Immiscible Oil–Water Interface: Dual Function of Electrokinetic

Article pubs.acs.org/ac

Immiscible Oil−Water Interface: Dual Function of Electrokinetic Concentration of Charged Molecules and Optical Detection with Interfacially Trapped Gold Nanorods Hye Soo Han,†,§ Jihwan Song,‡,§ Joohee Hong,† Dongchoul Kim,*,‡ and Taewook Kang*,† †

Department of Chemical and Biomolecular Engineering and ‡Department of Mechanical Engineering, Sogang University, Seoul 121-742, Korea S Supporting Information *

ABSTRACT: In this paper, we report that an immiscible oil−water interface can achieve the dual function of electrokinetic molecular concentration without external electric fields and sensitive optical detection without a microscope. As a proof-of-concept, we have shown that the concentration of positively charged molecules at the oleic acid−water interface can be increased significantly simply by controlling the pH. Three-dimensional phase field simulation suggests that the concentration of positively charged rhodamine 6G can be increased by about 10-fold at the interface. Surface-enhanced Raman spectroscopy (SERS) is utilized for label-free detection by taking advantage of this molecular accumulation occurring at the interface, since gold nanorods can be spontaneously trapped at the interface via electrostatic interaction. SERS measurements suggest that the immiscible oleic acid−water interface allows the limit of detection to be improved by 1−3 orders of magnitude.

I

modification. Seed solution is prepared by mixing 10 mL of 0.2 M CTAB and 25 μL of 0.1 M HAuCl4 followed by addition of 0.6 mL of ice-cold 0.01 M NaBH4. Growth solution is prepared by mixing 28.5 mL of 0.1 M CTAB, 255 μL of 0.01 M silver nitrate, and 1.5 mL of 0.01 M HAuCl4 followed by addition of 165 μL of 0.1 M ascorbic acid with mild stirring. Three hours after preparation, 36 μL of the seed solution is added to the growth solution and is left without disturbance overnight. The solution is centrifuged twice at 10 000 rpm for 15 min. The supernatant is removed, and the rest of the solution is redispersed in water. The aspect ratio and optical properties of gold nanorods are analyzed by transmission electron microscopy (TEM) and ultraviolet−visible (UV−vis) spectroscopy (Agilent, 8453). Gold nanorods are deposited on a carbon-coated 300-mesh TEM grid (Ted Pella.) by drop casting 20 μL of the solution and dried at room temperature in air for TEM analysis. Raman Spectroscopy and Surface-Enhanced Raman Spectroscopy (SERS) Measurements. A portable Raman spectrometer (QE65000, Ocean Optics Inc.) and a 785 nm laser module (I0785MM0350MS, Innovative Photonic Solution Inc.) are used without a microscope system for Raman and SERS measurements (see Figure S1, Supporting Information). A 785 nm laser is operated as an illumination source with 250 mW of power and 1 s of integration time. In rhodamine 6G (R6G) measurement, 1.0 mL of gold nanorod solution is mixed

ncreasing the number of target molecules in a probed volume is considered to be one of the fundamental techniques in molecular sensing.1−4 To this end, some of the techniques that have been developed include molecular trapping by the use of electric fields,2 thermodiffusion and convection,3 and the drying of water droplets on a small hydrophobic surface.4 However, these methods require electrical power to generate electric fields, a heat supply for temperature distribution, or lithography to fabricate a hydrophobic surface. Here we report that an immiscible oil−water interface is able to simultaneously achieve electrokinetic molecular concentration of charged molecules without external fields, energy supplies, and lithographic techniques as well as sensitive optical detection without a microscope. Carboxylic end groups of fatty acids can be easily deprotonated in water, uniformly rendering negative charges over the entire interface. Therefore, an immiscible fatty acid−water interface, for example, an oleic acid−water interface, could eletrokinetically concentrate charged molecules in water to the interface and trap charged gold nanoparticles at the interface. In addition, we would expect that the immiscible liquid−liquid interface where the molecules are more concentrated and the nanoparticles are trapped is clearly visible to the naked eye, making it possible to collect molecular fingerprinting information by directly illuminating the interface, thus eliminating the need for a microscope.

