Dielectrophoretic Nanoparticle Aggregation for On-Demand Surface

Subscriber access provided by Miami University Libraries

Article

Dielectrophoretic Nanoparticle Aggregation for on-demand SERS Analysis Reza Salemmilani, Brian D. Piorek, Rustin Yavar Mirsafavi, Augustus W. Fountain, Martin Moskovits, and Carl D. Meinhart Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.8b00510 • Publication Date (Web): 04 Jun 2018 Downloaded from http://pubs.acs.org on June 5, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Dielectrophoretic Nanoparticle Aggregation for on-demand SERS Analysis Reza Salemmilani,† Brian D. Piorek,† Rustin Y. Mirsafavi,‡ Augustus W. Fountain III,§ Martin Moskovits,⊥ and Carl D. Meinhart*† †

Department of Mechanical Engineering, University of California Santa Barbara, Santa Barbara, California 93106, United States ‡

Department of Biomolecular Science and Engineering, University of California Santa Barbara, Santa Barbara, California 93106, United States §

Research and Technology Directorate, Edgewood Chemical Biological Center, Aberdeen Proving Ground, Maryland 21010-5424, United States ⊥

Department of Chemistry and Biochemistry, University of California Santa Barbara, Santa Barbara, California 93106, United States

* address correspondence to [email protected]

1 ACS Paragon Plus Environment

Analytical Chemistry 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ABSTRACT

Rapid chemical identification of drugs of abuse in biological fluids such as saliva is of growing interest in healthcare and law-enforcement. Accordingly, a label-free detection platform that accepts biological fluid samples is of great practical value. We report a microfluidics-based dielectrophoresis-SERS device, which is capable of detecting physiologically relevant concentrations of methamphetamine in saliva in under 2 minutes. In this device, iodide-modified silver nanoparticles are trapped and released on-demand using electrodes integrated in a microfluidic channel. Principal component analysis (PCA) is used to reliably distinguish methamphetamine-positive samples from the negative control samples. Passivation of the electrodes and flow channels minimizes microchannel fouling by nanoparticles, which allows the device to be cleared and reused multiple times. ………………………………………………………………………………………………………………

The widespread use of illicit drugs is a major contributor to the costs of law enforcement, healthcare, and lost workforce productivity.1 A fast and low-cost clinical and forensic screening assay would be a valuable tool in healthcare and law enforcement. Ideally, such an assay should be sensitive, specific, rapid, inexpensive, noninvasive, field-deployable, and capable of accommodating biological fluids with minimal processing. Current screening techniques often rely on colorimetry for analysis, which typically only identifies classes of drugs.2 Moreover, as a single-parameter visual test, colorimetry is limited by subjective color perception. Consequently, a positive colorimetric test result often requires validation by more sensitive and specific laboratory techniques such as Gas Chromatography- Mass Spectroscopy (GC-MS), High Performance Liquid Chromatography- Mass Spectroscopy (HPLC-MS), etc.


Page 2 of 25

Page 3 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Surface Enhanced Raman Spectroscopy (SERS) is a vibrational spectroscopy technique that employs surface plasmon resonance to enhance the Raman scattering from analyte molecules. Metallic nanostructures that are capable of sustaining localized surface plasmon resonances (LSPR) are normally used as substrates for SERS.3 Examples of such substrates are nanowires, nanoshells, and nanoparticle aggregates.4-7 Enhancement factors can be as high as 10 , potentially capable of single molecule detection.8,9 A clear advantage of SERS over colorimetric techniques is its ability to yield a molecule-specific vibrational spectrum, eliminating in many cases the need for analyte-specific reagents and multiple test kits. Due to the naturally small size of the nanoscale substrates, SERS-based chemical detection platforms are amenable to miniaturization and can be used for label-free multi-analyte detection. Chip-based SERS devices have been used to detect trace concentrations of illicit drugs, explosives, cancer biomarkers, and other classes of molecules.10-13 Various biological fluids such as saliva, blood plasma, urine, and sweat are typically used as matrices for drug testing.14-16 Urine is the most common medium; however, saliva has several advantages. Unlike urine, saliva collection is considered noninvasive and potentially applicable in road-side drug testing. Its collection minimizes sample adulteration risk by direct supervision at the time of sample collection. Moreover, parent drugs are often detectable in saliva, while urine testing mainly quantifies metabolites of the target substance. Most importantly, detection and quantification of parent drugs in saliva can be good indicators of the level of impairment.17 Here we report a microfluidic chip-based SERS platform that employs dielectrophoresis (DEP) for highly localized trapping and release of SERS-active nanoparticles, on demand. The platform creates a pristine SERS substrate prior to each measurement, enabling rapid detection of physiologically relevant concentrations of methamphetamine. Since the detection mechanism is 3 ACS Paragon Plus Environment


