Open Access
Add to BibSonomy
Abstract
We reported an efficient fabrication strategy for a three-dimensional structure based on large-area flexible carbon fiber cloth decorated by Ag nanoparticles (AgNPs-CFC). AgNPs with different sizes and morphologies were assembled on the surface of the CFC via a vacuum thermal evaporation and high-temperature annealing method. Samples with Ag film thickness of 5 nm, 10 nm, 15 nm, 20 nm, 30 nm and 50 nm were prepared and analyzed. The forming mechanism of AgNPs on the surface of CFC was studied in detail. Finite element numerical simulations with COMSOL Multiphysics were carried on to show the electronic field distributions of different samples based on SEM images. Raman measurements with Rhodamine 6G (R6G) molecules as analytes indicated that the analytical enhancement factor (AEF) reaches to 2.4×1012 and a detection limit is as low as 1.0×10−14 mol/L.
© 2021 Optical Society of America under the terms of the OSA Open Access Publishing Agreement
1. Introduction
Surface-enhanced Raman scattering (SERS) has been widely used in bioscience [1], food safety and environmental science [2,3], due to its much higher sensitivity compared with Raman scattering. Metallic nanoparticles, nanorods, nanoholes, nanosheets and other nanostructures as SERS substrates have been proposed in [4–9], most of the traditional SERS structures are prepared on rigid substrates like SiO2/Si [10]. In some practical applications, rigid substrates may not be suitable for complex surface detection [11]. Fortunately, the flexible SERS substrates (using PDMS, PMMA, two-dimensional MoO3−x nanosheets ink and Nafion membrane, etc.) can adapt to these applications due to its property of winding and conformal attachment [12–15]. The flexible substrates could be easily detached from detected surface with little damage [16]. In recent years, with the property improvement of MoS2, graphene, carbon nanotubes (CNTs), flexible carbon fiber cloth (CFC) and other materials [17–22], which have many advantages of large specific surface area, good stability and strong adsorptive capacity, etc., several flexible composite structures (such as silver nanocrystals on carbon fiber cloth, graphene oxide/Ag nanoparticles based on PMMA and two-dimensional gold nanoplates on carbon nanotube sheet) have been investigated in Refs. [22,13,23], with a reported enhancement factor of ∼ 109 and limit detection (LOD) of ∼ 10−11 mol/L approximately. It is still necessary to further improve the performance of flexible SERS substrates.
Based on our reported carbon nanotubes/silver nanocomposite structure on SiO2/Si [24], there was a coupling effect of carbon nanotube and Ag nanoparticles, which contributes to Raman enhancement. In this paper, we propose a three-dimensional structure based on large area flexible carbon fiber cloth decorated by Ag nanoparticles (AgNPs-CFC) as SERS substrate. Its forming mechanism of AgNPs on the surface of CFC, electronic field distribution and Raman enhancement properties were analyzed in details as follow.
2. Materials and instruments
2.1 Materials
Carbon fiber cloth was obtained from Phychemi (HK) Co. Ltd. Concentrated sulfuric acid (AR, 98%) and nitric acid (AR, 65%) were obtained from Cheng Du Chron Chemicals Co. Ltd. Rhodamine 6G was purchased from Shanghai Aladdin Co. Ltd.
The surface morphologies of samples were characterized by an environmental scanning electron microscopy (ESEM, Thermo Fisher Scientific Quattro S). X-Ray diffraction spectra were characterized by multi-function high resolution X-ray diffractometer (XRD, Spectris Pte. Ltd PANalytical X’Pert Powder). Raman spectra were recorded with a laser confocal Raman spectrometer (Horbia Jobin Yvon LabRAM HR Evolution) equipped with a 50 × LWD objective of numerical aperture (NA) 0.5, at work distance (WD) of 10.6 mm. A 532 nm air cooled frequency doubled Nd: Yag green laser with a power of 50 mW and a grating of 600 gmm-1 was chosen for measurements. In order to reduce the heating effect caused by laser, the power filter was set at 10%, while an integration time was set at 5 s to reduce the noise signal. The baseline of spectral data was removed by the LabSpec software.
3. Preparation
The fabrication process of the three-dimensional Ag nanoparticles/carbon fiber cloth (3D AgNPs-CFC) hybrid structure is shown in Fig. 1. Firstly, in order to remove the organic and inorganic impurities on CFC surface, CFC was orderly cleaned by ultrasonic in acetone, anhydrous ethanol and deionized water for half an hour (Fig. 1(a)). Secondly, CFC was put into the hydrophilic agent (mixed with concentrated HNO3 and concentrated H2SO4 with a volume ratio of 1:1), heated and boiled for 1 hour in a thermostatic water bath pot, and then dried for 16 hours at 120 °C in a vacuum dryer (Fig. 1(b)). Thirdly, Ag film was deposited onto hydrophilic CFC by a vacuum thermal evaporation under condition of 10−3 Pa pressure and the rate about 1 Å s-1 (Fig. 1(c)). We prepared different thicknesses of 5 nm, 10 nm, 15 nm, 20 nm, 30 nm and 50 nm, respectively. Subsequently, CFC covered by Ag film (Ag film-CFC) was annealed at 600 °C for 60 minute in a quartz tube furnace, flowing mixed gas of H2 (30 mL/min) and Ar (120 mL/min) (Fig. 1(d)). AgNPs-CFC samples were completed. Finally, prepared samples were immersed in Rhodamine 6G solution with different concentrations for 24 hour before Raman measurement (Fig. 1(e)).
