Superhydrophobic-Oleophobic Ag Nanowire ... - ACS Publications

Sep 17, 2014 - The ultratrace detection and quantification of toxins in both water and organic liquids remains a challenge due to the random spreading...
0 downloads 0 Views 5MB Size
Article pubs.acs.org/ac

Superhydrophobic-Oleophobic Ag Nanowire Platform: An AnalyteConcentrating and Quantitative Aqueous and Organic Toxin SurfaceEnhanced Raman Scattering Sensor Xing Li,†,§ Hiang Kwee Lee,†,‡,§ In Yee Phang,‡ Choon Keong Lee,† and Xing Yi Ling*,† †

Division of Chemistry and Biological Chemistry, School of Physical and Mathematical Sciences, Nanyang Technological University, 50 Nanyang Avenue, Singapore 637371 ‡ Institute of Materials Research and Engineering, A*STAR (Agency for Science, Technology and Research), 3 Research Link, Singapore 117602 S Supporting Information *

ABSTRACT: The ultratrace detection and quantification of toxins in both water and organic liquids remains a challenge due to the random spreading and dilution of liquids on substrate-based sensors, especially for organic liquids with low surface tension. Herein, we fabricate a superhydrophobicoleophobic (SHP-OP) 3D Ag nanowire mesh-like surfaceenhanced Raman scattering (SERS) platform to overcome the random spreading issue, demonstrating ultratrace toxin sensing in both water and organic liquid. Our SHP-OP SERS platform is able to concentrate analyte solutions in water and toluene to 100-fold and 8-fold smaller areas, respectively, as compared to its omniphilic counterparts. The synergy of analyte-concentrating ability and intense SERS-enhancing properties on our SHP-OP SERS platform enables quantitative and ultratrace detection of melamine and Sudan I down to 0.1 fmol in water and toluene, respectively, using just 1 μL of analyte solution. These detection limits are 103-fold lower than the regulatory limits, clearly indicating our SHP-OP SERS platform as an appealing universal ultratrace toxin sensor. The ultratrace detection of spiked melamine in liquid milk down to 100 fmol also highlights the suitability of our SHP-OP SERS platform for the sensing of food toxins in real samples.

S

ideal SERS analytical platform should possess excellent detection sensitivity for both water- and organic-soluble toxins in their native solvents without the need for additional pretreatment/concentration processes that can otherwise cause the loss of precious analytes. We hypothesize that SERS substrates with antiwetting propertiesboth (super)hydrophobic and (super)oleophobicare able to address the aforementioned limitations of conventional liquid-wetting SERS platforms by confining both aqueous and organic analyte solutions into a tiny area. (Super)omniphobic SERS platforms bring about the “analyteconcentrating effect”, which together with the evaporation of solvent greatly enhances detection sensitivity for molecular sensing. (Super)omniphobic substrates refer to surfaces possessing antiwetting properties toward both water and organic liquids, with respective contact angles ≥ 150°, and typically comprise hierarchical micro- and nanoscale roughness grafted with omniphobic surface functionality.14−16 Although superhydrophobic SERS platforms (water contact angle ≥ 150°)17,18 have been reported to detect rhodamine 6G sensing

urface-enhanced Raman scattering (SERS) is a sensitive vibrational spectroscopic technique capable of providing a specific molecular fingerprint of an analyte, even at extremely low concentrations. The SERS sensitivity arises from the intense electromagnetic fields localized near the surfaces of noble metal nanostructures as a result of the excitation of their localized surface plasmon resonances.1−3 Hence, SERS is ideal for ultratrace detection of toxins, allowing identification of potential health hazards and enforcement of safety regulatory standards to address the increasing number of contaminated food products over recent years,4−7 for instance, the outbreak of water-soluble melamine contamination in milk powder7 and the abuse of organic-soluble Sudan I (1-phenylazo-2-naphthol), a carcinogenic industrial-used azo dye, as a food colorant. These cases pose serious health hazards to the general public.6,8,9 Despite the high sensitivity of SERS, the ultratrace toxin detections remain challenging, especially for the sensing of nonpolar analytes in their native organic solvents, where only a few studies have been reported.10,11 This is because most reported SERS platforms exhibit strong liquid-wetting behavior, contributing to random spreading and dilution of aqueous/ organic solutions across a substrate-based analytical platform.12 The SERS detection sensitivity for the sensing of nonpolar toxins is also hampered by the strong interfering signal originating from their native organic solvents.13 Hence, an © 2014 American Chemical Society

