Intense-Raman Scattering on Hybrid Au/Ag Nanoplatforms for the Distinction of MMP-9-Digested Collagen Type-1 Fiber Detection

Biosensors and Bioelectronics 72 (2015) 61–70

Contents lists available at ScienceDirect

Biosensors and Bioelectronics

http://d0956-56

n CorrNationaFax: þ8

E-m

journal homepage: www.elsevier.com/locate/bios

Intense Raman scattering on hybrid Au/Ag nanoplatforms for the dis-tinction of MMP-9-digested collagen type-I fiber detection

Kundan Sivashanmugan a, Jiunn-Der Liao a,b,n, Pei-Lin Shao a, Bernard Haochih Liu a,Te-Yu Tseng a, Chih-Yu Chang a

a Department of Materials Science and Engineering, National Cheng Kung University, 1 University Road, Tainan 70101, Taiwanb Center for Micro/Nano Science and Technology, National Cheng Kung University, 1 University Road, Tainan 70101, Taiwan

a r t i c l e i n f o

Article history:Received 12 February 2015Received in revised form28 March 2015Accepted 27 April 2015Available online 29 April 2015

Keywords:Au nanorodsNanoclustersSurface-enhanced Raman scattering4-aminobenzenethiolDigested type I collagen

x.doi.org/10.1016/j.bios.2015.04.09163/& 2015 Elsevier B.V. All rights reserved.

esponding author at: Department of Materil Cheng Kung University, 1 University Ro86 62346290.ail address: [email protected] (J.-D. Lia

a b s t r a c t

Well-ordered Au-nanorod arrays were fabricated using the focused ion beam method (denoted as fi-bAu_NR). Au or Ag nanoclusters (NCs) of various sizes and dimensions were then deposited on the fi-bAu_NR arrays using electron beam deposition to improve the surface-enhanced Raman scattering(SERS) effect, which was verified using a low concentration of crystal violet (10–5 M) as the probe mo-lecule. An enhancement factor of 6.92�108 was obtained for NCsfibAu_NR, which is attributed to thecombination of intra-NC and NR localized surface plasmon resonance. When 4-aminobenzenethiol (4-ABT)-coated Au or Ag nanoparticles (NPs) were attached to NCsfibAu_NR, the small gaps between 4-ABT-coated NPs and intra-NCs allowed detection at the single-molecule level. Hotspots formed at the inter-faces of NCs/NRs and NPs/NCs at a high density, producing a strong local electromagnetic effect. Ramanspectra from as-prepared type I collagen (Col-I) and Ag-NP-coated Col-I fibers on NCsfibAu_NR werecompared to determine the quantity of amino acids in their triple helix structure. Various concentrationsof matrix-metalloproteinase-9-digested Col-I fibers on NCsfibAu_NR were qualitatively examined at aRaman laser wavelength of 785 nm to determine the changes of amino acids in the Col-I fiber structure.The results can be used to monitor the growth of healing Col-I fibers in a micro-environment.

& 2015 Elsevier B.V. All rights reserved.

1. Introduction

The growth process during wound healing is of great interest.Skin tissue reconnection with active scaffolds or stem cells is oftenused to study cell differentiation and proliferation on the woundbed (Jain et al., 2014; Ngo et al., 2014; Ngo Thi et al., 2014). Bio-degradable and nonbiodegradable proteins and peptides are pro-mising supports for tissue reconnection owing to their usefulproperties, such as chemical structural diversity, large number ofarchitectures, mechanical strength, and biocompatibility (Desi-mone et al., 2011; Lin et al., 2012; Lv et al., 2014; Shoulders andRaines, 2009). A variety of natural proteins and peptides, such aslaminin, fibronectin, and collagen (Col) fibers, are often used toimprove the efficacy of tissue reconstruction (Silver et al., 2003;Stamov et al., 2008). Col fibers are most commonly used in tissueregeneration due to their excellent rates of biointegration, biode-gradability, and biocompatibility (Desimone et al., 2011; Lin et al.,2012; Lv et al., 2014; Silver et al., 2003; Stamov et al., 2008). The

als Science and Engineering,ad, Tainan 70101, Taiwan.

o).

triple helix structure of Col fibers may improve tissue growth viaseveral cross-linking procedures (Kadler et al., 1996; Shouldersand Raines, 2009). However, the degradation of Col fibers is also animportant process in wound healing, which generally occurs dur-ing the growth and development of adult connective tissueswithin pericellular and extracellular environments. In mature tis-sues, existing Col fibers are replaced by new Col fibers (Desimoneet al., 2011; Lin et al., 2012; Lv et al., 2014; Silver et al., 2003;Stamov et al., 2008).

Complex tissue reconstruction requires type I Col (Col-I) fiberprotein, which is an important structural component in bloodvessels, skin, tendons, ligaments, and bone (Bozec et al., 2007;Fang et al., 2013). The degradation of Col-I, still unclear in tissuereconstruction, can occur through either the extracellular or in-tracellular pathway (Fligiel et al., 2003; Orza et al., 2011; Sun et al.,2008b). The growth and degradation of Col-I fibers is tightlyregulated. From in vitro studies, matrix metalloproteinases(MMPs), including MMP-1, -9, -13, and -14, can effectively degradenative Col-I fibers surrounding a cell surface through the glycine inCol-I fibers (Aimes and Quigley, 1995; Bigg et al., 2007; Fini et al.,1992; Knauper et al., 1997). In addition, the degradation of Col-I,-II, and -III fibers using various temperature-dependent treat-ments has been investigated (Cárcamo et al., 2012; Jain et al., 2014;

K. Sivashanmugan et al. / Biosensors and Bioelectronics 72 (2015) 61–7062

Su et al., 2014). Methods for analyzing the degradation of Col-Ifibers include the enzyme-linked immunosorbent assay and high-performance liquid chromatography (Akhtar et al., 1999; Hudsonet al., 2012), which have excellent sensitivity for the detection ofspecific degraded enzymes in Col-I fibers. However, labelingtechniques usually require sophisticated procedures for samplepreparation and subsequent data analysis, as well as repetitivetests, making them high-cost. Therefore, a simple, quick, cost-ef-fective, and sensitive detection method for monitoring Col-I fibersis highly required.