Fig. 1. Schematic diagram of fabrication process of the 3D AgNPs-CFC nanostructures, (a) cleaned by ultrasound process, (b) hydrophilic treatment, (c) Ag film deposition, (d) annealing treatment, and (e) Raman measurements.
4. Characterization
4.1 SEM and XRD
Figure 2 shows SEM images of samples Ag film-CFC and AgNPs-CFC with different thicknesses of deposited Ag film. SEM images of sample 5 nm Ag film-CFC are shown in Figs. 2(a1) and 2(a2). We can see independent uneven Ag nanoparticle chains on the surface. As for sample 10 nm Ag film-CFC (Figs. 2(b1) and 2(b2)), Ag film becomes flat. When the thickness of deposited Ag film is 15 nm, 20 nm and 30 nm, we can see that CFC substrates are coated with smooth Ag film, and there are no AgNPs, shown in Figs. 2(c1), 2(c2), 2(d1), 2(d2), 2(e1), and 2(e2). When the thickness of deposited Ag film is up to 50 nm, we can see some Ag grain agglomerations appeared on the surface of Ag film.
Fig. 2. SEM images of Ag film-CFC substrates with different thicknesses of deposited Ag of (a1, a2) 5 nm, (b1, b2) 10 nm, (c1, c2) 15 nm, (d1, d2) 20 nm, (e1, e2) 30 nm, (f1, f2) 50 nm; SEM images of 3D AgNPs-CFC after annealing, while the thickness of deposited Ag of (a3, a4) 5 nm, (b3, b4) 10 nm, (c3) 15 nm, (d3) 20 nm, (e3) 30 nm, (f3) 50 nm; random statistics of AgNPs averaged radius of samples (a3) 5 nm, (b3) 10 nm, (c3) 15 nm, (d3) 20 nm, (e3) 30 nm, (f3) 50 nm; (g) XRD of sample CFC and Ag film-CFC; (h) calculated Ag coverage of samples.
Figures 2(a3), 2(a4), 2(b3), 2(b4), 2(c3), 2(d3), 2(e3) and 2(f3) show SEM images of samples after high temperature annealing. Three-dimensional surface of CFC is coated by Ag nanoparticles. The size of AgNPs depends on the thickness of deposited Ag film. A binarization grayscale image processing method was used to calculate the size and the coverage of AgNPs. The calculated average radius of top surface AgNPs, side surface AgNPs and Ag coverage is ∼ 33 nm, ∼ 25 nm and ∼ 12.1% for sample 5 nm AgNPs-CFC (Fig. 2(a3)), respectively. As for sample 10 nm AgNPs-CFC, three-dimensional surface of CFC is coated by Ag nanoparticles and Ag nanoparticle chains which have round ends of both sides (Fig. 2(b3)). The calculated Ag coverage is ∼ 12.5%. As for sample 15 nm AgNPs-CFC, most of Ag are nanoparticle chains and their gaps become larger (Fig. 2(c3)). The calculated Ag coverage is ∼ 26.5%. As for samples 20 nm AgNPs-CFC, 30 nm AgNPs-CFC and 50 nm AgNPs-CFC, the number of Ag nanoparticles is significantly reduced and the sizes of nanoparticle chains increases. Their calculated Ag coverages are ∼ 25.7%, ∼ 23.2% and ∼ 53.3%, respectively. The calculated average radius of side surface AgNPs for samples 30 nm and 50 nm AgNPs-CFC (Figs. 2(e3, f3)) is ∼ 55 nm and ∼ 150 nm. Figure 2(g) shows X-Ray diffraction spectra of CFC and Ag film-CFC, diffraction peaks at 38.1° and 64.5° correspond to (111) and (200) crystal faces of silver.
The calculated Ag coverages of all samples are shown in Fig. 2(h). For sample 5 nm AgNPs-CFC, Ag nanoparticle chains break up into larger isolated irregular round nanoparticles. For samples coated with smooth Ag film, such as 10 nm, 15 nm, 20 nm and 30 nm Ag film-CFC, flat Ag film breaks up into Ag nanoparticle chains at first. In this process, the coverage decreases dramatically until the nanoparticle chains rupture into nanoparticles. For sample 50 nm AgNPs-CFC, agglomerated Ag nanoparticles on Ag film surface enlarge by gathering Ag grain from flat film area, and then turn into isolated state. This process didn’t cause a significant decrease of the Ag coverage. Within this thickness range, if samples have thicker initial Ag film deposited, the final calculated Ag coverage will be lower. Because the thicker film hinders this process. These characterizations will be analyzed in the next section.
4.2 Forming mechanism of Ag nanoparticles
As for the forming of AgNPs-CFC, we analyze as follows.