Received: August 7, 2014 Accepted: September 17, 2014 Published: September 17, 2014 10437

dx.doi.org/10.1021/ac502955w | Anal. Chem. 2014, 86, 10437−10444

Analytical Chemistry

Article

at femto-/attomolar level,19,20 the ultratrace detections are limited to aqueous analytes only. The difficulty in nonpolar analyte detection persists. Herein, we fabricate a superhydrophobic-oleophobic (SHPOP) 3D Ag nanowire SERS platform and demonstrate its ultratrace detection and quantification of toxins in aqueous and organic liquids. To our knowledge, this is the first report exploiting oleophobicity to concentrate the number of organicsoluble analyte molecules per surface area. In combination with a highly homogeneous SERS platform, ultratrace detection and quantification of toxins in both water and organic liquids is demonstrated. Our strategy utilizes the large-area homogeneous electromagnetic “hot spots” imparted by 3D Ag nanowire mesh-like array and the analyte-concentrating effect from antiwetting properties to enhance the nanowire array’s SERS detection sensitivity. We first assemble Ag nanowires in a layerby-layer manner to form a highly roughened and nonclosepacked 3D mesh-like structure and use subsequent functionalization with perfluorodecanethiol to render antiwetting properties. The effect of Ag nanowire layers on the antiwetting properties will be systematically studied using various common liquids, including water, ethylene glycol, dimethyl sulfoxide, toluene, and n-dimethylformamide. Using our optimized SHPOP SERS platform, the analyte-concentrating effects on different aqueous and organic liquids, the detection limits of water-soluble melamine and organic-soluble Sudan I, and their corresponding SERS enhancement factors will be evaluated and compared with both regulatory limits and omniphilic SERS platforms (easily wetted by both water and organic liquids).

nanowires were dispersed in ethanol/chloroform mixture (1:1). For the fabrication of Ag nanowire monolayer, different surface densities (10, 24, 72, 97, and 131 μg/cm2) of Ag nanowires were dispensed on the water surface in the Langmuir−Schaefer trough. The O2 plasma-treated substrate was brought to contact with the Ag nanowires monolayer and then blown dry with nitrogen gas. Similarly, 3D Ag nanowire mesh-like arrays were fabricated using surface densities of Ag nanowires of 97, 90, 80, 70, 60, 50, 40, 30, 20, and 10 μg/cm2 for the first 10 layers, from bottom to top, and maintained at 10 μg/cm2 until the 40th layer. The Ag nanowires arrays were assembled at an alternating direction of ∼90° between each layer to form a mesh-like structure. The as-fabricated Ag nanowire arrays were then immersed in PFDT (5 mM, 5 mL) hexane solution for at least 15 h. Quantification of Concentrating Effect. Silica beads were synthesized using a modified Stober process. Water (1.5 mL) followed by ammonia (1.5 mL) was slowly added into tetraethyl orthosilicate ethanolic solution (0.35 M, 7.6 mL) under vigorous stirring. After 30 min, the crude product was washed with copious amounts of ethanol and water and subsequently redispersed in water. For toluene-soluble silica beads, the as-synthesized silica beads were functionalized with triethoxyoctylsilane (0.21 M, 7.5 mL) in ethanol for at least 12 h, washed with a copious amount of ethanol, and subsequently redispersed in toluene. The water-soluble silica beads (1.53− 1.53 × 10−5 g/L) and toluene-soluble silica beads (4.75−4.75 × 10−5 g/L) were prepared in water and toluene, respectively. Each solution (1 μL) was deposited and dried on a superhydrophobic-oleophobic platform under ambient conditions. The dried spots were characterized by SEM, and their surface areas were measured using ImageJ software. A similar experiment was also conducted on the 10-layer poly(vinylpyrrolidone)-grafted Ag nanowire arrays (omniphilic platform). SERS Characterization of Superhydrophobic-Oleophobic Platform using Melamine and Sudan I. Solutions of melamine were prepared in ultrapure water from 10−7 to 10−10 M. Solutions of Sudan I were prepared in toluene from 5 × 10−5 to 10−10 M. Each concentration (1 μL) for melamine and toluene was deposited on different regions of the superhydrophobic-oleophobic SERS platform and allowed to dry for subsequent characterization by SERS. Detection of Melamine in Commercial Liquid Milk using Superhydrophobic-Oleophobic SERS Platform. Liquid milk, containing 10−6 M spiked melamine, was treated using a modified method reported in the literature.21 Briefly, 200 μL of 1 M HCl was added to 4 mL of liquid milk and centrifuged for 30 min at 10 000 rpm. The supernatants were collected and filtered using a 0.22 μm PTFE filter. Added to the filtered solution was 60 μL of 1 M NaOH, which was subsequently subjected to similar centrifugation and filtration. The resulting pale-yellow clear solution was diluted 10-fold using ultrapure water, and a 1 μL droplet was dried on superhydrophobic-oleophobic SERS platform for characterization. Materials Characterization. Scanning electron microscope (SEM) imaging was performed using a JEOL-JSM7600F microscope. Contact angles were measured using Theta Lite tensiometer. Water, EG, DMSO, DMF, and toluene were chosen as the liquids for contact angle measurements. Static contact angles were measured with 4 μL droplets. Advancing and receding contact angles were measured by a drop-shape