Raman spectra provide structural and chemical informationabout target species in a chemical or biological sample (Luo et al.,2014; Xie and Schlucker, 2013). However, the very low efficiency ofRaman data collection leads to long acquisition time for detectingnormal analyte molecules. In addition, biological samples, e.g.,cells, proteins, and viruses, require a method with high resolutionor enhancement and sensitivity or selectivity. A strong Ramansignal is generated when a Raman-active analyte interacts with anoble metal, e.g., Ag, Au, or Cu, surface. A surface-enhanced Ramanscattering (SERS)-active substrate allows detection down to asingle-molecule level (Betz et al., 2014; Harper et al., 2013;Kleinman et al., 2013). The effect of SERS is mainly attributed totwo primary mechanisms: the chemical effect and the electro-magnetic (EM) effect (Betz et al., 2014; Harper et al., 2013;Kleinman et al., 2013). Very few studies have investigated Col-Ifiber structure using SERS, e.g., by detecting signals from an Au orAg nanoparticle (NP)-coated Col-I fiber surface (Orza et al., 2011;Sun et al., 2008b). Ag-NP-coated Col-I fiber produces strong Ra-man signals, but with low reproducibility. Alternatively, nano-lithography techniques such as electron beam lithography (Luoet al., 2014), focused ion beam (FIB) (Sivashanmugan et al., 2013a,2013b), and nanomechanical indentation (Chang et al., 2011; Yaoet al., 2012) have been employed to fabricate various types ofnanostructure (NS), e.g., nanorods (NRs), nanowires, nanocavities.Nanofabrication methods create highly precise and regular pat-terns that have a high enhancement factor (EF). SERS-active sub-strates have been used to detect nucleoproteins, oligonucleotides,and viruses due to the strong light scattering in the vicinity ofnanogaps and NSs (Shao et al., 2014; Sivashanmugan et al., 2015,2013a).

In our previous work, SERS-active substrates with well-ordered

Fig. 1. Schematic illustrations of experimental route for (a) as-fabricated NCsfibAu_NR,MMP-9-digested Col-I fibers on NCsfibAu_NR.

Au or Au/Ag multilayered NR arrays were fabricated using FIBtechnology (Sivashanmugan et al., 2015, 2013a, 2013b). It is hy-pothesized that a high EM effect due to the availability of multipleedges and a small curvature induced in the substrate (Costa et al.,2009; Shao et al., 2014; Sivashanmugan et al., 2014, 2015, 2013a).In addition, the localized surface plasmon resonance (LSPR) effectstrongly depends on NR shape and the spacing between NRs. TheLSPR effect is known to increase SERS sensitivity up to the single-molecule level. Notably, the SERS effect highly varies with the sizeof the target species due to laser interaction strength. Large bio-target species, e.g., cells and proteins, usually require a substratewith strong and highly reproducible LSPR. Therefore, a substratewith a high SERS effect is required for practical applications (Queet al., 2011, 2012).

In the present work, FIB and electron beam deposition werecombined for the fabrication of Au and Ag nanoclusters (NCs) on aFIB-made Au NR arrays (NCsfibAu_NR) as a SERS substrate, as il-lustrated in Fig. 1(a). The Au or Ag NCs on fibAu_NR influence thedetection of Raman-active species. This influence was examinedusing a low concentration (10–5 M) of crystal violet (CV) as theprobe molecule. As shown in Fig. 1(b), 4-aminobenzenethiol (4-ABT)-coated Au or Ag NPs were deposited on an optimizedNCsfibAu_NR substrate to create a large plasmonic effect via alarge number of hotspots. As shown in Fig. 1(c), Col-I, Ag-NP-coated Col-I fibers, and MMP-9-digested Col-I fibers were de-posited on an optimized NCsfibAu_NR substrate to determinechanges in Col-I fiber structure. The former application utilizes atarget chemical species bonded with NPs and attached to NCs®-bAu_NR, while the latter utilizes a target biological sample withNPs and attached to NCsfibAu_NR. These applications were used toconfirm whether the hybrid Au/Ag nanostructure can intensify theRaman scattering effect on the target species simply with NPs.

2. Experimental section

2.1. Fabrication of Au or Ag NCs on fibAu_NR

Au thin film was deposited onto polished single-crystal silicon(100) wafers primed with a 5- nm-thick adhesion layer of Ti usingan electron beam evaporator (VT1-10CE, ULVAC, Taiwan) at a

(b) 4-ABT-coated Au or Ag NPs on NCsfibAu_NR, and (c) Col-I, NP-coated Col-I, and

K. Sivashanmugan et al. / Biosensors and Bioelectronics 72 (2015) 61–70 63

controlled deposition rate. The optimal layer thickness for Au film(E420 nm) was maintained. Patterns were designed using Cor-elDRAW software. The NR designs were implemented by applyingan FIB (SMI 3050, SII Nanotechnology, Japan). The pattern size wasabout 120 mm�120 μm. The following beam conditions wereused: 30-kV acceleration voltage, 0.07-μm depth, 10-pA aperture,70-μs dwell time, þ0.56 optical lens, and 8-image scale (Siva-shanmugan et al., 2013a, 2013b). The spacing between adjacentNRs, diameter, and shape of Au NR arrays (fibAu_NR) were main-tained by adjusting the beam current. Au or Ag NCs were thendeposited on the fibAu_NR arrays to increase the effect of SERS, asillustrated in Fig. 1(a).

Au or Ag NCs on fibAu_NR were prepared using electron beamevaporation deposition with estimated diameters of 2, 5, and8 nm; the corresponding samples are denoted as G_1fibAu_NR,G_2fibAu_NR, and G_3fibAu_NR for Au NCsfibAu_NR andS_1fibAu_NR, S_2fibAu_NR, and S_3fibAu_NR for Ag NCsfibAu_NR.The size of the as-prepared NCs was based on the spacing betweenNRs (D-spacing, i.e., E35 nm). Information about all NCs andNCsfibAu_NR are given in Tables S1 and S2 (in Supporting In-formation), respectively. Field-emission scanning electron micro-scopy (FE-SEM, JSM-7001, JEOL, Japan), atomic force microscopy(AFM, Bruker, ICON SPM, USA) and high-resolution field-emissiontransmission electron microscopy (HR-FETEM, JEM-2100 F, JEOL,Japan) with energy-dispersive X-ray spectroscopy (EDS) wereemployed to analyze the morphologies of the as-prepared NCs,fibAu_NR, and NCsfibAu_NR.

