Journal of Applied Electrochemistry https://doi.org/10.1007/s10800-017-1139-1

RESEARCH ARTICLE

Surface enhanced Raman spectroscopy measurement of surface pH at the electrode during Ni electrodeposition reaction Takayuki Homma1,2 · Masahiro Kunimoto2 · Moe Sasaki1 · Tomoya Hanai1 · Masahiro Yanagisawa2 Received: 11 October 2017 / Accepted: 7 December 2017 © Springer Science+Business Media B.V., part of Springer Nature 2017

Abstract In this work, we developed a precise approach to analyze local proton concentration at the solid/liquid interface of electrodes, i.e. “surface pH”, during electrochemical reactions. For this, surface enhanced Raman spectroscopy (SERS) was applied to analyze pH-dependent structural changes of the –COOH group of p-mercaptobenzoic acid (p-MBA) modified onto Au nanoparticles (NPs) on the substrate close to a working electrode. Measurements using this system identified deprotonation of –COOH of p-MBA. Since preliminary experiments and density functional theory calculations suggest that the pKa of p-MBA attached to Au NPs is close to that in bulk solution, the SERS results indicate pH increase due to proton consumption by the cathodic overpotential of the working electrode. As an example, we applied this system to surface pH monitoring in electrodeposition process of Ni in an acidic bath, which indicated the validity of our method for precise detection of pH changes at electrode interfaces in situ. Graphical Abstract Surface pH around the W. E. changes…

Structure of a ached molecules changes Working electrode

Time dependent monitoring of electrode surface pH

SERS acve area

Keywords  Surface pH · Electrodeposition · Surface enhanced Raman spectroscopy · Au nano particles

1 Introduction

* Takayuki Homma [email protected] 1



Department of Applied Chemistry, Waseda University, 3‑4‑1, Okubo, Shinjuku, Tokyo 169‑8555, Japan



Research Organization for Nano & Life Innovation, Waseda University, 513, Wasedatsurumaki, Shinjuku, Tokyo 162‑0041, Japan

2

Electrochemical reactions proceeding on electrodes are influenced by pH in the interfacial area, and in many cases products cannot be predicted from the pH of the bulk solution. Therefore, to understand the details behind electrochemical behaviors and to control reactions more precisely, the pH near the electrode surface, which is termed here “surface pH”, must be precisely analyzed. Since a pioneering work by Dahms and Croll [1], many researchers have proposed techniques to measure surface

13

Vol.:(0123456789)



pH and have attempted to apply those strategies to practical analyses. Microelectrodes [2–4] and micromesh structures [5, 6] are major examples of conventional surface pH measurement setups. Romankiw et al. proposed the micromesh method and combined it with a rotating disk electrode method [6] to apply it to detailed analyses [7, 8]. Methods using rotating ring disk electrodes [9] and semiconductors [10] are also available. Han summarized these methods, classifying them into direct methods, such as the micromesh method, and indirect methods, such as the rotating ring disk electrode method [11]. Even though these useful methods are applicable to many research fields, they involve complicated setups that provide different reaction environments to those of normal electrochemical equipment. Thus, the development of simpler measurement methods that are applicable to normal electrochemical systems is still needed. In this work, we propose a novel approach for detection of surface pH changes, which utilizes the surface enhanced Raman scattering (SERS) and does not significantly interfere electrochemical reaction. This paper describes the details of the measurement system and its application to the analysis of Ni electrodeposition in an acidic bath.

2 Experimental section Figure 1 shows a schematic of the surface pH measurement setup used in the present study. In this setup, SERS detection of the structural changes in pH probe molecules, which are adsorbed onto Au nanoparticles (NPs) attached to an Au substrate close to a working electrode during Ni electrodeposition, detect surface pH changes around the electrode interface. Hence, pH changes across the pKa of the probe molecules can be measured. Here, the part of the Au substrate surface bearing the pH-probe-modified Au NPs is termed the “SERS active area” (Fig. 1). Preparation of the SERS active area is one of the key processes in this work. First, to introduce the SERS effect to the area, Au NPs were prepared by electrochemical deposition. For this, a bath containing 1.5 g/L H ­ AuCl4 was applied. Au NPs were deposited on sputtered Au on Si (10 mm × 10 mm)

Fig. 1  Experimental setup for the measurement of surface pH on the electrode in Ni electrodeposition by the SERS active area