based on DEP actuation to form the SERS-hot substrate for detection, the microfluidic chip can be reliably reused over multiple cycles in which the device is cleared and a new sample is introduced. DEP refers to migration of polarizable particles in electric field gradients. DEP has been extensively used in microfluidic devices for manipulation, sorting, and trapping of particles and cells. DEP theory and its applications in microfluidic technology have been extensively reviewed.18-20 Microfluidic-based DEP has been combined with SERS for characterizing WO3 and polystyrene nanoparticles.21 Chrimes et al. have reported controlled aggregation of silver nanoparticles in a microfluidic platform and subsequent detection of 1ppm dipicolinic acid in DI water using SERS.22 The same authors also used microfluidic DEP and SERS to capture and probe silver nanoparticle-coated yeast cells.23 Cherukulappurah et al. have reported dielectrophoresis-enabled assembly of metallic nanoparticles using microfabricated interdigitated electrodes for detecting benzenethiol using SERS. SERS-active nanoparticle aggregates were produced by DEP which together with analyte were drop-cast onto the electrodes. This technique was further extended for detecting adenine.24 Microfluidic-based DEP has also been used to trap bacteria with subsequent chemical analysis and identification using SERS.25-27 More recently an insulator-dielectrophoresis- based microfluidic device has been reported for detection of crystal violet using SERS.28 In the current study, we report the use of a DEP-SERS platform as a reusable probe for detecting drugs of abuse in saliva. Our device comprises a polydimethylsiloxane (PDMS) microchannel sandwiched between an electrode substrate and a top cover (Figure 1). Fluid flow is established with negative pressure by applying vacuum to the outlet of the channel. An AC electric potential is applied to the electrode contact pads to activate the DEP nanoparticle trap. Silver nanoparticles, premixed with the 4 ACS Paragon Plus Environment

Page 4 of 25

Page 5 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


analyte, agglomerate in the trap region to form SERS ‘hotspots’. Spectra are then acquired by scanning the trap regions using a Raman microscope.

Experimental Section Numerical Simulation. Migration of the nanoparticles in the device was modeled numerically using COMSOL Multiphysics V5.2 (Comsol, Stockholm, Sweden) to determine appropriate electrode geometries for effective nanoparticle trapping at low electrical potentials. The electric field in the device was computed by solving Laplace’s equation: ∇2 φ=0

(1)

where is the electric potential. Stokes’ law was used for approximating the drag force on the nanoparticles: Fdrag =6πrηv (2) where Fdrag is the drag force acting on the particle, η is the dynamic viscosity of the fluid, r is the radius of the nanoparticle, and v is flow velocity relative to the nanoparticle. Dielectrophoresis was modelled by the following equation: =2πr3 ϵm Re

ϵ*p -ϵ*m ∇|E|2 (3) ϵ*p +2ϵ*m

Where < FDEP > is the time-averaged DEP force on the particle, r is the radius of the particle, ϵm is the real part of permittivity of the medium, E is the electric field, and ϵ*p and ϵ*m are the complex permittivities of the particle and the medium, respectively.



Page 6 of 25

The migration velocity of the nanoparticles relative to bulk flow was calculated using force balance: + Fdrag = 0

(4)

Device Fabrication. The electrode geometry was lithographically patterned on 500µm-thick fused silica wafer (University Wafers). 100nm-thick gold electrodes were then formed by lift-off process, using a 15nm-thick titanium adhesion layer. Since tin-based solders readily dissolve gold, a 100nm-thick layer of nickel, covered by a 50nm-thick gold layer was deposited on top of the gold contact pads. Upon contact with hot solder, the top gold layer dissolves allowing the exposed nickel layer to form a strong bond with the tin-based solder. Microfluidic channels are made in PDMS (Sylgard 184, Dow Corning) using soft-lithography using an SU-8 mold. The microfluidic channels were 100 µm wide and 20 µm deep. Inlet and outlet vias were produced in the PDMS using a biopsy punch. Corresponding inlet and outlet vias were drilled on a glass microscope slide. The electrode substrate, the PDMS, and the microscope slide were ozonated and the PDMS was sandwiched between the electrode substrate and microscope slide and aligned under a microscope using methanol as lubricant to allow electrode-to-microchannel fine alignment. Pipette tips were attached to the inlet and outlet vias using epoxy adhesive to form chip-to-world fluidic connections. Aluminum wires were soldered to contact pads on the electrode substrate to form the electrical connection between the chip and the electrical function generator. To ensure device reusability, all wetted surfaces of the chip were passivated. A 5mM ethanolic solution of 11-Mercaptoundecyltetraethylene glycol (Sigma Aldrich) was incubated in the microfluidic channel to passivate the gold electrodes. Wetted PDMS and fused silica surfaces


