Ultrasensitive Plasmonic Platform for Label-Free Detection of

Subscriber access provided by La Trobe University Library

Article

An Ultrasensitive Plasmonic Platform For The LabelFree Detection Of Membrane-Associated Species Ian R. Bruzas, Sarah A. Unser, Sadegh Yazdi, Emilie Ringe, and Laura B. Sagle Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.6b00801 • Publication Date (Web): 20 Jul 2016 Downloaded from http://pubs.acs.org on July 24, 2016

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 free 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 accessible to all readers and 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.

Analytical Chemistry 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

An Ultrasensitive Plasmonic Platform For The Label-Free Detection Of MembraneAssociated Species Ian Bruzas*, Sarah Unser*, Sadegh Yazdiǂ, Emilie Ringeǂ, and Laura Sagle* *Department of Chemistry, College of Arts and Sciences, University of Cincinnati, 301 West Clifton Court, Cincinnati OH 45221-0172 ǂDepartment of Material Science and NanoEngineering, Rice University, 6100 Main Street, MS-325, Houston TX 77005

*Corresponding author Tel: +1 513 556 1034; Fax: +1 513 556 9239. E-mail: [email protected]

1 ACS Paragon Plus Environment


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

Page 2 of 25

ABSTRACT Lipid membranes and membrane proteins are important biosensing targets, motivating the development of label-free methods with improved sensitivity. Silicacoated metal nanoparticles allow these systems to be combined with supported lipid bilayers (SLBs) for sensing membrane proteins through localized surface plasmon resonance (LSPR). However, the small sensing volume of LSPR makes the thickness of the silica layer critical for performance. Here, we develop a simple, inexpensive and rapid sol-gel method for preparing thin conformal, continuous silica films and demonstrate its applicability using gold nanodisk arrays with LSPRs in the near-IR range. Silica layers as thin as ~5 nm are observed using crosssectional scanning transmission electron microscopy (STEM). The loss in sensitivity due to the thin silica coating was found to be only 16%, and the biosensing capabilities of the substrates were assessed through the binding of cholera toxin B to GM1 lipids. This sensor platform should prove useful in the rapid, multiplexed detection and screening of membrane-associated biological targets.

Keywords: Solid-Supported Lipid Bilayers, plasmonic biosensing, nanodisk array, silica sol-gel, localized surface plasmon resonance


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


Introduction Metal nanoparticles that host localized surface plasmon resonances (LSPRs) have proven useful in applications ranging from catalysis,

1,2

to biosensing,3,4 and

nanomedicine.5 Lipid membranes and membrane proteins are essential components of cellular metabolism and are key targets for more than half of pharmaceutical therapies.6,7 Interfacing surface-bound solid supported lipid bilayers (SLBs) with noble metal nanoparticle arrays is an attractive platform for refractive index-based biosensing as it offers label-free, facile optical detection. The small size of the metal nanoparticle transducers also allows their incorporation into microfluidic or multiplexed

systems.

However,

SLB/nanoparticle

constructs

demand

a

biocompatible interface. Silica is ideally suited for this application: it is a biocompatible material and retains the natural fluidity of the phospholipid bilayer, which enables multivalent interactions between biological species. However, whenever a coating is applied, it pushes the sensing species further from the nanoparticle surface, often reducing sensitivity. Minimizing the thickness of this coating is thus of paramount importance to maintain high sensitivity, while maintaining the mobility of the lipids and proteins within the bilayer. Several procedures have been developed to coat metal films and nanoparticles with silica. Physical8 and chemical vapor deposition9 as well as atomic layer deposition10 (ALD) are common methods of coating thin metal films and surface-bound nanoparticles with nanometers to microns of silica. For instance, nanohole array substrates encapsulated in silica through sputtering11 have been used in biosensing of membrane proteins. ALD-coated nanohole substrates have 3 ACS Paragon Plus Environment


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

Page 4 of 25

also been developed,12-15 yielding topographically flat substrates.16 However, the coating effect on sensitivity remains a difficult challenge, as exemplified by the large (almost 50%) decrease in sensitivity of gold nanodisks containing 10 nm of sputtered silica, most likely due to the particularly short field decay length of the nanoparticles employed in this study.17 While sputtering and ALD offer excellent control over film height, the instrumentation and processing can be expensive, and throughput and scalability are limited. Solution-phase silica coating is an inexpensive, scalable alternative. Sol-gel methods such as dip- or spin-coating can produce films with thicknesses ranging from microns18 to less than 10 nm,19 20 and have been applied to a variety of metallic systems, including thin metal films21,22 and nanoparticles.