13

Journal of Applied Electrochemistry

with a Cr adhesion layer, at − 0.25 mA for 120 s. Fig. 2 shows a scanning electron microscopy (SEM) image of the Au-NP-deposited plate, showing that the particle size is ~ 50 nm. This particle size and density are considered to be suitable for enhancing the SERS effect [12–14]. Then, pH probe molecules were attached to the Au-NP-deposited plate. p-Mercaptobenzoic acid (p-MBA) and l-cysteine (lCys) were selected as possible probes because they have the following three structure characteristics: (1) An –SH moiety to adsorb strongly onto the Au NP surface, (2) a –COOH group to protonate/deprotonate in response to the surrounding pH environment, thus providing information on local pH changes, and (3) a very simple molecular structure that should interfere very little with the profile of the focused Raman spectrum. To prepare the pH-probe layer, the AuNP-deposited plate was immersed in 1.0 × 10−2 mol/L l-Cys or p-MBA solution at 10 °C for 24 h. After immersion, the plate was rinsed with pure water. The pH environment was detected by measuring the Raman peak intensity of –COOH and/or –COO− of the pH probe molecules on this plate, i.e., the SERS active area, located close to the working electrode (Fig. 1). Raman spectra were measured using 3D laser Raman microspectroscopy with a confocal optical system (Nanofinder 30, Tokyo Instruments, Inc.). The pinhole size of the confocal system was set to 50 µm. Raman scattering was excited by a 632.8 nm line from a He-Ne laser. The data acquisition time was 10 s. A ×50 magnification objective lens with NA = 0.5 (Olympus) was utilized [except for the preliminary experiment to evaluate the capability of the SERS active substrate area, in which a ×100 magnification objective lens with NA = 0.9 (Nikon) was used]. The chemical states of the pH probe molecules attached to the SERS active area were theoretically analyzed using density functional theory (DFT) calculation. The calculation procedures

Fig. 2  SEM image of the Au-NP-deposited substrate

Journal of Applied Electrochemistry

depended on the scenario, the details of which are described in each section.

3 Results and discussion Raman spectroscopy of the SERS active area in solutions with different pH values was used to identify peaks attributed to protonated and deprotonated p-MBA (Fig. 3), revealing peaks at 1075 and 1588 cm−1 for benzene ring vibration, 1178 cm−1 for CH in-plane bending, 1440 cm−1 for –COO− stretching, and 1700 cm−1 for –COOH stretching [15–18]. In these cases, the pH of the solutions was altered using H ­ 2SO4 and NaOH. Figure 3 clearly shows the pH dependence of the peak intensities for –COO− and –COOH, indicating that our SERS active area concept is applicable to analyzing surface pH. Conversely, the Raman spectrum for the SERS active area substrate prepared with l-Cys as the pH probe molecule is not as clear as the spectrum for the SERS active area prepared with p-MBA (Fig. 4). The spectrum shows peaks that could be attributed to –COO− and –COOH at ~ 1400 and 1700 cm−1, respectively, but they are not sufficiently clear for use in pH measurement. Furthermore, the spectrum shapes in the l-Cys case are not stable, with many other unidentified peaks being observed for each pH solution. This is considered to be due to the insufficient stability of the molecular layer structure of l-Cys on the rough NP surface. In the case of p-MBA, the benzene ring provides sufficient interaction between the p-MBA molecules, stabilizing the molecular layer structure sufficiently despite the high surface roughness, while l-Cys is not capable of such strong intermolecular interaction. From this view point, the benzene ring is one of the key structures in the pH probe molecule. Furthermore, Fig. 3 shows the strong Raman peak from this benzene ring, which can be used as an internal standard to

Fig. 4  pH dependence of the Raman spectrum of L-Cys-modified SERS active area

estimate the comparative strength of the –COO− and –COOH peaks. These peaks for p-MBA were also found theoretically by DFT calculation performed with Gaussian09 [19], in which the exchange–correlation functional was B3LYP [20–22], the basis set for H and C was 6-31G**, and the basis set for O and S was 6-31 + G* [23]. Some p-MBAs are thought to interact with each other on the NPs, which might slightly change the electronic structure of p-MBA itself. However, DFT calculations for dimerized p-MBA show that such interactions influence the peak position little. In the calculations for the p-MBA dimer, the M06L functional [24] was applied because the intermolecular interaction between two p-MBAs, including the dispersion force effect, needed to be expressed accurately. These data suggest Raman spectroscopy of the SERS active area is sufficiently capable of providing chemical information on the p-MBA. It has been reported that p-MBA is suitable for probing pH environments within cells by attaching it to Au NPs contained in nanopores [25]. SERS measurements in this work also identified peaks attributed to –COO− and –COOH vibrations of p-MBA on the NPs, the intensities of which changed in response to the surrounding pH, which further demonstrates the ability of p-MBA to act as a pH probe molecule. To investigate the difference between the pKa values for p-MBA on Au and that in bulk solution, their deprotonation energies were analyzed theoretically. Deprotonation energy (ΔE) is defined in the following equations;