2.2. EF determination using molecular probe

The molecular probe CV was diluted in aqueous solution to aconcentration in the range of 10–5–10�15 M. To verify EF, fibAu_NRor NCsfibAu_NR with a molecular-probe-containing solution wascovered with a glass slide and then immediately measured usingRaman spectroscopy. Raman spectra were acquired using a con-focal microscopy Raman spectrometer (inVia Raman microscope,Renishaw, United Kingdom) with diode lasers at an excitationwavelength of 785 nm, and then scanned with an integration timeof 10 s over an area of 1 μm�1 μm (the size of the laser spot was�1 μm) using a 50� objective. Before each batch, the Raman shiftwas calibrated using the absolute intensity of signal at 520 cm�1

from a standard silicon wafer.The EF measurement was estimated according to the standard

equation (Sivashanmugan et al., 2013a, 2013b)

EFII

NN 1

sers

bulk

bulk

sers= ×

( )

where Isers and Ibulk are the SERS and normal Raman scatteringintensities, respectively, and Nsers and Nbulk are the numbers ofmolecules contributing to the inelastic scattering intensity re-spectively evaluated by SERS and normal Raman scattering mea-surements. The Raman intensity was averaged from 10 con-secutive measurements. The highest peak intensity for all sampleswas recorded as the specific band and treated as the Isers value inEq. (1). The ideal average Nsers value (E6.17�107 molecules) wasobtained from the volume of the uniform monolayer of CV ideallyadsorbed on NCsfibAu_NR. The Nsers value is correlated with theconcentration of 10�5 M of CV-containing solution. The spot sizeof the laser and the average value of Nbluk were obtained asE1 μm2 and E9.23�108 molecules, respectively. The Ibulk valueswere measured as E280 (arbitrary units). The above values weresubstituted into Eq. (1) to calculate the SERS EF values for samples.

2.3. Fabrication of 4-ABT-coated Au or Ag NPs on NCsfibAu_NR

Gold (III) chloride trihydrate (HAuCl4 �3H2O), silver nitrate(AgNO3), sodium citrate (HOC(COONa)(CH2COONa)2 �2H2O), and4-ABT (H2NC6H4SH) were purchased from Sigma-Aldrich. Ultra-pure water (E10 MΩ cm) was used as a solvent in the experi-ments. All glassware used in the experiments was cleaned withnitric acid before use.

The NPs to be coated with 4-ABT for deposition on NCs®-bAu_NR were prepared based on a well-established method (Si-vashanmugan et al., 2013b). Firstly, an Au sol was prepared bydissolving 1 mM HOC(COONa)(CH2COONa)2 �2H2O and 1 mL of0.01 mM HAuCl4 �3H2O in 20 mL of aqueous solution and thenboiled for 30 min. In this mixture solution, Au seed turned reddishafter 4 min. After continuous stirring for an additional 5 min, themixture was allowed to stand at room temperature, and thenwashed five times using water with some specific parameters asfollows: the radius and rotating speed were 5 cm and 12,000 rpm,respectively, to remove impurities and stored for further studies.

For the growth of Ag sol, 100 mL of aqueous solution containing1 mM AgNO3 was brought near boiling temperature. 3 mL of Agseed solution and HOC(COONa)(CH2COONa)2 �2H2O (with a finalconcentration of 1 mM in solution) was then added to the boilingsolution. The mixture was heated until the color of the solutionturned greenish yellow. The solution was removed and brought toroom temperature, washed five times using water with somespecific parameters as follows: the radius and rotating speed were5 cm and 12,000 rpm, respectively, to remove impurities, and thenstored for further studies. The morphology and structure of theNPs were characterized using HR-FETEM. The optical properties ofthe NPs were monitored as a function of time in 10- mm-optical-path-length quartz cuvettes with an ultraviolet-visible (UV–vis)spectrophotometer (UV–3600, Shimadzu).

The as-prepared Au or Ag NPs were then mixed with an opti-mized concentration of 5 mL of 4–ABT (10 mM ethanolic solution)and stabilized for 10 h. The solution was then washed with waterand ethanol in sequence. Au or Ag NCsfibAu_NR was then coveredwith a droplet containing 4-ABT-coated Au or Ag NPs and dried ina vacuum for 2 h, washed with ethanol, and then dried in air. The4-ABT-coated Au or Ag NPs on Au, fibAu_NR and NCsfibAu_NRsubstrates were examined at a Raman laser wavelength of 785 nm.The morphology and structure of 4-ABT-coated Au or Ag NPs onNCsfibAu_NR were respectively characterized using HR-FETEM andAFM.

2.4. Raman studies of Col-I fibers

Rat tail Col-I was purchased from Merck Millipore and thendiluted to 0.1 mg/mL in a 10-mL solution of phosphate-bufferedsaline (PBS) before use. The as-prepared samples were incubatedat 37 °C for 30 min to grow Col-I fibers. Col-I was diluted to0.1 mg/mL in a 5-mL solution of PBS with 5 mL (an optimizedconcentration) of Ag NPs and then incubated at 37 °C for 1 h tocreate NP-coated Col-I fibers. Col-I and Ag-NP-coated Col-I fiberswere then placed on an optimized NCsfibAu_NR. After 15 min, thesamples were dried with nitrogen gas and subjected to Ramanspectroscopic studies. The morphology and structure of Col-I andAg-NP-coated Col-I fibers on an optimized NCsfibAu_NR wascharacterized using AFM. MMP-9 was purchased from MerckMillipore. To activate MMP-9, MMP-9 with a concentration of0.005 or 0.01 mg/mL was diluted in a 10-mL solution of PBS. Todigest Col-I fibers, the fibers were mixed with various concentra-tions of MMP-9, and then incubated at 27 °C for 30 min. The MMP-9-digested Col-I fibers were then placed on an optimized NCs®-bAu_NR and subjected to Raman analysis. The morphology andstructure of the digested Col-I fibers on an optimized NCsfibAu_NR

K. Sivashanmugan et al. / Biosensors and Bioelectronics 72 (2015) 61–7064

were characterized using AFM.

3. Results and discussion

3.1. Characterization of as-fabricated Au or Ag NCsfibAu_NR

As shown in Fig. S1, the morphology of fibAu_NR was analyzedusing FE-SEM, AFM, and HR-FETEM. The specified spacing be-tween adjacent NRs was maintained by adjusting the workingcurrent and etching time during FIB fabrication. Fig. S1(a) and(b) show top- and side-view images of fibAu_NR, respectively. Thelength, diameter, and spacing between NRs for fibAu_NR wereE400, E170, and 35 nm, respectively. The AFM topography offibAu_NR is shown in Fig. S1(c). The as-formed NRs were uniformlyfabricated within the patterned area, as indicated by the AFM lineprofiles shown in the inset in Fig. S1(c).

The Au or Ag NCs were anticipated to slightly reduce the spa-cing between adjacent NRs and the tip ring diameter of NRs (Si-vashanmugan et al., 2013b). To confirm the size, dimension, anddistribution of Au or Ag NCs, the NCs were deposited on an Si (100)substrate and the resulting values were used as the reference. Fig.S2(a–f) show two-dimensional top-view AFM and FE-SEM imagesof the as-prepared Au or Ag NCs, evenly distributed on an Si (100)substrate. The gap between NCs significantly increased with

Fig. 2. FE-SEM top- and lateral-view (insets) images of Ag NCs on fibAu_NR: (a) S_1fibFETEM images of Ag NCs on fibAu_NR: (a-i) S_1fibAu_NR, (b-i) S_2fibAu_NR, and (c-i) S_regions in (a�c-ii) indicate thickness of Ag NCs on fibAu_NR surface in (a�c-iii).

increasing size and dimension of the deposited Au or Ag NCs. Inaddition, the average roughness (Ra) of Ag NC sample S_3(RaE4.33 nm) was higher than that of Au NC sample G_3(RaE2.68) due to the shape and physical properties of Au or AgNCs (Rycenga et al., 2009; Sivashanmugan et al., 2013b; Yamamotoet al., 2013; Yang et al., 2014), as listed in Table S1.