Page 7 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


were fluorinated in a 1mM solution of (Heptadecafluoro-1,1,2,2-tetrahydrodecyl) trichlorosilane (Gelest) in FC-40 (Sigma Aldrich). A vacuum pump was used to establish flow of the reagent in the channel. The sub-atmospheric pressure was adjusted such that the mean flow rate in the microchannel was approximately ~ 5 µl/min. After 30 minutes of reagent flow, the microfluidic channels were flushed with ethanol followed by water, then baked at 50 T overnight. Sample Preparation. Saliva was collected from a healthy adult volunteer. Food and beverages were avoided for 30 minutes prior to sample collection. The volunteer did not use prescription or illicit drugs prior to sample collection. Undiluted saliva was spiked with (+)-methamphetamine hydrochloride (Sigma Aldrich) to a concentration of 500 nM. A separate sample was prepared using the same saliva batch and spiked with heroin hydrochloride (0.1mg/ml in methanol-Sigma Aldrich) to a concentration of 50 µM to be used as the negative control. Saliva samples were then filtered using a 0.2 µm syringe filter (Millex®-FG) to remove cells and particles that could cause clogging of microfluidic channels. A 5 mM solution of (+)-methamphetamine was prepared in MilliQ DI-water to be used as the positive control. Citrate-capped 40-nm silver nanoparticles (Biopure, nanoComposix) were incubated in a 1 mM NaI solution in order to remove citrate from the surface of the nanoparticles to yield nanoparticles nearly free of background SERS spectra. The nanoparticles were then washed 3-times with MilliQ DI water and added to saliva samples at a concentration of 0.01 mg/ml. Experimental Procedure. Electrical connection was established between the function generator (Agilent, 33220A), AC voltage amplifier (FLE, F20A) and the chip. 25 µl of sample was loaded into the microfluidic device inlet reservoir, and the flow was mobilized, and adjusted to approximately 5 µl/min. The chip was loaded onto a LabRam Aramis Raman microscope (Horiba, Kyoto, Japan), equipped with a 633-nm laser as the excitation source. The laser was 7 ACS Paragon Plus Environment


focused on the DEP trap region using a 50X objective, exposing the sample to 0.78 mW of laser power. Microfluidic flow was subsequently stopped and an AC potential with a frequency of 2.5 MHz and an amplitude of 16 Vpp was applied to the electrode contact pads. SERS spectra were collected at 1-second integration intervals for a total of 150 seconds. After data acquisition, flow was re-established at a higher rate of approximately 20 µl/min. An AC potential of 12 MHz at 80 Vpp was then applied to the electrodes. The combination of high flowrate and secondary electrokinetally-induced flow at the surface of the electrodes cleared the agglomerated nanoparticles in the trap region, thereby preparing the device for a new detection run. The detection-clearance cycle was repeated three times with 500 nM methamphetamine-positive saliva samples. 50 µM heroin-positive saliva samples were used as negative controls for the experiments. Chemometric Analysis. Eigenvector PLS toolbox V7.9.5 (Eigenvector Research, Inc., Manson, WA, USA) was utilized to construct a principal component analysis (PCA) model for the analysis of the spectra. SERS spectra from 5 mM methamphetamine reference stock and pure saliva were used to construct the calibration model. The baseline was subtracted from each of the acquired spectra using a weighted least-squares filter. Spectra were then normalized by the area under the spectrum in the 500-1800 cm-1 range and finally mean-centered. Venetian-blinds with 10 splits were used for cross-validation. The portions of the SERS spectra in the 500-1800 cm-1 wavenumbers range were used for the analysis. Methamphetamine-positive saliva samples and heroin-positive saliva samples were then tested against the calibration model.


Page 8 of 25

Page 9 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Results and Discussion Preparation of Background-free Silver Nanoparticles.

Uncapped silver nanoparticles are not thermodynamically stable and tend to aggregate shortly after synthesis.29 Various capping agents such as citrate,30 polyvinylpyrrolidone (PVP),31 peptides,32 etc. are used to stabilize silver nanoparticles by electrostatic or electrosteric means. Capping agents are used to stabilize silver nanoparticles, thereby improving their shelf-life, but with the undesirable consequence of having the SERS spectrum of the capping agents interfere with that of the analyte, especially at low analyte concentrations. More importantly, the capping agent occupies surface adsorption sites that might otherwise be available to those analyte molecules unable to displace the capping agent. Methamphetamine is an analyte with low affinity for silver surfaces. As a result, we first removed as much of the citrate off the silver nanoparticles as possible by incubating the off-the-shelf citrate-coated 40-nm silver nanoparticles in a 1 mM sodium iodide solution. Xu et al. have previously used silver nanoparticles, modified with potassium iodide, for SERS studies of proteins and concluded that modification of silver nanoparticles with iodide removes surface impurities from particles while retaining stability and SERS activity.33 Here, the removal of the citrate from the surface of the nanoparticles was evidenced by the elimination of Raman bands associated with citrate. The iodide-modified silver nanoparticles retained their SERS activity and remained relatively stable for several days. Figure 2 shows SERS spectra collected from the silver nanoparticles before and after surface modification with sodium iodide, and the absorbance spectrum suggests minimal aggregation after surface modification of the silver nanoparticles with sodium iodide.