Step 1: Ag film deposition (Figs. 3(a1, a2))
Fig. 3. Schematic illustration of (a1) CFC, (b1) Ag film-CFC, (a3) AgNPs-CFC; detailed formation, (b1-d1) initial phase, Ag film deposition, (b2-d2) phase 1, Ag film on CFC, (b3-d3) phase 2, hole formation and growth, (b4-d4) phase 3, nanoparticle formation.
Figures 3(a1) and 3(a2) show Ag film deposition process via a vacuum thermal evaporation with silver atoms escaping from the surface of the silver substrate into the vacuum chamber (Figs. 3(c1)). As the surface temperature of CFC is lower than that of evaporation source, silver atoms coalesce, nucleate, grow and form a continuous thin film on the CFC surface, corresponding to the change from initial phase (Figs. 3(b1, c1 and d1)) to phase 1 on single carbon fiber. This process is mainly influenced by source-base distance and the mean free path of silver atoms λAg, which can be expressed as [25]:
In which ρ is the residual gas molecular density, dAg is the evaporation atomic diameter, P is the pressure (vacuum) in the cavity, k is Boltzmann's constant, and Te is evaporation temperature of silver. The thickness of silver film is controlled by the deposition time referenced to the silicon sample. Herein, the surface of carbon fiber is curved, which means that the distance from CFC side surface to silver substrate (Fig. 3(b1)) is larger than the distance from the CFC top surface (Fig. 3(b2)). Thus, the Ag film thickness on the top surface of CFC is a little larger than that on side surface, leading to more silver atoms deposited on CFC top surface (Fig. 3(c2)).
Figure 3(c2) indicates that the initial grain boundary of Ag films is a key factor for nanoparticles formation. A number of cracks were observed on the Ag film-CFC samples, affected by the nanoscale rough surface of carbon fiber, shown in Figs. 2(a2-f2).
Step 2: Annealing process (Fig. 3(a3))
Figure 3(a3) shows the solid dewetting process via a high temperature annealing with Ag film-CFC samples. During this process, if the annealing temperature is lower than the melting temperature of the silver, with the impel of external thermal energy, Ag film will rupture from the initial grain boundary (Fig. 3(c3)) [26].
Hole formation depends on the defects of the original film (Fig. 3(a2)). According to the grain configuration assumptions [27], Ag atoms coalesce as film by one-layer countless uniformed size Ag grains. The morphology of Ag grains is expressed in Fig. 3(c3). In which, H is the initial thickness of Ag film, L is the initial size of Ag grains, θ1 is the equilibrium groove angles of Ag film surface and θ2 is groove angle at the bottom surface of the film, expressed as:
Where γi is the interface energy between Ag film and the substrate, γb is the grain boundary energy, γs is the surface energy of Ag grain. Upon annealing process, Ag grains are enlarged to a critical status for hole formation on the Ag film, meanwhile grain boundary disappears. The size of the critical Ag grain Lc is:
When L < Lc, the hole will not appear on the Ag film. On the contrary, the hole will appear on the samples with thinner film thickness (H). Hole formation also depends on the defects, such as grain boundary and crack on the original film [28]. Cracks of Ag film-CFC samples increase the number of defects which lead to the hole formation on Ag film-CFC samples.
Once the critical size of the hole has been formed, the capillary force can drive the growth of nanoholes. The atomic transport is surface self-diffusion. Since the annealing time is constant, the hole growth rate is controlled by the shrinkage rate of hole edge which can express as [28,29]:
Where t is time, Bs is the metal constant, ${\bar{\nabla }_s}$ is the surface gradient operator; Ns is unit surface atomic number, Ω is atomic volume, TA is evaporation temperature of silver; Ds is surface diffusion rate, DS0 is the pre-exponential factor, EA is the activation energy for the diffusion process. The hole growth rate is inversely proportional to the film thickness [30]. Therefore, many wider nanoparticle chains were found on the samples deposited with thicker Ag films. This sample has fewer holes than other samples in the same annealing time (Fig. 2(d6)).
Phase 3: Nanoparticle formation (Figs. 3(b4, c4 and d4))
Figures 3(b3) and 3(b4) show that Ag nanoparticle chains break up from the long side edge, and turn into isolated round nanoparticles. The rate of nanoparticle formation is expressed as [31]:
Where Vα is the dimensionless speed and depends on the contact angle of silver. It can be regarded as constant. So the rate of nanoparticle growth is influenced by the thicknesses of Ag film.
SEM characterization in Fig. 2 can be analyzed by the effect of the thickness. In the process from phase 1 to phase 2, due to different thicknesses of individual samples, the difficulty of hole forming increases with the increase of film thickness, the number of holes per unit area is correspondingly reduced meanwhile the number of hole edges that have retracted is reduced. The result is that fewer nanoparticles or nanoparticle chains are eventually formed. For samples with larger thickness, more silver atoms will aggregate to form fewer, large-sized particles with larger gaps. For example, in Fig. 2 the Ag coverage of the 50 nm AgNPs-CFC sample is particularly large, because the number of silver particles per unit area is small, and most of the silver atoms are still agglomerated similar to the shape of Ag film. In Fig. 2(h), the Ag coverage of 15 nm and 20 nm AgNPs-CFC samples is larger than 30 nm AgNPs-CFC sample. Since the thickness of the initial silver film is much smaller than the height of the particles, when the silver atoms in a certain area agglomerate into larger particles, the aggregation of silver atoms causes the Ag coverage to be significantly reduced. However, the coverage rate of the 15 nm and 20 nm AgNPs-CFC samples has increased abnormally. This is because the cracks on the CFC surface promoting the formation of holes and transforming some large particles into small particles.