{ ( )} ΔEon Au = E(p-MBA− on Au) + E H3 O+ { ( )} − E(p-MBA on Au) + E H2 O

(1)

{ ( )} { ( )} ΔEin solution = E(p-MBA− ) + E H3 O+ − E(p-MBA) + E H2 O Fig. 3  pH dependence of the Raman spectrum of the p-MBA-modified SERS active area

(2) where E represents the total electronic energy of each system and p-MBA− represents the deprotonated form of p-MBA

13



containing the –COO− group. Each energy value was calculated by DFT (B3LYP), as implemented in GAMESS [26], which was used to perform energy density analysis (EDA) [27] as follows: the basis sets were 6-31G** for H and C, 6-31 + G* for O and S, and Hay-Wadt VDZ with effective core potential for Au [28]. The polarized continuum model (PCM) [29] was used to model the solvation effects of water in p-MBA systems using a dielectric constant of 78.36 [–]. A ­Au20 pyramid-shape cluster was used as the Au metal surface, which was applicable for full geometrical optimization [30, 31]. Figure 5 shows a schematic of the deprotonation of p-MBA on Au with optimized structures for each system. The deprotonation energy of p-MBA on the Au surface is + 200.5 kJ/mol, whereas that of p-MBA in solution is + 207.7 kJ/mol, indicating that the Au surface slightly stabilizes the deprotonated species p-MBA−. This suggests that adsorption on the Au metal surface slightly decreases the pKa of p-MBA. The value of each deprotonation energy, ΔE, seems to be slightly larger than the experimental values, as the pKa of p-MBA is reported to be 4.79 [18, 32]. This over estimation could be caused by the method used to model ­H3O+. ­H3O+ could form a proton network to stabilize itself, whereas PCM modeling is not capable of reproducing such an atomically discrete and kinetic view, suggesting that the model in this study is not suitable for such absolute evaluation. Nevertheless, it allows meaningful relative comparisons between the case on Au and that in solution, because the absolute value of the energy terms E(H3O+) and E(H2O), which appear in the calculation in both cases, cancel out in the comparison. Thus, we can conclude that Au slightly enhances deprotonation of adsorbed p-MBA. To elucidate the deprotonation enhancement effect by Au, the significance of the stabilization and destabilization effect by Au on each localized part of p-MBA were analyzed. The Fig. 5  Schematic image of the deprotonation of p-MBA on Au with a geometrically optimized structure obtained from DFT calculation

13

Journal of Applied Electrochemistry

total deprotonation energy, ΔE, was decomposed by EDA [27] into each local contributing part, i.e., the Au surface, the S of the thiol group, the benzene ring, the carboxylic group (–COOH), and ­H2O, as shown in Eq. 3: ΔE =

∑ Au

ΔEA +

∑

ΔEA +

S

∑

ΔEA +

benzene

∑

ΔEA +

COOH

∑

ΔEA + ΔES

H2 O

(3) where ΔEA represents energy change for each atom in the system upon the deprotonation of p-MBA, and ΔES represents the solvation energy change. Comparing the energy changes of each part upon deprotonation for p-MBA on Au and for that in solution, the difference in the energy change, Δ(ΔE), for the sulfur atom of the thiol is the most significant, suggesting that the sulfur atom is strongly influenced by the Au surface (see Table 1). This is because the sulfur atom of p-MBA provides an electron to the Au surface to directly interact with coordination bond. Furthermore, the table shows that this difference of sulfur atom is the predominant cause of the total deprotonation energy difference.

