sensors Article Article

Detection of of Copper(II) Copper(II) Ions Detection Ions Using Using Glycine Glycine on on Hydrazine-Adsorbed Gold Nanoparticles via Raman Spectroscopy 1,2, Nguyễn Ly 11, Chulhun Seo 22 and ** Nguy n Hoàng Ly and Sang-Woo Sang-Woo Joo 1,2, 11

DepartmentofofChemistry, Chemistry,Soongsil SoongsilUniversity, University,Seoul Seoul156-743, 156-743,Korea; Korea;[email protected] [email protected] Department DepartmentofofInformation InformationCommunication, Communication,Materials, Materials,and andChemistry ChemistryConvergence ConvergenceTechnology, Technology, Department SoongsilUniversity, University,Seoul Seoul156-743, 156-743,Korea; Korea;[email protected] [email protected] Soongsil * Correspondence: Correspondence:[email protected]; [email protected];Tel.: Tel.:+82-2-820-0434 +82-2-820-0434 22

Academic Editors: Jong Seung Kim and Min Hee Lee 19 August August 2016; Accepted: 13 13 October October 2016; Published: date 26 October 2016 Received: 19 2+ ions in environmental Afacile, facile,selective, selective,and andsensitive sensitive detection method Cuions Abstract: A Abstract: detection method forfor thethe Cu2+ in environmental and and biological solutions has been newly developed by observing the unique stretching peaks biological solutions has been newly developed by observing the unique CNCN stretching peaks at −1 upon the dissociative adsorption of glycine (GLY) in hydrazine buffer on gold at ~2108 cm ~2108 cm−1 upon the dissociative adsorption of glycine (GLY) in hydrazine buffer on gold nanoparticles (AuNPs). of of CuCu species on AuNPs was identified from X-ray nanoparticles (AuNPs). The Therelative relativeabundance abundance species on AuNPs was identified from photoelectron spectroscopy analysis.analysis. UV-Vis spectra indicated that the Authat particles aggregated X-ray photoelectron spectroscopy UV-Visalso spectra also indicated the Au particles 2+ complex. The to result in the color change owingchange to the owing destabilization induced by theinduced GLY-Cuby aggregated to result in the color to the destabilization the GLY-Cu2+ −1 could be observed to indicate the formation of the CN species CN stretching band at ~2108 cm −1 complex. The CN stretching band at ~2108 cm could be observed to indicate the formation of the 3+ , Fe2+ , Hg2+ , Mg2+ , Mn2+ , fromspecies GLY onfrom the hydrazine-covered AuNP surfaces. The surfaces. other ionsThe of Fe CN GLY on the hydrazine-covered AuNP other ions of Fe3+, Fe2+, Hg2+, 2+ 2+ 3+ 2+ 2+ 2+ 2+ + + + Ni , Zn , Cr , Co , Cd , Pb , Ca , NH , Na , and K at high concentrations of 50 µM not 4 2+, NH4+, Na+, and K+ at high concentrations Mg2+, Mn2+, Ni2+, Zn2+, Cr3+, Co2+, Cd2+, Pb2+, Ca ofdid 50 µM produce spectral detection based on the CN band determination of did not such produce suchchanges. spectralThe changes. Thelimit detection limit based onfor thetheCN band for the 2+ ion could be estimated to be as low as 500 nM in distilled water and 1 µM in river water, the Cu determination of the Cu2+ ion could be estimated to be as low as 500 nM in distilled water and 1 µM respectively. We attempted to apply our method estimate detection in cancer cells in river water, respectively. We attempted totoapply ourintracellular method to ion estimate intracellular ion for more practical detection in cancerpurposes. cells for more practical purposes.

Cu(II); Raman Raman spectroscopy; C≡N stretching stretching mode; mode; gold gold nanoparticles; nanoparticles; glycine; glycine; Keywords: Cu(II); Keywords: spectroscopy; C≡N intracellular imaging imaging intracellular

1. Introduction 1. Introduction Nanomaterial-based tools have been successfully implemented to monitor heavy metal ions [1]. Nanomaterial-based tools have been successfully implemented to monitor heavy metal ions [1]. By using nanomaterials, surface-enhanced Raman scattering (SERS) has provided one of the most By using nanomaterials, surface-enhanced Raman scattering (SERS) has provided one of the most sensitive methods on noble metal surfaces [2,3]. Selection of suitable reducing reagents is significant sensitive methods on noble metal surfaces [2,3]. Selection of suitable reducing reagents is significant in the preparation of well-dispersed particles to study nanometric objects [4]. SERS has opened up in the preparation of well-dispersed particles to study nanometric objects [4]. SERS has opened up novel opportunities in analytics to provide a vibrational fingerprint with high spectral resolution [5]. novel opportunities in analytics to provide a vibrational fingerprint with high spectral resolution [5]. Nanoparticles have been used to detect a trace amount of analytes [6,7]. Various techniques have Nanoparticles have been used to detect a trace amount of analytes [6,7]. Various techniques have currently been developed for the imaging of the metal ions [8]. Inductively coupled plasma mass currently been developed for the imaging of the metal ions [8]. Inductively coupled plasma mass spectrometry is practically a powerful method that has been developed for metal ion detection [9]. spectrometry is practically a powerful method that has been developed for metal ion detection [9]. However, it remains challenging to develop a novel analytical method for cheaper and easier metal However, it remains challenging to develop a novel analytical method for cheaper and easier metal ion monitoring. ion monitoring. At elevated concentrations, Cu2+ ions are toxic in fundamental biological processes and can cause At elevated concentrations, Cu2+ ions are toxic in fundamental biological processes and can damage to organs such as the liver and kidneys [10]. This potential hazard makes it indispensable to cause damage to organs such as the liver and kidneys [10]. This potential hazard makes it monitor the amount of copper in environmental water. Toward alleviating this issue, the selective indispensable to monitor the amount of copper in environmental water. Toward alleviating this detection of Cu2+ ions in aqueous solutions with high sensitivity is of great importance. To ensure safety, issue, the selective detection of Cu2+ ions in aqueous solutions with high sensitivity is of great a detection limit as low as 20 µM (1.3 ppm) is recommended by the Environmental Protection Agency importance. To ensure safety, a detection limit as low as 20 µM (1.3 ppm) is recommended by the Environmental Protection Agency (EPA) [11]. Recently, ZnO@ZnS core-shell nanoparticles [12] and Sensors 2016, 2016, 16, 16, 1785; 1785; doi:10.3390/s16111785 Sensors doi:10.3390/s16111785

