1. Introduction

Over the past few decades, surface-enhanced Raman scattering (SERS) has emerged as a highly promising technique for the detection of small concentrations of chemical and biological agents [1–6]. When chemical or biological analyte molecules are placed in close proximity of metallic nanostructured surfaces, the Raman signals are substantially enhanced, thereby enabling extremely sensitive detection of these analytes using SERS. The increase in the Raman cross-section of the coupled molecule-substrate complex is attributed to electromagnetic (EM) enhancement and chemical enhancement. While chemical enhancement results from the charge transfer between the analyte molecules and the nanostructured substrate, EM enhancement results from an increase in localized optical fields at electromagnetic hotspots (i.e., regions of high electromagnetic fields) in the nanostructured substrate [2,7] at plasmon resonance wavelengths. The electromagnetic enhancement of the optical E-field occurs at both the emission and excitation wavelengths. Consequently, the total electromagnetic enhancement of the Raman signals is approximated to be proportional to the fourth power of optical E-field enhancement in the electromagnetic hotspot where the analyte is present [1,2].

SERS offers an efficient way of detecting trace amounts of biochemical species and consequently, it has been widely used in sensing applications. However, producing a reliable, easy-to-fabricate and robust SERS substrate with tunable SERS activity has remained a challenge. Many tunable SERS substrates developed in the past have shown tunability of the SERS-activity by changing some parameter of the substrate during fabrication [8,9] although reports on post-fabrication active tunability of the SERS activity have been limited. Recently, active tunability of SERS substrates has been achieved via thermal activation of a temperature-sensitive polymer film [10], by mechanical stretching of plasmonic nanostructures on elastomeric membranes [11], or by opto-thermal actuation of the substrate [12]. In addition, an electromechanically-tunable, suspended two-wire plasmonic nanoantenna was employed by Hecht et al. [13] to achieve tuning of the plasmon resonance wavelengths by the application of voltage. However, active tunability of SERS signals ― by the application of voltage ― has not been demonstrated so far. In this work, we describe a simple but effective strategy to develop an active SERS substrate, in which the SERS signal from a cluster of dye-coated plasmonic nanoparticles can be tuned and increased by over 100% upon the application of an actuating voltage. The advantage of using the field-controlled SERS signal tunability is that we can potentially develop a closed-loop fully programmable system to control such devices.

In this paper, we employ electroactive polymers (EAPs) to achieve the active tunability of the SERS signal. Electroactive polymers constitute a class of synthetic macromolecules that undergo a dimensional change in response to an applied electric potential [14]. Of all EAPs, dielectric EAPs (D-EAP) exhibit the largest actuation strain upon exposure to an electric-field, efficiently coupling input electrical energy and output mechanical energy [14]. The electromechanical response of the D-EAPs is initiated by the field-induced charge separation that generates a compressive (Maxwell) normal stress, σ_M=ɛɛ_oE², on the film surfaces along the transverse direction (see Fig. 1). The variables ɛ₀ and ɛ represent the permittivity of free space and dielectric constant of the elastomer, respectively, and E denotes the magnitude of applied actuating electric-field. Additionally, repulsive like-charges that accumulate along both electrode surfaces also act to stretch the film in its plane.

 

Fig. 1. (a) Schematic showing a dielectric electro-active polymeric (D-EAP) membrane with two pairs of compliant deformable electrodes to form the active areas― AA1 and AA2. PA represents the passive region (PA) of the D-EAP membrane, spanning the two active areas. (b) SERS dye-coated Ag nanoparticles are drop-coated on the passive area (PA) of the D-EAP film when no actuating field is applied. (c) Due to an actuating voltage, an in-plane expansive strain in the active areas, AA1 and AA2 results in a contractile strain in the passive region, PA.