23-29

Coated planar gold

surfaces have subsequently been combined with SLB’s for surface plasmon resonance sensing,21,22,30 but this technology has yet to be applied to SLB sensing using LSPR of metal nanoparticles. Here, we show that a silica sol can be spin-coated over a gold nanodisk array, achieving thicknesses of less than 10 nm, and that this thickness can be controlled by sol concentration. Since this technique should be capable of accommodating more complex shaped nanoparticles, we investigate NIR resonant AuNDs for their improved sensing volume, longer field decay lengths and increased sensitivity.31,32 A negligible decrease of sensitivity occurs upon coating these AuNDs with silica yielding bulk refractive index sensitivities of 420 nm/RIU. Lipid adsorption and binding of the model system cholera toxin B (CTB) to GM1 lipids is monitored with the sol-gel film and compared with AuNDs coated by silica sputtering. The high 4 ACS Paragon Plus Environment

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


sensitivity of these sol-coated platforms is shown to be due to both the highly sensitive NIR-resonant AuNDs and the ultrathin silica layer, which minimizes the displacement of the measured species from the nanoparticle surface.

Methods Materials.

All lipids were purchased from Avanti Polar lipids. Carboxylated

polystyrene beads, 300 nm in diameter, were purchased from Life Technologies. Deuterium oxide (D2O) was purchased from Cambridge isotope laboratories. All other chemicals, unless indicated otherwise, were purchased from Sigma-Aldrich and were used as received. Gold Nanodisk Substrate Preparation. All nanodisk substrates were prepared with hole-mask colloidal lithography as described elsewhere.33 Briefly, glass coverslips (#2) were cleaned in piranha (3:1 concentrated H2SO4 to 30% H2O2) for 30 minutes before rinsing heavily with doubly distilled water and drying with N2. A 4% (w/v) solution of 350 kD poly(methyl methacrylate) (PMMA) in anisole was then spin-coated at 300 rpm for 30 sec and the thin film was subsequently hard baked at 180 °C for 5 min. The PMMA film was etched for 5 sec with O2 in a reactive ion etcher (March CS-1701) to render the surface hydrophilic. A 0.4% (v/v) aqueous poly(diallyldimethylammonium chloride) (PDDA) (MW = 350,000) solution was then drop-coated onto the glass slides, rinsed with doubly distilled H2O after 30 secs and dried with N2. Carboxylated poly(styrene) beads (0.08% (w/v)) with an average diameter of 300 nm were then drop-coated over the PDDA layer, rinsed with doubly distilled H2O after 2 minutes, and dried with N2. Electron beam deposition (Airco 5 ACS Paragon Plus Environment


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

Page 6 of 25

Temescal FC1800) was used to deposit 5 nm of gold over the poly(styrene) beads. Samples were tape-stripped to remove the poly(styrene) beads and expose the polymeric layer underneath. A 2-minute O2 reactive ion etching treatment was then used to etch wells into the polymeric layer and 20 nm of gold was deposited over each substrate. The mask was lifted-off by sonication in acetone, the samples were rinsed with doubly distilled H2O and then finally dried with N2 yielding nanodisks with dimensions of 20 nm in height and approximately 300 nm in diameter as determined by atomic force microscopy. Sol-Gel Coating. To prepare the silica sol 12 mg of concentrated HCl were mixed with 1.5 g doubly distilled H2O. This solution was added drop-wise with stirring to a mixture of 5 g tetraethyl orthosilicate (TEOS) and 11.05 g pure (≥99.5%) ethanol. This solution was allowed to stir for 15 minutes before refluxing at about 85 °C for 1 hour. This suspension was then allowed to cool and age for two days at room temperature to yield an 8% silica sol. This solution was diluted with pure ethanol to 0.25, 0.5, or 1% (v/v) and stored at 4 °C. Suspensions were then sonicated before use and spin-coated directly onto gold nanodisks on glass coverslips or on a silicon wafer at 2000 RPM for 60 secs. Samples were finally baked in a kiln (Paragon, Inc.) at 300 °C for 1 hour before use to calcinate the film. Small Unilamellar Vesicle Preparation. Lipids were prepared to a concentration of 2 mg/mL by mixing 1 mol% GM1 with egg PC and evaporating the solvent under gentle N2 flow. Lipid cakes were then placed under vacuum for at least 3 hours to remove residual solvent. Lipid cakes were then rehydrated in D2O buffer containing 10 mM TRIS-HCl (pH of 7.8) and 100 mM NaCl. Small unilamellar vesicles were 6 ACS Paragon Plus Environment