Table 1  Energy changes in each part of the p-MBA molecule upon deprotonation

Au S Benzene ring –COOH Solvation energy H2O Total deprotonation energy

ΔE on Au

ΔE in water

Δ(ΔΕ)

58.5 − 16.0 13.0 598.0 − 412.5 − 40.4 200.5

– 85.1 31.1 635.2 − 503.3 − 40.4 207.7

58.5 − 101.1 − 18.2 − 37.2 90.8 0.0 − 7.2

Bold shows the largest absolute value among the values in the table

Journal of Applied Electrochemistry

Fig. 6  Raman spectrum a from a spot 0.10  mm above the SERS active area, and b from the spot at the substrate

(a)

I(COO-)/I(benzene)

This result suggests that the stabilization of the sulfur moiety by the Au surface slightly decreases the pKa of p-MBA. Several studies have been conducted on the pKa of molecules on a metal surface. Aoki et al. suggested that selfassembled structures on Au(111) surfaces show increases of pKa values owing to the electrostatic repulsive interaction among deprotonated species and stabilization by hydrogen bonding between protonated species in the self-assembled system [33]. Furthermore, Futamata et  al. observed a decrease in pKa of p-MBA on Ag NPs due to the sterically dense hydrogen bonds formed among the small amount of protonated p-MBA molecules in the nano-gaps between NPs, which suppresses kinetic proton attack on the rest of the deprotonated species within the nano-gaps [18]. In the case of NPs, the metal surface is not necessarily monocrystalline. Polycrystalline surfaces should make the p-MBA adsorbate layer irregular, weakening intermolecular repulsive interactions among neighboring deprotonated species on the same surface. Furthermore, in this case, the effect of the metal surface on the electronic structure of the sulfur atom of p-MBA, shown in the DFT calculations in the previous section, is also considered to be an important factor. Hence, the pKa of p-MBA should decrease on the NP surface. Nevertheless, Fig. 3 indicates that this pKa decrease is slight, and that p-MBA should work as a pH probe, as is also claimed in other works [16, 25]. Based on this evaluation of the capability of our p-MBAmodified SERS active area, we applied it to pH analysis of an electrode surface using the setup illustrated in Fig. 1. The SERS active area was located ~ 0.16 mm away from the Ni deposition electrode in a Watts bath containing 242 g/L ­NiSO4, 41 g/L ­NiCl2, and 45 g/L ­H3BO3 to maintain the pH at 2.5. The working electrode was Au. To identify the influence from the electrode surface reaction, Raman spectra from two different spots in the SERS active area were measured; one spot close to the electrode (~ 0.16 mm away) and the other ~ 10.16  mm away (see Fig.  1). The Raman spectrum in Fig. 6a is taken from a spot 0.10 mm above the SERS active area, thus measuring the bulk solution, and shows peaks attributed to H ­ 3BO3 (886 cm−1) and ­SO42− (988 cm−1), while the spectrum in Fig. 6b obtained from the substrate surface shows the peaks for p-MBA, demonstrating that the selectivity of this measurement system is sufficiently high to detect the local chemical state around the electrode interface. The potential of the working electrode was maintained at -550 mV vs. Ag/AgCl. Figure 7 shows the changes in the peak intensity ratio I(COO−)/I(benzene) for the p-MBA-modified SERS active area with negative overpotential applied to the working electrode nearby. The peak at ~ 1400 cm−1 is assigned to I(COO−) and that at ~ 1080 cm−1 is assigned as I(benzene). The profile from the spot close to the electrode shows that ­COO− intensity increases during Ni deposition, reflecting

(b)

0

200

400

600

800

1000

Deposition time / s Fig. 7  Time-dependent peak ratio [I(COO−)/I(benzene)] of p-MBA on the SERS active area during Ni deposition. (a) Measured at the spot close to the working electrode, (b) measured ~ 10.16  mm from the electrode

the fact that electron acceptance of protons ­(2H+ + ­2e− → ­H2) and proton supply ­(H2O → ­H+ + O ­ H−) occur to increase the pH near the electrode. Such behavior is not indicated by the profile at the spot far from the electrode, suggesting

13



that pH increase takes place only near the electrode. The diffusion coefficient of ­OH−, ­DOH− = 5.30 × 10 −9 ­m2/s [34], provides an approximate value of the time required for ­OH− diffusion. According to the equation, t = L2/D, where t is the time required for diffusion and L is the diffusion distance, ­OH− requires only ~ 4.83 s to diffuse to the position on the end of the SERS active area that is ~ 0.16 mm from the electrode. Figure 7 illustrates this lag. However, ­OH− requires ~ 2.0 × 104 s to diffuse to the position ~ 10.16 mm from the electrode, which suggests that ­OH− generated on the electrode surface hardly influences the chemical states of p-MBA molecules on that spot, as seen in Fig. 7. In other words, the measurement spot on the SERS active area near the electrode is covered with a diffusion layer of O ­ H−, which is in the order of 1­ 01–102 µm, whereas the spot far from the electrode is not. These data suggest that the side of the p-MBA-modified SERS active area that is close to the electrode reflects the pH environment on the electrode surface with high sensitivity, thus working as a surface pH detector for the electrode during the deposition reaction. Such a measurement system is quite helpful, for electrode reactions often accompany the steep pH changes and their mechanisms can be found in a certain pH region. The potential deviation of measurement results lies in the signal/noise ratio (S/N) in SERS spectrum. Basically, S/N differs, depending on the difference in the condition of Au NPs on the SERS active area, which leads to the fluctuation of the value of the peak intensity ratio I(COO−)/I(benzene). Nevertheless, the signal intensity of ~ 1400 cm−1 should be strong enough to identify the dehydrogenation of pH probe, as shown in Fig. 3, which suggests that pH changes around the electrode is expected to be detectable whatever the condition of the NPs is like.