EXPERIMENTAL SECTION Chemicals. Silver nitrate (99+%), poly(vinylpyrrolidone) (PVP, average molecular weight (MW) = 360 000), anhydrous ethylene glycol (EG, 99.8%), 1H,1H,2H,2H-perfluorodecanethiol (PFDT, 97+%), 4-methylbenzenethiol (98%), triethoxyoctylsilane (97.5+%), tetraethyl orthosilicate (TEOS, 99+%), dimethyl sulfoxide (DMSO, 99.9+%), melamine (99%), and Sudan I (95+%) were purchased from Sigma-Aldrich; NaCl (99.5%) was from Goodrich Chemical Enterprise. n-Dimethylformamide (DMF, 99.8+%) and ethanol (99.99%) were from Fisher Chemical. Hexane (99%) was from Fulltime. Toluene (99.5%) was from J.T. Baker. Ammonia hydroxide was purchased from Sinopharm Chemical Reagent Co. Ltd. Commercial liquid milk was obtained from a local grocery store. All chemicals, except liquid milk, were used without further purification. Milli-Q water (>18.0 MΩ·cm) was purified with a Sartorius arium 611 UV ultrapure water system. Synthesis and Purification of Ag Nanowires. A solution of NaCl (12 mM) and PVP (0.45 M) in EG (10 mL) was heated at 160 °C for 5 min. AgNO3 (0.12 M, 5 mL) in EG solution was then added dropwise into the mixture at a rate of 5 mL/h by a syringe pump while stirring at 1000 rpm. The nanowire mixture was further heated for 30 min at 160 °C after the complete addition of AgNO3 and then cooled to room temperature. The crude Ag nanowires were washed with copious amount of acetone and ethanol and subsequently redispersed in ethanol and stored overnight in the refrigerator. To remove large colloidal particles, the top suspension was decanted and collected as the purified Ag nanowires while the sediment was discarded. Fabrication of Superhydrophobic-Oleophobic Platform. Silicon substrate was treated with oxygen plasma (Femto Science, Cute-MP/R, 100 W) for 5 min before assembly. Ag 10438

dx.doi.org/10.1021/ac502955w | Anal. Chem. 2014, 86, 10437−10444

Analytical Chemistry

Article

Figure 1. Fabrication and characterization of superhydrophobic-oleophobic (SHP-OP) 3D Ag nanowire SERS platform. (A) Schematic illustration of the fabrication of SHP-OP SERS platform via layer-by-layered Langmuir−Schaefer assembly of Ag nanowires into 3D mesh-like arrays and subsequent functionalization with perfluorodecanethiol (PFDT). To achieve nonclose-packed Ag nanowire arrays, the mass of nanowires per surface area is intentionally reduced with increasing buildup of Ag nanowire layers. (B−E) SEM images of 1, 2, 10, and 40 layers of Ag nanowires, respectively. (F) Root-mean-square (RMS) roughness, (G) static contact angles of various liquids, and also (H) contact angle hysteresis of water and toluene at different layers of Ag nanowire. “Control” refers to the thermal-evaporated Ag film. The blue and purple dotted lines denote phobicity (≥90°) and superphobicity (≥150°), respectively.

diameter and length of (85 ± 9) nm and (15 ± 2) μm, respectively, are synthesized using the modified polyol method, which exhibits characteristic localized surface plasmon resonances (LSPRs) at 349 and 384 nm (Supporting Information, Figure S1).24 Fabrication of Nonclose-Packed Ag Nanowire Meshlike Arrays. We begin by systematically fabricating nonclosepacked Ag nanowire mesh-like arrays of varying layers, via layer-by-layered Langmuir−Schaefer technique. Our unique design of a nonclose-packed Ag nanowire array allows greater exposure of underlying Ag nanowire layers for improved surface roughness and laser penetration depth, overcoming the low surface roughness and limited laser permeability (≤3 layers) typical in conventional closely packed woodpile-like structures.22 Briefly, a nonclose-packed Ag nanowire array is assembled using consecutive layers of orthogonally aligned Ag nanowires with increasing interparticle gap (Figure 1A), tunable by decreasing the mass of nanowires added per unit area during assembly (Supporting Information, Figure S2). Generally, we observe an approximate orthogonal alignment of Ag nanowire layers with increasing interparticle gaps as the layer of Ag nanowire increases until the 40th layer, highlighting the excellent reproducibility of our protocol to fabricate largearea mesh-like arrays (Figure 1B−E, Supporting Information, Figure S3). Although a nonperfect orthogonal alignment is obtained during assembly, the objective of creating a large number of vertically and laterally stacked Ag nanowires to

analysis of respective liquid at a flow rate of 0.1 μL/s. All contact angle measurements were repeated at least 5 times at different locations of the substrates. The root-mean-square roughness of the as-prepared substrates was measured using JPK Nanowizard3 BioScience atomic force microscopy (AFM) on a Zeiss inverted microscope. Silicon cantilevers from Budgetsensors (model Tap300-G with 30 nm aluminum back reflex coating) were used for noncontact mode operation. SERS characterizations were performed using both x−y and x−z imaging mode of the Ramantouch microspectrometer (Nanophoton Inc., Osaka, Japan) with an excitation wavelength of 532 nm under a 20× (N.A. 0.45) objective lens. An accumulation time of 50 s and a laser power of 0.065 mW were applied for the detection of melamine, while 10 s accumulation time and a laser power of 0.017 mW were applied for the detection of Sudan I. 4-Methylbenzenethiol was detected using 1 s accumulation time and a laser power of 0.065 mW. All SERS data were collected between 200 to 2000 cm−1. All SERS spectra were obtained by averaging at least 6 spectra.