The formation and distribution of Au or Ag NCs on varioussubstrates were respectively examined using FE-SEM and HR-FE-TEM. Fig. 2(a–c) and S3(a–c) show side and cross-sectional (insets)FE-SEM images of Au or Ag NCsfibAu_NR. Both Au and Ag NCswere uniformly distributed on fibAu_NR. Furthermore, FE-SEMEDS spectra and elemental mapping of Au or Ag NCsfibAu_NRconfirm the formation (as shown in Fig. S4) and distribution of NCson fibAu_NR. Ag NCs on fibAu_NR were highly aggregated com-pared to Au NCs on fibAu_NR due to their varied deposition rates,e.g., by measuring the decrease of Au element from fibAu_NR, aslisted in Table S2. In addition, Ra values, measured from AFM, forAg NCsfibAu_NR ranged from 18 to 33 nm and followed the se-quence S_1fibAu_NR4S_2fibAu_NR4S_3fibAu_NR. These valueswere higher than the corresponding values for Au NCsfibAu_NRsamples G_1fibAu_NR, G_2fibAu_NR, and G_3fibAu_NR (13–21 nm). Notably, the EM effect, and thus the SERS effect, is sig-nificantly increased by surfaces with a relatively high Ra (Costaet al., 2009; Kleinman et al., 2013; Rycenga et al., 2009; Shao et al.,2014; Sivashanmugan et al., 2014, 2015, 2013a, 2013b; Yamamoto

Au_NR, (b) S_2fibAu_NR, and (c) S_3fibAu_NR (inset image scale bar: 50 nm). HR-3fibAu_NR. (a�c-ii) HR-FETEM images of Ag NC distribution on fibAu_NR. Marked

Fig. 3. Raman-active peaks of 10�5 M CV molecules on (a) Au NCs on fibAu_NR(G_1fibAu_NR, G_2fibAu_NR, and G_3fibAu_NR) and (b) Ag NCs on fibAu_NR(S_1fibAu_NR, S_2fibAu_NR, and S_3fibAu_NR) examined at Raman laser wave-length of 785 nm (insets show corresponding Raman spectra). Enhancement fac-tors and relative Raman intensities for Au or Ag NCs on fibAu_NR were compared.CV peak at 1618 cm�1 was used as index for relation of relative Raman intensitieswith respect to Au or Ag NCs on fibAu_NR.

K. Sivashanmugan et al. / Biosensors and Bioelectronics 72 (2015) 61–70 65

et al. 2013; Yang et al. 2014).Fig. 2(a�c-i) and S3(a�c-i) show HR-FETEM images of Ag or

Au NCsfibAu_NR. A relatively high density of NCs on NRs was ob-tained by varying the deposition angle (e.g., the Au or Ag targetwas placed perpendicular to fibAu_NR). The spacing between ad-jacent NRs was not fully enclosed by Ag or Au NCs, as shown inFig. 2(a–c-ii) and S3(a–c-ii). As shown in Fig. 2(a–c-iii) and S3(a–c-iii), a small gap (e.g.,o2 nm) formed between NCs on the sidesurface of NRs, which may greatly enhance Raman scatteringaround NCs. The geometry of NCs on NRs is also an importantfactor: increasing the size and dimension of Ag or Au NCs on AuNRs tended to decrease the D-spacing of NRs, as summarized inTable S2. The optimized sample S_3fibAu_NR had lattice spacingsof 0.26 and 0.24 nm for Au and Ag, respectively, which match thelattice spacing of (111) face-centered cubic planes (Fig. S5) (Ry-cenga et al., 2009; Yang et al., 2014). Moreover, a tiny gap(e.g.,o1 nm) formed at the interface between NCs (intra-NCs) andNRs (NCs on NRs), greatly increasing the effect of LSPR around NCsand NCs on NRs and therefore increasing the sensitivity of SERS(Rycenga et al., 2009; Yamamoto et al., 2013; Yang et al., 2014;Yilmaz et al., 2014).

3.2. SERS effect on fibAu_NR and Au or Ag NCsfibAu_NR

The sensitivity and enhancement effect of CV were determinedat a Raman laser wavelength of 785 nm. The insets in Fig. 3(a) and(b) show 10�5 M CV solution dispersed on fibAu_NR, Au NCs®-bAu_NR (G_1fibAu_NR, G_2fibAu_NR, and G_3fibAu_NR), and AgNCsfibAu_NR (S_1fibAu_NR, S_2fibAu_NR, and S_3fibAu_NR). Forcomparison, Raman spectra of 10�5 M CV solution dispersed overAu or Ag NCs on an Si substrate are shown in Fig. S6(a) and (b). TheRaman enhancement effect was intensified by increasing the sizeof NCs. In addition, Ag NCs were adsorbed and resulted in a higherRaman scattering effect as compared to Au NCs, presumably at-tributable to the adjustments of surface roughness and distribu-tion of NCs. The strongest CV Raman scattering was observed forAg NCsfibAu_NR. The most intense Raman shift from the char-acteristic peaks of CV usually appears at 1618 cm�1 (CVpeak),which is assigned to the ring C–C stretching modes. The otherRaman band shifts are listed in Table S3 (Sivashanmugan et al.2015, 2013a, 2013b). Based on Eq. (1), the estimated EF values forthe samples are shown in Fig. 3(a) and (b). Ag NCsfibAu_NR sampleS_3fibAu_NR exhibited the highest Isers and EF (6.92�108). Sincethe intra-Ag NC and Au NR plasmon energies are most probablymatched that increase the electron oscillation at Raman-activesites, the intensity of Raman signals and the value of EF thus in-creased (Lee et al., 2013; Rycenga et al., 2009; Sivashanmuganet al., 2015, 2013a, 2013b; Yamamoto et al. 2013; Yang et al. 2014;Yilmaz et al. 2014). In Fig. S7, strong Raman intensities at1618 cm�1 for low concentrations of 10�7, 10�9, 10�13, and10�15 M of CV on S_3fibAu_NR can be seen. Presumably, the highRa value of NCs on the surfaces of NRs, along with the SPR effectaround NCs and the NC/NR interface and the LSPR effect aroundNRs, contributes to the improved low-concentration single-mole-cule detection.

