activity compared with GO/AgNPs, which is may due to that the rGO has poor properties in terms of molecular adsorption due to that the rGO was converted from GO by chemical and thermal treatment, so the contents of the functional groups are much less than that of GO. In order to further demonstrate the feasibility of the prepared OF-SERS sensor, Raman spectra of R6G with different concentration from 10⁻⁷ to 10⁻⁴ mol L⁻¹ were detected on GO/AgNPs/rGO as shown in Fig. 5(c). The intensities of the SERS signals of the R6G declining with the decrease of the R6G concentration was also observed. Fig. 5(d) shows the reasonable linear response between the Raman intensity at 613, 771 and 1361 cm⁻¹ and the concentration of R6G, where the value of R² reaches 0.98, 0.98 and 0.92 respectively, indicating the concentration of R6G could be determined by measuring the intensity of the SERS signals. It is evident that the prepared GO/AgNPs/rGO OF-SERS sensor is an effective platform for the SERS molecular detection.

Fig. 5 (a) Raman spectra of R6G (10⁻⁴ mol L⁻¹) and (b) intensity at 613, 771 and 1361 cm⁻¹ on OF-SERS substrates prepared with different films of bare optical fiber, GO, AgNPs, rGO/AgNPs, GO/AgNPs and GO/AgNPs/rGO. (c) Raman spectra of R6G detected by GO/AgNPs/rGO OF-SERS substrate with concentration from 10⁻⁷ to 10⁻⁴ mol L⁻¹. (d) Intensity at 613, 771 and 1361 cm⁻¹ on GO/AgNPs/rGO OF-SERS substrate as a function of concentration.

To investigate the stability of the GO/AgNPs/rGO for OF-SERS, we measure the SERS spectra on the GO/AgNPs and GO/AgNPs/rGO films with designated durations of 1, 3, 6, 10, 15 and 30 days. Obviously, in Fig. 6(a) and (b), the SERS spectra intensities of R6G on GO/AgNPs film decrease in a great scale, while that on GO/AgNPs/rGO film relatively maintains well. As can be seen in Fig. 6(c), after 10 days exposed to ambient condition, the relative intensity at 613 cm⁻¹ on GO/AgNPs film decreased by 94.6 ± 2.8% but that on GO/AgNPs/rGO film decreased by 32 ± 5.8%. What's more, this tendency maintains well even after 15 and 30 days for the case of GO/AgNPs/rGO. To further investigate the passivation role of the rGO film for the AgNPs, the GO/AgNPs and GO/AgNPs/rGO are analyzed using EDS for designated durations of 1, 3, 6, 10, 15 and 30 days. During the process of the EDS spot measurements, we held all the parameters constant (Oxford X-Max 50 electric refrigeration and X-ray energy spectrometer operating at 18 kV, the spot size is 200 nm, magnification is 10 000 and acquisition time is 60 s). What's more, to guarantee the precision of the EDS data, we randomly select four different locations of AgNPs on each substrate. The Au content introduced by the spray-gold treatment is not counted and other parameters are maintained constantly. The EDS spectra of the GO/AgNPs and GO/AgNPs/rGO at the designated 1 and 30 days are shown in Fig. 6(d). One can see that the content of oxygen obviously increases on GO/AgNPs substrate and no obviously change of the oxygen content on GO/AgNPs/rGO substrate from 1 to 30 days. Besides, as shown in Fig. 6(e), the Ag/O ratio of GO/AgNPs and GO/AgNPs/rGO films is plotted versus time. The initial values of the Ag/O ratio are 0.23 and 0.20 respectively for GO/AgNPs and GO/AgNPs/rGO films. The initial value of the Ag/O ratio for GO/AgNPs film is higher than that for GO/AgNPs/rGO film which is due to the presence of the rGO, which increases the content of oxygen on GO/AgNPs/rGO film. And, the values of the Ag/O ratio decrease to 0.13 and 0.17 respectively for GO/AgNPs and GO/AgNPs/rGO after the designated duration of 30 days. The decrease of the values of the Ag/O ratio indicates the oxidation of the AgNPs on both substrates versus the time. However, the decrease of the Ag/O ratio is more distinct on GO/AgNPs film compared to that on GO/AgNPs/rGO film. Also, the XPS results indicate the more serious oxidation of AgNPs on GO/AgNPs film compared to that on GO/AgNPs/rGO film, since the Ag 3d peaks on GO/AgNPs film shifted to lower binding energy evidently compared to that on GO/AgNPs/rGO film (see Fig. S1 in the ESI†).³¹ The combined results of our experimental studies prove that the long-term stability of prepared GO/AgNPs/rGO OF-SERS substrate is benefit from the stable chemical property of rGO which can protect the AgNPs from oxidation under ambient condition. The high stability of SERS signal of R6G detection indicates that the GO/AgNPs/rGO OF-SERS sensor has great potential for practical applications.

Fig. 6 Raman spectra of R6G (10⁻⁵ mol L⁻¹) detected on (a) GO/AgNPs and (b) GO/AgNPs/rGO film both after exposed to ambient condition for designated durations of 1, 3, 6, 10, 15 and 30 days. (c) The relative intensities at 613 cm⁻¹ on GO/AgNPs and GO/AgNPs/rGO film vs. exposure time. (d) EDS spectra of GO/AgNPs and GO/AgNPs/rGO films at the designated 1 and 30 days. (e) The atomic ratio of Ag/O of GO/AgNPs and GO/AgNPs/rGO films vs. exposure time. The error bars indicate the average and standard deviation of 4 measurements at different locations on each substrate.