RESULTS AND DISCUSSION Characterization of Ag Nanowires. Ag nanowires are used to form stable 3D nonclose-packed mesh-like superhydrophobic-oleophobic SERS sensing platform due to their ability to generate high density of electromagnetic hot spots at their intersections.22,23 Ag nanowires with monodisperse 10439

dx.doi.org/10.1021/ac502955w | Anal. Chem. 2014, 86, 10437−10444

Analytical Chemistry

Article

reported using hierarchical structures and/or engineered “reentrant” structures,14 their applicability as SERS sensor is greatly limited due to the need of Raman-interfering polymeric substrate and also costly and sophisticated equipment/ protocols.30,31 Therefore, our nonclose-packed 3D Ag nanowire mesh-like array excels as a simple yet efficient strategy for the direct fabrication of antiwetting SERS platform without the need for sophisticated protocols/equipment. Hereupon, water and toluene are chosen as the polar and nonpolar phases, respectively, for subsequent evaluations due to their wide usage as industrial/laboratory-based solvents. Contact angle hysteresis, the difference between advancing and receding contact angles, is investigated to quantify the extent of liquid−solid adhesion, which has a direct influence on the analyte-concentrating effects. The Ag nanowire arrays up to 10 layers exhibit a decrease and increase in water-based and toluene-based contact angle hysteresis (Figure 1H), respectively. Eventually, the contact angle hysteresis plateaus beyond the 10th layers of Ag nanowire at approximately 10° and 107° using water and toluene, respectively. This indicates that our 3D Ag nanowire arrays with ≥10-layers of Ag nanowires demonstrate a lotus-like (contact angle hysteresis < 10°) character in water and a rose-petal-like (contact angle hysteresis > 90°) behavior in toluene,29,32,33 denoting low and high extent of liquid−solid adhesion, respectively. Such contrasting phenomenon of water- and toluene-based contact angle hysteresis is due to the large differences in their surface tensions (refer to Supporting Information for detailed discussion).28 The lotus-like and rose-petal-like behavior toward water and toluene, respectively, are further complemented by the drying profiles of the corresponding liquids, whereby three-phase contact line recedes for water upon drying but remains pinned for toluene (Supporting Information, Figure S6). The “control” Ag film exhibits a moderate contact angle hysteresis of ∼37° for both water and toluene due to its low surface roughness. Hereafter, we only use 10-layered Ag nanowire arrays as the superhydrophobic-oleophobic (SHPOP) platform owing to its greater ease of fabrication without compromising both antiwetting properties and laser permeability (Supporting Information, Figure S7). Analyte-Concentrating Effect. The analyte-concentrating effect of our SHP-OP platform is studied by comparing its liquid−solid contact areas of 1 μL water and toluene with an omniphilic platform (contact angles < 90° for all liquids), in both wet and dried states. The omniphilic platform is represented by a 10-layered poly(vinylpyrrolidone) (PVP)functionalized Ag nanowire mesh-like array exhibiting water and toluene contact angles of approximately 37° and 5°, respectively. By casting 1 μL liquid droplets onto the respective platforms, our SHP-OP platform is found to confine water and toluene droplets into small liquid−solid contact areas of 0.13 and 1.77 mm2, respectively (Figure 2A). In contrary, water and toluene spread readily across the omniphilic platform, resulting in larger liquid−solid contact areas of 4.91 and 13.20 mm2, respectively (Figure 2C). Hence, prior to the drying of the liquid droplets, our SHP-OP platform demonstrates at least 38fold and 7-fold reduction of liquid−solid contact area for water and toluene, respectively. We further examine the analyte−solid contact area upon the drying of the liquid droplets to determine the overall analyteconcentrating effect of our SHP-OP platform, which reflects the actual condition during the SERS measurement. Silica bead and/or dye are added into water and toluene to aid in the