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


prepared by extruding with 21 passes through a 200 nm and 100 nm porous polycarbonate membrane using a mini-extruder (Avanti Polar Lipids). Lipids were stored under N2 at 4 °C and used within 1 week. LSPR Measurements. All NIR absorbance measurements were collected with a tungsten halogen lamp and a NIRQuest512 (InGaAs linear array CCD) spectrometer from Ocean Optics. Sensitivities were determined by measuring a sample in solutions of 0, 10, 20, 30, and 40% (w/w) glycerol in D2O. D2O is required as prominent vibrational modes of water above about 1400 nm preclude absorbance measurements beyond this region. Biosensing. GM1-CTB binding measurements were performed in a simple, homebuilt ~800 µl flow cell that was prepared by bonding glass slides together with poly(dimethylsiloxane) (Krayden, Inc.) through oxygen plasma etching. Lipids were diluted to a final concentration of 600 μg/mL in buffer containing 10 mM CaCl2. After incubating for at least 30 min, excess small unilamellar vesicles were washed away in buffer without CaCl2. In a typical experiment 200 nM CTB were flown in and allowed to bind for at least 30 min before washing with buffer to eliminate contributions to LSPR shift from the changes in bulk refractive index. Decay Length Measurements. To determine the decay length of the bare gold nanodisks and those coated with the sol-gel silica, the polyelectrolyte layer-by-layer technique described by the Rubinstein group was used.34 Glass coverslips or Si wafers were etched with O2 reactive ion etching for 6 minutes to clean the surface and render it hydrophilic. Poly(allylamine hydrochloride) (PAH) and poly(styrene4-sulfonate) (PSS) were used. The initial layer for both glass coverslips and Si 7 ACS Paragon Plus Environment


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

Page 8 of 25

wafers was PAH. Alternating layers of polyelectrolytes were then added by immersing the substrate for 12 minutes in a solution containing PAH or PSS (1 mg/mL) and 100 mM NaCl. Substrates were rinsed with doubly distilled H2O and immersed into a solution containing 100 mM NaCl before moving into the next polymer solution. Optical measurements were performed on substrates only after the PSS step. After PSS addition the substrate was rinsed with doubly distilled water and dried gently under a stream of N2 before measurement. Bilayer thicknesses were determined by measuring film heights with ellipsometry (Vase ellipsometer, J.A. Woollam Co.) after the addition of PSS. Atomic Force Microscopy. All atomic force microscopy measurements were performed on a Veeco Dimension 3100 (Bruker, Inc.) in tapping mode using a SiN tips (Micromasch, Inc.). Scan rates of 1 kHz and amplitude set points of about 1V were typical measurement conditions. Electron Microscopy. The shape, size and morphology of coated gold nanodisks were imaged by scanning electron microscopy (SEM) in an FEI Helios NanoLab 660 dual beam microscope equipped with a solid state concentric backscatter (CBS) detector and micromanipulator. The CBS detector was used to record low-kV backscattered electron images. By using the landing energy of 1.5 keV and stage biasing voltage of 2 kV, charging artifacts from the glass coverslip and silica layer were avoided in the backscattered electron images. For measuring the thickness of the silica layer, cross sectional TEM specimens from the coated gold nanodisks were prepared using focused ion beam (FIB) milling. The technique of in-situ lift-out35 was used in the FEI Helios microscope for preparing TEM specimens. A thick layer 8 ACS Paragon Plus Environment

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


of Pt (over 500 nm) was deposited on the region of interest using the electron beam induced deposition method to protect the silica layer from FIB-induced damage. Throughout the deposition and milling steps, a tungsten needle was brought in contact with the silica surface of the sample in the vicinity of the region of interest using the micromanipulator, to dissipate the charges induced by the electron and ion beams. At the final step of specimen preparation, the specimen surfaces were polished by a 2 kV Ga ion beam to minimize the damage caused by the 30 kV FIB. High angle annular dark-field scanning transmission electron microscope (HAADF STEM) images were acquired in an FEI Titan3 Themis S/TEM operated at 80 kV. Fluorescence Recovery After Photobleaching (FRAP) Measurements. FRAP small unilamellar vesicles were prepared as described above by the solvent evaporation and extrusion method. 2 mol% NBD-PE with egg PC in TRIS-HCl buffer (pH of approximately 7.8) was used and lipids were extruded through 200 nm and 100 nm polycarbonate membranes for 21 total passes each. A final concentration of 300 μg/mL lipids and 10 mM CaCl2 were used to prepare the SLB. Samples were then washed with doubly distilled H2O to remove excess fluorophores in the bulk and measured on a Zeiss confocal laser-scanning microscope with a 10x objective and a 488 nm laser line. A 20x20 µm square region was bleached with full power and allowed to recover for 2.5 min. Intensities in the bleached region were averaged, normalized to the background intensity and fit to a single exponential function in OriginPro to extract mobile fractions. Diffusion coefficients were determined for a square bleached region36 with: D = (L/2)2/(4t1/2), where D is the



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

Page 10 of 25

diffusion coefficient, L is the length of the box in μm, and t1/2 is the time for halfrecovery as determined from the fitting procedure.

