1. Introduction

Graphene based surface-enhanced Raman scattering (SERS) substrates have drawn a tremendous amount of attention recently, mainly resulting from the biocompatibility, corrosion resistance and chemical atomic surface test bed with the small-distance charge transfer between graphene surface and the adsorbed molecules, making the Raman signal more reliable and efficient [1–7]. Graphene with the large specific surface area of 2630 m²/g can also work as molecule enricher in SERS-active substrate, which could act as an excellent adsorbent towards organic molecules, especially the aromatic molecules [1–7].

To improve the performance of graphene towards high sensitive SERS applications, various hybrid structures of graphene and metal nanoparticles have been widely investigated due to the utilizable combined advantages [8,9]. Although, 2D graphene-based SERS hybrid substrates have been demonstrated very excellent SERS performance in regard to sensitivity and reproducibility, three-dimensional (3D) graphene based SERS substrates (called 3D SERS substrates) cannot only offer ultra-high surface area for the deposition of a large number of nanoparticles and the absorption of plenty of analyte molecules, but also achieve the full utilization of incident laser, which are the advantages that 2D graphene do not own. Besides, abrasion resistance, extensibility and hygroscopicity make the 3D graphene preferred in the practical application. Building of 3D SERS substrates has been the recent topic of intense study in SERS-based analysis.

For example, intentionally introducing crumples in flat graphene by means of inducing delamination and buckling of graphene [10,11] resulting from the stiffness mismatch of thin films on shrinking polymer substrates [12,13] to enhance the SERS performance of graphene-based substrates has been proved. But it missed the utilization of full 3D focal volume. Just recently, Chavis Srichan et al. developed AgNPs/GF (Ag nanoparticles/ graphene foam) as a novel SERS substrate to detect MB [14]. However, DC magnetron sputtering that was used to deposit silver nanoparticles was inefficient to form the all standing to GF branches due to the sags and crests of GF branches. Besides, the rigid substrates limit the application potential of AgNPs/GF under real-world conditions. It is necessary to develop a simple, low-cost and effective way to fabricate 3D flexible and sensitive graphene-based SERS substrate.

Food safety problem has penetrated into all aspects of social life and become the focus topic of common people [15]. Undeniably, a wide range of chemicals, such as antibiotics, pesticides, hormones, and anesthetics, were applied illegally to control or prevent disease outbreaks, which bears the brunt of aquaculture safety problems. Crystal violet (CV) [Fig. 11(b) in the Appendix], as a typical example, is forbidden but still widely used in aquaculture operation due to its low cost and high effectiveness against microbial-, fungal-, and parasitic-related fish diseases [16]. In recent year, SERS technology as an emerging detection means has drawn more and more attention. Although many studies clearly demonstrate that SERS substrates hosting closely separated metal nanostructures result in the large enhancements, the efficiency of sample collection in real-world applications, which is often overlooked, becomes the most urgent affair.

Here we develop a flexible SERS-based 3D AgNPs/G@Ni foam hybrid structure with high adsorptivity, sensitivity and reproducibility, which is composed of foam-like graphene macrostructure obtained by template-directed chemical vapour deposition on the nickel foam and uniform and high-density Ag nanoparticles wrapping around the foam graphene, via seed-mediated in situ growth process. (Fig. 1). On the basis of strong support from the mechanically robust Ni foam, which can avoid the collapse of 3D foam structure after the assemblage of silver nanoparticles on the surface of graphene skelecton, the homogeneous and high–density growth of silver nanoparticles spreading across the all surface area of the interior pore walls of the foam graphene was achieved via seed-mediated in situ growth process, which is simple, low-cost and fast. 3D graphene foam by template-directed chemical vapour deposition on the Ni foam plays multiple roles such as molecule adsorber [5] and stabilizer [6] and photoluminescence quencher [7]. Besides, the existence of thin graphene film improves the distribution uniformity of Ag nanoparticles later, contributing to the small fluctuations of different positions from the different batches and different branches of a same SERS substrate. And, the 3D focal volume and high-density and homogeneity hot spots in such 3D AgNPs/G@Ni foam create high detection sensitivity.

 

Fig. 1 (a) The schematic illustration of preparing the 3D G@Ni foam structure. (b) Processes of in situ growth of silver nanoparticles from electroless-deposited seeds in G@Ni foam. (c) The details of constructing AgNPs/G@Ni foam.

Download Full Size | PDF

Moreover, because the 3D substrate with special poriferous foam structure is flexible, bending and deformation almost has no influence on the integrality of the substrate, which is in urgent need for the in situ detection of SERS substrates in the practical application. So one of the distinct advantages of obtained AgNPs/G@Ni substrate due to high adsorptivity and flexibly is the ability to collect trace amount of analytes from real-world surfaces by swabbing across the surface by using such flexible 3D AgNPs/G@Ni foam, and we achieved the detection of CV molecules with a low concentration of 0.4ng/mL, which exhibited great application potential in food safety filed to achieve ultra-low trace for drugs on food surfaces.