4 Conclusions In order to understand the details of electrochemical interfacial reactions, surface pH on the working electrode must be measured precisely. We prepared a SERS active area with pH-probe-modified Au NPs for the investigation of such surface pH, applying it to the monitoring of pH changes on a working electrode surface during Ni electrodeposition. p-MBA, which has a –SH group to form a stable adsorbate layer on the Au NPs, –COOH to reflect the pH environment around the surface, and a benzene ring to provide strong Raman peaks as internal standards for the measurement, was applied as the pH probe. DFT calculation results indicated that the pKa of p-MBA slightly decreases upon adsorption on Au because the electronic structure of the sulfur moiety of p-MBA is influenced by the Au surface. The p-MBAmodified SERS active area close to the electrode for Ni electrodeposition (the gap between the SERS active area

13

Journal of Applied Electrochemistry

and the electrode was ~ 0.16 mm) was found to be capable of detecting the surface pH changes around the electrode surface which take place in the pH region including pKa of p-MBA. The Ni deposition reaction consumed protons to increase the pH at the electrode interface, which was identified by the increase in the peak intensity for the deprotonated species (–COO−) on the SERS active area. Our future work will focus on the attachment of various kinds of pH probe molecules to make this system applicable to variety of electrode reaction analyses. Acknowledgements  This research was financially supported in part by “Development of Systems and Technology for Advanced Measurement and Analysis” program from JST, a “Grant-in-Aid for challenging Exploratory Research (26600065)” of the MEXT, Japan, and Waseda University Grant for Special Research Project number 2017B-189.

References 1. Dahms H, Croll IM (1965) The anomalous codeposition of ironnickel alloys. J Electrochem Soc 112:771–775 2. Wei C, Bard AJ, Nagy G, Toth K (1995) Scanning electrochemical microscopy. 28. Ion–selective neutral carrier-based microelectrode potentiometry. Anal Chem 67:1346–1356 3. Park JO, Paik CH, Alkire RC (1996) Scanning microsensors for measurement of local pH distributions at the microscale. J Electrochem Soc 143(8):L174–L176 4. Klushmann E, Schultze JW (1997) pH-microscopy: theoretical and experimental investigations. Electrochim Acta 42:3123–3134 5. Romankiw LT (1970) Specific ion activity measurement at an electrode during electrolysis. IBM Tech Discl Bull 13:69 6. Deligianni H, Romankiw LT (1993) In situ surface pH measurement during electrolysis using a rotating pH electrode. IBM J Res Dev 37(2):85–95 7. Diaz SL, Mattos OR, Barcia OE, Miranda FJF (2002) ZnFe anomalous electrodeposition: stationaries and local pH measurements. Electrochim Acta 47:4091–4100 8. Koza JA, Uhlemann M, Gebert A, Schultz L (2008) The effect of a magnetic field on the pH value in front of the electrode surface during the electrodeposition of Co, Fe and CoFe alloys. J Electroanal Chem 617:194–202 9. Hessami S, Tobias CW (1993) In-situ measurement of interfacial pH using a rotating ring-disk electrode. AIChE J 39(1):149–162 10. Nakao M, Yoshinobu T, Iwasaki H (1994) Improvement of spatial resolution of a laser-scanning ph-imaging sensor. Jpn J Appl Phys 33:L394–L397 11. Han J, Brown BN, Young D, Nesic S (2010) Mesh-capped probe design for direct pH measurements at an actively corroding metal surface. J Appl Electrochem 40:683–690 12. Kneipp K, Kneipp H, Itzkan I, Dasari RR, Feld MS (1999) Ultrasensitive chemical analysis by Raman spectroscopy. Chem Rev 99:2957–2975 13. Ueno K, Juodkazis S, Shibuya T, Yokota Y, Mizeikis V, Sasaki K, Misawa H (2008) Nanoparticle plasmon-assisted two-photon polymerization induced by incoherent excitation source. J Am Chem Soc 130:6928–6929 14. Yoshida K, Itoh T, Tamaru H, Biju V, Ishikawa M, Ozaki Y (2010) Quantitative evaluation of electromagnetic enhancement in surface-enhanced resonance Raman scattering from plasmonic poperties and morphologies of individual Ag nanostructures. Phys Rev B 81:115406–11541-9