www.mdpi.com/journal/sensors www.mdpi.com/journal/sensors

Sensors 2016, 16, 1785

2 of 11

Sensors 2016, 16, 1785

2 of 11

(EPA) [11]. Recently, ZnO@ZnS core-shell nanoparticles [12] and fluorescent gold nanoclusters [13] 2+ ions. Functionalized gold have been introduced to achieve andintroduced selective detection of Cu fluorescent gold nanoclusters [13]efficient have been to achieve efficient and selective detection 2+ . SERS nanoparticles (AuNPs) [14] were used for the detection metal ions including of Cu2+ ions. Functionalized gold also nanoparticles (AuNPs) [14]ofwere also used for the Cu detection of 2+ 2+ and 2+ 2+ has also been used to discriminate Cu ions in aqueous solutions with an interference of Hg metal ions including Cu . SERS has also been used to discriminate Cu ions in aqueous solutions 2+ Co with [15,16]. an interference of Hg2+ and Co2+ [15,16]. Glycine amino acids, was known to adsorb on metal surfaces [17,18] via Glycine(GLY), (GLY),one oneofofthe thesimplest simplest amino acids, was known to adsorb on metal surfaces [17,18] 2+ ion2+as a dimeric form [19]. the carboxylate or the amino groups. GLY was predicted to bind the Cu via the carboxylate or the amino groups. GLY was predicted to bind the Cu ion as a dimeric form Using an infrared spectroscopy tool, GLY to dissociate to produce the cyanide group [19]. Using an infrared spectroscopy tool,was GLYreported was reported to dissociate to produce the cyanide under pH conditions, when awhen positive voltage was applied to a gold surface [20]. group alkaline under alkaline pH conditions, a positive voltage was applied to a electrode gold electrode surface −1 have been −1 Cyanide bands at ca. 2100 cm utilized as a marker band because of their discriminating [20]. Cyanide bands at ca. 2100 cm have been utilized as a marker band because of their position in the vibrational window without interference of the multiple overlapping bands discriminating position inspectral the vibrational spectral the window without interference of multiple −1 [21,22]. In view of such −1 between 1000 and 1700 cm electrochemical reactions [20], GLY may provide overlapping bands between 1000 and 1700 cm [21,22]. In view of such electrochemical reactions a[20], unique of the cyanide on of metal surfaces.group on metal surfaces. GLYmarker may provide a uniquegroup marker the cyanide 2+ In In this this study, study,we wereport reportaanew newCu Cu2+ ion detection method by means of SERS of GLY on AuNPs. 2+ UV-Vis absorption and a colorimetric UV-Vis absorption and a colorimetricmethod methodare arealso alsointroduced introducedtotocheck checkthe theCu Cu2+-induced -induced surface surface change. change. The variation of surface surface charge charge is is characterized characterized by means means of of zeta zeta potential potential measurements. measurements. The the increase increase of of the the Cu Cu2+2+ion ionininthe theGLY GLYcomplex. complex.This This The cyanide cyanide band band became prominent upon the is 2+ 2+ is the first study based on the selective detection of the Cu ion using the dissociative adsorption the first study based on the selective detection of the Cu ion using the dissociative adsorption of 2+ of GLY AuNPs that produces cyanide peaks whose intensities may correlated with Cuion GLY onon AuNPs that produces cyanide peaks whose intensities may be be correlated with thethe Cu2+ ion concentration. Moreover, micromolar sensitivity be achieved without the interferences concentration. Moreover, micromolar sensitivity couldcould be achieved without the interferences of 15 2+ 2+ 2+ 2+ of 15 ionic species including Hg Co andobserved Co observed in the previous These ionic species including Hg and in the previous RamanRaman studiesstudies [15,16].[15,16]. These results results be useful the development a novel detection schemetotoimprove improve the the efficacy of may bemay useful in thein development of aofnovel detection scheme of conventional methods. This study aims to develop a simple and inexpensive discrimination method conventional methods. This study aims to develop a simple and inexpensive discrimination method 2+ for for Cu Cu2+ ions by applying AuNP-based sensors.

Figure 1. Schematic diagram of detection of Cu2+ ions due to the redox properties to lead the 2+ ions due to the redox properties to lead the Figure 1. Schematic diagram detection of Cu dissociation of GLY into the CN of group on AuNP surfaces in the presence of hydrazine. The GLY-Cu2+ dissociation of GLY into the CN group on AuNP surfaces in thepositive presence of properties hydrazine. 2+ complex based on reference [19]. The increase of [Cu ] results in more redox of The the 2+ complex based on reference [19]. The increase of [Cu2+] results in more positive redox 2+ GLY-Cu GLY-Cu complex and stronger CN intensities. properties of the GLY-Cu2+ complex and stronger CN intensities.

Figure 1 shows our experimental scheme to detect Cu2+ ions with the complex of GLY in the Figure 1 shows our experimental scheme to detect Cu2+ ions with the complex of GLY in the presence of the hydrazine buffer, and subsequent adsorption on citrate-coated AuNPs. A dissociative presence of the hydrazine buffer, and subsequent adsorption on citrate-coated AuNPs. A reaction of GLY was found at higher concentrations of copper ions. By assembling and annealing dissociative reaction of GLY was found at higher concentrations of copper ions. By assembling and the Cu2+ and GLY2+ first in the hydrazine buffer, the complex can be ready to adsorb on AuNPs. annealing the Cu and GLY first in the hydrazine buffer, the complex can be ready to adsorb on We found that the Cu2+ ion made the surface charge of AuNPs more positive. On positively charged AuNPs. We found that the Cu2+ ion made the surface charge of AuNPs more positive. On positively Au surfaces, GLY was dissociated to produce the cyanide ions, which were adsorbed on AuNPs. These charged Au surfaces, GLY was dissociated to produce the cyanide ions, which were adsorbed on AuNPs. These reactions occurred under alkaline hydrazine buffer conditions at pH ~9.2. As control experiments, we tested the different amino acid such as alanine instead of GLY and performed independent trials using hydrazine, GLY, and NaOH with the same concentration and the pH value

Sensors 2016, 16, 1785

3 of 11

reactions occurred under alkaline hydrazine buffer conditions at pH ~9.2. As control experiments, we tested the different amino acid such as alanine instead of GLY and performed independent trials using hydrazine, GLY, and NaOH with the same concentration and the pH value of ~9.2 without the commercial hydrazine buffer containing 0.6 M GLY, 0.5 M hydrazinium sulphate and 0.0375% chloroform. The CN peaks could not be observed without either GLY or the commercial hydrazine buffer. We found that the CN peaks decreased significantly after 99.999% nitrogen-gas bubbling for 10 min, although not shown here. Despite the possible existence of the oxygen species, the ambient air may affect the redox reactions to produce the CN peaks under our experiment conditions. 2. Experimental Section 2.1. Materials GLY buffer solution (0.6 M, pH 9.2, Catalogue #G5418) containing 0.5 M NH2 NH2 and 0.0375% chloroform and the other ionic (mostly nitrate) substances of Cu(NO3 )2 , Fe(NO3 )3 , Fe(C2 O4 ), Hg(NO3 )2 , Mg(NO3 )2 , Mn(NO3 )2 , Ni(NO3 )2 , Zn(NO3 )2 , Cr(NO3 )3 , Co(NO3 )2 , Cd(NO3 )2 , Pb(NO3 )2 , Ca(NO3 )2 , NH4 NO3 , NaCl, and KNO3 , as the forms of either the hydrates or the nonhydrates, were purchased from Sigma Aldrich (St. Louis, MO, USA). GLY (99%) and hydrazine solution (35 wt % in H2 O) were also obtained from Sigma Aldrich. Human cervical carcinoma HeLa (ATCC® CCL-2) cells were grown on Dulbecco Modified Eagle Medium in a steri-cycle CO2 incubator (Thermo Fisher Scientific) in 5% CO2 /95% humidified air at 37 ◦ C. Cell culture media were supplemented with 1% penicillin–streptomycin antibiotics (Gibco)/0.2 ppm plasmocin and 10% fetal bovine serum (FBS). 2.2. Instrumentations We performed dark-field microscopy (DFM) [23] in cancer cells to check the possibility to apply our method under physiological conditions. We used the zeta potentials, transmission electron microscopy (TEM), energy dispersive X-ray analysis (EDAX), UV-Vis absorption spectroscopy, and SERS to characterize the AuNP suspensions. UV-Vis absorption spectra of the AuNP suspension were used to observe spectral changes with a Mecasys 3220 PC spectrophotometer. Surface charges were measured by zeta potential kits (Otsuka ELZ-2). High-resolution (HR)-TEM (JEOL JEM-3100) was used to observe the morphology of the AuNP Aggregates. EDAX spectrometer measurements were performed to find the Cu species assembled on Ni-coated grids by GLY-Cu2+ on metal nanoparticles using a Tecnai TF30 ST field emission TEM (300 kV). X-ray photoelectron spectroscopy (XPS) was conducted using a Sigma Probe Instrument Thermo VG spectrometer with an X-ray source of monochromic Al X-ray sources (Al Kα line: 1486.6 eV). The resolution of the full width at half-maximum is 0.76 eV in the Au 3d5/2 peaks. 2.3. Preparation of AuNPs and GLY-Cu2+ Complex We found that the 40 nm AuNPs (Catalogue # EMGC40/4) from BBI solutions (Cardiff, UK) would show the strongest CN species under our experimental conditions. For the UV-Vis experiment, in the first step, GLY (0.6 M, 10 µL), Cu2+ (1 mM, 50 µL), and ultrapure water (800 µL) were put into a 1.5 mL Eppendorf tube, stirred and stabilized over 60 min at room temperature. In the second step, 140 µL of AuNPs was added to the GLY-Cu2+ complex. Excessive amounts of GLY and hydrazine in the sample solutions were not found to disturb the presented ion detection scheme. For the zeta potential experiment, 10 µL Cu2+ (0.1 µM) and 50 µL GLY (0.6 M) were mixed and stirred for over 60 min at room temperature. Subsequently, 140 µL of the AuNP solution was added to the sample mixture and the surface charges of AuNPs appeared to shift toward the positive value after the addition of 10 nM of the Cu2+ ion by measuring zeta potentials. 1.0 mL of the resulting mixture solution was placed in a 1.5 mL centrifugal tube for the XPS experiment. The AuNPs-GLY-Cu2+ sample was collected by 4 ◦ C centrifugation at 10,000 rpm for 10 min. After the supernatant parts were removed up to a residual volume of 20 µL, the 20 µL