2. Experimental section

2.1 3D AgNPs/G@Ni foam formation

To fabricate G/Ni foam, a scaffold of porous nickel foam is applied as a template for the deposition of carbon atoms. The large-area and multilayer graphene was synthesized by CVD technology as described in previous work [17–19]. Briefly, after the pressure of quartz tube was pumped to under 10⁻³ Pa by a double-pump system, H² with the flow of 50 sccm was piped in the tube to remove the remaining nickel oxide when the temperature reached 300 °C. The nickel foam was annealed for 10 min with flowing hydrogen at 1050 °C to increase the copper grain size. Then, the growth process of graphene was carried out with the mixed flowing of H₂ (50 sccm) and CH₄ (50 sccm). After that, The G/Ni foam successively was immersed in acidic aqueous solution of SnCl₂ (0.2M) and AgNO₃ solution (0.2M) for two minutes to generate silver seeds embedded in the spongy G/Ni foam and then was rinsed by acetone and pure water respectively. This process was carried out three times to improve the density of silver seeds on the G/Ni foam. Subsequently, the silver seeds on the G/Ni foam then grew into larger Ag nanoparticles by immersing the foam into the mixed solution of AgNO₃ and ascorbic acid for a certain time, followed by air-drying.

2.2 Characterization Of 3D AgNPs/G@Ni foam

Scanning electron microscope (SEM, Zeiss Gemini Ultra-55) was conducted to elucidate the surface morphology of the AgNPs/G@Ni foam substrate. SERS spectra were collected with a Horiba HR Evolution 800 Raman microscope system (laser wavelength: 532nm, laser spot: 1μm) on the same conditions. The TEM images of GO were observed with a transmission electron microscopy system (JEOL, JEM-2100).

3. Results and discussions

3.1 Preparation and characterization of AgNPs/G@Ni foam

Nickel foam, a porous structure with an interconnected 3D scaffold of nickel [Fig. 2(a)] [20], was chosen as a template for the growth of GFs. As shown in Fig. 2(b), the raw nickel foam surface consists of closely connected nickel block body which was dotted with large numbers of patches of gray. Figure 1 illustrates the CVD reactor used for the synthesis of graphene foam in this work. In brief, carbon was introduced into a nickel foam by decomposing CH₄ at 1050 ^◦C under ambient pressure by pumping in a mixture of gases with a composition of CH₄: H₂ = 50: 50 sccm, and graphene films were then precipitated on the surface of the nickel foam. From the SEM images of scaffold of porous G@Ni foam shown in Fig. 2(c), after 5-min growth, original nickel foam surface was successfully wrapped by a thin graphene film completely, where the surface of nickel block body become smooth and uniform but the seams between nickel block body still are clearly visible. The graphene film growing on the entire surface of the nickel scaffold are interconnected with each other and there is no interface or physical breaks in the framework and thus they inherits the 3D interconnected scaffold structure of original nickel foam template, which providing more adsorption sites for the AgNPs and the detected organic molecules. When the growth time was extended to 10 min, thicker graphene film was deposited on the surface of nickel foam. As shown in Fig. 2(d), ripples and wrinkles became obvious on the graphene films with the increase of graphene thickness because the thermal expansion coefficients of nickel and graphene was different [21,22].

 

Fig. 2 SEM images of foam Ni sample (a) and at larger magnification (b). SEM images of foam Ni sample after graphene growth for 5 minutes (c) and for 10 minutes (d). Inset: the corresponding TEM images of foam graphene samples. (e) and (f) Raman spectra of graphene collected from the samples shown in (c)and (d) respectively.

Download Full Size | PDF

The Raman spectra [Figs. 2(e) and 2(f)] were randomly collected from the substrates displayed in Fig. 2(c) and 2(d) respectively. The spectra reveal the presence of D, G, and 2D bands at 1350, 1590 and 2710 cm⁻¹, respectively, which are concordance with sp², hybridized graphitic structures [23–25]. And the ratio of the peak intensities of the G peak to 2D peak in Fig. 2(c) is about two to three, confirming the presence of three- to four-layered graphene [26]. While the ratio of peak intensities of the G peak to 2D peak in Fig. 2(c) is as high as four to six, proving the presence of multilayered graphene [27]. This conclusion is further proved by the HRTEM analysis of folds at the edges of 3D foam graphene sheet, which gives a visualized exhibition about the number of graphene layers, where the number of dark lines represents the number of graphene layer. As shown in the insets of Figs. 2(c) and 2(d), the HRTEM images derived from the edge of graphene sheet exhibit four dark lines and seven dark lines, indicative of four-layered and seven-layered graphene.