Dielectrophoretic Trapping. Each device has four trap regions. A gap between the electrodes has a width of 7 µm (Figure 1). During device operation, the laser beam is focused by the Raman microscope on one of the four traps. It might then be moved to another trap after the spectra are acquired. DEP-induced agglomeration of the silver nanoparticles occurs primarily at the trap sites. The direction of the DEP force is determined by the sign of the Clausius-Mossotti factor (depicted in brackets in Eq. 3), which accounts for permittivity contrast between the particle and medium. During the trapping step, the AC potential frequency does not exceed 3 MHz. At these low frequencies, silver has very large dielectric constants,34 allowing the Clausius-Mossoti factor to be approximated by unity. This results in positive DEP, where nanoparticles migrate towards the electrode edges. Secondary electric field gradients, which are induced by the nanoparticles attached to electrode edges, further enhance the positive DEP effect and result in additional agglomeration of nanoparticles at the trap site, a phenomenon known as pearl-chaining.35 In the current device, approximately 2 minutes are required to agglomerate the silver nanoparticles by DEP and acquire a SERS spectrum of nanomolar concentrations of methamphetamine in saliva with a signal-to-noise ratio of at least ten. A larger electric potential results in faster trapping and detection but can also cause water electrolysis, electrode damage, nanoparticle melting, and subsequent short-circuiting of the electrodes. Hence, the lowest workable electric potential was used. Appropriate design and dimensions for such electrodes were determined using numerical simulation. Figure 3 shows the cross-sectional view of the microfluidic channel at mid-plane of one of the trapping zones. Contours of constant E2 are


Page 10 of 25

Page 11 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


shown as black lines. Red lines, showing the migration path of the nanoparticles towards the trap, are orthogonal to the contours of constant E2 as expected from equation (3). The flow is stopped shortly after the sample is loaded in the microfluidic device and remains stopped during signal acquisition. Migration of the nanoparticles is solely due to diffusion and electrokinetics during each acquisition cycle. Since the location of the DEP-aggregated SERSactive substrate is spatially determined by the design of the electrodes, the optimal spatial location for collection of SERS spectra is also determined by these design parameters. As a result, SERS spectra were successfully collected from trapping zones that were determined by computation without the need to first raster scan the device to find the maximal signal intensity. Sample Preparation and SERS Measurements. Saliva samples were premixed with freshly prepared iodide-modified silver nanoparticles and loaded into the device. In a typical detection run, sample flow is established, the device is placed on the microscope platform of the Raman spectrometer, and the laser focused with a 50X objective lens onto one of the four trap regions. The fluid flow is stopped and a 16Vpp, 2.5 MHz AC-potential is applied to the electrodes. The SERS signal intensity increases with time, due to the agglomeration of the nanoparticles by DEP, reaching a maximum intensity after approximately 2 minutes. Additional nanoparticles are likely accumulated beyond this time but do not significantly contribute to the signal collected. Although a higher concentration of nanoparticles agglomerate faster and improve the timeresponse of the device, they tended to foul the microchannels after long detection runs. For this reason, the relatively low nanoparticle concentration of 0.01 mg/ml was used in all of the experiments.



Experiments were carried out with: (i) pure saliva, (ii) 5 mM methamphetamine in DI water as reference samples, (iii) 500 nM methamphetamine in saliva as test samples, and (iv) 50 µM heroin in saliva as negative controls. A methamphetamine cut-off limit of 500 nM is lower than concentrations routinely observed in oral fluids of impaired individuals.15 Figure 4 shows a comparison of time-averaged raw spectra for the aforementioned samples. Saliva is a complex fluid with inter-individual and intra-individual variability in composition. The many SERS bands observed here likely originate from small protein molecules present in saliva.36 The 1004 cm-1, 1030 cm-1 and 1600 cm-1 known methamphetamine Raman vibrational modes (phenyl ring modes),37,38 effectively identify methamphetamine-positive saliva samples from pure saliva even without chemometric analysis. Figure 5 summarizes the time-evolution of SERS spectra for the Methamphetamine-positive saliva sample. Within a few seconds of application of the AC potential, SERS spectral features corresponding to saliva and methamphetamine appear as the nanoparticles agglomerate at the trap. Once the AC potential is applied, the signal intensity grows monotonically until it reaches a maximum intensity in approximately 120 seconds. Chemometric Analysis. A principal component analysis model with two principal components was found to capture 69.57 % of the overall variance of the spectra. The rest of the variance is attributed to noise and chemical intra-variability of the highly complex saliva matrix. A comparison between principal component loadings and SERS spectra of methamphetamine and saliva reveals that PC1, accounting for 66.80% of the overall variance of the data, is mainly comprised of spectral features that separate saliva from methamphetamine. PC2, which captures 2.77% of overall