Sensors 2016, 16, 1785

4 of 11

sample of AuNPs-GLY-Cu2+ was dropped on a 5 mm × 5 mm cover glass purchased from Deckgläser (Sondheim/Rhön, Germany). The sample was then dried at 75 ◦ C overnight. For the SERS experiment, 10 µL of Cu2+ (1 mM, in either distilled water or river water) and 50 µL of GLY (0.6 M, GLY buffer solution, pH 9.2) were mixed and stirred at room temperature for over 1 h. Subsequently, 140 µL of AuNPs was added to the mixture to examine the adsorption of GLY-Cu2+ on AuNPs by recording the SERS spectra of the samples. The pH value was measured to be 8.9 in all the SERS measurements. Since we had used the GLY buffer solution with the fixed values of initial GLY 0.6 M and 0.5 M hydrazine at pH 9.2, it was difficult to control the relative concentrations of GLY and hydrazine to maintain the identical pH values and reaction conditions. GLY exhibited an increasing Raman intensity correlated to [Cu2+ ] in water (50 µM) on AuNPs. Response behaviors were examined for GLY in the potential interference of various metal ions on AuNPs. The Raman spectral features of GLY-Mn+ on AuNPs were unchanged in the presence of Fe3+ , Fe2+ , Hg2+ , Mg2+ , Mn2+ , Ni2+ , Zn2+ , Cr3+ , Co2+ , Cd2+ , Pb2+ , Ca2+ , NH4 + , Na+ , and K+ (all at concentrations of 50 µM). Of all the tested metal ions, only Cu2+ ions increased the Raman intensity in the presence of GLY on AuNPs. Inductively coupled plasma optical emissions spectrometry (ICP-OES) was used to estimate the atomic compositions of the prepared samples with an ICAP-7400 analyzer from Thermo Scientific (Waltham, MA, USA) except the Hg measurements by a CETAC M-7500 mercury analyzer from Parma Company (Omaha, NE, USA). Triply-distilled water was used to test the ionic detection. In order to check the applicability of our analytical methodology, the river water samples were obtained from Han river (Seoul, Korea), respectively. Raman spectra for ionic detection in environmental samples were obtained as in the method of our recent report [24–26]. 3. Results and Discussion 3.1. Physical Characterization To understand the surface reactions, AuNPs were characterized by HR-TEM and XPS. The sizes of the AuNPs are ~40 nm, as shown in the TEM image in Figure 2a. We could also find a trace amount of the Cu species on AuNPs by EDAX. The relative abundance of the adsorbed Cu species on metal NPs could be found to be as low as 0.2%–0.6%. As shown in Figure 2, XPS data [27] also confirmed that the percentage of Cu is as low as 0.2%. Because of a small amount of the Cu species below 1%, we could not map the Cu distribution on a single nanoparticle. By adding GLY, the surface charge became slightly more negative. Zeta potential measurements indicated that the surface charges appeared to become positive and almost neutral to destabilize the colloidal conditions and induce aggregation in the presence of Cu2+ ions, as is consistent with the UV-Vis data. This result suggests that the high concentration of may oxidize the GLY complex on the metal nanoparticle surfaces. On the basis of the surface characterization on AuNPs, we attempted to monitor Cu2+ ions by Raman spectroscopy. Figure 2c shows the color change of AuNPs upon the increase of the Cu2+ ion concentrations. As the concentration of Cu2+ ions increased in the complexes, the UV-Vis absorption band at 600–900 nm became stronger, indicating the aggregation of AuNPs. It has to be mentioned that AuNPs intrinsically have negative charges. As shown in Figure 2c, the SPR bands of AuNPs are newly presented instead of those of AgNPs. The decrease in the SPR bands at 520 nm may be caused not by the destruction of the nanoparticles but by the increase in particle sizes owing to aggregation. Instead, we observed extensive shifts of the SPR band between 600 and 900 nm to support the evidence of interparticle aggregation. To address the aggregation issue, we newly performed the dynamic light scattering measurements as shown in Figure 2d. Since the static TEM measurements may not provide good information on the aggregation during the sample preparation process, as solution evaporation may induce particle aggregation owing to dryness on the sample grid, we performed light scattering measurements to

Sensors 2016, 16, 1785

5 of 11

estimate the hydrodynamic diameters of the aggregated AuNPs in aqueous solutions to better address the issue of aggregation in relation to SERS intensities. At [Cu2+ ] = 1 and 50 µM, the diameters became 240 and 640 nm, respectively, to suggest an extensive aggregation mediated by the Cu ion species, which should result in strong SERS effects. Sensors 2016, 16, 1785 5 of 11

(b) 800 600 400

(c)

- G NP LY s -C u( II) 1u M A uN -G P LY s -C u( II) 50 uM

A

uN

Ps

0

u -H N P yd s ra zi ne A uN -G P s ly ci ne bu ffe A u r

200

A

Relative Diameter (nm)

(a)

(d)

Figure 2. (a) (a)TEM TEMimage image and spectrum of GLY-Cu on AuNPs; (c) UV-Vis absorption 2+ on 2+ Figure 2. and (b)(b) XPSXPS spectrum of GLY-Cu AuNPs; (c) UV-Vis absorption spectra. spectra. The inset shows a photo of the AuNP solutions in the (1) absence and (2) presence of The inset shows a photo of the AuNP solutions in the (1) absence and (2) presence of hydrazine, 2+ with [Cu2+] = 1 and 50 µM in (4) and (5), 2+ 2+ hydrazine, (3) glycine buffer and glycine buffer-Cu (3) glycine buffer and glycine buffer-Cu with [Cu ] = 1 and 50 µM in (4) and (5), respectively; 2+ on AuNPs in the respectively; (d) Stick of relative hydrodynamic (d) Stick diagram of diagram relative hydrodynamic diameters diameters of GLY-Cuof2+GLY-Cu on AuNPs in the [Cu2+ ] 2+ [Cu ] concentration range of 50 1 and concentration range of 1 and µM.50 µM.