In nanoparticle formation process, 5 nm AgNPs-CFC sample has the highest shrinkage rate and nanoparticle formation rate among all of samples. So almost all the original uneven nanoparticle chains turn into round nanoparticles in just 1 hour. The film thickness impacts the above two rates significantly. In a constant annealing time, the size of the nanoparticle chains formed on the thick film is increased and the number is decreased. Therefore, the number of nanoparticle chains on 10 nm, 15 nm, 20 nm and 30 nm AgNPs-CFC samples gradually increases. For 50 nm AgNPS-CFC sample, the thicker silver film hinders the hole forming of the silver film. Due to the CFC cracks, many nanoparticle chains of different sizes were formed on the sample surface. However, these nanoparticle chains are still in an irregular state because of the low edge shrinkage caused by the thickness, especially on the top surface.
The difference between the top and the sides can be attributed to the effects of thickness. Compared with the side surface, the thicker film on the upper surface inhibits the formation of nanoparticles. Therefore, the size of nanoparticles and nanoparticle chains on the side of all samples is smaller than that on the upper surface.
5. Raman measurements
Figures 4(a), (b) and (c) show Raman measurements with our 3D AgNPs-CFC as SERS substrate for R6G molecules. Raman characteristic peaks of R6G molecules include ∼ 613, ∼ 771, ∼ 1187, ∼ 1314, ∼ 1361, ∼ 1509 and ∼ 1650 cm-1 [24]. Raman peaks of CFC are ∼1350 cm-1 and ∼1580 cm-1, which overlap with some characteristic peaks of R6G (in Fig. 4(a)). Therefore, the characteristic peak at 613 cm-1 is used to evaluate Raman enhancement characteristics of our samples. Shown in Fig. 4(b), the limit of detection (LOD) of 5 nm, 10 nm and 15 nm AgNPs-CFC samples is 1.0 × 10−14 mol/L (-14 M), 1.0 × 10−11 mol/L (-11 M) and 1.0 × 10−13 mol/L (-13 M), respectively. The LOD of 20 nm,30 nm and 50 nm AgNPs-CFC samples is 1.0 × 10−14 mol/L (-14 M). Analytical enhancement factor (AEF) of SERS substrate can be calculated as:
Where ISERS and INormal are the Raman peak intensities of SERS and without SRES respectively, CSERS and CNormal are molecular concentration to be measured during SERS and without SRES measurement, respectively.
Fig. 4. (a) Raman results of R6G with concentration of -14 M for 50 nm AgNPs-CFC samples, (b) the LOD of 5 nm, 10 nm, 20 nm, 30 nm and 50 nm AgNPs-CFC samples, (c) the calculated AEF of 5 nm, 10 nm, 15 nm, 20 nm, 30 nm and 50 nm AgNPs-CFC samples; (d) Raman results of R6G with concentration of -14 M for 5 nm AgNPs-CFC samples (after 200 days) and -14 M for fresh 5 nm AgNPs-CFC sample; (e) Raman results of crystal violet (CV) with concentration of -10 M for 5 nm AgNPs-CFC sample, (f) intensity distribution of the peaks at 913 cm−1 corresponding to (e) with the RSD of 0.338.
Shown in Fig. 4(c), the calculated AEF of 5 nm, 10 nm, 15 nm, 20 nm, 30 nm and 50 nm is 2.40 × 1012, 1.36 × 109, 2.85 × 1011, 2.42 × 1012, 1.25 × 1012, 9.07 × 1011, respectively. The maximal calculated AEF of our samples is 2.42×1012 and the lowest limit of detection of our samples is 1.0 × 10−14 mol/L (-14 M).
Due to the effects of oxidation, the AEF of 5 nm AgNPs-CFC sample which was stored about 200 days is 6.02 × 107 and the detection limit is 1.0 × 10−10 mol/L for R6G molecules (Fig. 4(d)). While for fresh 5 nm AgNPs-CFC sample, its AEF is 2.40 × 1012 and the detection limit is 1.0 × 10−14 mol/L for R6G molecules (Fig. 4(d)). We also measured Raman spectral of CV (concentration of 10−10 mol/L) with our sample (5 nm AgNPs-CFC) as SERS substrate, shown in Fig. 4(e). We can see clearly Raman characteristic peaks of crystal violet (CV) molecules at ∼ 416, ∼ 722, ∼ 913, ∼ 1175, ∼ 1357, ∼ 1389, ∼ 1579 and 1614 cm-1 [17]. The relative standard deviation (RSD) at 913cm-1 is 0.338 for -10 M CV molecules (Fig. 4(f)).