The embeddedness fulfillment of silver nanoparticles in the surface of G@Ni foam structure is required for suitable preparation technology. Traditional preparation methods for metal nanoparticles such as high-temperature annealing of thin metal films [28] and self-assembling process of metal collosol [29]etc. are just limited to flat surface substrates, which is inapplicable for 3D foam substrates. In our work, a two-steps growth strategy including silver seed deposition and in situ growth of silver nanoparticles was adopted [Fig. 1(b)]. Firstly, by immersing the G@Ni foam in an acidic aqueous solution of SnCl₂ (0.1M) for two minutes, Sn²⁺ in the solution aggregates and deposits on the graphene scaffold from all quarters to achieve the all-directional cover. The G@Ni foam were rinsed in pure water and in acetone respectively and dried. Subsequently, the G@Ni foam soaked in aqueous solution of AgNO₃ (0.02M) for the same time to replace Sn²⁺ on the pore walls with silver seeds depositing on the surface of graphene scaffold via the following reduction reaction of Sn²⁺(aq)+2Ag⁺ → Sn⁴⁺(aq)+2Ag(s). After that, Silver seeds were successfully immobilized along the surface of graphene scaffold. Finally, the in situ growth process of silver seeds on the walls of the 3D graphene foam substrates was carried out to further optimize the size and particle-particle space for improving the SERS performance of samples by immersing the substrate in the mixed solution of 0.01mM AgNO₃ and 0.1M ascorbic acid for a certain time. During this in situ growth process, silver seeds served as cores inducing the epitaxial growth of Ag shell to form larger silver nanoparticles. So, improving the distribution density of silver seed on the surface of graphene scaffold was very necessary for the formation of final high-density silver nanoparticles in the in situ growth process. The attempt of increasing the numbers of deposition part to change the distribution density of Ag seeds was made. The SEM images shown in Figs. 11(a)-11(d) exhibit the variations of distribution of Ag seeds located on the foam branches which are generated with four different depositing times of one, two, three and four times respectively. The gaps between AgNPs are too large for the substrate to be SERS-active owing to insufficient SERS hotspots as shown in Figs. 11(a) and 11(b) deposition for process with one and two times. After the deposition process of silver seeds is carried out four times [Fig. 11(d)], the density of Ag seeds significantly increase, but the content of stack and agglomeration increases likewise, which is also disadvantage for high-performance SERS substrates due to the relative lack of hot spots. Compared with the other three conditions, the Ag nanoparticle seeds on the surface of foam are dense and uniform almost without any agglomeration after three-time deposition, shown in Fig. 11(c). So this substrate is of the greatest potential to exhibit the excellent SERS ability. To verify the conclusion obtained through analyzing the SEM images, SERS signal intensity mappings of the R6G molecules at 613cm⁻¹ with concentration of 10⁻⁵M are detected as shown in the inset of Figs. 11(a)-11(c), The brighter color and darker color are corresponded to higher and lower intensity of the SERS signal respectively. It is obvious that the SERS signal of R6G molecules on the inset of Fig. 11(a) is extremely weak due to the too little distribution of metal nanoparticles. With the increase of deposition times, color of SERS mapping images is more and more bright. In contrast, Raman mapping collected from the substrates shown in Fig. 11(c) exhibit much brighter color distribution, which is coincide with the above SEM analysis.

The size of Ag seeds is too small to display satisfying SERS performance. So adjusting the size of Ag nanoparticles to improve the SERS performance of substrate is very necessary. Here, we explored the size controllability of decorated AgNPs by placing Ag seeds/GF substrates into a growth solution consisting of 0.1M ascorbic acid solution and 0.01M AgNO₃ solution for further chemical reduction growth. After the G@Ni foam substrate deposited Ag seeds via three times was immersed in the growth media for the further in situ reduction growth for different time, R6G was applied as probe molecule to measure the SERS performance of AgNPs/G@Ni foam. The characteristic Raman peaks of R6G deposited on the substrates are detected in the region from 500 to 1800cm⁻¹ under the same measurement condition for the 10⁻⁵M concentration. The SERS performance of AgNPs/G@Ni foam versus in situ growth times was recorded in Figs. 3(a) and 3(b). Obviously, with increasing in situ growth time, the SERS signal intensity gradually falls after transiently rising. Through comparing the relative Raman intensities of each SERS substrate, the substrate containing silver nanoparticles grown for 3-5 min was observed to provide the largest SERS effect, which is most potential to be the high-performance SERS substrate applied in the practical condition. Morphology of AgNPs/G@Ni foam after 4-min growth is recorded in the SEM images shown in Figs. 3(d)-3(f). Figure. 3(e) is the enlarged view of one random branch of AgNPs/G@Ni foam shown in Fig. 3(d). The uniform and dense Ag nanoparticles with average diameter of ∼70 nm distribute on the surface of all branches. Figure. 3(c) shows EDS spectrum from the AgNPs/G@Ni foam. All the peaks are associated to Ag and Ni element, demonstrating a existence and high purity of Ag particles. (Due to the existence of crinkly nature, Ag nanoparticles deposited on the thick foam graphene layer were apt to form stack and agglomeration structure, which is undesirable as SERS substrate [Fig. 12(a) in the Appendix].

 

Fig. 3 (a) Raman spectra of R6G on AgNPs/G@Ni substrates at various silver nanoparticle growing periods. (b) Relative Raman intensity of 612 cm⁻¹ at different growth time. (d)-(f) SEM images of AgNPs/G@Ni substrates under different magnification.