Results and Discussion To prepare thin silica films over surface-bound metal nanostructures, we combine hole-mask colloidal lithography (HCL) with sol-gel spin-coating. In HCL, the size, spacing, and height of nanodisks, and hence LSPR energy, can be tuned by altering the amount of deposited metal as well as polystyrene bead size and concentration. Here, gold nanodisks of 300 nm in diameter and 20 nm in height, as measured using AFM (see Figure S4), were employed. These gold nanodisks were chosen because their near infrared resonance have larger refractive index sensitivities and longer field decay lengths than nanoparticles resonant in the visible range.32 The latter is particularly beneficial for sensing species binding to SLBs, since the displacement from the sensor surface is larger than in most biosensing applications, where molecules are typically tethered directly to the nanoparticle surface.37 The refractive index sensitivity of bare and silica coated gold nanodisks was determined using glycerol/D2O mixtures, see Table S1. The LSPR peak wavelength vs. refractive index displayed a linear relationship for all samples, and the sensitivity values reported are the slope of these lines. The data shows a clear trend in that reduced refractive index sensitivities were observed for the higher sol concentrations, see Figures S1 and Table S1. Further investigation of this trend was carried out by spin-coating the sols of different concentration onto silicon wafer 10 ACS Paragon Plus Environment

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


substrates followed by ellipsometry measurements, see Table S2. Indeed, higher sol concentrations led to increasingly thicker films on the surface. In addition, density values for the silica films on the gold nanodisk substrates were calculated from the observed LSPR shifts and applying the Lorentz-Lorenz relationship, see Table S2. Interestingly, the density values of films generated from higher sol concentrations were not significantly different. This implies the change in sensitivity with increasing sol concentration is due mainly to increasing thickness.

Figure 1. Representative LSPR spectra of 300x20 nm gold nanodisks coated with 0.25% silica sol (A) and without silica coating (C) in varying glycerol-D2O solutions. Representative sensitivity plots of 300 nm gold nanodisks coated with 0.25% silica sol (B) and without silica (D). Average sensitivities for the 0.25% sol-coated gold nanodisks and bare gold nanodisks are 420 and 500 nm/RIU, respectively, see Table S1. The insets of A and C are zoomed in regions around the peak maxima. 11 ACS Paragon Plus Environment


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

Page 12 of 25

Sensitivities were determined by varying solution refractive index with 0, 10, 20, 30, and 40% glycerol (w/w) in D2O. Spectra in panel A and C are normalized.

By adjusting the original sol to a concentration of 0.25% (v/v) with ethanol, the sensitivity approaches that of the bare gold nanodisks. Representative spectra of the 0.25% silica sol-coated gold nanodisks and a linear plot of LSPR shift vs. refractive index can be found in Figure 1. Note that the silica coated particles have LSPRs significantly blue-shifted from bare particles (Table S1), a result of thermal annealing during the drying process, where solvent is removed by baking 1 hour at 300 °C. A decrease in the refractive index sensitivity of ~16% was found for the 0.25% silica sol-coated gold nanodisks compared to the bare gold nanodisks. A similar decrease in sensitivity, ~14%, was observed for the uncoated gold nanodisks alone after baking at 300 °C (Figure S3). These results suggest the majority of the sensitivity loss for the 0.25% sol-coated sample can be attributed to the annealing process, with minimal sensitivity loss due to the thin silica film.


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


Figure 2. SEM images of gold nanodisks coated with 0.25% silica sol (A) and 10 nm sputtered silica control (C). Cross-sectional HAADF STEM images of gold nanodisks coated with 0.25% silica sol (B) and 10 nm sputtered silica control (D).

The highest sensitivity samples, the gold nanodisks coated with 0.25% silica sol, were further characterized and compared with gold nanodisks coated with 10 nm of sputtered silica, as shown in Figure 2. The particle dimensions measured with SEM matched those measured by AFM (Figure S4). The AFM images of silica solcoated gold nanodisks also demonstrate that a smooth continuous film of silica is formed. A clear difference in thickness of the silica over the nanodisks is observed in the FIB-prepared cross-sections shown in Figure 2. Indeed, silica thicknesses above the 0.25% silica sol-coated and sputtered gold nanodisks are in the ranges of 4.2 to 7.0 nm and 7.3 to 9.4 nm, respectively. The thicknesses between the nanodisks were found to be 6 nm and 13 nm for the sol-gel and sputtered film (Figure S5), respectively. This topography and thicknesses result from the interplay of



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

Page 14 of 25

evaporative film thinning and capillary leveling forces resulting during spincoating.38