Download Full Size | PDF

The actuated media in this research is a commercially available acrylic elastomer (VHB 4905, 3M Co., Minneapolis, MN), appropriately prestrained with two pairs of deformable electrodes strategically placed on the opposite sides of the film segments AA1 and AA2 in order to form the active areas (See Fig. 1). Prestraining of D-EAPs is essential to reduce the voltage needed to create the electric-field necessary for actuation [14]. The deformable electrodes were deposited on opposite sides of the active areas AA1 and AA2. Carbon black (CB) filled polydimethylsiloxane (PDMS) elastomer composite was sprayed, using a stencil mask, to create the stretchable and uniformly thin electrodes (See details of Experimental Methods in Supplement 1). A voltage applied between the opposite electrodes, i.e., between E1a and E1b, and between E2a and E2b ― results in the actuating electric-field necessary to activate the active areas AA1 and AA2, respectively (See Fig. 1). The synthesized SERS dye-coated silver nanoparticles (See details of synthesis of nanoparticles in Supplement 1) were drop-coated on the passive area of the D-EAP film, spanning the two active areas, labeled PA in Fig. 1. The active areas are positioned and activated to work in consort upon application of an actuating voltage, such as to generate an in- plane expansive areal strain in the active areas. The isochoric deformation of these segments (AA1 and AA2) results in a contractile strain in the passive middle portion, PA, of the D-EAP membrane (See Fig. 1). Consequently, the spacing betwen the silver nanoparticles is expected to decrease with an increase in the actuation voltage applied to the active areas, which can lead to an increase in the electromagnetic (EM) enhancement in the corresponding electromagnetic hotspots. In this paper, we have, thus, demonstrated voltage-controlled actuation of a dielectric electroactive polymer membrane containing SERS dye-coated plasmonic nanoparticles to achieve voltage-tunable, active, and reversible modulation of SERS signals.

2. Results and discussion

First, the effect of the applied voltage (applied in consort to the electrode pairs in active areas AA1 and AA2) on the in-plane actuation strain in the D-EAP membrane was evaluated for various values of the pre-actuation spacing, ‘S’ as shown in Fig. 2(a). The in-plane actuation strain (in the x-direction) in the passive segment (PA) of the D-EAP membrane as a function of applied voltage is plotted in Fig. 2(a) for the pre-actuation spacing values of 1 mm, 2 mm, and 5 mm. It was found that, for the proposed substrate, a higher actuation strain in the passive area is achieved for lower pre-actuation spacing between the active areas AA1 and AA2. Hence, a value of 1 mm was chosen as the pre-actuation spacing between the active areas for all the measurements carried out in this paper.

 

Fig. 2. (a) Effect of pre-actuation spacing, ‘S’, between conformable electrodes pairs on the actuation strain in the passive segment of the D-EAP membrane, (b) Top view of blue plastic frame on which D-EAP membrane was suspended. Electrodes were deposited on front (E1a, E2a) and back (E1_b, E2_b) of the film. Ag nanoparticles were drop-coated on the D-EAP membrane between the two electrode pairs. (c) SEM image of Ag nanoparticles on the D-EAP membrane and (d) TEM image of Ag nanoparticles before drop-coating on D-EAP membrane.

Download Full Size | PDF

The silver nanoparticles were synthesized using a seed-mediated method published previously (See details in experimental methods, Supplement 1). Subsequently, the silver nanoparticle solution and the aqueous solution of the SERS-active dye CVP (Cresyl Violet Perchlorate) were mixed in a 1:1 (v/v) ratio. A 2 mM concentration of CVP was employed to prepare the solution containing dye-coated silver nanoparticles. These dye-coated silver nanoparticles were then drop-coated (in the passive region) onto the prestrained polymeric film suspended on a plastic frame.