Download Full Size | PDF

3.2. SERS activity of AgNPs/G@Ni foam

To demonstrate the SERS sensitivity of optimal substrate, SERS spectra of R6G molecules from AgNPs/G@Ni foam with various concentrations of 10⁻⁵ to 10⁻¹⁰M are shown in Fig. 4(a). The SERS signal intensity gradually decreases with decreasing of molecule concentration. The SERS characteristic signals of R6G molecules still can be detected even at 10⁻¹⁰M. To investigate the capability of AgNPs/G@Ni foam to detect R6G molecules quantitatively, the linear fit calibration curves corresponding to concentration are illustrated in Fig. 4(b). An excellent linear response in log scale that the SERS signal intensity of characteristic peak (613, 1365 and 1510.5 cm⁻¹) versus the R6G concentration with concentration from 10⁻⁵ to 10⁻¹⁰M is obtained with the high coefficient of determination (R² = 0.987, 0.968 and 0.972).

 

Fig. 4 (a) Raman spectra of R6G with different concentrations from 10⁻⁵ to10⁻¹⁰M. (b) Raman intensity of R6G at 613, 1365 and 1510.5 cm⁻¹ as a function of the R6G concentration.

Download Full Size | PDF

3.3 Reproducibility of the AgNPs/G@Ni foam

It is well known that high homogeneity and reproducibility for SERS-based substrates is important in practical application as a real-time analytical tool [17]. So in our wok, the SERS spectra of R6G molecules at 10⁻⁵M on the joints from 6 different branches in the same AgNPs/G@Ni substrate were collected [Fig. 5(a)]. The profile of Raman bands of R6G from different branches is very similar and the unified major Raman characteristic peaks distribution with tiny change in the Raman intensity indicates that the prepared AgNPs/G@Ni SERS substrate possesses well reproducibility for SERS signal in the large area scope. To obtain a statistically meaningful result, we carried out the statistics about the branch-to-branch intensity distribution of the characteristic peak 613cm⁻¹ of R6G at the concentration of 10⁻⁵M presented in Fig. 5(b). The red line presents the average intensity of Raman spectra measured from 8 different branches. It is obviously that fluctuation about the data around the average value is slight with the RSD of only 10.55%.

 

Fig. 5 (a) SERS spectra were collected from 6 different batches of AgNPs/G@Ni substrates. (b) Intensity distribution of the 612 cm−1 peak in the 8 branches from a same AgNPs/G@Ni substrate. (c) SERS mapping of vibration modes at 612 cm−1 of R6G molecules at 10-5 M dispensed on the AgNPs/G@Ni foam after growing for three minutes. (d) Average SERS spectrum of the R6G molecules from 8 positions on a same branch of AgNPs/G@Ni substrate (red line).

Download Full Size | PDF

Besides, we collected the Raman mapping of R6G at 613cm⁻¹ with concentration of 10⁻⁵M on the branch of AgNPs/G@Ni substrate over 10 × 10μm area with a step of 2.5μm to evaluate the SERS signal homogeneity. The uniform and bright color distribution on the mappings shown in Fig. 5(c) demonstrates the homogenous distribution of both the analyte molecules and the hotspots on the AgNPs/G@Ni foam. Raman mappings of R6G with concentration reduced at 10⁻¹⁰M were also measured on the AgNPs/G@Ni foam [Fig. 12(b), in the Appendix]. Due to the lack of analyte molecules deposited on the branches, the spatial coincidence of the molecules with the hot spots of the AgNPs/G@Ni foam is hard to come by which lead to the not uniform and not light mapping image.

Meanwhile, the SERS spectra of R6G molecules at 10⁻⁵M from 8 different batches of AgNPs/G@Ni foam also are displayed in Fig. 5(d). The red curve indicates the average SERS spectrum of the 8 spectra measured on substrates from different batches. The green shaded area wrapping with the red curve was formed by the 8 spectra which overlap with each other and the width of the area can reflect the intensity deviation of all the SERS spectra intuitively. As shown in Fig. 5(d), shaded area is narrow indicating the minor intensity deviation of various vibration modes of the SERS spectra and proving the excellent homogeneity of the AgNPs/G@Ni foam.