3.3. Quality of 4-ABT-coated Au or Ag NPs on S_3fibAu_NR

The optimized Ag NCsfibAu_NR sample S_3fibAu_NR was uti-lized as the substrate for attachment of 4-ABT-coated Au or AgNPs. The shape-controlled Au or Ag NPs were analyzed using UV–vis absorption spectroscopy and HR-FETEM. Fig. S8(a) and(b) shows the UV–vis spectra peak position for Au or Ag NP so-lution, which appeared at 525 nm for Au and 425 nm for Ag. Theresults correspond well to characteristic nanometer-scale sizes ofAu or Ag NPs. HR-FETEM images of Au or Ag NPs are shown in the

insets of Figs. 4(a) and (b) and S8. Based on the HR-FETEM images,Au or Ag NPs were spherical with an average diameter of 20 nm.

In Fig. 4(a) and (b), the likely interaction (or slight aggregation,as shown in the insets) of 4-ABT-coated Au or Ag NPs is shown inthe HR-FETEM images. The 10�7 M 4-ABT mixed with Au or Ag NPsolution after 10 h is shown in the insets of Fig. 4(a-i) and (b-i).The interactions of thiol and NH2 groups of 4-ABT with Au or AgNPs are illustrated in Fig. 4(a-i) and (b-i) (Abdelsalam, 2009; Kimet al., 2012). In general, the newly formed bonds between sulfur/Au and NH2/Ag are very strong; however, the bond strength differsfrom that of metals. Notably, 4-ABT-coated Ag NPs exhibit moreaggregation than that of 4-ABT-coated Au NPs, since a relativelystrong covalent bond is formed between Ag and NH2. On the otherhand, few layer of surfactant citrate may significantly influence theaggregations of sample with 4-ABT.

Fig. 4(c) and (d) show HR-FETEM images of the top surface of4-ABT-coated Au and Ag NPs, respectively, on S_3fibAu_NR. Thearea within the dotted lines has small gaps between 4-ABT-coatedAu or Ag NPs and Ag NCs/Au NR (o2 nm). The 4-ABT-coated AgNPs exhibited a relatively large aggregation on S_3fibAu_NR ascompared in Figs. 4(c-i) and (d-i) owing to the increased surface

Fig. 4. HR-FETEM images of as-fabricated 4-ABT-coated (a) Au and (b) Ag NPs. Insets show HR-FETEM images of Au and Ag NPs. Schematic diagrams of 4-ABT-coated (a-i) Auand (b-i) Ag NPs. Insets show 4-ABT-coated Au or Ag NPs in solution. HR-FETEM images of 4-ABT-coated (c) Au and (d) Ag NPs on S_3fibAu_NR. High-density NPs distributedon top surfaces of NR (c�d-i) are shown. The two inter-junction gaps between (d-ii) Ag NCs on Au NR and (d-iii) Ag NPs on Ag NCs attached on Au NR surface are shown.(e) Two- and (e-i) three-dimensional AFM images of optimized 4-ABT-coated Ag NPs on S_3fibAu_NR (scale bar: 170 nm).

K. Sivashanmugan et al. / Biosensors and Bioelectronics 72 (2015) 61–7066

chemistry of Ag (relatively dark) caused by 4-ABT (Kim et al.,2011a, 2011b). In Fig. 4(d-ii) and (d-iii), the inter-junction gapbetween Ag NCs and Au NR (o3 nm) and that between 4-ABT-coated Ag NPs and Ag NCs (o3 nm) can be seen. In Figs. 4(e) and(e-i) and S9, two- and three-dimensional AFM and HRTEM imagesof 4-ABT-coated Ag NPs on S_3fibAu_NR are shown, respectively;the distribution of 4-ABT-coated Ag NPs was suggested to the re-sulted HR-FETEM image in Fig. 4(d).

In the 4-ABT-coated Au or Ag NPs on Ag NCs (or S_3) fibAu_NRsystem, the effect of SERS with low-concentration (10�7 M) 4-ABTwas evaluated at a Raman laser wavelength of 785 nm. In Fig. 5(a) and (b), strong Raman bands appear, which are attributed tothe b2-type bands of 4-ABT (as listed in Table S4), in particular forsample N_6 (i.e, 4-ABT-coated Ag NPs on S_3fibAu_NR). The mostintense Raman shift appeared at 1586 cm�1, which is assigned tothe ring C–C stretching modes (Abdelsalam, 2009; Cao and Che,2014; Kim et al., 2012, 2011a, 2011b; Shao et al., 2014; Uetsukiet al., 2010), most probably due to a strong electronic chargetransfer between 4-ABT-coated Ag NPs and Ag NCsfibAu_NR at thesmall gaps, as shown in Fig. 4(d-ii) and (d-iii) (Abdelsalam, 2009;Cao and Che, 2014; Kim et al., 2012, 2011a, 2011b; Shao et al.,2014; Uetsuki et al., 2010). In Fig. 5(c) and (d), Isers at 1586 cm�1

and they integrated surface areas are compared. The results

indicate that the increase in the SERS effect is most likely due tothe excitation of electrons oscillation at the nanosized gaps of NPs/NCs and NCs/NR (Abdelsalam, 2009; Dendisová et al., 2013; Kimet al., 2012; Kim and Lee, 2005; Shao et al., 2014; Sivashanmuganet al., 2015, 2013b).

3.4. Quality assessment of Col-I fibers

As illustrated in Fig. 6(a), Col-I fibers form a triple helicalstructure that consists of three polypeptide chains (R) that repeat atriplet amino acid sequence (Jalan et al., 2014; Kadler et al., 1996;Shoulders and Raines, 2009). The three R chains twist togetherinto a unique triple helical structure, whose length and diameterrange from 100 to 500 nm and 30 to 100 nm, respectively. Thelength and diameter of Col-I fibers strongly depend on the in-cubation time (Capaldi and Chapman, 1984; Jalan et al., 2014).

To verify the sensitivity of SERS on Ag NCsfibAu_NR (S_3), AgNP-coated Col-I fibers were placed on the substrate and measured.Since the peptides are easily incorporated through their bio-chemical immobilization at the Ag surface without additionalmodification of the Col-I fiber structure (Orza et al., 2011; Sunet al., 2008a, 2008b), Ag NPs and the Col-I fiber surface exhibitcross-linking, as shown in Fig. S10(a) and (b). The incorporation of

Fig. 5. Raman-active peaks of 10�7 M 4-ABT and that coated on (a) Au and (b) Ag NPs on S_3fibAu_NR examined at Raman laser wavelength of 785 nm. (c, d) Peak position at1586 cm�1 (4-ABT) was used as index for comparison between integrated surface area of Raman peak position and relative Raman intensities of samples. T_1 and N_1: 4-ABTon Au; T_2 and N_2: 4-ABT-coated Au and Ag NPs on Au; T_3 and N_3: 4-ABT on fibAu_NR; T_4 and N_4: 4-ABT-coated Au and Ag NPs on fibAu_NR; T_5 and N_5: 4-ABT onS_3fibAu_NR; T_6 and N_6: 4-ABT-coated Au and Ag NPs on S_3fibAu_NR.