Figure 3. FRAP measurements carried out on SLBs formed over 0.25% sol-coated gold nanodisks. Fluorescence images taken after bleaching (A) and after recovery (B) of SLB containing 2 mol% NBD-PE. Normalized fluorescence recovery over time and single exponential fit are shown in C In order to further demonstrate the successful coating of silica over the 0.25% sol-coated gold nanodisks, a SLB was formed on these samples and characterized using fluorescence recovery after photobleaching (FRAP). The FRAP results confirm the presence of fluorescently labeled lipids on the coated arrays, and, importantly, indicate that the lipid bilayer formed is composed of mobile lipids which are able to diffuse across the bilayer. Fluorescence recoveries and fits to the raw data for SLBs formed on the 0.25% sol-coated gold nanodisks and on a clean glass coverslip, used as a positive control, are shown in Figure 3, Figure S6, and Table S3. Diffusion coefficients for SLBs on these substrates are also shown in Table S3. The similar recoveries for lipids on sol-coated gold nanodisks and the bare glass coverslip, 91.7 +/- 0.7% and 89 +/- 2%, respectively, strongly suggests the sol


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


coated nanodisks are completely embedded in a smooth silica layer, since a mobile bilayer is formed. Indeed, lipids formed directly on gold show significantly decreased fluorescence recoveries.

39,40

Additionally, the diffusion coefficients of

SLBs on these substrates are similar and agree with literature values.11,16 This further supports that the silica layer formed by the sol is continuous, as changes in surface area or morphology impact lipid diffusion coefficients.30 Being thin, conformal, and biocompatible, the 0.25% silica sol-coated gold nanodisk arrays make optimal platforms for the sensing of membrane-associated species. To assess the biosensing capabilities of the sol-gel coating strategy, small unilamellar vesicles composed of 1 mol% GM1 and egg PC were used to prepare SLBs and measure CTB binding. Kinetic traces of lipid adsorption and CTB binding to 0.25% silica sol-coated gold nanodisks as well as representative spectra are reported in Figure 4. Liposome adsorption and bilayer formation occurs within a few minutes, leading to a LSPR redshift that saturates at ~10 nm. Rinsing with buffer removes excess vesicles and leads to a small blueshift; the total shift due to lipids is 8.4 ± 3.0 for the 0.25% sol-coated gold nanodisks (Table 1). The addition of CTB (200 nM followed by rinsing) leads to a redshift of 2.2 ± 0.8 nm. Nonspecific adsorption to 0.25% sol-coated gold nanodisks was assessed by mixing CTB with SLBs prepared with only egg PC and no GM1 over 0.25% sol-coated gold nanodisks (Figure S7). In these experiments, no significant red-shift above instrumental noise was observed after washing away excess protein from solution. This suggests that nonspecific adsorption does not significantly contribute to the shift observed. LSPR shifts as high as 1.2 nm have been reported for CTB binding to GM1 over Ag 15 ACS Paragon Plus Environment


nanohole arrays coated with 15 nm of silica.12 At 2.2 nm, the LSPR shifts found here are the largest reported for the CTB-GM1 system, a result of the high sensitivity of the NIR-resonant particles and the optimized coating procedure developed herein.

A)

B)

Buffer wash

200 nM CTB

0.20

12

Absorbance (a.u.)

14

LSPR Shift (nm)

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

Page 16 of 25

Buffer

10 8

Final wash

6 4 2

Lipids

0 0

10

20

30

40

50

0.15

0.10 Initial Buffer After Lipid Addition And Washing After CTB Addition And Washing 0.05 1100

60

Time (min)

1200

1300

1400

Wavelength (nm)

Figure 4. Kinetic measurement of lipid adsorption and 200 nm CTB binding to 1 mol% GM1 lipids in egg-PC SLBs formed over 0.25% sol-coated 300 nm gold nanodisks (A). Representative LSPR spectra (B) were collected in initial buffer, after washing excess lipids, and finally after washing excess CTB from solution. Spectra are not normalized to show increasing intensity as lipids and CTB are adsorbed to the surface. Table 1. Lipid Adsorption and CTB Binding For Different Silica Supports. Sample

LSPR

LSPR shift

shift due

due

to lipids

protein

(nm)

(nm)

to


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


Nanodisks

-

8.4 ± 3.0

2.2 ± 0.8

0.25% Sol

Nanodisks

-

7.7 ± 1.3

1.3 ± 0.4

Sputtered SiO2 *Values reported are average ± standard deviation for at least 4 samples. a – values were determined by using equation 1.