3.4 Simulation results and comparison analysis

To further verify the SERS activity of AgNPs/G@Ni foam substrates, we compared the performance of AgNPs/G@Ni foam substrates with that of AgNPs@Ni foam, AgNPs/G@Ni plane, graphene and Si substrates. By maintaining the same measurement conditions, the CV aqueous solutions with the concentration of 10⁻⁶M were used as the probe molecule. Figure. 6(b) shows the SERS spectra of CV molecules at 10⁻⁶ M from six positions in the single branch of AgNPs@Ni foam substrates. In the SERS spectrum, many peaks showed at about 424, 725, 806, 915, 1053, 1174, 1371, 1538, 1587 and 1619 cm⁻¹ are observed, which correspond to the Raman signals of CV [18]. The profle of Raman bands observed in SERS spectra of CV from different positions is very similar, but change in the Raman intensity is significant with the RSD of 24.08%, indicating low reproducibility. To obtain the statistical data, the spot-to-spot intensity variation of the Raman characteristic peak 915 cm⁻¹ is quantified and displayed in Fig. 6(d). The wide intensity distribution still demonstrates that the SERS signals from AgNPs@Ni foam substrates have poor reproducibility. In contrast, the quality of SERS signals in terms of reproducibility and signal-to-noise ratio was improved on the AgNPs/G@Ni foam substrate. As shown in Figs. 6(a) and 6(c), the fluorescence background and intensity distribution in the SERS spectra are obviously reduced with the RSD of only 10.02%. Because of the additional graphene layer, the well-ordered AgNPs are produced on the G@Ni foam substrate and the CV molecules direct contacting with Ni substrate can be avoided. The narrow intensity distribution demonstrates the homogenous distribution of both the analyte molecules and the hot spots on the AgNPs/G@Ni foam. The uniform distribution of absorbed molecules and hot spots can be attributed to the well-ordered structure of AgNPs [Figs. 3(d)-3(f)] and good affinity of graphene for target molecules [19]. The formation of well-ordered structure of AgNPs benefits by flat graphene film which provides homogeneous ground for fostering well-ordered AgNPs.

 

Fig. 6 (a) SERS spectra were collected from 6 randomly selected spots on the single branch from the AgNPs/G@Ni foam. (b) SERS spectra were collected from 6 randomly selected spots on the single branch from the AgNPs @Ni foam. (c) Intensity distribution of the 912cm⁻¹ peak in the 6 spectra shown in Fig. 6(a). (d) Intensity distribution of the 915cm⁻¹ peak in the 6 spectra shown in Fig. 6(b)

Download Full Size | PDF

By maintaining the same measurement conditions, the R6G aqueous solutions with the concentration of 10⁻⁶M were used as the probe molecule to investigate the SERS ability of different substrates shown in Fig. 7(a). Higher SERS signals intensity on graphene substrates with and without AgNPs was given than those of corresponding silicon substrates. In the case of Si substrate, the SERS spectra of R6G can be negligible. As for the graphene substrate, because its SERS enhancement mainly results from a chemical mechanism (CM), the peak intensity of R6G is quite weak. Obviously, the foam AgNPs/G@Ni substrate provides significantly higher SERS enhancement compared with plane AgNPs/G@Ni substrates due to geometric effects. As for the explanation about SERS enhancement effect of foam AgNPs/G@Ni, multiple cascaded amplification mechanisms (localized surface plasmon mode and optical standing waves [30,31] or optical refraction [30,32]) of the 3D foam structures is widely accepted just shown in Fig. 7(b). The photon incident on an AgNP will be first amplified by localized surface plasmon mode and then further intensified by the charge transfer on the surface of graphene. The scattering process is over for the plane AgNPs/G@Ni substrates. In contrast, for foam AgNPs/G@Ni structure, the scattered photon derived from above processes will be further scattered and amplified through other interfaces of AgNP/G@Ni and until it finally exits to the photodetector, which effectively makes the incident laser oscillate among the branches so as to give rise to local enhancement of the incident laser and improve the response rate.

 

Fig. 7 (a) Comparison of SERS intensity of 10⁻⁵ M R6G on different SERS substrates including AgNPs/G@Ni foam, AgNPs/G@Ni plane, graphene and Si. (b) Amplification scheme for AgNPs/G@Ni foam.

Download Full Size | PDF

In the foam AgNPs/G@Ni structure, nickel foam as an important part in the whole product was used as the support to keep three-dimensional construction. But this is not all. In order to give a better understanding for the G@Ni foam as the platform, we calculated the electric field distributions of AgNPs/G@Ni, AgNPs/G and AgNPs/Si structure using comsol analysis (laser: 532nm, incident angle: 90°). The radius of AgNPs was set as 110 nm with a gap of 20 nm, by reference to scanning electron microscope (SEM) images and each silver nanoparticle was considered a spherical shape for the 3D COMSOL analysis

It is well known that graphene belongs to a chemical mechanism (CM), which is further demonstrated by the electric field distributions analysis again shown in Fig. 8, where graphene make nearly no contribution to electric field enhancement. In contrast, when Ni is used as the substrate supporting the AgNPs/G structure, electric field intensity around the AgNPs increases by leaps and bounds, which can attribute to intense resonant coupling phenomenon between AgNPs and nickel substrate with only nanoscale-level interval (thickness of four-layer-graphene).

 

Fig. 8 (a-c) are respectively the y-z x-y views of the electric feld distribution on the AgNPs/G@Ni, AgNPs/G and AgNPs/Si substrate.