Page 12 of 25

Page 13 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


variance, describes some of the intra-sample variations of the raw saliva. After construction and cross-validation of the PCA calibration model, heroin-positive and methamphetamine-positive saliva samples were tested against the model. Figure 6 summarizes the results of the principal component analysis of the test datasets together with the calibration datasets. As shown, the plot of PC1 vs PC2 clearly discriminates methamphetamine-positive samples from pure saliva wherein the associated ellipses correspond to 95% confidence intervals. Heroin-positive saliva samples, which play the role of negative-control for the PCA model, are also separated well from the other samples. It is worth noting that since the model is mean-centered, a sample with a PC1 score of zero would have equal contributions from spectral features of methamphetamine and saliva; whereas a positive PC1 score indicates a greater methamphetamine contribution over that of saliva. Device Passivation and Clearance. It is highly desirable to have a microfluidic chemical detection platform that resists fouling by nanoparticles and chemicals, as this would make the system inherently more resistant to crosscontamination and would make device reusability a viable option. Accordingly, DEP-induced aggregation of SERS-hot nanoparticles has a clear advantage over the use of aggregationinducing agents such as salts. Aggregation of the nanoparticles with DEP is spatially restricted to the pre-arranged trap sites, whereas chemically-induced aggregation using salts produces SERSactive aggregates at multiple sites in the microchannel making it difficult to reuse the channel for subsequent analyses. The localized and controllable nanoparticle trapping by DEP enabled us to restrict nanoparticle fouling to trap zones, making it also easier to develop protocols for flushing the aggregates after use.



To further reduce microchannel fouling by nanoparticles, and to ensure reusability of the device, the gold electrodes were functionalized with polyethylene glycol (PEG), and the wetted PDMS and glass surfaces were passivated by fluoroalkanes. These long hydrocarbon and fluorocarbon chains on the wetted surfaces tend to minimize adhesion of nanoparticles to the flow surfaces. The passivated devices were significantly superior to the unmodified devices in terms of overall signal clearance and reusability. The device is cleared as follows: flowrate is increased to 20 µl/min and a large AC potential of 80 Vpp at 12 MHz is applied to the electrodes. This produces a temperature gradient near the electrode edges by Joule heating the liquid surrounding the electrodes, resulting in vortex motion likely due to electrothermal flows.39 The frequency of the applied potential is chosen to be high, in order to minimize electrolysis-induced electrode damage.40 The existence of vorticity-induced mixing was verified by observing the motion of 20-nm fluorescent polystyrene nanoparticles. Electrokinetic mixing combined with high rate of the Poiseuille flow caused by the low pressure at the channel’s outlet, clears the nanoparticle clusters from the microchannels. Figure 7 shows the trap region after a clearance cycle. Minimal fouling and cycle-to-cycle chemical crosscontamination was observed after three detection-clearance runs.


Page 14 of 25

Page 15 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Conclusion An integrated microfluidic-dielectrophoresis and SERS platform is described that is capable of on-demand, highly-localized trapping of SERS-active silver nanoparticles. This assay is able to detect physiologically relevant concentrations of methamphetamine in saliva. Off-the-shelf citrate-coated silver nanoparticles are modified with sodium iodide to remove citrate, hence eliminating spectroscopic interference caused by presence of the capping agent. A PCA model was constructed and used to identify methamphetamine-positive samples. The specificity of the detection system was benchmarked by using heroin-positive saliva samples as negative controls. The PCA model successfully distinguishes methamphetamine-positive samples from negative controls with good accuracy and repeatability. Reusability of the chip was demonstrated by three consecutive detection-clearance cycles with minimal residual signal from the analyte used in previous detection cycles. To achieve this high degree of robustness and to minimize device fouling by the nanoparticles, the wetted surfaces were passivated with 11mercaptoundecyltetraethylene glycol (gold electrodes) and fluorinated polymers (glass and PDMS surfaces). The advantages of this novel DEP-SERS assay include compatibility with complex biological samples such as saliva, label-free detection of the drug of interest at physiologically relevant concentrations in under two minutes, reagent and aggregation-inducing agent free operation, and demonstrated reusability.

Acknowledgements Authors would like to thank Dr. Tracy Chuong for her contributions in discussions and guidance. The fabrication was partly undertaken at the UCSB nanofabrication facility, part of the National Science Foundation-funded national nanofabrication infrastructure network. Funding was provided by the U.S. Army via the Forensics Applied Research Program (PE 0602622A Project 552) at the Edgewood Chemical Biological Center, and by the Institute for Collaborative Biotechnologies through grant W911NF-09-0001 from the U.S. Army Research Office. The content of the information does not 15 ACS Paragon Plus Environment


necessarily reflect the position or the policy of the Government, and no official endorsement should be inferred.