4. Conclusion

An improved SERS optical fiber sensor based GO/AgNPs/rGO hybrid films was successfully investigated and demonstrated. R6G was selected to investigate the SERS and the antioxidation ability of the developed antioxidative SERS sensor. According to the photochemical deposition time optimization, best SERS enhancement emerged at 3 min on AgNPs film in our experiment. The relative intensity of R6G at 613 cm⁻¹ on prepared SERS sensor decreased to 79 ± 8.61% compared to 5 ± 2% on GO/AgNPs film after exposed to ambient condition for 30 days, which duo to the protection effect of rGO film. This work may offer a novel and practical method to fabricate efficient, low-cost and long-term stable SERS substrate on optical fiber end face.

Acknowledgements

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

References

1. M. Fleischmann, P. J. Hendra and A. J. Mcquillan, Chem. Phys. Lett., 1974, 26, 163–166.
2. D. L. Jeanmaire and R. P. Van Duyne, J. Electroanal. Chem., 1977, 84, 1–20.
3. C. Viets and W. Hill, Sens. Actuators, B, 1998, 51, 92–99.
4. E. J. Smythe, M. D. Dickey, J. Bao, G. M. Whitesides and F. Capasso, Nano Lett., 2009, 9, 1132–1138.
5. T. Liu, X. Xiao and C. X. Yang, Langmuir, 2011, 27, 4623–4626.
6. X. Yang, N. Ileri, C. C. Larson, T. C. Carlson, J. A. Britten, A. S. Chang, C. Gu and T. C. Bond, Opt. Express, 2012, 20, 24819–24826.
7. H. Sato, T. Tanaka, T. Ikeda, S. Wada, H. Tashiro and Y. Ozaki, J. Mol. Struct., 2001, 598, 93–96.
8. Y. S. Yamamoto, Y. Oshima, H. Shinzawa, T. Katagiri, Y. Matsuura, Y. Ozaki and H. Sato, Anal. Chim. Acta, 2008, 619, 8–13.
9. H. Mitsuoka, S. I. Morita, T. Suzuki, Y. Matsuura, Y. Katsumoto and H. Sato, Appl. Phys. Express, 2009, 2, 027001.
10. X. D. Wang and O. S. Wolfbeis, Anal. Chem., 2012, 85, 487–508.
11. C. Y. Liu, S. Y. Wang, C. C. Fu, H. B. Li, S. P. Xu and W. Q. Xu, Chem. Res. Chin. Univ., 2015, 31, 25–30.
12. X. L. Zheng, D. W. Guo, Y. L. Shao, S. J. Jia, S. P. Xu, B. Zhao and W. Q. Xu, Langmuir, 2008, 24, 4394–4398.
13. J. F. Li, Y. F. Huang, Y. Ding, Z. L. Yang, S. B. Li, X. S. Zhou, F. R. Fan, W. Zhang, Z. Y. Zhou, D. Y. Wu, B. Ren, Z. L. Wang and Z. Q. Tian, Nature, 2010, 464, 392–395.
14. K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, Y. Zhang, S. V. Dubonos, I. V. Grigorieva and A. A. Firsov, Science, 2004, 306, 666–669.
15. A. K. Geim and K. S. Novoselov, Nat. Mater., 2007, 6, 183–191.
16. Z. Chen, A. Umar, S. W. Wang, Y. Wang, T. Tian, Y. Shang, Y. Z. Fan, Q. Qi, D. M. Xu and L. Jiang, Nanoscale, 2015, 7, 10259–10266.
17. A. Bagri, C. Mattevi, M. Acik, Y. J. Chabal, M. Chhowalla and V. B. Shenoy, Nat. Chem., 2010, 2, 581–587.
18. X. Ling, L. Xie, Y. Fang, H. Xu, H. Zhang, J. Kong, M. S. Dresselhaus, J. Zhang and Z. Liu, Nano Lett., 2010, 10, 553–561.
19. X. Li, J. Li, X. Zhou, Y. Ma, Z. Zheng, X. Duan and Y. Qu, Carbon, 2014, 66, 713–719.
20. P. Wang, O. Liang, W. Zhang, T. Schroeder and Y. H. Xie, Adv. Mater., 2013, 25, 4918–4924.
21. S. C. Xu, S. Z. Jiang, J. H. Wang, J. Wei, W. W. Yue and Y. Ma, Sens. Actuators, B, 2016, 222, 1175–1183.
22. S. Chen, X. Li, Y. Zhao, L. Chang and J. Qi, Carbon, 2015, 81, 767–772.
23. Y. Zhao, W. C. Zeng, Z. C. Tao, P. H. Xiong, Y. Qu and Y. W. Zhu, Chem. Commun., 2015, 51, 866–869.
24. L. L. Zhang, C. L. Jiang and Z. P. Zhang, Nanoscale, 2013, 5, 3773–3779.
25. A. P. Liu, T. Xu, J. Tang, H. P. Wu, T. Y. Zhao and W. H. Tang, Electrochim. Acta, 2014, 119, 43–48.
26. X. Gong, J. Tang, Y. X. Ji, B. B. Wu, H. P. Wu and A. P. Liu, RSC Adv., 2015, 5, 42653–42662.
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. E. J. Bjerneld, K. V. G. K. Murty, J. Prikulis and M. Käll, ChemPhysChem, 2002, 3, 116.
29. E. J. Bjerneld, F. Svedberg and M. Käll, Nano Lett., 2003, 3, 593.
30. P. Wang, O. Liang, W. Zhang, T. Schroeder and Y. H. Xie, Adv. Mater., 2013, 25, 4918–4924.
31. G. I. N. Waterhouse, G. A. Bowmaker and J. B. Metson, Appl. Surf. Sci., 2001, 183, 191–204.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra16869h

---