Download Full Size | PDF

Summary, the excellent enhancement behavior and high reproducibility of the SERS signals from AgNPs/G@Ni foam can be attributed to at least three aspects: first, the special poriferous foam structure is conducive to oscillation and amplification the incident laser, giving rise to the local enhancement and improved collection of the laser. Second, the homogeneous graphene foam serve as the abundant attachment points for the Ag nanoparticles and analysis molecules, which is beneficial for the spatial coincidence of the analyte molecules with the hot spots. Third, three-dimensional and high-density hot spots deriving from the well-ordered silver nanoparticles array and intense resonant coupling between AgNPs and nickel substrate exist.

3.5 Practical applicability of the AgNPs/G@Ni foam

Thanks to the flexibility and porosity of AgNPs/G@Ni foam, it has the potential to absorb analyte on the arbitrary surface as completely as enough, just exhibited in Figs. 9(a) and 9(d) thus can be used in actual application. Crystal violet (CV) [Fig. 9(b)] powder was mixed with deionized water to prepare CV standard solutions of different concentration. The cleaned fish was sprayed the new-prepared CV standard solution and conducted SERS detection after a week under natural condition. Prior to the detection of CV, the fish surface was dropped ethanol on to disassociate the CV molecules from the fish and then covered the flexible and porous AgNPs/G@Ni substrate to swab the CV solution [Figs. 9(a) and 9(c)]. The process of collecting CV molecules was conducted on the whole surface of the fish, which can creatively achieve gathering of CV molecules in the detection substrate to create concentration effect. At the same time, the dried fish with the CV was directly exposed to the Raman laser and was performed the Raman detection.

 

Fig. 9 (a) and (d) Schematic of swabbing and concentrating process. (b) Chemical structure of CV. (c) The photo of swabbing process.

Download Full Size | PDF

The creative and convenient method of swabbing the CV molecules on the surface of fish, which can largely avoid the disturbances of fluorescent background arising from the fish and is most likely achieve the detection of ultra-low CV on fish surface by virtue of concentration effect. SERS spectrum of CV with the concentration of 4ng/mL and 0.4ng/mL and Raman spectrum directly collected from fish were shown in Fig. 10. 0.4ng/mL (1) represents the Raman spectrum that the substrate swabbed one side of the fish and 0.4ng/mL (2) represents two sides of the fish. In sharp contrast, Raman spectrum collected on the surface of fish directly is negligibly detected. Besides, in virtue of concentration effect achieved by swabbing both sides of the fish, as low as 0.4ng/mL CV was detected, which provides new opportunities for simply and efficiently tracing ultra-low CV in the real situation.

 

Fig. 10 Raman spectra of CV molecules of different concentration swabbed from the fish surface.

Download Full Size | PDF

This significant increasement benefits by the following two factors. First, the uniform and high-density Ag nanoparticles fabricated via the in situ growth method generate the abundant hot spots along the 3D branches of foam substrates. Then, in particular, the flexible 3D porous foam structure can absorb the disassociated organic pollutions molecules from arbitrary shape curved surface fully. What is more, 3D foam structure due to high specific surface area and enhanced optical scattering can integrate the scattered laser cumulatively, resulting in higher Raman enhancement effect. Besides, the foam substrate can collect the organic pollutions molecules conveniently and efficiently, so the flexible AgNPs/G@Ni substrates own the bright prospect in the practical application of detecting the organic pollutions on arbitrary shape curved surface.

4. Conclusions

We have presented the fabrication of 3D AgNPs/G@Ni hybrid foam structures as the flexible SERS-active substrate and successfully introduced structural three-dimensionality and flexibility to Raman enhancement. Using R6G molecules as a probe, the 3D AgNPs/G@Ni foam structures display excellent SERS activity with a minimum detected concentration of 10⁻¹⁰M. A good linear relationship between SERS peak intensity and R6G concentrations obtained in a range from 10⁻⁵ to10⁻¹⁰M. Besides, this 3D AgNPs/G@Ni foam structure shows higher Raman enhancement of at least two order magnitudes, compared to 2D AgNPs/G@Ni plane structures in experiment. Finally, the flexible 3D AgNPs/G@Ni foam structure can swab arbitrary curvilinear surfaces in the practical applicability to achieve molecules concentration effect, which promotes the development of universally adaptable SERS substrate with high sensitivity.

Appendix

 

Fig. 11 (a)-(d) show the distribution density of the silver nanoparticles after different number of times of growing. The inset is SERS mapping of vibration modes at 612 cm⁻¹ of R6G molecules at 10⁻⁵ M over 10 μm × 10 μm area. Raman intensity ranges from 2542 to 39127 counts.

Download Full Size | PDF

 

Fig. 12 (a) The distribution situation of Ag seeds thick graphene. (b) SERS mapping of vibration modes at 612 cm⁻¹ of R6G molecules at 10⁻¹⁰ M dispensed on AgNPs/G@Ni foam substrate.

Download Full Size | PDF

Funding

National Natural Science Foundation of China (NSFC) (11474187, 11274204, 61205174); Shandong Provincial Natural Science Foundation, China (ZR2016AM19)

Acknowledgments

Great thanks to Professor Yanyan Huo for her helpful suggestions in our COMSOL analysis. Thanks to Professor Chuansong Chen for his great help in the manuscript writing.