generate a high density of SERS hot spots has been achieved, which is in agreement to the literature.22,25 Hence, such an alignment of Ag nanowires will be simplified as “orthogonally aligned” for subsequent discussion. Effect of Ag Nanowire Layer on Surface Roughness. The effect of surface asperities on antiwetting properties is determined by quantifying the root-mean-square (RMS) roughness of the as-fabricated Ag nanowire layer(s) using atomic force microscopy (AFM). We observe an initial increment of surface roughness from (39 ± 4) to (124 ± 16) nm when the layer of Ag nanowires increases from 1 to 10 layers, respectively, and the roughness subsequently plateaus beyond the 10th layer at ∼139 nm (Figure 1F). Such rapid initial increase of surface roughness with the buildup of Ag nanowire layer is attributed to the large densities of interparticle gaps present along both the lateral and vertical planes, highlighting the importance of nonclose-packed structure in enhancing the exposure of underlying Ag nanowire layers (refer to Supporting Information, Figure S4, for detailed comparison with closely packed woodpile-like arrays). The subsequent saturation of surface roughness is possibly due to the limitation of AFM cantilever to probe underlying Ag nanowire layers efficiently. Nevertheless, the importance of nonclose-packed 3D Ag nanowire mesh-like array to produce large surface roughness for antiwetting properties is clearly exemplified. Furthermore, the orthogonal alignment of Ag nanowires also makes it more structurally ordered, which is crucial for homogeneous antiwetting and SERS-enhancing properties, compared to other fabrication methods such as via parallel alignment or drop-casting (Supporting Information, Figure S5). Effect of Ag Nanowire Layer on Antiwetting Properties. The as-fabricated 3D Ag nanowire arrays are then functionalized with 1H,1H,2H,2H-perfluorodecanethiol (PFDT) to impart antiwetting properties, which are subsequently evaluated using a variety of liquids, including water, ethylene glycol (EG), dimethyl sulfoxide (DMSO), ndimethylformamide (DMF), and toluene (Figure 1G). We use a 100 nm thermal-evaporated Ag film as a control for comparison with our Ag nanowire arrays. When water, a common solvent for polar analytes, is used for the surface wettability measurement, the Ag film exhibits a contact angle of (118 ± 1)o. In comparison, a single-layered Ag nanowire array exhibits a significantly higher water angle of (151 ± 2)o, which further increases to (170 ± 3)o on the fourth layer of the Ag nanowire. Further extension of Ag nanowires to the 40th layer maintains a relatively constant contact angle of ∼170°. EG, DMSO, DMF, and toluene are also observed to demonstrate similar rapid increments in antiwetting properties until the 10th layers of Ag nanowire, followed by the plateauing of contact angles when extended to the 40th layers (Figure 1G). Maximum contact angles of EG, DMSO, DMF, and toluene are determined as 158°, 130°, 111°, and 112°, respectively. The smaller maximum contact angles of organic liquids versus those of water are expected due to their significantly lower surface tensions (28.5 ≤ γ ≤ 47.3 mN/m) compared to water (γ = 72.7 mN/m).26,27 When using the control Ag film, limited antiwetting properties (contact angles ≤ 84°) are observed for DMSO, DMF, and toluene. Such observations on the increase in contact angles, for both water and organic liquids, with surface roughness agree well with the Wenzel and Cassie−Baxter antiwetting models, which state that surface roughness is essential for liquid-repelling properties.28,29 We note that, although superomniphobic platforms have been 10440

dx.doi.org/10.1021/ac502955w | Anal. Chem. 2014, 86, 10437−10444

Analytical Chemistry

Article

the ultratrace detection of polar and nonpolar analytes in their native solvents. SERS Detection of Food Toxins. SERS responses of water-soluble melamine and toluene-soluble Sudan I are determined to evaluate the performance of our SHP-OP platform for ultratrace detection and quantification of toxins in their respective native solvents, without any preconcentration/ treatment processes. Various aqueous melamine droplets (1 μL) are deposited and dried on the SHP-OP SERS platform prior to SERS measurement. All SERS spectra exhibit characteristic vibrational modes of melamine (Figure 3A and

Figure 2. Analyte-concentrating effect of SHP-OP platform demonstrated using water and toluene. (A) Contact angle images of 1 μL liquid droplets and (B) SEM images of respective dried silica bead aggregation spots on SHP-OP platform. (C) Contact angle images of 1 μL liquid droplets and (D) digital images of dried dyes aggregation spots on omniphilic platform, which is a 10-layered PVP-functionalized Ag nanowire mesh-like array. (E) Plot of dried aggregation areas of silica beads, dispersed in water and toluene, on SHP-OP platform with respect to silica bead concentrations. Insets are highmagnification SEM images within the respective dried aggregation spots of toluene (top) and water (bottom), using a silica bead concentration of ∼10 mg/L. (F) Comparison of analyte-concentration factor of SHP-OP platform with an omniphilic platform.

visualization of the dried aggregation spots. By drying 1 μL of silica bead-containing water and toluene droplets on our SHPOP platform, we notice a further 2.2-fold reduction of aqueous analyte−solid contact area to ∼0.05 mm2, whereas toluene analyte−solid contact area remains relatively unchanged at 1.84 mm2 (Figure 2B). Such analyte-concentrating effect of our SPH-OP platform is also observed to be independent of analyte concentrations (Figure 2E and Supporting Information, Figure S8). In contrary, the analyte−solid contact area of the omniphilic platform remains unchanged before and after drying of both water and toluene (Figure 2D). The results therefore clearly demonstrate that our SHP-OP platform prevents the random spreading of liquids by confining liquid−solid interaction into approximately 100-fold and 8-fold smaller area and, therefore, more concentrated than on an omniphilic platform when aqueous- and toluene-based analyte solutions are used, respectively (Figure 2F). The greater reduction of analyte−solid contact area for water than for toluene is mainly attributed to the low and high adhesion of water and toluene, respectively, to SHP-OP platform such that water readily undergoes a recession of 3-phase contact line upon drying (Supporting Information, Figure S6). It is noteworthy that our SHP-OP platform exhibits aqueous analyte-concentrating effect at least 7-fold superior than reported superhydrophobic platforms,34,35 and more importantly, it is the first demonstration of the ability to concentrate nonpolar analytes in organic liquids. Hence, it is clearly evident that antiwetting properties are essential for imparting excellent analyteconcentrating abilities to greatly enhance the detection sensitivity on substrate-based sensing platforms, especially for

Figure 3. SERS analysis of melamine and Sudan I on the SHP-OP SERS platform using water and toluene as solvents, respectively. Typical (A) SERS spectrum and (B) SERS image of dried melamine aggregation spot. (C) SERS intensities of melamine as a function of its concentrations. Typical (D) SERS spectrum and (E) SERS image of dried Sudan I aggregation spot. (F) Correlation of SERS intensities of Sudan I with its concentrations. The blue and purple dotted lines denote the linearity ranges of (C) and (F), respectively.