A comparison of the biosensing capabilities of the 0.25% sol-coated samples was compared to silica sputtered samples and the results are reported in Table 1. Although the observed LSPR shifts of the 0.25% sol-coated sample is higher than that of the sputtered sample, due to the increased thickness of the 10 nm sputtered film, the difference is rather modest, see a statistical analysis in Table S4. A possible explanation for this small difference is that the decay lengths are considerably longer than the thicknesses in question. In order to measure the decay lengths of the NIR-resonant gold nanodisks, a previously described34 layer-by-layer technique was employed. Briefly, alternating polyelectrolyte layers were applied to the nanoparticle surface and the LSPR shift was determined as a function of film thickness (Figure 5). The thickness and refractive index of the polymer film after addition of polyelectrolyte bilayers were measured using ellipsometry (Figure S8). The electromagnetic field decay length, l, was then calculated according to:34 ೏

∆ߣ = ݉∆n(1 − ݁ ି ೗ )

(1)



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

Page 18 of 25

where m is the bulk refractive index sensitivity, ∆݊ is the difference of the adsorbent layer and the ambient medium (air as a reference) refractive indices and d is the thickness of the adsorbent. Note that while Equation 1 was originally developed with a factor of 2 in the exponential, to account for a |E|2 dependence of the electromagnetic field,41,42 recent reports have omitted this factor and have shown good empirical agreement with experimental data.34,43 The factor of 2 was omitted here for ease of comparison with the current literature. The decay lengths for the samples shown in Figure 5 were found to be 51 ± 7 nm for the bare gold nanodisks and 56 ± 8 nm for the sol-coated gold nanodisks, respectively. The experimentally obtained decay lengths agree well with literature values for nanoholes43 (60 nm decay length) and are larger than what has been reported for smaller gold nanodisk structures: 150 nm diameter gold nanodisks43 (29 nm decay length) and 113 nm gold nanoislands34 (22 nm decay length). Since the field decay lengths are relatively long for the NIR-resonance gold nanodisks employed herein, the height of both assemblies lies in a similar region of the decay curve, yielding comparable sensitivities and a small difference in sensor response. However, it is important to note that many nanoparticles currently used for sensing lipid bilayers possess visible resonances hence shorter field decay lengths, often 10-20 nm;44,45 in these cases the dependence on distance is expected to be more significant, particularly if the height of the assembly is similar to the field decay length.


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 5. LSPR shifts obtained with increasing the thickness of polyelectrolyte bilayers on 0.25% silica sol-coated gold nanodisks and bare gold nanodisks. Points without error bars had standard deviations smaller than the symbol.

Figure 6. Binding of cholera toxin B-subunit (CTB) to solid supported bilayers containing 1 mol % GM1 on gold nanodisks with 0.25% silica sol. The data was fit to 19 ACS Paragon Plus Environment


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

Page 20 of 25

a Langmuir isotherm, black line, and the KD was determined to be 0.625 µM The maximum shift for CTB binding obtained with these substrates was 5.9. The shaded pink region towards the bottom of the graph represents shifts that are below the detection limit,.

To assess the limit of detection, a binding curve for the CTB-1 mol% GM1 system was generated using the 0.25% sol-coated gold nanodisks (Figure 6). The KD for the pentameric CTB was determined from fitting the data to a Langmuir isotherm.46 The Langmuir isotherm may be used to approximate the interaction of the multivalent CTB-GM1 system, provided that each binding event is independent, with an effective dissociation constant KD.46 The KD was determined to be 0.625 µM with a LOD of 20 nM. The limit of detection value was determined as the concentration giving a shift larger than four times the instrument noise, which is measured in Figure S10. The KD values obtained fall in the range of those reported in the literature with SPR (1 μM) and fluorescence techniques (370 nM).47,48 Since the limit of detection concentration is roughly 14 times lower than the KD concentration, where exactly half of all sites are occupied, this data implies that these sensors are able to detect approximately 7% surface coverage of CTB. Lastly, the sol-gel coated gold nanodisks samples appeared to be quite stable over time. Samples coated with SLBs were left out for one month, cleaned with detergent and oxygen plasma etching to remove lipids, and then used once again in biosensing measurements. Lipid absorption and CTB binding measurements (Figure S11) show minimal to no degradation of the sample, indicating that the sensors are robust and reusable.


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


Conclusions We report a simple, versatile, inexpensive sol-gel spin-coating method to coat surface-bound structures. This method enables uniform, thin film formation with thicknesses of less than 10 nm, which is considerably less expensive and more scalable than sputtering and ALD. In addition, this platform should be applicable to structures of arbitrary and asymmetric shape to improve sensitivity further. The silica coating was found to produce a fluid SLB and yield a negligible loss of sensitivity resulting in significantly larger shifts for lipid and protein species than current platforms. This high sensitivity is due to both NIR resonant gold nanodisks, which are shown to be highly sensitive structures with extended field decay lengths, in addition to the ultrathin silica coating. These larger field decay lengths and increased sensitivity should prove useful for sensing of larger species such as bacteria and viruses. The system herein combines properties typical to surface plasmon resonance (extended field decay lengths and higher sensitivities) with the ease of measurement and propensity for miniaturization that is required for many microfluidic and multiplexing applications. Thus, these substrates should find tremendous use in the screening of membrane-associated species in a multiplexed manner. Lastly, this method is also not restricted to silica and should easily be expanded to materials of different compositions (e.g. NiO or TiO) to facilitate new interfaces for biosensing or materials development.