The ability of the SERS substrate to modulate the SERS signals from the dye-coated silver nanoparticles drop-coated on the D-EAP membrane was experimentally demonstrated by applying a voltage across the deformable electrodes. A portable Raman system (See Details of Experimental Methods in Supplement 1) with a 785-nm excitation wavelength was employed to measure the SERS-signals from the cluster of SERS dye-coated silver nanoparticles placed on the D-EAP membrane, as the voltage applied across the electrodes was reversibly varied from 0 kV to ∼4.5 kV. The experimental set-up employed for the SERS measurements is shown in Fig. S1 in Supplement 1. A close view of the electrodes deposited on the D-EAP membrane is shown in Fig. 2(b). Figure 2(c) shows the Scanning Electron Microscopy (SEM) image for the silver nanoparticles deposited on the D-EAP membrane. The Transmission Electron Microscopy (TEM) image of the nanoparticles employed for SERS measurements is shown in Fig. 2(d) (See Fig. S2 in Supplement 1 for more SEM/TEM images).

The SERS data was collected from the passive segment of the D-EAP membrane at various operating voltages, using CVP (Cresyl Violet Perchlorate) as the SERS dye (See Figs. 3(a)–(b)). As the voltage was increased from 0.0 kV to 4.5 kV, the baseline corrected SERS signal (i.e., the height of the Raman peak) at 593.71 cm⁻¹ increased from approximately 1400 to about 2800 counts (See Fig. 3(b)) which is an increase of ∼ 100%. As the applied voltage is increased, the in-plane expansive areal strain in the active areas increases. The isochoric deformation of the active segments results in an increase in the contractile strain in the passive middle portion of the D-EAP membrane. Consequently, the silver nanoparticles dropped on the passive portion of the membrane moved closer as the applied voltage was increased. It is well known that a decrease in the gap between plasmonic nanoparticles leads to an increase in the optical electric-field (E-field) enhancement between the nanoparticles.¹² Thus, it is expected that the optical E-field enhancement between the silver nanoparticles would increase on increasing the applied voltage. Further, as the electromagnetic enhancement in the Raman signal is approximately proportional to the fourth power of the optical E-field enhancement [2], the electromagnetic SERS enhancement shown in Fig. 3(b) shows an increase with the applied voltage. Therefore, it can be concluded that a 100% increase in the Raman signals obtained in our experiment is primarily attributable to the reduction of spacing between the Ag nanoparticles on the substrate due to the contractile strain in the passive segment of the D-EAP membrane on application of voltage. When the applied voltage was brought back to 0.0 kV, the SERS signal from the substrate reduced back to 1850 counts. On removal of the applied potential, the strain in the passive and active segments of the D-EAP is restored to a value close to its original level, leading to an increase in the inter-particle spacing and hence, lowering of the SERS signal as shown in Figs. 3(a)–(c). The lack of complete recovery is due to the hysteretic loss in D-EAP film due to its viscoelastic behavior. It must be noted that the viscoelastic behaviour of dielectric electoactive polymers depends on the coupled phenomena of creep and hysteresis [15,16].

 

Fig. 3. (a) SERS spectra obtained from Ag nanoparticles on the D-EAP membrane for different actuation voltages, (b) Baseline corrected peak SERS signals (for the primary CVP peak at 593 cm⁻¹) versus actuation voltage, showing an increase in SERS signals with actuation voltage (red curve) and a decrease in SERS signals as the voltage difference is removed (blue curve).Additional results of SERS measurements are given in Supplement 1.

Download Full Size | PDF

Figure 4 (Visualization 1) shows the behavior of the D-EAP membrane — with the silver nanoparticles drop-coated on the passive segment in between the electrodes — on the application and subsequent removal of the actuating voltage. It can be observed from Fig. 4 (Visualization 1) that as the applied voltage increases, the boundaries of the carbon black conformal electrodes can be seen expanding due to the areal strain produced in the active segments sandwiched in the electrodes. Consequently, the passive region with the drop-coated silver nanoparticles ― seen as the dark spot in Fig. 4 ― can be seen undergoing a contractile strain. Therefore, the silver nanoparticles sticking to the D-EAP membrane are expected to move closer and thus lead to higher SERS signals which agrees well with the results shown in Fig. 3.