3.2. on AuNPs AuNPs 3.2. Raman Raman Spectra Spectra of of Hydrazine Hydrazine and and GLY GLY on Figure shows the the normal normal Raman Raman (NR) (NR) and and SERS SERS spectra spectra of of GLY GLY on on AuNPs. AuNPs. The Figure 33 shows The Raman Raman spectrum of GLY is consistent with that reported previously [28–30]. It was found that most spectrum of GLY is consistent with that reported previously [28–30]. It was found that most strong strong bands could be ascribed to the hydrazine peaks, as previously reported [31–33]. The SERS spectrum bands could be ascribed to the hydrazine peaks, as previously reported [31–33]. The SERS spectrum of of GLY appeared to relatively be relatively weak its amino group interacting Au surfaces. the GLY appeared to be weak withwith its amino group interacting with with Au surfaces. In theIn SERS SERS spectrum of hydrazine on AuNPs, we could not observe a strong hydrazone linkage-related spectrum of hydrazine on AuNPs, we could not observe a strong hydrazone linkage-related N-N −1 or N-H bending modes at 1174 and 1222 cm−1 [34]. N-N stretching 919 − 1cm stretching modemode at 919atcm or N-H bending modes at 1174 and 1222 cm− 1 [34]. Since hydrazine shows weak basic with a Since hydrazine shows weak basicproperties propertiesininaqueous aqueoussolution solutiontotoform formstable stablesalts salts with number of mineral acids such as hydrazine sulfate N 2H5·HSO4 and hydrazine hydrochloride a number of mineral acids such as hydrazine sulfate N2 H5 ·HSO4 and hydrazine hydrochloride N mayinteract interactwith with amino acid efficiently. As suggested SERS N22H H44·HCl, it may thethe GLYGLY amino acid efficiently. As suggested from thefrom SERSthe spectrum ·HCl, it spectrum of GLY in the presence of hydrazine as shown in Figure 3e, GLY can be decomposed to of GLY in the presence of hydrazine as shown in Figure 3e, GLY can be decomposed to yield a CN 2+. A magnified view of Figure 3f in the presence of Cu2+ yield a CN peak even in the absence of Cu peak even in the absence of Cu2+ . A magnified view of Figure 3f in the presence of Cu2+ also exhibited also exhibited peaks. the hydrazine peaks.experiment As a control experiment without hydrazine, we could not the hydrazine As a control without hydrazine, we could not observe any CN observe any CN peaks. This result indicates a role of hydrazine to bind the amino acid as a base peaks. This result indicates a role of hydrazine to bind the amino acid as a base after adsorption on the after onsurfaces. the metal nanoparticle The other amino acids methione, metaladsorption nanoparticle The other aminosurfaces. acids including methione, lysine,including and leucine would lysine, and leucine would not provide strong CN peaks as in the case of GLY. The simple structure of not provide strong CN peaks as in the case of GLY. The simple structure of GLY along with the strong GLY along with the strong interaction with the Cu(II) ion may facilitate its dissociation to yield CN interaction with the Cu(II) ion may facilitate its dissociation to yield CN peaks on metal surfaces, peaks onthe metal surfaces, whereas reaction appears to be unfavorable to the other acid. whereas reaction appears to be the unfavorable to the other amino acid. Considering theamino high initial Considering the high initial concentrations of both GLY (0.6 M) and hydrazine (0.5 M), we cannot exclude the interactions between hydrazine and Cu2+. As suggested from the SERS features of hydrazine in Figure 3f, it is likely that hydrazine adsorbed fairly strongly on Au to interact mainly with GLY. The increase of the Cu2+ ion was thought to lead a stabilization of the hydrazine-modified AuNPs.

Sensors 2016, 16, 1785

6 of 11

concentrations of both GLY (0.6 M) and hydrazine (0.5 M), we cannot exclude the interactions between hydrazine and Cu2+ . As suggested from the SERS features of hydrazine in Figure 3f, it is likely that hydrazine adsorbed fairly strongly on Au to interact mainly with GLY. The increase of the Cu2+ ion Sensors 2016, 16, 6 of 11 was thought to 1785 lead a stabilization of the hydrazine-modified AuNPs.

Figure (a)NR NR of of and Raman scattering (SERS) of hydrazine (wt. 35% H2O) Figure 3. 3.(a) and (b) (b)surface-enhanced surface-enhanced Raman scattering (SERS) of hydrazine (wt. in35% inon 2+] = 0; (c) NR of SERS GLYof onGLY AuNPs; (e) SERS GLY of (hydrazine buffer) with [Cuwith H2AuNPs; O) on AuNPs; (c)GLY; NR of(d) GLY; (d)ofSERS on AuNPs; (e)ofSERS GLY (hydrazine buffer) 2+ ] The ] = 50[Cu µM. a magnified view of theview weak (f)2+SERS (hydrazine buffer) with [Cu2+with [Cu ] = 0; of (f)Gly SERS of Gly (hydrazine buffer) = 50inset µM. shows The inset shows a magnified bands of hydrazine; SERS spectrum of spectrum 0.1 M KCNofon of vibrational the weak vibrational bands of (g) hydrazine; (g) SERS 0.1AuNPs. M KCN on AuNPs. −1 noteworthythat thatthe the CN CN band newly appeared in the spectra. In the 1 was It It isisnoteworthy band at at ~2108 ~2108cm cm−was newly appeared in SERS the SERS spectra. 2+ presence of 50 µM of Cu ions, the CN band became more intensified. The band position 2+ In the presence of 50 µM of Cu ions, the CN band became more intensified. The band position at −1 on Au Au surfaces, surfaces,considering consideringthat thatthis thisis is almost at ~2108 ~2108cm cm−1corresponds correspondstotothe theCN CN species species adsorbed adsorbed on almost −1 identical CN band 2115 when0.1 0.1MMKCN KCNis istreated treated AuNPs, demonstrated −1 when identical toto thethe CN band at at 2115 cmcm onon AuNPs, asas demonstrated inin Figure 3g. Figure 3g. Our result consistent with the stoichiometric reaction [20] equations that the CN group may Our result is is consistent with the stoichiometric reaction [20] equations that the CN group may formed the surfaces under alkaline conditions. Furthermore, the decompositionofof GLY the bebe formed onon the surfaces under alkaline conditions. Furthermore, the decomposition GLY at at the positive surface charge is in agreement with this report [20]. Referring from the previous reports positive surface charge is in agreement with this report [20]. Referring from the previous reports ofof the formation constants[35,36] [35,36] GLY-Cu and metal-CN, it was expected that CN would strongly the formation constants ofof GLY-Cu and metal-CN, it was expected that CN would strongly coordinate to the Cu ions. However, considering that only less than 1% of Cu(II) ions existonon coordinate to the Cu ions. However, considering that only less than 1% of Cu(II) ions exist nanoparticle surfaces in XPS measurements as suggested in Figure 2b, most CN bands may be nanoparticle surfaces in XPS measurements as suggested in Figure 2b, most CN bands may be ascribed to mode. the AuCN mode. toascribed the AuCN We obtained SERS spectra GLY alkaline AuNPs presence various ions—Fe 3+ ,3+, We obtained thethe SERS spectra of of GLY onon alkaline AuNPs in in thethe presence of of various ions—Fe +—at the same , Hg , Mg Mn Zn Cr Co Cd Pb2+2+,, Ca Ca2+2+,, NH NH+4+,, Na Na++,, and andKK 2+2+ 2+2+ 2+2+ 2+2+ 2+2+ 2+2+, ,Cr 3+3+, ,Co 2+2+,,Cd 2+2+,, Pb + —at FeFe , Hg , Mg , ,Mn , ,NiNi , ,Zn the same 4 2+ concentrations µM, comparison with shown Figure It was found that only 2+ asas 2+ 2+ concentrations of of 5050 µM, in in comparison with CuCu shown in in Figure 4a.4a. It was found that only CuCu ions strongly enhanced CN formation illustrated the stick diagram Figure reported ions strongly enhanced CN formation asas illustrated inin the stick diagram ofof Figure 4b.4b. It It is is reported that and form a bimetallic structure Presumably, due to similar structures that AuAu and CuCu cancan form a bimetallic structure [37]. [37]. Presumably, due to similar crystalcrystal structures and 2+ ion may easily adsorb on AuNP surfaces instead of the other ions. and electronic properties, the Cu 2+ electronic properties, the Cu ion may easily adsorb on AuNP surfaces instead of the other ions.