K. Sivashanmugan et al. / Biosensors and Bioelectronics 72 (2015) 61–70 67

Ag NPs leads to an increase of structural disorder on Col-I fibers.For comparison, Fig. S11(a) shows Rama spectra for Col-I fibers, AgNP-coated Col-I fibers on S_3fibAu_NR, Col-I fibers on a glasssubstrate, Ag NP-coated Col-I fibers on a glass substrate, Col-I fi-bers on fibAu_NR, and Ag NP-coated Col-I fibers on fibAu_NR.Among them, Ag NP-coated Col-I fibers on S_3fibAu_NR exhibitedthe highest Raman signals, i.e., those for amides I and II, Pro orHyp, and (Gly-X-Y)n (Gullekson et al., 2011; Lee et al., 2010), andthe highest integrated area of the Raman peak at 810 cm�1 (Fig.S11(b)). The characteristic Raman peak at 810 cm–1 is assigned tothe ʋ(C–C) in Pro and Hyp rings (other Raman bands are listed inTable S5) (Gullekson et al., 2011; Lee et al., 2010). The resultssuggest that the coupling of Ag NP-coated Col-I fibers attached toNCsfibAu_NR enhances the LSPR near the NPs/NCs-NR system.

The sensitivity of SERS for the detection of digested Col-I fiberswas verified. In general, the proteolysis of the MMP family cancleave Col and native Col- I, II, and III fibers at a specific site in allthree chains of the triple helix structure, with the cleavage startingat approximately the end of C- and N-terminus (Knauper et al.,1997; Kridel et al., 2001; Tam et al., 2002; Verma and Hansch,2007). Among the MMP family, MMP-9 is most important becauseit is generated on diseased cell surfaces (Kridel et al., 2001). MMP-9 has three common domains, namely propeptide, catalytic, andhemopexin-like domains, the latter of which is linked to the cat-alytic domain by a flexible hinge region, as illustrated in Fig. 6(b).In this study, Col-I fibers were digested by MMP-9 at concentra-tions of 0.005 and 0.01 mg/mL.

Fig. 6(c) and Fig. S12 shows the two-dimensional AFM topo-graphy of Col-I fibers (i and ii), 0.005 mg/mL MMP-9-digested Col-I

fibers (iii and iv), and 0.01 mg/mL MMP-9-digested Col-I fibers (vand vi). The inset of Fig. 6(c-i and ii) shows an AFM image of theas-formed twisted triple helical structure of a Col-I fiber. SinceMMP-9 contains catalytic and hemopexin C domains, the ability tobind with Col-I fibers and slowly start to cleave Col-I fibers arethought to be mediated (Aimes and Quigley, 1995; Knauper et al.,1997; Sun et al., 2008a; Verma and Hansch, 2007), as shown inFig. 6(c-v and vi).

Fig. 6(d) shows Raman spectra of Col-I fibers, 0.005 mg/mLMMP-9-digested Col-I fibers, and 0.01 mg/mL MMP-9-digestedCol-I fibers on S_3fibAu_NR. Strong Raman signals from amide I(1664 cm�1), Pro and Hyp (1562 cm�1), amide III (1230 cm�1),and ʋ(C–C) in Pro and Hyp rings (810 cm�1) for Col-I fibers onS_3fibAu_NR can be seen in Fig. 6(d-i) and are listed in Table S5(Cárcamo et al., 2012; Gullekson et al., 2011; Lee et al., 2010). Thisconfirms that S_3fibAu_NR is the most SERS-active substrate,which likely creates a strong local EM effect between intra-NCsand at the NCs/NR interface.

In Fig. 6 (d-ii and iii), Raman signals of amides I and III and Proand Hyp from Col-I fibers slightly decreased with increasing con-centration of MMP-9. When MMP-9 was added to Col-I fibers, theygradually lost their integrated structure due to the breaking ofhydrogen bonds between Gly peptide chains, as indicated in Fig. 6(c) and (d) (Aimes and Quigley, 1995; Fini et al., 1992; Knauperet al., 1997; Verma and Hansch, 2007). Raman spectra for0.005 mg/mL MMP-9 interacting with Col-I fibers (Fig. 6 (d-ii))show strong aromatic ring breathing at 1004 cm–1 and reducedamides I and II and Gly-X-Y Raman bands. Furthermore, the endsof C- and N-terminal peptides in Col-I fibers were cleaved when

Fig. 6. Schematic diagrams of structures of (a) Col-I fiber and (b) MMP-9. (c) Two-dimensional and color quantitative AFM topography images of (i and ii) Col-I fibers, (iii andiv) Col-I digested by 0.005 mg/mL MMP-9, and (v and vi) Col-I digested by 0.01 mg/mL MMP-9 on S_3fibAu_NR. Triple helix structure of Col-I fiber and its possibly denaturedstructure are respectively shown in insets of AFM images (i) and (v) (scale bar: 60 nm). (d) Raman-active peaks of (i) Col-I fibers, (ii) Col-I fibers digested by 0.005 mg/mLMMP-9, and (iii) Col-I digested by 0.01 mg/mL MMP-9 on S_3fibAu_NR.

K. Sivashanmugan et al. / Biosensors and Bioelectronics 72 (2015) 61–7068

the MMP-9 concentration was increased to 0.01 mg/mL, as shownin Fig. 6 (d-iii). It is known that an increased interaction of theMMP-9 hemopexin domain with Col-I fibers leads to highly ex-pressed Col-I fiber end terminals as hydrogen bonds between thethree chains are thoroughly broken (Aimes and Quigley, 1995; Finiet al., 1992; Knauper et al., 1997; Verma and Hansch, 2007), whichis confirmed by the disappearing Raman band, shown in Fig. 6 (d-iii), i.e., amide bands.