Supporting Information



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

Page 22 of 25

Additional characterization data on the sol-coated samples, along with results from stability studies and control experiments are available free of charge via the Internet at http://pubs.acs.org.

Acknowledgements This work was supported by University of Cincinnati and Rice University start-up funds.

References (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14)

Linic, S.; Aslam, U.; Boerigter, C.; Morabito, M. Nat. Mater. 2015, 14, 567–576. Sambur, J. B.; Chen, P. Annu. Rev. Phys. Chem. 2014, 65 (1), 395–422. Mayer, K. M.; Hafner, J. H. Chem. Rev. 2011, 111 (6), 3828–3857. Unser, S.; Bruzas, I.; He, J.; Sagle, L. Sensors 2015, 15 (7), 15684–15716. Dreaden, E. C.; Alkilany, A. M.; Huang, X.; Murphy, C. J.; El-Sayed, M. A. Chem. Soc. Rev. 2012, 41, 2740–2779. Cooper, M. A. J. Mol. Recognit. 2004, 17 (4), 286–315. Tiefenauer, L.; Demarche, S. Materials 2012, 5 (11), 2205–2242. Szunerits, S.; Boukherroub, R. Langmuir 2006, 22 (4), 1660–1663. Szunerits, S.; Das, M. R.; Boukherroub, R. J. Phys. Chem. C 2008, 112 (22), 8239–8243. Im, H.; Wittenberg, N. J.; Lindquist, N. C.; Oh, S.-H. J. Mater. Res. 2012, 27 (04), 663–671. Jonsson, M. P.; Jönsson, P.; Dahlin, A. B.; Höök, F. Nano Lett. 2007, 7 (11), 3462–3468. Lee, S. H.; Lindquist, N. C.; Wittenberg, N. J.; Jordan, L. R.; Oh, S.-H. Lab Chip 2012, 12, 3882–3890. Junesch, J.; Emilsson, G.; Xiong, K.; Kumar, S.; Sannomiya, T.; Pace, H.; Vörös, J.; Oh, S.-H.; Bally, M.; Dahlin, A. B. Nanoscale 2015, 7 (37), 15080–15085. Im, H.; Lee, S. H.; Wittenberg, N. J.; Johnson, T. W.; Lindquist, N. C.; Nagpal, P.; Norris, D. J.; Oh, S.-H. ACS Nano 2011, 5 (8), 6244–6253. 22 ACS Paragon Plus Environment

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


(15) (16) (17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27)

(28)

(29) (30)

(31) (32) (33) (34) (35) (36) (37) (38)