Sensors 2016, 16, 1785 Sensors 2016, 16, 1785 Sensors 2016, 16, 1785 Cu(II) Fe(III) Cu(II) Fe(III) Fe(II) Hg(II) Fe(II) Hg(II) Mg(II) Mn(II) Mg(II) Mn(II) Ni(II) Zn(II) Ni(II) Zn(II) Cr(III) Co(II) Cr(III) Co(II) Cd(II) Pb(II) Cd(II) Pb(II) Ca(II) NH4(I) Ca(II) NH4(I) Na(I) K(I) Na(I) K(I) AuNPs-GLY AuNPs-GLY

1000 2000 -1 1000 2000 Wavenumber (cm -1) Wavenumber (cm ) (a) (a)

3000 3000

Cu(II) Cu(II) Fe(III) Fe(III) Fe(II) Fe(II) Hg(II) Hg(II) Mg(II) Mg(II) Mn(II) Mn(II) Ni(II) Ni(II) Zn(II) Zn(II) Cr(III) Cr(III) Co(II) Co(II) Cd(II) Cd(II) Pb(II) Pb(II) Ca(II) Ca(II) NH4(I) NH4(I) Na(I) Na(I) K(I) K(I)

RamanIntensity Intensity(a.u.) (a.u.) Raman

n+

AuNPs-Gly-M n+50uM AuNPs-Gly-M 50uM

RelativeIntensity Intensity Relative

7 of 11 7 of 11 7 of 11

(b) (b)

Figure 4. (a) SERS spectra of GLY in the presence of 50 µM of various ions. The SERS spectrum of GLY Figure4.4. (a) (a) SERS spectra spectra of of GLY GLYin in thepresence presence of50 50 µMof of variousions. ions.The TheSERS SERSspectrum spectrumof of GLY Figure 2+ ionµM on AuNPs is SERS presented in the middle.the Only the Cuof couldvarious induce a strong CN peak as listed atGLY the 2+ ion could induce a strong CN peak as listed at the on AuNPs is presented in the middle. Only the Cu 2+ on AuNPs is presented in the middle. Only the Cu ion could induce a strong CN peak as listed at the top of the spectra; (b) Stick diagram of the relative intensities of the CN bands for the tested ions. Three topof ofthe thespectra; spectra;(b) (b)Stick Stickdiagram diagramof ofthe therelative relativeintensities intensitiesof ofthe theCN CNbands bandsfor forthe thetested testedions. ions.Three Three top independent measurements were performed to yield the error bars in order to check the independentmeasurements measurements were performed to the yield bars in order to check the independent were performed to yield errorthe barserror in order to check the reproducibility. reproducibility. reproducibility.

RamanIntensity Intensity(a.u.) (a.u.) Raman

2+ on 3.3. CN CN Formation Formation Reactions Reactions of of GLY-Cu GLY-Cu2+ on AuNPs AuNPs and and Selective Selective Test Test 3.3. 3.3. CN Formation Reactions of GLY-Cu2+ on AuNPs and Selective Test 2+ concentration. It seems certain that Figure 55 shows shows SERS SERS spectra spectra of of GLY Figure GLY depending depending on on the the Cu Cu2+ concentration. It seems certain that Figure 5 shows SERS spectra of GLY depending on the Cu2+ concentration. It seems certain that − 1 −1 the band at ~2108 cm can be ascribed to the CN species, referring from the unique band position. position. the band at ~2108 cm −1can be ascribed to the CN species, referring from the unique band the band at ~2108 cm can be ascribed to the CN species, referring from the unique band position. 2+ 2+ concentrations was was magnified magnified between between 2000 2000 and The CN stretching stretching region region depending depending on on the the Cu Cu concentrations The CN−stretching depending on the Cu2+ concentrations was magnified between 2000 1 in Figureregion 2+ and 2200 cm−1 5a. It revealed an increase in the intensities depending on the Cu 2200 in Figure 5a. It revealed an increase in the band the Cu2+2+ion −1 2200 cm in Figure 5a. It revealed an increase in the band intensities depending on the Cu ion concentration. ItIt was was found found that that the the CN CN stretching stretching band band showed showed aa rather rather asymmetric asymmetric shape shape that concentration. concentration. It was found that the CN stretching band showed a rather asymmetric shape that − 1 could be decomposed into two bands not absolutely absolutely certain certain whether whether could bands at at ~2070 ~2070 and and ~2108 ~2108 cm cm−1. .ItItisis not could be decomposed into two bands at ~2070 and ~2108 cm−1. It is not absolutely certain whether the weaker weaker band band may may be ascribed ascribed to the CN mode on Cu and Au surfaces, depending depending on their close the weaker band may be ascribed to the CN mode on Cu and Au surfaces, depending on their close vibrational band bandpositions positions[38]. [38].Considering Considering that only KCN exhibited an asymmetric shape, vibrational that only KCN alsoalso exhibited an asymmetric shape, the vibrational band positions [38]. Considering that only KCN also exhibited an asymmetric shape, the the weaker shoulder peaks suggest a different binding Au surfaces. the previous weaker shoulder peaks may may suggest a different binding modemode on Auon surfaces. In the In previous study weaker shoulder peaks may suggest a different binding mode on Au surfaces. In the previous study studythe [17], the peak positions of band CN band appeared to blueshift theapplied appliedpotential potentialbecome become more more [17], peak positions of CN appeared to blueshift as as the [17], the peak positions of CN band appeared to blueshift as the applied potential become more − 1 − 1 −1 −1 The CN CN peak peakbecame becameblueshifted blueshiftedfrom from~2108 ~2108cm cm to to ~2115 as concentrations the concentrations of positive. The ~2115 cmcm , as ,the of the positive. The CN peak became blueshifted from ~2108 cm−1 to ~2115 cm−1, as the concentrations of the 2+ ion the2+2+ Cu increased, presumably to more cationic environments. Cu ion increased, presumably due due to more cationic environments. Cu ion increased, presumably due to more cationic environments.