Figure 1. Depiction of the architecture of the DEP-SERS chip. (a): PDMS microchannel is sandwiched between electrode substrate and glass cover. Yellow pipette tips form fluidic reservoirs/connections. Aluminum wires, soldered to contact pads form the electrical connection between the chip and the AC signal generator. Chip is loaded onto the Raman microscope, upside down, to prevent laser light from transmitting through PDMS layer, to effectively eliminate PDMS Raman interference. (b) Each device consists of four trap zones. Traps are aligned to the center of the channels for maximum flow shear during clearing of the microfluidic channels for subsequent detection cycles. (c): Close-up view of a single trap using agglomeration of nanoparticles. (d) Typical SERS spectra acquired using the DEP-SERS chip.


Page 16 of 25

Page 17 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 2. Comparison of background SERS spectra of off-the-shelf 40-nm citrate capped silver nanoparticles before (red) and after (blue) surface modification with sodium iodide. Surface modification with sodium iodide effectively removes citrate from the surface of the nanoparticles, resulting in elimination of background signal in the 400 cm-1 – 1800 cm-1 wavenumber range. Nanoparticles retain their SERS activity and remain stable for days. Inset shows slight red-shift, without significant aggregation, after surface modification of nanoparticles with iodide consistent with the literature.33



Figure 3: Numerical simulation of silver nanopartical electromigration. Cross-sectional view of the microfluidic channel at mid-plane of one of the trap zones. Black lines correspond to contours of constant | | . Red lines show pathlines of the nanoparticles towards the trap. Highly polarizable nanoparticles migrate from the regions of low electric field intensity to regions of high electric field intensity.


Page 18 of 25

Page 19 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 4. Raw spectra acquired using the DEP-SERS chip. Bands at 1004 cm-1 , 1030 cm-1 and 1600 cm-1 discriminate methamphetamine-positive saliva samples from pure saliva samples. (a): 500 nM methamphetamine in saliva, averaged over 5s (b): Methamphetamine reference (5 mM in DI water), averaged over 5s (c): Saliva reference, averaged over 20s (d): 50 μM heroin in saliva, averaged over 3s



Figure 5. Evolution of spectra for 500 nM methamphetamine-positive saliva sample. Linear baseline correction is applied to the spectra. As nanoparticles agglomerate in the trap, the signal intensity increases until it plateaus at around 120s.


Page 20 of 25

Page 21 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 6. PCA scores plot for the calibration and test datasets. Calibration model was constructed using pure saliva and methamphetamine reference datasets. Methamphetamine-positive saliva samples and heroin-positive saliva samples were tested against the model. A 2-PC model provides good separation of the samples. The ellipses correspond to 95% confidence intervals.



Figure 7. Nanoparticle aggregation and clearance characteristics of the DEP-SERS chip. After application of AC – potential, silver nanoparticles start migrating towards the electrodes. Images (a), (b) and (c) illustrate agglomeration of nanoparticles at the trap site at 10s, 60s and 120s, respectively. After acquisition of spectra, the trap is cleared, as shown in (d). Inset (e) shows the value of signal clearance ratio ( for 3 consecutive detection-clearance cycles of three separate DEP chips. is defined as: I ( I1600,ON -I1600,OFF )I1600,ON , where I1600,ON is the intensity of 1600 cm-1 band approximately 120s into the detection cycle and I1600,OFF is the intensity of the 1600 cm-1 band after the nanoparticle trap has been cleared. Entire trap region is raster scanned after every clearance cycle and the highest residual signal is chosen for assessment of clearance performance. Intensity of the peak at 1600 cm-1 is chosen as


Page 22 of 25

Page 23 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


a signal clearance metric because it is one of the strong bands associated with SERS of methamphetamine in saliva (Figure 5). Supporting Information. Two consecutive trapping/clearance cycles of silver nanoparticles in the DEP chip.