■

EXPERIMENTAL SECTION

Received: April 15, 2014 Accepted: May 20, 2014 Published: May 20, 2014

Preparation of Gold Nanorods. Gold nanorods are synthesized by the seed-mediated growth method5 with © 2014 American Chemical Society

6160

dx.doi.org/10.1021/ac501378x | Anal. Chem. 2014, 86, 6160−6165

Analytical Chemistry

Article

Figure 1. Schematic illustration of an immiscible oil−water interface and photographic images of an oleic acid−water interface with R6G. Charged molecules that are well dispersed in water become more concentrated at the oleic acid−water interface upon the addition of oleic acid. The illustration shows the electrostatic interactions between the deprotonated carboxylic ends and the positively charged molecules. The photographic images also show a significant concentration of R6G molecules at the oleic acid−water interface.

■

with 0.25 mL of R6G solution and 1.25 mL of DI water in a cuvette. Modeling. A three-dimensional (3D) dynamic model based on the phase field approach6 was developed. The mass flux (Jtot) of charged molecules to an oil−water interface in the system can be expressed as a sum of two fluxes, the diffusion flux (Jd) and the electrostatic interaction (Je): Jtot = Jd + Je. The flux caused by the diffusion of target molecules is represented by Jd = −D∇c, where D is the diffusivity of the molecule. In addition, the flux induced by electrostatic interactions between the uniformly charged interface and the molecules is given by Paunovic and Schlesinger as Je = −zFuc∇ϕ, where z, F, u, and ϕ represent the valence of the ionic species, Faraday’s constant, the ion mobility, and the electrostatic potential, respectively.7 Note that the ion mobility is approximated by u = D/RT.8 Consequently, the flux of the system can be expressed as Jtot = −D∇c −(zFDc/RT)∇ϕ. The flux combined with the conservation of mass requirement gives the governing equation: ∂c/∂t = ∇·{D∇c + (zFDc/RT)∇ϕ}. The governing equation is further normalized with the characteristic time tc, length lc, and diffusivity Dc = lc2/tc for computational convenience. The normalized diffusivity, Dn = D/D c, is defined by the characteristic diffusivity. In particular, the dimensionless constant α that represents the significance of the electrostatic interactions is defined as zFϕc/RT, where the characteristic electrostatic potential, ϕc, comes from the normalized electrostatic potential, ϕn = ϕ/ϕc. The normalized governing equation is given by ∂c = ∇·(Dn∇c + Dnαc∇φn) ∂tn

RESULTS AND DISCUSSION

As a proof-of-concept (Figure 1), oleic acid was selected since it is a liquid at room temperature and its carboxylic end can be easily deprotonated upon mild pH control (pKa = 5.02).9 To qualitatively confirm the electrokinetic molecular concentration of charged molecules at the oleic acid−water interface, a positively charged dye, R6G, was used. The R6G molecule is positively charged at an acidic as well as at a neutral pH10 and disperses well in water. However, upon addition of oleic acid, it becomes more concentrated at the interface with time (Figure S2, Supporting Information). To further elaborate on this observation, the effect of the solution pH on the molecular concentration at the interface is systematically examined (Figure S3, Supporting Information). At neutral and basic pH conditions (pH 7 and 8), the accumulation of R6G molecules at the interface is significant after a few hours. However, when the solution pH is decreased, this accumulation becomes less significant (pH 5). Under more acidic conditions, the accumulation of R6G at the interface is not observed since the carboxylic end cannot be deprotonated (pH 3). Note that this molecular concentration is quite reversible (Figure S4, Supporting Information). As the pH is decreased, the R6G molecules accumulated at the interface become redispersed into the aqueous phase. In addition to R6G, two more positively charged dyes, Nile blue A and pseudocyanine iodide, were also tested and produced similar results. On the other hand, negatively charged dyes (bromothymol blue and fluorescein disodium salt) are not concentrated at the same interface. Only positively charged dyes are found to be more concentrated at the interface, confirming that the significant increase in concentration of charged molecules at the oleic acid−water interface can be attributed to electrostatic interactions (Figure S5, Supporting Information). To estimate the extent of this molecular concentration of charged molecules at the oleic acid−water interface, a 3D dynamic model was developed. The dynamics of the

(1)

More details about numerical implementation and scale can be found in the Supporting Information. 6161

dx.doi.org/10.1021/ac501378x | Anal. Chem. 2014, 86, 6160−6165

Analytical Chemistry

Article

Figure 2. Computational simulation of molecular accumulation at the oleic acid−water interface: (a) schematic illustration of the simulation domain considered, (b) snapshots of accumulated R6G at five different time intervals, with an initial concentration of 10 nM, (c) time-resolved plots of the concentration of R6G molecules at the oleic acid−water interface, with three different initial concentrations.