AuNPs-GLY-Cu(II) AuNPs-GLY-Cu(II)

1000 2000 1000 2000-1 Wavenumber (cm -1) Wavenumber (cm )

(a) (a)

[Cu(II)]/uM [Cu(II)]/uM 50 50 25 25 10 7 10 57 35 13 1 0.5 0.5 0.1 0.1

3000 3000

(b) (b)

2+] concentration. Figure 5. (a) Concentration-dependent Concentration-dependent SERS spectra of GLY depending on on the [Cu [Cu2+ 2+ Figure ] ]concentration. Figure5.5.(a) (a) Concentration-dependent SERS SERS spectra spectra of of GLY GLYdepending depending onthe the [Cu concentration. Baseline corrections were made to compare the spectral bands; (b) Intensity plot of the CN intensities intensities Baseline Baseline corrections corrections were were made made to tocompare comparethe thespectral spectralbands; bands;(b) (b)Intensity Intensityplot plotof ofthe theCN CN intensities 2+] concentration of 0–50 µM. The inset shows fitting the linear region between 0 and versus the [Cu 2+ versus concentrationofof0–50 0–50µM. µM.The Theinset inset shows fitting linear region between 0 versus the the [Cu [Cu2+]] concentration shows fitting thethe linear region between 0 and 10 µM. and 10 µM. 10 µM.

Figure 5b shows an intensity plot of the CN stretching bands depending on the [Cu2+2+] Figure 5b shows an intensity plot of the CN stretching bands depending on the [Cu ] concentration range of 0–50 µM in river water. The detection limit on the basis of the CN band concentration range of 0–50 µM in river water. The detection limit on the basis of the CN band is as low as 2 µM of Cu2+2+ions. The proposed detection limit was estimated from the linear fitting is as low as 2 µM of Cu ions. The proposed detection limit was estimated from the linear fitting

Sensors 2016, 16, 1785

8 of 11

8 of2+ 11] an intensity plot of the CN stretching bands depending on the [Cu concentration range of 0–50 µM in river water. The detection limit on the basis of the CN band is as 2+ ions. change curve, the could begin to occur on the linear Y (CNcurve, low aswhere 2 µM of Cuspectral The proposed detection limitbased was estimated fromequation the linearoffitting 2+ concentrations). intensities) = B (intercept) + A (slope) × X (Cu where the spectral change could begin to occur based on the linear equation of Y (CN intensities) = B (intercept) + A (slope) × X (Cu2+ concentrations). 3.4. CN Bands in River Water Samples Contacting Highly Concentrated Na, Ca, Mg, and K Ions 3.4. CN Bands in River Water Samples Contacting Highly Concentrated Na, Ca, Mg, and K Ions As proof of application, the concentrations of Ca, K, Na, and Mg, were found to be ~31.04 ppm, ~6.37 As ppm, ~25.19 ppm, andthe ~5.77 ppm for theofriver sample [24–26]. As to demonstrated in proof of application, concentrations Ca, K,water Na, and Mg, were found be ~31.04 ppm, 2+ determination could be applied in real water samples Figure 6, our analytical methodology of Cu ~6.37 ppm, ~25.19 ppm, and ~5.77 ppm for the river water sample [24–26]. As demonstrated in Figure 6, only with higher limit of detection. It has to be mentioned that ~2 in µMreal is water still lower thatonly the EPA our analytical methodology of Cu2+ determination could be applied samples with recommendation. We could clearly observe that the~2marker band of that ~2108 cm−1 recommendation. at micromolar higher limit of detection. It has to be mentioned µM is still lower the EPA 2+ . 2+. The detection limit may be expected concentrations of Cu improve toward around 1 of µM We could clearly observe the marker band of ~2108 cm−1 attomicromolar concentrations Cuby optimizing thelimit experimental conditions, although therearound is a turning point to yieldthe a higher slope The detection may be expected to improve toward 1 µM by optimizing experimental above ~5 µM. conditions, although there is a turning point to yield a higher slope above ~5 µM. SensorsFigure 2016, 16,5b 1785shows

River water

2050 2100 2150 -1 Wavenumber (cm )

(a)

Relative Intensity

10 7 5 3 1

Raman Intensity (a.u.)

2000

2 R =0.94148

[Cu(II)] / uM

2200

2

4 6 [Cu(II)] / 

8

10

(b)

Figure between 2000 2000 and and 2200 2200 cm cm−−11; ;(b) Figure 6. 6. (a) (a)A Amagnified magnified view view of of the the CN CN stretching stretching region region between (b)Intensity Intensity 2+ ] concentration range of 1–10 µM for river water. plot fitting the linear region on the [Cu 2+ plot fitting the linear region on the [Cu ] concentration range of 1–10 µM for river water. Three Three independent independent measurements measurements were were performed performed to to yield yield the the error error bars. bars.

3.5. Cu2+2+Ion Detection in Cancer Cells 3.5. Cu Ion Detection in Cancer Cells 2+ ion in cellular media as depicted We attempted to detect the intracellular concentrations of Cu2+ We attempted to detect the intracellular concentrations of Cu ion in cellular media as depicted 2+ ion concentration ranges in Figure 7. We could observe the increase of the CN peaks in the Cu2+ in Figure 7. We could observe the increase of the CN peaks in the Cu ion concentration ranges between 5 and 20 µM, whereas we could not find any CN peaks for the control tests at 20 µM of K+ between 5 and 20 µM, whereas we could not find any CN peaks for the control tests at 20 µM of K+ ion. ion. Since the free Cu2+ ion inside cells may be much lower than that of the total copper ions, it is Since the free Cu2+ ion inside cells may be much lower than that of the total copper ions, it is admitted admitted that our methods may be limited in the case of the usages for the intracellular cellular2+free that our methods may be limited in the case of the usages for the intracellular cellular free Cu ion Cu2+ ion measurements under our experimental conditions. measurements under our experimental conditions. Our current Raman spectroscopy-based ion detection approach may provide several Our current Raman spectroscopy-based ion detection approach may provide several advantages advantages including better spectral resolution and multiplexing, which can be applied to the areas including better spectral resolution and multiplexing, which can be applied to the areas of cancer of cancer cell research, pharmaceutical identification, and characterization of dynamic interactions cell research, pharmaceutical identification, and characterization of dynamic interactions between between nanoparticles and cells to overcome the limitation of the colorimetric method as depicted nanoparticles and cells to overcome the limitation of the colorimetric method as depicted in Figure 2c. in Figure 2c. Considering the unique spectral position of the CN band, the potential of our detection method is Considering the unique spectral position of the CN band, the potential of our detection thought to be promising in comparison with the conventional SERS method, which may suffer from method is thought to be promising in comparison with the conventional SERS method, which may the spectral overlap in the congested wavenumber region. We shall keep improving the sensitivity suffer from the spectral overlap in the congested wavenumber region. We shall keep improving the by introducing novel nanostructures and optimizing the experimental parameters. Our method may sensitivity by introducing novel nanostructures and optimizing the experimental parameters. Our provide a unique tool for detecting Cu2+ ions in an aqueous solution by means of Raman spectroscopy method may provide a unique tool for detecting Cu2+ ions in an aqueous solution by means of in a selective and sensitive manner. Raman spectroscopy in a selective and sensitive manner.

Sensors 2016, 16, 1785 Sensors 2016, 16, 1785

9 of 11 9 of 11

(b) Raman Intensity (a.u.)