References and links

1. M. Pumera, “Graphene-based nanomaterials and their electrochemistry,” Chem. Soc. Rev. 39(11), 4146–4157 (2010). [CrossRef]   [PubMed]  

2. Z.-S. Wu, S. Yang, Y. Sun, K. Parvez, X. Feng, and K. Müllen, “3D nitrogen-doped graphene aerogel-supported Fe₃O₄ nanoparticles as efficient electrocatalysts for the oxygen reduction reaction,” J. Am. Chem. Soc. 134(22), 9082–9085 (2012). [CrossRef]   [PubMed]  

3. W. Xu, X. Ling, J. Xiao, M. S. Dresselhaus, J. Kong, H. Xu, Z. Liu, and J. Zhang, “Surface enhanced Raman spectroscopy on a flat graphene surface,” Proc. Natl. Acad. Sci. U.S.A. 109(24), 9281–9286 (2012). [CrossRef]   [PubMed]  

4. J. Mertens, A. L. Eiden, D. O. Sigle, F. Huang, A. Lombardo, Z. Sun, R. S. Sundaram, A. Colli, C. Tserkezis, J. Aizpurua, S. Milana, A. C. Ferrari, and J. J. Baumberg, “Controlling subnanometer gaps in plasmonic dimers using graphene,” Nano Lett. 13(11), 5033–5038 (2013). [CrossRef]   [PubMed]  

5. W. Xu, J. Xiao, Y. Chen, Y. Chen, X. Ling, and J. Zhang, “Graphene-Veiled Gold Substrate for Surface-Enhanced Raman Spectroscopy,” Adv. Mater. 25(6), 928–933 (2013). [CrossRef]   [PubMed]  

6. G. Lu, H. Li, C. Liusman, Z. Yin, S. Wu, and H. Zhang, “Surface enhanced Raman scattering of Ag or Au nanoparticle-decorated reduced graphene oxide for detection of aromatic molecules,” Chem. Sci. 2(9), 1817–1821 (2011). [CrossRef]  

7. Y. Wang, Z. Ni, H. Hu, Y. Hao, C. P. Wong, T. Yu, J. T. L. Thong, and Z. X. Shen, “Gold on graphene as a substrate for surface enhanced Raman scattering study,” Appl. Phys. Lett. 97(16), 163111 (2010). [CrossRef]  

8. C. Yang, C. Zhang, Y. Huo, S. Jiang, H. Qiu, Y. Xu, X. Li, and B. Man, “Shell-isolated graphene@Cu nanoparticles on graphene@Cu substrates for the application in SERS,” Carbon 98, 526–533 (2016).

9. S. Xu, S. Jiang, J. Wang, J. Wei, W. Yue, and Y. Ma, “Graphene isolated Au nanoparticle arrays with high reproducibility for high-performance surface-enhanced Raman scattering,” Sens. Actuat. Biol. Chem. 222, 1175–1183 (2016).

10. J. Zang, S. Ryu, N. Pugno, Q. Wang, Q. Tu, M. J. Buehler, and X. Zhao, “Multifunctionality and control of the crumpling and unfolding of large-area graphene,” Nat. Mater. 12(4), 321–325 (2013). [CrossRef]   [PubMed]  

11. M. C. Wang, S. Chun, R. S. Han, A. Ashraf, P. Kang, and S. Nam, “Heterogeneous, three-dimensional texturing of graphene,” Nano Lett. 15(3), 1829–1835 (2015). [CrossRef]   [PubMed]  

12. C.-C. Fu, A. Grimes, M. Long, C. G. L. Ferri, B. D. Rich, S. Ghosh, S. Ghosh, L. P. Lee, A. Gopinathan, and M. Khine, “Tunable nanowrinkles on shape memory polymer sheets,” Adv. Mater. 21(44), 4472–4476 (2009). [CrossRef]  

13. M. D. Huntington, C. J. Engel, A. J. Hryn, and T. W. Odom, “Polymer nanowrinkles with continuously tunable wavelengths,” ACS appl. Mater. Inter. 5(13), 6438–6442 (2013). [CrossRef]  

14. C. Srichan, M. Ekpanyapong, M. Horprathum, P. Eiamchai, N. Nuntawong, D. Phokharatkul, P. Danvirutai, E. Bohez, A. Wisitsoraat, and A. Tuantranont, “Highly-Sensitive Surface-Enhanced Raman Spectroscopy (SERS)-based Chemical Sensor using 3D Graphene Foam Decorated with Silver Nanoparticles as SERS substrate,” Sci. Rep. 6, 23733 (2016).

15. R. Bhat and V. M. Gómez-López, eds., Practical food safety: Contemporary issues and future directions (John Wiley & Sons, 2014).