mobility of the molecule can be altered, this alteration may affect the extent of the accumulation at the interface (Figure S6). A molecule with 1.5 times higher diffusivity than that of R6G reaches the maximum concentration at the interface about 1.5 times faster. However, the concentration maxima are the same for three cases with different diffusivities. Even if molecules with higher diffusivities are accumulated at the interface in less time under the same electrostatic potential, the diffusion of the accumulated molecules from the interface to the oil region will also be faster. Thus, it is understandable that the maximum concentration at the interface is independent of the diffusivity of the molecule. To realize the benefit of this significant molecular concentration at the oil−water interface, SERS was exploited, since charged gold nanoparticles, for example, gold nanorods, could be spontaneously trapped at the interface through electrostatic interactions (Figure S7, Supporting Information). Gold nanorods are positively charged due to the cetyltrimethylammonium bromide (CTAB) bilayer. When oleic acid is added to a gold nanorod solution, a thin black layer appears at the interface with a slight concomitant decrease in absorbance of the prepared gold nanorods in water (both 513 and 764 nm). Possible orientations of individual gold nanorods at the interface and corresponding surface plasmon resonance modes with respect to the direction of laser illumination are shown in Figure 3a. On the basis of electrostatic interactions between the interface and gold nanorods, it is reasonable to assume that gold nanorods can be oriented either vertically or horizontally at the interface. Prior to detection, the effect of gold nanorods on SERS measurement is examined. Without the R6G molecule, the Raman measurement from the interface between oleic acid and a solution of gold nanorods is obtained. The Raman transition at 1445 cm−1 is monitored as the characteristic peak for oleic acid (Figure S8, Supporting Information). The peak intensity remains nearly constant over the observation time, indicating that the increase in Raman intensity by the accumulation of the gold nanorods at the

concentration of R6G molecules at the interface are investigated (Figure 2) while assuming that the gradient of the electrostatic field is generated only along the χ3 direction, perpendicular to the interface (Figure 2a). The normalized diffusivities of R6G in water and oleic acid are 1.00 and 0.05, respectively. The dimensionless constant α approaches 16.11 It is further assumed that the electrostatic potential of oleic acid is maintained at a constant during the simulation. The simulation results for 10 nM R6G are shown in Figure 2b. R6G molecules are apparently more concentrated at the oleic acid−water interface with time. Figure 2c shows a quantitative analysis of the accumulation of R6G for different initial concentrations. At the oleic acid−water interface, R6G molecules accumulate rapidly to reach a concentration that is approximately 7 times higher than the initial concentration within 2 h. Following that, the rate of accumulation of R6G decreases, with the maximum concentration being reached at about 8 h. The R6G molecule can be accumulated at the oleic acid−water interface at concentrations that are up to 10 times higher than the initial concentration in 8 h. Initially, R6G molecules are evenly distributed in the water. It is the electrostatic potential that can cause an increase in the number of R6G molecules at the interface. As the number of R6G molecules at the interface increases, a large concentration gradient is generated across the interface, leading to diffusion of R6G molecules into the oil phase, thereby reducing the number of molecules at the interface. The extent of accumulation of charged molecules at the interface can be affected by the type of ionizable oil (i.e., electrostatic potential) (Figure S6, Supporting Information). For fatty acid methyl esters and fatty alcohols, a lower electrostatic potential can be developed at the oil−water interface.11 Therefore, with these oils, R6G concentrations at the interface are increased by about 5- and 2.5-fold, respectively. It is noteworthy that the interfaces at lower potentials take a longer time to reach concentration maxima. Since, depending on the type of charged molecule, the ion 6162

dx.doi.org/10.1021/ac501378x | Anal. Chem. 2014, 86, 6160−6165

Analytical Chemistry

Article

Figure 3. (a) Schematic illustration of SERS measurement at the interface. A 785 nm laser illuminates the interface. The sizes of gold nanorods, targets, and the laser beam are exaggerated for clarity. The top and side views of the interface represent possible orientations of gold nanorods and the corresponding surface plasmon resonance mode. (b−d) Time-resolved SERS spectra and (e−g) changes in peak intensities for R6G at the oleic acid−water interface are shown: (b, e) 1 μM, (c, f) 100 nM, (d, g) 10 nM R6G. Yellow bands show the indicator peaks at 614 and 1513 cm−1. Each point corresponds to the average of measurements repeated three times. The error bars are the standard deviation. If the intensity measured from the interface is increased over the background fluctuation (i.e., the background intensity measured from the oleic acid−water interface without analyte) by 100 counts, the signal is determined to be detectable in this study.