Temperature-denatured Col-I fibers on S_3fibAu_NR were stu-died. The Col-I fibers were rapidly denatured by temperature dueto lowered strength and fusion. Their morphology and Ramanspectra are shown in Fig. S13. It is difficult to distinguish tem-perature-denatured Col-I fibers from MMP-9-digested Col-I fibersfrom Raman spectra. Further study is required. The proposed EMmechanism of Ag NP-coated Col-I fibers on NCsfibAu_NR is illu-strated in Fig. 7. In Fig. 7(a), a relativity strong LSPR effect is ob-tained on the fibAu_NR surface, and the spacing between adjacentNRs and the length of NRs contribute to LSPR. The high SERS maybe induced by the combined NC SPR and NR LSPR effects fromNCsfibAu_NR, which occurred at the interface of NCs and NRs, asshown in Fig. 7(b). In addition, the intra-NC effect increased thelocal field effect around NCs and NRs. Interestingly, multi-polar

excitations are mostly generated at the surfaces of Ag NP-coated4-ABT or Col-I fibers on NCsfibAu_NR, as shown in Fig. 7(c) and (d).The generations of large plasmonic fields around nanogaps in NSssystems, contributing to the detection of single molecules, werealso theoretically and experimentally evidenced and thereaftersuggested that around nanogaps, the EM effect is significantlyenhanced (Qian et al., 2010; Chen et al., 2013). A similar effect wasobtained for 4-ABT-coated Ag NPs on NCsfibAu_NR. Notably, AgNP-coated 4-ABT on NCsfibAu_NR SERS effect systematically en-hances the LSPR effect as compared to the Ag NP-coated Col-I fi-bers on NCsfibAu_NR due to the size effect of target species.

4. Conclusion

Highly SERS-active substrates, Au or Ag NCs on fibAu_NR, arefabricated by combining FIB and electron beam deposition meth-ods. The diameter of clusters and the formation of NCs on NR affectthe SERS effect, which was evaluated using CV as a molecularprobe at low concentration. In this work, the EF of S_3 (Ag NCs)fibAu_NR significantly increased with an Ag cluster size of �8 nm(to 6.92�108). It is very likely that the intra-NC LSPR effect on the

Fig. 7. Proposed schematic diagrams of SERS effect on (a) Au NR, (b) Ag NCs on Au NR, (c) 4-ABT-coated Au NPs on Ag-NC-coated Au NR, and (d) Ag-NP-coated Col-I fibers onAu NR.

K. Sivashanmugan et al. / Biosensors and Bioelectronics 72 (2015) 61–70 69

surface of NRs contributes to the detection of single molecules. Inaddition, a double inter-juction hotspot effect resulted from4-ABT-coated Au or Ag NPs on S_3fibAu_NR, which provides morehotspot areas at the interface of NPs/NCs and NCs/NRs and sig-nificantly increases the number of SERS-active sites. The optimizedSERS-active substrate S_3fibAu_NR was applied to detect MMP-9-digested Col-I fibers under various MMP-9 concentrations. Theresults indicate that the Raman band signals of Col-I fibers de-crease with increasing MMP-9 concentration. Presumably, MMP-9-digested Col-I fibers may gradually degrade the Col-I fiberstructure through the scission of hydrogen bonds between Glypeptide chains. Au or Ag NCsfibAu_NR thus has potential for in-vestigating tissue growth in various bio-environments.

Acknowledgment

This work was financially supported by the Ministry of Scienceand Technology, Taiwan under grant 103-2221-E-006-067-MY3.

Appendix A. Supplementary material

Supplementary data associated with this article can be found inthe online version at http://dx.doi.org/10.1016/j.bios.2015.04.091.

References

Abdelsalam, M.E., 2009. Cent. Eur. J. Chem. 7 (3), 446–453.Aimes, R.T., Quigley, J.P., 1995. J. Bio. Chem. 270 (11), 5872–5876.

Akhtar, S., Meek, K.M., James, V., 1999. Cardiovasc. Pathol. 8 (4), 203–211.Betz, J.F., Yu, W.W., Cheng, Y., White, I.M., Rubloff, G.W., 2014. Phys. Chem. Chem.

Phys. 16 (6), 2224–2239.Bigg, H.F., Rowan, A.D., Barker, M.D., Cawston, T.E., 2007. FEBS J. 274 (5), 1246–1255.Bozec, L., van der Heijden, G., Horton, M., 2007. Biophys. J. 92 (1), 70–75.Cao, Q., Che, R., 2014. ACS Appl. Mater. Interfaces 6 (10), 7020–7027.Capaldi, M.J., Chapman, J.A., 1984. Biopolymers 23 (2), 313–323.Cárcamo, J.J., Aliaga, A.E., Clavijo, E., Brañes, M., Campos-Vallette, M.M., 2012. J.

Raman Spectrosc. 43 (2), 248–254.Chang, C.-W., Liao, J.-D., Shiau, A.-L., Yao, C.-K., 2011. Sens. Actuators B-Chem. 156

(1), 471–478.Chen, J., Qin, G., Wang, J., Yu, J., Shen, B., Li, S., Ren, Y., Zuo, L., Shen, W., Das, B., 2013.

Biosens. Bioelectron 44, 191–197.Costa, J.C., Ando, R.A., Sant'Ana, A.C., Rossi, L.M., Santos, P.S., Temperini, M.L., Corio,

P., 2009. Phys. Chem. Chem. Phys. 11 (34), 7491–7498.Dendisová, M., Havránek, L., Ončák, M., Matějka, P., 2013. J. Phys. Chem. C 117 (41),

21245–21253.Desimone, M.F., Helary, C., Quignard, S., Rietveld, I.B., Bataille, I., Copello, G.J.,

Mosser, G., Giraud-Guille, M.M., Livage, J., Meddahi-Pelle, A., Coradin, T., 2011.ACS Appl. Mater. Interfaces 3 (10), 3831–3838.

Fang, M., Goldstein, E.L., Matich, E.K., Orr, B.G., Holl, M.M., 2013. Langmuir 29 (7),2330–2338.

Fini, M.E., Girard, M.T., Matsubara, M., 1992. Acta Ophthalmol. Suppl. 202, 26–33.Fligiel, S.E., Varani, J., Datta, S.C., Kang, S., Fisher, G.J., Voorhees, J.J., 2003. J. Invest.

Dermatol. 120 (5), 842–848.Gullekson, C., Lucas, L., Hewitt, K., Kreplak, L., 2011. Biophys. J. 100 (7), 1837–1845.Harper, M.M., McKeating, K.S., Faulds, K., 2013. Phys. Chem. Chem. Phys. 15 (15),

5312–5328.Hudson, D.M., Kim, L.S., Weis, M., Cohn, D.H., Eyre, D.R., 2012. Biochemistry 51 (12),

2417–2424.Jain, R., Calderon, D., Kierski, P.R., Schurr, M.J., Czuprynski, C.J., Murphy, C.J., McA-

nulty, J.F., Abbott, N.L., 2014. Anal. Chem. 86 (8), 3764–3772.Jalan, A.A., Jochim, K.A., Hartgerink, J.D., 2014. J. Am. Chem. Soc. 136 (21),

7535–7538.Kadler, K.E., Holmes, D.F., Trotter, J.A., Chapman, J.A., 1996. Biochem. J. 316, 1–11, Pt

1.Kim, K., Kim, K.L., Shin, K.S., 2012. Analyst 137 (16), 3836–3840.Kim, K., Lee, H.B., Shin, D., Ryoo, H., Lee, J.W., Shin, K.S., 2011a. J. Raman Spectrosc.