References (1) NIDA. Trends and Statistics 2017, https://www.drugabuse.gov/related-topics/trends-statistics (accessed May 31 2018). (2) National Institue of Standards and Technology (NIST). NIJ Standard-0604.01 2000. (3) Moskovits, M. Persistent misconceptions regarding SERS. Phys Chem Chem Phys 2013, 15, 53015311. (4) Wustholz, K. L.; Henry, A. I.; McMahon, J. M.; Freeman, R. G.; Valley, N.; Piotti, M. E.; Natan, M. J.; Schatz, G. C.; Van Duyne, R. P. Structure-activity relationships in gold nanoparticle dimers and trimers for surface-enhanced Raman spectroscopy. J Am Chem Soc 2010, 132, 10903-10910. (5) Tao, A.; Kim, F.; Hess, C.; Goldberger, J.; He, R.; Sun, Y.; Xia, Y.; Yang, P. Langmuir−Blodgett Silver Nanowire Monolayers for Molecular Sensing Using Surface-Enhanced Raman Spectroscopy. Nano Letters 2003, 3, 1229-1233. (6) Jackson, J. B.; Halas, N. J. Surface-enhanced Raman scattering on tunable plasmonic nanoparticle substrates. Proc Natl Acad Sci U S A 2004, 101, 17930-17935. (7) Schwartzberg, A. M.; Grant, C. D.; Wolcott, A.; Talley, C. E.; Huser, T. R.; Bogomolni, R.; Zhang, J. Z. Unique Gold Nanoparticle Aggregates as a Highly Active Surface-Enhanced Raman Scattering Substrate. The Journal of Physical Chemistry B 2004, 108, 19191-19197. (8) Le Ru, E. C.; Blackie, E.; Meyer, M.; Etchegoin, P. G. Surface Enhanced Raman Scattering Enhancement Factors: A Comprehensive Study. The Journal of Physical Chemistry C 2007, 111, 1379413803. (9) Kneipp, K.; Wang, Y.; Kneipp, H.; Perelman, L. T.; Itzkan, I.; Dasari, R. R.; Feld, M. S. Single Molecule Detection Using Surface-Enhanced Raman Scattering (SERS). Physical Review Letters 1997, 78, 1667-1670. (10) Andreou, C.; Hoonejani, M. R.; Barmi, M. R.; Moskovits, M.; Meinhart, C. D. Rapid detection of drugs of abuse in saliva using surface enhanced Raman spectroscopy and microfluidics. ACS Nano 2013, 7, 7157-7164. (11) Piorek, B. D.; Lee, S. J.; Moskovits, M.; Meinhart, C. D. Free-surface microfluidics/surfaceenhanced Raman spectroscopy for real-time trace vapor detection of explosives. Anal Chem 2012, 84, 9700-9705. (12) Pallaoro, A.; Hoonejani, M. R.; Braun, G. B.; Meinhart, C. D.; Moskovits, M. Rapid identification by surface-enhanced Raman spectroscopy of cancer cells at low concentrations flowing in a microfluidic channel. ACS Nano 2015, 9, 4328-4336. (13) Wang, C.; Yu, C. Analytical characterization using surface-enhanced Raman scattering (SERS) and microfluidic sampling. Nanotechnology 2015, 26, 092001. (14) Drummer, O. H. Drug Testing in Oral Fluid. Clinical Biochemist Reviews 2006, 27, 147-159. (15) Huestis, M. A.; Cone, E. J. Methamphetamine disposition in oral fluid, plasma, and urine. Ann N Y Acad Sci 2007, 1098, 104-121. (16) Allen, K. R. Screening for drugs of abuse: which matrix, oral fluid or urine? Ann Clin Biochem 2011, 48, 531-541.