visible becomes clearly detectable at 1513 cm−1 after 3 h (Figure 3b−d). For comparison, the background intensity (i.e., baseline) is also shown in Figure 3e−g. It should be noted that there is no increase in SERS intensity measured from the aqueous mixture of gold nanorods and R6G without the oleic acid (Figure S11, Supporting Information). For control experiments, both a negatively charged molecule, bromothymol blue (BTB) (Figure S12, Supporting Information), and a neutral molecule, 1,2-bis(4-pyridyl)ethylene (BPE) (Figure S13, Supporting Information), were also tested. There is no noticeable increase in SERS intensities for both cases over the observation time even at high concentration (10 μM).

interface can be ruled out. Thus, the increase in Raman intensity of molecules from the interface over time is due to an increase in the number of molecules at the interface alone. SERS spectra from the oleic acid−water interface are collected as a function of time (Figure 3; Figure S9, Supporting Information) for three different initial R6G concentrations (1 μM, 100 nM, and 10 nM). Among the intrinsic Raman transitions of R6G (Figure S10, Supporting Information), 614 and 1513 cm−1 are selected as the indicator peaks for R6G, having no overlap with those of oleic acid. The SERS intensities at 614 and 1513 cm−1 significantly increase for all cases, and notably, the Raman peak for 10 nM R6G that was not initially 6163

dx.doi.org/10.1021/ac501378x | Anal. Chem. 2014, 86, 6160−6165

Analytical Chemistry

Article

Figure 4. Time-resolved SERS intensities at the oleic acid−water interface for three different initial concentrations of (a) 10 μM, (b) 1 μM, and (c) 100 nM adenine at 735 cm−1 and (d) 100 μM aniline at 533 and 1001 cm−1.

To further validate the dual function of the immiscible interface, electrokinetic concentration and label-free detection at the interface, SERS for adenine (pKa = 4.2)12 is also measured at the interface (Figure 4). Among the intrinsic Raman transitions of adenine (Figure S10, Supporting Information), the intensity at 735 cm−1 is chosen as the indicator. Interestingly, SERS signals for three initial concentrations that were not detectable without oleic acid become detectable after 1 h. Our method was also successfully applied to detect aniline, which is environmentally relevant, cationic, and spectroscopically transparent. SERS intensities at 533 and 1001 cm−1 are selected as the indicator of aniline. A 100 μM concentration of aniline becomes detectable at the interface after 5 h, further supporting the versatility of our method (Figure 4d; Figure S14, Supporting Information).

aniline, which is cationic and spectroscopically transparent. SERS measurements suggest that the immiscible oleic acid− water interface can increase the limit of detection by 1−3 orders of magnitude. Since our proposed method allows for significant molecular concentration at the interface without external electric fields, and also simultaneous optical detection from the visible interface without a microscope, we are confident that it can be applied to a wide variety of sensing applications from environmental monitoring to molecular diagnostics.

■

ASSOCIATED CONTENT

S Supporting Information *

Experimental details including methods, spectroscopic data, photographic images, and computational details as discussed in the text. This material is available free of charge via the Internet at http://pubs.acs.org.

■

CONCLUSIONS In summary, we have demonstrated that an oil−water interface can achieve the dual function of electrokinetic molecular concentration of charged molecules and label-free detection of the molecules with interfacial gold nanoparticles. When oleic acid is added to water, the oleic acid−water interface becomes uniformly negatively charged over the entire interface via the deprotonation of the carboxylic ends and forces positively charged molecules to diffuse to the interface without the application of external electric fields. The accumulation of positively charged dye molecules is clearly visible to the naked eye and is reversible by controlling the pH. A 3D phase field simulation was carried out to estimate the extent of accumulation of charged molecules at the interface. Simulation results show that the interface concentration of R6G increases to 10 times the initial concentration. According to the simulation, the type of ionizable oil and the diffusivity of the target molecule also affect the concentration maxima and time required to reach the maximum concentration. To realize the benefit of this molecular concentration at the oleic acid−water interface, SERS for label-free detection was performed, since positively charged gold nanorods are also trapped at the interface. The SERS signal for 10 nM R6G, which was not detectable in solution, is clearly appreciable from the interface. Interestingly, SERS signals for three initial adenine concentrations that were not detectable without oleic acid become detectable. Our method was also successfully applied to detect

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. Author Contributions §

H.S.H. nad J.S. contributed equally to this work.

Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This research was supported by the International Research & Development Program of the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT & Future Planning (MSIP) (Grant 2013K1A3A1A32035444), the NRF funded by the Korean government (Ministry of Education, Science and Technology, MEST) (Grant 2011-0015749), the Leading Foreign Research Institute Recruitment Program through the NRF funded by the MSIP (Grant 2013K1A4A3055268), the Basic Science Research Program through the NRF funded by the MSIP (Grants NRF2013R1A1A2011263 and 2013056561), and Sogang University Research (Grant SRF-201314004). 6164

dx.doi.org/10.1021/ac501378x | Anal. Chem. 2014, 86, 6160−6165

Analytical Chemistry

■

Article

REFERENCES

(1) Nam, J. M.; Thaxton, C. S.; Mirkin, C. A. Science 2003, 301 (5641), 1884−1886. (2) (a) Cho, E. S.; Kim, J.; Tejerina, B.; Hermans, T. M.; Jiang, H.; Nakanishi, H.; Yu, M.; Patashinski, A. Z.; Glotzer, S. C.; Stellacci, F.; Grzybowski, B. A. Nat. Mater. 2012, 11 (11), 978−985. (b) Cho, H. S.; Lee, B.; Liu, G. L.; Agarwal, A.; Lee, L. P. Lab Chip 2009, 9 (23), 3360−3363. (3) Kang, T.; Hong, S.; Choi, Y.; Lee, L. P. Small 2010, 6 (23), 2649−2652. (4) (a) De Angelis, F.; Gentile, F.; Mecarini, F.; Das, G.; Moretti, M.; Candeloro, P.; Coluccio, M. L.; Cojoc, G.; Accardo, A.; Liberale, C.; Zaccaria, R. P.; Perozziello, G.; Tirinato, L.; Toma, A.; Cuda, G.; Cingolani, R.; Di Fabrizio, E. Nat. Photonics 2011, 5 (11), 683−688. (b) Shao, F. F.; Ng, T. W.; Liew, O. W.; Fu, J.; Sridhar, T. Soft Matter 2012, 8 (13), 3563−3569. (5) Gou, L.; Murphy, C. J. Chem. Mater. 2005, 17, 3668−3672. (6) (a) Lu, W.; Kim, D. C. Nano Lett. 2004, 4 (2), 313−316. (b) Lu, W.; Kim, D. Acta Mater. 2005, 53 (13), 3689−3694. (c) Song, J.; Kim, D. Phys. Rev. E 2010, 82 (5), 051902. (7) Paunovic, M.; Schlesinger, M. Fundamentals of Electrochemical Deposition, 2nd ed.; Wiley-Interscience: Hoboken, NJ, 2006; x, 373 pp. (8) Griffiths, S. K.; Nilson, R. H.; Ting, A.; Bradshaw, R. W.; Bonivert, W. D.; Hruby, J. M. Microsyst. Technol. 1998, 4 (2), 98−101. (9) Riddick, J. A.; Bunger, W. B.; Sakano, T.; Weissberger, A. Organic Solvents: Physical Properties and Methods of Purification, 4th ed.; Wiley: New York, 1986; xv, 1325 pp. (10) (a) Daniels, C. R.; Tauzin, L. J.; Foster, E.; Advincula, R. C.; Landes, C. F. J. Phys. Chem. B 2013, 117 (16), 4284−4290. (b) Yan, R. X.; Liang, W. J.; Fan, R.; Yang, P. D. Nano Lett. 2009, 9 (11), 3820− 3825. (c) Milanova, D.; Chambers, R. D.; Bahga, S. S.; Santiago, J. G. Electrophoresis 2011, 32 (22), 3286−3294. (11) Vogel, V.; Möbius, D. J. Colloid Interface Sci. 1988, 126 (2), 408−420. (12) Camargo, A. P. M.; Baumgartel, H.; Donner, C. Phys. Chem. Chem. Phys. 2003, 5 (8), 1657−1664.

6165

dx.doi.org/10.1021/ac501378x | Anal. Chem. 2014, 86, 6160−6165