According to the SEM characterization, for sample 5 nm AgNPs-CFC, the hot spots are mainly contributed by Ag nanoparticles on the top and side surface. While for sample 10 nm AgNPs-CFC, the hot spots are contributed by Ag nanoparticles and nanoparticle chains on the top surface and Ag nanoparticles on the side surface. For samples 15 nm, 20 nm and 30 nm AgNPs-CFC, the hot spots are mainly caused by Ag nanoparticle chains on the top surface and Ag nanoparticles on the side surface. In addition, for 50 nm AgNPs-CFC, Ag nanoparticles and nanoparticle chains that exist on both top and side surface are the principal source of hot spots.
6. Simulation
Finite element software COMSOL Multiphysics 5.5 was used to simulate the spatial distribution of electromagnetic field intensity of our samples. Shown in Fig. 5(a), the wavelength of laser is 532 nm, polarizing along the Y direction and propagating in the -Z direction. The complex refractive index of Ag was set as 0.054007 + 3.4290·i [32], and the refractive index of carbon fiber was set as 2.6913 + 1.4557·i [33]. From SEM images in Fig. (2), Ag nanoparticles were set as two kinds of structures: silver nanoparticle (Figs. 5(b1) and (b2)) and silver nanoparticle chain (Figs. 6(a1), (a2), (a3)). The radius of the silver nanoparticles was set as 25 nm, 33 nm, 43.5 nm, 55 nm, 60 nm, 80 nm, respectively; silver nanoparticle chains were composed spheres with radius of 25 nm and 30 nm or spheres with radius of 50 nm and 60 nm, respectively; the radius of the carbon fiber was simplified as 300 nm, based on SEM images. All of silver particles were evenly distributed on the surface of the carbon fiber cloth. In the Y-Z plane, the 25 nm-particles were evenly distributed along the curved surface, and the spacing was ∼ 1 nm. Note that the key parameter we studied is the influence of the radius of the particles on the electric field distribution, therefore, we set the gap to a fixed value. It was evenly distributed along the X-axis in the X-Z plane, with an interval 0.5 times the particle radius (Figs. 5(b1) and 5(b2)). The electric field distributions of Y-Z plane monitor are shown in Fig. 5(c). The maximal local electric field intensity of other Ag nanoparticles is shown in Fig. 5(d).
Fig. 5. (a) Schematic model of local electric field effects; (b1) and (b2) Schematic model of 3D AgNPs-CFC with Ag nanoparticles radius set as 25 nm; (c) the local electric field intensity of Ag nanoparticles with radius set as 25 nm; (d) the maximal local electric field intensity at different positions of Ag nanoparticles with radius set as 25 nm, 33 nm, 43.5 nm, 55 nm, 60 nm and 80 nm; the unit of local electric field intensity is V/m.
Fig. 6. (a1-a3) Schematic model of 3D Ag nanoparticle chains spheres with radius of 25 nm and 30 nm; (b) and (c) the local electric field intensity of different Y-Z planes corresponding to monitor 2 and 3; the unit of local electric field intensity is V/m.
The enhancement factor induced by the electromagnetic field effects can be calculated by:
Where E0 is the incident electric field intensity, Eout(ω0), Eout(ωs) is electric field intensity at frequency ω0 and ωs, respectively. In Fig. 5(d), for Ag nanoparticles radius of 25 nm, 33 nm, 43.5 nm, 55 nm, 60 nm and 80 nm, the maximal local electric field is 281 V/m, 362 V/m, 92.9 V/m, 54.2 V/m, 363 V/m and 347 V/m, respectively, and the corresponding maximal EFEM is 6.2 × 109, 1.7 × 1010, 7.4 × 108, 8.6 × 106, 1.73 × 1010 and 1.4 × 1010. When the particle radius is less than 60 nm, the strongest hot spots is located on the side surface. If the radius is greater than 60 nm, it will be located on the top surface. There is an electric field distribution of 10 ∼ 30 V/m between Ag and CFC. The increase in this value is accompanied by the weakening of the local electric field between the nanoparticles. The additional effective dielectric loss introduced by the CFC [10] which results in partial electric field coupling into the CFC and leads to the decrease of the hot spots, especially for AgNPs with radius of 43.5 nm and 55 nm corresponding to 10 nm and 15 nm samples, respectively. This is reason for the Raman measurements of R6G. The parameters of 33 nm, 43.5 nm, 60 nm, and 80 nm are the average particle radius data of samples 5 nm, 10 nm, 15 nm, and 20 nm AgNPs-CFC. Therefore, the size of Ag nanoparticles will greatly affect the LOD and AEF of 3D AgNPs-CFC samples.
In Fig. 6, silver nanoparticle chains were composed spheres with radius of 25 nm and 30 nm, the lower carbon fiber and the upper carbon fiber were undulated along the Z axis, which was used to simulate the three-dimensional CFC surface. In the Y-Z plane 2, adjacent nanoparticle chains were arranged staggered with an interval of ∼ 1 nm, and were evenly distributed along the X axis in the X-Z plane, with an interval of 0.5 times the particle radius (Fig. 6(a1)). Monitor 2 (Y-Z plane 2) is set on the plane where the center of 30 nm radius sphere is located, and Monitor 3 (Y-Z plane 3) is set on the plane where the center of 25 nm radius sphere is located (Fig. 6(a3)).