Supporting Information (SI), Table S1).5,36 Using the most intense SERS band at 685 cm−1, the SERS image of the dried aggregation spot demonstrates a surface area of ∼0.05 mm2 (Figure 3B), which is similar to the spot area previously determined using SEM. Correlation of the SERS intensities with melamine concentrations reveals a linear decrease of SERS intensities from (26 425 ± 1 093) counts to (1 040 ± 145) counts between melamine concentrations from 0.1 pmol to 0.1 fmol, respectively (Figure 3C and SI, Figure S9). The detection limit for melamine is determined to be 0.1 fmol, where the SERS band at 685 cm−1 is still clearly distinguishable (signal-tonoise ratio > 3; SI, Figure S9A). In the absence of melamine, the SHP-OP SERS platform gives a featureless SERS spectrum (“control” in SI, Figure S9A), affirming that the SERS band at 685 cm−1 originates from melamine, and the perfluorodecanethiol-grafted Ag nanowires do not interfere with ultratrace 10441

dx.doi.org/10.1021/ac502955w | Anal. Chem. 2014, 86, 10437−10444

Analytical Chemistry

Article

detection due to the weak/moderate SERS responses from C− F moieties.37 The analytical enhancement factor is determined to be (1.76 × 1011) (SI, Figure S9), indicating that our SHP-OP SERS platform is able to enhance Raman signal by ∼1011-fold compared to solid melamine. Similarly, our SHP-OP SERS platform is also capable of quantitative and ultratrace sensing of Sudan I in toluene. Characteristic vibrational fingerprints of Sudan I are observed in all SERS spectra acquired (Figure 3D and SI, Table S2).8 We note that the dried toluene aggregation spot is larger than the maximum scan area and therefore cannot be fully imaged using SERS (Figure 3E). Employing the most intense band at 1233 cm−1, a linearity range between 0.1 and 50 pmol is observed (Figure 3F). Further dilution of analyte concentrations to eventually 0.1 fmol gives a relatively constant SERS intensity of ∼10 counts, which we attribute to the submonolayer Langmuir adsorption of Sudan I on our SERS platform (SI, Figure S10).19,35 Thus, the detection limit of Sudan I is determined as 0.1 fmol (SI, Figure S11), and we again exclude any potential background interferences from our SHP-OP SERS platform (control in SI, Figure S11A). Consequently, an analytical enhancement factor of (1.87 × 108) is determined (SI, Figure S11), which is ∼103-fold less compared to the analytical enhancement factor calculated using water-soluble melamine due to the difference in analyte concentration factor for both liquids, and also the Raman cross sections of the respective analytes. Nevertheless, the high analytical enhancement factors of 108−1011 assert the presence of a large density of intense electromagnetic hot spots required for ultratrace sensing application.22 We would like to emphasize that the detection limits of 0.1 fmol (equivalent to 0.1 nM), for both melamine and Sudan I, using our SHP-OP SERS platform is 103-fold less than the current regulatory limits of 0.1 μM.7,38 Comparison with Control SERS Platforms. Our SHPOP SERS platform is also determined to be superior to other SERS platforms: 100 nm thermal-evaporated Ag film and PVPfunctionalized 10-layered Ag nanowire array. SERS evaluation of these two control SERS platform are performed under the same experimental conditions, while concentrations of melamine and Sudan I applied are 10 fmol and 1 pmol, respectively. For our SHP-OP SERS platform, we again emphasize that characteristic and distinct SERS bands are observed even at ultratrace melamine and Sudan I concentrations of 0.1 fmol (Figure 4A), which are their detection limits. However, for both the control SERS platforms, no distinct SERS band of melamine and Sudan I is observed despite using 100-fold and 104-fold higher concentrations of analytes (Figure 4B−C), respectively. Furthermore, PVPfunctionalized SERS platform also demonstrates strong and broad background spectrum from the PVP moiety,39 used for Ag nanowire synthesis, near the 685 and 1233 cm−1 regions of interest. Due to the absence of distinct SERS bands of melamine and Sudan I, analytical enhancement factor for both control SERS platforms cannot be accurately determined. Nonetheless, the importance of having strong SERS-enhancing single-crystalline Ag nanowires, antiwetting properties and nonRaman-interfering surface moiety in our SHP-OP SERS platform are again highlighted through these control SERS platforms. In addition, our SHP-OP platform also possesses large, homogeneous and intense SERS-active area, extending over 400 μm × 142 μm, essential for ultratrace and quantitative molecular detection, making it superior over random Ag

Figure 4. SERS evaluation of (A) SHP-OP SERS platform with (B) PVP-functionalized 10-layered Ag nanowire array and (C) Ag film. Schematic illustration of respective SERS platforms (left) and SERS spectra (right) of melamine (blue) and Sudan I (orange) are included for comparison. The concentrations of melamine and Sudan I used are both at 0.1 fmol for (A), and 10 fmol and 1 pmol, respectively, for (B) and (C). AEF refers to analytical enhancement factor.