Journal of Applied Electrochemistry 15. Creager SE, Steiger CM (1995) Conformational rigidity in a selfassembled monolayer of 4-mercaptobenzoic acid on gold. Langmuir 11:1852–1854 16. Kudelski A (2009) Surface-enhanced Raman scattering study of monolayers formed from mixtures of 4-mercaptobenzoic acid and various aromatic mercapto-derivative bases. J Raman Spectrosc 40:2037–2043 17. Michota A, Bukowska J (2003) Surface-enhanced Raman scattering (SERS) of 4-mercaptobenzoic acid on silver and gold substrates. J Raman Spectrosc 34:21–25 18. Yu Y, Handa S, Yajima T, Futamata M (2013) Flocculation of Ag nanoparticles elucidating adsorbed p-mercaptobenzoic acid by surface enhanced raman scattering. Chem Phys Lett 560:49–54 19. Frisch MJ et al (2009) Gaussian 09, revision A.01, Gaussian, Inc., Wallingford 20. Lee C, Yang W, Parr RG (1988) Development of the Colle-Salvetti correlation-energy formula into a functional of the electron density. Phys Rev B 37(2):785–789 21. Miehlich B, Savin A, Stoll H, Preuss H (1989) Results obtained with the correlation energy density functionals of Becke and Lee, Yang and Parr. Chem Phys Lett 157(3):200–206 22. Becke AD (1993) Density-functional thermochemistry. III. The role of exact exchange. J Chem Phys 98(7):5648–5652 23. Hehre WJ, Radom L, Schleyer PvR, Pople JA (1986) Ab initio molecular orbital theory. Wiley, New York 24. Zhao Y, Truhlar DG (2006) A new local density functional for main-group thermochemistry, transition metal bonding, thermochemical kinetics, and noncovalent interactions. J Chem Phys 125:194101–1941-18 25. Cong VT, Ganbold EO, Saha JK, Jang J, Min J, Choo J, Kim S, Song NW, Son SJ, Lee SB, Joo SW (2014) Gold nanoparticle silica nanopeapods. J Am Chem Soc 136:3833–3841

26. Schmidt MW, Baldridge KK, Boatz JA, Elbert ST, Gordon MS, Jensen JH, Koseki S, Matsunaga N, Nguyen KA, Su S, Windus TL, Dupuis M, Montgomery JA Jr (1993) General atomic and molecular electronic structure system. J Comput Chem 14(11):1347–1363 27. Nakai H (2002) Energy density analysis with Kohn-Sham orbitals. Chem Phys Lett 363:73–79 28. Hay PJ, Wadt WR (1985) Ab initio effective core potentials for molecular calculations. Potentials for K to Au including the outermost core orbitals. J Chem Phys 82:299–310 29. Cancès E, Mennucci B, Tomasi J (1997) A new integral equation formalism for the polarizable continuum model: theoretical background and applications to isotropic and anisotropic dielectrics. J Chem Phys 107(8):3032–3041 30. Wang J, Wang G, Zhao J (2003) Structures and electronic properties of C ­ u20, ­Ag20, and A ­ u20 clusters with density functional method. Chem Phys Lett 380:716–720 31. Zhao L, Jensen L, Schatz GC (2006) Pyridine-Ag20 cluster: a model system for studying surface-enhanced raman scattering. J Am Chem Soc 128:2911–2919 32. Kim KB, Han JH, Choi H, Kim HC, Chung TD (2012) Dynamic preconcentration of gold nanoparticles for surface-enhanced raman scattering in a microfluidic system. Small 8(3):378–383 33. Aoki K, Kakiuchi T (1999) pKa of an ω-carboxylalkanethiol selfassembled monolayer by interaction model. J Electroanal Chem 478:101–107 34. Atkins PW (1998) Physical chemistry, 6th edn. W. H. Freeman, New York

13