(a) AuNPs-GLY-Mn+

(c) Cu(II) - 20 uM Cu(II) - 10 uM Cu(II) - 5 uM K(I) - 20 uM Control

1000 2000 -1 Wavenumber (cm )

3000

(d) 2+ (20 µM) (a); AuNPs-GLY-K+ + Figure 7. DFM DFM cell cell images images of of HeLa HeLa cells cells treated treated with with AuNPs-GLY-Cu AuNPs-GLY-Cu2+ Figure 7. (20 µM) (a); AuNPs-GLY-K 2+ (20 µM) (b); and AuNPs-GLY (c) for 24 h and subsequently the SERS spectra were observed. Only (20 µM) (b); and AuNPs-GLY (c) for 24 h and subsequently the SERS spectra were observed. Only Cu 2+ treated cells achieved the CN peak (d). The red arrows indicated the positions where the SERS Cu treated cells achieved the CN peak (d). The red arrows indicated the positions where the SERS spectra spectra were taken. were taken.

4. 4. Conclusions Conclusions 2+ ions by referring to the CN band increase by means We report aa new new detection detection method method for for Cu Cu2+ We report ions by referring to the CN band increase by means of SERS of the dissociation from GLY on hydrazine-adsorbed absorption spectra of SERS of the dissociation from GLY on hydrazine-adsorbed AuNPs. AuNPs. UV-Vis UV-Vis absorption spectra are are 2+ also to check check the the [Cu [Cu2+]-induced also introduced introduced to ]-induced surface surface change change to to indicate indicate the the Au Au particle particle aggregation aggregation to to result in the the color color change. change. The variation is characterized by means of of zeta zeta potential potential result in The surface surface charge charge variation is characterized by means −1 became prominent upon the increase of Cu2+ ions. measurements. band at at ~2108 ~2108 cm cm− 1 became prominent upon the increase of Cu2+ ions. measurements. The The cyanide cyanide band 2+ ion concentration. The other 15 ions of The cyanide peak intensities may be correlated with the Cu Cu2+ The cyanide peak intensities may be correlated with the ion concentration. The other 15 ions of 3+ 2+, Hg2+, Mg2+, Mn2+, Ni2+, Zn2+, Cr3+, Co2+, Cd2+, Pb2+, Ca2+, NH4+, Na+, and K+ have not produced Fe Fe3+, ,Fe Fe2+ , Hg2+ , Mg2+ , Mn2+ , Ni2+ , Zn2+ , Cr3+ , Co2+ , Cd2+ , Pb2+ , Ca2+ , NH4 + , Na+ , and K+ have not such spectral changes aschanges Cu2+. The limit based on thebased CN band is estimated to estimated be as low as produced such spectral asdetection Cu2+ . The detection limit on the CN band is to 2+ 2+ ions under 2beµM of Cu ions in river water. Our method was found to correlate with the Cu 2+ as low as 2 µM of Cu ions in river water. Our method was found to correlate with the Cu2+ ions intracellular conditions in cancer cells. cells. under intracellular conditions in cancer

Acknowledgments: This Basic Research Research Laboratories Laboratories (BRL) NRF grant grant Acknowledgments: This work work was was supported supported by by Basic (BRL) through through an an NRF funded by the MSIP (No. 2015056354). Author N.H.L. conceptualized the study. C.S. provided the insight the ionic detection Author Contributions: Contributions:S.J. S.J.and and N.H.L. conceptualized the study. C.S. provided the of insight of the ionic using novel materials. N.H.L. designed and conducted the experiments. N.H.L.N.H.L. and S.J. analyzed the data. detection using novel materials. N.H.L. designed and conducted the experiments. and S.J. analyzed the All authors read and reviewed the manuscript. data. All authors read and reviewed the manuscript. Conflicts of Interest: The authors declare no conflict of interest. Conflicts of Interest: The authors declare no conflict of interest.

References References 1. 1. 2. 2. 3.

March, G.; Nguyen, T.D.; Piro, B. Modified electrodes used for electrochemical detection of metal ions in March, G.; Nguyen, T.D.;Biosensors Piro, B. Modified electrodes used for[PubMed] electrochemical detection of metal ions in environmental analysis. 2015, 5, 241–275. [CrossRef] environmental analysis. Biosensors 2015, 5, 241–275. Tong, L.; Wei, H.; Zhang, S.; Xu, H. Recent advances in plasmonic sensors. Sensors 2014, 14, 7959–7973. Tong, L.; Wei, H.; Zhang, S.; Xu, H. Recent advances in plasmonic sensors. Sensors 2014, 14, 7959–7973. [CrossRef] [PubMed] Doria, G.; Conde, J.; Veigas, B.; Giestas, L.; Almeida, C.; Assunção, M.; Rosa, J.; Baptista, P.V. Noble metal nanoparticles for biosensing applications. Sensors 2012, 12, 1657–1687.

Sensors 2016, 16, 1785

3. 4.

5.

6.

7. 8.

9.

10. 11. 12.

13. 14.

15.

16.

17. 18. 19.

20.

21.

22.