structures fabricated via drop-casting of Ag nanowires (Supporting Information, Figure S12). SERS Detection of Melamine in Liquid Milk. To evaluate the applicability of our SHP-OP SERS sensor for real sample detection, we spike 1 μM of melamine into commercial liquid milk (Figure 5A) and remove most of the

Figure 5. SERS detection of melamine in liquid milk. Digital images of (A) commercial liquid milk and (B) treated milk. (C) SERS spectra of treated milk originally spiked with 1 μM melamine (blue), melamineabsence-treated milk (red) and SERS platform in the absence of milk and melamine (control; black). Both blue and red spectra are obtained from the treated milk solution after a 10-fold dilution of (B). Hence, the effective melamine concentration of the blue spectrum is estimated to be ≤0.1 μM instead. 10442

dx.doi.org/10.1021/ac502955w | Anal. Chem. 2014, 86, 10437−10444

Analytical Chemistry

Article

proteins using a modified reported protocol.21 The resultant pale-yellow supernatant (Figure 5B) was then diluted 10-fold using ultrapure water and subjected to SERS detection. A distinct band at 685 cm−1, with intensity of (342 ± 37) counts, is observed in the SERS spectrum, which clearly indicates the presence of melamine within the treated liquid milk (Figure 5C). The indexing of the SERS band at 685 cm−1 to melamine is again confirmed from the SERS analysis of melamine-absence treated milk and also the SERS platform in the absence of milk, where both reveal rather flat and featureless spectra. We note the significant discrepancy of SERS intensity in treated milk from the melamine-in-water case, which was also previously observed due to the incomplete recovery of melamine as a consequence of adsorption on removed proteins via hydrogen bonding, and also competitive adsorption of low molecular weight species, such as proteins and vitamins, on the electromagnetic hot spots.5 Nevertheless, the ability to detect ≤100 fmol of melamine in the complex liquid milk matrix, despite the aforementioned issues, is a clear demonstration of the suitability of our SHP-OP SERS platform for ultratrace detection of food toxins in real samples.

University’s start-up grant. H.K.L. thanks the A*STAR Graduate Scholarship support from A*STAR, Singapore.



(1) Ko, H.; Singamaneni, S.; Tsukruk, V. V. Small 2008, 4 (10), 1576−1599. (2) Rycenga, M.; Cobley, C. M.; Zeng, J.; Li, W.; Moran, C. H.; Zhang, Q.; Qin, D.; Xia, Y. Chem. Rev. 2011, 111 (6), 3669−3712. (3) Romo-Herrera, J. M.; Alvarez-Puebla, R. A.; Liz-Marzan, L. M. Nanoscale 2011, 3 (4), 1304−1315. (4) Mulvihill, M.; Tao, A.; Benjauthrit, K.; Arnold, J.; Yang, P. Angew. Chem., Int. Ed. 2008, 47 (34), 6456−6460. (5) Kim, A.; Barcelo, S. J.; Williams, R. S.; Li, Z. Anal. Chem. 2012, 84 (21), 9303−9309. (6) Puoci, F.; Garreffa, C.; Iemma, F.; Muzzalupo, R.; Spizzirri, U. G.; Picci, N. Food Chem. 2005, 93 (2), 349−353. (7) Ingelfinger, J. R. N. Engl. J. Med. 2008, 359 (26), 2745−2748. (8) Kunov-Kruse, A. J.; Kristensen, S. B.; Liu, C.; Berg, R. W. J. Raman Spectrosc. 2011, 42 (6), 1470−1478. (9) Rebane, R.; Leito, I.; Yurchenko, S.; Herodes, K. J. Chromatogr. A 2010, 1217 (17), 2747−2757. (10) Peron, O.; Rinnert, E.; Toury, T.; Lamy de la Chapelle, M.; Compere, C. Analyst 2011, 136 (5), 1018−1022. (11) Cecchini, M. P.; Turek, V. A.; Paget, J.; Kornyshev, A. A.; Edel, J. B. Nat. Mater. 2013, 12 (2), 165−171. (12) Saito, Y.; Wang, J. J.; Smith, D. A.; Batchelder, D. N. Langmuir 2002, 18 (8), 2959−2961. (13) Kim, K.; Han, H. S.; Choi, I.; Lee, C.; Hong, S.; Suh, S.-H.; Lee, L. P.; Kang, T. Nat. Commun. 2013, 4, 1−9. (14) Kota, A. K.; Kwon, G.; Tuteja, A. NPG Asia Mater. 2014, 6, e109. (15) Kobaku, S. P. R.; Kota, A. K.; Lee, D. H.; Mabry, J. M.; Tuteja, A. Angew. Chem., Int. Ed. 2012, 51 (40), 10109−10113. (16) Golovin, K.; Lee, D. H.; Mabry, J. M.; Tuteja, A. Angew. Chem., Int. Ed. 2013, 52 (49), 13007−13011. (17) Wu, Y.; Hang, T.; Komadina, J.; Ling, H.; Li, M. Nanoscale 2014, 6 (16), 9720−9726. (18) Gentile, F.; Coluccio, M. L.; Zaccaria, R. P.; Francardi, M.; Cojoc, G.; Perozziello, G.; Raimondo, R.; Candeloro, P.; Di Fabrizio, E. Nanoscale 2014, 6 (14), 8208−8225. (19) Lu, L.-Q.; Zheng, Y.; Qu, W.-G.; Yu, H.-Q.; Xu, A.-W. J. Mater. Chem. 2012, 22 (39), 20986−20990. (20) De Angelis, F.; Gentile, F.; Mecarini, F.; Das, G.; Moretti, M.; 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), 682−687. (21) Giovannozzi, A. M.; Rolle, F.; Sega, M.; Abete, M. C.; Marchis, D.; Rossi, A. M. Food Chem. 2014, 159, 250−256. (22) Chen, M.; Phang, I. Y.; Lee, M. R.; Yang, J. K. W.; Ling, X. Y. Langmuir 2013, 29 (23), 7061−7069. (23) Liu, J.-W.; Wang, J.-L.; Huang, W.-R.; Yu, L.; Ren, X.-F.; Wen, W.-C.; Yu, S.-H. Sci. Rep. 2012, 2, 987. (24) Goh, M. S.; Lee, Y. H.; Pedireddy, S.; Phang, I. Y.; Tjiu, W. W.; Tan, J. M. R.; Ling, X. Y. Langmuir 2012, 28 (40), 14441−14449. (25) Tao, A. R.; Yang, P. J. Phys. Chem. B 2005, 109 (33), 15687− 15690. (26) Jasper, J. J. J. Phys. Chem. Ref. Data 1972, 1 (4), 841−1010. (27) Milling, A. J. Surface characterization methods: Principles, techniques, and applications; Taylor & Francis: New York, 1999. (28) Shirtcliffe, N. J.; McHale, G.; Atherton, S.; Newton, M. I. Adv. Colloid Interface Sci. 2010, 161 (1−2), 124−138. (29) Guo, Z.; Liu, W.; Su, B.-L. J. Colloid Interface Sci. 2011, 353 (2), 335−355. (30) Pan, S.; Kota, A. K.; Mabry, J. M.; Tuteja, A. J. Am. Chem. Soc. 2013, 135 (2), 578−581. (31) Chu, Z.; Seeger, S. Chem. Soc. Rev. 2014, 43 (8), 2784−2798. (32) Bhushan, B.; Nosonovsky, M. Philos. Trans. R. Soc., A 2010, 368 (1929), 4713−4728.