In Figs. 6(b) and (c), the hot spots are located between the ends of the Ag nanoparticle chain with a radius of 25 nm, and they are on the interface of two adjacent carbon fibers. Meanwhile, the electric field distribution on the top and side surfaces is relatively small. This is because the distance between the nanoparticle chains on the sides of two adjacent carbon fibers was controlled by the relative position of the carbon fibers. If their gaps were smaller than that of other nanoparticle chains, the maximal local electric field will be stronger. On the contrary, if the distance between nanoparticles was increased, the intensity of the hot spots will be greatly reduced, shown in Fig. 6(c).
For Ag nanoparticles (Fig. 5(d)) the maximal local electric field is 363 V/m, EFEM = 1.73 × 1010; for Ag nanoparticle chains (Fig. 6(b)), the maximal local electric field is 60.3 V/m, EFEM = 1.3 × 107; for Ag nanoparticle chains on the interface of two adjacent carbon fiber, the maximal local electric field is 237 V/m, EFEM = 3.1 × 109. Therefore, the Raman enhancement effect of the nanoparticle sample is better than that of the nanoparticle chain sample, but the enhancement effect of the nanoparticle chain is also greatly enhanced due to the influence of the relative position difference caused by the three-dimensional spatial distribution of carbon fiber. These results are similar to Raman measurements.
According to SEM images, some samples, such as 10 nm, 15 nm, 20 nm and 30 nm AgNPs-CFC samples, have silver nanoparticles and silver nanoparticle chains distributed on the side and top surface of the CFC, respectively. In Figs. 7(a1), 7(a2) and 7(a3), the mixed sample covered by Ag nanoparticle chains which were composed of spheres with radius of 50 nm and 60 nm on the top surface and staggered with an interval of ∼1 nm in Y-Z plane. On the side, a number of rows of silver nanoparticles with a radius of 25 nm were located with an interval of ∼1 nm in different Y-Z planes. Monitor 4 (Y-Z plane 4) is set on the plane where the center of the 60 nm radius sphere is located; Monitor 5 (Y-Z plane 5) is set on the plane in which the center of the 25 nm radius nanoparticle is located; and Monitor 6 (Y-Z plane 6) is set on the plane where the center of the 50 nm radius sphere is located.
Fig. 7. (a) Schematic model of 3D AgNPs-CFC with Ag nanoparticles on the side surface and Ag nanoparticle chains on the top surface; the local electric field intensity of Monitor 4 (Y-Z plane 4) (b), Monitor 5 (Y-Z plane 5) (c), Monitor 6 (Y-Z plane 6) (d); the unit of local electric field intensity is V/m.
In Fig. 7(b), Y-Z plane 4 is mainly used to analyze the influence of the top Ag nanoparticle chains on the hot spots. The maximum local electric field is 90.2 V/m, EFEM(max) = 6.6 × 107. Hot spots are located on the interface of two Ag nanoparticle chains. In Fig. 7(c), Y-Z plane 5 is mainly used to analyze the influence of side Ag nanoparticles on hot spots. The maximal local electric field is 327 V/m, EFEM(max) = 1.1 × 1010. Hot spots are located between 25 nm radius Ag nanoparticles on the side surface. Because there is also a synergistic effect of silver nanoparticles at the interface of two adjacent carbon fibers, the result is greater than the phenomenon in Fig. 5(c). In Fig. 7(c), the Y-Z plane 6 is mainly used to analyze the influence of the top Ag nanoparticle chains on the hot spots. The maximal local electric field is 90.2 V/m, EFEM(max) = 6.6 × 107. In Fig. 7(d), the Y-Z plane 5 is mainly used to analyze the synergistic influence of the side Ag nanoparticles and the top Ag nanoparticle chains on the hot spots, the maximal local electric field is 110 V/m, EFEM(max) = 1.5 × 108. The hot spot distribution shown in Y-Z plane 5 is more common in this structure. Although the maximal EFEM is not as large as specific position, the advantage of these hot spots is that they can be densely distributed on the substrate surface. For Raman measurement, the probe molecule can effectively fall inside the hot spots. In this structure, the size and position of particles and chains, and the relative position of CFC have an important influence on the simulation results. The morphology and size of the particles are mainly controlled by the original film thickness. These factors can effectively improve the SERS capability of 3D Ag-CFC samples.
7. Conclusion
In summary, we have developed a flexible three-dimensional AgNPs-CFC via depositing a layer silver film on flexible carbon fiber cloth substrate and high-temperature annealing. This method reported in this article can be used to prepare large-area flexible substrates, which could be bent without breaking and attached conformally to many complex surfaces of interest. In addition, we studied the forming mechanism of AgNPs on the surface of CFC in detail and explained the influence of CFC structures and the morphology of AgNPs on the field distribution using electromagnetic solvers. The AEF of our samples can reach 2.42 × 1012 and a detection limit is 1.0 × 10−14 mol/L for R6G molecules. In addition, we explained the influence of CFC structure and the morphology of AgNPs on the measurements through simulation. It is potential to be used in applications of the requirements on flexible substrates and large areas with high sensitivity.