42 (12), 2112–2118.Kim, K., Lee, H.S., 2005. J. Phys. Chem. B 109 (40), 18929–18934.Kim, K., Yoon, J.K., Lee, H.B., Shin, D., Shin, K.S., 2011b. Langmuir 27 (8), 4526–4531.

K. Sivashanmugan et al. / Biosensors and Bioelectronics 72 (2015) 61–7070

Kleinman, S.L., Frontiera, R.R., Henry, A.I., Dieringer, J.A., Van Duyne, R.P., 2013. Phys.Chem. Chem. Phys. 15 (1), 21–36.

Knauper, V., Cowell, S., Smith, B., Lopez-Otin, C., O’Shea, M., Morris, H., Zardi, L.,Murphy, G., 1997. J. Biolog. Chem. 272 (12), 7608–7616.

Kridel, S.J., Chen, E., Kotra, L.P., Howard, E.W., Mobashery, S., Smith, J.W., 2001. J.Biolog. Chem. 276 (23), 20572–20578.

Lee, H.M., Jin, S.M., Kim, H.M., Suh, Y.D., 2013. Phys. Chem. Chem. Phys. 15 (15),5276–5287.

Lee, S.M., Pippel, E., Moutanabbir, O., Gunkel, I., Thurn-Albrecht, T., Knez, M., 2010.ACS Appl. Mater. Interfaces 2 (8), 2436–2441.

Lin, J., Li, C., Zhao, Y., Hu, J., Zhang, L.M., 2012. ACS Appl. Mater. Interfaces 4 (2),1050–1057.

Luo, S.C., Sivashanmugan, K., Liao, J.D., Yao, C.K., Peng, H.C., 2014. Biosens. Bioe-lectron 61, 232–240.

Lv, J., Chen, L., Zhu, Y., Hou, L., Liu, Y., 2014. ACS Appl. Mater. Interfaces 6 (7),4954–4964.

Ngo, M.-H.T., Liao, J.-D., Shao, P.-L., Weng, C.-C., Chang, C.-Y., 2014. Plasma Process.Polym. 11 (1), 80–88.

Ngo Thi, M.-H., Shao, P.-L., Liao, J.-D., Lin, C.-C.K., Yip, H.-K., 2014. Plasma Process.Polym. 11, 1076–1088.

Orza, A., Soritau, O., Olenic, L., Diudea, M., Florea, A., Rus Ciuca, D., Mihu, C., Cas-ciano, D., Biris, A.S., 2011. ACS Nano 5 (6), 4490–4503.

Qian, L., Das, B., Li, Y., Yang, Z., 2010. J. Mater. Chem. 20, 6891–6895.Que, R., Shao, M., Zhuo, S., Wen, C., Wang, Suidong, Lee, S.T., 2011. Adv. Funct. Mater.

21, 3337–3343.Qu, L.L., Li, D.,W., Xue, J.Q., Zhai, W.L., Fossey, J.S., Long, Y.T., 2012. Lab Chip 12,

876–881.Rycenga, M., Hou, K.K., Cobley, C.M., Schwartz, A.G., Camargo, P.H.C., Xia, Y., 2009.

Phys. Chem. Chem. Phys. 11 (28), 5903–5908.Shao, F., Lu, Z., Liu, C., Han, H., Chen, K., Li, W., He, Q., Peng, H., Chen, J., 2014. ACS

Appl. Mater. Interfaces 6 (9), 6281–6289.Shoulders, M.D., Raines, R.T., 2009. Annu. Rev. Biochem. 78, 929–958.

Silver, F.H., Freeman, J.W., Seehra, G.P., 2003. J. Biomech. 36 (10), 1529–1553.Sivashanmugan, K., Liao, J.-D., Yao, C.-K., 2014. Appl. Phys. Exp. 7 (9), 092202.Sivashanmugan, K., Liao, J.-D., Yao, C.-K., 2015. Sens. Actuators B-Chem. 206,

415–422.Sivashanmugan, K., Liao, J.-D., You, J.-W., Wu, C.-L., 2013a. Sens. Actuators B-Chem.

181, 361–367.Sivashanmugan, K., Liao, J.D., Liu, B.H., Yao, C.K., 2013b. Anal. Chim. Acta 800, 56–64.Stamov, D., Grimmer, M., Salchert, K., Pompe, T., Werner, C., 2008. Biomaterials 29

(1), 1–14.Su, L., Cloyd, K.L., Arya, S., Hedegaard, M.A.B., Steele, J.A.M., Elson, D.S., Stevens, M.

M., Hanna, G.B., 2014. J. Biophoton. 7 (9), 713–723.Sun, Y., Wang, L., Sun, L., Guo, C., Yang, T., Liu, Z., Xu, F., Li, Z., 2008a. J. Chem. Phys.

128 (7), 074704.Sun, Y., Wei, G., Song, Y., Wang, L., Sun, L., Guo, C., Yang, T., Li, Z., 2008b. Nano-

technology 19 (11), 115604.Tam, E.M., Wu, Y.I., Butler, G.S., Stack, M.S., Overall, C.M., 2002. J. Biol. Chem. 277

(41), 39005–39014.Uetsuki, K., Verma, P., Yano, T.-A, Saito, Y., Ichimura, T., Kawata, S., 2010. J. Phys.

Chem. C 114 (16), 7515–7520.Verma, R.P., Hansch, C., 2007. Bioorg. Med. Chem. Lett. 15 (6), 2223–2268.Xie, W., Schlucker, S., 2013. Phys. Chem. Chem. Phys. 15 (15), 5329–5344.Yamamoto, Y.S., Hasegawa, K., Hasegawa, Y., Takahashi, N., Kitahama, Y., Fukuoka, S.,

Murase, N., Baba, Y., Ozaki, Y., Itoh, T., 2013. Phys. Chem. Chem. Phys. 15 (35),14611–14615.

Yang, Y., Zhang, Q., Fu, Z.W., Qin, D., 2014. ACS Appl. Mater. Interfaces 6 (5),3750–3757.

Yao, C.-K., Liao, J.-D., Chang, C.-W., Lin, J.-R., 2012. Sens. Actuators B-Chem. 174,478–484.

Yilmaz, M., Senlik, E., Biskin, E., Yavuz, M.S., Tamer, U., Demirel, G., 2014. Phys.Chem. Chem. Phys. 16 (12), 5563–5570.