SUMMARY In summary, we have introduced the fabrication of a large-area, homogeneous superhydrophobic-oleophobic 3D Ag nanowire SERS platform for ultratrace and quantitative detection of toxins in both water and organic liquids. Analyte concentration factors of approximately 100-fold and 8-fold for water and toluene, respectively, have been achieved, which, together with large density of electromagnetic hot spots, enables the detection of melamine and Sudan I down to 0.1 fmol, ∼103fold below regulatory limits. Only 1 μL of analyte solution is required for detection, which is essential to reduce toxin exposure and can potentially serve as a miniaturize analytical platform. Our superhydrophobic-oleophobic SERS platform is also suitable for the sensing of food toxins in a real sample, demonstrated via the ultratrace detection of ≤100 fmol of melamine in liquid milk. Together with the simple and costeffective fabrication protocol, our superhydrophobic-oleophobic SERS platform offers an attractive, universal sensor crucial in the field of safety, where ultratrace toxins may exist or require extraction into water or organic solvents. In addition, our fabrication protocol is generic and can be extended to other high aspect ratio particles and materials to tailor the mechanical and optical properties of the detection platform.



ASSOCIATED CONTENT

S Supporting Information *

Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org.



REFERENCES

AUTHOR INFORMATION

Corresponding Author

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

X.L. and H.K.L. contributed equally to this work.

Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS X.Y.L. thanks the support from National Research Foundation, Singapore (NRF-NRFF2012-04), and Nanyang Technological 10443

dx.doi.org/10.1021/ac502955w | Anal. Chem. 2014, 86, 10437−10444

Analytical Chemistry

Article

(33) Fürstner, R.; Barthlott, W.; Neinhuis, C.; Walzel, P. Langmuir 2005, 21 (3), 956−961. (34) Xu, F.; Zhang, Y.; Sun, Y.; Shi, Y.; Wen, Z.; Li, Z. J. Phys. Chem. C 2011, 115 (20), 9977−9983. (35) Lee, H. K.; Lee, Y. H.; Zhang, Q.; Phang, I. Y.; Tan, J. M. R.; Cui, Y.; Ling, X. Y. ACS Appl. Mater. Interfaces 2013, 5 (21), 11409− 11418. (36) Huang, H.; Shende, C.; Sengupta, A.; Inscore, F.; Brouillette, C.; Smith, W.; Farquharson, S. J. Raman Spectrosc. 2012, 43 (6), 701−705. (37) Larkin, P. Infrared and raman spectroscopy; principles and spectral interpretation. Elsevier Science: 2011; p 109. (38) Calbiani, F.; Careri, M.; Elviri, L.; Mangia, A.; Pistarà, L.; Zagnoni, I. J. Chromatogr. A 2004, 1042 (1−2), 123−130. (39) Mahmoud, M. A.; Tabor, C. E.; El-Sayed, M. A. J. Phys. Chem. C 2009, 113 (14), 5493−5501.

10444

dx.doi.org/10.1021/ac502955w | Anal. Chem. 2014, 86, 10437−10444