Funding
National Natural Science Foundation of China (61875024); Chongqing Outstanding Youth Fund (cstc2019jcyjjqX0018); Funds of Central Universities (CQU2018CDHB1A07).
Acknowledgement
We would like to thank Dr. Gong Xiangnan at Analytical and Testing Centre of Chongqing University for his help in Raman measurement. We also want to thank Dr. Zhang Cong for his help in Ag deposition.
Disclosures
The authors declare no conflicts of interest.
Data availability
Data underlying the results presented in this paper are not publicly available at this time but may be obtained from the authors upon reasonable request.
References
1. X. G. Zhang, X. L. Zhang, C. L. Luo, Z. Q. Liu, Y. Y. Chen, S. L. Dong, C. Z. Jiang, S. K. Yang, F. B. Wang, and X. H. Xiao, “Volume-enhanced Raman Scattering detection of viruses,” Small 15(11), 1805516 (2019). [CrossRef]
2. F. Y. H. Kutsanedzie, A. A. Agyekum, V. Annavaram, and Q. S. Chen, “Signal-enhanced SERS-sensors of CAR-PLS and GA-PLS coupled AgNPs for ochratoxin A and aflatoxin B1 detection,” J. Agric. Food Chem. 315, 126231 (2020). [CrossRef]
3. Y. Xu, F. Y. H. Kutsanedzie, M. Hassan, J. J. Zhu, W. Ahmad, H. H. Li, and Q. S. Chen, “Mesoporous silica supported orderly-spaced gold nanoparticles SERS-based sensor for pesticides detection in food,” J. Agric. Food Chem. 315, 126300 (2020). [CrossRef]
4. Z. P. Zha, R. C. Liu, W. Yang, C. Li, J. J. Gao, M. Shafi, X. W. Fan, Z. Li, X. J. Du and S, and Z. Jiang, “Surface-enhanced Raman scattering by the composite structure of Ag NPs-multilayer Au films separated by Al2O3,” Opt. Express 29(6), 8890–8901 (2021). [CrossRef]
5. J. Zhang, X. L. Zhang, C. H. Lai, H. J. Zhou, and Y. Zhu, “Silver-decorated aligned CNT arrays as SERS substrates by high temperature annealing,” Opt. Express 22(18), 21157–21166 (2014). [CrossRef]
6. K. Sivashanmugan, J. D. Liao, and C. K. Yan, “Elimination of gallium concentration on focused-ion-beam-fabricated Au/Ag nanorod surface to recover its Raman scattering characteristic,” Sens. Actuators, B 206, 415–422 (2015). [CrossRef]
7. A. Martin, C. Schopf, A. Pescaglini, J. J. Wang, and D. Iacopino, “Facile formation of ordered vertical arrays by droplet evaporation of Au nanorod organic solutions,” Langmuir 30(34), 10206–10212 (2014). [CrossRef]
8. J. Yu, M. S. Yang, Z. Li, C. D. Liu, Y. S. Wei, C. Zhang, B. Y. Man, and F. C. Lei, “Hierarchical particle-in-quasicavity architecture for ultratrace in situ Raman sensing and its application in real-time monitoring of toxic pollutants,” Anal. Chem. 92(21), 14754–14761 (2020). [CrossRef]
9. R. C. Liu, Z. P. Zha, C. Li, M. Shafi, Q. Q. Peng, M. Liu, C. Zhang, X. J. Du and S, and Z. Jiang, “Coupling of multiple plasma polarization modes in particles–multilayer film system for surface-enhanced Raman scattering,” APL Photonics 6(3), 036104 (2021). [CrossRef]
10. T. C. Gong, Y. Zhu, J. Zhang, W. J. Ren, J. M. Quan, and N. Wang, “Study on surface-enhanced Raman scattering substrates structured with hybrid Ag nanoparticles and few-layer graphene,” Carbon 87, 385–394 (2015). [CrossRef]
11. X. F. Liu, J. M. Ma, P. F. Jiang, J. L. Shen, R. W. Wang, Y. Wang, and G. L. Tu, “Large-scale flexible surface-enhanced Raman scattering (SERS) sensors with high stability and signal homogeneity,” ACS Appl. Mater. Interfaces 12(40), 45332–45341 (2020). [CrossRef]
12. Y. Ma, Y. Du, Y. Y. Chen, C. J. Gu, T. Jiang, G. D. Wei, and J. Zhou, “Intrinsic Raman signal of polymer matrix induced quantitative multiphase SERS analysis based on stretched PDMS film with anchored Ag nanoparticles/Au nanowires,” Chem. Eng. J. 381, 122710 (2020). [CrossRef]
13. X. F. Zhao, J. Yu, C. Zhang, C. S. Chen, C. S. Xu, C. H. Li, Z. Li, S. Z. Zhang, A. H. Liu, and B. Y. Man, “Flexible and stretchable SERS substrate based on a pyramidal PMMA structure hybridized with graphene oxide assivated AgNPs,” Appl. Surf. Sci. 455, 1171–1178 (2018). [CrossRef]