(17) Engblo, C.; Gunnar, T.; Rantanen, A.; Lillsunde, P. Driving Under the Influence of DrugsAmphetamine Concentrations in Oral Fluid and Whole Blood Samples. Journal of Analytical Toxicology 2007, 31, 276-280. (18) Pethig, R. Review article-dielectrophoresis: status of the theory, technology, and applications. Biomicrofluidics 2010, 4. (19) Cetin, B.; Li, D. Dielectrophoresis in microfluidics technology. Electrophoresis 2011, 32, 24102427. (20) Li, M.; Li, W. H.; Zhang, J.; Alici, G.; Wen, W. A review of microfabrication techniques and dielectrophoretic microdevices for particle manipulation and separation. Journal of Physics D: Applied Physics 2014, 47, 063001. (21) Chrimes, A. F.; Kayani, A. A.; Khoshmanesh, K.; Stoddart, P. R.; Mulvaney, P.; Mitchell, A.; Kalantar-Zadeh, K. Dielectrophoresis-Raman spectroscopy system for analysing suspended nanoparticles. Lab Chip 2011, 11, 921-928. (22) Chrimes, A. F.; Khoshmanesh, K.; Stoddart, P. R.; Kayani, A. A.; Mitchell, A.; Daima, H.; Bansal, V.; Kalantar-zadeh, K. Active control of silver nanoparticles spacing using dielectrophoresis for surfaceenhanced Raman scattering. Anal Chem 2012, 84, 4029-4035. (23) Chrimes, A. F.; Khoshmanesh, K.; Tang, S. Y.; Wood, B. R.; Stoddart, P. R.; Collins, S. S.; Mitchell, A.; Kalantar-zadeh, K. In situ SERS probing of nano-silver coated individual yeast cells. Biosens Bioelectron 2013, 49, 536-541. (24) Cherukulappurath, S.; Lee, S. H.; Campos, A.; Haynes, C. L.; Oh, S.-H. Rapid and Sensitive in Situ SERS Detection Using Dielectrophoresis. Chemistry of Materials 2014, 26, 2445-2452. (25) Cheng, I. F.; Chen, T. Y.; Lu, R. J.; Wu, H. W. Rapid identification of bacteria utilizing amplified dielectrophoretic force-assisted nanoparticle-induced surface-enhanced Raman spectroscopy. Nanoscale Res Lett 2014, 9, 324. (26) Cheng, I. F.; Lin, C. C.; Lin, D. Y.; Chang, H. C. A dielectrophoretic chip with a roughened metal surface for on-chip surface-enhanced Raman scattering analysis of bacteria. Biomicrofluidics 2010, 4. (27) Lin, H. Y.; Huang, C. H.; Hsieh, W. H.; Liu, L. H.; Lin, Y. C.; Chu, C. C.; Wang, S. T.; Kuo, I. T.; Chau, L. K.; Yang, C. Y. On-line SERS detection of single bacterium using novel SERS nanoprobes and a microfluidic dielectrophoresis device. Small 2014, 10, 4700-4710. (28) Almeida, G. B.; Poppi, R. J.; da Silva, J. A. Trapping of Au nanoparticles in a microfluidic device using dielectrophoresis for surface enhanced Raman spectroscopy. Analyst 2017, 142, 375-379. (29) Sato, T.; Ruch, R. Stabilization of colloidal dispersions by polymer adsorption. Surfactant Science Series 1980, 9. (30) Bastús, N. G.; Merkoçi, F.; Piella, J.; Puntes, V. Synthesis of Highly Monodisperse Citrate-Stabilized Silver Nanoparticles of up to 200 nm: Kinetic Control and Catalytic Properties. Chemistry of Materials 2014, 26, 2836-2846. (31) Danilczuk, M.; Lund, A.; Sadlo, J.; Yamada, H.; Michalik, J. Conduction electron spin resonance of small silver particles. Spectrochim Acta A Mol Biomol Spectrosc 2006, 63, 189-191. (32) Graf, P.; Mantion, A.; Haase, A.; Thunemann, A. F.; Masic, A.; Meier, W.; Luch, A.; Taubert, A. Silicification of peptide-coated silver nanoparticles--A Biomimetic soft chemistry approach toward chiral hybrid core-shell materials. ACS Nano 2011, 5, 820-833. (33) Xu, L. J.; Zong, C.; Zheng, X. S.; Hu, P.; Feng, J. M.; Ren, B. Label-free detection of native proteins by surface-enhanced Raman spectroscopy using iodide-modified nanoparticles. Anal Chem 2014, 86, 2238-2245. (34) Lorrain, P.; Corson, D. Electromagnetic Fields and Waves. W.H. Freeman&Co Ltd 1970. (35) Hermanson, K. D.; Lumsdon, S. O.; Williams, J. P.; Kaler, E. W.; Velev, O. D. Dielectrophoretic assembly of electrically functional microwires from nanoparticle suspensions. Science 2001, 294, 10821086. (36) Feng, S.; Huang, S.; Lin, D.; Chen, G.; Xu, Y.; Li, Y.; Huang, Z.; Pan, J.; Chen, R.; Zeng, H. Surface-enhanced Raman spectroscopy of saliva proteins for the noninvasive differentiation of benign and malignant breast tumors. Int J Nanomedicine 2015, 10, 537-547. 24 ACS Paragon Plus Environment

Page 24 of 25

Page 25 of 25 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


(37) Tsuchihashi, H.; Katagi, M.; Nishikawa, M.; Tatsuno, M.; Nishioka, H.; Nara, A.; Nishio, E.; Petty, C. Determination of Methamphetamine and Its Related Compounds Using Fourier Transform Raman Spectroscopy. Applied Spectroscopy 1997, 51, 1796-1799. (38) Farquharson, S.; Shende, C.; Sengupta, A.; Huang, H.; Inscore, F.; Rapid Detection and Identification of Overdose Drugs in Saliva by Surface-Enhanced Raman Scattering Using Fused Gold Colloids. Pharmaceutics 2011, 3, 425-439. (39) Lian, M.; Islam, N.; Wu, J. AC electrothermal manipulation of conductive fluids and particles for lab-chip applications. IET Nanobiotechnol 2007, 1, 36-43. (40) Morgan, H.; Green, N. G. AC Electrokinetics: colloids and nanoparticles. Research Studies Pr Ltd, 1st ed 2002.

……………………………………………………………………………………………………………...... TOC graphic:


Dielectrophoretic Nanoparticle Aggregation for On-Demand Surface

Recommend Documents