 

Fig. 4. Top-view images of the electrically-actuated plasmonic substrate on application of voltage: (a) 0 kV, (b) 1.8 kV and (c) 3.3 kV captured using a video camera. As the applied voltage increases, the boundaries of the carbon black conformal electrodes can be seen expanding due to the areal strain produced in the active segments. The passive region with the drop-coated silver nanoparticles can be seen undergoing a contractile strain as V is increased from 0 to 3.3 kV (Visualization 1).

Download Full Size | PDF

In order to elucidate the effect of inter-particle gaps on the electromagnetic fields and SERS signals from a cluster of metallic nanoparticles, numerical modeling for a cluster of silver nanoparticles deposited on a D-EAP membrane, was carried out using Finite Difference Time Domain (FDTD) modeling (see Supplement 1 for details of Numerical Methods). We consider particles with diameters (D) of 60 nm and 75 nm in accordance with the average size of the nanoparticles in the acquired SEM and TEM images. The particles are assumed to be periodically arranged in the x- and y- directions with a gap, G, between them in the x-direction and a fixed inter-particle gap of 10 nm in the y-direction (see Fig. 5). As the inter-nanoparticle gap ‘G’ is expected to decrease with increase in the actuation strain (in the x-direction) on application of voltage to the electrodes, the effect of decreasing the gap (in the x- direction) on the EM SERS enhancement was simulated.

 

Fig. 5. SERS enhancement (|E|⁴/|E₀|⁴) versus the contractile actuation strain at a wavelength of 785-nm for periodic nanoparticle clusters with D = 60 nm and D = 75 nm on a D-EAP membrane. The pre-actuation gap between the nanoparticles in the x-direction is assumed to 10 nm and is reduced to 7.5 nm at an actuation strain of 25%.

Download Full Size | PDF

Figure 5 shows the change in the electromagnetic SERS enhancement (which is proportional to the fourth power of |E|/|E₀| ― where E is the E-field in the center of two nanoparticles and E₀ is the incident E-field) as a function of the actuation strain (%). It can be seen from Fig. 5 that as the actuation strain is increased from 0 to 25% (i.e., the gap between the nanoparticles in the x-direction decreases from 10 nm to 7.5 nm), the SERS signals increase by ∼ 200% and 154% from the zero-actuation strain values for D = 60 nm and D = 75 nm, respectively. This happens due to the strengthening of the dipole interactions between the nanoparticles on reducing the inter-particle gaps which results in an increase in the optical E-field enhancement. In our experiments, a pre-actuation spacing of 1 mm between the electrode pairs was chosen, which lead to a maximum actuation strain of 16% on application of voltage of 4.5 kV, as shown in Fig. 2(a). It can be seen from the simulated results in Fig. 5, that for an actuation strain of 16%, the SERS signals increase by ∼ 96% and 76% from the zero-actuation strain values for D = 60 nm and D = 75 nm, respectively. As the nanoparticles dropped on the membrane had a distribution of particles close to the simulated values of diameters, these results agree well with the experimentally obtained value for the increase in the SERS signal of ∼ 100% for an actuation voltage of 4.5 kV. The SERS enhancement versus the contractile actuation strain for the different values of diameters of the silver nanoparticles are given in Supplement 1 (See Supplementary Information, Note 4).

3. Conclusion

In conclusion, we have proposed a voltage-tunable SERS substrate that allows active modulation of SERS-intensity over a broad range of actuating voltage. The SERS-intensity is tuned by varying an actuating voltage applied to the D-EAP substrate. These substrates are inexpensive and can be easily fabricated using drop-coating nanoparticles onto a variety of easily available D-EAP films types (acrylic, silicone, polyurethane, etc.) based on the application requirements. These substrates offer the possibility of large-area fabrication if multiple parallel electrode pairs, in the proposed and other geometries, are developed on a large-area polymeric film. These voltage-tunable SERS substrates can be employed for efficient chemical and biological sensing applications. Hence, the tunable SERS substrates proposed in this paper exhibit active modulation of the SERS-signals, as well as an increase of the SERS signals, by the application of an actuating voltage.