14. L. L. Lan, X. Y. Hou, Y. M. Gao, X. C. Fan, and T. Qiu, “Inkjet-printed paper-based semiconducting substrates for surface-enhanced Raman spectroscopy,” Nanotechnology 31(5), 055502 (2020). [CrossRef]
15. X. F. Guo, D. P. Wang, and R. Khan, “Nafion stabilized Ag nanopillar arrays as a flexible SERS substrate for trace chemical detection,” Mater. Chem. Phys. 252, 123291 (2020). [CrossRef]
16. Z. Y. Li, X. Huang, and G. Lu, “Recent developments of flexible and transparent SERS substrates,” J. Mater. Chem. C 8(12), 3956–3969 (2020). [CrossRef]
17. X. F. Zhao, C. D. Liu, J. Yu, Z. Li, L. Liu, C. H. Li, S. C. Xu, W. F. Li, B. Y. Man, and C. Zhang, “Hydrophobic multiscale cavities for high-performance and self-cleaning surface-enhanced Raman spectroscopy (SERS) sensing,” Nanophotonics 9(16), 4761–4773 (2020). [CrossRef]
18. C. H. Li, S. C. Xu, J. Yu, Z. Li, W. F. Li, J. Wang, A. H. Liu, B. Y. Man, S. K. Yang, and C. Zhang, “Local hot charge density regulation: Vibration-free pyroelectric nanogenerator for effectively enhancing catalysis and in-situ surface enhanced Raman scattering monitoring,” Nano Energy 81, 105585 (2021). [CrossRef]
19. C. Zhang, C. H. Li, J. Yu, S. Z. Jiang, S. C. Xu, C. Yang, Y. J. Liu, X. G. Gao, A. H. Liu, B. Y. Man, and A. B. Chemical, “SERS activated platform with three-dimensional hot spots and tunable nanometer gap,” Sens. Actuators, B 258, 163–171 (2018). [CrossRef]
20. J. Zhang, P. Y. Zhang, Y. M. Ding, X. L. Zhang, J. M. Quan, and Y. Zhu, “Ag-Cu nanoparticles encaptured by graphene with magnetron sputtering and CVD for surface-enhanced Raman scattering,” Plasmonics 11(6), 1495–1504 (2016). [CrossRef]
21. J. Zhang, Z. H. Yin, X. L. Zhang, and Y. Zhu, “Quantitative SERS by electromagnetic enhancement normalization with carbon nanotube as an internal standard,” Opt. Express 26(18), 23534–23539 (2018). [CrossRef]
22. S. C. Lu, T. T. You, N. Yang, Y. K. Gao, and P. G. Yin, “Flexible SERS substrate based on Ag nanodendrite-coated carbon fiber cloth: simultaneous detection for multiple pesticides in liquid droplet,” Anal. Bioanal. Chem. 412(5), 1159–1167 (2020). [CrossRef]
23. W. Xin, J. M. Yang, C. Li, M. S. Goorsky, L. Carlson, and I. M. De Rosa, “Novel strategy for One-Pot synthesis of gold nanoplates on carbon nanotube sheet as an effective flexible SERS substrate,” ACS Appl. Mater. Interfaces 9(7), 6246–6254 (2017). [CrossRef]
24. J. Zhang, X. L. Zhang, S. M. Chen, T. C. Gong, and Y. Zhu, “Surface-enhanced Raman scattering properties of multi-walled carbon nanotubes arrays-Ag nanoparticles,” Carbon 100, 395–407 (2016). [CrossRef]
25. S. K. Prabha, P. D. Sreehari, M. G. M. and S, and P. Sathian, “The effect of system boundaries on the mean free path for confined gases,” AIP Adv. 3(10), 102107 (2013). [CrossRef]
26. C. V. Thompson, “Solid-state dewetting of thin films,” Annu. Rev. Mater. Sci. 42(1), 399–434 (2012). [CrossRef]
27. T. P. Nolan, R. Sinclair, and R. Beyers, “Modeling of agglomeration in polycrystalline thin films: Application to TiSi2 on a silicon substrate,” J. Appl. Phys. 71(2), 720–724 (1992). [CrossRef]
28. W. W. Mullins, “Theory of thermal grooving,” J. Appl. Phys. 28(3), 333–339 (1957). [CrossRef]
29. J. M. Quan, J. Zhang, X. Q. Qi, J. Y. Li, N. Wang, and Y. Zhu, “A study on the correlation between the dewetting temperature of Ag film and SERS intensity,” Sci. Rep. 7(1), 14771 (2017). [CrossRef]
30. K. T. Miller, F. F. Lange, and D. B. Marshall, “The instability of polycrystalline thin films: experiment and theory,” J. Mater. Res. 5(1), 151–160 (1990). [CrossRef]
31. H. Wong, P. W. Voorhees, M. J. Miksis, and S. H. Davis, “Periodic mass shedding of a retracting solid film step,” Acta Mater. 48(8), 1719–1728 (2000). [CrossRef]
32. P. B. Johnson and R. W. Christy, “Optical constants of noble metals,” Phys. Rev. B 6(12), 4370–4379 (1972). [CrossRef]
33. A. B. Djurisic and E. H. Li, “Optical properties of graphite,” J. Appl. Phys. 85(10), 7404–7410 (1999). [CrossRef]