16. D. Hurtaud-Pessel, P. Couëdor, and E. Verdon, “Liquid chromatography-tandem mass spectrometry method for the determination of dye residues in aquaculture products: Development and validation,” J. Chromatogr. A 1218(12), 1632–1645 (2011). [CrossRef]   [PubMed]  

17. M. J. Natan, “Surface enhanced Raman scattering,” Faraday Discuss. 132, 321–328 (2006). [CrossRef]   [PubMed]  

18. C. Li, Y. Huang, L. Pei, W. Wu, W. Yu, B. A. Rasco, and K. Lai, “Analyses of trace crystal violet and leucocrystal violet with gold nanospheres and commercial gold nanosubstrates for surface-enhanced Raman spectroscopy,” Food Anal. Methods 7(10), 2107–2112 (2014). [CrossRef]  

19. W. Xu, N. Mao, and J. Zhang, “Graphene: a platform for surface-enhanced Raman spectroscopy,” Small 9(8), 1206–1224 (2013). [CrossRef]   [PubMed]  

20. S. Drieschner, M. Weber, J. Wohlketzetter, J. Vieten, E. Makrygiannis, B. M. Blaschke, V. Morandi, L. Colombo, F. Bonaccorso, and J. A. Garrido, “High surface area graphene foams by chemical vapor deposition,” 2D Mater. 3(4), 045013 (2016).

21. S. J. Chae, F. Gunes, K. K. Kim, E. S. Kim, G. H. Han, S. M. Kim, H. Shin, S. Yoon, J. Choi, M. H. Park, C. W. Yang, D. Pribat, and Y. H. Lee, “Synthesis of large-area graphene layers on poly-nickel substrate by chemical vapour deposition: Wrinkle formation,” Adv. Mater. 21(22), 2328–2333 (2009). [CrossRef]  

22. Z. Chen, W. Ren, L. Gao, B. Liu, S. Pei, and H. M. Cheng, “Three-dimensional flexible and conductive interconnected graphene networks grown by chemical vapour deposition,” Nat. Mater. 10(6), 424–428 (2011). [CrossRef]   [PubMed]  

23. A. C. Ferrari, J. C. Meyer, V. Scardaci, C. Casiraghi, M. Lazzeri, F. Mauri, S. Piscanec, D. Jiang, K. S. Novoselov, S. Roth, and A. K. Geim, “Raman spectrum of graphene and graphene layers,” Phys. Rev. Lett. 97(18), 187401 (2006). [CrossRef]   [PubMed]  

24. A. C. Ferrari, “Raman spectroscopy of graphene and graphite: disorder, electron–phonon coupling, doping and nonadiabatic effects,” Solid State Commun. 143(1-2), 47–57 (2007). [CrossRef]  

25. S. Xu, S. Jiang, J. Wang, J. Wei, W. Yue, and Y. Ma, “Graphene isolated Au nanoparticle arrays with high reproducibility for high-performance surface-enhanced Raman scattering,” Sens. Actuat. Biol. Chem. 222, 1175–1183 (2016).

26. X. Wei, Z. Meng, L. Ruiz, W. Xia, C. Lee, J. W. Kysar, J. C. Hone, S. Keten, and H. D. Espinosa, “Recoverable slippage mechanism in multilayer graphene leads to repeatable energy dissipation,” ACS Nano 10(2), 1820–1828 (2016). [CrossRef]   [PubMed]  

27. J. Zhao, Y. He, Y. Wang, W. Wang, L. Yan, and J. Luo, “An investigation on the tribological properties of multilayer graphene and MoS₂ nanosheets as additives used in hydraulic applications,” Tribol. Int. 97, 14–20 (2016). [CrossRef]  

28. J. Guo, S. Xu, X. Liu, Z. Li, L. Hu, Z. Li, P. Chen, Y. Ma, S. Jiang, and T. Ning, “Graphene oxide-Ag nanoparticles-pyramidal silicon hybrid system for homogeneous, long-term stable and sensitive SERS activity,” Appl. Surf. Sci. 396, 1130–1137 (2017). [CrossRef]  

29. K. Zhang, J. Zhao, H. Xu, Y. Li, J. Ji, and B. Liu, “Multifunctional paper strip based on self-assembled interfacial plasmonic nanoparticle arrays for sensitive SERS detection,” ACS Appl. Mater. Interf. 7(30), 16767–16774 (2015). [CrossRef]   [PubMed]  

30. A. M. Gilbertson, Y. Francescato, T. Roschuk, V. Shautsova, Y. Chen, T. P. H. Sidiropoulos, M. Hong, V. Giannini, S. A. Maier, L. F. Cohen, and R. F. Oulton, “Plasmon-Induced Optical Anisotropy in Hybrid Graphene-Metal Nanoparticle Systems,” Nano Lett. 15(5), 3458–3464 (2015). [CrossRef]   [PubMed]  

31. D. B. Farmer, D. Rodrigo, T. Low, and P. Avouris, “Plasmon-Plasmon Hybridization and Bandwidth Enhancement in Nanostructured Graphene,” Nano Lett. 15(4), 2582–2587 (2015). [CrossRef]   [PubMed]  

32. M. Asl, P. Bristowe, K. Koziol, and T. Heine, “Effect of compression on the electronic, optical and transport properties of MoS2/graphene-based junctions,” 2D Mater. 3(2), 025018 (2016).