Im, H.; Wittenberg, N. J.; Lesuffleur, A.; Lindquist, N. C.; Oh, S.-H. Chem. Sci. 2010, 1 (6), 688–696. Jose, J.; Jordan, L. R.; Johnson, T. W.; Lee, S. H.; Wittenberg, N. J.; Oh, S.-H. Adv. Funct. Mater. 2013, 23 (22), 2812–2820. Zan, G. H.; Jackman, J. A.; Kim, S.-O.; Cho, N.-J. Small 2014, 10 (23), 4828– 4832. McDonagh, C.; Sheridan, F.; Butler, T.; MacCraith, B. D. J. Non-Cryst. Solids 1996, 194, 72–77. Faustini, M.; Louis, B.; Albouy, P. A.; Kuemmel, M.; Grosso, D. J. Phys. Chem. C 2010, 114 (17), 7637–7645. Kambhampati, D. K.; Jakob, T.; Robertson, J. W.; Cai, M.; Pemberton, J. E.; Knoll, W. Langmuir 2001, 17 (4), 1169–1175. Linman, M. J.; Culver, S. P.; Cheng, Q. Langmuir 2009, 25 (5), 3075–3082. Phillips, K. S.; Han, J. H.; Martinez, M.; Wang, Z. Z.; Carter, D.; Cheng, Q. Anal. Chem. 2006, 78 (2), 596–603. Liz-Marzán, L. M.; Giersig, M.; Mulvaney, P. Langmuir 1996, 12 (18), 4329– 4335. Lin, L.; Zapata, M.; Xiong, M.; Liu, Z.; Wang, S.; Xu, H.; Borisov, A. G.; Gu, H.; Nordlander, P.; Aizpurua, J.; Ye, J. Nano Lett. 2015, 15 (10), 6419–6428. Liu, S. H.; Han, M. Y. Adv. Funct. Mater. 2005, 15, 961–967. Sendroiu, I. E.; Warner, M. E.; Corn, R. M. Langmuir 2009, 25 (19), 11282– 11284. Floris, F.; Figus, C.; Fornasari, L.; Patrini, M.; Pellacani, P.; Marchesini, G.; Valsesia, A.; Artizzu, F.; Marongiu, D.; Saba, M.; Mura, A.; Bongiovanni, G.; Marabelli, F.; Quochi, F. J. Phys. Chem. Lett. 2014, 5 (17), 2935–2940. Irit Ruach-Nir; Tatyana A Bendikov; Ilanit Doron-Mor; Zahava Barkay; Alexander Vaskevich, A.; Israel Rubinstein. J. Am. Chem. Soc. 2007, 129 (1), 84–92. Wu, H.-J.; Henzie, J.; Lin, W.-C.; Rhodes, C.; Li, Z.; Sartorel, E.; Thorner, J.; Yang, P.; Groves, J. T. Nat. Methods 2012, 9 (12), 1189–1191. Goksu, E. I.; Hoopes, M. I.; Nellis, B. A.; Xing, C.; Faller, R.; Frank, C. W.; Risbud, S. H.; Satcher, J. H., Jr.; Longo, M. L. Biochim. Biophys. Acta 2010, 1798 (4), 719–729. Miller, M. M.; Lazarides, A. A. J. Phys. Chem. B 2005, 109 (46), 21556–21565. Larsson, E. M.; Alegret, J.; Käll, M.; Sutherland, D. S. Nano Lett. 2007, 7 (5), 1256–1263. Fredriksson, H.; Alaverdyan, Y.; Dmitriev, A.; Langhammer, C.; Sutherland, D. S.; Zäch, M.; Kasemo, B. Adv. Mater. 2007, 19 (23), 4297–4302. Kedem, O.; Tesler, A. B.; Vaskevich, A.; Rubinstein, I. ACS Nano 2011, 5 (2), 748–760. Langford, R. M.; Clinton, C. Micron 2004, 35 (7), 607–611. Deschout, H.; Hagman, J.; Fransson, S.; Jonasson, J.; Rudemo, M.; Loren, N.; Braekmans, K. Opt. Express 2010, 18 (22), 22886–22905. Anker, J. N.; Hall, W. P.; Lyandres, O.; Shah, N. C.; Zhao, J.; Van Duyne, R. P. Nat. Mater. 2008, 7 (6), 442–453. Rawlings, C.; Wolf, H.; Hedrick, J. L.; Coady, D. J.; Duerig, U.; Knoll, A. W. ACS 23 ACS Paragon Plus Environment


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

(39) (40) (41) (42) (43) (44) (45) (46) (47) (48)

Page 24 of 25

Nano 2015, 9 (6), 6188–6195. Sofou, S.; Thomas, J. L. Biosens. Bioelectron. 2003, 18 (4), 445–455. Peng, P.-Y.; Chiang, P.-C.; Chao, L. Langmuir 2015, 31 (13), 3904–3911. Jung, L. S.; Campbell, C. T.; Chinowsky, T. M.; Mar, M. N. Langmuir 1998, 14 (19), 5636–5648. Haes, A. J.; Van Duyne, R. P. J. Am. Chem. Soc. 2002, 124 (35), 10596–10604. Mazzotta, F.; Johnson, T. W.; Dahlin, A. B.; Shaver, J.; Oh, S.-H.; Höök, F. ACS Photonics 2015, 2 (2), 256–262. Haes, A. J.; Zou, S.; Schatz, G. C.; Van Duyne, R. P. J. Phys. Chem. B 2004, 108 (1), 109–116. Whitney, A. V.; Elam, J. W.; Zou, S.; Zinovev, A. V.; Stair, P. C.; Schatz, G. C.; Van Duyne, R. P. J. Phys. Chem. B 2005, 109 (43), 20522–20528. Winter, E. M.; Groves, J. T. Anal. Chem. 2006, 78 (1), 174–180. Williams, T. L.; Jenkins, A. T. A. J. Am. Chem. Soc. 2008, 130 (20), 6438–6443. Moran-Mirabal, J. M.; Edel, J. B.; Meyer, G. D.; Throckmorton, D.; Singh, A. K.; Craighead, H. G. Biophys. J. 2005, 89 (1), 296–305.


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


TOC Figure:


Ultrasensitive Plasmonic Platform for Label-Free Detection of

Recommend Documents