10 of 11

Doria, G.; Conde, J.; Veigas, B.; Giestas, L.; Almeida, C.; Assunção, M.; Rosa, J.; Baptista, P.V. Noble metal nanoparticles for biosensing applications. Sensors 2012, 12, 1657–1687. [CrossRef] [PubMed] Khandelwal, P.; Singh, D.K.; Sadhu, S.; Poddar, P. Study of the nucleation and growth of antibiotic labeled Au NPs and blue luminescent Au8 quantum clusters for Hg(2+) ion sensing, cellular imaging and antibacterial applications. Nanoscale 2015, 7, 19985–20002. [CrossRef] [PubMed] Kassu, A.; Farley, C., III; Sharma, A.; Kim, W.; Guo, J. Effect of pore size and film thickness on gold-coated nanoporous anodic aluminum oxide substrates for surface-enhanced Raman scattering sensor. Sensors 2015, 15, 29924–29937. [CrossRef] [PubMed] Yaman, Y.T.; Abaci, S. Sensitive adsorptive voltammetric method for determination of bisphenol A by gold nanoparticle/polyvinylpyrrolidone-modified pencil graphite electrode. Sensors 2016, 16, 756. [CrossRef] [PubMed] Li, D.; Lv, P.; Zhu, J.; Lu, Y.; Chen, C.; Zhang, X.; Wei, Q. NiCu alloy nanoparticle-loaded carbon nanofibers for phenolic biosensor applications. Sensors 2015, 15, 29419–29433. [CrossRef] [PubMed] Braidy, N.; Poljak, A.; Marjo, C.; Rutlidge, H.; Rich, A.; Jayasena, T.; Inestrosa, N.C.; Sachdev, P. Metal and complementary molecular bioimaging in Alzheimer’s disease. Front. Aging Neurosci. 2014, 6, 138. [CrossRef] [PubMed] Majedi, S.M.; Kelly, B.C.; Lee, H.K. Toward a robust analytical method for separating trace levels of nano-materials in natural waters: Cloud point extraction of nano-copper(II) oxide. Environ. Sci. Pollut. Res. Int. 2014, 21, 11811–11822. [CrossRef] [PubMed] Brewer, G.J. Copper toxicity in the general population. Clin. Neurophysiol. 2010, 121, 459–460. [CrossRef] [PubMed] U.S. Environmental Protection Agency. Risk Assessment, Management and Communication of Drinking Water Contamination; U.S. EPA 625/4-89/024; Environmental Protection Agency: Washington, DC, USA, 1989. Sadollahkhani, A.; Hatamie, A.; Nur, O.; Willander, M.; Zargar, B.; Kazeminezhad, I. Colorimetric disposable paper coated with ZnO@ZnS core-shell nanoparticles for detection of copper ions in aqueous solutions. ACS Appl. Mater. Interfaces 2014, 6, 17694–17701. [CrossRef] [PubMed] Yuan, Z.; Cai, N.; Du, Y.; He, Y.; Yeung, E.S. Sensitive and selective detection of copper ions with highly stable polyethyleneimine-protected silver nanoclusters. Anal. Chem. 2014, 86, 419–426. [CrossRef] [PubMed] Weng, Z.; Wang, H.; Vongsvivut, J.; Li, R.; Glushenkov, A.M.; He, J.; Chen, Y.; Barrow, C.J.; Yang, W. Self-assembly of core-satellite gold nanoparticles for colorimetric detection of copper ions. Anal. Chim. Acta 2013, 803, 128–134. [CrossRef] [PubMed] Tsoutsi, D.; Guerrini, L.; Hermida-Ramon, J.-M.; Giannini, V.; Liz-Marzán, L.M.; Wei, A.; Alvarez-Puebla, R.A. Simultaneous SERS detection of copper and cobalt at ultratrace levels. Nanoscale 2013, 5, 5841–5846. [CrossRef] [PubMed] Li, F.; Wang, J.; Lai, Y.; Wu, C.; Sun, S.; He, Y.; Ma, H. Ultrasensitive and selective detection of copper (II) and mercury (II) ions by dye-coded silver nanoparticle-based SERS probes. Biosens. Bioelectron. 2013, 39, 82–87. [CrossRef] [PubMed] Ernst, K.H.; Christmann, K. The interaction of glycine with a platinum (111) surface. Surf. Sci. 1989, 224, 277–310. [CrossRef] Han, J.W.; James, J.N.; Sholl, D.S. Chemical speciation of adsorbed glycine on metal surfaces. J. Chem. Phys. 2011, 135, 034703. [CrossRef] [PubMed] Colaneri, M.J.; Vitali, J.; Peisach, J. Aspects of structure and bonding in copper-amino acid complexes revealed by single-crystal EPR/ENDOR spectroscopy and density functional calculations. J. Phys. Chem. A 2009, 113, 5700–5709. [CrossRef] [PubMed] Zhen, C.-H.; Sun, S.-G.; Fan, C.-J.; Chen, S.-P.; Mao, B.-W.; Fan, Y.-J. In situ FTIRS and EQCM studies of glycine adsorption and oxidation on Au(111) electrode in alkaline solutions. Electrochim. Acta 2004, 49, 1249–1255. [CrossRef] Hu, Q.; Tay, L.L.; Noestheden, M.; Pezacki, J.P. Mammalian cell surface imaging with nitrile-functionalized nanoprobes: Biophysical characterization of aggregation and polarization anisotropy in SERS imaging. J. Am. Chem. Soc. 2007, 129, 14–15. [CrossRef] [PubMed] Lin, L.; Tian, X.; Hong, S.; Dai, P.; You, Q.; Wang, R.; Feng, L.; Xie, C.; Tian, Z.Q.; Chen, X. A bioorthogonal Raman reporter strategy for SERS detection of glycans on live cells. Angew. Chem. Int. Ed. Engl. 2013, 52, 7266–7271. [CrossRef] [PubMed]

Sensors 2016, 16, 1785

23. 24. 25. 26. 27. 28. 29. 30.

31. 32. 33. 34. 35. 36. 37. 38.

11 of 11

Cong, V.T.; Ganbold, E.-O.; Saha, J.K.; Jang, J.; Min, J.; Choo, J.; Kim, S.; Song, N.W.; Son, S.J.; Lee, S.B.; et al. Gold nanoparticle silica nanopeapods. J. Am. Chem. Soc. 2014, 136, 3833–3841. [CrossRef] [PubMed] Ly, N.H.; Oh, C.H.; Joo, S.W. A submicromolar Cr(III) sensor with a complex of methionine using gold nanoparticles. Sens. Actuators B Chem. 2015, 219, 276–282. [CrossRef] Ly, N.H.; Joo, S.-W. Zn(II)-concentration dependent Raman spectra in the dithizonecomplex on gold nanoparticle surfaces in environmental water samples. Appl. Surf. Sci. 2015, 356, 1005–1011. [CrossRef] Ly, N.H.; Joo, S.-W. Silver nanoparticle-enhanced resonance Raman sensor of chromium(III) in seawater samples. Sensors 2015, 15, 10088–10099. [CrossRef] [PubMed] Moulder, J.F.; Stickle, W.F.; Sobol, P.E.; Bomben, K.D. Handbook of X-ray Photoelectron Spectroscopy; Physical Electronics Inc.: Chanhassen, MN, USA, 1995. Kumar, S.; Rai, A.K.; Singh, V.B.; Rai, S.B. Vibrational spectrum of glycine molecule. Spectrochim. Acta A 2005, 61, 2741–2746. [CrossRef] [PubMed] Shi, Y.; Wang, L. Collective vibrational spectra of α-and γ-glycine studied by terahertz and Raman spectroscopy. J. Phys. D Appl. Phys. 2005, 38, 3741–3745. [CrossRef] Podstawka, E.; Ozaki, Y.; Proniewicz, L.M. Part I: Surface-enhanced Raman spectroscopy investigation of amino acids and their homodipeptides adsorbed on colloidal silver. Appl. Spectrosc. 2004, 58, 570–580. [CrossRef] [PubMed] Durig, J.R.; Griffin, M.G.; Macnamee, R.W. Raman spectra of gases. XV: Hydrazine and hydrazine-d4 . J. Raman Spectrosc. 1975, 3, 133–141. [CrossRef] Braibanti, A.; Dallavalle, F.; Pellinghelli, M.A.; Leporati, E. The nitrogen-nitrogen stretching band in hydrazine derivatives and complexes. Inorg. Chem. 1968, 7, 1430–1433. [CrossRef] Sheets, R.W.; Blyhokler, G. Isocyanate formation from adsorbed carbon monoxide and ammonia or hydrazine on vanadium, iron, and nickel. J. Phys. Chem. 1975, 79, 1572–1573. [CrossRef] Ly, N.H.; Joo, S.-W. Surface anchoring mode-dependent hydrolysis reactions of hydrazone groups on gold examined by pH-dependent Raman spectroscopy. J. Raman Spectrosc. 2015, 46, 1082–1086. [CrossRef] Oraby, E.A.; Eksteen, J.J. The selective leaching of copper from a gold–copper concentrate in glycine solutions. Hydrometallurgy 2014, 150, 14–19. [CrossRef] Cho, K.; Jang, Y.S.; Gong, M.-S.; Kim, K.; Joo, S.-W. Determination of cyanide species in silver and gold plating solutions by Raman spectroscopy. Appl. Spectrosc. 2002, 56, 1147–1151. [CrossRef] Li, G.; Luo, Y. Preparation and characterization of dendrimer-templated Ag-Cu bimetallic nanoclusters. Inorg. Chem. 2008, 47, 360–364. [CrossRef] [PubMed] Bowmaker, G.A.; Kennedy, B.J.; Reid, J.C. Crystal structures of AuCN and AgCN and vibrational spectroscopic studies of AuCN, AgCN, and CuCN. Inorg. Chem. 1998, 37, 3968–3974. [CrossRef] [PubMed] © 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC-BY) license (http://creativecommons.org/licenses/by/4.0/).